AESCHYLUS

AGAMEMNON
 AESCHYLUS

AGAMEMNON
EDITED

WITH A COMMENTARY BY

EDUARD FRAENKEL

VOLUMEI
PROLEGOMENA, TEXT
TRANSLATION

OXFORD
AT THE CLARENDON PRESS
 This book has been printed digitally and produced in a standard specification
in order to ensure its continuing availability

OXFORD
UNIVERSITY PRESS

Great Clarendon Street, Oxford OX2 6DP

Oxford University Press is a department of the University of Oxford.
It furthers the UniversityG objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
5.0 Paulo Shanghai Taipei Tokyo Toronto

Oxford is a registered trade mark of Oxford University Press

in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York

© Oxford University Press 1950

The moral rights of the author have been asserted

Database right Oxford University Press (maker)
Reprinted 2003

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose this same condition on any acquirer
ISBN 0-19-927170-4 (Vol.1)
ISBN 0-19- 814101-7 ( Set )
 ΤΟ
J. Ὁ. BEAZLEY
AND
RUTH FRAENKEL
 PREFACE
THE man who is unwise enough to add still another commentary on
the Agamemnon to the many already in existence must expect first
of all to be asked what type of reader he had in mind when planning
his book. Faced with this embarrassing question I should like to
borrow my answer from Lucilius, who, we are told, dicere solebat
neque se ab indoctissimis neque a doctissimis legt velle, quod alteri
nihil intellegerent, altert plus fortasse quam ipse. In other words, I
have written neither for those who have only a rudimentary know-
ledge of Greek nor for the specialists. My favourite reader, whose
kindly and patient face would sometimes comfort me during the
endless hours of drudgery, looked surprisingly like some of the
students who worked with me for many years at Oxford in our happy
seminar classes on the Agamemnon. Without the inspiring, and often
correcting, co-operation of those young men and women I should not
have been able to complete the commentary. If they thought a
passage to be particularly difficult, that was sufficient reason for me
to examine and discuss it as fully as I could; and more than once it
was their careful preparation, their inquisitiveness, and their persis-
tent efforts that made it possible to reach what seemed to us a satis-
factory solution. Thus, ἀεὶ yap ἡβᾶι τοῖς γέρουσιν εὐμαθεῖν, I was
fortunate enough late in life to profit from the high standard of work
in the classics at many schools in England and Scotland. If this
standard can be maintained in years to come, my commentary will
for a generation or two find the readers for whom it is primarily
meant, and then, I trust, be replaced by a better one.
When, about twenty-five years ago, I first came to grips with some
of the major problems of the Agamemnon, I thought that I could deal
with them in a series of articles. But soon it became clear that such
treatment would involve the danger of not being wholly honest,
since there would be a strong temptation to dwell on certain fascinat-
ing aspects of the play and to neglect many thorny issues which,
though perhaps less exciting, yet have an equal claim to be properly
investigated. It therefore proved necessary to expound the play at
full length.
About one fundamental point I made up my mind at an early
stage (cf. C.R. li, 1937, 63): if the commentary was to serve the pur-
pose which I had in mind, it would have to follow, though of course
in a modernized form, the example of the old editiones cum notis
variorum. The arrangement resulting from that decision is bound to
be repulsive to many readers, but I do not want to apologize for it.
I will, however, try to explain briefly my main reasons for preferring
vi
 PREFACE
this clumsy scheme to the shorter and perhaps more attractive
presentation of my own views only.
One of my motives may or may not be called sentimental, but in
any case it is a strong one: I should hate the idea of being forced to
take over from others a great deal of the most valuable material and
the most decisive statements in my commentary without acknow-
ledging my debt in detail. I think it wiser to adopt the method of the
ancient commentator who, when confronted with the same problem,
decided thus: cum . . . liceret usquequaque nostra interponere, maluimus
optima fide quorum res fuerant eorum etiam verba servare (Aelius
Donatus in the letter dedicating his commentary on Virgil to L.
Munatius). But it is not for the sake of moral satisfaction alone that
substantial quotations, or paraphrases, from the commentaries of
earlier scholars seem to be desirable. On many controversial points
of the Agamemnon the arguments in favour of, or against, any parti-
cular view have been put forward with great lucidity and force by
eminent scholars. If I were to try to summarize their reasoning I
should merely blunt the edge of their arguments. In the discussion
of a text such as that of the Agamemnon the antagonism of two, and
sometimes more, contrasting views is not due, as the inexperienced
might think, to the obscura diligentia of classical scholars, but is the
true expression of the complexity of the object itself. For a moment
it may be possible by ingenious dialectical manceuvres to concentrate
the light on one possible interpretation of a passage and darken the
opposite one, but the apparently defeated side will in course of time
come to life again and take its revenge on those who have neglected
it. In these circumstances it is far better for the student to have
each aspect of a controversial question put to him, even to the point
of one-sided exaggeration, in the words of its most capable and
energetic champions rather than in the diluted summary of a late-
comer. I hope I have in most cases put the reader in a position to
reject my own interpretation, if he chooses to do so, with the help
of my excerpts from, or references to, other scholars’ comments.
It is for this reason above all that I have given so much space to the
reproduction of what has been said before. It is neither likely nor
desirable that in the near future another scholar should feel tempted
to plod once more through so many commentaries, translations,
books, and articles in order to bring together a representative selec-
tion of his forerunners’ views. In reading recent commentaries and
papers on the Agamemnon I have often found that the writer puts
forward as brand-new an opinion which has more than once been
recommended and then subsequently been refuted. This is pardon-
able enough, for a good deal of the relevant literature is not easily
accessible, especially after the last war. Even at Oxford there exist,
so far as I know, only two copies of Philologus, Suppl.-Bd. I (1860),
vil
 PREFACE
which contains Ahrens’s indispensable comments (see p. 55 f. below).
I have always borne in mind this state of affairs. In many cases I
have also tried to give some idea of the extent to which a particular
interpretation was followed in modern times ; sometimes I may have
overdone it and adduced more names than was absolutely necessary,
but if this is a mistake I do not think it will unduly delay the
reader.
As regards the discussion of controversial passages, I must make
one further observation. It is a widespread belief that in the case of
a so-called crux only one of two roads is open to the conscientious
scholar: either he feels capable of understanding the disputed passage
as it stands or he has to assume a corruption of the text. To me this
belief seems to be based on a fallacy. We have only to pause for a
moment and consider, first, the enormous gulf between our ways of
life and thought and those of ancient Greece, then the sadly fragmen-
tary nature of our whole tradition, and, finally, the solitary boldness
of Aeschylus, to realize that it would be a sign of megalomania if we
fancied it to be possible for us fully to understand the words of this
poet wherever we have them in their original form. More than once,
therefore, I have had to state that I regard the text of a certain line
as probably sound but am nevertheless unable to grasp its meaning.
This conviction must not, of course, serve as a pretext for slackening
in our exertions. Every possible effort should be made to understand
a difficult passage; but when a careful examination of the language
and the style has produced no indication of a corruption and yet the
sense remains obscure, then there may be a case, not for putting a
dagger against the passage, but for admitting the limits of our com-
prehension.
The commentary includes a number of discussions on points of
grammar, syntax, semasiology, word-order, and the like, and on
prosody and metre. These sections may at first sight look like mere
digressions, especially those which contain a larger number of
examples than would have been required to decide the point in hand.
It might be argued that I ought to have taken them out of the com-
mentary and published them separately. It was after careful con-
sideration that I decided to leave them where they are. Not only does
the starting-point (and in most cases a great deal more) of these
apparent digressions form an important part of my argument, but
it seems to me that the interest of the reader who cares for such
problems at all is best served by my way of presenting the material.
In this I have been following the model of Bentley, Porson, Elmsley,
Lachmann, Madvig, Wilamowitz (especially in his commentary on the
Herakles), Housman, and others. It is my experience, and probably
that of others as well, that it is far easier to remember 'X or Y has
discussed such and such a phenomenon in connexion with that vexed
ix
 PREFACE
passage in that play’ than to have present in one’s mind the precise
number of the volume of one of the far too many periodicals in
which X or Y may have dealt with the matter. If the discussion of
a linguistic or metrical problem is intimately interwoven into the
investigation of a passage which has often worried you, then you
will easily recall the two things together.
The commentary is not preceded by an introduction. A book of
this kind does not seem to me the proper place for a recapitulation
of what can be found elsewhere about the story of Agamemnon and
his house, or the life and work of Aeschylus, or any other of the many
important topics connected in one way or another with this play.
Everything that I regard as the commentator’s immediate concern,
such as the appreciation of a scene or a song as a whole, an attempt
to grasp the essential features of the characters, remarks on the
relation of the play to the epic versions of the story, illustrations of
the traits linking up this drama with contemporary events, and so
forth, has been incorporated into the commentary.
My account of the readings of the MSS is based in the case of the
Mediceus on the facsimile published in 1896 (see p. ı below), and in
the case of the other MSS on photostats in my possession. When in
1940 we were cut off from communications with Italy, I had photo-
stats of F only, which had kindly been sent me by Dottoressa Teresa
Lodi, Librarian of the Biblioteca Medicea Laurenziana. At that junc-
ture I should have been most seriously hampered in my work had not
the generosity of Professor Alexander Turyn and of the authorities
of Harvard University Library come to my assistance : they provided
me with photostats of the other MSS.
In the critical apparatus I have in many cases adopted ‘la méthode
de l’apparat négatif” (see the pamphlet Emploi des signes critiques,
etc., published for the ‘Union académique internationale” first by
7. Bidez and A. B. Drachmann, then, 2nd ed., Brussels and Paris
1938, by A. Delatte and A. Severyns, p. 40 f.); so, for instance, the
note on 12, νυκτίπλακτον Tr, means that the other MSS have the
reading of my text, vuxrimAayxrov; the note on 29, ἐπορθριάζειν MV,
means that the other MSS have the ἐπορθιάζειν of the text, etc. But
I have given up this method whenever I thought that a different
notation would be clearer or more convenient, e.g. on 23, φάος MV:
dal pap.: νῦν φῶς ΕἼΤ. I may therefore be guilty of inconsistency,
but I hope there is nowhere a real ambiguity.
I now regret that I have included in the apparatus certain minutiae
of spelling and accentuation, 1.6. things which have little, if anything,
to do with the ancient παράδοσις. 1 am, of course, in full agreement
with the principle of a condensed apparatus which nowadays is
generally accepted. But I thought that this was a good opportunity
of reminding young students of the fact that many forms which they
X
 PREFACE
rightly expect to find in modern editions of classical texts are not
based on direct evidence in every single case but have been restored
from our general knowledge of the typical changes taking place in
the course of the transmission. I think now that it might have been
better to mention these things in the Prolegomena. However, no
great harm has been done, for when the MS tradition is so slender
asin the case of the Agamemnon, the reader is in no danger of having
his attention diverted from the main points in the apparatus by the
occasional insertion of some less important items.
The ‘testimonia’ for the text of this play are so few that it seemed
hardly worth while to place them in a separate section and thus
encumber the pages.
In the text I have in the lyric parts followed Wilamowitz’s example
and adopted the ancient system of notation as modified by Triclinius
(for his metrical signs see p. 17 below). Whereas ancient manuscripts,
.as is shown by the papyri, put the παράγραφος, the κορωνίς, etc. on
the left-hand side of the text column, Triclinius puts them on the
right-hand side. In my text — signifies the end of a strophe, == the
end of an antistrophe, and —c (a modification, for typographical
convenience, of Triclinius’ already greatly simplified form of the
kopwvis) the end of an epode.
The translation opposite the Greek text is to be regarded as part
of the commentary and nothing more. It makes no literary claim
whatsoever. For one who only recently learned a modest amount of
English it would be a piece of stark impertinence to try his hand at
such a task if he undertook it for its own sake. I wish I could have
done without the translation. But since not even the fullest com-
mentary can be exhaustive, I consider it the commentator’s duty to
supplement his detailed exegesis by an attempt to render the whole
text in prose as accurately as possible. About the severe limitations
of any translation, let alone a translation from Aeschylus into a
modern language, there can be no disagreement among sensible
people. But in the present instance the thorny question of how far
a translation, standing by itself, can be expected to reproduce its
original does not arise at all. Our issue is a much simpler one. My
version fills a few, though by no means all, of the gaps in my com-
mentary, and that is its justification. It may also every now and
then serve as a pointer to some of my notes. For in certain cases
my translation deliberately over-emphasizes a meaning or a shade
of meaning that seems to me to be important, but not to have been
sufficiently noticed. Several printed English translations have done
a good deal to facilitate my task ; nevertheless, this part of my work
would have come to nothing if some of my friends had not been
at pains to polish and re-polish my drafts over and over again.
The stylistic unevenness of the commentary is a consequence of
xi
 : PREFACE
its history. When I began to work on it, and for many years after, I
had no choice but to write in German. Therefore the bulk of the
commentary had subsequently to be translated into English. Other
parts of the book, viz. the Prolegomena, the general surveys in the
commentary and some special notes, and the Appendixes, were
written directly in English. Although I have spent much time in
trying to remove the resultant inequalities, I have achieved but
little. I hope the reader will forgive the chequered form, or rather
the formlessness, of the whole thing and also put up with some minor
inconsistencies such as the changes between ‘colon’ and ‘kolon’,
‘Iphigeneia’ and ‘Iphigenia’, and the like.
After the manuscript of the commentary had been delivered to
the Press in the summer of 1946, no major additions or alterations
could be made. Consequently it has not been possible (except in a
few particular cases) to take into account anything published after
that date. But far worse is the curtailment of my literary resources
caused by the war and by the conditions in the years immediately
after the war, when only a small portion of the output of foreign
scholarship reached this country (even now some of the gaps in our
sets of continental periodicals have not been filled). In these circum-
stances anything like complete information proved impossible. I
especially regret that Dr. P. Groeneboom’s commentary on the
Agamemnon (Groningen 1944) came into my hands too late.

From the tale of my shortcomings I now turn to the more agree-

able task of thanking my many benefactors. There was a period
during the war when I thought that I should have to give up all hope
of having the commentary translated within any foreseeable time,
for the scholars who might have been both able and willing to take
part in such a work were then either serving in the forces or engaged
in some important war work or overburdened with teaching. At that
juncture Professor Roger Mynors took the initiative and convinced
me by his own generous action that my gloom was not justified.
For several months he sacrificed, after the day’s heavy duties at the
Treasury, many night hours and translated a large and particularly
tricky piece, thus getting my book out of the worst of the deadlocks
it has had to pass through. It was with equal spontaneity and
unselfishness that a few years later Professor T. B. L. Webster and
Mrs. Webster (Miss A. M. Dale) gave up for my sake the precious
spare time left over by their exacting work for the Foreign Office ;
they made it possible, under exceedingly unfavourable circumstances,
to produce a translation of some extensive sections. No less great is
the service done to my book by Miss Christina Barratt’s readiness,
in the face of serious obstacles, to render some long and difficult
xii
 PREFACE
pieces. Other helpers to whom I am very much indebted for their
share in Englishing the commentary are Mr. G. J. Boyden, Mr. F. C.
Geary, and one who wants to remain unnamed. Miss Barbara Flower
contributed some valuable suggestions to the translation of the
Greek text.
In the last ten years the name of Miss Margaret Alford has ap-
peared more than once in the prefaces of works of Greek scholarship
published by the Clarendon Press ; the present preface is fortunately
in a position to follow suit. An indefatigable worker, Spartan in the
austerity of her life, a rigid and accomplished grammarian, Miss
Alford is at the same time the soul of gentleness. Of her genuine
understanding of great poetry I have ample proof. She spent a long
time going through the English version of my commentary, correcting
errors of the translators and blunders of my own, and giving me,
always with disarming modesty, all sorts of invaluable advice.
As for the reading of the proofs, I have been particularly lucky in
enjoying therein the persistent support of two friends, Professor
Rudolf Pfeiffer and Mr. W. S. Barrett. Their efforts were by no
means confined to the correcting of misprints and the ironing out
of inconsistencies. Pfeiffer, whom I used to consult long before it
came to printing, has put all his learning and judgement ungrudg-
ingly at my disposal; what that means will be clear to all classical
scholars. One point at least I must mention specially. But for
Pfeiffer’s perfect familiarity with the Greek lexicographers, etymo-
logists, paroemiographers, and so forth, I should never have been
able to find my way through those mazes. Barrett, for his part,
worked his way into the commentary with such patience and zeal
that he soon remembered all its minutiae far better than I did myself.
But while tidying up countless minor inaccuracies, Barrett, being
the kind of scholar he is, never lost sight of the major issues. His
manner of tackling problems that had baffled me is shown by a few
samples in the Addenda, but the instances in which his knowledge
of Greek and his critical acumen came to my rescue are many more.
His greatest contribution to my book, however, lies in his overhaul-
ing of the first volume, which he read, and corrected, in the manu-
script ; the translation especially has been improved by him almost
out of recognition.
From other friends at Oxford I often had to seek information on
points of detail, and they never let me down. I am in various ways
indebted to Dr. R. W. Hunt, Professor Paul Jacobsthal, Professor
Felix Jacoby, Mr. Edgar Lobel, and Professor Paul Maas.
Outside Oxford it is to two sons of that great mater studiorum,
Basle, that I owe a debt of profound gratitude. At the initial stage
of my struggles with the problems of the Agamemnon Professor Peter
Von der Mühll often discussed difficult passages with me, applying
xiii
 PREFACE
to them his rich scholarship and his mellow wisdom. He confirmed
me in the conviction that there was still a great deal to be done for
the interpretation of the play. Later on, when the commentary
began to take shape, Jacob Wackernagel followed its growth with
the most animating sympathy. By word of mouth as well as in letters
and postcards (the last of them was written in February 1938, a few
months before his death) he urged me to persist in my efforts. He
never tired in his willingness to reply to any question I might put
to him ; some of his answers will be found in the commentary.
Several libraries have been most liberal in granting me every
facility I could wish for. So my sincere thanks go to the authorities
in charge of the Bodleian Library, the Ashmolean Museum, the
Taylor Institution, Corpus Christi College (Oxford), New College,
and Christ Church ; moreover, to the librarians of Cambridge Univer-
sity Library, the Biblioteca Medicea Laurenziana at Florence, the
Biblioteca Nazionale at Naples, the Bibliothèque Nationale at Paris,
and the Bibliotheek der Rijks-Universiteit of Leyden (where Dr. H. J.
Drossaart Lulofs was a kind and efficient mediator).
What Dr. Colin G. Fink, Head of the Division of Electrochemistry
in the Department of Chemical Engineering at Columbia University,
New York, has done for me may be seen from Ὁ. 204 f. of the com-
mentary.
I am deeply indebted to the Delegates of the Clarendon Press, who
at a very difficult time did not hesitate to undertake the publication
of this bulky book, and no less deeply to the Secretary to the Dele-
gates and the staff of the Press, who have carried through an exceed-
ingly troublesome piece of work with great patience and admirable
skill. I am also very thankful to the Trustees of the Jowett Copyright
Fund for making a generous grant towards the expenses arising
from the translating of some parts of my German manuscript.

The dedication names those two who have made it possible for
me to write the book. Professor Sir J. D. Beazley began to take an
interest in my work soon after I had settled down at Oxford ; he has
never failed me since. The encouragement given me by him and by
Jacob Wackernagel determined me to persevere in what I believed
to be a far too ambitious plan. For many years Beazley, with un-
changing generosity and endurance, continued to read twice the
draft of every section of the commentary and the translation of the
Greek text. He would then write his criticisms and suggestions in
the margin and afterwards discuss all the difficulties with me at
great length, often returning to a point with which we had been
dealing before. These talks alone were an abundant reward for all
I was able to do. Beazley’s name appears in the commentary in
many places, but my debt to him goes far beyond anything I owe
χὶν
 PREFACE
him in detail. Of my wife it must suffice to say that she has made
greater sacrifices for this book than anyone else, and, moreover, that
at all stages of my work she has given me the kind of help which
only she could give.
E. F.
CORPUS CHRISTI COLLEGE, OXFORD
2 January 1950

XV
 CONTENTS
VOLUME I
PROLEGOMENA
I. The Manuscripts. . . . . . . I
11. Some Editions and Commentaries . . . . 34
APPENDIX I. The Evidence for Casaubon’s Work on
Aeschylus . . 62
II. John Pearson's Share in Stanley" S Aeschylus 78
SIGLA LIBRORUM . . . . . . . . 86
TEXT AND TRANSLATION . . . . . . 87
PLATES
I. MS Tr (Farnes. Neap. ii. F. 31): Ag. 1-14 at
11. MS F (Laur. xxxi. 8): end of hypothesis of Ag., Ag. 13) end

VOLUME II
A SELECT LIST OF ABBREVIATIONS. . . . vi
COMMENTARY ON I-I055 . . . . . . I

VOLUME III
COMMENTARY ON 1056-1673 . . . . . . 481
APPENDIXES
A. On the Postponement of certain Important Details in
Archaic Narrative . . 805
B. On the Weapon with which, according to the Orestei,
Agamemnon was murdered . 806
c. Cho. 991-1006 . . . . . 809
D. The Footprints in the Choephoroe . 815
E. Short Syllables before Initial Mute and Liquid in
i the
Lyrics of Aeschylus . 826
F. The Word-order in Ag 1434 οὔ μοι μου lady Deis
épmaTei . 827

ADDENDA ET CORRIGENDA TO COMMENTARY . . 829

INDEXES . . . . . . . . . 833

xvi
 PROLEGOMENA
I
THE MANUSCRIPTS
Our oldest manuscript of the Agamemnon is a very small fragment
on papyrus, Pap. Oxy. 2178, ascribed by Lobel to the second century
after Christ. It contains a few letters (varying from one to ten a line)
of the beginnings of ll. 7-30. As one would expect, it agrees at 23
with MV (φάος) against ΕἼΤ (νῦν φῶς) and contains 1. 7, an interpola-
tion which is presumably pre-Alexandrian.
Next comes M (the Mediceus), codex Laurentianus xxxii. 9, parch-
ment, written as it seems at the beginning of the eleventh century,
one of the most illustrious of Greek manuscripts. There is no need
to describe it again here or to outline its history. To peruse the
splendid facsimile published by the Italian Ministero dell’ Istruzione
Pubblica in 1896 is a continuous delight; any student of Aeschylus,
however young and inexperienced, should attempt to make himself
familiar with the clear and easy script of this great book. A good
guide to the part of the MS which contains the text of Aeschylus,
the only part that concerns us here, is provided in Rostagno's
succinct and learned introduction to the facsimile; to this should
be added Wilamowitz's remarks in the preface to his edition of
Aeschylus. For further information on M cf. H. W. Smyth, Harvard
Studies in Class. Philol. xliv, 1933, 17 ff., and A. Turyn, The Manu-
script Tradition of the Tragedies of Aeschylus (Polish Institute of
Arts and Sciences in America, New York City 1943), 17 ff., and the
books and articles quoted by him.
As regards the corrections in the Mediceus, we are not concerned
here with the several correctors of the Renaissance (14th or τοί
century), who confined themselves almost entirely! to the plays of
the ‘Byzantine triad’, i.e. Prom., Sept., and Pers. All that is relevant
to our present purposes is the body of corrections made by the first
διορθωτής. These corrections are contemporary with the text of the
codex.? Here, as in many similar cases, the person responsible for

τ ‘quasi esclusivamente' says Rostagno, op. cit. 14 (in his select examples he quotes
a few corrections by Renaissance correctors from the Choephoroe, but none from the
Agamemnon and the Eumenides). Perhaps this point should some time be examined
afresh by another scholar, who ought to be no less experienced a palaeographer than
was Rostagno,
2 This would be the case, as far as the bulk of the text of Aeschylus (i.e. all save the
first eight leaves containing Pers. 1—705) is concerned, even if Rostagno was right when
he assumed, though with noticeable hesitation, that the text of the first eight leaves
was written in the second half of the roth century and the rest in the rith century. On
this point I cannot venture a judgement of my own, but for the reasons put forward by
4872.1 I B
 PROLEGOMENA
a fresh copy of a classical text intended from the outset that a
διορθωτής should join the copyists and supplement their work: the
book would not have been regarded as complete until his corrections
were entered. I have used the sign m for the corrections and additions
of the διορθωτής.
It was the opinion of Rostagno (op. cit., p. 12) that the διορθωτής
of the Mediceus had access to another exemplar, with the help of
whichhe emended his text and added the scholia and also variants
and interlinear glosses. This idea was emphatically repudiated by
Wilamowitz (Aeschyli tragoediae, p. xif.). For a further discussion
of the problem see H. W. Smyth, op. cit. 45 f., who on the whole
seems inclined to agree with Wilamowitz. In a case like Ag. 1127
it would be helpful if a clear answer to this question could be given.
When, between 1421 and 1423, Giovanni Aurispa acquired the
Mediceus, the codex had already suffered the damage which has had
such serious consequences for the text of the Agamemnon and even
worse consequences for the prologue of the Choephoroe.! ‘At a period
which nothing in the book itself enables us to fix, the quires of
Aeschylus became unsewn, and one whole one, and three quarters
of the next were lost’ (T. W. Allen, J. Phil. xxii, 1894, 183). This loss
of the entire 18th quaternion (eight leaves, or sixteen pages) and six
leaves of the 19th meant the disappearance from M of Ag. 311-1066
and 1160-end, and of the ὑπόθεσις of the Choephoroe and the beginning
of its prologue.
The ‘codex Bessarionis’, Venetus Marcianus 468, paper, in the
Biblioteca Nazionale di San Marco,” was, as it seems, written in the
thirteenth century. With Wilamowitz and others I call it V (in
Weil’s and Mazon’s editions it is B). It contains the Byzantine triad
with scholia, followed by a list of the plays of Aeschylus, the ὑπόθεσις
to the Agamemnon, and Ag. 1-348 (written in three columns to a page,
whereas the text of the triad is written in two columns to a page; cf.
Turyn, op. cit. 28). There are in V no scholia to the Agamemnon.
‘Apparet scribae praeter triadis librum commentario instructum in-
notuisse codicem, e quo supplementa illa [the list of plays, the
ὑπόθεσις to Ag., and Ag. 1-348] sumpsit’ (Wilamowitz, p. xv). In the
text of the Agamemnon the words are often rather crowded together
(in consequence, apparently, of the arrangement in three columns) ;
Wilamowitz in the preface of his edition, p. xi, Rostagno’s hypothesis seems to me im-
probable. However, leaving aside this minor issue, we are justified in regarding the
whole pre-Renaissance Aeschylus matter in M, i.e. the text, the scholia, and the correc-
tions of the διορθωτής, as a product ‘eiusdem aetatis et eiusdem scholae’ (Wilamowitz,
loc. cit.); cf. also T. W. Allen, J. Phil. xxii, 1894, 168, and Turyn, op. cit. 18.
1 See the letter written by Ambrogio Traversari to Niccold de’ Niccoli in May 1424
(reprinted by Rostagno, op. cit. 8 n. 2), from which it appears that the 14 leaves had
already been lost at that date.
2 ‘The number now used in the Library, but not published in any catalogue, is 653’
(H. W. Smyth, op. cit. 29 n. 1).
2
 THE MANUSCRIPTS
they are, however, perfectly legible, except in the upper lines of the
pages, where the ink has faded.
The next MS in order of time is presumably Tr,! in the Biblioteca
Nazionale in Naples, cod. II.F.31, formerly called Farnesianus ;?
paper. It is written by the hand of Demetrius Triclinius;? conse-
quently its date is about the first quarter of the fourteenth century.*
It contains the Byzantine triad and Ag. (complete) and Eum. (with
two large lacunae; see p. 7). There are rich scholia. These fall into
two groups: (1) σχόλια παλαιά, 'introduced by capital letters (which,
in the commentary, occur only for this purpose) and by the projection
of their initial part from the rest of the column',5 and (2) Triclinius'
own scholia, marked as such by the note ἡμέτερον (or ἡμέτερα) and
11 sympathize with the point of view of D. S. Robertson, who declares himself
(C.R. lvii, 1943, 111) in favour of the symbol T instead of Tr; but there have already
been so many embarrassing alterations in Aeschylean nomenclature that I prefer to
continue the practice introduced by Wilamowitz and accepted by Mazon, Murray, and
G. Thomson. This has also the advantage of reminding the reader of the fact that the
editor of this book, Demetrius Triclinius, has a distinct individuality of his own.
* For further details (such as size, etc.) of this and other MSS, and for modern publica-
tions dealing with them, I refer once and for all to H. W. Smyth's thorough monograph,
‘Catalogue of the Manuscripts of Aeschylus', Harvard Studies in Class. Philol. xliv, 1933,
1 ff., and to Turyn’s book quoted above.
3 This has been stated by several scholars; recently by Turyn, op. cit. 102 f. E. Lobel
and P. Maas have examined with me many pages of the photostats of the Naples MS
(Tr) of Aeschylus and compared them in the Bodleian Library with the original of
MS New College 258, containing Aphthonius and Hermogenes, which bears the sub-
scription διὰ χειρὸς δημητρίου τοῦ τρικλίνη and the date August 1308 (cf. Turyn, op. cit.
103 n. 89); we have also compared the facsimile (Wattenbach-Velsen, Exempla codd.
graec., pl. 21) of the last page of Triclinius' dated (1319) autograph of Hesiod (in Venice).
Despite certain differences in detail, the general character of the script is very much
the same. Lobel and Maas are agreed that Tr was in all probability written by Triclinius ;
at any rate (as Lobel puts it) anyone who tried to deny the identity of the hands would
have to produce very strong arguments to prove his point. No such argument is pro-
vided by an assertion which was first put forward by Friedrich Kuhn, 'Symbolae' etc.,
Breslauer Philol. Abhandlungen, vi, 1892, 101, and repeated by W. I. W. Koster, Scholia
in Arist. Plut. et Nub. (Leyden, 1927), p. ii f., and by K. Holzinger in X APIZTHPIA
Alois Rzach dargebracht (Prague 1930), p. 75 n. 21. These scholars say that some of the
metrical scholia on the Agamemnon in Tr, though obviously Triclinian, are marked as
σχόλια παλαιά, and that since such an error could not possibly be ascribed to Triclinius
himself the MS must be regarded as a copy of his autograph. This assertion is based
on an inaccuracy in van Heusde's edition: all the metrical scholia on the later part of
the play (from 930 [852 van Heusde] on; for the earlier part van Heusde's notations are
correct) which he alleges to be παλαιά are in the MS clearly marked as Triclinius’ own;
in this respect Dindorf’s publication (Philol. xx, 1863, 46 f.) is more reliable.—Of the
differences between Tr and the other two Triclinian MSS referred to above I will mention
one which struck my eye at once: in the New College MS of the rhetoricians and in the
Venice Hesiod the breathing signs, for both asper and lenis, have the common rounded
form of the later period, but in Tr they are invariably angular, and, furthermore, of
the earlier, ‘completer’ (cf. V. Gardthausen, Griech. Palaeographie, 2nd ed. ii. 386)
angular type, i.e. F and 4, not L and , J. According to Gardthausen, op. cit. 388, this
is a feature of the archaizing minuscule of the later period; I do not know whether in
the case of Triclinius’ Aeschylus this detail could be ascribed to the influence of his
exemplar.
+ Cf. Turyn, op. cit. 103 f.
5 H. W. Smyth, Harvard Studies, xxxii, 1921, 93.

3
 PROLEGOMENA
a cross or simply by the cross.’ On the left-hand margin of the page
reproduced as plate I the two types can be easily distinguished:
beneath the ornament opposite the interval between the first two
lines of the text is the heading σχόλια παλαιά, and underneath it are
the two ‘old’ scholia, the one beginning with Aorelws and the other
with Kuvos δίκην; while at the top of the marginal column is the
heading ἡμέτερον, and underneath it a cross and then ἡ εἴσθεσις τοῦ
δράματος κτλ.
The next MS is F, in the Biblioteca Laurenziana in Florence,
cod. xxxi. 8, paper.? For evidence of its date we do not have to
depend on palaeographical indications, which for MSS of the four-
teenth and fifteenth centuries at any rate seem to be rather elusive.
An approximate terminus post quem, viz. after the first two or three
decades of the fourteenth century, is provided by the fact that the
metrical scholia, written by the same hand as the text, are derived
from the commentary of Triclinius (see below, pp. 16 ff.); and a
precise terminus ante quem is provided by a note on the last page of
the MS (after the subscriptio at the end of the text of Lycophron)
in which somebody whose hand is entirely different from that of the
copyist, presumably the owner of the book, mentions the death of
his wife in the year 1374. This important entry,? which was pub-
lished in the eighteenth century by A. M. Bandini, Catalogus codicum
Graecorum Bibliothecae Laurentianae, ii. 84, is never mentioned by
editors of Aeschylus ;* there is, however, a reference to it in M. Vogel
and V. Gardthausen, ‘Die griech. Schreiber des Mittelalters und der
Renaissance’, Bethefte zum Zentralblatt für Bibliothekswesen, xxxiii,
1909, 259 n. 4, and in Turyn, op. cit. 70 (for H. W. Smyth, see foot-
note 4 below) ; let us hope, then, that future editors will not forget
it again. The MS contains the Byzantine triad, Ag. (complete), and
Eum. (with the same two lacunae as Tr).
The corrections in F, and its scholia (metrical and others), glosses,
and variants, will be discussed in detail later on. Here it will be suffi-

1 For the manner in which Triclinius in his edition of the bucolic poets distinguished
his own scholia from the παλαιά see C. Wendel, ‘Uberlieferung und Entstehung der
Theokrit-Scholien', Abhandl. d. Gesellsch. d. Wissensch. zu Gottingen, Phil.-hist. KL,
N.F. xvii. 2, 1921, 31 ff.
2 Since Tr is older than F, it may seem strange that I have generally referred to the
two MSS in the order FTr. My reason for doing so is that the text of F is pre-Triclinian,
as will be shown below.
3 I have a photograph of it.
4 The consequences of this oversight are remarkable. Blass, Die Eumeniden des
Aischylos, p. 18, criticized Wecklein for dating F to the first half of the 14th century,
and added that he himself did not know why the MS should be older than the 15th
century. Wilamowitz (p. xix) said 'saec. XIV, vergente, opinor, scriptus' ; Mazon (vol. ii,
p. xx) ‘xive ou xv* siècle’ ; Murray ‘saec. X1V vergente scriptus’ ; and even H. W. Smyth,
Harvard Studies, xliv, 1933, 16, though he quotes from Vogel-Gardthausen the entry
of the year 1374, dates the MS 'xiv-xv'. So here, once again, ‘non impune Bandinium
neglexerunt editores' (Wilamowitz, Anal. Eur. 4 n. 6).
4
 THE MANUSCRIPTS
cient to state briefly that at any rate in the section containing the
Agamemnon not only the variants but also the scholia and most of
the glosses are by the same hand as the text,! but that some of the
glosses show a different, and perhaps later, hand.
The last MS to be used by me is the Venetus Marcianus 616 (now
663),2 which I call G with Wilamowitz and others (in Mazon’s edition
it is V) ; parchment. The more recent editors ascribe it to the fifteenth
century. It contains the Byzantine triad and Ag. and Eum. In the
Eum. there are the same two large lacunae as in Tr and F, but unlike
those MSS G does not contain the whole Ag. : eight leaves, containing
ll. 46-1094, are missing. It has metrical scholia on Pers., Ag., and
Eum., but no other scholia. On the relation between F and G see
p. 30f.; from what is said there it will appear that in most cases it is
unnecessary to mention the reading of G, but that, on the other hand,
this MS is by no means without its value for the reconstruction of the
hyparchetype.
About the ‘codex Romanus’ (E in Wilamowitz’s edition), in the
Biblioteca Nazionale in Rome, cod. graec. 5, one sentence will
suffice. After Pasquali’s re-examination of the MS and his clear ver-
dict (in the article quoted in footnote 1 below) this book should meet
with its deserved fate: οὔτ᾽ ἐν λόγωι οὔτ᾽ ἐν ἀριθμῶι.

From the time of G. Burges,? Cobet,* and W. Dindorf until fairly

recently it was commonly believed (though there were a few eminent
dissenters)® that all our extant MSS of Aeschylus, including those
which contain only the Byzantine triad, are derived from the
Mediceus (M). This belief has now been shaken, mainly, so it seems,
as a result of Wilamowitz’s edition. As far as the text of Prom.,
Sept., and Pers. is concerned, one need hardly fear that, if the
fashion changes again, sensible scholars will relapse into the former
error. But the problem is different, and the risk of a reactionary
movement greater, in the case of Ag. and Eum. For here, as Wilamo-
witz? admitted in 1889 (Euripides Herakles, vol. i, 1st ed. = Ein-
leitung in die griechische Tragödie, 204 n. 166), there seems to be a
1 This conclusion, at which I arrived after a long study of the photostats, has met
with the consent of both Lobel and Maas. Thus we agree with Pasquali, Rendicontt
dell’ Accad. dei Lincei, Ser. VI, vol. vi, 1930, 39, and Turyn, op. cit. 1or, against Wilamo-
witz, P. Xix.
2 Cf. H. W. Smyth, Harvard Studies, xliv, 1933, 29 n. 1.
3 See his edition of the Suppl. of Aeschylus (1821), p. 42, and of the Eum. (1822),
. 39 f.
P 4 The relevant sentence from his Oratio de arle interpretandt (1847) is quoted by
A. C. Pearson in his Oxford edition of Sophocles, p. iii f.
5 See especially Philol. xviii, 1862, 55 f.
6 To those whom I shall mention presently we may add Th. Bergk, Neue Jahrbücher
für Philol. und Paedag. xcvii, 1868, 374.
7 The caution of this early remark is very different from the easy optimism in his
later treatment of the same subject.

5
 PROLEGOMENA
lack of really decisive passages. This being so, it is not only useless
but positively harmful to amass inconclusive instances in an attempt
to prove that the later MSS are independent of M. No purpose is
served by referring, for example, to Ag. 297 (παιδίον ὠποῦ MV, πεδίον
ἀσωποῦ FTr), as was done by H. L. Ahrens (he regarded it as ‘incon-
ceivable’ that the correct reading here should have been the result
of a Byzantine conjecture), Wilamowitz,' p. xix, and Turyn, op. cit.
111. For it would be very difficult to reject Weil’s argument (p. iv
of his Teubner edition): 'potuit vera lectio felici coniectura reponi
a grammatico Byzantino Persarum versus 805 [μέίμνουσι δ᾽ ἔνθα πεδίον
Ἀσωπὸς ῥοαῖς | ἄρδει] memore'. Fortunately we need not rely on
such doubtful support. Even for Ag. and Eum. (and it is the text of
these plays only that concerns us here) the evidence, though not
ample, is yet sufficient to show that VF(G)Tr are independent of M;
in a later section some instances from the scholia will be added.
The main credit for refuting with substantial arguments the
assumption that the later MSS containing Ag. and Eum. are derived
from M goes to Ahrens (Philologus, Supplementband i, 1860, 216 f.),
F. Heimsoeth (in his book Die indirecte Überlieferung des aeschy-
leischen Textes, 1862, 180 ff., and in several ‘Indices lectionum’ of
Bonn), C. Brennan (‘On the Manuscripts of Aeschylus’, J. Phil. xxii,
1894, 49 ff.), and F. Blass (introduction to his commentary on Eum.,
p. 19 f.). Although each of these scholars included in his list passages
which do not in fact prove his point, each of them nevertheless was
able to produce really conclusive evidence, e.g. Ahrens Ag. 1133
and 1152, Brennan Ag. 79, 137, and 1143. Most of the instances dis-
cussed below were adduced by Blass.
I begin with V (including readings of V shared by FTr). Here the
harvest is bound to be meagre, for we have only 348 lines to go upon,
and, moreover, the readings of V agree to a large extent with those
of M (see below). But what we do find is not to be despised. At Ag. 79
τόὀθιπεργήρως in VF is a step nearer to the original reading than
τίθιπεργήρως in M, but since it is equally unintelligible it cannot be
regarded as an attempt to emend τίθιπεργήρως, and consequently
must be derived from a source independent of M (for an explanation
of the wrong ending see the commentary); when Dindorf, Philol.
xviii, 1862, 64, ascribes the reading τόθιπεργήρως to mere accident
('beruht wohl nur auf Zufall’), he shows clearly to what subterfuges
he is driven by his prejudiced view that M is the only source of the
other MSS. At Ag. 72 V’s áríra«a, again independent of M, is in all
probability conflated from the variants ἀτίται and ἀτίτᾳ. An even
more interesting varia lectio is preserved by V at Ag. 137 (rightly
appreciated by Brennan, op. cit. 53): πτάωνκα is presumably the
! The conclusion which in the same context Wilamowitz draws from Eum. 959 κύριες
has been shown to be wrong by P. Maas: see my note on Ag. 562 (vol. ii, p. 284 n. 2).
6
 THE MANUSCRIPTS
result of JITAKA in the archetype,' whence M chose πτάκα and the
source of FTr πτῶκα.
I now come to the relation between M and the group F(G)Tr. It
is generally recognized that this group represents a single hyparche-
type; the two large lacunae in Eum. common to the three MSS, from
582 to 644 and from 778 to 807, are alone sufficient to prove the
common origin. The following passages show that the hyparchetype
was not derived from M. In Ag. 1133 the meaningless δὴ αἱ of ΕἼΤ
would not have been written by anyone who had the διὰ of M before
him: it goes back to the original διαὶ. In Ag. 1143 F has φιλοίκτοις
ταλαΐναις (i.e. the genuine φιλοίκτοις together with the variant or
gloss ταλαίναις), and so obviously had the exemplar of Tr (Triclinius
himself having clipped it metri gratia); in M, on the other hand,
φιλοίκτοις has disappeared.
We now turn to a number of passages in which the readings of
FTr coincide with readings of the first scribe of M which were
subsequently erased, or altered in some other way, by the διορθωτής
(m). The instances which come under this head are of such a kind
that while the trained eye of a modern palaeographer finds no
difficulty in making out the writing of the first hand, a Byzantine
copyist would not have noticed anything but the correction of the
διορθωτής.Σ For our present purpose only those instances are to be
regarded as relevant in which it is virtually impossible that the
reading of FTr should be due to the conjecture or mere error of a
scribe who had the corrected reading of m before him. In Ag. 245
VF have the εὐπόταμον which was originally written in M (Triclinius
gives a metrical conjecture of his own): they (or their hyparchetype)
would certainly not have produced this non-existent and nonsensical
word if they had found the perfectly good εὔποτμον which we now
read in the Mediceus (one has to look rather carefully to discover in
the erasure a dim trace of the a). In Ewm. 262 (not included in Blass's
list) the δυσαγκόμιστρον of FTr coincides with the original reading
of M, where the p has completely disappeared in the erasure. If in
Eum. 299 the scribe of the hyparchetype of FTr had had in front of
him οὔτοι, i.e. the reading by which the corrector of M supplanted
the earlier οὔτις, there would have been no reason why he should
have changed it to οὔτις. Nor was there in Eum. 365 any conceivable
cause for giving up Ζεὺς (the o which the corrector of M added at the
end is perfectly clear) in favour of Ζεῦ (this instance is absent from
Blass's list): we must therefore conclude that the exemplar of FTr
had Ζεῦ as had the exemplar of M. In Eum. 950 the correct? reading
! For the intrusion of a gloss into the text of V see 153.
2 Blass, Eum. p. 19, wisely puts between square brackets the passages in which the
letters originally written in M are still recognizable enough to have struck the eye of
a Byzantine reader, viz. Eum. 272 and 424.
3 Cf. my remark in vol. iii, p. 628.
 PROLEGOMENA
ἐπικραίνει, in which FTr agree with the first hand of M (where the
t before the ν has been completely erased by the corrector), cannot
be accounted for by assuming that the scribe of the hyparchetype
of ΕἼΤ objected to ἐπικρανεῖ on metrical grounds, for from Ag. 1340
it is clear that he scanned this form as © ~~—; nor would he have
taken exception to the meaning of ἐπικρανεῖ, since the future tense
of the verb in its ordinary sense would seem to suit the context
very well.
Certain other instances in which the reading of M before its cor-
rection is identical with that of FTr do not provide an absolute
proof that FTr are independent of M in its final form, since there we
must admit the possibility, though not the probability, that a scribe
with M before him altered its reading either by mistake or as a
deliberate correction, in which case the coincidence with the original
reading in M would have to be ascribed to mere chance. Thus it is
possible that in Ag. 1152 someone starting from ἐπὶ φόβωι produced
ἐπίφοβα by a lucky conjecture. In Eum. 211 (the first scribe of M
wrote τίσ yap, but the o has been completely erased) the reading
tis yap in FTr might be due to the error of a copyist whose eye strayed
to the beginning of 209 ris nôe. The weakest item in Blass's list is
Eum. 330 (παράφρονα FTr and perhaps! the first hand in M, παραφορὰ
M after the correction), for here it would be natural for a copyist
to alter his text to agree with the antistrophe (343), where the MSS
have παράφρονα.
As regards the σύναρθρον of FTr in Ag. 254, I am inclined to regard
it with Blass* as a corruption of the original σύνορθρον rather than
a conjecture based upon συνορθὸν (M), for it is only as a grammatical
term that σύναρθρος is commonly used.
In our task of demonstrating that the group FTr is independent
of M we are not very considerably helped by the scholia, for most
! See, however, as against Blass and Wilamowitz, the app. crit. of Vitelli-Wecklein.
I cannot make out from the facsimile what letter has disappeared in the erasure
between ¢ and o. * Brennan, op. cit. 53, reserves judgement on this point.
3 Blass (Eum. p. 20) quotes two passages from the F scholia on the Ag. in order to
show that they are independent of M. One of them, on 283 (adduced for the same
purpose from Tr by F. Heimsoeth, ‘De scholiis in Aesch. Ag. . . .', Index lectionum
of the University of Bonn, 1868/9, ix), does not prove the point, for from the name
Ἑρμαῖον anyone could infer that this was a place ἐν ds ἐτιμᾶτο ὁ ‘Epuñs; the other, on
58, makes a much better case, and will be discussed in our examination of the F scholia.
Wilamowitz (p. xix) mentions four scholia in Tr which in his opinion prove that the
group FGTr is independent of M (all of them had been used for the same purpose by
Heimsoeth, Die indirecte Überlieferung des aesch. Textes, 181 f., and ‘De scholiis in
Aesch. Ag.’, vii); but none of them is really conclusive. One of them is the note on
Ag. 1093, where the σχόλ. παλ. in Tr have ws κύων εὗρις ζητεῖν (slightly misquoted by
Wilamowitz). Here Wilamowitz admitted himself (Hermes, xxv, 1890, 162 n. ı, where he
also acknowledges his debt to Heimsoeth) the possibility that the reading in Tr is the
result of a Byzantine conjecture, and this is in fact very likely : no scribe who had his
eye on I. 1093 as a whole (and, perhaps, remembered the beginning of the Ajax) could
fail to recover the original reading edpis (or εὕρις) from the slight corruption εὑρίσκοι
8
 THE MANUSCRIPTS
of them have come down to usin a rather attenuated form. We shall
be in a slightly better position when we come to analyse the non-
metrical scholia in F (p. 24 f. below). For the moment I shall confine
myself to drawing attention to two remarkable pieces of evidence
which provide a welcome confirmation of what has been stated above.
It is surprising that the strange θυοσκεῖς (or θνοσκινεῖς as the MSS
have it) in Ag. 87 is in M left without any explanation ; and when we
then find in the σχόλ. wad. of Tr a variant (εὕρηται καὶ κτλ.) and added
to it a substantial comment (see my note), there can hardly be any
doubt that this comes from the original stock of the scholia and
consequently shows the hyparchetype of FTr to be independent of
M. We are probably justified in drawing the same conclusion from the
scholion on Ag. 177 μάθος. M has no note at all, but in the oyóA. παλ.
in Tr we read this: Avoxpnoros μὲν ἡ λέξις, ἀλλ᾽ Ἀττική" ὡς γὰρ τὸ
ῥέος καὶ τὸ βλέπος, οὕτω καὶ τὸ μάθος καὶ ἔτι τὸ δίψος καὶ τὸ βλάβος καὶ
τὸ κλέπος. For similar comments cf. e.g. Prom. 400, where (on ῥέος)
M has the scholion ῥεῦμα. παρὰ τὸ ῥέω ῥέος ὡς κλέπτω κλέπος KTA.,
and cf. also Schol. Ar. Peace 528 (on πλέκος). It is not of course
inconceivable that a Byzantine scholar whose knowledge of the text
and scholia of Aeschylus was confined to what he found in the
Mediceus or one of its descendants should have added at Ag. 177
a note on μάθος which he compiled from the lexicographers, but this
is most unlikely, especially as we do not find the words μάθος, δίψος,
βλάβος, κλέπος in the list of similar formations which comes nearest
to that in our scholion, viz. the one in Suid. s.v. πλέκος.
Before examining in detail the relation between F and Tr and the
changes which the text and scholia of their common hyparchetype
have undergone in each of them, something must be said about the
position of V in regard to M on the one hand and to FTr on the other.
This position has been well summed up by Brennan, who says
(p. 55) ‘V sides now with M against FGTr, now with the latter against
M’, and by Wilamowitz, p. xxi, who in discussing the text of Ag.
1-348 as presented by V says ‘neque ad Mediceum haec redeunt
neque ad idem exemplum quo FGTr usi sunt, sed modo ad hoc modo
ad illum se applicant'. So V shares with M (against FTr) the genuine
readings 5 θέρος βροτοῖς, 23 φάος, 48 κλάζοντες, 80 τρίποδας, etc., and
the corruptions 2 μῆκος δ᾽ ἦν, 29 ἐπορθριάζειν, 82 ἡμερόφατον, 123
λογοδαίτας, etc.,! and on the other hand shares with FTr (against M)
of the scholion in M. Again at Eum. 54, where the last word of the line was corrupted
in the παράδοσις, the scholion in M, οἷον αἱματηράν (cf. Cho. 1058), would enable even
a poor scholar to elaborate it and write αἱματηρὸν araAaypóv (σχόλ. παλ. in Tr). The last
item in this group (for Ag. 1672, where M is missing, is of no use here) is Eur. 560; here
again Heimsoeth and Wilamowitz have not proved their point, for Triclinius, as was
his wont, changed the θερμοεργῶι of his exemplar’s text to Qepuóx, to square the metre
with that of the antistrophe, and then added the former word asa gloss, ἤγουν θερμουργῶι.
! At 198 V provides a typical example of accumulating corruption, for the starting-
9
 PROLEGOMENA
the genuine readings 79 τόθ. (cf. p. 6), 119 ἐρικύμονα, 141 ἀέπτοις, etc.,
and the corruption 26 σημανῶ.
The first few items in this brief list, besides illustrating the relative
value of V, may also serve a different purpose: they reveal certain
changes which the text has undergone in the hyparchetype of
F(G)Tr and thus lead to the unwelcome conclusion that for the part
of the Agamemnon in which only F(G)Tr are extant, i.e. for 1,232
lines out of 1,673, the foundations of the text are far from reliable.
The worst feature is not deterioration through inadvertent miscopy-
ing such as is bound to happen in the course of transmission, but
a number of deliberate and wanton alterations, as e.g. at 5 βροτοῖς
θέρος instead of θέρος βροτοῖς, at 23 νῦν φῶς instead of φάος, at 48
«Ady£avres instead of κλάζοντες, to which we may add, e.g., 98 εἰπεῖν
instead of αἰνεῖν, 286 ὑπεὶρ ἔλης instead of ὑπερτελής, 1094 ἐφευρήσει
instead of ἄν edpyjon (or ἀνευρήση), 1095 μαρτυρίοις μὲν yap instead of
μαρτυρίοις yap (interesting because it shows a pre-Triclinian! metrical
interpolation in lyrics), 1106 Bods πόλις instead of πόλις Bode (cf. 5).
In some of these instances it would be very difficult to detect the
mischief without the assistance of MV or M. At 340 (where M is not
preserved) the ἀνελόντες of V reproduces the archetype, and ἄν γ᾽
ἑλόντες in FTr is a metrical conjecture.
Now that I have discussed the degree of reliability of our MSS of the
Agamemnon, I will dwell on this question for a moment and will add
an observation that takes us beyond the description, comparison,
and valuation of the extant MSS. The precariousness of the authori-
ties on which in the main our text of the Agamemnon rests comes
out even more strikingly when we look at the quotations furnished
by ancient writers, lexicographers, etc. In the case of this play they
are not very numerous, but nevertheless are quite sufficient to give
us a salutary warning. It is to the text of Aristophanes’ Frogs that
we owe at 109 the genuine ἦβας and at 111 καὶ χερὶ (which in the MSS
of Aeschylus has been ousted by a gloss, and could hardly have been
recovered by conjecture). At 141 the gap in our MSS has been filled
with the help of the Etymologicum Magnum; at both 282 and 284
a rare word has been replaced by an ordinary one, which would
have been accepted without hesitation if the lexicographers (and
Athenaeus) had not enabled us to recover the original reading. At
448 the true form διαὶ, before it was found in the quotation of the

point of its xardgevov was obviously the κατέξενον which we find also in M.—Among the
features which link M and V together, the colometry should not be overlooked: in the
ode 104 ff. the arrangement of the metrical κῶλα in V agrees to a considerable extent
with that in M against the arrangement in FTr.
! The proof will be given below.—It seems to me probable that at 1096, too, the
change in FTr of τάδε to τὰ is due to an attempt to make the line correspond to 1091
(αὐτοφόνα κακὰ καρτάναι). Such ‘corrections’ do not presuppose an elaborate metrical
system: a mere counting of the number of syllables was sufficient for the purpose.
Io
 THE MANUSCRIPTS
"Epimerismi', was restored by Hermann; but at 1356 it was only the
discovery of the so-called Tryphon that revealed the nature of a
particularly insidious corruption, which had been glossed over by
many editors, though S. Butler had the sense to protest ‘sed sanum
esse hunc locum nemo mihi persuadebit’. Of these seven instances,
the first five occur in a part of the play which is preserved in all our
MSS of the Agamemnon, and so prove irrefutably that we are by
no means on safe ground even where we have MV as well as FTr:
the corruptions in question are not peculiar to one or another group
of MSS, but go right back to the archetype. It may be useful here
to repeat the warning in which Wilamowitz (p. xxix) sums up the
inference to be drawn from a longer list of passages from all Aeschy-
lean plays, passages in which faulty readings of the MSS are corrected
with the help of quotations: ‘sufficiunt haec, ut intellegamus vitia
in archetypo infuisse multa, nec pauca ex his coniectura probabilis
tollere numquam poterat. Apparet etiam subinde voces rariores
glossematis expulsas esse, perniciosum corruptelae genus, quoniam
verum deficiente testimonio perraro recuperatur atque studio rariora
captandi multi in avia abrepti sunt.’ The last remark applies parti-
cularly to the large group of scholars who, especially in the course
of the last hundred years, have endeavoured to find a place in the
text of Aeschylus for some homeless gloss in Hesychius. The tempta-
tion to do so and to toy with ingenious combinations will always be
great, but anyone who strives in earnest to deserve the name of
κριτικός Should at least form a clear idea of the pitfalls of this game.
One of the considerations which ought to make us pause was recently
emphasized by Latte, Mnemos. Ser. III, x, 1941, 87: the quantity
of tragic texts from which excerpts were made by Didymus and
Diogenianus, the lexicographers from whose works the bulk of the
material contained in Hesychius is derived, was more than ten times
as large as what is now preserved. I therefore prefer to err on the
side of exaggerated caution. Even in the case of ll. 639 and 677 (see
the commentary on these two passages) it does not seem to me beyond
doubt that we are justified in replacing the text of FTr, poor though
the authority of these MSS often is, by readings gathered from
anonymous quotations in Hesychius.
To return now to the mutual relationship of the extant MSS: it
is of vital importance to the task in hand to form a clear idea of the
character of F(G), and especially of the relation between F(G) and
Tr. We have seen that the text of F(G)Tr in many places shows the
marks of arbitrary interference. That is bad enough. But our position
would be worse, and indeed almost desperate, if those critics were
right who hold that F(G) as well as Tr represents a substantially
Triclinian text.! In that case all we should have to go on for about
τ See T. Blass, Die Eumeniden des Aischylos, p. 18; G. Thomson, The Oresteia, i. 76;
II
 PROLEGOMENA
three-fourths of the Agamemnon would be a text into which, from
what we know about Triclinius, we should have to expect that many
violent changes had been introduced by that learned and ingenious
but extremely reckless doctrinaire. Fortunately we may comfort
ourselves on this head. |
It is possible to obtain a small-scale picture of the typical difference
between the Triclinian and the pre-Triclinian text (F) by glancing,
for instance, at the first part of the first stasimon (355 f.). In 356
Triclinius gives way to his dislike for the paroemiac and consequently
barbarizes the line. To secure corresponsion between strophe and
antistrophe he inserts a stopgap in 379 (cf. 397), adds ὡς at the
beginning of 387 (cf. 369), prefixes év- to θείς in 395 (cf. 377), reads
πολλὰ δ᾽ ἔστενον instead of πολὺ δ᾽ aveorevov in 408 (cf. 425), and adds
a preposition in 383 and the article in its corresponding line 401. The
‘rhythmical refrain’ of the second strophe and antistrophe he gives
in the following form:
str. (416 ff.) εὐμόρφων γὰρ κολοσσῶν
ἔχθεται χάρις Typi!
ὀμμάτων δ᾽ ἐν ἀχηνίαις
ἔρρει πᾶσ᾽ ἀφροδίτα.
ant. (433 ff.) οὖς μὲν γὰρ πέμψεν οἶδεν"
ἀντὶ δὲ βροτῶν τεύχη
καὶ σποδὸς πρὸς ἑκάστου τοὺς
δόμους εἰσαφικνεῖται.
The changes in 433 ff. are especially violent; yet the only flaw in
these eight lines as they were written in the hyparchetype of FTr
was the omission of rıs in 433. It would be easy to go through all
the lyrics of the play and collect a rich harvest of meddlesome re-
writing of this kind ;2 but it seems more profitable to dwell for a
moment on the eight lines we have just noticed and to use them to
demonstrate that the text which was so boldly altered by Triclinius
was very similar to the text of F not only in its words but also in
the arrangement of the lines, i.e. the κωλομετρία. If we want fully to
understand the textual manœuvres of Triclinius, we have to realize
and especially A. Turyn, op. cit. 11o ff. D. S. Robertson says in his review of Turyn's
monograph (C.R. lvii, 1943, 111) ‘no one can doubt that FGT have a common source,
and that this source had been edited in a thoroughly Triclinian spirit’. It is the second
part of this sentence which has to be examined here, for the first part is obviously true.
Ben E. Perry (Class. Phil. xl, 1945, 260) and R. Cantarella (Dioniso, x. 1947, 149 and 152)
also accept Turyn’s results, including his assumption that FG represent Triclinius’ first
edition.
! Interpolation of the article, as here and in front of δόμους in 436, is one of the best-
known characteristics of Triclinius’ editorship. The insertion is often prompted, as here,
by metrical reasons.
2 The right explanation of the changes which the lyrics as they stand in F have
undergone in Tr was given long ago by Petrus Victorius (p. 2 of his preface).
I2
 THE MANUSCRIPTS
that to him the colometry which he found in his exemplar was
invested with the authority of an authentic παράδοσις, and, moreover,
that the colometry of the hyparchetype of FTr is on the whole
faithfully preserved in F. It was equally well preserved in the
exemplar of Tr, and where Triclinius did not have a special reason
for altering it he left it as he found it. This can be seen, for instance,
from the beginning of the lyrical part of the parodos, 104 f., where
the colometry in FTr is different from that in MV. The same appears
e.g. at 1134.f. That passage in M is arranged thus:
πολυεπεῖς τέχναι
θεσπιωιδὸν φόβον
| φέρουσι μαθεῖν.
but in FTr thus:
moAveneis τέχναι θεσπιωδὸν
φόβον φέρουσι (-σιν Tr) μαθεῖν.
To return now to the rhythmical refrain 433 ff., we find that in F
the lines are arranged as follows:
ods μὲν γὰρ ἔπεμψεν οἶδεν"
ἀντὶ δὲ φωτῶν τεύχη
καὶ σποδὸς εἰς ἑκάστου
δόμους ἀφικνεῖται.

A comparison of this form of the lines with the form in which they
appear in Tr (see p. ı2) puts it beyond doubt that what we have in
F, so far from being a Triclinian product, is in fact the foundation
on which the metrician Triclinius erected the edifice of his conjectures.
Of the many passages by which the same point could be illustrated
I will for a special reason mention 1015-17 = 1030-4. These lines
appear in F and Tr in the following form:
1015 ff. in F: 1015 ff. in Tr:
πολλά τοι δόσις ἐκ λιὸς πολλά τοι δόσις διός
ἀμφιλαφής τε κἀξ ἀμφιλαφής τε καὶ
ἁλόκων ἐπετειᾶν ἐξ ἀλόκων ἐπετειᾶν,
νῆστιν ὥλεσεν νόσον. νῆστιν ὥλεσεν νόσον"
1030 ff. in F: 1030 ff. in Tr:
νῦν δ᾽ ὑπὸ σκότω βρέμει νῦν δ᾽ ὑπὸ σκότῳ βρέμει,
θυμαλγής τε καὶ οὐδὲν ἐλπομένα ποτὲ
ἐπελπομένα ποτὲ καίριον ἐκτολυπεύσειν,
καίριον ἐκτολυπεύσειν, ζωπυρουμένας φρενός.
ζωπυρουμένας φρενὸς.
The reason why Triclinius deleted ἐκ in 1015 and omitted θυμαλγής
τε καὶ οὐδὲν ἐπ after 1030 is one and the same: he wanted to achieve
13
 PROLEGOMENA
exact corresponsion between strophe and antistrophe. This was
not very easy, for in the strophe, either after 1004 or after 1005,
—uu-.0- had dropped out. Nevertheless, by means of a few
minor alterations and a bold cut after 1030, he succeeded in producing
two symmetrical stanzas of fifteen κῶλα each ; and having done this
he states, in his metrical scholion on 975, καὶ ἔστι τῆς μὲν πρώτης
στροφῆς τὰ κῶλα vy' . . . τῆς δὲ δευτέρας ιε΄ καὶ τὰ τῆς ἀντιστροφῆς
τοσαῦτα. This makes it clear that the omission in Tr of the eight
syllables after 1030 is due, not to an inadvertence as was assumed
by Blass (p. 18 f. of his edition of the Eum.) and Wilamowitz (on
Ag. 1031; he wrongly concluded that the arrangement of the lines
in the exemplar of Tr was different from that in F), but to a deliberate
act of harmonization on the part of Triclinius.
Compared with the lyrics, the trimeters did not provide Triclinius
with quite so many opportunities for alterations on metrical grounds,
but here too he found enough instances of faulty scansion to reward
his labours.! See e.g. the following lines: Ag. 1139 (οὐδέποτ᾽ FG:
οὐ δή ποτ᾽ Tr), 1231 (here he had τολμᾶ(ι) before him and consequently
took τοιάδε as neuter accusative), 1295, 1356, Eum. 306 (where the
absence in his exemplar of δ᾽ induced him to change ὕμνον... τόνδε
to ὕμνων... τῶνδε), and probably Ag. 539 (οὐκ ἀντερῶ F: οὐκέτ᾽
ἀντερῶ Tr; for the schol. vetus see the commentary). At Ag. 1652
he produced a correct trochaic tetrameter by writing πρόκοπτος
instead of the πρόκοπος which he found in his exemplar.
It would, however, be wrong to suppose that Triclinius confined
himself to correcting the metre. To show that this was not so it
will suffice to quote a small number of passages the text of which
(as preserved in the hyparchetype of FTr) was altered by him on
grounds other than metrical. At Ag. 184 ff. he missed (or did not
like) the anacoluthon, common though it is in Aeschylus, and con-
sequently in 187 changed συμπνέων to συμπνέει. At 231 he sub-
stituted ἐν ὄσσοις for ἀόζοις, which he probably did not understand.
At 304 he attempted to emend the hopeless μὴ χαρίζεσθαι by writing
δὴ instead of μὴ. At 455 he failed to see the point of ἔχοντας (and
perhaps mistook ἐχθρά for the neuter plural), and therefore wrote
ἐχθρῶς, with the gloss ἐχθρωδῶς διακειμένους περὶ αὐτήν. At 980 ff.
again (cf. above on 184 ff.) he thought the anacoluthon objectionable,
so he removed the apparent harshness by the conjecture ἀποπτύσαι,
explaining the construction in his paraphrase: οὐδὲ i£e ἐπὶ τὸν φίλον
θρόνον τῆς φρενός μου θάρσος εὐπιθὲς. . . ὥστε ἀποπτύσαι καὶ ἀπο-
βαλεῖν τοῦτο κτλ. At 1084 he altered the meaningless παρὲν to παρὸν.

1 Trimeters which did not scan until Triclinius corrected them existed even in the
much more intensely studied plays of the Byzantine triad, although, of course, con-
jectures on metrical grounds were made there at an earlier period too (cf. e.g. Pers. 326,
687, 782). In Ag. 340 the metrical conjecture ἄν γ᾽ ἑλόντες (FTr) is pre-Triclinian.

14
 THE MANUSCRIPTS
At 1611 he put ἰδόντι in the place of ἰδόντα and thus destroyed a fine
idiomatic peculiarity.
In all these instances, and in several similar ones, the readings
which gave rise to Triclinius’ conjectures are still preserved in F. It
is therefore quite wrong to call the text of F Triclinian or to regard
it with Turyn as a first immature edition by Triclinius, for a Tri-
clinius who would have let by all the things which, in striking con-
trast to the alterations in Tr, we find in F, would be no Triclinius
at all. I am not even prepared to confer on the text of F the doubtful
honour of calling it ‘proto-Triclinian’, since it is definitely pre-
Triclinian.
This can best be illustrated by a comparison of the primary
readings of F with its interlinear corrections. The following list con-
tains a selection of passages in which a word in F has been corrected,
by the superscription of a letter or two,! in such a way that the
corrected form agrees with the reading of Tr against the original
reading in F: Ag. 511 ἦλθ᾽ F, ἦλθες TrF?; 680 κλύων F, κλύειν TrF?;
718 f. οὗτος P, οὕτως TrF?; 745 πικροῦ F, πικρὰς TrF?; 920 βόαμα F,
βόημα TrF?; 957 δόμους F, δόμων TrF?; 974 μέλῃ F, μέλοι TrF?;
1252 παρεσκόπεις F, παρεσκόπης TrF?; 1477 γέννης F, γέννας TrF?;
Eum. 217 μόρσιμοι F, μόρσιμος TrF?; 218 ὅρκους τὶ F, ὅρκους re TrF?;
435 σεβόμεναι F, σέβοιμέν TrF? (this, at any rate, is what was intended
by F2, though actually the scribe, while writing ov above the o,

1 For details of the interlinear letters see my app. crit. and Blass’s app. crit. to the
Eumenides. I have not included in this list Ag. 1030 (βλέπει F: βρέμει TrF?) and 1267
(ἀμείβομαι F : ἀμείψομαι GTrF?), for I think it likely that in these two instances all that
happened is that the scribe of F made a slip and then corrected it, and that the correction
represents what he found in his exemplar; in the case of 1267 the reading of G is in
favour of this assumption (1030 belongs to the part lost in G). Whereas almost all
the ‘Triclinian’ corrections are made by writing fresh letters in a tidy script above the
letters written in the text, these two alterations look quite different: the letters of the
text have been transformed in a crude and ugly manner, so that (e.g.) the B is still
clearly distinguishable under the #. The same is true of the correction of ἔχουσαν to
ἔχουσ᾽ at 367; for this cf. Pasquali, ‘I codici inferiori della trilogia eschilea’, Rendic.
dell’ Accad. det Lincei, Ser. VI, vol. vi, 1930, 41: ‘Il tratto di penna che ha cancellato
av ἃ senza dubbio dello stesso inchiostro. Si ha l'impressione che F si sia accorto dell’
errore tra lo scrivere e abbia subito riparato. Ora una identica impressione si ha per
il βλέπει--βρέμει del v. 1030 [see the beginning of this note], tranne che qui rimane qualche
dubbio, perché l'inchiostro della correzione pare più scuro. This clumsy method of
changing one letter into another is often employed by the scribe of F in cases where
there can be no doubt that the form first written down was due to a mere slip or over-
sight ; e.g. at l. 19 of the ὑπόθεσις (ἐθιδάχθη corrected to ἐδιδ-) and in the text of the play
at e.g. 66 (κύμακος to xdp-), 108 (Bírpovov to δίθρ-), 230 (βραχεῖς to flpaBeis), 242 (mpove-
νεέπειν to προσεννέπειν), 622 (πὰς to πῶς), 716 (uéAawov to μέλεον), 746 (δύσερ- to δύσεδ-),
762 (καλλέππ- to καλλίπα-), 809 (πολιῶν to πολιτῶν), 881 (ἀμφίλεκκα to ἀμφίλεκτα; the
original letter is indistinguishable in the photostat), 1486 (πανεργάταν to -yérav).—It is
possible that this cruder method of transforming the letters themselves instead of
writing the correction above them was exceptionally used also for the purpose of
inserting a Triclinian reading. That seems to be the case at 1279, where the fact that G
as well as ΕἸ has ἄτιμόν makes it probable that this was the reading of the hyparchetype,
and that the change of the final v to «in F is due to the influence of a Triclinian MS.

I5
 PROLEGOMENA
forgot to delete the final αι); 567, where the verb πέλει, with which
Triclinius patched up the incomplete trimeter (εἴτ᾽ οὖν διάκτορος
πέλει τυρσηνικὴ Tr), is added in F in the right-hand margin; 674
τούσδ᾽ F, τάσδ᾽ TrF?.
Two of the passages of the Agamemnon mentioned in the preceding
list are quoted in the metrical scholia of F: 680 in the scholion on
489 (494 Wecklein), which ends ὧν τελευταῖος ' τοσαῦτ᾽ ἀκούσας ἴσθι
τἀληθῆ κλύειν ',! and 974 in the scholion on 810 (801 Wecklein), which
ends ὧν τελευταῖος ‘ μέλοι δέ σοι τοι τῶνπερ ἂν μέλλης τελεῖν ᾿. In other
words, we find in these F scholia the κλύειν and μέλοι of the text of
Tr as against the κλύων and μέλῃ of the text of F before its correction.
That is precisely what we should expect: just as the interlinear
corrections in F are derived from a text edited by Triclinius, so are
the metrical scholia.
It is the character of these metrical scholia in F that has been
chiefly responsible for the assumption that this manuscript is in all
respects Triclinian, and for Turyn's view that it is a copy of 'Tri-
clinius’ first recension’. Now there can certainly be no doubt that the
metrical system underlying these scholia is the system of Demetrius
Triclinius, and (more particularly) that the metrical scholia in F
are closely akin to the metrical scholia in Tr. But it seems to me no
less obvious that the metrical notes in F, so far from belonging to
an earlier recension, are in fact simply an abridged and simplified
version, or rather rearrangement, of Triclinius’ metrical commentary.
Any section could serve to illustrate this; I select, on account of its
comparative simplicity, the metrical analysis of the first stasimon
(355-488). Opposite 1. 355 Tr has the following scholion, marked with
the note ἡμέτερα and a cross: ὦ Ζεῦ βασιλεῦ: εἴρηται ἡμῖν ἐν τῷ τῶν
Περσῶν δράματι περὶ τοῦ εἴδους τούτων (τούτου Tr) τῶν χορῶν’ ὅμοιον
γάρ ἐστιν ἐκείνοις καὶ τοῦτο" ἔχει γὰρ ἐν ἀρχῇ μὲν σύστημα ἐπιφθεγματικὸν
εἰς δύο διῃρημένον περιόδους, κώλων ἀναπαιστικῶν iB", ἑξῆς δὲ στροφὰς τρεῖς
καὶ ἀντιστροφὰς τοσαύτας καὶ ἐπῳδόν: καλεῖται δὲ ταῦτα ἑπτὰς ἐπῳδική. ἐπὶ
ταῖς ἀποθέσεσι παράγραφος. ἐπὶ δε τῷ τέλει τῆς ἐπῳδοῦ κορωνὶς καὶ παρά-
ypados. Then, opposite 1. 367, and marked in the same way : Atds πλα-
γάν: ἐντεῦθεν ai κατὰ σχέσιν ἄρχονται στροφαΐ" καὶ εἰσὶ τῆς μὲν πρώτης
στροφῆς τὰ κῶλα ιη΄, καὶ τὰ τῆς ἀντιστροφῆς τοσαῦτα' τῆς δευτέρας ιζ΄.
καὶ τὰ τῆς ἀντιστροφῆς τοσαῦτα. τῆς τρίτης ὁμοίως ιζ΄. καὶ τὰ τῆς ἀντι-
στροφῆς τοσαῦτα: τῆς δὲ ἐπῳδοῦ ιε΄. ἃ καὶ μετρήσεις τοῖς προτέροις
ἑπόμενος. εἰσὶ γὰρ τὰ μὲν ἀντισπαστικά' τὰ δὲ τροχαϊκά' τὰ δὲ χοριαμβικά'
1 Vitelli-Wecklein, in their account of the F scholia (Wecklein, vol. i, p. 337), print
κλύων here ; but F has κλύειν, clearly written and with no trace of a correction.
2 This refers to Triclinius’ own scholion (ἡμέτερον) on Pers. 532, the second sentence
of which reads: ἔχει γὰρ πρῶτον ἐν τάξει προῳδοῦ σύστημα ἐπιφθεγματικὸν ὀνομαζόμενον
ἀνομοιομερές, κώλων ἀναπαιστικῶν ts’, εἶτα στροφὰς καὶ ἀντιστροφὰς γ΄. Cf. also Triclinius’
note on the parodos of the Ag. (40): τὰ τοιαῦτα εἴδη τῶν χορῶν, ὡς εἴρηται καὶ ἐν τῇ
ἀρχῇ τοῦ τῶν Περσῶν δράματος, ἑτέρως εἰσὶν ἐσχηματισμένα κτλ,
τό
 THE MANUSCRIPTS
τὰ δὲ ἰαμβικά. ἐφ᾽ ἑκάστης στροφῆς καὶ ἀντιστροφῆς παράγραφος. ἐπὶ
δὲ τῷ τέλει τῆς ἐπῳδοῦ κορωνὶς καὶ παράγραφος. The substance of
these two scholia is given in F in the following scholion (on 355):
ὦ Ζεῦ βασιλεῦ: 6 παρὼν χορὸς συνέστηκεν ἐξ κώλων ρλς΄. ὧν τὰ πρῶτα
ιβ΄, ἀναπαιστικὰ δίμετρα ἀκατάληκτα καὶ καταληκτικὰ ἤτοι ἐφθημιμερῆ:
καὶ povóuerpa:! τὰ δὲ ἑξῆς pxd’, χοριαμβικὰ δίμετρα ἀκατάληκτα καὶ
καταληκτικὰ ἤτοι ἐφθημιμερῆ καὶ πενθημιμερῆ καὶ ἡμιόλια' καὶ τρίμετρα
βραχυκατάληκτα καὶ καταληκτικά, ὧν τελευταῖον: ᾿Ιγυναικογήρυτον
ὅλυται (sic) κλέος ᾿. Without going into too many details or discussing
some minor alterations, it may be said that the scholion in F is
a typical example of the boiling-down of a fairly learned commentary
into the bare minimum of notes indispensable for the purposes of
a school edition. The information concerning the general structure
of the ode has been ruthlessiy cancelled, and along with it the
reference to the analogous structure of an ode in the Persae; more-
over, all avoidable technical terms have been cut out, and all that is
left is a very reduced description of the actual metres. In particular
there is no longer any mention of the colometrical σημεῖα such as
παράγραφος and κορωνίς ; this omission is characteristic of the metrical
scholia in F throughout.? Their retention there would have been
senseless: in Tr all the παράγραφοι, kopwvides, etc. referred to in the
metrical notes are to be found in the text column, but the editor
of F did not furnish his text with any such signs and consequently
could not refer to them in his metrical notes. This system? of carefully
differentiated signs, with reference regularly made to them in the
metrical scholia, is perhaps the most characteristic feature of
Triclinius’ editions of the dramatists. Other considerations apart,
1 καὶ povdperpa is left out by Vitelli-Wecklein (vol. i, p. 337).
2 T will give two more examples, from the first two metrical scholia on the Agamemnon.
At the end of the very first scholion (that which begins ἡ εἴσθεσις τοῦ παρόντος δράματος
[τοῦ δράματος Tr] Tr has ἐπὶ τῷ τέλει κορωνὶς εἰσιόντος τοῦ χοροῦ; at the end of the
scholion at the beginning of the anapaests (40) Tr has ἐπὶ ταῖς ἀποθέσεσι πάσαις παρά-
ypados. Neither of these remarks appears in F.
3 Triclinius took it over from Hephaestion epi σημείων; see the Triclinian scholion
on Ar. Plut. 253 (p. 338 Dübner), and cf. K. Holzinger in the Festschrift XAPIZTHPIA
Alois Rzach dargebracht (1930), 70 and 72.
* A general picture of these scholia can most readily be obtained from the following
sources: for the first three plays of Euripides (Hec., Or., Phoen.), the relevant sections
in Dindorf's unreliable edition of the scholia (vol. i. 206 ff. ; vol. ii. 10 ff.; vol. iii. 12 ff.);
for Sophocles, Turnebus’ pretty little volume bearing the title * Δημητρίου τοῦ Τρικλινίου,
Eis τὰ τοῦ Σοφοκλέους ἑπτὰ δράματα, Parisiis, MDLIII, apud Adrianum Turnebum typo-
graphum regium"; for Aristophanes, the scholia printed between double square brackets
(I D) in the old monograph by C. Thiemann, Heliodori colem. Aristoph. quantum
superest, Halle, 1869, or (for more accurate information) Bachmann's appendix to
Zacher's edition of the Peace (1909), where (pp. 1o9 ff.) a special column is allotted to
the metrical scholia of Triclinius as printed in the Aldina. The scholia in the codex
Farnesianus (Tr) of Aeschylus have been published by the following scholars: Prom.
by H. W. Smyth in Harvard Studies in Class. Philol. xxxii, 1921, 1 ff. ; Sept. by W. Din-
dorf (unsatisfactory), in Philol. xx, 1863, 385 ff. (the Thoman scholia), and xxi, 1864,
193 ff. (Triclinius' own scholia); Pers. by Lydia Massa Positano, in Collana di studi
4872-1 17 C
 PROLEGOMENA
their presence in Tr and their complete absence from F might make
us hesitate to accept Turyn's assumption (op. cit. 112 f.) that the
exemplar of FG ‘represented Triclinius’ first recension’ and that he
‘afterwards wrote down his second recension of the Aeschylean text in
the Naples codex, elaborating more exactly the metrical scholia’, etc.
The metrical scholia in F, compared with those in Tr, show not
only the omissions which we have just noticed but also the regular
addition of a certain item. In both the Tr and the F scholia the
metrical summary of a piece of dialogue is always followed by the
quotation of its last line; e.g. on Ag. ı they have ὧν τελευταῖος
* μαθοῦσιν αὐδῶ κοὐ μαθοῦσι λήθομαι ’, on 258 ὧν τελευταῖος ' χάρις γὰρ
οὐκ ἄτιμος εἴργασται (F, εἴργεσθαι Tr) πόνων ', etc. In F the same
practice is followed in the summaries of the lyrical sections; e.g.
on 40 ὧν τελευταῖον ' γαίας μονόφρουρον ἕρκος ', on 355 ὧν τελευτ(αϊον)
' γυναικογήρυτον ὅλυται [sic] κλέος ’, etc. But hardly ever! in the
colometrical scholia in Tr is there a similar reference to the con-
cluding line of lyrics. The same practice as in Tr is noticeable in the
Triclinian metrical scholia on Sophocles and Euripides; these too
always quote the last lines of the trimeter parts but never those
of the lyrics. The reason for this difference is obvious. No one who
glances over the pages of a Triclinian manuscript such as our Tr
can be in doubt for a moment as to the precise point at which a
canticum comes to its end: not only does the very conspicuous
κορωνίς attract the reader’s attention, but the very shape of the
slender columns of lyrics distinguishes them from the broader
columns of trimeters. But while Tr is wasteful of space, F is not:
where Tr gives each of two lyrical cola a separate line, F often has
them on the same line (though separated by a small blank). This has
happened, for example, at the end of the lyrical part of the parodos
(down to 257) and at the end of the first stasimon (down to 487);
so that here, quite apart from the absence of a kopwris, the column
of the canticum and that of the following dialogue look (at any rate
at first sight) very much alike. Therefore, to spare his reader trouble,
the editor of F added a note which in Tr would have been quite
unnecessary.
The colometrical signs such as παράγραφος and kopwvis are not
the only Triclinian σημεῖα that are conspicuously absent from F.
Triclinius, who in his Hephaestion (pp. 3 ff. Consbr.) found a chapter
περὶ κοινῆς (συλλαβῆς) which discussed among other things the
greci diretta da V. De Falco, xiii, Naples s.a. [1948] (I am very grateful to Signora Massa
Positano for sending me a copy of her edition); Ag. by Dindorf, Philol. xx, 1863, 16 ff.
(the παλαιά and the Triclinian scholia separate), and by van Heusde (more accurately),
in his edition (1864) of the play. For my own quotations from the Tr scholia on the Ag.
I have always consulted the photostats. The Tr scholia on the Eum. are edited in
A. Turyn's book The Manuscript Tradition ... of Aeschylus, 125 fl.
1 The remark in schol. Ag. 975, ὧν τελευταῖος * ζωπυρουμένας φρενός ’, is exceptional.
18
 THE MANUSCRIPTS
shortening of a diphthong in front of a vowel (τοιοῦτος =~ — -,
παλαιός =~~v, etc.) and the treatment of muta cum liquida,
thought it would be a good thing in the case of such prosodically
ambiguous syllables to use a special sign to show the reader whether
he was to scan τοιοῦτος as — — ~ Or as ~ — ~, ὅπλα as — ~ Or as ~ ~, etc.
So he added two new symbols to the traditional prosodic symbols,
accents, and breathing signs. He gives an account of this invention,
of which he is exceedingly proud, in the short treatise which he
incorporated in the introduction to his commentaries on Aristo-
phanes! (Proleg. xvii, p. 43f. Dindorf, p. xxxf. Dübner) and
Aeschylus (cf. H. W. Smyth, Harvard Studies in Class. Phil. xxxii,
1921, I, and Turyn, op. cit. 106) under the title περὶ σημείων τῆς
κοινῆς συλλαβῆς τῶν ἐντὸς κειμένων τῆς βίβλου. After saying that it
was he who introduced a sign for the κοινὴ συλλαβή ᾿ διὰ τὴν τῶν
πολλῶν πλάνην ’, he continues: ἐπενοήθη δὲ διπλοῦν τὸ σημεῖον, διὰ
τὸ διπλῆν τινα καὶ ταύτην (i.e. τὴν κοινὴν συλλαβήν) ἔχειν τὴν δύναμιν.
ὅτε μὲν οὖν ἀντὶ βραχείας ὀφείλει λαμβάνεσθαι, σημεῖον ἐπενοήθη τοδὲ -,
μακρὰ δηλονότι καταρχὰς ἄνω βλέπον τὸ τοῦ ἰῶτα στοιχείου σημεῖον
ἔχουσα, ὅτε δ᾽ ἀντὶ μακρᾶς, τοῦτο ἀντεστραμμένον οὑτωσὶ 7, μακρὰ
δηλονότι ἐν τῷ τέλει κάτω νεῦον τὸ τοῦ ἰῶτα σημεῖον ἔχουσα.2 In Tr these
signs are often to be seen above the vowel of a syllable whose quantity
is ambiguous. For example, on the plate facing p. 1 of H. W. Smyth’s
article (in Harvard Studies, xxxii) the sign — can be seen at line 5
(Prom. 795) above the v of κυκνόμορφοι, and at line 3 from the bottom
(Prom. 803) above the first a of ἀκραγεῖς, while at line 8 (Prom. 798)
has been written above the e of τῶνδε to indicate that it is not
lengthened by the vp of the following τρεῖς;? similarly plate II of
Wilamowitz’s edition, at line 5 (Pers. 949), shows - above the
second a of dpiSaxpuv. The sign + is also used to mark the shortening
of a final vowel in front of an initial vowel, e.g. above the ἡ of κακῶν
yàp δὴ αἱ at Ag. 1133; this is in keeping with the fact that Hephaestion
begins his discussion of the κοινὴ συλλαβή with instances of such
shortenings. There are no such signs in F, at any rate in the part
containing the Agamemnon (of which alone I have photostats).
Nevertheless, the editor of this text was not wholly insensitive to
the advantages of Triclinius’ innovation. In the case of muta cum
liquida he would let his readers fend for themselves as best they
could, and whereas Tr uses one sign to warn them that (e.g.) ἔθρισεν

τ For the MSS containing this treatise cf. Holzinger, in the article cited on p. 17 n. 3.
2 Cf. e.g. the Triclinian scholion on Ar. Plut. 14 published from a Paris MS of the
15th century (Coislinianus 192) by W. I. W. Koster, Scholia in Aristophanis Plutum et
Nubes (Leyden 1927), p. 2: Ἢ προσῆκ᾽ αὐτῷ πδὶιεῖν : κοινὴ συλλαβή ἐστι τὸ ποι, ὡς "Hoar-
στίων φησίν, ἀντὶ βραχείας λαμβανόμενον. Οὐ δεῖ οὖν ποεῖν γράφειν.
3 Conversely, at Ag. 492 (ἐφήλωσε [FTr] φρένας) Triclinius put the sign ^ above the
ending of the verb, thus implying a false prosody ; he did not take advantage of the
reading ἐφήλωσεν of his σχόλ. aA. which he copied in the margin.
I9
 PROLEGOMENA
(536) and diôpus (1105) have their first and second syllables respec-
tively long, and another to indicate that (e.g.) the rp of Τροίαι (529)
does not make position and πάτριον (1157) has its first syllable short,
F gives them no assistance either here or in any similar instance.
But it is different when it comes to the shortening of the first syllable
of τοιοῦτος and the like. Obviously the editor of F was too shrewd
a schoolmaster not to see that this might become a most unpleasant
stumbling-block and upset the scanning of the whole line. So some-
thing had to be done about it. But it was against his general principle
to use a technical symbol which he would have had to explain at
great length. Consequently at the six places in which the o of forms
of τοιοῦτος (315, 593, 1075, 1352, 1360) and of οἷον (1256) is in Tr
marked as short by the sign = written above it, F has the word
κοινή written above the o..! It appears, then, that of the large amount
of information on prosody which Triclinius had provided, only the
indispensable minimum found its way into F, and that in this MS
the notation by means of special signs is given up altogether, ob-
viously because it was thought unsuitable for an edition which laid
no claim to methods of technical learning.
Since the Triclinian sign = was correctly interpreted by the
scholar responsible for F in its final form, it follows that this man,
probably a pupil of Triclinius or at any rate brought up in the
tradition of his doctrine, did not merely rely on what he found in
Triclinius’ edition of this particular text but was able to supplement
it from his knowledge of his master’s system in general. This observa-
tion is corroborated by another feature of the metrical scholia in F,
namely the formula with which they introduce the description of
dialogue parts (in trimeters and in tetrameters, and including the
ἀμοιβαῖον between Cassandra and the Chorus). In F these descriptions
begin regularly εἴσθεσις διπλῆς ἀμοιβαίας or εἴσθεσις διπλῆς povo-
στροφικῆς or simply εἴσθεσις διπλῆς: see (following Wecklein’s
numeration) the scholia on 270, 494, 801, 1019, 1056, 1177, 1342, 1577,
1649, printed in Wecklein’s edition on pp. 337 ff. Tr, however, follows
a quite different practice; e.g. the metrical scholion on 258 (270
Wecklein) begins ai ἑξῆς αὗται συστηματικαὶ περίοδοι στίχων εἰσὶν
ἰαμβικῶν τριμέτρων ἀκαταλήκτων ol”, that on 489 (494 Wecklein)
begins a£ συστηματικαὶ αὗται περίοδοι στίχων εἰσὶν ἰαμβικῶν τριμέτρων
ἀκαταλήκτων, and so on, and the same or a very similar formula is
used for the colometric description of dialogue parts throughout the
manuscript, not only in the scholia on Ag. and Eum. but in those
on the Byzantine triad as well. The formula which we have found

1 It should be noticed that κοινὴ at 1360 is by a later hand than the other instances
of this gloss. The κοινὴ at 1075 (1059 Wecklein) has been overlooked by Vitelli-Wecklein,
vol. i, p. 338.— The one instance in this play (1663) of the οι in τοιοῦτος being shortened
in tetrameters is marked neither in Tr nor in F.
20
 THE MANUSCRIPTS
in F, εἴσθεσις διπλῆς xrA., is completely absent from the scholia in
Tr ;? nor does it seem to be found anywhere else in Triclinius’ com-
mentaries on tragedies. In Triclinian metrical scholia on Aristophanes,
however, it does occur; cf. e.g. the scholia on Clouds 314, 889, 961,
etc. (Heliodori colometriae, ed. Thiemann, pp. 34, 42f.), on Peace
124, 301, etc. (in Zacher-Bachmann’s edition, p. 110, right-hand
column), and again on Clouds 314 in the specimen from cod. Vat. 1294
published by Zacher, ‘Die Handschriften . . . der Aristophanes-
scholien’, Jahrb. f. class. Philol., Suppl. xvi, 1888, 316. On this late
use (or rather abuse) of διπλῆ and related technical terms, cf. Thie-
mann (op. cit. οὗ f.; cf. also 105), who points out that ‘usus ille,
quem apud scholiastas Byzantinos cognoscimus, nominandi totas
periodos signorum nominibus’ is quite alien to the practice of the
earlier commentators, who use the word διπλῆ only to denote τὸ
σημεῖον τῆς διπλῆς.2 We cannot say why the editor of F, when he
adapted the Triclinian colometrical notes to his own more elementary
requirements, preferred the formula εἴσθεσις διπλῆς to the expressions
which Triclinius himself used in his commentaries on tragedy, but
the fact that he did choose it shows that he was familiar with certain
variants in Triclinius’ terminology and could therefore do a little
more than copy him slavishly.
The simplification which is characteristic of F’s treatment of
questions of metre and prosody can be seen also in its practice of
packing into a single paragraph the whole of the colometrical in-
formation about a choral ode, however long and varied. An instance
of this practice can be seen in the scholia on the first stasimon quoted
above (p. 16f.): at 355 Tr, after describing the general arrangement
of the ode, discusses only the introductory anapaests, and reserves
the more detailed examination of the lyrics proper for the scholion
on 367; F, on the other hand, gives right at the beginning, at 355,
the whole of the information which the editor of this school edition
wishes to impart, and then has done with the many metrical
problems of the ode. Similarly, F includes the whole of its metrical
comment on the parodos (ll. 40-257) in the note on 40, whereas Tr
has separate pieces of metrical analysis in the scholia on 40, 104, and
160. The latter method is in keeping with Triclinius’ general prac-
tice ;? that of F is not.
1 Compare, for example, the Tr scholia on Sept. 181, 422, 486, 792 (Dindorf, Philol.
xxi, 1864, 200 ff.) and on Eum. 179, 276, 397 (Turyn, op. cit. 130 ff.) with the corresponding
scholia in F (see the eclectic publication by Dindorf in vol. iii of his Aeschylus, Oxford,
1851, pp. 513 ff. and 528 ff., which may suffice for our purpose). The former always begin
αἱ μονοστροφικαὶ αὗται περίοδοι and the like, the latter always εἴσθεσις διπλῆς κτλ. and
the like. .
2 In the Triclinian scholion on Plut. 253 quoted by Thiemann we find this remark
about a περίοδος of lines spoken by the actors : ὁ τοιοῦτος σχηματισμὸς καλεῖται διπλῆ.
3 Compared with what is found in Triclinius’ commentaries on Sophocles, Euripides,
Aristophanes, and the Byzantine triad of Aeschylus, his metrical scholia on Ag. and
21
 PROLEGOMENA
But enough of the metrical scholia in F. Great as is their importance
asa clue to the problem of the relation between F and Tr, they are
not our only clue, and it is unfortunate that Blass and Turyn based
their belief in the Triclinian character of F entirely on the evidence
of its metrical scholia, to the exclusion of its text and of its non-
metrical scholia.! Of the latter something must now be said.? In
number they are not many: whereas the metrical scholia in F extend
over the whole of the Agamemnon (they begin at 1 and end at 1649
with the colometry of the trochaic tetrameters), the non-metrical
. marginal scholia (as distinct from interlinear glosses) cover only the
prologue and the parodos (the last is that on 252 [264 Wecklein],
σύναπτε τὸ προχαιρέτω eis τὸ προκλύειν), and only on the prologue are
they at all full; on the parodos all that we have (apart from mere
glosses) is a few pieces of paraphrase (on 58, 104, and 124 [128 Weck-
lein]), and no explanatory comment at all. Nevertheless, despite
their paucity, these non-metrical scholia provide important evidence
for the relationship between F and the other MSS.
The F scholia taken as a whole (metrical and non-metrical together)
fall into two groups: those which are marked with a cross before the
first word, and those which are not.? The notation is on the whole
executed with great accuracy : not one of the very numerous metrical
scholia lacks its cross (this, if nothing else, would leave us in no
doubt about their Triclinian origin). Now we also find the cross
affixed to two scholia which in Tr are marked as παλαιά, viz. that
on 2, ἀστείως εἴρηται τὸ κοιμώμενος κτλ. (for Tr see plate I, for F see
plate II), and that on 21/22, δεῖ διαστήματος ὀλίγου κτλ. The former
case is simple: the wording of the F scholion differs only in negligible
details* from that of the σχόλ. παλ. in Tr, from which it is obviously
derived. But the F scholion on 21/22 cannot derive from Tr, for the
stage-direction with which it begins (δεῖ διαστήματος ὀλίγου... τὸν
πυρσόν [cf. vol. ii, p. 15, on 1. 217) is followed by these words: τὸ δὲ
νυκτὸς ἀντὶ τοῦ ἐκ νυκτὸς ἡμέραν ἡμῖν διδούς. This gloss is almost

Eum. give the impression of being rather sketchy. By the time he had reached these two
plays he was probably tired of his job. Cf. also the many back-references such as ἃ καὶ
μετρήσεις τοῖς προτέροις ἑπόμενος (on Ag. 160), οἷα πολλάκις εἴρηται (on Eum. 244), etc.
! From Wilamowitz’s description (p. xix of his edition) we might infer that F con-
tained metrical scholia only.
2 My observations are of necessity confined to the Agamemnon, since I have neither
photostats nor collations of the F scholia on the other plays. An accurate description
of their contents would be welcome. (At Florence in September 1948 I could but glance
through F.)
3 This applies only to the marginal scholia; none of the interlinear glosses has a
cross attached to it.
4 The first sentence runs in Tr Aoreiwos εἴρηται τὸ κοιμώμενος ἐπὶ μὴ ὑπνοῦντος, but
in F ἀστείως εἴρ. τ. x. οὐκ ἐπὶ ὕπνου. In the next sentence Tr has φρουροίη and τις ὑπνῶν,
F φρουροῖ and ὁ ὑπνῶν. Then Tr has ἀλλ᾽ ἐπὶ ἀνακλίσεως ἁπλῶς εἴρηται, F ἀλλ᾽ ἐπὶ ἁπλῆ
(sic) ἀνακλίσεως. The last sentence, iv’ (Tr; ἵνα F) εἴη ὁ φύλαξ... νυκτερινῇ, is identical
in Tr and F.
22
 THE MANUSCRIPTS
exactly the same as that in M, ἐκ νυκτὸς ἡμέραν ἡμῖν δοῦσιν (this ought
to be διδούς), whereas the version in the σχόλ. παλ. of Tr is entirely
different: ὁ διδοὺς ἡμῖν φωταυγῆ ἡμέραν. Nevertheless we find in F
the cross in front of this scholion. What presumably happened is this.
When the scribe of F ran his eye over the margins of the Triclinian
copy, he noticed the identity of the stage-direction in Tr's oxdA. aA.
(δεῖ διαστήματος ὀλίγου... τὸν πυρσόν) with that in his own copy,
where he had entered it from his primary exemplar before adding
the Triclinian scholia. He therefore appended the cross to this
scholion, regardless of the discrepancy in its last sentence. It appears,
then, that the editor of F denoted by a cross all the notes that he
took over from the commentary of Triclinius, whether they belonged
to the older scholia which Triclinius called παλαιά and distinguished
by the way in which he wrote their beginnings (see above, p. 3 f.),
or whether, like all the metrical scholia, they formed part of Tri-
clinius’ own additions (ἡμέτερα). In this detail too the editor of F,
whom we have seen to be a follower of Triclinius’ doctrine, employed
that scholar’s methods: just as Triclinius affixed a cross to the
scholia (the ἡμέτερα) which he added to those provided by his
exemplar (or one of his exemplars), so also the editor of F used the
cross to mark the difference between his additional matter (in this
case the excerpts from Triclinius’ commentary) and the scholia
which he found in his own primary exemplar. It is a lucky accident
that for the beginning of the Agamemnon, i.e. the only part of the
play for which the F scholia are not too meagre, we also possess
the evidence of M and its scholia; for this enables us to prove the
reliability of the differentiation in F between Triclinian and non-
Triclinian scholia by demonstrating that certain scholia which F
does not mark with a cross, and which are absent from Tr, are found
in M. These scholia are the following (for F cf. plate IT): on 2, τῶν
κατὰ τὸ μῆκος τῆς ἐτείας φρουρᾶς. ἐπὶ μῆκος δὲ ἦν κοιμώμενος (so also
in M, except that there the last clause, written as a separate scholion
in the other margin, has the form ἦν ἐπὶ μῆκος κοιμώμενος) ; on 6
(preceded by a sign corresponding to that above δυνάστας), τοὺς
δυναμένους παρὰ τὰ ἄλλα σημᾶναι τοὺς καιρούς (exactly the same in
M) ; and on 11 (again preceded by a sign corresponding to that above
ἀνδρόβουλον), τὸ μείζονα 7) κατὰ γυναῖκα βουλευόμενον" γενναῖον" ἢ κατὰ
ἄνδρα βουλευόμενον (so also in Μ, except that there the last clause
* Turyn, op. cit. τοῦ, speaking of the part of F which contains the Agamemnon,
asserts that there ‘the Triclinian scholia, designated as such in T by a cross t and the
word ἡμέτερον, or only by a cross, appear in F with a cross ; the scholia παλαιά, designated
in Agam. and Eum. in T by a general title at the beginning of each of both plays and
differentiated by a capital initial letter . . ., appear in F without any special qualifica-
tion’ ; this assertion is based on an error.—It is to be regretted that in Vitelli-Wecklein’s
edition of the F scholia on the Agamemnon the important differentiation by means of
the cross has not been taken into account.
23
 PROLEGOMENA
reads ἢ κατὰ ἀνδρὸς βουλευομένης, which is obviously the genuine
version). It would be wrong to exclude from this survey some of
those shorter scholia (or excerpts from scholia) which in Wecklein’s
edition are marked as glosses by the note ‘(GI.)’.2 Some of these
also are found in M in more or less the same form as in F, but are
absent from Tr, viz. 3 ἀνέκαθεν ἄνωθεν" ἐξ ἀρχῆς, 32 οἰκειώσομαι, 56 ὃ
Opetos.
The fact that in all these instances the scholia and glosses in F
are closely related to those in M and have nothing corresponding to
them in Tr must not tempt us to assume that the basic material of
the F scholia is derived from M. On the contrary, some of the non-
Triclinian scholia in F provide excellent confirmation of the view
expounded above (pp. 7 ff.) that the hyparchetype of FTr was inde-
pendent of M. The scholion on Ag. 33 appears in F, M, and Tr in
the following forms: in F (in the margin, not marked with a cross,
i.e. not derived from the Triclinian stock) παροιμία τὸ τρὶς ἐξ βάλλειν
ἐπὶ τῶν ἄκρως εὐτυχούντων; in M καὶ παροιμία ‘ ἀεὶ yap εὖ πίπτουσιν
οἱ Διὸς κύβοι; in Tr (σχόλ. παλ.) Παροιμία τὸ τρὶς ἕξ βάλλειν ἐπὶ τῶν
ἄκρως εὐτυχούντων: ἐπεὶ καὶ τοῖς κυβευταῖς οὐκ ἔστι πλείω τούτων
εὐθυβολῆσαι. λέγουσι δὲ ταύτην καὶ ἑτέρως" ‘det γὰρ εὖ πίπτουσιν οἱ
Διὸς κύβοι ’. It is obvious that the oldest form of the scholion (perhaps
with a few slight alterations) is preserved in the παλ. of Tr, and that
what we have in F and M is two different excerpts, that in M of a
particularly unfortunate kind. A similar relation between F, M, and
Tr (παλ.) may be observed in the case of the scholion on 36: F has
(in the margin, with no cross, but preceded by a sign corresponding
to a sign in the text above βοῦς) παροιμία ἐστὶν ἐπὶ τῶν ἀλάλων ἢ
! All that Tr has here is τὸ κατὰ ἄνδρα βουλευόμενον, written as an interlinear gloss
above ἀνδρόβουλον. It appears that the hyparchetype of FTr contained the full scholion,
with the last clause already corrupted (partly because of the preceding xara γυναῖκα)
from κατὰ ἀνδρὸς βουλευομένης to κατὰ ἄνδρα βουλευόμενον, and that F copied the scholion
in its entirety while Tr was content with its last clause.
2 This distinction between ‘glosses’ and other scholia is (at any rate in many instances)
inevitably arbitrary. For example, at 3 M has the two notes πρὸς τὸ φυλακτικὸν κτλ, and
τὸ δὲ ἄγκαθεν κτλ. written in the margin, and Wecklein consequently prints them (p. 255)
as ordinary scholia; on p. 335 he marks the corresponding notes in F as ‘Gl.’ because
they are there written above the words of the text (interlinear). But this differentiation
is not consistently maintained in Wecklein's notes. On p. 335 he adds ‘Gl.’, e.g., to the
interlinear glosses at Ag. 8 (above φυλάσσω) μηνύω and at Io (above βάξιν) φήμην, but
not to those at 11 (above éAmifov) ἐλπέδα ἔχον and at 16 (above puvdpecBat) λέγειν kAatew.—
The choice between the marginal and the interlinear position depends often not on
the nature of the note but on the space at the disposal of the copyist.—It is especially
unfortunate that Dindorf, in his publication of Triclinius’ commentary on the Ag.
(Philol. xx, 1863, 30 ff.), should have left out altogether all interlinear scholia and
glosses without exception. This results in the absurdity that a scholion like that on 104,
ἤγουν δυνατός εἶμι εἰπεῖν τὸ συμβὰν αὐτοῖς σημεῖον ἐν ὁδῷ ἐξιοῦσιν, ἀγαθόν, which is sub-
stantially pre-Triclinian and is important because it shows that the F scholion here is
not derived from Tr (see p. 26), cannot be found in Dindorf’s article at all. In such cases
(and they are many) one has to have recourse to a photograph or to van Heusde’s badly
arranged edition.
24
 THE MANUSCRIPTS
ἀντὶ τοῦ βάρος ἐπίκειται 7 φοβοῦμαι ζημίαν ἐπικεισομέμην μοι; M has
ἢ βάρος ἐπίκειται" ἢ φοβοῦμαι ζημίαν ἐπικεισομένην μοι; Tr (σχόλ. παλ.)
has παροιμία ἐστὲ τὸ βοῦν ἐπὶ γλώττης φέρει, ἐπὶ τῶν μὴ λαλούντων
διά τινα αἰτίαν. λέγει οὖν καὶ οὗτος ἢ ἀντὶ τοῦ βάρος μοι ἐπίκειται, ἣ
φοβοῦμαι ζημίαν ἐπικεισομένην μοι. Here too the form in Tr comes
nearest to the original form of the scholion ; the omissions in F are
this time less considerable than in the scholion on 33, but in M only
the tail-end is preserved. The same kind of relation between the
Tr scholia (the fullest version and that which comes nearest to the
original) and those in M and F (excerpts) can be noticed at 49, where
the σχόλιον παλαιόν which Triclinius had before him, although it is
not copied by him verbatim, can, up to a point, be recovered from
his own scholion. F has, without a cross, ἐκπατίοις λεχέων ἀντὶ τοῦ
ἔξω τῆς αὐτῶν οἰκίας, M has in the right-hand margin (as a gloss to
ἐκπατίοις) τοῖς ἔξω τῆς ὁδοῦ, and in the left-hand margin the para-
phrase οἵτινες ὕπατοι ὄντες, ὀλέσαντες δὲ τὸν πόνον τῶν ὀρταλίχων, τὸν
ἐν τοῖς δεμνίοις τηρούμενον, ἐπὶ τῶν λεχέων στροφοδινοῦνται, and Tr
has above ἐκπατίοις the interlinear gloss τοῖς ἔξω τῆς ὁδοῦ, and in
the margin this scholion (Triclinius’ own, marked by a cross):
ἐκπατίοις- τοῖς ἔξω τῆς ὁδοῦ. πάτος γὰρ ἡ ὅδός. ἢ ἐκπατίοις λεχέων,
ἀντὶ τοῦ τοῖς ἔξω τῆς οἰκίας αὐτῶν. τὸ δὲ ὅλον" οἵτινες ἀπολέσαντες τὸν
πόνον τῶν ὀρταλίχων τῶν ἐν τοῖς δεμνίοις τηρουμένων" οὕτω γὰρ ὦφελεν
εἰπεῖν, στροφοδινοῦνται ἐν ἄλγεσι τῶν παίδων ἐκπατίοις (the rest of the
scholion does not concern us here). It appears that the scholion which
Triclinius used and slightly modified provided two alternative inter-
pretations of ἐκπατίοις : (I) τοῖς ἔξω τῆς ὁδοῦ (in agreement with the
gloss in M and Tr), and (2), taking ἐκπατίοις λεχέων together, ἀντὶ
τοῦ τοῖς ἔξω τῆς οἰκίας αὐτῶν (in agreement with the F scholion).
Moreover, it is obvious that the paraphrase of δεμνιοτήρη πόνον
ὀρταλίχων ὀλέσαντες, viz. ὀλέσαντες τὸν πόνον τῶν ὀρταλίχων, τὸν Ev
τοῖς δεμνίοις τηρούμενον (M), belongs to the old stock and had its
place also in the σχόλια παλαιά used by Triclinius ; it was he himself
who changed τὸν... τηρούμενον to rÀv . . . rnpovuevwv, with the
comment οὕτω yap ὥφελεν εἰπεῖν. This correction and possibly a few
other minor alterations probably account for his listing this scholion
with the ἡμέτερα. That the scholia which the editor of F found in his
exemplar (we are not dealing here with those which he added from
the commentary of Triclinius) were not derived from M may be
seen also from 58, where the full paraphrase found both in F (mar-
ginal, without a cross) and in Tr (σχόλ. παλ.), viz. ὑπὲρ τῶν μετοικι-
σθέντων νεοσσῶν πέμπει τοῖς παραβᾶσι καὶ μετοικίσασιν αὐτοὺς ὕστερον
τιμωρίαν, is not likely to be an expansion of the brief note in Μ, ὑπὲρ
τῶν μετοικισθέντων νεοσσῶν. |
1 Here I have kept the punctuation of Tr. |
2 In the four instances mentioned in this paragraph, what has happened is that only
25
 PROLEGOMENA
After this digression we return again to the problem of the two
elements of which the scholia in F are composed, viz. the pre-
Triclinian groundwork and the material derived from Triclinius’
edition. By examining certain scholia found in F and M but not in
Tr, we have been able to demonstrate that in any marginal scholion
in F the absence of a cross must be regarded as an indication that the
scholion in question is not derived from the commentary of Tri-
clinius. This conclusion is confirmed by the only two marginal
scholia in F with no cross added to them which we have not yet
examined. One is the scholion on 104, δυνατός εἰμι εἰπεῖν τὸ συμβὰν
αὐτοῖς σημεῖον ἀξιοῦσι (it ought to be ἐξιοῦσι)" τὸ δὲ ὅδιον ἀντὶ τοῦ
ἐνδοξότατον ; i.e. a piece of paraphrase followed by an interpretation
of ὅδιον. M has the paraphrase (8vvarós . . . ἐξιοῦσι) in exactly the
same form, though the interpretation of ὅδιον is different. In Tr,
however, although the paraphrase is substantially the same, an
interpretation of ὅδιον has intruded into the middle of it: ἤγουν
δυνατός εἰμι εἰπεῖν τὸ συμβὰν αὐτοῖς σημεῖον ἐν ὁδῷ ἐξιοῦσιν, ἀγαθόν.
In other words, the hyparchetype οἱ FTr had the paraphrase in
exactly the same form as M ; the editor of F copied it as he found it,
but Triclinius introduced a slight change. The other scholion is that
on 252 (264 Wecklein), where F has (in the margin) ovvarre! τὸ
προχαιρέτω εἰς TO mpoxAvew; this cannot possibly derive from Tri-
clinius, since the words τὸ δὲ προκλύειν (which have nothing corre-
sponding to them in the strophe) are omitted by him from his text.
Now at last we are able to sum up our various observations and
form a fairly accurate idea of what happened. The editor of F used
as his main exemplar a pre-Triclinian or at any rate non-Triclinian
MS. This MS, for the part containing the Agamemnon, was furnished
with excellent scholia, at least at the beginning. Whether the rapid

a fragment of the original explanation or paraphrase (in the first two instances the tail
end) has been preserved in M. We find a similar relation between the σχόλ. wad. of Tr
and the M scholia in the following examples, which were adduced by Heimsoeth, ‘De
scholiis in Aesch. Ag.’, vif.: at Ag. 260 (272 Wecklein) the exóA. aA. in Tr reads ἤγουν
ἐρήμου καταλειφθέντος τοῦ θρόνου τοῦ βασιλέως, ws ἀποδημοῦντος ἐκείνου. παρόντος μέντοι
οὐ δεῖ ἐντυγχάνειν αὐτῇ πάντα καὶ τὸν τυχόντα, but in M all that is left is παρόντος μέντοι
οὐ δεῖ συντυγχάνειν αὐτῆι; at 1133 (1125 Wecklein) the σχόλ. wad. reads Διὰ τὰ προσόντα
κακὰ τοῖς eis μαντείαν εἰσερχομένοις καὶ τὰ λεγόμενα φοβερά ἐστι, but Μ' omits the words
εἰς μαντείαν, without which the verb εἰσερχομένοις does not make sense. Instances like
these are sufficient to show that Heimsoeth was right in rejecting Dindorf’s view that
all the scholia in the later MSS were derived from M. Unmistakably, in Pasquali’s
words, ‘scoli di M non sono spesso intelligibili, perchè ritagliati arbitrariamente da un
contesto maggiore’ (op. cit. [on p. 15 n. x], p. 38 n. 1). The point has been elaborated
with regard to the scholia on Prom., Sept., and Pers. by Wilamowitz, Hermes xxv, 1890,
161 ff., who produced some very striking examples of genuine ancient matter in the
scholia of the later MSS; cf. also Pasquali, Storia della tradizione (1934), 27 f.
ı K. Zacher, Die Handschriften . . . der Aristophanesscholien, 618, observes that
phrases like σύναπτε πρὸς κτλ. are much more common in those Aeschylus scholia which
go back to Thomas Magister than in those added by Triclinius. For another σύναπτε-
scholion (in Tr) see vol. ili, p. 561.
26
 THE MANUSCRIPTS
decrease of real scholia after the prologue was a feature of this
exemplar' or is due to the dwindling zeal of the copyist we cannot
determine with certainty ; the latter, however, seems far more likely
when we consider that the man had an ample supply of Triclinian
scholia (both παλαιά and others) at his disposal throughout the
Agamemnon and yet, with the exception of the colometrical notes,
hardly copied them at all after he had reached the end of the ana-
paests of the parodos.? He evidently thought that for the purposes
which his edition was to serve short interlinear notes were quite
sufficient as far as the understanding of the text was concerned. It
was different, however, with the metre. No section of the Agamemnon
is in F left without a colometrical comment. It is easy to guess the
reason for this persistence, which forms such a striking contrast
to the indolent treatment of the non-metrical notes. The activities
of Triclinius and his followers seem to have had the result that in
the practice of the schools some kind of colometrical illustration,
though in a simplified form, was regarded as necessary. It was mainly
to satisfy this demand that the editor of F added to the copy of his
primary exemplar considerable excerpts from the commentary of
Triclinius. In doing so he also incorporated, to begin with, a certain
amount of non-metrical Triclinian matter, but, as we have seen, he
soon gave it up except for corrections of the text and short glosses.
It was only after copying the text and scholia of his first exemplar
that he made use of Triclinius’ edition ; the proof of this lies in the
arrangement of the scholion on 2 (one of Triclinius’ παλαιά), ἀστείως
εἴρηται κτλ., lines 3-5 of which have been pushed aside from the
normal left-hand edge of the scholia column because the earlier
scholion διὰ τὸ φυλακτικὸν κτλ. was in their way (see plate II).?
The pre-Triclinian text of F and the Triclinian colometry in its
margin do not always live together in perfect harmony. There is
at least one place where we can observe a clash between the two
heterogeneous elements. The F(G) scholion on Ag. 1537 ff. (1539 ff.
Wecklein), duly marked by a cross in F, runs ἀναπαιστικὰ κῶλα +”,
ὧν τὸ a’, τὸ γ΄, τὸ ς΄ καὶ τὸ 8’ μονόμετρα," τὰ δὲ λοιπὰ δίμετρα ἀκατά-

! This was the view of Blass (Die Eum. des Aisch., p. 20: ‘an Scholien hat das Original
von f g h [FGTr] nur wenig gehabt’); but he had not noticed the indications in the
F scholia which enable us to form a more precise idea of what lies behind them.
2 As regards the Eumenides, F contains a few marginal glosses and the Triclinian
metrical scholia, but no other scholia. |
3 It can also be observed elsewhere in F, e.g. fol. 28 recto (beginning of the Septem),
fol. 30 recto, fol. 32 recto, that the beginnings of some of the lines of the Triclinian
scholia in the margin are not, as is the rule, arranged in a perpendicular line, each
beginning strictly beneath the first letter of the preceding one, but recede to the right-
hand side because some interlinear scholia or glosses protrude into the margin. In such
cases the late-comers, viz. the Triclinian scholia, had to squeeze themselves into what
space was left. ΝΣ . .
* This description is quite accurate, since the first, third, sixth, and ninth κῶλα, viz.
27
 PROLEGOMENA
Ankra, τὸ δὲ ı' ἐφθημιμερές. The statement τὰ δὲ λοιπὰ δίμετρα ἀκα-
τάληκτα is perfectly correct as far as Tr is concerned, for in that MS
the fourth colon is disfigured by a horrible Triclinian interpolation
and appears as Ópoíras νῦν κατέχοντα χαμεύναν, i.e. an acatalectic
dimeter; but the original text, as preserved in F, has δροίτας kare-
xovra χαμεύναν, 1.6. a ἐφθημιμερές, in flagrant contradiction to the
colometrical description. We are fully justified in insisting on this
detail, as the statements of the colometries on points like this are
very accurate: the Byzantine metrician, when describing an ana-
paestic system that contains more than one (i.e. the concluding)
paroemiac, will not say (as he does at 1537 ff.) ra δὲ λοιπὰ δίμετρα
ἀκατάληκτα, but will speak of a certain number of ἀναπαιστικὰ
δίμετρα ἀκατάληκτα καὶ καταληκτικά (cf. e.g. schol. F on Ag. 355
[367 Wecklein] and 681 [686 Wecklein]).
Of many of the interlinear scholia and glosses in F we cannot say -
whether they come from the pre-Triclinian stock of the hyparchetype
of FTr or are additions derived from the commentary of Triclinius.
It is certain, however, that both types are to be found in F. There is
a clear instance of a Triclinian gloss, e.g., at 512 (517 Wecklein) : in F
the gloss ἀπόμαχος, written as it is above καὶ παγώνιος, does not make
sense, but in Tr, where the reading of the text is κἀπαγώνιος, it is
perfectly intelligible. On the other hand, there is a gloss of a quite
different type at 513, where over dywvious θεοὺς F has τοὺς ἅμα Evi
τόπῳ ἱδρυμένους (Tr has instead the commonplace rendering τοὺς
ἐφόρους τῶν ἀγώνων) ; this remarkable interpretation is probably
correct (see the commentary), and in any case looks like an excerpt
from some very good lexicographical authority.’ Again, the short
paraphrase of 1365 in F (quasi-interlinear, i.e. written on the narrow
inner margin, beginning between 1364 and 1365, and not above 1365,
because the space there was already filled by the gloss ὡριμοτέρα,
γλυκυτέρα), ἡδύτερος, φησίν, 6 θάνατος τῆς τυραννίδος, possibly repre-
sents the original version, while the less sensible version in the σχόλ.
maA. of Tr, ‘Hdvrepds ἐστι, φησίν, 6 θάνατος 6 ὑπὲρ τῶν δεσποτῶν τῆς
τυραννίδος, may be the result of an enlargement.
Our close examination of F has led us to the conclusion which was
briefly formulated by Wilamowitz (p. xix f.), viz. that the text of
this MS is free from Triclinian elements, although these are to be
found in its corrections and its scholia and glosses. By this statement
the assumption that the hyparchetype of F(G)Tr went back to
Triclinius or was influenced by his edition is implicitly rejected.
I now come to an important point, which does not seem to have
(1) ἰὼ γᾶ γᾶ, (3) äpyuporoixov, (6) ἢ σὺ τόδ᾽ ἔρξαι, (9) χάριν ἀντ᾽ ἔργων, are separated from
the rest and written as monometers not only in Tr but also (with less spacing, but
clearly enough) in F.
τ For the use of the word τόπος in the interpretation of ἀγών cf. Schol. Hom. Z 376 =
Etym. gen. B = Etym. M. 15. 47.
28
 THE MANUSCRIPTS
received sufficient attention :' the hyparchetype of FTr had variants
in the margin, marked by the usual γρ(άφεται). They cannot be
expected to have survived in this form in Tr, for H. W. Smyth’s
remark (Harvard Studies in Class. Philol. xxxii, 1921, 83) about the
Tr scholia on the Prometheus seems to be true of these scholia in
general: ‘statements about various readings are contained in the
body of the scholia and are never indicated by an interlinear or
marginal yp.’ F, however, still preserves sufficient instances of such
marginal variants indicated by yp.? One occurs at Ag. 45, where in
the margin of F we find, exactly as in the margin of M, yp. ἴλιον
αὐτάν͵,3 another, an abortive one, at Ag. 3, where, above the word
φυλακτικόν of the interlinear gloss διὰ τὸ φυλακτικὸν «rÀ., one can
clearly read yp. a (cf. plate II) ; the smirch on this yp. a shows that
the scribe tried afterwards to wipe it out ; what he intended to write
before he stopped was obviously yp. ἀνέκαθεν. Both these variants
proceed from the hand of the scribe of F himself, as do certain
marginal variants in the Byzantine triad, e.g. on Prom. 56 (there the
reading of F is πασσάλεν᾽ ἐρρωμένως," yp. πρὸς πέτραις, and on Sept. 16,
yp. βροτῶν."
It is « priori the most probable assumption that these variants
marked by yp. which we find in F are a faithful reproduction of
corresponding features in the hyparchetype of FTr, which in this
respect will in principle not have been different from M. A valuable
confirmation is provided by a few passages where F does not contain
! Mazon, vol. ii, 2nd ed., p. xxiv, considers the possibility of variants in the hyparche-
type of FTr, but does not follow up his surmise.
2 In the part of V containing the Ag., since the text is written in three columns to
a page and had to be written very closely in order to be got in at all, there is no room
to spare for any marginal notes.
* Not mentioned by Wecklein.
* Not mentioned in our editions. |
5 It is important to distinguish these early entries from certain others of an apparently
similar type. In the Eumenides we read in the margin of F the following four notes:
at 222 (οῦτοι F) yp. olôa, at 245 (μηνυτῆ, with a blank after it, F) yp. μηνυτῆρος, at 299
(οὔτις σ᾽ ΕἾ yp. οὔτοι σέ, and at 476 (οὔκουν εὔπεπλον F) yp. οὐκ εὐπέμπελον. A cursory
examination of F in September 1948 showed me that these four marginal notes, all of
which are mentioned in the editions of Blass and Wilamowitz, are entirely different
from the neatly written variants (yp. Üuov αὐτὰν etc.) of which I have just spoken:
they are in an ugly late hand. The inference is that they were written after the com-
pletion of the MS; this is confirmed by the fact that the note yp. μηνυτῆρος (fol. 112
verso, l. 3 from the top) stands a good deal below the line (245) to which it belongs. The
reason for this displacement is clear: the whole upper part of the margin was already
occupied by the Triclinian metrical scholion (it has been shown above, p. 27, that the
incorporation of the Triclinian scholia marks the latest stage in the work of the scribe
of F). The readings of these four notes in the margin of F agree with the text
of M (or m) against the text of F. I regard it as practically certain that the Renaissance
scholar who made these entries had access to M or a copy of it. These four ‘variants’,
then, have no claim to be mentioned in the apparatus of a modern edition. The same
late hand can be recognized in the text column of Eum. 299, where we find underneath
the o« of ἀθηναίοις three dots and above it the correction as (again in agreement with
the reading of M).

29
 PROLEGOMENA
an interlinear or marginal variant and where nevertheless it can be
inferred with certainty that such a variant existed in the hyparche-
type. Ag. 215 is of particular interest. There can be no doubt that
ὀργᾶι (MVF) is the genuine reading. Now when the curious αὐδά(ι),
. which we find, with yp. before it, in the margin of M, appears in the
text of Tr,! this can only mean that the hyparchetype of FTr had,
like M, ópyà: in the text and adda in the margin, and that the editor
of F (or his exemplar) chose the former and Triclinius (or his
exemplar) the latter. Then there is the discrepancy at Ag. 1152,
where the original reading of FTr agrees with the original reading
of M (see above, p. 8) and their corrected reading with the correc-
tion in M ; in this case it is not quite certain but highly probable that
the hyparchetype had the variant, though not necessarily as a
marginal note with yp. Again at Ag. 64, where FTr concur in the
false reading épeuropévov, the correction do written above the word
in F seems to point to a variant in the hyparchetype. The same is
probably true of Ag. 197, where F has πολυμήκη with the correction
παλιμμ written above: that πολυμήκη is not due to an individual error
of this MS is clear from the fact that it is the reading of V as well;
on the other hand, it is unlikely that Triclinius, who reads παλιμ-
μήκη, should have hit upon this word (a ἅπαξ λεγόμενον) by conjecture
if his exemplar had had only πολυμήκη. Everything becomes simple
if we assume that the hyparchetype of FTr had πολυμήκη in the
text and παλιμμήκη as a variant: F reproduces its exemplar faith-
fully, while Triclinius decides at once in favour of the reading which
is metrically satisfactory.
These considerations perhaps have a bearing on one of the most
awkward problems of textual criticism in the Agamemnon. In the
commentary on 1041 it will be shown that there is just the possibility
(nothing more) of accounting for the curious discrepancy between
the readings of F and Tr by assuming that it is the result of a variant
which the hyparchetype of FTr preserved in addition to the reading
which we find in F.
I now come to G. The most conspicuous feature of this MS is the
large extent to which its text agrees with that of F, for long stretches
almost to the point of identity. This is the reason why in most cases
the readings of G need not be mentioned in the apparatus: my
silence about any of them means that it is identical with the reading
of F. But there are, nevertheless, a number of cases where G agrees
with Tr against F: Ag. 30 ἀγγέλλων, 38 λέξειεν,2 1232 δυσφιλὲς, 1284
ἄξει, 1288 ἐν θεῶν, 1367 μαντευσόμεσθα, 1379 ἔπαισ᾽, 1383 περιστιχ-

1 In the margin of M the variant yp. αὐδᾶ(ι) is followed by the gloss d μάντις δῆλον ὅτι,
and in Tr αὐδᾷ is glossed (above the line) by λέγει ὁ μάντις.
2 Wilamowitz’s note (λέξει ἘΞ) is inaccurate: F never had anything but λέξει (not
λέξει᾽, as Franz and Hermann say).
30
 THE MANUSCRIPTS
(περιστοιχ- ΕἾ, 1446 φιλήτωρ, 1464 ἐκτρέψης, 1517 ἀσεβεῖ, 1571 δύστλητα,
1594 χερῶν, 1617 veprepa, 1658 ἔρξαντα, 1665 προσσαίνειν. From the
ὑπόθεσις the following instances may be added: three times (16, 18,
27) GTr have aiyıod- where F has aiyıor-; twice (19 and 27) the δὲ
which is wrongly added in F is absent from GTr. If we were to
assume that in these cases F alone has preserved the reading of the
hyparchetype, we should be obliged to conclude either that Tr and
G hit independently on the same corrections or that G took them
over from a Triclinian text. The latter alternative is extremely
unlikely, since all the characteristically Triclinian alterations which
we have discussed above are absent from G ; the former, involving the
assumption that the copyist of G was occasionally as much bent on
conjectural criticism as Triclinius, is ruled out by the general char-
acter of G, which is a dull and mechanical piece of work.' The correct
conclusion to be drawn from our list is that in all these cases the
readings of GTr are those of the hyparchetype, and that those of F
are individual blunders of this MS, several of them due to great
carelessness or haste. Thus G (and this is its chief value)? warns us
against the belief that wherever Tr exhibits a reasonable reading
where F has a corruption this reading should be regarded as the
result of a conjecture by Triclinius. G, then, was not copied from F.
This is confirmed by Ag. 1279, where G exhibits the ἄτιμόν which was
originally written in F (cf. p. 15 n. 1). In F the final ν is hardly
recognizable, having been obliterated in the correction to ἀτιμοί;
if.G had been copied from F, we should find ἄτιμοί in G.
The possibility must be admitted that sometimes, when GTr agree
against F, the reading in F may be due not to an individual error
in F but to a variant in the hyparchetype: 1658 ἔρξαντες is possibly
a case in point.
The manner in which the metrical scholia in G (it has no others)
are badly crowded together in its narrow margins seems to show that
originally it was not intended that scholia should be added here at
all. It is possible, though it cannot be proved, that when the text
of G was completed (its source, as we saw, was not F), the scribe
used F, or a ‘gemellus’ of F, for the metrical notes which, on second
thoughts, he had come to regard as useful. In the part of G following
the large lacuna, i.e. from 1095 to the end, the scholia are identical
with those in F, but at the beginning of the play, where the metrical
scholia in F are rather detailed, G has a shorter version: in the first

1 In the part of G containing the Agamemnon there are no signs of original conjectures.
It is true that at 1652 πρόκοπος has been changed (by the same hand?) into πρόκωπος,
but this even ἃ stupid man could do if he had his eye on the preceding line. However,
it is perhaps to the credit of the scribe of G that he sometimes leaves a blank when he
does not understand the reading of his exemplar; cf. 1221 (γέ instead of γέμος) and 1664.
2 I cannot therefore agree with Mazon (vol. ii, and ed., p. xxi n. 3), who wants to
eliminate G completely from future editions of the Oresteia.
31
 PROLEGOMENA
metrical scholion (p. 335 Wecklein) the words ὧν τελευταῖος to the
end of the scholion are missing in G; in the next (on 40: 6 παρὼν
χορὸς «rA.) G agrees with F only up to the words ἤτοι ἐφθημιμερῆ
καὶ novöperpa, and then (where F continues ἑξῆς δὲ μεταβὰς eis
ἑτέραν ὑπόθεσιν κτλ.) has merely the crude abridgement ὁμοίως δὲ
καὶ ἑτέρως τοσαῦτα. τὰ δὲ λοιπὰ χοριαμβικά.

Fate has not been too kind to the text of the Agamemnon. Above
all, the loss of the fourteen leaves of the Mediceus entailed irreparable
harm. The verdict of Wilamowitz (p. xxii), grim though it is, is none
too pessimistic: ‘Agamemnonis ea pars, quam deficiente Mediceo
haec [scil. memoria, i.e. FGTr] sola tradit, tam incerto nititur funda-
mento, ut aperta damna coniciendo reparari vix possint, alia ne
animadvertantur quidem.’ But still, looking at the history of the
text as a whole, there is a good deal to be grateful for. To begin
with, mankind will for ever be indebted to the wise decision of the
unknown editor who, at the beginning of the Roman Empire, in-
cluded the Oresteia in his selection of seven Aeschylean plays.’ In
the next place it was a most fortunate circumstance that a copy of
this selection, written probably not earlier than the fifth century
(see Wilamowitz, p. xxv), survived the Dark Ages, so that it became
possible for a Byzantine scholar of the ninth century to transcribe
this uncial codex into the new minuscule. ‘Ignotus hic vir et est et
semper erit, sospitator haud dubie Aeschyli appellandus, nam ad
exemplar ab eo ita transcriptum ut suae aetatis necessitates et
desideria postulabant codices nostri redeunt, per hoc exemplar
demum ad librum vere archetypum [i.e. the uncial codex]' (Wilamo-
witz, p. xxiv). One of the descendants of this earliest minuscule text
of Aeschylus is the codex Mediceus, another was the ancestor of V,
and a third was the MS whose remote and rather degenerate descen-
dant was to become the hyparchetype of FGTr. At that late stage
there happened something that proved providential for the preserva-
tion of the greater part of the Agamemnon. The compiler of the book
which is here called the hyparchetype of FGTr was not content to
follow the common practice of his time and copy only the Byzantine
triad, but added to it the Agamemnon and the Eumenides. But for
this lucky accident the 400 lines extant in the Mediceus and the
additional 38 lines in V would have been all that was left of the
Agamemnon. But even when the preservation of the complete play
1 Cf. Wilamowitz, Einleitung in die griechische Tragödie (= Eur. Her. i, 1st ed.), 195.
* Several arguments have been put forward in order to demonstrate that the common
source of M and all the other MSS was a minuscule codex. Here it will suffice to mention
two groups of errors common to all extant MSS: (1) wrong division of words, e.g.
Eum. 177 éx(e)ivou, 204 δ᾽ ἔκτωρ, 269 δ᾽ ἐκεῖ τίς; (2) typical misreadings of minuscule
letters, e.g. Zum. 246 νεκρόν. These examples are taken from the Oresteia ; many analogous
instances could be added from the Byzantine triad.
32
 THE MANUSCRIPTS
had thus been secured there was still a serious danger ahead. It
might easily have happened that the text of the parts missing in the
Mediceus and V survived only in the form into which it had been
brought by the violent critical manœuvres of Triclinius. That would
be the case if only Tr or a similar MS, and not F, had come down to
us. Fortunately we possess F as well, and, as has been demonstrated
above, F, though of later date than Triclinius’ edition, contains
a pre-Triclinian text, a text which in the course of time had con-
tracted all kinds of disfigurement, including a number of arbitrary
alterations, but was at least free from the particularly reckless,
though often ingenious, conjectures of Triclinius.

ema 33 D
 II
SOME EDITIONS AND COMMENTARIES
NEITHER Robertello's! skill nor the genius of Turnebus? was given
sufficient scope in the text of the Agamemnon, of which at that time
only the ll. 1-310 and 1067-1159 were known. It was not until Petrus
Victorius (Pier Vettori), using the codex F, included the whole play
in his edition (1557) that scholarly work on the Agamemnon could
begin in earnest. Victorius had also at his disposal a collation of Tr,
then in the library of Cardinal Alessandro Farnese. He rightly con-
cluded that of the two forms of the text the purer one was on
the whole that of F (cf. p. 12 n. 2). But he was so much biased in
favour of his Florentine manuscript that in many places he preferred
to the correct readings of Tr the obvious errors of F, e.g. 331 νῆστις,
556 κακοτρώτους, 577 τροίην, 791 δεῖγμα, 907 ἄναξ, 937 αἰδεσθεὶς, 1255
δυσπαθῆ, etc. Considering that Victorius, as far as the greater part
of the Agamemnon is concerned, produced the editio princeps, it is
surprising how little he did to emend the text. Even such obvious
corrections as 587 ἀνωλόλυξα μὲν, 999 ψύθη had to be introduced by
his 'printer', Henricus Stephanus (Henri Estienne), whose appendix
to Victorius' edition is a very important contribution. In the only
three cases (apart from mere orthographica) where Victorius diverged
from the MS reading he did nothing more than make an all but
inevitable change in the mood or the voice of a verb: 1381 ἀμύνεσθαι,
1654 δράσωμεν, 1658 ἐπράξαμεν.
Victorius' preface shows that he had some very sound ideas about
the genesis of our scholia; he realized e.g. that the scholia in the
margins of the MS texts are ultimately derived from ancient books,
ὑπομνήματα (he says ‘iusti commentarii", and he also saw that the
merely periphrastic scholia are of late origin (cf. on this point Wila-
mowitz, Aeschyli tragoediae, p. xxiv). But when he added a selection
of scholia to his text of Aeschylus, he was bent solely upon the
practical purpose of providing the reader with pieces of information
that might help him to understand better what the poet meant.*
Consequently he made no attempt to keep separate the pieces of
! Some scholars call him Robortelli. I have followed the guidance of E. Rostagno,
L’Eschilo Laurenziano, Facsimile, το, and of the Enciclopedia Italiana, xxix (1936), 519.
2 Of his edition of Aeschylus Wilamowitz (Aeschyli tragoediae, p. vi) says "ingenium
et doctrina editoris tantum praestitit, ut superare eum me quidem iudice potuerit
nemo’.
3 Wilamowitz in his brilliant appreciation of Victorius’ work (‘Geschichte der
Philologie’ in Gercke-Norden, Einleitung in die Altertumsw. i, 3rd ed., p. 14) praises
him for discovering an important branch of the scholia on Homer (cf. E. Maass, Scholia
in Hom. Il. Townl. i, p. viii f.).
4 Musurus’ practice in compiling his scholia on Aristophanes seems to have been very
much the same.

34
 SOME EDITIONS AND COMMENTARIES
different provenance which he found in Tr, i.e. the σχόλια παλαιά,
Triclinius’ own additional scholia (ἡμέτερα), and the interlinear
glosses, but conflated them freely. When Dindorf reprinted this con-
glomerate in his edition of the scholia (pp. 504 ff.) without indicating
its source, this was merely useless ; it was immensely more harmful
when Wecklein mixed up the σχόλια παλαιά with the scholia ‘quae
Victorius exhibet’ (p. x; what he says there to justify his procedure
shows that he had no clear idea of the nature of Victorius’ scholia).
Consequently we read in Wecklein’s edition, e.g. at Ag. 838 (829
Weckl.), 926 (917 Weckl.), 937 (928 Weckl.), 938 (929 Weckl.), scholia
which are entirely Triclinius’ own (ἡμέτερα) and were taken over by
Victorius, and, worse, at 822 (813 Weckl.) Wecklein prints Suidas’
article πάγας" δίκτυα, παγίδας κτλ., because Victorius had inserted
it into his scholia for the benefit of his readers.
Victorius’ ultra-conservative treatment of the text of Aeschylus
offered a wonderful chance to his contemporary, the true initiator
of the study of Greek poetry in France, Auratus (Jean Dorat).'
This enthusiastic and highly influential teacher was not interested
in publishing the fruits of his learning; his emendations were pro-
pagated by his admiring friends and pupils,” above all by the greatest
of them, Scaliger. In the Agamemnon Auratus’ emendations (counting
only those which either have been almost generally accepted or at
any rate are highly probable) amount to 29, that is to say they far
outnumber any other set of successful corrections proposed here by
one single scholar. Among them are of course some which anyone
might have found on what was practically virgin soil, but there are
others which show the touch of the master critic and make us see
the reason why Hermann said (on Ag. 1434 [1396 Herm.]) 'ille omnium
qui Aeschylum attigerunt princeps Auratus’.
In 1580, i.e. twenty-three years later than Victorius’ edition,
Willem Canter’s text of Aeschylus was published after the premature
death of the editor (born 1542, died 1575), ‘quem, si fata iuveni illi
pepercissent, inter summos omnium temporum philologos numerare-
mus’ (Bruno Keil in his edition of Aelius Aristides, vol. ii, p. xxxv).
In his Aeschylus the progress beyond Victorius is considerable. That
Canter should be praised because he paid special attention to the
arrangement of the lines in the lyrics and to their strophic responsion
(cf. Wilamowitz, Einleitung in die griech. Tragödie, 222, and Aeschyli
tragoediae, p. vi) is just, but it is not enough to acknowledge this
feature of his editions of Euripides (completed by himself) and
ı A lively portrait of this fascinating man is drawn by Mark Pattison, Essays, i. 206 ff.
2 Consequently we cannot in every case be sure whether a particular emendation
really belongs to Auratus. I have not done anything to check the authorship of those
conjectures which, on the basis of earlier marginal notes, were ascribed to Auratus by
Butler, Blomfield, Hermann (cf. M. Haupt in the preface to Hermann's edition, p. xvi f.),
and others, and which appear under his name in Wecklein's app. crit. or in his ‘Appendix’.
35
 PROLEGOMENA
Aeschylus. A fairly adequate idea of his capacity for textual criticism
may be obtained even from going through a single play. Canter knew
that faulty division of words is one of the main sources of trouble
in most of our MSS. This knowledge he turned to very good account
when he restored in Ag. 701 f. ἀτίμωσιν, in 963 δ᾽ εἱμάτων, in 1229
κἀκτείνασα, and in 1599 ἀμπίπτει. Observation of the laws of the
trochaic metre led him to substitute in 1671 ὥστε for ὥσπερ. His
ἐνέβης in 1567 and, perhaps, his τείνοντες in 1362 may seem pretty
obvious, but the emendations in 1211 (dvaros) and in 1418 (ἀημάτων)
are truly admirable.
The giant Scaliger never worked through a Greek or a Latin text
without jotting down in the margins a number of conjectures, often
mere freaks, sometimes real pearls. In the Agamemnon he hit the
mark in six passages ;! special praise is due to him for restoring
οἴκτωι in 134 and continuing the speech of the coryphaeus at sor.
But what he did for this play in his odd moments is as nothing com-
pared to the sustained effort of his younger friend Casaubon.
To the part played by Isaac Casaubon in the study of Aeschylus
general opinion does less than justice. The reason is that the docu-
ments which furnish the chief evidence of his activities in this field
have not met with the attention which they deserve. I shall deal
with these documents and the inferences to be drawn from them
in Appendix I (pp. 62 ff.) ; here I shall confine myself to an attempt
to outline a few characteristic features of Casaubon's treatment of
Aeschylus and in particular of the play on which, as time went on,
he concentrated more and more, the Agamemnon.
When Casaubon first set himself in earnest to work on Aeschylus,
his attention was primarily, though not exclusively, directed to
textual criticism, i.e. to the improvement of the text which he found
in the printed editions, especially those of Petrus Victorius and of
Canter. Nor did he at any later stage neglect what he justly regarded
as the backbone of any scholarly effort. This greatest of the editors
of Athenaeus recovered from that author the genuine reading zavóv
at Ag. 284, which in the MSS of Aeschylus had been ousted by $avóv.
At 69 he hit the mark with the simple change to ὑποκαίων. At 336,
unlike his predecessors, he did not tolerate the nonsensical δυσ-
δαίμονες, but emended the passage once and for all. At 1092 (ἀνδρο-
σφαγεῖον) he recognized the bold Aeschylean compound, at 1122 he
achieved what is likely to be the final emendation. On 1410, where
the editors tell us that it was left to Hermann's pupil Seidler to set
the dochmiacs right, Casaubon makes this comment: ‘lego ἀπόπολις
vel ἀπόπτολις ex contrario versu ἀντιστροῴ.᾽, and at 1430 he restored
both sense and metre by reading réuua<r).?
1 For the copy of Victorius" Aeschylus annotated by Scaliger see p. 67.
2 That he was in the habit of paying attention to the metre of the lyrics may be seen,
36
 SOME EDITIONS AND COMMENTARIES
One of the most important tasks of the editor of a dramatic play
may be included in the category of textual criticism without unduly
stretching the term, viz. the task of ascribing every line to its proper
speaker. Here Casaubon’s penetrating interpretation led him to real
triumphs. Before him not even Canter had found fault with the
notation of the MSS at 258-354, which turned this scene into a
dialogue between an ἄγγελος! and Clytemnestra, and thus made
havoc of the structure of the whole tragedy. A no less silly mistake,
though on a minor scale, disfigured the dialogue 622-35, and here it
was again Casaubon who first spotted and corrected the blunder.
Only those who degrade the noble craft of textual criticism to
a plaything fancy that it is easier or less important to give the right
interpretation of a difficult passage than to change the text by a
brilliant conjecture. In the centuries after Casaubon the phrase
Ag. 2 φρουρᾶς Ereias μῆκος has been exposed to all sorts of distortions
or far-fetched suggestions. He himself, with his firm grasp of Greek
and his unfailing common sense, kept clear of violence as well as
artificiality and so understood the construction of the sentence far
better than many later critics. In the margin? at 237 φθόγγον apatov
οἴκοις he observes: ‘haec verba add. ut sint ἐξηγ. praecedentium, ita
prorsus ut fabulae initio adiecit φρουρᾶς Ereias u.’ As another instance
of Casaubon's ars interpretandi I choose Ag. 931, which Casaubon,
clinging strictly to the fixed meaning of a set phrase, correctly
rendered "Responde mihi ex animi tui sententia'. Here several
eminent scholars have missed the sense and the tone of the sentence
and consequently have taken as a plain expression of obstinacy what
is in reality a significant piece of Clytemnestra's subtle cunning.
Casaubon is fully alive to certain stylistic devices of which
Aeschylus is particularly fond. One instance must suffice here. He
says (Paris MS, fol. 15) ‘debemus notare . . . Aeschylum solitum esse
quod dixit obscuris verbis postea quid intelligat explicare' and again
(ibid., fol. 107) ‘ut iam diximus semper solet Aeschylus illa quae satis
obscure dixit postea illustrare clariori sententia'; for the pheno-
menon in question, cf., for example, my notes on 136 and 238.
A sixteenth-century scholar cannot perhaps be expected to give
much thought to the details of the action on the stage when some
effort is needed to work them out; it is the more gratifying to find
Casaubon making on 83 ff. the observation 'senes Clytaemnestram
compellant absentem', etc. (cf. vol. ii, p. 51 n. 1).
Casaubon's keen interest in all branches of Realien, institutions of
private life, sacred rites, details of political and military organization,
e.g., by his remark on 427 (Paris MS, loose sheet fol. 20) ‘doctissimi viri [i.e. Auratus]
censuerunt tollendum esse illud ἐφ᾽ éoréas . . . sed metrum corrumpunt".
T Consequently the ἄγγελος obtained a place in the list of dramatis personae after
the ‘Ymd@eats.
2 ie. of the ‘Cambridge Aeschylus' (see p. 62 f.).

37
 PROLEGOMENA
and so on, is as manifest in his notes on the plays of Aeschylus
as in anything else he wrote; it is unnecessary to quote particular
instances. Of greater interest is the deep religious feeling that this
fervent Calvinist applied to the tragedies of an intensely religious
poet :! the instances quoted in Appendix I, pp. 65 and 76 f., will be
sufficient to illustrate the point.
But far more important than any details is the general character
of the commentary on the Agamemnon which Casaubon planned and
to a large extent executed. Its characteristic features emerge clearly
from the notes and manuscripts that have survived. We see here the
endeavour of a great and good man to blend the kind of instruction
that would be welcomed by an all but Greekless reader with the
communication of the highest technical knowledge, to combine the
discussion of choice grammatical and antiquarian problems with
moral and religious edification, and above all to do full justice to the
dominating ideas and the artistic qualities of the Agamemnon (as
the commentator saw them) and at the same time not to shirk the
minutest detail, however thorny. Such an effort devoted to such
a subject was a novelty in the history of European scholarship.
The first Englishman to leave his mark on the study of Aeschylus
was Thomas Stanley. In this amiable, gifted, and industrious man
there is nothing that could justly be called great; and yet the
enormous success with which his edition of Aeschylus met and the
influence which it has exerted and is still exerting after nearly three
centuries are not undeserved. Stanley represents in a pure form the
type of the learned amateur,? with his undeniable advantages over
the professional scholar and also his inevitable shortcomings. The
son of a respectable and well-to-do house (both his parents belonged
to the gentry), Thomas Stanley had never to worry about the
necessities of life. His mother came from a family several members
of which distinguished themselves by their literary activities. In his
youth he enjoyed all the advantages of an expensive and carefully
directed education, including the ‘grand tour’. Throughout his life
‘he cultivated literary society, and his wealth enabled him to aid
many less fortunate men of letters’.? He first gained a reputation as
a writer of lyric poems and a translator of ancient and modern
(Italian, Spanish, French) poetry. A competent judge pronounces the
verdict that Stanley’s work ‘possesses very considerable charm’ and
™ Throughout Casaubon’s life and work it is true that his ‘literary ardour was liable
to be checked by a controlling religious sentiment’ (Mark Pattison, Isaac Casaubon,
and ed., 49). .
2 This classification is not based on the general circumstances of the man’s life or
the fact that he never held an academic post. No one would, e.g., call Musgrave, one of
the masters of textual criticism in Euripides, an amateur because he was by profession
a physician, nor should we apply that term to the excellent Homeric scholar Walter
Leaf, who was a highly efficient and influential banker.
3 Dictionary of National Biography, liv. 79.
38
 SOME EDITIONS AND COMMENTARIES
that in many of his translations and throughout his original verse
‘he has succeeded in maintaining a very high level of favour and of
prettiness’.!
Before he had reached the age of 30, Stanley turned to a far more
ambitious scheme. His History of Philosophy, even when judged
merely by the exertions which the compilation of its materials and
the actual writing required, must be considered a very creditable
effort. He cast his net widely and brought together an impressive
body of evidence, which he shaped into a very readable and ex-
tremely well-arranged account. Methods of historical criticism or
anything like an analysis of the sources cannot be expected in a book
of that period. The History of Philosophy is linked up with Stanley's
earlier production by the translations (all of them in verse and most
of them rhymed) of certain poems which it contains, e.g. of Solon’s
elegies, Plato’s epigrams, Aristotle’s hymn on Hermias, the carmen
aureum of ‘Pythagoras’, the greater part of Aristophanes’ Clouds,
and the whole of Ausonius’ Ludus septem sapientum. But what here
concerns us most is the remarkable instinct for the needs and tastes
of an educated public which Stanley showed when he embarked on
his bold undertaking. The success of the History of Philosophy (it
maintained its position as a standard book well into the eighteenth
century) proved that this bulky work with its mixture of scholarship
and deft popularization was excellently calculated to satisfy the
demands of many generations of readers. The same instinct prompted
Stanley to plan the work with which his name will for ever be con-
nected, his Aeschylus. One can easily understand the enthusiasm
with which this edition was greeted by his contemporaries. Here was
to be found, in the compass of a single folio volume of moderate
size, the whole of Aeschylus, including the fragments, accompanied
by a translation in straightforward Latin prose and followed by a
commentary. Thus the work of one of the greatest poets, which up
to that time had been reserved for a small minority of highly skilled
scholars, was suddenly made the common possession of thousands
of readers all over the civilized world.
Apparently it was not until the first volume of his History of
Philosophy (published in 1655) drew near its completion that Stanley
began serious preparations for editing Aeschylus.? At that time his
1 G. Saintsbury in The Cambridge History of English Literature, vii. 83 f.
2 It gives an entirely wrong impression of the book when J. E. Sandys, History of
Class. Scholarship, ii. 351 (the paragraph is an abridged reproduction of Sidney Lee's
article in the Dict. Nat. Biogr.), says that ‘it is mainly derived from Diogenes Laertius’.
3 In Appendix II, in which the evidence for many of the following statements will
be found, it is shown that the copy of the Victorius edition of Aeschylus into which
Stanley's early notes and parts of a translation had already been entered wasin Pearson's
hands in 1654/5. Stanley's excerpts from Stephanus Byzantius (‘Ex MSS Palat. Bib.
Vat. Stephani de Urbibus, which he made primarily for the sake of his Aeschylus,
bear the date ‘anno 1653’ (Camb. Univ. Libr., Stanley MS Gg. iii. 15, pp. 167 ff.).
39
 PROLEGOMENA
equipment for so ambitious a task was utterly inadequate. His early
marginal notes are of a very elementary kind, he struggles hopelessly
even with points of no great difficulty, and his acquaintance with
the general apparatus of Greek scholarship, such as the ancient
lexicographers, the technical writers, and so on, is sadly limited.
However, he soon made determined efforts to amend this by widening
the range of his reading and collecting materials in a systematic
manner.' But the wisest thing he did was to turn for assistance to
the man who was in a better position to help than anybody else, John
Pearson, afterwards Bishop of Chester. Pearson, though primarily
known as a theologian, was perhaps England’s greatest classical
scholar before Bentley (his only possible rival being Gataker); in
range of learning and critical power he is probably inferior to no
English scholar save Bentley.” Both in the choice of his subjects
(many of them on the borderline of theological and classical studies)
and in the manner in which he tackled them he showed himself the
true heir to Casaubon’s scholarship (in his Prolegomena to Hierocles,
De providentia, London 1655, p. 5, he calls Casaubon ‘criticorum
princeps’). To Stanley’s request Pearson responded with the gener-
osity of a μεγαλόψυχος : he filled the margins of Stanley’s copy of
Aeschylus with a wealth of rare learning and acute criticism. Stanley
took the whole of it over, lock, stock, and barrel.
It is unlikely that the value and the stimulating effect of Stanley's
edition would have been nearly so great if Pearson’s anonymous
contributions had not formed an important part of it. It is to the
clandestine collaboration of these two wholly dissimilar partners that
we owe the particular character of one of the most influential editions
of a Greek poet. For what through the medium of this book has for
a long time exerted its influence on the life of classical scholarship
is neither Stanley’s graceful versatility and humanistic enthusiasm
alone, nor isit entirely the vast learning and the searching intensity of
the great scholar who served as an invisible helper, but rather the
happy combination of the two forces, neither of which would in
itself have been sufficient, since for the task of a solid and tactful
interpretation of great poetry both are required and each has to
rely on sympathetic support from the other side. The first substantial
contribution made by English scholarship to the study of Greek
! See the MS in Cambridge (cf. the preceding footnote) that contains his excerpts
from various authors. The large mass of his notes on Euripides is particularly instruc-
tive: there it is obvious that all the time Aeschylus is uppermost in his mind (the
marginal note ‘Agam.’ occurs very often). It is also noticeable that he constantly draws
on the ancient lexicographers, the Etymologica, the paroemiographi, and the scholia
(including Eustathius). |
2 I cannot agree with Bywater, who in his inaugural lecture Four Centuries of Greek
Learning in England (1894, published posthumously at Oxford in 1919), p. 14, classes
Pearson with a number of scholars of whom he justly says that they were not of the
first order.
40
 SOME EDITIONS AND COMMENTARIES
poetry is truly representative: it points to the future development
of this branch of learning and to its divergent but not necessarily
antagonistic tendencies.
In addition to the information which Stanley received from Pearson
directly, he seems to have derived considerable profit from his
example. During the seven or eight years that passed between the
time when Pearson entered his notes into Stanley’s copy and the
publication of the edition (1663) Stanley must have worked very
hard. The notes in his commentary (I am not, of course, speaking of
those which he borrowed from Pearson) are on a much higher level
than his early marginalia. And when the book was at last published,
its author did not by any means regard it as final. He continued to
collect ample materials, which he hoped to incorporate in a second
edition. For this purpose he had a copy of his Aeschylus interleaved
and divided into eight parts; the addenda and corrigenda contained
in these volumes are copious.’ The increase of material is not, how-
ever, the most impressive feature of these later notes. Again and
again we see Stanley pondering afresh over passages where his
former solution does not satisfy him any more. 'Cogita' or 'cogita
an potius' and the like recurs in many places. His translation, too,
is now subjected by him to severe criticism. Moreover, he plans to
enhance the usefulness of his work by the addition of certain fresh
categories of illustrative material and by certain changes in the
general arrangement. On the fly-leaf of the first fascicle (Adv. b. 44. 1)
we read among other entries: 'Hephaestion: f(ortasse) praeponen-
dus oxoAiis (aut saltem quae ad Tragicorum rationem carminum
spectant) et Latine reddendus’, 'Notae variorum seorsim reponendae,
inter Canteri Notas, et Commentarium nostrum. sub hac epigraphe
Variorum Excerpta’, 'Affigenda argumenta scenarum et chori, in
margine prout fit ab Interprete Sophoclis', ‘Charta Geographica.
in prima punctis designetur Itinerarium Iüs, in Prometheo. et in
Supplic. Regiones Persicae et Graecanicae, in Persis. Faces Agamem-
noniae, in Agam.' It is gratifying to see how Stanley, far from being
spoiled by the success of his book, strives to do his utmost in order
to make the work worthy of its subject. What he here sets down as
necessary elements of ἃ commentary on a dramatic poet goes far
beyond the ideas of his own time: it anticipates conceptions of the
nineteenth century.
There remains the vexed question of Stanley's plagiarism. What
we now know about the somewhat callous manner in which Stanley
exploited Pearson's emendations and interpretations and every
1 Cambr. Univ. Libr., Adv. b. 44. 1-8. The bulk of these notes was published by
Butler in his re-edition of Stanley. But Butler printed only what he considered impor-
tant, leaving out not only the long quotations from earlier scholars (P. Victorius,
Casaubon, etc.), but also remarks by Stanley himself. Sometimes the manner in which
Butler abridges makes it impossible to see Stanley's point.
41
 PROLEGOMENA
detail of his felicitous phrasing seems at first to add plausibility to
the aspersions cast by Blomfield upon his honesty. We should,
however, beware of rash conclusions. In Appendix II it is demon-
strated that in using Pearson’s materials without restriction Stanley
certainly did not act as a thief. Therefore the accusation rests on
the character of his borrowings from earlier scholars. To begin with,
it must be conceded to Butler (Aeschylus, vol. viii, p. xxi) that no
charges of dishonesty should be based on the notes which Stanley
wrote down after the publication of his Aeschylus, for we do not
know in what form and with what kind of acknowledgement he
intended to use them in the second edition. Then we have to take
into account the literary conventions of the period. A scholar of the
sixteenth or seventeenth century, when he says ‘legendum’, ‘lego’,
or something to the same effect, does not imply that he claims the
reading in question as a conjecture of his own, but merely states that
he adopts a reading different from that which he finds in the edition
he uses. Moreover, at that period, when many suggestions of many
scholars drifted about in the margins of classical texts, the author-
ship of a conjecture was, broadly speaking, regarded as a matter of
much less consequence than at a time when the means of registration
had become easier and the ambitions of most scholars were different
from what they used to be. Finally, in the case of Stanley ample
allowance must be made for carelessness, haste, and slips of memory.’
A good illustration is provided by the way in which Stanley mentions
certain readings which had in fact been advocated by Canter. On,
e.g., Ag. 803 (812 St.) Stanley's commentary says ‘lege θάρσος ἀκούσιον᾽,
on 1211 (1220 St.) ‘legendum censeo ἄνατος. Hesychius, Avaros,
ἀβλαβής᾽, on 1474 (1483 St.) ‘mallem ἐπεύχεαι᾽, on 1637 (1646 St.)
‘lege iv’, on 1672 (1681 St.) ‘manca haec sunt, et ex Scholiaste sup-
plenda hoc modo. . . .” All these suggestions were made by Canter,
and all are printed under Canter’s name (i.e. excerpted from Canter’s
edition) in Stanley’s edition, p. 688 f. ; therefore it would be fantastic
to assume that Stanley wanted to conceal that they were Canter’s.
But when he uses precisely the same expressions in recommending
conjectures made by Auratus or Scaliger, he is called a plagiarist.
Still, it may be argued that he had a much better chance of getting
away with his ‘theft’ if the emendation in question was not taken
from a printed book but from some marginalia or had been made
known to him in some other way, and that therefore in such a case
there is a strong suspicion of mala fides. Against this it must be said
that there are many instances where he carefully records the author
of an unprinted conjecture, e.g. on Ag. 547 (556 St.) ‘legimus cum
Aurato στρατοῦ᾽, 1172 (1181 St.) ‘legit Jacob . . .', 1261 (1270 St.)
‘laborat sensus, qui tamen melius constabit si legamus . . . vel, cum
! This was rightly pointed out by Butler, loc. cit.
42
 SOME EDITIONS AND COMMENTARIES
Jacob, . . .’, 1555 (1564 St.) ‘lego . . . quod et ad oram libri sui obser-
vasse Jacob video’, and similarly in the ‘Addenda’, e.g. on 427
(437 St.) ‘Auratus glossema putat’, 980 (989 St.) ‘ad oram cod.
Scaligeriani, ἀποπτύσαν᾽, 1061 (1070 St.) ‘f. καρβάνου, et ita Auratus’,
1664 (1673 St.) ‘sunt qui malint αἱρούμεθα. Auratus αἰρούμεθα᾽,
etc.
In this connexion two factors ought not to be overlooked. First,
it was almost inevitable that during the early stages of the study of
Greek Tragedy, ὅτ᾽ ἀκήρατος ἦν ἔτι λειμών, in certain places certain
emendations should occur to any scholar who knew a sufficient
amount of Greek and read his author with some attention. Stanley,
as we have seen, made considerable progress in scholarship after the
time when he gave his copy of Aeschylus to Pearson. There is there-
fore nothing improbable in the assumption that he hit independently
on a number of corrections which, without his knowing it, had been
made by other scholars before him. Secondly, many scholars have
had the unpleasant experience of fancying that they have emended
a corrupt passage entirely on their own, only to discover afterwards
that they had previously seen the emendation somewhere else. What
happens in such cases is that the improved reading sinks into our
mind, but that we forget all about the circumstances under which
we first met with it. The picture of Stanley’s personality which has
gradually formed itself in my mind makes me inclined to believe that
his memory played him such tricks more than once. I do not think
that he was dishonest when he made such comments as in his note
on Ag. 1595 (1604 St.)—I deliberately choose an instance where
appearances are strongly against him—‘suspicabar legendum ἔκρυπτ᾽
ἄνω θεὶς ἀνδρακὰς καθημένοις". I do not deny that he must somewhere!
have seen that this version of the line, involving three alterations,
belonged to Casaubon, but I very much doubt whether he remem-
bered it when he wrote his note. Carelessness, yes ; dishonesty, no.
The discussion of Stanley's so-called plagiarism has also suffered
from a lack of sense of proportion. This becomes clear as soon as his
commentary is viewed in its entirety and judged by its author's
intentions. Stanley does not make much of his emendations. To
him the correction of the text is but a minor, if necessary, ingredient
of his work. His main objective is a full and scholarly illustration of
the plays. He is often remarkably successful in penetrating beyond
mere details; sometimes he even catches a glimpse of what a later
age would have termed dramatic technique. On Ag. 258 (266 St.) he
observes 'non lac lacti magis simile atque hic locus illi est in Persis
[155], ubi senes Persici (ex quibus constituitur Chorus) de expeditione
Xerxis valde solliciti (ut Graeci nostri de Agamemnone) longa
* For a possible source see Blomfield, Museum Criticum, ii (Cambridge 1826),
488 f.

43
 PROLEGOMENA
adhibita oratione, tandem ingredientem Reginam, mutato genere
carminis, salutant : quod videntur non animadvertisse qui Nuntium
hic ingressum, et Trojae expugnationem quam ab accensa face
didicerat exponentem, commenti sunt’. This is excellent. The refer-
ence to the close analogy in the Persae would be extremely valuable
even if Stanley knew that the lines had already been restored to
their proper speakers by Casaubon. It seems, however, obvious that
a scholar who studied the structure of Aeschylean scenes in this
manner needed nobody’s assistance to reject the monstrous arrange-
ment of Ag. 258 ff. in the MSS and the earlier editions. In his notes
on 271 (279 St.) and 277 (285 St.) he adds some sound arguments on
the same point. Nor is this an isolated instance of Stanley’s readiness
to interpret the dialogue in the light of the dramatic action. His
discovery that Ag. 1650 must be spoken by Aegisthus was a fruit
of the same kind of alert attention. Perhaps even our generation has
still something to learn from a man who brought so adequate an
understanding to the reading of ancient Drama.
Stanley’s edition was re-edited by Jan Cornelis de Pauw (in 1745),
‘adiectis etiam suis, hoc est audacissimis, saepe etiam inanissimis
adnotationibus’ (Butler). This scholar, like Canter, hailed from
Utrecht, but there the similarity ends. Pauw, a very unpleasant
character, was in the habit of making a fool of himself, though he did
not invariably do so.' That he sometimes hit on excellent emenda-
tions is shown by e.g. Cho. 532 οὖθαρ ἦν and 734 τοῖς ξένοις. In the
Agamemnon his alterations of the text are confined to the subordin-
ate, if necessary, process of restoring in the lyrics syllabic responsion
(e.g. in 165 and 758), mostly by adding a final ı or w to datives in -ats
or -oıs (683 f., 777, 1095, 1173, 1536). However, Pauw went beyond
such trifles when he suspected Ag. 7, and every now and then he
understood correctly some not quite obvious construction, though
on the whole there is little useful information to be found in his
boastful notes.
Two years before Pauw’s edition the first two parts of F. L.
Abresch’s Animadversionum ad Aeschylum libri tres (the third book,
dealing with Eum. and Suppl., appeared in 1763)? were published in
Holland, where Abresch, a German by birth, had settled (like his
great contemporary Ruhnken). These Animadversiones are in fact an
extensive commentary on a large selection of passages in Aeschylus
(many of them difficult ones). A glance through the book will at
once assure the reader that he is in the company of a good scholar
1 Even in Pindar, where Pauw’s name has become a byword, some of his suggestions
are valuable; see the balanced judgement of Tycho Mommsen in the Praefatio to his
edition (1864), p. ix. ~
2 ] have used the reprint published at Halle in 1832, which was supervised by the
great Ritschl, who at that time had to stoop to such drudgery in order to make some
money (cf. O. Ribbeck, Friedrich Wilhelm Ritschl, i. 92, 98).

44
 SOME EDITIONS AND COMMENTARIES
distinguished by wide reading and a firm grasp of the Greek language.
Abresch showed better judgement than Stanley and others when he
recognized the effect of the parenthesis in Ag. 14 f. on the construc-
tion of the whole period, and consequently rejected the attempt to
alter the beginning of 12.' Among the rich parallel material furnished
in his notes the collections of certain thought-patterns and of
formulas of ordinary speech are particularly valuable, e.g. on Ag. 37
ei φθογγὴν λάβοι, and on 67, where he pointed out the common
element in, and the meaning of, expressions like ἐσμὲν οἷόν ἐσμεν,
εἴσ᾽ οὗπέρ εἰσι etc. In his note on Ag. 664 he demonstrated that
θέλουσα in such a context goes back to the language of prayers, and
thus showed the wantonness of the conjecture vaveroAoóc', which
nevertheless found admirers long afterwards. His restoration of
καρδιόδηκτον in 1471 is generally accepted. Abresch deserved well of
Aeschylus ; therefore he may be forgiven the crude misinterpretation
of 1428 λίπος Em’ ὀμμάτων αἵματος ed πρέπει which has endeared itself
to almost all his successors.
Benjamin Heath’s book, Notae sive lectiones ad . . . Aeschyh,
Sophoclis, Euripidis quae supersunt dramata deperditorumque relli-
quias, was published about the same time (1762) as Abresch’s Observa-
tiones, and its general plan is very similar; but in learning and
intensity of interpretation Heath is inferior to Abresch. Some of
his interpretations are of course correct (not only when he rejects
Pauw’s wild suggestions), but, without passing judgement on his
whole book, it may be said that the chapter on the Agamemnon is
of no great importance.
The notes of Pauw, Abresch, and Heath dealt with select passages
only. It was not until the publication of Schütz’s fully annotated
edition of Aeschylus (from 1782 on) that a substantial advance
beyond Stanley was made on the whole front. To call the mind of
Christian Gottfried Schütz ‘lentum ingenium’ (Wilamowitz, Aeschyli
iragoediae, p. vii) is hardly fair. Hermann’s verdict, which I have
quoted in vol. ii, p. 254, seems to be nearer the truth, even if we bear
in mind that this compliment was written during the lifetime of ‘the
veteran Schiitz’ (as Hermann calls him) and was obviously meant
to please the old gentleman. Speaking for myself, I gratefully remem-
ber that at an early stage of my struggles with Aeschylus Schiitz’s
clear, honest, and tactful commentary helped me a great deal to
find my way through the text. Schiitz is not content with a selective
method, but tries to comment on every section of a play in its
entirety. Moreover, his contribution to textual criticism is far from
negligible. Here is a list of successful emendations of his, culled from
the Agamemnon only (omitting minor corrections): 69 ἐπιλείβων,
᾿ 307 f. ὑπερβάλλει... φλέγουσα, 312 τοιοίδε τοί μοι, 948 δωματοφθορεῖν,
* I ought to have quoted him in the commentary ad loc.

45
 PROLEGOMENA
1012 πλησμονᾶς, 1084 περ ἐν, 1165 deletion of κακὰ, 1174 κακοφρονῶν,
1332 δακτυλοδείκτων thus rightly understood, 1563 θρόνωι, 1655 θέρος.
Several of these changes are emendationes palmares. Furthermore,
Schütz recognized the interpolations of 871 and 1290. In the latter
case he afterwards changed his mind for the worse and decided in
favour of an unsatisfactory transposition, for as he grew old his
courage began to fail him. The same decline of critical vigour can
be observed in his later treatment of the lines Cho. 205-10 and 228,
which he had once suspected to be post-Aeschylean insertions (for
details see vol. iii, Appendix D). This discovery, which paves the road
_ to the true appreciation of one of the finest scenes in Greek Tragedy,
is a good example of Schiitz’s sympathetic understanding of what is
characteristically Aeschylean.
A short anonymous review (written in May 1783) of the first volume
of Schiitz’s Aeschylus was the earliest publication of an unknown
young Fellow of Trinity College, Cambridge, Richard Porson.! This
man was soon to do for the text of Greek plays what none of his
contemporaries would have been able to do. Here, where we are
only concerned with his contribution to the study of Aeschylus, we
should bear in mind that if this part of his work did not exist, it
would make no difference to Porson’s fame. However, his approach
to textual criticism in Aeschylus is characteristic of the man. Porson
was probably the first scholar who fully realized that a truly critical
edition of Aeschylus was impossible unless it was based on reliable
information about the readings of the Mediceus. The tragicomic
failure of his attempt to obtain that information is well known :?
when the Syndics of the Cambridge University Press had entrusted
Porson with the editorship of a new edition of Aeschylus, he applied
to them for a small grant to enable him to go to Florence and collate
the manuscript, but was told ‘Let Mr. Porson collect his manuscripts
at home’. What follows is a sad story. ‘In 1795 a folio Aeschylus was
issued from the Foulis Press at Glasgow, with some corrections in
the text. These were Porson’s; but the book appeared without his
name, and without his knowledge. He had sent a text, thus far
corrected, to Glasgow, in order that an edition of Aeschylus for a
London firm might be printed from it ;* and this edition (in 2 vols.
8vo) was actually printed in 1794, though published only in 1806, still
without his name [and without a preface or notes]. This partly
corrected text was the first step towards the edition of Aeschylus
τ Reprinted in Tracts . . . of the late Richard Porson, collected by Thomas Kidd
1815), 4 ff.
2 os. Jebb in Dictionary of National Biography, xlvi. 160; M. L. Clarke, Richard
Porson, 15 f., and the same author's Greek Studies in England 1700-1830, 70.
3 This text, a copy of Pauw's edition, carefully corrected by Porson, is now in the
Library of Trinity College, Cambridge (Adv. b. 3. 1); ‘it was obtained with the other
books and papers bought by Trinity College at Porson’s death’,
46
 SOME EDITIONS AND COMMENTARIES
which he had meditated, but which he never completed'.! What we
have, then, is a mere torso, but it is adorned with some magnificent
emendations such as at Cho. 1052 φόβου νικῶ (cf. my note in vol. ii.
214 n. 1). Some of Porson’s most brilliant readings, however, are not
to be found in the edition (nor is his elimination of the sparrows
from Ag. 145), but were afterwards made known by Blomfield and,
partly from Dobree’s notes, by Kidd, Tracts, 208 ff., e.g. Ag. 850
πῆμ᾽ ἀποστρέψαι νόσον, 1391 f. διοσδότωι γάνει.
Porson’s greatest pupil, Peter Dobree, made a few very good
emendations in the text of Aeschylus; after his premature death
they were published in his Adversaria, vol. ii. 14 ff. Another member
of the Porsonian school, a less eminent figure, did a signal service
to the study of Aeschylus by producing a carefully planned and
executed work on a large scale. C. J. Blomfield began his edition in
1810 with the Prometheus and continued it until 1824, when he was
appointed Bishop of Chester and thus prevented from editing the
last two plays, Eumenides and Supplices.2 At Ag. 1356 he used a new
source discovered by him and was thus able to restore the genuine
text of a passage which had seemed hopelessly corrupt. ‘Haud
spernendas protulit coniecturas’ (Wilamowitz)—from the Agamem-
non I will mention 236 φυλακᾶι, 1590 αὐτός, and especially 1566 πρὸς
draı—but the most valuable part of his edition is the ‘Glossarium’
appended to each play. Here a wealth of linguistic information is
brought together, from literature as well as from the lexicographers,
scholiasts, etc. Blomfield very properly drew on materials collected
by his predecessors, but he greatly added to them from his own
extensive reading. The influence which these ‘Glossaria’ have exerted
on our dictionaries and almost all later interpretations of Greek
Tragedy can hardly be overrated; sometimes it requires a deter-
mined effort to disentangle oneself from the net of the fable convenue
that originates from Blomfield’s renderings.
Elmsley used a brief stay at Naples to make a collation of the text
of the Agamemnon in the ‘codex Farnesianus’ (Tr) ; he published it
in Museum Criticum, ii (Cambridge 1826), 457 ff. That the notes in
his editions of three plays of Euripides have done a great deal to
elucidate certain points in the language of Aeschylus may be seen
at several places of my commentary.
When we now turn to Gottfried Hermann, we meet for the first
time in the course of our survey a scholar of the first magnitude in
whose life (and a long life it was) the work on Aeschylus never ceased
to occupy a central position. Hermann’s book De metris poetarum

! Jebb, op. cit. 161; cf. Blomfield, Museum Criticum, i (Cambridge 1826), 110 f.;
M. L. Clarke, Richard Porson, 65 f.; Greek Studies in England . . ., 70.
* I have used the sth edition (1829) of Prom., the 3rd (1824) of Sept., the 4th (1830)
of Pers., the 3rd (1826) of Ag., and the 2nd (1827) of Cho.

47
 PROLEGOMENA
Graecorum et Romanorum (1796) is in several respects immature, very
naturally, since the author was 24 years old when he published it,
but it revealed for the first time the true nature of a number of
Aeschylean lyrics and, besides, contained some happy conjectures ;
to mention one instance, the reading suggested for Ag. 412 f. went
a long way towards restoring this difficult passage (cf. my note,
vol. ii. 217 n. 1). In 1799 he edited the Eumenides. The famous product
of Hermann’s mature age, Elementa doctrinae metricae (1816), yields
fewer decisions on questions .of textual criticism than might be
expected (at any rate so far as the lyrics of the Greek dramatic poets
are concerned, for in the large sections of Plautine cantica which
Hermann includes his observations on the text are both extensive
and important, and many a passage has there been emended by
him for good). However, those readers of the Elementa who hoped
that some fresh light might be thrown on the text of the lyrics in
Greek Drama were not entirely disappointed: they found e.g. on
P. 704 Ag. 131 (dya) and 1468 f. (διφυίοισι) corrected. But these two
emendations were by that time no longer strictly novel. In the same
year 1816, a few months before the publication of the Elementa
doctrinae metricae, Hermann’s publisher had issued a slender little
book, ‘Aeschylos Agamemnon, metrisch tibersetzt von Wilhelm von
Humboldt'.! The text on which this translation was based had been
thoroughly revised by Hermann (long discussions even on minute
details took place between the two men), and on the last three pages
Hermann gave a list, with a few very brief remarks, of the passages
where he diverged from the textus recepius. It is in this place, then,
and in this very modest form, that most of Hermann’s emendations
of the text of the Agamemnon were first made known; in later years
he added but little to them. The most conspicuous contribution was
of course made in the lyrics; thus we find here, in addition to the
two corrections mentioned above, at 254 αὐγαῖς, 369 deletion of the
first ws, 1111 dpéypara, 1133 διαὶ (practically in the MSS, but un-
discovered until then), 1455 mapdvovs, 1565 ἀραῖον. But the trimeters
were not neglected: Hermann stated here that 1203 was to be placed
after 1204, and that at 1271 μέτα was a corruption of μέγα.
After the Elementa Hermann published many important articles
on Aeschylus, but the edition of all the extant plays and fragments,
which was to be the crowning work of his life, was never completed.
After Hermann’s death in 1848 his pupil and son-in-law Moriz Haupt
published the edition from the materials Hermann had left behind.
The book is therefore different from what it would have been if
Hermann had been able to give it the finishing touch. Being in some
! The printing of Hermann's book at times delayed the progress of Humboldt’s
proofs; see Wilhelm von Humboldts Briefe an Gottfried Hermann, mitgeteilt von A.
Leitzmann (Weimar 1929), Ρ. 43.
48
 SOME EDITIONS AND COMMENTARIES
parts overladen with matter of little relevance and in others too
laconic, it lacks the easy poise of Hermann’s re-edition of Erfurdt’s
Sophocles. But it is not only these imperfections that make it
difficult for the modern reader fully to appreciate the value of the
book. The ‘Adnotationes’ are not, and do not pretend to be, what
we mean (and Casaubon meant) by a commentary. Like the notes
in Bentley’s editions of Latin poets and, for that matter, in most
eighteenth-century books of a similar form, but unlike the commen-
taries in Schütz’s Aeschylus or in Heyne’s Virgil, Hermann’s
‘adnotationes’ only very rarely turn straight to a difficult point of
interpretation or to the elucidation of the context; as a rule any
observation the editor wants to make is hung on a peg of textual
criticism, and among the witnesses quoted for a particular reading
there appear many that are to us absolute nonentities. Consequently,
the reader who intends to get at the core of Hermann’s argument has
to develop a special technique ; in course of time he will learn, while
skipping a good deal of cumbersome and antiquated detail, to con-
centrate on the priceless information intermixed with it.’ Despite
the apparent predominance of textual criticism the notes on the
Agamemnon show clearly that Hermann had not forgotten what he
had said in the opening paragraph of his appendix to Humboldt’s
translation, viz. that here, as in many other ancient texts, ‘ungleich
wenigeres einer Verbesserung, als einer verständigen Erklärung
bedarf’. The manner of his interpretation may be seen, e.g., from his
notes on the following passages: Ag. 3 ἄγκαθεν, το κρατεῖ, 57 (diffi-
culty of τῶνδε), 217 ed γὰρ ein, 613 f. (to be assigned to Clytemnestra),
637 χωρὶς ἡ τιμὴ θεῶν, 790 ff. (where, by changing 8’ to 7’, he recovers
the structure of the whole period), 900 (where he rightly takes
exception to κάλλιστον ἦμαρ in this context, though his conclusion is
wrong), 1096 f. (where both the construction of the sentence and the
true nature of the vision are well explained), 1127 μελαγκέρωι.. . .
μηχανήματι. It would, however, be a sad mistake to assume that
Hermann’s interest was confined to problems of language, style, and
metre. That this enthusiastic lover of great poetry? was capable of
visualizing a Greek play in the full setting of a dramatic performance
is shown by several of the articles in his Opuscula; here we must
especially notice his paper De re scenica in Aeschyli Orestea, reprinted
at the end of the second volume of his Aeschylus.’
! In my seminar classes at Oxford it was very gratifying to watch young under-
graduates working their way through Hermann’s Aeschylus; they soon spoke of the
book not merely with admiration but with delight.
2 ‘Der französischen, englischen und italienischen Sprache war er [Hermann] mächtig,
von ihrer Literatur fesselten nur Shakespeare und Dante ihn dauernd’ (Otto Jahn,
Biographische Aufsálze, 114).
* From this paper it may also be seen how much the vigorous old fighter had learnt
from his adversary Otfried Müller, not only in details but in general outlook.
4872-1 49 E
 PROLEGOMENA
W. v. Humboldt’s translation of the Agamemnon, which has been
mentioned in connexion with Hermann’s notes, has also a claim of
its own to be considered here. The translation, and even more so the
introduction, is a great monument of the deeper understanding of
Greek poetry and art that had been made possible by the ideas
of Winckelmann, Goethe, Friedrich Schlegel, and their followers, and
to no small extent by Humboldt himself. A typical example of the
way in which the new spirit freed men’s eyes from conventional
prejudices will be discussed when we come to Humboldt’s solution
of a vexed problem of interpretation (vol. ii, pp. 254 ff.).
The most momentous result of Humboldt’s translation was that
it brought Goethe into fresh contact with the Agamemnon. His
immediate reaction is contained in the famous letter to Humboldt
of ı September 1816 (Goethe’s Werke [Weimar edition], iv. Abtheilung,
27. Band, pp. 156 ff.), from which a few sentences must be quoted
here:
*. . . eine solche uralte Riesengestalt, geformt wie Ungeheuer, tritt über-
raschend vor uns auf, und wir müssen alle unsere Sinne zusammennehmen
um ihr einigermaßen würdig: entgegen zu stehen. In einem solchen
Augenblick zweifelt man keineswegs hier das Kunstwerk der Kunstwerke,
oder, wenn man gemäßigter sprechen will, ein höchst musterhaftes zu
erblicken. . . . Das Stück war von jeher mir eines der betrachtungswürdig-
sten... Verwundersam aber ist mir jetzt mehr als je das Gewebe dieses
Urteppichs: Vergangenheit, Gegenwart und Zukunft sind so glücklich
in eins geschlungen, daß man selbst zum Seher, das heißt: Gott ähnlich
wird. Und das ist doch am Ende der Triumph aller Poesie im Größten und
Kleinsten. . . . So hat mich auch wieder auf’s neue ergriffen daß jede
Person, außer Clytemnestra, der Unheilverketterin, ihre abgeschlossne
Aristeia hat, so daß jede ein ganzes Gedicht spielt und nachher nicht
wiederkommt uns etwa auf’s neue mit ihren Angelegenheiten beschwer-
lich zu fallen.’

The statement in the last sentence is fundamental: it illustrates

perfectly the unique οἰκονομία of this play in which all the characters,
including the central figure of the king, are given only a single scene,
with the one exception of Clytemnestra, ‘who links together the
chain of evil’. Goethe speaks as a poet who, familiar with the whole
of Europe’s dramatic literature, has also long been active as a pro-
ducer of theatrical performances. From the pages of the lofty, if
sometimes stilted, version there arises in Goethe’s mind the vision
of the play in its entirety, and his imagination finds no difficulty in
bringing it back to its natural place, the stage of the theatre.
There is fortunately no need to rake up again the quarrel between
Gottfried Hermann and Otfried Müller which centred in their dispute
about the Eumenides. Nowadays we all know that Müller’s commen-
tary on that play (1833) marked the beginning of a new era in the
50
 SOME EDITIONS AND COMMENTARIES
interpretation of Aeschylus and indeed of Attic drama in general.
Henceforth no conscientious scholar was at liberty to comment on
a Greek play without taking into account the monuments of art
as well as the literature, or to neglect the relevant problems of
religion, law, and political and social history, and the conditions of per-
formances on the Athenian stage. It is obvious that the interpretation
of the two other plays of the Oresteia owes a special debt to Müller’s
book.
Welcker’s name must at least be mentioned here, not on account
of any special contribution to the study of the Agamemnon, but
because he has done so much to deepen our views on Greek religion
and poetry in general and on Aeschylus in particular.
We have now reached the full current of nineteenth-century
activities, where it is no longer possible to follow up all the individual
rivulets. We shall have to be content with a small selection of works
(in addition to those which are of special value others will be men-
tioned on account of the influence they have had), and any such
selection is bound to be subjective.'
In autumn 1847 Hermann had a visit from a young Oxford
graduate, John Conington, who had come to Leipzig to do homage
to the prince of Greek scholars. In the following year Conington
published a text of the Agamemnon, ‘with a translation into English
verse, and notes critical and explanatory'. This edition, coming as
it did from a scholar 23 years old, was a very creditable achievement.?
It contained, of course, many immature assertions, but also some
very fine interpretations, and showed at more than one place the
instinctive grasp of certain delicate shades of poetry which is
characteristic of Conington's later work. Before Conington Heath's
transposition of 1. 1304 had been generally accepted (it is still found
in the text of Wilamowitz), with the result that the subsequent lines,
and especially Cassandra's outburst in 1305, became unintelligible.
Conington showed that the transposition was wrong;* and, from

! Some brief remarks on the commentaries of Plüss and Ubaldi and on the notes
which Sewell, Droysen, Lewis Campbell, and Platt added to their translations are made
in the commentary ; see the Index under their names.
2 H. J. Smith in his ‘Memoir’ in Miscellaneous Writings of John Conington, vol. i,
p. xxvii, says that ‘the visit . . . was embarrassed by difficulties of language (they had
to talk in Latin, and each of course pronounced it after the manner of his own country)’.
However, Conington himself (Preface to his edition of the Agamemnon, p. vii, footnote)
quotes a picturesque German idiom as used by Hermann in that conversation (he called
Droysen's translation of Aeschylus ‘luftig’).
? Henry Nettleship says of it (Dictionary of National Biography, xii. 15) “the notes,
though slight, contained one brilliant emendation, λέοντος tw for λέοντα σίνιν᾽. The
emendation λέοντος Inv (with its corollary, #0os, at 727) is not in the book (see my note
on Ag. 717), and several of the notes are anything but slight, whatever Conington
himself thought of them in later life.
* This, as Conington says in his note, was first pointed out to him by Dr. Arnold
(Conington's school was Rugby).
5I
 PROLEGOMENA
the firm ground of the order he had thus established, he was able to
follow the trend of Cassandra’s impassioned thought up to its climax.
In 1851 Conington published his unassailable emendation of Ag. 717
and 727; in 1857, when he was Corpus Professor of Latin, there
appeared his edition of the Choephoroe, ‘with notes critical and
explanatory’. This little book, though weak on the side of textual
criticism, seems to me to be one of the most useful commentaries
on a Greek play and, perhaps, the most tactful of all. The notes,
rich though they are in information on points of language and style,
yet do not make us forget that we are dealing with great poetry;
from time to time a few remarks couched in quiet and noble language
bring us back from the details to the main issues. Special attention
is always paid to the links with the Agamemnon.
The most recent publication on which Conington drew while
working on the notes of his Agamemnon was F. A. Paley’s commen-
tary in its first, Latin, form (the part containing the Oresteia ap-
peared in 1845). But it was in its renewed, i.e. its English, form that
Paley's book exerted its great influence. The English commentary
was first published in 1855; the last, fourth, edition appeared in
1879.1 Paley was certainly no genius, but a solid worker and an
honest man; whatever the shortcomings of his erudition and the
weaknesses of his other publications, we should not belittle his merits
as a guide to Aeschylus. If he often fails to reach the highest peaks,
he always takes us safely up to the foothills and sometimes beyond,
and that is far more than most of his fellow-interpreters achieve.
Many difficulties which in other commentaries are either passed
over altogether or dismissed with a casual remark are courageously
tackled by Paley. If we want to take the measure of his conscientious-
ness, we have only to observe the numerous and important altera-
tions which distinguish any subsequent edition of his commentary
from its predecessor. It is perhaps a forgivable weakness that, after
pondering so long over certain riddles and knowing of so many
different answers, he sometimes found himself unable to make up
his mind. What is worse is that it does not seem to have occurred
to him that these tragedies are not some kind of poetry to be read
as we may read an epic, but dramatic plays meant to be performed
on the stage. Nevertheless, some of his interpretations may be
regarded as final He also made a few convincing emendations,
although textual criticism was not his forte (for one thing, he was
too easily inclined to sail in the wake of Hermann). His prose-
translation of Aeschylus (2nd edition 1871) is frowned at by refined
English stylists, and to me, too, it seems rather wooden, but it is
a work of great honesty ; it prefers to come to grips with tbe harsh-
! The references in my commentary are to the 4th edition unless I say something
to the contrary.
52
 SOME EDITIONS AND COMMENTARIES
nesses and ambiguities of the Greek rather than go round them or
gloss them over. Besides, it is amusing to watch several later trans-
lators, fastidious though they are, lifting solid chunks, sometimes
with some slight modifications, from Paley’s version.
A word must be said here about B. H. Kennedy’s edition (2nd
edition 1882) of the Agamemnon, ‘with a metrical translation and
notes critical and illustrative’. His textual criticism is so wild that
the reader is almost reminded of Hartung,' but, as in the case of
Hartung’s editions, there are in the notes some observations which
make it worth while reading them. Moreover, Kennedy’s book is
permeated by a spirit of relentless inquisitiveness, and this, com-
bined with a passionate devotion to the poetry of Aeschylus, has a
stimulating effect. Above all, Kennedy deserves our gratitude for
unriddling the cryptic first few lines of the stichomythia 931 ff.
We have now to go back to Germany. F. H. Bothe, with his solid
knowledge of the Greek and Latin dramatic poets, commented on
the plays in a manner which is seldom exciting but often instructive ;
he was also a successful emendator. R. H. Klausen's commentaries
on the Agamemnon (1833) and the Choephoroe (1835) had at their
time a considerable influence in England too, but, though learned,
they are on the whole not distinguished by great originality or sound
judgement. The commentary on the Agamemnon was greatly im-
proved when R. Enger, a scholar who made valuable contributions
of his own to the study of Aeschylus, re-edited it in 1863.
Of Wilhelm Dindorf's slovenly edition of the scholia and his
sweeping statements about the MS tradition of Aeschylus something
has been said in the first chapter. His editions of the poet himself,
though hasty too, are on a somewhat higher level. Dindorf never
was, and could not possibly be, a reliable editor; it is difficult to
understand how, on top of his other activities, he managed to read
all the texts, from Homer to Eusebius (not to mention the many
volumes of scholia), which he edited. But he did read them; and as
he was a very able man and one who really knew Greek,? he will
surprise us every now and then, in the midst of the most unbelievable
carelessness, by a brilliant observation or a convincing suggestion.
His two textual corrections in the Agamemnon,* at 478 and 1472,
are of the type that attracted Pauw: they rid us of monosyllables
destructive to the metre. But in the text of the other plays of

! It is no accident that Kennedy speaks with unqualified admiration of S. Karsten's

edition of the Agamemnon (Utrecht 1855), the notes and commentary of which are
learned and by no means dull, but very extravagant indeed; it is only occasionally
that I have been able to profit from them. However, his emendation at 1163 is excellent.
* First Leipzig 1827 and Oxford 1832 (with extensive Latin notes 1841), then in the
Teubner collection (sth edition 1870).
? This even Wilamowitz acknowledges, Geschichte der Philologie, 65.
* I do not take into account mere orthographica.

53
 PROLEGOMENA
Aeschylus he succeeded in making some less easy emendations. Nor
was he insensible to the dramatic life in a scene or to a change of
style: he noticed, e.g. the major inconsistency in Orestes’ first speech
after the murder (Cho. 973-1006),' and he stressed the very odd
stylistic character of the exaggerated imagery in Ag. 895-902 ;? the
rash conclusions to which he jumped in both cases do not concern
us here. But his greatest service to Aeschylean studies was the
publication of his Lexicon Aeschyleum (1873). This book, if far from
perfect, is a very valuable instrument. It has considerable merits
of its own, although it owes a great deal to the earlier works of
Wellauer and Linwood (the latter scholar’s Lexicon to Aeschylus,
which shows in many places a remarkably independent judgement,
is still worth consulting).
Hartung’s commentaries are largely animated by his envious
hatred of Hermann, but some excellent observations compensate
the reader to a certain degree for the poison he has to swallow. His
emendations at Ag. 245 f. and 386 are among the best made in this
play.
Schneidewin in his commentary on the Agamemnon (1856) went
deeper into the secrets of the play than any editor had done before.
He detected in many passages an undertone or an implication that
had not yet been noticed. It was inevitable, on the other hand, that
so keen a searcher after hidden meanings should more than once be
tempted to find mysteries where there were none. But on the whole
the commentary profited greatly from Schneidewin’s vast learning
and especially from his intimate acquaintance with the iambo-
graphers and elegists, the lyric poets, Sophocles, and the collections
of proverbs. When O. Hense re-edited the book in 1883, he sobered
it down a good deal and added valuable notes of his own.
Of a very different character is Nägelsbach’s unpretentious edition
of the Agamemnon, posthumously published from his drafts in 1863.
Its honest prose-translation and straightforward notes often prove
helpful.
The Dutch scholar J. A. C. van Heusde’s Agamemnon (1864) is
a terrifying book; it displays its materials in the crudest possible
manner. But my debt to it is great: before I obtained photostats
of Tr, it was only here that I could look for reliable information
about all the Tr scholia, and the ‘commentary’ provides an almost
inexhaustible store of rare parallels (many of them have found their
way into Blaydes’s farrago). But the whole thing is like a charnel
house. It will be best to leave it quickly and walk back into the
sunshine.
Heinrich Ludolf Ahrens,? who once and for all placed the study of
Cf. vol. iii, pp. 813 f. 2 Cf, vol. ii, p. 410 n. 3.
3 This eminent scholar is meant wherever in the app. crit. or in the commentary I

54
 SOME EDITIONS AND COMMENTARIES
the Greek dialects and the textual criticism of the Bucolict Graect
and their scholia on solid foundations, was a very great grammarian
and a masterly editor, but he was far more than that. If we want to
form an idea of the wide range of interests, the keenness of observa-
tion, and the sympathy with Greek thought which are characteristic
of this pupil of Otfried Miiller’s, we may turn to his paper Uber die
Göttin Themis (Hanover 1862); there it can also be seen that the
archaeological material is as a matter of course taken into account
just as conscientiously as are the literary sources. Aeschylus was in
the centre of Ahrens’s studies from the days of his youth. A specimen
of his earliest article on the text of the poet (published in 1832;
Ahrens was born in 1809) is quoted in vol. ili, p. 676, n. 3. Then (to
leave aside minor contributions) he co-operated in J. Franz’s edition
of the Oresteia (1846), and finally published in Philologus, Suppl.-
Bd. I (1860), three articles, ‘Studien zum Agamemnon des Aeschylus’,
which contain a full commentary on select passages from the begin-
ning of the play to I. 1252. Several mistakes in the more recent treat-
ment of the Agamemnon might have been avoided if more attention
had been paid to this work of Ahrens; but none of the editors after
O. Hense (1883) and Wecklein has properly studied it, and most of
them do not even seem to know that it exists. The ‘Studien zum
Agamemnon’ belong to that high class of works of scholarship in
which, even where the author’s conclusions are wrong, his criticism
of his forerunners, his own line of argument, and the evidence pro-
duced by him will always be of the greatest value to anyone who
attempts to tackle the problems in a critical spirit. The first article
opens with a sound résumé of the MSS and their inter-connexion.
As has been stated above (p. 6), Ahrens was among the first to
challenge the then current assumption that the Mediceus is the
source of all our MSS. His conjectural criticism is on the whole
rather violent. However, among his successful emendations or correct
interpretations of MS readings are the following: Ag. 101 ds ἀνα-
φαίνεις, 170 λέξεται, 190f. παλιρρόχθοις, 542 ἦστε, 1194 κυρῶ, 1231
τόλμα, 1657 στείχετ᾽ αἰδοῖοι. He also saw that 863 is interpolated. But
brilliant emendations and convincing deletions have been made by
other scholars as well; Ahrens’s specific contribution to the inter-
pretation of the Agamemnon consists in his thorough examination
of certain lexicographic, and especially semasiological, facts relevant
to the meaning of difficult words and phrases in this play. Impressive
though the mere bulk of the accumulated material is, what Ahrens
gives is not dry lists but a lively and penetrating discussion carried

say simply ‘Ahrens’. On the few occasions where I have to mention E. A. J. Ahrens,
who happened to edit Aeschylus for Didot, I shall add his initials.
! Wecklein's Appendix has made it possible to quote Ahrens's conjectures without
reading his articles.
55
 PROLEGOMENA
out by a great philologist who had long been intimately conversant
with Greek literature from Homer and the lyric poets to the Byzan-
tine lexicographers and scholiasts. I will mention a few instances as
deserving special attention: on 17 évréuvew, 72 arirns, 275 (where, by
establishing the connotations of the phrase οὐκ ἂν λάβοιμι, Ahrens
made the meaning of the line clear), 276 ἄπτερος, 411 orißoı φιλάνορες,
806 εὔφρων (if the results of this investigation had been taken into
account in our lexica, we should have been spared some bad mis-
interpretations of certain passages of Greek poetry), 1196 ἐκμαρτυρεῖν.
Ahrens’s ‘Studien zum Agamemnon’ constantly refer to Weil’s
edition of the play. H. Weil, ‘der zum Franzosen geworden war, weil
ihm in Deutschland der verdiente Lehrstuhl unerreichbar war’
(Wilamowitz, Erinnerungen, 179), published a text ofthe Agamemnon,
with critical and explanatory notes in Latin, in 1858 (there followed
in 1860 the Choephoroe, and in 1861 the Eumenides, with important
‘Addenda et Corrigenda’ to the Agamemnon). This is a work of a high
standard, despite the overbold textual criticism which the editor
mitigated considerably in his later editions of the text (in the
Teubner collection). Weil’s succinct comments are almost always to
the point and help the reader, no matter whether he agrees with the
editor’s conclusions or not, to see more clearly where the difficulty
of a disputed passage lies. Many of his interpretations are convincing,
and it does not often happen that his remarks jar with the nature of
Aeschylus’ poetry. A clear thinker, a very learned scholar, and a
μουσικὸς ἀνήρ, Weil in his later essays says a great deal on Greek
Tragedy in general and on Aeschylus in particular that is worth
remembering.
We pass from a wide range of learning to narrow specialization
and, what is more distressing, from the harmony of sympathetic
understanding to the petty prejudices of Philistinism when we now
turn to Wecklein. Both language and thought in his annotated
German edition of the Oresieia (1888) often reach such a depth of
crude vulgarity that even the most hardened scholar may find it
difficult not to lose his temper. And yet it would be wrong and
indeed harmful to despise Wecklein’s work, the result of a lifetime’s
devotion to Attic Tragedy. The main value of his critical edition of
Aeschylus (1885-93) lies in its Appendtx. Shortly after its publication
Housman (Journ. of Philol. xvi, 1888, 244) greeted the Appendix with
a sigh of relief, and with a groan when he thought of the time and
labour which scholars had been forced to waste before such a com-
plete register of the conjectures of critics existed. But Wecklein was
no mere compiler. In the course of a long and industrious life he
acquired an uncommon familiarity with the language and the
dramatic technique of Tragedy, and, moreover, by dint of common
sense he often avoided the snares of improbable artificialities in
56
 SOME EDITIONS AND COMMENTARIES
which many of his betters were caught. Some of his conjectures are
very good, and some of his solutions of vexed problems carry con-
viction. The perusal of his commentaries on Aeschylus and Euripides,
though never enjoyable, is never entirely without its reward.
A great Greek scholar, whose many-sided activities spread over
a large field outside Tragedy, F. Blass, wrote towards the end of his
life a commentary on the Choephoroe (published in 1906) and one on
the Eumenides (posthumously edited in 1907), both of which are also
relevant to the interpretation of the Agamemnon. The chief value
of these books lies in their observations on language and style, but
other important issues are discussed as well. Two earlier papers of
Blass dealt with passages in the Agamemnon: he emended the end
of L 429 and gave an admirable interpretation of the sentence in
926 f., which had been generally misunderstood.
Towards the end of the nineteenth century the most original
contributions to the study of Aeschylus and Euripides in England
came from Verrall. His contemporaries were partly enthralled, partly
amused, and partly revolted, but no one could disregard or whittle
away the force of his influence. Nowadays there seems to be a certain
tendency to take no notice of him. If such a tendency exists, it ought
to be strongly resisted. It is probably true that to very young
students Verrall may become a danger, for they are usually not in
a position to protect themselves from the incantation of his brilliant
and enthusiastic sophistry. But adults should not forgo the benefits
that may accrue to them from listening to Verrall and, if possible,
refuting him. When all is said and done, there remains the fact that
we are here in the presence of a first-rate mind and a real artist’s
soul. Verrall was by nature out of sympathy with fifth-century
Athens and all it stands for (the Silver Latin poets on the one hand
and Dante on the other are far more congenial to him). Aeschylus
was too simple and solid for his taste; he wanted him more complex
and refined. Therefore he was always ready, by assuming an in-
nuendo or a double meaning, to read some psychological subtlety
into the words of the poet. Still Verrall knew a great deal about
things Greek, and—despite the long catalogue of his linguistic sins
drawn up by Headlam—he knew the language exceedingly well.
Moreover, he was endowed with a quality which made him parti-
cularly fit to expound the works of the Attic tragedians: he had
a lively idea of the conditions of dramatic poetry. No English scholar
before Verrall, and no earlier scholar at all save Otfried Müller (to
whom in this respect Verrall’s contemporary Wilamowitz was equal
from his youth), was possessed of a similar capacity of visualizing
a Greek tragedy as a play to be performed on the stage. Consequently
one vital aspect of this poetry was more adequately brought out in
Verrall's commentaries than in those of his forerunners. There are
57
 PROLEGOMENA
a great many other points where the freshness of Verrall’s approach
enabled him to shed time-honoured prejudices and reinstate the
truth; his interpretation of Ag. 1428 provides a good instance.
Unfortunately he had little patience and even less of that special
gift of scholarly perseverance that enables a man to swallow vast
clouds of dust in the faint hope that in the end his labour may be
rewarded by a small grain of gold. Verrall would not pause to con-
sider various possible answers to an intricate question; his lively
imagination carried him quickly to the point where a shining
phantom appeared, in shape and colour very much like the real
thing, but with a seductive glamour of its own. His textual criticism,
following as it does the uncompromising ‘conservative’ method (even
where all that we have to go upon is FTr), is the least interesting
part of his work:: it is so monotonous as to become dull, while every-
thing else in his books sparkles with life. It would not be surprising
if Verrall’s keen inquisitiveness and powerful fantasy should con-
tinue to stimulate serious scholars at a time when the methodical
conclusions of more judicious critics have long been absorbed in the
stagnant reservoirs of recognized opinions.
A violent invective against Verrall’s treatment of Aeschylus was
(apart from a small volume of translations from Meleager) the first
book published by Walter Headlam (On Editing Aeschylus, 1891).
Afterwards he wrote many articles on passages in Aeschylus! and
translated for the series of 'Bell's Classical Translations’ the Pro-
metheus, the Oresteia, and the Suppliants, with brief but valuable
notes. His death in 1908, at the age of 42, cut short the high expecta-
tions of what he might have done as editor and commentator of
Aeschylus. In 1910 A. C. Pearson published a text of the Agamemnon
from Headlam’s materials, with Headlam’s verse translation and
some of his notes.? Recently it has become possible to know far more
of Headlam’s work on the Oresteia thanks to the piety and patience
of George Thomson, who undertook the arduous task of deciphering
and transcribing Headlam’s notes in the margins of two of his copies
of the plays. To these materials and those published by Pearson,
Thomson added the relevant sections of Headlam’s printed articles,
and finally included the whole in his own edition of the Trilogy.
We should be guilty of gross injustice if we measured the torso
of Headlam’s work by the same standard which would be applicable
if he had been able to complete his edition. What we can safely do
is to try broadly to assess the value of his unfinished work. Headlam
was inferior to Verrall in brilliance and originality, but greatly
1 À bibliography (by L. Haward) is appended to the book Walter Headlam, His
Letters and Poems, With a Memoir by Cecil Headlam, London 1910.
2 On the defects of this posthumous edition see George Thomson, The Oresteia of
Aeschylus, vol. i, p. x; on the mistake by which the Introduction was inserted as if it
were Headlam’s see M. R. James in Atkenaeum, no. 4368 (15 July 1911), p. 73.
58
 SOME EDITIONS AND COMMENTARIES
superior in seriousness and devotion to the search for truth. For him
a pleasing possibility was not good enough; his mind would not be
at ease until he had exhausted every means of getting at the sense
of an obscure passage or emending a corruption. Again and again
he would return to the same crux and attempt a fresh solution. With
unflagging industry he piled up materials from the classics as well
as from remote corners of Greek literature in order to exploit them
for the interpretation of Aeschylus. In doing so he did not always
escape the danger of a method which, however laudable in itself,
requires careful discrimination and a strong historical instinct if the
issues are not to be confused. Headlam was sometimes apt to intro-
duce into the thought and the language of Aeschylus elements which
belong to a much later period in the life of the Greek people. Like
Verrall, Headlam was more attracted by psychological complexity
than by simple grandeur; consequently he preferred in many cases
the glittering fluctuation of a double meaning to the straight line
of a plain thought. But unlike Verrall, Headlam was not very much
interested in problems of dramatic technique and in the links by
which the plays are insolubly tied to the conditions of a given stage.
In his conception of the figure of Agamemnon he followed the con-
ventional misrepresentation which turns the great and noble king
into a boastful and weak little man.!
Headlam’s progress beyond the earlier commentators is not to
be found in a very different general view but rather in the more
correct and penetrating interpretation of certain details, some of
which are of great importance. Several felicitous emendations appear
as the natural result of his intense struggle for a deeper understanding
of the text. Of the passages in the Agamemnon where his interpreta-
tion seems to me particularly successful I will mention 345-7, 527
(where his argumentation should convince even the most conserva-
tive critic that Salzmann’s deletion of the line is necessary), 637
χωρὶς ἡ τιμὴ θεῶν, 934 (on εὖ εἰδώς), 1657 (on the absurdity of the
phrase πρὸς δόμους πεπρωμένους), and above all his recognition of the
meaning and origin of 900 and 902.
We have to go a long way back, at least as far as Hermann, and
perhaps still farther, to Bentley, if we want to find a classical scholar
whose stature is comparable to that of Wilamowitz. Although it is
not in his work on Aeschylus that his greatness is most clearly
expressed, yet the effect of what he has done in this particular
province is enormous ; in this domain, too, ‘there is after all no getting
away from Wilamowitz’.2 The three books which are the major
representatives of his Aeschylean studies, the commentary on the
Choephoroe, the edition of the seven plays, and the Interpretationen,
" On this point cf. Headlam’s remarks in Cambridge Praelections, 1906, pp. 126 ff.
2 A. Y. Campbell, The Agamemnon of Aeschylus, p. xix.
59
 PROLEGOMENA
but especially the last two, show an extraordinary unevenness:
pieces of magnificent scholarship are intermixed with things which
are thoroughly bad. τὰ μὲν ὧν οὐ δύνανται νήπιοι κόσμωι φέρειν, ἀλλ᾽
ἀγαθοί, τὰ καλὰ τρέψαντες ἔξω. The bad things are not merely due to
our common heritage of human frailty or to the great man’s admi-
rable courage,’ but are to a large extent signs of haste, haste that has
so often spoiled a fine work of research by forcing its author to use
violence where caution was required. Many a sixth-form boy at a
good English school would be able to spot a few passages in which
Wilamowitz prints a text that is not Greek, and many more passages
where he assumes a meaning that is not warranted by his text.
Besides, it is generally known that in Wilamowitz’s publications
there are always several errors in detail, misquotations, incorrect refer-
ences, inconsistencies, and so forth. The critics who think that they
have done with his books when they have pointed out some such
weaknesses may be safely left to their amusements while we for our
part try to gauge the extent of our debt to his work.
When we consider the history of the παράδοσις of our text and the
principles on which any apparatus criticus to an edition of Aeschylus
has to be based, we see that here Wilamowitz has drawn the lines so
firmly and at the same time with such exemplary circumspection
that it is not likely that they should ever be altered, except in details
of minor consequence. This applies not only to the Oresteia but also
to the Byzantine triad, where the problems of the recensio are far
more complex. It is no exaggeration to say that Wilamowitz’s preface
and apparatus criticus make any earlier edition of Aeschylus look
hopelessly out of date. The better insight into the nature of the
παράδοσις and the improved classification of the manuscripts are not,
however, the only or even the chief reasons for regarding this edition
as a very great work. And the editor's successful emendations, some
of which are of the highest order, are inevitably fewer than those of
some earlier scholars. The most characteristic feature of Wilamo-
witz’s Aeschylus is rather the intense breath of life that permeates the
book throughout. Again and again a difficult passage is made under-
standable by some pithy observation. Moreover, many subsidiary
details which in other books would seem to have been raked together
from a rubbish-heap appear in Wilamowitz’s apparatus as living links
with the past of the Greek people in all its ramifications. These notes
contain a great deal of choice learning, but without any heaviness.
Nor are there here (nor, for that matter, in anything that Wilamo-
witz wrote) any departmental barriers. For him there was no such
thing as a watertight compartment of textual criticism, another of
historical grammar, another of metre, another of history of religion,
1 Hermann, Elementa doctrinae metricae, p. xvi, says of Bentley, ‘audentissimus ille,
quod periculum non formidaret, saepe, sed κεῖτο μέγας μεγαλωστί᾽,
60
 SOME EDITIONS AND COMMENTARIES
another of ancient law, and so forth. No single subsection of the
technique of research was allowed to get the better of the rest: they
had all to be subservient and to co-operate to one purpose only, the
adequate interpretation of the text in hand. It is a natural conse-
quence of this view that Wilamowitz had to make two important
additions to the customary apparatus criticus: in his edition of
Aeschylus the apparatus proper is preceded by a detailed analysis
of the metre of all the lyrics, and by a special section headed Actio.
This section greatly helps the reader to take his mind off the pages
of the printed book and direct it towards the orchestra at the foot
of the Acropolis. To many questions arising out of the text an answer
is given in Wilamowitz’s Interpretationen and, in the case of the
Oresteia, in the introductions to his translations; some of these dis-
cussions go deeper into the thought of certain passages in Aeschylus
than any previous attempts. There are, however, other problems,
both special and general, where we look in vain for some assistance
from Wilamowitz. But what he has done is more than enough to
encourage and enable his successors to fill some of the gaps he left.
If the serious study of Greek survives, as we hope it will, then
Wilamowitz’s work on Aeschylus will maintain its stimulating and
enlightening power for many generations to come.

61
 APPENDIX I
THE EVIDENCE FOR CASAUBON’S WORK ON AESCHYLUS
THE book which is our main source of information for Casaubon’s study
of Aeschylus does not seem to have been properly examined as yet,
although it has not remained entirely unnoticed. About the middle of
the eighteenth century a Dutch scholar, Simon de Vries, excerpted from
it the notes on the Eumenides for the benefit of F. L. Abresch,! who
acknowledges his debt in the preface to Book III of his Animadversiones
ad Aeschylum. Then Butler, in his edition of Aeschylus, vol. viii (Cambridge
1816), p. xxx f., touched upon it with a few rather misleading remarks ;?
it was also mentioned by Blomfield, Museum Criticum, ii (Cambridge
1826), 489 n. ı, and by H. W. Smyth, Harvard Studies in Class. Philol.
xliv. 54 n. 1.3 The book in question is a copy of Petrus Victorius’ edition
of Aeschylus (published by Henr. Stephanus in 1557), bound together with
Apollonius Rhodius (by the same printer, 1574) and Callimachus (by the
same printer, 1577). From Bishop Moore’s library it came into the Library
of the University of Cambridge (shelf-mark now Adv. b. 3. 3); cf. A Cata-
logue of Adversaria . . . in the Libr. of the Univ. of Cambridge (1864), p. 34,
Nn. vi. 5. As is shown by Is. Casaubon’s name written in his own hand
on the title-page of the volume and by numerous marginalia, the book
belonged to Casaubon.
Glancing over the margins of this Stephanus edition one cannot but
feel bewildered. At first sight it looks as though several hands have been
at work. Inks of different colour and pens of different shape have been
used. The letters vary greatly both in size and in shape: there are,
especially in the outer margins, brief notes written in large, bold strokes
and stretching out comfortably as if no consideration of space disturbed
the writer’s mind; on the other hand, the longer comments and the
translations show small timid letters and are crammed into a minimum
of space, since the better part of the margin had already been taken up
by those entries which were fortunate enough to come first. Some notes
are written horizontally, while others slope; some seem to have been
jotted down with a rushing pen, while others indicate meticulous care.
And yet there can be no doubt that all these marginalia, save one small
group to be discussed later on, are by the hand of Isaac Casaubon.*
1 Cf. p. 44 f.
2 He also excerpted from it a number of Casaubon's readings, which he incorporated
into the ‘Varr. lectt.’ printed in the second part of each volume, but these excerpts are
far from complete. In several cases suggestions of Casaubon with which suggestions
made by Stanley coincide are omitted by Butler, even if they are very clearly written
on a page from which Butler copied other notes of Casaubon's.
3 Smyth also mentions Casaubon's marginalia in the Canter edition of Aeschylus
preserved in the University Library of Cambridge (Catalogue of Adversaria, p. 26,
Nn. iv. 39, now Adv. e. 3. 2). These entries are uninteresting: they consist of brief
notes of a very elementary kind. Apparently Casaubon used this Canter text mainly
at an early stage of his study of Aeschylus.
4 In the Cambridge Catalogue of Adversaria, loc. cit., it is asserted that ‘some of the
notes are by Bishop Pearson'. This misstatement is based on a blunder by Butler
62
 CASAUBON’S WORK ΟΝ AESCHYLUS
Anyone who has carefully examined the large store of Casauboniana in
the Bodleian Library! must reach this conclusion. The document that
clinches the matter is Casaubon’s copy of the Basle edition (1529) of
Polybius, Bodl. MS Casaub. 19. Its abundantly rich marginalia furnish
in the forms of the letters as well as in the arrangement of the notes exact
parallels to all the strange features and puzzling irregularities of Casaubon’s
notes in his copy of the Stephanus text of Aeschylus. In the latter book
I have rigidly scrutinized any entry that did not at first sight show un-
mistakable signs of Casaubon’s hand, and invariably I found the authenti-
city of every detail confirmed by notes in the Polybius.
In the Cambridge Aeschylus (as I shall cali Casaubon’s copy of Vic-
torius’ edition) several stages in the growth of the marginalia can easily
be discerned. The earliest stage is represented by certain short notes in
the outer margins written with a broad, soft pen and executed in swift
strokes; they keep a fairly regular distance from the column of the text.
By far the greater part of these notes consists of ‘variae lectiones’, i.e.
discrepancies from Victorius’ text, often introduced by ‘.’, which is
sometimes written in full, viz. ‘lege’, or by ‘f.’ (fortasse) or ‘al.’ (alii).
The variants are partly readings which had been adopted by earlier
scholars? and partly conjectures of Casaubon himself. These notes are
all as succinct as possible, e.g. at Ag. 468 (ὑπερκότως) Casaubon underlines
orc and writes in the margin πως, at 834 (καρδίαν) he underlines av and
writes in the margin δίᾳ, at 1092 (πέδον) he underlines Sov and writes in
the margin ‘f. Sov’, at 1134 (θεσπιῳδὸν) ov is underlined and ‘f. dav’ written
in the margin, at Eum. 581 (κυρώσων) wy is underlined and in the margin
we find ‘v. [= vel] ρῶσαι v. pwoor’. To this group belong most of the notes
on the greater part of the Choephoroe (from p. 245 on) and those on the
Eumenides and Supplices. Occasionally the notes of this earliest class
contain references to parallel passages, e.g. p. 183 (Ag. 233) ‘ edpemd.
Id. vp. p. 396’ (Casaubon quotes Euripides after Canter's edition, Ant-
werp, 1571; so the reference is to the Messenger’s speech Iph. A. 1540 ff.),
p. 191 (Ag. 494 κλάδοις ἐλαίας) ‘Festus [Paulus Festi, p. 192 M.]. Oleagineis
coronis ministri triumphantium utebantur etc.’, p. 204 (Ag. 914) ‘Sic
εὐριπιδ. p. 383 [Iph. A. 1106]. λήδας γένεθλον, ἐν καλῷ etc.’, p. 216 (Ag. 1293)
‘Ita Soph. Aiace [833]’, p. 270 (Eum. 2) ‘Eurip. p. 437 (Iph. T. 1259 ff.]’,
p. 322 (Suppl. 440 f.) ‘similis locus p. 338 [Suppl. 944 f.]’. Finally there are
in this group a few isolated quotations from works of modern scholars,
e.g. p. 177 (Ag. 32) ‘vide Hadr. Junii Animaduersat. 2 c. 4’.
(edition of Aeschylus, vol. viii, p. xxxi), who rashly identified Casaubon’s copy of
Victorius’ edition (in Cambridge) with the copy of the same edition to which Needham
(for his Aeschylean studies cf. E. B. Ceadel, C.Q. xxxiv, 1940, 55 ff.) had referred as
'Rawl.' and in which he had rightly recognized Bishop Pearson's notes. For a full dis-
cussion of this important book in the Bodleian Library, MS Rawl. G. 193, see pp. 78 ff.
Further specimens of Pearson’s marginalia are to be found, e.g., in two of his books
now in the University Library of Cambridge, viz. his copy of Iustinus Martyr (Adv.
ἃ. 41. 4) and that of Photius’ Bibliotheca (Adv. a. 41. 3).
1 [ have also compared the very rich and interesting notes in Casaubon's copy of
Pamelius' edition of Tertullian (Paris 1583), which is in the Library of Corpus Christi
College, Oxford, and the marginalia in his copy of Nonnus in the same library.
2 Some of them are marked ‘P.’, which probably means Franciscus Portus, Casaubon's
teacher at the Academy of Geneva.

63
 APPENDIX I
The notes of the second group may conveniently be styled ‘headings’.
They are written with a hard and pointed pen in much larger strokes
than those of group I; indeed, the letters are far larger than those of any
other notes in this book. These headings were entered later than the
marginalia of group I: see, e.g., p. zor, where the second heading (on
Ag. 834 ff.), ‘Inuidiae descriptio’, has been pushed to the right-hand half
of the outer margin because δίᾳ, the variant added to 834 καρδίαν (see
above), had already occupied the left-hand half; higher above on the
same page the heading (on 826) ‘Plejadibus occidentibus captum Ilium’,
with nothing in its way, fills the whole width of the margin. I now quote
some specimens of these headings. P. 35 on Prom. 547 f. ‘Homines ἐσό-
vetpot, p. 44 on 709 ‘Nomades Scythae’, on 716 'Chalybes', p. 45 on 726
‘Salmydessia gnathus’, on 730 ‘Bosporus Cimmericus’, p. 47 on 767 ff.
‘Iupiter throno pellendus, non per uxorem, sed per filium ex ea natum,
patre potentiorem’, on 772 ‘De eo qui Prometheum sit liberaturus. Sed
videtur Aeschylus duas fabulas in unam confundere. Nec cohaeret haec
res’ [this is impressive; cf. Wilamowitz, Interpr. p. 132f.], p. 53
on 871 ‘Hercules ex genere Epaphi’, p. 55 on 907 ‘Hinc incipit ἡ τῆς
tpaywodias Avots’, p. 65 (ὑπόθεσις of Septem) ‘Laius et Iocaste', then ‘Laius
ἀρρενοφθορίας author’, then ‘Oedipus’, then ‘Polybus’, etc., p. 100 on 592
‘Amphiarai prudentia’, p. 133 on Pers. τόρ ‘Oculus domus’, p. 154 on 611
‘Xodv descriptio', on 616 ‘Sacra ad eliciendam Darii umbram', p. 155 on
63) "Preces ad deos inferos, ut Darium remittant', p. 197 on Ag. 650
‘Naufragium’, p. 219 on 1387 'alludit ad poculum Jouis Zwrfpos', p. 225
on 1583 'Atreus'. Some such headings have gone to the upper margin,
e.g. pp. 31-3 "Promethei inuenta’, pp. 85 and 86 (Septem) ' Addcews
mala’, some to the lower margin, e.g. p. 1o2 (referring to Sept. 623)
᾿ποδῶκες ὄμμα mira loc.’, p. 110 (to Sept. 784) 'Oculi xpecoadrexvor’. Some-
times it is the scholia that give rise to such headings. Thus on p. 52 we
find written against the scholia in the typical letters of this group a list
(numbered by Arabic numerals) of the descendants of Epaphus, a few
lines farther down (referring to the words ναῦν... τὴν κληθεῖσαν πεντη-
κόντορον) ‘Nauis πεντηκόντορος᾽, and still farther down ‘Danai filiae
Aegypti filiis nubunt’. On p. 127, using the same type of letters, Casaubon
adds to the scholion (Pers. 41) ἁβροδίαιτοι δὲ οὗτοι the note 'Lydorum
luxus. No. [= nota]', similarly on p. 168 to Schol. Pers. 906 (909 Weckl.)
he adds in bold strokes ᾿Σημείωσαι Dio criticus’,! on p. 193 to Schol. Ag.
532 ᾿Σημείωσαι οὔτε repetendum ex seq. Sic apud Sophoclem pag. 38’
(the reference is to Schol. S. Aj. 628, printed on p. 38 of H. Stephanus’
edition [1568] of Sophocles), on p. 241 to Schol. Cho. 325 (324 Weckl.)
᾿'Σημείωσαι vocabula certa certis poetis propria’. Sometimes we find,
written in the same bold strokes, the note 'prouerb.' or 'prou.' or 'pro.',
e.g. on Ag. 969 ff., 1050 f., 1163, 1172 (which he completely misunderstood),
1233, 1624, 1668.
Occasionally among the notes of this group we meet with an observation
on the style of a passage, e.g. p. 226 on Ag. 1630-2, where in the text ἦγε
and ἄξῃ are underlined, and a marginal note in huge letters testifies that
1 Unfortunately this Δίων whom Casaubon was so glad to meet in Victorius' text is
but the ghost of a man: the real reading in the scholion is δίχα,
64
 CASAUBON’S WORK ON AESCHYLUS
Casaubon thought this ‘Elegantiss.’ But far more characteristic of this
profoundly religious scholar are such ‘headings’ as these: p. 8 on Prom.
49 f. ‘Deus solus liber’, p. 153 on Schol. Pers. 602 (605 Weckl.), where the
Homeric passage σ 136 f. is quoted, ‘No. quod verissimum est’, p. 181 on
Ag. 163 f. ‘Egregie dictum: Ad tranquillitatem unus patet portus, in Deo
acquiescere’, p. 241 on Cho. 325 f. ‘Animus post cineres viuit’, p. 320 on
Suppl. 381 ff. ‘Deus’. His moralizing outlook appears in his note on p. 203
on Ag. 887 (the large letters are those typical of the ‘headings’) ‘Mirae
impurissimae mulieris blanditiae’.
There follows a third group (perhaps it would be more correct to speak
of several groups) consisting of miscellaneous notes, both critical and
explanatory, written with a pen similar to that of group II, but in much
smaller letters. It is to this class that the bulk of the marginalia in the
Cambridge Aeschylus belong. The arrangement of these notes in the margins
shows that they are later than those of group II. By far therichest comments
of this type are appended to the Agamemnon ; next, but at a considerable
distance, come the notes on the Septem and Persae. These entries are
evidence of the more intense work which Casaubon devoted to Aeschylus
at a later stage. It will be sufficient here to illustrate the whole group by
one instance which is again characteristic of Casaubon’s outlook: p. 54
on Prom. 887 ff. ᾿Σημείωσαι. Apparet in multis Aeschylum non fuisse
ἀμύητον librorum philosophorum quorum dogmata et sententias saepe
perstringit ut pag. 181’; there follows a quotation of Horace, Odes 4. 11.
29 ff. The reference ‘pag. 181’ is to Casaubon’s earlier note on Ag. 163
οὐκ ἔχω προσεικάσαι κτλ., ‘Egregie dictum’, etc. (see above).
Finally, Casaubon added Latin prose translations of certain sections
which present particular difficulties, viz. a number of choruses and other
lyrics. These translations are almost always headed by the formula—
common also at the head of the excerpts in Casaubon's note-books—«.8.
(σὺν θεῷ). The bulk of them are written in the upper and the lower margins ;
sometimes there is an overflow from the upper margin into the outer or
the inner margin. On p. 179 the translation of κύριός εἰμι θροεῖν κτλ. is
crammed into the narrow inner margin, since the lower margin was already
filled with the long note 'Sequentia in hoc choro sunt difficillima’, etc.
(belonging to group III). There are more instances which show that the
translation was added after the explanatory notes had been written. l'or
example, on p. 206 the translation covers the whole lower margin, but its
first seven lines stop a few inches before the edge, for there the space is
occupied by the entry ‘Hipp. σφαλερὸν ἡ ἄκρα εὐεξία᾽ (referring to Ag. roor),
which is written in the huge strokes of the 'headings'. Again on p. 223 the
beginnings of the lines of the translations are twice pushed away from
their natural place towards the right-hand edge, in the first case by a
critical sign (a cross) put against the lacuna marked in Victorius' text
after 1498 ἐπιλεχθὴς (as printed here), and in the second case by the
variant oövros (on the same line with, and referring to, 1532 mírvovros),
which belongs to group I, and the paraphrase on top of it, ‘quo vertam
curam meam ut prospera sit', which belongs to group III. The continuous
pieces of the Agamemnon which Casaubon translated in this book are
the following: 104-257, 367-487, 681-809, 975-1034, 1448-61, 1468-80,
4872-1 65 F
 APPENDIX I
1497-1512, 1521-36, 1567-76. Of the Septem he treated in the same way
78-107, 288-335, 720-91, 832-53, 915-25, of the Choephoroe 22-83, 152-63,
306-71, 380-414 (κλυούσῃ). There are no translations of Prom., Pers., Eum.,
Suppl. No doubt these versions are the result of a carefully planned and
sustained effort. They superseded former attempts on a small scale such
as we find, e.g., on p. 189, where the continuous translation of Ag. 378 ff.,
written in a very small script in the upper and the inner margin, was
preceded by a translation of 381-9 (σίνος), written in much bigger letters
in the lower margin.
One or two general inferences may be drawn from the story of Casaubon’s
Aeschylean studies as told in the marginalia of his copy of Victorius’
edition. It is obvious that at the earlier stages of his occupation with
Aeschylus he tried to obtain as far as possible a correct text and, moreover,
to impress upon his memory the main contents of the plays and especially
passages of a strong religious or moral tone, and that, as was to be ex-
pected of a scholar of his rank, he also noticed many points of historical,
mythological, antiquarian, grammatical, etc., interest both in the text
and in the scholia. It was not, however, until a later stage that he began
seriously to grapple with the particular difficulties of these thorny poems,
to examine in detail possible solutions of an enigmatic passage, and, last
of all, ἴο embark upon the arduous task of translating the cantica of three
of the tragedies. Another fact emerges no less clearly from these notes:
as time went on, Casaubon concentrated more and more on the Agamemnon.
This is borne out by a comparison of his notes on this play with those on
all the others, and is amply confirmed by the observation that the trans-
lations of the lyrics in the Agamemnon are nearly complete (only those of
the Cassandra scene and the short pieces 1407~11 and 1426-30 are lacking
here), whereas the cantica of the play that in this respect comes next, the
Septem, show considerable gaps, and the translation of the lyrics of the
Choephoroe has not advanced beyond a little more than a third of the play.
Thus far I have said nothing about the one set of notes which does not
come from Casaubon himself.” The book contains about forty variae
lectiones written in a hand which in its neatness is as different as can be
from Casaubon’s ‘trés mauvaise lettre grecque’.? All these notes are marked
by an 'S.' at their end, with the one exception of that on p. 98 on Sept. 560
(the note is εἴσωθεν ἔξω), which is undoubtedly by the same hand. Only
one of them is of any length, viz. that at the bottom of p. 228 (prologue
of Cho.), where the quotation in Ar. Frogs 1126-8 and 1172 f. is written
out. ‘S.’ stands for Scaliger; the notes are transcribed from, and are a
! Short pieces of translation of the lyrics of the Cassandra scene (of which there is
no continuous translation in this copy) are written down in the outer margin, or, if
the room there was occupied by earlier notes, in the inner margin. Their script is that
of group ΠῚ. Every now and then we also find translations of short pieces of dialogue
(e.g. Ag. 10-15, 1180-5), which the forms of the letters and the places where they are
written show to belong to the same group.
2 Apart from the ‘S.’ notes discussed in the following paragraph I have found only
one entry of which I am certain that it is not in Casaubon’s hand: in the margin of the
ὑπόθεσις of the Ag. (p. 175), 1. 3 from the bottom, a fine firm hand has written ‘ ὀγδοη-
κοστῇ legendum’, and the words in the text to which this refers, εἰκοστῇ ὀγδόῃ, are
underlined in the same ink.
3 Scaliger, quoted by Mark Pattison, Casaubon, and ed., 184 n. 1.
66
 CASAUBON’S WORK ON AESCHYLUS
selection of, the notes which Scaliger entered himself into his copy of
the Stephanus edition (1557) of Aeschylus, now in the Library of Leyden
(756 D 21).! They contain not only Scaliger’s own conjectures but also
readings of other scholars (Auratus, Canter, etc.) of which he approved.
In the first two-thirds of the book the S. notes are scanty: there is only
one on Prom., two on Sept., none on Pers., one at the beginning of Ag.
(p. 179) ;and only from p. 227 on (Ag. 1639) do they become more numerous.
The S. notes are later than group I of Casaubon’s own entries, to which
in several cases they refer (see e.g. p. 319 on Suppl. 349, where after
Casaubon's note μὲ σὰν ἱκέτιν we read ‘Sic S.) ; moreover, there are cases
in which an S. note is placed too high or too low because its natural place
was already occupied by one of Casaubon’s variants. As regards the
chronological relation between the S. notes and groups II and III of
Casaubon’s own entries, I have no criterion to fix it. But it is absolutely
certain that the S. notes were entered into this book before Casaubon
entered into it his continuous translation of certain lyrical pieces. For
on p. 84 (Sept. 288 ff.) Casaubon’s translation in the inner margin turns
in a curve round the note ‘reipa S.’ (this refers to 292 δυσευνήτορας, where
ropas is underlined in the same ink). It is clear, then, that Casaubon, after
using his Stephanus text of Aeschylus for some time, but before the stage
of his most intensive work on it, had a number of Scaliger’s readings
entered in his own copy, which he then continued to use. That Casaubon
was in the habit of drawing in this manner on the resources of his illus-
trious friend appears e.g. from his letter to Scaliger written at Mont-
pellier, 7 July 1597 (Epist. cxliii in Almeloveen’s edition) : ‘Accepi tandem
quas ad me . . . dedisti literas . . . et cum illis tuas in aliquot veteres
Scriptores notas. Beasti me isto munere . . . animum etiam adjecisti, ut
jam suscipere editionem eorum scriptorum, ad quos illustrandos me
hortaris, non verear.'
In the Bodleian Library I have found the link between the S. notes in
the Cambridge Aeschylus and Scaliger's original marginalia in his copy
of Aeschylus (in Leyden). The bulk of Bodl. MS. Add. A. 176 (Summary
Catalogue of Western Manuscripts, vol. v, no. 24763) consists of excerpts
and copies (some of them transcribed from notes collected by Scaliger,
others made for Scaliger’s use) in the hand of Charles Labbé,? who signed
several of them at the end (fol. 119 recto 'Carolus Labbaeus', 143 verso
'Karolus Labbaeus', 299 recto ‘Caroli Labbaei Biturig.', 342 recto ‘Carolus
Labbaeus Bitur.’). Labbé, a pupil of Scaliger's, was often employed by his
master as an amanuensis and was very much liked by him ;? the friendship
which Casaubon felt for the able young man is manifest in the tone of the
many letters he wrote him.* Labbé's hand looks at first sight very much
like Scaliger's own hand; but it is prettier, less firm, and has a tinge of
1 J have photostats of pp. 165-229, i.e. the whole of the Ag. with the end of the
Pers. and the beginning of the Cho.
2 Born in 1582; cf. Didot's Nouvelle biographie générale, xxviii (1859), 342.
3 Cf. Scaligerana 22, p. 134, 'escrit fort bien en Grec, c'est un honneste jeune homme
docte et infatigable', quoted by Mark Pattison, Casaubon, and ed., p. 183. For the im-
portance of Labbé's collection of the Glossaria, in the publication of which he was
greatly assisted by Scaliger, cf. Goetz, RE vii. 1461 f. and Corpus Gloss. Lat. i (1923), 257.
+ Cf. Isaaci Casauboni Zpistolae ed. Almeloveen (1709), Index I.

67
 APPENDIX I
professional calligraphy. A detailed examination puts it beyond doubt
that the hand which entered the S. notes in the Cambridge Aeschylus is
Labbé’s. Moreover, the Oxford volume of miscellaneous excerpts by
Charles Labbé provides direct proof that it was he who transcribed from
Scaliger’s copy of Aeschylus those notes which, with 5. added to them,
are found in the margins of the Cambridge book. Fol. 205 recto of the
Oxford convolute contains under the heading ‘Aeschilus’ a list of readings
which agree in every detail with the S. notes in the Cambridge Aeschylus
(including the note on Sept. 560 εἴσωθεν ἔξω, where, as I have said, the
S. is inadvertently omitted there). All the S. notes of the Cambridge book
recur on this one page of the Labbé volume (here there is of course no 'S.'
appended to them), with the sole exception of the five notes taken from
the last few pages, viz. p. 336 (four notes) and p. 341. When Labbé copied
Scaliger’s notes, he did not follow the order of the pages of the text of
Aeschylus, but went to and fro in an odd way so that we find, e.g., after
a note from p. 312 one from p. 311, then one from 308, one from 303, one
from 302, one from 300, etc. Apparently he skipped over the pages of the
book, turning forward and backward, and only gradually made up his
mind as to what to copy and what not. For it is only a small portion of
Scaliger’s own notes that were transferred from the margins of the Leyden
book first to the Oxford leaf and then to the margins of the Cambridge
book. We may sometimes doubt the wisdom of Labbé's selection: e.g. he
omitted at Ag. 134 the excellent οἴκτῳ ; consequently Casaubon did not
know of it. Nor is Labbé’s accuracy that of a perfect copyist. In his
way of punctuating and of putting the accents he shows himself far less
careful than either Scaliger or Casaubon (in the S. notes of the Cambridge
Aeschylus there are a few more mistakes of this kind in addition to those
in Labbé’s first copy), and at Ag. 103 (p. 179) he wrote both on the Oxford
leaf and in the Cambridge Aeschylus λυπησίῴφρονας, which does not con-
strue (Scaliger’s own note is λυπησίφρονα).
When Casaubon had sent his copy of Aeschylus to Labbé in order to
have Scaliger’s notes entered, Labbé took from Casaubon something in
return, as we learn from his notes in the Oxford book. On the back of the
page containing the excerpts from Scaliger’s notes on Aeschylus, i.e. on
fol. 205 verso, we find the headings ‘Apollonius. H. Steph.’ and (on the
following line) ‘Casaub.’, and underneath many excerpts from Casaubon’s
notes in the margins of his text of Apollonius, i.e. the text which is bound
together with his Aeschylus (the ‘Cambridge Aeschylus’). The page that
precedes the excerpts from Scaliger’s notes on Aeschylus, i.e. fol. 204
verso, contains a few excerpts from Casaubon’s notes on Callimachus,
under the heading ‘Casaub. Callim.’: they agree completely with Casau-
bon’s notes in the margins of his Callimachus, which, like the Apollonius,
is bound together with his Aeschylus. It is therefore practically certain
that, though the present binding of Cambr. Adv. b. 3. 3 is modern, the
three books, Apollonius, Callimachus, and Aeschylus, when they formed
part of Casaubon’s library, were bound in one volume as they are now.
The exchange of marginal notes such as we see it here was a natural
expedient at a time when no classical periodicals existed.
The story of the Cambridge Aeschylus finds its continuation in a manu-
68
 CASAUBON’S WORK ΟΝ AESCHYLUS
script in Paris, Bibl. Nat. MS grec 2791 (formerly codex Dupuy-Reg.
3330. 2).! The title-page of this book says ‘Aeschyli | Agamemnon | Isaaco
Casaubono | interprete. | Mpcx’; a later hand has added ‘cum ejusdem
notis et observationibus erud.’ The MS contains the Greek text of the
play with a Latin interlinear translation (always written above the Greek
line to which it belongs), and in the outer margins notes of a fairly
elementary character, although discrepancies from the textus receptus and
especially corrections by Casaubon are also mentioned (the latter usually
with 'ex emendat. Is. Casauboni' added to them). The hand is that of
a professional copyist aiming at calligraphy and at a pleasing arrangement
of the page. But his dainty, if sometimes stupid, performance has been
interfered with by large additions in a firm and scholarly hand.? This
second scribe has obviously no time to waste on considerations of beauty ;
all he seems to be interested in is the substance of solid learning. He treats
the work of his predecessor with utter ruthlessness, adds fresh notes
wherever he finds room for them, at the end of a note written by the
other man, or at the bottom of the page, or quite often even in the column
of the text; nor does he mind drawing a line across part of the page in
order to connect a note quite unmistakably with the word to which it
refers. Even so the pages did not provide nearly enough room for all the
additional matter ; so he wrote further notes, some of them of considerable
length, on loose sheets of various sizes, and these, together with a sheet
on which he wrote an ‘Argumentum Agamemnonis Aeschyli' and some
blank sheets, were later inserted in the book.?
At the end of the original manuscript (fol. 96) we read, in the hand of
the first scribe, first '"TEAOZ ΣῪΝ @EQ’ and then, underneath, 'Absoluit
Isaacus Casaubonus 5. Kal. Mart. 1610’. This colophon has tempted some
scholars to believe that the MS is an autograph of the great man.* So far
as the first scribe, the calligrapher, is concerned, any suggestion of his
being identical with Casaubon can be dismissed at once.5 Nor does the
hand of the writer of the additional notes, though apparently a scholarly
hand, show any likeness to that of Casaubon.
But although the additional notes in the Paris MS are not written in
Casaubon's own hand, there can be not the faintest doubt about their
authenticity. Every single piece of them bears in diction as well as in
matter the unmistakable stamp of Casaubon's craftsmanship. This general
! The Bibliothéque Nationale has provided me with excellent photostats of the
whole MS, which I intend to deposit in the Bodleian Library.
2 Vauvilliers, who in Notices et extraits des manuscripts de la bibliothèque du roi,
tome i, Paris 1787, pp. 324 ff., gave a description of the MS which, despite a few errors,
is very good, says (p. 326) that the ink of the additional notes is different; cf. also
Boissonade in Butler's edition of Aeschylus, vol. viii, p. xxxii, fatramento atro magis'.
3 When the whole was bound together, the order of the original pages was badly
disturbed.
+ Cf. e.g. 7. Franz, Des Aeschylos Oresteia, 310, and H. W. Smyth, Harvard Studies in
Class. Philol. xliv, 1933, 54.
5 Vauvilliers, op. cit., saw on internal grounds that the scribe could not be Casaubon ;
he was misrepresented by Boissonade, loc. cit., who made him assert the very reverse
of what he did say.
$ Àn expert in the scripts of scholars of the early 17th century might be able to
identify this very striking hand.

69
 APPENDIX I
impression is amply confirmed by a comparison of these notes with
Casaubon’s marginalia in the Cambridge Aeschylus. Most of the longer
comments in the Paris MS show that they are developed from notes in
the Cambridge book. Here I must content myself with quoting a very
few instances. Ag. 78: Cambr. has ‘Sic dixit Suppl. γυνὴ μον. οὐδὲν" οὐκ
ἔνεστ᾽ ἄρης᾽; Paris has ‘idem dixit in Supplicib. Aeschylus οὐκ ἔνεστι
dpns'. 120: Cambr. has 'Torquet ista doctos. Primum sciendum est
βλάπτειν construi cum gen. Hom. βλάπτειν κελεύθου", etc.; Paris has
‘Phrasis Homerica quum βλάβεσθαι construitur cum gen. ut dixit Homer.
βλάβεσθαι κελεύθου impediri quominus perficias iter’. 163: Cambr. has
‘Egr(egie) dictum: Ad tranquillitatem unus patet portus, in Deo acquies-
cere'; Paris has 'erumpit Poeta in diuinam sententiam de Deo et dicit
quod si mortales velint ab omni sollicitudine bona fide liberari in solo
Ioue debere quiescere' (there follows a long diatribe). In the scene 264 ff.
Casaubon had in the Cambridge book corrected the ascription of the
dialogue to the wrong persons: on one of the loose sheets (p. 204) of the
Paris MS we read 'grauiter errant qui hic nuncium cum clytemnestra
loquentem inducunt. nequaquam enim est nuncius sed chorus', etc. On
281 "Héóaucros “Iôns there is in the Paris MS (inserted sheet, fol. 99) a very
long discussion about φρυκτοί and the like (with an amusing digression
about speed of transport in the ancient world as compared with modern
standards), and in it the remark 'de his multa Polybius': in Cambr., at
the beginning of the Ag. (p. 176), we read ‘De phryctis veterum vide
Thucy. oxoA. p. 8o et omnino Polybium pag. 231 et 232’; the reference is
to Casaubon's copy of the Basle edition (1529) of Polybius, Bodl. MS
Casaub. 19, where on p. 231 f. he has in long marginal notes (with
sketched drawings) discussed the πυρσεῖαι in Polybius and Aeneas
Tacticus.
1f the true spirit of Casaubon's scholarship comes out more clearly in
the additional notes of the Paris MS, it would be wrong to deny that in
many cases the marginalia of the first scribe also contain faithful expres-
sions of Casaubon's views. For example, on 141 μαλερῶν ὄντων, where
Cambr. has 'Sic fere apud Thucyd. πλωϊμωτέρων ὄντων pag. 3', the Paris
MS has by the first hand 'uaAepóv ὄντων. casus absolutus ad occasionem
et modum declarandum, Thucydides lib. xr πλοιμότερωνϊ ὄντων: quia
tempus aptius esset ad nauigationem'. Similarly, where Cambr. has (on
855 ff.) ‘Conspectus primus Clytemnestrae et Agamemnonis. Hac tota
parte friget Aeschylus misere: Seneca euitauit hanc partem' and (belong-
ing to the same context, on 877) 'hic primum compellat Clytemn. Agamem-
nonem, quam frigide! quam prolixe! quam ἀπίθανον quod tamdiu silet
Agamemnon', the Paris MS has by the first hand 'videtur in hac parte
parum decorum seruasse Aeschilum [sic], cum tam frigide Clytemnestra
Agamemnonem maritum excipiat, hanc partem praetermisit Seneca in
Agamemnone'. This last note in the Paris MS should be taken together
with the additional note on 855 (loose sheet, last page of the Paris MS),
where the severe criticism of this scene (‘nihil quicquam frigidius et
languidius potest excogitari quam hic locus Aeschyli is laboured at
1 [n their handling of accents and breathing signs and also in their spelling the two
scribes of the Paris MS are less accurate than Casaubon himself.
70
 CASAUBON’S WORK ON AESCHYLUS
great length, and also with the ‘Argumentum Agamemnonis Aeschyli’ on
the first inserted leaf of the Paris MS, in the middle of which we find a
brief comparison of the Aeschylean play with Seneca’s Agamemno and
then ‘hic autem egregie Aeschylus decorum obseruat: nam antequam
Agamemnonem introducat redeuntem multos praemittit', etc."
The Greek text of the Paris MS is on the whole based on Victorius’
edition, but in the choruses Canter’s division of lines is sometimes taken
into account, without, however, any consistency. Where the word-order
is slightly involved, the reader is assisted by Greek letters marking the
‘natural’ order,? not only in the lyrics, but even in dialogue sentences
which seem to us quite simple. So we find, e.g. at 504 a above σε, B above
ἀφικόμην, y above φέγγει, ὃ above ἔτους, e above δεκάτῳ, and 6 above τῷδ᾽.
That Casaubon’s edition of the Agamemnon was meant to serve not only
scholars but also readers with only a very elementary knowledge of Greek
becomes still more apparent from the translation. In it every Greek word
has its Latin equivalent written above it. The resulting word-order is
downright barbarous, and the whole thing reminds us of the Greek cribs
composed in the eastern half of the Roman Empire for the benefit of
schoolboys who had to plod through the Aeneid.? Nevertheless it is per-
fectly clear that the crude version in the Paris MS is nothing but the
disiecta membra of Casaubon’s fine and scholarly translation. To see this
we have only to glance at any piece of the continuous Latin rendering
in the margins of the Cambridge Aeschylus. Take, e.g. (p. 180 lower
margin) 108 ff. ‘quomodo ales impetuosa, i.e. aquila, miserit cum hasta
(exercitu) poenarum exactore, in terram Teucrida, i. Troiam, imperium ~
geminum, i. duos imperatores Atridas, Graeciae pubem, i. principes
Graecae iuuentutis, ducatum consentientem, i.e. duces belli inter se
consentientes. Volucrium rex [corr. ex reges], Rex et ille qui aquilini
coloris erat, et ille qui posticam partem albam habebat, h.e. aquilae duae,
cum apparuissent prope palatium regibus nauium Atridis, à manu fulminis
iaculatrice (h.e. missu Jouis), depascentes sobolem leporinam, i.e. leporem,
valde faetam h.e. multos lepusculos in utero habentem, laesum in extremis
cursibus: h.e. post longum cursum tandem captum,* vescebantur in
sedibus undique splendentibus i.e. in ipso palatio'. And this is the inter-
linear version of the same passage in the Paris MS: quomodo achiuorum
geminum imperium i.e. fratres Atridas pubem graeciae ducatum con-
sentientem i.e. duces belli consentientes miserit cum hasta i.e. exercitu
poenarum exactori [sic] impetuosa ales i.e. aquila in teucridem terram
auium rex i.e. duae aquilae regibus nauium Atridis una quae erat fusca,
aquilini coloris et altera candida a parte postica cum apparuissent prope
τ Casaubon had learned a lesson at the court of the French king ; cf. e.g. his observa-
tion on the parodos (Paris MS, loose sheet, fol. 6) ‘nam erat e regia maiestate Clytem.
ut non statim admitterentur in domum [sc, the Elders], sed pauxillum ad fores ex-
pectarent’.
2 This system corresponds to the practice in the scholia of straightening out the
word-order of complicated passages under the heading ro é£$s.
3 Cf. eg. Pap. Rylands 478 (Catalogue of the Greek and Latin Papyri in the John
Rylands Library, vol. iii, 1938, pp. 78 ff.) with C. H. Roberts’s introduction.
4 The words ‘laesum . . . captum' are added in the left-hand margin, which probably
accounts for the change of gender.
71
 APPENDIX I
palatia a manu fulminis jaculatrice i.e. missu Jouis in undique splen-
dentibus sedibus i. in ipso palatio depascentes leporinam i.e. unam
leporem faetam multis lepusculis pascebantur sobolem quae laesa fuerat
in extremis cursibus 1.6. post longum cursum tandem captam’.
The dependence of the first scribe of the Paris MS on Casaubon’s entries
in the Cambridge Aeschylus is no less obvious in his text and marginal
notes than in his interlinear translation. The evidence is plentiful; a few
specimens will suffice. 627 (ἄχθος) : Cambr. has in the margin ‘f. ἔχθος.
Hesych. exp. πόνον᾽ ; Paris has ἄχθος crossed out in the text, and in the
margin ‘Exdos potius quam ἄχθος. ex emendato Is. Casaubon’. 680:
Cambr. has κλύων underlined in the text, and in the margin 'vóv scr.’;
Paris has κλύων underlined in the text, and in the margin ᾿κλυὼν. melius’,
with ‘Ex emendatione Is. Casauboni’ above. 682 ff.: Cambr. and Paris
both have brackets round the words μή τις... ἐν τύχᾳ νέμων. 696: Cambr.
has xeAoavrwv underlined, and in the margin '«éAcav, τοῦ 2.’; Paris has
κελσάντων crossed out, and in the margin '«éAcav, roû”, with ‘Ex emendat.
Is. Casauboni’ above. 697: Cambr. has ἀξιφύλλους underlined, and in the
margin ‘Sc. αὐξιῴ. vel defid.’ ; Paris has ἀξιφύλλους crossed out, and in the
margin ᾿ἀεξιφύλλους᾽, with ‘Ex emend. Is. Casauboni’ above.
Sometimes the first scribe of the Paris MS was not very intelligent in
the way in which he copied what he had before him in Casaubon’s book
(Cambr.). This may be illustrated by two instances. At 165 he first wrote
μάταν, which he found in Victorius’ (or, for that matter, any other) text,
but then changed the final v to s. The reason is that in Cambr. Casaubon
underlined μάταν and wrote in the margin ‘Aur. μάτας᾽, then left a small
blank and went on ‘optime μάταν explicat Hesy. ματαιότητα ut hic schol.’.
The copyist in his haste noticed only the first part of the marginal note
and wrongly concluded that Casaubon accepted Auratus’ conjecture. At
718 the copyist crossed out οὗτος ἀνὴρ and wrote in the margin 'ó πατὴρ ex
emendat. Is. Casauboni’. Of this 'emendation' Casaubon is quite innocent.
In Cambr. we see that he underlined οὗτος in the text, added a sign to
it, and repeated the sign in the margin with the note ‘6 πατήρ᾽. In other
words, since the theme of the preceding sentences is Paris, and since
the scholiast also refers the stanza 717 ff. to him, Casaubon thought that
by οὗτος ἀνήρ Priam was meant (cf. also his translation, ‘Huius pater
leonem aluit exitiosum aedibus’, etc.) ; the idea that he wanted to alter
the text is excluded by the form of his note.
Probably the copyist of the Paris text, who, as we shall see, must have
acted on Casaubon’s orders, did not depend solely on the Cambridge
Aeschylus but also had at his disposal other materials collected by
Casaubon! for his ‘interpretation’ of the Agamemnon, and especially a
Latin version of the pieces of which there is no translation in Cambr.
I should hesitate to infer from an interesting self-correction of the Paris
scribe at 931 that Cambr. was his only exemplar. In the Paris MS the
interlinear translation of μὴ παρὰ γνώμην was first given in the form ‘non
contra sententiam meam’, which was then deleted by the scribe himself
and supplanted by ‘ex animi tui sententia’. From a marginal note in
Cambr. we see that Casaubon translated 931 f. ‘Responde mihi ex animi
1 Some of them may still be lurking in some library.
72
 CASAUBON’S WORK ON AESCHYLUS
tui sententia. Scito me sincero animo responsurum’ (for the correctness of
this interpretation see my commentary). It is tempting to assume that
the Paris scribe, left in this section to fend for himself, first translated
931 as best he could (in misunderstanding the line he is in the company of
Paley, Wilamowitz, and other good Hellenists), and then, after glancing
at the margin of Cambr., adopted Casaubon’s translation. But it is equally
possible that, although he had a complete translation by Casaubon before
him, he first gave inadvertently the translation of παρὰ γνώμην which
seemed to be suggested by the immediately following ἐμοί, and only then
looked at his exemplar and realized his mistake.
Now I turn once more to the far more interesting part of the Paris MS,
the additional notes. All of them, from the briefest comments crammed
into the pages of the original book to the longest digressions written on
loose sheets, are the genuine product of Isaac Casaubon.! Whether they
are of an edifying or an antiquarian character, whether they deal with
religion or with grammar or with institutions of ancient life, they reveal
the powerful approach, the critical acumen, the unfailing sense of responsi-
bility, and the almost boundless learning of one of the world’s greatest
scholars. Nor does their author attempt to conceal his personality. I
select a few typical examples which show this attitude of the commentator
and at the same time illustrate his extreme caution. On 1639 he observes
‘Tos. Scal. legit ἄρχειν τὸ λοιπὸν; sed nostra lectio [i.e. the MS reading]
melior est’ ; on 103 λύπης φρένα ‘hypallage pro λύπην φρενός ; hypallage dura
et plane Aeschyli. vel legendum λύπην φρενός: sed tamen nihil volo
mutare’; on 1498 μηδ᾽ ἐπιλεχθὴς ‘hic locus est corruptus sed non ausim
emendare quia nonnulla verba hoc loco desunt ; quam autem potui horum
versuum commodiorem explicationem ego dedi’. Perhaps the most
remarkable of the notes of this type is the one (fol. 73) on 1267 ἔτ᾽ ἐς φθόρον
κτλ. : ‘hic versus non multum mihi placet, et quoquo modo vertam nulla
expositio mihi arridet, quare crediderim corruptum esse, sed non ausim
corrigere; quam autem potui commodiorem explicationem dare dedi.’
This is the great critic’s last word on a corrupt passage with which he had
struggled desperately for many years, perhaps decades, as is demon-
strated by the evidence in the Cambr. Aeschylus: there we find first a
cross put against the line; the size of this cross, its place in the margin,
and the ink in which it is written (ἀγαθὼ δ᾽ ἀμείψομαι in the text is under-
lined in the same ink) prove beyond doubt that it belongs to the earliest
group of Casaubon’s notes; to the cross there was later added, in the small
letters and the ink typical of group III, the note ‘non intelligo. fort. ir’
ἐς φθόρον' πεσόντ᾽ ἐγὼ δ᾽ ἀμ. ego proiecta respuam'.? As little as on ques-
tions of textual criticism does Casaubon allow himself to comment on
1 It is only in subordinate details that the zpsissima verba are occasionally overlaid
with the provisional phrasing of the copyist. For example, the note on 32 (loose sheet,
fol. 6) concludes ‘plura vide in Comment. Casaub. ad Sueton.’ The way in which
Casaubon himself would have given such a reference is shown e.g. by the end of his note
on 498 f. (fol. 24), ‘de his multa vide ad Theophrasti caract. cap. περὶ δυσχερείας"
(Casaubon in his commentary on Theophr. Char. 19. 8 quotes the Ag. passage).
2 In the margin of the Cambridge Aeschylus we find at 1639 'rö λοιπὸν S.’, with
πολιτῶν underlined in the text.
3 Tt was in consequence of this note that the first scribe of the Paris MS crossed out
73
 APPENDIX I
any point of interpretation with greater confidence than was justified
by his searching scrutiny ; sentences such as (fol. 18, on 408 drAnra τλᾶσα)
‘duarum explicationum quam volueris elige, nam ambae huic loco bene
conueniunt’ occur several times.
The tone, the contents, and the very size of the additional notes in the
Paris MS make it abundantly clear that what we have here is neither a
mass of ‘private’ notes nor mere materials for a commentary, but to all
intents and purposes Casaubon’s commentary on the Agamemnon in as
nearly final a form as it was possible for him to produce in the harassed
conditions of the last few years of his life. To begin with, why should
Casaubon have gone to the trouble of working out an elaborate ‘Argu-
mentum Agamemnonis Aeschyli’ (cf. p. 71), had not his intention been
to give the general reader an introduction which would enable him better
to understand and appreciate the play as a whole? It is with the same
object in view that the bulk of the more extensive notes is written.
Casaubon is always at pains to make it easy for the reader to grasp the
gist of an ode or a section of the dialogue and to prevent him from losing
his way amidst the difficulties of the detail. Introductory formulae such
as ‘sensus horum versuum hic est’ are common. At the beginning of the
somewhat enigmatic ode 975 ff. we find (fol. 55) the note ‘quia hic chorus
est paulo obscurior ideo verum sensum et veram interpretationem hoc
loco dabimus ut omnia sint clara et perspicua’. The considerable number
of French phrases, especially proverbial sayings and the like, with which
the supplementary notes in the Paris MS are interspersed provides
additional evidence of Casaubon’s eagerness to win the interest of the
general reader, who, though he could be expected to have sufficient Latin
to understand the commentary, would react more readily to a striking
illustration from his native tongue. See, e.g., at the end of the note on 73
ὑπολειφθέντες (fol. 8) 'λείπεται οὗτος ἐκείνου il vaut bien moins que l'autre’;
on 395 πρόστριμμα ‘macula ex attritu, comme quand lon passe aupres d’une
roüe de charrette, et qu'elle touche au manteau, l'ordure et l'impression
que laisse la roüe s'appelle proprement en grec πρόστριμμα᾽; on 709
μεταμανθάνουσα κτλ. (fol. 37) ‘gall. leurs ris sont tournés en pleurs’; on 934
τέλος (fol. 51; the note has been wrongly inserted by the copyist between
a note on 942 and one on 945), after a comment on the variety of meanings
of τέλος, the remark 'et multae sunt ap. Graecos voces quae ita latam
habent significationem quae pro variis locis varie quoque sunt accipiendae
ut prouerbio gallico possent appellari selles a tous cheuaux'; on 1033
ἐκτολυπεύσειν 'proprie hoc verbum significat deuider sa quenoüille’; on
1421 ff. (fol. 82) 'Gallice recte verteremus, ie vous aprendrai a estre sage,
aliter ie vous ferai bien soustenir' (here, as elsewhere, Casaubon shows
a fine instinct for the tinge of colloquialism in certain parts of the Aeschy-
lean dialogue); on 1668 ἐλπίδας σιτουμένους 'gallice viure d'esperances'.
Cf. further the additional notes on 131 olov, 160 Ζεύς, 456 βαρεῖα δ᾽ ἀστῶν
φάτις (fol. 22), 805 dm’ ἄκρας φρενός (fol. 44), 1088 πρὸς τὴν Arpeudwv (fol. 61).
It is part and parcel of this tendency towards 'popularization' (though
Casaubon’s insatiable appetite for minute points of Realien also plays
ἀγαθώ, which he had written in the text, and added the note 'éyó ex emend. Is. Casau-
boni’.

74
 CASAUBON’S WORK ON AESCHYLUS
a part in it) when he more than once goes out of his way to compare or
contrast ancient institutions and customs with what corresponds to them
in the modern world. One of these digressions is especially welcome since
it enables us to fix the time at which Casaubon wrote the notes which were
to be his last contribution to the study of Aeschylus. Commenting on 519
and discussing the meaning of θᾶκος he says (fol. 26) ‘d&kos etiam aliud
significat, nam θᾶκοι exponuntur latrinae publicae! ut in multis urbibus
videre est ut Geneuae et Londini'. In other words, Casaubon wrote the
notes copies of which are inserted in the margins and on the separate
sheets of the Paris MS some time after his arrival in London (end of October
1610), where he died on r2 July 1614.
The date thus established for Casaubon's supplementary notes, i.e. his
real commentary, on the Agamemnon will be seen in its full significance
when we combine with it the colophon at the end of the text in the Paris
MS, ‘Absoluit Isaacus Casaubonus 5. Kal. Mart. 1610’. Now at last we
are in a position to form a clear idea of thc story that lies behind our
documents.
As a very young professor at the Academy of Geneva? Casaubon planned
an edition of the whole of Aeschylus with a full commentary. This intention
he announced in two places? in his commentary on Strabo (published
1587), viz. p. 15 of the first edition (p. 18 of the Paris edition of 1620)
'Comparat eum [Homer] Strabo cum Sophocle et Euripide: tertium
Aeschylum potes his addere: qui caetera summus, mirabilem tamen
Geographiam habet : quod nos aliquando Deo dante ostendemus, quando
ilum poetam cum nostris Annotationibus edemus', and p. 87 (ro4),
dealing with the Πελασγικὸν "Apyos, 'Aeschylus autem ὁ φιλόμηρος (ut et
Soph.) hoc non ignorabat, quum fecit Regem Argiuum gloriantem suae
ditionis fines esse Strymonem et Pindum, ac Perraeborum Paeonumque
fines: quem locum [Suppl. 254 ff] paucissimi intelligunt. Nos autem si
Deus dederit, in nostra illius poetae editione explicabimus.’* Many years
later he obtained from Charles Labbé Scaliger's readings of a number of
passages (cf. p. 67f.); he would hardly have asked for them if by that
time he had given up the idea of editing Aeschylus. Of the sustained
efforts which he continued to devote to the poet the Cambridge book
provides full evidence. But that book also shows that, as time went on,

! For the evidence and a detailed discussion see Casaubon on Theophr. Char. 14. 5.
3 There is just the barest possibility that at Geneva Casaubon sometimes lectured
on Aeschylus; he certainly did not do so at Montpellier (for the subjects of his lectures
there see Mark Pattison, Isaac Casaubon, 2nd ed., 99 ff.). In any case his work on
Aeschylus at its maturer stage was not the outcome of, or connected with, any lectures
on the poet.
3 Quoted, on the advice of John Pearson, in Stanley's Preface (see p. 78 below).
* This plan of an edition of Aeschylus would seem to deserve at least a brief mention
in a biography of Casaubon. But the name of Aeschylus is not to be found in Mark
Pattison's Casaubon. 'This book is justly famous; it is brilliantly written and highly
instructive so far as it goes. It fails, however, to convey an adequate idea of the true
nature of Casaubon's scholarship and of the originality and greatness of his work.
There is very little in the book to show that Pattison was sufficiently familiar with even
the published editions and commentaries of Casaubon, to say nothing of the wealth
of unpublished materials in the Bodleian, the most important piece of which is the
annotated Polybius.

75
 APPENDIX I
Casaubon contributed less and less to the other plays and concentrated
more and more on the Agamemnon. Finally, as the horizon of his personal
life increasingly darkened and his health also began to give way, he seems,
presumably in the winter of 1609/10, to have resigned all hope of ever
completing an edition of Aeschylus on the scale on which he had planned
it in the buoyant days of his youth. After the wreck of his more ambitious
scheme he decided to salvage at least the Agamemnon. He therefore gave
one or two annotated copies of the text, and probably a draft of a complete
Latin translation and a collection of short explanatory notes, to an
amanuensis, whom he advised in general terms as to the manner in which
the work was to be executed. Apparently it was his intention to provide
not only for the experienced scholar but also for the reader who had little
Greek but was nevertheless keen to grope his way through the text of
Aeschylus. Hence the crude word-for-word translation, the notation that
was to help in disentangling the word-order, and the many notes of a rather
elementary character. However, features of deeper learning were by no
means excluded from the work at this stage. The colophon ‘absoluit
Isaacus Casaubonus’, etc., seems to imply that when Casaubon put his
materials into the hands of his amanuensis it was his intention that the
copy which he ordered him to produce should be the final form of his
‘interpretation’ (cf. the title-page) of the Agamemnon. And indeed the
Paris MS, leaving on one side the additional notes entered by the later
scribe, looks in every respect like a fair copy: it was in all probability
designed for the printer. But when Casaubon received it back he was not
satisfied. However, this was not the time to do anything about it. The
grave anxiety caused by the murder of Henry IV (May 1610), the restless-
ness preceding Casaubon’s emigration to England, then the difficulties of
making himself familiar with his new surroundings, the fresh obligations
inflicted on him by King James and others, all these were circumstances
utterly averse to the completion of a work which, however much he had it
at heart, seemed less urgent that the many demands of the day. Still, no
sooner had he somehow settled down in London than he returned to the
Agamemnon. At a period when all his time seemed occupied with rejecting
Baronius, fighting the king’s campaign against the Pope, plunging deeper
and deeper into all sorts of ecclesiastical writings, and complying with the
wishes of the Court and his many friends, the heroic scholar made it
possible to escape every now and then from the turmoil of his various
obligations and bring a serene and fresh mind to the study of the great
Attic masterpiece. No doubt in doing so he also satisfied his own religious
conscience. In this regard the latest form of his commentary bears witness
to the same spirit as several of his early notes in the Cambridge Aeschylus
(cf. p. 65). Sometimes he identifies, perhaps unconsciously, the Zeus of the
poet with his own Christian God,! e.g. fol. 8 on 65 ἐν προτελείοις 'sensus
autem horum versuum [60 ff.] hic est. Deus mittit Atridas contra Paridem
. . imponet tamen Deus Graecis et Troianis graues luctus’, etc. (in the
same paragraph he goes on to quote from the Bible the fate of Nineveh),
and fol. 10 on 104 ff. ‘ex hoc autem loco possumus cognoscere quam optime
de Deo sentirent Pagani qui omnem in eo spem et fiduciam habebant" ; on
1 For the similar outlock of Demetrius Triclinius see vol. ii, p. 102, n. 1.
76
 CASAUBON’S WORK ON AESCHYLUS
other occasions he stresses the Greek mode of expression, as fol. 5r on 928
θεοῦ ‘notandum est hoc loco Agamemnonem potius dicere θεοῦ quam
θεῶν. veteres enim qui de Diis male sentiebant tamen instinctu quodam
naturae in rebus maioris momenti Deum potius dicebant quam Deos.
hoc obserua’. But this keen interest in the religious thought of the poet
does not blind Casaubon’s eye to other aspects of the play or to the many
points of scholarship which its interpretation raises. Nor does he waver
in his determination to satisfy alike the demands of the interested amateur
and those of the professional scholar. Of this twofold purpose of the supple-
mentary notes there is evidence on almost every page. Casaubon’s vigour
seems to be quite unbroken, his mastery of innumerable details as un-
failing as ever. At the same time (as has been shown above) he spares no
words to lift his reader above the mere detail, to guide him safely through
the maze of many a difficult passage and direct his attention towards the
important issues. Perhaps the most noticeable feature of this commentary
is the historical sense (of course within the inevitable limitations of
Casaubon’s time) which permeates the whole and influences the treatment
not only of antiquities of all kinds but even of textual criticism. We late-
comers, with all the facilities of modern libraries at our disposal, may well
pause for a moment to admire the resourcefulness of Casaubon when
(fol. 40 on 767 φάους κότον) he wants to illustrate in truly historical style
what he regards as a peculiarity of ancient scribes, i.e. their apparent
economizing of letters by writing only a single letter if one word ended
with the same letter with which the next word began. As he has no
facsimiles of uncial manuscripts to which he can refer, he does the next
best thing and quotes in support of his assertion the practice of the scribes
of the Florentine Digest.
Some of Casaubon’s best emendations in the text of this play do not
reappear in the Paris MS. We do not find there 69 ὑποκαίωνἦ (on the con-
trary, the first scribe has appended to ὑποκλαίων the note ‘xAaiev minus
quam Saxpvew’); at 1411 there is no trace of Casaubon's fine correction
ἀπόπολις (ἄπολις is not only in the text but also in the lemma of the note
on a loose leaf, fol. 82) ; 1547 ἐπιτύμβιος alvos is left unaltered. It is possible
that these omissions are due to mere oversight, but it seems equally
possible that Casaubon, like other great scholars, grew more sceptical as
the years went by and consequently dismissed in later life some excellent
suggestions of his more daring period.
I do not know whether Casaubon carried the completion of his com-
mentary to the point where he did not want to make any further additions.
In any case, the supplementary notes in the Paris MS go right to the end
of the play (although they are comparatively short towards the end), and
the last of the inserted leaves (fol. 92) contains a note on 1587 προστρόπαιος
and another note on 1595 ἔθρυπτ᾽ ἄνωθεν. All the additional notes in the
Paris MS are undoubtedly copied from notes written by Casaubon.

1 Casaubon presumably used the Digest ('Pandectae Florentinae' as he quotes it)

in the edition of Taurellus (Florence 1553), who 'digesta in omnibus, etiam ortho-
graphicis, ad Florentinum codicem repraesentauit' (Mommsen, Preface to his edition,
p. xvii f.).
2 Butler recovered it from the margin of the Cambridge Aeschylus.

77
 APPENDIX I
Whether it was he himself who had these notes copied and inserted into
the Paris MS, or whether the copy was made after his death, it is impossible
to say so long as the hand of the second scribe has not been identified
(cf. p. 69 n. 6).

APPENDIX II
JOHN PEARSON’S SHARE IN STANLEY’S AESCHYLUS
THE MS Rawlinson 6. 193 in the Bodleian Library was briefly described by
Needham (repeated in Butler’s Aeschylus, vol. viii, p. xxx). It is a copy
of Victorius’ Aeschylus (published by H. Stephanus, 1557). Its fly-leaves
and margins are covered with notes, some of them very long, in the virile
and beautiful hand of John Pearson. Among the authors quoted in these
notes Hesychius, for whom Pearson did more than anyone before him,
is very prominent, nor is there a lack of other references, both to classical
and patristic texts, which bear the hall-mark of the great scholar and
theologian. Special attention is paid to the scholia, not only those on
Aeschylus.' Suggestions of readings different from those in Victorius’ text
are numerous. Severalof them have ‘Jac.’ added to them, i.e. they are
conjectures of, or readings recommended by, that erratic and original
scholar Henry Jacob (1608-52) ;? Pearson had presumably copied them
from the margins of a text belonging to Jacob.?
When I opened this book I saw to my amazement that whole para-
graphs of Stanley’s commentary, and as a rule those which contain the
most remarkable pieces of real erudition, are nothing but copies of
Pearson’s marginal notes, with hardly a word altered. One has not to go
far to find abundant proof of this. On the back of the title-page Pearson
wrote ‘Post Victorium et Stephanum Aeschylum edere et illustrare voluit
Is. Casaub. ipso teste ad Strab. p. 18’; in the last paragraph but one of
Stanley’s preface we read ‘Aeschylum etiam, post Victorium et Stephanum,
edere et illustrare voluit Vir incomparabilis /saacus Casaubonus, ipso
teste ad Strabonem paginis 18, et 104’ (the latter passage could easily be
added with the help of the ‘Index Auctorum" in the then current edition of
Casaubon's Strabo, Paris 1620). When we pass on to the BIOZ AIZXYAOY
in Victorius' edition, we see that the greater part of the learned material
in Stanley's notes on this chapter comes from Pearson. Nor does the
picture change when we reach the first play, Prom. There, right at the
beginning, Pearson entered in the lower right-hand corner of p. 4 the note
1 As I glanced over the pages of Pearson's Exposition of the Creed, my eye was struck
in Article II, ‘Our Lord’, by a huge footnote (p. 295 f. of the first edition of 1659) on
κύριος, κύρειν, κυρεῖν. It begins, characteristically, with a reference to Hesychius, and
passes on to a scholiast on Sophocles; then comes a quotation from A. Prom., two from
the scholiasts (differentiated according to their relative chronology) on Sept., one from
a scholion on Pers., and several from passages of Euripides with the scholia on them.
2 See Dictionary of National Biography, xxix. 118.
3 Stanley's note on Ag. 206 (214 St.) shows that he, too, had access to a copy of Canter's
Aeschylus annotated by Jacob.
78
 JOHN PEARSON’S SHARE IN STANLEY’S AESCHYLUS
‘Prometheum caput Jovis secuisse tradunt aliqui cum Minerva nasceretur.
Pind. Schol. 64. b’; in the inner margin of p. 5, with a sign referring to
Prom. 7, ‘non Vulcani solum sed et Minervae. Plato’; and in the lower
margin of the same page, referring to the same line, ‘Prometheum ignis
inventorem negabant Argivi, et Phoroneum fuisse asserebant. Paus.
p. 119.’ These three separate notes are nicely run into one in Stanley’s
commentary on Prom. 7, the only change being the replacement of
Pearson’s ‘aliqui’ by ‘sunt vero qui’ in order to attach the sentence to
the preceding one. To select at random a few similar instances from the
commentary on the same play: Stanley’s note on 323 down to the quota-
tion ‘Act. Apost. 9. 5’ is an exact copy of Pearson’s note; Stanley’s long
note on 362 ἐφεψαλώθη is taken in its entirety, and without any change,
from Pearson; Stanley’s whole learned comment on 480 (479 St.) οὐ
χριστὸν οὐδὲ πιστὸν is patched together from four different notes of
Pearson’s; the note on 710 (709 St.), with the quotation from Tertullian,
goes back to Pearson; the last four lines of the note on 793 (792 St.), with
the emendation of the scholion and the references to out-of-the-way
passages, belong to Pearson; and so does the note on 853 (852 St.) πεντη-
κοντόπαις. Nor are things different in the other plays. I will take a few
instances from the commentary on the Agamemnon. In the notes on the
dramatis personae compare Stanley’s remarks on the ‘Nuntius’ and on
the φύλαξ with Pearson’s notes, ‘hic nulla Nuntii mentio, ergo qui Arg.
scripsit N. non agnovit’ and ‘Hom. Οδυσσ. δ. in descriptione reditus
Agamemnonis. τὸν δ᾽ dp’ ἀπὸ σκοπιῆς . . . ποιμένι λαῶν". From the com-
mentary on the text of the play I will give the following selection ; I quote
Pearson’s notes, with which the corresponding notes of Stanley’s should
be compared. On 3: 'Hesy. legebat, opinor corrupte, ἀγρίαθεν. Aypiadev,
ἀνέκαθεν. Αἰσχύλος Ἀγαμέμνονι. Ita alibi idem Ayradev. ἀνέκαθεν. On 17:
'v. Eurip. Med. Eör’ ἀντίμολπον . . . κωκυτὸν. v. Schol.’ On 53: 'Hesy. de
pullis nidos servantibus, ego potius de vulturibus’ (neither here nor else-
where did Stanley blush to take over Pearson’s ‘ego’). On 79: “ἐσχατόγηρως.
Eus. Hist. 116. a. a.'. On 87: ‘v. Hesych. Eur. Bacchis μαινάδας θυοσκόους.᾽
On τοῦ: ‘mallem Bas. sed Aristoph. ἦβαν dixit. On 137 (140 St.): ‘ut
αὔτανδρον. σὺν αὐτοῖς τοῖς ἀνδράσι. Hesy. αὐτόπρεμνος.᾽ On 141 (145 St.):
“Αἰσχύλος ἐν Ἀγαμέμνονι. . . δρόσους κέκληκε. Etym. in "Epoo/ (in the
upper margin) and (on the scholion, where the text printed by Victorius
is χωρὶς δ᾽ ad τέρσαι) “1. χωρὶς δ᾽ αὖτ᾽ ἔρσαι. vel αὖθ᾽ ἔρσαι. Odys. « 222.
v. Etym.' On 234 (243 St.): ‘Hes. Adpönv. ἄνω, ἢ φοράδην. male scribitur
Aepdyv.’ On 304 (312 St.): 'ignis mandatum. i. ignem quem ut transmit-
terent in mandatu habebant. p.p. λαμπαδηφόρων νόμοι.᾽ On 513 (522 St.):
‘Suppl. 196 [189]’, on which passage Pearson gives a note referring to the
other instances of ἀγώνιοι θεοί in Suppl. and to Ag. 513. On 884 (893 St.):
‘consilium projectum iniret, sc. occidendi Orestis. Jacta est alea.’ On 908
(918 St.) : ‘Reg. 1x, 13. Matth. xxi, 8.’ On 942 (951 St.) : ‘vel τῆςδε legendum.
vel a δήρη. δήριος. pertinax.’ On 1061 (1070 St.): ‘Inde fortasse Barginus.
Lex. Graecol. προσφώνησις... Glossarium Vetus. Barginae. peregrinae'
(with this compare the second half of Stanley's note). On 1344 (1356 St.) :
"Peccat Aeschylus cum apud eum Agamemnon tanta celeritate et occiditur
1 The omissions here and in the following quotations are mine.

79
 APPENDIX II
et tumulatur ut actori vix respirandi tempus detur. Voss. Inst. Poet.
1. 1 p. 22.’ On 1382 (1391 St.) : ‘kara δὴ τοὺς τραγικοὺς . . . ἔνδυσιν τραχήλου.
Schol. Hom. JA. a. 7.” On 1462 (1471 St.): ‘haec ad Strophen primam
referenda’ (cf. also Pearson on 1464, ‘ista ad Anapaest. immediate
anteced.’). On 1476 (1485 St.): ‘qui per tres generationes huius familiae
gravis est.’ On 1602 (1611 St.) : Aepdany μὲν éymue . . . ἔτεκε. Apollod. 1. 3.
133. at quomodo iidem et Pelopidae et Pleisthenidae. Ayauéuvwv κατὰ μὲν
“Ὅμηρον... Πλεισθένους. Schol. Hom. ιλ. a. 7’ (it is evident that in Stanley's
note all that is not mere commonplace comes from Pearson). On 1640
(1649 St.) : ‘elegantissima haec sunt. Quadriga 4 equis constabat, quorum
2 Jugales, Funales 2. vid. Salm.’, etc.
Pearson’s emendations of the text and his suggestions of possible
readings were appropriated by Stanley in the same manner as were his
interpretations. Here again I must confine myself to a selection; again
I quote Pearson’s notes, with which Stanley’s should be compared. On
Prom. 55: ‘io’.! Βαλών νιν ἀμφὲ y. i. ἀμφιβαλὼν. ut ante περιβαλεῖν et p. p.
ἀλλ᾽ ἀμφὲ πλεύραις 71. Pers. 51.’ On Prom. 112: ‘si retineamus θηρῶμαι,
videtur legendum τοιῶν öe.’ On Prom. 187: ‘aut ἔμπας aut οἴω delendum
suadet carminis ratio. ἔμπας agnoscit Hic Schol. utrumque Arund.'* On
Prom. 1013 (1012 St.): ‘io. οὐδενὸς μεῖον. On Sept. 225 (231 St): ‘ic.
σωτήριος.᾽ On Sept. 667 (673 St.) : 'Schol. legisse videtur προσεῖδε.᾽ On Sept.
830 (836 St.): ‘ut Castores. nisi versus desit qui ad Eteoclem pertineret."
On Ag. 2: ‘io. μῆχος. vel μῆχαρ. 208.’ On Ag. 645 (654 St.): ‘io. ràv.' On
Ag. 1266 (1275 St.) : ‘ode.’ On Ag. 1625 (1634 St.) : ‘I. τοῦδ᾽ ἥκοντος᾽ (in the
next line Pearson placed a comma in front of ἅμα and deleted the comma
after dpa). On Ag. 1664 (1673 St.): δύσφορον vel tale quid deest."
Before leaving the subject of textual criticism I want to mention two
instances where Stanley's scholarship was not good enough to enable
him to see the value of Pearson's suggestions and where he therefore
thought it unnecessary to take them over. At the end of Prom. 945, where
Stanley gives the textus receptus τὸν ἐφημέροις, without any note, Pearson
observes ‘roy delendum' ; and on Sepi. 619 φιλεῖ δὲ σιγᾶν ἢ λέγειν τὰ καίρια
(no comment in Stanley's edition) Pearson has the note 'versus hic videtur
superfluus' and adds in the inner margin ‘Choe. 570 [582]’.*
Some entries of a more general character indicate the liveliness of
perception and the keen interest in literary history that are no less
noticeable in John Pearson than is his enormous learning. As an example
I quote his observation on the last item in the list of the dramatis personae
! Stanley, even when he leaves everything else unaltered, invariably replaces
Pearson's ἴσως by 'forsan' or 'fortasse'.
2 For these scholia see Stanley at the end of his preface.
3 This observation, like many others in this list, shows the vigilant mind of the true
κριτικός (cf. Hermann's discussion of the passage). For a conservative view see, e.g.,
Verrall, Mazon, Groeneboom ad loc.
* I have always regarded Sept. 619, which disrupts the context, as interpolated (the
same view was held, according to Wecklein's ‘Appendix’, by Jacobs and C. G. Haupt);
other scholars have resorted to transpositions. A particularly unfortunate excuse was
excogitated by Stanley in his posthumous note, 'quasi diceret Eteocles, Amphiaraus
ad commilitones suos dehortatione illa non usus fuerat, nisi eos victos fore praevidisset’ ;
this coincides almost exactly with Wilamowitz's note ‘talis vir tacuisset nisi certus esset
futuri".
80
 JOHN PEARSON’S SHARE IN STANLEY’S AESCHYLUS
at the end of the 'Yzó8eow of Pers. (p. 124), προλογίζει δὲ à χορός : ‘ita et
in Eurip. Rheso. male igitur Chorus a Scaligero! definitur, Pars fabulae
post Actum: et ipse Aristoteles, cum τὸ χορικὸν dividit et definit, ad τὸ
πολὺ, non τὸ dei respexit'.^ This note, without the slightest change, is
printed in Stanley's commentary, p. 755.
The section ‘Henrici Stephani Obseruationes' etc. which is printed as
an appendix to Victorius' edition of Aeschylus (pp. 359 ff.) called for
several marginal comments on the part of Pearson. These notes were
fused by Stanley with those which he found in the margins of the texts
of the plays. So e.g. in Stanley's note on Prom. 35 the whole of the criticism
of Stephanus ('Sed haec observatio est nihili' to the end of the paragraph)
comes from Pearson on p. 360; the note on Prom. 371 is copied from
Pearson on p. 362 ; the quotation from Themistius in Stanley on Prom. 378
was written down by Pearson on p. 363; and all the passages quoted by
Stanley on the ‘Yrddeoıs of Sept. in the section which begins ‘De Tragoediae
inscriptione lis est' (p. 737) can be found in Pearson's note on p. 366.
This is indeed an astounding affair. One may feel tempted to address
Stanley in Cicero's words, '(a Ioanne) vel sumpsisti multa, si fateris, vel,
si negas, surripuisti'. And yet, though no sign of a confession can be found
in Stanley's commentary, any idea of what we commonly mean by theft
or plagiarism is in this case out of the question. But we have interrupted
the book in the middle of its tale; we must wait until it has finished.
What, on fresh inspection, the MS Rawlinson G. 193 tells us about its
own history is this. It belonged to Thomas Stanley, who first used it for
a time, then asked John Pearson to enter his observations into it, then,
several years before the completion of his commentary, received it back
and continued to use it. We have now to examine the evidence in
detail.
Needham said in his description of MS Rawl. G. 193 (cf. p. 78) 'olim fuit
Th. Stanleii ut manus ejus abunde testatur'. This is perfectly correct, as
anyone will see who knows Stanley's hand from the Addenda to his
Aeschylus (Cambr. Univ. Libr., Adv. b. 44. 1-8) and from his 'Adversaria'
(ibid., Stanley MS, Gg. III. 15). Stanley first entered into the book in
large characters a number of 'headings', glosses, and fairly elementary
notes; later on he added in a smaller hand pieces of a Latin translation,
some of them very long. Then the book passed into the hands of Pearson,
who wrote down his own notes. The relative chronology of these three
stages is obvious throughout the book. A clear instance can be seen e.g.
on p. 129, where in the lower half (opposite the scholia) of the outer margin
we find one of Stanley's early notes, then, surrounding this note, his
translation, and finally, underneath Stanley's translation, Pearson's note
in the lower margin. Perhaps still more instructive is the case of the lower
part (opposite the scholia) of p. 242: in the outer margin Stanley first
wrote a few notes; round these notes he drew simple rectangular frames,
and going round these frames he wrote a translation of the whole text,
: Cf. Iulius Caesar Scaliger, Poetice, lib. i, cap. ix (p. 38 of the edition of 1594) : ‘Chorus
est pars inter actum et actum . . . tutior erit definitio quae dicat: post actum'.
3 For the difficulty of reconciling chapter 12 of the Poetics with the known facts of
earlier Tragedy see Bywater's commentary, p. 206 f.
4872-1 8r G
 APPENDIX II
which covers the outer margin and the greater part of the lower margin.
In the space left free by the translation, i.e. the lowest part of the lower
margin, we find, in Pearson’s hand, the correction of a detail of Stanley’s
translation, and, furthermore, in the narrow inner margin, very crammed,
Pearson’s note on the scholion ‘YzepBopéov (1. 5 from the bottom).
Pearson's notes are often written round one of Stanley's ; in other cases
they begin on the same line on which a note of Stanley's ends. Nor is it
uncommon for one of Pearson's remarks to refer to the note of Stanley's
that precedes it. See e.g. p. 142 top, where Stanley copies in the margin
the silly remark of the scholion (on Pers. 371) τὸ κράτος ἀντὶ τοῦ κράτους,
and Pearson, continuing on the same line, adds ‘male. ut ex quantitate
patet'. The result of this joint effort appears in Stanley's commentary:
'Scholiastes male notat hic ἀντίπτωσιν, ut ex quantitate patet.' On Ag. 7o
ἀπύρων ἱερῶν we find first a long note by Stanley, ᾿ἀπυρων pro δια ἀαπυρων
... Sic igitur intellige: neque flens per non ignea sacra erynnium (ignem
ad sua sacra non admittentium) iras graues mulcendo decipiat', to which
Pearson, continuing on the same line, adds this: 'recte. nisi quod διὰ non
sit intelligendum. sed sacra pro Diis ponuntur'. After receiving this crumb
from the great man's table, Stanley rides a high horse in his commentary:
‘per ἄπυρα ἱερὰ ipsae hic intelligendae Eumenides, sacra pro Diis: qua
metonymia non perspecta frustra se torsit vir eruditus." Still more
amusing are the antecedents of Stanley's comment on Ag. 103 as he
published it. Stanley wrote in the margin of his copy of the Stephanus
edition “τὴν θυμοβορον. anapaesticus est versus. quod superest λύπης φρενα
ex scholiaste irrepsit in contextum, cum apud illum legeretur λυποῦσαν
¢péva’, to which Pearson added 'neutiquam. Schol. enim non exponit
θυμοβόρον, sed constructionem aperit. vult enim Poetam dixisse λύπης
φρένα pro λύπην $pevós', and finally we read in Stanley's commentary this:
'Cave vero ne putes λύπης φρένα e Scholio irrepsisse in textum. Scholiastes
enim non exponit θυμοβόρον, sed tantum constructionem aperit : vult enim
Poetam dixisse λύπης φρένα pro λύπην $pevós. At Ag. 220 (229 St.) first
Stanley notes ‘leg. róre', then Pearson ‘immo ὅθεν᾽ ; the result in the
printed commentary is ‘Legimus, cum Scholiaste, 00e’. Of Ag. 637 (646 St.)
Stanley gives in the margin a paraphrase which agrees with P. Victorius'
interpretation, but Pearson vetoes this with ‘non. sed alius honor eorum
Deorum qui bona mittunt, alius eorum qui mala ut Erinnyes’; the result
is that Stanley in his commentary, after quoting Victorius, continues
‘Ego vero postremum illud aliter interpretandum censeo, nempe . . .' and
then gives Pearson's ifsa verba. At Ag. 1046 (1055 St.) Stanley writes in
the margin ‘fort. leg. e£eıs’, Pearson protests ‘imo. habes a nobis (dicta
sc.) quae fieri solent', and consequently Stanley says in his commentary
‘Poterit legi ἕξεις. . . Sed nihil exigit' and gives the translation ‘Habes
a nobis (dicta sc.) quae fieri solent’.? It will have become abundantly clear
that Stanley, like a diffident pupil, bows everywhere to the master's
superior judgement.
Fortunately we can fix with a fair degree of accuracy the time at which
Pearson wrote his notes in Stanley's copy. On p. 31 he illustrates διάδοχοι
! ie, Henr. Stephanus, whose note is printed on p. 378 of the edition of 1557.
2 For the correctness of this interpretation see my commentary.
82
 JOHN PEARSON’S SHARE IN STANLEY'S AESCHYLUS
(Prom. 464) by the marginal note ‘idem quod alibi [Aesch. fr. 194 N.]
ἐκδέκτορες ab eodem. v. Plut. de Fortuna p. 98’, to which he later added
in a different ink ‘et Hierocl. Fragm. 212’.! The latter reference is to the
edition of Hierocles De Providentia published in London in 1655 with
notes by Meric Casaubon and very learned Prolegomena by John Pearson.
The passage referred to, p. 212, is συμπαθεῖς... ἐν τοῖς ἐναντίοις καιροῖς
διάδοχοι τῶν ἀνιαρῶν. It is obvious that, in contrast to the Aeschylus frag-
ment quoted from Plutarch, this passage is not very much to the point;
it is the sort of 'parallel' which a reader likes to jot down if it is thrown
in his way. Only when his own and Meric Casaubon's edition of Hierocles
had reached the stage of page-proofs was it possible for Pearson to give
his reference in the form in which he did give it. In other words, this
particular entry was made in 1655 (or, at the earliest, late in 1654). The
fact that the reference to Hierocles was added afterwards to the original
note, showing as it does that this passage was not in Pearson's mind from
the beginning, makes it unlikely that he added it long after the publication
of his little edition of the Neoplatonist.
Since the reference to Hierocles is contained in an additional note, it
follows that Pearson, when he made this entry, had already been engaged
for some time in annotating Stanley's copy of the Victorius edition of
Aeschylus. Nor is this the only indication of a successive growth of
Pearson's marginalia in this book. To quote only one example, the outer
margin of p. 31 shows clearly that he began with writing a few brief notes,
especially variae lectiones (ypu . . .), and added the more extensive com-
ments afterwards, using a different ink and pen. A scholar of Pearson's
gigantic capacity for work is not likely to have spent an unreasonable time
on collecting and bringing to paper notes like these marginalia. Besides,
the years into which this πάρεργον fell were perhaps the most active in a
very active life:? from 1654 on Pearson delivered at St. Clement's, East-
cheap, 'the series of discourses which he published in 1659 under the title
of "An Exposition of the Creed" ', thus producing the work on which,
outside the circles of specialists, his fame primarily rests. Of a very
different kind of work undertaken during these same years we get a
glimpse through the entry on the title-page of his copy of Hesychius?
‘Hesychium integrum primo perlegi MDCLV. Oct. xv—Iterum MDCLXVII.
Mart. xxvi’. Probably there were many other activities, both ecclesiastical
and secular, that filled his days at that time. Át the end of his Prolego-
mena to Hierocles (1655) he says he must now conclude 'ut reliquis studiis,
quae me imperiosius avocant, vacem’. Taking all this into consideration,
it is safe to assume that Pearson returned Stanley's copy to its owner
not later than 1656. This means that Stanley received the book back six
or seven years before he completed his edition (published in 1663); he
therefore had plenty of time to profit from the gift of his benefactor, not
only by taking over almost everything that the great scholar had written
in the margins, but also by trying, to the best of his own capacities, to
follow the master's model in collecting further materials. His attempts
! Needless to say, these notes reappear in Stanley's commentary.
2 Cf. Dictionary of National Biography, xliv. 169.
3 [bid., 173.

83
 APPENDIX II
to add something to Pearson’s notes have sometimes left their traces in
this very copy of Aeschylus. For example, on p. 2 the whole note (now
in Stanley’s commentary, p. 707, right-hand column, middle) ‘immo vero
Geloi' down to αὐτὸ neraßdvras is Pearson's, but Stanley, continuing on
the same line, adds ‘et Athenaeus deipn: lib. 14 p. 627 . . . Mnôos ἐπιστα-
pevos’. Similarly on p. 24 (schol. on Prom. 323) a wealth of parallels to πρὸς
κέντρα Aaxritew is quoted by Pearson, but at the end of them Stanley
is at least able to make this contribution of his own: ‘vide Agam. p. 226
[l. 1624].’
The facts, then, as they emerge one by one from the margins of MS
Rawl. G. 193, are perfectly clear. But we are faced with a very different and
indeed puzzling problem as soon as we try to interpret these facts in the
light of our common experience of human relations or to apply to them
what we regard as normal standards of a scholar’s decent behaviour. One
thing, however, remains certain: the idea that Stanley in this matter acted
as a thief must be discarded. A thief, according to the Oxford English
Dictionary, is ‘one who takes portable property from another without the
knowledge or consent of the latter, converting it to his own use’, ‘one who
does this by stealth, esp. from the person’. Now nothing can be farther
removed from stealth than what Stanley did in this case. He gave his
copy of Aeschylus to Pearson and received it back with the great scholar’s
notes; he must have told him for what purpose he wanted his contribu-
tions. It is practically certain that Stanley, in using Pearson’s materials,
did it with ‘the knowledge and consent of the latter’. But even supposing
that Stanley did not at once inform Pearson of his plan of editing Aeschylus
with a commentary, it is inconceivable that at the time when he published
his edition (1663) he could have hoped to get away with his enormous
borrowings if he had not previously obtained Pearson’s consent. At that
time Pearson, who had for many years been known as one of the most
eminent English divines and scholars, was Master of Trinity College,
Cambridge. Unless you are a Junatic, you will not attempt to steal products
of scholarship from a man in such a position. That Stanley did not feel
any qualms about his appropriation of the fruits of Pearson’s learning
can also be inferred from the fact that he did not destroy the book which
might at any time have given its secret away (Pearson’s characteristic
hand must have been known to many contemporaries).
But why did Stanley not acknowledge, at least in a general way, his
debt to Pearson, which he could have done once and for all in his Preface?
The most probable answer seems to be that, for reasons unknown to us,
Pearson did not want his name to be mentioned in this connexion. Perhaps
his generosity was such that he preferred to make a complete and un-
qualified present of the precious things which he had put at the younger
man’s disposal: a king can afford to be lavish. However that may be,
by the manner in which Stanley used the notes he paid a very high tribute
to Pearson’s character: he must have felt that he could rely absolutely
on his benefactor’s silence. This fits in with what we are told about
Pearson’s excellent temper and equanimity.’ So far as Stanley is con-
cerned, we might perhaps wish that he had taken a little less advantage
1 Cf. op. cit. (p. 83 n. 2), 170.

84
 JOHN PEARSON’S SHARE IN STANLEY’S AESCHYLUS
of Pearson’s generosity and used a more modest language when he paraded
the great man’s observations as his own. But in order to be able to judge
the true nature of the relation between Pearson and Stanley we ought to
possess some direct documentary evidence, e.g. letters exchanged by
them; thus far I have not succeeded in finding any, despite all my
efforts.’
1 Butler in his edition of Aeschylus, vol. viii, pp. xvi ff., prints a number of ‘testi-
monia’ which Stanley had prepared for the second edition of his Aeschylus (Butler
copied them from the fly-leaves of the first volume of Stanley’s interleaved copy of
his Aeschylus in the University Library at Cambridge, Adv. b. 44. 1). The first of them
is Stanley’s copy of a letter written to him in February 1664 by Isaac Vossius; it con-
tains the following sentence: ‘quod de Χοηφόρων initio significavit tibi Dom. Pearson,
id male intellectum’, etc. Now in Stanley’s edition there is no hint of any such remark
by Pearson. The obvious inference is that Stanley had informed Vossius that Pearson,
either by word of mouth or in a letter, had given him his views on the beginning of the
Choephoroe.

85
 SIGLA LIBRORVM
frustula papyri Oxyrhynchiae cf. p. I
Laurentianus XXXII. 9 cf. p. I
eiusdem libri corrector vetustus cf. p. I sq.
: — Venetus Marcianus 653 (olim 468) . cf. p. 2 sq.
Triclinii liber, Farnesianus Neapolitanus II. F. 31 cf. p. 3 sq.
Laurentianus XXXI. 8 cf. p. 4534.
Venetus Marcianus 663 (olim 616) . cf. p. 5

86
AIZXYAOY ATAMEMNON
 ᾿Αγαμέμνονος ὑπόθεσις
᾿Αγαμέμνων εἷς Ἴλιον ἀπιὼν τῆι Κλυταιμήστραι, εἰ πορθήσοι τὸ Ἴλιον,
... τῆς αὐτῆς ἡμέρας σημαίνειν διὰ πυρσοῦ. ὅθεν σκοπὸν ἐκάθισεν
ἐπὶ μισθῶι Κλυταιμήστρα, ἵνα τηροίη τὸν πυρσόν. καὶ ὁ μὲν ἰδὼν
ἀπήγγειλεν, αὐτὴ δὲ τῶν πρεσβυτῶν ὄχλον μεταπέμπεται, περὶ τοῦ
5 πυρσοῦ ἐροῦσα' ἐξ ὧν καὶ ὁ χορὸς συνίσταται: οἵτινες ἀκούσαντες
παιανίζουσιν. μετ᾽ οὐ πολὺ δὲ καὶ Ταλθύβιος παραγίνεται καὶ τὰ κατὰ
τὸν πλοῦν διηγεῖται. ᾿Αγαμέμνων δ᾽ ἐπὶ ἀπήνης ἔρχεται" εἵπετο δ᾽ av-
τῶι ἑτέρα ἀπήνη, ἔνθα ἦν τὰ λάφυρα καὶ ἡ Κασσάνδρα. αὐτὸς
μὲν οὖν προεισέρχεται εἰς τὸν οἶκον σὺν τῆι Κλυταιμήστραι, Κασσάν-
10 Spa δὲ προμαντεύεται, πρὶν εἰς τὰ βασίλεια εἰσελθεῖν, τὸν ἑαυτῆς
καὶ τοῦ ᾿Αγαμέμνονος θάνατον καὶ τὴν ἐξ ᾿Ορέστου μητροκτονίαν, καὶ
εἰσπηδᾶι ὡς θανουμένη, ῥίψασα τὰ στέμματα. τοῦτο δὲ τὸ μέρος
τοῦ δράματος θαυμάτεται ὡς ἔκπληξιν ἔχον καὶ οἶκτον ἱκανόν.
ἰδίως δὲ Αἰσχύλος τὸν ᾿Αγαμέμνονα ἐπὶ σκηνῆς ἀναιρεῖσθαι ποιεῖ,
15 τὸν δὲ Κασσάνδρας σιωτήσας θάνατον νεκρὰν αὐτὴν ὑπέδειξεν,
πεποίηκέν τε Αἴγισθον kal Κλυταιμήστραν ἑκάτερον διισχυριζόμενον περὶ
τῆς ἀναιρέσεως ἑνὶ κεφαλαίωι, τὴν μὲν τῆι ἀναιρέσει ᾿Ιφιγενείας, τὸν δὲ
ταῖς τοῦ πατρὸς Θυέστου ἐξ ᾿Ατρέως συμφοραῖς.
ἐδιδάχθη τὸ δρᾶμα ἐπὶ ἄρχοντος Φιλοκλέους ὀλυμπιάδι ὀγδοη-
20 κΚοστῆι ἔτει δευτέρωι (a. Chr. n. 459/8). πρῶτος Αἰσχύλος ᾿Αγαμέμ-
νονι, Χοηφόροις, Εὐμενίσι, Πρωτεῖ σατυρικῶι. ἐχορήγει ZevokAñs
᾿Αφιδναῖος.

Codices: mVFG(raro notatur, cf. p. 30 sq.)Tr

Titulus ἀγαμέμνονος ὑπόθεσις MV: ὑπόθεσις ἀγαμέμνονος F: ὑπόθεσις τοῦ προκειμένου
δράματος Ττ I κλυταιμν- semper VF et (praeter 258) Tr; cf. comm. ad 84
πορθήσει FTr 2 verbum omissum aut post "IAov, ubi ὑπέσχετο FTr, aut post
σημαίνειν, ubi ἔφη V κατ᾽ αὐτὴν σημανεῖν τὴν ἡμέραν Tr διὰ τοῦ πυρσοῦ m
3 ἵνα τηροίη mVF: ὥστε σκοπεῖν Tr καὶ ὁ μὲν MVE: ὁ μὲν οὖν Tr 4 τὸν τῶν
mpeoß. ueram, ὄχλον ΕἼΤ . 4.834. ἐροῦσα περὶ τ. m. FTr 5 ἐξ ὧν δὴ καὶ Tr
6 δὲ om. G prius καὶ om. F 8 καὶ ἑτέρα Tr κασσάνδρα semper V:
κασ- rell.; cf. comm. ad 1035 9 προσήρχετο F: προσέρχεται Tr post κλυτ.
add. ὑπαντησάσῃ αὐτῷ Tr IO πρὶν εἰσελθεῖν τὰ Bac. Tr ἐλθεῖν F αὑτῆς
ν IO sq. ἑαυτῆς τε καὶ Tr II ἐξ om. Tr 13 θαυμαστόν ἐστιν
ὡς καὶ ἔκπληξιν καὶ οἶκτον ἱκανὸν ἐμποιεῖν FTr 15 ὑπέδειξεν (-£c V) mV: ἀπέδειξε F:
ἀποδείκνυσιν Tr 1688 (pro re) Tr αἴγιστον F (sed recte G) καὶ κλυταιμνήστραν
διισχυριζομένους Tr 17 Tv μὲν τὴν ἀναίρεσιν idıyeveias προβαλλομένην Tr 18 πατρὸς
Θυέστου Victorius : πατρὸς ὀρέστου M: πατρὸς αἰγίσθου (-ἰίστου F) VFG τὸν δὲ τὴν
θυέστου ἐξ ἀτρέως συμφοράν Tr 10 ἐδιδάχθη δὲ F (δὲ non habet G) 10 54. ὀγδοηκοστῆι
Meursius: xg codd. 20-2 verba πρῶτος κτλ. usque ad finem didascaliae om.
Tr 21 mpwret mV: πρῶτος F 22 ἀφιδνεύς codd.: corr. Wilamowitz de didascalia
(19-22) cf. fastos theatri Atheniensis, IG 11. 22. 2318, 1. 5o ἐπὶ Φιλοκλέους . . . τραγωιδῶν |
Ξενοκλῆς Adudva: ἐχορή | Αἰσχύλος ἐδίδασκεν

88
 τὰ τοῦ δράματος πρόσωπα
φύλαξ χορός ἄγγελος
Κλυταιμήστρα Ταλθύβιος κῆρυξ 25
᾿Αγαμέμνων Κασσάνδρα
Αἴγισθος
mVF(G)Tr
26 sq. κασσάνδρα: ἀγαμέμνων : alyıodos V 27 alyioros F post Aiyıodos
FGTr habent προλογίζει (δὲ add. F; cf. 19) ὁ φύλαξ, θεράπων (6 dep. G) ἀγαμέμνονος, post
fabulae nomen m inseruit θεράπων ἀγαμέμνονος ὃ mpoAoyılduevos, οὐχὶ ὁ ὑπὸ alylodov
raxbeis. haec olim argumenti fabulae partem fuisse apparet

89
 QYAAZ
Θεοὺς μὲν αἰτῶ τῶνδ᾽ ἀπαλλαγὴν πόνων,
φρουρᾶς ἐτείας μῆκος ἣν κοιμώμενος
στέγαις ᾿Ατρειδῶν ἄγκαθεν, κυνὸς δίκην,
ἄστρων κάτοιδα νυκτέρων ὁμήγυριν
καὶ τοὺς φέροντας χεῖμα καὶ θέρος βροτοῖς 5
λαμπροὺς δυνάστας ἐμπρέποντας αἰθέρι.
[ἀστέρας ὅταν φθίνωσιν ἀντολάς τε τῶν.]
καὶ νῦν φυλάσσω λαμπάδος τὸ σύμβολον,
αὐγὴν πυρὸς φέρουσαν ἐκ Τροίας φάτιν
ἁλώσιμόν τε βάξιν: ὧδε γὰρ κρατεῖ 10
γυναικὸς ἀνδρόβουλον ἐλπίφον κέαρ.
εὖτ᾽ ἂν δὲ νυκτίπλαγκτον ἔνδροσόν τ᾽ ἔχω
εὐνὴν ὀνείροις οὐκ ἐπισκοπουμένην
ἐμήν: Φόβος γὰρ ἀνθ᾽ Ὕπνου παραστατεῖ
τὸ μὴ βεβαίως βλέφαρα συμβαλεῖν ὕπνωι᾽ 15
ὅταν δ᾽ ἀείδειν ἢ μινύρεσθαι δοκῶ,
ὕπνου τόδ᾽ ἀντίμολτπον ἐντέμνων ἄκος,
κλαίω τότ᾽ οἴκου τοῦδε συμφορὰν στένων
οὐχ ὡς τὰ πρόσθ᾽ ἄριστα διαπονουμένου.
νῦν δ᾽ εὐτυχὴς γένοιτ᾽ ἀπαλλαγὴ πόνων 20
εὐαγγέλου φανέντος ὀρφναίου πυρός.
ὦ χαῖρε λαμττήρ, νυκτὸς ἡμερήσιον
φάος πιφαύσκων καὶ χορῶν κατάστασιν
πολλῶν ἐν Ἄργει τῆσδε συμφορᾶς χάριν.
ἰοὺ ἰού. 25
᾿Αγαμέμνονος γυναικὶ σημαίνω τορῶς
εὐνῆς ἐπαντείλασαν ὡς τάχος δόμοις
ὀλολυγμὸν εὐφημοῦντα τῆιδε λαμπάδι

Codices: MVFG(raro notatur, cf. p. 30 sq.)Tr (de papyro cf. ad vs. 7)

2 μῆκος ἣν FTr: μῆκος δ᾽ ἦν MV 3 cf. Hesychius ἄγκαθεν (ἀγρίαθεν cod.)*
ἀνέκαθεν. Αἰσχύλος Ἀγαμέμνονι; eadem glossa, sed fabulae nomine omisso, in Zuvay.
rEg. χρησ. (Anecd. Bekker 337. 25), unde Photius Berol. 17. 23 ἄγκαθεν" κατὰ συγκοπήν,
ἀντὶ τοῦ ἀνέκαθεν. οὕτως «Αἰσχύλος ; cf. schol. ad h.l. 4-6 toti afferuntur ab Achille
in Árat. p. 28 Maass 5 θέρος βροτοῖς MV Achilles: βροτοῖς θέρος FTr 6 αἰθέρι
MFTr: ἐν θέρει V et Achillis cod. Vat. versuum 7-30 initia mutila exstant in pap.
Oxy. 2178 7 versum Pauwio suspectum eiecit Valckenaer — ór' dv MV II
Ioannes Siceliota, schol. ad Hermog. de id. vi. 225 Walz, ὡς κἀν τῶι Ἀγαμέμνονι" ! γυναικὸς
ἀνδρόβωλον (sic) ἐλπίζων xéap' ἐλπίζον FIT: ἐλπίζων (sed in utroque libro super o
scr. o) MV 12 εἰύτ᾽ ἂν δὲ »[uxr etiam pap. super δὲ scriptum γε F νυκτί-
πλακτον Tr ἔχων Weil, fort. recte 13 εὐ]νὴν ὀνείροις etiam pap. 14 éluÿv: φόβος
yà[p etiam pap. 17 ἐντέμνων MVTr: ἐκτέμνων (superscr. ἐν F) FG 23 φάος MV:
φαΐ pap. : νῦν φῶς FTr 26 σημαίνω M : σημανῶ rell,
90
 The house of the Atridae at Argos. On the flat roof lies a
Watchman
Waichman. Of the gods I ask deliverance from this toil, from
my year-long watch, in which, lying at the house of the
Atridae on my arms, dog-fashion, I have become familiar
with the assembly of the stars of night, and those bright
potentates conspicuous in the sky who bring winter and
summer to man. And so now I am watching for the torch-
signal, the bright gleam of fire bringing a tale from Troy, the
tidings of her capture: for in such a manner rules a woman’s
heart of manly counsel, being full of expectation. And when
I keep my couch that is restless by night and drenched with
dew, this couch of mine that no dreams look upon—for
instead of Sleep Fear stands beside me, that I may not close
my eyelids firmly in sleep—and when 1 have ἃ mind to sing
a song or hum a tune, tapping (as sap from a root) this
medicine of song against sleep, then 1 weep and groan over
the misfortune of this house, not now, as of old, excellently
husbanded. But now may there come a happy deliverance
from my toil by the fire of good tidings appearing in the dark.

A short interval. Then the beacon appears ; the Watchman

jumps to his feet
O hail, thou light-giver, that showest a light of day by
night and (the signal for) the setting up of many a dancing
choir in Argos for this event.
Hurrah, hurrah!
To Agamemnon’s wife I give clear signal to arise from her
couch with all speed and lift up her voice, for the house to
hear, in a jubilant shout of thanksgiving in welcome to this
91
 ἐπορθιάφειν, εἴπερ ᾿Ιλίου πόλις
ἑάλωκεν, ὡς ὁ φρυκτὸς ἀγγέλλων πρέπει" 30
αὐτός τ᾽ ἔγωγε φροίμιον χορεύσομαοι,
τὰ δεσποτῶν γὰρ εὖ πεσόντα θήσομαι
τρὶς ἐξ βαλούσης τῆσδέ μοι φρυκτωρίας.
γένοιτο δ᾽ οὖν μολόντος εὐφιλῆ χέρα
ἄνακτος οἴκων τῆιδε βαστάσαι χερί. 35
τὰ δ᾽ ἄλλα σιγῶ’ βοῦς ἐπὶ γλώσσηι μέγας
βέβηκεν’ οἶκος δ᾽ αὐτός, εἰ φθογγὴν λάβοι,
σαφέστατ᾽ ἂν Akgeıev ὡς ἑκὼν ἐγὼ
μαϑοῦσιν αὐδῶ Kol μαθοῦσι λήθομαι.

ΧΟΡΟΣ
δέκατον μὲν ἔτος τόδ᾽ ἐπεὶ Πριάμου 40
μέγας ἀντίδικος,
Μενέλαος ἄναξ ἠδ᾽ ᾿Αγαμέμνων,
διθρόνου Διόθεν καὶ δισκήτττρου
τιμῆς ὀχυρὸν φεῦγος ᾿Ατρειδᾶν,
στόλον ᾿Αργείων χιλιοναύτην 45
τῆσδ᾽ ἀπὸ χώρας
ἦραν, στρατιῶτιν ἀρωγήν,
μέγαν ἐκ θυμοῦ κλάφοντες "Apn,
τρόπον αἰγυπιῶν
οἶτ᾽ ἐκπατίοις ἄλγεσι παίδων 50
ὕπατοι λεχέων στροφοδινοῦνται
πτερύγων ἐρετμοῖσιν ἐρεσσόμενοι,
δεμνιοτήρη
πόνον ὀρταλίχων ὀλέσαντες"
ὕπατος δ᾽ ἀίων ἤ τις ᾿Απόλλων 55
ἢ Πὰν ἢ Ζεὺς
οἰωνόθροον γόον ὀξυβόαν,
τῶν δὲ μετοίκων » « * s » κ, 57

29-45 MVF(G)Tr 46-57 MVFTr

29 ἐπορθριάζειν MV 30 ἑάλωκεν codd.: de papyri scriptura cf. comm. ἀγγέλλων
GTr: ἀγγέλων MVE 3256. totum distichon afferunt Photius et Suidas (vol. iv. 593
Adler) s.v. τρὶς &£ «rd. 33 Eustath. ad a 234 p. 1414. 5 (neque poetae neque
fabulae nomine addito) κατὰ τὸ ‘ τρὶς ξὲ βαλούσης τῆς Διὸς φρυκτωρίας ' βαλούσης
(βαλοῦσα Suid.) τῆς ἐμῆς φρυκτωρίας Phot., Suid. 38 λέξει F (sed λέξειεν G cum
reliquis) 39 αὐδῶν où V κακοὺ M (cf. Pers. 295, ubi καὶ εἰ M) 40 πριάμω
MV 45 χιλιοναύταν codd. (super av scr. n MTr) yp. ἴλιον αὐτάν in marg. mF
post χιλιοναύταν deficit G; cf. ad 1095 47 par Μ ἀρωγὰν codd. (superscr.
n MTr) 48 xAdy£avres FIT 52 ἐρετμοῖσι V 53 sq. Hesychius δεμνιοτήρη πόνον"
Αἰσχύλος Ἀγαμέμνονι, cum scholio de quo cf. comm. 57 τῶνδε codd. : distinxit
Hermann lacunam post ὀξυβόαν statuit Hermann, post μετοίκων ego τῶν δὲ
μετοίκων hinc affert schol. Soph. Oed. Col. 934; cf. comm.
92
torch, if the town of Ilion is taken indeed, as is the message
of the beacon that shines out of the dark; and I myself will
dance for prelude, for as my masters’ cast has been lucky,
I'll make my move accordingly, now that this beaconing has
thrown treble six for me. Well, may it then come to pass that
the master of the house return and I clasp his dear hand
in mine. As for the rest, I am silent: a great ox has set his
foot upon my tongue; the house itself, if it were given a
voice, would tell the tale most clearly ; I choose, for my part,
to speak to those who know and understand, and to those
who do not to forget.

He descends into the house.

Enter the Chorus of twelve old citizens of Argos

Chorus. This is now the tenth year since Priam's great

adversary at suit, King Menelaus and Agamemnon, the stal-
wart yoke of the Atridae, paired in the honour of two thrones
and two sceptres derived from Zeus, put out from this land
an Argive armament of a thousand ships to give fighting
help, shouting from an angry heart the cry for a mighty war,
like vultures that, in extreme (?) grief for their children, high
above their bed circle round and round, rowing with the oars
of their wings, having lost the couch-keeping labour they had
spent over their nestlings ; but one in the height, Apollo, it
may be, or Pan or Zeus, hearing the shrill cry of the birds'
lament, and (feeling great compassion for) the denizens in his
93
 ὑστερόποινον
πέμπει παραβᾶσιν ᾿Ερινύν.
οὕτω δ᾽ ᾿Ατρέως παῖδας ὁ κρείσσων 60
ἐπ᾽ ᾿Αλεξάνδρωι πέμπει ξένιος
Ζεύς, πολυάνορος ἀμφὶ γυναικὸς
πολλὰ παλαίσματα καὶ γυιοβαρῆ,
γόνατος κονίαισιν épeiSopévou
διακναιϊιομένης τ᾽ ἐν προτελείοις 65
κάμακος, θήσων Δαναοῖσιν
Τρωσί θ᾽ ὁμοίως. ἔστι δ᾽ ὅπηι νῦν
ἔστι" τελεῖται δ᾽ ἐς τὸ πειτρωμένον᾽"
οὔθ᾽ ὑποκαίων οὔτ᾽ ἐπιλείβων
οὔτε δακρύων: ἀπύρων ἱερῶν 70
ὀργὰς ἀτενεῖς πιαραθέλξει.
ἡμεῖς δ᾽ ἀτίται σαρκὶ παλαιᾶι
τῆς τότ᾽ ἀρωγῆς ὑπολειφθέντες
μίμνομεν ἰσχὺν
ἰσόταιδα νέμοντες ἐπὶ σκήπτροις. 75
ὅ τε γὰρ νεαρὸς μυελὸς στέρνων
ἐντὸς ἀνάισσων
ἰσόπρεσβυς, "Ἄρης δ᾽ οὐκ ἔνι χώραιΐ,
τό θ᾽ ὑπεργήρων φυλλάδος ἤδη
κατακαρφομένης τρίποδας μὲν ὁδοὺς 80
στείχει, παιδὸς δ᾽ οὐδὲν ἀρείων
ὄναρ ἡμερόφαντον ἀλαίνει.
σὺ δέ, Τυνδάρεω
θύγατερ, βασίλεια Κλνταιμήστρα,
τί χρέος; τί νέον; τί δ᾽ ἐπταισθομένη, 85
τίνος ἀγγελίας
πειθοῖ περίπεμττα θυοσκεῖς;
πάντων δὲ θεῶν τῶν ἀστυνόμων,
ὑπάτων χθονίων
MVFTr
50 ἐρινύν MV: ἐριννὺν FTr 64 ἐρειδομένου V : ἐριδομ- M: ἐρειπομ- (superscr. do
F) FTr 65 πρωτελείοις F 66 davaoicı V 67 ὅπη codd. (super n add. « M)
69 ὑποκλαίων codd.: corr. Casaubon οὔθ᾽ ὑπολείβων codd. : corr. Schütz
70 οὔτε δακρύων deleverunt Bamberger aliique; glossemate δακρύων inculcato genuina
scriptura expulsa esse videtur 72 ἀτίται σαρκὶ corr. ex ἀτίταᾳ σαρκὰ M: ἀτίταια
σαρκὶ V : ἀτίται σαρκὶ F : ἀτιταὶ σαρκὶ Tr 75 ἰσόπεδα T" 76 voepós V
77 ἀνάσσων codd.: corr. Hermann 78 ἔνι FTr: u M: é V. χώραι corruptum esse
viderunt multi, sanavit nemo 79 τό θ᾽ ὑπέργηρων Tr: τόθιπεργήρως VF:
τίθιπεργήρως M 80 τρίποδος FTr 81 8’ om. V 82 ἡμερόφατον MV
83 τυνδάρξῳ" (littera w ex ao facta et accentu super e lineola transfixo’ Vitelli) M
84 κλυταιμνήστρα VFTr (cf. ad argum. 1) 87 πυθοῖ F θυοσκεῖς Turnebus :
θυοσκοεῖς varia lectio in schol. vet. Tr: θυοσκινεῖς (priore ı in rasura scripto M) MTr:
θύοσκινεῖς (accentu super v lineola transfixo ΕἾ VF
94 .
realm, sends on the transgressors her who brings punishment
though late, Erinys. Even so the sons of Atreus are sent by
the mightier one against Alexandros, by Zeus, guardian of
guest-law, who over a woman of many husbands will bring
about wrestlings many and wearying, where the knee is
pressed in the dust and the spear-shaft snapped in the initial
offering, for Danaans and Trojans alike. The matter is where
now it is; it will be fulfilled to its destined end: neither by
burning sacrifices nor by libations (nor by the spell of (?))
fireless offerings shall he (i.e. the guilty man) soothe aside
the relentless wrath.
But we, insolvent (i.e. unable to serve as soldiers) with our
aged frames, were left out of the supporting expedition that
then was sent, and remain behind, moving a child-like
strength upon staffs. For as the young marrow leaping
upward within the breast is like (that of) an old man, and
Ares is not therein . . ., so extreme old age, its foliage now
withering, goes its way on three feet, and no stronger than
a child he (1.6. the old man) wanders, a dream-phantom
appearing in daylight.
But thou, daughter of Tyndareus, Queen Clytemnestra,
what is it now? what fresh tidings? what intelligence hast
thou received, what message has prevailed upon thee, that,
by sending (messengers) around, thou arrangest sacrifices ?
The altars of all gods that dwell in our city and rule it—gods
supreme and gods infernal, those of the sky and those of the
95
 τῶν τ᾽ οὐρανίων τῶν τ᾽ ἀγοραίων, 90
βωμοὶ δώροισι φλέγονται"
ἄλλη δ᾽ ἄλλοθεν οὐρανομήκης
λαμπὰς ἀνίσχει,
φαρμασσομένη χρίματος ἁγνοῦ
μαλακαῖς ἀδόλοισι παρηγορίαις, 95
πελανῶι μυχόθεν βασιλείωι.
τούτων λέξασ᾽ 6 τι καὶ δυνατὸν
καὶ θέμις αἴνει,
παιών τε γενοῦ τῆσδε μερίμνης,
ἣ νῦν τοτὲ μὲν κακόφρων τελέθει, 100
τοτὲ δ᾽ ἐκ θυσιῶν ἃς ἀναφαίνεις
ἐλπὶς ἀμύνει φροντίδ᾽ ἄπληστον
ἱτὴν θυμοφθόρον λύπης φρένα.
κύριός εἶμι θροεῖν ὅδιον κράτος αἴσιον ἀνδρῶν
ἐντελέων’ ἔτι γὰρ θεόθεν καταπνείει 105
πειθὼ μολιτᾶν ἀλκὰν σύμφυτος olov:
ὅπως ᾿Αχαιῶν δίθρονον κράτος, Ἑλλάδος ἥβας
ξύμφρονα ταγάν, 110
πέμπει ξὺν δορὶ καὶ χερὶ πράκτορι θούριος ὄρνις
Τευκρίδ᾽ ἐπ᾽ αἷαν,
οἰωνῶν βασιλεὺς βασιλεῦσι νε-
ὧν, ὁ κελαινὸς 6 τ᾽ ἐξόπιν ἀργᾶς, 115
φανέντες ἴκταρ μελάθρων χερὸς ἐκ δοριπάλτου
παμπρέτπτοις ἐν ἕδραισιν
βοσκόμενοι λαγίναν ἐρικύμονα φέρματι γένναν
βλαβέντα λοισθίων δρόμων. 130
MVFTr
91 δώροισι Tr: δώροις rell. 94 xptuaros M: χρήματος V : χρίσματος FTr 96
πελανῶ M: πελάνω (-w Tr) rell. 98 αἴνει Wieseler : αἰνεῖν MV: εἰπεῖν FTr IOI ds
ἀναφαίνεις Ahrens: ἀγανὰ φαίνεις M: dyava φαίνει V: dyava φαίνουσ᾽ FTr 102 ἄπλεισ-
ro» M 1Ο3 ϑυμοβόρον FTr (cf. schol. M ἥτις ἐστὶ θυμοβόρος λύπη τῆς φρενός) λύπης
φρένα MVF : λυποφρένα Tr; cf. schol. ipsius λυποφρένα δὲ γράφε. οὕτω γὰρ ἔχει πρὸς τὸ
μέτρον ὀρθῶς 104 versum affert Aristophanes Ran. 1276, ubi ὃς δῖον (in archetypo
o&kov fuisse conicias) R, ὅσιον rell.; in vetustis Aristophanis libris variam scripturam
fuisse e scholio quamvis corrupto apparet IOS ἐκτελέων codd. : corr. Auratus
καταπνείει Ald. : zaranveaeı (‘una post élittera erasa quae v fuisse potest’ Vitelli) M :
xaramveve rell. 106 ex μολπᾶν fecit μολπὰν M : μολπὰν rell. 108 sq. ὅπως...
ἦβας affert Ar. Ran. 1285 Bas Ar. : ἦβαν (corr. ex ἡβᾶν M) codd. IIO-I2
(évudpova ... alav) om. M spatio vacuo relicto, supplevit corrector (m) IIO σύμ-
¢pova VFTr τὰν γᾶν m (sed τὴν ὁμόφρονα περὶ τὰ τακτικά schol. M) III σὺν δορὶ
καὶ χερὶ πράκτορι Ar. Ran, 1289 (ubi θούριος ὄρνις quoque afferuntur): ξὺν (σὺν m) δορὲ
δέκας πράκτορι codd. 113 οἰωνῶν βασιλεὺς om. V in initio paginae 115 ἀργίας
codd. : corr. Heath (ἀργής) et Blomfield 116 où φανέντες F χειρὸς δορυπάλτου
codd.: corr. Turnebus 117 παμπρέπτοις (παμπρέποις m) ἐν ἕδραισιν (-σι V) MV:
παμπρέποισιν ἕδραις F: παμπρέπτεσιν ἕδραις Tr 118 βοοκομένην Tr 119 ἐρικύματα
Μ φέρματι MV : φέρβοντο FTr
96
market—are ablaze with gifts ; and torches, from here, from

there, rise up heaven-high, medicined by the gentle guileless

persuasions of hallowed ointment, the offering of the royal
household’s oil from the inner store.
Of these matters do thou consent to tell what thou both

canst and mayest, and become a healer of this anxiety,

which now at one time proves laden with thoughts of ill,

while at another from the sacrifices which thou makest

appear comes Hope and beats off the insatiate care and grief,

a bane that consumes the heart (?).!

I have power to tell of the auspicious command ruling the

expedition, the command of men in authority—for still from
the gods the age that has grown with me breathes down
upon me persuasiveness of song to be my warlike strength—
(to tell) how the twin-throned command of the Achaeans, the
concordant leadership of Hellas’ youth, was sped with
avenging spear and arm to the Teucrian land by the warlike
bird, king of birds appearing to the kings of the ships, the
black one and the one that is white behind,? appearing near
to the abode, on the spear-hand side, perched in full view,

feeding upon her of the hares’ race, big with many young,

! The last words cannot be restored with any degree of certainty.

2 Or ‘the black-tailed one and the white-tailed one’,

4872-1 97 H
 αἴλινον αἴλινον εἰπέ, τὸ δ᾽ εὖ νικάτω. —

κεδνὸς δὲ στρατόμαντις ἰδὼν δύο λήμασι δισσοὺς

᾿Ατρείδας μαχίμους ἐδάη λαγοδαίτας
πομποὺς [τ᾿ ἀρχάς" οὕτω δ᾽ eltre τεράιϊφων" 125
* χρόνωι μὲν ἀγρεῖ Πριάμου πόλιν ade κέλευθος,
πάντα δὲ πύργων
κτήνη πρόσθε τὰ δημιοπληθέα μοῖρα λαπάξει
πρὸς τὸ βίαιον" 130
οἷον μή τις ἄγα θεόθεν κνεφά-
oni προτυπὲν στόμιον μέγα Τροίας
στρατωθέν. οἴκτωι γὰρ ἐπίφθονος "Ἄρτεμις ἁγνὰ 135
πτανοῖσιν κυσὶ πατρὸς
αὐτότοκον πρὸ λόχου μογερὰν πτάκα θνομένοισιν᾽"
στυγεῖ δὲ δεῖπνον aletdv.’
αἴλινον αἴλινον εἰπέ, τὸ δ᾽ εὖ νικάτω. ==

“τόσον περ εὔφρων ἁ καλὰ 140

δρόσοις ἀέτττοις μαλερῶν λεόντων
πάντων τ᾽ ἀγρονόμων φιλομάστοις
θηρῶν ὀβρικάλοισι τερπνὰ
τούτων αἰνεῖ ξύμβολα κρᾶναι,
δεξιὰ μὲν κατάμομφα δὲ φάσματα [στρουθῶν]. 145
ἰήιον δὲ καλέω Παιᾶνα,
μή τινας ἀντιπνόους Δαναοῖς χρονί-
ας ἐχενῆιδας ἀπλοίας 150
τεύξηι σπευδομένα θυσίαν été-
ραν ἄνομόν tiv’ ἄδαιτον,

MVFTr
121 αἵλινον αἵλινον M (item 139, 159) atÀwov semel F (item 139, 159) εὐνικάτω
Μ (οἶμαι εὖ νικάτω in marg. m) 122 δύω Μ λήμμασι ΕἼΤ 123 λογοδαίτας
MV 1247’ del. Thiersch ἀρχάς MV: apxovs FTr πομπᾶς (πομπῆς Musgrave)
ἀρχούς Karsten: πομποὺς ἀρχᾶς Rauchenstein 125 δ᾽ οὖν εἶπε F τεράιζων
schol. M: τεράϊξζων M: τεράζων rell. 120 πρόσθε τὰ VFTr: προσθετὰ Μ δημιοπληθῆ
codd. : corr. O. Müller μοῖρ᾽ ἀλαπάξει codd, : recte distinxit Elmsley 131 dya
Hermann: ἄτα codd. - 131 sq. κνεφάσειε Tr 132 προτυφθὲν Tr 134 οἴκτωι
Scaliger : οἴκωι codd. 136 πτανοῖσιν M : -οἷσι V : -ois FTr κυσὶν M 137
πτάκα M: πτάωνκα V (cf. p.68q.): πτῶκα FTr 139 cf. ad 121 140 τόσσων M
á FTr: om. MV 141 Et. gen. B = Etym. M. 377. 37 ἔρσαι (spectat ad ε 222)
. καὶ Αἰσχύλος ἐν Ἀγαμέμνονι τοὺς σκύμνους τῶν λεόντων δρόσους κέκληκε, τοῦτο
μεταφράζων δρόσοις Tr: δρόσοισιν MVF ἀέπτοις VF: ἀέπτοισι Tr: ἀέλπτοις
M, ubi tamen schol. τοῖς ἕπεσθαι τοῖς γονεῦσι (μὴ δυναμένοις λεόντων Pearson ex
Etym. M.: ὄντων MV: om. FTr 142 φιλομάστοις ex φιλόμάτοις corr. M
143 Aristoph. Byz. p. 112 Nauck ὑστρίχων δὲ καὶ τῶν τοιούτων ὄβρια καὶ ὀβρίκαλα, ὧν
χρῆσις καὶ map’ Αἰσχύλωι ἐν Ἀγαμέμνονι ὀβρικάλοις ΕΣ 144 αἰνεῖ Lachmann: αἰτεῖ
codd. κρᾶναι FTr: xpava MV 145 στρουθῶν MV: τῶν στρουθῶν FTr — glossema
delevit Porson ; quid poeta hoc loco scripserit nescimus 149 Sq. xpovias post
ἐχενηίδας Tr 150 ἐχενηΐδας codd, ἀπλοΐας MV: ἀπλοΐδας FTr
98
prevented from running her final course. Say ‘woe, woe!',
but may the good prevail!

Now when the wise seer of the army saw the two Atridae,
twain in temper, he knew the warlike devourers of the hare
for the conducting chiefs ; and thus he spake interpreting the
portent: 'In course of time this expedition captures Priam's
town ; and all the herds before the walls, the plentiful posses-
sions of the people, shall fate lay waste with violence: let
only no envious grudge from the gods strike beforehand and
overcloud the great bit for Troy's mouth, the army on its
campaign. For out of pity pure Artemis bears a grudge
against the winged hounds of her father which slaughter for
a sacrifice the poor trembling hare with her young before the
birth; and she loathes the feast of the eagles.’ Say ‘woe,
woe!', but may the good prevail!

"The Fair One, kindly though she be towards the helpless!

dewdrops of ravening lions and pleasant to the suckling
young of all creatures that roam the field, yet consents to
fulfil that which the encounter portends for this undertaking,
the signs that appear favourable yet betoken disapproval.
But I invoke the blessed Healer? (to intervene and see to it)
that she prepare not against the Danaans a delay in port,
caused by adverse winds, time-consuming, ship-detaining,
in her eagerness to bring about another sacrifice, one without
precedent and law, without a feast, a worker of quarrels,

1 Literally perhaps 'unable to follow' ; see the commentary.

2 This is not a precise rendering. I cannot think of a satisfactory translation of ἐήεον
Παιᾶνα, which is derived from the invocation ἰὴ Παιάν.

99
 νεικέων τέκτονα σύμφυτον οὐ δεισ-
ἠνορα- μίμνει γὰρ φοβερὰ παλίνορτος
οἰκονόμος δολία, μνάμων Μῆνις τεκνόποινος.᾽ 155
τοιάδε Κάλχας ξὺν μεγάλοις ἀγαθοῖς ἀπέκλαγξεν
μόρσιμ᾽ ἀπ᾽ ὀρνίθων ὁδίων οἴκοις βασιλείοις"
τοῖς δ᾽ ὁμόφωνον
αἴλινον αἴλινον εἰπέ, τὸ δ᾽ εὖ νικάτω. —c

Ζεύς, ὅστις πότ᾽ ἐστιν, εἰ τόδ᾽ αὖ- 160

τῶι φίλον κεκλημένοι,
τοῦτό νιν προσεννέπω.
οὐκ ἔχω προσεικάσαι
πάντ᾽ ἐπισταθμώμενος
πλὴν Διός, εἰ τὸ μάταν ἀπὸ φροντίδος ἄχθος 165
χρὴ βαλεῖν ἐτητύμως. —

οὐδ’ ὅστις πάροιθεν ἦν μέγας

παμμάχωι θράσει βρύων,
οὐδὲ λέξεται πρὶν ὦν’ 170
ὃς δ᾽ ἔπειτ᾽ ἔφυ, τρια-
κτῆρος οἴχεται τυχών.
Ζῆνα δέ τις προφρόνως ἐπινίκια κλάφων
τεύξεται φρενῶν TO Tv — 175

τὸν φρονεῖν βροτοὺς ὁδώ-

σαντα τῶι πάθει μάθος
θέντα κυρίως ἔχειν.
στάξει δ᾽ ἀνθ᾽ ὕπνου πρὸ καρδίας
μνησιπήμων πόνος: καὶ παρ᾽ ἄ- 180
κοντας ἦλθε σωφρονεῖν.
δαιμόνων δέ που χάρις βιαίως
σέλμα σεμνὸν ἡμένων. —

καὶ τόθ᾽ ἡγεμὼν ὁ πρέ-

σβυς νεῶν ᾿Αχαιπκῶν, 185
μάντιν οὔτινα ψέγων,

MVFTr
153 σύμφυτον MFTr: συμμενεῖ φυτόν V (‘videtur librarius μενεῖ, interpretationem ad
μίμνει adscriptam, textui intulisse’ Hermann) 154 γὰρ om. FTr 156
ἀπέκλαιξεν M 157 oixos F 158 τοῖσδε ὠμόφωνον V ὁμόφωνον, sed,
ut videtur, primo o in rasura ex w facto, M 159 cf. ad 121 163 προσεικάσαι
ex προσηκᾶσαι corr. M 165 τὸ Pauw: τόδε MVF: γε Tr 170 οὐδὲ λέξεται
Ahrens: οὐδὲν λέξαι ΜΥῈ : οὐδέν τι λέξαι Tr 177 μάθος (sed super os scr. η) F
179 ἀνθ᾽ ὕπνου Emperius : ἔν (ἕν M) θ᾽ ὕπνωι codd. 182 δέ που FTr: δὲ ποῦ MV
100
born in the house and grown one with it, without fear of the
husband ; for there abides a terrible, ever re-arising, treacher-
ous housekeeper ; unforgetting, child-ayenging Wrath.’ Such
were the fated happenings which, together with great
blessings, Calchas cried to the royal house (as portended)
from the birds on the way; in harmony therewith say ‘woe,
woe!’, but may the good prevail!

Zeus, whoever he be—if to be called and invoked by this

name is pleasing to him, even thus do I address him. I have
nothing whereto to liken him, weighing all in the balance,
nothing save Zeus if there is need to cast the burden of vain
thought from the care-laden mind in real truth.

And he who aforetime was mighty, swelling with the bold-

ness of a victor in every contest, shall not even be reckoned,
since he is of the past; and he who afterward came into
being met his thrower and is gone. But anyone who gladly
shouts ‘Hail to Zeus the victor!’ shall hit full on the target
of understanding:

it is Zeus who has put men on the way to wisdom by

establishing as a valid law ‘By suffering they shall win under-
standing'. Instead of sleep there trickles before the heart
the pain of remembrance of suffering: even to the unwilling
discretion comes. There is, I think, a blessing from the gods,
who, using force, sit on the dread bench of the helmsman.

So it was then with the elder leader of the Achaean fleet ;

not finding fault with any prophet, letting his spirit go with
IOI
 ἐμπαίοις τύχαισι συμπνέων,
εὖτ᾽ ἀπλοίαι κεναγγεῖ βαρύ-
vovT' ᾿Αχαιικὸς λεώς,
Χαλκίδος πέραν ἔχων παλιρρό- 190
χθοις ἐν Αὐλίδος τόποις" ΞΞΞ

τινοαὶ δ᾽ ἀπὸ Στρύμονος μολοῦσαι

κακόσχολοι νήστιδες δύσορμοι,
βροτῶν ἄλαι, ναῶν (Te) καὶ 195
πεισμάτων ἀφειδεῖς,
παλιμμήκη χρόνον τιθεῖσαι
τρίβωι κατέξαινον ἄνθος ᾿Αργεί-
cov’ ἐπεὶ δὲ καὶ πικροῦ
χείματος ἄλλο μῆχαρ
βριθϑύτερον πρόμοισιν 200
μάντις ἔκλαγξεν, Trpogépov
"Ἄρτεμιν, ὥστε χθόνα βάκ-
TROIS ἐπικρούσαντας ᾿Ατρεί-
δας δάκρυ μὴ κατασχεῖν᾽ —

ἄναξ δ᾽ ὁ πρέσβυς τόδ᾽ εἶπε φωνῶν' 205

“ βαρεῖα μὲν κὴρ τὸ μὴ πιθέσθαι,
βαρεῖα δ᾽, εἰ τέκνον δαΐ-
ξω, δόμων ἄγαλμα,
μιαίνων παρθενοσφάγοισιν
ῥείθροις πατρώιους χέρας πέλας βω- 210
μοῦ. τί τῶνδ᾽ ἄνευ κακῶν;
πῶς λιπόναυς γένωμαι
ξυμμαχίας ἁμαρτών;
παυσανέμου γὰρ θυσίας
παρθενίου θ᾽ αἵματος ὀρ- o 215
γᾶι περιόργως ἐπιθυ- |
μεῖν θέμις. εὖ γὰρ ein. =

ἐπεὶ δ᾽ ἀνάγκας ἔδυ λέπαδνον

MVFTr
187 τύχαις F ovumvea Tr 190 sq. παλιρρόθοις (παλλι- F) codd. : corr. Ahrens
195 7e suppl. Porson 197 πολυμήκη (superscr. παλιμμ F) VF 198 κατέξενον
(super alterum € superscr. at) M: κατάξενον V 200 πρόμοισι VF 20I ἔκλαγξε
codd.: ν add. Porson (cf. 701 et vol. iii Append. E) 206 πιθέσθαι Turnebus:
πειθέσθαι MV: πείθεσθαι FTr 209 παρθενοσφάγοισιν M : -οισι VTr: -oıs F 210
ῥείθροις Tr: ῥεέθροις rell. πατρώας V βωμοῦ πέλας codd.: transposuit Blomfield
212 πῶς λιπόναυς Tr: τί πῶς λιπόναυς (λειπ- Ἐ) re rell. 215 sq. ὀργᾶι (-à VF)
MVF : yp. (ex yàp correxit Orelli) αὐδᾶ in marg. M: αὐδᾷ Tr 216 περιόργως hoc
accentu codd. 217 θέμις γὰρ εὖ FTr 218 λέδαπνον V
102
the blasts of fortune that fell upon him, when the Achaean
folk were sore pressed by famishing delay in port, while they
held the coast over against Chalcis, in the region of Aulis,
where the tides roar to and fro;

and gales coming from the Strymon, with harmful leisure,

starvation, bad lingering in port, wandering of the men,
(gales) unsparing of ships and cables, by lengthening the time
over again were wasting and wearing away the flower of the
Argives; and when the prophet had cried out to the chiefs
another and more grievous remedy for the sore storm,
revealing Artemis as cause, so that the Atridae struck the
ground with their staffs and could not keep back their
tears,—

then it was that the elder chief spake thus and said: ‘A
heavy doom indeed is disobedience, but heavy, too, if I rend
my child, the delight of my house, defiling a father’s hands
with streams from the slaughtering of a virgin at the altar’s
side. Which of these courses is without evil? How can I fail
in my duty to the alliance and thus become a deserter of the
fleet ? (I cannot,) for it is right and lawful that one should
with over-impassioned passion crave the sacrifice to stay the
winds, the blood of the virgin. (It shall be done;) for (my
hope is:) may all be well.’

And when he had slipped his neck through the strap of

103
 φρενὸς πνέων δυσσεβῆ τροπαίαν
ἄναγνον ἀνίερον, τόθεν 220
τὸ παντότολμον φρονεῖν μετέγνω.
βροτοὺς θρασύνει γὰρ αἰσχρόμητις
τάλαινα παρακοττὰ πρωτοπήμων. ἔτλα δ᾽ οὖν
θυτὴρ γενέσθαι θυγατρός, 225
γνυναικοποίνων πολέμων ἀρωγὰν
καὶ προτέλεια ναῶν. —

λιτὰς δὲ καὶ κληδόνας πατρώιους

παρ᾽ οὐδὲν αἰῶ te παρθένειον
ἔθεντο φιλόμαχοι βραβῆς. 230
φράσεν δ᾽ ἀόξοις πατὴρ μετ᾽ εὐχὰν
δίκαν χιμαίρας ὕπερθε βωμοῦ
πέπλοισι περιπετῆ παντὶ θυμῶι προνωπῆ
λαβεῖν ἀέρδην, στόματός
τε καλλιπρώιρου φυλακᾶι κατασχεῖν 236
φθόγγον ἀραῖον οἴκοις, ΞΞΞ

βίαι χαλινῶν τ᾽ ἀναύδωι μένει.

κρόκου βαφὰς δ᾽ ἐς πέδον χέουσα
ἔβαλλ᾽ ἕκαστον θυτήρων ἀπ᾽ ὄμ- 240
ματος βέλει φιλοίκτωι,
πρέπουσα τὼς ἐν γραφαῖς, προσεννέπειν
θέλουσ᾽, ἐπεὶ πολλάκις
πατρὸς κατ᾽ ἀνδρῶνας EUTPATTEZOUS :
ἔμελψεν, &yvài δ᾽ ἀταύρωτος αὐδᾶι πατρὸς
φίλου τριτόσπονδον εὔποτμον παι- 245
ὥνα φίλως ἐτίμα. —

τὰ δ᾽ ἔνθεν οὔτ᾽ εἶδον οὔτ᾽ ἐννέπω":

τέχναι δὲ Κάλχαντος οὐκ ἄκραντοι.
Δίκα δὲ τοῖς μὲν παθοῦσιν μαθεῖν 250
ἐπιρρέτει. τὸ μέλλον (5°)
MVFTr
219 δυσεβῆ F 222 βροτοὺς θρασύνει Spanheim : βροτοῖς. θρασύνει codd. 227
προτέλεια MFTr (ex πρωτ- corr. F): προτέλειαν V 229 αἰῶ τε O. Müller : αἰῶνα
codd. παρθένιον MV 230 βραβεῖς VITr 231 φράσε VF δ᾽ ἐν ὄσσοις Tr
232 χιμαίρας VF Tr: χειμ- Μ ὅὕπερθεν V 234 βαλεῖν Tr ἀέρδειν F 236
φυλακὰν codd. : corr. Blomfield 238 τ MVF:8 Tr ἀναύδων F 239 δ᾽ om.
Tr xeovo’ F 240 éBal Y éxaror V 240 sq. βέλει ἀπ᾿ ὀμμάτων φιλοίκτῳ
Tr 242 τὼς Maas: θ᾽ ὡς codd. 244 ἁγνᾷ Ir: ἁγνὰ MVF αὐδᾷ FTr:
αὐδὰ MV 245 εὔποτμον ex εὐπόταμον factum M: εὐπόταμον VF: εὔποτον Tr
245 sq. παιῶνα Hartung: αἰῶνα codd. 250 παθοῦσιν Tr: παθοῦσι rell.
251 δ᾽ suppl. Elmsley 251 sq. inter τὸ μέλλον et ἐπεὶ habent τὸ δὲ προκλύειν MVE
(verba interpolata ab antecedentibus verbis brevi intervallo separavit m)
104
compulsion’s yoke, and the wind of his purpose had veered
about and blew impious, impure, unholy, from that moment
he reversed his mind and turned to utter recklessness. For
men are emboldened by base-counselling wretched infatua-
tion, the beginning of woe. However, he brought himself to
become the sacrificer of his daughter, in aid of a war for
avenging the loss of a woman and as a preliminary rite on
behalf of the fleet.

Her prayers, her cries of ‘Father!’, and her virgin youth,

the war-thirsty commanders counted as nought. Her father,
after prayer, gave word to the attendants to take her
resolutely as she drooped forward, wrapped round in her
robes, and to hold her, like a kid, above the altar, and, by
guarding her fair mouth, to check a sound that would be
a curse upon the house,

by force and the voiceless power of the bridle. And she, as

she let fall to the ground her saffron-dyed raiment, smote
each one of the sacrificers with a pitiful arrow from her eye,
standing out as in a picture, wishing to speak to them by
name; for many a time in her father’s hall, where the men
were given rich feasts, she had sung, and with a chaste voice
the virgin lovingly celebrated her beloved father’s blissful
paean at the third libation.

What followed I saw not, neither do I tell; but the craft

of Calchas does not lack fulfilment. Justice weighs out
understanding to those who have gone through suffering.
And the future—when it comes thou mayst hear of it ; let it
105
 ἐπεὶ γένοιτ᾽ ἂν KAUOIS" προχαιρέτω᾽
ἴσον δὲ τῶι προστένειν"
τορὸν γὰρ ἥξει σύνορθρον αὐγαῖς.
πέλοιτο δ᾽ οὖν τἀπὶ τούτοισιν εὖ πρᾶξις, ὡς 255
θέλει τόδ᾽ ἄγχιστον ᾿Απίας γαί-
ας μονόφρουρον ἕρκος. ===

- ἥκω σεβίφων σὸν Κλυταιμήστρα κράτος"

δίκη γάρ ἐστι φωτὸς ἀρχηγοῦ τίειν
γυναῖκ᾽ ἐρημωθέντος ἄρσενος θρόνον. 260
σὺ δ᾽ εἴ τι κεδνὸν εἴτε μὴ πεπυσμένη
εὐαγγέλοισιν ἔλπίσιν θνηπολεῖς,
κλύοιμ᾽ ἂν εὔφρων’ οὐδὲ σιγώσηι φθόνος.

KAYTAIMHZTPA
εὐάγγελος μέν, ὥσπερ ἡ παροιμία,
ἕως γένοιτο μητρὸς εὐφρόνης πάρα. 265
πεύσηι δὲ χάρμα μεῖτον ἐλπίδος κλύειν"
Πριάμου γὰρ ἡιρήκασιν ᾿Αργεῖοι πόλιν.
ΧΟ. πῶς φής; πέφευγε τοῦπος ἐξ ἀπιστίας.
KA. Τροίαν ᾿Αχαιῶν οὖσαν À τορῶς λέγω;
ΧΟ. χαρά μ᾽ ὑφέρπει δάκρνον ἐκκαλουμένη. 270
ΚΛ. εὖ γὰρ φρονοῦντος ὄμμα σοῦ κατηγορεῖ.
ΧΟ. ἢ γάρ τι πιστόν ἐστι τῶνδέ σοι τέκμαρ;
ΚΛ. ἔστιν, τί δ᾽ οὐχί; μὴ δολώσαντος θεοῦ.
ΧΟ. πότερα δ᾽ ὀνείρων φάσματ᾽ εὐπειθῆ σέβεις;
KA. οὐ δόξαν ἂν λάβοιμι βριφούσης φρενός. 275
ΧΟ. ἀλλ᾽ ἦ σ᾽ ἐπίανέν τις ἄπτερος φάτις;
MVFTr
252 ἐπιγένοιτ᾽ M 253 loov MV 254 σύνορθρον Wellauer : συνορθὸν M :
σὺν ἀρθον V : σύναρθρον ΕἼΤ αὐγαῖς Hermann: αὐταῖς (äüraîs Tr) codd. 255
τούτοις F εὔπραξις codd.: distinxit Lobeck 257 ἕρκος MV: ἔρκος FTr
258 sqq. in scaena sequenti veras personarum notas restituit Casaubon 258
paragraphum (—) praefixit M, ἄγγελος mVF, ἄγγελος φύλαξ Tr κλυταιμήστρα
MTr: κλυταιμνήστρα ΝΕ 261 εἴ τι Auratus: εἰ τὸ M: εἴτε mVFTr 262
ἐλπίσειν M: ἐλπίσι V. - 263 oratio continuatur in M, «Avr. praefixerunt VFTr
σιγώσῃ V : σιγώιση (-σηι m) M: σιγῶντι (propter notam «Aur, praefixam) FTr 264
κλυταιμ. et paragraphum praefixit M, ἄγγελος VFTr 264 sq. Eustathius
ad A 9, p. 22. 32, postquam Soph. Trach. 94 sqq. attulit, addit haec : παρέοικε δὲ πάντως
τῆι τοιαύτηι τοῦ Σοφοκλέους ἐννοίαι καὶ Αἰσχύλου ἐν Ἀγαμέμνονι τὸ ‘ εὐάγγελος ἠὼς γένοιτο
μητρὸς εὐφρόνης mápa ' 266 paragraphum praefixit M πεύσει V 268-80
nihil nisi paragraphos praefixit M: «Avr. (268, 270, etc.) et dyy. (269, 271, etc.) prae-
fixerunt VFTr 269 τρόαν V 270 versum affert schol. Hom. 7 471 Σοφοκλῆς
“χαρά μ᾽ (sic Dindorf; χάρμ M, xdpu' vulgo) ὑφέρπει δάκρυον ἐκκαλουμένη ᾿; cf. Eust.
p. 1872. 65 (χάρμα... ἐκκαλούμενον) 271 φρονούσης (cum dyy. praefixum sit; cf.
263) FTr 272 1j γάρ τι Karsten: τί yàp τὸ codd. ἐστιν M 273 ἔστι V
274 εὐπειθεῖ in -05 corr. M 276 émiavé V cf. Hesychius ἄπτερος" αἰφνίδιος" παρὰ
“Ὁμήρωι ὃ προσηνὴς ἢ ταχύς. «Αἰσχύλος Ἀγαμέμνονι
106
be greeted in advance—but that is equal to being lamented
in advance, for it will arrive clear together with the rays of
dawn.
Clytemnestra appears at the door of the house

Howbeit, may for the rest achievement be prosperous, as

is the wish of this nearest sole-guarding bulwark of the land
of Apia.
I am come, Clytemnestra, reverencing thy power ; for it is
right to hold in honour the wife of the sovereign when the
man’s throne is left empty. But now, whether it is because
thou hast received some happy news, or whether thou hast
not and it is in hope of good tidings that thou offerest sacri-
fice, I would in loyalty hear; but if thou keep silence, I do
not grudge it.
Clytemnestra. Bearer of good tidings may the morning be,
coming—as the proverb goes—from (and taking after) its
mother night! But thou shalt learn of a joy which to hear is
beyond all hope: the Argives have taken Priam’s town.
Chor. How sayest thou? thy words have escaped me,
since I could not believe them.
Clyt. (I say) that Troy is in the hands of the Achaeans;
do I speak clear?
Chor. Joy steals over me, calling forth a tear.
Clyt. Yes, thine eye gives evidence of thy loyal heart.
Chor. Hast thou some sure proof of this?
Clyt. I have—assuredly I have—unless the god has
played me false.
Chor. Is it to persuasive visions of dreams that thou
payest regard ?
Clyt. I would let no one sell me the mere fancy of a
slumbering mind.
Chor. Can it be then that some swift-winged rumour has
made thy wit grow thick ?
107
 ΚΛ. παιδὸς νέας as κάρτ᾽ ἐμωμήσω φρένας.
ΧΟ. ποίου χρόνου δὲ καὶ πεπόρθηται πόλις;
ΚΛ. τῆς νῦν τεκούσης φῶς τόδ᾽ εὐφρόνης λέγω.
XO. καὶ τίς τόδ᾽ ἐξίκοιτ᾽ ἂν ἀγγέλων τάχος; 280
ΚΛ, Ἥφαιστος, Ἴδης λαμπρὸν ἐκπέμπων σέλας.
φρυκτὸς δὲ φρυκτὸν δεῦρ᾽ ἀπ᾽ ἀγγάρου πυρὸς
ἔπεμπεν Ἴδη μὲν πρὸς Ἕρμαιον λέπας
Λήμνου" μέγαν δὲ πανὸν ἐκ νήσου τρίτον
᾿Αθῶιον altros Ζηνὸς ἐξεδέξατο᾽' 285
ὑπερτελής τε πόντον Mote νωτίσαι
ἰσχὺς πορευτοῦ λαμπάδος πρὸς ἡδονὴν
* * * * * * * * *

πεύκη, τὸ χρυσοφεγγὲς ὥς τις ἥλιος

σέλας παραγγείλασα Μακίστου σκοττοῖς.
ὁ δ᾽ οὔτι μέλλων οὐδ᾽ ἀφρασμόνως ὕπνωι 290
νικώμενος παρῆκεν ἀγγέλου μέρος, |
ἑκὰς δὲ φρυκτοῦ φῶς ἐπ᾽ Εὐρίπου ῥοὰς
Μεσσαπίου φύλαξι σημαίνει μολόν.
οἱ δ᾽ ἀντέλαμψαν καὶ παρήγγειλαν πρόσω,
γραίας ἐρείκης θωμὸν ἅψαντες πυρί. 295
σθένουσα λαμτὰς δ᾽ οὐδέπω μαυρουμένη
ὑπερθοροῦσα πεδίον ᾿Ασωποῦ δίκην
φαιδρᾶς σελήνης πρὸς Κιθαιρῶνος λέτας
ἤγειρεν ἄλλην ἐκδοχὴν πομτποῦ πυρός.
φάος δὲ τηλέπομπον οὐκ ἠναίνετο 300
φρουρά, πλέον καίουσα τῶν εἰρημένων’
λίμνην δ᾽ ὑπὲρ γοργῶπιν ἔσκηψεν φάος
ὄρος τ᾽ ἐπ᾽ αἰγίπλαγκτον ἐξικνούμενον
ὥτρυνε θεσμὸν ἐμὴ χαρίφεσθαιΐ πυρός.
πέμπουσι δ᾽ ἀνδαίοντες ἀφθόνωι μένει 305

MVFTr
277 ὡς VFTr: ὡς M 280 καὶ πῶς 106° Tr 281 κλυταιμ quod iuxta
paragraphum praefixit M delevit m ct &yye(Aos) apposuit; ἄγγελος praefixerunt rell.
282-5 ex Aelio Dionysio (p. 87 Schwabe; cf. Eust. ad r 28 p. 1854. 27) toti afferuntur
in Photio Berol. p. 10. 22; cf. Et. gen. B — Etym. M. p. 7. 18 Αἰσχύλος γοῦν ἐν Ἀγαμέμνονι
τὸν ἐκ διαδοχῆς πυρσὸν ‘ am’ ἀγγάρου πυρὸς ’ ἔφη, Suid. s.v. ἄγγαροι 282 ἀγγάρου
Etym. M., Phot., Suid. (unde Canter hunc locum correxit), Eust.: ἀγγέλου codd.
283 ἕρμαιον M : ἕρματον V : ἑρμαῖον ΕἼ τ 284 cf. Athen. 15. 700 e πρότερος δὲ
τούτων (Menandrum et Diphilum dicit) Αἰσχύλος ἐν Ἀγαμέμνονι μέμνηται τοῦ πανοῦ
πανὸν (quod ex Athenaeo restituit Casaubon) Phot. Berol.: φανὸν codd. 285
Ἀθῶιον Blomfield : ἄδθωον MFTr (de accentu in V posito mihi non constat) 286
ὑπερτελής MV : ὑπεὶρ Ans FTr et Triclinii schol. vet. (cf. comm.) post 287 lacunam
statuit Paley 288 yAvoodeyyes V 289 σκοπάς codd.: corr. Turnebus
292 evpirmoy MV 293 μεσαπίου VFTr μολὼν F 294 ot à F 295
ἐρείκης mVFTr: ἐρίκης M 297 πεδίον ἀσωποῦ [Tr : παιδίον ὠποῦ MV
303 αἰγίπλακτον V 304 μὴ χαρίζεσθαι nondum sanatum : δὴ χαρίζεσθαι ΤΥ
108
 Clyt. Thou judgest my intellect to be faulty indeed, as
if I were a young child.
Chor. And since what time has the town been destroyed ?
Clyt. Since the night, I say, that has just now given birth
to the light of this morning.
Chor. And what messenger is there that could arrive with
such speed as this ?
Clyt. Hephaistos, sending forth from Ida a bright
radiance. And beacon ever sent beacon hither by means of
the courier fire: Ida (sent it) to the rock of Hermes in
Lemnos ; and a huge torch from the island was taken over in
the third place by Zeus’ peak of Athos; and paying more
than what was due (?), so as to skim the back of the sea (?),
the strength of the travelling torch joyously (went on...)
the pine-tree blaze, after (?) transmitting, like a sun, its
golden radiance to the look-out of Makistos. And he (i.e.
Makistos), not dallying nor heedlessly overcome by sleep,
did not neglect his share in the messenger’s duty, and afar,
over the streams of Euripus, the beacon’s light gave the
watchers of Messapion the sign of its arrival. They kindled
an answering flare and sent the tidings onward, by setting
fire to astack of aged heath. And the vigorous torch, not yet
growing dim, leaped, like the shining moon, over the plain
of Asopusto the rock of Kithairon and there waked a new
relay of the sender fire. And the far-sent light was not re-
jected by the watch-post, which burned more than it had
been ordered ; and the light shot down over the Gorgon-eyed
lake and reaching the mountain of the roaming goats urged
(the watch-post) not to neglect (?)! the ordinance of the fire.
And they with stintless might kindled and sent on a great
1 Text uncertuin.

109
 φλογὸς μέγαν πώγωνα, καὶ Σαρωνικοῦ
πορθμοῦ κάτοττον πρῶν᾽ ὑπερβάλλει πρόσω
φλέγουσα᾽ {εἴτ᾽ 7 ἔσκηψεν {εἴτ᾿ ἀφίκετο
᾿Αραχναῖον altros, ἀστυγείτονας σκοτάς"
κἄπειτ᾽ ᾿Ατρειδῶν ἐς τόδε σκήτττει στέγος 310
φάος τόδ᾽ οὐκ ἄπατιπον ᾿Ιδαίου πυρός.
τοιοίδε τοί μοι λαμπαδηφόρων νόμοι,
ἄλλος παρ᾽ ἄλλου διαδοχαῖς πληρούμενοι:
νικᾶι δ᾽ ὁ πρῶτος καὶ τελευταῖος δραμών.
τέκμαρ τοιοῦτον σύμβολόν τέ σοι λέγω | 315
᾿ ἀνδρὸς παραγγείλαντος ἐκ Τροίας ἐμοί,
(XO.) θεοῖς μὲν αὖθις, ὦ γύναι, προσεύξομαι"
λόγους δ᾽ ἀκοῦσαι τούσδε κἀποθαυμάσαι
διηνεκῶς θέλοιμ᾽ ἄν, ὡς λέγεις, πάλιν.
KA. Τροίαν ᾿Αχαιοὶ τῆιδ᾽ ἔχουσ᾽ ἐν ἡμέραι. 320
οἶμαι βοὴν ἄμεικτον ἐν πόλει πρέπειν.
ὄξος τ᾽ ἄλειφά τ᾽ ἐγχέας ταὐτῶι κύτει.
διχοστατοῦντ᾽ ἂν οὐ φίλως προσεννέτοις’
καὶ τῶν ἁλόντων καὶ κρατησάντων δίχα
φθογγὰς ἀκούειν ἔστι, συμφορᾶς διπλῆς" 325
οἱ μὲν γὰρ ἀμφὶ σώμασιν πεπτωκότες
ἀνδρῶν κασιγνήτων τε καὶ φυταλμίων
παῖδες γερόντων οὐκέτ᾽ ἐξ ἐλευθέρου
δέρης ἀποιμώφουσι φιλτάτων pópov:
τοὺς δ᾽ αὖτε νυκτίπλαγκτος ἐκ μάχης πόνος 330
νήστεις πρὸς ἀρίστοισιν ὧν ἔχει πόλις
τάσσει πρὸς οὐδὲν ἐν μέρει τεκμήριον,
ἀλλ᾽ ὡς ἕκαστος ἔσπασεν τύχης πάλον
ἐν αἰχμαλώτοις Τρωικοῖς οἰκήμασιν
ναίουσιν ἤδη, τῶν ὑπαιθρίων πάγων 335
δρόσων τ᾽ ἀπαλλαχθέντες, ὡς δ᾽ εὐδαίμονες
ἀφύλακτον εὑδήσουσι πᾶσαν εὐφρόνην.
306-10 MVFTr z311-37 VF Tr
306 péya V 307 κάτοπτρον codd.: corr. Canter πρῶνα (ex quo πρῶν᾽ fecit m) M
307 sq. ὑπερβάλλειν. ... φλέγουσαν codd. : corr. Schütz (ὑπερβάλλει iam Casaubon)
308 ely’... εἶτ᾽ vix sana; haec fere fuisse putes : κἀπέσκηψεν ὡς ἀφίκετο 309 αἶπος V
310 eis V τόδε σκήτιτει FTr: τόδ᾽ ἐνοκήπτει V : τόγε σκήπτει M post versum
310 deficit M; cf. ad 1067 . 312 τοιοίδε roi μοι Schütz : τοιοίδ᾽ ἕτοιμοι (ἔτυμοι ΕἾ
VFTr 315 τοιοῦτον FTr: τοι οὔτοι (sic) V 317 ΧΟ. Canter: om. codd.
319 λέγεις V : λέγοις FTr 320 KA. hic posuit Casaubon, ante 321 codd. τροίην
321 ἄμικτον codd. 322 üdeidar’ V ἐκχέας codd. : corr. Canter
323 δειχοτατοῦντ᾽ V 324 ἁλόντων δὲ καὶ F, sed δὲ lineola transversa deletum
325 ἐστὶν F 326 σώμασιν 329 ἀποιμώξουσι (£ ex ζ factum; cf. 443, 785,
1599) F 330 νυκτίπλακτος Tr 331 νήστεις Tr: νῆστις F: νήστισι V ἀρρίστοισιν
(altero p infra lineam addito) cum glossa διὰ τὸ μέτρον Tr 333 éoraoe V 334
οἰκήμασι V 336 ἀπαλλαγέντες FTr δ᾽ εὐδαίμονες Casaubon : δυσδαίμονες codd.
εἴ schol. vet. Tr
IIO
beard of flame, and it passed beyond the promontory that
looks down on the Saronic straits, blazing onward, and shot
down when (?)! it reached the Arachnaean peak, the watch-
post that is neighbour to our city; and then it shot down
here to the house of the Atridae, this light, the genuine
offspring of its ancestor, the fire from Mount Ida. Such, thou
seest, are the rules I arranged for my torch-bearers,—one
from another in succession supplied to the full; and victor
is he who ran first and last. Such is the proof and token that
I give thee, transmitted to me by my husband from Troy.
Chor. To the gods, lady, my prayers shall be addressed
hereafter ; but for this tale—I would fain hear it again com-
plete to the end, and wonder at it, the tale as thou tellst it.
Clyt. Troy is this day in the hands of the Achaeans.
Methinks cries that will not blend are clearly heard in the
city. When thou hast poured vinegar and oil into the same
vessel, thou wouldst address them as beings? at variance in
no friendly manner; so one may hear the voices of the
captured and the conquerors separately, voices of the
different fortunes that have befallen them: the one, having
flung themselves down upon the bodies of husbands and of
brothers, and children upon those of aged men whose off-
spring they are, from throats no longer free are bewailing the
death of their beloved ones; while the others—the battle’s
night-roving toil sets them famished down to make their
breakfast on what is in the town—not after any billet in due
apportionment ; but just as each drew the lot of chance, they
are now lodging in the captured Trojan houses, delivered
from the frosts and dews of the open sky, and like men of
blessed fortune they will sleep all the night without a watch
to keep. And if they reverence the gods who are the city-
1 Text uncertain. 2 i.e. ‘apply to them the name of being . . .’
III
 εἰ δ᾽ εὐσεβοῦσι τοὺς πολισσούχους θεοὺς
τοὺς τῆς ἁλούσης γῆς θεῶν θ᾽ ἱδρύματα,
οὔ T&v ἑλόντες αὖθις ἀνθαλοῖεν ἄν. 340
ἔρως δὲ μή τις πρότερον ἐμπίτττηι στρατῶι
πορθεῖν ἃ μὴ χρὴ κέρδεσιν νικωμένους"
δεῖ γὰρ πρὸς οἴκους νοστίμου σωτηρίας,
κάμψαι διαύλου θάτερον κῶλον πάλιν.
θεοῖς δ᾽ ἀναμπλάκητος εἰ μόλοι στρατός, 345
jeypnyopovf τὸ πῆμα τῶν ὀλωλότων
γένοιτ᾽ ἄν, εἰ πρόσπαια μὴ τύχοι κακά.
τοιαῦτά τοι γυναικὸς ἐξ ἐμοῦ κλύεις.
τὸ δ᾽ εὖ κρατοίη, μὴ διχορρόπως ἰδεῖν"
πολλῶν γὰρ ἐσθλῶν τὴν ὄνησιν εἱλόμην, 350
ΧΟ. γύναι, κατ᾽ ἄνδρα σώφρον᾽ εὐφρόνως λέγεις.
ἐγὼ δ᾽ ἀκούσας πιστά cou τεκμήρια
θεοὺς προσειπεῖν εὖ παρασκευάξζομαι"
χάρις γὰρ οὐκ ἄτιμος εἴργασται πόνων.

- ὦ Ζεῦ βασιλεῦ καὶ Νὺξ φιλία, 355

μεγάλων κόσμων κτεάτειρα,
fir ἐπὶ Τροίας πύργοις ἔβαλες
στεγανὸν δίκτυον, ὡς μήτε μέγαν
μήτ᾽ οὖν νεαρῶν τιν᾽ ὑπερτελέσαι
μέγα δουλείας 360
γάγγαμον, ἄτης παναλώτου.
Δία τοι ξένιον μέγαν αἰδοῦμαι
τὸν τάδε πράξαντ᾽, ἐπ᾽ ᾿Αλεξάνδρωι
τείνοντα πάλαι τόξον, ὅπως ἂν
μήτε πρὸ καιροῦ μήθ᾽ ὑπὲρ ἄστρων 365
βέλος ἠλίθιον σκήψειεν.

Διὸς πλαγὰν ἔχουσιν εἰπεῖν,

πάρεστιν τοῦτό γ᾽ ἐξιχνεῦσαι.
[ὡς] ἔπραξεν ὡς ἔκρανεν. οὐκ ἔφα τις
338-48 VFTr 349-69 FTr
339 7° ἰδρύματα V 340 οὔ τἂν ἑλόντες Hermann: οὐκ ἀνελόντες V : οὐκ ἄν γ᾽ ἑλόντες
r ἀνθαλοῖεν Auratus : dv θάνοιεν V : αὖ ϑάνοιεν FTr 341 πρότερον FTr:
mpárov V 341 ἐμπίπτη (super scr. ox) F: ἐμπίπτει V : ἐμπέπτοι Tr 342
τὰ μὴ V κέρδεσι VF 344 κάμψαι (super ı scr. e) F 346 ἐγρήγορον
corruptum : fortasse scribendum παρηγορεῖν (Headlaın) 347 τύχη Tr 348
κλύεις V : κλύοις FTr post hunc versum deficit V 351 dyy. (corr. Casaubon),
355 χορός praefixerunt codd. 356 τῶν μεγάλων ΤΥ (scilicet a paroemiaco, ut solet,
abhorrens; cf. comm. ad 1334) 361 cf. Pollux 10. 132 καὶ γάγγαμον, ἀφ᾽ οὗ καὶ
Αἰσχύλος τὸ δύσλυτον κακὸν ἢ δυσεξάλυκτον ἔφη ‘ yéyyauov ἄτης παναλώτον ' 36
ἔχουσιν Tr: ἔχουσαν mutatum in ἔχουσ᾽ F 368 πάρεστι FTr: v addidit Har-
tung τοῦτο γ᾽ Tr: τοῦτ᾽ F 369 prius ὡς delevit Hermann τίς FTr
112
patrons in the conquered land, and the shrines of the gods,
then they will avoid being taken in their turn after taking
the prey. But I fear that a desire may meanwhile seize upon
the army, to ravish what they ought not, overcome by love
of gain ; for they still need to win their way safe home again,
to turn back along the other limb of the double race-course.
But if the army should come home without having offended
against the gods, then it might happen that the harm done
to those who have perished could be assuaged,! if no sudden
disaster should befall.
Here thou hast heard what I, a woman, had to say! But
may the good prevail, so that it may be seen with no uncer-
tainty ; I prefer the enjoyment (of what I have) to wealth of
blessings.
Chor. Lady, thou speakest sagely like a wise and temperate
man. But now that I have heard from thee sure proofs, I am
ready to address the gods with due thanksgiving ; for a suc-
cess that is no inadequate return for the pains has been
achieved. Exit Clytemnestra

Chor. O Zeus the King, and befriending Night, that hast

won us possession of great glories, thou who didst fling on
the Trojan fortress a covering net, so that no one full-grown
nor any of the young could surmount the mighty drag-net of
slavery, of all-catching doom. The great Zeus of Hospitality
I hold in awe, him who has achieved this, long since bending
his bow on Alexandros in such wise that the shaft might not
alight without effect either short of the mark or above the
stars.

It is of the stroke of Zeus that they can tell; this at any

rate can be fully traced. He has achieved it as he decreed.
1 Text uncertain.
4872-1 113 I
 θεοὺς βροτῶν ἀξιοῦσθαι μέλειν 370
ὅσοις ἀθίκτων χάρις
ratoid’” ὁ δ᾽ οὐκ εὐσεβής.
πέφανται δ᾽ ἐγγόνους
ἀτολμήτων ἄρηΐ 375
πνεόντων μεῖφον ἢ δικαίως,
φλεόντων δωμάτων ὑπέρφευ
ὑπὲρ τὸ βέλτιστον. ἔστω δ᾽ ἀπή-
povtov ὥστ᾽ ἀπαρκεῖν
εὖ πραπίδων λαχόντι. 380
οὐ γάρ ἐστιν ἔπαλξις
πλούτου πρὸς κόρον ἀνδρὶ
λακτίσαντι μέγαν Δίκας
βωμὸν εἰς ἀφάνειαν. —

βιᾶται δ᾽ ἁ τάλαινα Πειθώ, 385

προβούλου παῖς ἄφερτος "Aras.
ἄκος δὲ πᾶν μάταιον. οὐκ ἐκρύφθη,
πρέπει δέ, φῶς αἰνολαμπές, σίνος"
κακοῦ δὲ χαλκοῦ τρόπον 390
τρίβωι τε καὶ προσβολαῖς
μελαμπαγὴς πέλει
δικαιωθείς, ἐπεὶ
διώκει παῖς ποτανὸν ὄρνιν,
πόλει πρόστριμμα θεὶς ἄφερτον᾽ 395
λιτᾶν δ᾽ ἀκούει μὲν οὔτις θεῶν,
τὸν δ᾽ ἐπίστροφον τῶν
gat’ ἄδικον καθαιρεῖ,
οἷος καὶ Πάρις ἐλθὼν
ἐς δόμον τὸν "Atpeidäv 400
ἤισχυνε ξενίαν τράπε-
ταν κλοπαῖσι γυναικός. ==

λιποῦσα δ᾽ ἀστοῖσιν ἀσπίστορας

κλόνους tAoyxivous τε Kalt ναυβάτας ὁπλισμούς, 405
FTr
374sq. quid hic lateat nescio 374 eyydvous (super prius y scriptum « Tr) FTr
379 dore κἀπαρκεῖν Tr 380 λαχόντα FTr: corr. Auratus 383 ἐκλακτίσαντι Tr
μέγαν Canter: μεγάλα FTr 386 προβουλόπαις FTr: corr. Hartung 387 ὡς
ἄκος (cf. 369) Tr παμμάταιον FTr: distinxit Musgrave 389 σίνος FTr (superscr.
gl. σέλας Tr): σέλας schol. vet. Tr 391 re Tr: om. F προβολαῖς FTr: corr.
Casaubon 394 moravd» Schütz : πτανὸν F: πτανόν τιν᾽ Tr 395 πρόστριμμα
θεὶς ἄφερτον Wilamowitz : πρόστριμμ᾽ ἄφερτον θεὶς F : πρόστριμμ' ἄφερτον ἐνθεὶς Tr
397 τῶν Klausen : τῶνδε FTr τῶνδ' ἐπίστροφον δὲ Weyrauch 400 ἐς δόμον τῶν
ἀτρειδᾶν F: eis οἶκον τὸν ἀτρειδᾶν Tr 401 τὴν ξενίαν Tr 402 κλοπαῖς F
405 locus nondum sanatus
114
Men have said that the gods disdain to care about such
mortals as trample on the grace of things not to be touched ;
but they who say it lack piety. It has become apparent . . . as
they! pant with a pride greater than is just, when the house
teems with wealth in excess, exceeding what is best. Be there
what is not harmful so that it suffices, for him to whom a
sensible mind has fallen as his lot. For there is no shelter for
a man when once in surfeit of wealth he has kicked the great
altar of Justice out of sight.

Force is employed by that wretched Persuasion, irresistible

child of her whose planning brings about the preliminary
decree, Infatuation. All remedy is vain. Not hidden is the
harm: it shows forth plainly as a terrible bright light;
and like base bronze, when rubbed and battered, so he (i.e.

the guilty man) becomes indelibly black when brought to

justice ; for a boy runs in chase of a flying bird after bringing
an intolerable affliction upon his people; to his prayers no
god will listen, but pulls down the unjust man that is con-

versant with such things. Such a man was Paris, who entered
the house of the Atridae and dishonoured the hospitable
board by theft of the wife.

She, leaving behind her, for the folk of her city, bustling
stir of men with shield and spear and armament of seamen,
τ Or ‘of those who’.

115
 ἄγουσά τ᾽ ἀντίφερνον ᾿Ιλίωι φθορὰν
βεβάκει ῥίμφα διὰ
πυλᾶν, ἄτλατα τλᾶσα πολλὰ δ᾽ ἔστενον
τόδ᾽ ἐννέποντες δόμων προφῆται"
“ἰὼ ἰώ, δῶμα δῶμα καὶ πρόμοι, 410
ἰὼ λέχος καὶ στίβοι piAdvopes.
πάρεστι σιγὰς ἀτίμους ἀλοιδόρους ἀπί-
στους ἀφειμένων ἰδεῖν.
πόθωι δ᾽ ὑπερποντίας
φάσμα δόξει δόμων ἀνάσσειν. 415
εὐμόρφων δὲ κολοσσῶν
ἔχθεται χάρις ἀνδρί,
ὀμμάτων δ᾽ ἐν ἀχηνίαις
ἔρρει πᾶσ᾽ ᾿Αφροδίτα. —

ὀνειρῦφαντων δὲ πενθήμονες 420

πάρεισι δόξαι φέρουσαι χάριν ματαίαν’
μάταν γάρ, εὖτ᾽ ἂν ἐσθλά τις δοκῶν ὁρᾶν,
παραλλάξασα διὰ
χερῶν βέβακεν ὄψις, οὐ μεθύστερον 425
πτεροῖς ὀπαδοῖς ὕπνου {κελεύθοις Τ᾽
τὰ μὲν κατ᾽ οἴκους ἐφ᾽ ἑστίας ἄχη
τάδ᾽ ἐστὶ καὶ τῶνδ᾽ ὑπερβατώτερα'
τὸ πᾶν δ᾽ ἀφ᾽ Ἕλλανος αἴας συνορμένοις ἀπέν-
θεια τλησικάρδιος 430
δόμοις ἑκάστου πρέπει"
πολλὰ γοῦν θιγγάνει πρὸς ἧπαρ’
oUs μὲν γάρ (Tig) ἔπεμψεν
οἶδεν, ἀντὶ δὲ φωτῶν
τεύχη καὶ σποδὸς els ἑκά- 435
στου δόμους ἀφικνεῖται. ==

ὁ χρυσαμοιβὸς δ᾽ “Apns σωμάτων

FTr
407 βεβάκει Keck: Beßaxe FT: Beßaxev Tr 408 ἄτλητα FTr: corr. Dindorf
πολλὰ δ᾽ ἔστενον Tr: πολὺ δ᾽ ἀνέστενον I 410 ἰὼ et δῶμα semel F 412 πάρεστι
Tr: πάρεστιν F σιγὰς ἀτίμους ἀλοιδόρους Hermann : σιγᾶς ἄτιμος ἀλοίδορος FTr
412 34. ἀπίστους Wilamowitz (anno 1889) Hermannum (ἄπιστος) et Schwerdtium
(ἀπίστως) secutus: ἄδιστος FTr 413 ἀφεμένων FTr: corr. Hermann 414 ὑπὲρ
ποντίας T 416 δὲ F: γὰρ Tr 417 rävôpi Tr 419 ἀφροδίτα (ultimae
litterae superscripsit ἡ ut solet) Tr: -r7 F 425 χειρῶν F 426 κελεύθοις valde
suspectum, fortasse ex glossemate ἀκολούθοις ortum ; exempli gratia πελῶσα scripserim
429 ἑλλάδος FTr: corr. Bamberger 429 sq. πένθεια FTr: corr. Blass 431
δόμοις Auratus (cf. schol. vetus Tr ἑκάστου τοῖς οἴκοις ὀδυνηρὰ πένθησις διαπρέπει) : δόμων
FTr 433 rıs suppl. Porson πέμψεν Tr 434 φωτῶν F: βροτῶν Tr 435 54.
πρὸς ἑκάστου τοὺς δόμους εἰσαφικνεῖται Tr
116
and bringing to Ilion, for dowry, destruction, was gone
stepping nimbly through the gates, having dared a deed not
to be dared; while the seers in the house, with many a groan,
spake thus: ‘Alas, alas for the house, alas for the princes!
Alas for the bed and the husband-loving steps! One may see
the silence of those who are forsaken, a silence without
honour, without reviling, without belief. Through longing
for her who is beyond the sea, a phantom will seem to rule
the house. The grace of shapely statues is hateful to the
husband, and when the eyes are starved, all charm of love

is gone.

And appearing in dreams fancies of mourning are present,

bringing a joy that is vain ; for in vain a man, when, fancying
that he sees what is dear—aside out of his arms the vision is
gone, never again to draw near on wings, the companions of
sleep. ’! Such are the sorrows at the hearth within the home,
and others, too, more than transcending these; while in

general, for those who went forth together from Hellas’ land,
non-mourning with enduring heart is manifest in each one’s
house; there is much, at any rate, that touches the very
heart: those whom they sent they know, but instead of the
men urns? and ashes come back to each one’s home.

The gold-changer Ares, changer of bodies, and holder of

1 The text of the last few words is uncertain.

2 Or ‘armour’?
117
 καὶ ταλαντοῦχος ἐν μάχηι δορὸς
πυρωθὲν ἐξ ᾿Ιλίου 410
φίλοισι πέμπει βαρὺ
᾿ψῆγμα δυσδάκρυτον, ἀντήνορος
σποδοῦ γεμίφων λέβητας εὐθέτους.
στένουσι δ᾽ εὖ λέγοντες ἄν- 445
Spa τὸν μὲν ὡς μάχης ἴδρις,
τὸν δ᾽ ἐν φοναῖς καλῶς πεσόντ᾽
ἀλλοτρίας διαὶ γυναικός.
τάδε σῖγά τις Posi,
φθονερὸν δ᾽ ὑπ᾽ ἄλγος ἕρπει 450
προδίκοις ᾿Ατρείδαις.
οἱ δ᾽ αὐτοῦ περὶ τεῖχος
θήκας ᾿Ιλιάδος γᾶς
εὔμορφοι κατέχουσιν" ἐχ-
θρὰ δ᾽ ἔχοντας ἔκρυψεν. — 455

βαρεῖα δ᾽ ἀστῶν φάτις σὺν κότωι"

δημοκράντου δ᾽ ἀρᾶς τίνει χρέος.
μένει δ᾽ ἀκοῦσαί τί μοι
μέριμνα νυκτηρεφές. 460
τῶν πολυκτόνων γὰρ οὐκ ἄσκοποι
θεοί’ κελαιναὶ δ᾽ ᾿Ερινύες χρόνωι
τυχηρὸν ὄντ᾽ ἄνεν δίκας
παλιντυχεῖ τριβᾶι βίον 465
τιθεῖσ᾽ ἀμαυρόν, ἐν δ᾽ ἀί-
oTois τελέθοντος οὔτις ἀλκά.
τὸ δ᾽ ὑπερκόπως κλύειν εὖ
βαρύ: βάλλεται γὰρ ὄσσοις
Διόθεν κεραυνός. 470
κρίνω δ᾽ ἄφθονον ὄλβον’
μήτ᾽ εἴην πτολιπόρθης,
μήτ᾽ οὖν αὐτὸς ἁλοὺς ὑπ᾽ ἄλ-
λων βίον κατίδοιμι. —
FTr
443 γεμίξων (£ ex ζ facto; cf. ad 329) F 444 τοὺς λέβητας Tr εὐθέτου FTr: corr.
Auratus 448 affertur in Epimer. Hom., Anecd. Oxon. i. 119. 10 παρὰ δὲ Ἀττικοῖς
προζο λαμβάνει τὸ ira, καὶ γίνεται διαὶ, καὶ συντάσσεται γενικῆι, καὶ ἰσοδυναμεῖ τῶι ἕνεκα
. καὶ ἐν Ἀγαμέμνονι" ‘ ἀλλοτρίας διαὶ γυναικός ᾽, ἀντὶ τοῦ ἕνεκα διαὶ (coniecerat
Hermann) Epim. Hom.: διὰ F: γε διὰ Tr 449 σιγᾷ F τίς hoc accentu FTr
451 προδίκοισιν F 452 οἵδ᾽ FTr: distinxit Victorius 454 sq. ἐχθρῶς Tr
457 δημοκράτου FTr: corr. Porson 459 μοι Karsten: μου FTr 462
ἄσκοποι Ir: ἀπόσκοποι F 463 δ' F: δ᾽ οὖν Tr ἐριννύες FTr 465 παλιντυχῆ
(χῇ Tr) FTr: corr. Scaliger 468 ὑπερκότως FTr: corr. Casaubon 472 μήτ᾽
Tr: a 8 F πτολιπόρθις F
118
his scales in the battle of the spear, sends from Ilion to the

kinsmen what has felt the fire, heavy gold-dust bitterly

bewailed, freighting the easily-stowed urns with ashes in

exchange for men. And they lament them, praising each

man, this one for being well-skilled in the fight, that one as

nobly fallen in the murderous battle, through another man’s

wife. These words are muttered low, and resentful grief
against the principals in the suit, the Atridae, comes over

(the people). Others there by the wall, in all their shapeliness,

occupy tombs of Ilian earth, and the hostile land conceals
its possessors.

Dangerous is the citizens’ talk, with anger in it; it pays the

debt arising out of a curse pronounced by the people. And an
anxious thought persists in my mind, to hear of something
covered in night. For the gods are not unwatchful of those
who cause much bloodshed; and the dark Erinyes in the
course of time, when a man is prosperous without justice,
by wearing away his life in a reverse of fortune render him
faint and dim ; and when he is among the vanished, there is

no help. To be praised exceedingly is dangerous, for a thun-

derbolt is hurled by the eyes of Zeus. My choice is prosperity
without envy: may I neither be a destroyer of cities, nor yet,
myself made captive by others, see my life (in ruin).
IIQ
 πυρὸς δ᾽ ὑπ᾽ εὐαγγέλου 475
πόλιν διήκει θοὰ
βάξις’ εἰ δ᾽ ἐτήτυμος,
τίς οἶδεν, À τι θεῖόν ἐστι [μὴ] ψύθος.
τίς ὧδε παιδνὸς ἢ φρενῶν κεκομμένος,
φλογὸς παραγγέλμασιν 480
νέοις πυρωθέντα καρδίαν ἔπειτ᾽
ἀλλαγᾶι λόγου καμεῖν;
[ἐν] γυναικὸς αἰχμᾶι πρέπει
πρὸ τοῦ φανέντος χάριν ξυναινέσαι.
πιθανὸς ἄγαν ὁ θῆλυς Spos ἐπινέμεται 485
ταχύπορος: ἀλλὰ ταχύμορον
γυναικογήρυτον ὄλλυται κλέος. —c

— τάχ᾽ εἰσόμεσθα λαμπάδων φαεσφόρων

φρυκτωρίας τε καὶ πυρὸς παραλλαγάς, 490
εἴτ᾽ οὖν ἀληθεῖς, εἴτ᾽ ὀνειράτων δίκην —
τερπνὸν τόδ᾽ ἐλθὸν φῶς ἐφήλωσεν φρένας"
κήρυκ᾽ ἀπ᾽ ἀκτῆς τόνδ᾽ ὁρῶ κατάσκιον
κλάδοις ἐλαίας" μαρτυρεῖ δέ μοι κάσις
πηλοῦ ξύνουρος διψία κόνις τάδε, 495
ὡς οὔτ᾽ ἄναυδος οὗτος, οὐ δαίων φλόγα
ὕλης ὀρείας σημανεῖ καπνῶι πυρός,
ἀλλ᾽ ἢ τὸ χαίρειν μᾶλλον ἐκβάξει Aéycov:
τὸν ἀντίον δὲ τοῖσδ᾽ ἀποστέργω λόγον'
εὖ γὰρ πρὸς εὖ φανεῖσι προσθήκη πέλοι. 500
ὅστις τάδ᾽ ἄλλως THIS’ ἐπεύχεται πόλει,
αὐτὸς φρενῶν καρποῖτο τὴν ἁμαρτίαν.

KHPYZ
ἰὼ πατρῶιον οὖδας ᾿Αργείας χθονός"
δεκάτωι σε φέγγει TOO” ἀφικόμην ἔτους,
πολλῶν ῥαγεισῶν ἐλπίδων μιᾶς τυχών’ 505
οὐ γάρ ποτ᾽ ηὔχουν τῆιδ᾽ ἐν "Apyelaı χθονὶ
θανὼν μεθέξειν φιλτάτου τάφον μέρος.
νῦν χαῖρε μὲν χθών, χαῖρε δ᾽ ἡλίου φάος
FTr
476 τὴν πόλιν Tr 477 ἐτητύμως FTr: corr. Auratus 478 7 (superscr. εἰ,
τοι F: ἤ τοι Tr: εἴ τι Hermann ἐστὶν F μὴ del. Dindorf 480 παραγγέλμασιν
Tr:-oı F 481 sq. ἔπει in fine versus 481, ἔπειτ᾽ in initio versus 482 F 482
λόγους F 483 ἐν del. Scaliger 489 «Aur. praefixerunt FTr: del. Scaliger
εἰσόμεθα F 490 φρυκτωριῶν FTr: corr. Wilamowitz 492 ἐφήλωσεν
schol. vet. Tr: -oe ΕἼΤ 496 οὗτος, οὐ Wilamowitz: οὔ τέ σοι FTr 500
προσθήκει Tr (sed in scholio ipsius προσθήκη καλοῦ τυγχάνοι dv) SOI χορός
praefixerunt FTr: del. Scaliger
120
 From the fire that has brought good tidings ἃ swift rumour
has spread through the city: but whether it be true, who
knows ? or whether it be some deception from the gods. Who
is so childish or so crazed of wit, the sort of man that would
let his heart be fired by a flame’s unexpected message and
then be distressed when the tale is changed ? It is fitting for
a woman's rule to agree to give thanks before the thing itself
has appeared. Too easily persuasive, a woman’s ordinance
spreads fast-travelling, but fast-dying does a rumour voiced
by women perish.

We shall soon know about the beaconings and fire-trans-

missions of the light-bearing torches, whether they be true,
or whether this light that came so pleasingly has beguiled
our minds like a dream: I see here a herald from the shore
beneath the shade of olive-branches, and mud's sister and
neighbour, thirsty dust, assures me of this, that this mes-
senger, not voiceless, not kindling the flame of mountain
timber, will not make signs by smoke of fire, but will either
by his words express more clearly the joyful news, or—but
I have no liking for the word that is the opposite of this: for
(I pray) may an addition happily be made to what has
happily appeared! If there be any that prays otherwise in
this matter for our city, may be himself reap the fruit of the
error of his mind.
Enter Herald
Herald. O earth of my fathers, O earth of the Argive land!
In this tenth year’s light I have come to thee; many hopes
have broken, but one has come my way; for I never trusted
that it would be here in Argive land that I should die and
have my share in a burial most dear to me. Now I say, Hail
I2I
 ὕπατός τε χώρας Ζεὺς ὁ Πύθιός τ᾽ ἄναξ
τόξοις ἰάπτων μηκέτ᾽ εἰς ἡμᾶς βέλη; 510
ἅλις παρὰ Σκάμανδρον ἦσθ᾽ ἀνάρσιος,
νῦν δ᾽ αὖτε σωτὴρ ἴσθι καὶ παιώνιος,
ἄναξ “AtroAAov’ τούς T' ἀγωνίους θεοὺς
πάντας προσαυδῶ τόν τ᾽ ἐμὸν τιμάορον
Ἑρμῆν, φίλον κήρυκα, κηρύκων σέβας, 515
Npws τε τοὺς πέμψαντας, εὐμενεῖς πάλιν
στρατὸν δέχεσθαι τὸν λελειμμένον δορός.
ἰὼ μέλαθρα βασιλέων, φίλαι στέγαι,
σεμνοί τε θᾶκοι, δαίμονές τ᾽ ἀντήλιοι,
εἴ Trou πάλαι, φαιδροῖσι τοισίδ᾽ ὄμμασιν 520
δέξασθε κόσμωι βασιλέα πολλῶι χρόνωι.
ἥκει γὰρ ὑμῖν φῶς ἐν εὐφρόνηι φέρων
καὶ τοῖσδ᾽ ἅπασι κοινὸν ᾿Αγαμέμνων ἄναξ.
ἀλλ᾽ εὖ νιν ἀσπάσασθε, καὶ γὰρ οὖν πρέπει,
Τροίαν κατασκάψαντα τοῦ δικηφόρου - 525
Διὸς μακέλληι, τῆι κατείργασται πέδον,
[βωμοὶ δ᾽ ἄϊστοι καὶ θεῶν ἱδρύματα]
καὶ σπέρμα πάσης ἐξαπόλλυται χθονός,
τοιόνδε Τροίαι περιβαλὼν zeukrrpiov
ἄναξ ᾿Ατρείδης πρέσβυς εὐδαίμων ἀνὴρ 530
ἥκει, τίεσθαι δ᾽ ἀξιώτατος βροτῶν
τῶν vive Πάρις γὰρ οὔτε συντελὴς πόλις
ἐξεύχεται τὸ δρᾶμα τοῦ πάθους πλέον.
ὀφλὼν γὰρ ἁρπαγῆς τε καὶ κλοπῆς δίκην
τοῦ ῥυσίου θ᾽ ἥμαρτε καὶ πανώλεθρον 535
αὐτόχθονον πατρῶιον ἔθρισεν δόμον"
διπλᾶ δ᾽ ἔτεισαν Πριαμίδαι θάμάρτια.
ΧΟ. κῆρυξ ᾿Αχαιῶν χαῖρε τῶν ἀπὸ στρατοῦ.
KH. χαίρω {τεθνᾶναι δ᾽ οὐκέτ᾽ ἀντερῶ θεοῖς.
ΧΟ. ἔρως πατρώιας τῆσδε γῆς σ᾽ ἐγύμνασεν; 540
KH. ὥστ᾽ ἐνδακρύειν γ᾽ ὄμμασιν χαρᾶς ὕπο.
ΧΟ. τερπνῆῇς ἄρ᾽ Forte τῆσδ᾽ ἐπήβολοι νόσου.
ΚΗ. πῶς δὴ διδαχθεὶς τοῦδε δεσπόσω λόγον;
FTr
511 ἦσθ᾽ Needham: 349° (superscr. es) F: ἦλθες Tr 512 καὶ παιώνιος Dobree:
καὶ παγώνιος F: κἀπαγώνιος Tr (in utroque libro superscripta est glossa ἀπόμαχος)
420 ei Auratus: ἦ FTr πάλαι (super à scr. v) F τοῖσιδ᾽ F: τοῖσιν Tr 521
δέξασθε F : δέξαισθε Tr 522 ὑμῖν Tr: ἡμῖν F 527 del. Salzmann; cf. Pers.
811 529 τοιούδε (sic) τροίᾳ περι" βαλὼν I 534 ὄφλων FTr $37 Erıoav
FTr $38, 540, etc. per totam stichomythiam «Aur. praefixerunt FTr, chori
notam restituit Heath 539 redväva corruptum est, nec non et sequentia verba
suspicionem movent οὐκέτ᾽ et in ipso versu et in scholio vetere ad 550 adscripto
Tr: οὐκ F 541 ἐκδακρύειν Tr 542 ἦστε Ahrens: ἴστε F: ἦτε Tr
122
O Land, hail O light of the Sun, and Zeus the country’s
highest, and thou lord of Pytho, who shootest forth, no
longer, I pray, against us, shafts with thy bow—we had
enough of thine enmity by Scamander’s stream, but now be
saviour and healer, Lord Apollo!—and all the gods in
assembly I hereby address, and chiefly my own protector,
Hermes the dear herald, whom heralds worship; and the
heroes who sent us forth, that they may receive back in
kindness the army, so much as the spear has spared.
O halls of our kings, beloved abode, and august seats, and
sun-facing deities ; if ever before, with this bright gladness in
your eyes welcome duly the king after long lapse of time. For
he has come bringing light in the night-time for you, and for
all that are here to share—Agamemnon our lord. Come, give
him good welcome, for so it is fitting, since he has dug down
Troy with the mattock of Zeus the bringer of justice, where-
with the soil has been thoroughly worked and the seed
perishes out of all the land. Such is the yoke which our king,
the elder son of Atreus, has laid upon Troy, and now he has
returned as a man of blessed fortune, most worthy of all men
of our time to be honoured ; for neither Paris nor his associate
city can boast that what they did is more than what they
have suffered. For, convicted of robbery as well as theft, he
has not only lost his plunder but has caused the house of his
fathers to be shorn off and utterly destroyed together with
the land ; and twofold is the penalty for their offence that the
sons of Priam have paid.
Chor. Herald returned from the Achaean army, I wish
thee joy.
Her. Joy I have; and I should no longer object to being
dead.!
Chor. Was it longing for this native land of thine that
exercised thee ?
Her. Yes, so much that my eyes fill with tears for joy.
Chor. It was asweet malady, then, this that ye had gotten.
Her. May I have an explanation and thus master this thy
saying ?
! The meaning of the line seems clear, though some details are uncertain.
123
 ΧΟ. τῶν ἀντερώντων ἱμέρωι πεπληγμένοι.
ΚΗ. ποθεῖν ποθοῦντα τήνδε γῆν στρατὸν λέγεις; 545
XO. ὡς πόλλ᾽ ἀμαυρᾶς ἐκ φρενός {μ᾽} ἀναστένειν.
ΚΗ. πόθεν τὸ δύσφρον τοῦτ᾽ ἐπῆν {θυμῶι στύγος [orparéi];
ΧΟ. πάλαι τὸ σιγᾶν φάρμακον βλάβης ἔχω.
ΚΗ. καὶ πῶς; ἀπόντων κοιράνων ἔτρεις τινάς;
ΧΟ. ὡς νῦν, τὸ σὸν δή, καὶ θανεῖν πολλὴ χάρις. 550
KH. εὖ γὰρ πέπρακται. ταῦτα δ᾽ ἐν πολλῶι χρόνωι
τὰ μέν τις ἂν λέξειεν εὐπετῶς ἔχειν,
τὰ δ᾽ αὖτε κἀπίμομφα' τίς δὲ πλὴν θεῶν
ἅπαντ᾽ ἀπήμων τὸν δι᾽ αἰῶνος χρόνον;
μόχθους γὰρ εἰ λέγοιμι καὶ δυσαυλίας, 555
σπαρνὰς παρήξεις καὶ κακοστρώτους, τί δ᾽ οὐ
στένοντες, fou λαχόντες ἤματος μέρος;
τὰ δ᾽ αὖτε χέρσωι καὶ προσῆν πλέον στύγος"
εὐναὶ γὰρ ἦσαν δηίων πρὸς τείχεσιν,
ἐξ οὐρανοῦ δὲ κἀπὸ γῆς λειμώνιαι ᾿ 560
δρόσοι κατεψάκαφον, ἔμτεδον σίνος,
ἐσθημάτων τιθέντες ἔνθηρον τρίχα.
χειμῶνα δ᾽ εἰ λέγοι τις οἰωνοκτόνον,
οἷον παρεῖχ᾽ ἄφερτον ᾿Ιδαία χιών,
ἢ θάλπος, εὖτε πόντος ἐν μεσημβριναῖς 565
κοίταις ἀκύμων νηνέμοις εὖδοι πεσών’
τί ταῦτα πενθεῖν δεῖ; παροίχεται πόνος"
παροίχεται δέ, τοῖσι μὲν τεθνηκόσιν
τὸ μήποτ᾽ αὖθις μηδ᾽ ἀναστῆναι μέλειν’
[τί τοὺς ἀναλωθέντας ἐν ψήφωι λέγειν, 570
τὸν τῶντα δ᾽ ἀλγεῖν χρὴ τύχης παλιγκότον;
καὶ πολλὰ χαίρειν συμφοραῖς καταξιῶ.]
ἡμῖν δὲ τοῖς λοιποῖσιν ᾿Αργείων στρατοῦ
νικᾶι τὸ κέρδος, πῆμα δ᾽ οὐκ ἀντιρρέτει.
ὡς κομπάσαι τῶιδ᾽ εἰκὸς ἡλίου φάει 575
ὑπὲρ θαλάσσης Kal χθονὸς ποτώμενα"
* Τροίαν ἑλόντες δή ποτ᾽ ᾿Αργείων στόλος
FTr
544 πεπληγμένος FTr: corr. Tyrwhitt 546 μ᾽ suppl. Scaliger 547
στρατῶι delevit, θυμῶι ante στύγος inseruit Schütz (ζύμῖν» στύγος; Wilamowitz)
549 κοιράνων Tr: τυράννων F $50 ὡς Auratus: ὧν FTr SSI ταῦτα δ᾽ nondum
expeditum 552 ἂν Auratus: εὖ FTr 553 τάδ᾽ FTr: distinxit Heath
556 de voce παρήξεις dubitari potest κακοτρώτους F 557 οὐ λαχόντες
corruptum; de verbis ἤματος μέρος nihil affirmare ausim 558 τάδ᾽ FTr: distinxit
Victorius 559 δαΐων Dindorf 560 δὲ Pearson: γὰρ ΕἼΤ λειμωνίέαι hoc
accentu FTr 561 κατεψάκαζον Dindorf: κατεψέκαζον FTr 563 λέγει Tr
569 μὴ δ᾽ FTr 570-2 eiecerunt Nägelsbach aliique 575 τάδ᾽ Weil 576
ποτωμένοις FTr: corr. Weil 577 rpoinv F
124
 Chor. Being smitten with desire for those who returned
the longing.
Her. Dost thou mean that the country here languished for
the army while the army languished for it ?
Chor. So much that I have often sighed from a dark and
weakened spirit.
Her. How came it that this dispirited gloom was on thy
heart ?!
Chor. Silence has long been my remedy against harm.
Her. How? Hadst thou someone to fear in the absence of
the rulers ?
Chor. So much that now, in thine own words, even death
were a great boon.
Her. Aye, for success has been achieved. Of all things?
that happen in a long time, one might say of some that they
have turned out well, of others that fault is to be found with
them; but who, except the gods, is free from hurt for the
whole time of his life? For were I to speak of our labours and
hard lodging, scanty (or ‘narrow’ ?) gangways where we had
wretched bedding—what was there that we had not to
lament, what that ...? Then again, on land, there was added
even more that was horrible: our couches were close against
the enemy’s walls, and from the sky—and from the ground
of the meadows—dews drizzled upon us, a constant damage,
making the hair of our clothes full of creatures. And if one
were to tell of the bird-killing winter, such intolerable cold
brought about by the snows of Ida, or the heat, when in its
noonday rest unruffled by a wind the sea lay sunk in slumber
without a wave—what need is there to mourn for this? the
distress is past; it is past, for those who are dead, so that
they will never care even to rise again; but for us who are
left of the Argive host, the gain prevails and the suffering is
no counterpoise. Therefore it is meet that, on this bright
day,? we should make this boast and let it soar over sea and
land: ‘Troy was once taken by the expedition of the Argives ;
! Details of the text are uncertain.
2 Text uncertain.
3 Or ‘to this light of the sun’.
125
 θεοῖς λάφυρα ταῦτα τοῖς καθ᾽ Ἑλλάδα
δόμοις ἐπασσάλευσαν ἀρχαῖον ydvos.’
τοιαῦτα χρὴ κλύοντας εὐλογεῖν πόλιν 580
καὶ τοὺς στρατηγούς" καὶ χάρις τιμήσεται
Διὸς τἀδ᾽ ἐκπράξασα. πάντ᾽ ἔχεις λόγον.
ΧΟ. νικώμενος λόγοισιν οὐκ ἀναίνομαι"
ἀεὶ γὰρ ἡβᾶι τοῖς γέρουσιν εὐμαθεῖν.
δόμοις δὲ ταῦτα καὶ Κλνταιμήστραι μέλειν 585
εἰκὸς μάλιστα, σὺν δὲ πλουτίφειν ἐμέ,

ΚΛΥΤΑΙΜΗΣΤΡΑ
ἀνωλόλυξα μὲν πάλαι χαρᾶς ὕπο,
ὅτ᾽ ἦλθ᾽ ὁ πρῶτος νύχιος ἄγγελος πυρός,
φράφτων ἅλωσιν Ἰλίου τ᾽ ἀνάστασιν,
καί τίς μ᾽ ἐνίπτων eire ᾿ φρυκτώρων διὰ 590
πεισθεῖσα Τροίαν νῦν πεπορθῆσθαι δοκεῖς;
3j κάρτα πρὸς γυναικὸς αἴρεσθαι Kap.’
λόγοις τοιούτοις πλαγκτὸς οὖσ᾽ ἐφαινόμην.
ὅμως δ᾽ ἔθνον, καὶ γυναικείωι νόμωι
ὀλολυγμὸν ἄλλος ἄλλοθεν κατὰ πτόλιν 595
ἔλασκον εὐφημοῦντες, ἐν θεῶν ἕδραις
θυηφάγον κοιμῶντες εὐώδη φλόγα.
καὶ νῦν τὰ μάσσω μὲν τί δεῖ σ᾽ ἐμοὶ λέγειν;
ἄνακτος αὐτοῦ πάντα πεύσομαι Adyov.
ὅπως δ᾽ ἄριστα τὸν ἐμὸν αἰδοῖον πόσιν 600
σπεύσω πάλιν μολόντα δέξασθαι: τί γὰρ
γυναικὶ τούτον φέγγος ἥδιον δρακεῖν,
ἀπὸ στρατείας ἄνδρα σώσαντος θεοῦ
πύλας ἀνοῖξαι; ταῦτ᾽ ἀπάγγειλον πόσει.
ἥκειν {δ᾽) ὅπως τάχιστ᾽ ἐράσμιον πόλει" 605
γυναῖκα πιστὴν δ᾽ ἐν δόμοις εὕροι μολὼν
οἵανπερ οὖν ἔλειπε, δωμάτων κύνα
ἐσθλὴν ἐκείνωι, πολεμίαν τοῖς δύσφροσιν,
καὶ τἄλλ᾽ ὁμοίαν πάντα, σημαντήριον
οὐδὲν διαφθείρασαν ἐν μήκει χρόνον. 610
οὐδ᾽ οἶδα τέρψιν, οὐδ᾽ ἐπίψογον φάτιν,
ἄλλου πρὸς ἀνδρὸς μᾶλλον ἢ χαλκοῦ βαφάς.
τοιόσδ᾽ ὁ κόμπος" τῆς (8°) ἀληθείας γέμων
FTr
584 εὐμαθεῖν Headlam : εὖ μαθεῖν FTr 585 κλυταιμνήστρᾳ ΕἼΤ 587
ἀνωλολύξαμεν FTr: distinxit Henr. Stephanus 590 ἐνίππων F διὰ recto
accentu FTr 593 πλακτὸς Tr 595 ὀλολυγμὸν (secundo ὁ superscr. w) F
596 ἐνθέων F 605 δ᾽ supplevit Weil 613 sq. Clytaemestrae continuavit
Hermann, κήρυξ praefixerunt FTr 613 δ᾽ supplevit Headiam
126
and these are the spoils which, to the gods throughout Hellas,
they nailed upon their temples as a glory like those of old.’
Hearing this, men must needs praise the city and her
generals; and the grace of Zeus, that has accomplished this,
shall be duly prized. Thou hast heard all I had to say.
Chor. I am conquered by thy words and am not sorry for
it, for to be teachable (is a thing that) remains always young
for those who are old. But it is meet that this should most
concern the house and Clytemnestra, while at the same time
enriching me as well.
Enter Clytemnestra
Clyt. I raised a cry for joy a while ago, when the first fiery
messenger came in the night, telling of the capture and the
razing of Ilion. And there were some who upbraided me and
said: ‘Have fire-signals prevailed upon thee to believe that
Troy is now laid waste? How like a woman to be uplifted in
heart!’ Such talk made me appear as one astray. But for all
that I made my sacrifices, and they, in woman’s fashion, one
here, one there, throughout the city, shouted out the jubilant
cry, lulling in the shrines of the gods the incense-fed fragrant
flame.
And at present, for the fuller story, what need is there for
thee to tell me? From the king himself I shall learn it all.
But I must hasten to receive in the best way my revered lord
on his return; for what light is sweeter for a wife to behold
than this, when, as the god has preserved her husband safe
from the field, she opens the gates for him? Take that as a
message to my husband! And bid him be back with all speed,
the people’s darling! But as for his wife—may he return and
find her in his home faithful, even such as he left her, a
watch-dog of the house loyal to him and an enemy to those
who wish him ill, and alike in all the rest, never having
broken any seal in all this while. And of joys from another
man—aye, or scandalous rumour—I know no more than of
tempering! metal. Such is my boast, and since it is full of
! The meaning is not quite certain; see the commentary.
127
 οὐκ αἰσχρὸς ὡς γυναικὶ γενναίαι λακεῖν.
ΧΟ. αὕτη μὲν οὕτως εἶπε, μανθάνοντί σοι 615
τοροῖσιν ἑρμηνεῦσιν εὐπρεπῆ λόγον.
σὺ δ᾽ εἶπέ, κῆρυξ’ Μενέλεων δὲ πεύθομαι"
el νόστιμός τε καὶ σεσωμένος πάλιν
ἥκει σὺν ὑμῖν, τῆσδε γῆς φίλον κράτος.
ΚΗ. οὐκ ἔσθ᾽ ὅπως λέξαιμι τὰ ψευδῆ καλὰ 620
ἐς τὸν πολὺν φίλοισι καρποῦσθαι χρόνον.
XO. πῶς δῆτ᾽ ἂν εἰττὼν κεδνὰ τἀληθῆ τύχοις;
σχισθέντα δ᾽ οὐκ εὔκρυτπτα γίγνεται τάδε.
ΚΗ. ἁνὴρ ἄφαντος ἐξ ᾿Αχαιικοῦ στρατοῦ,
αὐτός τε καὶ τὸ πλοῖον. οὐ ψευδῆ λέγω. 625
ΧΟ. πότερον ἀναχθεὶς ἐμφανῶς ἐξ Ἰλίου,
ἢ χεῖμα, κοινὸν ἄχθος, ἥρπασε στρατοῦ;
ΚΗ. ἔκυρσας ὥστε τοξότης ἄκρος σκοποῦ"
μακρὸν δὲ πῆμα συντόμως ἐφημίσω.
ΧΟ. πότερα γὰρ αὐτοῦ τῶντος ἢ τεθνηκότος 630
φάτις πρὸς ἄλλων ναυτίλων EKATIZETO;
KH. οὐκ οἶδεν οὐδεὶς ὦστ᾽ ἀπαγγεῖλαι τορῶς
TANV τοῦ τρέφοντος Ἡλίου χϑονὸς φύσιν.
ΧΟ. πῶς γὰρ λέγεις χειμῶνα ναυτικῶι στρατῶι |
ἐλθεῖν τελευτῆσαί TE δαιμόνων κότωι; 635
ΚΗ. εὔφημον ἦμαρ οὐ πρέπει κακαγγέλωι
γλώσσηι μιαίνειν’ χωρὶς À τιμὴ θεῶν.
ὅταν δ᾽ ἀπευκτὰ πήματ᾽ ἄγγελος πόλει
στυγνῶι TTPOGWTTWI πτωσίμου στρατοῦ φέρηι,
πόλει μὲν ἕλκος ἕν τὸ δήμιον τυχεῖν, 640
πολλοὺς δὲ πολλῶν ἐξαγισθέντας δόμων
ἄνδρας διπλῆι μάστιγι, τὴν “Apns φιλεῖ,
δίλογχον ἄτην, φοινίαν ξυνωρίδα:
τοιῶνδε μέντοι πημάτων σεσαγμένον
πρέπει λέγειν παιᾶνα τόνδ᾽ ᾿Ερινύων᾽" 645
σωτηρίων δὲ πραγμάτων εὐάγγελον
ἥκοντα πρὸς χαίρουσαν εὐεστοῖ πόλιν,
Trés κεδνὰ τοῖς κακοῖσι συμμείξω, λέγων
χειμῶν᾽ ᾿Αχαιοῖς οὐκ ἀμήνιτον θεῶν;
ΕἼΤ
616 εὐπρεπῶς FTr: corr. Auratus 618 re Hermann: ye FTr σεσωσμένος
FTr: corr. Wecklein (σεσωιμ- Wilamowitz) 619 ἥξει FTr: corr. Karsten 622
«Avr. hic et 626, 630, 634 praefixerunt FTr : corr. Casaubon τύχης FTr: corr. Porson
623 γίνεται FTr 624 ἁνὴρ Hermann: ἀνὴρ FTr 639 Hesychii σμοιῶε
(ίσμοιος cod.: corr. Musurus) προσώπωι" φοβερῶι 9 oruyvar, σκυθρωπῶι huc rettulit M;
Schmidt, fortasse recte 644 σεσαγμένων Fr: corr. Schütz 645 ἐριννύων
FTr 648 συμμίξω ΕἼΤ 649 ἀχαιῶν..... θεοῖς FTr: corr. Blomfield,
Dobree
128
truth, there is nothing shameful—for a noble lady—in my
proclaiming it aloud. Exit Clytemnestra
Chor. Thus she has spoken, a speech which, if thou under-
standest it through clear interpreters, looks fair. But tell me,
herald,—it is of Menelaus I would learn—is he on his way
home and back safe with you, the beloved ruler of this land ?
Her. It is impossible for me so to give a fair tale of what is
false that my friends might reap fruit of it for the long time
ahead.
Chor. Would then that thou mightest give a good tale of
what is true and hit the mark! When these things! are
severed it is not easy to conceal it.
Her. The man is vanished from the Achaean host, himself
and his ship as well. It is no false tale I tell.
Chor. Did he sail from Ilion in your sight, or did a storm
fall upon all alike and snatch him from the host ?
Her. Like a master archer thou hast hit the mark, and
hast voiced a length of suffering in a brief phrase.
Chor. Which was it, living or dead, that he was said to be
in the tale told by the other voyagers ?
Her. No one knows, so as to be able to give a clear report,
save Helios who fosters all life on the earth.
Chor. In what manner dost thou say that the storm came
to the fleet, and ended, by the anger of the gods?
Her. It is not fitting to defilean auspicious day with the
voice of evil tidings: such a celebration is apart from the
gods (of heaven). When a messenger with gloomy coun-
tenance brings to a city the dread? calamity of an army
fallen—how a wound has befallen the city, at once one
common wound for the whole people, while also many men
out of many homes are banished? by the double scourge
which Ares loves, a two-speared bane, a bloody pair—when
he carries on his back such a load of calamities, then it is
fitting that he should utter this paean of the Erinyes; but
when one comes with the good tidings that all is safe to a
city rejoicing in her well-being—how shall I mix the good
with such an ill, speaking of a storm which lacked not the
! j.e. the good and the true.
2 Lit. ‘such as one prays may not come to pass’.
3 Or ‘devoted (to the avenging deities)’?
4872-1 t K

129
 ξυνώμοσαν γὰρ ὄντες ἔχθιστοι τὸ πρὶν 650
πῦρ καὶ θάλασσα, καὶ τὰ πίστ᾽ ἐδειξάτην
φθείροντε τὸν δύστηνον ᾿Αργείων στρατόν.
ἐν νυκτὶ δυσκύμαντα δ᾽ ὠρώρει κακά"
ναῦς γὰρ πρὸς ἀλλήλησι Θρήικιαι πνοαὶ
ἤρεικον’ al δὲ κεροτυπούμεναι βίαι 655
χειμῶνι τυφῶ σὺν Ξάληι τ᾽ ὀμβροκτύποωι
ὦιχοντ᾽ ἄφαντοι ποιμένος κακοῦ στρόβωι.
ἐπεὶ δ᾽ ἀνῆλθε λαμπρὸν ἡλίου φάος,
ὁρῶμεν ἀνθοῦν πέλαγος Αἰγαῖον νεκροῖς
ἀνδρῶν ᾿Αχαιῶν ναντικοῖς τ᾽ ἐρειπίοις. 660
ἡμᾶς γε μὲν δὴ ναῦν τ᾽ ἀκήρατον σκάφος
ἤτοι τις ἐξέκλεψεν À ᾿ξηιτήσατο,
θεός τις, οὐκ ἄνθρωπος, οἴακος θιγών.
τύχη δὲ σωτὴρ ναῦν θέλουσ᾽ ἐφέξετο,
ὡς μήτ᾽ ἐν ὅρμωι κύματος τάλην ἔχειν 665
μήτ᾽ ἐξοκεῖλαι πρὸς κραταίλεων χθόνα.
ἔπειτα δ᾽ “AiSnv πόντιον πεφευγότες,
λευκὸν κατ᾽ ἦμαρ, οὐ πεποιθότες τύχηι,
ἐβουκολοῦμεν φροντίσιν νέον πάθος,
στρατοῦ καμόντος καὶ κακῶς σποδουμένον. 670
καὶ νῦν ἐκείνων εἴ τις ἐστὶν ἐμπνέων
λέγουσιν ἡμᾶς ὡς ὀλωλότας, τί μήν;
ἡμεῖς τ᾽ ἐκείνους ταῦτ᾽ ἔχειν δοξάζομεν.
γένοιτο δ᾽ ὡς ἄριστα. Μενέλεων γὰρ οὖν
πρῶτόν τε καὶ μάλιστα προσδόκα μολεῖν. 675
εἰ δ᾽ οὖν τις ἀκτὶς ἡλίου νιν ἱστορεῖ
καὶ τῶντα καὶ βλέποντα, μηχαναῖς Διὸς
οὔπω θέλοντος ἐξαναλῶσαι γένος
ἐλπίς τις αὐτὸν πρὸς δόμους ἥξειν πάλιν.
τοσαῦτ᾽ ἀκούσας ἴσθι τἀληθῆ κλνών, 680

ΧΟΡΟΣ
tts ποτ᾽ ὠνόματεν ὧδ᾽
ἐς τὸ πᾶν ἐτητύμως"
μή τις ὄντιν᾽ οὐχ ὁρῶμεν προνοί-
αἷἱσὶ τοῦ πεπρωμένον
FTr
654 ἀλλήλαισι Tr 655 ἤρειπον Tr κερωτυπούμεναι FTr 659 versum
afferunt Ammonius qui fertur (in excerpto 7. BapBapiouoë) p. 197 Valck., Phot. Berol.
p. 138. 16 660 ναυτικῶν τ᾽ ἐριπίων FTr: corr. Auratus 672 ri μή FTr: corr.
Linwood 677 Hesychii χλωρόν re καὶ βλέποντα' ἀντὶ τοῦ ζῶντα huc rettulit Toup,
fortasse recte 679 τις Tr: τίς F 680 κλύων (ultimis duabus litteris superscripto
ew) F (aoristum esse viderunt Casaubon, Kueck, 4}1}} κλύειν Tr et scholium metricum
in FTr ad 489 adscriptum 681 ὠνόμαξεν F 683 sq. mpovoiaıs FTr: corr. Pauw
130
wrath of the gods against the Achaeans ? For they who had
hitherto been utter foes, fire and sea, now swore alliance, and
displayed their covenant by destroying the unhappy Argive
host. In the night-time the disaster of evil waves arose:
Thracian blasts shattered the ships one against another ; and
they, being violently rammed beneath the storm of the
hurricane and the rush of drumming rain, sank from sight,
lashed round by an evil shepherd. And when the bright light
of the sun came up, we saw the Aegean sea aflower with
corpses of Achaean men and wrecks of ships. But ourselves
and our ship, uninjured in its hull, someone either stole
away or begged off, some god, not a mortal, who laid hand
upon the helm. And Saving Fortune sat graciously upon our
ship, so that it neither had to stand the welter of the waves
at anchor nor ran aground on a rocky shore. And afterwards,
having escaped the watery grave, in the clear bright day,
not trusting our good fortune, we let our thoughts dwell
upon the unexpected disaster, our fleet being wrecked and
miserably pounded. And at this moment, if any of them is
yet alive they speak of us as perished, of course; while we
imagine that it is they to whom that has happened. Well,
may it turn out in the best way possible: for as far as Mene-
laus is concerned, first and chiefly thou must suppose that he
is back again. But! if any ray of the sun does descry him safe
and sound, there is a hope that by the contrivances of Zeus,
who is not yet willing to destroy the race entirely, he may
return home again. Having heard thus much be assured that
thou hast been told the truth.
Exit Herald, in the direction of the sea
Chor. Who can it have been who gave the name with such
entire truth— was it someone whom we do not see, guiding
! Referring to an unexpressed thought ; see the commentary.
131
 γλῶσσαν ἐν τύχαι venwv' 685
τὰν δορίγαμβρον ἀμφινει-
κῇ θ᾽ Ἑλέναν; ἐπεὶ πρεπόντως
ἑλέναυς EAavEpos ἐλέ-
πτολις ἐκ τῶν ἁβροπήνων 690
προκαλυμμάτων ἔπλευσεν
Ξεφύρου γίγαντος αὔραι,
πολύανδροί τε φεράσπιδες κυναγοὶ
κατ᾽ ἴχνος πλατᾶν ἄφαντον 695
κελσάντων Σιμόεντος ἀκ-
τὰς ἐπ᾽ ἀεξιφύλλους
δι᾽ Ἔριν αἱματόεσσαν. —

Ἰλίωι δὲ κῆδος ὀρ-

θώνυμον τελεσσίφρων 700
Μῆνις ἤλασεν, τραπέφας àTi-
μώσιν ὑστέρωι χρόνωι
καὶ ξυνεστίον Διὸς
πρασσομένα τὸ νυμφότι- 705
μον μέλος ἐκφάτως τίοντας,
ὑμέναιον, ὃς τότ᾽ ἐπέρ-
ρεπε γαμβροῖσιν ἀείδειν.
μεταμανθάνουσα δ᾽ ὕμνον
Πριάμου πόλις γεραιὰ 710
πολύθρηνον μέγα Trou στένει KIKÄTIGKOU-
σα Πάριν τὸν αἰνόλεκτρον,
Ἱπαμπρόσθη πολύθρηνονΤ
Ἰαἰϊῶν᾽ ἀμφὶ πολίτανΤ 715
μέλεον αἷμ᾽ ἀνατλᾶσα. =

ἔθρεψεν δὲ λέοντος I-
νιν δόμοις ἀγάλακτον οὕ-
τῶς ἀνὴρ φιλόμαστον,
ἐν βιότου προτελείοις 720
ἅμερον, εὐφιλόταιδα
καὶ γεραροῖς ἐπίχαρτον"
FIr
688 ἑλέναυς Blomfield: ἑλένας FTr 690 ἁβροτίμων Tr T : corr. Salmasius
695 πλατᾶν Heath: zAdrav ΕἼΤ 696 sq. ἀκτᾶς F 697 én’ ἀεξιφύλλους
Abresch: em’ ἀξιφύλλους F : εἰς ἀεξιφύλλους (super ἀεξε scr. συνίζησις) Tr 700
τελεσίφρων F 701 ἤλασε FTr: v adiecit Porson (cf. ad 201) 701 sq. ἀτίμωσιν
Canter: ἀτέμως iv’ F: ἀτίμως Tr 707 SQ. ἐπέρρεπεν F: ἐπέπρεπεν Tr: v del. Porson
714 sq. quae poeta scripsit recuperari nequeunt 715 moAırav Auratus, pro-
babiliter, sed in versu corrupto ne hoc quidem certum est 716 μέλεον ex μέλαιον
corr. F 717 sq. λέοντα σίνιν FTr: corr. Conington 718 sq. οὗτος (superscr. ws) F
132
his tongue aright to the mark with thoughts anticipating
destiny ?—to her of the spear-wedding, for whom two sides
contended, Helen ? For fittingly it was that ship-destroying,!
man-destroying, town-destroying, she came out of her
luxuriously woven chamber-curtains and sailed forth, before
the blast of giant Zephyrus, and (after her) many men,
shield-bearers, hunters upon the vanished trail of oars, the
trail of those who had landed? on the leafy shores at the
mouth of Simois, by the will of bloody Strife.

And Wrath that brings her thought to fulfilment set afoot

for Ilion a marriage in the true meaning of the word, by
exacting account in after time for the dishonour of the guest-
table and of Zeus, guardian of those who share the same
hearth, from the folk who loudly (?) celebrated the strain in
honour of the bride, the wedding-hymn which it then fell to
the kinsmen to sing. But unlearning that song and learning
in its stead a song of much mourning, the aged town of
Priam must now be making loud lamentation, calling Paris
‘O thou of the fatal marriage!’ . . . having endured pitiful
bloodshed.

So, once, was a lion’s offspring reared by a man in his

house, getting no milk from its mother, still fond of the teat,

in the prelude of its life tame, a good friend of the children

! The play on words (‘EdAévy, éévaus) is untranslatable.
2 Or 'the oars, while they (i.e. Paris, Helen, and their companions) had landed . . .'.
3 The double meaning of κῆδος is untranslatable.

133
 πολέα δ᾽ ἔσκ᾽ ἐν ἀγκάλαις
νεοτρόφου τέκνου δίκαν,
φαιδρωπὸς ποτὶ χεῖρα σαί- 725
νῶν τε γαστρὸς ἀνάγκαις. —

χρονισθεὶς δ᾽ ἀπέδειξεν ἧ-
dos τὸ πρὸς τοκέων" χάριν
γὰρ τροφεῦσιν ἀμείβων
μηλοφόνοισιν (iv) ἄταις 730
δαῖτ᾽ ἀκέλευστος ἔτευξεν,
αἵματι δ’ οἶκος ἐφύρθη,
ἄμαχον ἄλγος οἰκέταις,
μέγα σίνος TroAukróvov:
ἐκ θεοῦ δ᾽ ἱερεύς τις "A- 735
τας δόμοις προσεθρέφθη. =

πάραυτα δ᾽ ἐλθεῖν ἐς Ἰλίου πόλιν

λέγοιμ᾽ ἂν φρόνημα μὲν νηνέμου γαλάνας 740
ἀκασκαῖόν (T’) ἄγαλμα πλούτον,
μαλθακὸν ὀμμάτων βέλος,
δηξίθυμον ἔρωτος ἄνθος.
παρακλίνασ᾽ ἐπέκρανεν
δὲ γάμου πικρὰς τελευτάς, 745
δύσεδρος καὶ δυσόμιλος
συμένα Πριαμίδαισιν
πομτπᾶι Διὸς ξενίου,
νυμφόκλαυτος Ἐρινύς. —

παλαίφατος δ᾽ ἐν βροτοῖς γέρων λόγος 750

τέτυκται μέγαν τελεσθέντα φωτὸς ὄλβον
τεκνοῦσθαι μηδ᾽ ἄπαιδα θνήισκειν,
ἐκ δ᾽ ἀγαθᾶς τύχας γένει 755
βλαστάνειν ἀκόρεστον olzüv.
δίχα δ᾽ ἄλλων μονόφρων el-
ni τὸ δυσσεβὲς γὰρ ἔργον
FTr
723 €&or’ Casaubon: ἔσχ᾽ ΕἼΤ 727 sq. ἔθος FTr: corr. Conington 728
τοκήων F 729 rpodedow Tr (quod editores afferunt, ἤγουν ἀμοιβὰς διδοὺς τοῖς
θρέψασιν αὐτόν, non scholium vetus est sed glossa interlinearis) : τροφᾶς F 730 ἐν
supplevit Bothe draw Tr 733 ἄμαχον δ᾽ F 734 πολύκτονον FTr: accen-
tum corr. Bothe 736 προσετράφη FTr: corr. Heath 737 παραυτὰ δ᾽ F:
παραυτά δ᾽ οὖν Tr 741 τ᾽ supplevit Hermann (δ᾽ Porson), sed dubitari potest
744 παρακλίνουσ᾽ Tr 745 πικροῦ (in fine superscr. as) F 747 πριαμίδαισι F
749 ἐριννύς Tr 750 ἐν τοῖς βροτοῖς Tr 754 μὴ δ᾽ FTr 756 ὀϊζύν FTr
(superscr. συνίζησις Tr) 758 γὰρ δυσσεβὲς FTr: traiecit Pauw

134
and a delight to the elders; and many a time it was in their
arms, like a nursling child, looking bright-eyed to the hand
and fawning under the constraint of its belly.

But matured by time it showed the character it had from

its parents: for, making a return to those who had reared it,
it made ready a feast as an unbidden guest in a horrid
slaughter of the flock, and the house was befouled! with
blood—to the house-folk an agony not to be warded off, a
vast havoc wherein many were killed ; by the will of the god
it had been reared? in the house to be a sacrificer in the
service of Ate.

At first, I would say, there came to the city of Ilion a

temper of windless calm and a gentle delight of wealth, a soft
arrow of the eyes, a heart-stinging flower of desire. But then
she swerved from this, and accomplished a bitter end of the
marriage-rites, rushing upon the sons of Priam as an evil
settler and an evil companion by the sending and guidance
of Zeus Hospitable, an Erinys bringing tears to brides.

Long has there been current among men an aged saying

that a man’s prosperity, when it has reached full growth,
begets offspring, and does not die childless ; and that out of
good fortune there springs insatiable woe for the race. But
I differ from others and am alone in my thought: it is the

1 The Greek verb includes the idea of confusion.

2 Literally ‘reared in addition (to those in the house)’.

135
 μετὰ μὲν πλείονα τίκτει,
σφετέραι δ᾽ εἰκότα γέννα: 760
οἴκων γὰρ εὐθυδίκων
καλλίτταις πτότμος αἰεί. =

φιλεῖ δὲ τίκτειν ὕβρις μὲν παλαι-

à νεάτουσαν ἐν κακοῖς βροτῶν γ65
ὕβριν, τότ᾽ ἢ τόθ᾽ ὅταν τὸ κύριον μόληι
Ἱνεαρὰ φάους κότον
δαίμονά Fre τὸν, ἄμαχον ἀπόλεμον ἀνίερον
θράσος, μέλαιναν μελάθροισιν “Atay, 770
εἰδομέναν τοκεῦσιν. —

Δίκα δὲ λάμπει μὲν ἐν δυσκάπνοις

δώμασιν, τόν τ᾽ ἐναίσιμον τίει 775
βίον’ τὰ χρυσόπαστα δ᾽ ἔσθλα σὺν πίνωι
χερῶν παλιντρόποισιν
ὄμμασι λιποῦσ᾽ ὅσια προσέμολε, δύναμιν οὐ
σέβουσα πλούτου παράσημον αἴνωι" 780
πᾶν δ᾽ ἐπὶ τέρμα vou. =

ἄγε δή, βασιλεῦ, Τροίας πτολίπορθ᾽,

᾿Ατρέως γένεθλον,
πῶς σε προσείττω; πῶς σε σεβίξω 785
μήθ᾽ ὑπεράρας μήθ᾽ ὑποκάμψας
καιρὸν χάριτος;
πολλοὶ δὲ βροτῶν τὸ δοκεῖν εἶναι
προτίουσι δίκην παραβάντες'
τῶι δυσπραγοῦντί τ᾽ ἐτιστενάχειν 790
πᾶς τις ἑτοῖμος" δῆγμα δὲ λύπης
οὐδὲν ἐφ᾽ ἧπαρ προσικνεῖται"
καὶ ξυγχαίρουσιν ὁμοιοπρετεῖς
FTr
766 örav nondum emendatum 767 νεαρὰ φάους corruptum ; κότον, licet falso
loco positum sit, genuinum esse videtur. exempli gratia scripseris φάος, κότον νεώρῃ
(φάος ex φάους fecit et transposuit Emperius; νεώρης vel vedpns Wilamowitz, νεάρη
Pasquali) 768 τε τὸν FTr: rirav Heimsoeth, fortasse recte ἄμαχον omisit
Tr 770 μέλαιναν... " Áray Auratus: pelaivas . . . dras FTr μελάθροις F
771 τοκεῦσι F 775 δώμασι F τ᾽ Hartung: δ᾽ ΕἼΤ 776 χρυσόπαστ᾽ ἐσθλὰ
Tr ἐσθλὰ hoc accentu ΕἼΤ: ἔσθλα Wilamowitz : ἔδεθλα Auratus 777 παλιν-
τρόποις F: παλίντροπ᾽ Tr: corr. Pauw 778 προσέμολε Hermann : προσέβα τοῦ FTr
783 πολίπορθ᾽ FTr: corr. Blomfield 785 σεβίξω (£ ex ζ factum; cf. ad 329) F:
σεβίζω Tr 789 παραβαίνοντες Tr; cf. ad 356 790-4 (βιαζόμενοι) afferuntur a
Stobaeo flor. 112. 12 (iv, p. τοῖο, 13 Hense) 790 τ᾽ Hermann: δ᾽ FTr Stob.
791 ἑτοῖμος Blomfield: ἕτοιμος codd., ut solent δῆγμα Tr Stob. : δεῖγμα F
792 προσεφικνεῖται Tr; cf. ad 356 793 καὶ νυκτὶ δὲ χαίρουσιν Stob.
136
impious deed that begets after it deeds which are more in
number but like to their own kind ; for the fate of the house
where justice is kept straight is always a fair offspring (of its
former fate).

But ancient insolence is wont to give birth to an insolence

which is young amid human misery, (to give birth to it)
sooner or later when the appointed day has come,! a fresh
rancour, an avenging daemon, an arrogant boldness against
which neither battle nor war avail and which has no share
in what is sacred, a black Ate for the house, resembling her
parents.

But Justice shines in sooty dwellings, and holds in regard

the righteous way of life; but the gold-bespangled mansions
where there is filth upon the hands she forsakes with eyes
averted and goes to what is clean, not revering the power of
wealth misstamped with praise; and everything she directs
to its proper end.
Enter Agamemnon in a traveller’s car; behind him,
half-hidden, sits Cassandra
Now tell me, King, sacker of Troy, offspring of Atreus,
how shall I address thee? How shall I pay thee homage,
neither overshooting nor falling short of the mark of graceful
behaviour ? Many in this world prize more the appearance of
reality, after passing the bounds of justice: anyone is ready
to echo the groans of him who is in distress—but the sting of
grief in no wise reaches to his heart; and resembling in
! The text of the following lines is extremely uncertain.

137
 ἀγέλαστα πρόσωπα βιαζόμενοι
„ar &

ὅστις δ᾽ ἀγαθὸς προβατογνώμων, 795

οὐκ ἔστι λαθεῖν ὄμματα φωτὸς
τὰ δοκοῦντ᾽ εὔφρονος ἐκ διανοίας
ὑδαρεῖ σαίνει φιλότητι.
σὺ δέ μοι τότε μὲν στέλλων στρατιὰν
Ἑλένης ἕνεκ᾽, οὐ γάρ (σ᾽) ἐπικεύσω, 800
κάρτ᾽ ἀπομούσως ἦσθα γεγραμμένος |
οὐδ᾽ εὖ πραπίδων οἴακα νέμων,
θράσος ἑκούσιον
ἀνδράσι θνήϊσκουσι kopígov:
νῦν δ᾽ οὐκ ἀπ᾽ ἄκρας φρενὸς οὐδ᾽ ἀφίλως 805
εὔφρων * & * * x *x
* X x πόνος EU τελέσασιν ?.
γνώσηι δὲ χρόνωι διαπευθόμενος
τόν τε δικαίως καὶ τὸν ἀκαίρως
πόλιν οἰκουροῦντα πολιτῶν.

ATAMEMNUN
πρῶτον μὲν "Apyos καὶ θεοὺς ἐγχωρίους 810
δίκη προσειπεῖν, τοὺς ἐμοὶ μεταιτίους
νόστου δικαίων θ᾽ ὧν ἐπραξάμην πόλιν
Πριάμον᾽ δίκας γὰρ οὐκ ἀπὸ γλώσσης θεοὶ
κλυόντες ἀνδροθνῆτας ᾿Ιλιοφθόρους
ἐς αἱματηρὸν τεῦχος οὐ διχορρόπτως 815
ψήφους ἔθεντο' τῶι δ᾽ ἐναντίωι κύτει.
Eis προσήιϊει χειρὸς οὐ πληρουμένωι,
καπνῶι δ᾽ ἁλοῦσα νῦν ἔτ᾽ εὔσημος πόλις.
ἄτης θύελλαι τῶσι' συνθνήισκουσα δὲ
σποδὸς προπέμπει πίονας πλούτου πνοάς. 820
τούτων θεοῖσι χρὴ πολύμνηστον χάριν
τίνειν, ἐπείπερ χάρπαγὰς ὑπερκόπους
ἐπραξάμεσθα, καὶ γυναικὸς οὕνεκα
FTr
794 ‘ dyélaora” δὲ ‘ πρόσωπα Αἰσχύλος ex Phrynicho ut videtur (fr. 61 de Borries)
afferunt Συναγωγὴ λέξ. xpno. Anecd. Bekkeri i. 336. 30, Photius Berol. 15. 11
post 794 lacunam statuit Hermann, paroemiacum excidisse ratus 795 πρωβα-
τογνώμων (B ex r corr.) F 798 σαίνειν FTr: corr. Casaubon 800 σ᾽ suppl. Musgrave
803 θράσος F : θάρσος Tr locus nondum expeditus est, quo fit ut ne de reliqua
quidem sententiae parte (804) certe iudicari possit 804 ἀνδράσιν εὖ θνήσκουσι Tr;
cf. ad 356 806 lacunam post εὔφρων statuit Hense (ante εὔφρων Schneidewin,
Headlam); luseris fere sic: εὔῴρων ζαἰνῶ τόδ᾽ ἔπος προτέρων" ' ἡδὺς) πόνος εὔφρων
τις Tr; cf. ad 356 814 κλύοντες FTr: aoristum esse vidit Wilamowitz; cf. 680 ἰλίέου
φθορὰς FTr: corr. Karsten 822 xäprayas Tyrwhitt: xai παγὰς FTr ὑπερκότους
FTr: corr. Heath
138
aspect men who share another’s joy, forcing their unsmiling
countenances, (they . . .). But if anyone is a good judge of
a flock, he cannot be deceived by the look in a man's eyes
which, while feigning to come from a loyal mind, blandishes
with a watery friendship.
Thou, in past time when thou ledst forth the army for
Helen’s sake, wast pictured in my mind—for I will not keep
it hidden from thee—in exceedingly ugly colours, and as not
wielding well the helm of the mind... ; but now deeply and
in true friendship loyal (do I approve of the old saying,
‘sweet) is labour to those who have brought it to a good end’.
In course of time thou wilt learn by inquiry who of the
citizens stays at home in the city with justice and who does
so out of season.

Agamemnon. It is right that I should first address Argos

and the country’s gods, who have helped me in bringing
about my home-coming and the satisfaction I have exacted
from the city of Priam. For the gods, when they had heard
by no spoken word the parties’ claims, cast with no wavering
verdict into the urn of blood their votes for the death of men,
the destruction of Ilion, while to the opposite urn (mere)
expectation of the hand came near, and the urn was not
filled. By the smoke the conquered city is easily recognized
even yet. The gusts of destruction are alive, but the embers
are dying with the city and sending forth rich breaths of
wealth. For this the gods must be paid with ever-mindful
gratitude, since we have exacted payment for a presumptuous
139
 πόλιν διημάθυνεν ᾿Αργεῖον δάκος,
ἵππου νεοσσός, ἀσπιδηφόρος λεώς, 825
πήδημ᾽ ὀρούσας ἀμφὶ Πλειάδων δύσιν'
ὑπερθορὼν δὲ πύργον ὠμηστὴς λέων
ἄδην ἔλειξεν αἵματος τυραννικοῦ.
θεοῖς μὲν ἐξέτεινα φροίμιον τόδε’
τὰ δ᾽ ἐς τὸ σὸν φρόνημα, μέμνημαι κλυὼν 830
καὶ φημὶ ταὐτὰ καὶ συνήγορόν μ᾽ ἔχεις.
παύροις γὰρ ἀνδρῶν ἐστι συγγενὲς τόδε,
φίλον τὸν εὐτυχοῦντ᾽ ἄνευ φθόνον σέβειν"
δύσφρων γὰρ ἰὸς καρδίαν προσήμενος
ἄχθος διπλοίζει τῶι πετπαμένωι νόσον᾽ 835
τοῖς τ᾽ αὐτὸς αὐτοῦ πήμασιν βαρύνεται
καὶ τὸν θυραῖον ὄλβον εἰσορῶν στένει.
εἰδὼς λέγοιμ᾽ ἄν’ εὖ γὰρ ἐξεπίσταμαι
ὁμιλίας κάτοπτρον’ εἴδωλον σκιᾶς
δοκοῦντας εἶναι κάρτα πρευμενεῖς ἐμοί. 840
μόνος δ᾽ Ὀδυσσεύς, ὅσπερ οὐχ ἑκὼν ἔπλει,
φευχθεὶς ἑτοῖμος ἦν ἐμοὶ ceipapdpos’
εἴτ᾽ οὖν θανόντος εἴτε Kal φῶντος πέρι
λέγω. τὰ δ᾽ ἄλλα πρὸς πόλιν τε καὶ θεοὺς
κοινοὺς ἀγῶνας θέντες ἐν πανηγύρει | 845
βουλευσόμεσθα: καὶ τὸ μὲν καλῶς ἔχον
ὅπως Xpovizov εὖ μενεῖ βουλευτέον"
ὅτωι δὲ καὶ δεῖ φαρμάκων παιωνίων,
ἤτοι κέαντες ἢ τεμόντες εὐφρόνως
πειρασόμεσθα Trfju' ἀποστρέψαι νόσου. 850
νῦν δ᾽ ἐς μέλαθρα καὶ δόμους ἐφεστίους
ἐλθὼν θεοῖσι πρῶτα δεξιώσομαι,
οἶπερ πρόσω πέμψαντες ἤγαγον πάλιν.
νίκη δ᾽ ἐπείπερ Eotret’, ἐμπέδως μένοι.

KAYTAIMHZTPA
ἄνδρες πολῖται, πρέσβος ᾿Αργείων τόδε, 855
οὐκ αἰσχννοῦμαι τοὺς φιλάνορας τρόπους
λέξαι πρὸς ὑμᾶς" ἐν χρόνωι δ᾽ ἀποφθίνει
FTr
825 ἀσπιδηφόρος Blomfield: ἀσπιδηστρόφος Y : ἀσπιδοστρόφος Tr 826 δρούσας F
827 ὑπερθορῶν FTr 828 ἄδην F : ἄδδην (cf. Niciam in schol. Hom. E 203) Tr
830 κλυὼν Wilamowitz : κλύων FTr; cf. 814 831 ταὐτὰ Auratus: ταῦτα ΕἼΤ
832 sq. afferuntur a Stobaeo 38. 28 (iii, p. 713 Hense) 833 φθόνου Tr; idem
voluit (cf. comm.), etsi ψόγου habet, Stobaeus: φθόνων F 834 καρδίᾳ Casaubon,
nescio an recte 835 πεπαμμένῳ YTr: corr. Porson 836 αὐτοῦ sic FTr
versus 834-7 in suspicionem vocat Beazley 842 ἕτοιμος FTr; cf. 791 850
πήματος τρέψαι νόσον FTr.: corr. Porson (an scribendum ἀποτρέψαι)
140
robbery, and for a woman’s sake the city was laid in the dust
by the fierce beast of Argos, the brood of the horse, the
shield-bearing host, which launched itself with a leap at
the setting of the Pleiades ; and springing over the wall the
ravening lion licked his fill of the blood of princes.
To the gods I have spoken this long preface; but with
regard to thy feelings, I have heard and remember them, and
I agree, and thou hast in me one who will speak on thy side.
Yes, rare among men are those to whom it is natural to
respect without envy a friend who is fortunate; since the
venom of malevolence, besetting the heart, doubles the load
to the owner of the disease: at one and the same time he is
weighed down by his own sorrows and groans at the sight of
the other’s prosperity. With knowledge—for I am well
acquainted with that mirror, intercourse—I may pronounce
image of a shadow those who seem most devoted to me.
Odysseus alone, the very man who did not sail of his own
will, when once in harness proved to me a ready trace-horse,
whether he of whom I am speaking be dead now or alive.
And for the rest, for what concerns the city and the gods, we
shall arrange general meetings and take counsel in full
assembly: and where a thing is well, our counsel must be how
it may endure so and abide; but where, on the other hand,
anything is in need of healing remedies, we shall endeavour
to turn to flight the harm of the disease by sage use either of
knife or of cautery. But now I will go to my house and my
home with its hearth and first salute the gods, who sped me
forth and have brought me back. And now that victory has
attended me, may it abide securely!
Enter from the house Clytemnestra, followed by
matdservants
Clytemnestra. Men of the city, noble elders of Argos
present here, I shall not be ashamed to describe to you my
love for my husband: in time men’s timidity fades away. It
141
 τὸ τάρβος ἀνθρώποισιν. οὐκ ἄλλων πάρα
μαϑοῦσ᾽ ἐμαυτῆς δύσφορον λέξω βίον
τοσόνδ᾽ ὅσονπερ οὗτος ἦν ὑπ᾽ Ἰλίωι. 860
τὰ μὲν γυναῖκα πρῶτον ἄρσενος δίχα
ἦσθαι δόμοις ἐρῆμον ἔκπαγλον κακόν,
[πολλὰς κλύουσαν κληδόνας παλιγκότους,]
καὶ τὸν μὲν ἥκειν, τὸν δ᾽ ἐπεισφέρειν κακοῦ
κάκιον ἄλλο πῆμα λάσκοντας δόμοις. 865
καὶ τραυμάτων μὲν εἰ τόσων ἐτύγχανεν
ἀνὴρ ὅδ᾽ ὡς πρὸς οἶκον ὠχετεύετο
φάτις, τέτρηται δικτύου πλέω λέγειν.
εἰ δ᾽ ἦν τεθνηκὼς ὡς ἐπλήθυον λόγοι,
τρισώματός T&v Γηρύων ὁ δεύτερος 870
[πολλὴν ἄνωθεν, τὴν κάτω γὰρ οὐ λέγω,
χθονὸς τρίμοιρον χλαῖναν ἐξηύχει λαβεῖν,
ἅπαξ ἑκάστωι κατθανὼν μορφώμοτι.
τοιῶνδ᾽ ἕκατι κληδόνων παλιγκότων
πολλὰς ἄνωθεν ἀρτάνας ἐμῆς δέρης 875
ἔλυσαν ἄλλοι πρὸς βίαν, λελημμένης.
ἐκ τῶνδέ Toi παῖς ἐνθάδ᾽ οὐ παραστατεῖ,
ἐμῶν τε καὶ σῶν κύριος πιστωμάτων,
ὡς χρῆν, 'OpéoTrns μηδὲ θαυμάσηις τόδε.
τρέφει γὰρ αὐτὸν εὐμενὴς δορύξενος, 880
Στροφίος ὁ Φωκεύς, ἀμφίλεκτα πήματα
ἐμοὶ προφωνῶν, τόν θ᾽ ὑπ᾽ ᾿ἸΙλίωι σέθεν
κίνδυνον, εἴ τε δημόθρους ἀναρχία
βουλὴν καταρρίψειεν, ὥς τι σύγγονον
βροτοῖσι τὸν πεσόντα λακτίσαι πλέον. 885
τοιάδε μέντοι σκῆψις οὐ δόλον φέρει.
ἔμοιγε μὲν δὴ κλαυμάτων ἐπίσσντοι
πηγαὶ κατεσβήκασιν, οὐδ᾽ ἔνι σταγών.
ἐν ὀψικοίτοις δ᾽ ὄμμασιν βλάβας ἔχω
τὰς ἀμφί σοι κλαίουσα λαμττηρουχίας 890
ἀτημελήτους αἰέν’ ἐν δ᾽ ὀνείρασιν
λεπταῖς ὑπαὶ κώνωπος ἐξηγειρόμην
FTr
860 ἐπ᾽ iw Tr 862 ἔρημον FTr ἐἔκπαγγλον Tr 863 ἡδονὰς FTr: corr.
Auratus versum interpolatum esse vidit Ahrens 867 φχετεύετο FTr
868 φάτις (non φάσις) etiam F τέτρηται Ahrens: τέτρωται FTr 869
ἐπλήθυνον FTr: corr. Porson 870 τὰν Wellauer: 7' ἂν FTr 871 eiecit
Schütz 872 λαβὼν FTr: corr. Paley 876 λελημμένης suspectum 878
πιστευμάτων FTr: corr. Spanheim 879 μὴ δὲ FTr 88I στρόφιος FTr de
voce ἀμφίλεκτα est quod dubites 882 τ᾽ ὑπ᾽ (superscr. glossa. ἐωνικὸν) Tr 884
ds τι Hartung (sententia parum intellecta): dore FTr 888 κατεσβήκασιν (super v
scr. 0, super f scr. 7, i.e. καθεστήκασιν) F 889 κλάβας F 890 ἀμφὶ σοὶ FTr
142
is not from others that I have learnt: it is my own life of
whose misery I shall tell, all the long while he was before
Tlion. In the first place it is a fearful grief that a woman
should sit at home all alone without the man, and that one
should come, and on top of him another (and another) should
bring fresh reports of evil, each worse than the last, which
they cry out for the house to hear. As for wounds, if this man
received as many as rumour thereof was led, like water in
conduits, to our home, he has holes in him more in number
than a net. And if his deaths had been as plentiful as were
the stories, he might truly have boasted, triple-bodied
Geryon the second, that he had got a threefold cloak of
earth, slain once under each form. In consequence of such
adverse rumours many a time by force did others loose the
suspended noose from my neck when I was caught in it.!
And this is why our son is not standing here by our side, the
warrant of thy pledges and mine, as he should be, Orestes ;
and do not think this strange. He is in the care of our
friendly ally, Strophius the Phocian, who warned me of two-
fold (?) disaster, thine own peril before Ilion, and the chance
that lack of a ruler, asserted noisily by the people, might
overthrow deliberation, as it is part of men’s nature to kick
a man all the more when he is down. In such an excuse as this
there can be no deceit.
For myself now, the gushing fountains of my tears have
run dry, and there is no drop left therein. And my eyes which
went late to bed are sore from weeping for the light-bearings
concerning thee that were ever neglected ; and in my dreams
I would be woken by the faint rushings of a gnat, and hear it
! I have attempted to render λελημμένης, but see the commentary.

143
 ῥιπαῖσι, θωύσσοντος, ἀμφί σοι πάθη
ὁρῶσα πλείω τοῦ ξυνεύδοντος χρόνον.
νῦν, ταῦτα πάντα τλᾶσ᾽, ἀπενθήτωι φρενὶ 895
λέγοιμ᾽ ἂν ἄνδρα τόνδε τῶν σταθμῶν κύνα,
σωτῆρα ναὸς πρότονον, ὑψηλῆς στέγης
στῦλον ποδήρη, μονογενὲς τέκνον πατρί,
γαῖαν φανεῖσαν ναυτίλοις παρ᾽ ἐλπίδα,
[κάλλιστον ἧμαρ εἰσιδεῖν ἐκ χείματος.) 900
ὁδοιπόρωι διψῶντι πηγαῖον ῥέος.
[τερπνὸν δὲ τἀναγκαῖον ἐκφυγεῖν ἅπαν.]
τοιοῖσδέ τοί νιν ἀξιῶ προσφθέγμασιν’
φθόνος δ᾽ ἀπέστω πολλὰ γὰρ τὰ πρὶν κακὰ
ἠνειχόμεσθα. νῦν δέ μοι φίλον κάρα 905
ἔκβαιν᾽ ἀπήνης τῆσδε, μὴ χαμαὶ τιθεὶς
τὸν σὸν πόδ᾽, ὦναξ, Ἰλίου πορθήτορα.
δμωιαί, τί μέλλεθ᾽, αἷς ἐπέσταλται τέλος
πέδον κελεύθου στορνύναι πετάσμασιν;
εὐθὺς γενέσθω πορφυρόστρωτος πόρος, 910
ἐς δῶμ᾽ ἄελπτον ὡς ἂν ἡγῆται Δίκη.
τὰ δ᾽ ἄλλα φροντὶς οὐχ ὕπνωι νικωμένη
θήσει δικαίως σὺν θεοῖς {εἱμαρμέναΐ.
AT. Λήδας γένεθλον, δωμάτων ἐμῶν φύλαξ,
ἀπουσίαι μὲν eltras εἰκότως ἐμῆι, 915
μακρὰν γὰρ ἐξέτεινας" ἀλλ᾽ ἐναισίμως
αἰνεῖν, παρ᾽ ἄλλων χρὴ τόδ᾽ ἔρχεσθαι γέρας.
καὶ τἄλλα μὴ γυναικὸς ἐν τρόποις ἐμὲ
ἄβρυνε, μηδὲ βαρβάρον φωτὸς δίκην
χαμαιπετὲς βόαμα προσχάνηις ἐμοί, 920
μηδ᾽ einacı στρώσασ᾽ ἐπίφθονον πόρον
τίθει’ θεούς τοι τοῖσδε τιμαλφεῖν χρεών,
ἐν ποικίλοις δὲ θνητὸν ὄντα κάλλεσιν
FTr
893 ἀμφὶ σοὶ FTr 897 ὑψιλῆς Tr 898 στύλον Tr: στόλον F 899 γαῖαν Blom-
field: καὶ γῆν FTr παρελπίδα Tr 900-2 eiecit Headlam (902 iam Blomfield) ;
goo, 902 huc non pertinere manifestum est, de 901 dubito 901 cf. Suid. s.v. πλέκος :
καὶ ἀπὸ τοῦ ῥέω ῥέος. Αἰσχύλος ὁδοιπόρωι... ῥέος 903 τοί νιν Schütz: τοίνυν FTr
προσφέγμασιν Tr 905 δέ μοι Bothe: 8° ἐμοὶ FTr 907 ἄναξ F 908 τέλος
F: τάδε Tr 909 orpwrvivas FTr: corr. Elmsley 913 εἱμαρμένα nondum emenda-
tum 919 μὴ δὲ FTr βαρβάθου (super 8 scr. p F 920 βόαμα (priori a superscr.
ἢ) Ἐπ βόημα Tr 921 μὴ δ᾽ FTr 923 versum attulerunt Eclogae, Cramer Anecd.
Oxon. ii. 455. 4 κάλλη" τὰ πορφυρᾶ iuäria. Εὔπολις" ' βάπτειν τὰ κάλλη ” (fr. 333 K.). καὶ
κέραμος καλλάϊνος. Αἰσχύλος" " ἐν ποικίλοις δὲ θνητὸν ὄντα κάλλεσιν ’ (Αἰσχύλος... . κάλλεσιν
omisit Cramer, teste A. Adler ad Suid. iii. 17 s.v. κάλλη), Suid. (s.v. κάλλη) τὰ πορφυρᾶ
ἱμάτια... Αἰσχύλος" * ἐν ποικίλοις κάλλεσιν ’, Suid. (s.v. KdAats) . . . «Αἰσχύλος * ἐν ποικίλοις
χρώμασιν᾽, Etym. gen. B s.v. κάλλαια (= Etym. Magn. p. 486. 49) . . . τὰ πορφυρᾶ γὰρ
κάλλη ἐκαλοῦντο. Εὔπολις" * βάπτειν ... ed’. καὶ Αἰσχύλος" * ἐν ποικίλοις... κάλλεσιν ’,
Etym. Gud. 5.ν. κάλλεα (cod. Par. 2631, Cramer Anecd. Par. iv. 23. 13; Par. 2636,
ibid. p. 63. 21; ‘ Sorbon.' ap. Gaisford ad Et. M. 1.1), ubi post Eupolidis locum «Αἰσχύλος
ἐν τῶι Ἀγαμέμνονι (dv τ. A. om. Par. 2636)" * ev ποικίλοις δὲ (δὲ Par. 2631, * Sorbon.’: γὰρ
Par. 2636) . .. κάλλεσιν"
144
trumpeting; since I saw there things befalling thee more
than could have passed in the time that slept with me.
Now, after enduring all this, with a mind freed from
mourning I would pronounce this man here the watchdog of
his abode, the saving forestay of a ship, the grounded pillar
of a high roof, a sole-born child to a father, land appearing to
sailors beyond their hope, to the thirsty wayfarer a flowing
spring.! Such are the terms wherewith I deem him worthy
to be addressed ; and far be envy,? for many are the evils of
the past that we endured. But now I pray thee, beloved,
come down from this car—but set not on the ground, O king,
thy foot that has destroyed Ilion. Handmaidens, why do ye
delay, ye upon whom the office has been enjoined to strew
the ground he walks upon with tapestries ? Straightway let
there be made a path spread with purple, that Justice may
conduct him into his unhoped-for home. And for the rest,
care not overcome by sleep shall arrange it in just fashion,
with the help of the gods.3
In the meantime the handmaids have begun to spread out
the tapestries
Ag. Offspring of Leda, guardian of my house, thy speech
was indeed well-suited to my absence, for thou hast drawn
it out to great length; but fitting praise—that is a gift of
honour that should come from others. And for the rest, do
not pamper me as though I were a woman, and do not adore
me as if I were a man of the East, with prostrations and
open-mouthed acclaim, nor, by strewing my path with
vestures, bring down envy upon it. It is the gods whom we
should honour with such ceremonies: to tread, a mortal,
! It is not quite certain whether this line (gor) is in its original place.
2 φθόνος gives one of the key-notes to the following scene. I have therefore kept the
one rendering, ‘envy’, although in some of the passages ‘jealousy’ (of the gods) might
seem more appropriate. 3 At the end of the line the text is uncertain.
4872-1 145 L
 βαίνειν ἐμοὶ μὲν οὐδαμῶς ἄνευ φόβου.
[λέγω κατ᾽ ἄνδρα, μὴ θεόν, σέβειν ἐμέ.] 935
χωρὶς ποδοψήστρων τε καὶ τῶν ποικίλων
κληδὼν Aurel καὶ τὸ μὴ κακῶς φρονεῖν
θεοῦ μέγιστον δῶρον" ὀλβίσαι δὲ χρὴ
βίον τελευτήσαντ᾽ ἐν εὐεστοῖ φίληι.
εἶπον, τάδ᾽ ὡς πράσσοιμ᾽ ἂν εὐθαρσὴς ἐγώ. 930
KA. καὶ μὴν τόδ᾽ εἰπέ, μὴ παρὰ γνώμην, ἐμοί:
AF. γνώμην μὲν ἴσθι μὴ διαφθεροῦντ᾽ ἐμέ.
KA. ηὔξω θεοῖς δείσας ἂν ὧδ᾽ ἔρξειν τάδε;
ΑΓ. εἴπερ τις εἰδώς γ᾽ εὖ τόδ᾽ ἐξεῖττεν τέλος.
ΚΛ, τί δ᾽ ἂν δοκεῖ σοι Πρίαμος, εἰ τάδ᾽ ἤνυσεν; 935
ΑΓ. ἐν ποικίλοις ἂν κάρτα μοι βῆναι δοκεῖ.
KA. μή vuv τὸν ἀνθρώπειον αἰδεσθῆις ψόγον.
ΑΓ. φήμη γε μέντοι δημόθρους μέγα σθένει.
ΚΛ, ὁ δ᾽ ἀφθόνητός γ᾽ οὐκ ἐπίφηλος πέλει.
AT. οὔτοι γυναικός ἐστιν ἱμείρειν μάχης. 940
KA. τοῖς δ᾽ ὀλβίοις γε καὶ τὸ νικᾶσθαι πρέπει.
ΑΓ. ἢ καὶ σὺ νίκην τήνδε δήριος τίεις;
KA. πιθοῦ’ κρατεῖς μέντοι παρεὶς [y’] ἑκὼν ἐμοί.
ΑΓ, ἀλλ᾽ εἰ δοκεῖ σοι ταῦθ᾽, ὑπαί τις ἀρβύλας
λύοι τάχος, πρόδουλον ἔμβασιν ποδός" 945
καὶ τοῖσδέ μ᾽ ἐμβαίνονθ᾽ ἁλουργέσιν θεῶν
μή τις πρόσωθεν ὄμματος βάλοι φθόνος,
πολλὴ γὰρ αἰδὼς δωματοφθορεῖν ποσὶν
φθείροντα πλοῦτον ἀργυρωνήτους θ᾽ ὑφάς.
τούτων μὲν οὕτω’ τὴν ξένην δὲ πρευμενῶς 950
τήνδ᾽ ἐσκόμιφε᾽ τὸν κρατοῦντα μαλθακῶς
θεὸς πρόσωθεν εὐμενῶς προσδέρκεται"
ἑκὼν γὰρ οὐδεὶς δουλίωι χρῆται φυγῶς!ι.
αὕτη δὲ πολλῶν χρημάτων ἐξαίρετον
ἄνθος, στρατοῦ δώρημ᾽, ἐμοὶ ξυνέσπετο. 955
ἐπεὶ δ᾽ ἀκούειν σοῦ κατέστραμμαι τάδε,
εἶμ᾽ ἐς δόμων μέλαθρα πορφύρας πατῶν.
FTr
925 eiecit Wilamowitz 930 εἶπον, τάδ᾽ ὡς Weil: ei πάντα δ᾽ ds FTr 932
διαφθεροῦντ᾽ ex διαφθαρεντ᾽ corr. F 933 mira collocatio verborum ; δείσας dv ηὔξω
θεοῖσιν ὧδ᾽ temptabat Hermann, fortasse recte épée Headlam : ἕρδειν F: ἔρδειν
Tr 934 ἐξεῖπον FTr: corr. Auratus 935 δοκῆ FTr: corr. Stanley 936
δοκῇ ex δοκεῖ factum F: δοκεῖ (superscr. n) Tr 937 μὴ νῦν F αἰδεσθῆς Tr: αἰδεσθεὶς
F 943 κρατεῖς Weil: κράτος FTr παρείς Bothe: πάρες FTr γ᾽ del. Wecklein
646 oiv τοῖσδέ μ᾽ Tr ἐμβαίνοντ᾽ ἀλ- FTr 948 σωματοφθορεῖν FTr : corr.
Schütz πόσιν FTr (superscr. gl. ἄνδρα ἁπλῶς Tr) 954 αὕτη Auratus: αὐτὴ FTr
956 κατέσταμαι (superscr. gl. ἰωνικὸν κατέστην) Tr, qui in scholio nota ἡμέτερον distincto
haec profert : εὕρηται καὶ κατέστραμμαι, ἤγουν κατεβλήθην κτλ, 957 δόμους (in fine
superscr. ων) F
146
upon embroidered fineries is to me by no means free from
fear. Different is the ring of the words ‘footmats’ and
‘embroideries’, and a mind without presumption is the god's
greatest gift ; one should praise a man’s fortune when he has
ended his life in welcome prosperity. I have said how I for
my part should act herein with good confidence.
Clyt. Aye, and tell me this too, of thine honest mind,—
Ag. My mind, be assured, I shall not allow to be falsified.
Clyt. Wouldst thou, in an hour of terror, have vowed to
the gods to do this as I request thee now?
Ag. Yes, if any man with full knowledge had prescribed
the performance of this ritual.
Clyt. And what dost thou think that Priam would have
done, if he had achieved such a deed ? (embroideries.
Ag. I think that for sure he would have walked upon
Clyt. Have no scruple then for the reproach of men.
Ag. And yet the voice of the people has great power.
Clyt. Aye, but if a man is not envied no one vies with him.
Ag. It is not a woman’s part to long for strife.
Clyt. Nay, but for the fortunate even to yield victory is
becoming. |
Ag. What? This ‘victory’ in this contest—does it mean
so much to thee?
Clyt. Yield; truly thou art the superior if of thine own
will thou hast left it to me.
Ag. Well, if this be thy will, let someone quickly loose my
shoes, which in slaves’ stead serve the feet to step on; and
as I walk on these purple draperies of the gods, may no
glance of envious eye strike on me from afar. For a strong
feeling restrains me from wasting our house’s substance with
my feet, spoiling therewith wealth, textures purchased for
silver.
In the meantime one of the handmaids has
finished untying and taking off Agamemnon’s
shoes. The king steps down from the car
So much, then, for that. But this stranger here, bring her
kindly into the house ; at him who uses his power gently the
god looks with favour from afar, for no one bears the yoke
of slavery of his own will. And she has come with me as the
flower chosen especially for me from among much wealth,
the army’s gift. But now, since I have been borne down and
must listen to thee in this, I will go into the halls of my house
treading purple underfoot. He walks slowly towards the door
147
 KA. ἔστιν θάλασσα, τίς δέ νιν κατασβέσει;
τρέφουσα πολλῆς πορφύρας ἰσάργυρον
κηκῖδα παγκαίνιστον, εἱμάτων Bapäs 960
οἴκοις δ᾽ ὑπάρχει τῶνδε σὺν θεοῖς, ἄναξ,
ἔχειν. πένεσθαι δ᾽ οὐκ ἐπίσταται δόμος.
πολλῶν πατησμὸν δ᾽ εἱμάτων ἂν ηὐξάμην,
δόμοισι προὐνεχθέντος ἐν χρηστηρίοις
ψυχῆς κόμιστρα τῆσδε μηχανωμένηι. 965
pizns γὰρ οὔσης φυλλὰς ἵκετ᾽ ἐς δόμους
σκιὰν ὑπερτείνασα ceipiou κυνός’
καὶ σοῦ μολόντος δωματῖτιν ἑστίαν
θάλπος μὲν ἐν χειμῶνι σημαίνεις μολόν᾽
ὅταν δὲ τεύχηι Ζεὺς [τ ἀπ᾽ ὄμφακος πικρᾶς 970
οἶνον, τότ᾽ ἤδη ψῦχος ἐν δόμοις πέλει
ἀνδρὸς τελείου δῶμ᾽ ἐπιστρωφωμένου,
Ζεῦ Ζεῦ τέλειε, τὰς ἐμὰς εὐχὰς τέλει"
μέλοι δέ τοί σοι τῶὥνπερ ἂν μέλληις τελεῖν.

ΧΟΡΟΣ
τίπτε μοι τόδ᾽ ἐμπέδως 975
δεῖμα προστατήριον
καρδίας τερασκόπον
πωτᾶταιϊ,
μαντιπολεῖ δ᾽ ἀκέλευστος ἄμισθος ἀοιδά,
οὐδ᾽ ἀποττύσας δίκαν 980
δυσκρίτων ὀνειράτων
θάρσος εὐπειθὲς 1-
get φρενὸς φίλον θρόνον;
χρόνος δ᾽, ἐπεὶ Trpuuvnoíov ξὺν ἐμβολαῖς
ψάμμος ἄμπτα, παρή- 985
βησεν, εὖθ᾽ ὑπ᾽ Ἴλιον
ὥρτο ναυβάτας στρατός’ —
FTr
958 personae nota deest in F 959 eis ἄργυρον FTr: corr. Salmasius 961
οἶκος FTr: corr. Porson 963 δειμάτων FTr: distinxit Canter εὐξάμην FTr
965 κομίστρα T μηχανωμένης FTr: corr. Abresch 967 ὑπερτίνασα F
969 μολών FTr: corr. H. Voss, Blomfield 970 τ᾽ del. Auratus 972
Emorpodwuevov Tr: ἐπιστρεφ- F : corr. Victorius 974 μέλοι Tr et schol. metr.
Triclinianum in FTr ad 810 adscriptum : μέλῃ (superscr. o) F τοι σοι Tr et
Schol. metr. (ad 810) quale exstat in Tr: σοι ro schol. metr. quale exstat in F (mero
lapsu): σοι F 976 δεῖγμα F 978 ποτᾶται F: ποτᾶτ᾽ Tr: corr. Meineke
979 ἄμισθος ἀοιδά, μαντιπολεῖ δ᾽ ἀκέλευστος Tr 980 ἀποπτύσαι Tr 982
εὐπιθὲς FTr: corr. Jacob 982 sq. ἵζει Scaliger: ἔξει F: te Tr 984 ἐπεὶ F:
em Tr ἐξυνεμβόλοις FTr: corr. Casaubon 985 ψάμμος Wecklein : ψαμμίας FTr
ἄμπτα Wilamowitz: axara F: ἀκάτας Tr 985 sq. παρήβησ᾽ Tr
148
 Clyt. The sea is there—and who shall drain it?—that
breeds an ever-renewed gush of abundant purple, precious
as silver, for the dyeing of vestures. Our house, by the gods'
grace, O King, has a supply of these things, and the house
knows not how to be poor. I would have vowed the treading
underfoot of many robes, if at the seat of an oracle I had heard

that task declared for the house when I was devising means
for bringing safely back this man's life. For as, when the root
remains, the foliage returns to the house, stretching over it
a shade against the dog-star, so, by thy coming home to the
hearth of thy house, thou dost signify that warmth has come
home in winter, and when from the sour grape Zeus is
making wine, then at once there is coolness in the house when
the consummate master is moving about his home.
Agamemnon goes within

Zeus, Zeus Consummator, consummate my prayers ; and

may'st thou take thought for what thou dost intend to con-
summate. She follows Agamemnon into the house

Chor. Why is it that so constantly this dread, as a

guardian in front of my auguring heart, flutters to and fro;
that a chant unbidden and unhired plays the prophet, and
I cannot spurn it, like dreams without clear meaning, and
let assuring trust settle on the throne of my mind ? Time has
grown old since with the throwing-in of the mooring-cables
the sand flew up, when the naval host set forth to Ilion ;
149
 πεύθομαι δ᾽ ἀπ᾽ ὀμμάτων
νόστον, αὐτόμαρτυς dv"
τὸν δ᾽ ἄνευ λύρας ὅμως 990
ὑμνωιδεῖ
θρῆνον Ἐρινύος αὐτοδίδακτος ἔσωθεν
θυμός, οὐ τὸ πᾶν ἔχων
ἐλπίδος φίλον θράσος.
σπλάγχνα δ᾽ οὔτοι ματάι- 995
zeı πρὸς ἐνδίκοις φρεσὶν
τελεσφόροις δίναις κυκλούμενον κέαρ.
εὔχομαι δ᾽ ἐξ ἐμᾶς
ἐλπίδος ψύθη πεσεῖν
ἐς τὸ μὴ τελεσφόρον. == 1000

μάλα γάρ τοι τᾶς πολλᾶς ὑγιείας

Ἰάκόρεστον᾽ τέρμα νόσος γὰρ
γείτων ὁμότοιχος ἐρείδει.
καὶ πότμος εὐθυπορῶν 1005
x * * X * * *

ἀνδρὸς ἔπαισεν ἄφαντον ἕρμα,

καὶ τὸ μὲν πρὸ χρημάτων
κτησίων ὄκνος βαλὼν
σφενδόνας ἀπ᾽ εὐμέτρον, 1010
οὐκ ἔδυ πρόπας δόμος
πλησμονᾶς γέμων ἄγαν,
οὐδ᾽ ἐπόντισε σκάφος.
πολλά τοι δόσις ἐκ Διὸς ἀμφιλα- 1015
φής τε καὶ ἐξ ἀλόκων ἐπετειᾶν
νῆστιν ὥλεσεν νόσον. —

τὸ δ᾽ ἐπὶ γᾶν πεσὸν ἅπαξ

θανάσιμον πρόπαρ ἀνδρὸς
μέλαν αἷμα τίς ἂν 1020
πάλιν ἀγκαλέσαιτ᾽ ἐπαείδων;
οὐδὲ τὸν ὀρθοδαῆ
FTr
990 ὅμως Auratus: ὅπως FTr 992 ἐριννὺς FTr: corr. Porson 995 sq.
ματάζει FTr: « inseruit Hermann 998 ἐξ ἐμᾶς F: am’ ἐμᾶς roe Tr 999
ψύδη FTr: corr. Henr. Stephanus 1001 sqq. initium strophae graviter corruptum
est ; res singulas in commentario excussi IOOI ydp τοι F: γέ (superscr. γὰρ) τοι δὴ
Τι 1004 544. ex antistropho apparet hemiepes dactylicum hic excidisse, sed
utrum post 1004 (Klausen) an post 1005 (Heath) incertum est 1007 ἕρμα FTr
1012 πλησμονᾶς Schütz : πημονὰς ΕἼΤ 1015 éc om. Tr Acs F 1016 καὶ
ἐξ Tr: κἀξ F 1018 πεσόνθ᾽ FTr: corr. Auratus IOI9 πρόπαρ F : προπάροιθ᾽ Tr
1020 sq. tis τ᾽ ἀγκαλ- Tr
150
and I learn their return from my eyes, myself the witness;

but still my soul within me chants, self-taught, the lyreless

dirge of the Erinys, not feeling at all the welcome confidence

of hope. Man's inward parts do not vainly bode—the heart,

in eddies that bring fulfilment, whirling against the mind

which is conscious of just retribution. But I pray that from

my expectation it may fall as a falsehood to the ground so as

to bring no fulfilment.

... of excessive health . . . for disease, its neighbour with

a common party-wall, presses against it. And human fortune,

sailing straight, . . . strikes upon a hidden reef. Then, if, to

save the possessions, cautious diffidence flings a part over-

board by a well-measured throw, the whole house does not

sink under an over-freight of abundance, nor does he (i.e. the

skipper) drown his ship in the sea. A large gift from Zeus,

ample and coming from yearly furrows, destroys the plague

of famine.

But a man's dark blood, once fallen aforetime in death on

the earth— who by incantation can call it up again? Else it

could not have happened that even he who knew the right
ISI
 τῶν φθιμένων ἀνάγειν
Ζεὺς ἂν ἔπαυσεν ἐπ᾽ ἀβλαβείαι.
εἰ δὲ μὴ τεταγμένα 1025
μοῖρα μοῖραν ἐκ θεῶν
εἶργε μὴ πλέον φέρειν,
προφθάσασα καρδία
γλῶσσαν ἂν τάδ᾽ ἐξέχει'
νῦν δ᾽ ὑπὸ σκότωι βρέμει 1030
θυμαλγής τε καὶ οὐδὲν ἐπτελττομέ-
va ποτὲ καίριον ἐκτολυπεύσειν,
ZWITUPOUNEVAS φρενός. ==

KAYTAIMHZTPA
εἴσω κομίφου Kal σύ, Κασσάνδραν λέγω, 1035
ἐπεί σ᾽ ἔθηκε Ζεὺς ἀμηνίτως δόμοις
κοινωνὸν εἶναι χερνίβων, πολλῶν μέτα
δούλων σταθεῖσαν κτησίου βωμοῦ πέλας.
ἔκβαιν᾽ ἀπήνης τῆσδε μηδ᾽ ὑπερφρόνει.
καὶ παῖδα γάρ τοί φασιν ᾿Αλκμήνης ποτὲ 1040
πραϑέντα τλῆναι δουλίας μάφης TPiat.
εἰ δ᾽ οὖν ἀνάγκη τῆσδ᾽ ἐπιρρέτοι τύχης,
ἀρχαιοπλούτων δεσποτῶν πολλὴ Xapıs'
oi δ᾽ οὔποτ᾽ ἐλπίσαντες ἤμησαν καλῶς,
&pol τε δούλοις πάντα, Kal πταρὰ στάθμην 1045
* ἘΞ x * EEK HK X

ἔχεις παρ᾽ ἡμῶν οἷάπερ vonizeran.

ΧΟ. σοί τοι λέγουσα παύεται, σαφῆ λόγον.
ἐντὸς δ᾽ ἁλοῦσα μορσίμων ἀγρευμάτων
πείθοι᾽ ἄν, εἰ Treo” ἀπειθοίης δ᾽ ἴσως.
ΚΛ. ἀλλ᾽ εἴπερ ἐστὶ μὴ χελιδόνος δίκην 1050
ἀγνῶτα φωνὴν βάρβαρον κεκτημένη,
Ἰέσω φρενῶν λέγουσα πείθω νιν λόγωι.
ΧΟ. &mov: τὰ λῶιστα τῶν παρεστώτων λέγει,
πείθου λιποῦσα τόνδ᾽ ἁμαξήρη θρόνον.
FTr
1024 ἄν ἔπαυσεν Martin: αὔτ᾽ ἔπαυσ᾽ FTr αὐλαβεία F : ἀβλαβείᾳ γε Tr 1030
βρέμει ex βλέπει factum F 1031 θυμαλγής re καὶ οὐδὲν Er om. Tr (cf. p. 13 54.)
1035 κασάνδραν FTr 1039 un δ᾽ FTr 1041 δουλείας μάζης βία F : καὶ ζυγῶν
θίγειν βίᾳ Tr: δουλίας μάζης θιγεῖν Keck; non liquet 1042 ἐπιρρέπει Tr
1044 οἵδ᾽ FTr: relativum agnovit Stanley 1045 παρὰ στάθμην Tr: παραστάθμων
F post 1045 lacunam statuit Hartung 1048 ἁλοῦσα C. G. Haupt: ἂν οὖσα FTr
1050 huc pertinere videtur Hesychii glossa χελιδόνος δίκην (χελιδόσι cod., ordine
alphabetico interrupto: corr. M. Schmidt): τοὺς βαρβάρους χελιδόσιν ἀπεικάζουσιν διὰ
τὴν ἀσύν[θ]ετον (corr. Wilamowitz) λαλιάν 1052 ἔσω φρενῶν λέγουσα non sanum ;
an corruptela latius pateat incertum est
152
way to raise up from the dead was stopped by Zeus that no
harm might be done. And did not established destiny prevent
my portion from winning more from the gods, my heart
outrunning my tongue would pour this out; but as it is, it
mutters in the dark, pained to the core and without hope
ever to accomplish any timely purpose, while my mind is
ablaze.
Enter Clytemnestra
Clyt. Get thee within, thou too; thou, Cassandra; since
Zeus without anger has made thee to share with our house
its lustral water, standing among our many slaves by the
household altar. Come down from this car, and be not over-
proud. They say that even Alcmena’s son once was sold to
bondage and endured to touch! the bread of servitude. But
if the constraint of that lot should indeed befall one, then
to have masters old in wealth is a thing to be deeply thankful
for; whereas those who have reaped a rich harvest that they
never expected are harsh to their slaves in all things, and
with exact strictness (allot to them no more than their due.
With us, on the other hand, the servants. . . .) Thou hast
heard from me what our custom is.
Chor. It is to thee she has spoken, and now she has
finished, a plain speech. Taken as thou art within the toils
of fate, pray obey her, if thou wilt obey; though perhaps
thou wilt disobey.
Clyt. Nay, if she is not, like a swallow, possessed of an
unintelligible foreign tongue, speaking . . . I attempt (?) to
persuade her by my words.
Chor. Go with her; she says that which is best as things
now are. Leave thy seat here on the wagon and obey.
1 Text uncertain,

153
 ΚΛ, οὔτοι θυραίαι τῆιδ᾽ ἐμοὶ σχολὴ πάρα 1055
᾿ τρίβειν τὰ μὲν γὰρ ἑστίας μεσομφάλου
ἕστηκεν ἤδη μῆλα ἱπρὸς σφαγὰς πυρόςΤ'
[ὡς οὔποτ᾽ ἐλπίσασι τήνδ᾽ ἕξειν χάριν]
σὺ δ᾽ εἴ τι δράσεις τῶνδε, μὴ σχολὴν τίθει-
εἰ δ᾽ ἀξυνήμων οὖσα μὴ δέχηι λόγον, 1060
σὺ δ᾽ ἀντὶ φωνῆς φράτε καρβάνωι χερί.
ΧΟ. ἑρμηνέως ἔοικεν ἡ ξένη τοροῦ
δεῖσθαι: τρόπος δὲ θηρὸς ὡς veatpétou.
KA. ἢ μαίνεταί γε καὶ κακῶν κλύει φρενῶν,
ἥτις λιποῦσα μὲν πόλιν νεαίρετον 1065
ἥκει, χαλινὸν δ᾽ οὐκ ἐπίσταται φέρειν
πρὶν αἱματηρὸν ἐξαφρίξεσθαι μένος.
οὐ μὴν πλέω ῥίψασ᾽ ἀτιμασθήσομαι.
ΧΟ. ἐγὼ δ᾽, ἐποικτίρω γάρ, οὐ θυμώσομαι.
ἴθ᾽, ὦ τάλαινα, τόνδ᾽ ἐρημώσασ᾽ ὄχον 1070
ἑκοῦσ᾽ ἀνάγκης τῆσδε καίνισον τυγόν.
ΚΑΣΣΑΝΔΡΑ
ὀτοτοτοῖ ποποῖ δᾶ"
“AttoAAov, Ἄπολλον.
ΧΟ. τί ταῦτ᾽ ἀνωτότυξας ἀμφὶ Λοξίου;
OÙ γὰρ τοιοῦτος ὥστε θρηνητοῦ τυχεῖν. — 1075

ΚΑ. ὀτοτοτοῖ ποποῖ δᾶ"

"Ἄπολλον, Ἄπολλον.
ΧΟ. ἡ δ᾽ αὖτε δυσφημοῦσα τὸν θεὸν καλεῖ
οὐδὲν προσήκοντ᾽ ἐν γόοις παραστατεῖν. ==

KA. "Απολλον, "Ἄπολλον, ἀγυιᾶτ᾽, 1080

ἀπόλλων ἐμός"
ἀπώλεσας γὰρ οὐ μόλις τὸ δεύτερον.
ΧΟ. χρήσειν ἔοικεν ἀμφὶ τῶν αὑτῆς κακῶν"
μένει τὸ θεῖον δουλίαι περ ἐν φρενί. —
1055-66 FTr 1067-84 MFTr
1055 ϑυραίαν FTr: corr. Casaubon τῆιδ᾽ Musgrave: τήνδ᾽ ΕἼΤ 1057 in verbis
πρὸς σφαγὰς πυρός corruptela latet; πάρος quod Musgravium secuti plerique edunt non
sufficit 1058 eiecit Wilamowitz 1064 ἦ Tr:3 F φρενῶν κλύει (sed
superscriptis β΄ et a’ corr.) F 1067 accedit M 1068 μὴν mFTr: μὴ M
1069 ἐποικτείρω codd.: corr. Kirchhoff ' . 1070 óyxov (ὄχλον fuisse suspicantur
Hermann, Vitelli) M IO7I εἴκουσ᾽ (ἀνάγκηι τῆιδε servato) Robortello ἀνάγκη()
τῆιδε codd.: corr. Casaubon 1072 xaodvipa codd. 1072 et 1076 ὀτοτοτοτοῖ
M πομποῖ FTr 1073 et 1077 ὦπολλον ὥπολλον M 1074 sq. Schol.
Eur. Phoen. 1028 Αἰσχύλος" ‘ τί ταῦτ᾽ ἀνωλόλυξας . . . rvxetv' 1078 ἡ δ᾽ M: ἥδ᾽
FTr 1080 ἀγυιάτ᾽ hoc accentu hic et 1085 codd., nisi quod 1085 ἀγυίατ᾽ F (1080 -d-
ex -εῦ- corr. F) 1083 αὐτῆς codd. 1084 περ ἐν Schütz: παρ᾽ ἕν M : παρὲν
F: παρὸν Tr

154
 Clyt. Come, I have no leisure to waste time here out of
doors. For they, the sheep, are already standing before (?)
the central hearth, for (?) slaughter ; and thou, if thou mean-
est to do aught of what I say, make no delay ; but if thou art
without understanding and my words reach thee not, then
instead of speech show thy meaning with outlandish hand.
Chor. It is an interpreter, a clear one, that the stranger
seems to need. Her behaviour is like that of a wild creature
newly caught.
Clyt. Mad she is, and hearkens to a wild mind, she who
has come hither from a newly captured city and yet has not
the sense to bear the bridle before foaming her spirit away
in blood. I will not waste more words to be disdained.
Exit Clytemnestra
Chor. But I, for I pity her, will not be angry. Come now,
unhappy one, leave this carriage empty, and of thine own
willtake on thyself the unaccustomed yoke of this constraint.

Cassandra. O woe, woe, woe! alas! Apollo, Apollo!

Chor. Why dost thou thus cry woe while invoking Loxias ?
he is not such that the chanter of a dirge should come his
way.

Cass. O woe, woe, woe! alas! Apollo, Apollo!

Chor. Once more with ill-omened sounds she invokes the
god whom it in no wise befits to be present at lamentations.
Cassandra leaves the car and begins to walk
towards the door, in front of which she sees the
stone symbol of Apollo Agyteus
Cass. Apollo, Apollo, Agyiates! a very Apollon to me! For
thou hast ruined me utterly for the second time.
Chor. She is going to prophesy, it seems, about her own
miseries. The divine gift abides in the mind, though it be
enslaved.
155
 ΚΑ. “AtroAAov, “AtroAAov, ἀγνιᾶτ᾽, 1085
ἀπόλλων ἐμός"
& ποῖ ποτ᾽ ἤγαγές με; πρὸς ποίαν στέγην;
ΧΟ. πρὸς τὴν ᾿Ατρειδῶν. εἰ σὺ μὴ τόδ᾽ ἐννοεῖς,
ἐγὼ λέγω cor καὶ τάδ᾽ οὐκ ἐρεῖς ψύθη. —

ΚΑ. [ἃ &]
μισόθεον μὲν οὖν, πολλὰ συνίστορα 1090
αὐτοφόνα κακὰ καρατόμα,
ἀνδροσφαγεῖον καὶ πεδορραντήριον.
XO. ἔοικεν εὔρις ἡ ξένη κυνὸς δίκην
εἶναι, ματεύει δ᾽ ὧν ἀνευρήσει φόνον. —

ΚΑ. μαρτυρίοισι γὰρ τοῖσδ᾽ ἐπιπείθομαι" 1095

κλαιόμενα τάδε βρέφη σφαγὰς
ὀπτάς τε σάρκας πρὸς πατρὸς βεβρωμένας.
ΧΟ. ἦμεν κλέος σοῦ μαντικὸν πεπυσμένοι
fhuevt, προφήτας δ᾽ οὔτινας ματεύομεν. ==

ΚΑ. ἰὼ πόποι, τί ποτε μήδεται; 1100

τί τόδε νέον ἄχος; μέγα
μέγ᾽ ἐν δόμοισι τοῖσδε μήδεται κακόν,
ἄφερτον φίλοισιν,
δυσίατον' ἀλκὰ δ᾽
ἑκὰς ἀποστατεῖ.
ΧΟ. τούτων ἄιδρίς εἶμι τῶν μαντευμάτων᾽ 1105
ἐκεῖνα δ᾽ ἔγνων: πᾶσα γὰρ πόλις βοᾶι. —

ΚΑ. ἰὼ τάλαινα, τόδε γὰρ τελεῖς;

τὸν ὁμοδέμνιον πόσιν
λουτροῖσι φαιδρύνασα' πῶς φράσω τέλος;
1085-94 MFTr 1095-1109 MFG(raro notatur, cf. p. 30 sq.)Tr
1088 el σὺ" τὸ, μὴ δ᾽ ἐννοεῖς F: el. τό περ μὴ δ᾽ ἐννοεῖς Tr 1089 καὶ τάδ᾽ M: κᾶτα
δ᾽ FTr ψφψύδη (cf. 999) Tr 1090 d d versus separati vicem obtinet in M, deest in
FTr £vviaropa M 1091 αὐτοφόνα mFTr: αὐτόφονα M καρατόμα Kayser: xapravat
M: κἀρτάναι F : κἀρτάνας Tr 1092 ἀνδρὸς σφάγιον codd. (« in duarum litterarum
litura M): corr. Casaubon et Dobree πεδορραντήριον M : πέδον ῥαντήριον (pavr- sine
spiritu m) mFTr 1093 εὗρις ex edpoos corr. M: εὗρις FTr IO94 μαντεύει M
ἂν εὑρήση M: ἐφευρήσει FTr: corr. Porson 1095 accedit G μαρτυρίοις codd.:
corr. Pauw μὲν γὰρ ΕἼΤ τοῖσδε πεπείθομαι codd. : corr. Abresch 1096
τὰ βρέφη FTr 1098 μην (superscr. ἦμεν.) M : ἦμεν FTr memvapévous (una
littera erasa) M 1099 ἦμεν M: ἦμεν FTr aut hic aut, id quod minus probabile
videtur, in versu antecedente verbo ἦμεν ex proximo versu inserto genuina vox
expulsa est μαστεύομεν codd. : corr. Schütz 1100 πόποι mFTr: ποποῖ M
I1OI ἄχθος (superscr. *&xos-) M μέγα del. Enger, fortasse recte; cf. 1108 1103
φίλοισιν Tr: φίλοισι MF ἀλκὰν F 1105 versum affert Photius Berol. p. 47. 13
1106 βοᾷ πόλις FTr 1108 πόσιν del. Enger; cf. 1101
156
 Cass. Apollo, Apollo, Agyiates! a very Apollon to me! Ah,
whither hast thou brought me? to what house?
Chor. The house of the Atridae. If thou art not aware of
that, I tell it to thee and thou shalt not call it false.

Cass. Nay, an ungodly one, that can bear witness to many

evil deeds of kindred murder and beheading, a slaughter-
place for men, a place where the ground is sprinkled.
Chor. Keen-scented like a hound the stranger seems to be,
and she is searching for the murder of those whose murder
she will find.

Cass. Yes, for here is the testimony that I trust: here are
babes crying because of their slaughter and their roasted
flesh that their father devoured!
Chor. We had heard of thy renown as a seer; but we seek
none of those who speak in the name of the gods.

Cass. O horror, what plot is this? what new grief is here ?

A mighty evil is being plotted here in the house, unbearable
to kinsmen and friends, beyond cure; and help stands far
aloof.
Chor. Of these prophecies I have no knowledge; the
former I recognized, for the whole city rings with them.

Cass. Ah, wretch! so thou wilt accomplish this? The

husband of thy bed, thou hast washed him clean in the bath,
157
 τάχος γὰρ τόδ᾽ ἔσται"
προτείνει δὲ χεὶρ ἐκ 1110
χερὸς ὀρέγματα.
XO. οὔπω ξυνῆκα’ νῦν γὰρ ἐξ αἰνιγμάτων
ἐπαργέμοισι θεσφάτοις ἀμηχανῶ. ==

KA. E É, παπαῖ παπαῖ, τί τόδε φαίνεται;

3j δίκτυόν τί γ᾽ "AiSou; 1115
ἀλλ᾽ ἄρκυς ἡ ξύνευνος, ἡ ξυναιτία
φόνου. Στάσις δ᾽ ἀκόρετος γένει
κατολολυξάτω θύματος λευσίμον.
ΧΟ. ποίαν Ἐρινὺν τήνδε δώμασιν κέληι
ἐπορθιάξτειν; οὔ pe φαιδρύνει λόγος. 1120
ἐπὶ δὲ καρδίαν ἔδραμε κροκοβαφὴς
-σταγῶν, ἅτε καὶ δορὶ πτωσίμοις
ξυνανύτει βίου δύντος αὐγαῖς"
ταχεῖα δ᾽ ἄτα πέλει. —

KA. dä, ἰδοὺ ἰδού ἄπεχε τῆς βοὸς 1125

τὸν ταῦρον’ ἐν πέπλοισιν
μελαγκέρωι λαβοῦσα μηχανήματι
τύπτει" πίτνει δ᾽ (Ev) ἐνύδρωι τεύχει.
δολοφόνου λέβητος τέχναν σοι λέγω.
ΧΟ. οὐ κομπάσαιμ᾽ ἂν θεσφάτων γνώμων ἄκρος 1130
εἶναι, κακῶι δέ τῶι προσεικάφω τάδε.
ἀπὸ δὲ θεσφάτων τίς ἀγαθὰ φάτις
βροτοῖς τέλλεται; κακῶν γὰρ διαὶ
πολνεπεῖς τέχναι θεσπιωιδῶν
φόβον φέρουσιν μαθεῖν. — 1135

ΚΑ. ἰὼ ἰὼ ταλαίνας κακόποτμοι τύχαι"

MF(G)Tr
1110 δὲ MFTr: γὰρ G ex χεὶρ factum χεῖρ᾽ M IIIT χειρὸς M Opeypara
Hermann e scholio : dpeyouérä corr. ex dpeydpeva M : ὀρεγμένα ΕἼΤ 1113 er’
ἀργέμοισι FTr ἀμνημονῶ G 1114 παπαὶ παπαὶ FTr IIIS ἦ Μ: ἢ FTr
* Audov Schütz : ἀΐδου codd. (superscr. συνέζησις Tr) 1117 ἀκόρεστος codd.: corr.
Bothe ante IIIS chori nota erasa in M 1119 ἐρινῦν M : ἐριννὺν FTr
1120 φαιδρύνει mFTr: φαιδρὺ M ante II2I Cassandrae nota erasa in M 1122
καὶ δορὶ Casaubon: καὶ δορία M : καὶ δωρία F: δωρία Tr πτώσιμος codd.: corr. Casau-
bon 1123 ξυνανυτεῖ codd. (ex £vvavvri corr. M) 1I12544M:G@aFTr βοῆς Tr
1127 μελάγκέρωι M (sed littera « in rasura ex v facta, et superscr. -v.): μελαγκέρων F:
μελάγκερων Tr: de scholio quod et μελάγκερων et μελαγκέρωι testatur v. comm.
λαβοῦσα codd. : λαθοῦσα schol. M 1128 πίτνει ex πιτνεῖ corr. M (teste Vitellio):
πιτνεῖ FTr ἐν supplevit Schütz 1120 rexvav Weil: τύχαν codd. 1130
ἄκρος eX ἄκρως corr. M 1132 τίς FTr: τις M: et τις (ἐνίοτε) et τίς interpretatur
schol. M 1133 βροτοῖσι F στέλλεται codd. (στέλεται, superscr. διὰ τὸ μέτρον,
Tr, quippe qui 1122 dre, superscr. καθὰ, pro pyrrhichio haberet) : corr. Emperius
διαὶ Hermann: δὴ αἱ FTr: διὰ M 1134 θεσπιωιδὸν codd.: corr. Casaubon 1135
φέρουσιν Tr: φέρουσι ΜῊ μαθεῖν suspectum
158
and then—how shall I tell the end? for speedily it will be;
hand after hand stretches forth and reaches out.
Chor. Not yet do I understand; for now the riddling
sayings leave me bewildered by their dim oracles.

Cass. Ah, ah! alas, alas! what is this that comes in view ?
some net of Hades? Nay, but the snare that shares his bed,
that shares the guilt of murder. Let insatiate Discord raise
to the race an exulting shout over the sacrifice that is to be
avenged by stoning.
Chor. What dost thou mean by thy Erinys whom thou
summonest to raise her voice over the house? Thy words
cheer me not. To my heart there rushes a drop of saffron dye,
the very one which to men fallen by the spear arrives to-
gether with the rays of setting life; and doom is swift at
hand,

Cass. Ah, ah! Look, look! Keep the bull from the cow!
In a garment she has caught him, with black contrivance of
the horned one, and strikes ; and he falls in a vessel of water.
It is the device of a treacherous murdering caldron whereof
I tell thee.
Chor. I would not boast high skill as a judge of oracles;
but this I liken to some evil thing. But from oracles what
good message ever comes to men? By (uttering) evil do the
wordy arts of prophets bring fear to learn.!

Cass. O woe for a wretched one’s hapless lot! For it is my

3 Text uncertain.

159
 τὸ γὰρ ἐμὸν θροῶ πάθος ἐπεγχύδαν.
τί δή pe δεῦρο τὴν τάλαιναν ἤγαγες;
οὐδέν ποτ᾽ ef μὴ Euvdavounevnv τί γάρ; |
ΧΟ. φρενομανής τις εἶ θεοφόρητος, ἀμ- 1140
φὶ δ᾽ αὑτᾶς θροεῖς
νόμον ἄνομον, οἷά τις ξουθὰ
ἀκόρετος βοᾶς, φεῦ, φιλοίκτοις φρεσὶν
Ἴτυν Ἴτυν στένονσ᾽ ἀμφιθαλῆ κακοῖς
ἀηδὼν βίον. — 1145

KA. ἰὼ Io λιγείας ἀηδοῦς μόρον᾽

περέβαλον γάρ οἱ πτεροφόρον δέμας
θεοὶ γλυκύν τ᾽ αἰῶνα κλαυμάτων &rep:
ἐμοὶ δὲ μίμνει σχισμὸς ἀμφήκει δορί.
ΧΟ. πόθεν ἐπισσύτους Beopöpous [τ] ἔχεις 1150
ματαίους δύας,
τὰ δ᾽ ἐπίφοβα δυσφάτωι κλαγγᾶι
μελοτυπεῖς ὁμοῦ τ᾽ ὀρθίοις ἐν νόμοις;
πόθεν ὄρους ἔχεις θεσπεσίας ὁδοῦ
κακορρήμονας; == 1155

ΚΑ. ἰώ,
γάμοι γάμοι Πάριδος ὀλέθριοι φίλων.
ἰὼ Σκαμάνδρου πάτριον ποτόν"
τότε μὲν ἀμφὶ σὰς ἀϊόνας τάλαιν᾽
ἠνντόμαν τροφοῖς"
νῦν δ᾽ ἀμφὶ Κωκυτόν τε κἀχερουσίους 1168
ὄχθους ἔοικα θεσπιωιδήσειν τάχα.
ΧΟ. τί τόδε τορὸν ἄγαν ἔπος ἐφημίσω;
νεογνὸς ἂν ἀΐων μάθοι.
πέπληγμαι δ᾽ ἅπερ δήγματι φοινίωι
1137-59 MF(G)Tr 1160-4 F(G)Tr
1137 ἐπεγχύδαν Headlam : ἐπεγχέασα M: ἐπαγχέασα ΕἾΤ 1138 τί Heimsoeth:
ποῖ codd. 1139 οὐδέν ποτ᾽ M: οὐδέποτ᾽ F: οὐ δή ποτ᾽ Tr 1141 αὐτᾶς FTr:
αὐτᾶς M 1142 ἄνομόν y Tr old Tr: ofa M: ota F 1143 ἀκόρεστος codd.:
corr. Aldina βοᾶς F: βοᾶις M : βορᾶς Tr φεῦ M: om. FTr φιλοίκτοις ταλαΐίναις
φρεσὶν F: φιλοίκτοισι φρεσὶν Tr: ταλαίναις (sic, ut videtur, M, teste Vitellio, ταλαίνᾶς m)
φρεσὶν M: φιλοίκτοις genuinum esse, ταλαίναις vero interpolatum perspexit Dobree
1146 ἀηδοῦς μόρον Dobree (μόρον ἀηδόνος Hermann): ἀηδόνος μόρον codd. 1147
περεβάλοντο M : περιβαλόντες FTr: corr. Hermann (nisi quod περίβαλον scripsit) 1148
γρ(άφεται) αἰῶνα in margine m: ἀγῶνα MFTr 1150 τ᾽ del. Hermann IISI
δοιάς M, corr. in δύας m 1152 ἐπίφοβα MFTr: ἐπὶ φόβωι (-w, scil. w supra a
scripto, F2Tr2) mF?Tr? 1153 μελοτυπεῖς (o supra prius e scripto) F : μολοτ- (e supra
prius o scripto) Tr 1154 ἔχη FG (ex ἔχει factum F) 1157 τόπον G 1158
τάλαινα F post 1159 deficit M 1163 ἂν ἀΐων Karsten: ἀνθρώπων ΕἼΤ
1164 πέπλημαι Tr ἅπερ Franz (ὅπως Hermann): ὑπὸ F: ὑπαὶ Tr
160
own sufferings that I cry aloud, adding them to the cup.
Wherefore hast thou brought me hither, poor wretch that
I am? for nothing but to share in death; what else ?
Chor. Frenzied art thou, with heaven-sent transports, and
about thyself thou criest aloud a tuneless tune, like a tawny
one, who, insatiate of lamenting cry, alas, with melancholy
mind bewails with ‘Itys, Itys’ her life that has woe flourish-
ing on either side, a nightingale.

Cass. Oh the death of the melodious nightingale: her the

gods cloaked in a feathered shape and a sweet life without
cries of woe; but what waits for me is cleaving by the two-
edged weapon.
Chor. Whence hast thou the rushing pangs of heaven-sent

transport, pangs that are vain, and shapest these fearful

songs alike with inarticulate sound and in loud notes?
Whence hast thou the limits of thy prophetic way, that tell
of evil?

Cass. O the wedding, the wedding of Paris, that brought

the destruction of his kin! O Scamander, my native stream!
Once on thy shores, unhappy maid, I was nursed and grew;
but now, it seems, it is by Cocytus and the banks of Acheron
that I shall soon chant my prophecies.
Chor. What word is this that thou hast uttered all too
plain? A little child that heard it might understand. I am
4872-1 16 I M
 δυσαλγεῖ τύχαι μινυρὰ [κακὰ] θρεομένας, 1165
θραύματ᾽ ἐμοὶ κλύειν. —

ΚΑ. ἰώ,
πόνοι πόνοι πόλεος ὀλομένας τὸ πᾶν.
ἰὼ πρόπυργοι θυσίαι πατρὸς
πολυκανεῖς βοτῶν ποιονόμων᾽ ἄκος
δ᾽ οὐδὲν ἐπήρκεσαν 1170
τὸ μὴ (oU) πόλιν μὲν ὥσπερ οὖν ἔχει παϑεῖν᾽
ἐγὼ δὲ Τθερμόνους τάχ᾽ ἐμ πέδωι βαλῶ.
ΧΟ. ἑπόμενα προτέροισι τάδ᾽ ἐφημίσω,
καί τίς σε κακοφρονῶν τίθη-
σι δαίμων ὑπερβαρὴς ἐμπίτνων 1175
μελίφειν πάθη γοερὰ θανατηφόρα"
τέρμα δ᾽ ἀμηχανῶ. =

ΚΑ. καὶ μὴν ὁ χρησμὸς οὐκέτ᾽ ἐκ καλυμμάτων

ἔσται δεδορκὼς νεογάμου νύμφης δίκην,
λαμπρὸς δ᾽ ἔοικεν ἡλίου πρὸς ἀντολὰς 1180
πνέων ἐσάιξειν, ὥστε κύματος δίκην
κλύειν πρὸς αὐγὰς τοῦδε πήματος πολὺ
ueigov' φρενώσω δ᾽ οὐκέτ᾽ ἐξ αἰνιγμάτων.
καὶ μαρτυρεῖτε συνδρόμως ἴχνος κακῶν
ῥινηλατούσηι τῶν πάλαι πεπραγμένων. 1185
τὴν γὰρ στέγην τήνδ᾽ οὔποτ᾽ ἐκλείπει χορὸς
ξύμφθογγος οὐκ εὔφωνος" οὐ γὰρ εὖ λέγει.
καὶ μὴν πεπωκώς γ᾽, ὡς θρασύνεσθαι πλέον,
βρότειον αἷμα κῶμος ἐν δόμοις μένει
δύσπεμττος ἔξω, συγγόνων ᾿Ἐρινύων. 1190
ὑμνοῦσι δ᾽ ὕμνον δώμασιν προσήμεναι
πρώταρχον ἄτην, ἐν μέρει δ᾽ ἀπέπτυσαν
εὐνὰς ἀδελφοῦ τῶι πατοῦντι δυσμενεῖς,
ἥμαρτον ἢ κυρῶ τι τοξότης τις ὥς;
ἢ ψευδόμαντίς εἶμι θυροκόπος φλέδων; 1195
F(G)Tr
1165 övoayyei FTr: corr. Canter μινύρα F: ulvvpa Tr κακὰ del. Schütz
1166 θαύματ᾽ Tr 1167 πόλεως F ὁλωμένας F: ὀλουμένας Tr: corr. Casaubon
1171 οὐ supplevit Hermann ἔχει Tr: ἔχειν F 1172 θερμόνους corruptum :
θερμὸν ῥοῦν Musgrave ἐμπέδω (-w Tr) FTr: distinxit Casaubon 1173 προτέροις
FTr: corr. Pauw 1174 καὶ τίς oe F: τίς oe καὶ Tr: affirmativam sententiam
esse viderunt Brumoy, Heath κακοφρονεῖν FTr: corr. Schütz 11745q. κακοφρονεῖν
δαίμων ποιεῖ ὑπερβαρὺς Tr 1175 ἐμπίτνων F: ἐμπιτνῶν Tr 1176 θανατηφόρα
Tr: θανατοφόρα F 1179 νύμφης GTr: νύμφας F 1181 ἐσάιξειν Bothe: ἐς
ἥξειν FTr 1182 κλύζειν Auratus: κλύειν FTr 1187 ξύμφθογγος Tr:
ξύμφογγος G: σύμφογγος F 1100 ἐριννύων FTr 1191 δώμασι FTr: v add.
Porson 1192 πρώταρχος F 1104 κυρῶ Ahrens: τηρῶ FTr τίς ὥς FTr
162
stricken as with a deadly sting at thy cruel fate, as thou
shrillest plaintive notes, shattering for me to hear.

Cass. O distress, distress of my city utterly destroyed! O

my father’s sacrifices before! the walls, prodigal in slaughter
of the grazing flock; yet they availed not for any cure to
prevent the city from meeting with what is in fact its fate;
while I shall soon pour the hot stream of my blood (?)? on the
ground.
Chor. These utterances follow in the same strain as those
before, and some malign power that falls exceeding heavy on
thee causes thee to chant of doleful death-fraught sufferings ;
but the end I cannot see.

Cass. Lo, now shall my oracle no longer be one that looks

forth from a veil like a newly wedded bride, but methinks
bright as a fresh wind it will rush in and blow towards the
sunrise, so that like a wave there shall surge toward the light
a far greater woe than this; I shall teach you now no more in
riddles. And bear ye witness unto me as in close pursuit I
scent out the track of the ills enacted long ago. I say that
from this house there never departs a choir that sings in
concert and yet with no pleasant sound, for not pleasant are
its words. Aye, and it has drunk—and grown bolder thereby
—of human blood ; and it abides, that revel-rout, within the
house, not to be turned away, the rout of the Erinyes bred
in the race. And they sing their song, besetting the chambers,
sing of the mind’s blindness that began it all, and each in
turn they vent their loathing for a brother’s bed, hostile to
him who trampled on it. Have I missed, or do I make a hit,
like an archer? Or am I a false prophet, a door-rapper, a
1 Or ‘for’.
2 Text uncertain.
163
 ἐκμαρτύρησον προὐμόσας τὸ μὴ εἰδέναι
λόγωι παλαιὰς τῶνδ᾽ ἁμαρτίας δόμων.
ΧΟ. καὶ πῶς ἂν ὅρκου πῆγμα γενναίως παγὲν
παιώνιον γένοιτο; θαυμάξω δέ σον,
πόντου πέραν τραφεῖσαν ἀλλόθρουν Ἱπόλινϊ 1200
κυρεῖν λέγονσαν, ὥσπερ εἶ παρεστάτεις.
ΚΑ. μάντις μ᾽ ᾿Απόλλων τῶιδ᾽ ἐπέστησεν τέλει.
ΧΟ. μῶν καὶ θεός Trep ἱμέρωι πεπληγμένος; 1204
ΚΑ. πρὸ τοῦ μὲν αἰδὼς ἦν ἐμοὶ λέγειν τάδε. 1203
ΧΟ. ἁβρύνεται γὰρ πᾶς τις εὖ πράσσων πλέον. 1205
ΚΑ. ἀλλ᾽ ἦν παλαιστὴς κάρτ᾽ ἐμοὶ πνέων χάριν.
ΧΟ. ἢ καὶ τέκνων els ἔργον ἤλθετον νόμωι;
ΚΑ. ξυναινέσασα Λοξίαν ἐψευσάμην.
XO. ἤδη τέχναισιν ἐνθέοις ἡἠιρημένη;
ΚΑ. ἤδη πολίταις πάντ᾽ ἐθέσπιφον πάθη. 1210
ΧΟ. πῶς δῆτ᾽ ἄνατος ἦσθα Λοξίου κότωι;
ΚΑ. ἔπειθον οὐδέν᾽ οὐδέν, ὡς τάδ᾽ ἤμπλακον.
ΧΟ. ἡμῖν γε μὲν δὴ πιστὰ θεσπίζειν δοκεῖς.
ΚΑ. ἰοὺ ἰού, ὦ ὦ κακά.
ὑπ᾽ αὖ με δεινὸς ὀρθομαντείας πόνος 1215
στροβεῖ ταράσσων φροιμίοις » κ = *.
ὁρᾶτε τούσδε τοὺς δόμοις ἐφημένους
νέους, ὀνείρων προσφερεῖς μορφώμασιν;
παῖδες θανόντες [ὡσπερεὶ πρὸς τῶν φίλων,
χεῖρας κρεῶν πλήθοντες οἰκείας βορᾶς, 1220
σὺν ἐντέροις τε σπλάγχν᾽, ἐποίκτιστον γέμος,
πρέπουσ᾽ ἔχοντες, ὧν πατὴρ ἐγεύσατο.
ἐκ τῶνδε ποινὰς φημὶ βουλεύειν τινὰ
λέοντ᾽ ἄναλκιν ἐν λέχει στρωφώμενον
οἰκουρόν, οἴμοι, τῶι μολόντι δεσπότη!. 1225
[iur φέρειν γὰρ χρὴ τὸ δούλιον Ξυγόν.]
νεῶν δ᾽ ἄπαρχος Ἰλίου τ᾽ ἀναστάτης
οὐκ οἶδεν οἷα γλῶσσα μισητῆς κυνὸς
F(G)Tr
1196 καὶ μαρτύρησον Tr μ᾽ εἰδέναι FTr: corr. Bothe, Dobree 1198 ὅρκος
πῆμα FTr: corr. Auratus 1200 πόλιν Vix sanum : πάνυ Schoemann: τὸ πᾶν Weil
versum 1203 (χορός praefixerunt FTr) post 1204 (xaodvöpa praef. FTr) collocavit
Hermann 1203 προτοῦ FTr 1205 ἀβρύνεται F : βαρύνεται Tr | I2II
dvaros Canter: ἄνακτος FTr 1212 οὐδέν᾽ Canter: οὐδὲν FTr 1216 φροιμίοις
ἐφημένους (in fine os superscr. Tr) FTr: ‘ crediderim, dum librarii oculus in sequens
[1217] ἐφημένους aberrasset, genuinum vocabulum excidisse, quod sine codicum ope
frustra reponere tentaveris ’ (Butler); φροιμίοις δυσφροιμίοις exempli gratia Hermann
1219 ὡσπερεὶ valde suspectum 1221 γέμος FTr: yé G 1224 λέοντ᾽ ἄναλκιν
quid sibi velit nescio; versum delevit Wilamowitz 1226 eiecitA. Ludwig
δούλειον Tr ζυγὸν ex ζυγω corr. F 1227 νεῶν τ᾽ FTr: corr. G. Vossius 1228
εὖ oldev Tr

164
babbler ? Bear witness, upon oath, that thou hast not heard,
and dost not know, the ancient wrong-doings of this house.
Chor. And how could the security of an oath, solidly
secured, avail for remedy? But I marvel at thee that, bred
beyond the sea, speaking a strange tongue, thou hittest full
on the mark with what thou sayest, as if thou hadst been
present here.
Cass. It was the seer Apollo who appointed me to this
office.
Chor. Being smitten—even he, a god—with desire?
Cass. Aforetime I was ashamed to speak of this.
Chor. Aye, we are all too fastidious when things are well
with us.
Cass. Well then, he was a wrestler who mightily breathed
his grace upon me.
Chor. Came ye also to the work of begetting children, as
the custom is?
Cass. I gave my consent, and then I played Loxias false.
Chor. When already thou wert seized by the art wherein
the god is present?
Cass. Already I prophesied to my people all that was to
befall them.
Chor. Surely then thou wast not unharmed by the wrath
of Loxias ?
Cass. 1 could make none believe anything that I said,
after that offence.
Chor. Yet to us thy divinations seem worthy of belief.
Cass. Ah! ah! oh misery! Once more a fearful pang of true
prophesying whirls me round, troubling me with . . . preludes.
Do ye see these young ones seated close to the house, in
semblance like the shapes of dreams? Children slain . . . by
their near ones, with their hands full of meat, the meat of
their own flesh served as food ; and, clear to see, they hold
their vitals and their entrails, a lamentable load, whereof
their father tasted. Hence vengeance is being plotted, I tell
you, by some craven lion,! wallowing in the bed, a stay-at-
home, ah me, who plots against the master on his return.
But the commander of the fleet and overthrower of Ilion
knows not, after the fine and lengthened speeches that the
lewd creature's tongue has uttered with radiant friendliness,?
1 Text doubtful; see the commentary.
2 More accurately ‘beaming disposition’.
I65
 λέξασα κἀκτείνασα φαιδρόνους δίκην
ἄτης λαθραίου τεύξεται κακῆι τύχηι. 1230
τοιάδε τόλμα' θῆλυς ἄρσενος φονεύς"
ἔστιν: τί νιν καλοῦσα δυσφιλὲς δάκος
τύχοιμ᾽ ἄν; ἀμφίσβαιναν, ἢ Σκύλλαν τινὰ
οἰκοῦσαν ἐν πέτραισι, ναυτίλων βλάβην,
θύουσαν "Ardou μητέρ᾽ ἄσπονδόν τ᾽ "Apn 1235
φίλοις πνέουσαν; ὡς δ᾽ ἐπωλολύξατο
À παντότολμος, ὥσπερ ἐν μάχης Tponflt
δοκεῖ δὲ χαίρειν νοστίμωι σωτηρίαι.
καὶ τῶνδ᾽ ὁμοῖον εἴ τι μὴ πείθω" τί γάρ;
τὸ μέλλον ἥξει" καὶ σύ μ᾽ ἐν τάχει παρὼν 1240
ἄγαν [y'] ἀληθόμαντιν οἰκτίρας ἐρεῖς.
ΧΟ. τὴν μὲν Θυέστου δαῖτα παιδείων κρεῶν
ξυνῆκα καὶ πέφρικα καὶ φόβος μ᾽ ἔχει
κλυόντ᾽ ἀληθῶς οὐδὲν ἐξηικασμένα"
τὰ δ᾽ ἄλλ᾽ ἀκούσας ἐκ δρόμου πεσὼν τρέχω. 1245
ΚΑ. ᾿Αγαμέμνονός σέ φημ᾽ ἐπόψεσθαι μόρον.
ΧΟ. εὔφημον, ὦ τάλαινα, κοίμησον στόμα.
ΚΑ. ἀλλ᾽ οὔτι παιὼν τῶιδ᾽ ἐπιστατεῖ λόγωι.
XO. οὔκ, εἴπερ ἔσται γ᾽" ἀλλὰ μὴ γένοιτό Tres.
ΚΑ. σὺ μὲν κατεύχηι, τοῖς δ᾽ ἀποκτείνειν μέλει. 1250
ΧΟ. τίνος πρὸς ἀνδρὸς τοῦτ᾽ ἄχος πορσύνεται;
KA. ἢ κάἀάρτ᾽ fap’ àvf παρεκόπτης χρησμῶν ἐμῶν.
ΧΟ. τοὺς γὰρ τελοῦντας οὐ ξυνῆκα μηχανήν.
ΚΑ. καὶ μὴν ἄγαν γ᾽ Ἕλλην᾽ ἐπίσταμαι φάτιν.
XO. καὶ γὰρ τὰ πυθόκραντα' δυσμαθῆ δ᾽ ὅμως. 1255
ΚΑ. παπαῖ,
οἷον τὸ πῦρ ἐπέρχεται [δέ μοι].
ὀτοτοῖ,
Λύκει᾽ ἼΔπολλον, οἱ ἐγώ [ἐγώ].
αὕτη δίπους λέαινα συγκοιμωμένη
λύκωι λέοντος εὐγενοῦς ἀπουσίαι
F(G)Tr
1229 καὶ κτείνᾶσα FTr: corr. Canter 1231 τοιάδε F: τοιαῦτα Tr τόλμα
Ahrens: τολμᾶ (-μᾷ Tr) ΕἼΤ post τόλμα interpunxit Enger, post φονεύς Elmsley
1232 ἐστιν ΕἼΤ δυσφιλὲς GTr: δυσφιλεὺς T 1235 ἀρὰν FTr: corr. anonymus
apud Blomfieldium et Butler (4pnr), deinde Franz (2p) 1239 ὅμοιον ΕἼΤ
1240 μ᾽ ev Auratus: μὴν ΕἼΤ 1241 γ᾽ del. Bothe οἰκτείρας ΕἼΤ 1242
παιδίων FTr: corr. Schütz 1244 κλύοντ᾽ FTr: aoristum esse vidit Wilamowitz
ἐξεικασμένα Tr 1247 κοίμισον Tr 1249 εἴπερ ἔσται Schütz: εἰ παρέσται FTr
1252 ἄρ᾽ ἂν corruptum : κάρτα τἄρα Hartung παρεσκόπεις (super εἰ scr. η) F:
παρεσκόπης ΤΥ: corr. Hartung 1253 τοῦ γὰρ τελοῦντος FTr: corr. Heimsoeth
1254 ἐπίσταται G 1255 δυσμαθῆ Tr: δυσπαθῆ F 1256 sq. παπαῖ extra
metrum posuit Weil et δέ μοι et alterum ἐγώ delevit, órorot extra metrum posuit
Wilamowitz, admodum probabiliter 1258 δίπλους FTr: corr. Victorius
166
like ἃ treacherous Ate, what work she will bring to pass, with
Evil’s blessing. Such is her daring: the female the slayer of
the male! She is—what hateful monster can I aptly call her ?
an amphisbaena, or some Scylla dwelling in the rocks, the
ruin of sailors, a raging hellish mother and one who breathes
truceless war against her nearest? And how she raised a
shout of triumph, the all-daring one, as though at the turn
of battle; yet pretends to rejoice that he is safe home again!
And if in what I have said there are things you will not
believe, it is all one; how could it be otherwise? What is to
be will come; and thou shalt very soon, a present witness,
call me in pity too true a prophet.
Chor. The feast of Thyestes on his children’s flesh I under-
stand, and shudder, and fear possesses me when it has been
told me in truth and not through images ; but in all else that
I have heard I have lost the track, and am coursing wide
of it.
Cass. Thou shalt look, I say, on Agamemnon's death.
Chor. Peace, unhappy one; lull thy voice to an auspicious
sound.
Cass. Nay, it is no Healer who presides over this that I am
saying.
Chor. Not if it is to be; but may it not chance to happen!
Cass. Thou utterest prayers, but ‘her thought is of
killing.
Chor. Who is the man by whom this woeful deed is being
brought about ?
Cass. Clearly thou hast in very truth lost the track of my
oracles.
Chor. Aye, for I understand not who they are who will
accomplish the design.
Cass. And yet I know the speech of Hellas all too well.
Chor. Aye, so do the Pythian ordinances; but still they
are hard to understand.
Cass. Oh, oh! how fierce is the fire that comes upon me!
Woe, woe! Lycean Apollo, ah me! She, the two-footed
lioness, lying with the wolf in the noble lion’s absence, will
167
 κτενεῖ με τὴν τάλαιναν’ ὡς δὲ φάρμακον 1260
τεύχουσα κἀμοῦ μισθὸν ἐνθήσει ποτῶν
ἐπεύχεται θήγουσα φωτὶ φάσγανον
ἐμῆς ἀγωγῆς ἀντιτείσεσθαι φόνον.
τί δῆτ᾽ ἐμαυτῆς καταγέλωτ᾽ ἔχω τάδε
καὶ σκῆπτρα καὶ μαντεῖα περὶ δέρηι στέφη; 1265
σὲ μὲν πρὸ μοίρας τῆς ἐμῆς διαφθερῶ.
it’ ἐς φθόρον’ πεσόντα γ᾽ ὧδ᾽ ἀμείβομαι.
ἄλλην τιν᾽ ἄτης ἀντ᾽ ἐμοῦ TTAOUTISETE.
ἰδοὺ δ᾽ ᾿Απόλλων αὐτὸς ἐκδύων ἐμὲ
χρηστηρίαν ἐσθῆτ᾽, ἐποπτεύσας δέ με 1270
κἀν τοῖσδε κόσμοις καταγελωμένην μέγα
φίλων ὑπ᾽ ἐχθρῶν οὐ διχορρόπως μάτην’
καλουμένη δέ, φοιτὰς ὡς ἀγύρτρια,
πτωχὸς τάλαινα λιμοθνὴς ἠνεσχόμην"
καὶ νῦν ὁ μάντις μάντιν ἐκπράξας ἐμὲ 1275
ἀπήγαγ᾽ ἐς τοιάσδε θανασίμους τύχας.
βωμοῦ πατρώιου δ᾽ ἀντ᾽ ἐπίξηνον μένει
θερμῶι κοπείσης φοίνιον προσφάγματι.
οὐ μὴν ἄτιμοί γ᾽ ἐκ θεῶν τεθνήξομεν᾽"
ἥξει γὰρ ἡμῶν ἄλλος αὖ τιμάορος, 1280
μητροκτόνον φίτυμα, ποινάτωρ πατρότ᾽
φυγὰς δ᾽ ἀλήτης, τῆσδε γῆς ἀπόξενος,
κάτεισιν ἄτας τάσδε θριγκώσων φίλοις
ἄξει νιν ὑπτίασμα κειμένου πατρός.
τί δῆτ᾽ ἐγὼ κάτοικτος ὧδ᾽ ἀναστένω; 1285
ἐπεὶ τὸ πρῶτον εἶδον Ἰλίου πόλιν
πράξασαν ὡς ἔπραξεν, of δ᾽ εἷλον πόλιν
οὕτως ἀπαλλάσσουσιν ἐν θεῶν κρίσει,
ἰοῦσα ἵπράξωΐ τλήσομαι τὸ κατθανεῖν.
[ὀμώμοται γὰρ SpKos ἐκ θεῶν μέγας.] 1290
F(G)Tr
1261 voci μισθὸν superscr. μνείαν F ἐνθήσει F : ἐνθήσειν (v finali post adscripto)
Tr ποτῶι Auratus: κότῳ (quod nescio an vindicari possit) FTr 1263
ἀντιτίσεσθαι Blomfield: ἀντιτίσασθαι FTr 1265 στέφη ex στέφει corr. Tr
1267 πεσόντα γ᾽ ὧδ᾽ Jacob: meadvr’ ἀγαθὼ δ᾽ FTr; in librorum scriptura duas lectiones
conflatas esse, scil. πεσόντα γ᾽ ὧδ᾽ et πεσόντα θ᾽ ὧδ᾽ (Verrall), perspexit Wilamowitz
ἀμείβομαι in ἀμείψομαι mutatum F : ἀμείψ- GTr I268 ἄτην FTr: corr. Stanley
1270 ἐπόπτευσας V : ἐπώπτευσας Tr 1271 μέγα Hermann: μέτα FTr 1272
de voce μάτην dubitari potest 1274 λιμόθνης FTr: oxytonon esse vidit Elberling
1277 ävremigyvov FTr: distinxit Auratus 1278 φοινίῳ FTr: corr. C. G. Haupt
1279 ἄτιμοί F2Tr: ἄτιμόν FG 1284 ἄξει GTr: ἄξειν F 1285 κάτοικος
FTr: corr. Scaliger 1287 οἵδ᾽ FTr: relativum agnovit Victorius εἷλον Mus-
grave: elyov FTr 1288 ἐν GTr: ἐκ F 1289 πράξω nemo dum explanavit :
κἀγὼ Heath, haud male, sed dubito 1290 hunc versum, quem interpolatum
esse olim suspicatus erat Schiitz, et ille curis secundis et alii varium in modum trans-
posuerunt versus in Epimerism. Hom., Crameri Anecd. Oxon. i. 88. 8, atque in
Etymol. Voss. s.v. dpapev sine poctae nomine allatus, dpape yàp ὅρκος ἐκ θεῶν μέγας,
a Dindorfio huc tractus est, sed incertum est an huc pertineat
168
slay me, wretched that I am, and, like one that prepares
a draught, will add a guerdon for me too! to her potion: she
boasts as she whets the sword against the man that it is
because I have been brought here that she will exact a
penalty of bloody death. Why then in mockery of myself do
I keep these trappings, and this sceptre, and the fillets of
divination round my neck? Thee at least will I destroy
before the hour of my fate. Go to perdition ; now at least, as
you lie on the ground, I thus requite you. Enrich some other
in my stead with curse and doom! And see, Apollo himself
stripping me of my prophetic garb ; and after he had watched
me, even in this attire, mightily mocked by friends who were
foes, and their delusion clear beyond doubt—and, like a
wandering mendicant priestess, I bore being called ‘beggar,
poor wretch, starveling ’_so now the Seer has exacted me,
the seer, as his due, and has led me off to die like this. And
instead of my father’s altar there is a chopping-block await-
ing me, red with my hot blood when I am slaughtered in
sacrifice before the burial.2 Yet our death shall not be
unavenged by the gods; for there shall come another to
avenge us, the offspring that slays his mother, he who exacts
atonement for his father. An exile and a wanderer, a stranger
to this land, he shall return to put for his kin the cope-stone
on these baneful doings: his father’s outstretched body as it
lies on the ground shall bring him back. Why then thus
piteously do I lament? Now that I have seen the city of
Ilion faring as it fared, and they that took the city come off
thus in the decision given by the gods, I will go . . . and bear
1 The Greek seems to be ambiguous between 'wages paid (to Agamemnon) for
(bringing) me' and *wages paid to me'.

* Literally ‘bloody with the hot preliminary sacrifice of me who have been slaughtered'.
169
 “Aidou πύλας δὲ τάσδ᾽ ἐγὼ προσεννέπω:
ἐπεύχομαι δὲ καιρίας πληγῆς τυχεῖν,
ὡς ἀσφάδαιστος, αἱμάτων εὐθνησίμων
ἀπορρυέντων, ὄμμα συμβάλω τόδε.
ΧΟ. ὦ πολλὰ μὲν τάλαινα πολλὰ δ᾽ αὖ σοφὴ 1295
γύναι, μακρὰν ἔτεινας. εἰ δ᾽ ἐτητύμως
μόρον τὸν αὑτῆς οἶσθα, πῶς θεηλάτου
βοὸς δίκην πρὸς βωμὸν εὐτόλμως πατεῖς;
ΚΑ. οὐκ ἔστ᾽ ἄλυξις, οὔ, ξένοι, χρόνωι πλέωτ.
ΧΟ. ὁ δ᾽ ὕστατός γε τοῦ χρόνου πρεσβεύεται. 1300
ΚΑ. ἥκει τόδ᾽ ἦμαρ’ σμικρὰ KepSaved φυγῆι.
ΧΟ. ἀλλ᾽ ἴσθι τλήμων οὖσ᾽ ἀπ᾽ εὐτόλμου φρενός.
ΚΑ. οὐδεὶς ἀκούει ταῦτα τῶν εὐδαιμόνων.
ΧΟ. ἀλλ᾽ εὐκλεῶς τοι κατθανεῖν χάρις Bporói.
ΚΑ. ἰὼ πάτερ σοῦ σῶν τε γενναίων τέκνων. 1305
XO. τί δ᾽ ἐστὶ χρῆμα; τίς σ᾽ ἀποστρέφει φόβος;
ΚΑ. φεῦ φεῦ.
ΧΟ. τί τοῦτ᾽ ἔφευξας; εἴ τι μὴ φρενῶν στύγος.
ΚΑ. φόνον δόμοι πνέουσιν αἱματοσταγῆ.
ΧΟ. καὶ πῶς; τόδ᾽ ὄφει θυμάτων ἐφεστίων. 1310
ΚΑ. ὁμοῖος ἀτμὸς ὥσπερ ἐκ τάφου πρέπει.
ΧΟ. οὐ Σύριον ἀγλάϊσμα δώμασιν λέγεις.
ΚΑ. ἀλλ᾽ εἶμι κἀν δόμοισι κωκύσουσ᾽ ἐμὴν
᾿Αγαμέμνονός τε μοῖραν. ἀρκείτω βίος.
ἰὼ ξένοι. 1315
οὔτοι δυσοίζω, θάμνον ὡς ὄρνις, φόβωι,
ἀλλ᾽ ὡς θανούσηι μαρτυρῆτέ μοι τόδε,
ὅταν γυνὴ γυναικὸς ἀντ᾽ ἐμοῦ θάνηι
ἀνήρ τε δυσδάμαρτος ἀντ᾽ ἀνδρὸς πέσηι.
ἐπιξενοῦμαι ταῦτα δ᾽ ὡς θανουμένη. 1320
ΧΟ. ὦ τλῆμον, οἰκτίρω σε θεσφάτον μόρου.
ΚΑ. ἅπαξ ἔτ᾽ εἰπεῖν ῥῆσιν ἢ θρῆνον θέλω
ἐμὸν τὸν αὐτῆς’ Ἡλίωι δ᾽ ἐπεύχομαι
πρὸς ὕστατον φῶς {τοῖς ἐμοῖς τιμαόροις
ἐχθροῖς φονεῦσι τοῖς ἐμοῖς τίνειν ὁμοῦτ 1325
δούλης θανούσης, εὐμαροῦς XEIPWHATTOS.
ἰὼ βρότεια πράγματ᾽" εὐτυχοῦντα μὲν
F(G)Tr
1291 τάσδ᾽ ἐγὼ Auratus: ras λέγω [Tr 1293 ἀσφάδαστος I'Tr 1295 δ᾽ αὖ
Tr: δὲ F 1296 éxrewas G 1297 αὑτῆς Tr: αὐτῆς F 1299 χρόνῳ πλέω
(πλέῳ Tr) FTr; nondum emendatum 1305 σῶν Auratus: τῶν FTr 1309
φόβον FTr (v supra P scripsit Tr) I3II ὅμοιος FTr 1317 μαρτυρεῖτέ FTr:
corr. Orelli 132I οἰκτείρω FTr 1322 ἔτ᾽ FTr: ἔστ᾽ G 1324 sq. in ea
sententiae parte quam crucibus saepsi graves corruptelac latent, quibus quot verba
affecta sint accuratius definiri nequit 1324 τοῖς ex τοὺς corr. F
170
my death. And this door I address as the gate of Hades ; and
my prayer is to meet with a mortal stroke so that I may close
these eyes without a struggle, the blood streaming forth in
easy death.
Chor. O woman much to be pitied, but of much wisdom
too, thou hast spoken at great length. But if thou knowest
thine own death in very truth, how is it that, like a cow
driven by a god, thou goest fearlessly to the altar?
Cass. There is no escape, there is none, strangers. . . .
Chor. Yes, but the last of one’s time is valued most.
Cass. The day is come: little shall I gain by flight.
Chor. Well, be assured that thou art an enduring sufferer
with a brave soul.
Cass. None of those who prosper is ever called that.
Chor. But to win fame in dying is a boon for man.
Cass. O alas, father, for thee and thy noble children!
As she ts about to approach the door, she shrinks back
Chor. What is it now? what terror turns thee back ?
Cass. Faugh! [fancy.
Chor. Why this ‘faugh’? unless it be some horror in thy
Cass. The house breathes murder, with dripping blood.
Chor. Nay, nay: this smell is that of sacrifices at the
hearth.
Cass. There is the same reek—unmistakable—as from
a grave.
Chor. Thou speakest not of the Syrian incense that gives
splendour to the house.
Cass. Well, I will go, to bewail within the house as well
my own and Agamemnon’s fate. Let life suffice.
She stops again
Oh, oh, strangers—I do not raise this woeful cry like the
bird that shuns the bush, from fear, but that you may bear
me witness of this when I am dead, at the time when for me,
the woman, a woman dies, and for the ill-mated man there
falls a man. 1 ask this favour (9) of you as one about to die.
Chor. Poor maid! I pity thee for thy foretold death.
Cass. I would make one speech more, or it may be a dirge
—my own. To the Sun I pray, turning towards his light, the
last, that the guilty ones may requite my avengers not only for
the king's murder but also! for the slaying of me, the slave, an
easy overthrow.
Alas for the state of man: when there is success, a shadow
1 The words printed in italics are only meant to give the general sense; the details
are quite uncertain. See the commentary.
171
 σκιά τις ἂν Tp&yetev: εἰ δὲ δυστυχῆι,
βολαῖς ὑγρώσσων σπόγγος ὥλεσεν γραφήν.
καὶ ταῦτ᾽ ἐκείνων μᾶλλον οἰκτίρω πολύ. 1330

ΧΟ. τὸ μὲν εὖ πράσσειν ἀκόρεστον ἔφυ

πᾶσι βροτοῖσιν: δακτυλοδείκτων δ᾽
οὔτις ἀπειττὼν εἴργει μελάθρων,
“ μηκέτ᾽ ἐσέλθηις ᾽, τάδε φωνῶν.
καὶ τῶιδε πόλιν μὲν ἑλεῖν ἔδοσαν 1335
μάκαρες Πριάμου,
θεοτίμητος δ᾽ οἴκαδ᾽ ἱκάνει"
νῦν δ᾽ εἰ προτέρων aly’ ἀποτείσαι
καὶ τοῖσι θανοῦσι θανὼν ἰἄλλωντ
ποινὰς θανάτων ἐπικράναι, 1340
τίς ἂν (ἐξ)γεύξαιτο βροτῶν ἀσινεῖ
δαίμονι φῦναι τάδ᾽ ἀκούων;

ΑΓ. ὦμοι, πέπληγμαι καιρίαν πληγὴν ἔσω.

ΧΟ. σῖγα' τίς πληγὴν Abtei καιρίως οὐτασμένος;
ΑΓ. ὦμοι μάλ᾽ αὖθις, δευτέραν πεπληγμένος. 1345
ΧΟ. τοὔργον εἰργάσθαι δοκεῖ μοι βασιλέως οἰμώγμασιν.
ἀλλὰ κοινωσώμεθ᾽ ἤν πως ἀσφαλῇ βουλεύματ᾽ (ἦι).
- ἐγὼ μὲν ὑμῖν τὴν ἐμὴν γνώμην λέγω,
πρὸς δῶμα δεῦρ᾽ ἀστοῖσι κηρύσσειν βοήν.
-- ἐμοὶ δ᾽ ὅπως τἀχιστά γ᾽ ἐμπεσεῖν δοκεῖ 1350
καὶ πρᾶγμ᾽ ἐλέγχειν σὺν νεορρύτωι ξίφει.
--- κἀγὼ τοιούτου γνώματος κοινωνὸς ὧν
wnoizopal τι δρᾶν τὸ μὴ μέλλειν δ᾽ ἀκμή.
--- ὁρᾶν πάρεστι: φροιμιάφονται γὰρ ὡς
τυραννίδος σημεῖα πράσσοντες πόλει. 1355
-- ypovígouev γάρ’ οἱ δὲ τῆς μελλοῦς κλέος
F(G)Tr
1328 ἀντρέψειεν FTr: distinxit Porson δυστυχῆ FTr: vel δυστυχῆι vel δυστυχῆ
sententiae satisfaciat 1329 βουλαῖς G ὥὦλεσε ΕἾΤ: v add. Porson
1330 οἰκτείρω FTr 1331 πράττειν FTr: corr. Porson 1332 βροτοῖς FTr:
corr. Pauw δακτυλοδεικτῶν FTr cum schol. vet.: adiectivum agnovit Casaubon
(-«rov), genet. plur. Schütz 1334 μηκέτι 8 εἰσέλθης FTr: corr. Hermann 1338
ἀποτίσει FTr, quod quoniam cum sequentibus non convenit, aut ἀποτείσαι (Keck)
scribendum esse videtur aut ἀποτείσηι (Sidgwick) 1339 ἄλλων non intellego
1340 θανάτων ἄγαν ἐπικρ- Tr (cf. ad 356) ἐπικράναι Keck: ἐπικρανεῖ FTr: ἐπικράνηι
Sidgwick 1341 ἐξεύξαιτο Schneidewin : εὔξαιτο FTr: ris ποτ᾽ ἂν εὔξ- E. A. J. Ahrens
1346 εἰργᾶσθαι ΕἼΤ 1347ἄν πως FTr: corr. Paley βουλεύματα. FTr: ἦι
suppl. Enger 1348-70 singulis dehinc distichis χορός praefixerunt FTr; duo-
decim choreutas verba facere viderunt Bamberger et O. Müller, cf. comm. 1348
ὑμῖν ex ἡμῖν corr. F 1356 Blomfield intellexit huc pertinere Tryphonis qui
dicitur περὶ τρόπων, Rhet. Graec. viii. 741. 9 Walz, κατὰ παρονομασίαν . . . οὕτως ὠνόμασται
καὶ παρ᾽ Αἰσχύλωι μελλώ" * χρονίζομεν ὧδε τῆς μελλοῦς χάριν ', unde τῆς μελλοῦς restituit
Blomfield: χρονίζομεν yàp: οἱ δὲ, τῆς (τῆς om. Tr) μελλούσης κλέος FTr
172
may turn it ; but when there is misfortune, a wet sponge with
one dash blots out the picture. And this I pity far more than
that. Cassandra goes within
Chor. Of prosperity mankind never has its fill; and no one
bars it from a hall whereat men's fingers point and turns it
away with these words, ' Enter no more'. So here is a man to
whom the blessed ones have given it to take Priam's town,
and he has come home honoured of heaven ; but if he should
pay for the blood of those slain aforetime, and give to the
dead, by his own death, a full recompense for . . .! deaths,
what man could boast himself born with fortune beyond
harm, on hearing this?

Ag. (from within). Oh! I am smitten deep with a mortal

blow.
Chor. Silence! who raises a cry of a blow, wounded with
a mortal wound ?
Ag. Oh! yet again; I am smitten a second time.
Chor. Methinks the deed is done, from the king's cries of
pain. But let us take counsel together if perhaps there is
some safe plan.
The twelve Elders in succession
I. 1 for my part tell you what I propose: to let criers
sound a summons to the citizens to bring succour hither to
the house.
2. And I think we had best rush in at once, and prove the
deed while the newly drawn sword is there.
3. I too join in a proposal such as this and vote for doing
something ; this is the moment to act without delay.
4. It is clear: their prelude is that of people whose doings
betoken tyranny against the city.
5. Yes, for we are dallying ; but they tread under foot the
τ Text uncertain,

173
 πέδον πατοῦντες οὐ καθεύδουσιν χερί.
— οὐκ οἶδα βουλῆς ἧστινος τυχὼν λέγω:
τοῦ δρῶντός ἐστι καὶ τὸ βουλεῦσαι [mépif.
--- κἀγὼ τοιοῦτός εἰμ᾽, ἐπεὶ δυσμηχανῶ 1360
λόγοισι τὸν θανόντ᾽ ἀνιστάναι πάλιν.
-- ἢ καὶ βίον τείνοντες ὧδ᾽ ὑπείξομεν
δόμων καταισχυντῆρσι τοῖσδ᾽ ἡγουμένοις;
- ἀλλ᾽ οὐκ ἀνεκτόν, ἀλλὰ κατθανεῖν κρατεῖ"
πειταιτέρα γὰρ μοῖρα τῆς τυραννίδος. 1365
— ἢ γὰρ τεκμηρίοισιν ἐξ οἰμωγμάτων
μαντευσόμεσθα τἀνδρὸς ὡς ὀλωλότος;
— σάφ᾽ εἰδότας χρὴ τῶνδε μυθεῖσθαι πέρι"
τὸ γὰρ τοπάτειν τοῦ σάφ᾽ εἰδέναι δίχα.
-- ταύτην ἐπαινεῖν πάντοθεν πληθύνομαι, 1370
τρανῶς ᾿Ατρείδην εἰδέναι κυροῦνθ᾽ ὅπως.

KAYTAIMHZTPA
πολλῶν πάροιθεν καιρίως εἰρημένων
τἀναντί᾽ εἰπεῖν οὐκ ἐπαισχυνθήσομαι.
Trés γάρ τις ἐχθροῖς ἐχθρὰ πορσύνων, φίλοις
δοκοῦσιν εἶναι, πημονῆς ἀρκύστατ᾽ av 1375
φράξειεν ὕψος κρεῖσσον ἐκπηδήματος;
ἐμοὶ δ᾽ ἀγὼν ὅδ᾽ οὐκ ἀφρόντιστος πάλαι
ἱνίκης παλαιᾶς ἦλθε, σὺν χρόνωι γε uv
ἕστηκα δ᾽ ἔνθ᾽ ἔπαισ᾽ ἐπ᾽ ἐξειργασμένοις.
οὕτω δ᾽ ἔπραξα, καὶ τάδ᾽ οὐκ ἀρνήσομαι, 1380
ὡς μήτε φεύγειν μήτ᾽ ἀμύνεσθαι μόρον.
ἄπειρον ἀμφίβληστρον, ὥσπερ ἰχθύων,
περιστιχίφω, πλοῦτον εἵματος koakóv:
tate δέ νιν δίς’ κἀν δυοῖν οἰμώγμασιν
μεθῆκεν αὐτοῦ κῶλα’ καὶ πεπτωκότι 1385
τρίτην ἐπενδίδωμι, τοῦ κατὰ χθονὸς
Διὸς νεκρῶν σωτῆρος εὐκταίαν χάριν.
οὕτω τὸν αὑτοῦ θυμὸν ὀρυγάνει πεσών"
κἀκφυσιῶν ὀξεῖαν αἵματος fopayñvi
Ἕ(Τε
1357 καϑεύδουσιν (ΤΥ: -σι F 1359 πέρι FTr: τί μή; (an potius τί μήν;Ὁ)
Wilamowitz 1362 «reivovres FTr: corr. Canter 1364 κρατεῖ Casaubcr :
κράτει FTr 1367 μαντευσόμεσθα GTr: -σόμεϑα F 1368 μυθοῦσθαι FTr: corr.
J. G. Schneider 1375 πημονὴν FTr: corr. Auratus ἀρκύστατ᾽ ἂν Elmsley :
ἀρκύστατον FTr 1376 φρέξειεν G 1378 νίκης FTr: veixns Heath, fortasse
recte: δίκης Pauw 1379 ἔπαισ᾽ GTr: ämeo’ F ἐξειργασμένοις, priore o eX p
facto, F 1381 ἀμύνασθαι FTr: corr. Victorius 1383 περιστιχίζω Tr:
περιστιχίζων (-στοιχ-ΕἾ FG 1384 οἰμωγμάτοιν Elmsley 1387 Διὸς Enger:
ἄδου FTr 1388 αὑτοῦ Schütz : αὐτοῦ FTr ópvyáre Hermann: ὁρμαίνει FTr
1389 σφαγὴν corruptum; nescio an fuerit ῥαγὴν

174
honoured name of Cautious Delay, and their hands are not
asleep.
6. Iknow not what plan I can find to put forward: it is to
the doer that the planning also belongs.
7. I am of like mind, for I see no way to raise the dead to
life again with words.
8. Say: shall we, trying to drag out our lives, bow down
thus beneath the rule of these defilers of the house?
9. Nay, it is not to be endured ; to die is better, for that
is a milder lot than tyranny.
10. What? shall we by inference from cries divine that he
has perished ?
rir. We should have sure knowledge before we talk of this
matter, for guessing is very different from sure knowledge.
12. What as the outcome of all my thoughts prevails in
me is to commend this opinion, that we should know with
certainty how it is with the son of Atreus.

While they are still standing about in indecision, the

doors are thrown open. Clylemnestra 1s seen standing
by the side of a bath-tub, in which the body of
Agamemnon lies, covered with a huge embroidered
garment; near by is the body of Cassandra
Clyt. Much have I said before to suit the moment: in
saying the contrary now I shall feel no shame. For how else
could one who has in hand acts of enmity against enemies
who pass as friends fence the nets of harm to a height past
overleaping ? For me this contest has come as one that I have
long had in mind, as one that brings to the issue a long-
standing feud,! but come it has at last; and I stand where
I struck, with the deed done. And so I achieved it—I will
not deny it—that he could neither escape nor ward off his
doom. An endless net, like a fish-net, I throw around him,
an evil wealth of dress, and I strike him twice ; and with two
cries there on the spot he let his limbs go slack ; and then,
when he is down, I add a third stroke, a welcome prayer-
offering to the Zeus beneath the earth, the saviour of the
dead. So he belches out his own life as he lies there, and
blowing forth a sharp . . . of blood, he strikes me with a
1 Or ‘claim’? See the commentary.

175
 βάλλει μ᾽ ἐρεμνῆι ψακάδι φοινίας δρόσου, 1390
χαίρουσαν οὐδὲν ἧσσον ἢ διοσδότωι
γάνει σπορητὸς κάλυκος ἐν λοχεύμασιν.
ὡς ὧδ᾽ ἐχόντων, πρέσβος ᾿Αργείων τόδε,
χαίροιτ᾽ ἄν, εἰ χαίροιτ᾽, ἐγὼ δ᾽ ἐπεύχομαι.
εἰ δ᾽ ἦν πρεπόντως ὥστ᾽ ἐπισπένδειν νεκρῶι, 1395
τάδ᾽ ἂν δικαίως ἦν, ὑπερδίκως μὲν οὖν’
τοσῶνδε κρατῆρ᾽ ἐν δόμοις κακῶν ὅδε
πλήσας ἀραίων αὐτὸς ἐκπίνει μολών.
ΧΟ. θαυμάφομέν σου γλῶσσαν, ὡς θρασύστομος,
ἥτις τοιόνδ᾽ ἐπ᾽ ἀνδρὶ κομπάξεις λόγον. 1400
ΚΛ. πειρᾶσθέ μον γυναικὸς ὡς ἀφράσμονος᾽
ἐγὼ δ᾽ ἀτρέστωι καρδίαι πρὸς εἰδότας
λέγω: σὺ δ᾽ αἰνεῖν εἴτε με ψέγειν θέλεις,
ópoiov: οὗτός ἐστιν ᾿Αγαμέμνων, ἐμὸς
πόσις, νεκρὸς δὲ τῆσδε δεξιᾶς χερὸς 1405
ἔργον, δικαίας τέκτονος. τάδ᾽ ὧδ᾽ ἔχει.
x

ΧΟ. τί κακόν, ὦ γύναι,

χϑονοτρεφὲς ἐδανὸν ἢ ποτὸν
πασαμένα ῥυτᾶς ἐξ ἁλὸς ὀρόμενον
τόδ᾽ ἐπέθου θύος δημοθρόους τ᾽ ἀράς;
ἀπέδικες ἀπέταμες: ἀπόπολις δ᾽ ἔσηι 1410
μῖσος ὄβριμον ἀστοῖς. —

ΚΛ. νῦν μὲν δικάφεις ἐκ πόλεως φυγὴν ἐμοὶ

καὶ μῖσος ἀστῶν δημόθρους τ᾽ ἔχειν ἀράς,
οὐδὲν τότ᾽ ἀνδρὶ τῶιδ᾽ ἐναντίον φέρων,
ὃς οὐ προτιμῶν, ὡσπερεὶ βοτοῦ μόρον 1415
μήλων φλεόντων εὐπόκοις νομεύμασιν,
ἔθυσεν αὑτοῦ παῖδα, φιλτάτην ἐμοὶ
ὠδῖν᾽, ἐπωιδὸν Θρηικίων ἀημάτων.
οὐ τοῦτον ἐκ γῆς τῆσδε χρῆν σ᾽ ἀνδρηλατεῖν
μιασμάτων ἄποιν᾽; ἐπήκοος δ᾽ ἐμῶν 1420
ἔργων δικαστὴς τραχὺς εἴ. λέγω δέ σοι
F(G)Tr
1391 sq. διὸς νότω | γᾶν εἰ FTr: corr. Porson 1392 σπόρητος FTr 1395
πρεπόντων FTr: corr. Is. Vossius 1397 τοσόνδε Blomfield 1401 μου om. G
1404 ὅμοιον FTr 1406 δικαίως Tr 1407 ἐδανὸν Tr: ἐδανὸν F 1408 ῥύσας
F: ῥυσᾶς Tr: corr. Stanley ópópevov Canter : ὁρώμενον (ὀρ- Τῇ) ΕἼΤ 1409 τόδ᾽
ἐπέθου (ἐπεύθου Tr) θύος non intellego 1410 ἀπέτᾶμες F : ἀπέταμες G: ἀπέτεμες
Tr; cf. 1153 ἄπολις FTr: corr. Casaubon 1411 ὄμβριμον FTr: corr. Porson
1414 οὐδὲν corr. ex οὐ σὺν F τότ᾽ Is. Vossius: τόδ᾽ ΕἼΤ 1417 αὐτοῦ Victorius:
αὐτοῦ FTr 1418 ἐπωδὴν G ἀημάτων Canter: τὲ (τε Tr) λημμάτων ΕἼΤ 1419
χρή F Tr: corr. Porson 1420 βιασμάτων G
176
darksome shower of gory dew; and I rejoiced no less than the
crop rejoices in the rich blessing of the rain of Zeus when
the sheath is in labour with the ear.
So stands the case, noble elders of Argos here: be glad, if
ye will be glad ; for me, I glory in it. And were it possible to
pour libations over the dead body in a manner that would
suit the circumstances, this (my doing) would be just, yea,
more than just: so many are the curseful evils wherewith this
man in his house has filled a bowl, a bowl which he now
drains himself on his return.
Chor. We marvel at thy tongue, at the boldness of its
speech, to utter such boasting words over thine own hus-
band.
Clyt. Ye make trial of me as if I were a woman of no sense;
but I speak with fearless heart to you who know it well, and
whether thou choosest to praise or blame me, it is all one:
this is Agamemnon, my husband, made a corpse by the work
of this right hand, a righteous craftsman. So stands the case.

Chor. Woman, what harmful thing hast thou tasted, a

thing thou hast eaten that the earth nourished, or a drink
sprung from the flowing sea, that thou hast put upon thy-
self (9) this... and the loud curses of the folk? Thou hast
cast away, thou hast shorn away, and away from the city
thou shalt be banished, the people’s heavy hatred on thee.

Clyt. Now, upon me, thou dost pass judgement of banish-

ment from the city, and the people’s hatred, and the bearing
of the loud curses of the folk ; though then thou didst not set
thyself at all against this man here, who, holding it in no
special account, as though it were the death of a beast where
sheep abound in the fleecy flocks, sacrificed his own child, the
dearest fruit of my travail, to charm away the winds of
Thrace. Is it not he whom thou shouldst have banished from
this land as a requital for the unclean thing he did ? but when
thou takest cognizance of my acts thou art a stern judge.
But I tell thee this: thou must utter such threats as these in
4872.1 177 Ν
 τοιαῦτ᾽ ἀπειλεῖν, ὡς παρεσκευασμένης
ἐκ τῶν ὁμοίων, χειρὶ νικήσαντ᾽ ἐμοῦ
ἄρχειν: ἐὰν δὲ τοὔμπαλιν κραίνηι θεός,
γνώσηι διδαχθεὶς ὀψὲ γοῦν τὸ σωφρονεῖν. 1425

ΧΟ. μεγαλόμητις εἶ,

περίφρονα 5° ἔλακες, ὥσπερ οὖν
φονολιβεῖ τύχαι φρὴν ἐπιμαίνεται᾽
λίπος ἐπ᾽ ὀμμάτων αἵματος εὖ πρέπει.
ἄντιτον ἔτι σε χρὴ στερομέναν φίλων
τύμμα τύμματι τεῖσαι. <= 1430

KA. καὶ τήνδ᾽ ἀκούεις ὁρκίων ἐμῶν θέμιν’

μὰ τὴν τέλειον τῆς ἐμῆς παιδὸς Δίκην
"Arnv Ἐρινύν θ᾽, αἷσι τόνδ᾽ ἔσφαξ᾽ ἐγώ,
οὔ uot Φόβου μέλαθρον ᾿Ελπὶς ἐμπατεῖ,
ἕως ἂν αἴϑηι πῦρ ἐφ᾽ ἑστίας ἐμῆς 1435
Αἴγισθος, ὡς τὸ πρόσθεν εὖ φρονῶν ἐμοί:
οὗτος γὰρ ἡμῖν ἀσπὶς οὐ σμικρὰ θράσους.
κεῖται γυναικὸς τῆσδε λυμαντήριος,
Χρυσηΐδων μείλιγμα τῶν ὑπ᾽ Ἰλίωι,
fj τ᾽ αἰχμάλωτος ἦδε καὶ τερασκότπος, 1440
fj κοινόλεκτρος τοῦδε θεσφατηλόγος,
πιστὴ EUveuvos, ναυτίλων δὲ σελμάτων
ἱστοτρίβης. ἄτιμα δ᾽ οὐκ ἐπραξάτην᾽
ὁ μὲν γὰρ οὕτως, ἡ δέ τοι κύκνου δίκην
τὸν ὕστατον μέλψασα θανάσιμον γόον 1445
κεῖται φιλήτωρ τοῦδ᾽, ἐμοὶ δ᾽ ἐπήγαγεν
εὐνῆς! παροψώνημα τῆς ἐμῆς χλιδῆς.
ΧΟ. φεῦ, τίς ἂν ἐν τάχει μὴ περιώδυνος
μηδὲ δεμνιοτήρης
μόλοι τὸν αἰεὶ φέρουσ᾽ ἂν ἡμῖν 1450
F(G)Tr
1422 sqq. huius sententiae structura nondum expedita est ; aut post ὁμοίων aliquid
excidisse suspicor aut in ultimis verbis (xeipi . . . ἄρχειν) graviorem latere corruptelam
1428 λῖπος FTr (accentum corr. Abresch) : MBos Casaubon, fort. recte 1428 sq.
πρέπειαν rierov F : πρέπει ἀτίετον Tr: ἄντιτον agnovit Weil 1430 τύμμα τύμμα ΕἼΤ:
corr. Casaubon τίσαι FTr 1433 ἐριννύν FTr 1435 ἐμῆς Porson (ἐμᾶς
Scaliger): ἐμὰς ΕἼΤ 1437 μικρὰ FTr: corr. Blomfield 1441 ἡ Karsten: καὶ
FTr 1443 ἱστοτρίβης quomodo intellegi possit nescio, corruptum esse affir-
mare non ausim; vulgo substituitur ἐσοτριβής (Pauw), quae vox nihili est 1444
sq. xóxvov . . . γόον affert Tzetzes ad Lycophr. 426 (p. 157. 14 Scheer) 1446 φιλήτωρ
GTr: φιλήτως F 1447 εὐνῆς corruptum esse viderunt Casaubonaliique παροψόνημα
FTr: corr. Casaubon τῆς ἐμῆς χλιδῆς non temptandum 1450 ἂν Emperius:
ἐν FTr
178
the knowledge that I am prepared no less than thou . . .
(that) he who has conquered me by force shall rule me; but
if the god ordains the contrary, thou shalt be taught a lesson
and learn—though late—discretion.

Chor. High and daring is thy thought, and haughty thy

ringing words, as in truth thy mind is maddened by the
blood-dripping thing that has come to pass: the blood-fleck
in thine eyes is clear to see. As a requital thou must yet,
bereft of friends, pay for blow with blow.

Clyt. This also thou hearest, this the solemn power of my

oath: by Justice accomplished for my child, by Ate and
Erinys, to whom I slew this man, for me it is not within the
house of Fear that Expectancy sets foot, so long as the fire
upon my hearth is lighted by Aegisthus, loyal to me as
heretofore ; for in him we have no slight shield of confidence.
Low lies the man who did outrage to this woman, the
charmer of each Chryseis before Ilion, and she too, this cap-
tive here and auguress, the prophesying bedfellow of this
raan, a faithful concubine, and . . . of (?) the benches of the
ship. They have fared after their deserts: he (died) as I have
told, and she like a swan, after singing her last lament of
death, lies here, his lover, and for me she has brought .. . an
added relish to my feast of luxury.

Chor. Alas! would that swiftly, without agony and without

long lying abed, some fate might come bringing us the
4872.1 179 N2
 μοῖρ᾽ ἀτέλευτον ὕπνον, δαμέντος
φύλακος εὐμενεστάτον [καὶ]
πολέα τλάντος γυναικὸς Era:
πρὸς γυναικὸς δ᾽ ἀπέφθισεν βίον. —
ἰὼ
παράνους Ἑλένα, 1455
μία τὰς πολλάς, τὰς πάνυ πολλὰς
ψυχὰς ὀλέσασ᾽ ὑπὸ Τροίαι,
νῦν [δὲ] τελέαν πολύμναστον ἐπηνθίσω,
[δι αἷμ᾽ ἄνιτττον. A τις ἦν τότ᾽ ἐν δόμοις 1460
Ἔρις, ἐρίδματος, ἀνδρὸς olgus. —c

ΚΛ, μηδὲν θανάτου μοῖραν ἐπεύχου

τοῖσδε βαρυνθείς"
μηδ᾽ εἷς Ἑλένην κότον ἐκτρέψηις,
ὡς ἀνδρολέτειρ᾽, ὡς μία πολλῶν 1465
ἀνδρῶν ψυχὰς Δαναῶν ὀλέσασ᾽
ἀξύστατον ἄλγος ἔπραξεν.

ΧΟ. δαῖμον, ὃς ἐμπίττεις δώμασι καὶ διφυί-

οἷσι Τανταλίδαισιν,
κράτος {τ᾽) ἰσόψυχον ἐκ γυναικῶν 1470
καρδιόδηκτον ἐμοὶ κρατύνεις᾽
ἐπὶ δὲ σώματος δίκαν [por]
κόρακος ἐχθροῦ σταθεὶς ἐκνόμως
ὕμνον ὑμνεῖν ἐπεύχεται *. =

KA. νῦν δ᾽ ὥρθωσας στόματος γνώμην, 1475

τὸν τριπάχυντον
δαίμονα γέννης τῆσδε κικλήσκων᾽"
ἐκ τοῦ γὰρ ἔρως αἱματολοιχὸς
νείραι τρέφεται, πρὶν καταλῆξαι
τὸ παλαιὸν ἄχος, νέος ἰχώρ. | 1480
F(G)Tr
1451 ὕπνον ex ὕμνον corr. F 1452 καὶ del. Franz 1453 πολέα C. G. Haupt:
πολλὰ FTr διαὶ F: διὰ Tr 1455 rapdvous Hermann: παρανόμους FTr 1457
ὀλέσασ᾽ Tr: ὀλέσας F 1458 δὲ del. Wilamowitz τελέαν Wilamowitz : τελείαν FTr
1460 δι᾽ del. Wellauer ἢ τις Schütz : ἥτις FTr 1461 ὀϊζύς FTr 1464 μὴ δ᾽
FTr ἐκτρέψης GTr: ἐκτρέχης F 1466 ὀλέσασ' Tr: ὁλέσαν F 1468 ἐμπίτνεις
Canter 1468 sq. διφυεῖσι FTr: corr. Hermann 1469 τανταλίδαισιν Tr:
«λίδεσιν F 1470 τ᾽ suppl. Hermann (δ᾽ Casaubon) ἰσόψυχον licet perplexe
dictum sit non corruptum esse videtur 1471 καρδίᾳ ön«röv FTr: corr. Abresch
1472 μοι del. Dindorf 1473 ἐκνόμως schol. vet. in Tr: ἐκνόμως ex ἐννόμως corr.
Tr: ἐννόμως F 1474 in fine v — deesse viderunt Pauw, Schütz, alii 1475
δ᾽ del. Headlam, nescio an recte 1476 τριπάχυιον ΕἼΤ: corr. Bamberger 1477
γέννης, in fine as superscripto, FG: γέννας Tr 1479 νείρει FTr: corr. Casaubon
(qui νείρῃ scripsit, unde νείρᾳ Wellauer) 1480 ἴχαρ Headlam
180
eternal sleep unending, now that our kindest protector is
laid low, after enduring much through a woman’s doing ; and
at a woman’s hand he has lost his life.

Oh, thou mad Helen, that didst alone destroy those many,
all those many lives, under the walls of Troy! now thou hast
crowned thyself with the last and perfect garland unforget-
table, blood not to be washed away. In truth there was an
Eris then in the house, strong-built (?), a husband’s bane.

Clyt. Nay, pray not in grief at this for death to be thy

fate; nor turn thine anger upon Helen, saying that she was a
man-destroyer, that she alone destroyed the lives of many
Greek warriors and wrought a woe that none might stand
against.

Chor. O daemon, that fallest upon the house and the two
Tantalidae, and holdest a sway . . .! that works from women,
and my heart is sore for it; he has settled, like a hateful
raven, on the body, and glories in singing tunelessly a
song. ...

Clyt. Now thou hast mended the opinion thou didst utter,
by invoking the thrice-gorged daemon of this race; for it is
by his doing that a craving for blood to lap is nourished in
the belly, new pus before the old woe has ceased.

1 I cannot understand this passage.

181
 XO. 1 μέγαν οἰκοσινῆ
δαίμονα καὶ βαρύμηνιν αἰνεῖς,
φεῦ φεῦ, κακὸν αἶνον,
ἀτηρᾶς τύχας ἀκόρεστον᾽
ἰὼ ἰή, διαὶ Διὸς 1485
Travaitiou πανεργέτα᾽
τί γὰρ βροτοῖς ἄνευ Διὸς τελεῖται;
τί τῶνδ᾽ οὐ θεόκραντόν ἐστιν;

ἰὼ ἰὼ βασιλεῦ βασιλεῦ,
πῶς σε δακρύσω; 1490
φρενὸς ἐκ φιλίας τί ποτ᾽ εἴπω;
κεῖσαι δ᾽ ἀράχνης ἐν ὑφάσματι τῶιδ᾽
ἀσεβεῖ θανάτωι βίον ἐκπνέων,
ὦμοι μοι
Koltav τάνδ᾽ ἀνελεύθερον,
δολίωι μόρωι δαμεὶς (δάμαρτος) 1405
ἐκ χερὸς ἀμφιτόμωι βελέμνωι. ---

KA. αὐχεῖς εἶναι τόδε τοὔργον ἐμόν'

μηδ᾽ ἐπιλεχθῆιςϊ
᾿Αγαμεμνονίαν εἶναί μ᾽ ἄλοχον.
φανταζόμενος δὲ γυναικὶ νεκροῦ 1500
τοῦδ᾽ ὁ παλαιὸς δριμὺς ἀλάστωρ
᾿Ατρέως χαλεποῦ θοινατῆρος
τόνδ᾽ ἀπέτεισεν
τέλεον veapois ἐττιθύσας.

XO. as μὲν ἀναίτιος εἶ 1505

τοῦδε φόνου, τίς ὁ μαρτυρήσων;
πῶ, πῶ; πατρόθεν δὲ
συλλήπτωρ γένοιτ᾽ ἂν ἀλάστωρ.
βιάζεται δ᾽ ὁμοσπόροις
ἐπιρροαῖσιν αἱμάτων 1510
F(G)Tr
1481 οἰκοσινῆ Wilamowitz : οἴκοις τοῖσδε FTr: οἰκονόμον Schneidewin 1484 ἀκορέ-
orov FTr: corr. Todt 1486 πανεργέτα Tr: mavepyerav FG (in F -yér- ex -yar- corr.;
cf. Christ. pat. 1457 ὦ παῖ παναιτίον θεοῦ mavrepydra) 1489 nulla nota in FG;
ἡμιχόριον praefixit Tr, scilicet ‘in exemplo paragraphus ephymnium distinguebat '
(Wilamowitz) ; cf. 1513 ἰὼ semel F 1491 ποτ᾽ dp’ εἴπω Tr; cf. ad 356 1493
ἐκπνείων Tr; cf. ad 356 1494 rävö' ἀνελεύθερα Tr 1495 in fine δάμαρτος
(suppl. Enger) vel aliquid simile excidisse si minus certum est at probabile videtur
1497 τοὔργον ἐμὸν τόδε; Tr 1498 μὴ δ' FTr ἐπιλεχθῆις, quidquid 510] viri
docti de hoc verbo finxerunt, perobscurum est; atque ne de voce μηδ᾽ quidem satis
constat, nisi quid forte ante hunc versum excidisse putes 1503 ἀπέτισεν FTr
1504 νεκροῖς Tr 1507 δὲ om. Tr
182
 Chor. Truly a mighty house-wasting daemon, one of

grievous wrath, is he whom thou praisest—alas, alas, an evil

praise—insatiate of baneful success ; woe, woe, through the

will of Zeus, the cause of all, the doer of all, for what is

fulfilled for men without Zeus? what is there herein that is

not ordained by the god ?

O my king, my king, how shall I weep for thee? From my

loving mind what shall I say ? Thou hast been brought to lie
in this woven spider-web, breathing out thy life by an
impious death, ah me, to lie thus ignobly here, subdued in
a guileful doom at thy wife's hand with a two-edged weapon.

Clyt. Thou art confident that this deed is mine, . . . that

I am Agamemnon's wife. But appearing in the shape of this
dead man's wife, the ancient fierce spirit that takes ven-
geance for the misdeed of the cruel feaster Atreus has now
rendered this full-grown man as payment to the young, a
crowning sacrifice.

Chor. That thou art guiltless of this murder, who shall

bear witness? how can it be, how? but an avenging spirit
sprung from a father's crime might lend a helping hand.
Violently, in fresh streams of kindred blood that swell the
183
 μέλας "Apre: ὅποι δὲ καὶ προβαίνων,
πάχναν κουροβόρον παρέξει.

ἰὼ ἰὼ βασιλεῦ βασιλεῦ,
πῶς σε δακρύσω;
φρενὸς ἐκ φιλίας τί ποτ᾽ εἴπω; 1515
κεῖσαι δ᾽ ἀράχνης ἐν ὑφάσματι τῶιδ᾽
ἀσεβεῖ θανάτωι βίον ἐκπνέων,
ὦμοι μοι
κοίταν τάνδ᾽ ἀνελεύθερον,
δολίωι μόρωι δαμεὶς (δάμαρτος)
ἐκ χερὸς ἀμφιτόμωι βελέμνωι. = 1520
ΚΛ, οὔτ᾽ ἀνελεύθερον οἶμαι θάνατον
τῶιδε γενέσθαι, * * * o x
* * 0k 0k 0k ko K KK καὶ

οὐδὲ γὰρ οὗτος δολίαν ἄτην

οἴκοισιν ἔθηκ᾽,
ἀλλ᾽ ἐμὸν ἐκ τοῦδ᾽ ἔρνος ἀερθέν, 1525
τὴν πολύκλαυτον {τ᾿ ᾿ἸΙφιγένειαντ
* * * 0X * K€ * 0 K K K

ἄξια δράσας, ἄξια πάσχων

μηδὲν ἐν “AiSou μεγαλαυχείτω,
ξιφοδηλήτωι
θανάτωι τείσας ἅπερ ἔρξεν.
ΧΟ. ἀμηχανῶ φροντίδος στερηθεὶς 1530
εὐπαλάμων μεριμνᾶν
ὅπαι τράπωμαι, πίτνοντος οἴκου.
δέδοικα δ᾽ ὄμβρον κτύπον δομοσφαλῆ
τὸν αἱματηρόν' ψακὰς δὲ λήγει.
δίκην δ᾽ ἐπ᾽ ἄλλο πρᾶγμα θηγάνει βλάβης 1535
πρὸς ἄλλαις θηγάναισι Μοῖρα. —
F(G)Tr
1511 mpooßaivav FTr: corr. Canter: προβαίνει vel -νοι Wilamowitz 1512 πάχνα
κουροβόρω (-ρῳ Tr) FTr: corr. Auratus 1513 χορός praefixit F, ἡμιχόριον GTr; cf.
1489 ἰὼ semel F 1515 τί ποτ᾽ dp’ Tr 1517 ἀσεβεῖ GTr : εὐσεβεῖ F
ἐκπνείων ΤΥ 1518 ἀνελεύθερα Tr 1510 cf. 1495 post 1522 γενέσθαι
lacunam statuit Wilamowitz, probabiliter 1526 τ᾽ ᾿Ιφιγένειαν hoc loco ab
Aeschylo non scriptum ante 1527 lacunam statuit Wilamowitz, probabiliter ;
desideratur verbum unde pendeat ἀλλ᾽ éuôv . . . τὴν πολύκλαυτον, exempli gratia ἔθυσε
πατήρ suppl. Wilamowitz 1527 ἀνάξια δράσας FTr: corr. Hermann 1529
rioas FTr ἦρξεν FTr: corr. Spanheim 1530 φροντίδων Tr 1531 εὐπάλαμ-
vov μέριμναν FTr: corr. Enger: εὐπάλαμον (servato μέριμναν) Porson 1532 ὅπα FTr
(superscr. n, ut solet, Tr) 1533 δημοσφαλῆ G 1534 ψεκὰς ΕἼΤΙ : corr. Blomfield
1535 δίκη F: δίκα (superscr. η) G: δίκᾳ (superscr. 7) Tr: corr. Auratus θηγάνει
ex Hesychii glossa θηγάνει" ὀξύνει (cf. Etym. M. p. 450. 13 θηγάνει" ὀξύνει, ἀκονᾶν) restituit
Hermann: θήγει FTr 1536 θηγάναις FTr: finale addidit Pauw
184
old, black Ares forces his way ; and wherever he advances he

will bring about a frost that devours the young.

O my king, my king, how shall I weep for thee? From my

loving mind what shall I say ? Thou hast been brought to lie
in this woven spider-web, breathing out thy life by an

impious death, ah me, to lie thus ignobly here, subdued in

a guileful doom at thy wife’s hand with a two-edged weapon.

Clyt. It was no ignoble death, methinks, that this man

died; (nor . . ..) For neither did he use guile to bring bane
upon the house; nay, my offspring that I conceived from
him, the much-lamented . . . Iphigeneia(?), (was slaughtered
openly by her father). What has been done to him is com-
mensurate with what he did; so let him not pride himself in
Hades, since with death by the sword he has paid for what
he wrought.

Chor. Bereft of meditation’s resourceful thought, I am at

a loss where to turn, now that the house is falling. I fear the
drumming of the rain of blood that will wreck the house: the
scattered shower is ceasing. And justice, to bring about a
new deed of harm, is being whetted on new whetstones by
the Moira.
185
 ἰὼ γᾷ γᾶ, εἴθ᾽ ἔμ᾽ ἐδέξω
πρὶν τόνδ᾽ ἐπιδεῖν ἀργυροτοίχου
δροίτης κατέχοντα χάμευναν. 1540
τίς ὁ θάψων νιν; τίς ὁ θρηνήσων;
N σὺ τόδ᾽ ἔρξαι τλήσηι, κτείνασ᾽
ἄνδρα τὸν αὑτῆς ἀποκωκῦσαι
ψυχῆι τ᾽ ἄχαριν χάριν ἀντ᾽ ἔργων 1545
μεγάλων ἀδίκως ἐπικρᾶναι;
τίς δ᾽ ἐπιτύμβιον alvov ἐπ᾽ ἀνδρὶ θείωι
σὺν δακρύοις ἰάπτων
ἀληθείαι φρενῶν πονήσει; —c 1550

ΚΛ, οὐ σὲ προσήκει τὸ μέλημ᾽ ἀλέγειν

τοῦτο’ πρὸς ἡμῶν
κάππεσε κάτθανε, καὶ καταθάψομεν,
οὐχ ὑπὸ κλαυθμῶν τῶν ἐξ οἴκων,
ἀλλ᾽ ᾿Ιφιγένειά νιν ἀσπασίως 1555
θυγάτηρ, ὡς χρή,
πατέρ᾽ ἀντιάσασα πρὸς ὠκύπορον
πόρθμευμ᾽ ἀχέων
περὶ χεῖρα βαλοῦσα φιλήσει.

ΧΟ. ὄνειδος ἥκει τόδ᾽ ἀντ᾽ ὀνείδους" 1560

δύσμαχα δ᾽ ἐστὶ κρῖναι.
φέρει φέροντ᾽, ἐκτίνει δ᾽ ὁ καίνων.
μίμνει δὲ μίμνοντος ἐν θρόνωι Διὸς
παϑεῖν τὸν ἔρξαντα' θέσμιον γάρ.
τίς ἂν γονὰν ἀραῖον ἐκβάλοι δόμων; 1565
κεκόλληται γένος πρὸς ἄται. ==

ΚΛ. ἐς τόνδ᾽ ἐνέβης ξὺν ἀληθείαι

χρησμόν: ἐγὼ δ᾽ οὖν
ἐθέλω δαίμονι τῶι Πλεισθενιδᾶν

F(G)Tr
1537 ἡμιχόριον praefixerunt FTr εἴθ᾽ ἔμ᾽ Ἐ: εἴθε μ᾽ Tr 1540 δροίτας FTr
cf. Eust. ad μ 357, p. 1726. 11 ἧς (scil. vocis Spotrn) χρῆσις καὶ παρ᾽ Αἰσχύλωι ἐν Ayayıdp-
vo νῦν κατέχοντα Tr; cf. ad 356 χαμεύναν FTr 1542 ἔρξαι GTr:
ἔρξαι F 1543 ἀποκωκύσαι F 1545 ψυχὴν ἄχαριν FTr: corr. E. A. 7. Ahrens
1547 ἡμιχόριον praefixerunt FTr ἐπιτύμβιος alvos FTr: corr. Casaubon 1549
δακρυοιν (v ex o facto) F 1551 οὔ oe F: οὔτε Tr μέλημα λέγειν FTr : recte
distinxit Karsten 1553 κάπτεσε xárÜavev Tr 1555 ἰφιγένειαν" ἵν᾽ FTr:
dist. Auratus 1558 ἀχέων ex ἀχαιῶν corr. G 1559 χεῖρε Porson
φιλήση FTr: corr. Stanley 1563 θρόνωι Schütz: χρόνω (-ῳ Tr) FTr 1565
dpatov Hermann: ῥᾶον F : ῥᾷον Tr 1566 πρὸς dra: Blomfield : προσάψαι FTr
1567 ἐνέβη FTr: corr. Canter o» G 1569 πλεισθενιδᾶν (superscr. wv Tr)
FTr
186
 Ο Earth, Earth, would thou hadst received me, or ever

I had seen him occupying the lowly bed of the silver-sided

bath! Who willit be that shall bury him ? who that shall sing
his dirge ? Wilt thou dare to do this—to bewail the husband
whom thine own hand has slain, and unrighteously to fulfil
for his shade an unkindly kindness in return for his great
deeds? And who will pour forth with tears the funeral praise
over the hero’s grave, and labour thereat in sincerity of
mind ?

Clyt. It is not for thee to concern thyself with this duty:

by my hand he fell, by my hand he died, and my hand shall
bury him—not with wailing of the household, but Iphigeneia
his daughter, as is fitting, shall with loving welcome meet her
father at the swift-flowing passage of the stream of woe, and
fling her arms round him and kiss him.

Chor. Taunt has now been met with taunt; hard is the

case to judge. The plunderer is plundered, the slayer pays

the price. And it abides, while Zeus abides upon his throne,

that to him who does it shall be done: for so it is ordained.

Who can expel the seed of curse from the house? The race
is glued fast to perdition.

Clyt. On this oracle thou hast entered with full truth ; but
187
 ὅρκους θεμένη τάδε μὲν στέργειν 1570
δύστλητά περ ὄνθ᾽, ὃ δὲ λοιπόν, ἰόντ᾽
ἐκ τῶνδε δόμων ἄλλην γενεὰν
τρίβειν θανάτοις αὐθένταισιν.
κτεάνων δὲ μέρος βαιὸν ἐχούσηι
πᾶν ἀπόχρη μοι μανίας μελάθρων 1575
ἀλληλοφόνους ἀφελούσηι.

ΑἸΙΓΙΣΘΟΣ
ὦ φέγγος eügpov ἡμέρας δικηφόρου᾽
φαίην ἂν ἤδη νῦν βροτῶν τιμαόρους
θεοὺς ἄνωθεν γῆς ἐποπτεύειν ἄχη,
ἰδὼν ὑφαντοῖς ἐν πέπλοις ᾿Ερινύων 1580
τὸν ἄνδρα τόνδε κείμενον, φίλως ἐμοί,
χειρὸς πατρώιας ἐκτίνοντα μηχανάς.
᾿Ατρεὺς γὰρ ἄρχων τῆσδε γῆς, τούτου πατήρ,
πατέρα Θυέστην τὸν ἐμόν, ὡς τορῶς φράσαι,
αὑτοῦ δ᾽ ἀδελφόν, ἀμφίλεκτος ὧν κράτει, | 1585
ἠνδρηλάτησεν ἐκ πόλεώς τε καὶ δόμων.
καὶ προστρόπαιος ἑστίας μολὼν πάλιν
τλήμων Θυέστης μοῖραν ηὔρετ᾽ ἀσφαλῆ,
τὸ μὴ θανὼν πατρῶιον αἱμάξαι πέδον
αὐτός: ξένια δὲ τοῦδε δύσθεος πατὴρ 1590
᾿Ατρεύς, προθύμως μᾶλλον ἢ φίλως, πατρὶ
τὠμῶι, κρεουργὸν ἦμαρ εὐθύμως ἄγειν
δοκῶν, πάρεσχε δαῖτα παιδείων κρεῶν.
τὰ μὲν ποδήρη καὶ χερῶν ἄκρους κτένας
ἔθρυπτ᾽ ἄνωθεν # * * Ἐν κῃ
“τὰ * ἀνδρακὰς καθήμενος' 1595
ἄσημα δ᾽ αὐτῶν αὐτίκ᾽ ἀγνοίαι λαβὼν
ἔσθει, βορὰν ἄσωτον, ὡς ὁρᾶις, γένει.
κἄπειτ᾽ ἐπιγνοὺς ἔργον οὐ καταίσιον
ὦιμωξεν, ἀμπίπτει δ᾽ ἀπὸ σφαγὴν ἐρῶν,
F(G)Tr
1570 θεμένα (in fine superscr. 5) Tr 1571 δύστλητά GTr : δύσπλητά F ὃ δὲ
Auratus: ὁ δὲ FTr 1573 om, Triclinius, mero lapsu, nam in scholio metrico
ad 1567 adscripto ἀναπαιστικὰ κῶλα «’ esse dicit, exhibet vero novem 1574 δὲ
Auratus: re FTr 1575 sq. πᾶν, ἀπόχρη μοι δ᾽ ἀλληλοφόνους μανίας μελάθρων ἀφελούση
(-on Tr) FTr: δ᾽ del. Canter, ἀλληλοφόνους post μελάθρων transposuit Erfurdt, pro-
babiliter 1577 εὔφρον FTr 1580 ἐριννύων FTr 1582 χειρὸς Tr: χερὸς F
1585 αὐτοῦ δ᾽ Elmsley : αὐτοῦ τ᾽ FTr 1588 εὕρετ᾽ FTr 1590 αὐτός Blomfield :
αὐτοῦ FTr ξενίᾳ Tr 1594 χερῶν GTr: χρεῶν F 1595 loci difh-
cillimi sive interpretationem sive emendationem nemo adhuc absolvit post ἄνωθεν
lacunam indicaverunt Hense et Wilamowitz an scribendum καθημένοις (Casau-
bon) vel -uévev? 1599 ὥμωξεν ex ὥμωζεν corr. F (cf. ad 329) duninre Canter:
ἄν' πίπτει FT τ σφαγῆς FTr: corr. Auratus
188
I for my part am willing to swear ἃ compact with the daemon
of the house of Pleisthenes and bear with all this, hard
though it be; while for the future he shall leave this house
and wear out some other family with deaths by hand of
kindred. And so long as I have a small part of ıny posses-
sions, I have enough of everything, once I have rid our halls
of the frenzy of mutual bloodshed.

Enter Aegisthus with a bodyguard!

Aegisthus. O kindly light of the day that brings retribu-
tion! Now at last I can say that the gods, as avengers of
men, look down from above with a watchful eye on the woes
of earth, now that I have seen this man here lying in the
woven robes of the Erinyes, to my delight, paying for what
his father’s hand contrived. For Atreus, being king in this
land, this man’s father, dealt thus with Thyestes, my
father—to put the matter clearly—and his own brother:
being questioned in his sovereignty, he drove Thyestes forth
in banishment from his city and home. And the unhappy
Thyestes, returning home as a suppliant at the hearth, found
a safe lot, safe from dying and staining his native soil with
blood—himself; but by way of hospitable welcome this
man’s godless father, Atreus, with more eagerness than
friendship, served up to my father, under pretence of cele-
brating a feast-day with good cheer, a banquet of his
children’s flesh. The foot-parts and the branching extremi-
ties of the arms he cut up small on top (?) (of the dish, and
with this dish Thyestes was served, Thyestes who was)
sitting apart by himself; but the indistinguishable parts of
it (i.e. of the flesh) he forthwith took in ignorance and ate, a
meal of ruin, as thou seest, to his race. And then, when he
perceived the monstrous thing that had been done, he
shrieked, and fell backward spewing out the butchery,
1 This is very likely though not certain. It is conceivable that the soldiers first hide
ee in the neighbourhood, and appear only when summoned by Aegisthus

189
 [μόρον δ᾽ ἄφερτον Πελοπίδαις ἐπεύχεται] 1600
λάκτισμα deinvou ξυνδίκως τιθεὶς ἀρᾶι,
οὕτως ὀλέσθαι πᾶν τὸ Πλεισθένους γένος.
ἐκ τῶνδέ σοι πεσόντα τόνδ᾽ ἰδεῖν πάρα.
κἀγὼ δίκαιος τοῦδε τοῦ φόνου papes:
τρίτον γὰρ ὄντα μ᾽ ἐπὶ δέκ᾽ ἀθλίωι πατρὶ 1605
συνεξελαύνει τυτθὸν ὄντ᾽ Ev σπαργάνοις’
τραφέντα δ᾽ αὖθις ἡ δίκη κατήγαγεν.
καὶ τοῦδε τἀνδρὸς ἡψάμην θυραῖος ὦν,
πᾶσαν συνάψας μηχανὴν δυσβουλίας.
οὕτω καλὸν δὴ καὶ τὸ κατθανεῖν ἐμοί, 1610
ἰδόντα τοῦτον τῆς δίκης ἐν ἕρκεσιν.
ΧΟ. Αἴγισθ᾽, ὑβρίξειν ἐν κακοῖσιν οὐ σέβω.
σὺ δ᾽ ἄνδρα τόνδε φὴις ἑκὼν κατακτανεῖν,
μόνος δ᾽ ἔποικτον τόνδε βουλεῦσαι φόνον; |
οὔ φημ’ ἀλύξειν ἐν δίκηι τὸ σὸν κάρα 1615
δημορριφεῖς, σάφ᾽ ἴσθι, λευσίμους ἀράς.
ΑΙ. σὺ ταῦτα φωνεῖς νερτέραι προσήμενος
κώπηι, κρατούντων τῶν ἐπὶ τυγῶι δορός;
γνώσηι γέρων ὧν ὡς διδάσκεσθαι βαρὺ
τῶι τηλικούτωι, σωφρονεῖν εἰρημένον, 1620
δεσμὸς δὲ Kal τὸ γῆρας αἵ τε νήστιδες
δύαι διδάσκειν ἐξοχώταται φρενῶν
ἰατρομάντεις. οὐχ ὁρᾶις ὁρῶν τάδε;
πρὸς κέντρα μὴ λάκτιξε, μὴ παίσας μογῆις.
ΧΟ. γύναι, σὺ τοὺς ἥκοντας ἐκ μάχης νέον 1625
olkoupós εὐνὴν ἀνδρὸς αἰσχύνων ἅμα
ἀνδρὶ στρατηγῶι τόνδ᾽ ἐβούλευσας μόρον;
ΑΙ. καὶ ταῦτα τἄπη κλαυμάτων ἀρχηγενῆ.
Ὀρφεῖ δὲ γλῶσσαν τὴν ἐναντίαν ἔχεις"
ὁ μὲν γὰρ ἦγε πάντ᾽ ἀπὸ φθογγῆς χαρᾶι, 1630
σὺ δ᾽ ἐξορίνας νηπίοις ὑλάγμασιν
&Enı κρατηθεὶς δ᾽ ἡμερώτερος φανῆι,
ΧΟ. ὡς δὴ σύ μοι τύραννος ᾿Αργείων ἔσηι,
ὃς οὐκ, ἐπειδὴ τῶιτδ᾽ ἐβούλευσας μόρον,
δρᾶσαι τόδ᾽ ἔργον οὐκ ἔτλης αὐτοκτόνως. 1635
F(G)Tr
1600 eiecit J. C. Schmitt 1601 dp&G 1602 ὀλέσθαι testatur Tzetzes (in scholiis
ad alleg. Iliad., Crameri Anecd. Oxon. iii. 378. 10, καὶ AtoxvAos λέγων" ‘ ἀρᾶτ᾽ ὀλέσθαι
πᾶν τὸ Πλεισθένους γένος ’): ὀλέσθη ΕἼΤ 1609 ξυνάψας Tr 1611 ἰδόντι Tr
1613 τόνδ᾽ ἔφης FTr: recte distinxit Pauw 1617 veprépa (-pa Tr) GTr: verepa F
1621 δεσμὸς Tr: δεσμὸν F : δεσμοὶ Karsten 1622 éfoxórare G 1624 παίσας
testatur schol. Pind. P. 2. 173¢ (Αἰσχύλος Ἀγαμέμνονι" ‘ πρὸς κέντρα μὴ λάκτιζε,
μὴ παίσας μογῆις Ὗ: πήσας ΕἼΤ 1626 αἰσχύνουσ᾽ FTr: corr. Keck (αἰσχύνας
Wieseler) 1631 ἠπίοις FTr: corr. Jacob 1634 τῷδ᾽ ἐβουλεύσας Tr: rade
βουλεύσας F

190
kicking over the table and making this act of equal right
with the curse he uttered: ‘so perish all the race of Pleis-
thenes’. Hence it is that thou canst see this man laid low.
And it is with justice that I am he that schemed this killing ;
for he drove me, the thirteenth (child), into exile with my
unhappy father, while I was yet a babe in swaddling-clothes ;
but when I had grown to manhood Justice brought me back
again. And I laid hands on this man though I was not
present there, by putting together the whole device of the
fatal plan. This being so, even death would now be welcome
to me, now that I have seen this man in the toils of Justice.
Chor. Aegisthus, to triumph in misfortune is a thing I care
not to practise. Thou avowest, dost thou, that thou didst wil-
fully slay this man, and alone didst plan this pitiable mur-
der? I say that in the hour of justice thy head, be very sure,
shall not escape the people’s pelting of stones and curses.
Aeg. Dost thou talk thus, thou that art seated below at
the oar, while those on the (helmsman’s) bench are masters
of the ship? Thou shalt learn now in thine old age how hard
a thing it is to be taught a lesson at thy time of life, when
discretion is enjoined upon thee. Prison-bonds, and the
pangs of hunger, are most excellent healer-prophets for the
mind to teach even old age. Canst thou see and seest not
this? Kick not against the pricks, lest in striking them thou
suffer pain.
Chor. Thou woman, thou: to do this to those newly come
from battle, and, while as a stay-at-home thou wast defiling
the man’s bed, to plot this death against the general in the
field!
Aeg. These words too are the breeders of a race of rueful
cries. Thy tongue is the opposite of Orpheus’ tongue ; for he,
by his voice, led all things after him in delight, but thou
stirrest up anger by foolish barkings and shalt be led away.
But once mastered thou wilt show thyself more tame.
Chor. As if I shall see thee ruler over men of Argos, thee
who after plotting death against this man hadst not the
courage to do this deed by killing him with thine own hand!
I9I
 Al. τὸ γὰρ δολῶσαι πρὸς γυναικὸς ἦν σαφῶς,
ἐγὼ δ᾽ ὕποπτος ἐχθρὸς À παλαιγενής.
ἐκ τῶν δὲ τοῦδε χρημάτων πειράσομαι
ἄρχειν πολιτῶν τὸν δὲ μὴ πειθάνορα
φεύξω βαρείαις, οὔτι μὴν σειραφόρον 1640
κριθῶντα πῶλον, ἀλλ᾽ ὁ δυσφιλὴς σκότωι
λιμὸς ξύνοικος μαλθακόν σφ᾽ ἐπόψεται.
ΧΟ. τί δὴ τὸν ἄνδρα τόνδ᾽ ἀπὸ ψυχῆς κακῆς
οὐκ αὐτὸς Tiväpızes, ἀλλά νιν γυνὴ
χώρας μίασμα καὶ θεῶν ἐγχωρίων 1645
ἔκτειν᾽; Ὀρέστης ἄρά trou βλέπει φάος,
ὅπως κατελθὼν δεῦρο πρευμενεῖ τύχηι
ἀμφοῖν γένηται τοῖνδε παγκρατὴς φονεύς;
Al. ἀλλ᾽ ἐπεὶ δοκεῖς τάδ᾽ ἔρδειν καὶ λέγειν, γνώσηι τάχα.
εἶα δή, φίλοι λοχῖται, τοῦργον οὐχ ἑκὰς τόδε. 1650
ΧΟ. εἴα δή, ξίφος πρόκωπον πᾶς τις EUTPETTIZETW.
ΑΙ. ἀλλὰ κἀγὼ μὴν πρόκωπος κοὐκ ἀναίνομαι θανεῖν.
ΧΟ. δεχομένοις λέγεις θανεῖν σε’ τὴν τύχην δ᾽ αἱρούμεθα,
ΚΛ, μηδαμῶς, ὦ φίλτατ᾽ ἀνδρῶν, ἄλλα δράσωμεν κακά’
ἀλλὰ καὶ τάδ᾽ ἐξαμῆσαι πολλά, δύστηνον θέρος. 1655
πημονῆς δ᾽ ἅλις γ᾽ ὑπάρχει: μηδὲν αἱματώμεθα.
στείχετ᾽ αἰδοῖοι γέροντες πρὸς δόμους ὑπεττρωμένους} [τούὔσδε],
πρὶν παθεῖν ἔρξαντες" αἰνεῖν χρὴ τάδ᾽ ὡς ἐπράξαμεν.
εἰ δέ τοι μόχθων γένοιτο τῶνδ᾽ ἅλις, δεχοίμεθ᾽ ἄν,
δαίμονος χηλῆι βαρείαι δυστυχῶς πεπληγμένοι. 1660
ὧδ᾽ ἔχει λόγος γυναικός, εἴ τις ἀξιοῖ μαθεῖν.
ΑΙ. ἀλλὰ τούὐσδ᾽ ἐμοὶ ματαίαν γλῶσσαν ὧδ᾽ ἀπανθίσαι
κἀκβαλεῖν ἔπη τοιαῦτα δαίμονος πειρωμένους,
σώφρονος γνώμης [δ ἁμαρτὼν τὸν κρατοῦντα (Acıdopeis).
F(G)Tr
1637 1j Porson (ἦν Canter): 3 FTr 1638 τῶνδε FTr: distinxit Casaubon — 1640
μὴν Wieseler : μὴ FTr 1640 sq. Pollux 7. 24 Αἰσχύλος... εἴρηκε σειραφόρον κριθῶντα
πῶλον᾽ 1641 σκότωι Auratus: κότω FTr 1642 σύνοικος G 1644 νιν Spanheim:
σὺν ΕἼΤ 1646 dpa mov F: dpa ποῦ GTr 1650 Aegisthum pergere vidit Stanley:
χορός praefixerunt FTr 1651 χορὸς ex aly correctum praefixit F: χορὸς G: nulla
nota in Tr 1652 πρόκωπος (w corr. ex o) G: πρόκοπος F: πρόκοπτος Tr κοὐκ
Lobelio monente (οὐδ᾽ ille) scripsi: οὐκ FTr 1653 αἱρούμεθα Auratus: ἐρούμεθα ΕἼΤ
1654 δράσομεν FTr: corr. Victorius 1655 θέρος Schütz : ó ἔρως FTr 1656 ὕπαρχε
FTr: corr. Scaliger πημονῆς ἅλις δ᾽ ὑπάρχει Hermann ἡματώμεθα FTr: corr.
Jacob 1657 oreixer’ αἰδοῖοι Ahrens: στείχετε δ᾽ οἱ FTr πεπρωμένους inane
τούσδε del. Auratus 1658 ἔρξαντες F : ἔρξαντα GTr αἰνεῖν Heath : καιρὸν FTr
χρὴν FTr: corr. Hartung ἐπραξάμην FTr: corr. Victorius 1659 y’ ἐχοίμεθ᾽
FTr: corr. Hermann 1660 πεπληγμένοις G 1662 τούσδ᾽ ἐμοὶ Is. Vossius:
τούσδε μοι FTr ἀπανθίσαι mire dictum 1663 δαίμονας FTr: corr. Casaubon
πειρωμένη G 1664 versus male mutilatus ut prorsus restituatur vix fieri potest ;
eo textu quem exempli gratia dedi aliquam saltem medendi rationem adumbratam esse
spero δ᾽ del. Schwerdt ἁμαρτὼν τὸν Schwerdt : ἁμαρτῆτον FTr inter γνώμης
et κρατοῦντα (quod in fine versus positum est) lacunam habet G λοιδορεῖς suppl. Vossius
192
 Aeg. Because the deceit was clearly the woman’s part,
while I was his suspected enemy from of old. But I will
endeavour to use this man's wealth to rule the citizens ; and
him who will not obey I will yoke under a heavy yoke—no
barley-fed young trace-horse, mark my words! No, hunger,
hateful housemate of darkness, shall see him softened.
Chor. Why then in the cowardice of thy heart didst thou
not slay the man thyself, but didst let the woman, to the pol-
lution of the land and the land’s gods, do the killing ? Oh, is
Orestes somewhere alive, that he may return hither with
auspicious fortune and kill this pair and prove the trium-
phant victor?
Aeg. Well then, since thou art determined so to act and
speak, thou shalt have thy lesson forthwith. What ho! my
friends of the guard, the work to be done here is at hand.
Chor. What ho! let every man make ready his sword, hilt
in hand!
Aeg. Nay, I too am hilt-in-hand, and do not refuse! to die.
Chor. Thou speakest of thy death: we accept the omen,
and choose to take what will come to pass.
Clyt. Nay, nay, my dearest, let us work no further ills:
even those that are here are many to reap, an unhappy
harvest. We have harm enough already: let us keep free
from bloodshed. Go your ways, reverend elders, to your
homes . . . ere doing bring you suffering: we must accept
our present lot, even as it has fallen to us. But should it come
to pass (that we could say) ‘enough of these afflictions’, we
should welcome it, sadly stricken as we are by the heavy
hoof of the daemon. That is what a woman has to say, if any
think fit to heed it.
Aeg. But that these should . . . such idle speech against
me and fling out such words, putting their fate to the test!
Thou lackest a sober mind, so to abuse thy master.?
! For the ambiguity, which cannot be rendered, see the commentary.
2 Text uncertain.

193
 ΧΟ. οὐκ ἂν ᾿Αργείων τόδ᾽ εἴη, φῶτα προσσαίνειν κακόν. 1665
ΑΙ. ἀλλ᾽ tye σ᾽ ἐν ὑστέραισιν ἡμέραις μέτειμ᾽ ἔτι.
XO. οὔκ, ἐὰν δαίμων (y') Ὀρέστην δεῦρ᾽ ἀπευθύνηι μολεῖν.
Al. οἶδ᾽ ἐγὼ φεύγοντας ἄνδρας ἐλπίδας σιτουμένους.
ΧΟ. πρᾶσσε, πιαίνον, μιαίνων τὴν δίκην, ἐπεὶ πάρα.
ΑΙ. ἴσθι μοι δώσων ἄποινα τῆσδε μωρίας χρόνωι. 1670
ΧΟ. κόμπασον θαρσῶν, ἀλέκτωρ ὥστε θηλείας πέλας.
KA. μὴ προτιμήσηις ματαίων τῶνδ᾽ ὑλαγμάτων᾽ ἐγὼ
καὶ σὺ δωμάτων κρατοῦντε * + θήσομεν καλῶς.
F(G)Tr
1665 προσσαΐνειν GTr: προσαίνειν F 1666 xAvr. praefixit F 1667 γ᾽ suppl.
Headlam 1670 χρόνωι Wecklein : χάριν FTr 1671 ϑαρρῶν FTr: corr. Porson
ὥστε Canter: ὥσπερ ΕἼΤ 1672 sq. schol. vet. Tricl. ᾿Εγώ, φησί, καὶ σὺ κρατοῦντες
τῶνδε τῶν δωμάτων διαθησόμεθα τὰ (Victorius: τὸ Tr) καθ᾽ αὐτοὺς καλῶς, unde Canter
ἐγὼ, Auratus καλῶς, quae perierunt in FTr, restituerunt 1673 καὶ σὺ θήσομεν
κρατοῦντε τῶνδε δωμάτων FIT: transposui, τῶνδε eieci nescio δὴ fuerit πάντα (νεῖ
ταῦτα vel τἄλλα) θήσομεν καλῶς.

194
 Chor. It would not be like men of Argos to fawn upon a
villain.
Aeg. But I in days to come will yet visit thee with my
vengeance.
Chor. Not if Destiny guide Orestes back hither to his
home.
Aeg. 1 know that exiles feed on hopes.
Chor. On with it, make thee fat, befouling justice, since
thou canst.
Aeg. Thou shalt make me amends, be sure, for this folly
in due time.
Chor. Boast and be confident, like a cock beside his hen.
Clyt. Take no account of these idle barkings: thou and I
as masters of the house shall order all things! well.
Clytemnestra and Aegisthus go into the house. The Chorus
leave the orchestra.
1 Text uncertain.

195
 AESCHYLUS
AGAMEMNON
 AESCHYLUS

AGAMEMNON
EDITED

WITH A COMMENTARY BY

EDUARD FRAENKEL

VOLUME II
COMMENTARY ON I-10£55

OXFORD
AT THE CLARENDON PRESS
 This book has been printed digitally and produced in a standard specification
in order to ensure its continuing availability

OXFORD
UNIVERSITY PRESS

Great Clarendon Street, Oxford OX2 6DP

Oxford University Press is a department of the University of Oxford.
It furthers the University@ objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
S<o Paulo Shanghai Taipei Tokyo Toronto

Oxford is a registered trade mark of Oxford University Press

in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York

© Oxford University Press 1950

The moral rights of the author have been asserted

Database right Oxford University Press (maker)
Reprinted 2003

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose this same condition on any acquirer
ISBN 0-19-927171-2 (Vol. 2)
ISBN 0-19-814101-7 ( Set of three Vol.@)
 CONTENTS
VOLUME II

A SELECT LIST OF ABBREVIATIONS vi

COMMENTARY ON I-I055 . . 1
 A SELECT LIST OF ABBREVIATIONS
(Most of the abbreviations used in the commentary need no explanation)
Ahrens. II. L. Ahrens, ‘Studien zum Agamemnon des Aeschylus’, Philologus,
Supplementband i, 1860, 213 ff., 477 ff., 535 ff.
Arch. Jahrb. Jahrbuch des Deutschen Archäologischen Instituts.
Ath. Mitt, Mitteilungen des Deutschen Archäologischen Instituts: Athenische
Abteilung.
Berl. Sitzgsb. Sitzungsberichte der Preussischen Akademie der Wissenschaften,
phil.-hist. Klasse.
Buck and Petersen, A Reverse Index. C.D. Buck and W. Petersen, A Reverse Index of
Greek Nouns and Adjectives. Chicago 1944.
Collitz-Bechtel. Sammlung der griechischen Dialekt-Inschriften, herausgegeben von
H. Collitz und F. Bechtel. Göttingen 1884-1915.
C.Q. Classical Quarterly.
C.R. Classical Review.
D. added to the number of a lyric fragment refers to Anthologia Lyrica Graeca ed.
E. Diehl, vol. i, 2nd ed., vol. ii, 1st ed.
Daremberg-Saglio. Ch. Daremberg et Edm. Saglio, Dictionnaire des antiquités grecques
el romaines. Paris 1877-1919.
Daube. B. Daube, Zu den Rechtsproblemen in Aischylos’ Agamemnon. Zürich
1938.
Dittenberger Syll. Syllage Inseriplionum Graecarum, ed. Wilh. Dittenberger ; ed. tertia.
Lipsiae 1915-19.
Ferrari, La parodos, W. Ferrari, ‘La parodos dell’ “ Agamennone”’, Annali della
IM ες rs

R. Scuola Normale Superiore di Pisa, Lettere, Storia e Filosofia, Serie II,

Vol. vii, Fasc. IV, 1938, pp. 355 ff.
Ed. Fr., Kolon u. Satz, ü. Eduard Fraenkel, ‘Kolon und Satz, Beobachtungen zur
Gliederung des antiken Satzes, II’, Nachrichten von der Gesellschaft der
Wissensch. zu Gottingen, phil.-hist. Klasse 1933, Fachgruppe I, Nr. 19.
Ernst Fraenkel, Denominativa. Ernst Fraenkel, Griechische Denominativa in ihrer
geschichtlichen Entwicklung und Verbreitung. Göttingen 1906.
Ernst Fraenkel, Nomina agentis. Ernst Fraenkel, Geschichte der griech. Nomina agentis
auf -rip, -τωρ, -rns (-r-). Strassburg 1910 and 1912.
F Gr Hist Die Fragmente der Griechischen Historiker, von F. Jacoby. Berlin-
Leiden 1923-.
Furtwängler-Reichhold. A. Furtwängler und K. Reichhold, Griechische Vasenmalerei.
München 1904-32.
Goodwin. W. W. Goodwin, Syntax of the Moods and Tenses of the Greek Verb;
rewritten and enlarged. Boston 189o.
Gölt. gel. Anz. Gôttingische gelehrte Anzeigen unter Aufsicht der Gesellschaft der
Wissenschaften.
Idg. Forsch. Indogermanische Forschungen.
IG Inscriptiones Graecae. Berolini.
J.Ph. Journal of Philology.
K. added to the number of a comic fragment refers to Comicorun Atticorum Fragmenta,
ed. T. Kock.
Kuhns Zeilschr. Zeitschrift für vergleichende Sprachforschung auf dem Gebiete der
indogermanischen Sprachen.
Kühner-Blass. Ausführliche Grammatik. der griech. Sprache von R, Kühner. I.
Elementar- und Formenlehre. 3. Aufl. von F. Blass. Hannover 1890
and 1892.
vi
 A SELECT LIST OF ABBREVIATIONS
Kühner-Gerth. Ausführliche Grammatik der griech. Sprache von KR. Kühner. II.
Satzlehre. 3. Aufl. von B. Gerth. Hannover 1898 and 1904.
Kumaniecki. C. F. Kumaniecki, ‘De elocutionis Aeschyleae natura’, Archiwum
Filologiczne Nr 12. Cracoviae 1935.
L-S A Greek-English Lexicon, compiled by H. G. Liddell and R. Scott:
A New [oth] Edition revised . . . by Sir Henry Stuart Jones. Oxford
1925-40.
P. Maas, Griech. Metrik. P. Maas, Griechische Metrik (Einleitung in die Altertums-
wissenschaft, herausgegeben von A. Gercke und E. Norden; i. Band,
7. Heft). Leipzig und Berlin 1929.
Matthiä, Griech. Gramm. Ausführliche Griechische Grammatik von A. Matthiä. 3. Aufl.
Leipzig 1835.
Meisterhans. Grammalik der Attischen Inschriften von K. Meisterhans. 3. Aufl.
besorgt von E. Schwyzer. Berlin 1900.
N. added to the number of a tragic fragment refers to Tragicorum Graecorum Fragmenta,
rec. À. Nauck, 2nd ed., 1889.
Nachr. Gôtt. Ges. Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen;
phil.-hist. Klasse.
Oesterr. Jahresh. Jahreshefte des Oesterreichischen Archäologischen Instituts.
Pap. Oxy. Oxyrhynchus Papyri, ed. B. P. Grenfell and A. S. Hunt, &c. London
1898-.
Pap. Soc. It. Papiri greci e latini (Pubblicazioni della Societa Italiana per la ricerca
dei papiri ...). Firenze 1912-.
Passow. Handwörterbuch der griechischen Sprache, begründet von F. Passow.
5. Aufl., bearbeitet . . . von V. C. F. Rost und F. Palm. Leipzig
1841-57.
Passow-Crónert. Passow's Wörterbuch der griech. Sprache, völlig neu bearbeitet von
W. Crónert (a-dvd). Góttingen 1912-14.
Pf. added to the number of a fragment of Callimachus refers to Callimacht fragmenta
nuper reperta, ed. R. Pfeiffer. Bonn 1923.
Philol. — Philologus ; Zeitschrift für das klassische Altertum.
Plüss. Aeschylos’ Agamemnon mil erlduternden Anmerkungen, herausgegeben
von R. Enger, 3. Auflage neu bearbeitet von Theodor Plüss. Leipzig
1895.
RE Paulys Real-Encyclopädie der classischen Altertumswissenschaft; Neue
Bearbeitung von G. Wissowa, W. Kroll und K. Mittelhaus. Stuttgart
1894-.
Röm. Mit. Mitteilungen des Deutschen Archäologischen Instituts: Römische
Abteilung.
Rhein. Mus. Rheinisches Museum für Philologie. Neue Folge.
Schuursma. J. A. Schuursma, De poelica vocabulorum abusione apud Aeschylum.
Amstelodami 1932.
Schwyzer, Exempla. Dialectorum Graecarum exempla epigraphica potiora, ed. E. Schwy-
zer. Lipsiae 1923.
Schwyzer, Griech. Gramm. I. Griechische Grammalik . . . von E. Schwyzer (Handbuch
der Altertumswissenschaft, herausg. von W. Otto, 2. Abteil., τ. Teil,
1. Band). München 1939.
Snell, Aischylos. B. Snell, ' Aischylos und das Handeln im Drama’, Philologus, Supple-
mentband xx, Heft 1, 1928.
Sokrates. Sokrates ; Zeitschrift für das Gymnasialwesen.
Stud. It. Studi italiani di filologia classica.
Wackernagel, Dehnungsgeselz. Jacob Wackernagel, Das Dehnungsgesetz der griechischen
Composita. Basel 1889.
vii
 A SELECT LIST OF ABBREVIATIONS
Wackernagel, Syntax. 1]. W., Vorlesungen über Syntax. Basel 1920 and 1924.
Wackernagel, Unters. 2. Homer. J. W., Sprachliche Untersuchungen su Homer (For-
schungen zur griech. und lat, Grammatik, herausg. von P. Kretsch-
mer und W. Kroll; 4. Heft). Göttingen 1916.
Wackernagel, Verm. Beiträge. J. W., Vermischte Beiträge zur griechischen Sprachkunde
(Programm zur Rektoratsfeier der Univ. Basel). Basel 1897.
Wilamowitz, Interpr. U. von Wilamowitz-Moellendorff, Aischylos, Interpretationen,
Berlin 1914.
Wilamowitz, Griech. Tragoedien. Griechische Tragoedien, übersetzt von U. von W.-M.
Berlin 1899-1923.
Williger, Sprachl. Unters. W. Williger, Sprachliche Untersuchungen zu den Komposila
der griech. Dichter des 5. Jahrhunderts (Forschungen z. griech. u. lat.
Grammatik, herausg. von P. Kretschmer und W. Kroll; 8. Ileft).
Góttingen 1928.

viii
 COMMENTARY
1. θεοὺς μέν: attempts to find in 8 καὶ νῦν or even in 20 νῦν δέ the clause
answering to μέν here are forced, and Denniston, Particles, 382, rightly rejects
them. He points to the tendency in early poetry and prose to start a speech
with μέν. Of the seven extant plays of Aeschylus all except Sept. and
Cho. open with μέν; so did the Myrmidons (fr. 131 N.). In Sophocles the
prologues of Aj., Trach., Phil. begin with μέν (in the four other extant plays
a vocative occupies the first place, as in A. Sept., Cho., Mvaot fr. 143 N.).!
In Euripides we find μέν openings in Hipp. and Hel. But in Ag. x and other
places of the kind we must also recognize that μέν is deprived of its answering
δέ by a change in the form of a long sentence’ (Jebb on S. Phil. 1; cf. the
convincing explanations of Wilamowitz on A. Eum. 636 and Suppl. 410 and
in his Pindaros, p. 467 f.). In particular, the μέν here gives the impression
that the Watchman intends, after his prayer to the gods, to pass on to
other matters. But many vivid details of his nightly troubles come flooding
into his mind, and he has hardly resumed his main theme when suddenly
(after line 21) he espies the beacon, and all is changed.
2. What is the construction of μῆκος Modern critics were not the first to
feel this difficulty, as we may perhaps infer from theattempted 'emendation'
in MV : μῆκος δ᾽ ἦν, which is clearly related to M's punctuation (a colon after
ereias). One of the scholia in M, ἦν ἐπὶ μῆκος κοιμώμενος, points the same way
(the other, τῶν κατὰ τὸ μῆκος τῆς ἐτείας φρουρᾶς, presupposes the uninter-
polated text). μῆκος δ᾽ ἦν κοιμώμενος neither scans nor makes sense. That it
takes ἦν as first person, for which Aeschylus and Sophocles use only the older
form 7, is a less serious objection; for ἦν might be altered to 7, as Elmsley
had to alter it in several passages of Sophocles.
To turn to modern discussions of the passage, Stanley’s conjecture μῆχος
(abandoned by himself, and later taken up by Valckenaer and many others)
is wrong: (1) the word does not occur in Aeschylus, nor for that matter in
Sophocles (others therefore have suggested μῆχαρ) ; (2) the force of ἀπαλλαγὴν
πόνων would be weakened if it were taken up by φρουρᾶς μῆχος. The MS text
is sound. But even of the editors who recognize this several have dis-
torted the sense by violence or over-ingenuity. The first point to get
clear is that φρουρᾶς Ereias μῆκος is in apposition to τῶνδε πόνων, and not
an adverbial qualification of αἰτῶ (‘per longitudinem annuae vigiliae posco
liberationem’) as it is taken by Klausen, Peile, Ahrens, Weil, Paley, Nägels-
bach, Headlam, Goodwin ($ 26), Sheppard, C.R. xxxvi, 1922, 6 n. 2, and others.
This interpretation takes φρουρᾶς ἐτείας μῆκος as a mere indication of
time, although the πόνοι consist precisely in the φρουρά. Besides, αἰτῶ
is then retrospective, summarizing the past (though it may also include the
present); and this is out of place in an announcement in the first line
of a prologue, which should refer to the moment when the speaker opens his
! That "Iw Κάϊκε «rA., spoken κατὰ τὴν εἰσβολὴν τοῦ mpoÀóyov, is the first line of the
play is clear; cf. Welcker, Die Griech. Tragódien, i. 53, and see below on 503. Wilamowitz,
Introd. to his translation of Soph. Phil. (Griech. Tragódien, xii), p. 9, n. 2, suggests that the
fragment of the Philoctetes of Aeschylus, 249 N., Σπερχειὲ ποταμὲ βούνομοί τ᾽ ἐπιστροφαΐ, was
the first line of the play. This is possible (cf. Hermann, Opusc. iii. 122), but not certain.
4872.2 B I
line 2 COMMENTARY
lips to speak. There is here an obvious identity of form with Eum. 1 πρῶτον
μὲν εὐχῆι τῆιδε πρεσβεύω θεῶν τὴν πρωτόμαντιν Γαῖαν. Moreover, the inter-
pretation of Klausen and his followers misses an important stylistic point.
The use of apposition to elaborate or intensify a detail of a preceding
statement is a favourite device of Aeschylus, and it seems to be specially
characteristic of the Watchman’s speech. Drop by drop,' one by one, ideas
and images form in his weary head, and as they come he gives them utterance,
advancing to clearer and more forceful expression of his experience. Thus he
says in 3 ἄγκαθεν, κυνὸς δίκην, in 8 f. λαμπάδος τὸ σύμβολον, αὐγὴν πυρὸς φέρουσαν
ἐκ Τροίας φάτιν κτλ., to which we may add the participles developing the sense
in 17, 19, and 33; cf. also on 10. From this it is clear that φρουρᾶς develops
and stands in apposition to τῶνδε πόνων. Again, it is certain that Ereios
cannot mean ‘which has lasted for many (or for several) years’ (Ahrens—
his full discussion of the word on pp. 219 ff. shows that his interpretation is
impossible—Karsten, Mazon), but only (since the sense ‘returning year by
year’ is out of the question) ‘lasting one year’. As Wilhelm Schulze observes
(Zur Geschichte lat. Eigennamen, 50 n. 3), ‘the singular does not as a rule
require a special expression . . . the word annus by itself means “one year” ;
therefore anniculus means “one year old" '. Cf. annuus. The detail of the
watch that lasts a year comes from the Odyssey (ὃ 526 φύλασσε δ᾽ ὅ γ᾽ εἰς
ἐνιαυτόν),2 as does the figure of the Watchman.? It only remains to explain
μῆκος. Here too an impartial balancing of the alternatives leads to the
conclusion that the obvious meaning is the right one: ‘ab his laboribus, a
vigilia sc., quae longitudine annua est’ (Wellauer). Hermann rejected this
interpretation with contempt, quite wrongly. The right view is found e.g.
in Heinrich Voss (‘Gebt, Götter, fleh’ ich, dieser Mühe Erledigung, der Hut,
ein Jahr an Lange’), Conington, Franz Beckmann, Bemerkungen zu Prol. u.
Parodos des aesch. Ag. (Braunsberg 1867), 13, B. Risberg, De nonnullis locis
Ag. Aesch. (Diss. Upsala 1891), 1 ff., Pliiss (φρουρᾶς ἐτείας in apposition to
πόνων; and μῆκος... goes with ereias’). It is wholly unprofitable to com-
pare the difficult passage S. Ant. 857 f. ἔψαυσας ἀλγεινοτάτας ἐμοὶ μερίμνας
πατρὸς τριπόλιστον οἶτον, which is cited with quite different interpretations
by Hermann (cf. also his Opuscula, i. 169 f.) and Wilamowitz (Verskunst, 519)
independently. The latter, both there and again in Hermes, lxi, 1926, 279,
repeats the assertion already made in his Isyllos von Epidauros (1886), p. 186,
that the accusative, to put it bluntly, stands in place of a genitive: μήκους
would be demanded by strict grammar, but is altered because there are so
many other genitives. In reality, the combination ἐτείας μῆκος is on the lines
of a very ancient and well-established usage, and only the particular meaning
of ἔτειος and the function that μῆκος derives from it give the expression any
peculiar character. In A 312 we read (of Otus and Ephialtes) ἐννεαπή-
xees ... εὖρος, ἀτὰρ μῆκός ye γενέσθην ἐννεόργυιοι, and in Ar. Ach. 909 μωκκός
ya paxos οὗτος. In the Watchman's mouth the addition of μῆκος, ‘one year
in length’, is not otiose, but marks the period of his watching as a μακρὸς
1 I sometimes call this mode of expression guttalim for short.
2 Mazon in his edition of the Oresteta, p. xxv, n. 2, seems to misinterpret both the Odyssey
passage and ἐτείας in Ag. 2.
3 The obvious difference is pointed out in the note in M: θεράπων ᾿Αγαμέμνονος ὁ
προλογιζόμενος, οὐχὶ 6 ὑπὸ Αἰγίσθου ταχθείς (from a fuller ὑπόθεσις).
2
 COMMENTARY line 3
χρόνος. “μῆκος defesso et impatienter exspectanti aptissimum est’ (Wellauer).
A single year need not seem much, measured against the length of the Trojan
War or many of the νόστοι, but for the Watchman himself it is cruelly long.
κοιμώμενος : Schneidewin wrongly calls φρουρὰν κοιμᾶσθαι an oxymoron,
on the ground that properly φρουρὰ φρουρεῖται. Passow, on the other hand,
s.v. κοιμᾶσθαι (followed by L-S), rightly puts this passage under the heading
‘sich lagern, um Wache zu halten, excubare’ together with Xen. Cyr. 1. 2. 4,9
of ἔφηβοι... . κοιμῶνται περὶ τὰ ἀρχεῖα. The verb κοιμᾶσθαι, normally meaning
‘to sleep’ (not merely ‘to lie’, though the etymological kinship with κεῖσθαι,
κοίτη is unmistakable),' often takes on a more general meaning. W. Schulze,
Quaest. eb. 73, and Wackernagel, Vermischte Beiträge, 4, have pointed out
that words meaning ‘sleep’ are freely used of lying awake at night, e.g.
iadeıw in Homer and καθεύδειν in Attic; cf. Wilamowitz on Ar. Lys. 282. A
sensible explanation is found in the scholia (ZxoA. παλ. in Tr and, with a
few irrelevant variations, in F) : ἀστείως εἴρηται τὸ κοιμώμενος ἐπὶ μὴ ὑπνοῦντος.
πῶς γὰρ ἂν καὶ dpoupoin καὶ ἄστρα βλέποι τις ὑπνῶν; ἀλλ᾽ ἐπὶ ἀνακλίσεως
ἁπλῶς εἴρηται, iv’ εἴη 6 φύλαξ οὗτος κοιμώμενος καθὰ καὶ κύων ἐν φυλακῆι
νυκτερινῆι.
3. στέγαις: can this really mean ‘upon the roof’, as it is rendered by most
modern editors and translators? This view is a survival from the time when
it was generally accepted that ἄγκαθεν was the same as ἀνέκαθεν (so e.g.
Stanley, ‘excubans super aedes Atridarum’), and remained after a different
interpretation of ἄγκαθεν (see below) had practically won the day. Misgivings
about taking στέγαις as ‘on the roof’ might perhaps be roused by the plural.
A more serious objection, however, is made by Verrall: ‘the use of the dative
is dubious; there is nothing in κοιμώμενος to determine the dative... to the
meaning on: κοιμώμενος στέγαις, if the dative were taken as quasi-local,
should mean sleeping [more correctly couched, as Sidgwick and others
render it] in the house.’ That is just what it does mean (though Verrall,
who assumes 'roof' in some form or other, disregards this meaning) ; in the
interlinear gloss in Tr it is rightly explained as ἐν τοῖς οἴκοις, and so Mazon
‘au palais’. The plural στέγαις is used as in Ag. 518 and often in the other
tragedians ; the locative dative as in S. Oed. R. 20 ἀγοραῖσι θακεῖ and elsewhere.
Aeschylus is content at this point to say 'in the dwelling of the Atridae'.
This is what the spectator, for whose guidance the prologue-speech is
primarily intended, must be told as soon as possible (the supplementary
information, ἐν Ἄργει, does not come till 24). We are not told explicitly that
the Watchman is on the roof, nor do we need to be :* the spectator sees it for
himself, and the reader, when he finds the Watchman lying at full length and
yet in a position to survey the whole heavens and a part of the horizon,
knows that he must be lying not in a room or courtyard but on the roof.
᾿Ατρειδῶν : Aeschylus makes the two brothers live in the same house.
For the probable reason, see on 400.

1 Cf. J. H. H. Schmidt, Synonymik der griech. Sprache, i. 450 ff., who has collected a large
number of passages (among them Ag. 2) in which κοιμᾶσθαι does not denote sleep itself but
refers to the position of the sleeper or the quality of his couch. But somehow there is
always a connexion with the idea of sleep, as Schmidt points out.
2 Homer says of Elpenor, who lay down on the roof of Circe's house, ἐν δώμασι Κίρκης...
κατελέξατο (x 554 f.) and Κίρκης δ᾽ ἐν μεγάρωι καταλέγμενος κτλ. (A 62).
3
line 3 COMMENTARY
ἄγκαθεν: it is excusable that ancient grammarians should misinterpret
a rare word (κατὰ συγκοπήν, ἀντὶ τοῦ ἀνέκαθεν), but not that Mazon (in a foot-
note to his translation) should champion the error (litt. de longue date’) and
attack the modern editors.’ The right explanation, formerly considered but
not maintained by S. Butler, was given by Hermann. We must start from
the only other passage where ἄγκαθεν is found, Eum. 8o, where ἄγκαθεν
λαβών has the same meaning as the Homeric ἀγκάσ᾽ ἑλών (Paley). Hermann
recognized the connexion of ἀγκάσ᾽ and ἄγκαθεν with ἀγκώνξ (cf. E. Risch,
Wortbildung der homer. Sprache, 1937, 53, 305), and further recalled that
ἀγκών means not only the elbow but the crook of the arm within the elbow:
S. Ant. 1236 f. ἐς ὑγρὸν ἀγκῶνα... παρθένωι προσπτύσσεται and elsewhere.
We can thus understand ἀγκάσ᾽ ἑλεῖν (or a corresponding verb) and similarly
ἀγκάζομαι. As far as the sense goes, therefore, the interlinear gloss on Eum.
8o in the Mediceus is correct: (dyxader) ταῖς ἀγκάλαις (on Ag. 3 Tr has the
gloss ἐν ἀγκάλαις). But it remains surprising that Aeschylus could use
ἄγκαθεν λαβών as synonymous with ἀγκάσ᾽ ἑλών. That ἄγκαθεν is a secondary
formation (Schwyzer, Griech. Gramm. i. 631 n. 5) we can hardly doubt. We
may perhaps conjecture that Aeschylus found it somewhere in later Epic
with the same meaning as ἀγκάσ᾽ or very nearly the same. This conjecture
may derive some support from the remarks of K. Lehrs, De Aristarchi studiis
Homericis, 3td ed., 135 n. 78, on the decay of the suffix -θεν, which began in
Homer and continued in the poetic vocabulary of post-Homeric times, and
especially on the secondary use of ἕκαθεν = ἑκάς and ὑψόθεν = ὑψοῦ (ὑψό-
θεν in this sense is post-Homeric). Lejeune, loc. cit. (see below, footnote 1),
says: ‘comme ἕκαθεν existait a côté de ἑκάς, ἀνέκαθεν a côté de ἀνεκάς, en
vertu d'une analogie toute formelle, ἄγκαθεν a été formé a côté de dyxds.’
Both in Eum. 80 and in Ag. 3 we must recognize the meaning in ulnis (so
Hermann, better perhaps ἐπ ulnas). The Watchman (like the watchdog) lies
as it were thrust forward ‘into his arms’, with the upper part of his body
between them.
κυνὸς δίκην takes up ἄγκαθεν and elaborates it in the way explained on 2:
‘the Watchman is lying flat on his elbows, and in this position he is like a
dog in the act of watching some particular object with his head on his paws.
In this position the Watchman with the least trouble has the widest look-out.
... The simile is exact both in attitude and purpose, as both the man and the
dog are watching, and have their heads between their fore-limbs’ (T. Maguire,
Notes on the Ag. of Aesch., 1868, 6). But the comparison is suggestive beyond
this detail. In Egypt and the Sudan the village dogs prefer to lie at night
on the roofs of the houses or huts even when they are not flat, as Brehm,
Tierleben, i, 2nd ed., 595 f., tells us with picturesque details (I owe this
reference to Beazley and P. Jacobsthal). Cf. too the experience of an R.A.F.
sergeant pilot, who ‘after ramming an Italian fighter over Malta with his
Hurricane baled out’ (Daily Telegraph, 11.xi.1941) and afterwards told the

1 It is due to Mazon's influence that M. Lejeune, Les Adverbes grecs en -8ev (1939), 323 f.,
attempts to justify the identification at Ag. 3 of ἄγκαθεν with ἀνέκαθεν in a manner which
bears the marks of an uneasy compromise; if this excellent philologist had followed his
own judgement, he would probably have avoided this mistake.
2 The linguistic argument for this was given much later by Joh. Schmidt, cf. F. Bechtel,
Lexilogus zu Homer 7; Schwyzer, Griech. Gramm. i. 631 n. 5.
4
 COMMENTARY line 6
reporter: ‘I just opened the lid, and I was out... . As I came down over
the island I saw I was going to land on a village. I hit a wall of a house and
landed on a flat roof right on top of a dog, which let off a terrific howl,
jumped off the roof and bolted up the road.’ ‘Dogs on the roof’ are recognized
by B. Schweitzer, Ath. Mitt. lv, 1930, 107 ff., in the decoration of a Mycenaean
wooden casket.! There is also the characteristic apophthegm of Empedocles
(A. 20a, Diels-Kranz, Vorsokr. i, sth ed., 285 f., [Arist.] Magna Mor. B 11,
p. 1208 b 11) κυνός ποτε ἀεὶ καθευδούσης ἐπὶ τῆς αὐτῆς κεραμῖδος κτλ.
4. κάτοιδα: (Π6 use of this word seems to have been at first confined to
Tragedy (our earliest example is Pers. 744), and thence to have found its
way into the language of Middle (Eubulos 43. 3, mock-tragic style) and New
Comedy. οἶδα and κάτοιδα are completely equivalent in meaning e.g. in
S. El. 922 f. This tragic word also occurs in late dialogues of Plato (only
kareıöws, once in the Sophist and once in the Laws), which is in keeping with
the general character of their style.
5. χεῖμα καὶ θέρος: it has been pointed out by Martin P. Nilsson, Die Ent-
stehung und religiöse Bedeutung des griech. Kalenders, Lund 1918, 22, that the
division of the year into two seasons was not abandoned by the Greeks even
when other seasons had been added. In Prom. 454 ff. the beginnings of winter,
spring, and summer are indicated by the stars.
The reason why the order of θέρος βροτοῖς has been changed in ΕἼΤ is
obvious?. In consequence of a development which began at latest in the first
centuries of our era (cf. P. Maas, Griech. Metrik, $ 21), the last word of a
normal Byzantine trimeter was required to be paroxytone. The alteration
at the end of 1106 (πόλις βοᾶι M, βοᾶι πόλις FTr) is to be explained in the
same way (cf. also the order of 1064 in F, where the mistake was corrected
afterwards). This observation may serve as a warning. For it is not unlikely
that in those sections where our text depends entirely on the authority of
FTr, i.e. in the greater part of the Agamemnon, some similar disturbances
of the original order of words remain undiscovered.
6. δυνάστας is claimed by Headlam as an astrological term, on which Housman
remarks (in his copy of Headlam’s prose translation, kindly lent me by Hugh
Last) : ‘Aeschylus knew no more of astrology than of Christianity or the pox.’?
δυνάστας derives its obvious meaning from what precedes; the stars referred
to (see below) bring round the seasons for mankind, and in this reveal them-
selves as mighty lords. The ancient world found it easy to believe that the
stars by their morning and evening risings and settings (cf. on I. 826) not
only indicate summer and winter but actually bring them (cf. Nilsson, p. 24) ;
the stars are not inanimate objects but living, active beings (cf. C.Q. xxxvi,
1942, 11). At the same time, to describe them as λαμπροὶ δυνάσται develops
with a vivid contrast the conception evoked by ἄστρων ὁμήγυριν : against the
1 He traces the origin of the motif back to Egypt. On this point cf. Axel W. Persson,
*New Tombs at Dendra’ , Skrifter ... Kungl. Human. Vetensk. i Lund, xxxiv, 1942, 179 fl.
2 The wrong assertion that the reading of V is βροτοῖς θέρος passed from the apparatus
of Wilamowitz into the notes of A. Y. Campbell, Murray, G. Thomson. Something similar
has ‚happened on 23, where V (with M) has φάος, and on 246, where the reading of V is
εὐπόταμον (here Campbell has no critical note at all).
3 Cf. W. Capelle, “Älteste Spuren der Astrologie bei den Griechen’, Hermes, Ix, 1925,
373 ff. G. Murray, Aeschylus (1940), 86, is in that context misleading (besides, Aeschylus
does not speak here of the planets).

5
line 6 COMMENTARY
gathered throng of the multitude stand out the leaders and great lords (cf.
Schneidewin ad loc., Pasquali, Stud. It. N.s. vii, 1930, 228). This accurately
expresses what we feel when we look up at the starry heavens; we are aware
of the great glittering throng, paying no attention to the separate stars that
make it up, and we also notice the individuals which are known to us
according to their constellations and some of them also by their own proper
names. The καί then in καὶ τοὺς φέροντας means simply ‘and’, ‘and besides
them’, not ‘and among them especially’ as Housman, J. Phil. xvi, 1888, 246
explains it (for this use of καί he rightly refers to Pers. 749 f., cf. Denniston,
Particles, 291, and see below on $13, at end). For ἐμπρέποντας cf. on 242.
There can be hardly any doubt what manner of heavenly bodies are referred
toin 5f. Blomfield may say: 'Dubium videri potest, num Solem ac Lunam
innuat poeta, quod censet Schutzius, an potius Zodiaci signa’, but the Sun
and Moon are out of the question, and even the second interpretation, adopted
by Housman, loc. cit. (‘the stars of the Zodiac’), seems inappropriate here. We
are in a region where the attitude to heavenly phenomena is very simple,
a world not essentially different in this respect from the world of Homer and
Hesiod; and in consequence the explanation which implies less technical
knowledge of astronomy has more chance of being right. We must therefore
agree with those who understand the passage as Klausen does: ‘dicit Sirium,
Arcturum, Orionem, Pleiades, Hyades, quorum ortus et occasus hiemem et
aestatem hominibus disterminat [cf. Hesiod, Erga 566, 609 f., 615]: ἦν δ᾽ οὐδὲν
αὐτοῖς οὔτε χείματος τέκμαρ οὔτ᾽ ἀνθεμώδους ἦρος οὔτε καρπίμου θέρους βέβαιον
. ἔστε δή σφιν ἀντολὰς ἐγὼ ἄστρων ἔδειξα τάς τε δυσκρίτους δύσεις. Prom.
454 seqq.’ The same view is followed e.g. by Schneidewin, Wecklein, Sidg-
wick, Wilamowitz (in his app. crit. and Glaube d. Hell. i. 261).
7. Since Pauw suspected this line and Valckenaer (on E. Phoen. 506 f.)
rejected it, the strife over its genuineness has never abated. In the dis-
cussion, however, few editors seem to have clearly grasped the principal
difficulty: what part in the whole sentence is played by the temporal clause
ὅταν hbivwow? We must grant that in Stanley’s translation it looks simple
enough: ‘cognovi . . . principes claros, eminentes in coelo stellas, occasus et
ortus ipsarum.' Very similar are the renderings of e.g. Headlam, Mazon
(‘dont je sais et les levers et les déclins), and Pasquali, Stud. It. N.S. vii, 1930,
227 (quando spuntano e quando tramontano"); Pasquali very neatly points
out how Aeschylean is the juxtaposition of ὅταν φθίνωσιν and avroAds, but
like most of the other critics he devotes no attention to the ὅταν. Several
commentators have attempted to support the conventional view with a
grammatical explanation, e.g. Scholefield: ‘xdroida ἀστέρας ὅταν φθίνωσιν
idem est quod κάτοιδα ἀστέρων φθίσιν᾽ and Headlam: ‘the construction ὅταν
φθίνωσιν is idiomatic for watching, observing, marking the time when.' This
assertion is completely unjustified; of the supposed examples adduced by
Headlam, Demosth. 4. 31 is quite irrelevant, since the subordinate clause
beginning with ἡνίκ᾽ dv ἡμεῖς (there follows an optative!) is not the object
of the verb but explanatory to τοὺς ἐτησίας 7) τὸν χειμῶνα, and of his other
parallels, which do consist of interrogative object-clauses, none contains the
ἄν which is precisely what makes his view of Ag. 7 untenable. ὅταν φθίνωσιν,
as Hartung, Karsten, and J. C. Lawson have already stated (they agree in
rejecting the line), cannot stand in an indircct question but can only mean
6
 COMMENTARY line 7
"whensoever they disappear’. Editors who do not overlook this fact or cannot
force themselves to ignore it, and yet regard the line as genuine, are driven
to the most uncomfortable devices. Some, passing over line 6, take orav
φθίνωσιν with φέροντας χεῖμα krÀ.; so Franz Beckmann, Bemerk. z. Prolog...
des Ag. (Braunsberg 1867), 14, Plüss (with marks of parenthesis before and
after 6), and G. Thomson (to judge by his translation). This interpretation,
which apart from anything else compels us to take avroAds τε τῶν = καὶ ὅταν
ἀνατέλλωσιν, was branded as 'absurdum' by Karsten; it assumes that in
spite of its position ὅταν φθίνωσιν can be incorporated with or subordinated
to the clause τοὺς φέροντας χεῖμα «rA., which produces chaos in the syntax,
of a kind unheard of in Greek at any rate of this period (for the principle,
see on 1127). Others, perhaps involuntarily, have slightly distorted the mean-
ing of the verb in order to provide a connexion for ὅταν φθίνωσιν. Thus
Ahrens translates (p. 227): 'und die den Sterblichen Sommer und Winter
bringenden glánzenden Fürsten habe ich kennen gelernt, wenn sie im Aether
prangen und wenn sie in Unsichtbarkeit [he alters ἀστέρας to ἄιστοί 6°]
schwinden und ihren Aufgang.' This is extremely ambiguous; looking to the
end of the translation, one is inclined to understand 'I have learnt to know
the time of their disappearance', which as we have seen would do violence
to the Greek; from the beginning one would suppose it to mean ‘I have
learnt to know them at the time when they disappear'. This, however, would
destroy the parallelism with ἀντολάς re τῶν. What the Watchman says is
this: One whole long year have I been lying here night after night, and now
I know all the stars. Taken by itself it would make good sense to say: I
recognize the stars as they rise, i.e. when such-and-such a star appears above
the horizon, I say to myself "There is Sirius' and so on. But (1) this idea suits
only the rising, and not the setting, (2) κάτοιδα does not mean this,’ (3) this
interpretation like the other abandons the parallelism with avroAds.
The line has been attacked with false arguments of various kinds, so that
we may sympathize with Housman when in hisattempt to rescue the victim of
persecution (J. Phil. xvi, 1888, 245 f.) he recalls the words σύγγονον βροτοῖσι τὸν
πεσόντα λακτίσαι πλέον. For example, the fact that Achilles in his commentary
on Aratus breaks off his quotation with 6 proves precisely nothing.^ Nor
should the use of φθίνειν, unusual though it is, be adduced as an argument
against its genuineness, for Ahrens, referring to Pers. 232, 377, has shown
that Aeschylus could use φθίνειν as equivalent to δύνειν, though it is not
synonymous with it. On other sham charges we need waste no time. A
really serious argument arises from the metre. The point is not, as is often
said, the dactyl in the first foot of the trimeter, but the dactylic word in
that position. From C. F. Müller's dissertation De pedibus solutis etc. (Berlin
1866), pp. 79, 85 f., 93 f., it appears that apart from this only one trimeter
1 The meaning recognize given by L-S s.v. 2 suits none of the four passages there quoted.
2 This has been said by e.g. Hermann, Ahrens, and Mazon. They might also have pointed
out that Achilles shortly before (p. 27. 11 ff. Maass) has copied out the whole passage
A. Prom. 454-60, but omitted 46r μνήμην ἁπάντων, μουσομήτορ᾽ ἐργάνην although it belongs
closely to what precedes.
3 Even by Pasquali, op. cit. 225. W. Schulze, Quaest. ep. 479, on the other hand, is clear
and accurate, only he should not have quoted Soph. fr. 782. 2 N. (= 866 P.) oixérw; in this
tattered fragment, quoted by Plutarch from memory, the word-order is quite uncertain,
cf. Pearson ad loc.
7
line 7 COMMENTARY
in Aeschylus begins with a dactylic word: Cho. 986, with the proper name‘
Ἥλιος (Müller adds fr. 138. 1 N. Avridoy’ ἀποίμωξόν με κτλ., but the elision
there makes a difference). As will be shown in the footnote, it seems possible
to make out the reason why Aeschylus insisted on placing the name “Haws
at the beginning of the line, although its prosodic form did not force
him to do so. In Sophocles we find the same practice: “Hise only,
Aj. 846 and fr. 523 N. (= 582 P.), for μήποτε (Oed. C. 1634) need not be
felt as a single word. (Euripides, on the other hand, has a large number of
miscellaneous dactylic words.) Is it really necessary, in order to save for
Aeschylus a line open to serious objections on grounds of language, to
swallow a metrical phenomenon entirely without parallel in the thousands
of trimeters of the older tragedians that have come down to us? Or ought
we to patch it up at our discretion (for this too has been done) and thus make
it somewhat more acceptable?
Against the rejection of the line Housman (loc. cit.) argues that 'the
Aeschylean archaism τῶν never came from the workshop of an interpolator’.?
To speak of an Aeschylean archaism is wrong, for the anaphoric* use of the
so-called article is often found in later Tragedy as well. Pasquali (op. cit. 226)
in his defence of the line makes a point of the position of anaphoric τῶν at
the end of the trimeter, and compares Eum. 137 and Septem 385, where, how-
ever, it is doubtful whether we should not prefer the reading of the Mediceus
ἐσώ (i.e. ἔσω) with e.g. Wecklein, Wilamowitz, Groeneboom, and Murray.

! It ought to go without saying that in making statements about resolved long syllables,
those proper names which are prosodically awkward must be treated separately. Ceadel,
C.Q. xxxv, 1941, 68 f. has some sensible remarks on the point.
2 [n regard to the position of “Ἥλιος in Cho. 986 and the corresponding position of a
name in several other Aeschylean trimeters, it is worth noting that the passages in ques-
tion show two peculiarities: (1) the proper name is separated by punctuation from the
remainder of the line in which it occurs, and (2) the preceding line contains a word which
is in apposition to the proper name, or (for it does not matter which way we take it)
a word or phrase to which the proper name is in apposition. Aeschylus seems to have
been fond of this arrangement, which enhances the weight of the proper name by placing
it at the beginning of a line, isolating it from what follows, and leading up to it by means
of appositional words. It will be sufficient to quote parallels from the Oresteia only. Ag.
514 Í. τόν 7’ ἐμὸν τιμάορον | 'Ἑρμῆν, φίλον κήρυκα, 1590 f. τοῦδε δύσθεος πατήρ, | ᾿Ατρεύς, προ-
θύμως «rÀ., Cho. 133 f. ἄνδρα δ᾽ ἀντηλλάξατο, | Αἴγισθον, ὅσπερ κτλ, (for the punctuation see
below on 1436), Eum. 6 f. Tiravis ἄλλη παῖς Χθονὸς καθέζετο, | Φοίβη" δίδωσι xrA., 89 f. σὺ δ᾽,
αὐτάδελφον αἷμα καὶ κοινοῦ πατρός, | 'Ερμῆ, φύλασσε, 455 f. πατέρα δ᾽ ἱστορεῖς καλῶς, | Aya-
μέμνον᾽, ἀνδρῶν ναυβατῶν ἁρμόστορα. Of the two Sophoclean examples of dactylic words at
the beginning of a trimeter, the one, Aj. 845 f. σὺ δ᾽, ὦ τὸν αἰπὺν οὐρανὸν διφρηλατῶν, | Ἥλιε,
πατρώιαν κτλ, falls into line with the type described here (cf. especially Eum. 89 f. quoted
above): in the case of the other, fr. 582 P. “Hate, φιλίπποις Θρηιξὶ πρέσβιστον σέλας, the
preceding line is unknown. Of other proper names collocated in the same way I will quote
from one Sophoclean play the following instances : Aj. 462 f., 574 ἴ.», 859 f. To the Aeschylean
instances we may confidently add Ag. 880 f. and Cho. 677 ff., where we find not a mere name
but Lrpodios ὁ Φωκεύς---ἰῆς arrangement, apart from this, corresponding exactly to that of
the other passages. The reason, then, why Aeschylus and Sophocles admitted Ἥλιος, Ἥλιε
at the beginning of trimeters lies probably in the unwillingness of the poets to part with a
favourite syntactical pattern. It would apparently not have satisfied them to put the
proper name in a different place, although by doing so they could have avoided an unusual
form of the first foot.
3 This is adopted by G. Thomson. He is unwilling to credit an interpolator even with
dvroAds for ἀνατολάς, although it is the quite normal poetic form.
+ For the meaning of this term see e.g. Monro, Homeric Grammar, and ed., p. 215.
8
 COMMENTARY line 8
In any case it would be rash to infer from the fact that anaphoric τῶι, etc. is
found in Aeschylus once or perhaps twice at the end of a trimeter, that an
interpolator familiar in general with this use of the ‘article’ in Tragedy might
not accidentally put a τῶν at the end of the line which he was manu-
facturing ; perhaps from metrical necessity (to name only one out of several
possibilities), because he wanted to compress the two complementary clauses
into one line and there was no room for a longer pronoun. But on this point
we can know nothing. I am prepared to admit the possibility that the words
ἀντολάς τε τῶν Come from some unknown Aeschylean context; the one thing
I will not believe is that Aeschylus wrote the whole line as we now have it.
Pasquali has used as a further argument in favour of the genuineness of
L 7 the way in which the riddling expression τοὺς $épovras . . . λαμπροὺς
δυνάστας is followed by the explanation that solves the conundrum, ἐμπρέ-
ποντας αἰθέρι ἀστέρας, this procedure being characteristic of Aeschylus. It
is quite true that in many passages of Aeschylus which have something of a
ypipos about them, every reader must be struck by his anxiety to append an
unambiguous solution, although this runs counter to the nature of the
ypipos and impairs its effect; cf. on 238, 494 f., 825, and elsewhere, and see
also Gilbert Murray, Aeschylus (1940), 62 f. But if 7 is omitted, so that τοὺς
.. . λαμπροὺς δυνάστας ἐμπρέποντας αἰθέρι go together, there is no question
of a ypi$os ; ‘the shining powers that bring to mortals winter and summer,
standing out clear in the sky’ means certain stars with no possible ambiguity,
especially as dorpwv . . . ὁμήγυριν has gone before.
How the interpolation arose we can only conjecture. To me, Valckenaer’s
assumption still seems the most probable, that first someone added ἀστέρας
in the margin to make more clear the meaning of what precedes and that
then somebody else felt tempted to fill this out to a full trimeter. Wilamo-
witz, praef. xxviii, takes the same view. Cf. on 871 and especially on 1226.
The supplement might have been made under the influence of Prom. 457 f.
ἔστε δή σφιν ἀντολὰς ἐγὼ ἄστρων ἔδειξα τάς τε δυσκρίτους δύσεις (compared
by Stanley), especially as a reader of Ag. 5 might well be reminded of Prom.
454 ff. (οὔτε χείματος rékuap . . . οὔτε καρπίμου θέρους). It is also possible that
the whole verse was added to make the meaning clearer. Jachmann, Philol.
XC, 1935, 338 n. 9, thinks it more likely that 7 was intended as a substitute
for 6. But however the interpolator set about his work, and whatever his
models, he did not understand his trade too well, for there can be no doubt
that he meant to say ‘their setting and their rising’.
Undoubted traces of 7 are to be seen on the scrap of papyrus Pap. Oxy.
2178 (probably 2nd century A.D.). The date of this piece of evidence accords
with what is now the current view of the origin of major additions intended
to explain and elaborate the text, viz. that they took their rise as a rule
before the Alexandrian age.
8. Rai νῦν: ‘and so now’, as often, e.g. A 109, 4 12, ὃ 193 (where Ameis-Hentze
comment: ‘xai νῦν introduces the application of the general idea to the
circumstances of the moment’), o 542, A. Cho. 696 (Wilamowitz ad loc. notices
that καὶ νῦν marks the special application of the preceding general sentence).
The general situation, under which the present is ranged with the help of
καὶ νῦν, has been described from 2 onwards. Cf. on 184 kai τότε.
On φυλάσσω = ‘watch for, lie in wait or ambush for, look out for’ (as early
9
line 8 COMMENTARY
as Homer), see L-S s.v. B. 2 and add Aesch. AixrvovAxoi, Pap. Soc. It. xi.
1209 a 3 (D. L. Page, Greek Lit. Pa. i, p. 10) τί σοι φυλάσσω. . .;
λαμπάδος: simply ‘torch’. The vivid image must not be weakened. The
Watchman says 'torch' because he assumes that that is what the beacon will
look like when he sees it in the far distance. Perhaps also the use of the word
here and in 28 (as certainly in 287 and 296) is meant to prepare the audience
for the λαμπαδηφόρων νόμοι in 312.
10. ἁλώσιμόν τε βάξιν : the clause added with τε is epexegetical, as often in
Tragedy ; cf. Denniston, Particles, 502, and see below on 215. He first gives
only the general expression, φάτιν, and then makes it more concrete with
ἁλώσιμον βάξιν. This corresponds in the order of the ideas, though not in the
manner of the addition, to the appositional phrases discussed on 2.
ὧδε yàp κρατεῖ: from Stanley down to Wilamowitz and Mazon many
editors have rendered the phrase by sic enim iubet and the like. This is not,
however, supported by the usage of κρατεῖν. In Passow (followed by Dindorf’s
Thesaurus) this passage is the only one cited for the sense ‘to command’;
L-S, s.v. V, following Wellauer (on Ag. 10), add E. Hec. 282, where even
Hermann (on Ag. 10) supports the false interpretation. But the words οὐ
τοὺς κρατοῦντας χρὴ κρατεῖν à μὴ χρεών contain no sense of iubere, aliquid
imperare; they mean rather 'rulers must not exert a wrongful rule' (ἃ μὴ
χρεών is ἃ cognate accusative like Ag. 1470 f. kpáros . . . κρατύνεις) and the
Scholiast rightly paraphrases: τοὺς ἐν ἀρχαῖς ὄντας kai δυναμένους πράττειν
ἃ βούλονται οὐ χρὴ εἰς κακόν τι τῆι ἐξουσίαι χρῆσθαι. C. G. Haupt was therefore
right when, in his note on Ag. 10, he protested: '«pare? proprie non est
1ubel ; potius: “sic dominatur, potentia sua sic utitur ut hoc ita instituerit’’.’
Similarly Hermann, Enger, Schneidewin, and others. So long as the queen
governs in the way she does, there can be no question of disobedience or
dereliction of duty in any of her subordinates. The sentence gives the reason
why the Watchman has fulfilled night after night for a whole long year a task
which is to him a weariness of the flesh.
For the adverb with κρατεῖ cf. 951 τὸν κρατοῦντα μαλθακῶς (Verrall).
For the expression ὧδε κρατεῖ κτλ. cf. 258, where the coryphaeus says ἥκω
σεβίζων σὸν κράτος.
11. ἀνδρόβουλον: this rare word (Phrynichus, Praep. soph. p. 31. 19 de
Borries, may perhaps refer to this line) was probably coined for this context
by the poet. γυναικόβουλος occurs only Cho. 626. The oxymoron γυναικὸς
ἀνδρόβουλον κέαρ, forceful both in sound and sense, impresses the hearer’s
mind from the outset with one of the principal features of Clytemnestra’s
character. When Sophocles (fr. 857 N. = 943 P.) speaks of an avöpddpwv
γυνή he is presumably following (as he so often does) the example of Aeschylus,
which had a fascination for him.
ἐλπίζον : only this or ἐλπίζων, which comes to the same thing,’ is based on
MS authority. ἐλπίζων is not only the original reading of M and V ; it is also
found in the quotation of the words γυναικὸς... κέαρ by Johannes Siculus?

1 The confusion of -wv and -ov (similarly -ws and -os) is already common in the literary
papyri. For the Mediceus of Aeschylus I give a few examples of -«v instead of -ov (ignoring
passages in which a different sense may be intended, such as Pers. 803, Cho. 822, Eum. 59,
“171, Suppl. 190, 449) : Sept. 444, 869, Ag. 140, Eum. 230, 893, Suppl. 267. Cf. also S. Ay. 452.
2 Johannes Siculus has been generally regarded down to very recent times, even by
IO
 COMMENTARY line 12

(Schol. on Hermogenes 7. ἰδ. i, Rhet. graec. vi. 225. 22 Walz, reprinted in

Vahlen’s edn. of the περὲ ὕψους on 3. 1, p. 5), who lifted the whole section, i.e.
the quotation of Ag. 11 and the following allusion to a passage in the
Oreithyia of Aeschylus, from the φιλόλογοι ὁμιλίαι of the Neoplatonist and
rhetorician Cassius Longinus (cf. Kaibel, Hermes, xxxiv, 1899, 113 f). The
recensio therefore leads us to ἐλπίζον. None the less, all editions from the
princeps down to Stanley read ἐλπίζω, preceded in 10 by κρατεῖν. Stanley
(in his Addenda) raised the question whether the MS reading should not be
preferred, but Elmsley again supported κρατεῖν... ἐλπίζω. His attack on
ἐλπίζον, in spite of Hermann's refutation, has retained its influence to the
present day. I shall not go into the various alterations which have been
proposed, since I believe that the text as it stands is sound. Some editors
have taken éAmifov as part of the predicate (with κρατεῖ), e.g. W. Sewell
(cf. his explanatory note), who translates 'for thus is strong in hope a
woman's heart', and Plüss, who takes it in the same way (Paley wavered).
But the order is against this view (cf. Sidgwick), nor is it recommended by
the meaning of κρατεῖν. On the other hand, it is hardly sufficient to treat
ἐλπίζον merely as an adjective, on the same footing with ἀνδρόβουλον. The
word seems to retain its full participial force, i.e. to denote not so much a
quality of the subject as its situation or behaviour in the circumstances under
which the action (κρατεῖ) takes place. A comparable passage, though with
a slightly more straightforward order of words, is Sept. 52 f. σιδηρόφρων yàp
θυμὸς ἀνδρείαι φλέγων ἔπνει. That ἐλπίζον is used without an object is, of
course, uncommon. ‘Ac fortasse consulto ita loqui putandus est excubitor,
quo innuat aliquid illam amplius quam excidium Ilii sperare. Intellegit enim
excubias eam propterea instituisse, ut ne ex improviso adveniens Agamemno
deprehenderet indicia adulterii (Hermann).
12. εὖτ᾽ ἂν δὲ κτλ.: the construction of the following sentence (destroyed
by Bentley and Wilamowitz) is correctly explained by Schütz: 'totus ille
locus est ὑπέρβατον, ubi oratio excurrit a verbis εὖτ᾽ ἂν δέ usque ad ἐπισκοπου-
μένην, ibi autem insistit, parenthesi φόβος yàp . . . συμβαλεῖν ὕπνωι interiecta ;
deinde constructione illa resumpta ὅταν δ᾽ ἀείδειν etc. usque ad finem decurrit.’
Cf. too Schneidewin and Headlam ad loc. and M. Berti, ‘Anacoluti Eschilei’,
Rendic. Ac. Lincei, Ser. VI, vol. vi (1930), p. 26o, a careful study, which,
however, following Havers, greatly exaggerates the frequency of the so-
called Nominativus pendens and consequently champions numerous cor-
ruptions. Naturally the sequence of ideas in this passage is not taken up
again mechanically in the precise place where the parenthesis broke in upon
it. The thought meanwhile has progressed a stage farther (fear of falling fast
asleep—means of preventing it). So what he is now describing (ὅταν δὲ κτλ.)
is one typical detail of his general situation.’
Radermacher, RE v. 1611, and Wilamowitz on A. Ag. 11, as being one and the same person
as Doxopatres. This error is refuted by Hugo Rabe in the preface to his Prolegomenon
Sylloge (1931), pp. lii and cxiii f., who shows that Doxopatres used and expressly cites his
predecessor Johannes. Rabe has also (cxiv) explained the origin of the mistake. For the
reference to Rabe's statements I am indebted to the kindness of the late Dr. A. B. Poynton,
and also for the chance of inspecting the photograph of codex L (Laur. 57. 5) of Hermogenes
περὶ ἰδεῶν with the Scholia of Johannes Siculus, presented by him to the Bodleian Library.
1 It is amusing to see how the editor of F (or the author of his exemplar) was shocked by
the anacoluthon and tried to get out of it: he put a comma after εὖτ᾽ dv δὲ (the comma is
II
line 12 COMMENTARY
νυκτίπλαγκτον: not ‘nachtumwittert’ (Wilamowitz) and certainly not
merely ‘= νύχιον or something similar’ (Gustav Meyer, Philologus, Suppl.
Bd. xvi, Heft 3, 1923, 99). The sense is better brought out by Blomfield’s ‘qui
noctu excitat vel errare facit’ (so L-S: ‘causing to wander by night, rousing
from bed’). The derivation from the active verb πλάζειν would in itself raise
no difficulty (whether A 308 ἀνέμοιο πολυπλάγκτοιο is to be taken transitively
is not certain). It is well known that verbal adjectives in -ros are often active
in meaning, cf. Wilamowitz on E. Hipp. 1346, Kaibel on S. El. 711 ff., A. C.
Pearson on E. Phoen. 209 and on Soph. fr. 534. 4. So e.g. Ag. 361 πανάλωτος
active, elsewhere δοριάλωτος passive ; see further below on 238. Ernst Fraenkel,
Nomina agentis, ii. 76, and Wackernagel, Syntax, i. 288, call attention to the
fact that originally these verbal adjectives had no direct relation to one
voice of the verb rather than another, and Wackernagel remarks in a similar
connexion (p. 136): ‘these formations are not in the strict sense passive
expressions ; they do not really indicate a state of passivity, but some quality
connected with the idea expressed by the verb’ (cf. also Gerschewitsch,
Stud. It. N.S. xv, 1938, 131 ἡ. But the active meaning which Blomfield gives
the word does not do full justice to the peculiar character of the expression.
It is perhaps more appropriate here to recognize, as we must often do else-
where in Aeschylus, a looser relation between the noun and its highly poetical
compound epithet. An εὐνὴ νυκτίπλαγκτος is a couch which is the scene of
a continuous compulsion to move about by night, νυκτὸς πλάζεσθαι. In simpler
language νυκτίπλαγκτος would probably be applied not to the εὐνή but to
the man who πλάζεται, sleepless like the Agamemnon of the Doloneia, who
says of himself (K 91) πλάζομαι ὧδ᾽, ἐπεὶ οὔ μοι ἐπ᾽ ὄμμασιν ἥδυμος ὕπνος ἑζάνει
krÀ. The Watchman’s meaning is clear. For him at his post there can
be no lying peacefully at rest; no, he πλάζεται, he must jump up to watch
the horizon or he rises in sudden alarm fearful of falling asleep; in a word,
he is ever in movement. The word vuxrimdayrros occurs elsewhere in extant
literature only once in Ag. (330) and twice in Cho. (524, 751) ; it is quite possible
that Aeschylus coined it for this passage in the prologue, and then repeated
it in this and the following play (cf. on 386). In the three later passages
νυκτίπλαγκτος might be taken as active with more justification than here,
but I believe that there too, where it is applied to πόνος, δείματα, κελεύματα,
it is adequately understood as ‘of such a kind that therein (or thereby) a man
wanders in the night-season’.
13f. εὐνὴν... ἐμήν: the wide hyperbaton (Sperrung), the effect of which
is intensified by the overflow into the next line, serves to lay emphasis on
the possessive, ‘bitterly, in contrast to the fortunate sleepers in house and
city’ (Plüss; Verrall too ad loc. rightly drew attention to the function of the
order in adding to the emphasis).' An exact parallel is Suppl. 365 f. οὔτοι
κάθησθε δωμάτων ἐφέστιοι | ἐμῶν. τὸ κοινὸν δ᾽ ei μιαίνεται πόλις KTA., except
that there the counterpart (τὸ κοινόν, πόλις) of the emphatic words ἐμὰ
δώματα is at once expressed, while in Ag. 13 f. the antithesis is left to be
understood. The types of hyperbaton (not only for the sake of emphasis) in
also in G) and added the gloss ὁπότε δὲ φυλάσσω, i.e. he took νυκτέπλαγκτον ἔνδροσόν τ᾽ ἔχω
κτλ, as apodosis.
! So did F. W. Newman, Comments on the Text of Aeschylus (1884), 61, but he thought
ἐμήν to be ‘unendurably emphatic’ and altered it.
12
 COMMENTARY line 17
earlier Greek deserve and would repay careful study. Τῆς import of just the
type with which we are concerned here may be illustrated by at least one
example. In Eur. Alc. 1072 ff., where Herakles is introducing to Admetos
the supposed stranger-lady, he says: ei γὰρ τοσαύτην δύναμιν εἶχον wore σὴν
ἐς φῶς πορεῦσαι νερτέρων ἐκ δωμάτων γυναῖκα. The soul of this sentence lies
in the powerful emphasis with which σὴν γυναῖκα is here pronounced.
14. It is perhaps not necessary to print Φόβος and Ὕπνου, but it probably
accords with the intentions of the poet that we should hear an undertone of
personification in the sentence: instead of the friendly god of Sleep the
Daemon of terror stands by his troubled pillow.' It often proves impossible
in early Greek to make a clear distinction between common noun and proper
name, and we may feel inclined to lament with Otfried Müller (Aesch. Ewm.
p. 166) that ‘we latter-day grammarians find thrust upon us by the capital
letters of proper names the unpleasant duty of having to make up our minds
on the point'. The expression ὀνείροις οὐκ ἐπισκοπουμένην seems to lend sup-
port to the view that ὄνειροι as well as φόβος and ὕπνος are thought of by the
poet as living, divine, or quasi-divine beings. ἐπισκοπεῖν is used of persons,
and especially of deities ; it is applied in S. Aj. 854 to the companion of Sleep,
Θάνατος (cf. also below on 1270 ἐποπτεύσας). “Ονειρος in Homer (B 8) is ad-
dressed as a person ;? in Hesiod (Theog. 21x f.) Νὺξ... τέκε δ᾽ Ὕπνον, ἔτικτε
δὲ φῦλον ' Oveipov ; in the late passage w 12 we read of a δῆμος ὀνείρων. Sappho
fr. 67 D. contains, unfortunately in very bad condition, a prayer (or at any
rate an address very like a prayer) to “Ovopos. For the dative ὀνείροις cf.,
besides the grammars, e.g. the commentators on S. 47. 539. (After writing
the above I see that Schneidewin had remarked on 13 ‘the expression is given
a tinge of personification, as though the Watchman were thinking of οὖλος
"Ovewos', and on 14 ‘comes to my side like some spirit’.)
15. τὸ py... ‘stands by to prevent my closing my eyes in sleep’ (Goodwin,
$ 811, with many parallels) ; cf. on 568.
16. The natural thing to do in such a position. Stephan Bergler has used this
passage to explain Ar. Clouds 720 καὶ πρὸς τούτοις ἔτι τοῖσι κακοῖς φρουρᾶς
ἄιδων (cf. the Scholia there καὶ οὕτως ἦν λεγόμενον ἐπὶ τῶν ἀγρυπνούντων
and Paroemiogr. Gr. ed. Gotting. vol. ii, p. 88 no. 26; on Ag. τό the
axoÀ. mad. in Tr have: τοῦτο κατὰ τὴν παροιμίαν φησὶ τὴν λένουσαν 'bpovpäs
dde’).
μινύρεσθαι: Ατ. Eccl. 880 μινυρομένη τι πρὸς ἐμαυτὴν μέλος.
δοκῶ: ‘I have a mind to . . .' (Linwood) ; so e.g. 1649.
17. ὕπνου ἀντίμολπον . . . ἄκος: not ‘song, sleep's substitute’ (L-S) but
*remedium quod contra somnum cantando fit' (Blomfield). For this function
of ävrı- a parallel was rightly drawn long ago with Horn. h. Cer. 229 ἀντίτομον
and especially Pindar, P. 4. 221 (Medea) σὺν δ᾽ éAaíox φαρμακώσαισ᾽ ἀντίτομα
στερεᾶν ὀδυνᾶν δῶκε χρίεσθαι; there ἀντι- (against) and ὀδυνᾶν are to be taken
together like avrı- and ὕπνου in Aeschylus, and the other part of the compound,
-τομα, ‘produced by cutting’, corresponds to -μολπον, which is used here
because of ἀείδειν and μινύρεσθαι. Similarly in E. Med. 1176 εἶτ᾽ ἀντίμολπον
1 Lewis Campbell translates: ‘for Fear usurps the place of Sleep beside me’; similarly
Headlam.
2 Cf. 7. Hundt, ‘Der Traumglaube bei Homer’, Greifswalder Beiträge zur Literatur- und
Stilforschung, Heft ix, 1935, 44 ff.
13
line 17 COMMENTARY
ἧκεν ὀλολυγῆς (refers to 1173) μέγαν κωκυτόν is rightly explained by the
scholiast : ἀντίφθογγον, ἐναντίον τῆι εὐχῆι. οὐ γὰρ ταὐτὸν μολπὴ καὶ θρῆνος.
évrépvuv: For the understanding of the ‘metaphor’ in general the passages
compared by Stanley and Blomfield are sufficient. Besides Pindar, P. 4. 221
(schol. κατὰ μεταφορὰν τὴν ἀπὸ τῶν ῥιζοτόμων), which has been quoted above,
especially important are E. Alc. 970 f. οὐδ᾽ ὅσα Φοῖβος AoxAnmaddas ἔδωκε
φάρμακα πολυπόνοις ἀντιτεμὼν βροτοῖσιν and Andr. 120 f. ei τί σοι δυναίμαν
ἄκος τῶν δυσλύτων πόνων τεμεῖν (schol. ἀπὸ τῶν ῥιζοτομουμένων ἡ μεταφορά).
Cf. also A. C. Pearson, The Fragments of Sophocles, ii. 172: ᾿ῥιζοτόμος was
applied as well to those whom we should call sorcerers as to the legitimate
followers of Asclepius.’ But on the sense of the particular expression ἐντέμνων
most of the commentators have made no remark. The entry in L-S, ‘cut
in, shred in, of herbs in a remedy’, agrees with Paley’s explanation (which is
followed by Sidgwick): ‘The compound ἐντέμνειν properly refers to the
“shredding in” of herbs in preparing a potion. He may mean “‘putting into
it (i.e. the φρουρά) a remedy ".' Against this Kennedy rightly says (at the
end of his translation): ‘this cannot be proved from any passage of any
author. . . . “Shredding in” is a mere guess invented to account for the word
in this place.” His own unlucky conjecture &v τέμνων won the approval of
Paley and is printed by A. Y. Campbell in his text. Kennedy’s survey of the
meanings of évréuvo (like that in L-S) omits the evidence which for our
passage is decisive; he might have found it in Dindorf’s Thesaurus (right at
the beginning of the article ἐντέμνω) or in Ahrens's careful discussion (p. 228)
of Ag. 17. ἐντέμνειν with the root of a plant as its object means ‘tapping it’
to get the sap out, cf. Theophr. Caus. plant. 6. 11. 14 παρασκευάζουσιν αὐτοὺς
(scil. rods ὀπούς) τὰ μὲν éyyuAilovres . . . τὰ δὲ ἐντέμνοντες, 6. 11. 15 ἔστι δὲ καὶ
τῶν μὲν καὶ ἄνω καὶ ἐν ταῖς ῥίζαις ἡ ἐντομή. . . καὶ γὰρ ἡ ῥίζα καὶ ὁ καυλὸς
ἐντέμνεται, cf. too Theophr. Hist. plant. 9. 1. 61., Pliny, Nat. hist. 13. 124
(radix) incisa lacte manat, 20. 254 e radicibus prima germinatione incisis.
Although Ahrens referred to this technical use, he yet preferred the reading
ἐκτέμνων, which has inferior authority: ‘If we are to defend the reading
ἐντέμνων. . . we must suppose that Aeschylus took the expression in an
unusual way from this process of tapping sap; the expression ἄκος ἐντέμνειν
would of course be a very bold one, because the remedy is not incised but
obtained by an incision.’ Ahrens surprisingly overlooked the fact that
precisely this bold usage is characteristic of the style of the tragedians. Cf.
Matthiä, Griech. Gramm. § 409. 5 n. 1; he quotes among other passages S. A7.
55 ἔκειρε πολύκερων φόνον (on which the late Scholia have: ἤγουν διὰ τὸ κείρειν
ἐποίει), Ant. 972 ἕλκος τυφλωθέν, E. Suppl. 1205 τρώσηις φόνον. Among these
phrases ἄκος ἐντέμνειν, ‘to produce a remedy by incision’, falls perfectly into
place. The phrase is bold enough, and when ὕπνου ἀντίμολπον is added, the
expression becomes even further removed from simple speech. A certain
tendency to the διθυραμβῶδες is noticeable elsewhere in this prologue (cf.
II, 12 f. νυκτίπλαγκτον . . . ἔχω εὐνήν, 33): in style it is as much the τηλαυγὲς
πρόσωπον of the majestic trilogy as the solitary utterance of a man of the
common people.
19. διαπονουμένου : although the word seems not to occur elsewhere in poetry,
it should not be tampered with, as it gives an excellent sense. ὧδε οὖν τὸ
πεδίον φύσει καὶ ὑπὸ βασιλέων πολλῶν ἐν πολλῶι χρόνωι διεπεπόνητο, says Plato
14
 COMMENTARY line 22
(Critias 118 b) of land which in addition to its natural qualities had been
‘admirably husbanded’ by a long process of the appropriate διακόσμησις.
The toilsome, unending, and successful activity that goes to foster the well-
being of land or household could not be more accurately indicated; it was
this fostering care that in the house of the Atridae had given place to
heedlessness and neglect.
20 repeats verbatim the emphatic expression of 1, and marks thereby the
conclusion of this part of the speech. For this manner of enclosing as it were
in a frame see Wilamowitz, Arist. u. Athen, ii. 335, and in his commentary on
the Choephoroe p. 149; cf. also on 1196 below and on 1611.
21. ὀρφναῖος occurs in Homer (only in the Doloneia and once in the Odyssey)
exclusively in the formula νύκτα δι᾿ ὀρφναίην, and elsewhere too is always an
epithet of night or something dark. Here, on the other hand, it is boldly
used of the beacon that appears in the darkness. Cf. Schuursma, 39, 168.
Between 21 and 22 stands a scholion! δεῖ διαστήματος ὀλίγου, εἶτα dvaxpa-
yeiv, ws θεασάμενον τὸν πυρσόν. In terms of stage direction: ‘Iacuerat cubito
nixus excubitor. Nunc viso igne exsilit, postquam paulum tacuerat’ (Her-
mann).
The device of an interval is also employed in the prologue of the Eumenides
(after 1. 33) ; there the effect is intensified by the contrast in the movements
of the priestess, who is composed as she enters the temple and horror-
stricken as she comes out of it again.
22. ὦ χαῖρε. The ὦ in this phrase (cf. Porson on E. Or. 470 [476]) seems to
belong to the language of ordinary life. To the examples in Tragedy, here and
A. Suppl. 602 ὦ χαῖρε πρέσβυ, S. Aj. 91 ὦ χαῖρ᾽ Adava, El. 666 ὦ χαῖρ᾽ ἄνασσα,
E. Med. 665 ὦ χαῖρε καὶ σύ, παῖ σοφοῦ Πανδίονος, Αἰγεῦ, Hipp. 1453, Heraclid.
574, 660, Her. 523 ὦ χαῖρε μέλαθρον πρόπυλά τε κτλ., El. 1334 (at parting)
ὦ χαῖρε πόλις, Hel. 616, 1165, Or. 477 ὦ χαῖρε καὶ σύ, may be added from
Comedy Ar. Ach. 872, Knights 1254, Peace 523, 1357, Birds 1586, Lys. 853, 1097
&* χαίρετ᾽ ὦ Λάκωνες, Menander Γεωργός 41 ὦ χαῖρε πολλά, Μυρρίνη. Con-
versational use οἱ ὦ with the imperative is indicated also by passages like
Ar. Frogs 269 à? παῦε παῦε, Lys. 1271 ὦ δεῦρ᾽ ἴθι, δεῦρο (δεῦρο... was...
felt as an imperative’, Wackernagel, Syntax, i. 71; on the use in the
Lesbian poets Lobel observes, Ἀλκαίου μέλη, p. xliv: ‘ôeüpo and its plural
δεῦτε... are imperatives’), 1304 ὦ εἶα κοῦφα πᾶλον. To the same class belong
passages like E. Cycl. 52 ὕπαγ᾽ ὦ ὕπαγ᾽ ὦ κεράστα, El. 112 (= 117) ὦ ἔμβα
ἔμβα (Soph. fr. 58 N. = 61 P. must be omitted as too uncertain ; in fr. 410 N.=
448 P. the lacuna should not be ignored, cf. Pearson ad loc.). Another mode
of expression derived unmistakably from the habitual language of daily
ı Wilamowitz’s note (under Actio) ‘pausa post 21 (monuit Tr)’ is incorrect, since the
scholion in question is much older than Triclinius: it is found in M, in Tr (where it is clearly
marked as one of the σχόλια παλαιά), and in F.
2 The editors, even Wilamowitz, accentuate ὦ, I do not know why. The Ravennas has
ὦ. Assuming that there is any point at all in distinguishing ὥ in exclamations from ὦ
before a vocative (for the evidence of ancient and Byzantine grammarians v. L-S s.v. 4),
then logically the ὦ which precedes an imperative and goes closely with it must be treated
like ὦ before a vocative. It is certainly not an independent exclamation.
3 Here too editors since Dindorf accentuate ὦ against the ὦ of the MSS, and the same is
true of Lys. 1271 and other passages. This arbitrary practice is perhaps partly responsible
for the fact that the article ὦ in L-S gives no idea how widespread is the usage under
discussion.

15
line 22 COMMENTARY
life is the insertion of a formula of adjuration between & and the imperative:
S. Aj. 371 ὦ πρὸς θεῶν ὕπεικε, Oed. R. 646 ὦ πρὸς θεῶν πίστευσον, Οἰδίπους,
τάδε, 1037 ὦ πρὸς θεῶν... φράσον, Ar. Eccl. 970 σὺ δέ μοι, φίλτατον, ὦ ἱκετεύω
(equivalent to ἃ πρὸς θεῶν), ἄνοιξον. Instead of πρὸς θεῶν we find quite a long
adjuration inserted in front of the imperative in E. Hypsipyle fr. 60, col. τ.
25 ff. ὦ πρός σε γονάτων ἱκέτις Ἀμφιάρεω πίτνω καὶ πρὸς γενείου τῆς τ᾽ Ἀπόλλωνος
τέχνης" καιρὸν γὰρ ἥκεις τοῖς ἐμοῖσιν ἐν κακοῖς" [p]icat με. Sometimes it may
happen that only the words ὦ πρὸς (τῶν) θεῶν are actually uttered, and
the imperative which was to follow is either dropped by the speaker himself
(since it could be understood from his preceding utterance) or left un-
spoken as another person breaks in. This is the case in Ar. Lys. 857, where
failure to understand the usage? has largely interfered with the appreciation
of a delightful scene. Right at the beginning (850) Kinesias has said to Lysi-
strata πρὸς τῶν θεῶν νῦν ἐκκάλεσόν μοι Muppivny; to the compliments of Lysi-
strata (853-7) he pays not the slightest attention, but in his burning im-
patience repeats the same demand: ὦ πρὸς τῶν θεῶν." Lysistrata, however,
cunningly pretends not to understand what he wants, and emphasizes again
what she has already said; whereupon the unhappy Kinesias for the third
time (861) entreats her: ἴθι viv κάλεσον αὐτήν. In a question taking the place
of an imperative ὦ πρὸς τῶν θεῶν is found in Ar. Wasps 484 dp’ ἂν ὦ πρὸς τῶν
θεῶν ὑμεῖς ἀπαλλαχθεῖτέ μου; It has fossilized into a formula (perhaps λέξον,
λέξατε, or something of the sort is to be understood) in passages like Ar.
Plut. 458, 1176, Alexis fr. 87. 3 K., Demosth. 9. 15 ὦ πρὸς τοῦ Aids and similarly
14. 12, 21. 98 ὦ πρὸς τῶν θεῶν. Whether in A. Cho. 942 ἐπολολύξατ᾽ ὦ (rightly
separated by Seidler) and in the similar passage E. Tro. 335 βοάσαθ᾽ Ὑμέναιον
ὦ the ὦ is to be taken with the preceding imperative or, as seems to me more
likely, should be regarded as the exclamation ὦ, it is hard to say. The fact
that ὦ can be used in exactly the same way with both vocative and impera-
tive is one more indication how close together the two are in function as
well as form (cf. Wilhelm Schulze, Kuhns Zeitschr. 111, 1924, 105 = Kl. Schriften
423).
λαμπτήρ: this word does not occur in true Attic (Comic Poets, Orators,
and inscriptions) ; cf. Ernst Fraenkel, Zdg. Forsch. xxxii, 1913, 112. Tragedy
probably took it over from the Odyssey. As the agent noun of λάμπειν it
is a general word for anything that gives light. In the Odyssey it is probably
a brazier or pan filled with dry wood and used as a ‘lamp’, as it is ex-
plained, after the ancient lexicographers, by e.g. Ebeling, Lex. Homericum, 969,
and Daremberg-Saglio, i. 873; it can also denote the lamp in a house (A. Cho.
537), a watch-fire in a camp (5. Aj. 286), a lantern (as early as Empedocles
! Similarly Ar. Birds 661 ὦ τοῦτο μέντοι νὴ Al’ αὐτοῖσιν πιθοῦ, where Coulon and O.
Schroeder wrongly print à.
2 Wrongly rendered e.g. by Chrestien (‘Florens Christianus’ ; his translation is reprinted
in Küster’s edition): ‘Dii vostram fidem’, and by J. G. Droysen: ‘wirklich? OY; similarly
Wilamowitz : ‘Er spricht zweifelnd, wie die Antwort lehrt.’ Rather differently, but equally
wide of the mark, Rogers: ‘The words are an ejaculation of love and pleasure drawn from
his excited passions by Lysistrata’s gratifying intelligence.’ On the other hand, Bothe,
whom Rogers attacks, quite rightly saw that what Kinesias is going to say when Lysistrata
interrupts him is something like ὦ πρὸς τῶν θεῶν ἐκκάλεσον αὐτήν.
3 We may compare E. Hipp. 219, where πρὸς θεῶν must not be connected with the fol-
lowing affırmative sentence (Wilamowitz’s translation is wrong), but takes up the request
of 215 πέμπετέ μ᾽ εἰς ὄρος.
16
 COMMENTARY line 24
fr. 84. 3; this is later the commonest meaning, so much so that it was taken
over by the Romans before the beginning of extant literature, perhaps
through the medium of the Etruscans, cf. Walde-Hofmann, Lat. etymol.
Wörterbuch, 761, G. Devoto, Storia della lingua di Roma, 128). Here no speci-
alized meaning is intended, as for instance Headlam’s ‘Thou blessed lantern!’
This does not suggest the picture of a beacon with its leaping flames. Less
inappropriate, but still not right, is the translation current since Stanley,
fax, ‘torch’, ‘Fackel’. That is of course the correct rendering of λαμπάς
(used by the Watchman in 8 and 28); but λαμπτήρ is not a mere synonym
of λαμπάς, but rather an alternative to it with the double advantage of a
more lofty sound, as a non-Attic word, and a meaning not limited to any
specific light-giver. The German word ‘Leuchte’ with its poetical colouring
combines both advantages in the same way.
In what follows there are two ways of construing νυκτός, The paraphrase
in the scholia (MF), which may go back to antiquity, runs: ἐκ νυκτὸς (for this
idiomatic expression cf. L-S s.v. νύξ i. 2) ἡμέραν ἡμῖν διδούς (so rightly F,
δοῦσιν M). νυκτός is here taken with πιφαύσκων as meaning ‘by night’. On
the other hand, the σχόλια παλαιά in Tr, ὦ λαμπτὴρ τῆς νυκτὸς χαῖρε, 6 διδοὺς
ἡμῖν φωταυγῆ ἡμέραν, indicate punctuation after νυκτός. There is after νυκτός
a stop (or colon) in V, a comma in F. Turnebus punctuated after λαμπτήρ,
Victorius after νυκτός. The latter punctuation was defended vigorously by
Hermann, the former by Ahrens (p. 229) and others, rightly in my opinion:
‘this punctuation is urgently required in view of the two precisely ana-
logous passages quoted by Wellauer, Ag. 522 ἥκει γὰρ ὑμῖν φῶς ev εὐφρόνηι
φέρων and Pers. 300 f. φάος μέγα καὶ λευκὸν ἦμαρ νυκτὸς ἐκ μελαγχίμου (both
metaphorical). The forcefulness of the Aeschylean expression, which illus-
trates the magnitude of the deliverance by the contrast between the light
and the darkness which prevailed before it, is sadly weakened by the punctua-
tion commonly accepted [i.e. λαμπτὴρ vuxrés,].” Schneidewin and Ahrens are
right in assuming that φάος here conveys the well-established secondary
meaning ‘salvation, deliverance’ (cf. on 522), although naturally the literal
meaning is also present.
23. φάος. The interpolated version νῦν φῶς may serve as a warning of the
dangers which lie about us when we have nothing but FTr to go upon.
πιφαύσκων: the word belongs to Epic and has passed thence into the
lyric poets; in Drama, as Schneidewin has already remarked on this passage,
it is so far recorded from Aeschylus only.
23 f. χορῶν κατάστασιν πολλῶν: in a Greek community of this period this is
the natural way to celebrate a piece of good fortune or a success achieved.
Cf. (Keck) E. Alc. 1154 f. ἀστοῖς δὲ πάσηι τ᾽ ἐννέπω τετραρχίαι χοροὺς ἐπ᾽
ἐσθλαῖς συμφοραῖσιν ἱστάναι. After the unhoped-for return of Herakles and the
resulting μεταβολὰ κακῶν the Chorus (E. Her. 763) sing χοροὶ χοροὶ καὶ θαλίαι
μέλουσι Θήβας ἱερὸν κατ᾽ ἄστυ.
24. év"Apye. For the problem of the dwelling-place of the Atridae see on 400.
C. P. Bill, Transact. and Proceed. Amer. Philol. Assoc. lxi, 1930, 114 ff., has
undertaken the thankless task of showing that when Aeschylus speaks of
Argos he means not Argos but Mycenae. Wilamowitz, on E. Her. 15, takes
the view that ‘Aeschylus in the Oresteia avoided the name Mykenai because
Athens was then on friendly terms with Argos (cf. the pertinent utterances
4872-2 ς 17
line 24 | COMMENTARY
Eum. 289 ff., 670 ff., 762-74], and the destruction of her famous rival [Mykenai]
had taken place only a few years before’.
συμφορά, in itself neither good nor bad but used most frequently of un-
favourable events, is not very rare in a favourable sense. In the Oresteia
Aeschylus so uses it five times.
25. ἰοὺ ἰού: on the accentuation of the MSS here and elsewhere, cf. L-S s.v.;
Körte, Hermes, Ixviii, 1933, 273 τι. 4.
26. σημαίνω M: σημανῶ rell. The continuation in 31 αὐτός τ᾽ éywye . ..
χορεύσομαι seems at first glance to support σημανῶ, which for this reason is
read by Hermann, Wecklein, Wilamowitz, and others (e.g. G. Pasquali,
Storia della tradizione, 27). But the passage undeniably gains in force if he
says ‘with this loud cry of mine (ἰοὺ ἰού) I give Agamemnon’s lady a clear
sign to arise... .' (so rightly Peile and Paley). If we take it in this way,
the asyndeton is given its full force. A confirmation may be found in the
precisely parallel passage 1315 f. ἰὼ ξένοι. οὔτοι δυσοίζω, θάμνον ws ὄρνις,
φόβωι, where Cassandra’s passionate outcry is immediately followed by the
explanation, like the Watchman’s ἰοὺ ἰού here. Another close parallel is
the explanation of the cry in Prom. 66 αἰαῖ, Προμηθεῦ, σῶν ὑπὲρ στένω πόνων.
The clause αὐτός 7’ ἔγωγε φροίμιον χορεύσομαι is not attached to 26 but to
27 ff.; see on 31. For the corruption to σημανῶ cf. 1267.
27. On ἐπαντείλασαν the scholia have ὡς ἐπὶ ἄστρου ἢ σελήνης. One can see
how this explanation is suggested by the predominant use of the verb. On
E. Phoen. τος ἀπὸ κλιμάκων ποδὸς ἴχνος ἐπαντέλλων there is a similar scho-
lion ἀπὸ τοῦ ἡλίου ἡ μεταφορά; on this Wecklein rightly remarks that the
transitive use of the word does not support this interpretation, and in fact
one can say without hesitation that it is impossible. Similarly here it is quite
arbitrary to suppose an astronomical sense. In Aeschylus this sense is not
inseparable from the word, as may be seen from Cho. 282, where ἐπαντέλλειν
is used just like ἀντέλλειν in Sept. 535. I cannot therefore agree with Keck:
‘the queen is, so to speak, the sun rising upon the house’, Headlam: εὐνῆς
ἔπαντ. is a reverent phrase, suggested by a comparison with the rising of the
sun or stars’, A. S. F. Gow, Journ. Hell. Stud. xlvii, 1928, 137 n. 12, who
renders it ‘dawns’ and regards the expression here as a more modest variation
of Pers. 150. In this passage, when the λαμπτήρ has just been addressed as
νυκτὸς ἡμερήσιον φάος πιφαύσκων, it would be most inappropriate in the next
breath to compare the queen with the rising sun. Wecklein’s note is purely
comic: ‘the Watchman’s mind is still full of his observation of the stars.’
27 f. Wecklein : δόμοις goes with ἐπορθιάζειν as in 1119... . τῆιδε λαμπάδι goes
with εὐφημοῦντα.᾽ But it is natural to take τῆιδε λαμπάδι as depending on
ἐπορθιάζειν. On the other hand, it seems inadvisable to understand δόμοις
with most editors as a locative, for the locative dative of common nouns is
normally found with verbs of tarrying, sitting, lying, laying down, and so on,
although collocations like E. Phoen. 931 ff. θαλάμαις . . . σφαγέντα are found
here and there. Karsten rightly says: 'ópow . . . domui, ut domum quasi
participem faciat gaudii sui. . . . Dativus λαμπάδι pendet a praepositione
verbi ἐπ-ορθιάζειν.᾽ So van Heusde and Verrall.
28. ὀλολυγμός is found as a cry of good omen in association with εὐφημεῖν
in 595 f. and Ar. Peace 96 f. On this and on ὀλολυγμός in general see S.
Eitrem's article (cited on 597), p. 47.
I8
 COMMENTARY line 32
29. It goes without saying that the true reading here is ἐπορθιάζειν, found in
two other passages of Aeschylus, not the ἐπορθριάζειν of MV, which is found
nowhere else. ὄρθριον is read for ὄρθιον by two MSS in Pers. 389.
30. ἑάλωκα (‘from FeFdAwxa, cf. ἑόρακα’ Passow-Crönert) is correctly formed;
cf. W. Schulze, Quaest. ep. 385 τι. 3, O. Schroeder on Pindar (ed. mai.) P.
3. 57. Pap. Oxy. 2178 may have contained a variant, but all that is preserved
is traces of the first two letters: ‘ea is excluded’, 'πείπτωκεν cannot be veri-
fied’ (Lobel). That ἑάλωκεν is genuine can hardly be doubted.
πρέπει: cf. ON 242.
31. +’. The conjecture δ᾽ (Blaydes) betrays an insufficient understanding of
the context. The ritual thanksgiving which Clytemnestra is to inaugurate
with the ὀλολυγμός leads in its ordinary course to a χορῶν κατάστασις. ‘And
in connexion therewith’ (if one may thus show by exaggeration the force of
re) ‘I myself will dance by way of prelude.’ The middle χορεύσομαι should
not be forced to yield a special meaning, as it is by Schneidewin: ‘He will
dance for himself (middle)’, and Headlam: 111 lead off myself with a dance
upon my own account.’ L-S correctly say ‘Med., in same sense’ ; the passages
which they quote in addition to Ag. 31, viz. E. Jon 1084 and Ar. Thesm. 103,
refer to dancing by more than one person.
In the nineteenth century it became the fashion to assume that the Watch-
man while speaking these words executed a few dance-steps on the roof.
This was first put forward, so far as I can make out, by Peile, who was
followed by Conington (‘And I will dance the prelude here myself’), Paley,
Schneidewin, Kennedy, etc., down to Wilamowitz, Headlam, and A. Y.
Campbell. The possibility that Aeschylus intended this cannot of course be
disputed, but I do not see that the text makes this assumption necessary,
or even likely. If the Watchman meant ‘I propose to dance here and now’,
one would expect him to give his reasons (32) in the present tense. But as
θήσομαι refers beyond a doubt to a later point in time, it is natural to suppose
that the same applies to χορεύσομαι. When he comes into the queen’s pre-
sence, before she can set on foot the ὀλολυγμός and the χορῶν κατάστασις
(on φροίμιον MFTr have the gloss: πρὸ τῆς Κλυταιμήστρας) he will dance as an
introduction his own dance of joy. He will thus celebrate the moment when,
with the bringing of his tidings, he rids himself at length of his heavy burden
and gets his share in the good fortune of his lord and master.
32. Anyone surveying the discussions of this line might well come to the
conclusion that this passage is one of those in which Aeschylus has expressed
himself in dark and riddling speech. Yet here the blame appears to lie en-
tirely with the commentators (1.6. in the last resort with Triclinius) ; to the
Athenian audience the sentence, which deals in familiar phraseology with
familiar things, must have been completely unambiguous. There is no longer
any doubt that εὖ πεσόντα go together, and that it is wrong to take ed alone
with θήσομαι, as used to be done. The problem of the line therefore centres
in the question whether θήσομαι is to be understood by itself or with εὖ
πεσόντα added as a predicate. The oldest interpretation known to us is the
gloss in MF on θήσομαι: οἰκειώσομαι. We need not waste any time over this
rendering ; it is clearly nothing more than an unlucky guess, made without
any knowledge of the technical meaning of θέσθαι which is here implied.
Undoubtedly the author of the gloss intended εὖ πεσόντα to be taken as
19
line 32 COMMENTARY
belonging to τὰ δεσποτῶν, not as supplementary to θήσομαι. The same inten-
tion is probably expressed by the comma in F after πεσόντα (seemingly in the
scribe’s hand, to judge by the amount of space left for it). Triclinius, on the
other hand (who marks this scholion as his own by the note ἡμέτερον), says:
διὰ τῆς ἡμετέρας εὐφροσύνης καὶ τῆς χορείας δείξω Kal ἐγὼ τὰ συμπεσόντα νῦν
τοῖς δεσπόταις ws καλῶς ἔπεσον καὶ ἐγένοντο. A somewhat similar interpreta-
tion was advanced by Stanley: ‘res enim heriles ut bene cedant efficiam’,
Hermann (Opusc. v. 342): ‘nam salva res herorum uti sit fecero’, Klausen
and others down to Wilamowitz: ‘efficiam ut bene cecidisse erae res videan-
tur.’ This view would in any case have little probability as completely
ignoring the technical meaning of θέσθαι (see below) applied to dice-play ;
but a further decisive argument against it is the use of the middle voice.
That this idea would demand θήσω will be clear to anyone who examines
in any of the larger lexica the material for τίθημι, especially — facto, efficio ;
the examples from Aeschylus confirm the general impression. Plüss did
justice to the middle: ‘ed πεσόντα predicative to θήσομαι; he intends to
make the assumption [cf. L-S s.v. τίθημι B. ii. 1] that the masters’ game is won’,
but this thought is too complicated to be probable. The true meaning was first!
expressed by Blomfield in his glossary; he gave adequate evidence, and one
wonders that his explanation did not win the day once for all. The result and
the relevant authorities are now to be found in many commentaries (also in
Pearson’s note on Soph. fr. 947) and in L-S s.v. τέθημι A. vii. 2: ‘in the game
of werreia, kuBeia . . . to place as skilfully as possible the pieces which have
been assigned to one by the luck of the dice’. The reference is to the game
(or perhaps more correctly to one of the group of games; for details see
Lamer, RE xiii. 1967 ff.) which consists in a combination of dice-throwing
and board-game: the two players move their counters on a board, but
the extent of each move is determined by a previous casting of the dice.
English and French scholars (e.g. Lafaye in Daremberg-Saglio, v. 127) com-
pare this and the similar games among the Romans with their own back-
gammon (French trictrac)*2 To move one's counter is τίθεσθαι τὴν ψῆφον
(the middle is here as clearly demanded as in τίθεσθαι τὴν ψῆφον of recording
one’s vote, cf. on 816); the full expression will have been θέσθαι τὴν ψῆφον
πρὸς τὰ (ἐκ) πεσόντα, but in the technical language of the players this seems
to have been shortened at an early stage to θέσθαι τὰ (ἐκ) πεσόντα. This
is clear, leaving our passage on one side, from Soph. fr. 861 N. (= 947 P.)
στέργειν δὲ τἀκπεσόντα Kai θέσθαι πρέπει σοφὸν κυβευτήν. The more complete
expression is to be found in Plato, Rep. 604 c ὥσπερ ἐν πτώσει κύβων πρὸς τὰ
πεπτωκότα τίθεσθαι τὰ αὑτοῦ πράγματα. There is probably an allusion to the
technical term in Thuc. i. 25. 1 ἐν ἀπόρωι εἴχοντο θέσθαι τὸ mapóv.? In real life

1 The connexion of θέσθαι with dice-playing had been noticed by Hemsterhuys on Lucian,
Nec. 21 (vol. i. 486 of the edition of 1743), but he was only prepared to recognize εὖ θέσθαι.
He therefore altered the text of the decisive fragment of Sophocles, and in Ag. 32 took εὖ
with θήσομαι.
2 Encycl. Brit.: ‘Backgammon, a game played with draughtsmen and a special board,
depending on the throw of dice. . . . The men are moved on from point to point, according
to the throws of the dice made by the players alternately.’ See there for further details.
3 θέσθαι is explained by Classen-Steup much more inaccurately than by Poppo and
K. W. Krüger. Precisely the same phrase occurs in Lucian, Nec. 21 τὸ παρὸν εὖ θέμενος ;
Thucydides himself says in 4. 59. 4 τὰ γὰρ ἴδια ἕκαστοι εὖ βουλόμενοι δὴ θέοθαι κτλ. To
20
 COMMENTARY line 33
one makes one’s moves according to what one has thrown oneself. In the
Watchman’s case it is his masters’ throw, but as he is compelled to share in
the game and the result has a decisive effect on his own position, it is possible
for him to speak in language used elsewhere by the player himself. His own
concern in the matter appears also in the μοι of 33. The move he proposes to
make (θήσομαι) is to go at once to Clytemnestra and give her news of the
appearance of the beacon. Then he will receive a handsome reward (though
probably not such an enormous one as his prototype in Homer, 8 525 f., the
hireling of Aegisthus, was at any rate promised). His most valuable recom-
pense, however, will be the certainty that never again need he undertake
his hated watch. This lucky throw enables him to make a move on the board
which he is quite sure will win the game (cf. on 33). Some commenta-
tors have found deliberate ambiguity in the sentence, so e.g. van Heusde
and in a different sense A. Platt (note to his translation): ‘This phrase is
designedly ambiguous; the Watchman means the success of Agamemnon,
but the words may also mean the success of the plot of Clytemnestra.' E.
Med. 54 ἴ. χρηστοῖσι δούλοις ξυμφορὰ τὰ δεσποτῶν κακῶς πίτνοντα was compared
long ago by Casaubon. If that is a reminiscence of our passage (Verrall), it
would show that Euripides took ra δεσποτῶν εὖ πεσόντα together just as we
have seen it is necessary to do.
It goes without saying that εὖ πεσόντα points forward to the xußeia which
receives full and vivid expression in 33. The scholia quote as a παροιμία
the trimeter dei γὰρ εὖ πίπτουσιν οἱ Διὸς «Bow! ascribed by Schol. E. Or. 603
to Sophocles (fr. 809 N. = 895 P.); the line is also given without author’s
name in the paroemiographers and lexicographers. Cf. Wilamowitz on E.
Her. 1228.
33. τρὶς €€ βαλούσης. Part of the evidence from literature and the lexicogra-
phers is given by Blomfield; for fuller references see Lafaye in Daremberg-
Saglio, v. 126, and Lamer, RE xiii. 1955. In antiquity the game was norm-
ally (and in the earlier period apparently always) played with three dice (cf.
Lamer, 1943). The oldest and best illustration of the expression τρὶς ἐξ βαλεῖν
—and vice versa—is the painted clay model of a gaming-board acquired in
Athens and now in the National Museum in Copenhagen. This has been
published by J. L. Ussing, Vidensk. Selsk. Skr., ste Række, hist. og Phil.
Afd. ste Bd. iii, Copenhagen 1884, 149 ff., with two good illustrations (plate I),
and discussed by Blinkenberg, Ath. Mitt. xxiii, 1898, 8, and (with a repeti-
tion of the latter’s poor sketch of the upper side) by Deubner, Die Antike,
vi, 1930, 171, 175 (cf. also Lamer, RE xiii. 1996). Beazley, who has examined
the original, tells me that the piece is ‘Attic on stylistic grounds, time second
quarter of the 6th century’. On the surface of the board are engraved nine
lines parallel to the shorter sides; the ends of these lines (apart from two
ends now destroyed) are occupied each by an oval counter, making eighteen
in all. Outside the lines to right and left lie two dice, the upper face showing
the 6. In the middle are traces of a third die now lost, and Blinkenberg has
rightly assumed that on this too the 6 lay uppermost. He has also drawn the
what extent a metaphor is felt in the passages from Thucydides and in others where
εὖ θέσθαι appears (cf. below on 913), it is impossible to discover; L-S are right.
1 Apparently Διὸς crept from the scholion into Eustathius’ quotation of Ag. 33 (see
app. crit.).
21
line 33 COMMENTARY
convincing conclusion that what is represented here is the ‘lucky throw’,
the τρὶς ἕξ, which enabled a player on his next move to occupy with his men
all the eighteen points at the ends of the lines, and so to win the game. What
corresponds to this last move (θήσομαι) in the Watchman’s game, we have
already seen.
To say of the φρυκτωρία that it τρὶς ἐξ ἔβαλε is very bold; cf. on 17.
34. δ᾽ οὖν: breaking off and passing on to something new (cf. Denniston,
Particles, 461 {.), making clear at the same time that the new idea rests on the
considerations that have already been put forward; cf. on 255, 1568.
koAövros: cf. on 675.
εὐφιλής first occurs here, and elsewhere only in Eum. 197. Similarly
δυσφιλής first in the Oresteia (Ag. 1232, 1641; Cho. 624, 637, 1058; Eum. 54),
then S. Oed. C. 1258 (probably a reminiscence of Ag. 1641, see note there);
κοινοφιλής (a certain! restoration) only in Eum. 985. On the other hand,
θεοφιλής (Eum. 869) occurs a little earlier; besides Aeschylus fr. 350. 3 N.
(date unknown) cf. Pind. Isthm. 6. 66 (probably 480 B.c.), Bacchyl. 11 (ro).
60. The first instance of προσφιλής is Septem 580. In the AukrvovAxoi, Pap.
Oxy. 2161, col. 1. 31 (anapaests), Aeschylus ventures the coinage ποσθοφιλής.
For the formation of these and similar fifth-century words see R. McKenzie,
C.Q. xiii, 1919, 148.
35. ἄνακτος οἴκων: cf. on 1225 (δεσπότης).
βαστάσαι: the mistake, fashionable down to our own time (‘touch’ L-S,
following Passow), of misinterpreting the word in this and similar passages
can be traced back to Schiitz and Blomfield ad loc. They, and others in their
train, e.g. Ellendt, Lexicon Sophocleum, Jebb on Oed. C. 1105, mutilated the
explanation preserved by Suidas: βαστάσαι où τὸ ἄραι δηλοῖ παρὰ τοῖς Arrıkois,
ἀλλὰ τὸ ψηλαφῆσαι καὶ διασηκῶσαι καὶ διασκέψασθαι τῆι χειρὶ τὴν ὁλκήν, by
quoting only as far as ψηλαφῆσαι. If one takes the interpretation of the
lexicographer as a whole, it is quite adequate : βαστάζειν means not a desultory
touching, grasping, or taking hold of an object (here of the hand and forearm)
but the holding and poising of it, e.g. for careful examination, as in Homer
$ 405 ἐπεὶ μέγα τόξον ἐβάστασε καὶ ἴδε πάντηι.2 The Watchman describes the
δεξίωσις, the affectionate gesture at the moment of greeting a friend, after
a long absence, when one holds his right hand and does not quickly let it go.
This is not the same as ‘shaking hands’ (Wilamowitz ‘zum Willkomm schüt-
teln’). Admetos (E. Alc. 917) on his wedding day enters his house φιλίας
ἀλόχου χέρα βαστάζων. Attic grave-reliefs show what is meant. Stanley was
right then in translating μὲ reducis . . . manum domini . . . hac sustineam
manu. No passage in Tragedy makes it necessary to assume the supposed
ı When Verrall, while accepting Hermann’s xowod«Aet, maintained that κοινοφελεῖ was
‘not impossible’ and that ‘it would be formed correctly from κοινός and ὄφελος᾽, he forgot
that it ought to be κοινωφελής, see Wackernagel, Dehnungsgeseiz, Ὁ. 50.
2 On 5. Phil. 656 f. dp’ ἔστιν dore κἀγγύθεν θέαν λαβεῖν καὶ βαστάσαι (the bow) Rader-
macher rightly refers to the Odyssey passage, but none the less glosses βαστάσαι with
ψηλαφῆσαι. From the sense ‘to hold an object while examining and weighing it’, the
extension to abstract things follows without any difficulty, as in A. Prom. 888 ἐν γνώμαι
τόδ᾽ ἐβάστασε (where the scholiast comments ἐδοκίμασεν and very intelligently quotes
¢ 405), Ar. Thesm. 438 (with the scholion in Suidas s.v. ἐβάστασεν, vol. ii. 189 Adler: avri τοῦ
ἐδοκίμασεν), Eupolis fr. 73, 303 K., Polyb. 8. 16. 4 (with the gloss in Suidas s.v. βαρυδαίμων,
i. 455 Adler : ἀντὶ τοῦ διεσκέπτετο).
3 Rightly understood also by F. A. Wolf, cf. W. von Humboldt to Hermann, 17.xii.1815
22
 COMMENTARY lines 36 f.

meaning of mere contrectare; Philoctetes' bow of destiny is no pocket pistol,

and for the meeting of Oedipus and Antigone (Oed. C. 1105) we may well
borrow Goethe's words 'dich halten dieser Arme Schranken' (cf. too Prom.
1019 πετραία δ᾽ ἀγκάλη σε βαστάσει).
Lines 34 f. are the more moving because the spectator knows, or at least
suspects, that in spite of the master’s homecoming the wish is destined not
to be fulfilled.
36 f. βοῦς ἐπὶ γλώσσηι μέγας βέβηκεν : the parallels were collected long ago:
Theognis 815 βοῦς μοι ἐπὶ γλώσσηι κρατερῶι ποδὲ λὰξ ἐπιβαίνων ἴσχει κωτίλλειν
καίπερ ἐπιστάμενον, Aesch. fr. 316 N. ἀλλ᾽ ἔστι κἀμοὶ κλὴις ἐπὶ γλώσσηι φύλαξ,
Soph. Oed. C. 1051 ff. (of the Mystai at Eleusis) θνατοῖσιν, ὧν καὶ χρυσέα κλὴις
ἐπὶ yAwooaı BéBake προσπόλων Εὐμολπιδᾶν, Strattis fr. 67 (i. 731 Kock) βοῦς
ἐμβαίνει μέγας. The attempt at an explanation (βοῦς = money) found in
Pollux 9. 61, in the Paroemiographi (Diogenianus 3. 48, i. 223 Leutsch-
Schneidewin; cf. the passage quoted in their note), in Hesychius s.v. βοῦς
ἐπὶ γλώσσηι (in the second place), and in Schol. BD on Homer ® 79, is fatuous.
The phrase βοῦς μέγας βέβηκε indicates in the vivid manner of popular
exaggeration! the oppressive force of a heavy weight ; the idea of being paid
to hold one's tongue is ruled out both here and in Theognis by the context.
What we have is clearly a widespread proverbial phrase ; and that the epithet
in βοῦς μέγας (characteristic of common speech; cf., in another context,
Callim. Epigr. 13. 6 Wil.) belongs to the original expression is made probable
by Menander's parody in the Adıns (fr. 21) παχὺς γὰρ ὗς ἔκειτ᾽ ἐπὶ στόμα.
βέβηκεν, which is found in the passages of Theognis, Aeschylus, and Sophocles,
undoubtedly forms part of the traditional phrase ; in Sophocles it is obviously
transferred from the ox to the key. Perhaps it is permissible to infer with
Schneidewin from the passage in Sophocles that this vigorous expression for
keeping silence belonged to the ritual of Eleusis,? and it is possible that
κλήις was substituted for Bods (so Ahrens, p. 232) to tone down its violence.
A parallel to this from the same class of ideas would be εὐκαμία, which is
cited as Doric in the Et. Gen. and has recently turned up in Sophron (cf.
Latte, Philol. Ixxxviii, 1933, 467; the text is now available in D. L. Page,
Greek Lit. Pap. i. 330); in parts at least of the motherland this may have
been the word originally used in the ritual? and have been replaced by
εὐφημία ἔστω when the gag (on κημός see on 238) came to sound too crude.*
The clause βοῦς ἐπὶ γλώσσηι κτλ. takes the form of an asyndeton because it
explains the fact of his silence by giving the reason which compels it (Schneide-
win rightly says: ‘Bods ἐπὶ γλώσσηι βέβηκεν is added to give the reason for
(W. v. H.s Briefe an G. Hermann, mitget. von A. Leitzmann, Weimar 1929, p. 32) : 'Baeráca:
χερί versteht er [Wolf] nicht vom Geben, oder Berühren mit der Hand, sondern vom
Unterstützen mit dem Arm.’ G. Hermann, however (Opusc. v. 342), says: ‘regis . . . caram
manum tandem hac manu contingere.'
* The ‘great ox’ is the biggest and heaviest creature known to ancient people. When
the Romans in their campaigns in south Italy first met the elephant, they called it Luca bos.
2 Cf. also G. Thomson, ad. loc. I see no reason for resorting to a formula from the
Cretan mysteries (with A. B. Cook, Zeus, ii. 345).
3 On I 171 εὐφημῆσαι, which is quite unparalleled in Homer, see Leaf, ad. loc., and
Wilamowitz, Ilias und Homer, 247 n. 1.
4 After writing the above, I see that Lobeck, Aglaophamus, i. 36, long ago established a
relation between εὐκαμέα and the formula βοῦς ἐπὶ γλώσσηι, and also linked both with the
mysteries of Eleusis.
23
 lines 36 f. COMMENTARY
the silence’). For this explanatory asyndeton cf. on 1261 and 1284, and also
G. Thomson on 36 f. (who, however, limits it unduly to one special type).
37. εἰ φθογγὴν λάβοι. λαβεῖν is is the verb used in all the passages with the
same motif collected by Peile and others: S. El. 548 φαίη δ᾽ ἂν ἡ θανοῦσά γ᾽, εἰ
φωνὴν λάβοι, E. Iph. T. 52 (στῦλος εἷς ἔδοξέ por...) φθέγμα δ᾽ ἀνθρώπου λαβεῖν,
Plato, Prot. 316 ἃ καί μοι δοκεῖ ἡμῶν ἡ ἄρτι ἔξοδος τῶν λόγων ὥσπερ ἄνθρωπος
κατηγορεῖν... καὶ εἰ φωνὴν λάβοι, εἰπεῖν ἄν κτλ. Cf. also Babrius 71. 5f. ἡ
θάλασσα... γυναικείην λαβοῦσα φωνὴν εἶπε. It is possible that φωνὴν λαβεῖν
is the regular phrase developed in the realm οἱ fairy story and folk-tale for
miraculous happenings of this kind. For a similar notion cf. E. Hec. 836 ff.
εἴ μοι γένοιτο φθόγγος ἐν βραχίοσι καὶ χερσὶ καὶ κόμαισι καὶ ποδῶν βάσει ἢ
Δαιδάλου τέχναισιν 7 θεῶν τινος κτλ.
39. What people are meant by μαθόντες ἢ On this point most commentators
remain silent. Gilbert, in the second edition of Enger’s commentary, says:
‘Il. 36-9 are addressed to the audience’ ; the same view is held by W. Nestle,
‘Die Struktur des Eingangs in der att. Tragödie’, Tübinger Beitr. z. Altertumsw.
X, 1930, 21, 106. A compromise was made by Plüss in his re-edition of Enger-
- Gilbert’s commentary: ‘36. The following words are accompanied by a
gesture . . . and addressed to absent participators in the events (these being
represented by the audience).” From what we know of Attic tragedy, any
thought of turning to the audience seems out of the question. The right
solution was given by F. Leo, ‘Der Monolog im Drama’, Abhdl. d. Göttinger
Ges. d. Wiss., N.F. x, no. 5, 1908, 8 n. 1: μαθοῦσιν αὐδῶ refers of course to the
imaginary listener (‘den als möglich gedachten Hörer’), not to the audience
in the theatre’. It is perhaps inherent in the act of speaking that whenever
a man utters words aloud, even though he is alone and speaking primarily
to himself, he has the feeling of addressing other persons.
κοὐ μαθοῦσι λήθομαι: first correctly understood by Blomfield.’ Cf. Hdt.
4. 75. 1 ὁ δὲ τῶν μὲν δὴ ἐκεῖνοι προσεδέοντο αὐτοῦ, τούτων μὲν ἑκὼν ἐπελήθετο,
ἀρξάμενος δὲ ἀπ᾽ ᾿Αχαιμένεος ἐγενεηλόγησε κτλ., 4. 43. 7 ἔχων χρήματα μεγάλα,
τὰ Σάμιος ἀνὴρ κατέσχε, τοῦ ἐπιστάμενος τὸ οὔνομα ἑκὼν ἐπιλήθομαι (cf. too
3. 147. I τουτέων μὲν τῶν ἐντολέων μεμνημένος ἐπελανθάνετο). These passages
- make it quite clear that in the passage in the Agamemnon, as we should have
inferred from the word-order and the symmetrical arrangement, ἑκὼν must
be connected with λήθομαι as well as with αὐδῶ. The expression ἑκὼν (ἐπι-)
λήθομαι preserves a valuable piece of early psychology. If the question is
(as in Hdt. 4. 43. 7 already quoted) how far a man’s silence can be trusted,
one has a far stronger guarantee if he knows the perilous secret and deliber-
ately ‘forgets’ it than if he is merely unwilling to utter it or by chance cannot
call it to mind.
The last sentence (36-9) is compact of emphatic expressions drawn from
the stores of common speech. On βοῦς ἐπὶ γλώσσης μέγας βέβηκεν and ἑκὼν
λήθομαι see above; the forceful personification of οἶκος δ᾽ αὐτὸς εἰ φθογγὴν
1 Stanley's translation ‘non intelligentes lateo'—followed by e.g. Schütz, W. v. Hum-
boldt, F. A. Wolf (reprinted in W. v. Humboldt’s Ges. Schriften, Akademie-Ausgabe,
viii. 230), Hermann, Opusc. v. 342—survives in astonishing fashion in Wilamowitz's ren-
dering : ‘den nicht Wissenden ein Rätsel ist’. Verrall quite arbitrarily says: ‘Ay8opa: is here
the passive answering to the active λανθάνει με τοῦτο, “I do not observe this”.’ A. Platt,
too, is very much open to question : “for of mine own will are my dark sayings dark to the
unlearned’,

24
 THE PROLOGUE AS A WHOLE
λάβοι, σαφέστατ᾽ ἂν λέξειεν stands on the same level. The man’s excitement,
his resentment at the doings of which he has so long been a mute and reluctant
spectator, and at the same time his fear of landing himself in trouble by
a thoughtless word against the ruling powers—all these find utterance in
expressive images and turns of phrase. The syntactical form of the four
concluding lines is in remarkable contrast to all the speech that has gone
before. Down to 35 there is a broad and even flow of sentences, some of
them built into periods and expanded by subordinate clauses; in the last
four lines the ear is arrested by brief κόμματα, fired off as it were in suppressed
passion. There is no question of rhetoric here, even in the antithesis and echo
of the concluding line; all is perfectly natural. It is worth noticing that at
this early stage of Attic literature the poet shows himself capable of expressing
strong emotions not only through the words but also by the form of his
sentences.
THE PROLOGUE AS A WHOLE
Each play of the Oresteia begins with a prayer by the προλογίζων (in the
Suppliants it is the Chorus who recite the opening prayer). During Orestes’
prologue in the Choephoroe Pylades is on the stage. The prologues of the
Agamemnon and the Eumenides, on the other hand, are real soliloquies. It is
characteristic of the Aeschylean conception of a monologue (cf. Prom. 88 ff.)
that the speaker does not address his own heart or mind but turns to the
gods or to divine beings; cf. F. Leo, Der Monolog im Drama, 7; W. Schade-
waldt, Monolog und Selbstgespräch, 52. The prologues of the Agamemnon
and of the Eumenides, which have some important features in common
(cf. Leo, op. cit. 8, and see on 21), are both delivered by a προτατικὸν πρόσω-
πον. In using this device Aeschylus, as Leo pointed out (p. 115), seems to
have followed the model of Phrynichus, in whose Phoenissae (cf. arg. A.
Pers.) the prologue beginning with τάδ᾽ ἐστὶ Περσῶν τῶν πάλαι βεβηκότων
was spoken by the eunuch who was στορνὺς θρόνους τινὰς τοῖς τῆς ἀρχῆς
παρέδροις, that is to say by a simple servant engaged in his daily work, a
figure very similar to the Watchman in the Agamemnon.
Whatever models may have influenced the form of this prologue, its con-
tent makes it appear as a work of high originality and indeed one of the great
masterpieces of dramatic speech. Seldom has monologue proved so singularly
appropriate a form of expression. Loneliness itself, protracted and tormenting
loneliness, seems here to have found its voice. Not once are we reminded, as
is so often the case with later prologues, of the fact that a prologue is pri-
marily a means of preparing the mind of the audience for what is to follow
by introducing them to, or reminding them of, facts and circumstances funda-
mental to the evolution of the plot. It could not occur to us to think of
the Watchman as a mere προλογίζων. It is true he disappears for good after
his one speech, but so long as he is on the stage he seems to be there, as it
were, in his own right, and to speak not as the dramatist’s mouthpiece but
for his own sake. In his limited sphere this character is perfectly presented.
A plain and honest man’s reactions to a monotonous and toilsome task, his
sufferings, his frustrations and hopes, are spread out clearly before us. So
real are all the elements of his troubles and the objects of his ponderings,
from the hard and damp couch on the roof to the bright stars overhead, from
25
 COMMENTARY
the. misgivings caused by the threat overhanging his master to the feeling
of immense personal relief when the period of exacting duty is at last over;
so strongly does everything the Watchman says grip our mind that weare
forced to put ourselves in his place. We cannot look on him as a subordinate
character, having ἃ merely preparatory function: while he is there, he 15 to
us a fellow man, a fellow sufferer.
At the same time the main objective of a dramatic prologue is firmly,
though unobtrusively, maintained. There is first (1. 9 f.) the expectation of
the capture of Troy. Then momentous words (1. 11) give the full measure of
Clytemnestra's gigantic figure. Soon afterwards (1. 18 £) the evil that is
brewing in the house of Agamemnon is hinted at in an impassioned sentence,
the thought of which is taken up and intensified in the concluding lines of
the prologue (36 ff.). Thus from the outset that sombre note is struck which
is soon to become a keynote of the whole play; the obsession of inescapable
doom begins to work on the hearer’s mind. Finally, the prologue discloses
an important aspect of the man who is the true centre of this tragedy. What-
ever Agamemnon’s faults in other respects may be, to the people of his
household he is a kind master, whom they not only respect but love.’ The
affectionate feeling and deep devotion in 1. 34 f. should not be disregarded.
The Watchman is longing for the king’s return with a fervent desire worthy
of Eumaeus. The first mention of Agamemnon in the play is bound to stir the
sympathy of the audience in his favour. This detail, more than anything else,
makes it clear how much the poet gained by substituting the faithful servant
of Agamemnon for the bribed hireling of Aegisthus (8 524 ff.).

THE PARODOS
Wilamowitz, in his note on 1. 39, says: ‘Custos in domum descendit. longior
pausa. The same view was held by Verrall. As far as the performance in
the Athenian theatre is concerned, we do not know how long an interval, if
any, separated the entrance of the Chorus from the exit of the Watchman.
The point is of no consequence. What does matter is the unmistakable fact
that in the imaginary time of the action (as distinct from the time of the
performance) a considerable interval must be assumed between the last words
of the Watchman and the first of the Chorus. In the meantime Clytemnestra
has received the message and dispatched her servants with orders for sacri-
fices on all the altars of the town. Cf. on 83 ff. An Athenian audience would
not have found it difficult to comply with such a simple demand on their
imagination. Cf. pp. 254 ff., on the time elapsing between the first part of the
play and the arrival of the Herald.
It seems natural to assume that it is during this interval that the sun is
supposed to rise. While the Watchman is delivering his speech it is night;
in 1. 279, on the other hand (cf. 264 f.), the morning is indicated as the time
1 It may be well to combine the Watchman’s attitude towards Agamemnon with the
phrase Cho. 54 f. σέβας δ᾽ ἄμαχον ἀδάματον ἀπόλεμον τὸ πρίν, which is rightly explained by
the scholiast : ἑκούσιον σέβας τὸ μὴ ἐκ φόβου, ἀλλὰ ἐξ αἰδοῦς αὐτοῖς γενόμενον... ἡ αἰδώς, ἣν
περὶ ᾿Αγαμέμνονος εἶχον οἱ δῆμοι, νῦν εἰς φόβον ἐτράπη. ἐκεῖνον γὰρ ἠιδοῦντο καὶ ἐφίλουν, τὸν
δὲ φοβοῦνται κτλ.
2 Verrall, Introd. liii, and Wilamowitz, Interpr. 163, worked out in detail what they
believed to happen on the stage during the interval. Wilamowitz's view, based on a wrong
interpretation of 88 ff., was rejected by Kranz, Hermes, liv, 1919, 301 n. 3.
26
 COMMENTARY line 41
at which the dialogue takes place. Wilamowitz, it is true, makes the end of
the lyrical part of the parodos coincide with the moment of daybreak, but
his assumption is based on a misinterpretation of 254, cf. ad loc.
40-103. The Chorus march in, singing (or reciting) anapaests, and get into
position for the delivery of the subsequent song. This preparatory movement
requires here and in the Persae a remarkably long time: the Chorus probably
walk several times round the orchestra before they take their stand (cf.
W. Kranz, Stasimon, 147). This form of the parodos, in which anapaests
precede the song in lyric metres, seems to be the oldest; it probably goes
back to the period before the invention of the prologue (which is, however,
at least as old as the Phoenissae of Phrynichus). We find this bipartite form
in those plays of Aeschylus which have no prologue, Supplices, Persae, and
Myrmidones ;' it follows the prologue in the Agamemnon, the Ajax, and the
Alcestts.
40. δέκατον μὲν Eros: here too the tendency to begin a speech with μέν (on 1)
may play its part; but it is also clear that the thought of the endless war and
the sufferings of both sides is for the Elders only a preliminary to what they
wish to express about themselves: 72 ἡμεῖς δέ. To give an account of one’s
own position is one of the principal themes of a parodos. μέν is intended not
to emphasize δέκατον but to indicate the relation of the first section (40-71)
to what follows: decimus quidem hic belli annus est . . . sed nos senes domi
mansimus. The first idea is spun out at length, but at the same time the
original intention is maintained.
41. ἀντίδικος : apart from the doubtful fragment Heraclitus B 133 Diels-
Kranz this seems to be the earliest occurrence of the word. There is, however,
little doubt that as early as the date of the Oresteta it had become, as it is in
Antiphon, part of the technical vocabulary of the Athenian courts; the verb
ἀντιδικεῖν is found in Ar. Clouds 776. It is important not to content oneself
like the scholiast (ἀντὶ τοῦ ἐχθρὸς ἀπὸ τῶν ἐν ταῖς δίκαις ἐχθρῶν, cf. Hesychius
ἀντίδικος" ἀντίπαλος, ἐχθρός, ἐναντίος) With a colourless rendering (‘meaning
simply antagonist’ Sidgwick, ‘generally [as opposed to the technical sense]
opponent, adversary’ L-S), but to grasp the unimpaired force of the legal
term (so rightly W. Sewell, at the end of his translation, with the excellent
comment ‘the legal metaphors, so profusely employed in this play, must be
strictly attended to’, Wecklein ‘Prozcssgegner’, Headlam ‘Priam’s great
accusing peer’; cf. also Daube 97). The word must have sounded strange to
the audience in this context, for it was by no means an accepted idea, nor
indeed a particularly obvious one to regard the war against Priam as an
action at law. But that is exactly what Aeschylus did, and he pursued the
consequences of his view into the most unexpected details (534 ff., 813 ff.).
So he uses the juristic word in a prominent place, right at the beginning of
the entry-song of the Chorus, to give the whole play the colouring which
essentially belongs to it.” As the legal claim is in the first instance the concern
of the man who has been primarily wronged, Menelaus is put first here. But
! That the play began with the anapaests of the Chorus (fr. 131 N., Pap. Oxy. 2163,
fr. 1) Τάδε μὲν λεύσσεις κτλ. is certain; that these anapaests were followed by the lyrics
(fr. 132) Φθιῶτ᾽ ᾿Αχιλλεῦ κτλ. has been made highly probable by G. Hermann, Opusc. v. 137.
2 The frequent occurrence of δίκη in the Orestera (as compared with the rest of Aeschylus
plays) was pointed out by Peretti, Stud. It. N.S. v, 1928, 202.
27
line 4τ COMMENTARY
the apposition to Πριάμου μέγας ἀντίδικος is twofold, Μενέλαος ἄναξ 78°
Ἀγαμέμνων, so that ἀντίδικος assumes a ‘collective’ meaning, or, to put it
differently, takes the place of a dual (see on 115). That Agamemnon is to be
regarded as a ‘claimant at law’ against Priam to the full extent, no less than
Menelaus, is a leading theme of the play, expressed with special clarity in
the first speech of the Herald and the King’s own words on entry.
43 f. διθρόνου Διόθεν καὶ δισκήπτρου τιμῆς ὀχυρὸν ζεῦγος ᾿Ατρειδᾶν:
both genitives depend on ζεῦγος, and both are in different ways epexegetical ;
first the ζεῦγος, the ‘pair’, is identical with the Atridae, and secondly the
ζεῦγος, the ‘two coupled together’, constitute the double kingship, the
δίθρονος καὶ δίσκηπτρος τιμή. On two genitives with one noun, cf. Wilamowitz
on E. Her. 170.
44. ὀχυρόν: for the variation in the spelling of the first syllable, ὀχ- and ἐχ-,
see the dictionaries. While in Pers. 78 thereis equally strong MS autho-
rity for ἐχυροῖσι and ὀχυροῖσι, in Pers. 89 the testimony of the MSS is perhaps
slightly in favour of exupois. Therefore Dindorf, Lex. Aesch., suggested that
in Ag. 44, too, we might write éyvpóv. This was put in the text by Wilamo-
witz. It does not seem possible to make out which was the spelling of
Aeschylus. óyvpós is considered by Gustav Meyer, Griech. Grammatik, 3rd ed.,
40, to be the older form.
That Arpaôäv should not be altered to Arpeidaw is shown by Wacker-
nagel, Unters. zu Homer, 57, who points out that in the language of the older
post-Homeric poets the two brothers are invariably referred to with the
plural of the patronymic; the change had already been rejected by Ahrens
(p. 234), H. W. Smyth, Harvard Studies in Class. Philol. vii, 1896, 148, and
A. Cuny, Le Nombre duel en grec (Paris, 1906), p. 134 n. 2. For the ending
-ἂν of a proper name in anapaests see on 1569.
45. χιλιοναύτης' ‘consisting of a thousand ships’ may well have been invented
by Aeschylus, but is just as likely to have occurred in post-Homeric epic ;
οἵ. Pers. 83 πολυναύτης. There is no difference in meaning between it and
χιλιόναυς, which is found several times in Euripides; cf. W. Schulze, Quaest.
ed. 445 n. τ; Ernst Fraenkel, Nomina agentis, i. 24. The commentators
observed long ago (cf. Schol. E. Or. 353 and Schol. E. Andr. 106, Eustathius
on B 760, p. 338. 35) that in Euripides and elsewhere in the poets (e.g.
Lycophr. 210, Plaut. Bacch. 928) 1,000 is. the usual number for the Greek
fleet that went to Troy, while the figure derived from the Homeric Catalogue
of Ships is 1,186, for which Thucydides (1. 10. 4) gives 1,200 as an approxima-
tion. Varro, rust. 2. 1. 26 numerus non est ut sit ad amussim, ut non est cum
dicimus malle naves 1556 ad Troiam (quoted by Stanley).
47. ἦραν: στόλον atpew here and Pers. 795 is probably best taken like τὰς
ναῦς αἴρειν (Thuc. r. 52. 2, ἀπαίρειν Hdt. 8. 57. 2); πόλεμον αἴρεσθαι seems to
be not quite so close (middle in A. Suppl. 338, Hdt. 7. 132. 2, Thuc. 4. 6o. 2,
Plato com. fr. 107 K., but active in E. El. 2 ὅθεν ποτ᾽ ἄρας ναυσὶ χιλίαις
Ἄρη «rA.; Ar. Birds 1188—paratragic—has πόλεμος αἴρεται in the passive).
ἀρωγήν: van Heusde well says ‘dpwyds et ἀρωγή vocabula iuris', referring
to Hom. Z soz, Ÿ 574; 'dicitur στρατιῶτις ἀρωγή Menelai, qui Priami ἀντίδικος
erat.' Cf. more recently Ferrari, La parodos, 356; Daube, 98. Naturally
1 For the hand by which the variant γρ(άφεται) Dorv αὐτὰν is written in the Mediceus cf.
Rostagno's introduction to the facsimile, p. 12.
28
 COMMENTARY lines 48 ff,

‘vocabulum iuris’ requires some qualification. In the scene of litigation on

the Shield of Achilles Z 502 λαοὶ δ᾽ ἀμφοτέροισιν ἐπήπυον, ἀμφὶς dporyot the
subject is 'the crowd of the people who with acclamation lend support now
to one party now to the other, and therefore are called helpers in the strife
(dpwyoi) (Lipsius, D. att. Recht i. 4, cf. also ἃ. Glotz, La Solidarité de la
famille, 293); Leaf, ad loc., and R. J. Bonner and G. Smith, The Administration
of Justice, ii. 189, take the word in a sense rather more definitely technical.
Similarly in V 574 ἀρωγή is used half technically of taking sides in an arbitra-
tion. Aeschylus was familiar with the special meaning of 'support given in
a court of law’, as is shown by Suppl. 726 ἐγὼ δ᾽ ἀρωγοὺς ξυνδίκους θ᾽ ἥξω
λαβών. Here in Ag. 46 we have both the military conception and the idea of
a lawsuit ; the expedition is at the same time a demand for legal redress.
στρατιῶτιν ἀρωγήν is in apposition to the sentence or (what comes to the
same thing) to the action of the verb (so Hermann, Schneidewin, Paley,
Plüss). The syntax of this has been well explained by G. Bernhardy,
Wissensch. Syntax der Griech. Sprache, 127, and is discussed at length by
A. Matthiä, Griech. Gramm. 910 f., who puts in its proper setting Ag. 226
(which shows the same use of ἀρωγάν), and by Wilamowitz on E. Her. 59.
48 ff. The editors have rightly remarked the influence on Aeschylus here
of two Homeric similes, ΠῚ 428 f. ὥς 7’ aiyvmol . . . μεγάλα κλάζοντε μάχωνται
and s 216 ff. κλαῖον δὲ λιγέως, ἁδινώτερον ἢ τ᾽ οἰωνοί, φῆναι ἢ αἰγυπιοὶ γαμ-
ψώνυχες, οἷσί τε τέκνα ἀγρόται ἐξείλοντο πάρος πετεηνὰ γενέσθαι. The principal
motif clearly comes from the Odyssey passage, but there the birds only lament
in pain; the element of war-cry is contributed by the /liad. But apart
from the noting of these parallels, hardly anvthing has been done to explain
the passage. Most commentaries are quite silent on κλάζοντες. Schütz re-
marks rather lamely: ‘magnum ex animo Martem clangentes, h.e. bellum et
caedem spirantes. ... xAdfovres videtur pertinere ad comparationem subse-
quentem.' This view was developed still more unsuccessfully by Schneide-
win, who explains that the normal Ἄρη πνεῖν is here replaced by κλάζοντες ‘for
the sake of the comparison with ὀξὺ βοῶντες αἰγυπιοί᾽. But this would deprive
the simile of any basis in the description of the event itself; precisely the
detail which furnishes the link with the simile would be foisted on the passage
'for the sake of the comparison'. It does not make things better if we take
κλάζοντες Ἄρη of the typical βοή of Epic (‘battle-cry’ Verrall, and similarly
Murray; 'Schlachtruf' Wilamowitz), for a battle-cry is completely out of
place in the situation expressed by στόλον ἦραν. The right interpretation
is rather that given by Paley: ‘the Atridae cry war! and call for vengeance
for the rape of Helen, as vultures fly screaming round their eyrie when their
young have been taken away’, and Mazon: ‘ils criaient la guerre.” κλάζοντες
Ἄρη ‘they cry “War!” ' as in Cho. 402 Bods Aovyós "Epwóv, E. Phoen. 1154 f.
βοᾶι πῦρ καὶ δικέλλας, Propertius 4. 8. 58 vicinas Teta clamat aquas (see
P. Maas, Kuhns Zeitschr. liii. 312). The general ancient practice of adapting
the content of the cry to the construction of the sentence is discuss»d by
Lehrs, Quaest. ep. 325, and others, then by O. Schroeder, Comm. on Pind.
Pyth. 6. 36 (δονηθεῖσα φρὴν βόασε παῖδα öv; a close parallel which Schroeder
might have mentioned is Bacchyl. 17 (16). 14 ff. βόασέ 7’ ᾿Ερίβοια χαλκοθώρακα
Πανδίονος ἔκγονον: ‘she cries in her anguish "Theseus!" , cf. also W.
Schulze, ΚΙ. Schr. 91 n. x. In regard to the subject-matter, it is worth
29
lines 48 ff. COMMENTARY
noticing how Aeschylus’ imaginary picture of the summoning of the host
in heroic times contains an echo of immemorial custom. The extent and
significance of the call to arms, the clamor armisonus, the ‘geriifte’, has been
penetratingly analysed by W. Schulze, Berl. Sitzgsber. 1918, 484 ff. (= ΚΙ.
Schr. 163 ff.) ; he starts from the cry for help of the man suddenly attacked,
but points out that the cry for help and call to arms merge into one another
and arma clamare could be used for both, so long as σιδηροφορεῖν was still
customary. In general Aeschylus like the other tragedians uses βοή for the
cry of alarm: Suppl. 730, Ag. 1349 (W. Schulze, Kl. Schr. 182) ; but he was also
familiar with the idea of βοή in the special sense involved here, viz. the cry that
calls the armed host together for defence, as is sufficiently clear from Suppl.
679 ff. μηδέ τις ἀνδροκμὴς λοιγὸς ἐπελθέτω τάνδε πόλιν δαΐζων, ἄχορον ἀκίθαριν
δακρυογόνον “Apn βοάν τ᾽ ἔνδημον ἐξοπλίζων. There is no thought there of
pestilence (so Wecklein), but—as ἐξοπλίζειν alone is sufficient to show (cf.
102 πρὶν ἐξοπλίζειν Apn)—neither can βοάν be understood with the scholion
as τὴν ἐπὶ τοῖς ἀπολωλόσι βοήν; Wilamowitz is right (Interpr. 40): ‘und es
komme kein mannerbezwingendes Unheil . . . das die Biirger gegen die
Biirger unter die Waffen ruft.’ Here the Atridae in the anger of their hearts,
ἐκ θυμοῦ (Schol. où γλώσσηι ἀλλ᾽ ἔργωι θυμούμενοι), cry "War!" ; μέγαν is then
added to enhance the force of the expression, μέγαν Ἄρη being a free and
loftier ‘paraphrase of the actual cry’ (I owe this formulation to D. 5. Robert-
son) as in the passage quoted above from Bacchylides Béage . . . χαλκοθώρακα
Πανδίονος ἔκγονον. It is wrong, of course, when some editors interpret μέγαν
as though they had in front of them the μεγάλα κλάζοντε of one of the Homeric
prototypes (see above on 48 ff.), as e.g. Paley (‘yelping loud war with all
their hearts’) and Headlam (‘and loud they uttered harsh and passionate
cries of War’).
50. ἐκπατίοις ἄλγεσι: the dative as e.g. in S. 47. 531 καὶ μὴν φόβοισί γ᾽ αὐτὸν
ἐξελυσάμην. As regards the meaning of ἐκπατίοις, in spite of all the acumen
expended on the passage, we cannot get beyond pure conjecture. But it
would be irresponsible to cut the knot and replace the strange word by a
familiar one (Blomfield conjectured ἐκπάγλοις). The adjective is found no-
where else; for the adverb there is evidence (already adduced here by Henr.
Stephanus) in Erotian, p. 41. τό Nachmanson (as a variant on the reading
ἐκπάγλως in Hippocrates de mul. aff. 2. 171, viii, p. 352. 1 Littré ra audi ra
αἰδοῖα ἐκπάγλως αἴθεται κτλ.) ἐκπατίως" extpönws? καὶ (. . .)? ὁδοῦ ἀγνοοῦντες.
ἔνιοι δὲ γράφουσιν ἐκπάγλως. This seems to indicate for ἐκπάτιος the meaning
enormis (so Dindorf, Lex. Aesch., though he withdrew it in his Addenda,
p. 426, and adopted the spatial meaning), excessive (L-S). The scholiast
(in M and Tr), on the other hand, glosses our passage with τοῖς ἔξω τῆς ὁδοῦ,"

1 Virgil, Aen. 7. 582 Martem . . . fatigant is rightly explained by Servius as proelium cum
clamore deposcunt.
2 ἐκτρόπως, which is called vox nihili in Dindorf's Thesaurus, is quoted only from this
passage by L-S.
3 For the way in which the lacuna may be filled see Nachmanson. Gregory of Corinth,
who in the third part (wept τῆς ᾿Ιάδος διαλέκτου) of his compilation on dialect makes use
of Erotian from $ 163 onwards (cf. Koen, ad loc.), had a more complete text of Erotian's
glossary at his disposal than we possess (cf. Nachmanson, Praef. xvii): his version (8 185)
of the gloss runs τὸ ἐκτρόπως καὶ παραδόξως' ἐκπατίως.
4 Τὴ F the gloss has this form: ἐκπατίοις λεχέων" ἀντὶ τοῦ ἔξω τῆς αὐτῶν οἰκίας,
30
 COMMENTARY . line 50
cf. Hesychius ἐκπάτιον" τὸ ἔξω πάτου. The connexion with πάτος can hardly
be doubted; the formation has been compared with ἐκτόπιος, which is
used several times by Sophocles. Is the proper spatial meaning adequate
here? πάτος 'the trodden path, the beaten track' seems to have disappeared
from the living language after Homer. Homeric expressions like Y 136 f.
κιόντες ἐκ πάτου ἐς σκοπιήν, Z 202 πάτον ἀνθρώπων ἀλεείνων, make it very likely
that ἐκπάτιος may have meant ‘off the beaten track, lonely’ (so Paley and
Verrall, whose treatment of this passage in other respects is quite arbitrary).
To adopt this sense in Ag. so has seemed particularly attractive to the many
editors who suppose that thereby the simile of the birds becomes appreciably
more vivid. Headlam in particular takes this opportunity (cf. C.R. xvi,
1902, 436) to make some excellent remarks on ‘Aeschylus’ habitual practice
of pursuing a similitude, of carrying a figure through’—remarks which
retain their value even if one cannot follow his judgement on this passage.
If we consider the spatial sense of the word, we can rule out in advance the
view of its syntactical position first expressed by Triclinius and often
adopted since. Triclinius says in a note marked by a cross (= ἡμέτερον) :
δέον δὲ εἰπεῖν ἐκπατίων παίδων, ἐκπατίοις εἶπε πρὸς τὸ ἄλγεσιν. Similarly e.g.
Pauw, Schütz (with the quite improbable rendering of ἐκπάτιος = ‘abactus,
e medio raptus’), Butler (quoted by Peile), and Peile. Such an ‘enallage’ is
out of the question. But even the translation ‘in solitary grief” (Paley), ‘in
remote anguish’ (Headlam), which is inoffensive grammatically, is unsatis-
factory in sense. In explanation Paley observes: ‘The poet seems merely to
describe the haunts of vultures in the wild and solitary places, far away from
man’, and in the same way Headlam: ‘eagles and vultures were notoriously
remote and solitary.’ But it seems absurd that if the poet wanted to emphas-
ize the remoteness of the birds, he should have attached a word meaning
‘remote’ to ἄλγεσι (‘the connection of the epithet, as merely indicating
locality, with ἄλγεσι, is very harsh’, says Sewell). Wilamowitz felt this;
though he adopts the interpretation ἐκπάτιος = ἔξω πάτου, ‘solitary, re-
mote’, in his translation he separates the epithet (‘im Frieden des Gebirges’)
completely from ἄλγεσι: ‘wie ein Geierpaar . . . klagend um die Brut. Ihr
Fittich rudert kreisend um die Felsen durch die Lüfte, wo das Nest im
Frieden des Gebirges ihre Jungen barg.’ Moreover, it seems doubtful whether
‘Aeschylus would be likely to speak of a place which has just been invaded
by the tread of the spoiler as ἐκπάτιον whatever it may have been considered
before’ (Conington). I think there is also a question of poetical effect. We
are here at the point where the poet makes a transition from the account of
a particular event to the scene in nature with which he compares it, and at
this point he would hardly anticipate the visual details, the local colouring
of the simile drawn from the life of the great birds. It is much more effective
if first we have the expression of the feeling common to both men and birds,
and only at a later stage (from ὕπατοι onwards) the details penetrate into the
picture: ‘they cry like vultures which in overmastering pain for their children
high above the eyrie. ..’. For these reasons I hope that those commentators
are right who like Klausen,' Karsten, and others (see above) interpret the
words as ingenti dolore de liberis. Whether Aeschylus could use, and did use,
1 He makes, however, 2 compromise with the commentators who take the word in the
local sense : ‘Ingens dolor vulturios huc illuc rapit, ut huc illuc supra nidum circumvolitent.’
31
line 50 COMMENTARY
ἐκπάτιος in this sense, remains admittedly uncertain. And so the summing-
up must be non liquet.
παίδων. If one can say (Ag. 571) τύχης ἀλγεῖν' and (E. Hec. 1256) παιδὸς
ἀλγεῖν, no exception can be taken to ἄλγη παίδων, ‘sorrow for one’s child-
ren’. παίδων here corresponds to τέκνα in the passage of the Odyssey (x 217)
on which Aeschylus is drawing (cf. on 48 ff). But the word is chosen of
set purpose. παῖς is not used elsewhere of the young of beasts, as is rightly
emphasized by Verrall (Appendix I B ; the arbitrary assertions in that section
need not be gone into), cf. also P. Menge, De poet. scaen. Graec. sermone (Diss.
Göttingen 1905), 7, W. Ferrari, La parodos, 357. Plüss correctly comments:
‘the expression transferred from human parents.” Thus the feeling of grief
is intensified, and by transferring the sorrow of the wild creatures to the
human sphere, a further link is formed between the comparison and the
thing compared. Naturally it occurs to no one that the Atridae have lost
children, but one near and dear has been torn from them, as from the birds,
and that is the point here. Cf. on 717 ἶνιν.
51. ὕπατοι λεχέων. The paraphrase of the older scholia (in M) makes λεχέων
depend not on ὕπατοι but (and this can hardly be defended) on στροφοδινοῦνται:
οἵτινες ὕπατοι ὄντες... ἐπὶ τῶν λεχέων στροφοδινοῦνται. The interlinear gloss
in Tr, on the other hand, has ὑπεράνω τῆς καλιᾶς. ‘Supra cubilia, duris-
sima constructio’, says Blomfield in his neat fashion. No even partially
adequate parallel has yet been produced; most editors compare completely
unrelated expressions like Prom. 846 ἐσχάτη χθονός. Others treat the passage
as though we were principally concerned here with the so-called use of
superlative instead of comparative. None of the suggested alterations or
deletions is convincing. Headlam’s ὑπατηλεχέων is beyond discussion.
λεχέων is vouched for by 5. Ant. 423 ff. κἀνακωκύει πικρᾶς ὄρνιθος ὀξὺν φθόγγον,
ὡς ὅταν κενῆς εὐνῆς νεοσσῶν ὀρφανὸν βλέψηι λέχος, where apart from the Homeric
motif common to both poets there may well be ἃ reminiscence of this passage
of Aeschylus. λέχος in this application is rare; the use of a word borrowed
from human life performs the same function as παίδων above (cf. also Sept.
291 ff. δράκοντας ὥς τις τέκνων ὑπερδέδοικεν λεχαίων [this is what the read-
ing of the MSS comes to, cf. on 1653] . . . πελειάς). While admitting the
syntactical boldness of the expression ὕπατοι λεχέων, we should not blind our-
selves to the fact that the meaning is made perfectly clear. Weil well re-
marks: ‘insolenter quidem dictum est, sed sine dubio ab Aeschylo profectum.
. ὕπατοι λεχέων ita ponitur pro ὑπὲρ λεχέων, ut simul insit notio superlativi :
vulturii supra cubile sublime volant.’ Similarly Ahrens (p. 238: ‘hoch über
ihrem Neste’), Wecklein, and others. It seems highly probable that the poet
did not shrink here from an unusual expression, perhaps trusting to the
analogy of the construction of ὑπέρ and ὕπερθε.
στροφοδινοῦνται. The singular Homeric (IT 792) στρεφεδίνηθεν (δέ οἱ ὄσσε)
is rightly regarded by Leaf ‘as a proof that the Greek language in its most
vital period was capable of forming compounds beyond the lines of its regular
development’ ; on this type of formation cf. Ernst Fraenkel, Glotia, iv, 1913,
1 The conclusiveness of the passage is not prejudiced by the fact that ll. 570-2 do not
belong to the context in which we find them. 571 is in itself free from objection (see the
note there).
2 Dante, Par. 19. 91 (of the stork) quale souresso 1l nido si rigira (στροφοδινεῖται).
32
 COMMENTARY lines 531.
32. One can understand how Aeschylus must have been struck by this
ῥῆμα βόειον; in its noble daring it breathes the true spirit of his own language.
It is a sublime transference, and the meaning in the new context is at least
as vivid as in the old. “When Aeschylus says στροφοδινοῦνται, he is re-
modelling the first element of στρεφεδινεῖσθαι on the analogy of the noun
stems’ (Ernst Fraenkel, loc. cit.; cf. Wilamowitz on this passage). A slight
variation of the Homeric στρεφεδίνηθεν δέ of ὄσσε occurs in Prom. 882 rpoxo-
δινεῖται δ᾽ ὄμμαθ᾽ ἐλίγδην.
52. ἐρετμοῖσιν : a word extends from the first anapaestic metron into the next,
as in 64, 75, 84, 95 (i.e. with unusual frequency in the parodos of this play);
elsewhere 790, 793, 1339, 1341, 1555, 1557, Cho. 340, 859, 1073, Eum. 1010, and
cf. Suppl. 625, Δικτυουλκοί, Pap. Oxy. 2161, col. 2. 24 γάμον ὁρμαίνωμεν, ἐπεὶ
τέλεος. In all these passages only one syllable overruns the end of the
metron; a variant appears, characteristically, in the Prometheus: 172 καί μ᾽
οὔτι μελιγλώσσοις πειθοῦς. In view of this peculiarity in a drama of the
Prometheus-trilogy I am not sure that on fr. 192. 4 λίμνην παντοτρόφον
Αἰθιόπων (from the Προμηθεὺς λυόμενος) Wilamowitz (ed. maior, p. 68) is right
in observing ‘vitiosis numeris (duae breves metrum nusquam excedunt)',
but it must be owned that the expression in the form in which we have it
seems doubtful on stylistic grounds. (In these two passages G. Hermann,
Elementa doctr. metr. 374, assumed that a quasi-diaeresis took place between
the two elements of the compounds μελι-γλώσσοις, παντο-τρόφον.) A Sopho-
clean instance of two syllables overrunning the end of the metron is Trach.
985 κεῖμαι πεπονημένος ἀλλήκτοις. As is well known, Sophocles also provides
several anapaestic dimeters of the type of Ag. 52, e.g. 47.146 ἧπερ δορίληπτος
ἔτ᾽ Av λοιπή, Trach. 1276 μεγάλους μὲν ἰδοῦσα νέους θανάτους.
53 f. δεμνιοτήρη πόνον ὀρταλίχων. In the interpretation of these words
many scholars have gone astray by (1) referring to the young of the vultures
the action described in δεμνιοτήρη, as Hesychius (see below) suggests, and
(2) taking πόνον, in a way considered but rejected by Schiitz, as meaning res
labore parta (so, e.g., Dindorf and L-S following Passow ; cf. Pearson on Soph.
fr. 793. 3). On the latter point I need only say that in passages where πόνος
means 'result of toil' the agent whose toil produces it is always expressly
named: μελισσᾶν... τρητὸν πόνον or ὑψηλὸν. . . τεκτόνων πόνον OT ἐμὸν
ὠδίνων πόνον. That is not so here; and besides, if πόνον here meant the
product, there would be no room for ὀρταλίχων. Now for the first point.
The scholion on our passage preserved by Hesychius s.v. δεμνιοτήρης runs!
καθότι où νεοσσοὶ ἔτι τοιοῦτοί εἰσιν ὡς τὰ δέμνια τηρεῖν καὶ κατέχειν, μηδέπω
πέτεσθαι δυνάμενοι. πόνον δὲ τὰ περὶ τὴν τροφὴν αὐτῶν. In this both the meaning
given to πόνον and the rendering of δεμνιοτήρης (taken by itself) are right.
The verbal element in the word is active (cf. Ernst Fraenkel, Nomina
agentis, ii. τος n. 2), and the whole represents ra δέμνια τηρῶν, as we Can see
from Ag. 1449, the only other passage where the word occurs. For the special
meaning of rnp- presupposed here cf., e.g., Ar. Frogs 15151. σὺ de τὸν θᾶκον
τὸν ἐμὸν παράδος Σοφοκλεῖ τηρεῖν καὶ διασώιζειν. The interpretation in
Hesychius is, however, wrong in referring δεμνιοτήρη to the unfledged young
cowering in the nest. The attribute is clearly intended to qualify the toil
itself, and must be understood of the parent birds guarding their brood.
The genitive ὀρταλίχων depends on πόνος (toilsome watch over . . ., toil
4872.2 D 33
lines 53 t. COMMENTARY
for ...) just as in Pers. 751 πολὺς πλούτου πόνος οὑμός. Hermann's inter-
pretation of the passage is entirely correct: 'δεμνιοτήρη πόνον Oprakixwv est
cubiliprema cura pullorum, sive labor quem parentes pullis incubando sus-
tinuerunt.' In support of this rendering (as opposed to the view that πόνος
denotes the young birds themselves) it may be pointed out that this simple
and obvious idea, 'all the toil of nurture was in vain, 15 now fruitless and for
ever lost’, is frequent in Tragedy, e.g. Cho. 752 (the Nurse's lament for Orestes’
death) καὶ πολλὰ καὶ μοχθήρ᾽, ἀνωφέλητ᾽ ἐμοὶ τλάσηι, E. Tro. 1187 f. οἴμοι τὰ
πόλλ᾽ ἀσπάσμαθ᾽ al τ᾽ ἐμαὶ τροφαὶ ὕπνοι τ᾽ ἐκεῖνοι φροῦδά μοι, S. El. 1143 ff.
ὀρταλίχων. The word ὀρτάλιχος, like the bird-name κόψειχος and like
ὀβρίχοισι, for which we now have evidence from Aeschylus (cf. on 143), is
formed with the diminutive-suffix -yos, which was particularly widespread
in Boeotian, though by no means limited thereto (cf. Bechtel, Griech. Dialekte,
i. 264; Locker, Glotta, xxii, 1934, 56 ff.; Buck and Petersen, A Reverse Index,
681). The living use of the word in Boeotian is proved by Ar. Ach. 871 with
the scholia (details below) and Strattis fr. 47. 4 K. 'you Thebans call τὸν
ἀλεκτρυόνα ὀρτάλιχον᾽. Here in Aeschylus we have the original meaning ‘young,
unfledged bird'; the other more specific sense ('chicken' and, among the
Boeotians, ‘cock’) is clearly a secondary development (L-S give the relation-
ship in inverse order). This view is supported by the second oldest occur-
rence of the word, Ar. Ach. 871, which again shows the non-specific use,! and
is in agreement with the evidence in ancient lexicography, derived in this
instance from the treatise of Aristophanes of Byzantium περὶ ὀνομασίας
ἡλικιῶν (cf. below on 141 and on 143); cf. A. Nauck, Aristophanis Byzantii
fragmenta, p. 127 (ὀρνίθων τὰ ἐν ὄψει ἤδη ὄντα veorrot, κατὰ δέ τινας ὀρτάλιχοι).
There the context shows that we must think not of the domestic fowl but
of birds in general. Cf. also Hesychius (ὀρτάλιχοι" οὗ μήπω πετόμενοι veoooot‘
καὶ οἱ ἀλεκτρυόνες). That the development here was from the general to the
particular, and not the other way round, is further made probable by the
well-known secondary use of ὄρνις as ‘domestic fowl’ (cf. Wackernagel,
Unters. zu Homer, 165 n. x), which provides an exact parallel. .
55. ὕπατος. The word’ seems here to be complex and to enshrine more than
one idea. The first sense that presents itself is ‘one most high, one most
mighty’, meaning a powerful god; to this corresponds the use of ὁ κρείσσων
in 60, which prepares the way for the naming of Ζεὺς ξένιος. To limit this
ὕπατος as many do, following Ahrens (see below), to the god’s seat on a
mountain peak (Wilamowitz, too, has ‘ein Gott, ein Herr des Hochgebirges’)
hardly does justice to the poet’s intent. Homer, whose usage more than
anything else will have been the model for Aeschylus, uses ὕπατος of Zeus
(and he applies the epithet to no other god) only in the metaphorical sense:
θεῶν ὕπατος καὶ ἄριστος and Κρονίδη ὕπατε κρειόντων, which is glossed by the
ancient interpreters with éfoywrare, μέγιστε. It is true that in many places
in the Greek world we find the cult name Ζεὺς Ὕπατος used for Zeus
. ' In Ar. Ach. 871 ὀρταλίχων denotes birds in general, not domestic fowls, as is clear
from 875 ff. (Bechtel therefore, Griech. Dialekte, i. 308, is not correct.) The scholiast, who
says ὀρταλίχων δὲ τῶν ἀλεκτρυόνων κατὰ τὴν τῶν Βοιωτῶν διάλεκτον, is speaking from sound
knowledge, which, however, in this case is misapplied. Sophocles’ use of the word
ὀρταλίχων for young kids in fr. 725 N. (= 793 P.) is a case of free poetic transference
(cf. Pearson, ad loc.).
2 Cf. Wackernagel, Unters. zu Homer, 213 f.

34
 COMMENTARY lines 55 f.

worshipped on mountain tops (references in O. Gruppe, Gr. Mythologie, 1103

n. 2 under B, and Roscher’s Lex. d. Mythol. vi. 570; in Attica, however, the
altar of Ζεὺς Ὕπατος was in the πόλις near the Erechtheion), and it has been
supposed that ‘there is reason to suspect that the epithet had originally
a literal rather than a metaphorical sense’ (A. B. Cook, Zeus, 11. 876). But in
Aeschylus this cannot overrule the Homeric usage any more than in Pindar
(Ol. 13. 24 ὕπατ᾽ εὐρὺ ἀνάσσων Ὀλυμπίας . . . Ζεῦ πάτερ); if it were necessary,
we could point to Ag. 509 ὕπατός τε χώρας Ζεύς. It is also clear that here
ὕπατος is used principally, if not exclusively, with reference to Zeus, for whose
introduction we are effectively prepared by the naming of two other gods
who, great though they are, are yet less than he; in the application of the
simile in 60 ff. it is Zeus alone who acts. But this use (derived from Homer),
‘one who is mightiest’, does not rule out here the other sense, ‘one who sitteth
on high’. The echo of ὕπατοι (51) demands that we should be conscious here
of this local sense as well. As Weil well says, ‘vulturii supra cubilia sublime
volant. In altiore etiam loco—üraros δ᾽ diwy—sedent dii vindices qui illis
opitulantur.’ ‘One most high, one high above’ is all the poet says; we may
think of the sky or of a high mountain, but we have no right to emphasize
one alternative rather than the other. What is probably true of Pan (see
below) need not for that reason apply to the other two gods also.
55 f. Why are these three gods and no others named as avengers of the birds
bereft of their young? Much acumen has been lavished by many commenta-
tors on the answering of this question. A critical survey of solutions put
forward down to his own time (1860) is given by Ahrens, pp. 234 ff. He him-
self takes a one-sided view and maintains that the three gods are here named
primarily as dwellers on the mountain tops. Apart from the weakness of the
evidence regarding Apollo, this cannot be right, for if the poet attached
so much importance to the fact that these gods dwell on mountains, he would
have added a word to make it clear. In the whole discussion (including
that since Ahrens) it has not been sufficiently noticed that the mention of
Apollo is the only point that really needs an explanation. The part played
here by Zeus, like his relation to eagles and vultures, is clear without more
ado. On Pan see below. As for Apollo, we cannot, as I see it, say with cer-
tainty what moved the poet to mention him. He is not one of the gods who
are usually thought of as dwelling on mountains. Ahrens's reference to
Parnassus and Cynthus does not alter the fact, nor does the evidence cited
by him from Pausanias (8. 38. 8) that on the Lykaion in Arcadia, besides the
shrines of Pan and of Ζεὺς Avkaws, there was also a temple of AméAAwy
ITappäoıos (this certainly has no bearing on Aeschylus). Welcker, Griech.
Gétterlehre, i. 487, sees a connexion between Ag. 55 and the mention by
Aeschylus in the Προμηθεὺς λυόμενος (fr. 200 N.) of ᾿ἄπόλλων ἀγρεύς (see
further in Gruppe, Mythol. 1247 n. 3), and Karsten, too, remarks 'memoratur
"AmóAAov ut deus venator' ; but in the Prometheus the context should not be
overlooked (Ἡρακλῆς ἕτερον θεὸν παρακαλεῖ μέλλων ἐπὶ τὸν ὄρνιν [sc. the eagle]
αἴρεσθαι τὸ τόξον) ; Apollo is of course past-master in shooting with the bow.
Some, e.g. Stanley and Schneidewin, assume that Apollo is named as god of
augury; but the vultures in this passage have nothing to do with omens.
G. Thomson sees in Apollo a ‘patron of birds’. It is well known that birds
are sacred to him, as to many other gods, but vultures (or eagles) are not
35
lines 55 f. COMMENTARY
among them. I have thought of H 58 ff., κὰδ δ᾽ dp’ A@nvatn τε καὶ ἀργυρότοξος
Ἀπόλλων ἐζέσθην ὄρνισιν ἐοικότες αἰγυπιοῖσιν φηγῶι ἐφ᾽ ὑψηλῆι, but that
Aeschylus should have inferred from this a special relation between Apollo
and the aëyumoi seems to me most unlikely. It is probably better to admit
that we do not know the poet’s special motive in naming Apollo. But it
is not surprising that this god should be named : he may well be moved at the
deed of violence, and help the great birds who cry for aid in their distress,
just as he helps mortal men who call upon Ἀπόλλων ἀποτρόπαιος.
ñ τις ᾿Απόλλων À Πὰν ἢ Ζεύς: the oldest known parallel to this form of
expression in the religious sphere comes in the dactylo-epitritic fragment in
the great papyrus of Favorinus (Pap. Vat. Gr. ii, ed. Norsa-Vitelli, Cittä del
Vaticano 1931), col. ii. 6 εἰσὶν γάρ, εἰσὶν ἀξιοσπάμονες ἄλλαι ταὶ μέλονται πρός
τινος ἢ Διὸς ἢ γλαυκᾶς Addvas; as author Vitelli doubtfully suggests Pindar,
P. Maas (Deutsche Literaturzeitg. 1931, 1211) Sophocles. In Ag. 55, ris must
be taken with all three gods. For the form in general cf. Semonides of
Amorgos fr. 17 D. ἢ τυφλὸς ἤ τις σκνιπὸς 7) μέγα βλέπων, S. Oed. C. 95 ἢ σεισμὸν
ἢ βροντήν τιν᾽ ἢ Διὸς σέλας, Virgil, Aen. x. 18x ff. Anthea st quem . . . videat
... aut Capyn aut . . arma Caici. The idea is ‘some one of the Most High,
no matter which: either Apollo or Pan or Zeus’.
Πάν: the Scholiast (in M) has the gloss ὁ ὄρειος. This is actually all the
explanation that is required. The god of the Arcadian mountains, and then
of mountains and forests in general, is naturally the protector of the wild
creatures that dwell there, including the vultures that nest upon their cliffs.
As lord of the mountain forest, Pan in common with the Mirnp ὀρεία receives
the homage of birds, which also live on the mountain side, with hymn and
dance (Ar. Birds 745) ; but it by no means follows that Pan was in any special
way a patron of birds. ‘The god who had his throne in the mountain heights
was a powerful lord, so that Aeschylus, who had scen his reception in Athens,
mentions him next to Zeus’ (so Wilamowitz, Glaube d. Hell. i. 247; O. Kern,
Religion d. Griech. ii. 38 n. 3, on this passage is not quite right). Beazley points
out (Der Pan-Maler, το) that while Pan was almost unknown in Athenian art
before the Persian Wars, he was very popular in Athens after it; cf. also F.
Brommer, ‘Panbilder des 5. Jahrhunderts’, Arch. Jahrb. liii, 1938, Anzeiger
376 ff.
57. yoov: ‘especially lamentation for the dead’, as Passow rightly says (fol-
lowed by the earlier editions of L-S, but now the remark is unfortunately
cut out); similarly, e.g., Klausen on Ag. 1079, Ebeling, Lex. Hom. ‘fletus,
eiulatus, in primis funebris’. For this almost technical use of the word, in
which it comes close to θρῆνος, see the remarks of Schadewaldt, Monolog
und Selbstgespräch, 160 n. 2, and Hermes, Ixvii, 1932, 320. In Aeschylus I find
only a single passage (Prom. 33) where the lament is not connected in
some way with a dead body or with death. Everywhere else the narrower
meaning is retained, even (as appears from the context) in Sept. 657 ; in Suppl.
116 ζῶσα γόοις με τιμῶ there is a marked verbal point (see below on 1322).
The close kinship with θρῆνος is clearly seen for instance in fr. 291 θρηνεῖ δὲ
γόον τὸν ἀηδόνιον and Ag. 1079 (see ad loc.). The birds lament πόνον ὀρταλίχων
ὀλέσαντες : for them their young are dead even though in fact they may have
been carried off alive.
τῶνδε μετοίκων; down to our own time several critics have substituted.
36
 COMMENTARY line 57
conjectures for the ‘finely imaginative word μετοίκων᾽ (Sidgwick). The
kernel of the matter was correctly understood in antiquity. In the scholion
on S. Oed. C. 934 (abbreviated in Suidas s.v. μέτοικοι, iii. 379. 4 ff. Adler) is
included the explanation, probably taken from a commentary on the Aga-
memnon: κέχρηται δὲ καὶ Αἰσχύλος ἐπὶ τῶν οἰωνῶν ἐν Ἀγαμέμνονι λέγων οὕτω"
'τῶν δὲ μετοίκων᾽ ἀντὶ ἐνοίκων: μετοίκους γὰρ εἶπε τῶν ὑψηλῶν τόπων τοὺς
oiwvoüs.! The reference to ὑψηλοὲ τόποι clearly gives us the direction in which
to look for the poet’s meaning. Somewhere high above our heads, in heaven
or on some heaven-kissing mountain, dwell the gods. And up above (seen
from the point of view of men) there also dwell the vultures, no matter
whether their eyrie be perched upon a cliff or in the hollow of some rock?
(cf. Eum. 23, E. Ion 155). But those lofty regions belong really to the gods;
they alone have full rights of citizenship therein. The birds that are permitted
to live there too are μέτοικοι in the heavenly πόλις : they stand accordingly
under the protection of the lords of heaven. This is a noble and a simple
idea, perhaps peculiar to Aeschylus, who (fr. 53) speaks of the swallow that
shares man’s house with him as πεδοίκου χελιδόνος (the passage is adduced by
Ahrens, but I cannot share his interpretation of μετοίκων).
The only difficulty is τῶνδε. Hermann, followed by Enger, Schoemann
(Opusc. iv. 157), and others, rightly took exception to it: ‘aut τῶν μετοίκων
aut μετοίκων ponendum erat.’ Subterfuges are of no avail. What is the use,
for instance, of Wecklein’s reference to S. El. 790 πρὸς τῆσδ᾽ ὑβρίζηι μητρός,
where Clytemnestra is standing close to the speaker and τῆσδε has its full
deictic force? Ahrens’s artificial explanation (‘the pronoun denotes the birds
as in the immediate neighbourhood, having made their way before the throne
of the god’) is likewise quite unconvincing. Headlam-Thomson suggest that
τῶνδε is ‘equivalent to a limiting epithet’ and cite as a parallel Eum. 405
(in that passage, which has long been recognized as an actors’ interpolation—
cf. Wilamowitz, Einleit. in d. griech. Tragödie, p. 152 n. 63, and Interpr. 181—
it is as clear as can be that τόνδ᾽ ὄχον refers to something visible on the stage)
and 500 (where the Chorus speak of themselves as μαινάδων τῶνδε). But apart
from the difficulty emphasized by Hermann, I feel also another objection.
Alongside of οἰωνόθροον, which denotes the source of the γόος and to a certain
extent takes the place of a genitive, τῶνδε μετοίκων is strange. It may have
been because they felt this difficulty that Stanley (‘delendum ὑπόστιγμα post
μετοίκων᾽, Pauw, and Schütz deleted the comma placed by earlier editors
from Turnebus onwards after μετοίκων (this comma is found in F but not in
M or V; and the scholion in M, ὑπὲρ τῶν μετοικισθέντων νεοσσῶν, clearly
connects τῶν δὲ (see below) μετοίκων with what follows, not with what pre-
cedes; why Hermann says ‘scholiastes quo rettulerit, non liquet’ I do not
understand). But in the interpretation ‘horumque abactorum [this false
rendering rests on the μετοικισθέντων of the scholion] poenas sero exigentem
mittit raptoribus Erinyn’ (Schiitz), the genitive is made to depend on
ὑστερόποινον *Epwiv. That cannot be regarded as possible here, where a god
! After this we read the words κἀκεῖσε ἀντὶ ἐνοίκων. Here and also in the words which
precede the sentence quoted above the scholion has gone badly astray, as has been re-
marked by Ahrens (p. 238), and by van Heusde in his apparatus on Ag. 57; cf. also V. de
Marco, ‘De Scholiis in Soph. trag. vet.’, Accad. dei Lincei, Serie VI, vol. vi, fasc. 2, 1937, 212.
2 Any early morning at Delphi the vultures may still be seen flying up high over Castalia
from their nests in the rocky clefts of the Phaedriades.
37
line 57 COMMENTARY
sends the Erinys, though it is quite natural to say πατρὸς ᾿Ερινύς when the
Erinys, the Curse and the Spirit of Vengeance, is conjured up and set in
motion by the father. It should also be clear that Hermann’s suggestions
must both be dismissed, right though he was in reading τῶν δὲ μετοίκων,
preceded by a stop: he wanted either to understand ἐστίν with ἀΐων (very
unnatural; nor would one think of an anacoluthon after the participle in
a sentence which is so smoothly carried forward and broken by no interrup-
tions) or to assume that some words have dropped out: γόον ὀξυβόαν (peya
θυμοῦται», τῶν δὲ μετοίκων κτλ. On the latter proposal Wilamowitz comments
‘probabiliter’. The definite gain in Hermann's treatment of the passage is the
recognition that we must read τῶν δὲ (cf. on 1638). In this connexion
attention might well have been called to the surviving indications that the
παράδοσις was interpreted in this way (τῶν δὲ) not later at any rate than
early Byzantine times and perhaps even in antiquity (that in passages
where the sense is open to doubt accents were sometimes written in ancient
manuscripts is clear from any number of papyri). In Schol. 5. Oed. C. 934
(see above) the quotation is given in the Laurentianus in the form τῶν δὲ
μετοίκων (noted by Papageorgiu and easily recognizable in the facsimile).
That in the text of the Agamemnon likewise the exemplar of the Mediceus
had the same accentuation seems to me to be made probable (though the
conclusion is not certain) by the above-quoted scholion in M: ὑπὲρ τῶν
μετοικισθέντων νεοσσῶν. However that may be, I believe Hermann’s division
into two words (τῶν δὲ) and the punctuation adopted by him and others
(comma in front of τῶν δὲ μετοίκων) to be certainly right. It also seems to me
inevitable to postulate a lacuna. I put it, however, not like Hermann before
τῶν δὲ μετοίκων but after. What Aeschylus wrote we cannot recover, but about
the necessary ideas I have no doubts. Exempli gratia we may fill the gap
with words which up to a point would account for the omission: τῶν δὲ
μετοίκων (uéyav οἶκτον ἔχων» torepdmowov κτλ. We find οἶκτον ἔχειν in
Sophocles, οἶκτος with the genitive often in Tragedy. For the expression
cf., e.g., S. 4j. 139 μέγαν ὄκνον ἔχω. But the details are quite uncertain; my
point is to establish the structure of the sentence. τῶν now regains its
proper and effective possessive function. For lacunae in anapaests see on
1521 ff.
58. ὑστερόποινον πέμπει παραβᾶσιν "Epiwóv: there is an exactly corresponding
expression in Cho. 382 (Zeus) κάτωθεν ἀμπέμπων ὑστερόποινον arav (on which
see Schadewaldt, Hermes, Ixvii, 1932, 332 f., who points out the part that this
late-following atonement plays in Solon), and a very similar one in Ag. 155
Mis rervomowos. These weighty new compounds emphasize one of the
central ideas of the Ovesteta. Cf. S. Ant. 1074 f. ὑστεροφθόροι "Epivves, probably
following Aeschylus' example.
59. παραβᾶσιν. So far we have only heard that the vultures have lost their
young, not of an outrage against the eyrie. It may be said that Aeschylus
from the beginning of his simile (48) expects his hearers to remember the
passage in the Odyssey (see on 49), where we are told οἷοί re τέκνα ἀγρόται
ἐξείλοντο πάρος πετεηνὰ γενέσθαι. But if he thought that this feature of the
story should be regarded as essential from the outset, he would expressly
refer to it; to presuppose, in the manner of Hellenistic narrative, that the
hearer is familiar with some earlier treatment of the subject is not his way.
38
 COMMENTARY lines 61 f.

For a real understanding of the passage there are two things we must bear
in mind. First, Aeschylus shares with many of his contemporaries the
tendency of archaic narrative not to display at the outset the most important
ὑποκείμενα in their entirety, but to introduce the details at the moment when
they give rise to a new element in the story and therefore cannot be sup-
pressed any longer. Cf. Appendix A. We shall see the same tendency at
work in Ag. ı90. In the passage before us the outrage is first mentioned
(rapaBâow) in the place where its expiation is strongly emphasized.
But there is a further point to be noticed here. As the image of the birds
of prey develops, it assumes a new function. The manner in which it is intro-
duced closely resembles the Homeric fashion: ὥς 7° αἰγυπιοὶ... κλάζοντε...
ds of κεκλήγοντες κτλ. But where the comparison flows back into the main
stream of narrative (60), the connexion, clearly marked by οὕτω, is no longer
the crying but the punishment of the criminal by Zeus. The Aeschylean
simile is not content like the Homeric’ with a self-contained reflection from
the world without, presented as a counterpart of the incident principally
depicted ; as it runs its course, almost unnoticeably it takes on new features
which belong to the main narrative. Cf. on 395. Confronted with what the
parent birds have lost, the hearer is meant from the outset to think at the
same time of the loss of the Atridae and the κλοπαὲ γυναικός (401) ; thus his
mind is already prepared for the παραβάντες even before they are mentioned.
The simile as a whole leads us by a gradual progression into a much deeper
level of thought ; it starts with external things, with crying and lamentation,
and ends with the central theme, the just vengeance of Zeus. For the inter-
action of comparison and thing compared cf. on 732 ff.’
Ἐρινύν: the mighty period that extends from the beginning of the parodos
(40) here at last finds its terminus in a word heavy with meaning placed at
the very end: 'Epwóv. In each of the four great choric songs (cf. 463, 749, 992)
of this play the Erinys takes a prominent place.
60. ὁ κρείσσων: substantival, as though it were ὁ θεός (it is so glossed in
Tr). As an explanatory afterthought ξένιος Ζεύς is added, but not until we
have been given the name of Alexandros, who broke the laws of hospitality.
The gods are often called κρείττονες (cf. Wilamowitz, Glaube d. Hell. i. 19),
and this is explicitly reported from the Aitna tragedy of Aeschylus (fr. 10).?
‘From Xenophanes [A 28, Vorsokr. i5. 117] downwards the very condition
which defined and limited τὸ θεῖον was τὸ kpareiv’ (Headlam, C.R. xv, 1901,
396). But here ὁ κρείσσων also expresses the certainty that Zeus, who watches
over the laws of hospitality, is mightier than Paris the offender and his sup-
porters, and will vanquish them. Thus in ὁ κρείσσων here, just as in ὕπατος
in 55, more than one group of associations is bound up.
61f. ξένιος Ζεύς. He is offended by the adultery committed beneath the
husband's roof. The kind of adultery which violates at the same time the
sanctity of the home and the mutual bond between host and guest has been
! This is of course a rough contrast. For a qualification see H. Fránkel, Die homerischen
Gleichnisse, 105.
2 I do not want to alter what I wrote many years ago, but I now add a reference to
Daube, Rechtsprobleme, 101 f., with whose interpretation I agree to a large extent.
3 We should not adduce Prom. goz f., for there Musgrave recognized that θεῶν was
a gloss, and Wilamowitz rightly explained κρεισσόνων by reference to the beginning of the
ode.
39
lines 61 f. COMMENTARY
reckoned since earliest times a particularly heinous wrong. Cf. G. Glotz,
La Solidarité de la famille, 317 n. 3.
62. πολυάνορος : the first occurrence of the word. On the numerous adjectives
and proper names formed in the same way from the Iliad onwards cf.
Wackernagel, Dehnungsgesetz 4o, Kretschmer, Οἰοίία, xxiv, 1936, 246 ff.,
©. Hoffmann, Glotia, xxviii, 1939, 28 ff. ἀνήρ here refers to the relation between
the sexes, while it has another sense (ἀνήρ — inhabitant, warrior, etc.) in
πολύανδρος, Which Aeschylus uses several times (Pers. 73, 533, 898, Ag. 693).
The distinction in accordance with which -avop is mainly used with reference
to a single man, -avópos mainly of a number of men, is not confined to this
particular compound, cf. O. Hoffmann, op. cit. 42 ff. In the passage before
us zoAÀvávopos is explained in the meagre gloss of the Mediceus: πολλοὺς
μνηστῆρας ἐσχηκυίας. On the other hand, the scholion preserved by Triclinius
(ZxoA. aA.) glosses the adjective by πολυάνδρου and quotes from Lycophron's
Alexandra (146) νυμφεῖα πεντάγαμβρα . . . γάμων, said of Helen (cf. ibid. 143
[also of Helen] τῆς πενταλέκτρου θυιάδος). Similarly Stanley referred to
Lycophr. 851 τῆς... τριάνορος κόρης (Helen)! and especially to the well-known
line of Stesichorus (fr. 17 D.) where Aphrodite in anger at the daughters
of Tyndareos διγάμους τε καὶ τριγάμους τίθησιν καὶ λιπεσάνορας ; accordingly
Stanley rendered it by multinubae. This is clearly right : first, it is in keeping
with the proper meaning of -dywp (cf. Sommer, Abhdl. Bayr. Akad. NF.
ix, 1934, 45: ‘die oftmals einen Gatten gehabt hat’; O. Hoffmann, Glotta,
xxviii, 1939, 41), and secondly, thus and only thus is brought out the full
bitterness of the reproach.
63. πολλὰ παλαίσματα καὶ yuroßapfj: for the order of words see note on 403 f.
‘The expression: γυιοβαρῇ παλαίσματα shows that the poet did not mean to
indicate the fall . . . of the severely wounded, but the pressing and holding
down of one wrestler by the other.” So Wecklein, Studien zu Aesch. 92
(following Casaubon, Blomfield, and others). Casaubon compares Pers. 929 f.
Aoia δὲ χθὼν... ἐπὶ γόνυ κέκλιται. It is thus clear that one must decide
against Ahrens’s preference for ἐρειπομένου (FTr) over the ἐρειδομένου of
MV; Karsten long ago compared the Homeric (e.g. H 145) ὕπτιος οὔδει ἐρείσθη.
63 ff. παλαίσματα θήσων: the two words belong closely together as one ex-
pression, into which the genitives γόνατος ἐρειδομένου and διακναιομένης κάμακος
are intruded.
moAvdvopos ἀμφὶ yuvaixes . . . γόνατος κονίαισιν ἐρειδομένου is reminiscent of
£ 68 f. 'EAévns . . . ἐπεὶ πολλῶν ἀνδρῶν ὑπὸ γούνατ᾽ ἔλυσε.
65. προτελείοις : Headlam, ad loc. : προτέλεια, as representing the ceremonies
previous to the consummation of marriage, was metaphorically used for
preliminaries to the completion, perfection, accomplishment of anything—of
a voyage in v. 227, of mature age in v. 720, and often in later authors’. This
is right in the main (cf. also Wackernagel, Dehnungsgesetz, 10), but does less
than justice to the use that Aeschylus makes of the word. It is highly
probable that he did not find the ‘metaphorical’ use already in existence, but
is himself responsible for this extension of meaning, as elsewhere also he
extends the use of terms belonging to the technical language of religious cult.
In Aeschylus the original ritual meaning is not faded to the degree that
Headlam's words suggest. On Ag. 227 see Passow and L-S s.v., where the
τ Cf, Schol, E. Andr, 229,
40
 COMMENTARY line 67

preceding θυτὴρ γενέσθαι is rightly taken into account. In the present passage,
too, the rendering ‘preliminaries’ misses some of the sense. It is an ‘opening
sacrifice’, and the offering proper will follow when one of the combatants
receives his death-blow ; ‘to sacrifice’ often comes to mean ‘to slay’, as is well
known, not only in θύειν but in ἱερεύειν and kindred words (cf. 735 ἱερεύς τις
Aras). But by the word προτέλεια an Athenian understands in the first
instance not a preliminary sacrifice in general but, in accordance with what
is by far the most frequent use of the word, a sacrifice offered before marriage."
The word in itself, therefore, suggests cheerful images and ideas. For this very
reason here and in 227 Aeschylus inverts it and gives it a sinister meaning.
This employment of bona verba to indicate something disastrous is very
characteristic of the poet, cf. on 336 (about Sept. 367) and 745.
It can hardly be doubtful here, in this close association with διακναιομένης
κάμακος, what is meant by the προτέλεια. The scholiast has got it wrong:
κυρίως ταῖς πρὸ γάμου θυσίαις (right so far), νῦν δὲ ταῖς πρὸ τῆς ἁλώσεως
μάχαις. This explanation, and still more openly the expanded version in
Triclinius’ ZyoÀ. παλ., νικωμένων ἐν ἀρχαῖς τῆς μάχης, μέχρις ἂν τελεσθείη τὸ
πεπρωμένον (this refers to 68, wrongly, for that clause belongs to a fresh train
of ideas), is influenced by a desire to bring προτέλεια into relation with the
τέλος of the expedition. This ignores the fact that in the intruded genitives
γόνατος ἐρειδομένου and κάμακος διακναιομένης we are not given a complete
picture of the war that preceded the sack of Troy ; a few characteristic details
are picked out from the battle-scenes, and nothing more. It is in παλαίσματα
that the whole of the fighting is indicated; there can be no doubt that here
‘wrestling’ is meant not in the literal sense but as a general term for a contest
(with spear, sword, or whatever it may be), a usage often found in Greek and
not unknown in modern languages. Wecklein, then, is on the wrong track:
‘the spear-fight forms the prelude to the wrestling-fight: when the spear is
shattered, then they turn to wrestling’ (so, too, E. Neustadt, Hermes, lxiv,
1929, 236 ἢ. 1). Schütz rightly says: ‘primum hastis pugnabant, deinde, illis
fractis, gladiis utebantur. We may be sure that Aeschylus had in mind
this order of events, which is typical of many scenes of fighting in the I/:ad.
Cf. also Hdt. 7. 224. ı (the Spartans at Thermopylae) δόρατα μέν νυν τοῖσι
πλέοσι αὐτῶν τηνικαῦτα ἤδη ἐτύγχανε κατεηγότα, οἱ δὲ τοῖσι ξίφεσι διεργάζοντο
τοὺς Πέρσας. The second stage, the battle with the sword, leads to the
θυσία in the sense which has been explained above.?
66. θήσων (παλαίσματα): cf. on 395 and on 16or. | .
67. Following the arrangement in the MSS, G. Hermann, Elem. doctr. metr.
379, makes a new anapaestic system begin after the paroemiac κάμακος
θήσων Δαναοῖσιν with Τρωσί θ᾽ ὁμοίως : ‘plena est enim et absoluta sententia
versu paroemiaco, sed egregie, quasi nunc demum Troiani in mentem veniant,
! The προτέλεια seem to have embraced still further ceremonies: the fact that they are
rightly defined in Hesychius s.v. as ἡ πρὸ τῶν γάμων θυσία, καὶ ἑορτή is vouched for by some
other pieces of evidence (not, it is true, very plentiful or exact); cf. Pernice in Gercke-
Norden’s Einl. i. d. Altertumswissenschaft, ii. 1, 4th ed., p. 55. In any case it is obvious that
the ϑυσία is a predominant element: it is plain from the usage of Aeschylus that he is
thinking chiefly of this element (cf. also the scholion quoted below). Cf. on 720.
2 There is nothing in the text to make us suppose ‘that the fighting was a bitter marriage
prelude for Paris’ (Sidgwick); besides, this would be quite untrue chronologically (cf.
699-716).
41
line 67 COMMENTARY
hi in principio novi systematis commemorantur. Est enim praecipua
quaedam vis in horum commemoratione.’ This is probably correct, cf. also
W. Kranz, Hermes, liv, 1919, 302. Similarly in A. Suppl. 4f. and 32 f. the
sentence continues over the end of the paroemiac, and at Suppl. 14 (ἐπέκρανε)
there is the end of a clause (κῶλον), but the sentence continues.
ἔστι δ᾽ ὅπηι viv: monosyllable at the end of the metron, following a long
syllable. Wifstrand, Hermes, lxix, 1934, 211, shows that in Aeschylus’ ana-
paests this occurs only five times in all, and always in places where, as here,
the preceding ‘longum’ is resolved. But his attempt to eliminate this passage
from the list of those five special cases, by introducing the enclitic νυν, is
mistaken. It is by no means clear that νῦν here must be ‘indeterminate’ ; it
serves to indicate the sharply emphasized point of transition from the events
of the past which have been narrated hitherto to the still unknown future
(τελεῖται δ᾽ ἐς τὸ πεπρωμένον).
Abresch long ago adduced parallels from which he inferred rightly that in
this ἔστι δ᾽ ὅπηι νῦν ἔστι we have a current manner of speaking obviously
derived from the speech of common folk, ‘cuius modi formulis utuntur
Graeci, quoties aliquidpiam non admodum probant, palam tamen repre-
hendere nolunt’, e.g. S. Oed. R. 1458 ἀλλ᾽ ἡ μὲν ἡμῶν μοῖρ᾽, ὅποιπερ elo’, ἴτω,
E. Med. 889 f. ἀλλ᾽ ἐομὲν οἷόν ἐσμεν, οὐκ ἐρῶ κακόν, γυναῖκες, and others of the
same kind. Blomfield in his glossary furnishes a rich collection of instances.
Cf. on 1171.
69 f. οὔθ᾽ ὑποκαίων κτλ. Casaubon’s emendation ὑποκαίων for ὑποκλαίων 15
almost universally accepted, and rightly so. Furthermore, it is very probable
that Schütz was correct in changing ὑπολείβων into ἐπιλείβων, relying princi-
pally on a very similar passage in the Niobe, (fr. 161) μόνος θεῶν γὰρ Θάνατος
οὐ δώρων ἐρᾶι, οὐδ᾽ ἄν τι θύων οὐδ᾽ ἐπισπένδων ἄνοις. What ὑπολείβειν might
mean here no one can explain (see also on this point Farnell’s article referred
to below), while ἐπιλείβειν makes perfect sense ; Homer (and Apollonius after
him) uses it of pouring the wine-offering over the burning sacrifice. The
confusion of ém- and ὑπο- is common; thus in the quotation from Hesiod
"Epya 122 f. in both Pl. Crat. 398 a and Aristeides, ii. p. 230 Dind., the MSS
give ὑποχθόνιοι instead of ἐπιχθόνιοι, in A. Sept. 588 the majority of MSS have
ἐπὶ for ὑπὸ, in Pers. 395 three MSS have ὑπέλεγεν for ἐπέλεγεν, in Cho. 715
ἐπευθύνωι is written for ὑπευθύνωι; cf. E. Hipp. 1089, Ar. Knights 459, and
Tucker’s app. crit. on A. Sept. 575 Tucker. Thirdly, the words οὔτε δακρύων
which stand in the MSS after ὑπολείβων, introducing an idea entirely foreign
to the context (this is well shown by Farnell), were recognized as a gloss by
Bamberger (Oduscula, 38), Paley, and Hermann, who says ‘interpretatio
adscripta aut ad οὔθ᾽ ὑποκλαίων, aut fortasse ad οὔθ᾽ ὑπολείβων᾽. The latter
assumption seems unlikely; the explanation of the Homeric δάκρυα λείβων
quoted by Hermann from Hesychius is not to the point.
As for ἀπύρων ἱερῶν, all a priori probability would lead one to suppose that
here as elsewhere it is a technical expression, meaning bloodless sacrifices,
just as it is in Pind. Ol. 7. 48 (6 years earlier than the production of the
Oresteia) and quite generally elsewhere (the material is collected by P. Stengel,
Griech. Kultusaltertiimer, 3rd ed., 1920, 102). Many commentators indeed
“have found in this expression a completely different sense: ‘dupa ἱερά apud
Aeschylum metaphorice dicta sunt, ut sacra igne carentia ea intellegantur,
42
 COMMENTARY lines 69.

quae irrita sunt impieque facta’, says Hermann, and he is followed by the
majority of editors down to our own time, including Kranz, Hermes, liv,
1919, 302 n. 1: ‘nichtbrennender Opfer strengen Groll.’ This interpretation
of ἄπυρα ἱερά as ‘a sacrifice that will not burn’ (Verrall), ‘des offrandes dont
la flamme ne veut pas’ (Mazon), seems to be essentially at variance with
Greek linguistic usage and Greek ideas of sacrifice. Editors have fre-
quently contented themselves (as Mazon does) with a reference, following
Hartung's example, to S. Ant. 1006 {.; but how can a passage which treats
of ἔμπυρα and βωμοὶ πάμφλεκτοι throw any light on arupa? Again, from the
gloss in Hesychius ἀπύρου: ἀθύτου. Σοφοκλῆς Mvoois (fr. 384 N. = 417 P.),
which was cited by Hermann and has repeatedly been cited since, no deduc-
tions whatever can be drawn regarding the technical term ἄπυρα ἱερά, as
Farnell, C.R. xi, 1897, 296, has demonstrated in an article ‘On the Interpreta-
tion of Aesch. Ag. 69-71’ (apart from Wilamowitz's brief note, this article is
much the most helpful contribution to the understanding of this vexed
passage). It is certain therefore that ἀπύρων ἱερῶν here is used not in any
metaphorical sense, but with its normal technical meaning, as indeed the
scholiast understands it: τῶν θυσιῶν τῶν Μοιρῶν καὶ τῶν ᾿Ερινύων, ἃ καὶ
νηφάλια καλεῖται (cf. Eum. 107). From this it follows that the offering of
ἄπυρα ἱερά stands on the same footing as ὑποκαίειν and ἐπιλείβειν, and that
it is therefore impossible to make ἀπύρων ἱερῶν dependent as a subjective
genitive on ὀργάς (‘nichtbrennender Opfer strengen Groll’), as is still often
done. Such a personification of the sacrifice is alien to Greck religious
thought, as Farnell, loc. cit., rightly says. There is, however, still another
argument by which it can be made probable that what is meant here is the
ὀργαί of the gods. In extant literature παραθέλγειν occurs only here, but
once elsewhere Aeschylus uses a derivative of this rare verb, and that,
too, in a passage which has the closest possible connexion in theme with
ours, Suppl. 385 f. μένει τοι Ζηνὸς ἱκταίου κότος, δυσπαράθελκτος παθόντος οἴκτοις.
The nominative here is a completely convincing correction οἱ Schütz’s for
the meaningless δυσπαρθέλκτοις of the MS; παθόντος is best explained by
παθὼν δέ τε νήπιος ἔγνω and Cho. 313 δράσαντι παθεῖν. If a man has offended
against one of the basic laws of religion and morals, against the rules
which are watched over by the supreme god as Ζεὺς ikeratos (Suppl. 385)
or Ζεὺς ξένιος (Ag. 61), then neither lamentation (οἴκτοις Suppl.) nor sacrifice
(Ag.) can appease the ill will and the wrath (κότος Suppl., ὀργαί Ag.) of
Zeus; it remains δυσπαράθελκτον. This agreement cannot be accidental in a
sentence that expresses so strongly one of the most fundamental convictions
of the poet (cf. Wilamowitz, Interpr. 202). It is quite possible that he
coined the word παραθέλγειν for contexts such as this (they may well have
occurred in other plays); this is the more likely as the new-formed word
seems to bear some polemical reference to the Homeric παρατρωπᾶν. These
sentences of Aeschylus are in most decided conflict, at least in regard to
certain definite offences, with the convictions of popular belief and of the
poet of the Atrai (I 497 ff.), whose lines are attacked on the same grounds
by Plato's Adeimantos (Rep. 2. 364 d): στρεπτοὶ δέ re kai θεοὶ αὐτοί... καὶ
μὲν τοὺς θυέεσσι καὶ εὐχωλῆισ᾽ ἀγανῆισι λοιβῆι τε κνίσηι re παρατρωπῶσ᾽ ἄνθρωποι
λισσόμενοι, ὅτε κέν τις ὑπερβήηι καὶ ἁμάρτηι. καὶ ydp τε λιταί eiat krÀ.! Cf., too,
1 The account in y 146 ff. is quite different: Agamemnon intended ῥέξαι ἱερὰς ἑκατόμβας
43
lines 69 f. COMMENTARY
Prom. 34 Διὸς γὰρ δυσπαραίτητοι φρένες and 184 f. ἀκίχητα γὰρ ἤθεα καὶ κέαρ
ἀπαράμυθον (cf. the δυσπαράθελκτος discussed above) ἔχει Κρόνου παῖς, and
Eum. 384 where the Erinyes are called δυσπαρήγοροι βροτοῖς ; see also the
note below on 396. A compound of deAyew is appropriate for the sense
aimed at here; Peitho is θέλκτωρ Suppl. 1041 (cf. Prom. 172 f., Eum. 886). The
idea that sacrifices are powerless to reconcile the gods to offences is later
very clearly expressed by Diphilos, Plaut. Rud. 22 ff. atque hoc scelesti in
animum (si) inducunt suom, Iovem se placare posse donis, hostiis, et operam
et sumptum perdunt, etc. ; further instances of the same conviction are given
by F. Marx, ad loc. (Abh. d. Sächs. Akad. d. Wiss., Phil.-Hist. Kl. xxxviii,
Nr. 5, 1928).
The idea therefore is simple: ‘neither libation nor burnt offering nor
sacrifice without fire can turn aside and appease the unyielding wrath.’
Farnell, loc. cit., and Wilamowitz agree in essentials with this explanation
and the restoration of the text on which it depends. In one point only
Farnell went astray, by taking ἐπιλείβων (as he also reads) ἀπύρων ἱερῶν
together: ‘making a libation of, or from, fireless offerings.’ (H. Stephanus
long ago seems to have sought for something of the kind; he says ‘valde
desideraverim verbum quod cum ἀπύρων ἱερῶν iungatur’.) A decisive objec-
tion to this, apart from the difficulty of the construction, lies in the fact
that ἐπιλείβειν (cf. p. 42) clearly means the pouring of a libation over burnt
sacrifice. It seems therefore likely that in our text an expression governing
ἀπύρων ἱερῶν is lost, in other words that the gloss δακρύων has not been simply
incorporated without injury to the original text (cf., e.g., A. Suppl. 552, Pers.
6) but has replaced part of it (as, e.g., in 111 the gloss δίκας has driven out the
original καὶ χερὶ). This is the view of Wilamowitz (cf. also his translation).
But his attempt to retrieve the original by keeping as closely as possible to
the ductus litterarum! of δακρύων is not very convincing, since this word arose
in all likelihood not as a misreading but as a gloss or perhaps an arbitrary
addition. From the point of view of grammar, Wilamowitz’s conjecture
(οὔτε δι’ ἁγνῶν ἀπύρων ἱερῶν κτλ.) is unobjectionable, but ἁγνῶν seems to me
otiose. 1 think it likely that a noun in the dative has fallen out on which
ἀπύρων ἱερῶν depended. I do not venture to reconstruct what is lost ; purely
exempli gratia one might insert οὔτ᾽ ἐπαοιδαῖς ἀπύρων ἱερῶν. The thrice-
repeated οὔτε places all three clauses on the same level, although the first
and second both refer to ἔμπυρα, but that need cause no difficulty ; the poet’s
purpose is strongly to emphasize the uniform futility of sacrifice of every
kind, so he introduces first the most widespread type of offering, burnt
offering, indicating it in two ways, and then adds the other type, sacrifice
without fire.
71. Schol. (on 69) λείπει τὸ τίς. Headlam wrongly says: 'παραθέλξει without
τις is strange: perhaps we should read παραθέλξεις᾽ ; for the right view see
J. Vahlen, Opusc. ii. 512, Wilamowitz, app. crit.: ‘indefinitum pronomen
in order to appease Athene's anger, νήπιος, οὐδὲ τὸ ἤιδη, ὃ οὐ πείσεσθαι ἔμελλεν" οὐ γάρ τ᾽
αἶψα θεῶν τρέπεται νόος αἰὲν ἐόντων. The contrast between this passage and στρεπτοὶ δέ τε
καὶ θεοὶ αὐτοί was noted in ancient times, cf. Porphyrius on I 497 (p. 140. 24 Schrader) and
on y 147 (p. 34 Schrader).
1 So, too, Ahrens (pp. 244 ff.): οὐδ᾽ ἀνρύων (or ἀναρύων). His proposed reading breaks
down, apart from other considerations, because it necessitates taking ἀπύρων ἱερῶν ὀργάς
together : his extremely artificial way of understanding the genitive need not be discussed.
44
 COMMENTARY line 72

latet in verbo.’ This usage is widespread in the early language; cf. Wilamo-
witz, Berl. Sitzgsber. 1909, 823 n. 2, and Interpr. 14 n. x and app. crit. on A.
Eum. 234 (also Blass, ad loc.);' H. Jacobsthal, Idg. Forsch. xxi, Beiheft,
P. 133; Ad. Wilhelm, Oesterr. Jahresh. xiv, 1911, 211; Wackernagel, Syntax,
i. 112 ; O. Schroeder, Appendix (1923) to Pind. Ol. 14. 22 ff. ; and my remarks,
Deutsche Literaturzeitung, 1927, 860. It should, however, be noticed that
often (and so here) the unexpressed agent. is not really ‘undefined’ but ‘the
person concerned’. As a rule, the reference is made clear by the context.
δὲ in tus vocat, ito. Ni it, antestamino eqs. is a case in point. Cf. below
on 391 ff.
72. ἀτίται. The corrector of the Mediceus, as his marking dréra shows, took
this as a dative, and he was followed by Robortello and the early editors and
again by Hermann; the reading in V, drirasa, seems to point to arira. in an
earlier copy. F and Tr give ἀτίται and ἀτιταὶ respectively as a plural (with
the gloss βλάβην ἔχοντες [deriving ἀτίται from ἄτη, cf. Hesychius ἀτίζων,
Etym. M. arw] dmó* τοῦ γήρως in F, and the glosses βεβλαμμένοι and τῷ
γήρᾳ in Tr), which is adopted by Turnebus and since Schiitz by the majority
of editors, and is clearly right. The meaning is disputed. The passive inter-
pretation ‘honoris expertes’, ‘unhonoured’, was put forward by Pauw and
again by Wellauer; Lobeck, Paralip. 428 n. 41, attempted to justify it
on grammatical grounds; it has passed into L-S, and is indeed accepted
by Ernst Fraenkel, Nomina agentis, i. 185 (followed by E. Williger, Sprachl.
Unters. 52). This interpretation is unlikely: it forces us to assume a passive
sense in an agent noun and, moreover, runs counter to the meaning of
ἀτίτης in the only other passage where it occurs, Eum. 257 (it is surprising
that Blass, who rightly understood the word there, remarked ‘anders
Ag. 72’). The truth was detected by Ahrens (p. 248 f.) ; his discussion, details
apart, is final, and the evidence for réras which has since come to light in
Crete can easily be fitted into it. In Et. 257 the sense of ματροφόνος ariras
is without question 'qui poenas non luit' (Weil), and the connexion with
rivo), τίνομαι is unmistakable. To the same root belongs Cho. 67 riras φόνος,
however one reconstructs in detail its derivation from the verb. The general
sense there is given in the scholia by ὁ τιμωρός; Ahrens (and Wilamowitz
in his commentary of 1896 on Cho. 67) regarded the original meaning as ‘he
that pays', but this is by no means convincing, and the derivation not from
τίνω but from τίνομαι is much more probable (cf. Ernst Fraenkel, Nomina
agentis, i. 183). The τέται in the inscriptions from Gortyn in Crete have
rendered intelligible the gloss in Hesychius riraı . . . κατήγοροι τῶν ἀρχόντων;
they form a judicial commission of control, which calls other officials to
account for failure to perform their legal functions during their term of office,
and can bring about their condemnation to monetary penalties (cf. Busolt-
Swoboda, Griechische Staatskunde, 748, where in note 4 the evidence is con-
veniently collected). In the name of these officials the derivation from
1 In that passage Weil, as early as 1861, interpreted the MS reading correctly as προδῶς
(he should, however, have left the εἰ alone) and stated: ‘haec in universum dicta esse,
neque ferri posse primam personam, satis ostendunt verba ἐν βροτοῖσι κἀν Geois.’ Verrall,
Sidgwick, Headlam, Murray (? perhaps a misprint: he says in the apparatus ‘cf. Ag. 71’)
acquiesce in the first person προδῶ, although the whole character of this concluding couplet
points clearly to a general maxim.
2 This is the reading of F, and not ἀντὶ as printed by Vitelli-Wecklein (p. 336).

43
line 72 COMMENTARY
τίνομαι can hardly be doubted, whether one prefers the interpretation
‘he who punishes’ (so Ernst Fraenkel, Nomina agentis, i. 181) or the other
*coactor, exactor, he who exacts payment! of the fine’ (so Schwyzer, Exempla,
P. 454; Bechtel, Griech. Dial. ii. 793); the latter might be supported by
phrases like Collitz-Bechtel 4979. 57 καὶ τὸν Tírav, ai μὴ ᾿στείσαιτο (cf. the
certain restoration in 4982. 5 róvs riravs [ἐστείσασθαι)). Aeschylus need have
known nothing of the existence of this official title or indeed of the word
τίτης, Tiras; it was open to him at any time to form from the stem rw- the
corresponding agent noun. Very much bolder is his ἀτίτης, 'one who does
not, or cannot, pay', because of the extreme rarity of the privative prefix
applied to such formations. For the language of poetry, however, these
limitations were not insuperable. Ahrens quotes a line of Hesiod which for
our purposes is very instructive, Erga 355 δώτηι μέν τις ἔδωκεν, ἀδώτηι 5
οὔ τις ἔδωκεν. Just as Hesiod by the side of an unusual δώτης set a very
peculiar ἀδώτης (in the same verse, of course, and to secure an antithesis),
so in all probability Aeschylus developed the rare word rirms to an even
bolder ἀτίτης (one would gladly know whether he used either or both outside
the Oresteia). The meaning ‘he that does not pay’ is in Ag. 72 applied to a
special sphere; on this point Ahrens is probably right in the main: 'As
τίνω, ἀποτίνω can be applied to the discharging of any kind of obligation,
SO drirns, le. 6 μὴ ἔχων ἀποτεῖσαι, can be used not only of the ἀδύνατος
χρήμασι but also of the ἀδύνατος σώματι, and this meaning is in the highest
degree suitable here, i.e. unft for military service.’ We are of course familiar
with the view that military service is the discharging of a debt (to the State),
but it is not likely that we should often find this idea expressed among the
Greeks. Headlam refers to Sept. 20 χρέος τόδε, where, however, other meanings
than the one which he assumes are possible. The conception is in itself so
natural that one scarcely needs a parallel. Yet we may perhaps (cf.
Wilamowitz, Interpr. 166) recall the related idea on the strength of which
Demosthenes (21. 166), though only for the sake of a pointed and spiteful
remark, is able to say: οὐδένα... τρόπον ἄλλον ἐν τοῖς ἱππεῦσιν αὑτὸν ἀτελῆ
ποιῆσαι στρατείας δυνάμενος ταύτην εὕρηκε Μειδίας καινὴν ἱππικήν τινα πεντη-
κοστήν. The expression ἀτελῆ στρατείας, with which we are concerned, is
illustrated by the words (165) οὗ ὁ νόμος προσέταττεν, ἐνταῦθα τοῖς σώμασιν
αὐτοὶ ληιτουργεῖν ἠξίουν. Cf. Hdt. 3. 67. 3 ἀτελείην. . . orparnins καὶ φόρου
(where, however, the second genitive may have helped to influence the choice
of the word) and from the Decree of the Amphiktiones of Delphi (first
τ These translations should perhaps be avoided since they suggest a confusion with
πράκτορες (or the ἐρευταί who in Crete correspond to them): the πράκτορες in Athens and
elsewhere are not judicial authorities but subordinate executive officials (cf. Wilamowitz,
Aristoteles u. Athen, i. τοῦ n. 20).
2 The equally Hesiodic ἀβούτης does not help our purpose at all; it is =*dBous as, e.g.,
χιλιοναύτης = χιλιόναυς. For the Homeric πολυβούτης, a poetical coinage, see E. Risch,
Wortbildung der homer. Sprache, 31.
3 Wilamowitz, ad loc., says: ‘ ἀδώτης is not really supported by linguistic analogy, but,
preceded as it is by δώτης, it is an ingenious if bold coinage.’ So he considers δώτης quite
normal. Exactly the opposite view is held by Ernst Fraenkel, Nomina agentis, i. 118, who
with good reason regards δώτης as very irregular and explains it as ‘only formed subse-
quently from the compound form [aöwrns]’. There is no need here to make a decision on
this point: δώτης, like ἀδώτης, may be an artificial coinage, and both words may have
been invented for the sake of mutual contrast in this passage.
46
 COMMENTARY line 77

quarter of the 3rd century B.c.) 70 11. 12. 1132. 14 εἶναι δὲ τοὺς τεχνίτας ἀτελεῖς
orparelias πεζικᾶς] καὶ ναυτικᾶς.
73. The expression τῆς τότ᾽ ἀρωγῆς confirms the conclusion that must be
drawn from the structure of the whole, namely that the sentence beginning
with 72 (ἡμεῖς δέ) refers back to the beginning (40 μέν) of the parodos
(cf. 47 ἀρωγήν).
75. ἰσόπαιδα: for the occurrence of compounds of this type in the tragedians,
cf. on 1442 f.
νέμοντες: it seems unnecessary to give the verb here the sense of ‘sup-
porting’, ‘keeping upright’; it means only ‘setting in motion’, ‘guiding’,
as in Pind. N. 6. τς πόδα νέμων, A. Ag. 802 οἴακα, 685 γλῶσσαν (cf. ad loc.).
The idea of propping up is not introduced until we come to ἐπὶ σκήπτροις.
76 ff. In the two clauses here related paratactically with re... . re all the
emphasis rests upon the second, so that we should naturally subordinate
them with ‘as ...so’. Many commentators have seen this. On this type of
‘copulative [better ''paratactic"] comparison’ cf. Hermann on A. Sept.
584 ff. (565 Herm.); J. Vahlen, Opusc. i. 303 ff.; Wilamowitz on E. Her.
ror f., in his commentary on A. Cho. 246-63, and in his Pindaros, 471; cf.
also Denniston, Particles, 515.
76. veapös: a favourite word with reference to the fresh tender youth of
both animals and men, cf. on 359.
77. Hermann’s interpretation of ἀνάσσων as ἀνάισσων was at once adopted
by Schiitz (in his second edition) and since then by the majority of editors.
Hermann (on Ar. Clouds 996) gave no explanation but contented himself
with the comment ‘verbum ἄισσειν saepius obscuratum librariorum imperitia’ ;
in support he might well have quoted (instead of an extremely doubtful
conjecture of his own in Pind. Ol. 13. 107) A. Pers. 96, where he himself, with
Turnebus, Victorius, Brunck, rightly regards the ἀνάσσων of the MSS as
ἀνάισσων. In Ag. 77 the attempt to defend ἀνάσσων has been made again,
following Paley, by Headlam; ‘dvdcowv’, he says, ‘is appropriate to the
marrow, regent in its frame of bone and dominating the vital functions.’
The idea is perhaps possible in itself, though none of the passages which
Headlam quotes can support it, for in Pl. Tim. 73 Ὁ τούτοις σύμπασιν ἀρχὴ μὲν
ἡ τοῦ μυελοῦ γένεσις the meaning of ἀρχή is not ‘rule’ but ‘beginning’; the
precious "Timaeus Locrus’, roo a, does not speak of the marrow in general
but of the brain (ἐγκέφαλον) when he says ἐν ὧι a ἁγεμονία (from Pl. Tim.
45 b, inserted into a context the rest of which is taken from Tim. 73 b, etc.) ;
and Pliny throws no light on it at all. In any case, however, such an idea
would be quite unsuitable here. If ἀνάσσειν is really the function of μυελός,
it cannot be limited to the tender period of youth; yet here, with ὁ νεαρὸς
μυελός preceding, we must have something specially characteristic of im-
maturity. Headlam’s translation well shows to what distortions he was
driven by loyalty to his own view; his rendering is logical, but substitutes for
the arrangement of the clauses in the text something quite different: ‘For
as the marrow, that holds sway within the breast, being youthful is but as
old age. . . .' Wilamowitz' adduces in support of ἀνάισσων a passage from
1 G. Thomson criticizes Wilamowitz : ‘Hippocrates is speaking of a fluid which is very
different from the marrow of the bones.’ But Wilamowitz was well aware of what was
thought about the nature of the marrow in antiquity; cf., e.g., Arist. De part. animal. 2. 2,

47
line 77 COMMENTARY
Hippocrates π. νούσων 1. 20, vi. 178 Littré ; but this is irrelevant, for there the
words τὸ ἀπὸ τῆς σαρκὸς ὑγρὸν ἀΐσσει διὰ τῶν φλεβίων simply resume the pre-
ceding sentence ὁκόταν... τοῦτο αὐτὸ τὸ ὑγρὸν ὑπερθερμανθῆι ὑπ᾽ αὐτῆς τῆς
LA é > M , 1
gapkos . .. σκίδναται ἐς τὰς φλέβας, and the process there described has
nothing in common with Ag. 77. But a passage which is thoroughly relevant
is Arist. Hist. anim. 3. 20, p. 521 Ὁ 8 (quoted by van Heusde, who nevertheless
attacked those who accept avdıoawv) : γίνεται δὲ ἐν μὲν τοῖς νέοις αἱματώδης
πάμπαν 6 μυελός, πρεσβυτέρων δὲ γενομένων ἐν μὲν τοῖς πιμελώδεσι πιμελώδης,
ἐν δὲ τοῖς στεατώδεσι στεατώδης. Cf. also in general Plato, Tim. 81 Ὁ νέα μὲν
οὖν σύστασις τοῦ παντὸς ζώιον... συμπέπηγεν δὲ ὁ πᾶς ὄγκος αὐτῆς ἁπαλός,
ἅτ᾽ ἐκ μυελοῦ μὲν νεωστὶ γεγονυίας, τεθραμμένης δὲ ἐν γάλακτι. Aeschylus is
thinking then of the fluid marrow of children which has not yet hardened,
and he uses of it the expression ἀνάισσει, as for instance Homer, X 147 f.,
says ἔνθα δὲ πηγαὶ δοιαὶ ἀναΐσσουσι Σκαμάνδρον. The verb ἀνάισσειν conveys
an impression of impetuous movement, as is fitting in this connexion: the
fluid marrow in children behaves like a child itself, leaping up and darting
about in continual excitement.' The poet's point is the soft and feeble
nature of childhood, especially in reference to Ἄρης δ᾽ οὐκ ἔνι. For this
reason, instead of confining himself with anxious pedantry to what is
anatomically correct, and speaking of marrow in the bones, he says στέρνων
ἐντός, and thus indicates in a picturesque manner the proper home of strength
and spirit (or their opposites), as, e.g., in Soph. fr. 196 N. (— 195 P.) ἀνδρῶν
γὰρ ἐσθλῶν στέρνον οὐ μαλάσσεται, E. Phoen. 134 Ἄρη δ᾽ Αἰτωλὸν ἐν στέρνοις
έχει.
78. "Apns δ᾽ οὐκ ἔνι χώραι. It is out of the question to regard ev? as ἃ
preposition (evi χώραι, as, e.g., Kirchhoff, Wecklein, Verrall, Headlam [cf.
also his book On editing Aeschylus 124], G. Thomson) ; for ἐνὶ equivalent
to ev does not occur in Drama. On the other hand, é = ἔνεστι is found, e.g.,
in Pers. 738, Prom. 294, and several times in Sophocles and Euripides. So
ἔνι must be taken as ἔνεστι. If this view needed any further support, it could
be readily found in the parallel expression Suppl. 749 γυνὴ μονωθεῖσ᾽ οὐδέν"
οὐκ ἔνεστ᾽ Ἄρης (to which there is possibly a reference in 5. El. 1243 f. ὅρα
γε μὲν δὴ κἀν γυναιξὶν ὡς Ἄρης Eveorw). The real difficulty of the passage lies
in the interpretation of χώραι. Hermann with his unsurpassed knowledge of
the language recognized this, and also quite clearly formulated the difficulty.
χώρα is found by no means rarely in the sense ‘the place in which a person
or thing normally is to be found or resides’, so that οὐκ ἐν χώραι ἐστίν by
itself well might mean ‘in suo loco non inest; is not in its proper place’. This
explanation (Stanley had already rendered ‘non inest in suo loco’) was put
forward and illustrated by Blomfield; he is followed by Passow and in sub-
stance by the earlier editions* of L-S. The idea ‘Ares remains not in the place
p. 647 Ὁ 13, where the μυελός is put in the list of ὑγρά, ibid. 2. 6, p. 651 b 20 ἔστι δὲ καὶ d
μυελὸς αἵματός τις φύσις, Hist. anim. 3. 20, p. 521 Ὁ 4, and the analogous passages quoted in
Bonitz’s Index. The same idea is found, e.g., in Seneca, Nat. quaest. 3. 15. 2.
1 Shakespeare's Venus says (Venus and Adonis, 141 f.): ‘My beauty as the spring doth
yearly grow, My flesh is soft and plump, my marrow burning.’
2 For the accents in the MSS see app. crit.
3 It looks as though Elmsley’s observations (on E. Heraclid. 893) were as much disregarded
by more recent scholars as the use of good dictionaries.
+ In L-S? the words have been added ‘or perh. in the land, cf. Ar. Lys. 524 [οὐκ ἔστιν ἀνὴρ
48
 COMMENTARY line 78
to which he belongs’, ‘is not at his post’, seems to me wholly inappropriate
here ; after what has gone before one could only speak like that if one meant
to suggest that the breast of children is the natural post for the god of
war, which he (like a deserter?) has abandoned. Hermann also rightly saw
that the interpretation ‘is not in his own proper place’, ‘dwells not in his
wonted seat’, is incompatible with ἔνε : ‘nam é si legitur, omnis vis in hoc
verbo est.” He therefore understood χώρα in the sense which we find in
Theognis 152 and 822, where χώρην μηδεμίαν and χώρη ὀλίγη are used with
reference to the man who belongs to no class, who has no place (in society),
counts for nothing, and is οὔτ᾽ ev λόγωι οὔτ᾽ ev ἀριθμῶι (Passow, L-S, Dindorf
in his Thesaurus, following Portus, class the Theognis passages with Xen.
Anab. 5. 7. 28 ev οὐδεμιᾶι χώραι ἔσονται). Yet the interpretation at which
Hermann thus arrived, ‘non inest, non censetur in loco aliquo et numero’,
in no way does justice to the sense here demanded (Ahrens, p. 250, is right).
The gloss in the scholion (in M) τῶι τόπωι ἐκείνωι would certainly fit the
context; the assumption of the glossator evidently is that χώραι takes up
the preceding στέρνων. But even if that were in keeping with the meaning of
χώρα, it would surely be necessary, in the case of such a strong anaphora,
to have the article, τῆι χώραι. Of the many conjectures none is really
convincing. Kayser read Ἄρει δ᾽ οὐκ Evi χώρα, and Gilbert (with a slight
variation) Apews δ᾽ οὐκ ἔνι χώρα; the latter conjecture is adopted by Weil
and Wilamowitz.! (With some astonishment one sees Wilamowitz appealing
for support to the same Theognis passage (822) called in by Hermann to the
defence of the MS reading.) Against this proposal it may be remarked, first
that the special force of ἔνι is again weakened in the way shown by Hermann
to be inadmissible, and secondly, that even so no satisfactory sense for χώρα
results. But apart from that, any proposal to attack the MS reading Ἄρης
δ᾽ οὐκ év is rendered doubtful a priori by the οὐκ ἔνεστ᾽ Ἄρης of Suppl. 749
and the corresponding expression (see above) in S. El. 1243í.; it seems
highly probable that in each of these three passages it is the same brief
and pregnant phrase that is used to express the unfitness (in Sophocles the
unexpected fitness) of women or children for war. Our result then is that,
unless a convincing explanation of χώραι can be found in some other direc-
tion, this word.—and this alone—is to be regarded as corrupt. In this I agree
with Ahrens, p. 250 f., though I cannot accept his conjecture (χωρεῖ, and
Ἄρης δ᾽ οὐκ ἔνι in parenthesis). Weil's idea (in the Addenda to his first
edition of the Eumenides) of replacing χώραι by χλωρῶι (χλωρᾶι M. Schmidt)
is at first sight tempting, especially in view of the following φυλλάδος ἤδη
κατακαρφομένης, but can hardly be maintained, for although when used
in certain contexts yAwpós is glossed by ἁπαλός and veapds, yet the viridis
aetas is not that of children under age but that of youths and men in the pride

ἐν τῆι χώραι], taking into account Murray's note. This explanation had already been
proposed by S. Butler: ‘os vero martius ex hac terra abiit (una cum Agamemnone profecti
sunt ii quorum viget aetas)', and revived by Arnold. Against this Conington argues quite
rightly : ‘but the whole tenor of the expressions used seems to shew that the thought of
those who are gone has been for the time dismissed, and that the mind of the old men is
engaged in pursuing a reflection on the parallel between childhood and old age generally,
suggested by the sense of their own weak and helpless state.'
i He paraphrases, however, Glaube d. Hell. i. 322, n. 1, ‘the man who is no longer fit for
military service has no Ares in him’.
4872.2 E 49
line 78 COMMENTARY
of their strength: Suid. χλωρόν" τὸ ἀκμάζον. Μένανδρος. Thus the difficulty
remains unresolved.'
79. ὑπεργήρων. Sidgwick and Wilamowitz, with the older critics (Schiitz,
Blomfield, Hermann, Schneidewin, Enger in his re-edition of Klausen,
and Nägelsbach), have rightly adopted the reading of Tr (‘Tricl. non e propria
coniectura’ says Wilamowitz; that seems to me unlikely, and I agree with
Ahrens, p. 251: ‘It is very probable that the neuter originates from Tri-
clinius, for such an alteration was of course not too much for his ingenuity’).
On the other hand, many editors extract τί θ᾽ ὑπέργηρως from the τίθιπεργή-
pws of the Mediceus, and put it in the text. Wilamowitz’s note, which
contests both the question itself and the re in the question,’ seems to me
irrefutable ; to continue with a question within the framework of the ‘copula-
tive comparison’ (cf. above on 76 ff.) is quite impossible. Headlam’s reference
to the questions in Pind. P. 8. 95 is unfortunate. The τίθιπεργήρως in M is
probably a further stage in corruption from the τόθιπεργήρως preserved in
VF. The false masculine form in MVF is convincingly explained by Enger
and Wilamowitz as an intrusion under the influence of the following ἀρείων.
It is true that others (from Wellauer onwards; cf. Ahrens, p. 251, who,
however, expresses himself with caution) have taken precisely the opposite
view, and found in ἀρείων clear evidence that the masculine ὑπεργήρως 15
correct ; so again recently G. Thomson (ii. 367) : 'The reading τό θ᾽ ὑπέργηρων,
accepted by Wilamowitz and others, cannot stand unless ἀρείων is altered
to ἄρειον᾽, but this argument is deceptive. It is so natural to continue with a
constructio κατὰ σύνεσιν (viz. ἀρείων) that no parallels are needed, but it may
be worth while recalling that in 738 ff. the subject of ἐλθεῖν is indicated by
four neuters in succession, which are immediately followed by zapaxAtvaca.
Both there and here, the transition to the natural gender of the subject is
made in a new clause connected by δέ.
That the paroxytone accent on ὑπεργήρως given in MVF must be retained
was emphasized amongst others by Hermann and Ahrens (p. 251), and
renewed attention was drawn to it by Wackernagel, Nachr. Gótt. Ges. 1914,
123 n. 1 (for an earlier discussion of the question, with the ancient authorities,
see H. W. Chandler, Introd. to Greek Accentuation, 2nd ed., p. 156, § 546).
φυλλάδος κατακαρφομένης : Archilochus 113 D., of the loved one growing
old, οὐκέθ᾽ ὁμῶς θάλλεις ἁπαλὸν χρόα" κάρφεται yàp ἤδη. Unquestionable
borrowing from Archilochus by Aeschylus is to be found in Pers. 763, Cho. 123.
80. τρίποδας ὁδούς: that we have here an echo of the riddle of the Sphinx
(for the evidence see Ed. Schwartz, Scholia in Eur. i. 243) was seen long ago
by Spanheim (his notes are printed in C. G. Haupt’s edition); cf. also C.
Robert, Oidipus, i. 57, and see below on 1258. τρίπους in this sense probably
comes from Hesiod, Erga 533, cf. on 149 (p. 90).
81. ἀρείων: an epic word, later taken over by the lyric poets (Pindar,
Bacchylides) ; its only other occurrence in Drama is in Sei. 305 (lyric
iambics).
* I find it impossible to accept the solution suggested in C.R. lviii, 1944, 35.
? None of the editors who read τί θ᾽ ὑπεργήρως has thought it necessary to say how in this
context re after τί can be accounted for. It can be seen, e.g. from Denniston, Particles,
533 f., that the Homeric τίς 7’ dpa and the like are quite different.
3 This was correctly understood by Triclinius, who says in his own scholion : πρῶτον δὲ
οὐδετέρως εἰπὼν τὸ ὑπέργηρων, εἶτα ἀρσενικῶς ἐπέφερε τὸ ἀρείων πρὸς τὸ σημαινόμενον.
50
 COMMENTARY lines 83 tf.

82. ἡμερόφαντον is only found here. In this context -$avrov is perhaps

meant to suggest the idea of φάσμα.
Wilamowitz on E. Her. 111 shows it to be probable that the description of
old age in 79 ff., though given in a nobler style, borrows its characteristic
elements from proverbs and other popular sayings.
83 ff. At first sight it may seem obvious to infer from the words σὺ δέ,
Τυνδάρεω κτλ. that the queen has just come out from within the house.
This is the interpretation of Hermann, which has been followed by Paley
and others: it was opposed by Enger! (see his revision of Klausen's commen-
tary on 103), and afterwards by T. Maguire, Notes on the Ag. of Aesch. (1868),
το, by Hense in the appendix to Schneidewin's commentary on 83 ff., and
by Wecklein (Studien zu Aesch., 98 ff., and in his annotated edition), who was
followed by Verrall; Headlam, too, in his translation makes the queen's
first entrance coincide with 254 (G. Thomson abandoned Headlam's view and
followed the lead of J. T. Sheppard), and so do Mazon and G. Méautis,
Eschyle (1936), 131, 145. W. Kranz, De forma stasimi (Diss. Berlin 1910), 63 ff.,
went thoroughly into the whole matter, with the result that Wilamowitz,
Interpr. 164, gave up his earlier belief in the correctness of Hermann's view.
If Clytemnestra were on the stage at the moment of the Chorus's questions
(83 ff.), she would be bound to answer (as Enger had already rightly argued), at
latest after the end of the anapaests; or if she did not, some reason must have
been given for her silence. Wilamowitz (in his earlier attempts) endeavoured
to meet this objection by assuming after 103 the loss of a number of lines in
which the Chorus suggested a motive for Clytemnestra's silence. Another
solution was advocated by Hermann, who supposed (on 1o3) that Clytem-
nestra first went up to the altars in front of the house, then, during the
anapaestic part of the parodos, went off into the city to offer sacrifice there
too, and finally returned at the end of the ode (256). Murray (in his translation ;
cf. also C.R. liii, 1939, 11 and his Aeschylus, 1940, 209 ff.) makes the same
assumption with a slight variation (Clytemnestra goes into the house at the
end of the anapaests, and returns at the end of the lyrical part) ; yet another
variation of the same view is A. Y. Campbell's (after 103: 'Silet regina, tergo
spectantibus obverso’ ; after 257: ‘Regina ad senes se convertit’). But if this
were the case, we should be assuming ad hoc a peculiarity which is scarcely
believable in Greek tragedy. For it never happens that a leading character
enters the scene for the first time and then leaves it again without having
uttered a word. Actually everything is perfectly simple, once it is recognized
that Clytemnestra's first entrance comes towards the end of the lyrical part
of the parodos. The words in 83 ff. are addressed to the queen while she is
still inside the house: the excitement of the elders makes it possible for them
to address Clytemnestra as if she were present. Enger has rightly referred to
the analogy of the parodos of the Ajax. Perhaps the instance of the parodos
of the Hippolytus, to which Kranz called attention, is still more striking.
There we have the women of Troezen, who form the Chorus, standing in front
of Phaedra's house. In the first pair of stanzas they say how they have been
! Before him le Pére Brumoy, Heinrich Voss (or his father as editor?), and W. Sewell in
their translations had placed the queen's first entry after the end of the choral ode (257).
Cf. also Casaubon's observation on 83: 'senes Clytaemnestram compellant absentem . . . et
videntur quasi praesentem interrogare “Quid novi?" etc.’
51
lines 83 ff. COMMENTARY
informed about the sufferings of their queen and what they have been told
of them; then they start afresh in a new stanza with the words (141) ἦ σύ γ᾽
ἔνθεος ὦ κούρα κτλ., and it is only at the end of the choral ode that Phaedra
enters with the Nurse. The fact that the excited questions in 4g. ὃς ff. are
particularly suitable as addressed to the queen in her absence was observed
by Hense (in the appendix to Schneidewin's commentary) ; cf. also Verrall:
"The form of apostrophe in both cases [5. Aj. 134 and Ag. 83] indicates the
like impatience for the presence of the person addressed.'
It is not until this passage that we are told the reason why the Chorus
appear in front of Agamemnon’s house. Αἱ the same time it now becomes
clear that we must imagine some time to have clapsed between the end of the
prologue and the beginning of the parodos. In this interval the queen has
received the Watchman's message and has immediately arranged for the
fire of burnt sacrifice to be kindled at all altars throughout the city (cf.
also 594 ff.). She herself has sent her attendants round to see to these sacri-
fices (87), and sent sacrificial oil from the royal store to be used for them (96).
The Elders have gone to the house of the Atridae, perhaps merely to inquire
the reason for these preparations, but more probably because Clytemnestra
has sent them a message to say that she wishes to see them and to make an
announcement to them. Cf. on 258.
84. Κλυταιμήστρα. This is the only ancient form of the name ; of that there
should be no further question (in spite of E. Bruhn on E. 75A. T. 208 and in
the introduction to his edition of S. EZ. p. 49). Inscriptions on Attic vases, on
"Homeric bowls', and on marble (/G xiv. 93o. 10) agree with the spelling in the
ancient MS of Philodemus, the universal testimony of Latin, and the forms
(brought out by Papageorgiu) in the Mediceus, which are almost unanimous
(for the exceptions see below). Cf. the survey given by Hófer in Roscher's
Lexikon der Mythol. ii. 1230 ff. ; P. Kretschmer, D. griech. Vaseninschriften,
166 ff., and Glotta, iii, 1912, 330 f. ; W. Schulze, Gótt. gel. Anz. 1896, 234 (= KI.
Schr. 697). The etymology put forward by the grammarians παρὰ τὸ κλυτὸν
kai τὸ μήδω [they ought to say μήδομαι] (Etym. M. 521. 17, etc.) agrees with
this evidence. The intrusion of the form with v, which 'is entirely due to the
etymologizing fancies of a late period" (W. Schulze), into the Byzantine
MSS can still be clearly followed. Laurentianus Mediceus 32. 9 (which we
call M in Aeschylus and L in Sophocles) is nearly free from it, as we have said,
but in all later MSS of Aeschylus (Ag. 258 Κλυταιμήστρα in Tr may be mere
accident), Sophocles (including A), and Euripides (including M) the v has won
the day, as it has in all MSS of the Odyssey. The oldest MS of Dio Chryso-
stomus, the Vaticanus V, has twice (11. 43, 46) the form without v, and once
(11. 51, if one may judge by the silence of v. Arnim) Κλυταιμνήστραν. In Pind.
P. 11. 17 all MSS give Κλνταιμνήστρας, while in Nem. το. 6 the oldest MS there
preserved (B) has ‘Yrrepunorpa (cf. O. Schroeder, Proleg. to his ed. maior,
p. 30) like the Mediceus in Schol. A. Sept. 569 (556 Wecklein). The Byzantines
seem to have wavered between the two forms fairly early: in Venetus À
of the Iliad, at A 113, Κλυταιμνήστρης is written but the v taken out with two
suprascript dots; in A. Eum. 116 even M has Κλυταιμνήστραξ (but in the
™ See, however, Von der Mühll's note on Odyssey y 266.
2 Papageorgiu, on Schol. S. El. 95 (p. 106. 11), is wrong when he reports Κλυταιμήστρα.
(In the same note he contrives to quote a reading of M at A. Ag. 585.)
52
 COMMENTARY line 90
dramatis personae of the Eumenides in M the ν is added by the corrector) ;
in S. El. 1473 in L the » of Κλυταιμήστρα is a correction in place of another
letter, as may be clearly seen in the facsimile (only an examination of the
MS itself can decide if a » was first written).
85. τί χρέος: for this expression see Blomfield, ad loc., and Elmsley on E.
Heraclid. 96; cf. L-S s.v. ii. 2. It is essentially equivalent to ri χρῆμα; (cf.
on 1306), ‘Quid rei est ?”, ‘What is the matter?’ (Elmsley) ; but τί χρέος; stands
stylistically on a higher level.
ἐπαισθομένη. This compound, which is very rare elsewhere in early Greek,
is a favourite in Tragedy (cf. on 4 κάτοιδα) ; the only earlier occurrence is in
Hesiod in the Catalogue of Helen’s suitors, fr. 96. 82 Rz., 3rd ed. (overlooked
by L-S).
87. The πειθοῖ of the MSS (in F only slightly distorted by itacism) is right.
Scaliger’s πευθοῖ, which was preferred even by Ahrens, and is still mentioned
by Murray in his apparatus, is not worth considering. The meaning of πειθοῖ
is, of course, ‘suasu’ (Stanley) or ‘persuasione’, not ‘fiducia’ (Schütz). The best
comment is provided by 590 f. φρυκτώρων δία πεισθεῖσα Τροίαν νῦν πεπορ-
θῆσθαι δοκεῖς;
The MS reading! is θυοσκινεῖς (though the first hand οἱ M perhaps wrote
θυοσκνεῖς, for which Ahrens compared Anacreon fr. 3. 3, where for Bergk’s
διοσκέω the MSS have διὸς κνέων or dioorvew). For this Turnebus, with the
help of θυοσκεῖν from Hesychius, restored θυοσκεῖς. This is presumably
right, but in the circumstances it is impossible to be absolutely certain.
What is behind the scholion (ZyoA. ra.) in Tr εὕρηται καὶ θυοσκοεῖς, ὡς ἀπὸ
τοῦ θυοσκόος, ὃ διὰ θυσιῶν τὸ μέλλον κοῶν, ἤτοι νοῶν (for the latter part cf.
Hesychius θυοσκόος, Et. M. 457. 41, Eust. in Od. p. 1601. 3) one cannot say.
Ahrens made an unsuccessful attempt, with much learning and great in-
genuity, to show that the true reading is θυοσκνεῖς. ‘Si recte se habet θυοσκεῖν,
eadem haec analogia est quae in βοηθεῖν, βοηθόος᾽, says Hermann. This
linguistic explanation is accepted, e.g. by Wilamowitz and by E. Boisacq,
Dictionnaire étymol. de la langue grecque, s.v. θυοσκόος. For the etymology
of θυοσκόος see Boisacq, loc. cit., and L-S, for the usage P. Stengel, Die
griech. Kultusaltertümer, 3rd ed., 6o, and L. Ziehen, RE vi A. 736 f. Ziehen is
perhaps right in assuming that the word originally denoted a person who
inspected the flame or smoke of a sacrificial fire. There is no need, however,
to relate this to offerings of incense only; cf. on 1409. It seems rather doubt-
fui whether in Homer ἃ θυοσκόος is concerned with the inspection of σπλάγχνα
(as was assumed by Wilamowitz, Heimkehr des Odysseus, 55) ; for arguments
against it see Stengel, loc. cit. But there would be nothing surprising in
a process by which the narrower meaning of θυοσκόος in course of time gave
place to a more general one, with the result that instead of one typical
detail eventually the whole ritual of offering burnt sacrifices was envisaged.
It is from θυοσκόος (used in this wider sense) that the verb θυοσκεῖν was
probably derived.
90. οὐρανίων: Enger’s conjecture τῶν τε θυραίων is put in the text by
Weil, Wecklein (in his annotated edition), Headlam, Mazon, A. Y. Campbell.
But the other categories of gods mentioned in this passage are all com-
ı What Wilamowitz (followed by A. Y. Campbell and Murray) says about the reading of
V and Tr is wrong.
53
line 90 COMMENTARY
prehensive ones (this, as is well known, is true of the θεοὶ ἀγοραῖοι, with
their xowoßwpia, no less than of the others); and it is not easy to see
how the narrow category of θεοὶ θυραῖοι could be named side by side with
these. The arrangement of the text as it stands is thus explained by Wila-
.mowitz : “ἀστυνόμοι sunt οὖς τὸ ἄστυ νομίζει, quos deinde δημοτελεῖς vocabant.
qui deinde per duo binorum paria explicantur, superi et inferi, et qui caelum
et qui forum tenent.’ Klausen compares Septem 271 f. χώρας τοῖς πολισσούχοις
θεοῖς (cf. Ag. 338 f.), πεδιονόμοις τε κἀγορᾶς ἐπισκόποις, where the clause
πεδιονόμοις τε κτλ. is clearly explanatory. On the contrast between οὐρανίων
and ἀγοραίων Sidgwick observes: ‘by ayopaioı he probably means those who
do specially preside over the city-life; by οὐράνιοι those who do not.’ The
reference of ὕπατοι, as χθόνιοι shows, is purely local; οὐράνιοι refers to a
sphere of activity different from that of the ἀγοραῖοι. Otfried Müller’s
inference that the altars of the market-gods were probably visible on the
stage is unjustified (cf. on sı3); the Elders report what they have seen
on their way to Agamemnon’s house.
95. ἀδόλοισι: the best commentary is provided by a regulation for the
policing of the market in the Mysteries-inscription from Andania (Ditten-
berger, Syll. 736), line 100 6 δὲ dyopavduos 6 ἐπὶ πόλεος ἐπιμέλειαν ἐχέτω, ὅπως
of πωλοῦντες ἄδολα καὶ καθαρὰ πωλοῦντι. “Genuine, unadulterated’ suits here,
too, but at the same time, in connexion with παρηγορίαις and alongside
μαλακαῖς, it preserves its original sense: the pure oil revivifies the flame
without deceit.
παρηγορίαις: this word does not occur again earlier than Apollonius
Rhodius, who uses παρηγορεῖν freely. Whether Aeschylus took the noun
from later Epic or formed it himself from the familiar παρηγορεῖν, like fr. 348
παρηγορήμασιν, we cannotsay. Fomenta (Stanley) is a good rendering, except
that the expression in Greek is more peculiar and suggestive of personifica-
tion.
96. πελανῶι. It has been suggested that the meaning which we have here
is the original one, cf. P. Persson, ‘Beitr. z. indog. Wortforschung’, 517.
Kungl. Hum. Veiensk.-Samf. Uppsala, x, 1912, 748 τι. 1: "The Greek word
denoted various thick fluids (blood, oil, honey, rheum in the eyes, etc.),
and in particular meant a kind of batter, which was thrown into the fire
as an offering or poured out as a kind of libation. . . . The use of the word
does not indicate that ''flat" was its basic meaning, although in course of
time it came to be applied to flat cakes as well.’ In the same way L-S give
as the basic meaning: ‘any thick liquid substance, of various consistency.’
If this view is correct, the rather artificial explanations given by Wilamowitz,
ad loc., and by P. Stengel, D. griech. Kultusaltertiimer, 3rd ed., 99, can be
ruled out; cf., however, Specht, Kuhns Zeitschrift, xii, 1934, 284 ff. Be that
as it may, there can be no doubt that the word here means the libation of
oil, as unequivocally as in other passages of Aeschylus and Euripides the
libation or shedding of blood; cf. also Blass on Cho. 92. Ahrens (p. 262) and
following him Dindorf in his Lex. Aesch. recommend the maintenance of the
1 Or perhaps rather of τὸ ἄστυ νέμονται or νέμουσιν (for the latter cf., e.g., Pind. Ol. 2. 13
ὦ Κρόνιε παῖ 'Péas, ἔδος Ὀλύμπου νέμων ἀέθλων τε κορυφάν xrÀ., where ἀέθλων κορυφάν shows
that νέμων has not merely its local meaning, but is also equivalent to διοικῶν, διέπων, as the
scholiast rightly remarks; thisis true also of ἀστυνόμοι).

54
 COMMENTARY line ror

accentuation πελανός, which the Mediceus almost always exhibits! (here,

Cho. 92, Eum. 265; first hand at Pers. 524, 816; but at Pers. 204 M has
πέλανον) ; the same MS gives πελανοὺς at Ap. Rhod. 1. 1077, 4. 712. As for
Herodas 4. 91, Bücheler, ad loc., at once pointed out that there the papyrus
gives meAavov?, thereby indicating the accentuation seAavóv ; Headlam-Knox,
ad loc. quote Ar. Plut. 661, where the Ravennas (not the Venetus) has
TeÀavós. This accentuation corresponds to the doctrine of Herodian (cf.
Arcadius p. 73. 11 Schmidt), which is followed by L-S.
On βασιλείωι Wilamowitz refers to Sappho 96. 21 D., but here, where the
reference is to the queen's sacrifice, the epithet does not, as there, indicate an
εἶδος μύρου, but has its full possessive sense. In thought, though not gram-
matically, μυχόθεν βασιλείωι go naturally together (interlinear gloss in Tr: ἐκ
τῶν μυχῶν, τουτέστι τῶν θαλάμων τῶν βασιλείων).
98. Following Schoemann (Opusc. iii. 160) and Ahrens (p. 263 f.), Wilamowitz,
Interpr. 163 n. ı (cf. also his note on E. Her. 186), has again recalled that
αἰνεῖν very often contains a 'iudicii voluntatisque adsignificatio' and has
shown that Wieseler's change of αἰνεῖν into αἴνει is necessary: placeat tibi
dixisse. Berti (cf. on 12), pp. 238, 273, attempts in vain to restore our belief
in the connexion Aé£aca . . . παιών re γενοῦ, which was accepted by Hermann
(discussed at length in his note on Vigerus, De Graec. dict. idiot. p. 772; the
rare καί following a participle in X 247—cf. Denniston, Particles, 308—is not
enough to support such a use of re) and others. Hermann referred to Cho.
557, Where most editors agree that δόλωι τε conceals a corruption ; Paley adds
Sept. 45, a verse which has not yet been satisfactorily either emended or
explained (see below on 1432 f.) ; Cho. 864 f. has been corrected. As an alterna-
. tive Paley (in the later editions of his commentary) suggested that ‘perhaps
another re or καί was intended to follow, but the poet was carried away by
the long relative sentence after μερίμνης" ; in this he followed Schneidewin ;
the same view is held by Wecklein (who compares the quite different series
in Pers. 482 ff.) and Headlam. This last refers to Suppl. 481 κλάδους re
τούτους krÀ., where it was long ago recognized that a preceding verse has
fallen out and taken the corresponding clause with it. Ahrens attempts to
maintain the MS reading by taking αἰνεῖν as equivalent to an imperative,
but this seems so ambiguous that Wieseler's easy correction is to be pre-
ferred.
99. παιών: the earliest evidence for the word in this sense, but the correspond-
ing use of παιώνιος occurs in Suppl. 1067.
100 f. τοτὲ pév ... τοτὲ δέ: this accentuation of the adverb in this formula,
which agrees with the doctrine of the ancient grammarians, is given here as
elsewhere by the MSS, cf. Kühner- Blass, i. 618 n. 5, L-S s.v.
101. ἐκ θυσιῶν dyavà φαίνεις : φαίνονσ᾽ in FTr is clearly an early attempt to
set right the unintelligible text—whether φαίνεις (M) or φαίνει (V) we cannot
Y The statement in Chandler, Greek Accentuation, 2nd ed., 83, ‘the word is always
proparoxytone in the books', is therefore incorrect.
2 For the use of the ‘gravis’ see Gardthausen, Griech. Palaeographie, ii. 391, and Schubart,
Einführung 1n die Papyruskunde, 5o.
3 A. Cho. 192 (Paley's ‘assent’ is right) is a particularly clear example of this use. Observa-
tions which are useful for the material they provide, though not always lucid, can be found
in E. Hoffmann, Qua ratione ἔπος, μῦθος, αἶνος, Aóyos . . . adhibita sint (Diss. Göttingen,
1922), 57, 76, and passim.
55
 line 101 COMMENTARY
say—and therefore must not be used as a basis for restoration; cf. on 319. If
we retain φαίνεις, the unmetrical ἀγανὰ (ayava is out of the question in these
anapaests) cannot be remedied by altering to ayavn, for that could only refer
to Clytemnestra, which is meaningless. Violent rearrangements, such as,
e.g., Headlam’s, can be ruled out as quite improbable. Therefore ATANA
is to be regarded as corrupt. This assumption is supported by the fact that
the adjective &yavós never occurs in Tragedy,’ and the adverb ἀγανῶς only
in the interpolated anapaests E. I^. A. 601 (on which see Weil's commentary,
and recently the careful discussion by D. L. Page, Actors’ Interpolations in
Greek Tragedy, Oxford 1934, 161). Ahrens's easy correction (in the edition
by Franz, who had also thought of it) to ACANA is probably right, although
Ahrens had doubts about it later (PAslol. Suppl. i. 267). The verb dva-
φαίνειν, often used by Aeschylus and the tragic poets who followed him, suits
admirably here; Wilamowitz refers to E. 75h. T. 466 θυσίας ἃς 6 παρ᾽ ἡμῖν
νόμος οὐχ ὁσίας... ἀναφαίνει. W. Ferrari, La parodos, 362 f., again defends
. the text of the MSS and reads (with Pauw) ἀγανὰ (ἀγανὴ would be better)
$avÜcio'. The deceptive argument that ἀγανή balances κακόφρων was used
long ago by Weil.
103. τὴν θυμοφθόρον λύπης φρένα: to run one's eye over the attempts made
to emend this passage during several centuries leads to the conviction that
so far no one has found the true remedy. Fortunately, the corruption is
confined to the last line; it is quite arbitrary to attack the preceding words
(down to and including ἄπληστον) with rearrangements or alterations of
other kinds. Whether the θυμοφθόρον of MV or the θυμοβόρον of FTr (with
the gloss in M: ἥτις ἐστὶ θυμοβόρος λύπη τῆς φρενός) is the genuine reading
can hardly be decided till we have attained certainty about the whole line;
both are Homeric words and both make equally good sense. The least
violent suggestion seems still to be that put forward by Ahrens and Weil
(p. xli of his Teubner ed.): λύπης, θυμοφθόρον. ἄτην (‘jedoch habe ich im
Grunde nie geglaubt mit ihr das Wahre getroffen zu haben’, Ahrens, p. 268;
he would then propose τὴν θυμοβόρον, φρενὸς ἄτην). This depends, however,
on the not entirely convincing assumption that λύπης φρένα has intruded as
a gloss on φροντίδα λύπης and partly driven out the true reading. τὴν is deleted
also by Wecklein, Stud. z. Aesch. p. 93, and Wilamowitz. According to
Wilamowitz? we are faced here with a metrical abnormality; this is very
unconvincing. À genitive governed by ἄπληστον (as in the two other passages
of Aeschylus in which the word occurs) would, of course, be suitable but cannot
be regarded as indispensable.
! The three passages in which the adjective, which belongs to the language of epic and
lyric poetry, is found in Aristophanes are significant. In Lys. 1109 we have σεμνὴν ἀγανήν
in a ‘highly poetic’ (Wilamowitz on 1110) song of praise in anapaests. Similarly we find
the word in a μακαρισμός (of ἃ purely lyrical type) in Wasps 1467, where the ἐπεμάνην
which follows is an evident reminiscence of the language of erotic poetry. In the same
manner the ecstatic lover Kinesias says in Lys. 886 ἐμοὶ μὲν αὕτη... δοκεῖ... κἀγανώτερον
λέπειν.
, 2 More detailed information on this and on similar compounds is to be found in W.
Ferrari, La parodos, 363 ff.
3 Cf. his Verskunst, 270 n. 1 and 368 n. 1 (where read ‘Fried. 947’ instead of ‘Frésch. 947’).
None of the passages quoted by Wil. can be compared with an anapaestic parodos ; nor
can the peculiar anapaests in E. Hec. which at 167f. and 209 f. (cf. Wil., Hermes, xliv,
1909, 449 n. I) seem to end in iambic dimeters.

56
 THE METRE OF THE PARODOS

The lyrical part! of the parodos (104-257)

Meire. From the point of view of metrical arrangement, this gigantic
chorus, the longest and richest? extant in Greek tragedy, falls into three
sections:
I. Dactylo-iambic: a triad, 104-59.
II. Trochaic: two pairs of stanzas, 160-91.
III. Iambic-choriambic: three pairs of stanzas, 192-257.
I
104-21 = 122-39 140-59
6 dact. 140 2 iamb.
105 5 dact. 3 lamb. (the second a choriamb)
5 dact. (the first three = 4 dact.
spond.) hipponacteum
iamb.-+4 dact. 4 dact. (the first two = spond.)
110 2 dact. (adon.) 145 —VV-UV-UUu—= US (2)
6 dact. 3 iamb.
2 dact. (adon.) 7 dact.
114-15 8 dact. 150 7 dact.
iamb. +4 dact. 9 dact.*
3 dact. 155 6 dact.
6 dact. 6 dact.
120 2 iamb. 6 dact.
5 dact. 2 dact. (adon.)
ς dact.
106 = 125. The three spondees at the beginning are worth noting. We
find exactly the same in the dactylic pentameter (in a purely dactylic con-
text) Pers. 902 τευχηστήρων, παμμείκτων τ᾽ ἐπικούρων," and two spondees at
the beginning of the dactylic pentameter Eum. 368 — 377.
108. For the caesura after ὅπως Ἀχαιῶν here and in many lines of similar
structure, see Rhein. Mus. Ixxü, 1918, 334 n. 1.
ııı ff. The most satisfactory arrangement is perhaps that which I have
adopted above, following Weil and O. Schroeder (Aeschyli cantica), viz. to
take together πέμπει... ὄρνις = 129 f. κτήνη... λαπάξει and scan it as a
hexameter. This seems to be recommended by the symmetry which it
yields: a dactylic line, καταληκτικὸν εἰς δισύλλαβον, followed by an adoneus
(Τευκρίδ᾽ em’ alav), here as immediately before (109 f.). It may, however, be
objected that as a rule the tragedians, even in their lyric dactyls, when
! It might be more consistent with the terminology of Aristotle to call this part ‘stasimon’.
T prefer, however, for the sake of convenience, to conform to the usual practice and confine
the term 'stasimon' to the following choruses.
? Cf. W. Kranz, Stasimon, 144: “The parodos of the Agamemnon, being the introduction
to a trilogy, was furnished forth with special magnificence.'
3 Cf. the commentary on the line.
* Or, perhaps, 6 dact. 3 dact. For the possibility of a hexameter without a caesura in the
third metron see p. 58.
5 Taken as a pentameter by Wilamowitz and O. Schroeder. Hermann attaches τευχη-
στήρων to the preceding line, which produces a pleasant coincidence of the syntactical
with the metrical unit, but it may be asked whether a hexameter spondeiazon (ἀκάματον...
τευχηστήρων) is likely to occur in these lyric dactyls.
57
lines 104-257 COMMENTARY
forming hexametric lines, insist on a caesura in the third metron (usually
the semiquinaria; Aeschylus also admits the caesura κατὰ τρίτον τροχαῖον,
e.g. Ag. 122, Suppl. 69). But the rule does not seem to be without exceptions.
In this very chorus the sequence of three hexameters at 155-7 cannot be
doubted, but 156 has the semiseptenaria and no caesura in the third metron.
In Eum. 365 the restoration of the text Ζεὺς αἱμοσταγὲς ἀξιόμισον ἔθνος τόδε
λέσχας is almost certain ;? if so, the verse has no caesura at all, whereas the
corresponding line 352 shows the normal semiquinaria.* In the dactylic
song E. Suppl. 271 ff. we find two hexameters without a caesura, 274 οὖς ὑπὸ
τείχεσι αδμείοισιν ἀπώλεσα κούρους, 282 μηδ᾽ ἀτάφους, τέκνον, ἐν χθονὶ Κάδμου
χάρματα θηρῶν. In order to provide these lines with a caesura, transpositions
were suggested at 274 by Hermann (and Nauck), at 282 by Wilamowitz
(Anal. Eurip. p. 91); but Wilamowitz, Die Ilias und Homer, 351 n. 1, recanted
his former view and in both lines restored the MS reading.
If we do not believe in hexameters without a caesura, we may regard
θούριος ὄρνις Τευκρίδ᾽ em’ alav either simply as four dactyls, with a slightly
unusual spondee in the second metron (cf. 144), or as two adonei, 1.6. four
dactyls formed in such a manner as to coincide with the double-adoneus,
a verse used not only by Sappho (fr. 126 D.), but also in the fifth century,
viz. as the concluding line of an Attic drinking-song, Τυδεΐδην τέ φασι τὸν
ἐσθλόν (cf. Wilamowitz, Arist. u. Ath. ii. 317), by Pindar, Nem. 2. 5 ἐν πολυ-
υμνήτωι Διὸς ἄλσει, and, with a short syllable at the end of the first adoneus
(as in Τυδεΐδην re), ibid. 20 ἀλλὰ Κορινθίων ὑπὸ φωτῶν, and by Euripides,
fr. 369. 4 N. Θρηΐκιον πέλταν πρὸς Adävas (cf. Wilamowitz, Verskunst, 388 f.).
Aeschylus provides an unmistakable instance of the double-adoneus being
used in a dactylic context: Eum. 1041 = 1045 Παλλάδος ἀστοῖς" Zeus (0)
πανόπτας.
117. The three dactyls, with the spondee in the first metron, are identical
with πλούτου πρὸς κόρον ἀνδρί, grato, Pyrrha, sub antro, a pherecratean. Cf.
also, e.g., Eum. 351 συνδαίτωρ μετάκοινος, in a dactylic context.
143. Here the hipponacteum, θηρῶν ὀβρικάλοισι τερπνά, may possibly be
regarded as one of those dactylic lines ending in — v— © which I have examined
Rhein. Mus. \xxii, 1918, 166 ff.; cf. Wilamowitz, Verskunst, 358 f.
145. If the text is sound (see the note ad loc.), this is probably a dactylic
line (equivalent to five dactyls?) ending with a cretic, for which form see
Rhein. Mus. xxii, 1918, 176 f.
In the extant plays of Aeschylus this triad is the only instance of
dactylo-iambics. But there were many more. Aristophanes in the scene of
the Frogs (1261 ff.) where he quotes κύριός εἰμε θροεῖν κτλ. also recalls examples
of songs in the same metre from six* other tragedies of Aeschylus. Further-
more we learn from Aristophanes, a highly competent expert, that those
1 In the fragment of Aeschylus’ Ξάντριαι (?), Pap. Oxy. 2164, fr. 1. τό ff., we have 15
dactylic hexameters (sung? recited ?). Of one, 1. 21, the caesura cannot be made out, owing
to the destruction of the middle part. Of the remaining 14 lines, 10 have the semiquinaria
and 4 the xara τρίτον rpoyaiov.
2 For the intrusion of γάρ Blass aptly compares Eum. 378.
3 I regard O. Müller’s d(aé)otpos as probable (cf., e.g., Ag. 1410 d(7d) ods), not certain,
but I have not the faintest doubt that the line is a hexameter.
4 At least two of the examples quoted by Aristophanes may with probability be ascribed
to πάροδοι: fr. 132 N. (cf. above, p. 27 n. 1) and fr. 273 a self-introduction of the Chorus.
58
 THE METRE OF THE PARODOS
songs, including the one from the Agamemnon, had important musical
elements in common with the κιθαρωιδικοὶ νόμοι (for details see Rhein. Mus.
Ixxii, 1918, 321 ff. ; cf. also W. Kranz, Stasimon, 139).

II
160-7 = 168-75. Simple trochees, interrupted by (165) five dactyls, which
form a link with the opening triad. Among the trochees catalectic dimeters
(lecythia) are conspicuous (τοῦτό νιν προσεννέπω, etc.). In the first line the
lecythion is preceded by a spondee so as to form a trimeter. It is fairly
common for a spondee to serve as an equivalent of a trochaic metron, not
only at the beginning of a trochaic stanza or period, as, e.g., Ag. 179, Cho.
602, E. Cycl. 356, Ar. Lys. 781, 785, 794, but also in other places, as, e.g., Ag.
182 Bılaiws," Cho. 614 Σκύλλαν (the corresponding line 603 is corrupt), Eum.
322, 323, etc.
176-83 = 184-91. Simple trochees. For the spondees see above.
Trochaic stanzas of this type are among the most noticeable elements in
the choruses of the Oresieia. They may in fact be regarded as one of the
links by which the unity of the trilogy is emphasized.

III
192-204 = 205-17. From 192 to 198 (δὲ καὶ πικροῦ) simple iambics, cretic
and bacchius being employed as equivalent to an iambic metron, as usual.
199—200. Two catalectic choriambic dimeters.
201-4. Pure choriambs, catalectic (bacch.) at the end. This scansion is
indicated in most modern editions, from Porson onwards; it is simple and
wholly satisfactory. Murray, on the other hand, prints at 201 an adoneus,
μάντις ékAay£ev, followed by five ionici a minore, προφέρων... Arpeidas,
and the clausula v v — v — —, which sometimes forms the end of a period of
ionici a minore, e.g. in the Helicon poem of Corinna (fr. 4 D.), Ar. Wasps
302, E. Bacch. 385, 536. In Murray's arrangement the ends of the metrical
units coincide with the ends of the words; that might recommend his view;
in favour of it one might also point to the conclusion (447-51) of the iambic
stanza Ag. 438 ff., where a similar metrical interpretation seems at any rate
possible. But a strong argument against it arises from the necessity of
introducing, in the midst of an iambic-choriambic (or, according to Murray,
iambic-ionic) stanza, the lonely adoneus at zor.
218-27 = 228-37. Simple iambics (from 225 onwards a few choriambs
replacing iambic metra).
238-46 = 247-57. Simple iambics (the last metron but one a choriamb).

104. κύριός eiu: the phrase has a legal ring (‘I have authority to do, am
entitled to do' L-S).
ὅδιον κράτος αἴσιον: according to the scholiast's paraphrase (in M) this
means τὸ συμβὰν αὐτοῖς σημεῖον ἐξιοῦσιν. This explanation was long ago
adopted by Stanley (‘ratum signum quod occurrit in via’) and has on the
τ It is, however, possible to take βιαίως σέλμα σεμνὸν ἡμένων as an iambic trimeter (at the
beginning of the stanzas Cho. 385 and 405 v -- — — v — υ — v — is doubtless to be regarded as
iambic). If this be correct, the last line of the trochaic section would lead up to the following
iambic stanza.
59
line 104 COMMENTARY
whole held the field from Heath (‘victoriae omen faustum in itinere oblatum’),
Schütz, and Blomfield to the present day ; Ahrens's protest (p. 272) has had
no effect. It is translated ‘des Kriegszugs sieghaft Zeichen’ (Wilamowitz),
‘the sign of the war-way’ (Murray), and so on, as though σημεῖον stood not in
the scholion but in the poet’s text.’ It is hard to see how κράτος could
acquire that meaning. Headlam, like the rest, translates ‘the auspicious sign
of victory’, but he does not distort the meaning of κράτος, for, like Wecklein in
his annotated edition, he adopts Francken’s conjecture τέρας for κράτος ;
these critics were not deterred by the fact that the line is quoted with
κράτος in the Frogs. Other commentators take the expression differently :
‘der unter glücklichen Götterzeichen unternommenen Heerfahrt Obmacht’
(Schneidewin), ‘die sieghafte Macht, welche der Heerfahrt durch Schicksal
und Götterwillen verliehen war’ (Pliiss), ‘the fatal victory’ (Sidgwick). This is
at least possible Greek. It is not, however, very probable that κράτος here
is equal to victorta. Where in Aeschylus and Sophocles the word has this
meaning, it is used in phrases such as κράτος ὀπάζειν, κράτος νέμειν τινί, κράτος
κατακτᾶσθαι, ἄρνυσθαι (and, perhaps, Cho. 490 80s . . . εὔμορφον κράτος, although
there some editors ascribe to κράτος an entirely different sense). The only
interpretation on correct lines is Hermann’s: ‘Aeschylum proprio significatu
ὅδιον κράτος dixisse puto vim viairicem, quo indicaret profectionem insti-
tuentes’, or, as he puts it in his version (Opusc. v. 344): ‘Fas mihi dis
carum robur celebrare virorum ductorum.' He has grasped the essence of the
thought ; it is of minor importance that κράτος in this passage means not so
much vis or robur as imperium (so rightly Wellauer and Dindorf, Lex. Aesch.).
What in my opinion settles the question is the pointed repetition of word
and thought in 109 Ἀχαιῶν δίθρονον κράτος. After the interruption of the
main thought by the parenthesis ér. yap . . . αἰών, the opening phrase is
resumed ; therefore κράτος Ἀχαιῶν and κράτος ἀνδρῶν ἐντελέων cannot mean
different things. 'It is my office to tell of the command that set on foot the
expedition, the command favoured of fortune.' The nouns (κράτος ἀνδρῶν) at
the beginning denote concisely and effectively the leaders in the action which
the Chorus is going to relate; ὅδιον is added to indicate the sphere in which
the κράτος is exercised, and αἴσιον to emphasize its most important feature.
Then intervenes a parenthesis (ἔτι γὰρ κτλ.) ; and after this interruption the
object of θροεῖν is repeated in a different form (ὅπως krA.), but in such a way
that the sentence is again headed by an expression denoting the two leaders,
Axadv δίθρονον κράτος, which now becomes the object of πέμπει ὄρνις. The
resumption of the opening phrase does not stop with κράτος xev; it
would be truer to say that ὅδιον finds its counterpart in πέμπει Τευκρίδ᾽ ἐπ᾽
αἶαν, and αἴσιον in θούριος ὄρνις with its series of weighty appositions.
A Greek would easily underständ ὅδιον κράτος, since he can say τῆς ὁδοῦ
(or τοῦ στόλου) κρατεῖν (ὁδός is used of the expedition against Troy in, e.g.,
£ 235). This initial sentence moves on very general lines; so the epithet αἴσιον

τ C. Thomson goes so far as to say : ‘However these words (ὅδιον κράτος) are to be trans-
lated, they must refer to the sign which was seen at the departure of the kings from home.’
He himself attempts a compromise solution by combining the meanings of κράτος and τέρας,
but it will not do. Murray's and Thomson's reference to 2 293 = 311 is misleading; there,
as elsewhere, the formula καί εὖ κράτος ἐστὶ μέγιστον indicates surpassing strength, in this
case the eagle's. |
60
 COMMENTARY line 105
takes into account only the μεγάλα ἀγαθά (156), while the adverse circum-
stances are passed over. This is in keeping with the course of the war viewed
as a whole, and at the same time makes it possible to open with bona verba.
105. ἐντελέων. The MS reading is ἐκτελέων. This word, which is found else-
where in Aeschylus (Pers. 218), means ‘full-grown, ripe, perfect’ (cf. Wilamo-
witz on E. Jon 780), and is quite unsuitable here as a description of warrior
manhood. The warriors, from the standpoint of the old men who are telling
the tale, might be described as ‘still in the strength of manhood’ or something
similar, but not as adulti, which implies a comparison with younger people,
boys for instance. The same objection would have to be raised to the con-
jecture evreAdwy, if the word were to be taken in its normal sense ‘full-grown’
as in Cho. 250 (cf. Blass, ad loc.) ; and similarly the closely related meaning
of ἐντελής (cf. τέλειος) ‘perfect, faultless, complete’ and the like, which is
found, for instance, in Sophocles and Thucydides, would be quite inappro-
priate here in the description of the army.
Attempts have been made to keep éxreAéov and ascribe to it a peculiar
meaning ; there is, however, nothing in the evidence to support the view that
the word could be used ‘in strained sense for "royal", “high” ', as Sidgwick
suggests. Headlam's note on his prose translation (‘ ἐκτελέων, which, how-
ever, would be more naturally contrasted with immaturity than with the
aged Elders’ own decay. évreAéov would be “men in power” ’) shows that he
was on the right track, and only reluctant to take the final step. The obvious
conjecture évreAéov (Auratus and others) early became the received text, and
later was vigorously supported by Hermann. Blomfield, Hermann, and
others refer to Timaeus, Lex. Plat. (= Suidas s.v.) ἐντελεῖς: οὗ ἄρχοντες, καὶ oi
ἄρξαντες, to which must be added Hesychius évreAéorarov ἐντιμότατοι. But
things are not quite so simple as might seem from this. We possess no evi-
dence for such a sense of ἐντελεῖς: it is given by the lexicographers on the basis
of material of which we know nothing. In the inscription from Miletus of the
last third of the fourth century B.c. quoted by Wilamowitz, ad loc.
(Dittenberger, Syll. 286 1. 10 = Schwyzer, Exempla, 735), ἐντελής (in contrast
to ἀτελής and possibly coined to express the opposite idea) denotes the man
who is qualified to undertake public munera. That is quite a different
conception, as Wilamowitz was well aware ; but since on his view. (and that of
other editors) the meaning in Ag. 104 f. is ‘die streitbare Mannschaft’, ‘qui
stipendia faciunt! , he welcomed the apparent parallel in the inscription. One
may in any case wonder whether ἐντελής, divorced from the connexion with
ἀτέλεια, could be used absolutely, and whether it could assume the specialized
sense qui stipendia faciunt. But leaving that on one side, the repetition of
κράτος in 109 makes it very improbable that ἀνδρῶν in 104 should refer to
anyone but the Atridae. Kranz, Hermes, liv, 1919, 3o3 n. 2, has rightly con-
cluded that we must explain ἐντελέων (with Blomfield and Hermann) as τῶν
ἐν τέλει övrwv.! For this there is support in the contemporary use of τέλος.
For the meaning of τέλος in Aeschylus see Blass on Ewm. 743 (Wila-
mowitz’s note on Sept. 1025 is incorrect). Pindar in a late poem, Nem. 11. 9,
uses τέλος δωδεκάμηνον of the prytany; Aeschylus uses τέλος in the sense of
! This interpretation is already anticipated in an interlinear gloss in Tr, overlooked by
modern scholars : τῶν ἐν τέλει ὄντων, ἢ τῶν βασιλέων. In flat contradiction to this, however,
Triclinius reads éxreAéov with the other MSS.
61
line 105 COMMENTARY
officium (in its early sense), munus, Ag. 908, Cho. 760, Eum. 743, though always
with reference to a definite duty, not an office; see also on 1202. It is far
more likely than not that he was also familiar with the expression οἱ ἐν τέλει,
though he may have had no cause or no inclination to employ it in the
extant plays (Euripides similarly does not use it). The term has an archaic
look, and it is found, moreover, in Herodotus and as far back as the earliest
extant plays of Sophocles, Ant. 67, Ajax 1352; also in the Ἀχαιῶν σύλλογος,
fr. 142. 14 P., Phil. 385, 925). οἱ ev τέλει means ‘men in authority, in office”,
and thus indicates rulers and commanders-in-chief ; the use of ἐντελέων here
corresponds to this. Although &vreAns had a fixed meaning for Aeschylus
(Cho. 250) as for the normal language in general (see above), he was at liberty
to provide it with a different meaning! by deriving it afresh, as it were, from
ἐν τέλει by means of 'hypostasis' (H. Usener, Kleine Schr. i. 250 ff.; cf., e.g.,
πρόχειρος from πρὸ χειρῶν, φροῦδος from πρὸ ὁδοῦ, perfidus from per fidem, etc.).”
καταπνεύει may perhaps have stood in M before the erasure of the letter
following the first «, and is the reading of the other MSS; the Aldine ed.
introduced καταπνείει, which has become the received text. Ahrens (p. 270)
supported the MS reading καταπνεύει as the Aeolic form, and was followed by
Wilamowitz. But this seems open to question. The evidence for this Aeolism
is weak? (cf. Lobel on Alcaeus fr. 145 and his Introd. to Sappho xxxiii),
and even were the support stronger than it is, there would be little likelihood
in Tragedy of a Lesbian form not also found in Homer. We should therefore
probably join the earlier critics and W. Schulze, Quaest. ep. 279, in preferring
the Homeric form karanveieı.* At the same time, it may be recalled by way
of warning that Timotheus in the Cyclops (fr. 7 Wil. = 2 D.) has éyxeve and
évéyeve, though in the Persae he ‘shows no Aeolic forms’ (Wilamowitz, p. 39).
105f. In order to understand ἔτι yap... αἰών, the first necessity is to take
σύμφυτος αἰών correctly, in other words to abandon the conjecture ἀλκᾶι
made independently by Schiitz and Hermann, and still put in the text by
Platt (J. Phil. xxxii, 1913, 44), Wilamowitz, and A. Y. Campbell. These
editors take ἀλκᾶι σύμφυτος αἰών together; it is, however, quite certain that
σύμφυτος αἰών must be taken by itself, and that the scholiast rightly under-
stood the sense to be ὁ yàp σύμφυτός μοι αἰών, 6 ἐστι τὸ γῆρας, . . . καταπιεῖ.
Here, as so often, αἰών is the lifetime; cf. Wilamowitz on E. Her. 669 and his
Aischylos Interpr. 170 n. 3: 'xpóvos is absolute, αἰών relative, defined by the
1 C£, e.g., the manner in which in Ag. 471 he uses ἄφθονος in a peculiar sense, laying strong
emphasis on φθόνος, whereas elsewhere he uses it with its conventional meaning.
2 An early example of ‘hypostasis’ with ἐν is Hesiod, Erga 344 εἰ γάρ rot καὶ χρῆμ᾽ ἐγκώμιον
(ancient varia lectio) ἄλλο γένηται, on which Wilamowitz remarks that the combination of
ἐν τῆι κώμηι into an adjective is good Greek, although it does not occur elsewhere.
3 Wilamowitz himself remarks, Kl. Schr. i. 405, that in Alcaeus fr. 81. 4 D. πλέην, ‘which
can only be taken as an infinitive’, has not become πλεύην,
4 It is one thing to assume that Aeschylus used καταπνείες in dactyls of the epic type and
another to introduce πνεέειν into tragic passages where there is no such justification. In
A. Cho. 621 πνέονθ᾽ à κυνόφρων ὕπνωι the true nature of the glyconic was understood as long
ago as Hermann (on S. Ant. 1146 [1132 Herm.]), and yet G. Thomson puts Heath’s πνείονθ᾽
in the text. In E. Iph. A. 578 Murray reads πνείων with Dindorf, but μιμήματα πνέων is
perhaps to be taken as a dochmius (see Wilamowitz, Verskunst, 572). A more difficult case
is 5. Ant. 1146; here we might straight away accept Brunck’s πνειόντων, were it not for the
general obscurity of the metre in this ode (see Wilamowitz, Verskunst, 123) and the presence
of several corruptions in the text.
62
 COMMENTARY lines 105 f.

person or thing whose air it is’; cf. also H. Frankel, ‘Die Zeitauffassung in
der archaischen griech. Lit.’, Zischr. f. Ästhetik, xxv, 1931, Beilageheft, p. 114.
It is clear that 6 σύμφυτός μοι αἰών, τὸ γῆρας is the right explanation, for
it gives us an idea which occurs several times in Aeschylus and Sophocles
—the idea that a man’s lifetime is born, grows up, and ages with him. The
commentators on our passage (e.g. Conington, Hartung, Keck, Nagelsbach,
Schneidewin-Hense) have illustrated the idea; good observations will also
be found in Radermacher on Soph. Aj. 623 and Oed. C. 7, cf. A. C. Pearson
on Soph. fr. 950. 2.! The nearest parallel to the image of the σύμφυτος αἰών
is perhaps that of the συγγενεῖς μῆνες in Oed. R. 1082; except that there the
habitual and basic idea is given a new point in order to bring out a special
conception : just as Tyche is the mother of Oedipus, so the months, too, take
on a family relationship to him, they become his “kinsmen’ (Jebb) or perhaps
even his ‘brothers’ (Wilamowitz). The same connexion is seen from the other
point of view when we are told of Ajax’s mother (S. Aj. 623) that she is
παλαιᾶι σύντροφος (Nauck, évrpodos MSS) ἁμέραι. In the Prometheus Zeus is
still young and knows nothing yet of any trouble, but as Prometheus says
(981) ἀλλ᾽ ἐκδιδάσκει πάνθ᾽ ó γηράσκων χρόνος ; be it man or god, as he grows
old, χρόνος itself grows old ; cf. Eum. 286 (a line obviously added in the margin
from another tragedy). This conception that a man and the epochs of his
life ‘run abreast’ is expressed with particular force by Aeschylus in fr. 362 N.:
ἀλλ᾽ οὔτε πολλὰ τραύματ᾽ ἐν στέρνοις λαβὼν θνήισκει τις, εἰ μὴ τέρμα συντρέχοι
βίου. For the related image of the ξυνεύδων χρόνος in Ag. 894 see ad loc. The
way of looking at things that we find in σύμφυτος αἰών, συγγενεῖς μῆνες, etc. 15
given a different turn in the πότμος συγγενής that belongs to every man
(Pindar, N. 5. 40) ; this is then directly continued in Philemon (fr. 10 Kock):
νῦν δ᾽ οἶδ᾽ ἀκριβῶς τὴν τύχην ὡς οὐ μία, οὐδ᾽ ἔστι πρώιην, ἀλλὰ μετὰ τῶν σωμάτων
ἡμῶν, ὅταν γιγνώμεθ᾽, εὐθὺς χὴ τύχη προογίνεθ᾽ ἡμῖν, συγγενὴς τῶι σώματι, and
more profoundly in the well-known lines οἱ Menander (fr. 550 Kock), now
more intelligible in their context (cf. Demiañczuk, Supplem. comicum, p. 60)
ἅπαντι δαίμων ἀνδρὶ συμπαρίσταται εὐθὺς γενομένωι μυσταγωγὸς τοῦ βίου,
ἀγαθός κτλ. In all these examples the notion οἱ simultaneous birth, growth,
progress, and so on,? always applies to the relationship between χρόνος, αἰών,
πότμος, δαίμων, etc. and the individual man as such. In the passage before
us, the interpretation σύμφυτος ἐμοί is in agreement with these ideas, while
the notion attached to the conjecture ἀλκᾶι σύμφυτος is unparalleled.
The fundamental idea of 105 f. has been rightly compared by Hartung with
E. Her. 678 ἔτι τοι γέρων ἀοιδὸς κελαδεῖ μναμοσύναν. It is indeed extremely
probable that Euripides here had the beginning of the Agamemnon chorus in

1 I cannot, however, agree with Headlam’s and Pearson’s view that in the fragment of
Sophocles θεία ἡμέρα is equivalent to ‘inspired vitality’, nor can I accept the interpretation
of σύμφυτος αἰών = ‘inbred vitality’ which Pearson advanced in his inaugural lecture
(Cambridge 1922) on ‘Verbal Scholarship’, p. 29. In our attempts to understand the Sopho-
clean words ἐν ols 6 νοῦς θείαι ξύνεστιν ἡμέραι τεθραμμένος we may perhaps do well to take
advantage of Wilamowitz’s comment (Berl. Sitzgsber. 1909, 826 ff. = Pindaros, 200 ff.) on
Márep ‘Aou πολυώνυμε Θεία. It will then appear that the notion of ‘divine daylight’
comprises physical brightness as well as the clearness and lucidity of a mature mind.
2 Cf. also Hippocrates s. νούσων 2. 73 (vii. 112 Littré) ταῦτα ποιέειν, καὶ ἅμα τῆι ἡλικίηι
ἀποφεύγει [the patient], καὶ ἡ νοῦσος καταγηράσκει σὺν τῶι σώματι" ἣν δὲ μὴ μελεδανθῆι, συναπο-
θνήισκει.
63
lines 105 f. COMMENTARY
his mind (so Hartung, and Wilamowitz, E. Her., 2nd ed., ii. 148) ; the thought
in both passages is substantially the same.
106. Obviously MOATIAN must be taken as gen. plur. It is less easy to
determine whether this genitive depends on πειθώ or on ἀλκάν. In examining
this point, we should not forget that πειθώ here no more than anywhere else
(cf. on 87) can mean ‘fiducia’, as it is rendered by Schütz, F. Bamberger
(Opusc. 39), Dindorf (Lex. Aesch. s.v.), and Wecklein. It means the ability
to tell a story in such a way that the hearer believes what he is told. When
Eumaeus rejects the narrative of Odysseus as a ψεῦδος, he says (ξ 363 f.)
οὐδέ με πείσεις εἰπὼν ἀμφ᾽ Ὀδυσῆι. ‘Suadela, as Blomfield rightly translates,
is a quality every bit as necessary in a song where a tale is told, like that
which follows here, as it is for the orator’ (Ahrens, p. 271). Consequently, one
can take together either πειθὼ μολπᾶν (so Blomfield, whose text, however, in
what follows is quite arbitrary) or μολπᾶν ἀλκάν (so Enger, Ahrens, Hense,
Ferrari, La parodos, 367 τι. 2) ; in the latter case μολπᾶν would have to be taken
as an epexegetic genitive, ‘the ἀλκή that expresses itself in song’, and the
phrase would be in apposition to πειθώ. I regard the latter interpretation
as not impossible, but prefer the former for two reasons: (1) what the
Chorus begin to do here is not πείθειν in general but πείθειν by means of
song, (2) ἀλκάν without qualification seems to do much more justice to the
special meaning of the word. In early Greek in general (see e.g. the good
article in Passow-Crönert), and so also in Aeschylus, ἀλκή has, besides its
principal meaning ‘defence, protection, defensive action’, the closely related
sense of ‘warlike strength’, but not of ‘strength’ or ‘power’ in all senses."
Now the beginning of this song is based on the contrast to that part of the
anapaests (72 ff.) in which the physical weakness of the Elders and their un-
fitness for war are emphasized. In continuation of that thought they now sing:
‘for still from God above my age breathes down upon me the power of singing
persuasively, so that this power becomes my militant strength’. ἀλκάν, taken
as merely in apposition to πειθὼ μολπᾶν, would hardly be intelligible; but as
soon as we regard it as predicative (to go with καταπνείει), it forms the
culmination of the sentence. ἀλκή in its original sense has long deserted
the Elders, but they still retain the power which can take its place, the
power which enables them once again in song to renew the great experience
of ten years ago. He who tells a tale that can be trusted, and hands on a
true story for all time to come, falls not far short in merit, if short at all, of
him who did the deed; this was a widespread view in the time of Aeschylus
(cf., e.g., Pindar, Ol. 11. 10, with ἐκ θεοῦ there as θεόθεν here)"
That the phrase θεόθεν καταπνείει as a whole is taken over from late epic
poetry (θεόθεν in Homer is found only in 7447) or the language of oracles is
made probable by its appearance in the parodic passage of the comic poet
Plato, fr. 173. 14 Kock, μή σοι νέμεσις θεόθεν καταπνεύσηι. It may seem
attractive to find in the θεόθεν καταπνείει of this chorus an early expression
of that view of the peculiar inspiration of the poet which meets us first in
! It is impossible to say whether the gloss in Hesychius, ἀλκή" δύναμις, ἰσχύς. ἡ ἀλέξησις.
3| μάχη. Αἰσχύλος Ἀγαμέμνονι (or at any rate the interpretation given in the first place) refers
to this passage.
2 Pind. Ol. 1. 112 should not be adduced as a parallel (although aAxäı occurs there), for
the thought is influenced by the common image of the βέλος of the Muse.

64
 COMMENTARY line 110

Democritus (B 18 ποιητὴς δὲ doca μὲν ἂν γράφηι per’ ἐνθουσιασμοῦ καὶ ἱεροῦ

πνεύματος, καλὰ κάρτα ἐστίν) and then in Plato (Meno 99 d θείους τε εἶναι καὶ
ἐνθουσιάζειν, ἐπίπνους ὄντας καὶ κατεχομένους ἐκ τοῦ θεοῦ; Phaedrus 262 d,
metaphorically of the cicada, tows δὲ καὶ οὗ τῶν Μουσῶν προφῆται οἱ ὑπὲρ
κεφαλῆς ὠιδοὶ ἐπιπεπνευκότες ἂν ἡμῖν εἶεν τοῦτο τὸ γέρας ; not to mention here the
other passages that speak of tlıe θεία μανία and so forth of the poet without
expressly indicating the element of ‘inspiration’). The influence of this idea
can be seen in certain well-known passages of Horace and elsewhere (cf.
R. Reitzenstein, Hellenist. Mysterienreligionen, zrd ed., 321 f.; W. Kroll,
Stud. z. Verständnis d. röm. Lit. 24 f.; W. Kranz, Ilbergs Neue Jahrb. lii, 1924,
72; H. Lewy, ‘Sobria ebrietas’, Beth. z. Zischr. f. neutest. Wissensch. ix, 1929,
47 1). But it would be wrong to class Ag. 105 with these passages simply
because of the phrase θεόθεν karanveieı. There is no suggestion here of a
ἱερὸν πνεῦμα, ἃ κατοχή, in the sense of the passages quoted from Demo-
critus and Plato. Nor should we read that idea into the passage in which
Hesiod says of the Muses (Theog. 31) ἐνέπνευσαν δέ μοι αὐδήν ‘they breathed
voice into me’; what Hesiod means is that previously he was no more
capable of utterance than the other shepherds. What happens is no less
wonderful, but at the same time the language in which it is described is no
less ordinary and non-technical, than when in the Jhad a god ἔμπνευσε μένος
into a hero, or even into his horses. The ‘breathing into’ a man of the power
of song, without any thought of ἑερὸν πνεῦμα and the like, is found in the age
of Aeschylus in e.g. Pind. Οἱ. 6. 82 f. δόξαν ἔχω rw’ ἐπὶ yAdoou . . . à μ᾽
ἐθέλοντα προσέρπει καλλιρόοισι πνοαῖς. Finally, in the interpretation of our
passage we must also bear in mind that, just as behind Hesiod’s ἐνέπνευσαν
αὐδήν stands the earlier ἔμπνευσε μένος, So too καταπνεῖν, which here is asso-
ciated with πειθὼ μολπᾶν, could in serious poetry govern other objects, and
thereby convey the idea of a gift of the gods ‘breathed down from on high’
without any specific πνεῦμα. This we can see from the anapaests in Ar. Lys.
551 ff. ἀλλ᾽ ἥνπερ 6 τε γλυκύθυμος "ἔρως x1) Κυπρογένει᾽ ᾿ἀφροδίτη ἵμερον ἡμῖν
κατὰ τῶν κόλπων καὶ τῶν μηρῶν καταπνεύσηι, where Wilamowitz in his com-
mentary has pointed out the poetical colouring and dignity of language. The
metaphor there is simple: Aphrodite breathes down ἵμερος as she does else-
where (E. Med. 839) gentle breezes. In the phrase of Aeschylus perhaps the
boldest feature is that old age is made the subject of καταπνεῖν, while yet
this ‘breathing down’ proceeds θεόθεν. Such bold compression is found else-
where in this ode. Here the concise phrase makes perfectly clear what the
Chorus want to say: ‘my old age gives me still the power to win my hearer’s
assent to what I sing, and this power of song is the gift of god.’ Cf. also on 131.
108. ὅπως ᾿Αχαιῶν κτλ. For the resumption of the sentence after the
parenthesis see on 104.
110. From the 'recensio' of our MSS! it must be deduced that the παράδοσις
was raydv. This has been defended against manifold attempts at alteration
by Enger (in his re-edition of Klausen): *ut duos Achivorum reges dixerat
! r10—12 were omitted by the first hand of M and added later, as is remarked by Dindorf
(ed. 5, praef. Ixxxi), Kirchhoff, and Weil (Teubner edition), cf. Vitelli-Wecklein, app. crit.,
and Rostagno's preface to the facsimile of the Mediceus, p. 13. Editors since Wilamowitz
have not mentioned this. Wilamowitz himself does not refer to it in his apparatus, but
see his preface, p. xii.
4872.2 F 65
line 110 COMMENTARY
δίθρονον κράτος, ita duos Graeciae iuventutis in bellum proficiscentis duces
aptissime dicit £üuppova raydv.' There is, however, a difficulty. As far as we
can judge from our scanty evidence, the first syllable of raya ought to be short
(therefore Dindorf conjectured ξύμφρονε rayo). In Ar. Lys. 105 ἐκ τᾶς rayás
fills an iambic metron (Wilamowitz, ad loc., regards raya as the Laconian
form), and ταγαῖσι also appears with the first syllable short in the trimeters of
Pseudo-Callisthenes,! Hist. Alex. magni 33. 11, p. 35. 19 Kroll; similarly in
Eum. 296 τἄγοῦχος indicates τἄγή. On the other hand, we must bear in mind
that the reason why the stem-syllable of rayos (similarly Sept. 58 τἄγευσαι) is
long is by no means clear (Chantraine, La Formation des noms, 12, says 'the
long a is surprising). ταγός and ταγά seem to have been alien to Ionic and
Attic; apart, perhaps, from Cyprus, these words were part of the spoken
language only in Thessaly,? where as early as the fifth century we have
evidence not only for rayds but also for the abstract nouns ταγά and ἀταγία:
Dittenberger, Syll. 55 = Schwyzer, Exempla 557, 7 kév raya «ev ἀταγίαι. On
the occurrence of rayós and its derivatives in the Attic tragedians, Eduard
Meyer, Theopomps Hellenska, 237 n. 2, conjectures that ‘we ought perhaps to
recognize in it an effect of the close connexion of Thessaly with Athens since
the days of the Pisistratids'.?
111. σὺν δορὶ δίκας πράκτορι is the reading of the MSS. Without the assis-
tance of Aristophanes, the true reading might perhaps have been discovered
by conjecture, possibly with the help of the Attic epigram adduced by Hense,
IG ü. 3. 2717 (= 24 Kaibel) κτώμενον εὔκλεαν δορὶ kal χερὶ τόνδε KrA., but would
hardly have won general acceptance. δίκας is an explanatory gloss on
πράκτορι, Which has slipped into the text, adopted the lyric termination,
and driven out xai χερὶ (cf. on 69). The man who added the gloss understood
πράκτορι correctly. πράκτωρ in Aeschylus has always (three times in all) the
same meaning—though naturally not the same technical limitation—as in
the language of administration in Athens and elsewhere (cf. on 72), — Latin
'exactor', the man who collects payment of a penalty or exacts an atonement ;
that is to say, the noun is based on that particular meaning of πράσσειν or
πράσσεσθαι (cf. on 705) which is common in Aeschylus in general and in the
Oresteia in particular. πράκτωρ in the same sense occurs in 5. El. 953, but
Sophocles also uses it with the more general meaning of ‘agent’ or ‘cause’.*
For the history of πράκτωρ in general see Ernst Fraenkel, Nomina agentis, i.
220. For πράκτορι with the feminine noun χερί cf. on 664.
For the use of the preposition in ξὺν 8opi καὶ χερὶ mp. cf., e.g., Sept. 31 σοῦσθε
σὺν παντευχίαι, Pers. 202 f. σὺν θυνηπόλωι χερὶ βωμὸν προσέστην, and see Tycho
1 The verses incorporated here and elsewhere derive from the author of the book; on
this see W. Kroll, RE x. 1712. 40 ff. The very different view in regard to the inserted
choliambs put forward by A. D. Knox (in the Loeb Library edition of Herodes, Cercidas,
etc., pp. 287 ff., and Gnomon, vi, 1930, 321) need not be discussed here.
2 To the instances which have long been known a very early one has been added in the
archaic inscription from Atrax reprinted by Kretschmer, Glotta, xxvi, 1938, 36 f., from
"Apx. "Ed. 1934-5, 140 ff.
3 Cf. also Daube, 31 (where, however, in the interpretation of κέν raya x&v ἀταγίαι account
must also be taken of Hiller v. Gaertringen in Dittenberger, Syll. 55 n. 2, and RE vi A.
119. 64 f.).
* There is, of course, no reason why Aeschylus, too, should not have used πράκτωρ in
its wider sense. In the AucrvovA«ot, Pap. Oxy. 2161, col. 1. 5, προπράκτορα is parallel with
πρόξενον. ‘It seems to mean “champion” but not “‘avenger” ' (Lobel).
66
 COMMENTARY line 115

Mommsen, Beiträge zu der Lehre von den griech. Präpos. 609. To the expres-
sion as a whole parallels (adduced by van Heusde) may be found in Tragedy,
E. Iph. A. 587 ff. ὅθεν (θεᾶν, suppl. Wilamowitz> ἔρις ἔριν Ελλάδα σὺν δορὶ ναυσί
τ᾽ ἄγει Τροίας πέργαμα, and in Satyric drama (after the model of Aeschylus ?),
Achaeus fr. 29 (p. 753 N.) Ἄρης ὁ ληιστὴς σὺν δόρει σὺν ἀσπίδι, the second half
of which line is twice quoted by Aristophanes."
112. θούριος ὄρνις : in Aeschylus, θοῦρος and θούριος always refer to the onrush
of battle, just as they do in Homer (predominantly θοῦρον “Apna and θούριδος
ἀλκῆς, with isolated examples of ἀσπίδα θοῦριν, αἰγίδα θοῦριν). At the outset
πέμπει... θούριος ὄρνις sounds as though the omen that set the expedition
under way was given by a single bird only, and that impression continues with
the beginning of the apposition οἰωνῶν βασιλεύς; but the further addition
ὁ κελαινὸς 6 τ᾽ ἐξόπιν dpyäs makes it quite clear that ὄρνις here indicates a
pair of birds, and that is made still more obvious when we reach the verb
φανέντες.
115. Hartung’s conjecture ὁ δ᾽ ἐξόπιν, still placed in the text by Weil in all
his editions and by Wecklein in 1888,? should really mislead no one for more
than amoment. Schneidewin rightly says that ‘the singular οἰωνῶν βασιλεύς...
is followed by à . . . à τε, loosely connected with it, and therefore we then have
φανέντες; οἷ. 41f., E. Iph. T. 3f. Ἀτρέως δὲ mais Μενέλαος ᾿ἀγαμέμνων τε
[where the text should not be tampered with)’.
It seems likely that the correct form is not ἀργᾶις but apyäs, the result of
contraction from *apyäevrs, as H. L. Ahrens, De Graec. ling. dial. ii. 196,
Kühner-Blass, i. 202, L-S s.v. assume. Some copies in ancient times perhaps
had apyaıs, since there was supposed to be direct contraction from apydeıs ;
this would account for the corruption to ἀργίας.
What are the two eagles between which a distinction is here drawn?
Clearly the second is the πύγαργος, which stands first among the γένη ἀετῶν
at the beginning of ch. 32 (p. 618518) of the ninth book of the Historia
animalium, the work of one of Aristotle’s pupils (cf. W. Jaeger, Aristoteles,
352 = 329 of the Engl. transl.). Further references are collected by D’Arcy
Thompson, A Glossary of Greek Birds, 2nd ed., 255, and by Pearson on Soph.
fr. 1085 (= 981 N.). The context in Sophocles is unknown, but we are told
that πύγαργος was used by him ἐπὶ τοῦ δειλοῦ. The indirect form of expression
in Aeschylus is worthy of remark: ‘aquilam πύγαργον decentius nuncupavit
ἐξόπιν dpyáv ' (Lobeck, Pathol. proll. 497; his defence of ἀργίαν may now be
neglected). In other words, Aeschylus (at any rate in the more exalted style
of our passage ; in a satyric drama he might well have been less strict) avoided
the vulgar word πυγή just as Homer did (cf. Wackernagel, Unters. z. Homer,
225 f.). On the force of the epithet as a characterization of Menelaus see
below. About κελαινός it is more difficult to reach certainty. The simplest
course is to understand it of the overall colour of the first eagle, as almost
f Tt is by no means certain that σὺν δόρει σὺν ἀσπίδι is an old popular formula (for the
form cf.,e.g., Petronius61.9 per scutum per ocream), taken over, and not invented, by Tragedy,
as Daube, 107, n. 40, assumes. It is artificial to understand ovv xepé in Aeschylus, as he does,
of the army instead of the hand and arm of the general. In any case there is no evidence
in Aeschylus for χείρ = ‘band’; the incorporation of Suppl. 958 σμικρᾶς χερί in L-S χείρν
rests on a misinterpretation.
2 Much worse is Karsten's violent change ὁ μὲν αἰθός, ὁ δ᾽ but this, tco, has recently been
put in the text.
line 115 COMMENTARY
all commentators do, and this receives adequate support from what we are
told of the μελανάετος (cf. D’Arcy Thompson, op. cit. 197) in the passage of
the Historia animalium cited above. To this we could add the evidence
of ® 252 (already adduced by Stanley), if we could there regard αἰετοῦ...
μέλανος τοῦ θηρητῆρος as the original text (but see below). κελαινός is, however,
taken differently by Schneidewin; cf. his note: ‘The largest were πύγαργοι
or νεβροφόνοι, the most powerful μελάμπυγοι or λαγωφόνοι (Arist. H.A. 9. 32,
Hom. IE. 21. 252) . . . . Hence Archilochus fr. 110 [= 93 D.] μή rev μελαμπύγου
τύχηις.᾽ This view is shared by Wilamowitz on E. Her. 237 [238 f.], where
he is discussing the σχῆμα ἀπὸ κοινοῦ: ‘A. Ag. 115 keAawós ὅ τ᾽ é£ómw apyós!
(i.e. μελάμπυγος and λευκόπυγος, cf. Porphyr. on 2 315)’; the same interpreta-
tion underlies his rendering: ‘Das Schwanzgefieder des einen weiss, des
anderen schwarz'. This interpretation, though not certain, must be taken
seriously. If it were accepted, the clause 6 «eÀawos . . . dpyäs would be
more tightly knit. We could‘speak with confidence if we could be sure
that in 252 Ahrens has rightly conjectured peAavópaov, which is approved
by Leaf and considered by Wackernagel, op. cit. 226 n. 3 (for ancient dis-
cussions of the passage cf. Philippson, Hermes, lxiv, 1929, 167 f., where,
however, Howald, Rh. Mus. \xxii, 1918, 415 f., ought to have been taken
into account) For the eagle ὅς θ᾽ ἅμα κάρτιστός τε kai ὥκιστος πετεηνῶν
(which there provides a comparison for Achilles) makes a very appropriate
counterpart for Agamemnon. But even if we leave the Homeric passage
out of account, there is a piece of evidence which shows that long before
Aeschylus μελάμπυγος was a common name for a special kind of eagle. It is
true that in the words in question, Archilochus fr. 93 D. (110 Bergk), μή rev
μελαμπύγου τύχηις, this precise point—the meaning of weAauruyov—has been
much discussed. Otfried Müller, Die Dorier, i (1824), 458 (expanded in the
English translation by Tufnell and Lewis, i. 447 f.), refers the expression solely
to Herakles, or to some person whose rump shows the same peculiarity, and
leaves the eagle out of account, while Lobeck, Aglaoph. ii. 1299, says quite
definitely ‘Archilochus non Herculem Melampygum significavit sed aquilam’,
and supposes that the reference of the ‘proverb’ μὴ μελαμπύγου τύχηις to
persons was a later development. .Lobeck’s view is followed, e.g., by O.
Gruppe, Griech. Mythol. i. 487 τι. 2, and E. Diehl on Archil. fr. 93 (to include
the fragment in the context of the fable about the Fox and the Eagle seems
to me quite arbitrary), while on O. Müller’s side are, e.g., Adler, RE xi. 310
(‘Kerkopen’), and Wilamowitz on Ar. Lys. 802 ({μελάμπυγος was applied to
Herakles in the story of the Kerkopes which was recounted by Archilochus
110 B.') ; Seeliger in Roscher's Mythol. Lex. ii. 1168 f. sits on the fence. But
besides the evidence quoted below, there are well-known analogies which
make it likely that the warning μή rev μελαμπύγου τύχηις points not to either
eagle or Herakles, but to Herakles and eagle too. Only when we understand
this can we do full justice to the exquisite homely humour of the expression,
humour which Archilochus must have revelled in just as much as the painters
and sculptors of the archaic period. That Archilochus described the eagle as
μελάμπυγος is certain ; it is expressly attested by Schol. B on 2 315 (whether
Porphyrius or no is uncertain) and by Zenobius Athous 2. 85 (Miller, Mél.
p. 367): both witnesses quote the line of Archilochus as evidence. On
! ἀργός is perhaps a slip of memory.
68
 COMMENTARY line 115

the other hand, not only is μελάμπυγος ἃ suitable epithet for Herakles, but
the tradition represented by the paroemiographers (see Zenob. vulg. 5. xo,
Corpus Paroem. vol. i. 119, with the note; the same source lies behind
Tzetzes on Lycophron 91, cf. H. Schultz, Gött. gel. Anz. 1910, 34) tells the story
of Herakles and the Kerkopes to explain the line μὴ σύ ye μελαμπύγου τύχηις
(or τύχοις ; σύ ye is an easy alteration of the original rev, which is preserved
in the sentence of the subsequent narrative μηδὲν ἄδικον ποιεῖν, ἵνα μή τινος
μελαμπύγου τύχηις). The naughty boys, who interfere with all passers-by, are
warned by their mother, clearly not 'to be careful they do not fall in with a
black-rumped man', but 'to mind they do not have to do with a μελάμπυγος,
and so get what they deserve'. She tries to frighten the young rascals by a
threat of a not uncommon type, seeming to say: ‘Be careful, or a great wicked
eagle will come and carry you off.' This is how the children understand her
to begin with. Later on, hanging head-downward from the branch or pole,
when they catch sight of the hairy hero's hinder parts, they suddenly under-
stand what their mother meant, and burst out in a hearty laugh in their
surprise. What makes this interpretation more probable is this (and with
our fragmentary tradition one cannot claim more than probability): that in
this way the prophetic warning of the mother of the Kerkopes falls into line
with one of the most widespread types of oracular utterance. The prophecy
speaks of an animal; when fulfilment comes it is discovered that a man was
meant. A good deal of the charm of such stories lies in the solving of the
riddle καὶ coi (or, in other cases, τῶι δεῖνι) τί θηρῶν ὀνόματος μετῆν; (E. Phoen.
412). The story told in the Phoenissae (cf. schol. on 409) of the oracle given
to Adrastus about his future sons-in-law (wild boar and lion) isa good example
ofthetype. Cf. further, e.g., Hdt. r. 55. 2 (the solution in r. 91. 5), 5.928. 3, the
many oracles in Aristophanes’ Knights (197 ff., 1017 ff., 1030 ff., 1037 ff., 1051 ff.,
1067 f.), Lys. 770 ff., and so on. Result: for ὁ «eAawós in Ag. 115, besides the
ordinary meaning ('black' or 'dark' all over), we must take into account as
possible the other sense, — μελάμπυγος. If there was any ancient evidence for
the statement of the grammarians (Schol. Lycophr. 91, hence Etym. M.
p. 695. 50) that μελάμπυγος is used ἐπὶ τοῦ ἐσχυροῦ, the choice of this species
of eagle as a symbol for Agamemnon is highly appropriate.
Precise zoological identification of the species of eagle named by Aeschylus
must not be attempted. Even in modern times it is an open question whether
eagles which show certain differences in colouring, and to some extent in the
form of their feathers, are to be classed in one and the same species, cf., e.g., in
Brehm's Tierleben, iv, 2nd ed., 611 f., the discussion on the distinction between
Rock Eagle and Golden Eagle (there treated as two species, but in later
editions as one). Since eagles were especially important in augury, it is
possible that the Greeks felt it necessary at an early date to draw sharp
distinctions between as many different types as possible, even on the basis of
comparatively insignificant distinguishing marks (c.g. black or black-edged
tail feathers while the rest of the plumage is brown). That in Ag. 115 the
difference of the birds of prey contains a reference to the different characters
of the two brothers would in any case be an obvious guess; such references
are common in the allusive style of oracular utterances. But it is made quite
certain by the interpretation of the seer, who at once (122) connects with
the eagles the δύο λήμασι δισσοὺς Arpeidas, and thus brings their differences
69
line 115 COMMENTARY
sharply into prominence. λῆμα in Aeschylus can mean character or disposi-
tion in general, but the far more common meaning is courage. To Agamemnon
corresponds the eagle that (whatever view we take of κελαινός in this
passage) is reckoned as of outstanding strength, and spoiling for the fray ; to
Menelaus the one which by Aeschylus’ time, we may be sure, if not before,
was regarded by popular opinion as δειλός or, at any rate, as poorer in spirit.
This suits the general history of the character attributed to Menelaus. The
description of him as μαλθακὸς αἰχμητής (P 588), unduly emphasized by
Plato, Symp. 174 c, is unique in Homer, ‘but the depreciation of Menelaus has
already begun in Epic, and for the tragedians his inferiority was an accepted
fact’ (Wilamowitz, Griech. Tragoedien, iii. 280 n. 1).
116. φανέντες: plural, for the use in Aeschylus of the dual of such participles
is limited, cf. on 323.
Tkrap μελάθρων: of the geographical position of the μέλαθρα we are here
told nothing at all; it remains for some time quite indefinite (so rightly
Eugen Petersen, Die attische Tragödie, 635). All that matters here is that the
omen of the eagles appeared near the dwelling-place of the Atridae. Nor
ought we to infer from μελάθρων the existence of a fixed royal palace;
Euripides calls the lodging of the Atridae in Aulis δώματα in Iph. A. 440, 820,
854, μέλαθρον or μέλαθρα in 612, 678, 685. In this use οἱ μέλαθρον, μέλαθρα the
tragedians perhaps follow Homer (cf. Leaf on J 204). For the manner in which
later the scene of the action is fixed in a definite locality cf. on 19o f.
ἵκταρ in the time of Aeschylus can hardly have still been a part of the
living language. So far as we know, he is the only author to use it after
Hesiod (Theog. 691) and Alcman (1. 80 D.)—he has it again in Eum. 99] and
in the Edomi, fr. 67; for the proverb οὐδ᾽ ἔκταρ βάλλει quoted by Plato,
Rep. 9. §75 c (and thence in the Plato-lexicon of Timaeus and in the lexico-
graphers ; the evidence is now conveniently accessible in Ada Adler's edition
of Suidas, on ii, p. 628. 277), which has enticed later writers into imitation, is
probably very ancient. Possibly the word belongs to religious or liturgical
language.
χερὸς ἐκ δοριπάλτου: 1.6. as a δεξιὸς ὄρνις. The side on which the omen
appears is of decisive importance, and thus, in spite of the admixture of signs
of ill, the favourable meaning wins the day: 145 δεξιὰ μὲν κατάμομφα δὲ
φάσματα.
δορίπαλτος occurs only here. For its use in an active sense cf. on 12.
118. παμπρέπτοις (for verbal adjectives with the prefix rav- see on 960) ἐν
ἕδραισιν : Hermann explains ‘caeli fausta portendens regio est’, and refers to
E. Her. 596 ὄρνιν ἰδών tw’ οὐκ ἐν αἰσίοις ἕδραις ("unterwegs war mir ein übler
Vogelflug begegnet’ Wilamowitz, cf. also his commentary). But it is unlikely
that Aeschylus imagined the two eagles as tearing the hare to pieces in
mid-air. Such an idea would be most unnatural. Compare also the celebrated
coins of Akragas—there is a good reproduction of a decadrachm and a tetra-
drachm in G. F. Hill, Coins of Ancient Sicily (1903), plate vir. 15, 17 with
text on pp. 119 ff., in K. Regling, Die anttke Münze als Kunstwerk, Berlin
1924, plate XXIV. 532 f., and in Beazley and Ashmole, Greek Sculpture and
Painting, fig. 242 (the decadrachm) ; a coarser illustration of the decadrachm
also in D’Arcy Thompson, A Glossary of Greek Birds, 2nd ed., title-vignette
and p. 14. There the two eagles tearing a hare to pieces are perching ; in one
79
 COMMENTARY line 119

type (the decadrachm) the hare clearly lies on a point of rock and they perch
on it, and even where this detail is lacking, they are at rest and not in flight.
Whether it be rock or tree-top Aeschylus does not care, and so he contents
himself with the non-committal éôpaiow; the important thing for him is
that the rending of the hare takes place where the whole host can clearly see
it. We are accustomed to think of birds which serve for omens as showing
themselves in flight. That is the normal thing; we find it in the rules for
observing and interpreting the flight of birds set down in Ephesus in
Aeschylus’ own day (Dittenberger, Syll. no. 1167). But the eagle at rest can
also serve on occasion as an auspicium: Xen. Anab. 6. 1. 23 καὶ ὅτε ἐξ ᾿Εφέσου
δὲ ὡρμᾶτο Κύρωι συσταθησόμενος, ἀετὸν ἀνεμιμνήισκετο αὑτῶι δεξιὸν φθεγγό-
μενον, καθήμενον μέντοι, ὅνπερ ὁ μάντις... ἔλεγεν ὅτι μέγας μὲν οἰωνὸς εἴη καὶ οὐκ
ἐδιωτικὸς καὶ ἔνδοξος, ἐπίπονος μέντοι (cf. δεξιὰ μὲν κατάμομφα δὲ φάσματα), for
which the seer then goes on to give the reasons.
119. βοσκόμενοι Aayivav ἐρικύμονα φέρματι γένναν. This passage was rightly
understood long ago by Schütz, Porson, Blomfield, and Hermann (followed
by Wilamowitz), but is still arbitrarily altered by other scholars. ‘Vescentes
leporino genere mulia prole fecundo. Ceperant enim aquilae leporem feminam,
eamque gravidam' (Schütz). M has ἐρυκύματα (with a marginal gloss πολυκύ-
μονα which perhaps points to ἐρικύμονα as the reading of the exemplar), VF Tr
ἐρικύμονα. The latter alone conforms to the analogy of ἀκύμων (E. Andr. 158)
and ἐγκύμων, and to the general rule by which compounds derived from
neuters in -ua are formed in -μων (e.g. ἀναίμων, πολυκτήμων, πολυπράγμων,
etc.); cf. Debrunner, Griech. Wortbildungslehre, p. 72. There is therefore no
reason to regard the reading of M as a rare metaplasm and not rather as a
scribal error induced by the following φέρματι (so rightly Hermann ; more slips
of the kind are quoted by P. Maas, Sokrates, iii, 1915, 234 n. 4, and, with
examples from the text of Aristotle’s Poetics, Lobel, C.Q. xxiii, 1929, 76 n. 2;
for a modern example see Housman, /. Phil. xxxv, 1920, 287).
λαγίνα γέννα denotes what springs from the hare and consequently belongs
to the race of hares, as in Prom. 164 οὐρανίαν γένναν, Ag. 760 σφετέραι δ᾽
εἰκότα γένναι, and so on. Here, in the deliberate obscurity of the oracular
style, the periphrastic expression denotes a single individual animal (the
scholion is right: Aayívav γένναν" τὸν Aaywóv). So Euripides in the exalted
language of his dactylic hexameters (Andr. 119) makes the Chorus say to
Andromache ®6ids ὅμως ἔμολον mori σὰν Aowjrida γένναν (Murray is wrong
in objecting to γένναν), but there the idea of origin is more strongly emphasized
(I who come from Phthia, you who come from Asia’). Aeschylus might
perhaps be blamed (not by me) for choosing here, where he wishes to
denote the pregnant hare, a word that has its own quite different associa-
tions with procreation and birth. But the text is neither obscure nor am- .
biguous ; ἐρικύμονα with its straightforward active sense makes the meaning
perfectly clear. The old suggestion $épuara, revived by Hartung, would
compel us to assume a sense of ἐρικύμων which, though not impossible, is very
doubtful. Further, we should then have to ask how far the ‘embracing’ type
of apposition Aayívav, ἐρικύμονα φέρματα, yévvav—that is, the word-order
exemplified in inter densas, umbrosa cacumina, fagos—is found in early Greek
poetry. Does it occur there at all, or do we find in the poetry of the pre-
Hellenistic period antecedents only to an arrangement like that of Colluthus
71
line 119 COMMENTARY
109 ποιμενίη δ᾽ ἀπέκειτο, βοῶν ἐλάτειρα, καλαῦροψ, in which the inserted apposi-
tion does not consist of noun plus adjective? E. Hipp. 1037 ὅρκους παρασχών,
πίστιν ov σμικράν, θεῶν represents a somewhat different type, since the
noun (ὅρκους) comes first and the attribute consists of the genitive θεῶν.
(H. Boldt, De libertore colloc. verb., Diss. Göttingen, 1884, roo ff., Leo on the
Culex, p. 37, Norden on Virgil, Aen. 6. 7 f., offer no materials for an answer to
this question.) In Aeschylus I do not recall meeting anything of the kind.’
More than anything else, however, the corresponding passage in 137 αὐτότοκον
πρὸ λόχου μογερὰν πτάκα θυομένοισιν makes it probable that the object of
βοσκόμενοι here, as of θνομένοισιν there, is the mother hare.
The adjective λάγινος does not occur elsewhere, but is quite correctiy
formed (see Lobeck, Proll. Pathol. 200 f.; on the wide use made of the suffix
-wos cf. Chantraine, Formation des noms, 201 f., and Buck and Petersen, A
Reverse Index, 261 f.). It is most easily regarded as derived from Aayós, which
besides being the ruling form in Ionic and recorded for Epicharmus, is also
found in Sophocles and several times in Attic Comedy (locus classicus:
Tryphon fr. 19 Velsen ap. Athen. 9. 400 a-d; cf. also Lobeck, Phrynichus,
p. 186); Aeschylus himself has λαγοθήρας, Ar. Lys. 789 λαγοθηρεῖν. The form
- Aayós is secondary (for the details see Schwyzer, Griech. Gramm. i. 557 n. 1),
the primary form being λαγώς (-ὥς), a compound meaning 'slack-ear', as
Schwyzer saw (Kuhns Zeitschr. 37, 146, cf. also his Griech. Gramm. i. 438 n. 3).
ἐρικύμονα is probably rightly glossed (in M) by πολυκύμονα; so, e.g.,
Blomfield ‘ad magnum numerum in utero concipiens’.
φέρμα is found only in the prayer A. Suppl. 69o (of the fertility and fruit-
fulness of the earth) and here, in the account of a portent in its oracular
setting. Schwyzer, Griech. Gramm. i. 523, accounts it one of the early forma-
tions that nevertheless first appear after Homer, and compares Old Church
Slavonic bréme, ‘foetus’. In the list of members of the Erechtheid phyle
killed in battle in the year 459/8, IG i? 929 (= Tod, Greek Hist. Inscr. 26), 46,
we read the Attic personal name Ζυνφέρμιος, from which Bechtel, Histor.
Personenaamen, 445, infers the name Σύνφερμος, ‘twin brother’. Aeschylus
very likely took φέρμα from the language of liturgy, and if so, we must class
it with the borrowings from ‘the style of oracles’ pointed out by Wilamowitz,
Interpr. 166 n. 2, in just this section of the parodos.
The dative φέρματι, depending on ἐρικύμονα, is no less natural in itself than,
for instance, S. Trach. 235 νόσωι βαρύν. But the ‘redundant’ addition here
‘gives to the expression that fullness which is characteristic of the exalted style.
! [t is true that Weil (1861) regards Eum. 302 as an example of this word-order and ex-
pressly compares it with Ag. 119. He punctuates ἀναίματον, βόσκημα δαιμόνων, σκιάν and
explains thus: “inter adiectivum et substantivum interposita sunt, quae rei rationem
aperiunt, quasi dixisset ἀναίματον (δαίμονες γάρ σε βοσκήσονται) σκιάν. Wilamowitz adopts
the same punctuation, and so Mazon: ‘ombre vidée de sang qui aura repu des déesses',
G. Thomson: ‘a bloodless shadow’. But this interpretation is not necessary. Assuming
that the conjecture σκιάν gives us the true text (which is by no means certain), it is more
natural to retain the usual punctuation (see the good note by Paley on ἀναίματον βόσκημα
δαιμόνωνλ and regard σκιάν (without an epithet) as a final and impressive indication of what
Orestes is destined to become. (I cannot agree with G. Hermann, Opusc. vi. 2. 6o, that
σκιάν is ‘pitifully flat’; like him, Blass, too, finds σκιάν ‘lame and unpleasing’.) Headlam (in
a note to his translation) conjectures that the line does not belong here but was added as a
parallel, which seems to me well worth considering; he retains σκιά, and translates ‘the
bloodless meat of Spirits below a shadow’.
72
 COMMENTARY line 121
BAaßevra: to be taken, apparently, with γένναν. If this assumption is
correct, this instance would have to be classed with the two or three others in
which the masculine of an aorist participle with stem ending in -vr- is used
as a feminine (cf. on 562). This, to be sure, is doubtful; but even so, it seems
less hazardous than to follow either Hermann’s ‘ad PAaßevra intelligendum
est Aaywov’, which would be unsuitable here where the sex matters so much
(so rightly Schneidewin), or Paley’s expedient ‘the construction is as if the
poet had said Adywa γεννήματα, or rather Aayw καὶ ra τέκνα αὐτῆς᾽. The latter
assumption, adopted also by Donaldson, New Cratylus, 3rd ed., 682, ‘things
(i.e. the hare and her young) stopped from running any more races’, would
in itself be welcome, but for this type of the ‘constructio κατὰ σύνεσιν᾽ I can
offer no parallel.
λοισθίων δρόμων: genitive of separation (Kühner-Gerth, i. 396; see also
below on 479 φρενῶν κεκομμένος), ‘prevented from running the final course’.
For the poetic use of βλάπτειν with a genitivus separativus Blomfield referred
especially to a 195 ἀλλά νυ τόν γε θεοὶ βλάπτουσι κελεύθου, which he suggested
(perhaps rightly) might be Aeschylus' model. Further evidence for the
Homeric βλάπτειν — 'to impede' is collected by O. Becker, 'Das Bild des
Weges', Hermes, Einzelschriften, iv (1937), 8 n. 6. There is probably also an
allusion to the proverbial περὶ ψυχῆς (or περὶ ἑαυτοῦ) δρόμος. Whether a
reference to the proverb is to be seen in Homer X 161 ἀλλὰ περὶ ψυχῆς θέον
Ἕκτορος ἱπποδάμοιο, we cannot say for certain. But Herodotus uses it with
various modifications (9. 37. 2 τρέχων περὶ τῆς ψυχῆς; the other passages
quoted by Stein on 7. 57), and says, actually with reference to a portent of
which a hare forms part, in 7. 57. 1 περὶ ἑωυτοῦ τρέχων. Cf. Ar. Wasps 375
τὸν περὶ ψυχῆς δρόμον δραμεῖν (further passages in Lobeck, Paralip. 510 f.).
Whether, as is generally assumed, Ar. Frogs 192 is alluding to the special
variation of the proverb (Zenobius 4. 85) λαγὼς τὸν περὶ τῶν κρεῶν τρέχων, 1
do not know; in any case there is no reason to think of that form of it in
connexion with the passage before us. It is clear that Aeschylus has ennobled
the proverbial expression by his choice of words. λοίσθιος, in general a rare
word, is found only in the most exalted poetic diction; in Tragedy we first
meet it in the Γλαῦκος Ποτνιεύς of Aeschylus, Pap. Oxy. 2160, fr. 2. 2. 12, then
once in each play of the Oresteia ; for the history of the word see Wilamowitz
on E. Her. 23. In δρόμων Platt, C.R. xi, 1897, 94, wishes to recognize a tech-
nical nicety of the language of the chase; this seems alien to the style of the
passage, nor is it suggested by the thought.
121. αἴλινον. For the history of the cry see Wilamowitz, on E. Her. 'Zweite
Gesangnummer' (ii, and ed., 84 f.); see also his Euripides Hippolytos (1891),
p. 28 f. The double cry atAwov atAwov comes also in S. Aj. 627, E. Or. 1395,
obviously from the ritual of mourning; for the doubling compare ὑμὴν ὑμήν
(and its variants), which is like atÀwov an ejaculation probably of non-
Greek origin whose meaning is lost (cf. Wilamowitz, locc. citt. and Griech.
Verskunst, 28 f.; P. Maas, RE ix. 131), used in the refrain of wedding-songs as
αἴλινον is in the refrain of dirges. Similarly θρίαμβε διθύραμβε (Wilamowitz) ;
it is also possible that we should ‘still hear in ie the remnant of the
doubling of in’ (Wünsch, RE ix. 145. 65). As atÀwov appears predominantly
(though not exclusively, cf. Wilamowitz, E. Her. ii. 85) as a cry of sorrow, it
forms here a strong contrast to the following εὖ.
73
line 121 COMMENTARY
It was observed by ancient grammarians (Et. M. 35. 1) that the cry aldıvov
has its place ἐν τοῖς ἐφυμνίοις, as we find it here. Such ἐφύμνια, refrains, belong
to the oldest (and from a religious point of view the most important) elements
of liturgical song. Otfr. Müller, A. Ewm. p. 91, remarked, in connexion with
the refrain of the δέσμιος ὕμνος of the Eumenides, that ‘such a repetition of
the words which express the especial purpose of the whole action was
characteristic of incantations and promises of destiny ; we find it, e.g., in the
love-magic of Theocritus . . . and in the song of the Fates at the wedding
of Thetis in Catullus'. P. Huvelin, ‘La notion de I’tnsuria dans le trés ancien
droit romain’, Mélanges Ch. Appleton, Lyons, 1903, 58 n. 5, asks whether
anyone has yet considered what influence the religious idea of repetition may
have had on the formation of the refrain in song. 'Generally speaking, all
refrains of ancient times are to be regarded as early stages in the development
of poetry, and as being in origin cries of invocation. It was from these
ejaculations, in so far as they were ritual in character, that the magic power
proceeded. Consequently, as art made progress and song came into being,
they could not be dispensed with ; they had to be shifted to the end, or some-
times the beginning, of the stanza produced by art' (L. Deubner, Neue Jahrb.
f. d. klass. Altert. xliii, 1919, 400). Cf. also W. Kranz, Stasimon, 130. Here in
the parodos the refrain undoubtedly fulfils an artistic purpose, bringing
out the structure more clearly and also increasing the solemnity of the
dactylic opening. But the old religious force still survives: the refrain
rounds off first the story of the τέρας (cf. 125) and finally the prophecy of the
seer, and thus with its cry of alarm and its prayer for a happy issue it serves
to heighten the effect of a ‘promise of destiny'.
τὸ δ᾽ εὖ νικάτω: this is transformed in the Ocdipus of Nicomachus(Lex. M essan.
ed. Rabe, Rhein. Mus. xlvii, 1892, 406) ὅτε μὲν λῶιστον, τόδε νικώιη (anapaests).'
τὸ εὖ here and in 349, as also in E. Her. 694 and Eur. fr. 918. 3 N., should be
regarded as a substantival use not of the adverb but of the old adjective, of
which the masculine form survives in the Homeric evs. Such is the opinion
of Wackernagel, Synt. ii. 142 (and of Bruhn before him, see below), although
E. Schwyzer, Berl. Sitzgsber. 1938, 81, is sceptical. Wackernagel's view is sup-
ported by the passages of Tragedy in which εὖ διδόναι means 'to give good
things’, S. Oed. R. 1081, Oed. C. 642, E. Alc. 1004, Andr. 750,0r. 667, used always
of the gods (three times in prayers). It is unlikely that εὖ in εὖ διδόναι should
be regarded as an adverb. (The expression in E. Suppl. 463 f. κακοῖσιν ὡς
ὅταν δαίμων διδῶε καλῶς [quoted to me by W. S. Barrett] is probably formed
after the model of εὖ διδόναι, and the same may be true of E. Med. 879
θεῶν ποριζόντων καλῶς). We may therefore agree with E. Bruhn (on Oed. R.
1081) that ‘the so-called adverb εὖ is really the neuter of the old adjective
evs, surviving here as a noun in the accusative singular neuter, as it does in
the nominative in the phrase τὸ δ᾽ εὖ mxdrw’. Perhaps we may assume
that the isolated noun εὖ has survived as a relic in the language of prayer
in such phrases as E. Alc. 1004 χαῖρ᾽, ὦ πότνι᾽, εὖ δὲ Soins, Oed. C. 642 ὦ Ζεῦ,
διδοίης τοῖσι τοιούτοισιν εὖ.

122. κεδνός : ‘si dice subito, in prima sede nel verso e nel periodo, che [the
seer] à xedvds, che per Eschilo ha un significato oscillante tra "buono" e
τ Nauck, Tragicae dictionis index, xxv f., gives the fragment to Nicomachus the tragedian
from Alexandreia Troas, A. Körte, RE xvii. 462, to Nicomachus the comic poet.

74
 COMMENTARY lines 122 f.

“prudente”, quasi a influenzare il giudizio che si dovrà dare della sua

predizione’ (Ferrari, La parodos, 368). To the Athenians the word xeövos is
of course a Homeric γλῶττα.
orparöpavrıs here only. In the well-known passages Poet. 22, p. 1450Ὁ,
Rhet. 3. 3, p. 1406°1, Aristotle tells us that διπλᾶ were characteristic of the
dithyramb; Aristophanes had already poked fun at its tendency in this
direction. For such compounds in the citharodic nomos of Timotheus see
Wilamowitz, Timotheos, p. 45 f.; for their frequency in the dithyrambic poets
cf. G. Meyer, ‘Die stilist. Verwendung der Nominalkomposition im Griech.’,
Philol. Suppl. 16, Heft iii (1925) 157 f. But Aeschylus has in general a prefer-
ence for διπλᾶ, so that it would be wrong to ascribe their frequency in this
ode to the influence,of the citharodic nomoi, which according to Aristophanes
appears in this and similar dactylic odes of Aeschylus (cf. p. 58 f.).
122 f. ἰδὼν δύο λήμασι δισσοὺς Arpeidas κτλ. This is the first we hear of
Calchas, but of course we are meant to suppose that his attention was at once
called to the eagles, and that he has been carefully observing first their flight
and then their rending of the hare. Now his eye returns to his immediate
surroundings; he sees before him the two Atridae, and forthwith he under-
stands the meaning of the omen. ‘When he beheld the two sons of Atreus
with their two different characters, then he knew that the valiant eaters of
the hare were the captains that led the host on its way.’ The general idea is
rightly understood in the paraphrase (in Tr and F): ἤγουν εἰκόνισεν εἶναι τοὺς
(τοὺς δύο F) ἀετοὺς τοὺς δύο Arpeidas. Now for the details. μαχίμους is taken
by Stanley, Schütz, and others as an epithet of Ἀτρείδας ; so, among more
recent scholars, e.g. Headlam. Blomfield took it with Aayodairas. This is
better in point of style, for it gives every noun its attribute, which, in
this carefully balanced sentence, is an advantage.! At the same time it
produces a stronger thought. Of the warlike expedition of the Atridae we
have already heard, and the addition of μαχίμους to their name would be
superfluous. On the other hand, if applied to the eagles, the adjective has
special point; their warlike deed, their act of pillage, has just been brought
under the very eyes of the army. The eagles have shown themselves to be the
warrior birds that they are by nature, and this fact makes it easier for the
seer to take them as representing the Atridae who are just going forth to war.
There is good reason for supposing that to the Athenian audience the arrange-
ment of the sentence was made still more intelligible by the style of delivery.
For we have plenty of evidence to show that Aeschylus likes to put syntactical
breaks in strophe and antistrophe in the same place in the line (cf., e.g., on
210). Here we have a case in point. In 105 there is a stop after évreAécv, and
in 123 Arpetôas, i.e. the end of the participial clause, marks the conclusion of
a distinct element of thought (a participial clause may serve as a self-con-
tained 'kolon', cf. Ed. Fr., Kolon und Satz, ii. 321 ff.).
δύο λήμασι δισσοὺς ᾿Ατρείδας. This use of δισσοί is unusual, but it is un-
fortunate that scholars have gone on echoing the verdict of Lobeck (on S. 47.
151) and Dindorf? that the word will not do here, although Hermann referred
! I have not forgotten that Aeschylus is not Ovid.
* His conjecture ἴσους is bad sense and false prosody (the ‘productio epica" in polysyllables
such as Prom. 185 ἀπαράμυθον, Eum. 387 ücodorairake is entirely different). Nevertheless,
ἴσους was again advocated by W. M. Edwards, C.Q. xxxiii, 1939, 204 f. [Cf. Addenda.]

75
lines 122 f. COMMENTARY
to S. El. 645 δισσῶν ὀνείρων, where one of the explanations in the scholia,
διπλῆν ἐχόντων φύσιν, is probably correct (cf. Kaibel and Jebb ad loc.;
the latter refers to Arist. Rhet. i. 15. 13, p. 1375?26, μάρτυρές εἰσι διττοί, of
μὲν παλαιοὶ of δὲ πρόσφατοι). We may apply the paraphrase διπλῆν ἔχοντας
φύσιν to λήμασι δισσούς as well Wilamowitz rightly has ‘die beiden
verschiedengemuteten Atreussóhne', cf. also Schuursma, 92 f. For the con-
struction with the dative we may compare Pindar, N. 7. 54 φυᾶι δ᾽ ἕκαστος
διαφέρομεν βιοτάν, λαχόντες ὁ μὲν τά, τὰ δ᾽ ἄλλοι. It is hardly possible with
Abresch, Ahrens (p. 275), and Headlam to take δύο λήμασι together, certainly
not if we are to understand it as Ahrens does: ‘the two Atridae are both
μάχιμοι, but not évi λήματι; their characters and motives are of two different
kinds’ ; for as we are reminded by Platt, J. Phil. xxxv, 1920, 90, δύο for δυοῖν
is not found in Tragedy.’ One way out is to take λήμασι with δύο as a kind of
‘dative of respect’, but this is not attractive. It is, however, altogether
unjustified when Ahrens and Headlam appeal to the order of words in
support of their method of construing the clause, for ôvo . . . Arpeidas repre-
sents a very common type of hyperbaton; for a provisional sketch see my
Iktus und Akzent, 165 f. For the difference in the courage of the brothers
see on IIS.
“Arpeidas Blomfieldius Monkio auctore. Recte fortasse.’ It seems im-
possible to go beyond Hermann's cautious comment and make out whether
Aeschylus here wanted the name to be pronounced as three syllables or as
four (for the problem in general cf. K. Meister, Die Homerische Kunstsprache,
51). The spondee at the beginning of the second line would not be more
objectionable than the κεδνὸς δέ with which the antistrophe opens.
ἐδάη: not only did he understand that it was the Atridae who were indi-
cated by the two eagles; the verb seems also to convey that the difference
of character in the two pairs, and their correspondence one with another,
became clear and significant to him. ἐδάη is specially appropriate to express
the grasping of such a difference as this, cf., e.g., Homer I” 208 ἀμφοτέρων δὲ
φυὴν ἐδάην καὶ μήδεα πυκνά, τ 325 Í. πῶς yap ἐμεῦ σὺ ξεῖνε δαήσεαι εἴ τι γυναικῶν
ἀλλάων περίειμι νόον καὶ ἐπίφρονα μῆτιν;
124. πομποὺς ἀρχάς. The τ᾽ written in the MSS after πομπούς is wrong, as
was seen by Fr. Thiersch, De analogiae Graecae capitibus, ii. 58,andby Karsten:
it destroys the relation between the object (Aayodairas) and the predicative
noun (ἀρχάς). It happens often enough that the succession of several words
in the same case puzzles a copyist, and leads to the mistaken insertion of a
connecting particle? Many editors have accepted Karsten's πομπᾶς apxovs
‘expeditionis duces’ (which—with retention of the r’—is as old as Musgrave),
for which the dpyovs of FTr supplies a foundation. It is quite possible that
this is what Aeschylus wrote. But in favour of the MS reading πομπούς
! The only exception cited by Kühner-Blass, 1. 633, happens to be our passage, and the
solution offered is, not the obvious correct interpretation, but a bad conjecture. The rule
is given implicitly in Passow’s dictionary, explicitly by L-S.
2 Examples are to be found everywhere. Besides Ag. 1452 and A. Suppl. 42 f. (see on
1526) I will only mention A. Suppl. 646 f. Δῖον ἐπιδόμενοι πράκτορα {re del. Bergk] σκοπόν
(I know that some editors persist in preferring an intolerable text to the simple and evident
emendation), Cho. 428 (where the deletion of the καί is demanded by the sense and accepted
by Enger, Wilamowitz, and Headlam), and Cho. 586. Pap. Rylands 547 in E. Phoen. 654
offers not χλοηφόροισιν ἔρνεσιν κατασκίοισιν but xA. ἔρν. καὶ κατασκ.
76
 COMMENTARY line 129

we should probably adduce the words used a few lines before (in 111) of
the pair of eagles Ἀχαιῶν δέίθρονον κράτος... πέμπει... Τευκρίδ᾽ én’ alav;
here, where the relation is established between the eagles and the Atridae,
we should not tamper with πομπούς unless we are forced to do it. It would
be possible to think of πομποὺς dpxäs (Rauchenstein), ‘those who send forth,
who prepare the way for, the chieftains’ (dpxd used of the two brothers
together, as we have had rayd), or perhaps ‘those who prepare the way for
the beginning of the war’. But it seems to me simpler and clearer to read
πομποὺς ἀρχάς, ‘the conducting chiefs’ (so, e.g., L-S). Calchas now realizes
that the eagles are—i.e. that they represent—the commanders that set the
train of war in motion. dpxai here as in E. Phoen. 973 λέξει γὰρ ἀρχαῖς (Schol.
τοῖς ἄρχουσι δηλονότι) καὶ στρατηλάταις. πομπός used as an adjective recurs
in 299.
125. οὕτω: cf. on 615.
τεράιζων: only here in extant literature. It may well be a coinage of
Aeschylus ; pardber 994 is probably another.
126. The speech begins with a fresh line (as in 206, 410), corresponding to tlıe
‘old epic technique still maintained by Apollonius Rhodius’ (Pfeiffer, PAslol.
xcii, 1937, 12 n. 22). For a speech beginning in the middle of a line cf. on 590.
χρόνωι: cf. Bacchyl. 11 (10). 120 Πριάμοι᾽ ἐπεὶ χρόνωι βουλαῖσι θεῶν μακάρων
πέρσαν πόλιν κτλ.
μέν belongs to the whole sentence; the corresponding clause follows in 131
οἷον μή τις dya κτλ. (so rightly Schneidewin). The uév-clause contains the
prophecy properly so called, the establishment of facts of which the prophet
possesses certain knowledge. In contrast with this, the clause οἷον μὴ κτλ.
expresses apprehension, something derived by Calchas from the baneful
element of the portent (the destruction of the still unborn young) and from
his knowledge of the nature of the πότνια θηρῶν, apprehension that comes very
close to probable conjecture.
ἀγρεῖ: the use of this ‘prophetic’ present instead of future is well known.
‘Oracular language in particular is fond of describing future events in terms
proper to the present (‘Gegenwartsausdriicken’)’, Wackernagel, Syntax, i. τότ;
cf. also Kiihner-Gerth, i. 138; Neil on Ar. Knights 127. ἀγρεῖν as an equivalent
of αἱρεῖν, an Aeolic usage which owes its place in literary language to Homer,
is found in Drama only here. Wackernagel, Unters. zu Homer, 166, reckons
with the possibility that in his text of Homer Aeschylus read, in the speech
of Calchas, B 329, τῶι δεκάτωι δὲ πόλιν ἀγρήσομεν εὐρυάγυιαν, where the MSS
have αἱρήσομεν. We must, however, bear in mind what Wilamowitz, Sappho
und Simonides, 275, remarks with reference to αὐτάγρετοι in the iambus of
Semonides of Amorgos.
127. κέλευθος of going forth to war, also Pers. 758 ; this is probably developed
from a Homeric use of the word (e.g. A 504 κέλευθος = ‘advance’), cf., besides
the lexica, J. H. H. Schmidt, Synonymik der griech. Sprache, iv. 630 f.
128 f. πάντα δὲ πύργων «rA.: capture is followed by sack. πύργοι, as often in
Homer, serves to denote walls as well as towers.
129. κτήνη: the older explanation is ‘opes’, following the scholia (xryvy
κτήματα, cf. Hesych. «rfvea: χρήματα) ; so, for instance, Blomfield and Her-
mann, and again in recent times Schuursma, 23. This is rejected by Paley,
who translates ‘the public flocks and herds'. If this is right, as I believe it to
77
line 129 COMMENTARY
be, then Aeschylus has adopted the Ionic word! for flocks and herds; cf.
Bechtel, Griech. Dialekte, iii. 311 1, Reference to wealth in the form of live-
stock yields a satisfactory sense. On the other hand, the possibility must be
admitted that Aeschylus knew κτήνη or κτήνεα as a literary word for ‘posses-
sions’, and may well have used it in that sense himself. It is so employed, as
Platt, J. Phil. xxxii, 1913, 46, has pointed out, in the Hesiodic Catalogue of
the suitors of Helen, fr. 94. 49 Rz., 3rd ed., where κτήνεσσι picks up the
preceding χρυσόν re λέβητάς τε κτλ. (the passage? is overlooked by L-S). But
πρόσθε πύργων (see below) makes it almost certain that livestock is meant
here.
Otfried Müller’s suggestion in Gölt. gel. Anz. 1834, 1982, δημιοπληθέα (the
MSS have δημιοπληθῆ) is highly probable.
Blomfield and Hermann took δημιοπληθῆ (for the formation of the word
see Chantraine, Formation des noms, 424 ff., Schwyzer, Griech. Gramm. 1. 513)
rightly as πολλὰ δήμια and compared the parallel use of ἀρσενοπληθῆ ἑσμόν
Suppl. 30 and γυναικοπληθὴς ὅμιλος Pers. 122: the herds (or the possessions ?
see above) consist of a quantity of beasts (or things) which are common
property, as the ‘swarm’ consists of a quantity of men and the crowd of a
quantity of women; Wilamowitz on E. Her. 1273 (where he also compares
E. Alc. 951 f. ξύλλογοι γυναικοπληθεῖς) paraphrases κενταυροπληθής with ev
ὧι πληθύουσι κένταυροι. In S. Aj. 175 (πανδάμους ἐπὶ βοῦς ayedaias) πανδάμους
indicates that the herds belong as booty to the commonalty of the Greek
host, so that the underlying conditions are different from what is meant
here ; but as far as the word goes, the herds of the Trojan community might
equally be styled πάνδημα.
In front of this we must certainly read πρόσθε rà. προσθετὰ in M is simply
due to wrong division of words; Pauw’s πρόσθετα, adopted by Hermann, is
impossible: ‘non esset congesta, sed additicia’ (Weil). πρόσθε πύργων belong
together (Klausen). It is perhaps no accident that of the four passages in
which Homer has the ‘prepositional’ use of πρόσθε following its noun, two
indicate what we have here, the space before a walled town: M τάς ἐκ δὲ
τὼ ἀΐξαντε, πυλάων πρόσθε μαχέσθην and (with the insertion of another word,
as in our passage) M 445 f. ὅς pa πυλάων ἑστήκει πρόσθε (cf. A. Sept. 525 πρόσθε
πυλᾶν). Aeschylus may have felt his order of words to be the more justifiable
in that Homer has once a still longer insertion between the genitive and its
‘preposition’ πρόσθε: À 54 τάων οὔ τοι ἐγὼ πρόσθ᾽ ἵσταμαι. Another adverb in -dev
used ‘prepositionally’ is divided from the preceding noun to which it belongs
by an intruding substantive (as in Ag. 129) in Pindar, Ol. 3. 12 “EAAavodixas
γλεφάρων Αἰτωλὸς ἀνὴρ ὑψόθεν... βάληι (explained in the paraphrase by
ἐπάνω τῶν ὀφθαλμῶν). The sense of πύργων πρόσθε in the present passage

' Plato says (Critias 109 c) καθάπερ ποιμένες κτήνη πληγῆι νέμοντες ; since this belongs to a
passage where Plato is alluding to Heraclitus B 11 Diels πᾶν ἑρπετὸν πληγῆι νέμεται, I infer
that he either found κτήνεα already used by Heraclitus in just this context, or inserted the
Heraclitean word (cf. B 29) from his general recollection; this is then not incompatible
with his dislike for Ionic words (Wilamowitz, Platon, ii. 412). In Xenophon the Ionism
κτῆνος is not surprising.
2 In the passage quoted by Platt from the same fragment I. 25, κτήνει is not quite certain.
3 An earlier protest against Hermann's rendering of πρόσθετα by ‘collata a populo' was
made by Hartung. W. M. Edwards’s translation (C.Q. xxxiii. 1939, 207) of πρόσϑετα, ‘con-
signed to her (scil. Moipa)’, is not happy.
78
 COMMENTARY line 131
is illustrated by Paley with the help of Suppl. 691 πρόνομα (schol. πρὸ τῆς
πόλεως νεμόμενα) Bora (a certain restoration), ‘the cattle that pasture without’;
cf. also Wilamowitz, Interpr. 40 n. 1, who takes a sounder view of Ag. 129
there than in the apparatus of his edition.
Good store of cattle represents an important, and to the besieging army a
particularly welcome, part of the enemy’s wealth ; that the Trojans preserved
considerable supplies of food throughout the siege is mentioned again in 331.
Troy was not so completely invested by the Greek forces (cf. the story of
Troilus) that it was impossible for the inhabitants to maintain herds of cattle
in some places outside the walls. Cf. also on 1168.
130. As is shown by the whole expression μοῖρα λαπάξει πρὸς τὸ βίαιον,
μοῖρα here does not mean ‘division’ (Murray) but (as in Homer) ‘doom’,
‘death’, and so most editors understand it. With πρὸς τὸ βίαιον compare
πρὸς βίαν and Prom. 212 πρὸς τὸ καρτερόν; cf. on 382. The expression as a
whole reminds us of the oath in Sept. 47 Aamd£ew ἄστυ Καδμείων βίαι. In view
of the other epic usages in this ode, Wellauer and Ahrens wish to preserve
the Homeric ἀλαπάξει (so the MSS, λαπάξει Elmsley); this is possible, but
cannot be proved. The initial a of ἀλαπάζειν has been regarded since Passow
as an example of ‘Vokalprothese’ (in general cf. Schwyzer, Griech. Gramm. 1.
411 f.); cf. Bechtel, Lexilogus zu Homer, 28. This view was held by ancient
grammarians, cf., e.g., Athen. 8. 362 f. (further evidence is given by Adler on
Suidas s.v. λάπαθον, to which may be added Eustath. on the //iad, p. 65. 28,
Ῥ. 838. 61).
131. οἷον μή = μόνον μή (so the paraphrase in the scholia), ‘modo ne’: this
use of οἷον seems to be rare, but that it is sound there is no call to doubt;
the construction is the same in E. Cycl. 219 μὴ ᾽μὲ καταπίηις μόνον and elsewhere
with μόνον μή. Headlam infers from the oracle-parodies in Anth. Pal. 11. 163 and
11. 365 (cf. also Rothstein on Propert. 4. 1. 147) that it was the normal practice
in oracles to anticipate some untoward circumstance with a phrase beginning
with μόνον μή or something like it.
Hermann's restoration dya has rightly met with general approval. The
best explanation is offered by the gloss on Hdt. 6. 61. 1 (φθόνωι καὶ ἄγηι
χρεώμενος) preserved in Phot. Berol. r5. 26, Suidas A 212, Etym. M. 9. 1
ἄγη παρ᾽ ᾿Ηροδότωι βασκανία.
It is, of course, true that θεόθεν here serves as attribute to dya, and there-
fore this expression is rightly compared with Sept. 841 πατρόθεν εὐκταία φάτις,
Pind. Ol. 3. 28 dvdyxa πατρόθεν, and the like by M. Lejeune, Les Adverbes
grecs en -θεν (1939), 155 ff., 176, but I cannot agree with him that ‘on n’y
sent plus une indication de provenance’.
The chief difficulty of the sentence lies in the last word στρατωθέν, which,
however, is certainly correct, so that none of the many conjectures need be
considered. Of the scholars who retain the MS reading, some have taken
refuge in very violent interpretations. Thus Ahrens (p. 280) explains στόμιον
στρατωθέν as frenum ab exercitu iniectum. This is impossible Greek; Ahrens's
allegedly analogous verbs in -οῦν are all quite different. Wecklein (comm.)
and Wilamowitz (Interpr. 166 n. 2, and Hermes, liv, 1919, 50) explain as ‘das
aus einem Heer verfertigte Gebiss’, ‘the bit made out of an army’; and so
Ernst Fraenkel, Griech. Denominativa, 160, and Wackernagel, Unters. zu
Homer, 125. This breaks down because for such a sense of a verb in -oüv

79
 tine 131 COMMENTARY
(‘to make one thing out of another’) there is certainly no analogy in earlier
Greek and probably none anywhere, as we can see from the full collectionsof
Fraenkel and Wackernagel in the books just cited. The verbs in -οῦν derived
from nouns are either ‘factitiva’ or ‘instrumentativa’. The purely 'factitive'
meaning (‘to make an army out of’, ‘to turn into an army’) assumed in
renderings such as 'the bit turned into an army' (Peile, quoted by Conington
on Cho. 549), ‘in exercitus speciem redactum" (Blaydes), is nonsense and must
be discarded. Nor does the function of the 'instrumentativa' come into the
picture; they mean 'to equip or furnish with the thing indicated by the
underlying word' (Fraenkel, Denominativa, 71), as ἀργυροῦν, θριγκοῦν, πυργοῦν,
and the like. No real parallel is offered by a usage like that of δεδοκωμένος,
' quoted by Wackernagel, op. cit. 126, following Mayser, from a papyrus of the
first century B.C. with the translation ‘aus Balken gefügt’ ; the more correct
rendering in L-S 'furnished with rafters' makes the normal instrumental
meaning perfectly clear. The most we can get from this, thercfore, is 'a bit
provided with an army', and however much scope we allow to the bold use
of metaphors in this ode, there is no sense at all in that. It seems that after
all scholars must be wrong in assuming that Aeschylus here broke fresh
ground and formed στρατωθέν directly from στρατός (Ernst Fraenkel, p. 77,
supposes ‘Augenblicksbildung’, a coinage ad hoc). It is far more likely that
in using this verb, which elsewhere occurs only in Homer, it was from the
Homeric usage that he started (so, too, Wackernagel, loc. cit., but it does not
affect his interpretation).! éorparówvro (once ἀμφεστρατόωντο), in the three
passages of the Iliad in which it occurs, means ‘they were to be found on
campaign (or in camp)'. We should therefore be well advised to interpret
our passage more or less as Hermann does: 'στρατοῦσθαι significat in castris
esse. ... Hic commorantes Aulide intelliguntur.' Since, however, the Homeric
usage makes it impossible to assume a sharp distinction between 'to be in
camp' and 'to be on the war-path', it may be the most prudent course
to accept a paraphrase like 'frenum quod consedit in castris’ (Nägelsbach)
anda translation like 'the mighty curb of Troy, the embattled host' (Sidgwick).
Taken in this sense, στρατωθέν shows that the object of κνεφάσηι is meant
really to be the Greek host. Although the army is described in a daring
ypidos? as στόμιον μέγα Τροίας, yet στρατωθέν is added: ‘participium non ad

! I may perhaps quote here a few sentences from the letter in which J. Wackernagel on
23. ix. 1936 replied to my interpretation of στρατωθέν : 'Mit Recht verwerfen Sie die Deutung,
die Wilamowitz für στρατωθέν gegeben hat und der ich mich in meinem Homerbuch ohne
die gebührende Prüfung der bei Verben auf -oóv vorhandenen Deutungsmóglichkeiten zu
Unrecht angeschlossen habe. . . . Evident scheint mir Ihre Auffassung dass στρατωθέν
auf dem homerischen ἐστρατόωντο beruht ; bei diesem haben wir aber die besondere Bedeu-
tung der Verba auf -oóv nicht vorauszusetzen, der Ausgang -owvro konnte zwar von Aisch.
auf ein Verbum auf -oóv bezogen werden, setzt aber eigentlich ein solches auf -ἂν voraus
[so Lobeck, PHMATIKON 185). (Vielleicht hat K. Meister, Die homer. Kunsisprache
S. 87, recht wenn er sagt “die Dichter haben unmittelbar von στρατός (u. dgl.) aus die
*zerdehnten' Formen abgeleitet".) . . . Dass aber Aisch. gerade ein Part. aor. pass. daraus
bildete, kann aus der Bevorzugung des Part. aor. pass. von Verben auf -οῦν erklärt werden,
die wir bei den Autoren des V. Jahrhunderts treffen (Sprachl. Unters. 122, wo etwa Hdt. 7.
139. 3, 9. 61. 2 [μουνωθέντες] beigefügt werden kénnen).’
2 When Ahrens (p. 279) describes ‘the confusion of three different metaphors in κνεφάσηι
... προτυπέν . . . στόμιον᾽ as ‘very hard to believe’, he shows little feeling for the style of
this part of the ode.
80
 COMMENTARY line 135

verbum στόμιον praecise, sed πρὸς τὸ σημαινόμενον, ad id quod subintelligitur,

referendum est’ (Schütz). At the same time, in a way characteristic of
Aeschylean γρῖφοι (cf. on 136, 495, 825, and elsewhere), the participle explains
the keyword of the riddle, or in Headlam’s phrase, ‘orparw6év is an epithet
"Jimiting" the metaphor’.
προτυπέν: the unfortunate rendering in L-S ‘forged beforehand’ derives
ultimately from Blomfield, who among other explanations doubtfully
suggested procusum. This was rejected by Hermann, and the true mean-
ing put forward: ‘prius percussum, i.e. ante belli clades immolatione Iphi-
geniae afflictum’. Wecklein’s comment is ‘struck beforehand as though by
lightning’, and we may accept his image ; but it is more important to remember
that the poet’s order of words indicates a close relationship between προτυπέν
and κνεφάσηι: the two verbs indicate only different manifestations of the
same catastrophe, which is brought to pass by the dya, the βασκανία. That
such an evil power τύπτει, 'strikes a man (or a city)’, corresponds to what we
are told in Zum. 155 ὄνειδος ἐξ ὀνειράτων μολὸν ἔτυψεν and to Eum. 509, where
a man is said to be ξυμφορᾶι τετυμμένος. And the act is properly called
προ-τύπτειν (Hermann is right here, too), because it may take place before the
capture which is the prerequisite of the application of the στόμιον Τροίας; a
very similar use of προ- occurs e.g. Cho. 647 προχαλκεύει δ᾽ ἶσα φασγανουργός.
There is clearly also a reference to the beginning of the speech: χρόνωι μὲν
ἀγρεῖ κτλ. (where we have seen that the wev-clause corresponds to the clause
beginning with οἷον) : ‘In the end this host will capture Troy, provided only
that not beforehand ....' The dya is made the subject, and at the same time
the result is described as coming to pass θεόθεν; the standpoint is similar to
that of τος where old age θεόθεν karanveicı, but in 131 the expression is less
strained, inasmuch as the ἄγα belongs directly to the deity.
135. οἴκτωι yap xrA.: the attempt to preserve the oixwı of the MSS proves
impossible. We cannot make both datives dependent on ἐπίφθονος, so that
they express the objects against which the φθόνος is directed, for there is no
question here of a σχῆμα ᾿Ιωνικόν or a καθ᾽ ὅλον καὶ κατὰ μέρος; Hermann’s
artificial explanation of πτανοῖσιν κυσὶ πατρός ‘quantum per aquilas cognosci
potest’ is its own refutation. Further, it is most unlikely that οἴκωι by itself
would suffice to denote the house of the Atridae; it is certainly wrong to
appeal to 157 οἴκοις βασιλείοις. Scaliger restored the true reading with his
οἴκτωι. The ground of the goddess’s reaction is emphatically put first and
marks the keynote of what follows: the sympathetic nature of the πότνια
θηρῶν ( 470) becomes for a space the centre of interest. It is surprising that
Wilamowitz of all scholars should have found the dative ot«rax difficult.
He used to draw our attention to the many instances of a verbal construction
preserved when the verb has been turned into an adjective. The dative in
οἴκτωι ἐπιφθονεῖ would belong to the same class as that in 5. El. 233 ἀλλ᾽ οὖν
εὐνοίαι γ᾽ αὐδῶ and the like.
ἐπίφθονος expresses her disapproval of what has taken place. Essentially
the meaning of φθόνος is unchanged ; the word here as elsewhere (see p. 349 1.)
expresses the attitude of a deity that says to man, as it were: ‘what lies
beyond a certain point you are not allowed to claim as your own, and if you try
ı Zielinski, Tragodumenon libri tres, 253 n. 1, fancied it to be possible that οἴκωι ἐπίφθονος
should mean ‘von Haus aus feindlich gesinnt’.
4872.2 G 8r
line 135 COMMENTARY
to do so I will make you smart for it.” This φθόνος appears before a transgres-
sion as a warning, after a transgression as a readiness to exact reprisals. For
ἐπίφθονος and its construction (xvoi) cf. Hdt. 9. 79. 1 κἀκείνοισι δὲ ἐπιφθονέομεν.
136. The eagles are here indicated by way of a ypidos, but in Aeschylus’
well-known fashion (for references see e.g. p. 9 and p. 81) the explanation
follows at the end of the sentence with αἰετῶν ; there is a close correspondence
with Prom. 1021 f. Aids δέ τοι πτηνὸς κύων, δαφοινὸς aierds. Probably eagles or
similar birds are referred to in fr. 282 N. ἐταμαῖς κυσὶν depodoírow (compared
by Blomfield). Other passages where κύνες means 'servants, attendants' are
collected by Wilamowitz on E. Her. 420.
137. αὐτότοκον: the scholion (in M) rightly σὺν αὐτῶι τῶι τόκωι. Stanley
compares Eum. 401 αὐτόπρεμνον, and also αὔτανδρον (Hesychius σὺν αὐτοῖς
τοῖς ἀνδράσιν), though our earliest evidence for that is Hellenistic. Cf. on 536
αὐτόχθονον.
λόχος in this sense is not found elsewhere except in A. Suppl. 677 ; in other
passages from Homer onwards the word means either ‘ambush’ or ‘company’
(of men, etc.). We cannot say whether Aeschylus found the word in earlier
poetry in the sense of ‘giving birth’ or gave it the new meaning himself (as
seems to be assumed by Schuursma 113), which in view of the obvious
etymology and the connexion with Adxios, Aoxeveodar, Aoxeia would be easy
enough. On the wrath of Artemis, which is the theme of these lines, the
scholiast comments: Aoyia γάρ ἐστιν ἡ θεός. There is good fifth-century
evidence for Artemis being called Aoxia ; when we now notice that the other
Aeschylean instance of λόχος meaning 'childbirth’, Suppl. 677 (again in a
choral ode), occurs in a prayer to Artemis, we may consider the possibility
that Aeschylus used λόχος in this sense to recall the Aoxia.
The note struck in oikrwı is echoed in 137 f.; every expression is chosen to
bring out as strongly as possible the calamity of the killing of the hare and its
unborn young. θυομένοισιν : the expression increases the effect of the unholy
deed, and at the same time prepares our minds for the sacrifice of Iphigeneia,
the ἑτέρα θυσία (151).
140. τόσον περ. Ahrens argues that we should retain the spelling of M
τόσσον. In support of that we may remember that in 5. 47. 184 the same
manuscript offers us τόσσον, which is proved right by the metre, and that in
Pers. 864 öooas, similarly demanded by the verse, is found in M and some
other copies (the majority have ὅσας). On the other sidewe have Pers. 121,
where M has the unmetrical τόσσοσδε, and Pers. 816, where some of our MSS
give τόσσος in a trimeter, which is certainly wrong.’ Cf. e.g. in M Cho. 370
πρόσσω, 450 dpeooiv, Eum. 135 ὀνείδεσσιν. τόσον περ εὔφρων in its relation to
what follows was rightly understood by Hermann: ‘quamvis tantopere favens
catulis ferarum, tamen bona vult portendi.’ This paraphrase also brings out,
quite rightly, that in the concluding clause of this long sentence, δεξιὰ μὲν
κτλ., the happy omen, qualified though it be, is yet far the more important,
cf. on 104 aiovov. We need therefore pay no attention to would-be corrections
and violent interpretations. In Homer τόσσον μέν often introduces the first
of two contrasted clauses, cf. Leaf on Z 378. For the attachment of τόσον to
an adjective, cf. ® 275 ἄλλος δ᾽ οὔ τίς μοι τόσον αἴτιος Οὐρανιώνων and the like.
1 ‘Scribes, unlike the imaginary scribe of scholars who do not read MSS, frequently alter
commoner forms to rarer’ (Housman, C.Q. xxii, 1928, 5).
82
 COMMENTARY line 141
Of the explanations and translations that attempt to separate τόσον περ from
εὔφρων, I will say nothing. Hermann also rightly recognized that the dative
δρόσοις κτλ. depends on εὔφρων ; we obscure this and upset the balance of the
whole clause if we put a comma after & καλά with Wilamowitz, which would
mean that both δρόσοις and ὀβρικάλοισι are governed by τερπνά. For the
order of words in eödpwv a καλὰ δρόσοις cf. H. Schöne, Hermes, lx, 1925, 160.
à καλά: ‘the article a [not in MV], although derived from a less trustworthy
source, does not look like a Byzantine emendation’ (Ahrens) ; in any case, we
cannot do without it. Wilamowitz in his apparatus has shown how much
more intelligible it makes the structure of the clause. The absence of the
article would be difficult here for another reason, in that we have not so much
a technical use of the cult-title as a free allusion to it. (For the former, cf.
for instance the dedication in IG v. 2. 398 [’Apjréuliros ᾿Ημέ]ρας Aovoov,
where there is certainly no room for the article, and Callim. Hymn. 3. 235 f.
τὸν (sc. νηόν) δ᾽ Evi Λούσοις 'Ημέρηι, also Bacchyl. 11 (10). 39.) For Καλλίστη
as a cult-title of Artemis there is plenty of evidence (besides the ordinary
reference-books, cf. Usener, Kl. Schr. iv. 11 f., and, for the elimination of an
apparent instance, Wilamowitz, Glaube d. Hell. i. 182 n. 3). It is alluded to
in E. Hipp. 66 f., whereas E. Iph. T. 36 "Aprejas . . . τοὔνομ᾽ ἧς καλὸν μόνον
perhaps refers rather to Καλή. The closest parallel to our passage is to be
found in Ar. Frogs 1359, when correctly restored : ἅμα δὲ Aikruvva παῖς à καλά. ἢ
141. δρόσοις: when Aristonicus in his explanation of the critical signs of
Aristarchus came to discuss the passage « 222 χωρὶς δ᾽ αὖθ᾽ ἔρσαι, he cited as a
parallel this passage (Etym. gen. B = Etym. M. 377. 37 épou). In doing so
he possibly followed the treatise of Aristophanes of Byzantium περὶ ὀνομασίας
ἡλικιῶν (cf. on 54 and 143), a long excerpt from which is preserved by Eusta-
thius in his commentary on the Odyssey passage, with a reference to Ag. 143.
From the same source comes Soph. fr. 725 N. (— 793 P.) ψακαλοῦχοι (from
ψάκαλον, itself derived from ψακάς) μητέρες alyés τε κτλ. Referring to the
passages from Homer and the two tragedians, Bechtel, AbA. Gölt. Ges., Phil.-
hist. KI.N.F. ii. 5, 1898, 11, observes that ‘the comparison of young creatures
with fresh drops of dew is peculiar to the Greek point of view’. It is possible
that Aeschylus has gleaned δρόσοι used in this sense from some earlier poet
(Wilamowitz, Interpr. 166 n. 2).
ἀέπτοις is slightly distorted in M (the same corruption is found elsewhere,
see below), but rendered certain by the scholion. It is a rare word, but found
in Aeschylus certainly in one other place, Suppl. 908, where Robortello's
conjecture has long held the field, but the MS reading (it does not matter
that the words are wrongly divided) was restored by Ahrens, p. 286, and
independently? by Wackernagel, Stud. It. N.s. v, 1927, 27. Another example
from Aeschylus is as good as certain, namely the fragment (213 N.) of the
ı So Kock, followed by Blaydes and by O. Schroeder, Ar. Cantica; the MS reading is
παῖς Ἄρτεμις καλά, which will not do, for two reasons. First, because in this exalted style
Ἄρτεμις so close on Δίκτυννα is hardly tolerable (compare the exactly similar foisting-in of
Ἄρτεμιν in E. Tro. 554 and of Ἄρτεμι to explain ‘Aypordpa Ar. Lys. 1262). Secondly, because
of the metre: the well-known clausula — v — « —, ὧι μέγας λιμήν, Pseudolum tuom, is out of
place among the simple cretics and trochees of this part of the ode, and seems moreover
not to be found in Aristophanes.
2 Wackernagel's suggestion that in hymn. Hom. Ap. 91 ἀέλπτοις should be changed to
dérrow had likewise been anticipated by Ahrens.
83
 line 141 COMMENTARY
Proteus preserved by Hesychius ἄελπτοι δεινοί, καὶ ἄαπτοι: Αἰσχύλος Πρωτεῖ.
There ἄεπτοι has been detected by Ahrens, loc. cit. But what does ἀέπτοις
mean here? No difficulty is raised by Suppl. 908, for ἄεπτ᾽ ἄναξ πάσχομεν
agrees perfectly with the meaning δεινός given by Hesychius in the passage
quoted, and also perhaps s.v. ἀελπάρεα' δεινά, where Ahrens conjectures that
there is some underlying corruption of derros. Ahrens has assumed the same
meaning in Ag. 141, and renders it ‘die argen Jungen wilder Löwen’ ; he thinks
the point is to show ‘how lovingly Artemis as Aoxia protects the young of
wild animals, seeing that she extends her protection even to those that work
mischief’. That seems to me out of key with the sentiment of the passage and
out of balance with what follows (see below) ; above all, however, it is ex-
tremely improbable that the poet who in an ode of this same play (717 ff.)
paints such a moving picture of the helpless, harmless, friendly nature of the
. lion's whelp (and therein takes up the φιλόμαστος of this passage) should here
have spoken of the lion’s ‘cruel young’. On the other hand, the explanation
of the scholion τοῖς ἕπεσθαι τοῖς γονεῦσι μὴ δυναμένοις agrees perfectly with the
facts. Arist. Hist. anim. 6. 31, p. 579°7 τίκτει δὲ καὶ d λέων πάνν μικρὰ οὕτως
ὥστε δίμηνα ὄντα μόλις βαδίζειν, Pliny, Nat. Hist. 8. 45 semenstres vix ingredi
posse nec nist bimenstres movert, ‘which is in entire agreement with modern
observations’ (Steier, RE xiii. 974). With this meaning, ἀέπτοις would exactly
correspond with the epithet of the following clause, φιλομάστοις, just as the
unusual word δρόσοις balances the unusual word ὀβρικάλοισι. The helplessness
of the young animals sets the loving care of the goddess in its true light. As
Artemis cares for the newborn so also she cares for the young as yet unborn ;
this idea is not explicit here, but the point has already been made in 137.
In an epigram by Philippus of Thessalonica (Anth. Pal. 9. 22; quoted by
Spanheim in his note on this passage), Artemis, in her quality of ὠδίνων
ἐπίσκοπος, refuses to accept the sacrifice of an animal that is big with young.
The motif is certainly taken over from earlier poetry, and can be used as
evidence for ancient belief.’
All the same, we must not minimize the objections that can be brought
against the interpretation of the scholion. It receives no support from Suppl.
908 (as regards fr. 213, Ahrens ingeniously conjectures that the φῶκαι of
Proteus might be called ἄεπτοι, but that is of course quite uncertain), and—
what is more serious—no support from etymology. That dezros could be
derived from ἕπεσθαι is categorically denied by Ahrens and Wackernagel ;
both support a derivation from ézew, Ahrens without reserve, Wackernagel
as a possibility. Nevertheless, we can at least regard it as possible that the
poet has used ἀέπτοις here in the sense ‘that cannot follow (their dam)’. We
should then be confronted with another of those transformations of meaning,
characteristic of Aeschylus, which are referred to in the note on 149 (ἐχενῆιδας)
and elsewhere in this commentary.”
1 On the other hand, the reference to [Xen.] Cyneg. 5. 14 (F. L. Lucas, C.R. xxxv, 1921,
28) is of no help towards the understanding of this passage. Neither god nor man can
expect of eagles the behaviour of φιλοκυνηγέται.
2 Wackernagel answered my letter of inquiry about ἀέπτοις on 18. 1. 1937 : ‘In der Deutung
des Wortes bin ich seitdem [i.e. his article i1 Stud. It. 1927] nicht viel weiter gekommen,
wenigstens nicht so, dass ich mich mit Sicherheit aussprechen kénnte. Doch schiene mir
jetzt die Schol.-Deutung in dem Sinne am ansprechendsten, dass A. ein der poetischen
Sprache längst angehöriges Wort in dem Orukelstile, der der ganzen Stelle eignet, will-

84
 COMMENTARY lines 1441.
μαλερῶν: here and elsewhere ‘devouring’, cf. Wilamowitz, Arist. u. Ath.
ii. 407 ; Bechtel, Lexilogus, 222.
143. ὀβρικάλοισι: this and related words were discussed with comprehen-
sive learning by Aristophanes of Byzantium in the section of his Ae£eıs
entitled Περὶ ὀνομασίας ἡλικιῶν ; fortunately, we possess copious extracts from
it on this very point (Arist. Byz. fragmenta coll. A. Nauck, pp. 111 ff.; for
ößpıa and ὀβρίκαλα cf. ibid., p. 125, and A. Fresenius, De “έξεων Aristoph.
... excerptis Byzant. 26). His quotation from the AucrvovAkot (fr. 48 N.) can
now be verified (Pap. Oxy. 2161, col. 2. 11): the new fragment gives the some-
what surprising form ὀβρίχοισι (for the suffix cf. on 54).
τερπνά: Hermann’s interpretation as = laeta, commended by Wilamowitz
(it goes back to Erfurdt on S. Ani. 419), is attractive, especially when we
think of ¢ 104 (Artemis) τερπομένη κάπροισι καὶ ὠκείηισ᾽ ἐλάφοισι. But the
point in our context is not so much that the goddess takes delight in the
wild creatures as that she is delightful and dear to them; so it is probably
best to retain the normal meaningof the word. Cf. also Critias B 6 Diels
(= fr. 4 Diehl) πρὸς τὴν τερπνοτάτην τε θεῶν θνητοῖς “Ὑγίειαν. Keyssner, Philol.
xcii, 1937, 277, points out that τερπνός in this sense is common in hymns of
the later period. In this construction with the dative, Aeschylus may have
taken the word (like εὔφρων) from the language of prayer. The decisive
argument for this view of τερπνά is the parallelism, worked out even in
details, of the two clauses, which are arranged in chiasmus εὔῤρων. . .
δρόσοις ἀέπτοις μαλερῶν λεόντων and πάντων ἀγρονόμων φιλομάστοις θηρῶν
ὀβρικάλοισι τερπνά. ἷ
Hermann objected to the metre of 143 ‘quod dactylicum esse debet’ and
therefore altered the text ; similarly in recent times, e.g., Platt, C.R. xi, 1897,
95 and Headlam. Their changes are unwarranted, see the metrical analysis.
144f. The words τούτων αἰτεῖ ξύμβολα κρᾶναι δεξιὰ μὲν κατάμομφα δὲ
φάσματα στρουθῶν are full of difficulties. The first point to be settled concerns
the main verb of this section. aire? has been attacked by many editors, and
is certainly difficult. The paraphrase in the scholia has acquiesced in some
sort of solution: τὰ σύμβολα αἰτεῖ με φάναι τούτων δεξιὰ μὲν διὰ τὴν νίκην,
ἐπίμομφα δὲ διὰ τὸν χόλον Ἀρτέμιδος, and αἰτεῖ is understood in the same way
by Wilamowitz, as is shown by his paraphrase: ἐπὶ τοσοῦτον εὐμενὴς ἡ
Ἄρτεμις κελεύει τὰ φάσματα, τοιαῦτα ὄντα, τῶν προκειμένων σύμβολα ποιεῖσθαι.
The assumption of this intensive meaning of αἰτεῖ = κελεύει is not objection-
able in itself, cf. Passow-Crönert, 182. 55 ff. But where αἰτεῖν is used in this
sense, it means not simply the ordering or commanding of some action no
matter what, but either a request for something, which is expressed in the
accusative, or, with an infinitive, a demand for the receiving of something,
e.g. E. Hec. 40 f. αἰτεῖ δ᾽ (Achilles) ἀδελφὴν τὴν ἐμὴν Πολυξένην τύμβωι φίλον
πρόσφαγμα καὶ γέρας λαβεῖν. As a rule, however, the infinitive construction
is found where we have, not the intensive meaning of the verb but the
very much commoner ‘ask, beseech, implore’, and so on; from Homer
onwards, y 173 ἠιτέομεν δὲ θεὸν φῆναι τέρας. Therefore the prevalent usage of
αἰτεῖν with an infinitive is not in favour of the meaning ascribed to the
kürlich umgedeutet und nach dem Muster der vielen aktiven Adjektive auf d—ros an
ἕπομαι angeschlossen hätte.’
τ This parallelism of structure has now been pointed out by W. Ferrari, La parodos, 371.
85
lines 144 f. COMMENTARY
expression αἰτεῖ ξύμβολα κρᾶναι by the scholiast and Wilamowitz. And there
is a further difficulty. αἰτεῖν with the infinitive, but with no accusative to
show from whom something is requested or demanded, seems never to occur
at all; at best the construction would be most unusual. Thus the normal
usage makes it difficult simply to understand the pronoun after αἰτεῖ, as the
scholiast does, taking his cue evidently from the sense that seems to be
demanded here: αἰτεῖ με φάναι. We have to admit, then, that the usual
meaning and the common construction of αἰτεῖν, though they do not abso-
lutely exclude the possibility that Aeschylus here wrote αἰτεῖ, make it rather
doubtful. But this is not all. The assumption that αἰτεῖ κρᾶναι could mean
αἰτεῖ με φάναι or something to the same effect is discredited by two fatal
objections: it is incompatible with the meaning of xpávaw and it does not
bring out the relation which we must expect between this clause and the
concessive expression τόσονπερ εὔφρων at the beginning of the sentence.
On the first of these two points: we can readily understand why Hermann,
who retained αἰτεῖ and also was strongly influenced by the interpretation
of the scholia, found himself driven to alter «páva« into κρῖναι, although this
conjecture is quite improbable both in itself and especially in the context
of Calchas' prophecy, of which in this very chorus it is said (249) τέχναι δὲ
Κάλχαντος οὐκ ἄκραντοι. But we may agree with Hermann that αἰτεῖ (ke)
κρᾶναι is unintelligible. The special sense of xpaívew discussed on 369 (‘to
ordain’, ‘to decree something that is valid, is to be fulfilled’) is out of the
question here, for κρᾶναι in that sense is a faculty of the god (cf. θεόκραντος,
πυθόκραντος, μοιρόκραντος) but not of the seer, whose province here can only
be to explain the portent—the portent which has indicated the divine will
and the divine decision that carries its own guarantee of fulfilment. On the
second point: the unmistakable concessive force of wep in τόσονπερ εὔφρων
«rA., to which Hermann himself gave its full weight (quamvis tantopere
favens catulis ferarum, tamen . . .’), excludes the possibility that in the princi-
pal clause in the sentence nothing was said except αἰτεῖ με φάναι OT κρῖναι,
or ‘heisst mich den Sinn des Zeichens enthüllen' (so Wilamowitz in his
translation; afterwards he took the passage differently). What we expect
as a contrast is something like this: 'although Artemis is so tender-hearted
towards the young of all beasts, although the eagles' slaughter of the unborn
hares is an abomination to her, yet in the main she does not set her face
against the fulfilment of this prodigy, which in spite of baneful details yet
promises a favourable outcome on the whole. It is true, she has in mind a
certain proportion of misfortune. . . .' We shall come back to this.
These difficulties have been taken into account by those critics who have
restored in place of αἰτεῖ another form of the same verb. So Blomfield with
hesitation : ‘an legendum αἴτει in imperativo ?', before him Schütz, suggesting
αἰτῶ, which has been followed by several editors, harsh though such a
parenthesis is, Verrall aörjı (as he should have written, not αἰτεῖ) ‘thou art
prayed'. But to say nothing of stylistic objections, these suggestions make
no sense; Gilbert's aîva, too, which is put by Sidgwick and Headlam in the
text (they translate ‘consent thou to fulfil' or something of the sort), pro-
duces the same wrong idea. It is impossible to assume that the seer in this
! Starting from the paraphrase of the scholion, Paley suggested that the original reading
was τούτων (u') αἰτεῖ, which was put in the text by Mazon.
86
 COMMENTARY lines 1441.
passage addresses himself immediately to Artemis with a prayer or a request.
For he continues: cov δὲ καλέω Παιᾶνα κτλ. The content, however, of this
prayer to Apollo is, we are surprised to find, a wish which not Apollo himself
but Artemis is to fulfil. This latter point is easily explained: Artemis has
been wronged by the eagles’ depredations, therefore she alone has a right to
exact atonement and it is only from her that complete or partial forgiveness
can come. And if the seer does not address himself to her directly, but uses
the mediation of her brother Apollo,! who is to intercede with his divine
sister,? that is explained by the position of Calchas or (to put it differently)
by the view which the poet holds of the seer’s competence as the servant of
one particular god. Aeschylus found Calchas in the traditional form of the
story a dependant of Apollo, from whom he had received his gift of prophecy
(Homer, A 72). Calchas is therefore not free to turn directly to any other
deity ; if he were to attempt such a thing, it would probably remain without
effect. The brief explanation of the scholion on 146 ‘ ὡς μάντις ’ is therefore
quite correct. And the form of the continuation mov δὲ καλέω Παιᾶνα shows,
no less clearly than its content, that a prayer to Artemis cannot precede.
Headlam noticed this difficulty and tried to avoid it ; he translates: ‘And the
power of the Healer also I invoke'. 'Also' — καί is just what we should be
bound to expect, if such a view is to be taken of the whole passage; but
'also' is not there, and cannot possibly be supplied out of a mere δέ. The sort
of connexion employed, when several gods are appealed to one after another,
can be seen without going farther afield (cf. on this point below on 509)
in the parodos of the Septem (97-165) and the Oedipus Rex (160 ff.): when
we pass on from one deity to the next, re or καί are the common particles.
Another group of commentators (Klausen, Peile, Paley, Nágelsbach,
Ferrari, La parodos, 371) retains αἰτεῖ as the indicative, but understands
Zeus as the object of it. The objection to this is that in the more immediate
context of our passage there is not the slightest indication that would suggest
a reference to Zeus.
Thus the defence of aire? appears to be a forlorn hope. The right reading
was restored long ago by Carl Lachmann, De choricis systematis tragicorum
Graecorum (1819), 56, with the simple change? aivet. Apparently no editor has
put this in the text. Only Lewis Campbell, who in addition to his knowledge
of the subtleties of tragic diction sometimes displays a feeling for the peculiar
ideas of Aeschylus which enables him to hold his own against generally
accepted views,* conjectured aivet without knowing of his predecessor (Am.
Journ. of Philol. i, 1880, 431, and in the appendix to his prose translation
of the Oresteia [1893], where the passage runs: “The Lovely Goddess, although
so kind to the tender cubs . . . consents to ratify the . . . fulfilment of the
sign. The omens are favourable, but not unmixed with bane’). aivet here
satisfies all requirements. This verb often expresses a 'iudicii voluntatisque
adsignificatio', cf. on 98. Here, as often elsewhere, it stands on the verge
! Hermann says ‘Non Apollinem ad avertendam Dianae iram invocat Calchas, sed
Paeanem, qui quisquis est, non est frater Dianae'. I mention his error only because others
have followed his lead.
? Apollo's intervention with Artemis on behalf of a mortal against whom the goddess
has a spite is also found e.g. in the story of Admetus and Alcestis ([Apollod.] Bibl. 1. 106).
3 In E. Phoen. 924 Elmsley has made the certain restoration of αἴνει for acri.
+ Cf. on 411 and on 1116.
87
lines 144 f. COMMENTARY
between 'to say yes to something, to agree with something' or the like and
‘to promise something, to pledge oneself to it’ and so on (thc latter, e.g.,
E. Alc. 12 f. ἤινεσαν δέ μοι θεαὶ Ἄδμητον dibqv τὸν παραυτίκ᾽ ἐκφυγεῖν, with
the aorist infinitive as in Ag. 144 κρᾶναι). It is quite natural that Artemis
should appear here as the subject of «paivew. This does not mean that it
is for Artemis herself to carry to completion the destiny which the sign
foretells ; the hope is that she may allow it to come to pass, that although
she has been wronged by what happened, she may not stand in the way of
the fulfilment of the ostentum which is in the main, if not wholly, favourable.
ξύμβολα. That σύμβολα or σύμβολοι (both are found with no difference of
meaning) is used by Aeschylus in a technical sense can be seen from Prom.
487, where the Titan, after speaking of his discovery of the interpretation of
dreams, and of the help that he has given humanity towards the understand-
ing of κληδόνες (cf. below on 1653), continues: (κληδόνας... ἐγνώρισ᾽ αὐτοῖς)
ἐνοδίους τε συμβόλους. What may be another example appears in the frag-
ment of the Glaukos Potnieus, Pap. Soc. It. 1210. 4 ]ev κελεύθωι ξυμβολῖ, cf.
Körte, Arch. f. Papyrusforsch. xiii, 1938, 97. For the meaning cf. Ar. Birds
721 ξύμβολον ὄρνιν (scil. καλεῖτε) with the scholion (cf. also the commentators
on Ar. Frogs 196) and especially Schol. Pind. Ol. 12. 10 with the line of
Archilochus (fr. 46 D.). For the evidence in later writers sce Lobeck, Aglao-
phamus, 828°. (The monograph of Walter Miri, ΣΎΜΒΟΛΟΝ, Wort- und
sachgeschichtliche Studie, Bern 1931, is only known to me from Snell's review
in Gnomon, viii, 1932, 388 f.) It is clear that σύμβολα in the first instance are
ἐνόδια σύμβολα. In this specific sense the word denotes here the omen dealing
with the κέλευθος (126), the expedition of the fleet and army. The addition
of τούτων,to which many have objected, offers no difficulty. Wilamowitz'
rightly explains: τῶν προκειμένων σύμβολα. ταῦτα by itself, without reference
to anything expressly named just before or just afterwards, but yet with
clear demonstrative force, indicates something ‘that lies before us’, ‘that is
happening here and now (or has just happened)’ ; this is in itself natural, and
actually not unusual in Aeschylus. Cf. e.g. Prom. 265 εὖ δὲ ταῦτα πάντ᾽
ἠπιστάμην, which is not to be referred to the γνώμη that immediately pre-
cedes; the paraphrase of the late scholia rightly explains: à viv πάσχω.
Similarly, e.g., Suppl. 410 ὅπως dvara ταῦτα... . ἐκτελευτήσει καλῶς, Eum. 199
αὐτὸς σὺ τούτων οὐ μεταίτιος πέληι.
To ξύμβολα the clause δεξιὰ μὲν κατάμομφα δὲ φάσματα is in apposition.
This explanation is no mere piece οὗ ornament, but impresses on us once again
the contrasted elements of the portent and so its peculiar force, and at the
same time binds into a whole the different sections of the speech of Calchas.
After beginning with the positive, favourable elements (126 ff), the seer
proceeds (131 ff.) to the negative and baneful. At the beginning of the next
section, marked off by metre as well as sense, the favourable view is again
prominent (140-4) ; then follows a brief contrast of favourable and unfavour-
able (145), as transition to the apprehension of threatening evil which fills
the last sentences of the prophet’s speech. When the speech is ended, the
Chorus (156 f.) once again bring out the two contrasted aspects; τοιάδε...
1 While accepting this detail, I find it impossible to agree with the highly artificial
interpretation of 144 f. that underlies his paraphrase, or with his transposition of δεξιὰ μὲν
κατάμομφα δὲ φάσματα, which significantly concludes the whole period.
88
 COMMENTARY line 146
ξὺν μεγάλοις ἀγαθοῖς... μόρσιμα is just like δεξιὰ μὲν κατάμομφα δέ. The
expression δεξιά refers back to 116 χερὸς ἐκ δοριπάλτου. The word κατάμομφα,
like other derivatives of μεμῴ-, has a religious connotation, cf. Latte, Archiv
f. Religionsw. xx, 1921, 264 n. 3.
Finally, we have the meaningless and unmetrical στρουθῶν (or τῶν arpov-
θῶν). It was recognized by Porson, and should never have been called in
question since, that this is an interpolation based on the prodigy of the
sparrows in B 311 ff. The addition was meant, one may suppose, to explain
τούτων misunderstood. It makes little difference whether τῶν στρουθῶν
(FTr) represents the earlier form of the interpolation, as Wilamowitz assumes,
or whether στρουθῶν was originally written in the margin and the article
added later in order to adapt the word, clumsily enough, to the dactylic
metre, which I think more probable. The view put forward by Casaubon
(‘avium i.e. aquilarum’) and Stanley, and supported by Schiitz and many
others, that στρουθοί can mean birds of any kind, gains nothing in correctness
from the confidence with which it has been repeated down to our own time.
Hermann has shown what lies behind the alleged testimony (quoted by
Schütz) of Eustathius (on B 311, p. 228. 38, cf. Aelii Dionysii . .. fragmenta,
ed. E. Schwabe, pp. 208. rff). There is no evidence for the theory that
στρουθός with nothing to define it could denote any bird at will; how, then,
could it be used of an eagle of all birds? στρουθός by itself always means
either the sparrow or the ostrich.'
Here, as elsewhere (cf. on 7o δακρύων), it is hard to decide whether the gloss
has simply taken its place alongside the true text or whether it has expelled
some of the original words. I must leave both possibilities open. So far as
we can see, the sense is complete, and the position of the qualifying clause
δεξιὰ... φάσματα at the end of this section is entirely right as was shown
above. Unfortunately the metre gives us no real help towards a decision,
for there is no correspondence, and the fact that this dactylo-iambic canticum
has no parallel in the extant plays of Aeschylus makes it impossible to
determine what variations in the conclusion of a dactylic period should be
regarded as probable in such a case as this. We should be able to pass
judgement with much greater confidence if we enjoyed even a part only of the
supply of similar or closely related lyrics which was available to Aristophanes
(Frogs 1264 ff.). As things are, we may regard it as at least possible that
W. Ferrari (La parodos, 373) is right, when on the basis of materials collected
by me, Rhein. Mus. lxxii, 1918, 177 f.,* he assumes that in Ag. 145 at the end
of ἃ dactylic period the cretic clausula φάσματα is permissible.
146. On the shortening of the first syllable of Παιᾶνα cf. Wilamowitz on

™ At the bottom of the famous Herakles relief in the Villa Albani (Otto Jahn, Griech.
Bilderchroniken, platc V ; cf. W. Helbig, Führer, ii, 3rd. ed., no. 1850, and, for the inscriptions,
IG xiv. 1293) there is a catalogue (in hexameters) of the hero's labours. L. 5 runs: πέμπτον
δὲ στρουθοὺς Στυμφαλίδος ἤλασε λίμνης. What induced the versifier to call the Stymphalian
birds στρουθοί, we do not know; possibly he pictured them to himself as ostriches (Gruppe,
RE, Suppl. iii. 1041, 36).
2 As regards Ar. Lys. 1287 ff., we should perhaps prefer the simpler analysis which was
later given by Wilamowitz in his commentary, though the traditional division of the lines
which I followed (as does O. Schroeder, Aristophanis Cantica, 56; it is found in the Ravennas)
has the advantage that the metrical and syntactical cola coincide. But in either case it is
certain that we have there a cretic ending to the dactylic period (1290) ἣν ἐπόησε θεὰ Κύπρις.

89
line 146 COMMENTARY
E. Her. 820, and add Bacchyl. 17 (16). 129 παζάνιξαν. On the reason why an
appeal is made to Apollo to intercede with his sister see p. 87.
147. In the clause beginning with μή τινας, the baneful element, which in
κατάμομφα δέ occupied only a subordinate place, comes to the fore with
greater intensity and in greater detail. This negative clause introduced by
μή stands, so far as the sense is concerned, in the same relation to the positive
clause a£vet . . . κρᾶναι as the py-clause 131 ff. οἷον μή τις dya κτλ. to the pre-
ceding positive sentence χρόνωι μὲν aypei κτλ. The ebb and flow of hope and
fear seen in the words of the seer dominate in the same way the later odes of
the Chorus. All through, the certainty that an evil deed must bring upon
itself inevitable expiation sets bounds to the expectation of ahappy outcome.
τινας (and similarly 151 rıva) calls attention to the purely suggestive,
foreboding character of the prophecy.
149. ἐχενῆιδας. It is most unlikely that the ἐἰχθύδιόν τι τῶν πετραίων, 6
καλοῦσί τινες ἐχενηΐδα (Arist. Hist. anim. 2. 14, p. 50518) should have first
acquired its popular name (among the Romans remora) between the time
of Aeschylus and Aristotle. (On the various kinds of echeneis cf. A. W. Mair
in the note on his translation of Oppian’s Halieutica 1. 212 in the Loeb
Library.) Nor is there any indication that in the living language of Athens
ἐχενηΐς meant anything other than this particular small fish. If so, then by
a bold and arbitrary ‘re-etymologizing’' Aeschylus has so transformed the
name of the tiny creature that it now vividly expresses the working of the
mighty powers of nature. Unless I am much mistaken, this particular
experiment encouraged Sophocles to hazard, in the same sphere of ideas,
something closely parallel: fr. 694 N. (= 761 P., quoted by Athen. 3. 99 d)
τὴν ἄγκυραν ἰσχάδα κέκληκεν διὰ τὸ κατέχειν τὴν ναῦν" ναῦται δ᾽ ἐμηρύσαντο
νηὸς ἰσχάδα᾽, ‘where an Athenian could only think of a dried fig’ (Wilamowitz,
Hellenist. Dichtung, ii. 150). A further parallel is to be found in Sophocles’
use of εὐνοῦχος (fr. 721 N. = 789 P.), and perhaps of ἀλέκτωρ (fr. 767 N. = 851
P.) “οὑμὸς δ᾽ ἀλέκτωρ αὐτὸν ἦγε πρὸς μύλην᾽, if there ἀλέκτωρ, as is generally
assumed, really means ‘husband’ ;2 elsewhere the meaning is always ‘cock’,
cf. Ernst Fraenkel’s very full discussion, Nomina agentis, i. 154 n. 8. The
same artificial device is kept up by Plato, when in T'heaet. 149 b he, in the
words of Pollux (3. 15), τὴν παρθένον (Artemis), ὡς οὔπω λοχευθεῖσαν, ἄλοχον
καλεῖ. Hesiod had already allowed himself a similar piece of word-play, Erga
533 τότε δὴ τρίποδι βροτοὶ too, on which Wilamowitz comments: 'rpimovs is a
γρῖφος, jokingly used, for the first thing the hearer thinks of is a tripod’.
ἀπλοίας: cf. on 188.
1 éxevnis should have been discussed by Schuursma, De . . . abusione, especially in his
third chapter (‘Abusiones secundum etymologiam’). In all periods poets have the right to
restore to a word its ‘original’ meaning, which in daily usage it has entirely or almost
entirely lost. Horace, Odes ı. 36. 20 lascivis hederis ambitiosior provides a good example.
The peculiar use in Horace, Odes 4. 4. 65 of evenit, to which several critics have objected,
belongs to the same category; Baiter ad loc. rightly says ‘Horatius saepius ad propriam
vocabulorum vim redire ausus est’.
2 The context is quite uncertain. Wilamowitz’s interpretation, ‘Hieron und Pindaros’,
Berl. Sitzgsber. 1901, 1296 n. 3, which Pearson might have mentioned, runs: ‘ οὑμός . . .
μύλην is said by Admetus of Apollo, his slave, who is called to the mill by the crowing of
the cock.’ This has the great advantage that we need not reject Admetus as the speaker,
who is attested by Plutarch; but is this view of ἦγε probable? On the supposed ἀλέκτωρ =
‘husband’ in Bacchyl. 4. 8 see Snell, ad loc. [Cf. the Addenda.]
go
 COMMENTARY line 153

150. σπευδομένα: the same unusual middle also Eum. 360 (dactyls as here).
151. θυσίαν ἑτέραν: since Stanley much misdirected learning and ingenuity
has been spent on the explanation of ἑτέραν; Ahrens even dragged in the
epulae Thyesteae. The right answer was given long ago by Pauw: ‘opponitur
praecedenti (v. 137) αὐτότοκον . . . πτάκα θνομένοισιν᾽. The choice of words
in the earlier passage makes it certain that Aeschylus intended this reference
back. Wilamowitz's deletion of ἄνομον is not justified; as Kranz rightly
says, Hermes, liv, 1919, 307 n. 2, τινα after ἄνομον seems to show that the two
attributes ἄνομος and ἄδαιτος are valid only for the θυσία ἑτέρα: 'a second
sacrifice, a lawless one, of which it is not good to eat', so e.g. also Headlam's
prose translation. Cf. Hdt. 1. 162. 1 τὸν (i.e. Aprayov) . . . Ἀστυάγης ἀνόμωι
τραπέζηι ἔδαισε.
153. νεικέων τέκτονα. τέκτων, at this stage of Greek thought and language, is
never used of a thing (as for the hand, Ag. 1405 f., it is readily treated even
in modern speech as equivalent to a person). Here as elsewhere we must
not weaken the word by substituting a mere activity or quality for the living
agent. We must not introduce anything ‘abstract’ into this very concrete
‘Zimmerer des Zwistes’, this ‘architect of discord’ or ‘master-craftsman of
strife’. As the seer contemplates, with eyes that see clearly into the future,
the baneful consequences of the sacrifice of Iphigeneia, the θυσία itself, which
first appears as a certain indication of ruin, becomes the expert contriver
of dissensions. The personification is then carried forward in σύμφυτον and
δεισήνορα. At bottom, however, this τέκτων νεικέων that issues from the evil
deed is hardly to be distinguished from the Mövıs, set out in the same way
with all the attributes of a person, that rules the following explanatory
clause. We can, if we will, make this ‘craftsman of strife’ more vivid with
the help of an Alastor, an Ate, an Erinys, or some similar being ; but whether
or no, its ‘daemonic’ nature is clear.
With great art the epithets of θυσίαν are so chosen that the Athenian
audience, who know the story, already at this stage understand the sacrifice
of Iphigeneia, while Agamemnon and the army, though they are bound to
infer that some sinister piece of ritual is afoot (ἄνομον for this purpose is
clearly indispensable), are quite unable to tell of what kind it will be. The
result of this is twofold: for the spectator the way is prepared for the hymn
to Zeus (160 ff.) by the suggestion of that frightful manifestation of the
god’s power, the father that must slay his own daughter; and on the other
side, a climax of the narrative is held in reserve for the moment (198) when
Calchas actually reveals the demands of Artemis without disguise.
σύμφυτον has often been misunderstood (cf. Paley, where earlier interpre-
tations are passed in review). It is a mere shift to conceal their embarrass-
ment when editors (Heath, Blomfield, and others) call to their aid the long-
suffering figure of hypallage or enallage, and understand τέκτονα συμφύτων
νεικέων; the ‘parallels’ adduced for this are inadequate, and the peculiar
force of τέκτονα (see above) is on this view completely ignored. Wilamowitz
evidently understands ‘the sacrifice that in accordance with its nature
worketh strife’, for he paraphrases (Interpr. 166 τι. 2) κατὰ τὴν ἑαυτῆς φύσιν
yap νείκη τεκταίνει καὶ συγκροτεῖ, and so others, e.g. Headlam, ‘worker of
feud in the very flesh’. A similar view, it is true, is suggested by certain
renderings in the current lexica (‘eingeboren, eigen, natürlich’ Passow,
OI
line 153 COMMENTARY

‘natural’ L-S). But it seems that, at least in earlier Greek, the relation of
a thing described as σύμφυτον is always brought out quite clearly; in other
words either we are told directly or we can infer without ambiguity from the
context σὺν ὅτωι δὴ τὸ σύμφυτον συμπέφυκε. Thus Pind. Isthm. 3. 14 ἀνδρῶν
δ᾽ ἀρετὰν σύμφυτον (scil. τοῖς ἀνδράσιν) οὐ κατελέγχει, A. Ag. 107 σύμφυτος
αἰών (see ad loc.), E. Andr. 954 (the Chorus to Hermione) ἄγαν ἐφῆκας γλῶσσαν
eis τὸ σύμφυτον (scil. got, schol. eis τὸ γυναικεῖον γένος), Lysias 10. 28 οὕτω
σύμφυτος αὐτοῖς ἡ δειλία. In Plato and Aristotle, too, so far as I can see, the
relation intended almost always comes out quite clearly. Thus the usage of
the adjective does not favour Wilamowitz’s interpretation. It is, on the
other hand, in keeping with general usage when other editors explain like
Enger, Schneidewin, and Sidgwick ‘genti innatum’, ‘a seed of strife clinging
to the race’. Against this interpretation it should not be objected that the
relation of σύμφυτον remains in doubt. In the clause immediately following,
of which the subject Mis as we have seen takes up νεικέων τέκτονα, all the
weight is laid on the μίμνειν of this Menis, of whom we are further told that
she holds sway in the house (and in the race). Where Menis dwells, there
dwells also the νεικέων τέκτων, an inmate and a native, grown together with
the house of the Atridae, σύμφυτος. Later the spirits of the Curse on this
same house are called σύγγονοι "Epivdes (1190), immediately after we have
been told (1189) that the swarm of them ἐν δόμοις μένει, as the Menis μίμνει.
δεισήνορα: the word occurs as an adjective only here. New formations
following the prototype of Homeric words like ῥηξήνωρ and so on were
naturally open to every poet; in the same way Stesichorus, for instance,
coined λιπεσάνορας. A list of related formations is given by Wackernagel,
Dehnungsgesetz, 40 f. We do, however, find the name Δεισήνωρ in a Homeric
list of names P 217 (cf. on this O. Hoffmann, Glotia, xxviii, 1939, 49), and its
use as an epithet may very well have derived from that, as happened with
ἀντήνωρ and other words (cf. on 443). The ἡ of these compounds, as compared
with πειθάνωρ (Ag. 1639), ποιμάνωρ (Pers. 241), etc., is probably to be accounted
for by the borrowing from Homer, cf. H. W. Smyth, Harvard Studies in
Class. Philol. vii, 1896, 146. The meaning of od δεισήνορα is correctly given
by the paraphrase in the scholia as οὐ φοβουμένην τὸν ἄνδρα. .
οὐ δεισήνορα goes not with θυσίαν, but with the phrase which is in apposi-
tion to it, viz. νεικέων τέκτονα (so rightly e.g. Nagelsbach), in which under the
form of a noun of the agent a new element is introduced and a further stage
is indicated in the inevitablé process of events. The second act of bloodshed,
the murder of Iphigeneia, builds up strife, and this is brought about without
any fear of the man, i.e. the husband, or any sense of respect for him (we find
the pair αἰδὼς καὶ δέος in the Iliad, and in the Cypria stands the saying often
modified later iva yap δέος, ἔνθα καὶ αἰδώς). The phrase, which is equivalent
to such a sentence as ἡ θυσία νείκη τεκταίνεται οὐ φοβουμένη τὸν ἄνδρα, points
quite unambiguously, although in veiled language, to the development that
culminates in the murder of Agamemnon.
154 f. μίμνει... τεκνόποινος. Stanley’s idea that by the μῆνις we are really
to understand Clytemnestra has not yet lost its influence, and many com-
mentators have found something especially elegant in the fact that most of .
the epithets in this clause seem equally applicable to her (Verrall actually
renders μῆνις ‘the wrath of the wife’). This might seem particularly tempting
92
 COMMENTARY line 155

in view of οἰκονόμος. But in this passage the poet has no thought of such play
with a double meaning; the idea of his μῆνις, or better Máws (cf. on 14), is
much more simple and grand. In order fully to understand οἰκονόμος here,
we need not resort to the housewife; we have only to remember A. Suppl.
415, where a brother of ῆνις, Alastor, is described as βαρὺς ξύνοικος (cf.
on 1641 f.). The fundamental idea is the same in both places, but what
stands out here is the bold and vivid way in which Μῆνις is represented as
mistress of the house. It is by no means true that all the epithets here are
equally applicable to the housewife; for to think of her in connexion with
παλίνορτος would be absurd.
μίμνει: she remains for all time, as we are told of the sin of Laios (and its
consequences), Sept. 744 αἰῶνα δ᾽ ἐς τρίτον μένειν. No supplication, no repen-
tance causes the Menis to relent ; it would be quite fruitless to try and induce
her, perhaps by means of the kind of prayer called ἀποπομπή (cf. on 1571-3),
to leave the house and pass to some neighbour or some enemy: xai μὴν
πεπωκώς γ᾽, ὡς θρασύνεσθαι πλέον, βρότειον αἷμα κῶμος ἐν δόμοις μένει δύσπεμ-
πτος ἔξω, συγγόνων ᾿Ερινύων. These lines (1188 ff.) form the best commentary
on 154 f., for Maus and ᾿Ερινύες belong closely together. If we understand
μίμνει aright, we escape the temptation to foist in the idea ‘while the husband
goes to the war, the wife stays behind watching faithfully over hearth and
home . .. here what stays behind is the desire for revenge’ (Wecklein, and
similarly others). Such an interpretation shows again the influence of the
idea we have already rejected, according to which the clause, and with it
the μῆνις, refers also to Clytemnestra. This entirely confuses the issue.
There is no question here of staying behind at home while the army is over-
seas. The point is that out of Agamemnon’s offence arises an evil thing, long
lasting, that makes its home in his house and ever renews its power, whose
memory never weakens. waAivopros and μνάμων support the idea of μίμνει.
On uiuvew in a related context cf. on 1563.
παλίνορτος is found only here. The scholia rightly explain it: ἡ ἐξ ὑστέρου
ὁρμωμένη. For a time things go well. The guilty—or their children—draw
breath, they fancy it is all over, the μῆνις is finished. Not at all: πάλιν
ὄρνυται. For the verbal element, with which editors have tampered, cf.
θέορτος found in Aeschylus as well as in Pindar, veopros used several times
by Sophocles, and πέδορτος in S. Ichn. (fr. 314 P.) 212.
155. οἰκονόμος: the word is always used as a substantive, including the one
passage earlier than Aeschylus where it appears, Phocylides, fr. 3. 7 oëko-
νόμος τ᾽ ἀγαθή. It is thus a priori unlikely that it should stand here on the
same footing as φοβερά, παλίνορτος, and the rest, as one of the many adjectival
epithets of Μῆνις (Sidgwick is wrong in saying 'here are six adjectives and
one substantive’). The relation of the parts of the sentence to one another
must be made clear in modern texts by punctuation ; it will not do to refrain
from commas altogether, which many editors have here regarded as the
safest course. Others enclose in commas οἰκονόμος δολία. This might be
paraphrased οἰκονόμος δολία οὖσα, which in itself is not bad but disrupts the
monumental grandeur of the sentence by straddling the series φοβερὰ παλίν-
opros μνάμων Μῆνις τεκνόποινος on either side of a parenthetical clause.
Similarly Headlam's attempt to punctuate before δολία does not prove satis-
factory (in his prose translation: 'a terrible recoiling keeper of the house,
93
line 155 COMMENTARY

a treacherous unforgetting child-avenging Wrath’); δολία beside μνάμων

Μῆνις τεκνόποινος would be feeble. If, on the other hand, we connect δολία
closely with οἰκονόμος, we see that the desirable state of an οἰκονόμος ἀγαθή
is here reversed (cf. on 1641 f.): a daemonic housekeeper has stealthily in-
truded to exercise her functions cum dolo malo. It is right therefore to punc-
tuate after δολία, and there only, as was done long ago by Stanley, Butler,
and Blomfield and among recent editors, e.g., by Verrall and Wilamowitz.
The preparatory description (foBepa . . . δολία), disguising rather than re-
vealing the nature of the thing, is an instance of a common Aeschylean
device (cf. e.g. Sept. 727-30): ‘so legt der Dichter ein Rätsel künstlich mit
Worten verschränkt oft der Versammlung ins Ohr . . . aber noch fehlet das
Wort, das die Bedeutung verwahrt.' The solution of the riddle follows in the
last three words of the sentence, which narrow down the idea to the relation
between crime and punishment, while the oixovópos-image is couched in
more general terms. μνάμων and rekvómowos are not co-ordinate, for τεκνό-
mowos specifies this particular Μῆνις.
pvápov: it may seem tempting to see in this a reference to an institution
in the sphere of law, namely the μνάμων, whose 'duty it was to guarantee the
authenticity of the proceedings and of the verdict, originally of course by
his powers of memory' (K. Latte, Heiliges Recht, 1920, 3o n. 4; these valuable
explanations are ignored in E. Weiss's article 'Mnemones', RE xv. 2261 ff.).
The assumption of such a reference seems to be favoured by the well-known
line Prom. 516 (more fully discussed below on 1535 f): Μοῖραι τρίμορφοι
μνήμονές τ᾽ "Epwées (which probably inspired S. Aj. 1390 μνήμων 7’ ᾿Ερινύς,
for τελεσφόρος, which follows there, is also found in the same passage of the
Prometheus, 511). But the fuller form of the expression in Et. 383 κακῶν
τε μνήμονες Zeuvai warns us against regarding the epithet as an echo of the
judicial office, and caution is further recommended by Latte's observation
that the uvápoves ‘are only found in the Dorian sphere of influence and are
entirely absent in pure Ionian states'. The use of μνήμων in a general, non-
technical sense is Homeric, and is found in Prom. 789. That Aeschylus
should emphasize this quality in the Erinyes and the Menis is natural
enough. The τίσις of Zeus may be slow to act, but in the long run no crime
lies hidden from it, and it does not forget to punish the guilty (Solon fr. r.
25 ff. D.), οὐδὲ μήποτε λάθα κατακοιμάσηι, as Sophocles (Oed. R. 870) says of
the eternal ordinances. Cf. above on 58.
The last words of Calchas are extremely sinister.
156 f. τοιάδε... μόρσιμα: on the form of this very ancient type of separation
(ὑπέρβατον, ‘Sperrung’), where an adverb or adverbial group plus verb
stands between the separated elements of a coherent phrase cf. my book
Iktus und Akzent, 166; Joh. de Vries, Unters. über die Sperrung . . . in der
Sprache der ait. Redner (Diss. Freiburg i. Br. 1938), 15 f.;" for the separation
of τοιόσδε from its noun see Zktus und Akzent, 321 n. 3. That here special
emphasis lies on the separated word-group τοιάδε... μόρσιμα is clear enough.
Cf. 892 1.
1 A few instances of this particular type may be quoted: Lysias 25. 17 μεγίστην ἡγοῦμαι
περὶ ἐμαυτοῦ τῆι δημοκρατίαι πίστιν δεδωκέναι, Dem. 20. 141 εἶτα μεγίστας δίδοτ᾽ ἐκ παντὸς
τοῦ χρόνου δωρειάς, 18. 210 καίτοι πολλοὶ παρ᾽ ὑμῖν, ἄνδρες ᾿Αθηναῖοι, γεγόνασι ῥήτορες ἔνδοξοι
καὶ μεγάλοι πρὸ ἐμοῦ, Aeschines 2. 38 πλείστην δὲ εἰκότως ἐποιήσατο διατριβήν.

94
 COMMENTARY lines 158 f.
Κάλχας: the name is given here for the first time; before the beginning
of his speech he was introduced (122) as κεδνὸς στρατόμαντις. Similarly in the
second half of the ode we have 186 and 201 μάντις, at the end 248 Κάλχαντος.
In a similar way Pindar (Ol. 6. 33, 36) and Bacchylides (s. 7o, 77) in their
narrative begin with a patronymic that lightly conceals their man, and only
later give us his proper name.
ἀπέκλαγξεν ‘like ἔκλαγξεν (201) expresses the loud and excited tone of
voice which marked the spirit and exaltation of the μάντις᾽ ; so rightly
Headlam, and similarly Plüss before him.' Cf. Pindar, Paean 8. 20 Schr.
ἔκλαγξε (the object is a prophecy). To the force of the preposition ἀπο-
most commentators seem not to have given a second thought ; Plüss is an
exception, but his tentatively expressed conjecture ((ἀπο- as in ἀποδιδόναι,
because he gives back what he has received
?’) has little probability. Perhaps
it is justifiable to translate simply ‘ring or shout forth’ (L-S) ; for parallels
we should refer to hymn. Hom. in Merc. 280 μάκρ᾽ ἀποσυρίζων and 5. Ant.
Io21 οὐδ᾽ ὄρνις εὐσήμους ἀπορροιβδεῖ Bods. It is also possible that as a result
of the addition of οἴκοις βασιλείοις the word here acquires the sense of making
known ; then further analogies for the preposition are offered by ἀποδείκνυμι,
ἀποφαίνω. For all that, we may perhaps wonder whether Aeschylus, who in
all probability coined the compound for this passage, may not have intended
an intensification of the indication of origin given in ἀπ᾽ ὀρνίθων. If so, the
fresh coinage may be due to the same linguistic tendency which in Hdt.
4. 22. 2 produced the phrase ἐπεὰν δὲ (the huntsman) ἀπίδηι τὸ θηρίον ἀπὸ τοῦ
δενδρέου. To intensify the expression of direction as Aeschylus seems to do
would perhaps be somewhat bolder (though by no means too bold for the
language of this ode), in so far as what is denoted here is no locality but the
source and origin of the prophecy; it is from the eagles that the prophet
derives his knowledge of the future, and hence his power to proclaim it.
μόρσιμα: the expression is chosen with great caution; an unambiguous
indication of evil is avoided. μόρσιμον (the lexica are supplemented by C.
Arbenz, Adjektive auf -uos, Diss. Zürich 1933, 32) both in Homer and other
authors and also in Aeschylus is sometimes used in a neutral sense, i.e. not
in malam partem, e.g. fr. 13 σοὶ μὲν γαμεῖσθαι μόρσιμον, γαμεῖν δ᾽ ἐμοί, similarly
Eum. 217. But phrases of common occurrence such as μόρσιμον ἦμαρ and the
like do at least bring the sinister meaning very near the surface, and the
conclusion of Calchas' speech points in the same direction. Without a strong
admixture of the idea of misfortune there would be no ground for the addition
of ξὺν μεγάλοις ἀγαθοῖς, and the sentence rois δ᾽ ὁμόφωνον κτλ. would lose
its point altogether. But it is important that the negative content of the
double omen be not over-emphasized. Taking it as a whole, and from the
standpoint in particular of the men who experienced the outbreak of the war,
the decisive fact is that Troy will fall (126 χρόνωι μὲν ἀγρεῖ «rA.); to this
extent the αἴσιον of 104 is not an unduly favourable description. But in the
eyes of the citizens of Argos, and similarly of the spectators who stand on
the threshold of the trilogy, the evil that will live and be active afterwards
has the greater weight and is predominant in the prophet's speech.
158f. The variation in the treatment of the refrain from that found in
1 Similarly ἐνοπαί, which always denotes a loud or excited cry, a shrill or thrilling sound,
is sometimes used by Euripides of an oracle (cf. Denniston on E. El. 1302).

95
lines 158 f. COMMENTARY
strophe and antistrophe at this point, where the first part of the ode is brought
to a conclusion, is of great beauty. The words τοῖς δ᾽ ὁμόφωνον bring the
refrain into intimate connexion with the preceding sentence and give it a
new vividness."
We do not know whether Aeschylus was the first? to introduce into the
story of Iphigeneia’s sacrifice in Aulis the prodigium of the two eagles devour-
ing a hare. Supposing that the motif had not already been connected with
the saga, no probable guess can be made as to a tradition in which the poet
may have come across it. Nor do the representations on the famous coins
of Akragas (cf. on 118) throw any light on the matter.” All we may safely
infer is this. Since the action of two eagles feeding on one hare is, in the
opinion of experts, ‘utterly untrue to nature’* (D’Arcy Thompson, A Glossary
of Greek Birds, 2nd ed., 189), the idea of such a thing happening was perhaps
from the outset connected with an omen of particular gravity. Aeschylus
makes much of the τέρας. He takes great pains in working out the details,
and especially dwells on the pitiful aspect of the cruel feast. So the commisera-
tion (οἶκτος) and disgust (στυγεῖ δὲ δεῖπνον αἰετῶν) of Artemis are easily
understood. The sufferings of the mother-hare obviously point to the fate
of Troy. The symbolic meaning of the unborn young is less clear. Some
commentators, e.g. Schneidewin and Wecklein, find the explanation in |. 128
πάντα δὲ... κτήνη... τὰ δημιοπληθέα, and take the young hares in their
mother’s womb to represent the Trojan booty. Eugen Petersen, D. ait.
Tragödie, 638, holds that the sign indicates what the Atridae are going to do
to Troy and the Trojans, not only the men but also the women and children.
H. S. Dawson, C.R. xli, 1927, 214, assumes that 'the eagles devouring the
pregnant hares [sic] symbolize the extinction of the Trojan women and
their unborn children at the hands of Agamemnon and Menelaus; he finds
in the passage a direct influence of Z 57 ff. That interpretation had already
been implicitly rejected by T. Plüss, Die Tragödie Agamemnon und das
Tragische, Basel 1896, 6, who pointed out that in 128 f. it is not young human
beings that are mentioned but cattle and wealth. I should prefer not to work
out in minute detail the relation between the sign and its interpretation as
given by Calchas. From the fact that the eagles mercilessly devour the hare
with her unborn young the seer seems to infer that Troy and all that is in
the city will be completely and violently destroyed. The violence is strongly
emphasized in λαπάξει πρὸς τὸ βίαιον, and that it is to be a wholesale destruc-
tion is illustrated by a suggestive detail: not even the flocks outside will be
spared, much less anything within the walls of the conquered city.
1 The opposite device is skilfully employed in Catullus’ song of the Parcae. There the
burden currite ducentes subtegmina eqs. forms a separate unit detached from what precedes
it. Only where itfirst occurs (64. 327) is it different : sed vos, quae fata secuntur, currite ducentes
subtegmina, currile fusi. Here the words currile ducentes eqs. have not yet been shaped into
a refrain proper, but seem spontaneously to arise from the context of the prooemium.
2 This seems to be the opinion of Wilamowitz, Hermes xviii, 1883, 253, and of C. Robert,
D. griech. Heldensage, 1097 ; it is certainly the opinion of Welcker, Der epische Cyclus, ii. 143
n. 78, and of Eugen Petersen, Die attische Tragódie, 638. |
3 The statement of Wilamowitz loc. cit. that on the coins two eagles are tearing asunder
‘a female hare’ is without foundation.
+ This comment recalls the objection raised in antiquity to the simile N 198 ff., where
two lions seize one goat (see the scholia).
96
 THE FEAST OF THE EAGLES AND AGAMEMNON’S GUILT
The main problem, a much vexed one,’ of this whole section arises from the
fact that we are not told anywhere in the ode why the wrath of Artemis is
directed against the Atridae.” The naive assumption that the sole reason for
the anger of the goddess lies in her abhorrence of the eagles’ feast was long
ago rejected by Blomfield (Glossarium on 1. 147: ‘hoc portentum non tam
caussa quam signum fuit sacrificii ab Atridis mox consummandi’), whom
others followed, e.g. Welcker, Die Aeschylische Trilogie, 409; F. Bamberger,
Opusc. 39; but it has not died out. Conington’s surprising inference ‘Diana
had a reason for protecting the hare: but the hare in the language of sym-
bolism meant Troy; hence she hated the future destroyers of Troy’ was
almost literally repeated by C. Robert, Die griech. Heldensage, 1097: ‘Here
[in the chorus of the Ag.] Agamemnon himself is quite guiltless; it is only
the two eagles that are guilty; but as these typify the Atridae, it is against
them that the goddess directs her wrath, demanding the sacrifice of Iphi-
geneia ;’ the same view is held by T. Zielinski, Tragodumenon libri tres, 253 f.
Such an easy shifting of responsibility is inconceivable in a play whose
central problem consists in the connexion between guilt and atonement? and
which is dominated by the belief in the unerring sway of divine justice.
Conington would have been right if he had said ‘hence the seer concluded
that she hated... .'
It is of no avail to have recourse to the excuse urged by Blomfield and
others that Aeschylus did not need to explain the cause of the anger of
Artemis since the audience was perfectly familiar with that detail of a well-
known story. It must be regarded as an established and indeed a guiding
principle for any interpretation of Aeschylus that the poet does not want
us to take into account any feature of a tradition which he does not mention.
That is the more important when the issue is of such magnitude as in the
case of a punishment inflicted upon Agamemnon by a deity, with gravest
consequences for himself and his house. It is, however, perfectly true that
. Aeschylus and his audience had been brought up with a form of the story
which left no doubt about the reason of Artemis' wrath. The real problem
then is this: why does Aeschylus pass over in silence, that is to say why does
he deliberately discard, what was for himself and his audience the recog-
nized version of the story, and that not in some subordinate detail of the
saga, but in a matter of the deepest religious and moral significance?
According to the Cypria, Artemis had good reason for being angry with
Agamemnon. Ín the summary of Proclus we read this: καὶ τὸ δεύτερον
! In addition to the commentaries and the handbooks of mythology special monographs
have dealt with it, e.g. the pamphlets quoted by Wecklein in the introduction to his com-
mentary, p. 20, n. I.
2 The problem would not exist at all if T. Plüss, ‘Die Tragödie Agamemnon und das
Tragische’, Wissensch. Beilage zum Bericht über das Gymnasium, Basel 1896, 12, were right.
He asserted that Artemis is angry not with the Atridae, but only with the eagles, and
endeavoured to support his thesis by a very subtle juridical argumentation. He became
hopelessly entangled in sophisms, a not uncommon case with that fine scholar, who was
always conscientious, never dull, and often happy in his approach to great works of poetry,
but sometimes lacking in common sense.
3 For the same reason (apart from the artificiality of the hypothesis) I find it impossible
to agree with Daube, 147 ff., who regards the sacrifice of Iphigeneia as ‘an atonement pay-
able in advance for the destruction of Troy'. The fundamental maxim δράσαντι παθεῖν
cannot be supplanted by a δράσοντι παθεῖν.
4872-2 H 97
 THE FEAST OF THE EAGLES AND AGAMEMNON’S GUILT
ἠθροισμένου τοῦ στόλον ἐν Αὐλίδι Ἀγαμέμνων ἐπὶ θήρας βαλὼν ἔλαφον ὑπερ-
βάλλειν ἔφησε καὶ τὴν Ἄρτεμιν: μηνίσασα δὲ ἡ θεὸς ἐπέσχεν αὐτοὺς τοῦ πλοῦ
χειμῶνας ἐπιπέμπουσα" Κάλχαντος δὲ εἰπόντος τὴν τῆς θεοῦ μῆνιν καὶ Ἰφιγένειαν
κελεύσαντος θύειν τῆι Ἀρτέμιδι κτλ. The risk of uncritical confidence in the
excerpts of Proclus is well known (see especially Ed. Schwartz, RE i. 2884 f.).
But in this case we seem to be on firm ground. The summary of the Cypria is
considered reliable even by sceptical critics, cf. Wilamowitz, Die Ilias und
Homer, 278 n. 2; E. Bethe, Homer, ii. 206. Moreover, that the feature which
here concerns us, Ἀγαμέμνων ἐπὶ θήρας βαλὼν ἔλαφον ὑπερβάλλειν ἔφησε καὶ
τὴν Ἄρτεμιν, formed part of the narrative of the Cypria is rendered highly
probable by the fact that Sophocles, El. 566 ff., follows the same tradition,?
and that Callimachus, Dian. 263, alludes in passing to Agamemnon’s boasting
on that occasion. We need not, however, stress the point. For even suppos-
ing that in the Cypria the cause of Artemis’ anger was explained not on the
same lines as in Sophocles’ Electra but by one of the different motives of
which traces are preserved in the mythographical tradition (see footnote 2
below), e.g. ὅτι τὴν χρυσῆν ἄρνα οὐκ ἔθυσεν αὐτῆι Arpeis—even so the point
which alone matters here would remain unchanged. For we should at any
rate have to attribute to the Cypria a plain and straightforward causality.
Agamemnon (or, unlikely though this seems to me, his father) insults
Artemis by word or deed or both; hence the goddess brings about the
ἄπλοια and thus holds back the fleet. Such a motive, simple and consistent,
is in keeping with the spirit of fairy-tales, sagas, epic poetry. But with
Aeschylus it was different. Any attempt to accept a story of that type or
to compromise with it would have struck at the very heart of his tragedy.
In the narrative of the Chorus the climax is reached with the monologue
of Agamemnon (206 ff.) and the comments attached to it (218 ff.) The king
is faced with the alternative of two deadly evils. Whichever course he chooses
is bound to lead him to unbearable ἁμαρτία. After a violent struggle he
resolves to sacrifice his daughter, fully aware that what he is going to do is
an unpardonable sin and will have to be atoned for. His fatal step puts him
τ To this we must add the mythographical tradition which ultimately depends on the
Cypria, cf. Wilamowitz, Hermes, xviii, 1883, 250; C. Robert, Heldensage, 1096 f.
2 The discrepancy is slight (cf. Wilamowitz, Hermes, xviii, 1883, 250) and confined to
mere details. I do not see any reason for doubting that Electra’s account of her father’s
transgression in Aulis (misunderstood by O. Gruppe, Griech. Mythologie, 669 n. 4) is based
on the Cypria, as was assumed by Welcker (Aeschyl. Tril. 408, D. epische Cyclus, ii. 101,
143) and has been more or less generally accepted since, among others by Wilamowitz in
the above-quoted article. In his last book, however, Wilamowitz says (Glaube d. Hell.
i. 181 n. 1, ii. 233 n. 2) that Sophocles himself invented Agamemnon's hunting of the deer
and his boasting. That assertion is probably due to a slip of the memory; Wilamowitz
seems to have forgotten the evidence from which he had drawn his former conclusion.
However venerable the tradition preserved in E. Iph. T. 20 ff. and 209 ff. may be, it cannot
be regarded as the tradition followed by the poet of the Cypria. With Zielinski, Tragodum.
245 f., I am inclined to believe that in the epic narrative Agamemnon's offence did not
consist merely in his boast, but that there, too, as in Sophocles' Electra, he was said to have
killed the deer in the sacred grove of Artemis, Αὐλίδος ἀνάσσης (E. Iph. A. 434). Then, by
disparaging the goddess, he added insult to injury. The discrepant reasons for the wrath
of Artemis which we find in the Epitome of [Apollod.] Bibl. 3. 21 must not be ascribed to
the Cypria.
3 Here it has been necessary to anticipate, but the interpretation of the details (see
below) should be taken together with this brief survey.
98
 COMMENTARY line 160

under the yoke of compulsion ; there can be no way back; on and on he must
go, and the end, he knows as well as the Elders, will be utter ruin. Aeschylus,
by using unmistakable language (220 ff. róev . . . πρωτοπήμων), makes it
clear that all the evil that is to befall Agamemnon has its first origin in his
own voluntary decision. There is no need here to go into the general implica-
tions and to consider what in the work of Aeschylus is the essence of a tragic
conflict. For the purposes of the present interpretation it will suffice to state
that in this play the poet has carefully built up a religious and moral ἀφορμή
from which the whole movement of his hero’s doings and sufferings has to
take its start. From the point of view of Aeschylus it was all-important that
nothing but Agamemnon’s deliberate decision should appear as the primary
cause of his sufferings, πρωτοπήμων. That effect would have been impossible
if the king had first, by a comparatively minor offence,’ brought upon him-
self the revenge of Artemis and had consequently been forced to sacrifice
Iphigeneia. In that case the moral dilemma which Aeschylus wanted to be
the fountain-head of Agamemnon’s fate would have been degraded to
secondary importance. By a bold stroke the poet fought his way out of the
difficulty: he followed the traditional story in maintaining the wrath of
Artemis and her appeasement through the sacrifice of Iphigeneia but
eliminated the act of Agamemnon which had incensed the goddess. While
suppressing the cause, he elaborated the details of the sign whose unfavour-
able elements portended the disapproval of Artemis. This solution was
facilitated by the fact that Aeschylus was writing not a coherent narrative
but a retrospective song in a lofty strain and was therefore at liberty to
select and emphasize a few significant points, and moreover by the cryptic
character appropriate to the words of a seer. Still it would be wrong to
under-estimate the audacity of the poet who ventured to withhold from his
audience an important and familiar link of the story. Aeschylus might be
confident that the power of his song would keep the hearers firmly in its
grip and leave no room for idle speculation or curiosity about details.

160. On the Hymn to Zeus as a whole see p. 112 f. First of all we must
discuss certain details of expression and thought, although the division in
this case is even less satisfactory than usual.
ὅστις wor ἐστίν: in these words the many names and appellations of the
god are summed up in a formula which includes them all. The formula itself
is traditional, and so, too, in sense if not actually in wording, is the addition
εἰ τόδ᾽ αὐτῶι φίλον κεκλημένωι. Cf. Plato, Crat. 4ooe ὥσπερ ἐν εὐχαῖς νόμος
ἐστὶν ἡμῖν εὔχεσθαι, οἵτινές τε καὶ ὁπόθεν χαίρουσιν᾽ ὀνομαζόμενοι (the gods),
ταῦτα καὶ ἡμᾶς αὐτοὺς καλεῖν, ὡς ἄλλο μηδὲν εἰδότας. An impressive apprecia-
tion of these liturgical formulae is given by Eduard Norden, following
Usener and others, in a well-known section of his book Agnostos Theos
' Cf. the good remarks of Eugen Petersen (Die atiische Tragödie, 376 and especially
639), who fully appreciated the importance of Aeschylus’ divergence from the Cypria. In
the version of the story which we find in E. Iph. T. 20 ff. and 209 ff. the reason why Artemis
demands the sacrifice of Iphigeneia is quite different from that given in the Cypria (cf.
above p. 98 n. 2), but there, too, the sacrifice is ‘the dreadful consequence of a mere
rashness’ (T. Plüss, Die Tragödie Agamemnon und das Tragische, 11 n. 8).
2 The ritual formula is often used in humorous parody, e.g. Athen. 8. 334 b ὁ τὰ Κύπρια
ποιήσας ἔπη, εἴτε Κύπριός τίς ἐστιν ἣ Στασῖνος ἢ ὅστις δή ποτε χαίρει ὀνομαζόμενος.

99
line 160 COMMENTARY
(144 ff.). The credit of having traced these stock phrases back to the ‘vetus
cultus deorum’ and having discussed a few, but well-chosen, examples belongs
to Petrus Victorius (Variae lecttones, xiii. 2); he also quoted a passage the
expressions of which are very similar to those used by Aeschylus here,
Plat. Phil. 12° καὶ viv τὴν μὲν Ἀφροδίτην, ὅπηι ἐκείνηι φίλον, ταύτηι mpocayo-
ρεύω. An intelligent use of his discussion was made by Stanley on Ag. 160,
who added a reference to E. Troad. 884ff. as well as some Latin parallels
(see his posthumous notes). But what matters most is to understand the
use which Aeschylus makes of these formulae here. For him they are by no
means mere liturgical relics, which he employs according to custom as any
worshipper might when taking part in a traditional cult. Nor have they
become nothing but ornament, purely intended to add greater amplitude
and solemnity to the ode. Such radical ‘secularization’ of religious material,
which plays a large part elsewhere in the history of poetic forms, is foreign
to the spirit of Aeschylus. It is truer to say that for him the elements of
ancient prayers still have religious import, but compared with their tradi-
tional use it is in the service of a more sublime religious feeling that they
stand here (cf. Latte, Archiv f. Religionsw. xx, 1921, 275 n. 2). In the early
stages of religious thought a great expenditure of pious foresight and some-
times actual cunning is considered necessary in order to prevent a daemon
or a god—generally thought of as reluctant—from using some malicious
device so as not to comply with the wishes of the mortal who prays to
him and to evade the magic or almost magic compulsion of his worship
and his prayer. To know the name of the daemon is to acquire power over
him (‘Ei wie gut dass niemand weiss, dass ich Rumpelstilzchen heiss’) ; the
exhaustive enumeration, or if that is impossible the summarizing, of all
his names (ὅστις δή more χαίρεις ὀνομαζόμενος and the like) is therefore
necessary for the prayer to take effect. Aeschylus here takes over the heri-
tage of a more primitive belief because he can make it serve his own advanced
convictions. ὅστις ποτ᾽ ἐστίν : that means here not merely the god’s name and
identity but his real nature and character. Like Euripides (Troad. 885 f.)
Aeschylus might have said to the supreme deity, albeit in a different sense,
ὅστις ποτ᾽ el σύ, δυστόπαστος εἰδέναι, Ζεύς. It was the conviction of Aeschylus
that the manner in which Zeus arrives at his decisions and carries on his
government is hid from mortal knowledge. He moves in a dark thicket,
the paths of his thought are κατιδεῖν ἄφραστοι, as Suppl. 95 puts it (cf. below
on 182); one can see how close this expression comes to δυστόπαστος εἰδέναι
(although in other respects the speculation in that passage of Euripides,
which owes a good deal to sth-century philosophy,’ goes far beyond Aeschy-
lus). Of the true nature of the almighty Lord of Justice we possess no real
knowledge, just as we can only indulge in conjectures on his attitude towards
us mortal men: hence the που in the clause δαιμόνων δέ που χάρις (see the
fuller discussion on 182). Thus the opening words of this hymn, Ζεύς, ὅστις
mor ἐστίν, for all their apparent conventionality, make ready beforehand
for the conclusion which crowns the whole, the idea which is central in the
poet’s thought.
160~2. τόδε. .. τοῦτο: the frequent, though not universal, distinction in the
force of the two demonstrative pronouns is here unmistakable: ‘öde intro-
1 See e.g. F, Solmsen, Plato’s Theology (Cornell University 1942), 45 f.
IOO
 COMMENTARY line 163
duces a direct indication of the object . . . οὗτος gives nothing beyond its
mere identification’ (Kaibel on 5. El. 4 ff., p. 69, cf. ibid. on 371, p. 129).
This differentiation is very old; referring to the Homeric usage, C. Hentze,
Philol. xxvii, 1868, 512, recalls ‘the observation made by Nitzsch that οὗτος,
as distinct from ὅδε which points out something physically, defines the idea
and distinguishes one object from another’. The difference is well formulated
also by Windisch, Curtius Studien, ii (1869), 256: ‘odros is in the main a
pronoun that refers back, ὅδε genuinely deictic.’ So in a rather prolix para-
phrase: ‘If this appellation here is pleasing to him (the speaker surveys the
whole range of possible names of the supreme god, and points to one among
them expressly), then by that one I address him, namely Zeus.’
162. τοῦτό viv προσεννέπω: ‘with this name I address him’, cf. 1291 ‘Aiôou
πύλας δὲ τάσδ᾽ ἐγὼ προσεννέπω ‘and this gate here I address as the gate of
Hades’, but especially Cho. 110 τίνας δὲ τούτους (that is, as εὔφρονες) τῶν
φίλων προσεννέπω; where τίνας τῶν φίλων is the object and τούτους the noun-
predicate (Cho. 224 is not yet restored with certainty). προσεννέπειν is used
of addressing or calling-upon in prayer (as in Ag. 162) in Pind. Isthm. 6. 17
(cf. below on 323), 5. Aj. 857, E. Hipp. 99; an allusion to this usage is found
in Ag. 1291.
163. οὐκ ἔχω κτλ. The asyndeton has caused discomfort to many commenta-
tors, without their perhaps being fully aware of the reason. Thus Wunder
took the words from ὅστις ποτ᾽ ἐστίν to προσεννέπω as a parenthesis, and
received the express approval of Ahrens (p. 291), while Headlam regarded
as a parenthesis εἰ τόδ᾽ αὐτῶι to προσεννέπω. This is the more unjustified,
in that it was at the beginning of hymns and similar appeals to the deity
that ydp was most commonly used to introduce ‘parentheses’ of this kind.
No less arbitrary is Verrall’s assumption of a break in the construction
after Ζεύς. Wilamowitz punctuates rightly, but translates 163 ‘und ob ich
alles wage, zu leicht befind ich alles’, thus by his ‘und’ obscuring the asyn-
deton. The apparent difficulty disappears, as soon as one realizes that the
hymn here, as at the beginning, follows the formulae of prayer. In prayers
after the appeal to the divinity a fresh start with no connecting particle is
a perfectly legitimate, although not the predominant, form. Pind. N. 7. 1
’EAeißvia, πάρεδρε Μοιρᾶν βαθυφρόνων, mai μεγαλοσθενέος, ἄκουσον, Ἥρας,
γενέτειρα τέκνων. ἄνευ σέθεν οὐ φάος κτλ. (it is characteristic that the para-
phrase in the scholia with ἄνευ γὰρ σοῦ interpolates an in itself quite acceptable
connecting particle). That the invocation can be set at the head of the main
body of the prayer with no connexion may be seen with exceptional clarity
in the song of the phallos-procession Ar. Ach. 263 ff., where in 265 at the end
of the many ἐπικλήσεις comes the only pause (catalexis) of the whole ode, so
that ἕκτωε σ᾽ ἔτει προσεῖπον makes a fresh start.
προσεικάσαι. The expression προσεικάζειν τινί τι is several times found in
Aeschylus (and also elsewhere) in a sense which hardly differs from that of
the plain εἰκάζειν τινί τι (as in Sappho's öprarı Bpadivan σε μάλιστ᾽ ἐϊκάσδω,
etc.); cf. προσέοικα and Herodotus’ προσείκελος alongside of the Homeric
εἴκελος. ‘To liken’ gives the force of this εἰκάζειν better than ‘to compare’,
which has too much faded and has become too abstract. To ask ‘whom does
so-and-so resemble?’ came more naturally to the Greeks than to ourselves,
cf. on 1629 ff. Most commentators here, indeed, have avoided this simple
ror
line 163 COMMENTARY
meaning, and render the verb with ‘coniectando, cogitando assequi’ or some:
corresponding expression. Wunder, on the other side, rightly has: ‘Jovi
non possum quicquam comparare omnia perpendens praeter Jovem,’ and
Hartung: ‘seines Gleichen find’ ich riichts ausser ihm.’ If we worked out the
underlying question in accordance with the popular game, it would run:
τίνι τὸν Δία μάλιστα eikdlw;—the answer, after considering all possibilities,
is οὐδενὶ πλὴν Διός. I cannot agree with Ahrens (p. 292) that this is a very
artificial idea. The greatest Christian poet says of his God (Paradiso 15.
76 ff.) : ‘perd che il Sol (i.e. God) che v'allumó ed arse col caldo e con la luce,
è si iguali, che tutte simiglianze sono scarse.’
164. ἐπισταθμώμενος : this compound is found only here. E. Suppl. 202 has
διεσταθμήσατο, Ernst Fraenkel, Griech. Denominativa 158, quotes διασταθμώ-
μενος from Hippocr. π. τέχνης 12. Blomfield explains ‘ad amussim expendo.
a στάθμη, amussis’, but this is too narrow, for the verb is formed from
σταθμᾶσθαι, and though that, too, in the main means ‘to measure’, yet e.g.
in Ar. Frogs 797 we find καὶ γὰρ ταλάντω: μουσικὴ σταθμήσεται. So here ‘to
weigh and measure in comparison”.
165. ei τὸ μάταν ἀπὸ φροντίδος ἄχθος χρὴ βαλεῖν. This type of word-order
(cf. on 156) is very ancient, cf. Homer X 93 ὡς δὲ δράκων ἐπὶ χειῆι ὀρέστερος
ἄνδρα μένηισι, I 7 πολλὸν δὲ παρὲξ ἅλα φῦκος ἔχευεν, X 43 ἦ κέ μοι αἰνὸν ἀπὸ
πραπίδων ἄχος ἔλθοι (more of the kind in my book, Iktus und Akzent, 166),
cf. also on 814 ff. In the line of the J/zad last quoted we should hardly assume
‘tmesis’ (Leaf, ad loc.), and certainly not in the passage of the Ag. before us,
which Wecklein, e.g., thus explains: τὸ μάταν φροντίδος ἄχθος ἀποβαλεῖν (so
Headlam ‘cast off the strange vague burden on my mind’), while the para-
phrase in the scholia quite correctly takes together ei τὸ τῆς ματαιότητος
ὀφείλει τις ἀποσείσασθαι τῆς γνώμης. For alleged tmesis in a similar case cf.
on 1008.
φροντίς here, as in many other passages of Aeschylus, clearly indicates
contemplation and deep thought, especially when it is intense and anxious.
Cf. on 1531. It is less easy to grasp the precise meaning of τὸ μάταν ἄχθος.
Hermann explains: ‘vocat chorus suam sollicitudinem τὸ μάταν ἄχθος, quia
causam eius non idoneam esse videt’ (similarly Headlam: ‘causeless, un-
accountable, unwarranted’). But then the content of thought would be too
slight, and the limitation to the Chorus too narrow, for the requirements of
a passage carrying such religious weight.’ Others take the passage in a
deeper sense, but in so doing borrow to a dubious extent from the ideas of
Jewish and Christian writers and their modern followers. Thus Verrall
comments: ‘the burden, in the language of The Preacher, of “‘vanity’’, the
oppressive sense of futility which must accompany a belief that the moral
problem of the world is insoluble’, and Wilamowitz translates: “Von Sorgen
und von Sinnen und Zweifeln löst das Herze mir Zeus allein.’ I believe that
Aeschylus is here saying something simpler and less general. The μάταν
ἄχθος, which mortal man is to cast out in decisive fashion from his φροντίς,
1 A fine sensibility to this is shown by Triclinius. At the end of a long explanation (ex-
pressly said by him to be his own), in which he quotes Euripides and a passage from the
Psalms, he says: Ταῦτα δὲ ἄγαν ἐστὶν εὐσεβῆ. Cf. also Triclinius on 364 τείνοντα πάλαι τόξον,
where he compares Psalm 7. 13 τὸ τόξον αὐτοῦ ἐνέτεινεν καὶ ἡτοίμασεν αὐτό: no doubt he
remembered that this section of the Psalm begins ὁ θεὸς κριτὴς δίκαιος καὶ ἰσχυρός κτλ,
102
 COMMENTARY line 169
finds an accurate parallel in the μάταια φρονήματα of which Eteocles says
(Septem 438 1.) τῶν τοι ματαίων ἀνδράσιν φρονημάτων ἡ γλῶσσ᾽ ἀληθὴς γίγνεται
κατήγορος. This, and shortly afterwards (441) θεοὺς ἀτίζων κἀπογυμνάξζων
στόμα χαρᾶι ματαίαι, refer to the blasphemous frenzy with which Capaneus
has tried to slight the power of Zeus. Zeus being omnipotent, the specula-
tions of the man who denies his power are rightly called φρονήματα μάταια
and the boasting is uttered yap& ματαίαι. It is remarkable how often μάταιος
denotes that particular variety of the futile, the deluded, the ‘vain’, which
expresses itself in the denial of existing values and powers. Oceanus ends the
speech in which he tries to wean Prometheus from his τραχεῖς kai τεθηγμένοι
λόγοι against Zeus with the warning (329) ὅτε γλώσσηι ματαίαι ζημία προστρί-
βεται, and in the same sense Hermes calls him (999) ὦ μάταιε. This blind
refusal to recognize true worth is seen very clearly in Hdt. 7. 107 1 (Arta-
banus to Mardonius) σὺ δέ, ὦ παῖ Γωβρύεω, παῦσαι λέγων λόγους ματαίους περὶ
᾿ΕἙλλήνων οὐκ ἐόντων ἀξίων φλαύρως ἀκούειν (for the shade of meaning in
μάταιος here Stein adduces other passages in Herodotus). The sense thus
becomes quite clear in 5. El. 642 ματαίαν βάξιν, where Jebb’s explanation
is too indefinite and Kaibel’s wrong. When a man’s inward parts, notwith-
standing any misleading appearance of a happy issue, refuse to be deceived
about the only real working force, the öi«xn that strives to bring things to
their due consummation, then (Ag. 994) σπλάγχνα. . . οὗτοι ματάιζει, and
in Sophocles in a closely related context of the man who is dikas ἀφόβητος
we are told that τῶν ἀθίκτων ἕξεται ματάιζων (Oed. R. 891). Thus τὸ μάταν
ἄχθος is the burden of the folly which induces men to believe that Zeus is
not the almighty ruler, who directs all that is done among mankind. It may
be useful to illustrate this idea from the second choral ode, where we are told
(369) οὐκ ἔφα τις θεοὺς βροτῶν ἀξιοῦσθαι μέλειν, ὅσοις ἀθίκτων χάρις πατοῖθ᾽"
ὁ δ᾽ οὐκ εὐσεβής. Such ἃ misconception is here called ἄχθος because it op-
presses a man and whelms him in stupor, and because he must cast it from
him if he would reach his aim. If he is to succeed in really freeing himself
from his burden, there is only one course to take: to recognize that Zeus is
supreme and that there is nonc other like him.
166. ἐτητύμως (cf. on 682) is the exact opposite of μάτην.
167 ff. οὐδ᾽ ὅστις... οὐδὲ λέξεται. For the negation which is placed at the
beginning of the sentence and then repeated before the verb cf. on 1634 f.
The δέ of the former οὐδέ serves to attach the sentence to what precedes, the
latter οὐδέ means ne . . . quidem.
167. οὐδ᾽ ὅστις πάροιθεν κτλ. The gloss in Tr, ὁ Κρόνος, is wrong. Schütz
recognized that Uranos is meant here and Kronos in 171 (ὃς δ᾽ ἔπειτ᾽ ἔφυ),
and rightly compared Prom. 956 f. οὐκ ἐκ τῶνδ᾽ (the citadel of Zeus) ἐγὼ
δισσοὺς τυράννους ἐκπεσόντας ἠισθόμην;
169. παμμάχωι: Verrall and Headlam are probably right in detecting here,
besides the obvious general meaning, a reference to the παγκράτιον, corre-
sponding to the way in which Zeus in the next sentence is described as
τριακτήρ. The later technical use of πάμμαχος, παμμάχιον can be learnt from
the lexica, but as early as Bacchylides 13 (12). 76 παμμαχίαν (Kenyon, Jebb,
L-S read παμμαχιᾶν, against the accentuation of the papyrus) denotes the
pancration. L. Robert, Études épigraphiques et philologiques (1938), 91 f., in
his important discussion of πάμμαχος and kindred words, points out that in
103
line 169 COMMENTARY
a Theban epigram of the fourth century, IG vii. 2470, πάμμαχον is similarly
used of the pankration. In Arist. Lys. 1321 Athena Chalkioikos is called
adppayos; whether there is a reference there to the technical use, as there
probably is in the application to Zeus here, cannot be determined because of
the mutilation of the passage.
θράσει: for the meaning cf. on 803 f.
170. οὐδὲ λέξεται: Ahrens’s emendation is certain. On the voice cf. Wacker-
nagel, Syntax, i. 139 f.: ‘Originally the middle form served here [i.e. in the
future tense] also as a passive. . . . As late as the fourth century the forms in
-coua were frequently used to express the passive future.’ λέξομαι passive
is apparently confined to Tragedy, cf. Wilamowitz on E. Her. 582. οὐδὲ
λέξεται expresses a strong negation; its sense comes very Close to ἐν ovdevi
λόγωι ἔσται. The intensity of this οὐδὲ λέξεται would be weakened if we took
πρὶν av as a predicate, though that would be possible in grammar, cf. on 269.
It will therefore probably be best, with the majority of commentators, to
regard it as an addition intended to explain the preceding words and
enhance their effect : ‘shall not be reckoned, being one of the past’ (Headlam).
For the use of the present ὦν with πρίν see Wackernagel, Syntax, i. 158; the
phrase there quoted from Homer τά τ᾽ ἐσσόμενα πρό τ᾽ ἐόντα offers a welcome
parallel; cf. also, e.g., E. Ion 1609 οὐκ αἰνοῦσα πρίν, 1612 δυσμενῆ πάροιθεν ὄντα,
and below on 364. For a detailed discussion of the usage see B. Huebner,
Diss. phil. Halens. iv (1880), 128 f.
171. τριακτῆρος : Schol. νικητοῦ ἐκ μεταφορᾶς τῶν ἐν τοῖς πεντάθλοις ἀποτρια-
ζόντων. The word occurs only here. A list of nouns of the agent in -τήρ
found only in Tragedy is given by Ernst Fraenkel, Nomina agentis, it. 13 f.
A derivative of the same verb occurs Cho. 339 οὐκ arpiarros dra (Schol.:
οὐκ ἀνίκητος... amd τῶν παλαιστῶν, ot amorpıalovraı ὑπὸ τῶν ἀντιπάλων).
For the explanation of the word cf. Photius and Suidas (further evidence
in Adler’s edition) s.v. τριαχθῆναι" λέγουσιν οὗ παλαιστρικοὶ ἀντὶ τοῦ τρὶς πεσεῖν.
Cf. further the commentators on A. Eum. 589, A. C. Pearson on Soph. fr.
941. 13, Pfeiffer on Callim. Zamb. fr. 9. 274 ff. The image of Zeus as victor
in the wrestling match is in itself perfectly intelligible (just as is οὐκ ἀτρίακτος
dra), and surely contains no allusion to a particular form of the story of the
overthrow of Kronos. We should therefore do better to leave on one side
the alleged Elean tradition of the wrestling of Zeus and Kronos in Olympia
given by Pausanias 5. 7. 10 and 8. 2. 2, to which Schneidewin and Headlam
refer here," even were the authority for it really greater than is assumed by
Wilamowitz, Pindaros, 214 n. 1, in opposition to Pohlenz, RE xi. 1989. The
image of the wrestling match is then carried on in the choice of the expression
ἐπινίκια (174).
174. κλάζων, a strong word (cf. on 156), as is appropriate to τήνελλα καλλίνικε
and similar acclamations. A word very similar in sound and sense, κεχλαδώς,
is used by Pindar where he refers to the jubilant greeting τήνελλα καλλίνικε
κτλ, (Ol. 9. 1 f.).
175. τεύξεται φρενῶν τὸ πᾶν. The expression is faultless and should not be
1 Th. Bergk, Griech. Lit. Gesch. iii. 356, protests against this as a typical example of
blundering interpretation. He rightly demands that ‘the commentator [on Aeschylus]
should eschew all petty pedantry, drag nothing in by the heels, but steep himself entirely
in the simple grandeur of the work’,
104
 COMMENTARY line 176
tampered with. It is wrong to argue that one either possesses φρένες or does
not possess them, and cannot ‘obtain’ them, reach them as a goal, or come
into possession of them. Editors rightly refer to Prom. 444 φρενῶν ἐπηβόλους
(cf. Wecklein’s commentary and the ancient explanations cited by him).
The meaning of emnßoAos, which is as old as Homer, is given correctly in
Tim. Lex. Plat. 126 emnßoAoı οἱ ἐπιτυχῶς βάλλοντες" βάλλειν γὰρ τὸ τυχεῖν, and
this explanation shows very clearly how close to one another are the two
passages in Aeschylus.' The expression τεύξεται φρενῶν rests upon the idea
that there is a strictly limited area of φρένες which one can and must hit
(like a man shooting, e.g. Cambyses in Hdt. 3. 35. 2 ei μὲν γὰρ τοῦ παιδὸς
τοῦ aoû . . . βαλὼν τύχω μέσης τῆς kapôins). This imagery is used without
pedantic regard for the fact that the target is the φρένες of the marksman
himself. If he misses his mark, he is in the position of the man to whom one
can say (Prom. 472f.) ἀποσφαλεὶς φρενῶν πλανᾶι. So Headlam is right in
explaining τεύξεται φρενῶν as ‘the opposite of ἁμαρτήσεται φρενῶν᾽ (cf. on
1664). Cf. Theognis 1. 408 γνώμης οὐκ ἀγαθῆς éruyes. There is a related
expression, although not quite so bold, in 380 εὖ πραπίδων λαχόντι. Cf. also
S. Aj. 1256 ei μὴ νοῦν κατακτήσηι τινά.
τὸ πᾶν. The adverbial use οἱ τὸ πᾶν and ἐς τὸ πᾶν in ἃ strongly affırmative
sense (prorsus, ‘altogether’) is a favourite with Aeschylus, cf. on 429.
176. φρονεῖν ; the verb has, broadly speaking, two principal meanings, which
can be distinguished as early as Homer, ‘to think, to be minded’ and ‘to
be wise, to have understanding’ (cf. J. Böhme, Die Seele und das Ich im
homer. Epos, Leipzig, 1929, 44). Here we are concerned with the second only,
which in Homer is very much in the background (for the reason see Böhme,
p. 45) but ‘is the most frequent sense in Attic’ (L-S) ; it comes very close to
the meaning of σωφρονεῖν and of ὑγίεια φρενῶν (Eum. 535 f.). In what φρονεῖν
mainly, if not exclusively, consists here is made clear by the following
words: τῶι πάθει μάθος κτλ. The clause καὶ map’ ἄκοντας ἦλθε σωφρονεῖν
looks back to φρονεῖν, selecting and emphasizing out of its wider sphere
of meaning the most important mode of its activity, cf. on 180. The quint-
essence of these sentences is given in the briefest form in Eum. 520 ξυμφέρει
σωφρονεῖν ὑπὸ στένει, ‘conducit rebus angustis coactum sapere’ (Hermann) ;
Headlam there remarks (and it is equally true for Ag. 181): σωφρονεῖν is
synonymous with γνῶναι σεαυτόν, to know your place in relation to the Gods
and your fellow-men’. We can have no detailed knowledge of the way in which
God watches over the affairs of men and governs them (Eum. 530, in the same
context as the sentence just quoted): ἀλλ᾽ dAA& δ᾽ ἐφορεύει. But one thing
God has made clear to us: he demands that we shall respect τὸ μέσον, τὸ
μέτριον, τὸν καιρόν, and beware of all extremes, of anything that is too big
and goes too far (μηδὲν ἄγαν) : παντὲ μέσωι τὸ κράτος θεὸς wmaoev.* If we act
and behave accordingly, we shall display the capacity for σωφρονεῖν, ὑγίεια
φρενῶν (Eum. 535 f.). Whether Sophocles, as he wrote the concluding words of

! The strange 4 in ἐπήβολος is discussed by Wackernagel, Dehnungsgesetz, 36; he too

regards the word as a compound of the preposition and the stem βολο-.
2 The best commentary on Eum. 530 is to be found in the lines Suppl. 1057-61, where
καιρόν was properly understood by Schütz and μηδὲν ἀγάζειν by Paley. Generally speaking,
the odes of the Supplices, compared with those of Oresteia, show the consistency in Aeschylus’
fundamental conceptions regarding the ways of Zeus.
105
line 176 COMMENTARY
the Antigone (μεγάλοι δὲ λόγοι μεγάλας πληγὰς τῶν ὑπεραύχων ἀποτείσαντες γήραι
τὸ φρονεῖν ἐδίδαξαν) thought of our passage in the Agamemnon, we cannot
say, but at least in their idea of the way in which mortal men are brought
to φρονεῖν he and Aeschylus are at one. With φρονεῖν as the goal of man’s
journey we are on a much higher plane of ideas than is reached by the
traditional verdict παθὼν δέ re νήπιος ἔγνω or in proverbial expressions such
as παθήματα μαθήματα and the like.
ὁδώσαντα. The verb ὁδοῦν, ‘to set in the right path’, meets us first in
Aeschylus (twice in Prom.). Its occurrences are listed by Ernst Fraenkel,
Griech. Denominaliva, 145.
τῶι πάθει μάθος θέντα κυρίως ἔχειν. Schütz and Porson alter τῶι to τὸν,
but the anaphora thus produced would make θέντα parallel to ὁδώσαντα and
thus destroy the subordination which is demanded by the sense. ‘Recte se
habere τῶι πάθει μάθος θέντα vidit Wellauerus. Iuppiter dicitur βροτοὺς φρονεῖν
ὁδῶσαι θεὶς τῶι πάθει μάθος κυρίως Exew’ (Hermann), ‘eo enim quod experientia
malorum erudimur via nobis ad virtutem aperta est’ (Wilamowitz). (P. Maas,
Die neuen Responsionsfreiheiten, ii, 1921, 7 n. 2, on the other hand supports
Schütz's τὸν... θέντα, observing that repetition of the article is demanded
by the asyndeton.) But how are we to understand the rau? We cannot
accept Nägelsbach’s interpretation: ‘Construe: τῶι πάθει θέντα κυρίως ἔχειν
μάθος, der dem Leiden die Bestimmung gegeben hat, dass es unwiderruflich
enthalte Witzigung’. For κυρίως ἔχειν (with the intransitive meaning of the
verb) is ‘to be valid’, ‘to be good in law’, as is clear from the passage classed
with ours by Passow, Isaeus 7. 26 ὅτι τὰ πεπραγμένα Ἀπολλοδώρωι κυρίως
ἔχειν νομίζει καὶ κατὰ τοὺς νόμους. Thus τῶι πάθει, as we might expect, is an
instrumental or causal dative. With reference to the article Wilamowitz
says: 'articulus eandem vim habet atque ἐξ αὐτοῦ ἐκείνου ὃ àv πάθωμεν". I
would paraphrase it—in very prosaic style—‘through (or by means of) the
πάθος appropriate to any given μάθος᾽, 1.6. I would see here an example of
the so-called demonstrative or possessive use of the article, cf. in general
Kühner-Gerth, i. 593, and for Aeschylus Wilamowitz on Ag. 288 and below
on 350 and 896. A good parallel is offered, it seems to me, by a passage which
is cognate in thought as well, Democritus B 182 τὰ μὲν καλὰ χρήματα τοῖς
πόνοις ἡ μάθησις ἐξεργάζεται, τὰ δὲ αἰσχρὰ ἄνευ πόνων αὐτόματα καρποῦται.
The contrast of τοῖς πόνοις and ἄνεν πόνων is instructive: on the positive
side Democritus is not satisfied with a reference to toils in general but wishes
to bring out that for the winning of καλὰ χρήματα the only means is the
endurance of specific πόνοι which are inextricably bound up with them
(Meineke seems to have objected to the article in τοῖς πόνοις, and one may
sympathize with his difficulties; but his conjecture rote: ἀνθρώποισι is not
worth discussing).
179. στάζει: Verrall explains this: ‘the admonitory recollection of experience
is compared to a wound which long afterwards will ache at times and even
break out again, reminding the sufferer of the original hurt’. This is sym-
pathetic and vivid ; the idea of the wound is suggested by στάζει (cf. E. Med.
1199, S. Phil. 783, etc.), and in the passage compared by Schiitz, Ag. 1121
ἐπὶ δὲ καρδίαν ἔδραμε κροκοβαφὴς σταγών xrÀ., it is introduced in the following
relative clause. Similarly in Cho. 842 we have ἄχθος δειματοσταγές in a con-
1 When Heath and, more recently, Blass defended the MS reading of this passage, they
106
 COMMENTARY line 179
text where severe wounding is explicitly mentioned; ἄχθος δειματοσταγές is
well explained by Schütz. But here in Ag. 179 we must be careful not to
lay too strong an emphasis on the possible reference to a wound. It is also
possible to find in the passage the idea of constant dropping" (cf. E. Suppl.
79 ff. ἄπληστος ade μ᾽ ἐξάγει χάρις γόων πολύπονος, ὡς ἐξ ἀλιβάτου πέτρας
ὑγρὰ ῥέουσα σταγὼν ämavoros ἀεὶ γόων) ; this goes well with the μνησιπήμων
πόνος Of the man who has some guilt upon his conscience (and it is with him
in the first instance that we are concerned here): for him there may be
pauses between one drop and the next, but there is no release: μέμνει γὰρ
φοβερὰ maAivopros μνάμων Mis.
ἀνθ᾽ ὕπνου. Only by force or cunning can one attempt to preserve the ἔν
θ᾽ ὕπνωι of the MSS. Force was used for instance by Hermann: ‘ad στάζει
ex sequentibus, ut particula post ἐν posita ostendit, accusativus τὸ σωφρονεῖν
supplendus est’, cunning by Headlam when in support of this alleged use
ofre...xai he cited a great number of passages, not one of which is really
comparable. The ‘temporal parataxis’ of two verbs, as in 5. Ant. 1186 f.,
καὶ τυγχάνω τε κλῆιθρα..... χαλῶσα καί με φθόγγος... βάλλει BV ὦτων (cf. Jebb,
ad loc.) is well known, but cannot be found here. The ‘parallels’ adduced
bring out all the more clearly that in Ag. 179 we have not got what with
such a use of re... καί we must expect, namely στάζει τ᾽ ἐν ὕπνωι.... καί... .;
what we have is ἔν θ᾽ ὕπνωι.2 Plüss’s naive remark that ἔν θ᾽ ὕπνωι 'corre-
sponds to’ καὶ παρ᾽ ἄκοντας reveals the editor’s embarrassment. If we then
look up Stein’s note on Hdt. 6. 41 (lines 5 f.), quoted by Headlam, where a
number of passages in Herodotus are discussed, we arc told with reference
to them all: ‘In this parataxis re often serves at the same time, as it does
here, as a connecting link with the preceding clause’. But in Ag. 179 the
connexion is made with the help of στάζει δέ. Furthermore it can be shown
that the following καί cannot have a copulative function, cf. on 180. ἔν
θ᾽ ὕπνωι is therefore a corruption. Emperius’s admirable emendation ἀνθ᾽
ὕπνου was long ago accepted by Bamberger, Schoemann, Enger, and Ahrens.
Schoemann (Opusc. iii. 165) attempted to establish a quite close connexion
between ἀνθ᾽ ὕπνου and στάζει, maintaining that orafew is properly used of
sleep; but his Greek examples, ὕπνος... νήδυμος ἀμφιχυθείς and περὶ δ᾽
ἀμβρόσιος κέχυθ᾽ ὕπνος, are quite different, and even assuming that passages
from Virgil are evidence for early Greek ways of thinking, fessos sopor
irrigat artus and the like present a very different idea. Above all, it is quite
wrong to infer that, because the poet says πόνος στάζει ἀνθ᾽ ὕπνου, it must be
possible to say ὁ ὕπνος στάζει. For ὕπνος indicates a state of being; the real
contrast is not between one ‘subject’ of στάζειν and another which stands on
the same footing, but between sleep on the one hand and the ordlew of
πόνος on the other. Thus, to reproduce it with pedantic accuracy: ‘instead
of sleep’ (or ‘in the place where if things ran their untroubled course one
would find sleep’) ‘one finds the slow drip, drip of pain’.
used the parallel Ag. 179 f. as their chief argument (Tucker on Cho. 842 says ‘the thought
in Ag. 179 is quite different’). Convinced by Blass, Wilamowitz (cf. the note at the end of
his translation) recanted his former opinion.
1 This was written before I had seen the remarks of G. Méautis, Eschyle, 137 n. 1, with
which I fully agree.
2 For this reason the passages quoted by Denniston, Particles, 515, are clearly unsuited
to defend the MS reading in Ag. 179. |
107
line 179 COMMENTARY
The passages quoted by Headlam (cf. also C.R. xvii, 1903, 241) to support
ἐν ὕπνωι from the point of view of sense are surprising, for Job 33. 14 ‘God
speaketh . . . in a dream, in a vision of the night, when deep sleep falleth
upon men... then he openeth the ears of men and sealeth their instruction’
(the similar passage Job 4. 13 ff. had been compared by Scholefield) and
Psalm 15. 7 ‘I will bless the Lord who has given me counsel: yea, my reins
instruct me in the night seasons’ are parallel not to ἐν ὕπνωι πόνος στάζει but
to the familiar νυκτὶ βουλὴν δίδου (Plut. Them. 26. 2, Hdt. 7. 12. 1, cf. Men.
Epitr. 35). And if Dio Chrysostom (40. 21) says διασπᾶι καὶ διαφθείρει τὸν
ὕπνον and the narrator in Achilles Tatius 1. 6. 2 οὐδὲ ὕπνου τυχεῖν ἠδυνάμην,
and that is substantiated out of the experience of us all, is that any support,
as Headlam would clearly like it to be, for ev ünvaı? But why do we need
support at all? No man who is not stock and stone will doubt that trouble
and affliction pursue their victims even into the realm of sleep and dreams;
but does that make it any the less true that only too often the victim of fate
‘die kummervollen Nächte auf seinem Bette weinend sass’ ?'
πρὸ καρδίας: that the emotions are active before or near the heart is a
view we often meet in Aeschylus,? cf. below on 834 and on 976, also Cho. 1024
πρὸς δὲ καρδίᾷι φόβος ἄιδειν ἑτοῖμος, Sept. 288 f. γείτονες δὲ καρδίας μέριμναι
ζωπυροῦσι τάρβος. The idea may have its basis in physiological theories.
There is probably some connexion with the position allotted to the φρένες,
the diaphragm, as a seat of the emotions, and similarly among the Romans
to the praecordia (which properly, in the anatomical sense, is equivalent to
φρένες, cf. Plin. Nat. hist. 11. 197).
180. καὶ wap’ ἄκοντας : eiiam ad invitos, even to them. καί as e.g. in À 300
ὄφρα καὶ οὐκ ἐθέλων τις ἀναγκαίηι πολεμίζοι, A. Suppl. 83 ff. ἔστι δὲ κἀκ πολέμου
τειρομένοις βωμὸς ἀρῆς φυγάσιν ῥῦμα, and elsewhere. Obviously the καί, with
which the γνώμη (ἦλθε) emphatically opens, cannot have copulative force
(re . . . καῇ here. For the idea Platt, J. Phil. xxxii, 1913, 47, compares
Thuc. 6. 87. 4 ἀμφότεροι ἀναγκάζονται ὁ μὲν ἄκων σωφρονεῖν, ὁ δ᾽ ἀπραγμόνως
σώιζεσθαι. Adding ᾧ 13 (θεοὶ) καί re χαλιφρονέοντα σαοφροσύνης ἐπέβησαν,
A. Eum. 550 ἀνάγκας ἄτερ δίκαιος ὧν οὐκ ἄνολβος ἔσται, and especially Eur. fr.
524. 2 N. τὸ φῶς δ᾽ ἀνάγκην (cf. Thuc.) προστίθησι σωφρονεῖν (van Heusde),
one is tempted to ask whether behind all this may not lie some formula of
popular philosophy (developed on Stoic lines by Seneca, Dial. 7. 15. sf.
habebit illud in animo vetus praeceptum: deum sequere! quisquis autem queritur
et plorat et gemit, imperata facere vi cogitur et invitus rapitur ad iussa nthilo
minus). Besides, the passages quoted show that the idea of σωφρονεῖν
acquired under the compulsion of a higher power has again and again
occupied men's thoughts, while in the more general idea of φρονεῖν (176), to
which man is led by God in Aeschylus, we have perhaps to recognize a more
advanced conception.
181. ἦλθε σωφρονεῖν. On the construction cf. (against Housman, C.R. ii,
1888, 244) 584 and the instances quoted there.
182 f. δαιμόνων δέ που χάρις. . . . ἡμένων. This sentence has been much mis-
! Cf. Philol. Ixxxvi, 1931, 13 n. 35.
? There is not a word on this in the superficial compilation of J. Dumortier, Le Vocabu-
laire médical d’Eschyle (Paris 1935), although several pages there are filled with passages
where Aeschylus mentions the heart.
108
 COMMENTARY lines 182 f.
handled, but the sense can be quite certainly determined. Perhaps it is best
to start from the words βιαίως σέλμα σεμνὸν ἡμένων, and to interpret these it
is essential not to obscure the sense of σέλμα. For the view generally current
Triclinius seems to be ultimately responsible. In contrast to the older
scholion preserved in the Mediceus, (σέλμα) τὸν ζυγόν: ὑψίζυγος γὰρ ὁ Ζεύς
(see below), we find in Tr above σέλμα σεμνόν the interlinear gloss καθέδραν
σεμνήν, ἤτοι τὸν οὐρανόν. Accordingly Stanley, although he printed the correct
and early scholion below his text, translated: ‘sedili venerando (coeli) in-
sidentium’. This interpretation held sway with only negligible variations
through the centuries, appearing in Hermann ('tenentium verenda culmina',
Opusc. v. 346) and afterwards (e.g. Conington: ‘from their glorious thrones
of light’, Nägelsbach : ‘die gewaltig auf erhabenem Stuhle thronen’, Paley:
"who sit on their awful thrones’,' Murray: ‘lifted to the throne of life’,
McNeice: ‘throned inviolably’). Most of the lexica follow suit, e.g. Linwood,
Lex. Aesch.: ‘a seat or throne’, Dindorf: ‘improprie de throno’, L-S: ‘gener-
ally, seat, throne’. The correct interpretation was put forward as early as
Schiitz, whose thorough discussion of the passage and victorious refutation
of the view that ᾿᾽σέλμα = throne’ ought to have received general recognition.
Actually only a minority followed his lead, e.g. Blomfield (‘transtrum’),
J. H. Voss (Humboldt only half-way), Bothe, Passow, Schneidewin, Karsten,
Kennedy, Wecklein, Lewis Campbell, Plüss, E. Petersen, D. att. Tragödie,
640, Kranz, Hermes, liv, 1919, 305, Schuursma, De... abusione, 15. On the
negative side, it is an objection to the rendering ‘throne’ that in the early
period σέλματα or σέλμα (the singular in hymn. Hom. 7. 47) is never used
except in connexion with a ship. On A. Sept. 32 L-S have gone astray with
Blomfield (‘tabulatum’) and the editors who follow him, while the scholion
rightly says κατεχρήσατο; Wilamowitz, Interpr. 60 n., briefly mentions that
here ‘the imagery (σέλματα, θωρακεῖα) is taken from a man-o'-war; the ev
πρύμνηι πόλεως of line 2 is in harmony with it’. On the positive side, it is
in favour of the rendering ‘steersman’s bench’ that in general both Greeks
and Romans (hence our govern, government) use ‘to steer’ with remarkable
frequency of the management of city and state so that it means ‘to rule’ or
‘to be lord over’,? and in particular that Aeschylus often uses the metaphor,
both of human rulers (Sept. 2 f., 62, Ag. 1618, Eum. 16, 765) and of divine
(Suppl. τος fí.—about the uncertainty of that passage see p. 1r0o—Prom.
149, 515). At this point we derive valuable assistance from the scholion,
although apparently we have it only in a much abbreviated form: τὸν ζυγόν'
ὑψίζυγος yap ὁ Ζεύς. This explanation is linked up with the ancient inter-
pretation of Homer. Ζεὺς (or Κρονίδης) ὑψίζυγος occurs four times in the
Iliad, and on this phrase we have a substantial ancient note, the fuller form
of which is found not in the faded scholion on 4 166 (ἡ μεταφορὰ ἀπὸ τῶν ἐν
ναυσὶ ζυγῶν eb’ ὧν καθέζονται of épéaaovrec),? but in Eustathius on 4 166 and
ı This contrasts with Paley’s correct explanation of this passage in his notes on Ag. 1618
and on E. Jon 595.
2 Considerations of a more general character lead to the conclusion that in Greece in
early times the helmsman had command of the ship. On this problem and on the later
development of the position of the κυβερνήτης cf. Koester in Kromayer-Veith, Heerwesen...
der Griechen und Römer (Handb. d. Altertumsw. iv. 3. 2, 1928), 188.
* Leaf, who rightly refers to A. Ag. 183 and 1618, unfortunately contented himself, like
Ebeling in his Lexicon Homericum, with reprinting this less valuable version of the ancient
109
lines 182 f. COMMENTARY

on H 69, in a scholion on H 69 (much abbreviated), in Eiym. M., p. 786. 4

(‘Oros’, cf. R. Reitzenstein, Gesch. d. griech. Etymologika, 348 n. 2) ; it seems
likelythatwehave an echo ofthisHomeric exegesis in the admirable scholion on
E. Phoen. 75 μετενήνοχεν ἀπὸ τῶν κυβερνητῶν. τοιοῦτόν ἐστι καὶ τὸ παρ᾽ 'Ομήρωι
‘Zeus ὑψίζυγος᾽. ἐπὶ γὰρ τοῦ ὑψίστου ζυγοῦ καθήμενος ὁ κυβερνήτης περιφέρει
τοὺς οἴακας (the passage in the text to which this belongs, Phoen. 74 f., is
quoted by Eustathius on 4 166). The scholia which here underlie Eustathius
contained excellent learning, as can be seen from their refutation of the
opposite rendering ὁ ἐν ὕψει ζυγοστατῶν, 6 ἄνω σταθμίζων, which includes a
thorough discussion and definition of the Homeric use of ζυγόν. The sense
ὁ ἀφ᾽ ὕψους τὰ κάτω κυβερνῶν ws ἀπὸ ζυγῶν κυβερνητικῶν is shown by the
learned commentator to be the right one. This interpretation always
remained valid among educated readers of Homer in antiquity, although it
was perhaps not the only one current: Numenius (quoted by Eusebius,
praep. ev. ττ. 18, p. 539 c) κυβερνήτης μέν που Ev μέσωι πελάγει φορούμενος ὑπὲρ
πηδαλίων ὑψίζυγος τοῖς οἴαξι διιθύνει τὴν ναῦν ἐφεζόμενος, and similarly Eusebius,
De laud. Const. το, p. 223. 12 f. Heikel.? It is probable that by the time of
Aeschylus this meaning of ὑψίζυγος was instilled into Athenian schoolboys
when they had to learn their Ὁμήρου γλώττας : τί καλοῦσι κόρυμβα; τί καλοῦσ᾽
ἀμένηνα κάρηνα; and the like. The replacement of ὑψίζυγος by σέλμα σεμνὸν
ἥμενος is wholly appropriate. σεμνόν brings out the implication of the
Homeric expression.’ oéAua can mean any £vyóv* or transtrum of the ship,
and in particular that part of the quarter-deck on which the bench or seat
of the helmsman was to be found. That ζυγόν (or ζυγός) was a current name
for it can be seen from Ag. 1618 and the passages quoted there in the note.
It is possible that in Suppl. 595 ff. Aeschylus had used allusively of Zeus
the image of the helmsman sitting above all others and ruling all, as Schneide-
win on Ag. 183 assumes and Paley on Suppl. 595 thinks possible. But so
long as we have no really intelligible restoration and explanation? of the text
of Suppl. 596 f., we cannot decide that with certainty. As regards Ag. 182,
Schütz recognized that δαιμόνων, taken with its whole context, must refer
in particular to Zeus; Wilamowitz's reference to Suppl. 98 (100 Murray)
should also not be overlooked. This point of view, from which Zeus is seen
scholion. In L-S too, failure to make adequate use of the ancient commentary has blurred
the precision of the image : ὑψέζυγος, prop. of a rower, sitting high on the benches: metaph.
of Zeus, high-throned', this last containing the same arbitrary confusion that we find in the
prevailing translation of σέλμα σεμνὸν ἡμένων. .
* We can hardly doubt that this squares also with the idea of the epic poets. A seafaring
folk, at the time when there was as yet no phalanx or corresponding military formation,
saw in the relation of the helmsman to his crew, if not the only typical instance, certainly
the most impressive one of the unconditional subordination of many men to the command
of one. Consequently there could be no better symbol for absolute authority. The fact that
the steersman sat considerably higher than his rowers (cf. on 1617 f.) increased still further
the appropriateness of the metaphor. A. B. Cook, who understands 'steersman' rightly,
Zeus, ii. 830 n. 1, ought not to have considered even tentatively an interpretation of his
own, ‘who driveth his team aloft’.
3 Both passages in Passow s.v. ὑψίζυγος,
3 We should like to know what Pindar had in mind when he said (fr. 14 Schr.) Ata
εὐρύζυγον instead of the Homeric ὑψίζυγος, which Bacchylides like Hesiod before him simply
took over.
4 Cf. Pollux 1. 89 τὸ δὲ ζυγὸν καλεῖται καὶ κλὴις καὶ σέλμα, Hesychius ocAgara: τὰ ζυγὰ
τῆς νεώς. 5 Wilamowitz’s treatment (cf. also his Interpr. 36) seems too artificial.
110
 COMMENTARY lines 182 f.

as representative of all the gods, is contained by implication in the prayer

of Odysseus, v 97 ff.: Au δ᾽ εὔξατο... 'Ζεῦ πάτερ, ei μ᾽ ἐθέλοντες... ἤγετ᾽
ἐμὴν ἐς γαῖαν, ἐπεί μ᾽ ἐκακώσατε λίην᾽ κτλ.
Against βιαίως objections have been raised on inadequate grounds, or
attempts have been made to weaken its sense. But this ruler administers
his office, not in kindliness like the gods of Plato (Critias 109 ©) ἦι μάλιστα
edorpobov ζῶιον, ἐκ πρύμνης àrevbüvorres,! olov oları πειθοῖ ψυχῆς ἐφαπτόμενοι
κατὰ τὴν αὐτῶν διάνοιαν, but like the gods of Heraclitus, who, as Plato dis-
approvingly says (ibid. 1095), are σώμασι σώματα βιαζόμενοι,2 καθάπερ ποιμένες
κτήνη πληγῆι νέμοντες (the passage alluded to is Heraclitus, fr. 11 Diels, πᾶν
ἑρπετὸν πληγῆι νέμεται). In this hymn Zeus is a stern and violent over-
lord. ‘Bia, which in Homer is merely bodily strength, has become detestable
brute force’ (Wilamowitz on E. Her. 215). This aspect of the lordship of
Zeus is by no means predominant in Aeschylus, as anyone knows who has
read the Supplices. But in the series of events which form the background
to this hymn, all the emphasis is laid on the unsympathetic sternness of the
divine rule. For Agamemnon there is no escape, no way out to innocence and
happiness. ‘Ihr lasst den Armen schuldig werden, dann überlasst ihr ihn
der Pein: denn alle Schuld rächt sich auf Erden.’ Aeschylus forbids us to
water down his severe thought, or humanize his stern God to please our-
selves. With σέλμα ἡμένων, which indicates merely the exercise of authority,
βιαίως goes very well. Klausen’s (‘de imperio deorum vi condito et servato’)
and Verrall’s (‘who came by struggle to his majestic seat’) interpretation,
accepted, e.g., by Murray, reads into ἡμένων an arbitrary meaning ; moreover,
this whole section (from 176 on) is solely concerned with the relation between
Zeus and those who are ruled by him, and not (like 168 ff.) with the relation
between him and his predecessors.
With the construction odAua . . . ἡμένων Wellauer has compared S. 447. 249
εἰρεσίας ζυγὸν ἑζόμενον ; more examples will be found in Kühner-Gerth, i. 313 f.
Thus the phrase βιαίως σέλμα σεμνὸν ἡμένων indicates the supremacy of
the heavenly powers, and especially of the highest god, exercising itself with
unbridled force. The opposite of that is χάρις. Between these two contrasted
phrases there is an interplay like that between πάθος and μάθος, between
ἄκοντας and σωφρονεῖν. What is meant by χάρις is clear enough. The word
denotes here as so often the kind of disposition and of behaviour from which
springs the kind of action described by χαρίζεσθαι, and to a certain extent
such action is included in it; see the brief but instructive aphorisms on this
subject in Wilamowitz, Pindaros, 152. We are not of course at liberty to
falsify the meaning by making a kind of ‘interpolatio Christiana’ (L. Traube,
Vorlesungen, ii. 67), and rendering ‘Gott ist gütig' (Wilamowitz). ‘Grace’ is
in itself a good word for the behaviour of the god or man who volunteers to
do something to which he is not bound; but this rendering stresses unduly
one particular aspect, and has also the disadvantage of Christian associations.
The more general idea of ‘favour’ is adequate, but there is also an element of
‘graciousness’ or ‘kindness’. The δαιμόνων χάρις is in essence no other than
the χάρις Διός, which has brought about the punishment of Troy and the
fortunate homecoming of the army (581 f.).
1 Again an adaption of the Homeric Ζεὺς ὑψίζυγος (cf. Politicus 273 °-e).
2 The contrast of βία and πειθώ is common and is a special favourite with Plato.
III
 lines 182 f. COMMENTARY

wou: ‘surely there is... .'. Wackernagel, Kuhns Zeitschrift, xxxiii, 1895,
,

22, points out that as early as Homer «ov is used less often with a local
meaning than in the sense of 'certainly', ‘to be sure’, ‘in statements of whose
truth one is convinced, although one cannot prove them'. This is primarily
so ‘in sentences that refer to the rule of the gods', for which Wackernagel
adduces a group of examples from Homer (so e.g. also Pind. P. 1o. 11, Bacchyl.
5. 91 τὰ δέ που Παλλάδι ξανθᾶι μέλει, A. Pers. 740, S. 417. 489 θεοῖς yàp ὧδ᾽
ἔδοξέ που). Our passage takes its place with these. The cautious utterance,
avoiding an absolute statement, is quite essential. The crushing power of
the gods over us we know only too well and by many tokens ; that at the same
time a χάρις is to be detected in it the poet believes, but restrains himself
from stating it too absolutely as a fact. δαυλοὲ γὰρ πραπίδων δάσκιοί Te
τείνουσιν πόροι κατιδεῖν ἄφραστοι, so we are told in Suppl. 92 ff., to which we
must add 1057 f. ri δὲ μέλλω φρένα Δίαν καθορᾶν, ὄψιν ἄβυσσον; (about the
context cf. above p. 105 n. 2); and in the same play it is said of Zeus (100)
φρόνημά πως αὐτόθεν ἐξέπραξεν ἔμπας ἑδράνων ἀφ᾽ ayvav, where πὼς has the
same function as ποὺ here.
The poet leaves us in no doubt wherein consists the xapıs of Zeus, if we
may regard it as such (που). The juxtaposition of mercy and absolute power
sends our thoughts back to the earlier collocation of πάθος and μάθος. But
even apart from that, the unrelieved severity of the point of view which
dominates the whole stanza rules out the supposition that xapıs can mean
something vague and general; the more so, as tlıe central thought has just
again been impressed upon us with «ai παρ᾽ ἄκοντας ἦλθε σωφρονεῖν. As
δαιμόνων looks back to Zeus, as σέλμα σεμνὸν ἡμένων takes up the picture we
have already been given of his unlimited supremacy, so χάρις likewise must
refer to the only χαρίζεσθαι on his part of which there has been any mention:
τὸν φρονεῖν βροτοὺς ὁδώσαντα τῶι πάθει μάθος θέντα κυρίως ἔχειν. There might
be a world—so far the poet’s thinking seems to have progressed—in which
man had only to suffer, struck down by obscure powers without ever under-
standing why, without ever recognizing any connexion between doing and
suffering. As things are, this is not so, and the god who is pre-eminent in
eternity has opened a way to φρονεῖν (cf. on 176) and an insight into the ever-
valid principle δράσαντι παθεῖν : that this is so is, so far as man can make out,
a χάρις δαιμόνων. It is true that to see this χάρις for what it is, we must look
deep. In our first impression the suffering ordained by God seems anything
but χάρις; it is only in its results, in what we learn by it, that it proves to
be a favour. The same idea is not indeed clearly expressed but distinctly
- hinted at by Herodotus; for when, alluding to the proverbial πάθος μάθος,
he makes his Croesus say (1. 207. 1) τὰ δέ μοι παθήματα ἐόντα ἀχάριτα μαθήματα
γέγονε, strong emphasis is laid on the antithesis between ἐόντα ἀχάριτα and
μαθήματα γέγονε.

After the epode of the opening triad (159) there is a sharp break. Instead
of the rolling iambo-dactyls we hear trochees. The narrative with its luxuri-
ant details, delivered in a language both artificial and cryptic, is succeeded
1 1 suspect that both these passages of the Supplices and Eum. 530 are influenced by
Hesiod, Erga 483 f. ἄλλοτε δ᾽ ἀλλοῖος (cf. Eum. 530 ἄλλ᾽ ἀλλᾶι δ᾽ épopevet) Ζηνὸς νόος αἰγιόχοιο,
ἀργαλέος δ᾽ ἄνδρεσσι καταθνητοῖσι νοῆσαι,
112
 THE HYMN TO ZEUS
by a concise hymn of great lucidity and simple, if powerful, structure. Most
startling of all, the account of what happened in Aulis ten years ago is
broken off immediately before reaching its climax. Abruptness, then, there
is, but no loose arrangement, no wavering in the progress of thought either
here or anywhere else in this great chorus. For when, after the conclusion
of the hymn to Zeus,! the narrative is taken up again (184), it is made clear
by the connective particle (καὶ τότε) that what happened to Agamemnon is
a case in point, a παράδειγμα illustrating the sovereign power of Zeus over
men and the manner in which the god leads through suffering to wisdom.
But only gradually, as the song goes on, does the connexion of thought
reveal itself. When the first notes of the hymn strike the hearer’s ear, they
seem to come as an interruption rather than as a continuation. The effect
of this seeming harshness is intense. In the last sentences of the seer’s
speech there was (150 ff.) a clear allusion to Artemis demanding the sacrifice
of Iphigeneia. Thus the account of the events in Aulis has been carried to
the threshold of the insoluble dilemma with which Agamemnon will be faced.
A point of utter ἀμηχανία has been reached.” There the chant leaves the story,
at any rate for the moment, and turns to him who alone, in such a conflict,
is capable of relieving man’s mind from the burden of idle speculation, Zeus.
The praise of Zeus begins in ἃ manner that recalls solemn formulae of κλητικοὶ
ὕμνοι and prayers (cf. on 160), but voiced as it is here, it is no prayer. Zeus
is not addressed but spoken of in the third person only. In other tragedies,
e.g. the Suppliants, the Seven against Thebes, King Oedipus, the background
of the first song is the presence of a calamity under which the Chorus are
suffering. It is therefore natural for them to pray to the gods and implore
their help. The long lyrical part of the parodos of the Agamemnon, save the
last stanza (248 ff.), is wholly concerned with events that took place ten
years before. The vivid recollection of the terrible choice with which the
king had been faced produces in the minds of the elders an acute feeling of
perplexity and helplessness; hence their thoughts are directed towards the
omnipotent overlord. Since their impulse does not arise from any immediate
emergency, there is nothing they could pray for. Instead of asking for help
they glorify Zeus, attempt to fathom his nature, and describe the way in
which his will works. Thus the uncommon attitude of the hymn is rooted
in the peculiar conditions of this chorus-song as a whole. At the same time
the poet obtains an opportunity of going beyond anything that could be
expressed in a prayer. On many occasions prayers are justified; Aeschylus
would be the last man to deny or minimize their value. Yet when, at the
stage of his maturest speculation, he endeavours in a sublime effort to un-
riddle the ultimate cause of the fate and suffering of man, any form of
beseeching and imploring would be inadequate. Zeus has established his
eternal law, he has shown us the way to understanding, discretion, wisdom.
All that matters is that we should learn the hard lesson he wants to teach us:
πάθει pabos. Once we have failed, no sacrifices, no prayers will be of any avail

1 We cannot say whether it is intentional or mere accident that the gliding transition at
184 goes together with a continuation of the same metre whereas the fresh start at the
beginning of the hymn is marked by the different rhythm.
2 Cf. Kranz, Hermes, liv, 1919, 305; B. Snell, Aischylos, 48.
3 Cf. Philol. Ixxxvi, 1931, 14 ff.
4872-2 I 113
line 184 COMMENTARY
(cf. 68 ff.) ; we shall have to pay the full price for our trespasses. It is not
happiness, not even forgiveness, that the powerful god who rules by force
grants to mortals: at the end of the hard path there is nothing but φρονεῖν."
In the first stanza of the hymn προσεννέπω and ἔχω must of course be
understood of the elders of Argos. But there is no doubt that far more is
included in that ‘I’. The Chorus speak also for the citizens of Athens, to
whom they belong, and above all for the poet. It would be quite wrong
to assume that the hymn does not form an organic part of the surrounding
narrative (see above). But it is true that the culogy of Zeus is intended to
be valid beyond the limits of any particular situation. This widening of the
sphere is supported by the abruptness at the beginning (160) ; after that we
do not expect that the song will move on the same lines as before.
The hymn is a corner-stone not only of this play but of the whole trilogy.
Its connexions with the rest of the Agamemnon will be noticed in the com-
mentary. As regards the links with the Choephoroe and the Eumenides,
they could be duly appreciated only in a full interpretation of those plays.
Here I must be content with two hints. When Agamemnon’s murderers
have met with their punishment, the Chorus sing (Cho. 948 ff.) ἔθιγε δ᾽ ἐν
μάχαι χερὸς ἐτητύμως Διὸς κόρα' Δίκαν δέ νιν προσαγορεύομεν βροτοὶ τυχόντες
καλῶς. And throughout the Eumenides the poet takes great pains to make it
clear that both Apollo and Athena are the fulfillers of their father’s will and
that the final sentence reached by the Areopagites is perfectly consistent
with the ordinances of the supreme god.

184. καὶ τότε. With καί ‘the special case is ranged under the general law’
(Wecklein, Studien zu Aesch. 106). ‘So then . . .' (Paley). Similarly in the
following ode a καί serves to make the transition from a series of general
maxims concerning the law of crime and punishment to the special case
(399 ff.) : οἷος καὶ Hápis . . . ἤισχυνε κτλ. The whole structure there is parallel
to that of the ode before us. The anapaests 362 ff. recall the punishment
with which Ζεὺς ξένιος against whom Alexandros had offended visited him
and his house; the stanzas from 367 ff. onwards are concerned with Zeus’
inescapable power to execute justice and what happens when a man falls
into sin (this section? corresponds to the ‘Hymn to Zeus’ in the parodos) ;
finally the ἐφύμνιον 399 with οἷος καὶ Πάρις returns to the starting-point. So
here καὶ τότε clearly refers to the events of which we were told before the
hymn began. It is therefore wrong to deny to this τότε its function as a
reference back and to regard it as merely a support for the following εὖτε
(188) or the τότε which many editors read in 205.
ὁ πρέσβυς ‘the elder’ of the two, as in 205, 530. This use appears not to be
found elsewhere.
ı What D. 8. Robertson has said of Pindar’s attitude, that it is ‘inexpressibly Greek in
its dignity and in its refusal to accept any false consolation’, holds good of this Aeschylean
hymn.
É From 369 on the individual case recedes into the background, and the treatment is
gnomic. Surveying the entire context of the parodos, one might describe the Hymn to
Zeus as a monumental γνώμη.
3 This transition from the γνώμη to the narrative, or to the events with which we are
actually concerned, corresponds to an early form, cf., e.g., 0 329 f. οὐκ ἀρετᾶι κακὰ ἔργα" κιχάνει
τοι βραδὺς ὠκύν' ὡς kai νῦν Ἥφαιστος ἐὼν βραδὺς εἷλεν "Apqa κτλ.
114
 COMMENTARY line 188
From 184 onwards Agamemnon takes the foremost place (Ferrari, La
parodos, 382).
186. μάντιν οὔτινα. Schol.: περισσεύει τὸ τινα. But the weakened use of
nullus = non (cf. Wackernagel, Syntax, ii. 67, and J. B. Hofmann, Lat.
Umgangssprache, 80) has no analogy in Greek (on S. El. 276 cf. Kaibel, ad
loc.). Thus οὔτινα retains its full force: ‘he upbraids no seer’, says the poet,
‘embracing this particular case under the general rule’ (Verrall). In other
words: Agamemnon does not behave as others in corresponding circum-
stances usually behave towards the seer concerned, e.g. Agamemnon in the
Iliad (A 106): μάντι κακῶν, où πώ ποτέ μοι τὸ κρήγυον εἶπας, or Hector with
Polydamas (M 230 ff.).
187. ἐμπαίοις : ἃ similar conception is found in 347 εἰ πρόσπαια μὴ τύχοι κακά,
cf. also 5. El. 902 f. ἐμπαίει τέ μοι ψυχῆι σύνηθες ὄμμα. The word ἔμπαιος is
recorded elsewhere only from Empedocles Β 2. 2, where Emperius saw that
ἔμπεα is ἔμπαια (for the spelling e for αἱ and vice versa cf. on 1653).
ouumvewv: the verb appears to occur here only in Greek before Plato.
But we have just had (147) ἀντιπνόους, and it is certainly no accident that in
the very place (187) where the thought goes back to the opposing winds
which are the bearers of the first, and causes of the later, ἔμπαιοι τύχαι, it is
the ‘metaphor’ of συμπνέων that comes into the poet's mind. The imagery
in itself is readily understood; it belongs to the same sphere as 219 φρενὸς
πνέων δυσσεβῆ τροπαίαν, cf. note ad loc. What we are told is that Agamemnon
did not set his face against the winds of fortune encountering him, but let
himself be carried in the same direction as they. Taking as his starting-point
familiar passages in Homer, O. Becker, 'Das Bild des Weges', Hermes,
Einzelschriften Heft iv, 1937, 169 ff., shows how 'the direction of the will that
is resolved upon action is seen [by the Greeks] through the imagery of the
wind’.
188. εὖτ᾽ ἀπλοίαι κτλ. calls afresh to our remembrance what was fearfully
foretold in 147 ff. (150 ἀπλοίας); what was then a source of apprehension
is now actuality (Ferrari, La parodos, 384).
kevayyei: the available evidence is well set out in Blomfield’s glossary ; he
rightly opposes those commentators who see in the word a reference to the
emptying of jars of provisions (so again E. Williger, Sprachl. Unters. 18).
There is no case for referring the element dyyos here to anything other than
the ἀγγεῖα of the body, seeing that xevayyia (or -in) means ‘hunger’ not only
in Ionic but also in everyday Attic, as is clear from the passages of Comedy
cited by Pollux, 6. 31. Thus κεναγγής means ‘suffering from hunger’. The
expression closely corresponds to the following (192 f.) πνοαὶ... νήστιδες.
Every Athenian of any military experience was familiar with these results
of ἄπλοια, and with the menace they represent; thus in Thuc. 6. 22 Nicias

1 TINA was misread as ΠΝΑ͂, the very early compendium for πνεῦμα (L. Traube,
Nomina Sacra, 42), as E. Kueck, Studia . . . in Aeschylum (Diss. Göttingen, 1890), 41, and
Ileadlam saw.
2 That Aeschylus had this passage in mind when he described the opposite behaviour
of his own Agamemnon is practically certain; we cannot tell whether he found in the line
a special reference to Calchas’ demand that Iphigeneia should be sacrificed, as Zielinski
does (Tragodumenon libri tres, 1925, 243). Any reader of Homer may have felt tempted
to connect the μάντι κακῶν passage with the scene in Aulis; the BT scholia ad loc. think it
necessary to give a special warning against this: τὸ yàp ᾿Ιφιγενείας ὄνομα οὐδὲ οἶδεν ὁ ποιητής.
115
line 188 COMMENTARY
bases his request for a surplus of ships on the necessity of securing his com-
missariat, tva ἣν που ὑπὸ ἀπλοίας ἀπολαμβανώμεθα, ἔχηι ἡ στρατιὰ τὰ ἐπιτήδεια.
190. πέραν. Here, too, Blomfield in his Glossary is right, in saying that we
have the accusative of πέρα and referring to Suppl. 262 ἐκ πέρας Ναυπακτίας
(Wilamowitz emended his note in the ‘Corrigenda’, p. 382, of his edition ;
he had already given the right answer in his note on E. Her. 234). On the
other side, Schütz and Hermann (on 1. 697 of his numbering) took πέραν
as an adverb and consequently ἔχων as intransitive ; so did the author of the
marginal gloss (in M) ἀντὶ τοῦ ἐχόμενος. But for the latter usage, though it
is of course well known (e.g. Pind. P. 1. 72 ὄφρα κατ᾽ οἶκον ὁ Φοίνιξ ὁ Τυρσανῶν
T' ἀλαλατὸς ἔχηι), there seems to be no example in Aeschylus; at any rate
nothing can be made of the two instances adduced besides this passage by
Dindorf, Lex. Aesch. 145 left-hand column, (1) Sept. 102, which Dindorf
misunderstands like Tucker and L-S p. 750 B. i. 4 (the right explanation is
given in Wecklein’s commentary and by Wilamowitz, Interpr. 70), and (2)
Ag. 723 (v. ad loc.).
191. παλιρρόχθοις : this restoration by Ahrens is generally accepted. On the
‘powerful currents that change their direction repeatedly in the course of
the day’ in the Euripus cf. Philippson, RE vi. 1282 f. The phenomenon was
celebrated in antiquity (Philippson gives the references); the reasons for it
have only been explained in modern times.
190 f. Χαλκίδος πέραν ... ἐν Αὐλίδος τόποις. It is characteristic of Aeschy-
lus’ way of telling a story, and of early narrative technique in general (cf.
on 59), that no definite geographical place is indicated until the story has
reached the point at which the counter-currents in the Euripus and the north
wind, which is there particularly disastrous, have become of immediate
importance. But of course the audience, being familiar with the events,
must have identified the locality long before the Chorus expressly mentioned
the names of Chalcis and Aulis.
192. ἀπὸ Στρυμόνος: the same winds are described in 1418 as Θρήικια ἀήματα.
The reference is to the strong NNE. wind, Boreas, known to the inhabitants
of the north Greek coasts as ‘EAAnorovrias (Hdt. 7. 188. 2). Before the battle
of Artemisium in the year 480 the Athenian fleet lying before Chalcis saw
a storm getting up from the north (Hdt. 7. 189. 2), which naturally raged
much more unrestrainedly on the Magnesian coast where the Persian fleet
lay. This fateful occurrence may have deepened Aeschylus’ conception of
the violence of Boreas in the Euripus, and the foundation of the shrine of
Boreas as a result of it is likely to have been in the thoughts of the poet
of an Oreithyia.!
193. νήστιδες. It is customary to speak of an 'active' use, but it seems more
appropriate here as with other privative expressions (cf. on 238) to recognize
the neutral use, which should not be related to any particular voice of the
verb: ‘that which is characterized by the lack of anything to eat.’
! It is worth considering the suggestion made by Welcker, D. griechischen Tragödien,
iii. 1502 f., then hinted at by Furtwängler (in Furtwängler-Reichhold, ii, Text p. 189) and
C. Robert, D. griech. Heldensage, 169, and adopted by W. Schmid, Gesch. der griech. Lit.
ii (1934), 203 n. 9, that Aeschylus was moved to select the subject of his Oreithyta by the
founding of the temple of Boreas. Cf. the observations of Wilamowitz, Sappho und Simo-
nides, 207, on the tribute paid to Boreas by Simonides in the poem quoted as ἡ ἐπ᾽ ᾿ἀρτεμισίωι
#
ναυμαχία.

II6
 COMMENTARY lines 197 f.

övoopuos here is used in a different sense from Pers. 448 and elsewhere,
and must mean ‘in portu male detinens’ (Blomfield). Cf. also Schuursma,
159 f.
194. βροτῶν ἄλαι: the expression has often in recent times been doubted
or altered (cf. G. Thomson), without adequate reason. The scholia (in M)
quote 4 330 καὶ δὴ ἄγρην ἐφέπεσκον ἀλητεύοντες ; the preceding line u 329
(ἀλλ᾽ ὅτε δὴ) νηὸς (in the scholion νηῶν, to adapt it to the context of our
passage) ἐξέφθιτο Yia πάντα is cited in the scholion on 188 ἀπλοίαι kevayyet.
It is sensible to compare the ἀλητεύειν of the companions of Odysseus, who
lay a whole long month debarred from continuing their journey by the storm
(μ 325 ff.) so that their victuals ran out (the related passage 8 368 f. is referred
to by van Heusde) ; only, when thinking of the Greek host in Aulis, we must
not limit ἀλᾶσθαι too narrowly to the purpose of foraging. The crews wander
to and fro on land, partly looking for provisions, partly because they have
no serious occupation. Their discipline suffers no less than the ships and their
tackle. Aeschylus was an old soldier and all this was familiar to him. To
understand ἄλαι here of distraction of spirit is quite arbitrary. βροτῶν du
is placed immediately in apposition to mvoai, as though he said ‘inability
to sail, that means men wandering about’; this is forceful and no objections
should have been raised. To explain it by saying that ἄλαι is here used ‘in an
active sense’ (‘Irrefiihrung’ is appended in Passow-Crônert) adds nothing
to our understanding.
196. παλιμμήκη: this is a ἅπαξ λεγόμενον, like several words in this part,
but the meaning is quite clear and vivid: when they have waited a long time
and think it must all be over now and they can set out, they are thrown
‘back’ again and the trouble begins afresh.
197. τρίβωι: the interpretation διατριβῆι given in the paraphrase of the
scholia is slightly varied by Hermann: ‘rpißax exquisitius dictum pro rpiBije.’
He is clearly thinking, as his rendering mora shows (Opusc. v. 347), Of the
meaning ‘delay, postponement’ which τριβή or τριβαΐ not seldom has. This
interpretation has been accepted by many: Schneidewin supports it by a
reference to the interchange observed by Lobeck between βίοτος and βιοτή,
φθόγγος and φθογγή, and so on. This view is certainly preferable to Blom-
field’s, who assumes here the ordinary sense of ‘way, path’, etc., for he is
contradicted by the passages which he adduces as parallels (always the
material way is meant, the actual track, never ‘journey, expedition’). But
there is no reason to make Hermann’s assumption of a special meaning. The
sense of τρίβος is the same here as in 391. Klausen and Paley are right;
the latter translates ‘began to wear out by waiting’ and quotes Thuc. 7. 42. 5
(reflections of Demosthenes on the army besieging Syracuse) οὐ τρίψεσθαι
ἄλλως, cf. 6. 18. 6 τὴν πόλιν, ἂν μὲν ἡσυχάζηι, τρίψεσθαί τε αὐτὴν περὶ αὑτὴν
ὥσπερ καὶ ἄλλο τι, καὶ πάντων τὴν ἐπιστήμην ἐγγηράσεσθαι. Cf. below, p. 479 ἴ.
It is just this destructive attrition of the natural faculties as an effect of
idleness (here made inevitable by the storm) that is the main point of τρίβωι
κατέξαινον ἄνθος Apyelwv. The verb karafaivew is now to be found also in
the fragment of the Myrmidons of Aeschylus, Pap. Soc. It. xi. 1211. 2 (= D. L.
Page, Greek Lit. Pap. i. 138), πέτροις kara£avÜévra.
197 f. Denniston (in Greek Poetry and Life, 1936, 129) observes that Wilamo-
witz ‘produces an unbroken series of metra, but at the cost of ignoring the
117
lines 197 f. COMMENTARY
strong stop in str. and ant. after 197 == 210'. It is true that after Ἀργείων
and πέλας βωμοῦ (on the transposition see below) there is a full-stop. But
should that induce us here to ignore the continuous series of iambics? This
is not the place to demonstrate that the coincidence of syntactical and
metrical units, which is often maintained and not least by Aeschylus (cf.,
e.g., on 381-4), is far from an absolute rule, and that in this respect different
metres are very differently treated. It will suffice here to show by two
examples of a full-stop in lyric iambics in the Oresteia that no such general
principle existed: Ag. 369 ἔπραξεν ὡς ἔκρανεν. οὐκ ἔφα τις — 387 ἄκος δὲ πᾶν
μάταιον. οὐκ ἐκρύφθη, Cho. 45 ff. ἰὼ γαῖα μαῖα, μωμένα μ᾽ ἰάλλει δύσθεος γυνά.
φοβοῦμαι δ᾽ ἔπος κτλ. = 56 ff. 8U ὥτων φρενός τε δαμίας περαῖνον νῦν ἀφίσταται.
φοβεῖται δέ τις κτλ.
198. ἐπεὶ δὲ καὶ κτλ. : kac here needs an explanation. It cannot (with Schneide-
win) be referred to πικροῦ. One might think of connecting kai ἄλλο μῆχαρ,
but it is perhaps better to take this initial καί as intended to give emphasis
and to refer it to the whole following clause, ‘when he in fact . . .’, cf. on 558.
ἄλλο μῆχαρ. Several commentators here assume the well-known pleonastic
use of ἄλλος (thus Nägelsbach translates ‘ein Anderes, ein Mittel’), but their
parallels from Homer (B 191, ζ 84, ı 367) are all quite different, and the same
is true of the passages collected by E. Bruhn, Anhang zu Sophokles, § 182.
More help is to be had from Paley’s explanation: ‘It is implied that some
remedies had been recommended, tried, and found to fail, before this last
and terrible resource was enjoined.’ That would well suit the allusive style
of this lyrical narrative, which selects and concentrates on the principal
points.
200. πρόμοισιν. Whether or no πρόμος really arose as a shortening of
πρόμαχος, as Hentze following ancient authorities assumed (cf. W. Schulze,
ΚΙ. Schr. 310 ; Bechtel, Lexilogus, 285), the two words are in any case synonym-
ous in Homer, as Aristarchus saw (Lehrs, De Aristarch stud. Hom. 101 f.).
In Tragedy, however (first in A. Suppl. 9os, a certain restoration), it is
used merely for ‘leader, captain, lord’ (in Ag. 410 precisely equivalent to
dominus). We may at least wonder whether Aeschylus was familiar with an
interpretation of the Homeric passages like the one rejected by Aristarchus
(cf. Lehrs, op. cit. 37 f.), schol. A on I" 44: πρόμον" ἡ διπλῆ, ὅτι κατὰ συγκοπὴν
τὸν πρόμαχον εἴρηκεν, οὐχ ὡς ol γλωσσογράφοι τὸν βασιλέα.
201. ἔκλαγξεν: cf. on 156 ἀπέκλαγξεν. Since the initial mp cannot lengthen
the preceding syllable (cf. Appendix E), Porson's éxAay£e(v» is necessary.
προφέρων does not strike us as unusual, because we are accustomed to
‘bring forward’, ‘vorbringen’, in the sense required here. In Greek, however,
this use is not very common, especially in poetry. We should be more
correct in recognizing the same technical meaning, to 'make known’ an oracle
and the like, which is familiar to us from Herodotus, and also found, e.g.,
in Demosthenes. This is undoubtedly the sense in the only other passage
where the verb is recorded from Aeschylus, Ag. 964 (where Paley is right).
Cf. Stein on Hdt. 5. 63. ı and Wackernagel's observations (Syntax, ii. 238 f.)
on the primary meaning of προειπεῖν, προαγορεύειν, etc. (— edicere).
202 ff. ὥστε χθόνα... μὴ κατασχεῖν : the concise indications of overmastering
emotion must be seen as part of the same picture with the μάντιν οὔτινα
ψέγων of 186. It is precisely the passion here described which makes it possible
I18
 COMMENTARY line 205

to see the moderation of Agamemnon’s behaviour towards Calchas in its

true light: he is shown as a great and noble man, ἃ μεγαλόψυχος ; as such he
appears afterwards in the only scene in which he is on the stage. It is well
worth noticing how the Atridae in Aeschylus do not give full vent to their
passion in the same unbridled way as Homer’s Achilles, who in his dispute
with Agamemnon (A 245) ποτὶ... σκῆπτρον βάλε yaim. Aeschylus, who had
that very scene in his mind (cf. on 186), toned down the roughness of the
detail; we see the effect of Attic εὐσχημοσύνη.
205. ἄναξ δ᾽ ὁ πρέσβυς κτλ. Formally this is the apodosis of 198 ff. ἐπεὶ δὲ
καὶ πικροῦ χείματος κτλ. The scholiast rightly says: τὸ ἑξῆς (= ordo est):
ἐπεὶ δὲ πικροῦ, ἄναξ ὁ πρέσβυς τόδ᾽ εἶπεν. For the well-known use of δέ in the
apodosis after ἃ temporal protasis cf. Denniston, Particles, 179. The almost
verbal repetition of the phrase at the beginning of the last stanza but one
(184) καὶ τόθ᾽ ἡγεμὼν 6 πρέσβυς makes it easy for the audience to realize that
the sentence there left unfinished, which has almost slipped from our memories
through the flood of new images that has come between, is here taken up
again and at length is given its main verb. The ‘ring-composition’ or ‘con-
centric’ composition! of the narrative that runs from 184 to 205 can be made
clear by a time-scheme on the following lines. The latest point in the order
of events is the beginning of Agamemnon’s speech, which is already in view
at 184 καὶ τόθ᾽ ἡγεμὼν κτλ., although εἶπε is not introduced till 205. This
speech of Agamemnon’s results from the revelation of the seer which in
186 with μάντιν οὔτινα ψέγων is implied rather than expressed. The things
through which the seer’s utterance is called forth, namely the ἔμπαιοι τύχαι
(187), and more accurately the hardships caused by the dmAota (188-91), lie
a stage farther back. The ἄπλοια in its turn is caused by the unloosing of
the north wind (192). We have now reached the most distant point in the
narrative, which turns round and in the attributes of 193 ff. (κακόσχολοι κτλ.)
again describes the ἄπλοια with its concomitants; then (197) indicates their
effect on the army; goes on to the announcement (198-201) by Calchas of
a dreadful antidote, but this time in greater detail; next describes, also in
greater detail than before, how the speech of Calchas affected the Atridae
(for the relation between 202 f. dore χθόνα κτλ. and 186, see on 202) ; and finally
with 205 returns to its starting-point.
The purpose of this highly artificial arrangement will be fully grasped as
soon as we consider the special object it serves here. Immediately after the
hymn to Zeus, at 184, there was no room for a report on the ὑποκείμενα of
Agamemnon’s speech. For there the poet’s main concern was to make it
clear that the way of feeling and behaving to which man is bound by an
insight into the nature of Zeus’ almighty power is exemplified in Agamem-
non’s resolution to take upon himself the sacrifice of his daughter, in full
recognition of all the pain that must inevitably follow such a step. Conse-
! On this arrangement: of a poetical narrative see especially L. Illig, Zur Form der
pindarischen Erzählung (1932), 56 ff. W. A. A. van Otterlo, Beschouwingen over het arch-
aische element in den stijl van Aeschylus (Diss. Leiden, 1937), deals at considerable length
with ‘ringcompositie’ (pp. 18 ff. and 76 ff.), but he uses the term in a different sense, viz. to
denote the device by which at the end of a speech or a section of a speech the introductory
idea is repeated, as, e.g.. A. Suppl. 407-17; cf. above on 20. I have not been able to see van
Otterlo’s paper ‘Untersuchungen über Begriff . . . der griech. Ringkomposition’, published
by the Academy of the Netherlands in 1944.
119
line 205 COMMENTARY

quently, when the glorification of Zeus has been brought to a close, the ode
makes straight for the speech in which the king announces his decision.
But in order to make that speech intelligible the audience must be given,
even in this narrative of hints and brief touches, at least the essential
ὑποκείμενα; and the only remaining way to do this was to start from the
moment immediately before the king’s declaration and then work backwards.
At the same time, by means of the artifice of continuing at 184 with καὶ τότε
before he has yet explained in detail what lies behind it, the poet makes it
possible to give this τότε a very welcome ambiguity or at any rate uncer-
tainty. The temporal adverb has the function, after the interruption caused
by the praise of Zeus, of making contact again with what preceded and
‘continuing to a certain degree (“einigermassen’’) the account of events at
Aulis’ (Wilamowitz, Interpr. 166 n. 2). It does fulfil this function, but only
‘to a certain degree’, for although the hearer is carried back to the plane
of what he has already been told, ra ev Αὐλίδι, he is not taken to the parti-
cular point reached just before the hymn, viz. the speech of Calchas which
ends at 155. What is actually meant by that τότε becomes plain only after the
supplementary account of the ὑποκείμενα (188-204) has been given, and we
then realize that since the events narrated just before the hymn to Zeus,
another series of important events has taken place, among them a new
speech by Calchas ;' we may say that the action has been moving on while
we were listening to the praise of Zeus.
205. τόδ᾽ εἶπε. Stanley: ‘Malim τότ᾽, ut respondeat τῶι ἐπεὶ v. 198”, accepted
by Hermann on more adequate grounds: ‘haec verba quum aperte repetant
quod v. 184 dici coeptum est, καὶ τόθ᾽ ἡγεμὼν KrA., non τόδε sed τότε dici
postulant.’ But this, too, is not convincing, for ἄναξ δ᾽ 6 πρέσβυς by itself is
quite enough to make clear the resumption of 184. And when Hermann in
support of the conjecture adds ‘pronomine si uti voluisset poeta, credibilius
est τάδε dicturum fuisse’, that is not so, cf. 408 f. πολλὰ δ᾽ ἔστενον τόδ᾽ ἐννέποντες
δόμων προφῆται. |
εἶπε φωνῶν. Examples of this ‘pleonasm’ are given in Lobeck’s commentary
on S. 47. 757. He also points out the solemn character that such turns of
phrase sometimes have, as does Schneidewin in his note on our passage. A
characteristic example is Cho. 279 (Apollo) πιφαύσκων εἶπε. Cf. also Ennius,
Ann. 44 1. Vahlen exim compellare pater me voce videtur his verbis and similar
passages discussed by E. Lófstedt, Syntactica, ii. 185 f.
206. κήρ: ruin, grievous fate. The gradual development of this idea out of
the Homeric κὴρ θανάτοιο has been sketched by Wilamowitz, Glaube der
Hellenen, i. 270 ff. For the earlier, more strictly religious view cf. F. Jacoby,

! It is hard to understand that the clear evidence could have been distorted. R. Hölzle,
Zum Aufbau der lyrischen Partien des Aischylos (Diss. Freiburg i. Br. 1934), 99, 109, asserts
that the dora with its grave consequences has already been lasting for some time when
the sign of the eagles appears. According to him, the utterance of Calchas referred to at
198 ff. forms part of the speech which is recalled in 126-55. Daube, Rechtsprobl. 169 n. 14,
is impressed by Hölzle’s assertion ; Regenbogen, Gnomon, xv, 1939, 603 f., rightly rejects it.
There can be no doubt that 148 ff. μὴ . . . ἀπλοίας τεύξηι refers to the future. The sequence
of the events is clearly indicated: (1) prodigy of the eagles sent by Zeus; (2) speech of
Calchas, 126-55; (3) beginning of the ἅπλοια caused by Artemis; (4) another speech of
Calchas (summarized in 198 ff.); (5) Agamemnon's compliance with the will of the gods
and the sacrifice of Iphigeneia.
120
 COMMENTARY line 210

‘Theognis’, Berl. Sitzgsb. 1931, 101 n. 1; a survey of the occurrences in Tragedy

is given by Malten, RE Suppl. iv. 890.
The beginning with βαρεῖα «np is significant: the very first words of
Agamemnon emphasize the gravity of his deed. What Wilamowitz, Vers-
kunst, 181, says with regard to the words 209 f. juaivwv . . . πέλας βωμοῦ,
‘sentimus horrorem patris in taetri sceleris imagine commorati’, is true of
this speech as a whole. Almost every phrase that is here applied to the
sacrifice of Iphigeneia conveys the idea of something utterly unnatural and
horrible. Nowhere does the speaker attempt to make the crime appear in a
milder light. It is in merciless terms that Agamemnon describes, and im-
plicitly condemns, what he is going to do. And yet he cannot avoid doing it.
On this antinomy see below, p. 146 f.
208. Saigw. The verb δαΐζειν (‘mangle, rend asunder, hew in pieces’; on the
etymology cf. W. Schulze, Quaest. ep. 380) here serves to emphasize the horror
of the deed, which is brought out still further in what follows.
ἄγαλμα: Apollonius Soph. p. 6. 30 (further evidence is given by Ed.
Schwartz on Schol. E. Hipp. 631, to which add Schol. Arist. Wasps 313)
ἄγαλμα’ πᾶν ἐφ᾽ ὧι τις ἀγάλλεται. This, καὶ οὐχὶ τὸ ξόανον, as Apollonius adds,
is in Homer the only meaning of the word (cf. K. Lehrs, Kl. Schr. 60). On
the sense of the word here and in similar passages cf. Wilamowitz on E. Her.
49, and, for a comprehensive view, see H. Bloesch, Agalma (Bern-Bümpliz
1943), 12 ff. j
210. ῥείθροις. The MSS give ῥεέθροις, just as in 5. Ant. 1124 they give
ῥέεθρον. There, too, the metre peremptorily demands a disyllabic form (the
editors’ opinions differ only with respect to the termination). The word itself
in our passage has often been attacked, from Enger and Schoemann down to
Headlam. In this connexion, the responsion 197 rpißwı = 210 ῥείθροις is
constantly pointed to as objectionable, but wrongly. It is true that in lyric
iambics in Aeschylus correspondence syllable by syllable is far the commonest,
but the licence here objected to is by no means unusual. In the following
survey the acatalectic trimeters to be found among lyric iambics are omitted.
The letter : following the line-number of the strophe indicates that the free
responsion is found at the beginning of a metrical colon, as in Ag. 197 = 210.
We find v- — —— in Suppl. 139i πατὴρ 6 mavr. = 149 ἀδμῆτος ἀδμ., 374
θρόνοις χρέος = 385 Ἱκταίου κότος, 540 i λειμῶνα osx. = 549 περᾶι δὲ Τεύθ.,
560 ὕδωρ τὸ Νείλ. = 570 πᾶν δ᾽ αὖ γυναικ., 576 Ἱμάντωι σθένει = 584 γένος τόδε,
794 i ἢ λισσὰς aly. = 802 τὸ γὰρ θανεῖν, Pers. 10141 πῶς δ᾽ οὐ στρατόν = 1026
ἄγαν ἄρει., Sept. 287 i μέλει, φόβωι — 304 motor! δ᾽ ἀμείψ., 740 1 πόνοι δόμων =
748 χρηστηρίοις, 834 i κακόν με kapô. = 842 βουλαὶ δ᾽ dmor., 890 Ἰσπλάγχνων re
πλευρ. = 902 πέδον φίλανδρ., 912 i σιδηρόπληκτ. = 923 ws ἐρξάτην, 968 1 ἐντὸς
δὲ καρδ. = 979 ἔδειξεν (ἐδείξατ' codd.) ex, 970 i σὺ δ᾽ αὖτε καί = 981 σωθεὶς δὲ
πνεῦμ᾽. And this interchange of a short and a long syllable in the first place
of the iambic metron is found not only in the earlier plays but also in the
Oresteia, a thing which should not be supposed a matter of course, in view of
the greater strictness with which the responsion in lyric iambics is observed
1 If we are to be completely cautious, we must reckon with the possibility of the scansion
vu — for «otov; but it is not likely here, cf. below on 1256 f. On the other hand, in Cho. 42 i
τοιάνδε χάριν = 55 σέβας δ᾽ ἄμαχον, 428 i κροτητὸν du. = 450 τοιαῦτ᾽ ἀκού., the shortening of
the first syllable of τοιάνδε, τοιαῦτα would be perfectly normal.
121
line 210 COMMENTARY
in this late work (cf. Ρ. 451). To Ag. 197 i = 210 we may add 447 1 τὸν δ᾽ ἐν
φοναῖς = 466 τιθεῖσ᾽ ἀμαυρ., 11631 νεογνὸς ἂν d. = 1174 καί τίς σε κακοφρ.,
15331 δέδοικα δ᾽ ὄμβρ. = 1563 μίμνει δὲ μίμν., Cho. 24 φοίνισσ᾽ ἀμυγμ. = 34
]rov ἀμβόαμ., Eum. 383 τε μνήμονεςϊ Σεμναί = 391 ἐμοῦ κλύων θεσμόν, 550 1 ἐκ
τῶνδ᾽ (the MS reading is correct, cf. the note on Ag. 1223) ἀνάγκ. = 558 καλεῖ
δ᾽ ἀκού. Further, vus — —vv in Suppl. 135 re δόμος ἅλα = 145 Ἰγνά μ᾽
ἐπιδέτω, Cho. 2 5 Ἱμοῖς ὄνυχος
à dAok. — 35 μα μυχόθεν ἔλακ., 423 Ἱμὸν Ἄριονἔν =
444 ]ov μόρον. ἐγώ, 425 τα πολυπάλακτ. = 446 ]ros πολυσινοῦς. In Prom. 181,
too, as P. Maas has pointed out to me (see his Griech. Metrik, § 109), it will be
right to retain ἠρέθισε Siar. (= 162 ye Διός; 6 δ᾽ ἐπικ.) in this form as Hermann
does (though he arbitrarily alters the strophe), for the omission of the
temporal augment would hardly be permissible in Drama (cf. Kühner-Blass,
ii. 18) ; no support can be obtained from Pers. 915, where ὄφελεν isi borrowed
from a familiar epic formula and as such found elsewhere in Tragedy (cf.
L-S s.v. ὀφείλω ii 3 a).
As regards the word ῥείθροις, we must admit that it is an unusual expres-
sion ; this was pointed out by Hermann, and with his comment and Wilamo-
witz's note 'insolitum hoc de sanguinis effluvio; auget crudelitatem" all has
been said that needs saying. On the savagery of the expressions in this
sentence cf. on 206 and 208. The bold noun ῥείθροις is defined by παρθενο-
σφάγοισιν : to this peculiar collocation we shall do justice not so much with the
rendering 'streams of a slaughtered maiden’s blood’ (L-S) as with 'streams
(of blood) that flow at the slaughter of a maiden'. So E. Williger, Sprachl.
Untersuchungen, 34, who deals instructively with the verbal adjectives in
whose ‘verbal element a verbal abstract lies concealed'.
210. Blomfield's πέλας βωμοῦ in place of the MS reading βωμοῦ πέλας is ἃ
much easier change than the other proposals. The word-order of the MSS
can be explained by the observation that 'the genitive commonly stands
before πέλας ᾿ (L-S): the scribes here as in countless other places have fol-
lowed their preference for ‘simplex ordo' (Headlam, C.R. xvi, 1902, 244 ff.,
cf. below on 395). No further correction is needed, here or in the correspond-
ing line in the strophe. The punctuation after βωμοῦ corresponds exactly
with the stop after 198 Apyeiwv; cf. on 122 f.
211. τί τῶνδ᾽ ἄνευ κακῶν; A fundamental motif of the trilogy, cf. Cho. 339 τί
τῶνδ᾽ εὖ, τί δ᾽ ἄτερ κακῶν;
212. πῶς λιπόναυς κτλ. For the understanding of the whole it is decisive
that the answer implied in this ‘rhetorical’ question can only be ‘impossible’.
Gone is the ἐποχή which expressed itself just now (206 f.) in the balancing of
two evils equally intolerable (βαρεῖα μὲν κὴρ . . . βαρεῖα de): he has only two
possible lines of conduct, and the moment he calls one of them by its true
name, λιπόναυν γενέσθαι, his mind is made up and he is resolved to sacrifice
his daughter ; this is the point at which ἀνάγκας ἔδυ λέπαδνον. It is no ordinary
meditative individual that here speaks of deserting, but the king of the host,
Agamemnon. ‘Il balance en père et se détermine en roi’ (le Père Brumoy,
Theätre des Grecs, edition of 1785, 11. 6).
λιπόναυς: in the interlinear gloss in Tr (πῶς καταλείψω τὰς ναῦς) and in
Blomfield's Glossary rightly explained as active: ‘desertor classis'. This
1 We need hardly consider the remote possibility that initial x» here does not make
position (cf. below on 978).
122
 COMMENTARY lines 214f.

follows the explanation current in antiquity, cf. Hesychius λιπόνεως" ὁ τὴν

ναῦν ἀπολελοιπώς. Cf. also γραφὴ λιποταξίου and γραφὴ λιποναυτίου as well as
λιποστρατία. Hermann on the other hand has ‘quomodo a classe destituar?’,
basing himself on an arbitrary interpretation of the context without a shadow
of support from the language.’ None the less he is followed by Headlam
(‘deserted by my fleet’). But when the poet deliberately chose that clear-cut,
ugly word λιπόναυς, the Athenian audience could not feel the faintest doubt
that here a possible criminal act had been mooted, and rejected. Such an
action would be criminal for any member of the expedition, how much more
for the supreme commander.
213. ξυμμαχίας ἁμαρτών is well compared by Nagelsbach (similarly Passow—
Crönert, 337. 63 ff.; Platt, 7. Phil. xxxii, 1913, 47) with (2 68 ἐπεὶ où τι φίλων
ἡμάρτανε δώρων, where we have an unfamiliar use of the verb (cf. Schol. T
ἐπεὶ οὐχ ἡμάρτανεν ὁ “Exrwp τοῦ ἐμοὶ τὰ προσηνῆ ποιεῖν). Nägelsbach rightly
gives as the sense of the verb in both places ‘es fehlen lassen an . . .', and
translates: ‘Wie kann ich die Flotte lassen, die Bundestreue brechen?’
συμμαχία means here not, as it does in Pind. Ol. το. 72, in the epigram of the
year 457 expressing gratitude for the victory of Tanagra (Hiller v. Gaert-
ringen, Hist. Griech. Epigramme, no. 46; M. N. Tod, Greek Hist. Inscr. no. 27),
and commonly in later usage, the body of allies, but ‘alliance’, the relation
(with reference to rights and duties) in which a σύμμαχος stands to his fellows.
Our earliest instance (not mentioned in L-S) of συμμαχία used not in a
collective sense is the sentence in the treaty between Elis and Heraia (Ditten-
berger, Syll. 9; Solmsen-Fraenkel, Iscr. sel. 52; Tod, Greek Hist. Inscr. no. 5)
συνμαχία κ᾽ ἔα ἐκατὸν Férea. What Agamemnon says is: ‘when I have failed
in my duty as a ovppayos’. Probably Aeschylus has here taken advantage of
a rare use of ἁμαρτάνειν because he wanted once again (cf. on 212) with the
help of this verb to emphasize the moral failure of such an action, for in this
context ἁμαρτάνειν acquires a sense of moral guilt that it by no means always
has. The aorist participle ἁμαρτών indicates the cause of which λιπόναυν
γενέσθαι is the result; for this type of subordination cf. on 785.
214 f. παυσανέμου yàp θυσίας κτλ. The word παυσάνεμος occurs here only.
γάρ really does mark a reason, and shows very clearly that the question πῶς
Aurôvaus κτλ. is tantamount to the expression of a resolve (see above): ‘I
cannot become a deserter from the fleet, I must do what the goddess demands,
and the thought that thereby a maiden's blood must be shed ought not to
deter me, for it is lawful to crave the sacrifice that lulls the winds.’ To this
yap-clause and similarly to the one immediately following we can apply
Denniston's remark (Particles, 61) that in connexions with γάρ ‘compression
of thought is often the source of difficulty’.
παυσανέμου . . . θυσίας παρθενίου θ᾽ αἵματος : this special type of juxta-
position has been well described by Schadewaldt, Hermes, lxvii, 1932, 328
n. 3, and illustrated with examples from Aeschylus, including this passage.
In the sentences belonging to this category 'a single idea is expressed both in
the abstract and in the perceptible concrete symbol’, cf. on 238 and 949 (for
another type that has wrongly been classed with this, cf. on 1526). A similar
1 This interpretation, unfortunately, was chosen also by L-S, under the influence of
Hermann (he is actually named in the earlier editions) and, presumably, Paley, 2nd ed.
(afterwards Paley changed his mind).
123
lines 214 f. COMMENTARY
group arranged in the reverse order is found e.g. in Pers. 112 f. πίσυνοι
λεπτοδόμοις πείσμασι λαοπόροις T€ μηχαναῖς.
215 ff. ὀργᾶι περιόργως ἐπιθυμεῖν θέμις. For centuries these words have
been repeatedly questioned or altered. I choose at random Bothe: ‘quid hoc
sibi vult, aut quis ita locutus est Graecorum?’, Ahrens (p. 297): ‘the intoler-
able expression ὀργᾶι περιόργως (ὀργᾶι as a substantive)’, G. Thomson: ‘the
text is open to suspicion’. A refutation of all the errors that have been put
forward is unnecessary ; it must suffice to refer to some efforts that seem to
have had a specially tenacious life. To avoid the ‘tautology’ ὀργᾶι περιόργως,
many critics from Casaubon onwards have attempted to extract from ὀργᾶι
some form of the verb ὀργᾶν. So, e.g., Humboldt, who makes ὀργᾶι the govern-
ing verb (with ἐπιθυμεῖν depending on it) and θέμις the subject: ‘nach dem
Sühnopfer des Sturms heischet Begier [accusative] heftig das Recht’, and
similarly Plüss: ‘take together θέμις ὀργᾶι ἐπιθυμεῖν περιόργως θυσίας ... .
θέμις : the public law of the alliance, thought of as a personal force.’ Against
this we may say: (1) an infinitive depending on ὀργᾶν can indicate only the
object or content of the desire, not the desire—émbuueir—itself ; (2) to ascribe
to θέμις ‘as a personal force’ the crude emotion of ὀργᾶν is to show oneself
quite insensitive to the feeling of a Greek of the classical period when he
heard the name θέμις or Θέμις. Ahrens and Verrall likewise take ὀργᾶι as a
verb, but make ξυμμαχία the subject. This would in any case be objectionable
on stylistic grounds, for the sentence beginning with παυσανέμου makes a
completely fresh start; moreover, that interpretation is out of the question
because ξυμμαχία in 213 (see above) is not equivalent to of ξύμμαχοι. Ahrens
alters ἐπιθυμεῖν (see below); Verrall is compelled to make the genitive depend-
ent on ὀργᾶν (as others do, too), for which there is no parallel; this usage
might indeed not be impossible, but could certainly not be ‘according to
rule’, for ὀργᾶν, even where it governs an infinitive, retains its proper force,
and cannot simply be treated as interchangeable with verbs meaning ‘to aim
at’, ‘to aspire to’, etc. Another way of getting rid of the ‘tautology’ ὀργᾶι
περιόργως was tried by Hermann. He started from the varia lectio αὐδᾶιε,
which we find in the margin of M (yp. αὐδᾶι restored by Orelli and Hermann
for yap αὐδᾶι) and in the text of Tr; in accordance with the gloss ὁ μάντις
δῆλον ὅτι, which in M is connected with the variant αὐδᾶι, Hermann made
the seer the subject: ‘vates dicit fas esse avide expetere . . . sacrificium’,
taking θέμις as indeclinable (for this see the lexicons and Ernst Fraenkel,
Glotta, iv, 1913, 27 f.). Such a bare naked αὐδᾶι, ‘he says’, though Calchas is
not mentioned in the whole speech, is hardly credible. The variant αὐδᾶι is
indeed remarkable enough, and I know no serviceable conjecture as to its
origin, but for us the important point is that we can say, pretty certainly,
Aeschylus wrote not αὐδᾶι but ὀργᾶι: the proof of this is the excellence of the
expression ὀργᾶι περιόργως. The credit of rightly appreciating this phrase
belongs to Blomfield. He classes it with Prom. 944 τὸν πικρῶς ὑπέρπικρον
(compared with ὀργᾶι wepwpyws also by Heinrich Voss, Curae Aeschyleae
[Univ.-Progr., Heidelberg 1812], ro) and with the remarkable phrase in
1 As far as Aeschylus is concerned, there is an instance of the verb in the fragment of the
᾿Ελευσίνιοι (fr. 25 A in H. W. Smyth’s Loeb edition) preserved by Didymus, Commentary on
Demosthenes, col. 14. 12 ff., and probably in Cho. 454, where there can be hardly any doubt
that we have a verb-form, though the interpretation is controversial.
124
 COMMENTARY lines 215 ff.

H 97 aivößev αἰνῶς, in which connexion we should recall H 39, 226 οἰόθεν οἷος
and perhaps IT 776 (Σ 26, w 40) μέγας μεγαλωστί. We have here a peculiar
(archaic?) form of intensification, effected by placing next to an adjective an
adverb formed from the same stem! (the instrumental ὀργᾶι has the force of
an adverb). Cf. E. Or. 811 πάλαι παλαιᾶς ἀπὸ συμφορᾶς Séuwv.? The example
in Ag. 215f. is loftier than the others, which corresponds to the general
style of this ode. In Greek this mode of expression seems to be very rare.’
The nearest parallels known to me are to be found in considerable numbers
in Plautus :* Pseud. 13 misere miser sum, Rud. 977 esne impudenter impudens?,
Aul. 314 mortalem parce parcum praedicas, Cas. 522 nimium scite scitus es,
854 1, belle belliatula etc. In Aeschylus the περι- or ὕπερ- (Prom. 944) rein-
forces the repetition of the stem as a further element of emphasis (there is
something of the sort in nimium scile scitus es); this is in harmony with
Aeschylus’ tendency (to be discussed on 378) towards pleonastic description
of anything excessive or disproportionate.
ὀργή here is certainly not = τρόπος, as the gloss in M explains it, misled
by the well-known usage which is found as early as Hesiod, very common in
Pindar, and not unknown in Tragedy. But it is not = ‘anger’ either; rather,
it indicates in more general terms passionate emotion, as, e.g., in S. Oed. R.
1241. Thus ὀργᾶι περιόργως is ‘in passion most passionately’.
περιόργως : this is the accentuation of the MSS here, also of the MS of
Hesychius, s.v. (perhaps from a scholion on our passage), and the MS of
Photius. περιοργῶς was written long ago in Hesychius by Daniel Heinsius, in
Ag. 216 by Blomfield, who has been followed by Ahrens (p. 297) and others,
to all appearance plausibly, for elsewhere we find only περιοργής (once in
Thucydides, then in his imitators) and no περίοργος. But the unanimous
paroxytone accentuation of the adverb in the MSS of Aeschylus, Hesychius,
and Photius may, P. Maas tells me, go back to genuine grammatical tradition.
We have no idea how Aeschylus (or perhaps an earlier poet whom he followed)
formed the word. So it will be more prudent to retain the accent of the MSS.
! We may compare the compounds formed to express comparison by the doubling of an
adjective-stem, which are discussed by J. Wackernagel, Altind. Grammatik, ii. 1, 147 f.—
For οἰόθεν οἷος etc. cf. M. Lejeune, Les adverbes grecs en -ev (1939), 9o.
2 πολὺ πολυιδρίδας (lyr.) in col. 2. 1 of the new fragment of Sophocles’ Inachos (Pap. Tebt.
iii. 1, no. 692, cf. R. Pfeiffer, Sitzgsb. Bayer. Akad., Phil.-hist. Abt. 1938, Heft 2. 24; D. L.
Page, Greek Lit. Pap. i. 24) is only remotely similar to the instances quoted above.
3 I am not sure whether in the phrase S. Oed. R. 1469 ὦ γονῆς γενναῖε the dative is really
an ‘intensifying instrumental’ as Schwyzer assumes (Preuss. Akad., Phil.-hist. Abhdl. 1940,
No. 7, p. 6). The phenomenon discussed here is different from the well-known figure in
which the adverb belonging syntactically to the verb stands next to an adjective of the
same stem, as in Ar. Ach. 253 f. dy’, ὦ θύγατερ, ὅπως τὸ κανοῦν καλὴ καλῶς οἴσεις KTA.,
Clouds 554 ἐκστρέψας τοὺς ἡμετέρους ᾿Ιππέας κακὸς κακῶς, and the like. I think it possible that
in Ag. 215 f. also a close syntactical relationship between the instrumental ὀργᾶι and
ἐπιθυμεῖν was felt by the poet and by his audience (and the same may be true of the Homeric
μεγαλωστί in regard to κεῖτο). But even if that is so, épya must act as an element that
intensifies the immediately following περιόργως.
* The parallelism between the Homeric οἰόθεν οἷος etc. and these Plautine expressions
was noticed also by E. Hofmann, Ausdrucksverstärkung (Ergänzungsheft No. 9 zur Zeitschr.
f. vergleichende Sprachforschung, 1930), 92.
5 Cognate with this are the phrases where instead of the same word a synonym is added
to obtain the effect of a superlative, as, e.g., in Plaut. Most. 495 inepte stultus es and in the
collocations impie ingralus, mansuete oboediens, comiter facilis adduced by J. Vahlen, Ges.
Philol. Schriften, i. 513, from Cicero and Livy.
125
lines 215 ff. COMMENTARY
ἐπιθυμεῖν: this word is used in the lexicographers and scholia as an ex-
planation of ὀργᾶν, consequently since the seventeenth century repeated
attempts have been made to cut it out here asa gloss, and its opponents appeal
to the fact that it occurs nowhere else in Aeschylus. It is, however, found
in 8. Trach. 617 and E. Alc. 867. Here, with dependent genitive, it suits
admirably. That a virgin’s blood should become the object of anyone’s
ἐπιθυμία is most unnatural; the phrase is therefore very appropriate in this
context, in which the unnatural character of the sacrifice is emphasized again
and again (cf. p. 121 f.).
What is the subject of ἐπιθυμεῖν ἢ We may dismiss at once Artemis (gloss
in Tr; Casaubon, John Pearson, and others conjectured Ἄρτεμις for θέμις,
and Stanley translated accordingly). Heath rendered: ‘Nam aequum est
socios meos . . . expetere sacrificium'. This was taken up by Schütz (who
remarked about the noun ‘ex vocabulo ξυμμαχίας repeti’), and has been
adopted by nearly all translators ; cf. also the notes of Weil, Ahrens (p. 298),
Sidgwick, Verrall. That some scholars have felt qualms about it appears
from their conjectures, e.g. Schoemann’s (Opusc. iii. 167) σφ᾽ ἐπιθυμεῖν.
Wilamowitz regarded Agamemnon as the subject (‘und zu dem Blute der
Jungfrau . . .treibt es mich unwiderstehlich’). We should perhaps recognize
that the absence of a definite subject is intentional. Agamemnon chooses a
phrase which includes both his companions and himself. A few translators
have tried to bring this out, e.g. Davies (‘that we crave’) and Mazon (‘on
peut le désirer’).
217. θέμις = fas est (on this see, besides the lexicons, R. Hirzel, Themis,
Dike und Verwandtes, 51). Had Agamemnon chosen a term from the sphere
of merely human obligations and merely human laws, the contradiction
would not be so flagrant between what he would like to believe and the severe
norm ordained by God. He knows that the task he has in hand may be
necessary, but cannot possibly be θέμις. So, too, the εὖ yap εἴη that immedi-
ately follows may sound hopeful, but there is no real hope in it.
εὖ yap ein: for the compression of the thought see on 214. Hermann well
says: ' e? γὰρ εἴη ita explicandum est, ut omissa censeatur sententia, ad
quam yáp referatur : sic fiat. Ita quod brevissime dixit poeta, accipiendum est
hunc in modum: nam, quoniam evitari non potest, obtandum ut bene vertat.
It is a very delicate touch that Agamemnon should speak first of the grounds
and justification of his resolve and then of the result that he hopes will
follow from it, and in between leaves out ‘I am determined to sacrifice her”.
He cannot bring himself to utter the fatal words. Behind the last phrase
seems to lie a regular concluding formula from the language of prayer.
G. Thomson compares A. Suppl. 974 εἴη δὲ ra λῶιστα, E. Med. 89 εὖ yàp
ἔσται (the scholion is instructive : τὸ δὲ ' εὖ yàp ἔσται ' εὐχομένη λέγει, τουτέστιν"
ἐπὶ καλῶι δὲ ἀποβαίη krA., just as in Schol. A. Ag. 217 καλῶς ἀποβαίη), Herodas
4. 85 ἰὴ ἰὴ Ilaígov: ὧδε ταῦτ᾽ ein, taken up by the other character with εἴη
γάρ, ὦ μέγιστε, which is clearly liturgical. Conington on Persius 4. 20 com-
pares this εὖ εἴη with the formula bene sit, which is ‘a sort of grace’.
To the speech of Agamemnon (206-17) we may apply the words by which
A. Rivier, Essai sur le tragique d’Euripide (Lausanne, 1944), 33, developing
ideas of B. Snell, attempts to describe the fundamental conditions of action
in an Attic tragedy. After quoting A. Cho. 899 Πυλάδη, ri δράσω ; μητέρ᾽
126
 COMMENTARY line 219

αἰδεσθῶ κτανεῖν ; he says: ' "Que ferai-je?"" Cette question décrit la situation
cee

tragique par excellence, celle qui peut étre considérée comme l’archetype de
toutes les situations οὐ le drame attique, depuis Eschyle, place ses person-
nages. C’est la situation du héros qui, pour répondre a l’action des puissances
divines, se voit devant la nécessité d’agir lui aussi, et de commettre un acte
dans lequel il engage sa responsabilité d'homme ; sur cet acte les dieux vont
le juger et lui marquer plus fortement encore leur puissance redoutable.
C’est donc la nécessité d’une decision capitale, souvent mortelle, toujours
irrévocable. L’action ainsi congue a une dimension métaphysique; elle
atteste le lien de dépendance concrète qui rattache l'homme aux puissances
suprêmes du cosmos.”
218. ἐπεὶ δ᾽ ἀνάγκας κτλ. This passage is quoted in Euripides’ Iphigenia in
Aulis, 443 (Agamemnon is speaking) ès ol’ ἀνάγκης ζεύγματ᾽ ἐμπεπτώκαμεν
and 511 (the same) ἀλλ᾽ ἥκομεν γὰρ εἰς ἀναγκαίας τύχας θυγατρὸς . . . ἐκπρᾶξαι
φόνον. Cf. also below on 1071, further Prom. 108 ἀνάγκαις ταῖσδ᾽ ἐνέζευγμαι
τάλας, Soph. Tereus, fr. 532 N. (= 591 P.) ζυγὸν ἔσχ᾽ ἀνάγκας, Phil. 1025
ἀνάγκηι ζυγείς, Eur. fr. 475 N. τὸ τῆς ἀνάγκης... ζυγόν, Or. 1330 ἀνάγκης δ᾽ ἐς
ζυγὸν καθέσταμεν, Hdt. 8. 22. 2 εἰ... ὑπ᾽ ἀναγκαίης μέζονος κατέζευχθε ἢ ὥστε
ἀπίστασθαι (related metaphors are also used by Herodotus elsewhere in con-
nexion with dvayxatn). ᾿᾿ἀνάγκη quaevis condicio qua libertatis fines coercentur'
(Klausen). We should remember that ‘ ἀνάγκη is much more the compulsion
imposed on men in concrete circumstances than predestined rigid necessity ’
(Wilamowitz, Gr. Trag. ii. 26), cf. on 1535 f. On ‘the crushing fetters which
the Greeks originally found in ἀνάγκη ' cf. R. Hirzel, Themis, 427, with ex-
amples. The yoke of compulsion is spoken of with reference to a παρακοπή, a
stroke depriving one of one’s reason (223), also by Bacchyl. 11(10). 45 τὰς
(the daughters of Proitos) ἐφόβησεν... "Hpa . . . παραπλῆγι φρένας καρτερᾶι
ζεύξασ᾽ ἀνάγκαι. Passages of Tragedy like those quoted above are imitated
in the anapaestic poem (early Empire) in Berliner Klassikeriexte, v. 2, p. 135,
l. 13 f. (D. L. Page, Greek Lit. Papyri, 414): Hecuba ἦλθεν ὑπ᾽ αὐτὴν ζεῦγλαν
ἀνάγκης.
219. φρενὸς ... τροπαίαν : editors long ago compared Sept. 706 λήματος...
rporaiä, where the image of the wind (our πνέων) is clearly brought out in
the following θελεμωτέρωι (a certain restoration) πνεύματι. Lobeck, Parali-
pomena, 314, may be right in holding that rporaia was originally an adjective,
but if so, the three passages in which Aeschylus uses the word make it rather
improbable that it was still by him felt as such. Cf. also Wilamowitz on E.
Her. 681, and for the general problem of such a function of the feminine see
below on 916. O. Becker (in the monograph quoted on 187), 173 f., brings the
way of indicating the change in the wind by τροπαίαν (cf. A. C. Pearson on
Soph. fr. 1103) into relation with passages in Homer in which we read of a
τρέπειν φρένας, νόον, and the like. He rightly regards the τροπαΐα, which is
described as δυσσεβής, dvayvos, dviepos, as a preparation for the παρακοπά
(223). τροπαία, like μεταγνῶναι (221), here indicates a turning from what is
reasonable and moderate to the infatuate and disastrous.
πνέων: for the underlying idea cf. on 187 and, e.g., Cho. 390 ff. πάροιθεν δὲ
πρώιρας δριμὺς ἄηται κραδίας θυμός, ἔγκοτον στύγος, S. Ant. 929 f. ἔτι τῶν
αὐτῶν ἀνέμων αὑταὶ ψυχῆς ῥιπαὶ τήνδε γ᾽ ἔχουσιν, E. Phoen. 454 σχάσον δὲ
δεινὸν ὄμμα καὶ θυμοῦ πνοάς. ‘The “stormy blasts" represent the "irrational"
127
line 219 COMMENTARY
force of emotion, which in Greek is so often felt as a force pressing on one
"from without" ’ (Schadewaldt, Hermes, 1xvii, 1932, 333).
220. ἄναγνον ἀνίερον: together with δυσσεβῆ a very impressive and complete
description of something nefastum, for ἁγνόν refers to purity, especially in
man's relation to the divinity, ἱερόν to what belongs to the gods and is in
conformity with their nature (cf. the sketch of the meaning of both words
in Wilamowitz, Glaube der Hellenen, i. zı f., for ἁγνόν also Ed. Williger, 'Hagios',
Religionsgesch. Vers. u. Vorarb. xix. τ, 1922, passim).
rößev: the starting-point of the fateful change of mind is quite sharply
defined ; the poet is anxious to emphasize that all the disasters following are
to be definitely traced back to that. Cf. on 223.
221. vavróroA pov, instead of πάντολμος used by Aeschylus and others else-
where, occurs only here and 1237 (the apparent parallel fr. 192. 4 παντοτρόφον
is very uncertain, cf. above on 52). The article renders the phrase as definite
as if it were a superlative (e.g. τὸ ἔσχατον τόλμης). τὸ παντότολμον is taken up
again in 224 with érAa. With reference to the same situation E. 15. T. 862
τόλμαν ἣν ἔτλη πατήρ.
μετέγνω; the verb is first found in Suppl. 110 and here. The passage of the
Supplices stands very close to ours in its general tone. Of the man who takes
a resolve such as Agamemnon’s one might say, according to the judgement of
the Chorus, διάνοιαν μαινόλιν κέντρον ἔχων ἄφυκτον, ἄταν δ᾽ ἀπάται μεταγνούς,
where the concluding clause means ‘he has changed his mind, to his own
infatuation and destruction,’ because he is deceived (has let himself be
deceived, cf. Pers. 93 ἀπάταν θεοῦ, but also Ag. 385 βιᾶται δ᾽ ἁ τάλαινα πειθώ)".
222. βροτοὺς θρασύνει κτλ. There is no call to doubt that Spanheim’s
restoration is right. The MSS (including M) have βροτοῖς with a punctuation-
mark (colon) after it. Such wrong punctuation is to be observed in other
places where the unusual position of ydp caused difficulty, e.g. in S. El. 492
the Laurentianus as the facsimile shows has dAexrp’. ἄνυμφα γάρ, in Phil.
1450 Í. καιρὸς Kai πλοῦς 65°’ ἐπείγει ydp, and similarly in E. Her. 1126 the MSS
punctuate ἀρκεῖ" σιωπὴ (or σιωπῆ) yap. In other passages attempts were made
to bring ydp into its normal position by rearrangement, e.g. Ag. 758, Ar.
Plut. 1205 (here the Ravennas and the ®-group of MSS have ταῖς μὲν yap
ἄλλαις) ; cf. Pind. P. 6. 44, where τῶν viv δέ has been corrupted to τῶν δὲ νῦν.
An earlier form of the text may perhaps be inferred from the scholion ὅθεν
ἔγνω πάντα τοὺς ἀνθρώπους τολμᾶν. That looks as though the commentator
had read βροτούς, also with a stop after it. All attempts to connect βροτοῖς
or βροτούς or any other form of the word with the preceding sentence lead to
absurd results. The text generally accepted gives just the sense required by
the context. Verrall is wrong in saying that ‘the emphatic position of βροτούς
is incorrect’. The preceding sentence deals with Agamemnon. Then follows
in justification a general reflection : ‘(Agamemnon’s criminal change of heart
! Hermann, who is followed by Paley, wrongly renders perayvous sero cognoscens. The
passage is rightly explained by Bücheler, Rh. Mus. xli, 1886, 8 (= Kl. Schr. iii. 89), Wecklein
(annotated edition 1902), Wilamowitz, Interpr. 31. Murray has again returned to an inferior
form of the text, while Headlam (in the notes to his translation) actually adopts a monstrous
conjecture of Tucker's and rejects perayvous ; yet precisely this expression is characteristic
of the poet's point of view: for to pass from a normal state of mind into a condition in
which he is ready to commit a crime, a man must have undergone a μεταγνῶναι, ἃ παρακοπά,
ἃ παράνοια φρενώλης.
128
 COMMENTARY line 228
Was typical of the way men behave.) For mortal men are made foolhardy. . . .’
The only consideration that might arouse some slight doubt of the traditional
arrangement of the text concerns the position of γάρ. This is harsher than
in any other passage in Aeschylus.’ For the syntactical connexion between
βροτούς and θρασύνει is much less close than the partitive relationship between
μόνος and θεῶν in fr. 161. 1 μόνος θεῶν yap Θάνατος οὐ δώρων ἐρᾶι or that in the
‘figura sermonis’ of Prom. 29 θεὸς θεῶν yap οὐχ ὑποπτήσσων χόλον (with which
compare, e.g., S. 47. 522 χάρις χάριν γάρ ἐστιν ἡ rixrovo’ ἀεί, E. Tro. 621 κακῶι
κακὸν yap eis ἅμιλλαν ἔρχεται). But as the language of Comedy shows (e.g.
Ar. Lys. 489 διὰ τἀργύριον πολεμοῦμεν γάρ; and much else of the kind), the
unusual position of γάρ, which becomes still more common in the fourth
century, is ‘clearly no poetic licence’ (Wilamowitz, Menanders Schiedsgericht,
156) but belongs to the language of ordinary life. We need therefore feel no
difficulty if Aeschylus on occasion makes a somewhat freer use of it than he
usually does.
223. παρακοπὰ mpwrompwv: this isa central theme in the work of Aeschylus,
cf. Pers. 97 ff. φιλόφρων yap παρασαίνουσα τὸ πρῶτον παράγει βροτὸν εἰς ἄρκυας
ἅτα, τόθεν (cf. on 220; it is perhaps no accident that in the surviving plays of
Aeschylus τόθεν only occurs in these two passages which are so closely akin
in thought) οὐκ ἔστιν ὑπὲρ θνατὸν ἀλύξαντα φυγεῖν and Ag. 385 f. where the
πειθώ that drives a man to take a criminal resolve is described as προβούλου
παῖς ἄφερτος ἄτας. Cf. also on 1192.
224. ἔτλα δ᾽ οὖν κτλ. On this use of the double particle Denniston, 463 f.,
says: ᾿ δ᾽ οὖν leads back to the main topic, which has temporarily been lost
sight of.’
225 f. yuvarkomoivwv . . . ναῶν: this whole clause, as Heath saw, is in apposi-
tion to the preceding sentence (ἔσλα θυτὴρ γενέσθαι θυγατρός). Wilamowitz, in
his discussion of this type of construction (on E. Her. 59), has taken this
instance into account.
225. γυναικοποίνων πολέμων : the expression is so chosen as, at least from the
point of view of the Chorus, to throw his guilt into relief, cf. 448 ἀλλοτρίας
διαὶ γυναικός, 799 ff. σὺ δέ μοι τότε μὲν στέλλων στρατιὰν ᾿Βλένης ἕνεκα... κάρτ᾽
ἀπομούσως ἦσθα γεγραμμένος.
226. For προτέλεια cf. on 65. The relationship with the immediately
preceding θυτήρ is close. This phrase, too (see on 218), is echoed in the [pht-
geneia in Aulis, 718 προτέλεια δ᾽ ἤδη παιδὸς ἔσφαξας θεᾶι; and 433 Ἀρτέμιδι
προτελίζουσι τὴν νεανίδα, but there the reference of the term to the wedding-
ritual is fully maintained.
228. κληδόνας marpwious: κληδών in the sense of ‘cry, appeal’ seems to occur
only here and Eum. 397. πατρῶιος used as it is here, to some extent represent-
ing an objective genitive, comes a few times in Sophocles (φόνους πατρώιους
and the like).
From this πατρώιους onwards the theme of fatherhood becomes dominant:

! The position of yépin βροτοὺς θρασύνει γάρ may at first sight look as if it belonged to
the same category as that in Ag. 758 τὸ δυσσεβὲς γὰρ ἔργον, but in reality the latter is a
much milder instance, and so is even Ar. Ach. 581 ὑπὸ τοῦ δέους γὰρ τῶν ὅπλων (cf. on 653).
I would not dare maintain that in the new Niobe-fragment 1. 4 Reinhardt’s supplement
(cf. Hermes, ἸΧΙΧ, 1934, 237) [λάβρωἸς κακοῦ γάρ is impossible because of the word-order, but
taken on its own merits it is quite uncertain.
4872.2 K 120
line 228 COMMENTARY
231 πατήρ, 243 πατρός, 244 πατρὸς φίλου; this was pointed out by Ferrari,
La parodos, 390.
229 f. wap’ οὐδὲν... ἔθεντο: the phrase is discussed in Blomfield's Glossary ;
more material in Rau, Curtius Studien, iii, 1870, 76. Among the many related
passages, the identical collocation παρ᾽ οὐδὲν τίθεσθαι is found in E. Iph. T.
732, Plato, Phaedr. 252 a. Of E. Or. 569 παρ᾽ οὐδὲν... ἦν the scholiast gives an
adequate paraphrase: ἐν ὀλίγηι προφάσει καὶ πάρεργον ἦν.
αἰῶνα παρθένειον the MSS, αἰῶνα παρθένειόν 7’ Casaubon and Elmsley. This
change finds supporters even in recent times, although the apparently easy
addition of τ᾿ here is really rather a bold conjecture. For this would make a
simple connective re stand after (instead of between) the noun and adjectival
attribute which form a word-group. Of this there seem to be very few un-
doubted examples, not only in Aeschylus but in the earlier language as a
whole ; Porson’s protest against the reading E. Med. 752 (750 Porson) ὄμνυμι
Γαῖαν λαμπρὸν ᾿Ηλίου re φῶς, ‘copula ineleganter inserta’, is well founded, and
when Elmsley (ad loc.) retorts ‘cuius quidem inelegantiae exempla satis
frequenter apud tragicos occurrunt’, in this case Cambridge has the better of
Oxford. To take Aeschylus himself first, in Sept. 285 πρὶν ἀγγέλους σπερχνούς
τε καὶ ταχυρρόθους λόγους ἱκέσθαι is not the same, even if we retain ἀγγέλους.
For then we should have, not a simple connexion as in αἰῶνα παρθένειόν Te,
but a pair coupled with re . . . καί of which each member consists of noun plus
attribute (Tucker’s ingenious shift of taking σπερχνούς re καὶ ταχυρρόθους
λόγους together and ἀγγέλους as predicative with ἱκέσθαι may be dismissed).
It is, however, possible that with Wilamowitz we should prefer the reading
of the Parisinus ἀγγέλου. As for Suppl. 9 f., it is quite improbable that we
should simply try to retain the MS reading without altering or adding to it
(as, e.g., Wecklein does in his edition of 1902; he explains γάμον Αἰγύπτου
παίδων ἀσεβῆ τ᾽ ὀνοταζόμεναι ‘and in abhorrence of the marriage’, and adds a
misleading remark on τε ‘standing fifth’). It is more difficult to reach a
decision on Suppl. 282 Κύπριος (Κύπρις M') χαρακτήρ τ᾽ ἐν γυναικείοις τύποις
εἰκὼς πέπληκται κτλ. Wilamowitz's arguments against Kurpus' are hardly
relevant. He asserts: ‘incredibilia et correptio et anapaestus’, but his first
point is met by the short initial syllable of Κύπριος in Pers. 891 and his second
by a reference already made by Wecklein to two passages where anapaests
stand with a syllable long by position forming their longum, viz. Pers. 343
ἑκατὸν dis and Ag. 509 ὕπατός re (cf. Blass on Eum. 24). But in view of the
harshness of the construction, which is unusual even for that play, and of the
obscurity of εἰκώς, I am by no means convinced that ‘the words are un-
questionably sound’ (Kranz, Stasimon, 285; he is wrong with regard to the
metre, and indeed most of what is asserted about the suggestive force of
particular rhythms rests on pure imagination). On Ag. 404 f. see ad loc. In
Sophocles there seems to be no example which belongs here even remotely.”
In Euripides, Alc. 818 f. kai κουρὰν βλέπεις μελαμπέπλους στολμούς re may be
dismissed at once: the four lines (817-20), of which three were only imper-
fectly attested in antiquity (schol. on 820), are unquestionably interpolated,
as was shown (not for the first time; cf. Wecklein's appendix) by Wilamowitz
1 Korte, too, Hermes, lxiv, 1929, 73 n. 4, says: ‘ Kdmpios is corrupt".
2 Such an order as is produced in Ant. 1124 by Jebb’s insertion (accepted by Bruhn) rap’
ὑγρὸν ᾿Ισμηνοῦ ῥεῖθρόν {τ᾽ is extremely suspicious.
130
 COMMENTARY lines 229 f.

(in Adolf Gross, Die Stichomythie in der griech. Tragödie, Berlin, 1905, p. 39,
and briefly in Griech. Tragödien, iii. 156, at the end of his translation of the
Alcestis). We ought to be thankful that here the odd order of words helps to
unmask the intruder, and not set ourselves to wash the Ethiop white by
making κουράν and στολμούς change places as Murray does (with whom D. L.
Page, Actors’ Interpolations, 60 f., expresses agreement ; Denniston, Particles,
517, is much more cautious). In Tro. 1069 (lyric iambics), where the MSS
have τέρμονά τε πρωτόβολον ἁλίωι (in the antistrophe 1079 αἰθέρα re πόλεως
oAouevas is unexceptionable), Musgrave’s suggestion τέρμονα πρωτόβολόν θ᾽
is wrong for the same reasons that make us reject Elmsley's αἰῶνα παρθένειόν 7’
here.’ Wilamowitz (Verskunst, 169) with his répuová re πρωτόβολον ἕωι has
probably found the right answer. It is, however, noticeable that in the same
ode we have 1063 f. (προύδωκας . . .) ὦ Ζεῦ, καὶ πελανῶν φλόγα σμύρνης aifepias
τε καπνόν. Here the parallelism between (a) πελανῶν and σμύρνης αἰθερίας and
(b) φλόγα and καπνόν is apparently meant to be intensified by the fact that
σμύρνης and aifepias are not separated by the particle. Whether anything of
this kind is also to be found in earlier poetry must remain an open question,
until we have a study of the whole material that distinguishes between the
different grammatical types. Elmsley's remarks in Mus. Crit. ii. 282 are of no
assistance for the type ‘noun plus attribute plus re’; Denniston, Particles,
517, confines himself to a brief selection ; I myself have not collected systema-
tically.? It seems worth noting that the only certain parallel I have yet come
across to the passage in the Troades occurs in the Birds, which was produced
in the following year, and moreover in the monody of the Hoopoe, in which
the εὐριπιδαριστοφανίζων turns on all the tricks of the most modern metric
and music, 255 ff. ἥκει ydp τις δριμὺς πρέσβυς καινὸς γνώμην καινῶν ἔργων τ᾽
ἐγχειρητής (so the MSS, except the Ambrosianus, which like Suidas iii. p. 83.
29 Adler omits 7’; 7’ ἔργων in Parisinus A is perhaps a conjecture). This has
in common with the Troades-passage the placing of the genitive-group in
front of the noun that governs it; here again we might attribute the post-
ponement of the re to the tendency not to spoil the balance of καινὸς γνώμην
and καινῶν ἔργων, although it is here a parallelism in form rather than sense.
Other examples have I none,’ and those I have given are evidently not of a
kind to encourage us to add one to their number by a conjecture (παρθένειόν
7’) (the same is true of Scaliger's conjecture A. Suppl. 1042 ψεδυραὶ τρίβοι 7'
᾿Ερώτων, which still maintains its place in more recent editions). No objec-
tion can be raised, on the other hand, to αἰῶ re, suggested in place of αἰῶνα
by Otfried Müller (Gött. gel. Anz. 1834, 1978 = KI. Schr. i. 279); he took αἰῶ
! In Bacchae 571 the rearrangement proposed by Wilamowitz (see also his Griech.
Verskunst, 610) and accepted by Murray, Audiav πατέρα re κτλ. (the MSS have “υδίαν re τὸν
τᾶς εὐδαιμονίας βροτοῖς ὀλβοδόταν πατέρα τε τὸν ἔκλυον), cannot be regarded as successful.
Nor should we in Soph. fr. 775 N. (= 859 P.) follow Headlam and Pearson, who read σὺν
odreı κωδωνοκρότωι τε (σὺν σάκει δὲ κωδ. Plutarch, Mor. 640 a, where the most recent
editor, Hubert, puts δέ outside the quotation and adds the note " δὲ Plutarcho tribuit
Wilamowitz’).
2 In prose I will briefly mention at least Hdt. 8. 140 β 3, where Stein and Hude leave
ἐξαίρετον μεταίχμιόν re in the text, while with other editors J. E. Powell regards an alteration
as necessary, I think rightly.
3 In passing I may say that I do not share the confidence of Farnell who (The Works of
Pindar, Crit. Comm. p. 226) boldly suggests that Pind. P. 11. 9 ἱερὰν Πυθῶνά τε should be
taken together and asserts that ‘the slight postponement of the τε causes no trouble’.
131
lines 229 f. COMMENTARY
from the entry in the lexicographers (Zuvaywyn λέξεων χρησίμων, Bekker,
Anecdota, i. 363, now also Phot. Berol. p. 56. 18) αἰῶ τὸν αἰῶνα κατὰ ἀποκοπὴν
Αἰσχύλος, of which we cannot say whether it referred to this particular
passage. The explanation κατὰ ἀποκοπήν is of course nonsense: the accus.
αἰῶ is rightly classed with ‘words in -ws, gen. -oos’ by Buck-Petersen, A
Reverse Index, 26, and the nom. ais has been detected in a quotation from
Stesichorus in the glossary of Cyrillus by Latte, Mnemos. S. iii. 10, 1941, 84;
cf. also L-S p. 2045. The credit of having first made use of the form αἰῶ for
the text of Aeschylus belongs to Ahrens, who in the year 1832 restored it in
Cho. 350. This was tacitly appropriated by Hermann, while in his Opusc. vii.
44 he rejected Miiller’s restoration in Ag. 229 because the last syllable of
παρθένειον could not stand in place of a long syllable. Against this Ahrens,
p. 298, rightly refers to the syllaba anceps in 208 ἄγαλμα and 239 χέουσα (cf.
ad loc.); all three passages show a bacchiac ending to an iambic series,
similarly Ag. 777, where παλιντρόποισζινδ, with syllaba anceps at the end, is
an almost certain restoration. To these instances we should perhaps add 251
τὸ μέλλον, cf. ad loc.
230. βραβῆς : Aeschylus is for us the first person to use the word, yet here and
in Pers. 302 it looks as though it had already developed beyond a more original
and specialized meaning. For with Blomfield, who is followed by the lexicons,’
one would like to assume that the word had primarily the sense found in
Sophocles and Plato ‘arbiter in a prize-fight’ and then took on the wider
sense given it by Aeschylus, as in Horace, Odes 1. 3. 15, Notus appears as
arbiter Hadriae (‘arbitrum pro rectore posuit’ Porph.). In Euripides, too,
βραβεύς means not ‘commander’ but ‘judge’ (although not confined to
athletic contests). There is hardly any reason for assuming with Wilamowitz
(Griech. Lesebuch, ii. 34, on A. Pers. 302) that it is a ‘glossematic’ word for
‘commander-in-chief’.
231. ἀόζοις. ἀοζήσω- διακονήσω, ὑπουργήσω is quoted from the ᾿Ελευσίνιοι of
Aeschylus (fr. 54) by Hesychius.^ Wilamowitz, Hermes, xii, 1927, 287, refers
to an inscription of Corcyra IG ix. 1. 976 (an excerpt is given by Schwyzer,
Exempla, no. 139 g p. 66), where in a list of attendants at a sacrifice with
various duties an dofos is referred to. ‘Aeschylus is using a word that was
foreign to Athenians and Ionians’ (Wil.).
μετ᾽ edxav. From Stanley to Blomfield it was usual to punctuate after
εὐχάν. Hermann contested this: ‘nam per’ εὐχὰν cum λαβεῖν coniungendum
est’, and many have followed his lead. Among later editors, e.g. Conington,
Keck, Nagelsbach, Verrall, Wilamowitz (in his translation), Headlam have
taken jer’ edydv with what precedes. No arguments seem to have been
produced by either side. Naturally the poet did not bother his head about
‘punctuation’, but it can hardly have been a matter of indifference to him
whether the indication of time was taken as part of the account of what
Agamemnon did himself or as belonging to his directions to the servants;
and it is probable that the method of delivery made it clear to the audience
how the parts of the sentence were to be grouped. We, therefore, must try
to get this clear. As concerns the language, i.e. the order of words, examples
are not very common of such a placing of the adverbial expression between
the governing verb plus subject and the dependent infinitive clause; we
™ See now further Daube 29. 2 Cf. Phot. Berol. p. 153. 7.
132
 COMMENTARY line 232
shall find a precisely analogous and equally ambiguous example in 1036
(aumvirws)." As concerns the matter, it is undoubtedly preferable that prayer
should be offered first and then the person who is performing the sacrifice
give his orders to the ministrants. In the other case, i.e. if we make per’
εὐχάν part of the infinitive construction, we can hardly avoid the result that
the prayer appears as a concern of the attendants, whereas it is in fact the
business of the army commander (cf. Agamemnon’s prayer before the sacri-
fice in B 411 ff.) to utter the prayer that is so important,” possibly leaving the
others to join in with a concluding formula. We may find support for this in
the Messenger's report on the sacrifice of Polyxena (E. Hec. 534 ff.) : Neopto-
lemus, during the pouring of libations on the tomb, utters a solemn prayer to
Achilles, and then the narrative goes on (542): τοσαῦτ᾽ ἔλεξε, πᾶς δ᾽ ἐπεύξατο
στρατός. elra . . . φάσγανον... ἐξεῖλκε κολεοῦ, λογάσι δ᾽ Ἀργείων στρατοῦ
νεανίαις ἔνευσε παρθένον λαβεῖν.
232. ‘It was the custom at a sacrifice to raise the victim bodily on the shoulders
of the attendants to prevent any unseemly resistance, which would have
been ill-omened, and to ensure that the blood should fall upon the altar’
(A. C. Pearson on E. Hel. 1561, who compares E. Εἰ. 813 f. and Iph. T. 26 f.).
The lifting alone forms part of Agamemnon's instructions, not also the corn-
parison δίκαν χιμαίρας. The poet is no pedant ; he makes only one reference to
this particular incident in the action, and therefore at this point he adds
the comparison which strictly speaking should form part of the narrator's
description and does not belong to the king's commands (cf. on 233). The
comparison adds to the liveliness of the picture and makes it more moving.
The image is taken from a young goat called by the special name of χίμαιρα
(originally ‘that which has seen one winter, one year old’, cf. bimus, trimus) ;
nor is this without its point. In Athens, in Sparta, and in several other
Greek cities, it was the custom before a battle to offer to Ἄρτεμις Ayporepa
the sacrifice of a χίμαιρα or a number of χίμαιραι, cf. Passow s.v. (followed
by L-S, who among the supporting quotations omit Plutarch, Lycurgus 22. 4
and Aelian, Var. hist. 2. 25 [evidence for Athens]) ; E. Pfuhl, De Atheniensium
pompis sacris (1900), 34 n. 1; Neil on Ar. Knights 660; Farnell, The Cults
of the Greek States, ii. 434, 562 f.; P. Stengel, Griech. Kultusaliertümer, 133.
Iphigeneia is the substitute (that is the point of the comparison) for that
special sacrifice which otherwise would in similar circumstances be offered to
Artemis. φιλόμαχοι βραβῆς establishes a connexion with the sacrifice before
battle.
! When an adverb (to use an inaccurate term, not really suitable for a collocation like
ἐόντος dei) is ‘sandwiched’ between two ‘cola’, its connexion, and thereby the understanding
of the whole sentence, may be obscured, and that not only for the modern reader. Perhaps
the best-known example of this comes from the beginning of the work of Heraclitus
(B 1 Diels): τοῦ δὲ λόγου τοῦδ᾽ ἐόντος ἀεὶ ἀξύνετοι γίνονται ἄνθρωποι «rA., on which Aristotle
remarks (Rhet. 3. 5, p. 1407b 17) ἄδηλον γὰρ τὸ ἀεὶ πρὸς ποτέρωι δεῖ διαστίξαι, after adducing
the passage as an example of ἄδηλον εἶναι, ποτέρωι πρόσκειται, τῶι ὕστερον ἣ τῶι πρότερον.
To the discussions of the passage listed in Diels-Kranz, Vorsokr. i, 5th ed., 150, add O.
Gigon, Unters. z. Heraklit (1935), 1 ff. Gigon regards it as possible that the ambiguity was
deliberately intended by Heraclitus, but assumes (with Diels, Capelle, and others) that ‘in
the first instance’ we should take ἐόντος ἀεί together.
2 "The king who offers sacrifice in his camp or within the confines of his palace, and the
head of a household who offers in his own courtyard, utter the prayer themselves; in the
temple it is offered by the priest or priestess’, L. Ziehen, RE xviii. 606.

133
line 233 COMMENTARY

233. πέπλοισι περιπετῆ: an unparalleled expression, as Blomfield rightly

remarks. Even 5. 4j. 907, where Eustathius (cf. Lobeck, ad loc.) emphasizes
the boldness of the language, is easier to understand, for there the preposition
still retains its full value: Ajax has actually ‘fallen around’ the sword that
has pierced his body. ἡ δὲ ᾿Ιφιγένεια εἰσέπεσεν εἰς τοὺς πέπλους, in so far as
they enshroud her completely.’ Perhaps we ought to assume that the mean-
ing of περιπίπτειν is sometimes not very different from that of εἰσπίστειν or
ἐμπίπτειν. In favour of this is, e.g., the usage of Herodotus, ‘encounter’ and
the like. On the other hand, we must bear in mind the possibility that
Aeschylus, in accordance with his indifference in the coinage and use of
compound verbal adjectives, only intended to convey the idea that of πέπλοι
περὶ αὐτὴν πίπτουσι. As for πέπλοισι, it indicates here as often in Tragedy
garments in the most general sense. The vagueness with which the tragedians
use πέπλος has been pointed out by Otfried Miiller, and especially by F.
Studniczka, Beiträge zur Geschichte der aligriechischen Tracht (Vienna 1886),
133 f. He adduces among other things the frequent use of the plural and of
the formation πέπλωμα: ‘in Aeschylus these expressions denote not merely
the principal garment of women, including both the Dorian and the linen
Ionian—hence the promiscuous interchange with xeróv— but also the mantle
worn by men, and men's clothing in general’. πέπλοισι περιπετῆ describes the
appearance of Iphigeneia during the sacrifice, but this is not a part of the
order given to the ministrants. She is described as she is when the men lay
hands upon her, described by the poet and not by Agamemnon (cf. on 232).
παντὶ θυμῶι goes with λαβεῖν ἀέρδην, as one says elsewhere πάσηι τέχνηι or
πάσηι τέχνηι kal μηχανῆς Or παντὶ τρόπωι or, in the language of contracts (and
also, by a certain restoration, in A. Suppl. 147), παντὶ σθένει. Even closer is the
phrase in the Hesiodean Aoris 364 παντὶ μένει σπεύδων. Blomfield compares
Pind. N. 5. 31 πολλὰ ydp μιν παντὶ θυμῶι παρφαμένα Arávevev and Eum. 737 f.
τὸ δ᾽ ἄρσεν αἰνῶ πάντα... ἅπαντι θυμῶι.2
προνωπῆῇ (with the gloss mpoveveurviav): the precise meaning of the word
here is hard to determine. It may indicate no more than a falling forwards,
but in E. Alc. 143 ἤδη προνωπής ἐστι kai ψυχορραγεῖ it describes the slipping
away of consciousness and vital force, and in our passage the whole position
of Iphigeneia makes it likely that L-S are right with their rendering ‘to
take her as she fell fainting forward’. But Kuiper, loc. cit. (footnote 2 below),
goes too far when he gives the word by itself the meaning ‘unconscious, in a
dead faint’.
235. καλλιπρώιρου. The prototype of adjectives compounded with πρῶιρα
is clearly the Homeric xvavompanpos (νηὸς κνανοπρώιροιο). καλλίπρωιρος, which
in E. Med. 1335 is used of a ship, had already been applied by Aeschylus in
Sept. 533 to a handsome young man in the sense of καλλιπρόσωπον (so the gloss
in the scholia). Again in Prom. 423 f. o&umpanpowi . . . ἐν αἰχμαῖς, πρῶιρα
serves metaphorically to indicate simply the forward end of something. This

1 On the choice in E. Tro. 1181 between ἐσπίπτων πέπλους and ἐσπίπτων λέχος, Paley
judged very much more cautiously than Wilamowitz, Griech. Trag. iii. 362 (at the end of his
translation). In any case it is very true to life that while saying something affectionate the
child should bury itself in the πέπλοι of the old woman.
2 Ferrari, La parodos, 391, who refers to the last two passages, persists none the less in
taking παντὶ θυμῶι not with the verb but with προνωπῆ ; so, too, Kuiper, Mnemos. lv, 1927, 102 f.

134
 COMMENTARY lines 238 f.
weakened use of πρῶιρα is possibly older, and may well go back to the
‘kennings’ of oracular and sacrificial language. This is suggested by S.
Trach. 12. ἀνδρείωι κύτει βούπρωιρος, but more clearly by Empedocles
B 61. 2 βουγενῆ ἀνδρόπρωιρα. An inscription in Delphi of the beginning of
the second century B.c., Dittenberger Syll.? 604. 8, runs: συνετέλεσαν τῶι θεῶι
ἑκατόμβαν βούπρωιρον (see also what follows); its editors quote Hesychius
βούπρωρον" θυσία ἐξ ἑκατὸν προβάτων καὶ βοὸς ἑνός. This way of saying ‘with
an ox at their head’ looks like a very ancient liturgical expression. A deriva-
tion of this metaphorical use from the language of sacrifice would also square
with Soph. fr. 658 N. (= 726 P.) ὦ πρῶιρα Aoißns “Eoria, κλύεις τάδε; (cited
as evidence for the fact that τῆι ‘Eoria: τὰς ἀπαρχὰς ἔθος ἦν ποιεῖσθαι).
236. The change to φυλακᾶι, tentatively suggested by Blomfield, is necessary.
The meaning of φυλακὰν κατέχειν is shown by E. Tro. 194. As Headlam says,
‘if the MS reading is kept, it should be treated as subject to xaracyeiv’, and
so Klausen translated it, but this was rejected by Conington. In fact, such a
change of subject in a passage like this, where the two infinitives are joined by
re and the whole depends on φράσεν ἀόζοις, is extremely unlikely. The sense
is: ‘by keeping guard over her mouth to hold back (stifle) cries which would
bring a curse upon the house’.
φθόγγον ἀραῖον οἴκοις : “any ill-omened expression which might excite the
φθόνος of the gods’ says Paley, and he is right in principle, except that we
should not think here of φθόνος. G. Murray, The Rise of the Greek Epic,
4th ed., 88 n. 2, exactly takes the point: ‘not a spoken curse— which would
make the passage hideous—but the mere crying of a murdered daughter,
which necessarily involves an apa’.
238 f. βίαι χαλινῶν τ᾽ ἀναύδωι μένει. κρόκου βαφὰς δ᾽ és πέδον χέουσα κτλ.
This passage is so difficult, or at the least so peculiar, that even critics of the
eminence of Ahrens (pp. 299 ff.) and Wilamowitz have succumbed to the
temptation to prefer the simplifying version of Triclinius (either as a whole or
in select details) to the apparently less attractive παράδοσις. Ahrens recognized
that δ᾽ after χαλινῶν is a conjecture, but adopted it none the less; he also
followed Triclinius in deleting the δ᾽ given by the MSS after βαφάς and (like
other critics) putting the stop after οἴκοις instead of after μένει. Wilamowitz
retained χαλινῶν 7’, but in altering the punctuation and deleting δ᾽ after
Badds he, too, followed Triclinius. The right answer is given by W. Kranz,
Hermes, liv, 1919, 309 f. and Stasimon, 153, 302. He has also taken into account
the objection raised on a point of form by Ahrens, Wilamowitz (in the year
1895, reprinted Verskunst, 182 n. 1), and P. Maas (Sokrates, iii, 1915, Jahres-
bericht, 234), that such an enjambement into the new stanza (βίαι χαλινῶν τ᾽
ἀναύδωι μένει) as is implied by the MS reading is unexampled in Aeschylus
and therefore inadmissible.? Ag. 176 had been referred to by Ahrens, but
there, as he rightly says, there is a much greater degree of independence at
the start of the new stanza. But that this is so in Suppl. 582 (quoted, and
τ On the other hand, in A. Cho. 390 the metaphor of the ship in πρώιρας is fully felt
(ἄηται shows this, cf. above on 219).
2 For the occurrence of such an enjambement across the end of a stanza, especially in
Pindar, cf. Wilamowitz, Pindaros, 244 n. 4; O. Schroeder in his commentary (1922) on
Pyth. 9. 16; Rolf Nierhaus, Strophe und Inhalt im pindar. Epinikion (Diss. Leipzig 1936 =
Neue Deutsche Forschungen, Bd. 89, Abteil. Klass. Philol.), passim. Cf. also Fr. Dornseiff
Archäische Mythenerzáhlung (1933), 22, 71.

135
lines 238 f. COMMENTARY
excused as legitimate, by Maas) no one would maintain, and there is little
force in the objection that there we are concerned with an antistrophe but in
Ag. 238 with the beginning of a new strophe, for, as Maas himself says,
'enjambement in the antistrophe is itself in any case contrary to normal
usage’. We must also remember that in Septem 855 (though it is hard to
decide the details of the text and consequently of the correspondence) the
beginning of the antistrophe probably cuts into the sentence which has
begun earlier (cf. on this Wilamowitz, Interpr. 84, and W. Kranz, Stasimon,
153 f.). No doubt the rule is valid (cf. also Kranz, op. cit. 117 f.), but the poet
in rare cases has allowed himself to break it. The indications to be derived
in Ag. 238 from the meaning have been taken into account by Kranz. But
by far the best argument in favour of the correctness of the MS reading is
provided by the arrangement of the narrative as a whole, beginning from
218. This section opens with a general estimate of Agamemnon’s behaviour
from the moral and religious point of view (down to 223 πρωτοπήμων). Then
follow four stages of the actual course of events, in four clear-cut, independent
sentences, each with a connecting δέ as the second word: (1) 224 ἔτλα δ᾽ οὖν
xrÀ., Agamemnon the subject: his resolve (and, by implication, its being
made known to the army). (2) 228 Acras δὲ «rA., the chieftains are the subject
(with Iphigeneia secondary): an emphatic confirmation of Agamemnon’s
resolve. (3) 231 φράσεν δ᾽ ἀόζοις κτλ., Agamemnon again subject: detailed
instructions. (4) 239 κρόκου βαφὰς δ᾽ ἐς πέδον χέουσα κτλ., Iphigeneia the
subject: the sacrifice, down to the end of the narrative in 246. Diagram-
matically expressed, we have the arrangement abab. These stanzas lay
bare the roots of the tragedy of Agamemnon; his decision, i.e. the mpwro-
πήμων παρακοπή, the source of all that follows, dominates the narrative, and
therefore in the two leading sections (a) it is he who acts, while the two that
correspond to them (b) describe the results for which he is responsible. In the .
beginning of the four sections there is a close correspondence, which
Wilamowitz’s alteration entirely destroys.
238. Bids χαλινῶν τ᾽ ἀναύδωι μένει: a further example of the particular form
of epexegesis which was discussed on 214. The whole expression here serves
to explain the cryptic φυλακᾶι in 235 ; for the tendency of Aeschylus to follow
his γρῖφοι with an explanation, see Ὁ. 82.
χαλινῶν: Aeschylus says nothing of a gag put in the mouth (Ar. Knights
376 ἐμβαλόντες αὐτῶι πάτταλον . . . eis τὸ στόμα, and similarly Thesm. 222) ;
what is described here is the binding of the mouth externally with tightly
drawn bandages or thongs. This arrangement is called a ‘bridle’ because it
looks like one. The expression is apparently not poetical, but taken from the
language of ordinary life, cf. Hdt. 3. 14. 4 τούς τε αὐχένας κάλωι δεδεμένους καὶ
τὰ στόματα ἐγκεχαλινωμένους. An analogous expression is found in the passage
of Schol. Ar. Knights 1150 compared by Blomfield, Αἰσχύλος ἐν Λυκούργωι
(fr. 125 N.) ἀλληγορικῶς τοὺς δεσμοὺς κημοὺς εἴρηκε διὰ τούτων ‘ Kai τούσδε
κημοὺς στόματος '. κημός is ‘really the muzzle that prevents the horse from
biting’, used by Aristophanes in jest of the mouth-band, φορβειά, worn by
the flute-player (Wilamowitz, Sitzgsber. Berl. Akad. 1911, 525 = KI. Schr. i.
346), cf. also above on 36 f.
ἀναύδωι has often been questioned. Blomfield, Dindorf, and others, e.g.
Wilamowitz, Verskunst, 182 n. x, explained ἄναυδον μένος as ‘vis quae vocem
136
 COMMENTARY line 239

prohibet’ ; this agrees with the explanation given in the scholia : τῶι ἀναύδωι
μένει τῶν χαλινῶν τῶν μὴ ἐώντων αὐτὴν λαλεῖν. Clearly this answers in general
to the sense needed here. But when Wilamowitz adds the remark that ‘in
vetere Graecorum lingua adiectivis omnibus activam et passivam vim inesse’,
in the first place that is an exaggeration (‘omnibus’),’ and secondly we should
probably do well here to steer clear of the voices of the verb. ἄναυδος means
‘voiceless, without sound (αὐδή)᾽ from the Odyssey onwards, and so often in
Aeschylus. It has apparently, but only apparently, passive force in S. Aj.
947 f. ἄναυδον ἔργον ‘Arpeidav ;* that is a peculiar and bold usage and yet not
so very remote from the ordinary one: elsewhere ἄναυδον is something at
which or in which there is no speech, in the passage of the Ajax it is something
for which (for describing which) there is no speech, no means of expression.
ἄναυδον μένος, though perhaps equally bold, is really less harsh ; here, too, we
should be careful not to introduce a verbal notion; it is sufficient to para-
phrase ‘that wherein no voice is’. The context makes it quite clear that if the
violence is described as ἄναυδον, this refers to its effect.” It is open to anyone
to regard this as a ‘condensed’ expression. But it would perhaps be more
correct to recognize that these privative adjectives formed from nouns often
denote the lack of something in a wider sense, merely indicating that it is not
a quality of the subject in question, not associated with it. Hence, for in-
stance, Homer can say ἄυπνος ἀνήρ no less than ἀύπνους νύκτας ἴαυον. Good
remarks on this will be found in Gerschewitsch (see footnote 1 below), p. 144,
but he makes too much play with verbal paraphrases and consequently with
“passive and active sense’.
239. xpóxou ... χέουσα, The text is sound. Many editors have objected to
the hiatus after χέουσα, e.g. Hermann, Ahrens (p. 301), and Verrall (p. 235
n. 1 of his edition). But hiatus is also found at the end of 368 (there, it is true,
we have a full-stop) and 377, Cho. 441, and in a different but related metre
e.g. Eum. 372. At the end of a full μέτρον (u — o Ὁ) in the middle of an iambic
period there is hiatus and syllaba anceps in Ag. 1091, syllaba anceps in the
strophe with hiatus in the antistrophe in Suppl. 135 = 145; this is still more
surprising than when it accompanies catalexis. For the syllaba anceps cf. on
229 (p. 132).
κρόκου βαφάς: a garment dyed with saffron. Whether Aeschylus is
thinking in particular of the robe called κροκωτός we cannot say; that the
word is recorded first from Cratinus and Aristophanes may well be due to the
1 What is really a common phenomenon is the alternation of active and passive force in
verbal adjectives in -ros (cf. on 12), especially if they are privative compounds like ἄκλαυτος,
which is passive in X 386, active in à 494; dvaros, which Aeschylus uses several times in the
Supplices in an active sense but in Ag. 1211 as passive (cf. Wilamowitz on E. Jon. 701);
ἄποτος, used in both senses by Herodotus ; ἀφθόνητος, passive in Ag. 939 and Pind. Ol. 11. 7
but active in Ol. 13. 25; ἄϊστος, generally passive but used in an active sense by Euripides ;
ἀνέλπιστος, partly passive in Thucydides, partly active (cf. Classen-Steup on 3. 30. 2), and
so on. Cf. further R. A. Neil on Ar. Knights 292; the lists in Schuursma, De abustone 52 ff.,
which however include a good deal that does not belong; Chantraine, La Formation des
noms, 306 f.; and especially Gerschewitsch, Stud. It. N.S. xv, 1938, 134 f.; an attempt at
explanation ibid. 143 f.
2 On the text cf. Wilamowitz, Verskunst, 332 n. 4.
3 Blass rightly says on Eum. 785 λειχὴν ἄφυλλος drexvos : ‘ ἄφυλλος and ἄτεκνος might be
referred to the “lichen” itself that has neither leaves nor fruit, but are better understood
of its effect on plants and trees [and on the offspring of all living things)’.
137
line 239 COMMENTARY
nature of our surviving literature. The poet here is careful to keep at a
distance anything too closely technical; we must therefore not expect that
his words will help to satisfy our curiosity regarding the dominance of the
‘ionic’ chiton or the ‘doric’ peplos or a mixture of the two in the female
costume of the first half of the fifth century. What matters is that he means
no everyday garment but something splendid and costly, befitting a king’s
daughter: Iphigeneia here, Antigone in E. Phoen. 1491 (with a characteristic
addition oroAidos κροκόεσσαν ἀνεῖσα τρυφάν). Similarly the shoe of Darius,
Pers. 659, is κροκόβαπτος and the swaddling clothes of the royal child Herakles,
Pind. N. 1. 38, are κροκωτὸν σπάργανον..
ἐς πέδον χέουσα: Wilamowitz’s ‘vom Busen riss rohe Faust das Safran-
kleid’ is not in the Greek; it would be ruled out even if we had not the
parallel passage long’ ago adduced (by Peile and others) from /liad E 734 f.
(Athene) πέπλον μὲν κατέχευεν ἑανὸν πατρὸς Em’ οὔδει ποικίλον. The words can
only mean ‘let fall to the ground’. This is not ἃ picturesque and ornamental
detail, it is an essential feature that stands in the closest connexion with
ἔβαλλε. .. βέλει φιλοίκτωι and should be interpreted accordingly. Iphigeneia—
this we have to infer from a narrative that only brings out the important
things—has in the meantime freed herself from the arms that were holding
her (or the ministrants have let her go free) and has thrown herself to the
ground by the altar in front of Agamemnon and the captains of the host.
With a quick movement of back and shoulders she lets her robe slip to the
ground,? and there she kneels, with upturned eyes, naked before the men—
she the king’s daughter heretofore so modest, so closely guarded. In the
general picture of her appearance, the element of φίλοικτον rises to a level
that unnerves one. Between the simple grandeur of this scene’ and the
ostentatious εὐσχημοσύνη of Euripides’ Polyxena (Hec. 558-70) there is a
great gulf fixed.
240. The imperfect ἔβαλλε is noticeable in contrast to the three preceding
aorists ἔτλα (223), ἔθεντο (230), and φράσεν (231), which on each occasion
introduce a fresh fact. On the other hand, with ἔβαλλε we are told of a detail
subordinate to the main action (cf. Wackernagel, Syntax, i. 182).
ἕκαστον θυτήρων naturally refers not to the ministrants but to Agamem-
non (224 θυτὴρ θυγατρός) and to the other chieftains taking part in the sacrifice,
who have been known to her of old (243 ἐπεὶ πολλάκις krA.).

* I regard Porson’s alteration as certain; Wilamowitz, Berl. Sitzgsb. 1903, 596, like
Hermann, retains the MS reading oroAida . . . rpußäs. Incidentally, is it possible to regard
στολίδα κροκ. (v v — UV) as credible prosody in Tragedy, even in dactyls (cf. Appendix E)?
2 Something of the sort, only on a much more modest scale, is to be found some decades
later in the inner picture of a red-figured cup from Spina (S. Aurigemma, Il R. Museo di
Spina, and ed., 1936, p. 162 plate 78). There Cassandra has sunk on her knees by the altar
of Apollo in front of Clytemnestra, who is falling upon her swinging an axe in her uplifted
hands. Cassandra, looking her murderess straight in the face, flings out her right hand to
Clytemnestra’s chin; as a result of her passionate movement (Iphigeneia in E. Iph. T. 362
uses a vivid word : ὅσας γενείου χεῖρας ἐξηκόντισα), the peplos which is here her only garment
has slipped down so that her right breast is bare. The slipping down goes much farther
with the Iphigeneia of our chorus, but here, too, as in the vase-painting, it accompanies an
imploring look that appeals for rescue, and in both it is the result of an impetuous movement.
3 The substance of the matter has been correctly understood and brought out by Kranz,
Hermes, liv, 1919, 310. If we tried to imagine that Iphigeneia is wearing another garment
under the saffron-coloured one, the decisive words would become meaningless.
138
 COMMENTARY line 242

ἔβαλλε... βέλει. Her only weapon is a glance that is darted ἀπ᾽ ὄμματος
(in itself a common idea, cf. on 742) and is φίλοικτον, and this shows the full
measure of her helplessness.
242. πρέπουσα. In Homer πρέπειν, as Buttmann, Lexilogus, i.* 20, established
and the lexicons repeat, means ‘to stand out’, conspicuum esse (‘ins Auge
fallen’, Pohlenz, Antikes Führertum, 1934, Ὁ. 58; a good sketch is there given
of the further development of the meaning of the word, more fully by
Pohlenz, Nachr. Gétt. Ges., Phil.-hist. Kl. 1933, 53 ff.), though we must observe
that it is used exclusively for the purpose of describing something that is
conspicuous against its background or surroundings, that ‘stands out’ from
them whether by size, shape, or colour. It is significant that the simple verb
is found only in M 104 (of Sarpedon) ὁ δ᾽ ἔπρεπε καὶ διὰ πάντων, while μετα-
πρέπειν (sometimes in ‘tmesis’) is quite common (cf. also ἐκπρεπής and
μεταπρεπής), in the same sense as another compound in Sappho (fr. 98. 6 D.)
νῦν δὲ “ύδαισιν ἐμπρέπεται γυναίκεσσιν ὥς ποτ᾽ ἀελίω δύντος à βροδοδάκτυλος
σελάννα πάντα περρέχοισ᾽ ἄστρα. In the same image and with the same sense
πρέπειν is used of the full moon standing out against the starry sky in Sept.
389 f. In Suppl. 719 f. πρέπουσι δ᾽ ἄνδρες viov μελαγχίμοις γυίοισι λευκῶν ἐκ
πεπλωμάτων ἰδεῖν, the verb suits the effect of the contrasting colours; it is
well taken up by eërperros in 722. When in Pers. 239 the queen inquires after
the armament of the Athenian army πότερα γὰρ To£ovAkös αἰχμὴ διὰ χεροῖν
αὐτοῖς πρέπει ; she is perhaps thinking of the appearance of a line of battle,
where bow and arrow stand out clearly. In Ag. 30, 389 πρέπειν refers to the
light as seen against the surrounding darkness; cf. on 321. πρέπουσα here is to
be understood in the same way: ‘conspicuous, clearly standing out’. At this
moment the great complex group of men, chieftains and ministrants at the
sacrifice, is in the foreground ; away behind them the mass of the army; both
alike appear as a mere foil against which stands out the central figure of
Iphigeneia. Her passionate movement (cf. on 239) makes her form con-
spicuous, sharply silhouetted. It may be subjective, but for me this descrip-
tion is always associated with figures like the Actaeon of the Pan Painter and
the Penthesilea Painter’s Tityos, forms whose bold movement makes on the
beholder no less deep an impression than the φίλοικτον which they so fully
convey. Beazley reminds me of Quintil. Inst. 8. 5. 26 nec pictura, in qua nihil
circumlitum est, eminet." This passage of the Agamemnon is our earliest
evidence for the clear definition of the individual figures being regarded as
an essential quality in painting.
In @ ὡς, θ᾽ will not do, for πρέπουσα, which is purely descriptive, cannot be
put on the same level as χέουσα (239), which carries on the narrative. Pauw
long ago objected to the particle. Probably we should write rws, as is sug-
gested by P. Maas, Sokrates, iii, 1915, Jahresbericht, 236, with which Wila-
mowitz, Verskunst, 182 n. 2, agrees. On an Attic alabastron of about 480 B.C.,
British Museum E 718, is incribed Adpodiaia καλέ, τὸς δοκεῖ Εὐχίρο, cf. C.
Smith, Cat. Vases B.M. iii. 354; Klein, Lieblingsinschriften, 150 (reference
from Beazley). In Drama we find τώς = ὡς not only in the mouth of the
Megarian Ar. Ach. 762 and in a satyr play (Soph. Ichn. 296), but also A.
! For the contrast or standing out of a figure in a painting against its surroundings or its
background, eminere is the regular term, cf. Thes. 1. L. v. 2, p. 494. 40 with the references
there given.
139
line 242 COMMENTARY
Sept. 637 (Suppl. 718 is corrupt). A possible alternative suggestion is Wila-
mowitz’s πρέπουσ᾽ ὅπως (cf. Prom. 1001).
242 f. προσεννέπειν θέλουσα. This is the first time anything is made of her
inability to speak, for it was not her silence that prompted the comparison
with a figure in a painting. But now the poet, who likes to maintain and
develop his comparisons, obtains a fresh effect from the ἐν γραφαῖς (‘the one
is too like an image and says nothing’, Shakespeare, Much Ado, τι. i. 7).
προσεννέπειν: she wished to call on the individuals (ékaorov θυτήρων) by
name, and thus add force to her appeal (‘help me, Diomedes!’, ‘Ajax, save my
life !’). How the maiden who has hitherto been kept within the house was in
a position to know the captains of the host is explained forthwith in the
errei-clause. That is how I had always understood it, when I found the same
explanation given by W. Sewell. He remarks, in the notes to his translation,
that we must observe the precise sense of ἕκαστον and προσεννέπειν, and adds:
‘She longed to address them each severally, and by their name— since, how-
ever, females were at that period secluded from general society, she had had
the means of becoming acquainted with them individually when appearing
at her father's banquets to play and sing before them.' MacNeice, too, rightly
has: 'to speak to them by name'.
243. εὐτραπέζους : for τράπεζα and its compounds denoting meals and feasts
cf. Wilamowitz on E. Her. 385.
245. ἔμελψεν 'cantaverat', for 'in Greek the need is not felt to express what
we call the Vorvergangenheit' (Wackernagel, Syntax, i. 151).
245 ff. ἀταύρωτος 'is perhaps hieratic' (Wilamowitz on Ar. Lys. 216).
τριτόσπονδον occurs here only. The scholiast on Suppl. 27, who uses the
word to gloss Ζεὺς σωτὴρ τρίτος, shows good judgement. For the thought cf.
on 1586 f.
παιῶνα: this emendation of Hartung's is generally accepted. Blomfield in
his Glossary had put together the materials which made the emendation all
but inevitable. The corruption was facilitated by the fact that the form used
by Aeschylus was not παιάν but παιών ; on this point cf. Wilamowitz, Platon,
ii. 339 (his remark escaped the notice of Wackernagel, who in Glotta, xiv,
1925, 61 ff., thoroughly demonstrated that παιών, παιωνέζειν are the forms of
the genuine early Atthis).!
It was the regular custom in various places, Athens among them, after the
meal, when the dishes of food had been taken away and before the drinking
proper began, to pour libations to a trinity of gods and therewith to sing.
a paean; cf., besides the reference books, the survey given by L. Deubner,
Neue Jahrb. f. d. klass. Altert. xlii, 1919, 391 ff., and further H. Sjóvall,
Zeus im aligriech. Hauskult (Lund 1931), 86 ff.
πατρὸς φίλου: it is he who starts the paean. The master of the house
offers the libation and utters the prayer over it, ending with in παιών (cf., e.g.,
Ar. Peace 453, Thesm. 310), and to this ‘Amen’ either the whole company of
those present or one individual replies with the singing of a hymn. For the

! In Cho. 343, in spite of the express warning in the commentaries of Wilamowitz and
Blass, Murray has retained the alteration παιάν, on Ag. 247 he describes παιᾶνα as ‘formam
Atticam’, against which the cautious treatment of L-S 1286 might have warned him. A. C.
Pearson, too, disdains παιών in S. Oed. R. 186, although it is the original reading of the
Laurentianus and also preserved in the lemma of the scholion.
140
 COMMENTARY line 247
two stages cf. Plato, Symp. 176 a σπονδάς Te σφᾶς ποιήσασθαι καὶ ἄισαντας τὸν
θεὸν κτλ. That the children of the house were brought in to join in the singing
of the religious song is natural enough ; for Arcadia Polybius 4. 20. 8 is evidence
for it even down to his own day. Here the expression πατρὸς... παιῶνα (as
distinct from εὐχήν or the like) suggests that either Agamemnon sings the
start of the hymn and then his daughter joins in, or that immediately after
the prayer all strike up the hymn, and the maiden’s clear voice is heard above
the song of the men. Iphigeneia shows reverence and honour to the paean
by her share in it.
The part played by the daughter in the sacrificial ritual of a company of
men consisting by no means solely of her near relations (as we understand
from the explanation of προσεννέπειν [sc. ἕκαστον θυτήρων) θέλουσα by the
following érei-clause) would be hardly conceivable within the limits of
Athenian custom.’ It looks as though Aeschylus in the development of this
particular feature had deliberately ‘Homerized’. ‘Ci si sente un’ atmosfera
eroica’ (Ferrari, La parodos, 396). For the use made by Aeschylus of the
difference between the Homeric customs and those of his own age cf. on 1382,
and on 1595 (καθήμενος). The emphasis laid on Iphigeneia's ayvorns and her
virginity (ἀταύρωτος), which in itself heightens the pathos, may also be
intended to underline the contrast with the ¢dAtpeac or aÿAnrpides who
normally play their part in the σπονδή (Schneidewin).
247. τὰ δ᾽ ἔνθεν: ‘the sacrifice itself could not be more impressively told
than by this terrible hint’ (Sidgwick). At the same time the expression ra 8’
ἔνθεν is so comprehensive that it includes not only the sacrifice of Iphigeneia
but also its sequel,” i.e. those subsequent events which were predicted in the
prophecy of Calchas,? the undisturbed departure of the fleet, the final (126
χρόνωι) capture of Troy, and also the inevitable consequences of the sin
which hound the sinner on to punishment (154 f.). Thus τέχναι δὲ Κάλχαντος
οὐκ axpavrot follows without a break on the preceding sentence. In the
subsequent words the thought dwells on the last link in the chain, the
punishment of the crime. That this must come to pass the Elders know, not
only because the prophecies of Calchas are to be trusted, but above all
because they themselves are convinced that δράσαντι παθεῖν is ineluctable.
Cf. on the third stasimon, p. 451 f.
τὰ δ᾽ ἔνθεν κτλ. indicates that the Elders have themselves experienced in
Aulis the events which they have so far been telling. Naturally the spectator
has no right—and if he is the kind of spectator for whom Aeschylus wrote,
no reason*—to inquire why the old men followed the expedition from Argos
as far as Aulis.
1 Triclinius, a thoughtful man, was surprised at the fact. He observes: ἔθος δὲ ἦν ὡς
ἔοικεν ἄιδειν τε καὶ ὀρχεῖσθαι τῶν βασιλέων ras θυγατέρας ἐν τοῖς συμποσίοις, δῆλον δέ ἐστι τοῦτο
ἐντεῦθέν τε καὶ ἐκ τῆς ὀρχησαμένης Ηρωδιάδος (he means of course her daughter Salome).
2 Ferrari, La parodos, 398, says: ‘ τὰ δ᾽ ἔνθεν means, not “the events that followed the
sacrifice” but “the consequences of the sacrifice”. I fancy the expression includes both.
3 On the other hand, nothing suggests that the Chorus are hinting at a knowledge of the
rescue of Iphigeneia, as Wilamowitz, Hermes, xviii, 1883, 253, maintains. So, rightly,
L. Séchan, Etudes sur la tragédie grecque (1926), 372 n. 7, Ferrari, La parodos, 397. Nothing
is known of the contents of the ᾿Ιφιγένεια of Aeschylus, cf. on 1523 ff.
4 For the principle involved cf. my general observations, p. 255 f., about the difference
in the treatment of circumstances relevant to the dramatic action and of circumstances
which belong only to the general background.
141
line 250 COMMENTARY
250. Δίκα. . . ἐπιρρέπει. That ἐπιρρέπειν here is used transitively, as in
Eum. 888, was recognized in the paraphrase of the scholia : τοῖς μὲν πεπονθόσιν
ἡ δίκη δίδωσιν τὸ μαθεῖν. The simple verb is also transitive in a similar context
in Bacchylides 17(16). 25 ὅ τι μὲν ἐκ θεῶν μοῖρα παγκρατὴς ἄμμι κατένευσε καὶ
Δίκας ῥέπει τάλαντον. For the idea of the scales of justice cf. hymn. Hom.
Merc. 324 Sixns . . . τάλαντα, Bacchyl. 4. 12 (where now εἷλκε Δίκας τάλαν[τ is
certain ; cf. Addenda to p. 9o n. 2), Cho. 61 ῥοπὴ . . . δίκας, and see the observa-
tions of G. Björck, Eranos, xliii, 1945, 60.
This sentence, which shortly before the end of the parodos impresses on
us once again the fundamental theme of the central ‘Hymn to Zeus’ (cf.
177), only lightly conceals in the form of a γνώμη the prediction of the specific
evil that is to come. The context leaves no doubt which deed it is, and which
doer, that draws down the suffering. There is a strong emphasis on τοῖς
παθοῦσιν : suffering is allotted them by Dike, and only through suffering can
they reach wisdom. Thus we may perhaps regard μέν here as ‘emphatic’.
251 ff. The words τὸ δὲ προκλύειν (in M not by the first hand) are omitted in
Tr.! John Pearson and Elmsley recognized that there is no room for these
words here. Oddly enough, however, the intruding phrase although long ago
rejected still exercises a baleful influence on the understanding of the passage.
Stanley translated his text τὸ δὲ mpokAew . . . mpoyaipérw: ‘ipsum autem
praescire . . . valeat'. This corresponds to the advice given in the marginal
note in F: σύναπτε τὸ προχαιρέτω eis τὸ mpoxAvew. The interpretation of
προχαιρέτω == valeat (so the interlinear gloss in Tr φθαρήτω τὸ μέλλον) has
maintained itself in countless commentaries and translations as well as in the
lexicons, despite the warning provided by the passage in Plato, Phil. 39 d,
where ἡδοναὶ καὶ λῦπαι is taken up by τὸ mpoxaipew τε καὶ προλυπεῖσθαι. But in
any case it should be clear that προχαιρέτω and προστένειν stand in a relation
of antithesis to one another (cf. also fr. 266. 3 τῶι μήτε χαίρειν μήτε λυπεῖσθαι) ;
that disposes also of the artificial πρὸ χαιρέτω of Ahrens and Meineke which
many editors have adopted. The subject of προχαιρέτω is naturally τὸ μέλλον.
From this point of view, too, the traditional interpretation of the imperative
(‘das fahre hin’ Hartung, ‘let it go’ Headlam, and the like) is unsatisfactory.
When we are thinking of the appearance of something that has not yet
arrived but is still expected (τὸ μέλλον), χαίρειν puts us in mind in the first
instance not of valedicere but of greeting the new arrival. The current
interpretation was followed by Wilamowitz in his Greek paraphrase of our
passage, on E. Her. 575; but his translation, though rather free (it is accepted
by Kranz, Hermes, liv, 1919, 311), seems to imply a quite different and more
! Theoretically speaking, there are three possible ways of accounting for this omission.
(a) Triclinius used a good MS, where the interpolation was missing. (6) He skipped the
words inadvertently. This is the view of Blass (Introd. to his edition of Zum. p. 18 f.), who
compares Ag. 1031 (wrongly, for this line was deliberately omitted by Triclinius, see my
account of the MSS in vol. i) and 1573. (ὦ Triclinius, with his eye on the corresponding
passage of the strophe (241), realized the discrepancy and restored symmetry by deleting
τὸ δὲ προκλύειν. The last explanation seems to me by far the most probable. Cf., e.g., what
Triclinius did in the case of the famous interpolation Pind. Ol. 2. 27 (29) φιλέοντι δὲ Moivaı,.
in A. Prom. 400, Sept. 721, 883, in S. El. 856, where, on the evidence of the antistrophe, he:
cut out αὐδᾶις δὲ ποῖον, and in Oed. R. 507, where for the same reason he deleted én’ αὐτῶι.
(he might also have deleted yap, as Hermann observed. Incidentally, Pearson’s note ignores.
Hermann’s later (1833) re-edition of Erfurdt’s commentary, where he recanted his former
suggestion and assumed a lacuna after 493 ὅτου δὴ). Cf. further E. Hec. gıı.
142
 COMMENTARY line 254
appropriate view: ‘. . . die Zukunft hört man, wenn sie da ist, früh genug.
Ich biet’ ihr nicht Gruss voraus noch Fluch voraus.’ This points in the right
direction, but the antithesis in Aeschylus is sharper. With ἴσον κτλ. Headlam
compares E. Or. 426 τὸ μέλλον δ᾽ toov ἀπραξίᾶι λέγω (cf. also Hdt. 6. 86 y 2
ἔφη τὸ πειρηθῆναι τοῦ θεοῦ καὶ τὸ ποιῆσαι ἴσον δύνασθαι) ; cf. moreover Ag.
1403 f, σὺ δ᾽ αἰνεῖν εἴτε με ψέγειν θέλεις, ὁμοῖον. Naturally, although χαῖρε
had long ago crystallized into a formula of greeting, it still sometimes con-
veyed the original meaning; it is so used for instance not only in Comedy
(Ar. Ach. 832, Eupolis fr. 308 K., Philemon fr. 7. 3 f. K.) but also in A. Ag.
538 f., S. Trach. 227 f., E. Hec. 426 f., Phoen. 618. In our passage there is no
real play on the word, but Aeschylus uses it in order to bring out how the
Chorus even from a harmless and as it were neutral use of it are plunged back
again at once into their sombre expectations. One goes to meet a new arrival
with the word χαῖρε; thus the Chorus say: ‘the future will surely come in its
due time, so I may anticipate its coming and without more ado greet it with
xaipe’. That is said at first in the traditional colourless way, so that there is
hardly any thought of a definite rejoicing or welcome. But then the Elders
are overcome again by the feeling of all that their fear foresees ; they become
conscious in a deeper sense of what they have just been saying, and correct
their utterance. In ordinary life, the counterpart of a friendly greeting with
χαῖρε or xatpew λέγω is a malediction with οἴμωζε or οἰμώζειν λέγω (the latter,
e.g., in Ar. Plut. 58) or something of the sort. στένειν in mpoorevew is merely
an equivalent, suited to the more exalted style, for the vulgar οἰμώζειν. ἢ
Thus the Chorus say (to use a simplifying paraphrase): '. . . may be bid
welcome in advance; that is however all one with being cursed in advance’.
It is ἴσον, not because the two attitudes are the same in themselves, but
because for the future which is in any case unalterable it makes no difference
at all whether one says to it beforehand χαῖρε or στένε (οἴμωζε). What the
Elders seem to imply is this: in reality, in a case like this, lamentation would
be more appropriate than rejoicing. After this short excursus, the thought
returns with τορὸν yap ἥξει back to τὸ μέλλον ἐπεὶ γένοιτ᾽ ἂν κλύοις.
251. We might perhaps be content with excising the interpolation (see above)
and then reading τὸ μέλλον ἐπεὶ κτλ. as, e.g., Kirchhoff, Verrall, and Wilamo-
witz? do. The asyndeton could probably be justified, and the syllaba anceps
at the end of τὸ μέλλον would be unobjectionable (cf. on 229). My reasons for
preferring the slightly smoother text produced by the insertion of δ᾽ are not
particularly strong.
254. It is fairly safe to deduce from the discrepancy in the MSS that the
reading of the archetype was σύνορθρον. For the formation cf. Thuc. 2. 3. 4
τὸ περίορθρον.
Topóv ... σύνορθρον αὐγαῖς" ‘full clear with the rays of morning’ (Head-
lam). I cannot see why that should be ‘hopelessly vague’ as A. Y. Campbell
says (Cambridge Univ. Reporter, 9 December 1936), in support of Verrall.
1 Cf. in Sophocles’ snachus, Pap. Tebt, iii. 1, 692. 69 (= D. L. Page, Greek Lit. Pap. i.
p. 24) εἶπον... αἰάξαι, where the everyday οἰμώζειν or κλάειν ‘is replaced by αἰάζειν which
is usual only in Tragedy’ (R. Pfeiffer, Sitzgsb. Bayr. Akad. Phil.-hist. Abt. 1938, Heft 2. 53).
2 At any rate in his edition of Aeschylus (1914). At an earlier stage he, too, adopted τὸ
μέλλον <8"), see the note at the end of his translation, Griech. Trag. ii. 117.
3 The corruption αὐταῖς for αὐγαῖς is found, e.g., also in 5. Phil. 1199 (Pind. P. 9. 62 is
uncertain). |
143
line 254 COMMENTARY
Wilamowitz draws from the almost proverbial phrase an inference ('illucescit",
cf. also his Interpr. 164), which has been rightly rejected by Kranz, Hermes,
liv, 1919, 312 (also by G. Méautis, Eschyle, 126 n. 2), who says: ‘the words
[ropóv yàp ἥξει xrÀ.] belong . . . very closely to what precedes; thus they are
used in a metaphorical sense and mean “the future is always revealed by the
morrow”, a proverbial expression as is shown by line 1240’ (compared by
Paley). Above all we ought to compare with ropov ἥξει σύνορθρον αὐγαῖς
1180 f. λαμπρὸς δ᾽ ἔοικεν (sc. ὁ χρησμός) ἡλίου πρὸς dvrodds . . . ἐσάιξειν (cf. the
note there), where λαμπρός corresponds to ropóv in this passage.
255. δ᾽ οὖν 'however' (Paley, Headlam) indicates a breaking off and begin-
ning afresh, cf. Denniston, Particles, 461. At the same time it is made clear
by οὖν (‘if only that is so’) that the following idea grows out of what precedes ;
that is true also of 34, 1568, and related passages.
εὔπραξις was rejected as not Greek by Lobeck, Phrynichus, sor, and
there seems now to be universal agreement that he was right, and that we
must read εὖ as a separate word ; on the linguistic point cf. Ernst Fraenkel,
Kuhns Zeitschr. xlv, 1913, 162 n. 1. But it is still necessary to warn the reader
against concluding the optative sentence with εὖ (in which case τἀπὶ τούτοισιν
would have to be taken as subject). As the attempts of Nägelsbach, Weck-
lein, and others show, that leads to intolerable violence in what follows. The
subject of the sentence is πρᾶξις, ‘completion, success, result’, a usage found
in Homer, especially in the Odyssey (cf. Leaf on 92 524), and in the Oresteia,
Cho. 814. πέλοιτο εὖ is uttered here as a wish for πρᾶξις as 500 εὖ πέλοι is for
προσθήκη: in both passages the verbal element in the subject still makes
itself felt. The use ‘without a subject’ is simpler, as in A. Suppl. 454 γένοιτο
δ᾽ εὖ (compared by Klausen), Suppl. 122 πελομένων καλῶς, 'if all goes well’
(Wilamowitz, Interpr. 32); the latter passage was quoted by Miss H. L.
Lorimer, C.R. xlv, 1931, 211, who also points out how rare is the construction
of πέλω or πέλομαι with an adverb. For such a use without subject cf.
perhaps the absolute use of bene sit (Petronius 92. 3 oportet hodie bene sit,
more of the same kind in Thes. 1. L. ii. 2114. 42 ff.).
τἀπὶ τούτοισιν is rightly rendered by L-S, s.v. ἐπί B. ii. 2, as quod superest,
and classed with Hdt. 9. 78. 2 σὺ δὲ kai rà λοιπὰ τὰ ἐπὶ τούτοισι ποίησον, where,
however, the expression is simply the object of the verb, as it is (with slightly
different meaning) in Thuc. 1.65. 1 and 7. 62. 3; the similar phrases collected
by Blomfield in his Glossary also serve as object or subject. But the adverbial
function of τὰ ἐπὶ τούτοις here, as an 'accusative of respect', is perfectly good
Greek, cf., e.g., the expression τὰ νῦν τάδε, rightly recognized by the scholiast
on Ar. Peace 858 as idiomatic (references are given by Elmsley on E. Herachd.
641 and by Stein on Hdt. 7. 104 1. 6), which is used adverbially by Euripides
as well as by Aristophanes, further S. Oed. C. 133 τὰ νῦν.
ὡς at the end of the line (there is no συνάφεια here) is unobjectionable:
Aeschylus places it even at the end of a dialogue-trimeter (cf. on 1354).
256 f. τόδ᾽ äyxıorov . . . ἕρκος. The error of the scholion (ἐπεὶ μόνοι
γέροντες ἐφύλασσον τὴν ᾿Ελλάδα) long continued active; but that explanation
is quite out of place, and was strongly attacked by Hermann. Quite apart
from the thought, we need here on the analogy of the customary dramatic
technique some announcement of the queen before her first appearance
(Headlam). ἄγχιστον has often been understood spatially, so Schütz (‘Chorus
144
 COMMENTARY line 257
de regina loquitur, quam propius ad se accedere videt; eo pertinet τόδ᾽
ayxıorov’) and latterly Passow-Crönert, s.v., and Kranz, Hermes, liv, 1919,
312 n. 2. On the other hand, the scholion explains (in accordance with the
familiar use of ἀγχιστεῖς and ἀγχιστεία) τὸ συγγενικόν, and similarly, e.g.,
Schneidewin-Hense, Wilamowitz (translation), and Headlam (‘ ayxıorov
describes her relation to the throne’). With a metaphorical use of the terms
appropriate to inheritance, the poet describes things as though during the
king's absence the sovereignty had passed by some ayxıoreia—of a peculiar
sort, it must be admitted—to the queen. Neither here nor farther on in this
play are we given any unnecessary picture of the constitutional position, but
so much at least is clear that Clytemnestra exercises a power which may be
termed regency. This conception is in keeping with certain hints, faint though
they are, in Homer. Cf. on this H. M. Chadwick, The Heroic Age (1912),
359: ‘It has often been remarked that the position of women in the Homeric
poems appears to be one of greater influence and responsibility than any-
thing we find in later times. But nowhere is this responsibility made so clear
as in the absence of all evidence for the constitution of a regency when the
king is away from home’; ibid. 386: “The cases of emergency arising out of
the misfortunes of Odysseus and Agamemnon bring to our attention another
curious feature... namely that the king does not seem to appoint a regent in
his absence. . .. Are we to suppose that the queen is the person in authority?'
Perhaps we may also consider a suggestion put forward by Daube, 32: ‘The
justification of Clytemnestra’s rule on the ground of ἀγχιστεία certainly does
not square with Attic law. Aeschylus may have been aware however that
Semiramis and his own contemporary Artemisia of Halicarnassus succeeded
to their husband’s rule after their death. He invents a similar succession for
Clytemnestra, which is called into operation, however, not by death but by
the monarch’s absence from the realm.’ Nor should we forget that according
to the view adopted by Aeschylus the widow of Darius as queen-mother
seems to exercise the regency during the absence of Xerxes; the Chorus call
her not only (Pers. 152) βασίλεια ἐμή, but actually (173) γῆς ἄνασσα τῆσδε.
This, and possibly other oriental analogies, may have helped the poet to form
his picture of the position of Clytemnestra. In general cf. Conington on
Cho. 360: ‘in his [Aeschylus’] days the idea of sovereignty had become more
or less associated with that of Oriental despotism, so that Agamemnon would
appear to him a sort of reduced type of Xerxes’.
’Amtas: with reference to the land of Argos, already in the Supplices.
257. ἕρκος, used of the person, is Homeric.
Between the sombre thoughts about the future expressed in 249-54 and the
prayer that follows them there is apparently a contradiction; this has been
pointed out by A. S. F. Gow, C.Q. viii, 1914, 1 (he finds a solution by declaring
τὸ μέλλον corrupt). In actual fact the behaviour of the Chorus here is just as
full of contradictions and just as natural as under certain conditions the
behaviour of men at their prayers often is. Besides, the contradiction is by
no means confined to the final antistrophe but runs through the whole ode,
indeed through the whole play up to the catastrophe (Bamberger, Opusc. 50,
well says on the fourth choral ode [975 ff.]: ‘optat chorus ut metus suus
irritus sit; non irritum fore persuasum habet’.) The Elders know well
enough, as thinking men, experienced in God’s ordinances and the way of the
4872,2 L 145
line 257 COMMENTARY
world, that the sins which have been committed must bring in their train
inevitable punishment and ill; nevertheless they wish, as human beings filled
with apprehension, that all may yet turn out for the best, and wishfulness
and anguish together make up their prayer. atAwov αἴλινον εἰπέ, τὸ δ᾽ εὖ
νικάτω. When the emotions thus succeed for a time in silencing the intellect,
the result perforce can only be a ‘hoping against hope’. With complete
clarity this disagreement is brought out by Aeschylus in Suppl. 454, where the
king after ordaining sacrifices says γένοιτο δ᾽ εὖ παρὰ γνώμην ἐμήν. The
‘contradiction’ in Ag. 998 ff. is quite as sharp. There the content of two whole
stanzas is made up, not of forebodings only, but of definite expectations that
fresh evil must inevitably come; against this fear the apparent signs to the
contrary are of no avail. And yet there follows at the end the prayer εὔχομαι
δ᾽ ἐξ ἐμᾶς ἐλπίδος ψύθη πεσεῖν és τὸ μὴ τελεσφόρον.

Some points relating to this chorus as a whole have been mentioned in the
notes. In addition to them, attention may be drawn to a remarkable feature
which has been noticed by several readers’ and particularly well illustrated
by W. Ferrari, La parodos, 377 f., 398. The story is permeated and indeed
dominated by a strong antagonism. A portent assures the Greeks of the final
success of their undertaking, but at the same time forebodes a grave hindrance
caused by the wrath of a deity in consequence of which the leader of the army
will be forced to commit an unforgivable crime. Zeus himself sends the
portent, Zeus who sends the sons of Atreus against the Trojans (6o ff.) ; yet
the avengers whose cause he favours shall not be allowed to avoid pollution.
According to the will of Zeus the capture of Troy is a just punishment ; but
the cruelty of the execution is emphasized ; to this we may add from a later
song the still more general condemnation (461) τῶν πολυκτόνων ydp οὐκ
ἄσκοποι θεοί, which refers primarily, though not solely, to the losses among the
Greeks. When we look back at the anapaests of the parodos, we see that in the
very same sentence (6o ff.) in which the expedition is said to be the work of
Ζεὺς ξένιος it is also stated that Zeus in that war causes many deaths of Greeks
and Trojans alike zoAvdvopos ἀμφὶ γυναικός. No doubt, then, the Greeks are
fighting for a good and righteous cause (Arpews παῖδας... πέμπει ξένιος
Ζεύς), but that cause is linked up with a far less irreproachable motive.
‘For a woman's sake’ would be bad enough (cf. Suppl. 477), moAvavopos (cf.
the note on 62) makes it much worse. When the issue is put like this, there
can be no question about the verdict: it is the reverse of the gentle excuse
made by the bewitched old men on the wall of Troy, οὐ νέμεσις Τρῶας xai
ἐυκνήμιδας Ayaods τοιῆιδ᾽ ἀμφὶ γυναικὶ πολὺν χρόνον ἄλγεα πάσχειν. Thus,
from the first origin of the war, good and evil are intermixed everywhere:
τί τῶνδ᾽ ἄνευ κακῶν ; The clash of right and wrong, of fears and hopes, finds
a moving expression in the refrain αἔλινον αἴλινον εἰπέ, τὸ δ᾽ εὖ νικάτω. The
wish for the triumph of the good is echoed in the last stanza of the chorus
(255): πέλοιτο δ᾽ οὖν τἀπὶ τούτοισιν εὖ πρᾶξις, where it is preceded by its sin-
1 See, e.g., H. Weil, Etudes sur le drame antique, 31: ‘Les quatre scènes et les quatre
chœurs qui ouvrent cette grande tragédie ont cela de commun qu’à la joie du triomphe se
mèlent constamment de tristes souvenirs et de sombres pressentiments. Le spectateur ne
peut oublier un instant le refrain de la prophétie de Calchas: Αἴλινον αἵλινον εἰπέ, τὸ δ᾽ εὖ
# 3
νικάτω.

146
 COMMENTARY line 259

ister counterpart (250) : Aika δὲ τοῖς μὲν παθοῦσιν μαθεῖν émippére ; the wording
here directly recalls the most momentous passage of the hymn to Zeus. All
the bewildering antagonisms in the struggles of mortal men originate in the
rule of Zeus. He does not offer a soothing solution of the disharmonies:
βιαίως σέλμα σεμνὸν ἡμένων are the last words of the hymn that praises him.
But he is trustworthy; the law which he has set up shall never be broken
(178), and that law is just: Δίκα τοῖς παθοῦσιν μαθεῖν ἐπιρρέπει: Justice, ful-
filling the will of Zeus. This is the only certainty in the midst of-endless con- .
flicts and perplexities. The whole trilogy circles round this centre. The
ultimate triumph of Zeus’ justice, manifested in the trial of the Euwmenides,
will be the last link in the chain whose beginning unrolls in the monumental
sentences of the greatest of all odes in Greek tragedy.

258. Wilamowitz’s note runs: ἄγγελος praefigunt codd.’ But in M ἄγγελος

is an addition of m on the left hand of the ‘paragraphos’ (—) ; cf. in addition
to the note in Vitelli-Wecklein’s edition (‘ut videtur’) Rostagno in the Intro-
duction to the facsimile of M, p. 13 c. Originally M had here only the para-
graphos (as also often elsewhere in similar places, cf. on sor). Down to 281
the person-names are confused in the MSS, see critical apparatus.’ The
editors of the sixteenth century (including Canter) put up with the unbear-
able arrangement according to which the whole of the ensuing dialogue takes
place between Clytemnestra and a ‘Messenger’. Casaubon removed the error,
and Stanley called attention to the structure of the parodos of the Persae,
which is very similar in form and content ; there (155) the parodos is followed
by a speech of the coryphaeus in dialogue-metre.
σὸν Κλυταιμήστρα κράτος. The coryphaeus says politely : ‘I am come here
to show my respect for your sovereignty.’ He does not say outright : ‘because
you have given orders that we should appear here in front of the palace’,
but that is probably the sense of the words. Cf. on 10 κρατεῖ. It is a natural
assumption that Clytemnestra sent round a messenger to inform the Elders
that she wished to see them and give them information at the same time as
she made arrangements for the offering of sacrifices in the town (87 ff.).
The author of the Hypothesis understood the action in this way : αὐτὴ δὲ τῶν
πρεσβυτῶν ὄχλον μεταπέμπεται, περὶ τοῦ πυρσοῦ ἐροῦσα, ἐξ ὧν καὶ ὁ χορὸς
συνίσταται. This assumption is not contradicted by the words (263) οὐδὲ
σιγώσηι φθόνος, which need be nothing more than an expression of especial
politeness. The assumption that the old men appeared of their own volition
to ask the reason for the sacrifice in the town which the queen has decreed is
less probable, if not actually impossible. In this case also the sense of 258
would be: “Your κράτος (1.6. your royal position, in virtue of which you have
ordained the sacrifices) has caused my coming.’ This use of κράτος cannot
therefore be put on a level with the examples of respectful 'periphrasis'
quoted by Headlam (and A. D. Knox) on Herodas 5. 68.
258. The ensuing question (261 ff.) repeats the earlier questions which were
addressed to the queen before she appeared (85 ff.) ; 262 θυηπολεῖς takes up
87 θυοσκεῖς exactly.
259. δίκη: cf. on 811.
1 Owing to a mistake (or misprint) Wilamowitz asserts that M assigns 263 to KAvr. This
has unfortunately found its way into the Oxford text.

147
lines 259 f. COMMENTARY
259 f. φωτὸς ἀρχηγοῦ: cf. on 1627. The sentence is, as it were, an explana-
tion of ἄγχιστον 256. The speaker mentions in general terms the respect
which is due to the queen in a special degree during the temporary vacancy
of the throne.
261. In regard to the εἴ τι κεδνόν of Auratus, Hermann remarks: ‘estque id
aptius, licet defendi possit εἴτε. Nam potius hic est, num quid fausti acceperit
quaeri.’ Against this Wilamowitz, who keeps εἴτε κεδνόν, argues: ‘non
poterat dicere “sive boni aliquid sive non boni"; sed causa iusta sacrificii
nuntius faustus est; quem si non accepit, vana spe illa effertur'. But these
polemics miss the mark; the thought remains the same whether εἴτε xeóvóv or
εἴ τι κεδνόν is Written, and it is impossible to arrive at the meaning sive non
boni because in any case εἴτε μὴ πεπυσμένη remains unaltered. The normal
expression rz κεδνόν seems to be preferable. If, on the other hand, «e8vóv by
itself is to be defended, nothing is gained by Verrall's parallels, θαυμαστὸν
ποιεῖς, ἄτοπον λέγεις, 1.6. instances of internal object.
The sentence o? δ᾽ εἴ τι κτλ. is so constructed that at first the intention of
the speaker seems to be εἴ τε κεδνὸν πεπυσμένη εἴτε μὴ πεπυσμένη θνηπολεῖς,
and then the clause εἴτε μὴ πεπυσμένη θνηπολεῖς is elaborated by the addition
of the explanatory εὐαγγέλοισιν ἐλπίσιν, which disposes in advance of the
possibility of the queen being offended by the suggestion of μὴ πεπυσμένη
θυηπολεῖς.
262. ' εὐαγγέλοισιν ἐλπίσιν pro ἐλπίδι εὐαγγελίας ᾿ (Schütz). Blomfield compares
E. Med. τοιο δόξης δ᾽ ἐσφάλην εὐαγγέλου.
263. εὔφρων of loyalty to the ruler also 797, 806 (cf. ad loc.), also here in the
following 271 εὖ φρονοῦντος. Cf. Cho. 109 τοῖσιν εὔφροσιν (Schol. τοῖς εὖ
φρονοῦσι τῶι Ayapeuvovi δηλονότι).
οὐδέ: on this use οὗ οὐδέ ‘as a balancing adversative which sets a negative
idea in the scale against a preceding positive idea’ cf. Denniston, Particles,
191 ; W. L. Lorimer, C.R. xlviii, 1934, 221 ; xlix, 1935, 14.
264. εὐάγγελος : an echo of εὐαγγέλοισιν 262, just as 266 μεῖζον ἐλπίδος refers
back to 262 ἐλπίσιν. This method of linking choral ode or speech of Chorus
with the actor’s response is clearly very old; cf. W. Kranz, Neue Jahrbücher
f. d. Klass. Altert. xliv, 1919, 157, and the same author's Stasimon, 23. Cf.
note on 1412 f. For the beginning of her reply Clytemnestra chooses a word
that is bon; ominis.
ὥσπερ ἡ παροιμία as in Soph. fr. 260 N. (= 282 P.).1, Eur. fr. 668. τ N.,
cf. Plato Com. fr. 174. 3 K. καθάπερ ἡ παροιμία. 1 have been unable to find
anything in the grammars on the ‘ellipse’ which is normal in this kind of
phrase, ' cf., e.g., Alcaeus fr. 74. 1 D. ὡς λόγος (contrast fr. 129 D. ws λόγος ἐκ
πατέρων ôpwpe), À. Suppl. 230 ὡς λόγος (also E. Iph. T. 534, Phoen. 396), Eum.
4 ὡς λόγος ris, Hdt. 1. 75. 3 ὡς δὲ ὁ πολλὸς λόγος “Ελλήνων, Cratinus fr. 169 K.
ὡς ὁ παλαιὸς λόγος, fr. 228 K. ὡς μὲν ἀνθρώπων λόγος, Plat. Laws 715 e ὥσπερ καὶ ö
παλαιὸς λόγος, 948 Ὁ ws γε λόγος, Demosth. 23. 66 ὡς λόγος, S. Oed. R. 715
ὥσπερ γ᾽ ἡ φάτις, Ant. 829 ὡς φάτις ἀνδρῶν.
We do not know the proverb to which Aeschylus alludes. It may have
said quite simply ‘like mother like child in beauty’ or something of this
kind. It is natural to regard morning or sun or day as the child of the
1 Römer, Sitzgsb. Bayr. Ak., Phil.-hist. Kl., 1888, 240, notices the usage, but takes only
Aeschylus into account.
148
 COMMENTARY line 268

preceding night (Hesiod, Theog. 124 Νυκτὸς δ᾽ αὖτ᾽ Αἰθήρ re καὶ 'Hyépm
ἐξεγένοντο) ; 279 refers back to this idea. S. Trach. 94 ff. is compared by
Eustathius (see app. crit.). The theme was elaborated by Theodectes, from
whose Oedipus (p. 802 N.) Hermippus ap. Athen. τὸ p. 451 f quotes this riddle
(together with the solution): εἰσὶ κασίγνηται δισσαί, ὧν ἡ μία τίκτει τὴν ἑτέραν,
αὐτὴ δὲ τεκοῦσ᾽ ὑπὸ τῆσδε τεκνοῦται. H. Usener, Ki. Schr. iv. 387 f. (quoted
by Radermacher on S. Trach. 94 f.), compares a Slovak fairy-story with
Ag. 265 and Trach. 94: there the sun says ‘My mother gives birth to me anew
every morning as a beautiful boy and every evening she buries me as a weak
old man’. Cf. the riddle in Anatole France, Sainte Euphrosine (in 'L'Étui de
nacre’): ‘J’engendre ma mère et je suis par elle engendré. . . .' Solution:
‘Celui qui engendre sa mére et qui est engendré par elle n’est autre que le
jour’. Clytemnestra knows that the night has brought good news just now.
‘May the future correspond to this’ is her wish here and at the end of the
whole scene, 349 f. It is very doubtful whether Aeschylus is making in
εὐφρόνη an etymological allusion (cf. on 681 ff.) to the ‘cheerfulness’ or ‘good
will’ of night, as is assumed by, e.g., Schneidewin, Sidgwick, Kaibel on S. El.
17 (there Housman, C.R. ii, 1888, 244, also takes εὐφρόνη as εὐφροσύνη), and
Verrall. The use of the word elsewhere does not suggest this, cf. J. Vahlen,
Opusc. ii. 507 f. The bare gloss in Hesychius εὐφρόνη" νὺξ καὶ εὐφροσύνη does
not allow of reliable conclusions. There is no reason to suppose that
εὐφρόνης is used here in order to echo the εὔφρων of 263 (so Schneidewin).
266. Cf. the lyric poem of the classical period, Pap. Soc. It. 1181 (= D. L.
Page, Greek Lit. Pap. no. 84) 32 ἀέλπτωι περὶ χάρ[μαΪτι.
267. Πριάμου... πόλιν: the separation marks emphasis, cf. 1335.
ἠϊιρήκασιν : in early tragedy this perfect of result is still rare, cf. J. Wacker-
nagel, Studien z. griech. Perfectum (Programm Göttingen 1904), p. 11.
268. πῶς φής; πέφευγε τοὖπος ἐξ ἀπιστίας. The piling up of p-sounds? is
probably meant to express the breathless excitement of the questioner, as in
Ar. Birds 319 ποῦ; πᾶ; πῶς φής; and S. Oed. C. 1099 ποῦ ποῦ; ri φής; πῶς
εἶπας; It is therefore unlikely to be pure chance? that further questions of the
Chorus in this stichomythia also contain ?-alliterations, although to a
smaller degree: 274 πότερα φάσματ᾽ εὐπειθῆ, 276 Emiavev τις ἄπτερος φάτις, 278
ποίου... πεπόρθηται πόλις. Extensive -alliteration also 820, 1669, cf., e.g.,
Pers. 681 f. likewise in an excited question, Prom. 98 f. (including φεῦ φεῦ),
Ag. 489 f. λαμπάδων φαεσφόρων φρυκτωρίας τε καὶ πυρὸς παραλλαγάς, Eum.
249 f., S. A7. 866 f. (including φέρει), 1112, El. 210, 504 f., E. Bacch. 1298 (in
passionate excitement) τὸ φίλτατον δὲ σῶμα ποῦ παιδός, πάτερ; Ar. Birds 321,
Herodas 2. 28 f. (see Groeneboom, ad loc.). On alliteration in Aeschylus cf.
W. Porzig, Aischylos, 76 ff., W. B. Stanford, Aesch. in his Style, 81 ff. ; further
literature in W. Schmid, Gesch. d. griech. Lit. ii (1934), 297 n. 2, and T. B. L.
Webster, Introduction to Sophocles (1936), 161 n. τ. Blomfield on A. Cho. 51x
called attention to the strong r-alliteration (‘paene Sophocleum’).
! The hypothesis that εὐφρόνη in Aeschylus can be identical with εὐφροσύνη is the basis
of an extremely bold conjecture by Headlam on the very corrupt line Che. 417.
2 The alliteration is the stronger if the assumption is correct that in fifth-century Athens
$ was still heard as a true aspirated mute and not yet as a spirant. On this complicated
problem cf. Schwyzer, Griech. Gramm. i. 204 ff.
3 It must, however, be admitted that there are instances in Greek poetry where strongly
marked z-alliteration seems to be accidental, cf., e.g., Leaf on I" 50.
149
line 269 COMMENTARY
269. Τροίαν ᾿Αχαιῶν οὖσαν scil. λέγω. Answer to πῶς φής; On this accusa-
tive and participle after verbs of saying cf. Jebb on Soph. Oed. R. 463,
Goodwin, § 910, p. 361, Kühner-Gerth, ii. 72 n. 2; cf. also below on 583.
Some of the not very numerous instances give the impression that the
participial construction is much more emphatic than the infinitive: thus
Homer # 2 φίλον πόσιν ἔνδον ἐόντα, ‘Odysseus is really at home’. There seems
to be the same emphasis in our passage and in the similar S. El. 676, where,
again in stichomythia, the same question {τί φής, τί φής;) is answered:
θανόντ᾽ ᾽᾿Ορέστην νῦν τε kai τότ᾽ ἐννέπω.
271. εὖ φρονοῦντος: cf. on 263.
yap: ‘Ja, dein treuer Sinn spricht sich in deinen Augen aus’ (Nägelsbach).
On this common ‘assentient’ γάρ cf. Denniston, Particles, 86 f.
κατηγορεῖ: on κατήγορος (Sept. 439) and κατηγορεῖν = ‘indicator, indicate’
cf. Schadewaldt, Hermes, lxxi, 1936, 42. Headlam quotes later examples,
but it is not necessary to agree with him in seeing the influence of the technical
language of the physiognomists here.
272. The MSS have τί γὰρ τὸ πιστόν ἐστι without punctuation (Wilamowitz’s
apparatus is incorrect). Schütz saw that this was not in keeping with the
reply in 273 and suggested: ri γάρ; τὸ πιστὸν κτλ. Hermann followed him.
But then the article τό could not be accounted for, as was pointed out by
Ahrens (p. 479). The passage which has been quoted in support—S. Trach.
398 Ÿ καὶ τὸ πιστὸν τῆς ἀληθείας veueîs;—is obviously no parallel. The
punctuation of Prien: ri γὰρ τὸ πιστόν; ἔστι τῶνδέ σοι τέκμαρ; still finds favour.
But against this Ahrens has rightly remarked that πιστόν should not be
severed from τέκμαρ, since πιστὰ τεκμήρια in 352 is an express reference to
this passage. Only two of the conjectures which have been made (and it is
unlikely that other possibilities exist in this case) remain open: (r) to write
with Ahrens ri γάρ; τὶ πιστόν ἐστι τῶνδέ σοι τέκμαρ; Meineke’s doubt as to
the correctness of ri ydp; (Philol. xix, 1863, 194) is unjustified; he says, one
would expect ri δέ; quid vero? but not quid enim? In fact ri γάρ would be
entirely suitable here. Cf. Jebb on S. Oed. C. 542 (‘how then?’ to mark the
transition from one subject to another) and Denniston, Particles, 83 (‘purely
transitional’). For the word-order τὲ πιστὸν... τέκμαρ cf., e.g., Suppl. 679 τις
ἀνδροκμὴς Aovyós, Ag. 276 τις ἅπτερος φάτις. But that a sentence should begin
with the unemphatic rı is unlikely. (2) Therefore we must decide in favour of
the other conjecture and write with Karsten and Meineke: ἦ ydp τι πιστόν
ἐστι krÀ., cf. Prom. 745; Denniston, Particles, 282. On the common inter-
change of H and TI cf. Cobet, Novae lectiones, 368.
273. On τί δ᾽ οὐχί cf. on 557.
274. εὐπειθῆ: Blomfield, who here and in several other passages wrote
eom instead of the εὐπειθῆ of the MSS, is still followed by the editors. The
change is not, however, justified, as was shown by Ahrens, p. 480, cf. also
Blass on Cho. 259 and the cautious survey in L-S. In Ag. 983 εὐπειθές is
demanded by the metre, in Prom. 333 εὐπιθής. Probably Aeschylus could use
either form at will, just as his ἀπειθεῖν Ag. 1049 stands beside the Homeric
ἀπιθεῖν (Ahrens, Blass). Since in the majority of cases the middle syllable is
metrically indifferent, we have no means of distinguishing where the 'high-
grade’ form with εἰ and where the ‘low-grade’ form with « was used (cf.
Wackernagel, Vermischte Beiträge, 16). There is no reason to suppose that
150
 COMMENTARY line 276
the difference of form served to distinguish the two meanings of the word, i.e.
(1) ‘qui facile persuadet’, (2) ‘cui facile persuadetur, qui facile obsequitur’.
φάσματ᾽ εὐπειθῆ appears to be an instance of the former meaning, as Cho.
259 σήματ᾽ εὐπειθῆ. The Choephoroe passage also shows that εὐπειθῆ in Ag.
274 should not be taken predicatively (so Schneidewin and Verrall}. The
coexistence of active and middle-passive meaning seems to be very old, as
Ἐὐπείθης, the father of Penelope’s suitor Antinoos (like Εὐαγόρας, Apıoro-
πείθης, Ἀρισταγόρας etc.), is so called because he εὖ πείθει (in the assembly of
men), while Διοπείθης is so called because he τῶι Au πείθεται (list of names in
-πείθης in Bechtel, Histor. Personennamen, 366).
275. δόξαν: ‘fancy’; sometimes the meaning ‘vision’ is uppermost, e.g. 421,
Cho. 1051, 1053.
λάβοιμι is misunderstood by many. Headlam preferred to adopt Karsten’s
λάκοιμι. But ‘I should not scream out the mere fancy of a slumbering brain’
introduces a false idea; the sense of Clytemnestra’s answer to the question
in 274 must be that she rejects dreams as the source of her conviction. With
the expression οὐ... dv λάβοιμι Ahrens (p. 481) rightly compares Ar. Wasps
508 f. ἐγὼ yàp οὐδ᾽ dv ὀρνίθων γάλα ἀντὶ τοῦ βίου λάβοιμ᾽ àv οὗ με νῦν ἀποστερεῖς,
where λαβεῖν means ‘receive (as equivalent)’, almost = ‘buy, allow to be
sold to one’, as in Ar. Peace 1263 λάβοιμ᾽ ἂν αὔτ᾽ eis χάρακας ἑκατὸν τῆς δραχμῆς,
Frogs 1236 λήψει γὰρ ὀβολοῦ πάνυ καλήν τε κἀγαθήν and elsewhere. Nearly
allied to the passage in the Wasps is Clouds 1395 f. τὸ δέρμα τῶν γεραιτέρων
λάβοιμεν ἂν ἀλλ᾽ οὐδ᾽ epeBivOov." It is clear therefore that οὐκ ἂν λάβοιμι...
‘I would not buy it, permit it to be sold to me’, or ‘I would not give a farthing
for it’, is a turn of speech belonging to colloquial Attic.? The expression here
has a tinge of the vulgar and is certainly not very polite. Clytemnestra is
provoked by content and form (φάσματ᾽ εὐπειθῆ) of the preceding question,
and here we may sympathize with her pride in her cleverness and discern-
ment. In her anger she descends to a rather lower level of speech with οὐκ av
λάβοιμι and at the same time chooses definitely polemical expressions in
δόξαν and βριζούσης φρενός. The sharpness of tone is continued by both sides
in the two following lines, then the coryphaeus turns to something new (278).
276. ἀλλ᾽ 4: cf., e.g., Cho. 220 (and, if Blass’s conjecture is correct, Eum.
426) ; in general cf. Denniston, Particles, 27. Commonly used ‘in statements
or hypotheses to which the speaker wishes to represent himself as driven
rather against his will, as if they alone were possible’ (Conington on Cho.
774). ‘It generally means ‘‘Perhaps?’’, “I hope not”, asking a question in
hope of a negative answer’ (Neil on Ar. Knights 953). Notice that the in-
tended double question which began in 274 with πότερα is broken off and is not
continued by the expected 7 but by the more strongly deprecating ἀλλά. A
brief note on the point will be found in Schneidewin-Hense and Wecklein
on this passage, at greater length Kaibel on 5. El. 534 ff., E. Bruhn, Anhang
zu Soph. $ 217. They quote S. Aj. 460-8, El. 534 ff, Xen. Anab. 5. 8. 4.
! The passage in the Clouds agrees almost exactly with Plautus, Cas. 347 non ego istud
verbum empsim littibilicio; Mil. 316 non ego tuam empsim vitam viliosa nuce; cf. also Poen.
274 quotus ego nebulai cyatho septem noctes non emam. Incidentally a good parallel for the
meaning of λαμβάνειν (nearly = buy) is given by emere, which moves from ‘take’ (preserved
in the isolated imperative em(e) and the compounds) to ‘buy’.
2 A more elevated version of the same phrase is found in S. Aj. 477 f. οὐκ ἂν πριαίμην
οὐδενὸς λόγου βροτὸν ὅστις xevataw ἐλπίσιν θερμαίνεται.
I5I
line 276 COMMENTARY
For similar constructions in other languages cf. Ernst Fraenkel, Glotta, iv,
1913, 48 1.
ἐπίανεν. When this verb is used metaphorically of mind and spirit, it is
contemptuous ; e.g. Ag. 1669; Pind. P. 2. 56; Bacchyl. 3. 68: in these passages
the ‘fatty degeneration’ is also shown in the fact that the power to form
sharp, clear judgements has been impaired (the Roman speaks of pingue
ingenium). The depreciatory expression, in which association with pig
(Semonides of Amorgos fr. 7. 6) or ox (Ananios fr. 5. 9) seems still to be felt,
sets the tone of the whole verse: this must be taken into account in inter-
preting ἄπτερος. Clytemnestra’s answer is correspondingly sharp. Cf. on 275.
ἄπτερος. Thedifficulty of understanding this word here was felt inantiquity ;
that is shown by the fragment of a scholion to this passage which is preserved
in Hesychius s.v. It is probably impossible to say with certainty what the
poet here meant. Perhaps there were already current in his time as later on
(cf. Wilamowitz, Isyllos, 112 τι. 6) several interpretations of the phrase used
in the Odyssey τῆι δ᾽ ἅπτερος ἔπλετο μῦθος. The most detailed treatment of
the word and its meaning in this passage (with a synopsis of earlier literature)
is in Ahrens, pp. 481 ff.; Headlam and those who followed him in the dis-
cussion (C.Q. xxx, 1936, 1 ff., 105, 151 f.)' have considered neither Ahrens nor
the observations of Wilamowitz. Schuursma, De . . . abusione, 106, reserves
his judgement. The view that ἄπτερος must be understood here in the
normal sense? of unwinged was forcibly upheld by Hermann,’ who rejected
other explanations of the word. He used as his main argument the normal
1 In the first of these three articles J. A. K. Thomson endeavours to prove that the phrase
τῆι δ᾽ ἄπτερος ἔπλετο μῦθος in the four passages of the Odyssey where it occurs means ‘the
point of his remark escaped her’ because (as Thomson adds by way of explanation) she was
not in the secret, was not in the plot. This thought is neither expressed nor implied in any
of the Homeric passages. What the context in each of them clearly indicates is this: ‘she
did not reply (but acted as she was told)’. This the poet expresses by saying ‘to her the
thought was without wings, i.e. remained unuttered’. For a detailed discussion see Ameis-
Hentze on p 57 (cf. especially the ‘Anhang’ ad loc., fasc. iii, p. 108 f.) and Ernst Fraenkel,
Glotta, ii, 1910, 29 n. 1. The ancient interpretation which referred μῦθος to the preceding
speech of Telemachus was adopted by many modern scholars (among them Wilamowitz,
loc. cit.), but was rightly abandoned by Doederlein (quoted in Dindorf’s Thesaurus, i. 2,
p. 1831 at the bottom), who was followed by the two scholars I have just mentioned and,
e.g., Monro on p 57, Autenrieth-Kaegi and L-S s.v. ἄπτερος, Hermann Frankel, Die homer.
Gleichnisse, 80.
J. A. K. Thomson holds that in Ag. 276 as in the instances of ἄπτερος in the Odyssey the
underlying image is that of an unfeathered, i.e. a random-flying, arrow; he consequently
translates ἄπτερος φάτις by ‘aimless rumour’. As he regards it as ‘certain that ἄπτερος
φάτις is a synonym of ἄπτερος μῦθος ', the breakdown of his interpretation of the Homeric
passages would seem to be fatal to his view on Ag. 276. But it might be possible to maintain
this view by arguing that, although the image of the arrow is incompatible with ἄπτερος
μῦθος as used by Homer, and although there is no trace of this interpretation in the com-
ments of the ancient grammarians which we possess, perhaps Aeschylus may have taken
the difficult Homeric phrase in the sense suggested by Thomson. To me this seems very
unlikely.
2 ie. with d- as στερητικόν. The view that d- is to be taken as ἀθροιστικόν is clearly ex-
pressed, e.g., in the rendering ἐσόπτερος which the scholion on Ag. 276 has in common with
schol. Hom. p 57 and Hesychius s.v. ἄπτερα.
3 Clemm in G. Curtius’s Studien zur griech. u. lat. Grammatik, viii (1875), 83, and Ernst
Fraenkel, Glotta, ii, 1910, 29 n. 1, agree with Hermann's interpretation ; the latter separates
completely Hesiod’s ἀπτερέως (see below) from ἄπτερος, Pohlenz, too, decides for ‘unwinged”
(Gnomon, ix, 1933, 625).
152
 COMMENTARY line 282

meaning of the word in Tragedy; E. Her. 1039 provided him with a parallel
for the particular nuance of meaning which he assumed in this line, implumis.
Against this Ahrens argues: ‘an unfledged rumour could at best only be one
which was too weak to leave its immediate neighbourhood and could not
possibly therefore have come from Troy to the ears of Clytemnestra’. Ahrens
himself starts from the ancient explanation (cf. Apollonius Sophista, p. 41. ı)
preserved in Hesychius and elsewhere, αἰφνίδιος or raxVs,! and understands
here a swift-sped rumour. So also e.g. Th. Bergk, Griech. Lit. Gesch. iii. 356
n. 201, and Headlam (‘wing-swift’). If it is a necessary assumption that
Aeschylus here follows the authority of older poetry and its early inter-
preters, this meaning, which is perhaps supported by the ἀπτερέως of Hesiod
(Parmenides, Apoll. Rhod.), seems appropriate for our line. A ‘swift-sped’
rumour flies up somewhere and sweeps rapidly from place to place, but if
anyone tries to hold it, it is gone: ταχύπορος" ἀλλὰ ταχύμορον . . . ὄλλυται
κλέος (486 f.). Here, where we expect the adjective to express depreciation,
or at least some scepticism, this meaning seems more suitable than προσηνής,
which appears in ancient glossography as a rival explanation and is preferred
by Wilamowitz, with whom Wackernagel agrees, Kuhns Zischr. xxxiii, 1895,
49. But, as has been said, we are here dealing with pure conjecture.
277. Prom. 986 has long been adduced for comparison: Beazley recalls
Theogn. 254 ἀλλ᾽ ὥσπερ μικρὸν παῖδα λόγοις μ᾽ ἀπατᾶις.
278. On the idiomatic use ποίου χρόνου (for which Karsten and others wanted
to write πόσου) see Passow s.v. ποῖος, Dindorf in Thesaurus and L-S (ποῖος
iv), cf. also O. Schroeder on Ar. Birds 920. Note that Euripides, echoing this
line in Hel. 111 πόσον χρόνον yap διαπεπόρθηται πόλις, puts πόσος for ποῖος
and alters the case. On καί after an interrogative and the emphasis thereby
given to the question as a whole or to particular parts of it see Denniston,
Particles, 312.
279. τεκούσης: cf. on 264.
280. ἀγγέλλων (suggested by Stanley, who, however, himself says ‘sed non est
necesse’) spoils the sense. The audience was in no danger of joining ἀγγέλων
τάχος (so Wilamowitz in his note) : the Greek ear was accustomed to arrange-
ments of words much more intricate than this, where the two emphatic
hyperbata ris . . . ἀγγέλων and roôe . . . τάχος stand side by side.
281. For Hephaistos as the lord of every fire and for the very common
metonymic use of the name of this god cf. Wilamowitz, Kl. Schr. v. 2, p. 15 f.,
and the collection of examples in Malten, RE viii. 329.
282. Without the help of external evidence no one could have conjectured
that the MS reading ἀπ᾽ ἀγγέλου πυρός was wrong; the same is true of 284
(ravov). Much of this kind must be lost to us for ever. Cf. vol. i on the MSS.
Herodotus compares (8. 98. 2) the Persian ἀγγαρήιον with the Greek
lampadephoria in honour of Hephaistos.? I should like to infer from Hero-
dotus' description that the most important element in the picture which the

11 am not able to decide what is the authority of Fr. trag. adesp. 429 N. ἀπτέρωι τάχει
(in the editio princeps, but in none of the MSS, of Pollux 9. 152) nor do I know what con-
clusions can be drawn for the classical period from the Byzantine instances collected by
Nauck.
2 The passage in Herodotus has fallen a victim to the modern passion for discovering
interpolations everywhere.
153
line 282 COMMENTARY
Greeks of the period of the Persian war formed of the dyyapıjıov was the
handing on of the message in relays: this would make the expression peculiarly
appropriate here.
283. “ἕρμαιον: the accent in M and V is in agreement with Herodian's state-
ment (i. 369. 12). I have followed the practice of the editors of Homer
(7 471) and not the differentiation in LS s.v. ‘Eppatos, although it is arguable
that the word, combined with λόφος or λέπας, becomes an adjective.
The scholion, κοινὸν τὸ ἔπεμπεν, which has been erroneously inserted on 291,
belongs presumably to ἤδη μὲν «7A., not, as Dindorf and Wecklein assume,
to 289.
284. πανός is shown by Athen. 15. 700 d, e to be a particular kind of torch
(cf. also ib. 699 d).
284 f. rpirov . . . Ζηνός : Beazley's suggestion that this contains a pointed
allusion to the τρίτος (σωτὴρ) Ζεύς is very well worth consideration. The
great role that he plays in Aeschylus is well known (cf. on 1387). He is
named without the addition of σωτήρ in Cho. 244 f. σὺν τῶι τρίτωι πάντων
μεγίστωι Ζηνί. The triad Hephaistos, Hermes, and Zeus would stand sig-
nificantly at the beginning and the allusion to the Soter would give the
fire-message from the very start an implication of good omen to take on
its way.
285. The commentators compare Soph. (Thamyras) fr. 216 N. (= 237 P.)
Θρῆισσαν σκοπιὰν Ζηνὸς Abou. Wilamowitz, Glaube der Hell. i. 225, notes
that for sailors and islanders of the northern Archipelago Athos, as a weather
mountain! and therefore a seat of Zeus, had the same significance as
Arachnaion (below 309) for the people in the neighbourhood of Argos. Cf.
Wilamowitz, Interpretationen, 168: ‘Er (Aischylos) hat Ida, Athos, Makistos
vom Schiffe aus als Landmarken gesehen und verbindet sie nun durch seinen
heroisch-optischen Telegraphen.’ A journey in the reverse direction, from
Mount Athos to Lemnos and from Lemnos to Mount Ida, is made by Hera
(& 228 f. and 28: ff.).
almos: first instances here and 309, then in Sophocles and Euripides; in
prose only in the Corpus Hippocraticum ; therefore Aly, De Aeschyli copia
verb. 55 f., assumes that Aeschylus took the word over from Ionic. It may
have occurred in post-Homeric Epic and from there entered both the voca-
bulary of Tragedy and that of the Hellenistic poets (cf. Meineke, Anal.
Alexandrina, 289).
286. The variant (in FTr) ὑὑπεὶρ ἕλης i is strange ; it is expressly attested in the
Σχόλια παλαιά ın
i Tr: σημείωσαι ὡς ἡ "EAAq τὸ κύριον" διπλασιάζουσα τὸ λ.

ἀπέβαλε νῦν αὐτὸ διὰ μέτρου ἀνάγκην" τὸ δὲ ὑπὲρ προσέλαβε τὸ i διὰ τὴν αὐτὴν
aTriay.

. ὑπερτελής : probably rightly explained by Ahrens, 488, ‘what goes beyond

the τέλειον [or réAos]' ; he compares for the idea 301 φρουρὰ πλέον καίουσα τῶν
εἰρημένων. The same meaning is apparently found in S. Trach. 36: ἄθλων

! Cf. also Wilamowitz, Berl. Sitzgsb. 1912, 535 (on the Arsinoe ode of Callimachus, 1. 47):
‘Athos is not particularly high, 1000 m. lower than Olympus, but is particularly impressive
to seafarers because of its position and the shadow which near sunset it casts as far as
Lemnos’; see also his note, which mentions Soph. fr. 708 N. = 776 P. "Abws σκιάζει νῶτα
Anpvias βοός, a line connected with the Agamemnon passage by Is. Vossius (quoted in
Stanley’s commentary).

154
 COMMENTARY line 287

τῶνδ᾽ ὑπερτελὴς ἔφυ (Schol. ὑπὲρ τὸ τέλος γέγονεν οἷον τοῦ τέλους τῶν ἄθλων
κρείττων ὥφθη κατορθώσας αὐτούς). A retrograde formation from ὑπερτελεῖν
(only in Ag. 359) or the commoner ὑπερτέλλειν is also possible ; this assump-
tion suits E. Ion. 1549 f. τίς οἴκων θυοδόκων ὑπερτελὴς ἀντήλιον πρόσωπον
ἐκφαίνει θεῶν; which can hardly be separated from such expressions as Hdt. 3.
104. 2 6 ἥλιος... ὑπερτείλας μέχρι οὗ ἀγορῆς διαλύσιος, Eur. fr. 772 N. θερμὴ
δ᾽ ἄνακτος (sc. ἡλίου) φλὸξ ὑπερτέλλουσα γῆς καίει τὰ πόρρω κτλ., etc. It is
likely that the gloss in Hesychius ὑπερτελής : ὑπὲρ τὸ τέλος ἀφικομένη goes back
to a scholion on Ag. 286 as Hermann suggests. For the continuation of the
narrative with re cf. e.g. Pers. 417, where it has been challenged, and in general
Wilamowitz, Pindaros 469.
vorioas: the meaning of the verb causes difficulties which Porson's note
on E. Phoen. 663 (654) does not solve. The context seems to demand what
Blomfield gives as explanation: ‘der dorsum eo. Lux dicitur ire én’ εὐρέα
νῶτα θαλάσσης. Recte igitur schol. [marginal gloss in M] ὑπερβῆναι.᾽ So also
Kennedy and Headlam (who, however, considers also a quite different
explanation): ‘to skim the broad back of the ocean’. This interpretation
finds no support in the use of vwrifew and its composita elsewhere, but that
does not suffice to contradict Blomfield’s explanation, as all other instances
are later, and, even if the word was in more general use as early as the time
of the Oresteia, Aeschylus was the last man to shrink from giving it a new
meaning. It is not easy to find a true analogy for this use of a denominative
verb in -iZew, but the limits for this (-iZew) ‘Allerweltsdenominativsuffix’
(Debrunner, Griech. Wortbildungslehre, 129) must not be drawn too narrowly."
R. A. Neil, on Ar. Knights 273, rightly speaks of the ‘usual elasticity of verbs
in -{{w’. Ahrens, after carefully examining the use of νωτίζειν and νωτίζεσθαι,
rejected Blomfield’s interpretation. His own conjecture (pp. 493 ff.) is
grotesque: he reads ὥστε νωτίσαι ἰχθῦς, which he explains ‘sodass das Meer
die Fische auf den Rücken nahm, d. h. die Fische auf den Rücken des Meeres
kamen’. Apart from other reasons, ἰσχὺς λαμπάδος is corroborated by 296
σθένουσα λαμπάς (so Weil rightly). Therefore I see no need to discuss the
ideas of imaginative scholars who elaborate the activities of the fishes and
either arrange them in ‘choirs’ (A. Y. Campbell) and make them dance?
(G. Thomson) or picture them engaged in less sophisticated occupations
(Murray, cf. also Cornford, C.R. liii, 1939, 162).
287. πορευτοῦ λαμπάδος. Evidence for verbal adjectives in -rós having only
two endings in Tragedy is to be found in Lobeck on S. A7. 224; add also Ag.
593; for a discussion of the phenomenon see Wackernagel, Syntax, ii. 49.
λαμπάς prepares here and 296 for the special point in the conclusion 312.
Cf. on 8.
Paley recognized that before 288 there is a lacuna, in which the main verb
must have disappeared ; Wilamowitz arrived at the same conclusion. This
is much less daring than the attempts of Hermann and others to justify the

™ Fr. 436 N. proves that Aeschylus’ use of verbs in -é£e can be extremely unexpected :
the lexicographers (who unfortunately only gloss the single word without quoting the
whole line) show that he either used οὐρανιζέτω for πρὸς τὸν οὐρανὸν διικνείσθω or ὀὐρανίζετο
for πρὸς τὸν οὐρανὸν διικνεῖτο, This seems to imply a force in -/£ew sufficiently near to that
assumed by Blomfield.
2 On ‘dancing fish’ cf. Lillian B. Lawler, Class. Phil. xxxvi, 1941, 153.

155
line 287 COMMENTARY
text as it stands. Hermann connects πρὸς ἡδονὴν πεύκης (as he reads, follow-
ing Schütz), which contradicts the fixed use of πρὸς ἡδονήν, and maintains
that it means ‘ut pro lubitu luxuriaretur flamma'. Blomfield, Verrall, and
Blaydes regard πεύκη as in apposition to ἐσχὺς λαμπάδος, which is scarcely
tolerable and does not supply the missing main verb. Therefore the lacuna
is certain on grounds of language. The suggestion of W. M. Calder (CR.
XXXVI, 1922, 155 ff.) that the lacuna may also account for a difficulty arising
out of the subject-matter is worth considering. It had been pointed out
(Verrall, 2nd ed., Introd., p. xxif.; W. Riepl, Das Nachrichtenwesen des
Altertums, 1913, 51) that in contrast to the moderate distances from the
Makistos post onwards the first three intervals (Ida-Lemnos, Lemnos-Athos,
Athos-Euboea) are abnormally wide and the third is about 180 km. Diels’s
general reflection (Antike Technik, 3rd ed., 8o n. x), ‘however, the imagination
of a poet bridges even wider distances’, and in particular his reminder that in
the Arsinoe of Callimachus (fr. 1. 47 ff. Pfeiffer) Charis watches from Athos
the smoke of a pyre burning in Alexandria, gives little help. The Hellenistic
poem belongs to the fairy world of playful mythology, whereas there is
sufficient evidence to show that the transmission of a fire signal over the
islands was for Aeschylus and his audience something perfectly real of which
they would think in geographical terms (cf. e.g. Diels, loc. cit., in the text,
Wilamowitz, Griech. Trag. ii. 307 n. 1). The chief landmarks in the northern
Aegean were so vital in the decades before 458 for normal Athenian merchant-
men as well as for the conduct of the war that there could have been no
doubt about their approximate distances. The hazy and confused representa-
tion of the farthest north which meets us in the Prometheus trilogy is some-
thing completely different. Clytemnestra’s narration presupposes the
knowledge natural to seafaring Athenians, as Verrall has emphasized. For
such reasons Calder proposed to insert the name of an intervening post
between Athos and Euboea in the lacuna which, on different grounds, must
be assumed before 288; he chooses the island Ikos. Calder maintains his
position (C.R. lii, 1938, 171) against the objections of principle raised by
Horace L. Jones.’ It seems to me indeed possible that in the lacuna before
288 the mention of an intermediate post is lost with the rest; but I do not
regard this assumption as necessary because even with it the first two stages,
though somewhat shorter, remain uncomfortably long, if one insists on
applying to the queen's report the criteria of everyday experience. Most of
all, we do not know whether the poet has not contented himself with a more
summary treatment of the areas farther removed from Argos (and Athens)
to prevent his narration from becoming excessively long; moreover, the
insertion of an element of the θαυμάσιον here would hardly have decreased
the πιθανότης.
πρὸς ἡδονήν was meant to go with the verb which is now lost ; cf. p. 199.
288. πεύκη seems to be a verbal variation for λαμπάς (several times) and
πανός 284.

τ In Classical Studies presented to Edward Capps (Princeton 1936), pp. 182 ff. I agree with
Calder that the experiments with the modern heliotrope bear no relation to ancient experi-
ence. Jones’s essay, which seems to have been written for the Greekless reader, does not
touch on the other reasons which have led to the assumption of a lacuna before 288, quite
independently of the problem of the distance.
I56
 COMMENTARY lines 292 f.

288 f. τὸ χρυσοφεγγὲς ... σέλας : ‘articuli vis possessiva’ Wilamowitz, probably

rightly. Cf. e.g. Prom. 28 τοιαῦτ᾽ ἐπηύρου τοῦ φιλανθρώπου τρόπον, Menander
fr. 652 Kock (quoted in K. W. Krüger, Griech. Sprachlehre, $ so. 2 n. 3) τότε
τὰς γυναῖκας δεδιέναι μάλιστα δεῖ, ὅταν (rt) περιπλάττωσι τοῖς χρηστοῖς λόγοις
‘with their well-known.
ὥς τις ἥλιος. In comparisons τις often seems redundant. This usage is
fully illustrated by J. Vahlen, Opusc. acad. ii. 181 ff. Cf. 1142, 1194, and Fragm.
trag. adesp. 541. 3 N. (of the many-coloured wings of the Sphinx) κυανωπὸν
ws τις Ἶρις ἀντηύγει σέλας.
289. With the use of παραγγέλλειν A. S. F. Gow, Journ. Hell. Stud. xlviii,
1928, 141, compares Pers. 380 παρεκάλει; it indicates ‘the passage of the word
or cry from one squadron to another”.
Μακίστου oxomais: as far as I know, the scholion to this passage, Makıorov
ὄρος Εὐβοίας, is our only source for the localization. The scholion may depend
on other evidence but may be a simple deduction, since any reader moderately
familiar with geography must have said to himself that 292 ἐπ᾽ Εὐρίπου pods
presupposed the naming of a post in Euboea. As the next stage is Messapion,
which is known to be situated south of Anthedon, it seems natural to accept
as the preceding transmitting station the mass of Kandhili, with which
Makistos is now generally identified (cf. e.g. Kiepert’s maps; Baedeker’s
Griechenland, sth ed., 1908, p. 224; Philippson, RE vi. 852 f.). The conclusion
of Fr. Thiersch, Abh. Münch. Akad. Philol.-philos. Cl. v. 2 (1849), 39, that
ἑκάς 292 necessarily fixes the mountain in the farthest north of the island,
does not compel assent. Nor is the normal identification shaken by Ahrens’s
view (p. 496) that ‘the form Maxıoros instead of the Attic-Ionic Μήκιστος
would.be very surprising for an Eubocan mountain’. Who can say that the
peak, which would be seen from far away, was not given its name by the
inhabitants of the mainland across the straits? Perhaps, however, Aeschylus
merely assimilated the vowel to that of the common superlative and allied
forms, as in Tragedy (cf. L-S s.v. μήκιστος) only μάκιστος occurs in trimeters
(Aesch. fr. 281. 1 N.) as in lyric metres (E. Hipp. 818, S. Phil. 849) and ana-
paests (5. Oed. R. 1301). Add A. Pers. 698 (troch. tetrameter) μακιστῆρα
μῦθον. The other possibility which must also be mentioned is that μάκιστον
is not a traditional proper name but a descriptive adjective as probably 303
αἰγίπλαγκτον and possibly 302 γοργῶπιν (v. ad loc.).
290 f. Stanley was right: ‘hic autem non omnino aut cunctans aut incurie
somno victus nuntii partes praetermisit, Hermann and Paley wrongly take
the negative not with the whole sentence (cf. on 1312) but only with the
participles and understand παρῆκεν as the equivalent of παρήγγειλεν, which is
not supported by usage. The mountain is the subject of the sentence and is
masculine for Aeschylus. ' Makıoros mons masc. gen. ut ‘Yunrros, Ilap-
νασσός. Isque pro custodia ipsi imposita infertur in seqq. prorsus παθητικῶς:
ὁ δ᾽ οὔτι μέλλων. Ita G. Hermannus in scholis explicuit', Thiersch, loc. cit. (on
289). In his edition Hermann abandoned this correct view and read σκοπῶι.
292 f. With regard to the construction the opinions of editors differ. Some
(e.g. H. Voss, Hartung, Keck, Kennedy, Paley, Verrall, Mazon) take φρυκτοῦ
φῶς as subject, others (e.g. Conington, Nágelsbach, Wilamowitz) regard φῶς
as accusative, taking φῶς μολόν as the content of σημαίνειν and keeping as
subject à 3é from what has gone before. The order of the words seems to
157
lines 292 f. COMMENTARY
favour the first interpretation; there is no particular awkwardness in the
taking up of 6 δέ by φρυκτοῦ φῶς which is almost identical in meaning.
With this interpretation there remain two possible ways of understanding
σημαίνει μολόν (Headlam is correct). σημαίνει can either be taken absolutely
(‘give signs’), which is quite common both in Homer andin Tragedy (examples
in L-S s.v. I. 2); cf. S. Ant. 1208 καὶ δεσπότηι Kpeovrı σημαίνει μολών. Or
μολόν is to be understood as the content of σημαίνειν, which is perhaps prefer-
able as being a more forceful expression. (L-S do not commit themselves
and cite Ag. 293 for the participial construction, p. 1593 1. 3, as well as for the
absolute use.) I have no certain instance of a participle in the nominative
(i.e. relating to the subject) depending on σημαίνειν (Ag. 969 is doubtful, ct.
ad loc.), but this causes no difficulty. Goodwin rightly says, § 904, of the whole
group with which we are concerned: ‘The participle may belong to either the
subject or the object of these verbs, and agree with it in case.’ Cf. also
Kühner-Gerth ii. 52 f., in particular S. A7. 471 f. δηλώσω... y . . . ἄσπλαγ-
xvos ... γεγώς, Ant. 20 δηλοῖς γάρ τι καλχαίνουσ᾽ ἔπος, Thuc. 3. 64. 1 δῆλόν τε
ἐποιήσατε... ov μηδίσαντες etc. |
ἐπ᾽ Εὐρίπου pods: ‘over’, as ἐπ᾽ εὐρέα νῶτα θαλάσσης and in similar expres-
sions,
293. Aeschylus mentioned the Μεσσάπιον ὄρος in Γλαῦκος Πόντιος also, Pap.
Oxy. 2159. ro.
σημαίνει, present, as towards the end of the narrative (from 305), whereas
the majority of the verbs in this speech are in the past. On the presumable
force of this present see on 1383.
295. The first hand of M alone has ἐρίκης. If the deme Erikeia of the Phyle
Aigeis is really named after heather, the epigraphic evidence collected by
Meisterhans, 42 n. 261, proves that at least from the middle of the fourth
century the good Attic form was epıx- (see also op. cit. 53 n. 437 b). On the
other hand, L-S supports the reading épew- by an inscription (Schwyzer,
Exempla, 720) from Thebes at the foot of Mount Mykale, i.e. from the
neighbourhood of Priene, first half of fourth century, which reads L. 18 f. ἐν
τῶι ᾿Ερεικὄᾶντι λόφωι.
γραίας ἐρίκης : Sophocles fr. 784 N. (= 868 P.) has γραίας ἀκάνθης in the
same case and in the same place in the line. This may be one of the instances
in which he shows himself influenced by the bold phrase-coining of the
Agamemnon, cf. on 526.
296 f. This is not a case of piling up participles, because each of the three
participles has a quite separate syntactical function : ὑπερθοροῦσα marks the
beginning of the main action, od μαυρουμένη the state of the subject at the
time, while σθένουσα, which is more strongly attributive and has less verbal
sense, gives the quality which is maintained by οὐ μαυροῦσθαι.
οὐδέπω : Schneidewin and Wecklein disregard the temporal meaning when
they explain ‘durchaus nicht etwa’, ‘garnicht’, in conformity with the usual
interpretation of e.g. Soph. Oed. R. τος. Ancient interpretation of Homer
sometimes followed this path, e.g. Scholion A to & 143 παρέλκει ἐνθάδε τὸ πω.
Kaibel on S. El. 403 rightly protests: "λαμπὰς οὐδέπω μαυρουμένη is the torch
which has never gone out’; Conington: ‘not yet grown dim’.
297. If we want to find a reason for the strange mutilation in the common
source of MV, we may perhaps surmise that, after e had been miswritten as
158
 COMMENTARY line 301

at’ (cf. on 1653), a syllable-counting Byzantine used surgery on the verse

he could not scan. When Verrall thinks it necessary to put reôlov . . . wroü
in the text (although he himself admits: ‘on the other hand the later MSS
may represent a tradition’) and Murray mentions in the apparatus Turnebus’
conjecture "poro, they pay tribute to the belief in the unfailing superiority
of the Mediceus. Pers. 805 πεδίον Ἀσωπὸς is found at the same place of the
trimeter.
299. πομποῦ πυρός strikes with peculiar effect the keynote of the whole narra-
tion: 283 ἔπεμπεν, 300 τηλέπομπον, 305 πέμπουσι.
300. ἠναίνετο. For the ambiguity of this verb cf. on 1653. Commentators
differ in their interpretation here, as at Z 500. It is taken to mean ‘to decline
to do a thing’ by e.g. Schütz (‘non recusavit quominus et ipse flammam late
splendentem accenderet'), Nàgelsbach, Wecklein, Mazon; but the sense of
‘disown’, ‘refuse to accept’, is given by Stanley (lucem . . . longe-missam non
respuebat’), Conington (‘spurned not the fiery gift far sent’), Paley (‘nor did
the guard . . . disown the light that reached him from afar’), Verrall, Headlam.
The second interpretation with the nuance 'spurn, reject' seems to me the
more natural and effective rendering; the meaning would be in accord with
Homer's σὲ δ᾽ ἀναίνεται ἠδὲ σὰ δῶρα (I 679), and οὐ yàp καλὸν ἀνήνασθαι δόσιν
ἐστίν (σ 287). But perhaps Aeschylus intended to keep the sense suspended
between the two meanings of the verb, with some of the implications of each.
He could appeal to Homeric usage also for ἀναίνομαι with the accusative of
refusing (to do) an action, but in such a case Homer would probably use not
a noun such as φάος but one which expresses an action, as in y 265 dvaivero
ἔργον deines, or ὃ 651 χαλεπόν κεν ἀνήνασθαι δόσιν ein (where δόσιν picks up the
δῶκα Of 649).
301. τῶν εἰρημένων: Stanley’s rendering ‘supra dictis’ has long prevailed:
even Headlam gives as his second alternative ‘beyond the aforesaid’. This
gives a needlessly frigid turn to the sentence, which θεσμός 304 (cf. note ad
loc.) makes downright improbable. Weil’s ‘custodes dicuntur flammam
accendisse etiam maiorem quam iussi erant’ gives the correct sense. εἰρημένον
(or -uéva) means ‘command’, ‘order’, not only in Sophocles (Ant. 215) and
Thucydides (cf. also Hdt. 7. 26. 1 εἴρητο) but in Aeschylus, too (Ag. 1620).
Dindorf substitutes for the words πλέον... τῶν εἰρημένων the anonymous
fragment cited by Hesychius s.v. προσαιθρίζουσα (-Bepitovoa cod.) πόμπιμον
φλόγα. This has been widely approved (cf. Nauck on Trag. frag. adesp. 260) ;
Weil and Ahrens (p. 487) also attributed the quotation to this speech, though
at a different point. Hense is sceptical; Wilamowitz does not mention the
conjecture; both Headlam and Murray expressly reject it. The attribution
is in fact quite arbitrary.

! The same misspelling occurs in the same word e.g. in Helladios' fourth century A.D.
copy (preserved on stone) of the (expanded) epigram on the Megarians killed in the Persian
War (Hiller v. Gaertringen, Histor. Griech. Epigramme, no. 30; M. N. Tod, Greek Hist. Inser.
no. 20) rot δὲ καὶ ἐν παιδίω Βοιωτίω.
2. Aeschylus’ audience certainly knew where Oropus lay and that a line from Messapion
to Kithairon would not come anywhere near Oropus. What Verrall says in favour of this
conjecture is very poor support.
3 Wilamowitz on Prom. 1031 takes the MS reading εἰρημένος there in the same way and
compares Ag. 301. In spite of Pohlenz, Gnomon ix, 1933, 625, I have reached no certain
conclusion on Prom, 1031.
159
 lines 302 f. COMMENTARY
302 f. Poets are not always kind to their interpreters, or to topographers
looking for straightforward identifications. Aeschylus here plays hide and
seek with them. In this description, which undoubtedly as Wilamowitz says
' (Griech. Tragoedien ii. 307 n. 1) ‘reveals exact knowledge of the localities’, he
twice substitutes a poetic circumlocution for a proper name. αἰγίπλαγκτον,
‘ranged over by goats’, can hardly have suggested to Athenian ears the name
of a mountain, and it is quite possible that Aeschylus coined the word (cf.
On 12 νυκτίπλαγκτον) himself.’ It has been generally assumed, following
Ernst Curtius, Peloponnes, ii. 552, that Geraneia is meant, and this is un-
doubtedly right. It is true that the ὄρος Μεγαρίδος of the scholiast, which is
consistent with this assumption, could easily have been got from the context
(cf. on 289 Maxiorov). But a glance at the map shows that a line drawn from
Kithairon over the westernmost tip (cf. on 306 1.) of the Saronic Gulf to
Arachnaion must pass through Geraneia, or at least its outlying spurs. It
would in any case be absurd to go out of one’s way to find some insignificant
hill-top between Kithairon and the straits east of the Isthmus, with the
mighty mass of Geraneia there at hand, nepin Γεράνεια, familiar to every
Athenian from so many of its aspects. And throughout this speech it is
clearly the high mountain tops dominating their surroundings that are
selected for the beacons; in such a connexion any idea of a minor elevation
in the foothills is obviously inappropriate.
Much more awkward is the problem of the yopydms λίμνη. Here, too, it is
highly probable that the ‘gorgon-eyed lake’ is to be taken not as a proper
name but as a description. It remains uncertain whether this adjective had
been earlier applied to the lake or whether (as I am inclined to believe)
Aeschylus was the first to do so. The sketchy note of Hesychius (s.v.
γοργῶπις) on the occurrence of this word in Cratinus' Pylaza (ii, p. 115 Mein. =
fr. 178 K.) tells us nothing further about the context there and gives no clue
as to the relation to Aeschylus. The story in Hesychius, loc. cit., and in
Etym. M.s.v. ἐσχατιῶτις, that Gorgo daughter of Megareus and wife of Korin-
thos threw herself into the lake, which thereupon came to be called Γοργῶπις
instead of ᾿Εσχατιῶτις, does not prove that the name l'opyôms was ever
actually given to the lake in living speech ; the shadowy Gorgo (for whom see
Malten, RE vii. 1597, ignored by Eitrem, RE xi. 1399) and the trivial story of
her suicide may quite easily have been an etymological invention to account
for a single reference in some poet, possibly this very passage. More worthy
of serious attention is the identification (cf. Etym. M. loc. cit.) of the Γοργῶπις
with the ᾿Εσχατιῶτις λίμνη in the region of the Isthmus. The identification of
the λέμνη in the ἐσχατιά with the 'Vouliagmeni' (not to be confused with that
near Cape Zoster in Attica) on the southern side of the western finger of the
peninsula of Perachora, which juts out into the Corinthian Gulf, has been
generally accepted since it was put forward by Curtius (cf. Bólte, RE vii.
1659; Payne-Dunbabin, Perachora, i. 9). This is clearly right, since the

! I should like to take Pers. 309 ἀμφὶ νῆσον τὴν πελειοθρέμμονα in the same way, i.e. asa
poetic description closely resembling αἰγίπλαγκτον and similarly avoiding the proper name.
(The identification is much disputed, cf. A. S. F. Gow, Journ. Hell. Stud. xlviii, 1928, 141.)
I cannot therefore agree with Wilamowitz's phrasing (Griech. Lesebuch, ii. 34) : ‘The Athe-
nians knew which of the islets (there were several) was called after the wild doves’ ; instead
of *was called' I should say *was called by Aeschylus'.
160
 COMMENTARY line 304
position of this remarkable lake makes the name ’Eoyarıörıs peculiarly
appropriate. It does not follow from this, however, that the γοργῶπις λίμνη
of Aeschylus is the Vouliagmeni, since the identification of the Topyanıs with
the ᾿Εσχατιῶτις may be the invention of some commentator. The question
can be decided only on internal grounds. Undoubtedly the appearance of
the Vouliagmeni amply justifies the name γοργῶπις : the traveller descending
from the modern village of Perachora and catching his first glimpse of the
lake sees below him its round and shining blue eye somewhat deep-set
beneath its rocky brows.’ But the context of the passage is against this
identification. 302 serves precisely the same function as 286 πόντον wore
νωτίσαι, 292 Em’ Εὐρίπου ῥοάς, 297 ὑπερθοροῦσα πεδίον Ἀσωποῦ and 306 f.
Σαρωνικοῦ πορθμοῦ κάτοπτον πρῶν᾽ ὑπερβάλλει. That is to say, between one
mountain peak and the next the blaze of the beacon crosses (note the recur-
ring ὑπερ-) an intervening stretch of water: the sea, a strait, a river-valley.
We must not therefore follow Wilamowitz’s translation and conceal the clear
arrangement: ‘Zur Alm des Gaisbergs überm gorgoaüg’gen See schwang sich
der Schein’, as if the γοργῶπις λίμνη were named simply in order to indicate
a sheet of water near Geraneia which reflected the brightness of the fire as it
reached the summit of the mountain. It is in fact quite obvious from the
consistent arrangement of the speech that the γοργῶπις λίμνη should lie in a
hollow between Kithairon and the αἰγίπλαγκτον ὄρος, and this makes the
identification with the Vouliagmeni impossible. The point has been emphasized
again recently against the communis opinio by Humfry Payne (Perachora,
i. 9 n. 4). I cannot, however, subscribe to his further conclusions: it seems to
me idle to try to determine which particular lake—in any case of quite
insignificant size—in the neighbourhood, possibly, of Pagai, Aeschylus had in
mind. If indeed he did mention so trivial a piece of water in the same breath
with the great outstanding landmarks, then we can only put it down to his
caprice. The only other possibility, which Payne does in fact recognize, is
that by γοργῶπις λίμνη is meant the extreme eastern tip of the Halcyon
Gulf But the other places where λίμνη is used of the sea (e.g. A. Suppl. 529,
but in the language of lyric) do not favour this assumption ; and still less does
the epithet γοργῶπις. It is impossible to say why the poet uses fancy names
to designate lake and mountain ; perhaps it is just for the sake of variety.
302 ff. ἔσκηψεν on the one hand and ἐξικνούμενον wrpvve on the other cannot
refer to two stages, an earlier and a later in point of time, but simply indicate
in different words the descent and the arrival of the signal on the αἰγίπλαγκτον
ὄρος. Ahrens’s assertion (p. 503) that "σκήπτω does not in itself imply a
movement downward from above, but a thrusting and therefore rapid move-
ment straight forward’ is not borne out by the evidence.
304. ὥτρυνε (as often from Homer downwards) has a dependent infinitive, the
object of which is θεσμὸν πυρός. The object of ὥτρυνε did not need mention-
ing, since it can be understood without ambiguity from the preceding lines:
it is the watch on the αἰγέπλαγκτον ὄρος, or—it comes to the same thing—the

1 ] was fortunate enough to visit Perachora in 1937. The good map in RE xv. 166 gives
a convenient picture of the whole area concerned. .
2 So Otfried Müller, Die Dorier, ii, and ed., 421. His assertion that λίμνη in Plutarch,
Quaest. Graec. 59, p. 304 f ἐν Αἰγείροις παρὰ τὴν λίμνην means the gulf of the Halcyon sea
rests on a mistake, cf., e.g., Bólte, RE vii. 1659, Ernst Meyer, RE xv. 169.
4872.2 M 161
line 304 COMMENTARY
mountain itself (cf. 290 f.). The expression θεσμὸν... πυρός is irreproachable,
continuing as it does the idea of τῶν εἰρημένων (301). It would be incompre-
hensible that the obviously genuine θεσμόν should have been assailed again
and again, did we not know that many critics, when at a loss over a notorious
corruption, look round in the neighbourhood, seize upon a word which is
perfectly good in itself, find fault with it, and so patch up the whole in sorry
fashion. Here there can be no mistaking the seat of the corruption: μὴ
χαρίζεσθαι does not make sense. Early critics of Aeschylus were quite clear
about this. Older conjectures, from Triclinius’ δὴ χαρίζεσθαι downwards, are
sufficiently refuted by Ahrens, 497. Heath’s xarileoda: still finds supporters ;
it is supposed to mean ‘hortabatur ne constituta ignium lex desideraretur’
(Enger in Klausen, 2nd ed.). Th. Bergk's objection that there is no evidence
for a passive of yarifew has not been overthrown by Hermann. Even more
serious is the syntactical difficulty involved in a passive infinitive construc-
tion depending on ὥτρυνε. Weil is justified in his argument (1858) against
both Heath’s conjecture and Casaubon's μὴ xpovileoda:,' adopted by many
editors: ‘quae mihi videntur abhorrere a Graecorum usu loquendi, qui non
solent verba hortandi et iubendi, ut Latini ?wbere, cum infinitivo passivo
iungere'. Several editors seem as a matter of fact to take χρονίζεσθαι as a
middle, e.g. Paley (who regarded χρονίζεσθαι as ἃ pis aller, see below): ‘and
urged on the succession [?] of the fire not to linger in its course’; Headlam:
‘urged the fiery ordinance to make no tarrying'. But such a middle can hardly
stand in view of the general use of χρονίζειν, and in particular of Sept. 54 καὶ
τῶνδε πίστις οὐκ ὄκνωι χρονίζεται. None of the numerous conjectures really
commends itself, and I find myself in agreement with Paley: 'the reading is
so uncertain that it has been marked with an obelus.' What one expects is
an infinitive in the sense of 'obey', 'respect', or, if μή is the poet's own (and
this is by no means certain), 'not to neglect' or something similar.
306. φλογὸς... πώγωνα: cf. Eur. fr. 836 (Phrixus) πώγωνα πυρός, interpreted
by Poll. 2. 88 ἡ εἰς ὀξὺ ἀναδρομὴ τῆς φλογός. A picturesque and effective
phrase, if one thinks of the shape of beards at the time when Aeschylus was
a young man.
306-8. The difficulties in these lines have all been noted, but in part glossed
over, in part not seen in their proper connexion. The infinitive ὑπερβάλλειν
might just be defensible, not, indeed, as being in epexegetic or consecutive
relation to φλέγουσαν (Wellauer, Schneidewin among others) or to μέγαν
πώγωνα (so, e.g., Paley, Plüss) : ‘the flame was so bright that it could be seen
beyond the point farthest removed in that direction from Aegiplanctus'
(Paley), but as, e.g., Wecklein and Verrall take it, as a final-consecutive to
complete the sense of πέμπουσι or the whole of the preceding clause. Cf. in
general Goodwin, $8 770-5, and in particular Wackernagel's remarks (Syntax,
i. 262) on verbs of sending with the infinitive, and K. W. Krüger, Griech.
Sprachlehre, i, $ 55. 3, n. 20, on the example of such an infinitive at Eur. Jon
1559 ἡμᾶς δὲ πέμπει τοὺς λόγους ὑμῖν φράσαι. But there are obvious differences:
(1) φράσαι in the line of Euripides is much more the aim of πέμπειν than
ὑπερβάλλειν can be in the Agamemnon passage, and (2) the gap between the
infinitive and its governing verb here would be awkwardly wide. Still, one
1 The alteration is not improbable in itself. In Sept. 54, which is compared by the com-
mentators, all the MSS, save one, give χρονίζεται, but Stobaeus quotes χαρίζεται,
162
 COMMENTARY lines 306-8

might perhaps put up with this. But in the text as it stands the καί before
Σαρωνικοῦ is unintelligible. Commentators, if they notice it at all, get out of
the difficulty by Hermann’s adeo or the like, which is lame, to say the least.
When previously in the north such leagues of distance have lain between
one post and the next, there is really no need to mark by an ‘even’ the cross-
ing of the 5 miles or so of water between the north and south sides of the
Saronic Gulf at its westernmost tip. The next trouble is with the gender of
$Aeyovoav. From Wellauer onward almost all commentators have explained
this as making the participle agree with the φλόγα implicit in φλογὸς πώγωνα,
but I have been no more successful than Ahrens (p. 498) in finding a parallel
for this bold construction. Finally, εἶτ᾽ ἔσκηψεν, εἶτ᾽ ἀφίκετο. This figura,
scarcely tolerable here in itself, becomes, as Hermann rightly felt, completely
absurd when one reflects that σκήπτειν means ‘travel downwards’, ‘hit the
ground’, and that therefore no time can elapse between the moment of
ἔσκηψεν and that of ἀφίκετο, so that the anaphoric elra-phrases would not
describe a progressive action! but only a single event happening at one and
the same moment: ‘then it travelled downwards, then it arrived’. Verrall’s
rendering ‘alighting then and only then’, and his comment, are well calcu-
lated to bring out the impossibility of the text he defends. For an added
doubt on the subject of εἶτα see p. 164.
Schütz removed the first three of these four awkward points by writing
καὶ Σαρωνικοῦ πορθμοῦ κάτοπτον πρῶν᾽ ὑπερβάλλει πρόσω φλέγουσα, thus getting
tid of the dubious infinitive and giving a proper function to the hitherto
meaningless καί. (The other way of giving sense to καί, namely by assuming
a lacuna before it, is open only in theory, since in fact we cannot imagine
that any post was mentioned between Geraneia and the western end of the
Saronic Gulf.) For the third point, the feminine φλέγουσα, this picking up of
φλογὸς (πώγωνα) in a new clause with a new verb (cf. on 292 f. for the taking
up of the subject by φρυκτοῦ φῶς) is as unobjectionable as the φλέγουσαν of the
text is harsh. It is surprising that this self-evident emendation has received
so little notice. Schütz’s restoration of the following words is less happy. He
starts off correctly after φλέγουσα with καί (xdır’), an easy alteration which
improves the construction (cf. 310), but at the end of the line he follows
Stanley in writing ἔστ᾽ ἀφίκετο and translates ‘donec attingeret’, thus ignoring
the contemporaneity of ἔσκηψεν and ἀφίκετο noted above.” Nor is Hermann's
εὖτ᾽, which appears in most recent texts, quite satisfactory. In Tragedy εὖτε
with the aorist indicative, where the meaning is temporal and not causal,
normally means ‘from the time when’, e.g. Sept. 745, Ag. 986, S. El. 508,
not ‘as soon as’, ut primum, which is the sense required here. This would be
ὡς, cf., e.g., Pers. 361, S. Trach. 777, Phil. 271, 0ed. C. 1607. Perhaps Aeschylus
did write ὡς ἀφίκετο. There is besides these objections to εἶτα a further
point which should be noted, though it cannot be used as an argument.
1 Wilamowitz’s translation ‘und weiter strebt’ es, immer weiter’ is not warranted by his
text.
2 For a further objection to ἔστ᾽ ἀφέκετο cf. the survey by Gildersleeve, Am. Journ. of
Philol. xxiv, 1903, 398 (he is himself dependent on Fuchs, ‘Die Temporalsitze mit den
Konjunktionen “bis” . . .' in Schanz's Beiträge z. hist. Syntax der griech. Spr. No. 14,
Würzburg 1902). According to these scholars earlier Greek usage demands in the case of a
dependent clause with ἔστε an imperfect in the main sentence (instead of ἔσκηψεν). This is
precluded by the content here.
163
lines 306-8 COMMENTARY
While ἔπειτα is common in Aeschylus, εἶτα, apart from this passage,’ occurs
only once in the Prometheus (777), the language of which is peculiar in so
many ways, and there with adversative force (cf. infr. on 481) after a participle.
It can hardly be due to mere chance that so ordinary a word is completely
missing from all the remaining plays, particularly considering how common
εἶτα is in Sophocles. A more likely explanation is that those of the surviving
plays of Aeschylus which are nearest in point of time to the Sophoclean, that
is the Oresteia and the Prometheus, show in this point as in so many others
a ‘Sophoclean’ usage. An alternative suggestion, which I put forward with
some reserve, is that in the present passage the twice-repeated εἶτα has dis-
placed a different reading. Perhaps Aeschylus wrote φλέγουσα κἀπέσκηψεν
ὡς ἀφίκετο (for an intransitive ἐπισκήπτειν v. Eum. 482). The use of καί as
connecting particle (xai ἐπέσκηψεν) between 306 f. καὶ... πρῶν᾽ ὑπερβάλλει
and 310 κἄπειτα would be quite appropriate to the style of such a narrative,
cf., e.g., the polysyndeton Pers. 382-4 kai. . . καθίστασαν... καὶ νὺξ ἐχώρει, κοὐ
μάλ᾽ Ἑλλήνων στρατὸς... ἔκπλουν... καθίστατο. The possibility might be
considered that an original ΚΑΙΕΠΕΣΚΗΨΕΝ became first KAIEITEZKHYEN
and then NEITEZKHYEN. But this attempt to restore 308 after φλέγουσα
is much less reliable? than Schiitz’s alterations (φλέγουσα and the preceding
ὑπερβάλλει).
307. Canter’s κάτοπτον has been almost universally adopted, and the gloss in
M τὸ κατόψιον in conjunction with E. Hipp. 30 f. κατόψιον γῆς τῆσδε has long
been quoted in its support.” Headlam wanted to read κατόπτην as the more
strictly analogous form, but the technical significance of both karorrnp and
κατόπτης (for the relation of these two cf. Ernst Fraenkel, Nomina agentis,
ii. 3) by the time of Aeschylus makes this conjecture undesirable. For the
active as well as the passive function of verbal adjectives in -ros cf. on 12.
The πρών, ‘foreland’, mentioned here is no conspicuous mountain-peak (there
is no watchpost on it), but part of the lower foothills which, to the south of
the main ridge of Geraneia, slope down towards the sea between Krommyon
and the modern village of Kalamaki. It is presumably for the sake of varia-
tion that the headland overlooking the πορθμός rather than the πορθμός is
made the object of ὑπερβάλλει.
306 f. Σαρωνικοῦ πορθμοῦ. This was correctly rendered by Stanley freti (so,
too, Hartung ‘am Saronischen Sunde’); but from le Pére Brumoy onwards
most translators and commentators, perhaps led astray by the scholia, have
with remarkable unanimity taken it as the equivalent of κόλπου, thereby
obscuring the precision of the expression (Mazon, however, renders correctly
τ In Ag. 1089 κᾷτα is a corruption in FTr.
2 Despite these grave uncertainties no reader of the Agamemnon ought to be perturbed
by the ghost of a line which, following suggestions by Cobet and Meineke, was put in the
place of 308 by Wecklein (Rhein. Mus. xxviii, 1873, 626 f, and in the appendix to his com-
mentary [1888] p. 144).
3 κάτοπτρον is again championed by W. B. Stanford, Aeschylus in his Style, 136 n. 9,
and by J. A. K. Thomson, C.Q. xl, 1946, 56 ff.; the latter also keeps the MS reading ὑπερ-
βάλλειν and translates the passage as follows: ‘Kindling a great beard of flame . . . they
send the message that the peak (i.e. Aegiplanctus) cast it (i.e. the flame) even over the
mirror of the Saronic gulf, blazing on and on. . ..' The assumptions on which this rendering
rests have been explicitly or implicitly rejected in my notes. Moreover, it will be obvious
that the identification of the πρών in 307 with the αἰγίπλαγκτον ὄρος mentioned in 303 is
incompatible with the whole structure of this speech.
164
 COMMENTARY line 313
‘le détroit Saronique’). Since Aeschylus is thinking of the line Geraneia-
Arachnaion, he has to take into account only the most western part of
the Saronic Gulf. It forms almost a rectangle with its eastern side open; the
north side reaches from Krommyon (Hag. Theodori) to the entrance to the
Corinth Canal, the length from north to south being c. 7 to 8 km. πορθμός is
as apt a term as could be wished for this bight which is so like a strait. It is
by no means certain that Aeschylus took over πορθμὸς Zapuwırös as an
existing name for the western bight ; he may equally well have invented this
special term for κόλπος (or πόντος : E. Hipp. 1200) Zapwrırös to suit his con-
text here. The lengthy and learned discussion which Ahrens (499 ff.) devotes
to this passage in order to prove ‘that the Athenians could hardly have
understood by πρὼν κάτοπτος πορθμοῦ Σαρωνικοῦ any other point than
Methana’, is a wasted effort. For the origin of the name Σαρωνικὸς κόλπος
see Wilamowitz, Glaube der Hell. i. 387 ff.
309. Apaxvatov altos: Paus. 2. 25. xo ἔστι δὲ ὄρος ὑπὲρ τῆς “ήσσης (for its
place on the map cf. Boethius, RE xii. 2136) τὸ Apaxvaiov κτλ. There can be
no doubt that it has been rightly identified as the mighty ridge called to-day
Arna, which stretches from west to east on the north side of the highway
from Argos to Epidaurus. The main peak, Hag. Elias, is 1,199 m. high,
according to the map and Baedeker, Griechenland (1908), 323. As a landmark
it is visible for a great distance (for photograph see Hesperia, vii, 1938, 512).
311. οὐκ ἄπαππον: the interlinear gloss (in FTr) οὐ ξένον, ἀλλὰ συγγενές is
not bad.
312. τοιοίδε τοί μοι: Schütz was the first to divide the words correctly and
thus restore their sense; Hermann merely contradicted him without pro-
ducing an argument. Clytemnestra’s next speech ends (348) in an analogous
Way: τοιαῦτά τοι γυναικὸς ἐξ ἐμοῦ κλύεις, both phrases expressing her self-
assertive pride. In both there may be some nuance of the ‘boasting’ which
Denniston, Particles, 540, attributes to τοι. τοί μοι also in Prom. 1040.
313. ἄλλος wap’ ἄλλου xrA.: Ahrens’s discussion (303 f.) makes a good starting-
point here. He rightly rejects the explanation of those commentators who
refer ἄλλος παρ᾽ ἄλλου to νόμοι, and discusses the interpretation given by
Schütz and the translators who accept it, ‘leges alio ab alio vices excipiente
impletae’. Here Ahrens remarks: ‘The sense is natural and appropriate, but
can hardly be got from the words as they stand, since the nominative ἄλλος
can by no possible construction be made to refer to λαμπαδηφόρων but only
to νόμοι, which gives no reasonable sense.” He therefore advances a con-
jectural reading. No alteration, however, is required. The participial clause
1 The normal meaning of the common noun πορθμός = ‘strait’ (the meaning has already
become specialized in Homer from the original meaning which we conclude to have been
‘place for crossing’, cf. J. Holt, Glotta, xxvii, 1939, 192 f.) is completely confirmed by the
instances of the place-name Πορθμός quoted by Ferguson, Hesperia, vii, 1938, 68: (1) in
Euboea on the Euripus, probably the same place as the modern Skala Aliverioü, east of
Eretria, (2) on the narrow sound between Karpathos and the small island Saros north of it,
(3) the place near Sunion, the existence of which is known from the inscription published
by Ferguson, op. cit. (το f. rod ᾿Ηρακλείου τοῦ ἐπὶ Πορθμῶι, cf. 94 f. ἐφ᾽ ᾿Ηρακλείωι ἐπὶ
Σουνίου) ; Ferguson and Homer A. Thompson localize this Πορθμός on the bank of the small
bay north-west of the Poseidon temple at Sunion (cf. the map ‘Laurion’ in Baedeker’s
Griechenland, 5th ed., p. 126). This bay has on a much smaller scale almost the same shape
as the bay (already described) at the mouth of the Corinth Canal and bears the same
relationship to the true Saronic Gulf.
165
line 313 COMMENTARY

ἄλλος παρ᾽ ἄλλου κτλ. refers to the whole preceding clause τοιοίδε τοί μοι
λαμπαδηφόρων νόμοι. The binding arrangements (νόμοι) which Clytemnestra
has made for the torch-racers consist in their being ἄλλος παρ᾽ ἄλλου διαδοχαῖς
πληρούμενοι. The clause ἄλλος παρ᾽ ἄλλου κτλ. is clearly distributive, in
function not different from a differentiating clause such as of μὲν παραδιδόντες
τὴν λαμπάδα, οἱ de δεχόμενοι. This apposition of a nominative participle has
an analogue in such passages as Prom. 200 ff. στάσις δ᾽ ev ἀλλήλοισιν ὠροθύνετο,
οἱ μὲν θέλοντες ἐκβαλεῖν... οἱ δὲ τοὔμπαλιν σπεύδοντες κτλ. (the clause of μὲν
θέλοντες... οἱ dé . . . σπεύδοντες is in apposition not to ἀλλήλοισιν but to
στάσις δ᾽ ev ἀλλήλοισιν ὠροθύνετο), S. Ant. 259 1. λόγοι δ᾽ ἐν ἀλλήλοισιν ἐρρόθουν
κακοί, φύλαξ ἐλέγχων φύλακα (the participle illustrates the whole preceding
clause), E. Phoen. 1462 ff. ἦν δ᾽ ἔρις στρατηλάταις, οἱ μὲν πατάξαι πρόσθε
IToÂuveikn δορί, ot δὲ κτλ. (οἱ μὲν... οἱ δὲ... in apposition not to στρατηλάταις
but to ἦν δ᾽ ἔρις στρατηλάταις).
ἄλλος παρ᾽ ἄλλου : Plat. Laws 6. 776 Ὁ καθάπερ λαμπάδα τὸν βίον παραδιδόντας
ἄλλους ἐξ ἄλλων. For the echoing of the expression here in 490 πυρὸς παραλ-
Aayas cf. ad loc.
διαδοχαῖς : an appropriate word which recurs in the often quoted descrip-
tion of the Attic torch-race, best given in a Patmos scholion to Demosthenes
(Bull. Corr. Hell. 1, 1877, 11) οἱ ἔφηβοι, ἀλειψάμενοι παρὰ τοῦ γυμνασιάρχου, κατὰ
διαδοχὴν τρέχοντες ἧπτον τὸν βωμόν" καὶ 6 πρῶτος ἅψας ἐνίκα καὶ ἡ τούτου φυλή.
πληρούμενοι. If the signal failed to reach one of the watchmen, he would
go empty, οὐ πληρούμενος like the voting-urn 817. Since they all receive their
full share, they are πληρούμενοι. For πληροῦσθαι of persons Weil (Addenda)
compares E. Iph. T. 306 πολλοὶ δ᾽ ἐπληρώθημεν. The conception of each signal
as something quantitative comes out very clearly 301 (πλέον καίουσα). The
plural πληρούμενοι is in close relation to λαμπαδηφόρων, since πληροῦσθαι 15
said of them all ; ἄλλος παρ᾽ ἄλλου diadoxaisisaddedasa qualitative description.
314. It is improbable that complete certainty will ever be attained as to the
sense of this line. The difficulty! is in general a result of the compressed
brevity with which Aeschylus has played upon the connexion between the
torch-race with which his public was familiar and the transmission of the
fire-signals, which is here described as a kind of relay-race. There are in
addition particular difficulties caused by the ambiguity of some of the
phrasing : δέ may mean ‘and’ or ‘but’; ὁ πρῶτος δραμών can mean two quite
different things in relation to racing (see below); finally ὁ πρῶτος καὶ τελευ-
ratos may refer (see below) to two different people, but need not do so. We
shall have to be content with a more or less likely solution, if any can be
reached.
Paley found the main reason for this lack of clearness in the fact that
'there is so much obscurity on the real nature of the Athenian torch-race'.
This has cleared up since Paley's time. Details may still be uncertain, but
we are fairly well informed now on the main features of the Athenian
Lampadedromia, thanks to inscriptions and the exhaustive investigation by
N. Wecklein, Hermes, vii, 1873, 437 ff., the conclusions of which have been
followed up and attractively presented by P. Foucart, Rev. de philol. xxiii,
1 Platt, it is true, says (J. Phil. xxxii, 1913, 48) : ‘Really this is one of the simplest lines in
Aeschylus', and then goes on to produce a solution the simplicity of which is indeed amazing.
Í can but envy his confidence.
166
 COMMENTARY line 314

1899, 112 ff. The articles in Daremberg-Saglio and Pauly-Wissowa s.v.

‘Lampadedromia’, P. Stengel, Griech. Kultusaltertiimer (1920), 224, and also,
e.g., A. Körte, Arch. Jahrb. vii, 1892, 149 ff., rightly subscribe to the views of
those two scholars.’ It can thus be held as certain that the competition was
not between individual torch-racers, but between teams, each representing
the phyle? from whose members it had been chosen. The single runners in
each team were posted one after another at the appointed distances, abreast
of their opposite numbers in the other teams. The actual winner, i.e. the
runner who first kindled the fire on the altar with his torch, was held the
victor not on his own account but for his team and his phyle: καὶ 6 πρῶτος
ἅψας ἐνίκα καὶ ἡ τούτου φυλή (cf. on 313). It is quite clear from the inscriptions
that the phyle was regarded as the victor, e.g. IG ii. 1229 of the year 346/5
Akayavris ἐνίκα λαμπάδι Παναθήναια τὰ μεγάλα κτλ. Wecklein attempted to
bring out this point in his comment: ‘the first runner down to the last, as in
the Lampadedromia, in which several teams stood side by side and a whole
team was the winner’. Foucart, loc. cit. 113, rejects this, because it takes into
account only a single aspect of the ceremony and gives a wrong sense to the
line. He himself translates ‘la victoire est au coureur qui est arrivé le premier
et le dernier’, with the explanation: ‘le premier, par rapport aux concurrents
des autres tribus, le dernier, par rapport à la file des coureurs de sa tribu’.
This over-ingenious interpretation is improbable in itself, since it is too obvious
to need saying that the runner who reaches the goal is the one who comes
last, τελευταῖος, of his own team as reckoned from the start of the race; the
one, that is, who is posted nearest to the goal and so is the last to receive the
torch and run with it. With this meaning, the line, unmistakably intended
as the grand conclusion of the metaphor, would be entirely pointless. More
than this, such an interpretation is the result of ignoring the fundamental
difference between Clytemnestra’s own arrangement and the torch-race with
which she compares it (see below). The same applies to Wecklein’s explana-
tion quoted above, and he also sees the real content of the line in something
which any Athenian must have taken for granted; namely that victory,
indirectly at least, was accredited to all the team. (In this connexion one
might also doubt whether the thought vıräı 6 πρῶτος ἅψας---α briefer and
rather looser way of saying 6 πρῶτος δραμὼν καὶ ἅψας τὸν βωμόν----καὶ ἡ τούτου
φυλή could be properly expressed in the words ‘ νικᾶε 6 πρῶτος καὶ ὁ τελευταῖος
δραμών᾽.) In fact both scholars have here fallen into an error which is easily
made by just that school of interpretation which (laudably) keeps its eye on
the factual content of ancient texts and the premisses implicit in contempor-
ary allusions. The modern researcher is naturally interested in obtaining
further evidence on some detail of ancient life which we can recover only
with effort. This makes it easy for him to overlook how differently the ancient
poet was situated. The poet has a definite end in view, and it may be quite
unnecessary, perhaps even distracting, for him to emphasize a point of
detail which both he and his public can take for granted, while to the modern
antiquarian it appears as something worthy of careful note.
1 Further literature on the Lampadedromia in Broneer, Hesperia, xi, 1942, 149 f. |
2.1 am strongly in favour of A. W. Gomme’s attempts (C.R. lviii, 1944, 62) to ‘induce
scholars to drop the use of the word “tribe” for these highly artificial divisions of Greek
states’.
167
line 314 COMMENTARY
Unlike the above-mentioned commentators,! others see in the clause xd
δ᾽ 6 πρῶτος καὶ τελευταῖος δραμών a point not of likeness but of contrast,” and
therefore take δέ adversatively. So Klausen: ‘Qui ultimus erat in hoc cursu,
eiusdem est iuris et honoris atque qui primus: quum plerumque ultimus qui
ad terminum advenit, victoria et honore destitutus sit’ ; similarly in general,
but taking ὁ πρῶτος kai τελευταῖος as one and the same person, Schneidewin,
and, almost exactly as Klausen, Wilamowitz: ‘in Atheniensium lampade-
dromiis unus victor est ; hic vero victores dicendi sunt omnes, quia νικητήρια
πυρά incendunt’. Wilamowitz's last clause (‘quia . . .’), where he agrees with
Platt, loc. cit., is irrelevant to the main problem. As regards the decisive
point of Klausen’s and Wilamowitz’s interpretation, their view seems most
unlikely on account of the meaning which they ascribe to reAevraios. On
purely formal grounds there need be no trouble about taking τελευταῖος as =
ó τελευταῖος and thus ‘understanding’ a repetition of the article with the
second of a pair of contrasted things, as 324 τῶν ἁλόντων καὶ κρατησάντων, cf.
Jebb on S. El. 991 and on Oed. C. 606, Gildersleeve, Syntax of Classical Greek,
§ 605, p. 278. But the meaning thus given to ‘last’ does not fit the facts. To
realize this we must be clear about the sense of ὁ πρῶτος δραμών. Headlam
(whose own explanation is too artificial to be convincing) is quite right in
saying ' πρῶτος δραμεῖν, though it could mean “ to start first ” [6 πρῶτος
δραμών is in fact so used in Hdt. 8. 98. 2], usually meant “to finish first”.
The statement νικᾶν δ᾽ ὁ πρῶτος is in form and in matter parallel to the ὁ
πρῶτος ἅψας ἐνίκα of the Demosthenes scholion quoted above; from con-
temporary poetry one might compare Pind. Οἱ, 6. 75 πρώτοις περὶ δωδέκατον
δρόμον ἐλαυνόντεσσιν, i.e. in the words of the paraphrase: epi τὸν δωδεκά-
κυκλον δρόμον ἀγωνισαμένοις καὶ νικήσασιν. But this means that τελευταῖος is a
real correlative of πρῶτος, i.e. if πρῶτος is the first to reach the goal τελευταῖος
must be the last to reach it and cannot possibly refer either to the first runner
of the winning team who starts farthest from the goal, or to the last runner
who runs ‘home’, since this latter is à πρῶτος δραμών. There is a way out of
this dilemma, as we shall see presently. But first it should be noted that
even Klausen, Schneidewin, Wilamowitz, and others who correctly take 8é
as adversative still let Clytemnestra speak in terms which do not bring out
the fundamental difference between the Lampadedromia of Athenian festivals
and her own series of beacons. Of course a form of the metaphor could be
devised in which that difference could be treated as unimportant and even
pass unnoticed. This is clear from, e.g., the passage which is always quoted in
reference to these lines, Hdt. 8. 98. 2, where the writer, facing as it were in
the opposite direction, compares the Persian ἀγγαρήιον with the Greek torch-
race in honour of Hephaistos. All he is concerned with is the peculiar device
of the relay-race as a means of passing something on very fast; there is no
jarring element in his image since the question of victory does not arise.
But in Clytemnestra's speech victory is the final point to which all else is
t And, e.g., Phillimore, whose ingenious tour de force, C.R. xvii, 1903, 105 f., I need not
answer. It would be useless to discuss all the ill-considered suggestions which have been
published on this line.
2 On this crucial point L. Campbell is entirely right, Am. Journ. of Philol. i, 1880, 433:
‘Clytemnestra is not explaining to the Argive elders the nature of the λαμπαδηφορία which
they know well, but is pointing out the difference between her own and the ordinary
λαμπαδηφόρων νόμοι. His own interpretation that Hephaistos is victor is wrong.
168
 COMMENTARY line 319

leading up. A single team, posted from Ida to Argos, is running here and there-
fore there is no competition, and yet there is mention of a winner as in the
AtticLampadedromia. This paradox must have struck Athenian hearers as
something almost grotesque. From this I infer the possibility (nothing more)
that the point of 1. 314 may lie in this very paradox. ‘But the victor is he
who ran first and last’, since no other team was competing. She might just
as well have said ‘first and alone’, but that would be merely flat. She speaks
of a victory because her people, the watchman on the Arachnaion and also
the whole chain of watchers behind him (all represented by the one whose
beacon finally reaches the goal and lights the fire there, just as clearly as the
victorious runner in the torch-race represents the whole team of his phyle)—
because all these have played their part surpassingly well and achieved far
more than was demanded of them. All through the speech this was constantly
emphasized.
315. τέκμαρ: with this word she indicates that the long narrative is in fact an
answer to the questions put by the Chorus in 272 and all through the sticho-
mythia. Then there follows, more concretely, with re added (cf. on 214 f.),
σύμβολον, the agreed signal, referring to the beacon as in line 8; this idea leads
up to 316, since it was with Agamemnon that the agreement was made, and
he is the παραγγείλας at the other end of the message’s course.
τοιοῦτον: for the prosody cf. on 1256 f.
W. Sewell’s appreciation of the Beacon Speech (in the preface to his
translation) is worth reading side by side with the original. ‘Every word tells.
The fire becomes a courier, leaping from height to height. Instead of a vague
generalization, all the mountains are singled out and painted. The scaur
(Aéras) of Lemnos, with its sides scarred and seamed by torrents; the great
upheaved cone or hump (afzos) of Athos; the scaur again of Cithaeron ; the
watch-height of Macistus ; the beetling headland looking down on the Saronic
Gulf; the flame passing from pinnacle to pinnacle, now ridging the sea with
fire, now like “the golden sun’’, and now like “the pale clear moon" ; at one
time lifting itself like a gigantic torch, at another tossing and shaking its
"mighty beard of fire" ; now stooping over the Gorgopian lake, and now
blazing up from the pile of "grey old heath" : all these are exquisite poetical
touches, worthy of the first of sensuistic artists; that is, of artists who think
only of addressing their compositions to the senses.’
317. There is no indication of speaker, i.e. at some time the παράγραφος (cf.
on sor) was missed out ; 321 has it one line too late.
318. κἀποθαυμάσαι is interposed ‘dia μέσον between ἀκοῦσαι and διηνεκῶς
which go together; for a similar collocation cf. Lobeck on S. Aj. 475 f. (3rd
ed., p. 221 f.) ; Vahlen, Opusc. i. 113; Wilamowitz, Hermes, xviii, 1883, 246 n. 1,
and on E. Her. 222; Kaibel on S. El. 1358 (279 n. 1) ; Pearson on S. Ichn. 198;
Housman on Manilius 4. 534 (with ‘Addenda’, vol. v, p. 158); and below on
s6o and 611. For a phrase inserted between a verb and its adverb cf., e.g.,
WY 407 f. ἵππους δ᾽ Ἀτρεΐδαο κιχάνετε, μηδὲ λίπησθον, καρπαλίμως, and E. El.
617 φοβεῖται γάρ σε, κοὐχ εὕδει, σαφῶς, where Denniston rightly prefers the
meaning indicated by the punctuation. Such sandwiching of a phrase often
gives an effect of vivacity of speech (Wilamowitz).
319. διηνεκῶς : irrefutable evidence that the Attic form was duaver- is forth-
coming in the grammarians and in inscriptions (as early as the fourth century,
169
 line 310 COMMENTARY

cf. Meisterhans, 3rd ed., 16 n. 76), cf. also Wackernagel, Dehnungsgesetz, 40.
This is confirmed by the retention of διανεκῇ in a fragment of Anaxandrides
(ii. 137 Kock) quoted by Athenaeus 10. 455 f, and cf. also Plato, Hipp. Maj.
301 b and e. On the strength of this Blomfield put here διανεκῶς, perhaps
rightly. ‘Sed tragicis forma placere potuit Homerica’ (Dindorf). In Tragedy
(in which so far the adjective has not been found) the adverb occurs only
here. διηνεκέως occurs in three passages of the Odyssey, ὃ 836, n 241, μ 56,
always with ἀγορεύειν. Of the various interpretations in the scholia only one,
viz. ἕως τέλους τὰ πάντα, does full justice to the context of the passages (the
rendering of Ameis-Hentze on ὃ 836, ‘genau’, is wrong). Apparently Aeschylus
transferred the Homeric adverb from ἀγορεύειν to ἀκούειν and used it in the
same sense as Homer. It is therefore not sufficient to translate with Paley
and Sidgwick ‘at length’ or with Wilamowitz ‘ausfiihrlich’. What the queen
is asked to give is not a mere repetition with more elaborate details, but a
version that carries the tale on to its close. As the theme is on the whole to be
the same, the coryphaeus says ἀκοῦσαι... . πάλιν.
Although λέγεις and λέγοις, since the pronunciation had become the same,
would be easily confused, it is a priori likely that λέγεις represents the earlier
reading, while λέγοις seems to be due to the effort to obtain a more plausible
construction (cf., e.g., on 101, where the scribe of the hyparchetype of FTr had
before him the same φαίνεις which we find in M, but tampered with it in order
to improve the construction). The man who wrote ὡς λέγοις πάλιν probably
wanted it to mean wt dicas tterum, but that would not do ‘quia sic [with the
preceding λόγους ἀκοῦσαι διηνεκῶς θέλοιμ᾽ ἄν] inverteretur ordo sententiarum’
(Hermann). Wellauer and others took ὡς in a causal sense (= ‘as’) and λέγοις
as equivalent to an imperative. This interpretation, too, was rejected by
Hermann as not sufficiently polite. Kennedy and Sidgwick (followed by
Verrall) regard the optative λέγοις as assimilated to θέλοιμ᾽ av. Kennedy's
rendering (in his critical note), ‘but I should like to hear again . . . how you
tell this story to its close’, tears διηνεκῶς out of its context (it goes with
ἀκοῦσαι), and the interpretation of Sidgwick, ‘according as thou wouldst tell
it again’, seems unnatural. Nor does Headlam’s suggestion (ἕως λέγοις, ‘so
long as you should speak’) yield a satisfactory sense; besides, no instance of
monosyllabic &ws has as yet been found in Aeschylus. All these attempts
suffer from the difficulty of accounting for the optative λέγοις. A much
simpler arrangement of the sentence is obtained by Bothe’s οὗς λέγεις, which
Hermann and others accepted. The ‘pleonastic’ addition to λόγους τούσδε of
a postponed οὖς λέγεις may perhaps express a certain excitement (cf. on the
position διὰ μέσου in the preceding line). The same reason seems to apply to
the separation of πάλιν from the phrase to which it belongs. So far so good.
But the ὡς of the MSS should not be altered.’ The addition to λόγους τούσδε
of a pleonastic supplement by means of a ós-clause is idiomatic. Cf. H 406 f.
4 τοι μῦθον ᾿Αχαιῶν αὐτὸς ἀκούεις, ὥς τοι ὑποκρίνονται, i 62 ἀλλ᾽ οὐκ ἔσθ᾽ ὅδε
μῦθος ἐτήτυμος, ὡς ἀγορεύεις (and, slightly different, Xen. Cyr. 8. 2. 14 καὶ
λόγος δὲ αὐτοῦ ἀπομνημονεύεται, ὡς λέγοι παραπλήσια ἔργα εἶναι κτλ). We
may, if we like, regard such constructions as a ‘contamination’ of μῦθον Ἀχαιῶν
1 ὡς λέγεις (conjectured by Blomfield before the reading of V was known) was put in the
text by Franz and Keck and also by Wilamowitz when he published his translation (see,
at the end of it, his remark on l. 306 Kirchhoff).
170
 COMMENTARY line 323

ἀκούεις and ἀκούεις ws ὑποκρίνονται--50 here of λόγους τούσδε ἀκοῦσαι and

ἀκοῦσαι ὡς λέγεις.
320. Τροίαν Ἀχαιοί: the same opening effect as 269, here still further em-
phasized by the word-order of the whole sentence. This is an example of the
very old and widespread tendency to place side by side not only similar
words, or words of the same root, but also contrasted ideas, or ideas which
complete each other (Fr. Leo, Nachr. Gótt. Ges. 1895, 433, cf. also Hermes,
Ixvii, 1932, 370 n. 4). A good example is Cho. 89 f. πότερα λέγουσα παρὰ φίλης
φίλωι φέρειν γυναικὸς ἀνδρί, and again Sept. 695 f. φίλου γὰρ ἐχθρά μοι πατρὸς...
ἀρὰ ξηροῖς ἄκλαυτος (a good restoration) ὄμμασιν προσιζάνει (Menander fr. 245
Kock οὐδεὶς δι᾽ ἀνθρώπου θεὸς σώιζει, γύναι, ἑτέρου τὸν ἕτερον). Cf. 1272, Eum.
825 μηδὲ... θεαὶ βροτῶν κτίσητε δύσκηλον χθόνα, and on 1318.
ἔχουσι in this context is almost technical ; so Herodotus, referring to the
events which presumably contributed to Aeschylus’ invention of the victory
signal, says (9. 3. 1) ἅμα δὲ πύρσοισι διὰ νήσων ἐδόκεε (sc. τῶι Mapôoviur)
Baci δηλώσειν ἐόντι ἐν Σάρδισι ὅτι ἔχοι ᾿Αθήνας. Cf. A. W. Gomme, A
Historical Cominentary on Thucydides, 1. 304. Beazley recalls Sir Charles
Napier's dispatch 'Peccavi' (I have Sind).
321. πρέπειν is said mostly of the conspicuousness (cf. on 242) of visible things,
here of audible. The lexicons compare Pind. N. 3. 67 Boa δὲ νικαφόρωι σὺν
Apıorordeidaı πρέπει (Schol. ἐκπρεπής ἐστιν); cf. also Soph. Ichn. (fr. 314 P.)
224 Í. θαῦμα yàp karékAvov, ὁμοῦ πρέπον κέλευμά πως κυνηγετῶν krÀ. Cf. Lobeck,
PHMATIKON, 340.
322. For the re after ὄξος cf. on 324 f.
ἄλειφα : besides L-S cf. (for references in the ancient grammarians) Passow-
Crönert s.v. and F. Jacoby on Hesiod, Theog. 553.
323. The reading of the MSS is sound. Of Stanley’s conjecture od φίλω
Headlam said ‘it may well be right’, Wilamowitz put it in the text. Ahrens,
p. 504 (and others before him), had already rejected it, with the fitting com-
ment: ‘while οὐ φίλω contains an unnecessary tautology, où φίλως, i.e.
inimically, indicates by a typical litotes the irreconcilable feud between
vinegar and oil, the ὥμικτον.᾽ Objections to the conjecture φίλω cannot be
based on the grounds that διχοστατοῦντ᾽ must then be taken as -oóvre, dual,
since the dual of the participle, though rare in Aeschylus, is definitely estab-
lished for the Oresteia (nowhere else as yet): Ag. 652 φθείροντε, 1673 Kparoüvre,
Cho. 254 éyovre, 564 μιμουμένω, Eum. 428 mapôvrow ; this has been pointed
out by A. Cuny, Le Nombre duel en Grec (1906), 117, on the basis of the
instances collected by E. Hasse. Headlam rightly insists that où φίλως, if we
retain it, must be taken with διχοστατοῦντ᾽. It would, however, be wrong to
water down the peculiar expression by giving the verb a neutral sort of tone,
as Verrall tentatively does; Conington’s question is to the point: ‘what
would φίλως διχοστατεῖν be?’ Antagonism, amounting sometimes to hostility,
appears to have been always implied in early Greek in the word διχοστατεῖν
(the gloss in Hesychius S:yoorare?: μάχεται, πολεμεῖ is correct, as also the
gloss in Tr on διχαστατοῦντ᾽ Ag. 323: μαχόμενα) ; the earlier use of διχοστασίη
(Solon fr. 3. 37 D., Theogn. 78, Bacchyl. 11(10). 67) agrees with this. Thus
1 It is in the nature of ἔχειν and kindred verbs in other languages that they easily take
on the meaning ‘to hold, occupy, rule’ and the like. On this much-discussed phenomenon
see, e.g., H. Frisk, Studien zur griechischen Wortstellung (Gothenburg 1933), 157 f.
171
line 323 COMMENTARY
διχοστατοῦντ᾽ οὐ φίλως has a certain ‘fullness’ of expression. But such fullness
is exactly in Aeschylus’ manner when he is constructing a litotes with the
help of an adverb negatived by οὐ. This point has eluded the critics who
would alter the text, even as it has elsewhere, cf. Pers. 325 κεῖται θανὼν
δείλαιος οὐ μάλ᾽ εὐτυχῶς, where Weil has conjectured εὐθέτως and Wecklein
complains of the ‘tame and pointless edruyds’. In Cho. 645 τὸ πᾶν Διὸς σέβας
παρεκβάντας οὐ θεμιστῶς the adverb has likewise been challenged ; Blass justly
remarks: ‘here again the expression is very full.” Another instance of the
kind is Ag. 805 οὐδ᾽ ἀφίλως εὔφρων, where even Ahrens has altered ἀφίλως as
having ‘no reasonable sense’. Cho. 67 riras φόνος πέπηγεν (for the meaning
see below on 392) οὐ διαρρύδαν contains a similarly ‘superfluous’ addition of
the same type.
προσεννέποις. Here, too, a fine point is lost by attempts to blunt the sharp
edge of the sense; so the gloss in Tr εἴποις dv,’ Stanley: dixeris, and later
editors, as, e.g., Paley: ‘you would say’ (also incorrectly in his commentary:
“ προσεννέπειν is here “to speak of" ') and even Lobel (Pap. Oxy. xviii, p. 21,
on 2162, fr. 1, col. 1. 24 1): ‘Ag. 323 "say" '. The sense of the verb is here
what it always is? (cf. on 162 and 242 f.), ‘to address or call upon by name’.
It appears correctly under 'appello' in Wellauer and Dindorf, so also Lin-
wood: ‘to speak to or address’. The personification of the two antagonists,
vinegar and oil, apparent in διχοστατοῦντε, is here continued: they are no
inanimate objects, but hostile creatures in conflict with each other. The man
who is occupied in pouring them into a vessel looks down with wonder at the
passionate struggle of the two. He might ask: 'by what name should I
properly address you?’ The answer is: διχοστατοῦντ᾽ οὐ φίλως.
324 f. The connexion of these two lines with the preceding two in the form
of a ‘comparatio paratactica', the first part being in thought subordinate to
the second, was noted, e.g., by Paley, Schneidewin, Wecklein, Verrall. The
connecting particles are here ve (ὄξος re) . . . καί. Paley also protested,
rightly, against the full stop after 322, which is disruptive of the sense ; later
editors have sometimes disregarded this. δίχα 324 emphasizes the close
interdependence of these two pairs of lines.
324. Cf. (Blomfield) A 450 f. ἔνθα δ᾽ ἅμ᾽ οἰμωγή τε καὶ εὐχωλὴ πέλεν ἀνδρῶν
ὀλλύντων τε καὶ ὀλλυμένων.
For the article which serves the two participles cf. on 314.
325. Blomfield’s comma before συμφορᾶς διπλῆς points the sense clearly.
Aeschylus has a liking for the explanatory afterthought at the end of the
sentence. The addition gives clarity to the whole expression δίχα φθογγὰς
ἀκούειν ἔστι' ἤγουν ἐκ διπλῆς συμφορᾶς γιγνομένας καὶ διπλῆν συμφορὰν ἀπο-
φαινούσας.
3271. Το φυταλμίων παῖδες γερόντων exception was taken as early as
Auratus; a full discussion in Ahrens, 505 f., who adds to older conjectural
1 Triclinius’ paraphrase προσείποις ἂν αὐτά, ἤγουν καλέσεις κτλ. is much better.
2 So in Pind. Zsthm. 6. 17, where L-S assume a special meaning ‘entreat or command’.
Wilamowitz, Berl. Sitzgsb. 1909, 818, gives the right sense: ‘Ich richte mein Gebet an die
Moiren, dass sie seinem Rufe Folge leisten’ ; literally, ‘I call upon Clotho and her sisters by
name’; this is the act of praying, and the content of the prayer is given in the infinitive
clause which follows. For προσεννέπειν in the address of a prayer cf. on 162. In Pap. Oxy.
2162, fr. 1, col. 1. 25 (Aesch. Isthmiastae) I consider the supplement mpooev[vere: extremely
improbable.
172
 COMMENTARY line 331
emendations a particularly violent one of his own. Weil’s suggestion φυτάλμιοι
παίδων γέροντες was adopted by Wecklein and Headlam among others.
Most commentators, especially Hermann and Ahrens, have failed to realize
that the poet is not concerned here to table more or less exhaustively the
‘various categories’ of the captured and slain. He is merely selecting from
the mass of the conquered a few particularly impressive groups: women over
the bodies of their husbands, others (probably most of them, though not all,
women) over the bodies of their brothers, finally’ children over the bodies of
their φυτάλμιοι γέροντες. What does this expression mean? Almost all com-
mentators and translators take it as simply ‘fathers’. Not so Wilamowitz,
who in his translation, contrary to his usual custom, expands in a paraphrase
in order to get the full sense: ‘Die einen neben Leichen, hier des Gatten, dort
des Bruders, neben einem Greise hier ein Kind, das jüngste Reis des Stammes
der von ihm entspross’. This implies, without stating in so many words, that
grandchildren are in question. There is in fact nothing in the expression
παῖδες ἀμφὶ φυταλμίων γερόντων σώμασιν πεπτωκότες to exclude this idea. It
is arbitrary to render φυτάλμιος by ‘parens’. Soph. fr. 720 N. (= 788 P.)
φυταλμίωι πατρί is no argument in favour, since, as Pearson says in his note,
‘the adjective signifies procreative’. φυτάλμιος does not commit itself to any
particular generation in relation to the procreated, any more than γενέσιος
and γενέθλιος, which occur as cult-names of Poseidon (φυτάλμιος being also
thus used, cf. Wilamowitz, Glaube d. Hell. i. 213). Thus the poet’s expression,
intentionally vague no doubt, leaves us free to imagine the young children of
a grey-haired father as well as grandchildren. Every hearer will have before
his eyes above all the picture of the slaughtered Priam. The exact relation-
ship is here less important than the pitiful contrast of youth and age. The
point is well made by P. Menge, De poet. scaen. Graec. serm. (Göttingen 1905),
22. It isnatural enough that women and children are prominent in the groups
of captives. Conington quotes the Agamemnon lines in his note on the
passage in the Virgilian Iliupersis (Aen. 2. 766 1.) where after the description
of the Trojan booty come the words: pueri et pavidae longo ordine matres
stant circum. Literary reminiscence is hardly in question; it is the common
human experience in such a calamity.
328. For ἐλεύθερος as an adjective of two terminations the only parallel
quoted in the lexicons is E. El. 868.
329. ἀποιμώζειν so far as is known does not occur before Aeschylus (cf. fr.
138 N.) ; for ἀπο- cf. on 1544.
Though φίλος often indicates merely a close relationship (whether of blood
or other kind), in φίλτατος the emotional connotation is unmistakable.
330. αὖτε ‘freely taken over from Epic only by Aeschylus' (Wilamowitz on
Ar. Lys. 66). For its use elsewhere (seldom in other dramatic authors, never
in prose) cf. L-S. For the function of δ᾽ αὖτε cf. on 512.
νυκτίπλαγκτος : cf. on 12.
331. Troy is far from starving; the seer's prophecy 129 spoke of herds of
cattle which should be the spoil of the conqueror; here as elsewhere that
prophecy is meant to coalesce with the later account of actual events into a
single picture in the mind of the hearer.
ı For re καί, which here does not connect the two phrases between which it stands, cf.
on 403 f.
173
line 332 COMMENTARY
332. πρὸς οὐδὲν ἐν μέρει τεκμήριον : the preposition as in Hdt. 3. 52. 2 πρὸς
ὧν δὴ τοῦτο τὸ κήρυγμα οὔτε τίς οὗ διαλέγεσθαι... ἤθελε and elsewhere, cf.
Kühner-Gerth i. 520. “Without each being allotted a special billet ; an addi-
tion prompted by τάσσει᾽ (Schneidewin).
333. Cf. Sept. 376 ws τ᾽ ev πύλαις ἕκαστος εἴληχεν πάλον.
After πάλον there is no punctuation in V and F (a comma in Tr). A full stop
was first inserted here, as far as I can trace, by Blomfield, and has been
retained in some English editions down to the present day, though Paley,
Verrall, and Headlam, for instance, are here correct. The antithesis to
mpôs . . . τεκμήριον is developed in a quite natural manner through the series
of clauses 333-7. It might perhaps be said that the wealth of detail which
comes to swell the queen’s vivid narrative has led to an extension beyond the
limit originally intended.
335. ἤδη: cf. on 971.
335 f. For the content cf. 560 ft.
336. Casaubon’s emendation ὡς δ᾽ εὐδαίμονες has been adopted by most
editors.! Martin takes ws as exclamatory: ‘quam beati vero securam dor-
mient totam noctem’; so Schneidewin and Weil and of later editors, e.g.,
Headlam and Murray. Ahrens (p. 507), while rightly rejecting this, proceeds
to complicate the thought unnecessarily : ‘and they will . . . sleep in the belief
that they are now εὐδαίμονες, 1.6. under the protection of the gods’. Wecklein
follows this. ὡς here serves the same purpose as in, e.g., Pers. 710 f. ζηλωτὸς
ὧν βίοτον edaiwva Πέρσαιςὡς θεὸς διήγαγες ---ἃ simple comparison. The parallel
which clinches the matter seems to have escaped the critics: Demosth. 19.
235 ἑώρων αὐτοὺς καὶ ἐπὶ τοῖς τοιούτοις ἐκεῖ σεμνυνομένους ὡς εὐδαίμονας Kal
λαμπρούς.
εὐδαίμονες : ‘who live a grand life’; here presumably in a somewhat less
restricted sense than Cho. 700 or the well-known passages in Herodotus,
Thucydides, etc., where of εὐδαίμονες means simply the rich, but that use
points the way for interpretation here. Wilamowitz, Glaube d. Hell. i. 369,
brackets Ag. 336 with Hesiod, Erga 826, and Theognis, 1013, where ‘ εὐδαίμων
and ὄλβιος occur together as synonyms'. 'He's having a good time’ is ex-
pressed in ordinary Attic by εὐδαιμονεῖ γ᾽ ἄνθρωπος (Ar. Ach. 836). In E. Or.
1606 εὐδαίμων ἔφυ practically amounts to 'he is to be envied'. Iocasta's bold
definition of tyranny as ἀδικίαν εὐδαίμονα points in the same direction. It is
of course possible that in the mind of Clytemnestra as of the Chorus and the
thoughtful spectator the εὐδαιμονία of the conquerors in the short breathing-
space between past troubles and the awful home-going that awaited them
appeared at best dubious, but the poet has not emphasized the point. There
is in Aeschylus a close parallel, in matter as well as form, to the expres-
sion ὡς εὐδαίμονες, but as the passage, even after Wilamowitz's satisfactory
reconstitution of the text, has been continually misunderstood and then
altered unnecessarily, I must set it out here with a brief explanation. Sept.
363 ff. δμωίδες δὲ καινοπήμονες νέαι" | τλήμον᾽ αἶσαν αἰχμάλωτον | ἀνδρὸς εὐτυ-
χοῦντος ὧς | δυσμενοῦς ὑπερτέρου | ἐλπίς ἐστι νύκτερον τέλος μολεῖν. Wilamowitz
has here adopted Prien’s emendation of the MS reading τλήμονες εὐνάν,
ı F. M. Cornford, Thucydides Mythistoricus (1907), 156, defends δυσδαέμονες, in which he
sees ‘a magnificent stroke of irony’. Verrall’s retaining of the MS reading will surprise no
one familiar with his ‘conservative’ textual criticism.

174
 COMMENTARY lines 338 f.

suggesting in the apparatus a probable explanation of the corruption, and

has also—this is the cardinal point—given ws 365 an accent, thus making it
refer to the preceding words (for the position after substantive plus attri-
bute cf., e.g., Ag. 277). ‘And the young handmaidens have new grief: it is to
be expected that the unhappy fate of captivity, if the foe like a man favoured
of fortune should prevail, will in the night be fulfilled.’ ὑπερτέρου is correctly
rendered κρατήσαντος in the scholiast’s paraphrase. What the enemy over-
powers is not so much the conquered city in general as the special booty here
in question, the young slave-women. The comparison which describes the
victor’s attitude is pitched lower than in the Agamemnon passage: the enemy
takes possession of his booty like one whose enterprise is attended with success
(εὐτυχεῖ), but the difference is only a matter of nuance. The prospect which
the frightened girls have in mind is the same as that depicted by the captive
Trojan women (E. Tro. 563 ff.) ἔν τε δεμνίοις καρατόμος ἐρημία νεανίδων
στέφανον ἔφερεν “Ἑλλάδι κουροτρόφον, Φρυγῶν δὲ πατρίδι πένθος, but Euripides
is here again more crudely outspoken. In Aeschylus the worst that awaits
them is expressed by the veiled yet unambiguous νύκτερον τέλος, in which
every Greek must have heard an echo of τέλος γάμου. It is a significant
illustration of the poet’s tendency to give bona verba a twist which turns
them to horror (cf. on 65).
For the analogy between the Aeschylean description (Ag. 320-37) and the
'IMov πέρσις of Polygnotus see Beazley in Beazley and Ashmole, Greek
Sculpture and Painting, 38. Within the play the description of what happened
after the capture of Troy supplements the king’s account of the capture
itself (825 ff.).
338. Most editors! have followed Scaliger in dividing εὖ σέβουσι, doubtless
influenced by Valckenaer's and Porson’s observations on E. Phoen. 1321
(i331 Valck.), cf. also K. W. Krüger, Griech. Sprachlehre, ii, § 46. 8 n. 3.
Dindorf, Thesaurus, iii. 2443, has some careful arguments on the other side;
so also Ahrens (507) and Headlam on this passage. Blass on Eum. 1020 and
L-S s.v. σέβομαι incline, though with some hesitation, to keep εὐσεβοῦσι. In
E. Tro. 85 transitive εὐσεβεῖν is certain? (εὖ σέβειν would be absurd in view of
the line which follows’), and similarly Antiphon 3 y 11 οὔθ᾽ of θανατώσαντες
ἡμᾶς... εὐσεβοῖντ᾽ ἂν ὑπὸ τῶν ἀπολυσάντων τοὺς ἀνοσίους. Less certain are
E. Phoen. 1321 and Frag. trag. adesp. 350 N. There seems no very good
reason for rejecting εὐσεβοῦσι τοὺς θεούς in Aeschylus, but it is of course
impossible to say with certainty how he meant his EYZEBOZI to be spoken.
338 f. τοὺς πολισσούχους θεοὺς τοὺς τῆς ἁλούσης γῆς: Cf. Sept. 271 χώρας τοῖς
πολισσούχοις θεοῖς, on which see Regenbogen, Hermes, Ixviii, 1933, 59 n. 1.
For the sentiment in 338 ff. cf. Pers. 809 ff. Beazley, Der Kleophrades-
Maler (1933), 27, brings the lines Ag. 338 f. into his description of the Vivenzio
vase and recalls how the altar of Zeus and the statue of Athena, both of
1 Including Hermann, notwithstanding his own remark on S. Ant. 731 (727 Herm.).
2 I do not understand why A. C. Pearson on E. Phoen. 1321 says of Tro. 85 ‘even there
εὖ σέβειν may be right’.
3 For the common habit of picking up the compound by the simple form of a verb without
any modification of the meaning cf., e.g., Elmsley on E. Med. 1219 f., Housman on Manilius
3. 328, Wackernagel, Syntax, ii. 177. In the so-called ‘Urfaust’ Goethe adınitted the fine old
usage of living speech : ‘Wer Recht behalten will und hat nur eine Zunge, Der hält’s gewiss’,
but in the later form of the play this was watered down to the timid 'Behált's gewiss’.
175
lines 338 f. COMMENTARY
which were desecrated by the victors, are sometimes found side by side in
representations of the Sack of Troy belonging to the period when Aeschylus
was a young man.
340. The two emendations are certain. The repetition of dv in this and many
similar instances is probably to be regarded as the result of a compromise
between the very old order of words according to which ἄν occupies the
second place in the sentence and the later tendency of putting it beside the
verb; cf. Wackernagel, Idg. Forsch. i, 1891, 399 ff.
The proverbial tone of the expression was recognized by Schneidewin,
Wecklein, Headlam, and parallels adduced, e.g. Cho. 499, S. Oed. C. 1025
γνῶθι δ᾽ ὡς ἔχων ἔχηι (I would add here the remark of Aristippus about his
relations with Lais, Athen. 12. 544 d, ἔχω καὶ οὐκ ἔχομαι, translated by Cic.
ad Fam. 9. 26. 2), Antiphilos Anth. Pal. 9. 14. 7 εἷλε δ᾽ ἁλούς (inadequately
explained by Karl Miller, Die Epigramme des Antiphilos von Byzanz, 1935, 69).
341. ἐμπίπτοι looks like a Byzantine conjecture, but given the similarity in
sound of the different endings it is useless to defend -« or -n as being the
better attested; only sense and grammar can decide. As far as sense goes,
the optative, upheld by, e.g., Dindorf, Conington, Wilamowitz, is scarcely
what is wanted ; a wish is too weak here, while an expression of fear is what
one expects after the description of the victors’ heedless indifference, and
is in keeping with the tone of foreboding which marks the conclusion of the
speech. Commentators refer appropriately to 131, which, in conjunction with
Suppl. 357 f., 399 f., shows that Goodwin, ὃ 264 (cf. also his Appendix II,
p. 392), is wrong in his statement that ‘after Homer we find no examples of
the independent clause with either μή or μὴ οὐ until Euripides’; Kühner-
Gerth i. 224 is also very inadequate on this use. The peculiarity of the
present in ἐμπίπτηι struck Hermann’s fine linguistic sense : ‘metuendi verbum
si intelligitur, recte se habet coniunctivus praesentis, quo tempore utitur
Clytaemnestra, quod praesentia, non futura, metuit’ (this would make it a
case like € 356 ὦ μοι ἐγώ, μή τίς μοι ὑφαίνηισιν δόλον αὖτε ἀθανάτων) ; so, too,
Headlam, with the comment ‘she is still imagining the scene’. This may be
right, since in later instances of such j}-clauses, where the fear refers to the
future, we do in fact find the aorist. It is, however, possible that the addition
of πρότερον (temporal, not to be taken as = potius, as Klausen and Ahrens)
has influenced the choice of tense. Moreover, in Homer o 19, π 87 (Goodwin,
$ 261) the present subjunctive occurs in a similar independent un-clause to
denote fear of something to come.
ἐμπίπτειν expresses the overmastering force with which covetousness
seizes upon men, cf. the ἐμπίτνειν of the evil spirit 1175 (cf. ad loc.) and 1468,
the invincible Eros (5. Ant. 782) ἐν κτήμασι πίπτει, Thuc. 6. 24. 3 ἔρως ἐνέπεσε
τοῖς πᾶσιν ὁμοίως ἐκπλεῦσαι (with significant force).
344. With the infinitive added to complete the sense of Se? . . . σωτηρίας
Paley aptly compares Suppl. 407 ff., where an acc. c. inf. is added to δεῖ...
φροντίδος σωτηρίου.
The warning delivered in ll. 338-44, as Paley among others points out,
hints in the first place at the younger Ajax’s desecration of the cult-statue of
Athena, as described in Epic (Κασσάνδραν δὲ Αἴας ὁ ' IAéws πρὸς βίαν ἀποσπῶν
συνεφέλκεται τὸ τῆς ᾿Αθηνᾶς ξόανον), and his consequent fate on the homeward
voyage; thus it prepares the way for the description of the storm in 649 ff.
176
 COMMENTARY lines 345-7
(cf. Wilamowitz, Griech. Tragoedien ii. 32 n. 1; P. Friedlander, Die Antike, i,
1925, 16f.). But there are other similar acts of sacrilege and destruction
which we are also meant to recall. Nothing points here to any impiety for
which Agamemnon in particular might be held responsible, since these
transgressors are to forfeit their νόστιμος σωτηρία (they do in fact perish in the
storm later described), while Agamemnon reaches home.
345. ἀναμπλάκητος : for the widespread use of ἀμπλακεῖν and its derivatives
and the religious significance of these words cf. Wilamowitz, Glaube d. Heil.
ji. 120 f.
μόλοι: cf. on 675.
345-7. A very difficult passage. Wilamowitz's treatment of it is singularly
unhappy, but can usefully serve as a starting-point for discussion. There
should be no tampering with the perfectly good clause ei πρόσπαια μὴ τύχοι
xaxd, least of all to make it into a dubious asyndeton. This conditional
clause 'if only no new (sudden) evil supervene' clearly indicates that in the
preceding apodosis ending γένοιτ᾽ dv some comforting possibility, however
slight, must have been expressed, in sense something like 'then there would
still be hope', 'then the issue might yet be good'. All commentators who in a
desperate effort to save in 346 ἐγρήγορον or at least ἐγρηγορός (in Eum. 706 the
correct ἐγρηγορὸς of M appears as ἐγρήγορον in FTr) get a presage of evil out
of these words—resurrection of the sufferings of the killed men or something
of the kind—are compelled to give the simple εἰ which follows a concessive
twist,” translating it as 'etiamsi', ‘even if’, ‘although’, etc. ; so, e.g., following
Stanley, whose etiam is clearly a misprint for etiamsi, Heath, Schütz, Nägels-
bach, Sidgwick (‘there still might wake the woe of the dead’), and recently
P. Friedlánder, Die Antike, i, 1925, 17, and E. Neustadt, Hermes, lxiv, 1929,
257. Wecklein, Stud. z. Aesch. (1872) 11o, avoided this mistake. It is clear
from 345 and from the antithesis here, marked by δέ, to 341 f., that 346 must
contain a favourable possibility ; but here again, in an attempt to evade the
logical sequence, the εἰ (in 345) has been improperly translated as if it were
εἰ δὲ καί (Headlam, C.R. xvii, 1903, 290, is good on this point). 346, then, is
quite indispensable; there are no grounds for assuming that the corruption
extends beyond the first word. ἐγρήγορον is surely discredited by the fact
that the ‘wakefulness of the sufferings of the dead’ cannot but force the
thought in a direction contrary to that so clearly indicated by the surround-
ing uncorrupt lines. Headlam's discussion, C.R. xii, 1898, 245 f., seems to me
decisive both for a general understanding of the passage and for a limitation
of the corruption to ἐγρήγορον. He makes in this article only cautious
suggestions as to the word which originally stood in the place of ἐγρήγορον:
Kirchhoff had proposed παρήγορον, which is nowhere attested in a passive
sense; Headlam accordingly modifies it to παρηγορεῖν, which would thus
! Wilamowitz's exchange of εἰ for εὖ Suppl. 86 is no more successful.
2 The reader who checks the passages quoted by Kühner-Gerth ii. 489, Anmerk. 2, will
see that they provide no justification for assuming a ‘concessive’ use of εἰ here.
3 I had reached the position here set out, including the assumption of a probable
παρήγορον, or, in default of other instances with the required sense, παρηγορεῖν, long before
I became acquainted with Headlam's article or his text. I mention this because it seems to
me to show that in this passage, if violent remedies such as the transpositions put forward
by several critics are rejected, one is driven step by step to the conclusions reached by
Headlam, which have unfortunately not had the success they deserve.
4872.2 N 177
lines 345-7 COMMENTARY

reproduce exactly, except for the tense of the infinitive, the construction
which we find in 34 f. For the meaning see 95 and also L-S παρηγορέω ii. 2.
Later (C.R. xvii, 1903, 289) Headlam declared for εὐήγορον = εὔφημον: ‘the
sore hurt of the perished should grow gentle in its language’, a conception
which I find difficult to realize. Certainty about the word to be restored here
is unattainable.
347. πρόσπαια: cf. on 187.
348. γυναικός; we must be careful not to destroy the deliberate vagueness
of the word by confining it to one particular notion. Paley paraphrases:
‘Though I am only a woman, I have such advice and such sage precepts to
offer.” But since a concluding τοιαῦτα usually embraces the whole of a
preceding speech (cf. on 613 f.), it is unlikely that it refers here only to the
last section and not to the description of the captured town (321 ff.) as well.
Neither does κλύεις favour a narrow limitation of the meaning. Clytemnestra
is probably calling attention to her superior, man-like insight into the nature
of human affairs, including her knowledge of the reverence due to the gods,
and also her experience of what life is like in the midst of the turmoil of war.
This latter is particularly remarkable in a woman.
κλύεις : ‘at the end of a speech, implying that all is said’ (Conington on Cho.
443 κλύεις πατρώιους δύας ἀτίμους, where he compares also Prom. 683 κλύεις
τὰ πραχθέντα). Cf. on 1431. For the tense see Wackernagel, Syntax, i. 166.
348 f. Since 348 concludes the whole address, the following τὸ δ᾽ εὖ κρατοίη
no longer refers, like the section beginning 338, simply to the home-coming of
the warriors, but is directed to the future in general. It is a prayer of dreadful
ambiguity. Clytemnestra has every reason to pray in such terms for herself, but
in the ear of the spectator, who guesses what slic has in mind to do, the words
must have rung like a travesty of the repeated τὸ δ᾽ εὖ νικάτω of the parodos.
349. μὴ διχορρόπως: cf. on 1272.
350. Bothe, Klausen, and many other commentators have pointed out that
εἱλόμην here is an example of the familiar (cf. the grammar-books and also
Wackernagel, Syntax, i. 176) idiomatic aorist like ἐπήινεσα, etc. The main
difficulty of the line lies in the ambiguity of the relation between πολλῶν...
ἐσθλῶν and the remainder of the clause. A very sensible explanation is given
in the scholion : ὅτι πολλῶν χρημάτων προκριτέα ἡ ὄνησις * εἰ γὰρ ἐξ ὧν ἔχει τις
ἀγαθῶν ὄνησίν τινα μὴ λαμβάνει, μάταια πάντ᾽ ἂν ein. Ihe interpretation of
Hermann (cf. Ofusc. i. 180; on his conjecture τήνδ᾽ see below) and Franz
agrees in substance with that of the scholion: 'hunc ego fructum multae
prosperitati praefero' (Hermann), ‘statt vieler Güter wähl’ ich solchen
Vollgenuss' (Franz). These scholars take αἱρεῖσθαι as meaning προαιρεῖσθαι,
make πολλῶν ἐσθλῶν (genitive of comparison) depend on the verb, and assume
a contrast between πολλὰ ἐσθλά, further blessings, and the enjoyment of the
blessings which are already at hand. I believe this interpretation to be highly
probable. ‘ αἱροῦμαι, "I make my choice”, is used frequently in Aeschylus of
choosing or closing with a piece of fortune, nearly like the Latin οῤίο᾽
(Conington on Cho. 933). The shading off of αἱρεῖσθαι = ‘choose’ into
αἱρεῖσθαι = ‘prefer’ is almost imperceptible. Dindorf in the Thesaurus and
L-S quote S. Phil. 1099 f. τοῦ Awiovos' δαίμονος εἵλου τὸ κάκιον αἰνεῖν (ἑλεῖν
1 A, C, Pearson’s ‘unattractive’ (Housman, C.R. xxxix, 1925, 78) conjecture λώιονος ἐκ
is quite unnecessary; the MS reading is probably correct (cf. Radermacher, ad loc.;
178
 COMMENTARY line 350

codd. : corr. Hermann) as an instance of αἱρεῖσθαί ri τινος, ‘to prefer something

to something else’. With this we may compare (cf. Passow-Crönert, 173. 60)
Pind. N. 10. 58 f. τοῦτον, 7] πάμπαν θεὸς ἔμμεναι οἰκεῖν τ᾽ οὐρανῶι, eiÂer” αἰῶνα,
Antiphanes fr. 187 K. 7j μὴν ἑλέσθαι τοῦτον ἂν ζῆν τὸν βίον 7) τὴν Σελεύκου τοῦ
βασιλέως ὑπεροχήν, Theocr. 11. 49 ris κα τῶνδε θάλασσαν ἔχειν 7 κύμαθ᾽ ἕλοιτο ; As
regards the antithesis of πολλῶν ἐσθλῶν and τὴν ὄνησιν, it is well illustrated by a
prayer in Menander’s Kolax (fr. 292 Kock) which concludes: διδόναι σωτηρίαν,
ὑγίειαν, ἀγαθὰ πολλά, τῶν ὄντων Te νῦν ἀγαθῶν ὄνησιν πᾶσι" ταῦτ᾽ εὐχώμεθα. It
is natural to pray for ἀγαθὰ πολλά (cf. e.g. Antiphanes fr. 163. 6 K. τούτωι
πολλὰ κἀγάθ᾽ οἱ θεοὶ... δοῖεν), and so the phrase πόλλ᾽ ἀγαθὰ γένοιτό σοι could
become a mere formula of well-wishing or gratitude as, e.g., in Ar. Eccl. 1067.
But sensible people will pray not only for blessings to come, but also for
being allowed the enjoyment of the good that is at present theirs, τῶν ὄντων
τε νῦν ἀγαθῶν ὄνησιν. In other cases any desire to obtain further blessings may
be disclaimed and all that is asked of the gods may be the undisturbed use of
what is there already. So Admetus, protesting that he does not wish to have
children by a second wife, says (E. Alc. 334) ἅλις δὲ παίδων: τῶνδ᾽ ὄνησιν
εὔχομαι θεοῖς γενέσθαι. Similarly E. Hec. 996 f. σῶσόν νυν αὐτὸν μηδ᾽ ἔρα τῶν
πλησίον.---ἥκιστ᾽- ὀναίμην τοῦ παρόντος, ὦ γύναι. Clytemnestra, too, declares
that she prefers the enjoyment of what she now possesses to an abundance of
good things to come. To bring out this contrast, Hermann thought it neces-
sary to write τήνδ᾽ ὄνησιν, and many editors' have followed him. But in the
language of Aeschylus τήν seems quite sufficient for the purpose. The
‘article’ often (cf., e.g., 177, 637, 896) retains its stronger force as a demon-
strative pronoun, indicating that the thing to which it refers is not merely a
representative of its class but rather the one particular object with which the
speaker is directly concerned. The general expression πολλὰ ἐσθλά makes it
the more easy in the counterpart to grasp the force of the restrictive τήν,
‘this enjoyment here’, ‘the enjoyment related to the issue in hand’, ‘the
enjoyment of what I have now obtained’.
If this interpretation is correct, we are entitled to recognize in 350 a prayer-
like theme and see in it an adequate continuation of the preceding wish τὸ δ᾽
εὖ κρατοίη κτλ. It is probable that adherence to the prayer formula illustrated
above? has led Aeschylus to use here the word ὄνησις, which does not occur
elsewhere in his extant plays or fragments.
Wilamowitz’s translation, Griech. Tragödien, xii, and his note at the end of it, p. 113;
Housman, loc. cit.).
ı Wilamowitz, who in his edition accepted τήνδ᾽, afterwards changed his mind. The
reinterpretation of the passage which he published in Hermes, Ixii, 1927, 284, is too artificial
to be convincing and does not take into account the contrast between πολλὰ ἐσθλά and
ὄνησιν, which we have seen to be a recurrent idea.
2 The prayer which I have quoted from Menander's Kolax hasa close parallel in the prayer
mentioned in the funeral regulations of Gambreion (3rd cent. B.C.), Dittenberger, Syil.
1219. 20 ff. ἐπεύχεσθαι τοῖς ἐμμένουσιν... εὖ εἶναι καὶ τῶν ὑπαρχόντων ἀγαθῶν ὄνησιν. Cf. also
ibid. 997. 14 (Smyrna) βίου καὶ ἐργασίας καλῆς γένοιτο παρὰ τῆς θεοῦ ὄνησις, further the prayer
Demosth. 18. 141 ἐπεύχομαι πᾶσι τούτοις, εἰ μὲν ἀληθῆ πρὸς ὑμᾶς εἴποιμι... ., εὐτυχίαν μοι
δοῦναι καὶ σωτηρίαν, εἰ δὲ. . . αἰτίαν ἐπάγω τούτωι ψενδῆ, πάντων τῶν ἀγαθῶν ἀνόνητόν με
ποιῆσαι, and the adaptation of the prayer formula in Demosth. 28. 20 ἀντιβολῶ πρὸς παίδων,
πρὸς γυναικῶν, πρὸς τῶν ὄντων ἀγαθῶν ὑμῖν, οὕτως Övauode τούτων, μὴ περιίδητέ pe κτλ. Tam
inclined to find in E. Hel. 935 ὄντων ἐν οἴκοις χρημώτων ὀνήσομαι an echo of the prayer.
The thought that it might be better to pray for the enjoyment of what we possess than
179
line 350 COMMENTARY

The progress of thought in Clytemnestra’s concluding sentences is not

difficult to trace. ‘Troy has fallen, the ten-year struggle is crowned with
success; there are still perils threatening, but may the good prevail with no
uncertain victory ; for I do not pray for further blessings, but shall be content
to enjoy what has been obtained.’ This the queen speaks aloud, without
prejudice to that which she secretly intends and plans. Cf. on 500.
A different, though in my opinion far less suitable, interpretation of 350
must at least be mentioned, for it is grammatically correct and has been
adopted by many critics. According to this view πολλῶν ἐσθλῶν is objective
genitive, to be taken with ὄνησιν. The scholars who favour this interpretation
vary a good deal in their opinions as to the precise meaning of πολλῶν
ἐσθλῶν and the relation of the clause πολλῶν... εἱλόμην to the preceding
sentences (cf., e.g., the notes of Paley and Verrall).
351. κατ᾽ ἄνδρα σώφρονα: he pays her the compliment for which her pointed
remark at 348 had asked. Apparently the coryphaeus takes Clytemnestra’s
words in the wide sense in which they were meant (cf. on 348); by using the
epithet σώφρων he praises her sober judgement and her wisdom as well as her
discretion and piety.
εὐφρόνως : ‘with good sense’, ‘with understanding’, cf. on 806.
353. Paley’s conjecture αὖ, later given up by Paley himself, has often been
put in the text. Hermann argued against it, Headlam strongly supports it.
He says that εὖ is never found with verbs like προσειπεῖν or προσαυδᾶν. The
argument does not count for much, since if ed is correct here it is not used in
the sense of bene dicere or any meaning of the kind, that is to say, it indicates
not the content of προσειπεῖν but the manner in which it is carried out,
‘piously’, ‘in accordance with the rite’ (the opposite of such an invocation
is Catullus 4o. 3 deus . . . non bene advocatus). And what would αὖ refer to?
Surely not to the circumstance that 'the Chorus has already in the first ode
called upon the gods' (Schneidewin)? It hardly seems possible to relate it
to 317. If the coryphaeus says there 'I shall pray to the gods later', he can
hardly look back to that announcement with the words θεοὺς προσειπεῖν αὖ
παρασκευάζομαι. Headlam compares Ar. Birds 226 ovzroys μελωιδεῖν αὖ παρα-
σκευάζεται (he also quotes Thesm. 99, where however, αὖ is merely a dubious
conjecture), but there the situation is that the hoopoe, having paused a while
after his first solo aria, is about to begin on his second. In the same way
in Pers. 947 κλάγξω δ᾽ αὖ γόον ἀρίδακρυν the ad looks back to the πέμψω
for an increase of blessings was not, of course, confined to Greek prayers. Val. Max. 4. 1. 10
Africanus . . . censor, cum lustrum conderet inque solilaurilium sacrificio scriba . . . sollemne
ei precationis carmen praeirel, quo di immortales ut populi Romani res meliores amplioresque
facerent rogabantur, ‘satis’ inquit ‘bonae et magnae sunt; itaque precor ut eas perpetuo inco-
lumes servent’, ac protinus in publicis tabulis ad hunc modum carmen emendari tussit. Cf.
also Augustus’ own words (from the records of the Senate) in Suet. Aug. 58. 2 compos
factus volorum meorum, p.c., quid habeo aliud deos immortales precari quam ut hunc con-
sensum vestrum ad ultimum finem vitae mihi perferre liceat? Vor Horace this type of prayer
acquires special significance, in accordance with his ideal of deorum muneribus sapienter
uti, cf. Carm. 1. 31. 17 frui paratis . . . mihi, Latoe, dones (wrongly taken by G. Carlsson,
Philol. xc, 1935, 391; the pregnant sense of frui is demonstrated by the passage Epist. 1.
4. 7 quoted just below), cf. Sat. 2. 6. 4 f. nil amplius oro (cf. Carm. 2. 18. 11 nihil supra deos
lacesso), Maia nale, nisi ul propria haec mihi munera faxis, Epist. 1. 4. 7 di tibi divitias
dederunt artemque fruendi, 1. 18. 107 f. (prayer) sit mihi, quod nunc est, etiam minus, et
mihi vivam quod superest aevi, siquid superesse volunt di.
180
 COMMENTARY line 354

πολύδακρυν ἰαχάν of the strophe. Thus it is safer to leave προσειπεῖν εὖ as it

stands.
354. χάρις οὐκ ἄτιμος πόνων : ‘no inadequate return for all our trouble’, Paley,
on the whole correctly. χάρις is used on the one hand of the delightfulness,
the ‘favour’, in short everything through which a χαρίζεσθαι manifests itself
(cf. on 182) ; on the other hand of the effect of the favour on the receiver (S. 47.
522 χάρις χάριν γάρ ἐστιν ἡ τίκτουσ᾽ dei, E. Hel. 1234 χάρις γὰρ ἀντὶ χάριτος
ἐλθέτω). ᾿ χάρις is reciprocal in Greek; χάριν δίδωσι qui gratum facit, χάριν
λαμβάνει cur gratum. fit, χάριν ἀποδίδωσι qui gratiam. reddit’ (Wilamowitz,
Griech. Lesebuch, ii. 97). From the point of view of the effect χάρις τινός can
mean either ‘gratitude for a thing’ or ‘the favourable outcome of a thing;
the return, the reward for a thing’. So, e.g., E. Hec. 828 ff. ποῦ... τῶν ev
εὐνῆι φιλτάτων ἀσπασμάτων χάριν τιν᾽ ἕξει παῖς ἐμή, κείνης δ᾽ ἐγώ ; Suppl. 1135 f.
ποῦ δὲ πόνος ἐμῶν τέκνων, ποῦ λοχευμάτων (a certain emendation) χάρις; Eur. fr.
446 N. ὦ μάκαρ, οἵας ἔλαχες τιμάς, "Ϊππόλυθ᾽ ἥρως, διὰ σωφροσύνην, οὔποτε
θνητοῖς ἀρετῆς ἄλλη δύναμις μείζων" ἦλθε γὰρ 7) πρόσθ᾽ 7) μετόπισθεν τῆς εὐσεβίας
χάρις ἐσθλή. “Sooner or later εὐσέβεια gets its reward.’ This γνώμη, on which
is based the preceding maxim οὔποτε θνητοῖς ἀρετῆς ἄλλη δύναμις μείζων, 15
unmistakably parallel to the pronouncement on the special case: oias ἔλαχες
τιμάς, “Ιππόλυθ᾽ ἥρως, διὰ σωφροσύνην, with εὐσεβίας χάρις corresponding to
τιμὰς (ἔλαχες) διὰ σωφροσύνην. The particular nuance of χάρις discussed here
is indeed not identical with τιμή, but a closely related concept (cf. Plato,
Euthyphr. 15 a). It comes out clearly in Ag. 581 and also Ag. 354: ‘since a
return for the trouble has been effected, not without worth and value (for the
trouble).’ As regards the place of πόνων in the sentence I think Schneidewin
(‘der Lohn der Mühen ist als ein deren nicht unwerther geschafft’) and Verrall
(‘ πόνων depends directly upon χάρις, though relative in sense to ἄτιμος ’)
have taken it correctly. Wecklein connects πόνων only with ἄτιμος : ‘eine
Gnade ist erwiesen, die der (um sie aufgewandten) Mühen nicht unwert ist’.
In view of the instances quoted above of χάρις with the genitive this seems
to me unlikely, nor does his rendering of εἴργασται give it quite the right
sense,

The two speeches of Clytemnestra and the replies of the coryphaeus (281-
354) present a problem which must be discussed as a whole. When the queen
first announces the capture of Troy, the spokesman of the Elders, very
naturally, dares not trust his ears (268). To her repcated statement he
answers with an expression of boundless joy (270). But immediately after-
wards (272) he reveals serious doubts. The same mood prevails in the
subsequent utterances of the old man (274, 276): he makes no attempt to
conceal his scepticism and suggests the possibility of easy credulity in terms
which can hardly be called polite. No wonder Clytemnestra is offended. The
coryphaeus then moderates his tone, but his last question (280) is still full
of suspicion. It is pretty much as if he said: ‘It is impossible that any
messenger could arrive here so quickly’. Then Clytemnestra describes the
arrangement of the beacons. She ends her speech (315) with a pointed refer-
ence to the challenge thrown down (272 ἢ γάρ τι πιστόν ἐστι τῶνδέ σοι τέκμαρ ;):
τέκμαρ τοιοῦτον σύμβολόν τέ σοι λέγω κτλ. The coryphaeus in his reply (317 ff.)
does not take any notice of this last remark. He postpones the thanksgiving
181
 COMMENTARY
which in the circumstances might be expected of him, and asks for a more
complete report. It does not follow from his utterance that he is not con-
vinced by what he has so far heard.’ He would not be a Greek if he did not
enjoy listening as long as possible to a fine tale beautifully told. So we need
not even excuse the poet by saying that, for the benefit of the audience, he
wanted Clytemnestra to add a second piece of description in the grand style.
When the queen has finished her speech the coryphaeus, by using practically
the same phrase as in 272, makes it plain that his former doubts have been
overcome and he is now fully satisfied: ἐγὼ δ᾽ ἀκούσας πιστά σου τεκμήρια
θεοὺς προσειπεῖν εὖ παρασκευάζομαι. The sincerity of these words is confirmed
by the following anapaests and the stasimon to which they are the introduc-
tion : there the fall of Troy and the victorious end of the war are accepted as
a certain fact. Now this ode is sung after the queen has left ; so there can be
no ground for suspicion that the former words of the coryphaeus (352 ff.)
may have been influenced by some reluctance to speak freely in the presence
of Clytemnestra. Only in the epode of the stasimon (475 ff.) and the subse-
quent trimeters is the attitude of tne Chorus completely reversed. The
reason for the change will be discussed later on. At present we are concerned
with a difficulty in the preceding scene. There the optimistic confidence of the
Elders is represented as the result of a speech or a couple of speeches, large
sections of which, when judged by the standards of a modern reader, seem to
contain very little to justify such an effect. Before, however, tackling this
problem, it is important to remember that the poet clearly wishes the two
speeches of Clytemnestra to be taken as a unit or at any rate as parts of one
whole. They are designed by way, not of contrast, but of supplement to
each other. This appears from the fact that 317 ff. express a wish for continua-
tion rather than a protest or criticism, and still more from the manner in
which 352 (πιστά σου τεκμήρια) harks back both to the stichomythia (272) and
to Clytemnestra’s words at the end of her first speech (315). One would not
wish to stress a plural like τεκμήρια unduly, but here it seems obvious that
the coryphaeus by πιστὰ τεκμήρια refers both to the account of the beacons
and to the description of captured Troy.
We now turn to the peculiar feature which has shocked sagacious critics,
the seemingly odd discrepancy between the main contents of Clytemnestra’s
second speech (320-50) and the effect it produces in that it completely con-
vinces the coryphaeus, who styles it πιστὰ τεκμήρια. In the words of Verrall
(2nd ed. p. 207) ‘her [the queen’s] reflexions may or may not be very laudable
and wise, but what have they to do with the "evidence" of the victory?’
Wilamowitz uses even stronger language (Interpr. 167 f.): ‘this speech cannot
possibly be intended to allay the incredulity of the Chorus; for what does
Clytemnestra know of the things she is relating?’ . . . ‘it is really not suitable
and proper that Clytemnestra, the woman who has been sitting at home,
should describe the captured town, and that she, who is planning the murder,
1 Verrall (2nd ed. p. 206 f.) asserts that the coryphaeus at 317 ff. ‘treats the queen's
proffered "proof" of the Greek victory with a reserve which is barely saved from dis-
courtesy’, and that ‘his behaviour is in fact distinguished from the open incredulity of the
speakers at the close of the following ode (475) only by such a decent disguise as the queen's
presence necessarily commands’. This interpretation of 317 ff. may be conducive to
winning the reader's acceptance of Verrall’s second Chorus of ‘conspirators’ (see below),
but it is certainly not borne out by the words of the text.
182
 THE EFFECT OF CLYTEMNESTRA’S FIRST TWO SPEECHES
should utter a warning against the destruction of the temples’. The latter
charge is wholly unjustified: Clytemnestra, acting as vice-gerent of the king
and representing the dignity of the highest office, must be allowed to voice
religious and moral convictions suitable to the position she holds. In her
comments on the danger of sacrilege there is nothing that makes the assump-
tion of a hidden meaning necessary or even probable. We should therefore
be content to take them at their face value and not try to probe into Clytem-
nestra’s ‘real’ feelings.” As for the object of the dramatist, it is undeniable
that the section 338 ff. serves also as a preparation for the description of the
storm at 649 ff. (cf. the note on 344), but it would be quite wrong to assert
that for that purpose alone it has been dragged into the queen’s speech
without being an organic part of it.
The other objection of Wilamowitz seems to be better founded. Clytem-
nestra cannot be possessed of any evidence concerning the state of affairs in
Troy, which she so vividly describes. How, then, is the effect of her speech
on the Elders to be accounted for?
The problem would disappear altogether if Verrall’s expedient could be
taken seriously. He invented (cf. his Appendix I. J) a second Chorus of
accomplices of Clytemnestra or ‘conspirators’, to whom he assigned ll. 351-66
and a good many more lines in the following scenes. But this freakish idea
cannot help us.
Wilamowitz (Griech. Tragoedien, ii. 32 n. 1; Interpr. 167 f.) took a different
line. So far from eliminating the difficulty, which he viewed very much in the
same light as Verrall, he blamed the poet for not having done better, but
mitigated his censure by a slightly patronizing attitude ('etwas sehr naive
Technik’, ‘eine naive Dramaturgie’). The crucial point in his criticism (‘what
does Clytemnestra know of the things she is relating?' etc.) is based on a
consideration that is not likely to have occurred to the Elders of Argos or, for
that matter, to an Athenian audience. It is not very helpful to subject
Aeschylus to an inquisition of our own, and to divert our attention from
the way in which he has deliberately chosen to handle the subject.
The sceptical questions À γάρ τι πιστόν ἐστι τῶνδέ σοι τέκμαρ ; and καὶ τίς
τόδ᾽ ἐξίκοιτ᾽ ἂν ἀγγέλων τάχος ; have been thoroughly answered in Clytem-
nestra’s beacon speech. The leader of the Chorus, as we have seen, does not
doubt or minimize the weight of the evidence offered by the queen, but
merely desires a completion of her tale. At the end he declares himself
fully satisfied. Does Clytemnestra’s second speech then (320 ff.) add nothing
to the τέκμαρ contained in the first and are we to regard it as a mere ornament,
which, decorative though it be, does not serve any structural purpose? The

1 Schneidewin, whose approach to these utterances of Clytemnestra was very similar

to that of Wilamowitz, saw in them ‘bitter irony’.
2 It is amusing to observe how those who here plunge into psychological guesses reach
exactly opposite conclusions. On the one hand Keck (Comm. p. 270) : ‘Clytemnestra, whose
aim it is to destroy her husband, is full of anxiety in thinking out everything that might
bring harm too soon upon the army, and put Agamemnon beyond the reach of her
vengeance’, and G. Méautis, Eschyle, 152: ‘elle souhaite le retour d’Agamemnon, car elle
se sentirait frustrée du bonheur méme de sa vie si son mari périssait d’une autre main que
la sienne’. On the other hand Murray, Aeschylus, p. 212: ‘Obviously her fears are largely
her disguised wishes. If only Agamemnon could be destroyed before reaching home!’—
despite her burning thirst for direct and personal revenge and her jubilant triumph 1377-98.
183
 COMMENTARY
question must be answered in the affirmative so long as, following Verrall and
Wilamowitz, we insist on that conception of evidence which regulates the
procedure of a Bow Street magistrate. The scene of Aeschylus presents a
different picture. The Elders have learnt that the queen’s assertion is based
not on wishful thinking but on a solid first-hand message. They do not in the
least question the reliability of the beacon-sign (for their changed attitude at
475 ff. see below). Still, after so many years of frustrated hopes, they are not
at once capable of grasping the full meaning of what they have heard. They
know indeed that Troy has been taken, but the fact has not yet come home
to them as a reality. To be able to ‘realize’ what has happened, they need
some aid. This aid is given them in the queen’s second speech. She does not
know any more than anyone else what is actually going on in the conquered
city, but, from careful observation of human life, she has present in her
mind all the typical features of such a catastrophe. She works them out in
her powerful way, and carries with her the Elders, the Athenian audience,
and endless generations of latter-day readers—save a few professional critics.
She is not in a position to tell ra γενόμενα, but in describing οἷα av γένοιτο she
provides exactly what is wanted. Her suggestive imagination lends colour
to what before were bare outlines, and adds reality to the sound of words.
Under the spell of her speech the Elders feel that all she has said is true to life;
it cannot have been otherwise, it must in fact have happened as she says.

THE FIRST STASIMON (355-488)

A. The Anapaestic Prelude (355-66)
For the form of a stasimon in which the lyrical part is preceded by
anapaests see W. Kranz, De forma stasimt, 48 f. and Siasimon, 135, 166. This
form, like the corresponding one of the parodos (cf. on 40-103), is apparently
peculiar to early tragedy: the only instances are in the plays of Aeschylus,
cf. Suppl. 625, Pers. 532, 623, Sept. 822, Eum. 307. More than once a close
parallelism between the anapaests and the following lyrics can be observed
in that the former anticipate or outline some of the main contents of the
latter. In this respect Pers. 532 ff. and Eum. 307 ff. are particularly similar
to Ag. 355 ff.
B. The Lyrical Part (367-488)
Metre
This ode consists almost entirely of simple iambics of the type which we
have found in section III of the lyrical part of the parodos (192 ff.). So e.g. 367
U=- —v- v——is the sameas 197, 370v —v— -u- -—v-is the same
as 238, etc. In the first stanza the line (380) before the ephymnium, - vv —
v — 9, seems to form a transition to the pherecrateans of the ephymnium,
since that particular form of choriambic dimeter, - v v — v — o, Lydia dic per
omnis, may be regarded as equivalent to a pherecratean, cf. Wilamowitz,
Verskunst, 323 n. x and 396.
The last four lines of the third stanza (not counting the ephymnium),
448-51 = 467-70, are divided in F and Tr in the same manner in which they
are printed by Stanley, Porson, Hermann, Murray, and many others, save
that in F (but not in Tr) the first of these lines begins with the first syllable
of ἀλλοτρίας. This arrangement is undoubtedly attractive. If we accept it, the
184
 THE METRE OF THE FIRST STASIMON (367-488)
endings of the metrical cola coincide not only with the endings of words but
also, in the strophe completely, in the antistrophe partly, with the endings of
syntactical units. The metre, too, would be pleasant enough. τάδε σῖγά τις
Baile: and φθονερὸν δ᾽ ὑπ᾽ ἄλγος ἕρπει are ionici a minore of the anaclastic type
(φέρ᾽ ὕδωρ, φέρ᾽ οἶνον ὦ παῖ), to which the following colon προδίκοις Arpeidars
would form an equally normal clausula (σὺ δὲ σῦκά μ᾽ αἰτεῖς, cf. p. 50,
metrical note on 201-4).' As for the first of those four lines, ἀλλοτρίας διαὶ
γυναικός, it is a form of the hipponacteum, -vu-u-v--; exactly the
same form occurs e.g. Ag. 687 between a choriambic-iambic dimeter and
ionics, Pers. 648 after two choriambs and before ionici a minore, Pers. 659
ἔλθ᾽ ἐπ᾽ ἄκρον κόρυμβον ὄχθου followed by ionici a minore,? and perhaps also
Sept. 721 after a choriambic dimeter and before ionici a minore. At Ag. 743
we find the hipponacteum? δηξέθυμον ἔρωτος ἄνθος after a choriambic-iambic
dimeter and before ionici a minore, at E. Her. 677 the ‘normal’ hipponacteum
——-vu-v-- between a glyconic (choriambic dimeter in the antistr.) and
ionici a minore.* No exception should be taken to the hipponacteum Ag. 448
being attached to iambics. Apart from the instances where a hipponacteum
follows a choriambic-iambic dimeter (Ag. 687, 743), we find it attached to
iambics at Pers. 976, Cho. 382, and especially Ag. 1488 (7 iamb., hippon.),
and at E. Hec. 631 (5 iamb., hippon., choriamb. dim., cf. Wilamowitz,
Verskunst, 546). Finally, we should not fail to observe that with the end of
the hipponacteum both at 448 and at 467 we reach the conclusion of a unit
which not only syntactically but in its thought is distinctly separated from
what follows. This is very much in favour of admitting a hipponacteum here,
since that verse often serves as a concluding line.
This κωλίζειν of 448-51 is not, however, the only possibility. A different
one was introduced by Blomfield, whom Weil, Kirchhoff, Sidgwick, Wila-
mowitz, and others followed. They break the lines after yuvat-, βαῦ-, and
ép-, so that we obtain three acatalectic choriambic-iambic dimeters, followed
by the catalectic form (- v v -- uv--). I leave it to the reader to make his
choice. The argument that the second method yields a greater simplicity
and uniformity in the metre of the whole song does not seem to me very
strong when we bear in mind the character of the other stasima of this play
! I think it probable that we have to recognize the same clausula Prom. 134 θεμερῶπιν
αἰδῶ, after ionici a minore. But I cannot demonstrate in passing why I should prefer there
to follow the analysis of the excellent metrical scholion (130 Wecklein) and assume two
periods, each beginning with the colon Ὁ — v — — (μηδὲν φοβηθῆις and κτύπου yàp dx) and
continued in ionics. For the beginning of ionici a minore with Ὁ — v — — cf., besides Anacreon
fr. 49 D., which the scholiast quotes, S. El. 1058. The same colon occurs Prom. 398 between
the initial choriambic dimeter and a period of ionici a minore (or, perhaps, v — v — - forms
the beginning of the stanza, which then would go on with ionics).
2 It will be clear that (and why) I disagree with Wilamowitz here.
3 It seems to be by an oversight that Wilamowitz, Aeschylus, praef. xxxiv and Verskunst,
249, produces the pattern o--uu—u-u, but not -v-vu-v-5, From Verskunst,
249, it appears that he regarded Ag. 743 as a hipponacteum; to this we should add his
commentary on E. Jon 113, where the ‘normal’ (cf. Hephaestion, p. 32. 20 Consbr.) form
——-—vv-u-- in the strophe corresponds to the -v-vv—~—-— of the antistrophe.
The variation at the beginning (even v v v — v v — v — —-, Bacchyl. 4. 2) is the same as in
the kindred glyconics etc.
* Icannot admit that Wilamowitz's later arrangement of the passage, Griech. Tragoedien,
i. 368 and Verskunst, 555, is an improvement on his commentary. On Jon 113 he rightly
says: ‘aber ich beachte lieber die Wortabteilung' (for the principle cf. Versk. 548).
185
line 355 COMMENTARY

and especially of 681-716 and 737-63. Moreover, there is ἃ rather serious

objection to the second method. If we accepted it, we should be left with a
group of τὸ metra without a pause between them (from 447 τὸν δ᾽ ἐν φοναῖς
to 451 Ἀτρείδαις). This would be a violation of the rule discovered by Paul
Maas (Griechische Metrik, ὃ 65, cf. O. Schroeder, Grundriss der griech. Vers-
geschichte, ız) that tlıe maximum size of such uninterrupted groups is 8 metra
unless there is a very special reason for an extension beyond that length.
At the end of each stanza we find the same ephymnium : two pherecrateans
followed by ἃ priapeum. This ephymnium has been compared by Wila-
mowitz (comment. on E. Her., survey of the stasimon 348 ff., vol. 112, p. 80 f.)
with two very similar ones, A. Suppl. 640 ff. and E. Her. 359 ff. The stasimon
in the Supplices consists of four couples of stanzas (no epode) ; the first three
end with 2 pherecr.+priap., exactly like the present ode of the Ag. The
three couples of stanzas E. Her. 348 ff. (no epode) differ only in that there are
at the end of the first two 3 pherecr.+priap., and at the end of the third
4 pherecr.+priap. There is always strong punctuation before the beginning
of the ephymnium so that it is separated from the bulk of the stanza by
syntactical structure as well as by metre. But the words are never repeated,
not even in strophe and antistrophe. Wilamowitz termed this form ‘ephym-
nium rhythmicum’. He thought that it originated in early sacred songs. We
find in stasima of Aeschylus very similar types of ‘ephymnium rhythmicum’
several times employed on a smaller scale: Sept. 295 ff. (cf. Wilamowitz,
Verskunst, 398), after an iambic period, 6 pherecr. +pentameter choriambicus ;
Cho. 609 ff., after trochees, 2 priap. (i.e. only the first line, glyc. instead of
pherecr., differs from the one in the first stasimon of the Ag., but in this
instance the sentence overlaps from the preceding trochees) ; Ag. 696 ff., after
ionics, priap. (- vu — v—— being equivalent to pherecr.) +pherecr. ;' and
perhaps Ag. 725 f., after lecythia, priap. (syntactically separated from the
preceding part).
The epode (475 ff.) consists of simple iambics, uninterrupted by any catalexis;
hence Wilamowitz, Interpr. 168, styled it πνῖγος.
355. ὦ Ζεῦ βασιλεῦ : for the history and frequency of the designation Ζεὺς
βασιλεύς cf. Wackernagel, Unters. z. Homer, 210, Wilamowitz, Glaube der
Hell. i. 140, and Ziegler in Roscher’s Lex. d. Mythol. vi. 608 f.; for the rich
epigraphical evidence cf. the indexes of each volume of A. B. Cook, Zeus
(Index I, s.v. ‘Zeus’, Epithets). ‘In Homer βασιλεύς and βασίλεια are never
used of gods and goddesses, either attributively or predicatively . . . Ζεὺς
βασιλεύς (with or without a genitive) is peculiar to Hesiod, the Epic cycle
(Thebais fr. 3. 3), the Hymns (h. Cer. 358), lyric poctry in general, and Attic
poets from Solon, fr. 31 (28 D.).1, onwards’? (Wackernagel). Aeschylus has
the vocative in one other passage, also as the beginning of choral anapaests,
Pers. 532. It seems probable that he has borrowed this form of invocation
from Attic everyday speech, in which it seems to have been used under
stress of marked excitement. The six passages in Aristophanes where ὦ Ζεῦ
βασιλεῦ occurs belong to a strictly limited range of expressions: the vocative
is found either with exclamatory sentences introduced by ὅσον (Clouds 2,
1 Cf. Aesch. fr. 355 N.: priap. (the first two syllables missing)+ pherecr. We do not know
what preceded, nor whether the preserved lines served as *ephymnium rhythmicum'.
2 “From Solon . . . onwards’ should be cancelled, cf. below, p. 348 n. 1.
186
 COMMENTARY line 356

Frogs 1278), οἷον (Wasps 625), or ws (Plut. 1095), or with an exclamatory

genitive (Clouds 153, Birds 223); the emotions conveyed by the expression—
astonishment, pain, fear—are in every case particularly strong. Its use by
way of supplement to definitely colloquial constructions (τὸ χρῆμα τῶν
κόπων ὅσον Or τοῦ φθέγματος τοὐρνιθίου) makes it unlikely that Aristophanes
in this exclamation is following the pattern of more exalted poetry. It is
much more likely that Aeschylus has adopted an everyday phrase in the two
passages which bear so close a resemblance to each other: in the one the
Persian elders give excited expression to the news of defeat, in the other
the elders of Argos acclaim the news of victory. Here ὦ Ζεῦ also prepares the
hearer for 4ia at the beginning of the second period of the anapaests, and for
Διός at the beginning of the lyrical part (367).
Νύξ: for the matter cf. on 826. νὺξ φιλία is here personified, as the following
words show still more clearly: an exact parallel is the invocation ὦ Νὺξ ἱερά
(cf. Wilamowitz, Glaube der Hell. i. 22) with which the Andromeda of Euripides
(fr. 114 N.) begins the anapaestic prologue’ of the play.
356. «reareıpa: found only here. Stanley translated it domina, and his re-
marks on Ag. 6 indicate that he took κόσμων κτεάτειρα to refer to the star-
spangled robe of Night. Schütz objected to this: κτεάτειρα non domina est,
sed quae paravit seu conciliavit μεγάλους κόσμους h.e. victoriae gloriam,
praedae magnitudinem’, etc. Long before Schütz Casaubon had glossed
κτεάτειρα by ‘acquisitrix’. This is in keeping with the explanation given by
Triclinius: ἥτις τὰ κτέανα καὶ τὰ κτήματα καὶ τοὺς μεγάλους ἡμῖν παρέδωκας
κόσμους τῶν Τρώων. The interpretation of Schütz was followed by Humboldt
(‘des unendlichen Glanzes Erkämpfrin’) and Blomfield (‘ κτεάτειρα in posses-
sionem tradens') and afterwards by most commentators, as also by Ernst
Fraenkel, Nomina agentis, ii. 29, while others, e.g. Plüss, Wilamowitz (‘im
Glanz der Sternenkrone’), Headlam, Kranz, Philol. xcili, 1938, 432, have
reverted to Stanley’s interpretation. In the dictionaries we find Passow and
Dindorf explain the word as ‘paratrix’: similarly the older editions of L-S
interpret it as ‘winner’, while in the ninth edition we find only ‘possessor’
(see also under κόσμος IV). If for the moment we leave aside the linguistic
point, it should be recognized that Triclinius’ and Schütz’s interpretation is
by far the better in itself. It gives to the line, instead of a vague lyricism, an
idea which is expressed with Acschylean terseness, and leads as forcibly as
possible to the very heart of the situation. It is of no use to compare Prom.
24. There the sufferings of the hero are brought out in the thought that after
the parching heat of the daytime he will welcome the approach of the
sublime deliverer, the ποικιλείμων νύξ, who in her turn is to cause fresh tor-
tures for him before long. But are the night’s starry mantle and crown any
concern of the Elders, who are engrossed in the fall of Troy? The direction
1 So I maintain and with good reason. According to the express testimony of a man who
had the play in front of him, the anapaests are τοῦ προλόγου τῆς ᾿Ανδρομέδας εἰσβολή. The
weight of this evidence has been ably pointed out by Welcker, Griech. Tragoedien, ii. 647.
On the other hand, Carl Robert, Arch. Zeitg. xxxvi, 1878, 18, has followed Hartung in
trying to depreciate its value; and v. Arnim, Wilamowitz, Pohlenz, and others have since
taken the same view. I hope on another occasion to go into this important point and also
to demonstrate that in the Iphigenia in Aulis it is only the fine anapaestic prologue which
is the genuine work of Euripides, not the dull trimeters ἐγένοντο Andaı Θεστιάδι τρεῖς
παρθένοι κτλ,
187
line 356 COMMENTARY

in which the meaning of μεγάλων κόσμων κτεάτειρα must be looked for is

evident from the fact that this attribute stands between two others which
are also applied to night. Νὺξ φιλία, “friendly, kindly Night’, implies helpful
service rendered by the night, and the relative clause 77’ ἐπὶ Τροίας πύργοις
«TA, indicates more specifically what service the helpful night has rendered.
Between these two expressions connected in thought stand the words
μεγάλων κόσμων κτεάτειρα: we are therefore entitled to expect a parallel
thought here as well.
It was necessary to start from the function of the expression as a whole,
for the sense of κόσμων is ambiguous and the word κτεάτειρα is completely
isolated. An attempt at a linguistic explanation of κτεάτειρα was made by
Ernst Fraenkel, Nomina agentis ii. 29. According to him the word might
have been coined by Aeschylus by way of ‘deliberate archaism’, though the
possibility cannot be excluded that he found it in late epic poetry. We are
chiefly concerned here with the question of the kind of words with which the
linguistic instinct of the poct and his audience was likely to associate «redreipa.
κτήτωρ may be set aside, for the word is not found before Hellenistic times,
and, besides, a word actually containing «rea- would be more probable. This
could be found in the Homeric κτεάτεσσιν and cognate words. But since
κτεάτειρα (like σώτειρα etc.) clearly appears to be an agent-noun, it is highly
probable that an association was felt with a verb (not necessarily a derivation
from it). The only verb available, so far as we can judge from what survives,
is κτεατίζω. And in point of fact a passage like I7 57 δουρὶ δ᾽ ἐμῶι κτεάτισσα
(sc. κούρην), πόλιν ἐυτείχεα πέρσας, should suffice fully to support the sense of
κτεάτειρα Which is assumed here.
It remains to examine κόσμων. It is obvious that the stars could be spoken
of as the ornaments of the night, but if Headlam states that later Pythagor-
eans regarded the stars as κόσμοι, such erudition is inapposite here. The
meaning which is assumed for κόσμοι in Schütz's paraphrase (see above) is at
least as probable as that which Headlam and others preferred. It may
suffice to glance at Pindar's use of the word, e.g. Οἱ. 8. 82 ἐνέποι κεν Καλλιμάχωι
λιπαρὸν κόσμον ᾿Ολυμπίαι (schol. τὸν κόσμον τὸν ἐν ᾿Ολυμπίαι, τούτεστι τὴν
νίκην. ἡ γὰρ ἑνὸς νίκη πᾶν τὸ γένος κοσμεῖ). Ihe plural κόσμων presumably
indicates that the praise of Night is expressed in general terms, though it is
the particular success which is especially thought of among the various
κόσμοι.
357. πύργοις: cf. on 128 f.
358. ereyavóv: a ‘covering’ net. It can hardly be assumed that we have
here an instance of the special meaning of 'close-woven ; proof (water-proof
etc.)’ of which there is evidence from the fourth century onwards (Kennedy
translates here: ‘a net so closely meshed’, similarly Verrall ; Headlam: ‘a net
of proof’). It is more likely that the poet was guided to his choice of the
adjective by the same idea as that which makes him use στέγαστρον ἀνδρός in
Cho. 984 for the death-robe in which Agamemnon was entangled, and which
is so often called a ‘net’ in the Trilogy. Wilamowitz renders oreyaorpov by
‘Hiille’. If this is correct, one might perhaps think of it in this passage too as
an all-round enveloping and enclosing net. However, it is safer to assume
that oreyavdv means nothing more than covering from on top. In every
instance where the adjective occurs in the fifth century it carries the notion,
188
 COMMENTARY lines 358 f.

roughly, of a horizontal roof-like or lid-like covering. So in Eur. Cretans,

fr. 472. 4 ff. (cf. Wilamowitz’s restoration and explanation, Berliner Klassi-
hertexte, v. 2.77 n. 1) ναοὺς. . ., which the timber (cut into rafters) areyavods
παρέχει (cf. Fr. trag. adesp. 115. 2 οἶκον στεγανόν), Thuc. 3. 21. 4 ‘at night, for a
shelter against rain, ἐκ τῶν πύργων, ὄντων... ἄνωθεν στεγανῶν, τὴν φυλακὴν
ἐποιοῦντο᾽. Sophocles too would certainly not have used the word στεγανός
absolutely to describe the shelter afforded by a shield, but in Ant. 113 f. he
says αἰετὸς eis γᾶν ws ὑπερέπτα, λευκῆς χιόνος πτέρυγι oreyavos (here used
passively), because the picture is kept up of the broad and level spread of the
eagle’s wings. In Ag. 357 f. the phrase ἐπὶ Τροίας πύργοις ἔβαλες indicates
quite plainly a covering put on from above, not just an encirclement of the
walls. It follows that the imagery—and this only, not anything correspond-
ing to it in reality—is rightly illustrated by Paley with his reference to the
comic incident in Aristophanes’ Wasps, where a net is spread from above
over the house of Bdelykleon.’ This snaring is very appropriate as an act of
Nv£, which always comes οὐρανόθεν : Night covers the city from above, but
not with the kindly veil which brings rest every night to mortals, but with the
net of doom. So in the clause ending with öixrvov we have a sublime vision of
superhuman range, unconnected with any probability in ordinary life (cf.
note on 365 ὑπὲρ ἄστρων). In the following clause ὡς μήτε κτλ. a single detail
is worked out of the idea of the net, bringing us closer to a familiar process
either of fishing or of hunting. Compare too the description given by Hero-
dotus 6. 31 of caynvevew, the process adopted by the Persians with the in-
habitants of the islands they conquered.
358 f. μήτε μέγαν μήτε νεαρῶν τινα: for this very common form of expression,
which ‘exhibits the comprehensiveness of a general notion by propounding
it in disjunctive form’ (A. C. Pearson on E. Hel. 1137), cf. Wilamowitz on
E. Her. 1106 ; Vahlen, Opusc. ii. 279 ff.; Leo, Analecta Plautina ii (Göttingen
1898), 36 ff.
οὖν: for οὖν added to either member in οὔτε... οὔτε (μήτε. . . μήτε) cf.
Denniston, Particles, 419 f., ‘it is found with equal frequency in the first, and
in the second, limb’. ᾿οὔτ᾽ οὖν is simply a more sonorous οὔτε᾽ (Wilamowitz,
Berl. Klasstkertexte, v. 2. 65).
veap@v: obviously the primary reference is to the young Trojans (328
παῖδες), but in relation to the metaphor of the net there is also the suggestion
of young wild animals; this is a proper, but by no means exclusive, use of the
word: [Xen.] Cyn. 9. 1o (in a description of fawn-hunting) τὰ yap σώματα
αὐτῶν διὰ τὸ ἔτι νεαρὰ εἶναι τῶι πόνωι οὐ δύναται ἀντέχειν (ibid. 9. 8 of μὲν νέοι
τῶννεβρῶν..... οἱ δὲ ἤδη μεγάλοι... with the same contrast as here, as already
observed by Abresch).
ὑπερτελέσαι: found only here; presumably it means (cf. note on 286) ‘to
rise above, to get over’ something. The underlying notion of the ὑπερ- is
explained by Klausen, probably correctly, with the help of 1375 f. πημονῆς
ἀρκύστατ᾽ ἂν φράξειεν ὕψος κρεῖσσον ἐκπηδήματος and of Pers. 98 ff. παράγει
βροτὸν εἰς ἄρκνας ἄτα (a certain emendation), τόθεν οὐκ ἔστιν ὑπὲρ (Hermann
1 * A net is stretched over the house and the outbuildings (164, 208, 368), which, however,
never prevents passage through the front door, so it does not reach so far downwards,
though it does obstruct the window.’ Wilamowitz, Berl. Sitagsb. 1911, 473 (= KI. Schr. i.
300).
189
lines 358 f. COMMENTARY

and Wilamowitz should not have substituted ὑπὲκ for this!) θνατὸν ἀλύξαντα
φυγεῖν, and Paley explains similarly :‘the notion is that of young and nimble
creatures leaping over an enclosure, ἀρκύστατον, covered by a net' (the last
phrase should be 'enclosed by a net'. Paley, mistaking the fanciful character
previously ascribed to the action of Night, has attributed to the process of
netting in actual hunting a feature which is alien to it). This brings us at
once to the question as to what kind of net and what kind of creatures were
in the poet's mind. In asking this question we hope not to over-scrutinize
the text: however, considering the manner of Aeschylus in dealing with such
matters, and the expert knowledge and appreciativeness of his audience, one
feels inclined to believe that he had a definite picture before his eyes, either
of game or of fish, not of both together, nor of anything indefinite and
general. Schütz decides in favour of fish, Schneidewin, and, following him,
Paley, in favour of game: most commentators say nothing. γάγγαμον 361
(see note) seems to indicate fish, as the word is only used of a fishing-net, and
γαγγάμη and γαγγαμεύς are used correspondingly. Pollux 1o. 132 quotes this
passage from the Agamemnon under the heading of τὰ ἁλιέως σκεύη. However,
all the examples of γάγγαμον etc. occur many centuries later than Aeschylus,
and it is possible that the word had a more general meaning in his time.
That the passage Ag. 358 ff. as well as 1375 f. refers to game, which does not
succeed, like the νεβρός Eum. 111 f., in escaping ἐκ μέσων ἀρκυστάτων, might
be vouched for by the $rep-, especially if we compare [Xen.] Cyn. 6. 8, where
these words are used of the setting of the hunting-net : στοιχιζέτω δὲ μακρά,
ὑψηλά, ὅπως dv μὴ ὑπερπηδᾶι. Besides, the distinction made in μήτε μέγαν
μήτε νεαρῶν τινα seems more appropriate if we are thinking not of fish but of
game. See above on γεαρῶν. But I cannot arrive at a definite conclusion.
We can point to the passage in Homer which toa considerable extent helped
to suggest the imagery of these Aeschylean lines (Stanley quotes it for one of
its details). In E 487 ff. Sarpedon says to Hector: μή πως, ὡς ἁψῖσι λίνον
dAóvre maváypov, ἀνδράσι δυσμενέεσσιν ἕλωρ καὶ κύρμα γένησθε" οἱ δὲ τάχ᾽
ἐκπέρσουσ᾽ ἐὺ ναιομένην πόλιν ὑμήν. There can be no doubt that Aeschylus
started from Homer’s word wavdypov and changed it to παναλώτον (see
p. 191); for the slight alteration (ravaAwrov epithet to ἄτης, not to the net)
compare note on 653. But more important than this detail is the fact that
Aeschylus has adopted the Homeric idea as a whole, and brought the simile
and the third line (οὗ δὲ τάχ᾽ ἐκπέρσουσι) into a more intimate connexion.
What in a particular emergency has been feared by the ally of the Trojans
as a possible fate is now turned into a terrible reality by the sentence of
Zeus. But we do not know what kind of net Homer had in mind, nor how
Aeschylus interpreted the passage. In later antiquity E 487 f. was usually
taken to refer to fishing: in addition to the BT Scholia and Eustathius cf.
Plutarch, De soll. anim. 26 (Mor. 977 f), Athen. 1. 25 b, Tryphiod. 674 f.
1 For this they refer to the paraphrase in the ®-scholia διὸ οὐκ ἔστιν ἄνθρωπον ὑπεκδρα-
póvra τὴν ἄτην φυγεῖν. But then they should also have mentioned that in the scholia of the
same MSS on 94 ff. we find, with reference to 95 f. (p. 42 Daehnhardt) τίς οὖν ὁ ἐν ταχυτάτωι
ποδὶ ἀνάσσων kai κρατῶν τοῦ εὐπετέος xai συντόμου πηδήματος τῆς ἄτης, ἤτοι ὑπερπηδήσας
αὐτῆς τὰ θήρατρα καὶ ἐκφυγὼν ταχέως; where the last clause, as is clear from the words
θήρατρα (= dpxvas) and ἐκφυγών, is made up out of 98 ff., and therefore the use of ὑπερπηδήσας
is not in favour of the assumption that the commentator in 99 had anything else than
ὑπέρ in front of him.
190
 COMMENTARY line 365
Athenaeus and Eustathius show that it was well known how seldom fishing _
with nets is found in Homer. Leaf adduces as tlıe only other instance the
metaphor in x 384 ff."
361. yayyapov: for the dredge-net so named, and employed in fishing for
sponges, oysters, and sea-urchins, see A. W. Mair, Oppian (Loeb Library,
London 1928), Introd. xliii: he remarks there that the word is still used on
the Black Sea coast; cf. Dindorf in the Thesaurus for the modern Greek
use of γαγγάμη.
πανάλωτος active: cf. on 1. 12. The (un-Homeric) type of a verbal adjective
with the prefix wav- is a great favourite with Aeschylus, see on 960.
362 ff. Cf. (Paley) the threatening speech of Menelaus to the Trojans Ν 623 ff.
οὐδέ τι θυμῶι Ζηνὸς ἐριβρεμέτεω χαλεπὴν ἐδδείσατε μῆνιν ξεινίου, ds τέ mor
ὕμμι διαφθέρσει πόλιν αἰπήν.
362. Δία ξένιον : the name forms a unit to which μέγαν is the attribute.
363. Rightly punctuated by Hermann (cf. Humboldt’s translation) and by
Blomfield with the note: 'Vulgo interpungitur τὸν τάδε πράξαντ᾽ em’ AdeEdv-
δρωι, quod vix Graece dici arbitror. For the meaning of πράξαντα cf. 369,
for the construction of the following words A 370 (Ἀλέξανδρος) Τυδεΐδηι ἔπι
τόξα riraivero and elsewhere.
364. τείνοντα πάλαι: ‘In the Attic writers words like πάλαι, ἄρτι, Or πρόσθεν
are found with the present when there is reference to the past’ (Wackernagel,
Syntax, i. 158) ; cf. above on 170. We may be allowed to work out the allusive
image and say that Zeus stretched his bow immediately or soon after the
crime of Paris, and then waited. The passage is an unmistakable echo of
60 ff. Arpews παῖδας ὁ κρείσσων En’ Ἀλεξάνδρωι πέμπει ξένιος Ζεύς : the stretch-
ing of the bow coincides with the sending out of the avenging expedition.
364 ff. ὅπως ἂν... σκήψειεν. Hermann observes: ‘Ut rarius ita loquantur
Attici, recte tamen optativo iunguntur hae particulae: v. de particula ἄν
iii. 4 [Opusc. iv. 151 sqq.]. Nec si quis ὅπως ἄν hic eo modo quo interpretari
malit, quicquam obstat.’ Accordingly Headlam attempts to combine the two
syntactical aspects (‘that so his bolt should neither fall short of the mark ...’),
while Wecklein (‘ ὅπως, wie’) and Wilamowitz in his apparatus (‘ ὅπως dv
σκήψειεν minime finale est (nec poterat ea vi dici) sed modale’), insist on the
modal function. For ὅπως dv with the optative cf. K. W. Kriiger on Thuc. 7.
65, Goodwin, §§ 349 ff., Kühner-Gerth, ii. 386.
365. πρὸ καιροῦ : in local sense ‘short of the mark’. On καιρός in early Greek
generally and particularly in such phrases as this see Wilamowitz, Hermes,
xv, 1880, 507 ff. = Kl. Schr. i. 43 ff. (cf. also his summary in Sappho und
Simonides, 247 n. 1).
ὑπὲρ ἄστρων: there are no grounds for changing this; Hermann in spite of
his strong admiration for Auratus (on Ag. 1396 Herm.) has rightly declined
to accept his conjecture here. Stanley refers to the proverbial expression ets
οὐρανὸν rokevers (ἐπὶ τῶν διακενῆς κατ᾽ αὐθάδειάν τι ποιούντων). This would
mean bringing Zeus down to the level of an earthly marksman. ΤῸ appreciate
the particular beauty of the words we have to realize that the viewpoint is
™ Hermann Frankel, Die homer. Gleichnisse, 62, without engaging in a discussion of the
point, sees in E 487 f. ‘a battue with nets’. That would be more appropriate, certainly, in
the context in which it stands. How far this interpretation is supported by the Vaphio
gold cup and other evidence, I cannot judge.
I9I
line 365 COMMENTARY
taken from the world of the gods: Zeus is sitting on a high mountain or in
heaven. ‘If the shot of Zeus hits above the stars it is the same as if a man’s
shot goes into the air or into the ground : sublimely ironical’ (Plüss). We find
a bold conception of a similar type in the fragment (87 Schr.) of a Pindaric
hymn (fully illustrated by Wilamowitz, Sappho und Simonides, 131 and
Reden und Vorträge, i, 4th ed., 127). There the gods in heaven are represented
as looking down at the earth: it lies before their eyes as a blue expanse (as
their heaven is seen by our eyes), and in this blue the little island of Delos,
Asteria, appears as a star.
366. ἠλίθιον : a brief, and partly hypothetical, history of the word is given by
Wilamowitz, Sappho und Simonides, 176 τι. 1; to this we should add the
passages of the lexicographers collected by Drachmann (ii. 66. 9) on Schol.
Pind. P. 3. 11 (20a). The gloss found in the Pindar scholiast, ἄπρακτος, suits
the Agamemnon passage too. That Aeschylus borrowed the word from
Sicily (cf. on 1507), as Wilamowitz supposes, is in itself improbable, consider-
ing that it occurs in choral lyric and in Ionic. Moreover, the supposition is
refuted by the occurrence of the word in a sense agreeing with Ag. 366 (of
something unperformed, fruitless, unsuccessful, missing its purpose) in an
inscription (IG 13. 975) on a marble sepulchral discus from Attica, about the
end of the sixth century, figured in Marshall, Journ. Hell. Stud. xxix, 1909,
153, and better in P. Jacobsthal, ‘Diskoi’, 93. Berliner Winckelmannspro-
gramm, 1933, 26 (cf. ibid. 28 for the date): Γνάθονος τόδε σέμα θέτο δ᾽ αὐτὸν
ἀδελφὲ heAÜwov νοσελεύσασα. Whether the writer took the word ἠλίθιος from
the language of poetry or of everyday speech one cannot tell: νοσηλεύειν
seems to be normally a prose word (but in Sophocles’ Euryfylos, fr. 215 P.,
it is restored with probability). The verb ἠλιθιοῦν occurs Prom. 1061.
The paroemiac ending in a molossus is uncommon ; G. Hermann, Elem.
doctr. metr. 378, instances also A. Suppl. 8 and Pers. 32 (proper name), cf. also
Wilamowitz on Sept. 826. We must also include Pers. 152 βασίλεια δ᾽ ἐμή,
προσπίτνω, Since it cannot be doubted that the line has been correctly trans-
mitted so in the majority of MSS, and that in M a gloss has found its way in.
367. Blomfield (followed by, e.g., Headlam and A. Y. Campbell) puts a stop
after ἔχουσιν, 'dictionis numerorumque concinnitate contempta' (Hermann).
In the first four lines of this choral ode the syntactical construction shows the
closest correspondence between strophe and antistrophe: not only does each
line consist of a full colon (370 too; for dependent infinitive constructions as
separate syntactical cola cf. Ed. Fr., Kolon und Satz, ii. 326 ff.) but also the
structure of the phrases is exactly parallel in detail. In each case a full-stop
comes at the end of the second line, then another at the same point in the
third line (ἔπραξεν ὡς ἔκρανεν — ἄκος δὲ πᾶν μάταιον), the remainder of which
in both cases consists of οὐκ followed by a trisyllable. Further instances of
words of equal length are εἰπεῖν = πειθώ, πάρεστιν = προβούλου, ὡς ἔκρανεν —
πᾶν μάταιον, ἀξιοῦσθαι μέλειν = αἰνολαμπὲς σίνος. The use of ἔχω with in-
finitive = potestatem habeo, scio (Dindorf) is common in Aeschylus as in
Sophocles.
Διὸς πλαγάν: Kranz compares this (on Heraclitus B 11) with Heraclitus’
phrase πᾶν ἑρπετὸν πληγῆι νέμεται and (as van Heusde before him) with 5, Aj.
137 πληγὴ Διὸς (cf. ibid. 278 f. μὴ '« θεοῦ πληγή τις Tien).
368. τοῦτό γε: this act at least of Zeus’ coercion we can trace, although many
192
 COMMENTARY line 369
of his paths run in darkness undiscoverable (Suppl. 93 ff. SavAoi γὰρ πραπίδων
δάσκιοΐ τε τείνουσιν πόροι κατιδεῖν ἀφραστοι).
369. The first ὡς is destructive both of sense and metre: perhaps it found its
way in because someone had in mind the pattern ὡς ἴδον, ὡς ἐμάνην. Hermann
rightly bracketed the word ; only he should not have read ἔπραξαν. This still
finds supporters, though ' it is hardly conceivable that πρᾶξαι used of Zeus
just before in 1. 363 should here refer to something quite different’ (Ahrens,
512). ‘He has fulfilled it in action as he duly ordained it.” ἔπραξεν (for its
meaning in 363 and here cf. on 1380) finds its adverbial complement in ὡς
ἔκρανεν as in 1380 οὕτω δ᾽ ἔπραξα. The verb πράσσειν is used absolutely in
1669 and Suppl. 515 (contrasted with λέγειν as here with «paivew).
The meaning of «paivew here needs more detailed discussion. «paivew in
the sense of ‘rule’ (for the origin of this meaning cf. Wackernagel, Unters. 2.
Homer, 157) is indeed found in Sophocles, but not in what is preserved of
Aeschylus. But when Dindorf, Lex. Aesch., lists all passages under ferficto,
this is not justified by the evidence. Passow more correctly renders several
passages by 'beschliessen, bestimmen’; in L-S, too, a special meaning is
noted ‘to ordain” and illustrated by A. Ag. 369, E. Suppl. 139, El. 1248
(τἀντεῦθεν δὲ χρὴ πράσσειν ἃ Μοῖρα Ζεύς τ᾽ Expave σοῦ πέρι, which in its co-
ordination of ideas closely resembles Ag. 369). But this usage extends con-
siderably farther. For the simple verb «paivew in this sense we may add the
following examples from Aeschylus: Ag. 1424 (see note), Suppl. 92 κορυφᾶι
Διὸς ei κρανθῆι (‘im Haupte des Zeus beschlossen ist’, Wilamowitz, Interpr. 31,
Headlam with greater exactness ‘if the fulfilment of it be ordained’), 608
τόνδε κραινόντων λόγον, 622 (as restored with certainty in accordance with the
scholion) χερσὶν Ἀργεῖος λεὼς ἔκραν᾽ ἄνευ κλητῆρος ws εἶναι rade, where χερσὶν
éxpave amounts to an intensified ἐχειροτόνησε, 943 τοιάδε δημόπρακτος ἐκ
πόλεως μία ψῆφος κέκρανται. From Euripides we may add: Andr. 1272 πᾶσιν
γὰρ ἀνθρώποισιν Ade πρὸς θεῶν ψῆφος κέκρανται κατθανεῖν τ᾽ ὀφείλεται, Tro. 785
ὅθι σοι πνεῦμα μεθεῖναι ψῆφος ἐκράνθη, Ion 464 Φοιβήιος ἔνθα γᾶς μεσσόμφαλος
ἑστία... μαντεύματα Kpaiver, 570 ἐς μὲν σὴν ἀνεύρεσιν θεὸς ὀρθῶς ἔκρανε. In
Pindar and Sophocles, though xpaivew is common, it is not used with this
meaning. In all the passages here cited the particular shade of meaning is
quite clear: ‘to pronounce and establish in binding and valid form with the
guarantee of fulfilment in the future’. The object of κραίνειν so used is, in the
case of human beings, a resolution (e.g. of the assembly), a sentence, and
the like; in the case of gods, a decree of fate, an assurance of a future event.
The common element in the various applications of xpaivew can easily be re-
cognized ;? ‘to make valid, effective’ is an approximate rendering of what the
verb signifies in ψῆφος κέκρανται and the like, as well as in sentences like τόδε
μοι κρήηνον ἐέλδωρ. However, if we want to arrive at a precise interpretation
of passages that to some extent have been insufficiently appreciated, a render-
ing that is rather too narrow would be preferable to one that is vague. It
may therefore be permissible to distinguish roughly two ‘meanings’, while
This is wholly appropriate. For a phrase such as Burke’s ‘It is ordained in the eternal
constitution of things that men of intemperate minds cannot be free’ there could be no
better rendering than «éxpavrat in the sense in which it is sometimes used by Aeschylus.
2 ‘C’% sempre, nel verbo [xpaivar], l'idea di un fine che deve essere raggiunto’ (W. Ferrari,
La parodos, p. 369 n. 3).
4872.2 Oo 193
line 369 COMMENTARY

recognizing that the shifting of meaning depends mainly on the notions

expressed in other parts of the sentence.
The shade of meaning here observed is clearly also inherent in the nominal
compounds Ag. 458 δημόκραντος, 1255 πυθόκραντος, 1488 θεόκραντος, Cho. 612
and Eum. 392 μοιρόκραντος (and a less clear example, Sept. 162 πολεμό-
kpavros). Finally we may add as important confirmation the use of ἐπικραίνεινἶ
in several passages of Aeschylus: Suppl. 14 (why éréxpaver should be rejected
there I cannot understand), Eum. 348, 950, 960.
Of course both ‘meanings’ of κραίνειν often closely approximate to each
other : e.g. in the case of the Homeric ἀκράαντος sometimes the one meaning is
more prominent, sometimes the other. Thus in B 137 f. ἄμμι δὲ ἔργον αὔτως
ἀκράαντον the rendering ‘imperfectum’ will suffice, while in B 202 8eonporins
eO qv... μύθεαι ἀκράαντον or T 565 (ὄνειροι) ere’ ἀκράαντα φέροντες the shade
of ‘irritum’, ‘something not rendered valid’, is preponderant. Wilamowitz
rightly remarks on A. Cho. 65 ἄκραντον : ‘the word is used of the sentence or
counsel for which there is no corresponding fulfilment’ ; it would perhaps be
still more exact to say ‘whose fulfilment is not guaranteed’,
To return to Ag. 369, ἔκρανεν would have to be rendered in Latin by the
pluperfect tense. ‘In Homer, but also in much later writers, an action which
precedes another and is shown to do so by the whole context of the sentence,
may be expressed by an imperfect or an aorist, whether it is in an independent
or in a subordinate sentence. . . . The Greeks attached no importance to
making a distinction between various degrees of past time’ (Wackernagel,
Syntax ; i. 151 f.); ‘the pluperfect is of rare occurrence in Pindar (0. 6. 54) as
in Aeschylus’ (Gildersleeve, Pind. Ol. and Pyth., p. ciii).
οὐκ ἔφα: for the close connexion of od φημι and the like, which is ‘not
peculiar to Greek’, cf., besides the grammar books, Wackernagel, Syntax,
ii. 262.
tis: ‘quo loco τις ad nescio quem atheum referre, ut Blomf. fecit [and, e.g.,
Headlam and W. Nestle, Neue Jahrb. f. d. klass. Altert. 1907, 328 n. 2], plane
absonum est; significat enim nonnullos, multos, ut saepissime', Bamberger,
Opusc. 46. Schütz had already made the same point, similarly Hartung on the
passage, and Wilamowitz, Interpr. 195. Cf. 449 and 590 καί τίς μ᾽ ἐνίπτων
εἶπε, Cho. 58 f. φοβεῖται δέ τις. That it is not a question in this case of an
individual but a typical denial follows moreover from the allusion in 1578 f.
370. θεούς is related in exactly the same way to the preceding statements
about Zeus as in 182 δαιμόνων is to the corresponding statements in that
context. A similar function of θεοί is found several times in Homer, cf.
Nägelsbach, Hom. Theologie, 2nd ed., 129, 17o.
οὐκ ἀξιοῦσθαι : ‘considered it beneath their dignity’ (cf., e.g., Hdt. r. 199. 1).
This is not said by way of any sublimated philosophical conception in the
manner of Epicurus, but as if the gods were playing the part of 'grands
seigneurs' and not disposed to bother about petty affairs.
371. Scots: such sinners are not few. On ἀθίκτων χάρις Wilamowitz, Interpr.
194 f., remarks: ‘What a choice and delicate idea it is to stress the charm and
! The particular shade of meaning of the word in these passages is not noted in the
dictionaries : only Eum. 348 is correctly rendered in Linwood, Lex. to Aesch., ‘were ordained
unto us’. (What is given in the ‘Addenda’ of L-S p. 2071 is not satisfactory : the connexion
with the use of xpaívew here discussed should have been stressed.)

194
 COMMENTARY line 374
attractiveness of what one is not allowed to touch.’ This is a fine thought,
but rather narrow in conception and perhaps not quite accurate. χάρις (cf.
on 182 and 354), ‘charm, grace’, is the vehicle, or the origin, of a χαρίζεσθαι
(in some connexions the act of χαρίζεσθαι), consequently it should evoke in
the recipient a corresponding thankfulness, piety, etc., and also delight (cf.
χαίρειν from xap-). The man who, by opposing its power, shows himself
thankless, impious, etc., is represented as one who tramples xapıs underfoot.
For the xapıs which operates in things that may not be touched, things
‘sacrosanct’, one may aptly compare E. Med. 439 Beßaxe δ᾽ ὅρκων χάρις, the
ὅρκοι being a particular instance of such ἄθικτα.
372. warotro: as is well known, πατεῖν often means ‘tread under foot, trample
on’ (L-S). The verb, as is recalled by Daube, 115, has this connotation of
‘abusing, destroying’ in most passages of Aeschylus and so here, too. One
might say even more definitely, πατεῖν has this sense in all passages of
Aeschylus where it is used transitively, Ag. 372, 957, 1193, 1357, Cho. 642,
Eum. 11o (it is intransitive in Ag. 1298, Cho. 732). Outside the Oresteia no
instance of the word has so far been found in Aeschylus. This typical use of
πατεῖν gives more point to the meaning of ἄθικτα and to the idea of profana-
tion. Cf. on 957, 1193. The man who thus violates the inviolable, making it
βέβηλον in the original sense of the word, miscet sacra profanis.
εὐσεβής : ‘pious’ in the sense that one performs in the right way the duty
of σέβεσθαι τοὺς θεούς, renders to the gods what is due to them in ritual
observance, and acknowledges it in his thoughts and general behaviour. In
the context of this passage the essential idea of εὐσέβεια is that the honour of
the gods and above all of Zeus (cf. on 370) may not be injured by imagining
that they could tolerate neglect of the principle that every act of wickedness
must be atoned for sooner or later. εὐσεβής is by no means a worn-out
expression in the time of Aeschylus. The range of meaning which it has in
his language can be illustrated on the one hand by Suppl. 340 (336 Wil.) πῶς
οὖν πρὸς ὑμᾶς (the Danaids) εὐσεβὴς ἐγὼ πέλω; and the pypav . . . εὐσεβὴς
ὁμιλία (fr. 136 N.), and on the other by the ‘Platonic’ line Sept. 610 σώφρων
δίκαιος ἀγαθὸς εὐσεβὴς ἀνήρ, on which see Jaeger, Die Antike, iv, 1928, 163,
and Wilamowitz, Glaube der Hell. i. 15 n. x.
374 ff. Here we meet the first instance of a difficulty which from this point
onwards is all too frequent, in a passage where we have only the slightest
hopes of recovering the genuine text. The text of many of the choral odes
is naturally in a far worse state of preservation than that of the dialogue;
moreover, here we have to reckon with the unfortunate circumstance that
now we are deserted by M and even by V and have to depend on FTr alone.
374. Casaubon proposed to change éyydvous! to ἐγγόνοις. This easy alteration,
which provides a dative for πέφανται and seems to give a suitable sense, was
adopted by, e.g., Hermann, and also by Wilamowitz. Yet the objection still
holds good which was raised by Bamberger, Opusc. 46, Schoemann, Opusc.

1 There are no grounds for writing ἐκγόνους. In the fifth and fourth centuries ἔγγονος very
often occurs in Attic inscriptions (but ἔκγονος, too, is often found), see Meisterhans 107 n. 972.
In papyri the spelling ἔγγονος is also extremely common, see W. Crönert, Memoria Graeca
Herculanensis, 55. Cf. Vahlen, Opusc. i. 271 n. 10, Neil on Ar. Knights 786, Wilamowitz,
Sappho u. Simonides 92 n. 2 and Platon ii. 338 (on Rep. 364 a, where ἐγγόνων is the MS
reading).
105
line 374 COMMENTARY
iii. 169 f., and Ahrens, 513: everything that is said in this opening section of
the ode about transgressors in general refers primarily to Paris and his people ;
and in his case it is the generation of the guilty themselves, and not their
descendants, who have to atone for the transgression. But even apart from
this the attractiveness of ἐγγόνοις is deceptive. No doubt every reader thinks
at once in this connexion of Homer4 160 ff., and of Solon’s elegy (1. 27 ff. D.),
where we are told that if the guilty man escapes punishment, it is visited
upon his children or his children’s children (cf. in general Theognis 203 ff.) ;
one could add, e.g., A. Suppl. 433 ff., and the oracle delivered to Glaukos by
the Pythia, Hdt. 6. 86 y 2; for. further evidence see Nägelsbach, Die nach-
homerische Theologie, 34 f. But it should also be clear that it is not at all like
Aeschylus to point by a passing hint, in fact a single word, to so important a
thought, and not to give it the slightest further consideration in the whole
extent of an ode of such wide range. At least the complementary idea which
would give this thought its proper place in the system of divine justice (‘he
himself may escape punishment, but . . .’) should have been briefly touched
upon. However, that is not the case, nor does anything else lead up to the
notion of descendants. Headlam praises Hartung’s conjecture ἐκτίνουσα and
makes it the basis of his own restoration of the passage: he writes ἀρή (‘ruin,
destruction’; cf., besides L-S, Bechtel, Lextlogus, 59 f.), but his rendering
of the whole sentence πέφανται δ᾽ ἐκτίνουσ᾽ ἀτολμήτων ἀρὴ πνεόντων μεῖζον ἢ
δικαίως, ‘manifest as penalty for wicked sin hath here been shown us havoc
of presumptuous arrogance’, is not convincing either in what is taken together
or in what is separated. Verrall has accepted Hartung’s τόλμη: and this
form, which is a wrong one for Tragedy, now stands in the Oxford text. I
pass over other suggestions.
To Ἄρη πνεόντων several critics have rightly objected on the grounds that
its content is unsuitable to this passage (a fresh attempt to justify it was
made by Daube, 120 n. 83) ; we may add that the expression, in the connexion
in which we find it here, is proved to be impossible by a point of grammar,
over which even Wilamowitz, Interpr. 195, has passed lightly, in company
with Schütz, Paley, and many others. Ἄρη πνεόντων μεῖζον ἢ δικαίως, ‘of
those who cherish a spirit of rebellion more than is permitted’, ‘die mehr als
gerecht Krieg schnauben’, does not seem to be Greek, at least not fifth-
century Greek. For the so-called adverbial use of μέγα, μεῖζον, μέγιστα is
limited to expressions in which the function of μέγα etc. as an internal object
remains operative, though perhaps somewhat weakened in force. So, besides
the familiar μέγα σθένειν etc., those constructions are also perfectly regular
where an object affected (an ‘external object’) is added, as in Cho. 255 καί σε
τιμῶντες μέγα, Eum. 12, 992 f., E. Heraclid. 1012 f. τὸν θεὸν μεῖζον τίουσα τῆς
ἐμῆς ἔχθρας πολύ. But an expression like πόλεμον μέγα πνεῖ, i.e. a combination
of two effected objects (since μέγα stands for μέγα πνεῦμα), is never found
(cf. on 711), and seems to me to be hardly conceivable. The objection raised
here also holds good against Wilamowitz’s ἀτολμήτων ἔρον πνεόντων μεῖζον κτλ.
(the fact that neither ἔρος nor ἔρον occurs in Aeschylus might be merely
accidental). There are no objections from the point of view of syntax to
Hermann’s restoration of the text. But when he writes “Apy πνεόντων μείζον᾽
(already suggested by Abresch) ἢ δικαίως and translates ‘maiorem quam fas
erat Martem spirantium’ (cf. Stanley’s ‘Martem spirantium maiorem iusto’),
196
 COMMENTARY lines 378-80
such a qualification of πνεόντων "Apn is hardly imaginable. On the other hand,
μεῖζον πνεόντων is in itself excellent, cf. E. Andr. 189 οἱ γὰρ πνέοντες μεγάλα,
Bacch. 640 κἂν πνέων ἔλθηι μέγα. Here the half-adverbial μεῖζον is supple-
mented by ἢ δικαίως, as in E. Heraclid. 1013 (vide supra) it is by the genitive
of comparison.
377. $Aeövrwv: in extant literature the verb is found only here and in 1416
in the same form.
δωμάτων : here as often elsewhere, e.g. 1468, and like δόμοι frequently, used
to denote not only the house itself, but also the family with their way of life
and their destiny.
ὑπέρφευ: Wilamowitz on E. Her. 1321: ‘The word was coined by Aeschylus’:
not very convincing. It is of course possible to assume that when Cratinus
(fr. 359), as we are told by Phrynichus, Praep. soph. Ὁ. 89. ı de Borries, used
μηδὲν ὑπέρφευ in the sense of μηδὲν ἄγαν, that was a case of παρατραγωιδεῖν.
But it is equally possible that this μηδὲν ὑπέρφευ was a common variation of
the favourite maxim; that would well account for the earliest passage in
which we find it, Pers. 820 (ὡς οὐχ ὑπέρφευ θνητὸν ὄντα χρὴ φρονεῖν), and our
passage, which turns on a divine precept, points in the same direction ; while
the use Euripides makes of the word (Her. 1321, Phoen. 550) is further removed
from this.
For the hiatus at the end of the catalectic trimeter see note on 239."
378. On ὑπὲρ τὸ βέλτιστον Schütz observes: ‘haec verba otiosam vocabuli
ὑπέρφευ explicationem sapiunt’. Others have assumed corruption of the text
here. Wilamowitz’s comment (on E. Her. 1321 and Interpr. 195) is also based
on that of Schütz, except that he omits the 'otiosam': ‘ ὑπέρφευ was not
enough: hence the explanation ὑπὲρ τὸ βέλτιστον ᾿. This is inadequate. It
should be recognized that Aeschylus forcibly over-emphasizes the note of
excess, because it is for him an extremely important idea: excess, τὸ λίαν, τὸ
äyav, etc., is in his opinion the very thing which imperils human happiness
and peace more than anything else. It is not an exact parallel, but a case
closely resembling this in thought, when Darius, whose chief mission consists
in constantly renewed rejection of impious excess, uses twice in quick suc-
cession a very similar redundance of expression, Pers. 794 κτείνουσα λιμῶι
τοὺς ὑπερπόλλους ἄγαν and 827 Ζεύς τοι κολαστὴς τῶν ὑπερκόμπων ἄγαν φρονη-
μάτων ἔπεστιν, cf. also on Ag. 1013, and 5. Aj. 951 ἄγαν ὑπερβριθὲς ἄχθος
jvvcav (for text see Wilamowitz, Verskunst, 506). Similarly Homer o 405
οὔ τι περιπληθὴς λίην τόσον. For other redundant expressions of the same kind
in Aeschylus cf. 428 τῶνδ᾽ ὑπερβατώτερα, and note on 215 ff.
ὑπὲρ τὸ βέλτιστον : ‘beyond what is best for them’ is Paley’s paraphrase,
who aptly compares Plat. Phaedr. 233 a ἐκεῖνοι μὲν yap καὶ παρὰ τὸ βέλτιστον
τά τε λεγόμενα Kal τὰ πραττόμενα ἐπαινοῦσιν.
378-80. ἔστω... λαχόντι. In 379 of course only the text of F represents the
παράδοσις. Triclinius in his characteristic way wrote κἀπαρκεῖν, in order to
restore the correspondence with 397 τὸν δ᾽ ἐπίστροφον τῶνδε (--οα--ωο -- — v),
cf., e.g., on 998.
! Murray, who, by transposing the words, spoils the word-order (for ὑπέρφεν ὑπὲρ τὸ
βέλτιστον forms one phrase, cf. on 378), employs the same method where he regards a hiatus
in dochmiacs as unpleasant, e.g. Eum. 259, 784 (and yet in Eum. 840 πνέω τοι μένος ἅπαντά
τε κότον he leaves the syllaba anceps alone).

197
lines 378-80 COMMENTARY

ἀπαρκεῖν: the dictionaries, from Wellauer's Lex. Aesch. to L-S?, give the
special meaning contentum esse, to be contented, for this passage. The responsi-
bility for this lies, as it so often does, with Stanley's translation. He took the
λαχόντα of the MSS as the subject to ἀπαρκεῖν and translated: μέ contentus
sim recte sapiendi compos. This is completely inconsistent not only with the
usage of Tragedy (Pers. 474, S. Oed. C. 1769, Eur. fr. 892. 4 N.) but also with the
whole of Greek as far as one can see from the Thesaurus. Lycophron's (1302)
ἀπηρκέσθησαν, which is taken in the same way by Passow and L-S, even if
it were not an artificial expression, could not support the meaning of 'be
content’ for ἀπαρκεῖν ; nor could we for such a use rely upon the middle or
passive ἀπαρκεῦμαι in the anonymous iambographer of the Heidelberg
papyrus I. 79 (Diehl, Anth. Lyr. i. 3, p. 114), where the context is unknown,
and Α. D. Knox (Loeb Library) conjecturally translates 'I am content with
little’. Nor is this meaning supported by ἀρκεῖν: this does not mean ‘to be
content’ but ‘to suffice’, in both personal and impersonal constructions.'
Auratus avoided this mistake by writing λαχόντι. Hermann objects: 'non
opus esse Schützii coniectura λαχόντι, hodie nemini ignotum est'. As is
evident from his punctuation (wore κἀπαρκεῖν between commas) and presum-
ably, too, from Humboldt’s translation which he had supervised ("Doch
harmlos, und so, dass der Habe Mass still gnügt, sev es, bei Sinnes Weisheit"),
he subordinated λαχόντα to ἔστω ἀπήμαντον. So he evidently had in mind
cases like E. Med. 659 f. ὅτωι πάρεστιν μὴ φίλους τιμᾶν καθαρᾶν dvoi£avra
(-ἔαντι inferior MSS) κλῆιδα φρενῶν. But in this and in other comparable cases
(cf. Elmsley on E. Med. 727 in his numbering [= 744] and on Heracltd. 693)
we are concerned as a rule with an infinitive construction (cf. Jebb and
Kaibel on S. El. 479 f., Blass on A. Cho. 410 f., and my note on Ag. 1611),
and even where this is missing, as in Pers. 913 f., Cho. 41o f., S. El. 479 f.,
there is (cf. note on 1055 f.) a preceding dative of the personal pronoun or
something corresponding (in Pers. 913 Schütz is probably justified in writing
ἐμοί for éuóv,* otherwise one would have to suppose that ἐμῶν γυίων is a
sufficient substitute for the personal pronoun). The infinitive as well as the
pronoun makes it easier for the hearer to refer the accusative of the participle
either to the 'logical subject' of the action denoted by the infinitive or to the
person indicated by the pronoun. But in this instance λαχόντα has no such
support. λαχόντι is therefore necessary. Its grammatical connexion will be
discussed presently.
ἔστω δ᾽ ἀπήμαντον: ‘the difficulty is to find a definite subject to ἔστω,
which some make τὸ πρᾶγμα, others τὸ τῆς τύχης ' (Paley). The arbitrariness
of 'supplying' such subjects is obvious. Nor does this method of approach do
justice to the strong emphasis on ἔστω expressed by its position at the
beginning (cf. on 958). ἔστω is not merely copula but means 'let there be'
(viz. 'available' or 'allotted' and the like). For the force of the verb placed
like this cf., e.g., Eum. xo12 f. εἴη δ᾽ ἀγαθῶν ἀγαθὴ διάνοια πολίταις, Suppl. 974
εἴη δὲ rà λῶιστα, and on the other hand for the copula cf. Suppl. 663 f. ἦβας
δ᾽ ἄνθος ἄδρεπτον ἔστω, 686 f. εὐμενὴς δ᾽ ὁ Δύκειος ἔστω πάσαι νεολαίαι. Verrall's
1 L-S s.v. ἀρκέω IIT wrongly translate S. Aj. 76 ἀλλ᾽ ἔνδον ἀρκείτω μένων: ‘let him be
content to stay within’.
2 Murray’s punctuation (full stop after ῥώμη) is unfortunate, removing as it does εἴθ᾽
ὄφελε from its place at the beginning of the next sentence.
198
 COMMENTARY lines 381-4

punctuation, adopted by Headlam, with a comma after ἔστω δ᾽, rests on the
assumption that ‘the subject of ἔστω is τὸ βέλτιστον ’ (Verrall ; cf. Headlam's
translation '. . . beyond the Best ; and let that be, with a good endowment of
the mind to have sufficience without hurt’). This is bad, for it blurs the
clearness of the sharp contrast: first, up to ὑπὲρ τὸ βέλτιστον, the behaviour
of the presumptuous and vainglorious man, then, from ἔστω onwards, the
wish of the pious and sensible one. The difficulties disappear when we take
ἀπήμαντον as the subject. ἀπήμαντον in this context is well illustrated (van
Heusde) by Theogn. 1153 f. ein μοι πλουτοῦντι κακῶν ἀπάτερθε μεριμνέων
ζώειν ἀβλαβέως μηδὲν ἔχοντι κακόν. ‘Be there allotted that which is harmless.’
Joined to this by way of explanation or qualification is ὥστ᾽ ἀπαρκεῖν (Daube,
121 n. 84, compares Solon fr. 5. ı δήμωι μὲν γὰρ ἔδωκα τόσον γέρας ὅσσον
ἀπαρκεῖ). As regards ὥστε, it may well in this and similar cases be taken, with
Kühner-Gerth, ii. 502, in the sense that ‘the notion of τοσοῦτος is implied in
the connexion of thought’.
The contrast with the preceding passage suggests taking together the
words ἔστω ed πραπίδων λαχόντι (cf. the passage quoted above from Eum.
1012 f. ein δὲ... διάνοια πολίταις). But the definition ‘for the wise’ also suits
well with awapkeiv, cf. (van Heusde) E. Phoen. 554 ἐπεὶ τά γ᾽ ἀρκοῦνθ᾽ ἱκανὰ
τοῖς γε σωφρόσιν. So perhaps it will be best to connect εὖ πραπίδων λαχόντι
both with ἔστω and with ἀπαρκεῖν.
380. The construction of Aayxavw with the genitive, which goes back to
Homer, occurs only here in Aeschylus, occasionally in Sophocles and Euri-
pides. For the thought cf. της, for the expression 802.
The prayer éorw . . . λαχόντι is parallel to the avowal and wish in 471 ff.,
except that in this case the first person is not so strongly emphasized, though
the elders count themselves among the εὖ πραπίδων λαχόντες.
381-4. The text of this ‘rhythmical refrain' (cf. p. 186) is firmly established,
and it is only the grammatical relation of some of its expressions that is
controversial. We start with πρὸς κόρον. This is sufficiently explained by the
instances quoted from Tragedy in Blomfield's Glossary for πρὸς βίαν, πρὸς τὸ
καρτερόν, πρὸς ἡδονήν etc., and others of the same kind in L-S s.v. πρός
C. III. 7 (p. 1498 f.). We must accordingly take πρὸς κόρον λακτίσαντι together.
It seems doubtful whether πλούτου depends on ἔπαλξις or on πρὸς κόρον. The
great majority of commentators join ἔπαλξις πλούτου together, 'the protection
afforded by wealth'; on the other hand Pauw, Wecklein, and Plüss make
πλούτου depend on κόρον ; similarly Headlam ('For there is no strong fortress
to be found for him that in surfeit of wealth kicks into obscurity . . .’) and
Verrall in his translation (for in his notes he proposes as an alternative taking
ἔπαλξις πλούτου together). It is, to start with, an argument in favour of the
latter interpretation that the thought 'there is no defence for the man who is
spoilt by too much prosperity and... . overthrows . . . the altar’ is in this
passage far more suitable in its unqualified assertion than 'for prosperity
affords no protection for the man who in insolence . . . overthrows’. Moreover,
the absolute assertion οὐ ydp ἐστιν ἔπαλξις is more forceful, and one would be
reluctant to sacrifice the correspondence with οὔτις ἀλκά in 467 : both passages
are perhaps a development of the Homeric (u 120) οὐδέ τίς ἐστ᾽ ἀλκή (cf.
ᾧ 528, x 305). The addition of the genitive πλούτου to the phrase πρὸς κόρον,
which serves as an adverb, is not open to objection, cf. S. Ant. 3o πρὸς χάριν
199
lines 381-4 COMMENTARY

βορᾶς (with Jebb’s note) and Phil. 534 πρὸς ἰσχύος κράτος. The function of eis
ἀφάνειαν still remains to be settled. Strangely enough Heath, Musgrave,
Schütz, Hermann, Schneidewin, and others make it go with οὐ γάρ ἐστιν
ἔπαλξις πλούτου, and take eis = adversus. On the other hand, Pauw, who has
a right understanding of the whole of this sentence, takes Aakrioavrı eis
ἀφάνειαν together, as do Paley, Wecklein, Sidgwick, Verrall, and Headlam.
The order of words points to this construction, which is further confirmed if
we observe that in the other five cases of the ‘rhythmical refrain’ in this ode
the concluding priapeus forms a self-contained syntactical unit, and in no
case is the last line closely bound with line 1 or 2. The connexion between
λακτίσαντι and εἰς ἀφάνειαν is a comparatively loose one in that the result of
the action is briefly expressed by a prepositional phrase, cf. on 998 f. As an
expression of complete annihilation εἰς ἀφάνειαν corresponds to the idea in
466 f. The βωμὸς δίκας recurs in Eum. 539, which continues in exact corre-
spondence with our present passage: μηδέ vw κέρδος ἰδὼν ἀθέωι modi λὰξ
ations, Cf. Cho. 641 ff.
For the idea that κόρος πλούτου (or ὄλβου) gives rise to wickedness and
disaster, cf., besides the familiar passages in Solon (fr. 3. 9 ff., 3. 34, 5.9 D.),
Pindar, Ol. 1. 56 f., 13. το, and Theognis 153 f. (on which see P. Friedlander,
Hermes, xlviii, 1913, 588), Hdt. 3. 80. 3 ἐγγίνεται μὲν γάρ οἱ ὕβρις ὑπὸ τῶν
παρεόντων ἀγαθῶν, φθόνος δὲ ἀρχῆθεν ἐμφύεται ἀνθρώπωι. δύο δ᾽ ἔχων ταῦτα
ἔχει πᾶσαν κακότητα" τὰ μὲν γὰρ ὕβρι κεκορημένος ἔρδει πολλὰ καὶ ἀτάσθαλα, τὰ
δὲ φθόνωι.
For 383 f. cf. Lycophron 136 f. (Paris is being addressed) ἔτλης θεῶν dAovrós
ἐκβῆναι δίκην λάξας τράπεζαν κἀνακυπώσας Θέμιν.
385. τάλαινα: I see no reason for abandoning the ordinary meaning (‘miser-
able, wretched’ etc.) in this and other passages of Tragedy in favour of a
more closely etymological one (Denniston on E. El. 1171).
386. Nowadays nobody doubts Hartung’s emendation. πρόβουλος occurs
here for the first time, the adverb in Cho. 620 (for the text cf. on 1613); in a
more definitely technical sense πρόβουλος is used in the end of the Septem
(1006), which probably dates from the fourth century. Here the idea clearly
is that Ate, like, e.g., the βουλή in Athens, is responsible for the προβουλεῦσαι:
the προβούλευμα which she authorizes is then passed on to Peitho, who in her
turn takes the necessary measures for its execution.
ἄφερτος we may with a high degree of probability call a new coinage of
Aeschylus. In the whole of Greek literature (apart from late imitators) it
occurs only in the Oresteia, where it is found in nine passages, of which five
are in Ag. and two each in Cho. and Eum.* This probably indicates that the
poet invented the word for the first play of the trilogy, and to begin with
freely employed this happy coinage, then as he went on used it more sparingly.
The same can often be observed in the rise of new words in poetry. Cf., e.g.,
my book Plautinisches im Plautus, 106 n. 3; and for Aeschylus, Schade-

1 Usener's hypothesis, Kl. Schr. iv. 454, that the picture must have suggested itself to the
poet through some ritual custom (the upsetting of the sacrificial altar) seems to me to be
doubtful.
2 Its meaning in all passages is ‘unbearable’ and nothing else. P. Von der Mühll, if his
view is correctly reported by Daube 7 n. 25, takes it in Ag. 386 to mean ‘nicht ausgetragen’,
of a child whose birth is untimely. But in ἄφερτος there is no trace of the aus-element.
200
 COMMENTARY line 387
waldt, Hermes, lxvii, 1932, 328 n. 3, and W. Ferrari, La parodos, 376, and the
notes on 12, 34, 450, 525, 629, 639, 883, 887. Herodotus, and later Thucydides
and others, use ἀφόρητος : this word, but not ädepros, is found in the list in
Pollux 3. 130; and it is also used as a gloss on ἄφερτος, e.g. in the scholion
on Ag. 386 and on Cho. 469.
mais . . . "Aras: the powers of evil genealogically connected, as in Eum.
533 δυσσεβίας ὕβρις τέκος, cf., e.g., Solon fr. s. 9 D. τέκτει γὰρ κόρος ὕβριν,
Pind. Ol. 13. το Ὕβριν, Κόρου ματέρα θρασύμυθον, Bacchyl. Paean 4. 23 f.
Snell τίκτει δέ τε θνατοῖσιν εἰρήνα μεγαλάνορα πλοῦτον κτλ, oracle in Hdt. 8.
77. 1 δῖα Δίκη σβέσσει κρατερὸν Κόρον, Ὕβριος υἱόν. Cf. on 753. To call Peitho
the daughter of πρόβουλος “Arn looks like a variation in malam partem,
preserving the important προμηθές element, of the genealogy in Alcman fr.
44 D. which makes Peitho, together with Tyche and Eunomia, /Ipouadeias
θυγάτηρ (a thoroughly rationalizing interpretation is given by Wilamowitz,
Hermes, lxiv, 1929, 487); however, each of these conceptions may have
arisen spontaneously. ary (cf. on 1192) is the blind infatuation, the παρακοπὰ
πρωτοπήμων (223) which leads to crime and ruin. The description is given in
Pers. 97 ff. of how "Ar carries out her acts of rapacaivew and παράγειν on man,
whom she entangles in the net, and how all further evil proceeds from this.
To this action πείθειν corresponds; thus we can well understand the concep-
tion which calls ‘wretched Peitho’ the child of Ate. As a result of arn Peitho
overpowers a man by persuading him that he is obliged to do what it is not
right for him to do, and at the same time talks him out of his resistance.
This is the situation of Agamemnon as it is shown in his words 206 ff. and
subsequently explained (217 ff.). The other case, perhaps still commoner, is
that in which the πειθώ of another man persuades the doer of the crime to
his action : thus, e.g., 5. El. 562 πειθὼ κακοῦ πρὸς ἀνδρός is given as the cause of
Clytemnestra’s crime. The βιᾶσθαι of Peitho brings the man who comes in
her way ‘under the yoke of compulsion’.
387. Musgrave’s separation of the words is obviously right, for what is
required here is the idea ‘every cure is vain’, cf. 1170 ἄκος δ᾽ οὐδὲν ἐπήρ-
xeoav. For the position of πᾶν cf. Suppl. 375 χρέος πᾶν ἐπικραίνεις. No ob-
jection to this separation of the words is to be founded on the fact that,
whereas such assimilations are usual in inscriptions and papyri (cf. W.
Crönert, Memoria Graec. Hercul. 61 f.), they have only rarely survived in our
medieval MSS (cf. Eduard Schwartz in the Berlin edition of the works
of Eusebius, ii. 3, p. cxcvii, and, in general, Starkie’s critical note on
Ar. Clouds 973 and Schwyzer, Griech. Gramm. i. 40) 1.).} For usually
(in cases of ἐμ etc.) the correct separation of words was self-evident,
but in the present case it is possible for the four syllables to have been
taken as rauparauov.*
1 Whether the ξὺμ of Q and L in Sept. 605 really goes back to ancient παράδοσις I cannot
venture to decide; in Sept. 764 the reading of M! is not σὺμ βασιλεῦσι but συμβασιλεῦσι. In
Eum. 987 Blass and Wilamowitz state that M originally had ἐμ βροτοῖς, Wecklein says
nothing (in the facsimile it looks as if there were an erasure here, but that could only be
decided by an examination of the original). For ἐμ παιδοτρίβου and the like being preserved
in MSS of Aristophanes, cf. Von der Mühll, Gnomon, 1925, 319.
2 Similarly Cho. 1050, where ἀμμείναιμ᾽ may have been taken as a form from ἀναμένω.
A. Suppl. 137 συμπνοιαῖς is an obvious misinterpretation. Similarly in Sept. 661 σὺμ (as in
MF) φοίτωι may owe its retention to an association with συμφοιτᾶν and συμφοιτητής (it is
201
lines 387 ff. COMMENTARY
387 ff. Here the development of the thought step by step can be easily
observed. To the clause οὐκ ἐκρύφθη (gnomic aorist) σίνος, the expression
πρέπει δέ is added to provide a positive complement to the negative verb, and
this πρέπει, ‘stands out, shows brightly out of darkness’, used of flame and
light e.g. in 30 and Sept. 390 (cf. on 242), leads to the image of φῶς αἰνολαμπές.
This stylistic device by which an idea first presented in the verb is then
intensified in a substantival expression may be observed elsewhere. Cf. on
827 and 842, and, e.g., Sept. 498 βακχᾶι πρὸς ἀλκήν, where the idea outlined in
βακχᾶι is further developed in the succinct simile θυιὰς ὥς (Stanley, Schütz,
Blomfield, and others rightly place this between commas); to this there is
attached, independently of the immediately preceding words, a fresh supple-
mentary expression with φόβον (not to be altered) βλέπων, the whole being a
good example of the guttatim style, cf. on 2.
σίνος : ‘Ionic word, very rare in Attic prose’ (L-S). In Drama it has so far
been found only in the Agamemnon, where it occurs three times, a further
instance of what has just been noted in the case of ἄφερτος (386).
391. ΄ προσβολαῖς vel ut Heathius exp. attritione ad lapidem Lydium, vel
omnino quacumque allisione . . . (nummi) crebro usu affricti et allisi' (Schütz).
The 'touchstone' interpretation, which started from Heath, still enjoys a
wide popularity in spite of Hermann's warning (vereor ne non de lapide
Lydio loquatur poeta, sed aes adulterinum dicat, quod usu tritum amisso
falso splendore cognoscitur"); it must be firmly rejected, tempting though it
may be to connect zpifex with such expressions as Theogn. 450 χρυσὸν...
τριβόμενον Baadvox (cf. ibid. 417 f.), Hdt. 7. 10 a 1 ἐπεὰν δὲ παρατρίψωμεν (sc.
τὸν χρυσόν) ἄλλωι χρυσῶι, διαγινώσκομεν τὸν ἀμείνω and the like. For in this
case there is no question of gold, as in every instance where the βάσανος is
mentioned, but of χαλκός, and bronze cannot be tested with the touchstone.
An explanation like that of Sidgwick: ‘he is found “black smutched" like
bad copper (instead of being bright like gold)’ is a mere subterfuge, for (1)
in the comparison χαλκός is not what appears as the result of a change, but
the original material in which the change takes place (μελαμπαγὴς πέλει), (2)
the antithesis to κακὸς χαλκός would not be gold, but ἀγαθὸς χαλκός, or better
in the words of Sophocles (fr. 780 N. = 864 P., a passage already quoted by
Musgrave) λάμπει γὰρ ἐν χρείαισιν ὥσπερ εὐγενὴς (cf. Pearson’s note) χαλκός.
Nor is Headlam's translation ‘base metal’? admissible here, where the point
is exactly the particular kind of metal. Those who cling to the touchstone
would at least be consistent if they accepted Paley’s suggestion ‘perhaps we
should read χρυσοῦ, in allusion to the use of the touchstone’. The ‘touchstone’
idea would probably not have persisted so tenaciously had it not been sup-
ported by the alleged meaning of δικαιωθείς. The rendering of this word has

worth noticing that in M this is written συμφοέτω, and, e.g., at Eum. 1024 ξυμπροσπόλοισιν,
not ξὺμ as it is given in Blass’s apparatus), and in Pollux 4. 130 ἐμ ψυχοστασίαι may have
been kept because of some reminiscence of ἔμψυχος. Cf. E. El. 752 ἐμφόνιον for ἕν" φόνιον,
Ion 1581 ἔμφυλον for ἕν φῦλον, Phoen. 589 (MBV) οὐμμέσωι (the error in the breathing-sign
does not matter). In Bion’s Adonts 55 the MSS have πάγκαλον (= πᾶν καλόν), which may
be due to a misunderstanding facilitated by the πανάποτμος of the next line; preservation
of an original πᾶγ καλόν is not likely. Cf. on 1172.
! It shows, however, that he was at pains to connect a definite idea with the words. The
same cannot be said of G. Thomson. His note is confused, and his translation ‘like to false
bronze betrayed by touch of sure-testing stone’ does not make sense.
202
 COMMENTARY line 391
had to suffer even to the present day (L-S? ‘proved, tested’) from the conse-
quences of Stanley's wrong translation probatus.! W. Sewell translated
rightly ‘blacken’d with bruise and many a blow, to sentence is he brought’,
but this remained unnoticed, like so much else which that good and honest
man understood better than most. Then in 1860 Ahrens expressly protested
against foisting on to the word a meaning quite alien to it. His own explana-
tion is not a happy one, but it points in the right direction. His admonition
was followed by only a few, e.g. Wecklein ; independently of him Verrall says
rightly : ‘the rendering ''tested" is not supported, so far as I can discover, by
any example or analogy.’ The latter point is fundamental, for actually the
meaning of δικαιοῦν as βασανίζειν would not only be unsupported by any
other instance of the use of the word, but would contradict everything that
we know about the function of verbs in -οῦν derived from -o-stems generally.
Thus Ernst Fraenkel, Griech. Denominativa, 124, reasonably puts the passage
under the heading ‘provide justice’, ‘sentence’, ‘punish’. The nearest parallels
are Hdt. x. 100. 2 ei τινα πυνθάνοιτο ὑβρίζοντα, τοῦτον... κατ᾽ ἀξίην ἑκάστου
ἀδικήματος ἐδικαίευ, 3. 29. 3, 5. 92 B 2 (oracle) ; for Thuc. 3. 40. 4 ὑμᾶς δὲ αὐτοὺς
. . . δικαιώσεσθε cf. Classen-Steup. The comparatively rare meaning was
noticed by ancient grammarians: Suid. δικαιοῦν" δύο δηλοῖ, τό τε κολάζειν καὶ
τὸ δίκαιον νομίζειν: οὕτως ‘Hpddoros. R. Hirzel, Themis, Dike und Verwandtes,
137 n. 6, quotes for the meaning ‘punish’ besides Plato, Laws 934 b (δικαιού-
μενον) also Cic. Verr. ii. 5. 148 ἐδικαιώθησαν inquit, hoc est, ut Siculi loquuntur,
supplicio adfecti et necati sunt. Beazley reminds me of the Italian ‘giustiziato’
— 'executed'. The δικαιωθείς of the Agamemnon line is to be added to the
numerous cases (collected by Wackernagel, Unters. z. Homer, 122 ff.) of verbs
in -oöv first found in the aorist passive, cf. on 440.
προσβολαῖς : misinterpreted in LS, Stanley correctly ‘allisionibus’.
προσβάλλειν is used of damaging impacts of one object upon another, e.g. E.
Hipp. 1233 ἁψῖδα πέτρωι προσβαλὼν ὀχήματος, Plato, Lach. 183 d προσβαλούσης
yàp τῆς νεὼς ἐφ᾽ ἦι emeßdrevoev πρὸς ὁλκάδα τινά. To strike or knock against
(or to batter) gives exactly the right sense here. A knock against a solid
object causes scratches or other damage and disfigurement? on the surface of
the bronze; that is a second possibility of damage besides rpißos, wearing
away by continual use. The combination of the two nouns, τρίβωι τε καὶ
προσβολαῖς, gives a more detailed picture of what Sophocles, in the above-
cited fragment, calls ἐν χρείαισιν. Of the sinful man it could not be said
λάμπει yàp ἐν χρείαισιν ὥσπερ εὐγενὴς χαλκός. At first it may indeed seem as if
all were well with the evil-doer, and people will be imposed upon by his fine
appearance as they are by the shining surface of a bronze vessel so long as it
isnew. But in course of time the sinner’s real nature will come out, and when
he has met with his punishment, everyone will say : of course, he was made of
1 Even Blomfield, who is as a rule careful in his statements about the use of words, has
not freed himself from Stanley’s influence: ‘In hoc loco δικαιωθείς videtur significare
probatus.’
2 This helps us to understand in what sense προσβολαί is used Eum. 600: δυοῖν γὰρ εἶχε
προσβολὰς μιασμάτων. There also it denotes any disfigurement of the surface, and not merely
‘stains’, although the latter is of course, included. With regard to this use it is instructive
to note that in Antiphon Tetral. 2. 3. 8 the phrase τὰς θείας προσβολάς undoubtedly picks up
the preceding θεία κηλίς, whether we approve Wilamowitz’s arrangement of the text,
‘Comment. gramm. IV’, Index schol. Göttingen 1889/90, p. 19 n. 2, or another.
203
line 391 COMMENTARY

bad stuff. So Paris, after committing his crime, was first greeted with jubila-
tions by his kinsfolk and countrymen, only to be cursed by thern when he,
and they with him, had been condemned by the gods (705-17). The phrase
κακοῦ χαλκοῦ τρόπον obviously refers to bronze utensils, such as cauldrons,
tripods, etc., not to ‘base coin’ (as, e.g., Schütz, Schneidewin, and Wecklein
take it). The minting of bronze coinage was unknown in Athens in the time
of Aeschylus and for a long while afterwards (cf. Head, Hist. Num. 2nd ed.,
376), nor were bronze coins in circulation yet in other parts of Greece at that
time.’ A clue to the special meaning of κακοῦ χαλκοῦ is provided by its close
connexion with the fate of the sinful man. This connexion makes it unlikely
that the ‘badness’ of the bronze should in this case be supposed to be caused
by a lack of skill on the part of the craftsman or an unlucky accident in the
process of casting. This ‘baseness’, as opposed to the noble nature of εὐγενὴς
χαλκός, is rather to be regarded as the result of a dishonest manœuvre by which
a trustful customer might be deceived while the artisan makes an improper
profit. The obvious assumption, then, is that the κακὸς χαλκός contains an
unduly high proportion of a cheap material, namely lead. This admixture
has the effect that when the bronze object has been used for some time and
suffered from wear and tear as well as battering, there appears instead of the
fine lustre of the surface an unsightly blackness which cannot be removed.
The fact that κακὸς χαλκός in this context suggests a considerable admixture
of lead was recognized by the expert whom A. Platt, J. Phil. xxxv, 1920, gı
quotes: “The real explanation, as I learn from Professor Collie, is this: good
bronze has no lead in it ; if lead be present, oxide of lead is formed by contact
with the air, and this is black; even if no lead be there, but the bronze is
insufficiently smelted, it may blacken through the formation of oxide of
copper. Bad bronze blackens at once after polishing, and that is what
Aeschylus refers to by rpißwı καὶ προσβολαῖς. This comment, though it -
started from a right proposition, was bound to miss the implication of the
image since the meaning that it ascribes to τρίβωι καὶ προσβολαῖς is untenable
(see the notes above). I therefore looked for a different explanation of the
technical process. Through the kindness of Dr. Gisela Richter of the Metro-
politan Museum of Art in New York I was put in touch with Dr. Colin G.
Fink, Head of the Division of Electrochemistry in the Department of Chemical
Engineering of Columbia University. He applied to the problem his great
metallurgical knowledge and a keen interest in the interpretation of the
passage, and so reached what seems to me a most satisfactory solution. This
is what he writes:
‘Whereas copper and tin form alloys within wide limits of composition,
copper and lead form no such alloys. Only a relatively small percentage of
lead will alloy with copper. We have examined a large number of ancient
bronzes and we have frequently detected lead globules included or “‘sus-
pended” in the bronze or copper. The X-ray photograph readily revealed
! Even in Sicily bronze coinage only begins c. 420-410 B.C., cf. Head, op. cit. 117: refer-
ences to the individual cities ibid. p. 122 (Akragas), p. 130 (Kamarina), p. 146 (Himera),
Ρ- 175 (Syracuse), etc.—Wilamowitz has not of course committed the mistake of thinking
of bronze coins, but when he translates ‘black like counterfeit coin, from which time has
worn off the false gleam of silver’, by which he obviously means plated or otherwise debased
silver coins, this is not compatible with χαλκοῦ (see above).
204
 COMMENTARY line 392
the presence of these lead globules, lead being relatively opaque to X-rays
as is well known.
‘A true 9o: το copper-tin bronze is a noble alloy, highly resistant to
atmospheric corrosion and to wear. Bronze tools, spears, etc. were far
superior to those of copper.
‘In our experience with “leady’’ bronzes we have found that any lead
globules imbedded in the surface of the bronze article will spread or “flow”
(as we say) by "rubbing and attrition” so as to cover much of the bronze
with a film of lead surface. Now, whereas the pure, noble, 90: 10 bronze
is relatively stable in the atmosphere, retaining its golden lustre, the lead-
bearing bronze, after it has been in use for a relatively short time and the
lead, due to its softness and malleability, has spread over the surface, soon
turns black due to the formation of lead sulfide—the sulfur content of the
atmosphere being derived from the many open fires, oil lamps, torches, etc.
‘Any mechanical means such as rubbing or wear or attrition or polishing,
that will cause the lead globules in the surface to flatten out and ''flow"
over the bronze surface, will soon give rise to a black patina. On the other
hand, the noble go: 10 bronze improves by wear, rubbing, etc. "for it
shines like noble bronze”.
‘The black sulfide-of-lead patina is not easily eliminated. It is, for all
practical purposes, “indelibly black"; rubbing will give it an ebonite
appearance.
‘The addition of lead to bronze was practised by the ancients as well as
by the “metallurgists” of the renaissance.’

391 ff. The subject of the sentence is the presumptuous evil-doer, to whom the
whole of the preceding passage refers. It does not matter that the immedi-
ately preceding subject from the purely grammatical standpoint was σίνος.
The language at this early stage allows the person about whom a statement
is made to be understood simply from the context ; cf. on 71.
392. μελαμπαγής (for the fact that the ‘high’ [or ‘strong’] grade of the vowel
of the root is quite normal in such compounds cf. Wackernagel, Vermischte
Beiträge, 16 ; he classes μελαμπᾶγής and Sept. 642 καινοπηγής with the Homeric
εὐπηγής) : elsewhere the word is found only in Sept. 737 of the blood of the
slain man. In Ag. 392 the -πᾶγής element is neglected by several modern
scholars (including L-S). On the other hand, Headlam rightly says: ‘a dark
stain fixed in him’. Tucker remarks on Sept. 737: ‘In -παγές is implied more
than mere clotting. There is the same allusion to the supposed indelibleness
of the bloodstain of murder as in Cho. 67 τίτας φόνος πέπηγεν οὐ διαρρύδαν.᾽ In
this passage, too, ‘indelibleness’ is indicated: the damaged bronze vessel and
the discovered and punished evil-doer remain dark-coloured ; the former fair
colour that showed on the surface does not return. The choice of the pictur-
esque μελαμπαγής is presumably due in the first place to the comparison with
bronze of bad quality. However, the word may also have a direct reference
to the evil-doer. Thus in Pindar’s famous lines about Theoxenos (fr. 123
Schr.) the wretch who is proof against the boy’s beauty is described thus: ἐξ
ἀδάμαντος À σιδάρου κεχάλκευται μέλαιναν καρδίαν ψυχρᾶι φλογί, in the σκόλιον
of ‘Solon’ quoted by Lobon (Scol. anon. 32 D.), which probably dates from
the fifth century (Wilamowitz, Hermes, lx, 1925, 300), it is said of the traitor
205
line 392 COMMENTARY

that he speaks deceitfully ex μελαίνης φρενός; Diog. Laert. 8. 34, following

Aristotle (= Diels-Kranz, Vorsokr. i, sth ed., 463.19), comments on the
Pythagorean ritual thus: τὸ μὲν λευκὸν τῆς τἀγαθοῦ φύσεως, τὸ δὲ μέλαν τοῦ
κακοῦ; a quotation from a corresponding Pythagorean doctrine in [Plut.]
De lib. educ. 17, p. 12 d runs: ‘ μὴ γεύεσθαι μελανούρων ' (paroemiac) τουτέστι
μὴ συνδιατρίβειν μέλασιν ἀνθρώποις διὰ κακοήθειαν (quoted by Schneidewin).
393 ff. δικαιωθείς : for the meaning see above (p. 202 f.). The close connexion
with μελαμπαγὴς πέλει is essential. Only after the guilty man has received due
punishment will those who were deceived in their opinion of him in the time
of his prosperity see clearly the blackness and worthlessness of his character.
We may indeed say, if we like, that the effect of punishment on the sinner and
his reputation is the same as that of τρίβος and προσβολαί on the bronze vessel :
the fair appearance of the outside is gone. But until we are given the key in
δικαιωθείς, we are not carried beyond the imagery of the bronze vessel and a
quite general reference to the evil-doer: it is the participle that completes the
meaning of the sentence. This expression, ‘he received his punishment’, leads
straight on to the ἐπεί clause. ‘For a boy runs after a bird... .’ Properly it
might go on: ‘he enjoys his sport for a while, but then very soon....’ But
here, where the ode is pressing on to the main point and is for the rest content
with mere hints, there is no room for such explicitness (for proverbial sayings
being briefly alluded to cf. on 1316). The preceding maxim is illustrated by
the sentence ‘for a boy pursues’, etc. It is usual to take (as Blomfield does,
who says 'subaudiendum ws ante παῖς ’) as the subject of διώκει not παῖς but
the subject of the previous sentence. Now it is no doubt true, as Sidgwick
on this passage says, that ‘the omission [a not quite accurate’ though under-
standable term] of “‘as’’ is both Latin and Greek usage in comparisons’ ; there
are some instances in Vahlen, Opusc. ac. i. 344, and in my Plautinisches im
Plautus, 51 f., and more in Kock, Com. Ait. fragm. i, p. 28 (on Cratinus 52),
A. Nauck, Bulletin de l' Acad. de St.-Pétersbourg, N.S. i (xxxiii), 435, Shorey,
Class. Phil. iv, 1909, 433 ff., Groeneboom on Herodas 6. 14. But here there is
a strong probability in favour of Aeschylus’ having made the metaphor of
the boy chasing the bird not subordinate but independent: for the details
see below on 399 οἷος κτλ.
394. This passage has been related to the proverbial expression τὰ πετόμενα
διώκειν by Casaubon, who quoted Plato, Euthyphro 4 ἃ τί de ; πετόμενόν τινα
διώκεις ; Stanley illustrated the idea with many instances (the passages
adduced by Wilamowitz, Hermes, lxii, 1927, 287, are contained in Stanley’s
list), to which Musgrave added Eur. fr. 271 N. πτηνὰς διώκεις, ὦ τέκνον, τὰς
ἐλπίδας. Cf. also Plato, Euthyd. 291 Ὁ ὥσπερ τὰ παιδία τὰ τοὺς κορύδους διώ-
κοντα, ἀεὶ ὠιόμεθα... λήψεσθαι, αἱ δ᾽ ἀεὶ ὑπεξέφευγον. The general sense of the
expression to denote vain and fruitless trouble is sufficiently clear. But what
is the relevance ofit here? This question can be answered better in connexion
with the following words (πόλει... dpeprov), the details of which will have to
be discussed first.

™ Because usually in such cases we are concerned with identifications. I have discussed
elsewhere (Plautin. im Plautus, 38 ff.) the distinction between identification and comparison
(or, in the terminology of Aristotle, Rhet. 3. 4 p. 1406> 20, between μεταφορά and εἰκών). The
customary description follows the same line as the scholia, in which we often find λείπει
TO ὡς.

206
 COMMENTARY line 395
395. πρόστριμμα: ‘attritionem’ (Stanley). Rightly explained by Blomfield
as ‘id quod προστρίβεται (infligitur)’, who compares Prom. 329 γλώσσηι
ματαίαι ζημία προστρίβεται. Weil wrongly objects, and states that ‘est vox
propria qua Aeschyli aequales sceleris contagionem dicebant.' Dindorf (Lex.
Aesch.) follows his lead. Headlam on similar lines translates by 'fearful
contamination’. The only instance given by Weil is Antiphon Teiral. 3. 2. 8
ὑμῖν kai οὐ τούτωι τὸ μήνιμα τῶν ἀλιτηρίων προστρίψομαι, but there it is only
the object and not the verb which carries the special meaning, as can also be
seen from Prom. 329, Ar. Knights 5, Demosth. 25. 52, and later examples
(Thesaurus). The passages show that in colloquial language it is customary
to say that one προστρίβει or προστρίβεται something damaging or harmful
upon somebody. We must avoid the temptation of trying to find in πρόσ-
τριμμα a continuation of the idea of τρίβωι in 391! as, e.g., Schneidewin does.
The idea of the bronze vessel is finished with; with the ‘boy and the bird’ a
new thought is introduced, and πρόστριμμα θείς must have been intelligible
to the audience independently of any further reference.
θεὶς ἄφερτον : Wilamowitz’s transposition is by far the least drastic change,
and has rightly been adopted by Headlam, who classes this together with the
numerous passages previously collected by him in C.R. xvi, 1902, 243 ff., in
which the copyist has been led to make transpositions by his desire to
produce a normal order of words.?
πόλει πρόστριμμα Geis: this special meaning of τίθημι, predominant in
facio, which corresponds to it etymologically, is of common occurrence in
earlier Greek. Among the instances of it in Aeschylus, Sept. 46 f. πόλει
κατασκαφὰς θέντες bears a particular resemblance to the present passage. Cf.
also on 1601.
Verrall and Headlam (see his prose translation) have treated the words
émei . . . ὄρνιν as parenthetical, and thereby upset the arrangement of the
whole section. Most editors and translators do not actually break up the
intimate connexion of the sentences from ἐπεὶ to θεὶς ἄφερτον (or whatever
their reading is), but several of them have failed to observe that the participle
is in the aorist (cf. on 785 and 789) and have therefore missed the sense.?
1 Even the same word is not seldom used twice in close proximity without any special
effect being intended. There is a good example in Eum. 568 σάλπιγξ βροτείου πνεύματος
πληρουμένη, 570 πληρουμένου yàp τοῦδε βουλευτηρίου.
2 Wilamowitz (Gótt. gel. Anz. 1914, 109 = KI. Schr. v. 1. 466) has made it highly probable
that a corruption of this kind must be assumed in an epitaph of imperial times from
Miletus: as he supposes, it was the author’s intention that the line should read: στὰς
πρόσθε τύμβου τὴν ἄνυμφον δέρκεο (the violation of ‘Porson’s rule’ has no significance at this
period, see p. 238 n. 2 and cf., e.g., Kaibel, Epigr. 352. 3, 362. 1, 1002. 5), but on the stone
it runs: στὰς πρόσθε τύμβου δέρκε τὴν ἄνυμφον.
3 Similarly in Cho. 396 (κάρανα δαΐξας) many commentators have failed to note the aorist,
and have consequently misunderstood the whole sentence, e.g. Paley : “But is it likely that
Zeus... will ever lay on them a wrathful hand, cleaving the murderers’ heads?’ ; similarly,
e.g., Verrall (as is clear from his punctuation) and Tucker. Weil (1860) has made the
objection to this: ‘Vulgo post daifas interpungitur, quo fit ut Jupiter dicatur manum ad-
movere postquam capita perculerit’; he has put the question-mark rightly after βάλοι,
but his reading is wrong. The correct explanation was given by Wilamowitz (first in his
edition of 1896). He saw that there is an anacoluthon after δαΐξας or, to put it differently,
this is to be taken as a ‘nominativus absolutus’. This was accepted by Headlam in his prose
translation (in the note, however, he expresses some doubt about the construction, and
offers the conjecture βαλών for consideration). Like Weil, Wilamowitz rightly insists on
207
line 395 COMMENTARY
The words are unambiguous: ‘for a boy runs after a winged bird (i.e. [see on
304] pursues the unattainable), after he has brought unbearable harm upon
his city.’ According to the whole context of the passage it must be expected
that this sentence contains primarily a statement about the sinner in
general terms, the sinner whose punishment follows inevitably at the end,
but that a veiled reference can also be felt to Paris. There can be no thought
here of the rape of Helen: Paris has won her, so it is certainly not a case of ra
πετόμενα διώκειν. The point of the thought was rightly seen by Hermann:
‘eius (proverbii) hic haec vis est, ut inania sperasse dicatur Paris, quum se
bello victorem fore credidit’. But, as was pointed out by Ahrens (p. 519),
Hermann laid too much stress on the special reference to Paris, to whom the
general statement is not applied explicitly until the words οἷος καὶ Πάρις.
We may therefore elaborate the sentence thus: ‘The sinner (and a sinner like
Paris) hopes in his childish presumption for the unattainable, i.e. escape from
the consequences of his guilt, after he has brought trouble on his fellow citizens
by his action.” ‘Escape from the consequences of his action’ is not an arbi-
trary interpolation but rather formulates the centralthought which dominates
the whole passage from 387 ἄκος δὲ πᾶν μάταιον to the end of the period at
καθαιρεῖ in 398. Only one point remains to be cleared up. It seems disturbing
to the sense that in such close connexion with the picture of the boy there
should be mention of harm to the city, with which the silly child is certainly
not concerned. The sensitive taste of the modern reader may perhaps be
offended by the intrusion of anything so ponderous as πόλει πρόστριμμα upon
the light-hearted gaiety of διώκει παῖς ὄρνιν. We must, however, remember
that these Aeschylean comparisons are not meant to be taken as independent
units, but that the central thought which they illustrate should be continu-
ously present in the hearer’s mind, and that towards the end, just before the
simile (or alvos) joins up again with the original context, it presses strongly
to the fore. This we have seen in 59, where in the simile itself nothing leads
up to παραβᾶσιν, which nevertheless is indispensable to the main thought.

the force of the aorist : ‘ Satéas praeteritum ad χεῖρα βάλοι referri nequit.’ Sidgwick makes
the quite misleading remark in answer to Weil: ‘It is a mistake to suppose that the aor.
part. is necessarily past ; it simply describes the act, the time being fixed by the sense and
context ; e.g. Od. 5. 374 ἁλὶ κάππεσε χεῖρε πετάσσας᾽. How can we treat on the same footing
as καὶ πότ᾽ dv... βάλοι... δαΐξας a sentence in which πετάσσας stands next to an indicative
of narration in past time? In the latter case, as is well known, the aorist participle need
not denote priority of time, cf., e.g., K. W. Krüger on Thuc. 2. 68. 2 ὀνομάσας (with parallel
passages), Goodwin $ 15o, Kühner-Gerth i. 197 ff. (though some of the examples quoted
there from post-Homeric language do not belong to this group, i.e. those in which the
aorist participle denotes the action from which the action of the main verb results).
Because it is important to define exactly the limitation of the use of the aorist participle
I will mention that Wilamowitz in 1896 (while giving the correct view of Cho. 396) made
an unnecessary concession to the customary interpretation: ‘It might seem natural
to take it "und wann mag wol Zeus zur Vernichtung der Fürsten Hand anlegen ?”,
in which case one would have to regard δαΐξας as attracted in tense to the main verb’.
Such an ‘attraction’ would be admissible if the main verb were ἔβαλε, but not in the
case of πότ᾽ dv... βάλοι. Schadewaldt, Hermes, Ixxvii, 1932, 334 n. 2, wrongly attacked
Wilamowitz's own view of the tense of δαΐξας ; with a misapplication of terms he speaks,
in the case of this subordinate participle, of an 'effective aorist' (in Kühner-Gerth i. 154 the
term is used in a quite different sense). Schadewaldt is also wrong in finding difficulty
about assuming a ‘nominativus absolutus' ; whereas M. Berti, p. 245 (see above on 1. 12),
correctly classes the construction of Cho. 396 with the many similar passages in Aeschylus.
208
 COMMENTARY line 400
πόλει in 395 serves the same purpose: before Paris is mentioned, an element
intrudes into the picture which directs the thoughts of the audience towards
him. Again in the next choral ode, just before the transition from the alvos
to the main train of thought, we find details in 732 ff. which would be out of all
proportion in the alvos taken simply by itself, but are of great significance for
what the αἶνος has to illustrate. Cf. also on 1011 δόμος. In the following lines
(396 ff.) the thought of the boy is entirely dropped, and the evil-doer who was
being spoken of before comes back into his place once more.
396. λιτᾶν δ᾽ ἀκούει μὲν οὔτις θεῶν: again a strong repudiation of the
belief classically formulated in the Homeric Acrat, and already challenged in
69 f. (see note).
397. The reading éwiorpopov should not be challenged. Stanley's rendering
horum auctorem is essentially correct, Paley’s is more exact: ‘the man who
engages in these things’. This is in keeping with the meaning of the rare word
in the one passage in which it occurs before Aeschylus, a 177: Telemachus
says of Odysseus ἐπίστροφος ἦν ἀνθρώπων. This is explained in one version of
the scholion as ἐπίστροφος 6 φροντιστής (‘the man who concerns himself with
something’ : how obvious this explanation was from the point of view of later
Greek is clear, e.g., from [Demosth.] 10. 9 οὐδὲν ἐφροντίσατ᾽ οὐδ᾽ ἐπεστράφητ᾽
οὐδὲν τούτων)... ἤγουν ἐπιστροφὴν καὶ ἐπιμέλειαν ποιούμενος τῶν ἀνθρώπων, CF.
Apollon. Lex. 74. 18 ἐπίστροφος ἐπιστρεπτικός, οἷον ἐπιμελής (in the scholion
and also in Eustathius there is a second explanation giving a passive sense to
ἐπίστροφος). There is no ground for reading ἐπίστροφος and making it refer to
the god, as Heath did, with the rendering 'deus iste, cui horum facinorum
vindicta curae est', and more recently Wilamowitz. It is quite wrong to
quote the gloss of Hesychius (ἐπίστροφος ἐπιστροφὴν ποιούμενος καὶ φροντίζων)
as evidence for this interpretation, for, as is shown by the above quotation,
it is simply a scholion on the passage in the Odyssey. The only doubt one can
have is whether to read τῶνδ᾽ ἐπίστροφον δέ with Weyrauch, or τὸν δ᾽ ἐπί-
στροῴφον τῶν with Klausen, Sidgwick, Headlam, and others. The latter perhaps
is the slighter alteration (τῶνδε intruding as an explanation of τῶν), and,
moreover, one is somewhat reluctant to make a conjectural restoration pro-
ducing a syllaba anceps, even though syllaba anceps in such a place is war-
ranted elsewhere (cf. on 229; for hiatus cf. on 239).
398. καθαιρεῖ: ἃ θεῶν τις can easily be supplied out of οὔτις θεῶν as subject,
especially inside a μὲν... δέ construction. For the positive continuation of
οὔτις Sidgwick compares Hor. Sat. 1. 1. 1 ff. ut nemo . . . contentus vivat, laudet
diversa sequentis (parallels in Orelli-Baiter, ad loc.). But above all we should
compare S. Aj. 482 (with Schneidewin's note), Ani. 263 κοὐδεὶς ἐναργής, ἀλλ᾽
ἔφευγε μὴ εἰδέναι (see Jebb, ad loc., and also on El. 72, 650), Plato, Ref. 2.
366 d τῶν ye ἄλλων οὐδεὶς ἑκὼν δίκαιος, ἀλλὰ... ψέγει τὸ ἀδικεῖν, ἀδυνατῶν αὐτὸ
δρᾶν (Bernhardy, Griech. Syntax, 458).
399. οἷος kai Πάρις: οἷος is the connecting link between the simile, within
which no special reference was made to Paris, and the first part of the ode.
In 49 ff. and 717 ff. the devices by which the similes are embedded in their
surroundings are of a different type, cf. on 737. The passage 394 ff. διώκει
mais... ὄρνιν κτλ. appears in itself not really as a comparison but as a little
alvos, though rather alluded to than narrated explicitly.
400. ἐς δόμον τὸν ᾿Ατρειδᾶν: cf. on 3. It is possible that this particular
4872-2 P 209
line 400 COMMENTARY
passage and the conception that is brought out in it explain to a certain
extent the strange invention of the single home shared by the Atridae. For
here the deliberate intention of the poet is clear. According to the summary
given in the Chrestomathia of Proclus, it was related in the Cypria how
Alexandros éevilera . . . παρὰ Mevelauı,' and this is in agreement with the
later vulgate tradition, [Apollod.] Bibl. epit. 3. 3 ἐφ᾽ ἡμέρας δ᾽ ἐννέα ξενισθεὶς
παρὰ Μενελάωι κτλ. Various considerations may have decided Aeschylus to
deviate from the traditional form of the story. One of his motives was
probably that he felt it important that in the great lawsuit, the Trojan War,
the plaintiff should not be solely or mainly represented by Menelaus, but
that both the brothers should appear equally as ἀντίδικοι of Priam (40 f.).
Agamemnon especially was to become a legitimate avenger of the wrong. So
it may have seemed essential to the poet that the violation of the rights of
hospitality and the stealing of the wife should have been committed ‘in the
house of the Atridae’.?
402. The word κλοπή, which certainly goes back to an earlier date, is found
first in the Agamemnon. Cf. 534, where the same thing is emphasized in
technical language: it is an indication worth noting of the treatment through-
out of the Trojan War as a criminal trial.
403. ἀσπίστορας: the word occurs only here (instead of the older form
ἀσπιστής), an artificial coinage, as in other cases where the tragic poets allow
themselves arbitrarily to append the suffixes -τήρ and -τωρ to noun-stems
(cf. Ernst Fraenkel, Nomina agentis, 1. 22, ii. 28 f.); it serves presumably ‘to
give an archaic colouring to the line’ (Williger, Sprachl. Unters., 49).
ἀσπίστορας κλόνους : paralleled by Pers. 105 f. ἱππιοχάρμας κλόνους.
403 f. The MS reading ἀσπίστορας kAóvovs λογχίμους τε Kal ναυβάτας ὁπλισμούς

1 There is no need here to enter into a discussion of the dwelling-places (Argos, Mycenae,
Sparta) assigned to the Atridae in the various pre-Aeschylean traditions. A useful survey
of the more recent literature on this problem, the difficulties of which are well known, is
contained in an article by W. Ferrari, ‘Studi Stesicorei', Athenaeum, 1938, 5 ff.
2 The idea that the transference of the dwelling-place of Menelaus to Argos was ınade
particularly for the sake of the satyr-play (Wilamowitz, Griech. Tragoedien, it. 302) seems
unlikely. The contention of E. Petersen, Rhein. Mus. lxvi, 1911, 12: ‘Paris’ visit and the
abduction of Helen are quite inconceivable in a house shared by the brothers' stands in
contradiction to the unequivocal words of the text. His rationalis leads him to this
consideration : ‘how would Paris have profited from the absence of Menelaus [which forms
part of the story in the Cypria but not in the trilogy of Aeschylus] if Agamemnon was on
the spot?’ That is just the kind of question that we are not justified in asking here or
anywhere else, because in doing so we are trying to get behind the work of the poet as he
has given it. Aeschylus leaves us quite in the dark about these circumstantial details: he
treats them as irrelevant, whereas the common ownership of the house and its bearing on
Agamemnon's legal claim are, from the poet's point of view, of primary importance. The
foregoing note was written before the publication of Daube's Rechtsprobleme. He has noted
the crucial point (p. 24 n. 53 'Agamemnon is more directly wronged if Paris violates the
law of hospitality towards him too’), but he regards as an incidental fact what was obviously
for Aeschylus a prime motive for his invention of the shared dwelling-place. Moreover,
Daube's account does not make it clear that the fact of the brothers’ residence in the same
city, even supposing it could be ascribed with probability to some branch of the epic
tradition, still does not necessitate the assumption that they lived in the same house. But
it is the latter point that is so strongly stressed in Aeschylus. If we assume that he derived
this particular detail from an epic source (which I think unlikely), then certainly his only
reason for adopting it was that the idea of the common residence was so admirably suited
for the purpose of stressing the common legal claim of the two brothers.
210
 COMMENTARY lines 403 f.

is indefensible. It is supposed to mean ‘clipeatas hastilesque turbas et

navales armaturas’ (Stanley). Hermann laconically remarks: ‘non eleganter
hic re positum est.” He is right. For when a word-group is arranged thus:
first adjectival attribute, then noun, then second adjectival attribute, there
is no example, at all events in Aeschylus and Sophocles (in whose works I
have examined this type of grouping, see below), where the second attribute
is connected by re. On the other hand, there is no justification for Ahrens’s
taking exception to the fact that ' re does not correspond to the καί which
stands close to it’. Cf., e.g., Ag. 326 ff. of μὲν γὰρ ἀμφὶ σώμασιν πεπτωκότες
ἀνδρῶν κασιγνήτων Te, καὶ φυταλμίων παῖδες γερόντων krÀ., where re only
connects ἀνδρῶν and κασιγνήτων, while καί joins a new group of mourners to
those already mentioned,’ Eum. 188 ff. 48 ἀκρωνία λευσμός Te, καὶ μύζουσιν
οἰκτισμὸν πολὺν ὑπὸ ῥάχιν mayevres, 905 f. ἐξ οὐρανοῦ Te, κἀνέμων anyara . ..
ἐπιστείχειν χθόνα. The transposition proposed by Hermann ἀσπίστοράς τε
καὶ Aoyxipous κλόνους is not to be considered, because, as Ahrens shows, we
must regard the undoubtedly sound line of the antistrophe (421) πάρεισι
δόξαι φέρουσαι χάριν ματαίαν as the basis for establishing the metre. Hermann
and Ahrens are in agreement in so far as they are trying to find the defect
not in any disturbance of the words but only of their sequence. This pro-
cedure seems to be well justified in this case, for every individual word in the
group from ἀσπίστορας to ὁπλισμούς seems significant and appropriate to the
context, and the difficulties arise simply from the arrangement of the words.
‘Emendation is limited to a very narrow range by the fact that the metre is
fixed, as we have stated. To imagine a high degree of irregular correspon-
dence (the MS text would make v-- |u-u- in this line correspond to
v—v-|--v- in the antistrophe) is impossible because of the particular
regularity in the correspondence of these pure iambics in the Oresteia (see
Wilamowitz, Verskunst, 293 f.; Denniston in Greek Poetry and Life, Oxford
1936, p. 143). Of course if we follow Verrall (see his 2nd ed. of the Agamemnon,
p. 231 f.), and call in question again slight corrections which have long since
been accepted, such as the addition of re in 195 or the παλιρρόχθοις of 191, we
shall be ready to tolerate submissively any number of irregularities.” Most of
the more recent editors have followed Ahrens’s arrangement of the text (first
published in Franz's edition, 1846): κλόνους τε καὶ Aoyxinovs ναυβάτας θ᾽

! Ahrens assumed that Ag. 327 was corrupt. No parallel for a re which is not connected
with the καί following is afforded by Ag. 1014 f., for in that passage (cf. ad loc.) Verrall’s
punctuation is certainly wrong.
2 The controversial construction of the sentence needs a brief explanation. The three-
fold phrase καὶ ταῦτα γῆθεν ἔκ re ποντίας δρόσου ἐξ οὐρανοῦ re explains the preceding line
ὁποῖα... ἐπίσκοπα and at the same time provides the first and more particular answer to
the question in 902. Then comes a fresh clause: κἀνέμων ἀήματα εὐηλίως πνέοντ᾽ ἐπιστείχειν
χθόνα, in which the acc. and inf. is parallel to the preceding nominal phrase. This is how
Otfried Müller takes the passage (see his translation) and so, e.g., Franz, Wecklein, Sidgwick,
L. Campbell (translation), Headlam, Murray; cf. also E. Kienzle, Der Lobpreis von Städten
etc. (Diss. Basel 1936), 53. Wilamowitz’s punctuation after ταῦτα spoils the arrangement ;
Blass’s alteration τἀνέμων is equally wrong.
3 Verrall found no difficulty in accepting the anomaly, on which stress has been laid
above, of the use of re. He explains: ‘ re couples the adjectives ἀσπέστορας and Aoyxipous,
καί couples ναυβάτας ὁπλισμούς to the whole phrase preceding.’ That is of course the arrange-
ment which the sense demands, but Verrall’s more detailed remarks (Appendix II. 1,
p. 232) show that he has no example for the order ‘first adjective, noun, second adjective, re”.
211
lines 403 f. COMMENTARY

ὁπλισμούς. Apart from the transposition, the only change made here is the
addition of 8, which is particularly easy before the O. But this version suffers
again from an irregular word-order. It is Verrall’s merit to have pointed this
out, remarking (p. 232) on Ahrens’s arrangement: ‘the first re, however it be
taken, is both useless and misplaced’. The former objection is untenable ;
one need only refer to Denniston’s remarks (Particles, 512 f.) on the redundant
use of τε καί in the poets, and to some of the examples mentioned below. But
Verrall’s second criticism is fully justified. Denniston joins him in cautiously
expressed agreement (Greek Poetry and Life, 144 n. 1): ‘Ahrens’ transposition
puts τε in an odd place.’ But this does not go far enough. However, before
we can examine the argument, a word must be said about the structure of the
words in Ahrens’s text. He himself takes together κλόνους ἀσπίστοράς τε καὶ
Aoyxipovs (a different, but very infelicitous, grouping is suggested by Williger,
Sprachl. Unters., 52 n. 1), ‘turmoil of shield and spear’. This is in fact required
by the context. The close connexion between ἀσπίς and λόγχη is obvious,
and, moreover, the expression κλόνους ἀσπίστορας kai λογχίμους (for the moment
we may disregard the particular type of connecting particles) has an exact
parallel in Aeschylus in the phrase (quoted by Schneidewin) Suppl. 182
ὄχλον δ᾽ ὑπασπιστῆρα kai Gopvaaóov λεύσσω. The fact that λογχίμους belongs
not to ὁπλισμούς but to xAóvovs is further shown by the Homeric phrase,
which probably suggested the words here used, κλόνον ἐγχειάων (E 167, Y 319).
Accepting then this interpretation we have to put up with a remarkable type
of word-order in Ahrens's text. If for the moment we disregard re, we may
state that the sequence 'first attribute, noun, καί, second attribute', i.e. in
our Case ἀσπίστορας κλόνους kai Aoyxipovs, occurs indeed only rarely in
Aeschylus and Sophocles, but is sufficiently attested. The normal sequence is
either to put both attributes before the noun as, e.g., in Pers. 33 9 μέγας καὶ
πολυθρέμμων Νεῖλος, Prom. 311 ὧδε τραχεῖς kai τεθηγμένους λόγους, 690 ὧδε
δυσθέατα καὶ δύσοιστα πήματα, Ag. 42 f. διθρόνου Διόθεν καὶ δισκήπτρου τιμῆς,
Eum. 927 f. μεγάλας καὶ δυσαρέστους δαίμονας, etc., or to put the noun first as,
e.g., in Pers. 142, φροντίδα κεδνὴν καὶ βαθύβουλον, 532 f. Περσῶν τῶν μεγαλαύχων
καὶ πολυάνδρων, etc. Similarly with connexion by means of re . . . καί, either
like Sept. 595 σοφούς re κἀγαθοὺς àvrnpéras etc. or like Eum. 862 f. Ἄρη
ἐμφύλιόν re καὶ πρὸς ἀλλήλους θρασύν etc. However, we also find the arrange-
ment noted above as unusual: Ag. 63 πολλὰ παλαίσματα καὶ yvwoflapij, 556
σπαρνὰς παρήξεις kal κακοστρώτους, Cho. 373 μεγάλης δὲ (re Kirchhoff) τύχης
καὶ ὑπερβορέου μείζονα, Eum. 152 ἄθεον ἄνδρα καὶ τοκεῦσιν πικρόν, for Ag. 1481
see note. (Different and, to some extent, easier is a case like S. Ant. 1001 f.
κακῶι κλάζοντας οἴστρωι kai βεβαρβαρωμένωι, because here the grammatical
colon is carried to a conclusion, and then καὶ feBapBapwpévo is added as an
afterthought.) So the only doubtful point remaining is that in the text as
restored by Ahrens we are concerned not with ἀσπίστορας kAóvovs καὶ λογχί-
μους but with ἀσπίστορας kAóvovs τε καὶ λογχίμους. The objection to this
arrangement is a strong one and is really threefold. First of all the placing
of τε after a word-group consisting of noun and adjectival attribute is very
uncommon in earlier Greek, as is pointed out above on l. 229. But the passage
discussed there, Sept. 285, if ἀγγέλους is to be retained, might well be quoted
here in support. The second difficulty is much harder to get over. The result
of Ahrens's arrangement is that the expression ἀσπίστορας kAóvovs as a whole
212
 COMMENTARY line 407
is connected with what follows, whereas what the sense requires (see above)
is that only the epithets, ἀσπίστορας and λογχίμους, should be joined together.
This was Ahrens’s intention too, for he says expressly: ‘I have meant the
construction to be κλόνους ἀσπίστοράς τε καὶ Aoyxipous.’ Thirdly, as the
examples quoted above show, we do find, at least in Aeschylus and Sophocles
(I have not extended my investigation beyond these), the joining of two
adjectival attributes by τε καί certainly in cases where both attributes pre-
cede, or both follow, the noun, but where the noun is placed between the
adjectives I have found no instance of the re καί joining the adjectives. The
coincidence of such peculiarities makes me feel such misgivings about
Ahrens’s transposition, which is in itself attractive, that I do not venture
to accept it. Enger’s transposition (in his revision of Klausen) ἀσπίστοράς τε
καὶ KAdvous λογχίμους does indeed avoid the wrong placing of re after doni-
oropas κλόνους, but there remains the irregularity of the epithets being joined
by τε καί where the noun comes between them, and as the order adopted by
Enger seems altogether very harsh, his conjecture fails to inspire confidence.
After all this I am obliged to let the text of the MSS stand unaltered, though
I cannot consider it to be sound.
ναυβάτης is found again adjectivally in 987 and in a few other instances in
Tragedy.
406. ἀντίφερνον : the only instance in literature.
407. BéBaxev: ‘ οἴχεται, she is gone and out of sight in a moment’ (Paley).
It is not easy to appreciate how it is possible for βέβηκε, the plain ‘resultative
perfect’, to have the qualifying phrase ῥίμφα διὰ πυλῶν added to it, which
obviously does not describe the situation which is brought about (like ἔξω
δόμων and similar phrases added to βέβηκε), but the manner of the action
which has led up to the present situation. It may be doubted whether it is
possible for the result of ῥίμφα διὰ πυλῶν ἔβη to be indicated by the phrase
ῥίμφα διὰ πυλῶν βέβηκε (5. Ant. 766 ἁνήρ, ἄναξ, βέβηκεν ἐξ ὀργῆς ταχύς does not
seem to be exactly comparable; for βέβηκεν ταχύς cf. my note on 605 ἥκειν
ὅπως τάχιστα). But in this passage βέβακεν is also extraordinary in itself.
Elsewhere in the narrative parts of the choruses in the Agamemnon past
events are reported and described throughout in the aorist or the imperfect
(στένει in 712 is a genuine present tense, referring to the time after the destruc-
tion of Troy, and πέμπει in 111 is hardly an instance to the contrary: it
presents in timeless form the πομπή as the theme of the song before the
narrative proper begins). It is true that in the chorus in Suppl. 542-56 the
narrative starts with a series of present tenses (Murray should not have
followed Burges in altering φεύγει in 542), and it is not till 565 ff. that it
continues in the past tense. But is this a valid analogy? Further, is the
assumption allowable that in Ag. 407 the perfect βέβακεν, which elsewhere
denotes the situation brought about, is assuming the function of a present
tense descriptive of the action?? I very much doubt it, and I think it likely
that we ought to follow Keck in writing βεβάκει. For the peculiarity of such
a ‘Homeric’ pluperfect being used by Aeschylus see note on 653, and for
1 E. Phoen. 1329 ἀλλ᾽ οἴχεται μὲν σὴ κασιγνήτη πάλαι takes up 1322: βέβηκ᾽ ἀδελφὴ σή,
Κρέων, ἔξω δόμων.
? This would clearly be different from such a use of the perfect as, e.g., E. ZI. 9 ff. θνήισκει
(Agamemnon) . . . xo pèv. . . ὄλωλεν, Αἴγισθος δὲ βασιλεύει χθονός,
213
line 407 COMMENTARY

βεβήκει itself cf. instances like a 360 ἡ μὲν θαμβήσασα πάλιν oikövde βεβήκει,
‘rursus domum profecta est’ (Wackernagel, Syntax, i. 186). In the Hesiodic
Catalogues (fr. 93 Rz.), in the account of the adulteries of the daughters of
Tyndareus, we read about Helen’s sister Timandra: Τιμάνδρη μὲν ἔπειτ᾽
"Exepov προλιποῦσ᾽ ἐβεβήκει, ἵκετο δ᾽ es Φυλῆα.
ῥίμφα is elsewhere found only in Epic and Pindar. ‘Homer’s ῥίμφα combines
the meanings (cf. Hesychius and Suidas s.v.) of ταχέως and ῥαιδίως, εὐχερῶς '
(Wackernagel, Sitzgsb. Berl. Ak. 1918, 395). Helen walks away with divinely
untroubled swiftness, like Artemis in Pindar's dithyramb for the Thebans
(Pap. Oxy. xii. 1604) τό ῥίμφα δ᾽ εἶσιν Ἄρτεμις κτλ. Though she is ἄτλατα
τλᾶσα, that does not weigh heavily upon her nor retard her steps. The
Helen of Euripides (1528) leaves the palace of Theoclymenus ἁβρὸν πόδα
τιθεῖσα.
408. We cannot make out whether Triclinius found πολλὰ δ᾽ ἔστενον in his
exemplar, or whether it is a conjecture of his own, based on πολὺ δ᾽ ἀνέστενον,
with the object of making the metre correspond syllable by syllable with that
of the antistrophe (425) ; the latter is much more likely as it would be quite in
keeping with his general practice. However that may be, πολὺ δ᾽ avéorevov
is wrong. For Aeschylus, who very commonly uses the adverbial πολλά (546
πολλὰ ἀναστένειν), uses πολύ only in connexion with comparative and super-
lative expressions, in which we may include Cho. 1052 un φόβου νικῶ πολύ,
where νικῶ is equivalent to ἥσσων γενοῦ. A further objection to πολὺ δ᾽
ἀνέστενον is the extreme rarity, to say the least, of a correspondence like
vuu- =-—v-— in the lyric iambics of the Oresteia, cf. p. 351.
409. δόμων προφῆται (for the meaning of προφῆται cf. on 1099) is compared
by Blass on Cho. 37 f., and others before him, with the κριταὶ ὀνειράτων of
that passage: ‘any kind of trustworthy men called in in such cases’. (Beazley
tells me that there is often a servant in Greek houses nowadays who is known
to have the gift of interpreting dreams, and who is in consequence regularly
consulted on such occasions.) For δόμων προφῆται together cf. Cho. 32 δόμων
ὀνειρόμαντις. Cf. also G. Thomson, C.Q. xxx, 1936, 105. E. Petersen, Die
attische Tragödie, 243, discusses the nameless seers who are occasionally
found in the place of those definitely named, such as Calchas and Teiresias,
uttering warnings to powerful princes. Cf., too, O. Kern, Rel. d. Griech. ii. 140,
for ‘the great number of soothsayers, seers, and prophets who from about the
seventh century rule the streets and even make their way into the houses of
the great, often influencing the decisions of statesmen and commanders’.
Obviously the prophets’ word in this case is just as infallible as in the speech
of Calchas in the parodos: what is prophesied here (415 δόξει) actually comes
to pass.
410. πρόμοι: cf. on 200.
! It is regrettable that Porson's striking emendation is nowadays rejected by his country-
men in favour of the μὴ φοβοῦ, νικῶν πολύ of the MS. I have just put forward a linguistic
argument for the necessity of the alteration. Even without such an argument a reader who
is sensitive to the tragic force of this powerful scene might feel that so crude an expression
as μὴ φοβοῦ, νικῶν πολύ is intolerable in this passage, especially after Orestes (1023) has de-
scribed himself as νικώμενος. The parallelism with Zum. 88, where Apollo is encouraging the
man tormented by the Furies, μὴ φόβος ce νικάτω φρένας, is surely not accidental. In the
list of examples from Tragedy of νικᾶσθαί τινος in L-S s.v. II. 3 the oldest is not given, A.
Suppl. 1005.
214
 COMMENTARY line 412
411. ἰὼ λέχος καὶ στίβοι φιλάνορες: on the whole correctly taken by the
older commentators (Stanley, Pauw, Schütz, Bothe, etc.) to refer to Helen
and Menelaus. This is suggested by the parallelism with the preceding line,
and also by the most obvious interpretation of φιλάνωρ, which means, in both
the other passages where it occurs in Aeschylus, Pers. 136 and Ag. 856,
‘husband-loving’ (cf. O. Hoffmann, Glotta, xxviii, 1939, 39). Klausen’s in-
felicitous interpretation ‘gressus amatorii, fuga Helenae cum Paride’ (simil-
arly taken by Wilamowitz: “Weh die Frau floh mit fremdem Manne’) has
been attacked by Ahrens (p. 521). He also objects to the rendering of orißoı
proposed by, e.g., Schütz (and similarly by Verrall and Headlam: ‘Alas for
the husband's bed and the wifely print of her form upon it’)': a misrepresenta-
tion because it refers to a different type of ‘vestigia’. Actually, as Ahrens
says, ' orißoı can only mean steps, paths, or footmarks’. But he is not quite
right in understanding στίβοι diAdvopes to mean the entry into the bride-
groom's house of the loving bride on the wedding-day. A closer relation to
λέχος is needed, just as in the preceding line the πρόμοι are the πρόμοι of the
house. orißos serves among other things as the ‘noun of action’ for στείβειν,
and it is also used to indicate the effect of oreiBew, so occasionally it means
‘treading’ or ‘tread’. S. Phil. 29 καὶ στίβου γ᾽ οὐδεὶς κτύπος, 206 φθογγά του
στίβον κατ᾽ ἀνάγκαν ἕρποντος. The στίβοι φιλάνορες in association with λέχος
therefore mean the steps with which the loving wife approaches the bed of
her husband.? It conveys the thought not so much of eisavaßaivew itself
(Θ 291 ἠὲ γυναῖχ᾽, ἥ κέν τοι ὁμὸν λέχος eivavaßaivoı, Hesiod, Theog. 939, Hymn
io Aphrodite 161 λεχέων εὐποιήτων ἐπέβησαν), as of the moving towards the
bed, from which the πλᾶτις derives her name (cf. Ernst Fraenkel, Nomina
agentis, i. 42; he might have quoted A. Prom. 897, E. Andr. 25}: cf., e.g.,
Hdt. 3. 69. 6.
412. πάρεστι governs the infinitive as in 368, Sept. 922, Cho. 977 (wrongly
altered by Wilamowitz), Eum. 160, 167. The basis for the restoration of what
follows was provided by Hermann. He recognized ovyas as accusative,
restored to dripos ἀλοιδορος the endings in -ovs? required by sense and metre,
and similarly restored the long second syllable to ἀφεμένων. Here he read
ἀφειμένων (for the frequent confusion of ἀφέμενος and ἀφειμένος cf. Cobet,
Nov. lect. 642) with this interpretation : ‘ ἀφειμένα sunt haec ipsa, quae turpiter
missa fecit Helena, domum, torum, maritum.’ The reasoning of Ahrens
(p. 522) against this interpretation is weak, his assertion that ‘ ἀφειμένος
cannot possibly mean desertus’ is wrong: cf., e.g., Hdt. 1. 166. 3 ἀπέντες τὴν
Κύρνον ἔπλεον ἐς '"Pijyvov, S. Phil. 486 f. ἀλλὰ μή u^ ἀφῆις ἔρημον οὕτω, E. Hec.
115 ποῖ δή, Δαναοί, τὸν ἐμὸν τύμβον στέλλεσθ᾽ ἀγέραστον ἀφέντες; (the passages
in Dindorf’s Thesaurus). Hermann’s interpretation is probably correct, with
the slight modification (by Franz and others) of taking ἀφειμένων as masculine.
One must beware of taking ἀφειμένων as middle. It is not true that Menelaus
1 This interpretation may have been influenced by unconscious reminiscence of such
' passages as Prop. 2. 29. 35 f. apparent non ulla toro vestigia presso . . . nec tacutsse duos and
the like.
2 Since writing this, I have found that Lewis Campbell translates : ‘Alas! for the couch,
whither she came lovingly to meet her lord!’ So he has taken the passage in the same way
as I have.
3 Cf., if it is worth while, e.g. Pind. N. 1. 24 EZAOZ correctly interpreted by Aristarchus
as ἐσλούς.
215
 line 412 . COMMENTARY

lets Helen go of his own accord. Moreover, Jebb observes, on S. Ant. 1165,
that the forms of the perfect ἀφεῖμαι etc. are always used only in the passive
sense: the examples in the Thesaurus are numerous, and as regards Demosth.
23. 157, Cobet’s correction is rightly adopted by the more recent editors.!
Perhaps, however, besides Hermann’s proposal we should also consider
Dindorf’s ἀφημένων, which has been accepted by Wecklein (in his com-
mentary), Sidgwick, and Headlam. In its favour is the single passage where
the word occurs, Ο 106, where Hera says of Zeus: ὁ δ᾽ ἀφήμενος οὐκ ἀλεγίζει
οὐδ᾽ ὄθεται. One can easily imagine that Aeschylus might have borrowed this
vivid description of such unconcerned ‘sitting apart’ and used it to describe
᾿ an indifference of quite another kind. Yet it is doubtful whether ἀφημένων is
suitable here. What the Homeric Zeus is doing, at least according to the
statement in Hera’s inflammatory speech, is perfectly clear: he is sitting
apart from the assembly of the gods, which his wife is addressing. The phrase
which Headlam quotes for comparison from Hdt. 4. 66, ἠτιμωμένοι ἀποκατέαται,
also describes those who are sitting apart from the main body of the rest, who
are taking part in the warriors’ carousal. In this passage, however, there is
no mention of any circle of people from which Menelaus has isolated himself,
and only by way of fanciful elaboration would it be possible to infer it.
Quite apart from this, ἀφημένων as the only description of Menelaus in this
sentence would be strangely indefinite, while ἀφειμένων, after the words
Beßareı κτλ. (407) have just been used of Helen, is quite unambiguous.
Perhaps we may detect in this only slightly disguising plural some tactful
consideration for their master on the part of the δόμων προφῆται. In any case
it seems that we have here an instance of that special type of the ‘general’
plural which is occasionally called ‘allusive’ (see on 1618), cf., e.g., Cho. 886
τὸν ζῶντα katvew τοὺς τεθνηκότας λέγω (where in the riddling words—cf. 887
ἐξ aivıynarwv—the veiled allusion is particularly significant), S. Ant. 1263 f.
ὦ Kravovras re καὶ θανόντας βλέποντες ἐμφυλίους, and possibly, too, E. Her.
1309, where Herakles speaks of himself as τοὺς εὐεργέτας “Ελλάδος after the
preceding words (1306) ἄνδρ᾽ ᾿Ελλάδος τὸν πρῶτον. For the plural oıyas we may
recall E. Bruhn’s remark, Anhang zu Soph. ὃ 3. ii: ‘the plural is often favoured
by the connexion of the word concerned with another which is in the plural’:
this applies here to the connexion with ἀφειμένων.
| A ἀτίμους ἀλοιδόρους : the explanation ‘transferred from the person who is
keeping silence’ (Wecklein) does not quite do justice to the peculiarity of the
expression. Certainly the word ἄτιμος may be applied to Menelaus, as his
wife has been stolen from him. But in the ‘prophets’ ’ description the stress
is laid not so much on the husband himself as on his amazing state of mind
at the time. He is completely under the spell of silent stupefaction. This
silence finds no place for abusive language, or for any word calling for
vengeance or hinting at measures of any kind. From this kind of silence
τιμή, and the thought of it, is very far removed; hence the silence itself is
called devoid of τιμή. Clearly the word is here again taken in the sense in
1 The supplement suggested by Wilamowitz in Menander, Epitr. 585 (649 Koerte, 3rd ed.),
[ἀφ]ειμένον (middle, see his commentary, p. 104), is very uncertain. Jacobi's Index to Meineke's
Comici yields Alexis fr. 219 Kock, 1. 12, τό ἀφειμένους, ἀφεῖται, both passive, and Poseidippos
fr. 23 (iii, p. 342) K. ἀφεῖσαι (where Dobree’s restoration of the ending seems certain),
passive.
216
 COMMENTARY line 412
which Menelaus uses it when he says (P 92) Πάτροκλόν θ᾽, ὃς κεῖται ἐμῆς Ever’
ἐνθάδε τιμῆς (cf. also A 159).
It would be mere hair-splitting to object that the parallelism in ἀτίμους
ἀλοιδόρους is incomplete, because ἄτιμος has to be taken in a passive and
ἀλοίδορος in an active sense. They are held closely together by the privative
prefix : in such cases (cf. on 238) we need not mind about active and passive,
but feel that ‘without honour’ and ‘without reviling’ are privatives belonging
to the same category. The beginning of the word which has been corrupted
into dówros must also be the privative d-, as has been rightly recognized
by Hermann (appendix to Humboldt’s translation), W. Sewell, Karsten,
Dindorf, Schwerdt, and Blass, Hermes, xxix, 1894, 634; and these have been
followed in the essential point by Wilamowitz. The use of the privative τρί-
κωλον is very old and widespread, and perhaps we can find one of the roots
of it in such phrases as ἀφρήτωρ, ἀθέμιστος, ἀνέστιος (I 63), of which the
form and content are associated with solemn imprecations.' It will suffice
here to quote from Aeschylus Ag. 769 ἄμαχον ἀπόλεμον ἀνίερον, Cho. 54
ἄμαχον ἀδάματον ἀπόλεμον, as well as S. Ant. 876 ἄκλαυτος ἄφιλος ἀνυμέναιος,
1071 ἄμοιρον ἀκτέριστον ἀνόσιον νέκυν, Oed. C. 130 f. παραμειβόμεσθ᾽ ἀδέρκτως
ἀφώνως ἀλόγως TÓ . . . στόμα... ἱέντες, 1236 f. ἀκρατὲς ἀπροσόμιλον γῆρας
ἄφιλον, E. Andr. 491 ἄθεος ἄνομος ἄχαρις 6 φόνος, Ar. Frogs 838 ἀχάλινον
ἀκρατὲς ἀπύλωτον στόμα, and in more comic form ibid. 204 ἄπειρος ἀθαλάττωτος
ἀσαλαμίνιος, Phrynichus Com. ap. Phot. Berol. p. 118. 25 ἄσιτος ἄποτος
ἀναπόνιπτος. The figure is one which invites imitation: in tragic language,
e.g. Paradise Lost, ii. 185 unrespiled, unpitied, unreprieved (ibid. v. 895 f. in
quadruplet form, unmoved, Unshaken, unseduced, unterrified), in comic (with
παρατραγωιδεῖν of Shakespeare's line in Hamlet, i. 5. 77 unhouseled, disappointed,
unaneled), L. Sterne, Tristram Shandy, chap. 35 unwiped, unappointed, un-
annealed. The conjecture driorws? is so easy and, when given the proper
ending, so suitable for the sense, that we may believe we have thus recovered
the genuine reading. ἄπιστος, ‘unbelieving, distrustful’, as here, is found in
the formula in the Odyssey θυμὸς δέ τοι αἰὲν ἄπιστος. Thus the meaning of
‘silence without honour, without reviling, without belief’ is (in our own more
pedantic manner of speaking) ‘a silence in which there is dishonour, in which
no word of reviling is uttered, in which one (i.e. the person keeping silence)
cannot believe what has happened’. The explanation given here shows my
reason for preferring Hermann’s emendation (or rather interpretation of the
MS reading) ἀτίμους ἀλοιδόρους to that of Dindorf and others which is just
™ There is a much later instance (IV/III cent. B.c.), but conceived in the same spirit, in
rhe curse recorded on the Attic lead tablet, Dittenberger, Syll. 1175 1. 10 dywpa καὶ ἄμοιρα
καὶ ἀφανῆ αὐτῶι ἅπαντα γένοιτο. Cf. ibid. I. 28f. R. Hirzel, ‘Die Strafe der Steinigung’,
Abh. Sachs. Ges. d. Wiss., phil.-hist. Kl. xxvii, 1908, 253 n. 7, observes: ‘In Hom. II. 9. 63 in
the words spoken against the man who stirs up civil discord, ἀφρήτωρ ἀθέμιστος ἀνέστιός
ἐστιν ἐκεῖνος, I cannot help recognizing an ancient formula of proscription or of permanent
and complete banishment, only, as happens elsewhere too, it is proleptically transformed
from a wish or curse into an affirmation. “He is tribeless, hearthless", instead of “he
deserves to be. ..."' This coincides exactly with my own assumption. For the asyn-
detic tricolon in general cf. F. Leo, Analecta Plautina, iii (Göttinger Univ.-Programm,
1906), 6 ff.
2 This conjecture (apart from the vowel of the final syllable) was made before F. I.
Schwerdt (De nova Aesch. Ag. recensione, Bonn 1860, 20) by G. Hermann, De metris poetarum
Graec. εἰ Rom. (1796) 432; afterwards he rejected it.
217
line 412 COMMENTARY

as easy, viz. ἀτίμως ἀλοιδόρως. Wilamowitz also preferred ἀτίμους κτλ. in

1885 and in the appendix to his translation, whereas later he wrote ἀτίμως κτλ.
I fail to see what would be the meaning of ἀλοιδόρως ἀφειμένων.
415. béopa:i.e. Helen’s (as Triclinius took it and later on, e.g., Stanley, Pauw,
and Heath). This, the simplest and most obvious explanation, has always
been followed by men with a sensitive feeling for poetry, e.g. le Père Brumoy,
W.v. Humboldt, Klausen, Welcker, Bamberger (Opusc. 48), Sewell, Conington,
Paley, Dean Milman, Lewis Campbell, Wilamowitz, Platt,’ Murray, and others.
But prosaic precision makes its protest: ‘nam etsi Menelai animo obversari
potuit ac debuit Helenae absentis imago, nullo modo tamen ista imago
animo et cogitatione concepta videri potuit aedibus imperare! Immo potius
h.l. φάσμα, ut alibi εἴδωλον, notat hominem adeo macilentum, adeo ex-
sanguem, ut non vivum corpus, sed umbra potius aut spectrum esse videatur'
(Schütz) ; similarly φάσμα is taken as referring to Menelaus by, e.g., Blomfield,
Scholefield, Ahrens, Nágelsbach (expounding in detail in 'Quaest. Aeschyleae',
Progr. Erlangen, 1858, p. 9, 22), Wecklein, Verrall, Plüss, Headlam. That
φάσμα by itself with no qualifying word could adequately indicate the man
who has become a 'shade' of his former self is almost inconceivable. Nor do
the 'parallels' quoted by the commentators seem to be pertinent. In one of
them Philoctetes (946 f.) says about himself κοὐκ οἶδ᾽ (sc. Neoptolemus)
ἐναίρων νεκρὸν ἢ καπνοῦ σκιάν, εἴδωλον ἄλλως, and in the other Oedipus (Oed. C.
109 1) οἰκτίρατ᾽ ἀνδρὸς Οἰδίπου τόδ᾽ ἄθλιον εἴδωλον" οὐ yap δὴ τόδ᾽ ἀρχαῖον
δέμας, but the difference is marked in both passages by the fullness of the
expression and also by the use of εἴδωλον. εἴδωλον ‘image’ can, it is true, refer
to the same matter as φάσμα, but the two words reveal entirely different
aspects. After the departure of his beloved, whether she has died or been
carried off, the bereaved husband is disturbingly confronted by her ‘appari-
tion’ in every room of the home, every time he looks at bed and chair, every
time he meets the members of the household. If anyone cannot appreciate
this either from personal experience or the power of imagination, then even
the intense pathos of lines like E. Alc. 944 ff. will be lost on him. It is said of
the φάσμα: δόξει δόμων ἀνάσσειν, because the apparition is taking over the
role formerly assigned to the actual person: Hermione, e.g., is called δόμων
ἄνασσα in E. Andr. 897. Should any further indication be needed of what
Aeschylus meant by the φάσμα, we could get it from the fact that after the
apparition that seems to rule over the house in the daytime (φάσμα δόξει... .),
we have in 420 f. the description of dverpddavror . . . δόξαι following in the closest
conceivable connexion of thought and language.
The question so often raised (e.g. by Welcker, Rhein. Mus. ix, 1854, 194 f.),
whether Aeschylus had Stesichorus’ εἴδωλον of Helen in his mind here, can
hardly be answered. In any case, the thought of such a ‘reference’ must not
be allowed to blind us to the complete dissimilarity : in Stesichorus we have a
peculiar and mythical miracle, here a piece of general and purely human
experience.
416. κολοσσῶν : probably a pre-Greek word, and a stranger to Attic as far as
is known to us (this was pointed out by Blomfield). The development of its
meaning has been outlined by Wilamowitz in Ferri, ‘Alcune iscrizioni di
ı J. Phil. xxxii, 1913, 51: ‘The old notion that the φάσμα may be Menelaus himself,
which has been recently revived, seems to me quite ridiculous.’
218
 COMMENTARY line 418
Cirene’, Abhdlg. Preuss. Akad. d. Wiss. 1925, phil.-hist. Kl. No. 5, p. 39, and
Berl. Sitzgsb. 1927 (‘Heilige Gesetze’), p. 169. κολοσσός is used, in the oath of
the Theraeans, of a wax figure ‘which must have been anything but “colossal” ’,
and, generally speaking, the notion of exceptional size has no place in the
original meaning of the word. So, e.g., Herodotus says κολοσσοὶ μεγάλοι if he is
concerned with great size, or adds the measurements. Klausen remarked on
Ag. 416: ‘significatio magnitudinis quae humanam excedat (Etym. M.) serius
est recepta, Wilamowitz has shown that it was only the reputation of the
gigantic statue of Helios by Chares of Lindos which, not in fact till the late
Hellenistic period, brought about the particular meaning of ‘colossus’. The
results of Wilamowitz's investigation have been summarized by E. Ben-
veniste, Revue de philol. 1932, 119 ff., who also deals with the pre-Greek
origin of the word and the beliefs revealed in the customary use of such
effigies; further information on the subject is to be found in Lauffer, Ath.
Mitt. Ixii, 1937, 94 f. In the passage before us we should not have in mind
small figures of wood or wax or clay, but ordinary statues ‘like those of the
Attic κόραι. Perhaps Aeschylus transferred to the house of the Atridae that
abundant array of female statues which was to be seen in the sanctuaries of
his own day and city.’ There is no need to think of portraits of Helen (as is
considered possible by Schiitz, and taken as a fact by Ch. Picard, Rev. de
phil. lix, 1933, 351 f.). Every statue of a beautiful woman’ is more than the
deserted husband can bear, for it reminds him of her whom he has lost. It
does not matter whether only one such statue is there or several: the text
does not say definitely, for the plural can have a ‘general’ significance here in
a sentence which does indeed refer principally to Menelaus, but all the same
(cf. the words following) approximates closely to a generalization. There is
hardly anything in common between the practice here supposed and the
statue-motif which was introduced by Euripides as a definite dramatic
element in the story of Protesilaus and Laodamia (cf. Wilamowitz, Stud. It.
N.S. vii, 1929, 91 f. = KI. Schr. v. x, 525 ff.) ; nor is there any connexion with the
motif of the statue of the dead wife as it appears in the lines E. Alc. 348 ff.,
where perhaps the poet utilized a passage from his own Protesilaus, as is
supposed by Wilamowitz (loc. cit. and introduction to his translation of
Alkestis, Griech. Tragoedien, iii. 91 n. 1).
417. éx8era χάρις: ‘di tutt’ altre cose, qual mi torse più nel suo amor, più
mi si fe’ nimica’ (Dante, Purg. 31. 86 f.).
ἀνδρί remains to some extent indefinite: on the one hand it refers back to
the particular husband (Menelaus) to whom the preceding lines applied and
the word φιλάνορες alluded in 411 ; on the other hand it denotes, as ἀνήρ often
does, ‘man’ or ‘husband’ in a quite general sense. This is in keeping with the
whole character of this ‘refrain’ 416-19: the particular case (indicated, it is
true, only in veiled terms) is extended so as to suit a wide application.
418. aynviats: for the prosody (conflicting with ἄχήν, ἠχήν) of this word,
' E. Langlotz, in the introduction to his publication of the Korai of the Acropolis (1939),
p. 9, observes: ‘Most of them (the Korai) belong to the last three decades of the sixth
century, especially to the years before 500 [the period of Aeschylus' youth], while they
become scarce again after that date.’
? Itis entirely arbitrary to render, as Picard does, loc. cit., εὐμόρφων by ‘exact’ (cf. on
454) ; moreover, he attributes to the poet features of primitive civilization of which nothing
is to be found in the text.
219
line 418 COMMENTARY

adopted from another dialect, see Wackernagel, Vermischte Beiträge, 17 f.

For the plural cf., e.g., 726 γαστρὸς ἀνάγκαις.
ὀμμάτων δ᾽ ἐν ἀχηνίαις : Schütz’s far-fetched explanation (‘the statues are
eyeless, and cannot therefore inspire love’) still finds supporters (e.g. L-S s.v.
ἀχηνία) ; it was rejected by Klausen and Hermann, who says: 'simplicius est
dici adspectu Helenae deficiente omnem voluptatem cessare'. They under-
stand ὀμμάτων to be the eyes of Menelaus (as did Butler, and later, e.g., Ahrens,
Kennedy, Nägelsbach, Wecklein, Wilamowitz), while Heath (followed by,
e.g., Sidgwick and Headlam) makes it refer to Helen. The former seems to me
far the better, as providing a much stronger connexion with the immediately
preceding passage: the statues of young women are hateful to him, because
his eyes are starved with longing for the real Helen, and he cannot bear to
look at the statues which for all their loveliness are not the beloved herself.
Connexions with Greek physiology such as Donaldson (quoted in Paley) and
Headlam have tried to see here should not be read into the passage. To refer
ὀμμάτων to Menelaus and therefore to take it as subjective genitive is not
ruled out by the fact that in Cho. 301 (and Ar. fr. 20 K.) ἀχηνία governs an
objective genitive. Verrall, by way of compromise, suggests that we may see
an intentional vagueness in the expression, in which the eyes of both Menelaus
and Helen are equally concerned. His interpretation of the passage shows
delicacy of feeling, but does not seem necessary.
419. ἔρρει: this word 'for "going away" is mostly used with the idea of
"going to destruction" ' (Wilamowitz on Ar. Lys. 335); so epperw usually
expresses a curse (Wilamowitz, Berl. Sitagsb. 1927, 14): cf. also K. Latte,
Heiliges Recht, 63.
πᾶσ᾽ ᾿Αφροδίτα: there is an unmistakable connexion with the preceding
χάρις, as there is a close relation between ἐπίχαρι and ἐπαφρόδιτον ; in the
former phrase the meaning was not so much that of xapıs in general as of
the charm of lovely statues of women: and here the same idea is continued.
The train of thought in this passage leads us to think above all of the primary
activity of the great goddess: thus Headlam ‘all passion of love is lost”.
420. ὀνειρόφαντοι: for the resumption of the idea expressed earlier cf. note
on 415 φάσμα. With regard to the connexion with the words immediately
preceding (εὐμόρφων δὲ κολοσσῶν κτλ.) we find a similar sequence of thought
in E. Alc. 354, which is due not to any literary dependence on this passage
but to the nature of human experience: there Admetus has been speaking
of the statue of Alcestis (348 ff. σοφῆι δὲ χειρὶ τεκτόνων δέμας τὸ σὸν εἰκασθὲν ἐν
λέκτροισιν ἐκταθήσεται, ὧι προσπεσοῦμαι κτλ.) and he continues: ἐν δ᾽ ὀνείρασι
φοιτῶσά p! εὐφραίνοις ἄν.
πενθήμονες : Housman’s elegant conjecture πειθήμονες (J. Phil. xvi, 1888,
259 f.), which had already been considered by Karsten, was put into the text
by Wecklein, Weil, Verrall, Headlam, and others. “Visions which persuade
him that they are Helen’ would give an attractive sense here. The conjecture
gains force from ὀνείρων φάσματ᾽ εὐπειθῆ in 274 (Headlam). It need be no
objection that the word is only found otherwise in Nonnus and his successors,
for they may have been following an earlier example; nor would the case be
1 ἔρρει πᾶσ᾽ ᾿Αφροδίτα might also be applied to one who was trying (which is far from
being the case here) to console himself elsewhere for the loss of his beloved : his experience
would in that case be admonuit dominae deseruttque Venus.
220
 COMMENTARY lines 425 f.
different with the πενθήμων of the MS text, which is not found elsewhere
earlier than in Nonnus’ imitator Christodorus (Ecphrasis 148). But whatever
can be said in favour of πειθήμονες, I do not think that the doubts expressed
about πενθήμονες justify us in regarding the word as corrupt. Of course
Housman is right in objecting to the interpretation luctificus (Dindorf)
‘causing sorrow’, which seems to go back to Schiitz. But the older translation
luctuosus (Stanley, cf. L-S ‘mournful’) should not be dismissed by para-
phrasing it ‘of sorrowful aspect’ and then ridiculing it. πενθήμων can appar-
ently express a variety of relations to πένθος or πενθεῖν (for the type of the
suffix and its functions cf. P. Chantraine, La Formation des noms en grec,
p. 173 f., Buck and Petersen, A Reverse Index, 216) ; therefore we may at least
be allowed to ask whether it does not here mean in a quite general sense
‘mournful, connected with, or belonging to, mourning’. If this is right, then
δόξαι πενθήμονες are fancies that usually come into the mind of one who is
mourning. This makes excellent sense. Perhaps ὀνειρόφαντοι here also in-
fluences the meaning of πενθήμονες, as, in Aeschylus and elsewhere, parallel
attributes not uncommonly affect each other's shades of meaning: 'fancies
and visions which appear in the mourner's dreams'. It is easily understood
that the πένθος is a contributory cause of the dreams. Therefore we shall not
be too ready to give up πενθήμονες, especially as the word sums up so con-
cisely the state of the ἀφειμένοι, and seems to connect the beginning of the
antistrophe very closely with the end of the strophe.
421. δόξαι: ‘imaginings’, ‘figures born of fancy’, cf. on 275.
χάριν: χάρις is here, as often, what somebody welcomes, what delights,
gratifies him ; similarly in 550, 786.
423. To put the construction right, many editors have adopted Scholefield's
ὁρᾶι. Against this Headlam rightly argues that one does not say δοκῶν ὁρῶ,
but δοκῶ ὁρᾶν, cf., e.g., Pers. 188. Hermann keeps the MS reading, but his
explanation 'ne idem verbum repeteretur, omissum est ὁρᾶι. Plena enim
oratio esset εὖτ᾽ dv ἐσθλά τις δοκῶν ὁρᾶν ὁρᾶι ' is too artificial to be convincing.
Matthià, Griech. Gramm. $ 556 n. 2, rightly recognizes in this passage an
anacoluthon of the participle ; so does Berti in the essay quoted on 12, p. 241
(bringing in unnecessarily the ‘Nominativus pendens"). Cf. on 1511. Schneide-
win, P/niol. ix, 1854, 137, and Headlam find ‘dramatic effect’ here in the
omission of a finite verb, as expressing the sudden breaking off of the dream
before the fulfilment of what the dreamer desires; this seems probable.
After the beginning 'for in vain, when fancying that he beholds good' one
expects the apodosis to be continued by something like 'he attempts to
catch it'. But after the sentence has been interrupted, μάταν is no longer
taken into account.
tis: here we find expressed the wider scope of the thought for which we are
already prepared (cf. on 417): what is said here does not apply only to the
particular case of Menelaus. For the same reason the object of the verb is
not 'his absent wife' or the like, but the general tcrin ἐσθλά, Housman
(J. Phil. xvi, 1888, 260) did not realize the very good reason (cf. on 429) for
what he called 'drifting off into a generalization'. Therefore his conjecture
és θιγάς, which has been put into the text by Murray and (with the slight
variation es Aaßas) by A. Y. Campbell, mars the sense of the whole sentence.
425 f. οὐ μεθύστερον . . . κελεύθοις. As the MS text stands, οὐ μεθύστερον
22I
lines 425 f. COMMENTARY

can only be connected with Beßarev. Accordingly in F a comma is put after

μεθύστερον, and in Tr the gloss ἀλλὰ παραχρῆμα is written above οὐ μεθύστερον.
That the phrase should have this meaning seems strange. But, in their
desperate efforts to fit οὐ μεθύστερον into the sentence somehow or other, the
commentators have had recourse to such renderings as ‘haud ita post’
(Stanley), ‘visum illud e manibus elapsum paulo post discedit’ (Schütz), ‘non
postea vel sero, sed statim’ (Bothe, in complete agreement with Triclinius),
and soon. This has been the prevailing interpretation right up to the present
time, cf., e.g., Enger (in his revision of Klausen) and Wecklein ‘ilico’, L-S
(s.v. μεθύστερον) ‘in a moment’, Verrall ‘that instant’, Headlam ‘is gone a
moment afterwards’, Platt ‘is gone in a moment’, etc. Here it is not only
the distortion of the meaning of οὐ μεθύστερον that is objectionable, but also
the addition of an adverb like ‘in a moment’ to what is not an aorist, but an
unequivocal perfect, viz. βέβηκε ‘it is gone’ (cf. on 407).
But there is more here that is curious. Quite apart from the difficulty of
finding a word to which οὐ μεθύστερον could be related, the phrase πτεροῖς
énadois ὕπνου κελεύθοις is far from clear. For this reason the conjecture
ὀπαδοῦσ᾽, which seems to do away with both difficulties, has had a considerable
success. It was proposed by Dobree (Adversaria, ii. 23, with the explanation
‘non amplius somni comes’) and independently of him by J. Franz, who
translates: ‘das nicht zum zweiten Mal befliigelt nachschwebt dem Gang des
Schlummers’, and by Hermann, who paraphrases by ‘non rediturum posthac
cum somno’. The passage was similarly treated by, e.g., Kennedy, Davies,
Pliiss, Wilamowitz (translation). But certain considerations tell against this
reading. It is perhaps not a decisive objection, though well worth noticing,
that the verb ὀπαδεῖν occurs nowhere in Tragedy (the ‘versus nondum emen-
datus'—this is Nauck's verdict—in a fragment of quite uncertain origin, Fr.
trag. adesp. 493 N., may be disregarded), while ὁπαδός occurs a few times in
Aeschylus, frequently in Euripides, and is not missing in Sophocles either.
As regards this point, it might be argued that Aeschylus was at liberty,
especially in odes of lofty style, to take over a verb from epic diction. There
is, however, a real difficulty about the tense. Hermann translates, as the
sense seems to require, non rediturum. But this meaning cannot be got out of
ὁπαδοῦσα. Lastly, there is a serious difficulty about making κελεύθοις (or
κελεύθων as others write it) depend on ὀπαδοῦσα or ὀπαδοῖς. The object of the
'following, accompanying, attending', indicated by the dative or the genitive
with these words, is invariably a person (or thing) with whom (which) one
goes or is, whose companion one is, etc.: no word belonging to the category
of ó8ós seems to occur in this connexion. I do not regard it as impossible for
Aeschylus to have ventured on so bold an extension of usage, but I do not
find sufficient grounds for assuming that he has done so in this case. On the
other hand, the expression πτεροῖς ὁπαδοῖς ὕπνου is excellent: ‘on wings which
are sleep's companions (escort)’, ie. together with sleep the dream-vision
approaches us, to fly away again in its company. So I see the seat of the
corruption in «eAeUfors.!
1 I do not wish to deny that ὕπνου κελεύθοις is per se a good poetic image, though no
expression exactly parallel to it seems to occur elsewhere (cf. O. Becker, ‘Das Bild des
Weges', Hermes, Einzelschriften, Heft. 4, 1937, 13). But the point here is the difficulty of
connecting it with ὁπαδ-.
222
 COMMENTARY lines 427-8
Up to this point we have been establishing facts. What follows is an
hypothetical attempt first to explain the corruption, and then to recover the
poet’s thought, if not his actual wording. I suspect that κελεύθοις is due to
somebody who had ἀκολούθοις before him, and was trying to make it fit into
the metre (about which the strophe left no doubt). ἀκόλουθος is the stock
gloss for ὀπαδός, not only in the Etym. Gen. (AB) ὀπηδεύω" ἀκολουθῶ and
Hesychius (ὀπαδός" ἀκόλουθος, συνεργός, δοῦλος), but also in, e.g., the interlinear
gloss in T on Hom. B 184 ὀπήδει: ἠκολούθει, Schol. S. Trach. 1264 αἴρετ᾽,
omadoi‘ οὐκ οἰκέταις, ἀλλὰ τοῖς ἀκολουθοῦσιν αὐτῶι... φησί, Schol. Theocr. 2.
166 Νυκτὸς ὀπαδοί' τουτέστιν ἀκολουθοῦντες τῆι Νυκτί. It is particularly
apposite that in the Marcianus οἱ Euripides at Hipp. 108 we find written
above ὀπαδοί the interlinear gloss ἀκόλουθοι. I assume that in Ag. 426 an
interlinear gloss of this kind has, to start with, displaced the original ending
of the line v — — (Ag. 111 provides evidence to illustrate this process), and
that it was afterwards adapted to fit the metre. I would insert exempli
gratia πελῶσα in place of κελεύθοις. The future πελῶ (pres. πελάζω) is found in
Aeschylus (Prom. 281) as well as in Sophocles. The hiatus at the end of the
corresponding line in the strophe (409) shows that a metrical period ends here,
and therefore the syllaba anceps is allowable.
For the thought of 423-6 cf. (Schneidewin) Ÿ 99 f. ὧς ἄρα φωνήσας (Achilles
in his dream addressing the apparition of Patroclus) ὠρέξατο χερσὶ φίληισιν,
οὐδ᾽ ἔλαβε.
426. Pauw and Schütz make the ‘prophets’’ words end after κελεύθοις ; most
editors have followed them, probably rightly. For (1) after 409 τόδ᾽ evverovres
a longer speech is expected (while Paley, in his translation, and Wilamowitz
limit the prophets’ words to 410-ı1) ; (2) it follows from the future δόξει that
at least 415 also belongs to the speech, but, in view of the connexion noted
above, one would be reluctant to separate from it the following passage down
to 426 ; (3) 427 τὰ μὲν κατ᾽ οἴκους clearly indicates a conclusion and a transition
to something new.’ While admitting the possibility of a different arrange-
ment, I have adhered to the view held by most editors. If this view is correct,
the speech ends at the corresponding point in the antistrophe to that at which
it begins in the strophe (409), a point which is marked by the conclusion of a
metrical period.
427. κατ᾽ οἴκους of the one royal house as in 35, 237, and elsewhere. ἐφ᾽
ἑστίας “at the hearth’: the genitive is correct here as is shown by the lexicons
and grammars, and recognized by Hermann; so he need not have considered
Vossius’ conjecture edeoriovs, which has actually been put into the text by
other critics before and after Hermann.
428. The bold and forceful ὑπερβατώτερα should not be questioned. If
Housman, J. Phil. xvi, 1888, 262, says ‘you would expect τῶνδ᾽ ὑπερβατά....,
not τῶνδ᾽ ὑπερβατώτερα ’, he fails to appreciate Aeschylus’ preference, noted
above on 378, for redundant duplication in phrases denoting excess.
427-8. τὰ μὲν κατ᾽ oikous ... ὑπερβατώτερα. There is a striking similarity

1 For both of the last-mentioned reasons it seems to me wrong to put the end of the
prophets’ speech after 419 with Lewis Campbell (prose translation, 1893) or after 415 with
Murray and G. Thomson (Thomson does not discuss the point in his commentary, nor
does he mention the fact that Headlam puts the end of the prophets’ speech after 426, cf.
his prose translation).
223
lines 427-8. COMMENTARY
of form in the sentence Thuc. 6. 17. 6 r re οὖν ἐκεῖ (in Sicily) ἐξ ὧν ἐγὼ ἀκοῆι
αἰσθάνομαι τοιαῦτα καὶ ἔτι εὐπορώτερα ἔσται κτλ.
429. τὸ mäv: always adverbial in Aeschylus (cf. on 175, Blass on Cho. 331,
where obviously no alteration must be made, Kaibel on S. El. 1009) : ‘on the
whole, to the full extent, totally, generally’, etc. ‘In the house (of Menelaus)
... thus it is, but in general. . . .' ‘From the private grief of Menelaus while
he sat at home we pass now to the general multitude at large... .' (Headlam).
. The subtle transition in this section deserves close study.
The phrase 414 πόθωι (cf. on 431) δ᾽ ὑπερποντίας is the pillar on which the
structure of the subsequent thoughts mainly rests. These thoughts are con-
cerned with Menelaus in the first place, but not with him alone. The word
ἀνδρί in 417 (cf. note) is chosen so as to include other men in similar distress.
A still more considerable widening of the perspective takes place when at 423
the Chorus, instead of saying ‘he’ and ‘his absent wife’ or something to the
same effect, speak of τις and ἐσθλά. The grievous experience of ὀνειρόφαντοι
πενθήμονες δόξαι is common not only to deserted or bereaved husbands but to
all who miss one very dear to them, wife, parent, child. We have only to add:
miss a dear one across the sea; for the initial πόθωι ὑπερποντίας (the latter
word with a gradually widening application, so that the sex matters no
longer) remains the keynote throughout. The final touch to the generalization
of an experience which was first illustrated by the special case of Menelaus
is given in the words τὸ πᾶν δὲ κτλ. Stripped of their accidental features, the
sufferings of Helen’s husband are the same as those of the unnamed mourners
in thousands of Greek houses; of them it is said: πολλὰ θιγγάνει πρὸς ἧπαρ,
they, like the son of Atreus, have to pass through τάδ᾽ ἄχη καὶ τῶνδ᾽ ὑπερ-
βατώτερα. To them, too, the illusory consolation of passionate dreams will
come: the τις who in 433 is the subject of ἔπεμψεν may also appear as ἐσθλά
τις δοκῶν ὁρᾶν.
Thus, by smooth and almost imperceptible transitions, we are led from the
picture of the departing Helen and the sorrows of her husband back to the
wretched victims of her rashness, the citizens to whom she left armament and
war (403 ff.). But what at the beginning of the second stanza showed itself
as an imminent peril has in the meantime become a horrible fulfilment : for
many years the war has gone on, the gold-changer Ares plying his trade
unceasingly ; and every house in Greece has its share of grief. The way is now
open to the grave charges against the Atridae ; we hear of the grumblings and
curses of the people, and a climax is reached with the appeal to the highest
power: τῶν πολυκτόνων γὰρ οὐκ ἄσκοποι θεοί. This war was ordained by the
divine champion of justice, and yet.... The searching mind has been carried
to the verge of an abyss, an antinomy as irreconcilable as those which are
revealed in the parodos.
The slight alteration "EAAavos (cf. 1254) has deservedly been accepted.
συνορμένοις : ‘for those who have departed together (for the war)’. With-
out the article, as 413 ἀφειμένων.
429 f. Between συνορμένοις and πένθεια there is a syllable missing: this is
shown by the strophe (412 £). That πένθεια is not a Greek word at all is
implicitly shown by Lobeck, Paral. 322, when in desperation he classes it with
Homer's eAeyxein and Nicander's ὀνειδείη. Housman, J. Phil. xvi, 1888, 263,
stigmatizes it as ‘a mere collocation of letters’, cf. also Platt, J. Phil. xxxii,
224
 COMMENTARY line 431

1913, 53. From P. Chantraine’s survey (La Formation des noms, 88) it appears
that there is no sort of analogy to be found for wevdera." The passage was
emended by Blass, Hermes, xxix, 1894, 633, and rightly expounded by
Wilamowitz in his Commentariola metrica of 1895 (= Verskunst, 184). For
ἀπένθεια τλησικάρδιος cf. Hermes, lxviii, 1933, 242 ff. In L-S, Addenda s.v.,
ἀπένθεια is rendered by ‘non-lamentation’. Obviously the pointed expression
is only appropriate in a case where there is cause for sorrow, but where the
visible and audible manifestations of mourning (πένθος, luctus?) are absent,’
because the feelings are under control In the phrase of Protagoras (B 9
Diels) about Pericles, who νηπενθέως ἀνέτλη after losing his young sons, the
characteristic attitude is indicated with special clearness. An illustration of
this attitude is afforded by the numerous burial scenes we find on Athenian
lekythoi: "There are no tears on these vases: to be exact, in all the hundreds
of lekythoi that have come down to us, there are only one or two weeping
figures. Grief is not rendered by a distorted face, but by bended head, by
gesture of hand or arm, by the attitude of the whole body’ (Beazley, Attic
White Lekythoi, 1938, 21 £). No contradiction of the steadfast endurance
stressed in 429 f. is contained in 445 ff. : in the στένειν of that passage and its
content we have no ὀλοφύρεσθαι, and certainly no πένθος in the more restricted
sense which is intended here. Cf., too, the description of the manly bearing in
Sept. 50 f. δάκρυ λείβοντες, οἶκτος δ᾽ οὔτις ἦν διὰ στόμα.
τλησικάρδιος here clearly takes up the Homeric phrase τέτλαθι δὴ κραδίη
(there is a slightly different shade of meaning in Prom. 159). As far as the
formation of the word is concerned, it is compared in the Thesaurus with
τλάθυμος (Pindar); Blomfield refers to ταλακάρδιος, Headlam to Hesychius
τλασίφρονα' ὑπομονητικόν, and Wilamowitz to names like Τλησιμένης.
431. The scholiast’s paraphrase (ZyoA. wad. in Tr) τῶν συνηγμένων ἀπὸ τῆς
“Ἑλλάδος ἁπάντων ἑκάστου τοῖς οἴκοις ὀδυνηρὰ πένθησις διαπρέπει makes it very
probable that, at the time when it was written, δόμοις still stood in the text,
and this was restored by Auratus. What the attempt to justify δόμων leads
to is shown by Conington’s careful but inconclusive note, at the end of which
he admits that he prefers δόμοις really. It is perhaps as well to give a special
warning against Dobree’s öduwı "v. Ahrens, De crasi et aphaeresi (Stolberg
1845), 24 f. (= Kl. Schr. i. 79), observes that aphaeresis of ἐν is found only
twice in Tragedy.* One of the two instances, Eur. fr. 491. 5 N., needs no

1 As recently (1939) Schwyzer, Griech. Gramm. i. 469, attributes ‘ πένθεια for wevdos’ to
Aeschylus with no suggestion of doubt, or reference to the textual problem, I should like to
mention that Wackernagel wrote to me on 3 May 1933 with reference to the article in
Hermes, \xviii: ‘Many thanks for the ἀπένθεια τλησικάρδιος. I am glad the grammarian has
been able in this case to clear the way for the commentator, whom everybody will now be
thankful to follow.’
2 Horace in one of his earliest epodes (16. 39) renders Archilochus’ (rAjre) γυναικεῖον
πένθος ἀπωσάμενοι with the accuracy one expects from him by muliebrem tollite luctum, and
in another of the early epodes (10. 17) he has a variation of it: lla non virilis heiulatio (this
is a technical term for the mourning cries of women at funerals).
3 The scholiast on E. Phoen. 1028 says ἀπενθὴς δὲ θεὸς ᾿Απόλλων and thus quite correctly
uses ἀπενθής to make it clear that mourning is alien to the nature of the god.
* He rightly leaves out of consideration instances like A. Suppl. 89 κἀν or the Sophoclean
examples of 7) ἐν, μὴ ἐν etc. It matters little whether we regard these simple cases, which are
hardly to be compared with a possible δόμωι 'v, as ‘crasis’ with Ahrens, or as 'aphaeresis'
with A. Lucius, Dissert. philol. Argentor. ix (1885), 391.
4872-2 Q 225
 line 431 COMMENTARY
further word, for in that passage the conjecture ᾿᾽γκαλεῖσθαι has long dis-
appeared from the editions. For the other passage, Prom. 741 εἶναι δόκει σοι
μηδέπω "v (μηδ᾽ ἐπῶν MSS) προοιμίοις, Ahrens refers to E. Hec. 1195 where the
MSS give the metrically impossible καί μοι τὸ μὲν σὸν ὧδ᾽ ἐν φροιμίοις ἔχει and
the omission of the preposition (ὧδε φροιμίοις) has been generally accepted.
Consequently Ahrens raises the question whether the preposition should not
be regarded as an intruder also in the closely related passage of the Prome-
theus. The suggestion certainly deserves to be considered. But even if the
aphaeresis is retained in that passage, it is far too rare a phenomenon to
be introduced by way of conjecture anywhere else.
The scholion is right, too, in paraphrasing πρέπει by διαπρέπει (similarly
taken, e.g., by Linwood, Dindorf, Schneidewin), and Headlam is wrong in
translating ‘is proper’: the verb means ‘stands out prominently, is clearly
recognizable’ (cf. on 242). ovvopuevors belongs to ἀπένθεια. . . πρέπει, as
dative of the person interested; then δόμοις is probably to be regarded as
supervening, a case of the so-called σχῆμα ’Iwvirov, cf. Wilamowitz on E.
Her. 162. It would also be possible to take δόμοις as locative.
δόμοις ἑκάστου is in sharp contrast to 427 κατ᾽ οἴκους, the royal house: it
therefore further stresses the general character (already indicated by τὸ πᾶν)
of the 8é-clause in contrast to that introduced by ra μέν. The idea that each
individual house (eis ἑκάστου δόμους again shortly afterwards in 435) is hit by
the sufferings of the war is important. We find it also in the passage in the
parodos of the Persae, which has so much in common with the present
passage in the Agamemnon (from 414 to 433) not only in content but also in
the use of significant words, Pers. 133 ff., λέκτρα δ᾽ ἀνδρῶν πόθωι πίμπλαται
δακρύμασιν: Περσίδες δ᾽ ἁβροπενθεῖς ἑκάστα πόθωι φιλάνορι τὸν αἰχμήεντα θοῦρον
εὐνατῆρα προπεμψαμένα λείπεται μονόζυξ. Here the keynote is provided by the
repeated πόθωι, as in the Agamemnon passage πόθωι (414), though it is not
repeated, expresses the idea which links strophe and antistrophe together.
πόθος, of course, like ἵμερος etc., expresses the desire of the lover, Ag. 414 the
yearning for the lost one carried off overseas, but also the yearning for those
who are gone from us for ever: e.g. ἄνδρας μὲν πόλις ἦδε ποθεῖ καὶ δῆμος
᾿Ερεχθέως (IG 13. 945 = Tod, Greek Hist. Inscr. 59, on the dead of Poteidaia
in 432), Gorgias Epitaphios (B 6 Diels) τοιγαροῦν αὐτῶν ἀποθανόντων 6 πόθος οὐ
συναπέθανεν, Lysias 2. 71 ἀξιον τοῖς ζῶσι τούτους ποθεῖν, Plato, Phil. 48 a τὰς
ev τοῖς θρήνοις καὶ πόθοις ἡδονάς.
432. γοῦν: ‘at any rate’. Denniston, Particles, 451: ‘Much the commonest
use of γοῦν is to introduce a statement which is, pro tanto, evidence for a
preceding statement. This has been well termed “part proof”. He rightly
quotes this passage under this head.
θιγγάνει: the verb is not uncommon in Tragedy, though apparently alien
to pure Attic: cf. Rutherford, New Phrynichus, 391 n. 1, and for its occurrence
in the dialects Wackernagel, Unters. z. Homer, 222. For the thought cf. P 564
μάλα γάρ pe θανὼν ἐσεμάσσατο θυμόν.
435. τεύχη: it is tempting to take this as meaning the armour, especially as
this Homeric use of the word occurs not only in Sophocles and Euripides, but
in Aeschylus, too, cf. the new fragment of the Myrmidons,' Pap. Soc. It. xi,
No. 1211, 1. 17 =D. L. Page, Greek Lit. Papyri, i. 140 (cf. Schadewaldt,
1 I regard the attribution as almost certain, in spite of the lack of external evidence.
226
 COMMENTARY line 435
Hermes, 1xxi, 1936, 27). This was how Schütz took it (cf. the considered dis-
cussion in Paley) and also, e.g., Wilamowitz (besides his translation, cf. his
Griech. Verskunst, 185 n. 1, and his summary in Griech. Tragoedten, ii. 307) ;
Headlam considered it as a possibility. But there seems to be no evidence
that the weapons or armour of the fallen were sent home with their ashes
after battles overseas; and though the incompleteness of our information
may account for this, it is wiser, perhaps, to interpret τεύχη as meaning urns
for the ashes (5. El. 1114, 1120 have often been quoted as parallels). The fact
that the plural τεύχη generally means not ‘vessels’ but ‘armour’ (cf. Wacker-
nagel, Syntax, i. 88) need be no objection, for in Eum. 742 we find τευχέων
used of voting-urns and Aesch. fr. 180. 5 N. μυρηρῶν revyéwv. The meaning
‘urns’ may find some further support in the occurrence a few lines down of the
words λέβητας εὐθέτους in close connexion with an almost literal repetition of
the phrase ἀντὶ δὲ φωτῶν... σποδός. However, it seems impossible to decide
with certainty the meaning of τεύχη here.
In the later fifth century the bodies of Athenians who fell in battles abroad
were as far as possible cremated on the battlefield (in this an earlier custom
was maintained); the ashes were afterwards collected and, again as far as
was practicable, sent back home. So the commentators on Thucydides have
rightly concluded from the accounts in 2. 34 taken with 6. 71. 1; in a some-
what later period we have similar evidence from Isaeus 4. 19 and 9. 4 (Wyse
produces further evidence in his commentary on these passages). The words
of this ode of Aeschylus provide the earliest testimony to the custom described
in these accounts. In mentioning the transfer of the warriors' ashes to their
own cities Aeschylus allows himself what the scholiasts on similar occasions
(e.g. on Sept. 277 and Eum. 566 ff.) call an anachronism, for the practice of
collecting the bones and ashes for the purpose of taking them back to the
homeland was unknown in the period of the epic poems: the lines H 334 f.
which were obelized by Aristarchus stand 'in glaring contradiction to the
practice of the Homeric poems’ (Wilamowitz, Die Ilias und Homer, 55).
F. Jacoby's careful re-examination of all the available evidence (Journ. Hell.
Stud. lxiv, 1944, 37 ff.) has led him to the conclusion that in Athens the custom
of burying in the Kerameikos the ashes of those who had fallen in battles
abroad was introduced after the disaster at Drabeskos, in 464 B.C., i.e. a few
years before the performance of the Oresteia (cf. on 1547).
The significance of these lines for the state of feeling in Athens at the time
of the production of the Oresieia was fully appreciated by J. G. Droysen
(Aeschylus, translated by J. G. D., 3rd ed., p. 558) : he sees in them a reminis-
cence of the fearful losses sustained by Athens in the campaigns of the year
459 (to illustrate the gravity of the losses, he quotes the casualty list! of the
Ἐρεχθηΐς, IG i? 929 = Tod, Greek Hist. Inscr. no. 26): similarly Merriam,
‘Telegraphing among the Ancients’ (Papers of the Archaeol. Institute of
America, Classical Series, iii, No. 1), 20 ff., Wilamowitz, Griech. Tragoedien, ii.
307; cf. also G. De Sanctis, Storia dei Greci (1939), ii. 91. This observation
must not be understood to imply that the words οὖς μὲν γάρ τις ἔπεμψεν κτλ.

1 It had already been referred to, in connexion with the Oresteia, by Otfried Müller,
Eumenides, 124. For the controversial problem of the date of the list cf. Kolbe, Hermes
Ixxii, 1937, 267 (with references to the earlier discussion) ; the view which he rejects is again
held by R. Meiggs, Journ. Hell. Stud. lxiii, 1943, 29 n. 42.
227
line 435 COMMENTARY
and their continuation in the following stanza become significant only when
seen in relation to contemporary events: the phrase ἐξ ’IAlov (440) marks the
idea which is predominant in the thoughts of the Chorus. But here, as so
often in this ode, the individual experience broadens into something more
general, and our knowledge of the background of the years 459/458 allows us to
guess what response must have been awakened in the hearts of the audience
by such words.
437 ff. Without being fanciful we may perhaps indicate the point out of
which the magnificent imagery is developed. The words ἀντὲ φωτῶν σποδός
are repeated in the next stanza in the more powerfully compressed phrase of
443 ἀντήνορος σποδοῦ, in which the idea of ἀμείβεσθαι is contained. It is out of
this idea that the daemonic figure of χρυσαμοιβὸς Ἄρης σωμάτων arises. While
Ares is here practising a very special trade, we find him in a chorus of the
Supplices (638) occupied in the commonest and most important of the tasks
of peaceful mankind, though indeed after his own gruesome fashion (as in the
song ‘Es ist ein Schnitter, heisst der Tod’): τὸν ἀρότοις θερίζοντα βροτοὺς
ev ἄλλοις. It is characteristic of Aeschylus that in order to heighten the effect
of terrible happenings he does not borrow his imagery from the realm of the
unreal and the fantastic, as many romantic poets do, but from the familiar
processes of everyday life or the peaceful incidents in nature: while he
depicts with minute exactness little details innocent enough in themselves
(e.g. in this passage ταλαντοῦχος, ψῆγμα, γεμίζων λέβητας), he gives them at the
same time a metaphorical relation to terrible powers, and it is just this
! The last word is altered or misunderstood by many editors. Pauw wrote ἐνάλλοις, ‘non
male’ in Hermann's opinion. This now stands in Murray’s text (‘corr. Lobel’,—in Cho. 505
in Murray’s text a conjecture is attributed to Lobel which was made by Schütz, adopted by
Wecklein (1888), and mentioned with approval by Verrall, though it had been refuted by
Hermann; to Lobel again in Eum. 769 a proposal of Heath’s is ascribed) ; in L-S p. 2068 it
is indicated as ‘prob.’ I confess I cannot understand at all what ἀρότοις ἐνάλλοις is supposed
to mean here. Pauw’s explanation is ‘praeposteris, et quae contra naturae leges sunt’,
But this is not the meaning of the word. The context of πάντα δ᾽ ἔναλλα γένοιντο in Theoc. 1.
134 (from 132 viv ἴα μὲν φορέοιτε βάτοι onwards) shows quite clearly that the scholiast’s gloss
οἷον ἐνηλλαγμένα is right. It means not simply ‘changed’ (so L-S), but ‘exchanged’,
“inversus et contrarius’ (Budaeus, quoted in the T'hesaurus). The correct interpretation of
Suppl. 638 is given, as to the general sense, by Tucker (‘fields not his own’), and, with
greater verbal precision, by Wilamowitz, Interpr. 37 (‘der Menschenernte schneidet wo
andere gesät hatten’). ἄλλος cannot actually be the equivalent of ἀλλότριος (for Pind. P. 4.
268, where it is usually so taken, see O. Schroeder, Pindars Pythien, 1922, 47: ' ἄλλοις ἐν
τείχεσιν : the well-known idiom instead of ἄλλοθι, ἐν τείχεσιν, as opposed to ἐὸν χῶρον, out in
the country’). Also dporos does not in itself mean ‘corn-field’ (so L-S ad init.) either here
or in Hom. ¢ 122 (dpdrocow ‘by ploughings, i.e. agriculture’ Ameis—Hentze, correctly). Its
meaning here as elsewhere is rather ‘ploughing’ (and ‘sowing’). The addition of ἐν, to which
Pauw objected, but which is quite indispensable here, signifies ‘in the place where the
ploughing and sowing is done’. If we say ‘he reaps the harvest in other ploughing-places’,
it is really clear enough that a contrast is implied with the normal procedure whereby the
farmer reaps in the identical field which he himself has ploughed. The idea of the corn
which is reaped by the intruding reaper is further developed in 663 ff. Bas δ᾽ ἄνθος ἄδρεπτον
ἔστω" μηδ᾽ ᾿Αφροδίτας edvdrwp βροτολοιγὸς "Apns κέρσειεν ἄωτον. To Athenian ears there was
probably a particularly sinister sound about the phrase ἀρότοις ἐν ἄλλοις in which human
lives are the corn-crop, because in Athens every legal marriage was contracted γνησίων
παίδων Em’ ἀρότωι (see below on 1207). The parents have laboured for the ἄροτος, but now
comes the god of destruction, θερίζων ἀρότοις ev ἄλλοις. For the word-order dpéras . . . ev
ἄλλοις, which is very unusual in Tragedy, but which is not to be suspected here in a choral
ode, cf. on 964.
228
 COMMENTARY line 440
contrast that intensifies the horror. This appears, e.g., in the dry technical
terms of legal procedure in the descriptions of the merciless wiping out of the
enemy’s city and state (534 ff. and 813 ff.), and especially in the grim travesty
of the gay figures of the κῶμος (1186 ff.) and of the young growing corn
(1391 f.), cf. ad locc.
437. ὁ χρυσαμοιβὸς δ᾽ Ἄρης σωμάτων: ‘Ares, the gold-changer who
exchanges men’s bodies’. For the word-order cf. on 633. The construction, in
which the second element (ἀμειβ-) of the compound has not only the first part
(χρυσ-) depending on it as object but also the genitive σωμάτων, has a remark-
ably close parallel in Sept. 729, where the brutal destroyer, the XaAuBos Σκυθῶν
ἄποικος, is designated as κτεάνων χρηματοδαΐτας πικρός. It is certainly not
accidental that in that passage, as in this, the bold phrase describes in a very
condensed form the activity of a murderous daemon : the similarity of the idea
forces the language into a similar mould. Another point of resemblance
between the passages is that in Sept. 729, too, the picture is drawn from the
sphere of the transaction of business: this affords (together with the slightly
varied expression! in Sept. 941-5) a further example of what was pointed out
in the preceding note.
χρυσαμοιβός: the word only occurs here; the scholia identify it with
ἀργυραμοιβός and also note the analogy with ἀλφιταμοιβός.
438. In regard to the expression ταλαντοῦχος κτλ. the scholiast (ZyoÀ. «aA.
in Tr) observes : εἴληπται δὲ τὸ νόημα τῶι Αἰσχύλωι ἐκ τῆς κατὰ τὸν Ata.‘ Ομηρικῆς
ζυγοστατήσεως. This seems to be a case of misapplied acumen. The detail of
the ‘holder of the scales’ forms an integral part of the image beginning with
ὁ χρυσαμοιβός, and that image is derived from everyday life (see above) and
has nothing to do with the Homeric conception. Perhaps the commentator
was lured into this wrong combination by his knowledge of the fact that
Aeschylus in his Vvyooraota was indeed influenced by the motif of the Homeric
Zeus weighing the δύο κῆρε of the two heroes? and especially by its variation
in the Aethiopis,> where the ‘keres’ were replaced by ψυχαί. When Aeschylus
in the Phrygians works out the implications of a ‘hyperbolical’ expression in
Homer, and actually has Hector’s dead body weighed against gold in the
scales (Schol. A and T on Iliad X 351), this is again quite different from the
imagery of the passage in the Agamemnon.
ἐν μάχηι Sopds: similarly [Eur.] fr. 1109. 4 N. (restored with certainty),
μάχη δορός also in E. Cycl. 5, fr. 360. 24, 1048. 5, in a satyr play Pap. Oxy. viii.
1083, fr. 1. 9 f. ἔστι μὲν τὰ πρὸς μάχην δορός. For S. Ant. 674 see Jebb, ad loc.,
cf. further E. Heraclid. 159 f. és πάλην δορός.
440. πυρωθέν: cf. 481. Wackernagel, Uniers. z. Homer, 124, includes it in
the considerable number of verbs in -oöv which make their first appearance
in the form of the aorist passive. Headlam would like to see in it a secondary
τ In this connexion it should be remembered that the perversion of the normal process
of dividing thc inherited estate, the κλῆρος, provides one of the main themes for the last
part of the Seven against Thebes.
2 Cf. the passages from Plutarch and the Iliad scholia quoted by Nauck, Trag. Graec.
Fragm. 88 f., in which it is stated that the fundamental idea of Aeschylus’ Yuxoorasia was
suggested by the weighing of the κῆρε in Homer.
3 Cf., e.g., Studniczka, Jahrb. d. Inst. xxvi, 1911, 132, Beazley, Der Kleophrades-Maler,
12, A. B. Cook, Zeus, ii. 733 ff. and iii. 1148 f., and G. Bjórck's article ‘Die Schicksalswage',
Eranos, xliii, 1945, 58 ff., in which a survey of the more recent discussion is also given.
229
line 440 COMMENTARY

meaning ‘fined in the fire’, but there seems to be no instance of this earlier
than the language of the Bible.
441 f. βαρὺ ψῆγμα: nothing, I am afraid, can be done to help those fanatics
of logic who would remove or at least suspect this magnificent oxymoron ;
nor can we prevent them from reviving Schiitz’s unfortunate suggestion
βραχύ (put forward in his later edition) instead of following his correct
explanation: βαρύ quia amicorum ossa cineresque graviter amicos afflige-
bant’. If parallels are needed, they can be found in E. Suppl. 1123 ff. φέρω...
ἐκ πυρὸς πατρὸς μέλη, βάρος μὲν οὐκ ἀβριθὲς ἀλγέων ὕπερ κτλ. (and this is more
than a ‘parallel’, for there is in the whole stanza—note especially 113o—an
obvious reference to Ag. 434-44) and S. El. 1139 f. οὔτε παμφλέκτου πυρὸς
ἀνειλόμην, ws εἰκός, ἄθλιον βάρος. Schütz was aware of the relation between
the Agamemnon passage and its echo in E. Suppl., but he wrongly used
Suppl. 1130 to support his conjecture.
442. ψῆγμα: this means the ashes of the dead men, but the imagery of
χρυσαμοιβός is kept up: ψῆγμα is, as was recalled by Bamberger, Opusc. 48,
the technical term for gold dust (there is a collection of instances in H.
Blümner, Technologie, iv. 119 n. 3), and as such is perhaps of Ionian origin, cf.
W. Aly, De Aeschyli copia verborum (1906), 42 f.
443. ἀντήνορος σποδοῦ ; ‘ashes in place of the man’. The adjective is found
only here, and presumably was coined by Aeschylus. The possibility that in
this case (in spite of the difference in meaning) as in others the invention of
a new poetic compound was suggested by a Homeric personal name is
pointed out especially by E. Williger, Sprachi. Unters. 5 n. x (where earlier
literature on the subject is quoted), and supported by O. Hoffmann, Glotta,
xxvili, 1939, 36.
444. εὐθέτους (the ending is restored with certainty): Schol. M on Sept. 642
yp. εὔθετον, iv’ ἦι εὐβάστακτον. θεῖναι yap τὸ ἀναλαβεῖν λέγουσιν 'Arrwot. The
meaning obviously is that their shape and light weight render the urns
convenient for transport (in the fifth century room for the urns had probably
to be found on board the dispatch boats which maintained the connexion
between the war zone and Athens, cf. F. Jacoby, Journ. Hell. Stud. lxiv,
1944, 62 n. 118). The special technical sense quoted by van Heusde (from
Phryn. Praep. soph. p. 71.9 de B., cf. the inscription from Phrygia quoted in
L-S s.v. εὔθετος) εὐθετεῖν νεκρόν" τὸ εὖ κοσμεῖν ἐν τάφοις νεκρόν is irrelevant
here, for the bodies have been cremated.
448. διαί: the form, attested here by an ancient quotation, is found in 1453,
1486, and Cho. 656, and now also in the papyrus of the Myrmidons, 1. 8 (D. L.
Page, Greek Lit. Pap. i. 140). But are we justified in accepting the interpreta-
tion of the Epimerismoi, ἀντὶ τοῦ ἕνεκα, or, for that matter, of Triclinius
(ἤγουν ἕνεκα), as several editors do (e.g. ‘for sake of another’s wife’, Headlam,
‘fiir das Weib eines andern’, Wilamowitz; cf. also his note on Sept. 233)? It
cannot be denied that Dindorf’s (Lex.) general explanation ' διά cum genetivo
personae vel rei per quam quid fit’ covers this passage, too; similarly L-S
διά A. III. a ‘by her doing’, Verrall ‘through another’s wife’. We only need
add that this διά c. gen. has not only an instrumental function in Aeschylus,
but also serves to indicate the causa efficiens ; whether the one sense or the
other is preponderant, or both carry equal weight, depends in every case on
the context. Kühner-Gerth, i. 485, give too narrow an interpretation when they
230
 COMMENTARY line 449

quote Ag. 447 as a case of ‘causal διά c. gen.’ and explain thus: ‘merely:
through the fault of another’s wife, not by her hand’. Not by her hand,
certainly, but by all her faults of commission and omission, for example by
her having lived throughout the ten years in Troy as the wife of Paris.
Plüss has a good comment: ‘Helen is not merely the guilty cause, but also
medium and instrument.’ The causal-instrumental sense is at least strongly
suggested in the parallel passage 1453, where the words πολλὰ rAavros γυναικὸς
διαί are used of Agamemnon, for the Chorus immediately continues παράνους
eva, μία τὰς πολλὰς... ψυχὰς ὀλέσασο᾽ ὑπὸ Τροίαι, and Helen could not
possibly be more clearly indicated as the efficient cause and at the same time
the instrument of evil. The same applies to Ag. 1486 διαὶ Διὸς παναιτίου
mavepyera (see note), where the poet himself explains the double function;
Sept. 233, too, must be taken in the same way. The old grammarian who
started the explanation ἀντὶ τοῦ ἕνεκα seems in this case (and also in Aesch.
fr. 296 N., cf. Schadewaldt, Hermes, lxxi, 1936, 46 n. 1) to have mistaken the
particular shade of meaning ; he, like his modern followers, may perhaps have
had at the back of his mind some familiar Homeric tags like (I 339) ‘EAévns
ἕνεκ᾽ ἠυκόμοιο OT (A 438) Ἑλένης μὲν ἀπωλόμεθ' εἵνεκα πολλοί etc. Aeschylus
himself in 800 says that the campaign was undertaken ‘EAevns ἕνεκα, cf. also
62. Obviously ‘for the sake of another’s wife’ would be appropriate here, too;
that does not, however, justify us in altering the thought actually expressed
by the poet. It is bad enough for a hero like Amphiaraos to be ruined διὰ
yuvarkas; that it should come about in this case ἀλλοτρίας διαὶ γυναικός
heightens the indignation of the bereaved."
449. ns: cf. on 369.
Batlew, properly to utter the sound Bad Bas (for the formation cf. on 1316
δυσοίζω), of a dog's barking, then figuratively used of yapping vituperation,
so Ar. Thesm. 173; here ‘to grumble'. This use is natural enough; cf. Latin
gannire (cf. Thes. 1. Lat. vi. 2, p. 1691 f.), properly of a dog = latrare, then
figuratively — murmurare, obstrepere, maledicere. For the difficulty in Pers.
13 see Wilamowitz, ad loc.
τάδε σῖγά τις βαὔζει: Wecklein asserts: ' 7d8e refers only to the words
ἀλλοτρίας διαὶ γυναικός. Cf. τάδε 1334.’ Accordingly Verrall (in his translation),
Headlam, and Murray print the words ἀλλοτρίας διαὲ γυναικός in inverted
commas. But while in 1334 the whole expression to which τάδε refers con-
sists of μηκέτ᾽ ἐσέλθηις, in this case the phrase ἀλλοτρίας διαὶ γυναικός would be
cut off from the sentence (445 ff.) to which it not only essentially belongs, but
of which it is the culmination, voicing as it does extreme bitterness. Thus
Wilamowitz in his translation is right in leaving the indirect speech unbroken:
he renders τάδε σῖγά τις faite. by: “Leise murrt man so’. This was obviously
1 The foregoing note was already written when I became acquainted with P. T. Stevens's
valuable article AAAOTPIAZ AIAI TYNAIKOZ, which is also instructive for the later use
of διά, C.R. 1, 1936, 162 ff. I agree with him on one chief point : he, too, rejects the meaning
‘on behalf of’, ‘for’, because, as he rightly stresses, this use of διά with the genitive is not
found elsewhere in the classical authors. But it seems to me that Stevens, like Kühner-
Gerth, loc. cit., draws too sharp a distinction between the causal function, which he assumes
in this and kindred passages in Aeschylus, and the instrumental function. Though this
distinction may be justified from a purely logical point of view, yet some of the passages
in Aeschylus seem to show that the poet deliberately used an expression which denoted
both the originator and the main instrument of a given action or event.
231
line 449 © COMMENTARY

the poet’s meaning: it is the characteristic attitude of those who, being

completely or temporarily powerless, cannot entirely suppress their criticism
of the conduct of their rulers although they can only give furtive expression
to it.’ The meaning of σῖγα here, and the opposite of it, can be illustrated
from a passage of Herodotus which vividly depicts the typical phenomenon
of just such discontent as this: 8. 74. 2 τέως μὲν δὴ αὐτῶν ἀνὴρ ἀνδρὶ παραστὰς
σιγῆι λόγον ἐποιέετο, θῶμα ποιεύμενοι τὴν Εὐρυβιάδεω ἀβουλίην- τέλος δὲ ἐξερ-
ράγη ἐς τὸ μέσον. We need not resort to the expedient of Hermann, who takes
σῖγά τις βαὔζει to be in antithesis to στένουσι δ᾽ ed λέγοντες, and accordingly
divides the pronoun, writing ra δὲ oiya τις βαὔζει and translates alia tacite
quis mussttat. There is a similar passage in S. Ant. 700, where the discontented
talk of the citizens is repeated, ending with some words of direct speech, and
concluding τοιάδ᾽ ἐρεμνὴ (cf. 692 ὑπὸ σκότου) city! ἐπέρχεται φάτις, cf. on 460.
450. ‘The people’s grief for their losses in war is φθονερόν, they would like to
be spared them and are angry about them’ (Daube, 133, cf. also 135). Cf. on 135.
If we consider 5. Aj. 157 πρὸς yap τὸν ἔχονθ᾽ ὁ φθόνος ἕρπει, we might be
tempted to take ArpeiSats as depending directly on ὑφέρπει, ‘secretly draws
near to the Atridae’; this use of the dative with a verb of motion would be
quite regular (Kühner-Gerth, i. 406f.). So, e.g., Stanley, Schütz (‘subrepit
Atridis’), Conington (‘jealous hate begins to creep on those avenging brothers’),
and Jebb on 5. Oed. R. 786. But the interpretation followed by the majority
of more recent commentators seems better: ‘Grief full of disfavour towards
the Atridae comes over the mind (of those who stay at home).’ For in the
first place this picture of what goes on in the minds of those concerned
elaborates the idea of σῖγά ris Bate: Secondly, with this interpretation we
can keep to the full meaning of ἄλγος. Thirdly, where Aeschylus uses ὑφέρπειν
elsewhere (Ag. 270 χαρά μ᾽ ὑφέρπει, Cho. 463 τρόμος μ᾽ ὑφέρπει), it denotes the
subire of emotions in just the sense which we are here assuming. Here an
accusative governed by ὑπὸ... ἕρπει, viz. the people at home, may be
supplied from the context. This verb, which is always rare, does not occur
earlier than the Oresteia, and may possibly belong to the group discussed
on 386.
φθονερός like φθονεῖν (and ἐπίφθονος above, 135) governs the dative.
451. πρόδικος is found only here in a non-technical use; if the word was not
invented by Aeschylus, the meaning which he gives it in this passage was
probably a novelty. The öixn- element reminds us once more that the cam-
paign is viewed as an act of legal procedure (cf. on 41), while the προ- indicates
the position of the leading men, as the Atridae are called πρόμοι in 200 and
410. Paley renders correctly ‘the principals in the suit’. For a different view
see Daube, 20 f.
The poet points here to the typical reactions upon loyalty which a long
war produces in the home country; the consequences become even worse in
case of a reverse: οὐδ᾽ Erı γλῶσσα βροτοῖσιν ἐν φυλακαῖς" λέλυται γὰρ λαὸς
ἐλεύθερα βάζειν, ὡς ἐλύθη ζυγὸν ἀλκᾶς (Pers. 591 ff.).

τ Ennius, Ann. 182 Vahlen intus in occullo mussabant, Shakespeare, King John, Act IV,
Sc. ii, 187 ff. “Young Arthur’s death is common in their mouths: And when they talk of him
they shake their heads And whisper one another in the ear; And he that speaks doth gripe
the hearer’s wrist, Whilst he that hears makes fearful action, With wrinkled brows, with
nods, with rolling eyes,’
232
 COMMENTARY line 457
454, εὔμορφοι, unreasonably challenged by earlier critics, and misinterpreted
by others (‘in voller Gestalt’ Wecklein, ‘corps intacts’ Mazon), was explained
by Wilamowitz, Isyllos, 167 n. 19. He (cf. also his Pindaros, 202) compared
this passage with Pind. Ol. 6. 76, where Charis is said to shed over the victors
εὐκλέα μορφάν, on which the scholiast comments ἐπεὶ of νικῶντες δοκοῦσιν
εὐειδεῖς εἶναι. In his commentary on Cho. 490! Wilamowitz recalls the fact
that in Ag. 454 it is the heroes who are spoken of as εὔμορφοι; similarly
Pliiss. So the meaning may be taken as something like ‘transfigured’,
‘glorified’, ‘verklart’ (this has not been refuted by Kranz, Stasimon, 277).?
The view that the poet here really meant the heroes, whose cult was still
observed in the Troad right up to his own day, finds welcome confirmation
in the close correspondence of the words θήκας γᾶς κατέχουσιν with those of a
prayer in Suppl. 25 f. ὕπατοί τε θεοὶ καὶ βαρύτιμοι χθόνιοι θήκας κατέχοντες.
ἐχθρά: sc. γῆ.
455. ἔχοντας : the soil of which they have taken possession (cf. on 320 ἔχουσι)
and which they have defended against the enemy, the soil in which they now
lie, belongs to them (Wilamowitz, Glaube der Hell. ii. 17 ; cf. his more detailed
illustration of this idea in his commentary on E. Her. 1016).
The thought of the ‘refrain’ (452-5) is intimately connected with the
preceding description of discontent in the homeland. Similarly in the Persae
the passage just quoted (on 451) about ἐλεύθερα βάζειν is followed by the
momentous words (595 ff.) which conclude the whole stasimon : αἱμαχθεῖσα δ᾽
ἄρουραν Alavros περικλύστα νᾶσος ἔχει τὰ Περσῶν. The analogous evolution of
thought in the two choral odes is certainly no accident. As the war-wise
poet sees it, the restless murmuring in the homeland and the everlasting rest
of the fallen abroad are directly connected together as parts of one and the
same happening. The subtlety of his treatment of details is revealed in the
fact that he uses the words Αἴαντος... νᾶσος ἔχει τὰ Περσῶν about the defeated
warriors who lie buried in Salamis,? but κατέχουσιν and ἔχοντας of the con-
querors before Troy.
456. σὺν κότωι: ‘when they are angry’. For this use of σύν, which is character-
istic of Aeschylus, to denote an accompanying circumstance, see Tycho
Mommsen, Beitr. z. d. Lehre v. d. griech. Präpos. 607. Cf. below 776.
457. δημοκράντου : Porson's correction is as neat as the ‘mixture’, which is
still preferred by some critics, is untidy and, when applied to ἀρά, unsuitable
(in spite of Verrall. Aeschylus has a predilection for compounds with
-kpavros ; for the meaning of «paívew which underlies these formations cf. on
! His interpretation of εὔμορφον κράτος is very uncertain. Personally I adopt with many
others the view of Erfurdt and Hermann (on S. Oed. R. 189).
* Wilamowitz in Glaube der Hell. ii. 531 returned again to the idea that the dead man
‘appears "beautiful", glorified’, but the epitaph in JG i.? 982 should not have been quoted
in this connexion, for there (μνῆμ᾽ ἐσορῶν οἴκτιρ᾽ ὡς καλὸς ὧν ἔθανε) the point is that the
young man has died in his youthful beauty, and this is the thought which has been read into
our passage by the scholia (rd δ᾽ εὔμορφοι πρὸς πλείονα οἶκτον προσέθηκεν). As Wilamowitz,
loc. cit., connects the idea of the beauty of the dead with the beautiful monuments of the
Kerameikos, so Eugen Petersen, D. att. Tragódie, 642, wrote: 'Aeschylus is thinking of
reliefs on tombstones, like those outside the gates of Athens.' This is an attractive sug-
gestion, which may be taken as an alternative to the interpretation given above.
3 This implication is not, however, inherent in the phrase itself, cf., e.g., the epitaph of the
Corinthians who fell at Salamis (IG i.* 927, Tod, Greek Hist. Inscr. no. 16) νῦν δ᾽ παμὲ Αἴαντος
νᾶσος ἔχει Σαλαμίς.

233
line 457 COMMENTARY
369. δημόκραντος dpa here, δημόθρους ἀράς 1413, and Önuoppideis ἀράς 1616 are
clearly variations of one and the same notion. The xpaivew-element brings
out the idea that the curses are to be regarded as valid utterances, that they
carry with them the guarantee of fulfilment. This is in fact an essential
characteristic of dpaí, so that they can be regarded in the light of legally
binding obligations.
Snpokpávrou ... χρέος. The scholion (ZyoA. rad. in Tr) takes Agamemnon
to be the subject: ἤγουν ὀργίλως ὑπὸ τῶν πολιτῶν σκώπτεται 6 Ἀγαμέμνων καὶ
ἀποδίδωσί τι χρέος τῆς δημοσίας κατάρας, and so does, e.g., Butler, and similarly
Wecklein : ‘ "the party concerned pays the debt" .... the subject to rive like
the subject in 71 is to be supplied from the context ; the person they have in
mind is the man who has aroused the people's ill-will.’ I do not regard this
interpretation as impossible. But since in 451 the πρόδικοι Arpeidaı are the
objects of the malevolent attack and in 461 the πολυκτόνοι are blamed, it is
not very probable that in between these there should be an anonymous
singular subject to τίνει, and a singular which would in fact be the equivalent
of the plural. Klausen's and Hermann's interpretation is far more con-
vincing: 'sermo civium infensus imprecationis quam populus ratam facit
negotium solvit' (Klausen). The uttering of the curse is to be imagined as
having happened in the past, possibly during the earlier stages of the war:
it is here treated as equivalent to the contracting of an obligation. The
obligation to pay is redeemed at the moment when the stored-up resentment
of the people finds voice in the φάτις σὺν κότωι: this is the first step towards
revolt, thus the φάτις brings about the payment. So, too, the curses of Oedipus
contain the demand for fulfilment (Sept. 766 f., text uncertain); there the
curse itself exacts the payment due: 840 ἐξέπραξεν οὐδ᾽ ἀπεῖπεν πατρόθεν
εὐκταία φάτις.
Several commentators, while taking φάτις as subject of τίνει, understand
δημοκράντου ἀρᾶς differently, as epexegetic genitive to χρέος, e.g. Enger
(1855), ‘pays the debt of a people’s curse, 1.6. is punished as a people’s curse,
counts for as much’; Nägelsbach, ‘it (the talk of the citizens) takes the place
of a curse ordained by the people’; Paley, ‘it performs the part of a ban
solemnly ratified by the people’; Hcadlam, ‘it performs the office of a curse
publicly decreed’. This interpretation, as compared with that discussed
above, has the great disadvantage of weakening the technical sense of χρέος
τίνειν. It would imply that the φάτις raises a claim to a later fulfilment, 1.6.
punishment of the guilty by the gods. But rive χρέος suggests that a final
satisfaction for the creditor follows forthwith.
459 f. The reading of the MSS (μου) is translated by Stanley: ‘exspectat
autem audire mea sollicitudo aliquid occultum’, so, e.g., Verrall (‘And I am
expecting in trouble to hear of some secret of the dark’) and Headlam (‘My
anxious thought awaits the news ...’). In this case, as Verrall did not fail to
notice, the position of μου is suspicious: at any rate I have not found any
example in Aeschylus of a μου preceding the noun on which it depends, not
even a case so allowable as S. Oed. C. 582 ὅταν θάνω ᾽γὼ καὶ σύ μου ταφεὺς
γένηι (in Prom. 133 μου depends on ἐξέπληξε; in Suppl. 785 Dindorf's bracket-
ing is rightly adopted by Wilamowitz, but even if μου καρδία were genuine,
it would be a case of μον being put between the attribute and the noun, and
therefore of a different type of word-order). This, however, might be mere
234
 COMMENTARY lines 461 f.
accident. The most weighty objection is that the context here suggests
emphasis on the continuance and persistence of fear. The sphere of curse and
requital holds a central place in Aeschylus' thought, and the connexions of
ideas in it are very strong and uniform. This being so, one must not overlook
the fact that the beginning of the third stasimon works out with great power
the idea of the persistence of anxiety and fear which cannot by any means be
chased away (975 ff.): τέπτε μοι τόδ᾽ ἐμπέδως δεῖμα... πωτᾶται κτλ. ‘Und
mir bleibt die Sorge nächtlich-dunkle Kunde zu vernehmen’ (Nägelsbach),
‘and for me the anxiety abides to hear night-darkened tidings’—this is the
sense required here, therefore Karsten’s μοὶ is necessary. Aeschylus has a
liking for μένει with dative: Suppl. 434 f. παισὶ τάδε καὶ δόμοις... μένει "Ape
τίνειν ὁμοίαν θέμιν, Ag. 1149 ἐμοὶ δὲ μίμνει σχισμός, Eum. 894 τίς δέ μοι τιμὴ
μένει; cf. Eum. 496 f. πολλὰ... πάθεα προσμένει τοκεῦσιν μεταῦθις ἐν χρόνωι.
The examples show also that an infinitive or a noun can be the subject for
this μένει μοι. In the present instance we may perhaps suppose that ἀκοῦσαι
does not depend on μέριμνα but on the phrase μένει μοι μέριμνα.
460. νυκτηρεφές : found only here.
In the scene of the Antigone quoted above on 449 Haimon says (692) epoi
δ᾽ ἀκούειν ἐσθ᾽ ὑπὸ σκότου rade . . . then (700) concluding: τοιάδ᾽ ἐρεμνὴ σῖγ᾽
ἐπέρχεται φάτις. There is in that passage perhaps ἃ reminiscence of phrases of
this ode: 449 σῖγά ris βαὔζει, 456 ἀστῶν paris, 459 f. ἀκοῦσαΐ τι vuxrnpebés. If
this is correct, then Sophocles has given quite a different sense to the last of
these passages, in order to fit it to the context of the φάτις : in Aeschylus it is
the avenging stroke of destiny which is the object of ἀκοῦσαι.
461 f. οὐκ ἄσκοποι as in Homer £2 157 οὔτ᾽ ἄσκοπος, belonging perhaps to
the group of privative compounds which in Homeric and post-Homeric
Greek are mainly found in connexion with a negative particle (cf. Wacker-
nagel, Syntax, ii. 297 f.). In Parmen. B 7. 4 ἄσκοπον ὄμμα (for the meaning cf.
H. Frankel, Gott. gel. Anz. 1928, 271 n. 2, and Nachr. Gótt. Ges. 1930, 170) the
adjective is subservient to the oxymoron of the phrase. ἄσκοπον by itself
in passive sense is found in Cho. 816 and so in Sophocles; for the active as
well as passive function of privative verbal adjectives cf. on 238. As here
the avenging gods keep an eye on the evil-doers, so the Erinyes, to whom the
thought passes on directly, are called in Eum. 499 βροτοσκόποι μαινάδες. The
regard or consideration of a god for men or his places of worship is commonly
expressed by ἐπισκοπεῖν (E. Phoen. 155 σκοπεῖν), ἐφορᾶν, ἐποπτεύειν, and the
like (corresponding to respicere); evidence from Aeschylus is given in W.
Schmid, Gesch. d. griech. Lit. ii. 272 n. 6, cf. also on 14 and on 1270.
It would be quite contrary to the poet's intention if we were to conclude
from the words τῶν πολυκτόνων γὰρ οὐκ ἄσκοποι Geot, together with the pas-
sages in which the grimness of war is deplored, that the war against Troy is
‘una impresa in cui Agamennone ha sacrificato all’ orgoglio della casa e alla
propria ambizione l' interesse dei suoi popoli' (G. De Sanctis, Storia dei
Greci, 1939, ii. 90). It is in fact twice said most emphatically (60 ff., 362 ff.)
that Zeus, the protector of justice, has dispatched the Atridae on the errand
of vengeance. Nevertheless the gods are angry because of the much blood-
shed. This is only one of the irreconcilable contradictions that present them-
selves in the dilemma of Agamemnon (206 ff.), in that of Orestes, and of other
heroes in Aeschylus.
235
line 464 . COMMENTARY

464. ruxnpös according to the dictionaries is in the classical period found only
in this passage and twice in prayers in Aristophanes (Ach. 250, Thesm. 305;
the latter in prose), which indicates perhaps that the word originated in
religious language : it occurs afterwards once in Aristotle, and more frequently
in Hellenistic Greek.
Tuxnpôv . . . δίκας : ‘who is successful without justice’.
465. τριβᾶι: in the lexicons of Wellauer and Dindorf 'attritus' is rightly given
as the only meaning for this passage,! which is compared with Cho. 943. Most
commentators, however, take τριβή here as 'cursus vitae' (Klausen), 'Lebens-
führung' (Nágelsbach), or the like. This is certainly wrong, if only because
τρίβειν βίον (much the same as aetatem terere, etc.) as a rule does not indicate
a way of living in a neutral sense, but is disparagingly used of a dragging life
which is wearisome, miserable, purposeless, etc. (e.g. S. El. 602, E. Heraclid.
84, Ar. Peace 589, Plut. 526, Eubulos fr. 68. 2 Kock, Herodas 3. 52) : and this
makes no sense here, where it is a question of the catastrophic change to
disaster. τριβὴ βίου corresponds to the κτεάνων τριβή which the house of the
Atridae has to suffer under the usurping tyranny (Cho. 943). τριβᾶι is dative
of the instrument ; that such a means is especially appropriate to the Erinyes
is shown by 1573, where the Spirit of the curse is entreated ἄλλην γενεὰν
τρίβειν θανάτοις αὐθένταισιν. It seems possible that the imagery employed in 391
in τρίβωι (for 395 πρόστριμμα see p. 207) is echoed in this τριβᾶι: 1t is given
a somewhat different turn, but here, too, the basic idea is that of ‘wearing
away'.
466. τιθεῖσι : this passage is the only one quoted in Kühner- Blass, ii. 192, and
L-S s.v. τίθημι for the occurrence of this Epic-Ionic form (cf. H. W. Smyth,
Ionic Dialect, 575; Bechtel, Griech. Dial. iii. 179; Schwyzer, Gr. Gramm. i.
688) in Attic poetry.
ἀμαυρόν: in certain passages of the tragedians the meaning 'dark' seems to
be predominant, in others the meaning ‘weak’ (cf. Wilamowitz on E. Her.
124), but sometimes the word might contain an idea of both combined. Here
anyhow the sense is: 'cause him to grow faint and dim' (Headlam). The
context shows that the point here is weakness; on the other hand, the
continuation ἐν δ᾽ ἀΐστοις τελέθοντος shows clearly that the lack of visibility
is essential.
466 f. ἐν δ᾽ ἀίστοις τελέθοντος : Wecklein, Stud. z. Aesch. (1872) 23, well
observes ‘it broadens and heightens the preceding statement (ὥστε ἐν ἀίστοις
TeÀéÜew) . . . τελέθοντος is not to be taken generally, but as referring to the
person punished'.
467 f. οὔτις ἀλκά: cf. on 381.
469 f. ὄσσοις was regarded by Bothe as corrupt, and even so great a gram-
marian as Ahrens (p. 528) agrees with him. Hermann and Meineke (PAilol.
xix, 1863, 196) have indeed retained it, but put a strain on the syntax by
taking it as a dative of the object aimed at: ‘praestringit oculos Iovis fulmen’,
so Sidgwick ‘upon his eyes’. The right interpretation was given by Klausen
(whom Paley followed) : he regarded ὄσσοις as the dative of the instrument
and compared 510 τόξοις ἰάπτων . . . βέλη. Zeus needs no external instrument,
the divine eye is weapon enough. Grammatically Διόθεν goes probably with
1 On the general sense Linwood is right (‘the destruction of life, reversing his fortune") ;
so is Headlam.
236
 COMMENTARY line 478

κεραυνός (see M. Lejeune, Les Adverbes grecs en -θεν, 156: ‘le tonnerre lancé
par Zeus’), cf. Suppl. 437 Διόθεν κράτη, but in this context it has the effect of
qualifying the eyes as the eyes of Zeus. Klausen has seen that we have here
a transference of the notion of an evil eye: cf. on this point Wilamowitz,
Verskunst, 187, and Interpr. 37 n. 4. On the other hand, in 947 (compared by
Klausen), where the words μή τις πρόσωθεν ὄμματος βάλοι φθόνος are used with
reference to the gods, there is unmistakably an allusion to the lightning-
stroke. The fact that the idea of φθόνος is dominant in the sentence Ag. 469 f.
is proved by the words directly following on it, κρίνω δ᾽ ἄφθονον ὄλβον. A
striking counterpart of the Aeschylean passages is provided by Hdt. 7. το. e
ἐπεάν aft ὁ θεὸς φθονήσας φόβον ἐμβάληι ἢ βροντήν. For the religious back-
ground cf. A. B. Cook, Zeus, ii. 503. What is a live conception in the poet of
an early age becomes a simile for later writers: Lucr. 5. 1125 f. et tamen e
summo, quasi fulmen, deicit ictos invidia interdum contemptim in Tartara
taetra (quoted by Schneidewin), Culex, 341 f. (in the description of the Return
from Troy) omne $ropinquo frangitur invidiae telo decus.
471. κρίνω: cf. L-S s.v. II. 7 ‘decide in favour of, prefer, choose’, with several
illustrations.'
ἄφθονον : used as ἀφθόνητος 939, ‘unenvied’. Though ἄφθονος = ‘free from
envy' is occasionally found (L-S s.v. I. 1), yet here the poet indulges in a
somewhat bold reinterpretation of the etymologizing type, for in general
usage ἀῴθονος had long acquired the sense of ‘abundant’, and in association
with a word like ὄλβος this meaning was almost bound to suggest itself to
an ancient audience, cf., e.g., Solon fr. 23. 5 D. πλοῦτον ἄφθονον. Cf. Schuursma,
De... abusione, p. 153.
472. μήτ᾽ εἴην πτολιπόρθης (taken up in ἁλούς following): this thought per-
forms a twofold function here: to begin with, it offers simply an example
(hence the explanatory asyndeton) of the ὄλβος μετὰ φθόνου, but next it leads
the imagination of the audience, after the long series of general considera-
tions, back again to the Τροίας ἅλωσις, and thus forms a transition to the
epode.
πτολιπόρθης with this termination only occurs here; the usual Homeric form
appears in 782.
473 f. The plain βίον karióou would certainly be strange, if it stood by itself;
but here this is not really the case, for the audience will have learned from the
words ἁλοὺς ὑπ᾽ ἄλλων what kind of life is intended as the second of the two
extreme cases. No alteration therefore is needed.
478. Hermann made the certain correction τι, and further adopted the εἰ of
the corrector of F: εἴ τι θεῖόν ἐστι μὴ ψύθος; this he translates: ‘verane, quis

! A. Suppl. 396 does not belong to this category. In L-S the explanation of Paley is
adopted : *prefer that which is righteous in the sight of the gods’, so, e.g., Hartung, Wecklein
(1902) : ‘decide in favour of what the gods hold to be holy.’ But this interpretation breaks
down because, as commonly happens in such 'epirrhematic' passages (cf. my note on Ag.
1412 f.), the beginning of the King's rejoinder directly refers to the concluding lines of the
preceding section of the Chorus; and there the words εὔκριτον, κρῖμα, κριτής can only be
understood of actual judgement and judicial decision. Wilamowitz's paraphrase of 395f.,
δίκαζε ξύμμαχον ἑλόμενος τὴν εὐσέβειαν, involves an impossible order of words: ξύμμαχον
ἑλόμενος Δίκαν belong together. Headlam is right: ‘Take thou Justice upon thy side and
judge that is righteous before heaven' (Tucker wavers). σέβας depends on κρῖνε as cognate
accusative (— κρίσιν εὐσεβῆ).
257
line 478 . COMMENTARY
scit? nisi divina aliqua fraus est.” Headlam agrees with Hermann about the
text and the interpretation, and adds by way of explanation ' et ru... ψύθος
is added as an afterthought'. Such a correcting addition would be perfectly
suitable if the preceding words expressed a confident belief, or at least a
strong hope, that the information was true. But after the undisguised doubt
(εἰ δ᾽ ἐτήτυμος, τίς οἶδεν), the words nis: aliqua fraus est make a very lame
continuation (as was rightly stressed by Schoemann, Ofusc. iii. 173). On the
other hand, if we try to give the reading ἤ τι θεῖόν ἐστι μὴ ψύθος a chance,
we might compare the apparently superfluous use of a negation found in
dependent clauses following a negative (or an implied negative) in the main
sentence, for which see, e.g., Kühner-Gerth, ii. 206 n. 6, Wackernagel, Syntax,
ii. 308. But the genuineness of this μή is rendered suspect by its position ;!
the examples from Sophocles which have been collected by the commentators
on Oed. R. 329, El. 993, Oed. C. 1365, etc., are by no means parallel. Ahrens's
ane is wrong, apart from other reasons, because an: indefinite is nowhere
found in Aeschylus or Sophocles, or indeed anywhere in Tragedy as far as the
lexicons tell us. (For the same reason Enger’s my in Ag. 347 cannot be
considered.) The simplest emendation is Dindorf's deletion of the μή. We
cannot tell what caused its intrusion (Wilamowitz rightly gave up his earlier
suggestion, published in his Comment. metr. il. 10 n. 1, ἐστιμ ψύθος, cf. on 387) ;
it is conceivable, e.g., that someone may have thought a regular trimeter was
required here before that of the following line.*
479-82. τίς ὧδε... καμεῖν; the construction is very unusual. Klausen is
not really helpful: ὧδε sequente infinitivo, ut οὕτω E. Phoen. 380’, for there
(οὕτω yàp ἤρξατ᾽, ἄνομα μὲν τεκεῖν ἐμέ κτλ.) the infinitive depends on the verbal
expression οὕτω ἤρξατο. Nor is Wecklein, 'for the epexegetic infinitive after
ὧδε cf. K. W. Krüger, Sprachlehre, i. 57. 10. 9', for in the passages discussed
there and in Krüger's commentary on Herodotus a phrase composed of a
principal verb and οὕτως or ὧδε is explained by a subsequent infinitive clause,
as, e.g., in Hdt. 5. 95. 2 κατήλλαξε δὲ ὧδε, νέμεσθαι ἑκατέρους τὴν ἔχουσι, Xen.
Cyr. 8. 7. 10 ὑμᾶς... οὕτως ἐξ ἀρχῆς ἐπαίδευον, τοὺς μὲν γεραιτέρους προτιμᾶν
κτλ. In the case before us, however, the ‘epexegetic’ or, if you like, the
‘consecutive’ infinitive forms the complement to the definition of quality
ὧδε παιδνός. Parallels to this construction, as far as I have been able to find
out, are fairly common only in Homer, where definitions of quality look for
their completion to an infinitive, e.g. p 20 οὐ γὰρ ἐπὶ σταθμοῖσι μένειν ἔτι
τηλίκος εἰμί, B 60 ἡμεῖς δ᾽ οὔ νύ τι τοῖοι ἀμύνεμεν, n 309 f. οὔ μοι τοιοῦτον ἐνὶ
στήθεσσι φίλον κῆρ μαψιδίως κεχολῶσθαι (cf. also Leaf on T 140). In addition
there is as least one instance that I have observed in Tragedy (possibly there
are more) which may be said to stand half-way between these Homeric
examples and those discussed above with groups of οὕτω (or ὧδε) and a verb:
E. Hypsip. fr. 60, col. 1, 1. 45 f. (v. Arnim, Suppl. Eur. p. 61, Ὁ. L. Page,
! Poppo drew attention to the strange position of this μή as early as 1821 in the Pro-
legomena to his Thucydides, i. 303.
2 Cf. Eum. 275 δελτογράφωι «rA., where in FTr one syllable has been added to the two
dochmiacs, preserved in M, so as to tum them into an iambic trimeter, though a faulty one
(Porson's law was apparently unknown between Seneca and Porson; see P. Maas, ‘Text-
kritik’, in Gercke-Norden, Einleit. in d. Alterstumsw. i. 2, $ 27). An iambic trimeter precedes
in 273, and 275 is followed by a dialogue in trimeters. Tr turns the beginning of the stanza
Ag. 737 = 750 into a regular trimeter by inserting in 737 οὖν and in 750 τοῖς,
238
 COMMENTARY lines 479-82
Greek Lit. Papyri, p. 100 1. 230 f.) καὶ πέφυχ᾽ odrws,' γύναι, κοσμεῖν 7’ ἐμαυτὸν
καὶ τὰ διαφέρονθ᾽ ὁρᾶν. These instances, however, are so to speak less offensive,
though similar in principle, because in them the infinitive-complement is far
more succinct. It is therefore understandable that in Ag. 479 ff. grammarians
have had recourse to the common expedient of scholiasts ‘ λείπει τὸ Gore ’
(in Tr there is an interlinear gloss ὥστε at 480 above φλογός). Thus Goodwin,
§ 600: ‘ ὥστε seems to be omitted’, while Paley complains: ‘the omission of
wore is very harsh.’ To assume such an omission can hardly pass for a proper
explanation. Nor is the tentative suggestion of Paley really satisfactory, that
‘we might place a question at κεκομμένος, and regard what follows as an
infinitive of exclamation’. ris ὧδε παιδνὸς ἢ φρενῶν κεκομμένος is incomplete
in itself and finds its completion in what follows: it is the ‘childish’ behaviour
described in 481 f. that justifies the sharp expressions used in 479. Now if we
consider once more the expedient of ‘ λείπει τὸ wore’, We can see that even
the addition of &ore would not settle the matter. For in that case we should
be bound to expect πυρωθείς instead of πυρωθέντα, at least according to the
rules of Attic prose (cf. also, e.g., the nominatives in Ag. 1589 f., not after
ὥστε but after τὸ μή), because the person that is the subject of πυρωθέντα....
καμεῖν is the same that is called παιδνός, whether we are thinking of an
individual or a typical character. Lobeck’s statement in his edition of
Phrynichus,p. 750 n. (cf. also Krüger on Thuc. 1. 12. 1), does not afford suffi-
cient ground for questioning the rule for good Attic prose ; in cases like Thuc.
7. 34. 6 (αὐτοὺς ἑκατέρους) the special reason for the accusative is immediately
obvious. Wecklein's comment, 'the acc. πυρωθέντα in accordance with the
meaning παιδνόν ἐστι mupwdevra . . . καμεῖν ', is merely an excuse made up
ad hoc. Therefore it might be as well not to couple the instance before us
with uses of the infinitive introduced by wore. As compared with these, the
addition of a plain infinitive with an epexegetic function presumably repre-
sents an older stage of language. In this connexion we may recall that, as
Thiersch and afterwards Lehrs (De Aristarchi stud. Hom., 3rd ed., 157)
observed, the use of ὥστε in Homer in the sense of ‘ita ut’ is of extremely rare
occurrence: accordingly we find in Homer, as we have seen, the plain in-
finitive used to indicate result or explanation? Here Aeschylus has merely
extended the use of the infinitive clause beyond the limits customary in
Homer. The infinitive in this passage has the same epexegetic function as the
relative clause in Prom. 159 f. : τίς ὧδε τλησικάρδιος θεῶν ὅτωι τάδ᾽ ἐπιχαρῆ ;
As far as the accusative πυρωθέντα is concerned, there is probably a parallel
to it in the Oresteia itself ; one cannot, however, claim it with certainty, for
the passage, Cho. 481 f., is a highly controversial one. I think myself, with
Conington, Blass, and others, that there, apart from the missing word at the
end (πόνον Enger, though there are other possibilities), the MS gives us the
authentic reading: κἀγὼ πάτερ τοιάνδε σου χρείαν ἔχω, φυγεῖν μέγαν προσθεῖσαν
Αἰγίσθωι (móvov) ; but other commentators, e.g. Wilamowitz, Headlam (note
to his prose translation), and Dodds, C.Q. xxxii, 1938, 2 f., suppose the corrup-
tion to lie deeper. The particular detail with which we are here concerned,
1 Plain πέφυκε (or ἔφυ) with an explanatory infinitive is of course common, cf. L-S
φύω B. II. 2.
2 There is no instance of dere with infinitive in the extant remains of the Lesbian poets
(Snell, Hermes, lxvi, 1931, 73 n. 1).
239
lines 479-82 COMMENTARY

the question of accusative or nominative, enters into the discussion in Cho.

482 also: e.g., Tucker says: ‘the accusative participle is ungrammatical’,
Headlam (and similarly Dodds): ‘if the accusative participle is correct, I
think [as Canter and many after him had done] the sentence must have
contained pe’, whereas Wilamowitz writes θεῖσαν with Emperius without
adding a pronoun, thus keeping to the construction given in the MS. The
infinitive here does not directly serve to express a prayer (which would in
itself be possible), but is joined epexegetically to χρείαν σου ἔχω (τοιάνδε does
not look forward, but is related to the preceding prayer of Orestes); the
scholion gives the customary kind of paraphrase: ὥστε φυγεῖν τὰς ἐπιβουλὰς
Αἰγίσθου τιμωρησαμένην αὐτόν. The accusative προσθεῖσαν and the accusative
πυρωθέντα Ag. 481 perhaps give each other mutual support.
479. φρενῶν κεκομμένος : one would like to know what is the function of the
genitive here. Of course a fifth-century Greek would not have classified the
case according to grammatical categories, but all the same we may ask how
he felt about it, i.e. with what analogous constructions he would, perhaps
unconsciously, connect this one. ‘ “Stricken in mind”, gen. of reference’, is
Sidgwick’s comment ; Plüss is more cautious: ‘Gen. of separation? Gen. of
the sphere within which the predicate applies?’ The separative function
seems to me more likely (so also Kühner-Gerth, i. 396). This is supported by
kindred expressions, cf. on 120 βλαβέντα λοισθίων δρόμων and, e.g., Theognis
223 νόου βεβλαμμένος ἐσθλοῦ (quoted by Blomfield), γος βλάπτουσα νόοιο, Pind.
Ol. 1. 58 εὐφροσύνας ἀλᾶται, and, made easier by the prepositional compound,
Phryn. Praep. soph. p. 103. ı de Borries παρακεκροῦσθαι τῶν φρενῶν ἢ τοῦ νοῦ,
and παρακόπτομαι middle in Ar. Knights 807 οἵων ἀγαθῶν... παρεκόπτου:
‘swindle out of a thing’ (L-S). For other separative genitives cf. on 1023.
The meaning of φρενῶν is here the same as in 175.
481. νέοις : ‘unheard-of’ as often.
πυρωθέντα (cf. 440): ‘set on fire, inflamed’, here used contemptuously, in
connexion with the generalidea. The continuing of the fire imagery from 475
onwards has been well expounded by Headlam; similarly Neustadt, Hermes,
Ixiv, 1929, 260 n.
ἔπειτα after the participle here, just as, e.g., in Sept. 267, keeps its full
temporal force, and so it does in Eum. 654; but in these passages of the
Oresteia we can see how the way is being prepared for ἔπειτα to acquire the
adversative meaning (‘and then still’, ‘then however’) which is well known
from idiomatic expressions such as Ar. Ach. 498, Birds 29, and the like.
483. αἰχμᾶι: here Blomfield started the trouble from which many com-
mentators have not yet recovered. He finds fault with Stanley and Schiitz,
who understood it to mean muliebre imperium, and explains ‘ γυναικεία αἰχμή
est indoles feminea’, with a reference to Cho. 630. Now in that passage (riw...
γυναικείαν ἄτολμον αἰχμάν), in accordance both with the context of the words
and with the assumed situation in general, it is perfectly clear that the Chorus
is reproaching itself for paving respect to the ‘rule of a woman’, by which is
meant that of Aegisthus, cf. infra on 1625 ff. (τίειν is used in an exactly corre-
sponding connexion in Ag. 259). Both passages were rightly understood by
Wellauer, who rejected Blomfield’s interpretation. Prom. 405 has long been
cited for the meaning of ‘rule’ for αἰχμή. But Paley, Sidgwick, Verrall,
Headlam (also in his translation of Cho. 630), L-S (though hesitatingly) stick
240
 COMMENTARY line 485
to the rendering ‘a woman’s temper’, which not a single passage actually
requires; some indeed regard it as quite a different word from αἰχμή, ‘hasta’
(after Ahrens, 530). The transition is a simple one from the designation of the
weapon to the meaning ‘rule’, especially in elevated style. Weil (Addenda at
the end of his first edition of the Eum. p. 130) refers appropriately to E. Hipp.
975, Hec. 9, where δόρυ stands for ‘sceptre, rule’, cf., e.g., Lucan 5. 61 saevum in
populos puer accipis ensem. Of course αἰχμή in association with γυναικός
sounds a malicious note, which is in keeping with the bitterness of the whole
passage. For the particular function of the expression cf. on 484.
484. χάριν ξυναινέσαι: Schneidewin rightly says: ‘Dank zugestehen'. For
συναινεῖν cf. Ahrens 532, L-S also correctly ‘concede’. χάριν, directed towards
the gods, as, e.g., in 821. Clytemnestra is being blamed because immediately
after the arrival of the beacon-message, and before the true facts have
appeared (πρὸ τοῦ φανέντος), she has sacrificed thank-offerings (87 ff., 262) and
thereby represented something to the citizens (476) as certain which is
perhaps only an illusion. Her sacrificing cannot be regarded as a private
action: coming as it does from one who takes the ruler’s place in his
absence (259f.), it is a demonstration of authority. A lawful sovereign,
experienced and conscious of responsibility, would have behaved with greater
caution and would have waited a while before arousing dangerous illusions.
It may now have become clear why the phrase γυναικὸς aiyu& is used here.
485. πιθανὸς ἄγαν κτλ. The real difficulty of this sentence, due to the
obscurity of ὅρος, has been needlessly increased by careless treatment of
πιθανός. The great majority of commentators (including Wilamowitz, Griech.
Verskunst, 186) as well as the lexicons of Wellauer, Passow, and L-S have
adopted Stanley’s translation credulus, which 87 and 591 seem to support.
Schiitz on the contrary takes it: ‘ad persuadendum accommodatus’, so
Bothe, Linwood, Franz, Weil (Addenda at the end of his first edition of the
Eum. p. 130), and Dindorf in the Thesaurus and the Lex. Aesch. This is the
only meaning which accords with the bulk of the evidence from the earlier
period. Euripides and Aristophanes, Herodotus and Thucydides use the
word solely as corresponding to πείθειν, not πείθεσθαι. For the ‘passive’ use
there seems to be no certain example before Xenophon.’ For though L-S
quote Plato, Gorg. 493 a, with Ag. 485 under the heading ‘easy to persuade,

™ There is, however, evidence in Hellenistic poetry (not mentioned in L-S), Pap.
Tebtun. i. p. 3 (= Diehl, Anth. lyr. ii. 297 l. 5) πιθαναὶ δ᾽ épyariBes σιμοπρόσωποι, ξουθόπτεροι
μέλισσαι. In contrast with the correct translation of Grenfell and Hunt ‘the willing busy
bees', Wilamowitz's comment is unfortunate, Timotheos, 83 n. (taken over by W. Schubart,
Einführung in die Papyruskunde, 132, and Diehl, loc. cit.): ' πιθανός no longer = πείθειν
δυνατός, but = χαρίεις, as Asclepiades (possibly Posidippus) calls a hetaira, Anth. Pal. 5.
158.' For the latter he might have quoted the well-known prologue of Menander's Thais
(fr. 217 Kock), where the speaker describes the hetaira as θρασεῖαν ὡραίαν δὲ καὶ πιθανὴν
ἅμα xrÀ. ; the passages may be classed under the heading I. 3 in L-S (‘of manners, winning,
plausible). The transition of meaning can easily be understood: the effect of her πείθειν,
not merely by means of words, is such that nothing can resist her, thus πιθανός becomes
something like ‘irresistible’, ‘bewitching’. But as far as the ἐργατίδες μέλισσαι are concerned,
they are termed ‘obedient, submissive’ like the men of whom Xen. Cyr. 2. 2. 10 says πιθανοὶ
δ᾽ οὕτως εἰσί τινες ὥστε πρὶν εἰδέναι τὸ προσταττόμενον πρότερον πείθονται. The discipline of
the worker-bees is continually being stressed ; πιθανός is very appropriate, as is seen, 6.8.»
from Aelian, Nat. anim. 5. 11 6 μὲν βασιλεὺς (of the bees) μιᾶι προστάττει ὑποσημῆναι κατα-
δαρθάνειν" καὶ ἡ μὲν πεισθεῖσα τοῦτο ἐκήρυξεν κτλ.
4872.2 R 241
line 485 COMMENTARY
credulous’, yet a glance at the context shows that although τὸ πιθανόν is meant
to carry on the notion of ἀναπείθεσθαι, yet a middle-passive force of the
adjective, if it is to be thought of at all, can only be regarded in that passage
as a momentary freak. Probably τὸ πιθανόν there is simply to be taken in the
ordinary active sense. Plato is preparing the way for his etymologizing play
on the word idos, in which, by way of securing a more exact formal corre-
spondence, he inserts in front of τὸ πειστικόν the word πιθανόν, which is
intended here perhaps as complementary: ‘on account of the action of the
πείθειν and of the πείθεσθαι resulting from it’. In any case this transient play
on words does not allow us to infer a use of πιθανός which is as inconsistent
with that of Plato himself (who uses the word very often) as it is with that of
the earlier authors. Now as regards the Agamemnon passage, it is of no use
to try to make a case for an isolated meaning on the ground that this is really
the oldest! extant example of the word and that we must not tie the poet
down to the later usage. For it can hardly be doubted that in the year 458
πιθανόν had its place in the Attic vocabulary within the sphere in which it is
used by Aristophanes and Thucydides. In Thuc. 3. 36. 6 Cleon is described,
with reference to his success in the Assembly, as τῶι δήμωι παρὰ πολὺ ἐν τῶι
τότε πιθανώτατος and, exactly parallel, 4. 21. 3 τῶι πλήθει πιθανώτατος. In
6. 35. 2 in the description of the Assembly at Syracuse, the word is applied
to the leading politician Athenagoras : ev τῶι παρόντι πιθανώτατος τοῖς πολλοῖς.
Apart from this stereotyped ‘formula’ Thucydides does not use the word.
We find it in Aristophanes, first Knights 629, used in the superlative exactly
as in Thucydides about an oratorical success of Cleon’s, this time in the
Council. Then in Thesm. 268 to speak πιθανῶς is included among the rules for
conducting oneself in a quasi-Assembly, and in 464 πιθανά is used in approval
of a speech delivered by one of the women citizens. The same applies to
πιθανός in E. Or. 906 as a characteristic of a demagogue who appears in the
Ecclesia of the Argives (Eur. fr. 396. 2 approximates more closely to Hero-
dotus’ usage). Herodotus (in seven passages according to Powell) speaks
of a λόγος, a piece of news, etc., as πεθανός either in itself or for some particular
recipient. We may conclude from the passages of the Attic authors that the
word was chiefly used, in fifth-century Athens, to describe the effect of
speeches delivered in the Assembly or the Council, especially by politicians
with whom Πειθώ τις ἐπεκάθιζεν ἐπὶ τοῖς χείλεσιν. So it is highly appropriate
that Aeschylus should use it here with reference to the act or conduct of the
ruler’s vicegerent. But even if we do not take into account the general usage
of the word, the conclusion of the sentence, γυναικογήρυτον ὄλλυται κλέος,
should leave no room for doubt as to the meaning required here for πιθανός.
The sentence ‘a rumour that is put about by a woman’s utterance makes its
way quickly, and as quickly perishes’ would be inappropriate for continuing
a statement about woman’s credulity, for the medium of undiscerning belief
is not a γηρύεσθαι. On the other hand, if γυναικογήρυτον is referred back to
the πιθανότης of a woman’s speech, a fine point is brought out: a woman can
easily talk men round, but this γυναικογήρυτον κλέος soon passes away.
ἐπινέμεται, too, rightly understood (see below), is only suitable for an active
1 An additional instance has recently been discovered from Aeschylus’ Γλαῦκος Ποτνιεύς,
Pap. Oxy. 2160, fr. 8. 16, but the context in which πιθανας stood there cannot be ascer-
tained,
242
 COMMENTARY line 485

sense of πιθανός : 'the . . . (the substantive must be left unspecified for the
moment) of a woman spreads abroad quickly, like a sweeping flame’: it
spreads abroad, not because the woman who disseminates it is credulous, but
because she succeeds in πείθειν, so that the number of those who are overtaken
by her persuasion (as by a flame or a pestilence), i.e. those who believe it, is
always increasing. Lastly, with regard to the sentence πιθανὸς ἄγαν κτλ.
independently of its particular context here, it is obviously easy enough to
provide illustrations of the credulity of women, e.g. Tac. Ann. 14. 4 facili
feminarum credulitate ad gaudia (Paley). But the feminine aptitude for
deceptive πιθανόν is at least as common an idea with the ancient Greeks: we
find it contained in a nutshell in the well-known proverb in Hesiod, Erga 375
ὃς δὲ γυναικὶ πέποιθε 'πέποιθ᾽ 6 γε φιλήτηισιν or ibid. 77 f. in the account of
woman's coming into the world: Hermes fashions in her breast ψεύδεά θ᾽
αἱμυλίους τε λόγους.
ἐπινέμεται: the verb is often used of the spreading (grassarı) of fire (and
that is clearly appropriate here, cf. on 481) or of a pestilence. The traditional
view that this use of the word applies to the present passage was challenged
by J. W. Donaldson, New Cratylus (Cambr. 1839), 269 ; according to him the
passage means: ‘from excessive credulity the boundaries of a woman’s mind
are easily encroached upon’. He was followed by Paley, Dindorf (Lex. Aesch.),
L-S, but wrongly, for the combination ἐπινέμεσθαι ὅρον is doubtful, and still
more so the passive use of the verb with this meaning (for the different phrase
ἐπινέμειν βοσκήματα and the like see the lexicons). Headlam, C.R. xvi, 1902,
441, conclusively refutes Donaldson and gives fresh support to the old inter-
pretation grassatur.
ὁ θῆλυς ὅρος: very difficult. The scholion (ZxoA. aA. in Tr) says ἤγουν
περιφραστικῶς ἡ γυνή" ὡς ταὐτὸν ὃν γυναῖκα εἰπεῖν καὶ ὅρον αὐτῆς ἐκθεῖναι
(‘definition of woman’ periphrastic for ‘woman’), which is a fairly stupid
piece of pedantry : the commentator was simply using the favourite! formula
of explanation, περιφραστικῶς.... Even so the meaning here assigned to
ὅρος, though belonging to a much later period (for preparatory stages cf.
Wilamowitz on E. Her. 669), is at least one that is found in Greek, while the
commentators of the seventeenth and eighteenth centuries invented an ad hoc
meaning (muliebre iudicium Stanley, similarly Schütz). Blomfield refused to
adopt this and preferred to follow the scholiast, but then added : 'sed forte
legendum ὁ θῆλυς ἔρος '. This proposal has had a considerable vogue, and is to
be found still, e.g., in the texts of Wecklein, Headlam, and Mazon. Yet épos
introduces not only a form without example in Aeschylus (cf. on 375), but also
with the sense 'desire' an idea which is quite out of keeping with the context
of this passage, though it could be combined with it by way of elaborating
detail as was attempted, e.g., by Ahrens: ‘the women of Argos had of course
the liveliest ἔρως for the homecoming of their menfolk' ; a similar explanation
is given by Neustadt, Hermes, lxiv, 1929, 260 n. Moreover this conjecture is
connected with the middle-passive sense of πιθανός which has been shown to
be untenable. P. Von der Mühll, 74g. Forsch. 1, 1932, 139, has postulated an
1 Cf., e.g., Schol. E. Phoen. 1167 ἐπάλξεων ἐρίπνας" περιφραστικῶς τὰς ἐπάλξεις, Schol. Pind.
Ol. 3. 24 κόσμον ἐλαίας" περιφραστικῶς τὴν ἐλαίαν. Numerous examples of this type are
collected from the scholia on Euripides in Ed. Schwartz's Index (ii. 314 s.v. περίφρασις),
from those on Pindar in Drachmann (iii. 361 s.v. περιφραστικῶς).

243
line 485 COMMENTARY
unattested word ὄρος (from ὄρνυμι) = ὁρμήϊ for Ag. 485 and 1154; with the
majority of commentators he regards Hermann’s interpretation of the MS
reading as untenable. Hermann retains ὅρος, but gives it a new interpretation:
* ὅρος est decretum, referturque ad sacrificia Clytaemnestrae iussu per urbem
instituta'. This satisfies the requirements of the train of thought, yet Her-
mann has perhaps passed too lightly over the difficulty which arises out of the
meaning he assumes for ὅρος. Ahrens, 534, remarks that öpos = decretum
‘might be to some extent justified’ by Hesychius' ὅρος" νόμος, θεσμός, and by
the use of ópíL (Ahrens himself, however, decides in favour of Blomfield's
ἔρος). Now as regards the gloss in Hesychius, we should do well not to rely on
it, since we have nothing to tell us in what sphere the word may have
possessed that meaning. It is different when we turn to ὁρίζειν. A pertinent
use of this verb does actually occur in the Ores/eia and in one of the earliest
extant plays of Sophocles. ópífew^ means ‘to set a norm, to make a rule, to
establish a line of conduct as binding', etc. Thus Cho. 927: to Clytemnestra's
complaint that she can find no mercy Orestes replies πατρὸς yàp αἶσα τόνδε
σῴζοι ὁρίζει μόρον (Elmsley's restoration is convincing, cf. Conington and
Blass, ad loc., and Headlam's footnote to his translation; H. L. Ahrens,
Kl. Schr. 56, thought ὁρίζει ‘paullo languidius',? which seems to show that he
was not aware of the characteristic meaning in this passage): ‘my father’s
fate determines this doom for you’. In S. Ant. 452 the laying down and
establishing of laws is called ὁρίζειν νόμους. We can, then, regard it not indeed
as beyond doubt, but as highly probable that ὅρος is the genuine reading in
the present passage and that ὁ θῆλυς ὅρος means the same as τὸ ὑπὸ γυναικὸς
ὡρισμένον, ‘woman’s order (decree), the rule for conduct which is prescribed
by a woman'. What the queen does, and sets as a model for the others,
becomes a ópos for the citizens because they 'regulate themselves by it'.
Inasmuch as Clytemnestra has believed the beacon-message, sacrificed thank-
offerings, and announced the fall of Troy, she has set jn motion the θοὰ βάξις
(476), which, or its source, is called in 487 γυναικογήρυτον κλέος. This establish-
ing of a rule of conduct on her part carries conviction, is πιθανόν, for it leads
the people in Argos to believe what they would be only too glad to believe.
Consequently the citizens join in the sacrifices and other manifestations of
thankfulness and joy, not from compulsion, but from conviction and of their
own accord, as follows also from 594 ff. The example set by the queen has its
effect on the others, it 'spreads out’ ; this is pointedly denoted by ἐπινέμεται.
486. raxümopos: predicative, like ταχύμορον in the next clause; the anti-
thesis between these two words is full of bitterness. There is a further sharp
contrast between ἐπινέμεται and ὄλλυται.
487. κλέος : rumor (Blomfield), so Passow, who well compares the usage in
Homer. Cf. especially 4 137 μὴ πρόσθεν κλέος εὐρὺ φόνου κατὰ ἄστυ γένηται
ἀνδρῶν μνηστήρων. On this Homeric κλέος = φήμη (Eustath. on % 137) cf. now
M. Greindl, ‘ KAéos κῦδος εὖχος etc.’, Diss. München 1938, 16 ff.; he recalls,
1 A similar suggestion had already been made by F. W. Newman, Comments on the Text
of Aeschylus (1884), 71: ‘If ὅρος could have the sense of ὁρμή, that would not be amiss for
these two passages [Ag. 485 and 1154]’.
2 Cf. Wilamowitz on E. Jon 1222: ‘the tragedians use ὁρίζειν for “ordain, decree" ; in the
language of everyday life it is only ‘‘delimit” ’.
3 But above all he objected to the crasis, as others have done, and recently Denniston
on E. El. 413 and E. Harrison, Cambridge University Reporter, 11 Nov. 1941.

244
 COMMENTARY lines 486 f.
with Schroeder, the fact that Pind. P. 4. 125 is taking up a Homeric use in
ἤλυθον κείνου γε κατὰ κλέος.
486 f. ταχύμορον γυναικογήρυτον ὄλλυται κλέος : an exactly parallel idea in
Cho. 845 f. (only there it is a case of over-hasty fear as here it is of over-hasty
hope) ἢ πρὸς γυναικῶν δειματούμενοι λόγοι πεδάρσιοι θρώισκουσι, θνήισκοντεςϊἷ
μάτην.
The language of the old citizens is here very acrimonious: the fact that they
never say ‘Clytemnestra’ or ‘the queen’ outright, but speak, in thinly-veiled
language like a malcontent though powerless opposition, about woman’s
government, woman’s orders which are willingly accepted, and woman’s
speech, makes their words none the more innocent. Everything is tuned to
one and the same note; the repetition at such close intervals of γυναικός,
θῆλυς, γυναικογήρυτον, and the pointed antitheses all serve to express passion-
ate disapproval. Comments of this kind are obviously typical of whole
sections of the population of Argos: because they are typical, Clytemnestra
knows about them and replies to them in 590 ff.

Most modern editors do not even pause to find out whether there is any
foundation for their belief that the epode 475 ff. should be divided among
different members of the Chorus; they just take it for granted.? The common
practice of splitting the epode is due to the influence of Otfried Müller,
Göttinger gel. Anz. 1834, No. 198 f., p. 1980 f. (= KI. Schr. i. 280 f.), and of Her-
mann. Hermann (Opusc. vii. 45 ff., Aesch. ii, 2nd ed., 407), while rejecting
Müller's treatment of the epode as a whole,’ agreed with him‘ as to the neces-
sity of taking it as 'commatic', i.e. performed by subsections of the Chorus
singing in succession.5 Since it was well known that Hermann was not given
to easy acceptance of any view of Müller's, their agreement on this one point
had a fatal influence on their successors, including Ahrens (p. 532). And yet
! βθνήισκοντες should be protected by the parallel passage, above all by ὄλλυται, against
alteration to θνήισκοντος, yet this is put into the text even in quite recent editions (for the
‘redundant’ μάτην cf. Housman on Manilius 3. 108).
? The more recent among them might at least have heeded the solemn warning given
by Wilamowitz as long ago as 1895, Comment. metr. ii. 14 (= Verskunst 190).
3 Hermann's criticism is particularly successful where he shows that Müller was wrong
in denying that 475-88 is an ἐπωιδός. To Hermann’s observation concerning the identity
of the metre in the epode with that of the whole stasimon we may add what Wilamowitz,
Interpr. 168, says about the epode having the character of a πνῖγος. Müller has found
belated followers in R. Arnoldt, Der Chor im Ag. des Aesch. 41 ff., and Wecklein (1888), and
recently G. Thomson, The Oresteia, i. 22. H. Weil in his first edition (1858) called the piece
an epode and left it undivided, but he unfortunately recanted in his Teubner text.
4 Neither Müller nor Hermann himself seems to have remembered that long before
Müller's article Hermann (Elem. doctr. metr. 728) had anticipated Müller's view when he
said it was obvious ‘epodum non a toto choro cantatar esse . . . in epodo nihil nisi dubitatio
aliqua inicitur de fide nuntii . . . quae res non modo mirum quantum diversa est ab iis quae
ante dicta erant, sed omnino ita comparata ut ineptum videatur huiusmodi suspicionem,
quam ab uno alterove in medium afferri decebat, magnifice coniunctis universi chori
vocibus decantari.’
5 Müller assigned 475-8 to Α, 479-84 to B, 485-8 to C; Hermann 475-8 to À, 479-82 to B,
483-4 to C, 485-8 to D. Most editors followed Hermann. The arrangement in the editions
of A. Y. Campbell and Murray is the same as in Weil's Teubner text. Since any kind of
division is here arbitrary, we need not speculate whether ζυγά or στοῖχοι would be preferable
or ἡμιχόρια (Wecklein). A strong protest against Müller's ‘commatic’ arrangement of the
epode was soon raised by Klausen in the preface to his edition of the Choephoroe (1835) xi f.
245
 COMMENTARY
the assumption that the epode should be divided among three or four groups
of choreutae would have proved quite improbable had some attention been
paid to a basic formal rule. In no extant Greek tragedy is there an instance
of a stanza or an epode of a stasimon! divided among three or more sections
of the Chorus, although a division among ἡμιχόρια does occur: E. Suppl.
598 ff.2 This formal criterion settles the question: Ag. 475-88 must not be
divided. Nor is there, from the point of view of the contents, any reason for
splitting the epode into smaller units. But for a wretched modern delight in
deceptive liveliness, such a device would hardly have appealed so strongly to
editors. With ris οἶδεν... ris ὧδε maudvos . . ., without a change of the speaker,
cf., e.g., Prom. 160 ff. Nor do the asyndeta (cf. Otfried Müller, op. cit. 1980)
indicate any such change; see the pertinent warnings of Wilamowitz, Griech.
Verskunst, 190. It should be obvious that the passage Ag. 479 ff. marks an
increase of scepticism beyond 477 f., and 483 ff. a further strong increase. It
seems perfectly natural that, while the Elders are thus pondering over the
news, doubt and indeed disbelief more and more get the better of them.
But we cannot leave it there. For the feelings expressed in 475 ff., however
natural in themselves, seem to be quite inconsistent with the tenor of the
bulk of the stasimon, which is based on the certainty of the conquest of Troy.
This consideration above all induced Otfried Müller to sever the epode from
the main body of the canticum. He also believed he had detected a special
element in the action which led to the reaction of the Chorus voiced in 475 ff.
He found it (op. cit. 1981) in the ὀλολυγμός which was raised throughout the
city (594 ff.) : on hearing that jubilant shouting the Elders were provoked to
their critical utterances. This baseless hypothesis, though rejected by Her-
mann, was accepted by many commentators, down to Wecklein, and recently
by G. Méautis, Eschyle, 157. Ahrens (p. 533) saw in the words 476 f. πόλιν
διήκει θοὰ βάξις a strong confirmation of Müller’s view: how else, he asks,
could the Chorus, who at 261 ff. had themselves been quite unaware of the
victory, now know that such a rumour was spreading? This way of looking
at a play is a glaring example of misapplied rationalism. The Elders need
hardly strain their imagination to picture what now, with more and more
sacrifices offered and the excitement increasing in proportion, is going on all
over Argos and in what sort of talk the people is indulging.
1 D. L. Page, C.Q. xxxi, 1937, 94, neglected the difference between stasima and other
choric songs, else he would not have regarded the conventional division of Ag. 475 ff. as
‘probable’.
e For a detailed discussion of this stasimon cf. Joseph Lammers, Die Doppel- und
Halbchóre in der antiken Tragödie (Diss. Münster 1931), 95 ff. My statement about the
stasima would be too sweeping if we had to read E. Jon 695 ff. in the form in which it is
printed by Murray. But of his suggestion that the lines ‘confusas et interruptas praebent
singularum ancillarum sententias' no notice is taken in Wilamowitz's edition of the play,
and Α. S. Owen (note on 695-712) cautiously rejects it. Neither in matter nor in form can
E. Alc. 213 ff. be regarded as a stasimon. Besides, it seems very doubtful (cf. Lammers,
p. 87) whether it is at all necessary to divide that ode as has been customary since Matthiae
and Hermann. In any case, all that might be conceded would be division among ἡμιχόρια ;
division among smaller sections of the Chorus (Hermann and others) is quite arbitrary. As
for Rhesus 527 ff., it is certainly not a stasimon (Wilamowitz, Hermes, lxi, 1926, 288, rightly
speaks of 224 ff. and 342 ff. as 'the only two stasima', see also W. Kranz, Stasimon, 2641.) ;
so, as far as the present issue is concerned, it does not matter whether we accept Murray's
division among several sections of the Chorus or are content with ἡμιχόρια (cf. Lammers,
op. cit. 126 f.).
246
 THE EPODE 4758. AS A WHOLE
But the discrepancy between the bulk of the stasimon and its epode re-
mains. Hermann thought that the gulf might be bridged by the expedient
of a subtle psychological interpretation : ‘laeto nuntio accepto facile omnium
animi eriguntur ad gaudendum: sed mox, praesertim tali nuntio accepto qui
facile fallere potuerit, alius atque alius [this is based on the assumption of a
division of the Chorus, see above] dubitare incipiunt’. Wilamowitz, on the
. other hand, did not admit the existence of that gulf: in his Comment. metr. ii
(Göttingen, 1895) 10 (= Griech. Verskunst, 185 f.) he constructed a chain of
thoughts connecting the epode with the preceding section of the song. This
is how he paraphrased it :! ‘this expedition for the sake of a lewd woman is
sin, and if it now has led to a proud triumph, then the realm of the Atridae
has become as wanton and sinful as Troy, and the outcome will be in accord-
ance with it [437-74]. Then he goes on: ‘the song that was begun in thanks-
giving for the victory concludes with the doubting question: is it really
possible that the news of this undeserved victory should be true?’ and finally
pronounces his verdict : “The reversal in the mood of the Chorus has a suffi-
cient foundation in the progress of thought, but it also serves the dramaturgic
purpose’, etc. I confess that I cannot find any evidence for the progress of
thought which Wilamowitz assumes; in particular the ‘undeserved’ victory
of which he speaks strikes me as an interpolation and, worse, as utterly in-
compatible with Δία τοι ξένιον μέγαν αἰδοῦμαι τὸν τάδε πράξαντα and Διὸς
πλαγὰν ἔχουσιν εἰπεῖν.
Like Wilamowitz, W. Kranz (Stasimon, 159 1.) contrives, by assuming
gradual changes in mood and thought, to make the epode appear a psycho-
logically intelligible result of the whole stasimon. After summing up the
contents of the third stanza and its antistrophe (‘doubts arise whether the
expedition which demanded so many victims was really just, and anxiety
lest the conqueror should provoke the envy of the god’), Kranz describes the
reversal in the epode thus: ‘After so many heavy thoughts the joy of the
victory seems to have flown: we believe the Chorus when at the end, as on
former occasions (97 ff., 261 ff.), they again utter doubts whether that message
of the fall of Troy, the foundation of the whole song, is really true and not
perhaps the rash belief of a woman.’ Now not to be able any longer to enjoy
the good news is one thing, to disbelieve it is another. The thread which the
learned interpreter has spun is too delicate; it will break under the strain.
Kranz then enters on the road to compromise : ‘One may say that the dramatist
forces the song to this close for the sake of the tension with which we are to
look forward to the following scene, the arrival of the Herald ; none the less the
song remains a marvellous whole containing indeed contrasts, but in such a
form as to appear a unity.’ It is of course obvious that the sceptical attitude
of the Chorus (475 ff.) provides a very effective foil to the speech of the Herald
who finally confirms the fall of Troy (cf. Lesky, Hermes, lxvi, 1931, 200). But
does that mean that, dissatisfied with the flimsy psychological speculations
which endeavour to regard the scepticism at the end of the choric ode as
arising from ideas of the preceding parts, we have to assume that the epode
has been crudely patched on to the stasimon, regardless of its general out-
! In the introduction to his translation (Griech. Tragoedien, ii. 33). Afterwards, Interpr.
168, he only gave a brief hint: ‘So here the moral reflections finally lead the Chorus back
from the joy of victory to the doubt which in the epode is uttered undisguised.'

247
lines 475-88 COMMENTARY
look? In assuming this we should be looking at tlıe choreutae as though they
were on the same footing as the main dramatic characters. But they clearly
are not. Their participation in the action of the play is far less intense and
consistent. On the other hand, it is no mere exaggeration but substantially
wrong when Kranz, Stasimon, 169, asserts that ‘the Chorus of the Elders in
the Agamemnon have by now become merely an accompanying instrument’
incapable of acting themselves, even when at the end they are drawing swords
against Aegisthus (1651)'. True, they are not in a position at any moment to
give a decisive turn to the course of events. But it does not follow that they
are merely an accompanying instrument in the sense in which this could be
said of many a Chorus in later Tragedy. On thecontrary, they play an integral
part in the action.” Their activity, however, consists not in acting on their
own initiative but in reacting. In so doing they do not serve as a mere sound-
ing-board. They in their turn provoke important utterances and thus open
fresh aspects and more than once help to bring the dramatic tension to a
climax. It is the incredulity of the Elders that elicits from the queen her
speeches on the beacons and on the conditions in captured Troy. Similarly
the inquiry of the coryphaeusleads to the Herald’s report on the storm and the
fate of Menelaus. Still more important, because intimately connected with
the central issue of the tragedy, is what the Elders do to put Agamemnon on
his guard against the peril that threatens him. When they speak tothe Herald,
who is going back to the king, they have to confine themselves to a carefully
guarded hint (cf. on 615 f.). But as soon as Agamemnon appears himself,
they make their warning perfectly clear and dwell on it with great intensity
(788-98) ; the king's reply (830 ff.) does full justice to the seriousness of their
advice. We shall see later on that the apparently strange behaviour of the
Elders after Agamemnon’s murder is in fact not contradictory to their previous
attitude, and that there, too, they are far from forming an ‘accompanying
instrument’. But for the moment we have to return to the epode 475 ff. The
reason why at this particular juncture the Chorus give vent to their scepticism
cannot be found in any consistent course through which their thought may
be supposed to have passed during the former parts of the song. We have
seen that the text does not bear out any such construction. The moment
which the poet has chosen for the utterance of the Elders’ doubts was dictated
to him by considerations of dramatic structure, that is to say the need for an
effectual foil to the Herald’s speech. The possibility, however, of the old
Argives’ making such utterances at all is based on true psychology. To that
extent Hermann’s interpretation holds good: ‘laeto nuntio accepto facile
omnium animi eriguntur ad gaudendum: sed mox . . . dubitare incipiunt.’
The Elders, like anyone who, after a long period of desperate waiting and
disappointed hopes, is suddenly told that the end of his troubles has at last
come, wil naturally be the prey of contradictory emotions, of cheerful
confidence followed by distrust and gloom. If they had to function as ordin-
ary characters in the play, the sequence of their changing moods would
probably be brought into a consistent line of psychological progress. As a
1 ‘Nur noch begleitendes Instrument.’ By saying ‘noch’ he suggests that a later develop-
ment is anticipated.
2 Cf. Th. Plüss, Die Tragödie Agamemnon und das Tragische (Basel, 1896), 23; G. Méautis,
Eschyle, 130.

248
 COMMENTARY lines 489 f.
Chorus, however, they are possessed of less spontaneity ; their words serve in
the main as a reaction to the acts and words of the actors or as a means of
leading them on. There is, as it were, a certain looseness in the psychological
texture of the Chorus. Of this looseness Aeschylus takes full advantage. He
does not indeed ascribe to his Elders anything that would be altogether
inconsistent with their general condition and their outlook. But he does feel
at liberty, whenever it suits the demands of the dramatic structure, to select
a particular aspect of that general condition and outlook without providing it
with a special foundation. We shall notice the same treatment of the Chorus
in other parts of the play.

489. For the wrong indication of speakers in the MSS see on sor.
τάχ᾽ εἰσόμεσθα is similarly used in transition from a choral ode to a dia-
logue-scene in S. Ant. 631.
489 f. If we keep to the reading of the MSS, the juxtaposition of the three
nouns in the genitive, λαμπάδων, φρυκτωριῶν, πυρός, presents a difficulty, not
merely because of the accumulation, but because it is not clear how they are to
be arranged, i.e. whether they are all parallel, or, in case that does not work,
which of them depends on which. In the abstract there are two possibilities :
(1) To co-ordinate the three genitives, making all three depend on παραλλαγάς
(so e.g. Schütz and Nägelsbach). This is improbable because such an arrange-
‘ment of a threefold expression (a, b--7e, καὶ -4-c) appears to be unparalleled
at any rate in Aeschylus;! among the numerous examples cited by Dindorf,
Lex. Aesch. 349, section 2 (‘ re respondens sequenti xa "), there is not a single
one that is really comparable. (Ag. 327 has only a superficial similarity, for in
that passage a new governing noun (παῖδες) is introduced for the third genitive,
so that it is really not a case of three co-ordinated parts of an expression, but
only of two, the first of which is subdivided.) (2) A different arrangement is
proposed by Conington, who translates 'Soon shall we know of these light-
bearing lamps The passage of the beacons and the flame', and explains:
‘Kap. φαεσφόρων is the genitive denoting the general matter about which
something is to be known ; παραλλαγὰς φρυκτ. x. r. the special object, the thing
to be known'. This is too laboured to be convincing. The right solution was
put forward by Wilamowitz, who (in his edition of 1885) reads φρυκτωρίας:
this emendation was adopted by Wecklein in his annotated edition, and
Headlam did in fact recognize the necessity of the emendation in his prose
translation: *we shall soon know about these beaconings of light-bearing
torches and these passings-on of fire.' However, the text as restored by
Wilamowitz may be understood in a way different from Headlam's: I assume
that λαμπάδων φαεσφόρων is governed by what is to be regarded as a single
expression φρυκτωρίας τε καὶ πυρὸς παραλλαγάς. This provides a stronger
structure for the whole and a more natural position for the re (from Head-
lam's translation we should suppose that the re came after λαμπάδων).
πυρὸς παραλλαγάς, taken closely together and practically equivalent to a
single compound, balances ¢puxrwpias. φρυκτωρίας is here genuinely a noun
1 In Ag. 1014 f. πολλά τοι δόσις ἐκ Διὸς ἀμφιλαφής τε καὶ ἐξ ἀλόκων ἐπετειᾶν it would be
quite wrong to suppose that πολλά, ἀμφιλαφής, and ἐξ ἀλόκων ἐπετειᾶν are all on the same
footing. It is rather the case that ἀμφιλ, and ἐξ dA. ἐπετ. are closely connected with each
other, while πολλά is added to the whole expression (cf. ad loc.).
249
lines 489 f. COMMENTARY
of action, as it is elsewhere in the fifth century, e.g. Thuc. 3. 22. 8 (‘Feuer-
signalisierung’ Classen-Steup). For the accumulation of almost synonymous
expressions here see on 491, for the alliteration see on 268.
490. παραλλαγάς : the word is found only here in poetry. In this passage it
serves to compress into one noun the process described in 313 by ἄλλος παρ᾽
ἄλλου διαδοχαῖς πληρούμενοι.
491. eire . . . εἴτε must not here be taken, as they have often been,
as uirum . . . an, but are conditional, sive . . . sive. The addition of οὖν
puts this beyond doubt, cf. the instances collected by Denniston, Particles,
418 f., to which more may be added from Dindorf, Lex. Aesch. 108, and
Ellendt-Genthe, Lex. Soph. 219. This observation disposes of Bothe’s attempt
to cut out I. 490, which is improbable in itself. Bothe objected to what he
calls a rhetorical and unpoetical accumulation of expression in 489 f., and
similarly Emperius, Opusc. 122, who regarded the 'inutilem verborum cumula-
tionem' as unworthy of Aeschylus, and tried to get rid of it by condensing the
two lines into τάχ᾽ εἰσόμεσθα λαμπάδων παραλλαγάς. There was some point in
what these critics felt. The accumulation of practically synonymous words
produces a peculiar effect and is intended to do so. Schneidewin's explanation
is substantially correct though he puts it in a somewhat crude form: 'the
accumulation [in λαμπάδων... παραλλαγάς] is due to the fact that the leader
of the Chorus models his words on the ''pathetic" style of Clytemnestra in
describing the signals by means of a variety of expressions, and his purpose
is to parody it.' It is quite obvious that the coryphaeus does refer directly to
the queen's statement and repeats her expressions: with λαμπάδων cf. 287
λαμπάδος, 296 λαμπάς, 312 λαμπαδηφόρων, with φαεσῴ. cf. 300, 302, 311 φάος,
with φρυκτωρίας cf. 282 φρυκτὸς δὲ φρυκτόν, 292 φρυκτοῦ, and for παραλλαγάς
see the reference to 313 explained on 49o. The repetition of the expressions so
often used by Clytemnestra herself, and the bombastic accumulation of
them within the compass of a short sentence, can only have a contemptuous
effect here. The tone, even more aggressive if possible than in the preceding
epode, hardly leaves room for doubt which of the alternatives expressed in
491 f. the speaker is inclined to believe.
491 f. ὀνειράτων δίκην τερπνὸν τόδ᾽ ἐλθὸν φῶς. Hermann says rightly: 'co-
haerent ὀνειράτων δίκην ἐλθὸν τερπνόν, somniorum ritu veniens ut oblectaret
(where τερπνόν is attached predicatively to ἐλθόν, as, e.g., in 530 εὐδαίμων ἀνήρ
is attached predicatively to ἥκει). But ἐφήλωσεν φρένας also goes with
ὀνειράτων δίκην, since both phrases contain the idea of ψεῦδος.
492. φῶς ἐφήλωσεν φρένας : continuing the alliteration of 489 f.
493. κήρυκ᾽ ἀπ᾽ ἀκτῆς κτλ. gives the reason for the statement τάχ᾽ εἰσόμεσθα
κτλ. For this kind of explanatory or amplifying asyndeton cf. Kühner-Gerth,
ii. 344; Denniston, Particles, p. xliii f.
493 f. κατάσκιον κλάδοις ἐλαίας: in itself κατάσκιος means only ‘shaded,
covered with’, and thus, e.g., in Suppl. 345, 354 there is no idea of 'crowning
with garlands' ; the sense is rightly expressed in the paraphrase of the Schol.
on 345 τοῖς θαλλοῖς ἱκετηρίων κλάδων πεπληρωμένους (cf. Wilamowitz, Interpr. 7
‘the branches of the suppliants waving from the altars’). But here we can
hardly help supposing that the Herald is crowned with a garland, for if he
were just holding a branch in his hand, he could certainly not be described
as κατάσκιος. Besides, in passages of other plays where the arrival of someone
250
 COMMENTARY line 495
coming from a distance is announced in a similar way (S. Trach. 178 f., Oed. R.
73-83) the wearing of garlands is clearly indicated. It is harder to see what
Aeschylus intended to be the significance of the garland in this case. Weck-
lein : ‘because he brings the message of victory. Cf. Soph. Oed. R. 82’. But in
that passage there is the important circumstance that it is from the Delphic
oracle that Creon is coming back, garlanded with the god’s laurel leaves:
moreover it is expressly said that the garland betokens news of success (on
this point cf. Jebb). In Trach. 178 the sole reason for the garland is the
victory and the triumphant homecoming which is announced by the mes-
senger.! Cf. further (van Heusde) Chairemon Trag. fr. 6 (p. 783 N.) στεφάνους
τεμόντες ἀγγέλους εὐφημίας. But if that were the significance of the garlanding
here, then those present should, as they do in Sophocles, draw conclusions of
victory or at least indicate victory as probable; but this is contradicted not
merely by 491 f., but also by 498 f. Verrall’s explanation is: ‘what is inferred
is that he comes ἀπ᾽ ἀκτῆς [for this inference the garland is unnecessary: it
can be made from the direction from which he is coming]. The herald is
wreathed, as the ship itself was wreathed, in sign of gratitude to the gods for
the safe conclusion of a voyage’ ; similarly Plüss. I do not know of any really
cogent explanation. In the reference-books s.v. corona etc. and in the article
‘Die Bedeutung des Kranzes im klassischen Altertum’ by L. Deubner,
Archiv für Religionswiss. xxx, 1933, 70 ff., I have not found anything to
elucidate this passage.
494 f. κάσις πηλοῦ Edvoupos, Supia κόνις : in Aeschylus’ usual manner first
comes the ypigos, then its solution. «dos: ‘una enim eademque ex terra,
madore lutum, calore pulvis gignitur' (Schütz). For «dows πηλοῦ cf. in general
Sept. 494 λιγνύν. . . αἰόλην πυρὸς xdow, Hipponax fr. 38 D. συκέην μέλαιναν
ἀμπέλου κασιγνήτην, Hipponax Pap. Oxy. 2175, fr. 2. 10 ἀσβόλου κασιγ[νήτη],
and in particular Hipponax fr. 70 A Bergk (= 37 Knox) βολβίτου κασιγνήτην.
One would like to see in this mode of expression an instance of the manner
in which the dialogue style of Tragedy has adopted certain elements of the
Ionic iambus (cf. Wilamowitz, Sappho und Simonides, 275, and Interpr. 247) ;
in this case, however, it is also possible that at the back of such turns of
expression there are ‘kennings’ of hieratic language. Phrases that one com-
monly finds elsewhere, where direct descent is signified, such as répewav
ματέρ᾽ οἰνάνθας ὀπώραν (for other instances see Dornseiff, Pindars Stil, 51,
and Schadewaldt, Der Aufbau des Pindarischen Epinikion, τό n. 4), are of a
different type.
While the relationship between πηλός and κόνις is stressed here, the two are
contrasted with each other in the farmers’ saw (carm. pop. 16 Diehl) σῖτον ἐν
πηλῶι dureve: τὴν δὲ κριθὴν ἐν κόνει, as in the 'rusticum vetus canticum’
(Macrob. 5. 20. 18) hiberno pulvere, verno luto eqs. Cf. Seneca, EPist. 57. 2 duo
incommoda inter se contraria simul pertulimus: eadem via eodem die et luto et
pulvere laboravimus.
495. ξύνουρος : ‘vicinus, quia ibi est pulvis, ubi desinit luti humor’ (Klausen).
διψία kóvis: the same phrase in S. Ant. 246.
1 Only remotely comparable is the fact that victorious generals are crowned with a
στέφανος ἐλαίας, as, e.g., Eurybiades and Themistocles by the Spartans after the battle at
Salamis (Hdt. 8. 124. 2). For the olive-wreath as sign of victory cf. Wilamowitz, Reden und
Vortráge, i, 4th ed., 376 n. 2.
251
line 495 COMMENTARY
What is meant here is of course the dust raised by the approach of the
Herald. Blomfield’s different interpretation (‘pulvis quo ex longo itinere
illitus est’) was recently repeated by Gilbert Murray, Aeschylus, 213; it had
been well rejected by Lewis Campbell, C.R. iv, 189o, 302.
496. Headlam's explanation of the MS reading σοι is linguistically correct,
but he does not touch upon the difficulty of the subject-matter. ' σοι cannot
be right ; for it is as certain as anything about Greek plays can be certain that
Clytaemestra is not now on the stage’ (Housman, J. Phil. xvi, 1888, 265).
Statements like Nägelsbach’s ‘per oof appellatur Clyt. absens'—it should be
noticed that in 83 ff. the absent queen is directly addressed by name—or
Plüss's ' cov is some person unspecified’ are not helpful. οὗτος, on which both
Wilamowitz (1885) and Housman have hit, is excellent. Wilamowitz's οὐ
δαίων is better than Housman's ἀνδαίων, as making it clearer that οὐ δαίων
κτλ. provides the explanation to οὔτ᾽ ἄναυδος. For this is how we must under-
stand the relation between the two clauses. G. Thomson, who adopts
Wilamowitz’s text, is wrong in comparing this with cases of οὔτε... οὐ in real
correspondence, like those cited by Denniston, Particles, 510, e.g. Prom.
450 f. κοῦτε πλινθυφεῖς δόμους προσείλους wav, οὐ Evdoupyiav. In Ag. 496 f. it is
plain that ἄναυδος and δαίων φλόγα refer to one and the same type of messenger.
In 498 we get with ἀλλ᾽ ἐκβάξει λέγων ἃ sharp contrast to ἄναυδος σημανεῖ. So
we have here the same type of additional sentence as in the examples cited
by Kühner-Gerth, ii. 292 b, and Denniston, 511, for oùre . . . δέ (ἦμος δ᾽ οὔτ᾽
dp πω ἠώς, ἔτι δ᾽ ἀμφιλύκη νύξ etc.), except that ἀλλά stresses the adversative
relation more strongly than δέ.
499. ἀποστέργω : this is the first instance we have of the word (for the
date of the 'Terpander' lines, fr. 4 D., in which it occurs see Wilamowitz,
Timotheos, 64 n. 1). In this euphemistic sentence Aeschylus uses a verb of
euphemistic form (as, e.g., one says 'discontinue' in breaking off relations or
stopping payment of a subscription). There is a corresponding euphemism
(following closely on 636 εὔφημον ἦμαρ κτλ.) in 638 ἀπευκτὰ πήματα.
500. ‘For I pray that an addition may happily be made to what has already
happily appeared (or, been realised)' is Paley's rendering, which is probably
correct, cf. on 255; the assumption that εὖ, ‘good’ (cf. on 121), might be the
subject here and προσθήκη a predicative noun seems too risky. This con-
cluding prayer is just as different from Clytemnestra's (349 f., see the note) as
is the whole attitude of the Chorus from hers. Clytemnestra pretends that
she is now entirely content: there now remains, so she says, nothing more to
wish for except τῶν ὄντων νῦν ἀγαθῶν ὄνησις. The Elders, on the other hand,
still distrustful and full of sinister forebodings, pray for an addition to what
has already been happily achieved. In the central scene of the play the
sequence of two concluding prayers and the discrepancy between them will
be found again, and the effect of the parallelism will be deepened : there (854)
Agamemnon wishes νίκη δ᾽ ἐπείπερ ἕσπετ᾽, ἐμπέδως μένοι, whereas Clytem-
nestra prays (973 1.) Ζεῦ... τὰς ἐμὰς εὐχὰς τέλει" μέλοι δέ τοί σοι τῶνπερ ἂν
μέλληις τελεῖν.
501. The marking of a change of speaker (XO.) here in the MSS is clearly
connected with the ΚΑ. at 489. Wilamowitz rightly supposes that in the
archetype there were merely paragraphoi. For this one need not even appeal
to the general practice in ancient texts of dramatic works (for the parabases
252
 COMMENTARY line 501

of Comedy Hephaestion περὶ σημείων το, p. 75 Consbruch, gives a very

detailed account): the Mediceus too has often enough preserved the para-
graphos. The change in 489 (transition from choral ode to trimeters spoken
by the coryphaeus) has an exact parallel in 258. There the first hand of M
puts only a paragraphos at the beginning of the line.! Similarly in M the
paragraphos? is found, e.g., at Cho. 730 (the passage resembles Ag. 489: there
the coryphaeus, after preceding anapaests from the Chorus, announces the
approach of a fresh character) ; at Septem 369 (trimeters by the coryphaeus
after the choral ode) there is a paragraphos and in front of it ἡμιχ., at Pers.
155 (transition from the parodos to dialogue lines by the coryphaeus) the
paragraphos has in front of it ἔτιοχ (ἔτι ὁ χορός). So in all probability it is only
the period of the later Middle Ages that is responsible for the nonsensical?
introduction of Clytemnestra at 489. At 5or as well as at 489 we have to infer
that the παράδοσις of the earlier period gave only the paragraphos. Un-
fettered by the MSS Scaliger let the speech of the coryphaeus continue to the
end of 502, and he was followed by Hermann and others, e.g. by Wilamowitz
in his translation, Kranz, Stastmon, 276, and A. Y. Campbell, and this is
obviously right. For ‘it is quite without parallel that after the entry of a
fresh character has been announced by the coryphaeus, another member of
the Chorus should speak’ (Hense in the Appendix to Schneidewin’s edition).
The lyric choral passages, Sept. 100, 104, E. Her. 815 ff., cited by Wilamowitz,
Interpr. 169, in support of the view that a second member of the Chorus is
speaking in Ag. sor f., are entirely inappropriate for comparison with the
dialogue lines here. On the other hand, there is a real analogy between our
present passage and the two concluding lines of Teucer’s speech in 5. 47.
1038 f. ὅτωι δὲ μὴ τάδ᾽ ἐστὶν ἐν γνώμηι φίλα, κεῖνός τ᾽ ἐκεῖνα στεργέτω κἀγὼ
τάδε, cf. also the end of Creon’s speech in Ant. 209 f., where a threat against
unpatriotic citizens is followed by the words: ἀλλ᾽ ὅστις εὔνους τῆιδε τῆι πόλει,
θανὼν καὶ ζῶν ὁμοίως ἐξ ἐμοῦ τιμήσεται, and the end of the ῥῆσις in E. Anar.
419 f. ὅστις δ᾽ αὔτ᾽ ἄπειρος ὧν ψέγει, ἧσσον μὲν ἀλγεῖ κτλ. There it is true the δέ
(or ἀλλά) makes the connexion more smooth. But the asyndeton in ὅστις
τάδ᾽ ἄλλως κτλ. is unobjectionable: ‘emphatic expression of contrast by way
of supplement’ (Plüss) ; for Aeschylus’ liking for asyndeton in general cf.
Kranz, Stasimon, ıso. There is an asyndetic ending to a poem with a similar
introduction to this in Bacchyl. ıı(ro). 123 ff. δωκαίας ὅστις ἔχει φρένας, εὑρήσει
σὺν ἅπαντι χρόνωι μυρίας ἀλκὰς Ἀχαιῶν. The wrong attribution to a new
speaker of the closing lines of a speech is not uncommon: we shall find it
again at 613 f.; and in the Mediceus a paragraphos has been wrongly put at
Eum. 305 and 913 before the last two lines of a speech and the last three lines
respectively. A like mishap has caused a disturbance in the three closing
lines in S. El. 428-30: in this case, as I’ shows, the mistake had already been
observed before Triclinius, and nowadays nobody doubts that the lines, the
meaning of which is difficult (cf. T. v. Wilamowitz, Dram. Technik des Soph.
1 Similarly, e.g., Zum. 117 ff. in M there is only the paragraphos, in the later MSS ‘Aréa.
and Xop.
2 The paragraphos is also used where lyrics succeed trimeters spoken by the same person,
e.g. Prom. 114.
3 Of course every piece of nonsense that occurs in a MS finds its defenders at some time
or other: in this case they are Karsten, Dindorf, and A. O. Prickard, C.R. xiv, 1900, 434 ff.
Lawson ascribes 489 ff. to the Watchman!
253
line 501 COMMENTARY
177 n. 1), form the conclusion of Chrysothemis’ speech. The converse mis-
take, so to speak, is found in Ant. 278 f., where the MSS attribute two further
lines with the preceding passäge to the Sentry (instead of to the Chorus); cf.
also E. Ion 832 ff.
502: a modified form of curse instead of αὐτῶι ἐς κεφαλὴν τράποιτο, which is
characterized as ἀγροικότερον in Plato, Euthyd. 283 e. (There is a correspond-
ing modification in E. Hec. 1276, where Hecuba replies to a prophecy of evil
directed against her daughter with the words: ἀπέπτυσ᾽" αὐτῶι ταῦτα σοὶ
δίδωμ᾽ ἔχειν). Perhaps similar imprecations were sometimes directed against
ill-wishers at the conclusion of genuine prayers, cf., e.g., Ar. Peace 441 (after
a prayer for peace) ὅστις δὲ πόλεμον μᾶλλον εἶναι βούλεται κτλ. Cf. also the
sentence with which the mother of Demaratos ends her long speech in Hdt.
6. 69. 5 ἐκ δὲ ὀνοφορβῶν αὐτῶι τε Aeuruxidm καὶ τοῖσι ταῦτα λέγουσι τίκτοιεν ai
γυναῖκες παῖδας.

The often-quoted scholion on 503 (ZxoÀ. wad.) runs: τινὲς μέμφονται τῶι
ποιητῆι ὅτι αὐθημερὸν ἐκ Τροίας ποιεῖ τοὺς "EAAnvas ἥκοντας. The different
stages of modern reaction to this piece of Hellenistic rationalism are highly
interesting. Schütz not only countenanced but elaborated the verdict of
those unknown ancient critics: he added a special digression (‘Excursus I’)
to prove that the blunder of contracting a long period into the space of a
single day? was the only flaw in the οἰκονομία of the play. Now Schütz cer-
tainly deserved the praise of Hermann, Opusc. vi. 1. 97, who says that among
the Aeschylean scholars of those days he was the only one ‘dem poetischer
Sinn nicht abzusprechen ist’. Still, in a case like this, it is not surprising that
he should have proved faithful to eighteenth-century rationalism. A few
decades later (1816) W. v. Humboldt wrote (Aesch. Ag. xii f. = Gesammelte
Schriften, viii. 126 f.) in regard to the criticism mentioned by the scholiast :
‘The fact that the sighting of the beacon-signal and the return of Agamemnon
are separated only by a few hundred lines, spoken and sung without inter-
ruption, will not surprise anyone who is familiar with the works of the ancient
writers. It would in fact be a mistake to assume with confidence that
Aeschylus compressed the homeward journey into a single night, or intended
1 This obviously goes back to a criticism made in the Hellenistic period. Comparing the
many similar remarks in the scholia on Euripides (cf. Wilamowitz, Einleitung in die griech.
Tragoedie, 158 ff.), we may consider it likely that in the original commentary the attack was
followed by a defence of the poet. For the phrasing cf., e.g., schol. E. Phoen. 4 ἔτι δέ τινες
ἐγκαλοῦσι τῶι Ἐὐριπίδηι ὡς οὐκ ἀκολούθως yeveadoyjaavre . . . πρὸς οὖς ῥητέον ὅτι, εἰ μὲν ἄνωθεν
ἤρξατο, μακρὸς ἂν ἦν ὁ λόγος «rA., schol. Or. 396 ἐγκαλοῦσί rives ... ἀγνοοῦσι δὲ örı.... For
the arrangement by which after the unfavourable criticism a justification is given cf., 6.5.»
schol. Phoen. 31, 47, 202, 267, 301, 409 (p. 297. 17 Schwartz), 507, schol. Andr. το, 32 (p. 254.
11 Schw.), 224, 616, 734, schol. Tro. 31, etc.
Dio’s comparison of the three Φιλοκτῆται enables us to form a general idea of that ration-
alistic criticism of Aeschylus of which a specimen is preserved in the scholion. He says (Dio
Chrys. 52. 5): Aeschylus did not trouble to have the appearance of Odysseus changed by
Athena, ὥστε τυχὸν dv τις ἐγκαλέσαι τῶν οὐ φιλούντων τὸν ἄνδρα, ὅτι οὐδὲν αὐτῶι ἐμέλησεν
ὅπως πιθανὸς ἔσται ὁ ᾽Οδυσσεὺς οὐ γιγνωσκόμενος ὑπὸ τοῦ Φιλοκτήτου, and later on (7): καὶ
γὰρ εἰ μὲν ἐδύναντο πάσας διαφεύγειν τὰς ἀλογίας ἐν ταῖς τραγωιδίαις, ἴσως ἂν εἶχε λόγον μηδὲ
τοῦτο παραπέμψαι" νῦν δὲ πολλάκις ἐν μιᾶι ἡμέραι παραγιγνομένους ποιοῦσι τοὺς κήρυκας πλειόνων
ἡμερῶν ὁδόν.
2 ‘Quae in rerum natura aliquot mensium spatio vix absolvi poterant, ea Aeschylus in
scena unius diei spatio fieri iussit.’
254
 THE HERALD’S ARRIVAL AND THE NATURAL LAPSE OF TIME
to let it take its natural time. The former idea is contradicted in no ambiguous
terms when he describes the scattering of the fleet by astorm.... The latter
would utterly spoil the splendid swift movement of the play, in which after
the uncertainty and tension caused by the beacon-flame a prompt solution
is demanded. The question itself could not arise with a poet of Aeschylus’
time, and there was nothing incompatible with his conception of a tragedy in
allowing Agamemnon and his army to appear immediately, without giving
an account of the length or shortness of their journey. Works of ancient art
very often disdain to be so scrupulous about connecting by visible means and
in accordance with natural happenings (“auch gewissermassen äusserlich,
und wie es in der Natur zu seyn pflegt") the several parts of that which they
represent.’ Humboldt then goes on to illustrate his point by examples from
Greek decorative art.
This is the voice of a new epoch. In Humboldt we find the historical out-
look of Winckelmann and his followers, who showed how to view works of
art in relation to the conditions of their own time, acting upon the genius
of an individual, the friend of Schiller and Goethe, an unsurpassed judge
and interpreter of their most difficult works, a man with an extraordinary
capacity for the sympathetic understanding and lucid explanation of great
poetry. Humboldt’s view was more or less exactly adopted by his country-
men.! It was no improvement when some commentators made the addi-
tional point that ‘the stasimon, which, as we should say, separates two acts,
makes it easier for the audience to imagine that a considerable time has
passed’ (Enger, in his edition of 1855, note on 503 = 486 Enger). This opinion
was shared, e.g., by Wilamowitz, Griech. Tragoedien, ii (Introd. to Ag.), 33,
cf. also ibid., xiv (‘D. griech. Tragoedie und ihre drei Dichter’) 56, with
special reference to the present instance in the Ag. (his similar remarks
Hermes, 1x, 1925, 288 f., where he is speaking mainly of Euripides, are more
appropriate, though not entirely so), and by M. Croiset, Eschyle, 184. There
are certainly choral odes in Greek Tragedy which may be compared to an
entr’acte, but the stasimon Ag. 355-488 is different. W. Gilbert was right
when, in his re-edition of Enger’s commentary (1874), he explicitly rejected
his predecessor’s view ; ‘for’, he argued, ‘at the end of this stasimon (475 ff.)
the Chorus say that the news of the beacon-post is now rapidly spreading
through the city, which would naturally happen shortly after its arrival’.’
The stasimon, then, not only does not suggest, but makes it difficult to
imagine that while it is being sung a period of many days or weeks elapses.
It will therefore be better to return to Humboldt’s more general explanation.
We may add a really helpful remark of Wilamowitz, Hermes, lx, 1925, 288:
‘Attic tragedy is not bent upon faithfully reproducing the reality of everyday
life in its irrelevant features.’ Whenever circumstances are relevant to the
dramatic action in its proper sense, they are rendered unambiguous and can
be easily discerned without the effort of peeping into what are supposed to be
hidden allusions or making complicated calculations. Where, on the other
hand, facts belong only to the general background or where they are prim-
arily connected with things that play no direct part in the evolution of the
1 Cf., e.g., the introductions of Klausen (p. xvii f.) and of Wecklein (annotated edition,
1888, p. 23).
2 The same point was made by Lesky, Hermes, lxvi, 1931, 199.
255
line 503 COMMENTARY
plot or the themes of the dialogue, they remain indifferent: we are not to
think of them at all. This applies for instance to the locality of the plays and
also to the time required for the action and its various parts. In the Agamem-
non the speedy transmission of the beacon-post is very much stressed, for
that speed is essential to the wonder roused by Clytemnestra’s speech. As to
the king’s and the Herald’s return, on the other hand, it is sufficient to learn
that they have arrived after all. No feeling member of the audience will,
while exciting speeches are being delivered on the stage, sit back and ponder
over the length of time it must have taken the fleet to cross the Aegean Sea.
The rule that the action of a tragedy should take place within the space of
one day was of course unknown to the contemporaries of Aeschylus.
Recently Gilbert Murray (see his translation of the Oresteia, his edition of
Aeschylus, and his book Aeschylus, Ὁ. 213) and A. Y. Campbell (see his
‘annotatio critica’ p. 88, and his translation) have endeavoured to put the
clock back by reviving the hypothesis of Blomfield (3rd ed., p. xiii f.), who,
starting from the notion of ‘unitas temporis',* excogitated the λύσις that the
Chorus, by way of μετάστασις,2 left the orchestra after 488. Since it is obvious
from the very phrases that the lines 489-92 form a direct continuation of
475 ff., the assumption of a few days 'interval (Murray and A. Y. Campbell)
before 489 seems impossible.

503. ἰὼ πατρῶιον οὖδας : Aeschylus uses ἰώ fairly often, but not in trimeters
except in this speech (again in 518) and in a cry of Cassandra's (1305, cf.
1315), and again fr. 143 N. (discussed below). It must therefore be regarded
as expressing strong excitement. The Herald's first word shows him as a man
quite overcome by his feelings and indulging them freely. How well this acts
as a foil to the self-controlled demeanour of the king will be seen. The Herald
belongs to a class of society? which does not feel bound to practise the
εὐσχημοσύνη and the restrained utterance of the real gentry. It is significant
that the prologue of Aeschylus’ Mvoot (fr. 143 N.) began in just the same way:
ids Kdixe Μύσιαί τ᾽ ἐπιρροαί. This is the salutation spoken on arrival in Mysia
by the same servant of Telephus who shortly afterwards says (fr. 144) ποταμοῦ
Καΐκου χαῖρε πρῶτος ὀργεών, εὐχαῖς δὲ σώιζοις δεσπότας παιωνίαις.
From the form of the expression we may gather perhaps the way in which
the passage was acted. ‘From his first words (οὖδας) it would seem that he
throws himself down, like Shakespeare’s Richard II, to salute the beloved
earth’ (Verrall). I have been led to the same conclusion* by observing the
stylistic level of ἰώ, and also by remembering certain passages in the Odyssey,
1 Blomfield might be excused for clinging to rules which, when he was young, were by
most people considered canonical. He admitted that Humboldt’s book was of little use to
him, ‘quum Germanice parum sciam’ (p. xiv). It is doubtful whether, if he had known
German, he would have been able to appreciate Humboldt’s view on the disregard of the
real time.
2 Murray repeats Blomfield’s error ‘ μετανάστασις ᾿. He also follows Blomfield in com-
paring Eum. 231, where, however, the Chorus make it perfectly clear that they are leaving
the stage; he further adds a comparison with Eum. 565, but there the Chorus remain in the
orchestra, whatever Verrall says to the contrary.
3 For the status of heralds from the point of view of Athenian society cf. Wilamowitz,
Eur. Herakles, i, 2nd ed., p. 122 n. 18.
4 I have since found that W. v. Humboldt took the same view. His wife writes to him
(29 July 1809) : ‘if I do not kiss the soil of my native land like the herald in the Agamemnon’.
256
 COMMENTARY line 504
ὃ 521 f. ἦ τοι ὃ μὲν (Agamemnon) χαίρων ἐπεβήσετο πατρίδος αἴης, καὶ κύνει
ἁπτόμενος ἣν πατρίδα (the scholion here is ἔθος εἶχον οὗ ἀποδημοῦντες τῆς
πατρίδος, ὅταν ἐνδημήσωσι, κυνεῖν αὐτὴν καὶ κατασπάζεσθαι), ε 463, ν 354
(Odysseus) κύσε δὲ ζείδωρον dpovpav, and finally the monologue, with its
obviously ‘para-tragic’ colouring, at the entry of Plutus when he returns
home after his cure (Ar. Plut. 771 ff.) καὶ προσκυνῶ γε πρῶτα μὲν τὸν ἥλιον,
ἔπειτα σεμνῆς Παλλάδος κλεινὸν πέδον χώραν Te πᾶσαν Κέκροπος ἥ u’ ἐδέξατο.
Prayers like this of the returning traveller play a large part in the whole of
Attic drama,? cf., e.g., Aeschylus in a papyrus fragment to be published in
Pap. Oxy. Ὁ - μὲν εὐχαῖς πρῶτα πρεσβεύων σέβω, Ὁ — ἱκνοῦμαι φέγγος ἡλίου τὸ
νῦν, Eur. fr. 558 N. ὦ γῆς πατρώιας χαῖρε φίλτατον πέδον Καλυδῶνος κτλ. (pro-
logue, Diomedes speaking), Or. 356 ff., Ar. Ach. 729 ff. (the man is not return-
ing to his home, but the form is the same), Menander fr. 13 Kock χαῖρ’, ὦ
φίλη γῆ, διὰ χρόνου πολλοῦ σ᾽ ἰδὼν ἀσπάζομαι «rA., Plaut. Most. 431 ff. (Phile-
mon), Bacch. 170 ff. (Menander) erilis patria salue, quam ego biennto, post-
quam hinc in Ephesum abit, conspicio lubens. saluto te, vicine Apollo, qui aedibus
propinquos nostris accolis, Stich. 649 f. (Menander). In form the prologue to
the Euripidean Telephus is to be classified with these, ὦ γαῖα πατρίς, ἣν Πέλοψ
ὁρίζεται, χαῖρε κτλ. (fr. 696 N., to which can now be added a piece of the
continuation from the Milan papyrus published by Calderini, cf. Körte,
Archiv für Papyrusforsch. xiii, 1939, 98, and D. L. Page, Greek Lit. Papyri, i.
130), although Telephus is not coming back to his former homeland in the
usual sense (after a campaign or a journey). Owing to the fact that the pas-
sage Ag. 503 ff. is couched in terms of prayer the speaker pays no attention to
the presence of the Chorus (the same applies to 810-28).
504. δεκάτωι σε φέγγει τῶιδ᾽ ἀφικόμην ἔτους. The conjecture δεκάτου made
by Henry Jacob and adopted by Stanley is easy enough in itself and has
always found some approval, but without sufficient reason. However, the MS
reading must not be defended in the way Schütz defends it : ‘nihil mutandum,
nam e vulgata idem sensus oritur’. This is not the case. If we write δεκάτου,
it can only mean ‘on this day of the tenth year’, Headlam translates quite
correctly, though with a significant addition : ‘on this blessed day in the tenth
year’. This would be laying undue emphasis on the day in the year. The
point that matters is: ‘after ten years’ or ‘in the tenth year’; we do not need
Dikaiopolis’ prayer (Ar. Ach. 266 Exrwı σ᾽ ἔτει προσεῖπον εἰς τὸν δῆμον ἐλθὼν
ἄσμενος) and like passages in order to understand what is obvious here. The
words of the MSS mean ‘with (or ‘in’) this tenth year’s light’ (so Voss and
Nägelsbach correctly, and Verrall, see below). From the point of view of
syntax and style the expression is quite normal: it is an instance of the
τ We find the corresponding gesture upon saying good-bye to the place where one has
lived for a long time in 5. Phil. 533, 1408. (The view of Kock and Neil, on Ar. Knights 156,
that even in the case of τὴν γῆν προσκυνεῖν it is not a question of stooping down to the
earth, but of throwing it a kiss with the hand, should be discarded. In general cf. against
this widely held view the sound observations of H. Bolkestein, “Theophrastos’ Charakter
der Deisidaimonia’, Religionsgesch. Versuche u. Vorarbeilen, xxi. 2, 1929, 23 ff.)
2 Their influence extends still farther afield. In certain expressions used by Catullus in
poem xxxi (where of course the strength of the poet’s personal feeling is undeniable) we
may still find some connexion with such regular patterns of prayers and monologues in
drama on the part of homecoming travellers: 12 f. salve o venusta Sirmio atque ero gaude;
gaudete vosque . . . (Ag. 508 νῦν χαῖρε μὲν χθών, χαῖρε δ᾽ ἡλίου φάος), 5 f. vix mt ipse credens . . .
videre te in tuto (Ag. 506 οὐ γάρ ποτ᾽ ηὔχουν «rA.).
4872-2 S 257
line 504 COMMENTARY

particular form of ‘enallage’ which has been dealt with by Lobeck on 5. Aj. 7,
and subsequently e.g. by Wecklein on A. Eum. 292, Jebb on S. Ant. 794,
Wilamowitz on E. Her. 468, Kaibel on 5. El. 1374, Williger, Sprachl. Unters.
9 ff. Wilamowitz among others stresses the point that in expressions to be
classified under this head we must ‘think of the substantives as coalescing
into a single compound’; similarly Jebb characterizes νεῖκος ἀνδρῶν as ‘one
notion’; A. S. Owen renders δεκάτωι... μηνὸς ἐν κύκλωι in E. Jon 1486 by ‘in
the tenth month’s-circle’ ; cf. Mooney on Apoll. Rh. 4. 1519 λαιὸν... ταρσὸν
ποδός : ταρσὸν ποδός is treated as a compound, "'footsole" '. In our passage
the underlying idea is remarkable. Aeschylus is apparently transferring to
the year what is properly applicable to the day, and therefore speaks of the
‘year-light’. Verrall is substantially correct : ‘With this tenth annual dawn, if
the expression may pass. φέγγος ἔτους is an imitation of the common phrase
φέγγος ἡμέρας. Similarly Plüss: ‘The year has its light like the day.’ We
could understand this well enough without any parallels. But it is no un-
welcome confirmation when we read in Hdt. 1. 190. 1 καὶ τὸ δεύτερον ἔαρ
ὑπέλαμπε and in 8. 130. 1 ἔαρος δὲ ἐπιλάμψαντος, the latter instance being all
the more to the point because elsewhere (4 times) in Herodotus ἐπιλάμπειν is
used, with ἡμέρη as subject, only of the dawn of day.
505. Schol. ἡ λέξις ἐκ μεταφορᾶς τῶν ἀγκυρῶν, dv πολλῶν ῥαγεισῶν εἰς τὸ ὕδωρ
μία τις περισώιζει τὴν ναῦν. This is in the main right, but ῥαγεισῶν suggests
that the thought is of anchor-cables, or more accurately, since there is no
mention of an anchor, of mooring-cables (see account in Torr, Ancient Ships,
13 1). The ‘metaphor’ is an obvious one, cf. A. C. Pearson on E. Hel. 277
(evidence from sth-century texts); van Heusde quotes Stob. Flor. 110. 22
(vol. iv. 1001 Hense) οὔτε ναῦν ἐξ ἑνὸς ἀγκυρίου οὔτε βίον ἐκ μιᾶς ἐλπίδος
ὁρμιστέον. “Ihe Christians regarded the anchor as the emblem of hope’
(Eitrem, Symb. Osl. xix, 1939, ὃς, producing evidence from early Christian
times). The implication of the image is the same in the case of the mooring-
cables as in that of the anchor. [See Addenda.]
506. ‘For I never trusted (I had no confident belief) that I should . . .”. For
the meaning and construction of αὐχεῖν see on 1497.
507. Examples that occur elsewhere of μετέχειν μέρος τινός (cf. L-S μετέχω 2,
and add Cho. 291 f.), as well as the sense, are against understanding τάφου as
epexegetic genitive.? There is a clear partitive relation between μέρος and
τάφου. The lucky man who returns home gets his share in the τάφος which as
a whole is allotted to all those who die in the home-country.
509. Wilamowitz (at the end of his translation of the Oresteia, Griech.
Tragoedien, ii. 309) refers to Livy, who says (32. 25. 2) with regard to the
popular assembly of Argos in 198 B.C. mos erat comitiorum die primo velut
ominis causa practores pronuntiare Iovem Apollinemque et Herculem. Accord-
ing to Wilamowitz Aeschylus has deliberately named the city gods of Argos
1 I entirely disagree, however, with Verrall’s inferences, which are connected with the
whole system of his chronological reconstruction. His inexact translation ‘tenth ...dawn’
instead of ‘tenth . . . light’ is also influenced by his chronological speculation : Verrall (cf.
his Introd. p. xli n. 3) attaches importance to the beginning of the year, of which Aeschylus
says nothing.
2 On the other hand, in the second epigram for the Athenians killed at Potidaea (7G i.?
945, Tod, Greek Hist. Inscr. no. 59) 1. 3 ἐχθρῶν δ᾽ of μὲν ἔχουσι τάφου μέρος, of δὲ φυγόντες κτλ.
the genitive τάφου is clearly epexegetic.
258
 COMMENTARY line 511
here, because of his knowledge of local circumstances. This is certainly
possible, though not cogent, for e.g. in Athens Ζεὺς Epreios and Ἀπόλλων
πατρῶιος (the Pythian Apollo as here, cf. Wilamowitz, Aristoteles und Athen,
ii. 45, and again in his edition of Eur. Jon, p. 2, and Glaube der Hell. ii. 34;
Ferguson, Hesperia, vii, 1938, 30 f.) are closely connected (Pl. Euthyd. 302 ἃ,
Arist. 40. πολ. 55. 3, Demosth. 57. 67, Harpocration s.v. ἑρκεῖος Ζεύς; see
Homer A. Thompson, Hesperia, vi, 1937, τος, for the possible partnership of
Zeus and Apollo in the temple on the west side of the Agora). For the special
emphasis on AmoAAwv ὁ Πύθιος in a prayer addressed to all gods ὅσοι τὴν
χώραν ἔχουσι τὴν 'Árrucjv see Demosth. 18. 141. We should very much like
to know whether we are to suppose with Wilamowitz that the altars of Zeus
and the Pythian Apollo are to be seen in the orchestra in front of the palace
of the Atridae. As will be shown on 518, the first portion of the prayer (to
517 inclusive) contains nothing that necessitates such a supposition. Nor does
ı08ı enable us to settle the question whether in that passage an altar of
Apollo is assumed to be in front of the palace.
The polysyndetic series with re... re... re etc. (twice in 509, then 515 f.,
516, and twice in 519) is a common arrangement in such invocations of
companies of gods, cf. Philol. Ixxxvi, 1931, 9 f., to which might be added e.g.
A. Sept. 271 ff., E. El. 674 ff., Herodas 4. 3 ff., cf. too the prayer-parody in
Ar. Knights 634 ff.
For the rare form of the ‘anapaest’ at the beginning of 509 cf. p. 130 (about
Suppl. 282).
510 f. The fusion of the homeland cult in the poet’s own time with the deeds
of the Homeric god is characteristic of the world in which Aeschylus’ charac-
ters move. For a prayer as a setting for pieces of narrative cf. on 813.
510. ' μηκέτ᾽ eis ἡμᾶς go together, hence the order’ (Headlam). Cf. on 931.
At first it looks as if the Herald were going to address Apollo with but a
slight variation of one of the god’s usual ἐπικλήσεις : ὁ Πύθιός τ᾽ ἄναξ, τόξοις
ἐάπτων.. . . βέλη (one could easily conceive a form of the clause in which an
epithet was added either to τόξοις or to βέλη), cf., e.g., hymn. Hom. Ap. 140
ἀργυρότοξε ἄναξ ἑκατηβόλ᾽ Ἄπολλον. Then, in the midst of his phrase, he
suddenly seems to recall the sufferings that were brought upon the army by
Apollo’s archery, so he inserts the words μηκέτ᾽ εἰς ἡμᾶς and elaborates them
in the following lines.
511. ἅλις : ‘enough and more than enough’, cf. on 1659.
Needham’s (cf. Ceadel, C.Q. xxxiv, 1940, 60) correction ἦσθ᾽ was made
known by Butler, and put into the text by Blomfield. Examples of wrongly
written ἦλθε (E. Ion 577) and other intruded forms of ἐλθεῖν are given by
A. S. Owen on E. Jon 828 ; add e.g. A. Cho. 536.
The allusion to the plague carries the Herald away, so he turns in a paren-
thesis direct to the god, in a personal outburst, with a tinge of naive familiar-
ity. On the one hand he has not yet completely forgiven Apollo, on the other
hand he wants to flatter him into benevolent behaviour. The parenthesis
extends beyond 511, to which Wilamowitz and others limit it, as far as
Ἄπολλον in 513, then the interrupted list of gods is resumed. Only at the end
of the first part of the prayer, 517, is a direct petition (δέχεσθαι) expressed
(in 510 the wish is expressed by way of implication) ; for the rest this first
section is confined to mere invocation.
259
“line 512 COMMENTARY
512. αὖτε: not ‘be once more our saviour’ (L. Campbell), but αὖτε here only
strengthens the adversative δέ, as also e.g. in 330, 553, 558.
513. τούς τ᾽ Gywvious θεούς. There is a certain amount of confusion in the
explanations that have been advanced. Stanley based his interpretation on
Hesychius’ ἀγώνιοι θεοί" of τῶν ἀγώνων προεστῶτες. This has strongly influenced
later views. Hermann and other scholars who related the gloss of Hesychius
. to this passage understood ἀγῶνες, in an obviously arbitrary way, to mean
war. Stanley observes on 515 'Mercurium ἐναγώνιον intelligit ’, in spite of the
fact that Hermes is quite clearly invoked here as the protector of κήρυκες ; the
fact that ἐναγώνιος “Ἑρμῆς is found elsewhere in Aeschylus (fr. 384 N.) and
that he is also called ἀγώνιος at this period (Pind. Zsthm. 1. 60) does not affect
the matter. Still greater harm has been done by the dragging in (disapproved
by Blomfield) of the ancient explanation of Q 1 (Scholia BT and Eustath.
p. 1335. 55), Which must be quoted in full: ἀγών’ ἢ τὸ πλῆθος, 'Üctov δύσονται
ἀγῶνα᾽ (Z 376): ἢ αὐταὶ ai διαγωνίσεις. παρὰ δὲ Βοιωτοῖς ἀγὼν ἡ dyopd: καὶ τὸν
ἀγορανόμον ἀγώναρχον; καλοῦσιν. καὶ ἀγωνίους θεούς Αἰσχύλος τοὺς ἀγοραίους.2
καὶ ᾿Ησίοδος (Theog. 91) ἐρχόμενον δ᾽ av’ ἀγῶνα᾽. In this comment on Aöro
δ᾽ ἀγών we have as full a survey as possible of the meanings of ἀγών. The
statement about the Boeotian agonarchs has been corroborated by the
Hellenistic inscription from Thespiae, IG vii. 1817 = Collitz-Bechtel, i, no.
812a (an excerpt in Schwyzer, Exempla, no. 550g), which begins rd aywvapxv
τὺ ἐπὶ Πολέαο ἄρχοντος, for it is clear that the agonarchs here correspond to
the ἀγορανόμοι, for whose annual term of office also (at Thespiae and else-
where) we have the evidence of inscriptions (cf. Fiehn, RE vi. A. 42. 8; he
should have mentioned the agonarchs as well). In the title of these magis-
trates the meaning of ἀγών is ‘market’, i.e. the public place in which com-
mercial business is transacted. It is very possible, in fact the arrangement of
the different explanations makes it probable, that the scholiast, or his author-
ity, who understands Aeschylus’ ἀγώνιοι θεοΐ as gods of the ἀγορά, was
thinking of the market as a business centre (this side comes out particularly
strongly in ‘Epuñs ἀγοραῖος) at least as much as of the political assembly,
which may have been suggested e.g. by a recollection of Ζεὺς ἀγοραῖος (men-
tioned by Aeschylus in Exum. 973). Otfried Müller (Anhang zu Aesch. Eum.,
p. 38), referring to the scholia on 21, finds both meanings in Ag. 513; his
treatment of the passage has had a strong influence right down to the present
time. He says: ‘The orchestra, in which the worthy elders, the πρέσβος
Ἀργείων, are gathered together, must represent a public place, a market-place
for assemblies, which in ancient Greek cities was probably often in front of
1 Since the plural rù ἀγώναρχυ has come to light in the inscription from Thespiae cited
farther down, the suggestion made by E. Maass (in his edition of the Scholia Townleyana)
that ἀγωνάρχην should be read in accordance with Eustathius must be dismissed.
2 It was incautious of E. Maass to insert the Agamemnon passage after Αἰσχύλος, and of
Wilamowitz to produce the Homer scholion as ‘Testim.’ on Ag. 513. It might equally well
refer to one of the four passages in the Supplices (as Hermann thinks), to say nothing of the
lost plays. The use of the accusative of course proves nothing at all, cf., e.g., schol. O 27
ὡς Εὐριπίδης (Phoen. 210 ὑπὲρ ἀκαρπίστων πεδίων) dxdpmora πεδία λέγων τὴν θάλασσαν.
3 So Foucart and, following him, Oehler, RE i. 883, 40, Bechtel, Griech. Dial. i. 303,.
Schwyzer, loc. cit., similarly the dictionaries of Passow-Crönert and L-S (though in the
latter with the misleading addition ᾿ἀγών = assembly’). R. Meister in Collitz-Bechtel, i, .
no. 812a is mistaken: “There were four agonarchs (= agonothetae) mentioned in the:
inscription’.
260
 COMMENTARY line 513
the palace of the ἄνακτες. Only here could the altars of the market-gods be
found, and according to the anapaests of the parodos it is likely that these
were visible [cf. on 90]; these gods are substantially identical with the
ἀγώνιοι θεοί. Neither in the Agamemnon nor in the Supplices can the ἀγώνιοι
θεοί be gods of contest, but they are (as correctly interpreted by I. G. Schneider)
gods of the Assembly’, etc. The last expression, which is somewhat indefinite,
is found still in Sidgwick and Headlam on Ag. 513: ‘gods of the gathering’,
‘gods of assembly’, whereas the term ‘market-gods’ is used by Wilamowitz
(translation and Interpr. 169) and Passow-Crönert (even in the passages of the
Supplices). Now it is clear that in A. Suppl. 189, 242, 333, 355 the ἀγώνιοι θεοί
have nothing to do with the market-place : their τέμενος (for this cf. especially
Wilamowitz, Interpr. 6 f., and for the θεῶν ἀγορά in general see his observa-
tions ibid. 72 f.) is on a πάγος (189) outside Argos. They are correctly defined
by Wecklein! (on Suppl. 189 [195 Weckl.] and on Ag. 513 [518 Weckl.]) as ‘the
deities linked together in a company of gods (θεῖος ἀγών Hom. H 298 [cf. the
Aristonicus scholion in A and Leaf ad loc., who also interprets A. Ag. 513
quite correctly]), to whom sacrifices are offered in common on one altar, and
who therefore form a ξυντέλεια (Sept. 251, Schol. τὸ κοινὸν ἄθροισμα τῶν θεῶν).
Cf. Suppl. 222 πάντων δ᾽ ἀνάκτων τῶνδε κοινοβωμίαν". Cf. also Sept. 219 f.
θεῶν ἄδε πανάγυρις. These gods are therefore not the ‘gods of assembly’, but
rather ‘gods in assembly’,? di consentes in the proper sense (“who are together’).
So the rendering in the interlinear gloss in F τοὺς ἅμα évi τόπωι ἱδρυμένους is
probably correct. It is, however, obvious that, as far as the actual cults are
concerned, such a company of gods often coincides with the gods of the
market. 'The ἀγορά is the favourite place for the worship of the twelve gods'
(Weinreich in Roscher's Lexikon, vi. 837, with many references in detail).
The Americans have recently excavated the altar of the twelve gods in the
Agora of Athens; as for Rome, everyone will immediately recall eos urbanos
(i.e. duodecim deos consentis), quorum imagines ad forum auratae stant, sex
mares et feminae totidem (Varro, v. rust. 1. 1. 4). But in the Supplices at any
rate the ἀγώνιοι θεοΐ cannot possibly be equated with market gods (see above),
and it is very doubtful whether in Ag. 513 Aeschylus' use of the expression is
different from that in the earlier play. We have to reckon with the possibility
that he was the first to speak of ἀγώνιοι θεοί, following the model of Homer's
θεῖον ἀγῶνα" or something of the kind ;5 in any case no other evidence for the
term is available.$ It is therefore likely that Aeschylus used the term con-
sistently. There is no reason at all to expect a mention of the market in
Ag. 513. O. Müller is of course right in supposing that in ancient Greek
τ Verrall rightly followed him.
2 This correct explanation is given first place in L-S. The meaning next proposed as
alternative follows the note by Tucker (i.e. really Stanley) on Suppl. 189, which had already
been proved wrong by Otfried Müller. The phrase τοὺς ἀγωνίους θεούς of our passage is
rightly rendered by Groeneboom on Sept. 219 as 'de verzamelde goden'.
3 In order to appreciate the quality of this interpretation, it is sufficient to turn to the
gloss in M on Suppl. 189 (ἀγωνίων θεῶν) : στρογγύλα γάρ ἐστι rà ἱερεῖα καὶ γωνίας οὐκ ἔχοντα
(cf. schol. BT on 2 r, Etym. M. p. 15. 55, Paulus Festi p. 10.8M.).
4 Quoted as Aeschylus’ model in schol. Sept. 219 (205 Wecki.).
5 ἀγών is used in the Homeric sense in Ag. 845.
6 It is quite irrelevant in this connexion that Plato, Laws 783 a, uses ἀγωνίοισι θεοῖς for
the gods of the ἀγῶνες (‘games’). The gloss in Hesychius quoted above ἀγώνιοι θεοί: of τῶν
ἀγώνων προεστῶτες suits that passage.
261
 line 513 COMMENTARY

cities the market ‘may often have been in front of the palace of the ἄνακτες":
a good example is afforded in Pindar’s description of Cyrene (P. 5. 03 ff.),
where the Agora lies near the king’s palace (cf. Wilamowitz, Pindaros,
379 f.). But in the Agamemnon no point is made of the market’s being near
at hand,’ so it was of no consequence to the poet. A decisive argument comes
from the structure of the speech: only from 518 onwards is any notice taken
of the locality, which was partly represented in the background of the
orchestra, while many details were left to the imagination of the audience.
The invocations of the gods in the earlier part of the speech are not related to
anything actually visible (cf. on 518). So a localizing touch such as is implied
in Wilamowitz’s translation ‘Und euch auch grüss’ ich alle, die ihr rings den
Markt beschützend thront’ (‘and I greet you all who are enthroned as pro-
tectors around the market-place’) has no sufficient foundation.
But what place has the invocation of the ἀγώνιοι θεοί in the context of this
prayer? What is its function? The answer must show whether the meaning
adopted above for ἀγώνιοι θεοί is adequate. A simple consideration leads to
our end. It would be absurd to exclude from the community of ‘all the
assembled gods’ those who were mentioned before, viz. Zeus, the highest lord
of the land, and the Pythian Apollo (sog ff.). Rather is it obvious that the
speaker who is here praying ‘post specialem invocationem transit ad generali-
tatem, ne quod numen praetereat'.? Due attention has recently been given
to this formula of completion in prayers. A good illustration is provided by
the parody prayer in prose Ar. Birds 865 ff. Εὔχεσθε τῆι ‘Eorias τῆι ὀρνιθείωι
καὶ τῶι ᾿Ικτίνωι τῶι ἑστιούχωι καὶ ὄρνισιν ᾿Ολυμπίοις καὶ ᾽᾿Ολυμπίηισι πᾶσι καὶ
πάσηισιν.. We hear in the parodos of the Septem (116 ff.) a long prayer to
Zeus, Athene, Ares, Aphrodite, Apollo, Artemis, and Hera. Then these
words follow: io παναρκεῖς θεοί, ἰὼ τέλειοι τέλειαί τε γᾶς τᾶσδε πυργοφύλακες
«TA. ; these gods are not different from those previously invoked ; so this, too,
is a generalis invocatio by way of completing the special one. Ag. 513 approxi-
mates still more closely to the common formula of prayers with the πάντας
in τούς τ᾽ ἀγωνίους θεοὺς πάντας. An individual colouring is given to the
Herald's prayer by the fact that in the same context as the generalis invocatio
he ends up with a special emphasis on the divine protector of his own calling.
Conscientious commentators have observed that τόν 7’ ἐμὸν τιμάορον ‘Epuñv
must be taken as forming part of what has preceded it, e.g. Wilamowitz:
! Euripides, on the other hand (El. 708 ff.), not only makes it clear that the palace of the
. Átridae is supposed to be directly adjoining the market-place, but uses this detail to give
greater liveliness to an important episode in the earlier history of the family.
2 Servius on Virgil, Georg. i. 21; he adds the substantial statement: ‘more pontificum,
quoniam ritu veteri in omnibus sacris post speciales deos, quos ad ipsum sacrum quod fiebat
necesse erat invocari, generaliter omnia numina invocabantur.' Cf. on this point Wissowa,
Religion u. Kultus, 2nd ed., 38, and Hermes, lii, 1917, 92 f.
3 In addition to Fr. Jacobi’s dissertation ΠΑ͂ΝΤΕΣ @EOI (Halle 1930), which is of
fundamental importance for thc Hellenistic period, cf. Wilamowitz, Glaube d. Hell. ii. 344 f.,
and ©. Kern, D. Religion der Griech. ii. 38 ff. Kern rightly relates these formulae in prayers
to the idea of the ἀγορὰ θεῶν; A. Ag. 513 (and Sept. 166 ff.) he does not mention.
* Kock's insertion of a bird representing Poseidon after πάσηισιν has been put into the
text by van Leeuwen, O. Schroeder, and Kern (see previous note), but it is wrong. It is
quite in the manner of Peisetairos to draw an obvious conclusion from the mention of
ὄρνιθες "OAyumoı πάντες and to break in with: ὦ Zovwépaxe (cf. Knights 560), χαῖρ᾽ ἄναξ
Πελαργικέ.
262
 COMMENTARY line 519

‘vor allen Hermes’ and Headlam : ‘and chiefly my own patron, Hermes’. For
this form of copulative inclusion cf. on 5.
To sum up. Strong arguments are in favour of the interpretation of
ἀγώνιοι θεοί which has been adopted above. However, the possibility that the
expression here means ‘gods of the gathering’ cannot be excluded. It might
be supported by 845 κοινοὺς ἀγῶνας θέντες ἐν πανηγύρει. The inclusion of
Hermes in the number of the divine patrons of the assembly would be natural
enough.
514. προσαυδῶ, the Homeric word, found also in a salutation in Pers. 154
and elsewhere in Tragedy. The word used in everyday language for such an
address of greeting (and especially at reunion after long parting, as here) is
προσειπεῖν, e.g. Ar. Ach. 266 ἕκτωι σ᾽ ἔτει προσεῖπον, Peace 557 προσειπεῖν
βούλομαι τὰς ἀμπέλους. Thus Agamemnon himself in his first speech, which is
parallel to the Herald's speech, 810 f. πρῶτον μὲν Ἄργος καὶ θεοὺς ἐγχωρίους
δίκη προσειπεῖν, SO, e.g., E. Her. 599 f. ἐσελθὼν (Wilamowitz, προσελθὼν MS)
νῦν πρόσειπέ θ᾽ ἑστίαν καὶ δὸς πατρώιοις δώμασιν σὸν ὄμμ᾽ ἰδεῖν, 608 f. οὐκ
ἀτιμάσω θεοὺς προσειπεῖν πρῶτα τοὺς κατὰ στέγας, Phoen. 633 οὐ γὰρ old’ ei
μοι προσειπεῖν αὖθις ἔσθ᾽ ὑμᾶς (his home and the θεῶν ἀγάλματα) ποτε.
τιμάορον : here and in Suppl. 43 rightly explained in the scholia as βοηθόν;
‘properly “watching over the position (τιμή), the office" . . . hence "keeper,
protector", also ‘‘avenger” [so Ag. 1280]’ (Wackernagel, Syntax, i. 126 f.).
515. Cf. the pedestal from the Acropolis of an ἄγαλμα offered to Hermes by an
Athenian κῆρυξ with dedicatory inscription (in στοιχηδόν script) IG 1.2 631.
516. For the inclusion of the heroes in prayers of this kind cf., e.g., Thuc. 2.
74. 3 θεοὶ ὅσοι γῆν τὴν Πλαταιΐδα ἔχετε καὶ ἥρωες, 4. 87. 2 μάρτυρας . . . θεοὺς καὶ
ἥρως τοὺς ἐγχωρίους ποιήσομαι, Lycurg. I εὔχομαι... τῆι ᾿άθηνᾶι καὶ τοῖς
ἄλλοις θεοῖς καὶ τοῖς ἥρωσι τοῖς κατὰ τὴν πόλιν καὶ τὴν χώραν ἱδρυμένοις, with the
comic travesty in Ar. Birds 881. In Hdt. 8. 109. 3 Themistocles says the
θεοί τε καὶ ἥρωες had achieved the victory over the Persians (cf. also 8. 143.
2); this is in keeping with the idea of this passage (ἥρως τοὺς πέμψαντας). ‘In
contrast to the epic hero-less world there can be no consciousness or idea of
the Polis without fellowship with the ἥρωες. As well as the gods the spirits
of those fallen in battle are above all the indispensable guarantors of the
Polis’ (K. Reinhardt, Hermes, lxxvii, 1942, 234).
518. ἰὼ μέλαθρα κτλ. With the repetition of the passionate exclamation (cf.
503) a new section begins, i.e. the Herald now turns with a determined move-
ment towards the front of the house. The following words of address are all
evoked by what he sees directly in front of him: μέλαθρα, στέγαι, θᾶκοε,
δαίμονες ἀντήλιοι. That was not so in the first part of his speech. It is true
that there he probably began by touching the soil of his homeland, but then
he greeted the sunlight, Zeus, Apollo, and all the other gods. They are his
homeland’s gods, but all the same the manner of the greeting in the first part
is not conditioned by anything directly perceptible on the spot.
519. θᾶκοι: the form θᾶκος is normally used in Athens, cf. Meisterhans, 16;
Wilamowitz, Platon, ii. 339. For the subject-matter Paley appositely com-
pares y 406 ff. ἐκ δ᾽ ἐλθὼν (Nestor) κατ᾽ ἄρ᾽ ELer’ ἐπὶ ξεστοῖσι λίθοισιν, of of
ἔσαν προπάροιθε θυράων ὑψηλάων λευκοί, ἀποστίλβοντες ἀλείφατος" οἷσ᾽ ἔπι μὲν
πρὶν Νηλεὺς tleaxev . . . Νέστωρ αὖ τότ᾽ edile.... σκῆπτρον ἔχων. It is the seat
in front of the house, on which the king in the heroic age is throned in state
263
line 519 COMMENTARY

when in the exercise of his office he is administering justice, presiding in the

men’s assembly, etc. On the François vase (Furtwängler-Reichhold, pl.
11 Í) we can see Priam outside Troy in front of the city gate sitting on a
@AKOZ. Here the plural θᾶκοι is used either with reference to the 8íüpovos
καὶ δίσκηπτρος τιμή (42 f., cf. 109) of the Atridae, or simply on the analogy of
θρόνοι (found in Aeschylus even of the seat of a single individual, cf. λέχη,
eivai, etc.) and the like.
ἀντήλιοι: for the unaspirated form in this and other compounds see
Lobeck’s detailed note on S. Aj. 805; for the rarity of this phenomenon in
Attic cf. K. Meister, D. homer. Kunsisprache, 215 (on p. 216 he gives a not
particularly convincing conjecture that 'considerations of perspicuity' account
for the unaspirated ἀντήλιος). The forms ἀντήλιος, ἀπηλιώτης, and the like are
attested as Attic by Phrynichus, Praef. soph. p. 25. 19 and fr. 249 (p. 164. 11)
dé Borries. On the strength of this and the occurrence in Aristophanes of
forms such as ἀπηλιαστά (Birds 110) etc., Lobeck refused to take these forms
as Ionisms: they are again stated to be such by H. W. Smyth, Jon. Dial. 333,
and (with reservations) by Gustav Meyer, Griech. Gramm., 3rd ed., 326, and
L-S. But it is precisely in Ionic inscriptions of the earlier period that we find
καθημένδ, μεθέληι, κάθοδον (Wackernagel, Uniers. z. Hom. 24 n. τ, Bechtel,
Griech. Dial. iii. 38). For further discussion see the books and articles quoted
by Ernst Fraenkel, Nomina agentis, ii. 207 n. 1.
The sense of ἀντήλιος is clearly 'facing the (rising) sun, turned towards the
East’, in two passages of tragedy, viz. S. Aj. 805 f. of δ᾽ éamépovs ἀγκῶνας, oi
δ᾽ ἀντηλίους ζητεῖτ᾽ ἰόντες κτλ. and E. Ion 1549 f. τίς οἴκων θυνοδόκων ὑπερτελὴς
(i.e. over the east front of the temple of Apollo) ἀντήλιον πρόσωπον ἐκφαίνει
θεῶν; In Ag. 519 we may be content to assume (with Klausen and Otfr.
Müller, Anhang zu den Eumeniden, 38) that some undefined deities are referred
to, whose statues stand in front of the façade of the royal house, which we
are to suppose faces eastwards. But it is at least as probable that divine
beings of a special type are meant, whose name, δαίμονες ἀντήλιοι, conveyed
something quite definite to the Athenians. Unfortunately we know hardly
anything about them. Stanley (on this passage) has quoted the evidence;
Welcker, Götterlehre, iii. 214, deals with these particular δαίμονες in the briefest
way conceivable ; other handbooks of Greek religion do not mention them at
all. Tümpel, however, RE i. 2349 f., says pretty well everything there is to be
said about them. We should like to know the context of Eur. fr. 538 N., but
all we learn about it is from Hesychius, s.v. ἀντήλιοι θεοί: οἱ πρὸ τῶν πυλῶν
ἱδρυμένοι. Εὐριπίδης Μελεάγρωι.Σ To this we may add Tertullian, De idol.
15 (middle) Etiam apud Graecos Apollinem θυραῖον et Antelios daemonas
ostiorum praesides legimus; De corona 13 (end) ‘Roman gods, whom the
Devil ostits affinxerit . . . et apud Graecos Thyraeum Apollinem. et Antelios
daemonas’. It is unlikely that Hecate to whom someone prays in Aeschylus
fr. 388 N. δέσποιν᾽ Ἑκάτη τῶν βασιλείων πρόδομος μελάθρων, should be in-
cluded among the δαίμονες ἀντήλιοι (Ar. Wasps 802 ff. κἀν τοῖς προθύροις
ἐνοικοδομήσοι. . . Sixagrypisiov . . . ὥσπερ 'Exárewv πανταχοῦ πρὸ τῶν θυρῶν);
we do not know what deities belonged to that category. When Orestes and
Pylades (S. El. 1374 f.) go into the royal house πατρῶια προσκύσανθ᾽ ἔδη θεῶν
ὅσοιπερ πρόπυλα ναίουσιν τάδε, the cult-statues and altars there mentioned are
1 Dan. Heinsius, ad loc., quoted the passages from the Agamemnon and from Tertullian,
264.
 COMMENTARY line 522
perhaps not different from the δαίμονες ἀντήλιοι. But though we know very
little of these δαίμονες, yet we have two points settled for Ag. 519. In the first
place the δαίμονες ἀντήλιοι cannot simply be identical with any of the gods
addressed in 509-15, for in that case, quite apart from the different turn in the
two parts of the prayer (cf. on 518), the concluding generalis invocatio 513 f.
would be meaningless. In the second place, as results from the context here
and particularly from 520, their cult-statues must be imagined as present on
the spot (no matter whether they were actually visible in the orchestra),
while this is unnecessary in the case of the gods addressed in the first half of
the prayer.
520. ei (Auratus’ excellent emendation) wou πάλαι: a variant of the familiar
formula (ei more . . .), whereby the prayer appeals to previous instances of
divine kindness and help.'
τοισίδε :? Bechtel, Griech. Dial. iii. 167, has tried to narrow down the extent
of the use of this form even in Herodotus, and in Tragedy it certainly leads
a precarious existence. I have not checked the material again, as one would
need to do. However, it is certain that in S. Phil. 956 we must keep τοισίδ΄,
which is the reading of the MSS (further corroborated by Suidas s.v. ὀρει-
Barns); in E. Hipp. 400 τοισίδ᾽ is probably right, instead of the τοῖσιν of the
MSS. So we shall not venture to object to τοισίδ᾽, as Hartung and others have
done, because the form is without parallel in Aeschylus. Nor need there be
any objection to the sense. Klausen's explanation of τοισέίδε: “δεικτικῶς de
tempore: nunc, ut alibi de loco' (so Headlam again) was rightly rejected by
Ahrens, 537. The deictic function of ὅδε is bound to be local here as elsewhere,
and this is in keeping with the general character of this section (cf. on 518).
On the one hand there is a reference in φαιδροῖσι τοισίδ᾽ ὄμμασι to the morning
sunshine, which is now lighting up the δαίμονες ἀντήλιοι; this was recognized
by Schneidewin, Verrall, Wilamowitz (cf. his edition and also Griech.
Tragoedien, xiv, 1923, 41) ; on the other hand, the phrase indicates the facial
expression of the deities. Verrall says rightly : ᾿φαιδροῖσι bright both literally
and in the derived sense of "glad" ; a predicate’. φαιδροῖσε is in fact clearly
'predicative' (or, if one prefers, proleptic): 'let these eyes of yours look
brightly shining on the king to make him welcome'. Of the emotional
meaning of φαιδρός, which is combined here with the physical, there is a good
example in Cho. 565 θυρωρῶν οὔτις dv φαιδρᾶι φρενὶ δέξαιτο. With ὄμμασι
δέξασθε cf. Xen. Anab. 7. 7. 46 σὺ δὲ... ἐμὲ ἐδέξω καὶ ὄμμασι καὶ φωνῆι kai
ξενίοις. Headlam refers to Bacchyl. xx (10). 15 f. ἵλεώι νιν ὁ Δαλογενὴς viós . . .
Δατοῦς δέκτο βλεφάρωι.
521. With κόσμωι here Abresch has compared Pind. P. 3. 83 τὰ μὲν ὧν οὐ
δύνανται νήπιοι κόσμωι φέρειν, and Schütz points out that κόσμωι is taken up
in 524 by καὶ γὰρ οὖν πρέπει.
522. It matters little whether ὑμῖν is a conjecture of Triclinius or whether
ἡμῖν is a mistake peculiar to F. ὑμῖν is in any case necessary, for one who has
himself taken part of the campaign cannot say ἥκει ἡμῖν φέρων. Only with
the words τοῖσδ᾽ ἅπασι does the Herald take any notice of the presence of the
1 Verrall's (4 που πάλαι), put in brackets like this, is a marvellous instance of his ‘con-
servative’ criticism.
2 The error of Wilamowitz, who says that τοῖσιν is found in F (confusion between F
and Tr), has been taken over by Murray and G. Thomson.
265
line 522 COMMENTARY

Chorus, to whom he then turns with ἀσπάσασθε in 524; up to this moment

he had only been addressing himself to the land and to its divine patrons
(cf. on the prologue, p. 25). There is an exact parallel in the arrangement of
the speech of Agamemnon at his entrance (cf. on 829 f.).
$s: ‘the Greek heard in φῶς an undertone of rescue from danger’ (Wila-
mowitz on E. Her. 563, cf. ibid. on 531). Particularly striking examples of
this usage are O 741 τῷ ἐν χερσὶ φόως, À 538 ai δὲ (sc. πύλαι) πετασθεῖσαι
τεῦξαν φάος, where the opponents of Zenodotus accused him of the stupidity
of having misunderstood the use of φάος = σωτηρία. In E. Med. 482 and
elsewhere we find φάος σωτήριον, cf. S. El. 1354 ὦ φίλτατον φῶς, ὦ μόνος
σωτὴρ δόμων Ἀγαμέμνονος.
523. ᾿Αγαμέμνων ἄναξ: after the extensive preparation this comes effectively
at the end; cf. on 1436.
525 ff. δικηφόρος here occurs for the first time, and was perhaps coined for
this passage, in keeping with the fundamental conception discussed on 41.
Elsewhere it is found only in 1577 and Cho. 120. For other 'Oresteia-words'
cf. on 386. The Διὸς μάκελλα caught the fancy of Sophocles: fr. 659 N. =
727 P. μακέλληι Ζηνὸς ἐξαναστραφῆι. As remarked by G. G. Salzmann, Observ.
in Aesch. Ag. (Berlin 1823) 11 (cf. also Headlam, C.R. xvi, 1902, 439) the
imagery expressed in μάκελλα extends not only to κατασκάψαντα but also to
τῆι κατείργασται πέδον and the whole of 1. 528: ‘he has levelled Troy with the
mattock of Zeus Justicer, whereby the soil hath been broken and the seed of
all the land destroyed’ (Headlam, but the tense of ἐξαπόλλυται needs to be
rather more precisely realized: ‘and thus the seed of the whole land is being
destroyed’). Into this consistent series of ideas extending from κατασκάψαντα
to χθονός, and all alike borrowed from the sphere of agriculture, 1. 527 βωμοὶ
κτλ. intrudes like a foreign substance,’ as was recognized by Salzmann and
others. The line would therefore be open to grave objections, even if it did
not recur almost word for word (only with δαιμόνων θ᾽ instead of kai θεῶν) in
Pers. 811 (not 511 as Murray states, taking over Wilamowitz’s misprint). But
even if we disregard the context, the line is shown by its content not to be
genuine. To the poet and his contemporaries? the destruction of holy places
by the enemy seemed an unparalleled atrocity. ‘Is it conceivable’, Headlam
rightly asks, ‘that Aeschylus . . . could have put this statement as a proud
boast in the mouth of a religious herald?' The objectionable nature of the
thought was rightly taken into account by Salzmann: he referred to E. Tro.
15 f. where the deity hostile to the Greeks says ἔρημα δ᾽ ἄλση kai θεῶν ἀνάκτορα
φόνωι karappet. In our passage the line was presumably written in the margin
‘as a parallel’ by a reader, quoting from memory,? ‘to illustrate the devasta-
! H. D. F. Kitto, Journ. Hell. Stud. lx, 1940, 111, making a fresh defence of the line,
which for him 'tragically is one of the most effective ones in the play’, does not discuss at all
this matter of the breaking up of a consistent series of images. Hermann on the contrary,
as can be plainly seen from his note, was really convinced by Salzmann that 527 disturbs the
context : his only defence consists of the characteristic argument that cutting out the line
‘non eam habet veri similitudinem, ut sine librorum fide delendus esse versus videatur".
2 Not only to them however ; such feelings are widespread. Cf., e.g., Livy 1. 29. 6 Egressis
urbe Albanis Romanus passim publica privataque omnia tecta adaequat solo . . . templis tamen
deum—ita enim ediclum ab rege fuerat —temperatum est.
3 Knoche, Rhein. Mus. Ixxxv, 1936, 32, it is true, quotes the line as one of those which
*have undergone an alteration, in order to make them fit their new context in point of
grammar, or the new situation in point of content’, but one is bound to ask, with what
266
 COMMENTARY line 532

tion of the land’ (Headlam). Probably the point at which the reminiscence
suggested itself was in 528 καὶ σπέρμα πάσης ἐξαπόλλυται χθονός : this might
easily have recalled the memory of Pers. 812 πρόρριζα φύρδην ἐξανέστραπται
βάθρων (note especially the second halves of the lines). For such marginal
reminiscences! Wilamowitz, Hermes, lxii, 1927, 285, compares Et. 105 (where,
incidentally, the rejection of 105 should be ascribed to Prien, and that of
104-5 to Schütz) ; cf. infra on 1290. Probably there would not have been so
many who refused to recognize this particularly obvious interpolation, if the
defenders of the line had not been only too glad to find here an extra touch
in the delineation of the godless villain Agamemnon in addition to the con-
clusions they felt bound to draw from the wrong interpretation of 811 and
above all from their misunderstanding of the carpet scene. They do not seem
to have asked themselves whether it would really be in the manner of
Aeschylus to give a mere hint; in the passing remark of 527, at a circumstance
which is, in their view, so essential in showing the guilt of the central figure.
Should we not expect that the point would be taken up again somewhere else,
either in a choral ode or in one of Clytemnestra’s accusations after the
murder ??
530. πρέσβυς: cf. on 184. The fact that εὐδαίμων ἀνήρ belongs predicatively
to ἥκει has long been recognized (by Paley, Karsten, and others) ; there is a
parallel in 1337 θεοτίμητος δ᾽ οἰκάδ᾽ ἱκάνει.
532. τῶν viv: passed over unnoticed by nearly all commentators? Verrall
did notice it, and says: “The ominous effect of these lines is aided by their
ambiguity. The intention is that Agamemnon, having more than avenged
his honour upon Troy, has now no rival in the world.” The innuendo thus
assumed would perhaps be appropriate in the mouth of the Chorus, but it
conflicts entirely with the attitude of the Herald, who is the one and only
character in the play who gives himself up quite simply and without any
gloomy forebodings to the feeling of joy in the present good fortune. A. S. F.
Gow, C.Q. viii, 1914, 3, has justly refused to disregard τῶν viv, yet his sup-
position that it is corrupt shows that even he has failed to see in this passage
justification. Would δαιμόνων θ᾽ have been more unsuitable in Ag. 527 than καὶ θεῶν»
Knoche’s list calls for a good deal of further criticism, e.g. with regard to the passages from
E. Phoen. Also A. Sept. 549 is certainly genuine, and certainly forms the end of the speech
(Wilamowitz has treated the passage quite wrongly ; there should be no transposition) : all
the reports of the Scout, with the significant exception of that which concerns the wise seer
(568-96), finish with a threat against Thebes on the part of the enemy. 426, on the other
hand, as Lachmann saw, is an interpolation, a clumsy watering down of 425 ὁ κόμπος δ᾽ οὐ
κατ᾽ ἄνθρωπον φρονεῖ, after which the κόμπος itself, i.e. 427 ff. θεοῦ re yàp θέλοντος «rÀ., must
follow directly. The interpolator presumably wanted to have an introductory verb of
speaking before the oratio obliqua with its parenthetic φησιν, cf. infra on 1600.
1 So I maintain, undeterred by Jachmann, Nachr. Gótt. Ges., Phil.-hist. Kl., Fachgruppe 1,
1936, 125, 137. See what happened at A. Pers. 253, and for the problem in general cf. my
remarks in Eranos, xliv, 1946, 87.
2 [n the note on 344 it was observed that no reference to Agamemnon can be found in
lines 338-44.
3 Since the writing of the following observations, which I am leaving unaltered, G.
Thomson's edition has been published, which shows that Headlam collected some parallels
and interpreted the phrase correctly. I had not noticed Simonides fr. 21 D. and E. Med. 948.
A. Y. Campbell at the end of his translation quotes the second line of the Spartan inscrip-
tion composed in praise of Damonon in the sth century (Solmsen-Fraenkel, Inscr. Graec. ad
inlust. dial. sel., no. 22, p. 35) vırdhas ravra har’ οὐδὲς mémoxa TOV νῦν.

267
line 532 COMMENTARY

a characteristic Greek outlook. There is a particularly instructive example

in Pind. Ol. 1. 103 ff. πέποιθα δὲ ξένον μή τιν᾽ ἀμφότερα καλῶν τε ἴδριν aude καὶ
δύναμιν κυριώτερον τῶν γε νῦν κλυταῖσι δαιδαλώσεμεν ὕμνων πτυχαῖς (cf. Bacchyl.
5. 3 ff., also addressed to Hiero, yodont . . . τῶν γε νῦν αἴ τις ἐπιχθονίων, ὀρθῶς),
for obviously Pindar is far from wishing to diminish the praises of Hiero.
But like the Argive herald he is evidently prevented, by an αἰδώς natural to a
Greek, from assigning a superlative attribute to a human being in such a way
as to preclude the possibility of its ever being surpassed. Cf. Akusilaos
fr. 40 a (Diels-Kranz, Vorsokr. i, sth ed., 60) about Kaineus ἄνδρα ἄτρωτον,
ἰσχὺν ἔχοντα μεγίστην τῶν ἀνθρώπων τῶν τότε, in the epitaph quoted in Thuc.
6. 59. 3 on Archedike, the daughter of Hippias, ([Simon.] 85 D.) ἀνδρὸς
ἀριστεύσαντος ev 'EAadı τῶν ἐφ᾽ ἑαυτοῦ ‘Immiov, and further where a positive
superlative is intended, Thuc. 7. 86. 5 (about Nicias) ἥκιστα δὴ ἄξιος ὧν
τῶν γε ἐπ᾽ ἐμοῦ ᾿Ἑλλήνων ἐς τοῦτο δυστυχίας ἀφικέσθαι διὰ τὴν πᾶσαν ἐς
ἀρετὴν νενομισμένην ἐπιτήδευσιν ; so also Pl. Charm. 157 d ἴσθι ὅτι πᾶσι (so
Wilamowitz, Comment. gramm. iv. 27, cf. his Platon, ii. 364, for the MS reading
πλείστων Or πάνυ) δοκεῖ πολὺ σωφρονέστατος εἶναι τῶν νυνί, Phaedr. 228 a
δεινότατος ὧν τῶν νῦν γράφειν, Euthyd. 274 € ὑμεῖς dpa . . . τῶν νῦν ἀνθρώπων
κάλλιστ᾽ ἂν προτρέψαιτε εἰς φιλοσοφίαν κτλ. (cf. 305 a). Cf. too (where the
content implies a superlative) Pericles on Athens in Thuc. 2. 41. 3 μόνη γὰρ
τῶν νῦν ἀκοῆς κρείσσων ἐς πεῖραν ἔρχεται. So in Pl. Prot. 309 ἃ no ironical
tone is to be inferred (which would in itself be possible). The same kind of
qualification of a superlative is found as early as Homer: 1 558 f. “Idea θ᾽ ὃς
κάρτιστος ἐπιχθονίων γένετ᾽ ἀνδρῶν τῶν τότε, and a Verrall of earlier days was
hardly justified in detecting here a peculiarly subtle point: Schol. BT
ἐπαινετέον τὸ ἀκριβὲς περὶ τὴν ἑκάστου ἠθοποιίαν. ὁ μὲν Φοῖνιξ κάρτιστον εἰπὼν
τὸν "Ἴδαν, ἵνα μὴ δόξηι καὶ τοῦ Ἀχιλλέως αὐτὸν προτετιμηκέναι, ἐπήνεγκε ‘rev
tore, τῶι δὲ Νέστορι ἀθωπεύτως εἴρηται (A 262) ‘où γάρ πω τοίους idov’. Cf.
also (in a passage influenced by Homer) Hdt. 9. 72. 1 Καλλικράτης. .. ἐλθὼν
ἀνὴρ κάλλιστος ἐς τὸ στρατόπεδον τῶν τότε Ελλήνων. A Homeric equivalent of
the formula τῶν γε νῦν is found in ἡ 67 f. καί μιν ἔτιο᾽ ὡς οὔ τις ἐπὶ χθονὶ τίεται
ἄλλη (this performs the same function as a comparative or superlative),
ὅσσαι νῦν γε γυναῖκες ὑπ᾽ ἀνδράσιν οἶκον ἔχουσιν. Also ı 263 f. λαοὲ δ᾽ Arpeidew
Ἀγαμέμνονος εὔχομεθ᾽ εἶναι, τοῦ δὴ νῦν γε μέγιστον ὑπουράνιον κλέος ἐστί
(Aeschylus may possibly have had this passage in mind). Different in syntax
but parallel in sense is the well-known hexameter on the Dipylon vase, IG
i.2 919, hós νῦν ὀρχεστὸν mavröv ἀταλότατα παίζει.
In order to understand the feeling which underlies the ‘formula’ τῶν νῦν
we may perhaps quote the οἷοι νῦν βροτοί εἰσι used in the Iliad to make a
differentiation in merit not only between the generation of the men who
fought at Troy (e.g. E 304) and the poet's own contemporaries, but also between
the heroes of Nestor’s youthful days and those at Troy. The application of
a standard of this kind is clearly expressed in 0 216 ff., where Odysseus
scrupulously limits his claim to be a first-class archer in two respects:
πρῶτός κ᾽ ἄνδρα βάλοιμι ὀϊστεύσας ἐν ὁμίλωι ἀνδρῶν δυσμενέων... οἷος δή με
Φιλοκτήτης ἀπεκαίνυτο τόξωι δήμωι ἔνι Tpdwy . . . τῶν δ᾽ ἄλλων ἐμέ φημι πολὺ
προφερέστερον εἶναι, ὅσσοι νῦν βροτοί εἰσιν ἐπὶ χθονὶ σῖτον ἔδοντες. ἀνδράσι δὲ
προτέροισιν ἐρίζεμεν οὐκ ἐθελήσω, οὔθ᾽ ᾿Ηρακλῆϊ οὔτ᾽ Εὐρύτωι Οἰχαλιῆϊ κτλ.
Pindar in Pyth. 6. 40 ff. contrasts his own generation with that of the Trojan
268
 COMMENTARY line 533

warriors (τῶν πάλαι) with the words τῶν νῦν (with a superlative following).
But leaving aside the comparison of different generations, what seems to be
expressed in all these passages is a definite shrinking from the use of the
unrestricted superlative of praise :! the measure of human modesty is pre-
served by limiting oneself to what can be asserted from one’s personal
knowledge. This is the meaning of the expression at least in the passages
quoted from Plato, as can be seen from its fuller form in the closing sentence
of the Phaedo, where these words are used of Socrates: ἀνδρός, ὡς ἡμεῖς
φαῖμεν ἄν, τῶν τότε ὧν ἐπειράθημεν ἀρίστου καὶ ἄλλως φρονιμωτάτου καὶ δικαιο-
τάτου. However, the addition of τῶν νῦν to a laudatory superlative seems
soon to have become a mere matter of form: in A. Pers. 184 the two super-
human apparitions, the queens of both the lands or continents, are described
thus: μεγέθει τε τῶν viv ἐκπρεπεστάτα πολὺ κάλλει 7” ἀμώμω κτλ. The formal
nature of the phrase can be seen too in Hdt. 8. 8. 1 Σκυλλίης Σκιωναῖος, δύτης
τῶν τότε ἀνθρώπων ἄριστος, Ar. Wasps 954 ἄριστός ἐστι τῶν νυνὶ κυνῶν, Phry-
nichus Com. fr. 3. ı (1. 370 K.) ἔστιν δ᾽ αὐτούς γε φυλάττεσθαι τῶν νῦν χαλεπώ-
τατον ἔργον.
Πάρις γὰρ οὔτε ao. π΄: Schol. ἀπὸ κοινοῦ τὸ οὔτε" ληπτέον καὶ eis τὸ Πάρις. For
the range of this familiar figure of speech cf. in addition to the grammarians
Wilamowitz, Pindaros, 469.
ouvreAns: it seems tempting, especially considering 537, to understand
συντελής here, in accordance with its usual technical meaning, of the paying
of a joint reckoning (cf., e.g., Tod, Greek Hist. Inscr. 52), as e.g. Paley, Weck-
lein, Verrall, and Pliiss do, but that would perhaps be too strong a ‘prolepsis’.
Moreover, Sept. 251 is a warning to be cautious, for there ξυντέλεια is used of
the θεῶν πανάγνρις (220) : the company of the gods is bound together through
common functions and interests, but also bound to the citizens under its
protection. Thus here the meaning seems to be: ‘bound together in a common
cause (and for a common destiny)’.
533. In addition to the idea of δράσαντι παθεῖν, which permeates the trilogy,
special emphasis is laid here on the principle of toa πρὸς ἴσα (Hdt. 1. 2. 1).
The significance of this principle for the legal thought of the fifth century is
stressed by Daube, 137 f. Cf. also Pers. 813 κακῶς δράσαντες οὐκ ἐλάσσονα
πάσχουσι, S. Ant. 927 f. ei δ᾽ οἵδ᾽ ἁμαρτάνουσι, μὴ πλείω κακὰ πάθοιεν 7) Kal
δρῶσιν ἐκδίκως ἐμέ. It may be partly due to the proverbial (cf. Plato. Laws 9.
872 d-e) δράσαντι παθεῖν that τὸ δρᾶμα is here put first, and thus is bound to
have πλέον for predicate. But also it was of the δρᾶσαι that Paris boasted ;
therefore we must not reverse the position, as Lewis Campbell does in his
translation: ‘that the punishment hath fallen short of the deed’.
1 This αἰδώς is Hellenic. The Great King feels no compunction about immortalizing him-
:self in an inscription (Hdt. 4. 91. 2) as ἀνὴρ ἄριστός τε καὶ κάλλιστος πάντων ἀνθρώπων,
Aapeios ὃ Ὑστάσπεος krÀ., and the flatterer Mardonius says to Xerxes (Hdt. 7. 9. 1) ὦ
«δέσποτα, οὐ μοῦνον εἷς τῶν γενομένων Περσέων ἄριστος, ἀλλὰ καὶ τῶν ἐσομένων... .. Yet even
the Hellenes, or rather their descendants, in course of time lost this fine feeling which was
based ultimately on religion: thus we read later in many laudatory inscriptions on the
‘pedestals of statues of athletes or musicians a formula which is the exact opposite of the
ancient Greek, μόνος καὶ πρῶτος τῶν an’ αἰῶνος κιθαρωιδῶν νικήσας and the like, and Nero
after his successes in the festival contests is extolled in the following terms (Cassius Dio 63.
20. 2) ὅτι Νέρων Καῖσαρ πρῶτος πάντων τῶν ἀπὸ τοῦ αἰῶνος 'Ρωμαίων ἐνίκησεν αὐτό. For these
late formulae I am indebted to L. Robert, Études épigraphiques et philologiques (1938), 109 f.,
where further instances are given.
269
line 533 COMMENTARY
The word δρᾶμα, which occurs here for the first time, was rarely used in this
sense (‘the doing, the deed’) after the meaning ‘Drama’ had become preval-.
ent, cf. H. Richards, C.R. xiv, 1900, 388 ff. ; B. Snell, Azschylos, 5.
The highly technical character of lines 534 ff. makes it probable that the
discrimination which we find here between the delicts of ἁρπαγή and κλοπή
is borrowed from regular legal terminology, and that therefore the distinction
between ‘robbery’ and ‘theft’—easily understandable but not self-evident—
as between crimes of greater and less seriousness was just as familiar in Attic
law in the time of Aeschylus as it was.in that of Plato (Laws 12. 941 Ὁ κλοπὴ
μὲν χρημάτων ἀνελεύθερον, ἁρπαγὴ δὲ ἀναίσχυντον), and of Aristotle (Eth. Nic.
5. 5, p. 1131°6 ff., where κλοπή is cited among the λαθραῖα, whereas ἁρπαγή
belongs to the βίαια, cf. H. D. P. Lee, C.Q. xxxi, 1937, 130 f.). This is note-
worthy with regard to a differentiation which comes comparatively late, as
far as we can see, in Roman law. ‘The act of expropriation, even when
perpetrated with the use of force and by the action of a mob, was in the
earlier period of Roman law treated and punished merely as theft’ (Mommsen,
Röm. Strafrecht, 660). ‘During the classical period [of Roman law] robbery
(rapina), i.e. the violent appropriation of the property of others, was re-
garded as a qualified case of theft. From the end of the Republic [i.e. from
the praetorship of M. Licinius Lucullus 76 8.c., cf. Cic. Tull. 8] the Praetor’s
Edict [cf. Lenel, Edictum, 3rd ed., 391 ff.] provided a special actio vs bonorum
raptorum for it’ (Jórs-Kunkel, Röm. Privatrecht, 2nd ed., 256). However,
the use Aeschylus makes of the term has no technical connotation. The
meaning can hardly be that the κλοπαὲ γυναικός (402) had exhibited in this
case the special features of a ἁρπαγή, for Aeschylus’ Helen, like the Helen of
Homer (cf., e.g., I' 173 f. ὁππότε δεῦρο vidi ods ἑπόμην, θάλαμον γνωτούς τε
λιποῦσα krA., and in general see Leaf on B 356), had gone off with Paris (407)
without any compulsion and without the slightest reluctance. So what
matters most here seems the sound and sense conveyed by the word ἁρπαγή:
ἅρπαξ δὲ κακή, θανάτοιο δότειρα. Herodotus, as a matter of course, includes
the abduction of Helen in his enumeration of ἁρπαγαὶ γυναικῶν (1. 3. 1),'
whereas Gorgias (B 11 Diels, 6, 20, cf. 7) suggests as merely one of several
alternative possibilities that Helen did what she did βίαι ἁρπασθεῖσα.
535. τοῦ ῥυσίου ἥμαρτε: the context leaves no doubt that these words mean
‘he lost his booty’. Schütz comments rightly: ‘pvocov h.l. nec redemptionis
pretium nec pignus proprie dictum, sed praedam significat, intelligiturque
Helena quam una cum magna opum vi Paris abegerat.’ This view was
attacked by Hermann; his learned exposition provides all kinds of valuable
information about other uses of ῥύσια, but his interpretation of the Agamem-
non passage, ‘Helena a Paride rapta, a Graecis vindicata, ῥύσιον vocatur',

! Daube, 110, shows that in Herodotus, and presumably elsewhere too, “ἁρπάζειν can
quite generally mean the abduction of women, either with or without their consent'. I
agree with Daube in being unable to follow those commentators who find in the expression
ἁρπαγῆς re xai κλοπῆς a reference to Paris’ having carried off any property at the same time
as he carried off Helen. This feature of the tale, it is true, comes out in Homer more than
once, but nothing points to it in the Oresteia. Further, which of the two words should
appropriately denote the theft of property? ἁρπαγή! In that case the less important thing
would be given first place : moreover ἁρπάζειν is the usual term for the abduction of women
and therefore of Helen. Or κλοπή in this passage, where the words κλοπαῖσι γυναικός still
echo in our ears from the preceding choral ode?
270
 COMMENTARY line 535
based as it is upon the definition ‘pvo.ov . . . proprie significat rem ademptam,
ee

quae vi vindicanda est atque ita salva praestanda', obscures the sense by
bringing in the vindicatio. Attempts are often made to effect a compromise
in Ag. 535 with the usual meaning of ῥύσια, e.g. L-S: 'stolen property taken
back as compensation for the theft' (in earlier editions it was given correctly,
following Passow, as ‘booty, plunder, prey’, though with a certain doubt and
with reference to Hermann), Schuursma, De . . . abusione ap. Aesch. 116:
‘plerumque notat pignus, ut Aesch. Ag. 535’. Such explanations are to me
quite unintelligible, for how can Helen be designated as ‘compensation’ or as
‘pignus’, when she was certainly not seized by way of securing some other
legal claim, but was herself the object of abduction effected without a shadow
of legality and in violation of the laws of hospitality? Here commonsense is
on the side of Sidgwick, who says: 'the word is used a little [I should say
**very"'] freely, as she was the thing stolen.’ Elsewhere ῥύσια or ῥύσιον always
denotes a thing which is seized on the ground of a legal claim (actual or
alleged), whether by way of securing this claim (as ‘pledge’ or the like), or
with a view to compensation. In the only passage in Homer where ῥύσια
occurs, A 674, it is quite clear (and rightly noted in the A-Scholia) that when
Nestor takes possession of the cattle of the people of Elis, ῥύσι᾽ ἐλαυνόμενος,
he is doing so on the ground of a legal claim: Augeias had stolen a valuable
four-horse-chariot team from Nestor's father Neleus (the statement of this
fact is delayed until 698 ff.). Similarly it is obvious in A. Suppl. 412, 728, S.
Oed. C. 858 that something is seized as ῥύσια or ῥύσιον in order to obtain
security for an alleged legal claim. In S. Phil. 959 ῥύσιον means 'compensa-
tion’. Equally clear is the relation to an actual legal obligation in the law of
Argos (sth century), Dittenberger, Syl. 56 (Tod, Greek Hist. Inscr. 33), 40 f.,
αἱ δὲ μὲ δοῖεν ξένια, βδλὰ ἐπαγέτω ῥύτιον δέκα orarépov. Accordingly 'ῥυσιάζειν,
to seize as ἃ ῥύσιον, is predominantly used if one takes forcible possession
of a man or a thing on the ground of a legal claim’ (Wilamowitz on E. Ion
523). The use of the verb in this sense in E. Jon 523, 1406 (the conjecture in
E. Heraclidae 163 I pass over) agrees with its use in inscriptions’ and in the
historians of the Hellenistic period, as has been pointed out in detail particu-
larly by Ad. Wilhelm, Oesterr. Jahresh. xiv, 1911, 197 fl. ;? “fvoraleıw means
taking away the person or the property of another as security for one’s own
legal claim’ (Wilhelm, 198). In a different and non-technical sense ῥυσιάζειν
is used by Aeschylus for ‘removal, snatching away’ in Suppl. 424 and fr. 258
(Phineus), where it is applied to the Harpies: καὶ ψευδόδειπνα πολλὰ μαργώσης
γνάθου ἐρρυσίαζον κτλ. (for the form of the text and its interpretation cf.
Wilamowitz, Aesch. ed. mai., p. 177). On A. Suppl. 610 ἀρρυσιάστους cf.
Wilamowitz, Hermes, xxii, 1887, 257 (= Interpr. 12). It is perhaps not merely
accidental that Aeschylus uses ῥυσιάζειν in the same non-technical way as he
uses τὸ ῥύσιον in Ag. 535. Did the word ῥύσια, ῥύσιον in his time have a more
general meaning, ‘the thing stolen’, besides the special meaning which was
unquestionably current at that period? That is possible, but hardly probable
in view of the agreement between its use in Homer and that of Tragedy
1 The occurrence of the verb in the decree about the Locrian maidens, Schwyzer,
Exempla, no. 366, is much older than the examples from inscriptions to be found in L-S
s.v. βυσιάζω.
2 Wilhelm’s results are summarized in Busolt-Swoboda, Griech. Staatskunde, 1241 f.
271
line 535 COMMENTARY

generally (including A. Suppl.), and of the Argive inscription of the sth

century. It is a natural assumption that the more general meaning ‘the thing
seized’ is the original one rather than the more special meaning ‘the thing
seized by way of securing a legal claim’. We have two ancient etymologies
of the word, one deriving it from ἐρύομαι, ῥύομαι = “protect, guard’, so
Hesychius s.v. ῥύσιον (after giving the other explanation)- ἢ τὸ ἕνεκα ἐνεχύρου
κατεχόμενον, παρὰ τὸ ῥύεσθαι τὸν κατεχόμενον. This etymology is accepted by
L-S. The other derives ῥύσιον from ἐρύω, ‘to draw’ (cf. ῥυτήρ ‘strap’):
Hesychius loc. cit. eAxvoröv; cf. Schol. BT on Hom. A 674 ῥύσια' τὰ ἀντί
τινων ἑλκόμενα, οἱονεὶ ἅπερ ἀντὶ τῶν ἡρπασμένων ἁρπάζομεν. This etymology is
adopted e.g. by Jebb on S. Phil. 959, and similarly by W. Schulze, Berl.
Sitzgsb. 1918, 483 n. 2 (= Kl. Schr. 162 n. 4), who says: ‘the object upon which
the ἕλκειν is or may be practised is called ῥύσιον or ῥύσια᾽ ; E. Risch also,
Wortbildung der homer. Sprache (1937) 114 ff., makes ῥύσιον come from ἐρύω
‘to draw’. For the meaning, this etymology seems preferable. But even if we
assume that it is correct, it remains more than doubtful whether the general
meaning, ‘the thing dragged away, stolen’ and the like, survived right up to
the time of Aeschylus alongside the specialized meaning. It is more likely
that Aeschylus boldly gave the word a fresh meaning (cf., e.g., on 149). The
context here left no doubt about the sense, and it was possible too that the
audience was conscious of the word’s affinity with ép/ew. The half-technical
tone of the word, recalling the phrasing of legal claims, must have appealed
to the poet particularly in this sentence which abounds in juridical terms.
So he did not trouble to confine himself strictly to the exact technical
meaning of ῥύσιον (cf. the remarks p. 270 about his use of ἁρπαγή side by
side with κλοπή).
535 f. καὶ πανώλεθρον .. . δόμον. The simple verb ἔθρισεν (schol. ἐθέρισεν,
ἔκειρε) is found only here, but arredpıve occurs in Archilochus and Euripides,
παρέθρισε Ap. Rhod., διέθρισε Quint. Smyrn. G. Boesch, De Apollonii
Rhodii elocutione (Diss. Berlin, 1908) 64, observes that only aorist forms are
found. The expression καὶ πανώλεθρον κτλ. is derived from formulas of
cursing : ἐξώλης εἴην αὐτὸς καὶ τὸ γένος and the like (examples e.g. in Ditten-
berger’s Index, Syll. iv. 333 s.v. ἐξώλης, cf. the oaths of the Amphictyons in
Aeschines 3. 111). In the oaths of Dreros (Schwyzer, Exempla No. 193) the
expression is used 1. 81 καὶ κακίστωι ὀλέθρωι ἐξόλλυσθαι αὐτός τε kai χρήϊα
τἀμά. The word αὐτόχθονον, of which this is the only instance (the form is
defended by Lobeck, Paralip. 202), was rightly understood by Scholefield,
Klausen, and Hermann: “αὐτόχθων quum significet eum qui ipsa in terra
natus est, alia forma opus esse videbat poeta, si dicendus esset qui ipsa cum
terra sua aliquid pateretur’ (Hermann). Similarly Wilamowitz, Glaube d.
Hell. i. 2x1 n. 2: ‘ αὐτόχθονος is a bold spontaneous coinage for αὐτῆι τῆι
xBovi ’. Sophocles (Ant. 714) uses the Aeschylean word αὐτόπρεμνος (glossed
in Hesychius by αὐτόρριζος) in a similar context : ra δ᾽ avrıreivovra (scil. δένδρα)
αὐτόπρεμν᾽ ἀπόλλυται, where Jebb compares the Homeric αὐτῆισιν ῥίζηισι. To
the same range of ideas belong the imprecations in S. Aj. 1177 f. κακὸς.
κακῶς ἄθαπτος ἐκπέσοι χθονός, γένους ἅπαντος ῥίζαν ἐξημημένος, E. Hipp.
683 f. Ζεύς ce . . . πρόρριζον ἐκτρίψειεν, and (probably the closest of all to the
actual curse-formula) Ar. Frogs 587 f. πρόρριζος αὐτός, ἡ γυνή, τὰ παιδία,
κάκιστ᾽ ἀπολοίμην. For the formation of αὐτόχθονος cf. 137 αὐτότοκος. In the
272
 COMMENTARY line 537
apotropaic prayer in Sept. 71 f. μή μοι πόλιν γε πρυμνόθεν πανώλεθρον ἐκθαμνί-
ante δηιάλωτον where πανώλεθρον also recurs, the πρυμνόθεν corresponds to
αὐτόχθονον in our passage.
537. ra ἁμάρτια : only found in Pers. 676 (a very corrupt passage), here, and
in the comedian Plato. The explanation of the scholion τὸν μισθὸν τῆς
ἁμαρτίας is vindicated by Ahrens, 539, who, after rejecting other interpreta-
tions, says: ‘Blomfield has correctly observed that this [the scholiast’s ex-
planation] is supported by the ending -ıa: for this has a similar force in
θρεπτήρια, τροφεῖα, ζωάγρια [cf. also 0 332 μοιχάγρι᾽ ὀφέλλει, on which see
Latte RE xv. 2446. 41, and S. Trach. 542 otxovpta], εὐαγγέλια, etc., words which
are actually used in connexion with rivew, see the lexicons. The analogy with
ζωάγρια, εὐαγγέλια is particularly important, for there just as in dpdpria no
other suffix appears before the -ἰα.᾿ L. Schmidt (see Dindorf’s Lex. Aesch.
422) compares ἁμάρτια with ἀμπλάκιον (only in Pind. P. 11. 26). We might
also take notice of a suggestion put forward by Walter Petersen, Greek
Diminutives in τιον (Weimar 1910), 20 n. 1, in connexion with ἀδέκιον (it
occurs from Hat. 5. 89. 2 onwards; for further instances see Passow-Crónert
and L-S): ‘It was probably the use of ἀδίκιον in legal phrases . . . that gave
rise to a group of action nouns in -wov which were confined to legal use, mostly,
though not always, in certain stereotyped phrases with the genitive', etc.
For ἀδίκιον Petersen should also have referred to Dittenberger's note 15 on
the law of Oropos, Sylloge, 2nd ed., 589 (= 3rd ed., 1004). It is of course
doubtful whether adixıov is to be regarded as the prototype of the whole
group. For the old Attic word ὑποφόνια used in the laws of Solon (τὰ ἐπὶ
φόνωι διδόμενα χρήματα τοῖς οἰκείοις τοῦ φονευθέντος᾽ Harpocration s.v.) cf.
Latte, RE xvi. 284. 52 ff. In regard to the passage of the Ag. Wilamowitz,
Glaube d. Hell. ii. 120 n. 1, observes: ‘Aesch. Ag. 537 forms a word τὰ ἁμάρτια
on the lines of ῥύσια [this is quite uncertain] for the penalty for which the
guilty party is liable in reparation for the loss which his ἁμαρτία has caused
to the plaintiff. . . . Pers. 676 is not to be relied upon’.
διπλᾶ: here the difference between a wider and a more specialized meaning
of the word is made the basis of a significant jeu de mots. In the first place,
in accordance with the ordinary sense of ‘twofold’, διπλᾶ serves to sum up the
two facts which in the two preceding lines were expressed separately in a re-
and a xai-clause (the loss of Helen, and the utter destruction of Troy): ‘and
the sons of Priam have paid the price of their misdeeds twice over’. But
διπλοῦν also denotes (like the more ordinary διπλάσιον) the duplum, the
double value in regard to damages. In this connexion the word was used at
least as early as the time of Aeschylus, as appears from the Athenian law
about the Eleusinian mysteries (Dittenberger, Syll. 42 = IG 1.2 6, ‘ante a.
460’) 45 f. ra μὲν μΠακόσια hamAé&, τὰ δὲ hexdara διπλέι (‘de damnis resarciendis
hic dictum fuisse apparet’ [Kirchner]; so it is used exactly in the sense in
which we have it here). The Athenians were doubtless familiar with the law
of Solon whereby, in a δίκη κλοπῆς concerned with simple theft, the thief, in
the event of his restoring the thing stolen, was sentenced (Demosth. 24. 114 f.,
quoted by Spanheim in his note on A. Ag. 534) διπλάσιον ἀποτεῖσαι τὸ τιμηθέν,
1 T cannot accept the transposition in the text of Phot. Berol. p. 88. 7 which Wilamowitz
(on Ag. 537) assumes. Why should we doubt the occurrence of ἁμαρτάδας in Aeschylus when
Sophocles (fr. 999 P., from Phot. Berol. p. 88. 6) used duaprds?
4872-2 T 273
line 537 COMMENTARY

cf. G. Glotz, ‘Klopé’ in Daremberg-Saglio iii. 1. 829, J. H. Lipsius, D. att.

Recht und Rechtsverfahren 440. The same regulation is found outside Athens,
e.g. in the laws about the mysteries of Andania (Dittenberger, Syll. 736 =
Schwyzer, Exempla, 74) 75 ff. dv 8é rig . . . ἁλῶι εἴτε κεκλεβὼς εἴτε ἄλλο τι ἀδίκημα
πεποιηκώς. .. ö μὲν ἐλεύθερος, αν
ἃ κατακριθεῖ, ἀποτινέτ L6] διπλοῦν, ὁ δὲ δοῦλος

μαστιγούσθω καὶ ἀποτεισάτω διπλοῦν τὸ κλέμμα. After the preceding words

(534) ὀφλὼν... κλοπῆς δίκην, the expression διπλᾶ ἔτεισαν was bound to recall
to the audience the thought of the law about theft. So we have in addition
to the sense explained above a secondary one: ‘and the sons of Priam have
paid a double amount as penalty’. For an analogous double function of
διπλοῦς in Cho. 375 see below, note on 642.
538. ‘notanda est attractio, qua duae locutiones in unam confusae sunt,
κῆρυξ Ἀχαιῶν τῶν ἐν στρατῶι et κῆρυξ ἀπὸ στρατοῦ. V. ad Soph. El. 135 [137]’
(Hermann). Instances of this construction, which is more common in prose,
are given in Kühner-Gerth, i. 546; a typical case in Aeschylus is Cho. 507 τὸν
ἐκ βυθοῦ κλωστῆρα σώιζοντες λίνου. There is no reason to assume with Head-
lam, C.R. xvii, 1903, 290 ἢ. 2, that ἀπὸ στρατοῦ here has the same meaning as
ἀπὸ στρατείας elsewhere.
539. The difficulty of this line as it stands is most strikingly shown by the
non-existent form redvävaı, which Ahrens did not succeed in defending any
better than others after him. Enger's way of filling up the line by the
addition of ye after χαίρω would produce the correct form τεθνάναι: and this
has been approved by many. All the same Headlam's comment holds good:
‘yaipw ye . . . cannot be right, as this is the answer to χαίρεις, not to xaipe.’
This objection is not removed by Denniston, Particles, 131. He quotes only
two passages for ye 'answering a command or wish’, E. Alc. 420 and Rhesus
219. In Alc. 420 the re found in the better class of our MSS is probably to be
preferred :! but even if ye is taken from LP, the connexion with the preceding
words does not justify comparison with the passage here. In Res. 219 re is
the only reading of the MSS (probably correct, see footnote 1), but even
if ye were read instead of it, the relation of σωθήσομαι to the preceding wish
of the Chorus would offer only a distant analogy to the χαῖρε... χαίρω γε
here proposed by Enger. If ye were really the proper word for taking up an
imperative, we should expect to find at least a few cases of its being so used
in Aristophanes, which I have not succeeded in doing. However, quite apart
from the completion of the line by the missing syllable, Headlam is right in
saying that ἀντιλέγω with the infinitive cannot yield the sense required here.
Among all the passages quoted in the Thesaurus, there is no construction
that is comparable in any way. Headlam's own text (for other attempts of
his see J. Phil. xx, 1892, 299) χαίρω" τεθναίην δ᾽ οὐκέτ᾽ ἀντερῶ θεοῖς splits up
the sentence in a most improbable way. Other proposals are no better, and
so at present one can go no further than Verrall's despairing remark: 'this
line is hopeless.' The sense in general is fairly clear all the same, the more so
as 550 is also related to this line: ‘Yes, I do rejoice: and I would not? (or ‘no
! re was put in the text by Kirchhoff. I have not purposely looked for parallel expres-
sions in an analogous reply, only I happened by chance on S. Trach. 626 (in reply to
ἐξεπίστασαι in 624; so there is no wish expressed here) ἐπίσταμαί re καὶ φράσω σεσωμένα
and Oed. C. 113, where, in reply to the ofya of Antigone, Oedipus says: σιγήσομαΐ re xai
ov ne... κρύψον.
2 As regards the second element of οὐκέτ᾽ here, there is considerable probability in

274
 COMMENTARY line 542
longer’) object to being dead.’ With xaipw, as often elsewhere, there is an
insistence on the original sense of the formula of greeting χαῖρε, cf. on 251 ff.
τεθνάναι, ‘to be dead’, is in itself unobjectionable :! instead of the misquota-
tion in the app. crit. of Wilamowitz cf., e.g., Mimnermus fr. 2. 10 D. τεθνάναι
βέλτιον 7 βίοτος, Hdt. 7. 46. 3 τῶι οὐ παραστήσεται πολλάκις καὶ οὐκὶ ἅπαξ
τεθνάναι βούλεσθαι μᾶλλον ἢ ζώειν, but especially the instances collected by
I. Bekker, Hom. Blätter, ii. 9 of the wish often expressed in Homer τεθναίην,
redvain etc. Note also the examples from the language of laws? in Bekker
p. 10, where the death penalty is laid down with the expression τὸν ἀστὸν
τεθνάναι or the like, instead of what we should expect, namely ‘he must die’.
540. In F there is a question-mark at the end of the line, which has been
adopted by Petrus Victorius and most of the editors, whereas a full-stop is put
by, e.g., Weil, Dindorf (sth ed.), Kirchhoff, Verrall, Wilamowitz (though in his
translation he gave the sentence as a question), Mazon. A question seems
the natural thing. If 540 was a statement, one would expect a particle. 541
(γε) is obviously an answer. The treatment of the interrogative sentence
without interrogative particle in Kühner-Gerth, ii. 523, is very poor, and does
not take drama into account at all. From stichomythia in the Oresteia besides
Ag. 540 and perhaps 545 (see note) the following instances may be quoted:
Ag. 933 (see note), 1209, Cho. 222, 912, Eum. 206, 595, 896, 898.
542. ‘Mihi ἦτε a Triclinio profectum, vera autem scriptura esse videtur
codicis Florentini, ut hoc dicat, scite vos compotes esse huius suavis morbi
(Hermann). But this interpretation breaks down with ἄρ᾽ iore. For as far
as I can see ἄρα is not associated with imperatives. To check the evidence
collected in the Thesaurus, in the indexes to the tragic poets, and by Dennis-
ton, Particles, 32 ff., I have also compared the numerous passages given by
Todd, Index Aristophaneus, without, however, finding an instance of an
imperative with dpa. (As to the optative in Thesm. 887, this is in close con-
junction with the future indicative which follows, and that may account for
ἄρα.) On the other hand, if we accept the imperfect, we have the familiar
idiom. But it does not follow that we are bound to read ἦτε, of which Her-
mann rightly said that it looked like one of Triclinius' simplifying conjectures.
The true reading was recovered by Ahrens, 539 f. He restored the form
ἦστε, Which in ΑΓ. Peace 821 (where we have several good MSS) is attested
by the 'consensus librorum’ and in Eccl. 1086 is preserved in such MSS as
there are (the slight misspelling in R is of no consequence), cf. also Plat.
Symp. 176 ἃ παρῆστε (all MSS); the three passages are quoted by Kühner-
Blass, ii. 222. The only other passage in Aeschylus where we find the second
person plural is Suppl. 288 : there, it is true, the form is ἦτε, but in the facsimile
of the Mediceus it looks as if originally something else had been written
Hermann's assumption that Triclinius had before him nothing but the οὐκ ἀντερῶ which
we read in F, and filled it up by a conjecture of his own. The quotation of the line in the
pre-Triclinian scholion on 550 may easily have been adapted to the text of the MS (Tr) on
the margin of which it is written (for such adaptations cf. W. Jaeger, Hermes, lii, 1917, 483,
and Pasquali, Gnomon, v, 1929, 498 n. 1).
1 E. Petersen is quite wrong, Rhein. Mus. lxvi, 1911, 20: 1. 539 the thought does not
admit of the perfect, for it requires “die” here, not "be dead”.’
2 To Bekker’s list add, e.g., the Athenian decree of about the same period as the Oresteia,
IG i.? 10 = Tod, Greek Hist. Inscr., no. 29, ll. 30, 33 τεθνάτω. Ar. Frogs 1012 also belongs to
this category.
275
line 542 COMMENTARY
instead of nr, possibly nor ; the final decision of this point depends on a fresh
inspection of the MS. However that may be, it seems pretty clear that in
Ag. 542 ἴστε is but a slight corruption of the original ἦστε. The evidence in the
MSS of Aristophanes and Plato proves sufficiently that the earlier Attic
form was Fore,’ which was ousted later on by re as it had been ousted in
Homer (cf. Monro, Homeric Grammar, znd ed. ı3; Gustav Meyer, Griech.
Gramm., 3rd ed. 568). [Cf. the Addenda.)
543. Conington says rightly in his note on A. Cho. 171 (in stichomythia as
here) ‘With mds . . . μάθω comp. Ag. 543 πῶς δὴ διδαχθεὶς τοῦδε δεσπόσω λόγου;
(wrongly broken up by some of the later editors [following Schütz]) where
πῶς goes with διδαχθείς᾽. It seems to me to go with 58. δεσπόσω : ‘I should
like to have an explanation and so to master thy saying’, cf., e.g., Eum. 678
πρὸς ὑμῶν πῶς τιθεῖσ᾽ ἄμομφος ὦ, where πῶς goes with both τιθεῖσα and ἄμ. ὦ.
From a stylistic point of view the sentence is far better unbroken; there is
further support for this in the fact that questions of the πῶς δεσπόσω type
(aorist subjunctive) are remarkably numerous in the Oresteia, see below on 785.
As to the function of this question in its context in the dialogue, Sidgwick’s
remark on Cho. 171 is applicable here too: ‘ “How shall I hear it?” is only a
more gentle way of suggesting "let me hear it”. For a modern translator
it is hardly possible to render such ‘delicacies of Greek’ (Sidgwick) with
complete precision.
τοῦδε δεσπόσω λόγου: a bold and unusual expression. S. Phil. 1048 viv δ᾽
ἑνὸς κρατῶ λόγου (Blaydes) has an entirely different sense.
545 has been taken as a question by Heath, probably correctly. In sticho-
mythia it is customary to use the form of a question to carry on the dialogue.
In this case the question may seem justified up to a point by the cryptic
brevity of the expression τῶν ἀντερώντων ἑἱμέρωι, which is then taken up in the
slightly varied form of ποθεῖν ποθοῦντα. It is not, of course, impossible to
take 545 as a statement, but in that case there would be less smoothness in
the connexion with the following line.
The question as to the subject of ποθεῖν has been differently answered by
various commentators. Humboidt, Hartung, Conington, Nagelsbach, Ver-
rall, Pliiss, and others take τήνδε γῆν to be the subject, while Stanley, Schiitz,
Wilamowitz, Headlam, and others suppose ποθεῖν to be said of the λέγων, 1.6.
of the Chorus, and therefore regard τήνδε γῆν as the object of ποθοῦντα (te
exercitum patriam hanc desiderantem vicissim desiderasse, Schütz). The point
cannot be decided with absolute certainty: all the same the balance which is
achieved by taking ποθεῖν τήνδε γῆν ποθοῦντα στρατόν together seems far more
suitable for the terse style of this stichomythia than the construction which
requires ποθοῦντα τήνδε γῆν to be taken as attribute to στρατόν, thus making
this one part of the sentence carry practically the whole weight by itself.
For the ‘personification’ of τήνδε γῆν cf., 6. 8. S. Oed. R. 47 ὡς σὲ νῦν μὲν ἥδε γῆ
σωτῆρα κλήιζει.

540. ἀμαυρᾶς ἐκ φρενός: ‘aus gramverdunkelter Seele’ (Wilamowitz on Cho.

157). Here the sense is clear, while the similar expression in Cho. 157 is not
unambiguous: itistaken by, e.g., Conington and Blass to refer to the Chorus,
but in one explanation in the scholion and by, e.g., Wilamowitz and Headlam
to refer to the dead king. For ἀμαυρός cf. on 466.
! Therefore the brief statement in E. Schwyzer, Griech. Gramm. 1.677, ‘att. jore is justified.
276
 COMMENTARY line 550

It does not follow from the interpretation of 545 recommended above that
we need abandon Scaliger’s supplement φρενός (u^? which has been accepted
by most critics. For it is quite natural that the feeling of ‘this land’ should
be expressed in the attitude of its inhabitants, in this case the Elders, and no
one would object to Conington's rendering: ‘Her. What, that the land we
longed for longed for us? CHo. Aye, so that oft Igroaned from gloom of soul.’
<y’> was suggested by Heath and adopted by, e.g., Headlam. The position
of the particle is not in favour of this conjecture : cf. the survey in Denniston,
Particles, 146 ff. Against Boissonade's (o'» or Plüss's {σφ᾽ (his explanation
is: ‘ode: object ; subject γῆν᾽, i.e. he refers ode to στρατόν) it may be objected
that in the only other passage where avaorevew occurs in Aeschylus, Ag. 1286,
it is used absolutely.
547. orparó is very ambiguous in the mouth of the Herald. It is not worth
while to enumerate the suggestions which have been made. The splitting of
the line into two sentences (Verrall) should be avoided, as also the conjecture
στυγοστράτωι Which is as unsatisfactory for the sense of the passage as it is
for the style of this dialogue. Schütz's supposition is an attractive one, that
orparó arose from an erroneous gloss (στρατοῦ) on στύγος, and that the
genuine word has fallen out. I have adopted Schütz's {θυμῶι στύγος, but
only exempli gratia. No really obvious way of filling up the text has so far
presented itself. Moreover, I do not feel quite sure that ἐπῆν is sound, for it
looks odd in conjunction with πόθεν. |
548. These veiled words convey the same gloom as the corresponding utterance
ofthe Watchman in 36.
549. καὶ müs; standing alone as in 1310, Cho. 776! (Ag. 549 was rightly punctu-
ated by Stanley, 1310 by Pauw, Cho. 776 as early as the Mediceus). καὶ πῶς;
has the full interrogative force here, as in Ar. Knights 128 (we may concede
‘the shade of objection or incredulity’ there noted by Neil, though the pas-
sage is clearly different from those in which καὶ πῶς serves practically as a
negative), Clouds 1434, while on the other hand in Ag. 1310, Cho. 776, Ar.
Clouds 717, Pl. Theaet. 163 ἃ, and in many other cases kai πῶς approximates
to an expression of a strong doubt or denial, cf. Porson on E. Phoen. 1373
(1354) and below on 1507.
Here τυράννων is written in F, at Ar. Ach. 472 the Ravennas has τυράννους
for κοιράνους, at Prom. 958 VF have τυραννοῦντ᾽ for κοιρανοῦντ᾽.
550. The conjecture ws is obviously right (for its attribution to Auratus see
the extracts published by Blomfield, Museum Crit. ii, Cambridge 1826, 490).
τὸ σὸν δή referring to 539, similarly (Blomfield) Pl. Soph. 233 Ὁ τὸ σὸν δὴ
τοῦτο referring to Theaetetus’ statement in 232 d. δή has a firmly established
place in the formulas of quotation τοῦτο δὴ τὸ τοῦ λόγου, τὸ λεγόμενον δὴ τοῦτο,
and the like. Abundant illustration is to be found in Headlam on Herodas
2. 45, to which we may add Menander Phasma 42, fr. 402. 8 Kock, and now
from Aeschylus himself, in the fragment of the Myrmidons on the Florentine
papyrus, Pap. Soc. It. 1211, 1. 6 (cf. Hermes, Ixxi, 1936, 26, and D. L. Page,
Greek Lit. Papyri, i. 140), ©, τοῦτο δή, βροτοῖσιν ἰατρὸν πόνων, see also
P. Maas, C.R. liii, 1939, 59 (in passing he gave the correct punctuation and
1 It is probably due to an oversight that these examples are not recorded in Denniston,
Particles, 310, where it is stated that ‘the elliptical form καὶ πῶς; is common in Plato’ and
instances are also quoted from Euripides and Aristophanes.
277
line 550 COMMENTARY
interpretation of Pl. Gorg. 522 c). For the order (ὡς νῦν, then the apposition
τὸ σὸν δή, then the clause for which the apposition prepares the way) cf. E.
Cycl. 32 f. καὶ νῦν, τὰ προσταχθέντ᾽, ἀναγκαίως ἔχει σαίρειν κτλ.
χάρις: cf. on 421. The words of the coryphaeus xai θανεῖν πολλὴ χάρις
surpass in intensity what has been said by the Herald. The audience, which
is in the secret and can remember previous words of the Chorus, may possibly
gather from this strong profession of readiness for death an implication such
as ‘before some fresh evil befall as is bound to happen’. It is characteristic of
the Herald, whose vision goes no deeper than the surface of the situation, that
he fails to mark the sinister tone underlying the last words of the old man,
and only concerns himself with what fits in with his own satisfied frame of
mind. This is brought out with almost comic clearness in the brevity with
which he makes some sort of reply to the last words of his interlocutor, before
launching out under the full sail of the common man’s garrulity on the subject
of his own thoughts and the tale he is anxious to tell: εὖ γὰρ πέπρακται “Yes,
for things have gone well’.
551. ταῦτα cannot be right, as had been seen by Auratus and many others
after him (cf. Schütz, Ahrens, 540) ; Verrall’s note shows that he was unable
to justify ταῦτα δέ. The πάντα δ᾽ of C. G. Haupt (and Wilamowitz) is better,
but not really convincing.'
552. εὖ must be altered to ἂν not only because the particle is necessary here
(cf. on 1328)—there is no question here of Verrall’s imperative use of the
optative—but also because εὖ λέγειν in this context would be meaningless.
For the type of corruption cf. on 119 ἐρικύματα (M).
εὐπετῶς ἔχειν apparently does not occur elsewhere. εὐπετῶς here not
‘easily’ as Cho. 1047 and often elsewhere, but retaining the meaning we have
in εὖ πεσόντα 32 (so Stanley). For τὰ μὲν... ra de... καί cf. Denniston,
Particles, 305.
553 f. τίς δὲ πλὴν θεῶν κτλ. Cf. Pind. P. το. 21 θεὸς εἴη ἀπήμων κέαρ.
554. τὸν δι᾽ αἰῶνος χρόνον : ‘a good example of the meaning of the words;
χρόνος is time absolutely, αἰών relatively, determined by the subject to whom
the αἰών belongs: in the case of the Herald it would be his lifetime, in the
case of the gods it is eternity, cf. on E. Her. 669’ (Wilamowitz, Interpr.
170 n. 3). This is right in the main, but the idea of ‘eternity’ may be dis-
regarded here. It should be noted that αἰών by itself in Aeschylus does not
mean ‘tempus aeternum’ (so e.g. Dindorf, Lex. Aesch.). The meaning ‘life-
time’ (cf. on 105 f.) is quite unmistakable in the expression à αἰῶνος here, as
in Pers. 1008, Cho. 26, Eum. 563. Attributes such as Suppl. 582 8 αἰῶνος
μακροῦ, 574 9v aidvos . . . ἀπαύστου show clearly the limitations of the meaning
of the word by itself.
555. After the philosophizing of the opening, he now comes to the point.
The word, μόχθους, which he puts at the beginning by way of a headline,
indicates that he sums up his experiences on the campaign first with the
greatest possible generality : it was all μόχθοι for the ten years. Also δυσαυλίας
still applies to the whole time, referring to a detail but bringing out much the
1 In Prom. 275 too Wilamowitz reads πάντα with Herwerden instead of ταῦτα, but in that
passage the ταὐτά of the scholia which most of the more recent editors have adopted might
well be correct. In Ar. Wasps 798 Reiske made the highly probable correction πάνθ᾽ for the
unintelligible ταῦθ᾽,
278
 COMMENTARY lines 556 f.

worst aspect of the almost unending war: scarcely ever a decent night’s rest.
Aeschylus, who himself had served in the army, knows what that means,
and so the queen, speaking κατ᾽ ἄνδρα σώφρονα, emphasizes (336) that the
troops in the conquered city ὡς εὐδαίμονες ἀφύλακτον εὐὑδήσουσι πᾶσαν εὐφρόνην.
The lexicographers have called attention to the Attic idiom αὐλίζεσθαι —
κοιμᾶσθαι: Harpocration s.v. (= Antiphon Sophistes fr. 68 Diels-Kranz,
Vorsokr.), for further evidence see Adler on Suid. i. 414, πο. 4441. δυσαυλίαι
then refers to the nights on shipboard as well as to those at Troy; it is un-
justifiable to understand the word as meaning only their quarters on land,
as Schütz and others have done, and equally to confine it to the sea-voyage,
as Wilamowitz does. The word övoavAia, for which this passage was the only
pre-Hellenistic evidence, has now been restored with probability by Lobel
in another passage of Aeschylus, Pap. Oxy. 2162, fr. 2(a), i. 7.
556 f. Here the thought is narrowed down: the recollection of the discom-
forts of their sleeping accommodation first of all recalls to memory the suffer-
ings of the journey. That this is the line of thought we can see from 558 ra δ᾽
αὖτε yépoux, in spite of the obscurity of 556 f.
It is impossible to determine with certainty that παρήξεις is correct or what
it means. Schol. (ZyoÀ. ru. in Tr): τὸ δὲ παρήξεις ἀντὶ τοῦ παραδρομὰς ἐπὶ
τοῦ καταστρώματος τῶν νεῶν. This explanation with the help of the word
παραδρομαΐ, obviously used as a technical term, though not, apparently,
attested in this meaning elsewhere, looks much more like the remains of an
old learned explanation than an improvisation based only on the text. So
the scholion should not lightly be cast aside, as has often been done from the
eighteenth century onward. The interpretation it gives suits the sense of the
passage. As Causaubon saw, it suggests that πάρηξις here is what in Hellen-
istic authors is called πάροδος. The chief authorities for the use of πάροδος in
a ship are conveniently collected by Schweighaeuser in his note on Athen. 5.
203 f, i.e. Kallixeinos and Moschion (both quoted by Athenaeus), Plutarch,
Demetr. 43. 5, Pollux 1. 88. ‘There was a parodos or gangway on either
side of a Greek war-ship; and as [cf. Plutarch l.c.] combatants were posted
on the gangways as well as on the hurricane-deck, these gangways formed
part of the upper decking’ (Torr, Ancient Ships, 49 f.). Cf. Wilamowitz
(Griech. Lesebuch, note on p. 266. 17): ‘wdpodos is . . . a "Zugang" which
runs along the ship and on which the crew of marines (ἐπιβάται) moves.’
Possibly παραδρομή is a third, perhaps later, name for the same thing.
Whatever is meant by παρήξεις(}) in particular, the general sense must lie in
the direction suggested by the scholion. The interpretation given to it by
Schütz (raras ad terram appellendi occastones) and followed by Blomfield and
others is proved impossible by 558. About the derivation of πάρηξις, in view
of the fact that it does not occur elsewhere and that its meaning is not quite
clear, the most that can be done is to hazard a conjecture. The uncertainty
is increased by the fact that there seem to be no comparable derivatives of
ἥκω or its compounds (ἧξις Bekk. Anecd. 99. 4 is plainly an itacistic mistake
for its at E. Tro. 396). Supposing παρήξεις is the right reading here and is
derived from ἥκειν, it can hardly mean ‘accessus, adventus’ (Dindorf),
because we cannot connect this with any known use of παρήκω. It may seem
preferable to think of the meaning several times attested in Herodotus and
the Corpus Hippocraticum, ‘to stretch out along something’, cf. Erotian
279
lines 556 f. COMMENTARY
p. 71. 16 Nachm. παρήκουσι" mapareravraı. From this could easily be derived
πάρηξις in the sense of ‘that which runs along something’, cf. Chantraine,
La Formation des noms, 283: ‘the suffix (-t-) serves to form words of concrete
meaning D... ξύνεσις "junction", in « 515 πέτρη τε ξύνεσίς τε δύω ποταμῶν,
“un rocher et le confluent des deux fleuves" '. Such ἃ πάρηξις would be a
suitable description for the gangway along the sides of a ship. But the
doubts I have suggested remain, and unless new material comes to our help,
we must allow for the possibility that παρήξεις has displaced another word
from the text (Wackernagel, Glofta, xiv, 1925, 59, suffices to warn us against
considering Wecklein's παρίξεις, accepted by Verrall). But whatever Aeschylus
wrote here will have been either a nautical expression and familiar to the
Athenians or the rendering of a sailors’ term in the style of high poetry. To
this noun are attached the rare owapvas and moreover κακοστρώτους, which
occurs nowhere else: Aeschylus delights, as Wilamowitz, Interpr. 170 n. 3,
remarks, in this passage in dignifying very ordinary things with elaborate
expressions, see on 562. The same tendency is noticeable elsewhere, cf.
Wilamowitz in: his commentary on Cho. 756 (p. 223), and Griech. Tragoedien,
xiv (Ὁ. griech. Trag. u. ihre drei Dichter’), 51 n. i. [Cf. the Addenda.]
σπαρνάς: cf. Hesychius σπαρνάς" σπανίους, ἀραιάς, διεσπαρμένας, possibly
from a scholion on this passage. |
κακοστρώτους seems to be a new formation on the model of the late epic
εὔστρωτος.
For the order of the words in σπαρνὰς π. καὶ κακοστρώτους cf. on 404f.
The clause is apparently (we cannot go further owing to the uncertainty
concerning παρήξεις) in apposition to δυσαυλίας : to illustrate his point the
Herald picks out a particularly striking form of the general run of övoavdiaı ;
others might follow, but he then breaks off.
557. Several points are uncertain, but ri δ᾽ où στένοντες; quid vero non
gementes? must be kept. This phrase marks effectively the fresh start: the
Herald suddenly stops recounting further details and consequently breaks off
the construction and leaves out the required apodosis (as Schütz has correctly
pointed out). The question here is not of exactly the same type, but in effect
comparable with Pers. 1016, where the Chorus, which has just given long lists
of the lost Persian nobles (957-1001), after some interchange of lamentations,
in a way sums up the total with the question with which it restarts, τί δ᾽ οὐκ
ὅλωλεν; A still nearer parallel is the passage compared by Paley, E. Andr. 450,
where Andromache, after a string of abuse hurled at the Spartans, begins
again τί δ᾽ οὐκ ἐν ὑμῖν ἐστιν; οὐ πλεῖστοι φόνοι; ; οὐκ αἰσχροκερδεῖς; ; κτλ. At Ag.
556 Wilamowitz punctuates after τί δ᾽ où;? but quite apart from the difficulty
of his conjectures in the following words, this gives a completely false sense,
as is obvious from these passages, which are parallel among themselves:
Ag. 273 ἔστιν, τί δ᾽ οὐχί; Aesch. fr. 310. 1 λευκός" τί δ᾽ οὐχί; καὶ καλῶς ἠφευμένος,
S. Ant. 460 θανουμένη γὰρ ἐξήιδη, τί δ᾽ οὔ; (where the editors refer to 448).
The expression everywhere is equivalent to ‘of course’ and comes on top of a
strong affırmation for additional emphasis; consequently it would be quite
inappropriate at Ag. 556 after or in the course of the detailed description.
1 Supposing Triclinius found in his exemplar κακοτρώτους (so F),i it was easy for him to
recover the genuine reading, since in his ZyoA. wad. he read ἐπὶ τοῦ καταστρώματος τῶν νεῶν.
2 Cf. the Corrigenda at the end of his editio maior.
280
 COMMENTARY line 558

οὐ (or μή) placed at the end of a line so as to produce enjambement causes no

ground for objection in a late play of Aeschylus, cf. Cho. 1005 f. μὴ | γένοιτο,
Eum. 914 f. τὸ μὴ où | τήνδε... τιμᾶν πόλιν, Prom. 865, 918, and for the
phenomenon in general see below on 1354.
οὐ λαχόντες κτλ: the MS reading makes it practically unavoidable that
the hearer should from the striking anaphora infer an intended parallelism of
thought and therefore expect the second clause, οὐ λαχόντες, to convey a
positive statement as much as the first (quid non gementes? = omnia
gementes).' Cf. the passage quoted above from E. Andr. 450 ri δ᾽ οὐκ ἐν ὑμῖν
ἐστιν; οὐ πλεῖστοι φόνοι; «rA., or Eur. fr. 467. 1 N. τί γὰρ ποθεῖ τράπεζα; τῶι δ᾽
οὐ βρίθεται; (Blaydes), or in a longer form e.g. E. Andr. 298 τίν᾽ οὐκ ἐπῆλθε,
ποῖον οὐκ ἐλίσσετο dauoyepovrwv κτλ., Hdt. 7. 21. 1 ri γὰρ οὐκ ἤγαγε. . . ἔθνος
ἐπὶ τὴν Ἑλλάδα Ξέρξης; κοῖον δὲ πινόμενον ὕδωρ οὐκ ἐπέλιπε; So Klausen and
Paley had good grounds for giving its proper force to the anaphora: ‘what
was there that we had not to lament, what that we did not receive as our
daily portion?’ But it is scarcely permissible to follow Klausen and Paley in
understanding ἤματος μέρος as ‘id quod unusquisque dies fert laboris'. It
might be possible to understand from ἥματος μέρος the daily rations of the
soldiers (cf. E. El. 429 τῆς ἐφ᾽ ἡμέραν βορᾶς), though that is in itself doubtful.
But when we translate with Murray 'gemebamus cum cibum diurnum non
accipiebamus’, the force of the anaphora τί δ᾽ οὐ στένοντες, οὐ λαχόντες is
ruined by the subordination of the second clause to the first. Verrall's
parenthesis is sheer violence. To sum up. We must still agree with Ahrens
(542) : ‘the various attempts to save οὐ λαχόντες by interpretation seem to me
all to have failed'. No convincing emendation has yet been suggested.
Moreover, we cannot yet say whether ἤματος μέρος is sound. Thus the whole
line, except for the first word, remains uncertain. Wackernagel, Unters. zu
Homer, 254, judiciously remarks about Wilamowitz's ἵσματος : ‘one must
hesitate in ascribing such a formation as iopa to Aeschylus.’
558. Hermann observes ‘kai iungendum cum zAeov’,* Wilamowitz, Sonny,
Blaydes and Headlam prefer to alter the position of πλέον to follow καί, No one
will dispute the possibility of a disturbance of the original order here, such as
is common elsewhere. But it is questionable if such a supposition is necessary.
xai giving emphasis appears sometimes in an odd position, and the whole
question needs new investigation, which cannot be given here. The aphoristic
treatment by Wilamowitz (on E. Hipp. 391 and Her. 217) is not adequate, as
is shown e.g. by his wrong interpretation of Hipp. 1326; nor is Denniston’s
criticism sufficient (Particles, 326). Presumably in a large number of cases
we are not justified in looking for one single word to which καί belongs. (In
a preceding section Denniston suggests that the writer sometimes treats two
words as an indivisible unit and places καί before the first even when it bears
less emphasis than the other ; but this does not go far enough.) Often it looks
as if the emphasizing function of καί extends itself over a whole clause, no
matter of how many words it consists, and that then καί comes at the begin-
1 On rives οὐ = ἅπαντες and its Latin equivalent qui non see J. B. Hofmann in Schmalz,
Lat. Gramm., sth ed., 833, and Gnomon, ix, 1933, 518 f.
2 He rejects the traditional punctuation, viz. comma after προσῆν (so in F). In Tra
comma is put after χέρσῳ (so e.g. Paley). It should, however, be remembered that these
copyists read τάδ᾽ at the beginning of the line.
281
dine.558 COMMENTARY
ning of that clause: see Pearson on E. Hel. 79, 1069, 1200; and cf., e.g., Ar.
Lys. 1226. Here the phrase ra δ᾽ αὖτε χέρσωι clearly forms an organic kolon,
which, as a kind of heading, introduces the following sentence, cf. Ed. Fr.,
Kolon u. Satz ii. 327 f., 336 ff. At the head of the next kolon comes the
emphasizing καί. Cf. on 108.
'559ff. This naturally sticks as a particular hardship in the memory of
soldiers on active service. Thus the Chorus sings in Ar. Peace 347 f. πολλὰ γὰρ
ἀνεσχόμην πράγματά τε καὶ στιβάδας (Schol. ὅπερ ἦν πάσχειν ἐπάναγκες τοῖς
πολεμοῦσιν, ὑπαιθρίους ταλαιπωρεῖν καὶ χαμαικοιτεῖν).
559. δαΐων: so Dindorf, because the word always appears with a elsewhere in
Tragedy even in dialogue (Prom. 352,1 S. Aj. 784), except Cho. 628 ;? cf. also
Sept. 918 δαϊόφρον (MSS Saippwv). Ar. Frogs 1022 (cf. also 897) seems to show
that δάϊος was recognized as the tragic form (Clouds 335 δάϊον is naturally
used in a parody of Doricized lyric). Secondary intrusion of the Homeric
form may be at work: Suidas s.v. δηΐα quotes Ajax 784 ὦ δηΐα Τέκμησσα. But
on the other hand it is possible that the tragedians themselves occasionally
used δήϊος as well as δάϊος, especially as the verb is always δηιοῦν, and
δηιάλωτον occurs in trimeters in A. Sept. 72 and also in a dactylic hexameter
in E. Andr. 105, and ἀδῆιον in trimeters in S. Oed. C. 1533. Cf. K. Zacher,
Diss. phil. Halens. iii (1878), 28f. In the anapaestic fragment from the
beginning of Aeschylus’ Myrmidons (P. Oxy. 2163, fr. 1. 5) ônil is preserved
at the beginning of a line. We cannot determine with certainty whether the
tragedians used only the form with a or admitted that with n as well.
560. John Pearson saw that δὲ had been displaced by yap, though Hermann
defends the latter: see the parallel instances collected by Headlam, On
Editing Aeschylus, 119 f., e.g. Ag. 416 (δὲ F, yap Tr.). It is not clear whether
Denniston, Particles, 64 f., wishes to recognize here a case of ‘successive yap’s’:
he regards Verrall’s parenthesis (from καὶ προσῆν to reiyeouw—Hermann had
made a similar suggestion) as defensible.
ἐξ οὐρανοῦ ... κατεψάκαζον : ‘non perversa est oratio, si sic verba intelli-
guntur, ἐξ οὐρανοῦ δρόσοι καὶ ἀπὸ γῆς δρόσοι λειμώνιαι᾽ Hermann, who neverthe-
less prefers Schütz’s λειμωνίας, which destroys the antithesis ἐξ οὐρανοῦ καὶ
ἀπὸ γῆς. No doubt the words κἀπὸ γῆς λειμώνιαι (sc. δρόσοι) are thrust in
interjectionally upon the clause ἐξ οὐρ. δὲ κτλ. (so Verrall) and the verb
κατεψάκαζον follows as though the insertion had not been made. Such slight
inconsistencies, common in lively speech or narrative, are well illustrated by
Wilamowitz on E. Her. 222. Cf. above on 318. The peculiarity of this passage
lies in the fact that the insertion cannot be isolated so easily as usual from its
setting, because λειμώνιαι δρόσοι are closely connected and δρόσοι has to be
taken both with ἐξ οὐρανοῦ and with ἀπὸ γῆς. The ‘dripping’ of the moisture
from above and below must be regarded as one thing: that is precisely what
makes the picture striking. δρόσοι as one of the chief plagues of campaigning
were mentioned in 336. For the bivouacking of the army in the meadows
before Troy (᾿1δάιδι μίμνων molar λειμωνέδι) cf. S. Aj. 601 (quoted by van
Heusde, see also Wilamowitz, Verskunst, 509 f.).

1 δαΐων in a trimeter also in Sept. 277, if the passage is correctly restored by Hermann
(cf. 278); cf., however, Regenbogen, Hermes, Ixviii, 1933, 6o f.
2 No confident judgement can be given on δήιοις here so long as the following words are
not understood with certainty (cf. below, p. 351 n. 2, for this point).
282
 COMMENTARY line 562

561. The form κατεψάκαζον was adopted by Dindorf in accordance with the
Atticists (cf. on 1534).
Regarding the construction of ἔμπεδον σίνος there are three possible ways,
always supposing that nothing has dropped out and that the following line is
sound, 1.6. that τιθέντες agrees with δρόσοι (see below). The most improbable
is the punctuation adopted by Heath, followed still by Murray as by many
of his predecessors : Heath commas off ἔμπεδον σίνος ἐσθημάτων! and translates
‘certa vestium pernicies’. That separates ἐσθημάτων from τρίχα, which seems
to belong to it ; for other objections see Headlam. Other editors, e.g. Weiland
Wilamowitz, place a comma only after σίνος. If they mean by this that
ἔμπεδον σίνος is to be taken as accusative governed by κατεψάκαζον, this
should be rejected, since an intransitive κατεψάκαζον is here much more
effective (for intransitive καταστάζειν cf. Körte, Hermes, lxx, 1935, 434 n. 4).
The only remaining alternative is to take ἔμπεδον σίνος as accusative in
apposition to the sentence: thus Headlam encloses the words between
commas.
562. The structure of both sentence and line suggests taking ἐσθημάτων τρίχα
together, governed by τιθέντες with ἔνθηρον as predicate; we should not take
the words otherwise without good reason.
€vÜnpos is formed exactly like ἔνθεος (cf. Radermacher, Rk. Mus. Ixxxv,
1936, 2, and my note on 996 ἐνδίκοις). 'évÜnpov bene interpretatur Plüss
“bestiolis infectam" ' (H. Weil, Teubner edition, 2nd ed., preface xliv). It is,
in harmony with the general style of this description (see on 556), a 'dignified
periphrasis’ (Wilamowitz, Interpr. 170 n. 3) but playful and not entirely
serious, 'for θηρίον was applied in more or less humorous horror to the smallest
creatures’, as Headlam says, who rightly explains: ‘causing mildew and
making the hair or wool of our garments ['the σισύρα or κατωνάκη᾽ Wilamo-
witz, loc. cit.] verminous’. Schütz, Hermann, and many others have taken
τρίχα to mean the hair of the soldiers, putting a comma after ἐσθημάτων (see
above). Headlam believes that Sophocles has taken the passage in this way
when he, in the passage quoted already by Abresch, Aj. 1206 ff., makes the
Chorus say κεῖμαι δ᾽ ἀμέριμνος οὕτως del muxwais δρόσοις τεγγόμενος κόμας,
λυγρᾶς μνήματα Τροίας. But that this trait (reyy. κόμας) is borrowed from
Aeschylus and not from the common experience of every soldier would be
difficult to establish.
τιθέντες : it has been discussed for more than a hundred years whether this
participle can be taken with the feminine δρόσοι (besides the commentaries
cf. e.g. Boeckh on Pind. Ol. 6. 15, Pindars opera, ii. 2, p. 155, and Matthiä,
Griech. Gramm. ὃ 436. 2, p. 981). If the possibility is excluded, we must regard
the passage as defective or corrupt, since to understand ὄμβροι or something
similar is purely arbitrary. In order adequately to answer the question
whether τιθέντες here can legitimately function as feminine, we must not
appeal to material of a different type, e.g. to Lobeck’s collection of late
solecisms (Aglaophamus, 217) : Nicander too should be rejected as an author-
ity, in view of his well-known arbitrary treatment of the language (cf. W.
Kroll, RE xvii. 259). But even if we confine ourselves to older Greek usage, it
would be more prudent to compare with this passage only such as are really
of the same kind, i.e. participial uses, and even among them passages of
1 After ἐσθημάτων there is a colon in F, a comma in Tr.
283
line 562 COMMENTARY
doubtful interpretation should be excluded.! Wilamowitz, Interpr. 195, for
instance quotes among examples of the feminine use of the masculine parti-
ciple Eum. 960, but there his arrangement of the text (cf. also Interpr. 227 f.)
is very uncertain ;? nor should he have used Ag. 696 as evidence (see my note).
The final residue is discouragingly small, but too large to allow us to deny
the possibility of the occurrence out of hand. Besides τιθέντες here, we are
concerned principally with Ag. 120 BAaBévra, which possibly is a feminine,
and with Pind. Ol. 6. 15 ἑπτὰ δ᾽ ἔπειτα πυρᾶν νεκρῶν τελεσθέντων. Here indeed
Wilamowitz himself, like others, doubts the MS reading (Jsyllos, 163 n. 3
and later Pindaros, 310 n. 3), while O. Schroeder finally (Appendix of 1923 to
his editio maior, p. 512, and 3rd ed. of the editio minor) has left τελεσθέντων
in the text and taken it as feminine, as also one of the greatest and most
cautious experts of the last few generations, Wackernagel, Unters. zu Homer,
59 n. 2, and Syntax, ii. 47. To this should possibly be added a passage from a
different type of Greek in the recently found fragment of Sophron (Pap. Soc.
It. xi. 1214 = D. L. Page, Greek Lit. Papyrt, i. 330 f.). In 1. 5 the following
order is given to those taking part in the magic performance: ποτιβάντες νυν
not τὰν ἱστίαν Üwkeire. The fragment comes from the mime entitled ταὶ
γυναῖκες at τὰν θεόν avr. ἐξελᾶν (fr. 3 ff. Kaibel). ‘It is not very probable, from
all we know of rites of this kind, that men took part as well as women in
the exorcism’ says Latte, Philol. Ixxxviii, 1933, 261 n. 4; he concludes that the
masculine stands here as the general form of the verbal adjective, as in the
examples quoted in Kühner-Gerth, i. 82 f. ;? i.e. especially passages in Tragedy
where a woman speaks of herself in the masculine plural.* These passages
1 The aphoristic remarks of A. v. Blumenthal, Hermes, lxxv, 1940, 127 f., are not
helpful.
2 ] have not so far reached any quite satisfactory conclusion about this thorny passage.
Even if one regards Hermann's text, which most editors have adopted, as only a provisional
solution, it should be admitted that it is much less violent and his interpretation less
involved than that of Wilamowitz. ‘A completely otiose 8eai, and a simply unheard-of
position for the &', says Wilamowitz, Interpret. 227, with reference to Hermann's text. I
cannot agree with either of these points. If we write θεαί 7’ ὦ Μοῖραι ματροκασιγνῆται, then
θεαὶ Μοῖραι are not to be taken together (as many translators take them), but θεαὶ ματρο-
κασιγνῆται is in apposition to Μοῖραι, and this corresponds to the type of word-order illus-
trated by H. Schóne, Hermes, lx, 1925, 155 (θεὰ Θέτις ἀργυρόπεζα and the like). And as
regards the position of the ὦ, careful investigation would probably discover some com-
parable instances in addition to the Homeric πάτερ ὦ £eive (ἀγακλεὲς ὦ Μενέλαε and the
like being rather different) or [Hesiod] Scutum 78 ἥρως ὦ ᾿Ιόλαε βροτῶν πολὺ φίλτατε
πάντων and Pind. Ol. 8. 1 μᾶτερ ὦ χρυσοστεφάνων ἀέθλων ᾿Ολυμπία (both in the Thesaurus),
Apoll. Rh. 4. 1411 δαίμονες ὦ καλαὶ καὶ ééppoves, Callim. hymn. Del. 118 Πήλιον ὦ Φιλύρης
νυμφήϊον. Wilamowitz's inference of a varia lectio κῦρος from the reading in FTr (κύριες)
must be abandoned: cf. P. Maas, 'Griech. Paláographie', in Gercke-Norden, Einl. 1. d.
Altertumsw. i. 9, 3rd ed. 1924, ὃ 21, p. 80.
3 Express agreement with Latte's view is to be found in Körte, Archiv für Papyruskunde,
Xi, 1935, 266 n. 4. E. Andr. 712 τίκτοντας ἄλλους, noted by Körte as ‘particularly striking',
cannot be treated on a level with the cases with which we are concerned here.
* For the passages in Tragedy where a woman uses the singular of the masculine participle
in speaking of herself cf. Wilamowitz on E. Hipp. 1105 (p. 231). Among the examples
quoted there E. Phoen. 1724 should be cancelled (cf. Wilamowitz, Berl. Sitzgsber. 1903,
593 n. 1), in A. Sept. 565 Wilamowitz overlooked the difficulty of the text, in Zum. 297 he
misunderstood the meaning (it is a general statement). Whether τίων in Cho. 629 belongs
to this category, as Wilamowitz assumes in his commentary on the passage (p. 215), is
uncertain owing to the extremely corrupt state of this stanza, and it is still more uncertain
whether one ought to follow Wilamowitz in altering φρενῶν to φέρων in 626.
284
 COMMENTARY lines 565 f.
may at a pinch be regarded as analogous to roriBävres,' but the latter might
also be compared with instances like δρόσοι τιθέντες (expressly separated from
the other group by Kühner-Gerth, i. 83 n. 1). The upshot is that we have
no very good grounds for assuming that τιθέντες can go with δρόσοι but
perhaps enough for hoping that this assumption is justified. Wackernagel,
Unters. zu Homer, 59 n. 2, speaks of the ‘tendency of -vr- stems not to indicate
the difference of gender’. So supposing the text here is correct, ἐσθημάτων...
τρίχα must be taken as explanatory of the clause ἔμπεδον σίνος, which itself is
in apposition to the process specified in κατεψάκαζον. For this method of
adding successively one explanatory and picturesque detail to another
cf. on 2.
563. χειμῶνα again placed first as a ‘headline’ as 558 τὰ δ᾽ αὖτε χέρσωι and
565 ἢ θάλπος.
οἰωνοκτόνον does not occur elsewhere.
564. οἷον... ἄφερτον: this use of οἷος with an explanatory adjective is
Homeric (see the instances collected by Ebeling, Lex. Hom. ii. 40). Passow
(followed by L-S s.v. οἷος II. 6) expressly warns us against taking the
adjectival οἷος as purely adverbial, = οἷον, ὡς, in passages such as A 653 f.
: οἷος ἐκεῖνος δεινὸς ἀνήρ, T 493 οἷον ἐμὸν μένος ἔμπεδον οὐδ᾽ ἐπιεικτόν, and the like.
L-S rightly say ‘what manner of man, namely dread’, Ameis-Hentze on
T 493 ‘ οἷον is explained by ἔμπεδον «ré '. It seems to me to be going too far
when many editors (including Leaf) at A 653 put a comma after οἷος ἐκεῖνος :
that makes it look as though ofos was complete in itself. Aeschylus here
conforms with the Homeric usage of putting at least one word between οἷος
and the following adjective.
This passage is very similar to Herodotus’ description of the Scythian
winter 4. 28. 1 δυσχείμερος δὲ αὕτη ἡ... χώρη οὕτω δή Ti ἐστι, ἔνθα τοὺς μὲν
ὀκτὼ τῶν μηνῶν ἀφόρητος οἷος γίνεται κρυμός κτλ. The last words quoted are
strikingly similar (though in the case of οἷος it 15 ἃ mere accident, for the word
is used in a different way).
On ἄφερτος see on 386.
Here we see the effect of experiences such as those of the Persian army on
the Strymon (Pers. 495 ff.), which had certainly been shared by many
Athenian expeditions in the north. The Athenians remembered the winters in
the region of Potidaea (Plato, Symp. 220 a δεινοὶ αὐτόθι χειμῶνες), cf. Ar.
Ach. 138 f.
565 f. In two lines Aeschylus conveys a full picture of the southern sea under
the midday sky. So does Goethe:? ‘Ein weisser Glanz ruht über Land und
Meer, Und duftend schwebt der Äther ohne Wolken’. But in the words of the
Herald there is no touch of the pleasure which the northern artist takes in the
1 Chantraine, Revue de Philologie, XIIe serie, ix (1935), 24, rightly remarks that these
passages of Tragedy ‘présentent un caractère de généralité qui ne se retrouve pas ici [in the
passage of Sophron]’. It would be a way out of the difficulty if we could assume that male
attendants took part in such cults. But the objections to this assumption are more serious
than D. L. Page, loc. cit., seems to think. The treatment of the Sophron passage by H. J.
Rose, ‘The Eclogues of Vergil’, Sather Classical Lectures, xvi, Univ. of California Press,
1942, p. 2 f., does not take into account Latte’s article and the discussion which has arisen
from it.
2 It is well known that in this passage of the ‘Nausikaa’ Goethe freely adapted motifs
from & 44 f.
285
lines 565 f. COMMENTARY

brilliance and beauty of the colouring: the weight of the motionless sultriness
rests on everything ; if only something would awake to life, a wave or a light
puff of wind! The sailor is not in a position to admire the calm sea, he wants
to go on and suffers from the lull. No less powerful is the brief picture of the
sea moved by the wind rising about sunrise in 1180 ff.
565. Cf. Ar. Birds 1096 θάλπεσι μεσημβρινοῖς, Callim. Lav. Pall. 72 uecauflpwa
δ᾽ εἶχ᾽ ὄρος ἁσυχία. Obviously ἐν μεσ. κοίταις νηνέμοις means ‘in windstiller
Mittagsruhe’ (so Nägelsbach and other editors), not ‘in his midday couch’
(Conington, similarly Paley, L. Campbell, Headlam), ‘sur sa couche’ (Mazon).
‘The sea does not lie down on a bed’ says Wilamowitz, Platon, ii. 359, where
he compares this line with the expression ἀνέμων κοίτην (Sympos. 197 c) which
he correctly explains. ‘«oirn condicio iacendi’ (Klausen), cf. on 1494.
566. πεσών as Eum. 68 ὕπνωι πεσοῦσαι.
567. τί ταῦτα «rÀ.: the lively question beginning with τί cuts off the con-
struction, cf. 556.
568. παροίχεται δέ sc. πόνος. Examples of δέ in anaphora without preceding
μέν in Denniston, Particles, 163. 2. The following infinitive is consecutive
(= ὥστε). This interpretation, which Stanley had already given in his trans-
lation, was rightly followed by Schütz, Hermann, and most other editors;
Fr. Birklein, Beitr. z. hist. Syntax der griech. Sprache, hsg. von M. Schanz
(Würzburg 1888), 19, supported it by parallels with special reference to the
presence of the article. Cf. Pers. 292, Ag. 15, 1171, 1589, Cho. 302, Eum.
220 etc. On the other hand Blomfield, Kennedy, Wecklein (Stud. z. Aesch.
20 and in his commentary) and Plüss wish to make 76... . μέλειν the subject of
παροίχεται. What the last two say about μή is misleading, when they compare
sentences in which the infinitive comes as object of the verb (‘verbs of
forbidding’ Pliiss, ‘as after ἀπομνύναι etc.’ Wecklein). Above all: it would not
be sensible to say of μὴ μέλειν that it παροίχεται.
570-2. Objection has long been taken to these lines because they plainly
break the connexion between τοῖσι μὲν τεθνηκόσιν and 573 ἡμῖν δὲ τοῖς λοι-
ποῖσιν. Some scholars have suggested transpositions, whereas F. Bamberger,
Opusc. 56, attempts to attach the lines to their surroundings with the usual
psychological spider’s web. Ahrens says more honestly (p. 543): ‘the lines
must be regarded as a parenthesis.’ But what sense is there in speaking of a
parenthesis in a speech like this which is so lively and marked throughout
with breakings-off and restarts? And after this one the speech goes straight
on as though nothing had intervened! Consequently Nägelsbach (in his
translation he puts 570-2 in square brackets) and later Blaydes and Wilamo-
witz! eliminated the lines as an interpolation ; Weil considered this possibility
together with an alternative solution.” But Wilamowitz has (Hermes, lxii,
1927, 285) returned to the passage with a thorough rediscussion, and defends
the lines. In this he bases his judgement entirely on the quality of the lines,
while taking no notice of the main difficulty, the connexion with the context,
simply saying: ‘undeniably 573 follows naturally on 569; the three lines
between are unnecessary (‘entbehrlich’). But only Aeschylus can have
written them: the words τύχη παλίγκοτος have the true Aeschylean ring’, etc.
1 He too had previously (1885) transposed the three lines, placing them after 579.
2 In the preface to his Teubner edition he remarks: ‘Post 569 Elberling inserebat 573-574.
Sed videntur potius delendi esse aut 568-569, aut 570-572.’
286
 COMMENTARY lines 570-2

The nature of the intrusion must therefore be examined more exactly. It is

in itself always conceivable that an independent clause may intrude between
a pév ... and Ôé ..., e.g. Pind. P. 5. 15-20 τὸ μέν, ὅτι βασιλεὺς ἐσσὶ μεγαλᾶν
πολίων" ἔχει συγγενὴς ὀφθαλμὸς αἰδοιότατον γέρας τεᾶι τοῦτο μειγνύμενον φρενί"
μάκαρ δὲ καὶ viv, κλεεννᾶς ὅτι εὖχος ἤδη παρὰ Πυθιάδος ἵπποις ἑλὼν δέδεξαι
τόνδε κῶμον ἀνέρων, with the explanation offered by Wilamowitz (Pindaros,
381 n. 1): ' μέν 15 and δέ zo join kingship and the victory in the race together.
Between them is an unconnected sentence in which the preceding general
maxim [l. 12f.] is applied to Arcesilaus: ἡ πατρώια βασιλεία αἰδοιοτέρα
γίνεται προσγιγνομένης τῆς παρὰ σοὶ σοφίας. But in the Pindaric passage the
intruding sentence serves as an explanation of the preceding uév-clause, as
Wilamowitz shows in his paraphrase: 'firstly because thou art a mighty
king; this inherited prerogative has a worth which compels respect because
thy personal qualities are combined with it; secondly thou receivest this
festal procession for thy Pythian victory'. Here, on the contrary the 'rhe-
torical' question τί τοὺς ἀναλωθέντας κτλ. breaks the sequence of the sentence
and claims independence for itself. But, above all, the thought of 570 ff.
("Why count the dead, why should the living trouble themselves over the
adverse fortune?') takes a diametrically opposite line to what precedes and
what follows ('the labour is over, over for the dead so that they . . .' and 'the
gain outweighs the loss for us who survive’). The change of standpoint from
the dead for whom it is now over to those who count the dead is harsh, but
still more harsh and in fact scarcely tolerable is the effect of the insertion on
ἡμῖν δὲ rois λοιποῖσι, which is anticipated and weakened by τὸν {üvra.! The
thoughts of 568 f. and 570 f. are related to one another but are given an
entirely different turn in the two places. That reminds us strongly of the
many instances where in plays of Euripides, especially those most read, a
general maxim from a different play, often another of Euripides, has been
intruded, always roughly resembling the subject of the passage and some-
times similar in tone and thought, but pointless where it has now been
inserted.? Here the insertion may have been facilitated or caused partially
by the parallelism of τί... χρή with τί... δεῖ in 567. For I consider it at any
rate probable that 570 ff. is not a supplement made ‘ad hoc’ but a parallel
passage written in the margin. Whether the addition comes from Aeschylus,
I do not know:* τύχης παλιγκότου (for the meaning of the adjective cf. on 874)
is not sufficient for a diagnosis, especially as παλιγκότως is attested for
Euripides; for 570 cf. Rhesus 309 f. οὐδ᾽ àv ἐν ψήφου λόγωι θέσθαι δυναίμην.
There is no fault to find and no difficulty in 570 f. χρή stands ἀπὸ κοινοῦ, as
Stanley saw ; Schütz thought that τὸν ζῶντα also was ἀπὸ κοινοῦ, and so did
Wilamowitz in his edition, but he has recanted loc. cit. For the construction
ἀλγεῖν τύχης Hermann compared E. Hec. 1256. Ahrens tries in vain to get rid
! Moreover it would be heartless to address words like these to fathers who have lost
their sons (Beazley).
? See my remarks, Eranos, xliv, 1946, 83 ff.
3 This possibility was considered (but then rejected) by Wilamowitz in the preface to
his edition p. xxx and Hermes, lxii, 1927, 285, where he compared the interpolations at Ag.
527 and Eum. τος (in his app. crit. to the latter passage the note 'delevi' should be corrected
to ‘del. Prien’ and inside the brackets ‘Kirchhoff’ should be replaced by ‘Schiitz’).
+ Wilamowitz (p. xxx of his edition) said of these lines ‘versus optimi et Aeschylo non
indigni'.
287
 lines 570-2 COMMENTARY

of this parallel. This use of the genitive fits in without trouble among the
instances! collected by Kühner-Gerth, 1. 365 f.
572 is difficult. χαίρειν συμφοραῖς is explained in the scholion (Σχόλ. παλ.
in Tr): ἐπὶ ταῖς εὐποτμίαις χαίρειν, which is linguistically correct ; for συμφορά
(or cuupopat) used ‘de laeto eventu' (better ‘neutral’ as Blass on Eum. 897)
Dindorf, Lex. Aesch., s.v., quotes five passages (not including this). This
meaning is attested in Aeschylus only in the Oresteia. Schütz translates in
agreement with the scholion ‘immo vero aequum puto multum super his quae
acciderunt laetari’. This rendering does justice also to the meaning of
καταξιοῦν ; it has been followed by Bothe and others, e.g. Paley (with hesita-
tion), Schneidewin-Hense, and, with inconsiderable alteration of the text,
by Ahrens (p. 513). But this explanation is open to serious objection in spite
of its verbal correctness. For it is unlikely that an Athenian would under-
stand by πολλὰ χαίρειν anything other than the everyday ‘good-bye’, which
in the form πολλὰ χαίρειν λέγω (or κελεύω) quite often means 'begone with him’,
so e.g. E. Hipp. 113, Ar. Ach. 200, Pl. Phaedr. 272 e. So Stanley translates:
‘immo calamitatibus valedicendum censeo’. But there are difficulties in
respect to καταξιῶ (so far Ahrens is right). L-S have a separate heading
‘command, bid’: under this they quote besides the present passage only S.
Phil. 1095 σύ τοι κατηξίωσας and add ‘thou didst decree it so’, i.e. they adopt
Jebb’s translation (this agrees with Ellendt-Genthe s.v. : ‘decrevisti’), instead
of following the right explanation in his note: ‘hast thought it right (to have
it so). Blomfield's συμφοράς does not relieve the difficulty that lies in the
verb. The material which Headlam has collected is not adequate to support
his perhaps correct assertion that 'xaipew καταξιῶ is merely one of the many
variations of the phrase χαίρειν λέγω or κελεύω᾽, since all the passages he
quotes definitely contain a word of saying, which is precisely what we here
miss. But it is doubtful whether καταξιῶ can stand ‘shortly for εἰπεῖν ἄξιον
ἡγοῦμαι᾽ (so Wilamowitz and before him Hermann: 'malis aequum censeo
valeatis dicere"). Consequently we must reckon with the possibility that 572
was written by a bungler, as well as with the possibility of an arbitrary
tragic use of the verb. If the former alternative is true, l. 572 has not
the same origin as the two previous lines, but may have been added by
the ‘interpolator’. A further possibility is that in its original context the
sentence ran on and that some such infinitive as Aéyew followed in the next
line.
574. νικᾶι τὸ κέρδος : here the function of the Herald's scene as a short pause
in the action to relieve the strain finds concentrated expression. Hitherto
there has been heard no utterance of assured confidence, only wishes and
prayers of the depressing type of γένοιτο δ᾽ εὖ παρὰ γνώμην ἐμήν (see note
on 255), from the unfulfillable wishes of the Watchman (34 f.) and the in-
sistent repetition of aldıvov . . . τὸ δ᾽ εὖ νικάτω and the méAovro . . . εὖ (255) on
. the part of the Chorus to the ambiguous prayer of the queen (349) τὸ δ᾽ εὖ
κρατοίη and the renewed gloom of the Elders in the later part of the first
stasimon. Only the Herald can utter words of joyful satisfaction ; see the
observations on this speech as a whole (pp. 293 f.).
1 There is no need to decide here whether in S. Oed. R. 233 f. we should follow Schneide-
win, Kühner-Gerth, Bruhn, and Wilamowitz (translation) in taking φίλου δείσας together,
or adopt Jebb's explanation which seems to me to be highly artificial.
288
 COMMENTARY lines 575 f.

τὸ κέρδος : the article defines (one might well take it as ‘possessive’) : ‘the
particular prize that we have earned’.
ἀντιρρέπει: occurs first here, and only here in poetry. The adjective
ἀντίρροπος occurs in the epigram (Tod, Greek Hist. Inscr. no. 59) on the men
who fell at Potidaea (432 B.C.) ψυχὰς δ᾽ ἀντέρροπα θέντες ἠλλάξαντ᾽ ἀρετήν and
S. El. 120.
575 f. Long ago Heath felt a difficulty about ποτωμένοις. It cannot really
mean ‘qui per mare per terras incolumes reversi sumus’ (Hermann). F.
Bamberger’s rendering (Opusc. 246) is linguistically more exact, but his
explanation implies a highly improbable thought: ‘‘‘We who escaped may
to-day boast as we are swiftly flying over land and sea". ποτωμένοις stands
simply for the swift return home, and the sense is that they left behind in all
the temples they came to as they returned trophies as a memorial of the
great deed they had accomplished’. Imagine the dropping of trophies (com-
plete with dedicatory inscriptions) from the air! That fame or its recipient
(Theogn. 237) flies out over the land is a not unusual idea (see the instances
collected by Pearson on Soph. fr. 941. 11 and C.R. xl, 1926, 184), but we should
not elaborate the picture as Headlam does: ‘so that we may well go flying
over land and sea upon the wings of fame this bright day boasting thus’ or
Platt : ‘Therefore may we mount up like eagles to spread our wings over land
and sea.’ Heath's conjecture ποτωμένωι has been adopted by Ahrens and
Wilamowitz among others. For Helios being represented as the bringer of a
message Ahrens refers to S. Aj. 845 ff. (Trach. 94 ff. is different). We might
also refer to Ag. 632 f., where, it is true, the chief emphasis lies on the fact
that Helios πάντ᾽ ἐφορᾶι καὶ πάντ᾽ ἐπακούει, but it would be possible for him
ἀπαγγεῖλαι τορῶς. Cf. also E. Heraclid. 748 ff. Above all we must not forget
0 270 f. ἄφαρ δέ ot ἄγγελος ἦλθεν "HAuos.! But it is questionable whether
ποτᾶσθαι is a good word to use in describing the motion of the sun (in passages
like E. Ion 122 dp’ ἡλίου πτέρυγι θοῆι, Orphica 62. 3 Kern 'HéAe, χρυσέαισιν
detpopeve πτερύγεσσιν we should possibly? only think of the winged horses, cf.
E. El. 466, Or. 1001). But, above all, while the proud victor may κομπάσαι, the
expression would hardly be appropriate of the god. Ahrens seems to have
felt the difficulty: he says: ‘xoumdoat means that the sunlight itself enjoys
that glory, and therefore boastingly announces it', but of this thought there
is no trace in the passage nor is it in itself satisfactory. More probably Weil
(Addenda to his edition of Eum., 1861, p. 131) has hit on the right reading:
ποτώμενα. If we accept this, the passage would be closely related to Pind.
N. 6. 48 πέτεται δ᾽ ἐπί τε χθόνα kai διὰ θαλάσσας τηλόθεν ὄνυμ᾽ αὐτῶν (quoted by
Paley). It is well known that Aeschylus, who has a liking for the word (which
only once occurs in Sophocles), and other poets used ποτᾶσθαι in the same
sense as πέτεσθαι.
It is difficult to decide how τῶιδ᾽ ἡλίου φάει should be taken syntactically.
Generally it is regarded as temporal dative. For the strong emphasis laid
1 It is natural to think of the sun as the quickest messenger, cf. Shakespeare, Tempest,
Act II, sc. 1. 247 ff. ‘she that from Naples can have no note, unless the sun were post,—
the man i’ the moon’s too slow,—till new-born chins be rough and razorable'.
2 But not necessarily. Cf. the words from the Peirithoos of Kritias (fr. 18 Diels-Kranz,
Vorsokr. ii, sth ed., 384 = Eur. fr. 594 N.) δίδυμοί 7’ ἄρκτοι ταῖς ὠκυπλάνοις πτερύγων
ῥιπαῖς τὸν ᾿Ατλάντειον τηροῦσι πόλον, on which Wilamowitz observes, Hermes, Ixiv, 1929,
464: ‘It is an unparalleled conception that the stars have wings.’
4872-2 U 289
lines 575. COMMENTARY
on the notion ‘on this day’ with regard to the victory, we may compare
1. 320, and for the use of ἥλιος = ‘day’ cf. Martin P. Nilsson, Entstehung . ..
des griech. Kalenders (Lund 1918), 11, and the same author, ‘Primitive Time-
Reckoning’ (Acta Soctet. hum. litt. Lundensts, i, Lund 1920), 13 n. 1, further
Weinreich, Hermes, lvi, 1921, 329 n. 1. Others, e.g. Franz, Schneidewin,
Paley (transl.), Sidgwick, Platt, take τῶιδ᾽ ἡλίου φάει to mean ‘before (the
face of) this light of the Sun’, approximately in the sense of πρὸς τόδε φάος.
This is unquestionably possible in general, in particular we may quote Eum.
590 οὐ κειμένωι πω τόνδε κομπάζεις λόγον, on which Blass remarks ‘the dative
as with Aéyew'. This way of taking the words would equally allow room for
the motive which Ahrens finds in the conjecture ποτωμένωι: ‘it is as if the
sunlight reads the inscriptions on the λάφυρα which the Argives will dedicate
after their return.' It is difficult to decide which of the two meanings ('hoc
sole' or 'adversus hunc solem') is more appropriate. However, I am not sure
that Aeschylus wrote τῶιδ᾽, for if we accept Weil's τάδ’, the sentence gains in
perspicuity ; moreover 'apud Aeschylum quoties verba pronuntiata vel pro-
nuntianda per orationem directam afferuntur, non facile deest pronomen
demonstrativum' (Weil). In that case ἡλίου φάει would of course be the dative
of the person addressed.
577-9. The dependence of these lines on the typical form of dedicatory
inscriptions is obvious and has long been recognized, e.g. by Weil: 'Inde
praeco, spoliorum titulum componens, de rebus praesentibus quasi de
praeteritis loquitur: inde ἐπασσάλευσάν more.’
577. ποτε: in a corresponding context e.g. at the beginning of the inscription
on the altar of Zeus Eleutherios at Plataea, 479/8 B.c. (Plut. Aristid. 19. 7 =
Mor. 873 b; Hiller v. Gaertringen, Hist. griech. Epigramme, no. 26) τόνδε ποθ᾽
“Elmves . . . ἱδρύσαντο Διὸς βωμὸν ᾿Ἐλευθερίου. For ποτε especially in epitaphs
cf. Wade-Gery, J. H. St. liii, 1933, 72 ff., who takes into account this passage
of the Ag. (p. 77 n. 23). more is also found in an ironical epitaph similarly
embodied in the iambics of a play at E. 770. 119o.
578. ταῦτα : forms of ὅδε are more usual in dedications, but cf. e.g. the offering
of the Corinthian Diodorus for Salamis, 480/79 B.c. (Simonides fr. 108 D. ;
Hiller v. Gaertringen, op. cit., no. 21) ταῦτ᾽ ἀπὸ δυσμενέων Μήδων ναῦται
Διοδώρου ὅπλ᾽ ἀνέθεν.
579. ἐπασσάλευσαν : the aorist here follows the established use of ἀνέθηκε
etc.; cf. J. Wackernagel, Studien z. griech. Perfektum (Göttingen, 1904), 8.
It remains to discuss some details in these three lines. 577. Tpotav: for the
position of the word cf. 269, 320. For the dative θεοῖς 578 cf. Kühner-Gerth,
i. 419 (‘acts of a religious cult in honour of a god’). With 578 f. Stanley has
compared Rhesus 180: θεοῖσιν αὐτὰ πασσάλευε πρὸς δόμοις. This is a less harsh
expression than the σχῆμα ᾿Ϊωνικόν ‘they nailed it for the gods for their
houses’, but should at any rate suffice to protect the Aeschylean δόμοις from
being weakened down to δόμων. 578. θεοῖς. . . τοῖς καθ᾽ Ἑλλάδα: ‘the
gods throughout Greece’ (Verrall). The idea seems to be that parts of
the booty will be dedicated by the individual partners in the campaign in all the
chief sanctuaries of the Greek world. In the temple of Athena at Lindos there
were, as ἀκροθίνια τῶν ἐκ Τροίας, nine panoplies given by Tlepolemus and his
men, Alexandros’ helmet dedicated by Menelaus, a silver quiver as the gift
of Meriones, the quiver of Pandarus as the gift of Teucer (Die Lindische
290
 COMMENTARY line 583
Tempelchronik B 54-86, ed. Blinkenberg, Bonn, 1915, pp. 12 ff.). 579. ἀρχαῖον
yavos: Objection has often been made to the adjective (from the time of
Pauw and Heath), as equally at 5. Oed. C. 1632 (δός μοι χερὸς σῆς πίστιν
ἀρχαίαν τέκνοις), where Jebb alters the text, following many others. Hermann
already connected the two passages and adhered to the MS reading. But his
rendering is too artificial: ' πίστιν ἀρχαίαν dicit quae firma maneat, olim
antiqua futura', even if we connect it, as a number of editors do, with the
practice common in inscriptions of taking the viewpoint of the future (cf.
above on sore); it might also be asked whether παλαιάν would not be the
better word to express such a sense. Wilamowitz says rightly (in T. v.
Wilamowitz's book, Dramat. Technik des Soph. 365 n. 1): 'Here (Oed. C. 1632)
the χερὸς πίστις is not old because Theseus has given it previously, or because
it is in any way ἀρχαία with him, but obligation by grasping the hand is an
old symbol, hallowed and made binding by age, κατὰ τὸν ἀρχαῖον τρόπον. In
exactly the same way the victors hang up in the temples the weapons they
have captured, ἀρχαῖον γάνος, χαριστήριον ἐκ παλαιοῦ νομιζόμενον, Aesch. Ag.
579. Presumably ἀρχαῖον στέφος (Aesch. fr. 235, from the Sphinx) means ‘a
crowning or wreathing such as is customary from of old’, cf. besides for the
language and probable sense of the fragment Wilamowitz, Aesch. ed. mai. 71.
γάνος (cf. on 1392) here obviously is what would be called ἄγαλμα in the
technical language of dedications.
580. κλύοντας (Wilamowitz accentuates κλυόντας according to the principle
enunciated by him in the Preface, xxxiii, cf. on 680, 814, 830, but here without
adequate grounds) : the idea that the quasi-inscription spreads vocally (ὑπὲρ
θαλάσσης Kai χθονὸς ποτώμενα) is maintained.
581 f. καὶ χάρις «rA.: ‘and the favour (or grace) of Zeus which has brought
this to pass will be appreciated as it should be’.
τιμήσεται: the only old form of the future passive ; it is not until Aeschylus
that forms in -θήσομαι gradually take their place beside it (cf. Wackernagel,
Syntax, i. 204). For the close relation between χάρις and τιμή cf. on 354.
582. πάντ᾽ ἔχεις λόγον. Similar expressions for closing a speech (πάντ᾽
ἀκήκοας λόγον etc.) in Tragedy are collected in Blaydes’s note on this passage
and in the commentaries on S. 47. 480, to which may now be added Aeschylus’
satyr-play Δικτυουλκοί (Pap. Oxy. 2161, col. τ. 21) ἔλεξα. πάντ᾽ ἔχεις λόγον,
Menander Epitr. 75 (116 Körte, 3rd ed.) εἴρηκα τόν γ᾽ ἐμὸν λόγον. Cf. p. 473 f.
A similar concluding formula is found in Hom. Hymn. (5) Ven. 289 εἴρηταί
TOL πάντα.

583. The translation ‘dass deine Rede mich überwunden hat, leugn’ ich
nicht’ (Nägelsbach) or ‘defeat in argument I do not deny’ (Verrall) is no more
satisfactory than ‘nego, abnuo, recuso’ (Dindorf, Lex. Aesch.) or the classifica-
tion in K. W. Krüger, Griech. Sprachlehre, ii, $ 56. 7 n. 4 (“Verbum der
Äusserung’) and Kühner-Gerth, ii. 72.18 n. 2. The observation that avaive-
σθαι sometimes ‘pudoris ac paenitentiae notionem adiunctam habet’ (Stall-
baum on Plato, Phileb. 57e) was made long ago (see e.g. Schol. Aristid.
Panath. 183. 3, vol. iii, p. 308. 28 Dind. οὐκ ἀναίνεται" ἀντὶ τοῦ οὐκ ἀπαρνεῖται,
οὐκ ἐπαισχύνεται) and has rightly found its way into our dictionaries. The
passage is correctly rendered by Goodwin, ὃ 881: ‘I am not sorry (non piget)
to be overcome by your words.” The same construction and meaning are
found in E. Her. 1235 εὖ δράσας δέ σ᾽ οὐκ ἀναίνομαι (‘I repent not of what
291
line 583 COMMENTARY
I did to thee’, Wilamowitz), Bacch. 251 f. ἀναίνομαι, πάτερ, τὸ γῆρας ὑμῶν
εἰσορῶν νοῦν οὐκ ἔχον (‘I am troubled at the sight . . .' Goodwin, loc. cit.,
cf. Sandys, ad loc.), cf. also 12}. A. 1503 and for the meaning of the verb
Her. 1124.
584. Margoliouth’s conjecture ἤβη' threatens to become the accepted reading
since it has been put into the text by Wecklein (1888), Housman (C.R. ii,
1888, 244), Wilamowitz, and others. Housman formulated the rule: ‘the
infinitive sans article, though it can be the subject of certain verbs, such as
ἐστίν, γίγνεται, and ξυμβαίνει, cannot be the subject of a verb outside this well-
defined class.’ Consequently 584 is supposed to be faulty in the form in which
it is given in the MSS. But Housman’s rule is too narrow. Of one contra-
dictory instance, Ag. 181 ἦλθε σωφρονεῖν, Housman rids himself by means of
a hazardous conjecture; another in the same play, 1364 κατθανεῖν κρατεῖ, has
escaped his notice. The latter passage has a parallel in 5. Ant. 233 f. τέλος γε
μέντοι δεῦρ᾽ ἐνίκησεν μολεῖν σοι. Jebb ad loc. asserts that the infinitive there
and in Hdt. 6. τοι. 2 ἐνίκα μὴ ἐκλιπεῖν τὴν πόλιν and 8. 9. 1 ἐνίκα... πορεύεσθαι
κτλ. should not be regarded as the subject to ἐνίκησε and ἐνέκα, but he does not
produce the shadow of evidence to support this petitio principu. K. W.
Krüger rightly observes on Thuc. 2. 54. 3 ἐνίκησε δὲ ἐπὶ τοῦ παρόντος εἰκότως
λοιμὸν εἰρῆσθαι : ‘the subject is (τὸ) εἰρῆσθαι᾽ (cf. also J. E. Powell on Hdt. 8.
9. 1). Generally speaking, the same seems to be true in these cases as was
remarked by Kaibel, in treating of another type of infinitive, on S. El. 1030:
'since the infinitive was always felt to be an indeclinabie noun, in post-
Homeric poets [and occasionally also in older prose] the article may be put
in or left out as with other nouns.'
εὖ μαθεῖν is hardly correct. In order to make it intelligible Paley says: ‘ “to
learn well’, i.e. good news, a sort of play on the proverb "never too late to
learn" '; on the same lines, only much too freely, Wilamowitz translates :
‘denn an Glück glauben zu lernen wird der Mensch niemals zu alt’, similarly
Wecklein. Perhaps, in view of εὖ λέγειν etc., we should not deny the possibility
of an occasional use of εὖ μαθεῖν in a similar sense. But if we understand it in
that sense here, we introduce into the passage an inappropriate idea. 'I grant
I was wrong with my doubts': after that it is impossible to go on ‘one is never
too old to hear good news’ (or “accept good news’); we must have ‘one is
never too old to be teachable'. That requires εὐμαθεῖν, as Headlam has seen
(he refers to Cho. 225 δυσμαθεῖν), formed of course from εὐμαθής ('praeceptum
Scaligeri’, cf. Lobeck, Phrynichus 266 and 560 ff.). That εὐμαθής does not
occur in the active meaning before Plato (passively it is found in A. Eum.
442) may be due to chance; besides, the same is true of δυσμαθής: in the
passive sense Ag. 1255, actively first in Plato; yet Aeschylus has δυσμαθεῖν'
(used transitively).
It is better not to mix up the sentence with recollections of γηράσκω δ᾽ αἰεὶ
«TA., with which it is connected only in the most general way. It is a powerful
and striking thought that whereas a man grows older, his capacity for
learning does not do the same but keeps itself perpetually young.
585. ταῦτα: the Herald's tidings.
Κλυταιμήστραι: now the coryphaeus has caught sight of the queen, who is
coming out of the house.
1 E, Wunder, De Aesch. Agamemnone (Lipsiae 1857), 30 f., read ἦβαι,
292
 THE PURPOSE OF THE HERALD'S SPEECHES
586. μάλιστα: so placed that it supports the qualifying clause av . . . ἐμέ.
ταῦτα is the subject of this clause too.
It is difficult to decide the question whether σύν should be taken as an
adverb or as a prefix separated from its verb by ‘tmesis’, both here and in
many similar cases. In general see Wilamowitz on E. Her. 53 and especially
Wackernagel, Syntax, ii. 170 ff., particularly p. 170 concerning Attic Greek of
the fifth century; for the long-maintained independence of ovv even in appa-
rent compounds, see Wilamowitz on E. Her. 832, Wackernagel, 177, as well -
as the careful handling of the subject in L-S s.v. σύν C.

Although the Herald remains on the stage, and will deliver another long
speech in the scene that follows Clytemnestra's exit, this seems to be a
suitable place for attempting a brief appreciation of his part in the οἰκονομία
of the play.
Far from being a ‘messenger’, a mere instrument of communication without
a personality of his own, the Herald is a living and well-rounded figure.! Like
the Watchman, he is a plain man, a representative of the common people, but,
unlike the other, not harassed by evil forebodings and a feeling of approach-
ing doom. On the contrary, he is the only character in this tragedy who dis-
plays an unqualified optimism. His cheerful attitude contrasts sharply with
the overcast mood of the Elders and the weary restraint of the king. He is
capable of summing up a series of grave events in these words (573 f.) : ἡμῖν
δὲ τοῖς λοιποῖσιν Ἀργείων στρατοῦ vırdı TO κέρδος, πῆμα δ᾽ οὐκ ἀντιρρέπεϊ,
He thoroughly enjoys being alive and safely back after so many toils and
perils. Blissfully ignorant of what has happened in Agamemnon’s house, he
does not in the least catch the meaning of the dark hints of the coryphaeus:
εὖ γὰρ πέπρακται (551).
The appearance of such a character upon the stage has the effect of a lull
before the breaking of the storm. At no later moment would there be room
‚for a breathing-space, for with the return of Agamemnon we are irrevocably
on the way to the catastrophe. This was probably one of the considerations
that induced the poet to elaborate the scenes of the Herald to such length and
thus considerably delay the arrival of the king. Another object, achieved by
the same device, seems to be of still greater importance. The first scene of
the Herald (503-82) serves both as a supplement and as a foil to the scene of
the king.
Agamemnon is visible only during one long scene. The play called after
him is well advanced when at last he enters, and he disappears long before
the end. But that one scene is the centre of the tragedy: not only is it placed
in the middle of the whole fabric but it is also as it were the centre of gravity.
Its density its enormous. The more we study it, the more we realize that here
the poet has concentrated all his creative power on one objective. Nothing
was to be admitted that would not accentuate some essential feature of the
great central figure, the king, and have a bearing upon his tragic fate. Any
accidental detail, anything that might divert attention from the main issue,
had to be excluded. Supposing now the king entered without being preceded
by the Herald, he would be the only representative of the returning army. In
. that case it would have been almost inevitable that the common experience
1 The difference is well emphasized by Wilamowitz, Interpr. 160 f.

293
 COMMENTARY
which was so near to the hearts of Aeschylus and his countrymen, home-
coming after a long absence, warfare in remote lands, sea-voyage, and so
forth, should at any rate to a certain extent find expression in the speeches of
the first man to return, the king. The concentration on Agamemnon and
what is relevant to him would then have been impossible. As it is, the poet
has employed the Herald not only to delay the beginning of the catastrophe,
but above all to relieve the speeches of the king of all those elements which,
though important and exciting in themselves, are less material to his own
character and fate.
In the talk of the plain man, the Herald, there is full opportunity for such
details as were only too familiar to the generations which had fought at
Marathon and Salamis and Plataea and Mycale, in Thrace and Pamphylia,
and recently in Cyprus and Egypt. It would, however, be wrong to suppose
that the poet, singula dum captus circumvectatur amore, loses sight of the
οἰκονομία of his play. The fate of Agamemnon is partly rooted in the Trojan
War, the greatness and the horrors of which are always present in this
tragedy. So the speeches of the Herald round off the picture of the war that
is drawn in the songs of the Chorus and the imaginative descriptions of the
queen. His final report (650-80), with its fine borrowing from the Nóoro:,
answers the question which could hardly be avoided in a play having as its
theme the murder of Agamemnon: ποῦ MevéAaos ἔην; At the same time
this speech prepares the audience for the satyr-play, Πρωτεύς.
But the first speech of the Herald is a foil as well as a supplement to the
great oration which Agamemnon delivers on entering the stage. The two
opening prayers are parallel; so is the almost harsh use of juridical technicali-
ties: the Herald employing terms of criminal law (534-8), and the king
describing the high court of divine justice with details borrowed from Attic
procedure (813 ff.). By this parallelism the profound contrast in temper and
manners, in education and discretion is made the more striking. The Herald
is jubilant with wild joy (cf. 503 ἰὼ πατρῶιον οὖδας and 518, also the outburst
508); the king is a model of perfect restraint. His reserve may look like
coldness and has indeed often been mistaken for it. His attitude is dictated
by the strictest εὐσχημοσύνη such as befits a great gentleman. The maxims
with which the Herald proudly adorns his talk are all commonplace ; Agamem-
non's reflections on the ways of men are the fruit of mature wisdom. There is,
however, much more in the contrast of the two characters than the difference
of manners in men belonging to different social classes. Every word that
Agamemnon speaks seems to be weighed down by the burden of fate which
has been on his shoulders for so many years and also by the feeling of hostility
and fresh dangers facing him at home. No such thoughts are likely to cross the
Herald's mind. As far as he is concerned, all that matters is the unhoped-for
happiness of the present hour. There are no complications about him, no
hidden uncertainties. Representing a simple type of man, he is convincing
and pleasant enough so long as he is there. He could no more be regarded
as a mere piece of the dramatic machinery than could the Watchman of the
prologue ; he is much too lifelike for that. All the same the whole character is
meant also to set off the much greater figure of the king.
1 Cf. Boeckh, Graecae tragoediae principum, Aesch., Soph., Eur., num ea quae supersunt,
etc. (1808), 268.

294
 COMMENTARY line 594
587. μέν: cf. on 1. Denniston, Particles, 383, rightly rejects the attempts to
find here in what follows a particle to pick it up.
589. Ἰλίου: the attributive genitive used ἀπὸ κοινοῦ exactly as in Prom. τοῖς,
E. Her. 1091, Or. 406. We may compare also Eum. 9 λιπὼν δὲ λίμνην Andiar τε
χοιράδα, where AnAtav belongs to λίμνην as well.
590. ns: see on 369.
ἐνίπτων : this is the only occurrence of this Homeric word in drama, and
therefore to be regarded as inspired by contexts like {2 768 εἴ τίς με καὶ ἄλλος
ἐνὶ μεγάροισιν ἐνίπτοι.
"That the speech begins in the middle of the line . . . is against the usage of
Epic’ (Wilamowitz, Hellenist. Dicht. ii. 53 n. 2), cf. above on 126. The same
in Pers. 402, Prom. 647, Cho. 130. So Tragedy is more free in this respect.
φρυκτώρων διά. Stanley: ‘ab excubitoribus persuasa’, Schiitz: ‘facibus
accensis persuasa’. The latter is without doubt more vigorous (cf. 475, 480).
and consequently has been widely followed; Paley, however, explicitly
objects: ‘ φρυκτωροί are “‘beacon-watchers”.’ διά with the genitive used
of a thing is, of course, normal, but the preposition would probably be
no less correct if we took φρυκτωροΐ as persons, since these beacon-posts
function as messengers, and since διά with the genitive in the instrumental
sense is found especially with ἑρμηνεύς, ἄγγελος (cf. Stevens, C.R. 1, 1936, 162).
The dictionaries follow Estienne’s Thesaurus and Stanley’s translation, as
they often do, so Passow, Dindorf’s Thesaurus, L-S, and the Aeschylus-
lexicons of Wellauer, Linwood, Dindorf. Estienne was influenced by Thuc.
8. 102. 1 ὡς αὐτοῖς of re φρυκτωροὶ ἐσήμαινον καὶ ἠισθάνοντο τὰ πυρὰ...
φανέντα, where Budé gave the translation excubitores. On the other hand,
there is no doubt that Lycophron, who took so much material from tragedy,
used the word for a fire-signal in 345 (ὅταν) λάμψηι κακὸν φρύκτωρον. But there
L-S tell us that in the MSS the accentuation is φρύκτωρον, whereas they
quote Ag. 590 without question under φρυκτωρός, although the MSS give
φρυκτώρων. I do not know which was the original accentuation (real or
apparent similar changes of accent are mentioned in Chandler, Greek Accentua-
tion, and ed., 137, ὃ 455). But I am sure that here, as in Lycophron, the word
means the fire-signal, not the fire-watcher. Provisionally I adhere to the
accentuation of the MSS; possibly new material may sooner or later make a
decision possible. [Cf. the Addenda. ]
592. The use of πρός with the genitive to describe that ‘which lies in the
nature of a person (or a thing)’ (Wilamowitz, on E. Her. 585, cf. Matthia,
Griech. Gramm. § 316 d note) is not attested before the Oresteia.
κέαρ: plainly referring to 481 καρδίαν. The utterances at the end of that
choral song are typical of the unfriendly remarks of the citizens, such as had
come to the ears of the queen also. For the police-court methods applied to
this passage by Verrall see his Introduction, liii n. 2.
593. τοιούτοις: for the prosody see on 1256 f.
πλαγκτός: for the form of the feminine see on 287. ‘Such talk made me
appear as one astray’ (for the construction cf. Goodwin, § 914, 5).
594. ἔθυον: Stanley, followed almost universally, took this as singular.
Hermann regards it as plural; this is not absolutely impossible, but would
make the passage very harsh, since it would force us to understand an
1 φυκτωροὶ C: φρυκτώριοι AEF.

295
line 594 COMMENTARY

unexpressed but necessary preliminary clause (‘I refused to be diverted from

my purpose but brought thank-offerings, and the others followed my ex-
ample’) ; besides, the end of the sentence (596 f.) ev θεῶν ἕδραις... φλόγα would
become rather lame if ἔθυον had been said of the citizens before. To take the
imperfect in the full imperfect meaning (‘but still I went on sacrificing’
Paley, and Headlam ‘yet Icontinued offering’) seems attractive, but is possibly
not justifiable: see in general the emphatic warnings given by Wackernagel,
Syntax, i. 182 f., and in particular for the imperfect in an adversative clause
after a preceding aorist e.g. Pers. 372 ff. τοσαῦτ᾽ ἔλεξε... οἱ δὲ... δεῖπνόν τ᾽
ἐπορσύνοντο, ναυβάτης τ᾽ ἀνὴρ Tpomoüro κώπην, Prom. 656 ff. ἔστε δὴ πατρὶ ἔτλην
γεγωνεῖν... ὁ dé... ἴαλλεν.
γυναικείωι νόμωι: the meaning which νόμος, in itself equivocal, has here
is not at once clear. We must first of all exclude the idea of the musical term
(‘in a feminine strain’ Paley) since ὁλολύζειν cannot be described as a tune or
a melody, however much one extends the idea ; no one who notes {πὸ context
of the passage will quote Ar. Birds 222 θεία μακάρων ὀλολυγή in support of this.
The only two meanings that come into question are ‘iussu’ (Casaubon, ‘by
female law’ Kennedy) and ‘more muliebri’ (Stanley, ‘after woman’s use’
Headlam). The second is in every way conformable to regular usage (for
νόμος = ὃ νομίζεται in general cf., e.g., Wilamowitz, Kl. Schr. v. i. 120, and
Glaube d. Hell. ii. 86), and in particular it is recommended in the modal
dative νόμωι by parallels like 1207 (νόμωι = ὡς νομίζεται), Suppl. 220 τοῖσιν
᾿Ἑλλήνων νόμοις (Ar. Ach. 773 'EMávov νόμωι), cf. also Cho. 424. It is well
known that ὀλολύζειν is predominantly, if not exclusively, used of women’s
cries (instances! in Stengel, Kultusaltert. 112; Dittenberger, Syll. 982 n. 19;
Neil on Ar. Knights 616; C. Theander, Eranos, xv, 1915, 114, who also takes
Ag. 594 γυναικείωι νόμωι rightly ; and in L. Deubner's article, [quoted on p. 297,
n. 1], p. 5£). Here too it is probably the case that the oAoÀvyuós is chiefly
raised by the women, but in their excitement at the approaching homecoming
of their sons and grandsons the old men join in. Clytemnestra scoffs at those
who were foremost in mocking conduct which was πρὸς γυναικός, and shortly
afterwards have followed their queen's example and indulged in a feminine
habit. Besides, "iussu mulieris' would not really suit the sense of the passage ;
the idea here is not that Clytemnestra has definitely ordered something to be
done, but that the citizens have followed her example of their own accord
(cf. on 485). That only increases the triumph of the queen.
595. ἄλλος ἄλλοθεν: as earlier (92 ἄλλη δ᾽ ἄλλοθεν) sacrifices have been lit
everywhere in and in front of the house of the Atridae, so now everywhere in
the city.
597. θυηφάγον occurs nowhere else.
κοιμῶντες early provoked objection: starting with Casaubon editors have
continually tried new conjectures, none of which is convincing. θοινῶντες
suggested by Emperius is pretty in itself (cf. alere ignem, flammam ; the idea
would suit ll. 92-6 well; for the occasional use of 8owáv in the active even in
Attic cf. Wilamowitz on E. Jon 982). κοιμῶντες is in fact strange, for, as the
sense here must be in general 'in spite of all I made offerings, and the people
of the city followed my example', we do not expect to find the latter process
ı There can now be added Alcaeus, Pap. Oxy. 2165, col. 2, fr. 1. 26 ff. περὶ δὲ βρέμει
ἄχω θεσπεσία yuvalxwv ipas ὀλολύγας ἐνιαυσίας,
296
 COMMENTARY line 597

described by a word that refers specifically to the conclusion of the sacrifice.

Over and above this there is the doubt raised by the fact that the oAoAvyr,
the θυστὰς βοή (Sept. 269), in the ordinary course follows after the end of the
opening prayer (but before the prayer e.g. Hom. Z 301), and at the beginning
of the actual sacrifice: this is plain e.g. in a Homeric description (y 450)
which goes thoroughly into the details of the ritual; it is also presupposed in
the allusive brevity of E. Or. 1137 ὀλολυγμὸς ἔσται, πῦρ 7’ ἀνάψουσιν θεοῖς. Must
we then look on κοιμῶντες as corrupt? That would imply the highly improb-
able assumption that a technical and by no means hackneyed term has been
intruded by the chance of a corruption : we can scarcely imagine an intentional
‘interpolation’. Blomfield has quoted the line of the comedian Phrynichus'
Movörpomos (fr. 24 K.), ἔπειτ᾽ ἐπειδὰν τὸν λύχνον κατακοιμίσηι, where the verb
may well be a novel quip. Now our fuller text of Pollux (7. 178) gives further
evidence for the original idiom κοιμῆσαι : καὶ τὸ κατασβέσαι τὸν λύχνον ὅπερ oi
νῦν, κοιμῆσαι.2 καὶ κατακοιμίσαι Φρύνιχος εἴρηκεν ἐν Μονοτρόπωι κτλ. To this
must be added the passages collected by van Heusde, from Meleager of
Gadara, Anth. Pal. xii. 49. 1 f. σοῦ $Aóya . . . koysdoet . . . Bpouos, Plutarch, and
Nonnus 3. 43 Lapins . . . ἀκοιμήτου φλόγα πεύκης. Consequently if κοιμῶντες
here is sound, the passage can only mean: ‘lulling in the shrines the spice-fed
odorous flame’ (Headlam). How this ‘lulling’ was managed, the poet refrains
from telling us: possibly it was done, as Paley guesses, by pouring wine over
the ashes. Concerning the difficulty mentioned above, that according to the
MS reading the ὀλολύζειν comes at the end of the sacrifice, Paley remarks: ‘the
custom alluded to, of raising a shout at that particular point of the ceremony,
though probable in itself, does not appear to be known from other sources.’
But the passage quoted by Paley himself (Xen. Anab. 4. 3. 18 f. καὶ ot μάντεις
ἐσφαγιάζοντο eis τὸν ποταμόν... ἐπεὶ δὲ καλὰ ἦν τὰ σφάγια, ἐπαιάνιζον πάντες
οἱ στρατιῶται καὶ ἀνηλάλαζον, συνωλόλυζον δὲ καὶ αἱ γυναῖκες ἅπασαι) at least
suggests that the ὀλολυγμός occasionally followed later in the proceedings.
It is all the easier to imagine this happening in the case of bloodless sacrifices
such as those mentioned by Clytemnestra. However that may be, it is not
difficult to understand why here the ὀλολυγμός is singled out for special
mention. Clytemnestra could not convince herself with her own eyes of the
completion of the sacrifice in the various households, because she has been
the whole time in or in front of her house; but she cannot avoid hearing the
ritual cries which, now here, now there (ἄλλος ἄλλοθεν), rise from the open
courtyards of the city.
1 Cf. 5. Eitrem, ‘Beitr. z. griech. Religionsgesch. III’, Vid.-Selsk. Skrifter 1 Kristiania,
ii, Hist.-Filos. Klasse, 1919, No. 2. 45: ‘it is only at the moment when the sacrificial victim
receives the stroke (and not at the σφάττειν, which does not take place till L 454), at the
culminating point of the whole ceremony, that the women’s ὀλολυγή rings out. This accords
precisely with the ritual of the sacrifice as we know it in historical times.’ Eitrem provides
ample information on the ὀλολυγμός, but neither his discussion of the subject nor the
contributions made by others (cf. above on 594) are mentioned in the article ‘Ololyge’, RE
xvii. 2493 f., which can now be dismissed altogether after the publication of L. Deubner’s
paper, ‘Ololyge und Verwandtes’, Abhandl. d. Preuss. Akad. d. Wissensch., Phil.-hist. KI.
1941, No. 1.
2 This is my way of punctuating, which diverges from that of Bethe. κοιμῆσαι scil. of
"Arrixot ἐκάλουν.
3 It seems that in Statius, Theb. 6. 235 f. instant flammis multoque soporant imbre rogum
the meaning of soporare was borrowed from κοιμᾶν (‘Bedeutungslehnwort’).

297
line 597 COMMENTARY

θυηφάγον... εὐώδη: it would be artificial to take one of the adjectives as

predicative: both are placed together on an equal footing.
598. The epic comparative μάσσων (borrowed from Epic by Pindar) occurs
several times in Aeschylus, otherwise not in drama.
“καὶ ri δεῖ λέγειν; An Attic formula when something is suppressed which it
is superfluous or disagreeable to add’ Paley on Eum. 826, with instances. Add
E. Phoen. 761 ri δεῖ μακρηγορεῖν; Thuc. 1. 68. 3 νῦν δὲ τί δεῖ μακρηγορεῖν; 1. 73.
2; I. 123. 1; 4. 59. 2; Pl. Phaedr. 241 e καὶ τί δεῖ μακροῦ λόγου; Menex.
244 À καὶ μηκύνειν μὲν τί δεῖ; Aeschines 3. 141 καὶ τί δεῖ τὰ πλείω λέγειν;
Antiphanes fr. 33. 5 K. τί μακρὰ δεῖ λέγειν; In τὰ μάσσω the article is used as
in the passage from Aeschines and at Demosth. 18. 152 ri δεῖ ra πολλὰ λέγειν;
(Wilamowitz, Griech. Lesebuch, Erläut. p. 48, comments on it: ᾿πολλὰ λέγειν
would imply the sense ‘‘to talk what is not necessary”, ra πολλά = ra πολλὰ
ἐνόντα); Wecklein compares S. Phil. 576 μή νύν μ᾽ ἔρηι τὰ πλείονα, Oed. C. 36
πρίν νυν τὰ πλείον᾽ ἱστορεῖν.
τί δεῖ κτλ. is preceded by two short cola, (1) καὶ νῦν, (2) τὰ μάσσω μέν (to
which δέ 600 answers). The antithetical arrangement of μέν... δέ has here,
as often, the result that the interrogative pronoun is forced to the beginning
of the following colon, see on 1020. Here the effect of the first place being
given to τὰ μάσσω μέν is to produce something like a ‘headline’. Verrall
appositely translates: ‘and now, for the fuller tale, what need... .’.
Many editors lay a particular emphasis on σε by writing od μοι with
Wieseler. I think this upsets the balance of the sentence. The emphasis lies
on νῦν and τὰ μάσσω: for the present moment the brief summary of the news
suffices.
599. πάντα πεύσομαι Adyov: this phrase too (cf. on 598) belongs to colloquial
language, cf. Cratinus fr. 144. 2 K. καὶ πάντα λόγον τάχα πεύσει, Hát. 1. 111. 5
πυνθάνομαι τὸν πάντα λόγον, 4. 152. I, 9. 9. 1 ἐπύθετο πάντα λόγον, 9. 13. 2 πυθό-
μενος πάντα λόγον, 9. 94. 3 πυθόμενος τὸν πάντα λόγον etc. (cf. Stein on 1. 21. 1);
A. Pers. 246 τάχ᾽ εἴσηι πάντα νημερτῆ λόγον, S. Trach. 484 πάντ᾽ ἐπίστασαι λόγον,
Kritias Πειρίθοος (B 16. 5 Diels) οὐδεὶς ὄκνος πάντ᾽ ἐκκαλύψασθαι λόγον. The
phrase πάντα λόγον is rendered in a higher style by Callimachus Epigr. 1. 8
Wil. keivoi τοι πᾶν ἐρέουσιν ἔπος, Aitia fr. 9. 39 Pf.7 δ᾽ av’ Eros πᾶν ἐκάλυψεν ἔπος.
600 f. For the various ways of taking ὅπως see Paley. Hermann rightly says:
‘non dicit σπεύσω δέξασθαι ὅπως ἄριστα, sed ὅπως σπεύσω δέξασθαι ἄριστα.
V. Porsonum ad Hee. v. 398.’ Similarly Goodwin, §§ 272 and 275, but neither
there nor in Kühner-Gerth ii. 376 f. has full use been made of the instances
which Porson collected as evidence for ὅπως with the first person of the future
indicative.
αἰδοῖος here used in the Homeric sense. In Aeschylus the adjective occurs
eight times (including Ag. 1657, where it is a certain restoration) ; it is foreign
to living Attic (Plato uses it only in quotations from poets) and up to the present
not attested for Sophocles or Euripides; Aristophanes does not use the verb
αἰδοῦμαι or its compounds either (for the apparent exception Clouds 1468 see
Gnomon, ix, 1933, 615).
601 ff. τί γὰρ... ἥδιον «rA.: this form of question using a comparative
to express an emphatic superlative is common, cf., e.g., Pindar’s (fr. 89a
Schroeder) beginning of a procession-song Τί κάλλιον ἀρχομένοις 7 kararavo-
μένοισιν ἣ βαθύζωνόν τε Aarw καὶ θοᾶν ἵππων ἐλάτειραν ἀεῖσαι; Aesch. fr.
208
 COMMENTARY line 605

309 N. ri γὰρ ὄψον γένοιτ᾽ ἂν ἀνδρὶ τοῦδε βέλτερον; E. Alc. 879 f. (see below),
Horace, Epist. x. 4. 8 quid voveat dulci nutricula maius alumno? Cf. on
899-902. As regards the apparent omission of 7 before 603 f., Blomfield was
right in comparing this instance with E. Alc. 879 f. τί yàp ἀνδρὶ κακὸν μεῖζον,
ἁμαρτεῖν πιστῆς ἀλόχου; though he was not justified in speaking of 'ellipsis
particulae 7’, as was pointed out by Stallbaum on Plato, Gorg. 519 d and by
Hermann on this passage of the Agamemnon. In the Agamemnon passage
the infinitive is in apposition to τούτου, exactly as in Plato, Gorg. 519 d (καὶ
τούτου λόγου τί ἂν ἀλογώτερον εἴη πρᾶγμα, ἀνθρώπους . . . ἀδικεῖν... .;) the
infinitive is in apposition to τούτου τοῦ λόγου. But even if the infinitive is not
supported by a preceding genitive of comparison, the expression ‘what greater
evil is there?’ means much the same as ‘the greatest evil is . . .', so that it
can be naturally followed by an infinitive without 7: see the Alcestis passage
quoted above. For Latin parallels see Lófstedt, Syntactica, ii. 167 ff.
602. φέγγος: cf. 504, 1577.
603. ἀπὸ στρατείας : ‘revertentem ab expeditione', Hermann, who compares
Eum. 631.
Blomfield's conjecture avöpi! has been refuted by Hermann: the words
from ἀπὸ στρατείας to θεοῦ form one kolon, and from the context it is readily
understood for whom the act of πύλας ἀνοῖξαι is to take place.
605. 'Scribendum videtur ἥκειν δ᾽... Nam ταῦτ᾽ ἀπάγγειλον πόσει, sic sine
particula coniunctiva positum, ad antecedentia referendum est', Weil appro-
priately remarks (Addenda to Eum., p. 132). There was no need for him to
base his statement about the meaning of ταῦτ᾽ ἀπάγγειλον only on the
asyndeton. For the present purpose it is sufficient to examine the instances
in which Aeschylus and Sophocles use demonstrative pronouns in a similar
context. τοῦτο Or ταῦτα, as is well known, very often refers back to what has
been mentioned before, but is never found in the two poets as a prefatory
formula in the sense of 'as follows', in which sense τάδε etc. is normal (see on
615). For obviously passages like Prom. 377 οὔκουν... τοῦτο γιγνώσκεις ὅτι
krÀ., Pers. 333 f. φράσον μοι τοῦτο... πόσον δὲ πλῆθος ἦν νεῶν ‘EAAnvidwr, Cho.
933 f. τοῦθ᾽ ὅμως αἱρούμεθα, ὀφθαλμὸν οἴκων μὴ πανώλεθρον πεσεῖν and the like
are not comparable with the present passage: in them and e.g. S. 47. 1370,
Ant. 61, 188, Oed. R. 584, 1512, El. 591, Phil. 77, 232, and elsewhere (cf. also
Pind. P. 2. 21 ff.) there is always an appositional relation between the pre-
paratory τοῦτο and the immediately following simple clause; similarly in
Cho. 93 τοῦτο roûros serves to support the following infinitive of command.
The same kind of thing holds good even where exceptionally (cf. the passage
quoted above Pers. 333 f.) the clause to which τοῦτο looks forward appears in
the form of an independent main sentence: 5. Phil. 435 ff. λόγωι δέ σ᾽ ev
βραχεῖ τοῦτ᾽ ἐκδιδάξω" πόλεμος οὐδέν᾽ ἄνδρ᾽ ἑκὼν αἱρεῖ πονηρόν κτλ. All this is
different from a ταῦτα supposed to introduce a real oratio recta (as τόδε
205, 409, and often elsewhere) or an oratio obliqua, as almost all editors and
translators have assumed, since Stanley, in the case of Ag. 604. There is no
parallel for this in Aeschylus or Sophocles. It is a question here of the
peculiarity of a strictly defined usage in speech, where general assertions,
though in themselves valid (e.g. Jebb on Oed. C. 987 f.: ‘In poetry οὗτος often
refers to what follows . . . and ὅδε to what has just preceded’), are irrelevant.
1 Sidgwick and Murray put it into the text without mentioning that it is a conjecture.
299
line 605 COMMENTARY
Nor do I believe it is possible to justify an isolated instance of a substantival
ταῦτα (= ‘the following’) before a speech by pointing to the fact that Aeschylus
(Sophocles offers nothing similar, Euripides has El. 329) has once used τοῦτ᾽
ἔπος (Eum. sıo) as introduction to a short exclamation, and perhaps once,
Sept. 579, as introduction to a speech (the line is likely to be interpolated, cf.
P. 756 n. 1). Therefore Weil is right ; ταῦτ᾽ ἀπάγγειλον πόσει forms the con-
clusion of the preceding section. Consequently in 605 δ᾽ must necessarily be
inserted. What precedes she has said to the Herald without describing it as a
message or commission; at the end she adds: 'tell him that!' and goes
on immediately with vigour, as if what follows had just occurred to her in an
afterthought : ‘and he had better hurry about being here’ etc. (I am at a loss
to understand what Wilamowitz means when he says of 606 ff.: 'Agame-
mnoni haec nuntiari omnino non volt Clytaemestra’.)
Kennedy's assertion that “ἥκειν ὅπως τάχιστα is sheer nonsense’ is rash.
Cf. e.g. Ατ. Peace 275 ἧκέ νυν ταχύ, Lys. 924 ἧκέ vuv ταχέως πάνυ, and especially
Frogs 1507 ff. καὶ φράζ᾽ αὐτοῖς ταχέως ἥκειν ὡς ἐμὲ δευρὶ kai μὴ μέλλειν. Cf. the
note on 407 about βέβηκεν ταχύς.
ἐράσμιον : the only occurrence of the word in Attic before Plato. The limits
of the currency of this rare word cannot now be quite certainly determined,
but it seems to be at home only in Ionic and there, though it in no way has
an indecent sense, originally in the sphere of the erotic in particular : Semon-
ides fr. 7. 52 D. κείνηι γὰρ οὔ τι καλὸν οὐδ᾽ ἐπίμερον πρόσεστιν οὐδὲ τερπνὸν οὐδ᾽
ἐράσμιον (he goes on to describe her lewdness), Anacr. fr. τ8 D. (in connexion
with dancing) ἐρασμίην τρέψας θυμὸν ἐς ἤβην. Plato three times uses the
superlative in phrases which are characteristic of his glorification of &pws:
Rep. 3. 402 ἃ kai μὴν τό γε κάλλιστον ἐρασμιώτατον, Phaedr. 250 ἃ, speaking of
things which cannot be perceived by the senses, φρόνησις... καὶ τἄλλα ὅσα
epaora (nota bene!). νῦν δὲ κάλλος μόνον ταύτην ἔσχε μοῖραν, ὥστ᾽ ἐκφανέστατον
εἶναι καὶ ἐρασμιώτατον, Tim. 87 d πάντων θεαμάτων... κάλλιστον καὶ ἐρασμιώ-
τατον, then, obviously under the influence of Plato, Xen. Mem. 3. 10. 3
(within a series of superlatives referring to the soul) ro . . . φιλωκώτατον καὶ
ποθεινότατον καὶ ἐρασμιώτατον and similarly Conv. 8. 36 τῶι τὴν ψυχὴν épa-
σμίωι. Surely it is going beyond the limits of what the sensitive Athenian
society considers proper, when Clytemnestra dares to say of the grown man,
the king and commander-in-chief, that he will be on his homecoming
ἐράσμιος" πόλει. This one word betrays the same measureless hypocrisy that
is shown later (897 ff.) in the long list of laudatory comparisons heaped on
Agamemnon.
606. εὕροι: Hermann, followed by many others, has taken this as optative
of indirect speech. Such an optative would be quite legitimate in dependence
on an historic tense: of that type are all the instances of the ‘independent’
optative in continuation of an oratio obliqua quoted from Attic authors by
K. W. Kriiger, Griech. Sprachlehre, § 54. 6 n. 4, Goodwin, § 675, and Wyse on
Isaeus 8. 22; in Herodotus too (cf. Stein on 7. 3. 3 οὐκ ὦν οὔτε οἰκὸς εἴη) such
continuation-optatives occur in historic sequence. Madvig (Adv. Crit. i. 198)
has emphatically called attention to this limitation of the use of the oblique
optative, and has rejected Hermann’s interpretation as linguistically impos-
1 Headlam shows a delicate appreciation of the sphere to which the word properly
belongs when he translates ‘the country’s darling’. I have adopted his rendering.
300
 COMMENTARY line 606

sible: on this point his violent critic Cobet fully agreed with him (Var. lect.
znd ed., 404). The definite establishment of the grammatical rule has pre-
vented later editors (from Paley to Wecklein, Verrall, Headlam, and Wila-
mowitz) from giving any consideration here to an oblique optative. Is it then
necessary to put in the text the conjecture made by Schütz, and indepen-
dently by Cobet, eöpjoes? Does not the simple way out remain of understand-
ing εὕροι as expressing a wish? A number of critics had long taken it as
utinam inventat, ‘quo’, according to Hermann's judgement, ‘sententia paene
ridicula evadit’. This severe rejection is justified if we understand the pas-
sage to mean: ‘and may he find a faithful wife at home’, since that creates the
impression that Clytemnestra is talking of her faithfulness in the weakened
form of a wish and not as an established fact as she must according to her
whole attitude and as she does in the continuation of her speech with par-
ticularly high-flown (613 κόμπος) definiteness. With psychological refine-
ments (‘she avoids, as in itself suspicious, the direct assertion’ Paley, ‘observe
the terrible irony of the wish, sent as a loving message to Agamemnon’
Sidgwick) we get no further. The difficulty is removed as soon as we re-
member that εὑρεῖν is as equivocal as ‘find’, that it means not only ‘light upon’
but also ‘discover’, i.e. ‘convince oneself that something is so’. Only the
latter sense is satisfactory in passages like A. Suppl. 952 f. ἀλλ᾽ ἄρσενάς τοι
τῆσδε γῆς οἰκήτορας εὑρήσετε, Pers. 472 ff. (here I must quote at greater length
to make the point clear) ὦ στυγνὲ δαῖμον, ὡς dp’ ἔψευσας φρενῶν Πέρσας"
πικρὰν! δὲ παῖς ἐμὸς τιμωρίαν κλεινῶν Ἀθηνῶν ηὗρε. In both cases it is a question
of εὑρεῖν being used for the discovery of a thing quite unexpected and dis-
illusioning, in both cases the adjectives (ἄρσενας, πικράν) must be taken
predicatively. πιστήν, similarly, is predicative here, and therefore means not
‘come to find in his horne a faithful wife’ (Headlam) but ‘come to find in his
home the wife faithful’ (Verrall) ; this is supported by the amplifying addition
οἵανπερ οὖν ἔλειπε. It should not be objected that, in the two passages quoted,
ἄρσενας and πικράν are placed emphatically at the beginning of the sentence,
while the words here are γυναῖκα πιστὴν δὲ κτλ. The structure of the message
from 605 on is clear. ‘Tell him to come as quickly as possible. But so far as
his wife is concerned, I hope he may .. .'. γυναῖκα forms the keyword for all
‘that follows, and the word is placed in front in the well-known manner as a
sort of ‘headline’. Finally we must meet the objection that the position of δέ
is an argument for the attributive character of πιστήν. To rebut this, it is
sufficient to call attention to Sept. 41 αὐτὸς κατόπτης δ᾽ εἴμ᾽ ἐγὼ τῶν mpa-
γμάτων, where Wecklein gives the note: “δέ in the third place, without the
two preceding words forming a single notion’. With this interpretation the

ı We have herean instance of what might be called a common formula of emotion (the
hyperbaton contributes to the effect), cf. Prom. 739 f. πικροῦ δ᾽ Exupoas ὦ κόρῃ τῶν σῶν
γάμων μνηστῆρος, S. Aj. 1239 f. πικροὺς ἔοιγμεν τῶν ᾿Αχιλλείων ὅπλων ἀγῶνας ᾿Αργείοισι κηρῦξαι
τότε, El. 470 f. πικρὰν δοκῶ με πεῖραν τήνδε τολμήσειν ἔτι, E. Alc. 257 f. οἴμοι, πικράν γε τήνδε
μοι ναυκληρίαν ἔλεξας, Med. 300 πικροὺς δ᾽ ἐγώ σφιν καὶ λυγροὺς θήσω γάμους, Ar. Birds 1045
πικροὺς ἐγώ σοι τήμερον δείξω νόμους, 1468 πικρὰν τάχ᾽ ὄψει στρεψοδικοπανουργίαν, Thesm. 853
πικρὰν ᾿Ελένην ὄψει τάχ᾽, εἰ μὴ κτλ., Eubulus fr. 120. 6 (ii. 207 Kock) πικρὰν στρατείαν δ᾽ εἶδον
(in the last two passages the expression is nearer to plain language, so too e.g. E. Bacch.
357 πικρὰν βάκχευσιν ἐν Θήβαις ἰδών). Cf. Cobet, Variae lectiones, 573: ‘saepe solet πικρός in
iracundi et minaci oratione poni de ea re quae magnum aliquod malum et infortunium alicui
allatura esse dicatur.’
301
line 606 COMMENTARY
optative retains its full meaning. ‘But for his wife—may he come and find
her to be faithful, as he left her.’ She speaks of her ‘miorÿr εἶναι᾽ not as
something uncertain, and does not avoid the direct assertion (Paley), but
implies that the εὑρεῖν that she has in mind, the discovery, can only be made
by Agamemnon when he has come back home again: the wish that she
professedly expresses is that he may be soon in a position to do so.
607. οἵανπερ οὖν ἔλειπε: Denniston, Particles, 421, draws attention to the
fact that the use of περ οὖν in relative clauses is commonest in Aeschylus and
Plato and serves particularly ‘to stress the correspondence between idea and
fact, the objective reality of something which in the main clause is merely
supposed’; the relation between the wish εὕροι and the statement in the
relative clause is of a similar type here.
608. For ἐσθλός used as approximately equivalent to εὔφρων cf. Mazon’s
commentary on Hesiod, Erga 286.
It is perhaps not unnecessary to sound a warning against Hermann’s
interpretation: πολεμίαν τοῖς δύσῴφροσιν ambigue dicit Clytaemnestra, de
Agamemnone cogitans' ; Verrall extends the ‘ambiguity’ to the whole passage
from 606 onwards. This completely upsets the character of this speech:
Clytemnestra is availing herself of the imagery and ideas of a conventional
moral code, and if there is 'dishonesty' at all, it might be seen in the emphatic
protestation of what ought to be a matter of course. The antithesis expressed
in this line is of common occurrence, starting from Solon fr. 1. 5 D. εἶναι δὲ
γλυκὺν ὧδε φίλοισ᾽, ἐχθροῖσι δὲ πικρόν (to which Blaydes refers) ; parallels are
collected in the commentaries on E. Med. 8o9 and in G. Thomson's note on
the present line.
609. ὁμοίαν: the word is susceptible of two meanings. Either: ‘like (cor-
responding to) the loyal attitude which I have just described'. Thus e.g.
Stanley (‘et cetera prorsus similem), Schütz, Kennedy, Headlam (‘and for
the rest, alike at all parts’), Platt. Or, ‘remaining the same, unchanged’
(instances of this use from Hesiod onwards are given in L-S ὅμοιος I. 2).
Thus e.g. Conington, Paley (‘and in all else the same as she ever was’),
Verrall, Mazon ('toujours la méme en tout'). With this second interpretation
this expression would repeat the idea of 607 οἵανπερ οὖν ἔλειπε. I do not regard
this as impossible, but I prefer the first meaning, because it seems to me better
that she should say: ‘I don't want to go into details about all the rest; it is
exactly what would be naturally expected after what I have said.' It is
therefore a concluding formula. And then, as so often happens in Aeschylus,
she adds by way of afterthought a particularly characteristic detail (σημαντή-
piov KrA.).
σημαντήριον: in Tr we read the interlinear gloss σφραγῖδα τῆς πρὸς τὸν
ἄνδρα εὐνῆς. This absurd explanation still haunts us in Schneidewin and even
in Headlam, who after giving examples for the sealing of treasuries and
storehouses makes the surprising remark: 'so here σημαντήριον includes the
seal of chastity'.! Triclinius rightly recognizes the σφραγίς in σημαντήριον, for
which σήμαντρον is found in Herodotus and Euripides with the same meaning,
cf. also σημαντρὶς γῆ. The custom alluded to in this passage requires an
explanation. Paley refers to E. Or. 1108, but, when the words xai δὴ πάντ᾽
ἀποσφραγίζεται are there used of Helen, ‘Helen is setting the seal upon the
I I do not know what Headlam means by referring to Anth. Pal. 5. 217 (or 216).
302
 COMMENTARY line 6:1

property left by her sister, after her children have been condemned to death’
(Wilamowitz on E. Her. 53), and in the same way ‘Lykos [Her. 53] sets a
seal upon the confiscated palace’; so the legal position in these cases is an
entirely different one. The fact that husbands seal up storerooms to keep
their wives out of them is made the subject of humorous complaint in Ar.
Thesm. 421 ff.; however, that may not have been the rule in respectable
households: one can hardly imagine Xenophon’s Ischomachos, for instance,
treating his young wife in this way. But we can see from Plato, Laws 12.
954 a-b that it was the regular thing not only during the master’s absence on
a journey but on other occasions too for part of his property to be kept under
seal, cf. Theophr. Char. 18. 4 and Studniczka, ad loc. (Theophrasts Charak-
tere, herausgegeben von der Philol. Gesellsch. zu Leipzig, 1897, p. 141 f.).
This was no doubt always the case with objects of especial value. In the
household of the Olympian gods this is the practice of Zeus with his lightning :
only his favourite daughter knows where to find the keys to the sealed room
in which it is kept (Eum. 828). Clymene says in Euripides’ Phaethon (fr. 781.
8 ff. N.) κρύψω δέ νιν ξεστοῖσι θαλάμοις, ἔνθ᾽ ἐμῶι κεῖται πόσει χρυσός" μόνη δὲ
κλῆιθρ᾽ ἐγὼ σφραγίζομαι. There is no need to conclude from the present
passage that Clytemnestra had no ‘Schliisselgewalt’ (cf. W. Erdmann, ‘Die
Ehe im alten Griechenland’, Münchener Beiträge zur Papyrusforschung,
20. Heft, 1934, p. 278); she is only concerned with stating that she has left
everything of value undisturbed.
611. The transition from indirect speech to oratio recta is very natural
here. For the old and widespread habit of relapsing into direct speech see
Wackernagel, Unters. zu Homer, 167, E. Mayser, Gramm. der griech. Papyri
aus der Piolemäerzeit, ii. 3, p. 112 f., and the literature quoted by them.
Schütz proposes οὐκ οἶδα, which looks tempting at first sight, for from 606
onwards she has apparently been speaking about her fidelity as a wife, and so
a summarizing asyndetic conclusion seems probable. However, on closer
examination we realize that πιστήν is intended in a quite general sense and
ll. 606-10 are merely concerned with her merits as οἰκοφύλαξ and οἰκονόμος
(this alone should prevent us from referring ompavrnpıov to the ‘seal of
chastity’). It is not till 611 that she passes on to ‘conjugal fidelity’ in a special
sense, and οὐδ᾽ οἶδα is therefore necessary. If the author of the Χριστὸς
πάσχων quotes this line twice (65, 522) with οὐκ οἶδα, the reason presumably is
that he wished to let the Θεοτόκος express in a strongly isolated sentence her
immaculate purity.
οὐδ᾽ ἐπίψογον φάτιν: the old interpretation (‘haud novi voluptatem,
neque inhonestum sermonem ab alio viro magis quam’ eqs., Stanley) has still
found defenders in Conington (though with a certain reserve), Dindorf (Lex.
Aesch. ἐπίψογος), Paley (‘Nor know I either pleasure or slanderous word from
any other man’), and Verrall. But it is not quite easy to take φάτις ἄλλου πρὸς
ἀνδρός in the sense of ‘speech coming from another man’, ‘the fact of being
addressed by one who is not one’s husband’. The meaning which is the
predominant one in Aeschylus for φάτις (fama, rumor) inclines one rather to
understand ἐπέψογον φάτιν as the censorious talk of a third party. This sense
is highly appropriate in the context: Pindar, speaking of Clytemnestra’s
infidelity (P. 11. 26 ff.), says such an ἀμπλάκιον is καλύψαι ἀμάχανον ἀλλοτρίαισι
γλώσσαις κακολόγοι δὲ πολῖται. This points to the way in which the sentence
393
line 611 COMMENTARY
should be understood. Nägelsbach says rightly : “οὐδ᾽ ἐπίψογον φάτιν positum
est διὰ μέσου, cf. 318 [see the note there]. Nam ἄλλου πρὸς ἀνδρός referendum
est ad repıbw’, so also Wecklein, Wilamowitz (translation), Headlam (whose
interpretation is wrongly questioned by G. Thomson in his Supplement, ii.
374). In F and Tr commas (or colons) are put after τέρψιν and after φάτιν so
that the words οὐδ᾽ ἐπίψογον φάτιν are separated from the rest of the sentence.
612. χαλκοῦ Babäs. Here again (cf. on 312 ff.) the brevity of the allusion to
something which was perfectly familiar to the ancient audience makes it
difficult, if not impossible, for the modern reader to understand it exactly.
Schol. (ZxoA. wad. in Tr) ὥσπερ οὐκ οἶδα τὰς βαφὰς τοῦ σιδήρου, οὕτως κτλ.
This is in fact the meaning that suggests itself, for from Homer downwards
βάπτειν regularly denotes the tempering of iron or steel, and similarly βαφή
from S. 47. 651! right down to the late treatise περὶ βαφῆς σιδήρου (cf. Rommel,
RE iii A. 2127). This meaning, however, is only possible if we assume, as the
scholiast does, that Aeschylus has used χαλκοῦ here for σιδήρου, since there is
no process of ‘tempering’ in the case of copper or bronze, cf. Blümner,
Technologie, iv. 334. For this reason Welcker, Nachtr. zur Trilogie, 42 n. 6,
referred the expression to the process of colouring bronze, and was followed
by Klausen and others, and Wilamowitz translates ‘die Kunst das Erz zu
färben’; Wecklein (cf. his Stud. z. Aesch. 112 f.), too, accepts the meaning
of ‘colouring’, but takes ‘colouring of iron’ to mean an ἀδύνατον. This last
can be ruled out, for it would not make sense if Clytemnestra, by way of
emphasizing her statement that she knows nothing at all about intimacy
with any other man, should choose the comparison to an ἀδύνατον, 1.6. to
something which nobody knows how to do. On the contrary she is clearly
putting the knowledge of the τέρψις ἄλλου πρὸς ἀνδρός on the same level as
knowledge possessed by others (for it is a fact that certain other wives have
some knowledge of infidelity) in which she has no share and never can
have any. L. Campbell, Am. Journal of Philol. i, 1880, 434 f., rightly says:
‘The tempering of metal was a mechanic process, known to a class of crafts-
men, and to few or none beyond it ... the last thing therefore which a delicately
nurtured princess could be expected to know. It is much as if a modern fine
lady were to say, “I could no more think of doing such things than of shoeing
a horse".' The inappropriateness of an ἀδύνατον disposes also of the inter-
pretation of Blümner (loc. cit.), who understands βαφάς as ‘tempering’, ‘but
in the sense that in the case of copper this is indicated as something
impossible and non-existent’. On the other hand, Welcker's interpretation
(‘colouring metal’) satisfies the context and appears possible in itself. There
seems unfortunately to be no early evidence for a colouring process, for
with regard to the fragment of the sophist Antiphon (B 40, Diels-Kranz,
Vorsokr. ii, 5th ed., 345) βάψιν χαλκοῦ καὶ σιδήρου Blümner, 334 n. 1, is perhaps
too optimistic, since the list in Pollux 7. τόρ affords no reliable clue to the
meaning of the details. Plutarch’s observation (De Pyth. orac. z, Mor. 395 b)
on the Lacedaemonian votive offering in Delphi has probably nothing to do
with the βάπτειν of bronze, as I am informed by Beazley, who does not feel
so certain as Blümner (pp. 327 ff.) that the account of Silanion’s Iocasta
(Plutarch, De aud. poet. 3, Mor. 18 c, and Quaest. conv. s. 1, Mor. 674 a) τὴν
πεπλασμένην ‘lokdornv, ἧς φασιν εἰς TO πρόσωπον ἀργύρου τι συμμῖξαι τὸν
! The well-known difficulties of the context in that passage need not concern us here.

304
 COMMENTARY lines 613 f.
τεχνίτην, ὅπως ἐκλείποντος ἀνθρώπου καὶ μαραινομένου λάβηι περιφάνειαν 6
χαλκός, is to be rejected. Beazley regards it as conceivable that Silanion
used a different alloy for casting the head. This would of course only be
evidence for a later period, but such a technique may have been known to
Aeschylus and his audience. In any case the hypothesis that what is meant
here is the colouring of metal remains based on slender grounds.’ We should
therefore not rule out the possibility that the passage is rightly explained
by the scholiast and by, e.g., Headlam. We might perhaps allow that in this
context χαλκός means ‘metal’ and is used of iron, the more so when we
consider’ that we have here a reference to the activity of the craftsman who
is called a χαλκεύς in Homer (e.g. in ı 391, where he πέλεκυν μέγαν ἠὲ σκέπαρνον
εἰν ὕδατι ψυχρῶι βάπτη!) as he regularly is in Attic,? σιδηρεύς being exceptional.
It is possible that χαλκοῦ Badds in the mouth of Clytemnestra acquires a
secondary meaning and conveys some hint of the murder to an audience who
know the story, cf. especially Cho. 1011 (suggesting that the queen had
planned beforehand to do the deed with Aegisthus’ sword,* see Appendix B).
6131. Assigned in the MSS to the Herald. Hermann’s recognition of the fact
that the lines belong to Clytemnestra’s speech is worthy of him. Headlam’s
discussion of the passage (C.R. xvii, 1903, 242 f.), taking into account as it
does minutiae of grammar, quite ignores the fact that for Greek tragedy
there exists also something like a grammar of dramatic technique.’ Head-
lam’s supposition (and that of others® who proceed on similar lines) is incon-
ceivable from the point of view of ancient stage practice, to say nothing of
the gross impropriety of the Herald’s launching a violent criticism of her
behaviour at the queen as she goes out (Headlam translates: ‘Such is the
boasting ; though brim-full of truth, unseemly, surely, for a noble dame to
trumpet?’). The arguments of Hermann are sound: ‘nec decet praeconem
respondere, nec potuit ille τοιόσδ᾽ 6 κόμπος dicere, sed dicere debuit τοιόσδε
κόμπος, nec iustus finis factus erat orationi Clytaemnestrae', but we can
reinforce them further. The article in 6 κόμπος has a fully deictic (or, what in
this case comes to the same thing, a possessive) force: 'of such a sort is this
my boast (as revealed in the preceding words)’. MacNeice correctly trans-
lates: ‘Such is my boast.’ Clytemnestra concludes her speech with a dix of
an individual colouring. In this kind of summing-up τοιόσδε Or τοιοῦτος is
! An explicit protest against this interpretation was raised by Th. Bergk, Griech. Lit.
Gesch. iii. 356 n. 201.
? On this point I agree with G. Thomson.
3 For an instructive instance of the close connexion between σίδηρος and χαλκεύς see the
excerpt from a dialogue in the Θησεύς of Diphilus, Athen. ro. 451 b.
4 ‘Bronze is the conventional metal of war... and “the bronze" proverbially means “the
sword" ' (G. Murray, The Rise of the Greek Epic, 4th ed., 160).
5 Headlam offends still more seriously against this grammar when (op. cit. 245) he
assigns the speech in Cho. 691 ff. to Electra, following Turnebus, Weil, and others. The
truth had been stated long before, and especially well by Conington, ad loc. Another
striking example (there are many others) of Headlam's disregard for the rules of dramatic
convention may be seen in his treatment of Cho. 624 ff. ἀκαίρως δὲ κτλ. where, in the midst
of a sentence in a stasimon, he introduces 'another voice, interrupting' to which he gives
the remainder of the stanza. His disruption of A. Suppl. 85 ff., which he quotes as a parallel,
makes things even worse.
6 Verrall introduces an invention of his own. I find it difficult to take his ‘conspirator’
seriously, although I feel by no means sure that the wretch will not rise again some day from
his well-deserved place in Hades.
4872-2 X 305
lines 613 f. COMMENTARY

generally used: Clytemnestra herself in 348 says τοιαῦτά τοι γυναικὸς ἐξ

ἐμοῦ κλύεις, and at the end of the previous speech (312) she sums up the whole
of the description just given with the words τοιοίδε τοί μοι λαμπαδηφόρων
νόμοι. Cf. Cho. 149, Eum. 638, 867, S. El. 761, where Kaibel remarks: ‘Epilogue
introduced by τοιαῦτά σοι ταῦτ᾽ ἐστίν (as in Trach. 943 τοιαῦτα τἀνθάδ᾽ ἐστίν)",
and similarly Phil. 314 and elsewhere. There is a parallel e.g. in the funeral
oration of Pericles, Thuc. 2. 41. 5, where the central part concludes with a
sentence beginning: περὶ τοιαύτης οὖν πόλεως . . ..
Moreover, Hermann rightly felt that without these two lines Clytemnestra’s
speech would lack a suitable conclusion. If, on the other hand, the lines were
assigned to the Herald, the result would be something unheard of in Greek
dramatic technique. Where can we find any instance of a third party break-
ing in like this, between the end of a long ῥῆσις and the brief reply of the
coryphaeus which follows it, so as to rob the representative of the Chorus of
his traditional right of saying the first word in answer? The normal arrange-
ment (taking only Aeschylus and the earliest extant plays of Sophocles into
account) may be illustrated by the following passages: Pers. 515, Prom. 819,
1036, Ag. 1047 (the fact that Cassandra does not reply cannot prevent us from
recognizing in this case the conventional form with which we are here con-
cerned), Soph. Aj. 481, 583, 1091, 1264, Ant. 278 (wrongly assigned to the
Guard in the MSS), 471, 681, 724, 1091. Finally, as regards the mistake of the
MSS in this case, see note on sor for other instances of the concluding lines
of a speech being assigned to the wrong speaker.
Now for the explanation of the two lines. As they stand, the only possible
way of doing justice to the structure of the sentence is: ‘talis mea gloriatio
est, ea quidem, cum veritatis plena sit, digna quam nobilis mulier sine ullo
dedecore personet’, i.e. the clause οὐκ αἰσχρὸς κτλ. (to which the clause τῆς
ἀληθείας γέμων is subordinated) is to be regarded as in apposition to 6 κόμπος.
It is possible that this was the structure intended by Aeschylus and that the
text is sound. But the passage seems to gain a great deal in force if οὐκ
αἰσχρὸς κτλ. instead of being regarded as placed in apposition is taken as a
second predicate on a level with τοιόσδε. But this involves the necessity of
writing τῆς (8°) ἀληθείας, as Wilamowitz did, and Headlam (C.R. xii, 1898,
247) long before him and before his own mistaken treatment of the whole
passage. 4 can easily drop out before A.
ὡς γυναικὶ γενναίαι. “When ws is used with a substantival expression and
attached to a statement, it limits the validity of the latter to the former: a
relative validity, which we often express by “for (considering)”. Thuc. 4. 84.
2 ἦν δὲ οὐδὲ ἀδύνατος, ὡς “ακεδαιμόνιος, eimeiv’ (KK. W. Krüger, Griech.
Sprachl. § 69. 63. 6); cf. Kühner-Gerth, ii. 493 f., where there are further
examples. In the passage of Thucydides quoted by Kriiger the point of the
ws clause is this: ‘not a bad speaker, considering the fact that he was a
Lacedaemonian' ; so here: ‘not disgraceful, considering the fact that I am a
noble lady’. If Brasidas had been an Athenian, one might have called him
ἀδύνατος εἰπεῖν; if Clytemnestra had been a woman of humble condition, her
κόμπος Might have been considered αἰσχρός and have brought disgrace upon
her. Her meaning is clear: as a lady of her rank she is above the suspicion of
idle boasting.
But this highly compressed sentence (or clause) contains in its beginning a
306
 COMMENTARY lines 615 f.

still more emphatic statement of the reason why the κόμπος for once does not
carry αἰσχύνη with it: it is ἀληθείας γέμων. When a speech, or part of a speech,
contains something which is bound to seem like a κόμπος, this may be
robbed of offensiveness by adding a remark that what has been said is true;
and this was perhaps a device not uncommon in the early oratory of the fifth
century: cf. Prom. 1030 ff. ws 68° od πεπλασμένος ὁ κόμπος, ἀλλὰ Kai λίαν
εἰρημένος. ψευδηγορεῖν yap οὐκ ἐπίσταται στόμα τὸ Δῖον, ἀλλὰ πᾶν ἔπος τελεῖ,
Thuc. 2. 41. 2 (Pericles’ funeral oration) kai ὡς οὐ λόγων ἐν τῶι παρόντι κόμπος
τάδε μᾶλλον ἢ ἔργων ἐστὶν ἀλήθεια, αὐτὴ ἡ δύναμις τῆς πόλεως . . . σημαίνει. It is
significant for the character of Clytemnestra that she of all women stresses
the words οὐκ αἰσχρός so emphatically. Thus later, when she is turning her
back upon the conventional rules of propriety, she says (856) : οὐκ αἰσχυνοῦμαι,
and after the deed, when at length she lays all pretence aside (1373): οὐκ
ἐπαισχυνθήσομαι. Here we have a further argument in favour of Hermann’s
attribution of the two lines, if it were needed. Cf. also S. El. 254, 257.
The use of λακεῖν (cf. J. H. H. Schmidt, Synonymik der griech. Sprache, i.
72; Headlam: ‘to trumpet’) intensifies the κόμπος.
γυναικὶ yevvaiat: cf. (Schneidewin) the reminiscence in S. El. 287 (Electra
speaking of Clytemnestra) ἡ λόγοισι yevvaia γυνή.
615 f. The many misunderstandings of this passage need not detain us long.
They are at least as old as Triclinius: αὕτη μὲν οὕτως εἶπεν ὥστε σὲ μαθεῖν, οὐχ
ἡμᾶς, ἀκριβῶς ἑρμηνεύσασα. It should be clear from the outset that ἑρμηνεύς
simply means ‘expounder, interpreter’ and ropoiow ἑρμηνεῦσιν is exactly the
same aS 1062 ἑρμηνέως ... τοροῦ, ‘clear interpreters’. After vain attempts
I made my escape from the maze of proposals for the interpretation of
μανθάνοντι when I came across Hdt. 3. 38. 4 τῶν ‘EAAjvaw . . . δι᾿ ἑρμηνέος
μανθανόντων τὰ λεγόμενα (scil. ὑπὸ τῶν ᾿Ινδῶν). I therefore take ἑρμηνεῦσιν to
be the equivalent of δι᾿ ἑρμηνέων (in this I agree with Hermann). For this kind
of instrumental dative, where people are designated as instruments, cf. in
general Kühner-Gerth, i. 436; in particular note E. Heraclid. 390 ff. ἐγώ vw
αὐτὸς εἶδον, ἄνδρα yap χρεών... οὐκ ἀγγέλοισι τοὺς ἐναντίους ὁρᾶν, where δι᾽
ἀγγέλων might just as well have been used, cf. Kühner-Gerth, 1. 483.
Lastly εὐπρεπῶς. It seems necessary to write εὐπρεπῇ with Auratus: we
never find simply λόγον λέγειν (or an equivalent verb), without the addition
of an attribute or a pronoun to λόγον, among the considerable number of
such expressions in both Aeschylus and Sophocles. Aeschylus has the following
examples of similar concluding lines for the Chorus after a speech by one of
the actors: Suppl. 246 εἴρηκας audi κόσμον ἀψευδῆ λόγον, Ag. 1047 cot τοι
λέγουσα παύεται σαφῇ λόγον, Cho. 510 καὶ μὴν ἀμεμφῆ τόνδ᾽ ἐτείνατον λόγον (here,
however, ἀμεμφῆ is predicative as is νημερτῇ Pers. 246 τάχ᾽ εἴσηι πάντα νημερτῆ
λόγον), to which we may add from a stichomythia in Eum. 420 μάθοιμ᾽ ἄν, εἰ
λέγοι τις ἐμφανῆ λόγον. A predilection for a particular sound-pattern for the
end of a line is unmistakable in four of the five examples here quoted. In the
earlier instances of its use εὐπρεπής is often what ‘looks well’, in the literal
sense (Xenophanes B 3. 5 Diels xairnow . . . εὐπρεπέεσσιν, and occasionally
in Herodotus), or in the figurative : Hdt. 3. 72. 3 σκῆψιν εὐπρεπεστάτην, denoting
in this case something which only looks well, which is ‘plausible’ without
1 To the discussion on the correctness of the MS reading add Pohlenz, Gnomon, ix, 1933,
625.

307
 lines 615 f. COMMENTARY
possessing any corresponding merit of truth ;' it is exactly in this sense that
Thucydides likes to use both the adjective and the adverb, e.g. 3. 38. 2 τὸ
εὐπρεπὲς τοῦ λόγου (in this and several other passages the scholia give the
- gloss πιθανόν to bring out the special shade of εὐπρεπές). πιθανόν is applicable
also in E. Tro. 951 ἔνθεν δ᾽ ἔχοις àv eis ἔμ᾽ εὐπρεπῇ λόγον, where Wilamowitz
gives the somewhat crude but essentially apt rendering : ‘Noch eins, mit dem
du Eindruck machen kannst’. In the Cretans of Euripides (Berl. Klassiker-
texte v. 2, p. 75 = v. Arnim, Suppl. Eurip. p. 24 1. 31, D. L. Page, Greek Lit.
Papyri, i. 31) Pasiphae says derisively : εὐπρεπῆ γὰρ κἀπιδείξασθαι καλά. The
word further denotes a thing which is, in given circumstances, ‘suitable,
seemly’, thus A. Cho. 663 f. ἐξελθέτω... γυνή... ἄνδρα δ᾽ εὐπρεπέστερον. And
in Pers. 833, where Darius instructs the queen κόσμον ὅστις εὐπρεπὴς λαβοῦσ᾽
ὑπαντίαζε παιδί, it is not primarily a question of clothing which ‘looks well’
(though this too may be implied), but of what is ‘seemly’, for the clothes
which Xerxes is wearing have been rent by him in his mourning, and it is
unseemly for the king to go about in such a state. Obviously the first mean-
ing intended in our present passage is that of a suitable, seemly speech. But
we may be allowed to surmise that a secondary meaning is also intended in
this sentence so packed with innuendo, and to find here as an undertone
that sense of the ‘speciosum’, the deceptively convincing, which the word
possesses in Herodotus, Euripides, Thucydides, particularly as attribute to
Adyos and similar nouns. If this is so, the expression is ambiguous, though
not more so than the whole sentence.
The words of the coryphaeus are veiled, as is always the case when anyone
conceals, under cover of an apparently loyal utterance, some dangerous
thought which still he does not wish to leave completely unsaid. Perhaps the
speaker hopes that the Herald will pass the hint on to Agamemnon, perhaps
. he merely wants to relieve his heart of a little of its burden. The obscurity of
the two lines is enhanced by the accidental circumstance that the formal
conventions of tragic dialogue left only a very small space available for this
sort of concluding utterance by the coryphaeus, and on this occasion an
unusual amount of material had to be packed into it. This appears particu-
larly in the fact that εἶπε is accompanied by two expressions which are
strictly speaking in competition with each other: οὕτως and εὐπρεπῆ λόγον.
αὕτη μὲν οὕτως εἶπε would by itself be a complete clause both in grammar and
sense. If we attempt to connect οὕτως as a purely modal qualification with
εἶπε εὐπρεπῇ (or, if preferred, εὐπρεπῶς) λόγον, ‘she has spoken in this way
properly’, it will be seen that this is meaningless. The only way of connecting
it which makes sense is: ‘thus she has spoken’ (sic dixit = haec dixit). It is
certainly surprising that in this connexion οὕτως is used as the successor, so to
speak, of the Homeric ὥς (in ds ἔφατο and the like). For the usual formula in
referring back to a direct speech immediately after its conclusion is ταῦτα or
τοιαῦτα Or τοσαῦτα ἔλεξε etc., both in tragedy (Pers. 372 etc.) and in Herodotus,
Thucydides, and elsewhere (e.g. Lysias 1. 17 ταῦτα εἰποῦσα) ; cf. on 605. But
the rarity of this use of οὕτως εἶπε does not warrant an objection to the
1 To εὐπρεπές in this sense εὔμορφον is very closely akin: E. Cycl. 316 f. ὁ πλοῦτος... τοῖς
σοφοῖς θεός, τὰ δ᾽ ἄλλα κόμποι καὶ λόγων εὐμορφία. Perhaps still closer to it comes the phrase
in Hdt. 7. 168. 2 ὑπεκρίναντο μὲν οὕτω εὐπρόσωπα.
2 Butcf.,e.g., Xen. Anab. 2. 3. 24 ὁ μὲν οὕτως εἶπεν" ἀκούσας δὲ ὁ Τισσαφέρνης krA.,Cyrup. 6.
308
 COMMENTARY lines 615 f.

interpretation which is obviously required. For before the beginning of a

speech, too, the normal pronoun is ὅδε, both in tragedy and in prose: τάδε
e.g. Pers. 356, as often in Euripides, τόδε Ag. 205, 409, elsewhere τοιάδε,
τοσόνδε, and the like; while on the other hand οὕτως (cf. the Homeric ὧδε in
ὧδε δέ τις εἴπεσκε and the like) is fairly rare; but we do find it in Ag. 125 and
e.g. E. Hel. 1578 (and apart from drama e.g. Hipponax fr. 27 D., Pind. P. 4.
11, Hdt. 8. 140 a, Callimachus, Epigr. 1. 1 Wil., Babrius 95. 60, 122. 3). If,
then, οὕτως εἶπε means ‘thus she has spoken’, we must admit that there is
some redundancy in the sentence αὕτη μὲν οὕτως εἶπε... λόγον; the manner
in which we may try to analyse the structure in terms of syntax is of minor
importance. Probably it is best to regard the expression μανθάνοντι... λόγον
as a kind of explanatory afterthought to the terse clause preceding it: ‘she
has spoken thus, to you, if you can reach understanding through clear
interpreters, a speech which appears proper’. An exactly similar ‘after-
thought’ occurs in 1047 σαφῆ λόγον.
A word must be added about the dative construction μανθάνοντί σοι. It
belongs to the category discussed in Kühner-Gerth, i. 424 B. Of the instances
given there we may mention Hdt. r. r4. 2 ἀληθέϊ δὲ λόγωι χρεωμένωι οὐ
Κορινθίων τοῦ δημοσίου ἐστὶ ó θησαυρός, ἀλλὰ Κυψέλου, 7. 143. 2 és τοὺς πολε-
μίους τῶι θεῶι εἰρῆσθαι τὸ χρηστήριον συλλαμβάνοντι κατὰ τὸ ὀρθόν, but especi-
ally we must add from tragedy S. Oed. R. 616. Here too a fois is followed
by a concluding couplet of the coryphaeus: καλῶς ἔλεξεν (Creon) εὐλαβου-
μένωι πεσεῖν, ἄναξ (Oedipus is addressed) - φρονεῖν γὰρ of ταχεῖς οὐκ ἀσφαλεῖς:
‘recte haec dicta esse fateberis, o rex, si caves ne labaris’ (Brunck, adopted by
Erfurdt-Hermann). The passage closely resembles that with which we are
dealing here, not only in its syntactical and dramatic function, but also in
the speaker’s thought and feeling: he shrinks from giving the king a plain
warning to be careful, and so has recourse to the ambiguous expression
εὐλαβουμένωι, though everyone understands the σοί which is deliberately
suppressed, and so does Oedipus himself, as his answer shows. The repre-
sentative of the Elders of Argos, speaking to a man of humbler rank, has no
need to shrink from the direct reference in oot, but he is obliged, to an even
greater extent than the Theban Elder, to disguise his real feeling: πάλαι τὸ
σιγᾶν φάρμακον βλάβης ἔχω. He cannot therefore say outright: ‘she has
spoken ; but if you would get at the meaning underlying her smooth words,
you must have them translated to you by interpreters, people who under-
stand the language of the things that have been going on here in this house,
οἶκος δ᾽ αὐτὸς εἰ φθογγὴν λάβοι, σαφέστατ᾽ ἂν λέξειεν.᾽ This he cannot say, but
instead he gives the essential warning (as does the coryphaeus in the Oedipus)
in the more innocuous form of the subordinate expression μανθάνοντι. This
veiled manner of speaking extends to the choice of the epithet for λόγον:
instead of that which would really express his meaning he chooses the word
εὐπρεπῆ, which the audience was at liberty to understand as 'specious'. The

2. 22 6 μὲν οὕτως εἶπεν' οἱ δὲ σύμμαχοι κτλ. In the puzzling dramatic fragment schol. S.
Oed. C. 1375, fr. trag. adesp. 458 N. (cf. C. Robert, Oidipus, ii. 67 ff.), the brief direct speech
(l. 11) τυφλός- (the punctuation is Dindorf's) οὔτι (thus Brunck for οὔτοι, as read in the
Vaticanus [R] 2291, cf. De Marco, ‘De Scholiis in Soph.’, Accad. Linc. 1937, Ser. VI, vol. vi,
Fasc. ii, p. 217) γνώσεται is referred to by the words (1. 12) οὕτω λέγοντες, but in this case the
phrase belongs to the same sentence as the words which precede the oratio recta.
| 309
 lines 615 f. COMMENTARY
enigmatic language of anxious warning is familiar to those who know how
people disguise their words under the oppression of a tyranny ; and Aeschylus
was well acquainted with that state of mind; cf. on 449.
617. The two preceding lines, despite the μανθάνοντί σοι, were mainly con-
cerned with Clytemnestra, being spoken as she leaves the stage. Only now
does the speaker turn himself and his thoughts completely to the Herald. A
similar arrangement occurs in the concluding words of the Chorus in Rhesus
983 ff. οὗτος μὲν ἤδη μητρὶ κηδεύειν μέλει" od δ᾽ εἴ τι πράσσειν τῶν προκειμένων
θέλεις, "Exrop, πάρεστι.
|». Μενέλεων δὲ πεύθομαι. Wilamowitz’s laconic note, ‘cf. Pers. 334’, is ap-
parently meant to draw attention to the δέ, which is all to the good. How-
ever, Pers. 334 πόσον δὲ πλῆθος is of no help to us here, for in that passage
(cf. Denniston, Particles, 175) it is a case of the δέ which accompanies the real
question after some words of introduction, and that is quite different from
what we find here. We should rather compare this passage, though it is not _
exactly of the same kind, with those of which Denniston says, p. 189: ‘when
a sentence opens with a vocative, δέ is often postponed, and follows the first
word in the main body of the sentence”, as e.g. Prom. 3 “Πφαιστε, σοὶ δὲ χρὴ
μέλειν ἐπιστολάς. In Ag. 617 the δέ, in the loose manner of lively speech,‘is
put both after the vocative of the pronoun and after the first word in the
main body of the sentence. [This is wrong; Mev. δὲ v. is a parenthesis.] :
618. σεσωμένος: for the form cf. Kühner-Blass, ii. 544 f., Meisterhans, 185,
L-S, and Schwyzer, Griech. Gramm. i. 736. It seems impossible to make out
whether in the fifth century the spelling of this tense was with wı or w.
619. Karsten’s emendation ἥκει is wholly convincing, cf. 522, 531. ἥξει can
only mean ‘will be here’, this being the established usage at least in Aeschylus
(the meaning is of course modified when the point which the ἥκων reaches is
indicated). As may be gathered from νόστιμός τε καὶ σεσωμένος and, moreover,
from the whole situation and the scene which follows, the speaker must be
anxious to find out whether Menelaus has come safely home with Agamem-
non and his followers or whether anything has befallen him on the voyage.
He is therefore inquiring about something that the Herald was bound
to have known at the time of his landing, not about something still in the
future.
620. For the construction οὐκ ἔστιν ὅστις (or ὅπως etc.) with the optative
without ἄν, which is confined to poetry, cf. Goodwin, Moods and Tenses,
§ 241, Kühner-Gerth, ii. 429, Blass on A. Cho. 172 (Sidgwick in his annotated
editions of A. Ag. and Cho., Appendix i., ought not to have included Plat.
Euth. 296 e οὐκ ἔχω ὑμῖν πῶς ἀμφισβητοίην under this head. That passage
would in any case belong to a different type ; moreover, Heindorf has rightly
inserted dv in front of audıoß., cf. Stallbaum, ad loc.).
τὰ ψευδῆ. The observations of ancient? grammarians that ψευδής does not
occur in Homer and (Apollon. Dysk., Adv., p. 137. 9 f. Schn.) that ἀψευδής
is older, previously noted by Lobeck and others, have been recalled by P.
Maas, Mélanges Boisacq (Brussels 1938), 129 f.; he further shows that there
are no certain instances of ψευδής earlier than the Attic of the fifth century.
620-2. The three lines contain three elements of which the syntactical con-
nexion involves some ambiguity, because from a purely grammatical point
of view two constructions seem possible for each. Moreover λέξαιμι is taken
310
 COMMENTARY lines 620-2

by some commentators to mean dicam (‘call, declare, designate as’), by others

to mean narrem (‘report’). The ambiguities of syntax are these:
(1) 620 καλά might be predicative, as taken by Stanley, Heath, and, among
more recent editors, e.g. Klausen, Conington, Wecklein, Verrall, Headlam
(translation). Or τὰ ψευδῆ καλά means ‘falsa bona’, as understood by Schütz,
. Hermann, and e.g. Paley, Nägelsbach, Wilamowitz (‘falsche Freude’), G.
Thomson, C.Q. xxx, 1936, 106.
(2) 621: the infinitive might be final (so Hermann: ‘non narrem falsa bona
ad id, i.e. eo consilio ut diuturnus ex 115 fructus sit’). Orit might be epexegetic,
as most commentators take it, from Bothe (‘intelligendum ὥστε ἢ down to
Wilamowitz and G. Thomson, loc. cit.
(3) 622 εἰπών is understood by Wellauer, Hermann, and others to belong
both to «eóvd and to the words which follow, i.e. πῶς δῆτ᾽ dv, εἰπὼν κεδνά,
τἀληθῆ εἰπὼν τύχοις ; while Klausen, Conington, Donaldson (quoted by Paley),
Kennedy, Verrall, Headlam, G. Thomson, loc. cit. take κεδνά as predicative.
The first thing to aim at is to have a clear idea about τὰ ψευδῇ καλά: the
rest will to a great extent follow automatically from this. I do not really
believe that any audience hearing the line οὐκ ἔσθ᾽ ὅπως λέξαιμι τὰ ψευδῆ καλά
without concerning themselves with the question of construing what might
follow afterwards would have the slightest hesitation in taking καλά to be
predicate. For ra ψευδῇ λέγειν cf., e.g., Soph. fr. 326 N. (= 352 P.) καλὸν μὲν
οὖν οὐκ ἔστι τὰ ψευδῆ λέγειν, Phil. 108 (where, however, the text is not quite
certain), Antiphon 1. τὸ τοὺς τὰ ψευδῇ παρεσκευασμένους λέγειν. Moreover,
very strong emphasis must be laid on the insuperable difficulties'in which the
interpretation of ra ψευδῆ καλά as falsa bona lands us with regard to the
infinitive construction in 621. It is all very well to explain this by ‘intelli-
gendum wore’, but Bothe, who does so, takes καλά as predicative, and in
that case the construction is plain (see below). But can the least shadow of
a parallel be produced to justify construing falsa bona and yet regarding the
infinitive in 621 as consecutive or epexegetic, so that it would have to form
a supplement to the preceding sentence as a whole? Nothing in any way
comparable is to be found in K. W. Krüger, Griech. Sprachlehre, $ 57. το,
n. 6 ff., Goodwin, ὃ 775, Kühner-Gerth, ii. 4; nor is the epexegetic infinitive
in 482 (see Ὁ. 238 f.) parallel, for it depends on ὧδε παιδνός. It may have been
this difficulty which made Hermann (see above) decide to take the infinitive
as final, but the resulting sense is impossible. On the other hand, if we take
621 as epexegetic to the predicative καλά, we have before us the familiar type
of construction (Kühner-Gerth, ii. 13) where ‘the verb in the infinitive, its
noun form, is added to the adjective’ (Wilamowitz on E. Her. 1002). The
conclusive proof that this is the correct interpretation lies in the fact that
when we combine it with the equally necessary interpretation of κεδνά (622)
as predicate, the unmistakable parallelism of 620 and 622 extends to this
detail too. The poet wanted this parallelism to be clearly accentuated:
otherwise he would not have made ἂν εἰπὼν τύχοις in the coryphaeus's reply
correspond to λέξαιμι and τἀληθῆ to τὰ ψευδῆ and xedva to καλά. It is only
those commentators who have clearly brought out this symmetry, e.g.
Klausen, Conington, Kennedy, Sidgwick, Verrall, Headlam, who have done
justice to the elaborate structure of statement and reply. It further follows
that λέγειν in λέξαιμι is not used in the sense of ‘declare that . . .’, but in the
3II
lines 620-2 COMMENTARY

sense of ‘tell, report’. ‘I could not possibly tell the false and make it good, in
such a way that my friends might for a long time gather fruit of it.’
The simple inquiry about the fortunes of Menelaus is answered ‘by the
Herald with a sententious phrase. The thought of which he makes use was
possibly a commonplace for Aeschylus’ audience. Similar phrases occur often
enough, e.g. Hdt. 7. 101. 3 κότερα ἀληθείηι χρήσωμαι πρὸς σὲ 1) ἡδονῆι; (Schneide-
win), S. Ant. 1194 f., Eur. fr. 1036 N. πότερα θέλεις σοι μαλθακὰ ψευδῆ λέγω ἢ
σκλήρ᾽ ἀληθῆ; (Wecklein), Agathon fr. 12 N. εἰ μὲν φράσω τἀληθές, οὐχί 0°
eddpava: εἰ δ᾽ εὐφρανῶ τί ο᾽, οὐχὶ τἀληθὲς φράσω (Blomfield), Nicias' letter in
Thuc. 7. 14. 4 τούτων ἐγὼ ἡδίω μὲν ἂν εἶχον ὑμῖν ἕτερα ἐπιστέλλειν, οὐ μέντοι
χρησιμώτερά γε... καὶ ἅμα τὰς φύσεις ἐπιστάμενος ὑμῶν, βουλομένων μὲν τὰ
ἥδιστα ἀκούειν, αἰτιωμένων δὲ ὕστερον... ἀσφαλέστερον ἡγησάμην τὸ ἀληθὲς
δηλῶσαι (Blomfield), Aeschines 3. 12] πότερα τἀληθὲς εἴπω 7) τὸ ἥδιστον ἀκοῦσαι;
(Schneidewin). By his recourse to generalization the Herald escapes for the
moment the announcement of bad news (στέργει γὰρ οὐδεὶς ἄγγελον κακῶν
ἐπῶν), for he does not want to tell lies and says so. At the same time this man,
as portrayed by the poet, has a natural inclination to homely philosophizing,'
cf. ssı ff, 636 ff. The coryphaeus sounds the same note in his answer.
Though he fully realizes that there is bad news to be told, he begins by giving
a turn to the Herald’s words, as he takes them up, in the direction of εὐφημεῖν.
But then (623) with greater urgency he charges the man, who has already
shown himself reluctant to tell lies, to stick to the truth, which inevitably
must be revealed.
622. was... dv... τύχοις : an instance of the familiar use by which πῶς
(or τίς) with the optative and ἄν is equivalent, or nearly equivalent, to a wish:
in the second person, e.g. 5. Phil. 794 f. πῶς ἂν ἀντ᾽ ἐμοῦ τὸν ἴσον χρόνον
τρέφοιτε τήνδε τὴν νόσον; Attention was called to this usage as early as
Valckenaer’s note on E. Hipp. 208; cf. Kühner-Gerth, i. 235, L-S πῶς II.
1 b, Jebb on S. Phil. 531. Most of the commentators rightly recognize Ag.
622 as expressing a wish (Nägelsbach : ‘O könntest du doch erfreulich reden
und wahr zugleich’, Verrall: ‘O that thy true tale might be happily told!’).
We need not have any doubts about the occurrence of this wishful type of
question in Aeschylus, for in Ag. 1448 ff. no other interpretation is possible.
Obviously πῶς ἄν in Aeschylus sometimes serves to introduce a genuine
question, e.g. Suppl. 509, Pers. 243, 788, Ag. 1374 f. (cf. on 1198), but so it does
in Sophocles too, e.g. Oed. R. 765. And it is not surprising that the usage,
which is found in a rudimentary form in Homer, and is common in Sophocles
and Euripides, should also occur in the older tragedian. The combination
with δῆτα, which appears so frequently after interrogative words (Denniston,
Particles, 269 ff.), is no ground for objection: the sentence remains interro-
gative in form, and similarly we find δῆτα in S. Aj. 879 ff. τίς ἂν δῆτά po...
ἀπύοι; where a wish is unmistakably intended. We should, by the way, be
careful in such cases not to draw the line of distinction between question and
wish too sharply. ‘Even in wish-questions the interrogative word serves
to express the possibility of the notion, so that the underlying idea is: in
1 Cf. Arist. Rhel. 2. 21, p. 1395 a 6 of γὰρ ἀγροῖκοι μάλιστα γνωμοτύποι εἰσὶ καὶ ῥαιδίως
ἀποφαίνονται (καθόλου). Obviously Aeschylus does not want to characterize the Herald as
ἀγροῖκος, but the poet and his audience would have felt there was something typical of the
‘man of the people’ in this kind of talk.
312
 COMMENTARY line 629

what way is it conceivable? is there no way at all by which you might bring
it about? could you not somehow or other?’ (Ameis-Hentze, Anhang zu
Homer, 3rd ed., on o 195). This is to a great extent true also of the post-
Homeric usage. Such wishes have as a rule a tone of resignation about them;
in Ellendt-Genthe, Lex. Soph. 676 (end of the article πῶς), stands this com-
ment: ‘est eius, qui vix confidat fore, quod optaverit'.
τύχοις : τυγχάνω with the participle is at times used by Aeschylus with the
weakened meaning which it commonly has, e.g. Cho. 688 εἰ δὲ τυγχάνω τοῖς
«vpto,t . . . λέγων, Eum. 726 χῶτε δεόμενος τύχοι. Yet in other cases we find it
unmistakably with the full meaning of ‘making a good shot (at the mark, or
the truth)’,’ thus Ag. 1232 f. τί νιν καλοῦσα δυσφιλὲς δάκος τύχοιμ᾽ ἄν, Cho. 14 f.
ἢ πατρὶ τὠμῶι τάσδ᾽ ἐπεικάσας τύχω χοὰς φερούσας, 315 ff. τί σοι φάμενος ἢ τί
ῥέξας τύχοιμ᾽ ἂν ἕκαθεν οὐρίσας, 418 τί δ᾽ ἂν φάντες τύχοιμεν, 997 τί νιν προσείπω
«ai? τύχω μάλ᾽ εὐστομῶν; No doubt this is also the case here. ‘Might you
somehow or other state the truth as good and hit the mark !’. The passages
last cited from the Oresteia belong to the same sphere of ideas as Ag. 685
γλῶσσαν ἐν τύχαι νέμων (cf. note) and Cho. 949 f. Aikav δέ νιν προσαγορεύομεν
βροτοὶ τυχόντες καλῶς. Cf. also Suppl. 588 f. καὶ τόδ᾽ ἂν γένος λέγων ἐξ 'Enddov
κυρήσαις, Ag. 628 ἔκυρσας, 1201 κυρεῖν λέγουσαν.
623. Only one further instance of εὔκρυπτος is quoted in the lexicons (from
Hippocrates) ; it is surprising that a compound so natural should not have
been used more often.
626 ff. Schneidewin points out that in making the Atridae depart together
Aeschylus diverges from the epic form of the story. According to the Νόστοι
(cf. Proclus’ summary and [Apollod.] epit. 6. x, Soph. fr. 479 N. = 522 P.) and
y 136 ff. the Atridae quarrelled after the capture of Troy and, while Agamem-
non stayed behind, Menelaus and several other leaders sailed home. Cf. C.
Robert, Heldensage, iii. 1291. Apparently Aeschylus wished to avoid any-
thing that might seem inconsistent with the ξύμῴρων rayd of the Atridae
(Schneidewin). The sympathy and respect with which Menelaus is treated
in the Agamemnon serve also to increase the effect of the Proteus. It seems
likely, as Hense and Headlam suggest, that 626 contains an allusion to the
epic tradition.
627. ἥρπασε στρατοῦ. The lengthening of a short syllable by an initial
double consonant is mostly avoided, but is normal if it is the last word but
one in the line that has its final syllable lengthened, so in Pers. 192, Sept. 75,
Prom. 804, 863, etc. Cf. I. Hilberg, Princip der Silbenwdgung (1879), 218 ff. ;
P. Maas, Griech. Metr. § 125.
628. ἄκρος: ‘pre-eminent’, ‘outstanding’, used of a person who excels in
some activity or quality, as in Herodotus, Plato, etc. As in this line it is also
used in 1130 and S. El. 1499 of great accuracy of (mental) aiming.
629. ἐφημίσω: similarly at the end of a line (lyr.) in 1162 and 1173. No middle
form is found elsewhere in dramatic literature.
1 Cf. on this point and on the original connexion between the two meanings H. Fränkel,
Hermes, lx, 1925, 188.
2 The «dv of the MS is unsatisfactory. The passages quoted above show that Porson and
Weil were right in reading xai for κἂν (no further change is needed, see Blass, ad loc.). If
κἂν be kept, τύχω would be a mere auxiliary verb : ‘What name shall I give the thing, even
if I use the mildest terms?' (Paley). What both the sense and the Aeschylean manner of
speaking require is: ‘What shall I call it and hit the mark while using mild language?’

313
line 630 COMMENTARY

630. For ydp in questions in which the supplementary information required

is in no sense explanatory see Denniston, Particles, 82.
633. τοῦ τρέφοντος Ἡλίου χθονὸς φύσιν: for this ‘sandwiching’ of a proper
name between the elements of an ‘attributive group’ see the discussion by
H. Schöne, Hermes, lx, 1925, 154; among other instances to be added to his
second category is Ag. 437 6 χρυσαμοιβὸς δ᾽ "Apns σωμάτων; cf. also Aesch.
fr. 304 N. τοῦτον δ᾽ ἐπόπτην ἔποπα τῶν αὑτοῦ Kakdv.'
634 f. The point last touched upon (‘Is Menelaus dead or still alive ?”) being
one about which nobody knows, the coryphaeus goes back to the last actual
occurrence he had learned of (628 f.) : a storm had separated Menelaus from the
fleet on the voyage home. This reference to a point farther back is well
marked by the use of yap.?
635. τελευτῆσαί re is inserted into the phrase ἐλθεῖν δαιμόνων κότωι as, e.g.,
in 318 f. κἀποθαυμάσαι is inserted into the phrase ἀκοῦσαι διηνεκῶς.
637. χωρὶς ἡ τιμὴ θεῶν. The dispute about the meaning of these words goes
back to the early days of Aeschylean study. The sense assumed by the
interlinear gloss in Tr ἤγουν ταῦτα λέγοντες ἀτιμάζομεν τοὺς θεούς seems to be:
'separated (from the announcement of disastrous news) is the τιμὴ θεῶν (i.e.
of the gods of heaven). This meaning was explicitly adopted by Petrus
Victorius, Var. lect. xx. 22: 'Non oportere festum diem, dicatumque honori
deorum, verbis tristioribus nuntiisque gravioris mali contaminare: et liberum
remotumque esse debere ab omni huiuscemodi re, quicquid ad Deos colendos
honoreque afficiendos pertinet.' The alternative view to that of Victorius is
taken by Stanley; he explains 'alius est honor eorum deorum qui bona
mittunt, alius eorum qui mala, ut Erinyes’. The subsequent course of the
controversy may be traced in Butler's excerpts from the earlier editions, and
for the nineteenth and twentieth centuries in the commentaries and transla-
tions. Here it may suffice to cite two representatives of each of these inter-
pretations. That of Victorius is adopted e.g. by Wecklein: 'geschieden ist
die Ehre der oberen Gótter (von der Verehrung der Wesen der Unterwelt)',
and by A. Platt: 'It is not meet to mar the songs of gladness with tidings of
heaviness: far be that from the giving of thanks to heaven’; that of Stanley
(taking θεῶν in this case to mean both the gods of heaven and those of the
underworld) is adopted by Verrall: ‘the functions belong to different gods,
literally “the religious province (τιμὴ θεῶν) is distinct in the two cases”, the
one belonging to joyous or friendly gods, the other to gloomy or adverse
gods’, and by Wilamowitz, as far as may be gathered from his rather free
rendering: ‘Den Tag der Freude sollte Trauerbotschaft nicht entweihen,
Himmelsgóttern und Erinyen derselbe Mund nicht opfern.’? The interpreta-
1 This was attributed by Welcker to the Tereus of Sophocles, fr. 581 P. ; ‘highly plausible,
but the last word has not yet been said’, Wilamowitz, Interpr. 28, n. 3; Welcker's attribution
is regarded as certain by Oder and other scholars mentioned in Pearson’s introduction to
the fragments of the Tereus and also e.g. by C. Robert, Griech. Heldensage 156, n. 5, and
W. Buchwald, Stud. z. Chronol. d. att. Tragödie, Diss. Königsberg, 1939, 41, whereas Wacker-
nagel, Stud. z. griech. Perfektum (Progr. Göttingen, 1904), 11, makes the cautious observa-
tion: ‘the perfect πεποικίλωκε too would be rather more suitable for Sophocles than for
Aeschylus.’
2 Cf. Denniston, 63 (§ 4). Denniston himself includes the passage in the examples (on
p. 82) of γάρ used when the further information required concerns the cause or origin of the
facts already known.
3 Here the relation between the two clauses is somewhat changed by an element of the

314
 COMMENTARY line 637
tion has thus been pursued along the old paths. Things would have been
different if the protest made by Hermann on a point of grammar had been
heeded. Hermann says: ‘Laborarunt interpretes in his verbis propterea quod
ἡ τιμὴ θεῶν pro ἡ θεῶν τιμή dictum putarunt. At ea soloeca constructio foret.
Sententia haec est, ἡ τιμὴ χωρὶς θεῶν ἐστί, praemium sine diis est’, etc. He
took τιμή to mean the reward received by the messenger. This obvious
misunderstanding should not, however, have prevented subsequent com-
mentators from verifying the validity of his statement on the point of
grammar. Van Heusde alone attempted a feeble compromise with Hermann’s
grammatical interpretation (χωρὶς θεῶν), and it was not until Headlam (C.R.
XV, 1901, 23), who seems to have been independent of Hermann, that a
resolute statement of the point was made. Headlam regards the interpreta-
tion χωρὶς θεῶν ἐστὶν ἡ τιμή as ‘the only way the Greek admits’, obviously
(though he does not say so) for the same reason as Hermann. But, in con-
trast to Hermann, Headlam has understood ἡ τιμή correctly: ‘that observ-
ance is separated from the Gods of Heaven.’ Before going any farther we
must substantiate the contention that ἡ τιμὴ θεῶν could not be said for ἡ
θεῶν τιμή. Such groupings as the following are normal: A. Eum. 422 τὸ
τέρμα τῆς φυγῆς (a sure restoration), 759 τἀρίθμημα τῶν πάλων, 5. El. 263 τοῖς
φονεῦσι τοῦ πατρός, 1382 f. τἀπιτίμια τῆς δυσσεβείας, Phil. 755 τοὐπίσαγμα τοῦ
νοσήματος, Ar. Clouds 160, 168 τοὔντερον τῆς ἐμπίδος, 233 τὴν ἰκμάδα τῆς φροντί-
δος, Birds 1538 τὸν κεραυνὸν τοῦ Διός, Lys. 241 τὴν ἀκρόπολιν τῆς θεοῦ, Plui. 45
τὴν ἐπίνοιαν τοῦ θεοῦ, 63 τὸν ὄρνιν τοῦ θεοῦ, 1193 τὸν ὀπισθόδομον ἀεὶ φυλάττων
τῆς θεοῦ, and the like (for prose cf. Kühner-Gerth i. 618 n. 1). But it is quite
unusual (if indeed admissible at all) that a corresponding word-order should
be found in cases where the genitive has no article, for example, that τὸν
φθόνον θεῶν should be used instead of τὸν θεῶν φθόνον (Pers. 362), or τῶν
λόγων πατρός instead of τῶν πατρὸς λόγων (Prom. 40), or τὸν φόνον πατρός
instead of τὸν πατρὸς φόνον (Eum. 623). In the present case we must not
satisfy ourselves with the consideration that ὁ θεός and θεός, οἱ θεοί and θεοί,
are to a great extent used as equivalents, so that we occasionally find e.g. in
Aristophanes (Peace 9, Eccl. 1095) πρὸς θεῶν (perhaps influenced by the
diction of tragedy?) in place of the far commoner πρὸς τῶν θεῶν. That is
exactly why we might expect now and then to find such an arrangement as
τὸν φθόνον θεῶν or the like if the order of words permitted. Yet this type is
entirely avoided. There is therefore not the slightest ambiguity about an
expression like that in Ar. Wasps 43 τὴν κεφαλὴν κόρακος ἔχων: it cannot
possibly mean ‘having the raven’s head’, but only ‘having his head a raven’s
head’ (it might be said that from τὴν κεφαλήν we have to supply a second xe$.,
belonging to κόρακος ἔχων, cf., e.g., Wasps 794 ἀλεκτρυόνος μ᾽ ἔφασκε κοιλίαν
ἔχειν). In A. Sept. 369 f. ὅ τοι κατόπτης, ὡς ἐμοὶ δοκεῖ, στρατοῦ πευθώ τιν᾽
ἡμῖν, ὦ φίλαι, νέαν φέρει it is obvious that στρατοῦ belongs to πευθώ. Nor is
there any doubt where the genitive belongs in the difficult lines S. E7. 686 f.
(of which a particularly unfortunate hash is made in Pearson’s Oxford text)
δρόμου δ᾽ ἰσώσας τῆι φύσει τὰ τέρματα νίκης ἔχων ἐξῆλθε πάντιμον γέρας. Here
the right interpretation of 686 is given by Kaibel and Bruhn: ‘seiner glanz-
vollen Erscheinung (φύσει) machte er das Ende des Rennens (δρόμου ra
first clause being brought into the second: in the Greek the second clause supplies an ex-
planation or a reason for the first.
315
 line 637 COMMENTARY

τέρματα) entsprechend. Denn auf den Endspurt kommt es an; mit diesem
rechnet Orest auch 734 ff.” (Bruhn) ; ᾿λαμπρὰ μὲν ἡ φύσις ἦν αὐτοῦ, οὐχ ἧττον
δὲ λαμπρὸν τὸ τοῦ δρόμου τέλος᾽ (Kaibel). Without any doubt νίκης belongs
to γέρας. ἷ
I have examined all the extant plays and fragments of Aeschylus and
Sophocles, and there is no real parallel to the supposed ἡ τιμὴ θεῶν, though
one may have escaped my notice. Obviously we cannot class in this category
partitive genitives such as 5. 47. 682 f. τοῖς πολλοῖσι γὰρ βροτῶν, Phil. 304
τοῖσι σώφροσιν βροτῶν, Oed. C. 279 πρὸς τὸν εὐσεβῆ βροτῶν, etc. The addition
of an attribute to the genitive probably makes a difference to the relation
between the genitive and its governing noun, so that 5. El. 955 τὸν αὐτόχειρα
πατρώιου φόνου (as compared with the normal τὸν αὐτόχειρα τοῦ φόνου Oed. R.
266) need not surprise us; moreover it is likely that the particular nature of
the governing noun determines the connexion between the elements of the
group, cf. Trach. 1000 f. (anap.) τίς ὁ χειροτέχνης iaropias: the person who is
denoted as producing the effect and the effect itself are very closely related
to each other, more closely than possession and possessor (ἡ τιμὴ ἡ τῶν θεῶν) ;
thus it is a matter of the ‘Mordhandhaber’= ‘murderer’, ‘Heilkunsthand-
haber’ = ‘physician’. In another case, 5. Aj. 617 f., the close connexion is
properly appreciated by Hermann and Lobeck, who remark on ra πρὶν δ᾽
ἔργα χεροῖν μεγίστας ἀρετᾶς : ‘épya χεροῖν pro uno vocabulo est’ and regard it
as the equivalent of xeıpovpynuara. In Ant. 365 f. (lyr.) too τὸ μηχανόεν τέχνας
the relation between the governing word and its genitive is different from
what we should have in our hypothetical ἡ τιμὴ θεῶν. In Trach. 527 (lyr.)
the phrase τὸ δ᾽ ἀμφινείκητον ὄμμα νύμφας does not prove that τὸ δ᾽ ὄμμα
νύμφας would be admissible ; similarly we may not conclude from Oed. C. 596
ἦ τὴν παλαιὰν ξυμφορὰν γένους that τὴν ξυμφορὰν γένους could equally well have
been used. A remarkable case occurs in 5. Oed. C. 1721 (lyr.) ἀλλ᾽ ἐπεὶ ὀλβίως
γ᾽ ἔλυσεν τό τέλος, [ὦ] φίλαι, βίου : perhaps the insertion of the vocative here is
a contributory cause ; besides, the text is not quite certain: the ὦ was deleted
by Wilamowitz, Verskunst 525. The only relevant parallel? to ἡ τιμὴ θεῶν

1 Wilamowitz, Hermes, lix, 1924, 252, should not have adopted δρόμον in 686, which he,
like Jebb, asserted to be the reading of Triclinius. (From Pearson’s apparatus, however,
it can be seen that δρόμον is not to be found in any of the MSS known to him, and further-
more that the Parisinus 2711 (T), the only representative of the Triclinian MSS which has
been collated by Pearson in its entirety—cf. his Praefatio, xiii f.—, has the reading δρόμου.)
If δρόμον is written, then τὰ τέρματα νέκης ἔχων must inevitably be taken together, with the
result that “ravrıpov γέρας is nothing but empty bombast’ (Kaibel).
2 In the new fragment (Pap. Soc. It. 1208 = D. L. Page, Greek Literary Pap. 8) of
Aeschylus’ Niobe 1. 5 it is uncertain whether τοὐπιτέρμιον (or τοὐπιτίμιον) γάμου is to be
taken by itself or to be connected with the following words Ὁ --ἶον fuap.
3 A. Pers. 604 rdvrata . . . θεῶν would have to be regarded as a real parallel if it were
provable, or at any rate probable, that τὰ dvraia is what Aeschylus wrote. That, however,
15 doubtful. The three lines 603-5 are not easy. The one thing that seems to me certain is
that those commentators of whom the scholiast of the one group of MSS (D) says τινὲς εἰς
τὸ ‘mÂéa’ στίζοντες ὡς ἐξ ἄλλης ἀρχῆς τοῦτο (i.e. 604) ovvrácaovow are right. ἐμοὲ yap...
πλέα is in itself a terse sentence of a well-known type (whereas ἐμοὶ yàp . . . θεῶν, taken as
one clause, drags painfully); moreover it is recommended by its echoing and illustrating
the general maxim 600 πάντα δειμαίνειν φιλεῖ, The μέν in 603, then, is ‘solitarium’; the
detailed description that follows it diverts or interrupts the original thought. As for the
particles in 604 f., I speak with less confidence, although I feel strongly that what is wanted
after ἐμοὶ... φόβου πλέα is an asyndeton explicativum. The combination of... re... δέ
316
 COMMENTARY line 637
which I have come across in either of the older tragic poets is Eum. 1029
τὸ φέγγος ὁρμάσθω πυρός (Headlam, C.R. xvii, 1903, 287, has objected to
πυρός because of the word-order). Cho. 658 ἄγγελλε τοῖσι κυρίοισι δωμάτων
need not disturb us, for there it is not a case of a. simple possessive relation
between οὗ κύριοι and δωμάτων, but of a closer connexion, ‘Hausherren’.
On 1224-6 it will be shown that the οἴκων δεσπότης of tragedy corresponds
to οἰκοδεσπότης, a word which occurs in Xenophon (just as aedium dominus
is the forerunner of domnaedius, cf. my book, Iktus und Akzent, 63). Side by
side with τοῖσι κυρίοισι δωμάτων we find the phrase, of just the same meaning,
in Cho. 716 τοῖς κρατοῦσι δωμάτων, which clearly illustrates the nature of
the relation (certainly not purely possessive).
To sum up. The evidence for phrases like ἡ τιμὴ θεῶν being used as equi-
valent to ἡ θεῶν τιμή is, to say the least, extremely scanty,' at any rate in
the language of Attic drama. We shall therefore have to look for a different
way of construing these words. Such a way is presented by Headlam’s
interpretation : ‘that ceremony is apart from Gods of Heaven.’ An excellent
example of the prepositional χωρίς in this position (for the adverbial χωρίς
which often plays the part of predicate at the beginning of a sentence cf.
p. 419) is to be found in Plato, Lach. 195 a χωρὶς δήπου σοφία ἐστὶν ἀνδρείας (cf.
also Eur. fr. 391 N. οὐκ ἔστιν οὐδὲν χωρὶς ἀνθρώποις θεῶν). As regards the
thought, the passage quoted by Headlam from Plato, Laws 8. 828 c is in
general instructive (ἔτι δὲ καὶ τὸ τῶν χθονίων καὶ ὅσους αὖ θεοὺς οὐρανίους
ἐπονομαστέον καὶ τὸ τῶν τούτοις ἑπομένων οὐ συμμεικτέον, ἀλλὰ χωριστέον), and
in particular one should bear in mind Sept. 720 ff. τὰν... θεὸν οὐ θεοῖς ὁμοίαν
|. Ἐρινύν and most of all the following passages from the Oresteia, Eum.
69 f. (the Erinyes) αἷς οὐ μείγνυται θεῶν τις οὐδ᾽ ἄνθρωπος οὐδὲ θήρ ποτε, 386,
where the Erinyes’ function is described as λάχη θεῶν διχοστατοῦντα (exactly
like χωρὶς θεῶν), 721 f. (Apollo speaks to the leader of the Erinyes) ἀλλ᾽ ἔν re
τοῖς νέοισι καὶ παλαιτέροις θεοῖς ἄτιμος el σύ. For the meaning of θεῶν in Ag.
637, which applies only to the gods of heaven, excluding the powers of evil
such as the Erinyes and the like, we may also compare E. Hec. 2 ἵν᾽ "Auöns
seems impossible here, ‘quum recte sibi re et δέ respondeant, ubi a partitione in opposi-
tionem transitur’ (Hermann on S. Phil. 1313). The simplest way out is perhaps to keep 7’
= re) in 604 and alter δ᾽ in 606 to τ᾽, as has been suggested by Weil (who afterwards re-
turned to the MS reading) and Wecklein.
1 This is not affected by the observations of Nauck in the supplement to his commentary
on S. Oed. C. on 62 f. (I have used the 7th and 8th editions), referred to by Hense in the
supplement to Schneidewin (on Ag. 637). Nauck produces only very few examples of such
an arrangement as ἡ τιμὴ θεῶν (he does not quote Ag. 637); some of these have been dealt
with above, others are irrelevant, e.g. S. Aj. 501 f. τὴν ὀμευνέτιν Αἴαντος and the like, for
obviously a proper name does not stand on the same level as a common noun; nor is E.
Hel. 500 τὸ δεινὸν προσπόλου a suitable parallel for ἡ τιμὴ θεῶν. But there are two passages
cited by Nauck from Euripides which are really remarkable: El. 368 αἱ φύσεις βροτῶν (where
transposition would be simple enough, but since I have not examined the whole of the
evidence from Euripides I should not like to pronounce an opinion on that point), and
Bacch. 29 τὴν ἁμαρτίαν λέχους, where Sandys, following Tyrrell, remarks that 'the two words
combine to form one idea, and are therefore treated as practically equivalent to a single
word’. This explanation, which agrees with the one quoted above from Hermann's note on
S. Aj. 617 f., is perhaps the right one; the difference between this and our hypothetical
ἡ τιμὴ θεῶν is unmistakable.—Ar. Ach. 863 $vaeire τὸν πρωκτὸν κυνός is certainly not a
comparable instance, for there πρωκτὸν κυνὸς is apparently to be understood as the name of
a tune; in his note Starkie quotes a good parallel from Ach. 980 τὸν Ἁρμόδιον ἄισεται.

317
line 637 COMMENTARY
χωρὶς ὦικισται θεῶν (quoted by Spanheim in his note on Ag. 637, reprinted in
C. G. Haupt’s edition, which shows that he understood our passage correctly).
There is nothing unusual in designating as τιμή the reverence, the worship,
the yepas, accorded to higher beings (cf. on 690) ; in Eum. 894 f. there is men-
tion of a τιμή of the Erinyes, their τιμαΐ are spoken of several times in the
play which they dominate (cf. Pohlenz, Géit. gel. Anz. 1930, 448). I am pre-
pared for the objection that ἡ τιμή in Ag. 637 is too obscure to denote the
announcing of male ominata verba. It is certainly an oddly terse expression,
but the poet has taken sufficient precautions to ensure that his audience shall
have no doubt about the meaning. In the first place the ‘article’ (or rather
the ‘pronoun’) ἡ, which most commentators completely ignore, with its full
deictic-anaphoric force signifies the τιμή as something already mentioned, and,
since we cannot consider εὔφημον ἦμαρ in this connexion, that can only be the
uttering of bad news. In the second place the very peculiar expression in 645
παιᾶνα τόνδ᾽ ᾿Ερινύων resumes ἡ τιμή and explains it.
In the clause χωρὶς κτλ. the reason for what precedes is given in asyndetic
form, cf. on 951.
638. ἀπευκτά: cf. Suppl. 790, besides this form Aeschylus uses ἀπεύχετος.
For this kind of euphemism cf. on 499.
639. M. Schmidt referred to this passage the gloss in Hesychius σμοιῶι
προσώπωι" φοβερῶι ἢ στυγνῶι, σκυθρωπῶι, and consequently substituted
σμοιῶι for στυγνῶι in the text.! This is a very attractive suggestion. We
should be put on our guard against blind faith in the excellence of our MSS,
even where they are more numerous and of better quality than here, by
instances such as dyydpov in 282 and many others throughout the text of
Aeschylus, and on the other hand against blind ᾿Ησυχιασμός (C.R. li, 1937,
60) by the many horrors which haunt the textual criticism of the past; cf.
vol. ion the MSS.
πτωσίμου : the word occurs only here and in 1122, cf. on 386 ἄφερτος.
640. We need not be disturbed by the fact that τὸ δήμιον is also used in
Aeschylus as a substantive (‘the community, the state’). Nor is there any
justification for regarding τὸ δήμιον τυχεῖν as a separate clause. (Both mis-
takes are to be found, e.g. in Paley, who paraphrases ὥστε τὸν δῆμον τυχεῖν
αὐτοῦ.) τυχεῖν is to be taken with the dative πόλει in the sense ‘to happen to
one, to befall one’ as elsewhere (Pers. 706, Prom. 346). Lastly we must not
look, as almost everyone does, for an antithesis in πόλει uév . . . πολλοὺς dE.....
Abresch, followed by Schütz, rightly comments: ‘peév . .. ἕν est pro ἕν uev’;
we have thus an example of the word-order discussed below on 759. πόλει
here is not contrasted with the individuals: it denotes all those at whom the
blows of fate are aimed, including both community and individuals, and it
carries on the πόλει of 638. What seems to be intended is an antithesis like
ἕλκος Ev μὲν τὸ δήμιον... πολὺ δὲ τὸ τῶν πολλῶν οἴκων, though actually, as so
often happens, in the second part of the antithesis strict symmetry is not

1 E. Lobel tells me that a variant of the same glossematic word can be used to emend
another Aeschylean passage. In the gloss in Hesychius AMOIOZ- κακός, Σικελοί, which was
connected with σμοιός etc. by Is. Vossius, Lobel restores AMOIOZ (for the transition from
Sy to ou see, e.g., Schwyzer, Griech. Gramm. 1. 208), and accordingly he reads in Cho. 1048,
where δμωαὶ is indubitably wrong, δμοιαὶ γυναῖκες aide: practically no change at all. Iam
inclined to follow him.
315
 COMMENTARY line 642
maintained. The meaning of 640 is therefore: ‘(when a messenger . . . brings
word) that a wound has befallen the city, a single wound that concerns the
community, but that also many men . . .'. ἕλκος is given a forward position
as the main idea, because, according to the original intention, it was to have
been the subject of the second part of the antithesis as well.
641. ἐξαγισθέντας : rendered by ἐξορισθέντας in the interlinear gloss (in Tr
and F): presumably a mere guess. Pauw, Blomfield, and others take it more
adequately as ‘devoted to the gods of destruction’; cf. Jebb on S. Oed. C.
1526, where he compares ἐξοσιοῦν, which, however, does not occur until late
times in a sense corresponding to this. The problematic ἐξάγιστα does not
provide a clue. Ed. Williger, who discusses e£ayileıw exhaustively in his
monograph ‘Hagios’, Religionsgesch. Vers. u. Vorarb. xix. 1, 1922, p. 32,
suggests two alternative possibilities for the meaning: ‘The meaning can only
be one derived from a religious ‘“‘ban’’ . . . either ‘‘to remove from the sphere
of *dyos, the holy” (formed like efopifew “to drive beyond the frontier" . . ἡ;
or "to consecrate” (in which case the preposition would have a merely in-
tensifying force asit hasin ἐξαπατᾶν, ἐξαπολλύναι). In the passage in Aeschylus,
where the preposition still has its proper force, the simplest course is to be
content with the (much weakened) first meaning: with the second we should
have to interpret as ‘‘carried off from their homes and put under the ban of
Hades".' Quite recently another instance of e£ayilew has come to light:
E. Siegmann, Untersuchungen zu Soph. Ichneutai (Diss. Hamburg, 1941), 14
and 64, states that at 5. Ichn. (fr. 314 P.) 137 the papyrus has in the text
olwı πλαγέντες ἐνθάδ᾽ e&nyioueda, and this verb, which he takes to mean
‘we are bewitched’, is definitely superior to the ἐξενίσμεθα adopted by the
corrector of the papyrus from the copy of Theon. Siegmann also discusses
A. Ag. 641 f. and rightly insists on the religious meaning of ἐξαγίζειν.
642. διπλῆι μάστιγι: ‘translata hic est ad duplex malum, ut in Choeph. v. 375’
is the explanation of Hermann and many others ; of more recent commentators
cf., e.g., Headlam: ‘with two lashes, one inflicting public and the other private
wound’, Mazon: ‘Il semble que l'image de ce double aiguillon reprenne simple-
ment l’idée des vers précédents: Arès frappe à la fois et le pays et chacun des
foyers du pays.’ Such a symbolical relation had been found as early as
Blomfield in the imagery of this passage, not indeed in διπλῆι μάστιγι, but in
the following words δίλογχον ἄτην, ‘quae nihil aliud significat quam duplicem
calamitatem, sc. publicam et privatam’. This meaning does seem to suggest
itself: it would be in keeping with the poet’s style to take up the dichotomy
of 640 f. and develop it in διπλῆι. In the passage Cho. 375, adduced as a
parallel by the commentators, the phrase διπλῆς μαράγνης does of course in
itself indicate the terrible nature of the blow, but also the word ‘double’
is unmistakably the cue for the antithesis which follows: τῶν μὲν ἀρωγοὶ
κατὰ γῆς ἤδη" τῶν δὲ κρατούντων χέρες οὐχ ὅσιαι. In the note on 537 it has
been shown that in that line διπλᾶ, though possessing a significance of its
own (referring to the penalty of the duplum), also takes up the re- and
xai-clauses of the two preceding lines. In spite of these tempting analogies
it is wrong to take διπλῆι μάστιγι in Ag. 642 as referring to the double evil
which affects the community on the one hand, and the individual on the
other, ‘as it is made the instrument with which the private calamity is
inflicted’ (Conington). διπλῆι μάστιγι denotes here as it does in 5. Aj. 242 (cf.
319
line 642 COMMENTARY

also the scholia and commentaries on Ar. Birds 1463 f.) merely the instru-
ment by which particularly painful blows are inflicted. For the μάστιξ with
two sharp points at the end, a ‘combination of whip and goad’, see Leaf on
Ψ 387. .
643. The parenthetical mention of Ares leads to an intensifying of the notion
already presented in διπλῆν. μάστιγι. One would like to read into φοινίαν
ξυνωρίδα an allusion to the war-chariot harnessed to the rayé” ἵππω (E 356)
in which Ares the χαλκάρματος (Pind. P. 4. 87) goes raging about the field,
cf., e.g., in the oracle quoted in Hdt. 7. 140. 2 Ἄρης, Zupinyeves ἅρμα διώκων.
But this interpretation cannot be regarded as certain, for in Cho. 982 the
meaning of ξυνωρίς, as so often of συζυγία, is modified into that of ‘pair’
simply, instead of ‘pair in harness’, though in fr. 381 N. (cf. fra on 842) the
word retains its full force. &{Aoyxov here is generally understood of two
spears, and probably this is right. The two spears of the warrior are character-
istic of the Homeric age (but not of the time of Aeschylus), cf. H. L. Lorimer,
Annual Brit. School Athens, xxxvii, 1940, 173: there (with reference to S. 47.
408 δίπαλτος) it is rightly stated that Sophocles took for granted in his
audience a knowledge of the methods of fighting in the Iliad. The same
observation applies to Aeschylus (‘similitudo sumpta de more Homerico
gerendi duas hastas’ Klausen), and such a ‘Homeric’ feature would be very
suitable for Ares. Others interpret δίλογχος as meaning ‘furnished with two
sharp points (barbs?)’ (a meaning derived from a well-known use of
λόγχαι, cf., e.g., Hdt. 1. 52), and regard it as a further illustrative detail of the
μάστιξ. This is improbable, because the mention of Ares has introduced fresh
images and ideas (see above). The notion of ‘twofold’ is maintained, but the
subject to which it is attached is now different.
644. Resumption after the ὅταν κτλ. of 638. In the meantime the protasis has
been expanded by the influx of new and significant details, and is now
brimfull ; so the only way of avoiding a cumbrous and too protracted apodosis
is to have recourse to a resumptive phrase (τοιῶνδε πημάτων). After the
parenthesis (640-3) πημάτων harks back to 638 πήματα and πρέπει λέγειν to
636 f. πρέπει... γλώσσηι μιαίνειν.
μέντοι: some editors, e.g. Verrall, Plüss, and Platt (J. Phil. xxxii, 1913, 57),
quiet their grammatical consciences by writing μέν τοι, which provokes the
question of what this really amounts to in the case of an enclitic.? But this
μέντοι which is merely strengthening (not adversative) is quite in order here
after τοιόσδε; cf. Denniston, Particles, 398.
On σεσαγμένον Beazley observes: ‘homely image, enhancement of φέρηι
639’. Wecklein compares E. Iph. T. 1306, where the Messenger brings to the
house καινῶν φόρτον... κακῶν. Horace says to the bearer of his light volumina
in Epist. 1. 13. 6 st te forte meae gravis uret sarcina chartae . . ..
645. πρέπει takes up the πρέπει of 636, in the same way as the expression there
ἡ τιμή is answered by παιᾶνα... ᾿Ερινύων here (cf. on 637).
! On the other hand, I doubt whether in Pindar, P. 4. 78 f. ὁ 8’ (Jason) ἦρα χρόνωι ἵκετ᾽
aixpatow διδύμαισιν ἀνὴρ ἔκπαγλος we should really assume an influence of the Homeric
model, as Boeckh did (followed by Gildersleeve, Christ, and Schroeder), and not rather,
as the whole context suggests, think of the typical traveller with two spears, so familiar
from Attic vase paintings.
2 Pace Wilamowitzii dictum sit, see his remarks Platon, ii. 414, and in his commentary
on Menander's Epiirepontes, 293 (p. 81).
320
 COMMENTARY line 649
παιᾶνα (παιῶναῦ cf. on 245 ff) τόνδ᾽ ᾿Ερινύων: cf. Cho. 151 παιᾶνα τοῦ
θανόντος, where the Alexandrian edition had a x in the margin, with the
explanation contained in the scholion: ὅτι ἐπὲ ἀποθανόντος παιᾶνα εἶπεν
κακῶς. καὶ Εὐριπίδης (Alc. 424) 'παιᾶνα τῶι κάτωθεν ἀσπόνδωι θεῶι". Similarly
the diaskeuast! Sept. 867 ff. τὸν δυσκέλαδόν θ᾽ ὕμνον ᾿Ερινύος ἰαχεῖν Aida 7’
ἐχθρὸν παιᾶν᾽ ἐπιμέλπειν. As far as Aeschylus himself is concerned, it is
characteristic of him deliberately to use such a blasphemous paradox and
speak of a paean of the Erinyes, cf. on 1144 and on 1386 f.
τόνδε: with its full deictic force,” ‘this (such a) paean of the Erinyes’. The
Herald indicates the news of disaster as that with which they are concerned
here and now, which the coryphaeus has wanted him to announce, which he
is momentarily refusing to reveal, and which he will actually be revealing
shortly. Plüss rightly says: ‘rovöe: that which is demanded by the Elders
and declined by the Herald’.
647. εὐεστώ, found three times in Aeschylus, seems, like areorw, to have had
its proper home in Ionic, cf. Bechtel, Griech. Dial. iii. 279, 299. ‘edeorw is
formed from εὖ ἐστι after the model of the feminine abstract nouns like
raw (E. Schwyzer, in the article by E. Arend, Kuhns Zeitschrift, xv,
1938, 245).
648. This anacoluthon, i.e. a fresh beginning after the participle, presents no
difficulty ; cf. besides the commentaries Berti (in the article quoted above on |
12) 263 f. and in general Wilamowitz on E. Her. 186.
Tois with demonstrative force: the explanation follows.
649. As the words stand in the MSS, no intelligible sense can be made either
οἱ χειμῶν᾽ Ἀχαιῶν (these words would of course have to be taken together, for
in this context there can be no thought of 'Achaean gods' in Aeschylus)? or
of ἀμήνιτον θεοῖς. For the latter there is no help to be got from Conington's
reference to A. Suppl. 929, or Tucker's* comparison (C.R. vii, 1893, 341) with
A. Supl. 561 νόσοις ἄθικτον ; nor can I grasp the point of Murray's comparison
with E. Hipp. 1146 (obviously a woman can say μανίω θεοῖσιν, but can a
storm?) and with Ag. 1036. The right solution was found by Blomfield
(Glossarium, with unfortunate punctuation), Dobree (Advers. ii. 24 ; he trans-
lates non expertem irae deorum in Graecos), and Hermann in proposing
᾿Αχαιοῖς and θεῶν. For the corruption Paley appropriately compares Suppl.

11 refuse to argue against those who persist in denying the interpolation detected long
ago by Th. Bergk (Griech. Literaturgeschichte, iii. 304). But as I have had to quote a passage
from these anapaests, I should like to draw attention to the sentence that follows it. This
provides a glaring example of the clumsy method of this botcher, who in all probability is
no other than the author of the finale. In 871 ff. we read ἰώ, δυσαδελφόταται πασῶν ὁπόσαι
στρόφον ἐσθῆσιν περιβάλλονται, a grotesque periphrasis, especially in this situation. The poor
devil was dazzled by the brilliance of what seemed to him a ‘figura sermonis' in Aeschylus
that was worthy of imitation, 926 ff. δυσδαίμων σφιν à τεκοῦσα προπασᾶν γυναικῶν ὁπόσαι
τεκνογόνοι κέκληνται.
2 G. Thomson asserts "τόνδ᾽ has no meaning’ and introduces into the text the τόν γ᾽
which Headlam proposed, without justifying it, in 1898; but he does not mention that
later on Headlam (in his prose translation : “then proper is it that the hymn he utters should
be this to the Erinyes’) understood τόνδ᾽ correctly.
3 Some remarks on this which are very much to the point are made by Farnell, C.Q. iv,
1910, 181.
E He translates ‘a storm which ought to (or ‘will’) stir the wrath of the gods of (= favour
able to) the Achaeans' and calls this ‘sound’.
4872-2 Y 321
line 649 COMMENTARY
369, where instead of the ἀστῶν... τοῖσδε of the MS the editors since Scaliger
have written ἀστοῖς... τῶνδε. |
Examples of ἀμήνιτος are quoted in the lexicons from Aeschylus with one
passage from Herodotus, and after that nothing earlier than Plutarch. It is
rightly translated in Passow-Crónert by 'without anger' (as it had already
been by Stephanus in the Thesaurus and Blomfield). This meaning, or still
more precisely ‘that in which there is no μῆνις᾽, suits all the passages, whereas
if we takeitasa verbal adjective proper we encounter considerable difficulties.
Such a verbal adjective could possibly in Suppl. 975 and Hdt. 9. 94. 2 be taken
in the active sense of 'not being angry with', but that is unsuitable here (in
Ag. 1036 the exact bearing of ἀμηνίτως is not free from ambiguity, see note).
A passive force is out of the question, since μηνίω is intransitive (S. Oed. C.
1274 οὐδ᾽ à μηνίεις φράσας affords of course no evidence to the contrary).
ἀμήνιτος is therefore equivalent in meaning to ἄμηνις, which does not occur
until late times. In other words, ἀμήνιτος belongs to a type common in
poetic diction of 'apparent derivatives from denominative verbs', which
Wilamowitz has discussed exhaustively on E. Her. 290 (cf. also Kaibel on
S. El. 186, p. 99 f., Pearson on Soph. fr. 249, 1014). Thus Homer twice uses
the ordinary word ἄπυρον to describe a new vessel which has not yet been put
on the fire, and also uses ἀπύρωτον once, and we find ἀμάθητος as well as
ἀμαθής, ἀκάρπιστος as Well as ἄκαρπος, and so on. The interpretation of
ἀμήνιτος here commended is further supported by the fact that ὀξυμήνιτος,ἷ
which is once found, means exactly what would be conveyed by ὀξύμηνις
(which does not occur). It is presumably due to the form of ἀμήνιτος that a
genitive (θεῶν) is added to it, as elsewhere to verbal adjectives used in a
genuinely passive sense. The comparison drawn by Hermann and by many
others after him with S. Trach. 691 ἀλαμπὲς ἡλίου and the like is not relevant,
for the sun is itself light (λάμπει δ᾽ ἠελίοιο μένος). We should rather refer to
such combinations as S. Ant. 848 φίλων ἄκλαυτος, Aj. 910 ἄφρακτος φίλων,
Trach. 685 f. ἀκτῖνος. . . θερμῆς ἄθικτον. On the analogy of this type of
expression θεῶν might somewhat boldly be made to go with ἀμήνιτον, so that
the audience would have found no difficulty in taking the meaning to be
"without μῆνις (without μηνέειν) on the part of the gods'. To explain the
dative Axaıotis Hermann rightly said that it depends on ἀμήνιτον (as else-
where on μηνίειν), and referred to Plutarch, De gen. Socr. 5, Mor. p. 578a
ἀλλ᾽ οὐδ᾽ αὐτοῖς “ακεδαιμονίοις ἀμήνιτον ἔοικεν εἶναι τὸ δαιμόνιον.
The cause of the gods’ μῆνις has been indicated by anticipation in
338-44.
The storm during the homeward voyage of Agamemnon and the others
who fought at Troy forms part of the story as told in the epic poems (cf.
δ 514 ff. and the summary of the Nöoro:). For the persistence of the storm-
motif in the other tragedians and the later tradition cf. O. Gruppe, Gr.
Mythologie, 699 n. 1.
ı Eum. 412 φόνου (a certain restoration, in spite of Verrall) . . . ὀξυμηνίτου, i.e. ‘murder
with which there is connected sharp μῆνιο᾽, ‘caedes quae ὀξεῖαν μῆνιν excitet' (Wilamowitz ;
Stanley had taken it in the same way). Otfried Müller's, Paley's, and Headlam's translation
*murder caused by keen resentment' is wrong. The μῆνις in this case is the anger of those
to whom falls the exacting of vengeance and above all, as the scholiast had realized, the
anger of the Erinyes (for the close connexion between μῆνις or Μῆνις and 'Epwós see above
on 155).
322
 COMMENTARY line 653

650. ξυνώμοσαν : this is the earliest appearance of the verb, the earliest of
συνωμότης is in Eum. 127 (also figurative).
ὄντες ἔχθιστοι τὸ πρίν. Cf. Hippocrates π. νούσων 4. 49 (vii. 580 Littré) ro
ὑδρωποειδές, ὅ τι ἐστὶ τῶι mupl πολεμιώτατον. The passage in a tragedy (Trag.
inc. 155 Ribb.) prius undis flamma signifies an ἀδύνατον as appears from the
context in Cic. Phil. 13. 49.
651. τὰ πιστά: ‘the covenant’, as in Eum. 673. The article here has a pos-
sessive or a deictic force (‘of such a kind as could now be recognized by its
effect’).
ἐδειξάτην with its full force: they are staging an ἐπίδειξις : ‘look, now we
are in league together’.
With 650 ff. cf. Ovid, Ibis 341 f. (referring to the same story, viz. the fate of
the Locrian Ajax on the homeward voyage of the Greek fleet) wique ferox
periit et fulmine et aequore raptor, sic te mersuras adiuvet ignis aquas. The
passage may, directly or indirectly, be influenced by Aeschylus.
653. Hermann proposed (as an alternative to a transposition) to put a stop
after ἐν νυκτί, This has met with an extraordinary amount of approval, even
from Hermann’s venomous critic Hartung, and more recently from Wecklein,
Verrall, Plüss, Platt (J. Phil. xxxii, 1913, 58), who indicates a lacuna before
it, and from Murray. Hermann produces only a single argument : ‘displicent
haec verba, ἐν νυκτί, in principio posita, tamquam si primaria essent. At non
possunt esse primaria: nam etsi per noctem atrocius factum est malum,
tamen non nox, sed tempestas classem afflixit.' What does this mean? For
the subject-matter cf. μ 286 f. ἐκ νυκτῶν δ᾽ ἄνεμοι χαλεποί, δηλήματα νηῶν,
γίγνονται, for both matter and form cf. Pers. 495 f. νυκτὶ δ᾽ ἐν ταύτηι θεὸς
χειμῶν᾽ ἄωρον ὧρσε κτλ. The general description 650-2 embraces the first
hours following the embarkation (in day-time obviously), and then in 653 ff.
comes the account of the worsening of the situation through the onset of the
hurricane from the north. The bad weather begins before sunset (cf. Pacuvius
411 ff. Ribbeck, in a description of the Greeks on their voyage home from
Troy : interea prope iam occidente sole inhorrescit mare, tenebrae conduplicantur,
etc.), then, during the night, the storm turns into a hurricane. The position
of δέ (cf. Denniston, Particles, 187 f.) as third word in the sentence (for ev
νυκτί ‘quasi una vox est’), where there is no close connexion between the
two preceding words, can be paralleled in Aeschylus; cf. 606, 745, 1320, also
Pers. 749 Í. θνητὸς ὧν θεῶν δὲ; πάντων ὥιετ᾽ οὐκ εὐβουλίαι καὶ Ποσειδῶνος
κρατήσειν (in Suppl. 791 it is probably right to delete the δ᾽; in Cho. 1059
Blass’s treatment is wrong, cf. Wilamowitz, Interpr. 216).
ὠρώρει: the context shows clearly that this does not denote a point of time
anterior to that of the main narrative. Forms of the pluperfect are extremely
rare in Aeschylus: apart from this passage I have only discovered one in-
stance, προὐτεθεσπίκει in Prom. 211 (to which I would add Ag. 407 ; cf. ad loc.).
The line Ag. 653 as a whole is unmistakably an adaptation of the Homeric
1 Wilamowitz adopts re from less reliable MSS, and in his apparatus he tries to justify
the asyndeton. I should prefer not to split up the long relative sentence expressing the
accusation (from 745 onwards). Probably re is an alteration by a Byzantine scholar who
objected to the position of δέ, Wecklein’s critical note (in the fourth edition of Teuffel-
Wecklein’s commentary) on Pers. 749 shows less discernment than his interpretation (1902)
of Sept. 41. The question may also be raised how far it is reasonable to speak of ‘ δέ in
the fourth place’ when θνητὸς ὧν belong so closely together.
323
line 653 COMMENTARY

phrase ὀρώρει δ᾽ οὐρανόθεν νύξ, with the same slight modification that we find
elsewhere when an expression is borrowed from Homer, e.g. Ag. 361. No
doubt ὠρώρει was regarded as a Homeric word by Aeschylus, as it was by
Sophocles, Oed. C. 1622 οὐδ᾽ ἔτ᾽ ὠρώρει (the only instance in his plays) βοή, on
the lines of βοὴ δ᾽ ἄσβεστος ὀρώρει (A 500, N 169, and elsewhere) and βοὴ δ᾽
ἐπὶ πόντον ὀρώρει (w 48). It is a very interesting point that in adopting the
word Aeschylus also adopts its syntactical function. After other scholars
H. Meltzer, Idg. Forsch. xxv, 1909, 351, has illustrated the ‘very common
type of expression’ in Homer in which ‘the pluperfect serves to carry on the
narrative’, and Wackernagel, Syntax, i. 186, has given a helpful exposition
of the main point. He actually quotes ὀρώρει δ᾽ οὐρανόθεν νύξ, and translates
“es brach vom Himmel her die Nacht herein’. Such borrowing in the matter of
tense-syntax was not seriously hindered by the different use which Aeschylus
in his own language made of the pluperfect, for, as we have seen, he very
rarely used it at all.
654. ἀλλήλησι or (less likely) ἀλλήληισι: see Meisterhans, 3rd ed., 120 f,;
Schwyzer, Griech. Gramm. i. 559.
Θρήικιαι πνοαί: cf. 192, 1418.
655. κεροτυπούμεναι: the word occurs only here. Schneidewin recalls & 3
νεῶν ὀρθοκραιράων, on which schol. A says: λέγει δὲ διὰ τὸ τὰς πρώιρας καὶ
πρύμνας ἀνατετάσθαι, ἐκ μεταφορᾶς τῶν βοῶν (the scholion on our passage has
a similar comment, see below on 657) ; for the subject-matter cf. Leaf ad loc.,
for the word-formation and meaning W. Schulze, Quaest. ep. 35 n. 2.
656. σύν governs both χειμῶνι and ζάλης (cf. Wilamowitz on E. Her. 237).
For the force of σύν cf. on 456.
657. στρόβωι: this is the only instance of the noun in extant literature (cf.
Hesychius στρόβοι: avorpopai). Cf. Cho. 202 f. otoww ἐν χειμῶσι ναυτίλων
δίκην στροβούμεθα.
The commentators are not agreed as to the interpretation of ποιμένος
κακοῦ. Memories of Suppl. 767 ναῶν ποιμένες tempted Schütz and others into
thinking of the steersman here (in his second edition, however, Schütz took it
to mean the storm). This is also Paley’s explanation, though he mentions as
an alternative the other interpretation. Headlam translates it well: ‘by the
wild handling of that evil shepherd’. ποιμένος κακοῦ, in the same position in
the line, occurs once again in Aeschylus: in the recently discovered fragment
of the Myrmidons (Pap. Soc. It. xi, 1935, 1211, D. L. Page, Greek Lit. Pap. i.
140) 1. 8 the angry Achilles describes Agamemnon in these words. Here in the
description of the storm the imagery is continued in the way characteristic
of Aeschylus (cf. on 50): the μεταφορὰ ἀπὸ τῶν ταύρων (Schol.) presented in
655 (see note) leads to the kindred idea of the shepherd.
659. ὁρῶμεν : the only present tense in this narrative. Cf. on 293.
ἀνθοῦν : ‘the ἀτρύγετος θάλασσα blooms like a meadow; but the grassblades
[‘flowers’ would be more correct] are the wrecks of ships’ is Wilamowitz’s
comment on A. Pers. 420 (Griech. Lesebuch, Erläut. 36), as he there inserted
ἀνθοῦσα which he abandoned later ; however, his remark applies to Ag. 659.
661. ἡμᾶς ye μὲν δή: as opposed to those who perished, cf. on 887.
ναῦν ἀκήρατον σκάφος is compared by Lobeck, Paraltp. 263, to E. Cycl.
505 σκάφος ὁλκὰς ws yeuobeis.
662 f. ‘We were either spirited away θεῶν κλοπαῖς (E. Or. 1498) or saved by
324
 COMMENTARY line 672
the intercession of some divinity who begged us off’ (Headlam). The two
lines are a good example of the ‘guttatim’ manner of expression (cf. on 2).
To begin with he merely indicates the nature of the occurrence (ἐξέκλεψεν
ἢ ᾿ξηιτήσατο) and for this purpose only the unemphatic indefinite pronoun is
used. But next the question obtrudes itself as to who it may have been that
brought the event about: it was θεός τις, after which follows by way of
complement and corroboration οὐκ ἄνθρωπος (for a man would have been
incapable of the action), and finally the expression of modality is added. This
arrangement of the phrases should be made clear by the punctuation. For the
thought and the expression cf. Eum. 153 τὸν μητραλοίαν δ᾽ ἐξέκλεψας ὧν θεός.
On ἤτοι... cf. Denniston, Particles, 553.
664. τύχη ... σωτήρ. Ernst Fraenkel, Nomina agentis, ii. 49, discussing ‘the
habit, which was particularly common in tragedy, of combining the agent
nouns with feminine substantives without indicating the change of gender',
quotes this passage (also 111 and 1446). The suggestion that it might be
better to spell τύχη with a capital T in modern print leads to difficulties in
this case as in many similar ones (cf. on 14). Wilamowitz, who discusses the
passage (Heimkehr des Odysseus, 10 n. 4) to illustrate the religious feeling,
says 'eine rettende Tyche'.
θέλουσα has been unnecessarily challenged. Abresch recognized it as a
term of religious language and produced good parallels from Aeschylus and
Pindar. Cf. Cho. 19 γενοῦ δὲ σύμμαχος θέλων ἐμοί, 793, 814 (where even Wila-
mowitz has altered the θέλων of the MS), Suppl. 144. These four passages
come from prayers, so does Pind. Isthm. 6. 42 f. εἴ ποτ᾽ ἐμᾶν, ὦ Ζεῦ πάτερ,
θυμῶι θέλων ἀρᾶν ἄκουσας κτλ. This use of θέλων corresponds to that of volens
propitius (for mere volens in a prayer cf., e.g., Horace, Odes 3. 30. 16). Plüss
compares the prayer of Odysseus, v 98 f. Ζεῦ πάτερ, εἴ μ᾽ ἐθέλοντες ἐπὶ τραφερήν
τε καὶ ὑγρὴν ἤγετ᾽ ἐμὴν ἐς γαῖαν κτλ.
667 f. has been punctuated correctly by many editors. The two phrases, (a)
"Any . . . πεφευγότες, (b) λευκ. «. ἦμαρ, are subordinate to the words following :
‘we had escaped from death at sea, it was bright daylight, and yet... '.
668. οὐ πεποιθότες. Many translators think it necessary to weaken this
strong expression, thus e.g. Paley: ‘hardly trusting’, similarly Verrall, L.
Campbell, Headlam (‘hardly crediting’).
669. éBouxoAodpev: Blomfield was led astray by renderings in Hesychius
which have no reference to this passage, and translated the verb here ‘to
beguile’. This error forced its way into the dictionaries of Passow and L-S
(but in the ‘Addenda’ to the gth ed., p. 2057, it is corrected). The gloss in Tr
ἤγουν ὥσπερ ἐθεραπεύομεν ἐν λογισμοῖς τισι is correct, so is Stanley ‘meditabamur
curis recentem cladem’, cf. Paley. Exactly similar is Suppl. 929 ἀβουκόλητον
τοῦτ᾽ ἐμῶι φρονήματι, cf. also Hesychius βουκολήσομεν᾽ μεριμνήσομεν.
670. σποδουμένου : ‘a strong word from popular language’ says Verrall, who
discusses it more fully on Sepé. 809 (794 Verrall) ; the same remark was made,
without reference to Verrall or Wilamowitz (Interpr. 86 n. 5), by Regenbogen,
Hermes, lxviii, 1933, 53 n. 2.
672. It is difficult, if not impossible, to decide between τί μή and τί μήν in the
four passages from tragedy discussed by Dittenberger, Hermes, xvi, 1881,
334 n. 3. Cf. also Jebb on S. 47. 668. On the whole I think there is more to
be said for reading τί μήν in all the passages (the equivalent of the Megarian
325
line 672 COMMENTARY
σά μάν), cf. Blass on Eum. 203, Denniston, Particles, 333. How easily μήν
could be corrupted to μή in this phrase can be seen in 5. El. 1280, where
nobody doubts that Seidler’s correction ri μὴν (μὴ codd.) où should be adopted ;
cf. below on 1640.
673. ταῦτ᾽ : Hermann opposed Stanley’s interpretation (= ταὔτ᾽), then
Wecklein (annotated edition) and Wilamowitz drew attention to Triclinius’
explanation ἤγουν τὸ ἀπολωλέναι.
675. πρῶτόν τε καὶ μάλιστα: cf. Plat. Charm. 157 ἃ δεῖν οὖν ἐκεῖνο καὶ πρῶτον
καὶ μάλιστα θεραπεύειν, Rep. 3. 406 Ὁ ἀπέκναισε πρῶτον μὲν καὶ μάλιστα ἑαυτόν,
ἔπειτ᾽ ἄλλους ὕστερον πολλούς, Menex. 237 € ἔστι δὲ ἀξία ἡ ywpa . . . ἐπαινεῖσθαι
. πολλαχῆι μὲν καὶ ἄλληι, πρῶτον δὲ καὶ μέγιστον ὅτι κτλ.
μολεῖν has repeatedly been altered, after the precedent set by Hartung and
Ahrens. As far as the tense is concerned, the theory of Wilamowitz (on Sept.
429) of the potential force of the second aorist seems to me to rest on too slight
a foundation (on the difficult line E. Hel. 289, where, by the way, we have
ἐλθεῖν, cf. Paley and A. C. Pearson). One might rather feel tempted to com-
pare Andoc. 3. 27 ἐκ yap τοῦ moAduov χρονισθέντος Κόρινθον ἑλεῖν προσδοκῶσι
(not ‘expect’, but, as Dalmeyda translates, ‘ils comptent s'emparer de
Corinthe’). It is, however, more important to grasp exactly the particular
meaning of μολεῖν. In several passages of this play (which is enough for our
purpose; though the use is common elsewhere too, cf. Denniston on E. El.
519) μολών, without the addition of an adverb or an indication of place, means
‘come home, returned’ : 34, 1225, 1398 (with regard to the meaning, not to the
force of the tense, cf. also 345 ei μόλοι στρατός ‘in case the army comes home’).
So the sense here is: ‘assume that he is back again.’ One of the examples
cited by the lexicons for προσδοκᾶν in this sense (cf. the passage just quoted
from Andocides) used as governing verb of a dependent present infinitive is
Xen. Anab. 6. τ. 16 καὶ Χειρίσοφος ἐνταῦθα ἦλθε τριήρη ἔχων. καὶ οὗ μὲν στρα-
τιῶται προσεδόκων ἄγοντά τι σφίσιν ἥκειν" ὁ δὲ ἦγε μὲν οὐδέν κτλ. The infinitive
μολεῖν in our passage corresponds closely to ἥκειν in that. For the thought
see Verrall's remarks: ‘ μολεῖν that he arrived, 1.e. that his ship, like that of
Agamemnon, got home, that he reached the Peloponnese after the storm,
only, being carried to a greater distance, at some other part of the coast. He
would make for the nearest accessible point, not necessarily for Argos. It is
natural that this not improbable and consoling supposition should be enter-
tained, till it is disproved.’?
676. For the function of δ᾽ οὖν after εἰ see Paley, who carefully examines this
combination of particles in his note on 1042 (1009 Paley). As is well known
(cf. Neil on Ar. Knights 423), it always indicates the stressing of the adversa-
tive conditional clause (‘but if really’); cf. also Denniston, Particles, 465. In
this passage the implied antithesis (δέ) is somewhat obscured by 'the dis-
connected way in which alternate hopes and fears are expressed' (Paley).
The natural reason for this 'disconnectedness' is to be found in the reluctance
of the speaker to utter the word of ill omen straight out: he evades it and
passes on as quickly as he can. In his paraphrase Paley inserts an explanatory
parenthesis between μολεῖν and ei δ᾽ ov: ‘(though even about Aim there is a
! I cannot agree with G. Thomson that μολεῖν as interpreted by Verrall ‘makes nonsense
of the sequel’. For the connexion of the thoughts, which intentionally are intimated only
in part, cf. also on 676.
326
 COMMENTARY line 680
painful uncertainty)’; this is all right, only it is significant of the speaker’s
mood that the thought remains unexpressed, while its general outline can be
gathered from δ᾽ οὖν. ‘In his omnibus nihil turbatum est, sed alacriter
procedit oratio, neque ad regulas logicas descriptae sunt sententiae, sed ita
ut ex ipsa mente oriuntur’ (Klausen).
ἱστορεῖ: ‘aufspiirt’ (finds out). Cf. B. Snell, Phtlol. Unters. xxix (1924),
62 n. 1.
677. Toup's correction based on Hesychius' χλωρόν τε καὶ βλέποντα" ἀντὶ τοῦ
ζῶνταξ has been accepted by many editors since Hermann. ‘It is not certain
that the gloss relates to this passage, but it is highly probable, and the
improvement is great' (Verrall). The decision raises a question of principle,
cf. on 639. We may, perhaps, in favour of the MS reading compare Pers. 299
Ξέρξης μὲν αὐτὸς ζῆι τε kai βλέπει φάος. For the strengthening double phrase
καὶ ζῶντα (or χλωρόν τε) καὶ βλέποντα cf. E. Iph. A. 1225 ζῶσάν τε καὶ θάλ-
Aovcav, and, at greater length, S. Trach. 234 f. ἰσχύοντά τε καὶ ζῶντα καὶ
θάλλοντα κοὐ νόσωι βαρύν. This intensification is probably due in such cases
to the strong desire, which is at bottom religious, to emphasize the εὐφημεῖν,
cf. also vivit valet (Plaut. Bacch. 246 and elsewhere), di me salvom et servatum
volunt (Plaut. Aul. 677), ‘è sano e salvo’, ‘he is safe and sound’, etc.
677 f£. The clause μηχαναῖς Avs . . . γένος is taken in close connexion with the
preceding words by Paley, Wecklein, and others. But it seems far better to
regard it as belonging to the apodosis (ἐλπίς τις... ἥξειν πάλιν), as e.g. Verrall,
Headlam, and Wilamowitz do. The clause introduced by εἰ δ᾽ οὖν merely
states the condicio sine qua non; the important point, for which Zeus’ special
μηχαναΐ are desired, consists in the happy homecoming from unknown foreign
parts. 'At the end of the speech the Herald, in order to prevent the κακόν of
his report having more far-reaching effects, invokes a blessing (γένοιτο δ᾽ ὡς
ἄριστα 674) and mentions ἐλπίς, which is allowed the last word' (Neustadt,
Hermes, lxiv, 1929, 253).
680. Here it is certainly best (cf. on 580) to accent κλυών as Wilamowitz does:
the aorist has also been claimed for this and like passages? by Wilhelm Schulze,
Kuhns Zeitschrift, xxix, 1888, 240 (= Kl. Schr. 337), and by E. Kueck, Studia
... in Aeschylum, Diss. Göttingen 1890, 14.

THE SECOND STASIMON (681-781)

Four pairs of stanzas, no epode.
Metre
I. 681-98 = 699-716
First (681-5 νέμων) simple trochees, all lecythia save the third line where
the addition of a cretic produces a trimeter. For the type of these trochees
1 A like interpretation was given by Platt, 7. Phil. xxxii, 1913, 59: “The speaker shrinks
from continuing his thought, because it were ill-omened to blurt out, “but as it is, the pre-
sumption is that he is dead”. Then after this ellipse he goes on with δ᾽ οὖν to suggest that
if he is alive after all he will turn up some day.’
2 Cf. schol. S. Aj. 962 καὶ νῦν βλέποντα εἶπεν ἀντὶ τοῦ ζῶντα' διὸ τὸ x πρόσκειται, schol. E.
Tro. 632 ἔθος ἐστὶ τῶι Εὐριπίδηι τῶι βλέπειν χρῆσθαι ἀντὶ τοῦ ζῆν.
3 InE. Hel. 117, too, the aoristic character οὗ κλυὼν seems obvious: εἶδες σὺ τὴν δύστηνον;
ἢ κλυὼν λέγεις; (Pearson, ad loc., is hardly justified in comparing this participle with the
use of the indicative κλύω). In the same play 930 «Avovres εἰσιδόντες makes it almost in-
evitable to ascribe one and the same tense to the two elements of the asyndeton bimembre.

327
lines 681 ff. COMMENTARY

and the important part which they play in the Oresteia cf. the note (p. 59)
on 160 ff. Then (686 ff.) a choriambic-iambic dimeter plus a hipponacteum
of the form - vv — o — o — — (cf. p. 185).
688-95 (éAévavs . . . ἄφαντον) : ionici a minore. These ionics are mostly
anaclastic (cf. p. 185): προκαλυμμάτων ἔπλευσεν etc., but a plain dimeter,
vu —— uv —-—, occurs at 690. At 693, πολύανδροΐ re φεράσπιδες κυναγοί, the
anaclastic dimeter is preceded by an ionic; the line is thus a trimeter. At
the beginning of the ionic section (688 éAévaus EAavöpos ἐλέπ-) the form of the
dimeter, vo— o —vv-, is uncommon. It is perhaps to be regarded as a
link between the choriambic elements of the preceding section and the
anaclastic ionics, the prevalent form of the following part.
696 f. is rightly termed 'priapeus' by Wilamowitz, for the glyconic is fol-
lowed by -Uv— υ----, ie. the form of the catalectic choriambic dimeter
which is equivalent to a pherecratean (cf. p. 184). The last line of the stanza,
δι᾽ "Ἔριν αἱματόεσσαν, is a normal pherecratean ; for the beginning with o o vo
instead of — v cf. e.g. Bacchyl. 4. 17 δύο 7’ ὀλυμπιονίκας and for similar begin-
nings of glyconics in Bacchylides see the preface of Snell, p. 28. The con-
clusion of the stanza (priap.+pherecr.) forms a sort of 'ephymnium rhyth-
micum', cf. p. 186.

II. 717-26 = 727-36

Three clearly marked periods: (1) 2 glyc.+pherecr., (2) 3 dactylic ἡμιεπῆ
(ἄνδρα μοι ἔννεπε Μοῦσα), (3) 2 lecythia+priap.
Every line of this stanza is a dimeter, acatalectic or catalectic. What we
here call a dactylic colon occurs elsewhere as equivalent to a pherecratean,
e.g. Ar. Thesm. 1139 παρθένον ἄζυγα κούρην.

ΠῚ. 737-49 = 750-62

3 iamb., 4 iamb., 3 iamb., choriambic-iambic dimeter, hipponacteum (cf.
p. 185). Then ionici a minore, viz. 4 dimeters, the second (745) anaclastic, the
others plain. Finally a choriambic dimeter (- — u τοῦ -—)--pherecr.
IV. 763-72 = 773-82
Iambics. The last two dimeters have the form — vv — o — —, which renders
the end of this last stanza similar to the conclusions (pherecr.) of the three
others.

681 ff. The riddle is deftly constructed. The subject, the giver of the name,
remains hidden, the object is kept back as long as possible; when it should,
at the latest, be named, a parenthesis of weighty significance in itself
(μή Tis...véuwv) intervenes. Further, after the end of this parenthesis,
we have first two adjectives, which barely serve as a veil, and then at last
the name to support the etymology which follows. This postponement of the
proper name (so that it is preceded by preparatory notions) here serves the
particular purpose of heightening the tension caused by the riddle, but
the practice is not uncommon (cf. e.g. on 877 ff. and on 1436) and seems to
have been particularly favoured in choral lyric (cf. F. Dornseiff, Pindars Stil,
1921, 107 ff., and the same author's Die archatsche Mythenerzählung, 1933, 22).
681. ὠνόμαξεν F (even after Wecklein’s edition several editors have failed to
328
 COMMENTARY line 683
notice the point). Wecklein compared 443 where F has γεμίξων (but there £
has plainly been changed to £ by the scribe). E. 755. A. 416 (ὠνόμαξας MS:
ὠνόμαζες Markland) is compared by Hermann, who tentatively suggested
ὠνόμαξεν in this passage, before the reading! of F was made known, on the
ground that the imperfect tense was questionable ;? supposing, however, that
it should be kept, he would understand it as coepit nominare. This is as
unsatisfactory as when Wecklein speaks of 'repetition' or when Plüss explains:
‘the imperfect because in the heroic age names are given as the result of a
habit, as in the case of Hector's child Skamandrios or Astyanax.’ It would
be preferable to remember the general remarks of Wackernagel, Syntax, i.
183, who warns us against drawing too sharp a line between imperfect and
aorist, and in particular to notice the use of the imperfect in the case of
ὀνόμάζειν. Wecklein compares E. Suppl. 1218 παῖς de... Τυδέως, ὃν ὠνόμαζε
Διομήδην πατήρ and Cycl. 692 (ὄνομα) ὅπερ γ᾽ ὁ φύσας ὠνόμαζ᾽ ᾽Οδυσσέα, A. C.
Pearson on E. Phoen. 27 ὅθεν νιν ᾿Ελλὰς ὠνόμαζεν Οἰδίπουν quotes (besides E.
Hipp. 32 f.) Ih. A. 281, 416 (see above) and the fragment of the Aegeus of
Euripides (fr. 2 N.) ri oe μάτηρ ἐν δεκάται τόκου ὠνόμαζεν; (fr. 1048. 4 doris Kar’
ἰσχὺν πρῶτος ὠνομάζετο had better be left out of account owing to its un-
certainty). On Corinna fr. 11 D. χώραν 7’ an’ ἑοῦς πᾶσαν ὠνούμηνεν Wilamowitz,
Berl. Klassikertexte, v. 2, 45 n. 2, remarks: “The imperfect ὠνόμαινεν is
strange, one expects the aorist’. It does not seem strange any longer, nor is
there a need to see in ὠνούμηνεν an aorist as was suggested by P. Maas, RE
xi. 1396. 4x (there is sufficient evidence for 7 = at in the Corinna quotations
of Apoll. Dysk., see fr. 8 D.).?
682. ἐτητύμως. In connexion with a discussion of A. Suppl. 584 φυσιζόου
γένος τόδε Ζηνός ἐστιν ἀληθῶς, R. Pfeiffer, Sitzgsb. Bayer. Akad., Phil.-hist.
Abt. 1938, Heft 2, 9 n. 2, points out that in Seft. 829 ὀρθῶς, Ag. 682 and Cho.
948 ἐτητύμως, fr. 6. 3 N. εὐλόγως, and Suppl. 314 ἀληθῶς are similarly used,
when character, action, or destiny are indicated by the name. Ag. 699 ff.
᾿Ιλίωι δὲ κῆδος ὀρθώνυμον τελεσσίφρων Μῆνις ἤλασεν belongs to the same
category.
683. The ordinary way of taking this type of μή is formulated by Kühner-
Gerth, ii. 524, who speak of the μή which is used ‘when a negative answer is
expected’, and in note 3 add ‘when an affirmative answer follows a question
introduced by μή, this is always contrary to the expectation of the asker of
the question’. Similar explanations are given in regard to the present pas-
sage e.g. by Paley and Wecklein (‘doch nicht wer, am Ende jemand’). This
gives the wrong tone to the question in this passage. Wilamowitz avoids this
mistake in his note on A. Suppl. 294, referring also to Prom. 247 μή πού τι
προὔβης τῶνδε καὶ περαιτέρω; where the asker of the question expects an
affirmative answer or at least thinks it probable (that passage is wrongly
treated in L-S s.v. μή C. i. x in spite of Paley, Wecklein, and Headlam, though
Ag. 683 is rightly explained).
Precisely as here the initial question is developed 5. Oed. C. 1500 f. τίς ad
1 It is not likely to be more than a slip. Cf. on 785.
2 Similarly Verrall remarks on ὠνόμαξεν ‘the tense is more suitable’.
3 In the light of the passages quoted above the question seems justified whether in Pind.
P. 2. 44 τὸν ὀνύμαξε τράφοισα Κένταυρον the well-attested reading ὀνύμαζε should not be
preferred. (I now see that Snell has put ὀνύμαξε in the text.]
329
line 683 COMMENTARY
παρ᾽ ὑμῶν κοινὸς ἠχεῖται κτύπος . . . μή τις Διὸς κεραυνός, ἢ τις ὀμβρία χάλαζ᾽
ἐπιρράξασα;
τις ὅντιν᾽ οὐχ ὁρῶμεν : see p. 346 f. for the significance of this unseen namer
and sender of Helen. We do not know whence his influence comes ; it seems a
case of one who operates φανεὶς ἀλάστωρ 7) κακὸς δαίμων ποθέν (Pers. 354).
πρόνοιαι τοῦ πεπρωμένου belong only to a god or a daimon. “The name con-
tains the destiny . . . Since the destiny of the bearer of the name only unfolds
itself with time, it must be that spirits with a knowledge of the future have
power over the giving of the name’ (Neustadt, Hermes, lxiv, 1929, 246).
πρόνοια occurs in literature for the first time here and Cho. 606. In the
re-recording made in the year 409/8 of parts of the Draconian law ‘or rather
of what in the sth cent. counted as such in Athens’ (Latte, RE xvi. 281), on
murder, we have the words (IG 1.2 115. 11): ἐὰμ μὲ ’« mpovolas! κτ[ένει τίς
twa]. How old is the phrase ἐκ προνοίας ἡ And is that technical use of πρόνοια
the oldest use?? E. Phoen. 636 f. (quoted by Blomfield) ἀληθῶς (cf. Ag. 682
ἐτητύμως) δ᾽ ὄνομα Πολυνείκη πατὴρ ἔθετό σοι θείαι προνοίαι νεικέων ἐπώνυμον
is probably borrowed from this chorus.
685. ἐν τύχαι: cf. on 622. Here, where we are concerned especially with
ὀνομάζειν, we may add E. Zph. T. 1321 πῶς σε μεῖζον ὀνομάσας τύχω; Alexis
fr. 108. 4 f. K. τί ἂν τύχοιμ᾽ ὀνομάσας; Cf. below on 1653.
νέμων : ‘directing, guiding’. R. Pfeiffer (p. 47 of the article quoted on 682)
compares with this passage and 802 οἴακα νέμων the passage from Soph.
Inachos (Pap. Tebt. iii. x, No. 692, 1. 36 = D. L. Page, Greek Lit. Pap. i. 26)
ἐπί με πόδα νέμει. Cf, on 75.
686. δορίγαμβρον does not occur elsewhere. It is much less probable that
the poet wished to attach his bold new formation to one of the meanings of
γαμβρός (‘quam sponsi armis petunt’ Blomfield, which hardly seems linguistic-
ally conceivable) than that he has employed the second element of the word
as a poetical equivalent of -yauos, not, however, without the particular
colouring which results from the associations of the word γαμβρός with the
wedding rather than with the married state. The sense which is necessary
here seems to be practically that given by the gloss in Tr τὴν διὰ πολέμων
γαμουμένην (dopı- as in δοριάλωτος), or, as later commentators put it, ‘hastis
in matrimonium ducta’; so B. Todt, ‘De Aeschylo vocabulorum inventore’,
Progr. Paedag. Halle, 1855, 39; cf. also K. Zacher, De prioris nomin. compos.
Graec. partis formatione, Halle 1873, 36, who discusses the traditional rendering
with thoughtful criticism. Since the wedding preceded the battles, it did not
quite literally take place διὰ πολέμων ; but what matters is the inevitability
with which the wedding entailed war. At least in retrospect the union of
Helen and Paris constitutes ἃ part of the events of the war.

1 This, the earliest instance of ἐκ προνοίας (φόνος and the like), ought not to have been
omitted in LS.
2 The first evidence for the cult of "405vá Πρόνοια is not found until considerably later
(cf. Höfer in Roscher’s Lexikon, iii. 3120) ; in the case of Delphi the secondary transforma-
tion of Προνάια into Πρόνοια (in the fourth century) was established by Otfr. Müller (2nd
appendix to the Eumenides, p. 14, cf. Hermann on Eum. 21, F. Jacoby on F Gr Hist 269,
Staphylos F 2); O. Gruppe’s supposition (Griech. Mythologie, 103, 1096 τι. 2) that Pisistratus
had established in Delos a temple of "405và Πρόνοια would be without any foundation
even if the source of Macrobius 1. 17. 55 did not simply confuse Delos with Delphi (as is
assumed by Wilamowitz, Hermes, liv, 1919, 54).

330
 COMMENTARY line 690

ἀμφινεικῆ : Schneidewin refers to the fragment conjecturally ascribed by

Bergk (Mel. adesp. 44) to Stesichoros (fr. ro a Diehl) ἀλλ᾽ à πολυνεικὴς 87”
Ἑλένα. The exact meaning of ἀμφινεικής is doubtful. In the scholia on S.
Trach. 104 and 527 the attributes ἀμφινεικής and ἀμφινείκητος (both used of
Deianeira) are glossed by περιμάχητος. It is presumably on the ground of
this that Blomfield in his glossary translates the word de qua contenditur ; so
e.g. Linwood, ‘made a subject of contention’, and many editors. Opposed to
this is the gloss ın Tr on Ag. 686 περὶ ἣν δύο ἐφιλονείκουν and Triclinius himself
in his paraphrase: epi ἣν ἀμφότεροι ἐφιλονείκουν, MevéAaós re καὶ Πάρις. This
is followed e.g. by Paley: ‘a cause of dispute between two lovers’. For this
interpretation Paley refers to S. Trach. 527, for there it is obvious from
507 ff. that the strife of the two opponents is meant ; the same holds good for
Trach. 104, where Jebb is wrong in attacking Paley. This rendering seems to
me preferable on grounds both of sense and of language. ‘The meaning “on
both sides" is plainly preserved in many nominal compounds of ἀμφί᾽
(Wackernagel, Syntax, ii. 159, with examples) ; cf. also E. Williger, Sprachl.
Unters. 21 n. 1, and below on 881. But we need not think, with Triclinius,
only of Menelaus and Paris, but may preferably extend the reference to the
two sides represented by Menelaus and Paris. Thus the notion conveyed by
ἀμφινεικής is closely connected with δορίγαμβρος : both refer to the Trojan War.
687. Ἑλέναν. For the postponement of the name which solves the riddle cf.
above on 681 ff. Wilamowitz, Interpr. 77, remarks on Sept. 472 ff.: ‘He has
the name on the tip of his tongue, but wants to give first the momentous
characterization κόμπον ἐν χεροῖν ἔχων and thus make the hearers expect the
name more eagerly. In exactly the same way the messenger says 631 f. τὸν
ἕβδομον... τὸν αὐτοῦ σοῦ κασίγνητον, and eight lines intervene before the name
Polynices is mentioned. There the intention is absolutely plain’, etc. Cf.
also the carefully arranged γρῖφος at the beginning of the antistrophe Sept.
727 ff., where we are given first three veiled descriptions: (1) ξένος, (2)
KaAußos Σκυθῶν ἄποικος, (3) κτεάνων xpnparodairas πικρός, and only after
these periphrases comes the answer to the riddle in the κυρία λέξις : σίδαρος.
In Ag. 1186-90 it is only in the last word of a long sentence that it becomes
plain what sort of χορός is meant.
πρεπόντως : ‘so as to match the name’. For the occurrence of the adverbial
form cf. on 1395.
For Aeschylus’ etymologies a reference to W. Kranz, Stasimon, 287 ff., and
W. Schmid, Gesch. d. griech. Lit. i. 2 (1934), 297 n. 3, may suffice here. For the
historical understanding the important hint in Wilamowitz, Glaube d. Hell.
ii. 135 n. 1, should be added. It is well known how widespread was the
tendency to draw omens from names. Cf. A. S. Pease on Cicero, De divin.
(Illinois 1920), 1. 102, E. Riess, RE xviii. 376 ff., and the literature they quote.
688. Obviously &A&vaus (Blomfield), in spite of Lobeck, Paralip. 228.
690. Salmasius’ suggestion ἁβροπήνων is a palmary emendation, the corrup-
tion being easy (Paley refers to Prom. 223 where P? offers τιμαῖς instead of
mowais). It is probable that Lycophron 863 borrowed afpomyjvous . . . πέπλους
from Tragedy ; we cannot be sure whether he took it from Aeschylus. Hermann
calls ἁβροπήνων 'aptissimum', but keeps dßporiuwv and refers in its defence
to E. Iph. T. 1148 ἁβρόπλουτος, but that is not a compound of -τιμ-, and
when Paley in support of ἁβροτίμων quotes Suppl. 24 βαρύτιμοι (the heroes),
331
line 690 COMMENTARY
709 Δίκας μεγιστοτίμου, Cho. 357 σεμνότιμος ἀνάκτωρ (the hero)—cf. also Eum.
833—these passages only show who, like the πολυτίμητοι θεοί, has a claim to
rai (cf. on 637), and how questionable, totally apart from the compounding
with áfpós, such a compound would be as an epithet applied to προκαλύμ-
para. On the other hand, there are real parallels to ἁβρόπηνος in E. Iph. T.
814 εὐπήνοις ὑφαῖς (Blomfield), and in the same play 312 πέπλων re mpourd-
λυπτεν (N.B.) εὐπήνους ὑφάς.
692. The gloss in Hesychius γίγαντος" μεγάλου, ἰσχυροῦ, ὑπερφυοῦς may perhaps
refer to this passage, as was assumed by Wakefield, Hermann, and others;
at least it fits the meaning of the word here: ‘desumptum nomen de vi
immani ventorum’ (Klausen). The idea of Zephyros being a giant seems
remarkable when we remember that the Athenians in the time of Aeschylus
represented Zephyros as a beardless youth, e.g. on the famous skyphos in
Vienna by the Zephyros painter (H. R. W. Smith, Der Lewismaler, 1939,
pl. 27). There is no need to drag in the myth of Zephyros being the father of
Eros (Headlam). Verrall’s remark too misses the poet’s purpose: ‘in fact the
wind itself typifies the wild and monstrous passion’, for which e.g. the varia-
tions of Kerkidas (fr. 2a D.) on a theme of Euripides might be quoted.
Aeschylus obviously is concerned to emphasize that Helen’s fateful journey
was wonderfully, but also uncannily (see on 701), quick and easy. Just as
she left the home of her husband (βεβάκει ῥίμφα διὰ πυλᾶν, 407) without
finding the least hindrance either in herself or in anyone else, so the powerful
west wind helps her on the most speedy journey to Troy. Aeschylus is
following the Cypria in which we are told (Hdt. 2. 117) ὡς τριταῖος ἐκ Σπάρτης
Ardéavdpos ἀπίκετο ἐς τὸ "IArov ἄγων ᾿Ελένην, etaëï τε πνεύματι χρησάμενος καὶ
θαλάσσηι λείηι (for the variant account in the excerpt of Proclus and possible
explanations cf. C. Robert, Heldensage, 1084 f.). Pindar (N. 7. 27 ff., quoted
by G. Thomson) says of the Greek army that it θοαῖς ἂν ναυσὲ πόρευσαν
εὐθυπνόου Ζεφύροιο πομπαὶ πρὸς " Mov πόλιν, and in Sophocles’ Ἀχαιῶν σύλλογος
(fr. 142 P., col. 2. ı f.) Notos and Zephyros are mentioned as the winds one
of which shall carry the Greek fleet to Troy.
693. πολύανδροί τε κτλ. The verb ἔπλευσε (691) is used ἀπὸ κοινοῦ as with the
simple re e.g. in E. Phoen. 33 ff. παῖς oùuos . . . ἔστειχε... πρὸς δῶμα Φοίβου,
Adis θ᾽ οὑμὸς πόσις. We find the addition of a plural subject to a singular
predicate in K 563 (τόν pa) “Ἕκτωρ τε προέηκε kai ἄλλοι Τρῶες ἀγαυοί, and
elsewhere.
πολύανδροι = πολλοὶ ὄντες, cf. E. Williger, Sprachl. Unters. 17.
694. κυναγοί: cf. Lycophr, τοῖς (Aietes in pursuit of the Argo in which
Medea is escaping) τὴν vuudaywyov ἐκκυνηγετῶν τρόπιν.
695. The image of the hunt is maintained, but the fugitive quarry is a ship
and so the track is an iyvos ἄφαντον.
696. It is clear that κελσάντων refers to Paris and Helen. There are doubts
about the relationship between mAaräv and κελσάντων. Wilamowitz, Interpr.
195, saw in it ‘only another example of the feminine use of the masculine
participle’, but that is rather uncertain after what has been said on 562, and
should at the most only be taken into consideration if every other explanation
fails. Therefore we must consider whether κελσάντων is attached to πλατᾶν
as possessive genitive, or is put in parallel to it and consequently is possessive
genitive to txvos : ‘on the invisible track of oars, (the track) of those who had
332
 COMMENTARY line 698
landed’; perhaps we may even take xeAodvrwv as genitive absolute (‘while
they had landed’). A certain degree of independence in the structure of
696 ff. seems to give greater force to the concluding thought, moreover it is
desirable that this ‘rhythmical refrain’ should be syntactically separated from
the preceding lines. For the corresponding passage in the antistrophe (714 ff.)
we have unfortunately, owing to the corruption of the text, no quite definite
criterion, but everything goes to show that after aivéAexrpov we should put
a strong syntactical division. The four lines 723-6 = 733-6 are clearly
separated syntactically from the preceding dactyls, and the concluding
priapeum also is clearly divided from what precedes, Similarly in 748 f. =
761 f. the quasi-priapeum which closes the strophe is syntactically indepen-
dent. These glyconic-pherecratean codas belong apparently to the larger
company of those ‘rhythmical refrains’ of which I have spoken in dealing
with the first stasimon. Almost all those pherecratean-glyconic refrains
form independent syntactic kola (381 ff. = 399 ff., 416 ff. = 433 ff., 452 ff. =
471 ff., Suppl. 639 ff. = 652 ff., 663 ff. = 674 ff., 684 ff. — 694 ff., Sept. 295 ff. =
312 ff., E. Her. 359 ff. = 375 ff.), and, while the character of a refrain is more
obvious in the passages quoted, it is unmistakably present in the second stasi-
mon of the Agamemnon too. So it must not be considered accidental that the
corresponding syntactical arrangernent is maintained here, so far as we can see.
698. What I have just said should suffice to disprove the view taken by most
editors that δι᾽ ἔριν αἷμ. should be connected with 693 f. πολύανδροί τε κτλ.
across the frontier of the ‘refrain’. And, quite apart from the considerations I
have suggested, such an arrangement seems intolerable: ‘and after her a
swarming host of men, fierce hunters armed with shield upon the trail—the
unseen oar-blade trail of a quarry that had landed at the leafy banks of
Simois—in bloody rivalry’ (Headlam). The artificial parenthesis does not
remove the difficulty that, as soon as we have heard κελσάντων «rA., the
picture of the sea and the pursuing ships sinks into the background and a new
picture, that of the wooded Trojan shore and the guilty pair landed there,
dominates our fancy. No printed dash in the world can prevent us from
connecting & ἔριν αἷμ. with the immediately preceding words, with which
they are rhythmically so closely connected. Wecklein seems to have felt
something of the kind, but his explanation is grotesque: ' δι ἔριν αἷμ. goes
with ἀεξιφύλλους : the soil fertilized with blood will produce luxuriant vegeta-
tion'. The natural thing is to connect the adverbial qualification with the
verb, i.e. κελσάντων. But the thought cannot be that the landing of Helen was
caused by some quarrel that had taken place in the past, yet that would seem
to be the meaning of δι᾽ ἔριν. It follows that ἔρις is here the spirit of strife, i.e.
"Epis with a capital E, according to our habit of writing proper names. Thus
we restore to the preposition its particular significance. We have here the use
found from Homer onwards (cf. Kühner-Gerth, i. 484) according to which an
accusative with διά expresses a deity or divine action thanks to whom or
which something comes about. This use has been discussed with reference to
Pind. Isthm. 5. 11 κρίνεται δ᾽ ἀλκὰ διὰ δαίμονας ἀνδρῶν by Wilamowitz, Berl.
Sitzgsb. 1909, 828 n. 1; he remarks: ‘For this accusative with διά, equivalent
to ἕκατι, the best example is the inscription of Selinus IG xiv. 268 [= Schwyzer,
Exempla 166] διὰ τὸς θεὸς τόσδε vırövrı τοὶ ΣΖελινόντιοι etc.'; for the use in
prayers cf. Rhein. Mus. lxxii, 1918, 178 n. 1. So Mazon is right in printing δι᾿
333
line 698 COMMENTARY
“Epw, which Blass! had considered (on Cho. 474), cf. also G. Thomson. For
the ‘personification’ of the παιδολέτωρ “Epis in Sept. 726 cf. Wilamowitz,
Interpr. 79 n. x. The idea of the goddess of Strife standing behind the rape of
Helen which leads to the war, and therefore also facilitating her landing in
"Troas, is natural in itself and may have made the stronger appeal to Aeschylus
because it was “Epis in the Cypria (cf. the summary of Proclus) who caused
the quarrel of the goddesses and consequently the judgement of Paris. The
disastrous landing on the shore is ascribed to the intervention of "Epis, just
as immediately afterwards the disaster of the marriage in Troy is the work of
her near relative and companion Minis.
699. ’IAlur: for the meaning of the dative v. on 7or. The word placed emphatic-
ally at the beginning sharply marks the transition to the theme of the anti-
strophe: hitherto attention was centred upon Helen, what follows is concerned
with the fate of Troy.
«nöos:scarcely translatable, ἐπειδὴ τὸ κῆδος σημαίνει καὶ τὴν ἐπιγαμβρίαν
καὶ τὸ πένθος (schol). H. Voss: ‘Eh’ und Weh'.
699 f. ὀρθώνυμον occurs only here: possibly coined by Aeschylus (who wrote
in Sept. 405 ὀρθῶς ἐνδίκως τ᾽ ἐπώνυμον) on the model of ψευδώνυμος which he
uses several times ; a noticeable forerunner of the ὀρθοέπεια of Protagoras, cf.
Wilamowitz on E. Her. 56. For ‘genuine, true, and correct’ etc. naming of
the powers of fate and their actions (here an action of Myus is described) in
Aeschylus see on 682.
700. τελεσσίφρων: for the striving towards the τέλος which belongs to the
character of ῆνις and ’Epwus, cf. 996 ff.
701. Again (see pp. 91 ff.) Mus is closely related to ’Epwus (finally [749] Helen
herself plays the part).
Headlam’s ἤνυσεν, which he glosses with ἐτέλεσεν, ἐτελείωσεν, ἐξέπραξεν,
spoils the sense of the whole passage (his statement, cf. C.R. xiv, 1900, 116,
‘the MS reading κῆδος ἤλασε would mean "drove away", 'dispelled",' is
arbitrary). There can be no question at the present moment of completion
and fulfilment: although Μῆνις means to accomplish (τελεσσίφρων) what she
has been bearing in mind all the time, she causes it to become a reality only
with the conquest of the town. What is required here in continuation of the
story of the strophe is that after the landing in the district (Simois) there
followed the wedding in Troy. Just as the former was brought about à
"Epw, so Minis herself caused the latter (the fact that the giant Zephyros had
given a favourable journey was a mild prelude to the entrance of these
powers). With reference to the sense of ἤλασεν Plüss refers to S. Oed. R. 255
πρᾶγμα... θεήλατον and 992 θεήλατον μάντευμα, where the compound has the
same sense ('driven on, set in motion, by the god’) as in Ag. 1297 f. θεηλάτου
βοός. Verrall translates here, though with hesitation, ‘chased’ and supposes
an echo of the hunting metaphor of the strophe. This interpretation is
excluded by the object κῆδος. ἐλαύνειν means not only 'drive in front of
oneself' but also 'spur on, set in motion'. Quite in the spirit of the older
language Demosthenes says in one of the pathetic passages of his third
1 Wilamowitz too was on the right track. He says Interpr. 198 f.: ‘With ἔρις [Ag. 1461]
the Chorus is referring back to δι᾽ ἔριν αἱματόεσσαν, so there the noun is meant to be at least
half personification.’ I cannot, however, agree, as far as Ag. 698 is concerned, with his
further inference that Helen is identified with Eris.

334
 COMMENTARY line 705

Philippic speech (9. 54): ἀλλ᾽ eis τοῦτ᾽ ἀφῖχθε μωρίας ἢ παρανοίας ἢ οὐκ ἔχω τί
λέγω: πολλάκις γὰρ ἔμοιγ᾽ ἐπελήλυθε καὶ τοῦτο φοβεῖσθαι, μή τι δαιμόνιον τὰ
πράγματ᾽ ἐλαύνηι. A milder instance of the use of ἐλαύνειν in the sense of
‘spur on, set in motion’ has been badly handled by lexicographers and com-
. mentators: Pindar, N. 3. 74 ἐλᾶι δὲ καὶ τέσσαρας ἀρετὰς 6 θνατὸς αἰών (here
too a non-human agent). The paraphrase in the scholion weakens this down
to φέρει (hence Boeckh 'affert' ; Wilamowitz goes still farther, Pindaros 279:
‘das Menschenleben verlangt vier Tugenden’); similarly Passow 'hervor-
treiben, hervorbringen’ and, following him, Dindorf in Thesaurus, iii. 681a,
who renders it by 'produco, gigno, emitto', although he has no parallels to
offer (for Ar. Ach. 995 ἀμπελίδος ὄρχον ἐλάσαι he rightly suggests that it may
refer to the ‘drawing’ of the rows, like A 68 ὄγμον ἐλαύνωσιν, Pindar, P. 4. 228
αὔλακας. . . ἤλαυνε etc.). L-S follow suit, s.v. ἐλαύνω III 2 (plant, produce’),
and even O. Schroeder (ed. mai.) translates ‘procreat’. But Pindar says: ‘the
Aidv sets the ἀρεταί in motion, sends them on their way." In Ag. 699 ff. the
exact interpretation of ἤλασεν helps us also to give the right interpretation
to ᾿Ϊλίωι. Some editors have taken it as dative of the end to which, so e.g.
Bothe ( ἤλασεν ᾿Ιλίωι, h.e. eis " Dov, impulit κῆδος, velut telum, in Ilium’) and
Wilamowitz on E. Her. 837, giving it a slightly different sense ('A. Ag. 701
ist ἐλαύνειν noch sinnlich “‘hintreiben” ᾽). Possibly we might be satisfied with
this explanation. But the available evidence is not in favour of it; see the
examples of the use of the dative to answer the question 'whither?' given in
Kühner-Gerth, i. 443, in E. Bruhn's collection (Anhang zu Soph. ὃ 51), and
Jebb's notes on S. Phil. 67 and similar passages. So it seems preferable to
regard the dative as belonging to the neighbouring κῆδος, and signifying to
whom it is a κῆδος, as in E. Tro. 567 Φρυγῶν δὲ πατρίδι πένθος and many other
passages. It then means: ' a κῆδος for Ilion, one that exactly answered to its
name, was set in motion by Mis.’
The v of ἤλασεν is here required, cf. 201 app. crit. and Appendix E.
701 ff. πρασσομένα governs in the ordinary way a double accusative, (a)
τραπέζας καὶ ξυνεστίου Διὸς ἀτίμωσιν, where ἀτίμωσιν is equivalent to τὸ διὰ τὴν
ἀτίμωσιν ὀφειλόμενον, (b) (τοὺς) τὸ νυμφότιμον μέλος τίοντας. The order of the
words would be quite simple if we had only τραπ. ἀτίμ. ὑστέρωι χρόνωι πρασσ.,
but because after ὑστέρωι χρόνωι a second genitive depending on ἀτίμωσιν is
joined on by xai, the words xai Evveoriov Διός seem like an afterthought or
insertion. τὸ vuudor. μέλος... riovras comes at the end of the kolon because
the marriage-song, taken up again by ὑμέναιον, is to provide the keynote for
what follows.
701 f. τραπέζας as 401.
702. ὑστέρωι χρόνωι, in accordance with the character of the ὑστερόποινος
᾿Ερωύς (58 f.).
704. ξυνεστίου. Not, of course, a cult-name. The word describes Ζεὺς ξένιος
as guardian of the συνέστιοι (cf. Suppl. x Ζεὺς... ἀφίκτωρ, Eum. x19 προσί-
«ropes, Said of the gods in charge of suppliants, and see, for the meaning of
these epithets, p. 572, n. 1).
705. πρασσομένα: the present (of course, πραξομένα has been conjectured)
τ Beazley suggests that we might be meant to think of the image of a four-horse chariot.
Cf. Farnell's translation ‘four virtues are yokéd to the car of our mortal life’. This seems to
me the correct interpretation.
335
line 705 COMMENTARY
because πράσσειν (or πράσσεσθαι) χρέος is a prolonged process carried out in
stages: it includes not only the final attainment of the payment (that takes
place only with the victory) but also the making of the demand. So Pind.
Ol. 3. 7 στέφανοι mpáocovri με τοῦτο... χρέος, P. 9. 103 f. ἐμὲ δ᾽ οὖν τις ἀοιδᾶν
δίψαν ἀκειόμενον πράσσει χρέος κτλ. The Trojans by joining in the celebration of
Paris’ wedding make themselves jointly guilty of the rape; ῆνις, who has
helped to set this marriage-tie going, can now fasten on those who share the
guilt: already at the wedding she lays her claim against them and prepares
the way for the recompense.
706. ἐκφάτως. The adverb does not occur elsewhere, and the adjective seems
not to occur anywhere in early literature. Hence we can only conjecture its
meaning. Blomfield: ‘modo ineffabili. idem videtur significare atque adarws’,
Hermann (with Stanley): ‘supra modum dicendum’, Ahrens (p. 548):
“ungeheuer, übermássig, und von d$aros eben nicht verschieden’, Paley:
' "with loud voice" after Homer's ἐκφάσθαι ἔπος [this occurs only in two
similar beginnings of lines « 246 οὐδέ τι ἐκφάσθαι... . ἔπος and v 308 μηδέ τωι
ἐκφάσθαι)᾽, Wecklein's paraphrase wavers between several meanings ('em-
phatically, uncommonly', and then he quotes Paley's interpretation), Head-
lam in the main like Paley, though he is more inclined to Karsten's εὐφάτως
which Ahrens rejects (Headlam adds: ‘if ἐκφάτως is sound, it means “out-
'spokenly", in loud and bold avowal), similarly Wilamowitz: “ διαρρήδην,
ἀπροφασίστως᾽. Let us hope that the word will sometime be found in another
context in some papyrus of a classical poem.
707. ἐπέρρεπε (the deletion of -v ἐφελκυστικόν is demanded by the metre) :!
‘fell to’, intransitive as 1042.
708. γαμβροῖσιν: Pollux 3. 31 γαμβροὶ δ᾽ οἱ ἐκ τῆς τοῦ γήμαντος οἰκίας, olov
πατὴρ καὶ μήτηρ καὶ οἱ ἄλλοι πάντες κατὰ ταὐτὰ οἱ πρὸς γένους τῶι ἀνδρί. Cf.
M. Haupt, Opusc. ii. 401.
710. πόλις γεραιά: the same turn in Demades quoted by Demetrius, De
eloc. 285 πόλιν, οὐ τὴν ἐπὶ προγόνων τὴν ναύμαχον, ἀλλὰ γραῦν, σανδάλια ὑποδεδε-
μένην κτλ. Here: old as she is, she must still unlearn. A tragic and ironic
adaptation of a proverbial saying. Cf. 1425, 1619. More innocently and play-
fully Suppl. 361 σὺ δὲ παρ᾽ ὀψιγόνου μάθε γεραιόφρων, Cho. 171 πῶς οὖν παλαιὰ
παρὰ νεωτέρας μάθω; The recalling of the hymenaios, when later a dirge has
taken its place, occurs also in Prom. 555 ff. and, in a particularly beautiful
way, E. Alc. 915 ff.
711. πολύθρηνον belongs undoubtedly to what precedes, since the point of
this thought is that the former ὕμνος, 1.6. the hymenaios, becomes a θρῆνος.
Besides, to connect πολύθρηνον μέγα στένει (as e.g. Verrall) is scarcely admissible
(see on 375 f.). Moreover, the position of που, in the second place in the kolon,
is in favour of the division of the clauses which I have adopted. μεταμανϑάνειν
signifies both the unlearning of the old and the accompanying learning of the
new in its place; it makes little difference here whether ὕμνον πολύθρηνον is
the simple object of μεταμανθάνουσα in the sense of learning instead of some-
thing else, or πολύθρηνον, possibly with greater force, is taken predicatively :
‘she relearns the ὕμνος so that it becomes a dirge’ (Wecklein).
! The final ν is kept by G. Thomson (ii. 332), who uses it, together with two passages
which are admittedly corrupt, to build upon it a new theory of free strophic correspondence
in Aeschylus.

336
 COMMENTARY lines 714 f.
712. κικλήσκουσα: the verb (cf. on 1477) implies that the στένειν is directly
addressed to Paris.
712 f. A word overstepping the end of a dimeter is not uncommon in such
ionics, cf. Suppl. 870 χῶ]μα πολύψαμμον ἀλαθείς = 880 Νεῖλος ὑβρίζοντά σ᾽
. ἀποτρέϊψειεν, Pers. 85 f. ἐπάγει δουρικλύτοις ἀνδράσι τοξόδαμνον "Apo, 105
διέπειν ἱππιοχάρμας = 112 πίσυνοι λεπτοδόμοις πείσίμασι, Prom. 399 δακρυσί-
στακτον δ᾽ (Ὁ) am’ ὄσσων = 408 μεγαλοσχήμονά τ᾽ ἀρχαι[οπρεπῆ.ἷ
713. τὸν αἰνόλεκτρον : for this quite normal use of the article with the predicate
with verbs of calling cf. Paley, who quotes Prom. 834 προσηγορεύθης ἡ Διὸς
κλεινὴ δάμαρ, Jebb on S. A7. 726, Page on E. Med. 207, Kühner-Gerth, i. 592
n. 4, A. Svensson, Der Gebrauch des bestimmten Artikels . . ., Lund 1937, 23 f.
Headlam's translation, ‘terming Paris “O thou fatal wedded/" ', catches the
tone of the passage.
D. S. Robertson has defended his division of the words κικλήσκουσ᾽ Arapıv,
which Murray adopts in his text, in C.R. li, 1937, 162. He does not actually
take objection to the traditional reading, but nevertheless in view of the
beginning of the chorus he considers it likely that a play on the name is
intended also in the case of Paris. In this I cannot follow him. As will be
shown, the invention of the name Helen, which the Chorus is inclined to
ascribe to an invisible power with a knowledge of the future, is very closely
linked with her daemonic character, her being something more than a mortal
woman (see on 743). Paris, on the other hand, is quite an ordinary mortal.
As throughout this play he takes second place behind the great and incompre-
hensible figure of Helen, so he has here no claim to a significant mystery
concealed in his name. Besides, there is no correspondence between the
enigmatical and prophetic εὑρετής of the name Helen and the city of Priam,
which in its misery and disillusionment curses the man whom it had once
received with jubilation.
Euripides has varied the expression: Hel. 1120 Πάρις aivéyapos (the whole
conclusion of that stanza, 1118-21, is an echo of this passage of the Ag.,
moreover 1121 πομπαῖσιν ‘Appoôtras is a variation of Ag. 748).
714 f. I will not discuss the many wild suggestions that have been made in
these two badly corrupted lines. For a considered judgement see Schneide-
win-Hense's appendix ; the best treatment of the passage as a whole will be
found in Wilamowitz, Interpr. 196. He is right in maintaining that the end
of the strophe (696 ff.) is sound: consequently the metre of that passage
(priapeum J-pherecr. holds good also for the antistrophe, and therefore
714 f. must be regarded as corrupt. As for the details, it is at least very
probable that “πολύθρηνον has been repeated from 711 or assimilated to it’;
the change to παμπορθῇ (Seidler) or παμπόρθην ‘is attractive, but where the
rest of the passage is so corrupt the alteration of a single word lacks proof’.
On the other hand, the metre of 716 is sound, and there is no objection to the
wording. ἀνατλῆναι does not elsewhere occur in Aeschylus, but both the verb
and the construction are supported by the authority of passages in the
Odyssey. Wilamowitz thinks it probable that the last clause is complete in
itself in the same way as δι᾽ " Epw αἱματόεσσαν in the strophe, and consequently
ἀμφὶ πολιτᾶν (as Auratus has interpreted the MS πολίταν) does not belong to
1 I differ from Wilamowitz as to the dividing of the stanzas Prom. 128 ff., 397 ff., and
Sept. 720 ff. My reasons for doing so I hope to give at some other opportunity.
4872.2 Ζ 337
lines 714, COMMENTARY
it ; moreover ‘it is nonsense that the city ‘has to endure bloodshed on account
of the citizens’’’,
717. λέοντος ἵνιν: one of the best emendations in the text of Aeschylus,!
published by Conington (‘J. C.’) in an out-of-the-way place (in the collection
Terminalia, Oxford 1851, 83 £), and made generally known through the
appendix to Schneidewin’s edition (1856). Perhaps the corruption is partly
due to the influence of 734 σίνος and to a reminiscence of Y 164 f. λέων...
σύτης, À 480 f. Atv... σίντην. About fis Wilamowitz on E. Her. 354 says ‘an
obsolete word which [before Hellenistic times] occurs only in Aeschylus and
Euripides;? where they got it from is uncertain. It was kept in use in the
Cyprian dialect [cf. Bechtel, Grtech. Dial. 1. 449], but since this is most closely
connected with the Homeric dialect, it is probable that Tragedy has borrowed
tus from Epic.’ )
718. &yáAakrov: this lion-cub has not been weaned in the usual way but
‘taken away from its mother’s milk’ (Passow-Crönert, s.v.) too soon: that
is obvious from the following φιλόμαστον. The helplessness of the little
creature, still incapable of any misdeed, is an important trait in the story;
cf. 141 f. If we wish to elaborate the picture, Verrall’s explanation of aydAaxrov
may be not far from the truth: ‘the dam being killed by the huntsmen who
took the whelp' (cf. Statius, Achill. 1. 168 ff. fetam Pholoes sub rupe leaenam
perculerat ferro vacuisque reliquerat antris ipsam, sed catulos adportat et incitat
ungues); another possibility is suggested by S. A7. 986 f. μή τις ὡς κενῆς
σκύμνον λεαίνης δυσμενῶν ἀναρπάσηι. Valckenaer unfortunately connected
with this passage Hesychius (1) dydAaxros- ἡ ὁμόθηλος, (2) ἀγάλακτες σύγγονοι,
ἥλικες, ὁμογάλακτοι (‘hoc si probas, verteris a foster-brother' Blomfield) and,
in spite of Ahrens’s vigorous protest, the ghost is not yet laid. Wecklein
and Headlam misuse the form ἀγάλαξ, ayadaxres preserved in Hesychius and
other lexicographers (also in Call. Hymn. in Ap. 52) in order to alter the MS
ἀγάλακτον to -«ra and so obtain the short syllable which they require for
Wecklein’s Bovras (in place of οὕτως). Wecklein explains his &ydAakra . . .
φιλομάστων (sic) : ‘als Milchbruder der Säuglinge (seiner Herde)’, soapparently
the lion-cub was sucking the cow’s udder, which the poor creature would
presumably find difficult to reach.
οὕτως : how thoroughly the word is in place has been recognized by
! Tucker observes on A. Suppl. 42: ‘tus is apparently only used of human beings, a
consideration which makes Conington's λέοντος inv very doubtful’. (Walde in his article
on lus, Glotta, xiii, 1924, 127 ff., ignores Ag. 717.) Tucker’s argument could be used as an
objection to the unparalleled use of παίδων in Ag. 50 (and with greater plausibility since
mais is of far commoner occurrence). It is perhaps more than mere coincidence that ἦνιν
is used here in the same unusual way as παίδων there. For just as the carrying off of the
young birds in that passage illustrates the carrying off of Helen, so here the story of the
lion’s whelp is not being told for its own sake, but as a background for what is next to be
said about Helen. This might partly account for the anthropomorphic phrase, while at the
same time the fact that fs was a more exalted word than the ordinary σκύμνος is likely to
have influenced the poet in his choice. A transiator should not disregard the peculiar nuance
of lus. We must not render it by ‘cub’ or ‘whelp’ as if it were σκύμνος, but by something
like ‘offspring’ (Headlam). [Tucker’s objection had been anticipated by van Heusde.]
2 It is quite uncertain whether it was really ms that was written in the margin of fr. 2,
col. 1 of the papyrus of Sophocles’ Σκύριοι (Pap. Oxy. 2077, cf. R. Pfeiffer, Philol. Ixxxviii,
1933, 2, 13), as Pfeiffer informs me, who re-examined the papyrus after the publication of
his article. Pfeiffer, 13, quotes A. Suppl. 251 (trimeter) and thus corrects the mistake in
L-S s.v. tus (‘Trag. only in lyr.’).

338
 COMMENTARY line 721
Klausen (‘similitudinem introducit’) and notably by Nägelsbach (‘cf. Arist.
Vesp. 1182’). Wilamowitz has established it more exactly first C.R. xx, 1906,
446 and later in his edition: “οὕτω in apologi principio Arist. Ves. 1182,
Lysistr. 785, Plato Phaedr. 237 b'.! ‘The story is told as evidence; Aeschylus
does the same. This is later forgotten and so it [οὕτω] corresponds to our
“once upon a time" ' (Wilamowitz on Ar. Lys. 785). This idiomatic use was
rightly noted in antiquity: Schol. Ar. Wasps 1182 πρὸς τὴν συνήθειαν, ὅτι
τὸν μῦθον προέταττον οὕτως, οἷον “ἦν οὕτω γέρων καὶ γραῦς", a reference to Plat.
Phaedr. 237 Ὁ follows. Phoenix in the Iliad (I 524) begins a long story in this
way: οὕτω καὶ τῶν πρόσθεν ἐπευθόμεθα κλέα ἀνδρῶν ἡρώων. In Theocritus’
Cyclops (11), after the introductory γνώμη, the story starts: (1. 7) οὕτω γοῦν
paiora diay’ ὁ Κύκλωψ κτλ. Horace starts the paradeigma of his Europa story,
presumably in imitation of old lyric (Carm. 3. 27. 25): sic et Europe niveum
doloso credidit tauro latus, in the same way EPist. τ. 18. 41 gratia sic frairum
geminorum Amphionis aique Zethi dissiluit. It is surprising that Ag. 718
οὕτως is kept back so late in the sentence, but that may be connected
with the fact that here the alvos has been transformed into the exalted
lyric style.
719. ἀνήρ. Among the arguments in favour of the conjecture βούτας (see on
718) we find ‘the need of a more particular qualification of avjp’ (Wecklein,
Appendix to his annotated edition). In reality the unqualified ἀνήρ, with
which a character of the story is introduced, is a typical feature of the
terse style of the old alvos. This has maintained itself still in our late
version of the fables of Aesop, e.g. (I quote, in spite of the well-known objec-
tions, from Halm's edition) 635 Eis ἔριν ἀλλήλοις ἀνήρ τε καὶ λέων προῆλθον,
64 "Ανθρωπόν ποτε λέγεται πρὸς Σάτυρον φιλίαν σπείσασθαι. This way of begin-
ning an αἶνος can be established for the fifth century: Antiphon the Sophist
(Diels-Kranz, Vorsokr. ii, sth ed., 361) ἔστι δέ τις λόγος, ws dpa ἰδὼν ἀνὴρ
ἄνδρα ἕτερον ἀργύριον ἀναιρούμενον πολὺ ἐδεῖτο κτλ. Cf. Aelian, Nat. an. 8. 7
Ἀριστόξενος δέ πού φησιν (fr. 133 Wehrli) ἄνδρα ταῖς χερσὶν ὄφιν τινὰ ἀποκτεῖναι
κτλ. Luckily it can be shown that it makes absolutely no difference whether
in such an introduction the person of whom the alvos will tell is referred to as
ἀνήρ (or ἄνθρωπος) or as ἀνήρ tis. In Athenaeus 13. 606 a the beginnings of
two parallel alvoı follow one another, one from a comedy of Alexis, the other
from Philemon (for the form of the latter cf. Rhein. Mus. Ixxiii, 1920, 368).
Philemon says: ἀλλ᾽ ἐν Σάμωι μὲν Tod λιθίνου ζώιου ποτὲ ἄνθρωπος ἠράσθη τις,
while Alexis says: λιθίνης ἐπεθύμησεν κόρης ἄνθρωπος ἐγκατέκλεισέ θ᾽ αὐτὸν τῶι
νεώι. Besides, it may probably be assumed that in the present passage, just as
in some of the beginnings of stories I have quoted, the pointed contrasting
of the two distinct characters in the story (i.e. in this case: lion-cub, man)
accounts for the employment of the plain ἀνήρ.
720. In contrast to 65 (see ad loc.) and 227, προτέλεια, the “Voropfer’, here
contains no undertone of horror. It is questionable whether we get a hint
of offerings and sacrifice to come, in the sense that the preliminaries indeed
are free from them, but the actual life of the lion will be spent in bloody
slaughter and he will be proved to be ἱερεύς τις “Aras.
721. Paley quotes Arist. Hist. anim. 9. 44, p. 629 Ὁ 10 ἔστι δὲ (scil. 6 λέων) τὸ
1 These three passages are rightly explained in Passow’s dictionary s.v. οὕτω 1 ἃ (and
subsequently in L-S οὕτως I. 4).

339
line 721 COMMENTARY

ἦθος οὐχ ὑπόπτης οὐδενὸς οὐδ᾽ ὑφορώμενος οὐδέν, πρός τε τὰ σύντροφα καὶ συνήθη
σφόδρα φιλοπαίγμων καὶ στερκτικός.
722. γεραροῖς. Here the word means simply ‘the old men’. For the change
in meaning from ‘the venerable’ to ‘the old’ cf. Wilamowitz, Interpr. 39 (with
reference to Suppl. 667).
723. wodéa: this form occurs only here in the MSS of Aeschylus, but is rightly
restored in 1453: we find πολεῖ in Suppl. 745.
The MS ἔσχ᾽ has been taken by Hermann ‘neutrali significatu’ as before him
by Pauw (‘multum vero haesit in ulnis’) and by Schiitz. The objection to this
is not so much the fact that no sure proof of this use of ἔχειν is to be found in
Aeschylus (see on 190) as the tense; at any rate I am unable to see how the
use of the aorist to express a condition could be justified, more especially
since the temporal extension or the repetition of this condition is emphasized
by πολέα (243 ff., e.g., are quite different, ἐπεὶ πολλάκις... ἔμελψεν : there the
action lies in the distant past, and ἔμελπε would, as 240 ἔβαλλε, refer to the
time of the sacrifice in Aulis). Verrall understands it imaginatively ‘and many
a thing it got’, but this weakens ἐν ἀγκάλαις, and, besides, it is next door to
unavoidable, when a form of ἔχειν comes next to ἐν ἀγκάλαις, to take together
words which are elsewhere so closely connected with one another (on this
point Headlam is right). Time after time the conjecture of Auratus, φαιδρω-
mov ... oaivovra, has found approval, and, supposing we may translate the
clause thus altered ‘and often took it in his arms’, with Headlam, that
disposes of any doubts raised by the tense, but it is open to another
objection. As the MSS give it, the alvos is very deftly constructed. We do not
get the story of a man and his unhappy experiences but the story of a lion-
cub. It begins ‘once upon a time the offspring of a lion was reared’, and it is
considerably later that the less important rearer comes in, and even after
ἀνήρ is mentioned all the words to the end of the sentence (ériyaprov) apply
exclusively to the animal. Then the lion becomes formally the subject, and
remains so not only till the end of the strophe, but throughout the whole
antistrophe as well: it is from him and his transformation to a frightful
slaughterer that we pass in the next stanza to Helen and her transformation.
Such a structure should not be disturbed. Probably Casaubon and Pearson
have correctly restored éox’. The Homeric form, used Pers. 656 in a chorus,
seems particularly suitable in this tale.
725 f. σαίνων τε γαστρὸς ἀνάγκαις: 'contrarie refertur ad δαῖτ᾽ ἀκέλευστος
ἔτευξεν (731). Leo catulus eblanditur cibum quem adultus per vim rapit’
(Nägelsbach).
727. xpovioßeis: an unusual, though comprehensible, meaning of the passive
of χρονίζειν. I find the translation ‘matured by time’ in MacNeice.
ἦθος τὸ πρὸς τοκέων : since Blomfield this has been compared with Pindar,
Ol. 11. 19 f. τὸ yap éupuës οὔτ᾽ atÜwv ἀλώπηξ οὔτ᾽ ἐρίβρομοι λέοντες διαλλάξαντο
ἦθος.
728. xápw . . . ἀμείβων : so paying his τροφεῖα.
730. In itself the MS μηλοφόνοισιν ἄταις would be perfectly satisfactory, cf.
Pers. 652 (οὐδὲ... dvdpas . . . ἀπώλλυ) πολεμοφθόροισιν ἄταις, 1037 (καὶ σθένος
γ᾽ ἐκολούθη) φίλων ἄταισι ποντίαισιν. But the metre shows that a short syllable
is lacking. Bothe’s (ev) äraıs seems to me the obvious addition (σὺν is less
probable). Ahrens, whose suggestion dirais has been adopted by Wecklein,
340
 COMMENTARY line 736
says (p. 553): ‘the objection to ἐν ἄταις or σὺν draus, alterations which would
be most probable otherwise in spite of the repetition of ἄτας in 735, is the gloss
πολέμοις in Fa (ie. Tr], which refers unmistakably to another word, as
Bamberger has rightly seen’. This is a wrong conclusion. The commentary
of Proclus on Hesiod, Erga 230 f. οὐδέ ποτ᾽ ἰθυδίκηισι per’ ἀνδράσι λιμὸς
ὀπηδεῖ οὐδ᾽ ἄτη, runs ταῦτά ἐστι τὰ μάλιστα πιέζοντα τὸν βίον, πόλεμοι καὶ Mot.
731. ἀκέλευστος in place of the commoner ἄκλητος, ‘uninvited’, cf. Lysias rz.
22 ἐκέλευον συνδειπνεῖν. Dindorf (Thesaurus) quotes Plato, Laws 953 d irw μὲν
νῦν πᾶς ἀκέλευστος ὁ τοιοῦτος ἐπὶ τὰς τῶν πλουσίων καὶ σοφῶν θύρας.
733. οἰκέταις: Triclinius puts the two possibilities sensibly in his own
("Nperepov') paraphrase: οἰκέτας λέγει 7) τοὺς τῶν θρεψαμένων αὐτὸν δούλους,
ἢ καὶ αὐτοὺς ἐκείνους τοὺς θρεψαμένους. The first has been adopted e.g. by
Stanley (‘famulis’), Paley (‘the servants’), Kennedy, Wilamowitz (‘das
Gesinde’), and Blass (on Cho. 737); the second e.g. by Humboldt (‘den
Bewohnern grauenvoll’), Verrall, Headlam, Platt. The rendering ‘the
servants’ could at a pinch be defended by saying that they are in charge of
the creature and would therefore be the first to suffer from its getting out of
hand. But this seems a petty detail, whereas the other interpretation suits
the context extremely well. It is natural that the thought should proceed
from οἶκος to οἰκέται (‘inhabitants of the house’, ‘members of the household’).
ἄμαχον ἄλγος oix. and μέγα σίνος πολυκτ. are in apposition to the sentence
αἵματι δ᾽ οἶκος ἐφύρθη. It is difficult to reproduce this in translation.
732 ff. αἵματι δὲ... προσεθρέφθη. If it was intended that these sentences
should be applied only to the lion, who has mangled a few of his foster-
father’s sheep, in that case the expressions ἄμαχον ἄλγος οἰκέταις and especially
ex θεοῦ δ᾽ ἱερεύς τις Aras κτλ. would indeed be sadly disproportionate to the
subject. But it is obvious that underneath and along with the richly orna-
mented theme of the comparison the other theme of the arrival of Helen and
the effect she produces is perceptible and gradually gains in importance;
it claims for itself the attention of the hearer with ever increasing force, until
finally at the beginning of the next stanza only this theme is heard. We have
already called attention, 59, to a similar phenomenon on a more modest scale,
cf. also on 395.
735. ἐκ θεοῦ (= ἐκ θεῶν, θεόθεν), ‘by the will of the gods, by divine decree’,
cf. Headlam, On Editing Aesch. 106 f., A. C. Pearson on Soph. fr. 326.
ἱερεύς is explained by σφαγεύς in E. Her. 451, cf. Wilamowitz ad loc. on the
development of the meaning.
736. προσεθρέφθη. For the presumable cause of the corruption cf. Headlam,
On Editing Aesch. 105 f. The distribution of ἐθρέφθην and ἐτράφην as well as
of ἐστρέφθην and ἐστράφην in earlier Greek is discussed by Solmsen, Glotta, ii,
1910, 309 ff., who looks on the forms ἐθρέφθην ἐστρέφθην in Aeschylus, Euripides,
and Plato as Ionisms; cf. H. W. Smyth, The Ionic Dialect, 524 f. With the use
of προστρέφειν here Hase (followed by Dindorf, Lex. Aesch.) had compared Teles
(p. 40. 8 Hense) κἂν ἄλλον προστρέφειν. Headlam correctly explains (On
! The observation that the scholia put πόλεμος occasionally as a gloss on dry provides
the clue to the correct form of the scholion on A. Cho. 74 Weckl. There the words 6 ἐστι
πόλεμον break the connexion; Wecklein, vol. ii, Append. propag. 369, notes: ‘secludit
Buresch’. The gloss has only got into the wrong place, as so often happens: it belongs to
74 (73 Weckl.) ἄτην.
341
line 736 COMMENTARY

Editing Aesch. 108) ‘was reared as an additional inmate in the house’. The
word rounds off the αἶνος beginning with ἔθρεψεν 717.
Th. Nöldeke (‘Das Gleichnis vom Aufziehen eines jungen Raubtiers’, in A
Volume of Oriental Studies presented to Professor E. G. Browne, Cambridge,
1922, pp. 371 ff.) has made it very probable that the knowledge of this alvos
came to the poet from the East.! This fable was quite as well known in
Athens of the fifth century as the Alowrreioı λόγοι and others of the same kind.
There is nothing to show that the advice of Aeschylus in Ar. Frogs 1431 οὐ
χρὴ λέοντος σκύμνον ἐν πόλει τρέφειν alludes to the employment of the αἶνος in
the chorus of the Agamemnon and not rather to the underlying alvos itself.
There a well-known γνώμη is referred to and the warning added: ἣν δ᾽
ἐκτραφῆι τις, τοῖς τρόποις ὑπηρετεῖν. (For the strange account in Valerius
Maximus 7. 2 ext. 7 cf. Wilamowitz, Arist. u. Ath. i. 180n.; of course the
tempting attribution to the Ajo. of Eupolis remains uncertain.) It seems to
have passed unnoticed by the commentators? that the same αἶνος was familiar
to Plato, plainly from his childhood, and that he presupposes a knowledge of
it in his reader: Gorgias 483e (Kallikles) πλάττοντες τοὺς βελτίστους καὶ
ἐρρωμενεστάτους αὐτῶν, ἐκ νέων λαμβάνοντες, ὥσπερ λέοντας, κατεπάιδοντές τε
καὶ γοητεύοντες καταδουλούμεθα λέγοντες ὡς τὸ ἴσον χρὴ ἔχειν καὶ τοῦτό ἐστιν τὸ
καλὸν καὶ τὸ δίκαιον. ἐὰν δέ γε οἶμαι φύσιν ἱκανὴν γένηται ἔχων ἀνήρ, πάντα ταῦτα
ἀποσεισάμενος καὶ διαρρήξας καὶ διαφυγών, καταπατήσας τὰ ἡμέτερα γράμματα
... καὶ νόμους τοὺς παρὰ φύσιν ἅπαντας, ἐπαναστὰς ἀνεφάνη δεσπότης ἡμέτερος 6
δοῦλος, καὶ ἐνταῦθα ἐξέλαμψεν τὸ τῆς φύσεως δίκαιον. The words ὥσπερ λεόντας
are completely comprehensible only on the background of the αἶνος : they are
put in brusquely, exempli gratia, at the beginning, as if in Athens at the time
of Socrates and Kallikles it was the most usual thing in the world for a man
to rear a lion-cub, as the Persian king might do (Hdt. 3. 32. 1) or a Roman
courtesan of the Empire (Juvenal 7. 75 ff.) or Marshal Goering. That the
picture of the lion continues in the speaker’s imagination is plain from the
words διαρρήξας καὶ διαφυγών which follow, and is established by the exact
parallelism of the sequel in Aeschylus and Plato: xpoviodeis δ᾽ ἀπέδειξεν
ἦθος τὸ πρὸς τοκέων wea δὲ... φύσιν ἱκανὴν γένηται ἔχων àvijp . . . καὶ ἐνταῦθα
ἐξέλαμψεν τὸ τῆς φύσεως δίκαιον (the last, of course, one of the particular
points which Kallikles introduces into the alvos). It is impossible to say how
far the turn ἐπαναστὰς ἀνεφάνη δεσπότης ἡμέτερος ὁ δοῦλος is connected with the
second sentence in Aristophanes (ἣν δ᾽ ἐκτραφῆι τις, τοῖς τρόποις ὑπηρετεῖν) and
how far both refer to a thought in the original αἶνος.
737. πάραυτα. Stanley translates statim in agreement with Hesychius
πάραυτα" παραχρῆμα, εὐθέως, παραυτίκα (so the gloss in Tr on this passage:
ἤγουν αὐτίκα). Schütz has seen that it signifies a contrast with 744 ff. (mapa-
κλίνασ᾽ ἐπέκρανεν δὲ KrA.): ‘initio ... postea vero’. Klausen and Paley quote
the striking parallel Eur. fr. 1079. 5 N. πάραυτα δ᾽ ἡσθεὶς ὕστερον στένει διπλᾶ,
where δέ (in apodosi) can scarcely be right, but πάραυτα should not be normal-

1 Possibly the main motif of the fable of the crested lark, which according to Ar. Birds
471 was found ἐν Αἰσώπου λόγοις, also came originally from the East, as is asserted by Aelian
Nat. anim. 16. 5 (ἐξ ᾿Ινδῶν), cf. van Leeuwen’s edition of Ar. Birds pp. 266 ff., Hausrath RE
vi. 1727 (‘sounds un-Hellenic’), A. B. Cook, Zeus, iii. 44 n. 5.
2 Nor does F. M. Cornford make any comment on it, although in his book Thucydides
Mythistoricus, 188 ff., there is a special chapter, "The Lion’s Whelp'.

342
 COMMENTARY line 739
ized into παραυτίχ᾽ (L-S are right in giving it s.v. πάραυτα). In opposition to
this explanation Heath wished to read here παρ᾽ αὐτά, which he translates
similiter. This view is supported by Nägelsbach, Kennedy, Wecklein, Lewis
Campbell among others, and lately by Schuursma, De . . . abusione, 135.
Their linguistic argument (Wecklein: ‘the meaning "gleich dem", “ebenso”,
corresponds with the original sense of παρ᾽ αὐτά᾽ ; Schuursma: 'praepositio
παρα- hac abusione vim comparativam nacta est’) finds no support in any-
thing I have been able to establish about the use of παρά (cf. especially the
types of usage, apparently corresponding but really quite distinct, in L-S
1303 C. i. 6 and 7). On the other hand, the interpretation of πάραυτα as
παραχρῆμα, παραυτίκα is recommended by phrases like Demosth. 18. 13, 20.
139 παρ᾽ αὐτὰ τἀδικήματα (‘sur-le-champ, au moment même de l'accomplisse-
ment de ces delits’, Weil), 18. τς φυγὼν τοὺς παρ᾽ αὐτὰ ra πράγματ᾽ ἐλέγχους
(contrasted with τοσούτοις ὕστερον χρόνοις αἰτίας... συμφορήσας), etc., cf.
L-S 1303 C. i. 1o and especially Rehdantz-Blass, Demosth. Phil. Reden, ii.
2, 4th ed., p. 117 (‘apd with the accus.’). The linguistic evidence is un-
ambiguous and confirms the ancient interpretation of πάραυτα. It would be
a mistake to allow oneself to be diverted from this course by speculations
concerning what is apparently required to mark clearly the connexion of the
stanza 737 ff. with what precedes. Instead of postulating the employment of
a definite scheme of connexion, we should notice that the poet in three
choruses, the themes of which are closely interconnected, where he inserts a
simile into the lyric narrative of the past, arranges its link with its surround-
ings in three different ways. The first instance, 49-60, maintains the custom-
ary form: attachment of the comparison to what precedes by τρόπον
αἰγυπιῶν, at its end resumption by οὕτω. In the second example, 393-9, the
picture is first inserted by itself without any definite mark of comparison, and
only after the end of the comparison does ofos 399 bring out the application.
Finally, in the passage with which we are at present concerned, at the begin-
ning of the simile 718 οὕτως (which we may hope no one will any longer
question) points to a connexion, and then after the end of the simile, without
any reference back, the narrative at 737 proceeds quite simply.! Suppose we
struck out the whole pair of stanzas 717-36, then the following strophe
(737 ff.) would indeed completely lose that wonderful transparency thanks
to which, as it is, the comparison with the lion persistently shines through the
picture of the beautiful fiend; but we should not get what is commonly
called a break in the connexion or even a roughness in the joining of the
sentences. |
739. λέγοιμ᾽ ἄν: cf. on 838.
φρόνημα: easier to understand than to translate adequately. Aeschylus’
1 Wilamowitz, Interpr. 187, rightly remarks : ‘An image from the animal world . . . which
without further explanation illustrates the real character of Helen’. The connexion
between the illustration and the main theme can safely be left to the audience (or the
reader) : we may quote as an example of this, from a work in an entirely different style, the
beginning of Horace’s letter to Florus (Epist. 2. 2). There in ll. 2-19 we have an illustration
drawn from the world of business and the law, which is followed in l. 20, without any
formal connexion, by its application, dixi me pigrum etc. Afterwards we find precisely the
same arrangement repeated : 1. 26 introduces a fresh anecdote with Zuculli miles, continuing
till 40, after which the application of it follows, again without formal connexion, and a return
to the theme proper with Romae nutriri mihi contigit.
343
line 739 COMMENTARY
usage shows as the dominant meaning ‘mind, spirit, disposition’ and the like,
in which especially the two opposite poles, usually distinguished by epithets
(not in Prom. 953), insight and arrogance, or, regarded in another light,
mildness and hardness (scorn, pride), play a part. The essence of Tragedy,
which so often portrays excess and presumption, brings it about that in
Tragedy the second, so to speak negative, pole predominates, so, e.g., Pers. 808,
Sept. 438, 536 f. (though there the opposite pole is mentioned too), Prom. 207,
Cho. 191, 595, 996. It occurs besides in a neutral sense, Suppl. 100, Cho. 323.
In Ag. 739 f. the addition of νηνέμου yaAdvas draws the word to the side of
the milder sense.
741. äkaoxatov. Hesychius glosses the adverb ἄκασκα by ἡσύχως, μαλακῶς,
βραδέως (cf. Phot. Berol. 58. 19). Its occurrence in anapaests in Kratinos
fr. 126 K. is probably to be attributed to the influence of more elevated
poetry (with the thought there cf. Ag. 72 ff.). This is the only place where
the adjective occurs; in the same way as the following μαλθακόν it carries on
the characteristic conveyed in νηνέμου γαλάνας.
ἄγαλμα πλούτου, i.e. ὧι ὁ πλοῦτος (or à πλούσιος οἶκος) ἀγάλλεται, is very
similar to 208 δόμων ἄγαλμα.
The line contains an unsolved difficulty. The four nominal clauses from
740 dpövnpa . . . yaldvas to 743 δηξίθυμον. . . . ἄνθος show a striking regularity in
their parallel arrangement. It is only natural that almost all translators
render (unintentionally, we may well think) the four components of the list
asyndetically, and that Headlam remarks: “This passage affords a remarkable
instance of a common formula of description, in which the details are
accumulated without any connecting particles.” The correct feeling of trans-
lators and editors for what is expected here is all the more remarkable
because they almost all accept a text which offers the addition of δ᾽ or 7’ in
741, with the consequence of spoiling the asyndeton. The addition is enforced
by the responsion (741 = 754 τεκνοῦσθαι μηδ᾽ ἄπαιδα θνήισκειν) though not by
the metre in itself. For in itself v--uu-u-- would here be not only
tolerable, but very good. It is the hipponacteum, for which Hephaestion ro.
2 p. 32. 20 Consbr. quotes καὶ κνίσηι τινὰ θυμιήσας, i.e. an example with long
first syllable, but a short first syllable is quite legitimate, e.g. Ag. 1488
(= 1512) τέ τῶνδ᾽ οὐ Ücókpavróv ἐστιν; S. Oed. C. (133 =) 165 κλύεις ὦ πολύμοχθ᾽
dÀára.! The hipponacteum ἀκασκαῖον ἄγαλμα πλούτου would be in itself not
only unexceptionable, but also very suitable for the place in which it stands,
following on seven iambic metra and forming the transition to the two con-
cluding lines of this period which are to be discussed immediately. A similar
form of the hipponacteum syllable for syllable follows, as here, upon seven
iambic metra also in Ag. 1488. So far as the period Ag. 737-43 is concerned, it
cannot pass unnoticed first how nearly related (to say the least, for the two
lines are probably to be regarded as metrically equivalent)? ἀκασκαῖον ἄγαλμα
πλούτου
— — » v — v — — and the closing line 743 - v- vu — o — — are to one
another, and secondly that in the first strophe of this same stasimon the first
1 A. M. Dale, in Greek Poetry and Life (Oxford 1936), 196, rightly agrees with Wilamowitz,
Verskunst, 249, in recognizing the hipponacteum in this line.
2 The relation between the two lines is exactly the same as between τὸν ἀργῆτα Κολωνόν,
&0' and ἃ λίγεια μινύρεται or between ἐνικήσαμεν ὡς ἐβουλόμεσθα and φίλταθ᾽ ‘Apydé&’ οὔ τί
που τέθνηκας.

344
 COMMENTARY line 743

metric period is closed with a couplet (686 f. = 705 f.) τὰν δορίγαμβρον ἀμφινεικῆ
θ᾽ “Ἑλέναν; ἐπεὶ πρεπόντως, whose first line is identical with 742 μαλθακὸν
ὀμμάτων βέλος, while the next is, superficially, only distinguished from 743
δηξίθυμον ἔρωτος ἄνθος by the position of the dactyl, just as πολλὰ τὰ δεινὰ
κοὐδὲν ἀν(θρώπου) is distinguished from τοῦτο καὶ πολιοῦ πέραν, in other words
represents an equivalent variation. But these amusing metrical speculations
would only be helpful if it were possible to establish that the line of the
antistrophe (754) τεκνοῦσθαι μηδ᾽ ἄπαιδα θνήισκειν (3 iamb.) has not been
preserved in its original form. But there is no sign whatever of that: he
would be bold who would venture to meddle with these vigorous words
(Hartung’s τέκνων μήποτ᾽ ἄπαιδα θνήισκειν carries its own antidote). So
nothing remains but to adjust 741 metrically to 754. It is easy to understand
that efforts have been made to achieve this without spoiling the asyndeton.
But the attempts have failed, whether the beautiful ἄγαλμα (see above) has
been defaced or πλούτου changed to πλούτων and ἀκασκαίωνϊ written. Porson
inserted δ᾽ before ἄγαλμα. But then, especially with the parallelism of
φρόνημα yaldvas and ἄγαλμα πλούτου, it becomes almost inevitable to connect
μέν (740) and δέ together. μέν, however, is picked up by the δέ in 745 (for the
antithesis of 744 ff. and 737 ff. in general cf. above on 737 πάραυτα), as e.g.
Hartung, Karsten, Paley, and Keck have remarked. Consequently I prefer
Hermann’s 7’ and put it in the text, admittedly as a last resort, for I cannot
see why in the series of four expressions the first and second should be
more closely connected than one or more of the other elements with others.
742. ὀμμάτων βέλος. Following Homer's (8150) ὀφθαλμῶν... . BoAai Aeschylus
says fr. 242. 2 N. Breppdrwyr . . . βολή (a certain emendation). A full collection
of similar expressions is given by A. C. Pearson, C.R. xxiii, 1910, 256 f. and on
Soph. fr. 157 and 474. van Heusde cites Theocritus' (18. 37) 'EAéva τᾶς πάντες
ἐπ᾽ ὄμμασιν ἵμεροι ἐντί.
743. δηξίθυμον: Blomfield compares Soph. fr. 757 N. (= 841 P.) ὅτωι δ᾽
ἔρωτος δῆγμα παιδικὸν (for the expression cf. Wilamowitz, Kl. Schr. 1. 178)
προσῆι. Headlam, On Editing Aeschylus, 102 f., gives plenty of instances of the
‘metaphorical’ use of δάκνειν: the use as here with reference to desire is much
less frequent than that of pain, grief, etc.
ἔρωτος ἄνθος needs no explanation. But in connexion with it we are glad
to remember the picture of the rape of Helen which Makron, a contemporary
of Aeschylus, has given in his ‘masterpiece’ (Beazley, Attic Red-figured Vases
in American Museums, ı01) on the skyphos in Boston (Furtwängler-Reich-
hold, pl. 85). There Eros floats near Helen as she follows Alexandros, busying
himself as he flies in arranging the locks of the fair one to the best advantage.
The feminine? breaks in first 744 with παρακλίνασα ; till then we have only
heard that what came to Ilion was a φρόνημα, an ἄγαλμα, etc. We need not
lay emphasis on the ‘abstractness’ of these expressions, for that term is
scarcely in the case of φρόνημα and certainly not of ἄγαλμα adequate to the
! The word ἀκασκαῖον itself has happily been left undisturbed. The gloss on the word
(found also in ΕἾ, λίαν κεκοσμημένον, is merely an unfortunate attempt to guess at the
meaning. Platt’s suggestion (7. Phil. xxxii, 1913, 60) ἀκασκαίοι᾽ ἄγ. need not be considered.
2 This transition to the real gender of the subject is in itself natural: Hor. Od. 1. 37. 20 f.
(of Cleopatra) daret ut calenis fatale monstrum; quae generosius perire quaerens etc. Cf. also
above on 79. The same thing in a nutshell is ζ 157: λευσσόντων τοιόνδε θάλος χορὸν εἰσοι-
χνευσαν.

345
line 743 COMMENTARY
early Greek conception ; besides, the ‘concretes’ βέλος and ἄνθος are parallel
with them. The important thing here is the avoidance of the person. The
contour of the human figure is effaced, the bounds of her personality seem to
be widened while first a mood, then a delight, then the arrow of the eyes, and
lastly the fullness of the charm living in the flower take her place. The whole
of what is here described has no shade of indistinctness, but rather is con-
ceived and expressed with perfect clearness: the individual traits combine to
form a figure whose many-coloured bewitchingness, gentle and at the same
time powerful, lays the hearer under its spell. The air of impersonality, of
super-personality, raises Helen above the merely human, removes her from
among her kind, and brings her close to unknown Powers. The effect pro-
duced by the riddle at the beginning of the chorus is intensified here, and so
our minds are prepared for the Erinys at the end of this stanza.
745. For the position of δέ cf. on 653.
The τέλος γάμου (τελευτή in Suppl. 1051, Pindar, P. 9. 66) is here reversed
into its unholy opposite in a way characteristic of Aeschylus (see on 65). E.
Med. 1388 πικρὰς τελευτὰς τῶν ἐμῶν γάμων ἰδών contains perhaps a reminiscence
of this passage.
746. δύσεδρος. καὶ δυσόμιλος : the only occurrence of these adjectives in pre-
Hellenistic Greek. There is no need to look for a far-fetched meaning of
δύσεδρος (with Kennedy, p. 221, and Bonner, Class. Philol. xxxvii, 1942,
263 f., whose interpretations move on the same lines as Hermann’s note on
118 παμπρέπτοις ἐν ἕδραισι).
747. Πριαμίδαισιν : Wilamowitz’s statement, Ar. und Ath. ii. τ82, that here
the children of Priam cannot be meant but only the race as a whole, seems
arbitrary. Helen’s arrival is a κακὸν Πριάμοιο τέκεσσιν, cf. also Alcaeus fr. 74.
2 D. The prosody of Πρῖαμ. results ‘ex epicorum auctoritate’ (W. Schulze,
Quaest. ep. 151, cf. also A. Seidler, De vers. dochm. 21 n., Wilamowitz, Hom.
Untersuch. 325 n. 41) ; cf. 537 for the same syllable short.
748. πομπᾶι: ᾿πομπή is the act of escorting. The god who sends someone
stays and remains with the person through whom he is operating’, Wilamo-
witz on E. Her. 580, with examples.
749. νυμφόκλαυτος occurs only here, probably coined for this passage.
In view of the context we would be glad to translate it with Blomfield by
the reinterpreted Horatian phrase 'flebilis sponsa’, similarly e.g. Wecklein :
‘eine Tränenbraut’, Wilamowitz : ‘der Braut, der hóllischen, tränengefreiten’,
L-S: ‘a bride bringing woe’. That seems obvious since all we are concerned
with here is the wedding and marriage of Helen; she is the bride in whose
honour, along with Paris, the νυμφότιμον μέλος (705) has been sung, which still
rings in our ears. But it seems questionable whether such an interpretation
is linguistically permissible. In a possessive compound no hesitation would
be felt in assuming a free relationship of this kind (cf. Williger, Sprachl.
Unters. τό ff.), but the formation with a verbal adjective ending in -ro- seems
to make it necessary to understand it in the sense either of ‘which will be
wept for by the bride (or brides)’ (cf. νυμφόληπτος) or perhaps (though this
does not apply here) ‘she who weeps for brides’. Consequently it may be
more prudent to follow in substance the rendering which Blomfield suggested
as an alternative to the interpretation mentioned above, ‘sponsis deflenda’,
so e.g. Conington, Nägelsbach, B. Todt (De Aesch. vocab. inventore, Halle
346
 COMMENTARY line 753
1855, 41), the older editions of L-S, and Headlam: ‘bringing tears to brides’.
The slight modification suggested by Ch. Edw. Bishop, De adiectiv. verbal. . .
usu Aeschyleo (Diss. Leipzig 1889) 10, ‘quae a sponsis deflebatur', seems com-
mendable. If it at first seems strange that ‘bride’ here does not refer to Helen,
it is true on the other hand that, whenever she is mentioned, this thought may
be suggested: ᾿Ελένα, μία τὰς πολλὰς... ψυχὰς ὀλέσασ᾽ ὑπὸ Τροίαι.
As the lion showed itself in the end (735) as ἱερεύς τις "Aras, so Helen appears
finally as Erinys. The beginning of the whole chorus as well as of this stanza
had prepared the way for regarding her as adaemonic being. The identification
should not be weakened by an arbitrary rendering. It is of the same strength
of immediate vision as when later (1500 ff.) the deeply agitated Clytemnestra
takes refuge in the idea that the ancient avenging spirit is appearing in the
. figure of Agamemnon’s wife. Aeschylus attributes to both daughters of
Tyndareos, at the root of their personality, something belonging to the sphere
of the superhuman, something of the essence of evil spirits.’ Euripides has
adopted this conception in his Alexandros if, as is probable, the words of
Ennius, trag. 56 Ribb. (part of Cassandra’s prophecy) quo iudicio Lacedae-
monia mulier, Furiarum una, adveniet go back to the original.?
750 f. τέτυκται going with a predicate (παλαίφατος), as in Homer (cf. L-S
τεύχω III), simply in place of the substantive verb.
γέρων λόγος: cf. Cho. 314 τριγέρων μῦθος, Aesch. fr. 331 N. ὡς λέγει γέρον
γράμμα.
751 f. The gloss γεγονότα in Tr on τελεσθέντα expresses the view that μέγαν
is predicative. This is also the view of modern editors, e.g. Paley, Nägelsbach
(μέγ. τελ. prolepsis est: ὥστε μέγαν elvac’), Verrall, Plüss, Headlam, Mazon.
Cf. e.g. Cho. 791 μέγαν ἄρας (262 f. ἀπὸ σμικροῦ δ᾽ ἂν ἄρειας μέγαν δόμον). The
word-order is in favour of this interpretation.
μέγας ὄλβος is a common combination, e.g. Pers. 826, Cho. 865, Pind. Ol.
1. 56, E. Ion 703, Or. 340, 807. Here μέγαν τελεσθέντα, denoting maturity
(τέλος) and full growth, serves to make ὄλβος appear as a living creature, a
notion which is carried on and elaborated in the subsequent words. μέγας
of the full-grown animal, opposed to veapds, cf., e.g., 358.
753. For the genealogical conception in general cf. on 386, for τεκνοῦσθαι κτλ.
cf. Cho. 649, Eum. 532 f., Aesch. fr. 315 N. τῶι πονοῦντι δ᾽ ἐκ θεῶν ὀφείλεται
τέκνωμα τοῦ πόνου κλέος.
μηδέ: the use of this negative after λόγος τέτυκται is surprising, but no
editor seems to have thought it worthy of attention, and it is neglected by the
grammarians. Even Gildersleeve, Am. Journ. Phil. i, 1880, 49 f., and Wacker-
nagel, who in Syntax, ii. 282 f. carefully explains how un with the infinitive
steadily gains ground even where the infinitive expresses something which is
a fact, do not mention this passage, nor Jebb on S. Oed. R. 1455 (where he
should have quoted A. Pers. 431, 435). Possibly we should see in this λόγος
τέτυκται... μηδὲ... θνήισκειν a notable and isolated forerunner of the pas-
sages in Plato and Xenophon which Gildersleeve quotes loc. cit. (followed by
Goodwin § 685, p. 270; Kühner-Gerth, ii. 193 f. is unsatisfactory) : Plat. Rep.
346 e, Theaet. 155 a, Xen. Mem. 1. 2. 39, Cyrup. 7. 1. 18, where, in a statement
1 Headlam’s reference to Paris being identified with Erinys in Pind. Paean 8. 30 must be
given up, according to the conclusive exposition by C. Robert in Hermes, xlix, 1914, 315 f.
2 Cf. Snell, Hermes, Einzelschriften, Heft 5 (1937), 7, 28.

347
line 753 COMMENTARY
depending on φάναι or λέγειν, an infinitive with μή is found, e.g. καὶ ἄρτι
ἔλεγον μηδένα ἐθέλειν ἑκόντα ἄρχειν. The same occurs with ὁμολογεῖν Plat.
Prot. 336 Ὁ Swxpdrns . . . ὁμολογεῖ μὴ μετεῖναί oi μακρολογίας. But the distance
between the language of these writers and that of Aeschylus is considerable.
Cases like Thuc. 1. 139. ı are completely different. Cf. also (for something only
remotely related) A. C. Pearson on E. Phoen. 1175.
755. This seems to be the first occurrence! of ἀγαθὴ τύχη although we can
scarcely doubt that the conception is earlier. Sappho already has τύχαι σὺν
ἔσλαι (fr. 31. 4 D.). ἀγαθὴ τύχη does not occur elsewhere in Tragedy, cf. Gerda
Busch, Unters. z. Wesen der Τύχη in d. Trag. des Euripides, Diss. Heidelberg
1937, τό, 61, but Ag. 1230 seems to presuppose a common use of ἀγαθῆι τύχηι.
The earliest instance in Attic inscriptions is apparently the formula ἀγαθῆι
τύχηι τῆι ᾿Αθηναίων in the well-known decree about the Chalcidians 446/5 B.c.,
IG. i.* 39 (= Dittenberger, Syll. 64; Tod, Gk. Hist. Inscr. 42) 1. 40.
γένει: generi. (Stanley, Schütz), ‘to the race’ (Kennedy), ‘yéve zu BAa-
στάνειν, erspriesse dem Geschlechte' (Wecklein). This is the natural explana-
tion. Disaster comes to the children and children's children of the man who
had prosperity. Otherwise taken e.g. by Conington ('the generation of weal
and woe’), Verrall (‘yéve by kind, according to nature’), Headlam (‘that woe
insatiable is the natural heritage of a good fortune").
758. For the alteration of the word-order in the MSS cf. on 222.
759. μετὰ... τίκτει: ‘tmesis’, cf. on 586.
μέν ought normally to follow πλείονα ; it has, however, ‘been shifted to the
second place in the sentence [or colon] according to a widespread usage
which often is not recognized' (Wilamowitz, Platon, ii. 392, cf. also Denniston,
Particles, 372), cf. on 640. τὸ δυσσεβὲς yàp ἔργον is to be taken as a syntactical
colon, cf. Ed. Fr., Kolon und Satz, ii. 338: ‘an element in the sentence is
frequently placed at the beginning when it is intended to be the basis common
to both parts of a two-fold—often antithetical—sentence; in such cases it
can often be shown that this initial element functions as a colon.”
761. yáp has been excellently explained by Hermann: 'nam illud quidem
non dubium est, iustae domus prosperam sortem esse.' Cf. Vahlen on Aristot.
Poet., 3rd ed., p. 248: ‘membrum . . . per ydp particulam subnectitur, eo modo
quo notum est saepe in utraque lingua per causalem particulam adici quod
ita evidens esse significatur, ut vix videatur afferendum fuisse'. For the
corresponding employment of nam cf., e.g., Vahlen, Opusc. ac. i. 101 and Ges.
Schriften, ii. 658 ff., Meusel on Caes. B. Gall. 6. 13. τ, Housman on Lucan s. 749.
εὐθυδίκων. We do not know whether the adjective εὐθύδικος (first attested
in Bacchylides and Aeschylus) or the proper name Εὐθύδικος first came into
use (I consider the latter quite possible). The beautiful kore dedicated by
Euthydikos (IG. i^ 589) on the Acropolis (H. Payne and G. M. Young,
Archaic Marble Sculpture, plates 84 ff.) belongs to the second decade of the
fifth century, so the name was used in Athens in the sixth century.
762. καλλίπαις πότμος. The translations of Stanley and Schütz are vague.
Wellauer in the Lex. Aesch. translates ‘pulchram prolem habens", i.e. he takes
! It is plain from the way they are brought in by Plutarch, Sol. 3. 5, that the hexameters
‘Solon’ fr. 31 Bergk (28 D.) have nothing to do with Solon. Bergk was not deluded. Diehl
refers to the treatment of the matter by his colleague Kern, who does not take the context
in Plutarch into account.

348
 COMMENTARY line 762

the word as a ‘possessive compound’ (like e.g. ἀργυρότοξος) ; this has become
communis opinio not only in the dictionaries (Passow, and following him
L-S, Linwood, Dindorf) and in most commentaries and translations, but
also in books of modern grammarians (e.g. Debrunner, Griech. Wortbildungs-
lehre, ὃ 117; Schwyzer, Griech. Gramm. 1. 429). This brings in an idea alien to
this passage. Against it Bothe has correctly said that καλλίπαις must be taken
in the same sense here as in E. Or. 964 where it means the same as καλὴ παῖς
(‘determinative compound’, like, e.g., ἀκρόπολις), cf. also E. Her. 689 τὸν
Λατοῦς εὔπαιδα γόνον, Ih. T. 1234 εὔπαις 6 Λατοῦς γόνος. Verrall propounds
a compromise solution: ‘«aAAiraıs πότμος combines in one phrase the ideas
that the prosperity of the house is reproduced in successive generations, and
. that this prosperity is itself the child of righteousness ; as misery is of sin.’ I
think the interpretation as καλὸς παῖς quite sufficient.
On 757-62 as a whole. δίχα δ᾽ ἄλλων κτλ. : this powerful disclaimer of the
conventional impious thought corresponds completely with the protest in
the first stasimon (369 ff.) against what people (vis) say: ὁ δ᾽ οὐκ εὐσεβής. For
the theme of 758 ff. a passage in the Eumenides (532 ff.) has long been cited
as a parallel.
Wilamowitz, Griech. Lesebuch, Erläuterungen, p. 35, thought that there was
a discrepancy between the conviction of the Chorus in Ag. 757 ff. and what in
Pers. 362 is said about the φθόνος θεῶν. According to Wilamowitz the poet ‘in
his last work’, the Oresteza, ‘sharply contradicted the belief expressed in the
Persae’. It seems to me that the two passages are perfectly consistent and
that that is exactly what we should expect. Of course the thought of Aeschylus
matured and deepened as his insight into human nature and his creative
power grew, but there were in his religious and moral outlook certain fixed
principles not subject—-at any rate not to any perceptible degree—to a
process of evolution. Such thoughts as are uttered in Ag. 369 ff. and 757 ff.
have deep roots in the poet’s mind. It seems most unlikely that at any time
he should have doubted this fundamental truth: the gods (or Zeus) see to it
that sooner or later the impious man is punished while the righteous will be
spared. In regard to this conviction there is no difference between the choruses
of the Suppliants and those of the Oresieta. There is nothing to show that
Aeschylus ever entertained the cruder conceptions of φθόνος θεῶν. The pas-
sage Pers. 362 is in complete harmony with the belief which the poet expresses
elsewhere ; therefore we need not resort to the excuse that those are but words
of a messenger and consequently represent a lower standard. The sentence
τὸ δυσσεβὲς yàp ἔργον μετὰ μὲν πλείονα τίκτει κτλ, is perfectly illustrated by the
case of Xerxes. Xerxes, we are told by the messenger (and through him by
the poet), did not understand that it was the φθόνος θεῶν that turned against
him and led him to destruction. Blinded by a μέγας δαίμων (725), the king
endeavoured to upset the rule by which the gods have arranged the limits of
land and sea: θνητὸς ὧν θεῶν δὲ πάντων dur! οὐκ εὐβουλίαι καὶ Ποσειδῶνος
κρατήσειν (149 f.). The ghost of Darius prophesies (816 ff.) that the heaps of
slaughtered men on the battlefield of Plataea will tell future generations ὡς
οὐχ ὑπέρφευ θνητὸν ὄντα χρὴ φρονεῖν. ὕβρις yap ἐξανθοῦσο᾽ ἐκάρπωσεν στάχυν ἄτης
κτλ. Τὸ crave for the λίαν, the ἄγαν, is ὕβρις in the true sense of the word.
1 As is well known, the interpretation of Eum. 1033 Νυκτὸς παῖδες ἄπαιδες is highly
debatable.
349
line 762 COMMENTARY
Such trespasses of a mortal are invariably stopped by the god. God has drawn
limits and fixed them for ever; he will not allow any man to overstep them.
This, and this alone, is the purified idea of φθόνος θεῶν, as we find it in Pindar
(cf. O. Schroeder, commentary, 1922, on Pyth. 10. 1 and Neue Jahrbücher für
d. klass. Altert. 1923, 143) and Aeschylus.” ‘The φθόνος not only gives satisfac-
tion to the injured honour of the gods but consistently works for the realiza-
tion of δίκη (Daube, Rechtsprobleme, 133). Cf. p. 463.
764. νεάζουσαν: not ‘which wantons’ (Paley), but simply ‘which is young’.
Also in the passage which is always compared with this, Suppl. 104 ff. ἰδέσθω
δ᾽ εἰς ὕβριν βρότειον, ota νεάζει, πυθμὴν . . . τεθαλώς, the verb has the same
simple meaning (‘the old stem is young again’). So E. Phoen. 713 νεάζων
means ‘being young’, whatever may be the undertone suggested by the
context (cf. A. C. Pearson, ad loc.).
765. ἐν κακοῖς βροτῶν. Many editors take κακοῖς as masculine, e.g. Schiitz,
Schneidewin, Kennedy, Wecklein, Verrall, Headlam, Platt, Mazon. On the
other hand Linwood, Lex. Aesch. vedlew: “ἐν κακοῖς βροτῶν is not the same
as ἐν τοῖς κακοῖς βροτῶν, or ἐν κακοῖς βροτοῖς, but refers to the misfortunes of
those men who are made to suffer by the crimes spoken of’. Similarly e.g.
Paley (‘in the misfortunes of men’), Wilamowitz (‘mit fremdem Leide’),
Murray (‘amid the tears of men’). It is obvious that this way of taking the
word (κακά, not κακοί) gives a much better meaning in the context,? because
thus the sphere in which the ὕβρις νεάζουσα works is indicated. But since this
consideration may appear subjective, it will be better to give the decisive
weight to the usage of the word. As will be pointed out on 1612, ἐν κακοῖς,
‘in (amid) misfortunes, misery’, is a very commonly used expression (also
attested along with ὑβρίζειν) ; in combination with the genitive e.g. S. Aj. 1151
ἐν κακοῖς... τοῖσι τῶν πέλας. We should not here unnecessarily suppose a
departure from this established usage.
So far as the text of 763 ff. is concerned, it appears at first that everything
is simple down to 766 μόληι. At any rate we should not, as e.g. Wilamowitz
and A. Y. Campbell have done, meddle with τότ᾽ ἢ τότε 766. Blomfield com-
1 This is now common knowledge. The brief but fundamental observations by W. v.
Humboldt, Aesch. Ag. p. vf. (= Gesammelte Schriften, viii. 121), were elaborated by
Welcker, Griech. Gétterlehre, iii. 31 f. To this we may add the pertinent remarks of Erwin
Rohde, ΚΙ. Schr., ii. 329. A. B. Drachmann evolved a complicated theory concerning the
various aspects of Nemesis ; his view was partly adopted, partly modified, by the sociologist
Svend Ranulf, The Jealousy of the Gods and Criminal Law at Athens, 1933/4. Ranulf seems
to regard the highly artificial ‘types’ of his classification as something real, an objective
criterion by which to judge the evidence contained in the Greek texts. He speaks over
and over again of ‘confusion of Types I and III in Aeschylus’, ‘confusion of Types II and
III in Herodotus’, etc. (vol. i, pp. 90 ff., 119). I acknowledge the seriousness of his question-
ing, but feel quite unable to follow him. The chief fact to remember is that the φθόνος
ascribed to the gods was originally something crude and primitive. How else could it have
been called ¢8évos? The belief that certain actions of the gods originated from so base an
instinct must go back far beyond the period at which we learn of it from our literary sources.
When the idea first comes to our notice, it seems already well on the way to a more dignified
and purer conception.
2 Perhaps in Herodotus too; cf. Regenbogen, D. humanist. Gymnasium 1930, 12; F.
Hellmann, ‘Herodots Kroisos-Logos’, Neue Philol. Unters., Heft 9, 45; Pohlenz, Herodot
(1937), 114 f.
. 3 This was also Triclinius’ interpretation of ἐν κακοῖς, although he makes the mistake of
referring it to the conquest of Troy, whereas the words are used in a quite general sense.

350
 COMMENTARY line 766
pares E. Andr. 851 f. συμφοραὶ θεήλατοι πᾶσιν βροτοῖσιν ἢ τότ᾽ ἦλθον 1) τότε.
The idea of a dies certus an, incertus quando is important. Either the birth of
the younger Hybris, who embodies Ate and final destruction of the house
(770) and race, follows on the working of the elder immediately (or
possibly after a short interval); or else the proverb μενετοὲ θεοί (ὀψὲ θεῶν
ἀλέουσι μύλοι) proves true: Solon fr. 1. 29 ff. D. ἀλλ᾽ ὁ μὲν αὐτίκ᾽ érewev, 6 δ᾽
ὕστερον (and, supposing the sinners themselves escape the θεῶν μοῖρα, ἤλυθε
πάντως aûris* ἀναίτιοι ἔργα τίνουσιν, ἢ παῖδες τούτων 7) γένος ἐξοπίσω) ; cf. Eur.
fr. 979 N. dikn . . . σῖγα καὶ βραδεῖ ποδὲ στείχουσα μάρψει τοὺς κακόυς, ὅταν
τύχηι. But emendation is not easy for ὅταν, which is proved to be corrupt by
the metre. Here nearly everybody has been satisfied with Klausen’s ὅτε (on
the syntactical point cf. Kühner-Gerth, ii. 449 n. 4), but this is open to strong
objection. It causes, supposing no word is cut out (see below on this point),
the responsion 766 ὕβριν, τότ᾽ ἢ τόθ᾽ ὅτε = 776 βίον, rà χρυσόπαστία, i.e.
V—-u-uvucu-u-vc-, Now in lyric iambics the responsion between a
long and two shorts is in itself by no means alien to Aeschylus, cf., e.g., Pers.
256 dv’ dvıa κακά, νεόκοτα = 262 7) μακροβίοτος ὅδε γέ τις, 1042 tule μέλος ὁμοῦ
τιθείς = 1050 ἐπορθίαζέ νυν γόοις, Sept. 779 ἐγένετο μέλεος ἀθλίων = 786 ἔφηκεν
ἐπίκοτος τροφᾶς, 834 κακόν με καρδίαν τι περιπίτνει κρύος = 842 βουλαὶ δ᾽
ἄπιστοι “Λαΐου διήρκεσαν (in some respects Sept. 850 — 858 is doubtful; in Pers.
269 = 275 the text of the antistrophe is quite unreliable ; with Wilamowitz’s
ingenious but arbitrary reconstruction there would actually be in one dimeter
three instances of this inexact correspondence). But in the cantica of the
Oresteia no certain instance of a responsion of the kind is to be found,’
although there periods of pure iambics, some of them of considerable length,
play a great part and in them numerous instances occur of a long resolved
into two shorts. For Ag. 408 (πολὺ δ᾽ aveorevov F) and 1453, where πολλὰ
(= 1473 κόρακος) has been emended to πολέα, see ad locc. The single apparent
exception is Cho. 628 ἐπικότωι σέβας = 636 οἴχεται γένος, but there the im-
possibility of ἐπικότωι σέβας (in an otherwise very corrupt stanza) has been
recognized by many critics. So anyone who in Ag. 766 reads ὅτε and makes
it equivalent to the long syllable in the antistrophe is introducing into the
text by conjecture a peculiarity which departs from the established practice
of the Oresteia. I would not venture to do that, although I admit the possi-
bility that the absence of a responsion of this type in the Trilogy, where there
are so many opportunities for its occurrence, may nevertheless be due to
chance. The difficulty would be avoided if, while adopting óre in 765 and
ἔδεθλα in 776, we deleted fiov in 775; then we get a correspondence syllable
by syllable between ὕβριν τότ᾽ ἢ τόθ᾽ ὅτε τὸ κύριον uóXq and the line of the
antistrophe τὰ χρυσόπαστα δ᾽ ἔδεθλα σὺν πίνωι χερῶν. But this smoothness is
1 With this we may perhaps compare another peculiarity. As Wilamowitz has shown,
Verskunst, 293 f. (cf. also Denniston in Greek Poetry and Life, Oxford 1936, 143), there is in
the lyric iambics of the Oresteia (as distinct from the Septem at any rate) no instance of
responsion of ‘syncopated’ and complete metra. As for the irregularities defended by
Verrall in Appendix ii of his commentary on the Ag., they are in the main the result of
corruptions long eliminated by the critics.
2 Perhaps I may mention that long before I had become aware of the peculiarity of the
responsion I regarded the reading ἐπικότωι σέβας with extreme scepticism on purely
syntactical grounds (and also because I remembered Cho. 157). Wilamowitz’s attempt
(Interpr. 254 n. 2) to justify it does not touch the real difficulty.
351
line’ 766 COMMENTARY
secured at too high a price since βίον cannot be given up (see below) ; besides,
it will be shown (on 775 £.) that χερῶν had much better remain outside the
trimeter extending from βίον to πίνωι. As for ὕβριν in 766, we can hardly do
without it (Heyse has deleted it), nor without the preceding βροτῶν (deleted
by Kayser) ; otherwise we might make the trimeter in 766 begin with τότ᾽
ἢ and read : τότ᾽ ἢ τόθ᾽ ἡνίκ᾽ ἂν τὸ κύριον μόληι (for ἔσθλα in the antistrophe see
below). This seems to me, like everything else that has been proposed (e.g.
ἣν for ὅταν, which does not give the required sense), so unsatisfactory that I
think the only alternative is to leave the corruption ὅταν in the text, in spite
of my suspicion that one of the considerations I have suggested may be over-
cautious and therefore wrong.
767 ff. After μόληι begins one of the deeply-rooted corruptions I have men-
tioned on 374 ff. J. Conington (Misc. Writings, i. 283 1.) gives a judicious
appreciation of the difficulties. It seems pointless to discuss the various
attempts at restoration (besides the editions cf., e.g., Emperius, Opusc. 300 f. ;
G. F. Schoemann, Opusc. iii. 175 f.; Ahrens, 554 ff.; Wecklein, Studien zu
Aesch. 117; Wilamowitz, Interpr. 196 ff.—note especially the well-grounded
scepticism of his conclusions’). I regard the following as a possible, though in
detail by no means certain, arrangement of the text: [ὅταν τὸ κύριον μόληι
φάος, κότον νεώρη, δαίμονα τίταν, ἄμαχον, continuing with the MS reading. In
this the transposition and the change of φάους into φάος have been suggested by
Emperius (Opuscula, 301). κύριον often denotes the day appointed by estab-
lished custom for the fulfilment of a business; here τὸ κύριον φάος would, of
course, primarily refer to the time appointed for childbirth (Oppian, Cyn. 3.
156 πρὶν τοκετοῖο μολεῖν ὥρην, πρὶ νκύριον ἦμαρ was quoted by Stanley ; Pindar,
Ol. 6. 32, uses κύριος peis in the same sense), but it might also contain a
reference to the day appointed for the guilty man’s punishment. νεαρὰ is
difficult. It is quite improbable that it should be deleted (as does Ahrens: ‘a
gloss on νεάζουσαν᾽, and similarly Sidgwick and Headlam). Wilamowitz has
conjectured that it conceals a form of vewpns, which I consider possible,
though I am unable to follow his assumption that in a chorus there is perhaps
no objection to the a in vedpns ;2 my grounds being that vewpns (first attested
in Sophocles) is a compound of ὄρνυμι (cf. Ritschl, Opusc. i. 401, and Wacker-
nagel, Dehnungsgesetz 49) and has nothing to do with formations ending in
-ἥρης or -dpns (Wackernagel 41). Wilamowitz reads νεάρης and thus destroys
the concinnity of the passage, whereas Pasquali, Stud. It. N.s. vii, 1929, 229 f.,
1 ‘The alterations [suggested by Wilamowitz] certainly are numerous though slight ; the
archetype must have been difficult to read, and the Byzantines must have tried to patch it
up;... here the text of our MSS is so uncertain that corruptions may have taken place
which make powerless any criticism that is more than mere trifling.’ Wilamowitz had
previously expressed himself with even greater scepticism, Griech. Tragödien, ii. 117 : "The
following couple of stanzas [763-81] have suffered such damage that even to reach suitable
thoughts you have to use violent means'.
2? Wilamowitz (in T. v. Wilamowitz's book D. dramat. Techn. des Soph. 364 n. 2) refers to
‘Hermann, to whom people will not listen’, for the conjecture in Oed. C. 702 νεαρής (sic) τις.
There is nothing of the kind to be found in (Erfurdt-) Hermann's edition of the Oed. C.
Leipzig 1841, nor have I been able to find anything in his Opuscula or the Elem. doctr. metr.
In that passage νεώρης has been suggested by Schneidewin.
3 Kaibel, it is true, in his note on S. El. gor considers it possible that the second element
in νεώρης may have been felt to be no more than a suffix, and that Sophocles could have
used it simply for véos.
352
 COMMENTARY lines 770 f.

does better in putting in the accusative. The excellent conjecture rirav for re
τὸν is due to Heimsoeth. riras, which probably is related to riveoda: (cf. on
72), occurs in Cho. 67, approximately in the sense of τιμωρός. The same sense
would be suitable here too.
However the text is reconstructed, κότον should not be tampered with as
has been done by so many editors from Hermann to Wecklein, Headlam, and
Wilamowitz (Conington in the article quoted above wisely remarks: ‘if κότον
is to be emended at all’). On this point and the accumulation of parallel
expressions put in apposition see Pasquali, loc. cit.: ‘L’ ößpıs figliola, che &
insieme 2175, é chiamata “ira novella" . . . E questo κότος annunzia già il
θράσος Aras. Eschileo é il cumulo delle apposizioni: ad Eschilo per esprimere
certi concetti un' immagine sola non basta, ed egli ne cerca sempre altre nuove'.
769. ἄμαχον ἀπόλεμον ἀνίερον : for this particular type of trikolon cf. on 412.
770 f. Presupposing that the text here is sound, the way of constructing the
words suggested by Hermann (in his appendix to Humboldt's translation)
seems possible: . . . ἀνίερον, θράσος μελαίνας μελάθροισιν Aras, εἰδομέναν
τοκεῦσιν. This has been followed by Blomfield, Klausen, and many others.
Then θράσος Aras is explained by θρασεῖαν Arav (Klausen, Hermann); ‘that
εἰδομέναν is joined on to θράσος Aras does not cause a difficulty’ (Wilamowitz,
Interpr. 196, following Hermann); cf. on 744 παρακλίνασα. But the clause
θράσος... Aras is anything but simple in itself. We must not forget that the
phrase contains far more than a simple θράσος Aras. I feel as Sidgwick does:
‘to take θράσος μελαίνας ped. ἄτας together, “the boldness of a black Curse", is
very harsh’. The harshness becomes even greater when we take into account
μελάθροισιν, which Sidgwick did not do. How can the difficulty be overcome?
Donaldson’s suggestion, adopted by Paley and commended by Verrall, to
read the dual ueAaiva . . . Ara (i.e. Képov and Θράσος) εἰδομένα τοκεῦσιν, may
be set aside because the whole train of thought here, beginning with 751,
suggests one single child of ὕβρις as well as of ὄλβος ; with this accords the
eidouevav τοκεῦσιν of the MSS. For the same reason Sidgwick’s variation of
Paley's interpretation is inadmissible: he takes Aras as plural and reads
εἰδομένας With Casaubon and refers the plural to ὕβρις and θράσος. On the
other hand, the suggestion of Auratus, μέλαιναν... ἄταν, is possibly right. If
we adopt it, the result is (assuming the correctness of the very hypothetical
rearrangement of the preceding passage as suggested above) a satisfying
articulation of the long series of accusatives that (768) follow on φάος, all in
apposition to ὕβριν : we get first two parallel elements, κότον νεώρη and δαίμονα
τίταν, then the longer element ἄμαχον ἀπόλεμον ἀνίερον θράσος, then after these
preparatory periphrases the name of the daemon of evil (cf. for the same type
of structure 154 f., to which Pasquali also refers), μέλαιναν... . Arav, and finally
the closing element eiö. rox. which emphasizes once again the genealogical
relationship by picking up the thought of 760. But this rearrangement cannot
be regarded as certain. With regard to the dative μελάθροισιν, many editors
have taken it as locative, e.g. Nagelsbach and Wecklein (so too Conington,
who, however, understands pe. μελάθ. ἄτας to mean ‘in the halls of black
Ate’; which seems to me quite impossible). It is much more natural to take
μελάθ. as a genuine dative (so e.g. Schütz, Bothe, Headlam), ‘as an Ate for
the house’, similar e.g. to 733 ἄμαχον ἄλγος οἰκέταις. To prove that μέλαθρα
can stand not only in the concrete sense of the building but also of the house
4872-2 Aa 353
lines 770 f. COMMENTARY
and its inhabitants (with reference to their destiny) in the same way as
οἶκος etc. it is sufficient to compare Ag. 1575, Cho. 1065.
771. eiööuevos is here a Homericism as in Pindar and Herodotus.
On the stanza 764—71 as a whole: Dobree (Advers. ii. 24) points out that we
find in the author of Περὶ ὕψους a passage depending on a similar sententia
(Dobree thinks on this very passage) : he speaks (44. 7) of the moral degenera-
tion which results from ἄμετρος καὶ ἀκόλαστος πλοῦτος : xpovicavra δὲ ταῦτα ἐν
τοῖς βίοις νεοττοποιεῖται κατὰ τοὺς σοφοὺς (1.6. Pl. Rep. 573e, whence the
lumen orationis νεοττοπ., in the place of the simple rixrew, is borrowed as was
seen long ago) καὶ ταχέως γενόμενα περὶ τεκνοποιΐαν ἀλαζόνειάν TE γεννῶσι Kal
τῦφον καὶ τρυφὴν οὐ νόθα ἑαυτῶν γεννήματα ἀλλὰ καὶ πάνυ γνήσια (= εἰδομέναν
τοκεῦσιν).
773. Stanley quotes Trag. fr. adesp. 500 N. Δίκας δ᾽ ἐξέλαμψε θεῖον φάος. For
similar ideas cf. A. C. Pearson on Soph. fr. 12.
774. δυσκάπνοις : ‘unpleasantly smoky’. van Heusde refers to E. El. 1139 f.,
where Electra says χώρει πένητας ἐς δόμους" φρούρει δέ μοι μή σ᾽ αἰθαλώσηι
πολύκαπνον στέγος πέπλους. Cf. also Virgil, Ecl. 7. 50 adsidua postes fuligine
nigri, Martial 2. 9o. 7 f. me focus et nigros non indignantia fumos tecta iuvant,
Ps. Quintilian, Decl. mai. 13. 4 the poor man, with whom the rich neighbour
has gone to law, pointedly calls his dwelling nutriculam casam . . . pauperem
focum et fumosa tecta. In the poor n.an's cottage kitchen and parlour are the
same, so that the smoke and soot cover the whole place. Justice shines bright
in the midst of all the filth.
775. Most recent editors (but not e.g. Wilamowitz, Ubaldi, and A. Y.
Campbell) delete βίον as Ahrens did, not because he found any objection in
the word itself, but, as he himself conscientiously says (p. 554) ‘in order to
obtain a faultless metre’. That is in itself a doubtful proceeding: in the
present case it is all the less justifiable because strophe and antistrophe in the
MS reading up to 766 ὕβριν = 776 βίον inclusive and even beyond that (τότ᾽
ἢ — τὰ xpvo-) correspond syllable for syllable, and nobody, including Ahrens,
casts any doubt on the correctness of these simple iambics. Metrical diffi-
culties begin in what follows, i.e. where it is agreed that the strophe is badly
corrupted in several places. To cure the trouble there Ahrens starts by appiy-
ing his surgeon’s knife to healthy flesh. But there are also other grounds for
making the retention of βίον probable. A syntactical break follows βίον, just
as one follows ὕβριν in the corresponding place (766) in the strophe (rightly
punctuated by many editors; the following temporal adverbs are, as an
independent kolon, tacked on to a clause that is in itself already complete).
Before the break there stand in both cases two words with the quantities v —.
Such a marking of the syntactical responsion, sometimes, as here, emphasized
by the equal length of the words, is often to be found in the choruses of
Aeschylus (cf. on 123, and p. 515 n. 1—on Eum. 163 ff.—, furthermore notice
369 — 387, 448 ff. = 467 ff., 740 = 751). It is extremely improbable that this
phenomenon should be caused in the present instance by the intrusion of a
gloss and not be attributable to the poet. A very similar symmetrical break
in strophe and antistrophe is obtained, in the arrangement of the text here
postulated, also after the first o — of the following catalectic dimeter : φάος, =
χερῶν. Both words, ὕβριν and βίον, are closely connected with what precedes, but
syntactically separated from what follows (rà χρυσόπαστα.. . . χερῶν forms an
354
 COMMENTARY line 780

“extended object’, while παλιντρόποισιν ὄμμασι adverbially modifies λιποῦσα).

Bios means here a way of living, qualified by the adjective as Eum. 525 μήτ᾽
ἄναρκτον βίον μήτε δεσποτούμενον αἰνέσηις. Above all it seems to me unmistak-
able that ἐναίσιμος βίος forms a much better complement to the δύσκαπνα
δώματα than a person denoted by τὸν ἐναίσιμον.
775. The easy change of δ᾽ to 7’ seems to be required, otherwise it would be
almost inevitable to take τὸν δ᾽ ἐναίσιμον τίει βίον as antithesis to Δίκα...
λάμπει μὲν κτλ.
ἐναίσιμον. The meaning οὗ αἴσιμον often comes near that of aequum et
bonum, ‘was (für jemanden) recht und billig ist’ (Wilamowitz, Glaube d. Hell.
i. 358). That is how Ag. 916 ἐναισίμως is used. Homer offers several instances
of the use of the word in a sense still not coloured with a moral significance
(cf., e.g., Leaf on Z 519, Ameis-Hentze on « 383). But the change to the
meaning ‘just, righteous’ naturally takes place very easily: it is obvious e.g.
in p 363 γνοίη θ᾽, of τινές εἰσιν ἐναίσιμοι of τ᾽ ἀθέμιστοι. Consequently it is com-
monly glossed by δίκαιος in the Homeric scholia and elsewhere (Suid. ἐναίσιμος"
δίκαιος, καθήκων). That Aeschylus has taken the word, which he has borrowed
from Homer, in that sense (τὸν δίκαιον gloss in Tr) is obvious from the anti-
thesis that follows in σὺν πίνωι χερῶν.
776. ἔδεθλα conjectured by Auratus has rightly been commended by
Hermann and adopted by almost all editors. Auratus must be credited with
having found the sense that is required here, even if Wilamowitz is right in
surmising that ἔσθλα should be kept (P. Maas, Gnomon, ix, 1933, 664 accepts
this). Perhaps we should expect ἔθλον (for the ψέλωσις see W. Schulze, Kl.
Schr. 391) rather than ἔσθλον (see Schwyzer, Gloita, vi, 1915, 86 n. 1), but from
Schwyzer’s remark, Griech. Gramm. i. 839, it seems to follow that ἔσθλον, ἴοο,
could be accounted for linguistically. What I have said above (on p. 351)
about the correspondence of — to vv in these iambics is a strong argument
in favour of a disyllabic form, unless we suppose a deeper corruption here.
σὺν πίνωι χερῶν: ‘when there the hands are dirty’; cf. on 456.
777. παλιντρόποισιν : the ending has been restored by Pauw and others; for
the syllaba anceps at the end cf. on 229. For the expression cf. Suppl. 173
παλίντροπον ὄψιν, the adjective was presumably formed on the model of N 3
πάλιν τρέπεν ὄσσε φαεινώ (cf. Leaf ad loc., Kumaniecki 12).
778. It is impossible to attain complete certainty, but Hermann’s προσέμολε
is the most likely correction. As is well known, confusion between u and f is
very common in minuscule, and then the further corruption is easily ex-
plained. With the thought rà . . . σὺν πίνωι χερῶν... λιποῦσ᾽ ὅσια προσέμολε
Paley compares Cho. 308 ἧι τὸ δίκαιον μεταβαίνει. For the conception of the
ὅσιον cf. R. Hirzel, Themis, 43 n. 2 and 160 n. 1; Wilamowitz, Platon, i. 61 f.
(also K. Latte, Heiliges Recht, 55 n. 16) ; Wilamowitz, Berl. Sitzgsber. 1927, 159,
and especially his Pindaros, 317 for ὅσια δρῶν: 'refraining from polluting
actions, nothing positive’; that exactly fits the present passage, where the
antithesis to the soiled hands has immediately preceded, cf. also Cho. 378
χέρες οὐχ ὅσιαι. ὅσια stands in sharp antithesis to ἄδικα in Suppl. 404. We
must remember the strength of this opposition if we wish to appreciate fully
the magnificent oxymoron S. Ant. 74 ὅσια πανουργήσασα.
780. παράσημον. To Blomfield's collection of the evidence we must now add
Bywater's copious list of the occurrences of this word and related expressions,
353
line 780 COMMENTARY
C.R. liv, 1940, 10 ff. παράσημον when used of coins means κίβδηλον, the silver
being debased by an admixture of brass or lead. The picture of counterfeit
money is still fully alive in its association with δύναμιν πλούτου. Heath
recognized that αἶνος here has the same meaning as ἔπαινος (so already in
Homer, then in Pindar); cf. on 1482.
781. πᾶν δ᾽ ἐπὶ τέρμα νωμᾶι. Hermann objects to the traditional interpreta-
tion ‘omnia dirigit ad finem’, saying ‘dici hoc sensu debuit πάντα '. This is
prejudiced. πᾶν, ‘everything’ or ‘each thing (of those to which she turns, with
which she is concerned) she brings to fulfilment’, is an unexceptionable
expression. Cf., e.g., S. Oed. C. 465 ws viv πᾶν τελοῦντι προξένει, Soph. fr. 60 N.
(= 63 P.) ἐν δεσμοῖσι δραπέτης ἀνὴρ κῶλον ποδισθεὶς πᾶν πρὸς ἡδονὴν λέγει,
E. Andr. 448 1. ἑλικτὰ κοὐδὲν ὑγιές, ἀλλὰ πᾶν πέριξ φρονοῦντες. Triclinius mis-
understood the application of πᾶν: πᾶν γὰρ τοιοῦτον, ἤγουν πάντα τοιοῦτον
πλοῦτον, μὴ Er δικαίων συνηγμένον, ἐπὶ τέλος ἄγει ἄξιον αὐτῶι. More probably
πᾶν refers to each of the various activities of Aikn as they are described from
772 on. The rule of Δίκη as τελεσφόρος (cf. p. 449) is alluded to in ἐπὲ τέρμα
νωμᾶι, as R. Hirzel remarks, Themis 136 n. 4. With the expression and also
the thought as a whole cf. E. Suppl. 614 ff. δίκα δίκαν ἐκάλεσε καὶ φόνος φόνον,
κακῶν δ᾽ ἀναψυχὰς θεοὶ βροτοῖς νέμουσι, πάντων τέρμ᾽ ἔχοντες αὐτοί.
On 776 ff. it should be noted that not only the general point of view but also
the particular imagery finds a close analogy in Sept. 767 ff.

The four great choric odes of the Agamemnon, viz. the parodos and the
three stasima, all take us back to events that occurred ten years ago, about
the time of the outbreak of the war. The manner in which the past is recalled
and the extent to which this is done vary considerably. The long chant
κύριός εἶμι θροεῖν consists almost entirely of an account of what happened at
Aulis ; only the hymn to Zeus is detached from, though not unconnected with,
the narrative. In the first stasimon (367 ff.) the middle part (399-426) is
occupied by scenes of the past, but on the whole general meditations, and
ideas concerned with the present situation, prevail. The second stasimon
(681 ff.) is more neatly divided: the first five stanzas dwell on the departure
of Helen, on the arrival of the guilty couple, and on the consequences of their
crime, while the last three are devoted to the general problems involved in
those happenings. The third stasimon (975 ff.) shows the Chorus completely
absorbed in reflections on divine retribution and in dread of some imminent
catastrophe. Here too, however, there is at least one secondary clause
(984 ff.) which brings before us by a picturesque detail the Greek fleet setting
forth for Troy.
These shorter or longer retrospective passages are clearly no mere orna-
ments. The more carefully we study the great choric songs of the Agamemnon,
the more we marvel not only at the consistency of each of them but also at
the cumulative effect of unity and concord produced by the series in its
entirety. In this perfect structure there is no room for any detail that would
not at the same time support the plan of the whole. The fundamental idea
is simple: man’s blind folly leads him to crime, crime that will inevitably be
punished by divine justice. The poet did not, however, want to write a
sermon. Nor do the singers of his Chorus act as preachers; they belong as
living members to the dramatic organism. All their thoughts seem to arise
356
 THE FOUR GREAT CHORIC ODES
spontaneously from the tragic plot and its background. Any deed of dry in
this play, any misery of Agamemnon, his house, and his people, as well as the
disaster of the Trojans, can and must be traced back to its origin. That origin
lies in the events that immediately preceded or accompanied the beginning of
the war against Troy. Unless in every single case the παρακοπὰ πρωτοπήμων
is recalled and its dreadful implications are made clear, the process of relent-
less retribution, the work of the god’s supreme will, cannot be sufficiently
understood.
This, then, is the primary, or at any rate a primary, motive for the exten-
sive use of narrative elements in the odes. But here again it is well not to
forget that we are dealing with poetry. As a poet, Aeschylus availed himself
of the opportunity to widen the prospect of his play both in space and time,
and to include in it other sections of the old story besides those that were
required by the demands of the dramatic action. No Greek poet, least of all
Aeschylus, was in a position to think in a detached and abstract way of such
events as the departure of Helen, her and her lover’s voyage across the sea,
the reception given by the citizens of Troy to the adulterous couple. Famous
episodes like these contained too many seeds of indestructible life, too power-
ful potentialities of irresistible charm, to be treated merely as ultimate causes
of events resulting from them. Once the magic wand of the poet touched
them, they were bound to spring again into full life. That is what has hap-
pened in these choric songs. Behind the characters and conflicts of the actual
play there opens another plane, on which well-known figures appear, Paris
and Helen and Menelaus, the Priamids and the seers of the royal house at
Argos. As they emerge one by one from the various passages of the odes, we
obtain only glimpses; but when we look at the whole, the separate pieces
coalesce into a unity. The arrival of Paris in the house of the Atridae and
Helen's light-hearted desertion of her home and husband (399-426) are fol-
lowed by the picture of her swift passage to Ilion, Ζεφύρου γίγαντος αὖραι, and
the rejoicings of the Trojans at the fateful wedding (690-708). In the first ode
we hear nothing of what happened to the fleet in Aulis after the drAosa had
ceased, but in the third stasimon the eyewitnesses mention the moment when
the many ships at last left the shore (984 ff.).
Among those figures on the second plane there is one more accomplished,
more lifelike, and more mysterious than any other, Helen. She is to grow into
a νυμφόκλαυτος 'Epwós and to be cursed as zapávovs ‘EAeva. And yet her
irresistible power and the gentle sweetness of her very sin, which seems as
natural and innocent as a happy child's playing, have never been more
movingly praised. When Aeschylus designed the magnificently wide frame-
work of these choric odes, he had in mind the decisive issues of the tragedy.
But he also seized eagerly upon the chance of creating a picture worthy of
Helen and of himself.

783. A recently discovered fragment of Aeschylus (to be published in Pap.

Oxy.) begins ἄ]γε δή, βασιλεῦ.
πτολίπορθ᾽ : Blomfield's remark that in this Homeric word πτολ- is the only
correct form is confirmed by I. Bekker, Hom. Blätter, ii. 19. The irregular
accent of πτολίπορθος is explained by Wackernagel, Nachr. Gött. Ges. 1914, 130,
as an Aeolism.
357
 line 785 COMMENTARY
785. Wilamowitz and the editors who depend on his critical apparatus (A. Y.
Campbell, Murray, G. Thomson) do not mention that in F we have σεβίξω
(£ clearly altered into £, see on 681). The reading is recorded in Vitelli-
Wecklein, and had been fully discussed by Hermann: 'Doricam formam
σεβίξω Flor. habet, nescio an recte. Usitatior certe est in deliberatione de eo
quod agendum sit aoristi coniunctivus. Quamquam aliquando etiam prae-
senti utuntur, τί δρῶ; τί δρῶμεν; τί λέγω; et similia dicentes. Hermann’s
grammatical note is relevant. A closer inspection of the instances shows that
Aeschylus employs the question expressed by πῶς with the subjunctive with
a striking preference in the Oresteia: Ag. 212 πῶς λιπόναυς γένωμαι; 543 (cf. my
. note there) mas... τοῦδε δεσπόσω λόγου; 648 πῶς . . . συμμείξω λέγων; 785 πῶς
σε προσείπω; 1109 πῶς φράσω τέλος; 1490 = 1514 πῶς σε δακρύσω; Cho. 88
mds... εἴπω, πῶς κατεύξωμαι; 171 πῶς... μάθω; 187 müs... ἐλπίσω; 844
müs... δοξάσω; (Heath, Paley, and Wilamowitz are wrong in putting a
question-mark or a comma after ταῦτ᾽) 858 más . . . ἀνύσωμαι; 925 πῶς
φύγω; In other words there are, not counting the possible σεβίξω here,
thirteen instances, all with the aorist. There is no other instance in the
Oresteia of πῶς with the present subjunctive. In the diaskeuast Sept. 1058
we find πῶς τολμήσω. For the present there is only one example in Aeschylus,
Suppl. 505 ἐγὼ δὲ πῶς δρῶ; (there may, perhaps, be added Suppl. 339 [336
Wil] mas . . . εὐσεβὴς ἐγὼ πέλω; Eum. 678 màs . . . ἄμομφος ὦ; but the
particular nature of these copulative verbs must be taken into account).
This form of the deliberative question in the subjunctive is found much less
often in Sophocles. Of the five instances four show the aorist: Ant. 377, Aj.
1024, Phil. 1350, fr. 197 N. (= 196 P.), one the present: Aj. 214 πῶς δῆτα λέγω
λόγον ἄρρητον; This conspectus is decidedly in favour of σεβίξω. Besides, it
is supported by the consideration that in such a sharp anaphora a change of
tense seems very surprising, though not totally impossible, whereas πῶς ce
προσείπω; πῶς σε σεβίξω; gives a perfect parallelism. Cf. also the above-
mentioned passage Cho. 88 πῶς εὔφρον᾽ εἴπω, πῶς κατεύξωμαι πατρί; So far as
concerns the £ in σεβίξω, which is not confined to the Doric (cf. Schwyzer,
Griech. Gramm. i. 737 f.), A. Suppl. 39 σφετεριξάμενοι (similarly anapaests)
would suffice to defend it. We may add probably Cho. 954 f., where in the
meaningless ἐπ᾽ ὄχθει d£ev it may be supposed that ἐπωρθίαξεν (Meineke)
lurks. Ina fragment from Oxyrhynchos (probably Aeschylus) which is still
unpublished we find eppvÜja£a. All these considerations justify us in putting
σεβίξω in the text. Of course in itself the alteration of £ into £ might be simply
a corruption. Thus the Mediceus wrongly gives στενάξω. τί ῥέξω Eum. 788,
whereas the repetition of these words 819 appears as στενάζω. τί ῥέξω. In the
same way in F Ag. 443 we get wrongly γεμίξων (ζ altered into £), Ag. 681
probably wrongly (v. ad loc.) ὠνόμαξεν.
Finally a word on the relation of the participles ὑπεράρας and ὑποκάμψας to
πῶς σε σεβίξω (and possibly also to πῶς σε προσείπω, since it is conceivable
that the participles should be taken with both questions). Modern feeling
would be inclined to find a simple modality expressed in such participles. But
that is not to give its full meaning to the aorist. The Greek rather seems to
regard σεβίξαι as the result of the activities expressed in the participles. The
same is true e.g. of Pers. 333 ἀτὰρ φράσον μοι τοῦτ᾽ ἀναστρέψας πάλιν, 700 ff.
δίομαι μὲν χαρίσασθαι, δίομαι δ᾽ ἀντία φάσθαι, λέξας δύσλεκτα φίλοισιν (here too
358
 COMMENTARY line 786
we can scarcely regard λέξας as meaning the same as λέγων, much more
probably should ἀντία φάσθαι appear as the result of λέξαι δύσλεκτα), Eum.
921 f. dir’ ἐγὼ κατεύχομαι θεσπίσασα πρευμενῶς κτλ. The temporal relation is
quite clear Ag. 1240 f. καὶ σύ μ᾽ ἐν τάχει παρὼν ἄγαν ἀληθόμαντιν οἰκτίρας
ἐρεῖς : the οἶκτος is the impulse which causes the Elders to call Cassandra
aAndöuavrıs." For the function of the aorist participle cf. also on 395 and 789.
With 783 ff. (first the solemn vocatives, then πῶς σε προσείπω κτλ.) we may
compare the words with which Trygaios salutes the goddess Eirene ori her
appearance (Ar. Peace 520 ff.) ὦ πότνια βοτρυόδωρε, τί προσείπω σ᾽ ἔπος ; πόθεν
ἂν λάβοιμι ῥῆμα μυριάμφορον ὅτωι προσείπω σε; Cf. h. Hom. Ap. 19 πῶς rap
o ὑμνήσω πάντως εὔυμνον ἐόντα; Callim. Iov. 4 πῶς καί μιν. . . deicopev . . .;
It is not unlikely that the introductory clauses of these anapaests were sug-
gested to Aeschylus by formulae used in invoking gods.
786. ὑπεράρας : the verb seems not to be attested elsewhere before Plato and
Xenophon, and consequently it is scarcely possible to say with certainty
whether it still conveyed an image or has been weakened already to a simple
‘metaphor’ to express ‘to go beyond’. It is certainly used in the latter way in
Pl. Laws 717d, where we find a ‘polar’ opposition which exactly corresponds
with the present passage: (referring to the burial of the parents) μήτε ὑπερ-
aipovra τῶν εἰθισμένων ὄγκων μήτ᾽ ἐλλείποντα ὧν of mporäropes . . . ἐτίθεσαν.
Here and sometimes elsewhere drepaipew is completely identical in meaning
with ὑπερβάλλειν, which is much earlier attested and much more widely used,
cf. e.g. Hesiod, Erga 489 (the amount of rain) μήτ᾽ dp’ ὑπερβάλλων Bods ὁπλὴν
μήτ᾽ ἀπολείπων. The participle of ὑπερβάλλειν without an object is found very
early where there seems to be no imagery: Pind. N. 7. 66 ἔν re δαμόταις ὄμματι
δέρκομαι λαμπρόν, οὐχ ὑπερβαλών, and possibly the same is true here of
ὑπεράρας. In the passages of Hesiod and Plato quoted above ἐλλείπειν or
ἀπολείπειν forms the antithesis of ὑπεραίρειν or ὑπερβάλλειν. In the present
passage ὑποκάμπτειν signifies precisely the same, and Blomfield has appro-
priately translated it ‘to fall short of’. He has, however, remembering
another use of κάμπτειν, joined with this the idea of rounding the mefa on the
race-track, and thus misled many of his successors into over-refinements.
For supposing that a picture is to be found in ὑπεραίρειν at all (see above), it
would be the one drawn from shooting, while the idea of the rounding of the
mela would involve a mixing of metaphors (so e.g. Paley). But this is very
improbable, since μήτε... μήτε marks a strong parallelism in thought, and
furthermore ὑπερ- (‘beyond’, ‘over’) and ὑπο- (‘short’, *under') answer one
another so exactly that we are almost compelled to take both expressions on
one and the same level. It does not follow because [Xenophon] Cyn. 5. 16
ὑποκάμπτειν is used of the 'doubling' of the hare, that the word elsewhere
must denote a swerve in a sideways direction, and that it does not rather
denote a swerve in general from the direction that leads to the goal. Conse-
quently, supposing that imagery is still to be found in these expressions, in
that case ὑπεραίρειν means shooting beyond the mark, while ὑποκάμπτειν
means what Pindar expresses as follows (Ol. 9. 12) : οὔτοι χαμαιπετέων λόγων
ἐφάψεαι (Schol. χαμαιπετῆ δὲ κυρίως ἐστὶ δόρατα ἢ βέλη, ὅταν ἀποτυχόντα τοῦ
1 We may regard as similar e.g. the aorist in Hdt. 7. 141. 3 σοὶ δὲ τόδ᾽ αὖτις ἔπος ἐρέω,
ἀδάμαντι πελάσσας : the god sets about the process of ἀδάμαντι πελάζειν before he utters his
oracle.
359
line 786 COMMENTARY
σώματος χαμαὶ meonı). But it is just as possible that the verbal constituents of
ὑπεράρας and ὑποκάμψας have lost their full meaning, while the emphasis laid
on the prepositional elements with their contrasting of nimis and darum
remains unweakened. Paley well compares with this passage in which excess
and defect of expressions of respect are so strongly emphasized Cho. 857 f.
ὑπὸ δ᾽ εὐνοίας πῶς ἴσον εἰποῦσ᾽ ἀνύσωμαι;
787. καιρὸν χάριτος. For the meaning of καιρός cf. on 365. χάρις means here,
not as one might think the thankfulness of a recipient for a favour, but (cf.
on 421) the pleasure of the person to whom προσειπεῖν and σεβίξαι are directed,
or the pleasingness of the act itself, i.e. of the utterance. Cf. on this point
and on 785 f. as a whole the passage which has always been quoted by the
editors E. 15h. A. 977 f. πῶς ἄν σ᾽ ἐπαινέσαιμι μὴ λίαν λόγοις, μηδ᾽ ἐνδεὴς τοῦδ᾽
ἀπολέσαιμι τὴν χάριν; Here we have obviously a direct reminiscence of the
Agamemnon, as elsewhere in 15h. A., cf. on 218 and 226.
788. τὸ δοκεῖν εἶναι: that holds true quite generally, whatever quality may
be understood with εἶναι, true friends, loyal citizens, etc., cf. in the answer
840. It is a grim echo when later (1374 f.) Clytemnestra, with reference to her
revenge on Ágamemnon, speaks of the person who acts as ἐχθροῖς ἐχθρὰ
πορσύνων, φίλοις δοκοῦσιν εἶναι. Sept. 592 has been compared by many editors:
οὐ yap δοκεῖν ἄριστος ἀλλ᾽ εἶναι θέλει, and also Philemon fr. 94. 8 K. (ὅστις)
εἶναι δίκαιος κοὐ δοκεῖν εἶναι θέλει. |

789, δίκην παραβάντες. It is tempting to see in this expression a variation

of the preceding ὑπεράρας (καιρόν), similar to what we find e.g. Prom. 30
βροτοῖσι τιμὰς ὥπασας πέρα δίκης. Several editors have succumbed to the
temptation and have seen in δίκην παραβάντες a modality of προτίουσι. So e.g.
Schneidewin : ‘manche ziehen dem Sein den Schein vor und überschreiten das
Rechte (τὸν καιρόν) ; Lewis Campbell: ‘who prefer appearance and overpass
the right’; Platt; Mazon: ‘Tant de mortels, soucieux surtout d’apparences,
dépassent la juste mesure’. The same grammatical construction is assumed,
with a slightly different nuance of the sense, by Verrall: ‘Many rate semblance
above reality, and do injustice so’; and by Headlam: “There are many that
dishonestly affect mere outward show.’ But this way of taking παραβάντες
does not do justice to the subordination of the aorist participle to προτίουσι ;!
cf. on 395. The grammatically correct way of rendering the passage is shown
by translations such as these: “when they have passed the bounds of justice’
(Paley) ; ‘so viele verlassen des Rechtes Bahn und ziehen den Schein vor’
(Nägelsbach) ; ‘after violating justice’ (Kennedy) ; ‘molti violando il giusto
amano meglio parere che essere buoni’ (Ubaldi). Of course, the sentence
πολλοὶ δὲ κτλ. is primarily uttered as a general maxim. But it is certain that,
besides this, there is implied in it a hint of Clytemnestra and warning against
her, as Schiitz has seen.
790. That δ᾽ as read in the MSS and in Stobaeus must be altered to 7’ has been
recognized by Hermann ; that is the only way it can be brought out that the
general sentiment 788 f. πολλοὶ... παραβάντες finds its explanation in what
follows (790-4), i.e. the description of the two contrasted attitudes towards
! The interpretation rejected above (apaß. a qualification of the act of mporiew) would
fit in with the reading of Triclinius παραβαίνοντες. But it is unlikely that Triclinius was
concerning himself here about any change in the sense: his purpose was rather (cf. on
1334) to expand the paroemiac into a complete dimeter, as Hermann recognized.
360
 COMMENTARY line 795

another’s fortune or misfortune. This elaboration, internally bound together

by re... καί, is attached, as so often is the case with explanatory phrases,
asyndetically to the main clause which requires the explanation. For this
arrangement Hermann has appropriately compared 834-7.
τῶι δυσπραγοῦντι ἐπιστενάχειν is taken by Linwood, Lex. Aesch., and many
translators, e.g. Paley, Verrall, Headlam = ‘to mourn (or ‘sigh’) over’.
Persae 727 καὶ τί δὴ πράξασιν αὐτοῖς ὧδ᾽ ἐπιστενάζετε; may serve to support
this, cf. also E. Med. 929 τοῖσδ᾽ ἐπιστένεις τέκνοις. But, on the other hand,
when we compare À 153 f. τοῖς δὲ βαρὺ στενάχων μετέφη . . . ἐπεστενάχοντο δ᾽
ἑταῖροι, T 301 ds ἔφατο κλαίουσ᾽, ἐπὶ δὲ στενάχοντο γυναῖκες, we see that it is also
possible to take ἐπιστεν. in a corresponding way, i.e. ‘in addition to, together
with’. So the gloss (in Tr) συλλυπεῖσθαι, and among modern scholars e.g.
Schütz (‘infelicibus adgemere’), Nägelsbach (‘mit den Unglücklichen zu
seufzen’), Platt (‘to utter the voice of mourning with him that is in trouble’).
This is the only meaning that can be intended here, as is obvious from the
parallelism with 793 Evyyaipovow ὁμοιοπρεπεῖς.
791. For the change of the accent from early Attic ἑτοῖμος to late Attic
ἕτοιμος Cf. Schwyzer, Griech. Gramm. i. 383.
791 f. δῆγμα... προσικνεῖται is a true parenthesis.
792. οὐδέν adverbial.
794. Hermann, whom most critics have followed, has taken exception to the
structure of the sentence as well as to the hiatus after βιαζόμενοι which is
unexampled in similar anapaests in Aeschylus,’ and consequently decided
that a lacuna must be assumed here. Probably the main verb has disappeared
in the lacuna, and ξυγχαίρουσιν should be taken as dative. Ata pinch, it would
be possible to take évyyaip. as indicative, and to supply from the context the
dative which is required by ópowmpezeis, but that is far less likely. It is
possible that the interpolation which underlies the unmetrical line as quoted
by Stobaeus was made by an ancient reader who took £vyyaipovow as indi-
cative and consequently looked for a dative to go with ὁμοιοπρεπεῖς. This is
the view held by Enger (app. crit. in his re-edition of Klausen): 'Scriptura
apud Stobaeum nata est ex vulgata et librarii correctione, qui quum ad
ὁμοιοπρεπεῖς dativum requireret, νυκτὶ 86 . . . adscripsit'. The idea may have
been suggested by the Homeric vuxri ἐοικώς (Hermann).
With 793 f. compare O tox ff. ἡ δὲ γέλασσε χείλεσιν, οὐδὲ μέτωπον ἐπ᾽ ὀφρύσι
κυανέηισιν ἰάνθη, and the skolion of the fifth century (cf. Wilamowitz, Hermes,
Ix, 1925, 300 f.) preserved by Lobon in Diog. Laert. 1. 61: πεφυλαγμένος ἄνδρα
ἕκαστον ὅρα μή (σε) κρυπτὸν ἔγχος ἔχων κραδίηι φαιδρῶι προσεννέπηι προσώπωι κτλ.
795. προβατογνώμων does not occur elsewhere and was possibly coined for
this passage. With respect to the formation Bothe compares Aesch. fr. 243. 3
θυμὸν ἑππογνώμονα, which is possessed by the man who has learnt from his
experience with mares to read the sexual desire expressed in the eye of the
female. The sense of προβατογν. is completely clear ; it is surprising to read
not only in Paley’s translation but also in L-S ‘metaph., good judge of
character’. It has long been recognized that 'knower' (or ‘judge’) of the flock’,?

! Hermann, it is true, says that the hiatus might perhaps be defended 'propter aliquam
pausam in dicendo factam’, cf. also his Elementa doctrinae meiricae 373.
2 “πρόβατα in early Greek is not confined to sheep, as was often observed by the ancient
grammarians' (Wilamowitz on Hesiod, Erga 558), cf., e.g., Schol. ABT on Hom. 8 124.
361
line 795 COMMENTARY
as said of the king, is connected with the Homeric ποιμὴν λαῶν. The 'very
bold’ (Debrunner, Griech. Wortbildungslehre, 43) new formation ποιμάνωρ A.
Pers. 241 (also 74 ποιμανόρων)
had better, possibly, here be left on one side,
because there the poet may have intentionally stylized the unlimited author-
ity of the rulers among the Asiatics in an unhellenic manner (cf. W. Kranz,
Stasimon, 88). προβατογνώμων here, just like its Homeric forerunner, has a
claim to be considered as a step on the way to that ideology of kingship
which sees the true king as the good shepherd. In germ it is to be found in
Plato (Politikos 265 ἃ, 268a, etc.) and in Xenophon (Cyrup. 1. 1. 2, etc.), and
later develops in the way described by E. R. Goodenough, ‘Hellenistic King-
ship’, Yale Class. Studtes i (1928), 84.
For the principle laid down in 795 ff. that the genuine ruler must have the
faculty of seeing through flattery that arises from impure motives, even, as
in the present instance, at the moment of the proudest triumph, a good
example is offered by the behaviour of Pausanias to Lampon of Aegina after
the battle of Plataea, as described by Herodotus 9. 78 f.
796. ὄμματα φωτός. We should not connect φωτός with ὅστις... mpoßaro-
γνώμων. My warning against this way of taking the passage is due to the fact
that Hellenists like Kennedy, Wilamowitz (‘Aber des erfahrenen Mannes
Auge unterscheidet, wo der Schein der Treue . . . schmeichelt’), Murray (‘But
the wise shepherd knoweth his sheep, and his eyes pierce deep the faith . . .’)
hold this view. The greater number of interpreters, from Triclinius on, have
rightly looked on ὄμματα φωτός as subject of λαθεῖν. Unmistakably ὄμματα picks
up the preceding πρόσωπα. It is a very good trait that the caivew comes from
the eyes; cf. S. Oed. C. 319 f. φαιδρὰ γοῦν ἀπ᾽ ὀμμάτων σαίνει με προσστείχουσα.
797 f. Casaubon’s σαίνει has in the easiest way put the sentence right. τά is
relative as Suppl. 265 ; further examples of the relative use of the demonstra-
tive pronoun which we call the article are to be found in Dindorf, Lex. Aesch.
234 under 5. Wecklein says, rightly in general: ‘to δοκοῦντα ἐξ edppovos
διανοίας an idea like ξυγχαίρειν is supplied’, only I would say ‘an idea like
βλέπειν᾽. The passage is akin to those discussed by Kühner-Gerth, ii. 566 f. :
‘in antitheses often something must be supplied out of one element to supple-
ment the other’, and: ‘from a following verb of a specialized meaning a verb
of a generalized meaning must be understood’. Cf. also the commentaries on
S. El. 435 and Rothstein on Prop. 2. 31. 13. Especially this latter form of a
difficult zeugma is to be found in connexion with an antithesis (as here
εὔφρονος ἐκ διανοίας : ὑδαρεῖ φιλότητι) in Aeschylus: Suppl. 1006 f. ὧν πολὺς
πόνος, πολὺς δὲ πόντος οὕνεκ᾽ ἠρόθη δορί, Prom. 21 f. iv’ οὔτε φωνὴν οὔτε rov
μορφὴν βροτῶν ὄψηι. Cf. on 1148.
διάνοια here as in several other passages of Aeschylus indicates the feeling
between two persons (cf. Verrall on Eum. 985 ff.).
798. ὑδαρεῖ: Schol. μεμιγμένηι καὶ οὐ καθαρᾶι καὶ ἀκράτωι. Blomfield has
collected the parallels, among them Arist. Pol. 2. 4, 1262 Ὁ 15 ἐν δὲ τῆι πόλει
τὴν φιλίαν ἀναγκαῖον ὑδαρῆ γίγνεσθαι διὰ τὴν κοινωνίαν τὴν τοιαύτην. Obviously
ὑδαρές is something watered down, adulterated, in contrast to what is ἄκρατον.
800. Ἑλένης ἕνεκα: À 438 (of which we are again reminded in Ag. 1455 ff.)
Ἑλένης μὲν ἀπωλόμεθ᾽ εἵνεκα πολλοί, Hesiod, Erga 165 (the war took place)
‘EXévns Ever’ ἠυκόμοιο.
Many editors follow Hermann in deleting γάρ and reading oö(«). There is
362
 COMMENTARY lines 803 f.
no doubt that the particle is sometimes wrongly inserted, e.g. Pers. 550, Eum.
365, 378, but here there is little probability of such a supposition. The Homeric
passages in which οὐκ (or οὐδ᾽) ἐπικεύσω occurs are on the side of syntax not
suitable parallels, since in Homer it is always a case of self-sufficient clauses
standing by themselves (similarly in Apoll. Rhod. 3. 332 où σ᾽ ἐπικεύσω). Here
on the contrary we have a genuine parenthesis: the words où . . . ἐπικεύσω
cut right into the middle of the sentence où δέ pos . . . ἦσθα γεγραμμένος κτλ.
In the case of such an intrusion cutting a sentence in half ydp is particularly
in place, e.g. Pers. 341 ξέρξηι dé, καὶ γὰρ οἶδα, χιλιὰς μὲν ἦν krÀ., Ag. 838, 1069,
5. Trach. 479 f. καὶ ταῦτα, δεῖ γὰρ καὶ τὸ πρὸς τούτου λέγειν, οὔτ᾽ εἶπε κρύπτειν
KTÀ., Phil. 603 f. ἐγώ σε τοῦτ᾽, ἴσως γὰρ οὐκ ἀκήκοας, πᾶν ἐκδιδάξω, 938, Ar. Lys.
1137 εἶτ᾽ ὦ Λάκωνες, πρὸς γὰρ ὑμᾶς τρέψομαι, οὐκ ἴστε κτλ. In particular we
should compare passages in which, as here, the incidental remark excuses
and at the same time slightly postpones something unpleasant which the
speaker means to say. So e.g. Pl. Apol. 22a καὶ vn τὸν κύνα, ὦ ἄνδρες
᾿Αθηναῖοι, δεῖ γὰρ πρὸς ὑμᾶς τἀληθῆ λέγειν, ἦ μὴν ἐγὼ ἔπαθόν τι τοιοῦτον, there
follows a very biting remark about the μάλιστα εὐδοκιμοῦντες, Isocr. 12. 100
δύ᾽ ἢ τρεῖς τῶν στρατηγῶν τῶν ἡμετέρων, οὐ γὰρ ἀποκρύψομαι τἀληθές, ἐξήμαρτον
(G. Thomson quotes this, but nevertheless deletes γὰρ with Hermann).
Accordingly I prefer to adopt Musgrave’s supplement, especially as c could
so easily drop out before ε.
801. κάρτ᾽ ἀπομούσως is connected with οἴακα νέμων by Elmsley (who
however prefers ἀπόμουσός 7’ . . . οὔτε), Ahrens (p. 558), and e.g. by Schneide-
win-Hense and Wecklein : ‘as one who perversely and not for safety directed
the helm of the mind’ (Ahrens). The majority of editors connect ἀπομ. with
γεγραμμένος. This is obviously right, for not only is the word-order in favour
of it, and also the feeling that a kolon complete in itself is to be expected
before οὐδέ (what follows is added as an explanation), but it is also sup-
ported by the fine appropriateness of the resulting expression. The category
of μουσικόν or ἄμουσον has no place in the province of κυβερνητικὴ τέχνη, but
has a place in γράφειν, drawing and painting. Some editors and above all
W. Kranz, Stasimon 74, have compared the passage with other expressions
in Aeschylus that have reference to painting. ἀπόμουσος ó ἄμουσος: SO
Eustathius on Ihad © 518 (= Cramer, Anecd. Paris. iii. 232 f.) in a digression
about the privative use of ἀπο-, which like the whole treatise on the names of
the various stages of life, in which it occurs, goes back to Aristophanes of
Byzantium (pp. 89 f., 95 ff. Nauck), cf. above on 143. ‘In Greek ἀπό and ἐξ
are interchangeable with a-. So Sophocles describes the Pontos Euxeinos
(originally A£ewos [see on this H. Jacobsohn, Kuhns Zeitschr. liv. 254 ff.]) as
ὅρμος ἀπόξενος Oed. R. 196, and Hesychius glosses ἀπόδειπνος with ἄδειπνος,
ἀπόθεα (Soph. fr. 245 N. = 267 P.) with ἄθεα᾽ (Wackernagel, Syntax, ii. 296).
So the expression ‘you were then to me painted wholly without art (or ‘skill’)’,
means in the first place a picture that was made invita Minerva, but that must
be an ugly picture, and that is the point here.
802. εὖ πραπίδων οἴακα νέμων : cf. 380.
νέμων: cf. on 685 and Sept. 3 οἴακα νωμῶν.
803 f. one of the awkward cruces of the play. Hermann has established a
secure starting-point by defining the use of κομέζειν : ‘aliud enim κομίζειν est,
quod significat apportare vel adducere ad aliquem ita ut quod affertur vel
363
lines 803 f. COMMENTARY
adducitur prope illum statuatur, aliud φέρειν, quod latius patet atque sic
dicitur ut res allata etiam inseri ei ad quem affertur significari possit’. This
has been ignored by Ahrens among others ; Wilamowitz, who follows Ahrens,
translates ‘Mut dem Heer zu machen’, which, as Hermann has shown, is
precisely what θράσος... . κομίζων cannot mean. In the Thesaurus not even a
remote parallel is to be found to a use of κομίζειν such as is here assumed,
whereas we can certainly say in Latin animum adferre (Cic. Phil. 8. 22 animum
nobis adferre legati debuerunt : timorem altulerunt). In L-S s.v. κομίζω under
ii. ro. ‘bring, give’ Ag. 804 stands all but isolated. If we stick to the
attested meanings of κομίζειν, we can leave out of account, besides other
objections, a whole series of attempts to explain the MS reading, e.g. that of
Bamberger, Opusc. 248 f.: ‘du erschienest uns wie einer, der Sterbenden Mut
einsprach, den sie freudig annahmen', and a variation on this in Wecklein,
Stud. z. Aesch. 120, where he reads ἐτώσιον and translates : ‘du kamst mir vor
wie jemand, welcher Sterbenden eitele, nichtige Tróstungen zubringt'. Weck-
lein indeed quotes Hermann's note on κομίζειν with approval, but his
own rendering runs counter to it, unless he was thinking of the Sacrament
administered to the dying, for which 'Tróstungen' is the ordinary term. The
establishment of the use of κομίζειν also upsets Ahrens's elegant conjecture ἐκ
θυσιῶν. In itself the thought implied in this reading is consonant with Greek
ideas: 'the ἑερά gave the warriors courage and hope, and offered the soldiers
the right trust in the help of Heaven’ etc. (Eitrem, Symb. Osloens. xviii, 1938,
14, cf. also p. 29 f. for the point that offerings of this kind could be called
θυσίαι just as well as ἱερά). With ἐκ θυσιῶν the Chorus would be first, appar-
ently innocently, hinting at the καλλιερεῖν usual at the start of an expedition
as it was before battle: of course, neither Agamemnon nor the audience could
fail to understand the ominous underlying (or rather principal) meaning, i.e.
the reference to the sacrifice of Iphigeneia. I should find no difficulty in the
fact that this horror is touched on so lightly, considering the conciliatory and
respectful attitude which the Chorus is now taking up towards the king. But
all these considerations are of no avail so long as θράσος κομίζειν is not eluci-
dated in agreement with linguistic usage.
An unassailable interpretation of κομίζειν is the foundation of the rendering
which Weil, and, following him, Margoliouth, Verrall, and Headlam (also
Murray) have adopted: 'feminae audaciam voluntariam, h.e. feminam per-
fidam, virorum morte recuperare conans' (Weil), or, in Verrall's refined ver-
sion : ‘that thou for a willing wanton wouldst spend the lives of men’. Against
this it can hardly be objected that here the word θάρσος, not θράσος, is used
in a depreciatory sense. It is true that there seems to be no evidence for this
use in Tragedy, whereas the opposite, θράσος used in a good sense, undoubtedly
occurs several times (cf. Elmsley on E. Med. 456 in his numbering of the
lines) ; cf. also R. A. Neil on Ar. Knights 303. But the differentiation of the
meanings is secondary (cf. Wilamowitz on E. Her. 624) : it is first expressly
attested by Arist. Eth. Eud. 1234 b 12, and then by the Atticists (cf. also
Schol. E. Med. 469 and the quotations given there by Schwartz), and no one
can say whether it held good for Aeschylus. However, it seems questionable
if a person could be called ἑκούσιον θράσος. It cannot be doubted that plain
θράσος, even when not used in the vocative, could be so employed, and also
E. Andr. 261 ὦ βάρβαρον σὺ θρέμμα καὶ σκληρὸν θράσος is quite easy, since
364
 COMMENTARY line 806
σκληρόν and θρασύ are analogous ideas, so that the adjective here only
intensifies θράσος. On the other hand, ἑκούσιον θράσος would be at the best
much more difficult, since it is not her nature but her action that is ἑκούσιον.
But the decisive argument against the interpretation adopted by Weil is
the rendering of ἀνδράσι θνήισκουσι by ‘virorum morte’, ‘at the cost of good
men’s lives’ (Headlam). Even in itself it is difficult to believe in such an
instrumental of the person in sentences where the addition of an instrumental
qualification to the verb is not immediately suggested by the context, as
Ag. 616 and in the cases I have mentioned there. Moreover, in the present
sentence such an interpretation is feasible only on paper, since every hearer
must immediately connect together ἀνδράσι κομίζων, exactly as in Pind. N.
7. 28 ξανθῶι Μενέλαι δάμαρτα κομίσαι and in other passages, so that the dative
necessarily suggests the description of the person to whom the thing is being
returned. But, as no satisfactory explanation of ddpoos ἑκούσιον has yet been
propounded, we are probably not justified in adopting the metrical conjecture
of Triclinius, but shall have to return to the θράσος of F. Since I know of no
really convincing emendation of the whole expression, I must content myself
like Sidgwick with recognizing the grave corruption as such. This being so,
it seems impossible to determine with any degree of certainty the meaning
(and the authenticity) of θνήισκουσι. ‘Quid sit dvöp. θνήισκ. vix satis intelligo’
(Blomfield). Ahrens (560) has referred it to the casualties in Aulis caused by
‘starvation and sickness’. But this puts things too much out of perspective.
Here where they are speaking of the reproaches that were levelled at the king
for sending thousands of men to the war overseas just “EAevns ἕνεκα, they must
be thinking of the gigantic losses to be expected in the field, and not of the
relatively insignificant casualties before sailing. It would be a very bold
expression to call the warriors who are going to their death ἄνδρες θνήισκοντες :
I could not assert that it is too bold for Aeschylus. But since it is not possible
to make out the original context, this detail, too, remains uncertain.
805. οὐκ ἀπ᾽ ἄκρας φρενός: finally explained (‘non ex superficie mentis;
ex ima mente, sincere’) and illustrated by parallels by Blomfield, who quotes
e.g. E. Hec. 242 ob γὰρ ἄκρας καρδίας ἔψαυσέ μου. Wilamowitz, in his com-
mentary on Cho. 55, uses the expression διὰ φρενὸς περαΐνειν which occurs
there to illustrate the idea that underlies the present passage.
οὐδ᾽ ἀφίλως (= καὶ οὐκ ἀφίλως) εὔφρων is defended in the note on 323. The
words νῦν δ᾽... εὔφρων are correctly understood by Blomfield: ‘hodie vero ex
alta mente et vere benevolus’. We have here two adverbial phrases intensify-
ing the notion of εὔφρων, both in the form of ‘litotes’: οὐκ am’ ἄκρας φρενός
and οὐκ ἀφίλως.
806. εὔφρων: translated ‘graius’ by Stanley (Heath disagreed), and it has
often been maintained since his time that the word here means gratus,
acceptus ; so, among others, Hermann, the Aeschylus-lexica of Wellauer and
Dindorf, and with particular emphasis Headlam, who says: “εὔφρων means
pleasant, agreeable, welcome’, for which he quotes a number of instances from
Aeschylus and Pindar. The understanding of this and other passages in
which εὔφρων occurs would not have become so confused, and their treatment
in the dictionaries' would be more adequate, if more attention had been paid
* A particularly unfortunate specimen is the article in L-S. There sections I and II
contain several wrong interpretations, and there is no justification whatever for the
365
line 806 COMMENTARY
to Ahrens’s (561 ff.) excellent discussion ; Hense, in his appendix to Schneide-
win, is a shining exception. Ahrens has, in agreement with Weil, proved that
the supposition of the meaning gratus acceptus is not required in any passage
at all: ᾿εὔφρων has in the main three meanings: (1) friendly disposed, like
εὐμενής, (2) in good spirits, cheerful, like εὔθυμος, (3) reasonable.’ In order to
test these conclusions, the instances quoted by Headlam for the meaning
pleasant, agreeable, welcome may first of all serve. In Suppl. 20 (rightly
translated by Headlam himself: ‘what friendlier land’) χώραν εὔφρονα means
‘well-disposed’, *well-wishing', and 972 τόπος εὔφρων exactly corresponds;
Suppl. 534 is rendered correctly by Wecklein ‘erneuere . . . die uns freundliche
Kunde’. Pind. Ol. 2. 36 τὸν εὔφρονα πότμον has been well compared by Ahrens
to Ol. 14. 14 em’ εὐμενεῖ τύχαι, Pind. N. 7.67 ὁ δὲ λοιπὸς εὔφρων ποτὶ χρόνος ἕρποι
(Schol. A «páios ὧν καὶ εὔφρων) means ‘may the future be gracious’ (Wila-
mowitz, Berl. Sitzgsb. 1908, 338). All that is left of Headlam’s list is Suppl. 378
οὐδ᾽ αὖ τόδ᾽ εὗφρον, where as in Cho. 88 the word is taken rightly by Wellauer,
Dindorf, Wecklein, Blass as ‘besonnen’, 'verstándig' (‘reasonable’) ; Wilamo-
witz has corrected (Hermes, lxii, 1927, 284) the different rendering (to which
the translations of Headlam and G. Thomson have adhered) which he pro-
posed in his commentary (p. 162) on Cho. 88. We find the same meaning in
Pers. 772 and with εὐφρόνως in Ag. 351 and 849. εὔφρων could at any time take
the meaning 'sensible, reasonable', and the like, because the 'sense' meant by
the -ῴρων part of the compound includes both the side of ‘feeling’, ‘disposition
towards someone’, and that of ‘sensibleness’. Itis worth noting that Aeschylus
uses εὖ φρονεῖν in the meaning ‘be sensible’ in Prom. 385, and also as ‘to be
well-intentioned’ in Ag. 271, 797, and elsewhere,’ and that δύσῴφρων means
‘senseless’ (Sept. 874) as well as ‘ill-intentioned’ (Ag. 608 and elsewhere). But
to return to Ag. 806: here εὔφρων is understood by Dindorf in the Thesaurus,
Linwood, and L-S in the sense of ‘cheering, making glad’. Ahrens again has
seen that no instance makes it necessary to assume this meaning, which goes
on occupying a considerable space in our dictionaries ; I differ, however, from
Ahrens in the interpretation of the passages in question. The whole trouble
arises from the inappropriate interpretation of a passage in Homer, which is
not only not typical but exceptional, Γ 246. Quite normally and understand-
ably (even if we do not take into account the well-known meaning of εὐφρο-
σύνη, which is to be found in the Odyssey and later on) εὔφρων describes the
person who is in a cheerful mood, so O 99 εἰ πέρ ris ἔτι viv δαίνυται εὔφρων
(Schol. BT εἴγε ὅλως φιλεῖτε ὑμεῖς εὐωχεῖσθαι εὐφραινόμενοι τὴν ψυχήν). Similarly
in Pindar, N. 5. 38, where εὔφρονες {a is rightly referred by the scholia to the
kwudlovres, the festal choir, and connected with εὐφροσύνη (cf. Bacchyl.
11(1o). 12 of the celebration of victory: κῶμοί τε καὶ εὐφροσύναι). In spite of
these passages Diels (who is followed by Bury, C.A.H. iv. 491) translates
Xenophanes fr. 1. 13 εὔφρονας ἄνδρας (where incidentally the mention in 1. 4
of the κρατὴρ. . . μεοτὸς ἐυφροσύνης helps out the point) as ‘verstandige
existence of section III ‘= εὔφημος᾽, where the older editions also quote Xenophanes
fr. 1. 13 (see my following remarks in the text), while the ninth quotes only A. Suppl. 378,
Cho. 88. Wilamowitz’s remark (Heimkehr des Odysseus 192) on εὔφρων is too one-sided and
therefore misleading, perhaps only because of its brevity. J. Boehme too, Die Seele und
das Ich im homer. Epos, 38 n. 3, only touches on the problem of εὔφρων with a reference or
two; at least he might have done more in the case of I’ 246 than quote Autenrieth-Kaegi.
! Cho. 774 is wrongly translated and punctuated by Wilamowitz.
366
 COMMENTARY line 806

Männer’ (‘sensible men’) ; Kranz translates it: ‘wohlgesinnte Männer’ (‘well-

intentioned men’). The word is used by ‘Theognis’ in precisely the same way
as by Xenophanes, 765 eüdpova θυμὸν ἔχοντας (developed in the pentameter by
νόσφι μεριμνάων εὐφροσύνως διάγειν) in a section which possibly draws on that
very passage of Xenophanes (cf. F. Jacoby, Berl. Sitzgsb. 1931, 100). So when
we find in I’ 246 the expression οἶνον ἐΐῴρονα, καρπὸν ἀρούρης, we shall see in
this a solitary instance of the boldness of a poet who says of wine, the chief
bringer and begetter of εὐφροσύνη at a feast, what primarily applies to those
on whom it produces its effects. We ought not to soften down the peculiarity
by translating ‘that makes glad the heart’, but must cling to the meaning
‘cheerful’, ‘glad’ wine. Hesiod, Erga 775 εὔφρονα καρπόν is borrowed from
T 246, and consequently no independent evidence; E. Cycl. 507 ὁ φόρτος
(Wilamowitz, χόρτος MS) εὔφρων carries on the Homeric notion in a parod-
istic manner. We should forgive the interpreter whom Eustathius follows
when he helps himself out of the difficulty in face of the strange εὔφρων in
F 246 with the paraphrase εὐφραντικός (correspondingly e.g. schol. b, p. 96.
17 Drachm. on Pind. N. 5. 38: ai la... εὔφρονες οὖσαι, τουτέστιν εὐφραντικαί,
the gloss in Tr on A. Ag. 1577: εὐφραντικόν) ; but that ought not to have been
made the foundation of lexicographical wisdom. Shall we really in Ag. 1577
ὦ φέγγος εὗφρον ἡμέρας δικηφόρου fail to recognize the well-intentioned, kindly
light? Everything else that is listed under the head ‘cheering’ has just as
little claim to be included. So far as Ag. 806 in particular is concerned, there
was here no reason whatever to cast a doubt on the interpretation put
forward by Heath, ‘well-intentioned’, ‘well-wishing’. Not only is the word
εὔφρων used in the sense of benevolus by the Chorus when speaking of their
loyalty to the king and his house in an exactly analogous context (263), but
in the present passage the phrase eddpovos ἐκ διανοίας used in this sense a few
lines earlier (797) re-echoes in our ear. We may go farther and say that the
connexion between 805 f. and 797 f. is so obvious that it is difficult to imagine
how it can have been overlooked. The Chorus have only just remarked that
the capable ruler will be able to see through the person whose facial expression
only in outward appearance comes εὔφρονος ἐκ διανοίας, and now they go on
to say about themselves (so far as the sense goes ; for the wording in detail see
below) that they are profoundly and in true friendship εὔῴρονες. Supposing
we here introduce a meaning of εὔφρων which differs from that in 797, we are
destroying a carefully arranged connexion. The gist of the section 799-806
is: 'I was critical of you at the time of the start of your expedition, but
now I am thoroughly good-hearted towards you.’ The contrast between
the attitude of false friends as characterized 788-98 and the trustworthy
loyalty of the old men (799-806) is picked up again in 808 f. by the words τόν
τε δικαίως καὶ τὸν ἀκαίρως πόλιν οἰκουροῦντα πολιτῶν, and so rounded off.
πόνος: this, as Hermann saw, should be kept and the usual ‘ellipse’ of
ἐστί (i.e. in reality the nominal form of the predicate) assumed. Others have
read πόνον with Auratus and understood eiui with εὔφρων. ‘Ellipse’ of ei is
in itself not impossible, although it is rare in Aeschylus as in other writers too
(in the Aeschylean examples ἐγώ is always present! except for Cho. 412: there
1 Cf. £ 372 αὐτὰρ ἐγὼ wap’ ὕεσσιν ἀπότροπος. This particular type of sentence with a noun
predicate seems (like the very common type ὁ παῖς καλός) to belong to the inherited stock of
the Indo-European languages; A. Meillet, Introduction à l'étude compar. des langues indo-

307
line 806 COMMENTARY
we have μοι in the preceding clause and also in the parallel kolon which fol-
lows) : Cho. 135, 412, Prom. 246 (no lacuna follows, cf. besides the more recent
editions Denniston, Particles 353 f., Pohlenz, Gnomon, ix, 1933, 625); a
doubtful case is Eum. 780 ff., where text and interpretation are very contro-
versial. Cf. on Ag. 1652 ; examples from Sophocles and Euripides in Denniston
on E. El. 37.
van Heusde, Verrall, and Headlam hold that probably a proverbial ex-
pression underlies the words . . . πόνος εὖ τελέσασιν, conveying approximately
‘all is well that ends well’ ; for the thought cf. the passages quoted by Headlam
and G. Thomson (also in the Supplement). This conjecture finds a support
in the fact that when the old men utter such a thought now that the king
and his host are really returned re bene gesta, their feeling comes pretty near
to that which in 567 f. was expressed in the words of the Herald though their
tone is of course more restrained. The attempt to make εὔφρων the immediate
predicate of πόνος breaks down under the difficulty that then a meaning must
be ascribed to εὔφρων for which there is no sure evidence (see above). There-
fore εὔφρων must be connected with what precedes and consequently a lacuna
marked before πόνος. The assumption of a lacuna is in any case probable
from the moment that the proverbial character of what is said of the πόνος
in 806 is recognized; one then expects something to make it clear that this
θρυλούμενον expresses the present state of feeling of the speaker. So Headlam
was on the right track when he inserted ἔστιν ἐπειπεῖν, only he placed the
lacuna, as Schneidewin had done before, in front of εὔφρων, because he
thought that εὔφρων and πόνος should be taken together. Hense was probably
nearer the truth (in his appendix to Schneidewin’s edition) when he said that
‘the words οὐκ ἀπ᾿ ἄκρας φρενὸς οὐδ᾽ ἀφίλως εὔφρων hang together, and it is
preferable to assume the dropping out of four or six anapaests between
εὔφρων and sróvos'. It would be idle to try to restore the words ; I imagine the
thought to have been something like this: viv δ᾽ οὐκ da’ ἄκρας φρενὸς οὐδ᾽
ἀφίλως εὔφρων (αἰνῶ τόδ᾽ ἔπος προτέρων ‘450s (van Heusde quoted Eur. fr. 133
N. ἀλλ᾽ ἡδύ τοι σωθέντα μεμνῆσθαι πόνων, and G. Thomson Democritus fr. 243)
or else φροῦδος (cf. 567 ff.) or ἄπονος) πόνος εὖ τελέσασιν᾽, For the metrical
form of the proverb, the paroemiac, cf. on 1527.
A special warning must be given against an apparently simple method of
avoiding the marking of a lacuna, 1.6. altering πόνος to λόγος (Kayser), νόος,
πόλις (both M. Schmidt), πνόος (Weil), and several similar suggestions. The
final objection against them lies in the fact that, as has been shown, the MS
reading πόνος εὖ τελέσασιν points to a widespread ‘sententia’ and one which is
particularly appropriate here. But even in itself none of the proposed nouns
really is apposite. If we had to choose one, we might find a soft place in our
heart for νόος, against which no formal objection holds good (νόον occurs even
in trimeters, Cho. 742) and, so far as the meaning goes, possibly my remarks
on 1229 (φαιδρόνους) may suffice to defend it, although in tragedy the noun
νοῦς does not occur in a corresponding sense.
807. διαπεύθομαι only occurs here, διαπυνθάνομαι is first attested in Plato.
808. ἀκαίρως: the opposite of κατὰ καιρόν, ‘appropriately’ (Wilamowitz,
europ., 3rd ed., 341, quotes the old Persian adam navama ‘moi, je suis le neuvième”, ‘où la
1re personne étant exprimée par le pronom adam “moi”, le verbe “être” à la 17* personne
ne figure pas parce qu'il était superflu”.
368
 COMMENTARY line 809
Hermes, xv, 1880, 509 = Kl. Schr. i. 44), so ‘in an inappropriate way’, in a way
unsuitable to the given circumstances. ‘A somewhat mild and modified way
of expressing döikws’ (Paley).
809. oikoupoüvra. The simple and clear word is here rendered either vaguely
(e.g. Stanley: 'quis civium iuste . . . se gesserit in civitate', so still Verrall:
"who here at home hath done his duty', and Wilamowitz) or definitely wrongly
(e.g. Paley: 'has managed state affairs', Headlam : ‘whose stewardship of the
state among thy folk is honest’). The sound remark of Schütz ('oikovpetv
πόλιν h.l. non est administrare sed tantum versari in civitate") has not been
heeded. The dictionaries too lead us astray. So far as the exact meaning of
οἰκουρός, οἰκουρεῖν, etc. in classical times is concerned, it is plain that the
etymology as well as the conventional paraphrase of the scholia and lexico-
graphers is a hindrance rather than a help. Doubtless the word is compounded
from οὖρος, ‘watcher, guardian’. Recognition of the obvious derivation has
influenced the ancient grammarians: Hesychius οἰκουρός ὁ φροντίζων τὰ τοῦ
οἴκου Kai φυλάττων. οὖρος yàp ὃ φύλαξ λέγεται. The scholia correspond with
this, e.g. on Ar. Lys. 759 (τὸν ὄφιν... τὸν oikovpóv) τὸν φύλακα τοῦ ναοῦ (cf.
Hesychius and Photius s.v. οἰκουρὸν ὄφιν) and elsewhere. But if we look closer
into the question, we see that in none of the passages quoted in support from
the older literature does the word convey in itself the idea of watching; it
simply signifies the person who 'keeps at home’, habitually stays inside the
house and does not go out. In Ar. Wasps 97o the one dog, who οὐδέποτ᾽ ἐν
ταὐτῶι μένει, is contrasted with the other who is oikovpós μόνον (rightly para-
phrased by the scholion ἐνδομυχῶν). The passage S. Phil. 1327 f. ὃς τὸν ἀκαλυφῆ
σηκὸν φυλάσσει κρύφιος οἰκουρῶν ὄφις shows οἰκουρῶν by no means a synonym
of φυλάσσει, but tells us that the snake never leaves the inside of the sanctuary.
The same holds good of the oikovpós ὄφις, mentioned in Ar. Lys. 759 and else-
where, that belongs to Athena on the Acropolis. Certainly it is φύλαξ τῆς
ἀκροπόλιος (Hdt. 8. 41. 2), but that is not conveyed by the term οἰκουρός in
itself, at any rate for the Athenians of the fifth century. The original (as
proved by its etymology) ‘watcher’, ‘guardian of the house’, had by then
been weakened down to the meaning ‘one who keeps the house’ or ‘one who
keeps at home', though it lies in the nature of things that the wife, who
remains at home while her husband is abroad, or the οἰκουρὸς κύων, often has
the task of watching the house. For οἰκουρεῖν being chiefly a concern of
women cf. on 1626. There I shall discuss the development which has brought
it about that οἰκουρός is used in a depreciatory sense of the ‘stay-at-home’ in
contrast to those who do their duty in the field as soldiers.! This is the sense
of οἰκουρός in the only two passages in which it is attested for Aeschylus, Ag.
1225 and 1626, both times applied to Aegisthus. Here too we have a clear
hint at Aegisthus : he is ἀκαίρως οἰκουρῶν because he misuses his remaining at
home for shameful acts. In contrast with him stand the old men, δικαίως
οἰκουροῦντες, since they are long past the age for service and also because
during the absence of the king and his host they have kept their loyalty. The
addition of πόλιν to oikovpoóvra expresses the fact that it is not a case here of
someone who remains in the house and never goes into the street, but of the
man who stays in Argos while the others are fighting overseas.
The expressions in 808 f. are kept generalized and are carefully chosen,
1 This is quite clearly the meaning of οἰκουρεῖν in Ar. Ach. 1060.
4872.2 Bb 369
line 809 COMMENTARY
but the Chorus may well hope that the king will understand the warn-
ing against Aegisthus: the old men do not dare to express themselves
more clearly.

There can be no doubt that Agamemnon enters shortly before, or at the

moment when, the Chorus begin the anapaests (783 ff.). Some details of the
manner in which he makes his appearance are less certain. In the ὑπόθεσις
of the play we find this summary of the scene: Ayauéuvwr δ᾽ ἐπὶ ἀπήνης
ἔρχεται" εἵπετο δ᾽ αὐτῶι ἑτέρα ἀπήνη, ἔνθα ἦν τὰ λάφυρα καὶ ἡ Κασσάνδρα. This
statement has been repeated by many editors and translators in their stage-
directions (recently e.g. by Murray, Mazon, A. Y. Campbell, G. Thomson).
They seem to regard it as a piece of evidence, either for the intention of the
poet or for the manner in which the scene was (regularly? at one time?)
performed in antiquity. Now Wilamowitz, Interpr. 171 f. and 175 n. 1, has
noted that nothing in the ὑπόθεσις points to its autlıor being acquainted with
a stage performance of the play ; on the contrary, the writer, rather a stupid
man,’ depended entirely on the words of the text, which at least in one
instance? he grossly misunderstood. His assertion must therefore be treated
as a mere guess. When we now turn to the text of the play, we find that it
nowhere indicates the number of wagons used, which is not surprising since
this detail is irrelevant to the dialogue. There is nothing left, then, but to
weigh general probabilities. This does not take us very far. All we can safely
say is that there is no need for assuming more than one wagon ; in it Cassandra
would be seated behind or beside Agamemnon. I am inclined to believe that
Franz, Kennedy, Keck, Wecklein, Lewis Campbell, Wilamowitz, E. Petersen
(D. att. Tragödie, 231), and others are right in being content with one wagon
only. This arrangement seems to be recommended by its simplicity, but that
is perhaps a matter of taste. It might also be argued in favour of the one-
wagon hypothesis that when (somewhere about 950) Agamemnon dismounts,
he enables the spectators to have a full view of Cassandra, who, seated near
him, has up to this moment been partly hidden by him. But such an argu-
ment would be fallacious, for, whatever the arrangement on the stage, accord-
ing to the rules of early Attic drama no notice should be taken of Cassandra
before attention is drawn to her by some explicit phrase.
If the existence of a second wagon is doubtful, the assertion that upon it,
besides Cassandra, there were τὰ λάφυρα is quite arbitrary. If in this scene
pieces of booty mattered at all, they ought to be mentioned (cf. Wilamowitz,
Interpr. 172). But the superfluities with which the ὑπόθεσις credits the play
of Aeschylus have not been sufficient to satisfy the appetite of some modern
readers, who have invented further details of a spectacular show, regardless
of the difference between the time of Aeschylus and a period when on the
tragic state esseda festinant, pilenta, petorrita ... captivum portatur ebur. The
1 Aristophanes of Byzantium cannot be made responsible for this ὑπόθεσις, not even
indirectly ; his manner of presenting the summary of a play is quite different from what we
find here. Cf. Achelis, Philol. Ixxiii, 1914-16, 147.
* ἰδίως δὲ Αἰσχύλος τὸν ‘Ayauéuvova ἐπὶ σκηνῆς ἀναιρεῖσθαι ποιεῖ. Even in this case the
generosity of Verrall (Introd., p. lvii) provides a nice loop-hole.
3 The craving for display of unwarranted pomp at the performance of classic tragedies
was of course much older than the age of Cicero or Horace. Cf. Wilamowitz, Einleitung
in die griech. Tragödie (Eur. Herakles I, ıst ed.), p. 152 n. 62 (Schol. E. Or. 57). It is, in

370
 THE ENTRANCE OF AGAMEMNON
gorgeous triumphal procession amidst which Agamemnon is supposed to make
his entrance is still highly popular although it was exploded as long ago as 1846
by G. Hermann, Opusc. viii. 162 f. (= Aeschyli tragoedtae, ed. G. H., ii, p. 651).
He replaced the alleged ‘currus triumphalis’ by a more homely vehicle, ‘com-
mune vehiculum, quali iter facientes utuntur, idque nescio an mulis iunctum'.
In this he was rightly followed e.g. by Kennedy, Headlam, A. Y. Campbell.
Nothing in the text points to Agamemnon's coming with a retinue. There-
fore Wilamowitz, whose translation contained the stage direction 'Some
warriors accompany the wagon', explicitly recanted afterwards: 'currum
satellites non comitantur’, cf. also his Interpr. 172: ‘if Agamemnon brought
a retinue with him, the scheme for his murder would be endangered.'

810. πρῶτον μέν ... (cf. Eum. 1), extended to great length, picked up again
in 829 θεοῖς μὲν κτλ., then in 830 the δέ clause follows.
811. δίκη here as in 259, Sept. 866, and Eum. 277 approaches a somewhat
colourless ‘duty’ (R. Hirzel, Themas, 105 n. 2).
τοὺς ἐμοὶ μεταιτίους xrÀ.: many commentators have been reluctant to
accept the obvious meaning of the phrase, and have either given perairios a
‘weakened’ meaning or discovered in it the expression of a feeling which is
completely foreign to it. Stanley comments on μεταιτίους : 'Pro simplici
αἰτίους. Aliter Eum. 199.’ The mistaken equation of μεταίτιος With αἴτιος
originates, according to Dindorf's Thesaurus, with Budaeus. It was adopted
by Hermann: ‘kerarriovs dixit, ut saepe solent, pro simplici αἰτίους. Luculen-
tum extat exemplum in Soph. Trach. 1234.' That is all. The 'luculentum
exemplum' proves the opposite of what it is supposed to prove. It is true
that the words of Hyllos τίς γάρ ποθ᾽, 7 (Iole) μοι μητρὶ μὲν θανεῖν μόνη
μεταίτιος κτλ. are paraphrased by the Scholiast τὴν τῶν κακῶν αἰτίαν οὖσαν, but
it should be clear that only the general line of thought is indicated, and no
detailed rendering is intended. Modern commentators, certainly, have seen
that μεταίτιος bears here the same meaning of ‘accessory’ as it does in the
only two other passages in Sophocles where it occurs. They are in the same
play, in 260, where, just as in 1233 f., μόνον stands next to perairiov, and in
447. But, of course, even Hermann’s interpretation! is preferable to the
artificialities of L. Campbell (‘ "sole sharer of the blame" with Nessus’) and
Jebb (‘means that she alone shared the blame with Heracles’: and this from
a most loving son to his father dying in agonies!). The implication of the
passage seems obvious. Deianeira has taken her own life: she herself was
primarily αἰτία τοῦ θανεῖν so that the person who, through being her rival, has
driven her to death appears only as μεταίτιος.2 Nowhere either in Aeschylus
or in Sophocles is μεταίτιος equivalent to αἴτιος : it always denotes a share of
responsibility. The distinction is illustrated not only by Eum. 199 f. but also
by Hdt. 4. 200. τ ἐκδιδόναι τοὺς αἰτίους τοῦ $óvov . . . τῶν δὲ πᾶν yap ἦν τὸ πλῆθος
theory, not inconceivable that the author of the hypothesis to the Agamemnon was in-
fluenced by recollections of later performances in a more luxurious style, but from what has
been said above about his method the possibility seems remote.
1 Hermann's view was slightly modified by Zielinski, Iresione, i. 319, who rendered
μεταιτία by ‘die aktive Urheberin’, and explained in a footnote that the word really meant
‘die Mitschuldige’, but ‘only in so far as the main responsibility rests with the deity’.
2 A similar situation is viewed in a different way S. Ant. 1173 τεθνᾶσιν' of δὲ ζῶντες αἴτιοι
θανεῖν.

371
line 811 COMMENTARY
μεταίτιον, οὐκ ἐδέκοντο τοὺς λόγους. The misinterpretation customary with
μεταίτιος has also obscured the meaning οὗ παραίτιος, only here the error has
persisted more stubbornly, cf. for example Dindorf, Lex. Aesch. s.v. Even so
great a Hellenist as Blass commented on Cho. 910: “rapæria = αἰτία, but by
no means = perairia.’ In support of this he quotes, besides the passage
from Aeschylus’ Danatds (see below), the Hellenistic usage, although he
knows quite well that the weakening of the force of many prepositions which
is characteristic of compounds in that period cannot indiscriminately be
transferred to classical times. Cho. 910 has been correctly understood by
most of the recent commentators and translators: ‘Clyt. does not . . . venture
to throw the whole blame on Fate, but merely alleges that it was accessory’
(Conington). It is a fact, then, that παραίτιος, too, in the fifth century retains
its full meaning and is not merely a synonym of αἴτιος. The case of the famous
words of Aphrodite in the Danatds (Aesch. fr. 44 N.) τῶν δ᾽ ἐγὼ παραίτιος
forms no exception to this rule. Nowadays we tend to look for omnipotence
in a deity, even in a Greek one, wherever possible. Aeschylus, however, does
not make his Aphrodite? say : ‘whereof the cause am I’ (Murray, The Suppliant
Women, 13), but only that in any act of procreation, bearing, and bringing to
birth she works as a participant cause. It is, however, enough for this goddess
that nowhere in the whole wide world does life arise ἄνευ Ἀφροδίτης. Under-
standing of the important but not unlimited function here assigned to a great
deity will enable us to reject the misinterpretations of those scholars who,
while their approach to rods ἐμοὶ μεταιτίους κτλ. is linguistically sound, have
failed to appreciate its implications. We need not consider Paley’s idea ‘in
common with the heroes’ nor that of Pliiss: ‘uera:riovs: in the distant land,
on the sea, where other gods have power’ (Schneidewin had written earlier
on the same lines, but even more fantastically). But a real danger to the
understanding of Agamemnon’s attitude and so of the whole play arises from
the interpretation which in England,’ as I have heard from the lips of dis-
tinguished scholars, already seems to have become a commonly accepted
opinion.* It leads to judgements such as the following: ‘it suits the pride of
the speech to take it “helpers with me in our return” ’ (Sidgwick), ‘who with
me have contributed to etc., a strange form for the expression of religious
gratitude’ (Verrall, who in the preceding note says ‘the whole harangue is
haughty and repulsive’), ‘rods ἐμοὲ μεταιτίους recalls the arrogant inscriptions
* Consultation of earlier editions of L-S helps one to understand by what confusion
this passage in L-S, oth ed., came under the heading of ‘in bad sense, accessory to a crime’.
2 For a limitation of Aphrodite’s power in the same trilogy, at the end of the Supplices,
cf. Wilamowitz, Glaube d. Hell. ii. 152 n. 2.
3 It is not, however, confined to England. Cf., e.g., Eugen Petersen, Die attische Tragódie
(1915), 231; Mazon, Introd. (‘Notice’) to the Ag., p. 5.
* Of course in England, too, there have not been lacking readers who have been able
to understand the character as the poet created it and so to put the proper meaning upon
these opening lines. The right view was particularly well expressed by W. Sewell in the
introduction to his translation (and ed., p. xxxiv): ‘And yet all that is seen of Agamemnon
is noble, and calculated to inspire precisely that feeling of respect, which constitutes man
an object of just and rightly tempered pity. His religious appeal to the gods; his acknow-
ledgment of their hand in the just revenge which he had been enabled to take upon Troy',
etc., cf. on 904 and 915. On 948 I have again assembled for the sake of comparison some of
the conflicting modern judgements about the conduct of Agamemnon ; they have a bearing,
too, on the appreciation of Agamemnon's first speech.

372
 COMMENTARY line 811

set up by Pausanias' at Delphi and Byzantium’ (Headlam, who, to provide

a contrast, resorts to the lines of the diaskeuast Sept. 1074 f.), ‘the arrogant
opening words of Agamemnon A. Ag. 810-11’ (Denniston on E. El. 890 ff.).
These judgements? are based on the idea that humility is the most reliable
measure of true religious feeling. The view of the Greeks of the fifth century
(and I deliberately confine myself to that period) can be gathered with
certainty from the evidence, particularly that of Aeschylus himself. Orestes
says to Athena (Eum. 457 f.) ξὺν ὧι (Agamemnon) σὺ Τροίαν ἄπολιν ᾽Ιλίου
πόλιν ἔθηκας. When Darius says with bitterness in regard to the deluded con-
duct of his son (Pers. 742) ἀλλ᾽ ὅταν σπεύδηι τις αὐτός, χὠ θεὸς συνάπτεται,
(cf. also 724 γνώμης δέ πού τις δαιμόνων ξυνήψατο), he is quoting a well-known
general maxim, which is, of course, chiefly concerned with the helpful co-
operation of a god: Aesch. fr. 395 φιλεῖ δὲ τῶι κάμνοντι συσπεύδειν θεός, Eur.
fr. 432. 2 τῶι γὰρ πονοῦντι χὠ θεὸς συλλαμβάνει, Ar. Knights 229 (ot βοηθήσουσί
σοι... ἡ) κἀγὼ μετ᾽ αὐτῶν χὠ θεὸς συλλήψεται, Wasps 733 Î. σοὶ δὲ νῦν τις θεῶν
παρὼν ἐμφανὴς ξυλλαμβάνει τοῦ πράγματος καὶ δῆλός ἐστιν εὖ ποιῶν (see Wila-
mowitz, Kl. Schr. i. 335). This conviction is not peculiar to the Athenians ;3
the Delphic Apollo answers the Spartans (Thuc. i. 118. 3): κατὰ κράτος πολε-
μοῦσι νέκην ἔσεσθαι, καὶ αὐτὸς ἔφη συλλήψεσθαι καὶ παρακαλούμενος καὶ ἄκλητος.
The divine co-operation may sometimes prove harmful (as in the case of
Xerxes, see above): Ag. 1507 f. πατρόθεν δὲ συλλήπτωρ γένοιτ᾽ ἂν ἀλάστωρ (this
is preceded by ws μὲν ἀναίτιος el τοῦδε φόνου, τίς 6 μαρτυρήσων ; that means
that Clytemnestra is αἰτία τοῦ φόνου, the ἀλάστωρ is μεταίτιος OT παραίτιος).
E. Her. 1134 f. Question: who killed the children? Answer: σὺ καὶ σὰ τόξα
καὶ θεῶν ὃς αἴτιος. But such cases are, of course, rare. The confidence that
the god will take his share in fighting or working alongside man is deeply
rooted in Greek religious feeling. This is shown by a typical form of prayer:
Sept. 266 εὔχου τὰ κρείσσω, ξυμμάχους εἶναι θεούς, Cho. 2 (Ἑρμῆ) σωτὴρ γενοῦ
μοι σύμμαχός τ᾽ αἰτουμένωι (cf. Cho. 19), Cho. 812 ξυλλάβοι δ᾽ ἐνδίκως παῖς 6
Μαίας, E. Hel. 1443 f. (prayer to Zeus) ἕλκουσι δ᾽ ἡμῖν πρὸς λέπας τὰς συμφορὰς
σπουδῆι σύναψαι (for the verb cf. A. Pers. 742 quoted above), Rhesos 230
(Ἄπολλον) ξύλλαβε AapSavidas, Ar. Peace 416 f. πρὸς ταῦτ᾽, ὦ φίλ᾽ Ἑρμῇ (he is
standing near by) ἐύλλαβε ἡμῖν προθύμως, τήνδε καὶ ξυνέλκυσον, the last phrase,
too, is significant. It is not beneath the dignity of a Greek god to lend a hand
in a hard piece of work. But the main effort must come from man.’ This
conception is well illustrated by the Atlas metope of the temple of Zeus at
Olympia and many other works of Greek art. A Greek hero has not such a
! Pausanias is a great favourite with commentators who are bent on misrepresenting
the figure of the Aeschylean Agamemnon, cf. Kennedy, Journ. οὐ Philol. vii, 1877, 14;
Verrall, Aesch. Ag., and ed. 220.
2 Contrast with them the simple and wholly adequate statement of H. Weil, Études sur
le drame antique, 32: ‘Le vainqueur paraît enfin en personne; il proclame sa victoire avec
une grandeur exempte de présomption.’
3 Cf. Pindar’s paean (2) for Abdera, 64 f. τλάντων δ᾽ ἔπειτα θεοὶ συνετέλεσσαν.
* We may compare E. Hel. 1441 ff. ὦ Ζεῦ... βλέψον πρὸς ἡμᾶς καὶ μετάστησον κακῶν.
ἕλκουσι δ᾽ ἡμῖν πρὸς λέπας τὰς συμφορὰς σπουδῆι σύναψαι: κἂν ἄκραι θίγηις χερί, ἥξομεν κτλ,
Cf. too E. Hipp. 522 f. μόνον σύ μοι, δέσποινα ποντία Κύπρι, συνεργὸς eins, Ar. Knights,
586 ff. (prayer to Athena) δεῦρ᾽ ἀφικοῦ λαβοῦσα τὴν ἐν στρατιαῖς τε καὶ μάχαις ἡμετέραν
ξυνεργὸν Νίκην.
5. Cf, also the proverb (Zenob. 5. 93) σὺν ᾿Αθηνᾶι καὶ χέρα κίνει" παροιμία ἐπὶ τοῦ μὴ χρῆναι
ἐπὶ ταῖς τῶν θεῶν ἐλπίσι καθημένους ἀργεῖν.
373
line 811 COMMENTARY
good time of it as a hero of the Old Testament. ‘And the spirit of the Lord
came upon him, and he went down to Ashkelon, and slew thirty men of
them, and took their spoil’ (Judges xiv. 19). The Greek has to do what he
can with his own human strength, and the most the god does, if he happens to
be well disposed, is to assist (ξὺν dua θεὸς ὁ φίλος ἀνέτω πόνους, as the Chorus
sings in the Ichneutai of Sophocles, 1. 70). If Samson said ‘God has done this’,
he would be speaking the truth ; if Heracles said the same, he would be lying.
Undivided responsibility, therefore, for human actions, whether for good or
ill, cannot be cast upon the gods. And so, in Aeschylus, the human being
can no longer say, like Agamemnon in the {Πα (T 86 f. ; of course this attitude
is not typical of Homer, cf. Wilamowitz, Glaube d. Hell. ii. 116 ff.) : ἐγὼ δ᾽ οὐκ
αἴτιός εἶμι, ἀλλὰ Ζεὺς καὶ Μοῖρα καὶ ἠεροφοῖτις 'Epwis. The impassable
boundaries of the Aeschylean conception of human responsibility are most
clearly marked in the words of Clytemnestra in Cho. gıo ἡ Μοῖρα τούτων, ὦ
τέκνον, maparria. She is fighting for her life; how much better it would be for
her if she could credibly put the whole undivided blame upon Moira. But
that is just what she cannot do, because it would be contrary to the belief
which we have been discussing. Of course, abnormal circumstances may arise
or at any rate circumstances which may be regarded as contrary to the rule
at least from the point of view of one of the parties. This is the case with the
allegations of the Furies, in regard to the murder of his mother by Orestes;
they accuse Apollo (Eum. 199 f.): αὐτὸς où τούτων οὐ μεταίτιος πέληι, ἀλλ᾽ εἷς
τὸ πᾶν ἔπραξας ὧν mavairıos. But it is precisely here that the form and con-
tent of the sentence make it absolutely clear that the plaintiffs start from
what even to them is the normal situation, that in human action the god’s
part is that of μεταίτιον εἶναι. The same conception is at the bottom of a
sentence from the Niobe of Aeschylus (fr. 156 N.) which we now read in its
original context: θεὸς μὲν αἰτίαν φύει βροτοῖς, ὅταν κακῶσαι δῶμα παμπήδην
θέληι, which Schadewaldt (‘Die Niobe des Aisch.’, Sıtzgsb. Heidelb. Ak., Phil.-
hist. Kl., 1933-4, 3. Abhdlg. p. 10), explains as follows: ‘The god does not
send the disaster out of the bright sky, but only plants (φύει) the seed of the
αἰτία... it rests with the man whether he allows himself to be lured into
Hybris through his good fortune.’ The preceding interpretations of the most
relevant passages agree in all essential points with the detailed and adequate
picture of this attitude to be found in W. Kranz, Stasimon, 67 ff.
It appears, then, that the phrase τοὺς éuoi μεταιτίους νόστου κτλ., so far
from betraying pride or presumption, is, if measured by the standard of the
religious outlook of the poet and his times, a full and proper expression of
gratitude, exactly what in the circumstances was to be expected of an ἀνὴρ
εὐσεβής. Since we find similar expressions in prayers and elsewhere, it would
be quite wrong to see in the words of Agamemnon an indication that the
poet wants to make him appear in an unfavourable light.
812. δικαίων θ᾽ dv ἐπραξάμην κτλ. : ‘the satisfaction of the legal claims which
I have exacted’. Hermann compared E. Iph. T. 559 ὡς εὖ κακὸν (adjective)
δίκαιον εἰσεπράξατο. For the use of the plural in this meaning cf. Demosth. 2.
! For the transformation of this idea, to be found as early as the time of the later Epic,
cf. W. Jaeger, Berl. Sitzgsb. 1926, 73 ff., and especially Pfeiffer, Philol. Ixxxiv, 1929, 146 f.,
and also Jacoby, Die Antike, ix, 1933, 186 f.; G. Pasquali, Pagine meno stravaganti, 75 ft. ;
Theiler, Gnomon, xii, 1936, 170 f., and in Phyllobolia für P. Von der Mühll (Basel 1946), 56 ff.

374
 COMMENTARY line 814

24 Aaredayuovios . . . ὑπὲρ τῶν ᾿Ελληνικῶν δικαίων ἀντήρατε and immediately

afterwards iv’ οἱ ἄλλοι τύχωσι τῶν δικαίων, 24. 3 τῶν ὑμετέρων δικαίων προ-
ίστασθαι, Aeschines 2. 21 περὶ τῶν δικαίων τῶν ὑπὲρ Ἀμφιπόλεως καὶ τῆς ἀρχῆς
τοῦ πολέμου.
813. δίκας : here, of course, the legal claims, pleas, of both parties (for this
meaning! in general cf. R. Hirzel, Themis, etc. 105 f.). For this reason alone
the proposal of Ahrens, accepted by Verrall, to take ἀνδροθνῆτας with δέκας,
falls to the ground.
οὐκ ἀπὸ γλώσσης : brought forward with emphasis to provide strong con-
trast. The point is that the gods, by virtue of their own divine insight, hear
the claims direct, and not as a human judge does by way of speeches from the
parties involved and the examination of witnesses. van Heusde compares
Aeschines 1. 92 (about the council of the Areopagus) οὐ yap ἐκ τοῦ λόγου μόνον
οὐδ᾽ ἐκ τῶν μαρτυριῶν, ἀλλ᾽ ἐξ ὧν αὐτοὶ συνίσασι καὶ ἐξητάκασι, τὴν ψῆφον
φέρουσι. To avail himself of the evidence of human witnesses in forming his
judgement would be unworthy of the Lord of all Justice, Zeus, who παρ᾽
ἑαυτῶι τὸ δίκαιον ἔχει (Prom. 187), not only in the bitter sense intended by his
opponent. Here the urge towards the purification of the idea of God is quite
akin to that in Pindar, P. 3. 28 diev . . . “Λοξίας, κοινᾶνι map’ εὐθυτάτωι γνώμαν
πιθών, πάντα ἰσάντι νόωι (now compared by Daube also, Rechtsprobleme, 105
n. 34), Where the attack on the 'Hesiodic' story of the raven is unmistakable
and was noticed in antiquity.” There is a connexion, too, with Suppl. 934 f.
οὔτοι δικάζει ταῦτα μαρτύρων ὕπο Ἄρης «rÀ., although there, in keeping with
the uncouth nature of the herald, it is not so much a case of superior insight
as of superior strength ; the god of War makes no bones about it, 'er macht
kurzen Prozess'. For the train of thought and the attitude of Agamemnon it
is important to notice that in his opinion the gods, whose stern sentence he
and his army have carried out, act by virtue of superior knowledge.
814. κλυόντες : for the accent cf. note on 680.
ἀνδροθνῆτας is linguistically unobjectionable ; cf. Ernst Fraenkel, Nomina
agentis, 1. 79 n. 2, E. Williger, Sprachl. Unters. 45. “Voting pebbles which
condemn men to death'. For the qualifying attribute with ψήφους (ἔθεντο)
cf. Sept. 198 ψῆφος κατ᾽ αὐτῶν ὀλεθρία βουλεύσεται. The reading 'JAiov φθοράς
still finds its champions. For a long time (Paley follows Hermann in this) the
defence has rested upon the appeal to the frequent occurrence of an object in
the accusative depending on a periphrastic expression in the same way as it
would on the simple verb, the place of which roughly speaking the peri-
phrasis has taken: e.g. Sept. 290 f. (with regard to the usage in general cf.
Jebb on Oed. C. 223, Wilamowitz on E. Her. 690 and on Aesch. Suppl. 150,
Kaibel on S. El. 123 f.). But, as has been pointed out by Emperius, Karsten,
and Ahrens (565), the assumption of such a construction here (= ’IAiov
φθορὰς ἐψηφίσαντο) is completely ruled out by the addition of és αἱματηρὸν
1 A particularly clear example is Cho. 461 Ἄρης "Ape: ξυμβαλεῖ, δίκαι δίκα : ‘just claim with
claim’ (L. Campbell). No less clear is Eum. 491 ff. εἰ κρατήσει δίκα (re? καὶ βλάβα τοῦδε
μητροκτόνου, where most commentators understand correctly, with Stanley, 'si vicerit
causa noxia huiusce matricidae', while Blass, curiously enough, explains: δέκα xai BAáfa
must be the just penalty’.
. ? A similar idea is found in Prom. 286f., where it is said of Oceanus that he is τὸν πτερυγωκῆ
τόνδ᾽ οἰωνὸν γνώμηι στομέων ἄτερ εὐθύνων : in this his divinity is manifested as opposed to
human riders and charioteers.

375
line 814 COMMENTARY
τεῦχος ; it compels us to take ψήφους ἔθεντο literally as a description of the
action of hand and arm in reality and not as a periphrasis for ‘they decided’
or anything of that kind. If it is desired to retain the MS text, the only
possibility that remains is to take ῴθορας as genitive. This could be supported
by quoting E. Or. 1013 ψήφωι θανάτου κατακυρωθείς. But the easy emendation
of Karsten 'IAvdópovs! provides something stylistically much more satis-
factory. For the parallel formation of the two adjectives helps to give the
phrase the full Aeschylean weight,” cf. for example Suppl. 700 προμαθὶς
εὐκοινόμητις ἀρχά, and especially the phrase coined in Eum. 186 f., καρανι-
στῆρες ὀφθαλμωρύχοι δίκαι. For the formation of ᾿Ιλιοφθόρος cf. Timotheos,
Pers. 132 ᾿Ιλιοπόρος, and in the Altar of Dosiadas 17 ᾿Ιλιοραιστᾶν ἀρδίων. It is
true that Dosiadas may have invented mannerisms of his own (cf. Wilamo-
witz, Texigesch. d. Bukol. 246 f.), but ᾿Ιλιοραίστης has probably been taken
over from earlier poetry (in any case the Homeric words θυμορραίστης, kvvo-
paiorns are likely to have had some influence, cf. Ernst Fraenkel, Nomina
agentis, i. 44) ; in this connexion the Aeschylean τέκνωμα in the same poem of
Dosiadas (l. 4) should be noted.
814-16. Between the initial and the concluding elements of the hyperbaton
ἀνδροθν. ᾿Ιλιοφθ. ψήφους there is first a local adverbial phrase, then an adverb
of manner. The type of hyperbaton, ‘Sperrung’, is the common one which
has been discussed on 165. Here the insertion of two adverbial phrases ex-
pands the hyperbaton, with the result that the emphasis laid upon avôpobr.
’IAı060. ψήφους is strongly marked. Apart from its own significance this
clause together with aigarnpov forms one part of the antithesis of which the
rest is expressed in τῶι δ᾽ ἐναντίωι κύτει κτλ.
815. οὐ διχορρόπως: cf. on 1272.
816. ψήφους ἔθεντο. For the well-established use of the Middle in this con-
nexion cf. L-S τίθημι A IL 5. Cf. on 32.
816 f. τῶι δ᾽ ἐναντίωι κύτει κτλ. The Scholiast failed to understand this. He
imports into this passage the practice of the time of Demosthenes, when each
juryman received two ψῆφοι, one πλήρης the other τετρυπημένη. The pro-
cedure of the fifth century is, roughly speaking, as follows: each member of
the jury receives only one ψῆφος, which he has to put either into the urn of
condemnation, or into the urn of acquittal; cf. on the subject Blass’s note on
A. Eum. 734-43 and J. H. Lipsius, Das ait. Recht u. Rechtsverfahren, 924 fl.
With regard to the present passage Ahrens (565 f.) rightly insisted that here
only the older procedure can be meant. Then he proceeds to depict the
details: ‘If the κρύβδην ψηφίζεσθαι is not to be stultified by this [the voting
with only one ψῆφος], the juryman, while actually putting his voting pebble
into one urn, would also have to go to the other and make as if to put a pebble
into that, too. This is expressed here by the words τῶι δ᾽ evavrian . . . πληρου-
τ For this kind of corruption cf., e.g., 1092, 1471 (καρδίαι δηκτὸν instead of καρδιόδηκτον),
Suppl. 198, where Porson’s (Pref. to E. Hec. xxxvi) μετωποσωφρόνων is certain, ‘a thoroughly
Aeschylean ῥῆμα γομφοπαγές᾽ (Headlam). Ar. Ach. 376 is instructive, where the genuine
ψηφηδακεῖν, now established by the lemma of the ancient scholion (Pap. Oxy. 856, col. 1. 23)
was at first (in R and A) corrupted to ψηφοδακεῖν, then further to ψήφωι δακεῖν; in the
Renaissance MS B the reading ψηφηδακεῖν is due to a conjecture (cf. in general Wilamowitz,
Berl. Sitzgsb. 1911, 506 f. — Kl. Schr. i. 323 £.; Coulon, Introduction, p. xviii n. 3).
2 The pattern is not, of course, peculiar to Aeschylus, cf., e.g., E 639 ἐμὸν πατέρα θρασυ-
μέμνονα θυμολέοντα.

376
 COMMENTARY line 819
μένωι, 1.6, the hand approaches the urn and arouses the expectation that a
voting pebble will be put in, without actually doing so, with the result that
the urn is not filled’. This idea has been adopted by Wilamowitz (Arist. und
Ath. ii. 332 and note on Ag. 817) ; but it is not required. Complete justice is
done to the text if we take it that during the voting the poor acquittal-urn
is always hoping that a hand will appear to fill it, but all in vain: it is ap-
proached only by expectation and never by the hand itself. The choice of the
verb contributes to make the bitterness of the disappointment doubly per-
ceptible. To enter into the feelings, asit were, of the acquittal-urn we need not
suppose that the juryman comes and stands directly over the aperture,
pretending to drop his pebble. It ought to be clear that the expectation of
the urn is aroused (that the ἐλπίς approaches it) at the moment when the
man gets up from his seat and walks to the urns. The reconstruction,
attempted by Ahrens and Wilamowitz, of voting procedure in the fifth
century is definitely refuted by Ar. Wasps 987 ff., where Bdelykleon leads his
father up to only one xadioxos, the wrong one, and there is no place at all for
any pretence near the other.! It is not known whether in the fifth century
the secrecy of voting was safeguarded at all, and if so, by what means that
purpose was attained.
818. For νῦν ἔτι Headlam quotes Menander and later writers ; earlier parallels,
however, are provided by Sept. 708, S. Oed. R. 557, Phil. 950, Ar. Wasps
1089, etc.
The fact that the Achaeans set fire to Troy before their departure was men-
tioned in the ᾽Ιλίου πέρσις : see the excerpt of Proclus ἔπειτα ἐμπρήσαντες τὴν
πόλιν κτλ., [Apollod.] Est. 5. 23, cf. Quint. Smyrn. 14. 393 f., Tryphiod. 680.
Cf. R. Heinze, Virgils ep. Techn. 3rd ed. 27, Robert, Heldensage, 1274 n. 6.
This detail is alluded to in, e.g., E. Hec. 1215, Sen. Ag. 459 Iliacus atra fumus
apparet ποία. What the reference is in Pindar fr. 185 Schr. ἔτι δὲ τειχέων
(restored with probability but not with certainty) κακίει καπνός, is not
known; Boeckh (ii. 2, p. 653) supposed that the words were said by a hero or
god with reference to the walls of Troy; he compared Pind. Ol. 8. 36 (he
could also have cited Pyth. 5. 84), but did not think of the passage in the
Agamemnon.
eüonpos: it is by the smoke (and by that alone) that the conquered city
is even now still discernible, τῶι καπνῶι οὗ ναῦται τὴν Τροίαν σημαίνονται.
819. ἄτης θύελλαι ζῶσι. “There is no need to alter with Hermann to
θνηλαΐ, “sacrifices”; the strong bold imagery of Aeschylus might easily call
the Fire of Troy “the storms of Destruction" ' (Sidgwick). ‘ärns θύελλαι,
because the wind fans and spreads the destroying fire' (Wecklein). G.
Thomson's quotation of p 68 πυρός 7’ ὀλοοῖο θύελλαι is excellent, but he,
nevertheless, accepts the conjecture. ἄτης θύελλαι becomes still more easily
comprehensible if my supposition is right that the whole passage is influenced
by Homer, ® 522 f. ὡς δ᾽ ὅτε καπνὸς ἰὼν εἰς οὐρανὸν εὐρὺν ἱκάνει ἄστεος aio- _
μένοιο, θεῶν δέ € μῆνις ἀνῆκε. It is well known how often ἄτη and μῆνις are
used interchangeably. ζῶσι is in keeping with the picture given in the pre-
ceding line. Troy has been destroyed (814), and now its site is recognizable
only through the smoke. The gales of destruction live on; in all other
respects, and this is the implied antithesis, the city is dead. So the hearer
1 Cf. Lipsius, op. cit. 924 n. 92, who follows Ross.

377
line 819 COMMENTARY
can hardly [41] to understand σὺν τῆι πόλει with συνθνήισκουσα. Hermann’s
objections to this might perhaps occur to a reader, certainly not to an
audience susceptible to the sequence of strong and suggestive images. The
destroyed city dominates the whole passage from 812 to 820.
819 f. συνθνήισκουσα . . . mvods. There is no sentimental lamentation in
this fine sentence but a true note of profound sympathy. For the repetition
of z-sounds cf. on 268. G. Hermann, Opusc. viii. 365, considers the possibility
that these lines may have been in the mind of Callimachus, Hymn. Del. 179 f.
ἴδωσι δὲ πίονα καπνὸν (restored with certainty) γείτονος αἰθομένοιο.
821 f. τούτων . . . τίνειν : harking back to 8ıo f. The framing in this way of
rather large sections of a speech, particularly in its early part, is customary in
Aeschylus, cf. Suppl. 711-24 f., Pers. 179 f.-200, Ag. 1-20 f., Eum. 1-20.
πολύμνηστον : as far as can be seen from the lexicons, this adjective is to be
found outside dactylic poetry only in two passages from this play, 1459 in a
passive sense, and here in an active sense (cf. on 238) ; in an active sense it
occurs also in an epigram of Perses (at the end of the 4th century B.c., cf.
Pomtow in Dittenberger, Syil. 300 n. 3; Wilamowitz, Hellenist. Dichtung, i.
137), Anth. Pal. 6. 274. 3 moAvuväoroıo ... Τεισίδος. For in this last example it
appears to me that in connexion with the dedication to Eileithyia the inter-
pretation of Passow, 'thankful',' and Walz-Guillon (edition of the Ash.
Pal. Paris, 1928), ‘d’une profonde gratitude’, is much more appropriate than
that of Dindorf in the Thesaurus and of L-S who for this passage give the same
meaning ‘wooed by many’ as for πολυμνήστη in the Odyssey. On the other
hand, in Empedocles fr. 3. 3 (Diels-Kranz, sth ed.) πολυμνήστη . . . Μοῦσα
is correctly given by Karsten (in his edition) and L-S as ‘much-remembering,
mindful’, whereas Diels and Kranz translate ‘much celebrated’, Kranz refer-
ring to the Homeric ‘much wooed’. The name /IoAduvaoros is attested for an
earlier period.
822 ff. The MS reading kai mdyas . . . ἐπραξάμεσθα cannot be retained ; on this
point scholars are agreed. Of the proposed conjectures only two deserve
serious consideration. The first was made independently by Hermann and
Paley. They retained mdyas and changed the verb to éppañaueofa. The
objection of Ahrens (567) is not convincing. He does not accept 1375 f.
πημονῆς ἀρκύστατ᾽ ἂν φράξειεν as a parallel, ‘for the expression φράσσειν is
appropriately used of the net, which contains exactly the idea of surrounding
and penning in, but not of πάγαι, snares’. But this is not so; cf., e.g., Hdt. 2.
121 f 1 ráyas . . . ἐργάσασθαι καὶ ταύτας περὶ rà ἀγγήια ἐν τοῖσι τὰ χρήματα ἐνῆν
στῆσαι. The second objection, too, of Ahrens, ‘the Middle would provide
difficulties’, is open to suspicion, in view of the promiscuous use of φράξαι and
φράξασθαι in Homer, Aeschylus, and Thucydides. And so the expression
would be unobjectionable. The thought, too, appears to be suitable, for, as
the next lines speak of the cunning attack with the help of the wooden horse,
it would be a good introduction to say that Troy has been set about with
snares, and that the town and its inhabitants have ‘walked into the trap’.
Taken by itself, then, the conjecture ἐφραξάμεσθα seems very satisfactory, and
yet it is not right. The true reading was restored by Tyrwhitt,? who altered
πάγας to ἁρπαγάς and retained ἐπραξάμεσθα. The expression is excellent;
! Passow, it is true, quotes the passage also under πολυμνήστη, ‘much wooed’.
2 His acumen is highly praised by Wilamowitz, Kl. Schr. i. 351.

378
 COMMENTARY line 826

Ahrens compared Hdt. 5. 94. 2 ὅσοι Ελλήνων συνεπρήξαντο Μενέλεωι τὰς

“Ἑλένης ἁρπαγάς, and from Aeschylus himself we now possess (on an un-
published papyrus from Oxyrhynchus) the sentence ‘EAAddos Aoyayérais . . .
Μενέλεωι τὴν βίαιον ἁρπαγὴν [Ἑλένης συνεκ]πράσσουσι. The conjecture καὶ
ἁρπαγὰς (with regard to the crasis Headlam compared S. Phil. 644 χάρπάσαι)
is recommended also by the comparison with 534 (concerning the parallelism
of the Herald’s speech with that of the king in general see p. 294). What is
decisive, however, is the special train of thought in the present passage.
Inside the causal clause introduced by ἐπείπερ (822), the parallelism of the
two halves (ἐπραξάμεσθα and διημάθυνεν) is brought out strongly by kai...
καί... Therefore, in thought, too, a close correspondence is at least prob-
able. In the second half strong emphasis is laid on γυναικὸς οὕνεκα. If
ἁρπαγάς stands in the first half, then this clause, too, refers to Helen. The
corruption can be explained, perhaps, by supposing that at first kai ἁρπαγάς
was written (for scriptio plena of kai cf. app. crit. on 39 and e.g. on the Berlin
parchment of Euripides’ Cretans, Berl. Klassikertexte v. 2, pp. 73 ff. 1. 40 καὶ
οὐδὲν ἠδικηκότες) and that then someone attempted rather clumsily to make
the passage scan.
The necessity of emending ὑπερκότους to ὑπερκόπους seems to have received
almost general recognition. Hermann’s assertion, which was adopted by
Ahrens, 567, that ὑπερκόπους would not suit ἁρπαγάς, is not borne out by the
usage of the adjective.
824. διημάθυνεν : the verb occurs only in Aeschylus, here and in fr. 244. This
passage is clearly modelled on I 593 πόλιν δέ τε πῦρ ἀμαθύνει.
δάκος : Aeschylus uses the word (5 times) only in the sense of ‘dangerous
animal”, whereas his contemporary Pindar (Pyth. 2. 53) gives it the meaning
that we expect, i.e. ‘bite’. This 'kenning', which Euripides subsequently took
over, is probably pre-Aeschylean. This is indicated not so much by the gloss
in Hesychius Sdxn- θηρία, καὶ δάκος" ἑρπετόν as by the usage in Callimachus
(Hymn. 3. 84) and Nicander, who are likely to follow pre-tragic models.
825. The MS reading ἀσπιδηστρόφος was challenged by Blomfield and Lobeck,
Phrynichus, 683; in order to defend it Klausen adduced S. 47. 575 f., but the
detailed instruction given there for managing the shield does not support the
assumption that ἀσπιδηστρόφος is suitable for denoting a warrior as such.
The second half of the line provides, in the well-known fashion of Aeschylus,
the solution of the ypidos contained in ἕππου νεοσσός.
826. πήδημ᾽ ὀρούσας. A bold extension of this idea, making the wooden
horse itself, not the men, the ἵππου veoooos, take the leap, is to be found in the
Alexander of Ennius (60 f. Ribbeck) nam maximo saltu superabit? gravidus
armalıs equus, qui suo partu ardua perdat Pergama. It is possible that this
was derived from the Ἀλέξανδρος of Euripides. It was to Ennius (cf. Macrob.
Sat. 6. 2. 25) that Virgil was indebted for Aen. 6. 515 f. cum fatalis equus saltu
super ardua venit Pergama et armatum peditem gravis aituht alvo. Conington
! The use of the plural ἁρπαγαί when only one act of robbery is meant is common. It is
hardly adequate to call it an ‘allusive plural' (so A. C. Pearson on E. Hel. 5o and 1322).
It seems that a corresponding use of abstract nouns such as σφαγαί, φόνοι, etc. (cf. on 1329)
goes back to an early stage in the history of the Indo-European languages, cf. P. Maas,
Archiv für lat. Lexikogr. xii, 1902, 493 f. (following B. Delbrück and others).
? Snell, Hermes, Einzelschriften, Heft 5 (1937), 28 n. 1, tries to retain the MS reading
superavit, which I find difficult to understand.

379
line 826 COMMENTARY
(on Aen. 2. 237) established the connexion between Ag. 825 f. and the lines
of Ennius and Virgil.
ἀμφὶ Πλειάδων δύσιν : these words have long provided a tilting-ground for
contests fought in the heavy armour of well-trained chronologists. The
reader who is greedy for details will find them in Ahrens’s extensive survey
(569 ff.) as well as in the commentaries. It was very tempting to find in
ἀμφὶ Πλειάδων δύσιν the designation of the early setting or morning setting
(for the meaning of this term cf. Boll, RE vi. 2423 ff. ; Martin P. Nilsson, Die
Enistehung u. relig. Bedeutung des griech. Kalenders, Lund 1918, p. 7, and the
same writer’s Primitive Time-Reckoning, Lund 1920, pp. 5 ff.), i.e. the first
visible setting before sunrise, of the Pleiades at the beginning of November.
In many similar passages from Hesiod (Erga 615) onward this is doubtless the
meaning (for the subject-matter cf. Ilberg in Roscher’s Mythol. Lexikon iii.
2553 f.). Moreover, Casaubon and Stanley, carried away by their enthusiasm
for Scaliger’s chronological research, wanted to extract evidence, where there
is none to be found, for a date. The result of their discussion (cf. too Stanley’s
note on 40) is that Aeschylus places the capture of Troy in the late autumn,
in contradiction to the bulk of ancient tradition, according to which it took
place in Thargelion (cf. F. Jacoby, Marmor Parium, 149; Usener, Kl. Schr.
iv. 447 ff. ; C. Robert, Heldensage, 1289). Stanley’s conclusion was reproduced
in almost all the older commentaries; it is also to be found in Headlam’s
note on ı. Boeckh, however, maintained an independent judgement. In
1843 he wrote (Corp. Inscr. Graec. ii. 330): ‘Aeschylus hoc ad noctis tempe-
statem, non anni rettulit manifesto’, which, coming from that quarter, Her-
mann dismissed in a high-handed fashion. Boeckh had certainly laid himself
open to the criticism of the diversae scholae auctores by proposing an im-
possible interpretation for audi Πλειάδων δύσιν : ‘nocturnum tempus, ante-
quam ortae erant matutino Vergiliae’. This goes to show that not even
Boeckh had entirely freed himself from chronological speculation. However,
he was substantially right in recognizing that night-time was implied. The
fact that Troy was captured at night was taken by Aeschylus from the
᾽Ιλίου πέρσις (Ἰλιὰς μικρά fr. 11, Bethe, Homer, ii. 2, 2nd ed., p. 176) νὺξ μὲν
ἔην μεσάτη, λαμπρὴ δ᾽ ἐπέτελλε σελήνη, hence E. Hec. 914 (cf. Robert, Helden-
sage, 1253 n. 4, R. Heinze, Virgils epische Technik, 3rd ed. 24). But it is not
a desire to bring the passage into line with the epic tradition which drives us
to Boeckh’s interpretation, nor the circumstance that earlier in this play
(355 ff.) emphasis has been laid upon night as the time of Troy’s capture, but
a consideration of a more general character. Aeschylus 15 not an historian who
relates things past from a distance (let alone a writer for chronologers) but
a dramatic poet who stages things present as they are happening here and
now. H. Keck is quite right in describing (Neue Jahrbücher f. Philol. u.
Pädag. lxxxv, 1862, 520) as insipid the belief that in the king's speech which
follows immediately upon the announcement of the fall of Troy the time of
year could be mentioned.! On the other hand, the fact that the assault took
ı Ahrens (572) starts from the correct statement that Aeschylus has ‘wisely avoided the
stupidity of trying by definite expressions to make the audience believe that the return
journey has really been made at telegraphic speed’, but he draws a wrong conclusion:
‘It is, therefore, not only permissible but very suitable that the indication of the time of
year should impress upon the audience the fact that many a day has passed since the fall
of Troy.’
380
 COMMENTARY line 826

place at night is certainly vital; this detail is not only picturesque

in itself
but also throws into relief the ruthlessness of the punishment ordained by the
gods and the misery of the city which is overwhelmed in its sleep. Night is
suitable, too, as Keck likewise pointed out, to the assault and the bloody meal
of the lion (827 f.). If the audience did not know his «ither from first-hand
experience or from general knowledge of the habits of beasts of prey, they
might remember it from Homer (e.g. P 657 ff.). Keck, who here at any rate
showed sound appreciation of the conditions of a dramatic speech, attempted
nevertheless to compromise with the views of his predecessors. He writes
(p. 521): ‘the poet was under the necessity of naming a constellation which,
on the actual days when the pl»y was performed, set at the hour he wished
to indicate. Otherwise the poetical periphrasis, the purpose of which was to
elucidate, would have failed for the audience in this purpose . . . towards the
e :d of March the Pleiades set in Greece at τὸ o’clock in the evening’. It can-
not be regarded as impossible that Aeschylus, while writing the lines long
before the performance, asked himself: ‘what conspicuous constellation is
there which in Athens at the time of the great Dionysia sets an hour or two
before midnight?’ but such an assumption is not very probable or at all
necessary. It is going too far to try to guess what induced Aeschylus to
choose the Pleiades? for special mention. Perhaps, for him (cf. fr. 312 N.), too,
they were τὸ ἐνδοξότατον" τῶν ἀπλανῶν ἄστρων (Asclepiades of Myrlea apud
Ath. 11. 490 c, where Homer's testimony, 2 486, is adduced); it is sufficient
for our understanding of the passage to remember that Πλειάδων δύσις served,
upon occasion, if not perhaps very often, to depict a more or less advanced
hour of the night, the time when human kind is asleep, concubia nocte. The
connexion of the setting of the Pleiades with the idea of the time about
midnight is clearly shown in the lines quoted on this passage by Wilamowitz,
δέδυκε μὲν à σελάνα καὶ Πληιάδες, μέσαι δὲ νύκτες. The homely conceptions of
this anonymous song are much more suited to the explanation of a detail in
Agamemnon’s picturesque sentences than are the learned computations con-
cerning the Greek calendar. Thomson on Ag. ı could not depreciate the value
of this parallel. When he says ‘the position of a star varies from night to
night’, this is undeniably true. But this statement cannot be used as an
argument, unless we assume, what first needs proving, that the poet was
bent upon designating a particular night according to the season and so
indirectly pointing to a particular season. The wording and context of the

! Verrall's observation to the contrary at the end of his commentary (p. 219) is superficial.
2 In Rhesus 529 f., where the action takes place at a late hour of the night (towards
morning: 551 ff., following K 251 ff.), it is said that the Pleiades (probably shortly after
their rising) are in the sky. This is the likely interpretation of Parmeniscus in the scholia
(αἰθέριαι predicate, so too Paley: ‘the seven Pleiads are now in the heaven’). Whether the
poet thought of a definite time of year it is impossible to say. There is no parallel here to
ἀμφὶ Πλειάδων δύσιν. A somewhat later hour is depicted at the beginning of the chorus in
Eur. Phaethon (v. Arnim, Suppl. Eur., p. 70) ἤδη μὲν ἀρτιφανὴς "Euls ἱππεύει] κατὰ γᾶν’ ὑπὲρ
δ᾽ ἐμᾶς κεφαλᾶς Πλειὰ[ς . . . ‘As the Pleiad, which represents the stars, stands high in the
sky at the approach of Eos, there must have been a verb there indicating their fading or
vanishing in the heavens. It will be impossible to restore the verb with certainty’ (Wilamo-
witz, in his edition of the fragment, Berliner Klassikertexte, v. 2, p. 82). Here there is no
question of δύσις. .
3 On the important part which the Pleiades play because they are easily recognized
as a separate group cf. Martin P. Nilsson, Primitive Time-Reckoning (Lund 1920), 130.
381
line 826 COMMENTARY
whole passage seem to lead to no further conclusion than that an indication
is given here of an advanced hour of the night."
827. ὑπερθορών merely continues the idea of πήδημ᾽ ὀρούσας, but with λέων
a fresh element is added to the picture. The attacker, described just before as
ἀσπιδηφόρος λεώς, becomes now, in the development of ὑπερθορεῖν, a lion; cf.
on 387 ff. It is probable that this thought is mainly suggested by the idea of
the leap over the bastioned walls (cf., e.g., E 138 Aéovra . . . αὐλῆς ὑπεράλμενον) ;
moreover, the picture of the raging lion who licks the blood (cf. above,
730 ff.) presents itself as a natural feature. As the images, one after another,
boldly thrust their way forward, the poet does not seem to be hampered by
any meticulous concern about the compatibility of their details. Wilamowitz's
translation (‘da duckte sich zum Sprunge der grausame Leu, die Brut des
Rosses, Argos’ schildbewehrtes Volk, schwang hoch sich über Troias Zinnen,
weidete satt seine Gier an Priams königlichem Blut’) not only destroys the
order, but by throwing into confusion the different elements brings into the
passage a τέρας, which, however effective in its baroque contour, is certainly
foreign to the original. ἵππου veocoós fulfils its purpose, which is to tell, by
way of ypidos, where the armed men came from; the lion is not yet in the
picture, nor is there, at this point, any allusion to him. The scholiast’s prim
dilution ἤγουν ὥσπερ λέων ὠμοφάγος (to understand ὥσπερ is in itself admis-
sible, cf. on 393 ff.) would not be worth mentioning if it had not been per-
petuated up to most recent times in commentaries and translations (though
Hartung and Conington, for example, were not deceived). The subject of
ὑπερθορὼν.... ἔλειξεν is, of course, the lion, and not a noun to be supplied from
the preceding lines.
828. ἅδην: I do not know why the word is robbed in so many texts of the rough
breathing, which the grammarians attest (Lehrs, De Aristarchi stud. Hom.,
3rd ed., 327) and etymology demands (cf. W. Schulze, Quaest. ep. 452 n. 2).
τυραννικοῦ : the earliest occurrence of this adjective is in the Oresteia (again
Cho. 479). The foreign word τύραννος (cf. Schwyzer, Griech. Gramm. i. 491),
apparently borrowed from the East, occurs in Semonides of Amorgos,
Alcaeus, and Pindar {τυραννίς is found as early as Archilochus). In Attic it is
used from Solon onwards; in Tragedy it first occurs in late plays of Aeschylus,
the Prometheus and the Oresteia. As the history of the word shows, it was a
more successful rival of πάλμυς (cf. also the use of βαλλήν). Hippias of Elis
(fr. 9 Diels) had a correct notion of the immigration of the word: ὀψέ ποτε
τοῦδε τοῦ ὀνόματος εἰς τοὺς Ἕλληνας διαδοθέντος, κατὰ τοὺς Ἀρχιλόχου χρόνους.
τύραννος, as is well known, does not necessarily imply ‘the slightest undertone
of hatred’ (Wilamowitz on E. Her. 29); this undertone is certainly alien to
the word in its original use and is absent in many passages of Tragedy. On
the other hand, at 1355 and 1365 τυραννίς is used with a hostile tone.
829 reverts again (cf. 821) to the beginning of the speech and rounds off its
first part. φροίμιον is here the equivalent of πρῶτον in 810, just as Eum. 20
τούτους ἐν εὐχαῖς φροιμιάζομαι θεούς harks back to 1 πρῶτον μὲν εὐχῆι τῆιδε
πρεσβεύω κτλ., although in that passage πρῶτον also distinguishes the prayer
to Gaia from the rest. For the form of this framing cf. on 1196.
1 I have formulated without any compromise the conclusions which seem to me inevit-
able. Beazley, however, after carefully examining my arguments several times, remains
firm in his opinion that here, as elsewhere, Πλειάδων δύσις must indicate the time of year.
382
 COMMENTARY line 830
When in Aeschylus new characters make their entry, they sometimes begin
their speeches with lengthy sections which take little if any account of the
persons already present on the stage. This looks like an anticipation of the
“Auftrittsmonolog’ (monologue delivered upon the first entry) in later drama
(cf. F. Leo, Monolog im Drama, 8; Wilamowitz, Interpr. 169, who gives illus-
trations). Of the scenes which are comparable to this, the analogy is most
noteworthy in the opening speech of Agamemnon’s herald (cf. on 522), since
it is constructed altogether on lines parallel to this speech of the king. Here
the words ἐξέτεινα (cf. on 916) φροίμιον τόδε apologize to some extent for the
length of this introduction! or prelude.
830. Now, at last, he answers the sentences addressed to him by the Chorus.
τὰ δ᾽ es τὸ σὸν φρόνημα is compared by van Heusde with 5. Oed. R. 706 and
E. Ih. T. 691, but in both passages as in E. Her. 171 (where Wilamowitz has
explained the use of eis) the apparently similar expressions are fitted much
more neatly into the construction of the sentence and do not form separate
κῶλα by themselves. In the present passage we have a completely prosaic
‘quod autem attinet ad . . .', strongly marked off (F rightly puts a comma after
φρόνημα). The parallel adduced by Blaydes, E. El. 945 à δ᾽ és γυναῖκας, is
suitable ; Headlam adds Eubulus fr. 7.6 f., ii. 166 Kock.” These two examples,
from the late fifth century and from the fourth, show the dry formula of
transition in a well-ordered speech. It is exactly the same in Ag. 83o. It
does not often happen that we can establish with certainty, as in this case,
that Aeschylus has borrowed from the style of contemporary Attic speech.
Our material for comparison comes, of course, from a slightly later period,
but the evidence provided by the speeches in Thucydides puts it beyond all
doubt that Aeschylus owes to the law-court or the Assembly this formula for
transition to a new section. Cf. Thuc. 2. 40. 4 (funeral speech of Pericles) «ai
τὰ ἐς ἀρετὴν ἐνηντιώμεθα τοῖς πολλοῖς, 3. 62. 6 kal rà μὲν és τὸν μηδιομὸν τοσαῦτα
ἀπολογούμεθα, similarly 3. 64. 5. In a speech, too, in Xen. Cyruf. 5. 4. 25 eis
δὲ τὴν τοῦ καρποῦ κομιδήν... 6 ἐπικρατῶν οἶμαι καρπώσεται. The extremely
prosaic character of this formula is again clearly indicated by its use in ἃ
passage which closely follows the language of actual laws, Plato, Laws 6.
174 b eis μὲν οὖν χρήματα ὁ μὴ ᾿θέλων γαμεῖν ταῦτα ζημιούσθω.
φρόνημα refers, in general, to the loyal thoughts professed by the old men

1 The fact that φροίμιον has this meaning here is in no way changed by the dialectical
niceties of R. Böhme, Das Prooimion (1937), 67 f. and 83 f. n. 184.
2 This passage (ἃ δ᾽ eis τ᾽ ἐδωδὴν πρῶτα καὶ ῥώμης ἀκμὴν καὶ πρὸς ὑγίειαν, πάντα ταῦτ᾽
ἐδαινύμην) invalidates the conjecture of Musgrave, recommended by Denniston, τὰ δ᾽ ἐς
γυναῖκας for E. El. 945. The same formula of transition seems to be used in one of the long
fragments of Euripides’ Μελανίππη ἡ δεσμῶτις (Berliner Klassikertexte, v. 2, p. 125 {.; Hunt,
Fragmenta. Tragica Papyracea, no. vi; v. Arnim, Suppl. Eurip. 33; D. L. Page, Greek Lit.
Papyri, i. 112), 1. 12 (8 Page), for I am not convinced that Hunt (followed by v. Arnim and
Page, who does not even mention the variant) was right in choosing the reading of Satyros
τὰ δ᾽ ἐν θεοῖς αὖ instead of that of the Hellenistic anthology on the Berlin papyrus, ἃ δ᾽ eis
θεοὺς ad. Accepting this reading, we have in the sentence ἃ δ᾽ eis θεοὺς αὖ" πρῶτα γὰρ κρίνω
τάδε: μέρος μέγιστον ἔχομεν an arrangement exactly parallel to that in E. El. 945 f. ἃ 8’ ἐς
γυναῖκας" παρθένωι γὰρ οὐ καλὸν λέγειν: σιωπῶ. I have of course considered the possibility
that the reading ἃ δ᾽ eis θεοὺς αὖ might be merely an interpolation from 1. 18 ἃ δ᾽ eis τε
Μοίρας, but the awkwardness of τὰ δ᾽ ἐν θεοῖς αὖ seems to tell against this alternative.
At the end of l. 7 Hunt and v. Arnim (followed by Page) have preferred the ἔχει of Satyros
to the φέρει of the Berlin papyrus; against this see Wilamowitz, KI. Schr. 1. 444 τι. 2.
383
line 830 COMMENTARY
(782-809) and, in particular, to their assurance (806) that they are truly
speaking as eüppoves ; in the same connexion 834 Svcdpwv, 840 πρευμενεῖς.
μέμνημαι κλυών (for the accent required here for the sake of clarity cf. on
680) is not μέμνημαι cum participio’ as e.g. Xen. Cyrup. x. 6. 6 μέμνημαι...
τοιαῦτα ἀκούσας σου, for here κλυών does not express the content of the
μεμνῆσθαι (Schütz, among others, is wrong: ‘me audire memini’, so is Paley:
‘I remember hearing them’), but κλυών is added like any other participle with
a causal or modal function: ‘I have heard and so remember them’. The
object of μέμνημαι can be readily supplied from ἐς τὸ σὸν φρόνημα. The clauses
from μέμνημαι to ἔχεις are bound together as a τρίκωλον, and it makes no
difference that the second and third are more closely parallel with each other.
It is particularly in expressions of assent that such an almost tautological
emphasis is sought, e.g. Ar. Birds 851 (it cannot be made out in detail to what
extent the passage depends on Soph. fr. 449 N. = 489 P.) ὁμορροθῶ, συνθέλω,
συμπαραινέσας ἔχω, Cf. also S. Ant. 536 f. ὁμορροθῶϊ καὶ ξυμμετίσχω Kai φέρω τῆς
αἰτίας, the proverb (Diogenian 8. 50) οὐ σύμφημ᾽ οὐδὲ συναινῶ (for the metrical
form cf. on 1527). In the fetial formula in Livy τ. 32. 13 we find quod ...
senatus popult Romani Quiritium censuit consensit conscivit (cf. also 1. 32. 12).
832. συγγενές : in meaning and construction exactly corresponding to 884 f.
σύγγονον βροτοῖσι κτλ.
833. φίλον τὸν εὐτυχοῦντα : Paley successfully rejected the attempt of some
commentators to take φίλον predicatively, 1.6. to construe σέβειν τὸν εὐτυ-
xodvra (ws) φίλον. Cf. Gildersleeve, Syntax of Classical Greek, 280 (8 609): ‘In
poetry ... the difference between ἀνὴρ 6 ἀγαθός and ὁ ἀνὴρ 6 ἀγαθός cannot be
insisted on.’
ἄνευ φθόνων was defended by Hermann, who quoted Plato, Laws 7. 8ore.
He has been followed among others by Verrall and Murray. But Stob. 38. 28
bears witness to the singular, and it is a probable guess (Gesner) that ψόγου
in Stobaeus is simply a slip for φθόνου, as the subject of the chapter is Περὶ
φθόνου and all the adjoining quotations from the poets contain φθόνος or a
derivative of it.
834. Whether Casaubon, Grotius, Pearson, Hermann (detailed discussion),
and others are right in changing καρδίαν to the dative it is difficult to decide.
The argument for the dative is that elsewhere in Aeschylus πρόσημαι always
has this construction (four times, including the two instances in Ag.), cf.
also Sept. 696 ὄμμασιν προσιζάνει, For the retention of the accusative, besides
general considerations (cf. Kühner-Gerth, i. 313 f., n. 13), it may be adduced
that προσίζω A. Suppl. 189 and E. Hec. 935 governs the accusative (in Eur.
fr. 910. 9—in a metaphorical sense—it governs the dative).
Aeschylus says of a passion ‘seated close to or by (not in) the heart’; cf. on 179.
835. πεπᾶσθαι is regarded by Wilamowitz, on E. Her. 1426, as a word which
the Athenians had borrowed at an early date from their neighbours the
Megarians and Boeotians, whereas G. H. Mahlow, Neue Wege durch d. griech.
Sprache (1926), 138, considers it good Attic.
836 f. The γνώμη 834 f. δύσφρων . . . νόσον is followed, without a connective
particle, by an explanation composed of the two clauses τοῖς re . . . πήμασιν...
καὶ τὸν θυραῖον ὄλβον, cf. on 790.
' This striking emendation by Nauck, which Bruhn adopted, received the express
approval of Wilamowitz (in a class, ‘Proseminar’, 1908).

384
 COMMENTARY lines 838 ff.
836. In FTr we read αὐτοῦ (Wecklein’s app. crit. is misleading). αὐτοσαυτοῦ,
αὐτοσαυτόν etc. serves as an intensified form of the reflexive pronoun. Its
second element is in MSS of the tragedians sometimes written with what
commonly, though not very sensibly, is called 'spiritus lenis’ ; cf. Elmsley on
E. Heraclid. 144; Kühner-Blass, i. 600 n. 5; Kühner-Gerth, i. 564 f.; Wila-
mowitz on A. Sept. 194. Wilamowitz, however, in spite of his own emphatic
warning, printed without comment Cho. 221 αὐτὸς καθ᾽ αὑτοῦ in his edition of
Aeschylus; whereas earlier in his special edition of the Choephoroe he had
retained the κατ᾽ αὐτοῦ of the Mediceus. Pers. 415 we find in M and in a number
of MSS αὐτοὶ δ᾽ ὑπ᾽ αὐτῶν; on Prom. 762 πρὸς αὐτὸς αὐτοῦ Vitelli-Wecklein
make the cautious comment: αὐτοῦ M ut videtur’ (a glance at the facsimile
shows the ambiguity of the stroke written there). It is possible that in these
and other passages the smooth breathing follows the rules of the best tradition
among the grammarians. It is very probable that the linguistically unjusti-
fied transformation of αὐτοσαυτοῦ, still preserved in Doric, into αὐτὸς αὑτοῦ
(= ἑαυτοῦ) did not take place as early as the fifth century. This was the
opinion of Bernhardy, quoted with approval by H. L. Ahrens, De Graec.
ling. dial. ii. 274, cf. also, besides Blass, loc. cit., Wackernagel, Dehnungs-
gesetz 32 f., Syntax, ii. 89; Wilamowitz on E. Her. 961 (also concerning the
type of word-order τοῖς αὐτὸς αὐτοῦ πήμασιν, for which cf. Wilhelm Schulze,
ΚΙ. Schr. 631) ; Schwyzer in K. Deichgräber’s book Hippokrates über Entstehung
... des menschl. Körbers, 74 n. 5.
On 834-7 Beazley observes: ‘these lines seem to me intrusive. Why on
earth should Agamemnon be interested in the psychology of the envious man,
or care whether he suffers or not?’ I cannot answer this nor can I deny that
the viewpoint of 834-7 is entirely different from that of 832 f. All that the
two γνῶμαι have in common is that they are both περὶ φθόνου. It looks as
though 832 f. had induced a reader to jot down in the margin the other
γνώμη, ‘ob adfinitatem quandam sententiarum’ as Valckenaer says with
regard to a similar case (E. Phoen. 558), and that then the quotation found its
way into the text. Its original place was possibly in another play of Aeschylus.
Cf. on 570-2.
838-40. Apart from impossible artifices, there have long been two distinct
ways of interpreting this passage. Pauw, Blomfield, Hermann, Conington,
Paley, Schneidewin, Sidgwick, Verrall, Wilamowitz, etc., regard ὁμιλίας
κάτοπτρον as parallel to εἴδωλον σκιᾶς, and see in both phrases a description of
those who δοκοῦσιν πρευμενεῖς εἶναι. Those commentators differ about certain
details. Some, e.g. Paley, make the words from ὁμιλίας κάτοπτρον to πρευ-
μενεῖς ἐμοί depend as predicate on λέγοιμ᾽ av, i.e. take εὖ yap ἐξεπίσταμαι as
parenthetical. Others regard εἰδὼς λέγοιμ᾽ av as a Separate clause and con-
nect ἐξεπίσταμαι with du. κάτ., εἴδ. σκιᾶς (ὄντας). But they all see in κάτοπτρον
the symbol of the unreal, of mere appearance: ‘non ὁμιλίαν, sed ὁμιλίας
κάτοπτρον᾽ (Pauw), ‘imaginem specularem amicitiae’ (Hermann), ‘an un-
substantial picture, as reflected by the mirror’ (Schneidewin). The word
κάτοπτρον, and so the whole passage, is taken in quite a different sense by
Triclinius, Casaubon (‘scio uti consuetudine, ceu speculo, ad explorandos
animos hominum’, etc.), Heath, Schiitz (‘expertus enim dico, quod in eorum
consuetudinem tamquam in speculum inspexi, eos qui mihi maxime benevoli
videantur, nil nisi umbrae quandam imaginem esse’), Hartung, Nagelsbach,
4872.2 cc 385
lines 838 fi. COMMENTARY

van Heusde, Wecklein (mistaken in the detail), Headlam, etc. This latter
group take κάτοπτρον as something from which we receive a true likeness (‘the
glass in which the associate’s true character is shown’ Headlam). As is well
known, this usage of κάτοπτρον in Aeschylus is attested by fr. 393 κάτοπτρον
εἴδους χαλκός ἐστ᾽, οἶνος δὲ νοῦ (compared by Abresch) ; cf. further E. Hipp.
428 ff. κακοὺς δὲ θνητῶν ἐξέφην᾽, ὅταν τύχηι, προθεὶς κάτοπτρον ὥστε παρθένωι
νέαι χρόνος (van Heusde). Alcidamas in Arist. Rhet. 3. 3 p. 1406 b 12 calls
τὴν ᾿Οδύσσειαν καλὸν ἀνθρωπίνου βίου kdromrpov,' a metaphor of which Aristotle
disapproves. The same Alcidamas says Περὶ σοφιστῶν 32 eis δὲ τὰ γεγραμμένα
κατιδόντας ὥσπερ ἐν κατόπτρωι θεωρῆσαι τὰς τῆς ψυχῆς ἐπιδόσεις ῥάιδιόν ἐστιν.
Further examples of this usage could be quoted. On the other hand, no
evidence can be produced in favour of the interpretation which takes
κάτοπτρον as imago specularis (Hermann), ‘that ghost of friendship’ (Sidg-
wick), ‘eitles Spiegelbild’ (Wilamowitz) and the like. κάτοπτρον always means
‘mirror’, never ‘reflection’. That settles the question. It is not permissible to
get out of a fix as Conington does. He was aware of the difficulty (cf. his
note), and though he takes óp. κάτ. together with εἴδωλον σκιᾶς as predicate
of the following line, he translates ‘a mere glass of friendship, a shade’s
shade’; what he means, however, is not ‘glass’ but the image produced by
the mirror. Cf., too, the naive paraphrase of Paley’s: ‘the mirror (the unreal
semblance) of friendship’. Aeschylus’ thought is clear. ‘I am well acquainted
with the mirror, the intercourse of men.’ The ὁμιλία, 1.6. the way in which
people approach him and others, associate with him, and in all their relations
remain themselves or not, all this and the like is the instrument, the mirror
in which the king sees the image of the true disposition of his fellow beings.
It is not far-fetched to view ὁμιλία as a tool, as a touchstone for testing moral
conduct ; Blomfield made a very happy comparison with E. El. 384 f. τῆι δ᾽
ὁμιλίαι βροτοὺς κρινεῖτε καὶ τοῖς ἤθεσιν τοὺς εὐγενεῖς. The meaning of κάτοπτρον
which has thus been established makes clear the phrasing also. The clause
εὖ γάρ to κάτοπτρον is put in parenthetically (for γάρ in a parenthesis which
interrupts the sentence cf. on 800), while the words λέγοιμ᾽ ἂν εἴδωλον σκιᾶς...
πρευμενεῖς ἐμοί go together. In εὖ γὰρ ἐξεπίσταμαι κτλ. we have the reason (or
the more detailed elaboration) of εἰδώς placed emphatically at the beginning.
838. λέγοιμ᾽ ἄν: the restrained form of expression (cf. on 1049) reflects the
urbanity of Attic society (cf. in general Kühner-Gerth, i. 233). It is character-
istic that this form is chosen just when the speaker is quite sure, as here
! When I pointed out (Plautin. im Plaut. 388) how well these words illustrate a certain
kind of outlook at the end of the sth century and in the 4th, I associated with them Cicero’s
definition (ap. Donat. de com. 5. 1) which has made for us a commonplace out of what was
once an extremely bold image: comoediam esse imitationem vitae, speculum consuetudinis,
imaginem veritatis. (I should also have recalled Cic. p. S. Rosc. Am. 47, where Cicero, with
reference to the Ὑποβολιμαῖος of Menander translated by Caecilius, and to New Comedy in
general, says: haec conficta arbitror esse a poelis ut efficlos nostros mores in alienis personis
expressamque imaginem vitae cotidianae videremus, and I should have quoted most parti-
cularly Menander himself: Ter. Ad. 414 ff. denique inspicere, tamquam in speculum, in vitas
omnium iubeo alque ex aliis sumere exemplum sibi.) 1 drew the necessary conclusion that
this definition must be derived from the language of Greek aesthetics (or ‘poetics’). If,
however, speculum consuetudinis looks exactly like a translation of the Aeschylean ὁμιλέας
κάτοπτρον, the resemblance is no more than apparent: consuetudo could in itself be ἃ
translation of ὁμιλέα, but its meaning here is much more comprehensive, and moreover the
function of the genitive is quite different (in Aeschylus epexegetic, in Cicero objective).
386
 COMMENTARY line 843
(εἰδώς), cf., e.g., Sept. 375 λέγοιμ᾽ ἂν εἰδὼς εὖ τὰ τῶν ἐναντίων, Pers. 266 f. καὶ
μὴν παρών γε κοὐ λόγους ἄλλων κλυών, Πέρσαι, φράσαιμ᾽ ἂν κτλ., S. Oed. R. 95
λέγοιμ᾽ ἂν of? ἤκουσα τοῦ θεοῦ πάρα. Cf. also Ag. 739, 896, S. Oed. R. 282. A
polite phrase of a similar kind is sometimes found at the beginning of a
formal speech, addressed to a large audience, i.e. λέγοιμ᾽ ἂν ἤδη (without
object) : Ar. Knights 40, Lys. 97 (picked up in 119 by λέγοιμ᾽ ἄν) ; this evidently
reflects a forensic usage at Athens. [Cf. the Addenda.)
839. εἴδωλον σκιᾶς denotes, as it were, the superlative of the unreal. Parallel
passages have often been quoted ; but most of them, while akin in expression,
refer to the unreality of human existence as a whole (the meaning in Soph.
fr. 598 N. = 659 P. 1. 6 is entirely different).
840. δοκοῦντας εἶναι... πρευμενεῖς takes up the theme of 788 and 797.
841. That Odysseus did not voluntarily join the war against Troy, but was
only brought to do so by the cunning of Palamedes,' was related in the
Cypria, cf. Robert, Heldensage, 1091, Bethe, Homer, ii. 2, p. 234 f. Aeschylus
gives the story a completely new turn here or, more correctly, wins a new
point from it, which is important for the characterization of his Agamemnon.
In the epic the purpose of the pretty device of Odysseus’ feigned madness is
to display the artifices of the πολύμητις. But here the king, unobtrusively
and, as it were, by the way, gives an example of that ripe experience in
judging human nature as in political affairs, of which he has just spoken.
All the great lords who at the beginning in their noble enthusiasm for war
appeared ready for any undertaking proved completely unreliable when
things went wrong, when year after year passed and all that remained of the
inspiring adventure was privation. One man, however, endured faithfully to
the end, the one man who had been against the war from the very beginning
and had tried everything to escape it. Besides this, there appears to exist a
parallelism between the change in the attitude of Odysseus to Agamemnon
and the change in the attitude of the Chorus (799 ff.) : first a firm rejection of
the king’s actions, later genuine loyalty. In Sophocles also (Aj. 1331) Aga-
memnon names Odysseus as his greatest friend in the army. Cf. 4 360 f.
οὐχ ἑκών : elsewhere (E. Hel. 396) Menelaus points out that the Greeks went
to war against Troy ἑκόντες.
842. For the heavy task of the σειραφόρος in the chariot-races cf. on 1641.
Apparently here, as in 1640, the notion of ζεῦξαι is the starting-point for the
image of the ‘trace-horse’ ; cf. also the similar continuation fr. 381 N. ὅπου yap
ἰσχὺς συζυγοῦσι καὶ δίκη, ποία ξυνωρὶς τῆσδε καρτερωτέρα;
843. It is natural enough that Agamemnon should make such an observation:
the uncertain fate of a loyal comrade-in-arms is bound to cause anxiety to
one who has just landed safely at home. But the poet appears to have a
further purpose. The resemblance of the words about Menelaus and his
company 671 ff. could not fail to strike the audience. This leads us to assume
that in the Proteus, as was natural from the connexion of the two vooro: in the
Odyssey, the return of Odysseus also was mentioned. If this was so, both
hints would prepare the ground for the Satyr-play (cf. K. O. Müller, Aeschylos
Eumeniden, 199 ; Schneidewin on Ag. 843; Wilamowitz, Griech. Trag. ii. 302 f.).
1 It is not known whether this story played a part in the Palamedes of Aeschylus (cf.
A. C. Pearson, The Fragments of Sophocles, ii. 116 n. 2); it may have been briefly mentioned
there.
387
line 843 COMMENTARY
On etre . . . εἴτε καί see Denniston, Particles, 305. It is unnecessary to
supply, with Paley and Verrall, a second λέγω out of the clause εἴτ᾽ οὖν.
λέγω and make this the apodosis of the conditional sentence.
844. ra δ᾽ ἄλλα πρὸς πόλιν τε καὶ θεούς. This may be regarded, with Her-
mann, as a ‘contaminated’ form of expression, combining τὰ δ᾽ ἄλλα and ra
πρὸς πόλιν τε καὶ θεούς. Similarly in Cho. 512 the speaker, after reference to
the preceding speech, turns with ra δ᾽ ἄλλα to what remains to be done;
likewise Cho. 552, after the concluding wish, τἄλλα δ᾽ ἐξηγοῦ φίλοις κτλ.
For τὰ πρός (particularly in connexion with θεούς and πόλιν) cf. Headlam,
ad loc.
θεούς : the ἱερά have their fixed position on the agenda of any Ekklesia, in
Athens as elsewhere.
845. ἀγῶνας: in the meaning familiar from Homer and Hesiod (as also Pind.
P. 10. 30), cf. on 513. The passage quoted there, Sepi. 219 f., shows that in Ag.
845 πανήγυρις is more or less synonymous with ἀγών. This use of ἀγών was
presumably felt as a Homerism by Aeschylus and his audience, although at
that time a similar use was still alive outside the Attic dialect, cf. M. N. Tod,
Greek Hist. Inscr. p. 30, under (c), L-S Addenda, p. 2044.
Agamemnon is just as much at pains to emphasize the constitutional
checks to his authority as is Pelasgus, who says (Suppl. 368 f.): ἐγὼ δ᾽ ἂν où
κραΐνοιμ᾽ ὑπόσχεσιν πάρος, ἀστοῖς de πᾶσι τῶνδε κοινώσας πέρι. The use of the
word κοινόν there (xowwoas) and here (κοινοὺς ἀγῶνας) is significant. ‘Aeschylus
introduces into Argos a constitutional king ; there was a democracy there in
his own time’ (Wilamowitz, Staat u. Gesellschaft der Griechen, znd ed. 58).
848. ὅτωι δὲ καὶ det: one of the many instances where, after δέ, καί 'approxi-
mates in sense to ad’, cf. Denniston, Particles, 305.
849. For ἤτοι... cf. on 662.
κέαντες ἢ τεμόντες : cf. Heraclit. fr. 58 of γοῦν ἰατροὶ... τέμνοντες, καίοντες
κτλ. Parallel passages from the Corpus Hippocraticum are quoted by J.
Dumortier, Le Vocabulaire medical d'Eschyle (1935), 47f., cf. also PI.
Gorg. 52re, Polit. 293b, Xen. Mem. 1. 2. 54; ‘a popular idea’ (O. Gigon,
Unters. z. Heraklit [1935], 26).
εὐφρόνως : ‘judiciously’, cf. on 806. Correctly understood by e.g. Nägels-
bach, Headlam: ‘by sage use of the knife’ ; mistakenly Paley: ‘with friendly
hand’, Verrall : ‘do our kind endeavour’, G. Thomson: ‘kindly knife’. eddpdvws
is to be taken ἀπὸ κοινοῦ.
850. Porson’s restoration is magnificently simple and, in the main, right (it
would have been a good thing if Hermann had kept silent on this point) ;
there remains only the small question whether ἀποτρέψαι would not be more
suitable. Cf. Hdt. 1. 207. 1 εἶπον... τὸ dv ὁρῶ σφάλμα ἐὸν οἴκωι τῶι σῶι, κατὰ
δύναμιν ἀποτρέψειν, Pl. Gorg. τοῦ Ὁ ταύτην (scil. τὴν βοήθειαν) ἥτις ἀποτρέψει
τὴν μεγίστην ἡμῶν βλάβην, A. Pers. 216 ff. θεοὺς δὲ προστροπαῖς ἱκνουμένη,
εἴ τι φλαῦρον εἶδες, αἰτοῦ τῶνδ᾽ ἀποτροπὴν τελεῖν, τὰ δ᾽ ἀγάθ᾽ ἐκτελῇ γενέσθαι
σοί τε καὶ τέκνοις σέθεν (the division into two complementary sections is the
same as in Ag. 846 ff., but there man undertakes for himself what in the
Persae is to be begged from the gods). The sacrifice of the o would be of little
consequence;! in X 197 nearly all the MSS have ἀποστρέψασκε instead of
1 On the danger of ‘anxious adherence to the ductus litterarum’ see Housman, Manilius,
v, p. xxxiv f.
388
 COMMENTARY line 854
ἀποτρέψασκε, in Pl. Gorg. 509b ἀποστρέψει is wrongly written in F. Else-
where, too, in the MSS the two verbs, in forms of the future and aorist, are
often confused. I have not been able to find an incontrovertible parallel
from classical times for ἀποστρέφειν in the sense required here; in Ar. Clouds
776 editors now read, with Meineke, ἀποστρέψαι {-ψαις codd.) av, in Antiphon
6. 15, with Dobree, ἀποτρέψαι (Tetral. y y 7 has a different meaning). If,
however, Aeschylus did write ἀποστρέψαι, it probably implies a rather strong
metaphor: the disease in the house shall be driven back luc unde malum
pedem attulit.
The conception of a νόσος of a πόλις is widespread in the fifth century;
in Athens it is found as early as Solon, cf. Wilamowitz on E. Her. 542,
W. Jaeger, Paideia (Engl. transl), i. 139, iii. 5; Momigliano, C.Q. xxxvi,
1942, 118 n. 1.
851. μέλαθρα καὶ δόμους. On Ar. Birds 1247 f. μέλαθρα μὲν αὐτοῦ καὶ δόμους
Ἀμφίονος καταιθαλώσω πυρφόροισιν αἰετοῖς the scholiast notes ἐκ Νιόβης
Αἰσχύλου (fr. 160 N.). The editors of the fragments of Aeschylus, from Stanley
on, have taken the view that the quotation does not start until xai δόμους.
Certainty, of course, is impossible here, but it seems to me much more
probable that the passage in the Niobe went o— μέλαθρα καὶ δόμους Aupiovos!
(the words occupying the same position in the line as in Ag. 851), and that
Aristophanes has only inserted μέν (on account of πέμψω δέ in 1249) and
αὐτοῦ (of Zeus, demanded by the context of Peisetairos’ speech, whereby
Ἀμφίονος becomes humorous nonsense). It was just things like μέλαθρα καὶ
δόμους that sounded particularly ‘Aeschylean’ to an Athenian audience: δὲς
ταὐτὸν ἡμῖν εἶπεν ὃ σοφὸς Αἰσχύλος. The ‘tautology’ is less in evidence if the
words are distributed as in E. Her. 864 καὶ καταρρήξω μέλαθρα καὶ δόμους
ἐπεμβαλῶ, but there, too, 'μέλαθρα and δόμοι are not clearly distinguishable
. . . the parallelism of the two clauses has an exalted effect as in Hebrew
poetry’ (Wilamowitz, ad loc.).
ἐφεστίους : here and Sept. 73 (likewise δόμους ἐφεστίους) Aeschylus uses the
adjective in the sense of ‘(provided) with a hearth’,? in which the preposition
has a similar function to that in ἐπέκοτος, ἐπίπονος, and the like. This looks
like an arbitrary change of meaning. For elsewhere, in Aeschylus as well as in
other writers, the word either means simply ‘by the hearth’ (Ag. 1310, Eum.
169) or is used, after the model of Homer ( 248 ἀλλ᾽ ἐμὲ τὸν δύστηνον ἐφέστιον
ἤγαγε δαίμων, ψ 55 ἦλθε μὲν αὐτὸς ζωὸς ἐφέστιος, cf. Ed. Schwartz, Die Odyssee
22), predicatively to verbs of motion or staying, e.g. Suppl. 503 ναύτην
ἄγοντας τόνδ᾽ ἐφέστιον θεῶν (ἐφέστιον ἄγοντας go together, cf. sor ἡγεῖσθε
βωμοὺς ἀστικούς, θεῶν ἕδρας), 365 οὔτοι κάθησθε δωμάτων ἐφέστιοι ἐμῶν, similarly
S. Oed. R. 32 ἐζόμεσθ᾽ ἐφέστιοι, Trach. 262 ἐλθόντ᾽ ἐς δόμους ἐφέστιον (cf. Rhes.
201), E. Med. 713 δέξαι δὲ χώρᾶι καὶ δόμοις ἐφέστιον, Ion 654 f. ἄγων σ᾽ ἐφέστιον
δείπνοισι τέρψω, etc.
852. θεοῖσι δεξιώσομαι: the use of the dative with δεξιοῦσθαι to indicate the
person greeted seems to be without parallel.
854. ἐμπέδως : watered down by Cobet and Meineke (Philol. xix, 1863, 200)
τ Pickard-Cambridge, too, in Greek Poetry and Life (Oxford 1936), 118, includes μέλαθρα
in the quotation from the Niobe.
2 Paley's explanation (note on his translation): ‘lit. “rooms (halls) by the altar" ’ is
improbable.
389
line 854 COMMENTARY
to ἔμπεδος, although the latter himself quotes (and alters!) Suppl. 945 ws
μένειν ἀραρότως.
855. πρέσβος occurs only in the three similar forms of address here, in 1393,
and in Pers. 623. As to Pers. 623 βασίλεια γύναι, mpéoBos Πέρσαις, the correct
rendering is given by the scholiast with ὡς θεὸς γὰρ παρὰ Πέρσαις τιμᾶται and
the gloss τιμία παρὰ Πέρσαις, cf., ἰοο (on the Homeric πρέσβα θεά and the like),
Hesych. πρέσβα" ἔντιμος, πρεσβυτάτη, σεμνή (cf. Etym. M. 687. 3 ff.). But in
Ag. 855 and 1393 the sense seems to be more complex, so as to include the
notion of old age (therefore in translating I have followed Paley). The choice
of the abstract word, which is at the same time a rare one, gives the speech
more solemnity than the simple πρέσβεις (Pers. 840) would have done. For
the close association of the meanings of ‘venerable’ and ‘old’ (or vice versa),
not only in the case of mpeoß-, but elsewhere, cf. on 722.
856. οὐκ αἰσχυνοῦμαι: cf. on 614.
φιλάνορας ‘husband-loving’. The purpose of this speech and the speaker’s
attitude are quite clear, but have been continually obscured by artifices of
interpretation. Therefore the reader must be warned, at the very beginning,
against any concessions to the playful ingenuity shown by Headlam, among
others: ᾿φιλάνορας is chosen for its double meaning.’ Schneidewin, who,
throughout, was keen on the scent of such double entendres, took the same
view (so also recently B. Snell, Aischylos, 122). φιλάνωρ no more possesses a
secondary meaning here than in 411 or Pers. 136.
858. ro: ‘possessive’.
858 f. οὐκ ἄλλων πάρα κτλ. The introductory announcement is developed in
the asyndetic sentence which follows. οὐκ... μαθοῦσο᾽, ἐμαυτῆς «rÀ.: ‘not
what 1 have heard from others, But my own troubled life’ (Wecklein).
ἐμαυτῆς is brought forward emphatically to mark the second half of the
antithesis. .
861. τὸ μὲν πρῶτον, a phrase belonging together, has a word sandwiched
between its parts, like τὸν μὲν πλεῖστον χρόνον in 5. Aj. 311 ff. καὶ τὸν μὲν
ἧστο πλεῖστον ἄφθογγος χρόνον" ἔπειτα κτλ. In the present sentence γυναῖκα
pushes itself as near as possible to the beginning because it is on this word
that the emphasis is laid: woman’s fate is the theme. ro μὲν πρῶτον is estab-
lished as early as Homer: n 237 τὸ μέν σε πρῶτον ἐγὼν εἰρήσομαι, cf. PI.
Prot. 333 d τὸ μὲν οὖν πρῶτον eraAAwrrilero.... . ἔπειτα μέντοι κτλ.
From Clytemnestra’s words here and in the following lines it is clear that
the poet, out of his experience of long campaigns, knows of the sufferings of
women left alone and is human enough to have a sympathetic understanding
of their misery. Later (Cho. 920) Clytemnestra defends herself with the same
reflection : ἄλγος γυναιξὶν ἀνδρὸς εἴργεσθαι, τέκνον.
863. Auratus’ restoration οἱ κληδόνας is generally accepted (the acrobatics
performed by van Heusde, Verrall, and Plüss in defence of ἡδονάς can be dis-
regarded). That the line is an interpolation was seen by Ahrens (573). He
begins by observing that the enumeration of Clytemnestra’s woes falls into
two sections. The first (861 f.) is introduced by τὸ μὲν... πρῶτον. To this
section the second (864 f.) is attached by «ai (cf. in general Denniston,
Particles, 374). Ahrens then proceeds to point out that the retention of 863
‘confuses intolerably the first and second of the woes in her statement’. This
is actually the case; for in its content 863 belongs to what follows while
390
 COMMENTARY lines 864 f.
formally the participle clause is a subordinate part of the first section (861 f.
γυναῖκα... ἦσθαι. . . ἔρημον), although translations into modern languages
naturally tend to obliterate the subordination of κλύουσαν to the ἦσθαι clause.
Ahrens also recognized the fact that the interpolator borrowed his chief
phrase (κληδόνας παλιγκότους) from 874. The reason for the amplification, or
the possible purpose of the interpolator, can only be a matter of guesswork.
Mere mechanical repetition of the end of 874 and its subsequent expansion
into a whole line, together with the necessary alteration of the word-endings,
does not seem probable. It may be better, perhaps, with Mazon, to consider
whether 863 should not be regarded as an abbreviating interpolation, to take
the place of 864-76. This presupposes, of course, considerable crudity on the
part of the interpolator, in that he was on this supposition content to leave
τὸ μὲν πρῶτον without continuation. A good example of an abbreviating
interpolation (the type is fairly common in the text of Plautus) is provided
by E. Phoen. 428 (again in an unusually long play), composed by someone
who presumably wanted to carry on with 426 immediately after 407. Cf. also
Wilamowitz on A. Sept. 514.
864-73. For the wife of a soldier fighting on a distant front, rumours of bad
news (xAnddves) are a particularly hard trial when reliable information is
scanty and delayed. Cf. Propert. 3. 12. gff. Illa (1.6. the wife of Postumus,
who is just about to join a military expedition to the East) quidem interea
fama tabescit inani, haec tua ne virtus fiat amara tibi, neve tua Medae lactentur
caede sagiliae, ferreus aurato neu cataphractus equo, neve aliquid de te flendum
referatur in urna: sic redeunt illis qui cecidere locıs.
864f. To say nothing of the punctuation of earlier editors, even Wilamowitz,
and, following him, Mazon and Murray, destroy the coherence of the sentence
by putting a comma unnaturally after ἄλλο. A step in the right direction was
taken by Schütz, who remarked: ‘post ἥκειν supplendum κακόν vel πῆμα
λάσκοντα.᾽ This is certainly no ‘hopelessly confused expression’ (Ahrens,
573). All we need to do is to follow the sequence of thought as it presents
itself. It has long been stated (Wilamowitz on E. Her. 237; F. Leo, Analecta
Plaut. i, Göttingen 1896, 5) that the so-called σχῆμα ἀπὸ κοινοῦ ‘non altiori
tantum et artificioso sermoni convenit sed cottidiano et solutiori’ (Leo).
Ag. 864 f. is a case in point. We notice here the somewhat loose mode of
lively colloquial style. καὶ τὸν μὲν ἥκειν : this is not only incomplete, but, for
the moment, without any essential content. But the speaker does not round
off this first clause, because she is so eager to make it clear how close one
messenger with bad tidings follows upon another: ‘and one man comes and
on top of that (ém-) another brings ....’ It is exactly as in the story of Job
(i. 14 ff.), where in continuous succession ἔτι τούτου λαλοῦντος ἦλθεν ἕτερος
ἄγγελος καὶ εἶπεν πρὸς ᾿Ιώβ κτλ. It is just because the phrase τὸν μὲν ἥκειν by
itself still lacks meaning, that the hearer has no difficulty in understanding
from the following words what is necessary to complete the sense. A greater
inelegance is apparently to be found in the fact that the subject of ἐπεισφέρειν
is the second messenger (τὸν δέ), and that in spite of this the plural Aaoxovras
is used. In fact, however, this is completely adequate, for τὸν δέ is really, by
implication, not a true singular. Rather it stands for the long sequence of
messengers (if there were in all only two messengers, the πάθος of the passage
would give way to βάθος) : ‘another and yet another and yet another’.
391
lines 864 f. COMMENTARY
Perhaps the choice of the plural was, to some extent, determined by the fact
that the first messenger, too, was πῆμα Adokwv δόμοις. Of course κακοῦ κάκιον
ἄλλο πῆμα is the object both of ἐπεισφέρειν and of λάσκοντας, but the effect
depends upon the fact that at first we hear only ἐπεισφέρειν : ‘and one man
comes, and then another brings (something) in addition into the house’. For
the theme of 864 f. Beazley recalls £ 126 f. (words of Eumaeus) ὃς δέ κ᾽
ἀλητεύων ᾿Ιθάκης ἐς δῆμον ἵκηται, ἐλθὼν ἐς δέσποιναν ἐμὴν ἀπατήλια βάζει (news
about Odysseus).
866. καὶ τραυμάτων μὲν κτλ. Verrall aptly translates: ‘as for wounds’. It
serves as a kind of heading corresponding to ei δ᾽ ἦν τεθνηκώς in 869, when she
turns to something else.
867. Hermann wrote ἁνὴρ, which has been adopted by many editors. The
protest raised by Ellendt-Genthe, Lex. Soph. 64, against this normalization
seems to be justified. It would be 1416 to look for a special reason for the
‘omission’ of the article in instances such as Sept. 647 κατάξω δ᾽ ἄνδρα τόνδε,
Cho. 561 ἥξω σὺν ἀνδρὶ τῶιδε, and the like. Of course ἀνὴρ, which implies no
change of the παράδοσις, would be perfectly good, too. In a case like this it is
impossible to make out what the poet wrote.
867 f. @xereuero φάτις. van Heusde compares Empedocles B 35. 2 λόγου
λόγον ἐξοχετεύων (also 3. 2 ἐκ δ᾽ ὁσίων στομάτων καθαρὴν ὀχετεύσατε πηγήν).
868. τέτρωται (after 866 τραυμάτων) seems to be the achievement of a ‘think-
ing’ copyist.
πλέω : cf. 1068, Soph. Inachos, Pap. Tebt. 692 1. 68 (=D. L. Page, Greek Lit.
Pap. 24) ταῦτα μὴ λέξηις πλέω (see R. Pfeiffer, Sitzgsb. Bayer. Ak., Phil.-hist.
Abi. 1938, Heft 2. 54).
866-8. Clytemnestra’s urge to overstate the case brings her descriptions
almost to the point of the grotesque: in what follows (Geryon) this goes still
farther.
869. Headlam again (cf. on 856) indulges in ingenious fantasies: ‘A shade of
intonation in the Greek, as in English, would make a wish of this, "If only
he had been killed!" and I fancy this is the suggestion, that he deserved to
die three times over' etc. Clytemnestra's exaggeration originates from the
zealousness of a bad conscience, but she reveals nothing of the subconscious,
nor is there any ambiguity in the words εἰ δ᾽ ἦν τεθνηκώς.
ἐπλήθυον is a necessary emendation since only πληθύειν is used intransi-
tively in classical times, cf. Ernst Fraenkel, Griech. Denominativa, 65.
ὡς ἐπλήθυον λόγοι has been correctly understood by most commentators
since Schütz: ‘or if his deaths had been as numerous as the stories’ (Headlam).
The exact correspondence, in sense as well as form, with the preceding ὡς
clause (867 f.) is clear, as is its connexion with the main clause. So the gloss
in Tr ἤγουν ὡς oi πλείονες ἔλεγον λόγοι (similarly G. Thomson: ‘if he had died,
the prevalent report, he was a second Geryon’) is wrong ; in L-S, too, s.v.
πληθύω, the passage is put under the wrong heading.
871. Schütz saw that this line is spurious. Hermann's attempt to save it was
refuted by Ahrens (576). Ahrens himself assumes a corruption. In τὴν κάτω
γὰρ οὐ λέγω many commentators see the formula for averting the omen
(absit ut... dicam"), following Hermann, who drew attention to this idio-
matic usage (in his re-edition of Viger, De idiotismis, 756) : ‘où λέγω formula
est male ominatum quid proferentis, quod abominari se significat', for which
392
 COMMENTARY line 872
he quotes Ag. 871, Eum. 866, S. El. 1467. Headlam takes a different view:
““not to speak of that which is below”: a piece of laboured jocularity,
banter’. For the supposed idea in τὴν κάτω γὰρ οὐ λέγω (‘terrae portio infra
triplex cadaver πλείστη) reference has been made since Klausen to Sept.
949 f. ὑπὸ δὲ σώματι γᾶς πλοῦτος ἄβυσσος ἔσται. 1 prefer to the many over-
elaborated attempts at defending the line Verrall’s statement ‘this is non-
sense’, although I am unable to discover in this nonsense, as Verrall does, a
particularly nice point of psychology. The poet’s line of thought is bold but
clearly marked. It can be clarified in this way : ‘If he had died as often as was
reported, then, like a second Geryon (one body after the other cut down), he
would have had to be buried three times.’ In the Heraclidae, fr. 74 N. (for the
text cf. Wilamowitz, Verskunst, 460, 611), Aeschylus, while depicting the
three-bodied Geryon, τρίπτυχος, seems to revel in repetitions of the number
three. The extraordinary spectacle of one of the monster’s bodies being cut
down and falling to the ground, while the other two fight on as if nothing
had happened, must have been as familiar to the Athenians as it is to us
through the metopes of the Athenian treasury at Delphi and of the temple of
Zeus at Olympia (there two bodies have already fallen) and from vase-
painting. Clytemnestra in her frenzy of exaggeration follows up the triple
death to its grotesque sequel of a threefold burial. In the closely woven
texture of this clause there is certainly no room between τρισώματος and
χθονὸς τρίμοιραν χλαῖναν for πολλήν... .. How 871 got into the text cannot be
determined with certainty. I would like to suggest that originally for 875
πολλὴν ἄνωθεν ἀρτάνην was written in the margin as a variant (and perhaps
not a bad one, with πολύς in the singular to mark a repeated action) and
then after the third word had fallen out the other two were made up by
someone into a trimeter.' Cf. on 7. But about the details of the process
nothing can be known. Those who defend the line have also to reckon with
the following consideration. ἄνωθεν occurs only six times in Aeschylus, apart
from this passage. Now if the line is genuine, we are to believe that the poet
here wrote ἄνωθεν twice in five lines (this is by no means impossible but
certainly very remarkable), and, moreover, that he prefaced ἄνωθεν once with
πολλήν, and the second time with πολλάς, without seeking to establish any
close connexion between the two expressions, which concern entirely different
matters and are clearly not intended as formal parallels. Verrall regards this
repetition as another piece of psychological refinement: ‘in the same spirit’
as the ‘nonsense’ [mentioned above] ‘immediately afterwards (875) she makes
upon this πολλὴν ἄνωθεν a sort of forced and far-fetched play.’
872. x8ovós . . . χλαῖναν : the passages which illustrate this idea, I’ 56 f. ἦ τέ
κεν ἤδη λάϊνον &ooo χιτῶνα, Pind. N. 11. 16 καὶ reAevràv ἁπάντων γᾶν ἐπιεσσό-
μενος, have long been quoted ; as for Theogn. 428 see Diehl’s note. Cf. also
E. Hel. 851 ff. οἱ θεοὶ... εὔψυχον ἄνδρα πολεμίων θανόνθ᾽ ὕπο κούφηι karapm-
ἔσχουσιν (cf. ἀμπεχόνη) ἐν τύμβωι χθονί. Now we may add Alcaeus, Pap. Oxy.
2165, fr. x col. 1. 17 θάνοντες γᾶν ἐπιέμμενοι κείσεσθε.
λαβών MSS. Analogies for the use of the participle could at a pinch be
1 In E. Suppl. 252 ἀλλ᾽ ὡς ἰατρὸν τῶνδ᾽, ἄναξ, ἀφίγμεθα it appears that the first three
words were originally written in the margin as a variation on the beginning of 256 ἀλλ᾽ ὡς
ὀναίμην and then completed to make a trimeter. (Since writing this I have come upon
Dobree, Advers. ii. 80: ‘nisi mavis larpdv meram esse variam lectionem pro ἀναίμην 256’.]

393
line 872 COMMENTARY
found, cf. Kühner-Gerth, ii. 52 ff. However, not only is the rare word ἐξαυχεῖν
always constructed with an infinitive elsewhere (twice in Soph., once in Eur.),
but the simple αὐχεῖν, which occurs very frequently in Aeschylus and else-
where with the infinitive, never governs a participle. So it is safer to accept
Paley’s suggestion λαβεῖν.
873. μορφώματι: cf. on 1558 πόρθμευμα.
874. With τοιῶνδε (cf. on 613) the whole of the last section (from 864 on) is
summed up, cf. Prom. 331.
παλιγκότων : ‘hostile, injurious, adverse’, and the like. Of the renderings in
Suidas s.v., ἐναντίος is the most suitable and the most comprehensive.
᾿παλίγκοτος est adversus, adversarius, odiosus, infestus’ is Boeckh's explana-
tion on Pind. Ol. 2. 22. Car. Guil. Elberling, Observ. in aliquot locos Ag. Aesch.
(Copenhagen 1828), 9 ff., undertook a thorough examination of the usage of
the word. He showed, as did afterwards W. Aly, De Aeschyli copia verb.
(1906), 48 f., that in the poets, as in Herodotus, the meaning ‘nova mala’ is
not implied. Therefore Nágelsbach ('on account of such ever-recurring
reports’), Wilamowitz (‘to desperation rumour often drove me with terror
ever-new’) and others are wrong while e.g. Paley (‘of such adverse reports’)
is right. The word, which occurs in Archilochus, Sappho, Pindar, and a
number of times in Tragedy, is frequently met in Ionic prose; it appears to
be alien to normal Attic.
876. From Blomfield to the most recent times many have raised objections
to λελημμένης; so emphatically Ahrens. If it is to be retained, πρὸς βίαν
must probably be dissociated from it (in Tr there is a comma after βίαν) and
taken with ἔλυσαν (πρὸς βίαν without a genitive occurs four times in Prom.
and several times in Sophocles, Euripides, and Aristophanes, with a genitive
Eum. 5 and in Sophocles and Euripides). It seems very natural to connect
ἔλυσαν πρὸς Biav: the attempt at suicide is prevented by physical violence, cf.
Bacchyl. 11 (1o). 85 ff. τὸν (Proitos) δ᾽ εἶλεν ἄχος xpadiay . . . δοίαξε δὲ φάσγανον
ἄμφακες ἐν στέρνοισι πᾶξαι. ἀλλά viv αἰχμοφόροι μύθοισί τε μειλιχίοις καὶ Bias
χειρῶν κάτεχον. As to the meaning οἱ λελημμένης, there is the temptation to
translate it ‘when I was caught in the act’ (Ar. Plut. 455 em’ αὐτοφώρωι δεινὰ
δρῶντ᾽ εἰλημμένω) particularly in view of E. Tro. 1012 ff. ποῦ δῆτ᾽ ἐλήφθης ἢ
βρόχους ἀρτωμένη ἢ φάσγανον θήγουσ᾽, ἃ γενναία γυνὴ δράσειεν ἂν ποθοῦσα τὸν
πάρος πόσιν; But if that were so, the aorist would probably be required
instead of λελημμένης. Therefore it may be preferable to take it ‘when my
neck (or ‘I’) was (already) caught in the noose’, cf. Cho. 556 ff. ὡς av δόλωι
κτείναντες ἄνδρα τίμιον δόλοισι καὶ ληφθῶσιν ἐν ταὐτῶι βρόχωι θανόντες. Perhaps
the participle is to be understood as ἃ supplement to complete the picture,
tacked on at the end after βίαν (cf. on 893 θωύσσοντος) : in her eagerness to
depict the situation? as vividly as possible and to make her own danger .
appear as great as possible Clytemnestra adds AeAnupevns. However, the
dissociation of the participle from πρὸς βίαν seems to me rather harsh. I am
uncertain, therefore, whether Aeschylus really wrote λελημμένης. Of the

1 Cf. also Elmsley (on E. Bacch. 1100 [1102 vulgo]): ‘Adde Aesch. Ag. 876, si sana est
scriptura."
2 This situation is quite typical, cf., besides the passage in the Troades quoted above,
E. Andr. 815 f., Andoc. 1. 125 ἀπαγχομένη μεταξὺ κατεκωλύθη (with regard to other attempts
at restoration cf. the editions), καὶ ἐπειδὴ ἀνεβίω «TA,

394
 COMMENTARY line 880
conjectures offered none is satisfactory. With regard to the parallel forms
λέλημμαι and εἴλημμαι cf. Pearson on Soph. fr. 750. Ahrens raised objections
to ἄλλοι; the explanation of Wilamowitz 'dAAo. dictum videtur, ut ipsam
. nexuisse nodum intellegamus’ hits the mark.
877. ἐκ τῶνδέ vov: also Cho. 1056.
877-9. ἐκ τῶνδε... . Ὀρέστης. If the clauses of such a sentence are re-
arranged after the smooth manner of polished prose, a rendering of this kind
is produced : "This is the reason that we have not standing by our side here,
as we should, the boy Orestes, who in his person owns the pledges of love
between us two' (Headlam). But if the translator takes care to follow the
deliberate order of the text, he writes: 'And this is why our son is not stand-
ing here, the guarantee of your pledges and mine, as he should be, Orestes'
(MacNeice). In itself the extended hyperbaton ('weite Sperrung) παῖς...
*Opéorns would not be so very extraordinary, but it would increase the force
of the expression more than is natural here. It is especially the delayed ὡς
χρῆν that points the way to a correct understanding of the whole. Drop by
drop the phrases trickle out, and only their sum has the effect of a coherent
thought : ‘our son is not here, the pledge of our union, as he ought to be here,
Orestes.' The setting of the name significantly at the end is in keeping with
a widespread practice (cf. on 681 ff.) ; here it has a more ominous sound for
the audience, who know the sequel of the story, than for the characters on
the stage. The hesitant and piecemeal confession provides the dexterous
Clytemnestra with a means of appearing reluctant to utter her excuses;!
at the same time, to mitigate the disappointment, she soothingly pushes to
the front ἐμῶν re καὶ σῶν κύριος πιστωμάτων. :
878. πιστωμάτων : for the necessity of this emendation cf. Ernst Fraenkel,
Griech. Denominativa, 150.
879. μηδέ: cf. on 1498.
880. δορύξενος: in Aeschylus this word occurs only in the Oresteia (Cho. 914;
Cho. 562 is an interpolation, cf. on 1226) and only in connexion with the
relation between the house of Orestes and that of Pylades ; from there it was
taken over by Sophocles, El. 46, to describe the same relation. The use of the
word in Tragedy is not, however, confined to the tie between these two
families ; cf.,e.g.,S. Oed. C. 632, E. Med. 687, Andr. 999. The meaning supplied
for δορύξενοι by Aristophanes of Byzantium (Miller, Mél. 433, cf. Arist. Byz.
191 ff. Nauck) suits all the tragic passages: of κατὰ πόλεμον ἀλλήλους φιλο-
ποιησάμενοι. The word, therefore, designates the ‘establishment of hospitable
rejations as a result of brotherhood in arms’ (Wilamowitz on Cho. 562). The
fact that ‘the simple bond of hospitality may be supplemented by various
other things, and especially, in the case of a covenant between communities,
by agreements about war and peace, truce (indutiae), and military alliance -
(foedus)’ [this last, as we see, also between individuals and their descendants],
was noted by Th. Mommsen in his article on ‘Das römische Gastrecht’ (Röm.
Forsch. i. 331, cf. also, for early Greek conditions, the excellent sketch of the
rights of hospitality by Wilamowitz, ‘Staat u. Gesellschaft der Griechen’,

1 Something comparable, though in quite a different style, occurs e.g. Ar. Ach. 1051 ff.
ἐκέλευε δ᾽ ἐγχέαι σε τῶν κρεῶν χάριν, ἵνα μὴ στρατεύοιτ᾽, ἀλλὰ κινοίη μένων, eis τὸν ἀλάβαστον
κύαθον εἰρήνης ἕνα : quite at the end, when it must at last be said, he names the precious
object of his request, which he knowsit will be almost impossible to get the owner to part with.
395
line 880 COMMENTARY
2nd ed., 39 ff., Kultur der Gegenwart, Teil II, Abt. IV. τὴ. It is likely that
Aeschylus found the term in current use somewhere (not necessarily at
Athens). The remarkable custom, however, which Plutarch, Quaest. Graec. 17,
Mor. 295 Ὁ, c, assigns (source unknown) to archaic Megara, and on account of
which he discusses δορύξενος, presupposes a different relation from that which
must be presumed between Strophius and Agamemnon and in the other
passages in Tragedy.
881. XIrpopios: as the Medicean gives Στροφίος at Cho. 679 (cf. Blass,
Aischylos’ Choephoren, 24), S. El. 1111 and in the ὑπόθεσις of the Electra, there
is no point in following the later MSS here and in other passages in accenting
Στρόφιος, especially in view of the teaching of the grammarians (Etym. M.
521. 10; cf. Lehrs, De Aristarchi stud. Hom., 3rd ed., 265 f.; Chandler, Greek
Accentuation, 2nd ed., 69 f.) τὰ εἰς os λήγοντα τριβράχεα ἐπὶ κυρίων παροξύνεται
οἷον Σχεδίος, Χρομίος, KAvrios krÀ.!
ἀμφίλεκτα : the scholiast (Zy. rad.) did not know what to do with this; he
explains: τὰ ἀμφιβαλλόμενα περὶ σοῦ πήματα, ἤτοι τὰ ὑπὸ τοῦ δήμου κατὰ σοῦ
πεπλασμένα δεινὰ μηνύματα, 1.6. improvising arbitrarily he gives ἀμφίλεκτα
something of the same meaning as περιβόητα. This interpretation is clearly
influenced by 864 ff. It is not easy to find the same sense in the word here as
it has elsewhere. There is one and the same meaning in Ag. 1585, Sept. 809,
E. Phoen. 500; the rendering ἀμφίβολος preserved in Hesychius s.v.
ἀμφίλεκτος is adequate in all these passages. Cf. the Homeric ἀμφήριστος.
Linguistically this usage is quite clear ; ‘audi means, in compounds, “on both
sides” ' (Wilamowitz on E. Her. 1274, cf. above on 686) ; accordingly ἀμφί-
λεκτος means ‘maintained on both sides (in opposition)’, exactly as ἀμφίλογος
S. Ant. 111 νεικέων ἐξ ἀμφιλόγων, E. Med. 637 ἀμφιλόγους ὀργάς. ἀμφίλ.
πήματα here has been taken in the same way by a number of commentators,
e.g. by Peile and by Conington, who translates ‘the mischief of disputes’ and
notes that the expression is exactly parallel to S. Ant. 111 νεικέων ἐξ ἀμφιλόγων
and E. Phoen. 500 ἀμφίλεκτος ἔρις, overlooking the fact that this meaning of
ἀμφίλεκτος is closely akin to that of ἔρις and νείκη but not to that of πήματα.
Moreover, the thought ‘the mischief of disputes’, ‘the evil of controversy’
(Plüss), is not at all suitable here, apart from the fact that the adherents of
this interpretation see themselves compelled, with Stanley, to make the
following clauses (τόν re . . . κίνδυνον and ef re . . . καταρρίψειεν) co-ordinate
with ἀμφίλ. πήματα. Pauw, however, saw that the two clauses following on
προφωνῶν, bound together as they are by re... re, form an explanatory
apposition to πήματα. Other scholars (as far back as Passow) understand
‘doubtful, uncertain woes’ (Schneidewin—Hense with ‘bedenkliche’ and Mazon
‘des périls inquiétants’ are very vague), but this diverges considerably from
the usual meaning of the word, and from what is indicated by the general
function of ἀμφί. Since the second half of the nineteenth century, a third
meaning which had been put forward by Stanley (‘ancipitia mala’) and later
e.g. by Linwood (‘of double import, twofold’) has come to the fore (e.g.
Keck, Paley, Wecklein, Sidgwick, Wilamowitz, Verrall, Headlam, LS),
according to which ἀμφίλεκτα here means merely ‘twofold, double’. That is
probably correct (I hesitate to say more) although in assuming it we ascribe
1 For the accentuation 'Exéos, KAovios, Σχεδίος, etc., in the older MSS of the Iliad cf.
T. W. Allen, Homer's Iliad, Prolegomena (1931), 239 f.

396
 COMMENTARY line 883
to Aeschylus a rather bold alteration in the meaning of the word. Schuursma,
De... abusione, 132 adopts the meaning ‘double’; the term ‘abusio’ seems
better justified in this case than in many of his examples. There is, however,
one difficulty left. Those who take ἀμφίλεκτα = ‘double’ have not said what
they believe to be the function of the verbal element. I can only suppose
that -Aexra intensifies the idea contained in the verb προφωνῶν and that the
expression calamitates bifariam memoratas praedtcens is to be associated with
that well-known type of tragic speech which is grouped, often unsuitably,
under the heading of ‘prolepsis’, cf., e.g., S. Aj. 69 ἐγὼ yap ὀμμάτων ἀποστρό-
gous αὐγὰς ἀπείρξω. All things considered, the possibility (but not probability)
of a corruption in ἀμφίλεκτα must be admitted; the conjectures offered,
however, will not stand closer examination.
882. προφωνῶν: since the main verb (τρέφει) indicates an action extending
from the past into the present, there is no difficulty in understanding po-
φωνῶν of the time when the τρέφειν started. The relation between 889
βλάβας ἔχω and 890 κλαίουσα, which is concerned with the past, is similar.
883. δημόθρους ἀναρχία. As far as the main point is concerned, Stanley’s
translation, ‘populi concitati rebellio’, holds the field unchallenged: variations
and refinements in detail matter little (‘seditio cum clamore populari’ Blom-
field, others somewhat differently). The spreading of this misconception has
probably been promoted by the unconscious inclination to find here the later
meaning of ἀναρχία (‘popular anarchy’ Conington, ‘popular tumultuous
anarchy’ Headlam, etc.). So, following Wellauer and Linwood, Dindorf
(Thesaurus) and L-S separate this passage from the only other genuine
Aeschylean passage where the word occurs, Suppl. 907 (see below), and put it
under the heading ‘anarchy’ with the later instances. As for the adjective
dvapyos, it is in Homer confined to the Catalogue (B 703 = 726), where it
means ‘without a leader’. Aeschylus uses it in the same way (Eum. 696, cf.
526). To this use A. Suppl. 906 f. corresponds: πολλοὺς ἄνακτας... ὄψεσθε...
οὐκ ἐρεῖτ᾽ ἀναρχίαν, ‘you will not assert lack of a master’. This corresponds
exactly to δημόθρους ἀναρχία, ‘the people’s clamour (or ‘the assertion made by
the people’s rumour’) that they lack a ruler’. avapyin has the same meaning
in the only passage in Herodotus in which it occurs: 9. 23. 2 refers to the
cavalry who have just lost their hipparch: ἐδόκεε δέ σφι ἀναρχίης ἐούσης
ἀπελαύνειν παρὰ Μαρδόνιον. The same meaning underlies the term ἀναρχία in
the Attic list of archons (cf. Wilamowitz, Arist. u. Ath. i. 6). The earliest
example of the word used to denote ‘unruliness, disobedience to authority,
anarchy’ is S. Ant. 672 ἀναρχίας δὲ μεῖζον ox ἔστιν κακόν (referring back to the
immediately preceding words,' 669 καλῶς μὲν ἄρχειν, εὖ δ᾽ ἂν ἄρχεσθαι θέλειν).
It would be conceivable that Aeschylus too, in a late play, used the word in
this meaning. But as far as his extant works are concerned, he has not done
so. It isa pretty point, however, that in the scene which we now read at the
end of the Seven against Thebes ἀναρχία (1030) is used in the sense in which it
occurs in 5. Ant. 672. It is well known that this scene depends to a large
extent on the Antigone of Sophocles.
As soon as the meaning of ἀναρχία in Ag. 883 is correctly understood,
δημόθρους also appears in a much clearer light. It is exactly parallel to the
! This alone is enough to refute Seidler's transposition, which appears once more in
Pearson's text.
397
line 883 COMMENTARY
δημόκραντος of 457. δημόθρους does not occur outside the Agamemnon, cf. on
386. In 938 it is the attribute of φήμη, in 1409 (again 1413) of dpai. Here, 883,
it does not qualify ‘speech’, ‘rumour’, or some such word, but the contents of
the speech ; the people say: ‘there is no ruler here’ (cf. above on Suppl. 907
ἐρεῖτ᾽ ἀναρχίαν). Decisively in favour of this interpretation is the fact that it
alone brings out clearly the close connexion of the two sections of Strophius’
argument which are joined together by τε... 7€. . .: Agamemnon’s danger at
Troy leads at home to the kind of talk indicated by δημόθρους ἀναρχία: ‘we
have no master here and who knows when, or whether at all, he will return.”
A. Platt saw the point when he translated ‘seeing that there was no ruler in
Argos’.
There is here a strong compression of phrase reminiscent of some of the
poet’s lyrical passages ; the same is true of βουλὴν καταρρίψειεν which follows.
This conciseness forms a striking contrast to the prolixity of those parts of
Clytemnestra’s speech in which she depicts her sufferings or flatters her
husband.
884. βουλὴν καταρρίψειεν. This is another case in which a misconception of
Stanley’s has held its ground for centuries, almost completely unchallenged.
The great majority of commentators (and also L-S) follow Stanley’s transla-
tion ‘senatum deturbaret’. Severalscholars elaborated the point by supplying
a background of constitutional law. This view dominates the commentaries
from the time of Wellauer, Otfried Müller (Eumenides, Ὁ. 76: “The Chorus in
the Agamemnon represents a high Council (γερουσία), which the prince left
behind to administer the state in his absence; see ll. 855, 884’), and Hermann,
ad loc.—in striking contrast to his former virulent attack on Müller,
printed in his Opusc. vi. 2. 136 f.—down to Wilamowitz (who speaks in
Griech. Trag. ii. 32 f. of ‘the Chorus, which formed the Council of regency at
Argos’, cf. also his Interpr. 166) and his followers. In his commentary,
however, Stanley takes an entirely different view : 'consilium protectum iniret.
Sc. occidendi Orestis.’ It is not clear here how he means to construe the
words ; possibly he is compromising with Scaliger’s conjecture καταρράψειεν,
which was repeated by Abresch and adopted by Wecklein (Comm.),' Head-
lam, and A. Y. Campbell. Agreeing with Stanley’s comment, Heath takes the
passage to mean ‘ausum aliquod periclitaretur’, similarly (though not clear in
detail) Blomfield, who, on account of pirrew, assumes a metaphor from
dicing, Linwood, J. G. Droysen (‘verruchte Plane schmiede’), Verrall (‘and
the chance that noisy rebellion from below might risk a plot against us’),
Platt. These explanations quoted from a minority of commentators represent
in the main a common view in that they presuppose more or less clearly that
βουλή alone can mean ‘a plot against someone’. This is, however, extremely
doubtful. It is true that Verrall quotes, to support it, Andoc. 1. 61 ἐξήλεγξα
τὰ γενόμενα, ὅτι εἰσηγήσατο μὲν πινόντων ἡμῶν ταύτην τὴν βουλὴν Εὐφίλητος,
ἀντεῖπον δὲ ἐγώ, but, supposing that the text as quoted is correct,? βουλή is

ı Wecklein's remark (Introduction to his commentary, p. 24 n. 1) on the absence in the

Agamemnon of any hint that the Chorus represents a γερουσία is perfectly sound.
2 It is generally recognized that Bekker's excision of the MS γενέσθαι after rjv βουλήν
is correct. My question is whether the sentence would not gain considerably in compactness
if ταύτην τὴν βουλὴν γενέσθαι were taken as a gloss and omitted ; this would at once do away
with the awkwardness of assigning to ταύτην τὴν βουλήν an entirely different meaning from

398
 COMMENTARY line 884
qualified there by the demonstrative pronoun and the word itself means, as
always, nothing more than ‘resolution, plan’. The case is no different in
Hdt. 1. 63. 2 (in Powell’s Lexicon it is put under a special heading ‘plot’),
where its particular shade of meaning is imparted by σοφωτάτην ἐπιτεχνᾶται,
while the meaning of βουλή in itself is the same as in 4. 134. 2, 5. 118. 2 (in-
fluenced by the Homeric ἀρίστη φαίνετο βουλή), 8. 100. 3. To grasp the true
meaning of βουλὴν καταρρίψειεν we have only to remember that βουλή not only
expresses the result of βουλεύειν or βουλεύεσθαι, as in ἡ δὲ κακὴ βουλὴ τῶι
BovAedoavrı κακίστη (Wilamowitz, Hesiod Erga, p. 164, is wrong) and e.g. Ag.
1358, S. El. 1047, Plato, Crito 46 a, but also has quite commonly the function
of a ‘nomen actionis’, i.e. means ‘the making of a plan, the consideration’,
etc. Thus e.g. Thucydides i. 138. 3 τῶν παραχρῆμα 9v ἐλαχίστης βουλῆς
κράτιστος γνώμων, 5. τοὶ οὐκ, ἦν γε σωφρόνως βουλεύησθε... περὶ δὲ σωτηρίας
... ἡ βουλή, 111. 5 (where βουλή again picks up βουλεύεσθαι), 6. 9. 1 οὕτως
βραχείαι βουλῆι, Plato, Phaedr. 237 € μία ἀρχὴ τοῖς μέλλουσι καλῶς βουλεύσεσθαι"
εἰδέναι δεῖ περὶ οὗ ἂν ἦι ἡ βουλή, Epist. 7. 346 d ἐβουλενόμην δή... πρῶτος δ᾽ ἦν
μοι τῆς βουλῆς ἡγούμενος ὅδε λόγος, Arist. Eth. Nic. 3. 5. 1112 ἃ 18 βουλεύονται
δὲ πότερα περὶ πάντων, καὶ πᾶν βουλευτόν ἐστιν, 7) περὶ ἐνίων οὐκ ἔστι βουλή; 6. το
from 1142 a 31 to the end of the chapter. This use of the word is also found
in tragedy. Cf. e.g. Ion, fr. 63. 3 (p. 745 N.), with reference to Sparta, εὖτ᾽ ἂν
"Apns veoxpös ἐμπέσηι στρατῶι, βουλὴ μὲν ἄρχει, χεὶρ δ᾽ ἐπεξεργάζεται. Soph.
fr. 772 N. (= 856 P.) οὐ γάρ τι βουλῆς ταὐτὸ καὶ δρόμου τέλος is still more
instructive. For it brings us to an idea widespread in this period. Haste and
speed are inconsistent with and indeed harmful to true βουλεύεσθαι, to genuine
βουλή. Not only had the Abyvatoı ταχύβουλοι (Ar. Ach. 630) to be warned by
their poets and statesmen of the dangers of acting without due deliberation
(Diodotus in Thuc. 3. 42. 1 νομέζω δὲ δύο τὰ ἐναντιώτατα εὐβουλίαι εἶναι, τάχος
τε καὶ ὀργήν, Nicias ibid. 6. 9. 1 μὴ οὕτως βραχείαι βουλῆι περὶ μεγάλων πρα-
γμάτων.... πόλεμον οὐ προσήκοντα αἴρεσθαι); the whole Greek world is full of the
same warning. ‘First sleep on it’ is the advice given in a popular aphorism, or
in its Greek form: ‘leave the βουλεύεσθαι to the night’, νυκτὶ φωνήν, νυκτὶ
βουλήν, νυκτί τὴν νίκην δίδου (Plut. Themist. 26. 2; cf. Hdt. 7. 12. ı, Phocylides
fr. 8 D. νυκτὸς βουλεύειν κτλ., Menander, Epitr. 35 = 76 Koerte, 3rd ed.).
Emphatic warnings against ταχυβουλία are given in Theognis 633 f. βουλεύειν
δὶς καὶ τρίς, ὅ τοι x” ἐπὶ τὸν νόον ἔλθηι" ἀτηρὸς γάρ τοι λάβρος ἀνὴρ τελέθει, 1051 ff.
μήποτ᾽ ἐπειγόμενος πράξηις κακὸν, ἀλλὰ βαθείηι σῆι φρενὶ βούλευσαι (cf. A. Pers.
142 φροντίδα κεδνὴν καὶ βαθύβουλον θώμεθα, on this see Riemschneider, Hermes,
Ixxiii, 1938, 349) σῶι ἀγαθῶι τε vou: τῶν γὰρ μαινομένων πέτεται θυμός τε νόος
τε, βουλὴ δ᾽ εἰς ἀγαθὸν καὶ νόον ἐσθλὸν ἄγει. Here again, we find βουλή by itself
in the sense of (quietly) thinking something over, deliberating. This is
exactly the case in Ag. 884."
that of βουλή in the immediately preceding εἶπον τῆς βουλῆι. If an object for εἰσηγήσατο was
supplied from the context, this would correspond to the usage of Thucydides (4. 76. 2;
6. 90. 1; 6. 99. 2), Plato (Crito 48 a), and decrees recorded in inscriptions (Hellenistic examples
in Dittenberger, Syll. iv. 310). Section 67 of the speech provides an extension of the concise
version of 61.
t Some of the passages quoted in the previous lines have been discussed in a wider
connexion by E. B. Stevens, ‘The Topics of Counsel and Deliberation in Prephilosophic
Greek Literature’, Class. Phil. xxviii, 1933, 104 ff. I wrote without any knowledge of this
essay, and have let everything stand just as it was written.

399
line 884 COMMENTARY
καταρρίψειεν : ‘if the ruler’s absence, asserted noisily by the people, should
overthrow (quiet) deliberation'.! This seems to be the only known example
of καταρρίπτειν in pre-Hellenistic Greek, but it should be observed that its
use here is strong and very fine, as a more elevated equivalent of καταβάλλειν.
The metaphorical use? of the latter to describe the overthrowing and setting
aside of obstacles other than physical is well known; besides the examples
collected by Kranz, Vorsokr. iii, 5th ed., 229, cf., e.g., Hdt. 8. 77. 1 οὐ βουλό-
pevos évapyéws λέγοντας (scil. χρησμούς) πειρᾶσθαι καταβάλλειν, E. Her. 759,
Bacch. 201 f. πατρίους παραδοχάς, ds θ᾽ ὁμήλικας χρόνωι κεκτήμεθ᾽, οὐδεὶς αὐτὰ
καταβαλεῖ λόγος. Aeschylus likes to use images of this kind when describing
how passion or wickedness tries to overthrow and destroy the strong founda-
tion which guarantees moderation and morality, e.g. Ag. 382 ἀνδρὶ λακτί-
cavrı μέγαν Δίκας βωμὸν eis ἀφάνειαν, Eum. 538 ff. βωμὸν αἴδεσαι Δίκας" μηδέ
νιν κέρδος ἰδὼν ἀθέωι ποδὲ λὰξ ἀτίσηις, and with a different, though related,
image, 698 καὶ μὴ τὸ δεινὸν πᾶν πόλεως ἔξω βαλεῖν (this would be upsetting the
stable condition described as necessary in 516 ff). Only Plüss among the
commentators known to me has understood βουλὴν καταρρίψειεν correctly.’
Schoemann, Opusc. iii. 178, who unfortunately resorted to an unwarranted
alteration of the text, objected at least to the predominant interpretation of
βουλή = senatus. On this head there can be no doubt. Nowhere in the
whole play (cf. on 256 f.) has the poet given the slightest indication that
the Chorus of old men was to be regarded as representing an advisory body.
The decisive arguments for the interpretation put forward here are two-
fold. First, the sequence of thought thus produced runs smoothly and con-
sistently, secondly, the thought is in agreement with the view taken by
Aeschylus in this play and elsewhere of the consequences for the safety of the
ruling house resulting from a long war with many setbacks. As to the first:
if the people at home are always hearing and speaking about the danger
which threatens the king in the field, if they say: ‘we have no master, the
king will probably never return’, then their readiness to think and deliberate
quietly will give way and the licence let loose will turn against the heir
apparent, still a minor. All that matters in this context is the relation between
the people and the royal house ; the mention of a Council of regency (possible
under other circumstances) would here only serve to hinder and weaken the
direct impact of these two on each other. Moreover, we have to consider the
whole Aeschylean conception of such typical occurrences. In 449 ff. and
456 ff. the Chorus describes how out of the murmuring of the people at home
there arises a danger to the royal house. Exactly the same picture is drawn
1 If the objection is raised that it is not the ruler's asserted absence that can overthrow
anything, but only the fact that his absence is asserted, I accept this more circumstantial
rendering.
2 Originally derived from the language of wrestling, but the metaphorical use is quite
common, cf. Wilamowitz, Platon, ii. 415 n. I.
3 Recently I have found (in the library of New College) the modest little paper of a
young scholar, H. N. Bosker, Specimen literarium inaugurale, Groningen 1823. There
(p. 7) he comments on Ag. 883 f.: ‘Hic βουλὴ non senatum significare mihi videtur, sed
consilium, reclam agendi normam, quam periculum erat ne popularis seditio abjiceret. Nam
de quonam senatu hic sermo esse possit, non video ; cum absente Agamemnone rege summa
auctoritas non penes aliquem senatum, sed penes reginam esset. Anyone who has con-
ducted seminar classes knows that the common sense of the young often shatters the subtle
devices of their elders, and that only bad teaching can deter them from speaking their mind.
400
 COMMENTARY line 885
in the concluding stanzas, already referred to, of the great choral ode in the
Persae (584 ff.), where some of the consequences of a defeat are described thus:
βασιλεία γὰρ διόλωλεν ἰσχύς. οὐδ᾽ ἔτι γλῶσσα βροτοῖσιν ἐν φυλακαῖς" λέλυται γὰρ
λαὸς ἐλεύθερα βάζειν, ὡς ἐλύθη ζυγὸν ἀλκᾶς. Such a general ‘dissolution’
necessarily includes βουλὴν καταρρίπτειν.
ὥς τι σύγγονον. The MS reading ὥστε has been taken, almost universally,
as causal, e.g. by Nägelsbach ‘because it is natural to man’, Headlam: ‘it
being natural impulse’, Mazon: 'piétiner l'homme à terre étant un besoin
inne aux mortels’. Denniston, however, has shown, C.R. xlvii, 1933, 163 f.
(cf. Greek Particles, 527), that the causal ὥστε, frequent in Herodotus, is
almost completely absent from Tragedy. So he writes here ὥς re and trans-
lates ‘and saying that’. The separation of ὥς re had already been made by
R. Klotz; it was accepted by Plüss, with the explanation ‘and the thought
that’. This coordination of the oéyyovov-clause with the preceding clauses
produces a thought which leaves much to be desired: who would think of
putting the two special elements in Strophius’ warning, 1.6. the danger
menacing Agamemnon and the morale at Argos, on the same level as the
general observation on human nature σύγγονον βροτοῖσι τὸν πεσόντα λακτίσαι
mÀéov? And is it possible to treat this last as an integral part of the πήματα
npobwveiv? There remains the further difficulty of reconciling the three parts,
coordinated in each case by τε, with the introductory word ἀμφίλεκτα. To
Denniston's note: ‘it would be pedantic to object that the third re intrudes
upon the duality of ἀμφίλεκτα πήματα᾽, my answer is that, in this case, I will
gladly submit to the charge of pedantry. But since ἀμφέλεκτα, as already
shown, is not completely above suspicion, let the main argument be rather
that the subject-matter of the third clause is far more suitable for subordina-
tion in a causal sentence. Denniston himself admits as much when he says
twice without any modification 'the sense required is causal, "Quippe cum
innatum sit...” ' and ‘a causal force is required here.’ Therefore I write
ὥς τι (so did Hartung, who, however, did not take ὡς as causal and missed
altogether the connexion of the clauses; Keck followed him, but wrote ós
τὸ... ἡ. The τι has the effect of gently toning down, almost in the same way
as a πως, and it is suitably added to an adjective which serves as predicate
when the subject of the sentence is an infinitive, cf. Prom. 536 ἡδύ τι θαρσα-
λέαις τὸν μακρὸν τείνειν βίον ἐλπίσι, Alexis fr. 210. 3 K. οὕτω Tt τἀλλότρι᾽ ἐσθίειν

ἐστὶ γλυκύ (cf. also & 80 οὐ γάρ τις νέμεσις φυγέειν κακόν). The placing of τι in
front of the adjective is customary in Aeschylus, cf. Dindorf, Lex. Aesch., the
end of p. 358, beginning of p. 359. For the ‘ellipse’ of ἐστε in the subordinate
clause cf., e.g., Prom. 763 ei μή τις βλάβη. If the conjecture τι is rejected, we
should have to assume that here, as in Herodotus, ὥστε is equal to ὡς ; I do
not consider this impossible though one would expect to find more instances
in Tragedy.
σύγγονον: cf. 832, Arist. Poet. 4. 1448 Ὁ 5 τό τε γὰρ μιμεῖσθαι σύμφυτον τοῖς
ἀνθρώποις ἐκ παίδων ἐστί κτλ. Quite similar in expression is Tacitus, Hist. 2.
20 insita mortalibus natura recentem aliorum feltcitatem acribus oculis 1ntro-
spicere etc.
885. There may be here an allusion to some proverbial expression. The
τ G. Thomson has in his commentary Headlam’s note: “Two things might happen’, etc.,
but as in the text he follows Denniston, he must translate ἀμφίλ. πήμ. by ‘various dangers’.
4872.2 pd 40I
line 885 COMMENTARY
commentators quote, among other passages, 5. 447. 988 f., 1348, El. 835 f.,
Ar. Clouds 550, Lucret. 5. 1140 nam cupide conculcatur nimis ante metutum
(1139 sub pedibus vulgi). There appears to be an allusion to the same proverb
in 5. Ant. 1029 1. ἀλλ᾽ εἶκε τῶι θανόντι, μηδ᾽ ὀλωλότα κέντει. Cf. also Plutarch,
Demosth. 22. 4 ἀγεννὲς ζῶντα μὲν τιμᾶν... πεσόντα de... ἐπισκιρτᾶν τῶι νεκρῶι.
886. τοιάδε μέντοι: cf. on 644.
Schütz comments: ‘Falso Stanleius: Equidem tale monitum non est fraudu-
lentum. Sensus est: Et haec quidem excusatio mea omni dolo vacat.’ Stanley's
translation was followed by Klausen and others and also by Wilamowitz :
‘Das war gewiss ein redlicher und guter Rat’ and Headlam: ‘this was the
representation, carrying no guile in it.' Wecklein and others follow Schütz.
The predominant meaning of σκῆψις (which here occurs for the first time),
corresponding to a well-known usage of σκήπτεσθαι, is 'grounds for excuse'
(Wilamowitz on E. Ion 721), both in Attic (including the technical σκῆψις, the
objection against the undertaking of a liturgy) and in Herodotus. Its glossing
by πρόφασις (Hesychius, Suidas s.v. σκῆψις, where see A. Adler's note for
further evidence) covers roughly the whole usage. The expression in Ar. Ach.
392 ὡς σκῆψιν ἁγὼν οὗτος οὐχὶ δέξεται is well illustrated by the comparison
(Küster) with the proverbial (Zenob. 2. 45) ἀγὼν πρόφασιν οὐκ ἐπιδέχεται.
Sophocles' one instance (El. 584) of σκῆψις and the frequent instances in
Euripides suit this meaning. Therefore the word cannot mean 'warning,
advice, representation', and Schütz is right. The whole section from 877 was
an excuse for the absence of Orestes. It is this section in its entirety, and not
only the report which it contains about the warnings of Strophius, that
Clytemnestra now concludes by emphasizing that her excuse contains no
δόλος ; it is just in this matter, of such consequence for the future, that her
conscience is particularly troubled.
φέρει is here, as often in Tragedy, ‘stronger than ἔχει" (L-S, s.v. A. t).
It would be clear to the audience, from the very eagerness with which
Clytemnestra presents her excuses for the absence of Orestes, that she is
lying. If any doubt about her dishonesty remained, it would be removed by
the concluding line 886. Her lie is as subtle as befits such a woman: the fact
that she herself sent her son to Strophius is true (cf. Cho. 913 ff.), but of
course she did it to get him out of the way; the alleged advice of Strophius
is a guileful invention of her own. There is no indication in Aeschylus as to
the time when Orestes was sent away;' concerning the treatment of this
point in other branches of the tradition, cf. O. Gruppe, Mythologie, 701 f.,
Lesky, RE xviii. 975 f., and Jacoby's commentary on F Gr Hist go (Nicolaus
of Damascus) F 25.?
1 T. D. Seymour, C.R. viii, 1894, 438 ff., attempts to fill the gap with ingenious speculations.
? Jacoby's treatment of the passages of Pindar and Aeschylus is not happy. In Pindar
Orestes is rescued by the nurse when Agamemnon is murdered; in Aeschylus the nurse has
nothing at all to do with taking Orestes away to Strophius. The point of Kaibel's remark
(Commentary on Soph. El., p. 46 n. 1) ‘but a child at least ten years of age does not sleep
any longer with his nurse' escapes me. Is it supposed to discredit Pindar's account as not
true to life? Of sleeping with the nurse the poet says nothing. And is it unnatural that in
the turmoil of a murderous attack the nurse should snatch her former nursling from any
part of the house where he might be and get him out of harm's way? Moreover, there is no
justification for assuming, even in general, that in the time of Pindar there was always
ἃ strict separation of almost grown-up boys from the female servants; young Heracles
appears, even in the street, accompanied by an old woman (nurse?): skyphos of the
402
 COMMENTARY lines 890 f.
887. ἔμοιγε μὲν δή. For γε μὲν δή in Aeschylus cf. Blass on Eum. 419,
Denniston, Particles, 395. In all passages except Eum. 419 (which Blass
alters) it marks a contrast, though not a very strong one, cf. 661, 1213. Here
Clytemnestra after her long digression about Orestes returns to her own life
during Agamemnon’s absence. ἔμοιγε μὲν δή marks the beginning of a new
section as clearly as τοιάδε in 886 marked the end of the previous one.
ἐπίσσυτοι: cf. on 1150.
888. κατεσβήκασιν: cf. 958, Sept. 584.
890 f. τὰς ἀμφί oo... αἰέν. Most of the earlier commentators, and some of
the more recent, referred this to the beacon-signs. Blomfield, on the other
hand, suggested: ‘Quid si intelligamus per Aaparnpovyias nocturnas focu-
lorum accensiones, quibus veteres cubicula illustrare solebant? ut λαμπτῆρες
sint vigiles lucernae, ad quas Clyt. se adsedisse dicat, dum coniunx frustra
exspectaretur.’ This interpretation was adhered to e.g. by Hartung, Con-
ington, Enger, Keck, Sidgwick, Headlam, Platt, A. Y. Campbell. Headlam
elaborated it with colours borrowed from Hellenistic erotic epigrams (as if
one botched up a fresco of Michelangelo with patches from paintings of
Watteau). The decision between the two interpretations cannot be based
upon the meaning of λαμπτηρουχίαι, for this word would in either case yield
a suitable sense. It is a ἅπαξ λεγόμενον. No instance of λαμπτηροῦχος was
known until in a recently published inventory of sacred objects from Thespiae
(Bull. Corr. Hell. \xii, 1938, 150 ff.), drawn up between 395 and 380 B.c., the
following item appeared: λανπτερῶχοι σιδάριοι τρῖς, ‘lamp-stands’." It is
possible that λαμπτηροῦχος was used in this sense in Athens at the time of
Aeschylus; but whether this was so or not, it was certainly not this usage
that was the starting-point for the coinage of λαμπτηρουχία, but λαμπτηροῦχος
said of a person, ‘light-bearer’, or, perhaps, λαμπτηρουχεῖν (cf. δαιδουχεῖν).
Now ἔχειν λαμπτῆρα or λαμπτῆρας might as well denote the function of
servants engaged in holding lamps or torches (cf. on 22) in the house (cf.
Cho. 537) as the office of those ‘torch-bearers’ whom the queen praises in 312.
The first alternative cannot be completely excluded but seems most im-
probable. Why should Clytemnestra expect that Agamemnon, on his way
from Nauplia or the Temenion to Argos, would arrive by night? It seems
odd, too, that of the various preparations for a worthy reception of the king
the comparatively small detail of the lights should be singled out for special
mention. On the other hand, the device of the beacon signs, which long ago
was arranged between Agamemnon and Clytemnestra and the execution of
which was supervised by the queen with great care, is of primary importance
to all concerned. From the point of view of the audience, nothing would be
more natural than to connect Clytemnestra's phrase λαμπτηρουχίαι with the
elaborate description which she herself has given of the novelty of her
λαμπαδηφόροι.Σ This would be in keeping with the fact that in the only
passage of this play in which λαμπτήρ occurs, 22, it is used of the light of
a beacon.

Pistoxenus-painter in Schwerin (reproduced e.g. in Pfuhl, Meisterwerke griech. Zeichnung

u. Malerei, pl. 48; H. Diepolder, Der Penthesilea-Maler, pl. 4).
1 T owe this reference to Beazley.
2 The variation in the second element of the compound is easily understood. So, e.g.,
Lycophron 734 calls the λαμπαδηφορία in Naples λαμπαδοῦχον δρόμον.

403
lines 890 f. COMMENTARY

"The light-bearing that concerned you, was related to you.’ ἀμφί with the
dative indicates in poetry, as is well known, quite generally ‘connexion with’,
‘relation to’, and the like. The root idea of lying round a central point is
quite clear. So linguistically there is no difference between this ἀμφί σοι and
the example which follows at once (893) ἀμφί σοι πάθη, although there
Agamemnon is the man whom the sufferings have befallen whereas here he
is the man whom the ‘light-bearing’ message mainly concerned.
κλαίουσα, referring to the past, cf. on 882.
891. drnpeAjrous: the group to which the word belongs is not Attic;
perhaps Ionic? This would account for its occurrence, on the one hand, in
Xenophon and in Hellenistic literature, on the other (through the channel of
late Epic?) in Tragedy and in Apollonius Rhodius. ἀτημελής is used by
Euripides fr. 184. 2 N. (μοῦσαν... χρημάτων ἀτημελῇ) in an active sense;
G. Thomson (ii. 375) maintains that ἀτημελήτους is here, too, active, but with
no good reason. The holding of the fire-brands is neglected, i.e. not put into
effect.
892 f. λεπταῖς . . . θωύσσοντος. The order and connexion of the words here
has given rise to a great deal of argument. Matters will be made easier if, to
begin with, we cling to the meaning of θωύσσειν which it has everywhere else.
It always indicates a loud shout, cry, etc. It would, therefore, be strange if it
meant here 'the light-buzzing gnat' (Conington), 'der schwáchste Ton, des
Mückenflügels leises Summen' (Wilamowitz, in any case not very exact), and
the like. No one who has been kept awake in the middle of the night by
even a single mosquito is likely to favour the ‘pianissimo’ idea. The inference
of Strepsiades (Ar. Clouds 165), σάλπιγξ ó πρωκτός ἐστιν ἄρα τῶν ἐμπίδων, sug-
gestively describes the impression made by such a sound; cf. Babrius 84. 1 f.
κώνωψ... etre ταῦτα Boufhjoas, Aristides Rhetor 46 p. 309 (ii. 403 Dind.)
τῶι Δωδωναίωι μὲν οὐκ ἂν εἰκάσαις αὐτοὺς yaÂkelun . . . ταῖς δ᾽ ἐμπίσι ταῖς ἐν τῶι.
σκότωι βομβούσαις, Tertullian adv. Marc. 1. 14 p. 308. 12 sustine . . . culicis et
tubam et lanceam, Hieronymus tract. in psalm. 91, p. 122 (culex) habet . ..
tubam vocis (both quoted in Thes. I. L. iv. 1286 f.). In the Oxford Engl. Dict.,
vol. x p. 423, I find a quotation from modern English (1900): ‘anopheles, a
mosquito that does not trumpet’. θωύσσειν is a much louder noise than could
be produced λεπταῖς ῥιπαῖς, no matter what conception Aeschylus had of the
gnat’s method of producing the sound (for modern explanations see RE xvi.
451 and Brehm, Tierleben, Volksausg. iii, znd ed., 603). This is another of
those passages in Aeschylus the understanding of which has been obscured
by the practice of taking together what the poet intended to be apprehended
piece by piece until the whole was built up. A comma after ῥιπαῖσι may
provide a useful help in reading (for the subsequent addition of the participle
cf. 876, 1503, and in general the notes on 2 and ı581). Clytemnestra is par-
ticularly anxious to leave no doubt how light was her troubled sleep; even
the hardly perceptible whir of the gnat’s wing roused her. But at the same
time the agony of such moments presents itself to her so vividly that she
re-lives them ; so she adds a mention of the horrid hum with which the insect,
backwards and forwards, but ever nearer, approaches its victim. The word-
order from λεπταῖς to ῥιπαῖσι follows the normal type (cf. on 156), in which an
adverbial expression (often a noun plus preposition, as here) together with
the main verb is inserted between an attribute and its noun. Here, as
404
 COMMENTARY line 896

elsewhere, the separation (λεπταῖς... ῥιπαῖσι) serves to emphasize the phrase

the elements of which are separated. [Cf. the Addenda.]
894. τοῦ £uveudovros χρόνου: for the general conception underlying this
expression cf. on 105 f. In particular, a passage in Pindar is appropriately
recalled here, P. 9. 23 ff. (Cyrene) τὸν δὲ σύγκοιτον γλυκὺν (any other bedfellow
is unknown to the maiden) παῦρον ἐπὶ γλεφάροις ὕπνον ἀναλίσκοισα ῥέποντα
πρὸς ἀῶ.
On πλείω τοῦ ξυνεύδ. χρόνου Paley's comment is to the point: ‘It is a well-
known peculiarity of dreams, that the clearly defined events of a day or a
week may be conjured up by the mind and pass in array before it in the space
of even a few minutes.’
895. ἀπενθήτωι φρενί was connected by the copyist who is responsible for
the punctuation in F (comma after φρενὶ, and no punctuation within the line)
and by Stanley with τλᾶσα (also later by Conington and recently by G.
Thomson) ; on the other hand Triclinius (comma after rAäo’, and no punctua-
tion after φρενὶ), Pauw and Schütz took it correctly with what follows, and I,
Hermes, lxviii, 1933, 244, ought not to have contradicted them. Hermann,
who, following Elberling, encloses ταῦτα πάντα τλᾶσα in commas, explains
appropriately: ‘Ouoniam se nunc tandem omni sollicitudine liberatam quasi
ex intimo animo dicere videri cupit, vöv sine copula in principio dicti huius
positum est, ut supra praeco dixit [508] νῦν χαῖρε μὲν χϑών.᾽ Cf. for that e.g.
Φ 160 νῦν αὖτε μαχώμεθα. For ἀπενθήτωι φρενὶ λέγοιμ᾽ ἄν cf. 5. Trach. 763 f.
ἵλεωι φρενὶ... κατηύχετο.
896. λέγοιμ᾽ ἄν: cf. on 838.
τῶν σταθμῶν κύνα. Hermann found such difficulty in the article that he
altered the text; many editors have followed him. But the function of the
article here is clear: ‘of those σταθμά of which he is the proper protector’, i.e.
in the customary terminology, possessive. Nägelsbach takes it correctly. It
is worthy of note that Hermann Sept. 573 μέγιστον Ἄργει τῶν κακῶν διδάσκαλον
took objection to the article in exactly the same way (‘valde displicet arti-
culus röv’ ; he rejected the line as spurious, but see below on 1440-3). Actually
Sept. 573 provides a particularly good parallel to the present passage: the
line forms part of an asyndetic sequence of clauses which depict a man’s
character, some of them in bold imagery ; each clause is composed of a noun
with a dependent genitive, and all the genitives, except τῶν κακῶν, are
without article. What is meant by τῶν κακῶν is presumably ‘of those evil
arts in which he himself is a master’ ; it is also possible that it means ‘of the
ills from which Argos is suffering now’ (cf. 575 τῶνδε).
σταθμῶν is often taken to mean ‘stable’ or ‘folds’ here; perhaps it would
be better to give it the wider meaning of ‘farmstead’, ‘dwelling-place’, which
includes the human habitation. A house without its guard and protector,
the house-dog, is defenceless since its security depends on him. That is the
thought, exactly parallel to the comparisons of the following lines. In a
series of images Clytemnestra depicts her husband as defence and bulwark,
as the one whose absence would bring everything to rack and ruin. She no
more says that he actually exercises the function of a house-dog than that he
is a ship’s rope, a supporting pillar, or an only child. So it was quite mistaken
of Ahrens to find a contradiction between this first phrase of a sequence of
images which is altogether typical in aim and the fact that Clytemnestra
405
line 896 COMMENTARY
describes herself (607) as δωμάτων κύνα ἐσθλὴν ἐκείνωι, πολεμίαν τοῖς Svabpocır
and is addressed by Agamemnon (914) as δωμάτων ἐμῶν φύλαξ.
897. vpórovov: the 'fore-stay', the particularly strong rope which is the
principal means of supporting the mast, binding it to the forepart of the
ship. Examples in Torr, Ancient Ships, 94 n. 202. Aeschylus is referring here
to the later type of ship with only one forestay, as against the two πρότονοι
of the Homeric craft (A. Köster, Das antike Seewesen, 167).
σωτῆρα provides the keynote to all the phrases of this enumeration, cf.
on 896.
898. στῦλος is just as much 'pillar' as 'column', cf. the dictionaries and
D. S. Robertson, Greek and Roman Architecture, 382, s.v. 'Column'.
ποδήρη : Schol. ποδῆρες κυρίως λέγεται τὸ μέχρι ποδῶν διῆκον ἱμάτιον (chiton
would have been better). It is probable that Aeschylus intended a figurative
use here (much bolder is 1594 τὰ ποδήρη). It is doubtful whether Schneidewin
(whose note was taken over by Wecklein) is right in translating ‘mit festge-
fügtem Fusse’ and comparing Pindar’s (fr. 88. 6 Schroeder) ἀδαμαντοπέδιλοι
κίονες, which uses a different image. Ahrens (580 f.) brings much etymological
learning to bear upon the explanation of ποδήρης, but takes no account of the
way in which the word is used elsewhere. It seems that here it simply
denotes the column (or pillar) ‘which reaches down to the ground’.
Both the idea implied in στέγης στῦλον and its close association with the
next item in the list are well illustrated by the passage quoted by Butler,
E. Iph. T. 57 (interpretation of the vision 50 f. μόνος δ᾽ ἐλείφθη στῦλος...
δόμων πατρώιων) στῦλοι γὰρ οἴκων παῖδές εἰσιν ἄρσενες.
μονογενὲς τέκνον πατρί. Here, too, the owrnp-idea (cf. on 897) is clear:
παῖδες γὰρ ἀνδρὶ KAnddves' σωτήριοι θανόντι (Cho. 505 f., also 507 σώιζοντες).
This idea is the stronger in the case of an only son, on whom alone depends
the preservation of the family name and the sacra familiae and also of the
κλῆρος, since πλοῦτος ὁ λαχὼν ποιμένα ἐπακτὸν ἀλλότριον θνάισκοντι στυγερώτατος.
899--902. It is.convenient to deal with the following lines before discussing
899. The most obvious difficulty lies in 902 τερπνὸν δὲ κτλ. For it is not at all
clear how such a statement (iucundum vero est etc.) could form the con-
clusion to the sequence of parallel predicative nouns? which was introduced
896 with λέγοιμ᾽ ἂν ἄνδρα τόνδε. Schütz (1811) perceived the difficulty, and to
remove it placed 902 after 903. Against that Hermann rightly observes that
φθόνος δ᾽ ἀπέστω must not be separated from τοιοῖσδε... προσφθέγμασιν. He
himself says: ‘refertur ad illa v. 895 ἀπενθήτωι φρενὶ λέγοιμ᾽ av’, 1.6. he separates
the line from the sequence of which it apparently forms the conclusion and
gives it a reference to which nothing in the wording points. Blomfield wanted
to excise the line (in this he was right as will be shown), others have trans-
ferred it in various ways. Another method of rendering 902 harmless has
been pursued by Verrall and Wilamowitz, who take the sentence as an
et cetera-formula. It may well be doubted whether the fresh start of τερπνὸν
δὲ κτλ. is a suitable form for expressing such an et cetera at the end of an
enumeration, as would e.g. kai τὰ ἄλλα πάντα (Ag. 609, S. El. 657 [Ar. Frogs
621, after a long enumeration, πάντα τἄλλα], cf. Ar. Peace 599 τἄλλα θ᾽ dado”
1 For the benefit of those who have to use the Oxford text of Aeschylus I recall Hermann's
note,
2 A similar sequence is found in Sept. 572-5 and, considerably shorter, Ag. 1233 ff,
406
 COMMENTARY lines 899-902

ἐστὶ φυτά) or an equivalent relative clause (Aesch. fr. 70 Ζεύς ἐστιν αἰθήρ, Ζεὺς
δὲ γῆ, Ζεὺς δ᾽ οὐρανός, Ζεύς τοι τὰ πάντα χῶτι τῶνδ᾽ ὑπέρτερον, 5. Oed. R.
1406 ff. πατέρας, ἀδελφούς, παῖδας, alu” ἐμφύλιον, νύμφας γυναῖκας μητέρας Te,
χὠπόσα αἴσχιστ᾽ ἐν ἀνθρώποισιν ἔργα γίγνεται). Moreover the meaning assigned
to ἅπαν by Wilamowitz and Verrall is wrong: it is to be taken closely with
τἀναγκαῖον. The γνώμη of this line is obviously connected with another
γνώμη (and may be regarded as its continuation), quoted by Paley; it is
found in different forms; one is (Euenos fr. 8 D. = Theogn. 472) πᾶν γὰρ
ἀναγκαῖον πρᾶγμ᾽ ἀνιαρὸν ἔφυ.
No less of ἃ crux is line goo. Its difficulty has been obscured by the trans-
lation which has become traditional since Stanley, diem visu pulcherrimum
post tempestatem or the like. A closer examination, however, will show that
κάλλιστον εἰσιδεῖν, taken as an attribute, would be of a completely different
class from the other attributes in this sequence. For they all indicate not
only a characteristic peculiarity of the object concerned, but one which is
definitely connected with what is here its special function, i.e. protection,
preservation, and the keeping from destruction. κάλλιστον εἰσιδεῖν, on the
other hand, is vague, and in contrast to the objectivity of the preceding
attributes its effect is almost sentimental! When, many years ago, I first
noticed the peculiar character of goo and its affinity with 902, I came to the
conclusion that κάλλιστον is not an attribute of ἦμαρ but the predicate of the
sentence, just as τερπνόν is in 902. Moreover I inferred that neither 900 nor
902 belonged to the context of this passage but originally formed part of a
series of variations on the theme κάλλιστον τὸ δεῖνα (see below). Afterwards
I saw that Headlam had reached the same conclusions. He exemplifies the
theme of both goo and 9o2 by quoting Theogn. 255 f. κάλλιστον τὸ δικαιότατον᾽
λῶιστον δ᾽ ὑγιαίνειν" πρᾶγμα δὲ τερπνότατον, τοῦ τις ἐρᾶι τὸ τυχεῖν (where the
place of κάλλιστον is taken by τερπνότατον as it is by repsvóv? in Ag. 902),
Soph. fr. 329 N. (= 356 P.) κάλλιστόν ἐστι τοὔνδικον πεφυκέναι, λῶιστον δὲ τὸ
ζῆν ἄνοσον, ἥδιστον δ᾽ ὅτωι πάρεστι λῆψις ὧν ἐρᾶι καθ᾽ ἡμέραν (beside these two
it is now possible to set, for form and subject-matter, the much older passage
Sappho fr. 27a Diehl = 5 Lobel) and Asclepiades, Anth. Pal. 5. 169 (already
quoted by Blomfield). In all these instances we have answers to the question
which, from remotest antiquity, occupied the minds of the Greeks as well as
of other peoples (cf. W. Schultz, ‘Rätsel’, RE i A. 87): ri τὸ κάλλιστον OT τὸ
ἥδιστον" and the like (cf. e.g. also Pind. fr. 89 a Schr. ri κάλλιον ἀρχομένοις . . .
t Anyone who, after considering the whole passage, has grasped this point clearly, will
have to admire Hermann's feeling for style when he says: “κάλλιστον ἦμαρ non scripsisse
videtur Aeschylus. Friget enim nomen superlativum, multoque deterius est positivo, ubi
non idonea de caussa positum est. Quare γαληνὸν ἦμαρ scripsi. His conjecture may be
dismissed at once, but it shows that he was looking for an adjective which, in sense, should
be parallel with the preceding epithets. Meineke, Philol. xix, 1863, 202, and Housman,
J. Phil. xvi, 1888, 269, agree with Hermann in his objection.
2 0. Schroeder, Commentary on Pind. P. 8. 93, points out that τὸ τερπνόν there picks up
the xaAóv of 88. van Otterlo, on p. 155 of the article quoted in the next footnote, draws
attention to Pind. fr. 221 Schr. τέρπεται δὲ καί τις ἐπ᾽ οἶδμ᾽ ἅλιον κτλ. as a characteristic
element of this ‘priamel’ and similar ones.
3 Cf. Hesiod fr. 163 f. Rz. 950 . . . ἐστ᾽ ἐν δαιτὶ... τέρπεσθαι μύθοισιν... . ἡδὺ δὲ καὶ τὸ
πυθέσθαι κτλ. ; like some of the examples quoted above, this obviously comes from a lengthy
enumeration of the type of ‘priamel’ (= pracambulum), i.e. a series of detached statements
which through contrast or comparison lead up to the idea with which the speaker is
407
lines 899-902 COMMENTARY
N... ἀεῖσαι; Bacchyl. 4. 18 ff. τί φέρτερον 7) . . . παντοδαπῶν λαγχάνειν do
μοῖραν ἐσθλῶν; A. Ag. 601 f. τί γὰρ γυναικὶ τούτου φέγγος ἥδιον δρακεῖν, Ar.
Peace 1140 οὐ γὰρ ἔσθ᾽ ἥδιον 7 τυχεῖν μὲν ἤδη ᾽σπαρμένα krÀ., Birds 785 οὐδέν
ἐστ᾽ ἄμεινον οὐδ᾽ ἥδιον 1) φῦσαι πτερά, Philetaerus fr. 6. 2, ii. 232 Kock, ἥδιστόν
ἐστιν ἀποθανεῖν βινοῦνθ᾽ ἅμα, Philemon fr. 22 K. ἤδιον οὐδὲν οὐδὲ μουσικώτερον
ἔστ᾽ ἢ δύνασθαι λοιδορούμενον φέρειν). Of the great number of passages which
are relevant here, I will only mention E. Heraclid. 892 ff. ἐμοὶ χορὸς μὲν ἡδύς
. . τερπνὸν (for this word see above) δέ τι καὶ φίλων dp’ εὐτυχίαν ἰδέσθαι and
Eur. fr. 316 N. γύναι, καλὸν μὲν φέγγος ἡλίου τόδε, καλὸν δὲ πόντου χεῦμ᾽ ἰδεῖν
εὐήνεμον.. . . ἀλλ᾽ οὐδὲν οὕτω λαμπρὸν οὐδ᾽ ἰδεῖν καλὸν ws . . . παίδων νεογνῶν
ἐν δόμοις ἰδεῖν θάλος, on account of the appearance of ἐδεῖν, For this was
clearly a favourite word in this special type of ‘priamel’:' Ag. 900 shows the
same, as does also Ag. 602 (8paxeiv), similarly the epigram of Asclepiades
already mentioned, Anth. Pal. 5. 169, ἡδὺ δὲ ναύταις ἐκ χειμῶνος ἰδεῖν εἰαρινὸν
Zrébavov,* and the beginning of the second book of Lucretius, Suave (ἡδύ)
mari magno turbantibus aequora ventis e terra magnum alterius spectare (here
we have one equivalent to ἰδεῖν, and another in 1. 4 cernere suave est) laborem
... Suave eliam belli certamina magna tueri (N.B.) per campos tnstructa, tua
sine parte pericli ; sed nihil dulcius est etc.
The conclusion, then, is this: lines 900 and 9o2 do not belong here. Far
from having been composed to round off a sequence, they must, on the
contrary, have belonged originally to a series of introductory parallels, a
‘priamel’, which leads up to the instance in hand, just as in the passages of
Sophocles and Euripides quoted above. 9oo and 9o2 were presumably written
in the margin? by a reader, who was possibly tempted by a certain similarity
in subject between 899 and goo.* So far I agree with Headlam. But however
preferable a simple cut maybe to a complicated hypothesis, it seems to me
impossible to regard gor in the same light as 900 and 902. The purely nominal
phrase of goı cannot, as it stands, be associated with κάλλιστόν (ἐστιν) εἰσιδεῖν
and τερπνόν (ἐστιν) ἐκφυγεῖν. If we wanted it to belong to the same context,
it would be necessary to assume a lacuna between goo and gor, in which a
καλὸν δέ or the like had disappeared. On the other hand, there is an exact
correspondence between gor and the images enumerated in 897-9, not only
in form, but also, it appears, in matter: for the traveller in a thirsty land it is

primarily concerned, e.g. *Of all things A is the most valuable ; many regard B as the greatest
boon; to a man in such and such a position C is more precious than anything else; but in
my case D is the most desirable’ (cf. W. Schultz, op. cit. 67, F. Dornseiff, Pindars Stil,
1921, 97 ff.; id., Die archaische Myihenerzählung, 1933, 3f., 18 1.; Dodds on E. Bacch.
9o2-11, van Otterlo, Mnemos. S. ITI, viii, 1940, 145 ff.).
1 Cf. the preceding footnote.
2 Wilamowitz was presumably influenced by the recollection of this fine passage (cf. also
Theocr. 18. 27) in his translation of Ag. 900 : ‘der Tag der Frühlingssonne nach dem Winter-
sturm'. But if κάλλιστον ἦμαρ was meant to denote that, it would, all other considerations
apart, be intolerably vague.
3 Cf. on 525 ff.
4 It is well known, particularly from the text of Euripides, that not infrequently maxims
which do not belong to the original have been jotted down beside lines with which they
have a rough (often quite superficial) similarity ; a reference to the remarks of Wilamowitz,
Hermes, xl, 1905, 134 and of Jachmann, Nachr. Gött. Ges., Phil.-hist. KL, Fachgr. 1, N.F. i
(1936), 140, and to my article in Eranos, xliv, 1946, 81 ff., must suffice here. Cf. also on
834-7 (p. 385).
408
 COMMENTARY lines 899-902

not a case of something more or less pleasant, but of preserving life itself.
Of course, this could have been added to the lines of Aeschylus from some
other source, particularly as this τόπος appears to belong to the usual stock
of such sequences: the epigram of Asclepiades, previously referred to, begins
ἡδὺ θέρους διψῶντι χιὼν ποτόν (where only the particular phrase is more
choice than Ag. gor). But it is at least as possible that gor originally formed
a part of Clytemnestra’s predications and then, when the two spurious lines
crept in from the margin, got sandwiched between them. If we are to accept
this, I do not know how to determine the proper place of gor with certainty.
It may originally have followed 899 or 898; however, there are other possi-
bilities as well.
There remains to be said a word about 900. After the grammatical structure
has been made out and κάλλιστον given back its function as predicate of the
sentence, the meaning is quite clear (Headlam is right) : "The finest thing is
the sight of day (morning) after the storm.” The reference is patently to the
experience of the seafarer ; in itself significant, its significance is increased by
the thought of how often the storm abates at sunrise. ε 390 ff. ἀλλ᾽ ὅτε δὴ
τρίτον ἦμαρ ἐνπλόκαμος τέλεσ᾽ "Has, καὶ τότ᾽ ἔπειτ᾽ ἄνεμος μὲν ἐπαύσατο ἠδὲ
γαλήνη ἔπλετο νηνεμίη, Bacchyl. 13 (12). 128 ff. λῆξεν δὲ (scil. Βορέας) σὺν
φαεσιμβρότωι Aot, στόρεσεν δέ τε πόντον οὐρία, chorus in Eur. Phaethon
(Berliner Klassskertexte v. 2, p. 82, v. Arnim, Suppl. Eur. 70 £), where at
sunrise the mariners ask the morning breeze for a homeward journey
ἀκύμονι πομπᾶι σιγώντων ἀνέμων.
Finally a few remarks about 899. Here Blomfield found difficulty in the
καί and suggested γῆν ἐκφανεῖσαν Or γαῖαν φανεῖσαν or alternatively the ex-
cision of the line. Klausen and Schneidewin thought that καί marked the
beginning of a new sequence, ‘because from here onwards come phrases
signifying the sudden appearance of good fortune after a long period of
distress’. It has been shown above that there is no new sequence here. But
leaving that aside, «ai is quite unsuited for the function they assign to it, ‘for
it rather connects than divides’ (Ahrens, 581). Headlam defends καί with the
explanation povoyevés τέκνον πατρὶ (παρ᾽ ἐλπίδα φανὲν) καὶ γῆν φανεῖσαν
ναυτίλοις παρ᾽ ἐλπίδα. To say nothing of the fact that the arbitrary embellish-
ment of the povoyevés-phrase introduces an idea which is quite redundant
here, the case for such an ἀπὸ κοινοῦ breaks down for the reason that in the
whole sequence the phrases are simply strung together without connexion :
moreover, the case is exactly the same in Set. 572 ff. So it is very probable
that Blomfield with γαῖαν has restored the original, as in Eum. 755 Dindorf’s
γαίας for καὶ γῆς has been accepted by many editors (for other passages where
γαῖα etc. of the MSS has been displaced by γῇ etc. see Denniston on E. El.
678). For vavriAoıs in 899 Blomfield compares E. Andr. 891 f. ὦ ναυτίλοισι
χείματος λιμὴν φανεὶς Ayauéuvovos παῖ. But as far as Ag. 899 is concerned, I
consider it very probable that, in addition to the general τόπος, there is in
particular the influence of the ‘wonderful simile’ (Von der Mühll, RE Suppl.
vii. 763. 53) V 233 ff. ws δ᾽ ὅτ᾽ ἂν domdoros γῆ νηχομένοισι φανήηι, ὧν τε Ποσει-
δάων εὐεργέα vij ἐνὶ πόντωι paiom . . . ἀσπάσιοι δ᾽ ἐπέβαν γαίης, κακότητα
φυγόντες" Ws ἄρα τῆι ἀσπαστὸς ἔην πόσις εἰσοροώσηι. Penelope’s true love
becomes a lie in the mouth of Clytemnestra.
Viewed as a whole, the sequence of predications shows great firmness of
409
lines 899-902 COMMENTARY
outline. The Tuscan lovesong (‘L’é rivenuto il fior di primavera, l'é ritornata
la verdura al prato' etc.) which Wilamowitz, Interpr. 172, compares, is in
fact only useful to point a contrast! and quite unsuited to 'show from what
sphere the poet took his colours for the picture of hypocrisy'. It is far from
the mark to say that Clytemnestra 'conjures up these many images of the
preciousness of his appearance'. It is not a question here of precious objects,
in the sense of that love-poem, but of a much more limited idea, yet at the
same time one that is singularly appropriate to the hero and ruler, namely the
idea of protection and preservation. Even in the exaggeration of her clever
hypocrisy there speaks from Clytemnestra still the voice of a strong and
virile mind, not the cooing of an enamoured female. Therefore the compari-
son drawn later (Hermes, lxii, 1927, 287 f.) by Wilamowitz with the ancient
Egyptian song to Sesostris III (A. Erman, Die Literatur der Aegypter, 180 1.)
is wholly apposite. The relevant section of that hymn runs: 'How great (i.e.
what a bestower of blessings?) is the lord to his city: he alone is a million . . .
he is like a dike, which holds off the river in flood . . . like a cool house which
lets a man sleep on into the daytime . . . a bulwark which protects the fearful
from his enemy . . . the shadow of the flooding season to bring coolness in
summer...a warm, dry, corner in the winter... a mountain that holds off
the storm, at the time of heaven’s fury . . . he is like Sechmet [a goddess of
war] towards the enemies who cross his borders.' I have not omitted a single
comparison, in order to show that every one is concerned with protection and
defence. This makes the similarity to the Agamemnon passage much greater
even than Wilamowitz noticed. (Two parts of the sequence quoted appear in
Clytemnestra's later speech, cf. on 970 f.) However, there is no need to
follow W. Kranz, Stasimon, 102, 294, in supposing that Aeschylus was in-
debted to some definite Egyptian panegyrics. In subject-matter there is no
indication of any background other than Greek, and the form of such a
sequence of predications, as has already been shown, is used by Aeschylus
elsewhere (Sept. 572 ff.). The close resemblance in thought is explained by the
quite natural fact that for the Egyptians also, at least for the composer of the
hymn, the ruler was primarily φύλαξ, just as in the case of many Indo-
Germanic peoples, cf. on 1452. The lack of discretion? in this string of eulogies
is exactly appropriate to the manner of speech of the lady who doth protest
too much.
903. τοιοῖσδε marks the conclusion of the whole preceding passage, just as
in 312, 348, 874, and elsewhere. As regards the contents the three short
clauses (down to 905 ἠνειχόμεσθα) belong closely together, then with νῦν δέ
something new begins. The lines must be punctuated accordingly, with a
full stop before νῦν and a weaker punctuation mark at the end of 903.

1 Asa true parallel to the Tuscan canzone one might quote not Clytemnestra’s speech
but the beginning of Theocritus 12 (Airns).
? 'This is Erman's paraphrase and particularly noteworthy in our present context.
3 It so annoyed W. Dindorf that he (Aesch., editio quinta, praef., p. xcii), by one of
the sweeping strokes characteristic of his rash criticism, bracketed ll. 895-902 altogether
(the reader is left wondering what may be the meaning of 903 if it follows immediately
after 894). But it is to Dindorf's credit (he was often hasty but hardly ever dull) that he
sensed in these lines something very different from the general discretion of Aeschylus
and the normal standards of Hellenic self-restraint (he says "Theodorum Prodromum
similesque poetas audire videmur").
410
 COMMENTARY line 905
904. φθόνος δ᾽ ἀπέστω, directly connected with the panegyric. In this way
Pindar concludes his detailed glorification of Corinth with the prayer (Ol.
13. 25) ἀφθόνητος ἔπεσσιν γένοιο χρόνον ἅπαντα, Ζεῦ πάτερ, and P. το. 20 after
the praise of the victor and his father he adds μὴ φθονεραῖς ἐκ θεῶν μετατρο-
mis ἐπικύρσαιεν. Elsewhere, too, in Pindar (cf. P. 8. 71, Isthm. 7. 39 6 δ᾽
ἀθανάτων μὴ θρασσέτω φθόνος) ‘bold, extolling words demand an “absit
omen!" ' (Schadewaldt, ‘Aufbau des Pindarischen Epinikion’, Schriften der
Königsberger gel. Ges. 1928, 288) ; cf. for that idea Pohlenz, Herodot (1937), 111.
Cf. also E. Heraclid. 202 καὶ γὰρ οὖν ἐπίφθονον λίαν ἐπαινεῖν ἐστι, Titinius com.
110 Ribb. pol tu ad laudem addito praefiscini.' Headlam and G. Thomson
illustrate φθόνος ἀπέστω with a few similar expressions.
Most likely it is exactly φθόνος which Clytemnestra, for the furthering of
her plans, wishes to bring down upon him by her strong words of praise, just
as later by the spreading of the carpet (this is the view of Sewell,? Kennedy,
J-Phil. vii, 1877, 14, and A. Platt in the note on his translation; cf. also
Neustadt, Hermes, lxiv, 1929, 264). Aeschylus has chosen his words so as to
make us realize that from now on the $0óvos-motif dominates the scene: 921,
939, 947 (cf. Pohlenz, Herodot 111).
904 f. πολλὰ . . . ἠνειχόμεσθα perhaps reflects a phrase of everyday lan-
guage, cf. Ar. Peace 347 πολλὰ yàp ἠνεσχόμην πράγματα κτλ.
905. φίλον κάρα. The whole of the first main section of the speech (down to
876) was addressed only to the old men, and although she turns 877 ff. to
Agamemnon, she does not use any form of address, similarly 893. Then in the
sequence of predications (896-903) she speaks of him again in the third person
(cf. 860, 867). Here for the first time, when she wants to flatter him into
decisive defeat, Clytemnestra uses a full formula of address (cf. Th. Wendel,
‘Die Gesprächsanrede’, Tübinger Beiträge zur Altertumsw. 6. Heft, 1929, 117)
and calls him φίλον κάρα. This form of affectionate address, the prototype of
which is as early as Homer, occurs in Aeschylus only here (Wendel, p. 33),
whereas in Sophocles and Euripides it is to be found a number of times. Of
all the means of deception employed by Clytemnestra this expression of
deepening tenderness, carefully saved up, is perhaps the most sinister.
μοι is taken by some commentators as possessive (cf. on this in general
Wilamowitz on E. Her. 626; Wackernagel, Syntax, ti. 77), going with φίλον
κάρα, thus Stanley (‘carum mihi caput’) and e.g., among the more recent
scholars, Headlam (‘Now, dear my lord, step from that wagon’). But in that
case the pronoun would surely be put between the two words, cf., e.g., E. Hec.
409 (likewise Ion 1324, 1443) ὦ φίλη μοι μῆτερ, Iph. T. 795 ὦ φιλτάτη μοι σύγγονε,
and the like. In this passage μοι goes with the whole sentence as ethic dative,
as E 249 ἴ. μηδέ μοι οὕτω θῦνε διὰ προμάχων and in other examples of the kind
discussed by Kühner-Gerth, i. 423 d.

t The fear of provoking the βάσκανος by unconcealed praise is understandable and

widespread. 'Tabaqui knew as well as anyone else that there is nothing so unlucky as to
compliment children to their faces; and it pleased him to see Mother and Father Wolf
look uncomfortable’ (R. Kipling, The Jungle Book).
2 In the passage quoted in the note on 1. 811 Sewell sums up the character of Agamemnon
in these words: ‘the humility with which he bears his victory ; even his final concession to
her request of treading on her purple tapestries (a request not made without the hope of
inducing him to commit what might seem a crime, as indicating pride, and so expose him
to an “envious eye” from heaven, and facilitate the execution of her vengeance)’.
4II
line 906 COMMENTARY
906. ἔκβαιν᾽ ἀπήνης τῆσδε. He does not alight till 946 ff.; the interval is
filled by the contest which ends with Agamemnon’s defeat.
907. τὸν σὸν πόδα . . . Ἰλίου πορθήτορα. In the art of the ancient Orient, of
Egypt as well as of western Asia, representations of the enemy lying beneath
the foot of the victor are quite customary ; but they are foreign to Greek art
(cf. Rodenwaldt, Arch. Jahrb. xxxvii, 1922, 22 ff., and also B. Schweitzer,
Arch. Jahrb. xlvi, x931, 216). For Egypt cf. further e.g. the hymn to Thut-
mosis III, composed about 1470 B.c. (Erman, D. Literatur der Aegypier,
318 ff.) : ‘I make your adversaries to fall beneath the soles of your feet, that
you tread down the . . . of the rebels. . . . Tam come that you may tread down
the princes of Palestine; I spread them beneath your feet' etc. Cf. LXX
Joshua xo. 24 προπορεύεσθε καὶ ἐπίθετε τοὺς πόδας ὑμῶν ἐπὶ τοὺς τραχήλους
αὐτῶν, Psalm 109. 1 ἕως ἂν θῶ τοὺς ἐχθρούς σον ὑποπόδιον τῶν ποδῶν σου. It
may perhaps be supposed that Aeschylus makes conscious use of this theme
as a prelude to the acts of homage suggested in the following lines ; for upon
the ‘barbaric’ character of these acts special emphasis is laid. Works of art
from Egypt or western Asia, knowledge of which is in a number of places
evident in Aeschylus (cf. Kranz, Stasimon, 73 f. and 83 at the bottom) may
have supplied him and his audience with a picture of that brutal gesture of
victory. S. El. 456 ('Opéarqv) ἐχθροῖσιν αὐτοῦ ζῶντ᾽ ἐπεμβῆναι ποδί does not
prove anything with regard to normal Greek feeling: it is the prayer of ἃ
woman consumed by thirst for vengeance.
908. τέλος : ‘munus’, cf. on 105; associatedin Eum. 743 with the same verb
as here. What whim or oversight caused Triclinius to write τάδε here, I
cannot say.
909. Elmsley’s passing reference (on E. Med. 1209 f., n., where he quotes the
line and only adds to στρωννύναι ‘an στορνύναι ’) was rightly brought to notice
by Hermann. As far as our present knowledge extends, the Hellenistic
στρώννυμι is foreign to ancient Attic.! The example in L-S (cf. also Wila-
mowitz on Ag. 909), Com. adesp. 1211 K. (= Trag. adesp. 90, p. 857 N.) ap.
Athen. 2. 48a, should be cancelled, for there στρώννυ is a conjecture of
Musurus: of the two fifteenth-century MSS of the Epitome of Athenaeus the
one has στρωμνὰς and the other orp@\,* but there is no doubt that the reading
of the Epitome was στρῶσον, for this is what Eustathius, in all probability
himself the author of the Epitome,? quotes from it* in his commentary on the
Odyssey, p. 1887. 46. The incursion of the kow1-forms of the present stem into
late MSS is shown by S. Trach. 9o2, where the correct στορνύνθ᾽ is preserved
in L and A, whereas later MSS have altered it to στρωννύνθ᾽ ; Wilamowitz
notes that in the ὑπόθεσις of the Persae (p. 131. 6 Wil, Scholia in Aesch.
Pers. ed. Daehnhardt, p. 2 f.) the reading in M is στορνύς, but in all the other
MSS erpovws.
πετάσμασιν: in this sense in older Greek, apparently, only here. In
everyday speech παραπετάσματα appears to have been the usual word, thus
1 Schwyzer, Griech. Gramm. i. 697 ὃ, ‘att. στρώννῦμι᾽ ought to be modified accordingly.
2 στρῶσον is described by Peppink, Observ. in Athen. (Leyden 1936), 3, as the solution of
the abbreviation in E, whereas P. Maas, Byz. Zeitschr. xxxvii, 1937, 186, calls it a conjecture
of Peppink.
3 Cf. P. Maas, Byz. Zeitschr. xxxvi, 1936, 30.
4 Even Nauck does not mention στρῶσον although in his notes on the fragment he refers
to Eustathius.
412
 COMMENTARY line 913

Herodotus, Aristophanes, etc., in Hellenistic Greek predominantly περιστρώ-

ματα (so, too, Plautus and Cicero). The same objects are called later in 921,
960, 963 quite generally εἵματα, in 949 ὑφαί, more specially in 923 ποικίλα κάλλη
(in 926 and 936 simply ποικίλα), in 946 ἁλουργῆ, in 957 πορφύραι. They are,
therefore, carpet-like fabrics, patterned and made with a lavish use of purple.
‘These costly stuffs, of colourful patterns, adorned with figures, were quite
generally called βαρβάρων ὑφάσματα [E. Jon 1159], their home is ‘essentially
the classic land of carpet-weaving and of weaving in colours generally:
Persia,’ and that in its wider sense to include Asia Minor and Syria’ (Fr. v.
Lorentz, Rém. Mitt. lii, 1937, 210). For the oriental custom of spreading
garments on the path of a king cf. on 918-21. The connexion, therefore, with
the barbaric, or more particularly Persian, acts of homage subsequently
rejected by Agamemnon is quite close.
911. He is to be shown the way into his house by Aixy. Here there has been
no preceding strife, and to that extent the function of ΖΔέκη is different from
what she does in leading back a claimant Sept. 646f. and Ag. 1607, but
Agamemnon, too, has a legal claim to return to his own property. This is
the innocent meaning of the sentence as it is meant for the king. The business
of Δίκη, as Clytemnestra means it herself, is not difficult for the audience to
guess. Cf. on 1378. Here, with her wish framed like a prayer, the woman
speaks in fact with sinister ambiguity, just as in the prayer at the end of the
scene (973 f.); she is like the man who immediately after praying with all
propriety to the gods labra movet metuens audiri. The concluding sentence
(gı2 f.) is of the same kind and in it δικαίως is a significant repetition. The
fact that Clytemnestra entrusts her true desire to the κρείττονες, Δίκη and
(973) Zeus, shortly before the deed, is of the highest importance for the
success of her plan. It is on that account that she here speaks to them as she
does. If, on the other hand, she had allowed herself allusions to Aegisthus in
the speech by which she hoped to ensnare Agamemnon, this would have been
pointless and unwise.
913. εἱμαρμένα. Karsten and Meineke (Philol. xix, 1863, 202) rightly raised
objection to this word. ‘Quae fato constituta sunt non pendent a mortalium
arbitrio nec penes homines est ea recte vel prave ordinare (δικαίως τιθέναι).
Porro ξὺν θεοῖς non potuit iungi cum εἱμαρμένα, cum quo tamen proxime
cohaeret: perfici ξὺν θεοῖς recte dicuntur hominum consilia, non vero fati
decreta; potius dixisset πρὸς θεῶν εἱμαρμένα, ut Soph. Tr. 168 [there it is a
question, however, of a decree of Fate in the full sense]’ (Karsten). Meineke's
σὺν θεοῖσιν ἄρμενα has met with much approval. But the fact that ἄρμενος
never seems to occur in Tragedy may make us pause. And it is certain that
the passage when altered in this way cannot be taken as Meineke takes it:
'all the rest we will set in order with the help of the gods', similarly Wila-
mowitz: ‘Das weitre wird mein Eifer, den kein Schlaf bezwingt, mit Gottes
Hilfe fügen, wie es ihm gebührt.’ This is taking ἄρμενα predicatively with
θήσει. But there is no room here for that, for θήσει δικαίως clearly go together:
‘will so arrange it as is just’. For similar association of τιθέναι with an adverb?
! It is a nice point that Themistocles (Plut. Them. 29. 4 = Reg. apophth., p. 185 e),
just when he was in conversation with the Persian king, is supposed to have produced the
comparison τὸν λόγον ἐοικέναι τοῦ ἀνθρώπον rois ποικίλοις στρώμασιν κτλ.
2 ] am leaving aside as much too uncertain the supposition of Schütz and Wilamowitz

413
 line 913 COMMENTARY
cf. Pers. 283 (text uncertain in detail), Ag. 1601, 1673, on which Wilamowitz
refers to Eur. Antiope B τὸ (v. Arnim, Suppl. Eur. 19) ἡμεῖς καὶ σὺ θήσομεν
καλῶς, cf. further e.g. S. Trach. 26 τέλος δ᾽ ἔθηκε Ζεὺς ἀγώνιος καλῶς, E. Hifp.
521 ταῦτ᾽ ἐγὼ θήσω καλῶς, Hec. 875 πάντ᾽ ἐγὼ θήσω καλῶς, El. 648 καὶ μὴν
ἐκεῖνά γ᾽ ἡ τύχη θήσει καλῶς, Iph. A. 401 τἄμ᾽ ἐγὼ θήσω καλῶς, Eur. fr. 287 N.
ἀλλ᾽ οὐντυγχάνων τὰ πράγματ᾽ ὀρθῶς ἣν τιθῆι, πράσσει καλῶς, similarly with the
Middle, εὖ (καλῶς) θέσθαι (cf. above on 32): as early as Hesiod, Erga 23, in
Tragedy e.g. S. Oed. R. 633, El. 1434, Soph. fr. 324 N. (= 350 P.), E. Hipp. 709,
Her. 605, 938, Iph. T. 1003, Bacch. 49, Iph. A. 672," cf. also E. Andr. 378 f.
εἰ μὴ θήσομαι τἄμ᾽ ὡς ἄριστα.2 The monstrous σὺν θεῶν εἱμαρμένηι (Margoliouth),
the unnatural punctuation of Verrall, and other wild ventures need not be
discussed. The function of σὺν θεοῖς in the expression of a wish or confident
expectation at the end of a speech had already been grasped by Stanley, who
pointed to Cho. 782 γένοιτο δ᾽ ὡς ἄριστα σὺν θεῶν δόσει ; and indeed, something
like dis bene suvantibus seems needed here. Cf. the concluding words Sept. 450
Ἀρτέμιδος εὐνοίαισι σύν 7’ ἄλλοις θεοῖς. But I do not know how to restore the
last four syllables of the line. I did think of the possibility of σὺν θεοῖς δ᾽
εἰρήσεται (E. Med. 625, Ar. Plut. 114 ξὺν θεῶι δ᾽ εἰρήσεται, Hdt. 1. 86. 3 τῶι δὲ
Κροίσωι.... εἰσελθεῖν... τὸ τοῦ Σόλωνος, ὥς οἱ ein σὺν θεῶι εἰρημένον), but it is
too far removed from the MS reading.
914. Λήδας (for the use of this form in Attic see W. Schulze, Gótt. gel. Anz.
1896, 245) γένεθλον: respectful form of address in solemn style. Verrall,
with his flair for the piquant, discerns here a venomous allusion, in spite of
E. Iph. A. 116, 686, 1106, and in spite of the fact that Pindar calls the Dioscuri,
for whom he has the highest respect, “ήδας παῖδας ; so, too, the ‘Homeric’
hymn to the Dioscuri (33. 2) Τυνδαρίδας “ήδης καλλισφύρου ἀγλαὰ τέκνα (cf.
now, also, Alcaeus fr. 78. 2 D.).
'δωμάτων ἐμῶν φύλαξ would have made another woman wince’ says
Headlam. Penelope, however, reflects (r 525 f.): ἠὲ μένω παρὰ παιδὲ kai
ἔμπεδα πάντα φυλάσσω (cf. A 178), κτῆσιν ἐμὴν Suwids τε καὶ ὑψερεφὲς μέγα
- δῶμα krÀ Agamemnon here refers to the account of Clytemnestra’s func-
tions during his absence, which she had sent him by the Herald (607 ff.).
915. Mild banter, but not at all unkind, let alone irritable.* It is the harmless
jest which takes the edge off the remark. The king is, at least up till now,
completely composed, he speaks with the gracious dignity ofa great gentleman.
916. μακρὰν γὰρ ἐξέτεινας : cf. E. Med. 1351 μακρὰν dv ἐξέτεινα τοῖσδ᾽ ἐναντίον
λόγοισιν, A. Ag. 1296 μακρὰν ἔτεινας, S. Aj. 1040 μὴ τεῖνε μακράν, E. Iph. A.
that in the difficult passage A. Suppl. 86 we should interpret the letters of the MS as εἰ
dein... εὖ or read εὖ θείη... et.
1 Or. 512 καλῶς ἔθεντο ταῦτα (they established this νόμος, this custom) πατέρες of πάλαι is
rather different.
2 Similar to the θεῖναι (θέσθαι) καλῶς is E. El. 75 f. εἰσιόντι δ᾽ ἐργάτηι θύραθεν ἡδὺ τἄνδον
εὑρίσκειν καλῶς.
3 In the so-called ‘Laudatio Turiae’ (Dessau, Inscr. Lal. sel. 8393; Bruns, Fontes iuris,
7th ed., 126) of the time of Augustus the husband, addressing his dead wife, says (i. 38 f.):
officia ita partiti sumus, ut ego tutelam tuae fortunae gererem, lu meae custodiam sustineres
(the latter clause = δωμάτων ἐμῶν φύλαξ). This division of responsibility is in agreement
with a fundamental principle of Roman life, see Mommsen, Gesammelte Schriften, i. 415.
4 W. Sewell (see above, p. 372 n. 4) is right: ‘the cool and quiet distrust with which
he listens to the elaborate overstrained professions of Clytemnestra and rebukes her with
gentle irony’.

414
 COMMENTARY line 916

420 ἀλλ᾽ ὡς μακρὰν ἔτεινον. It appears that we have before us in all these
passages a phrase from everyday language : it is slightly varied e.g. Hdt.7. 5x. 1
πλεῦνα λόγον ἐκτεῖναι, E. Hec. 1177 ὡς δὲ μὴ μακροὺς τείνω λόγους, Plat. Prot.
361a (quoted by G. Thomson) μακρὸν Aóyov . . . ἀπετείναμεν, Rep. 605d
(Homer or a tragic poet is portraying a hero) ἐν πένθει ὄντα καὶ μακρὰν ῥῆσιν
ἀποτεΐνοντα ἐν τοῖς óBvppots.!
There is no noun to be supplied with μακράν. This negative conclusion was
reached by Lobeck, Paralip. 363, for a whole list of fixed phrases formed with
a feminine adjective without a noun; the positive statement of Wilamowitz
on E. Her. 681? is more appropriate ; he speaks of the ‘extensive use in Greek
of the feminine to denote indefinite abstracts’.” Of the examples from
Aeschylus which he quotes, one, like Ag. 916, refers to a preceding speech:
Eum. 638 ταύτην τοιαύτην εἶπον. This is of particular importance, for Wila-
mowitz’s treatment implies the correct interpretation of the sentence, which
has been almost universally misunderstood and made the subject of reckless
conjecture on the part of many critics. Unfortunately almost all later
editors and translators of the Eumenides have failed to note the remark of
Wilamowitz and his translation.* Cf. on 1386.
1 For this use of ἐκτείνω Blomfield rightly quotes Ath. 13. 573 b, but here we must first
of all restore to Aeschylus his own property, filched from him in the 19th century. Myrtilos,
the speaker in Athenaeus, says: καταλέξω δέ σοι, Κύνουλκε, ᾿Ιωνικήν τινα ῥῆσιν ἐκτείνας κατὰ
τὸν Αἰσχύλου... περὶ ἑταιρῶν κτλ. Every reader who is not prejudiced will take the words
᾿Ιωνικήν . . . ἐκτείνας as a quotation from Aeschylus, and so they appeared (with the
omission of τινα) in Stanley’s collection of the fragments. To be sure, he did not understand
the expression, as his note, reproduced in Butler (Aeschylus, Tom. viii, p. 121 on fr. 75),
shows: 'Ionicum verbum, pro dissoluto ac molli extulit Hesychius. ’Iwvxdv- τρυφερόν᾽, etc.
Butler himself (p. 238) only remarks that the words can be scanned in different ways.
Schweigháuser, Animadv. in Ath. vii (1805), 103, hit upon the unfortunate idea that only
ἐκτείνας Was borrowed from Aeschylus. This was taken over by Meineke, who added his
assumption that the title of the play which had dropped out after xarà τὸν Αἰσχύλου was
Ayauéuvova, so that the allusion was to Ag. 829, 916. In this way the fate of the quotation
was sealed, for Hermann (cf. his note on Ag. 796 Herm.) was convinced by Meineke, and
so he (or M. Haupt) excluded it from the collection of fragments (Aesch. i, 2nd ed.,
p. 412). He was followed by Nauck (who, Praef. p. ix sq., attacks Dindorf for accepting
the fragment). Wecklein was more cautious (see fr. 477 in his edition). It is certain that
τινα is added by Athenaeus for his own purpose, but just as certain that ᾿Ιωνικήν | ῥῆσιν...
ἐκτείνας belongs to Aeschylus. As a comment, Schol. Pind. Isthm. 6. 87a should suffice:
μακρολόγοι μὲν οὖν of “Iwves, σύντομοι δὲ οὐ μόνον Λάκωνες, ἀλλὰ καὶ ᾿Αργεῖοι (there follows
Soph. fr. 424 N. = 462 P.). Like Pindar and Sophocles Aeschylus (Suppl. 274) speaks of
the Argives’ propensity to brevity of speech ; and in the same context (273) he says μακράν
γε μὲν δὴ ῥῆσιν οὐ στέργει πόλις, 1.6. has the same use of ῥῆσιν as in the passage quoted by
Athenaeus. The true sons of the Peloponnese are men of few words, the more luxurious
Ionians of many. ‘In Ionian fashion you have made an endless speech and . . .' says some-
one reproachfully. Athenaeus, wittily enough, gives to ᾿Ζωνικήν its other well-known
meaning appropriate to his topic, the hetaerae.
2 One of his examples Plat. Euthyd. 273b ἄλλην καὶ ἄλλην ἀποβλέποντε eis ἡμᾶς may be
supplemented by Pind. Ol. 7. 82 ἐστεφανώσατο dis, kAewài τ᾽ ἐν ᾿Ισθμῶι τετράκις εὐτυχέων,
Νεμέαι τ᾽ ἄλλαν ἐπ᾽ ἄλλαι, where even so good an authority on syntax as Gildersleeve explains
‘the ellipsis of νίκαν is not violent’.
3 For ἃ discussion of the principle involved here cf. Joh. Lohmann, Genus und Sexus
(Göttingen 1932), 17. On the prepositional phrases ἐξ ὑστέρας, διὰ κενῆς, dm’ ἴσης, and the
like cf. besides Lobeck, loc. cit., Stein on Hdt. 1. 108. 16, the commentaries on Thuc. 1.
14. 3; 4. 126. 4; Kaibel on S. El. 1062, Rehdantz-Blass, Demosthenes’ Philippische Reden,
il. 2, 4th ed. (Indexes), p. 68 f.
4 Mazon, however, translates correctly: ‘J’en ai dit ce que j’ai dit.’ Some idea of the
415
lines 916 f. COMMENTARY
916 f. ἀλλ᾽ ἐναισίμως (cf. on 775) . . . γέρας. The scholiast (Σχολ. aA.)
explains: παρ᾽ ἄλλων... kal μὴ παρὰ τῶν οἰκείων 7) παρ᾽ ἑαυτῶν. Agamemnon
clearly means the former. Clytemnestra is a member of his household; and
it is not right for her, being in this position, to praise the master of the house.
Cf., e.g., Pindar fr. 181 Schr. 6 γὰρ ἐξ οἴκου ποτὶ μῶμον ἔπαινος κίρναται. Aga-
memnon couches his refusal in the form of a maxim. Here, too, he is not
caustic, but calm and dignified. We may well suppose that Clytemnestra’s
exaggerations were displeasing to him, but of this he says nothing.
917. ἔρχεσθαι γέρας: a Homeric phrase: A 120 ö μοι γέρας ἔρχεται ἄλληι,
with an entirely different meaning (van Heusde).
918-21. In form the three imperatives are simply on a level, without any
particularly close connexion of the separate clauses (as e.g. by μήτε. . .
pare...) being indicated: ‘do not...nor...nor...’. Closer observation
shows that, on the one hand, a parallelism between a and ὁ is intended (‘do
not pamper me like a woman’: ‘do not prostrate yourself in homage before
me, as if I were a barbarian’, for details see below), on the other hand there
is a parallelism between b and c. The connexion between the last two clauses
is probably a case of paratactic comparison (cf. on 76 ff.): ‘just as I should
have to refuse homage paid to me, as to a barbarian, with acclamation and
obeisance, so, too, I should dislike the strewing of garments.’ The interpreta-
tion of Schneidewin seems to me impossible. He thinks that Agamemnon
with the words μηδὲ βαρβάρου φωτὸς δίκην... προσχάνηις ἐμοί is rejecting a
humble gesture directed towards him while he is sitting in the chariot,
906 f. ; ‘the actor [who is playing the part of Clytemnestra] must be thought of
as speaking these lines bowing down and bent over almost to the ground.’
But the words μηδὲ... προσχάνηις ἐμοί are hardly suitable for expressing
reproach for a gesture already past; most important, however, is the fact
that the whole of Agamemnon’s speech, after a short introduction, is con-
cerned with parrying Clytemnestra’s intentions for the future and not with
criticism of what is past. G. Thomson tries a different explanation: 'pre-
sumably the δμωιαΐ are now prostrating themselves.’ But, quite apart from
the unseemliness of the king’s taking notice, in the middle of his speech, of
the conduct of the maidservants, does that suggestion really explain the
words he addresses to Clytemnestra? On the other hand, if it is correct to
suppose that the clause has a comparative force, then it is used by the poet
to brand the strewing of fine draperies on the ground as a barbaric, Oriental
ritual, as something that must be just as repugnant to a Greek and wounding
to his αἰδώς as if he had to allow himself to be the object of Persian obeisance.
It is clear that in 920 this ritual is meant. Just as we have χαμαιπετές here,

perversities which have destroyed the thought and structure of this passage can be ob-
tained from, e.g., the commentaries of Weil, Headlam-Thomson, Blass, and Verrall. The
whole trouble arose from following blindly the stupid explanation τὴν Κλυταιμήστραν
(interlinear gloss in M). The passage was correctly understood by Bothe: ‘ravrnv, ῥῆσιν,
quod intelligendum ex εἴρηται, verbo eiusdem originis [it is not necessary to understand a
noun, see above]’, but it appears that no notice was taken of his comment. With ταύτην
τοιαύτην εἶπον Apollo concludes the first part of his speech; this phrase (somewhat like
haec hactenus) is very close to ταῦτα... τοιαῦτα and the examples collected below on 950.
Apollo intends to deal with a further point after the conclusion of this section ; this is also
clear from μέν 636, which was rightly explained by Wilamowitz (cf. also his note on Suppl.
940).
416
 COMMENTARY line 925

so Aeschylus, Pers. 588, describes the Persian προσκυνεῖν as ἐς γᾶν προπίτνοντες,

‘exactly corresponding to the monuments’ (Kranz, Stasimon, 87). From the
association of the adjective with βόαμα we have to assume cries of greeting
and benediction connected with the obeisance. The poet undoubtediy knew
of the spreading of garments in someone’s path as an oriental custom.
Blomfield compares Ev. Marc. 11.8 καὶ πολλοὶ τὰ ἱμάτια αὐτῶν ἔστρωσαν eis τὴν
ὁδόν, for other instances see van Heusde on gog (831 v. H.), e.g. Reges 4. 9. 13
festinaverunt ilaque, et unusquisque tollens pallium suum posuerunt sub pedibus
eius . . . alque dixerunt: regnabit Iehu (an example of χαμαιπετὲς βόαμα),
Herakleides of Kyme ap. Athen. 12. 514c καὶ διήιει (the Persian king) διὰ τῆς
τούτων αὐλῆς πεζὸς ὑποτιθεμένων ψιλοταπίδων Σαρδιανῶν, ἐφ᾽ ὧν οὐδεὶς ἄλλος
ἐπέβαινεν 7 βασιλεύς (quoted by Klausen).
919. GBpuve: ‘the word ἁβρότης signifies for Aeschylus a characteristic
quality both of barbarians and of women’, B. Snell, Aischylos, 80.
βαρβάρου φωτὸς δίκην taken correctly by the scholiast: μὴ ὡς βαρβάρωι
μοι κέλευε θρύπτεσθαι; similarly most editors (cf. C.R. li, 1937, 61). The
minority of scholars who like Wecklein, L. Campbell, and Plüss take the
words with the subject have been followed recently by Kranz, Stasimon, 102,
and, as it appears from his translation, G. Thomson. The parallelism is
decisively in favour of the other interpretation, which also avoids the in-
congruity of a comparison between the conduct of the queen and that of a
barbarian man. An exact parallel to the connexion between βαρβ. φωτὸς
δίκην and the dative which comes later may be seen in Eum. 25 f. βάκχαις
ἐστρατήγησεν θεὸς Aayw δίκην Πενθεῖ καταρράψας μόρον.
The three y7j-clauses (918-20) proceed from the general to the particular;
first ‘do not require of me delicacy, as if I were a woman’, then he mentions
the προσκύνησις and the acclamation and alludes to the barbaric character of
Clytemnestra’s actual intention, and finally he points to the intention itself.
The triple ‘do not’ expresses strongly his reluctance.
922. rıpaAgeiv: this verb, which in the earlier literature occurs only in
Pindar (N. 9. 54) and Aeschylus (Eum. three times), is used by Aristotle, Pol.
7. 17. 1336 Ὁ 19 (ἀφέησιν ὁ νόμος τοὺς τὴν ἡλικίαν éyovras . . . τιμαλφεῖν τοὺς
θεούς), in a passage where he is quoting the νόμος which regulates ἃ cult. So it
appears at least possible that the word once belonged to the language of
religious ceremonies. If this is so, then the word is employed here and
similarly in Eum. 15 and 807 in the sense of its original usage.
923. ποικίλοις : ‘patterned’, cf. on 909.
κάλλεσιν: the principal instances are cited in Blomfield’s Glossary ; cf.
also L-S κάλλος 3. For the special meaning with which we are dealing
here Eupolis fr. 333 K. ra κάλλη τὰ περίσεμνα τῆι θεῶν is particularly signifi-
cant.
923 f. ἐν... κάλλεσιν βαίνειν. Cf. 945 τοῖσδέ u’ ἐμβαίνονθ᾽ ἁλουργέσιν θεῶν.
924. ἐμοὶ μὲν: for “μέν solitarium’ after a personal pronoun cf. Denniston,
Particles, 381.
925. Rejected as spurious by Wilamowitz. Pohlenz, Griech. Trag. ii. 35, takes
the opposite view: “1. 925 can certainly be dispensed with after 922-4, but it
provides so good a summary of what is really for Aeschylus the main factor
in this scene, that I cannot believe it an interpolation.’ This is all very fine,
but it misses the real issue. Wilamowitz excised the line not because it can
4872.2 Ee 417
line 925 COMMENTARY
be dispensed with, but because it destroys the clear structure of the thought:
‘926. 27 causam afferunt cur invidiam excitaturus sit calcata purpura’; of
course, this is based on Blass’s correct interpretation of 926 f. (see below).
The asyndeton χωρὶς κτλ. has its full force if this line follows immediately
upon 924, to explain it. 925, on the other hand, Adyw . . . σέβειν ἐμέ, can under
no circumstances serve as an explanation of Agamemnon’s fear, because it is
a command to others and not, like the rest that follows (down to the end of
929), a statement of maxims derived from common experience. So 925
interrupts the direct line of thought which, aiming at a single point only,
extends from 923 to 930. “That I, a mortal, should step over the splendid
draperies reserved for the service of the gods, this inspires me with fear
indeed. It is only necessary to hear the difference between the two terms,
footwipers, embroideries, in order to know that the one may not be used for
the other. And for a man to remain sensible (not foolishly to exalt himself)
is the greatest gift of the god. Only he who prospers right up to the end of
his life should be called happy.’ In the last two sentences fear is again the
predominating motive, fear that a man, befogged by the success of the
moment, should forget the limits of his humanity, usurp in ἀφροσύνη what is
not for him, and so hazard his τελευτῆσαι ἐν εὐεστοῖ. In the familiar Aeschylean
fashion the beginning of this section is taken up at the end, where in 930
εὐθαρσὴς ἐγώ recalls the ἐμοὶ μὲν οὐδαμῶς ἄνευ φόβου.
We cannot say what the origin of the spurious line was, nor how it found
its way into the text. One possibility, though not subject to proof, is that in
a later performance someone wished to shorten the speech (cf. on 863). 925
could be made to follow immediately upon 914.
926. ποδοψήστρων: schol. ἀπὸ μεταφορᾶς τῶν πέπλων, 8v ὧν τοὺς πόδας
ἐκμάσσοντες λαμπροτέρους ποιοῦσιν οὗ τούτους ἐκπλύνοντες συνεχῶς. ποδόψηστρα
γὰρ κυρίως ταῦτα λέγεται. νῦν δὲ τὰ ὑπὸ τοὺς πόδας ἁπλῶς πέπλα οὕτως εἴρηκε.
The formation (from ψῆν) is clear; for the use of the suffix -rpov to form
instrument nouns from verbs cf. Schwyzer, Griech. Gramm. i. 532, and Buck
and Petersen, A Reverse Index, 314. It seems fairly certain that ποδόψηστρον
is the correct reading in Herodas 5. 20; 11 this is so, the word, which is used
there as a term of contempt, may be confidently assumed to belong to
everyday speech, as, presumably, it did already in the time of Aeschylus.
As early as Hipponax we read (Pap. Oxy. 2175, fr. 3. 18) [πό]δας περιψήσας.
926 f. The first to understand this passage correctly was Blass, Mélanges
Henri Weil (1898), p. 11. The fact that A. Y. Campbell calls his interpretation
'imbecillam sordidamque' makes no difference. So long as the critics were
satisfied with the traditional interpretation ('absque pedum substragulis
pictisque tapetibus satis fama mea clamat’ Schütz, ‘no foot-cloths, no false
refinements, need proclaim what rumour cries’ Verrall), they were bound to
suppose that ποδόψηστρα and ποικίλα are treated as more or less equivalent
terms. That is to say that, whereas Agamemnon elsewhere speaks with awe
of the ποικίλα and wishes them to be reserved for the gods, he is to be regarded
here as putting them on the same level as ποδόψηστρα; and the audience is
not only to close its ears to the clearly contemptuous ring of that word, but
also to forget what as Greeks they knew very well and had just (922 ff.)
heard to remind them, viz. that the ποικίλα were not meant for the feet of a
mortal Furthermore, according to this interpretation Agamemnon is sup-
418
 COMMENTARY line 926

posed to say: ‘even without these things my reputation (or ‘reputation such
as I can claim’) raises aloud its voice’, as if he were Odysseus among the
Phaeacians, or his own herald (575 ff.), or anyone but the Agamemnon of this
play ; for where does this great king speak of reputation? Moreover, the old
interpretation is quite incompatible with the train of thought, which is
completely dominated by the theme of οὐκ ἄνευ φόβου (see above). Blass’s
interpretation is correct: ‘aliud sonat nomen ποδοψήστρων, id est pannorum
quibus pedes abstergentur, aliud purpurarum; non igitur oportet has pro
1115 habere.’ For the use of κληδών simply for ‘name, term’ Blass compares
Eum. 418. ἀυτεῖ is a strong word, but for that reason very suitable here:
only the completely deaf could fail to hear what the words themselves cry
aloud. The word ποικέλα by itself turns the thoughts to the service of the
gods, the word ποδόψηστρα conveys the idea of something used for wiping the
shoes. Finally, Blass’s explanation also falls into line with the common
linguistic usage according to which a χωρίς of this kind, coming in front of
two quite different things at the beginning of the sentence, generally forms
either the predicate itself or part of the predicate, cf., e.g., S. Oed. C. 808
χωρὶς τό τ᾽ εἰπεῖν πολλὰ Kai rà καίρια, Trag. fr. adesp. 560 N. χωρὶς rà Μυσῶν
Kai Φρυγῶν ὁρίσματα, Pl. Ewthyd. 289 d χωρὶς ἡ τοῦ ποιεῖν τέχνη kai ἡ τοῦ
χρῆσθαι, and, with the addition of the copula, Ar. Thesm. 11 χωρὶς γὰρ
αὐτοῖν ἑκατέρου στιν ἡ φύσις, Pl. Prot. 336 Ὁ χωρὶς γὰρ ἔγωγ᾽ wıumv εἶναι τὸ
συνεῖναί τε ἀλλήλοις διαλεγομένους καὶ τὸ δημηγορεῖν. Sometimes ἃ different
verb takes the place of the copula, with no essential change of word-order,
e.g. E. Alc. 528 χωρὶς τό τ᾽ εἶναι καὶ τὸ μὴ νομίζεται. This ywpis . . . νομίζεται
(aliud ... existimatur . . .) is very close to Ag. 926 f. χωρὶς... ἀυτεῖ (aliud . . .
sonat...). To forestall the objection that χωρίς, coming as it does immediately
before the genitive, must be taken as a preposition, cf., e.g., Semonides of
Amorgos fr. 7. 1 χωρὶς γυναικὸς θεὸς ἐποίησεν νόον and Ar. Thesm. 11 (quoted
above).
926. In ποδοψήστρων τε καὶ τῶν ποικίλων it may well be assumed that the
article is to be taken ἀπὸ κοινοῦ as in the examples from Tragedy? given by
E. Bruhn, Anhang zu Soph. $ 171. vii, e.g. E. Phoen. 495 f. καὶ σοφοῖς καὶ τοῖσι
φαύλοις ἔνδικα, Cf. also Pearson on Soph. fr. 188 (his judgement on the point,
however, is not correct) and add E. Hec. 294 f. Aóyos . . . ἔκ τ᾽ ἀδοξούντων
ἰὼν κἀκ τῶν δοκούντων. The peculiarity of the ἀπὸ κοινοῦ in Pindar, Of. 13. 98
᾿Ισθμοῖ τά τ᾽ ἐν Νεμέαι is pointed out by Wilamowitz, Pindaros, 370 n. 2.
Examples from prose? have been collected by Vahlen, Artist. Poet., 3rd ed.,
105, in a note on 1449 a 1 'IXuàs καὶ ἡ ᾿Οδύσσεια (no account of it has been
taken in the quite inadequate survey in Kühner-Gerth, i. 613) ; he adds: 'ad
poetarum usum nolo excurrere, apud quos multa similia extant, non semper

1 Here, as in Ag. 926, the two expressions are joined by re . . . καί, although it is the
difference between them that is emphasized. This is the case also in the passage quoted
later in the note, E. Alc. 528. Cf., e.g., 807 ff. γνώσηι δὲ χρόνωι διαπευθόμενος τόν τε δικαίως
καὶ τὸν ἀκαίρως πόλιν οἰκουροῦντα πολιτῶν. It appears that two diametrically opposite things
can be conceived as component parts of a closely bound whole.
2 Denniston's treatment of the matter, on E. El. 1351, is not satisfactory, and to read
there ‘Ooia, which would destroy the parallelism of the passage (οἷσιν δ᾽ ὅσιον καὶ τὸ δίκαιον
φίλον ἐν βιότωι), seems to me quite unacceptable.
3 Svensson, Eranos, xliv, 1946, 253, adds Xen. Ages. 11. 10, but there καλῶν ἔργων is
indefinite, and τῶν καλῶν σωμάτων refers to éraipots.

419
line 926 COMMENTARY
recte aestimata.’ An example from Aeschylus can now be added, @ewpoi,
Pap. Oxy. 2162, fr. 2 (a) col. 2. 1 κοὐδεὶς παλαιῶν οὐδὲ τῶν vewrépov.!
928 f. ὀλβίσαι . . . φίληι. Radermacher, Gnomon, xiv, 1938, 296, selects from
the tragic passages in which we find the topic ‘call no one happy before his
death’ (e.g. E. Heraclid. 865 f., Tro. 509 f., Soph. fr. 601 N. = 662 P.) several
in which, as here, the word ὀλβίζειν or ὄλβιος is used, viz. S. Oed. R. 1528 ff.,
Soph. fr. 588 N. (= 646 P.) οὐ χρή ποτ᾽ εὖ πράσσοντος ὀλβίσαι τύχας ἀνδρός,
πρὶν αὐτῶι παντελῶς ἤδη βίος (9) διεκπεραθῆι καὶ τελευτήσηι βίον (9) (it will be
noticed that in this extended form most of the essential phrases of Ag. 928
recur, either in the same words or in synonyms), E. Andr. 100 ff. These
passages (which had already been quoted by van Heusde) have been con-
nected by Radermacher and others with the words of Solon in Hdt. ı. 32. 7
(cf. on this Regenbogen, Das humanist. Gymnasium, 1930, 11 ff., and Hell-
mann, Neue Philol. Unters. ix, 1934, 47 ff.). For the μακαρισμός indicated
by ὀλβίζειν or ὄλβιος ὅστις cf. Regenbogen, loc. cit., Snell, Hermes, lxvi,
1931, 75.
929. εὐεστοῖ: cf. on 647.
930. From Blomfield’s note it would seem that in Tr the end of the line is
ἄνευ θάρσους ἐγώ. This is not true: the text in Tr is the same asin F. Her-
mann stated the facts accurately. Nevertheless Blomfield’s error was re-
peated in Vitelli-Wecklein’s app. crit. and from there made its way into the
editions of A. Y. Campbell and G. Thomson. Incidentally, in 933 the reading
of Tr is ηὔξω, not εὔξω.
For a desperate attempt to defend the MS text by assuming a highly
artificial construction, see Sidgwick’s commentary. As regards the sense
thus produced, Headlam is right when he says: ‘this can hardly be called a
meaning.” But even on purely linguistic grounds, the attempt to defend the
MS reading collapses before two objections: (1) the examples of ei with the
optative plus ἄν to which Sidgwick refers (more in Kühner-Gerth, ii. 482, to
which should be added the thoughtful discussion by W. Wyse on Isaeus 5.
32, p. 451), all belong to fourth-century prose.” (2) There is no other example
in Aeschylus of ὥς by itself = οὕτως (Dindorf, Lex. Aesch. 403); this was
expressly emphasized by Karsten and by Blass, Mélanges Weil 11, with
reference to the present passage. For Wilamowitz’s incorrect reading ws in
Ag. 1354 cf. ad loc.; in Pers. 565 Murray should not have adopted ὥς from
Heath and Schiitz; neither is Murray right in reading at Suppl. 622 ds
εἶναι τάδε, as against the majority of editors who accept the ὡς of the Mediceus
(Wecklein ad loc. ‘os = ὥστε). Moreover, even in Sophocles there is no
example of ὥς by itself = οὕτως (οὐδ᾽ ὥς and the like are quite different) ; in
both El. 65 and Oed. C. 1242 ws is used relatively. In general cf. Denniston
on E. El. 155.
The reading which Weil produced by altering one letter, εἶπον τάδ᾽ ὡς xrA.,
is free of grammatical difficulty and provides satisfactory sense. Weil him-
! ἀνδριὰς καὶ τὸ σφέλας on the 6th-century basis from Delos (Schwyzer, Exempla, 760) is
totally different: ‘the statue and its pedestal’.
2 The example Pind. N. 7. 89 εἰ δ᾽ αὐτὸ καὶ θεὸς ἂν ἔχοι, which was still defended by
Hermann, has long been emended.
3 The construction ἐκφυγεῖν ἄνακτ᾽ αὐτὸν ὡς ἀκούομεν (cf. Pers. 188 f.) was correctly
explained by G. Hermann in his notes on Vigerus, De tdiol., pp. 745 and 894, Paley ad loc.,
Vahlen, Ges. Schriften i. 373, and Wilamowitz, Platon, ii. 420.

420
 COMMENTARY lines 931-4
self put a comma after 7d9',! but it is better to put it after εἶπον, as Headlam
(‘I have said how I should act herein with a good conscience’) and Wilamo-
witz do. With regard to εἶπον at the end of a speech, Blass, Melanges Weil
11, compares Cho. 688, Eum. 638. The word εὐθαρσής, as noted by Weil, was
chosen in order to make the line hark back to 924. The ἐγώ, also, corresponds
to the pointed éuoi μέν in 924. It is a desirable, though minor, result of the
emendation (this, too, was pointed out by Weil) that εἰπέ in Clytemnestra’s
first words in 931 picks up the preceding εἶπον, a repetition very much in the
style of stichomythia.
931-4. The explanation of these four lines is extremely controversial, and
that not only with regard to details. It is rather in their approach to the
main idea that scholars hold opinions diametrically opposite to each other.
It will be convenient to take a couple of translations to represent the tradi-
tional interpretation, ignoring slight differences in detail. Conington: 'Crv.
Still say not this against my judgement now. Ac. My judgement know that I
will never spoil. Cry. Thou must have vowed to Heaven in this, from fear.
Ac. If e'er man did, I spoke from sober choice.’ Nägelsbach (and almost
exactly the same e.g. Wilamowitz, who in his translation took 933 as a ques-
tion): ‘Und doch widersetze dich in diesem Stücke meinem Willen nicht.—
Meinem Willen, wisse es, werd’ ich nicht entsagen.—Hast du den Göttern aus
Furcht also zu tun gelobt?—Ein Kundiger, wenn irgend wer, sprach diesen
Beschluss ich aus.' The decisive line for the interpretation of the passage is
933. Unfortunately hardly any commentator deigns to say explicitly how
he takes the line. But the wording of most translations makes it certain that
ὧδ᾽ ἔρδειν τάδε is taken to mean something like 'to be determined not to walk
upon the precious fabrics'. This is clearly the interpretation in Hartung's
paraphrase (in the commentary) and in Schneidewin's commentary: 'she
asks .. . whether Agamemnon has made any vow to the gods to act in this
way, i.e. to show such σωφροσύνη, and this from fear of someone' and in
Verrall: ‘You vowed perhaps in some hour of terror so to perform this act? i.e.
to make a humble entrance, propitiating the gods by renunciation.' This
interpretation, however, of ὧδ᾽ ἔρδειν τάδε 15 clearly wrong. Against it Weil
noted (1858) : 'Sed haec ad eum pertinere videntur qui aliquid facit, non qui
facere veretur', and gave his own paraphrase of the line as follows: 'haec res
tantum abest ut dis ingrata sit, ut etiam vovisses te ita facturum esse, si in
periculo versatus esses.' This hits the mark, but it remained ineffective
because it was not put into its proper context. Kennedy's handling of the
passage goes deeper ; in fact he has once and for all made clear the meaning
of 931-3 (first in the J. Phil. vii, 1877, 14 ff., then, with some additions, in the
appendix to his translation, 2nd ed., 1882, pp. 194 ff. ; he took no account of
Weil). He rightly asks: 'Are we anywhere told that Greek warriors in time of
danger vowed to the gods that they would refrain from doing something? Do
we not read everywhere, that their vows were to do something involving
expenditure, to offer victims, to build shrines, or, as here, to walk on purple
embroideries in honour of the propitious deity? . . Could anyone be supposed
to make a vow, that, if a dangerous crisis occurred, he would not walk on purple
after being saved?’ Kennedy compares 963 ff., where Clytemnestra alleges
1 Later (in the Teubner edition) he put the comma in front of εὐθαρσής, which seems to
me very harsh.
421
lines 931-4 COMMENTARY

that, in order to bring about Agamemnon’s safe return, she would have been
ready to offer that same vow which she envisages as a possibility in her
question in 1. 933. Indeed the clear correspondence (it might almost be
called an echo) of the dv ηὐξάμην with the earlier ηὔξω. . . dv should never
have been overlooked. Blass, Mélanges H. Weil, 12 f., noted that the theme
which is particularized in these two passages recurs in a more general form
(sacrifice of precious personal property in order to rescue from ruin the whole
fortune of the house) in the following chorus, 1008 ff.
Now that the main point is correctly understood, it is clear that 933 should
be kept as a question (‘would you in some fearful crisis have vowed that you
would do the thing I am now inviting you to do?’, in Kennedy’s paraphrase) ;
in F and Tr there is a question-mark at the end of the line; cf. on 540. A
further consequence is that in 934 ἐξεῖπον cannot be retained. As long as the
commentators took ὧδ᾽ ἔρδειν τάδε in 933 as expressing Agamemnon’s resolu-
tion not to tread upon the fabrics, τόδ᾽ ἐξεῖπον τέλος seemed to refer to this
resolution, but now the foundation for this interpretation has been removed.
So the emendation of Auratus ἐξεῖπεν proves to be correct: ‘Yes, if anyone
with real understanding of the matter had prescribed this (ritual) perform-
ance.’ τέλος in this sense: L-S s.v. I. 6, cf. on 1202. The limiting ye placed
between εἰδώς and εὖ shows that this point (the guarantee that the order
comes from a seer who is possessed of real knowledge) is essential for the
fulfilment of the condition, cf., e.g., S. El. 1105 ἥδ᾽, εἰ τὸν ἄγχιστόν ye κηρύσσειν
χρεών (cf. Denniston, Particles, 142, where more examples of this kind can be
found), cf. also Prom. 379 ἐάν τις ἐν καιρῶι ye μαλθάσσηι κέαρ. The particular
suitability of εὖ εἰδώς for the seer or priest of an oracle is well demonstrated
by Z 438 ἢ πού τίς σφιν ἔνισπε θεοπροπίων ἐὺ εἰδώς and B 170 οὐ yàp ἀπείρητος
μαντεύομαι, ἀλλ᾽ ἐὺ εἰδώς (quoted by Headlam), to which Platt, J. Phil. xxxii,
1913, 60 adds A 384 ἄμμι δὲ μάντις εὖ εἰδὼς ἀγόρευε θεοπροπίας ἑκάτοιο, cf. also
the epigram Hdt. 7. 228. 3 μάντιος, ds τότε κῆρας ἐπερχομένας σάφα εἰδώς κτλ.
Decisive confirmation of the correctness of this reading (e£eiwev) and the
meaning assigned to it here is to be found in Clytemnestra’s utterance in 964,
already quoted. It is in the elucidation of 934 that Blass and Headlam (who
is followed by Mazon) have improved on Kennedy’s interpretation.
It will be convenient to proceed at once to the discussion of the remaining
details of 933, before attending to the opening lines of the stichomythia.
Headlam’s reading ἔρξειν for ἔρδειν seems to be correct ; his observation that
‘I vow to do something’ is expressed by εὔχομαι with the future infinitive
(4 101 f. etc.?) is cogent as long as no parallel for the use of the present can be
adduced (with regard to S. Phil. 1032 f. Headlam’s judgement is, of course,
correct, and the manner in which L-S s.v. εὔχομαι II. x quote the passage is
‘suggestio falsi’). It is more difficult, however, to decide about the word-
order in the line. That dv ηὔξω must be taken together is clear, other con-
siderations apart, from the ἂν ηὐξάμην in 963, which in meaning (see above)
is strictly parallel. And indeed in sentences where there is a participle, which
is conditional in sense, subordinated to the main verb, av sometimes appears
to be placed next to the participle, cf. Kühner-Gerth, 1. 242 n. 1. But a closer
1 Cf. also Wilamowitz, Hellenist. Dichtung, ii. 237 n. 4: ‘The impression is given that
Idmon is introduced first as a prophet with a name to fit his calling.’ Cf. Πολύιδος (W.
Schulze, Quaest. ep. 118). 2 Cf. Latte, Hermes, lxvi, 1931, 129 n. I.
422
 COMMENTARY lines 931 f.

examination of the comparable passages, i.e. those in which ἄν is not repeated,

makes it clear that their word-order is completely normal: Thuc. 1. 11. 2
πολιορκίαι δ᾽ àv προσκαθεζόμενοι ἐν ἐλάσσονί re χρόνωι xai ἀπονώτερον τὴν
Τροίαν εἷλον, 2. 18. 4 καὶ ἐδόκουν où Πελοποννήσιοι ἐπελθόντες ἂν διὰ τάχους
πάντα ἔτι ἔξω καταλαβεῖν. In the first example ἄν is placed in accordance with
a very ancient tendency of word-order (Wackernagel, Idg. Forsch. i, 1891,
333 ff.) as close as possible after the word which begins the sentence (8é often
excludes it from the second place ; it should, however, be noticed that δ᾽ here
has been cut out by Krüger, and the whole passage is strongly contested, cf.
Steup's appendix) ; in 2. 18. 4 àv is again in a typical position, coming second
in the kolon which begins with ἐπελθόντες. In the passage quoted by Good-
win ὃ 224, Demosth. 6. 20 ἢ λέγοντος dv τινος πιστεῦσαι scil. οἴεσθε; (7 is out-
side the kolon as is frequently καί when it connects a sentence, cf. my Kolon
und Satz, ii. 341 n. 1) the word-order is of course quite normal. On the other
hand, in Ag. 933, as the line stands, viz. ηὔξω θεοῖς δείσας ἄν «rA., the ἄν is
neither next to the verb the conditional force of which it marks, in this case
ηὔξω, nor does it come second in the sentence or a κῶλον. Whether such a
position for ἄν, as it has here, is permissible, must remain at least open to
question. I have not found time to read through again a great number of
texts with a view to settling this particular point and would not like to rely
upon my memory, but I may at least say that I cannot recall having come
across ἄν in so abnormal a position. This fact makes me inclined to adopt the
suggestion which Hermann casually put forward, only to give it up at once.
He and Wilamowitz (the latter clearly without recollection of Hermann)
thought of altering the beginning of the line to δείσας ἂν ηὔξω (then, of course,
θεοῖσιν). It is possible that the word-order in the MSS, pushing forward the
main verb as it does, is another example of 'simplex ordo'; from Headlam's
list, C.R. xvi, 1902, 244 ff., I would pick out E. Or. 497, where instead of κάρα
θυγατρὸς τῆς ἐμῆς πληγεὶς ὕπο (Brunck) the MSS give: πληγεὶς τῆς ἐμῆς θυγατρὸς
(or θυγατρὸς τῆς ἐμῆς) ὑπὲρ κάρα,2 and S. Phil. 156, where Hermann has restored
μὴ προσπεσών με λάθηι ποθέν from μή με λάθηι προσπεσών. A pretty example of
the pushing forward of ἃ verb is to be found in the first line of the ὑπόθεσις of
Menander’s Heros, where the papyrus gives : ἄρρεν τεκοῦσα παρθένος θῆλύ θ᾽ ἅμα.
931 f. We have come to the conclusion that 933 is not an impulsive and
contemptuous utterance, called forth by the supposed stubbornness of
Agamemnon, but a serious question put for the purpose of bringing the king
to admit that even he, in spite of his religious scruples, would, under certain
circumstances, have actually bound himself, by a vow to the gods, to the
very same action which is now being suggested to him. The question is a
masterly thrust in Clytemnestra’s dialectical fencing. It is very important
for her that Agamemnon should grasp it clearly and answer without hedging.
The correct interpretation of 933 makes it possible to understand the prepara-
tory remark in 931: ‘Now answer me what I am now going to ask, openly and
honestly!’ In this context there is no room for the view generally held since
Schütz (see above): 'noli ea adversus sententiam meam dicere.’ We must,
1 Such infinitive constructions often form a separate kolon, cf. Ed. Fr., Kolon und Saiz,
ii. 326 ff., where sufficient examples are quoted from Thucydides and other authors.
2 Cf., with regard to Brunck’s emendation and the confusion of the original word-order
here and elsewhere, Wilamowitz, Hermes, lix, 1924, 256 n. 1.

423
lines 931 f. COMMENTARY
rather, return to Stanley’s translation: hoc dicas non contra animi sententiam
tui (this was anticipated by Triclinius, who put a comma after γνώμην and
wrote above παρὰ γνώμην the gloss ἀλλ᾽ ἑκουσίως, and Casaubon: ‘Responde
mihi ex animi tui sententia. —Scito me sincero animo responsurum.’). van
Heusde considered this as a possibility and it was adopted without qualifica-
tion by Kennedy and those who followed him. Now for the details: εἰπεῖν
παρὰ γνώμην is to speak contrary to one's own γνώμη, deliberate opinion,
conviction, advised judgment’ ; thus Headlam, who appositely quotes Thuc.
3. 42. 6 παρὰ γνώμην τι καὶ πρὸς χάριν (dv) λέγοι, 6. 9. 2 οὔτε ἐν τῶι πρότερον
χρόνωι... εἶπον παρὰ γνώμην oùre νῦν ἀλλ᾽ 1) ἃ ἂν γιγνώσκω (the way in which
παρὰ γνώμην is picked up is instructive, and the addition of βέλτιστα re-
inforces the point) βέλτιστα ἐρῶ, and also later examples. With regard to the
rest, the whole of our argument shows that τόδε in τόδ᾽ εἰπέ must refer to
what follows, i.e. to what she has in mind. ‘id quod dicam. Male vulgo ad
superiora refertur, quasi legeretur τόδε μὴ εἶπέ. Est familiariter et blande
alloquentis’ (van Heusde). ‘68’ εἰπὲ... ἐμοί differs from the usual formula
for asking a question, εἰπέ μοι, only in the appealing emphasis thrown upon
ἐμοί᾽ (Verrall. Against this Headlam rightly remarks that the position of
ἐμοί at the end of the sentence is not emphatic (this is the old feud between
Verrall and Headlam, cf. the latter’s On Editing Aeschylus, 8 f. and elsewhere),
but then he says himself: ‘nevertheless in the use of ἐγώ, ἐμοί, ἐμέ at the end
of three successive lines we hear an undertone of strife between the two wills.’
Headlam has supplied a sufficiency of examples of εἰπὲ δή νύν μοι τόδε and the
like, serving simply as introduction to a question. It may, however, be use-
ful to quote an instance of a similar preparatory clause from an Aeschylean
stichomythia: Prom. 616 ff. οὔκουν πόροις ἂν τήνδε δωρεὰν ἐμοί; ---λέγ᾽ ἥντιν᾽
αἰτῆι' πᾶν γὰρ ἂν πύθοιό μου. ---σήμηνον ὅστις ἐν φάραγγί σ᾽ ὥχμασεν. The
insertion of μὴ παρὰ γνώμην into the main thought τόδ᾽ εἰπὲ... ἐμοί was
recognized by Verrall; for the comparative independence of the inserted
phrase Headlam compares 510, 5. Ant. 446 σὺ δ᾽ εἰπέ μοι, μὴ μῆκος ἀλλὰ
συντόμως (punctuated correctly by Hermann), etc. ; for the insertion of such
a prohibiting phrase cf. further Sept. 200 f. Finally we come to καὶ μήν at the
beginning of the line. Our interpretation of the whole context goes to show
that Denniston, Particles, 352, was correct in cataloguing it as an example of
the ‘progressive’, not the adversative, use. The latter (Denniston, 357 f.) is
certainly used by Aeschylus upon occasion, e.g. Ag. 1254, but not very often.
‘Normally καὶ μήν marks a new departure: it is mainly used after a strong
stop’ ; this is the case here.
932. γνώμην μὲν κτλ. For the omission of the antithesis Blass on Cho. 400
compares the νόμος μέν there, ‘the law it is in any case’, with the μέν here.
Denniston, Particles, 380, puts Ag. 932 among the examples of μέν soli-
tarium’, where ‘the speaker uses μέν, like ye, in contrast with something which
he does not, even in the first instance, intend to express in words’.
διαφθεροῦντα. This seems to me an instance of ‘the Greek idiom, by which
a person is said himself to produce the changes of physical and mental state
which take place in him’ (Rutherford on Babrius 22. 1).
Perhaps the most important consequence of the interpretation followed
here is that we are now able to ensure a correct understanding of the tone set
at the beginning of this dispute which is of such consequence for the course
424
 COMMENTARY line 937

of the play. According to the traditional view (cf. the translations quoted
above) Agamemnon’s words are stubborn and obstinate, and Clytemnestra
(933) makes an offensive insinuation. Now, however, the question of the
queen appears as a cautious and thoroughly skilful attempt to elicit from her
husband in an apparently friendly way the decisive admission. Once he has
agreed to the proposition that treading on the costly fabrics is not necessarily
ἀσεβές, then, she calculates, the path is clear for further concessions. With
regard to the words of Agamemnon, Kennedy’s statement is adequate: ‘we
cannot suppose that Aesch. means Agamemnon simply and brutally to
reply here: ‘‘I am not the man to change my will, I can tell you." His language
is the sentiment of a true gentleman: ‘be sure I willsay what I really think’’.’
The gentilezza of the utterances of king and queen should make the serious-
ness of their conflict all the more perceptible ; but it is not surprising that an
age deaf to the notes of true nobility read into their words assertiveness,
contention, and the scorn of the plebeian.
935. τί δ᾽ ἂν δοκεῖ σοι Mpiapos: understand δρᾶσαι or the like. Such ‘ellipses’
clearly belong to everyday speech. Cf., e.g., Ar. Clouds ı54 f. τί δῆτ᾽ ἄν,
ἕτερον εἰ πύθοιο Σωκράτους φρόντισμα; 769 ff. φέρε, τί δῆτ᾽ dv, ei... τὰ γράμματ᾽
ἐκτήξαιμι τῆς ἐμῆς δίκης;
936. He answers the question in the affirmative, very emphatically (κάρτα)
but quite quietly, without a trace of contempt or irritation, which would
certainly have been understandable, since he had not long before (oro ff.)
made it quite clear that he did not wish to be honoured like a barbarian.
Clytemnestra seems purposely to overlook the fact that the conduct of a
non-Hellene cannot provide a standard for the behaviour of Agamemnon ;
she speaks here as if all that matters is to manœuvre her husband by means of
any conceivable admission, however thin the argument, from his uncondi-
tional refusal. Such tactics are in keeping with the character of this pre-
liminary skirmish (cf. on 944 ff.).
937. μή νυν κτλ. Triclinius: καθ᾽ ὁμαλισμὸν ἀναγνωστέον τό νυν Kal ἄνευ τόνου,
ἵνα ἦι ἀντὶ τοῦ δή. Cf. schol. E. Med. 584 τὸ δὲ "un νυν’ κατὰ ᾿Αττικοὺς ἀναπέμ-
πεται. Here, as often, νυν no longer has the full temporal force but only the
weakened meaning ‘then’, implying a quae cum ita sint. ‘Do not, then,
regard....’ For the reference of this νυν to what has immediately gone before
a few random examples should suffice, all belonging to prohibitions of a
similar kind. A. Suppl. 207 (also stichomythia) μή νυν σχόλαζε, Cho. 770
(conclusion of the stichomythia) μή vvv σὺ ταῦτ᾽ ἄγγελλε krA., E. Med. 584
ὡς Kal σὺ μή νυν eis ἔμ᾽ εὐσχήμων γένηι, S. Phil. 576 μή νύν μ᾽ ἔρηι τὰ πλείον᾽,
ἀλλὰ krÀ., Ar. Plut. 413 μή νυν διάτριβ᾽, ἀλλ᾽ dvve κτλ. In the present passage
it is not immediately clear what the ‘then’ points to. Presumably Clytem-
nestra draws her conclusion not only from the answer to her last question but
also from the one before. ‘If in certain circumstances an action of this kind
can result from pious motives, and if a king like Priam would have dared it
without hesitation—why do you, the king, need to regard the reproach of
men?’ For the thought in general cf. Archil. fr. 9 D. Αἰσιμίδη, δήμου μὲν!
ἐπίρρησιν μελεδαίνων οὐδεὶς ἂν μάλα πόλλ᾽ ἱμερόεντα πάθοι.
1 This elegant emendation (by Elmsley), which has been rejected by Bergk and others,
seems certain; the passage Ag. 937 f., where ἀνθρώπειος ψόγος is identified with φήμη
δημόθρους, may be regarded as a valuable parallel.
425
line 938 COMMENTARY
938. Agamemnon, calm and firm, meets the arrogance of his wife, who would
like to make light of the judgement of the masses, with an ancient saw.
Triclinius, ad loc., recalls Hesiod, Erga 763 f. φήμη δ᾽ οὔτις πάμπαν ἀπόλλυται,
ἦν τινα πολλοὶ λαοὶ φημίξωσι" θεός νύ τίς ἐστι καὶ αὐτή (add also ibid. 760 ff.
ὧδ᾽ ἔρδειν" δεινὴν δὲ βροτῶν ὑπαλεύεο φήμην KrA.), cf. also Ag. 456 βαρεῖα δ᾽
ἀστῶν φάτις σὺν κότωι. I should like to believe that 937 f. (notice also αἰδεσθῆις)
was influenced by Homer 75 (= τ 527): εὐνήν τ᾽ αἰδομένη πόσιος δήμοιό τε
φῆμιν (Penelope is not only a better woman but also wiser than Clytemnestra).
δημόθρους: cf. on 883.
939. The sentence was correctly punctuated by Stanley, whereas the MSS
and the earlier editions put a question-mark at the end.
'8é ye or 86. . . ye is common in retort, where the second speaker, accepting
the statement of the first, wishes to cap it or to bring in a consideration on the
-other side’ (Neil, Ar. Knights, Append. i, p. 191). Cf. 941.
939 f. His fear of φθόνος stands in her way, so she seeks to change his mood
with a platitude appealing to his pride. "Where there is no φθόνος, there is
no ζῆλος ', i.e. you cannot win ζῆλος, without at the same time taking φθόνος
into the bargain: Thuc. 2. 45. x φθόνος γὰρ τοῖς ζῶσι πρὸς τὸ ἀντίπαλον, τὸ δὲ
μὴ ἐμποδὼν (this has its particular point in the Funeral Speech, but apart
from that, this maxim—cf., e.g., Horace, Episiles 2. 1. 13 f.—applies to every-
thing that is not in the way) ἀνανταγωνίστωι εὐνοίαι τετίμηται. An increase of
emphasis expressed by means of οὐδ᾽ (suggested by Platt, considered neces-
sary by G. Thomson) is not demanded by the sense. The ἐπίζηλος τύχα
(Bacchyl. 5. 52) is one of the blessings which a great lord and king particularly
desires (cf. also A. Pers. 710). Agamemnon picks up ζῆλος from Clytemnestra’s
words but gives it the sharpest of its possible meanings (cf. Hesiod, Evga
τος f. ζῆλος δ᾽ ἀνθρώποισιν ὀιζυροῖσιν ἅπασι δυσκέλαδος κακόχαρτος ὁμαρτήσει
στυγερώπης, Democritus B 191 ταύτης γὰρ ἐχόμενος τῆς γνώμης εὐθυμότερόν τε
διάξεις καὶ οὐκ ὀλίγας κῆρας ἐν τῶι βίωι διώσεαι, φθόνον καὶ ζῆλον καὶ δυσμενίην) :
by substituting μάχη for ζῆλος and pointing to the danger of ἃ conflict with
the people if their feelings are hurt too much,’ he is able to reproach his wife
with an unwomanly aim. Here for the first time he rebuffs her with some
asperity.
941. πρέπει: Pohlenz, Nachr. Gótt. Ges., Fachgruppe I, 1933, 53, quotes the
line among the passages which show how πρέπειν (cf. on 242) in the time of
Aeschylus and Pindar was beginning to denote value and standard, since
what was primarily a conspicuous quality came now to be regarded as
suitable to its possessor and normally associated with him. With νικᾶσθαι
Clytemnestra admits, involuntarily it appears, that a contest is taking place
between her husband and herself. She is so preoccupied with her struggle
and at the same time so intent on pressing Agamemnon with new arguments
that she distorts his ikeipew μάχης into a reference to the contest taking place
between them. And in point of fact, if his maxim οὔτοι γυναικός ἐστιν ἱμείρειν
μάχης has general validity, it must also be valid for the struggle which

1 T have no doubt that this is what Agamemnon means by μάχη, i.e. he insinuates that
Clytemnestra is seeking to involve him in such a conflict ; for throughout this dialogue he
refers closely to what she has said before. He would, however, be failing to do this if (as
‚it appears generally to be taken) he simply said: do not seek a quarrel with me! For the
meaning which Clytemnestra gives to μάχη, cf. on 941.
426
 COMMENTARY line 943
Clytemnestra has in mind, but not Agamemnon. ‘If I, the woman, should
not be intent upon a struggle, then you, the fortunate and successful one,
must also be generous for once and allow yourself to be defeated.’
942. Only now does Agamemnon fully grasp what her attitude really is, and
so he asks with unfeigned astonishment: ‘is this victory in the struggle really
worth anything to you?’ Paley translates ‘that sort of victory’, similarly
Headlam (‘Do you set store on ''victory" of that kind?’). That can hardly be
contained in τήνδε. Wecklein takes a different view: ‘do you really esteem
that as victory in a contest?’; 1.6, he takes νέκην δήριος predicatively and
τήνδε as ‘attracted’ object. An attraction of this kind would be grammatically
possible, cf. Kühner-Gerth, i. 74, and also e.g. E. El. 757 σφαγὴν dureis τήνδε
pot, Menander, Epitr. 236 (277 Korte, 3rd ed.) βιασμὸν τοῦτον εἶναι παρθένον
(Wilamowitz : “τοῦτον should be neuter, but is attracted to βιασμόν, “this is a
βιασμός᾽᾽ "). But τίω or τιμάω is not, as far as I can see, found with a double
accusative such as Wecklein assumes here, and the meaning of the verb (‘to
consider valuable’, but not ‘to esteem as something’) does not favour the
assumption. τήνδε is quite simple: ‘the victory of which you speak’ (when
you say νικᾶσθαι), in the way that ὅδε so often refers to the immediately
preceding words of the other speaker in a dialogue, e.g. A. Suppl. 297, Prom.
249, etc.
At the beginning of the sentence the text has been the subject of much
conjecture (a survey of the earlier suggestions can be found in Ahrens,
584 f.), and exception has especially been taken to kai. And yet we are con-
fronted here with nothing more than the ἦ καί, which so often (cf. Denniston,
Particles, 285, 316) introduces a lively question (‘do you really consider it of
value?’). It is clearly the whole question that is intensified and not any
single word, cf., e.g., S. 47.44 ἢ kai τὸ βούλευμ᾽ ὡς ἐπ᾽ Apyeioıs τόδ᾽ ἦν; It does
not appear to me necessary to suppose (as e.g. Paley, Verrall, and Headlam
do; Wellauer, too, had taken καὶ ov together) that there is any particular
emphasis on ov. In any case it cannot be concluded that the pronoun is
emphatic merely because it is expressed, cf., e.g., Eum. 896 ov τοῦτο πράξεις
ὥστε pe σθένειν τόσον;
δήριος does not seem to occur elsewhere ; the Ionic form of the genitive is
probably due to δῆρις being an epic word (Wecklein); Pl. Rep. 390 e τῆς
μήνιος (compared by Schneidewin) is a good parallel.
943. Objection to the MS reading πιθοῦ" κράτος μέντοι πάρες γ᾽ ἑκὼν ἐμοί has
been raised by Thiersch, Hartung, Karsten, Enger, and later scholars, but not
altogether on solid grounds. However, the fact remains that y’ is pointless
and that the employment of μέντοι to strengthen a demand has no adequate
support in usage (Denniston, Particles, 403). Many alterations have been
suggested. Some of them were adopted by Wilamowitz, who read the line as
"follows: πιθοῦ" κρατεῖς μέν, τὸ δὲ πάρες γ᾽ ἑκὼν ἐμοί. This sentence certainly
has great liveliness of expression. Of course it would have been better if
1 Q. Schroeder, Euripidis cantica (1928), p. 215, conjectures in E. Andr. 467 δήριας, which
is attractive although I have no parallel for an accusative in -c-as in an Attic poet (Wilamo-
witz's suggestion, Verskunst, 427 n. I, ἤριδας is inadmissible).
2 It seems to me that Denniston’s apparent exception, Hdt. 9. 79. 2, is invalid. With
od μέντοι Pausanias, after finishing with the theme ‘Revenge for Leonidas’, turns again to
Lampon and his proposal: so it is simply an example of the ‘progressive’ use, which
Denniston deals with on pp. 406 ff.
427
line 943 COMMENTARY
Wilamowitz had pointed out in the apparatus that he was thereby intro-
ducing a metrical singularity into the text of Aeschylus. An examination of
the passages collected by C. F. Müller, De pedibus solutis, etc. (Berlin, 1866),
shows that the trimeters of Aeschylus provide no example of either a tribrach
or a dactyl (as here according to Wil. μέν, τὸ δὲ) consisting of three mono-
syllables. This makes the conjecture doubtful, although it does not rule it
out altogether. On quite different grounds, however, Weil’s emendation
can be shown to be preferable. Weil read: πιθοῦ" κρατεῖς μέντοι παρείς
(Bothe's reading in his first edition) γ᾽ ἑκὼν ἐμοί. A further improvement was
provided by Wecklein, who omitted the MS y’, of which Paley had said that
it looked like a metrical insertion. This reading is supported by three con-
siderations. First, it does not deviate much from the MSS. Secondly, the
tension between κρατεῖς. and παρείς (see below) provides an excellent point,
very much in the style of stichomythia. Thirdly, the line compared with it
by Weil (Addenda to his first edition, p. 134) and also by Nauck, S. 47. 1353
παῦσαι" κρατεῖς τοι τῶν φίλων νικώμενος, gives the very strong impression
(there is, of course, no absolute certainty) that Sophocles is following the
pattern of Ag. 943 both in the peculiar formation of his line and in the thought
(with regard to the influence of the Ag. upon Sophoclean plays, especially
the Ajax and the Antigone, cf. on 1293, 1389, for Sophocles’ debt to Aeschylus
in the rhythm and sound of a particular passage cf. on 1446 f.). Metrically the
line, as thus emended, is undeniably harsh (this applies also to the MS
reading, cf. G. Hermann, Elem. doctr. metr. 112). But when in a trimeter with
a break in the middle a first half such as Zum. 26 λαγὼ | δίκην | Πενθεῖ |
καταρράψας μόρον occurs and a second half such as Pers. 465 Ξέρξης δ᾽
ἀνώιμωξεν | κακῶν | ὁρῶν | βάθος, it is difficult to see why an arrangement
such as πιθοῦ" | κρατεῖς | μέντοι || παρεὶς | ἑκὼν | épot should be completely
ruled out.
κρατεῖς : not present, but ‘timeless’ (cf. Wackernagel, Syntax, i. 157 f.), ‘you
are victorious’, as Eum. 741 νικᾶι δ᾽ ᾽Ορέστης, κἂν ἰσόψηφος κριθῆι and often
elsewhere.
μέντοι (against μέν τοι cf. on 644) : emphasizing, ‘truly’ or the like. Against
taking it as an adversative μέντοι is the fact that there is in Aeschylus no
unequivocal example of this usage (Denniston, Paritcles, 404).
rapeis : with this verb the object must often be supplied from the context,
e.g. S. Oed. C. 591 ἀλλ᾽ οὐδ᾽, ὅτ᾽ αὐτὸς ἤθελον, παρίεσαν, Pl. Symp. 214e Τἀληθῆ
ἐρῶ. ἀλλ᾽ ὅρα εἰ παρίης.----λλὰ μέντοι... τάγε ἀληθῆ παρίημι καὶ κελεύω λέγειν. In
the present passage something like τὴν νίκην may be supplied. Paley compares
Hdt. 6. 103. 3 καὶ τὴν νίκην παρεὶς τούτωι and E. Tro. 655 f. ἤιδη δ᾽ ἀμὲ χρῆν
νικᾶν πόσιν, κείνωι T€ νίκην ὧν ἐχρῆν παριέναι.
944ff. With πιθοῦ Clytemnestra suddenly gives up the arguments which she
had developed through ever new artifices, and, with most skilful calculation,
! On this point the late E. Harrison (Cambridge) kindly gave me the benefit of his
expert knowledge. In reply to my question he began by stating (letter of 30. 11. 1938), on
the basis of his own research, that in the Aeschylean trimeter a tribrach or dactyl is never
made up of three monosyllables, and then he goes on with exemplary circumspection :
‘For my own part, however, I attach more importance to what I call the couple than to the
whole foot, and on metrical grounds I should not condemn πιθοῦ" κρατεῖς-μέν, τό-δὲ πάρες γ᾽
in a poet who has πάσχουσι, rà-96 μέλλουσι (Pers. 814) and τήμῆι τὸ-σὸν (Eum. 446)’. On the
two latter lines the same conclusion was reached by C. F. Müller, op. cit. 22 and 85.
428
 COMMENTARY line 945
turns to entreaty. Her dialectics were resisted with quiet determination ; the
moment she speaks beseechingly Agamemnon gives way. Whatever his mis-
givings and however repugnant her behaviour may be to him, she remains
his wife and the mistress of his house." Even in the moment of his defeat
Agamemnon appears as the true gentleman he always is.
944. ἀλλ᾽ ei δοκεῖ σοι ταῦτα. When Menander’s Polemon is coming round
after an outburst of passion, he says (Peric. 257 ff.) : ἀλλ᾽ εἴπερ οὕτω σοὶ δοκεῖ
πράττειν... ἐλθὼν διαλέγου κτλ. Cf. also Ismene’s resigned reply 5. Ant. 98
ἀλλ᾽ ei δοκεῖ σοι, στεῖχε. Hoopoe, yielding to the insistence of the two Athe-
nians, replies (Ar. Birds 665): ἀλλ᾽ ei δοκεῖ σφῶιν, ταῦτα χρὴ δρᾶν. It appears
that this phrase was commonly used when after an argument one party
realized that further resistance was useless.
ὑπαί... Avot: the tmesis is possibly to be taken as a Homerism like ὑπαί;
the linguistic form may be intended to dignify the description of the humble
service involved (cf. on 556 f. and on 562) ; the elevated style of πρόδουλον
ἔμβασιν ποδός is in keeping with this.
tts: the order is probably addressed to the maids mentioned in 908. There
can be no doubt that the king’s command is immediately obeyed and that
the untying of his shoes actually takes place on the stage. C. Robert, 22.
Hallisches Winckelmannsprogramm (1898), 23 f., wished to substitute a make-
believe for the real act because he thought that, if Agamemnon’s buskins
were taken off, this would diminish his stature and he would not appear as
tall as Clytemnestra. In point of fact the κόθορνος used in the performances
of the fifth century had no very thick soles and did not serve to make the
actors appear considerably taller. This has been demonstrated by K. K.
Smith, Harvard Studies in Class. Philol. xvi, 1905, 123 ff. (who on p. 141 f.
deals with Ag. 944 ff.). Cf. also A. Korte, Festschrift zur 49. Philol.-Ver-
sammlung, Basel 1907, 198 ff.
apBuAas. At least one particular kind of ἀρβύλαι is a shoe which reaches
up to the ankles, as is clear from its picturesque name: Hippocrates #. ἄρθρ.
ἐμβ. 62 (II 214. το Kühlewein) : ἀρβύλαι ἐπιτηδειόταται al πηλοπάτιδες καλεό-
μεναι. This shoe, as it goes up high enough and encloses the whole of the foot,
gives a firm hold : τοῦτο γὰρ, the doctor continues, ὑποδημάτων ἥκιστα κρατεῖται
ὑπὸ τοῦ ποδός, ἀλλὰ κρατεῖ μᾶλλον. On this, Galen xviii. 1. 680 Kühn remarks:
κοῖλον ὑπόδημα καὶ περιεσφιγμένον ἀκριβῶς ὅλωι τῶι ποδὲ μέχρι τῶν σφυρῶν. This
special type of shoe, very suitable for travellers, may be meant here. On the
other hand, a careful examination of the evidence has shown that ἀρβύλη was
used in Tragedy as a general word for shoe; see A. A. Bryant, Harvard
Studies in Class. Philol. x, 1899, 73 ff.; cf. also K. K. Smith, ibid. xvi, 1905,
130 f.
945. πρόδουλον (only here): for the meaning of πρό cf. L-S s.v. D. i. 3:
‘standing in another’s place, πρόμαντις, mpögevos’ and Wilamowitz, Sappho
und Simonides, 140 n. 3.
ἔμβασιν: Doubtless Aeschylus’ audience knew ἐμβάς (described in Pollux
7.85) as the familiar name of a particular type of common shoe; Hdt. 1. 195. 1
testifies to the word for Boeotia, Aristophanes for Athens. Aeschylus certainly
1 It is, however, a complete misrepresentation to assume with J. T. Sheppard, Cambridge
Ancient History, v. 124, that ‘when she urges him again, he thinks she loves him, and he
yields’.
429
line 945 COMMENTARY
does not intend to allude to this technical usage, but &ußas may well have
suggested the bold, and therefore exalted, use of ἔμβασις.
946. καί is connective: the thought which follows is closely connected with
the preceding sentence. So Verrall’s ‘even with these if I tread’ is wrong.
ἁλουργέσιν : similar words denoting purple are collected in Blümner,
Technologie, i, 2nd ed., 234 n. 2.
θεῶν : most commentators agree with Stanley in taking it with the next
line, whereas Heath and e.g. Nägelsbach, Verrall, Wilamowitz rightly take
θεῶν dAoupyeaı together. The antithesis between what belongs to the gods
and what to man colours all this part of the scene,' so it is not sufficient to
compromise as Paley does: “θεῶν virtually belongs both to ἁλουργέσιν and to
ὄμματος. The fact that the gods perceive and act πρόσωθεν is mentioned
several times in the Oresteia (Ag. 952, Eum. 297, 397) and clearly taken for
granted as a current belief (cf. also Ag. 1579), so that the audience, even
supposing they were at all inclined in the case of l. 947 to think of human
vision, would not be in any doubt whose eye was meant.
947. Cf. on 469 f.
948. The correct meaning of Schiitz’s excellent emendation? δωματοφθορεῖν
was given by Blomfield: ‘quod verbum idem valet atque οἰκοφθορεῖν apud
Platonem et Herodotum. Huc respicere videantur Clytæmnestræ verba in
vv. 961.2.’ Headlam compares further the humorous use of ὠλεσίοικος,
ἀπωλεσίοικος: Bekker, Anecd. 25. 15 = Phryn. Praep. soph. 44. 3 de B.
ἀπωλεσίοικον μειράκιον κτλ. So δωματοφθορεῖν means the bringing of material
ruin upon the house as a result of great extravagance. Schütz correctly
observed that the main thought of the sentence is that Agamemnon is afraid
that such waste of valuable property would be regarded by the gods as wicked
arrogance. Hence his αἰδώς. Therefore he treads on the carpets with his bare
feet to avoid spoiling them, as far as possible, and also (for this αἰδώς, too,
comes into play) to show his respect for that which is appointed for the gods
and their worship, just as the Moslem takes off his shoes on entering a mosque.’
The reason for his caution is as clear as was earlier his profound reluctance.
And yet we meet comments such as the following: “The king’s vanity gives a
ready consent: “1 will e’en tread on purple rather than on vulgar shoe--
leather.” ... He allows his scruples to be removed one by one by Clyt., and
ends by a wretched compromise between piety and pride, in consenting to:
walk, in barbaric splendour, upon purple garments, but without his shoes"
etc. (Paley) or, with no essential difference, this: ‘He tries to resist, but the:
splendour of an oriental homecoming seduces him and he yields . .. Agamem-.
non is going through the process of temptation. He protests rather too often.

1 For the fact that anything dedicated to the gods may not be trodden on (as opposed’
to βέβηλα), cf., e.g., Callim. fr. 9. 235 f. Pf., where the laurel tree says: ἁγνὴ γάρ εἰμι, xov
πατεῦσί μ᾽ ἄνθρωποι, ἱρὴ γάρ εἰμι.
2 Neustadt's defence of σωματοφθορεῖν (Hermes, lxiv, 1929, 245 n. 1) has not been successful..
3 For the widespread prohibition of the wearing of shoes on holy ground or in certain
ritual acts (these prohibitions may be based on reasons differing widely from each other),.
cf. in general O. Gruppe, Griech. Mythol. 912 n. 6, Ad. Wilhelm, Oesterr. Jahresh. xiv, 1911,
179 (with bibliography) ; and for particular cases e.g. Frazer, The Fasti of Ovid, vol. ii. 237 f.,.
Marbach, RE xvii. 1240 f. (with bibliography). From among the instances it is worth.
singling out the injunction in the law of Ialysos for the sanctuary of Alektrona (about.
300 B.C.), Dittenberger, Syll. 338. 25, μηδὲ ὑποδήματα ἐσφερέτω μηδὲ ὕειον μηθέν.

430
 COMMENTARY line 950
and yields . . . (on 945:) These bound slaves—i.e. his shoes. The metaphor
shows the trend of his unconscious mind’ (Murray),! similarly H. W. Smyth,
Aeschylean Tragedy (1924), 164, M. Croiset, Eschyle (1928), 191: ‘Faiblesse et
vanité tout à la fois prévalent en lui. Il se défend mollement, il finit par se
laisser faire’ etc., much the same W. Nestle, ‘Menschl. Existenz . . . in d.
Trag. d. Aisch.’, Tübinger Beitr. z. Altertumsw., 23. Heft (1934), 38 f.; W. C.
Greene, Moira (Harvard Univ. Press 1944), 127. The reader’s understanding
of the whole play depends in great measure upon what he decides here. The
correct view has been concisely and effectively expressed by Otfried Müller,
Eumenides, p. 193, similarly by John Symmons,? and W. Sewell in the passage
quoted p. 372 n. 4. The essential features of Agamemnon’s behaviour have
been appreciated by Wilamowitz, Griech. Tragoedien, ii. 34 f., and P. Fried-
lander, Die Antthe, i, 1925, 17 f. Cf. on 811.
949. Hermann’s verdict ‘non est credibile φθείροντα post δωματοφθορεῖν ab
Aeschylo scriptum esse’ has impressed many subsequent editors, although it
is based on a mere prejudice.
πλοῦτον ἀργυρωνήτους θ᾽ ὑφάς : on this form of epexegesis cf. on 214.
950. τούτων μὲν οὕτω. A noteworthy example of the ‘genitive of respect’,
which has been particularly well dealt with by Matthiä, Griech. Gramm.
§ 342. 3 (cf. also K. W. Krüger, Sprachlehre, ὃ 47. 3 n. 3, Kühner-Gerth, i.
363 n. 11, E. Löfstedt, Syntactica, i, 1942, 129 n. 1)? Matthiä appears to me to
understand correctly E. Andr. 361 f. ἡμεῖς μὲν οὖν τοιοίδε: τῆς δὲ σῆς φρενός,
ἕν σου δέδοικα: ‘but with regard to your disposition’, whereas Hermann
unhappily combines the two κῶλα and translates ‘tuae mentis unum (mulierosi-
tatem) a te metuo’. The genitive τῆς δὲ σῆς φρενός is similar in function to the
well-known Περὶ δὲ τῆς ᾿Αθηναίων πολίτειας and to the expressions collected
from the Hippocratic writings by M. Kupferschmid, Zur Erklärung der
pseudoxen. 'A0. πολ., Diss. Hamburg 1932, 15 and to those cited by me from
Plato and from letters in Kolon u. Satz, ii. 326, 350; in all those instances a
short kolon of its own marks the transition to a new subject. Similarly e.g.
Thuc. 5. 18. 8 Σκιωναίων δὲ καὶ Τορωναίων καὶ Σερμυλιῶν καὶ εἴ τινα ἄλλην πόλιν
ἔχουσιν Abnvaioı, ᾿Αθηναίους βουλεύεσθαι περὶ αὐτῶν κτλ., Plato Laws 7. 794 ἃ
τῶν δὲ τροφῶν αὐτῶν καὶ τῆς ἀγέλης συμπάσης, τῶν δώδεκα γυναικῶν μίαν ἐφ᾽
ἑκάστηι τετάχθαι κτλ., Xen. Oec. 3. 11 τῆς δὲ γυναικός, εἰ μὲν... κακοποιεῖ, ἴσως
δικαίως ἂν ἡ γυνὴ τὴν αἰτίαν ἔχοι, Arist. De part. anim. 4. 13 p. 696 ἃ 21 αὐτῶν δὲ
τῶν πτερυγίων, τὰ ἐν τοῖς πρανέσιν ἔχει τὰ δύο ἔχοντα πτερύγια μόνον, well ex-
plained by Ingemar Düring, Aristotle’s De partibus animalium (Göteborg
1943), 214: ‘First comes in the genitive the matter concerned “as to the
collocation of the fins" . . . the genitive serves as a sort of heading to the
following clauses.’ Ag. 950 only differs from such passages* in that the colon

The quotations are taken from Murray’s notes on his translation; similarly in his book
Aeschylus (1940), 218, he writes ‘He would not have walked on the tapestries if left to him-
self, but secretly he longs to do so’.
2 See the note on his translation of 922: ‘This entry of Agamemnon must have produced
a fine spectacle on the stage. His piety, magnanimity, and modesty require no comment.’
3 The reference given by Sidgwick and others to S. El. 317 is not really helpful, for there
it is a case of the special use of the genitive with verbs of saying, cf. Kühner-Gerth, 1. 363 c,
Jebb on 5. Trach. 1122 and Phil. 440.
4 In the wider sense, A. Eum. 913 f. also belongs here; τῶν ἀρειφάτων δὲ... πρεπτῶν
ἀγώνων (transition to a new point, after the conclusion of what has gone before has been
431
line 950 COMMENTARY
τούτων μὲν οὕτω does not lead on to what follows, but concludes what has
gone before. This does not affect the nature of the genitive (= περὶ τούτων,
as already observed by Pauw). Emperius’ τοὐμὸν μὲν οὕτω still has its
adherents. But it could not possibly occur to Agamemnon to sum up what
he has said so far by referring to it as to ‘his own affair’ and contrast it with
what he has to say in favour of Cassandra. What is required here is something
like haec hactenus, cf. Prom. 500 τοιαῦτα μὲν δὴ ταῦτα, Hdt. 3. 12. 4 ταῦτα μέν
νυν τοιαῦτα, 9. 90. I οὗτος μὲν οὕτω ἀπενόστησε, S. El. 696 καὶ ταῦτα μὲν τοιαῦτα,
Ar. Plut. 8 καὶ ταῦτα μὲν δὴ ταῦτα (Birds 1304 τοιαῦτα μὲν τἀκεῖθεν), PI.
Laws 676 ἃ ταῦτα μὲν οὖν δὴ ταύτηι, and the like, cf. also p. 416, footnote.
Exactly what we should expect is found in the MS reading.’ τούτων μὲν οὕτω
is a dry, businesslike formula of transition of the kind with which we have
met in a speech of Agamemnon (830), cf. on 1045 f. and on 1393. Conington
(in his edition of the Choephoroe) rightly compares this phrase with Cho. 453
τὰ μὲν γὰρ οὕτως ἔχει, ‘a conventional form of dismissing a subject as suffi-
ciently dwelt on’. The same can be seen in Eum. 453: there Orestes concludes
the first, and as such sharply separated (cf. 443 f.), part of his speech with the
words ταύτην μὲν οὕτω φροντίδ᾽ ἐκποδὼν λέγω. Cf. also Septem 422 f. τούτωι
μὲν οὕτως... Καπανεὺς δὲ κτλ.
τὴν ξένην δὲ κτλ. The servant has now finished her task and Agamemnon
descends from the wagon ;? not until 956 f. does he begin to walk slowly to the
gate of the house. The words τὴν ξένην «rd. turn the attention of the audience
for the first time to the significance of the veiled woman now sitting alone
in the chariot. Up to now she will not have been noticed as anything but a
component part of the entering procession: she was not to be allowed to
divert attention from the main issue, the dispute between Agamemnon and
his wife. And even now the words τὴν ξένην serve up to a point to conceal her
identity. ‘Spectatorum animos non sinit praecoci Casandrae miseratione
permoveri. tandem ubi vaticiniis eius opus est a Clytaemnestra adpellatur
1035. This statement of the young Wilamowitz (Anal. Eurip. 200) seems to
me more in conformity with the words of the text than his later hypothesis
(Interpr. 171), which makes ‘Cassandra immediately recognizable to the
audience by her dress as a prophetess’ as soon as she enters in the chariot.
How far it was possible at first to recognize anything at all of the ‘dress of the
prophetess’ (1264 ff.) in the heavily veiled and presumably cowering figure,
we cannot tell. If the poet had wished to make her identity clear to the
audience before the beginning of the Cassandra scene, he would presumably
have written 950 in a different form.
951. τὸν κρατοῦντα κτλ. This general maxim, which is to serve as a setting
marked by τοιαῦτα oovorı) depends, it is true, from a purely grammatical standpoint on
the ἀστύνικον which follows, but at the same time, as Sidgwick rightly notes, it functions
principally as a ‘genitive of respect’. The examples of the genitive of respect which E.
Nachmanson, APATMA Martino P. Nilsson dedic. (Lund 1939), 328 ff., has quoted from
the Hippocratic writings, although they may be somehow related to those discussed here,
do not mark the heading of a paragraph or conclude a section of a speech with a sum-
marizing formula.
ı Von der Mühll, RE Suppl. vii. 728. 29, notes that the formula of transition in Circe’s
speech μ 37 ταῦτα μὲν οὕτω πάντα πεπείρανται is equivalent to καὶ ταῦτα μὲν τοιᾶντα,
2 ] follow Wilamowitz. It seems far more likely that Agamemnon’s shoes are taken off
while he is still seated than that he should dismount at 944 as Verrall assumes. Wecklein
thinks that the king descends at 946.

432
 COMMENTARY line 959
for the particular request of the previous sentence, is introduced without a
conjunction; if one were used, a γάρ or γοῦν would be appropriate (cf.
Denniston, Particles, xlvi. 3). Similarly, in terser maxims, N x15 ἀλλ᾽ ἀκεώ-
μεθα θᾶσσον’ ἀκεσταΐ τοι φρένες ἐσθλῶν, S. Ant. 1194 f. τί ydp σε μαλθάσσοιμ᾽
ἂν ὧν ἐς ὕστερον ψεῦσται φανούμεθ᾽; ὀρθὸν ἁλήθει᾽ ἀεί, E. Or. 233 f. ἦ κἀπὶ γαίας
ἁρμόσαι πόδας θέλεις, χρόνιον ἴχνος θείς; μεταβολὴ πάντων γλυκύ. An instance
very close in form to these brief general maxims can be seen in Ag. 637
χωρὶς ἡ τιμὴ θεῶν.
954 f. w. x. ἐξαίρετον ἄνθος, στρατοῦ δώρημα: the use of ἐξαίρετος here and
Eum. 402 ἐξαίρετον δώρημα and similarly in Herodotus is a continuation of
that in Homer: B 226 f. γυναῖκες... ἐξαίρετοι, ds τοι Ἀχαιοὶ πρωτίστωι δίδομεν,
εὖτ᾽ av πτολίεθρον ἕλωμεν. The verb ἐξαιρεῖν occurs frequently from Homer
onwards in the corresponding sense.
Not with a single word does Aeschylus indicate whether the king has any
other feeling towards his captive beyond that of pity for her fate. There is
no room in this tragedy for the idea ἔρως ἐτόξευσ᾽ αὐτὸν (Agamemnon) ἐνθέου
κόρης (E. Tro. 255), arsit Atrides medio in triumpho virgine rapta (Horace,
Odes 2. 4. 1 f.). It is quite hopeless to attempt to convince those who are in
the habit of reading into any work of poetry what they want to find there
and, when challenged to produce some evidence from the text, will fall back
on the argument ‘but is it not simply human nature . ..?' Unless we allow a
great poet to draw a line precisely where he chooses, we shall never be able
to understand his work.
956. καταστρέφειν, 'make subject': Agamemnon sharply describes the
νικᾶσθαι (941 f.) for what it is for him (Schneidewin has seen the point). For
the addition of an epexegetic infinitive to κατέστραμμαι Blomfield compares
Hdt. 7. 51. 1 Κῦρος... ᾿Ιωνίην πᾶσαν... κατεστρέψατο δασμοφόρον εἶναι
Πέροηισι. It is true that the addition of ‘to obey you in this’ could, in itself,
serve to tone down the κατέστραμμαι; but as the audience realizes more
clearly every moment the symbolic meaning of the action, the strong word
will go on ringing in their ears. ‘I have been borne down,’ says the king, before
he walks to his house. We shall not see him again.
957. πατῶν should not be watered down. Headlam’s translation ‘trampling
purples’ is good. Agamemnon once more gives vigorous expression to what
he has to do, desecration. Cf. Daube, Rechtsprobleme, 115 nn. 71, 72, and above
on 372. Even in his last words Agamemnon is still making clear his reluctance
to do what the queen demands.
958. ἔστιν θάλασσα: Clytemnestra begins in a tone of magnificent emphasis
and displays at the same time an almost scornful superiority (as in the
following question), as if Agamemnon in his lack of spirit were afraid that
the sea could suddenly dry up.
τίς δέ νιν κατασβέσει; cf. Prom. 532 παρ᾽ ᾿Ὠκεανοῦ πατρὸς ἄσβεστον πόρον,
‘the multitudinous seas’.
959. πορφύρας here ‘purple’, not, as Dindorf in the Thesaurus and L-S take
it, ‘purple-fish’.
ἰσάργυρον : Salmasius, who recognized this word in the letters of the MSS,
compared Theopompus (F Gr Hist, 115 F 117) apud Athen. 12. 526 c, where, in
discussing the extensive use of purple garments by the people of Colophon,
the author says: ἰσοστάσιος yap ἦν ἡ πορφύρα πρὸς ἄργυρον ἐξεταζομένη.
4872.2 εἴ 433
line 960 COMMENTARY
960. κηκῖδα. Blomfield’s rendering ‘id quod tingit’ still has a strong influence,
even now; it rests upon the later usage of the word. Schütz rightly says:
‘suspicor antiquiorem fuisse huius vocabuli notionem, qua quicquid pro-
pullulat, πᾶν τὸ κηκῖον, significaret’; this meaning alone is suitable in the
remaining examples from Tragedy: Cho. 268, 1012 (Conington compares with
φόνου κηκίς 5. Phil. 697, 784), S. Ant. 1008. For the primary meaning of
κηκίω = ‘I begin to sweat’ cf. Wackernagel, Philol. Ixxxvi, 1931, 137 f.
παγκαίνιστον occurs only here. In Tr it is glossed with διόλου νεάζουσαν.
Blomfield: ‘Difficile est vim huius vocis accurate exprimere. καινίζειν est re
quavis primum uti... et hinc παγκαίνιστος est purpura, cuius talis est copia,
ut ea semper recenti quivis uti possit.’ It is true that καινίζειν, a verb of not
very frequent occurrence, is used predominantly in the sense given by
Blomfield or something like it and not in the purely factitive sense; cf. on
1071. This is not surprising since 'throughout almost all periods in the
language -ifew appears to have been used as suffix of-all-work in forming
denominative verbs' (A. Debrunner, Griech. Wortbildungslehre, 129). On the
other hand, there was always the possibility of taking kawi£euw (or a deriva-
tive of it) as factitive, for ‘the extremely frequent factitive meaning of -ἔζειν
competes with -οὔν᾽ (Debrunner, 134; ἃ factitive καινίζειν would exactly cor-
respond to oodilew ‘make wise’ [used only in the passive in the classical
period], which he quotes). In view of this, the rendering 'quite renewed,
ever new' (Passow), 'constantly renewed' (Linwood), 'ever renewed, ever
fresh' (L-S) appears to be linguistically correct; the idea of the supply of
purple constantly renewing itself in the inexhaustible sea is in keeping with
the spirited style of Clytemnestra’s speech. Paley, who translates ‘wholly
renewable', wrongly introduces the idea of the purple fabrics recovering their
original hues under the influence of the sun's rays. Karsten (with the in-
admissible conjecture βαφήν) understood παγκαίνιστον in the active sense as
‘omnia renovans' ; this is, in itself, linguistically possible (in general cf. on 12,
in particular cf., e.g., πάμφθαρτος, πανάλωτος). Wilamowitz, too, seems to
have taken it like this: ‘Es birgt des goldeswerten Purpursafts genug, zu
färben immer neuer Prachtgewande Schmuck’, ie., if I understand him
correctly, he takes παγκαίνιστον to mean 'ever producing anew' (— ever
providing with fresh purple). In view of the prosperity of the house, stressed
throughout the passage, there can be no thought of restoring the colour to
fabrics which have become faded through use (πορφυρίδες ἐξίτηλοι). I prefer
the passive meaning. 'Tamen vis vocabuli incerta' (van Heusde). In any
case its derivation from καινίζειν is indisputable. To derive it from πάγκαινος
(not attested) and take it as a superlative in -«eros (‘novissimum’ Klausen)
is in itself doubtful, nor does it give a satisfactory sense. The type of verbal
adjective in -ros, compounded with παν-, is in earlier Greek poetry! repre-
sented by the following instances: πάγκλαυτος (Aesch., Soph.), παγκόνιτος
(Soph.), πάγχριστος (Soph.), πάμβοτος (Aesch.), πάμμεικτος (Aesch.), πάμπληκ-
Tos (Soph.), πάμπρεπτος (Aesch.), πάμφθαρτος (Aesch.), πάμφλεκτος (Soph.),
πάμφυρτος (Soph., παντόφυρτος Aesch.), πανάλωτος (Aesch.), mavBákpvros
(Aesch., Soph., Eur.), πανδείμαντος, πάνδυρτος (Aesch., Soph., Eur.), πάνεφθος
! No example in Homer. It is significant that Aeschylus transposes the Homeric
mávaypos into πανάλωτος (above p. 190). The Homeric παν- compounds are examined and
assessed etymologically by Hoenigswald, Language, xvi, 1940, 183 ff.

434
 COMMENTARY line 961
(this perhaps does not strictly belong here), πάνθυτος (Soph.), πάνουρτος
(Soph.). It is clear that Aeschylus and Sophocles have a very strong liking
for words of this kind ;' Euripides takes over only a very few of the older
words coined in this way and does not appear to have coined a single new
one. This is in harmony with a more general tendency to be noticed in him.
In the formation of new adjectives made up of ‘adverbial’ and nominal
components (including the verbal adjective), he limits himself to very few
types in the first element (cf. W. Breitenbach, ‘Unters. z. Sprache der Eurip.
Lyrik’, Tübinger Beitr. z. Altertumsw., zo. Heft, 76). In Pindar and Bacchy-
lides the type παν... ros is not found at all.
961. It has long been realized that the MS reading οἶκος κτλ. cannot be
retained (Hermann's defence is unfortunate; Headlam tries the explanation
‘the house affords no store of these’, which is not sufficiently supported by
Theocr. 22. 222). οἶκος ὑπάρχει ἔχειν could only mean either ‘the house is there
to have or hold something’ or ‘the house is there for someone to hold’.
Porson’s οἴκοις is satisfactory and probably correct. For the use of an in-
finitive with ὑπάρχει (as with ἔστι, πάρεστι and the like) cf. 5. El. 1340 ὑπάρχει
γάρ σε μὴ γνῶναί τινα, ‘Domui suppeditat horum? (i.e. purpurarum) habere’.
It has been asserted again and again that the idea of horum affatim habere
(Schneidewin) is to be expected here; accordingly Karsten and Wilamowitz
read ἅλις instead of ἄναξ, although the sentiment expressed by this vocative
is just as important as in 907, and other scholars have made other alterations ;
Housman (C.R. iv, 1890, 9), who with Tycho Mommsen read ὄγκος for οἶκος,
joined the critics who raised the objection to Porson’s οἴκοις that ‘ ὑπάρχει
τῶνδε fails to convey the notion of abundance’. The truth is that the notion
of abundance is not at all essential here ; it is not primarily a question of how
great a supply of purple fabrics the royal house possesses (if it has any, it
will certainly have a good supply), but of their being kept in the house, and
not purchased only when needed. From remotest antiquity it has been a
matter of justifiable pride? to be in a position ui domo sumeret neu foris
quaereret (Plaut. Bacch. 648, similar examples are frequent in Plautus, the idea
is also transferred to the mental plane as in E. Tro. 652 f., Fragm. trag. adesp.
116 N.). For οἴκοθεν = ‘from my store’ see Leaf on Ÿ 558 and M. Lejeune,
Les Adverbes grecs en -θεν, 86. It is in accordance with this idea that the
axiom (or double axiom) in Hesiod, Erga 364 f. οὐδὲ τό γ᾽ ἐν οἴκωι κατακείμενον
ἀνέρα κήδει" οἴκοι βέλτερον εἶναι, ἐπεὶ βλαβερὸν τὸ θύρηφιν has been understood
by a part of the scholia (Tzetzes) : χαλεπὸν γὰρ τὸ μὴ ἀποκεῖσθαι ταῦτα ἐν τῶι
οἴκωι, ἀλλ᾽ ἐκ τῶν θύραθε καὶ ἔξω ταῦτα ὠνεῖσθαι (Proclus and Wilamowitz, ad
loc., take it differently).* Priam says (X so): ‘if my sons are still alive,
χαλκοῦ τε χρυσοῦ 7” ἀπολυσόμεθ᾽ " ἔστι γὰρ ἔνδον ”, cf. Καὶ 378f. Headlam, who

1 The ἀπανδόκευτος of Democritus (fr. 230) does not belong to this category, for it is
derived, not from δοκεύω, but from πανδοκεύω,
2 This use of the partitive genitive is not classical in Latin but it is old and very prolific
(cf. Löfstedt, Syntactica, i, and ed., 142 ff.).
3 ‘Wealth consists of the accumulations stored up within the walls of the house’ (A. C.
Pearson on E. Phoen. 552, with supporting quotations). Beazley recalls Petronius 38. 1 nec
est quod putes illum quicquam emere. omnia domi nascuntur : lana, credrae, piper, etc.
* The view held by Tzetzes seems to me to have the advantage of linking the lines much
more firmly with their context, especially with 366 f., than if they were taken to mean
*outside it is easily lost.'
435
line 961 COMMENTARY
cites the latter passage, also quotes Ar. Peace 521 f. πόθεν ἂν λάβοιμι ; . . . οὐ
γὰρ εἶχον οἰκόθεν and Rhes. 170, where Dolon replies to Hector’s offer of gold:
ἀλλ᾽ ἔστ᾽ ἐν οἴκοις " οὐ βίου σπανίζομεν, similarly 178 ἔστι χρυσὸς ἐν δόμοις.
As before in the stichomythia, Clytemnestra makes use of two arguments
to reinforce a single contention. First: ‘the sea produces endless masses of
purple (why are you so fearful of wasting and using up some of it?)’, next:
‘and as for our house, it keeps, thank heaven, a supply of purple rugs (if a
few are spoiled, it is not necessary to buy new ones immediately)’. Instead of
Porson’s οἴκοις, Headlam’s οἴκοι might possibly be considered, ‘giving the
construction ὑπάρχει (ἡμῖν) ἔχειν τῶνδε᾽, but οἴκοις seems simpler; it also
makes the transition to the new point more marked.
963. warnopov: she picks up his word πατῶν (957) ; it did not escape her that
it contained a protest against the desecration.
964. προὐνεχθέντος : cf. on 201.
Almost all the older commentators took ev χρηστηρίοις δόμοισι together, so,
too, e.g. Verrall, Plüss, Platt, Mazon, Murray (transl.). On the other hand,
following the example of Butler (in Peile), Conington recommended separating
χρηστ. (noun) from δόμοισι and making the latter depend on προὐνεχθέντος,
‘the constructions προὐνεχθέντος δόμοις ἐκ χρηστηρίων and 7. ἐν χρηστηρίοις
being mixed’. It has been taken in this way by Heath, Paley, Kennedy,
Wecklein, Lewis Campbell, Blaydes, Headlam. Wecklein notes: ‘ δόμοισι is,
as the order shows, dependent on προὐνεχθέντος.᾽ Assuming provisionally
that δόμοισι ἐν xpnormpioıs went together, I have examined dialogue in
Aeschylus and Sophocles for examples of an order of words corresponding to
the type δόμ. mpoûv. ἐν ypnor., i.e. the insertion of a word (or of several) into
an expression consisting Οἱ ἃ noun plus preposition and its adjectival attribute.
The result is as follows. Insertion of a participle: Eum. 806 λιπαροθρόνοισιν
ἡμένας ἐπ᾽ ἐσχάραις, 851 ὑμεῖς δ᾽ es ἀλλόφυλον ἐλθοῦσαι χθόνα, Aj. 575 f. διὰ
πολυρράφου στρέφων πόρπακος ... σάκος, Oed. R. 1073 f. ὑπ᾽ ἀγρίας ἄιξασα λύπης.
Main verb inserted: Suppl. 198 ἐκ μετωποσωφρόνων' ἴτω προσώπων, Aesch.
fr. 78. 2 ἐν τριγώνοις ἐκπεραινέτω ῥυθμοῖς, fr. 99. το ἐκ τῶν μεγίστων δ᾽ ἠρξάμην
φυτευμάτων, Trach. 273 an’ ἄκρας ἧκε πυργώδους πλακός, 1242 σὺ γάρ μ᾽ ἀπ᾽
εὐνασθέντος ἐκκινεῖς κακοῦ, Oed. C. 100 f. κἀπὶ σεμνὸν ἐζόμην βάθρον τόδ᾽ ἀσκέ-
παρνον, ἵτ. 588 N. (= 646 P.). 4 ἐν γὰρ βραχεῖ καθεῖλε κὠλίγωι (?) χρόνωι. Main
verb and other parts of speech: Cho. 89 f. παρὰ φίλης φίλωι φέρειν γυναικὸς
avöpi,” Eum. 634 f. ἐν δ᾽ ἀτέρμονι κόπτει πεδήσασ᾽ ἄνδρα δαιδάλωι πέπλωι, 857
ὅσην map’ ἄλλων οὔποτ᾽ ἂν σχέθοις βροτῶν. Various parts of speech (no verb)
inserted: Ag. 923 ev ποικίλοις δὲ θνητὸν ὄντα κάλλεσιν βαίνειν, Ant. 463 ὅστις γὰρ
ἐν πολλοῖσιν ὡς ἐγὼ κακοῖς ζῆι, Trach. 301 f. ἦσαν ἐξ ἐλευθέρων ἴσως ἀνδρῶν, 683
χαλκῆς ὅπως δύσνιπτον ἐκ δέλτου γραφήν. For the sake of completeness I add
a few examples in which the enclosing phrase does not contain an adjective
proper but a pronoun or pronominal adjective although, strictly speaking,
the group does not belong here: Trach. 304 πρὸς τοὐμὸν οὕτω σπέρμα, 1133 ἐξ
ἐμῆς θανεῖν χερός, Oed. C. 988 f. ἐν τοῖσδ᾽ ἀκούσομαι κακὸς γάμοισιν, Trach. 144 f.

! Porson’s completely satisfactory emendation; ‘a thoroughly Aeschylean ῥῆμα γομφο-

rayés’ (Headlam).
2 For the ‘association of words which are the same or of the same stem’ and the ‘associa-
tion of parts of speech of the same kind’ (both occur here), cf. F. Leo, Nachr. Gölt. Ges.,
1895, 433.
436
 COMMENTARY line 965
τὸ γὰρ vealov ἐν τοιοῖσδε βόσκεται χώροισιν αὑτοῦ. From these examples it
appears that, while the order in A. Suppl. 199 ὄμματος παρ᾽ ἡσύχου is normal
as is also a hyperbaton like S. El. 267 f. ὅταν θρόνοις Αἴγισθον ἐνθακοῦντ᾽ ἴδω
τοῖσιν πατρώιοις, a group such as * ὅταν θρόνοις Αἴγισθον ἐνθακοῦντ᾽ ἴδω Ev τοῖς
πατρώιοις, i.e. the sequence ‘Noun, verb (or some other part of speech),
preposition, adjectival attribute',! would be without parallel, at least in the
dialogue of the two older tragedians. In all the examples quoted above the
adjectival attribute comes at the beginning.? Accordingly it must be regarded
as at least very improbable that in Ag. 964 we are to assume a word-order
which would be unique in the dialogue of Aeschylus and unknown to that of
Sophocles. Therefore 8ópow: must be taken as referring to the house to
which the oracle is communicated.
965. ψυχῆς... τῆσδε. Ahrens (587) says that linguistic usage? requires that
this phrase should be referred not 'to Agamemnon, the person spoken to',
but ‘to the speaker, Clytemnestra' (Pauw had already commented ‘ ψυχῆς
τῆσδε pro ἐμοῦ "). It is of course impossible here to refer ψυχῆς τῆσδε to Cly- -
temnestra*—Ahrens’s alteration of the end of the line is fantastic—but that
τῆσδε cannot refer to the person spoken to is correct. Agamemnon, however,
is not spoken to here; translations such as 'means for getting thy dear life
safely back again' (Paley) or 'deines teuren Lebens Preis' (Wilamowitz) are
not admissible. Clytemnestra is certainly not addressing the sentences 963-7
(πολλῶν... κυνός) to her husband (in the sense that she looks at him and
talks to him), nor indeed to anyone in particular (for it is difficult to imagine
that she is talking to the Chorus). She speaks as if soliloquizing, feigning an
involuntary surge of passionate devotion to her husband. In 968 she turns
again to Agamemnon. From her attitude in 963 ff. we may perhaps infer that
the idea of a vow becomes in her mind so powerful (or at any rate so she
pretends) that for a moment she addresses herself to the gods. For passages
! Not so, when the attribute is a genitive: Sept. 185 βρέτη πεσούσας πρὸς πολισσούχων
θεῶν.
2 Seeking to come to a decision about the text of A. Suppl. 638 (cf. above on 437 ff.,
p. 228 n. 1) I subsequently extended my investigation to the lyrics (including all the
anapaests) of the three tragedians. The result was that for the type of hyperbaton with the
word-order ‘Noun, inserted phrase, preposition plus adjective’ there is even in the lyrical
parts of Tragedy, if I have not missed anything, only a single reliable example, i.e. A. Suppl.
638 τὸν ἀρότοις θερίζοντα βροτοὺς ἐν ἄλλοις. Considering the great uncertainty of the text,
I cannot venture to take into account the passage in the opening anapaests of the same
chorus, 629 τέρμον᾽ ἀμέμπτων πρὸς ἅπαντα. Again, for Prom. 1059 f. (anapaests) τόπων pera
ποι xwpeir’ ἐκ τῶνδε I have not found any other parallel, although in the case of the pronoun
greater latitude might have been expected than with an ordinary adjective. In spite of
this result there is no reason to doubt the text of Suppl. 638 τὸν apdraıs . . . ἐν ἄλλοις on
account of the word-order. For in choral lyric this type of word-order is not uncommon,
cf. e.g. Pind. Ol. 7. 91 ἅ re of πατέρων ὀρθαὶ φρένες ἐξ ἀγαθῶν Expeov, P. 3. 38 ἀλλ᾽ ἐπεὶ τείχει
θέσαν ἐν ξυλίνωι σύγγονοι κούραν, 4. 187 ἅλιξιν εὑρέσθαι σὺν ἄλλοις, 9. 68 θαλάμωι δὲ μίγεν ἐν
πολυχρύσωι Λιβύας, το. 2 πατρὸς δ᾽ ἀμφοτέραις ἐξ ἑνὸς... γένος... βασιλεύει. So Aeschylus
has in a chorus, exceptionally, allowed himself a word-order (perhaps felt to be bold)
which was not considered out of place in lyric poetry but was in general avoided by
Tragedy.
3 Ch eg. Ag. 1438 γυναικὸς τῆσδε, Diktyoulkoi (Pap. Oxy. 1261), col. 2. 12 f. dv μητρὶ καὶ
πατρὶ τῶιδε. For this use cf. Schadewaldt, ‘D. Niobe des Aisch.’, Sitzgsb. Heidelb. Akad.,
1933/4, 3. Abhdl., 14 n. 1.
* It probably is only a slip when Wackernagel, Unters. z. Homer, 220, translates the
passage : ‘means of saving my life’.
437
line 965 COMMENTARY
in Aeschylus which are half-way between a prayer and a soliloquy see on
1648.
xöpiorpa: ‘means for getting back’ (Paley), ‘means of rescue’ (Wacker-
nagel, see previous footnote, cf. also Reinhardt, Hermes, lxix, 1934, 240), in
certain circumstances simply ‘the bringing’, ‘the rescue’. This is Körte’s
view, Hermes, lxviii, 1933, 259, where he discusses the word κόμειστρα in the
new fragment of Aeschylus’ Niobe, cf. also Pfeiffer, Philol. Ixxxix, 1934, 11.
Paley compares E. Her. 1387.
μηχανωμένης of the MSS is one of the frequent examples of assimilation to
the preceding word. Now that it is clear that in 964 δόμοισι refers to the
recipient of the oracular utterance (see above), there may be an inclination to
adopt the conjecture of Auratus and Scaliger, μηχανωμένη. But the thought
would be distorted if she said: ‘I would have vowed it in my efforts to save
him, if it had been prescribed by oracles.’ Her consulting the oracle is part
of her μηχανᾶσθαι for the purpose of his safe homecoming, or, to express it
grammatically, the μηχανᾶσθαι is coordinate or subordinate to the zpoeve-
χθῆναι. Consequently she must say : ‘I would have promised it, if it had been
prescribed to me by oracles when I was thinking of a means of saving him.’
So the emendation of Abresch and Hermann, μηχανωμένηι, is preferable. ‘The
dative unxavwuérm depends as well as δόμοισι on προὐνεχθέντος᾽ (Paley) ; the
queen is in the absence of the master the representative of the δόμοι. So this
is an example of the σχῆμα ᾿Ιωνικόν, cf. Elmsley on E. Med. 961 f., Wila-
mowitz on E. Her. 162.
For the thought of 963-5 cf. 933 f. In S. Trach. 610 ff. Deianeira asserts that
she had vowed, in the case of Herakles' safe return, the gift of a splendid
garment which he was to put on when he made the thank-offering. Cf.
Catullus 66 (Callimachus). 33 ff. aique ib: me (the lock) cunctis pro dulci
coniuge divis non sine taurino sanguine pollicita es, si reditum letulisset.
966. ‘Verba ἐς δόμους cum imagine non quadrant’ says Karsten ; Schütz had
already raised objection to it and proposed ἐς κλάδους, Wilamowitz reads
ἵκετο στέγης. These critics are trying to force into the passage a clearness of
the pictorial image which is foreign to it. Vahlen, Opusc. i. 299, puts the
passage into its proper setting ; he points out that in ancient poetry the image
which serves as a comparison and the thing compared are not always (as
modern taste prefers) kept strictly apart,’ that, on the contrary, the image
frequently assimilates elements characteristic of the thing compared (here
the ἱκέσθαι ἐς δόμους: of Agamemnon). This sense is given correctly by
Wecklein, ad loc., and Verrall ('the comparison and the thing compared mix
together'); cf. p. 542. The predilection for transferring to the comparison
peculiarities of the thing compared, even when they impair the pictorial
quality of the image, can be noticed sometimes in Aeschylus, cf. on 1or1.
The blurring of the picture would not be felt so strongly by the ancient
audience, because in Tragedy δόμοι, as is well known, meant predominantly
not the material structure made of stones, the building in itself, but the
whole notion of house and household. The life of the household and its
1 Cf. also H. Frankel, Gott. gel. Anz., 1928, 274, and Wilamowitz in his commentary on
A. Cho. 246—63 (p. 180).
2 Cf. the passage quoted a few lines above, S. Trach. 610f. εἴ wor’ αὐτὸν ἐς δόμους
ἴδοιμι σωθέντ᾽ ἢ κλύοιμι.

438
 COMMENTARY line 969
dwellers will, if the root is still sound, be protected by the leafy shade which
once more comes to the house. The image is presumably derived from a vine
or a fig-tree, which, with a little help, spreads a covering of leaves over the
farmstead ; cf. Peile and Paley.
967. σκιὰν ...ceipiou κυνός : as σκιά contains the idea ‘defence, protection’,
it correspondingly takes a genitive, with a bold conciseness of expression.
σειρίου κυνός occurs also in Soph. fr. 735 N. (= 803 P.). The adjectival use of
σείριος (first in Hesiod, Erga 417 σείριος ἀστήρ) meaning the dog-star is com-
paratively rare, cf. Gundel, RE iii A. 314.
969. The emendation μολόν is generally accepted. But it is doubtful whether
onpaives can be retained, or whether it is necessary to accept Karsten's
conjecture, σημαίνει. In the latter case μολόν, which supplies the content of
the σημαίνειν, would have to refer to the subject θάλπος. For this construc-
tion there appears to be no incontrovertible parallel with the particular word
σημαίνειν, but exactly analogous instances justify it, as has been shown on
292 f. It is more difficult to decide whether in Aeschylus the construction σοῦ
μολόντος . . . σημαίνεις is probable. Wecklein refers to his note on 648 (653
Weckl.). But this is misleading, for the examples of anacoluthon which he
quotes there contain simply a participle in the nominative, whereas here it is
a question of a participle in a genitive absolute the subject of which would be
the same as that of the main verb. Kühner-Gerth, ii. 79 n. 4, deny without
any reservation the occurrence of such a construction. A more cautious
opinion is given by K. W. Krüger, Griech. Sprachlehre, § 47, 4 n. 2 at the end,
and particularly in his note on Thuc. 2. 83. 3, where Classen-Steup's com-
mentary is also helpful. In the Hellenistic papyri, from the third century B.c.
onwards, the construction is extraordinarily common : παραγενομένου δὲ αὐτοῦ
. εὗρεν Πτολεμαῖον and the like, and the place of the genitive absolute is
mostly at the beginning of the sentence, cf. E. Mayser, Gramm. d. griech.
Papyri aus der Piolemäerzeit, ii. 3. 68 ff. Among the editors of Aeschylus,
some hold that the construction is permissible, e.g. C. G. Haupt, ad loc., who
says: 'genitivorum absol ad subiectum pertinentium exempla haud rara
videsis apud Grammaticos', but unfortunately does not quote any ; Hermann
was clearly of the same opinion, for he reproaches H. Voss ‘quod non Graece
dictum esse putaret σοῦ μολόντος σημαίνεις '. Paley does not consider that it
needs a comment at all. What principally induces me to retain ompaives is
the usage of Thucydides. Of the relevant passages I select only one: 3. 13. 7
βοηθησάντων δὲ ὑμῶν προθύμως, πόλιν τε προσλήψεσθε κτλ. On this Classen-
Steup comment: ‘This second part of the alternative [βοηθησάντων κτλ.] is so
much the dominating idea here and indeed throughout the whole speech,
that it is put into the genitive absolute, although the subject of προσλήψεσθε
is the same as that of Bondnoavrwv.’ This comment could be applied almost
word for word to καὶ σοῦ μολόντος κτλ. With regard to this particular case it
must be added that the kolon σοῦ μολόντος corresponds to the preceding (966)
ἵκετ᾽ ἐς δόμους, 1.6. it stands too firmly on its own feet to take on the form of a
participle in the nominative subordinated to onpaives; and moreover it
serves, if not grammatically, certainly in its thought, as a basis for all that
follows up to the end of 972. In accordance with a well-known Aeschylean
practice, the conclusion (972) of this section, dvópós . . . δῶμ᾽ ἐπιστρωφωμένου,
harks back with only slight variations to its opening words (968) καὶ σοῦ
439
line 969 COMMENTARY
μολόντος δωματῖτιν ἑστίαν. Lastly, it seems to me that the hypocritical
adulation comes out much more powerfully if the king himself is the giver of
the sign instead of his arrival being made only the condition or circumstance
and θάλπος the subject.
970 f. This gives the other side of the picture of 969. We have met the
‘shadow of the flooding season to bring coolness in summer’ side by side with
the ‘warm, dry corner in winter time’ in the sequence of metaphors of pro-
tection and preservation quoted above on 899 ff. (p. 410) from an Egyptian
hymn. Another parallel, also from oriental poetry, has been quoted by
Schneidewin from Goethe’s translation of an Arabian poem (Noten zum
Westöstlichen Divan, ‘Araber’): ‘Sonnenhitze war er am kalten Tag; und
brannte der Sirius, war er Schatten und Kühlung.’
Here, too (cf. on 856, 869, 890 f.), Headlam finds ‘an ironical side-reference
to her realintention. ὄμφαξ was used in the sense of a "virgin" as well as of an
unripe grape’, etc.; G. Thomson has elaborated these unwarranted psycho-
logical niceties.
971. “ἤδη est 1am, referturque ad praecedentia’ (Hermann).
972. ἀνδρὸς τελείου : the meaning cannot be determined with certainty.
Probably Schneidewin is right : ‘he who has the τέλος [in the sense explained
on 105] and without whom the house is ἀτελής '; the second part of this
explanation had already been given by Blomfield. Verrall irrelevantly intro-
duces one of his double meanings (a secondary reference to ‘the perfect
victim, fit for the sacrifice”). The epithet is very suitable to the king and
lord of the house, but not commonly used in this way. It is clearly chosen
for the sake of the echo in Ζεῦ τέλειε. It is followed then (973 f.) by τέλει
and τελεῖν.
At 972 the door of the palace shuts behind Agamemnon; Clytemnestra,
who has accompanied him, pauses a moment outside and speaks her prayer,
before she, too, goes in (Wilamowitz).
973. Ζεῦ τέλειε. The cult of ‘Zeus, the Fulfiller, Accomplisher’ was wide-
spread in the Greek world. A list is given by Höfer in Roscher’s Lexikon d.
Mythol. v. 255 (see also A. B. Cook, Zeus ; the relevant passages can be found
with the aid of the index of vol. ii. 1338), the passages in poetry ibid. 256. 66 ff.
The meaning in A. Suppl. 525 f. is the most comprehensive (cf. Philol. Ixxxvi,
1931, 12 n. 30). In the present passage the explanatory imperative lays
emphasis upon the fulfilment of definite wishes and prayers. For the phrase
as a whole cf. Theognis 341 ἀλλὰ “Ζεῦ, réAeaóv μοι... . εὐχήν. In a concluding
prayer Ζεῦ réAew is also found in Pind. Ol. 13. 116. Cf. further the prayer in
Sept. 116 f. ὦ Ζεῦ πάτερ mavreMs . . . ἄρηξον κτλ. and Pind. N. 1o. 29 Ζεῦ
πάτερ, τῶν μὰν ἔραται φρενί, avyá ot στόμα" πὰν δὲ τέλος ἐν τὶν ἔργων. The form
of the prayer Ζεῦ τέλειε.. . . τέλει is the same as e.g. Menander, Epitr. 523 (587
Koerte, 3rd ed.) Ζεῦ σῶτερ, εἴπερ ἐστὶ δυνατόν, σῶιζέ με. Ζεῦ Ζεῦ also Cho. 246,
as e.g. E. Hipp. 1363. For the repetition, in general, of the name of the god
invoked cf. Norden, Commentary on Virgil Aen. 6. 46 (3rd ed. p. 136). ‘The
Agamemnon has more repeated vocatives than any other Greek tragedy’,
states I. A. Scott, Am. Journ. of Phil. xxvi, 1905, 37.
Clytemnestra alludes, in veiled but unmistakable terms, to the τέλος upon
which her whole mind is set. In dealing with the god, on the threshold of her
fatal deed, she can do naught else; this is in no way inconsistent with her
440
 WHY DOES AGAMEMNON YIELD?
behaviour on other occasions, cf. on gır. For the relation of the prayer
973 f. to the prayer or wish at the end of Agamemnon's speech (854) cf. on 500.
974. There is a similar enigmatical expression in Cho. 780, where, as here, the
success of the decisive action is entrusted to the gods. Here it is clear from
the words that Clytemnestra feels completely certain. She does not doubt
for a moment that Zeus, the guardian of retributive justice, intends to let
Agamemnon be murdered as a punishment for what he has done.
τελεῖν : future.
For the tone in which the poet required 973 f. to be spoken cf. on 1236.

Agamenmon abhors the impious and sacrilegious act which Clytemnestra

wants him to commit ; his whole nature revolts against it from the beginning
to the end. He loathes what eventually he is compelled to do and feels de-
graded by it, as a warrior, as a Greek, as a man respectful of the gods. Even
after yielding, he endeavours to mitigate the offence as much as possible and
emphasizes his unalterable αἰδώς. Why, then, does he yield at all? It would
be a poor excuse to say that what happens to the king in the carpet scene is
meant to be a symbol of his physical overthrow, and since the murder was a
foregone conclusion, settled once and for all in the saga, the former defeat
was as inevitable as the later. It is the poet’s main task to give fresh plausi-
bility to the old story which he has chosen for his subject. The clue to
Agamemnon’s behaviour must be found in the play itself. As he nowhere
reveals his motives explicitly, we must draw our inferences from the general
picture of his personality. This may seem a risky undertaking and exposed
to the fallacies of subjective interpretation. All that a conscientious reader
can possibly do is to try to outline the characteristic features as they present
themselves to his mind. Other readers must be allowed to reach different
conclusions, provided, however, that they do not neglect or distort any piece
of clear evidence in the words of the play.!
As it seems to me, one, though not the most important, reason why
Agamemnon gives way is his reluctance to get the better of a woman. When
he realizes that she has set her heart on reaching her end, his resistance falters
(940-5). In this, as in everything else, he proves a great gentleman, possessed
of moderation and self-control (the first reaction of Agamemnon to be men-
tioned in the play, 186 μάντιν οὔτινα ψέγων, is highly significant).
Yet there is a more profound motive underneath. The king’s unwilling-
ness to persevere in what he considers a right and proper refusal appears to
be grounded in a feeling which in plain language might be expressed by the
phrase ‘what is the use?’ It is not worth while to resist any more. For ten
years he has been struggling hard. At the very beginning of the war he was
forced to commit a most horrible sin, knowingly and without any illusion.
Then he had to contend not only with the enemy but also with disloyalty in
his own camp (838-42). At last Troy fell, but the homeward voyage was full
of disaster. Now, on the threshold of his house, he finds himself faced with
ill-concealed hostility and treacherous lies. The warnings of his faithful
subjects are not lost on him: he fully grasps their implications. Then
Clytemnestra speaks. Almost every word she says jars upon his sense of
decency, and, worse, lays bare her insincerity. To crown all, she bids him do
For the sort of biased view which I regard as unjustifiable cf. the notes on 811 and 948.
441
 lines 975-1034 COMMENTARY
a thing that he feels to be utterly humiliating and incompatible with his
. dignity and his piety. And yet he yields. He is tired to the utmost, worn out
by the unceasing struggle, overpowered by the slings and arrows of out-
rageous fortune. In a beautiful passage of Aeschylus (fr. 362 N.) we read this:
ἀλλ᾽ οὔτε πολλὰ τραύματ᾽ ἐν στέρνοις λαβὼν θνήισκει τις, εἰ μὴ τέρμα συντρέχοι
βίου κτλ. For Agamemnon now the moment seems to have come when τὸ
τοῦ βίου τέρμα συντρέχει.
ΑἹ this point we must be careful not to go too far. It would be wrong to
assume that Agamemnon suspects that Clytemnestra is plotting against his
life. If he knew or suspected that he was to be lured into a death-trap and
still walked on, he would be a fool, not a tragic hero. What he does perceive
is the atmosphere of hostility and danger. He knows that what his wife so
insistently asks him to do will not be for his own good. Without seeing through
Clytemnestra’s real scheme, he is not wholly deceived by her. He gives up
struggling and is thus defeated. It would be an understatement to call this
a symbolic defeat. It is a real one ; hence the strong expression κατέστραμμαι
(956). The scene of this defeat occupies the centre of the play. The actual
murder afterwards seems to be but the final confirmation of the king’s
decisive downfall.’

THE THIRD STASIMON (975-1034)

A choric ode consisting of only two pairs of stanzas is the commonest form
in the plays of Sophocles and Euripides. Aeschylean instances are found in
the works of his latest period: Prom. 526,* Ag. 975, Cho. 935; cf. W. Kranz,
Stasimon, 175.
Metre
I. 975-87 = 988-1000. Three lecythia (cf. 160 ff., 681 ff), molossus (see
note on 978), 5 dactyls (catalectic), 2 lecythia, 2 cretics+lecythion, acatalectic
iambic trimeter, 2 cretics+lecythion, lecythion.
The dactylic pentameter in this context may be compared with Cho. 591,
where it occurs between lecythia. .
Wilamowitz regards the whole stanza, apart of course from the dactylic
line, as trochaic. This uniformity may look attractive, but it would have to
be bought at a high price. To obtain a series of seven trochaic metra (from
982 θάρσος to 985 ἐμβολαῖς), we should lose the pleasant symmetry by which
the iambic trimeter, χρόνος... ἐμβολαῖς, is placed in a frame of 2 cretics+
lecythion on either side. We should furthermore be compelled to regard as a
mere accident the fact that the beginning of the iambic trimeter in both
1 A supplement to the last remark will be found in the note on 1129. In general cf.
Beazley (in Beazley and Ashmole, Greek Sculpture and Painting, 38), who observes that the
Polygnotan Troy Taken gave the time after the action and the east pediment at Olympia
gives the time before the action, and adds the following remark: “Two things are to be
said: first, that what we here observe in narrative art, the recession of interest from the
action itself to the mental condition previous or subsequent to action, is parallel to some-
thing already noticed in the substantive statue; and secondly, that long and conflicting
suspense, a knife-flash of action and that unseen, and then strange revulsion and unease, is
the very formula, the very soul, of Aeschylean tragedy.’
2 For the reason why Kranz does not regard this stasimon as an Aeschylean instance see
his bold hypothesis, Stasimon, 226 ff. (cf. also Th. Bergk, Kleine philol. Schriften, ii. 743,
no. 38).

442
 COMMENTARY line 978
stanzas coincides with the beginning of a word (in the strophe, 984, also with
the beginning of a fresh sentence). |
If we here assume an iambic trimeter within a trochaic (or trochaic-cretic)
stanza, we need not enter into a general discussion of the complex problem of
possible transitions from trochees to iambics and vice versa. For the decision
of the issue in hand it is sufficient to compare the trochaic stanza Eum.
916 ff. = 938 ff., where, after several lecythia, at 924 there follows an acata-
lectic iambic trimeter, ἐπισσύτους βίου τύχας ὀνησίμους. Wilamowitz, in this
case too, advocates the continuation of trochees, but he is again forced to
disregard in both stanzas the coincidence of the beginning of the trimeter
with the beginning of a word.
II. 1001-17 — 1018-34. ‘Prima incerta [because of the destruction of the
text]. antistrophus speciem obfert: 2 cret.+pherecr., cf. Eum. 329 f. [τόδε
μέλος, παρακοπά, παραφορὰ $pevobBaXjs] (Wilamowitz). Then (to take the
better preserved antistrophe, 1020 f. μέλαν alua . . . ἐπαείδων) apparently a
catalectic anapaestic trimeter, twice a hemiepes (μῆνιν ἄειδε θεά), decasyllabus
alcaicus, 6 lecythia, 8 dactyls, lecythion.
975. The only other example of the Homeric rérre in Aeschylus is Pers.
554, in the same metre. The preceding chorus (681) also began with a question
in the same form, τίς ποτε.
976. δεῖγμα in F, as e.g. Aratus 629 in C instead of the δεῖμα in the other
MSS. There is no need to refute those who defend δεῖγμα in spite of what the
context obviously requires.
For προστατήριον here L-S offer a choice between the local meaning,
‘hovering before’, and the other, ‘lording it over’. On the other hand, Wila-
mowitz, following Blomfield and others, explains: ' kapô. poor. nihil est nisi
πρὸ καρδίας." This local sense is undoubtedly in keeping with the ideas of
Aeschylus, cf. on 179. But if the poet did not wish his hearers to think at the
same time of the function of a προστάτης, it is unlikely that he would have
chosen the uncommon προστατήριος to express the simple local meaning. In
Sept. 449 we perceive a local meaning but at the same time something more
(patrona) in the naming of the προστατηρία Ἄρτεμις (see Wilamowitz, Hermes,
xxvi, 1891, 21x n. 1 = KI. Schr. v. 1. 46); for the general nuance cf. e.g. S.
Oed. R. 882 θεὸν οὐ λήξω ποτὲ προστάταν ἴσχων. This is confirmed by Eum. 518,
where it is said of τὸ δεινόν that φρενῶν ἐπίσκοπον δεῖ μένειν καθήμενον ; the
associations (cf. below on 1270) brought into play by ἐπίσκοπον lead up to the
idea of divine sovereignty, as in the case of προστατήριον. The analogy is not
fortuitous ; there the theme is τὸ δεινόν, of which the Erinyes are the guardians,
here it is the δεῖμα which is incessantly chanted to the heart and imprinted
upon it by the θρῆνος ᾿Ερινύος.
978. The MS reading ποτᾶται is retained by most of the editors. Hermann
(cf. his Elem. doctr. metr. 46) dealt with the corresponding word of the anti-
strophe (991), ὑμνωιδεῖ, by taking its first syllable as short, relying on the
testimony of Hephaestion r. 8 (p. 6. 6 Consbr.) that the middle syllable of
εὔυμνος is short in Epicharmus fr. gr Kaibel! The possibility of such a

* The possibility that in the Laconian chorus Ar. Lys. 1305 the first syllable of duviwpes
is short is considered by Wilamowitz, ad loc. E. Bacch. 71 ὑμνήσω ought not to have been
quoted by Hermann as an example of a short duv-. However, he was right in not comparing
the examples in which pv at the beginning of a word does not make the previous syllable
443
line 978 COMMENTARY
quantity in Aeschylus cannot be ruled out altogether, but it is very unlikely
that he took a liberty only one example of which could be found by ancient
metricians, and that outside the poetry of Greece proper. On purely metrical
grounds there would be no objection to v — — in this position ; it would follow
on the preceding lecythion (καρδίας repaokómov) exactly as in 182 we have
δαιμόνων δέ που χάρις βιαίως. But in the last metron of such a trimeter (ποτᾶται
could not be connected with the next line, for there the dactyls begin) the
responsion of a bacchius with a molossus (to keep the normal scansion of
ὑμνωδεῖ) seems very doubtful; it is true that O. Schroeder, Aesch. cantica,
accepts it, but the three examples which, besides this one, he quotes on p. 102
under 'Licentiae antistrophicae' for the responsion —— — τευ -- — do not
come from the end of a period, and two of them are extremely doubtful.
Therefore I consider it much more probable that Meineke, Philol. xix, 1863,
203, and Wilamowitz were right in reading πωτᾶται. For this epic form (cf.
Schwyzer, Griech. Gramm. i. 719 n. 3) Wilamowitz compares Prom. 645 πωλεύ-
pevaı,! where it is significant that the ὦ is only preserved in the Medicean ; in
his note on Ar. Lys. 1013 Wilamowitz concludes that πωτᾶσθαι was still
current in Laconian ; but this is not certain. Eum. 250 πωτήμασιν should also
be borne in mind. A single molossus is found in Sept. 368 παγκλαύτων (the
corresponding passage in the strophe, 356, is uncertain) in front of a lecythion
and after two lecythia and a trochaic trimeter.
979. μαντιπολεῖ only here; μαντιπόλος (first found in E. Hec. 121, but it
may be older) was probably formed after the pattern of οἰωνοπόλος, ὀνειρο-
πόλος (both Homeric) and the like.
For the thought cf. the words of Orestes, just before the Erinyes gain power
over him, Cho. 1024 πρὸς δὲ καρδίαι φόβος ἄιδειν ἑτοῖμος, ἢ δ᾽ ὑπορχεῖσθαι κότωι.
ἀκέλευστος ἄμισθος : the spontaneity of this song of fear is emphasized in
this line with regard to any possible commission and payment and in ggı
(αὐτοδίδακτος) with regard to the origin of the song itself. There is here an
important association of ideas. Just as the song of horror of the Erinyes
(Eum. 333) is ἀφόρμικτος, so here the song which they inspire in the heart
(992) is not only ἄνευ λύρας (990), but is altogether different from the song
which is sung at meal-times or on some festal occasion. A professional singer
sings neither unbidden nor unrewarded. Demodocus, who is asked ἀλλ᾽ ἄγε
δὴ μετάβηθι καὶ ἵππου κόσμον ἄεισον δουρατέου, receives (among other things,
one may suppose) a choice piece οὗ roast meat (θ 474 ff.), and the normal
position of the commissioned singer of later times is implied in the words
Μοῖσα. . ei μισθοῖο συνέθευ παρέχειν φωνὰν ὑπάργυρον (Pind. P. τι. 41). As
the word ἀφόρμικτος in the δέσμιος ὕμνος of the Eumenides, so here the
phrases ἄνευ λύρας, ἀκέλευστος ἄμισθος recall the many festal occasions on
which ἀοιδή is the delight of gods and men. It is by the comparison thus
implied that Aeschylus makes the gruesomeness particularly effective (cf.
on 1186 ff). Wecklein’s note on ἄμισθος ‘not as the χρησμολόγοι' introduces an
idea which is foreign to the passage.
long (in Hephaestion, loc. cit. and elsewhere); in this he was more cautious than W. Christ,
Metrik, 2nd ed., 14 and Wecklein on Ag. 991 (980 Weckl.). Hermann (he was followed by
Wecklein in his commentary and by G. Thomson, ad loc.) scanned Ag. 1459 πολύμναστον
with the second syllable short, but this depends on the mistaken supposition that it corre-
sponds to 1547 and therefore on false textual criticism and misunderstanding of the metre.
1 Cf. F. Ritschl, Opusc. i. 383 (anticipating Hermann).
444
 COMMENTARY lines 983 ff.
980. οὐδ᾽ ἀποπτύσας κτλ. 'Scaligeri et Casauboni coniecturam, ἀποπτύσαν,
constructionis mutatione non perspecta, receperunt Porsonus et Blomfieldius’,
observes Hermann correctly; the participle is ‘construed according to the
sense [of what follows] = ἀναθαρρῶ ᾿ (Schneidewin). The similar example
which occurs shortly afterwards, 1008 ff., should have been a warning against
conjectures (the earliest of which is found in Tr). ‘Figura huic auctori
familiaris’ (Stanley). More of such Aeschylean anacolutha are to be found in
M. Berti (cf. on 12) ; this passage ibid., p. 242.
The object of ἀποπτύσας can be supplied without difficulty from the pre-
ceding words.
982. εὐπειθές : cf. on 274.
ἵζει (an excellent emendation): in the passage quoted on 976, viz. Eum.
518 f., it is said of τὸ δεινόν : φρενῶν ἐπίσκοπον δεῖ μένειν καθήμενον. Cf. (van
Heusde) E. Alc. 604 πρὸς δ᾽ ἐμᾶι ψυχᾶι θάρσος ἧσται.
983. ‘ φίλον sensu Homerico’ C. G. Haupt, cf. Conington on Cho. 110, Blass
on Cho. 410 f., Wilamowitz, Interpr. 64 n. 1: ' φίλον fjrop simply possessive’.
$pevós . . . θρόνον: Passow and Dindorf (Thesaurus) compare Pl. Rep.
8. 553 b δείσας οἶμαι εὐθὺς ἐπὶ κεφαλὴν ὠθεῖ ἐκ τοῦ θρόνου τοῦ ἐν τῆι ἑαυτοῦ
ψυχῆι φιλοτιμίαν κτλ.
983 ff. χρόνος . . . παρήβησεν. This passage has suffered a slight corruption
in ξυνεμβόλοις and two serious corruptions in the following words ψαμμίας
ἀκάτα. But the way in which many editors have laid hands upon the sound
words, while some have thrown the whole passage into confusion, is quite
irresponsible; not even so Aeschylean a phrase as χρόνος παρήβησεν has
escaped their experiments. Ahrens (590 ff.), while retaining παρήβησεν, has
upset the meaning by taking χρόνος δ᾽ ἐπεὶ as a formula with no verb expressed
= diu autem est ex quo (Wecklein followed him in this). He based this view
upon S. 4j. 600 ἐγὼ δ᾽ ὁ τλάμων παλαιὸς ἀφ᾽ οὗ χρόνος κτλ. and the parallels,
mainly from prose writers, quoted for it by Lobeck. The phrases of this type,
as Lobeck has noted, are used predominantly as parentheses, though this is
not so e.g. in Xen. Anab. 3. 2. 14 ἀλλ᾽ οὔπω πολλαὶ ἡμέραι ἀφ᾽ 00... ἐνικᾶτε, cf.
Wackernagel, Vermischte Beiträge, 27. But the real objection to Ahrens’s
view is that in these instances there is never a simple χρόνος émet . . ., but there
is added always a παλαιός, πολύς or similar qualification. To these phrases
χρόνος παρήβησεν, ἐπεὶ... is closely parallel in meaning, indeed it might
perhaps be assumed that it was coined as an equivalent to a χρόνος παλαιός
(or πολύς), ἐξ o5 .... That παρήβησεν performs the function of a ‘perfect of
result' is completely normal for this period and particularly in this style, cf.
Wackernagel, Stud. z. griech. Perfektum (Progr. Göttingen 1904), 6 ff., 12.
It has been conclusively demonstrated by Ahrens that the correct reading,
obtained by a slight emendation, is ξὺν ἐμβολαῖς (Casaubon) and that it means
‘the hauling-in of the mooring-ropes when a ship leaves its landing-place' ;
he refers to ἐμβολή and ἐμβάλλειν as the technical term for the loading of
cargo.
ψαμμίας ἀκάτα is certainly corrupt; the metre of the completely sound
antistrophe shows that — o — — is required here: furthermore ἀκάτη is not a
Greek word (this was insisted upon by Blomfield and especially by Ahrens).
ψάμμος was recovered by Wecklein, Studien zu Aesch. 123. For the next word
(ἀκάτα) there are extremely narrow limits to the possibilities of emendation.
445
lines 983 ff. COMMENTARY
A verb is needed to go with ἐπεί and it must be a spondee. This being so, it
is difficult to explain (unless one wishes to enter upon moral judgements) why
the emendatio palmaris ἄμπτα, after it was once published (1885), was treated
with contempt. The form ἀνέπταν, ἄμπταν is established with certainty for
the choruses of Tragedy: 5. Ant. 1307 ἀνέπταν, E. Med. 440 ἀνέπτα, Ion 796
ἀμπταίην (restored with certainty from ἂν πταίην), cf. A. Prom. 555 προσέπτα,
5. Ant. 113 ὑπερέπτα. The idea of sand or dust ‘flying up’ is quite suitable,
cf. A. Suppl. 781 f. τὸ πᾶν δ᾽ ἄφαντος ἀμπετὴς ἀϊδνὸς (very satisfactorily
emended)! ὡς κόνις ἄτερθεν πτερύγων ὀλοίμαν (S. El. 714 ἴ. κόνις δ᾽ ἄνω φορεῖτο).
This verb is extremely appropriate here. As soon as the wind had changed at
last after an excessively long period of waiting and the fleet could sail from
Aulis, the mooring-ropes, which had been fastened on stones or poles, were
drawn into the ships at top speed: not a moment to lose! The ropes run
through the sand on the shore and it whirls into the air. From the context
of the great epic event a minor detail? is detached and fondly elaborated by
the poet ; just as in a similar passage in the preceding chorus (690) it was said
of Helen’s departure : ἐκ τῶν ἁβροπήνων προκαλυμμάτων ἔπλευσεν. This method
of dwelling on minor details at the expense of the well-known features of the
story is characteristic of many treatments of epic themes in lyric poetry.
986. ὑπ᾽ Ἴλιον after the Homeric ὑπὸ ἤίλιον ἦλθεν (Blomfield).
989. αὐτόμαρτυς only here in pre-Christian literature.
990 ff. Here it becomes particularly clear (cf. on 976, 979) that a fundamental
theme of the whole trilogy is expressed in these two stanzas. It is carried
farther in 1191 ff. and especially in the choruses of the Eusnenides.
992. αὐτοδίδακτος clearly points to the origin and nature of the knowledge
of the moral law which occupies the central position in this chorus as in the
whole Oresteia. At the same time the word helps to produce the effect dis-
cussed on 979 ἀκέλευστος ἄμισθος. In the only passage in which αὐτοδίδακτος
occurs in Homer (x 347) Phemius, who significantly enough is called Τερπιάδης,
the singer ὅς re θεοῖσι καὶ ἀνθρώποισιν ἀείδει, Says: αὐτοδίδακτος δ᾽ εἰμί, θεὸς δέ μοι
ἐν φρεσὶν οἴμας παντοίας ἐνέφυσεν. Here again, then, the awful chant which
the heart sings as a θρῆνος of the Erinys is set against the background of
that festal song which is the delight of all.
992 f. ἔσωθεν θυμός : this idea is rooted deep in the spirit of the poet; else-
where it breaks out where the heart raises its voice, whether in grim fore-
boding (Pers. 10 f. κακόμαντις ἄγαν ὀρσολοπεῖται θυμὸς ἔσωθεν) or in present
trouble (Pers. 991 βοᾶι βοᾶι μοι μελέων ἔντοσθεν ἦτορ). In the latter passage
the influence of similar phrases in Homer such as ὃ 467 μινύθει δέ μοι ἔνδοθεν
ἦτορ (cf. e.g. also A 243 σὺ δ᾽ ἔνδοθι θυμὸν ἀμύξεις) is quite clear; and such an
influence may probably be assumed for the other passages. Cf. also Trag. fr.
adesp. 176 N. πηδῶν δ᾽ ὁ θυμὸς ἔνδοθεν μαντεύεται.
993. τὸ πᾶν: cf. on 175. Here τὸ πᾶν strengthens the negative, as Prom. 215;
so Paley is right: ‘not at all feeling the . . . confidence of hope’, Kennedy and
others are wrong: ‘having not hope’s . . . courage to the full’.

! It is true that Wilamowitz in his edition of Aeschylus says: ‘Phot. Berol. ἀΐδνόν ro
ἀφανιστικόν [the important οὕτως «Αἰσχύλος he omits here] huc non pertinet’, but earlier
(Berl. Sitzgsb. 1907, 4) as well as later (Verskunst, 609) he came to a different conclusion.
2 The detail is, however, significant enough to represent the whole of the departure, cf.
Propert. 1. 8. 11 nec tibi Tyrrhena solvatur funis harena.

446
 COMMENTARY line 997

995 ff. ' Σπλάγχνα explicat poeta per κέαρ, quod τελεσφόροις δίναις κυκλούμενον
πρὸς ἐνδίκοις φρεσίν dicit . . . non fallitur animus, cor eventum ferentibus
fluctibus ad veracia agitatum praecordia, i.e. cor motu suo, qui non vano ex
timore nascitur, recte praesaga pulsat praecordia’ (Hermann).
ματάιζει used absolutely, as S. Oed. R. 891.
997. κυκλούμενον: the sense seems to be clear: ‘circumactum’ (Stanley),
‘whirling round’ (Conington). However, Headlam (first in C.R. xvi, 1902,
437 n. 15) has pointed out that κυκλοῦσθαι never means 'eddying', but always
“eircling round [encircling]'. I have not found any example to the contrary.
Headlam, therefore, read κυκώμενον' and was followed by Wilamowitz. It is
surprising that Headlam limited his investigation to κυκλοῦν, excluding
κυκλεῖν, for the latter is as early as Homer and occurs several times in Tragedy :
though not elsewhere in Aeschylus. Most dictionaries, Wellauer (Lex. Aesch.),
Passow, Dindorf (Thes. and Lex. Aesch.) L-S, quote the passage under
κυκλόωξ, but it is quoted correctly under ‘ κυκλεῖν to whirl round’ by Linwood,
Lex. to Aesch. The question of how far the Athenians of the fifth century felt
any phonetic difference between the κυκλούμενον (KYKAOMENON) from
κυκλοῦν and that from κυκλεῖν is one with which we do not need to bother
any more than with the origin of the differentiation between κυκλοῦν and
κυκλεῖν or the extent to which it was observed in detail.
The image in 996 f. (πρὸς ἐνδίκοις. . . κέαρ) is clear and consistent. It
illustrates effectively certain ideas fundamental not to this chorus alone but
to the play as a whole (see p. 451 f. below, on the first two stanzas). Anato-
mical details, however, should not be dragged in; they would obscure the
meaning. The heart, with its feelings and in particular its fears, is driven
round in a circle by the eddies which are called τελεσφόροι (discussed below).
The eddying movement takes place, as πρός shows, 'at' or 'upon' the φρένες,
i.e. the φρένες correspond to the solid reef or cliff, upon which the eddying
flood ever breaks anew, as in ε 401 δοῦπον ἄκουσε ποτὶ σπιλάδεσσι θαλάσσης.
The passage is customarily translated ‘the heart beats against the breast’ or
the like. Headlam places even stronger emphasis upon φρένες as an organ of
the body when he comments ‘the heart dashing against the midriff’ and
compares Prom. 881 xpadia δὲ φόβωι φρένα λακτίζει. In that passage Io is
certainly describing the physical phenomena of her suffering, but it is equally
certain that by φρένες in Ag. 996 we are not to understand the midriff. It is
true that Aeschylus, like Homer and others, sometimes uses the plural
φρένες in this way (Prom. 361, Ewm. 159), but far more frequently φρένες, in
the many passages in which it occurs in Aeschylus, means ‘mind, cogitation,
reflection, thought’ etc., cf. on ı75. This meaning is particularly clearly
marked in Pers. 767 φρένες γὰρ αὐτοῦ θυμὸν duakoorpépour. In that passage

t His reference to "Trag. fr. ap. Clem. Al. p. 486 [Potter] = Strom. 2. 20. 108. 1 (vol. ii,
p. 172 Stählin) is, apart from ‘Trag.’, perfectly correct. G. Thomson (ii. 376) speaks of it as
‘a false reference, which I have been unable to correct’. I wonder why. The lines, it is true,
are not to be found in Nauck’s collection, but for that there is a very good reason, see
Wilamowitz, Kl. Schr. i. 196.
2 But in L-S the reader is justly warned that ‘such forms as κυκλοῦνται, ἐκυκλοῦντο,
etc. may belong to κυκλόω or to κυκλέω '.
3 For the Homeric usage cf. the helpful remarks of Snell, Gnomon, i, 1931, 76 f. In the
remains of the Lesbian poets the meanings of φρήν and φρένες seem to be clearly differ-
entiated, cf. Lobel, Ἀλκαίου μέλη xxxvi.
447
line 997 COMMENTARY
the θυμός is the instrument of the passions and particularly of the desires
(ambition, etc.) ; in the present ode (993) it is the instrument of fear and sings
the dirge of the Erinys; its place is taken in the next sentence by σπλάγχνα,
which is again picked up by κέαρ. Here, as in the Persae, θυμός and κέαρ are
the tumultuous and unstable elements, φρένες the solid and fixed. It is clear
from ματάιζει that in 995 ff. σπλάγχνα and κέαρ are much more the seat of
spiritual and emotional functions than the bodily organs themselves. The
addition of ἐνδέκοις is in itself enough to prove that φρεσίν here does not refer
to the midriff or any other physical feature. No one doubts that this is the
case in examples such as Suppl. 106 δυσπαραβούλοισι φρεσίν, 750 Í. οὐλόφρονες
δ᾽ ἐκεῖνοι, δολομήτιδες δυσάγνοις φρεσίν (parallel to the οὐλόφρονες), Ag. 1064
κακῶν κλύει φρενῶν, or, in the singular, Ag. 275, 805, 895 ἀπενθήτωι φρενί, 1302,
and in many other instances. ἔνδικον means that in which there is δίκη, in
conformity with the normal type of these compounds, e.g. ἔναιμος, ἔνθεος, ‘in
which there is blood, a god’ (see Schwyzer, Griech. Gramm. i. 435, and cf. on
562 ἔνθηρον). The word is found in Aeschylus frequently ; elsewhere in Greek
literature it occurs first in Pindar, P. 5. 103 (composed three or four years
before the performance of the Oresieia), where he describes the celebration
and the song of victory as υἱῶι re κοινὰν χάριν ἔνδικόν 7° Aprceaikaı, because they
have provided satisfaction for the king’s legal claim (cf. on 813), in accor-
dance with the well-known idea of the obligation and the request for payment
which arises out of the victory; immediately afterwards (106) we read τὸ
καλλίνικον λυτήριον δαπανᾶν μέλος χαρίεν (with regard to this whole series of
ideas cf. Schadewaldt, ‘Aufbau des Pindarischen Epinikion’, Schriften der
Königsberger Gel. Ges., 5. Jahr, Heft iii, 1928, 278). In Aeschylus also, ἔνδικος
several times' means ‘in whom or which a legal claim stands’, ‘who or which
possesses a legal claim’. In the case of a person we find it e.g. Sept. 673 τίς
ἄλλος μᾶλλον ἐνδικώτερος ; ‘who else has a greater claim upon it?’ Of a thing:
Suppl. 590 f. riv’ ἂν θεῶν ἐνδικωτέροισιν κεκλοίμαν εὐλόγως ἐπ᾽ ἔργοις ; correctly
explained by Wecklein: ‘which god through his actions gives us a greater
claim duly to invoke him?’ Similarly Headlam : “What act on the part of any
other god affords me greater justification for appealing to him?’ Strictly
defined, ἔργα ἐνδικώτερα are deeds wlıich contain within themselves a greater
claim. We now turn to the phrase γόος Evöıxos? in Cho. 330, a passage which
* Of course, ‘in whom (or which) there is δίκη ' can always mean simply ‘just’, as in
Eum. 805, 966, and elsewhere.
2 Murray is not justified in turning to the scholion ζητεῖ... τὴν ἐκδίκησιν to support his
reading ἐκ δίκαν; for that phrase, just as ζητεῖ τὸ ἀντιτιμωρεῖσθαι in another version of the
same scholion, is simply a paraphrase of ματεύει. The words ἐκδίκησις, ἐκδικεῖν etc. are
favourites with the writers of these (and other) scholia, cf. schol. Cho. 287 (286 Weckl.),
where ἐκ προστροπαίων is explained: ἐκ τοῦ ‘Ayapéuvovos ἱκετεύοντος τοὺς θεοὺς ἐκδικήσεως
τυχεῖν, schol. Cho. 278 ὡς μὴ ἐκδικήσαντας, schol, Sept. 684 (671 Weckl.) ἕν κέρδος τὸ αὑτὸν
ἐκδικῆσαι ἀδικούμενον" ἐὰν δὲ ἀνεκδίκητος ἀποθάνηι krÀ., on Eum. 624 (627 Weckl.) ἐκδικήσαντα,
on Eum. 840 (842 Weckl.) where ἀνεκδέκητον is the gloss on arierov. Cf. also schol. S. Oed. R.
377 τάδ᾽ ἐκπρᾶξαι: ἀντὶ τοῦ τὰ κατ᾽ ἐμὲ ἐκδικῆσαι, schol. E. Phoen. 260 (on ds μετέρχεται
δόμους) ὃς ἀπαιτεῖ τὴν βασιλείαν καὶ ἐκδικεῖ, 935 (on δράκοντι τιμωρεῖ φόνον) ἐκδικεῖ τῶι δράκοντι
τὸν φόνον, schol. Apoll. Rhod. 3. 338 ποινάς : ἐκδικίας. In the first hypothesis to E. Or. 1
in one or two MSS " ἐκδικῶν interpretationis causa ad proximum verbum peramopeudpevos
adscriptum’ (Wecklein).
Altogether a great deal of harm is caused by scholars who consider themselves justified
in dragging out a word from the scholia, to support conjectures or interpretations, without
attending to the context or the linguistic practice of the scholia. (‘Scholiis ad artem criticam
448
 COMMENTARY line 997
has been ill treated by many editors, but on which much light has been
thrown by Schadewaldt's intensive interpretation of the whole xoppdst
(Hermes, \xvii, 1932, 319). The sphere of this κομμός is one of magic rituals:
the γόος is here represented as something (perhaps ‘someone’ would be more
adequate) that picks up the trail of the murderer, finally to track him down.
It is the right of the γόος to bring this about. δίκη (‘right’, ‘claim’) is operative
in the γόος, therefore it is called γόος ἔνδικος. In Eum. 135 dAynoov ἧπαρ
ἐνδίκοις ὀνείδεσιν, the ὀνείδη appear as ‘justified’, ‘entitled’: through the
sacrifices which she has made to the Erinyes (106 ff.) Clytemnestra has gained
the right to claim their help, so the reproaches with which she urges on the
slothful Erinyes to their duty are ἔνδικα. In Ag. 996 the phrase ἔνδικοι φρένες,
then, means the thoughts in which there is δίκη, in this case no special legal
claim, but simply τοὐφειλόμενον πράσσουσα Δίκη (Cho. 310). It is with this
δίκη that the hearts and minds of the old men are filled. In their φρονεῖν the
conviction of δράσαντι παθεῖν is contained. Upon and around the conscious-
ness that every wrong demands its expiation, there circle (as upon a firm
rock, if we wish to take up the image again) the trepidations of the heart.
The eddies which drive these emotions of the heart are called τελεσφόροι.
The meaning of this word presents no difficulty ; there are good references in
R. Hirzel, Themis, 136, who quotes, as the oldest example of the underlying
idea, Hesiod, Erga 217 f.: δίκη δ᾽ ὑπὲρ ὕβριος ἴσχει ἐς τέλος ἐξελθοῦσα. He also
points out that when in 5. 47. 1390 the τελεσφόρος Δίκη is invoked (together
with the μνήμων ’Epwvs, cf. above on 155), this corresponds to the idea in
Ag. 781 (Δίκα) πᾶν ἐπὶ τέρμα νωμᾶι. Confirmation is provided by the fact that
in Eum. 543 κύριον τέλος picks up, with almost the same meaning, the im-
mediately preceding word ποινά. Dike herself, who is called τέλειος in Ag.
1432, is not described as τελεσφόρος in the extant remains of Aeschylus, but
this may be mere chance. The apai in Sept. 655 are τελεσφόροι, just as in
Ag. 700 the μῆνις is described as τελεσσίφρων, and μοῖρα, too (Prom. 511), is
τελεσφόρος. There is no need to go into further details here (for the whole
group of ideas cf. Daube, 117 f). In the present passage the use of the
adjective with δίναις presents the idea that the eddies will not cease until the
goal of the movement, 1.6. the conclusion of the divinely willed trial, has been
exercendam nemini uti licet nisi qui totum scholiorum corpus non semel vel bis, sed iterum
iterumque diligenter pertractaverit’, Ed. Schwartz, Schol. in Eurip. ii, p. v). So Schade-
waldt, Hermes, Ixvii, 1932, 317 n. 2, seeks to show that ὁμοέως, which in fact never means
anything but ‘in the same way’, can mean ‘however’, ‘nevertheless’ in Cho. 320 and ‘comes
near to ὅμως '. For this purpose he first misinterprets (as do e.g. Conington and Sidgwick
on Cho. 320) the ὁμοίως in Pers. 214 (correctly understood by Teuffel-Wecklein : ὥσπερ καὶ
πρόσθεν, and Wilamowitz, Interpr. 52: ‘just as always’), then he asserts: ‘ ὅμως is also used
in the scholiast’s paraphrase of our passage’. This is refuted by the very words of the
scholion : ὅμως δὲ... ὁ γόος ὁ εὐκλεὴς ὁμοίως χάριτες κέκληνται: i.e. ὅμως is not a paraphrase
of ὁμοίως, but serves to strengthen the adversative δέ. This use (or the use of ὅμως alone to
mark the adversative connexion) is characteristic of these scholia, as may be seen from a
passage which follows immediately afterwards, schol. Cho. 327 (ὀτοτύζεται δ᾽ ὁ θνήισκων) :
δεῖ δὲ ὅμως τὸν ἀποθανόντα θρηνῆσαι, similarly e.g. on Sept. 447 ff. (434 ff. Wecki.), 1045 (1036),
Ag. 107 (109), 1130 (1122), Cho.61 ff. (59 ff.) ὅμως à φόνος πέπηγεν κτλ. (paraphrase of 66 ff. =
64 ff. Weckl.), 389.
! The sceptical note of Morel, Bursians Jahresbericht cclix (1938, i), 21, about the passive
ὀτοτύζεται 327, upon which the understanding of the whole passage depends, was invalidated
in advance, in Hermes, lxviii, 1933, 242 n. 9, by reference to the Labyadae-inscription ; cf.,
too, Latte, RE xvi. 283. 57.
4872.2 Gg 449
line 997 COMMENTARY
reached and the wrong expiated. The epithet ἐνδίκοις makes the meaning of
τελεσφόροις more easily intelligible.
The sentence 995-7 (σπλάγχνα . . . κέαρ) provides in the form of a general
statement the reason (Stanley: 'quippe praecordia haud fallunt’ eqs.) for the
preceding assertion. The universal character of the sentence was correctly
understood by e.g. Wilamowitz: ‘Nicht trügt das Zeichen, wenn das treue
Herz... warnend... pocht’ ; ‘my’ is wrongly inserted by Conington (‘my
bowels’), Schneidewin, Paley, and other translators. But, of course, the
general statement applies to the speaker among the rest.
998. Only the reading in F εὔχομαι δ᾽ ἐξ ἐμᾶς ἐλπίδος can be regarded as
παράδοσις ; the εὔχομαι δ᾽ ἀπ᾽ ἐμᾶς τοι ἐλπ. of Triclinius is clearly an attempt
to make it correspond to 985 ψαμμίας axdra(s) παρήβ. (cf. on 379) ; τοι is just as
bad a stopgap as ye frequently is.
998 f. ἐξ ἐμᾶς ἐλπίδος cannot mean: ‘contrary to my expectations’ (Enger) ;
that would be ἐκτὸς ἐλπίδος (S. Ant. 330, cf. ibid. 392) or ἔξω as E. Tro. 345
ἔξω TE μεγάλων ἐλπίδων. Neither is the purely local function of the preposition
admissible here, although it may seem particularly apposite in conjunction
with πίπτειν (cf. e.g. Pers. 313, Ag. 1245, Eum. 147). There is one ‘parallel’
against which I should like to give warning expressly: Eur. fr. 420 (Ino),
41. ὑπόπτερος δ᾽ ὁ πλοῦτος" οἷς “γὰρ ἦν ποτε, ἐξ ἐλπίδων πίπτοντας ὑπτίους ὁρῶ.
The image there (whether it is of a chariot or of something else) is just as
striking as A. Suppl. 96 f. ἐάπτει δ᾽ ἐλπίδων ἀφ᾽ ὑψιπύργων πανώλεις βροτούς. In
both cases the ἐλπίδες are to be taken in the common positive sense of good
hopes. The ἐλπίς here is quite different. It is on this account and also because
of the interpretation of ἐς τὸ μὴ τελεσφόρον which I regard as necessary (see
below) that I reject the explanation of Weil (' πίπτειν ἔκ τινος ets τι dictum est
per translationem perspicuam', similarly Plüss). So the only possibility
which remains is to understand ἐξ ἐμᾶς ἐλπίδος as ‘from my expectation,
starting from my expectation', i.e. instead of assuming a full local reference,
to take it in a weaker sense, which is not a clearly drawn 'out of . . .' in space,
but indicates the point of departure in a more metaphorical or abstract
sense. For this I can at least quote one contemporary parallel: Pind. P. 8.
88 ff. 6 δὲ καλόν τι νέον λαχὼν ἀβρότατος ἐπὶ μεγάλας ἐξ ἐλπίδος πέτεται ὑποπτέ-
ροις ἀνορέαις (for the necessity of taking μεγάλας with ἁβρότατος, as the
scholia paraphrase it, cf. Gildersleeve’s and Schroeder’s commentaries, and
also Wilamowitz, Bakchylides, 1898, 20). Cf. further Archilochus fr. 56. 3 D.
κιχάνει δ᾽ ἐξ ἀελπτίης φόβος, Xen. Anab. 2. 5. 12 ἐρῶ γὰρ καὶ ταῦτα ἐξ ὧν ἔχω
ἐλπίδας καὶ σὲ βουλήσεσθαι φίλον ἡμῖν εἶναι. Instead of merely saying ‘that
from my expectation nothing may come’, the Chorus says: ‘that from my
expectation it may fall as a falsehood to the ground’. ‘It’ is that which I
expect and fear, as everyone in the audience will understand at once. Any
attempt to make ψύθη the subject of πεσεῖν results in absurdity ; it is predica-
tive, as Schneidewin and Verrall rightly noted. For such a predicative noun
or adjective with recetv (and in exactly the same sense as here ‘to fall to the
ground, to be given over to destruction’) cf. 5. Aj. 616 ff. ra πρὶν δ᾽ ἔργα
χεροῖν (of Ajax) μεγίστας ἀρετᾶς ἄφιλα παρ᾽ ἀφίλοις ἔπεσ᾽ ἔπεσε μελέοις Arpeidaus.
! The ἐξ ἐλπίδων, which had sometimes been doubted, has now been found in the quota-
tion (taken from an early anthology) of the same five lines in a school copy-book of the
third century B.C., cf. O. Guéraud et P. Jouguet, Un livre d'écolier (Cairo 1938), 16 f.

450
 THE THIRD STASIMON AS A WHOLE
To the phrase ψύθη πεσεῖν, although sufficient by itself, there is added, to
clarify and complete the picture, ἐς τὸ μὴ τελεσφόρον, just as in 383 f. Aaxri-
cavrı μέγαν Δίκας βωμὸν eis ἀφάνειαν. The word τελεσφόρον, in the context in
which it was used just before (997), was unpropitious. To avoid the conse-
quences of such a malum omen the Chorus revokes the word at once (cf.
Neustadt, Hermes, lxiv, 1929, 252). The contradiction between the words
εὔχομαι κτλ. and the contents of the preceding stanzas as a whole is inevitable,
cf. on 255 ff. (p. 145 f.).
If we want to understand the first two stanzas of this ode and thereby one
of the basic conceptions of the whole play, we should be careful not to read
into the words of the poet any romantic tones or ideas. W. v. Humboldt
says in general (Introduction to his translation, p. xiv — Gesammelte Schriften
viii. 128) : 'In the Agamemnon the anxieties of the Chorus, the obscure, but
always horrifying, allusions of Clytemnestra, the lamentations and prophecies
of Cassandra fill the mind of the reader with sombre, though vague, fore-
bodings (“unbestimmte Ahndungen”), as with a melancholy music (“‘schwer-
müthige Melodien")'. He obviously has in mind this chorus in particular, as
is clear from his translation: “Wie doch schwebt mir immer vor unverrücket
jene Furcht, meinen ahndungsschwangern Sinn umflatternd?' and (998 ff.) :
* Möge, fleh' ich, entgegen nur meines Ahndens Bangigkeit hin es sinken ganz
in Nichts!’ It is just this word 'Ahnden', although it is quite foreign to
Aeschylus, that Wilamowitz! has introduced into both passages : "Was wollen
diese Grauenbilder, die von der ahnungsvollen Seele nicht weichen wollen'
and 'ich kann nur beten, dass der Ausgang mir die Ahnung Lügen strafe'.
In the words of this chorus everything is sharp and distinct : Aeschylus keeps
the δεῖμα clear of vague forebodings (‘Ahnungen’) and dim visions of horror.
How gladly would the old men put off their fear δίκαν δυσκρίτων ὀνειράτων,
but that is just what they cannot do. A firm line is drawn between dreams,
which admit of this or that interpretation, and the δεῖμα, which is so frightful
because it is so certain. The words in 994 σπλάγχνα δ᾽ οὔτοι ματάιξει correspond
exactly in thought to the où . . . δίκαν δυσκρίτων ὀνειράτων. The basis upon
which the certainty of the inner voice rests is stated in no less unmistakable
terms. This will be seen by following the clearly marked trend of thought in
the strophe and antistrophe.
The lines in the strophe 979 ff. and those in the antistrophe 990 ff. about the
song which comes from the heart are clearly meant to supplement and con-
firm each other. By its contents, as by the occasion of its performance and
the method of its presentation, the song of fear is the opposite of everything
customary in human society (cf. on 979 and 992). It begins of itself, un-
requested and unrequited ; the θυμός, which learned it from no one, sings it
from within. It is infallible (980 f. οὐδ᾽ ἀποπτύσας δίκαν δυσκρίτων ὀνειράτων,
995 σπλάγχνα δ᾽ οὔτοι ματάιζει) and cannot be turned aside, nor can it be
diverted by even the most glittering external success (990 ὅμως) ; it is the
θρῆνος ’Epwvos. Such are the Greek conceptions out of which translators in
thrall to romanticism make 'Hóllenlieder' (Wilamowitz) or evolve images
such as this: "like some Fury sorrow-ridden, weeping over things that die'
1 In several passages he owes more to Humboldt's translation than would be expected
in view of his severe criticism of it (Griech. Tragódien, ii. 4).
451
lines 1001 ff. COMMENTARY
(Murray), with the effect that the outlines of the clear-cut thought are dimmed
by an emotional haze. The θρῆνος "Epivvos does not deal with a vague horror,
a mere foreboding, but with the certainty of a universal law. This is once
more made clear in the penultimate sentence of the antistrophe, cf. on 996 f.
(pp. 447 ff.) and in particular the explanation of ἔνδικοι φρένες. Whatever
happens, the claim of δέκη must and will be met. This certainty can never
fade from the inmost heart of the Elders who earlier in the play professed
their creed by praising Zeus, the guardian of justice. The heart is called (997)
τερασκόπος ; ἃ τερασκόπος does not have forebodings of the τέρατα, he 5665
and recognizes them clearly; otherwise how could Apollo (Eum. 62) be so
called? No one is needed to teach the old men of Argos that which cannot be
averted (992), and no stroke of good fortune, not even the destruction of
Troy, for which they had almost ceased to hope, and the return of the vic-
torious king can mislead them in regard to those forces which they have
learnt to respect as being valid for ever. The fear which is based on the firm
ground of such a certainty is more awful than any foreboding.
1001 ff. As far as can be judged with any confidence in the case of an in-
sufficient supply of MSS and a combination of metrical forms which is, at
least, unusual, the text of the antistrophe (1019 f.), except for the slight cor-
ruption of πεσόν to πεσόνθ᾽, seems sound and its metre clear.! From this it
follows that the beginning of the strophe is very corrupt. ‘Emendation . . . is
almost hopeless’, says Hense, and Wilamowitz passes similar judgement.”
The extent of the corruption can be defined with a certain degree of prob-
ability. 1004 γείτων ὁμότοιχος ἐρείδει should not be touched (Wecklein 15
wrong: ‘ γείτων ex interpretatione ortum videtur’) ; its defence rests not only
upon the metre of the antistrophe (1021) but also upon the sense: with
ὁμότοιχος Paley compared Antiphanes fr. 295 K. λύπη μανίας ὁμότοιχος εἶναι
μοι δοκεῖ (cf. Plutarch, De garrul. 4, p. 503 4), Blomfield quoted Callimachus,
Cer. 116 f. μὴ τῆνος ἐμὶν φίλος... εἴη μηδ᾽ ὁμότοιχος " ἐμοὶ Kaxoyetroves ἐχθροί,
and for the personification οὗ νόσος conveyed by the word γείτων Williger,
Sprachl. Uniers. 24, quotes Suppl. 587 νόσους ἐπιβούλους, 684 f. νούσων δ᾽
ἑσμὸς ... Lor, Ag. 1017 νῆστιν... νόσον, and draws attention to the fact that
‘personification of the νόσοι is a part of the daemonology’ of the poet. It is
hardly likely that anyone will wish to touch νόσος (1003). The words τέρμα.
νόσος γὰρ are metrically unobjectionable as they stand, provided that μέλαν
... ἐπαείδων in the antistrophe 1020 f. can be taken as three anapaestic metra.
If a different metrical analysis be adopted, it would be necessary to assume
a corruption in the words τέρμα. νόσος γὰρ and to consider whether Porson
and others are right in putting a lacuna after νόσος γὰρ. There is no internal
evidence to settle the issue (apart from the established fact that νόσος must
have stood somewhere in the sentence) ; as regards τέρμα no certain judgement
is possible, on account of the serious corruption of the preceding words.
τ That Wilamowitz felt no unqualified confidence even in the beginning of the antistrophe
is clear from his discussion of it under Numer: (p. 218 of his edition) and from Verskunst,
332 n. 2. I myself cannot see anything to which objection can really be taken in 1018 ff.
2 The corruption in this passage, the equally serious one in 714 f,, and the mistaken
addition of a suffixed ν in 707 are the three pillars on which G. Thomson, The Oresteia, ii.
331 f., has built a new metrical theory. According to him an occasional break of the anti-
strophic correspondence is legitimate in the choruses of this play.

452
 COMMENTARY line 1005

1004. ἐρείδει. L-S s.v. II. ı (‘of an illness or pain, settle upon a particular
part’) compare two instances from physicians of the period of the Empire.
We should not, however, conclude that this is a particular technical expres-
sion or characteristic of medical language. The passages in question are
Rufus in Oreibasios 45. 30. 27 (vol. iii, p. 193 Raeder), where the author is
speaking of swellings in the neighbourhood of the ears, πολλοὶ γὰρ δὴ σφόδρα
ἐρεισάντων (‘if they, the swellings, have exerted strong pressure”) διεφθάρησαν,
and Galen, vol. xi, p. 61 K. εὑρήσεις . . . ei δι᾿ ὅλης αὐτῆς (scil. τῆς κεφαλῆς)
ἐκτέταται τὸ ἄλγημα πυνθανόμενος, ei κατά τινα τῶν μορίων ἐρείδει σφοδρότερον,
with an exactly similar use of ἐρείδειν. L-S’s rendering, therefore, does not
fit these passages ; and it is not acceptable for Ag. 1004 either. In all three
passages the normal meaning ‘prop, press, thrust’, frequent from Homer
onwards, is satisfactory. Here in Ag. 1004 the disease, the bad neighbour,
afflicts its victim at close quarters. ὁμότοιχος (recalling the houses of the
Plataeans, who ξυνελέγοντο διορύσσοντες τοὺς κοινοὺς τοίχους παρ᾽ ἀλλήλους,
Thuc. 2. 3. 3) not only strengthens the impression of immediately threatening
proximity but also presents the idea that the obnoxious fellow near by will
break down the flimsy wall of the house by pressing and thrusting and then
fall upon his victim.
In spite of the mutilation of the text at the beginning, the outline of the
thought in the first sentences of the strophe (down to 1004 ἐρείδει) is dis-
tinguishable. An excess of good health and vigour contains within itself the
danger of immediately imminent disease. Casaubon compared Hippocrates,
Aphorism. 1. 3 ai en’ ἄκρον εὐεξίαι odadepai; for further examples of this
‘widespread belief’ (Wilamowitz, Glaube d. Hell. ii. 137) see van Heusde and
G. Thomson, ad loc. ‘In exactly the same way’, the Chorus goes on (1005 ff.),
‘does a course which is too direct lead to shipwreck upon a hidden rock’.
The next idea (1008 ff. ‘too large a cargo’) is still intrinsically parallel to those
which preceded it and remains within the compass of 1005 ff., but it is given
a turn (‘safety is still possible by sacrificing a part’) which makes it serve as a
foil to the thought of the antistrophe (1018 ff.), in which the whole stasimon
culminates.
1005. With εὐθυπορῶν we may now compare the εὐθύπορον in a new fragment
of Aeschylus, Pap. Oxy. 2164, fr. ı. 11. A form of the verb εὐθνπορεῖν has
been restored with some probability in Pind. fr. 343 c ro Bowra (= Paean
13 Ὁ. 10 Snell).
The seafaring metaphor in εὐθυπορῶν is continued in ἔπαισεν... ἕρμα.
Blomfield quotes the chief passages for ἕρμα; to these there can now be
added Alcaeus fr. 46 b. 6 D. &pparı τυπτομέναν. Cf. also Pl. Rep. 8. 553 Ὁ
(Schneidewin).
With regard to 1005 f. καὶ πότμος . . . ἕρμα Hermann says correctly that, as
far as these words alone are concerned, no change is needed ; for παίειν with
the plain accusative (without eis or πρός) he compares 5. El. 744 f. Schoemann
(Opusc. iii. 182) indeed maintained that the thought was incomplete : ‘Quippe
non quaevis prospera fortuna ad scopulum impingit, sed ea tantum, quae
immodica est aut qua quis non recte utitur.’ So, too, Wilamowitz : ‘deside-
ramus nimii notionem in cursu directo.’ This view, however, was rightly
opposed by Ahrens (p. 601): ‘There is really no objection to the idea that
good fortune, while travelling under full sail, is often wrecked.’ But in spite
453
line 1005 COMMENTARY
of all this, Hermann’s seemingly facile reconstruction of the text was only
made possible by the most violent cutting down of the antistrophe; the text
there shows that a dactylic hemiepes has fallen out in the strophe. It is
difficult to decide whether the lacuna is to be assumed before 1005 (Klausen)
or after (Heath and others), (‘incertum ubi versus exciderit’ Wilamowitz ;
Sidgwick had already made a somewhat similar observation). Nothing can
be inferred from the fact that ἀνδρός seems to go with πότμος, for it is equally
possible that in front of ἀνδρός something on which the genitive depended
is lost.
1008 ff. Here, as in Sept. 769 ff., it is a question of jettison, ἐκβολή (cf., e.g.,
Arist. Eth. Nic: 3. 1 p. 1110a 9 ff., Act. Ap. 27. 18), the legal regulation of
which was taken over by Roman law (Dig. 14. 2 De lege Rhodia de iactu) as a
legacy from the Greek (for detailed information cf. U. E. Paoli, Stud: d:
dirillo Attico, Florence 1930). The formula (cf. Paoli p. 82) is retained in the
συγγραφή of the speech against Lacritus [Demosth.] 35. 11 ἀποδώσουσιν... τὸ
γιγνόμενον ἀργύριον... évreAés πλὴν ἐκβολῆς, ἣν àv où σύμπλοι ψηφισάμενοι
κοινῆι ἐκβάλωνται.
The general issue is thus settled. But in one point of detail scholars have
been led astray by false ingenuity and the misuse of a supposed parallel.
Instigated partly by Housman’s protest (see below) against the traditional
interpretation of σφενδόνας ἀπ᾽ εὐμέτρου, Verrall (ad loc. and Appendix I T)
and W. Wyse, C.R. xiv, 1900, 5, maintained that σφενδόνη indicates an actual
appliance. Headlam and Mazon (‘manceuvrant prudemment la grue’) ac-
cepted unconditionally Wyse's solution; in L-S it is regarded as possible.
Wyse finds in this passage another example of the meaning which σφενδόνη
has in the record of sums of money paid by the Council of Delphi to the
ναοποιοί, Dittenberger, Syll. 241 A 46. Among the payments there for the
year 353-2 is entered the sum for a μαχάνωμα (cf. also ll. 56 and 62) erected in
Kirrha upon ἃ χῶμα. Α component part of this μαχάνωμα is ἃ σφενδόνα.
According to the generally accepted interpretation of the first editor of the
inscription, Bourguet (he was followed by Pomtow in Dittenberger), the
engine was something corresponding to a crane for the unloading of building-
stones shipped to Kirrha from the Peloponnese; the σφενδόνη, the ‘sling’, is
‘la pièce de la machine où l'on place les fardeaux à soulever’. It is surprising
that so careful a scholar as W. Wyse found it possible to believe that this had
any bearing upon Ag. roro, and still more so that others accepted this. The
ey Κίρραι μαχάνωμα is not used on board ship, but in the harbour, upon a
χῶμα specially piled up for the purpose, a pier. In respect of a crane on board
ship for the unloading of freight, it proves nothing at all. Even assuming that
Greek sailors of the fifth century took such engines with them, how can it be
supposed that in the moment of greatest peril on the sea such an appliance
was put into action, with no other purpose than that of throwing parts of the
cargo overboard? The ship was not loaded with heavy containers or railway
engines but with goods such as could at any time be brought up from the hold
and thrown overboard by a few strong members of the crew. Verrall supposes
an appliance for weighing heavy goods during the loading, and accordingly
the situation depicted is for him not the moment of greatest peril on the sea
but the peaceful preparations before departure, and this in spite of βαλών,
and in spite of Sept. 769 ft.
454
 COMMENTARY line 1008

Now we can return to Schütz’s and Hermann’s interpretation of σφενδόνη,

undeterred by Housman’s criticism (J. Phil. xvi, 1888, 272). What the con-
text demands is: ‘ σφενδόνας poetice pro ἐκβολᾶς, instrumentum, quo res
peragitur, rei ipsius loco’ (Schütz), ‘bene attemperato iactw' (Hermann). This
use of σφενδόνη (‘i.q. σφενδόνησις ' Dindorf in the Thesaurus) may be con-
sidered bold, but it is linguistically sound. ἀγχόνη in a number of passages,
in Semonides of Amorgos, in Tragedy and elsewhere means ‘a rope for
hanging',! it also means ‘the hanging’, a fact emphasized by the ancient
grammarians who differentiate between the meanings ἡ ἄγξις and τὸ σχοινίον,
ὁ βρόχος.2 The analogy of words of this type may have facilitated the poet's
extension of the meaning ‘sling’? to that of ‘slinging’.
εὐμέτρου : probably not ‘modico, quantum ad levandam navem sufficit,
actu’ (Schütz and others), but = εὐστόχου (gloss in Tr), so as to get it well
overboard.
With ἀπό here cf. e.g. Thuc. 7. 71. 6 6 δὲ πεζὸς οὐκέτι διαφόρως, ἀλλ᾽ ἀπὸ μιᾶς
ὁρμῆς (‘elsewhere μιᾶι ὁρμῆι᾽, K. W. Krüger, ad loc.) . . . of μὲν... mape-
βοήθουν, où δὲ κτλ. Cf. on 1630 ἀπὸ φθογγῆς.
1008 f. καὶ τὸ μὲν... βαλών. Many have made conjectures, from Stanley
onwards. Since the style is both artificial and terse, and the MS evidence
scanty, the possibility of a corruption cannot be excluded. But there is no
word in the two lines which raises really serious doubts. So Headlam was
probably right in his earlier judgement when he said (footnote to his prose
translation) that he believed the MS reading to be correct, ‘though the
grammatical construction of the words τὸ μὲν πρὸ χρημάτων is uncertain’. He
seems later to have changed his mind with regard to τὸ μὲν πρὸ (posthumous
edition of the Ag.). If τὸ μέν is retained, it cannot, as Klausen wished, mark
the antithesis to 1018 τὸ δ᾽ emi γᾶν «rA., but ‘partem significat' (Hermann).
This would be taken ‘as if to imply τὸ δὲ (the other part) σώσας οὐκ ἐπόντισε
σκάφος. But as the independent clause οὐκ ἔδυ πρόπας δόμος is substituted for
a phrase such as τὸ δὲ σώσας, the principal statement is joined on by οὐδ᾽ '
(Wecklein, following Schneidewin-Hense). By the position of τὸ μέν at the
beginning of the sentence the idea of ‘part’ is emphasized. This idea is appro-
priate to the jettisoning (the treatment of the subject in the Digest, 14. 2,
contains many examples of sacrificing a part of the cargo) and also plays an
important part in the application of the simile : in cases where no blood-guilt
has been incurred, a man can mitigate the consequences of the evil which he
has brought upon himself if he is wise enough to sacrifice a part in order to
avoid complete destruction.
1008. πρό: Stanley took this as ‘tmesis’ (προβαλών) and this view has had its
adherents right up to recent times. But such an order of words would seem
particularly harsh in this passage,* where it is natural to take πρὸ χρημάτων
! In the oth edition of L-S, as against the earlier editions, this use is no longer mentioned.
The article in Passow-Crönert is much better.
2 They associate, perhaps arbitrarily, a difference of accent with the difference in mean-
ing, cf. Chandler, Greek Accentuation, 2nd ed., $ 141, p. 41. Protesting against this Schwyzer,
Griech. Gramm. i. 490. 5, writes: "The abstracts ἀγχόνη (sic) αὐονή ἡδονή '.
3 This meaning occurs as early as Archilochus fr. 3. It is therefore practically irrelevant
whether Cuny, Bull. Soc. Lingu. xxxvii, 1936, τ ff. is really right in his supposition that
σφενδόνη originally meant nothing but ‘band’ and only secondarily ‘thong, sling’.
4 πρό and χρημάτων would in that case stand together only by chance, as it were. It is
455
line 1008 COMMENTARY
together (this is what Weil said in his edition of 1858, cf. also Ahrens, 602).
It is indeed asserted 'simillimum huic exemplum legimus v. 165 ei τὸ μάταν
ἀπὸ φροντίδος ἄχθος χρὴ βαλεῖν ἐτητύμως ' (Nägelsbach), but the assumption
of tmesis there is based on a wrong interpretation. As for the meaning of the
supposed προβάλλειν, the commentators are at variance. Hermann renders
it by protcere, which is linguistically correct, while Wecklein explains ‘casting
overboard beforehand, i.e. before the sinking of the ship, at the right time’,
which can hardly be supported from the usage of προβάλλειν. Others, e.g.
van Heusde and Housman, J. Phil. xvi, 1888, 273, regard πρό as an adverb;
but there is no evidence in Aeschylus for the adverbial use of πρό. The sense
is correctly understood in Linwood’s Lexicon to Aesch. : ‘in behalf of [similarly
Verrall ad loc.] h.e. to save the rest of the property’. πρό is used here as in πρό
twos ἀποθανεῖν etc. ; cf. on 1168. ‘In defence of the goods’ refers of course only
to the main part, after taking away the part to be jettisoned; only a pedant
would require this to be specially mentioned. The meaning is all the more
clear because τὸ μέν stands in front.
1008 f. χρημάτων κτησίων : exalted expression for κτημάτων. Only this pas-
sage and S. Trach. 690 κτησίου βοτοῦ are quoted in the lexicons for κτήσιος
except for its use as applied to deities (cf. however on 1038).
1009. ὄκνος : it is questionable whether it is necessary to follow Blomfield,
Hermann, and others in assigning to the word simply the meaning ‘fear’,
which it has e.g. in several passages in Sophocles. The renderings ‘caution’
(L. Campbell), ‘cautious diffidence’ (Headlam) appear more appropriate ; the
passages collected by Headlam (in several of which ὄκνος is opposed to
θράσος) also show that the meaning sometimes borders on that of ‘fear’.
What is meant, at all events, is reluctance and unwillingness to continue the
course with the whole freight despite the heavy seas.
1011. The anacoluthon after the nominative of a participle will hardly be
questioned now by anyone. See Wecklein, Stud. z. Aesch. 14 f., Wilamowitz on
E. Her. 186 and Berti’s monograph (cf. on 12). Cf. e.g. Suppl. 4461. καὶ
γλῶσσα τοξεύσασα μὴ τὰ καίρια, γένοιτο μύθου μῦθος ἂν θελκτήριος, Cho. 520 1.
τὰ πάντα γάρ τις ἐκχέας ἀνθ᾽ αἵματος ἑνός, μάτην ὁ μόχθος.
δόμος (household, family, race, as so often) in spite of ἔδν and in spite of
what immediately precedes it. It is characteristic of Aeschylus not to depict
the object used for comparison (in this case the ship) separately, but, even at
the expense of the pictorial effect, to intermingle it with features of the thing
which it serves to illustrate. Cf. on 59, 395, 732 ff., 966. It is an important
point here that, in the case of a lesser fault and damage, it is not the whole
clear at once that it is quite a different matter in Prom. 878 f. ὑπό u’ αὖ σφάκελος καὶ φρενο-
πληγεῖς μανίαι θάλπουσι (Ag. 1215 f. ὑπ᾽ αὖ με δεινὸς... πόνος arpoßei) where preposition,
object, and verb are closely bound together by the sense. In 5. Ant. 1272 ff. ἐν δ᾽ ἐμῶι
κάραι θεὸς... ἔπαισεν, the words ἐν ἐμῶι xapaı go together, even if this is a case of tmesis
as is held by Nauck and Bruhn, but Wunder, Schneidewin, and Jebb are probably right in
construing ὁ δὲ θεὸς ἐν τῶι ép κάραι ἔπαισέ με. There is no tmesis in Cho. 573 f., 1024. For
a usage like Xenarchos fr. 2. 1 (ii. 468 Kock) ὡς ὑπό τι (Porson, εὐποτι Ath.) νυστάζειν γε
καὐτὸς ἄρχομαι there is a special reason, cf. Wackernagel, Syntax, ii. 174.
1 It may be doubted whether Wackernagel, Unters. zu Homer, 45 n. 1, is right in his
interpretation of Prom. 682 γῆν πρὸ γῆς ἐλαύνομαι, which seems to have been accepted by
L-S s.v. πρό C. For a different view see van Leeuwen and Starkie on Ar. Ach. 235. But in
no case would that fixed phrase, used as an almost proverbial formula, justify the assump-
tion of an isolated πρό in Aeschylus serving as an adverb.

456
 COMMENTARY lines 1014 f.

race which perishes, whereas, when blood-guilt has been incurred (1018), it is
visited upon children and children’s children. What is here stated negatively
in the strophe must be borne in mind to supplement positively what is said
in the antistrophe.
1012. When Headlam followed Hermann, Paley, and others in clinging to
πημονᾶς, he apparently did not appreciate at their true value the objections
brought against it. These objections were formulated most pungently by
Housman, J. Phil. xvi, 1888, 274: ‘the phrase πημονᾶς γέμων ἄγαν is ridiculous :
as if there were such a thing as πημονᾶς γέμειν μετρίως! But further: .. . it is
certain that the ship is [1008 f.] represented as laden not with πημονή but with
χρήματα κτήσια. Therefore we have to say that the χρήματα themselves are
here called πημονή as leading to disaster by their too great abundance. Now
perhaps there are places where wealth can be called πημονή, but this is a place
where it cannot: it cannot be called πημονή when it has just been called
ὑγίεια ᾿. He then refers to the passage Sept. 768 f. πρόπρυμνα δ᾽ ἐκβολὰν φέρει
ἀνδρῶν ἀλφηστᾶν ὄλβος ἄγαν παχυνθείς, which Schütz had made the main
support of his emendation πλησμονᾶς. To these considerations it should be
added that Aeschylus certainly (and others, too) counted an excess of pros-
perity, riches, etc., as one of the main causes of ruin. Wilamowitz is in agree-
ment with Housman when he writes, Hermes, xxxiv, 1899, 612: ‘It is obvious
that the ship's cargo is the opposite of πημονή ; only the idea of ‘“‘overloaded”’
is suitable.’ Both critics confirm, as far as the main issue is concerned, that
Schütz has hit the mark with πλησμονᾶς. Karsten (p. 242 of his commentary)
and Housman ought not to have introduced in its place a non-existent
παμονά, nor Wilamowitz an equally unsupported πλημονή. His observation
‘ πλησμονή always means being over-full, not simply full’ is quite correct but
cannot be used as an argument against reading wAnopovds here. “Being over-
full’ is exactly what is required, and for the use of pleonasm (γέμων ἄγαν
together with πλησμονᾶς) in ideas of this kind cf. on 378.
1013. ‘Rediit in constructionem poeta, ἐπόντισε referens ad ὄκνος ’ Hermann,
in substance rightly. But it is perhaps better to abstract from the context
the appropriate subject (here the ‘seafarer’), cf. on 71.
The manner in which in the section 1008-13 (καὶ τὸ μὲν... σκάφος) one
train of ideas leads to a kindred though somewhat different one is a good
instance of the type of almost imperceptible transition from one stage to the
next which we often notice in Aeschylus (cf. e.g. on 429). The connexion with
the beginning of this stanza lies in the idea ‘an excess of goods’, note especially
1012 δόμος πλησμονᾶς γέμων ἄγαν. In addition the comparison of the sea-
voyage begun in roos f. is continued. But, on the other hand, a new idea
appears here: the determined sacrifice of a part of the goods may avert
complete destruction in the moment of extreme peril. This theme, ‘rescue
at the last moment’, is continued in the lines that follow and illustrated by a
further comparison.
1014 f. In Hermann, in Vitelli-Wecklein, and consequently in most of the
more recent editions it is not noted that in F the reading is clearly Aids instead
of διὸς (van Heusde and Mazon do mention it), perhaps a survival of an uncial
corruption, which could have been easily corrected.
πολλὰ . δόσις: of the same thing Sept. 360 f. πολλὰ... γᾶς δόσις
(Blomfield).
457
lines 1014 f. COMMENTARY

τοι: Denniston, Particles, 538, notes on this and several other passages:
‘The Aeschylean chorus often uses ro: in the air, as it were, without any
obvious personal reference.” This present passage may be regarded as an
instance of the frequent use of ro. in general maxims (Denniston, 542 f.)
in which ‘the tense is usually present or gnomic aorist’.
Verrall puts a comma after ἀμφιλαφής re! and translates the following words
thus: ‘and rids the plague of hunger out of the annual field’, 1.6. takes ἐξ
ἀλόκων with ὥλεσεν (so, too, evidently misled by Verrall, A. Platt: ‘and
banishes famine from the fields in their season’). This makes the thought
absurd. The usual interpretation, which takes «ai ἐξ ἀλόκων ἐπετειᾶν with the
preceding words, should be retained. As far as the grouping in detail is
concerned, it is natural to take δόσις ἐκ Ais together. And we should not be
diverted from doing so by the desire to bring out as strongly as possible the
parallelism between ἐκ Aids and ἐξ ἀλόκων (as e.g. Wecklein: ‘ ἀμφιλαφής
from two sides, from Zeus, who as lord of the seasons gives prosperity, and
from the fertility of the soil’). Rather, the definition of πολλὰ δόσις ἐκ Aids
is provided by the phrase ἀμφιλαφὴς . . . ἐπετειᾶν, which is closely bound
together by re καί; the rich harvest is ‘abundant and coming from the
ploughed fields which bear their produce year by year’. So πολλὴ δόσις ἐκ
Aids is defined first purely quantitatively (ἀμφιλαφής), secondly by the state-
ment that the blessing of the harvest is yearly renewed. It has been said
that πολλά is incompatible with ἀμφιλαφής. But this is not correct. First
πολλά characterizes the subject in general, then there follows its more
elaborate description by means of suggestive details in the phrase ἀμφιλαφής
. ἐπετειᾶν.2 It may also be pointed out, by the way, that if ἐκ Aids were
taken with ἀμφιλαφής, the position of re would not be beyond question.
1015. ἀμφιλαφής: W. Aly, De Aesch. cop. verb. 46 f., makes it appear prob-
able that the word was of Ionic origin, cf. also Bechtel, Griech. Dial. iii. 276.
1017. Schütz: ‘ ὥλεσεν ferri quidem et explicari potest. Fortasse tamen ab
Aeschyli manu venit ἤλασεν. The conjecture, suggested with laudable
caution, has been accepted by many editors. But is not ὥλεσεν more forceful?
The destroying plague is, by grace of Heaven, itself destroyed. When the
new harvest comes in, the famine has already been raging for some time, just
as the ship, when saved at the last moment, was in extreme peril and had
lost part of its cargo.
1014—17. With the famine, the question is no longer, as it had been up to this
point, one of excessive prosperity and success, but only of an evil the worst
consequences of which could still be averted at thelast moment. This thought,
then, prepares the way for the antithesis which follows (1018 ff.) ; it does not,
however, come in here (1014) abruptly, but arises from the idea of the pre-
ceding sentences and then takes an independent turn. The circumstances
which lead to the jettisoning of cargo offer implicitly an example of safety
snatched from disaster just before it 15 too late.
1018 ff. There is a faint possibility of a corruption of the text, cf. p. 452 n. I.
11 do not know whether the punctuation in F (a colon after ἀμφιλαφής re) is due to the
same idea which guided Verrall or whether the agreement is fortuitous.
2 It is natural that when a poet is depicting the bounteous yield of the soil, he should not
be sparing of words expressing bounty. Cf. e.g. Horace, Odes 1. 17. 14 fl. hic tibi copia
manabit ad plenum benigno ruris honorum opulenta cornu.

458
 COMMENTARY line 1020

The use of ἅπαξ here! and in the closely related passage Eum. 648 (see below)
is correctly listed in L-S ἅπαξ II in the group of temporal and conditional
conjunctions (ἣν ἅπαξ. . ., εἴπερ ἅπαξ etc., ‘when once’, ‘if once’).
For the thought cf. Cho. 48 ri γὰρ λύτρον πεσόντος αἵματος πέδοι, 66 f., Eum.
261 f., 647 f. (there is in the words a strong echo of this passage in the
Agamemnon) ἀνδρὸς δ᾽ ἐπειδὰν αἷμ᾽ ἀνασπάσηι κόνις (cf. Eum. 980) ἅπαξ
θανόντος, οὔτις ἔστ᾽ ἀνάστασις. τούτων ἐπωιδὰς οὐκ ἐποίησεν πατὴρ oùuds κτλ.
It is one of the themes, then, which bind the trilogy into a single whole. Of
course, the idea is not confined to the Oresteia. Cf. especially Suppl. 443-51
and Sept. 734 ff. (in the details, too, there are contacts with the passages
quoted here), but also e.g. E. Suppl. 775 ff. τοῦτο γὰρ μόνον βροτοῖς οὐκ ἔστι
τἀνάλωμ᾽ ἀναλωθὲν λαβεῖν, ψυχὴν Bporeiav' χρημάτων δ᾽ εἰσὶν πόροι. In A. Suppl.
443-51 the arrangement, too, is similar to that of our ode:? two preparatory
propositions (evils for which there is repair), then the principal thought
(bloodshed is irreparable).
1019. θανάσιμον. For the occurrence of this word see C. Arbenz, Die Adjek-
tive auf -yos (Diss. Zürich 1933), 70f. Here it is used predicatively as a
complement to πεσόν. Cf. S. Oed. R. 959 εὖ ἴσθ᾽ ἐκεῖνον θανάσιμον βεβηκότα,
E. Hec. 1032 f. ἐλπὶς 7 σ᾽ ἐπήγαγεν θανάσιμον πρὸς Ἅιδαν.
πρόπαρ (the accent is uncertain, the word ‘is possibly abbreviated from
Homer’s προπάροιθε in accordance with some pattern or other’, Wackernagel,
Syntax, ii. 232): it is to be hoped that this is what Aeschylus wrote. Curious
explanations have been put forward. Paley: ‘at a man’s feet’; Wecklein:
‘the blood pouring from the breast flows to the ground in front of the man’;
Pliiss: ᾿πρόπαρ ἀνδρός "he who was before a real living man’’.’ Headlam,
however, is probably right: ‘But human blood once fallen aforetime’ ; i.e.
πρόπαρ is adverbial as it is in the only other passage in which it occurs in
Aeschylus, Suppl. 791 πρόπαρ θανούσας ‘Aidas ἀνάσσοι. There too πρόπαρ is
clearly temporal (cf. the immediately preceding θέλοιμι δ᾽ ἂν μορσίμου βρόχον
Tuxeiv . . . πρὶν krÀ.), but at the same time the meaning 'prius' passes into
‘potius’, as it does so often: ‘before it comes to that, sooner, rather’. In Ag.
1019 the temporal sense seems to be stressed. ‘When blood has once been
shed previously’, then there is no escape afterwards, never in all the rest of
time.
1020. τίς ἂν κτλ. The order of words is quite normal. The ‘extended’ object,
which stands in forceful antithesis to the preceding words, stretches from τὸ
δ᾽ ἐπὶ γᾶν down to αἷμα, after that there begins the new kolon introduced by
the interrogative pronoun, and ἄν comes in its proper place, i.e. second in
this kolon. Cf. Ed. Fr., Kolon und Satz, ii. 341; what there is said about the
effect of the antithetical structure upon the formation of this sentence is
supplemented, although from a different starting-point, by G. Thomson,
C.Q. xxxiii, 1939, 147 f. He shows that when uév . . . δέ- kola are put first and,
as is often the case, emphasis is laid upon them, the interrogative word is very
frequently forced away from the beginning of the sentence to the beginning

! This passage was presumably in the mind of the commentator who paraphrased (schol.
ad loc.) Eum. 261 with the words τὸ ἅπαξ χυθὲν αἷμα δυσίατον.
2 Cf. van Otterlo, Mnemos. S. III, viii, 1940, 164 ff., who compares these two Aeschylean
passages with I 406 ff. ληϊστοὶ μὲν γάρ re Bdes . . . κτητοὶ δὲ τρίποδες... ἀνδρὸς δὲ ψυχὴ πάλιν
ἐλθεῖν οὔτε Actor? οὔθ᾽ ἑλετή κτλ.
459
line 1020 COMMENTARY
of the second kolon, as e.g. in Ag. 598. Cf. on this collocation Schadewaldt,
‘Die Niobe des Aischylos', Sitzgsb. Heidelb. Akad., 1933-4, 3. Abhdl., 19 n. 2.
1021. ἐπαείδων, as the flow of blood can be stopped, in the case of a wounded
man who still lives, by ἐπαοιδή (Blomfield refers to 7 457 f. ἐπαοιδῆι δ᾽ αἷμα
κελαινὸν ἔσχεθον and the scholia on it); cf. in the parallel passage (see above
on 1018 ff.) Eum. 649 ἐπωιδάς. The verb ἀνακαλεῖν (or ἀνακαλεῖσθαι) was, as
early as Aeschylus (Pers. 621), used to denote the calling up of the spirit of
a dead person by means of magic and particularly by incantation. Aeschylus
uses the same verb here, where it is a case not of a spirit being summoned
but of a dead man being brought back to complete physical existence.
In these lines (1018 ff.) the Chorus speaks of murder only in a general way,
in the framework of the γνώμη. But we may ask (and perhaps an ancient
audience may have asked the same question) what is it that leads to this
train of thought, and which murder have the old men in mind? They know
about the bloodshed in the past, the sacrifice of Iphigeneia, and also of the
slaughter of the many before Troy, and they are convinced that τῶν πολυ-
κτόνων οὐκ ἄσκοποι θεοί (461 f.) and that wrong must bring retribution (first
strophe and antistrophe of this chorus). In their fear they may envisage, as a
threatening possibility, new bloodshed to requite the old. But our eagerness
to discover a particular application of the general statement can hardly
venture farther, without, by rough treatment, injuring the delicate substance
of what the poet has hinted rather than said. The utterances of the Chorus
here are still far from the clarity of its declaration in 1335-40 ; in between lie
the revelations of Cassandra. Cf. on 1151. The law of wrong and retribution
dominates this ode, but there is in it no reference to any particular wrong of
the past, and as regards the future, it shows itself in one thing alone: fear of
something which cannot be averted, though its nature is still completely
obscured. At this juncture the audience may be expected to realize that the
words of the Chorus somehow refer to past and future bloodshed and subse-
quent revenge in the house of Agamemnon. A greater definiteness will not
be demanded by the hearer who is willing to follow where the poet leads.
1022-4. οὐδὲ τὸν ὀρθοδαἢ . . . Ζεὺς αὔτ᾽ ἔπαυσ᾽ is the MS reading. The
emendation of the last words and the syntactical interpretation of the
sentence are very controversial. Hermann, followed, unfortunately, by
Wilamowitz in his edition,’ has cut into what is sound, by removing Ζεὺς
from 1024 and putting it at the beginning: Ζεὺς δὲ τὸν κτλ. (Hermann has
curtailed 1024 in a wholly arbitrary fashion). To refute him it is quite suffi-
cient to point out that 1022 is an example of a very old type of introducing a
παράδειγμα (cf. on 1040), cf., e.g., Z 130 f. (previous line οὐκ dv ἔγωγε θεοῖσιν
ἐπουρανίοισι μαχοίμην) οὐδὲ yap οὐδὲ Δρύαντος vids, κρατερὸς Avköopyos, δὴν
ἦν, Z 117 οὐδὲ γὰρ οὐδὲ βίη ᾿Ηρακλῆος φύγε κῆρα. So the introduction of the
παράδειγμα by οὐδὲ τὸν ὀρθοδαῆ should not in any circumstances be impugned.
οὐδέ differs from οὐδὲ γάρ only in that the explanation is not especially marked,
i.e. that a mere ‘not even’ has taken the place of ‘for not even’.
Zeus did indeed check the temerity of Asklepios. But to elicit a statement
to this effect from the words οὐδὲ τὸν... Ζεὺς éravo' (or ἀπέπαυσεν) is beyond
the skill of any commentator. It was not quite straightforward of Stanley
τ But not in his article, Hermes, xxxiv, 1899, 56 n. 2 (cf. also Griech. Trag. ii. 117), where
he writes οὐδὲ τὸν ὀρθοδαῆ . . . Ζεὺς ἀπέπαυσεν ἐπ᾽ aßAaßeiaı.
460
 COMMENTARY lines 1022-4

to translate his text οὐδὲ... Ζεὺς αὖτ᾽ ἔπαυσ᾽ as ‘Neque illum . . . Jupiter
prohibuisset’ (Nägelsbach translates similarly), but we may sympathize with
the predicament of the translator. Schütz attempted a different solution by
taking 1022 ff. οὐδὲ τὸν κτλ. as a question (cf. Schütz’s exactly corresponding,
but equally unjustified, interpretation of the sentence beginning with οὐδὲ
in 1523 as ἃ question) ; he has been followed by Ahrens (p. 605) and others and
later by Headlam. It may well be doubted whether the question would be
sufficiently clearly marked. Besides, if we took οὐδὲ «rA. as a question, we
should have to sacrifice altogether the connexion with the formula of transi-
tion quoted above from Homer οὐδὲ yap ὁ δεῖνα ‘even this or that hero did
not...'. The correct solution, I think, has been put forward long since, first
of 411, as far as I know, by Friedrich Martin, Observ. crit. in Aesch. Orest.
(Schulprogramm Berlin 1837) p. 6. In 1024 he changed αὔτ᾽ ἔπαυσ᾽ by an easy
emendation into dv ἔπαυσεν. This has been accepted by Peile, Conington,
Paley, and more recently by Lawson and G. Thomson. The emendation
provides the text which, as we saw, was translated by Stanley. As far as the
meaning is concerned, the conjecture is excellent. In form, too, it is un-
objectionable. For, in the transition from a general γνώμη to the παράδειγμα,
this particular variation of the purely narrative form is found in choral
lyric: Bacchyl. 5. 94 ff. χαλεπὸν θεῶν παρατρέψαι νόον ἄνδρεσσιν ἐπιχθονίοις. καὶ
γὰρ ἂν πλάξιππος Οἰνεὺς παῦσεν . . . σεμνᾶς χόλον Ἀρτέμιδος ‘else would horse-
smiting Oeneus have appeased the wrath of Artemis’ (Jebb). Cf., too, Pind.
Ol. 9. 29 ff., where the general maxim ἀγαθοὶ δὲ καὶ σοφοὶ κατὰ δαίμον᾽ ἄνδρες
ἐγένοντο is followed by the παράδειγμα in the form of the apodosis of an unreal
conditional sentence: ἐπεὶ ἀντίον πῶς av τριόδοντος 'HpakAéns σκύταλον τίναξε
χερσίν, ἁνίκα κτλ. Against the text recommended here it must not be ob-
jected that av should (as is often the case) stand nearer the beginning of the
sentence. Cf. in general e.g. Cho. 700 ff. ἐγὼ μὲν οὖν ξένοισιν ὧδ᾽ εὐδαίμοσιν
κεδνῶν ἕκατι πραγμάτων ἂν ἤθελον γνωτὸς γενέσθαι, Suppl. 509 καὶ πῶς βέβηλον
ἄλσος ἂν ῥύοιτό με, Soph. fr. 513. 3 N. (557 P.) ὁ χρυσὸς ἧσσον κτῆμα τοῦ κλαίειν
ἂν ἦν, and in particular, in the case of a negative introducing the sentence, as
here, e.g. 5. Oed. C. 45 ὡς οὐχ ἕδρας γῆς τῆσδ᾽ av ἐξέλθοιμ᾽ ἔτι, 941 ff. γιγνώσκων
δ᾽ ὅτι οὐδείς ποτ᾽ αὐτοὺς τῶν ἐμῶν ἂν ἐμπέσοι ζῆλος ξυναίμων. Moreover it is
evident that the sentence begins with a kolon which consists of an ‘extended’
object (cf. on 1020). The position of this object indicates that it is strongly
emphasized? so as to make it clear that the reference is to the story of
! It should, however, be noticed that Hermann in the appendix to Humboldt’s transla-
tion (1816) read as his text Ζεὺς dv αὖτ᾽ ἔπαυσεν én’ ἀβλαβείαι, where ἂν is correctly restored
but αὖτ᾽ is wrong, to say nothing of the metre. He later revoked this with some vehemence.
2 A similar example in a different style is Mimnermus fr. 11, where οὐδέ kor' ἂν μέγα
κῶας ἀνήγαγεν αὐτὸς ᾿Ιήσων ... οὐδ᾽ ἂν ἐπ᾽ ' Ὠκεανοῦ καλὸν ἵκοντο poor ‘is probably a mytho-
logical exemplum to illustrate some special experience’ (Pfeiffer, Philol. Ixxxiv, 1929, 143)
or, perhaps, some γνώμη. In the passage y 218 ff., which was recognized by ancient critics
as an interpolation, οὐδέ κεν "Apyein 'EAévg . . . ἀνδρὶ παρ᾽ ἀλλοδαπῶι ἐμίγη φιλότητι καὶ εὐνῆι,
εἰ ἤιδη «7À., the form is similar, yet there is a difference: for there the παράδειγμα introduced
in the form of the apodosis of an ‘unfulfilled’ conditional sentence is not simply related to a
γνώμη, but connected with a protasis. On the other hand, an exact parallel to the examples
from lyrics quoted above is to be found in Heraclitus fr. 40 πολυμαθίη νόον ἔχειν οὐ διδάσκει"
*HaíoBov γὰρ ἂν ἐδίδαξε καὶ Πυθαγόρην κτλ. : first the γνώμη, then (in its tersest form) the
πάραδειγμα.
3 This is a further argument against the conjecture Ζεὺς δὲ τὸν ὀρθοδαῆ.
461
lines 1022-4 COMMENTARY

Asklepios: τὸν ὀρθοδαῇ τῶν φθιμένων ἀνάγειν. Thus if Martin's emendation

(Ζεὺς ἂν ἔπαυσεν) is accepted, ἄν will have a normal position, coming second
in the second kolon of the sentence.
After the question (1017-21), which is equivalent to a negative statement,
οὐδέ is perfectly natural.
1022. ὀρθοδαῆ here only.
1023. τῶν φθιμένων : genitive of separation (cf. on 479), governed by ἀνάγειν.
Cf., e.g., fr. 258 N. καὶ ψευδόδειπνα πολλὰ μαργώσης γνάθου ἐρρυσίαζον, ‘eripiebant
(Harpyiae nimirum) multas epulas, quibus nemo frueretur, cupidis malis
(Phinei)’, according to Lobeck’s restoration and Wilamowitz’s (Aesch. trag.
p. 177) convincing explanation.
1024. Only ἐπ᾽ ἀβλαβείαιϊ is supported by the evidence of the MSS: not only
does Triclinius write this, but also the slight misspelling? in F, caused by the
later? pronunciation (avlavia), points to it (and not to εὐλαβ.). The function
of ἐπί with the dative is suitably paraphrased in the scholion: wore μὴ
ἕτερον βλαβῆναι. For the consecutive force of the construction cf., e.g., Sedt.
879 f. μελέους θανάτους ηὕροντο δόμων ἐπὶ λύμηι. But here, as in other cases,
care must be taken to avoid drawing too sharp a distinction between con-
secutive and final. There is an excellent parallel to the expression here, with
an almost synonymous substantive, in Theogn. 757 f. Zeus μὲν τῆσδε πόληος
ὑπειρέχοι αἰθέρι ναίων αἰεὶ δεξιτερὴν χεῖρ᾽ em ἀπημοσύνηι, where Hudson-
Williams explains ἐπί by ‘for, to secure’. It would also be possible to render,
with more consecutive force: ‘so that ἀπημ. prevails’. ἐπ᾽ ἀβλαβείαι means
then: ‘in order that no harm should happen’ or ‘with the result that no
(further) harm happened’. The ‘unharmfulness’ is concerned with a particular
sphere. In an inscription of Gortyn (not in the great law, but in a regulation
of the right of mortgage), Collitz-Bechtel, iii no. 4986, we read ı1 f. ὀμνύμεν
δὲ À μὰν τούτω μέν ἐστι ἀβλοπίαι δικαίως krA., upon which Blass notes:
‘ ἀβλοπίαι = ἀβλαβείαι, without damage to the rights of a third party, i.e.
synonymous with δικαίως ’.. Cf. also Wackernagel, Syntax, ii. 288: ‘for
Erythrai we have the evidence of an inscription of the third century B.c.
[Dittenberger, Sy. 1014. 67] mentioning the AßAaßia:; these have been
regarded as something akin to the Erinyes [cf. Hiller v. Gaertringen in
Dittenberger, loc. cit., n. 22; A. B. Cook, Zeus, ii. 1113 n. 1]; it may be re-
membered that in the hymn to Hermes 393 ἐπ᾽ ἀβλαβίηισι νόοιο means
“upright” and Cicero, Tusc. 3. 16, renders innocentia by aßAdßera, and also
that in Cretan ἀβλοπίᾶι with or without δικαίως means “in a just fashion" ’.
In the present passage, too, the meaning is unmistakably this: ‘so that the
order of law, given by God, should not be disturbed’; we might say, the
natural order, in accordance with which every creature has its part assigned
δικαίως, and with which it would be incompatible that a man should escape
death or be rescued from it afterwards.* Should anything of that kind occur,
1 The statement in L-S s.v. εὐλάβεια is correct, but the unfortunate formulation s.v.
ἀβλάβεια has had the consequence that in Buck and Petersen’s Reverse Index, 128, the earliest
instance of ἀβλάβεια is said to be in Cicero.
2 Wecklein in Silzgsb. Bayer. Ak., Phil.-hist. Kl. 1921, 5. Abhdl., p. 24, still maintained
that em’ αὐλαβεία in F was a slip of the pen for ἐπ᾽ εὐλαβείαι.
3 Here it is irrelevant that similar misspellings, such as pavdous for ῥάβδους, occur at an
early period, cf. Schwyzer, Griech. Gramm. i. 198.
* It is on this point particularly that the relationship of ἀβλάβεια and ’Epwos is clear. Cf.
462
 COMMENTARY lines 1025-9

then the whole order, which is endangered by whatever is ἄγαν, must suffer
serious harm. For this reason Zeus intervenes ἐπ᾽ ἀβλαβείαι.
The story of the punishment of Asklepios at the hands of Zeus was presum-
ably taken by Aeschylus, as by Pindar (P. 3), from the Hesiodic (frr. 122-7
Rzach, 3rd ed.) Eote (reconstructed by Wilamowitz, Zsyllos 70 ff., and Griech.
Tragódien, ii. 71 ff.). Aeschylus must have had a high regard for that tale,
since it provided powerful testimony for the justified φθόνος θεῶν, which is
based upon the eternal order of things. To resist any encroachment upon the
divine order, the god is bound to intervene with his restraining force, his
φθόνος, cf. p. 349 f. Aeschylus refers to the continuation of the story in Suppl.
214 ἁγνόν τ᾽ Ἀπόλλω φυγάδ᾽ an’ οὐρανοῦ θεόν and Eum. 723 f.
1025-9. The understanding of this sentence, particularly of the subordinate
clause, is rendered difficult by the complex meaning of some of its phrases.
But there is no reason for altering the text. The older conjectures have been
convincingly refuted by Ahrens (p. 606). He has also established the inter-
pretation at two decisive points: ‘First πλέον φέρειν can hardly mean anything
else but proficere, as Blomfield explained, or rather “to gain”, in which mean-
ing πλέον φέρεσθαι is more commonly found, but the active is used in S.
Oed. R. 1189 f. Further, the word-order makes it incredible that ἐκ θεῶν is to
be taken with μοῖρα τεταγμένα, as most commentators wish; it must either
be construed with εἶργε, as Karsten proposes, or with πλέον φέρειν, as the
sense seems to me to require.’ In spite of this cogent reference to the word-
order, the majority of commentators! have persisted in taking together
μοῖρα ἐκ θεῶν τεταγμένα. In support of ἐκ θεῶν being construed with πλέον
φέρειν it may be well to refer to S. Oed. R. 590 νῦν μὲν γὰρ ἐκ σοῦ πάντ᾽ dvev
φόβου φέρω. Now the path is clear for the rest of the interpretation. The
juxtaposition οὗ μοῖρα μοῖραν makes use of different shades of meaning of the
word. The τεταγμένα μοῖρα is the lot of man in general, established and
ordered by God (or by ‘the natural order’) ; it is his fate or destiny. τεταγμένα
as in Eum. 943 ff. μῆλά τ᾽ εὐθενοῦντα ya ξὺν διπλοῖσιν ἐμβρύοις τρέφοι χρόνωι
τεταγμένωι, where τεταγμένωι similarly refers to the natural order. On the
other hand, the following μοῖραν indicates the portion or lot (powers, possi-
bilities etc.) with which the individual is endowed ;? the μὴ πλέον φέρειν
which is tacked on in explanation makes this clear. “But if the established
destiny did not hold (our) endowment in check, so that it (the endowment)
cannot carry away ἃ surplus (increase its gain)’. εἴργειν means exactly the
action with which a higher power, a universal law, holds in check man and
the famous passage T 418 ds dpa φωνήσαντος (Xanthus, Achilles’ horse) "Epwves ἔσχεθον
αὐδήν, on which the BT scholia note: ἐπίσκοποι γάρ εἰσι τῶν παρὰ φύσιν, in expression
anachronistic, but in fact essentially apposite. The corresponding idea is fully developed
in Heraclitus fr. 94 Ἥλιος yap οὐχ ὑπερβήσεται μέτρα" εἰ δὲ μή, ᾿Ερινύες μιν Δίκης ἐπίκουροι
᾿ἐξευρήσουσιν. ‘Just as heaven and earth and sun and moon . . . without respite and in the
same way everywhere operate in eternal and immovable stability : so too the Erinyes are
to be regarded like a natural law in the moral world’ (Otfried Müller, Aesch. Eum. 181).
™ Conington had already noted: ‘The order is perhaps against joining ἐκ θεῶν with
τεταγμένα, yet it is not to be taken with μοῖραν... but connected generally with elpye.’
Paley construes as Ahrens does: ‘But if the appointed law of fate did not hinder fate from
getting further assistance [this is not right] from the gods.’
2 Cf. in general Nägelsbach, Hom. Theologie, 2nd ed., 124 f., Wilamowitz, Berl. Sitzgsb.
1908, 333: ‘Pindar and his audience hear . . . in μοῖρα the μέρος.᾽ The meaning ‘part’ is
clear, e.g., in Prom. 631, Cho. 238; cf. on 1588.

463
lines 1025-9 COMMENTARY

his allotted endowment, in order that he should not extend himself beyond
the measure assigned to him once and for all. Cf. Pind. N. 7. 5 f. ἀναπνέομεν
δ᾽ οὐχ ἅπαντες ἐπὶ ica: εἴργει δὲ πότμωι ζυγένθ᾽ ἕτερον ἕτερα, N. 6. x ff. ἕν
ἀνδρῶν, ἕν θεῶν γένος . . . διείργει δὲ πᾶσα κεκριμένα δύναμις (this corresponds,
though the point of view is different, to the τεταγμένα μοῖρα of Aeschylus),
ὡς τὸ μὲν οὐδέν, ὁ δὲ... οὐρανός. For the infinitive attached epexegetically
(see the paraphrase above) cf. Hdt. 8. 98. ı τοὺς οὔτε νιφετός, οὐκ ὄμβρος, οὐ
καῦμα, οὐ νὺξ Epyeı μὴ οὐ κατανύσαι τὸν προκείμενον αὐτῶι δρόμον τὴν ταχίστην.
That the infinitive here, introduced by μή, is best taken as epexegetic, seems
to me probable from Xen. Anab. 3. 3. 16 ἡμεῖς οὖν ei μέλλομεν τούτους εἴργειν
ὥστε μὴ δύνασθαι βλάπτειν ἡμᾶς πορευομένους κτλ. If, however, the reader
prefers to make the infinitive depend simply on etpyew (as on κωλύειν etc.), it
makes little difference.
1028 f. A fine and vivid picture. In spite of all the hesitations which bind
the tongue, the heart in the fullness of its fear would not wait till the tongue
was ready to speak, but would anticipate it and, as it were, in the tongue’s
place, pour out that with which it is filled. Schiitz, who originally under-
stood it correctly, later reversed the relation of tongue and heart by his
alteration of the text and thereby impressed Hermann, that implacable
logician, and also Platt (J. Phil. xxxii, 1913, 63) and A. Y. Campbell. Ahrens
has no part in this, but he, too, declares (607): ‘the καρδία προφθάσασα γλῶσσαν
is not to be tolerated.’ A poet has now and then a hard time of it with his
learned commentators. It is not necessary to read the chorus particularly
carefully, to realize how indispensable καρδία is as the subject of what follows
(1030 ff.). The idea which lies closer to normal experience, namely that the
tongue often outruns the mind, is expressed in the popular saying in [Isocr.]
I. 41 πολλοῖς yàp ἡ γλῶττα προτρέχει THs διανοίας (quoted by Schneidewin),
and in the sayings of the Seven Wise Men, Chilon 14 (Diels-Kranz, Vorsokr.
i, 5th ed., 63) ἡ γλῶσσά σου μὴ προτρεχέτω τοῦ νοῦ. It is possible that Aeschylus
had this in mind (Schadewaldt, Hermes, 1xxi, 1936, 42 n. 3).
1029. τάδε: what is on our minds now. |
The thought of 1025-9, then, is this: ‘If my portion and the possibility of
my success were not so narrowly and strictly limited, my heart would at
once pour out what is on my mind.’ There is no indication whether this
‘pouring out’ would consist of a warning to the king or of a prayer for the
prevention of ill. But it is no matter, for the decisive factor here is the
thought that all such utterances are bound to be completely useless, for they
would aim at a πλέον φέρειν, and it is just that which is denied us human
beings. The καρδία of the old men is filled with the knowledge of the basic
conditions of our existence, so it can only be now οὐδὲν ἐπελπομένα ποτὲ καίριον
ἐκτολυπεύσειν.
1030. βρέμειν frequently of seditious (cf. Eum. 978) or indignant murmuring.
Here it is perhaps directed against Clytemnestra. ὑπὸ σκότωι corresponds to
the (449) σῖγα Baëlew. Pind. P. 11. 30 ὁ δὲ χαμηλὰ πνέων ἄφαντον βρέμει.
1033. ἐκτολυπεύσειν. We need not go into the question whether the verb, as
is generally supposed, meant originally ‘unravel, wind off a skein of wool’ or
was only a mildly strengthened form of τολυπεύειν. Possibly the language of
the later Epic showed here and there the tendency to that amplification of the
verb by e£, without any appreciable alteration or intensification of the sense,
464
 COMMENTARY lines 1025 ff.

which has been demonstrated from Ionic and the language of Tragedy by
Rutherford, The New Phrynichus, p. 7, cf. also, for Tragedy, Wilamowitz on
E. Her. 155 and the note below on 1244 ἐξηικασμένα. It is in the Aoris, where
the only other example of ἐκτολυπεύειν occurs, that we find (329) the only
instance of ἐξεναίρειν instead of evaipeıw. But whatever the truth is about its
original meaning, everything goes to show that for Aeschylus the late Epic
verb ἐκτολυπεύειν was synonymous with the Homeric roAvrevew. The gram-
marians, too, understood it in this way, as can be seen from schol. T Hom.
& 86 (τολυπεύειν ἀργαλέους πολέμους) ἐκτελεῖν 'Nd ὁπόσα τολύπευσε σὺν αὐτῶι
καὶ πάθεν’ (2 7)" ἐξ οὗ καὶ τολύπη τὸ κατειργασμένον ἔριον as compared with
Hesychius ἐκτολυπεύσας: τελειώσας. Whether or not the gloss in Hesychius
comes from a scholion on Aowis 44 χαλεπὸν πόνον ἐκτολυπεύσας, the meaning
there is simply τελειώσας. From this it is clear that too ‘literal’ a translation
of ἐκτολυπεύσειν (e.g. Humboldt: ‘und nicht das Gespinnst zur gebührenden
Zeit zu entknäueln noch hoffend’, Wilamowitz : ‘was immer dem Krampf der
Brust sich entwinde’) is doubtful. The translator must be content with a
less picturesque rendering, even if he thereby sacrifices an image which is
attractive to the modern mind.! In any case we cannot be certain that for
Aeschylus and his audience this word, a “Ounpov γλῶττα, meant anything
more than ‘perform, complete, accomplish’.”
It follows from what has been said about ἐκτολυπεύειν that it is wrong to
adopt the view of the many modern commentators who make ζωπυρουμένας
φρενός dependent upon éxroA. as gen. separ., as e.g. Paley ‘to unravel any-
thing to the purpose from a mind that is bursting into a flame’, Wecklein,
Wilamowitz (see above), Headlam: ‘evolving . . . from a fevered brain’.
ζωπ. φρενός is rather a genitive absolute. This is how it is taken in the
scholion of Triclinius. The comma which in F and Tr is placed after exroAv-
πεύσειν is retained in the older editions. Following this punctuation, Stanley
translated correctly ‘et nihil sperans amplius tempestivum se perfecturum,
mente inflammata’, so, too, e.g. Franz, Conington, Schneidewin, Nägelsbach,
Sidgwick, Platt.
The vagueness of the last sentences of the stasimon (from 1025 on) is
probably intentional. The guarded language of the Chorus is in accordance
with its own situation, i.e. its position in regard to the events represented on
the stage, where the old men are placed between Clytemnestra and Agamem-
non, between the knowledge of ancient crimes and the uncertainty of new
crimes planned ; but it is also in accordance with the inevitable limitations of
the poet’s dramaturgy, owing to which the Chorus, whatever its anticipations
1 Verrall elaborates it with loving care: he perceives ‘the figure, homely but vivid, of a
woman with her wool. working in the winter against time, as we say, with no better light
than she gets by stirring her fire [you see, ζωπυρουμένας is not lost on Verrall]'. Qui Aeschylus
fuit, factus Tibullus est. .
2 Wilamowitz on Ar. Lys. 587 says: ‘roAvmevew occurs only in Homer [he did not need to
mention specially the Homerizing use as in the Rhesus etc.] and there the τολύπη is for-
gotten'. This is perhaps an exaggeration. What Wilamowitz seems to have had in mind
was better expressed by Matthew Arnold, On Translating Homer, 86: ‘Our knowledge of
Homer's Greek is hardly such as to enable us to pronounce quite confidently what is idio-
matic in his diction, and what is not ... but I seem to myself clearly to recognize an
idiomatic stamp in such expressions as τολυπεύειν πολέμους... and many others . . . roÀv-
πεύειν ἀργαλέους πολέμους seems to me to have just about the same degree of freedom as the
“jump the life to come”, or the “shuffle off this mortal coil” of Shakspeare'.
4872:2 uh 465
lines 1025 fl. COMMENTARY

and apprehensions, could never be permitted strongly to oppose inevitable

destiny. Here particularly, shortly before the catastrophe, it is impossible
to speak in any other way than ὑπὸ σκότωι of the terrible nature of fate and of
the feelings which this arouses in the hearts of Agamemnon’s loyal subjects.
All the same, the poet has taken good care to make clear what is, in the main,
the object of these last sentences. To see this we have but to pay attention
to the way in which the salient points of the first two stanzas are picked up
again: 1028 καρδία, οὐδὲν ἐπελπομένα κτλ. (cf. 993 f. οὐ τὸ πᾶν ἔχων ἐλπίδος
φίλον θράσος) and, significantly at the end, φρενός. The ὑπὸ σκότωι βρέμειν οἵ
the καρδία, also, is reminiscent of 991 ff. ὑμνωδεῖ... ἔσωθεν θυμός. The central
theme of the first pair of stanzas was: the progress of the guilt to its τέλος, the
requital, continues unabated; no success of the moment can hamper the
ordained course, no delay can silence the heart’s prescient cry of woe. Since
the chief elements of this train of thought are clearly indicated in the con-
clusion of the chorus, we may well venture to draw on the beginning and
thus add definition to the vagueness of the ending. I am well aware of the
uncertainty of such an attempt. The result is somewhat as follows. For the
guilty person to be pardoned his guilt, for the δίκη not to be brought to its
τέλος, would be μεῖζον ἢ κατ᾽ ἄνθρωπον and far beyond what is assigned to
man as his portion by his destiny fixed once and for all, his τεταγμένα μοῖρα.
Enrichment, πλέον φέρειν, over and above this portion cannot be expected
from the gods by anyone. Were it otherwise, the old men would now be
praying aloud for protection against the imminent evil, and might perhaps
even intervene themselves by telling what they know of the past and what
they fear for the future; νῦν 8é.... Whatever they may try, it is all in vain,
it must be in vain, for there is no longer any hope of achieving something
that at the right moment might change the course of events. Only covert
hints and murmurings are left to the heart.
The conclusion of the ode, then, is in harmony with its beginning, but the
last sentences also continue what has immediately gone before. The story of
the guilt and punishment of Asklepios is a warning illustration of the universal
law that makes it impossible for the μοῖρα of man πλέον ἐκ θεῶν φέρειν.

1035-9. At first sight it may appear tempting to take the subordinate clause
introduced by ἐπεί 1036 as supplying the reason for the imperative in 1039,
and so to put a comma after πέλας (1038) and a stronger mark of punctuation
after 1035. This is how the sentence was punctuated by Schütz (ed. nova),
Blomfield, and among more recent editors by Headlam (cf. also his prose
translation), Mazon, A. Y. Campbell, and also by Wilamowitz in his trans-
lation (but not in his edition, later). This undoubtedly gives a strong effect
to 1035, which then stands alone and very tersely expresses what Clytemnestra
wants toconvey. Moreover, an abrupt sentence of this kind would effectively
introduce the theme of the subsequent scene. Nevertheless I do not believe
that this was the intention of the poet, but rather follow Hermann, who
protested against the comma which Schütz and Blomfield put after πέλας,
‘quo et inconcinna redditur oratio et vis perit. Quum immota manet Cas-
sandra, denuo eam compellat Clytaemnestra his verbis, ἔκβαιν᾽ ἀπήνης τῆσδε"
μηδ᾽ ὑπερφρόνει.᾽ If we accept this, we have to assume a short pause before
1039; then, as Cassandra does not move, a fresh start. This grouping may be
466
 COMMENTARY line 1035

recommended, too, by the contents of the causal clause (1036 ff.), with its
reference to the favourable treatment she can expect in the house, as she will
be allowed to have a share in the family cult. This idea seems to be more
closely bound up with εἴσω κομέζου than with ἔκβαιν᾽ ἀπήνης τῆσδε. But above
all, the brusque manner of the curt order εἴσω κομίζου καὶ σύ, Κασσάνδραν
λέγω, if it were not mitigated by its continuation, would not be in accord
with the attitude which Clytemnestra assumes here. Condescending she
certainly is, but she begins by using persuasive, if not gentle, language; it is
not until later that she loses control of herself.
1035. εἴσω κομίζου : perhaps not very polite. There is definite rudeness in
Suppl. 949 κομίζου δ᾽ ὡς τάχιστ᾽ ἐξ ὀμμάτων, impatience in Prom. 392 στέλλου
κομίζου. Creon’s instruction to the guard in S. Ant. 444 σὺ κομίζοις ἂν σεαυτὸν
ἧι θέλεις ‘gives a contemptuous permission’ (Jebb). Not so rude is E. Phoen.
1636 (Creon) κόμιζε σαυτήν, Ἀντιγόνη, δόμων ἔσω, but it certainly comes rather
near to the limits of good manners. The argument which, in the following
lines, Clytemnestra advances to support her demand detracts somewhat
from the harshness of her first words.
Not until this point, at the beginning of the great scene of which Cassandra
is the centre, is her name heard (cf. on 950). Of course, we should not ask
when and how Clytemnestra learned her name. It is taken for granted that
people at home were sufficiently well informed about events in and before
Troy and about the leading characters, cf. 1083 f., 1098, 1439.
Κασσάνδραν. The spelling of the name with only one o is very common
in many MSS of both poets and prose-writers. In the ὑπόθεσις, the index
personarum, and the text of the Agamemnon (including the person-name in
the margin) M F Tr have Kaodvöpa, while in V (which does not go beyond
l. 348) the name is constantly spelt with cc. In the MSS of Euripides the
spelling with one σ is much commoner than the other; as early as Pap. Oxy.
877 (‘probably written in the third century’) we find Kaod[vôpar] (E. Hec.
1275). In Pindar P. 11. zo there is better authority for Kacdvöpav than for
Κασσάνδραν. On the other hand, in the Jhad the original oo has been pre-
served by all MSS in 2 699, and in N 366 by the Venetus A and others, while
the Venetus B and others have Kaodvöpnv. There can be no doubt that
Kaoo- is the only genuine form. The spelling with one o in inscriptions on
early vases! does not signify anything, for at that period the doubling of
consonants was often neglected (cf. Meisterhans 94). But Attic KATANAPA
provides conclusive proof, for only Kaoo-, not Kao-, could be the basis of
Kar(r)-. The instances are these: black-figured amphora in Berlin, no. 1698
Furtw. (Kretschmer, Vaseninschr. 178) : KL AJTANAPA ; plate by the Cerberus
painter (late 6th century) in Yale (Beazley, Attic Red-Fig. Vases in Am.
Museums, ı3f., and Att. Red-Fig. Vase-Painters, 55. 4): KATAAPA. For
the fourth century we have the evidence of the dedication on a bronze tablet
from Dodona (C. Carapanos, Dodone et ses ruines, pl. xxI1; cf. Wilamowitz,
Verskunst, 373): KAZZANAPAZ. Further confirmation may be derived from
the spelling of King Cassander’s name with oo (cf., e.g., Dittenberger, Orient.
1 To Kretschmer, Griech. Vaseninschriften 28, add now his observations in ANTI42PON,
Festschrift für J. Wackernagel (1923), 193 f.—I dare not use the inscription Kaooavöpa
(Kretschmer, Vaseninschr. 178) on the cup decorated by the Codrus painter, for the vase is
‘restored’ (Beazley, Att. Red-Fig. Vase-Painters, 740 no. 10).

467
line 1035 COMMENTARY
Gr. Inscr. no. 5 passim, 6. 5); on the coins, too, King KAZZANAPOXZ is
always spelt thus.’ In documents of the late third and early second cen-
tury B.C. several Kaooavöpo. are spelt thus (e.g. Tebtunis Papyri 820. 6 and
24, 827. 19, 1045. 46). To return to King Priam’s daughter. In good ancient
book-texts her name appears in the form which we should expect, cf. Ibycus
fr. 3. 12 D. (Pap. Oxy. 1790, 1st century B.c.), Κασσάνδραν. We need not
enter here into an etymological discussion. Many philologists seem to agree
that the first element of the name is connected with κέκασμαι, etc. (cf., e.g.,
Fick-Bechtel, Die griech. Personennamen, 160; Kretschmer, Glotia, xxiv,
1936, 247; O. Hoffmann, Glotta, xxviii, 1939, 52; E. Schwyzer, Griech.
Gramm. i. 442 n. 6), whereas v. Blumenthal, Glotta, xx, 1932, 285 ff., sees in
Kaoo- an Illyrian element (Ernst Fraenkel, RE xvi. 1632, disagrees with this
view). As far as the spelling is concerned, the fantastic etymology which we
read in Tzetzes’s introduction to Lycophron (vol. ii, p. 7. 4 Scheer) is of some
interest: Κασάνδρα δὲ λέγεται παρὰ τὸ Kdow ἀνδρεῖον ἔχειν τὸν Ἕκτορα. Cf.
also Trag. fr. adesp. 243 N. ὁμόπαιδα κάσιν Kaodvöpas. Is it conceivable
that this etymology accounts for the popularity in the Middle Ages of the
form Kacdvdpa, a form not unknown at a much earlier period??
1036. ἔθηκε. Stanley understood the verb in the sense of ponere and took it
with δόμοις (‘Quoniam te statuit Iuppiter in aedibus ut particeps esses’ eqs.),
similarly, among more recent commentators, e.g. Verrall ('since Zeus hath set
thee in a house, where thou mayst' etc.). Most scholars, however, rightly
take it as — facere, efficere. τιθέναι used in this sense governs an infinitive in
two other passages in Aeschylus: both of them are in this play (178, 1174),
and in both, as here, a god or δαίμων is the subject. δόμοις goes with κοινωνὸν
εἶναι.
It is an old point of contention whether ἀμηνίτως is to be taken with Ζεὺς
ἔθηκε or with κοινωνὸν εἶναι. Casaubon and Butler, rejecting the conjecture
ἀμηνίτοις (Auratus), stated that dumvirws should go with κοινωνὸν εἶναι.
Hermann, without reference to Butler, took the same view, and he was
followed by Klausen, Weil, Nägelsbach, Schneidewin-Hense, Wecklein, and
Plüss. On the other hand, Linwood, Conington, Hartung, Kennedy, Platt,
Verrall, Wilamowitz (transl.), Headlam, Passow-Crönert s.v. aunvıros, Mazon,
G. Thomson etc. connect the adverb with Ζεὺς ἔθηκε. It is of no use for the
support of the latter interpretation to observe, correct though it is, that
μῆνις is preferably ascribed to the gods (with regard to the usage of Aeschylus
cf. Latte, Nachr. Gött. Ges., Fachgruppe I, 1933, 23, and Körte, Hermes,
Ixviii, 1933, 259 n. 1. ; for the general usage see Frisk, Eranos, xliv, 1946, 29 ff.).
For it would in itself be quite possible to relate dumvires, ‘so that no μῆνις is
involved' (cf. on 412 ἀτίμους, for ἀμήνιτος cf. on 649), to the attitude not of
Zeus, but of the people in the house, with whom Cassandra is going to live;
this possibility is clear from Suppl. 975 σὺν τ᾽ εὐκλείαι καὶ ἀμηνίτωι Pager
λαῶν. Nevertheless I am inclined to connect ἀμηνίτως with Ζεὺς ἔθηκε,
without being able to give any definite reasons. The word-order, as far as I
can see, gives no help: for the difficulty of finding out the connexion of an
1 I owe this information to Dr. J. G. Milne, whom I consulted because Boeckh's remark
(critical note on Pind. P. r. 20, vol. i. 505) concerning the coins puzzled me.
2 In the epigrams Anth. Pal. 7. 327 and 328 (what date?) the forms Kácavbpov and
Kdoavdpe are required by the metre.
468
 COMMENTARY lines 1037 f.
adverbial phrase! placed between the governing verb and the infinitive cf.
on 221 and also Prom. 671 f. ἀλλ᾽ ἐπηνάγκαζέ νιν Διὸς χαλινὸς πρὸς βίαν πράσσειν
τάδε, where the earlier commentators, probably correctly, take πρὸς βίαν
with émqvéykate, whereas Wecklein (in detail in the Athens edition, 1896, of
his commentary) and Groeneboom make it equivalent to πρὸς βίαν φρενῶν
and take it with the words that follow. In Ag. 1578 f. φαίην ἂν ἤδη νῦν βροτῶν
τιμαόρους θεοὺς... ἐποπτεύειν there is no doubt that the νῦν goes with the
words which precede it, in S. 47. 752 f. Κάλχας . . . εἶπε κἀπέσκηψε παντοίαι
τέχνηι εἶρξαι... Αἴαντα the παντοίαι τέχνηι certainly goes with the words that
follow it (it is the same in the passages quoted by Jebb, ad loc., Hdt. 1. 112. 1,
Xen. Anab. 4. 5. 16). Both interpretations seem to provide a pregnant
thought, and the result, as far as Cassandra's lot is concerned, is similar in
both cases, whether Zeus prepared this fate for her, refraining from further
μῆνις (the destruction of Troy was sufficient), or whether she may live in the
house without μῆνις rearing its head against her. What and how God ordains
for man, and what man has to bear and how, are but different aspects of the
same thing. But this consideration should not lead to the conclusion that
Aeschylus left apsviros hanging vaguely in the air.? Its connexion will have
been to him unambiguous and the phrasing of the actor will have made it
clear to the audience, only we cannot discover it.
1036-8. What Clytemnestra here makes appear as a special favour is in truth
nothing more than the common practice of antiquity. Cf. Eduard Meyer,
Kl. Schr. i, 2nd ed., 185: ‘The slave . . . is a member of the household—oikerns
(‘Hausgenosse’) was the Greek term. Therefore he had his part in the cults
of the house, in the household worship, and while he was cut off from the
actual religious ceremonies of the State, he participated in those festivals
of the community and the people which had grown out of family cults, e.g.
the Anthesteria, i.e. the festival of the dead, at Athens, and in the Saturnalia
at Rome.? This situation was to be found throughout the whole of the ancient
world.’ Cf. also G. Glotz, La Solidarité de la famille, 163 n. 3 (with biblio-
graphy), and Wilamowitz, Glaube d. Hell. i. 288 n. 3. It is mentioned as an
exception that the old Kiron οὔτε δούλους προσῆγεν οὔτε ἐλευθέρους ὀθνείους
(Isaeus 8. 16) to his household worship of Zeus Ktesios.
1037. χερνίβων : for the ritual which introduced the actual sacrificial process
and which ‘begins with the wetting of the hands and ends with the sprinkling
of the holy-water’, cf. L. Ziehen, RE xviii. 601 1.
1037 f. πολλῶν μέτα! | δούλων: cf. Pers. 460 f. τοξικῆς τ᾽ ämo θώμιγγος, Eum.
τ In Ag. 1184 (q.v.) there is the difficulty of ascertaining where an adverb belongs which
stands between a governing verb and a dependent participle.
2 In modern print dashes are cheap; so they have been used here (Class. Philol. xxxvii,
1942, 265 f.) to mark the ‘cruel irony’ of Clytemnestra, which is a great favourite with the
present generation (cf. on 970 f.).
3 For the participation of the slaves at Rome in the cults of the house cf., e.g., the well-
known statement of the elder Cato in his speech ‘in L. Veturium de sacrificio commisso’
(Festus, p. 234 M. = Orat. Rom. fragm. i, p. 176 Malcovati) : domi cum auspicamus, honorem
me deum immortalium velim habuisse. servi, ancillae, si quis eorum sub centone crepuit quod
ego non sensi, nullum mihi vitium facit.
4 μέτα F Tr, Pers. 460 ἄπο all MSS, Eum. 114 πέρι M (not noted by Wecklein, hence the
mistake in Wilamowitz's apparatus) ; similarly, e.g.. E. Alc. 66 f. ἵππειον μέτα (so the MSS) |
ὄχημα. Even in antiquity it was uncertain whether ἀναστροφή, i.e. the throwing back of the
accent, took place in such cases, i.e. where the preposition was interposed between adjective

469
lines 1037 f. COMMENTARY
114 f. τῆς ἐμῆς πέρι ψυχῆς. For the use of μετά with the genitive in Aeschylus
cf. Tycho Mommsen, Beitr. z. der Lehre von den griech. Praepositionen, 605 f. ;
for the development of the meaning of μετά in Greek, see Wackernagel,
Syntax, ii. 240 ff., who, starting from the statement that ‘in the middle of’ is
the older meaning, goes on to say (p. 242): 'Staying in the midst of a company
includes being together with it and its members, and there are many passages
where the two meanings quite coincide, as in Homer . . . z 140 f. μετὰ δμώων
. . πῖνε kai ἦσθε "drank and ate in the midst of the serving-men" or “with
the serving-men" ' etc. This is also suitable at Ag. 1037 f.
1038. κτησίου βωμοῦ. Wilamowitz, ad loc., warns us against relating κτήσιος
(cf. on 1009) directly with Ζεὺς κτήσιος ; he refers to S. Trach. 690, where Jebb
paraphrases *belonging to the household' and similarly rejects direct allusion
to Zeus. But in the present passage it is at least natural to think of the altar
of Zeus Ktesios, especially as just before (1036) Zeus had been named as
allowing Cassandra the privilege of participation in the ceremony. The
passage is, in general, understood in this way: cf., in addition to the com-
mentaries, Preuner in Roscher's Lexikon, 1. 2627; Nilsson, Ath. Matt, xxxiii,
1908, 280; Sjövall, Zeus im altgriech. Hauskult (Lund 1931), 56.
1039. ἔκβαιν᾽ ἀπήνης τῆσδε. The parallelism with 906 is symbolic of the
common fate prepared by Clytemnestra for Cassandra and Agamemnon.
ὑπερφρονεῖν indicates the pride of a person who considers a suggestion
beneath his dignity (as in Hdt. 1. 199. 1), but it also means the attitude of
mind which attempts to set itself too high for what is fated, for the δαίμων
of the occasion, instead of acquiescing (Pers. 825 ὑπερφρονήσας τὸν παρόντα
δαίμονα). Probably both notions are implied here.
1040. kai . . . γάρ: these are the words used, from very ancient times, for
attaching the παράδειγμα to the sentence which is to be illustrated by it, a
sentence consisting often, as here, of an exhortation, e.g. T 95, (2 601 f. viv δὲ
μνησώμεθα δόρπου. καὶ γάρ τ᾽ ἠύκομος Νιόβη ἐμνήσατο σίτου «rA., Alcaeus fr.
73. 41. D. ἀλλ᾽ ἄγι μὴ μεγάλων ἐπ |o — VO]. καὶ γὰρ Σίσυφος Αἰολίδαις βασίλευς
krÀ., Pind. Ol. 7. 27, the Lesbian grinding song in Plut. Mor. 157 ἃ ἄλει,
pura, ἄλει" καὶ γὰρ Πιττακὸς ἄλει krÀ., Arist. Rhet. 2. 20. 3 p. 1393 a 32 (in the
section about the παραδείγματα) δεῖ πρὸς βασιλέα παρασκευάζεσθαι καὶ μὴ ἐᾶν
Αἴγυπτον χειρώσασθαι" καὶ γὰρ Δαρεῖος οὐ πρότερον διέβη κτλ. Frequently οὐδὲ
γάρ has the same function, cf. on 1022 ff.
φασὶν . . . more also belongs to the stereotyped forms of popular narrative:
it is to be found in the introduction of αἶνοι of all kinds, e.g. Sappho fr. 105 D.
φαῖσι δή ποτα Anôav krÀ., Alcaeus fr. 101 D. ws γὰρ δή ποτ᾽ Apıorodauov dato’
οὐκ ἀπάλαμνον ἐν Σπάρται λόγον εἴπην, Xenophanes fr. 7. 2 f. καὶ ποτέ μιν...
φασὶν ἐποικτῖραι καὶ τόδε φάσθαι ἔπος, Pind. P. 6. 21 τά ποτ᾽ ἐν οὔρεσι davri...
Φιλύρας vióv . . . παραινεῖν, Bacchyl. 5. 56, A. Suppl. 291 κληιδοῦχον Ἥρας
φασὶ δωμάτων more "Im γενέσθαι κτλ., Pl. Rep. 359 € οἵαν ποτέ φασιν δύναμιν
τῶι τοῦ Λυδοῦ προγόνωι γενέσθαι, Eubulus fr. 119. 5 (ii. 206 K.) ὃν φασί ποτε κτλ.,
Terence, Eun. (Menander) 584 f. Iovem quo pacto Danaae misisse aiunt
and noun, cf. Lehrs, Quaest. ep. 79 ff.; Chandler, Greek Accentuation, and ed., § g15 (too
dogmatic) ; Kühner-Blass, i. 334; Wackernagel, Idg. Forsch. xlii, Anzeiger, 55. But in any
case, this accentuation has nothing to do with the position at the end of a line; so Wila-
mowitz’s observations (in the apparatus) on Pers. 460 and Eum. 114 are erroneous. For a
few other cases of the MS μέτα cf. on 1271.

470
 COMMENTARY line 1041

quondam in gremium imbrem aureum, Sotades fr. 7 D. "Hpwv ποτέ φασιν dia
κτλ., Callim. Attia fr. 9. 4 Pfeiffer (Pap. Oxy. 1o11) "Hp γάρ κοτέ bac,
Catullus 64. 1 ff.
The two expressions which we have examined? are among those most usual
for the introduction of παραδείγματα or alvoı. The verb τλῆναι also (1041) had
an important place within a special type of paraenetic παραδείγματα: cf.
E 382 ff. rerdadı . . . πολλοὶ γὰρ δὴ τλῆμεν.... τλῆ μὲν Ἄρης, ὅτε... τλῆ δ᾽
Ἥρη, ὅτε... τλῆ δ᾽ Aïômns . . . εὖτε. . ., and, after this Homeric model, Pany-
assis, ᾿Ηράκλεια fr. 16 Kinkel τλῇ μὲν Δημήτηρ, τλῆ δὲ κλυτὸς Ἀμφιγυήεις, τλῆ
δὲ Ποσειδάων, τλῇ δ᾽ ἀργυρότοξος ΑἈπόλλων ἀνδρὶ παρὰ θνητῶι θητεύεμεν εἰς
ἐνιαυτόν κτλ. (compared with Ag. 1041 by Headlam and R. Oehler, Mythol.
Exempla, Diss. Basel 1925, 34),? S. Ant. 944 ff. ἔτλα καὶ Aavdas οὐράνιον φῶς
ἀλλάξαι δέμας κτλ.
It is difficult to make out whether the ancient audience felt as significant
Clytemnestra’s having recourse in this way, both in form and subject-
matter, to θρυλούμενα. When a woman of Clytemnestra’s superiority employs
in her speech devices of this kind, it is of course quite a different matter from
the use which in choric songs, in accordance with their generalizing mode, is
made of παραδείγματα, whether their purpose is confirmatory (e.g. Cho. 602-
22) or hortatory (e.g. S. Ant. 944 f£). What prompts Clytemnestra is hardly
disdain, in the sense that she is fobbing off old tales upon her victim. Perhaps
the passage is to be interpreted as meaning that the queen wishes to make
some friendly remark and conceal her true feelings, and as she has no genuine
comfort to offer, resorts to the conventional phrases which are always ready
at hand. The audience may feel that the exhortation is somewhat stale.
However, I would certainly not like to exclude the possibility that what is
intended here is a suitable reference to a dignified heroic example.
1041. On the whole one is inclined, and for very good reasons, to assume
that, when the readings of F and Tr differ considerably, the version in Tr is
to a certain extent due to the interference of Triclinius and F is more faithful
to the παράδοσις. Consequently attempts have frequently been made to
emend the reading of F δουλείας μάζης βία by a slight alteration, such as
δουλίας μάζης βίον (Blomfield), 8. μ. βίαν (Thiersch), δ. μ. βίαι (Weil and others;
this, the addition of the ı of the dative, is practically no alteration at all). It
is obvious that the first two of these proposals are extremely lame, but
δουλίας μάζης βίαι seems more attractive. It would imply that πραϑέντα
τλῆναι go together, but this is unobjectionable, for although in Aeschylus
τλῆναι usually takes an infinitive or an object in the accusative, the construc-
tion with the participle, to be found in Homer (e 362), Sophocles, etc., occurs
1 A pretty adaptation of the homely ποτέ φασιν to the style of elevated lyrics is found
E. El. 700 f., where the Chorus begin their tale of the golden lamb: . . . ποτὲ κληδὼν ev
πολιαῖσι μένει φάμαις κτλ.
2 They are found both together, as here, e.g. in Catullus 64. 212 namque (καὶ γάρ) ferunt
olim (so, too, Propertius 1. 20. 17), cf. also Catullus 64. 76 nam perhibent olim; the line of
Callimachus quoted just before points to Hellenistic poetry as the region from which such
formulae are likely to have come into Latin poetry. Cf. also Norden on Virgil, Aen. 6. 14.
3 To regard these lines, as G. Thomson does, as the ‘source’ of the passage in the
Agamemnon, is not only superfluous, in view of what has been demonstrated above, but
also very doubtful in itself: the Herakles-epic of Panyassis, a contemporary of Aeschylus,
had, as far as we can see, no influence at all on Athens and its literature (cf. Wilamowitz,
Eur. Her. i, and ed., 67, and also his Hellenist. Dichtung, i. 100).

471
line 1041 ᾿ς COMMENTARY
in Sept. 754 ff. However, the sense of δουλίας μάξης βίαι is not satisfactory.
This is best brought out by the paraphrase of its defender Verrall, ‘in spite
of the slaves’ porridge’. Against this G. Thomson observes rightly that with
βίαι too much emphasis is laid on the coarse diet in comparison with the other
hardships of slavery. Since F yields no realiy convincing text, it is natural
that several editors turn to Tr for assistance. Among them is Wilamowitz,
but his note seems to show that he did not feel quite happy about his decision:
‘quamquam βία in archetypo fuit, θιγεῖν in interpolata tantum memoria
traditur, nihil hoc aptius excogitatum est, nec sapit Triclinium.'! His reading,
πραθέντα τλῆναι δουλίας μάζης θιγεῖν, was anticipated by Keck. If this reading
were found in F, it would probably be accepted without hesitation. δουλίας
μάζης sums up the position of the slave most suggestively ; for this typical
feature cf., e.g., Archil. fr. 79. 6 D. δούλιον ἄρτον ἔδων, Hipponax fr. 39. 6 D.
δούλιον χόρτον, E. Alc. 2, etc. ; in any society the man who is no longer his own
master has to experience ‘come sa di sale lo pane altrui. The expression
θιγεῖν, too, seems very apposite ; ἅπτεσθαι σίτου and the like is common from
Homer on. But are we at liberty to pick and choose in such a manner, i.e.
to take the word δουλίας that suits us from F and θιγεῖν from Tr, dropping in
either case what we like less? This is doubtless an 'eclectic' procedure as
Hense called it when, in the appendix to Schneidewin's edition, he protested
against Keck's reading. Keck, however, attempted to justify his method : he
supposed that Fand Tr ‘drew from a common source, in which several glosses
were intermingled with the text'. This, though not very lucidly expressed,
contains perhaps an inkling of the truth. In my Prolegomena (vol. i, p. 29 f.)
sufficient evidence has been collected to make it highly probable that in the
margin of the hyparchetype of FTr variants were entered in fairly large
numbers. It would therefore not be difficult to suppose that in the hyparche-
type δουλίας μάζης βία (or βίᾳ) was written in the text and δουλείας ζυγῶν
θιγεῖν added, perhaps with yp., and that the reading of Tr is a compromise
between the former and the latter. Possibly those two readings were them-
selves the product of a contamination, and behind them lay one version
δουλίας μάζης θιγεῖν and another δουλείας ζυγῶν βίαι. But this sort of guess-
work is very unsafe indeed. The possibility must be admitted that what we
read in Tr represents one of Triclinius' wilder experiments.?
1042. εἰ δ᾽ οὖν κτλ. The train of thought, which is not immediately clear, is
intelligently explained in the scholion : καλὸν μέν, φησί, μὴ πειραθῆναι δουλείας,
εἰ δὲ πειρῶιτό τις, κάλλιον ἀρχαιοπλούτοις δουλεύειν καὶ μὴ νεοπλούτοις. In cases
where ‘a speaker hypothetically grants a supposition which he denies, doubts
or reprobates’, the conditional clause is often introduced by ei δ᾽ οὖν, ‘but if,
in reality’ (Denniston, Particles, 464 f.; cf. also Neil on Ar. Knights 423 f.).
1 Cf. the similar comment of Weil (Preface to his Teubner edition, p. ix): ‘Nec tamen
omnia quae hic liber [Tr] habet peculiaria a Triclinio mutata esse puto: movet me potissi-
mum Agamemnonis versus 1041, quo in loco maior est lectionis diversitas quam quae
grammatici licentiae tribui posse videatur.’ It is perhaps worth noting that Triclinius
wrote above @iyew the gloss ψαύειν.
2 Cf. for δουλείας ζυγά S. Aj. 944 (cf. Soph. fr. 532 N. = 591 P. 1. 5); δούλιον ζυγόν occurs
several times in Aeschylus. .
3 A. Kirchhoff, Philol. ix, 1854, 163, thought that Triclinius, starting from the text as we
have it in F and remembering E. Hipp. 885 f. θιγεῖν βίαι, concocted the phrase καὶ ζυγῶν
Biyew βίαι. Cf. also Mazon’s edition, ii, 2nd ed., p. xxiv n. 1.

472
 COMMENTARY lines 1045 f.
1043. Arist. Rhet. 2. 16 p. 1391 a 15 διαφέρει δὲ τοῖς νεωστὶ κεκτημένοις καὶ τοῖς
πάλαι τὰ ἤθη τῶι ἅπαντα μᾶλλον καὶ φαυλότερα τὰ κακὰ ἔχειν τοὺς νεοπλούτους"
ὥσπερ γὰρ ἀπαιδευσία πλούτου ἐστὲ τὸ νεόπλουτον εἶναι (compared by Stanley).
The unpopularity of the νεόπλουτοι (a fine specimen of the class is ridiculed
by Anacreon fr. 54 D.) in the latter part of the fifth century B.c. is attested
by the so-called Anonymus Iamblichi,z. 8 (Diels-Kranz, Vorsokr. ii, sth ed.,
401) ἅμα δέ τις καὶ τῆι ἐξ ὀλίγου χρόνου εὐδοξίαι προσγίγνεται βλάβη τοιάδε" τοὺς
γὰρ ἐξαπιναίως καὶ ἐξ ὀλίγου χρόνου ἢ πλουσίους ἢ σοφοὺς... γενομένους οὐκ
ἀποδέχονται ἡδέως oí ἄνθρωποι, Cratinus fr. 208. 2 K. ἀνδρῶν νεοπλουτοπονήρων,
Ar. Wasps 1309 ἔοικας, ὦ πρεσβῦτα, νεοπλούτωι Φρυγί (Kock, τρυγί the MSS),
cf. also Denniston on E. El. 253, where E. Suppl. 741 ff. ὁ δ᾽ αὖ τότ᾽ εὐτυχής,
λαβὼν πένης ὡς ἀρτίπλουτα χρήματα, ὕβριζε might have been quoted. The
word ἀρχαιόπλουτος, of which S. El. 1393 was previously our second oldest
example, has now turned up in the ΤΠλοῦτοι of Cratinus (fr. 2. 13 Mazon,
D. L. Page, Greek Lit. Papyri, i. 200; cf. A. Korte, Archiv f. Papyrusforsch.
xi, 1935, 261 f., R. Goossens, Rev. des Etudes anc. xxxvii, 1935, 412, 426; it is
missing in the Addenda to L-S).
πολλὴ χάρις: a strong expression, cf. 550.
1045 f. The conclusion of the speech has been badly treated by the editors.
Everything turns on the correct understanding of ἔχεις. It is in fact quite
simple, and quite unambiguous, and yet, here as elsewhere, we see how a
seemingly brilliant suggestion, once it has gained currency, may persistently
darken judgement. The old conjecture ἕξεις (Auratus) is still the cause of
mischief, whether it is openly incorporated in the text (e.g. by Wecklein in
his edition of the Oresteia, Sidgwick, Blaydes, Gow, C.Q. viii, 1914, 4, A. Y.
Campbell), or modestly concealee in the apparatus criticus (Murray), or,
worst of all, where the editor keeps ἔχεις but tries to compromise with ἕξεις
(e.g. Headlam: ‘expect from us our usage customary’, Mazon: ‘de nous tu
peux attendre les égards coutumiers’, G. Thomson: ‘what is customary shall
here be yours’). Verrall, too, (‘from us thou art receiving . . .’), Wilamowitz
(‘allein bei uns hast du es, wie wir’s halten’), and others do only apparent
justice to the present tense of the Greek text. How much neater and
linguistically sounder is Schütz: “ἔχεις. Si vera est lectio, regina haec dicit:
Habes a nobis, quae apud nos fieri solent, h.e. haec tibi dicenda habui . . . Sin
ἕξεις ab Aeschylo scriptum fuerit, haec exsistet sententia: Habebis omnia,
quae tibi convenient.’ The correct translation was given long since by
Stanley: 'habes a nobis (dicta sc.)'. In his interpretation of ἔχεις he was
followed by Hermann, but the latter took zap’ ἡμῶν in a different way
(see below). [Stanley followed Pearson; see vol. i, p. 82.]
The meaning of ἔχεις in the concluding sentence of a speech is clear in
general from 582 and the note there; cf. further the last sentence of Lysias'
speech (12) against Eratosthenes ἀκηκόατε, ἑοράκατε, πεπόνθατε, ἔχετε"
δικάζετε and Arist. Rhet. 3. 19 (end of the book) τελευτὴ δὲ τῆς λέξεως ἁἁρμόττει
ἡ ἀσύνδετος, ὅπως ἐπίλογος ἀλλὰ μὴ λόγοςἦι εἴρηκα, ἀκηκόατε, ἔχετε" kpivare'.
But there are parallels, too, for the special type found in the present passage.
In E. Phoen. 953 Tiresias concludes his address to Creon with the words: ra
μὲν παρ᾽ ἡμῶν πάντ᾽ ἔχεις. A special piece of good fortune enables us to pro-
duce an example of this concluding formula direct from the practice of con-
temporary public speech: at the end of his defence of Palamedes, Gorgias
473
lines 1045 f. COMMENTARY

says (fr. 11a 37, Diels-Kranz) εἴρηται ra παρ᾽ ἐμοῦ, καὶ παύομαι. So once
more (cf. on 830 and 950) it is clear that Aeschylus borrows formulae of transi-
tion and conclusion from the practice of public speeches. The same turn of
expression with a slight variation occurs in A. Suppl. 271 ἔχουσα δ᾽ ἤδη ram’
ἐμοῦ τεκμήρια, and at the end of the speech of the Pythia in E. Ion 1367 f.
ἐξ ἡμῶν δ᾽ ἔχεις ἅπαντα Φοίβου re: ‘everything which I and Apollo (for whom
I speak) can tell you, you have now been told (the rest you must find out for
yourself)’. This frees us from the artificiality of Hermann, who put a comma
after ἔχεις and made παρ᾽ ἡμῶν dependent upon vopileraı; Meineke went a
step further still (ἔχεις, παρ᾽ ἡμῖν κτλ). When Clytemnestra says ‘you have
heard from us what kind of practice is customary’, we may perhaps complete
it by adding ‘in our house’. But it would be sufficient to understand it quite
generally (1.6. embracing the two opposite types of homes described before) :
‘what is customary (in the treatment of slaves)’. For this use of οἷάπερ cf.
S. Oed. C. 896 οἷάπερ πέπονθ᾽ ἀκήκοας.
1045. It is quite improbable that παραστάθμων in F is anything else but
παρὰ στάθμην slightly disguised. Here, too, many commentators have un-
necessarily obscured the understanding of the words, imposed upon by the
gloss παρὰ τὸ πρέπον in Tr and by Stanley, who translates ‘praeter normam’.
The same rendering is to be found, e.g., in Dindorf’s Lex. Aesch. (Linwood
s.v. στάθμη even says ‘irregular, violent in conduct’.) This interpretation is
responsible for the statement ‘there are two senses of the phrase παρὰ
στάθμην᾽ (Paley on E. Ion 1514 [this passage, in which στάθμην is qualified,
has, by the way, very little to do with the set phrase παρὰ στάθμην])) and in
more recent times has led to such perverse translations from excellent
Hellenists as ‘inequitable’ (Headlam) and stranger still ‘gnädig über's Mass’
(Wilamowitz). Other commentators’ have not let themselves be led astray
but have followed the established usage, which is particularly binding in such
semi-proverbial phrases. Hartung and Keck, for example, have energetically
protested against the sloppiness of the traditional explanation, and they have
been followed by Hense, Wecklein, and G. Thomson.’ In point of fact there
can be no doubt about the meaning of παρὰ στάθμην, ‘ad amussim, according
to measure, quite exactly’; cf. Theogn. 543, 945, Soph. fr. 433 N. (= 474 P.).
5 and Pearson’s comment on it. In Ag. 1045, where domestic matters are
concerned, παρὰ στάθμην corresponds exactly to Gretchen’s phrase in Goethe’s
Faust: ‘Und meine Mutter ist in allen Stiicken so accurat’ (although she does
not need to be, as she is quite comfortably off) ; there is a similar phrase earlier
in the play: ‘Die Mutter ist gar zu genau’.* Elsewhere παρὰ στάθμην is used
adverbially. That it should be taken predicatively here is in itself hardly

1 Quoted by Valckenaer on Phoen. 961. An echo of similar formulae of conclusion may

perhaps be discerned in the concluding sentence of Agathon’s speech, Plat. Symp. 197 e:
οὗτος, ἔφη, 6 παρ᾽ ἐμοῦ λόγος, ὦ Φαῖδρε, τῶι θεῶι ἀνακείσθω.
2 In Passow’s dictionary s.v. στάθμη the correct meaning is given alternatively with the
incorrect for Ag. 1045; in L-S s.v. παρά C. II. 2 the passage is sensibly dealt with.
3 F. H. Rau, too, in his careful monograph on παρά, Curtius Studien, ili, 1870, 85 n. 27,
rightly understood the meaning of παρὰ στάθμην. So did R. Enger, Emendationes Aeschyleae
(Jahresbericht des Gymnasiums zu Ostrowo 1861), το, but he misinterpreted Ag. 1045 and
therefore conjectured κοὐ π. or.
4 Beazley recalls the grandmother of Ausonius, who (Parent. 5. 7f., p. 33 Peiper) non
deliciis ignoscere prompta pudendis ad perpendiculum seque suosque habuit.
474
 COMMENTARY line 1048

credible, and moreover it would be very harsh to make ὠμοί and παρὰ στάθμην
parallel.! We require a verb to go with παρὰ στάθμην. This makes a lacuna
after 1045 probable. And this assumption is supported by what has above
been said about 1046. The sentence ‘So now you have heard from me what is
customary’ does not connect up with the preceding words, for it suggests
that she has explained previously, in a few words at least, how things are
done in her house. The connexion must have been something like this: ‘they
are hard to their slaves in everything, and exactly according to measure (are
the rations apportioned. But we treat our servants in such and such a way).
So now you have heard’, etc. It is impossible to say how many lines have
been lost. After I had put a lacuna after 1045 I noticed that Hartung had
anticipated me; his judgement on 1045 is good, but he misunderstood 1046,
where he seized the occasion to bicker with Hermann once more.
With ὠμοὶ δούλοις Paley compares E. Hec. 359 f. (Polyxene) ἔπειτ᾽ ἴσως dv
δεσποτῶν ὠμῶν φρένας τύχοιμ᾽ ἄν (the situation is depicted in further detail
in what follows). For its opposite see 951 τὸν κρατοῦντα μαλθακῶς, in a
similar context.
1046. ἔχεις map’ ἡμῶν κτλ. The ‘epilogue’ is attached without a connecting
word, a good example of what Aristotle, in the concluding sentence of the
Rhetoric, quoted above (on ἔχεις), considers suitable.
1047. σοί τοι λέγουσα: the underlying colloquial phrases? (‘do you hear?
she is talking to you; now she has finished’) become clear as soon as we
replace the participle by the finite verb: coi λέγω or λέγει (Ar. Plut. 926) is
addressed to a person who does not react to what has been said, who does not
pay attention or pretends not to do so (Donatus on Ter. Hec. 523 HEUS TIBI
DICO ostendit M yrrinam avertentem se, quod nihil inveniat quod dicat de filia) ;
further passages are quoted by Headlam on Herodas 4. 42.
σοί Tot λέγουσα παύεται contains a complete thought in itself, or rather
two (cf. the paraphrase given above), and σαφῆ λόγον is added as a kind of
characterizing afterthought, cf. on 616. The necessity of taking it in this way
is clear from the fact that the set phrase oot λέγει (see above) does not really
admit of the addition of an object.
παύεται: cf. the conclusion of the speech of Gorgias (above on 1045 f.)
εἴρηται... kai παύομαι, Demosth. 4. 13 (at the end of a section) ὡς μὲν οὖν
δεῖ... παύομαι λέγων.
1048. ἐντὸς δ᾽ ἂν οὖσα is the reading of the MSS. Hermann, whose judge-
ment in the case of the particle ἄν carries particular weight, saw that if the
first dv is retained, ἐντὸς οὖσα also must necessarily be hypothetical in mean-
ing, cf. Goodwin, $ 224. An examination of the relevant examples in Kühner-
Gerth, i. 247, also confirms Hermann’s statement.” An hypothetical sense,

! G. Thomson, too, noted this. He was on the right track (‘perhaps a line containing a
finite verb has dropped out’), but then turns at once to consider Verrall’s παράσταθμοι.
2 Casaubon showed a fine ear for the colloquial tone of the sentence and the emphasis
laid on the pronoun: he wrote on the margin of his Stephanus text ‘c’est a vous qu’elle
vient de parler’, one of the very rare entries in French.
3 I have also looked through Wackernagel's copious collections of passages, Idg. Forsch.
i, 1891, 394 ff. Among them are the following examples of a dependent participle preceding
the main verb with a repetition of the dv: A. Suppl. 227 f., Ag. 340, Cho. 349 ff. ; S. 4j. 155 f.,
1058 f., Oed. R. 446; E. Alc. 72, Andr. 934 f., Suppl. 417 f., Tro. 1244, Hel. 948 f., rorr f.;
Hat. 2. 26. 2 (two examples); Thuc. 6. 18. 2; Xen. Anab. 2. 5. 20 (the series of conditional

475
line 1048 | COMMENTARY
however, (‘if you were caught’), is impossible here." Therefore Hermann read
ἐκτὸς δ᾽ ἂν οὖσα and was followed by Meineke (Philol. xix, 1863, 203), who
translates: ‘if you were not a captive, you would perhaps follow her, and
perhaps not.’ It is hardly possible to misunderstand more completely the
thought and tone (see below) of the passage. The probably correct and quite
simple emendation ἁλοῦσα occurred to C. G. Haupt (he did not stand by it
but defended the MS reading with empty phrases) and was adopted by
Schneidewin and others. With this text the kolon ἐντὸς δ᾽ ἁλοῦσα μορσίμων
dypevuarwv does not come inside the hypothetical (or potential) context,
which only begins at πείθοι᾽ ἄν. For the phrase ἐντὸς ἀγρευμάτων ἁλοῦσα cf.
Hdt. 6. 133. 2 ἐπολιόρκεε [Tapious κατειλημένους ἐντὸς τείχεος, E. Phoen. 263 f.
μή με δικτύων ἔσω λαβόντες οὐκ ἐκφρῶσι.
μορσίμων ἀγρευμάτων: of course, the leader of the Chorus has in mind
only the means of catching employed by fate against her as against Troy,
but the audience is quite at liberty to understand μόρσιμος here in the sense of
its familiar associations (cf. on 157), 1.6. to think of a deadly net in which
Cassandra is now entangled.
1049. What we expect is ‘now that you are captured, you must comply!’
But this is toned down to ‘Comply, please, if you do comply; but perhaps
you will not’. There is certainly politeness in this, but perhaps also a note
of true sympathy. Clytemnestra was not brutal, but she did say κομίζου,
ἔκβαινε. The old man avoids even the slightest appearance of wishing to give
a direct order to the unhappy princess. For the exactly similar form of
politeness used by Clytemnestra in 1394 see there. The politeness of the old
man finds expression first in the fact that he says πείθοι᾽ dv instead of an
imperative, according to a well-known usage? (cf. Kühner-Gerth, i. 233 f.;
Karl Lammermann, Von der attischen Urbanität, Diss. Göttingen 1935, 73 ff. ;
Zilliacus, Eranos, xliv, 1946, 266 ff.), and secondly in his addition of the con-
ditional clause. ‘It is worth noting that in polite conversation in Plato
people appear to be anxious to tone down a command expressed in the
imperative and to weaken it by a qualifying formula which leaves the
carrying out of the command to the convenience of the other person: .. . εἰ
μή τίς σοι ἀσχολία τυγχάνει oaa . . . εἰ οὖν Ti σοι μέλει krÀ. . . . ἐὰν dpa Kai σοὶ
συνδοκῆι ἅπερ éuoi. . .. The limitation εἰ βούλει occurs extraordinarily fre-
quently with the imperative’ (Lammermann, 72).
ἀπειθοίης δ᾽ ἴσως. ‘Recte omittitur ἄν in altero membro orationis, quod ita
comparatum est, ut pro parte eius sententiae, cui additum est av, haberi possit’,
Hermann, Opusc. iii. 190. Blomfield compared S. Oed. R. 937 ἤδοιο μέν, πῶς
δ᾽ οὐκ ἄν ; ἀσχάλλοις δ᾽ ἴσως ; for further discussion see Kühner-Gerth, i. 248 f.
With regard to the delivery of the line, one would like to think that the
suppositions begins at 2. 5. 17), 5. 6. 32, Cyrup. 1. 3. 11; Pl. Symp. 176 c, Sophist. 223 Ὁ;
Demosth. 4. 1 ; [Demosth.] 46. 13. In all these cases the participle shares in the hypothetical
sense, marked by ἄν, of the main verb. I have not found any example to the contrary.
1 T, for one, cannot make anything of an excuse such as that advanced by Plüss: “ἂν
οὖσα: causal, the reason being indicated tactfully as only an imaginary one.’
2 The polite form of order is to be found also in Eum. 1026 ἐξέκοιτ᾽ ἄν ; this was not under-
stood by Wecklein and Verrall; it is correctly translated by, e.g., Otfried Müller and Paley,
and correctly explained by Blass and Wilamowitz, Interpr. 228. A good parallel for this
usage in the third person is to be seen in the language of the laws in the older inscriptions of
Elis, cf. R. Meister, Griech. Dialekte, ii. 71 ; Kühner-Gerth, i. 234; Bechtel, Griech. Dial. 11.859.

476
 COMMENTARY line 1052

coryphaeus is watching Cassandra as he speaks, and pauses a moment after ei

πείθοιο, and then as no utterance or gesture results, continues : ἀπειθοίης δ᾽ ἴσως.
1050 f. Passages for comparison have long since been brought forward:
Hesychius (presumably from a scholion on Ag. 1050) χελιδόνος diknv"! τοὺς
βαρβάρους χελιδόσιν ἀπεικάζουσιν διὰ τὴν ἀσύνετον (Wilamowitz’s emendation
of ἀσύνθετον, the correct ἀσύνετα is to be found in schol. Ar. Frogs 93 =
Hesych., Suidas s.v. χελιδόνων μουσεῖον) λαλιάν, schol. Ar. Birds 1680 καὶ
Αἰσχύλος (fr. 450 N.) τὸ BapBapilew 'χελιδονίζειν᾽ φησί, καὶ "Iov ἐν ᾿Ομφάληι
(fr. 76 v. Bl.) τοὺς βαρβάρους χελιδόνας ἀρσενικῶς φησί, ὡς ᾿Ηρωδιανὸς κτλ.,
Hdt. 2. 57. 1 f. πελειάδες δέ μοι δοκέουσι κληθῆναι πρὸς Δωδωναίων ἐπὶ τοῦδε αἱ
γυναῖκες, διότι βάρβαροι ἦσαν, ἐδόκεον δέ σφι ὁμοίως ὄρνισι φθέγγεσθαι"... ἕως
δὲ ἐβαρβάριζε (ἡ γυνή), ὄρνιθος τρόπον ἐδόκεέ apr φθέγγεσθαι,2 cf. also S. Ant.
1001 f., where ἀγνώς (ἀγνῶτα φθόγγον) means ‘unknowable’ just as in Ag. 1051.
This description of the φωνή of the barbarians is not so radical as making
ἄγλωσσος (5. Trach. 1060) actually synonymous with βάρβαρος, where the
point of view is equally obvious (‘The Russian masses are incapable of dis-
tinguishing between any Western foreigners, whom they group together as
"the dumb ones"—that is, those who cannot speak Russian’, Moscow Corre-
spondent of the Daily Telegraph, 8 May 1939).
1052. The main difficulty of the lines was noticed by Stanley: 'nescio an
(non) satis ex usu πείθειν vel λέγειν ἔσω φρενῶν ' (published by Butler from
Stanley'snotes). Hermann explains: ‘coniungendum est λέγουσα λόγωι: dicendo
ei persuadeo intus in animo', and up to most recent times some editors have
remained content with this or a similar interpretation, e.g. ‘my words must
speak within her comprehension and persuade her to consent' (Headlam).
But the two passages which Hermann quotes, viz. S. Phil. 1325 γράφου
φρενῶν ἔσω and E. Med. 316f. λέγεις ἀκοῦσαι μαλθάκ᾽, ἀλλ᾽ ἔσω φρενῶν"
Oppwöia μοι μή τι βουλεύσηις κακόν, may not even conditionally (Hermann
says ‘comparari quodammodo possunt’) be reckoned as parallels (so far
Schoemann, Opusc. iii. 146, was right), for there the φρένες inside which the
action takes place are the φρένες of the person who performs the action,
whereas here Clytemnestra would be using her persuasive powers inside the
φρένες of Cassandra, without a possessive adjective to make this distinction
clear. I cannot believe that a phrase such as ‘I persuade her inside her
φρένες ' or the like could be an equivalent for the Homeric formula τοῦ (or
+a) θυμὸν ἐνὶ στήθεσσιν ἔπειθεν. Stanley attempted to evade the difficulty by
taking ἔσω φρενῶν with the words of the preceding line (similarly Schiitz).
Butler took λέγουσα as well with the preceding line, and this was adopted by
Pliiss (1895): his explanation and the justification of his embellishment
πειθώ (sic) vw λέγω can be read in his commentary. None of the numerous
conjectures in place of λέγουσα is convincing.* I am not able to define clearly
ı This lemma has been inserted here by M. Schmidt, and is probably correct. ‘The MS
has χελιδόσι, which does not accord with alphabetical sequence.
2 ‘The Kurds ... speak a tongue mocked by the Turks as a twittering of birds’ (David
G. Hogarth, A Wandering Scholar in the Levant, 107).
3 Correctly taken with βουλεύσηις by Paley and most of the commentators.
* I should like to issue a warning against θιγοῦσα, which occurred to me in view of Ag.
432 and E. Alc, 108 ἔθιγες ψυχᾶς, ἔθιγες δὲ φρενῶν and which I subsequently found in Wila-
mowitz’s appendix to his translation Griech. Trag. ii. 117): it goes no more suitably with
ἔσω φρενῶν than most of the other words proposed.

477
line 1052 COMMENTARY
the seat of the corruption, nor can I agree unconditionally with Wilamowitz’s
verdict ‘aut λέγουσα aut λόγωι corruptum'. It seems conceivable, e.g., that a
sentence such as φιλοφρόνως λέγουσα πείθω νιν λόγωι would be linguistically
correct. πείθω (‘dynamic’ present? cf., e.g., Hdt. 5. 104. 3 καὶ ἀνέπειθε πάντας
Κυπρίους συναπίστασθαι" τοὺς μὲν δὴ ἄλλους ἀνέπεισε, Auabovoious δὲ οὐ βουλο-
μένους of πείθεσθαι κτλ., Ar. Wasps 115 ff. καὶ πρῶτα μὲν λόγοισι παραμυθού-
μενος ἀνέπειθεν αὐτὸν... ὁ δ᾽ οὐκ ἐπείθετο) νιν λόγωι inspires greater confidence
than the text of the first half of the line, but this impression may be deceptive.
1053. τὰ λῶιστα τῶν παρεστώτων. Cf. (Blomfield) Prom. 216 f. κράτιστα δή
μοι τῶν παρεστώτων τότε ἐφαίνετ᾽ εἶναι κτλ. and (van Heusde) Ar. Knights 30
κράτιστα τοίνυν τῶν παρόντων ἐστὶ νῶιν κτλ. and also (idem) Pl. Symp. 193 c
εἰ δὲ τοῦτο (the complete fulfilment of Eros just described) ἄριστον, ἀναγκαῖον
καὶ τῶν νῦν παρόντων τὸ τούτου ἐγγυτάτω ἄριστον εἶναι" τοῦτο δ᾽ ἐστὶ krÀ. This is
then a set phrase to designate something which, although it is not absolutely
the best, as in ὑγιαίνειν μὲν ἄριστον ἀνδρὶ θνητῶι and in the sayings quoted on 900
(κάλλιστον τὸ δικαιότατον and thelike), yet is relatively the best in circumstances
where perfection is not attainable. The speaker gives a kindly hint: ‘yes, if
only you were free—but, as things are, her instruction is the best ; follow her!’
1054. Blomfield’s πιθοῦ has assumed the position of a textus recepius. Of
course this can hardly be called a change of the MS reading (the MSS have,
e.g., at Ag. 206 the unmetrical πείθεσθαι [πειθέσθαι ΜΊ and at Prom. 274 in
both places the incorrect πείθεσθε), but we should remember that no argu-
ment has been brought forward for the necessity of the alteration. Hermann
on S. El. τοῖς protested against the substitution of πιθοῦ for πείθου proposed
by editors both there and in other passages (Meineke on Eupolis fr. inc. 1. 7,
Fragm. Com. Graec. ii. 547, agreed with him). It is true, the difference in
meaning which Hermann wished to establish (‘fod est oboedt, quod est
statim mutari sententiam et fieri quod iubeat volentis: πείθου autem, sine
libi persuaderi’) is probably untenable.? But he rightly points out that in
Oed. C. 520 the metre (cf. the antistrophe) shows πείθου to be correct; it is
immediately preceded there by στέρξον. The form πείθου is certain also in
Menander fr. 929 Kock Κρωβύληι τῆι μητρὶ πείθου καὶ γάμει τὴν συγγενῆ (it is
immaterial whether the metre is trochaic or iambic), whereas in Eupolis
fr. 357. 7 K. ἀλλ᾽ ἐμοὶ πείθεσθε the metre is not decisive; Stobaeus reads
πείθεσθε, but here we depend on scanty MS evidence (cf. Hense’s note, vol. iii,
p. 226. 13). In our present passage πιθοῦ is possible but not necessary. πείθου
λιποῦσα: ‘leave your seat here in the wagon and then do as she says’ (i.e.
εἴσω κομίζου). The aorist indicates a previous action. The same relation in
time is to be found elsewhere where the aorist participle is used with a
present imperative, e.g. A. Suppl. 395 f., 955, Pers. 833 f., Cho. sor f. (incor-
rectly punctuated in the Mediceus, by the earlier editors, and later by Sidg-
wick and Murray, whereas Conington, ad loc., is correct), S. Aj. 685 f., etc.
1055 f. οὔτοι... τρίβειν. Everyone can see at once what the general sense
of this sentence must be. But it is more difficult to be clear about its con-

1 Cobet, Novae lectiones, 410, says about the MSS of the Attic prose-writers: ‘Nihil
frequentius est quam ut ἐπειθόμην scribatur ubi aoristus requiritur.
? Jebb (on S. El. 1015) has taken over Hermann's artificial distinction, but in El. 1207,
Trach. 470, 1228 he reads πιθοῦ against the MSS and in disagreement with Hermann.
Pearson does the same.

478
 COMMENTARY lines 1055 f.

struction in detail; some alteration of the text seems to be required. Of the

proposals which have been put forward we can at once dispense with Dobree's
σχολήν, adopted by Weil, Wecklein, Headlam (?), and G. Thomson. It is true
that oftwm terere does occur,' incidentally with a meaning quite different from
that postulated here for σχολὴν τρίβειν; but there is no trace of a phrase
corresponding to it in Greek. Most important is the fact that we not infre-
quently meet οὐ (μοι etc.) σχολή with the infinitive as a current colloquial
phrase in Aristophanes and elsewhere (examples in Dindorf's Thesaurus, vii.
1692 c); an echo of this everyday phrase can be perceived in the angry speech
of Clytemnestra.? Instead of the usual οὐκ ἐμοὶ σχολή Clytemnestra says in a
slightly exalted style οὔτοι... éuoi σχολὴ πάρα (cf. Prom. 818 σχολὴ δὲ πλείων
ἢ θέλω πάρεστί μοι, Plato, Politicus 272 Ὁ παρούσης αὐτοῖς οὕτω πολλῆς σχολῆς).
Several commentators have taken πάρα as a preposition (or rather as a
‘postposition’), and were doubly wrong if they linked it with the accusative
τήνδε (Hermann and Karsten protested against this); it should be remembered
that in Aeschylus and Sophocles, when πάρα comes after the noun it always
governs the genitive. τρίβειν here has no object (it is not used quite ‘abso-
lutely’, for it is completed to some extent by θυραίαι, see below), but is used
in the intransitive sense (i.e. the sense which morari has in classical, as
opposed to early, Latin), for which διατρίβειν is more common (frequently in
Aristophanes, sometimes in Thucydides, etc.) Against this Wecklein
(appendix to his commentary, p. 151) raised the objection: τρίβειν, where it
is used absolutely, means not to tarry, but to put off, protract.’ This note is
unhappily expressed (for in the first place there is hardly a reliable example
of τρίβειν used absolutely, and secondly, if the verb is used ‘absolutely’, it
cannot mean ‘put off, protract’), but on the negative side Wecklein gives a
useful warning. For there is indeed no trustworthy parallel for τρίβειν =
‘linger, hesitate’.? Still, there is no reason to assume that the poet could not
1 Virgil, Aen. 4. 271; Livy 1. 57. 5; Statius, Silv. 3. 5. 61, 4. 6. 2; Tac. Hist. 2. 34.
2 Headlam translates ‘I have no leisure to be wasting time . . .’, G. Thomson ‘I have no
time ...’, and yet they alter the οὔτοι... ἐμοὶ σχολή which they are translating.
3 In Dindorf’s Thesaurus and elsewhere, it is true, Thuc. 7. 48. 6 is quoted: τρίβειν οὖν
ἔφη (Nicias) χρῇναι προσκαθημένους καὶ yu}... ἀπιέναι, where τρίβειν is glossed in the scholia
by ἀντὶ τοῦ διατρίβειν καὶ παρέλκειν. It is this interpretation which influenced most of the
modern commentators in their treatment of 7. 49. 2, where Demosthenes in his reply says
ei δὲ dei μὴ ἀπάγειν τὴν στρατιὰν... ἀλλὰ τρίβειν αὐτούς. Here it has become the fashion to
follow K. W. Krüger’s proposal and read τρίβειν αὐτοῦ, But both the alteration and the
assumption of an uncommon meaning of τρίβειν, upon which it is based, have been cogently
refuted by Steup (Appendix to Classen’s commentary, vol. vii, ard. ed., 258f.). Only if the
MS reading is retained does the continuation in 7. 49. 2 ἔφη χρῆναι κτλ. keep its full force
(note particularly ὅθεν... θρέψονται πορθοῦντες τὰ τῶν πολεμίων καὶ ἐκείνους βλάψουσι). If
then τρίβειν αὐτούς (‘to wear them out’) is to stand here, it must have the same meaning
in the speech of Nicias, to which Demosthenes refers. This meaning alone, which is ex-
tremely well suited to the phrase of Nicias 7. 48. 2 χρημάτων yàp ἀπορίᾶι αὐτοὺς ἐκτρυχώσειν,
is in keeping, as Steup shows, with the whole usage of Thucydides (cf. above on 197) ; Steup
further notes appositely ‘dass τρίβειν, wenn es in der Bedeutung von διατρέβειν verstanden
werden könnte, jedenfalls auch wie das Kompositum [this does not of course apply to those
passages in which διατρίβειν with the definition of locality means ‘to stay somewhere’, as in
Ar. Knights 805, Clouds 1002, Pl. Lach. 180d, Philemon fr. 71. 6 K. and elsewhere, nor,
generally speaking, to those in which διατρίβειν is qualified by an adverb or predicative
adjective] einen gewissen tadelnden Nebensinn haben und mehr ein Säumen als ein einfaches
Verweilen ausdrücken würde, Nikias aber unmôglich das, was er getan haben wollte,
irgendwie hat tadeln können’. [In the passage of Demosthenes, however, which is to be

479
lines 1055 f. COMMENTARY

permit himself the use of the simple verb in the place of the compound
which was customary in everyday speech (no daubt in the time of Aeschylus
as well as at the later period for which we have the evidence of Aristophanes),
especially when the context makes the sense unquestionable. For the well-
known tendency of poetical language to use simple verbs in place of compounds
cf., e.g., Wackernagel, Syntax, ii. 188f. But perhaps it is not even necessary to
look to the compound verb. A phrase used by the irritated Creon (Ant. 577)
μὴ τριβὰς ἔτι, belonging clearly to colloquial speech, would in verbal form
become μὴ τρίβειν. Oedipus, too, in his impatience says (Oed. R. 1160)
ἁνὴρ 68, ὡς ἔοικεν, és τριβὰς ἐλᾶι. Clytemnestra's @upaiat . . . τρίβειν corre-
sponds in its rather more exalted style to the plain ἔξω διατρίβειν in the sentence
ἀλλ᾽ οὐχ οἷόν τ᾽ αὐτοῖσι πρὸς τὸν ἀέρα ἔξω διατρίβειν πολὺν ἄγαν ἐστὶν χρόνον
(Ar. Clouds τοῦ f.).
But what are we to do with θυραίαν τήνδ᾽ ἡ To take it with πάρα (as a pre-
position) is out of the question, as has been shown above. Bernhardy,
Wissenschaftl. Syntax des Griech. 190, saw an ‘ellipse’ (scil. τριβήν) in θυραίαν
τήνδε. He was followed among others by Karsten, who compared 219 πνέων
rporraiav and the like; Verrall also compares these expressions but rejects
the idea of an ellipse and takes θυραίαν as a substantive. I think both im-
possible, and I believe that πνέων rporaiav and the similar phrases quoted in
the note on 219 must not be compared with Bupaiar (still less with θυραίαν
τήνδε) τρίβειν. The form τήνδ᾽ cannot be correct; Musgrave restored the
correct reading with τῆιδ᾽ (‘here’). Some editors, including Wilamowitz,
believed that no further correction was required. As πάρα cannot be a pre-
position, θυραίαν can only go with τρίβειν, i.e. it must refer to Clytemnestra.
This yields a much more appropriate sense (as Hermann saw) than if it
referred to Cassandra: the queen, full of impatience to go in and do what she
has planned, says: ‘I have no time to stand outside any longer.’ But if
θυραίαν refers to the speaker, it seems to clash with ἐμοί, Blomfield wanted to
justify this by reference to apparently similar cases such as Cho. 4rof.
πέπαλται Sadré μοι φίλον κέαρ τόνδε κλύουσαν οἶκτον, which are more fully dis-
cussed on 378-80 and on 1611. Blomfield himself, however, had to raise the
objection : ‘fatendum tamen est accusativum in his locis verbo postponi’ ; so,
too, Paley, who says more precisely that the dative must precede. And this
is in fact true without exception! in the examples quoted p. 198. If then
Oupatay . . . ἐμοί is allowed to stand, it is necessary at least to be conscious of
its singularity. I prefer to read θυραίαι with Casaubon, i.e. 1 adopt the same
text as, e.g., Paley and Sidgwick. There is no question of an accumulation of
datives, for the local meaning of τῆιδε must have been clear to the audience
and it wouid not be felt as a dative. @upaiat placed emphatically at the
beginning provides the complement required for τρίβειν.
discussed immediately and in A. Ag. 1056 it is clear that the meaning ‘linger’ in a tone of
disapproval or depreciation is required.] The dictionaries quote as a second instance
Demosth. 23. 173 τριβόντων τούτων καὶ οὐδὲν ἁπλοῦν οὐδὲ δίκαιον ὑμῖν ἐθελόντων πρᾶξαι. where
the first hand in S together with the other MSS gives τριβόντων (accepted by, e.g., Sykutris),
whereas the corrector of S gives διατριβόντων (so, e.g., Butcher’s text).
τ Cf. also Kühner-Gerth, ii. 112, where, however, examples of the type mentioned above
are quite wrongly compared with anacolutha such as Pl. Alc. ii. 148 d ἥκουσα... τοὺς οὖν
᾿Αθηναίους ἀγανακτοῦντας τῶι πράγματι kai ἀπορουμένους τίνι χρὴ μηχανῆι τῶν παρόντων κακῶν
ἀποτροπὴν εὑρεῖν, βουλευομένοις αὐτοῖς δοκεῖν κτλ.
480
 AESCHYLUS
AGAMEMNON
 AESCHYLUS

AGAMEMNON
EDITED

WITH A COMMENTARY BY

EDUARD FRAENKEL

VOLUME III
COMMENTARY ON 1056-1673
APPENDIXES, INDEXES

OXFORD
AT THE CLARENDON PRESS
 This book has been printed digitally and produced in a standard specification
in order to ensure its continuing availability

OXFORD
UNIVERSITY PRESS

Great Clarendon Street, Oxford OX2 6DP

Oxford University Press is a department of the University of Oxford.
It furthers the University@ objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
So Paulo Shanghai Taipei Tokyo Toronto

Oxford is a registered trade mark of Oxford University Press

in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York

© Oxford University Press 1950

The moral rights of the author have been asserted

Database right Oxford University Press (maker)
Reprinted 2003

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose this same condition on any acquirer
ISBN 0-19-927172-0 (Vol. 3 )
ISBN 0-19-814101-7 ( Set of three Vol.G)
 CONTENTS
VOLUME III
COMMENTARY ON 1056-1673 . . . . 481
APPENDICES
A. On the Postponement of certain Important Details in
Archaic Narrative . . . 805
B. On the Weapon with which, according to the Oresteia,
Agamemnon was murdered . . 806
Cc. Cho. 991-1006 . . . . . . . 809
D. The Footprints in the Choephoroe . 815
E. Short Syllables before Initial Mute and Liquid in
à the
Lyrics of Aeschylus 826
F. The Word-order in AS. 1434 οὔ μοι dé μέλαθρον Anis
éuraret . 827
ADDENDA ET CORRIGENDA TO COMMENTARY . . 829
INDEXES . . . . . . . . . 833
 COMMENTARY
1056 f. τὰ μὲν yàp . . . σφαγὰς πυρός. Some commentators (cf., too, Con-
ington on Cho. 183) and Kühner-Gerth, i. 384, have taken ἑστίας μεσομφάλον
as a local genitive, and have adduced parallels some of which are quite
inappropriate. Paley rightly objects that ra μέν would in that case be very
clumsily placed. In his own interpretation, which was anticipated by Peile
and followed by Verrall, 7a . . . μεσομφάλου forms a kolon of its own, ‘for as
regards the family altar’. This seems to be extremely artificial, and the
passage which he and others have quoted for comparison (830 ra δ᾽ ἐς τὸ σὸν
φρόνημα etc.) is quite different because the grouping there arises naturally from
the subject-matter ; but who would think of regarding τὰ τῆς ἑστίας as a fresh
heading in this brief speech, which needs no particular grouping? Theobvious
way of construing the passage is to take τὰ μὲν yàp . . . μῆλα together, though,
of course, Hermann’s observation must be regarded, ‘respondent sibi ra μὲν
et σὺ δέ. Nam τά non articulus est, sed pronomen, 1lla’; cf. also Sidgwick’s
note: ‘article separated from subst., really a usage from Epic poetry [this is
perhaps rather too narrow], where the article is still a demonstrative or
pronoun: “they, the sheep".' Wackernagel, Syntax, ii. 130 ff., quotes among
other instances A 348 ἡ δ᾽ ἀέκουσ᾽ ἅμα τοῖσι γυνὴ κίεν, ‘but she, the woman...’,
and Ar. Thesm. 505 τὸ δ᾽ εἰσέφερε γραῦς ἐν χύτραι τὸ παιδίον, cf. also Ach.
179 f. oi δ᾽ ὥσφροντο πρεσβῦταί τινες Axapvıroi, στιπτοὶ γέροντες KrA.” But if
we accept this interpretation of ra, the genitive ἑστίας is left in the air.
Attempts have long been made to deal with this difficulty and another one
together. For σφαγὰς πυρός can hardly be sound; a paraphrase such as
‘mactatio igni hostiam decernens, ignis gratia instituta’ (Klausen) cannot be
allowed to count as an interpretation. The majority of modern editors be-
lieved that a way out of the double predicament was offered by Musgrave’s
σφαγὰς πάρος, and consequently they put it into the text. This may serve
as an emergency support for the construction, but I would not like to trust
its solidity. The fact that several heterogeneous parts of the sentence come
between ἑστίας μεσομφάλου and πάρος disturbs me, and my qualms are not
! Tt would be wrong, however, to regard this with Schneidewin as ‘an ominous combina-
tion’ (the same view is expressed in grosser form by G. Thomson: ‘the other victims are
ready; do you therefore make haste’). Before the appearance of Hermann’s edition W.
Sewell had pointed out (notes on his translation) that ‘the separation of τὰ μὲν from μῆλα is
important’. Unfortunately he, too, advocates a dark dAAnyopeiv: “The indefinite τὰ μὲν is
in Clytemnestra’s mind Agamemnon; the μῆλα is added afterwards as if to prevent sus-
picion.' The connexion seems to me quite simple: ‘the victim is ready, and you (you are
delaying the holy rite) make haste!’ The relation between the uév.. . and the δέ... clauses
is neither more nor less adversative than, e.g., Prom. 1-3.
2 I take A. Cho. 63 f. in the same way, τὰ δ᾽ ἐν μεταιχμέίωι σκότου μένει ypovilovras (a certain
emendation) dy»: ‘but other things in the twilight await those who linger on, woes’. Here
τὰ δέ prepares the way for ἄχη just as in Ag. 1056 τὰ μέν does for μῆλα. Of course this τὰ
δέ has the same function as if we said ‘in other cases’. The slight irregularity (ra δ᾽ ἐν
between 62 τοὺς μέν and 65 τοὺς δέ) serves the purpose of avoiding monotony.
4872.3 B 481
 lines 1056 f, COMMENTARY
allayed by Headlam’s reference (C.R. xvii, 1903, 287) to 129 πάντα δὲ πύργων
κτήνη πρόσθε or indeed to Theocritus' studied concluding clause (16. 108 f.) ri
γὰρ Χαρίτων ἀγαπατὸν ἀνθρώποις ἀπάνευθεν. There is still another troublesome
point, about which the commentators say nothing. éorqkev . . . μῆλα πρὸς
σφαγάς is translated ‘the cattle... are standing ... for slaughter’ (Headlam),
‘stehn schon die Schafe zur Opferung bereit’ (Wilamowitz), ‘les victimes sont
prétes, attendant le couteau’ (Mazon). Can πρὸς σφαγάς have this meaning?
I have found no tolerably corresponding parallel for this use of πρός, and the
heading ‘in reference to’, under which L-S (πρός C III. 3) class the passage,
is suitable for all the other passages quoted there but does not suit this one.
50 the suspicion of Paley and others that πρὸς σφαγάς has taken the place of
something else is certainly worth considering.
1056. ἑστίας μεσομφάλου means an altar in the middle (of the palace) and
does not indicate whether it is to be thought of as being in an enclosed room or
in the courtyard; probably it is in the courtyard.’ It is likely that the same
cult is meant as in 1037 f. (rejected by Süss, RE viii. 128o, without argument).
μεσομφάλου: the not infrequent examples of the word in the three great
tragedians (Ion and Agathon play with the pompous word) clearly support
the view that the epithet referred originally to the Omphalos in Delphi, 'the
midland navel-stone beside the singing fountain'. It had its origin presum-
ably in the sacred poetry of Delphi; it was probably from there and not from
Tragedy that Aristonoos (fr. 2 a. 3 D., Wilamowitz, Verskunst, 496) took it
over. From this we may conclude that Aeschylus purposely used the word
here in an arbitrarily generalized sense.” Verrall justly noted the peculiarity.
But unfortunately Aeschylus can never be ingenious enough for Verrall,
who here suggests that Clytemnestra wishes to mock Apollo's prophetess,
with a word from the vocabulary of Apollo’s cult.
1058. There is no decisive grammatical difficulty here;? the dative without
support in the preceding words could well be justified. Nevertheless, it is a
relief to find that Wilamowitz (as early as his edition of 1885) excised the
line, which he recognized as a pasticcio from 1043 f.; he was followed by
Mazon, Murray, and A. Y. Campbell. The queen is in a very great hurry: 'I
have no time to linger out here: the victim waits within.' The addition of an
emotional flourish is intolerable ; Clytemnestra has now no reason for making
any further display of gratitude or pleasure for the benefit of Cassandra or
the Chorus. The recurrence in a completely different context of οὔποτ᾽
ἐλπίσασι shortly after the οὔποτ᾽ ἐλπίσαντες in 1044 is just as improbable as
the κληδόνας παλιγκότους in 863 in view of 874, or as πολλὴν ἄνωθεν in 871
before πολλὰς ἄνωθεν in 875. I have not, however, succeeded in discovering
the purpose for which 1058 has been botched together.
1059. μὴ σχολὴν τίθει does not seem to occur anywhere else but is im-
mediately comprehensible. Compared with Suppl. 207 μή vvv σχόλαζε, 883
! The remarks of Wilamowitz about the palace in Euripides’ Herakles (Commentary on
the ‘Sixth Scene’, ii.? 206) may be applied to the palace in the Oresteia : ‘it has the plan of an
Attic house. . . . In the open space of the courtyard stands the altar, which is dedicated to
... Ζεὺς épxeios.’ Cf. Aristophanes fr. 245 K.; M. P. Nilsson, Greek Popular Religion, 66 f.
2 Schuursma would have been more justified in giving this word a place in his book De...
abusione ap. Aesch. than in the case of some of the words he actually discusses there.
* However, the connexion with ὡς does not seem to be satisfactory. This was pointed
out to me by Merton Dagut, a member of my seminar class.
482
 COMMENTARY line 1061
μηδέ ris oxoAalerw (van Heusde), the phrase is in a rather more exalted style.
But the tone of the demand as a whole is not quite that used in polite
conversation and therefore is well suited to express the queen’s impatience;
so much was clear to every ancient spectator. This effect is produced by the
particular form of the conditional sentence in association with the imperative,
with which Headlam on Herodas 7. 47 compares suitably a whole list of similar
phrases! such as E. Iph. A. 817 δρᾷ δ᾽ (cf. the editions), εἴ τι δράσεις, Pl.
Rep. 350 e ἀλλ᾽ εἴπερ τοῦτο ποιήσεις, ποίει, Theocr. 5. 78 εἶα λέγ᾽, ei τι λέγεις,
frequently in Plautus age si quid agis etc. For τι τῶνδε ‘a kind of litotes =
rade”, see Blass on Eum. 467, who also compares Ag. 1239, Eum. 142.
1060. ἀξυνήμων here only, instead of the usual ἀσύνετος. There is no con-
nexion with the epic words συνημοσύνη ‘treaty’ (as early as Homer, belonging
to συνίεμαι, cf. E. Risch, Wortbildung d. hom. Sprache, 1937, 138, beside which
there is in Homer μεθήμων, μεθημοσύνη and ἐφημοσύνη) and συνήμονες (Apoll.
Rhod.) ‘comrades’.
μὴ δέχηι: ‘si mea verba non intelligis’ (Blomfield), ‘if you do not take the
meaning of my words’ (Headlam), ‘si tu n’entends pas mes raisons’ (Mazon),
and the like are the usual translations of the passage (cf., e.g., Linwood, Lex.
to Aesch. s.v. ‘to understand’); they are not altogether wrong, but blur the
outlines considerably. Kennedy translated ‘thou repliest not’ and explained
it thus: ‘Such is the full force of od δέχει λόγον, a phrase used in dialectic
discussion. Its correlative is δοῦναι Aóyov.' This goes too far but is at least
an attempt to do justice to the strong expression used here. It seems best to
leave to εἰ μὴ δέχηι its ordinary meaning: ‘if you do not receive’. In the
situation which Clytemnestra assumes as a possibility, the main obstacle
would make itself felt before the question of understanding or failure to
understand arose. If to Cassandra Greek words are noises only, then the
words do not reach her: she, for whom they are intended, does not receive
them, therefore does not pick them up (to use a modern comparison, there
is a transmitting station but no receiver). The precision of this suggestive
though unusual thought should not be blunted.
1061. The assumption that this demand is addressed to someone other than
Cassandra originated, as far as I can see, in the nineteenth century. W.
Sewell (1846) makes σὺ δὲ κτλ. refer to the Chorus ; Conington rightly comments
on this ‘which can never be after δέχει immediately preceding’. He had a true
understanding of dramatic dialogue and knew that we are not allowed to
invent stage directions to which nothing in the words of the text corresponds.
Nowadays, of course, it is considered possible for an actor in a Greek play to
turn sharply round (Cassandra is still in the wagon) from one person to another
in the middle of a sentence,’ without even the vocative of a noun to mark the
transition. This was Wecklein's view, which was followed hesitatingly by
' Older than anything he quotes is K 96 f. ἀλλ᾽ εἴ τι Spaivers, ἐπεὶ οὐδὲ σέ γ᾽ ὕπνος ἱκάνει,
δεῦρ᾽ ἐς τοὺς φύλακας καταβήομεν, ὄφρα ἴδωμεν.
* Against this cf., e.g., Cho. 712, 983 or E. Alc. 545 f. (first a complete sentence concludes
the stichomythia, then comes a new sentence, ἀσυνδέτως, containing an order to the
servant).
. 3 Compared with such an assumption even Keck's attempt was better. He followed A.
Ludwig in putting the two lines 1060 f. after 1052 and took them as addressed to the Chorus,
reading μὴ δέχοιτ᾽ ἔπος. Of course neither the conjecture nor the rearrangement bears
closer examination.
483
line 1061 COMMENTARY
Verrall (‘in that case we should suppose that the sentence is impatiently
broken off at Aoyov’—schlecht und modern, to borrow Faust's phrase), —and
confidently by A. Y. Campbell, Murray, and G. Thomson. Now since the
Chorus happens to consist not of barbarians but of Greeks, Wecklein splits
the phrase «apfávox χερί (‘give the foreign woman a sign with your hand’),
while A. Y. Campbell, who construes in the same way (p. 103 of his edition),
introduces a pedisequus, and Thomson weakens the force of καρβάνωι and
paraphrases: σὺ δὲ ἀντὶ φωνῆς συνετῆς φράζε dovverwi καὶ βιαίαι χερί. Is it
really necessary to say that the weighty word καρβάνωι can only refer to the
woman who has just been described as φωνὴν βάρβαρον κεκτημένη, and upon
whose foreign origin and language so much stress is laid here and later
(1200) ? |
There is no linguistic obstacle in the way of referring où δέ to Cassandra.
Cf. in addition to the commentaries and Denniston, Particles, 177 ff. (in
particular 181), Rupprecht, Hermes, Ixxiii, 1938, 244. Eum. 885 ff. ἀλλ᾽ εἰ μὲν
ἁγνόν ἐστί σοι Πειθοῦς oéBas . . . σὺ δ᾽ οὖν μένοις av comes particularly close to
our present passage, for there the ‘you’ in the conditional clause is, to use the
traditional expression, the ‘logical subject’ (‘si tu sais respecter la Persuasion
sainte” Mazon). Moreover, to settle the point at once, the view of Karsten
(who reads ὧδ᾽) and Meineke (who reads ἀλλ᾽ that σὺ δ᾽ is doubtful after the
same beginning of the line in 1059 cannot be admitted as far as Aeschylus is
concerned. The ov δέ here corresponds to the natural use in conversation of
σὺ δ᾽ ἀλλά (cf. Wilamowitz on Ar. Lys. 904), ‘then . . . at least’.
The real difficulty for the critics of the naturalistic age lies in the thought.
G. Thomson first gives the traditional interpretation: “but if, not under-
standing Greek, you do not apprehend what I have said, then address me not
with voice but with barbarian hand’ and says then: ‘this interpretation is
flatly contradicted by common sense.’ This statement could be met even on
the grounds of a radical naturalism, quite inappropriate though that would
be. It is a common experience that, when we are making a frantic effort to
establish the necessary understanding with someone who has not the slightest
knowledge of our own language, we continue to use that language in an
attempt to force the information into the foreigner long after our ‘common
sense’ must have convinced us of the hopelessness of our efforts. Clytem-
nestra’s words have been appositely explained by Schiitz (followed by
Blomfield, Hermann, and others): ‘Ista dicens regina ipsa nutu, vultu, gestu
sermonem adiuvabat’; this is in accordance (except for the inappropriate
‘vultu’) with the general conception we must have of the performance on the
ancient stage. If Cassandra were any other person but herself, she could
react in a normal way and make clear καρβάνωι χερί her lack of understanding.
She would do it perhaps by a slight sideways movement of the lower arms,
with the hands open and the fingers apart. Therefore the two lines, taken as
addressed to Cassandra, are far from containing ‘pure nonsense’ (Mackail,
C.R. xix, 1905, 197) or ‘a patent imbecility’ (Campbell, C.R. li, 1937, 116),
even if we judge them by the crude criterion of truthfulness to life. More
helpful than such general considerations is G. Thomson’s reference to stage
convention. In Attic Tragedy the fact that non-Greeks do not speak Greek
is in general ignored, except, as Thomson rightly points out, when it is
mentioned ‘with a special purpose’. A good example of such a purpose is
484
 COMMENTARY line 1067
found in Suppl. 118 kapßäva δ᾽ αὐδὰν εὖ γᾶ kovveis. In the next scene the
daughters of Danaus converse with Pelasgus without the slightest difficulty;
in the prayer of the Chorus, however, emphasis is laid upon the forlorn state
of the girls who come from afar and beg for protection in a foreign land:
there only the earth, the land which reared their ancestress, understands the
language, καρβᾶνα αὐδάν, of the fugitives. Similarly in the passage in the
Agamemnon the emphasis placed upon the language barrier is to provide a
particular effect: it is clear and should not be artificially obscured.’ In the
present scene one of the poet’s chief concerns is to win sympathy for Cas-
sandra ; the wretchedness of her situation is increased by the alien environ-
ment, and it isa contributory factor that it can be assumed at least for a time
that she does not understand the language. Still more important perhaps,
both in itself and for the following scene, is the fact that, as long as Clytem-
nestra is present, Cassandra not only remains dumb but does not even make
a movement with her hands, as anyone in her situation would do who
wanted to establish any contact at all; but she is like an image of stone, as if
dead. Finally, to look at the situation from Clytemnestra's viewpoint: to
her it is a matter of great urgency to bring her victim as quickly as possible
into the house. After Cassandra has failed either to accede to her demands or
to speak a single word, the queen makes a last attempt to elicit from her at
least a gesture, which would either have to be acquiescent or alternatively
open the way for renewed pressure, whereas as long as Cassandra remains
motionless, she is unassailable. The words of the queen and the coryphaeus
continually turn the gaze and thoughts of the audience to the woman
cowering in the wagon, without any sign of life, in extreme loneliness; she
has, it appears, no part in anything that is said or done around her, for her
or against her.
With 1061 cf. (Blomfield) Hdt. 4. 113. 2 καὶ φωνῆσαι μὲν οὐκ εἶχε (οὐ yàp
συνίεσαν ἀλλήλων), τῆι δὲ χειρὶ ἔφραζε és τὴν ὑστεραίην ἐλθεῖν κτλ., and for the
relation between εἰπεῖν or φωνεῖν and φράζειν see J. H. H. Schmidt, Synonymik
der griech. Sprache, i. 90.
1062. ἑρμηνέως . . . ropoû: cf. 616.
1063. νεαιρέτου : Meineke, Phil. xix, 1863, 206, here again (cf. on 1061 σὺ 8’)
found the repetition (cf. 1065) of the same word in close proximity suspicious
and therefore made the conjecture νεαγρέτου ; this appears arbitrary in view
of the examples brought forward by Römer, Sitzgsber. Bayer. Ak., Phil.-hist.
Kl. 1888, 238, of similar repetitions in Aeschylus.
1067. Schol ἀπὸ τῶν στρηνιώντων ὑποζυγίων, ἃ οὐκ εἴκοντα τῶι xadude
ἀφρίζει μετὰ αἵματος. In accordance with this explanation the commentators
have rightly taken ἐξαφρίζεσθαι as middle and αἱματηρόν predicatively. The
cruelty of the picture is not to be reckoned against Clytemnestra but rather
against the practices of Greek horsemanship ; except that the fact of a woman
employing this image is perhaps significant of her character. Several Greek
horse-bits have come down to us; they have been thoroughly studied by

! I do not wish to go into all the mysteries which G. Thomson reads into this passage.
He reaches his interpretation partly by referring δράσεις 1059 (as Keck had already done) to
the sacral term δρώμενα; the correct interpretation of δράσεις should be based upon the
examples compared with it by Headlam (cf. on 1059), particularly E. Iph. A. 817; cf. also
B. Snell, Aischylos, 7 ff.

485
line 1067 COMMENTARY
E. Pernice in his article ‘Griechisches Pferdegeschirr’, 56. Berliner Winckel-
mannsprogr. 1896, in which also the relevant section of Xenophon's π.
ἑππικῆς, 10. 6 ff., is explained.’ The effects which might, but need not, be
obtained by the use of the sharp snaffle (ὁ τραχὺς χαλινός) can be unmistakably
deduced from the specimens which have been preserved:2 it ‘was very sharp
and cruel’, as Pernice? states on p. 18 (cf. also p. 25); he recalls in this con-
nexion the anecdote of Apelles,* who had difficulty in depicting the mixture
of blood and foam on the horse's mouth. There is no need to speak of
"incredible instruments of torture', as does W. Passow, 'Studien zum Par-
thenon', Philol. Unters. xvii, 1902, 55, nor should it be supposed that a Greek
rider constantly misused the sharp snaffle, but it is flying in the face of the
facts when A. Diehl, Die Reiterschépfungen der Phidiasischen Kunst (1921),
109 f., attempts in opposition to Pernice to represent this snaffle as a com-
pletely harmless implement. If the horse ever failed to obey, or if the rider
were unsympathetic or impatient, then the spiked rolls and the disks of the.
bit would in a moment tear the membrane of the lower jaw and the tongue
and draw blood. The earliest reference to this in literature is the present
passage. The force of this evidence cannot be argued away; the phrases
have remarkable points of contact with those of the Apelles story quoted
above. [See now Yalouris, Mus. Helv. vii, 1950, 30 ff.]
1068. οὐκ ἀτιμασθήσομαι occurs also in S. Oed. R. 1081.
1070. ὦ τάλαινα: this serves as a sort of leitmotiv on a small scale for the
Cassandra scene, cf. 1136, 1143, 1158, 1274, and, with increased emphasis,
1295. The effect is deepened when in 1197 Cassandra thus addresses the
Clytemnestra of her vision.
1071. καίνισον ζυγόν: Abresch referred to Hesychius καινίσαι'" καινῶι χρή-
σασθαι. The earliest examples of the verb are this and Cho. 492.5 Wacker-
nagel, Vermischie Beiträge, 38, points out that in Aeschylus throughout and
in some passages in Sophocles and Herodotus xawós with its derivatives
exhibits its original sense ‘unusual, extraordinary', cf. on 96o.
At the beginning of the line the slight alteration εἴκουσ᾽ has received almost
universal acceptance. Perhaps it is right. The MS reading ἑκοῦσ᾽ ἀνάγκηι
τῆιδε κτλ. leaves the dative without any connexion (even if Cho. 492 ὧι σ᾽
ἐκαίνισαν can be retained, it does not constitute a parallel). The artificialities
of Verrall (' ἀνάγκηι a possessive dative. The ἀνάγκη is personified as imposing
the yoke") and Plüss (similar, in the main, to Verrall) refute themselves. We
should, however, be reluctant to part with ἑκοῦσα. The difference is not very
! My attention was directed to this article and to those mentioned later by Paul
Jacobsthal.
2 The reader who has no access to Pernice's article will find a good photograph of a 'very
severe bit, for use in hard mouths, probably of the fifth or fourth century’ in the little book
The Daily Life of the Greeks and Romans, published for the Metropolitan Museum of Art
by Helen McClees (1941), p. 108, fig. 132.
* Wilamowitz, Pindaros, 372 n. 4, rightly agrees.
4 [Dio Chrys.] 63. 4 f. Gi, p. 146 v. Arnim) ἅπαντα δὲ ἐχούσης τῆς εἰκόνος ἐοικότα ἔλειπεν
(Apelles) ἀφροῦ χρῶμα, olov dv γένοιτο μιγέντος αἵματος καὶ ὑγροῦ κατὰ συνεχῆ μῖξιν... αἷμα
δὲ ἐπιρραινούσης τῶι ἀφρῶι τῆς ἐκ τοῦ χαλινοῦ ὕβρεως and later χρώματα ἐοικότα ἀφρῶι Yına-
γμένωι.
5 Ifin that passage we are to follow Conington’s proposal and read ὡς (Abresch, Blom-
field) exatvıoas (this was accepted by Paley, Sidgwick, Wecklein in his commentary, Verrall,
Murray, G. Thomson), the usage would be exactly parallel in both passages.
486
 AMOIBAION (1072-1177)

substantial ; but it seems more delicate, and more in keeping with the general
attitude of the leader of the Chorus, if he says not ‘yield to compulsion’ but
‘willingly take the yoke upon you as something unaccustomed’. So it is
worth asking afresh whether Casaubon’s ἑκοῦσ᾽ ἀνάγκης τῆσδε x. ζ. is not
preferable. ἑκοῦσα, placed emphatically at the beginning, implies ‘before she
has you forcibly dragged into the house’. If this is correct, the speaker is here
taking up again, and this time with success, his request of 1054, πείθου. The
phrase ἀνάγκης ζυγόν would have to be added to the examples quoted on 218.
The juxtaposition, too, of the ideas so frequently contrasted with each other,
ἑκών and ἀνάγκη, is desirable: if Cassandra follows the advice, she would, at
least according to the hopes and wishes of the Chorus, not suffer the same fate
as the person who ἄκων σωφρονεῖν ἀναγκάζεται (cf. on 180).

AMOIBAION (1072-1177)
Seven pairs of stanzas. In the first four all the singing is done by Cas-
sandra; the reply of the Chorus, or probably of the coryphaeus, is each time
confined to the recitation of two iambic trimeters. In the fifth stanza it first
seems as if the same pattern would be repeated, but after the two trimeters
(1119 f.) unexpectedly the whole Chorus burst into dochmiacs, a grandiose
effect. Then, in the last two stanzas, the part of the Chorus is entirely
lyrical, whereas Cassandra here each time begins with lyrics but concludes
with two trimeters. This insertion of trimeters must be taken together with
the occurrence of a single iambic trimeter in Cassandra’s earlier stanzas (at
the end of her part in IT and III, and in the middle of it in IV and V). For
the reason of the change in the distribution of lyrics and trimeters in the
last three pairs of stanzas see p. 539.
What we have here is a special variety of ‘epirrhematic’ composition. Its
normal type, such as appears here in the first four stanzas, consists of pairs
of lyrical stanzas sung by one party (primarily the Chorus, see below) and an
equal (2 or 3 or 5) number of trimeters by which the other party replies.
Wilamowitz, Interpr. 74, instances, besides the Cassandra scene, Suppl.
347 ff., Pers. 256 ff., Sept. 203 ff., to which we may add Suppl. 734 ff, cf.
W. Kranz, Stasimon, 15. The amoibaion Suppl. 347-417 is composed of three
pairs of stanzas sung by the Chorus, in answer to which Pelasgos always
recites five trimeters, save after the last antistrophe, where (407 ff.) he goes
on with a longer speech. A similar, though slightly bolder, ‘licence’ is to be
found at the end of the amoibaion Pers. 256 ff., where the messenger, who has
taken up each of the former stanzas of the Chorus with two trimeters, remains
silent after the third antistrophe, ‘ut Reginae tandem verba facere susti-
nentis gravitatem extollat’ (Wilamowitz, whose view seems convincing,
whereas Weil and, following him, O. Schroeder, Aeschyli cantica, 2nd ed.,
p. 18, hold that the trimeters of the ὑποκριτής precede the songs of the Chorus ;
they are forced to assume a lacuna before 255). The amoibaion Sept. 203-44
1s very regular: three pairs of stanzas from the Chorus, each stanza followed
by three trimeters from Eteocles. The form of Suppl. 734-61 is also regular:
two pairs of stanzas from the Chorus, each followed by two trimeters from
Danaus. There is, however, one special feature to be noticed here: in each
piece of the Chorus the lyrics (dochmiacs) are preceded by two trimeters.
This is an interesting parallel to Ag. 1119 ff.
487
lines 1072-1177 COMMENTARY

The typical, and doubtless older, form, then, shows this order: lyrics of
the Chorus, followed by trimeters from the actor. Here, in the Cassandra
scene, we find for the first time (cf. Kranz, Stasimon, 20) the reverse: the
ὑποκριτής is not answering but leading.
Metre
I. 1072-3 = 1076-7. Wilamowitz terms this ‘4 bacchei’. This is not
absolutely wrong, but it should be understood that what we have here must
not be identified with the normal bacchiac tetrameter (Prom. 115, Ag. 1103 =
1rro, Cho. 349 f. = 367 1., Aesch. fr. 23 and 341 N., Trag. fr. adesp. 144 N.,
S. Phil. 396, E. Ion 1446, Ar. Thesm. 1143 etc.),! for that line would certainly
not admit the syllaba anceps after the third metron (Ἄπολλον, see note) and
probably not v v v — (órororot) instead of o — -- O. Schroeder calls the first
metron a cretic, the three others bacchiacs. Terminology matters but little
here. It is more important to recognize that these are four separate ejacula-
tions, each of which, in terms of metre, may be regarded as roughly equi-
valent to a catalectic iambic metron (the possibility that in the first place we
have to write with M ὀτοτοτοτοῖ, i.e. a full iambic metron, cannot be ruled out).
II. 1080-2 = 1085-7. Three separate (cf. on I) bacchiacs+dochmiac,
iambic trimeter. -
III. xogo-2 = 1095-7. Two dochmiacs, 2 iamb., iambic trimeter.
IV. 1100-4 = 1107-11. Iamb.-+dochm., lecyth. (or, deleting 1101 μέγα
and 1108 πόσψ, see note on 1101, dochm.), iambic trimeter, bacchiac tetra-
meter + dochm.
V. 1114-24 = 1125-35. Cass. 2 dochm., 2 iamb. (catal), iambic trim.,
iamb. +dochm., dochm. --2 cretics.
Cho. 2 iambic trimeters, 4 dochm., cretic (paeon) +2 bacchiacs, iamb.+
cretic. The last line, iamb. 4-cret., is of course a common form of iambic
dimeter, see e.g. Cho. 429, 431 f., etc.
VI. 1136-45 = 1146-55. Cass. cret. (paeon)+bacch.+dochm., 2 dochm.,
2 iambic trimeters.
Cho. 3 dochm., 2 cretics+spondee, dochm. +2 cretics, 3 dochm.
At 1136 = 1146 the ὦ of the first & is probably shortened before the initial
vowel. We obtain then vv v— v—-, which is exactly the same as 1072
ὁτοτοτοῖ ποποῖ δᾶ.
The metre οὗ 1142 νόμον ἄνομον, οἷα τις ξουθά is self-evident (for the two
cretics cf. the other instances in V and VI) as analysed above (cf. also
©. Schroeder, Cantica; P. Maas, Sokrates, iii, 1915, 314 n. 2); therefore no
refutation of Wilamowitz (cf. also his Verskunst, 413 n. 2) is necessary.
VII. 1156-66 = 1167-77. Cass. interjection, 3 iamb., iamb.--dochm.,
3 dochm., 2 iambic trimeters.
Cho. 2 dochm., 2 iamb. 4-2 dochm., 3 dochm.

1072. Perhaps órororo: is to be adopted from F Tr; the one additional ro in

M may be due to a mistake such as often occurs when syllables are repeated
in this way. The same ejaculation drovo: causes complete confusion in the
MSS in Pers. 268 (cf., too, 274), 918, E. Phoen. 153o (on which Wilamowitz,
! The form in which the text of A. Zum. 789 is given by Hermann and Wilamowitz is
doubtful, |

488
 COMMENTARY line 1072

Berl. Sitzgsber. 1903, 598 n. 6, notes: ‘the interjection is, as usual, written in
different ways: the metre is decisive’), cf., e.g., the variations in MS readings in
Ar. Birds 227, 243, 260, etc. But it is equally possible that ὀτοτοτοτοῖ is the
genuine reading, cf. the note on the metre. For the invocation of the god
here and in the following lines the ‘recensio’ of the MSS points to AzoAAov
Ἄπολλον. This is preserved in F Tr at 1073, 1077, 1080, and 1085, and in the
last two places also in M. It should be clear that the invocation must be
precisely the same in all four passages: the monotonous repetition contri-
butes to the pathetic effect. It follows that at 1073 and 1077 ὦ is wrongly
added in M. To obtain a smooth bacchiac tetrameter Hermann! read here,
and consequently in the next strophe and antistrophe also, AméAAwv ; he was
followed by, e.g., Weil, Wilamowitz, Mazon, G. Thomson. But there is an
objection to this apparently insignificant alteration. It is improbable that the
nominative should occur in one single instance? in place of the usual Ξπολλον
which Aeschylus uses elsewhere. It is not, of course, impossible. K. W.
Krüger's remark (Griech. Sprachlehre, ii, $ 45, 2 n. 1) that ‘in the case of names
the vocatival nominative cannot be called frequent? even in the poets' is
correct in not ruling out altogether the possibility of such cases. In Aeschylus,
Pers. 649 f. Aidwveus and Sept. 105 παλαίχθων Ἄρης serve as vocatives. With
regard to the much-discussed use of Aias and Οἰδίπους in lieu of voca-
tives in Sophocles* cf. now particularly W. Schulze, Kl. Schr. 85 f. But the
fact remains that, as far as I know, not one example of Ἀπόλλων used as
a vocative can be produced, as against the numerous examples of Ἄπολλον.
So we should retain Ἄπολλον here as the MSS have it. AroAAov Ἄπολλον could
by itself be taken as a reizianum (v — συ — 5, like "ἔρως ὁ Διὸς παῖς), but this
is not at all probable, principally on account of the metrical context, and also
because the reizianum is very rare in Aeschylus.” So I prefer to follow
1 Hermann’s note describes ὦ -- το - - as ‘dochmius hypercatalectus’, cf. also his Elem.
doctr. metr. 254.
? Hermann does not explain how he justifies the nominative in his text. Perhaps he
reckoned this instance among those in which he (mistakenly) supposed ‘nominativum, quem
pro vocativo positum volunt, non vocantis, sed declarantis esse' (Preface to his edition
[1838] of Eur. Andr. xv ff).
3 With regard to the example quoted there Ζεῦ πάτερ... ᾿Ηέλιός re cf. Wackernagel,
Über einige antike Anredeformen (Progr. Göttingen 1912), 9, and Syntax, i. 7 and 307; for
further discussion of the problem cf. Lófstedt, Syntactica, 1, and ed., 97.
+ jebb and Pearson, although in the case of S. Ay. 482 and 1015 they note the variant
Alay found in Suidas, fail to do so on 289 (Suid. s.v. χρήματα, iv. 822. 5 Adler), in spite of the
reference in Hermann and Lobeck.
5 On ἀπόλλων ἐμός, which belongs to a different category, see p. 492.
$ R. Loewe, who in Kuhns Zischr. lv, 1927, 53 f., brings forward examples from Greek
poetry and particularly from Tragedy of the use of the nominative for the vocative in proper
names, includes in his list Ag. 1073, 1077 Ἀπόλλων, i.e. he accepts the text of Wilamowitz
(or of Weil) without taking any account of the MSS.
? There is possibly a reizianum in Ag. 1483 φεῦ φεῦ κακὸν αἶνον (= 1507), see Wilamowitz's
analysis, Verskunst, 249 n. 1 (he comes to a different conclusion in his edition and in
Verskunst, 390). As for ᾿Αθαμαντίδος " EMas (Pers. 70), which occurs in a passage composed
entirely of ionics (it is the same in Suppl. 868 — 878), it had better not be called a reizianum,
although it looks exactly like καλὰς ὧρας ἄγουσα or μεγάλα χάρις αὐτῶι. Again in the case
of Pers. 949-51 (Wilamowitz, Verskunst, 402, but cf. the cautious remark of O. Schroeder,
Aeschyli cantica, 2nd ed., p. 25) the relation to what we generally understand by reizianum
is doubtful. In Ag. 451 = 470 the ‘reizianum’ v v — u — — serves as clausula to the preceding
tonici a minore. In favour of taking Ἄπολλον “Ἄπολλον as a reizianum it could perhaps be
489
line 1072 COMMENTARY

O. Schroeder, Aeschyli cantica, and ed., 64, and scan AroAXov Ἄπολλον as two
bacchiacs,' assuming syllaba anceps in the middle, so that on the whole
Hermann’s metrical analysis (see above) remains unchanged. At the end of
a vocative serving as an interjection, syllaba anceps is probably as legitimate
as hiatus (cf. on 1537).
δᾶ. The scholia explain γῆ Awpıras‘ ὅθεν καὶ Δημήτηρ olov γημήτηρ;
the same explanation is given in schol. Eum. 842, and, in a more elaborate
manner, in schol. Prom. 568 τὸ δὲ d δᾶ ὦ γῆ. οἱ γὰρ Δωριεῖς τὴν γῆν δῆν καὶ
δᾶν φασιν, καὶ τὸν γνόφον δνόφον (a fuller form of this scholion is to be found
Etym. M. 60.8 s.v. dAevada). Besides this explanation, we find two others put
forward in schol. E. Phoen. 1296 φεῦ δᾶ φεῦ δᾶ: οἱ μὲν ὡς ἕν μέρος λόγου
ἀνέγνωσαν τὸ φεῦδα ὡς ἐν παρολκῆι τοῦ da’ ἔνιοι δὲ ἀντὶ τοῦ φεῦ δή" τινὲς δὲ ἀντὶ
τοῦ φεῦ γῆ, κατὰ πάθος μεταβληθέντος τοῦ ÿ εἰς d, ὡς ἐν τῶι Δημήτηρ κτλ. It is
clear (cf. O. Gruppe, Griech. Mythol. ii. 1165 n. τὴ that the pronouncements
about the supposedly Doric word da or δῆ = γῆ (cf. also Hesychius s.v. δῆ,
Etym. M. 265. 54 s.v. Δημήτηρ) are based on mere inferences and cannot be
regarded as evidence that the grammarians anywhere found a word δᾶ or
δῇ meaning ‘earth’.” That there was in fact no Doric word δᾶ = yn was
established by H. L. Ahrens (De Graec. ling. dial. ii. 80 and Kl. Schr. 419)
as the result of careful examination of all the available evidence.? A summary
of the extensive modern discussion of the subject is given by A. B. Cook,
Zeus, iii. 8 f. It must be left to the philologists to decide what is contained in
the first syllable of dauarnp. But whatever opinion is held about Kretsch-
mer's hypothesis (Wien. Stud. xxiv, 1902, 523 ff.; repeated in Glotta, i, 1909,
28 ; V, 1914, 307 ; XVii, 1929, 240; and, with rather less assurance, xxvii, 1938,
31), which finds in 46 ‘an ancient "Lallname" of the goddess of the earth’,
there is no reason to believe that a Greek of classical times, when he heard δᾶ,
understood it as meaning 'earth'.* As far as the passionate cry in Ag. 1072
is concerned, and (to pass over the uncertain passage Prom. 568) the similar
exclamations in Zum. 842, E. Phoen. 1296, Ar. Lys. 198 (with comic exaggera-
tion), Ahrens, Dial. ii. 8o, is right when he says 'Terrae invocatio in interiec-
tionibus istis profecto mira esset' and Kretschmer has in no way detracted
from the force of this statement by referring to passages which are not
similar, A. Suppl. 890, E. Phoen. 1290, and to the well-known oaths sworn by
said that it is metrically equivalent to ἐήϊε παιάν (the refrain occurs in this form, e.g., in
Pindar's paean [2] for Abdera, cf. Wilamowitz, Sappho u. Simonides, 247 n. 1; cf. also the
refrain Ar. Thesm. 972 χαῖρ᾽, ὦ 'Exdepye, ὄπαζε δὲ νίκην), but that is probably due to mere
chance.
1 O. Schroeder, Grundriss der griech. Versgeschichte, 67, conjectures that Cassandra’s
"Απολλον "Ἄπολλον is perhaps derived from ritual invocations.
2 It is not allowable to draw any conclusions from that etymology of Δημήτηρ which fitted
extremely well into the theological system of the Stoics (Stoic. vet. fragm. ii. 305. 24 v. Arnim,
cf. Cic. Nat. deor. 2. 67) but was much older, cf. E. Bacch. 275 f., and also Phoen. 685 f. (see
Wilamowitz, Pindaros, 42).
3 Ahrens also protested that in this connexion no reference should be made to δάπεδον
(for the etymology of which cf. Boisacq, s.v., Schwyzer, Griech. Gramm. i. 358); in spite of
that, it appears again in the altogether unfortunate note of Wilamowitz on Ar. Lys. 198.
Ahrens later, Phil. 35, 1876, 21 = KI. Schr. 198, weakened the point of his earlier argument.
* It is a retrograde step after the position established by Ahrens when Kretschmer says,
Wien. Stud. xxiv, 1902, 524, that we must recognize a Doric 84 ‘earth’, even if this is not to
be equated etymologically with the Doric γᾶ, Wilamowitz, Glaube d. Hell. i. 202, was more
careful: ‘it cannot . . . be exactly defined in what place δᾶ was said for ya.’
490
 COMMENTARY line 1081

the Earth. So I believe that we must follow those commentators who, like,
e.g., Headlam in his prose translation (he has, however, in his verse transla-
tion ‘O Earth’) and L-S (‘prob. an exclamation of horror’), see in δᾶ nothing
but an exclamation. Such half-barbaric cries may have had their place at
Athens, for example in the ritual of the Carian mourning-women. On the
subject of non-Greek cries in lamentation for the dead and other ritual acts
cf. Wilamowitz, Verskunst, 28 f. [Cf. the Addenda.]
1075 (1079). It is a familiar fact that Apollo is the ἀπενθὴς θεός (the expres-
sion is used in schol. E. Phoen. 1028, which quotes this passage). In E. Alc. 22
the god withdraws, μὴ μίασμά μ᾽ ἐν δόμοις κίχηι. Peile compares Stesichorus
fr. 22 D. (xopev)uard τοι μάλιστα παιγμοσύνας Te φιλεῖ μολπάς τ᾽ Ἀπόλλων"
κάδεα δὲ στοναχάς τ᾽ Aidas ἔλαχεν. There is, moreover, the very significant
invitation to lament the dead in Sept. 854 ff. (where the phraseology is
complicated by reference to the journey to the festival of Delian Apollo,
but the thought ‘Apollo keeps away from the θρῆνος᾽ is clear) ἀλλὰ γόων...
κατ᾽ οὖρον ἐρέσσετ᾽ ἀμφὶ κρατὶ πόμπιμον χεροῖν mirukov, ὃς αἰὲν δι᾽ Ἀχέροντ᾽
ἀμείβεται τὰν ἄστολον (cf. Wilamowitz, Interpr. 84 τι. 1) μελάγκροκον θεωρίδα,
τὰν ἀστιβῆ Ans: (Παιῶνι H. L. Ahrens, Kl. Schr. 81, perhaps correctly).
τοιοῦτος ὥστε: cf. on 1421-1424. For τοιοῦτος with the first syllable short
cf. on 1256 f.
1078. For δυσφημεῖν this was the oldest example until recently. Now we
have the word in Hipponax, Pap. Oxy. 2174, fr. 9. 8.
1079. For the use of the participle προσήκων in the sense of ὧι προσήκει cf.
Debrunner, Museum Helveticum, i, 1944, 38, who quotes similar instances.
γόοις picks up θρῆνος (1075), cf. on 57.
1081. ἀγυιᾶτα: ‘then her eye falls upon the ἀγυιεύς, the conical pillar of
Apollo, which was really only the curbstone in front of every house, of which
the customary μὰ τὸν Ἀπόλλω τουτονί in Menander [and also, e.g., Ar. Thesm.
748] has reminded us again’ (Wilamowitz, Interpr. 173; cf. Wilamowitz,
Menanders Schiedsgericht, 10). Stanley's explanation, based on schol. Ar.
Wasps 875 and Hesychius s.v. ἀγυιεύς, was substantially the same. Apart
from Menander, additional evidence has come to light in the gloss Ayvieds
of the ‘Photius Berolinensis’, p. 25. 26 Reitzenstein: ὁ πρὸ τῶν αὐλείων θυρῶν
κωνοειδὴς κίων, ἱερὸς Ἀπόλλωνος, καὶ αὐτὸς 6 θεός. For other testimonies cf.
Ed. Schwartz in the apparatus on schol. E. Phoen. 631 (add the excerpts
from ancient scholia on the same passage in a sixth-century papyrus pub-
lished and explained by Wilcken, Preuss. Akad. Phil.-hist. Abhdl. 1933, no.
6, pp. το and r4), and Reisch, RE i. gro ff. where there is also a discussion of
the question whether in some of the relevant passages the reference is not to
an altar rather than the pillar. Cf. also A. B. Cook, Zeus, ii. 160 ff., iii. 1120;
M. P. Nilsson, Greek Popular Religion, 79 f. Georg Kaibel in a lecture note on
Ar. Thesm. 489, after reference to Steph. Byz. s.v. Ayuid and Hesychius,
observed: “This Apollo (image and altar together) is greeted by Cassandra
in Ag. vo8o'. This interpretation is worth considering. No conclusion, how-
ever, can be drawn from Ag. 509 (see the note there). Sophocles (El. 634 ff.)
places a statue and an altar of Apollo in front of the house of the Atridae.
Cho. 583 must not be adduced in this connexion.'
! Cho. 583 rà δ᾽ ἄλλα τούτωι δεῦρ᾽ ἐποπτεῦσαι λέγω is taken by Schütz, Wellauer, Sidgwick,
Wilamowitz as referring to Apollo, by others (e.g. Schneider, Conington, Tucker) as referring

491
line 1081 COMMENTARY

Hermann (Opusc. viii. 164 = Aesch. trag. ii. 652), followed by Wecklein,
Verrall, and others, supposed that Cassandra descends from the wagon after
1071. This is possible, but it is more probable that she does not do so until
1080-1 (cf. Wilamowitz, ‘Actio’), so that in moving towards the house she
notices that Apollo, always her lord and invoked as such in her first cry of
woe, is now present before her eyes in the form of his stone symbol. Gilbert
Murray, in his book Aeschylus, 221, thinks it possible that Cassandra leaves
the wagon either at 1071 or at 1177 ; he puts forward a psychological argument
which might be in favour of the latter alternative. But as soon as we try to
visualize the scene as part of a fifth-century performance, the assumption
that Cassandra leaves the chariot as late as ı177 proves untenable. The
contents as well as the form (dochmiacs, etc.) of Cassandra’s songs require
the most excited gestures and unfettered movements, not as an accompani-
ment but as an integral part of the delivery. Her whole frame shakes, she is
driven up and down by the horrors of the visions which hold her in their grip.
It seems inconceivable that all this could take place while she is ‘cowering’
(Murray’s word) on the wagon.
ἀπόλλων ἐμός: 'vocatival nominative after the pattern of Homer's
γαμβρὸς ἐμός, "my destroyer" ' (Wackernagel, Anredeformen, Progr. Göt-
tingen 1912, 6 n. 2). For etymologies in general in Aeschylus see on 687. The
etymology presented here occurs again in the Phaethon of Euripides (fr. 781.
11 f. N.) : ὦ καλλιφεγγὲς "HM, ὥς μ᾽ ἀπώλεσας kai τόνδ᾽ - Ἀπόλλων δ᾽ ἐν βροτοῖς
ὀρθῶς καλὴν κτλ. Stanley compared this fragment and also Archilochus
fr. 30 D. ἄναξ Ἄπολλον, καὶ σὺ τοὺς μὲν αἰτίους σήμαινε καί σφεας ὄλλυ᾽ ὥσπερ
ὀλλύεις ; but Wilamowitz, Glaube d. Hell. ii. 114 n. 4, referring to fr. 75, is
probably right in rejecting the idea that an etymological play upon the name
is intended in the passage of Archilochus.' There is evidence in Pl.
Cratylus 404d and 4oge that the etymology ArsAAwr= ‘destroyer’ was
widespread ; cf. Macrobius 1. 17. 9 (= Apollodorus of Athens, F Gr Hist 244 F
95. 10 Jac.).
1082. The meaning of οὐ μόλις, ‘non parum’, was established by G. Hermann
in his notes to Vigerus, De Graec. dict. idiot. 788, for Ag. 1082, Eum. 864, and
E. Hel. 334. So, too, Schneidewin-Hense: ‘non parum, sed funditus’. (The
scholion here is not helpful: où μετὰ καμάτου, similarly Wilamowitz on Eum.
864 : ‘ où μόλις est où μογερῶς '.) Stylistically this usage can be compared with
that of οὐχ ἅπαξ, cf. Headlam, On Editing Aeschylus, 43; W. Schulze, Kuhns
Zeitschr. lvii, 1930, 117 = Kl. Schr. 222.
τὸ δεύτερον: as she herself is later (1211 ff.) to explain.
1083. χρήσειν ἔοικεν : the Chorus infers this from Cassandra’s ecstatic atti-
tude; the further conclusion ἀμφὲ τῶν αὑτῆς κακῶν is derived from the words
to Hermes, by Otfried Müller, Weil, Wecklein, and Blass as referring to Agamemnon.
Blass seems to me to have proved finally that the last view is correct ; the agreement of 489
with 583 f. is conclusive. Nor should we neglect what Blass says about the impossibility of
finding the ἀγνιεύς in the neighbourhood of the grave. Wilamowitz, Interpr. 177 (middle),
makes a very rash statement, for 583 is placed before, not after, the stasimon, and therefore
certainly does not belong to the ‘second Act’. It is hard to understand how Headlam could
follow the stupid idea of the scholiast and refer τούτωι to Pylades. This is made improbable,
apart from anything else, by the usage of ἐποπτεύειν in Aeschylus (cf. below, on 1270).
1 Nor does the combination Ἄπολλον... ἀπόλωλα in Menander, Peric. 440, seem to be
significant.

492
 COMMENTARY line 1090

ἀπόλλων ἐμός κτλ. 1081 f. But the trance of the seeress takes a different
course: not until 1137 ff. is she obsessed by the visions of her own fated end
(there the words of the Chorus, 1140 f. ἀμφὶ δ᾽ aórás θροεῖς, hark back to 1083).
1084. Asyndeton explicativum. The form of the sentence is that of a general
maxim, but, as usual, it is applied to the case in hand.
τὸ θεῖον : this passage and Cho. 958 seem to be the earliest certain instances
of a conception which was afterwards to play a prominent part in Greek
thought. It must not, however, be overlooked that here the meaning is
limited to the faculty and inspiration derived from, and given by, the god,
divinatio. The sense seems to be very much the same when a soothsayer says
(Ar. Birds 965) τὸ θεῖον ἐνεπόδιζέ με.
δουλίαι περ ἐν φρενί: an emendatio palmaris.
1083 1. It is assumed that Cassandra’s gift of prophecy is known to the
Chorus, cf. 1098 and, in general, on 1035.
1087. Stanley’s idea that there is a play on the etymology of ἀγυιεύς (ἀπὸ τοῦ
ἄγειν) should not be resuscitated.
1090 ff. From here onwards apparet domus intus et atria longa patescunt;
“toutes les images qui se présentent à l'esprit de la voyante, nous les voyons
avec elle, comme si les murs du palais devenaient transparents (H. Weil,
Études sur le drame antique, 37).
1090. & & is found only in M, and many editors have followed Hermann in
omitting it. It is absent from the antistrophe (1095), and to insert it there
before the yap-clause (as e.g. Blomfield, Conington, Karsten, Paley do) is
hardly justifiable. Wilamowitz says: 'interiectio extra responsionem posita",
Such a position of the interjection is of course possible, but the arrangement
of the subsequent interjections in this scene (1100 = 1107, 1114 = 1125,
1126 — 1146, 1156 — 1167) is not in favour of accepting d d here in the strophe
only. Moreover, the following μὲν οὖν in its familiar corrective function!
appears more forceful if the correction follows immediately upon the words
of the coryphaeus and not after an intervening exclamation of horror. It is
therefore probable that d d in M is wrong; perhaps its insertion is due to an
anticipation of 1125.
μισόθεον: quoted by Schwyzer, Griech. Gramm. i. 442, as the earliest
example of compounds formed with μισο- (‘more frequent towards the end of
the fifth century’). Unfortunately Stanley translated dis invisam (similarly
Schütz) and, since this seems to suit the context well enough, it has misled
translators right up to the present time, e.g. Paley, Wilamowitz, Headlam
(‘abhorred of heaven’), Platt, Mazon, Murray, A. Y. Campbell. The correct
rendering is given by the dictionaries (from Henr. Stephanus onwards), and
among the translators by, e.g., W. Sewell (‘one that hateth God’), Conington,
Kennedy, L. Campbell, Lawson; cf. also Dirlmeier, Philol. xc, 1935, 185 n. 31.
The meaning of the word is clear from Lucian, Timon, 35 and Pollux 1. 21
as well as from the whole group of words compounded with μισο-, which can
be conveniently examined in the list in Pollux 6. 172. μισόθεος is ἃ stronger
equivalent of ἄθεος, which in Aeschylus and other poets of the fifth century
not infrequently denotes the criminal (cf. Latte, Archiv f. Religionswiss. xx,
1921, 264).
! Wilamowitz used to impress upon us as the stock example E. Phoen. 551 περιβλέπεσθαι
τίμιον; κενὸν μὲν οὖν,

403
lines 1090 f. COMMENTARY

1090 f. πολλὰ ouvioropa κτλ. For the government of the accusative by the
verbal element contained in συνίστορα cf., in addition to the commentaries,
P. Menge, De poet. scaen. Graec. serm. (Göttingen 1905), 93; W. Schulze, Kl.
Schr. 654; W. Ferrari, La parodos, 365; Schwyzer, Preuss. Akad. Phil.-hist.
A bhdl. 1940 no. 7, 13. Wilamowitz (who was followed by O. Schroeder, Aesch.
cantica, and Groeneboom in his commentary) assumed that Sept. 975 was to
be construed accordingly, i.e. he took μογερά as accusative dependent on
βαρυδότειρα ; but the usual interpretation (Μοῖρα Bapuô. poyepd) satisfies both
sense and metre.
(στέγην) συνίστορα : cf. on 664. On συνίστορα Daube, 6, observes: 'there is a
reference to the legal word iorwp, the witness who is to give evidence of what
he has seen personally.’
1091 ff. Cassandra's prophecies in this stanza refer quite generally to the
horrible murders of kinsfolk in the house of the Atridae. In this first vision
it is the slaughter of the occupants of the house that rises before her eyes,
but no separate crimes are singled out. Throughout the whole scene the poet
has worked out with great consistency the development of the visions and in
particular their progress step by step to more concrete and distinct images.
It is for this reason that there is no word in 1090-2 which is not completely
apposite to the plurality of murders, the scene of which was and will be this
house. Those commentators who find here a description of particular crimes
(most of them anticipate the Thyesteae dapes from what follows later) destroy
alike the veracity of the picture and the artistic structure of the scene.
After Emperiusit was Ahrensin particular who demonstrated emphatically,
p. 698, that ἀρτάναι has no place at all in this passage and that no reference to
Aerope! is to be looked for here and still less any reference to Hippodameia.
kapravaı therefore is certainly corrupt. Emendation depends in part upon the
view we take of the corresponding line of the antistrophe, 1096. There
Hermann rightly adopted with Scholefield τάδε βρέφη from M (in τὰ βρέφη
he recognized an attempt to equate the metre of the antistrophe with that
of the strophe) ; Ahrens's attempt to refute it is unfortunate. τὰ βρέφη would
refer to the children as to something already known, and that is unsuitable
here; τάδε, on the other hand, is excellent, uttered as it is at the moment
when this particular vision emerges, and its strong demonstrative force is not
rendered at all redundant by the preceding μαρτυρίοισι τοῖσδε. A short
syllable, then, is lacking in xogr. Of the conjectures brought forward Kayser's
καρατόμα is by far the best;? it is probably what Aeschylus wrote. Here,
where no specific, detailed horror has yet appeared, we need a typical? word
for murdering or slaying to go with αὐτοφόνα. E. Troad. 564 shows how suit-
able καρατόμος (OT καράτομος) is to fulfil this requirement.
αὐτοφόνα : Blomfield's explanation is good (and better than that of many
In any case there is nowhere any evidence for her suicide, cf. Knaack, RE i. 678, C. Robert,
Heldensage, i. 296 n. 3 (his treatment of Ag. 1091 is inaccurate and he misinterprets αὐτοφόνα).
* kal ἄρταμα is inadmissible because prosodically incorrect (cf. P. Maas, Gr. Meirik,
$129). This proposal does not originate from Headlam but from Ahrens (p. 609, ‘If the
reading τάδε βρέφη is preferred in the antistrophe, καὶ ἄρταμα can be read here"). Headlam
and G. Thomson either did not read Ahrens's articles at all or only did so very cursorily ;
his readings they took from Wecklein.
3 Therefore it is no objection to this reading that ‘beheading was not a prominent feature
of the Pelopid horrors’ (H. J. Rose, C.R. lvi, 1942, 71).

494
 COMMENTARY line 1092

later commentators): ‘qui se vel suos perimit’. In Aeschylus (cf. Sept. 850,
and for the adverb Suppl. 65)! only the latter meaning is found ;? L-S are
right. For the function of adro- the commentaries refer to similar compounds,
particularly to αὐτοκτόνος (of fratricide), Sept. 681, 734, 805 ; cf. below on 1573,
also Cho. 472 f. οὐδ᾽ an’ ἄλλων ἔκτοθεν, ἀλλ᾽ ἀπ᾿ αὐτῶν. Williger, Sprachl.
Unters. 5 n. x, recalls the fact that in A 395 someone with a name to fit his
nature (cf. the context) is called Aërôdovos, cf. on 443. For the use of verbal
compounds in groups such as αὐτοφόνα κακὰ καρατόμα Williger (op. cit. 27 ff.)
has collected and explained a rich store of parallels also from Aeschylus; the
meaning, therefore, is: ‘the evil of the slaughter of kinsfolk and the beheading’.
For syllaba anceps and hiatus here (xaparôua) cf. on 239.
1092. It is wrong to start from πέδον ῥαντήριον, which is clearly a simplifying
reading, instead of from the πεδορραντήριον in M (for the formation of the
word cf. Ernst Fraenkel, /dg. Forsch. xxxii, 1913, 131; Schwyzer, Griech.
Gramm. i. 470. 7). Therefore Dobree’s interpretation of the MS reading—it
can hardly be called a conjecture—avöpoodayetov (Dobree, Advers. ii. 25,
meant only to give Casaubon’s ἀνδροσφάγιον its due credit) is necessary. The
juxtaposition of two artificial ἁμαξιαῖα ἔπη of similar function is completely
in keeping with the practice of Aeschylus; cf. on 814. avöpood. and πεδορραντ.
are of course nouns. Paley well compares avöpoodayeiov with Phrynich.
Ῥ. 49. 14 de Borries ἀνδροκτονεῖον: 6 τόπος, ἔνθα où ἄνθρωποι ἀποθνήισκουσι κτλ.
Certainly πεδορραντήριον is ‘something with which’ (or ‘where’) ‘one be-
sprinkles the ground’, a bold coinage, and bold also is its use to describe
the house. But in company with ἀνδροσφαγεῖον, 'Menschenschlachthaus', the
sense of the word is clear, and its application to the στέγη is facilitated by the
fact that many nouns in -rYpıov designate the place where a particular action
is undertaken.? The effect of this gruesome line is striking. For the picture
suggested by πεδορραντήριον cf. A 420 (the murder of Agamemnon and his
comrades) δάπεδον δ᾽ ἅπαν αἵματι θῦεν, v 394 ff. καί vw ὀΐω αἵματί τ᾽ ἐγκεφάλωι
τε παλαξέμεν ἄσπετον οὖδας ἀνδρῶν μνηστήρων, and for ῥαΐνειν in this connexion
v 354 αἵματι δ᾽ ἐρράδαται τοῖχοι καλαί τε μεσόδμαι, Lycophron, 1104 (the mur-
dered Agamemnon) τιβῆνα καὶ κύπελλον ἐγκάρωι ῥανεῖ. Presumably Aeschylus
based his neologism on the word περιρραντήριον which was the name of the
vessel used for sprinkling holy water (cf. Ziehen, RE xix. 856 f.) and is found
from Hdt. 1. 51. 3 on and frequently in inscriptions; with it goes ἀπορραν-
τήριον of the same meaning, found first IG 12. 256 (anno 434/3) and 262 (anno
428/7), then in E. Jon 435. In this case there is more than mere philological
interest in tracing what is presumably the prototype of the fresh coinage.
The awful word πεδορραντήριον must have evoked in the mind of the audience

1 This αὐτοφόνως has been the subject of much misunderstanding, in which Wilamowitz
has a part, Aisch. Interpr. 29 n. 2, whereas Lobeck on S. Aj. 842 is correct.
2 In the ‘Sacred Laws’ of Cyrene (Inscr. Graec. ad inlustr. dial. sel. ed. Solmsen-Fraenkel,
1930, p. 61) B 7, line 50, αὐτοφόνος seems to have the general meaning of ‘murderer’ in the
description of a third category of ἱκέσιοι (ἱκέσιος τρίτος, αὐτοφόνος), cf. Wilamowitz, Berl.
Sitzgsb. 1927, 171; Latte, Archiv f. Religionsw. xxvi, 1928, 48.
3 Sometimes one and the same word in -τήριον denotes not only a place but algo an
implement; e.g. ψυκτήριον is synonymous with ψυκτήρ ‘cooling vessel’ but means also,in
passages from poetry (including Aesch. fr. 146 N.) quoted by Nicander of Thyateira
ap. Athen. 11. 503 c, ‘a place for ἀναψύχεσθαι ' (Wilamowitz, Berl. Sitzgsb. 1904, 637, cf. also
Ernst Fraenkel, /dg. Forsch. xxxii, 1913, 142).

495
line 1092 COMMENTARY

the name of a sacral vessel: the περιρραντήριον was the instrument of ritual
purification, whereas here the most monstrous defilement is meant. This
suggestion of sacred things in connexion with what is grim and horrible is
quite in the spirit of Aeschylus, cf. on 1144.
1093. The essential qualities of a good dog include that of being εὔρις, cf.,
e.g., [Xen.] Cyn. 4. 6; Poll. 5. 60; in an image in S. Aj. 8.
1094. Scholion in M : ἀναζητεῖ ei γέγονεν ἐνθάδε παλαιὸς φόνος. On the strength
of this Platt, J. Phil. xxxv, 1920, 332, conjectured ματεύει δ᾽ ἔνθ' and A. Y.
Campbell adopted this. The alteration is tempting because it provides some-
thing which seems at first to be needed here, cf., e.g., Wecklein’s paraphrase
‘she is not on the wrong track but is looking where she will find something’
and Headlam’s translation ‘she is hunting on a track where she will find
killing’. But these renderings (and accordingly the conjecture ἔνθ᾽) do not
sufficiently take into consideration the close connexion with the preceding
lines. The house of Atreus has appeared to the prophetess as the scene of
murders, the murders of kinsfolk. The Chorus does not make the comment
‘she is hunting in the right place’, but ‘(it is of murder among members of the
house that she has spoken), she is on the track of the murder of such people
whose murder she will discover’. Only the personal element in ὧν provides
the precise reference to Cassandra’s words.
1095. Fr. Leo, Monolog im Drama, 8, rightly says: ‘The Chorus is not ad-
dressed in Cassandra’s monody, as might be supposed from its singing in
between ; but she in her ecstasy is conscious of its presence (1095 ydp, 1129).’
This is clear, too, from 1090 μὲν οὖν. True enough, Cassandra is engrossed in
her vision, and cut off from her surroundings, but this aspect of the situation
is exaggerated by Pluss when he explains γάρ as ‘confirming her own thoughts’
and comments rather arbitrarily on σοί 1129 ‘addressed to some listener or
other ; the actual vision is past’.
ἐπιπείθομαι (a certain emendation): an epic word, otherwise very rarely
used ; in drama also in S. El. 1472.
1096. For the text cf. on 1091 ff. κλαιόμενα τάδε βρέφη was taken as accusa-
tive by Stanley (‘subintellige video’), similarly by Hermann (on Ag. 2:
‘intellecto tali verbo, quo significaretur in mente habeo'). Elmsley, too, on
E. Heraclid. 693, took κλαιόμενα κτλ. as accusative, but regarded it as in
apposition to μαρτυρίοισι, which seems impossible. Elmsley’s theory that
‘non raro fit transitus e genitivo aut dativo ad accusativum’ should be dis-
missed ; in favour of it he quotes the very doubtful (cf. on Ag. 2) passage
S. Ant. 857 f. and Aj. 872, where his punctuation and interpretation are
wrong. Lobeck, on S. Aj. 136 (ed. tertia, p. 108 n. 1), while admitting the
possibility of the accusative as explained by Stanley, considered the nomina-
tive as an alternative: ‘utique ὁρῶσα suppleri potest, sed et μαρτυρεῖ et alia.’
It seems, however, most unlikely that μαρτυρεῖ should be understood here.
Finally, a warning should be issued against the supposition that in κλαιό-
peva . . . βρέφη we have an instance of the somewhat loose nominative which
in certain circumstances may be used in apposition to a preceding casus
obliquus (cf. Havers, Glotta, xvi, 1928, 105 ff.; E. Lofstedt, Syntactica, i,
2nd ed., 1942, 81 ff.). That usage, common to many languages, is indeed to
be found in Greek also, but, as far as I can see, there are no instances of it in
Attic poetry. Are we then to accept Stanley's (scil. video) or Hermann's
496
 COMMENTARY line 1099

(scil. in mente habeo) view? Hermann, after producing three (including Ag.
1096) highly questionable instances of accusatives where, according to him,
something like in mente habeo should be understood, continues: ‘sed quid
exemplis opus est in re pervulgata?’ I cannot share this optimistic belief,
and though I do not regard the accusative as absolutely impossible, I would
prefer to take κλαιόμενα τάδε βρέφη as nominative and translate it, with
Klausen, Peile, Conington, Paley, Wecklein, Plüss, and others, ‘(these) here
are infants ....’ She points to the vision which now appears to her. The
‘asyndeton explicativum’ would be normal even in a calmer speech; here,
where the prophetess is overwhelmed by a sudden apparition, such abruptness
is particularly suitable.
It was to be hoped that Hermann had once and for all disposed of Vic-
torius’ punctuation βρέφη, σφαγὰς; but it still persists in Wilamowitz’s
edition. The idea of the children weeping for their own slaughter should at
once be recognized as genuine ; Hermann referred to the parallel vision 1219 ff.
The lexicons quote 5, Trach. 153 for κλαίομαι with an object in the accusative.
1098. There is no room for doubt about the text. The reading of M Fur
taken as ἦ μὴν has certainly found favour with some commentators, but
Paley, who puts it into the text and at first defends it, goes on to say: ‘7
μὴν however is not usual in this kind of asseveration' ; Wilamowitz comments
with more vigour: 'gravissima illa affirmatio non convenit', and a glance at
the examples quoted by Denniston, Particles, 350 f., confirms this. If, on the
other hand, we adopt uev' the text (apart from the first two syllables in
1099, see below) is unobjectionable. The relation between the two sentences
in 1098 f. is clearly adversative or concessive. When this is the case, the
absence of a μέν in the first sentence and the emphatic position of the verb
at the beginning may sometimes be observed in this style, e.g. Ag. 1401 f.
πειρᾶσθέ μου γυναικὸς ὡς ἀφράσμονος, ἐγὼ δ᾽ ἀτρέστωι καρδίαι πρὸς εἰδότας
λέγω and, with a continuative 8é-sentence, Suppl. 950 f. ἔοιγμεν ἤδη πόλεμον
ἀρεῖσθαι νέον" ein δὲ νίκη καὶ κράτος τοῖς ἄρσεσιν. So it is probable that 1098 is
sound and the beginning of 1099, 1.6. the repetition of ἦμεν, corrupt. How-
ever, the alternative that the first word in 1098 is the seat of the corruption
cannot be completely excluded. No really convincing emendation has been
made; it does not seem possible to go beyond the diagnosis of Wilamowitz,
‘aut hic (1098) aut in versu proximo dittographia vocem aliam expulit’.
1099. As is well known, προφήτης does not primarily denote one who
‘prophesies’, i.e. foretells future events. The word, akin to προειπεῖν (edicere)
and the like (cf. on 201), means literally ‘pronouncer’. In actual use, however,
the general meaning was narrowed, and so the word signifies in the main a
person who makes pronouncements for or on behalf of a god, acting as his
spokesman or mouthpiece. When, as is often the case, a god manifests his
1 With regard to the value of the variae lectiones in the Mediceus cf. Blass, Introd. to his
commentary on the Choeph., p. 22 f., where the habit of putting a dot at either end of each
variant (as e.g. Ag. 1098 "ἦμεν", 1101 *dyos-) is discussed ; examples of the same practice from
the text of Sophocles in this codex (L) are quoted by T. W. Allen, J. Phil, xxii, 1894, 171.
2 Cf. Wackernagel, Syntax, ii. 239 f. His results, as far as the etymology and the earlier
history of προφήτης are concerned, have not been substantially modified in the elaborate
monograph by E. Fascher, ΠΡΟΦΗΤΗΣ (Giessen 1927). Fascher’s treatment (p. 14) of
A. Ag. 1098 f. is inconclusive. For the general notion of προφήτης cf. also O. Kern, Religion
der Griechen, ii. 112.
4872.3 C 497
 line 1099 COMMENTARY

will by means of prodigia or σύμβολα or some other veiled indications, the

prophet’s ‘pronouncing’ must of necessity include an interpretation of the
heaven-sent tokens. But even so the original meaning of προφήτης persists
wherever it occurs in earlier literature. It is not synonymous with μάντις.
The function of the μάντις and that of the προφήτης may fall to different
persons (Pindar fr. 150 Schr. μαντεύεο, Μοῖσα, προφατεύσω δ᾽ ἐγώ). But when
one and the same person is both μάντις and προφήτης, then two different
aspects are expressed by the two words. So in Eum. 17 ff. it is said of Apollo:
τέχνης δέ νιν Ζεὺς ἔνθεον κτίσας φρένα ἵζει τέταρτον τοῖσδε μάντιν ἐν θρόνοις"
Διὸς προφήτης δ᾽ ἐστὶ Λοξίας πατρός. No doubt Apollo is μάντις but at the
same time he is προφήτης of Zeus in that he serves as his mouthpiece and the
interpreter of his will. Exactly the same double function can be ascribed to
a human seer: Pind. Nem. 1. 60 f. ἐκκάλεσεν Διὸς ὑψίστον προφάταν ἔξοχον,
ὀρθόμαντιν Τειρεσίαν, A. Sept. 609 ff. 6 μάντις (Amphiaraus) . . . μέγας προφήτης.
In the present passage, too, μάντις and προφήτης, without having the same
meaning, refer to one and the same person. In 1099 προφήτας is of course a
generalizing plural, ‘people like you, the προφῆτις᾽ (the inept paraphrase of
the scholiast, τοὺς λέξοντας (Victorius, λέξαντας M) ἡμῖν περὶ σοῦ. αὐτοὶ yap
αὐτόπται γινόμεθα, rejected by Ahrens p. 612, was again adopted by Verrall).
The progress of thought is clear. By acknowledging Cassandra’s fame as a
μάντις, an inspired seeress, the leader of the Chorus admits implicitly that the
vision in which the slaughtered children of Thyestes appear to her corre-
sponds to therealevent. But the Elders do not wish the horrors of the past to
be dragged into the light. Cassandra might at any moment proceed to reveal
the details of that crime; so the old man attempts to stop her.
The plural in 1098 f. is remarkable. Elsewhere in this scene the Chorus
speaks of itself almost throughout in the singular: 1089, 1105, 1112 f., 1120,
1130 f., 1164, 1166, 1177, 1199, 1243 ff., 1253, 1321. Except in 1098 f. the plural
only occurs in 1213 ἡμῖν γε μὲν δὴ πιστὰ θεσπίζειν δοκεῖς. It is possible that in
both passages the ‘we’ means to include with the speaker his fellow citizens
of Argos, those who as the witnesses of past crimes live in fear of crimes to
come. But perhaps it is only a casual variation.
ματεύομεν. 'Schützius in prima editione recte ματεύομεν. Videntur hac
forma usi esse tragici ubi metrum primam syllabam corripi pateretur.’ So
Hermann, whose cautious mode of expression makes it clear that no certain
decision is possible. What is certain is first that the form with σ is several
times found in the MSS where the metre requires a short syllable, e.g. S. Εἰ.
1107 (L), E. Phoen. 416 (MABV), secondly that Euripides at any rate used
both μαστεύω (e.g. Hec. 754, Phoen. 36) and ματεύω, as did Pindar.
1100 ff. The subject of μήδεται remains concealed until it emerges in the
antistrophe (1107 ff.). The prophetess possessed by the vision knows who it is
she sees before her ; for the listeners, however, (in the first place the old men
of Argos) the uncertainty! makes the sinister effect of these sentences even
more painful. Cf. on 681 ff.
As regards the punctuation, i.e. the interpretation of the structure, of the
1 In some modern languages, e.g. in English and German, it is hardly possible to retain
the suspense of the riddle without using the passive, where the Greek has the deponent.
Le Pére Brumoy, however, (he was justly followed by Mazon), found an ingenious way out
of the difficulty : ‘Ah dieux! que prépare-t-on?'
498
 COMMENTARY line 1101

sentence (or sentences) from 1100 ἰὼ πόποι to 1104 ἀποστατεῖ, the editors vary
a great deal. One point, however, can be settled without much ado: κακόν
(1102) cannot possibly be regarded as an attribute on the same footing with
the much stronger expressions ἄφερτον φίλοισιν and δυσίατον, nor can it go
with ἄχος and thus, as an adjective, enter into competition with uéya. Conse-
quently κακόν is to be taken as substantive, to which μέγα forms an epithet of
the normal type, while the attributes ἄφερτον φίλοισιν and 8voíarov follow to
provide additional weight. It is less easy to determine whether or not the
greater part of Cassandra's stanza consists of interrogative sentences. Many
editors continue the second question down to δυσίατον. But the structure
seems to gain in force if only the words τί τόδε νέον ἄχος are taken as interro-
gative, and the remainder, μέγα... δυσίατον, as the answer. This is also
supported by the parallel arrangement in the stanza 1114 ff.; moreover, it
may perhaps be argued that the clause ἀλκὰ δ᾽ ἑκὰς ἀποστατεῖ attaches more
smoothly to a statement than to a question. The punctuation which I have
adopted is substantially that of Kennedy, Margoliouth, and Wilamowitz.!
μέγα... μήδεται κακόν. In y 261 we read of Aegisthus murdering Agamem-
non μάλα yàp μέγα μήσατο ἔργον. Here as elsewhere Aeschylus, while adopting
a Homeric expression, slightly modifies it. There is no reason to suppose
(with Papageorgiu, E. Bruhn, Introd. to his edition of Soph. El., p. 48, and
others) that in μήδεται any reference is intended to the etymology of KAvra-
μήστρα (cf. on 84) which was accepted by the grammarians in later antiquity.
Bacchyl. 16 (15). 3o (of Deianeira, who becomes her husband's murderess)
d δύσμορος, & τάλαιν᾽ (cf. Ag. 1107), olov ἐμήσατο; Perhaps a mere coincidence.
1101. Enger excised μέγα and correspondingly πόσιν in 1108; he was followed
by Wecklein (in his commentary), Weil (Teubner edition), Blaydes, and G.
Thomson. This provides a smooth text and an acceptable metre (iamb. 2
dochm.). Enger thought the reason for the additions was to be found in the
intrusion in 1108 of πόσιν, a gloss on ὁμοδέμνιον. The force of this argument
must be admitted, for it appears that words like ὁμευνέτης (-ris), ὄμευνος,
ópóAekrpos, σύλλεκτρος, συνευνέτης (-ris), σύνευνος are not generally used as
epithets with words meaning 'consort, husband, wife', thus expressing the
same idea twice (E. Or. 508 γυνή means ‘woman’, not ‘wife’). But the repeti-
tion of uéya? need not give rise to any doubt although it is known that 'the
repetition of the same word in Aeschylus occurs comparatively seldom’
(Wecklein, Stud. z. Aesch. 51; but in his collection of passages Eum. 782 ἰὸν ἰόν
is missing). For the double uéya, which stresses the extent of the evil, we
may compare Pers. 980 μυρία μυρία πεμπαστάν. Finally we come to the
question of metre. If it were necessary to take (as Wilamowitz does) τί τόδε
νέον ἄχος μέγα as an iambic dimeter, responsion with τὸν ὁμοδέμνιον πόσιν
would seem doubtful, as v ὦ — would have to correspond to vu ὦ vo? and
this has no certain parallel in the iambics of the choruses in the Oresteia
! Paley and Murray differ from the above-mentioned editors only in that they treat the
first μέγα as part of the question and begin the answer with the second μέγα.
? Paley and Murray put a question-mark after the first μέγα and then a colon after
Bvoíarov, so that the words from μέγ᾽ ἐν δόμοισε down to δυσίατον must be taken as the
answer. This seems to me most improbable.
3 It might seem possible to get rid of the inexact responsion by transposing the words in
1101, Le, τί τόδ᾽ ἄχος νέον, μέγα. This was considered by Hermann, but he justly rejected it,
because ‘ea minus apta est verborum collocatio".

499
line 1101 COMMENTARY

(cf. p. 351). But although the dochmiacs of the Cassandra scene are freely
interspersed with iambics, it is not certain whether vy o GO v -- vg, which
Schroeder correctly describes as a lecythion, is to be taken as iambic. In
Sept. 239-41 we have τίμιον &8os ἱκόμαν as the clausula of four dochmiacs ;
perhaps the doubling of the lecythion in Eum. 782 ἰὸν ἰὸν ἀντιπεν θῇ μεθεῖσα
καρδίας (also among dochmiacs) may be added. To sum up: no decisive
objection has been raised to the MS text, but there seems to be some proba-
bility in favour of Enger’s deletion here and in 1108,
1103. ἀλκὰ δ᾽ ἑκὰς ἀποστατεῖ is a stronger, more imaginative and lively
variant of οὔτις ἀλκά (467). The flat suggestion of the scholion τὸν ’Opdorny
φησί has found supporters up to recent times; against this Wecklein rightly
remarks: ‘there is no reference to any particular persons.’
1106. ἐκεῖνα : this is the object of the next clause also.
1107. ἰὼ τάλαινα: after the riddle of the preceding stanza the solution is
partly given in the exclamation and carried on in the following lines, cf.
on ııoo ff.
1107 ff. For examples of this use of yap in questions preceded by an exclama-
tion see Denniston, Particles, 80, under8.
Robortello and others rightly put a question-mark after τελεῖς, cf. Hermann.
After φαιδρύνασα Blomfield put a dash (a colon might do as well) to make it
clear that the main verb, which we should expect to come next, is supplanted
by the question πῶς φράσω τέλος ;
1108. For the text cf. on r101.
1109. Here and in the following lines Cassandra describes in sharp outline
everything that she discerns in her visions, but she does not go a hair’s
breadth farther. The prophetess sees the woman moving about as she assists
with the bath, but does not yet know what her further intentions are. πῶς
φράσω τέλος does not mean that she has difficulty in finding the right expres-
sion: the τέλος at which Clytemnestra’s actions are aimed is not yet per-
ceptible to Cassandra when she speaks these words, however much she may
fear the outcome of this action (1107 τόδε) which she sees before her.
λουτροῖσι φαιδρύνασα: Clytemnestra herself attends her husband in the
bath, as women and girls sometimes do in epic—even a king’s daughter.
This Homeric feature is the basis of the particular form of the murder in this
play, cf. on 1382. Here, too, (cf. on 245 ff.), Aeschylus is conscious of the
difference between the customs of the ‘Homeric’ world and those of his own
society. Wilamowitz in his commentary on Cho. 6 has shown that this is
probably true also of Orestes’ sacrifice of a lock of hair to the river-god.
φαιδρύνασα: the grammarians seem now to agree that the correct form of
the verb is φαιδύνω (cf. φαιδυντής), cf. Solmsen, Rhein. Mus. liv, 1899, 495 f.;
Ernst Fraenkel, Griech. Denominativa, 66; Schwyzer, Griech. Gramm. 1. 733.
In the earliest extant example the verb is used, as here, with reference to a
bath, Hesiod, Evga, 753.
1110. γάρ explains why she speaks of a τέλος. ‘For what I now see her doing
is not all: completion will come quickly.’
1111. épeyouéva, the reading of M (dpeyueva in FTr is probably a metrical
correction), is metrically faultless, cf. 1121 ἔδραμε κροκοβαφής = 1132 Tis ἀγαθὰ
φάτις. But προτείνει χείρ without an object, as read by several editors, appears
very doubtful; at any rate no satisfactory parallel is provided by the absolute
500
 COMMENTARY lines 1112 f.

use οἱ προτείνειν in geographical descriptions as in Plat. Critias, 111a μακρὰ

προτείνουσα eis τὸ πέλαγος and in Polybius (cf. Thesaurus, vi. 2057 B). Other
editors follow the accentuation of the corrector of M, χεῖρ᾽, so that Clytem-
nestra becomes the subject. Grammatically this is probably correct, for the
Middle, though uncommon with an object of this kind, does occur, cf. 2 506
ποτὶ στόμα χεῖρ᾽ ὀρέγεσθαι (correct in spite of the doubt expressed by Leaf
following earlier critics), Leonidas, Anth. Pal. 7. 506. 6 ναύταις χεῖρας ὀρεγνύ-
μενος. It is conceivable that the poet wrote χεῖρ᾽ ex χερὸς ὀρεγομένα. But the
alternative which is at our disposal is decidedly preferable. The scholion
runs: διαδέχονται δὲ ἀλλήλους τοῖς ὀρέγμασι τῶν χειρῶν Αἴγισθος καὶ Κλυταιμή-
στρα (misrepresenting the sense of the passage, see below). Supported by
this piece of evidence Hermann conjectured ὀρέγματα, which produces a fine
expression of true Aeschylean colour. Apart from this line ὄρεγμα occurs
for the first time in two passages of the Choephoroe. Since προτείνειν and
ὀρέγειν are very close to one another in meaning, ὀρέγματα serves as a cognate
accusative.
χεὶρ ἐκ χερός is one of the set phrases for describing an uninterrupted
sequence, examples of which are to be found in L-S s.v. ex I. 3, e.g. as early
as Homer T 290 ὥς μοι δέχεται κακὸν ἐκ κακοῦ αἰεί.
The scholiast (see above) referred the sentence προτείνει κτλ. to the murder
of Agamemnon (in which he actually makes Aegisthus take part). Similarly
Stanley comments: ‘Extenditur manus post manum appetens feriendi.’ This
gross misinterpretation has not completely disappeared yet, although
Klausen, Ahrens, and others have long since explained the passage correctly.
Paley translates ‘to reach him by repeated blows’, similarly Sidgwick (‘blow
on blow comes’), L. Campbell (‘blow after blow she deals at him’), Mazon
(‘se tendent pour frapper’). It appears that many readers are completely
blind to the arrangement, and consequently to the very spirit, of this scene.
They mingle together heterogeneous elements taken from its different parts,
neglecting the unmistakable fact that in Cassandra’s visions the doings of
Clytemnestra are revealed in successive stages, and that each stage is kept
strictly separate from the next. At the stage represented here (1107-11) there
can be no idea of blows; only much later does the τύπτειν take place. The
sentence προτείνει κτλ. describes nothing but the movements of Clytemnestra
as she is attending her husband in his bath. In using the present tense
(προτείνει) Cassandra returns from the possibilities of the future to what her
vision shows her now.
1112. “συνιέναι in Aeschylus means the understanding of words which have
a concealed meaning’ (Snell, Phil. Unters. xxix, 1924, 47), cf. 1243, 1253, Cho.
887, Suppl. 467.
1112 f. viv yàp .. . ἀμηχανῶ. Most editors follow Hermann, in the main, in
understanding : ‘nunc enim, si ante visa est aenigmata dicere, magis obscura
loquitur’, ie. take ἐξ αἰνιγμάτων in a temporal sense. Hermann’s introduction
of the comparative is arbitrary; a more faithful rendering is given by e.g.
Headlam, whose interpretation is substantially the same: ‘it was riddles
before ; now it is obscure blind oracles that perplex me.’ But this is not very
convincing, for the antithesis between aiviypara and ἐπάργεμα θέσφατα is
pointless here, where the whole context demands that the mysterious and the
obscure and the puzzling should be grouped together in contrast to what is
501
lines 1112f, COMMENTARY

clear and comprehensible. The interpretation of Hermann and Headlam

becomes even more doubtful when we consider Cho. 887 ξυνῆκα τοῦπος ἐξ
αἰνιγμάτων. This suggests that in our present passage ἀμηχανῶ (the opposite
of the preceding ξυνῆκα) is to be taken with ἐξ αἰνιγμάτων. The source of
comprehension (in this case lack of comprehension) is indicated by ἐκ (cf.
Conington's comment, who rightly compares 268) and the dative ézapy.
θεσφάτοις is added, just as σφαλῆναι, ἐξαπατᾶσθαι, and similar verbs are used
with the dative (Kühner-Gerth, i. 438). The sentence may seem to be unduly
compressed, but there are many such in Aeschylus; linguistically there can
be no objection to the phrase. Earlier commentators took ἐξ αἰνιγμάτων
closely with ἐπαργέμοισι (‘oraculis propter aenigmata verborum caecatis’) ;
but this seems to spoil the self-contained phrase ἐπάργεμα θέσφατα, which
requires no explanation and is not capable of being supplemented. νῦν
develops the idea of οὔπω: ‘I have not yet understood it, for now at the
moment I am... quite at a loss with the obscure oracles’. The sentence was
correctly understood by e.g. Conington, L. Campbell, Platt.
1113. ἐπαργέμοισι (the word occurs also in Cho. 665, Prom. 499) : schol. τοῖς
τυφλοῖς καὶ ἀσήμοις, ἀπὸ τῶν λευκώματα ἐχόντων ὀφθαλμῶν (the scholia on the
other two passages are similar). The disease of the eyes to which reference
is made here, ἄργεμον, is mentioned in the Corpus Hippocraticum (the in-
stances are quoted by Aly, see below) ; cf., too, Erotian, 56, p. 18. 23 Nachm.
äpyenov' πάθος τι περὶ τοὺς ὀφθαλμοὺς λευκωματῶδες, ὃ δὴ ἐκ τῆς παρεπομένης
λευκότητος ὠνομάσθη. W. Aly, De Aesch. cop. verb. 23 f., rightly concludes that
ἄργεμον was properly an Ionic medical term and that Aeschylus borrowed his
expression from it. But he ought not to have regarded ἐπάργεμος as a word
freshly coined by Aeschylus. This assumption is contradicted by the use of
the adjective in Aristotle, where it occurs in a literal and technical sense but
does not denote a disease. The reference is given in all the dictionaries but
mentioned neither by Aly nor by Bechtel, Griech. Dial. iii. 281. Aeschylus
has clearly taken over ἐπάργεμος from technical language and he provides us
here with an addition to our defective vocabulary of the oldest Greek medi-
cine (or, if the use of the word in Aristotle is the original one, of early Ionic
physical science).
1115. δίκτυον... Abou: cf. 1235 Ardov μητέρα.
y’ was excised by Dindorf, and he was followed by Weil and Wilamowitz.
I can find no example of hiatus after the indefinite r..! ἧ... “Ἅιδου, it is true,
is added as a further question (Ahrens and others are wrong in taking 7...
“Αἰδου affirmatively), but although given in interrogative form, it contains
in fact an answer to ri τόδε φαίνεται; this may perhaps explain the addition
of γε.2 But possibly we need not look for any special reason, since ‘in interro-
1 Wilamowitz, Verskunst, 99 n. 4, 409 n. 1, defends the conjecture proposed for S. Phil.
699 εἴ τι (τις MSS) ἐμπέσοι; he asserts that hiatus after 7: is legitimate in Sophocles (the same
erroneous statement is made by him in Sappho und Simonides, 88 f.). The case of ὅτι or ὅ τι,
after which (in comedy) hiatus may be allowed, proves nothing of course for rı alone. As far
as Attic drama is concerned, the main facts were extremely well set forth by Buttmann,
Griech. Sprachlehre, i, 2nd ed., 112 (§ 29. 1 n. 1).
2 Headlam’s reference to Cho. 418 f. is not to the point, for there ye does not, as it does
here, form part of the expression which denotes the main object of the question but belongs
to an additional clause. Neither should Cho. 994 ff. be quoted as a parallel, since the struc-
ture and interpretation of the passage are far from clear (see p. 813 f.).
502
 COMMENTARY line 1116

gative or exclamatory sentences, the ''pathetic" meaning of ye is natural, and

we find it in combination with dpa and 7’ (Neil, The Knights of Aristophanes,
198 ; he adduces a number of examples of dpd ye and dpa... ye). For... γε
Denniston, Particles, 125, quotes E. Cycl. 207 ἦ πρός γε μαστοῖς εἰσι krÀ., where,
however, Hermann, Paley, and Wecklein, following L. Dindorf, read re for ye.
1116. σύνευνος elsewhere is always used properly of the woman who sleeps
with the man (as in Ag. 1442) or more seldom of the man who sleeps with the
woman (as in Prom. 866). Accordingly the scholia here say: ἡ σύνευνος τοῦ
Ayayueuvovos αἰτία (this involves a notable and, once this interpretation of
ovvevvos is accepted, unavoidable ignoring of the special meaning of ξυναιτία)
ἔσται τοῦ φόνου. In accordance with this paraphrase almost all commentators
from Stanley up to the present time, unless they have altered the text,’ have
taken ἡ ξύνευνος (Clytemnestre) as subject and ἄρκυς as predicate (e.g.
Nägelsbach : ‘Aber die Gattin ist das Netz, die Mordhelferin !’, Headlam: ‘But
she is a snare that shares the bed, that shares the guilt of murder’, Mazon:
_ ‘Mais non, le vrai filet, c'est la compagne de lit devenue complice de meurtre!’).
Schutz protested against this; the substance of his own interpretation was
adopted by Bothe and Meineke, Phil. xix, 1863, 206.” The grammatical
correctness of the more generally accepted interpretation (“The wife is the
net’) has to be admitted. But, as Ahrens points out, who considers the pas-
sage corrupt (p. 614), the sense is ‘incredible, after the entangling πέπλος had
just been called a δέκτυον “Aiôov’. Nor is this all. The commentators do not
seem to have noticed what is in fact a fatal objection. Cassandra, surrender-
ing herself to her visions, endeavours to comprehend what is appearing before
her. ri τόδε φαίνεται: she is concerned with this only and with nothing else.
At the time when she puts the question, she has before her nothing but a
vague shape, while the nature of the object is not yet perceptible even in
broad outlines. There follows a stage in which the vision becomes much
more substantial: she believes, though still in doubt, that she can see a net
of death. In the next stage the vision must take substance to a still greater
degree—this is certain, for the poet’s conception is consistent and does not
admit of any break. It is inconceivable that Cassandra, at the moment when
she is struggling to depict the vision that is before her, should turn aside
from her description of the φαινόμενον and say: “The wife is the (or ‘a’) net.’
At this stage Cassandra concentrates entirely on finding out the shape and
character of the objects of her visions. She tries to delineate them as accur-
ately as she can, but makes no attempt to go further and interpret them by
stating ‘this thing stands for that’ or ‘the meaning of the thing in front of me
is this’. The real gist of her thought is easily recognized. ‘Some net of Hades ?
1 I do not need to say anything against the conjectures proposed by various scholars or
against Wilamowitz’s treatment of the passage, Interpr. 173 n. 1, as I hope to demonstrate
the correctness of the MS reading. In no case are we allowed to introduce the ‘axe’ here;
see Appendix B. .
2 Long after completing this section of my commentary I came across Lewis Campbell’s
prose translation of the Oresteia (London 1893) and the treatment of the passage in his essay,
Am. Journ. of Philol. x, 1880, 436 f. His interpretation (‘Is it a hunting-net of hell? Nay, it
is the snare that surprises him in his repose, that hath part in slaying him") is quite close to
my own. The interpretation of E. Petersen, also, Rhein. Mus. Ixvi, 1911, 31 f., seems to me
in the main correct.
3 Here perhaps a special warning is still desirable, since the unfortunate idea which
Conington put forward on Cho. 997 has been reproduced by Verrall and again commended
503
line 1116 COMMENTARY

Rather the net that is a bedfellow, that is in part responsible for the murder.’
At first she sees some net or other, then it becomes the particular ‘net’, the
garment that will entangle Agamemnon (Cho. 981 f., 998 ff., Eum. 634 f.) and
help to cause his death. We can clearly see how appropriately this ἄρκυς is
called ξυναιτία $óvov.! But Evveuvos is more difficult. There is nothing in the
trilogy to support Schiitz’s idea: ‘mihi quidem post ξύνευνος tacite supplen-
dum videtur ἐσθής, veste enim dormitoria irretiebatur Agamemno’ (for a
variation of this idea see Meineke, loc. cit., ‘of the blanket covering the mar-
ried couple, the frequently mentioned pia χλαῖνα, ὑφ᾽ ἣν of ἐρῶντες ἔπεσον᾽
etc.). To dispose of this interpretation we have only to remember that the
‘net’, the πέπλοι (1126, 1580, and also Eum. 635 δαιδάλωι πέπλωι), the εἶμα
(1383) or φᾶρος (Cho. 1011, Eum. 634), must mean the ceremonial garment
which Agamemnon was to put on after the bath and wear afterwards at the
festal banquet, if the normal course of events (see on 1382) had not been
broken off by the murder.” There is another reason for rejecting the inter-
pretation of Schiitz: the close parallelism between ἡ ξύνευνος and ἡ ξυναιτία
φόνου shows that the reference is the same in both, i.e. that ξύνευνος implies
participation in the murder. I know of no self-evident solution of the puzzle.’
Though it is far from certain, it should perhaps be considered as a possibility
whether the idea of ‘the embracing one’ could not be found in σύνευνος, which
normally means ‘wife’. Agamemnon cannot free himself from the embraces
of the ἄρκυς. But it must be admitted that the notion of embracing is not
clearly brought out by the word σύνευνος, and it is doubtful whether the
obscurity of prophetic speech is sufficient explanation. So long as we do not
completely understand from what point of view the garment is regarded as
advevvos, we cannot exclude the possibility of a corruption in the text. I had
thought of £/vepyos and then found it in Wecklein's appendix as a suggestion
of Risberg. But I am all but convinced that ξύνευνος, though I am unable to
grasp its implications fully, is what Aeschylus wrote, principally for the
reason that it once again brings together in close relation the two main

by Murray in the Oxford text. According to Conington’s view Orestes, in the scene of the
Choephoroe, identifies Clytemnestra with the ‘net’, the cloak of murder. ‘This identification
is doubtless a symptom of the frenzy which is beginning to work on him, at the same time
that it has its own imaginative truth. Precisely the same identification is made by Cas-
sandra, Ag. 1114 ff.” For the impossibility of this interpretation see p. 810 f. From
what has been said above it appears that W. Porzig, Atschylos, 139, is quite wrong when he
asserts: ‘At the beginning the prophetess is still capable of interpreting her visions; she
knows that the net is the wife (Ag. 1116). Then the floods carry her away’, etc. It is
significant that out of the whole scene Porzig could find only this one misinterpreted phrase
to support his view.
1 ] will not discuss the artifice of putting a stop after ξυναιτία (Weil [1858] and others). It
should be noticed that the end of the sentence after φόνου corresponds exactly to that after
τύπτει in the antistrophe (1128).
2 From the remarks below on 1382 it follows that E. Petersen’s interpretation (Rhein.
Mus. lxvi, 1911, 32) is wrong: ‘the great cloth or cloak in which the bather wraps himself
or is wrapped, not for drying, but for keeping warm and in which he is to rest on the
εὐνή ’; Petersen compares the procedure in a Turkish bath. The mistake of taking the
garment as ‘the insidious bath-cloak of Agamemnon’ was also made by Otfried Müller,
Aesch. Eum. § 98, p. 195.
3 I have thought of referring ξύνευνος to the use of the garment later for covering Aga-
memnon’s dead body (1492, 1580 f.), but that would point to a time too far ahead of the
events which fill Cassandra’s vision at the moment.
504
 COMMENTARY line 1117
instruments of the murder, Clytemnestra (airia) and the ἄρκυς (cuvatria) : they
both ‘share (Agamemnon’s) couch’, the woman in the normal sense, the net
in a sense that can only be guessed but has certainly an ominous meaning.
The repetition of the article, ἡ ξύνευνος, ἡ ξυναιτία φόνου, is necessary; it
shows that definite functions of the garment are in mind, i.e. that it is a
particular net and not merely ‘some infernal net’. Perhaps it would be better
to read ἡ in front of ἄρκυς also (this crasis would be quite unobjectionable
here in the trimeter, cf. Sept. 402, S. Ant. 1195, Aj. 1357, etc.).
ἀλλ᾽ ἄρκυς xrÀ.: a correcting ἀλλά after the question, cf. e.g. E. Med.
309 f. σὺ yàp ri μ᾽ ἠδίκηκας ; . . . ἀλλ᾽ ἐμὸν πόσιν μισῶ ('No, it is my husband I
hate’, Denniston, Particles, 5).
No importance is attached here! to the distinction between δίκτυον and
ἄρκυς (for such distinctions see [Xen.] Cyn. 2. 4 ff., Poll. 5. 26 ff.; cf. A. W.
Mair on Oppian, Cyn. 1. 150). The feminine ἄρκυς makes it possible to
personify the net as σύνευνος and ovvarria φόνου and to bring it into close
relation with Clytemnestra.
1117. Στάσις : Casaubon renders it by ‘factio’, Stanley offers a choice between
(x) ἔρις, simultas, (2) coetus. Similarly Blomfield: ‘coetus. Furiarum scilicet.
Cho. 458 στάσις δὲ πάγκοινος ἅδ᾽ ἐπιρροθεῖ. Vel potest esse seditio.’ The mean-
ing ‘coetus’ has been adopted by most commentators, while ‘discord, strife’
found favour with Klausen, Hartung, Ahrens (614 f.), Nagelsbach, Sidgwick,
Pliiss, Blaydes, Wilamowitz. It is obvious? that this latter interpretation is
supported by the answer of the Chorus, where ᾿Ερινύν picks up the preceding
Zráew just as δώμασιν ἐπορθιάζειν does the γένει κατολολυξάτω. Ahrens
rightly says: 'Only if this interpretation 15 accepted is the answer of the
Chorus ποίαν... ἐπορθιάζειν appropriate; the Elders of the Chorus quite
suitably think of the Zrdew as an Erinys unknown to them’ (Cratinus, fr.
240 K., makes τάσις in the political sense the mother of Pericles). Sidgwick
points out that where Aeschylus uses στάσις in the sense of a ‘troop’, he
always adds something, e.g. ἦδε or ἐμή, to make the meaning clear, whereas
‘to use στάσις absolutely, without article or defining gen., for "the band of
Furies’ would be very harsh’. áxóperos can hardly be regarded as sufficiently
defining. Moreover, it is clear from Eum. 976 ff. τὰν δ᾽ ἄπληστον κακῶν μήποτ᾽
ἐν πόλει στάσιν raid’ ἐπεύχομαι Bpéuew how Aeschylean the conception of the
àkóperos Στάσις is; in Passow-Crónert s.v. ἀκόρητος good parallels? to Ag.
1117 are collected, e.g. fr. mel. adesp. 123 Bergk (= fr. mel. mon. adesp. 13,
Anth. lyr. i. 4, p. 217 Diehl) καὶ τὰν ἀκόρεστον αὐάταν, E. Med. 638 ἀκόρεστα
νείκη. The word στάσις is not confined to the usual political meaning, cf.
e.g. Pers. 188.
The form ἀκόρετος is required by the metre here and in 1143. The linguisti-
cally correct form is ἀκόρεστος (six instances in Aeschylus) since the verbal
stem is κορεσ-, cf. Schwyzer, Griech. Gramm. i. 697 5. The form dxóperos
may be influenced by the epic form ἀκόρητος.
γένει: the dative denotes (as in the parallel phrase 1119 δώμασιν) simply
1 For Cho. 1000 cf. p. 743 n. 1.
? Meineke (Philol. xix, 1863, 206 f.), however, failed to notice it; he put a stop after
γένει, and read κατολολύξατ᾽ ὦ and in the next line made ποίαν ἐρινύν the object of ἐπορθιά-
Lew; he was followed strangely enough by Blass on Cho. 942.
3 For the conception in general cf. E. Norden, Aus altrömischen Priesterbüchern, 134 f.,
on "Apos dros πολέμοιο and the like.

505
line 1117 COMMENTARY

that to which the ὀλολύζειν is directed ; it was correctly taken in this way by
Schneidewin and Wecklein. γένει is certainly not to be given a local meaning
(this was the view of Hartung and Nägelsbach). Stanley and, among others,
Paley, Sidgwick, Verrall, L. Campbell take ἀκόρετος γένει together (‘insatiable
against the race’) ; this seems most unnatural.
1118. κατολ. θύματος: the genitive as e.g. in Sept. 393 ἵππος χαλινῶν κατ-
ασθμαίνων (quoted by Paley). For the part of the ὀλολυγμός in the act of sacrifice
cf. on 597; it is used in the same figurative manner Cho. 386 ἐφυμνῆσαι....
ὀλολυγμὸν ἀνδρὸς θεινομένου.
λευσίμου : several editors have unreasonably altered it. Those who retain
the MS reading adopt the explanation of Heath: ‘sacrificium lapidatione
dignum’. This covers the meaning of the word fairly well, cf. also (van
Heusde) Phot., Suid. etc. s.v. καταλεύσιμον : τὸν ἄξιον τοῦ καταλευσθῆναι εἶπε
Δείναρχος ἐν τῶι κατὰ Aukoëpyou. Of course the nuance implied by ἄξιον is
possibly only secondary. It is not found in the phrase 1616 ónuoppideis . ..
λευσίμους ἀράς (‘the objective of which is stoning, which leads to stoning’).
Here we must understand: ‘sacrifice, slaughter, which leads (or should lead)
to stoning’, just as θανάσιμον is that which causes death. C. Arbenz, Die
Adjektive auf -wos (Diss. Zürich, 1933), 79, makes it seem probable that
λεύσιμος is ‘a word formed loosely after the pattern of θανάσιμος᾽; the first
example of λεύσιμος occurs in the Ag. There is a close connexion between the
λεύσιμοι dpat which the Chorus in 1616 tell Aegisthus will await him from the
people and the δημόθροοι dpat which, as the Chorus say (1409, 1413), threaten
Clytemnestra as the result of her action. The murder of the king produces
the angry cry: destruction to the murderers! outlaw them (1410 f.)! stone
them! The deed in question is a slaughtering that brings stoning in its train,
a θῦμα λεύσιμον. The ὀλολύζειν to which the Fury of Dissension, 2’rdots, gives
voice is directed against this θῦμα λεύσιμον, and tlie curses of the people
which accompany the vengeance for this action are λεύσιμοι : the relationship
between the two ideas is unmistakable. The particular significance of stoning
is clear, too: ‘Aevoiuov: indicating that the murder is an action which severely
damages the whole community or defiles it with deeds of horror and must
therefore be righted by the indignant community as a whole’ (Pliiss). For the
nature of stoning among the Greeks as an expiation undertaken in the name
of the whole community cf. R. Hirzel, ‘Die Strafe der Steinigung’, Abk.
Sachs. Ges. d. Wiss., phil.-hist. Kl. xxvii, 1908, 254.
We must reject the interpretation followed by Paley, among others, in his
translation, 1871 (in his commentary he wavers): ‘a victim to be immolated
by stoning’, with the footnote: ‘At the prospect of Clytemnestra’s death’ ;
similarly e.g. L-S s.v. θῦμα I. 3 and s.v. λεύσιμος. This is contrary to the
observation that throughout this section (1090-1129) Cassandra, at each stage
of her visions, is concerned only with one subject at a time, i.e. only with
that which shows itself to her at any given moment. For this reason it is
impossible that θύματος should refer to anything else but φόνου 1117. Here
and in the following lines down to 1129 Cassandra’s mind is occupied entirely
with the death of Agamemnon.
The thought Στάσις δὲ κτλ. (‘And Dissension . . .’) emerges from the prophecy
of the murder contained in φόνου and follows on from it naturally enough.
Φόνος and Στάσις go together like brother and sister.
506
 COMMENTARY line 1122

1119. ποίαν Ἔρινύν κτλ. It is not likely that the Chorus is enquiring about
the nature of the Erinys (‘what sort of Fury is this that . . .?', L-S s.v.
ποῖος I. 3). The correct function of ποῖος here, and hence the tone of the pas-
sage, was perceived by Wecklein: ‘Was willst du mit der Erinys da, die dem
Hause ihren Gesang anstimmen soll? (What do you mean by talking about
this Erinys who is to address her song to the house?).’ Of course this is not
a case of the downright jeering to be found in the familiar ποῖος questions of
everyday speech, which pick up the words of the other speaker, as in ποίου
βασιλέως ; ποία ψίαθος ; and the like, nor is the colloquial phrase in its crude
terseness copied here as in S. Trach. 427 (‘to characterize a plebeian’, Wilamo-
witz on E. Her. 518) or in E. Hel. 567 (‘in liveliest displeasure’, Wilamowitz),
but an echo of the colloquialism is nevertheless audible. Wilamowitz, it is true,
draws a sharp distinction (loc. cit.): ‘this usage is not to be confused with
questions of surprise or horror as in A. Ag. 1119 and Suppl. 304.’ But as far
as the present passage is concerned, I cannot agree, and rather think that we
have here the same colloquial phrase modified by its transference to a more
exalted style. It is a matter of horror to the old men that Cassandra speaks
in this way of a murder, of a daemon of strife, and of a θῦμα λεύσιμον ; they do
not want to hear such things: ‘Do not speak of Erinvs and provoke her
᾿ Because they are deeply troubled they use the rather impolite ποίαν ;
then, instead of taking up the name of Zraoıs, as they would do in ordinary
speech, they use a choice variation and call the daemon ᾿Ερινύς. The audience
cannot be in doubt that the two are to be identified; this is made clear by
τήνδε, ‘this Erinys of whom you speak’.
It was noted by Blomfield that forms of the present stem of κέλομαι are
very infrequent in drama ; he quotes E. Hipp. 1283, in anapaests (W. Breiten-
bach, too, Unters. z. Sprache der Eurip. Lyrik, 1934, 44, quotes no other
passage.)
1120. οὔ pe φαιδρύνει (cf. on 1109) λόγος. Blomfield follows the scholion
(od σεσαφήνισται) and explains: 'clarum i.e. scientem reddit’. This would be
in keeping with the idea implied in ἐπάργεμον 1113, cf. 1178 ff. ἐκ καλυμμάτων
. δεδορκὼς... λαμπρός. Yet it seems rather hazardous to assume that
φαιδρύνει με could mean ‘it makes me (mentally) bright, it enlightens me’.
So it is better to adopt the meaning which is expressed in the interlinear
gloss in Tr (εὐφραίνει) and has been the most commonly accepted since
Stanley ('haud me serenat sermo’), i.e. to take φαιδρύνειν in the sense of
exhilarare 'to hearten, cheer' and the like. This conforms to the general
meaning of φαιδρός and an occasional later use of φαιδρύνειν or φαιδρύνεσθαι.
The thought ‘what you are saying does not cheer me’ is an excellent con-
tinuation of the impatient dismissal of the subject (see above) contained in
ποίαν ’Epwüv κτλ. Both phrases are expressive of the old men's strong
repugnance. But there is no escape for them: horror overwhelms them and
dominates their next words.
1122. Schol. M: ἡ σταγὼν τὸ αἷμα, the obvious interpretation. We need not
look for something more sophisticated. It is unlikely that Aeschylus should
have had in mind the gall (an idea assumed by Tucker and Wilamowitz and
rejected by G. Thomson). Headlam has a good parallel from Aristotle fr. 243
Rose (Gellius 19. 6. 1) διὰ τέ... o£. . . φοβούμενοι ὠχριῶσιν . . .; ὅτι τοῖς φοβη-
Üciow συντρέχει (scil. τὸ αἷμα) eis τὴν καρδίαν, Wore ἐκλείπειν ἐκ τῶν ἄλλων
507
line 1122 COMMENTARY

μερῶν. The poet with an easy transference of epithet calls the blood which
flows to the heart in fear κροκοβαφής, because yellow is the colour of fear.
John Symmons, in a note on his translation (published 1824), quotes Massinger
(1583-1640), The Emperor of the East, iv. 5: ‘My blood within me turns, and
through my veins, Parting with natural redness, I discern it Changed to a
fatal yellow." For the conception in general cf. also the passage from the
Alcmeo of Ennius (scen. 24 ff. Vahlen): ferribilem minatur vitae cruciatum et
necem ; quae nemo est tam firmo ingenio et tanta confidentia quin. refugiat
timido sanguen alque exalbescat metu.
The meaningless xai Sopia was changed by Dindorf to καιρία. This has
found much support, from Peile, Franz, and Conington down to Verrall and
J. €. Lawson. The conjecture is in form very elegant: καιδορια can easily be
explained as the product of two variae lectiones, καὶ δορὶ and καιρία (thus
Wilamowitz) ; cf. e.g. my app. crit. on 137. For the extension of two syllables,
-pía, beyond the end of the dochmiac, whereas in the corresponding place
1133 of the antistrophe there is the end of a word, cf. 1175. But as soon as the
sense is considered, καιρία seems far less satisfactory. καιρία πτώσιμος has
been taken in various ways. Peile: ‘which at the fatal moment, arrested by
the hand of death, finishes . . .’. Such an asyndeton between epithets of this
type is improbable. Others make καιρία dependent upon πτώσιμος, as e.g.
Nägelsbach : ‘caduca, ut letum afferat', Paley: ‘trickling from a fatal wound’,
Wecklein: 'hitting a dangerous spot', Verrall: "literally "shed so as to be
mortal" '. All these scholars take καιρία in the local sense, which is reason-
able, for in the passage compared by them, Ag. 1343, καιρίαν is undoubtedly
local. Sidgwick assumes the same syntactical construction of καιρία πτώσιμος,
but regards (with Peile, see above) καιρία as temporal; he is followed by L-S
s.v. καίριος 11. Since there are instances of predicative supplements (nouns or
adjectives) added to πίπτειν (cf. on 999), it may be possible to understand
καιρία πτώσιμος — ἣ καιρία ἔπεσε, though I think that the strong and rare
adjective πτώσιμος would be more forceful without a predicative supplement.
But there is another consideration of much greater weight : πίπτειν appears to
be quite an unsuitable word for saying that the blood flows to the heart
(1121) or to another part of the body. It is quite different when the blood
falls to the ground ; Hermann seems to have felt this, but his y& δορὶ πτώσιμος
provides a thought which is impossible in this context.
This is not a passage where I can presume to make a decision with absolute
confidence. But it now seems to me extremely probable that Casaubon suc-
ceeded in restoring the genuine reading. δορὲ πτωσίμοις fits in very well with
639, the only other passage where πτώσιμος is found. With the expression
δορὶ πτωσίμοις we can compare δορικανής, δουρικμής, δορίτμητος from Aeschylus
and δοριπετής from Euripides. And as for the καί, it is absolutely essential.
This is obscured by the fact that many editors have translated ἅτε (τε) here
as if it introduced a comparison.? The nature of dere does not allow this; cf.

1 I am inclined to believe that this impressive touch occurred spontaneously to the

English poet. However, it is well to remember that Massinger, who had the benefit of an
Oxford education, borrowed freely from ancient literature (Cambr. Hist. of Engl. Lit. vi. 163).
2 διτε (Wilamowitz in his edition of 1885, later recanted) should not be considered (the
form in the Laconian song Ar. Lys. 1308 does not provide an adequate parallel). More-
over, in a real comparison a new subject would be expected.
508
 COMMENTARY line 1125

besides G. Hermann’s Praefatio, p. vi f., to his edition (Erfurdt, zrd ed.) of

S. Oed. R., e.g. Kaibel on 5. El. 151, Kühner-Gerth, ii. 236, Denniston,
Particles, 523 f. Perhaps the modern reader expects a comparison here; but
that should not prevent him from realizing that the poet insists upon
identity. He is concerned with the characteristic quality not of blood in
general but of the κροκοβαφὴς σταγών. That liquid, the product of deadly
fear, runs to the heart of the Elders. ‘It is the very same’, they continue,
‘which also (xai) . . . to warriors mortally wounded.’ It is of no use to be more
logical than the poet and argue that the blood of those warriors could not
possibly be identical with the blood in the veins of the Elders.
δορὶ πτωσίμοις is dative of the person concerned. This dative occurs fre-
quently in Aeschylus, cf. on 1378, particularly with verbs of coming and
arriving.
1123. It is necessary first to establish the meaning of αὐγαῖς. The scholiast
took it as referring to the eyes: μηκέτι ὁρῶντος ταῖς αὐγαῖς. This use of
αὐγαί is found not only in Tragedy generally but also in Aeschylus (fr. 99.
13 N.; on the text cf. Wilamowitz, Interpr. 235) ; but it is difficult to see how
this meaning can yield any satisfactory result here. The connexion with
βίου δύντος points to the meaning ‘rays’ ('Quum proprie dicantur δυσμαὶὲ
ἡλίου, recte additur αὐγαῖς, ab ἡλίου αὐγαῖς petitum’ van Heusde). Abresch
compared with βίου δύντος Plat. Laws 770 ἃ ἐν δυσμαῖς τοῦ βίου (referring to
age), and Petrus Victorius in his commentary on Arist. Poet. 21, p. 1457 Ὁ 25
τὸ γῆρας ἑσπέραν βίου ἣ δυσμὰς βίου (not Empedocles, cf. Vahlen and Bywater
ad loc.; Diels-Kranz, Vorsokr. i, sth ed., 371, note on Β 152) quoted this
passage of the Ag. But in the words of the Chorus the phrase βίου δύντος
αὐγαῖς clearly means the sunset of life, i.e. death. For the aorist cf. e.g.
Hdt. 7. 149. 3 προειπεῖν... πρὸ δύντος ἡλίου ἀπαλλάσσεσθαι ἐκ τῆς... χώρης.
&uvavüre.: ‘whose arrival synchronises, coincides’ (Headlam).
1124. ταχεῖα δ᾽ ἄτα πέλει. This is the inference drawn by the Chorus from
what they are now experiencing (ἐπὶ δὲ καρδίαν xrA.). Such a fit of deadly
dread is no vain foreboder of ill: σπλάγχνα δ᾽ οὔτοι ματάιζει. In the last
stasimon the Elders declared themselves unable to shake off δεῖμα προστατή-
ριον καρδίας; however much they disliked it, they were forced to anticipate
the future. Here they undergo a similar experience. The images, too, are
kindred, especially the verbs: δεῖμα προστατήριον kapôtas . . . πωτᾶται there,
ἐπὶ καρδίαν ἔδραμε κροκοβαφὴς σταγών here.
1125. Shortening of ἃ long vowel or diphthong before an initial vowel occurs
in the dochmiacs of Aeschylus several times in the case of an interjection, cf.
Sept. 86, Eum. 143, 254, Prom. 576 (P. Maas, Sokrates, iii, 1915, 313).
A number of editors put a stop (or a colon) after Boos ; thus Karsten, Keck,
Margoliouth, Blaydes (hesitatingly), Headlam, Mazon. The scholion in M
(τίς ἂν ἀποστήσειεν αὐτῆς τὸν Ayauéuvova) takes ἄπεχε... τὸν ταῦρον together.
This is correct. There does not appear to be any example of active ἀπέχειν
used intransitively = ‘to keep clear of’ (Linwood, Lex. to Aesch., should not
have quoted under this heading Prom. 687, where ἄπεχε is transitive; Weck-
lein’s note in his commentary, Athens 1896, is correct; in Ar. Peace, 1053
Bachmann’s conjecture is quite arbitrary). Furthermore, the words ἄπεχε
T. B. 7. ταῦρον are to be taken together not only because this is the most
natural arrangement but also because the other interpretation, according to
509
line 1125 COMMENTARY

which ἄπεχε would be addressed to Agamemnon, becomes impossibly harsh

when τὸν ταῦρον follows immediately (in that case we should rather expect
‘you, the bull’).
ἄπεχε κτλ. Blass, Rhein. Mus. Ixii, 1907, 265, remarks: ᾿ἄπεχε is not ad-
dressed to anyone nor is anyone even thought of as being addressed, just as
in the case of ἰδού and our “look there" ; . . . So ἄπεχε simply means "away
with", as we should say.’ This observation of the general linguistic usage is
correct (cf. Wackernagel, Syntax, i. 106), but it fails to do justice to the special
pathos of this passage. Cassandra has reacted with the passivity of a tor-
mented sufferer to everything that has revealed itself up to this point; but
now that her vision has come to the verge of the fatal stroke, there breaks
from her breast a desperate cry for defence and help, just as even nowadays
the spectator in an Italian playhouse will burst into a warning shout, when
upon the stage he sees destiny looming over the hero. Cassandra is clearly
not addressing a particular person, but anyone who might come to the rescue.
She raises her cry although she knows that the portended fate is inevitable:
what she now sees coming is so terrible that she is swept away by her emotion.
As Schneidewin suggested, the words ἄπεχε τῆς Boos τὸν ταῦρον may be
derived from the language of the herdsman: ‘keep the bull away from the
cow, so that he does not do her any harm'.! Proverbial character had pre-
viously been claimed for this phrase by Boeckh on Pind. Pyth. 4. 142 (vol. ii.
2, p. 275). As Cassandra speaks these words (to leave the following lines aside
for the moment) she does not state which of the two animals is in danger.
But the audience familiar with country life must at first suppose that the
cow is threatened. It is only when λαβοῦσα follows? that the veil is torn
asunder and the unheard-of horror revealed: in this case it is the female
that is the unnatural aggressor. With the rapid development of the vision
this reversal of the natural roles strikes us like a sudden blow and is on that
account all the more powerful in its effect; in a later passage (1231) the
horror which the deed arouses finds its strongest expression in the words
θῆλυς ἄρσενος φονεύς (cf. also Eum. 627 ff.).
Schneidewin notes that the representation of man and woman by bull and
cow follows the tradition of oracular language. It may be said in general that
Cassandra’s visions, in that they sometimes (cf. 1223 ff., 1258 ff.) represent
human beings as animals,? draw upon the sphere of imagery which is charac-
teristic of the oracles,* as may be seen from e.g. Herodotus and Aristophanes
1 The mistake of Schneidewin in supposing that the herdsman's phrase must have been
ἄπεχε τὴν βοῦν τοῦ ταύρου and that Aeschylus reversed it is easily corrected as soon as the
arrangement of the cases of the nouns in similar expressions is taken into account. van
Heusde's comparison of Sept. 503 eip£et νεοσσῶν ὡς δράκοντα δύσχιμον is helpful.
2 Cf. on 1107. In 744 the feminine participle παρακλίνασα gives concrete meaning to the
preceding general notions, the grammatical form of which was neuter; there the change is
of course not so violent as here.
3 Jacob Burckhardt, Griech. Kult::rgeschichte (Kröners Taschenaugg. ii. 501): “In her
grandiose manner of describing persons (Agamemnon, Clytemnestra, Aegisthus) quite
clearly in the form of animals she demonstrates that she not only sees through the walls of
the palace’, etc.
+ This is of course not confined to the Greek world. An example picked up at random:
H. M. and N. K. Chadwick, The Growth of Literature, iii. 340 f., discuss a famous chant by
a great Hawaiian poet and prophet. It describes the horrors of contemporary civil war, and
prophesies the success and glory of the future king. “The poem is a riot of brilliant meta-
510
 COMMENTARY line 1127

and, with grotesque exaggeration, from Lycophron (cf. Wilamowitz in his

commentary on Cho. 275, p. 182 and his app. crit. on Prom. 857). In particular
the idea of representing Agamemnon by the bull readily suggested itself, for
it was familiar to every Greek that Agamemnon in the Nekyia (A 411) says:
'Aegisthus together with my wife killed me ὥς ris τε κατέκτανε βοῦν ἐπὶ
φάτνηι.
1127. μελαγκέρωι: precise information on the reading of M, which can without
difficulty be gathered from the facsimile, is supplied by Zacher, Hermes, xviii,
1883, 473, and Sidgwick (A. Y. Campbell took this into account) 'ueAdykepov
in pelayképar! mutatum, v iterum suprascripto M’; the other editors (not
only Hermann and Franz, but also Vitelli-Wecklein, although in their
edition, as in that of Kirchhoff, the erasure is recorded) and Blass, Introduc-
tion to his edition of the Choeph., p. 22, do not say that the last letter was
originally a v, which then was partly scraped out and an ı written in its place.
Thus μελάγκερων, which is also the reading of FG (accented there μελαγ-
κέρων) and Tr, can certainly claim stronger authority. It is in keeping with
this that the scholion in M starts from μελάγκερων: τὸν μελάγκερων ταῦρον
λαθοῦσα τῶι μηχανήματι τῶι διὰ τῶν πέπλων τύπτει. ἐὰν δὲ γράφηται μελαγκέρωι,
(μελαγκέρωι)ξ μηχανήματι τύπτει ἀντὶ τοῦ κεκρυμμένωι. From the general argu-
ments of Rostagno (in the preface to the facsimile, p. 12, under a) it seems
at least a likely supposition that the man who made the erasure on the line
in M and then wrote the » was the same person as the writer of the scholion ;
but this needs re-examining. For the controversial question whether this
διορθωτής had access to another exemplar of the text, see the remarks about
the MSS in vol. i. However that may be, the introduction of the divergent
reading by the words ἐὰν δὲ γράφηται μελαγκέρωι seems to point not to the
improvization of a Byzantine but to a varia lectio which is possibly very old.
Cf. e.g. schol. E. Or. 1287 ἐὰν ἧι “ἐκκεκώφηται' ἄνευ τοῦ ν... dav δὲ ἦι 'Enkerw-
dnvra’... Ἀριστοφάνης δὲ γράφει ᾿ἐκκεκώφωνται ξίφη᾽, schol. E. Phoen. 547 f.
ἐὰν γράφηται ‘éyew’, καὶ ‘arrovépew’, ἐὰν δὲ ᾿ἔχων᾽, καὶ ᾿ἀπονέμων᾽. From the
fair number of passages in the Pindar scholia where the scholiast writes ἐὰν
δὲ γράφηται κτλ. I will quote only schol. P. 3. 75 (on 3. 43) ἐὰν μὲν γράφηται
ἐν πρώτωι, ὡς νῦν γέγραπται (so, too, the Pindar MSS), ἔσται... ἐὰν δέ, ὡς ἔν
τισι, βάματι δ᾽ ἐν τριτάτωι, ἔσται... 6 μέντοι Ἀρίσταρχος γράφει τριτάτωι. Cf.
also the instances of ἐὰν μὲν. . (γράφηται or the like), ἐὰν de... which
W. G. Rutherford, A Chapter in the History of Annotation (1905), 68, quotes
from the Aristophanes scholia. The reading μέλαγκέρωι, thesefore, should
not be immediately dismissed in favour of peAdyxepwv. Supposing it the
original reading, its assimilation to τὸν ταῦρον, which is very close, i.e. its
alteration to the accusative, was, if not absolutely inevitable, certainly very
natural. The choice between the two variants can only be made by comparison
of the intrinsic merits of each. It is readily seen (and an examination of the
phors. ... The sustained image of the Hawaiian battle-chiefs as fighting cocks occupies the
whole of the first half, the eager impetuosity of the fighting birds being forcibly contrasted
with the quiet house of sleep in the second half.’
1 With regard to the accents Wecklein, Stud. z. Aesch. 1, notes that the accent over the
a is ‘am. recentiore', I am not in a position to check this statement. Wackernagel, Nachr.
Gott. Ges. 1914, 123, is inclined in this case, as with similar compounds (cf. above on 79
ὑπεργήρων), to regard the paroxytone as the original accentuation.
2 Added by Paley.
511
 line 1127 | COMMENTARY
examples of μηχάνημα and μηχανή used in the singular confirms the first
impression) that μηχανήματι requires a qualifying attribute. Some com-
mentators indeed defend μηχανήματι without an attribute and take it as
‘artfully’ and the like ; but this is quite unsupported by the usage of the word.
The object, on the other hand, can easily be supplied from the immediately
preceding clause and the whole context. So μελαγκέρωι is to be adopted.!
Before we turn to the explanation of the word, we must assign to the phrase
μελαγκέρωι μηχανήματι its proper place in the sentence. Competent Greek
scholars have found it possible to acquiesce in an interpretation like this (I
am selecting only a few representatives of this view): ‘Construe ἐν πέπλοισι
λαβοῦσα, ‘‘she has got him fast in his robes, and is goring him with her dark
horn" ' (Paley), ‘des Mantels Falten fangen ihn, die Tücke des schwarzen
Horns trifft ihn (Wilamowitz), ‘the Bull [in accordance with his punctuation,
see above] in raiment she has caught and with a black-horned engine smites’
(Headlam). This means that the words μελαγκέρωι μηχανήματι are dissociated
from their close connexion with λαβοῦσα, which, stands between them, and
taken with τύπτει. The scholiast did this, too (see above). The supposition
of such a word-order is linguistically unsound. It disregards the fact that ‘a
part of a sentence the verbal centre of which is a participium coniunctum,
provided that it serves to describe a self-contained action, forms a syntactical
κῶλον of its own'.? The explanation which I am refuting wholly destroys the
links that bind the kolon together internally, pulls down its fixed barriers,
and intermingles the component parts of two kola taken from their natural
context. Cf. also on 1511, and p. 794 n. 2. van Heusde was right, then, in his
protests against such a procedure, ‘quippe orationis partes miscens’ (he
decided, however, for peAdyxepwv) ; Verrall, too, felt some doubt on account
of the order of words. It appears, then, that the participial kolon held to-
gether by λαβοῦσα extends from ἐν πέπλοισιν down to μηχανήματι and is
followed by the 'Kurzkolon' (reduced clause) rémre.* This conclusion reached
so far on purely formal and grammatical grounds will presently be confirmed
by consideration of subject-matter. Finally it will be shown that the terseness
of τύπτει in contrast to the prolixity of the preceding words is characteristic
of similar passages in the trilogy.
ἐν πέπλοισιν λαβοῦσα goes naturally together. The idea that Clytemnestra
catches her husband (as fish are caught, 1382) in the garment which in her
hands becomes a net permeates the Oresteia, cf. Cho. 492 ff., 998 ff., Eum.
459 ff., 633 ff. (where we find a similar expression: ἐν δ᾽ ἀτέρμονι κόπτει
πεδήσασ᾽ ἄνδρα δαιδάλωι πέπλωι). The κυρία λέξις would be λαβεῖν ἐν δικτύωι,
cf. Plat. Euthyd. 302 Ὁ ὥσπερ ἐν δικτύωι εἰλημμένος (Blaydes), Arist. Hist. anim.
8. 2. 589 Ὁ 7 λαμβανόμενος ὃ δελφὶς ἐν τοῖς δικτύοις, and also E. Phoen. 263 1.

1 Some critics, although conceding the necessity of an attribute for μηχανήματι, reject the
MS variant μελαγκέρωι and instead alter ev πέπλοισιν. This is improbable in itself, and turns
out to be definitely wrong as soon as we are able to demonstrate that ἐν πέπλοισιν is un-
objectionable.
2 Quoted from my article ‘Kolon und Satz I’, Nachr. Gétt. Ges. 1932, 202, cf. ibid. 213 and
particularly ibid. ii, 1933, 321 ff. In order not to rely on my ‘Sprachgefühl’ or on anything I
have observed elsewhere, I have recently searched in all the lyrics of Aeschylus for a parallel
to the interlacing of κῶλα which the scholia and Paley, etc., assume here, but to no purpose.
3 Le Père Brumoy hit the truth: ‘Elle le surprend enveloppé dans un funeste vétement....
Elle le frappe ... . Il tombe dans son bain.’
512
 COMMENTARY line 1127

ὃ καὶ δέδοικα μή με δικτύων ἔσω λαβόντες οὐκ ἐκφρῶσι κτλ. Cf. further Cho. 557
ws... καὶ ληφθῶσιν ἐν ταὐτῶι βρόχωι. Consequently λαβοῦσα proves to be so
appropriate that λαθοῦσα need not be considered. It occurs in the scholion
of M, where the reading is not, as recorded in all the editions from Victorius
and Stanley to Vitelli-Wecklein, τὸν μελάγκερων ταῦρον λαβοῦσα, but quite
definitely and clearly λαθοῦσα (λαθοῦσα is noted by van Heusde as the reading
of the scholion in the Wolfenbüttel MS, saec. xv, which is a copy of the
Mediceus). This may be meant as the explanation of a text which contained
Aadoüca ; in that case we should be dealing here with a variant which might be
ancient. But it is equally conceivable that it is only a mistake made in the
copying of the scholion.! Whatever the truth may be, λαθοῦσαξ does not
yield any sense acceptable in the context ; Clytemnestra ensnares her husband
in the garment and does not conceal herself in it. We have seen that on
grounds of word-order μελαγκέρωι μηχανήματι must belong to the Aaßoöca-
clause and not to τύπτει. And everything else goes to show that μηχάνημα
refers to the festal robe which has become a net. This στέγαστρον ἀνδρός is
actually called τὸ μηχάνημα in Cho. 981 (cf., too, Cho. 1003 τῶιδε... δολώματι),
and, still more important, this unique instrument of murderous malignity is
throughout the Oresteia depicted so vividly and placed so much in the fore-
ground that the actual weapon used for the murder seems to play a subordi-
nate part; sometimes its exact nature is deliberately kept dark (see Ap-
pendix B). Many a murderer has set to work with a weapon (a sword or
whatever it may be), but it was reserved for Clytemnestra to turn a peaceful
festal robe into a net of death, and so to execute a μηχανᾶσθαι as cunning as
it was repulsive. The contrast between the snare which is all-important and
the weapon which is in comparison of minor consequence is clearly brought
out in the present passage. First the weapon is not indicated at all and
secondly the detailed λαβοῦσα clause is followed by the abrupt, if powerful,
verb, τύπτει. The same contrast may be noticed in 1380 ff. : first the impossi-
bility of flight or of self-protection (preparation for the net-theme), then two
whole lines depicting the festal robe and its use as a net, then concisely
παίω δέ νιν δίς. Very similar, too, except that the main verb is inserted, is
Eum. 634 f. φᾶρος περεσκήνωσεν, ἐν δ᾽ ἀτέρμονι κόπτει πεδήσασ᾽ ἄνδρα δαιδάλωι
πέπλωι. In the present passage μελαγκέρωι μηχανήματι is perhaps in apposi-
tion to ev πέπλοισιν. As an alternative it should be considered whether it is
to be taken as a modal (or, if preferred, instrumental) expression for the
purpose of defining the action more exactly. For μηχάνημα is not limited to
designating an object cunningly and skilfully thought of and devised: it can
refer directly to the process of cunning thinking and devising, i.e. it can just
as well denote the artful and insidious attack itself as the concrete instru-
1 Confusion of λαθ- and λαβ- is common, cf. e.g. the passages quoted by Denniston on
E. El. 545 f.
2 Moriz Schmidt, Verbesserungsvorschläge zu . . . Aeschylus’ Agamemnon (Jena 1864),
17 n., suggested λαθοῦσα as the reading of our passage.
3 On another occasion Aeschylus (fr. 375 N.) calls the garment in which Agamemnon
was ensnared ἀμήχανον τέχνημα (Nauck's emendation, probably correct, for τεύχημα) καὶ
δυσέκδυτον. The passage may come from the Proteus (cf. Wilamowitz at the end of his
commentary on the Choeph., p. 252 n. 3; Conington on Cho. 1-7 had already considered this
possibility) or it may belong to the lost part of the prologue of the Choephoroe (Conington,
loc. cit.). Wecklein wanted to put the line in the place of Cho. 999, with little probability.
4872-3 D 513
line 1127 COMMENTARY
ment. For this stronger ‘abstract’ meaning cf. Prom. 989 f. οὐκ ἔστιν αἴκισμ᾽
οὐδὲ μηχάνημ᾽ ὅτωι προτρέψεταί με Ζεύς krÀ., Eur. fr. 288 N. δόλοι δὲ καὶ σκοτεινὰ
μηχανήματα. Finally we come to the difficult μελαγκέρωι. The alterations
proposed by Ahrens and others attempt to eliminate the second part of the
compound. But it was just this part that Hermann recognized as essential
for the structure and connexion of the whole passage: ‘quoniam tauri et
vaccae appellatione usa erat, transfert cornua ad id de quo proprie dici non
potuerunt’ (in his further explanation he is less successful). In the note on
1125 it has been shown how a strong surprise effect is achieved when with the
word λαβοῦσα the customary relation between the sexes is suddenly reversed
and the female becomes the aggressor. The metaphorical colouring of the
preceding exclamation (ἄπεχε τῆς βοὸς τὸν ταῦρον) has not yet receded in
favour of purely human terms, although ἐν πέπλοισιν λαβοῦσα by itself might
make it seem so. Instead, the notion of the cow is continued in a characteristic
adjective ‘... horned’. The horns are the weapons of the cow; and the cow,
i.e. the woman, is going over to the attack. Here the adjective ‘dark-horned’
comes in,’ but it does not refer to her weapon of bronze or steel, for that is of
secondary importance (see above), but to that unique instrument of her guile
which really encompasses the man’s destruction. It is bold certainly, but not
too bold for the enigmatic speech of the ecstatic seeress. The attempt to see
in the epithet a vivid detail of the actual procedure leads inevitably to absurd
consequences or at least to a picture which, if conceivable at all, can be only
vaguely visuálized.? It should be clear also what an excellent touch it is that
in this context, as in the related passages of the trilogy quoted above, the
garment is described according to its decisive function, i.e. its effect as
Clytemnestra’s implement of destruction, and not by its outward appearance
at any given moment. The scholion paraphrases the epithet as κεκρυμμένωι,
and the idea that the malicious attack ‘comes out of the darkness’ is certainly
appropriate. When Orestes later (Eum. 459 ff.) reports the deed, he says:
ἀλλά νιν κελαινόφρων ἐμὴ μήτηρ κατέκτα, ποικίλοις ἀγρεύμασι κρύψασα. For both
passages.we may recall again Eur. fr. 288 δόλοι δὲ καὶ σκοτεινὰ μηχανήματα. I
would like to believe that there is also a natural association of ideas between
μελαγκέρωι and the description of the garment when it appeared in its first
1 Separate from the rest of the text of the scholion there stands in M opposite 1127 in
the right-hand margin this gloss: τῆς μελαγκέρου βοός. This explanation appears to pre-
suppose the reading μελαγκέρωι and to mean : ‘the μηχάνημα is called μελάγκερων because it
is the μηχάνημα of the black-horned cow.’ If this is what is meant, this interpretation
coincides in one of the main points with mine.
2 e.g. Wecklein: ‘The two outstretched hands are the black horns, so that the thing
appears to be black-horned.’ I do not think it is permissible to visualize the situation to
such a degree, but if so, the garment must consequently be envisaged as black, and this is
impossible in view of the δαίδαλος πέπλος (Eum. 635), the πλοῦτος einaros (Ag. 1383), the
ποίκιλμα (Cho. 1013), the ποικίλα ἀγρεύματα (Eum. 460; ποικίλα has presumably a double
meaning here, but in the first place means ‘patterned’). The idea, too, that the sword is
held under the garment in such a way as to form a kind of horn leads to the same difficulty
in view of the *black'-horned. There is another reason for regarding this assumption as
impossible. According to the idea which Aeschylus consistently maintains, the garment is
not used to conceal the murderous weapon (as e.g. in E. Or. 1125 κρύπτ᾽ ἐν πέπλοισι roid’
ἕξομεν ξίφη or in Schiller's ballad, ‘Zu Dionys dem Tyrannen schlich Damon, den Dolch im
Gewande’) but to be thrown over Agamemnon to prevent him from escaping and rob him
of free movement in the narrow δροίτη, so that he is defenceless when the woman strikes
him down.

514
 COMMENTARY line i129
vague form as ‘some net of Hades’. If the attempt is to be made to unfold
fully what is really quite untranslatable, we may render: ‘with black-horned
contrivance’ and paraphrase somewhat thus: ‘with the contrivance which
secretly works black mischief, the contrivance of the horned animal that uses
its horns to attack’.
1128. τύπτει properly denotes the thrust or blow with the sword or other
weapon of close combat (frequently in Homer), but here it also continues the
idea of the horned aggressor; cf. 655 κεροτυπούμεναι.
Blomfield and Hermann read κύτει for τεύχει simply on the ground that
they considered exact correspondence with γένει 1117 essential. Against this
Otfried Müller, Appendix to Aesch. Eum., p. 8, pointed out that in Bum. 157
μεσολαβεῖ κέντρωι = 164 φονολιβῆ θρόνον there is exactly the same kind of
responsion. He did not, however, convince Hermann (Opusc. vii. 12), who,
like many after him, accepted Wakefield’s θρόμβον in Eum. 164. The passage
in the Eumenides is difficult,’ but as far as θρόνον is concerned, no serious
charge has been made against it, save the metrical one, and that is based on
petitio principii. In the same chorus, Eum. 176, the MS reading ποτιτρόπαιος
δ᾽ ὧν (= 171 παρὰ νόμον θεῶν) is perhaps to be retained (most editors read with
Porson ὧν δ᾽, but this cannot be affirmed with certainty, for, as Wellauer
ad loc. observes, ‘saepe librarios δέ, quum tertio loco positum invenirent,
transposuisse iam saepius monitum est’; cf. Headlam, On Editing Aeschylus,
114 f.; the need for caution is demonstrated particularly by Eum. 615, where
Canter’s μάντις ὧν δ᾽ (δ᾽ dv MS) has rightly received general acceptance,
whereas in A. Suppl. 914 the excision of the δ᾽ is better than Porson's ὧν δ᾽
(cf. Wilamowitz ad loc.). In Ag. 1468 the MS reading δαῖμον, ὃς ἐμπίπζεις
(= 1448 φεῦ, τίς ἂν ἐν τάχει) is perhaps correct, though éurirras would be an
easy change. A further parallel for the loose correspondence under discussion
(besides the passage which may possibly be brought forward here, Ag. 1176,
q.v.) would be Sept. 699 κυρή]σας μελάναιγις δ᾽ οὐκ (= 706 τροπαΐαι χρονίαι
perad[Aaxrds), if δ᾽ could be retained, but this is uncertain. We can quote with
greater confidence, though unfortunately not with complete certainty, Sept.
698 ἀλλὰ σὺ μὴ ᾿ποτρύνου" κακὸς οὐ κεκλή[σηι = 705 viv ὅτε σοι παρέστακεν"
ἐπεὶ δαίμων, for it is rather unlikely that Aeschylus should have lengthened
the initial syllable of κεκλήσην in this metre.
1129. λέβητος: ‘a large, caldron-like vessel, used for all kinds of purposes.
It was generally made of metal . . . silver a 137’ (v. Lorentz, RE Suppl. vi.
218). The vessel, as represented by Aeschylus, is unmistakably identical with
ı A special difficulty arises out of the question how to punctuate 163 ff. In the scholia
and the text of M we find a colon after 163 (πλέον) but no sign of punctuation after 165
(κάρα), which means that φονολιβῇ θρόνον is to be taken as depending on προσδρακεῖν. This
was adopted by Otfried Müller and e.g. by Verrall and Wilamowitz. I prefer to follow
Hermann and others in continuing the sentence after 163 and putting a stop after 165
(κάρα). For I do not think we should neglect the guidance provided by the accurate
symmetry between the strophe and the antistrophe. This symmetry is brought out partly
by the repetition of the same words and sound patterns, and partly by the parallelism of the
syntactical structure and the equal length of words and groups of words. The compromise
to which Paley and Murray resort (they put a colon—Paley—or full stop—Murray—
after πλέον and a dash after κάρα) seems to me unsatisfactory. Paley (who, following Müller's
idea, divides the sentence among different speakers) takes φονολιβῇ θρόνον «rÀ. as the object
of the yet unspoken προσδρακεῖν. Murray probably means the text of his edition to be taken
in the same way, whereas in his translation he accepted the punctuation of Hermann.
515
line 1129 COMMENTARY

the ἀργυρότοιχος δροίτη of 1539 f. The name λέβης included among other things
vessels for washing oneself, or parts of the body at least, as in τ 386, 469 (for
the use of the word in Homer cf. Brommer, Hermes, lxxvii, 1942, 359 and
366 f.), and it had the same meaning in Athens (cf. e.g. Pollux 1o. 46, 9o). In
Pindar, Ol. 1. 26, the word denotes the bath in which the new-born child was
washed (cf. Wilamowitz, Pindaros, 234 n. 3). Aeschylus is likely to have left
to his audience the choice between thinking either of the vessels for washing
and bathing with which they were familiar themselves or of what they
imagined the ἀσάμινθοι of the Homeric poems to be like.
τύχαν: attempts have been made to explain it 'insidiosi lebetis casum'
(Stanley), or more definitely ‘einer meuchelmordenden Badewanne Geschichte’
(Wecklein), ‘the story of a... caldron' (Headlam). It may well be doubted
whether τύχη could have that meaning in the fifth century (there is nothing
comparable in Aeschylus and Sophocles, and the survey of Gerda Busch,
Unters. z. Wesen der τύχη in den Trag. des Eur., Diss. Heidelberg 1937, 14 f.,
does not show any parallel). But, even granting that τύχη could have that
meaning, the idea 'the story of a murderous caldron' is hardly suitable here,
where the caldron is mentioned only as instrumental to Clytemnestra's plot
and not in connexion with any former event. Weil's τέχναν is excellent.
Hense (in the margin of his private copy) refers to ὃ 529 δολίην ἐφράσσατο
τέχνην, said of the trap which Aegisthus lays for Agamemnon. The elaborate
contrivance with which Prometheus is fastened to the rock is called in
Prom. 87 τέχνη, and there the last word of the line ὅτωι τρόπωι τῆσδ᾽ ἐκκυλι-
σθήσηι τέχνης is, as Wilamowitz recalls in his note on Ag. 1129, watered down
to τύχης in some MSS. For the meaning of τέχνη, cf. e.g. Pind. P. 2. 32 οὐκ
ἄτερ τέχνας (schol. οὐ χωρὶς δόλου).
σοι: cf, on 1095.
It is through the succinct and allusive indications in the stanza 1125 ff.
that the audience experiences the full impact of Agamemnon's murder. We
are led up to this climax by Cassandra's preceding sentences (from ı100 on),
in which she first reveals the approaching crime in general terms and then
(1107 ff. and 1114 ff.) describes in detail the bath scene, with Clytemnestra
tending her husband, and the fatal ‘net’. When at last the brief but heavy
verb τύπτει is heard, it strikes us with the vigour of a final blow. We picture
the king, pitifully smitten down, long before the bath is actually shown to
our eyes. So great is the power of Cassandra’s words that we do not miss
anything when later on the deed itself is dealt with in the shortest possible
form (1343-5). Even the desperate cry of the dying man cannot match the
horror aroused by the vision of the seeress. What the Greek poet makes us
see with our mind's eye is infinitely more forcible in its effect than anything
actually shown on a stage could be.
1130. γνώμων is the man who, from experience, insight, and special knowledge,
is capable of deciding and judging, as in Thuc. 1. 138. 3, Xen. Mem. 1. 4. 5
(in Lys. 7. 25 ἐπιγνώμονας is probably rightly supplied from Harpocration,
who quotes it from this speech). Just as in the passage of Xenophon the
tongue which knows how to distinguish between sweet and bitter is called
γνώμων, so here the word is used of the person who hearing obscure prophecies
knows what their purport is and whether they promise good or ill.
ἄκρος: cf. on 628.
516
 COMMENTARY lines 1134 f.

1132. ἀπὸ δὲ θεσφάτων picked up from the preceding sentence and (as a
‘Kurzkolon’) placed emphatically at the beginning (cf. G. Thomson, C.Q.
xxxiii, 1939, 148), then the interrogative pronoun follows at the beginning of
the next kolon.
1133. τέλλεται has rightly met with almost universal acceptance (Sidgwick
defends στέλλετα). Emperius (Opusc. 297) and Hermann put forward the
emendation without giving reasons ; then Dindorf showed that this was con-
firmed by the scholion: ἀπὸ τούτων τῶν θεσπισμάτων τίς ἀγαθὴ φάτις γίνεται.
Forms of τέλλειν are not found elsewhere in Aeschylus, though they occur in
Pindar and Sophocles.
From the ‘recensio’ of our MSS it must be deduced that the reading of the
archetype was διαί, It is clear that the preposition here serves to mark the
instrument or means. It is taken in this way by Hermann: ‘per mala’; but
when he paraphrases ‘faciunt ut quis, quid significaverit timor, ipso eventu
malorum intelligat’ he is on more doubtful ground. We should expect a more
exact antithesis to ἀγαθὰ φάτις. So Headlam is presumably correct: “Terrns
of evil are the means by which the wordy arts convey their lesson of prophetic
fear.’ Paley quotes S. Trach. 1131 τέρας τοι διὰ κακῶν ἐθέσπισας (Schol.
ἄπιστον yap διὰ δυσφήμων ὥσπερ ἐμαντεύσω), and Headlam makes the attrac-
tive suggestion that both passages are based upon a proverbial phrase.
1134. πολυεπεῖς here only. Exactly what colouring Aeschylus intended to
give the word is unknown. The Homeric πολύμυθος gets us no further. It
must mean more than ‘viel tausend Sprüche’ (Wilamowitz) ; the context
points to something depreciatory. The meaning lies presumably in the direc-
tion of 5. Butler’s comment (quoted by Peile): ‘fallaciae multis verborum
ambagibus involutae ; ut sunt oraculorum pleraque'.
1134 f. Weil indeed believes that the MS reading, θεσπιωιδὸν φόβον, means
‘vaticinationum horrorem', and he adds the comment ‘a lyrica dictione mihi
non videtur alienum esse'. Yet the words as they stand ought to mean 'the
fear which sings prophetically'. This could well be said of the δεῖμα προστατή-
ριον καρδίας τερασκόπου (976 f.), but can it be applied to the φόβος which is
the result of sinister prophecies? If, however, emendation is necessary, it is
difficult to see what improvement Hermann's θεσπιωιδοί is supposed to
represent over Casaubon's θεσπιωιδῶν which was the reading of most editions
before Hermann. φόβον φέρουσιν μαθεῖν seems doubtful. Hermann translates
intelligentiam. timoris afferunt; Headlam ('convey their lesson of prophetic
fear”) and others render it similarly. But I have no parallel for φέρειν with a
dependent infinitive used like that. Lobeck understood the passage in the
same way as Hermann, for in his note on S. Aj. 799 he quoted Ag. 1135 as a
possible support for Bothe's conjecture ἐλπίζειν φέρει, ‘id est ἐλπίδα φέρει
metuere nos facit" (he did not, however, put this conjecture into the text as
did Hermann) Paley, Wecklein, and others take μαθεῖν epexegetically
(— ὥστε μαθεῖν). Infinitives tacked on in this way are, as is well known,
customary in Aeschylus, but the addition here would be redundant and
feeble. Nagelsbach gives a very artificial grammatical explanation and then
paraphrases: τέχναι φέρουσι φόβον μὲν τῶν μάντεων, φόβον δὲ τοῦ μαθεῖν τὰ
θεσπίσματα. 1 consider it certain that the Chorus means fear of the fate
prophesied and not of experiencing or understanding the prophecy. van
Heusde's 'iung. κακῶν διαὶ pabetv’ is the solution of despair. I have reached
517
lines 1134 f. COMMENTARY

no acceptable conclusion. Bothe’s μάταν provides a possible sense: ‘in vain’,

for the sinister prophecies are incapable of averting the disaster. But it is
hard to see how the common μάταν could have been corrupted to μαθεῖν.
1137. θροῶ must not be tampered with. Ahrens (p. 618) and, independently,
Wilamowitz conjectured θρόωι, but although θροεῖν is very frequent in
Tragedy, there is no instance there of 0póos. However, this is a minor point ;
what is really fatal to this conjecture is the fact that it completely destroys
the careful interlacing of the two halves of the stanza. As the text stands, the
clause in the answer of the Chorus (1140 f.) ἀμφὶ δ᾽ αὑτᾶς θροεῖς echoes Cas-
sandra’s words τὸ yap ἐμὸν θροῶ πάθος. Such an answer which picks up the
words of the other speaker has been shown to be a characteristic mark of
early tragic antiphony (and then of dialogue) between actor and Chorus, see
W. Kranz, Stasimon, 23; his illustrations are taken from the Supplices and the
Persae; cf. p. 693 f.
At the end of the line it may be taken as certain that there is a syllable too
many. For in the corresponding line of the antistrophe (1147) differences of
opinion should be confined to the restoration of the first dochmiac. Most
editors have rightly left the second dochmiac πτεροφόρον δέμας untouched.
Experiments such as Weil’s, who inserts two syllables into 1147/8 against the
MSS, or van Heusde’s transposition πτεροφόρον γάρ of περὶ δέμας βάλοντο
(Murray transposes similarly) have no claim to consideration. Therefore the
double dochmiac in 1137 may be regarded as certain. As for emendation,
Schoemann, Opusc. iii. 148, was on the right track when he supposed that the
gloss ἐπεγχέασα had displaced the correct reading; he read ἐπέγχυτον. But
Headlam's ézeyyvdav is far more convincing. Headlam compares Cho. 67,
where διαρρύδαν in the scholion is glossed by διαρρέων, and particularly Eum.
553, where for παρβάδανϊ we have the gloss παραβεβηκότα. The adverb κατα-
χύδην occurs as early as Anacreon, fr. 80. 3 D. Cassandra applies the same
image to the lament for her murder as later on (1260 f.) to the murder
itself.
1188. ποῖ... δεῦρο can hardly be retained. van Heusde's reference (he
was followed by Wecklein) to E. I$h. T. 77 is misleading, for there in ποῖ u’
αὖ τήνδ᾽ és ἄρκυν ἤγαγες the ποῖ has its full local meaning, as the whole context
and particularly 93 f. (ἥκω 8é . . . ἄγνωστον és γῆν) shows. But here, as is
clear from the answer οὐδέν ποτε κτλ., Cassandra is asking to what purpose
she has been brought to Argos. The question ‘where am I?' is appropriate at
the beginning (1087), where Cassandra is approaching the house, but not here.
Nor are the passages quoted by L-S s.v. ποῖ Il under the heading to what end ?*
of any help, for in Pers. 735 ποῖ τελευτᾶν and S. Oed. C. 227 ποῖ καταθήσεις the
local sense is clear, and even clearer in Cho. 732 ποῖ δὴ πατεῖς Κίλισσα δωμάτων
πύλας (‘pregnant construction, “Whither bound, dost thou pass the gates?" '
Sidgwick). The comparison with S. Oed. R. 2 (Plüss) is quite unjustified.
The emendation τί is easy ; the mistake was possibly caused by reminiscence
of 1087 d ποῖ ποτ᾽ ἤγαγές με.
Who is addressed here? Stanley and others understand Agamemnon.
1 The reading of M περαιβάδαν appears to have arisen from πεβιβ.
2 After this was written, the mistake was corrected in the Addenda to L-S (p. 2100);
but a warning against this interpretation may still be useful, for such ‘parallels’ possess an
uncanny power of attraction.
518
 COMMENTARY line 1142

S. Butler and many others think of Apollo.’ Hermann argues: ‘Agame-

mnonem dicit, ut ex sequente versu apparet’; Weil on the other hand com-
ments: '£vwÜavovuévqv non impedit quominus ἤγαγες de Apolline accipiatur.'
It is true that Agamemnon's death in any case dominates the thoughts of
Cassandra as of the audience, so that the reference in ξυνθαν. would be quite
clear even if the preceding sentence were not addressed to Agamemnon. It
may be wondered why the poet, if he wished to make Cassandra, after all
these portentous visions, turn again to the god as at the beginning, did not
show it clearly by a renewed invocation (cf. 1257). But perhaps this considera-
tion is not sufficiently weighty to exclude Apollo as the subject of ἤγαγες,
especially since 1136 ff. is sharply marked off from the preceding lines as a
new section by τὸ yàp ἐμὸν θροῶ πάθος κτλ. In favour of Apollo it may be
said that in two exactly similar phrases in this scene, 1087 and 1275 f. καὶ νῦν
ὁ μάντις μάντιν ἐκπράξας ἐμὲ ἀπήγαγ᾽ és τοιάσδε θανασίμους τύχας, Apollo is
the ἀγαγών (on the other hand ἐμῆς ἀγωγῆς 1263 refers to the bringing of
Cassandra by Agamemnon). It is then probable though not certain that here
too Cassandra is addressing Apollo. It may perhaps be felt that with this
interpretation the passage gains in depth and the whole scene in unity.
Apollo was and is Cassandra’s fate, so he dominates her mind completely : his
presence makes itself felt ever anew in her utterances.
1139. οὐδέν more: the addition to the negative of an indefinite pronominal
adverb belongs to the original stock of the Indogermanic languages, cf.
Wackernagel, Syntax, ii. 267 ff.
“τί yáp ;, like τί μήν;, when the question is rhetorical, comes to mean "of
course” ', Denniston, Particles, 85.
1140. ppevopavis: this is the only example.
θεοφόρητος (cf. 1150): a rare word in old Greek; cf. Wilamowitz's remark
(Glaube d. Hell. ii. 320 n. 2) on the title of Menander's Qeopopoupevy.”
1141. αὑτᾶς : for the use of ἑαυτοῦ etc. to cover the first and second persons
as well as the third, a use which in Greek as in kindred languages is presum-
ably secondary, cf. Wackernagel, Syntax, ii. 94 f., Schwyzer, Griech. Gramm. i.
606. Cf. also on 1672 f., p. 800. |
1142. νόμον. The scholion adds to the gloss ὠιδὴν ἀηδῆ the explanation παρὰ
τὸν κιθαριστικὸν δὲ νόμον φησί. The reference to the ‘kitharodic nomos’ is not
justified. There is here no special technical meaning, not even in the sense in
which such a meaning could be ascribed to the early Greek period (for the
original use of the term νόμος cf. Wilamowitz, Timotheos Perser, 90, and
Pindaros, 144 n. 1; Abert, Die Antike, ii, 1926, 142). Here as in Alcman and
elsewhere νόμος means simply ‘song’ or ‘melody’ (cf. Vetter, RE xvii. 842.
1 ff.) ; Alcman (fr. 93 D.) says significantly οἶδα δ᾽ ὀρνίχων vouws πάντων, just
as Aeschylus associates the song of the nightingale with Cassandra’s νόμος.
νόμον ἄνομον: the interpretation ὠιδὴν ἀηδῆ (see above) covers the mean-
ing fairly well, but what in fact the Chorus says is ‘a melody which is no
melody’. It is wellknown that oxymoron of this kind is widespread in tragedy.
Examples have often been collected, as by E. Bruhn, Anhang zu Sophokles,
129 ($ 222), and G. Meyer, ‘D. stilist. Verwendung der Nominalkomposition i.
ı Wilamowitz had a special reason for sharing this view: his treatment of 1137 τὸ γὰρ
ἐμὸν κτλ. led him to believe that there, too, Apollo was the subject of the sentence.
2 Cf. the title Θεοφόρητος of a comedy of Alexis.

519
line 1142 COMMENTARY
Griech.’, Phil. Suppl. xvi, Heft 3, 1923, 103 f. The history of this figure of
speech has been outlined by Wackernagel, Syntax, ii. 291 ; he refers also to the
unique instance in Homer *Ipos Aipos (a 73). The oldest example in Aeschylus
appears to be Pers. 680 νᾶες dvaes, then come Prom. 544, 904, and a number of
examples in the Oresteta.
With regard to τις in a comparison cf. on 288.
ξουθά: Rutherford’s extensive discussion of the word (Babrius 118. 1) has
not altogether superseded that of Blomfield in the glossary on this passage,
see also Conington ad loc. and the succinct statement of the main facts in
Wilamowitz’s note on E. Her. 488; special attention may be drawn to his
statement that the word, which we find first in the poetic language of the
fifth century, was ‘by that time scarcely understood’ (on this point cf. also
Rutherford). In general cf. A. C. Pearson on E. Hel. 1111. Passow’s article
shows sound judgement; it was followed by Dindorf and L-S down to the
8th edition. On the other hand, the treatment of the word in L-S? is ex-
tremely arbitrary,! presumably under the influence of Méridier's article, cf.
Kretschmer, Glotta, vii, 1916, 354. The statement of Blomfield and Wilamo-
witz, that in the fifth century the word was always taken to refer to colour,
remains true, whatever Rutherford says to the contrary. On the other hand,
the lexicographers (Hesychius, Suidas, s.v.) provide a confusing supply of
modifications in meaning, and it is a fact (cf. L-S?, s.v. 4) that later poetry
sometimes used ξουθός of sound. But there does not seem to be any certain
example? of the meaning ταχύς which is put forward by the lexicographers.
In whatever way the later reinterpretations of the word came about, ξονθή
as applied to the nightingale refers probably to the colour and not to the
song. The reason for the use of such an epithet for the inconspicuous little
bird is to be found perhaps in the fact that the nightingale appears in τ 518
as χλωρηὶς" ἀηδών and that Hesiod, Erga 203, calls it ἀηδόνα ποικιλόδειρον;*
1 Just one example. How is it possible to take the epithet ξουθαῖσι in Bacchyl. 5. 16 ff.
βαθὺν δ᾽ αἰθέρα ξουθαῖσι τάμνων ὑψοῦ πτερύγεσσι ταχείαις alerós as ‘whirring or steadily-
beating’, when there is ταχείαις there? Jebb ad loc. is correct, Pearson on E. Hel. 1111
wavers.
2 Even Wilamowitz, loc. cit., finds this sense in Hymn. Hom. 33. 12 ξουθῆισι πτερύγεσσι
(but why should a colour-description be unsuitable for the wings of the Dioscuri?) and
Chaeremon, fr.1.7N., p.781f. In the latter passage the meaning ‘quick’ is quite unthinkable ;
this poet of fine sensibility is really not the man to use here an epithet which would be not
only empty but in its context actually meaningless. A reading of the whole fragment with
its abundance of colour-nuances will show that a corresponding feature is to be expected
here. I quite agree with John Symmons, who, in a note on his translation of Ag. 1142,
comments on the phrase of Chaeremon: ‘The poet, talking of the beauty and yellowness of
those waxen tresses that dallied in the winds, says in effect that they communicated their
auburn hue to the winds that played amongst them’ (the particular shade denoted by
ξουϑός cannot, of course, be made out). H. Bartsch, De Chaeremone poeta tragico (Mainz
1843), 34, compares very appositely Antiphanes (fr. 217. 22 K.) ap. Athen. 14. 623 c, where it
is said of the τευθίς : ξανθαῖσιν αὔραις σῶμα πᾶν ἀγάλλεται, This is borrowed, if not from
Chaeremon himself, from a tragedian of similar style, and it clearly supports the supposition
that Chaeremon intended with ξουθοῖσιν ἀνέμοις to give a colour-effect.
3 It isan unfounded supposition of D'Arcy Thompson, A Glossary of Greek Birds, 2nd ed.,
17, that xAwpnis ‘may even be an ancient word of unknown origin, unconnected with
XAwpés”.
4 Before giving the correct meaning, Tzetzes ad loc. says τὴν ποικιλόφωνον λέγει, and
this is rightly rejected by Wilamowitz. I am unable to follow the attempt of O. Schroeder
(Hermes, 61,1926 426 f. and in his note on Ar. Birds 739) to refer ἀηδών here to the swallow.
520
 COMMENTARY line 1143

it is possible that in later epic the bird may have been sometimes described
in the same way and that this very adjective ξουθή may have been applied
to it. Simonides (fr. 45 D.) with his dnddves . . . χλωραύχενες provides a varia-
tion on the phrase in the Odyssey.
1143. ἀκόρετος Boäs. There can be no doubt about the first half of the line.
It is not permissible to consider βοαῖς, for axöperos here requires a genitive
to go with it (γόοις axopeororaroıs Pers. 545, on the other hand, is very good).
So we need not even refer to the Homeric use of ἀκόρητος with the genitive
(μάχης etc.) and the αἰχμᾶς ἀκόρεστον Pers. 999, which is derived from it.
But the remainder of the line is a problem. Wilamowitz adopts the commonly
accepted reading, i.e. that of M, but only faute de mieux, for in his apparatus
he says: ‘videtur Herm. recte propter importunum φεῦ miramque v.l. in M
non adquievisse. Sed antistrophus non est sollicitanda.' Hermann says that
the line in the antistrophe raises doubts, ‘et φεῦ minime hic aptum’, so he
reads here dxdperos Bods φιλοίκτοις ταλαίναις φρεσίν and inserts two syllables
into the antistrophe; neither of these suggestions is at all convincing. But
his note on φεῦ is indeed apt to make us pause. There is no example of φεῦ
in Aeschylus which is not provoked by an occurrence affecting the speaker
himself, by something which, whether he actually takes part in it or ex-
periences it in his imagination, is of immediate concern to him or his (e.g.
Pers. 568, 576). The use of φεῦ, as it seems to be employed here, with the
mention of a distant event which is subordinated to the main topic by way of
simile, is unusual. Moreover, the exclamation seems to be extremely strong
in itself: to judge from its use elsewhere in Aeschylus, we might suppose
that it expresses not mere sympathy but horror or a lament implying con-
demnation. Were it not for the metrical responsion, one would be inclined
perhaps to reject the φεῦ which is read only in M. But in the antistrophe
(1153) there is nothing really doubtful, as far as I can see; after long uncer-
tainty I have now come to the belief that Wilamowitz is right : ‘antistrophus
non est sollicitanda.’ So φεῦ is probably to be retained in 1143. Perhaps we
are to understand it like this: the Chorus, in conjuring up the sorrows of the
nightingale, are overwhelmed by the horror of the familiar story with its
scenes of murder, a mother’s murder of her child (Suppl. 65 ff). This would
perhaps give φεῦ its full force here too. To pass on to φιλοίκτοις ταλαΐψαις,
these words are hardly tolerable in this juxtaposition. But I consider it quite
improbable that φιλοίκτοις should have arisen as a gloss on ταλαΐναις, as most
editors since Heath suppose. τάλας is a very common word, whereas φίλοικτος
in earlier Greek is to be found, so far at least, only in Ag. 241. It is, moreover,
particularly suited to the mood of the lamenting nightingale, and precisely
the word we might expect from Aeschylus in this context; in Suppl. 57 ff.
he does actually write ‘if one of the augurs hears my lament (olxrov), he
would think he heard öra τᾶς Τηρεΐας Μήτιδος οἰκτρᾶς ἀλόχου, κιρκηλάτας
ἀηδόνος" dre . . . πενθεῖ μὲν οἶκτον ἠθέων... τὼς καὶ ἐγὼ φιλόδυρτος ᾿]αονίοισι
νόμοισι δάπτω τὰν... παρειάν. The triple οἶκτος and οἰκτρόν is resumed by
the equivalent of φίλοικτος, viz. φιλόδυρτος, which occurs nowhere else. Cf,
also S. Aj. 629 f. (Ajax’s mother) οὐδ᾽ οἰκτρᾶς γόον ὄρνιθος ἀηδοῦς ἥσει, El.
1076 f. orevaxovo’ ὅπως ἃ πάνδυρτος (correctly emended from πανόδυρτος)
ἀηδών. So it is probably right to follow Dobree and van Heusde in adopting
1 For the text see below on 1526, p. 723.
521
line 1143 COMMENTARY
φιλοίκτοις and bracketing raAaivats as a variant. The ¢-alliteration, too, may
possibly be mentioned as supporting φεῦ φιλοίκτοις φρεσίν (cf. on 268).
1144 f. “The grammatical relation of the accusatives "Irvv and βίον is not
certain’ says Headlam, and indeed several combinations have been tried.
But in this case a decision is possible. We can begin by stating that “Zru
“Irvv στένουσα go together. The combination of the onomatopoeic (and hence
readily repeated) name! with a verb of complaining or lamenting became a
favourite element of the nightingale motif, probably under the influence of
the passage in the Odyssey, r 522, παῖδ᾽ ὀλοφυρομένη "IrvÀov φίλον. Cf. S. El.
148 f. ἃ "Irwv, αἰὲν "Iruv ὀλοφύρεται ὄρνις, Eur. fr. 773 N. (cf. Wilamowitz,
Berliner. Klassikerlexte, v. 2, p. 81) 1. 23 ff. μέλπει δὲ δένδρεσι λεπτὰν ἀηδὼν
ἁρμονίαν ὀρθρευομένα γόοις (here a special turn is given to the idea of lamenta-
tion in regard to the early morning, the time when the action is taking place,*
cf. E. Suppl. 977) Ἴτυν “Iruv πολύθρηνον, Horace, Odes 4. 12. 5 Ityn flebiliter
gemens (the swallow), Culex 25x f. Pandionias . . . puellas, quarum vox Ityn
edit Ityn. So Hermann is right: ' "Irwv "Irw imitatio est vocis lusciniae et
pro adverbio construitur cum στένουσα.᾽ He is right, too, in his interpreta-
tion of the sentence as a whole: ' Ityn Ityn clamitando gemens affluentem
malis vitam’.? Most commentators, it is true, follow Stanley (‘per pullulan-
tem malis vitam") and take dud. x. βίον as accusative of duration of time; so
do Wilamowitz and Headlam in their translations. But the momentous
epithet ἀμφιθαλῆ κακοῖς makes it clear enough, one would think, that it refers
not to the limit of time but to the subject of the lament. It is astonishing
that the marked correspondence in thought between 1141 audi δ᾽ αὑτᾶς
(θροεῖς) and ἀμφιθαλῆ κακοῖς βίον has been missed. From a purely gram-
matical viewpoint we may say that in the relative clause (old τις κτλ) a
θροεῖ or similar verb must be understood from what goes before. But this
would not do justice to the tightly packed expression. The subordinate
clause would be understandable in itself without the addition of the βίον-
phrase: 'you sing of yourself, like the nightingale in her lament for Itys'.
But the sentence gains in beauty through the rich expression: 'as the
nightingale in her lament for Itys bemoans her ill-fated life'. Thus ἀμφ᾽
αὑτᾶς is given subsequently a more intense meaning and the true balance is
restored between comparison and compared.
ἀμφιθαλῆ: this word, which clearly belongs to cult-language, points
regularly to blessing and prosperity. The παῖς ἀμφιθαλής (as early asX 496)
is unsullied by the loss of parents, a fuer patrimus ei matrimus,* so he alone
τ Owing to the English pronunciation the ‘Tereu, Tereu!’ of the English Renaissance-
poets (e.g. Shakespeare, The Passionate Pilgrim, 21. 14) is just as effective, perhaps even
more so.—For duplication as a characteristic means of reproducing the sound of the voices of
birds and inventing names for them see Eduard Wólfflin, Sttzgsb. Bayer. Akad. 1882, iii.
445 (= Ausgewählte Schriften, 299).
2 In Hesiod, Erga 568, where the subject is not the nightingale but Πανδιονὶς χελιδών, the
variant ὀρθρογόη is worth noting ; perhaps it influenced Euripides (cf. O. Schroeder, Hermes,
Ixi, 1926, 425 f.; Wilamowitz’s comment on Erga 568 is unfortunate). The context of
Aeschylus fr. 291 N. θρηνεῖ δὲ γόον τὸν ἀηδόνιον is unknown.
3 Substantially the same rendering was given by Casaubon in the margin of his copy (in
Cambridge) of Victorius’ edition: 'velut aliqua flava (mollis, delicata) luscinia... Ityn
Ityn gemens, vitam nempe suam abunde florentem malis’.
+ As far as the element θάλος is concerned, it does not seem possible to account for the
meaning of ἀμφιθαλής, cf. Bechtel, Lexilogus, 41. Neither the treatment of the word by
522
 COMMENTARY line 1146

is suitable for certain functions in heavenly marriages (Ar. Birds 1737) as well
as earthly, and only he is entitled to perform certain religious offices. In the
other passages where it occurs the word is always boni ominis; this is also
clearly the case in Cho. 394 f., where a prayer for putting right the wrong is
addressed to ἀμφιθαλὴς Ζεύς (the exact meaning of the epithet in this passage
cannot be determined,'! cf. Wilamowitz and Blass ad loc.). Only here is the
auspicious word used to refer to evil. Aeschylus is fond of thus distorting the
meaning of a term of cult-language, cf. on 1387. The Chorus do not go into
the subject of the horrors to which Aëdon has been subjected, but the strong
and startling phrase ἀμφιθαλῆ κακοῖς gives some inkling of the dreadful things
they do not recount. It is not possible to say exactly what sense Aeschylus
meant to give to his reinterpretation of the old word; he was probably play-
ing on its etymology (cf. on 149). He need not intend it to mean ‘flourishing
with evil all around’, but may have had in mind ‘ill-fated on two sides’ (this
would come nearer to the best-known usage). Then it would be left uncertain
what were the two components of the double misfortune. Of the various
possibilities one emerges from Suppl. 64 ff. πενθεῖ μὲν" οἶκτον ἠθέων, ξυντίθησι
δὲ παιδὸς μόρον κτλ,
1146. For the restoration of the metre Dobree’s ἀηδοῦς μόρον should probably
be taken into account as an alternative to Hermann’s generally accepted
transposition. The form andoös occurs in lyrics S. Aj. 629 ;? as for the respon-
sion vuu—-u— = u--u-, there are several instances of it in Aeschylus,
some of them in the Oresteia (e.g. Cho. 953 = 965), cf. also Ag. 1164 δήγματι
φοινίωι = 1175 υ ——v-—.
Blass, Mélanges H. Weil (1898) 13, raised two objections to the MS reading:
(x) id is not used with the accusative elsewhere, (2) μόρος, which occurs very
frequently in Aeschylus, means ‘death’ in all other passages, but it cannot
have this meaning here in reference to the nightingale. The grave alterations
of the text which Blass consequently proposes need not bother us. But the
two observations of the eminent grammarian are valuable. As regards the
first, instead of saying 49‘tw7 cum accusativo minime iungi’, he would have
done better to point out that an accusative of exclamation is altogether very
rare in Greek.* But for the very reason that it is infrequent, this usage should
have been dealt with better than it is in Kühner-Gerth, i. 330. Bernhardy,
Farnell, C.Q. iv, 1910, 186 f., and A. Klinz, IEPOZ TAMOZ (Diss. Halle, 1933), 119 ff., nor the
article of A. Oepke, ᾿Ἀἀμφιϑαλεῖς, Arch. f. Religionsw. xxxi, 1934, 42 ff., which presents much
evidence for the later period (for considerable supplementary information from inscriptions
see L. Robert, Harvard Studies in Class. Phil., Suppl. Vol. i, 1940, 509 ff.) contributes
anything towards removing this linguistic difficulty. However, it is obvious that in
ἀμφι- the original meaning ‘on either side’ is preserved, cf. on 686 and 881.
! Tucker, ad loc., makes the ingenious suggestion (approved by A. B. Cook, Zeus, ii.
1072) that ἀμφιθαλής here means ‘the god of the ἀμφιθαλής᾽ and hence (a very bold step
indeed) ‘the god of both parents’. But leaving on one side the internal difficulties of this
explanation, nothing in the context of the stanza 394 ff. seems to support the assumption
of such an epithet in this passage.
2 μὲν instead of νέον, Haecker ; it improves the sense and the metre, and was adopted by
Headlam (note on his translation) ; Wilamowitz made the same correction.
3 The scholion on this says ἡ ἀηδὼ δὲ κατὰ Μιτυληναίους. Cf, Sappho, fr. 70. 7 D., where
ἄηδοι] is the most probable supplement, although not the only possible one.
* J. B. Hofmann in Stolz-Schmalz, Lat. Gramm., sth ed. (1928), 385, says of the accusative
of exclamation: ‘only in Latin, in contrast to other Indo-Germanic languages, is it of
fr»quent occurrence.’
523
line 1146 COMMENTARY
Wissensch. Synt. d. Griech. 134, was more thorough ; he lists (besides irrelevant
passages) Ag. 1146, and also Callim. lav. Pall. 89 ὦ ἐμὲ δειλάν and Bion, Adonts
28 αἰαῖ τὰν Κυθέρειαν (cf. ibid. 32 al τὸν Adwviv). To these last passages should
be added! Sappho, fr. 21 Ὁ. ὦ τὸν Ἄδωνιν as a much older piece of evidence for
the same ceremonial invocation, and Ar. Lys. 393 αἰαῖ Ἄδωνιν. Perhaps this
kind of accusative was originally used only in ejaculations in θρῆνοι and in
kindred ritual cries. It would be attractive to suppose, both for the passage
itself and also in view of the instance quoted above from Callimachus, that
the beginning of the poem of Alcaeus, fr. 123 D. "Eye δείλαν, ue παίσαν κακο-
τάτων πεδέχοισαν Was an exclamation. But whether this is so cannot be
determined, because the scanty remains of the next two lines do not allow us
to make out whether a verb governing the accusative followed. In any case
there is no reason to doubt the construction iw ... μόρον here. We now turn
to uópov.? As was stated by Blass, it cannot mean simply fatum (so Stanley,
and also Headlam and others). For everywhere in Aeschylus and quite
predominantly elsewhere (cf. Wilamowitz, Berl. Sitzgsb. 1927, 15 = Kl. Schr.
v. 1. 478) the word does not mean ‘fate in general’ but ‘the fated end’. The
divergent meaning assumed for Prom. 248 by older dictionaries and commen-
tators has now been rightly given up (Hermann referred to Plat. Gorg. 523 d) ;
for Cho. 441 see below on 16oo. So we must take μόρον here as e.g. Humboldt
(‘oder Nachtigall Tod’), Verrall, and Wilamowitz do. I can see nothing against
it. Aedon, the daughter of the Athenian king, was by the grace of the gods
preserved at the last moment from being killed by the sword of Tereus: she
lived on as a nightingale until she eventually died a peaceful death—an envi-
able contrast to what is in store for Cassandra (1149). If I still feel a very slight
doubt, it is solely because the ἀηδόνος μόρον of the MSS is not what the poet
wrote. So we have to admit the possibility (though it seems to me very re-
mote) that there was originally a different noun (e.g. πότμον) in front of ἀηδόνος.
1147. It is surprising that any difference of opinion still remains about the
emendation of the line. The reading of M περεβάλοντο ydp οὗ provides a syllable
too many ; moreover the Middle is meaningless here, as Schutz pointed out in
his first edition; he therefore read in the second edition παρέβαλον ydp οἱ
(Hermann's later defence of the Middle 'curaverunt ut vestiretur plumis' was
not successful). Hermann himself, before the appearance of Schutz's later
edition, had already found the correct solution (apart from the minor detail
of the augment) : περίβαλον γάρ ot κτλ.3 More recently English editors (Head-
lam, A. Y. Campbell, G. Thomson) have preferred Enger's (1855) περέβαλόν
γέ οἷ. Enger pointed out that the hiatus was legitimate* but said nothing to
1 The editor of the materials which Headlam collected for his commentary on the Oresteia
should have weeded out S. Ant, 1265 (nom. not accus.), E. Hipp. 870, Anth. Pal. 7. 468 and 9.
424. In the last three passages the accusative depends on a preceding verb, and the inter-
jection is placed διὰ μέσου.
2 G. Thomson’s translation “the sweet music of the nightingale', without a word in the
commentary, is fantastic.
3 This was Hermann's earliest restoration of the passage (published in his book De metris
poet. Graec. et Rom., 1796, p. 434), but he completely suppressed it later. It was mentioned
by Blomfield, then by Wecklein, and, with a slight inaccuracy, by Wilamowitz.
4 Hermann had already referred to S. Trach. 650 à δέ of φίλα δάμαρ, his conjecture S. El.
195 ὅτε οἱ (σοι codd.) has been accepted by the more recent editors. For the hiatus in front of
the dative o in Homeric and post-Homeric verse cf. now particularly Wackernagel, Unters.
z. Homer, 108 f., and Karl Meister, Homer. Kunstsprache, Index s.v. of.

524
 COMMENTARY line 1148
recommend the γε itself; later (app. crit. of the 2nd edition of Klausen’s
commentary, 1863) he abandoned it in favour of περέβαλον γάρ οἱ. The MS
γάρ, explaining the preceding exclamation, is just as excellent here as in the
corresponding passage of the strophe, 1137, where it has the same function,
cf. also 1107 ἰὼ τάλαινα, τόδε γὰρ τελεῖς.
περέβαλον: the elision of the ı before the augment! was successfully
defended by Hermann (in the commentary of his edition). He referred to
Eum. 634, where the MS περεσκήνωσεν ‘is necessary’ (Blass); the scholion,
which was suitably? restored by Wecklein: (ro X) πρὸς τὴν συναλοιφὴν τῆς
περί καὶ τὴν συζυγίαν τοῦ ῥήματος, establishes περεσκήνωσεν as certain for the
Alexandrian edition, so that it is difficult to imagine how this verb which is
so excellently suited to the ἀμφίβληστρον could be rejected (in L-S it has
failed to receive a lemma of its own, and the MS reading of Eum. 634 is
condemned as ‘falsa lectio’ at a place where the reader would not look for it,
s.v. περί G).
oi: for the occurrence of this form of the demonstrative in tragedy cf.
Denniston on E. El. 924.
1148. αἰῶνα: this variant? (cf. on 1127) was recognized as the correct reading
by the scholars of the Renaissance (for its replacement in the text of our MSS
by ἀγῶνα" van Heusde compares 512 παγώνιος in place of παιώνιος ; there,
however, the corruption may be due partly to ἀγωνίους in the following line).
In regard to γλυκὺν αἰῶνα van Heusde recalls ε 152 κατείβετο δὲ γλυκὺς αἰών,
Hdt. 7. 46. 4 ὁ δὲ θεὸς γλυκὺν γεύσας τὸν αἰῶνα φθονερὸς ἐν αὐτῶι εὑρίσκεται ἐών,
and Pind. P. 2. 26 γλυκὺν ἑλὼν βίοτον. The words γλυκὺν αἰῶνα pick up the
phrase of the Chorus ἀμφιθαλῆ κακοῖς βίον" and correct it. A rather more

1 Kuhner-Blass, ii. 32 f. needs correction. In this connexion Hermann refers to another

instance in Attic of the elision of the ε of περι- prefixed to a verb, i.e. the form περιών which
Porson and Dobree, Notae in Aristoph. 140 on Wasps 1020 (it was not Meineke or Dindorf
who saw this first), established as certain in place of the commonly accepted wepuwv for
several passages in comic fragments and for Thuc. 1. 30. 3 (cf. Hude, ad loc.). The shorter
form alone is correct in Attic, see, e.g., Blass on Demosth. 18. 44 (commentary [1890]) and
Wilamowitz, Griech. Lesebuch, Erldut., p. 47, Hermes, xl, 1905, 136, and Platon, ii. 338;
P. Maas, Gnomon, ix, 1933, 664 has again insisted on this point. The evidence of the Derno-
sthenes MSS has been collected by Fuhr in the preface to his Teubner edition (1914)
. XXX n. 8.
P 2 Cf.,e.g., Schol. A. Sept. 79 τὸ δὲ X πρὸς τὴν διαφορὰν τοῦ στρατοῦ καὶ τοῦ στρατοπέδου, Schol.
E. Hec. 323 and also the evidence produced by K. Lehrs, Die Pindarscholien, 110, and
J. W. White, The Scholia on the Aves of Aristophanes, 341 s.v. πρός, and Pap. Soc. It. 846
col. 1. 1o.
3 For the hand by which it is written in the Mediceus cf. Rostagno in the introduction to
the facsimile p. 12.
* This is presumably connected with the fact that less educated persons were prone at
an early period to pronounce intervocalic y as something like y. This was responsible, e.g.,
for the appearance of κλάγω and ὑγιγαίνις in a papyrus of the Ptolemaic period, cf. F. Blass,
Aussprache des Griechischen, 3rd ed., 107, Schwyzer, Griech. Gramm. i. 209. Cf. also the ancient
corruption O 680 συναγείρεται. Conversely we find oréos instead of στέγος in the Mediceus
in Pers. 141. How natural that pronunciation (or the complete neglect) of intervocalic y
was even to comparatively learned Byzantine scholars is shown by the fact that one of them
assumed that in the word εὐάγητον (Ar. Clouds 277) the a and the ys coalesced by synizesis
into one long syllable; see K. Zacher, ‘Die Handschriften der Aristophanesscholien',
Jahrbucher fur class. Philol., Suppl. Bd. xvi, 1888, 629.
5 The same is true of αἰῶνα in the Herodotus passage which I have just quoted : the whole
conversation of Xerxes and Artabanus turns upon the ἀνθρώπινος βίος.
525
line 1148 COMMENTARY
general verbal idea (‘assigned’ or the like) should be taken from περέβαλον to
govern αἰῶνα; the case is similar to those discussed on 797 f.
κλαυμάτων ἄτερ has been the subject of superfluous, nay impossible (C.R.
XXXIX, 1925, 148), conjectures and arbitrary attempts at explanation ; recently
it has even been supposed (by G. Thomson’) that Cassandra, in speaking of
the nightingale, brings forward a form or interpretation of the story com-
pletely different from that which the Chorus give in accordance with the
version current in the fifth century generally and also found in Aeschylus
(Suppl. 62 ff., fr. 291 N.): it is alleged that in Cassandra’s view the nightingale
would not lament and mourn at all. To understand our passage we must first
be clear about the meaning of κλαύματα. ‘Kennst nicht die Thrane’ renders
Wilamowitz with many others. κλαύματα may of course mean this and does
in fact have this meaning in Ag. 887, just as κλάειν may mean ‘to shed tears’,
e.g. Ar. Birds 342, without any thought of sound. Everyone knows that ‘to
cry’ has also taken on the meaning ‘to shed tears’ (the first example of this
use occurs in 1532 according to the Oxford Engl. Dict.), and it is equally well
known that in this case just as with κλάειν (and with plorare?) the reference
was originally to the sound. κλαύματα means the sufferer’s cry of pain; thus
it is used in conjunction with γόοι in Pers. 705. In everyday speech, as is
well known, the person threatened with blows or the like was told κλαύσει
μακρά; the κλάειν κελεύειν familiar in Aristophanes occurs as early as Hip-
ponax, Pap. Oxy. 2174, fr. 16. 18. In the heat of his quarrel with the base
Egyptian, Pelasgus is driven to use the vulgar phrase (Suppl. 925) κλάοις ἄν,
εἰ ψαύσειας, οὐ μάλ᾽ és μακράν. There is the same use in noun form in Ar.
Frogs 813 κλαύμαθ᾽ ἡμῖν γίγνεται. The tragic poets, too, transposing such a
phrase as κλαύσει μακρά into a more exalted style, use κλαύματα, as in Ag.
1628 καὶ ταῦτα τἄπη κλαυμάτων ἀρχηγενῆ, S. Ant. 931 f. τοιγὰρ τούτων τοῖσιν
ἄγουσιν κλαύμαθ᾽ ὑπάρξει βραδυτῆτος ὕπερ (both the threatening speech of the
extremely irritated usurper Aegisthus and the reproaches of Creon, beside
himself with rage, come very near to common abuse). κλαύματα then means
‘a cry or yell of pain’. It is in strong contrast to the καλὸν ἀείδειν of the
nightingale (τ 519), the ἀοιδός (Hesiod, Erga 208), which is ἱμερόφωνος (Sappho
fr. 121 D.), which Aristophanes (Birds 659) calls τὴν ἡδυμελῆ ξύμφωνον ἀηδόνα
Μούσαις and Euripides (Hel. 1109) τὰν ἀοιδοτάταν ὄρνιθα μελωιδόν. The life of
the little songstress, blessed by the Muses, is full of mourning, it is true, but
it is nevertheless a γλυκὺς αἰών ; she does not utter cries of pain, since she was
mercifully rescued at the last moment from the most terrible fate, death by
the sharp blade. There is another passage built up on the same contrast as
Ag. 1148, which happily confirms our interpretation: in 5. Aj. 629 ff. the
Chorus sing of the savage pain which will engulf the mother of the hero:
οὐδ᾽ οἰκτρᾶς γόον ὄρνιθος ἀηδοῦς ἥσει δύσμορος, ἀλλ’ ὀξυτόνους μὲν ὠιδὰς
θρηνήσει, χερόπλακτοι δ᾽ ἐν στέρνοισι πεσοῦνται δοῦποι κτλ. It would of course
be possible to emphasize a very different aspect of the nightingale’s lament.
But Jebb’s note on Aj. 627 ff. holds good for Ag. 1148, too: ‘The contrast
1 The substance of his interpretation, including the reference to Plato, Phaedo 85 a, was
anticipated in Pluss’s commentary (1895).
2 Cf. W. Schulze, Berl. Sitzgsber. 1918, 499 = ΚΙ. Schr. 177; he observes that ‘ “Heulen”
is only one step removed from “Weinen” all the world over’; for parallels from other
languages cf, also his Quaest. ep. 390 n. 3.
526
 COMMENTARY line 1150

meant here is plain enough; though in some other places, where the bitter
grief of the nightingale is the foremost thought, the bird’s note is sometimes
described as “piercing” ; e.g. Trach. 963.’
1149. It seems to be rather remarkable! that for once δόρυ should not mean
‘spear’ or ‘lance’ but ‘weapon’; and yet there is no doubt that this is the
meaning here. It is equally extraordinary that this same object, the weapon
with which Agamemnon (and Cassandra, too) is murdered, is called βέλεμνον
1496, a word used elsewhere only of missile weapons. The extension of mean-
ing was made possible or at least facilitated in the case of δόρυ by analogy
with ἔγχος which besides meaning ‘spear’ or ‘lance’ can bear secondarily the
sense of ‘sword’ or other weapons; in the case of βέλεμνον, the poet’s use of
the word possibly rested upon a similar extension of the meaning of βέλος.
Still we should do well not to minimize thé arbitrariness of the manner in
which Aeschylus uses δορέ here. There can be no doubt as to the intended
effect. The audience could not attach a precise meaning to δόρυ here or to
βέλεμνον in the later passage, i.e. these words would not stand for anything
more special than ‘weapon of attack’. This general expression served to
conceal the particular nature of the weapon, and this was exactly what was
important to the poet. As shown in Appendix B, Aeschylus wished to give
prominence to the garment, the ‘net’ (cf. on 1127), and to divert attention from
the weapon, except in passages where the instrument of murder, the sword
as such, was of special importance. In our present passage, 1149, we must
also take into account the mystifying character of Cassandra’s prophecies;
her manner is similar to that of the oracles: in the treatment of the subject-
matter there is a dislike for too precise details, and in the language there is
some elasticity in the terms employed. But this explanation is not sufficient
by itself, as ἀμφιτόμωι βελέμνωι used in 1496 in quite a different context shows.
In both passages Aeschylus has carefully avoided using an unambiguous
epithet, for although ἄμφηκες (and ἀμφίτομον) is extremely well suited to a
sword, it is not peculiar to it but could also be used of, e.g., an axe. For the
rendering in the gloss of the Mediceus and in Tr cf. p. 806 n. 1.
1150 (cf. 1154) πόθεν... ἔχεις. Cf. Io's question to Prometheus who shows
himself endowed with the seer’s vision (Prom. 593) πόθεν ἐμοῦ σὺ πατρὸς ὄνομ'
amveıs;
ἐπίσσυτος, except for 887 and Eum. 923, occurs only in E. Hipp. 574; this
may perhaps be due to the fragmentary preservation of Greek literature.
But the words of the Chorus in the Hippolytus τίνα θροεῖς αὐδάν, τίνα βοᾶις
λόγον; évere, τίς φοβεῖ σε φήμα, γύναι, φρένας ἐπίσσυτος ; show the same metre
and the same form of question (though this form is not of course uncommon) ;
and they contain, beside the rare ἐπίσσυτος, in φοβεῖ a similarity to ἐπίφοβα,
Ag. 1152. We must therefore reckon with the possibility that Euripides had
these lines in mind. If that is so, then ἐπίσσυτος could probably be regarded
as an ‘Oresteia-word’ (cf. on 386).

1 The assertion of the scholion on S. Aj. 906 πᾶν δὲ ἀμυντήριον καὶ δόρυ καὶ ἔγχος καλοῦσιν
of νεώτεροι is an odd generalization, as far as δόρυ is concerned.
2 Schuursma is wrong when he says (p. 78) that δόρυ is used in 1149 ‘aut gladii sensu aut
securis vicem’ and βέλεμνον similarly in 1496. This misses the main point altogether. He
was too much occupied by his efforts to come to a compromise between the two views
(axe and sword) which he found in the commentaries.

527
line 1150 COMMENTARY

The meaning of &mioovros in the three other passages where it occurs does
not suggest that it should be taken with πόθεν, as is assumed by, e.g., Paley,
Kennedy, Verrall (he actually takes θεοφόρους with it). Nor should we over-
look the strong anaphora with which 1154 móBev . . . ἔχεις repeats the πόθεν...
ἔχεις Of 1150. ‘From whence have you this anguish that sets in violently, god-
possessed (arising from possession by a god), and vain?' Hermann was
probably right in excising 7'. The reason for the intrusion of the particle is
obvious, cf. on 124.
θεόφορος is synonymous with θεοφόρητος 1140.
1151. ματαίους shows how far removed the Chorus is still from understanding
the prophetess and from appreciating the real situation. Cf. p. 460.
1152. ἐπίφοβος = di φόβος ἔπεστι, ‘something to which fear is attached’.
For the formation of the word see Schwyzer, Griech. Gramm. i. 435. Strangely
enough there does not seem to be another instance of émí$ofos before the
period of the Roman Empire, whereas e.g. ἐπίπονος, ἐπίφθονος are quite
common in the fifth century.
δυσφάτωι: the scholion! explains: ἀσαφεῖ, αἰνιγματώδει, This agrees with
the meaning of the word in the only other passage in which it occurs, Lycophr.
1o, where the speaker calls Cassandra's prophecies δυσφάτους αἰνιγμάτων
oïuas (the paraphrase δυσλύτους αἰνιγμάτων ὁδούς is right, and the scholion
δυσλέκτους δεινάς 1s wrong) ; it is not improbable that Lycophron was thinking
of this very line, Ag. 1152. But almost all the dictionaries and commentaries
(except Paley's) have scorned the explanation of the scholion and instead
followed more or less exactly the rendering in Estienne's Thesaurus and
Stanley's translation clangore infami. I can see no reason to depart from the
meaning upon which Lycophron and the scholion on the passage of Aeschylus
agree. It is more suitable than the other. In this sentence the manner of
μελοτυπεῖν is described in two ways: δυσφάτωι κλαγγᾶι and ὀρθίοις ἐν νόμοις.
These two phrases are brought into close relation with each other by ὁμοῦ re
(see on 1153). The relation, it seems to me, is much more understandable if
we find set over against the 'loud strains' (i.e. those which no one can fail to
take in) an 'indistinct, hardly comprehensible' note rather than a baneful,
dreadful note. With δύσφατος = ‘hardly articulate, hardly intelligible’ we
may compare E. Ion 782, where Creusa calls the pronouncement of the oracle,
which is incomprehensible to her, ἄφατον,2 and Plato, Tim. 50 € τρόπον τινὰ
δύσφραστον kai θαυμαστόν. Yet I have no doubt that the word δύσφατος could
in itself take on the meaning 'monstrous, terrible', just as adaros (e.g.
Pindar, N. 1. 47 μελέων ἀφάτων).
κλαγγᾶι: rightly understood by Passow s.v. (his interpretation was
adopted in Dindorf's Thesaurus) : ‘also used of articulated tones: of the song
1 In M in the left-hand margin. In the right-hand margin opposite 1152 (τὰ δ᾽ eridoßa . ..
&Àayyát) there is in M this note: διὰ τὸ δυστυχῇ μαντεύεσθαι, ἢ διὰ τὸ προειδέναι rà δεινὰ
ἀνωφελῶς. This comment does not belong to the words opposite it, but to 1151 ματαίους
δύας, cf., e.g., Hesych. δύη" δυστυχία «rA., Suidas δύη" ἡ κακοπάθεια, ἡ δυστυχία and Schol.
S. El. 642 σπείρηι ματαίαν' τὸ ματαίαν οὐκ ἔστι ψευδῆ ἀλλ᾽ ἀνωφελῆ. I see from Wecklein, ii.
360, that Risberg has assigned the scholion to its right place.
2 ἀναύδητον, which stands next to it, has a quite similar meaning and certainly does not
mean ‘dreadful’ here, as A. S. Owen, ad loc., again maintains (as did, e.g., Dindorf in the
Thesaurus, Jebb on S. Aj. 715). The immediate context (763-77 is quite different) makes
that impossible : what matters is the strangeness of the oracle.
528
 COMMENTARY line 1154

of the Chorus, S. Trach. 208; «A. δύσφατος, of the prophecy of Cassandra, Ag.
1152’. This deviates from the predominant usage, but the divergence is no
greater than when, e.g., κελαδεῖν is used in Pindar and other writers to mean
‘sing’. There is no reason then to take the word here to mean ‘crying’, as
most translators do. |
1153. μελοτυπεῖν is found only here. The meaning of the -rvz-element is not
obvious. There is probably no thought of κρούειν and πλῆκτρον. The closest
parallel appears to be the Aristophanic γνωμοτύπος, γνωμοτυπεῖν. So the first
of Blomfield's two renderings, carmina excudo, is presumably correct.
ὁμοῦ : Hermann's criticism, vehementer languet ὁμοῦ, nisi verbum addatur'
(he consequently makes an insertion of his own), has left a strong impression.
Ahrens (p. 618) talks of the ‘flat ὁμοῦ 7'', and Schneidewin-Hense make a
similar comment. But Nägelsbach is probably right: 'Contrarie inter se
referri videntur δύσφατος κλαγγή et νόμοι ὄρθιοι. Alioquin non video quid
velit ὁμοῦ. It is not, however, advisable to ascribe to ὁμοῦ re the adversative
function which in modern languages may be implied in phrases like 'and at
the same time',! ‘und zugleich’; I have found no instance of such a use.
'In association with', 'together with', is quite adequate. Cassandra's pro-
nouncements of horror are issued both in scarcely articulate sounds and
in loud strains: the δύσφατος κλαγγή and the ὄρθιοι νόμοι are not different or
alternative modes of delivery, but one and the same. They are characterized
first by the difficulty felt in grasping their meaning, then by the way in
which the song strikes the ear. There is a certain antithesis, but this rests
upon the meaning of the epithets: it would seem likely in itself that a pro-
nouncement made in a loud voice without any attempt at secrecy should be
clear and comprehensible. (This interpretation is not fully satisfactory;
there remains a lingering suspicion that something in the line may be
corrupt.)
ἐν: equivalent to a plain dative, cf. Blass on Cho. 423 and in general Wila-
mowitz on E. Her. 932.
It has often been maintained, as recently by W. Kranz, Stasimon, 140 (cf.
ibid. 139), that the ‘loud strains’ here have some reference to the ὄρθιος νόμος
(for which cf. particularly Wilamowitz, Timotheos Perser, 90) ascribed to
Terpander. This is improbable, both on account of the plural and for other
reasons; a protest against the assumption was raised by G. Hermann,
Opusc. vi, pars ii, 157.
1154. ὅρους. . . θεσπεσίας ὁδοῦ, Schütz is right: ‘quis tandem tibi male
ominatam. divinam viam definit; h.e. quis tibi male ominatorum carminum
modos praecipit? ὄροι ὁδοῦ h.l. poetice, quemadmodum apud Pindar. Ol. 9.
47 ἔγειρ᾽ ἐπέων σφιν οἶμον λιγύν᾽. The meaning of ὅροι (in general cf. on 485) is
quite clear here. Α ὅρος marks the boundary of a piece of ground (see, e.g., the
two boundary pillars from the Athenian Agora, erected in the latter part of
the 6th century, with the inscription hôpos εἰμὲ τὲς ἀγορᾶς, Hesperia, 8, 1939,
205 f. and 9, 1940, 266) as also of a road, e.g. IG 12. 878 λόρος ho86, 881 hópos res
686 rés ᾿Ελευσῖνάδε, 887 ἐμπορίο καὶ hodö hópos, to quote only a few early
examples from Attica. Similarly in a metaphor Bacchyl. fr. 11 Snell (7 Jebb)
1 Verrall: ὁμοῦ re, and αἱ the same time, marks an antithesis’. In Nägelsbach’s para-
.

phrase, too, the adversative element is overemphasized: ‘Formidolosa vates canit sono
male illo quidem ominato, sed eodem perquam claro atque acuto.’
4872-3 E 529
line 1154 COMMENTARY
εἷς ὅρος, μία βροτοῖσίν ἐστιν εὐτυχίας ὁδός, θυμὸν εἴ τις ἔχων ἀπενθῆ δύναται
διατελεῖν βίον, with the contrast following ὃς δὲ μυρία μὲν ἀμφιπολεῖ φρενί κτλ.:
there is only one way to success and happiness for mankind and that is
marked by a single boundary (the poet does not need to be pedantic and
remember that there must be ὅροι standing on both sides of the road) ; should
anyone try to walk at random off the road, he will not reach εὐτυχία, but
ἄκαρπον ἔχει mövov.! In the present passage ὅροι means the boundary-stones
and boundaries in general which mark the course of the road. ‘Whence
have you the limitation, the bounded track, of the prophetic song, the track
of your baneful speech?’ From öpovs we may take it that the Chorus is now
conscious of the steadiness and consistency of Cassandra’s prophecy of her
own fate: her lament does not roam at large but follows a definite path and
its direction is always the same. They talk of the ‘path’, because the ‘path
of the song’ is a generally familiar idea, whether or no the Greeks of the
classical period felt the connexion between oiu» (specialized in the Odyssey
so as to mean ‘song, lay’) and οἶμος.2
θεσπέσιος only here in Aeschylus, never in Sophocles (but he has, besides
the common θέσπιν αὐδάν Ichn. 244, θεσπιέπεια of oracles, and, as Aeschylus,
θεσπιωιδός), in Euripides only Andr. 296 παρὰ θεσπεσίωι δάφναι, where the
scholiast explains: κατέχουσα μαντικὴν δάφνην (cf. hymn. Hom. Ap. 396).
Similarly we have here in the scholia (M) the gloss τῆς μαντικῆς. This
explanation is possibly influenced by the ancient etymologies of θέσπις (for
the evidence see Lobeck, Pathol. Graec. seran. elem. i. 309, cf. also Eustath. on
B 600, p. 299. 23 ff.). Nevertheless it is probable, if not altogether certain,
that Aeschylus, who took over θεσπέσιος, as did Pindar and others, from
Homer as a γλῶσσα, intended it to mean 'oracular, prophesying’. It should,
however, be noticed that this meaning is not required in any of the numerous
passages in Homer where the word occurs, nor anywhere else,’ as faras I can
see. So there is the possibility that here (and similarly in E. Andr. 296) we
have simply the Homeric meaning (cf. Buttmann, Lexi. i. 166 f. ; Bechtel,
Lexil. z. Hom. 164 f.), which, broadly speaking, comes to the same as θεῖος,
‘the epithet of every great phenomenon’ (Buttmann), even of a terrifying one,
as in θεσπεσίη φύζα etc, But one would be reluctant to separate the θεσπεσία
ὁδός here from the immediately (1161) following θεσπιωιδήσειν, the adjective
θεσπιωιδός 1134, the verb θεσπίζειν 1210 and 1213 (we might also add the triple
θέσφατα: 1113, 1130, 1132), which are all used in the narrow compass of this
scene to describe Cassandra’s prophecies.
1 Jebb explains: ‘öpos is the canon, the rule or standard, by which true εὐτυχία is to be
measured’, But this takes no account of the close connexion, which existed in practice,
between ὅρος and ὁδός; moreover the meaning he gives ὅρος is hardly credible for such an
early date. In E. Iph. T. 1219 τοῦδ᾽ ὄρος τίς ἐστί por; the meaning ‘time limit’ is absolutely
clear: ‘Thoas would like to know how long he has to wait’ (Bruhn) ; in L-S the passage is
wrongly listed under ‘III. standard, measure’.
2 For the much-vexed problem cf. its recent treatment by Meuli, Hermes, lxx, 1935, 172 f.,
and O. Becker, ‘Das Bild des Weges’, Hermes, Einzelschriften, Heft iv, 1937, 68 f., where
there is a detailed discussion about the ‘way of song’. Becker only touches upon Ag. 1154
in passing, p. 176, without taking into account the ὅροι or comparing the fragment of
Bacchylides.
3 L-S,s.v. I. 2. b, unfortunately follow Gildersleeve in the interpretation of Pind. P. 12. 13;
against this cf. Wilamowitz, Pindaros, 147,and O. Schroeder’s commentary. Farnell, ad loc.,
recommends a compromise.

530
 COMMENTARY line 1164

The answer to the urgently repeated question πόθεν ἔχεις ; is known to the
audience long before the Chorus hears the full explanation (1202) : it is Apollo
who causes her anguish and gives her the foreknowledge of future destiny.
1157. πάτριον ποτόν : the rendering (L-S ποτός II. 2) ‘water of Sc. drunk by
my sires’ is linguistically correct, but in this context, where the audience
would have in mind expressions like γῆ warpia, etc., the meaning comes to
the same as ‘my native stream’ (Paley and others). This kind of reference to
the homeland is an old-established item of the poet’s stock-in-trade (cf.
Orelli-Baiter on Hor. Odes, 2. 20. 20; Norden, Berl. Sitzgsb., 1917, 673 τι. 2):
B 825 πίνοντες ὕδωρ μέλαν Αἰσήποιο, Pind. Ol. 6. 85 OnBav . . . τᾶς ἐρατεινὸν
ὕδωρ πίομαι. |
1159. ἠνυτόμαν : glossed by ηὐξόμην in the Medicean. For this meaning, ‘grow
up’, there is no further instance in L-S.
1161. ὄχθους. Casaubon’s conjecture ὄχθας was vigorously defended by
Wilamowitz. His English followers, Murray, A. Y. Campbell, G. Thomson
(Headlam knew better), would have done well to consult L-S. The alteration
was apparently inspired not only by Ammonius, 108 ὄχθαι... εἰσὶ ποταμῶν
χείλη" ὄχθοι δὲ ἐπάρματα γῆς, but also by the large number of passages in
which this distinction holds good. But it has long been known that this rule
is not without exceptions, cf., e.g., Ellendt-Genthe, Lexic. Sophocl. s.v. ὄχθη.
There is now an additional piece of evidence in the passage quoted in L-S,
Sappho fr. 97. 12 f. D. (p. 44 Lobel), which is decisive for Ag. 1161. There
Lobel's restoration ([o6]x[0]ow ἴδην Axeplovros] is ‘particularly elegant’
(Pfeiffer, Gnomon, ii, 1926, 315) and completely convincing.
1163. Karsten’s emendation deserves its place in the text, not ‘in default of
better’ (Sidgwick), but as an almost certain restoration of the original; it
satisfies grammar, thought, and style. The corruption arose presumably in
the way suggested by Karsten : avaiwv was misread as avwv and this was then
misinterpreted as avwy (ἀνθρώπων). For early instances of the ‘contraction’
ANOC etc. see L. Traube, Nomina sacra, 100 ff. We read in the Mediceus,
e.g., Eum. 56, 183 avwv, Sept. 425 avov, Prom. 445 avoıs (for traces of aver etc.
in the MSS of Plato see Wilamowitz, Platon, ii. 331 together with n. 1).
veoyvös. This compound, which is formed from the ‘weak grade’ of yer-,
presumably did not reach tragedy from the Ionic, where it continued its
active existence, but from the later epic (represented for us in this case by the
Homeric hymns).
1164. In place of the corrupt ὑπὸ (ὑπαὶ is the metrical conjecture of Tri-
clinius) we should read either ὅπως with Hermann or with Franz ἅπερ; the
latter is perhaps better, cf. Cho. 381.
Hermann altered δήγματι to δάκει, which editors have championed over
and over again although it destroys the consistency of the thought. The
conjecture is wholly arbitrary, evoked by Hermann’s belief that exact
1 The comment there was translated into English by Jebb on S. Ant. 1132; he added two
instances which he took from L-S, viz. Pind. P. 1. 64 and E. Suppl. 655. The latter passage,
"Ἰσμήνιον πρὸς ὄχθον, must not be adduced in this connexion; Wellauer, who quoted it in
order to show that the rule had no absolute validity and that Casaubon’s conjecture for
Ag. 1161 was wrong, did not recognize the Hismenion. The mistake was adopted by Her-
mann (on Ag. 1161= 1120 Herm.) and by Passow s.v. ; from Passow’s dictionary it found its
way into all editions of L-S, whereas Dindorf’s Thesaurus, v. 2473 A, gives the passage its
correct meaning.
531
line 1164 COMMENTARY

syllabic correspondence should be insisted upon without exception (cf. 1175).

Just as arbitrarily and for no other reason he alters, e.g., Sept. 453 βέλος ἐπι-
σχέθοι (= 418 δικαίως πόλεως) into βέλος κασχέθοι. Naturally Hermann was no
more tolerant when faced with the responsion — = v v in other places of the
dochmiac, so in Ag. 1165 he reshaped μινυρὰ θρεομένας (= 1176 γοερὰ θανατη-
φόρα) accordingly. The manner in which the Chorus here describe their
emotion, though ennobled by the dignity of tragic speech, corresponds to the
δέδηγμαι τὴν ἐμαυτοῦ καρδίαν of Dicaeopolis (compared by Blomfield); and
when, further, we read in this very play (791) δῆγμα δὲ λύπης οὐδὲν ἐφ᾽ ἧπαρ
προσικνεῖται, there is no need for a laboured paraphrase such as 'I have been
stung as by the bite of a poisonous viper' (Wecklein) or 'trifft mich wie
Schlangenbiss' (Wilamowitz).
1165. κακά has been rightly deleted by Schütz and others (cf. Hermann,
ad loc.).
8péouac occurs only in Aeschylus and Euripides out of all the literature
which has come down to us, 'always of women' (L-S), which is very noticeable.
The Ionic form Sept. 78 θρεῦμαι would suggest that Aeschylus took over the
word from the language of (post-Homeric) epic, to which he seems to be
indebted to a large extent; and we now have θρεύετο in an epigram of Epi-
daurus from the fourth century B.c., ZG Iv. i.? 616. 4 (quoted by L-S, p. 2077),
which, as far as it is preserved, shows a purely epic vocabulary (woAvaivnrov,
as a variant differing but slightly from the Homeric πολύαινον, cannot be
considered an exception).
1166. θραύματ᾽ after being often challenged was shown to be correct by
Bothe and Ahrens (p. 619 f.) : the one compared Ar. Birds, 466 6 τι τὴν τούτων
θραύσει ψυχήν, the other Hesychius θραῦσμα" λύπη, πληγή.
1168. πρόπυργοι θυσίαι. ‘Sacra pro turribus facta, ie. pro salute urbis.
Minus recte Stanleius, sacrificia antemoenialia’, says Blomfield ; he has been
followed by most scholars and by Wilamowitz explicitly on Pers. 860 (ὑπὲρ
τῆς πόλεως"); cf. also Williger, Sprachl. Unters. 5, who does not, however,
produce any arguments. Perhaps it would be more prudent to follow Head-
lam, who adds to the translation 'for his walls' the alternative 'or "before his
walls" ' (W. Sewell, Kennedy, Davies, L. Campbell translate ‘before’). There
was no need to say explicitly that the sacrifices of Priam were made for Troy;
and in Cassandra's nostalgic reminiscences it would not be unnatural that a
single feature of the scene (‘before the walls’) should stand out in her memory.
van Heusde, who translates 'sacra ante moenia facta in campo', compares
the epic fragment (in our texts of the Iliad it is © 548, 550-2) quoted in [Plat.]
Alcib. ii. 149 d; it has nothing to do with Homer, but is presumably older
than Aeschylus (Leaf, ad loc., and Wilamowitz, Die Ilias u. Homer, 3o n. 1,
think of a cyclic epic). There a hecatomb is offered to the gods by the Trojans
who are encamped before the city.
‘1169. πολυκανεῖς : this and δορικανής (A. Suppl. 987, a brilliant emendation)
are the only known compounds with -«av- (Buck and Petersen, A Reverse
Index, 721).
1170. ἐπήρκεσαν. Blomfield’s alteration ἐπήρκεσεν was disposed of by Her-
ı In Pers. 425 some of the MSS have θραύσμασι, but others, including M, have θραύμασιν.
For the development of such variants, which are discussed fully by Lobeck on S. Aj. 704
(Opadoua θραῦμα, p. 267 of the 3rd edit.), cf. Schwyzer, Griech. Gramm. i. 523 f.

532
 COMMENTARY line 1171

mann, who pointed out that ᾿ἐπαρκεῖν suppeditandi significatu cum accusativo
etiam in prosa oratione construitur’. As regards prose, Dindorf’s Thesaurus, iii.
1432 B, quotes, besides Plat. Prot. 321 a and Laws, 757 b, several passages from
the fourth century. But there is also an example from contemporary lyric:
Passow, s.v.,* Dindorf, loc. cit., and Paley, ad loc., compare Pind. N. 6. 6o,
where the poet says: ‘I announce πέμπτον ἐπὶ εἴκοσι τοῦτο... εὖχος ἀγώνων
dro... Ἀλκιμίδἄ τέ γ᾽ ἐπαρκέσαι (this reading, suggested by P. Maas, Re-
sponstonsfreiheiten, i. 15, has been adopted by Wilamowitz, Pindaros, 397 n. 2,
Schroeder, and, in substance, by Turyn) κλειτᾶι yevede’, ‘that you, Alcimidas,
have supplied the twenty-fifth victory for your family’. The meaning of the
construction in Ag. 1170 is clear :* ‘yet they availed not for any cure to save
my country from her present plight’ (Headlam).
1171. G. Hermann, De ellipst et pleonasmo (1808), 222 (= Opusc. i. 235),
inserted od after μή, but later revoked this emphatically. In his appendices
on Vigerus, De idiotismis, 800, he contrasts this passage with 5. Aj. 727 f.
and deduces that when the infinitive points to a future event, μὴ οὐ is used,
but when, as in Ag. 1171, the event has actually taken place, μή alone is
used. This distinction was adopted by Wecklein, Stud. z. Aesch. 19, but later
(Commentary and Appendix) he again felt uncertain. If μή alone is really
sufficient here (cf. the passages in Kühner-Gerth, ii. 218 n, though some of
them are not of the same kind) the reason should not be sought in Hermann’s
excessively subtle differentiation. On this particular point? it is impossible
to rely on MSS: ‘saepius in huiusmodi locis librariorum negligentia excidit
ov’ (Hermann, Oßusc. i. 234) ; this has happened in several MSS in the passage
quoted in n. 2 below, viz. Prom. 918, and also in Prom. 627, 787 (in both these
passages the first hand of M is among those which have μή alone), Eum. 914,
E. Hipp. 658, Phoen. 1176, etc. In Prom. 1056 even Wilamowitz, who was
satisfied with μή in Ag. 1171, follows Wecklein in inserting οὐ. And it is
Prom. 918 which makes me inclined to believe that here too we should read
μὴ od (so Headlam). Wecklein’s objection ‘that the poet avoided crasis in the
choral lyrics’ is irrelevant in the iambic trimeter (cf. on 1116).
ὥσπερ οὖν ἔχει παθεῖν. We do not know whether Triclinius found ἔχει in his
exemplar or whether (more probably) he made the slight emendation himself
from ἔχειν (FG). In any case the reading ἔχει is wholly satisfactory. I cannot
! He was followed, with a slight modification for the worse, by the older editions of L-S,
whereas in the ninth edition the passage is wrongly explained under III. 2.
2 It seems to me that the ἐπαρκεῖν in Prom. 918, which in the dictionaries is placed under
a different head, should be taken similarly and that there (οὐδὲν yap αὐτῶι ταῦτ᾽ ἐπαρκέσει τὸ
μὴ od πεσεῖν ἀτίμως) the construction is the same, including the infinitive used substantiv-
ally ; i.e. the οὐδέν there is equivalent to the οὐδὲν ἄκος in Ag. 1170.
3 Moorhouse, in his thoughtful article ‘The Construction with μὴ ob’, C.Q. xxxiv, 1940,
70 ff., takes note theoretically of the inconsistency in the MSS, but in practice does not
draw the necessary conclusions when he establishes the principle (p. 75 f.) ‘In editing texts
the MSS. evidence should be the only factor to consider in this particular, and μὴ οὐ or μή
is to be printed according to that evidence. . . . It is hardly to be expected that the MSS.
will have preserved the original presence or absence of οὐ with perfect fidelity ; but there is
at the present day no point in trying to improve on their testimony on this subject.’ The
effects of this conservative attitude may be illustrated by one example at least. In Eum.
914 Moorhouse agrees with the editors in accepting μὴ οὐ, for this is the reading in M, though
οὐ is omitted in F Tr. In Ag. 1171, on the other hand, he rejects the addition of (od) on the
ground that it does not appear in ‘the manuscripts’; but the MSS here consist only of
F(G) Tr, i.e. precisely those which, as Moorhouse agrees, mistakenly omit οὐ in Zum. 914.

533
line 1171 COMMENTARY

understand the objections raised by Wecklein and others. Wecklein first

notes appositely: ‘a form of expression which by its reticence hints at the
worst, cf. 1287 πράξασαν ὡς ἔπραξεν, S. Oed. R. 1376 BAaorodo’ ὅπως ἔβλαστε,
Oed. C. 273 ἱκόμην ἵν᾽ ἱκόμην, 336 εἴσ᾽ οὗπέρ efor’, etc. (there follow parallels
from Euripides ; cf. furthermore Cobet, Miscellanea critica, 43; A. C. Pearson
on E. Hel. 718 and on Soph. fr. 618; Denniston on E. El. 1141; and, e.g.,
Demosth. 3. 8 ἐχόντων μὲν ὡς ἔχουσι Θηβαίων ἡμῖν; the phrases mentioned
above on 67 are of the same kind). In the appendix, however, Wecklein says:
‘as can be seen from the examples quoted above, this form of speech demands
the same verb,’ and moreover παθεῖν is not suitable to ὥσπερ.᾽ This only
touches the surface of the matter. Every Greek would feel that the result
of a κακῶς παθεῖν is a κακῶς ἔχειν, and the substitution of an apparently
harmless ὡς or ὅπως or ὥσπερ for κακῶς is very suitable for this kind of
euphemism. We must of course understand from the preceding ὥσπερ ἔχει a
corresponding modal word (οὕτως) to go with παθεῖν : ‘(a means of preventing)
the city reaching the condition in which it is’. It is difficult to understand
how Headlam of all scholars? could tolerate such a serious disturbance of
word-order as τὸ μὴ οὐκ ἔχειν πόλιν μὲν κτλ. Obviously when Cassandra begins
τὸ μὴ (or μὴ οὐ) πόλιν μέν, she has already the intention of speaking about
herself immediately afterwards ; she does not, however, continue the infinitive
construction of the uév-clause but, with a perfectly natural break, makes the
öe-clause (ἐγὼ de... βαλῶ) independent, since she now feels overpowered by
the imminence of her own catastrophe. The function of the οὖν in this and
similar relative clauses of comparison, according to Denniston, Particles, 421,
is ‘to stress the correspondence between idea and fact, the objective reality
of something which in the main clause is merely supposed’.
Euripides (Tro. 1242) makes Cassandra’s mother say μάτην δ᾽ ἐβουθυτοῦμεν.
1172. ἐμ πέδωι βαλῶ : if the MS reading is correct (and I have no doubt of it),
then the words, however interpreted, must be taken as an instance of
‘tmesis’, i.e. = πέδωι ἐμβαλῶ (Verrall's defence of ἐμπέδωι has been as great
a failure as his interpretation of the whole line). As regards the preservation
of the final μ in front of , it is probably due to the belief of the copyist that
he was dealing with a form of ἔμπεδος (cf. on 387), as indeed in our MSS the
reading here is actually ἐμπέδω, which was first split into two words by
Casaubon. But what is the meaning of the phrase? Several scholars sup-
posed that βάλλειν was used intransitively; among them was Wilamowitz,
who wrote, Homer. Unters. 158: ‘Aeschylus read [A 423 f.] βάλλον ποτὶ yalyı
and understood it of falling to the ground’ ;3 he expressed the same view in his
ı The same rigid rule was set up by Headlam. He notes on his translation of Eum. 611
δρᾶσαι γάρ, ὥσπερ ἔστιν, οὐκ ἀρνούμεθα : ‘The MS gives ὥσπερ ἐστίν, which means nothing. It
was Greek to say δρᾶσαι γάρ, ὥσπερ ἔδρασα, or εἶναι γάρ, ὥσπερ ἔστιν, but not δρᾶσαι γάρ,
ὥσπερ ἐστίν. Davies restored εἶπον, but the editors have disregarded it.’ These polemics are
directed presumably against Sidgwick's note : ‘wozep ἐστίν, i.e. “as done it is"'. Blass, who
takes a similar view, compares Ag. 1171. I consider this interpretation correct and find
in it a special beauty of the passage. Only with horror can Orestes recall his deed; so the
addition of this veiled allusion (similar in function to the passages quoted above) fits
excellently. If ὥσπερ ἐστίν (or ἔστιν) meant only ‘that is true’ (Wilamowitz) or ‘as the truth
is’ (Verrall), its addition in this passage would be quite pointless.
2 It is probably due to the influence of Headlam's note that a freak of Stadtmuller's (his
judgement was hardly ever sound) disfigures the apparatus of the Oxford text.
3 Before Wilamowitz Lewis Campbell, Am. Journ. of Philol. i, 1880, 437, quoted this

534
 COMMENTARY line 1172

commentary on Cho. 574; cf. his note on Eum. 751, Verskunst, 151, and also his
general remark on E. Jon, 929: ᾿βάλλειν and its compounds are altogether
more commonly intransitive than is generally realised.' Blomfield had taken
the same view here ('praestat, opinor, intellegere éuaurÿr’),' and still earlier
Pauw. The latter explains: ᾿βαλῶ pro βληθῶ, activum pro passivo [this is
discussed below] ut saepe; Ego . . . brevi in solo proticiar.’ But when we look
at the evidence more closely, it turns out to be rather disappointing. Not a
single one of Wilamowitz's examples of the intransitive use of βάλλω and its
compounds is really certain. It was Wilamowitz's own pronouncement that
we do not understand? βάλλον in A 424; as for the rest, one only needs to
read a few commentaries on Cho. 574, Eum. 751 (where βαλοῦσα seems beyond
repair), and on E. Ion, 929; on this last passage Wilamowitz quotes E. El. 96,
but Denniston (ad loc.) is right when he says that Dobree's πόδα there is
excellently supported by fr. roro N. (where the χθόνα of the Apollonius scholia
is inferior to the πόδα of the lexicographers). I also agree with Denniston's
statement that the alleged intransitive use of ἐκβάλλειν cannot be justified?
by ἐκβάλλειν used of rivers, or by the intransitive εἰσβάλλειν (e.g. S. El. 719,
cf. also Bekker, Anecd. 112. 32, παραβάλλεις εἰς τὸν τόπον' ἀντὶ τοῦ φοιτᾶις
συνεχῶς), or βάλλ᾽ ἐς κόρακας (and similar phrases, cf. Neil on Ar. Knights,
1151). For Ÿ 462 see Leaf, ad loc. But even if we considered it possible that
βαλῶ or ἐμβαλῶ is used intransitively here, we should not be out of our diffi-
culties. For an ‘intransitive’ use of this kind implies a certain activity on the
part of the subject, as is unmistakabiy the case with the intransitive use of
εἰσβάλλειν and ἐμβάλλειν (‘attack’) ; cf. also the intransitive use of ῥίπτειν, ‘to
fling oneself’, illustrated by A. C. Pearson on E. Hel. 1325. But here we can-
not imagine that Cassandra throws herself on the ground or anything of the
kind, but only that something is done to her, that she passively suffers it.
This makes it clear, from the point of view of the sense, why Pauw explained
‘activum pro passivo’; but that cannot be justified linguistically. There are,
then, no very good grounds for taking the verb here as intransitive. However,
that assumption is more credible than the conjectures proposed, e.g. the
palaeographically elegant ἐμπελῶ βόλωι, which, being a combination of earlier
conjectures (ἐμπελῶ Hermann, βόλωι Bothe), was suggested by Ahrens (p.
620).* This proves impossible when we remember that Aeschylus nowhere
passage in the Nekyia to support the view that βαλῶ in Ag. 1172 should be taken intransi-
tively.
I This is adopted by L-S, s.v. βάλλω A. TH. 1.
2 Cf. also Von der Muhll’s note on A 423 ff.: ‘obscuri'. However, I admit the possibility
(though it seems to me to be a remote one) that Aeschylus modelled Ag. 1172 after A 423 f.
αὐτὰρ ἐγὼ ποτὶ yalyı... βάλλον and that he took βάλλον to be intransitive.
3 [n his short article ‘Intransitives BaAdew’, Rhein. Mus. 66, 1911, 626 ff., Stahl treats all
alike the Homeric eis dAa βάλλων and βάλλ᾽ és κόρακας and βαλῶ Ag. 1172 (about which he
comes to the same conclusion as Blomfield and Wilamowitz). In the case of Cho. 574 and
Eum. 751 he remains extremely sceptical, and very rightly so; but in E. Cycl. 574, in order
to gain another instance of the intransitive βάλλειν, he repudiates the self-evident emenda-
tion βαλεῖ. He, like others, was unable to produce any further example of this intransitive
use in older Greek.
+ Ὁ, Thomson ascribes Ahrens's conjecture to Headlam. He also ignores H. A. J. Munro's
discussion of the passage, 7. Phil. xi, 1882, 136 ff. (Sidgwick refers to it in his annotated
edition of the 4g.) : Munro, too, without knowing about Ahrens, conjectured ἐμπελῶ βόλωι.
In his treatment of the preceding line Thomson copies from Wilamowitz’s apparatus the
slip of the pen or printer's error ἔχει ἔχειν F (instead of ἔχει Tr ἔχειν F).

535
line 1172 COMMENTARY

suggests that Cassandra, too, is ensnared in the garment, the ‘net’, which is
destined for Agamemnon alone. And there can be no thought of a purely
metaphorical mention of a net in the midst of this tale of murder, where a real
net, the festal robe, plays a decisive part. In support of πέδωι I should like to
quote Lycophron, 1108, where Cassandra says ἐγὼ de δροίτης ἄγχι κείσομαι πέδωι,
but there does not seem to be any proof of its dependence on this passage.
Next θερμόνους must be examined. This word, which occurs only here, is
apparently analogous to the Euripidean (fr. 858 N.) ὦ θερμόβουλον σπλάγχνον,
a bold phrase ridiculed by Aristophanes ; it may also be compared (Hermann)
with S. Ant. 88 θερμὴν ἐπὶ ψυχροῖσι καρδίαν ἔχεις. This would mean that
θερμο- here, following the familiar usage, denotes ardour as opposed to a
cool head. A similar meaning was suggested by Pauw: ‘calida et accensa
mente, ut prophetissa’, and by Blomfield: ‘mentem inflammata, sc. afflatu
divino’ ; among later scholars, e.g., Nägelsbach comments: ‘furore, vaticinio
percita, explicatur ex v. 1256’ and Headlam ‘with fevered brain'.' It is not
permissible to retain θερμόνους and yet translate: 'noch wallt mein Blut:
bald lieg’ ich kalt am Boden’ (Wilamowitz) ; this looks like a compromise
with attempts made by earlier scholars to read θερμόν (either alone or with
the addition of a noun). It may be admitted that the prophetess seized by
the passion of her inspiration could perhaps be called 8epuóvovs ; although the
depreciatory tone which θερμόν has when applied to spiritual and mental
qualities and which it has in the phrases quoted above is quite unsuitable
here. But I fail to see how this trait can be at all relevant in this context,
where her wretched fate is presented to the audience in a terse word-picture,
cruelly realistic (ἐν réduc βαλῶ), which arouses pity and is to some extent
parallel to the destruction of Ilium. What this context seems to demand is
heard from Cassandra in 1278: θερμῶι κοπείσης . . . προσφάγματι (Paley quotes
it here). This brings us back to the position of the older editors who regarded
θερμόνους as corrupt, retaining θερμόν from it. Of their proposals the best is
probably Musgrave's θερμὸν ῥοῦν. Cf. 209 f. παρθενοσφάγοισιν ῥείθροις. There
is no objection to the monosyllable ῥοῦν in the trimeter, cf. Prom. 852 πλατύρ-
ρους, Aesch. fr. 300. 2 N. Νεῖλος Errrapovs, Soph. (?) fr. 649. 39 P. Ὁ — v xaAJAC
pov» ἐπ᾽ ‘AAdetod πόρον. Nor can any objection be raised to the use of βαλῶ (or
rather ἐμβαλῶ) in this connexion, as was pointed out as early as 1837 by
Fr. Martin (quoted in Hartung's commentary) ; he compared the line from
Aeschylus' Pentheus (fr. 183 N.) μήδ᾽ αἵματος πέμφιγα πρὸς πέδωι βάληις (for
the text cf. Wenkebach, Philol. Ixxxvi, 1931, 320). If we accepted Musgrave's
conjecture (or a similar one), there would be no need to take ἐμβάλλειν
intransitively.
1173. ἑπόμενα: it ‘follows’ what went before, i.e. does not contradict it, but
conforms to it and agrees. Cf. Pindar, Ol. 13. 47 ἔπεται 8’ ἐν ἑκάστωι μέτρον, on
which Boeckh remarks: 'aptus et conveniens certus unicuique rei modus est ;
ut alibi ἔπεσθαι absque ἐν: v. ad Olymp. 2. 22 sqq.’
The conjecture ἐπεφημίσω, adopted by Kennedy, Wecklein (in his anno-
tated edition), Verrall, Blaydes, and A. Y. Campbell, was first put forward
τ Plüss gives the same meaning: ‘whose eye [Ὁ] and mind are deviated by passion’ ; his
attempt to interpret βαλῶ transitively as ‘a picture of the marksman, whose spear, avoided
by his opponent, falls into the ground' is fantastic, like a good many other ideas in the work
of this scholar, who lacked neither ingenuity nor learning.
536
 COMMENTARY lines 1174 ff.

by Paley, who gave the following reasons: ‘not only because the ἐπὶ is
singularly appropriate to the idea of a second declaration, over and above
the former, which is described by the simple ἐφημίσω in the strophic verse—
but because it appears a better metrical correction than προτέροισι, which
editors have adopted from Pauw’. As for the latter argument: it is quite
common in Aeschylus for one syllable to run over the end of the dochmiac
(not to mention other types of the non-coincidence of the end of the metron
with the end of the word). This occurs not only 1129 δολοφόνου AéBn[ros
(= 1118 κατολολυξάτω), where two cretics follow the dochmiac, but also
where the continuation is in dochmiacs, e.g. Suppl. 392 = 402, 437 τάδε
φράσαι" δίκαια Διόθεν κράτη (= 432 πολυμίτων πέπλων τ᾽ ἐπιλαβὰς ἐμῶν), Sept.
155 = 163, 239 moridarov! κλυοῦσα πάταγον ἄμμιγα (= 233 διὰ θεῶν πόλιν νεμό-
μεθ᾽ ἀδάματον), Cho. 156 κλύε δέ μοι, κλύε σέβας (Hermann’s transposition is
unjustified), 954 f. (a likely restoration of Meineke’s, cf. above on 785), 960
ἄξια δ᾽ olpavoüxov ἀρχὰν σέβειν (= 971 μέτοικοι δόμων πεσοῦνται πάλιν), 968
καθαρμοῖσιν ἀτᾶν ἐλατηρίοις. In the last example the MS reading is καθαρμοῖς
and the missing syllable was supplied by Hermann, just as in Ag. 1173
προτέροισζι the final syllable was supplied by Pauw.’ It is clear from several
of the passages quoted that when a syllable overruns the end of the dochmiac
in the strophe, this need not happen again in the antistrophe. Paley’s other
argument, too, the linguistic, is open to objection. The sense which seemed
to him to recommend the ém- here,’ and which is so frequently found in the
Homeric ἐπεστενάχοντο and the like, is not known in the case of this particular
verb ἐπιφημίζομαι (Hdt. 3. 124. 2 ‘utter words ominous of the event’, L-S)
and its noun ἐπιφήμισμα (Thuc. 7. 75. 7). So it is clear that the minute change
of προτέροις to προτέροισι is decidedly preferable.
1174 ff. καί ris σε... θανατηφόρα. Up to the end of the eighteenth century
this sentence was printed and translated as a question ;* and it was still
believed to be such by some editors at a considerably later date. But Butler,
Blomfield, Hermann, and many after them regarded τις as indefinite and the
sentence as affirmative; they are right. It is clear from καί that this is not a
question. The examples of καί preceding the interrogative pronoun or
adverb (Denniston, Particles, 309 ff.) belong to types of sentence with which
this one has no similarity. The relation of the two sentences and their link with
the one before are quite clear. ‘In accordance with your earlier words have
you spoken this, and some god ill-disposed, by whom you are possessed,
makes you sing of death-bringing woes.’ In other words: ‘This shows the
same heaven-sent madness as your former speech.’ The last stanza of the
Chorus therefore corresponds exactly to the former one (1140 ff.), φρενομανής
τις εἶ, θεοφόρητος, ἀμφὶ δ᾽ αὑτᾶς θροεῖς νόμον ἄνομον. The Elders now give the
case up as hopeless, it goes beyond their understanding: τέρμα δ᾽ ἀμηχανῶ.
1 Heimsoeth’s conjecture; this may not be the correct reading in place of the corrupt
moraivıov, but κλυοῦσα, which is the important word here, is sound.
2 In Ag. 1536 also, Pauw is right with θηγάναισζ». It is well known that the final ı quite
frequently drops out.
3 Weil, too, who independently of Paley proposed the same alteration, notes ‘fere ut
Choeph. 457 ἐπιφθέγγομαι".
4 There was therefore no need for Α. Y. Campbell to distinguish xoi ris σε by the asterisk
which he, more Housmani, in his app. crit. makes to shine upon his own conjectures, for in
the MSS and the older editions the words are accented in the same way.

537
lines 1174 ff. COMMENTARY

In the context of this thought, which serves as a powerful foil to what comes
next, καὶ μὴν 6 χρησμὸς οὐκέτ᾽ ἐκ καλυμμάτων ἔσται δεδορκώς KrA., there can be
no point in their repeating the question πόθεν... θεοφόρους éxew . . . δύας
(1150) in another form. The vagueness (τις) of the reference to the δαίμων is
in keeping with the view, current from Homer on, according to which an
event was attributed to ‘some god or other’ (θεός τις, cf. also Ag. 663) when
its cause was hidden from man; for the particular case in point here, that of a
woman possessed, we may perhaps compare E. Hipp. 141 ff. 4j σύ γ᾽ ἔνθεος, ὦ
κούρα, εἴτ᾽ ἐκ Πανὸς εἴθ᾽ ‘Exdras 7) σεμνῶν Κορυβάντων φοιτᾶις ἢ ματρὸς ὀρείας:
there a sequence of possible deities have taken the place of θεός τις or τις δαίμων.
1174. According to the dictionaries kaxogpoveiv occurs only here in Greek
literature. Perhaps it would be more prudent not to register any form of the
verb at all except the participle, which was presumably formed as an opposite
to the eübpovewv (or ἐὺ φρονέων) current from Homer on (‘the verb εὐφρονέω
is not found’ L-S). With regard to the emergence of κακο- compounds as
opposites to εὖ- compounds cf. Wackernagel, Glotia, xiv, 1925, 51. We may
perhaps conclude from Aeschylus’ using κακοφρονῶν that he took the epic
ἐυφρονέων to be one word. Another possibility, although less likely, is that
he arrived at κακοφρονῶν from κακόφρων ; it would answer.to the rare ἀφρονῶν
beside ἄφρων and to σωφρονῶν beside σώφρων.
1175. ὑπερβαρής : ‘adjectival -us was (apparently) replaced in compounds by
15’, Schwyzer, Griech. Gramm. i. 449. For the idea underlying this word Pers.
515 f. has long since been compared: ὦ δυσπόνητε δαῖμον, ὡς ἄγαν βαρὺς
ποδοῖν ἐνήλου παντὶ Περσικῶι γένει. Cf. A. C. Pearson on E. Phoen. 1557 and
the note on 1660 below.
ἐμπίτνων : a lively representation of the manner in which a daemon takes
possession of the person or thing over which he achieves mastery. Cf. on 341
and also Cho. 36, where the daemon of the baneful dream, the δόμων ὀνειρό-
pavrıs, is depicted as γυναικείοισιν ἐν δώμασιν βαρὺς (corresponding to the
ὑπερβαρής here) πίτνων.
1176. θανατηφόρα. The form of the word in Tr is probably correct. Not only
is this particular word always θανατηφόρος (cf. Lobeck, Phrynichus, p. 651),
but it is also true in general ‘that whereas -η-φόρος is used after a long syllable
as well as a short, -ο-φόρος is apparently never found after a short syllable’
(Wackernagel, Dehnungsgesetz, 11). It is hardly possible to raise objection
to the long syllable on account of the responsion yoepà θανατηφόρα = μινυρὰ
θρεομένας, cf. on 1128. However, the possibility that the poet is here allowing
himself an arbitrary word-formation cannot be absolutely excluded."

Here, at the end of the first (lyrical) part of the Cassandra scene,? we may
pause to add a few general observations to what has been said in the notes.
! In spite of this consideration I would not venture to follow Wilamowitz in reading
θαλαμοπόλων in Sept. 359 against the evidence of the MSS (plentiful there) and despite the
fact that the word has always the form θαλαμηπόλος elsewhere, from Homer on. Certainly
Wilamowitz (who reads δόρει) cannot be refuted on the grounds that 8opi is the MS reading
in the strophe (347); in Suppl. 846, too, the metre requires δόρει (this oldest instance of
δόρει guaranteed by the metre is missing in the list in L-S, δόρυ, line 8 f.). The whole passage
in the Septem (from 1. 356) still awaits convincing restoration.
2 For a fuller exposition see my essay Die Kassandraszene der Orestie (Stuttgart 1937),
which has to be corrected in several details.

538
 THE LYRICAL PART OF THE CASSANDRA SCENE
In the preceding scene (1035-71) the sustained silence of Cassandra was very
much emphasized. Now, after the departure of Clytemnestra, the mute and
motionless figure on the wagon suddenly stirs, and breaks, not indeed into
speech or song, but into something between a song and those wild notes of
lamentation which were familiar to the Athenians from the ritual perform-
ances of the barbarian mourning-women from the East. What we hear are
not, of course, crude and formless cries but sounds ennobled by rhythms of
Hellenic music: still, they are distinct from articulate language. The first
real word that reaches our ear is the name of Cassandra’s divine overlord. He
is to remain in the centre of her thought throughout the long scene until at
last she hurls his sacred sceptre and fillets and garment to the ground and
thus finally tears herself away from him and his power (1264-76).
First Cassandra invokes the god who sways her whole being from within.
Then, as she steps down from the wagon, she sees his stone symbol in front
of the house and intensifies her invocation: ἀγυιᾶτ᾽, ἀπόλλων ἐμός. Only now
does she continue with a complete, explanatory sentence: ἀπώλεσας γὰρ κτλ.
Thus the gradual transition from torpor to involuntary disconnected cries,
then to articulate supplications, and finally to full speech is worked out in
the most impressive manner. It is possible, though not provable, that the
change from the lyrical metres to the iambic trimeter at 1082 marks a calming
down, in keeping with the more rational content of the line. I do not see
how we can decide whether the iambic trimeters in Cassandra’s stanzas were
sung exactly like the purely lyrical parts or were recited or delivered in a
manner between singing and reciting. Be this as it may, the lyrics greatly
prevail, and they, and especially the dochmiacs, point to highly emotional
gestures and movements (dancing is an awkward word) of the actor.
The coryphaeus, on the other hand, displays a detached reserve. He does
not sing but speaks, nay argues (1088 f., 1098 f., 1xo5 f.). There is a sharp
contrast between the excitement of the prophetess and the calm rationality
of the Elders, a contrast which finds its expression in the gestures no less than
in the words. But the aloofness of the old men cannot be maintained for
long. The power of the visions and revelations is too strong for them. They
know only too well the doings of that Erinys, Στάσις, in the house (cf. 154 f.),
and when Cassandra mentions the evil Spirit, they are horror-stricken. In-
fected, as it were, by the ecstasy of the unfortunate maiden, they are swept
off from their cool moderation and rushed into the excitement of dochmiacs
and wild movements (1121 ff.). The emotional elements are increased in the
next stanzas of the Chorus: there no trimeters are left at all. So the scene
reaches a climax of passion and movement immediately before the serene
notes of Cassandra’s speech (1178 ff.) take us into a completely changed
sphere.
E Cassandra's songs the lines 1136 f. clearly mark the transition to a
different theme: τὸ yàp ἐμὸν θροῶ πάθος. In the preceding vision Agamem-
non's fate has been revealed up to the moment of the actual murder (1128) ;
from now onwards to the end of the lyrical part the lamentations of the
prophetess are solely concerned with herself and her kin. But it is not only
the subject that has changed. Everything that Cassandra has said about the
disasters in the house of the Atridae from the slaughter of Thyestes' children
to the murder of Agamemnon was conveyed to her through real visions; we
539
lines 1072-1177 COMMENTARY

have seen that her words nowhere go beyond a description of what at any one
moment is being disclosed to her visionary mind. This particular type of
trance ends with the death of the king. What follows is not vision but
prophecy (1139 ξυνθανουμένην, 1149 ἐμοὶ δὲ μίμνει krÀ., 1160 f. νῦν 86 . . . ἔοικα
θεσπιωδήσειν, 1172 ἐγὼ... βαλῶ). It is likely that this different mental atti-
tude is emphasized by the change in the metrical form. Each of Cassandra’s
three pairs of stanzas from 1136 to the end of the lyrical part concludes with
two iambic trimeters. Here it seems worth while quoting a few sentences from
Henry Weil’s fine appreciation of the formal structure of the scene (Études
sur le drame antique, 270 f.). After describing the modification of the typical
form that takes place at 1121 (‘effrayé par la persistance des lugubres prophé-
ties, le chœur commence ἃ s’émouvoir, et son distique iambique est suivi de
vers chantés’), he continues: ‘Puis Cassandre commence à se recueillir : tout
en continuant de chanter ses visions [this has been corrected above], elle les
fait toujours suivre d’un distique déclamé; les deux trimétres passent du
chœur à la voyante. C'est ainsi que se prépare la prophétie claire et suivie en
trimètres non cadencés.'

1178ff. ‘Veramente oramai saranno nude le mie parole, quanto converrassi

quelle scovrire alla tua vista rude' says Beatrice (Dante, Purg. 33. 1oo ff.).
1178. καὶ μήν. Denniston, Particles, 355, includes this passage among his
examples of the ‘inceptive-responsive’ use of καὶ μήν, where ‘a person who
has been invited to speak expresses by the particles his acceptance of the
invitation: “Well”, “Very well”, “All right".' Denniston (with Paley)
regards Ag. 1178 as a ‘response to an implied appeal for plainer speaking',
implied, that is, in τέρμα δ᾽ ἀμηχανῶ. I do not know whether in these words
or in the last reply of the Chorus as a whole any such appeal is implied : all I
am able to perceive is the ring of complete resignation. It is perhaps another
argument against Denniston's interpretation that in the instances he quotes
the appeal is explicitly expressed. So I should prefer to see here an example
of the «ai μήν which marks a new stage (Denniston, 351 f.). The familiar use
of καὶ μήν in announcing the appearance of a new character on the stage (cf.
Denniston, 356) is only a special case of this type: it is quite customary in
Euripides and Sophocles, but there is an earlier example in A. Sept. 372 f.,
though not yet stereotyped into a formula. A good example of καὶ μήν when
the speaker introduces a new fact is provided by Prom. 1080, where Prome-
theus begins: xal μὴν ἔργωι κοὐκέτι μύθωι χθὼν σεσάλευται. There the contrast
with what went before (ἔργωι κοὐκέτι μύθωι) is indicated as in the present
passage (οὐκέτ᾽ ἐκ καλυμμάτων... λαμπρὸς δέ). There is something else which
marks the contrast with the preceding part of the scene more strongly than
do the words : as speech replaces song, not only does the music cease but the
movements of Cassandra and the Chorus, and indeed their whole appearance
on the stage, undergo a profound change.
1179. νεογάμου νύμφης δίκην. With a reference to Harpocration, s.v. dva-
καλυπτήρια, Casaubon and Stanley briefly explained the custom, which is
familiar also from Greek works of art. It is unnecessary to quote books of
reference.
1180. λαμπρός. À substantial explanation was given (after Porson and
Dobree) by Blomfield; cf. now L-S, s.v. I. 5. We may also compare λευκόνοτος.
540
 COMMENTARY line 1182
The word λαμπρός means 'clear, uncovered’ and at the same time when used
of the wind describes its force ;! so it forms here a connecting link between the
two heterogeneous ideas, (a) covering, (b) the wind atsunrise. If weattempt to
trace the evolution of the poet’s thought, we may infer with a certain degree
of probability that here the word λαμπρός first took shape as an opposite of
the idea contained in ἐκ καλυμμάτωνΣ and then gave rise, by virtue of the
current phrase λαμπρὸς ἄνεμος, to the extension of the image in ἡλίον πρὸς
ἀντολὰς κτλ. This process of evolving a coherent image out of the notion con-
tained in a single word is analogous to what has been observed in the note on
387 ff. The image is furthermore enriched with the idea that the prophecy,
which so far remains obscure, will in the course of time be clarified. Instead
of ‘in the course of time’ we have here ἡλίου πρὸς avr., because emphasis is
laid upon the contrast between the time when it grows light and the obscure
and obscuring present. The thought is the same in 250 ff. τὸ μέλλον ἐπεὶ
γένοιτ᾽ ἂν κλύοις (we do not yet know what the future will bring) . . . τορὸν yap
ἥξει σύνορθρον αὐγαῖς.
1181. πνέων belongs only to the sphere of the image; we should be careful
not to introduce the idea of ‘inspiration’ as Plüss and Headlam do.
There is no certain instance of εἰσήκειν, at any rate not in literary language,
before the time of the Empire. Editors of Ar. Wasps (606), starting from the
reading κἄπειθ᾽ (RI), have long since adopted κἄπειθ᾽ ἥκονθ᾽, In regard to
the present passage Ahrens (p. 622), following other scholars, demonstrated
the necessity of Bothe’s ἐσάιξειν and collected instances of ἄισσειν and its
compounds (e.g. Pind. Isth. 4. 6 οὖρος... &raioowv) used of the movement
of the wind.
1182. κλύζειν was restored with certainty by Auratus and, independently, by
Schütz. In view of & 392 (= ı 484, 541) ἐκλύσθη δὲ θάλασσα it would be pos-
sible to take the verb as transitive here, having as subject the wind (which
has taken the place of the χρησμός) ; this is the view of Nägelsbach, Headlam,?
and Mazon. But the whole sentence becomes much more forceful if we follow
Schütz and the majority of commentators in taking κλύζειν intransitively,
cf. Ψ 61 ὅθι κύματ᾽ ἐπ᾽ ἠϊόνος κλύζεσκον. Then the subject of κλύζειν is πῆμα (or,
as the preceding κύματος δίκην colours the thought of the whole passage,
κῦμα πήματος) τοῦδε πήματος πολὺ μεῖζον. Schütz presumes that Aeschylus
was influenced by 4 422 ff., where the west wind (here indirectly indicated) is
specifically mentioned ; this is possible, but not certain.
πρὸς αὐγάς, which in the past was heavily attacked, is now mostly left in
1 Blomfield gives an apposite illustration : ‘Virg. georg. x, 460 et claro silvas cernes aquilone
moveri. Nempe sic dicitur ventus vehemens, quia obscuro deterget nubila caelo (Horace,
Odes, τ. 7. 15). unde vis epitheti in hoc loco perspicitur. It may well be that clarus in the
passage of Virgil is a ‘Bedeutungslehnwort’ precisely after the model of λαμπρὸς ἄνεμος
(such reinterpretations play a great part in Virgil, cf. Ernst Steiner, Das Bedeutungslehn-
wort in Vergils Aeneis, Diss. Königsberg 1921). Virgil was not, however, the first to use
clarus in that sense: R. Pfeiffer, Philol. Ixxxvii, 1932, 189, has shown that Catullus, 66. 44
progenies Thiae clara xefers, not to the sun, but to the North wind.
2 That the word offered itself easily in this antithesis is clear from, e.g., Prom. 833 (of the
oracle, as here of the χρησμός) λαμπρῶς κοὐδὲν αἰνικτηρίως (cf. Ag. 1183), and particularly
Cho. 810 f. λαμπρὸν . . . ἐκ δνοφερᾶς καλύπτρας.
3 But in the copy of his prose-translation which he sent to Housman (see on l. 6 above)
he underlined the words ‘so as to wash up’ and wrote on the margin: ‘rather intransitive
with πῆμα subject, as Hom. 4 422, V' 61’.

541
line 1182 COMMENTARY

peace.’ Observation of nature provides the best justification. The point is

prettily illustrated by S. Butler (quoted in Peile's commentary) ; he refers to
one of his English predecessors: 'ait ille singularem huius metaphorae felici-
tatem illis ignotam fore, qui fluctus quasi ad solem orientem se convertere
videri nescierint. De hoc iudicent, qui solis ortum inter navigandum viderint.'
The passage in the Iliad (A 422 ff.) mentioned above contains nothing of the
atmosphere of the early morning which enlivens Cassandra's words and is
presented so vigorously in the simile of Catullus 64. 269 ff. (compared by
Wecklein). The Latin poet probably borrows his theme from Homer, but its
development is his own: hic, qualis flatu placidum mare matutino horrificans
Zephyrus proclivas incitat undas Aurora exoriente vagi sub limina Solis, quae
tarde primum clementi flamine pulsae procedunt, leviterque sonant plangore
cachinni, post vento crescente magis magis increbescunt purpureaque procul
nantes ab luce refulgent. The words ἡλίου πρὸς avroAds and πρὸς αὐγάς depict
the direction of the morning wind blowing from the west and the increasing
movement of the waves to the east, and these expressions also point to the
light which after the darkness of her cryptic words is now about to arise
upon the pronouncements of the prophetess.
τοῦδε πήματος : Hermann explains ‘suum dicit’, but this seems to miss the
special point. In this passage (from 1178 on) the main topic throughout is the
difference between the revelations now beginning and Cassandra's earlier
utterances. So she can hardly be referring to the contrast between her own
fate and that of Agamemnon, for both form the subject of the trimeter
section which now begins, as of the lyrical part which is now concluded.
τοῦδε seems rather to mean the anguish she feels here and now, the anguish,
‚that is, which has found expression in her words up to this point, in contrast
to the more intense suffering that awaits her. When she now begins to speak
undisguisedly and no longer softens the force of her meaning by enigmatic
words, the horror of her revelations will emerge in all its nakedness, and the
anguish of speaker and audience will become more intense. The increasing
surge of the waves and the increase of anguish are bound together as a single
idea.
In 1178-83, apart from the double comparison (explained on λαμπρός 1180),
we perceive that mutual penetration of the comparison and the thing com-
pared which should not be unfamiliar to the student of ancient poetry; cf.
on 966. Nagelsbach (on 1178) referred to the ‘comparatio compendiaria’ here ;
see also Wecklein, Stud. z. Aesch. 8, and Vahlen, Opusc. 1. 300.
1183. φρενοῦν appears first in the later plays of Aeschylus (cf. Cho. 116,
Prom. 335), then it is taken over by Sophocles and Euripides, then found in
Xenophon. We should like to know whether it was coined by Aeschylus.
The phrases οὐκέτ᾽ ἐκ καλυμμάτων (1178) and οὐκέτ᾽ ἐξ αἰνιγμάτων frame in
the usual manner (cf. on 20) the section 1178-83, which serves as a prooemium
to the following scene as a whole.
1184, With what part of the sentence does συνδρόμως go? The problem is
closely related to that discussed on 1036. The majority of translators seem
to follow Stanley in taking συνδρόμως with ῥινηλατούσηι ; similarly Passow, s.v.
σύνδρομος, and, following him, L-S. On the other hand several scholars take
1 The idea that the words dore... αὐγάς might be an interpolation is, I hope, implicitly
refuted by the explanation which I have adopted.

542
 COMMENTARY lines 1186 ff.

συνδρόμως with μαρτυρεῖτε, e.g. Peile (though only as an alternative), Hartung,

Paley, Kennedy, Headlam, MacNeice, G. Thomson. But in construing the
sentence like this they seem not to realize that the adverb adds a character-
istic trait to the picture of the sleuth-hound, which here as in 1093 f. impres-
sively portrays the action of Cassandra. Blomfield is right on συνδρόμως :
‘metaphora a canibus venatricibus sumpta’. Good hunting dogs διωκέτωσαν
ἐρρωμένως Kal μὴ ἐπανιεῖσαι σὺν πολλῆι κλαγγῆι Kai ὑλαγμῶι, συνεκπερῶσαι
μετὰ τοῦ λαγῶ πάντηι ([Xen.] Cyneg. 4. 5), in contrast to the worthless dogs,
which (ibid. 3. 6) κύκλοις πολλοῖς χρώμεναι καὶ πλάνοις ὑπολαμβάνουσαι ἐκ τοῦ
πρόσω τὰ ἴχνη παραλείπουσι τὸν λαγῶ κτλ. (cf. also the rest of the chapter).
Mazon’s rendering, ‘sans écart’, is apposite (‘rendez-moi témoignage que, le
nez sur la piste, j'ai suivi sans écart la trace des forfaits anciens”). συνδρόμως
describes as clearly as anything can the συνεκπερᾶν μετὰ τοῦ λαγῶ πάντηι.
1185. The participle ῥινηλατούσηι is not, as Pluss takes it, the equivalent of
an infinitive construction with μαρτυρεῖν τινι, but is used ‘in the sense “if I
do it, bear witness (confirm it) to me" ' (Wecklein).
μαρτυρεῖτε: as might be expected from the introduction of the speech
(1178-83), the attitude of Cassandra is completely changed. Now for the first
time she turns deliberately to the old men of the Chorus. More than that,
she actually calls upon them to co-operate with her. She is no longer enrapt
in her visions but endeavours to prove her case on rational lines. To be
believed at all, Cassandra must first free herself from the suspicion that, like
so many, she is a ψευδόμαντις ; she must prove herself a true prophetess. From
time immemorial it was essential to a seer to possess complete knowledge of
τά τ᾽ ἐόντα ré τ᾽ ἐσσόμενα πρό τ᾽ eovra.! Τῇ he shows himself capable of reveal-
ing certain facts of the past which could not have been known to him in the
natural course of events, people are ready to trust his prophetic gifts. This
is the reason for the test to which Croesus subjected the different oracles
(Hdt. x. 46 ff.), ἐπινοήσας τὰ ἦν ἀμήχανον ἐξευρεῖν τε καὶ ἐπιφράσασθαι (48. 2).
The cheap objection that knowledge of the earlier horrors in the house of the
Atridae could have reached the ears of the Trojan maiden through rumours
would not occur to the poet or his audience, as may be seen from 1199 ff.
1186 ff. Only in the last word (1190 ᾿Ερινύων) of the long enigmatic period
does she make it quite plain what kind of choir she has in mind (cf. on 687);
but the horrible character of this κῶμος emerges at the beginning of 1189, and
earlier still the expressions used in 1187 imply disaster. Here the poet, with
magnificent simplicity, has erected one of the supporting pillars of his great
edifice. In this passage the choir of the Erinyes makes its entry into the
trilogy, which it is to dominate until the end. The tale of the monsters who,
surfeited with the blood of their victims, chant their sinister song looks for-
ward to the choruses of the Eumenides, in particular to the δέσμιος ὕμνος.
1 It is characteristic of the practice of divination among many peoples all over the earth
that the prophet is expected to have a knowledge of the present and the past as well as of
the future, cf. H. M. and N. K. Chadwick, The Growth of Literature, i. 451 ff., 473, fi. 581 f.
The subject is well summed up by Mrs. N. Kershaw Chadwick, Poetry and Prophecy
(Cambridge 1942), 41: ‘Inspiration relates to revealed knowledge. Revelation covers the
whole field of human consciousness. It includes knowledge of the past and the hidden
present, as well as the future. Revelation of the future is generally less important than is
commonly supposed. Among backward peoples perhaps knowledge of the present plays
the most important part, especially of the hidden present.’
543
lines 1186 ff. COMMENTARY

The image of the κῶμος, taken from the homely sphere of everyday life, has
here assumed an unexpected and dreadful meaning (cf. on 437 ff.). The most
joyous custom of Attic life is transformed into an object of horror. This
κῶμος does not go round from door to door like an ordinary one, but remains
resident in the house of the Atridae, and no one can send it away; the drink
which inebriates it is human blood, and the song that it sings is ξύμφθογγος
οὐκ εὔφωνος. In another context Aeschylus takes again a gruesome image
from the setting of the συμπόσιον : the eagle, which is going to rend the body
of Prometheus and feast upon his liver, is ἄκλητος ἕρπων δαιταλεὺς πανήμερος
(Prom. 1024).
1188. πεπωκώς : for the conception of the Erinyes drinking human blood cf.
Cho. 577 f., Eum. 264 ff., and also Pearson on Soph. fr. 743. The perfect
πέπωκα does apparently not occur in any earlier instance (Sept. 821 is probably
interpolated).
1189. κῶμος : cf. (Paley) E. Phoen. 351 ff. εἴτ᾽ ἔρις εἴτε πατὴρ 6 σὸς αἴτιος, εἴτε
τὸ δαιμόνιον κατεκώμασε δώμασιν Οἰδιπόδα.
ἐν δόμοις μένει: a variant with very slight divergence from 1186; cf. the
note on 154.
1190. δύσπεμπτος ἔξω : according to the customs of Attic society it was on
the whole not difficult to get rid of a κῶμος, whether it stayed for a time in the
house and took part in the drinking (this is the idea here) or whether from
the first it was dissuaded from the attempt to settle down (Plato, Symp.
212 d, e).
συγγόνων Ἐρινύων is misunderstood by most commentators. The cogna-
tae Furiae of Stanley and Schiitz has had a long life. It was sometimes
taken to mean 'sisters',! as by Passow s.v. σύγγονος and in the translations of
Voss, Conington, Sidgwick, and Verrall, while others rendered the phrase
'kindred Erinyes' (this translation was fairly common from Humboldt to
Headlam). Against this it should be said that there is no conceivable reason
why any prominence should be given here to the relationship of the Erinyes
with each other. Rather, we must expect something very significant, which
sums up important points of the preceding description. The context of the
passage was taken into account by some commentators, among them Nágels-
bach, 'the swarm of Erinyes who revenge the blood of kinsmen', Wecklein,
'because they revenge through kinsmen the murder of kin', Wilamowitz
(on E. Her. 1076) ‘The 'Epwóes σύγγονοι (A. Ag. 1190) revenge the shedding of
kinsmen's blood, whether of ancestors or descendants. I would not deny
theappropriateness of such a thought, but the Greek word does not contain the
meaning which these latter commentators find in it. There is a third inter-
pretation that alone does justice both to the meaning of the word and to the
context: ‘Non cognatae, sorores dicuntur, sed domus, geni? cognatae; nam
praecessit δόμοις, sequitur δώμασιν᾽ (van Heusde); the expression was ex-
plained on the same lines by Enger, L. Campbell, Plüss. It is clear enough in
itself that the description of the avenging Furies as 'belonging in near
relationship to the house of the Atridae' is extremely suitable. Again, it is
most important to notice the correspondence, which cannot possibly be
ı This misunderstanding is not quite so funny as when in Pind. N. 11. 12 the phrase
drpepíay σύγγονον is explained by one scholiast ὡς ἀδελφοῦ αὐτοῦ ὄντος ᾿Ατρεμίου 1j ὡς ἀδελφῆς
(Arpepias supplevi), while another rightly says τὴν συγγενομένην αὐτῶι ἀφοβίαν,

544
 COMMENTARY line 1192
fortuitous, between this passage and 153 (it was pointed out by Schneidewin,
but only in regard to the words σύμφυτος and σύγγονοι ; in fact the resemblance
goes farther than that). There the sacrifice of Iphigenia, the ‘quarrel-making’
disaster, is called σύμφυτον, and this is at once supplemented and explained:
μίμνει γὰρ... οἰκονόμος... Mijvis τεκνόποινος. The strife is ‘rooted, long-
established’, and the Menis (who might also be an Erinys) dwelling in the
house and managing it ‘remains’. In 1190 it is the chorus of ouyyovor ᾿Ερινύες
itself of which it is said ἐν δόμοις μένει. In both passages the same poetic
conception is unmistakable. A variation of the idea that the urge for ven-
geance which springs up out of wicked deeds is in respect of the race σύμ-
φυτον or σύγγονον is found in the words of the Chorus in the Choephoroe (466) :
ὦ πόνος ἐγγενὴς καὶ mapduovoos ἄτης αἱματόεσσα mAayd. To consider, finally,
the linguistic usage, it is precisely σύγγονος which occurs in the sense of
‘grown up with’, ‘inherent’ in Ag. 884, similarly συγγενής 832, Eum. 691, cf.
σύμφυτον Ag. 153.
1191. δώμασιν mpoonpevar: "besieging the chambers (cf. [Rhesus 390 f.] πύργοις
προσῆσθαι) not ‘sitting in the house".' So Verrall; but despite this sensible
explanation he puts forward as ‘possible’ the conjecture πώμασιν, which was
adopted by A. Platt, who translates ‘sitting over their cups’, and by A. Y.
Campbell. Wecklein excogitated the horrible πτώμασιν. Verrall’s correct
interpretation of δώμασιν προσήμεναι is based on good observation: ' δώματα
being the inner rooms’, cf. Wilamowitz, Eur. Her. ii, znd ed., p. 207: ‘smaller
rooms, and particularly bedrooms (called in real life δωμάτια, cubicula, for
which Tragedy, which uses no diminutives, must substitute δώματα). This
does not, however, make it necessary to insist on a contrast between the
δώματα, the inner rooms, and 1189 ἐν δόμοις = ‘in the forecourt (αὐλή) or ‘in
the hall (μέγαρον). Rather does ἐν δόμοις designate the house in general as
the dwelling-place of this throng, while δώμ. προσήμ. with its greater pre-
cision leads us to picture the Erinyes chanting their awful song outside one
(or more) of the rooms and terrifying those within. It would be idle to imagine
any change of place as between 1189 and 1191.
1192. πρώταρχον here alone. ἄτη is, to quote O. Schroeder (Commentary on
Pind. P. 11. 54f.), ‘the word which comprises mental blindness, guilt, and
harm or damage [for the last cf. especially Bucheler, Rhein. Mus. xli, 1886, 8
= Kl. Schr. iii. 89 and also Latte, Archiv fur Religionswissenschaft, xx, 1921,
255 n. 3]. The theme is finely treated by Welcker, Griech. Götterlehre, i. 709 ff.
Here, as often in Homer (not always, cf. Buttmann, Lexilogus, i, zrd ed.,
223 Í.), the predominant notion is mental blindness, infatuation, though the
guilt which arises from the blindness is included. In the first stasimon
(385 f.) we heard that ἄτη, the blindness which leads to guilt, engendered
πειθώ, the persuasion which constrains men to foul deeds; there again dry
is the first cause. In Pers. 97 ff., too, ἄτη means ‘infatuation’, and there
also her action is the first link in the chain of calamity: φιλόφρων yap παρα-
σαίνουσα τὸ πρῶτον παράγει βροτὸν eis dpxvas dra. Mental derangement,
παρακοπά (Ag. 223), is hardly to be distinguished from ἄτη; its epithet is
πρωτοπήμων, a compound of which the second element expresses that which
is implied by the context here, where all that is said explicitly is that ἄτη
makes the first beginning.
ἐν μέρει: ‘allernis vicibus, suo quisque ordine’; so rightly Blomfield (on
4872.3 F 545
line 1192 COMMENTARY

332); 1.6. ‘alternately, one after another, each in her turn’. This is particularly
appropriate to the participants in a κῶμος. We find a somewhat different use
of ev μέρει in cases where A does in his turn what B has done to (or for or
against) him, as, e.g., in Eum. 198, 436. But there is no question of this here,
where the troop of Erinyes remains the subject throughout. We must, then,
think of a successive ἀποπτύσαι (expressed in modern terms and on a lower
level, an ‘ugh!’ uttered by individual Erinyes after they have all in concert
sung of the incestuous adultery) ; but nevertheless we must not try to picture
to ourselves with too great precision exactly how the succession of cries took
place,’ and certainly not attempt to find out what exactly corresponded to it
in the actual performance of a Greek choir. Such attempts are found, e.g., in
van Heusde: ‘ ἀμοιβαίως ἀπεστύγησαν ', and in Wecklein: ‘ ἐν μέρει (alter-
nately) must refer . . . to the alternation of strophe and antistrophe. The
song of the adultery of Thyestes forms, as it were, the antistrophe to the song
of Atreus’ abominable deed.’
ἀπέπτυσαν is here, as has long been observed, not past in sense, but is
rather used like the well-known colloquial ἀπέπτυσα.
1193 gives the explanation of mpwrapxos ἄτη. The initial deed of infatuation
is the adultery with the brother’s wife (so, rightly, e.g. E. Bruhn, Einleitg. zu
Soph. El. p. 4n. 2; G. Méautis, Eschyle, 193 n. 1) and not (as Klausenand many
after him suppose) the slaughter of Thyestes' children. Wilamowitz cham-
pions forcibly the view here rejected (Introduction to his translation of the
Ag., Griech. Trag. ii. 41) : ‘Formerly only the spilling of blood gave rise to the
hereditary curse; to Aeschylus adultery was a scarcely less heinous offence;
he makes the Erinyes of the house mention it along with the other, but for
all that the beginning of the guilt is the first murder, provoked by that
adultery (1192). This interpretation forces us to suppose that the veiled
expression πρώταρχον ἄτην is elucidated not, asis usually the case in Aeschylus,
by what follows immediately after it, but either by utterances in other parts
of this scene (one might feel tempted to think of 1217 ff., but there we have a
fresh start, once more a vision, something that in tone as well as matter is
quite different from the preceding part of the scene) or even by the presumed
general acquaintance of the audience with the events. That is not very likely.
Even if it be assumed that for the Erinyes the murder was of greater moment
than the adultery, it must yet be acknowledged that in the conception of the
poet both form a single complex of guilt and its requital by new guilt, and
that he has here designated quite distinctly the first origin (πρώταρχον) of the
infatuation and guilt. The emphasis laid on the ultimate cause of the long
process of transgression and atonement is significant for the whole of this
part of the scene; cf. on 1223. Heath, Schütz, Blomfield, etc. (so also Welcker,
D. Griech. Trag. i. 359), were wrong in understanding mpórapxos ἄτη as the
treacherous killing of Myrtilus. There is nothing whatsoever in the context
here which points to this; indeed, in the whole of the Oresteia (cf. especially
1 E, Norden, Aus altröm. Priesterbiichern, 185 ff., has pointed to indications on Greek as
well as Italian soil that ancient choirs pass from movements and the corresponding songs
performed 'in a united body' to those in which they move and sing alternis, in turn; he
might have quoted Apoll. Rh. 4. 1196 ff. νύμφαι δ᾽ ἄμμιγα mäcaı . . . ἱμερόενθ' ὑμέναιον
ἀνήπυον" ἄλλοτε δ᾽ αὖτε olödev οἷαι ἄειδον ἑλισσόμεναι περὶ κύκλον. I cannot say how far
Aeschylus had anything like that in mind when he made his σύμφθογγος χορός subsequently
break into successive cries.
546
 COMMENTARY line 1195
Cho. 1068 ff.) Aeschylus nowhere goes farther back than the crimes com-
mitted by Thyestes and Atreus. Cf. on 1468 f.
In the expression of 1193 there is an allusion to Hesiod, Erga 328 (‘Zeus is
wroth with him’) ὅς κε κασιγνήτοιο ἑοῦ ἀνὰ δέμνια βαίνηι. Daube, 115, rightly
remarks that in πατοῦντι we have a sharpening of the Hesiodic expression,
since the verb denotes not treading as such but a wrongful, often a destructive
treading! (cf. on 957). Wilamowitz, in his note on the passage of Hesiod,
remarks appositely : ‘the adultery of Thyestes, and, one might say, that of
Aegisthus also, falls under Hesiod’s law.’ Aeschylus took advantage of the
implication contained in the words of Hesiod.
τῶι πατοῦντι δυσμενεῖς : taken by Stanley as nominative agreeing with the
subject (Ἐρινύες) ;2 so, e.g., Schneidewin, van Heusde, Sidgwick, Verrall.
Schütz and Hermann connect δυσμενεῖς with εὐνάς ; so, e.g., Franz, Conington,
Paley, Nagelsbach, Wecklein, Mazon, MacNeice. Others (e.g. Headlam)
leave the question undecided. I have no doubt that the meaning is that the
bed became the foe of the impious brother who had hoped to make it an
instrument of his lust. From the formal point of view, too, a weightier
structure results if after the verb ἀπέπτυσαν there follows nothing but the
momentous object which with its epithet fills the whole line.
1194. It is obvious that τηρῶ is corrupt. Canter’s θηρῶ still finds adherents;
it stands in all editions of L-S s.v. (so in Passow and in Dindorf’s Thesaurus)
without a warning that it is but a conjecture; the translation given there,
‘have I missed or do I hit the quarry?’, is repeated by Headlam. But the
use of θηρᾶν in the fifth century gives no support for such an interpretation.’
Platt, J. Phil. xxxii, 1913, 65, rightly says: ‘Canter’s θηρῶ is here as absurd
as is τηρῶ itself.’ In contrast, Ahrens's κυρῶ is unimpeachable and probably
what Aeschylus wrote. On the phrase in general cf. 628, for κυρεῖν with the
accusative Ahrens referred to Cho. 707.
On τις in comparisons cf. on 288.
1195. The background is given by 1273 1. θυροκόπος" 6 τοῦ ἐπαιτεῖν ἕνεκα
Paley and Dindorf, Lexicon Aeschyleum, are correct ; the rendering in L-S is not happy:
s.v. πατέω II 2 frequent’ is given for Ag. 1193.
2 Those who wish to advocate this interpretation should not at any rate contend that
there is in the words an allusion to the name Εὐμενίδες, Time and again such jeux d'esprit
are foisted on Aeschylus although it is obvious that in the cases in which he wants to take
advantage of an etymology he manages to render his intention unmistakable. ‘Besides this,
it is exceedingly doubtful whether the Athenians (it was different in the Peloponnesus) were
familiar with the name “Eumenides’ for the Erinyes before the third play of the Oresteia
had produced its effect (cf. K. O. Muller, Aesch. Eum. p. 176, Anhang p. 29; Rapp in
Roscher's Lexikon, i. 1330f.; Wilamowitz, Griech. Trag. ii. 219; Gruppe, Griech. Mythologie,
766). I hope we are right in assuming (cf., e.g., Wilamowitz, loc. cit. ; Verrall, Introd. to Zum.
p. xxxvii; W. Schmid, Gesch. der griech. Lit. i. 2, 1934, 246 n. 5) that it was Aeschylus himself
who gave to the third part of the trilogy the title Edueviôes, Hermann’s guess (Opusc. ii. 133)
that the renaming of the Erinyes was mentioned in the lacuna after Eum. 1027 has met with
the approval of Wilamowitz, Interpr. 229 ; it is, however, quite uncertain. The words of the
ὑπόθεσις, ἡ ᾿Αθηνᾶ ras ᾿Ερινύας πραύνασα προσηγόρευσεν Εὐμενίδας, could perhaps be accounted
for in the way which K. O. Muller, op. cit. p. 28, suggested, 1.6. that a grammarian may
have elaborated Athena’s words at 992 τάσδε γὰρ εὔφρονας.
3 To say nothing of the more far-reaching errors of Passow, the instances (S. Phil. 1007,
Oed, C. 1026, etc.) quoted by L~S I. x under ‘metaph., catch or capture’ bear the ordinary
meaning ‘hunt, seek to catch’ ; a transition to the other meaning might perhaps be admitted
in Xen. Anab. 5. 1.9. Let us therefore stick to the linguistic implication of the nursery-song
(Pollux 9. 123): θηράσεις, ἀλλ᾽ οὐ λήψει.
547
line 1195 COMMENTARY

κόπτων τὰς θύρας (Phrynich. soph. p. 75. ı de Borries, quoted by Blomfield).

Klausen compares Plato Rep. 364 b ἀγύρται καὶ μάντεις ἐπὶ πλουσίων θύρας
LOVTES.

φλέδων, apart from this passage attested only in Timon's Silo, is per-
haps an Ionic word, like φλεδών and the medical term φλεδονώδεα (cf. Erotian
131, p. 90 with Nachmanson’s note). Words in -wv ‘often refer to qualities
that meet with disapproval’ (Buck and Petersen, A Reverse Index, 247).
‘ θυροκόπος φλέδων : both qualifying ψευδόμαντις ' says Plüss; possible but
not certain. It is equallÿ possible that θυροκόπος is attributive to (the sub-
stantival) φλέδων. Finally, it is conceivable that the three terms of abuse are
used substantivally and put all on a level; this seems attractive. Here, as in
1274, the asyndeton would give the effect of listening to a torrent of injurious
appellations.
1196 f. For the understanding of this much-disputed passage it is funda-
mental to recognize that λόγωι goes with εἰδέναι. Headlam denied it. Like
van Heusde, Sidgwick, Verrall, and others before them, he connects λόγωι
with παλαιάς and translates ‘storied’, ‘historic’, citing 5. Oed. R. 1394 f. τὰ
πάτρια λόγωι παλαιὰ δώματα, where, however, it is self-evident that λόγωι
qualifies πάτρια. Headlam’s chief objection (he does not specify his ‘other
objections’) to the connecting of εἰδέναι with λόγωι is that λόγωι would have
to come first. That would only be so if (as in the passages adduced by him,
E. Heraclid. 5, Antiphon 5. 75) the contrary ‘and not from direct experience’
were to be stressed. But here there can be no question of this.” Those who
connect εἰδέναι λόγωι refer the expression to the Chorus, and in this they are
right. There is no suggestion here that the Chorus could know of events far
back in the past in any other way than by hearsay. The meaning, then, is not
‘that you know it by report (and not directly)’, but ‘that you know it, as you
have heard reports of it’. But if εἰδέναι λόγωι belongs together, it follows that
what is in question is the knowledge of the old men of the Chorus. Cassandra
is obviously most desirous to prove that she is a true seer; it is therefore
imperative that she should claim her prophetic gift as the only source of her
knowledge. It would thwart her purpose if she declared that it was from
hearsay that she knew the old stories of the house of the Atridae. This
consideration (quite apart from ἐκμαρτυρεῖν) proves that μ᾽ before εἰδέναι
ı Headlam rebukes the editors of Sophocles for neglecting the parallel of Ag. 1197. This
censure seems odd when we remember Hermann’s wholly appropriate note on Ag. 1197:
* λόγωι, quod non magis hic referri potest ad παλαιάς, quam in Sophoclis Oed. R. v. 1394 . . .
ubi cum πάτρια coniunctum quae patria dicebamini significat. On the position of λόγων
after πάτρια Jebb remarks that this order of words is the less harsh since πάτρια is supple-
mented by παλαιά, It would seem more relevant that Oedipus is not saying here: Ὃ
Polybus and Corinth and house of my father’ (that is how he would have spoken before the
revelation of the truth), but is saying *. . . and house of my father (as I called thee)’. If he
had placed λόγωι first he would have laid stress on the implicit antithesis καὶ οὐκ ἔργωι,
which is alien to the emotion of this passage. In Hdt. 5. 20. 5 παρίζει Πέρσηι ἀνδρὶ ἄνδρα
Μακεδόνα ὡς γυναῖκα τῶι λόγωι the order of words is accounted for by the antithetical
arrangement,
2 Even less conceivable is an interpretation such as that propounded by Kamerbeek,
Mnemos. Ser. III, vol. viii (1939/40), 49 ff. ; on p. 52 he produces the following ‘translation’
of the passage : "Testare (= affirma) sodes hoc (hic et nunc) iuratus me fando audisse (nec
divinatione nosse) vetera facinora huius domus’; what ought to be the main point is here
supplied in brackets. Moreover, if such a contrast were intended, the word order would be
suspicious (see above).
548
 COMMENTARY lines 1196 f.

makes nonsense, and that with Bothe and Dobree! we must read μή (on the
crasis see below), so that the person challenged to do the ékuaprupetr becomes
the subject of μὴ εἰδέναι. This is what one would expect in this connexion.
It cannot often happen (outside the peculiar world of commentaries on the
classics) that a person is adjured to testify not to his own knowledge (or lack
of knowledge) but to that of the person who desires his testimony. Further,
it should be self-evident that Cassandra’s challenge is incomparably more
forcible if she says ‘bear witness under oath that you have never heard of the
old crimes of this house’ than if she said ‘Come, with an oath attest that I do
know the long and sinful history of this house’ (Headlam). The reader who
distrusts considerations of this kind may perhaps be impressed by a palpable
detail of grammar. Wecklein, Stud. z. Aesch. 18 ff., collected the infinitives
with τὸ μή which occur in Aeschylus, and Birklein, ‘Entwicklungsgeschichte
des substantivierten Infinitivs’, Schanz’s Beitr. z. hist. Synt. d. gr. Spr. iii.
1, 1888, 17 ff. followed this up by demonstrating that in Aeschylus the accusa-
tive of the infinitive with the article occurs without a negative only seven
times? and in the simplest form, i.e. with the unexpanded infinitive as
object of a verb, as, e.g., Ag. 498 τὸ χαίρειν μᾶλλον ἐκβάξει λέγων, 548 τὸ σιγᾶν
φάρμακον βλάβης ἔχω, similarly 788, 1289, Eum. 652. On the other hand,
negatived articular infinitives in the accusative are far more frequent (not
counting Ag. 1196, there are thirteen cases with τὸ μή, four with τὸ μὴ οὐ);
furthermore, they are often expanded so as to form sentence-like clauses, as
in, e.g., Ag. 1589 τὸ μὴ θανὼν πατρῶιον αἱμάξαι πέδον αὐτός, Cho. 302 ff. τὸ μὴ
πολίτας εὐκλεεστάτους βροτῶν . . . δυοῖν γυναικοῖν ὧδ᾽ ὑπηκόους πέλειν, Prom.
236 τὸ μὴ διαρραισθέντας eis Ardov μολεῖν and so on. Thus τὸ μ᾽ εἰδέναι λόγωι
. . . δόμων would be absolutely unique in the whole of Aeschylus, while there
is a large set of parallels to τὸ μὴ εἰδέναι κτλ.
Weil (already in 1859, see also the ‘Addenda’ to his first edition of the Eum.,
p. 136) and Ahrens (626 f.) have understood the sentence aright. The main
thing is the recognition that Cassandra so frames her challenge to the old
men to bear her witness that if they accepted it they would have to commit
an act of perjury. The coryphaeus immediately realizes the implication
(1198 f.) and evades the challenge. It is a less important question whether
ἐκμαρτυρεῖν means here simply ‘bear witness openly’ or whether it has some
more specific sense. Ahrens examined the instances of the verb and drew
attention to the technical term of legal procedure, viz. ἐκμαρτυρία, cf. Thal-
heim, RE v. 2213 f.; Lipsius, D. att. Recht u. Rechisverfahren, 886 ff. (the
relevant passages from the orators are given in extenso); R. J. Bonner and
G. Smith, The Administration of Justice etc., ii. 132 ff. If Pollux 8. 36 μαρτυρία
δὲ καλεῖται, ὅταν τις αὐτὸς ἰδὼν μαρτυρῆι, expaprupia δέ, ὅταν τις παρὰ τοῦ
ἰδόντος ἀκούσας λέγηι or Etym. M. 324. 3 = Suid. s.v. (vol. ii, p. 222. 27 Adler)
ἐκμαρτυρεῖν φασι τὸ λέγειν, οὐχ ἅπερ αὐτὸς εἶδεν, GAN ἅπερ ἑτέρων ἤκουσε
λεγόντων is set alongside the expression ἐκμαρτύρησον . . . μὴ εἰδέναι λόγωι,
it seems natural to find in the passage of Aeschylus an allusion to the legal
term. It may, however, be a fairly free allusion and not a technically correct
ı Adversaria, ii. 25. The two apparently arrived at the truth independently.
2 Here I modify Birklein’s statements after re-examining his evidence myself. Of the
articular infinitives without a negative Birklein notices only Eum. 652; of the negatived
ones Eum. 749 has fallen out both in Wecklein and Birklein. [See the Addenda.]
549
lines 1196 f. COMMENTARY

reproduction (for such a non-technical adoption cf., e.g., what has been said on
534 f. on ἁρπαγῆς τε καὶ κλοπῆς δίκην and τοῦ ῥυσίου). The éxuaprupia of Attic
legal procedure is described by Lipsius, loc. cit., and by Bonner and Smith,
loc. cit., as follows: ‘The eviuence of persons who were too ill to attend court,
or who were absent from the city at the time of the hearings, could be taken
in writing in the presence of a number of persons who afterwards on the
production of the document in court identified it by an attesting affidavit
as the statement of the original witness.’ Thus the ἐκμαρτυρῶν bore witness
before the court that an absent person had borne witness in his presence to a
certain state of affairs. But Cassandra is requiring the coryphaeus to bear
witness under oath that he has never received from anyone any information
on the events in question. Indubitably this negative content of the declara-
tion demanded makes its difference from the éexpaprupia of legal procedure
considerable. Nevertheless the striking phrase εἰδέναι λόγωι inclines me to the
belief that the association with ἐκμαρτυρία is intentional and that it would be
readily understood by an Attic audience, who ἐπὶ τῶν δικῶν ἄιδουσι πάντα
τὸν βίον. On the other hand, it is not impossible that ἐκμαρτυρεῖν is used here
in the same loose and non-technical manner as in Zum. 461 (‘simply an
intensification of μαρτυρεῖν᾽, Blass, ad loc.), Eur. fr. 553 N., Isaeus 3. 77,
Aeschines 1. 107 (cf. Lipsius op. cit. 887 n. 85). This is the view of, e.g., Head-
lam, K. Latte, Heiliges Recht, 36 n. 23, Daube, 2 f. Daube says that Ahrens’s
interpretation does not give its proper force to προυμόσας. I cannot under-
stand this, since the swearing in of witnesses naturally plays a large part in
many Greek legal systems, among them the Attic (there is no need to go into
details here; a survey is given by Latte, Heil. Recht, 33 ff., and in his article
‘Martyria’ RE xiv. 2035 ff). P. Von der Mühll (quoted by Daube, loc. cit.)
assumes that Aeschylus has expressed the idea of ἐξωμοσία in two separate
words. If we accepted this assumption, the words μὴ εἰδέναι would be in
keeping with the statement made by the witness who is called upon to
produce his evidence by means of ἐξωμοσία, for, as we learn from Pollux
(8. 55), ἐξώμνυντο δὲ καὶ of κληθέντες μάρτυρες, εἰ φάσκοιεν μηδὲν ἐπίστασθαι
τούτων ἐφ᾽ ἃ ἐκαλοῦντο ; a good example of this is provided by Demosth. 57. 59
καὶ ταῦτ᾽ οὐκ ἂν ἐξομόσαιτ᾽ Εὐβουλίδης οὐδ᾽ οἱ μετ᾽ αὐτοῦ μὴ οὐκ εἰδέναι (for
further details cf. Thalheim ᾿Εξωμοσία, RE vi. 1689; Lipsius, op. cit. 878 f.;
Bonner and Smith, op. cit. ii. 163 f). S. Ani. 535 (compared by Plüss with
Ag. 1196) may indeed, as Daube (with Jebb) supposes, contain an allusion to
the technical use of ἐξωμοσία : ἢ ᾽ξομῆι τὸ μὴ εἰδέναι. But in Ag. 1196 we do not
find ἐξόμνυσθαι, i.e. the word which would suggest that allusion, and Von
der Muhll’s ‘separation into two words’ seems an unwarranted hypothesis.
Moreover, εἰδέναι λόγωι is appropriate so long as there is a playful reminiscence
of ékuaprupia, but becomes pointless if we are supposed to think of ἐξωμοσία.
1196. μὴ εἰδέναι. On the use of μή (as early as Homer) to negative statements
of fact governed by verbs of swearing and others, cf. Goodwin, § 685, Wacker-
nagel, Syntax, ii. 282. The crasis μὴ εἰδέναι is unexceptionable. In Sophocles,
as Ellendt-Genthe, Lex. Soph. 448, establish, crasis of μή and εἰ never occurs
except with forms of εἰδέναι, but several times with them, Ant. 33, 263, 535,
Trach. 321, Oed. C. 1155. From Euripides Denniston on El. 961 (μὴ eioiôm)
cites Hipp. 1335 μὴ εἰδέναι, [on 313 μὴ εἰδότα, Or. 478 μὴ εἰδέναι. There is no
doubt about μή in any of these passages. That the Laurentianus reads μ᾽
550
 COMMENTARY line 1198

εἰδέναι in Ant. 535 and μ᾽ εἰδότ᾽ in Oed. C. 1155 shows how easy in this combina-
tion is the omission of the ἡ standing in crasis which we assume in Ag. 1196.
When we turn to the structure of the speech, we notice that ἐκμαρτύρησον
κτλ. is a slightly varied form of μαρτυρεῖτε in 1184. Aeschylus likes to produce
a kind of frame by rounding off a speech or a section of a speech by repeating
or recalling in a varied form something said at its beginning so that the
coherence of the included piece is brought out by the parallelism of the
opening and the concluding words.' Cf. 20, 829, 1183, 1611; Suppl. 417; Eum.
20, 453 (harking back to 444) and elsewhere.
1198. καὶ πῶς ἂν κτλ. Hermann wrongly saw in these words an instance of
the wishful questions with πῶς ἄν which were discussed on 622 ; he translated
‘atque utinam iusiurandum . . . medelam afferre possit)’.2 Against this (so
Paley, rightly) is the resultant diminution in the force of the reply ; and, in
point of detail, the presence of καί, which 'so often in dialogue introduces
questions that state a fact or draw a conclusion fatal to the opponent'
(Wilamowitz on E. Her. 509), its function, as generally in such questions, being
to ‘convey an emotional effect’ (Denniston, Particles, 309 f.; the present case
belongs to the type there characterized by the paraphrase ‘And what is the
good of that ?’). Denniston 310 also gives examples to show that in questions
introduced by καί ‘there is often an echo of a word from the previous speech’ ;
so here ὅρκος picks up προυμόσας (1196). The sentence beginning with πῶς ἄν
here conveys not a wish (however hopeless) but, as a ‘rhetorical question’,
a statement of impossibility like A. Suppl. 226 ff. ὄρνιθος ὄρνις πῶς ἂν ἁγνεύοι
φαγών; πῶς δ᾽ ἂν γαμῶν ἄκουσαν ἄκοντος πάρα ἁγνὸς γένουτ᾽ av; or, with καί,
as here, in rejoinder, E. Alc. 142 καὶ πῶς ἂν αὑτὸς κατθάνοι τε καὶ βλέποι;
(Paley), Ar. Peace 1076 Ὁ καὶ πῶς, ὦ κατάρατε, λύκος ποτ᾽ ἂν οἷν ὑμεναιοῖ;
ὅρκου πῆγμα γενναίως παγέν. Occasional attempts to keep πῆμα only go
to show the impossibility of defending it, e.g. the attempt of Weil, who (in the
Addenda to his Eumenides, 1861, p. 136) paraphrases πῆμα γενναίως παγέν in
almost exactly the same way as Pauw had once done: ‘mala penitus haerentia,
praeterita et immutabilia', or that of Verrall ('framed naturally to be a
hurt', intended by him to apply to the oath) ; utterly fantastic is the recent
attempt of Kamerbeek, loc. cit. (p. 548, n. 2). Auratus' ὅρκου πῆγμα has
rightly been generally adopted. For the expression cf. E. I$h. A. 395 τοὺς
κακῶς παγέντας ὅρκους (Markland, ad loc., connected this with Ag. 1198, and
C. G. Haupt on Ag. 1198, Paley, and others followed suit). It is natural enough
that the taking of the oath should be expressed by πηγ-; cf. factum, pepigi
of compacts (Catullus 62. 28 quae pepigere viri, pepigerunt ante parentes and
elsewhere), etc. If 1199 παιώνιος were the reading of the MSS, we might
attempt to keep ὅρκος despite a certain harshness (with πῆγμα y. v. in apposi-
tion) ; but, as it is, ὅρκου is necessary. Bothe objects 'dubites, qui factum sit,
ut tam plana oratio corrumpatur' ; we can share his wonder without taking
refuge with him in ὅρκους and an impossible construction.
γενναίως παγέν. Hermann: 'indicat, quamvis sanctissimum iusiurandum

1 Aeschylus does not, of course, stand alone in employing this simple device. ‘It is
Theocritus’ not infrequent habit to roundoff a paragraph, a song, a poem, and to indicate
that it is coming to an end, by an echo of its opening! (A. F. S. Gow, C.R. lvi, 1942, 111).
2 So Humboldt, perhaps following Hermann's advice: ‘O! konnte Schwur, ein fester,
fromm geknupfter Bund, Heilmittel werden!’
331
line 1198 COMMENTARY

tamen nihil profuturum esse’, similarly, e.g., Paley: ‘ “however solemnly

ratified” or ‘however generously the pledge is given" ', Wecklein: ‘mag der
Eid auch mit noch so aufrichtigem und wahrhaftem Sinne geschworen sein',
Headlam: 'an oath's plight, though genuinely plighted'. I doubt whether one
can get from the Greek such a quamvis . . . tamen, ‘however’, ‘though’, ‘auch
noch so'. These commentators wish, as it seems, to find in the passage the
description of an honourable oath; so also Wilamowitz: 'Und wenn ich
ehrlich leisten kónnte diesen Schwur, was würd' es helfen?' But that is not
the sense of γενναίως. Stanley's translation, though somewhat colourless, is
more correct: iuramenti compages valide pacta'. The best aid to the under-
standing of yevvatos is what Passow (after him L-S) put at the beginning of
his article, the definition of Aristotle, Hist. an. 488 b 18 εὐγενὲς μὲν γάρ ἐστι τὸ
ἐξ ἀγαθοῦ γένους, γενναῖον δὲ τὸ μὴ ἐξιστάμενον ἐκ τῆς αὑτοῦ φύσεως. The word,
then, means that which remains within its nature and type, the genuine.
From the notion of that which preserves its nature without falsification is
easily developed the notion of ‘stable, dependable, strong’. It is only these
shades of meaning which we have to consider here, omitting the more familiar
uses (for these see Wilamowitz on E. Her. 872). It is somewhat superficial,
but intelligible, that in the scholion on 5. Aj. 938 (χωρεῖ πρὸς ἧπαρ, οἶδα,
γενναία δύη) γενναία is glossed by ἡ ἰσχυρά; what is actually meant is a sorrow
which has kept to the full all the characteristic signs of sorrow. When Boreas
says (Aesch. fr. 281. 5 N.) νῦν δ᾽ où κέκραγά πω τὸ γενναῖον μέλος, the meaning is
that the true howl of Boreas has not yet begun, the northerly gale has not yet
reached its proper force. In E. Jon 935 the old servant says: ws ovorevaleıv
γ᾽ οἶδα γενναίως φίλοις, i.e. ‘I am a slave and feeble with old age (cf. 738 ff.),
no longer any good at any real service or help, but, as far as concerns showing
my sympathy for those who are near to my heart, in that at least (ye) I can
still play my full part.’ Plato, Theaet. 209 € τὸ γὰρ ἃ ἔχομεν ταῦτα προσλαβεῖν
κελεύειν, iva μάθωμεν ἃ δοξάζομεν, πάνυ γενναίως ἔοικεν ἐσκοτωμένωι. SO πῆγμα
γενναίως παγέν is the confirmation which has been given in a manner true to
type, with the utmost efficiency in establishing the object. Consequently,
there is here no special reference to a true oath, as opposed to perjury, but the
corroboration by oath is designated quite generally as a strong confirmation
of a statement. Thus in sense ὅρκου πῆγμα γενναίως παγέν is close to the
Homeric καρτερὸς ὅρκος. That provides a good foil to the coryphaeus’
assertion that in the present situation an oath is useless.
The coryphaeus evades the issue. Cassandra wished by her challenge to
force the old men to acknowledge her as a true prophetess. But the cory-
phaeus is arguing upon a completely different plane ; the mighty confirmation
of an oath could not undo any of the harm. Thus he avoids committing
perjury; at the same time his words are in accord with his usual attitude of
resignation. The declining of the oath is in itself an acknowledgement of the
truth of Cassandra’s revelations, though the first explicit acknowledgement
is contained in the following words, which naively admit that everything
happened exactly as she has said.
1199. The older editions keep cov and put a comma after it; so also Her-
mann, Paley, Wilamowitz, and others. Many editors follow Auratus in
reading oe. Schneidewin, who retains σου, rightly remarks: ‘the infinitive,
θαυμάζω σου τὸ λέγειν κυρίως, . . . takes the place of θαυμάζω τοῦτό σου ὅτι
552
 COMMENTARY lines 1202 ff.

λέγεις" ; similarly Verrall. Cf., e.g., Isocr. 8. 12 θαυμάζω δὲ rar... πρεσβυτέρων,

εἰ μηκέτι μνημονεύουσιν κτλ.
1200. πόλιν is probably corrupt. If it would have to mean ‘trans mare edu-
catam in urbe aliena lingua utente’ (Stanley), that is hardly Greek. Kuhner-
Gerth, i. 313 f. n..13 cite passages from poetry where the accusative with verbs
of rest denotes the place occupied, but this occurs only with such verbs as lie,
sit, stand, i.e. with those which express the result of a movement. Accord-
ingly Abresch tried to alter the text, while Schtitz assumed a lacuna between
ἀλλόθρουν and πόλιν. Hermann translates ‘te transmarinam de peregrina urbe
sic loqui ut si praesens fuisses’ ; so also L. Campbell, Verrall, Headlam, Platt,
G. Thomson. That is not even linguistically correct, since πόλιν λέγω, without
the addition of an infinitive, means not de urbe loquor, ‘I say something about
the city’, but! ‘I am speaking about the city, I mean the city (and not some-
thing else)’, and there can be no question of an analogy with κακῶς λέγειν
(so Hartung, rightly). But in regard to the thought, too, Hermann’s inter-
pretation is inappropriate, since, as Hartung, Schoemann, and others have
seen, after all that has been said in this scene on the non-Greek, barbarian
home and speech of Cassandra, ἀλλόθρουν must refer to her or to the Trojans
(this objection has by no means been deprived of its force by Meineke,
Philol. xix, 1863, 208, with his reference to S. Phil. 540). Enger conjectured
ἀλλόθρωι ᾽ν πόλει (adopted by Wecklein). But it should be remembered that
there is scarcely an instance in tragedy of the aphaeresis of ἐν except after a
monosyllable ending in n; cf. on 431. Proceeding from the probably correct
assumption that ἀλλόθρουν (referring to Cassandra) needs no addition,
Schoemann, Opusc. iii. 150, wrote πάνυ (for πόλιν) ; the adverb would go with
κυρεῖν λέγουσαν. Weil’s conjecture τὸ πᾶν, which comes to the same thing, is
perhaps preferable on stylistic grounds. A certain emendation is here im-
possible to attain.
An ingenious attempt to retain the text as it stands was made by Ubaldi:?
"Poiché πέραν & originariamente un accusativo come χάριν, cosi io intendo
ἀλλόθρουν πόλιν come apposizione di πέραν senza riferirlo a κυρεῖν λέγουσαν.
Dunque: io mi meraviglio di te, che allevata al di là del mare, in una città
d'altra lingua (ἀλλόθρουν), parli rettamente come se avessi assistito (a questi
fatti>.’ I wish this were the truth, but I dare not believe it, for if πέραν is to
be taken as an accusative as in 19o, it cannot be governed by τραφεῖσαν, and
if, on the other hand, it serves as an adverb, ἀλλόθρ. πόλιν can hardly be in
apposition to it. Or am I making the mistake of pressing archaic language
into the categories of over-rigorous grammar?
1201. κυρεῖν (cf. on 1194) λέγουσαν : ‘that you hit the mark with what you
say’; cf. on 622.
1202 ff. Hitherto Cassandra, absorbed in her visions, has given no answer
to the questions (1150-5) πόθεν... ἔχεις. Now she reveals the source of her
prophetic powers. It is evident, as throughout the non-lyrical part of the
scene, that a sharp stressing of causal relations has taken the place of a mere
sequence of visionary pictures and passionate outcries.
1 An object such as πόλιν is, of course, different from nouns denoting an action or event,
as, e.g., Cho. 444 λέγεις πατρῶιον μόρον : ‘our father's end was as you say’ (Headlam ; for the
context, which has often been misunderstood, cf. below on 1600).
2 Beazley arrived at the same conclusion. [For Übaldi see p. 571 n. 3.]

593
line 1202 COMMENTARY

1202. τέλος doubtless means here munus, ‘Obliegenheit’ (cf. on 105), but it is
possible that it also contains the notion of a ritual to be performed, cf. on
934 (p. 422) and Aesch. fr. 387 N. τοῦδε μυστικοῦ τέλους.
The stichomythia reveals the relation between Cassandra and Apollo.
Modern research (cf. W. R. Halliday, Greek Divination, 82, and especially
Latte, RE xviii. 840 and Harvard Theol. Rev. xxxiii, 1940, 15 f.) has made it
probable that the particular nature of that relation (for similar phenomena
cf. E. Norden, Vergilius Aeneis VI, and ed., p. 146) is a mythical reflection of
such rituals as we know, e.g., from Patara, where Apollo’s πρόμαντις was
regarded as the παλλακή of the god. After a somewhat vague hint by Bouché-
Leclercq, Histoire de la divination dans l'antiquité, ii. 148, it was suggested by
Latte that the author of the Cypria developed out of an Anatolian rite the
story of the seeress who promised to yield to the desire of Apollo and received
from him the gift of prophecy. From a sentence in Proclus’ summary (Homer:
opera, ed. Allen, v. 103. 1; Bethe, Homer, ii. 2, 2nd ed., 153) καὶ Κασσάνδρα
περὶ τῶν μελλόντων προδηλοῖ, it appears that in the Cypria Cassandra was
represented as a prophetess.! This testimony is generally trusted, cf., e.g.,
Wilamowitz, Griech. Tragoedien, ii. 271 n. 1; C. Robert, Heldensage, 997.
The scepticism with which Bethe, RE x. 2291 and Homer, ii. 2, 2nd ed.,
231 f., treats the Proclus passage is hardly justified; against his argument
(Cassandra’s prophecy a doublet of that of Helenos) see the remark of
Geffcken, Hermes, lxii, 1927, 5.
1203 f. Hermann’s transposition is almost universally recognized as neces-
sary ; instead of it we find in Verrall and Pliiss astounding feats of ingenuity
in favour of the pre-Hermannic fextus receptus. Anyone who feels tempted to
adopt Verrall’s text should not imagine that in so doing he is following the
‘MS tradition’; for in the MSS the indications of the speakers unmistakably
mark the whole section from 1202 onwards as stichomythia. Weil is more
cautious: he assumes the loss of single lines after 1202 and 1204. That is not
impossible but in the highest degree improbable. Hermann’s transposition
yields an uninterrupted progress of thought and a close connexion of question
and answer. Dobree (Advers. ii. 25) also saw that 1205 must follow 1203,
but put 1204 after 1205.
1205. ἁβρύνεται: often misinterpreted; so Stanley, ‘luxuriatur’, Schütz,
‘promptus est ad lasciviam', Dindorf, Lex. Aesch., 'delicias facto [with wrong
use of the Latin phrase, see T'es. 1. L. v. 447. 58 ff.], superbio, insolesco’,
others similarly. Right interpretations are given by Conington: 'folks are
delicate’, and Paley: ‘to be nice or particular’. Paley well compares E. 755. A.
858, 1343, Headlam Herodas 6. 45. Of the three senses given by the Berlin
Photius (p. 7. 18 Reitzenstein) for dßpüvera:, the second, θρύπτεται, is exactly
the shade of meaning which we have here.
For the use of πᾶς τις (cf. 1651) see Dindorf, Thesaurus, vi. 573. Cf. Wila-
mowitz, Arist. u. Ath, ii. 314 n. 12.
πλέον, placed as it is beside εὖ πράσσων, suggests a supplement such as ἢ
κακῶς πράσσων (the fuller form of expression in, e.g., Soph. fr. 186 N. — 188 P.
φιλεῖ γὰρ ἡ δύσκλεια τοῖς φθονουμένοις νικᾶν ἐπ᾽ αἰσχροῖς 7) "mi rois καλοῖς πλέον).
1206. ἀλλά: ‘Well then’ (Paley). Cf. Denniston, Particles, 18 (‘A person
asked to speak conveys his readiness to speak by speaking’).
Cf. Bacchyl. fr. 8a Snell and Wilamowitz’s note on it (Bacchyl. ed. Snell p. 46*).

554
 COMMENTARY line 1207
Modern prudishness has been busy weakening the force of this magnificent
line. Bothe: ‘ παλαιστής figurate dicitur is, qui pugnat ut impetret aliquid’,
Schneidewin : ‘Apollon habe ihrem Besitz nachgerungen, wie ein Ringer dem
Preise’, Paley: ‘he was a lover who inspired me with great affection’, L.
Campbell: ‘A wrestler for my heart’, Wilamowitz : 'Inbrünstig Liebeswerben
rang um meine Gunst’, Headlam: ‘he contended for me strenuously’. The.
use of παλαίειν and παλαιστής makes it beyond doubt that Apollo did not in a
metaphorical sense contend for her heart or her favour, but actually wrestled
with her. The god sets himself to overpower the maiden, who feels and acts
like a true maiden. Then she agrees, and that brings the physical wrestling
to an end; later she breaks her promise. But from the beginning it is not
merely brute force which is here at work; with all her resisting Cassandra is
susceptible to the power of the god’s χάρις. For all that she withdraws before
the consummation. How that could be the poet does not reveal; we must
rest content with knowing that all the misfortunes of the seeress sprang from
that breaking of her promise.
χάρις expresses here too (cf. on 182) a complex notion. If we had a plain
χάριν πνέων, like the Homeric μένεα πνείοντες, etc., it would merely indicate
that χάρις is an emanation of Apollo’s nature. This idea is indeed inherent
in the passage. But the addition of the dative ἐμοί implies that the divine
χάρις was directed towards, and at the same time worked upon, Cassandra ;
the god conferred his favour on her while she underwent the influence of that
irresistible radiance, that χάρις, which is particularly strong in Phoebus
Apollo: οὐδὲ yap θεοὶ σεμνᾶν Χαρίτων ἄτερ κοιρανέοντι χοροὺς οὔτε δαῖτας" ἀλλὰ
πάντων ταμίαι ἔργων ἐν οὐρανῶι, χρυσότοξον θέμεναι πάρα Πύθιον ᾿ἀπόλλωνα
θρόνους, αἰέναον σέβοντι πατρὸς ᾿᾽Ολυμπίοιο τιμάν (Pind. Ol. x4. 8 ff.).
1207. τέκνων εἰς ἔργον ἤλθετον: Blomfield aptly refers to Semonides of
Amorgos fr. 7. 48 f. D. ὁμῶς δὲ καὶ πρὸς ἔργον ἀφροδίσιον ἐλθόντ᾽ ἑταῖρον ὁντινῶν
ἐδέξατο. The plural is more usual, from the ἔργα γάμοιο, φιλοτήσια ἔργα of
epic on. As in Swppl. 1037 f. τίεται δ᾽ αἰολόμητις θεὸς (Aphrodite) ἔργοις ἐπὶ
σεμνοῖς, so here Aeschylus gives the proper force to ἔργον by characterizing it
with the aid of τέκνων, as well as by the addition of νόμωι. This expression
(cf. on 594), ‘as is customary’, raises the matter to the rank of institutions
hallowed by ancient custom and common consent! (even recent editors have
persisted in substituting for νόμωι the empty ὁμοῦ, which was rejected by
Hermann). The combination τέκνων eis ἔργον would recall to every Athenian
hearer the solemn marriage-formula (Menander, Perik. 435 f., cf. fr. 720 Kock
and the footnote below) γνησίων παίδων Em’ ἀρότωι (or ἐπὶ σπορᾶι), liberum
quaesendum causa, as Ennius (Trag. 97 Ribb., cf. 120) translates Euripides.
Aeschylus also alludes to it in the speech of his Europa (fr. 99. 5 f. N.; on the
text cf. Wilamowitz, Interpr. 235) : γυνὴ θεῶι μειχθεῖσα παρθένου σέβας ἤμειψα,
παίδων δ᾽ ἐζύγην ξυνωνίαι.
ἤλθετον; the normalizing doctrine of Elmsley (on Ar. Ach. 733 and
E. Med. 1041 [= 1073] was accepted by Hermann and many others.

1 When Clemens Alex. Strom. 2. 23. 1 (p. 188 Stählin) cites the marriage- -formula quoted
below he begins γάμος μὲν οὖν ἐστι σύνοδος ἀνδρὸς καὶ γυναικὸς ἡ πρώτη κατὰ νόμον (the point
relevant here) ἐπὶ γνησίων τέκνων omopäı (this variant of the formula is found in Menander
in the so-called fabula incerta, p. 88 Körte, 3rd ed., 29 f.; cf. Körte, Hermes, lxxii, 1937,
63 n. 3, 64 n. 4).
555
line 1207 COMMENTARY

Relying on the observation that in Attic drama metre often demands the
«τὴν forms for the second person of the dual of historic tenses (e.g. 5. Oed. R.
1511), Elmsley wished to expel the -rov forms wherever they occur. Protests
were entered by Conington on this passage and by others; there is a good
discussion in Kühner-Blass, ii. 69 f.; see in addition the treatment of the
problem by O. Lautensach, Gramm. Studien zu den griech. Tragikern u.
Komikern, i (Progr. Gotha 1896), 20 f. (further literature in Schwyzer, Griech.
Gramm. i. 667 n. 2). It must presumably be admitted that the Attic poets
used both forms.'
1209. The meaning of ἡιρημένη is not transparent. Stanley's ‘imbuta’ (so
Linwood, 'inspired with prophecy', and others) does not find any support
in the use of αἱρεῖν. Plüss's rendering, ‘chosen by the god [so Wellauer and
L-S αἱρέω CII] for the purpose of prophesying’, must also be rejected, not only
because with a passive not provided with any of the usual supplements the
dative cannot naturally be taken as anything but instrumental, but also
because the thought of Cassandra's being chosen is of small importance here
where the theme is the extent and degree of her subjection to the god.
Equally inappropriate is the comparison with phrases like ἄτη φρένας εἷλε,
δέος αἱρεῖ and the like, since the gift of prophecy is not comparable with
such emotions (or with émvos, etc.). Here, as so often, the verb bears the
meaning 'take captive, conquer, overpower', etc., but the place of the con-
quering god is taken by the τέχναι called ἔνθεοι because they contain the god
in themselves; their prize is the prophetess. It needs no saying that the
combination τέχναισιν ἐνθέοις ἡιρημένη, besides what is explicitly expressed,
implies the idea that the being so seized becomes herself ἔνθεος.
1211. ävaros is a certain restoration. With this word too (cf. on 238 dvav8os)
it is inappropriate to speak of an active or passive force ; ἄνατον is that which
is devoid of hurt, here the human being who comes off without damnum (on
this sense of ἄτη cf. on 1192). ‘How could you remain unharmed by the wrath
of Loxias?' So framed, the question comes to much the same as saying:
‘somehow or other Apollo must surely have punished you’. Cf. the question
framed in exactly the same way (also in stichomythia) Cho. 532 kai πῶς
ἄτρωτον οὖθαρ ἣν ὑπὸ στύγους; ‘somehow or other her breast must surely
have been hurt’. In both these passages πῶς in combination with the
privative ἀ- of the verbal adjective has an effect similar to that of πῶς οὐ in
cases where the sentence so introduced is really affirmative in force (cf.
Kühner-Gerth, ii. 522 n. 9, and more especially Wecklein on A. Prom. 41).
The coryphaeus has no doubt of the inevitability of punishment ; it is strange
enough in his eyes that Cassandra is still alive after she Aogiay ἐψεύσατο. It
might have been expected that Apollo would kill her: this is how he treated
Coronis, who deceived him, ἀποφλαυρίξαισά νιν ἀμπλακίαισι φρενῶν (Pind. P.
3. 12), just as Cassandra ἤμπλακε (1212, cf. Wilamowitz, Glaube d. Hell. ii. 121,
who compares the two passages). It is only in what follows that the old men
learn the peculiar and peculiarly cruel nature of Cassandra's punishment.
ı Wilamowitz's procedure is quite inconsistent. In Ag. 1207 he reads ἠλθέτην, with
Elmsley, but in S. Oed. C. 1746 (Verskunst 526) he keeps éAdyerov and in Ar. Thesm. 1158
(Verskunst 592) ἤλθετον, In A. Cho. 510 the form ἐτ(ελίνατον is preserved in M, except for the
wrong division of words (Murray's reference to Cho. 76o is inappropriate, since εἰχέτην is
there third person).

556
 COMMENTARY line 1216

1212. ἔπειθον οὐδέν᾽ οὐδέν. Those, at any rate, of Aeschylus’ contemporaries

who possessed insight deeper than the average were familiar with this as the
fate of persons with foreknowledge of the future, and realized what a depth
of misery it involved. The Persian in Herodotus 9. τό. 4 f. who predicts the
impending catastrophe says: ὅ τι δεῖ γενέσθαι ἐκ τοῦ θεοῦ, ἀμήχανον ἀποτρέψαι
ἀνθρώπωι" οὐδὲ γὰρ πιστὰ λέγουσι ἐθέλει πείθεσθαι ovdeis . . . ἐχθίστη δὲ ὀδύνη
τῶν ἐν ἀνθρώποισι αὕτη πολλὰ φρονέοντα μηδενὸς κρατέειν.ἷ
1213. γε μὲν δὴ stresses the contrast, cf. on 887. πιστά takes up ἔπειθον ; cf.
the passage of Herodotus just quoted.
1214. Abruptly the dialogue breaks off. The cries of Cassandra, the lyric
character of the shorter line (cf. Prom. 742), and the violent movements which
certainly accompany the cries of pain all of-a sudden transport the audience
from the quiet realm of διαλέγεσθαι and of questioning, instructing, inferring,
into a wild medley of strange visions. Similar, though not with so overpower-
ing a contrast, is Prom. 877 ff., which in expression also resembles Ag. 1215 f.
One would like to think that the musical accompaniment of Ag. 1214 con-
tained short but unmistakable echoes from the preceding songs of Cassandra
(or their instrumental accompaniment). Everything from here to the end of
1222 is vision; then, with the sharply causal ἐκ τῶνδε, begins once more the
rational procedure of genuine λόγος,
1215 f. ὑπ᾽ αὖ pe... orpoBet with Prom. 878 f. ὑπό μ᾽ ad... θάλπουσι shows
that of the words which tend to take the second place in the sentence some-
times one, sometimes the other must give place (cf. on the one hand, e.g., 5.
Aj. 787 f. τί μ᾽ αὖ τάλαιναν... ἀνίστατε; Trach. 1031 τόδε μ᾽ αὖ λωβᾶται, on the
other hand Anacr. fr. 49 D. οὐδ᾽ αὖ μ᾽ ἐάσεις... ἀπελθεῖν; Ar. Knights 336
οὐκ αὖ μ᾽ ἐάσεις; Sometimes it seems possible to detect the reason for the
order adopted; so in Prom. 567 χρίει τις αὖ με τὰν τάλαιναν οἶστρος, where με
τὰν τάλ. must be placed together). Here and in Prom. 878 f. the ‘tmesis’ is of
the same kind; it is the very old, presumably inherited type in which the
‘praeverbium’ is put at the beginning and between it and the verb comes often
a personal pronoun, as here, and equally often a particle; cf., e.g., Hipponax
(Pap. Oxy. 2176 fr. 1, col. x. 13) σύν roi (= σοί) τι βουλεῦσαι θέλω, Hdt. 2. 181. 3
κατά με ἐφάρμαξας, Ar. Wasps 784 ἀνά τοί με πείθεις, Birds 1506 ἀπὸ γάρ μ'
ὀλεῖς, Plut. 65 ἀπό ο᾽ ὁλῶ κακὸν κακῶς, E. Her. 1053 διά μ᾽ ὀλεῖτε, El. 1121 av’
αὖ σὺ ζωπυρεῖς νείκη νέα, etc.; so in the old formulae (Festus, p. 190. 2 M.)
ob vos sacro, sub vos placo; cf. Wackernagel, Syntax ii. 172 ff.
She does not say merely μαντείας (the word occurs in the hymn to Hermes)
or anything of the kind, but by using ὀρθομαντείας refers to 1195 ff. (ἢ ψευδό-
μαντίς εἶμι krÀ.). The genuineness of her gift of prophecy, which she has just
expressly vindicated, is of such moment to her that she stresses it despite the
agony of the beginning of the seizure ; cf. on 1241. The earlier visions (1095 ff.)
came to light simply as visions, without the speaker alluding to her prophesy-
ing (contrast the words of the coryphaeus 1098 f. κλέος μαντικὸν... προφήτας,
1105 τούτων τῶν... μαντευμάτων, 1113 θεσφάτοις, SO 1130, 1132, 1154); only
after the true visions have faded away does she say in her lament (1161) that
she will θεσπιωιδήσειν in Hades.
1216. ροιμίοις. The context here, as well as Aeschylus’ use of the word
in other passages where it occurs, leaves no doubt that in it the idea of a
I Cf. Gilbert Murray, The Oresteia (transl.), p. 13.
557
line 1216 COMMENTARY
beginning, a prelude, predominates. The expression is readily intelligible
here at the start of a new access of trance."
It is universally admitted that it is impossible to keep ἐφημένους, which has
made its way into this line from the end of the following one. Wilamowitz has
repeatedly (Hermes, xviii, 1883, 246 n. 2, Neue Jahrb. f. d. klass. Altert. 1912,
462 = Kl. Schr. i. 365, Atschylos Interpr. 160) expressed the view that
Aeschylus here wrote an iambic dimeter. But his discussion does not get at
the root of the matter. For shorter iambic elements interrupting the tri-
meters (disregarding interjections) he cites the following parallels: Ag. 1214,
1256 f., 1315; S. Tr. 865, 868, Oed. R. 1468, 1471, 1475, Phil. 219, 750, 787, 804,
Oed. C. 315, 318, Ichn. (fr. 314 P.) 101, 103, 132, 134; E. Iph. A. 1133 (for the
last passage cf. D. L. Page, Actors' Interpolations, 114). Dimeters are found
only in this Cassandra scene: 1214, despite κακά, scarcely differs from mere
exclamations ; and, for the purposes of our present examination, the form in
which Wilamowitz prints 1256 and 1257 may also be regarded as established
(details ad loc.). All the rest are monometers, some acatalectic, some cata-
lectic (on the exceptional τί φῶ Oed. C. 315 see below). It is common to all
the examples, first, that they are real exclamations and, secondly, that the
words contained in these short verses form separate syntactical units.^ The
instance which deviates farthest from the norm is S. Phil. 787 προσέρπει, one
of the cries of pain wrung from the tormented hero, but there, too, the
syntactical independence is clear (the τόδε of the following sentence should
not be made the subject of προσέρπει). In the rest we have, apart from the
obvious exclamations in Ag. 1256 f. (cf. ad loc.), Ag. 1315 iw ξένοι, Phil. 219,
and Oed. C. 318 τάλαινα, imperatival exclamations such as Oed. R. 1468 ἴθ᾽
ὦναξ, Phil. 750 ἴθ᾽ ὦ παῖ, Ichn. 101 ἰδού, ἰδού, 103 ἄθρει μάλα, 132 ἄκουε δή, 134
ἐμοὶ πιθοῦ, Iph. A. 1133 ἔχ᾽ ἥσυχος (uneuripidean, as Wilamowitz has rightly
insisted, cf. my footnote p. 580 n. 4), Trach. 868 ξύνες de (with a subordinate
clause attached), interrogative exclamations: Trach. 865, Oed. R. 1471 τί
φημί; 1475 λέγω τί; Phil. 804 τί φής, παῖ; Oed. C. 315 τί φῶ; (abnormal;
Elmsley read φημῶ. In other words, there is not the remotest parallel for the
interposition between trimeters of any chance piece of a statement of fact
in the form of a dimeter (or monometer), such as the orpoBet ταράσσων
φροιμίοις which Wilamowitz postulates here. We must therefore suppose
1 Recently this passage as well as others in which προοίμιον occurs has fallen into the
hands of apprentices who have practised on the word a sort of tight-rope-walking, with
considerable risk to their necks. Thus Herbert Meyer, Hymnische Stilelemente in der
‘frithgriech. Dichtung (Würzburg 1933), 29, alleges that the meaning ‘beginning, prelude’ is
inappropriate here. According to him ‘muss προοίμιον das Offenbarte heissen können’.
With equal gift of divination he translates Prom. 740 f. obs γὰρ νῦν ἀκήκοας λόγους εἶναι
δόκει σοι μηδέπω ᾽ν προοιμίοις : ‘die Worte, die du jetzt von mir gehört hast, scheint dir
bisher noch niemand offenbart zu haben’. Less foolish, but not less perverse, is the treat-
ment of Ag. 1216 by Robert Böhme, ‘Das Prooimion’, Bausteine zur Volkskunde u. Religions-
wissenschaft, vol. xv, 1937, 20,67. He asserts that prooimia are nomoi, for which (p. 83 n. 182)
‘this passage of Aeschylus is conclusive evidence’; a misinterpretation of Ag. 1153 nicely
fits in with the picture.
2 The same applies to the iambic monometers inserted ‘with a highly tragic effect’
(Wilamowitz, ad loc.) in a longish paratragic passage Ar. Lys. 710 f. On the other hand,
the anapaestic (paroem.), not iambic, kolon S. El. 1162 (cf. also 1160 f.) ὦ δεινοτάτας οἴμοι
μοι stands in the closest syntactical connexion with the following trimeter. The invocations
of Hades and the lightning of Zeus Trach. 1085 f. couched in three anapaestic metra are
syntactically isolated.
558
 COMMENTARY line 1219

with Hermann that the intrusive ἐφημένους has displaced the true ending of
the verse, as happened in Cho. 773 (particularly valuable because the Homeric
scholia have preserved the true reading), 875 (cf. 872), Sept. 463 (cf. 465), and
in a number of other places,! e.g. Agathon fr. 4 N. (Athen. ro. 454 d), where
at the end of l. 5 two words are wrongly repeated from the end of I. 2, Theo-
critus 14. 60 (admittedly a complicated case; cf. the Antinoe papyrus, with
Pohlenz's comments, Gött. gel. Anz. 1931, 369); for instances in the text of
Apollonius Rhodius see Wilamowitz, Hellenistische Dichtung, ii. 249. In
Lucretius 4. 990 the impossible ending of the line saepe quiete comes from 999.
‘Quid scripserit Aeschylus, certe dici non potest’, says Hermann; exempli
gratia he puts δυσφροιμίοις in his text; and certainly one would expect an
epithet with φροιμίοις ; cf. Sept. 7. (Cf. Culex 27, Stat. Theb. το. 113.]
στροβεῖ: so Cho. 1051 f. of the delusions which are driving Orestes dis-
tracted: τίνες ce δόξαι... στροβοῦσιν;
1217 f. In F there is a mark of interrogation after μορφώμασιν (in G and Tr
a comma). Hermann has nicely shown that the question is far more effective
than an imperative would be: ‘ipsa quaestio statim prodit ipsam videre quod
alii non videant: quo statim horror animos corripit.' In what follows she no
longer asks questions but states what she sees: παῖδες θανόντες... πρέπουσι.
δόμοις épnuévous: ‘sitting by, close to the house’. It may perhaps be
tempting to trace in these words the effect of the widespread idea that
ghosts (and especially the ghosts of those slain by violence)? haunt above all
the threshold, the door-jambs, or other places in the vicinity of doors.? But
there is nothing to justify so detailed a picture, since neither the door is
mentioned here, nor its threshold or jambs.
1218. ὀνείρων προσφερεῖς μορφώμασιν. This expression was long ago com-
pared with Prom. 448 ὀνειράτων ἀλίγκιοι μορφαῖσι. As regards the matter, this
is a true description of the spirits of the dead, which were known to the
living above all from their appearance in dreams; cf. Walter F. Otto, Die
Manen (1923), passim.
1219. The difficulty of ὡσπερεί did not escape Hermann ; against the tradi-
tional interpretation (pueri mortui quasi ab amicis' Stanley) he rightly
objects that ‘in occisis non apparet a quibus sint occisi’,* but Hermann's
1 The last three syllables of Ar. Wasps 1514 have come from 1504 (G. Hermann), the close
of 1507 perhaps from 1508 (Wilamowitz, Berl. Sitzgsb. 1911, 508 n. 1 = KI. Schr. i, 326). In
E. Phoen. 630 the beginning up to and including ἄκων δ᾽ is perhaps genuine, the end ἐξελαύ-
νομαι χθονός derived from 627 (so Geel, Schóne, A. C. Pearson, and Wilamowitz in T. v.
Wilamowitz, Dramat. Technik des Soph. 374 n. 1); but it is also possible that the whole line
is the pasticcio of an interpolator (so Valckenaer). .
2 On the βιαιοθάνατοι, who are debarred from all rest, cf. E. Rohde, Psyche, ii, sth ed.,
412; E. Norden, Vergils Aeneis Buch VI, 2nd ed., p. 11 ff.
3 In the books and articles cited by E. Norden, Aus altrómischen Priesterbüchern, 150 n. 1,
I have not found any conclusive evidence from Greek regions for this conception (the spirit
of the dead waiting outside the door); the general idea inferred from θύραζε κῆρες is well
known. On the whole the door is the 'critical point where good and evil spirits can throng
in and out’, ‘a place where spirits and ghosts whirl about’ (S. Eitrem, Hermes und die Toten,
Christiania 1909, 14, 40); Weiser-Aal in the article ‘Schwelle’, Handwörterb. des deutschen
Aberglaubens vii. 1518, says (of German regions) : 'Ghosts cannot cross the threshold.'
^ In spite of that Wecklein could write: "They look to her as though they have been
killed by their own kin', Plüss: 'the god knows that the children of Thyestes were killed
by their uncle Atreus, but with sublime irony he makes the prophetess express this only as
a conjecture. Both explanations are unintelligible to me alike in point of content (consider
559
line 1219 COMMENTARY

own solution seems desperate: ' worepei refertur ad παῖδες, quasi pueri
videntur’, On παῖδες... ὡσπερεί, as Hermann takes it, Paley remarks ‘the
words are out of their natural order’, but nevertheless he attempts to defend
this interpretation, Aivdov πλύνων. More recent editors have apparently
believed that they could make sense of ὡσπερεί without using violent means ;
cf., besides the passages quoted in the last footnote, e.g. Headlam: ‘children
slain as it were by their own kindred’, but it is hard to see what ‘as it were’
can mean here. Several critics who are quoted by Ahrens, 629, and Ahrens
himself regard ὡσπερεί as corrupt. A glance at the use of worepei confirms
their doubts and at the same time shows the wrongness of translations such
as ‘as it were’, ‘on dirait’ (Mazon). In Aeschylus and Sophocles (there is no
need to go farther afield) ὡσπερεί is found only in true comparisons: Ag. 1415,
Cho. 753, S. Oed. R. 264 f. ἀνθ᾽ dv ἐγὼ τάδ᾽, ὡσπερεὶ τοὐμοῦ πατρός, ὑπερ-
μαχοῦμαι (intended by Oedipus simply as a comparison; that in reality
it is more than this he does not know), Soph. fr. 33 N. (= 36 P.) ὑφηιρέθη σου
κάλαμος ὡσπερεὶ λύρας. Again, in thirteen of the fourteen places in which
(according to Todd's index) ὡσπερεί occurs in Aristophanes it introduces a
genuine comparison (only Birds 51 ἄνω κέχηνεν ὡσπερεὶ δεικνύς τί μοι is some-
what different). With this use of ὡσπερεί with a nominal expression agrees
the use of ὥσπερ εἰ with hypothetical clauses, e.g. Ag. 1201 ὥσπερ εἰ mape-
ordres, Ar. Birds 282 f. ὥσπερ εἰ λέγοις ᾿Ϊππόνικος Καλλίου κτλ., Frogs 1158
ὥσπερ γ᾽ εἴ τις εἴποι γείτονι κτλ., Eccl. x26 f. ὥσπερ εἴ τις σηπίαις πώγωνα
περιδήσειεν ἐσταθμευμέναις. There is no connexion, then, between this usage
and the meanings assumed for ὡσπερεί by the defenders of the MS tradition
in the passage before us. A corruption must therefore be taken as highly
probable. None of the conjectures proposed is convincing; Maehly’s sug-
gestion, to ascribe ὡσπερεί to the effect of προσφερεῖς, which is immediately
above it, and to read θανόντες νηλέως π. τ. $., is more in accord with the
style of the passage than are most others, but has no real probability either.
1221. yépos only here.
1222. πρέπουσι: ‘conspicui sunt’ (Blomfield) ; cf. on 242.
1223. ἐκ τῶνδε xrÀ.: a particularly clear indication of the emphasis laid
throughout this whole section of the play on the causal connexions of events
and, above all, on the idea that guilt inevitably results in retribution ; cf. on
1193 and on 1202 ff. In exactly the same way when Aegisthus is speaking
of these same things (1603) he uses ἐκ τῶνδε in passing from the slaughter of
Thyestes’ children to the retribution for it, and in Zum. 550, after giving
rules for the punishment of the guilty and for the conduct of good, godfearing
folk (in the sense of Eum. 270 ff. and Suppl. 707 ff.), the Chorus proceeds: ἐκ
τῶνδ᾽ 1 ἀνάγκας ἄτερ δίκαιος ὧν οὐκ dvoAßos ἔσται, again with reference to the

also how definite are all Cassandra’s other pronouncements about her visions) and in point
of language (on ὡσπερεί see above). Wilamowitz’s free translation 'slain children—was it
their uncle that slew them?’ seems to indicate that he took the passage in the same way as
Stanley and the editors just mentioned. Verrall translates ‘as infants slain by their parents
they appear’ and surmises ‘a belief that the children of infanticides haunted the house in
this way”. | |
1 Despite Verrall (‘according to this law’) and Wilamowitz (‘ex huius praecepti obser-
vantia’), editors still prefer the alteration ἑκών to the MS reading, which derives from the
very heart of Aeschylus’ thought. In his first edition (1861) Weil’s judgement was sounder
than later. For the metrical responsion see above on 210,
560
 COMMENTARY line 1224

causal connexion based on the moral code. So S. El. 570 ἐκ τοῦδε, where the
.punishment exacted by the goddess is the result of Agamemnon’s presump-
tion; Eur. Cretans, Berl. Klass. Texte v. 2, p. 75, 1. 25 f. = D. L. Page, Greek
. Lit. Pap. i. 74 (Minos has broken his word to the god) : ἐκ τῶνδέ τοί σ᾽ ὑπῆλθε
κἀπετείσατο δίκην Ποσειδῶν.
ποινάς put near the beginning of the sentence, since it is on this that the
strongest emphasis falls: ‘as a consequence of this comes retribution . . .’;
the manner and the instrument of the retribution are of comparatively minor
importance.
βουλεύειν : here probably still quite general, ‘plan’, but possibly it points
forward to the technical use in the accusation of Aegisthus in the concluding
scene, where a prominent place is given to the charge that he is answerable
for the murder, not indeed as χειρὶ ἐργασάμενος but as βουλεύσας (1614, 1627,
1634).
1224-6. That the text has here suffered badly is certain, but a definite
judgement on 1224 is, to me at any rate, unattainable. As a preliminary, we
may dispose of a detail which concerns not a real difficulty but the caprice of
many commentators, viz. the syntactical reference of τῶι μολόντι δεσπότηι.
The earlier editors (e.g. Stanley, Schütz) made the dative depend on ποινὰς
βουλεύειν, and so more recently, e.g., Wecklein, Pltiss, Wilamowitz. On the
other hand, Humboldt translates: ‘auflauernd, weh! im Hause meinem
kehrenden Gebieter’; like him nearly all English translators (W. Sewell,
Kennedy, Davies, Lawson are exceptions), from Symmons (quoted by
Conington) down to Headlam, Murray, MacNeice, G. Thomson, A. Y. Camp-
bell, though they differ as to the details of the interpretation, make τῶι
uoAövrı δεσπότηι depend on οἰκουρόν; Mazon does the same. Sidgwick ex-
pressly remarks: ‘the δεσπότηι is dative after οἰκουρόν, as the order requires’,
but gives no proof of the last statement. This way of construing the sentence
is certainly wrong. οἰκουρός (on the sense cf. on 1625 ff.) is always used
absolutely and never with a dependent dative. As regards the dative in
1225, it ought not to be overlooked that those passages to which this sentence
paves the way, 1627 and 1634, have respectively ἀνδρὶ στρατηγῶι and τῶιδε
with ἐβούλευσας μόρον. There is, besides, the general consideration that from
Homer on, if the object of βουλεύειν is, like ποινάς here, an expression of harm
(e.g. ὄλεθρον, πῆμα, etc.), it is usual to add a dative expressing the person
affected. τῶι pod. δεσπ. must, then, be connected with βουλεύειν. After read-
ing so many mistranslations, one appreciates the sound judgement of either
Triclinius or the unknown Byzantine scholar whom he followed : τῶι μολόντι"
τοῦτο πρὸς τὸ βουλεύειν σύναπτε.
1224. It is hard to swallow λέοντ᾽ ἄναλκιν. What I have seen of attempts to
vindicate it are at best mere subterfuges. Leaving on one side the particular
context here, it would be for a Greek one might say an offence against the
laws of nature to call a lion—of all creatures—-ävaArıs : Philemon fr. 89 K. τί
more Προμηθεύς, ὃν λέγουσ᾽ ἡμᾶς πλάσαι καὶ τᾶλλα πάντα ζῶια, τοῖς μὲν θηρίοις
ἔδωχ᾽ ἑκάστωι κατὰ γένος μίαν φύσιν; ἅπαντες ot Àéovrés εἰσιν ἄλκιμοι" δειλοὶ

πάλιν ἑξῆς πάντες εἰσὶν οὗ λαγώι κτλ, It should not be forgotten that in
Cassandra’s words, as in other oracles, the various beasts take the place of
definite individual men precisely because of their characteristic qualities. It
is impossible to discover any grounds on which Aegisthus could here be
4872.3 G 561
line 1224 COMMENTARY
termed a lion, in glaring contradiction of the verse which follows soon after,
1259. Wecklein, like Hartung before him, is quite beside the mark: ‘a lion
only in ferocity and destructiveness, not in courage, no true lion (λέοντος
εὐγενοῦς 1259)’. Does ferocity and destructiveness in lions really express itself
in scheming, since, as has been already demonstrated, βουλεύειν, which
excludes action proper, is here full of significance? ' λέων obviously sarcastic
and deliberately corrected in 1259’, says Pohlenz, Griech. Tragödie ii. 34. To
me this seems no whit less unnatural than to imagine that Hamlet (Im. ii. 278)
could ‘sarcastically’ say ‘this realm dismantled was of Jove himself; and now
reigns here a very, very—lion’. Even allowing for the irony one would still
have to discover some starting-point for the equation Aegisthus = lion, but
that is exactly what is missing. We are, indeed, assured (by Blaydes, Head-
lam, Mazon) that λέοντα applies to Aegisthus as a Pelopid ; so Paley (on 827 =
800 Paley), following Donaldson, alleges that the lion was the symbol of the
Atridae, adducing as evidence the Lion Gate of Mycenae (so also Headlam,
Mazon ad loc., Charlesworth, C.R. xl, 1926, 5, who strangely enough drags
in Cho. 938, W. B. Stanford, Aesch. in his Style, 90). But I have been able to
find no evidence that a king or a person of princely rank could as such be
termed a ‘lion’—not even in Lycophron, since Alexander, the fierce warrior,
the descendant of the lion Herakles, is, of course, a special case.
ἐν λέχει στρωφώμενον as it stands can mean only ‘der sich im Bette
wälzt’ (Nägelsbach), ‘that tumbles in the bed’ (Headlam). The use of the
verb here may be compared not so much with the Homeric instances as
with Oppian Cyn. 3. 425 f. (quoted in the Thesaurus), where the crocodile is
described thus: ἄλλοτε δ᾽ αὖ ψαμάθοισι κυλινδόμενος ποτὶ χέρσον, ἄγριον ἀσθμαΐί-
νων, στρωφώμενος ἀμφ᾽ ὀδύνηισιν. It is quite conceivable that Aeschylus found
something of the sort in later epic poetry. Strictly, then, it is only the in-
dolence of Aegisthus which is here denoted, not his adultery. To this, in
sharp contrast with 1258 f., nothing in the whole passage points. When in
L-S s.v. στρωφάω we find after ἐν λέχει στρωφώμενος the remark ‘i.e. claiming
a husband’s rights’, that is less a paraphrase of the actual text than a para-
phrase of the text the reader desiderates. The desire is intelligible enough,
not only for the sake of the subject-matter of the Oresteia. van Heusde com-
pared Menander fr. 535. 8 K. γαμηλίωι λέχει τε μοιχὸς ἐντρυφῶν. If we had
λέοντος ἐν λέχει, there would be no difficulty about this, and we should have
to consider whether the passage might not be a free translation of 8 333 ff.
(= p 124 ff.), particularly as there is an analogy in subject-matter (Penelope’s
suitors: Aegisthus) and as there, too, ἀνάλκιδες is very emphatic. But it is
only by expelling ἄναλκιν that one can get λέοντος into the verse ; one might
amuse oneself with, e.g., λύκον λέοντος €. À. orpwe.', for in conjunction with
the lion one would expect a beast, but ἄναλκιν, as has long been noticed, is
supported by y 310 ἀνάλκιδος Αἰγίσθοιο, S. El. 310 (Aegisthus) ὁ πάντ᾽ ἄναλκις
οὗτος. Taking everything together I can only say: I do not believe that the
poet wrote the line in the form in which we have it ; beyond that I cannot go.
Certainly, bracketing 1224, as Wilamowitz does, removes all difficulty and
yields the additional advantage that the dative τῶι μολόντι δεσπότης comes
closer to BovAevew. But does the line look like an interpolation?
1224-6. 1226, on the other hand, is indubitably interpolated. A. Y. Camp-
! But, other things apart, λύκον λέοντος would weaken 1259.
562
 COMMENTARY lines 1224-6

bell (C.Q. xxix, 1935, 27 f.) deserves our thanks for joining A. Ludwig and
Herwerden in the deletion of the line. One obviously offensive thing about it
is that it makes the relation between τῶι μολόντι δεσπότηι and other parts of
the sentence appear equivocal. It is, of course, possible to slur this over, as
has often been done, but with conscientious interpreters the inconsistency is
apparent ; cf., e.g., the translations of Voss (‘Darob Vergeltung, sag’ ich, wird
aussinnen wer... . Haushüter, weh! weh! ihm dem heimgelangten Herrn,
ach meinem’) and of Headlam (‘to keep it safe... for the returning master—
my master’). For the present purpose it does not matter whether the dative
is taken with oixovpóv (which has above been proved wrong) or with βουλεύειν;
the essential fact is that the hearer who reaches the expression τῶι μολόντι
δεσπότην via the particular sentence here, that is, via the words ἐκ τῶνδε...
oikoupov, cannot possibly help relating δεσπότης to what goes before it (on the
meaning in detail see below). But if ἐμῶν is subsequently added, the effect on
the hearer (though not on the reader who first reads forward from the
beginning to the end and then backward from the end to the beginning in
order to ‘construe’) will be that of a παρ᾽ ὑπόνοιαν of a kind unexampled in
this style.
No less offensive is the fact that 1226—and 1226 alone—makes Cassandra
deviate for ἃ moment from the course maintained unwaveringly throughout
this speech, and that in a way utterly unworthy of the figure drawn by the
poet with so magnificently firm a hand (good observations on this are made
by A. Y. Campbell. The words which, under the compulsion of the new
onset of trance, flow from the lips of the prophetess in unbroken succession
and coherence concern exclusively the chain of destiny forged from the links
of guilt and counter-guilt. Dominating all comes the murder of Agamemnon,
which Cassandra finally, provoked by the old men's lack of comprehension,
shortly and brusquely announces (1246). To tolerate 1226 is to assume that
the prophetess, who in this whole section serves merely as the instrument for
communicating a tremendous decree of destiny, could for a short moment
forget the noble austerity of her bearing to cast a side-glance at her own affairs
and, sighing like a sentimental girl, complain of her sad fate in terms of a
commonplace maxim. We should hardly be worthy of reading this great
scene if we allowed ourselves to acquiesce in such an inconsistency.
What the poet has here written is plain and simple: ποινάς φημι βουλεύειν
ruw® . . . oùkoupôr . . . τῶι μολόντι δεσπότηι. μολών of course means here ‘re-
turned’, not ‘war-gone’ (Murray); cf. on 675. The murder-plot, horrible in
itself, is the more horrible because the plotter is one who has stayed quiet at
home, and the victim is the master of the house who has returned home after
long absence at the war. ἔφθιθ᾽ οὗτος οὐ καλῶς, μολὼν ἐς οἶκον (Eum. 458 f.).
It is precisely the grievousness of this circumstance which calls forth the cry
of woe, to the poignancy of which so many critics have been deaf. The
position of οἴμοι in the middle of the trimeter is rare, but is found also in
S. Phil. 363 οἱ δ᾽ εἶπον, οἴμοι, τλημονέστατον λόγον. Cassandra speaks of Aga-
memnon as master not, probably, in the sense that Aegisthus is subject to
him, but because Agamemnon is the master and rightful head of the house
and household in which Aegisthus, the oëkoupés, has established himself
1 The dash which Browning and Headlam in their translations put after ‘master’ clearly
shows that they were aware of the difficulty.
563
lines 1224-6 COMMENTARY

during its owner’s absence. To the Watchman (35) Agamemnon is ἄναξ

οἴκων; in the absence of the mistress φάσμα δόξει δόμων ἀνάσσειν (415). Cf.
also Pers. 169 ὄμμα γὰρ δόμων νομίζω δεσπότου παρουσίαν, Eum. 60 τῶνδε
δεσπότης δόμων (Apollo in Delphi), E. Alc. 681 οἴκων δεσπότην,Ϊ 204. τούτους
(the children of our marriage) ἀνάσχου δεσπότας ἐμῶν δόμων, Ton 1535 f. καὶ γὰρ
ἂν φίλος φίλωι δοίη τὸν αὑτοῦ παῖδα δεσπότην δόμων.2
Now that 1226 has been proved spurious we may inquire into the origin of
the interpolation ; the question, which is after all of subordinate importance,
can only be answered conjecturally. A. Y. Campbell supposes that ἐμῶι was
first added as a gloss and then extended to a trimeter; cf. in general Wilamo-
witz, Praef. xxviii. Choosing at random, I quote as a Homeric instance of
such an interpolation developed from an explanatory addition a line which
is missing in the majority of good manuscripts, © 183.3 The same expansion
of a gloss into a complete verse is probable in Ag. 7 and in other analogous
cases.* Suffice it to refer to Cho. 562 (deleted by F. W. Schmidt) Πυλάδηι"
ξένος τε καὶ δορύξενος δόμων" and to S. Oed. R. 827, where the deletion suggested
by Wunder was adopted by Dindorf, Nauck, Herwerden, and Jachmann,
Nachr. Gott. Ges. 1936, Phil.-Hist. Kl. Fachgruppe i, N.F. Bd. 1, 196. No
matter, then, whether Ag. 1226 arose by repeated expansions or by a single
one; what is obvious is that the interpolator missed some expression of the
reference of δεσπότην (whose master?) and aimed at making the passage
clearer by his addition. Here again we have a case of ὑποβληθέντα πρὸς
σαφήνειαν τῶν λεγομένων (Schol. S. 47. 841, cf. D. L. Page, Actors’ Interpo-
lations in Greek Tragedy, 1934, 117, and the same author's Greek Literary
Papyri, 263; Jachmann, op. cit. 195 ff., 200 f.; Dodds on E. Baech. 1092;
examples of this type of interpolation are also given by Wilamowitz, Euri-
pides Ion, p. 28). The spurious stanza Horace, Odes 3. 11. 17-20, is a classic
example.
! L-S s.v. οἰκοδεσπότης quote no earlier instance than Xenophon for οἴκων δεσπότης and
for οἰκίας δεσπότης cite Plato's Laws.
2 No one in classical times would know that δεσπότης (from *&euarérns) originally denoted
the master of the house, but in connexions such as those cited the living use tallies with the
derivation.
3 Wilamowitz, on E. Her. 1140, put in the same category the line P 244, which had been
obelized by Heyne (cf. also Leaf, ad loc.). However, he was led by Hermann Fränkel’s
treatment of the passage (Die homer. Gleichnisse, 24) to recant his earlier view (Pindaros,
427 n. 1).
* In Ar. Frogs 168, too, I should like to think that first of all someone—a pedant—wrote
in τῶν ἐκφερομένων as an explanation of τινα and then someone else added the completely
inappropriate ὅστις ἐπὶ τοῦτ᾽ ἔρχεται. This, however, is not demonstrable. In any case,
there can be no doubt that the verse, which utterly ruins the surprise of the subsequent
little scene with the dead man on his bier, has been rightly deleted by Hamaker. J. Vahlen,
Opusc. acad. i. 286, defends the line with the usual artifice of ‘conservative’ criticism:
'ineptus esset, si haec diceret μίσθωσαί τινα neque rationem ostenderet qua id in hoc itinere
inferno fieri posset', just as if the delightful thing about the proposal made by Xanthias
were not that he speaks exactly as one might in the streets of Athens or some frequented
main road.
5 Conington, Wecklein (annotated edition), Sidgwick, Verrall, Tucker, and Blass are
thoroughly conscious of the difficulty, but nevertheless retain the verse. It is only in one
detail (* ξένος alongside δορύξενος is nonsensical' that Wilamowitz gave a more critical
verdict in his commentary than in his complete edition of the plays. H. J. Rose, who
(Annual Brit. School of Athens, xxxvii, 1940, 205) defends the verse and reads ξένος δὲ, does
not mention that this conjecture of Meineke was put by Wilamowitz into his text.

564
 COMMENTARY line 1227

Finally, so far as concerns the cheap maxim φέρειν γὰρ χρὴ τὸ δούλιον ζυγόν,
it is clearly a patch of the stuff which anyone even moderately acquainted
with tragic diction had always available! E. Or. 1024 φέρειν σ᾽ ἀνάγκη τὰς
παρεστώσας τύχας is of exactly the same stamp; this line (deleted by Kirch-
hoff) is an unmistakable interpolation; the scholia attest its absence.?
Whether there is any causal connexion between the insertion of Ag. 1226 and
the corruption (or interpolation?) of r224 I cannot say, but I do not think it
probable.
1227. Instead of νεῶν 7’ Gerard Vossius read νεῶν δ᾽, Hermann attempted to
vindicate 7’: ‘nam δέ posito γλῶσσα μισητῆς κυνός sic inferretur ac si iam
praegressa esset mentio mulieris. In mente habebat Cassandra dicere ποινάς
φημι βουλεύειν λέοντα ἄναλκιν καὶ μισητὴν κύνα. Nunc mutata orationis forma
copulam tamen retinere debuit.’ This is wrong in several respects. The
expression ἐκ τῶνδε (see above), which refers to the feast of Thyestes, pre-
pares the way for the mention, not of Aegisthus and Clytemnestra, but only
of Aegisthus as his father’s avenger. Moreover, μισητῆς κυνός 15 by no means
intended from the outset as a counterpart to λέων ἄναλκις (9), but arises
spontaneously as a reinforcement of ola γλῶσσα λέξασα (ef. on 1228). Above
all, the marked parallelism of νεῶν τ᾽ ἄπαρχος and ' IMov 7” avaorarns makes it
impossible to follow Hermann in taking the first re as connecting this clause
with the preceding sentence; e... re must go together as in πατὴρ ἀνδρῶν
τε θεῶν re, etc. Thus, if we retain νεῶν 7’, we have an asyndeton. This is how
the sentence has been translated, alike by older editors (e.g. Stanley, Hum-
boldt) and more recent ones (e.g. Wilamowitz, Headlam).? The asyndeton
would be of the type whereby something given in general terms in the pre-
ceding sentence is explained more fully. Accordingly Pliiss observes: ‘The
connexion with the preceding sentence is asyndetic: the thought “a recreant
avenger is lurking in the house on the watch for my master” is amplified with
striking effect: ‘‘the great warrior has no suspicion of death at a woman’s
hand".' I find it difficult to recognize in the second sentence an explanation
or elaboration of the first; on the contrary, it seems to me obvious that
Cassandra is now passing by a definite transition to Agamemnon’s lack of
suspicion ; I therefore agree with Schütz: 'adversativa hoc loco requiritur'.
ἄπαρχος was vindicated by Ahrens (p. 629 f.), who compared it with Pers.
327, Cho. 664 (uncertain), Pind. N. 4. 46 ἔνθα Τεῦκρος ἀπάρχειϊ (cf. Schroeder
in the apparatus to his editio mai.); Verrall also treats the word correctly ;5
Wilamowitz contributed a reference to the inscription of Thasos, JG xii. 8.
273 ἐπὶ τῆς πρώτης ἀπαρχῆς, and above all to the archaic bronze phiale from
! Cf. also p. 580 n. 4 (regarding E. Or. 478).
2 Just as there, so in E. Bacch. 1028 the aposiopesis after ἀλλ᾽ ὅμως led to an interpolation
to complete the sense; see among others Jachmann, op. cit. 208 n. 2. Here too the content
of the interpolation is a hackneyed maxim on the same plane of thought : χρηστοῖσι δούλοις
συμφορὰ τὰ δεσποτῶν. [Cf. Dodds ad loc.]
3 Mazon, on the other hand, inconsistently keeps νεῶν τ᾽ but translates ‘Et le chef de la
flotte, le destructeur de Troie ne sait pas.
4 Snell adds Bacchyl, 12 (11). 4 ff. ἐς γὰρ ὀλβίαν ξείνοισί με πότνια Nixa νᾶσον Aiyivas
ἀπάρχει ἐλθόντα κοσμῆσαι θεόδματον πόλιν, where the sense of the verb appears to be ‘orders’
or ‘orders away”.
5 In L-S s.v. drapyos we find only ‘falsa lectio for ἔπαρχος, A. Pers. 327’, which is less
than can be learned even from Passow’s treatment of the word. The article drépyw in L-S
shows that no use has been made of the index of Schwyzer’s Exempla.

565
line 1227 COMMENTARY

Thebes, the inscription on which can now be found most conveniently in

Schwyzer, Exempla, no. 440. 11 (p. 226 1.) : it contains the dating Φλόρακος
ἀπάρχοντος Aeprois Θεβαίοις, translated by Kretschmer, Jahresh. d. Oesterr.
Inst. iii, 1900, 137: ‘when Phlovax was at the head of chosen Thebans’.
Wilamowitz (Griech. Trag. ii. 205, cf. also? the note in his edition on A. Pers.
327) remarks on this: ‘aus der boeotischen Inschrift folgt, dass drapyos und
ἀπάρχειν im Sinne von Herrschen, Kommandieren gesagt worden ist, mag
die Nuance der Bedeutung auch diskutabel sein’. The qualification will
presumably apply to the Aeschylean passages as well; on Pers. 327 Wilamo-
witz, Griech. Lesebuch, ii. 34, observes: ' ärapxos without any noticeable
difference from apxos ’.
ἀναστάτης only here; ἀναστατήρ (presumably poetic, cf. Ernst Fraenkel,
Nomina agentis, ii. 8) only Cho. 303 and in a line of the diaskeuast Sept. rors.
1228-30. With οὐκ oldev ofa . . . τεύξεται cf. E. Tro. 431 δύστηνος, οὐκ οἶδ᾽
old νιν μένει παθεῖν. The MS reading καὶ κτείνασα makes nonsense. Canter’s
interpretation of it as κἀκτείνασα is necessary and almost universally accepted
(not by Ahrens). Tyrwhitt’s λείξασα must be rejected; Ahrens, who adopted
it (p. 631), endeavoured to force on the passage a wild confusion of images.
With ofa . . . λέξασα κἀκτείνασα Cassandra recalls the long hypocritical
speeches of the queen (from 855 on), which she had heard delivered; in using
ἐκτείνασα she, like Agamemnon (916 μακρὰν γὰρ ἐξέτεινας), stresses the pro-
lixity which must have struck every hearer. The reference to the scene during
which Cassandra and the Chorus alike heard with their own ears the
perfidious words of Clytemnestra should, one would think, be intelligible
enough; a superfluous but most welcome corroboration of it is provided
immediately afterwards by 1236 ws δ᾽ ἐπωλολύξατο recalling the conclusion of
that very scene (973 f.).
' ofa is understood with Adfaca as well as with τεύξεται as though it were
ola λέξασα οἷα τεύξεται "2 (Headlam). τεύξεται was derived by Stanley,
Schütz, and others from τεύχω; so, with detailed argumentation, Ahrens
(p. 632), and, of more recent editors, e.g. Wecklein, Wilamowitz, Headlam,
Blass (on Eum. 30 ff). On the other hand, Hermann protested against this,
deriving the word from τυγχάνω, and was followed by many scholars. But
the sense demands, not ‘she will obtain’, but ‘she will work, cause, bring
about’. Ahrens aptly recalls such combinations as τεύχειν θάνατον, ἄλγεα,
etc., as well as the fact that in À 430 κουριδίωι τεύξασα πόσει φόνον is said of
Clytemnestra herself. It is true that there seems to be no other certain
example of a future τεύξομαι used in an active sense, since in T 208 τεύξεσθαι
is doubtful (cf. Leaf, ad loc.), and it does not seem safe to rely on the gloss in
Hesychius rev&ouevn‘ ποιήσουσα (cited by Ahrens), as its origin cannot be
determined. But this is no real difficulty, since on the one hand other middle
forms with active sense are certainly attested, e.g. κ 182 τεύχονται, hymn.
Hom. Ap. 76 τεύξασθαι, and on the other ‘so viele Verba activa speziell im
Futurum mediale Flexion haben . . . Nicht alle Fälle brauchen gleich alt zu
sein; im Laufe der Entwicklung können Übertragungen und Nachbildungen
1 In both passages the reference should run ‘ ’E¢np. dpx. 1896, 243’, not ‘1897, 293’.
2 The substance of this explanation was anticipated by Fr. Thiersch, De analog. iii. 27
(quoted in Enger’s revision of Klausen’s commentary), only with the difference that he
derived τεύξεται from τυγχάνειν and therefore paraphrased οὐκ οἶδεν ola λέξασα οἵων τεύξεται,
566
 COMMENTARY lines 1228-30

stattgefunden haben’ (Wackernagel, Syntax, i. 134). Nor is the derivation of

τεύξεται from τεύχω discredited by the fact that Aeschylus elsewhere (Eum.
668) used the future τεύξω.
xuvös:asthetypeofshamelessness. In the Nekyia (A424) Agamemnon speaks
of Clytemnestra as ἡ κυνῶπις and says soon afterwards (427) ὡς οὐκ aivorepov
καὶ κύντερον ἄλλο γυναικός. Scylla, whose horrible deed (she.cut from the head
of her sleeping fathef the magic hair on which his life depended) is in Cho.
613 ff. set beside Clytemnestra’s crime, is there called à κυνόφρων. The
epithet μισητῆς is very strong. Little confidence can be reposed in the ancient
attempts to establish a difference of meaning between μισητή and μισήτη
(besides the lexica, cf. the material collected by Bethe on Pollux 6. 189, 1. 11,
and by Adler on Suidas M 1112). The word as such may not have had the
meaning ‘lewd’ ; it may only have received it on occasion from the context in
which it was employed by such authors as the iambist who says (Fragm.
iamb. adesp. ı Diehl = Archil. fr. 184 Bergk, 4th ed.) περὲ σφυρὸν παχεῖα
μισήτη γυνή or Cratinus (fr. 316 K. μισῆται δὲ γυναῖκες ὀλίσβοισιν χρήσονται) ;
but in the present passage it is clearly contemptuous. This is not the only
place where the excited speech of the inspired prophetess sweeps aside
seemly conventions; cf. on 1267.
In the transition from γλῶσσα... kakretvaca to the end of the sentence
'there is no mixture of metaphor' (Verrall). By a perfectly intelligible liberty,
as soon as ever we pass from speaking to acting the place of the speaking
tongue is taken by its possessor, 'as if she had said, ofa λόγωι λέξασα ἔργωι
ἐκπράξει ' (Paley).
φαιδρόνους has been frequently doubted and altered. With so un-
usual an expression the possibility of a corruption must of course be taken
into account, but I do not see that any really compelling objection has been
adduced. Obviously, in Clytemnestra's case the φαιδρότης of mind is only
appearance; ἀλλά vw (Agamemnon) κελαινόφρων ἐμὴ μήτηρ κατέκτα says
Orestes (Eum. 450 f.). It is to Clytemnestra in the first place that allusion is
made when (793 f.) the king is warned against those who behave fvyyaipovow
ὁμοιοπρεπεῖς ἀγέλαστα πρόσωπα βιαζόμενοι. But that by no means necessitates
such an alteration as φαιδρωπός (Wakefield and, independently, Weil).
Meineke, Philol. xix, 1863, 208, excellently compares, with reference also to
the ἄτη λαθραῖος of this passage, Pers. 97 ff. φιλόφρων yap παρασαίνουσα τὸ
πρῶτον παράγει βροτὸν εἰς ἄρκνας ara (restored with certainty). With regard
to the composition with -voos Munro, J. Phil. xi, 1882, 135, refers to Prom.
1003 θηλύνους and Prom. 383 κουφόνουν (Soph. also); it is perhaps not mere
chance that both occur in a play so close in date to the Oresteia. The meaning
of -vous in φαιδρόνους is the same as in εὔνους, an intensification of which may
be seen in the bold Aeschylean coinage. This sense of νοῦς and its derivatives
is practically alien to Homer.’ But in the post-Homeric compounds εὔνους
(as early as Alcaeus, Pap. Oxy. 2165, fr. 1 col. 1. 9, then A. Pers. 226, etc.),
! The careful examination of the usage of νοῦς etc. in Homer by Joachim Böhme, Die
Seele und das Ich im hom, Epos (1929), passim, has received the addition of some refinements
and corrections from Snell, Gnomon, vii, 1931, 77 f., 83, and from K. v. Fritz, Class. Philol.
xxxviii, 1943, 79 ff. The latter article may help us to understand how the post-Homeric
extension of the meaning which concerns us here was made possible and also why it kept
within narrow limits. For the meaning of νῶς in the remains of Sappho and Alcaeus οἵ,
Lobel, "4Axaíov μέλη, xxxvii n. 1.

567
lines 1228-30 COMMENTARY

εὔνοια (from A. Suppl. onwards), δύσνους (5. Ani. 212, etc.), κακόνους (in
extant texts not before the last quarter of the fifth century), the element -vovs
always denotes a friendly or unfriendly disposition towards somebody. The
same meaning of the simple noun νοῦς is found in a single Homeric passage,
v 229 μή μοί τι κακῶι vow ἀντιβολήσαις ; afterwards it is well attested in the
literature of the fifth century, although it does not seem to have become very
common. To quote at random a few examples, Pind. P. 3. 5 νόον &xovr’
ἀνδρῶν φίλον, 8. 18 and Paean 5. 45 εὐμενεῖ νόωι, Hdt. 1. 60. 5 δέκεσθε ἀγαθῶι
vour Πεισίστρατον. In the Attic σκόλιον (7 Bergk, 6 Diehl), too, the phrase
τὸν νοῦν ἐσιδόντα is Telated not to an intellectual function but to εὔνοια or its
opposite: it would be a good thing if we were able to find out our neighbour’s
disposition towards us.
1229 f. δίκην ἄτης Aaßpaiou probably goes in the first place with what pre-
cedes (cf. Meineke’s observation already quoted), but the idea of the opera-
tion of Ate extends to τεύξεται. At the end of the whole sentence kakfı τύχηι
shows the characteristic Aeschylean reversal of a formula of blessing (cf. on
1387); the dissertation of Gerda Busch cited in the note on 755 gives on
p. 16 n. 26 the few other tragic instances of forms of κακὴ τύχη.
1230. ἄτης Aaßpaiou: cf. S. Ant. 531 ff. σὺ δ᾽, ἢ κατ᾽ οἴκους ὡς ἔχιδν᾽ (cf. Ag.
1233) ὑφειμένη λήθουσά μ᾽ ἐξέπινες, οὐδ᾽ ἐμάνθανον τρέφων δύ᾽ ἄτα κτλ., Phil.
1272 (of one who, like Clytemnestra, hypocritically poses as a true friend)
ἀτηρὸς λάθραι.
1231 f. That τολμᾶ should be read as τόλμα was recognized by Ahrens. For
the structure of the sentence cf., e.g., 613 τοιόσδ᾽ 6 κόμπος. The correct punctua-
tion (not recognized by Ahrens) was suggested by Elmsley (on E. Heracl.
387 Elmsl.) and adopted by Blomfield, and, with the reading of Ahrens, by
Enger. Elmsley realized that ἐστι, when placed inside a sentence, cannot
stand at the beginning of a line! (Wilamowitz followed him, cf. also his note
on E. Her. 1297). Elmsley also saw that the sentence must be punctuated
after φονεύς, but he wrongly altered ἐστιν to elev. Cassandra begins: ‘she is
...’, but then, instead of naming any particular monster, in her agitation she
substitutes a question. This indignant aposiopesis and the interposition of a
question was long ago compared with Ar. Clouds 1378 ὦ---τί σ᾽ εἴπω; and
particularly Demosth. 18. 22 εἶτ᾽ ὦ---τί dv εἰπών σέ τις ὀρθῶς προσεΐποι ; As the
question here replaces a predicative noun, so it replaces an object in Sept. 435
τοιῶιδε φωτὶ meume—ris ξυστήσεται κτλ.2; the aposiopesis depicts the speaker's
perplexity. In Cho. 1073 the substantival part of the subject is replaced by a
question: νῦν δ᾽ αὖ τρίτος ἦλθέ ποθεν---σωτῆρ᾽3" ἢ μόρον εἴπω ;
θῆλυς ἄρσενος: cf. the oracle Hdt. 6. 77. 2 ἀλλ᾽ ὅταν ἡ θήλεια τὸν ἄρσενα
νικήσασα ἐξελάσηι κτλ. and the other in Diodorus 8. 23. 2 ἔνθα... τὸν ἄρσενα
θῆλυς ὀπυίει (Dindorf for orapei) ; Paley compares E. Iph. T. 621 αὐτὴ ξίφει
θύουσα θῆλυς ἄρσενας.
1232. τί... δυσφιλὲς δάκος : this type of hyperbaton in a sentence beginning
1 Römer, Sitzgsber. Bayer. Akad., Phil.-hist. Kl., 1888, 241, without referring to Elmsley,
once again stresses the fact that in Aeschylus ἐστι does not occur at the beginning of the
verse ; he ought to have discussed, or at least mentioned, Ag. 1232.
2 Wilamowitz’s treatment of the passage (cf. his Interpr. 76 n. 2) is not happy; his
punctuation was anticipated by Vahlen, Opuse. 1. 117.
3 The MS reading was already interpreted aright by Jacob, who was followed by Wila-
mowitz, Blass, and Mazon.
568
 COMMENTARY line 1235
with the interrogative pronoun is quite normal; vw (or a corresponding pro-
noun) is put as close as possible to the beginning ; then there follows the verb,
cf, e.g., Ar. Knights 1209 f. τῶι δῆτ᾽ ἂν ὑμᾶς χρησάμενος τεκμηρίωι δόξαιμι
κρίνειν... σοφῶς; Thesm. 849 τῶι δῆτ᾽ ἂν αὐτὸν προσαγαγοίμην δράματι; Andoc.
I. 104 τίνα αὐτοὺς οἴεσθε γνώμην ἕξειν περὶ σφῶν αὐτῶν ;
δυσφιλές : cf. on 34.
1233. τύὐχοιμ᾽ ἄν: cf. on 622.
ἀμφίσβαιναν : first mentioned here; the word also occurred in the ‘Storks’
of Aristophanes (Phot. Berol. p. 103. 22 = Demianczuk, Suppl. com. fr. 18,
p. 16). There is no need to consider the identifications made by modern
zoologists, ‘die eine Blindschlange, und zwar das in Griechenland haufige
Blodauge (Typhlops vermicularis Merr.) verstanden wissen wollen’ (Gossen-
Steier, RE ii A 524), since the context makes it clear that Cassandra has in
mind a terrible fabulous monster, about which, as about Scylla,’ sailors would
tell tales, and perhaps old women too, when they saw fit to scare (μορμολύτ--
reodaı) children in the dark. In the imitation of this passage (cf. Appendix
C), Cho. 994, μύραινα has taken the place of ἀμφίσβαινα. In Aristophanes the
context (cf. Photius, loc. cit.) probably showed that the creature wasimagined
as an ὄφις ὁ καὶ ἐπὶ τῆς οὐρᾶς κεφαλὴν ἔχων (Hesychius s.v. gives a fuller
definition, cf. Aelian, Nat. an. 9. 23); it is therefore called ἀμφικάρηνος by
Nicander ; so Lucan 9. 719 in geminum vergens caput amphisbaena (adapted by
Pliny, Nat. Hist. 8. 85). Lucan’s great admirer, Dante, rejoiced in the sonor-
ous name of the creature, which incidentally provided an excellent rhyme
(Inf. 24. 87).
1233. Σκύλλαν: this passage perhaps influenced E. Med. 1342 f. λέαιναν, οὐ
γυναῖκα, τῆς Tupomvidos Σκύλλης ἔχουσαν ἀγριωτέραν φύσιν.
1235. “Atdou μητέρ᾽ has been vigorously discussed for a century and a half.
The difficulty does not lie in Ἅιδου. Lobeck rightly notes on 5. Aj. 802:
‘Clytaemnestra dıdov μήτηρ dicitur pro αἰνομήτωρ vel κακομήτωρ᾽ ; he com-
pares E. Hec. 1077 Βάκχαις Ἅιδου, El. 143 Aida μέλος, Ar. Thesm. 1041 (= Eur.
fr. 122 N.) Aida γόον. Hermann rightly compares Ag. 1115 δίκτυον... Ἅιδου,
E. Cycl. 397 Aıdov μαγείρωι, wrongly A. Pers. 923 f. &ép£av . . . Ἅιδου σάκτορι
Περσᾶν (for there ‘Aiôov is an objective genitive dependent on odkropı). One
might add, e.g., E. Her. 562 Ardov τάσδε περιβολὰς κόμης. The last example is
suitable (as are others, too) to show that Ἅιδου in such combinations is a
substitute for an adjective signifying a quality. The Greeks did not derive
an adjective from the noun Ἅιδης. Instead, the poets employed the genitive,
which even in cases where a possessive adjective existed supplanted it more
and more, since ‘the genitive had gradually become the general casus ad-
nominalis’ (Wackernagel, Syntax, ii. 70). The phrase, then, was rightly
understood by Headlam among others: ‘raging, infernal, hellish mother’.?
The real crux of the passage lies in μητέρα. In relation to what person or
thing is the term ‘mother’ used in this context? While Hermann compares
passages which are by no means comparable, Headlam neglects the particular
difficulty altogether. The factor which must not be overlooked was nightly
1 Cf. A. von Salis, Sitzgsb. Heidelberger Akad., Phil.-hist. Kl., 1936/7, 1. Abhandlung, 39 f.
2 A good illustration of the underlying idea is provided by E. Phoen. 810 ἂν (the Sphinx)
6 κατὰ χθονὸς Ἅιδας Καδμείοις ἐπιπέμπει, with the scholion: ἣν ᾿Ερινύς τις 7 ᾿Αϊδωνεὺς 5
ἀλάστωρ ἐφῆκε τῆι πόλει" πάντα γὰρ τὰ δεινὰ χθόνια ἔλεγον οἱ ἀρχαῖοι.

569
line 1235 COMMENTARY
insisted on by Karsten (followed by A. Y. Campbell, C.Q. xxix, 1935, 33):
“μήτηρ tropice de re, non de persona sive femina usurpari solet’. Conse-
quently any ‘tropical’ or ‘metaphorical’ interpretation should be discarded
here where μήτηρ is applied to Clytemnestra. Are we, then, to resort again to
the device of the eighteenth century (quoted by Blomfield), ‘The Devil’s
Dam’, of which Ahrens (p. 633) dryly remarked ‘not bad as a joke’, or to its
well-mannered Victorian substitute ‘Dam of Death’ (Sidgwick)? Certainly
not, if it is to be taken as a mere façon de parler. There might, however, be a
possibility of taking the mother of Hades more seriously. An attempt in that
direction was made by Erwin Rohde, Psyche, ii, sth ed., 408 f. He, like
Karsten, rightly rejects the purely metaphorical interpretation of the whole
expression and asks: “Why, then, μητέρα of all words?’ He supposes that in
θύουσα “Aidov μήτηρ we must recognize a ‘figure of ancient Greek fairy-tales’
(‘Gestalt altgriechischer Märchen’) closely akin to Hecate; the Hades whose
mother is this terrible goddess is not to be identified with ‘the god of the
Underworld in Homer and Greek poetry generally, the brother of Zeus and
Poseidon’. This interpretation is linguistically unexceptionable and yields
an attractive thought; it deserves serious consideration. But everywhere in
Aeschylus Hades, i.e. the Ζεὺς τῶν κεκμηκότων, is a clearly defined deity of
the highest rank. I do not myself venture to set alongside that Hades this
unknown figure, possibly to be derived from obscure popular beliefs. If we
abandon Rohde’s surmise and attempt to understand μητέρα of a purely
human mother, we encounter considerable difficulties. For in this context we
do not expect a reference to Clytemnestra’s relation to her children,’ but
rather feel bound to refer φίλοις (1236) primarily, if not exclusively, to
Agamemnon. Since Wellauer this has often been stressed. This consideration
has led many critics to the conclusion that μητέρ᾽ is corrupt; thus Wilamo-
witz says in the Appendix to his translation ' μητέρ᾽, and μητέρ᾽ alone, is
incurably corrupt’. Of the conjectures proposed (e.g. μαινάδ᾽ Weil, λήιτορ'
Ahrens, udorop' Schoemann, Opusc. iv. 168, πρῆστιν A. Y. Campbell) none is
convincing. I believe that μητέρ᾽ should be retained. That becomes clear
as soon as full value is given to the connexion in which the phrase θύουσαν
Ἅιδου μητέρα stands in the sentence. As Nägelsbach well puts it: ‘A θύουσαν
separari non posse ἄσπονδόν τ᾽ ἀρὰν [this word is wrong ; see below] πνέουσαν,
et particula τὲ videtur evincere et participiorum κατὰ χιασμὸν collocatorum
necessitudo. . . . Quod autem θύουσα “Aidov μήτηρ superioribus adnexa est
ἀσυνδέτως, explicitur inde, quod a monstris ad aliud genus transitur simili-
tudinis.’ If, then, the particle re connects th& phrase ἄσπονδον. . . πνέουσαν
closely with θύουσαν Ἄιδον μητέρα, the relationship indicated by μητέρα and
that indicated by φίλοις must be apprehended together; consequently it is
to be expected that broadly speaking both expressions refer to one and the
same relationship, although it may possibly be looked at from a different
point of view in the two phrases. It follows not only that all proposed
alterations of μητέρ᾽ should be dismissed, but, positively, that we must have
here some word denoting relationship—that is, that in all probability μητέρ᾽
ı Mazon's suggestion (p. 8 of his edition) to refer the expression to what the murder of
Iphigenia has made of Clytemnestra (‘elle est une “mère en furie", qui venge son enfant’)
would be attractive did this interpretation not run clean counter to the context of the
passage.
570
 COMMENTARY line 1235
is genuine. It is pure caprice to remove re, as has recently been done by A. Y.
Campbell and G. Thomson, although Wilamowitz had once more warned the
reader against it, albeit somewhat obscurely and without drawing the full
consequences: ‘miramur dici de infenso in liberos animo Clytaemestrae, sed
quod per re adnectitur odium internecivum quod ad Agamemnonem neces-
sario refertur, persuadet, ut poetam hoc voluisse credamus.' Before proceeding
farther we must discuss shortly ἄσπονδον “Apn. Porson! saw that ἀρὰν is
corrupt; Butler and others, e.g. Lobeck on S. Aj. 802, restored "Ἄρη; Blom-
field ably elucidated the word. From the technical phrase πόλεμος ἄσπονδος
there is no way of reaching ἄσπονδος ἀρά, since wars are ended by σπονδαί but
curses are not. As is well known, a πόλεμος is often called ἄσπονδος in con-
nexion with ἀκήρυκτος ; it is a war fought without parley, truce, or conclusion
of peace until the enemy is completely annihilated or exhausted. If we take
the expression seriously, it is obvious that it is less fitted to qualify a single
act of murder, however barbarous, than the continuance of the will to destroy.
If we combine this observation with the fact which we have already noted,
that re unites the phrase ἄσπονδον "Apr πνέουσαν closely with θύουσαν Ἅιδου
μητέρα, we realize the significance of this latter phrase and the meaning of
what Cassandra indicates in veiled language indeed, yet clearly enough. The
πόλεμος ἄσπονδος which inspires Clytemnestra is directed against her φίλοι,
sui. This expression of course includes Agamemnon, whose murder domin-
ates this whole section of Cassandra's speech, but φίλοις does not mean him
alone, nor even, as the close connexion with μητέρα shows, him in the first
place. Clytemnestra is not here called ‘wife from the realm of destruction’,
but ‘mother from the realm of destruction, destroying mother’. First she
will murder her husband, θῆλυς ἄρσενος φονεύς, but then she will go on waging
this war of hers, in which there is no conclusion of peace, against her own
children ; soon her deed will drive Orestes to the murder of his mother ; then
the μητρὸς ἔγκοτοι κύνες (Cho. 924, 1054), the hellish pursuers, will seek to
hound him to death and destruction. Nor need μητέρα and φίλοις allude to
Orestes only, since afterwards Electra says (Cho. 190 1.) ἐμὴ δὲ μήτηρ, οὐδαμῶς
ἐπώνυμον φρόνημα παισὶ δύσθεον πεπαμένη. (In Euripides, Or. 195 ff., Electra
sings: ἔκανες ἔθανες, ὦ τεκομένα με μᾶτερ, ἀπὸ δ᾽ ὥλεσας πατέρα τέκνα re τάδε
σέθεν ἀφ᾽ αἵματος. This is, of course, no evidence of how Aeschylus conceived
the continuing effect of Clytemnestra’s deed, but in itself such a combination
is natural enough.) Cassandra constantly embraces in her prophecies the
more remote future along with imminent events, and soon after this (1280 f.)
she is to foretell Orestes’ murder of his mother.? Admittedly, the extension
in the range of thought is surprising in the present passage, but it continues
most forcibly what came immediately before. The murderess has just
(1233 f.) been called an inhuman, bestial monster ; now the name ‘destroying
mother’ is fastened upon her as having slaughtered the father of her children
and in so doing entered upon war to the death against those children and,
above all, her son.? More forcibly than any other name, this one banishes

! In the second Glasgow edition (1794 = Oxford 1806) a dagger is put against ἀρὰν.
2 Cf. P. Friedlander, Die Antike, 1, 1925, 20, on the Cassandra scene: "That Orestes is so
... emphatically indicated as the avenger brings in a factor which becomes important only
in the succeeding parts of the trilogy.’
3 It was not until my commentary had long been completed that I found a copy of the

571
line 1235 COMMENTARY
her from the realm of human nature; that the very being who gives birth and
nourishment should be associated with the powers of destruction points to a
frightful τέρας. The notion of the relation indicated in μητέρα is continued in
φίλοις ; in itself the word applies to the whole house, but here it is Orestes who
has the first place alike in the thoughts of the prophetess and of the audience
which knows the outcome. Clytemnestra’s εἴδωλον uses the same word to
denote her son in Eum. 100 παθοῦσα δ᾽ οὕτω δεινὰ πρὸς τῶν φιλτάτων, and
probably also in 119 φίλοις yap εἰσιν, οὐκ éuoi,! προσίκτορες.
1236. ws δέ: ‘and how she. . .’. Not a new beginning, but a continuation of
the line of thought started in 1227. Cassandra has first reminded her hearers
of Clytemnestra’s speeches (cf. on 1228-30), at which Cassandra herself and
the old men of the Chorus were present, and has exposed the true design and
schemes of the queen behind all her show ; now, to give the final confirmation
to her words, she lays bare the most treacherous single trait in the behaviour
of Clytemnestra during that scene.
ἐπωλολύξατο. What does this refer to? Older misinterpretations may be
dismissed, since Wilamowitz long ago recognized the true meaning of the
hint. In his translation he rendered 973: ‘Almighty Zeus, let my desire
triumph : take thought for that which you are making triumph this day’ and
1236: ‘how she raised a cry of triumph’, thus bringing the connexion into
prominence; in the introduction to his translation (Griech. Trag. ii. 35) and
later, in the critical apparatus on 1237 and in Interpret. 173, he expressly
explained: ‘This cry of exultation, for it is that rather than a prayer,” as is
confirmed by Cassandra’s verdict (1236) ...’. So far as the words go, Ζεῦ Ζεῦ
τέλειε κτλ. need not necessarily be an óAoAvypós; ‘c'est le ton qui fait la
musique'. Cassandra, motionless and intent, illuminated by the god, has
caught aright the undertone of jubilation ; now she is harking back to that.
We do not often enjoy the good fortune of being explicitly instructed by the
poet about the manner in which a particular passage should be recited?
modest but scholarly ‘school edition’ of the Ag. by P. Übaldi (Società Editrice Inter-
nazionale, Turin) He observes on 1235 f.: ‘con “Ac8ou μήτηρ (= madre che arreca morte,
madre funesta, cf. v. 1115 δίκτυον “Aiôov) Cass. vuol dire che Clit. & rovina anche dei figli,
coll’ ucciderne il padre. Cid che segue immediatamente, "Apr φίλοις (= ai congiunti)
πνέουσαν, appoggia questa spiegazione. I agree with every word.
! So Hermann for the MS reading éuoîs—to my mind a highly probable suggestion,
albeit not quite certain. Conjectures which tamper with προσίκτορες pronounce sentence on
themselves. Hermann has correctly explained the meaning of προσίκτορες and aptly com-
pared Ζεὺς ἀφίκτωρ at the beginning of the Supplices (i.e. the god who belongs to the
ἱκέτης and to whom the ἱκέτης turns; some more recent scholars prefer bolder interpreta-
tions); there is also a valuable discussion of these words by Williger, Sprachl. Untersuch-
ungen, 53 ff. ; he regards them, probably rightly, as new formations introduced by Aeschylus,
If Hermann's reading of Eum. 119 is accepted, the antithetic οὐκ ἐμοί in parenthesis
occupies the same position in the verse as S. Aj. 1136 ἐν τοῖς δικασταῖς, κοὐκ ἐμοί, τόδ᾽
ἐσφάλη, E. Or. 1348 ἡμῖν ydp ἤκεις, οὐχὶ σοί, σωτηρία, Verrall says that in Hermann's reading
the want of ἐμοῖς is objectionable; this 1 cannot admit, since the antithesis οὐκ ἐμοί makes
it absolutely clear of whose φίλοι she is speaking.
? Lesky, Hermes, lxvi, 1931, 201, rightly follows this interpretation.
3 In 973 f. a strong excitement is indeed expressed, but in a manner characteristic of the
Attic sense of discretion. The absence of noisy display apparently disappoints some modern
readers. So A. Y. Campbell makes the great queen break into an ἐλελεῦ ἐλελεῦ for everybody
to hear, and in his translation this is expanded to fill two whole verses, including a quota-
tion from Handel's Joshua. This found favour with Gilbert Murray, for though his transla-
tion had been free of anything of the kind, we read in his 'scenario' of the Agamemnon

572
 COMMENTARY line 1240

ὁλολύζειν (in general cf. on 594) is often used of the cry of exultation or
triumph after a victory or success; so as early as x 408; cf. Ag. 28, Cho. 386,
942, Eum. 1048, Ar. Knighis 1327, E. El. 691, Ar. Lys. 240, etc. The sense οὗ
emt- is as clear here as in Cho. 942 and, e.g., in E. Her. 348 f. αἴλινον μὲν ἐπ᾽
εὐτυχεῖ μολπᾶι Φοῖβος ἰαχεῖ. It is better to keep at a distance the thought that
the oApAvy is often raised at the beginning of the act of sacrifice proper (cf.
on 597), since it is obvious that the whole stress here falls on the complex
‘cry of exultation (to the gods) ve bene gesta’, as is shown by the addition
ὥσπερ ev μάχης τροπῆι, and above all since, apart from one blasphemous
utterance of the queen herself (1386 f.), Clytemnestra’s bloody deed is never
compared with an act of sacrifice. Blaydes and, following him, Blass,
Mélanges H. Weil, 14, have objected to the middle ἐπωλολύξατο ; both have
altered the text. But cf., e.g., Soph. fr. 491 N. (= 534 P.), 6 ἀλαλαζομένη.
1237. παντότολμος : cf. 221.
1238. δοκεῖ δὲ χαίρειν in this context must recall to the minds of the audience
the type of behaviour described in 793-8.
1239. ὁμοῖον : ‘(it is) all one’, ‘all the same’, understood aright by Casaubon
and Heath. Blomfield compared Ag. 1404, E. Suppl. 1069, Hdt. 8. 80. 2;
Dobree (Adv. ii. 326) added Alexis fr. 230. 4 K., where Dobree’s punctuation
and interpretation (‘all one’) are indeed disputed,” but are followed, probably
rightly, by Meineke and Wilamowitz (Menander, Das Schiedsgericht, p. 110).
τῶνδε... τι! cf. on 1059.
1240. An echo of 251-4 τὸ μέλλον... ἥξει.
μ᾽ ἐν isan admirable restoration : the division of xai... μήν is not admissible
(cf. Denniston, Particles, 358) in Greek of this period ;? moreover, the pronoun
is required.
παρών was altered by Ahrens and restricted by others with niggling pedantry
to a single detail (so Wecklein : ‘ παρών i.e. at the displaying of Agamemnon’s
corpse’). What we have here is merely the contrast: ‘You have now heard
from me of a future event ; soon you will yourself experience it’, i.c. the familiar
contrast which occurs in θ 491 ἢ αὐτὸς παρεὼν 1] ἄλλου ἀκούσας, B 485 f. ὑμεῖς
yap θεαί ἐστε πάρεστέ τε ἴστέ τε πάντα, ἡμεῖς δὲ κλέος olov ἀκούομεν οὐδέ τι
ἴδμεν, À. Pers. 266 παρών γε κοὐ λόγους ἄλλων κλυών. Since the point here is
the contrast with hearsay, Cassandra can assure the old men that they will

( Aeschylus, 1940, 219) the following stage direction :‘Exit Agamemnon into House. Ololugmos
from Clytemnestra and all the handmaids, which Cassandra hears (1236)’. ‘The production
would be rather like those with which Reinhardt made such an impression at the beginning
of this century’; thus, in a different context (p. 51), Murray formulates his idea of an
Aeschylean production. Campbell had a predecessor in Blass, who (Mélanges H. Weil, 15)
put an (dAoAvfe) as ἃ παρεπιγραφή after the lines 973 f. and a note at the end of the scene:
‘Quibus precibus nefariis pronuntiatis clamoreque sollemni sublato cum marito quem
trucidatura est domum ingreditur’.
1 If ὀλολυζομένη were the right reading here, as Ellendt held (so Nauck, Trag. dict. ind.
xiii; Pearson, ad loc., takes the contrary view), we should have evidence even more
conclusive for our purpose. In passing I would say that I doubt whether Sophocles wrote
βοῶσα side by side with ἀλαλαζομένη (or 0A0A,).
2 Kock’s remarks are irrelevant. Kaibel (Athen. 10. 431 Ὁ) returned to the traditional
punctuation (ὁμοῖον καὶ δίκαιον κτλ). But this can hardly be the equivalent of aequum
Tustumque est.
3 Ahrens (637) ought not to have compared Ag. 1652, where καὶ ἐγώ stands between ἀλλὰ
and μήν; see ad loc.

573
line 1240 COMMENTARY

be present at the murder of Agamemnon, by which she means the undivided

complex of it and its immediate result (the display of the king’s body) and
not the momentary act of his being struck down.
1241. Most editors follow Bothe and Blomfield in omitting y’. Blass, how-
ever, remarks on Eum. 121: ‘the γ᾽ after ἄγαν should be accepted without
hesitation from F G Tricl. (just as it should be left in the text in Ag. 1241);
cf. Ag. 1254’; but Wilamowitz’s note on Eum. 121 runs: ‘ dyav γ᾽ F Tr solita
interpolatione'.! It is well known that γ᾽ is often interpolated, even where no
point of prosody is involved (cf., e.g., R. A. Neil, The Knights of Aristophanes,
200 f.). But Neil himself (on Knights 667) expressly advocates the retention
of y’ in Ag. 1241, seeing in it, as in some other passages, the reproduction of a
turn of speech actually used by the speaker; so Denniston, Particles, 130
(‘You will say ἄγαν γ᾽ ἀληθόμαντις ’). I hesitate to question the judgement of
such authorities, but nevertheless I think it more likely that γ᾽ should be
omitted. That ye is often arbitrarily added after ἄγαν, not only by Triclinius,
but also in F, can be seen, e.g., in Pers. 520, where F alone has the wrong
reading γ᾽ (clearly because the quantity of the final syllable of ἄγαν was
mistaken ; see the footnote below).
ἀληθόμαντιν takes up ὀρθομαντείας (1215) ; the whole speech is set within the
frame of this idea, of decisive importance to Cassandra, who is struggling
with all her might for the recognition of the genuineness of her prophetic gift.
οἰκτίρας : on the tense cf. on 785.
1242. This double genitive and others of a similar type have been discussed
by Wilamowitz on E. Her. 170.
1243. ξυνῆκα καὶ πέφρικα καὶ φόβος μ᾽ ἔχει κλυόντ᾽ ἀληθῶς κτλ. : an example
of the widespread type of 'tricolon' in which the third member is expanded ;
similar, but simpler in form are, e.g., 7 Αἴας 7] ᾿Ιδομενεὺςἢ δῖος ᾿Οδυσσεύς or Ar.
Birds 851 ὁμορροθῶ, συνθέλω, συμπαραινέσας ἔχω (cf. above on 830). For the
special type in which the third member overlaps the end of the trimeter, cf.,
e.g., S. Ant. gor f. &Aovoa κἀκόσμησα κἀπιτυμβίους χοὰς ἔδωκα, El. 13 f. ἤνεγκα
κἀξέσωσα κἀξεθρεψάμην τοσόνδ᾽ ἐς ἥβης. E. Lindholm, Stilistische Studien (Lund
1931), following Behaghel, has thoroughly investigated this and kindred
forms, particularly in Latin poetry and prose, but with some consideration of
Greek. Cf. on 1412. We have here, as in the ‘tricolon’ and the pattern ‘der
wachsenden Glieder’ (Behaghel) generally, a very old and certainly pre-
rhetorical form of emphatic expression.
1244. kAvóvr' : since this clearly stands on the same plane as the contrasted
phrase (1245) ra δ᾽ ἀλλ᾽ ἀκούσας (cf. 680 τοσαῦτ᾽ ἀκούσας ἴσθι τἀληθῆ κλυών) and
must therefore be taken as aorist, Wilamowitz has very properly accented

1 One reason for the insertion of γ᾽ in Eum. 12x (and perhaps Ag. 1241, 1254) is presumably
the fact that the Byzantines scanned &yav as a pyrrhic (cf. the interpolation made by
Triclinius in Ag. 1340), as was already done (cf. Passow-Crönert and L-S s.v.) by the
epigrammatist Palladas (at the end of the fourth and the beginning of the fifth century).
In Sept. 238 Triclinius reads ἄγαν γ᾽ dreppp. for ἄγαν ὑπερφρ., in Eum. 340 ἄγαν γ᾽ ἐλ. for
ἄγαν &A., in S. Oed. R. 439 ἄγαν γ᾽ αἰνικτὰ for ἄγαν aiv. In S. Aj. 951 the true reading (cf.
Wilamowitz, Verskunst, 506) ἄγαν ὑπερβριθὲς is preserved in L, while A has a δ᾽ interpolated
after ἄγαν (for similar metrical remedies in A cf. J. Jackson, C.Q. xxxv, 1941, 166 f.), later
MSS γ᾽. In E. Ale. 809 VB have the genuine ἄγαν ἐκεῖνος, LP ἄγαν γ᾽. For the same reason
in Sept. 35 τιν᾽ has been inserted after ἄγαν by Triclinius and F2, producing a serjous metrical
blunder.

574
 COMMENTARY line 1244

the word in accordance with the principle propounded in his preface xxxiii:
cf. on 680.
ἀληθῶς οὐδὲν ἐξηικασμένα. The text has often been impugned, most for-
cibly by Ahrens, 637 ff. As sometimes happens to good grammarians he has
completely entangled himself in the meticulous establishing of a supposed
usage without paying sufficient attention to the sequence of thought as a
whole. This sequence of thought is plain here; I shall proceed from it. The
answer of the coryphaeus (1242 ff.) shows a very considerable advance, a
decisive success of Cassandra in her struggle for the recognition of the
genuineness of her seercraft and of the truth of her predictions. The very
first word (τὴν μὲν Θυέστου) goes far beyond all the admissions which the
Chorus has made, whether with reference to the visions of Cassandra in the
lyric part of the scene (1093 f., 1098 f., 1105 f.) or even after her first plain and
undisguised speech (1198 ff.). The old man is no longer admitting in general
terms and in half-evasive words that events such as those alleged by the
prophetess have happened in the royal house; his answer ‘calls a spade a
spade’, and for the first time in the trilogy we hear the name Thyestes. The
implicit contrast to the earlier attitude of the Chorus and the recognition
that now Cassandra’s words have left no room for misunderstanding find
direct expression in the phrase κλυόντ᾽ ἀληθῶς οὐδὲν ἐξηικασμένα. This is a
clear echo of the words with which Cassandra began her speech (1178 ff.)
οὐκέτ᾽ ἐκ καλυμμάτων... λαμπρὸς dé . . . οὐκέτ᾽ ἐξ αἰνιγμάτων. It is therefore
to be expected that the expression ἐξηικασμένα refers to Cassandra’s utter-
ances in the first part of the scene and points in the same direction as ἐκ
καλυμμάτων and ἐξ αἰνιγμάτων. Admittedly, ἐξηικασμένα cannot here mean ex
coniectura prolata, as it was translated from Stanley to Schütz and by some
scholars even later: nowhere does the prophetess utter conjectures, and the
old men never insinuate that she does. In the main Blomfield’s rendering is
correct : non modo levi [not a well-chosen word] similitudine expressa. On the
same lines, but still more exact, is Weil’s explanation (1861, Addenda at the
end of his edition of Eum., p. 137): ‘Cassandra dicitur res ipsas ex veritate
canere, detractis imaginum involucris, ἀληθῶς, οὐδὲν ἐξηικασμένα, i.e. ἐναργῶς
καὶ οὐ δι᾽ εἰκόνων... . Nihil mutandum.’ Ahrens maintained that ἐξεικάζειν
always means ‘make like’ and must therefore mean this here; he therefore
read κλύοντα μύθοις οὐδ. é£., which completely misses the idea required here
and is, moreover, obscure in expression. But Ahrens was also wrong as re-
gards the meaning of ἐξεικάζειν. Wherever the verb occurs (it is not common)
it is equivalent to a (slightly or considerably) stronger εἰκάζειν and so shares
the various shades of meaning of the simple verb. Thus the phrase (Ar.
Knights 230) οὐ γάρ ἐστιν ἐξηικασμένος, famous in the history of theatrical
masks, corresponds to such a phrase as εἰκόνα ἑωυτοῦ γραφῆι εἰκασμένην (Hdt.
2. 182. 1). In Aeschylus ἐξεικάζειν occurs only here and in Sept. 445," where
1 I am inclined to believe that Verrall, who was followed by Wecklein in his annotated
edition, was right in obelizing 1. 446. Despite what Wilamowitz says of it in his note, the
addition τοῖς ἡλίου looks like a miserable stopgap. Further, it is surprising to find μεσημ-
βρινοῖσι θάλπεσι repeated without alteration from 431. When I told E. Lobel that I regarded
the line as spurious, he drew my attention to the fact that in these speeches Eteocles never
repeats mechanically any of the expressions used by the scout, but always varies them, at
least by changing the word-order, cf. 511 with 493, but usually in a more elaborate manner,
cf. 400 f. with 387 ff., 404 with 387 and 391, 475 f. with 461 and 463 f., 558 θηρὸς ἐχθίστου with
579
line 1244 COMMENTARY

Groeneboom rightly remarks that it is not much more than a slightly more
sonorous εἰκάζειν (for the diminished force of ἐκ- cf. note on 1033 ἐκτολυπεύσειν
and on 1631 éfopivas). Just as εἰκάζειν, then, in the well-known parlour-
game τίνι σε μάλιστα εἰκάζω ; (e.g. Sappho fr. 127 D., Ar. Wasps 1308, Birds
804 ff., cf. below on 1629 ff.) means ‘describe by an image, a comparison’ (cf.,
e.g., Hdt. 4. 31. 2), so here ἐξεικάζειν means ‘indicate with the aid of images’.
It should, indeed, not be difficult to understand why such utterances as that
about the δίκτνον Ἅιδου or as ἄπεχε τῆς βοὸς τὸν ταῦρον are termed ‘speaking
in images’. The sense of ἀληθῶς, too, seems to be plain; cf., e.g., E. Or. 652 f.
ἀπέδοτο δ᾽, ὡς χρὴ τοῖς φίλοισι τοὺς φίλους, τὸ σῶμ᾽ ἀληθῶς ([1.6. ἔργωι ἀλλ᾽ οὐ
λόγωι ' Paley). So, also, in Hdt. x. 11. 4 ὥρα ἀναγκαίην ἀληθέως προκειμένην κτλ.
Gyges saw that (the queen had not merely alleged it, but that) the necessity
did really exist . . ., Hdt. 8. 110. 1 ἐπειδὴ γὰρ καὶ πρότερον δεδογμένος εἶναι
σοφὸς ἐφάνη ἐὼν ἀληθέως σοφός τε καὶ εὔβουλος κτλ. Everywhere the meaning
‘in reality’, ‘in truth’ predominates ; the context supplies the antithesis in each
case: ‘not deceptively’, ‘not merely in words’, ‘not merely in appearance’,
etc. In the present passage the antithesis to ἀληθῶς is expressly indicated:
οὐδὲν ἐξηικασμένα. ‘I have heard it in terms of reality, and not in figurative
language.’ The adverb is used with κλύειν with a perfect naturalness which
pedantic logic would concede only to its use with a verb of speaking. κλυόντ᾽
ἀληθῶς is made considerably easier by the fact that the opposite of ἀληθῶς is
forthwith given, as the object of the verb (οὐδὲν ἐξηικ.) ; if none the less it were
to be regarded as a bold phrase, I should reckon it among the compressed
turns of expression which are not infrequent in Aeschylus. Schneidewin’s
ἀληθῇ κοὐδὲν, which has recently become popular again, spoils the thought ;
it could never enter the head of the old man at this point to contrast true and
false predictions and in so doing to call in question the veracity of Cassandra’s
earlier prophecies; the point here is simply the difference in the manner of
her revelations, since now she has spoken οὐκέτ᾽ ἐξ αἰνιγμάτων.
1245, ἐκ δρόμου κτλ. is commonly understood as an image taken from chariot-
driving and compared with Prom. 883 ff. and Cho. 1022; on the other hand,
Conington (on Cho. 514) thinks of a bloodhound losing the trail, and compares
Ag. 1184; so Headlam (on 1251, according to his numeration). Beazley
observes: ‘Is not πεσών in favour of the second interpretation (hound)? Ina
chariot metaphor πίπτειν would be awkward, since πεσεῖν has a familiar
meaning in chariot-racing: 5. El. 747 τοῦ δὲ πίπτοντος πέδωι, A. Pers. 197
πίπτει δ᾽ ἐμὸς mais’, cf., e.g., W 467, Pind. P. s. 50, Ar. Clouds 1272. For ἐκ
539 TÓ . . . πόλεως ὄνειδος, 660 ff. with 644 ff. If 446 is obelized, τὸν πυρφόρον in 444 is, as
Verrall points out, said ‘in answer to the πυρφόρος of the shield (432)’. Nor is this all. It is
only if 446 is obelized that the sentence 444 ff. πέποιθα δ᾽ αὐτῶι κτλ. is directed against the
blazon of Capaneus (and not, like what comes before it, against his κυμαίνοντ᾽ ἔπη). And it
is precisely this which appears to be demanded by the symmetry of the whole scene; for
throughout (but for the obvious exception of Amphiaraus, whose shield is unemblazoned)
Eteocles is rebutting the pretensions of the attackers as expressed in the blazons on their
shields. I would therefore accept Verrall’s deletion and punctuate 444 f. as follows: πέποιθα
δ᾽ αὐτῶι ξὺν δίκηι τὸν πυρφόρον ἥξειν, κεραυνόν, οὐδὲν ἐξηικασμένον, i.e. ‘that his fire-bearer
(the πυρφόρος that he boastfully carries on his shield; hence the article) will come to him,
a thunderbolt (thus, a real πυρφόρος), and by no means an image of a fire-bearer (like the
πυρφόρος on the shield)’. If that is right, we must presumably suppose that 446 was fabri-
cated by someone who mistakenly desiderated a dative to go with ἐξηικασμένον ; in itself,
and especially so soon after 431, this error can easily be understood.

576
 COMMENTARY line 1252

δρόμου πεσών as said of the hound cf. [Xen.] Cyneg. 3. 5 ὅτε δὲ εἰσπίπτουσιν
(the hounds) eis αὐτά (scil. ra ἴχνη).
1246. 'Ayapépvovos . . . μόρον: an emphasizing separation (hyperbaton) of
the same type as in 267.
1246 f. A comparison with 1227-31 shows what a difference it makes that
here the name is mentioned: now the old men understand and cannot avoid
understanding what before they did not understand or could at least deny
that they understood. It is well known that the naming of a man in a sinister
context, as in this line, forthwith sets evil spirits on his track ; it is this which
makes the protest in 1247 so urgent.
εὔφημον : on the 'prolepsis' cf. Lobeck on S. Aj. 517.
1248. mawv: there can be no doubt that here, as in 99, the word means
‘healer, physician’ (cf. also 1199). That might suffice. But the fact that
Cassandra is directing her words against what the coryphaeus has just said
seems to suggest that Butler (in Peile’s commentary), Peile, and Paley were
right in recalling the ‘paean’, which could be regarded as a particular kind
of εὐφημεῖν; they refer to Sept. 267 f. κἀμῶν ἀκούσασ᾽ εὐγμάτων ἔπειτα σὺ
ὀλολυγμὸν ἱερὸν εὐμενῇ παιώνισον and Aesch. fr. 350.4 N. rdv’ ἐπηνφήμησεν.
‘The paean was sung (among other occasions) to avert threatened disaster’
(Deubner, Neue Jahrb. f. d. klass. Altert. xliii, 1919, 388). In Ar. Thesm. 310 f.
we find a fairly faithful reproduction of an actual prayer concluding with these
words: ταῦτ᾽ εὔχεσθε, καὶ ὑμῖν αὐταῖς τἀγαθά. in παιών, ἰὴ παιών, in παιών.
χαίρωμεν. ‘The cry “‘paean’’ obviously has here the same significance as the
preceding general wishes for blessings’ (Deubner, 389), ‘the paean was con-
sidered the shout of hailing in its most general form, through which one could
safeguard oneself against every conceivable ill’ (loc. cit.). The audience of
Aeschylus would certainly feel in the word παιών (the genuine Attic form for
‘paean’; cf. on 245 ff.) the force of a wish for blessings and of εὐφημεῖν ; thus
1248 is closely linked with the preceding verse in the way we expect in such
argumentative passages of stichomythia.
1249. ἀλλὰ μὴ γένοιτό πως : the old men are still desperately striving to ward
off the disaster by their wishes and prayers, albeit without any real hope.
Cf. the similar behaviour of the Chorus 255 and 997 ff.
1250. κατεύχηι. Conington and Paley translate ‘thou prayest against it’
or ‘against them’, but this rendering is not borne out by the other instances of
the verb in Aeschylus, for in Sept. 633 the implication of an adverse prayer
is due to other words in the sentence rather than to κατεύχεται. For the use
of the verb in general cf. Wilamowitz, Sappho und Simonides, 152 τι. 3.
1251. The conjecture ἄγος (Auratus, who made similar alterations of forms
of dyos in other passages of Ag.' and Cho.) ‘is perfectly unnecessary’ (Coning-
ton), for, as Wellauer says, ‘ dyos recte dicitur de eo quod dolorem excitat,
i.e. de malo’, cf. Conington on Cho. 586 and the commentaries of Wilamowitz
and of Blass on Cho. 635.
1252. It has long been pretty generally” agreed that in παρεκόπης the original
verb has been restored. Headlam illustrates it by referring to Suid. s.v.
ἀποκοπῆναι (= Anecd. Bachm. 125. 17) τῶν ἰχνῶν τὴν κύνα λέγουσιν, ὅταν
μηκέτι εὑρίσκηι τὰ ἴχνη (the same in an abridged version in Hesychius). dp’ ἂν
1 G. Thomson does not even mention that in Ag. 1579 the MS reading is ἄχη.
2 As is his habit, Verrall treats us to a piece of conservative criticism.
4872.3 H 577
line 1252 COMMENTARY

is corrupt,—‘adhuc infeliciter temptatum’, as Wilamowitz says. Murray’s

transposition, with its extraordinary order of words, ἦ κάρτα χρησμῶν παρε-
κόπης ἐμῶν ἄρα, is not justified by any of the passages which he adduces in its
support: in 5. Oed. R. 1099 (lyr.) τίς σε, τέκνον, τίς σ᾽ ἔτικτε TEV μακραιώνων ἄρα
Πανὸς . . . πελασθεῖσα; the word dpa has been challenged (Blaydes, followed
by Bruhn, wrote «opäv), but even if it is authentic it follows immediately on
the subject of the interrogative sentence (cf. Denniston, Particles, 39) and is
by no means comparable with the ἄρα which according to Murray would
occupy the last place of the sentence in our passage. Hdt. 4. 64. 3 ἦν dpa is of
course perfectly normal (cf. Denniston, 4r f.), and so is Eur. fr. 377. x N.
φεύγουσ᾽ dpa (moreover, this is not at the end of a sentence ; Murray’s ‘in fine’
is misleading) ; and equally normal is the passage which A. Y. Campbell (in
the notes to his translation) quotes to support Murray’s transposition, viz.
E. Ih. T. 1310 σὺ δὲ κατ᾽ οἶκον ἦσθ᾽ dpa (cf. what has just been said on Hdt.
4. 64. 3). These passages would be parallels if the text ran παρεκόπης dpa.
Murray might more plausibly have referred to Men. Epiir. 295 (336 Koste,
3rd ed.) dv συναρέσηι σοι τοὐμὸν ἐνθύμημ᾽ dpa, where Wilamowitz has pointed
out the unusual position of dpa (his remarks are not altogether correct).
The strong emphasis conveyed by ἦ κάρτα applies to the verb παρεκόπης;
the form of the sentence is very like that of A. Suppl. 452 7) κάρτα νείκους
τοῦδ᾽ ἐγὼ παροίχομαι, S. El. 1278 ἦ κάρτα κἂν ἄλλοισι θυμοίμην ἰδών. The close
connexion between κάρτα and παρεκόπης would be disturbed and the balance
of the sentence upset if we inserted an adverbial expression, with which κάρτα
would have to go (I once thought of κάρτα μακρὰν, but now give warning
against it ; it is also faulty because in cases of being driven from the course,
forced off the path, etc., what matters is the deviation, to left or to right,
from the proper line, whereas the amount of the deviation is irrelevant).
One might ask whether äpa, expressing realization of a fact, is at all appro-
priate side by side with the emphatic ἦ κάρτα (I have not been able to find
an instance, either in Herodotus, where κάρτα is so frequent, or elsewhere) ;
but this is perhaps no fatal objection. The sense seems to be: ‘greatly, as is
now apparent, have you failed to keep the track of my prophetic words.’
I consider Hartung’s ἦ κάρτα τἄρα as a possible emendation, though without
feeling confidence. For τἄρα = τοι ἄρα in Aeschylus? Dindorf, Lex. Aesch. 40,
adduces Cho. 112, 221, fr. 363 N. (let us hope that Aristophanes quoted
accurately).
1253. It is astonishing that Heimsoeth’s emendation can still be disregarded.
In 1251 the coryphaeus asked who would do the deed. Would he now, after
Cassandra’s taunt that he has completely failed to understand her, break off
and say: ‘because I cannot see the means whereby the doer should accomp-
lish it’ (Headlam)?3 There may be readers who do not find this kind of logic
disturbing. In that case they should at least remember the obstinacy with
which in a stichomythia a person who has asked a question will stick to his
purpose and insist on receiving an answer unless he is definitely denied it.

1 E, Petersen's conjecture (D. att. Tragodie 642) kápr' ἂν ἄκραν produces a wrong thought.
2 The error in Dindorf, p. 360, and Denniston, Particles, 555 1. 9, needs correcting.
3 On the left-hand page of his edition G. Thomson prints (with Headlam) the MS reading
τοῦ γὰρ τελοῦντος, while on the right-hand page he translates, not his own text, but the
conjecture which he does not mention: ‘But who shall do it? That escapes me still.’

578
 COMMENTARY lines 1256.

Wecklein takes τελοῦντας to be future, which is probably right.

μηχανήν : for the absence of the article cf., e.g., 1351 καὶ πρᾶγμ᾽ ἐλέγχειν.
1254. καὶ μήν: cf. on 931 f.
ἄγαν γ᾽: cf. on 1241. I am not sure whether γ᾽ should not be deleted here
also.
1255. πυθόκραντα : cf. on 369.
1256 f. These two lines are suspicious in several respects though some of the
apparent objections are perhaps of no great weight. The shortening of the
first syllable of ofov! is possibly defensible. As often in Aeschylus? and in
drama in general τοιοῦτος has a short initial syllable in this play 315, 593,
1075, 1352, 1360, 1663 ; in Sophocles the first syllable of οἷος is short, but only
(seven times, cf. Ellendt-Genthe, Lex. Soph. 520 1.) in the expression οἷός re,
which ‘quasi una vox est’ and should therefore be regarded as on the same
footing with τοιοῦτος and other trisyllables.? ὁποῖος has a short middle syllable
in an Attic oxdAıov (Athen. 15. 694 d), ποῖος a short first syllable in Ar. Wasps
1369,* but in Tragedy only in A. Suppl. 911,5 where Wilamowitz remarks
‘correpta dipthongus in bisyllabo rudioris artis signum’. Our material is
perhaps too scanty to justify us in regarding this prosodical rarity as an
indication of early technique. But supposing that Wilamowitz is right, the
scansion of οἷον in Ag. 1256 would provide an argument against the line having
been written by Aeschylus as it stands in the MSS. It is true that Homer
uses οἷος with a short first syllable: N 275, Σ 105, n 312, v 89, but it is scarcely
likely that the prosody of a very common word in the dialogue verse of
drama was influenced by the practice of epic. For a linguistic explanation
of this prosodical feature see Schwyzer, Griech. Gramm. i. 236.
It is harder to reconcile ourselves to the hiatus. In his treatment of
hiatus in trimeters (Kühner-Blass, i. 196) Blass rightly professes his doubts
about παπαῖ οἷον and points out the extraordinary character of the hiatus
ἐγὼ eya.° Extraordinary also is the inclusion in the trimeter, not indeed of
παπαῖ (this occurs in Sophocles and in comedy), but of ὁτοτοῖ. In Aeschylus
and Euripides, in whose works óroro? is by no means rare, as well as in
Sophocles (El. 1245) this is not found except in lyric metres.
1 Attention is drawn to it in F, where κοινὴ (scil. συλλαβή) is written above οἷον just as it
is at 315, 593, 1075, 1352, 1360 over the τοι of forms of τοιοῦτος. Cf., e.g., Schol. M on E. Her.
533; for the phenomenon itself and for the terminology see Hephaestion pp. 3 ff. Consbr.
2 Cf. fr. 155 N., quoted by Hephaestion 3. 14 Consbr. as an instance of τοιαύτας with the
first syllable short.
3 Cf. on this subject R. Sjölund, Meirische Kürzung im Griechischen (Uppsala 1938), 36;
unfortunately it is only in an arbitrary selection that he takes into account the instances
found in drama.
4 There is probably nothing wrong here with the ‘split anapaest’ κλέψαντα ποίαν, although
it seems to be made harsher by the fact that the two words are spoken by two different
persons; cf, Birds 1226 εἰ τῶν μὲν ἄλλων ἄρχομεν, ὑμεῖς δ᾽ οἱ θεοί, where we find the split
anapaest despite the punctuation ; see Maas, Griech. Metrik, § 111. Wilamowitz is too bold
in his note on Ar. Lys. 193: ‘Wasps 1369 runs ποίαν αὐλητρίδα, but ποίαν must be pronounced
as a monosyllable’; he expressed himself more cautiously earlier, Berl. Sitzgsb. 1911, 514
(= Kl. Schr. i. 333) and in his commentary on Menander’s Epitrep. p. 71 n. 1.
5 The passages adduced by Boeckh, Pindar, i. 2, p. 492, for ποῖος with the first syllable
short, E. Or. 155 (= 154), where he accepted a conjecture of Triclinius, and Phoen. 885
(= 878), are to be discarded.
6 J. Descroix, Le Trimetre iambique (1931), p. 26 f. (survey of hiatus), does not mention
Ag. 1257. .

579
 lines 1256 f. COMMENTARY
Moreover, there is the lame, not to say childish, continuation of οἷον τὸ
πῦρ by ἐπέρχεται δέ μοι. Editors have long tried to remove this awkwardness
either by punctuating the sentence in an artificial manner? or by altering the
text,? but without success. I consider it possible that Wilamowitz in his
early article (Hermes, xviii, 1883, 246 n. 2) propounded the right solution for
1256 f., though his treatment of Ag. 1214 ff. was not happy. If we assume
with him that δέ μοι and the second ἐγώ are stopgaps botched up by someone
who wanted complete trimeters,* we obtain the following result. First, the
interjections now have their appropriate place extra metrum ; secondly, οἷον
retains its normal prosody ; thirdly, the extraordinary hiatus disappears ; but
above all, the ugliness of the sentence (or sentences) in 1256 is replaced by the
fine and powerful οἷον τὸ πῦρ ἐπέρχεται. The structure of the sentence, first
παπαῖ, then ofov . . . ἐπέρχεται, falls in line with the type exemplified by
Wilamowitz in his commentary on A. Cho. 691 of ἐγώ, Kar’ ἄκρας... ὡς
πορθούμεθα; to the instances quoted by him (with Ag. 1256 cf. especially
E. Her. 1120 παπαῖ, τόδ᾽ ὡς ὕποπτον ἠινίξω πάλιν) we may add, e.g., Cho. 743 ff.
ὦ τάλαιν᾽ ἐγώ, ὥς pow... ἤλγυνεν, E. Or. 1020 f. of ἐγὼ μάλ᾽ αὖθις, ὥς σ᾽ ἰδοῦσα
! An examination of the examples of the use of δέ among which this line is included in
Denniston, Particles, 172, shows that not one of them is even distantly akin to it.
2 To say nothing of earlier attempts (Blomfield, etc.), Plüss assumes a break after ἐπέρ-
χεται δ᾽ ἐμοί [sic!]: ‘the subject comes late, the vision only gradually becomes distinct’; he
takes αὕτη δίπους λέαινα κτλ. to be the subject. That is, of course, impossible, but it shows how
discontented he was with the traditional arrangement of the text.
3 That Casaubon objected to δέ μοι is far more important than that he suggested the
improbable δέμας.
4 I will give an instance of the expansion of an interjection to a trimeter because this
instance is particularly instructive. In E. Or. 478 Wecklein rightly retained in his text only
ἔα of the verse £a: τὸ μέλλον ὡς κακὸν τὸ μὴ εἰδέναι. éa is ἐκπλήξεως ἐπίρρημα (Schol. A.
Prom. 114). Without exception in Euripides it expresses the surprise of the speaker at
some novel, often unwelcome, impression on his senses, the visual sense or another. For
ἔα in passages where a quasi-monologue is broken off see F. Leo, Monolog im Drama, 30, 67.
The rule applies to single as well as double ἔα, and to the instances in the Rhesus as well as
in the plays of Euripides. Cf. Aesch. Prom. 114, 298 (for ἔα, ri χρῆμα; see Wilamowitz on
E. Her. 524), Cho. 870, Δικτυουλκοί (Pap. Soc. It. 1209 fr. a, 8 = D. L. Page, Greek Lit. Pap.
i, p. 10) ἔα' τί φῶ τόδ' εἶναι (Prom. 687 alone is anomalous) ; 5. Oed. C. 1477 (the only instance
in Sophocles; the supplement of Mekler, which Pearson puts in his text in fr. 314. 39, is
probably wrong); Ar. Clouds 1260, Peace 60, Birds 1495, Thesm. 699, 1009, 1105 (= Eur.
fr. 125), Plut. 824 (the only instance in Aristophanes which does not fall under the rule is
Birds 327, where &a has exactly the same function at the beginning of the chorus as in
Prom. 687). Cf. also Plato, Prot. 314 d. This use clearly comes from the language of every-
day life. In Euripides the exceptions are Jon 540 (but here éa has long been emended to ἐκ),
Iph. A. 644 (this cannot with any degree of naturalness be brought into accord with
Euripidean practice) and 1132, where éa is unmistakably (as probably in 644) simply a cry
of distress. It is a stroke of luck that immediately after this uneuripidean éa we have
1133 ἔχ᾽ ἥσυχος, the only monometer of the type (cf. my note on Ag. 1216) in dialogue in
Euripides, ‘for this is not Euripides the son of Mnesarchus’, to quote Wilamowitz, Hermes
xviii, 1883, 247 n. 2. The same, I believe, is true of the ἔα of 644 and 1132; the imitator has
failed to follow the consistent practice of Euripides. But, to return to the starting-point
of these remarks, in E. Or. 478 éa is completely normal: the words which follow, 9 μητρο-
φόντης ὅδε «rA., show that the exclamation is provoked by the unexpected sight of Orestes.
A character in a Euripidean play has no time and no inducement to insert between éa and
its explanation a γνώμη such as τὸ μέλλον ὡς κακὸν τὸ μὴ εἰδέναι (there is no need to dilate
on the shallowness of τὸ μέλλον when all it comes to is ‘I did not think that I should find him
here’). The interpolator understood ξα as a cry of sorrow (see above); he constructed the
complete trimeter, which seemed to him an improvement, with the help of a commonplace
maxim (cf. on 1226).
580
 COMMENTARY line 1258
u... ἐξέστην. Cf. also Homer x 38 ὦ πόποι, ὡς ὅδε... τίμιός ἐστιν, m 364 ὦ πόποι,
ὡς τόνδ᾽ ἄνδρα... ἔλυσαν, σ 26. That the dimeters, first acatalectic, then
(Λύκει᾽ "Ἄπολλον, οἵ ’yw) catalectic, are here unexceptionable has already
been shown (on 1216). But despite all these advantages Wilamowitz’s
suggestion should not be regarded as the only possible way of emending
the passage.
1256. τὸ wip: she feels the approach of a new access of possession by the god
as the coming of a fire which will forthwith take hold upon her; cf. S. ET. 887 f.
ἐς τί μοι βλέψασα θάλπηι τῶιδ᾽ ἀνηκέστωι πυρί ; (with the scholion, παρακόπτεις
τὰς φρένας ὡς ἐν πυρετῶι). The Fury who drives Queen Amata mad does,
indeed, employ a particular instrument (the snake turned into a golden
necklace and a iaenia), but the effect is typical of the onset of madness:
Virgil Aen. 7. 354 ff. ac dum prima lues udo sublapsa veneno pertemptat sensus
alque ossibus implicat ignem necdum animus toto percepit pectore flammam eqs.
On ecstasy as the permeation of the possessed person by a fire-spirit see O.
Gruppe, Griech. Mythologie 849 (some of his instances are inconclusive).
1257. Λύκει᾽ “AroAAov. We cannot say with certainty why this particular
ἐπίκλησις is chosen (Schneidewin says "because of the designation of Aegisthus
as λύκος ’)" or what associations it would awaken in the minds of the hearers.
In Suppl. 685 the Δύκειος of Argos? is clearly invoked ‘because Apollo was
under that title the chief god of that city’ (Wilamowitz, Interpr. 40). The
reference is also clear in Sept. 145 “Δύκει᾽ ἄναξ λύκειος γενοῦ στρατῶι dalwı: ‘you
are called after the wolf, show yourself such against the enemy' (Wilamowitz,
Glaube d. Hell. i. x47 n. 2). As for Ag. 1257, it is worth considering the surmise
of R. P. Eckels, Greek Wolf-Lore (Philadelphia 1937), 61, that this is one of
those tragic passages in which ‘Apollo “ύκειος was invoked, not always
specifically against wolves, but more generally against a wide variety of
nuisances, misfortunes and calamities'; Eckels adduces also S. Oed. R. 203,
919, El. 645, 655, 1379. On the original meaning of “ύκειος much has been
written ; there is no need to discuss it here.
1258ff. Here again we have a vision at the beginning, exactly as at the
beginning of Cassandra's previous speech (1217 ff). 1258 αὕτη corresponds to
the τούσδε of 1217.
1258. In Lycophron 1107 Cassandra says ‘Agamemnon’s spirit will fly to the
underworld λυγρὰν λεαίνης εἰσιδοῦσ᾽ oikoupiay’.
δίπους : the context of the word is here ἃ γρῖφος, and it was presumably
used elsewhere in riddling and oracular language. The riddle of the sphinx
(cf. on 80) begins ἔστι δίπουν ἐπὶ γῆς κτλ. The form of this riddle containing
the word δίπουν is at least as old as the time of Aeschylus, since the famous
cup in the Vatican (Hartwig, Meisterschalen, plate 73; Beazley, Att. Red-
Fig. Vase-Painters, 296 no. ız) which shows the sphinx addressing Oedipus
and beside her the words [xJai rpi[zov} (from these words it follows that at
that time δίπουν too must have had its place in the riddle) belongs to some-
where about 470 B.c. (cf. C. Robert, Oidipus, i. 51, 56; I owe to Beazley the
correction of the date given by Robert). A. Suppl. 895 δίπους ὄφις is, like
! So also Welcker, Griech. Gótterlehre, i. 479: ‘Cassandra invokes Lykeios because in her
mind’s eye she sees Aegisthus as a wolf.’
2 Farnell, C.Q. iv, 1910, 181, assumes the same reference for Ag. 1257. To me it seems most
unlikely as spoken by Cassandra.
581
line 1258 COMMENTARY
Ag. 1258, a piece of abuse couched as a γρῖφος. Plat. Politicus 276 c τῆς
δίποδος ἀγέλης is possibiy an echo of oracular or similar poetry.'
1261. The text of the παράδοσις was ἐνθήσει. When Triclinius on second
thoughts wrote ἐνθήσειν, he obviously was endeavouring to secure a smoother
sentence-connexion by getting rid of the asyndeton at the beginning of 1262?
and drawing the whole passage from ὡς δὲ φάρμακον to 1263 φόνον into a single
sentence. Hermann (whose text, apart from κότωι, still stands in Murray's
edition) ought not to have allowed himself to be taken in by this easy device
for smoothing the expression, the less so as he aptly remarks on 1263:
‘explicatur hoc versu illud κἀμοῦ μισθὸν ἐνθήσειν körwı’. Nägelsbach drew the
right conclusion: ‘Quod hoc versu (1260 f. ὡς δὲ φάρμακον κτλ.) per imaginem
elocutus est poeta, id in sequenti versu planis ac propriis illustrat verbis.
Inde natum est asyndeton explicativum ἐπεύχεται KrA.' ; so Wilamowitz on
1262. Cf. on 36 and 1284.
ποτῶι (not mörwı): the defenders of κότωι, from Pauw, Schütz, Blomfield
on, take the matter too lightly. So, e.g., Hermann: ‘mei quoque mercedem se
irae admixturam gloriatur.’ admixturam? I can find no evidence that
ἐντίθημι with the dative ever means anything but ‘put into something’.
Others, while rejecting ποτῶι, tacitly make it the basis of their translation ;
so Blomfield ‘irae potioni admiscebit’, Headlam ‘she will make my wages an
ingredient in the rancorous brew’. If we read ποτῶι, there is no need to
distort the meaning of ἐντιθέναι, and at the same time the image begun with
ὡς δὲ φάρμακον becomes fully distinct :? as has long been seen, it corresponds
with Cassandra’s (1137) ἐπεγχύδαν. It ought not to be objected that the
alteration of κότωι entails the loss of something of value; in this section, the
theme of which is clearly indicated in 1260 with κτενεῖ ue τὴν τάλαιναν, Cas-
sandra is no longer concerned, as in her previous speech (1215-41), with
Clytemnestra’s disposition to herself, but is speaking of the connexion of her
own fate with that of Agamemnon. That is as sharply separated from what
goes before as was the second part of the lyric scene (1136 ff.) from its begin-
ning. Just as the speech 1258 ff. corresponds in general with that lyric passage,
so in detail does ἐνθήσει worms here with ἐπεγχύδαν there. I am therefore
inclined to accept words. If we retain κότωι we must assume with Schneidewin
that here, too (cf. on 966 and 1178-83), ‘image and object compared are fused
into one’. I do not think this impossible; in such cases there is always the
danger that our desire for consistent distinctness goes beyond the intention
of the ancient poet.
μισθόν: as payment, compensation (for more detailed discussion see on
1262 f.). Above μισθὸν is written in F the puzzling gloss μνείαν ; there is no
doubt about the reading ; Hermann, who said the gloss was un”, was misled
by the collation he used. Hermann added ‘quod nescio an μνῆσιν legi debeat’,
This, a non-existent word, was (in the place of μισθόν) put in the text by
Ortmann and, changed to μνῆστιν, by Lawson and A. Y. Campbell. It need
! On τρίπους in the context of a γρῖφος in Hesiod cf. above on 149.
2 A, D. Knox, Philol. Ixxxvii, 1932, 34 n. 23, disliked the asyndeton and still more the
comparatively rare but perfectly legitimate verse-ending of the type Zyvi πιστὸν ἄγγελον;
so he attempted to kill two birds with one stone. The result is horrifying.
3 Verrall’s allegation ‘ word (Auratus) leaves φάρμακον obscure’ is the opposite of the
truth. It is another question whether it is right to insist on the details of the image being
consistently worked out ; see below.
582
 COMMENTARY line 1263
not be taken into consideration, since μισθόν is sufficiently confirmed by
ἀντιτείσεσθαι (see below).
1262. On the asyndeton see on 1261. ἐπεύχεται: Stanley rightly
renders ‘gloriatur’. This meaning (as early as Homer), and not the other,
‘pray’ or ‘vow’, is demanded by the context, cf. 1394. That the aorist
infinitive? dependent on the verb used with this meaning has a past sense is
clear from K 368 μή ris . . . φθαίη ἐπευξάμενος βαλέειν (cf. Eustathius ad loc.),
hymn. Ven. 286 f. εἰ δέ κεν ἐξείπηις καὶ émeí£eac . . . ἐν φιλότητι μιγῆναι...
Kvfepeint, Plat. Laws 905 a μή more . . . ἐπεύξηται περιγενέσθαι θεῶν. It is
therefore safer to accept the easy emendation ἀντιτείσεσθαι than to classify
the passage with the doubtful cases (cf. Kühner-Gerth, i. 196 1.) where an
aorist infinitive dependent on a verb of speaking (stating) is supposed to
denote a future action.
φάσγανον: on the subject-matter see Appendix B. The word occurs here
alone in Aeschylus (Cho. 647 $acyavovpyós) ; it is the word used in A 424 for
the weapon with which Agamemnon was killed.
1263. φόνον: the cognate accusative as 5. Aj. 304 ὅσην κατ᾽ αὐτῶν ὕβριν
ἐκτείσαιτ᾽ ἰών (Wecklein). |
That which Clytemnestra will put into the draught of her revenge is termed
by Cassandra ἐμοῦ μισθόν. “These words are in themselves ambiguous, and
might mean either “pay for me” (to another), or “retaliation on me" (person-
ally)’ says Paley ; he decides for the translation ‘a requital for me’. In reality
the ambiguity, or rather the twofold meaning, of which Paley speaks must
not be interpreted away but recognized. Since κἀμοῦ μισθὸν ἐνθήσει follows
κτενεῖ με τὴν τάλαιναν it seems to apply primarily to the murder of Cassandra ;?
here, then, at the outset we must understand καὶ ἐμοῦ thus: ‘I, too, Cassandra,
shall have to pay the price (with my life)’. In the explanatory sentence
(ἐπεύχεται κτλ.) the idea evoked by ἐμοῦ μισθόν is unmistakably taken up and
developed by ἐμῆς ἀγωγῆς ἀντιτείσεσθαι φόνον (cf., e.g., E. [ph. A. 1169 κακῆς
γυναικὸς μισθὸν ἀποτεῖσαι τέκνα), but now it is Agamemnon who must pay and
atone for having brought Cassandra to Argos. It would be wrong to take
exception to this extension of the thought. Just as later on Clytemnestra
herself returns time and again to the point that by her deed she merely
secured the satisfaction of her claims to justice (cf. on 1378), so here does she
appear to the prophetess as the woman whose whole endeavour is directed
to exacting the payment of the price, restitution, retribution. Both of her
victims, Cassandra and Agamemnon, owe her compensation, but above all
Agamemnon. Here, as in ı137 and 1325 f., Cassandra holds fast to the idea
that hers is only a secondary part in the whole fearful story. It is in the light
of this idea that Clytemnestra’s deed is seen here: she will carry out the
murder of Cassandra as an addition to the murder of her husband. If in this
context Clytemnestra is said to declare that the death of Agamemnon is
retribution for bringing Cassandra home with him, no one can be deceived
about the real motives of the queen. [See the Addenda.]
1 On the Homeric use of εὔχομαι with the aorist infinitive although the object of the
εὔχεσθαι is a future event, see Latte, Hermes, lxvi, 1931, 129 n. 1. He indicates the limits of
the use and further points out that it is confined to cases in which εὔχομαι means ‘pray’.
2 Miss Alford observes: ‘I doubt this. I think that in ὡς δὲ φάρμακον the idea of mixing a
cup for Agamemnon cornes in and ἐμοῦ μισθόν is primarily payment exacted from Ag. or
punishment inflicted on him.'
583
line 1264 COMMENTARY

1264. καταγέλωτα : this form of the accusative occurs also in Menander fr.
245. 7 K., but karayeAwv in Ar. Ach. 76, Knights 319.
ἔχω: ‘keep, wear’, as so often.
1264 f. Almost all editors and translators put a comma after τάδε and take
σκῆπτρα and μαντεῖα στέφη asin apposition to τάδε. But after the preparatory
τάδε the copulation by καί... καί... seems to me intolerable in itself, and
particularly so in the context of this passionate speech. Among the instances
of kat... kat. . adduced by Dindorf, Lex. Aesch. 170 ὃ 5, I can find none even
remotely comparable. The usual manner of taking the passage is not the
only possible one. τάδε (without a comma after it) might be regarded as the
first of three terms.? This was the view of Triclinius (or a commentator whom
he followed), for he wrote τὰ ἐσθήματα as a gloss on τάδε. The correctness of
this interpretation does not necessarily follow from 1269 f., but if we accept
it, the two passages would agree excellently with each other; first Cassandra
tears parts of her garments and throws the rags to the ground along with her
staff and fillets, then she says that in this action of hers the god himself is
operative, just as throughout she feels herself his instrument and sacrificial
victim : τάδε would not be ambiguous to the spectators, who at the same time
as they heard the word saw the gesture which accompanied it. This pre-
supposes what is in itself natural enough, that not only the staff and fillets,
but also the dress worn by Cassandra belonged to the sacred gear of the
prophetess.? The nature of Cassandra's costume was recognized by Casaubon
(on Theophrastus, Char. 16. 2): 'Apud Aeschylum in Agamemnone habitus
vatis hic est, ut amictus veste vatum propria, quam poeta appellat χρηστηρίαν,
manu scipionem gestet, et coronam collo appensam.’
1265. σκῆπτρα: the commentators have compared Homer A 14; Casaubon
(see above) quoted Hesychius s.v. ἐθυντήριον : ὃ φέρουσιν oi μάντεις σκῆπτρον
ἀπὸ δάφνης.
στέφη: not ‘wreath’, since that is not worn round the neck, but the sacred
bands, ταινίαι (taken aright by, e.g., Verrall, in Headlam’s prose translation—
‘fillets —and by Wilamowitz).
1266. σὲ μέν: referred by the Byzantines (see above) to her robes; so Her-
1 We might regard as some sort of parallel the passage Cho. 670, where the general state-
ment πάρεστι yap ómoiámep δόμοισι τοῖσδ᾽ ἐπεικότα is subsequently particularized: καὶ
θερμὰ λουτρὰ καὶ πόνων θελκτηρία στρωμνή, but the sentence is continued : δικαίων 7’ ὀμμάτων
παρουσία. Thus we have there a kind of polysyndeton, the implication of which is something
like ‘all conceivable conveniences with which a house such as ours is accustomed to provide
its guests’. Furthermore, it should be remembered what a gulf there is between Clytem-
nestra’s well-constructed speech and Cassandra’s wild outburst.
2 J find that in recommending this I agree with Headlam’s prose translation (‘these robes
and wands and fillets of divination round my neck’), and in the main with A. Y. Campbell,
though he quite unnecessarily reads στολήν instead of τάδε, in accordance with the marginal
gloss in F on 1266 (σὲ μὲν) : ὦ στολή (Vitelli or Wecklein, p. 399, ought not to have omitted
this gloss, especially as Hermann had already mentioned it; it is quite clear in the photo-
graph). In Tr we read a scholion on 1266: πρὸς τὴν ἑαυτῆς ἐσθῆτα τοῦτο λέγει, σχίζουσα τὰ
ἑαυτῆς ἱμάτια; thus it is possible that the sense which Triclinius gives to 1264 τάδε is the
result of an interpretation of 1266 which he found in his exemplar.
3 Wieseler (cf. Wecklein on 774 ff. Weckl.) and J. Sommerbrodt, De Aeschyli re scenica,
Pars II (Jahresbericht ub. d. Kgl. Ritter-Akademie zu Liegnitz, 1851), p. Ixviii n. 9, made
the attractive suggestion that Cassandra's garment was covered by the dypqvóv: τὸ δ᾽ ἣν
πλέγμα ἐξ ἐρίων δικτυῶδες περὶ πᾶν τὸ σῶμα, ὃ Τειρεσίας ἐπεβάλλετο 1j τις ἄλλος μάντις (Pollux
4. 116), cf. Daremberg-Saglio, i. 165.

584
 COMMENTARY line 1268
mann; on the other hand, many modern commentators (from Bothe to
Wilamowitz) take it as referring to her prophet’s staff. For us a decision is
scarcely possible ; the spectators saw what was intended. The conjecture of
Munro, J. Phil. xi, 1882, 140, ‘an image of Apollo, I guess, which she wore on
her head or breast’, is quite improbable.
1267. ir’ ἐς φθόρον: the phrase is well illustrated in Blomfield’s Glossarium
on Sept. 252 (238 Blomf.); cf. also Kock on Ar. Clouds 789, and add Men.
Peric. 276. Closest to both the Aeschylean passages comes the fragment of
an Attic comedian quoted as coming from Epicharmus in Athen. 2. 63 c,-
where we find day’ és τὸν φθόρον. The expression comes dangerously near
‘go to hell’. The coarse turn of phrase accords admirably with the brusque
soldierly speech of Eteocles (Sept. 252). The furious anger of Oedipus (Oed. R.
430) is characterized by the unkingly οὐκ εἰς ὄλεθρον; from the lips of the
servant driven to despair (ib. 1146) the words are much less striking. In
E. Andr. 708 the enraged Peleus snarls at Menelaus: εἰ μὴ φθερῆι τῆσδ᾽ ὡς
τάχιστ᾽ ἀπὸ στέγης krA., in E. Her. 1290 οὐ γῆς τῆσδ᾽ ἀποφθαρήσεται ; is ‘a
vulgar .. . expression which Herakles deliberately chooses to describe people's
carping speech’ (Wilamowitz, ad loc.). The flagrant breach of the laws of
propriety is characteristic of Cassandra’s behaviour during this scene; cf. on
1228. It is not, however, here, as it is with other tragic characters, merely
that an overpowering passion makes her disregard the dictates of propriety ;
Cassandra’s whole being is already far beyond the bounds within which
manners and good form could have validity or meaning; like Antigone she
might say ἡ δ᾽ ἐμὴ ψυχὴ πάλαι τέθνηκεν.
From πεσόντ᾽ ἀγαθὼ Jacob (whose punctuation of the line was improved
upon by Blomfield) elicited πεσόντα γ᾽ ὧδ᾽ ; Verrall read πεσόντα θ᾽ ὧδ᾽ : these
two readings may be regarded as variae lectiones, the contamination of which
produced the text of our MSS, as Wilamowitz saw. It seems inconceivable
that after the passionate imperative ἔτ᾽ es φθόρον what follows could be
connected by re; on the other hand, ye, as understood by Wilamowitz, is
excellent: ‘si non antea, nunc certe cum humi iacetis, sic debita vobis reddo
mala’, ἀμείβεσθαι ‘to repay, requite’ is thoroughly appropriate here. The
present ἀμείβομαι, which was originally written in F, was possibly due to a
mere slip of the copyist, and the exemplar of F may have had ἀμείψομαι (so
G Tr)! but however that may be, we should put ἀμείβομαι in the text:
Cassandra performs the action while she is speaking (‘vittas proiectas pedibus
calcat’ Wilamowitz). The corruption to ἀμείψομαι is to be explained as
assimilation to διαφθερῶ (Verrall). In exactly the same way in 26 σημαίνω,
which M alone preserves, was corrupted to σημανῶ under the influence of
χορεύσομαι and θήσομαι (31 f.).
Long ago Canter noticed that the lines E. Tro. 256 ff. and 451 ff. were
inspired by 1264-7. It is worth observing how Euripides' version softens
down the original (e.g. 453 ἔτ᾽ dz’ ἐμοῦ χρωτὸς σπαραγμοῖς takes the place of
ἔτ᾽ és φθόρον) and, while elaborating it, makes it more sentimental. Canter
also compared Statius, Theb. 7. 784.
1268. The ἄτην of the MSS can only be defended by assuming some linguistic
monstrosity ({ ἄτην πλουτίζετε enrich Destruction, i.e. "be destroyed" ',
1 The note in Wilamowitz's app. crit. implies that the reading of G was originally
ἀμείβομαι, but in point of fact it was never anything else than ἀμείψομαι.

585
line 1268 COMMENTARY

Verrall ; the idea which he finds in the verse is worse than obscure) or else by
understanding the expression as meaning that Cassandra terms herself and
any successor she may have an drn. Naeke, Opusc. i. 173, interpreted it thus,
adducing ‘parallels’; the idea has recently been revived by Murray, who is
not impressed by Hermann’s decisive refutation. The correction ἄτης is self-
evident. On the construction (πλουτίζειν cum genet.) Hermann rightly
observes that the common πλουτεῖν cum genet. provides a satisfactory
analogy.
1269 f. ἰδοὺ... exöuwv . . . ἐσθῆτα. The participle stands by itself, without
a finite verb, ‘because the prophetess fancies the actual presence of the god
before her, in the act of stripping her of her attire ;—‘‘See, here is Apollo
himself stripping me, etc.” ’ (Paley).
1269. adrds . . . ἐσθῆτα: ‘he himself’; it is almost impossible not to under-
stand by this ‘the very being who once clothed me in them is now stripping
me of them’, though αὐτός here is not repeated and thus lacks the fearful
force it has in the famous complaint of Thetis, likewise directed against
Apollo (fr. 350 N.).
1270. At first sight the clause ἐποπτεύσας δέ με κτλ. appears to stand either
in simple apposition to the clause ἐκδύων ἐμέ κτλ. (‘See! Apollo himself is
stripping me—he too being the one that marked me’, Conington) or in a
temporal relation (‘and that too after he had coldly looked on’ etc., Paley) ;
a Greek was hardly in a position to draw a sharp line between the two modes
of subordination. But in either case δέ would seem to be highly objection-
able." Many commentators pass over the point in silence. Conington attempts
a justification : ‘ δέ stating an additional fact committed by the same person
at a different time, and so charging Apollo with both insults’; but this does
not do justice to the nature of this participial construction (subordinating or
coordinating, but not continuing). Triclinius took exception to it; to get rid
of it he conjectured ἐπώπτευσας (which he glossed with εἶδες) and this has
been adopted by many editors, despite the fact that in 1269 and 1275 f.
Apollo is spoken of in the third person. Others altered δέ; so, e.g., Auratus
(γέ με), Schmitt (more). Enger and Halm deleted the particle and read ἐμέ,
which was adopted by Wecklein (annotated edition) and Weil (Teubner text).
Hermann was on the right track in that he realized that émorr. κτλ. is not a
parallel addition to ἐκδύων, but the beginning of a new statement: he writes
‘illa verba ἐποπτεύσας δέ με, aut verbum aliquod, aut nomen, quod pro verbo
esset, requirebant’. His conjecture (1272 ματήρ pro μάτην) is, however,
impossible. Dindorf apparently held the same view on the structure of the
sentence and therefore assumed a lacuna after 1270; this is quite unwarranted.
As a matter of fact, we have the required continuation of ἐποπτεύσας δέ pe,
provided that our text is rightly punctuated, which entails first removing the
stop after 1272 μάτην (this stop goes back to F and Tr), and secondly recog-
nizing the parenthetical character of 1273-4, and the resumption of the main
sentence in 1275 καὶ νῦν κτλ. This analysis was in the main? achieved by
Heimsoeth, Die Wiederherstellung der Dramen des Aesch. 92, who was followed
by Keck; the right punctuation was then given, possibly independently of
1 Cf, Denniston, Particles, 164 under 3, on the two Euripidean passages in which he, along
with earlier editors, opposes the retention of δέ,
2 He spoilt his recognition of the truth by changing the δέ after ἐποπτεύσας to τε.
586
 COMMENTARY line 1271
their German predecessors, in the editions of Verrall, Sidgwick, and Headlam.*
Thus we have recovered a sentence-structure which is of striking vivacity
and, moreover, characteristically Aeschylean. ‘He witnessed all the scorn I
encountered in my prophet’s garb, and now, to crown the wrong, he has...’
Some such statement as this is what Cassandra would make in a calm
narrative. But she is not calm and she does not give a narrative; from her
troubled soul burst forth bit by bit complaints and recriminations, and when,
believing that she has everywhere been scorned, she thinks of the way in
which her nearest relatives have treated her, she cannot immediately con-
tinue what she was saying, because all the humiliations and injurious words
which she has endured rise up again in her mind, and she cannot but linger
over them and repeat to her own torment the hateful taunts. The intensity
of her angry recollections finds its expression in an independent sentence
(1273 1), with the result that the sentence beginning with éronr. δέ pe
remains unfinished and she has afterwards (1275) to start afresh. It is one of
the most familiar phenomena of Aeschylus’ style that the fullness of content
should thus overflow the containing vessel of the sentence-structure: cf. on
12, on 205 and elsewhere.
ἐποπτεύσας. In the Odyssey, v 140, the word is used of a man overseeing the
work of his subordinates (from this passage Hesiod, Erga 767 takes over the
expression ἔργα 7’ ἐποπτεύειν). But Aeschylus, as Schütz remarked on Cho.
583, uses ἐποπτεύειν exclusively (9 times) of gods and godlike beings (Cho. 489
of the hero who is to rise from the grave and look benignantly on the struggle
of his children ; so probably 583, cf. p. 491 n. ı above). In addition there is the
contemporary evidence of Pind. Ol. 7. 11 ἄλλοτε δ᾽ ἄλλον ἐποπτεύει Χάρις
ζωθάλμιος κτλ. In Eum. 220 we find in accordance with the character of the
wrathful daemons to whom Apollo is speaking the remarkable phrase émo-
πτεύειν κότωι, but elsewhere the verb, as one would expect, usually denotes the
benevolent, helpful look of the deity who turns ὄμματι μὴ Aofde to the man
imploring the protection of heaven. This use of ἐποπτεύειν agrees with that of
ἐφορᾶν, ἐφορεύειν, ἐπισκοπεῖν (see above on 14 and on 461 f.) and with the use
of ἐπίσκοπος, ἐποπτήρ, ἐπόψιος ; ci. A. B. Cook, Zeus ii. 1130, W. Kranz,
Stasimon, 50, 279. The rendering of ἐποπτῆρας by βοηθούς in the scholiast’s
paraphrase on Sept. 640 is appropriate. Thus there is probably a certain
bitterness in the expression chosen by Cassandra; it is almost as though she
said ‘and after he has looked kindly upon me, how I have been insulted, and
insulted in this sacred garb of his’.
1271. ‘By καὶ ἐν τοῖσδε κόσμοις she implies that what ought to have secured
respect only added to the ridicule’ (Paley).
καταγελωμένην: this is what she has experienced time and again in the
past ; because of this experience she has just (1264) felt as κατάγελως the very
fact that she is still wearing the garb of a prophetess.
pera (μέτα F Tr; for the accent see the note on 1037 f.) : the defence of the
MS reading has not been successful. I will not discuss the over-artificial
attempts of Verrall and Pliiss to take μέτα as an adverb. Many editors take
μετὰ φίλων together. The possibility of such a harsh enjambement cannot be
denied. Among others, Tycho Mommsen, Beitr. zur Lehre von den griech.
ı Mazon and Ubaldi (cf. on 1235) also adopt the right punctuation, and so does M. Berti
(in the article cited on 12), p. 243.

587
line 1271 COMMENTARY

Präpositionen, 772 f., gives a judicious examination of the relevant instances."

Sophocles not only puts the monosyllabic ev at the end of a line with its
noun at the beginning of the next line (Oed. C. 495), but also twice (Oed. R.
555, Phil. 626) has ἐπὶ in the same position (1. 6 of fr. 598 N. = 659 P. is doubt-
ful). In Euripides I can find one certain instance, El. 852 f. ἐγνώσθη δ᾽ ὑπὸ |
γέροντος, and we may perhaps add Phoen. 1317 f. ἐγὼ δ᾽ ἥκω μετὰ γέρων
ἀδελφὴν γραῖαν *Ioxdorynv.2 In the dialogue verse of Aeschylus? there is
only one case of a preposition (a monosyllabic one) in a corresponding posi-
tion, Eum. 238 f. mpoorerpuuévost τε πρὸς ἄλλοισιν οἴκοις." It would be
shortsighted to concede to the poet of the Oresteia the use of enjambement in
the case of monosyllabic prepositions, but deny it in the case of disyllabic
ones ; for either way we have a deviation from his normal practice (so far as
we can survey it). This deviation is, however, not surprising in the Oresteia,
but must rather be grouped with the other cases of 'Sophoclean' enjambe-
ment, which are by no means rare in the late plays, the Prometheus and the
Oresteia (cf. on 1354). However, these formal considerations cannot defend
μετὰ here, since it destroys the sense of the passage. It is contrary to the
rules of grammar when Paley translates 'inter ipsos amicos, sc. Troiae,
frustra rideri ab inimicis' or Hartung: ‘Ja, du hast mich oft geschaut in
diesem Anzug unter Freunden arg verhóhnt'. 'Inter amicos' (so already
Heath) would in itself be perfectly correct, but only if the connexion with the
verb were not broken, that is, if the statement to be made were that Cas-
sandra among and along with her friends or kin, who were also mocked, had
to endure the mockery of a third party (so Paley's later explanation). But
this yields an utterly impossible sense here. Cassandra’s misfortune is that
! ie. eliminating the passages in which the preposition at the end of tbe trimeter is
placed between an epithet and its noun. How the thing should not be done is shown by
J. Descroix, Le Trimètre iambique (1931), 291. He makes it impossible for the reader who
does not look up the texts to recognize that cases like S. 4j. 720 f. Μυσίων ἀπὸ | κρημνῶν,
Trach. 557 f. τοῦ δασυστέρνου παρὰ | Νέσσου, etc. belong to a different category from Phil.
626 f. ἐπὶ | ναῦν (cf. P. Maas, Griech. Metrik, $ 135). Contrariwise, there is not a trace in
Descroix of the instance which is closely related to the last, Oed. R. 555 f. ἐπὶ [τὸν σεμνό-
pavrıv ἄνδρα. His treatment (p. 290) of the corresponding phenomena in Aeschylus is very
unsatisfactory.
2 The word is accented μετὰ in all MSS except L, which has μέτα and, in the next line,
the interpolation στέλλων instead of γέρων. From Valckenaer to Murray many editors have
read μέτα; this was rejected by Hermann. I am inclined to prefer the explanation of Paley,
Wecklein, and À. C. Pearson. Wecklein writes (in his annotated edition of 1894, whereas
in his critical edition of 1901 he says: ‘ μέτα L, fort. recte"): ‘ μετὰ γέρων ἀδελφὴν ypatav: in
order to bring γέρων and γραῖαν close to each other [cf. above on Ag. 320] γέρων has been
placed between the prepositiun and the substantive. With ἥκω μετά τινα cf. Alc. 46 ἣν σὺ
νῦν ἥκεις μέτα, Suppl. 670 ἡμεῖς ἥκομεν νεκροὺς péra.'
3 In the Δικτυουλκοί, Pap. Oxy. xviii. 2161, col. 2. 12 f., ξὺν stands at the end of one
pherecratean and μητρὶ at the beginning of the next.
+ With Paley, Blass, Wilamowitz, and others I consider this alteration of Prien’s neces-
sary, since it greatly improves the structure of the sentence.
5 Hermann, ad loc., probably wrongly, pronounces it inadmissible to compare the Sopho-
clean instances with this line, relying on subtle differences in the relations of verse and
sentence structure to each other. He considers πρός corrupt. In this he is followed by
Wilamowitz, though with a different reading (cf. also his Interpr. 220). This seems to rne
a petitio principii, since πρός yields good sense. His treatment of Ag. 1354 shows how reluc-
tant Wilamowitz is to recognize ‘Sophoclean’ enjambement even in the late plays of
Aeschylus,
6 I should not be surprised if some commentators were prepared to retain the MS
588
 COMMENTARY line 1272

no one at all has believed her, that even those closest to her have turned
against her. That could perhaps be expressed by saying that she was.
mocked by friend and foe alike. This is the interpretation, e.g., of Headlam
(influenced to some extent by Paley) and (with a different text) of Wila-
mowitz. These explanations, as well as those of, e.g., Enger, Sidgwick, Verrall,
are further complicated by the fact that they give a wrong sense to οὐ διχορ-
ρόπως. This phrase does not mean ‘by one just as much as the other’,
‘by friends and foes alike’; nor does it mean what Wilamowitz implies:
‘amicos inimicosque distinctos esse docet οὐ διχορρόπως ᾿. Accurate informa-
tion is to be found in L-S: ‘Adv. διχορρόπως waveringly, doubtfully, used only
by Aesch. and always with a neg., où or μὴ dey.’ In the five passages where
the expression appears (Suppl. 605, 982, Ag. 349, 815, 1272) its use is quite
stereotyped and the sense is clearly ‘with an unambiguous result, in a
decision leaving no room for doubt’. On the position of οὐ διχορρόπως in this
sentence see below. Whereas for μετὰ no plausible explanation has been put
forward, Hermann’s μέγα has a high degree of probability. It goes excellently
with καταγελωμένην, cf., e.g., Prom. 1004 τὸν μέγα στυγούμενον, Eum. 113 ὑμῖν
ἐγκατιλλώψας μέγα. Adverbial μέγα often in the tragedians stands at the end
of the verse as was noticed by Peile. The interchange of μετα and μέγα is
common, e.g. in Cho. 137 μέγα is generally accepted in place of the MS reading
μέτα; conversely, in S. Phil. 515 the MSS have μέγα τιθέμενος instead of the
correct reading perarif., preserved in the paraphrase of the scholia of L.
φίλων ὑπ᾽ ἐχθρῶν, ‘by mine own people who were hostile’ is excellent and
in view of the whole situation far better than ‘by friends and foes alike’.
Clearly, in this section (from ἐποπτεύσας on) she 15 speaking only of her
experiences in Troy, as is shown above all by 1273 f.; only in 1275, with the
words καὶ νῦν, does she turn back to the present. For this type of word-order,
φίλων ὑπ᾽ ἐχθρῶν, at the beginning of a line cf., e.g., Pers. 611 Bods τ᾽ ἀφ᾽
ἁγνῆς, Prom. 366 κορυφαῖς δ᾽ ἐν ἄκραις. For the qualification by ἃ word follow-
ing the noun cf. S. Aj. 512 ὑπ᾽ ὀρφανιστῶν μὴ φίλων and especially, with a
paradox similar to the present, A. Suppl. 225 (κίρκων τῶν ὁμοπτέρων) ἐχθρῶν
ὁμαίμων καὶ μιαινόντων γένος, where Wilamowitz’s interpretation, ‘explicat
quos dicat accipitres, nempe inimicos, qui et consanguinei sint et cognatas
attrectent', produces a far more forcible sense than that of Wecklein ‘ ὁμαί-
μων is in apposition to κέρκων ᾿. With the oxymoron φίλων ὑπ᾽ ἐχθρῶν Paley
compares Sept. 695 φίλου yàp ἐχθρά μοι πατρός... dpa... προσιζάνει. One
might also compare the momentous sentence (cf. C. Robert, Osdipus ii. 121)
E. Phoen. 1446 φίλος yap ἐχθρὸς ἐγένετ᾽, ἀλλ᾽ ὅμως φίλος and Antiphon the
sophist fr. 49 Diels (Vorsokr. ii, 5th ed., 358) τοὺς φίλους ἐχθροὺς ποιῆσαι.
The interpretation of the second half of 1272 is difficult and much disputed.
There is no need to consider again the misinterpretations of οὐ διχορρόπως
which we have rejected. Some of the commentators take οὐ διχορρόπως with
ἐχθρῶν; so, e.g., Humboldt ‘sichtbar feindgesinnt’, Blomfield (coniungenda
sunt ἐχθρῶν où &iyopporws’ ; his reference to Suppl. 982 is misleading, since
there οὐ διχορρόπως really belongs to εἰσίν, which is to be ‘understood’),
reading, take it in a way which is grammatically unobjectionable, and then base on it a
piece of fiction. E.g. ‘there was in Troy a strong party opposed to the royal house. These
people would jeer: “There she goes, the daughter of Priam, the sister of Paris! Such a
beggarly false prophetess does the royal house turn loose amongst us”, etc.
589
line 1272 COMMENTARY

Wellauer (‘ab amicis, qui haud ambigue inimici erant’). But apart from the
grammatical harshness this is unsatisfactory because it does not add to the
powerful phrase φίλων ὑπ᾽ ἐχθρῶν anything in any way essential. The scribes
of F and Tr put a stop after ἐχθρῶν and therefore intended οὐ διχ. μάτην to be
taken together. This is how the passage is taken by Stanley, Schütz (‘quam-
quam haud ambiguum est istum risum frustra fuisse’), Weil, Hense (Schneide-
win considered giving ἃ completely impossible sense to μάτην). 1 am inclined
to believe that this is right. ‘Tam dubium non est me immerito irrisam esse,
quum omnia quae civibus praedixeram evenerint’ (Weil; the substance of his
rendering was anticipated by Schutz). That an adverbial qualification so
packed with meaning should follow and supplement a sentence or clause
already complete in itself would be entirely in the manner of Aeschylus; cf.
Wilamowitz on Suppl. 813 and below on 1581. où διχορρόπως adds weight to
the bitter μάτην (adequately rendered by Humboldt ‘wahnverblendet’) : now,
after the fall of Troy, the outcome removes all doubt of the ματαιότης of the
people who once laughed at Cassandra. Elsewhere also οὐ διχορρόπως relates
to a decision arrived at, e.g. in a court of law (Ag. 815) or in the assembly
(Suppl. 605). The collocation οὐ διχ. μάτην should presumably rouse no mis-
givings. One might perhaps cite as a parallel Eum. 952 f£. περί τ᾽ ἀνθρώπων
φανερῶς τελέως διαπράσσουσιν, Where φανερῶς has indeed often been challenged,
but, it seems, mistakenly.! But even without parallels οὐ διχορρόπως μάτην
should, if I am not mistaken, be regarded as an unexceptionable phrase. It is
perfectly intelligible that Cassandra should here not merely speak of the
mockery, but should add how unjust and undeserved was that mockery ; she
has proved herself the faithful mouthpiece of the all-knowing god ; if in spite
of that he has abandoned her, then his treatment of her is doubly hard and
her passionate rebellion entirely comprehensible. The slight uncertainty
which still remains in this passage extends only to the last three words; I
will not deny the possibility (not probability) that in the place of μάτην
Aeschylus wrote something else.
1273. φοιτὰς ὡς ἀγύρτρια goes together (Hermann and others, including
Mazon, wrongly put a comma after gourds). This was recognized by Heath,
who, however (and with him Schiitz), tampered with καλουμένη, as did also
Musgrave and recently Murray. It is difficult to understand how anyone can
fail to see that the injury consisted in ‘calling names’. It would be very odd?
if Aeschylus meant to say that Priam almost let his daughter starve to
death (‘Poor, starving and reviled, I endured all’ Murray, who reads κακου-
μένη). People called after Cassandra ‘starving beggar-woman’ and all the
other terms of abuse habitually showered on any φοιτὰς ἀγύρτρια. The com-
pletely normal use of ὡς in a comparison makes Wilamowitz's artificial
explanation (app. crit.) unnecessary.
ἀγύρτρια: this is the substantival part of the phrase and φοιτάς the attri-
ı If, as I believe, the text of the MSS is sound, φανερῶς and τελέως must not be taken as
standing in asyndeton on the same level (‘plain and full is their accomplishment’ Sidgwick),
but φανερῶς must be regarded as a qualification of the following word ; Headlam translates
accordingly ‘and in the affairs of Man most manifestly their dispose is absolute’.
2 Some scholars are aware of the oddity and yet resign themselves to it, as, e.g., Weil:
‘Miror regiam virginem fame enecatam; sed omnia interpretandi artificia . . . mihi non
videntur efficere ut Aeschylus hoc non dixerit.' All that was required was that he should
change his unfortunate punctuation.

590
 COMMENTARY line 1274

bute. Such translations of ἀγύρτρια as mendica (Dindorf), ‘Bettlerin’ (Passow—

Cronert) are not specific enough, while translations such as circulatrıx (Blom-
field and thereafter e.g. Verrall and Headlam ‘mountebank’), ‘fortune-teller’
(Conington), etc., emphasize too strongly a (possible) secondary connotation
and omit something essential. ἀγύρτης (besides the lexica cf. Stengel, RE i.
915 ff.), from dyeípew (on the formation of the word see Ernst Fraenkel,
Nomina agentis, i. 55 f.), denotes above all a person who as priest or prophet
collects gifts, at first really for the god or goddess, but later, as the institution
degenerated, more and more as a ‘beggar’ for himself. Cf. Stein on Hdt. 4.
35. 3 καὶ yap dyeipew σφι τὰς γυναῖκας, where the whole proceeding is per-
fectly honourable and there is no imputation of anything contemptible,’
just as in a fragment of Aeschylus (fr. 168 N., to which may now be added
Pap. Oxy. 2164, fr. 1. 16 ff.) Hera disguised as a priestess says in solemn
words, without any derogatory undertone, that she is engaged in dyeipew for
the goddesses of the springs of Argos. On the other hand, the context of
Ag. 1273 leaves as little doubt of the scornful note in the word ἀγύρτρια as
does S. Oed. R. 388 f. δόλιον ἀγύρτην, ὅστις ἐν τοῖς κέρδεσιν μόνον δέδορκε κτλ.
in the case of ἀγύρτης (here used of the seer, which is significant) ; again, in
Hippocrates #. ἱερῆς νούσου 1 the juxtaposition μάγοι re καὶ καθαρταὶ καὶ
ἀγύρται καὶ ἀλαζόνες shows that ‘ayyprns has already acquired the sense of
"fraudulent priest" ' (Wilamowitz, Griech. Lesebuch, ii. 168). In the earlier
instances the reference to someone who ‘collects’ in the service of a god,
whether as priest or seer, is everywhere present ; so also with μητραγύρτης,
which, probably by chance, is not attested until the first half of the fourth
century (anecdote in Arist. Rhet. 3. 2, p. 1405°20). The reference to the service
of the god is essential in Ag. 1273, too: that Cassandra appears in Apollo’s
sacred garb (ἐν τοῖσδε κόσμοις) brings the scorn of the mob on her. Thus the
sense is ‘begging priestess’ or ‘begging prophetess’. φοιτάς clearly applies in
the first instance to her wandering from place to place, but it probably also
includes the idea of wandering mentally,* as in Prom. 598 -κέντροισι φοιτα-
λέοισιν, S. Aj. 59 φοιτῶντ᾽ ἄνδρα pavidow νόσοις and elsewhere. So in E. Bacch.
165 φοιτάσιν eis ὄρος the word φοιτάσιν applies both to the movements of the
Bacchants and to being ‘out of their senses’,
1274. The question ‘What does τάλαινα go with?’ leads to the question
‘What is the general structure of the sentence?’ Hermann isolated τάλαινα
from the surrounding words: ‘rdAawa vix illa nisi de se dicere potuit. Sic
enim fere ubique hoc nomen doloris significandi caussa orationi inseritur.
In accordance with this he translates: ‘Sustinui, misera, quasi circulatrix,
vocari insana, mendica', etc. Similarly Paley: 'a beggar (wretch that I was)
and perishing by hunger’, Wecklein : ‘musste ich Ungluckliche mich Hunger
leidendes Bettelweib schelten lassen’, and Wilamowitz : ‘Und ich trug's, dass
man wie ein wabrsagendes Landstreicherweib hungrige Betteldirne mich
1 "The itinerant begging priests and priestesses became objects of scorn when they had
become a prey to Orphism; they once had been something more important’, remarks Ed.
Schwartz, Die Odyssee, 184, where he also discusses A. Ag. 1273 f.
2 Both naturally go together. Thus it is no merely external comparison when Horace
(Sat. 2. 3. 48 ff.) makes the popular philosopher Stertinius say in his sermon about the
nature of insania: ‘velut silvis ubi passim palantis error certo de tramite pellit, ille sinis-
trorsum, hic dextrorsum abit, unus utrique error sed variis inludit partibus, hoc te crede
modo insanum, nihilo ut sapientior ille qui te deridet caudam trahat.’

591
line 1274 COMMENTARY

unsel’ge schalt.’ It seems to me exceedingly harsh to detach. τάλαινα thus

from the complex of scornful terms (πτωχός, λιμοθνής), in the midst of which
it stands and to take it as an expression of self-pity. It would be natural for
all three terms in 1274 to express the scorn of the people. This consideration
is satisfied either by understanding the phrase as ‘being called beggar, poor
wretch, starveling’ (Headlam) or by regarding the three words as the expres-
sion of a single complex idea: ‘dass man mich . . . unseliges halbverhungertes
Bettelweib genannt’ (Nägelsbach). For the king’s daughter the expression
of pity (τάλαινα) in close conjunction with the social stigmas (πτωχός and
᾿λιμοθνής) only makes the scorn more bitter.
‚1275. The sense of ἐκπράξας is much disputed. A number of scholars, e.g.
Humboldt, Dindorf (Thes., Lex. Aesch.), Schneidewin, Kennedy, Headlam,
Mazon, G. Thomson take it as meaning ‘after making me a prophet’. Thomson
asserts that this meaning of the verb is ‘well attested’. I know no evidence
for this particular sense and construction (double accusative) of ἐκπράττειν =
reddere aliquem aliquid ; and even with the simple verb such a use as Pind. N.
5. 36 ὥστε... ποντίαν χρυσαλακάτων τινὰ Νηρεΐδων πράξειν ἄκοιτιν appears to
be very rare. We need not therefore consider the above explanation. Others
take ἐκπράξας ἐμέ as ‘who makes an end of me, kills me’; this has often been
advocated from Triclinius (ἤγουν φονεύσας) and Stanley to Wecklein, Verrall
(with a slight variation), Sidgwick, L-S?. For ἐκπράσσειν = ‘to kill’ the
passages E. Hec. 515 and 8. Oed. C. 1659 have long been adduced as evidence.
The objection that this sense of ἐκπράσσειν is not attested in Aeschylus, who
provides six instances of the verb, might be dismissed if the sense produced
were effective. But that is not the case; on the contrary, the ominous close
of the long period (1269-76 ; see above), θανασίμους τύχας, would be robbed of
its effect if it were almost immediately preceded by a verb meaning ‘kill’.
If we review the Aeschylean instances of ἐκπράσσειν, we find Suppl. 472 εἰ μὲν
yap ὑμῖν μὴ τόδ᾽ ἐκπράξω χρέος, where the interpretation ‘If I refuse to per-
form this service for you’ (Headlam; a very similar version had been given
by Paley) is unsatisfactory, since there is too strong an agreement with such
turns of speech as Pindar, Οἱ, 3. 7 πράσσοντί με τοῦτο... χρέος, P. 9. 104
πράσσει χρέος and with technical phrases like Thuc. 8. 108. 2 ‘AAwapracoéas . . .
πολλὰ χρήματα ἐξέπραξε (cf. also E. Her. 43 μή more . . . ἐκπράξωσιν αἵματος
δίκην etc.) ; it should therefore be taken as Tucker takes it: ‘If I shall not
succeed in collecting (i.e. making up) for you a quittance of this claim.’ It is
weli known that in several passages of the Oresteia πράσσειν (sometimes in the
middle) is used of exacting payment of a debt (cf. on 812, 823). Keck was
therefore right in requiring for ἐκπράξας here the sense ‘exact’.” So also
Droysen: ‘Nun führt der Seher mich, die Seherin, hieher, dass meine Schuld
ich zahle, her zum Untergang’, and Wilamowitz: ‘jetzt fordert sich der
Seher seine Gabe heim’. Cassandra herself is the debt owed to the god: he
has now called in (exegit) her person and her life. This interpretation alone
gives full value to the close connexion of ἐκπράξας with the double μάντις; ‘He
the seer has called in the seeress’. By bestowing on the maiden his own gift
1 Beazley remarks: ‘I can’t help thinking that λιμοθνής refers to leanness in Cassandra,
natural in one possessed (even in the midst of plenty), and is not merely a “social” word.’
Seeing that it stands beside πτωχός, this seems to me not very probable.
2 Keck could not know that Casaubon, too, had translated ‘poenam a me exigens’.
592
 COMMENTARY line 1278

of prophecy, in order to seal the bond of love for which he was striving,
Apollo made her his own ; now he has reclaimed his property : the consequence
is that she must die.
1276. ἀπάγειν here probably = ‘to arrest and carry off’, as, e.g., E. Bacch.
439 καὶ δεῖν κἀπάγειν ἐφίετο (cf. below 1632 déni).
1277. βωμοῦ πατρώιου δ᾽ avr’: the postposition of a ‘preposition’ seldom
occurs in Aeschylus save at the end of a line, but is unexceptionable; cf.
Wecklein, Stud. z. Aesch. 79 ff., and also the remarks of Denniston on E. El.
574. The postposition of ἀντι seems to be a relic of a very early stage of the
language; cf. Wackernagel, Unters. z. Hom. 181 f. n. 3, who, after pointing
out that in Homer ἀντὶ does not occur in real 'anastrophe', observes: ‘The
Γανυμήδεος ἀντί (ἄντι) of the Cypria [rather, of the Little Iliad, fr. 6], Hesiod’s
ἵνα μὴ βασιληΐδα τιμὴν ἄλλος ἔχοι Διὸς ἀντι θεῶν ἀειγενετάων (Theog. 893),
Aeschylus’ βωμοῦ πατρώιου δ᾽ ἀντ᾽ ἐπίξηνον μένει are scarcely likely to be due
to artificial archaizing. Rather should these passages . . . be reckoned among
the passages of post-Homeric poets which fortunately preserve ancient forms
of speech lost in Homer.’
ἐπίξηνον : rightly explained by H. Estienne in the Thesaurus, and also by
Stanley (‘truncus culinarius’), Blomfield (‘a chopping-block’), Wellauer,
Linwood, Dindorf in Lex. Aesch. Unfortunately, however, Passow assumed
the special meaning 'Henkerblock' for Ag. 1277 and the passages in the
Acharnians, and this (‘the executioner's block’) has been repeated in all
editions of L-S, and that though Schol. Ar. Ach. 318 (hence Etym. Gen.,
Hesych., Suidas) gives the correct explanation ἐπίζηνος καλεῖται ὁ μαγειρικὸς
κορμός, ἐφ᾽ οὗ τὰ κρέα συγκόπτουσιν. In agreement with this is the classifica-
tion of ἐπίξηνον among rà μαγειρικά in Pollux 6. go and το. rox. Elmsley
quoted the correct explanation on Ar. Ach. 318. It ought, indeed, to be self-
evident that Aristophanes is following his normal practice and turning the
merely metaphorical language of the Euripidean hero (fr. 706 N. ‘Ayaueuvor,
οὐδ᾽ εἰ πέλεκυν ἐν χεροῖν ἔχων μέλλοι τις εἰς τράχηλον ἐμβαλεῖν ἐμόν κτλ.) not into
a technical detail of a (pretended) execution, but into a trait from ordinary
domestic life." In her prophetic vision Cassandra is aware not only of the
horror but also of the shame of her fate in all its awfulness: she, who used to
play her part in the worship at the altar of her father’s house, is now herself
a Voropfer (preliminary victim) to be, not sacrificed, but slaughtered and
hacked to pieces like a beast, the flesh of which is cut up small for the
kitchen on the chopping-block. Miss Lorimer, who has given a good apprecia-
tion of the passage (C.R. xlv, 1931, 211 f.), remarks: ‘Aeschylus is not Seneca ;
the sinister hint conveyed by a single word is enough.’
μένει: immediately after ἐμὲ ἀπήγαγ᾽ és θανασίμους τύχας it need not be
expressly said whom the chopping-block awaits, but it is once again made
clear by κοπείσης which follows.
1278. It is obvious that of the two datives θερμῶι and $owiw one must be

* Wilamowitz, who has discussed the point in his commentary on A. Cho. 883 and pointed
out that a block was not used for executions in Athens, has rightly concluded that in Cho.
883 the conjecture ἐπιξήνου 1s wrong ; nevertheless, it has been put in the text by Headlam
(note on his translation), Murray, and G. Thomson. The derogatory note conveyed by
ἐπίξηνον, highly suitable though it is for Cassandra here, would be out of place in the words
of the servant in Cho. 883.
4872.3 I 593
line 1278 COMMENTARY
wrong (Verrall’s artificial explanation is of no avail) ; alongside two similar
case-endings the ending of a word in a different case is, as is well known,
often imperilled in the transmission. We have the choice between θερμὸν
(Schiitz) and φοίνιον (C. G. Haupt). The latter is preferable; it yields a
powerful picture. Haupt cites the translation of Heinr. Voss, who had
perhaps already adopted the reading φοίνιον : ‘ein Block, von meines Schlacht-
hiebs heissem Strom hellrot gefärbt’. φοίνιον was adopted by Wecklein,
Headlam (prose translation), Wilamowitz, etc. Accepting this, we have to
regard θερμῶι προσφάγματι as dependent on φοίνιον (in principle the dative
is the same as in ῥέε δ᾽ αἵματι γαῖα, etc.; the meaning of mpdcdayua—see
below—makes the construction here somewhat bolder, but it is perfectly
intelligible), while κοπείσης (regarded by other editors as genitive absolute)
is dependent on προσφάγματι. Taken thus, the tense of xometons causes no
difficulty. On the whole conception expressed in 1278 Headlam (in a note on
his prose translation) observes : ‘her second-sight prefiguring the block already
red with blood: compare the vision of Theoclymenus in Hom. v 348-357’.
προσφάγματι, The meaning of the term has been explained by E. Rohde,
Psyche, i, sth ed., 222 n. 1, proceeding from [Plat.] Minos 315 € ἱερεῖα mpo-
σφάττοντες πρὸ τῆς ἐκφορᾶς τοῦ νεκροῦ and taking into account the provision of
the burial-law of Ceos (Dittenberger, Syll. 1218, 1. 12) προσφαγίωι xpeodaı κατὰ
τὰ πάτρια, the Solonian legislation and the phrase which alludes to this custom
in E. Hel. 1255 προσφάζεται μὲν αἷμα πρῶτα νερτέροις. The πρόσφαγμα is thus
originally the blood-offering made to the dead before the burial proper, or,
more accurately, before the exdopa. Miss Lorimer, loc. cit. (on 1277), has
rightly brought Ag. 1278 into connexion with this:' ‘Cassandra alludes in
grim parody to the obsequies which Clytemnestra will provide for Agamem-
non: the πρόσφαγμα at least he shall have in her blood.' The meaning, then,
is this: ‘The chopping-block which is blood-red from the hot funeral sacri-
fice of (me,) the slaughtered woman.' Everything is compressed, hinting
(as was also the case with ézi£qvov) rather than describing fully. Time and
again in Aeschylus, above all in the Oresteia, do we encounter this ‘parodying’
of sacred rites.
It is, of course, conceivable that Aeschylus stripped πρόσφαγμα of its
special reference to the burial and used it simply in the sense of Voropfer
(preliminary sacrifice), just as, e.g., he used προτέλεια (cf. on 65, 720) without
particular reference to marriage rites. For this reason (after long hesitation)
I cannot pronounce Headlam's ingenious conjecture κοπέντος to be impos-
sible; the passage then would mean 'bloody with the still warm preliminary
slaughter of a butchered man'. Of the corruption of κοπέντος to κοπείσης
Headlam says: ‘it was a deliberate alteration made by a half-intelligent
corrector, who took the participle as referring to Cassandra, and therefore
made it feminine’. If we adopt κοπέντος, we undoubtedly secure a simpler
expression and one free from any veiled allusion; but in the language of
1 Though the article πρόσφαγμα in L-S came out in 1934, it makes no use of Miss Lorimer's
paper nor of the section in Rohde's book.
2 By it and the note on it (cf. the posthumous [1910] edition of his work on the 4g.) he
retracted his treatment of the text in his prose translation which I have discussed above.
3 The ‘parallels’ which Headlam adduces, viz. Ag. 263 and 271, are not to the point, for in
these two passages the alteration of the gender of the participle is the logical result of the
confusion by which the lines were ascribed to the wrong speaker.

594
 COMMENTARY line 1279

Aeschylus, particularly from the lips of this prophetess, this is not necessarily
any guarantee of correctness. What induces me to leave κοπείσης in the text
is not only the consideration that with it we can obtain an intelligible thought
by taking into account the original sense of πρόσφαγμα (see above), but also
the doubt whether the Aeschylean Cassandra could ever, even within the
narrowest limits, speak of Agamemnon’s death in relation to her own as a
preliminary sacrifice. Elsewhere (1137, 1260f., 1325 f.) she regards what
awaits her as an addition to the fate of Agamemnon, as an accompanying
circumstance; this seems to be characteristic of her whole attitude, and we
should therefore be reluctant to turn it into its opposite. If, on the other
hand, we retain the MS reading, so that Cassandra styles the murder which
is threatening her a funeral sacrifice in honour of the dead king, then the
typical relation between Agamemnon’s death and that of Cassandra prevails
here too.
1279f. Many commentators have referred the plural here to Cassandra
alone, on the analogy of such well-known cases as 5. Ant. 926 παθόντες av
ξυγγνοῖμεν ἡμαρτηκότες
ἡ etc. (cf. Kühner-Gerth, i. 83, and especially Wacker-
nagel, Syntax, i. 98 ff.) ; this use is not foreign to Aeschylus (cf. on 1552 f.).
In Cho. 20x ff. we find in Electra's speech a sudden transition from the singular
(199 éuot) to the plural! The sentence Ag. 1279 ff. was referred to Cassandra
alone by Schütz (‘non tamen inultae’, contrasting with Stanley’s nu/iz), and
later by, e.g., Nägelsbach, Kennedy, Wilamowitz, Platt, Mazon, MacNeice ;
they all translate τεθνήξομεν and ἡμῶν (this latter is omitted by Mazon) by
singulars. This is improbable. It is clear from 1281 that Cassandra, in accord-
ance with her whole conception (cf. on 1278), is here regarding her own fate
in connexion with that of Agamemnon. van Heusde is therefore right: ‘ego
et Agam.’; so explicitly Verrall. Cf. 1313 f. ἐμὴν ‘Ayauéuvovés τε μοῖραν. In
1318 f. Cassandra speaks of vengeance for herself and Agamemnon, that is,
of the fulfilment of her prophecy in 1279 ff.
1279. ἄτιμοι: correctly rendered by Stanley with inulti and glossed by
Abresch with ἀτιμώρητοι (on Cho. 1019 M has the gloss ἄτιμος" ἀτιμώρητος;
the same rendering is given in the ancient lexica to the orators: a collection
of the relevant passages can be found in Greene’s edition of the Scholia
Platonica, Philol. Monogr. publ. by the Am. Phil. Ass. viii, on Νόμοι 855 c,
p. 341). The next verse (with τιμάορος) expresses the same idea in a positive
form. Linguistically, of course, ἄτιμος is not ‘equivalent to’ ἀτιμώρητος, but
bears the general sense ‘that which has no τιμή᾽ (cf. on 238 and 412) ; here and
in kindred passages it has a somewhat specialized meaning : “that which has,
can obtain, no satisfaction (retribution)'.? Particularly apt is Abresch's
reference to the formula of the old law of murder preserved in Dem. 9. 44
ἄτιμος τεθνάτω ; the audience must have been reminded of it by Cassandra's
words; cf. also Daube, Rechtsprobleme, 198. Valckenaer on E. Hipp. 1417
compared the use of ἄτιμοι there? with that in our passage (he even suspects
borrowing on the part of Euripides, which seems unlikely).
1 Neither the attribution of 201-4 to the Chorus (Hermann) nor transposition (Weil) is
admissible; cf. Appendix D.
2 On the relation between τιμή and τιμωρός cf, Wilamowitz, Kl. Schr. v. 1. 127, and
Wackernagel, Synt. 1. 126f.
3 Wilamowitz ad loc. paraphrases ἄτιμοι with ἀτιμώρητοι.
595
line 1279 COMMENTARY
For οὐ μήν... ye Sidgwick quotes S. Oed. R. 810 οὐ μὴν ἴσην γ᾽ ἔτεισαν; for
further examples, see Denniston, Particles, 335.
1280. ἄλλος: ‘alius ac nos, qui mortui iacebimus’ (Klausen). Schneidewin
and others have referred this to the so-called pleonastic use of ἄλλος, but it is
not quite of the type to which this term is usually applied, since rıudopos
must be taken predicatively with ἥξει; cf. on 530. For ἄλλος ad cf. Ar. Thesm.
664 εἴ τις ἐν τόποις ἑδραῖος ἄλλος αὖ λέληθεν ὦν and ἕτερος ad, common in
Aristophanes.
τιμάορος:: cf. on 514.
It seems to me certain that the famous line (Virg. Aen. 4. 625) exoriare
aliquis nostris ex ossibus ultor was consciously borrowed from 1280.’ Even the
sequence of the words has been to a great extent preserved. The loving care
with which Virgil chooses ornaments from Greek tragedy to adorn the last
speeches of Dido is characteristic of his admiration for the great poetry of
Athens. This is not a matter of mere details. It has long been realized that
‘the episode of Dido is worked out very much in the spirit of the Greek
tragedy’ (H. Nettleship, Suggestions to a Study of the Aeneid, 1875, 34).
1281. φίτυμα (τέκνον, γέννημα Hesychius) appears not to be attested again
until Lycophron and the anecdote in [Plutarch] Apopth. Lac., Mor. 241 a,
with the epigram connected with it.
ποινάτωρ : elsewhere only in E. El. 23, 268, also with à; cf. in addition E.
Iph. T. 1433 (dialogue) ποινασόμεαθα. It is doubtful whether here and in
similar cases it is right to speak of Doric forms, as does, along with many
others, Ernst Fraenkel, Nomina agentis, ii. 16 and 23; cf. for the opposite
view Wilamowitz on E. Her. 377. G. H. Mahlow, Neue Wege durch d. griech.
Sprache (1926), 140, discusses ποινάτωρ together with other a-forms of similar
words found in dialogue,? namely θοινάσομαι (but he ought to have said that
in A. Prom. 1025 ἐκθοινήσεται is the reading of the MSS), @owarrp, θοινάτωρ
etc. and εὐνάτειρα, εὐνατήριον. In accordance with his general tendency
Mahlow regards these forms as Attic, in this case perhaps rightly.
1282. In Cho. 1042, where Orestes says ἐγὼ δ᾽ ἀλήτης τῆσδε γῆς ἀπόξενος, ‘the
poet deliberately recalls the prophecy of Cassandra by the device of verbal
repetition’ (Wilamowitz, ad loc.) ; in Orestes’ fate it is a frightful duplication
that after his deed as before it he is to be ‘im Elend’ (which means primarily
‘in exile’ but also ‘in misery’). ἀπόξενος is used with the same construction
again in Eum. 884, absolute and with a different meaning in 5. Oed. R. 196.
It is possible that Aeschylus coined the word for this passage.
1283. κάτεισιν : every reader will think of the sentence expounded by Aristo-
phanes (Frogs 1165) from the beginning of the Choephoroe ἥκω γὰρ ἐς γῆν τήνδε
καὶ κατέρχομαι; cf. 1647 and Eum. 462 κατελθών (in each case with reference
to Orestes); cf. also on 1607.
θριγκώσων : for the history of the word and the influence which this passage
had on Euripides see Blomfield. With the imagery and the sound-pattern of
x A. S. Pease, in his commentary on Aeneid iv (Harvard 1935), refers to Ag. 1280. Of
course Virgil does not borrow slavishly : Dido points emphatically to the definite avenger,
Hannibal, by using the form of the ἀποστροφή in exoriare, and to the continuity of Carthage
by saying nostris ex ossibus.
2 Mahlow himself, influenced by his conception of the linguistic form of the lyric parts of
tragedy, does not make sufficient distinction between instances occurring in dialogue and
those in lyric parts.

596
 COMMENTARY lines 1286-90
the verse-ending van Heusde compared Cho. 705 τοιόνδε πρᾶγμα μὴ καρανῶσαι
φίλοις.
Hermann’s transposition of 1290 to follow 1283 is wrong; see p. 600 f.
1284. Since we assume no lacuna before this verse, we must read ἄξει with
G Tr (déew F, i.e. a mistaken doubling of the v which follows).! The sentence
ἄξει νιν κτλ. is an asyndetic explanation of what precedes it: it states the fact
which makes the return of Orestes certain; for the underlying belief cf.
Cho. 329 ff. Examples of explanatory asyndeton are given in Kühner-Gerth,
ii. 344, cf. also on 36, 1261. The explanatory or confirmatory relation in which
this sentence stands to 1280 ff. also makes clear the shade of meaning of
ἄξει, which is obviously a complementary expression to ἥξει and κάτεισιν;
thus ἄγειν means ‘bring’, ‘cause to come’, cf. 853 οἷπερ πρόσω πέμψαντες
ἤγαγον πάλιν. Possibly, however, the idea is also present that the lying there
of his father’s corpse leads the son and sets him on his way as a divine power
might do (cf., e.g., 5. Oed. C. 252 εἰ θεὸς ἄγοι, 998 θεῶν ἀγόντων).
ὑπτίασμα : Casaubon explained ‘preces quas fudit pater cum occideretur ;
ὑπτιάζειν enim est expansis manibus orare',? and le Père Brumoy translated
‘l’imprécation d'un père mourant’. This interpretation is obviously influenced
by Prom. 1005; it was rejected by S. Butler, but adopted by Wilamowitz,
who adds the note ‘ ὑπτίασμα quid sit docet Pr. 1005’. Wilamowitz’s transla-
tion ‘ihn ruft der Vaterleiche handeringend Fleh’n’ suggests an appalling
picture. As for the idea of the ‘pére mourant’, it seems to be excluded by
κειμένου. It is entirely unjustifiable to infer from the clear sense of ὑπτιάσ-
μασιν χερῶν that ὑπτίασμα by itself could mean 'entreaty'. Almost universally
the word has been taken as the designation of ‘lying on the back on the
ground’ (so already Triclinius: ἡ daria κατάκλισις, ἤγουν τὸ νεκρὸν αὐτὸν εἶναι)
and the Homeric phrases ὁ δ᾽ ὕπτιος οὔδει ἐρείσθη and 6 δ᾽ ὕπτιος ἐξετανύσθη
have been compared with it. Plüss acutely observes: ‘ κειμένου : emphasizing
the powerlessness ; in contrast to dex’.
1285. κάτοικτος is the likeliest restoration of the corrupt κάτοικος. That it is
not attested elsewhere scarcely counts as an objection: Aeschylus uses
κατοικτίζω, and ἔποικτος has been rightly compared with κάτοικτος (e.g. by
Hartung and Schoemann, Opusc. ili. 144) ; it is attested only once, but that is
in this play, 1614 (cf. 1221 ἐποίκτιστος). On the active use of verbal adjectives
in -ros cf. on 12. So far as concerns the sense Wilamowitz’s objection ‘a
lugendo exempla ista eam non deterrent’ is invalid. If the sentence 'Agamem-
non's and my death will not go unexpiated' is followed by "Why then do I
mourn and lament?', the resulting connexion of thought is excellent. In the
next note it will be shown that Wilamowitz, like most others, has mistaken
the relation of the sentence τί δῆτ᾽ éyw . . . ávaorévo ; to its surroundings.
1286-90. Before going into details we must get the structure of the whole
clear. The punctuation which prevails in the editions makes it impossible for
the reader to understand that structure. What we may term the 'vulgate' is
Stanley's punctuation (Victorius and Canter had given practically the same):
a light stop (comma) after 1285 ἀναστένω, a heavy stop (full stop Victorius and
! Hermann, Vitelli-Wecklein, Wilamowitz, Murray, Thomson state that vw is the
reading of F; this is wrong: it is clearly wv (van Heusde and Mazon are right on this point).
2 But in an additional note he considered the alternative : *vel ὑπτίασμα κειμένου πατρὸς
est periphrasis pro "Agamemnon pater caesus" >» ’.

597
lines 1286 ff. COMMENTARY
Canter, question-mark Stanley) after 1288 κρίσει; 1.6. the sentence 1286 ff.
ἐπεὶ τὸ πρῶτον κτλ. is made subordinate to the question ri δῆτ᾽ ἐγὼ κτλ.
Stanley’s translation corresponds to this. This ‘vulgate’ is found in Porson,
Blomfield, etc., and, e.g., in Wecklein’s annotated edition (of 1888)! and sub-
stantially in the texts of Wilamowitz, Murray, and G. Thomson. Quite a
different view of the sentence-structure is taken by those editors who put
a question-mark after 1285 ἀναστένω and a comma after 1288 κρίσει, thus
making the ἐπεί clause the protasis to 1289 ἰοῦσα πράξω. This was done by
Schütz, as well as by Wellauer, Hermann, ? Peile, and later, e.g., by Kirchhoff,?
Wecklein (1885), Sidgwick, Weil in the Teubner text, Headlam, Mazon. Only
this way of taking the passage is correct, since a sentence beginning with
ri δῆτα draws the conclusion from a preceding statement (this need not be by
a different person ; for the continuation of an utterance of the speaker himself
cf. Ag. 1264, S. Phil. 1060, Oed. C. 1308, E. Alc. 960, Her. 1301, and other
instances in Denniston, Particles, 270). It is unexampled that a question
introduced by ri δῆτα should not refer to what precedes but have its reason
supplied by an erei or similar clause which follows.* Cassandra’s argumenta-
tion—from 1. 1279 οὐ μὴν ἄτιμοί γ᾽ to 1285 ἀναστένω ;—runs like this: ‘It is
quite certain that my death will be avenged; why then do I lament?’ The
question is equivalent to a negative statement (‘there is no need, then, to
lament’). What she intends to do instead is attached in the form of an
adversative asyndeton (as, e.g., S. Aj. 470 οὐκ ἔστι ταῦτα. πεῖρά ris ζητητέα krÀ.) :
‘after witnessing the fate of my city... I will endure...’.
1286. ἐπεὶ τὸ πρῶτον: ‘now that ...’. Instances of the use here exemplified
are given in L-S p. 1535 s.v. πρῶτος B. III. 3 e, under the heading ' πρῶτον,
πρῶτα are used after the relat. pron. and after relat. advbs., like English
once (= at all)’. The use is predominantly epic, cf. also Leaf on A 235. ‘The
sense as soon as is never necessary in Hom.’ (L-S); it is alien to this pas-
sage, too.
1287. πράξασαν ὡς ἔπραξεν: cf. on 1171.
Musgrave recognized and remedied the corruption in εἶχον. εἶχον has,
1 On the other hand, he gives the correct punctuation (see below) in his critical edition
(1885). Wilamowitz, too, in his translation, decided for the correct stopping. But in his
edition he puts, it is true, a question-mark after dvaorevw, but at the same time a full stop
after κρίσει; Murray and G. Thomson use the same punctuation. That Wilamowitz at the
time of his edition did in fact regard the érei-clause as the protasis to the question ri δῆτα
κτλ, is made clear by his saying: ‘nam a lugendo exempla ista [i.e. obviously the fate of
Ilion] eam non deterrent’. Weil, on the other hand, who in his edition of 1858 followed
Stanley’s punctuation, in his later Teubner text made the right decision.
2 We should not on this point go by the punctuation of the posthumous edition, for which
it is possible that M. Haupt is responsible ; it is the text printed in Opuse. ii. 80 which shows
Hermann’s own punctuation, and, as was to be expected, Humboldt’s translation agrees
with it.
3 The passage was treated accordingly by Martin P. Nilsson, ‘Die Kausalsätze im
Griechischen bis Aristoteles, I’, in Schanz's Beitr. 2. hist. Syntax der griech. Sprache, Heft xviii,
1907, 72, who based his investigation on Kirchhoff’s text.
4 G. Thomson missed the truth by a hairsbreadth. He says: ‘Why does Cassandra think
it unnecessary to lament (1285)? It would be natural to find the reason in the clause which
follows (1286-8), because such rhetorical questions were often explained by a clause begin-
ning with émet, That is true, but entirely leaves out of account δῆτα. Thomson accordingly
translates simply ‘why do I weep for this... .’; in contrast, Paley and Headlam rightly say
‘why then do I wail...’
598
 COMMENTARY line 1288
however, been retained by later editors, e.g. Hermann, Schneidewin, Kirch-
hoff, Wilamowitz in his translation (not in his edition). It would have to
apply to the Trojans (so the translations of Humboldt and Wilamowitz and
Schneidewin’s paraphrase), but such a distinction (introduced by οἱ δέ)
would be absurd after the phrase which affirms their total destruction, 'ZAiov
πόλιν πράξ. ὡς ἔπρ., while only the inclusion of the fate which now strikes
down the victor also gives Cassandra’s thought its full force. Others, e.g.
Conington, Nägelsbach, Verrall, Plüss, have understood εἶχον of the con-
querors. The verb would in itself be good in this sense (cf. on 320), but the
tense is wrong, notwithstanding Verrall’s subtleties. εἷλον is better and a
slighter alteration than ἔσχον (Bamberger). In favour of εἷλον is also the fact
that in a context very similar in motive (the atonement of Agamemnon) 1335
we find kai τῶιδε πόλιν μὲν ἑλεῖν ἔδοσαν μάκαρες Πριάμου.
1288. With οὕτως ἀπαλλάσσουσιν Paley compares Hdt. 8. 68 a 2 ἀπήλλαξαν
οὕτω ws κείνους ἔπρεπε [cf.1.16.2 amd... τούτων οὐκ ws ἤθελε ἀπήλλαξε, 5. 63. 4
ὁ μὲν δὴ πρῶτος στόλος... οὕτως ἀπήλλαξε], Ar. Peace 568 4} καλῶς αὐτῶν
ἀπαλλάξειεν ἂν μετόρχιον ; other similar examples may be found in Dindorf’s
Thesaurus, i. 2, p. 1153 D. As in the present passage so in Hdt. 8. 68 a ἀπήλ-
λαξαν οὕτω is used in malam pariem.
ἐν θεῶν κρίσει. Verrall’s! attempt to defend ἐκ θεῶν κρίσει as a syntactical
nicety fails; he misinterprets the construction in Ag. 1366 (see there), and
relies on one other passage alone, Sept. 821 ὑπ᾿ ἀλλήλων φόνωι, which would
have to be emended (so, e.g., Wecklein) were it not the botched work of an
interpolator. The reading ex θεῶν is possibly due to the fact that in the next
line but one (ὀμώμοται γὰρ κτλ.) this expression stands in the same position
in the verse. It seems that ev has not here the function closely related to the
instrumental (besides Kühner-Gerth, i. 464 f. and L-S s.v. ἐν A III cf.
Wilamowitz on E. Her. 932; instances from Aeschylus, not all of them
relevant, in Dindorf, Lex. Aesch. 118, $ 7), but its usual local-temporal force
‘in such-and-such a situation’, cf., e.g., 439 ἐν μάχηι δορός, 1237 ὥσπερ ἐν μάχης
τροπῆι, 1615 ἐν δίκηι, fr. 258 N. στόματος ἐν πρώτηι xapäı, S. Trach. 700 ev
τομῆι ξύλου. ‘Now, when the giving of judgement by the gods has come about’.
κρίσις occurs here alone in the text of Aeschylus, but in addition there is the
title of his play "OrAwv κρίσις.
Cassandra sees the fate which is now overtaking Agamemnon as the result
of a trial conducted by the gods in exactly the same way as Agamemnon
earlier (812 ff.) saw the destruction of the city of Priam. The play as a whole
shows that in both utterances we must recognize, not the partisan views of a
member of the other side, but the conviction of the poet; an important in-
dication of this is the fact that the king’s line of thought corresponds with
that of the first stasimon (367 ff., cf. 355 ff.) and Cassandra’s with the choral
anapaests 1335-42.
The exposition of the reasons which Cassandra gives for her decision to
resign herself to her fate has great power and beauty. ‘My city is utterly
destroyed; in accordance with the verdict of the gods the victor is now
falling victim to his fate (assassination) ; so I will (without further struggle)
1 He has a predecessor in Karsten, who did not, however, realize that ἐκ is the reading of
F, though Blomfield and Hermann had observed it. Karsten referred, like Verrall, to Ag.
1366, as well as to a few other passages which prove nothing.

599
line 1288 COMMENTARY

accept death.” Once again, as elsewhere, she sees herself as a secondary

figure in the tragic process. ‘After all this, what does it matter about me,
Cassandra?’ It is not the same thought as that by which the anxious young
women of Thebes rouse their courage (Sept. 263) σὺν ἄλλοις πείσομαι τὸ μόρσιμον,
but it is closely related to it. After renouncing complaints and lamentation
(1285), Cassandra now, by fixing her mind on the thought that fate hits the
conquered and the conqueror alike, musters the strength to meet her death.
1289. πράξω has not been satisfactorily explained by anyone. Hermann
writes: “πράξω refertur ad sequens τλήσομαι τὸ κατθανεῖν. Ita locutus est
Aeschylus etiam v. 1669 πρᾶσσε, πιαίνου μιαίνων τὴν δίκην. So Wilamowitz
on Her. 323, referring to the same passage (and to Cho. 779 ἄγγελλ᾽ ἰοῦσα,
πρᾶσσε, τἀπεσταλμένα). But clearly this has no special kinship to such a
preceding (or parenthetic) imperative (‘do!’ ‘on!’). Karsten (in opposition to
Hermann) rightly rejects 1669 as a parallel: ‘quod dictum tam constructione
quam sensu huic prorsus dissimile est’. Paley attempts a different explana-
tion, relating πράξω to πράξασαν ws ἔπραξεν, but he himself rightly objects
that when the meaning is to fare ‘from the nature of the case, some adverb
[or the corresponding accusative of a neuter pronoun or adjective] is almost
invariably added to specify the kind and manner of the faring’. Verrall
passed over this grave objection when he adopted Paley’s interpretation.
Bothe’s defence of πράξω, too, seems to me unacceptable ; he understands it as
‘agam, immo patiar mortem’ and explains: ‘omissis verbis lamentationibusque
se acturam esse dicit; verum illud suum agere est pati, si quidem abitura
est ad mortem’ (similarly Hartung: ' πράξω, synonymous with τλήσομαι᾽).
Thus we must agree with those scholars who regard πράξω as corrupt (so
already Auratus). None of the numerous conjectures is really convincing,
though Heath’s κἀγὼ is possible; an adverb, which many have looked for
here, is also conceivable.
1290. That the line does not fit the context in which it is set was noticed by
Schütz. At first he surmised, though with some hesitation, that it has come
into the text from somewhere else; later he placed it after 1284 (ἄξει vw xrA.).
Hermann put it before 1284 (and kept ἄξειν there) ; he has been followed by
most editors, including Wecklein, Headlam, A. Y. Campbell, Murray, etc.
Hermann translates: ‘magnum enim ab diis dictum est iusiurandum, re-
vocaturam eum esse necem prostrati patris’, but unfortunately he does not
say a single word about his conception of that great oath of the gods. Other
scholars, however, have discussed the point. Klausen, e.g., says: ‘iureiu-
rando Agamemnoni vindicem dii Orestem constituunt .... Hoc iureiurando
omnia nituntur, quae in Choephoris Oresti hoc iussum si neglecturus sit
minatur Apollo Cho. 270-96, 1032’. But in the passages of the Choephoroe
which he adduces there is not the slightest allusion to this mysterious oath.'
In spite of this, Paley is perfectly satisfied with Klausen's invention. Others,
e.g. Schneidewin, Wecklein, Murray, have taken refuge in a 37 ff., where the
gods warn Aegisthus and inform him: ἐκ yap ' Opéarao τίσις ἔσσεται Arpeidao
κτλ. But where is there a word of an oath? Furthermore, there would be a
! From time to time someone refers the εὐορκώματα of Cho. 9o1 to a supposed oath of
Apollo (Blass, whose treatment of the passage is particularly high-handed, combines it with
Ag. 1290). After what has been said against this by Verrall and especially Wilamowitz (he gives
an adequate interpretation of the rare word εὐορκώματα), renewed polemic is unnecessary.
600
 COMMENTARY line 1290

difference between an oath that Orestes should avenge his father and a
ὅρκος... ἄξειν νιν ὑπτίασμα κειμ. πατρός. But there is no need to embark on a
discussion of capricious freaks of fancy. The run of the poet’s sentence 1280-4
clearly shows that there is no room before 1284 for Hermann’s proposed
insertion. As has been demonstrated (on 1284), the asyndeton ἄξει νιν «rd.
explains what goes before, i.e. it gives the reason why Cassandra is in a
position to predict with such certainty the return of Orestes. The fact that
his father is lying there slain will bring the son back. The view that this
causal connexion is inevitable belongs to the living belief of the people. So
there is surely no need for Cassandra, in order to confirm the truth of her
prophecy, to refer to an oath of the gods. But there is a consideration which
carries even greater weight: in the whole of this long scene Cassandra utters
all her prophecies as coming from her own vision, her own irrefragable
knowledge, as flowing from her lips by virtue of the seercraft bestowed on
her by Apollo. It would blur the firm, clear outlines of the scene if in this one
passage the seeress referred to something different as the authority for her
predictions and the source of her knowledge, to say nothing of the fact that
this is the very place where she is announcing something that to her must be
self-evident (ἄξει νιν «rA.). Immediately before 1284 the line ὀμώμοται κτλ.
would be not merely superfluous but intolerable. Some editors, e.g. Bothe
and Wilamowitz, have left 1290 in the position in which it is in the MSS and
have regarded the content of the oath as the imminent death of Cassandra.
In this case, ἴοο, we should have to invent ad hoc an oath of Apollo (so Bothe)
or of other gods ; but even more serious is Hermann’s objection that with this
interpretation ‘Cassandra gravius de se et magnificentius loquitur, quam 1118,
in qua est, animi affectio ferre videtur’. It has been shown (cf. on 1278) that
Cassandra consistently regards her own death as a mere addition to the all-
important event, the murder of Agamemnon; it is then impossible that she
should load the fate that threatens her with the enormous weight of a great
oath of the gods. Finally, the complete absence of any clue to the content of
the oath would in itself be enough to bring grave suspicion on the line in this
setting. Wilamowitz realized this as keenly as Schütz ; his supplement dpape
γὰρ {τῶνδ᾽ ὅρκος seeks to supply the want. He has inserted in his text the
anonymous quotation preserved in the Homeric epimerismi, Cramer’s Anecd.
Oxon. i. 88 (= Etym. Voss. s.v. dpapev,' cf. Gaisford in the footnote to Ei. M.
134. 43) dpape γὰρ ὅρκος ἐκ θεῶν uéyas.? The sentence dpape κτλ. was first
brought into connexion with Ag. 1290 by Dindorf (not by Hermann, as
Wilamowitz states). One must leave open the possibility that the quotation
was taken from some other passage, but it is just as likely that it is in fact
derived from Ag. 129o. If this were the case then, as Dindorf held, the variant
ἄραρε yàp would perhaps deserve preference over ὀμώμοται yàp. But, what-
ever its original form, the line cannot have been written for this context;
instead of shifting it to a different position, which it fits just as ill as it does

1 ‘The composers of the Etymologica used as their source the same Homeric epimerismi
from which the composer of Cramer's collection compiled his lexicon', Cohn, RE vi. 18o,
stating a result of Reitzenstein's investigations. On the Etymologicum of cod. Voss. gr. 20
cf. Reitzenstein, Geschichte der griech. Etymologika, 220, 261 ff. ; RE vi. 816.
2 Wilamowitz in his 'Testimonia' inadvertently omitted μέγας, and consequently it is
lacking in Murray.
601
line 1290 COMMENTARY

that given it in the MSS, we should obelize it, as Schütz originally held." If
we reject this course, the only remaining possibility is to assume a lacuna;
this has of course been proposed, but the clear progress of thought from 1289
to 1291 makes it most improbable. If 1290 is ‘interpolated’, that may in this
case, where the line disturbs the sequence of thought, merely mean that some
reader took it from another tragedy, probably Aeschylean,? and wrote it in
the margin (cf. on 527). In that case it is not at all certain that the marginal
note—presumably written from memory—originally consisted of a complete
verse. We may at least wonder whether the original quotation was not simply
what we read in the Homeric epimerismi: dpape yap ὅρκος ἐκ θεῶν μέγας, from
which the complete trimeter of the MSS of Aeschylus was obtained by the
easy alteration of the verb. This, however, is beyond possibility of proof.
Still less is it possible to discover what gave occasion to the quotation. We
may imagine that the mention of the fate of Ilion (1286) reminded a reader of
a tragic passage speaking of an oath of the gods concerning the destruction of
the city, an idea perhaps suggested by Y 313 ff. ἤτοι μὲν yap νῶϊ πολεῖς
ὠμόσσαμεν ὅρκους πᾶσι per’ ἀθανάτοισιν, ἐγὼ (Hera) καὶ Παλλὰς Avr, μή ποτ᾽
ἐπὶ Τρώεσσιν ἀλεξήσειν κακὸν ἦμαρ. But this is mere idle fancy.
1291. Not ‘Ye Gates of Death, I greet you’ (Headlam), but ‘And this gate
do I address as the gate of Hades’ (Nägelsbach). Ἅιδου πύλας is the predicate
noun, τάσδε (scil. πύλας) the object, just as in 162 (τοῦτό νιν προσεννέπω) νιν is
the object and τοῦτο predicative. The phrase was understood correctly, e.g.,
by Droysen, Verrall, Wilamowitz, Platt. It is probable that there is present
in προσεννέπω, if only as an undertone, an echo of the use of the verb in
addressing or invoking a deity in prayer (cf. on 162). This produces a close
connexion with the following word, ἐπεύχομαι, where it is easy to understand
Hades as the god to whom the prayer is directed. Paley alleges: ‘the em-
phatic ἐγὼ is here not required’, and proposes a conjecture. Droysen rightly
understood the function of ἐγώ: ‘Dich, Pforte, grüss’ ich, Pforte mir ins
Schattenreich’ (‘Thee, gate, do I greet, for me the gate into the realm of
shades’) ; so Verrall (n. ad loc.) and Platt: ‘these gates are unto me as the
gates of hell”. .
“Αἰδου πύλας. The material collected by Blomfield, following earlier
scholars, in his glossary is sufficient to illustrate the importance of this con-
ception from Homer on; Usener, Kl. Schr. iv. 226 ff., gives a thorough survey.’
τάσδ᾽ éyw: for the type of corruption in the MSS see p. 655 n. 1.
1292. καιρίας : on the predominantly local meaning of καιρός in early Greek
cf. on 365.
ı In course of time scholars may perhaps come to admit that there are cases where it
is necessary to employ instead of the apparently milder measure of transposition the
apparently stronger measure of obelization. While turning over the pages of vol. x of
the Pap. Oxy. my eye was caught by 1249, a Babrius MS of the second century, where, in
the text of fable 118, 1. 5 is missing, ‘which was rejected by Gitlbauer (temere, Crusius thinks)
and transposed, with emendations, after 1. 6 by Seidler and Bergk’ (Hunt).
2 ἄραρεν occurs in Prom. 60.
3 Even the last line of the inscription above the gate of Dante’s Hell, ‘Lasciate ogni
speranza, voi ch’ entrate’, is not without an ancient forerunner: ‘tanuam hanc Orci . . . quo
nemo advenit, nisi quem spes reliquere omnes (Plaut. Bacch. 368 ff.; what follows is a piece of
comic παρὰ προσδοκίαν). The idea of Ἅιδου πύλαι did not, of course, originate only among the
Greeks, cf. Weinreich, Tübinger Beiträge zur Altertumswissenschaft, Heft v, 1929, 436 f.
(with references to recent discussions of the subject).
602
 COMMENTARY lines 1297f.
1293. It is probably best to follow the ancient grammarians and write
ἀσφάδαιστος; cf. A. C. Pearson on Soph. fr. 349 and 848, Schwyzer, Griech.
Gramm. i. 265 f.
εὐθνήσιμος (on the derivation cf. C. Arbenz, Die Adjektive auf -wos, Diss.
Ziirich 1933, 84) here alone; this, as it seems, is the earliest express reference
to ‘euthanasia’; edfavdrws occurs in Cratinus fr. 413 K.
The resemblance of this prayer to that of Ajax preparing for his death
(S. Aj. 832 ff.) has of course long been observed. That Sophocles really had
this passage in mind is proved by the fact that ἀσφάδαιστος occurs nowhere
else,’ and must thus have been extremely rare, even if we allow for losses
(it is curious that it is missing from the list of Aeschylean words in the Ajax
compiled by Jebb, Introd. lii). Cf. on 1323.
1296. μακρὰν Erewas: cf. on 916. In contrast with most interpreters, Verrall
rightly holds that the tone of the passage needs explanation: ‘ "You have
by long speaking deferred your fate for some time, it is true, but if you really
foresee it, why go to it at all?" This is the tone.’ This cannot be right, if
only because the decisive thought ‘deferred your fate for some time’ is
interpolated. After his expression of sympathy and wonder the coryphaeus
voices his discomfort, albeit only in the form of a statement of fact. μακράν
ye μὲν δὴ ῥῆσιν od στέργει πόλις (Suppl. 273); this dislike is not confined to
Argos. It is constantly felt necessary to give a special excuse for a ‘long
drawn out’ speech; so Artabanus (Hdt. 7. 51. 1) ἀναγκαίως γὰρ ἔχει περὶ
πολλῶν πρηγμάτων πλεῦνα λόγον ἐκτεῖναι or the old Oedipus (5. Oed. C. 1119 f.)
ὦ ξεῖνε, μὴ θαύμαζε πρὸς τὸ λιπαρές, τέκν᾽ εἰ φανέντ᾽ ἄελπτα μηκύνω λόγον (cf.
the reply of Theseus 1139 f.). In accordance with their general disposition,
the old men are really most painfully moved by the content of the speech,
with its terrible outburst of despair and its repeated clear prophecies of the
double deed of blood. So instead of saying, as he said several times before,
something like ‘What you have said is terrible’, the coryphaeus veils his
emotion and confines himself to the remark μακρὰν érewas. In just the same
way Agamemnon, who found the speech of his wife by no means to his liking,
says μακρὰν... ἐξέτεινας (916), though admittedly the effect is there softened
by the mild jest of the preceding line (915). What follows gives clear expres-
sion to the mood of the old men. They are still wishing that all they have
heard were not true, but now they scarcely dare to hope it (ei δ᾽ ἐτητύμως
«rA.). But then if Cassandra really has foreknowledge of her end, why does
she voluntarily hasten it on? The greatness of a noble soul baffles the under-
standing of the ordinary man. It is this contrast which gives its tension to
the stichomythia which follows.
1297. αὑτῆς: cf. on 1141.
1297 f. ‘ θεήλατον βοῦν dicit, quae se ultro offert ad immolandum; quales
memorant historici plus semel’, Stanley, rightly. Commentators cite material
from Roman sources, which is only indirectly relevant to this context; they
found the Latin passages ready to hand in the rich collections of Brissonius,
De formulis, i. 22 ff. He also cites Plutarch, Quaest. conviv. 8. 8. 3 (Mor.
p. 729 ἢ ἄχρι δὲ νῦν παραφυλάττουσιν ἰσχυρῶς τὸ μὴ σφάττειν πρὶν ἐπινεῦσαι
1 Wilamowitz’s conjecture on A. Suppl. 698 is quite uncertain: cf., besides the editions,
Pohlenz, Gött. gel. Anz. 1930, 448. For Sophocles (fr. 349 P.) we now have evidence for a
word not elsewhere attested, νεοσφάδαιστος.
603
lines 1297 f. COMMENTARY

κατασπενδόμενον and the oracle in Porphyrius, De abstin. 2. 9 ὃ δ᾽ ἑκούσιον ἂν

κατανεύσηι χέρνιβ᾽ Τἐπιθύειν τὸ δ᾽, émiokore,] φημὶ δικαίως. These passages are
evidence, not for the special importance of voluntary approach to the altar,
but for the general principle that the acquiescence (or what could be inter-
preted as such) of the victim was desired; cf. in addition Ziehen, RE xviii.
612, Meuli in Phyllobolia fiir Peter Von der Mühll (Basel 1946), 266 f. (for the
Greeks) and 254 f. (for ‘primitive’ tribes). Blaydes quotes Philostratus Her.
Ῥ. 329 (vol. II p. 215. 2 Kayser), where it is said of the cult of Achilles on
White Island τὸ δὲ ἱερεῖον αὐτόματον τῶι βωμῶι προσέστηκε.
1299. On the repetition of an οὐ by a following ‘ οὔ solitarium’ cf, Matthiä,
Griech. Gramm. ὃ 608. 6; I. Bekker, Hom. Blätter, i. 88: W. Schulze, Kl.
Schr. 700. Perhaps we can now add to the examples known earlier one from
Aeschylus himself (@ewpoi), Pap. Oxy. 2162, fr. 1, col. 1. 24 οὔτοὐτέρωσ᾽, οὔ,
δῆλος ἦσθ᾽ ὁδοιπορῶν], though admittedly there is some difficulty about the
beginning of the line (cf. Lobel ad loc.). The emphasis expressed by the
repetition (‘graviter repetita negatione’, Hermann) is unmistakable. The
phrase οὐκ ἔστ᾽ ἄλυξις is formed on the Homeric model: X 270 οὔ τοι ἔτ᾽ ἔσθ᾽
ὑπάλυξις, ἄφαρ δέ ae Παλλὰς ‘Abin ἔγχει ἐμῶι δαμάαι, cf. © 565 οὐκέτ᾽ ἔπειτ᾽
ἔσται θάνατον καὶ κῆρας ἀλύξαι.
χρόνωι πλέω. A satisfactory explanation or restoration has so far not been
put forward. We must from the outset exclude such interpretations as that
of Klausen : ‘ οὐ χρόνωι πλέω, scil. ἐστί, quod modo dictum in οὐκ ἔστ᾽ ἄλυξις.
Nihil lucramur eo, quod tempus lucramur’ (similarly Nägelsbach: ‘Es gibt
kein Entrinnen; Säumniss, Freunde, fruchtet nichts’); for in them, apart
from other considerations, the nature of the repeated οὔ (see above) has been
completely mistaken.! Hermann's χρόνον πλέω, which presumably is sup-
posed to mean ‘any longer’, ‘is objectionable as not giving that force to the
word χρόνος which the reply of the Chorus shows that it must have had in
Cassandra’s speech’ (Conington), but, above all, such an addition (‘non est
fuga diutius’) is an intolerable weakening of the unqualified statement οὐκ
ἔστ᾽ aAv£ıs ; cf. the Homeric passages cited above and 381 οὐ γάρ ἐστιν ἔπαλξις,
467 οὔτις ἀλκά, etc. This objection would be disposed of if the words could
mean ‘since (after) the time has been fulfilled’. But first, it is difficult to see
how that could be fitted into the sentence by any appropriate construction
(it is impossible to take seriously Verrall’s ‘absolute dative’; Weil felt the
difficulty and therefore, in the Addenda to his edition of Eum., 1861, p.
138, proposed an independent χρόνοι πλέω!) ; secondly, and this is decisive,
“mAdws χρόνος graece non videtur dici’, as Wilamowitz remarks. So far
as I can discover, this is true.^ Wilamowitz himself reads, as Wecklein
! The same applies to Kumaniecki, De elocut. Aesch. 48, who, contrary to both the explicit
indications in Wilamowitz’s app. crit., translates his text (with χρόνου πλέον at the end)
thus: ‘Nulla est salus; deest tempus, amici.’
2 Typical of the way in which some scholars bemuse themselves and their readers with
‘parallels’ is Blaydes’s reference: ‘Cf. S. Johann. Ev. 7. 8 ὅτι 6 ἐμὸς καιρὸς [nota bene] οὕπω
πεπλήρωται ᾽ (he might also have cited Ev. Marc. 1. 15). The modern reader who interprets
Ag. 1299 * ὅταν πληρωθῆι 6 χρόνος ' (Murray) is not conscious that he is misled by Biblical
recollections. To form a picture of the prevalent use of πληροῦν with χρόνος and kindred
expressions of time as object, we should add to L-S s.v. IIT. 3 the richer collection in W. ,
Bauer, Wörterb. zum Neuen Test. s.v. πληρόω 2. As in Hdt. 6. 63. 1 τοὺς δέκα μῆνας, so in
Pl. Laws 866 a τοὺς χρόνους... πληροῦν... τοὺς εἰρημένους, Tim. 39 d τὸν τέλεον ἐνιαυτόν, we
have to do with periods precisely fixed by date and capable of being expressed in numbers ;

604
 COMMENTARY line 1300

did,’ χρόνου πλέον (he says of it ‘corr. Voss’, but Gerardus Vossius left χρόνωι,
thus reading, like Pauw and many others after him, χρόνωι πλέον) ; he refers to
Eum. 163 δίκας πλέον, from which we have apparently to infer that he takes
the passage to mean ‘beyond the time there is no escape’. This would presup-
pose a meaning of χρόνος (without qualifying epithet or article) which I have
not been able to verify (‘the allotted span of time’ or ‘the fixed point of time’).?
The same objection applies, mutatis mutandis, to the proposal of Schiitz οὐκ
éar' ἄλυξις . . . χρόνου πλέων (translated by him ‘non magis est auxilii ac
salutis quam temporis ad elabendum copia’). I do not know what to do with
the passage except to recognize a corruption in χρόνωι πλέω. The presence of
τοῦ χρόνου in the next verse by no means proves that a form of χρόνος must
have preceded it. In itself the thought ‘there is no escape (why then should
I not offer myself willingly for sacrifice ?—cf. 1296 ff.)’ is a perfectly adequate
basis for the further thought: ‘but the last remnant of the time which one
has at one's disposal is valued most'.
1300. Here again we have an instance of the fatal effects of Stanley's mis-
takes. He paraphrased 'at qui postremus moritur, vel ipso tempore alios
superat' (on the other alternative which he considered see below). This
rendering, with unimportant variations, was accepted by many commentators.
So Heath: 'at qui postremus est, ipsam temporis moram praemium aufert',
similarly Schütz, Humboldt, Bothe, Hermann, Schneidewin, etc., many
English scholars, e.g. Blomfield, Peile, Conington, Paley, Sidgwick, and even
L-S, s.v. πρεσβεύω I. 2 (‘is first in point of time’). Klausen based a supposed
play of words on this interpretation, and was followed by Sidgwick, while
Paley saw an allusion to the lot which those condemned to death had to
draw. All this falls to the ground along with the interpretation. The truth
was recognized by Car. Guil. Elberling, Observ. in aliquot locos Agam. Aesch.
(Hauniae 1828), 25, who saw that ὁ ὕστατος τοῦ χρόνου belong together.
Hermann argued against this,? recalling the fact that Stanley had apparently
considered it along with the other interpretation (Stanley's words, 'vel tem-
pus est antiquissimum, id est, ἀξιώτατόν ἐστι τιμῆς᾽, are not quite clear).
the same applies to the documents, e.g. Pap. Oxy. 491 (A.D. 126), 7 ὅταν πληρώσηι rà εἴκοσι
ἔτη, Pap. Teb. 374 (A.D. 131), 10 ἧς ὁ χρόνος τῆς μισθώσεως ἐπληρόθη, etc., as well as to the
letter of Clement 1. 25. 2 πληρωθέντος τοῦ χρόνου (referring back to the preceding ἔτη mevra-
κόσια). These observations, then, confirm the correctness of Wilamowitz's remark. The use
outlined here corresponds exactly with that of complere tempus and the like, cf. Thes. I. Lat.
iii. 2094 f. Scholars who cite in connexion with Ag. 1299 Hesiod, Erga 778 ἤματος ἐκ πλείου,
192 πλέωι ἤματι may be asked to look at these passages once more.
1 Stud. z. Aesch. 140 (later he abandoned the view), with the translation: ‘there is no
escape that would be more than a postponement, would go beyond postponement'. That,
too, spoils the absolute (see above) οὐκ ἔστ᾽ ἄλυξις. Furthermore, it is not permissible to
translate unqualified χρόνος (and that in a text of the fifth century) as ‘postponement’; to
see this one has only to consult the section in Dindorf's Thesaurus viii. 1703 D-1704 A (a
short selection from it is given L-S xpóvos IV).
2 In S. Ant. 461 f. εἰ δὲ τοῦ χρόνου πρόσθεν θανοῦμαι ('antequam vitae tempus natura
praestitutum exierit! Ellendt-Genthe) the article provides exactly the determination which
is lacking in χρόνον πλέον.
3 Conington, too, declined to take ó ὕστατος τοῦ χρόνου together. In so doing this normally
very cautious scholar used an argument against which it may be well to say a word of
warning here, since it creeps in elsewhere also: 'It would be possible to join τοῦ χρόνου with
ὁ ὕστατος, but the rhythm is rather against it.' Against this cf. 1279 and in addition the
following random choice: Pers. 373, 477, Sept. 703, Prom. 484, Cho. 117, Eum. 276, 656.

605
line 1300 COMMENTARY

Elberling’s interpretation was adopted by Karsten, Enger (1863 in his re-

edition of Klausen), Nägelsbach, Kennedy, Hense, Wecklein, Wilamowitz,
Platt, Mazon, etc. Verrall, while preferring this view, nevertheless (so strong
is the power of traditional prejudice) declared the other interpretation to be
possible, although he rightly objected that it ‘introduces a comparison of
different persons, which is scarcely to the point’. It never occurred to him,
any more than to the others, to consider the scholion, although that had been
accessible in van Heusde’s edition since 1864 and more conveniently in
Wecklein’s since 1885. There we read (&xoA. mad. of Tricl.) τίμιός ἐστι map’
ἀνθρώποις, φησίν, ὁ ὕστατος χρόνος" ὅταν γὰρ πρὸς τὸ τέλος ἥκωσι, καὶ τὸ βραχὺ
ζῆσαι κέρδος ἡγοῦνται, which is clearly right and exactly what Elberling had
made out by himself. The expression ὁ ὕστατος τοῦ χρόνου is vouched for by
the passage which is set beside it in Passow’s dictionary s.v. χρόνος 2(a) and
in Dindorf's Thesaurus, viii. 1701, Xen. Rep. Lac. 1. 5 ἐπεί γε μὴν γυνὴ πρὸς
ἄνδρα ἔλθοι, ὁρῶν τοὺς ἄλλους τὸν πρῶτον τοῦ χρόνου ἀμέτρως ταῖς γυναιξὶ
συνόντας κτλ. Other instances of this construction (e.g. Eum. 428 ἥμισυς
λόγου, where the idiom is destroyed in the later MSS) can be found in Kühner-
Gerth, 1. 279 c, with the remark that ‘the construction is a genuine Attic one,
but more at home in prose than in the language of poetry’. In the passage of
Xenophon quoted the reference of τοῦ χρόνου is quite clear. What sense τοῦ
χρόνου (not χρόνου) must bear in Ag. 1300 may be seen from the paraphrase
given at the end of the note on 1299; cf., e.g., 5. Phil. 126 f. ἐάν μοι τοῦ χρόνου
(the time which one must allow for such an undertaking) δοκῆτέ τι κατα-
σχολάζειν κτλ.
With πρεσβεύεται Elberling has well compared Cho. 631 (‘is treated, re-
garded, as preeminent’); the paraphrase of the scholiast on Ag. 1300 (riuós
ἐστι παρ᾽ ἀνθρώποις) is quite precise.
1301. I do not understand why Wilamowitz, with Karsten, prefers the pre-
sent tense to κερδανῶ. κερδανῶ implies the thought ‘if, as you suggest, I try
to postpone my fate’. With the idea Stanley compares the words of Socrates
Plat. Phaed. 117 a οὐδὲν yap οἶμαι κερδανεῖν (ΒΖ, κερδαίνειν BT, unsuccessfully
defended by Stallbaum) ὀλίψον ὕστερον πιὼν ἄλλο ye ἢ γέλωτα ὀφλήσειν παρ᾽
ἐμαυτῶι, γλιχόμενος τοῦ ζῆν καὶ φειδόμενος οὐδενὸς ἔτι ἐνόντος.
1302-5. Allthat is necessary for the understanding of these lines in the order
in which they stand in the MSS has been said by Conington ;' his explanation
must therefore be quoted in full: τλήμων seems to imply resolution rather
than misery, though of course with the notion of the cause being a desperate
one, like τλήσομαι τὸ κατθανεῖν (1289): and dm’ εὐτόλμου φρενός is constructed
like dm’ ἄκρας φρενός (805) [cf. below on 1643], expressing the seat of the
1 It was at my request that Murray put in his apparatus the words ‘verum vidit Coning-
ton’; for I am convinced that the correct interpretation of a constantly mishandled passage
is worth more than many conjectures. It is a thousand pities that Wilamowitz either did
not know or disregarded Conington’s interpretation (he could have found it, e.g., in the later
editions of Paley). Conington says: ‘The words ἀλλ᾽ εὐκλεῶς τοι κατθανεῖν χάρις βροτῶι
could never have been put by him [Aesch.] into the mouth of Cassandra, who is as far as
possible from the common view of a glorious death, and indeed shows in her next speech
very plainly what feelings such a thought suggests to her.’ Wilamowitz felt that just as
strongly (‘non Polyxena loquitur aut Perpetua, sed Cassandra cui turpissima mors imminet
nec quisquam eius meminerit, cf. 1329"), but instead of leaving the verse to the coryphaeus,
to whom it is given in the MSS, he devised, under the spell of Heath’s transposition, the
conjecture εὐκόλως... κατθανεῖν. [For Dr. Arnold's contribution see vol. i, p. 51 n. 4.)
606
 COMMENTARY line 1306

feeling. The Chorus, finding that they cannot persuade Cassandra, attempt to
console her by their praise of her resolution. She, equally unwilling to avoid
her fate and to gloss over its misery, replies that this is not the way men
talk to the happy, implying that the very fact of consolation shows there is
wretchedness to be consoled. The Chorus, still wishing to comfort her,
argues that surely there is something in what men say when they talk of a
glorious death as a boon—Cassandra’s thoughts instantly fly to the death of
her father and his noble sons, as the experience she has had of what is called
so glorious, the death of the brave.”
1302. For the sense of τλήμων cf. Cho. 748 (τλημόνως), Cho. 384, 596 (there in
close connexion with ὑπέρτολμος and πάντολμος as here with εὔτολμος) ; in
both the last instances there is an undertone of censure, asin S. El. 275, 439;
on the other hand, in E. Heraclid. 570, Hec. 562, Suppl. 947 (τλημόνως), as
here, it expresses a positive quality. There is a good deal of truth in the
differentiation made by ancient grammarians: Schol. A on K 231 ὅτι τλήμονα
ol νεώτεροι τὸν ἀτυχῆ, ὁ δὲ “Ὅμηρος τὸν τλητικόν, τὸν ὑπομενητικόν (cf. Lehrs,
De Aristarchi stud. Hom., 3rd ed., 91), Schol. E. Or. 35 παρὰ μὲν τῶι ποιητῆι
τλήμων ὁ ὑπομονητικός, παρὰ δὲ τοῖς τραγικοῖς τλήμων ὁ δυστυχής, but it does
not apply without qualification. For the transition in the meaning from
‘patient, steadfast’ to ‘miserable’ cf. F. Jacoby, Hesperia, xiv, 1945, 204 n. 168.
1303. From what has been said above it follows that there is no longer room
for a rendering such as ‘in good fortune men are deaf to such teaching’
(Wilamowitz). What Cassandra means is this: the fortunate have no need of
τλῆναι and τολμᾶν; therefore one does not apply to them the corresponding
terms. ‘When we were happy we had other names’ (Shakespeare, King John,
v. 4. 8): this corresponds to the general thought, if not to the words, of the
Aeschylean passage. For ἀκούειν = ‘to have such and such things spoken of
one’ see the instances in Passow-Crönert s.v. ἀκούω I. 8. In Ar. Knights 820
οὔκουν ταυτὶ δεινὸν ἀκούειν, & Au’, ἐστίν u’ ὑπὸ τούτου; the addition of ὑπὸ
with the genitive is significant (as often with κακῶς ἀκούειν, etc.).
1304. The adverb εὐκλεῶς occurs in Aeschylus only here and Pers. 328
εὐκλεῶς ἀπώλετο. This usage is in conformity with that of Homer, where both
ἐυκλειῶς (X 110 ὀλέσθαι ἐυκλειῶς) and ἀκλειῶς (X 304, a 241 = ξ 371) are found
only in relation to death.
1305. ἰὼ πάτερ: exceedingly strong (cf. on 503), not a mere appeal but a
shriek. Conington has explained the connexion with the previous line (see
above). She is overpowered by the contrast between the usual view, that
εὐκλεῶς κατθανεῖν is a χάρις, and the stark reality of the end suffered by the
slain, as she sees it before her eyes, she who has witnessed the sack of Troy.
On 1305 Hermann rightly notes: ‘Haec ubi dixit Cassandra, pergens versus
aedes regias subito retinet gressum atque avertit se.’ The scholion on 1307
runs (ZyoÀ. ra. of Tricl.) : ἀπολοφυρομένη λέγει τοῦτο ἐν τῶι εἰσιέναι. ὀκνεῖ γὰρ
εἰσελθεῖν ὥς τι δεινὸν ὁρῶσα.
1306. τί δ᾽ ἐστὶ χρῆμα: so Cho. 885; see also the similar phrases Prom. 298
ἔα τί χρῆμα; and Cho. 10 τί χρῆμα λεύσσω; Apart from these instances, the
singular of χρῆμα has so far been attested for Aeschylus (he often uses the
plural) in one fragment only, Pap. Oxy. 2160, fr. 7. 3, which perhaps' belongs
to the Γλαῦκος Ποτνιεύς: οἷον τὸ χρῆμα roure . . . . Unfortunately the loss of
1 Cf. Lobel's introduction to Pap. Oxy. 2160.
607
line 1306 COMMENTARY

the end of the line makes it impossible to say whether it ran τοῦτ᾽ &...or
with the genitive τοῦ re . . .; in the latter case we should have a variation of
the turn of speech familiar from comedy ὅσον τὸ χρῆμα with a genitive follow-
ing. So far as concerns questions with ri χρῆμα, we find them in Sophocles as
well as in the plays of Aeschylus’ last period: Ant. 1049 ri χρῆμα; Aj. 288 ri
χρῆμα δρᾶις; Oed. R. 1129 ri χρῆμα δρῶντα; Phil. 1231 τί χρῆμα δράσεις ; and in
Euripides: Hipp. 909 τί χρῆμα πάσχει; he often has ἔα, ri χρῆμα; and several
times phrases such as Alc. 512 τί χρῆμα ("why?') κουρᾶι τῆιδε πενθίμωι πρέπεις ;
cf. P. T. Stevens, C.Q. xxxi, 1937, 190 f., and, for the meaning of χρῆμα in
these expressions, see Vendryes, Melanges Desrousseaux (Paris 1937), 477. It
is obvious that phrases of this type come from the speech of ordinary life,
and this is confirmed by comedy: Ar. Clouds 325 ri τὸ χρῆμα; 816 τί χρῆμα
πάσχεις; Wasps 834 τί more τὸ χρῆμα; We can hardly doubt that such turns of
expression were current as early as the time of Aeschylus’ youth; but so far
as our fragmentary tradition permits a judgement, for a long time he ex-
cluded them completely or all but completely from the strict style of his
tragedy. The same conclusion may apply to a number of linguistic peculiari-
ties to be found in the Prometheus! and the Oresteia. On ri xpeos ; equivalent
in meaning to ri χρῆμα, but less colloquial, see on 85.
1308. ἔφευξας : the verb is not attested elsewhere. We do not know whether
the Atticist Aelius Dionysius (fr. 332 Schwabe, p. 220) ap. Eustath. on X 447
p. 1279. 36 (φεύζειν τὸ φεῦ λέγειν) knew of other instances; for ὥζειν (also in
Ar. Wasps 1526) he referred to A. Eum. 124.
εἴ τι un: regular word-order,? for which E. Bruhn, Anhang zu Soph. 94,
quotes, besides the present passage, the following examples: Prom. 196,
5. Tr. 586, 712, Oed. R. 124, 969. In each case εἰ is followed by a verb. There-
fore Wecklein’s conjecture στύγει is very improbable. στύγος is here predica-
tive; for the so-called ellipse of the copula in ei-clauses cf., e.g., Suppl. 960
εἰ δέ τις μείζων χάρις, Prom. 763 ei μή τις βλάβη (very close to our passage),
816 τῶνδ᾽ ei τί σοι ψελλόν τε καὶ δυσεύρετον. Wilamowitz alters φρενῶν to φρενί,
wrongly: Plüss well compares 5. El. 1390 τοὐμὸν φρενῶν ὄνειρον. ‘A thing of
horror in your mind (your imagination)’. In spite of the deep impression
which Cassandra’s words have made on them, the old men, incapable of
grasping anything beyond what their senses perceive (1310), completely fail
to follow the course of her feelings and thoughts. In a milder form, with
their φρενῶν στύγος they take up again their φρενομανής τις el (1140), blind to
the reality and inevitability of the fated happenings.
1309. Verrall’s defence of φόβον (between 1306 and 1316 an intelligible error)
is of a piece with his retention of τῶν in 1305.
1310. καὶ πῶς : almost equivalent to an οὐδαμῶς ; cf. on 549 and on 1507.
θυμάτων ἐφεστίων : mentioned in 1056 f.
1311. With the construction ὁμοῖος ὥσπερ Blomfield compares E. Or. 697
ὁμοῖον ὥστε πῦρ κατασβέσαι λάβρον (cf. also S. Ant. 586), and Passow quotes
1 See the very helpful remark of Wilamowitz, Griech. Tragödien, xiv (‘Die griech.
Tragödie u. ihre drei Dichter’) 81: ‘in the long speeches of the Prometheus Aeschylus ad-
mitted certain forms of everyday speech which as yet had been avoided elsewhere [in
tragedy]’.
2 Wackernagel, Idg. Forsch. i, 1891, 367 ff., has set it against its background in the history
of the Indo-European languages (note specially p. 369 on the regularity of af τις etc. in
Doric).
608
 COMMENTARY line 1316
Xen. Symp. 4. 37 ὅμοια γάρ μοι δοκοῦσι πάσχειν ὥσπερ εἴ τις μηδέποτε ἐμπίμ-
πλαιτο.

πρέπει: cf. on 321.

1312. Schol (ZyoÀ. wad.): τὸν ἀπὸ Συρίας λιβανωτόν.----ὃ λέγεις, φησίν, οὐκ
ἔστι θυμίαμα, νῦν δὲ τεθυμίαται ὁ δόμος.
E. Bacch. 144 Συρίας δ᾽ ὡς MBávov καπνός was compared by Spanheim.
ἀγλάισμα first attested for us in the Oresteia (cf. Cho. 193), then in Sophocles
and Euripides.
δώμασιν is dependent on ἀγλάΐσμα, just as dyAailew has the dative of the
recipient in the cult song (Semos ap. Athen. 14. 622° = carm. pop. 48 D.)
σοί, Βάκχε, τάνδε μοῦσαν ἀγλαΐζομεν and in Theocritus, Epigr. 1.4. For datives
of this type cf. Kuhner-Gerth, i. 426 f.
The scholion just quoted understood the line rightly ; the point is that the
house is really pervaded by the smell of incense. The coryphaeus means that
Cassandra’s words are contradicted by the obvious fact. Headlam translates
‘no Syrian odour, truly, that you give! the house!’ and explains ' οὐ Σύριον
must be taken closely together κατ᾽ εἰρωνείαν : Ar. Frogs 1150 πίνεις οἶνον οὐκ
ἀνθοσμίαν ', etc. This is to mistake the function of od set at the head of the
sentence; it is the same as, e.g., 290, Sept. 717 οὐκ ἀνδρ᾽ ὁπλίτην τοῦτο χρὴ
στέργειν ἔπος (cf. Wilamowitz ad loc. and on Ag. 1634). On this word-order
cf. Leaf on © 164, Wilamowitz on E. Hipp. 390 (p. 204), where among other
instances he quotes the particularly striking passage E. Cycl. 537 ἠλίθιος
ὅστις μὴ πιὼν κῶμον φιλεῖ, Wilamowitz, Pindaros, 276 n. 3 (on Nem. 3. 15
οὐκ ἐλεγχέεσσιν Apıorordeidas τεὰν ἐμίανε κατ᾽ αἶσαν), the commentators on
Catullus 68. 39, Housman on Manilius 2. 41, on Lucan 1.145, and in C. Q. xxviii,
1934, 130 f. (it would be easy to add to the instances?). For the very old habit
of the Indo-European languages which is at the root of this word-order cf.
Wackernagel, Syníax, ii. 259 ff.
1313 f. Several editors have attempted to remove the two lines from this
place and put them somewhere else ; in doing so they show themselves blind
to the action on the stage. Urged by her own fate and no longer held back by
the desire to communicate her thoughts to the old men, who despite all her
efforts do not understand her, Cassandra now breaks off and once again (cf.
on 1305) moves towards the door, but then (1315) halts once again.
1314. ἀρκείτω βίος. The sense is rightly given by Stanley: hactenus vixisse
satis esto. Blaydes quotes the lines from a chorus of Euripides' (fr. 791 N.)
Philoctetes: ἅλις, ὦ βιοτά" πέραινε, πρίν τινα συντυχίαν ἢ κτεάτεσσιν ἐμοῖς ἢ
σώματι τῶιδε γενέσθαι. But the feeling there is widely different, cf. Ter. Eun.
(Menander) 551 f. nunc est profecto interfici quom perpeti me possum, ne hoc
gaudium contaminet vita aegritudine aliqua.
1316. οὔτοι δυσοίζω x7A.: in close relation to the preceding interjection; cf.
26. Cassandra explains her demeanour.
δυσοίζω : etymology and meaning should, one would think, be clear as the
1 This way of taking δώμασιν λέγεις is old and widespread, e.g. Humboldt : “Du ruhmest
nicht dem Hause Reize Syriens’, similarly Hartung and Nagelsbach, Kennedy: “Thou
claimest for the house no Syrian luxury.’ I know of nothing to justify the attribution of
this sense to λέγειν with the dative.
2 I cite one only, because of its apparent harshness: Servius Sulpicius (Cic. ad fam. 4. 5.
2) aul qui non in illis rebus exercitatus animus callere iam debet (i.e. callere iam non debet)
alque omnia minoris existimare?
4872-3 K 609
line 1316 COMMENTARY

day. Apollonius Dyscolus (Gramm. Graeci, i. 128. 7 Schneider), it is true, does

not produce any evidence to support his statement τάχα δὲ καὶ παρὰ τὸ oloi τὸ
οἴζω ἐστί." But the formation speaks for itself ; cf., e.g., 1308 &dev£as in relation
to φεῦ. Schwyzer, Griech. Gramm. i. 716, points out that ‘-£w as a mode of
forming verbs from interjections which end in vowels, etc.’ appears in litera-
ture for the most part from Aeschylus onwards ;? like A. Debrunner, Griech.
Wortbildungslehre, 117, he assigns δυσ-οίζω to this group, adopting the ety-
mology of Apoll. Dysc. For the compounding of such a verb of sound with
övo- (it intensifies the meaning of distress and outcry at disaster) cf. another
word attested only in Aeschylus, δυσβάυκτος Pers. 574 (Bailw from Bav belongs
to the same group of verbs). Lobeck's protest against the derivation from
ot, PHMATIKON 82, has little weight, since, as is shown by his words
“Svooilw ne significationem quidem lamentandi habet sed potius mala
ominandi', the great grammarian was here entirely dependent on the para-
phrases (ὑπονοεῖν, φοβεῖσθαι, ὑποπτεύειν, οἰωνίζεσθαι) which he found in
Hesychius. Blomfield in his glossary on Ag. 1316 also started from Hesychius ;
his rendering metuo partially or entirely dominates the lexicons s.v. övooiLw?
and the majority of commentaries and translations. But just as Humboldt
in his day translated 'Nicht wie ums Gebüsch der Vogel, jammr' ich, furcht-
bewegt', so later a minority of editors clung to the meaning 'give a cry of
distress’, e.g. Peile (‘uttering cries of distress’), Karsten, Schneidewin, Nagels-
bach, Wecklein, Verrall, Mazon. This sense of δυσοίζω alone accords with the
context. The interpretation adopted by the majority—I mistrust’ (Paley) or
something of the kind—omits the most essential trait: when we think of a
frightened bird we first of all call to mind πικρᾶς ὄρνιθος ὀξὺν φθόγγον. This
meaning is also required here because it is precisely the cry of appeal, ἰὼ
ξένοι, which prompts the explanatory sentence. Moreover δυσοίζω, under-
stood in this sense, provides the support for what follows (1317 ff.) ; see below.
Apart from this passage, δυσοίζω is found only in Rhesus 724 and 805. The
latter verse is so grievously corrupt that I can make nothing of it (Val-
ckenaer conjectures that behind the strange middle δυσοίζου lurks a different
verb). But in the case of 724 we may at least consider whether the current
interpretation, derived from Hesychius, is better than, or indeed as good as,
the other, which we find in the glosses of Vaticanus gog (A in Schwartz),
βλασφημῶν. ἢ ὀργιζόμενος καὶ λοιδορῶν, regardless of whether that is derived
from the lexicographical tradition (like, e.g., the gloss on διόπων which follows
soon afterwards, on 741) or is merely inferred from the text of the passage.
It appears to me that δυσοίζων if taken to denote a cry is a better answer to
the question ri λάσκων than it is if understood 'suspecting' (Paley) or ‘being
1 οἴξζομαι is not certain at the beginning of a fragmentary line of Sophocles’ Inachus,
Pap. Tebt. (vol. iii, 1933) 692, col. 3. 16. The first editors (and also R. Pfeiffer, Stizgsber.
Bayer. Ak., Phil.-hist. Abt. 1938, Heft ii. 23 ff., who examined this part of the papyrus in
London) put dots under the first three letters and observe: ‘ οἴζομαι, if right . . . y seems to
be excluded as the third letter, though not £ ; ἔρξομαι, however, is unsatisfactory.' Accord-
ingly Pfeiffer's translation (p. 51) lament' is intended by him to be only tentative.
2 This fact is, however, scarcely enough to vindicate in the Florentine fragment of the
Niobe, 1. 7, the conjecture ἐπωάξζουσα in the sense ascribed to it by Kloesel, Hermes, Ixxii,
1937, 467 (derivation from wa or a corresponding interjection).
3 Passow still noted ‘eigentl. klaglich jammern’, and in the older editions of L-S there
was at least a reference to a possible connexion with o£, but in the latest edition even that
has disappeared.
610
 COMMENTARY line 1316

distressed’ (L-S). Unfortunately, the corruption in 725 makes the question

harder to decide, but one can at least recognize the trend of thought. The
rendering ‘ βλασφημεῖν ’ goes perhaps a little too far, but ‘uttering a cry of
shocked agitation’ would fit very well. Distress and indignation wring the
cry from the commander of the army. The translation ‘utter a cry of distress,
dismay”! fits, I believe, Ag. 1316 as well as Rhesus 724. It is possible that in
other places (Rhesus 805?) the verb was used in the sense indicated by
Hesychius. His note that the word was Laconic may perhaps be derived from
a commentary on the Rhesus: so Parmeniscus (schol. Rhes. 523) claimed
mporauwt as a λέξις Βοιωτική (cf. Wilamowitz, Berl. Klasstkertexte v. 2, p. 28).
The temptation to introduce Hesychius’ interpretation into Ag. 1316 was
presumably increased by the fact that the accusative θάμνον could thus
readily be made dependent on a transitive verb. But that is unnecessary.
An explanation possible in itself would be the following. ‘I do not raise a cry
of distress as the bird in terror of the bush’—(‘screams’ or whatever other
supplement the context suggests). Theoretically speaking, the complex
expression, consisting of φόβωι and some verb, not necessarily a transitive
one, could govern an accusative as well as φοβεῖσθαι and the like. Wecklein
compares Demosth. 4. 45 of δὲ σύμμαχοι τεθνᾶσι τῶι δέει τοὺς τοιούτους ἀποστό-
λους, to which may be added 19. 81 6 γε δῆμος ὁ τῶν Φωκέων οὕτω κακῶς ...
διάκειται ὥστε... . τεθνάναι τῶι φόβωι Θηβαίους καὶ τοὺς Φιλίππου ξένους. On
this the commentators observe (e.g. Rehdantz-Blass, Demosth. Phil. Reden,
ii. 2, Indices, p. 40; H. Weil, Les Harangues de Démosthène, 103) that we have
here ‘a bold construction copied from the tragedians’. However, in Ag. 1316
the context demands that the comma which F and Tr put before φόβωι be
retained ; θάμνον ὡς ὄρνις is a parenthesis without a verb, in accordance with
the purely allusive character of such proverbial comparisons (see below).
Now for the content of the comparison. It is not enough to lay it down, as
Wilamowitz does (Interpret. 174 n. 1), that on any summer evening we can ob-
serve a bird twittering in alarm in front of a bush. We should remember the
realism of Aeschylean images and comparisons and, further, the momentous-
ness of this passage : Cassandra is in the act of going into the house, where, as
she knows, there awaits her murder contrived with insidious craft. We shall,
then, not be inclined to see in the fear of the bird merely something vague,
but shall rather suppose that it points to a definite danger (or dangers) which
the little creature has grounds for fearing, a danger which can naturally be
coupled in Cassandra’s thoughts with that which threatens herself. In the
scholia (ZxoA. παλ) we find the following explanation: où δυσχεραίνω, φησίν,
ὡς ὄρνις θέλουσα eis καλιὰν εἰσελθεῖν καὶ θῆρά τινα φοβουμένη. The solution here
given is possible, perhaps even satisfactory; at any rate the explanation (in-
cluding the reference to the nest, of which there is nothing in the text) shows
good observation of nature. ‘The bird stays, screaming, because her nest,
with the young, is in the bush: if it were not, she would not stay and scream,
but simply fiy off at once’ (Mrs. J. D. Beazley). We must, however, remember
that in the Attica of Aeschylus and Aristophanes, as in the England of
Shakespeare? (cf., e.g., Macbeth, ıv. ii. 34 f.) and in Italy to-day, far and away
1 Under the spell of the current view Housman, C.R, xxxix, 1925, 79 n. 2, said: “The
derivation of δυσοέξω is unknown, but its meaning shows that it is not from offw.’
2 And a good deal later. G. M. Trevelyan says in his survey of life in England at the
611
line 1316 COMMENTARY

the worst enemy, at least of all the many birds which can be served up as
‘uccellini di campagna’, is man. If this is what we must think of here, then
θάμνος gets its full value. van Heusde referred to x 468 f. ws δ᾽ ὅτ᾽ dv à κίχλαι
τανυσίπτεροι ἠὲ πέλειαι ἕρκει ἐνιπλήξωσι, τό θ᾽ ἑστήκηι ἐνὶ θάμνωι κτλ. Naturally,
bushes and undergrowth are the place indicated for nets as well as for liming,
cf., e.g., Palladius 13. 6 tempore hoc (December) per humiles silvas et bacis
fecunda virgulta ad turdos et ceteras aves capiendas laqueos expedire conveniet.
We have, then, to think in the first place, if not exclusively, of the devices of
ὀρνιθευταί (snares, nets, and bird-lime),! as did le Père Brumoy, who trans-
lated ‘comme l'oiseau qui pressent le piege’, Schütz in his paraphrase (note on
1320 ἐπιξενοῦμαι krÀ.): non tamen ego, tamquam avicula circa virgultum (in
quo scil. aucupis insidias metuit)', and then English commentators in the first
half of the nineteenth century (first, apparently, T. Medwin in the notes? to
his translation of the Agamemnon, London 1832). R. Shilleto (quoted by
Paley in his first edition, 1845) in his commentary on Demosthenes, De falsa
leg. (Cambridge 1844) supported this interpretation by referring to two pas-
sages of Shakespeare: the pithy maxim in the Rape of Lucrece 88 ‘Birds never
limed no secret bushes fear' and its truly tragic elaboration in 3 Henry VI,
V. vi. 13 ff. There the king, imprisoned in the Tower and left alone with
Gloucester, his murderer, says to him: “The bird that hath been limed in a
bush, With trembling wings misdoubteth every bush; And I, the hapless
mate to one sweet bird, Have now the fateful object in my eye, Where my
poor young was limed, was caught and kill'd.'
In all probability, θάμνον ὡς: ὄρνις is a proverbial turn of speech (C. G.
Haupt). This is suggested also by its linguistic form. The absence of a verb
is typical of a group of proverbs, e.g. (I select only a few instances with
animals as subject) βοῦς ἐν αὐλίωι (Cratinus fr. 32 K.), eis πάγας ὁ λύκος, ὄνος
eis ἄχυρα, ὄϊς τὴν μάχαιραν, 6 veBpôs τὸν λέοντα (schol. in Lucianum ed. Rabe,
259 sq.), ὄνος Avpas.3 From other spheres are taken, e.g., Arist. Rhet. x. 6,
p. 1363*; καὶ ἡ παροιμία δέ, τὸ ἐπὶ θύραις τὴν ὑδρίαν, as well as the trimeter
(explicitly cited by Pollux 9. 120 as ἃ παροιμία) ἥλωι τὸν ἧλον, παττάλωι τὸν
πάτταλον. Sometimes such a turn of speech is fitted into dialogue with the
help of a ὡς (ὥσπερ) asin Ag. 1316, so that in the place of the allusive proverb,
in which a person is symbolically represented by an animal (ὁ veßpös τὸν
λέοντα' ἐπὶ τῶν μὴ προσηκόντων), we havea full comparison. For this type a few
examples taken from tragedy may suffice ; in them, as in the present passage,
the verblessness typical of proverbs is preserved: E. Her. zog f. καί μ᾽
ἀφείλεθ᾽ ἡ τύχη, ὥσπερ πτερὸν πρὸς αἰθέρ᾽, ἡμέραι μιᾶι (cf. Paley and Wilamo-
witz ad loc.), ib. 869 ἀμπνοὰς δ᾽ οὐ σωφρονίζει, ταῦρος ws és eußoAnv.* Cf. E. Or.
beginning of the eighteenth century (England under Queen Anne, vol. i, Blenheim, p. 24):
* *Liming" by twigs, snaring and trapping birds of all kinds, not only pheasants and wild
duck but thrushes and fieldfares, had still a prominent place in manuals of The Gentleman's
Recreation.’
1 Oppian, Cyn. 1. 65 καὶ δολιχαὶ θώμιγγες ὑγρός τε μελίχροος ἰξός κτλ., and abundant
evidence in the notes of A. W. Mair (Loeb Library edition οὗ Oppian, 1928).
2 1 imagine it alludes to the manner of catching thrushes with limed twigs, common in
Greece and Italy at the present day.’
3 Expansions of this form of the proverb are found in Menander (fr. 527 K.) and elsewhere.
+ Neither Bergk (in Meineke's Fragmenta comicorum, ii. 1190) nor Kock noticed that this
(or the proverb on which it is based) is parodied by Aristophanes fr. 619 K. χωρεῖ ᾽πὶ γραμμὴν
λορδὸς ὡς εἰς ἐμβολήν.
612
 COMMENTARY line 1317
44 f. ποτὲ δὲ δεμνίων ἄπο πηδᾶι δρομαῖος, πῶλος ὧς ἀπὸ ζυγοῦ. I believe that
the same view should be taken οἱ S. Trach. 441f. as of the passages quoted
above from E. Her.: "ἔρωτι μέν νυν ὅστις ἀντανίσταται TÜKTNS ὅπως ἐς χεῖρας,
οὐ καλῶς φρονεῖ. Here we should put a comma before πύκτης and inter-
pret accordingly, for the usual way of taking the passage (‘ és χεῖρας with
dvravioraras ’ Jebb, so Radermacher) seems decidedly forced and makes the
structure of the sentence clumsy. Whether in the proverb which lay behind
Ag. 1316 θάμνον ὡς ὄρνις a verb of fearing, of avoiding, or something else was
to be supplied we cannot tell.
1317. ἀλλ᾽ ὡς «rA.: Hermann's interpretation of the MS reading as ἄλλως
produced so strong an effect that it forthwith became the accepted version.
It is certainly wrong. We need not regard it as an objection that ἄλλως =
μάτην (so, e.g., schol. Ÿ 144, schol. E. Hec. 626, etc.) is not attested in Aeschylus;
it occurs in Homer, Sophocles, Euripides, comedy, etc. But here it would
produce a completely wrong idea. At this juncture it cannot occur to
Cassandra to emphasize that it is not from vain or empty fear that she is
speaking (‘Non ego, ut avis virgultum, prae timore frustra metuo’, Hermann).
Long ago she has declared her fate to be unavoidable; and even the old men
have accepted it asa fact that she must die (1300 and especially 1304). Equally
absurd is the idea that the bird’s cries of distress before the bush are μάτην:
it has only too good grounds for fear. The misconception of δυσοίζω (together
with the slight corruption in 1317) has led to a mistaken view of the whole
passage. ἀλλά is in fact indispensable. The contrast οὔτοι... ἀλλά is found
elsewhere in Aeschylus (cf. Schadewaldt, Hermes, lxxi, 1936, 32) : Sept. 236 f.,
Eum. 48 ; furthermore we find οὔτοι... dein Suppl. 365 f. The present passage
as it stands cannot, however, be correct. For what would be the meaning of
ὡς davoven? (Hartung’s ‘as dying’ is doubly wrong: θανούσηι can mean only
‘when I am dead’. The explanation of Plüss is also impossible: ' ὡς θανούσηι:
in the sense that she has already been slain . . . ὅταν : the Elders are now called
upon to be witnesses of what has happened for a future contingency.’)
Wilamowitz does not say how he took his text (i.e. the text of the MSS) in
this regard, nor does his translation throw any light on the question. The
true reading was long ago restored by Joh. Casp. Orelli? by means of a very
slight alteration: ἀλλ᾽ &s . . . μαρτυρῆτε κτλ. This was adopted by Enger
among others (in his re-edition of Klausen) and recently by Mazon. The
corruption is scarcely worth wasting breath upon; anyone who likes may
compare Sepi. 237, ἀλλ᾽ ὡς πολίτας μὴ κακοσπλάγχνους τιθῆις, where the writer
of M wrongly wrote τιθεῖς (a legitimate form of the indicative for tragedy),
or S. Ant. 643 ὡς... ἀνταμύνωνται, where L has -ovraı, or Ant. 760 8. ὡς
θνήισκηι, where L has θνήισκει, or El. 889 f. &s . . . λέγηις, where later MSS
have λέγεις, or Phil. 636 ws . . . ὁρίζηι (ὁρίζει all MSS). The construction of the
sentence is admirable and needs no elucidation. But what is decisive is that
! Herwerden, whom Wecklein and Murray follow, interpolates the text from Athen. 3.
τοῦ b (= Eubulus fr. 75. 6 K.), where the easy corruption ὑπὸ has long been emended.
πῶλος ὑπὸ ζυγοῦ is supposed to mean νεοζυγὴς πῶλος. But the parallels quoted above make
it highly probable that in such a phrase some verb of motion should be supplied. With
πῶλος ἀπὸ ζυγοῦ we may understand either πηδᾶν or a more general expression such as ‘tries
to break away’.
2 In Philol. Beiträge aus der Schweiz (Zürich 1819), 206, edited by Döderlein and Bremi
(I have not been able to see a copy).
613
 line 1317 COMMENTARY

thus we obtain a chain of thought which is not only good in itself, but is also
in full accord with the ancient views about succour in an emergency,
‘Nothilfe’, and the process of securing atonement. Whoever wishes to under-
stand the background of the custom familiar to the poet and his audience
should read W. Schulze’s admirable article ‘Beiträge zur Wort- und Sitten-
geschichte II’, Berl. Sitzgsb. 1918, 481 ff.— Kl. Schr. 160 ff.; here I can only
take from it a few of the main points (cf. also above on 48 ff.). By old
German law ‘erhebt der durch unrechtmässige Gewalttat Angegriffene oder
Geschädigte... das Zeter- oder Mordgeschrei (clamor necessitatis, violentiae)
- ...Der Notruf... gilt den Nachbarn, den Bauern des Dorfes, den Bürgern der
Stadt und verpflichtet alle, die ihn hören, zur “Folge”... Die Hilfspflicht der
durch das “‘geriichte’’ herbeigezogenen ‘‘Schreimannen”’ (“Schreigenossen’’),
deren Versáumnis das Gesetz unter Strafe stellt, reicht über die Ergreifung
des Übeltäters hinaus bis in das anschliessende Gerichtsverfahren und
bekundet sich dort als Eideshilfe oder Zeugnispflicht’ (Kl. Schr. 163 f.). Only
if the 'gerüchte', the cry of distress, has been raised, can evidence of the deed
of violence be later laid before a court of law.' Schulze (Kl. Schr. 185) also
quotes (following Jacob Grimm) from a French statute of the year 1233: ‘St
puella dicit sibi fuisse violentiam illatam ab aliquo in iali loco ubt potuit
clamare et audiri ab aliquibus: si non clamaverit, non debel ei credi’, and sets
this beside the passage of Euripides (Tro. 998 ff.), where Hekabe says to
Helen: εἶεν" βίαι γὰρ παῖδα dis σ᾽ ἄγειν ἐμόν" τίς Σπαρτιατῶν ἤισθετ' ; 7) ποίαν
βοὴν ἀνωλολύξας ; κτλ. Since she uttered no 'gerüchte', Helen (so at least
argues Hekabe) has forfeited the right to claim that what she has suffered
at the hands of Paris should be regarded as injury. Attic tragedy in general
contributes much to our knowledge of this use: ‘there the motif has the
characteristics of real life, though . . . the remoteness from ordinary life is
increased by the solemn pomp of tragic diction’ (Schulze, 179 f., with numer-
ous examples). A. Suppl. 904 ἰὼ πόλεως dyot πρόμοι, δάμναμαι belongs here,
as does S. Oed. C. 822 id ξένοι κτλ. and many other passages from Tragedy.
Since Schulze was primarily concerned with the explanation of βοή and
βοηθόος, he had no reason to go into the connexion of the cry for help with
μαρτύρεσθαι, known from Aristophanes and the orators as well as from Tragedy
and Thucydides ; it is, however, obvious that they are closely connected. Cf.
e.g., Lysias 3. 15 ἦγον αὐτὸν βίαι, βοῶντα καὶ κεκραγότα καὶ μαρτυρόμενον. The
following passage of Antiphon (1. 29) may serve to elucidate what Cassandra
is saying : οἱ ἐπιβουλευόμενοι... ἐὰν μὲν δύνωνται καὶ φθάνωσι πρὶν ἀποθανεῖν, καὶ
φίλους καὶ ἀναγκαίους τοὺς σφετέρους ζαὐτῶν» καλοῦσι καὶ μαρτύρονται, καὶ
λέγουσιν αὐτοῖς ὑφ᾽ ὧν ἀπόλλυνται, καὶ ἐπισκήπτουσι τιμωρῆσαι σφίσιν αὐτοῖς
ἠδικημένοις.2
The sense of Cassandra’s words is thus quite clear. ‘I raise this grim cry of
woe (ἰὼ ξένοι, like Oedipus, see above), not from fear, but in order that after
1 Cf. also F. Wieacker, ‘Endoplorare’, Münchener Beitr. zur Papyrusforsch., 34. Heft
(1944), 129 ff., especially p. 156 f. on the development by which the ‘geriifte’, clamor, became
the opening stage of the action in court and the βοηθόοι became witnesses.
2 Related, though different in so far as it does not concern the victim of violence, is the
case of Orestes, who immediately after his deed secures a witness for the later proceedings
in court. In Cho. 983 ff. he has the robe which assisted in the murder of Agamemnon spread
out so that Helios may see it, ὡς ἂν παρῆι μοι μάρτυς ἐν δίκηι ποτέ, ὡς τόνδ᾽ ἐγὼ μετῆλθον
ἐνδίκως μόρον, τὸν μητρός.
614
 COMMENTARY line 1320

my death you may bear witness of this (this injury done me), at the time
when the guilty receive their punishment for my death and Agamemnon’s.’
Orestes (see the last footnote) assures himself of witnesses for the future pro-
ceedings in court: ὡς dy παρῆι μοι μάρτυς ἐν δίκην more. Cassandra does not
enter into the technical procedure to the same extent, and cannot do so,
because there will never be any actual legal proceedings against Clytemnestra
and Aegisthus. Thus arriving at a verdict and pronouncing it is fused with
the execution of the sentence, or, more exactly, the punishment is the expres-
sion of the fact that a trial has taken place and has issued in a verdict. This
poetic simplification has been consciously carried out by Aeschylus, as is
proved by 813 ff., where the trial with its technical details of voting according
to the custom of an Attic court of law results in the conquest, that is, the
punishment, of Troy.
1318 f. Striking expression is here given to the idea of the inevitability of
talio, as in so many other passages of Aeschylus; cf. Daube, 196 f.
1318. γυνὴ yuvaukôs : an old and widespread type of order; see 320. Cf., e.g.,
hymn. Hom. Merc. 154 μητέρα δ᾽ οὐκ ἄρ᾽ ἔληθε θεὰν θεός, Ar. Thesm. 538 f. ἵνα
διδαχθῆι γυνὴ γυναῖκας οὖσα μὴ κακῶς λέγειν TO λοιπόν (here the tendency to put
similar words next to each other leads to the separation of the words which
belong together, γυνὴ οὖσα ; such phenomena often occur), etc.
1319. δυσδάμαρ here alone; bolder is Suppl. 1064 γάμον Övodvopa, ‘marriage
with an evil man’.
1320. ἐπιξενοῦμαι. Here, too (cf. on 1316), many scholars have based their
explanations on Hesychius, with no very happy results; see, e.g., Stanley
(Addenda) and Blomfield, who, however, expresses doubt about the inter-
pretation found there. Hesychius emıfevovodar μαρτύρεσθαι, πορεύεσθαι.
Σοφοκλῆς Ἀχαιῶν συλλόγωι (fr. 149 N. = 146 P.) καὶ Αἰσχύλος Κρήσσαις (fr.
120 N.). On this gloss Pearson (on Soph. fr. 146) comments: ‘the inference to
be drawn is that in one of the passages cited ἐπιξενοῦσθαι was equivalent to
μαρτύρεσθαι, and in the other to πορεύεσθαι. This inference is perhaps not
fully justified, and all we can safely maintain is that from the context in those
two plays some ancient commentators drew the conclusion that the verb
there meant what the paraphrase preserved in Hesychius indicates. It would
be rash blindly to trust this interpretation. A salutary warning is provided
by the kindred gloss in Hesychius which precedes the one I have just quoted.
We may assume that the corrupt gloss ἐπιξενοδόκευμαι" ἐπιμαρτυροῦμαι has
been rightly altered by M. Schmidt to ἐπιξενοῦμαι" ξενοδοκοῦμαι, ἐπιμαρτύ-
ρομαι. If this is so, the second explanation is probably derived from a
scholion on Ag. 1320. In other words, an ancient interpreter rendered the
difficult verb ἐπιξενοῦμαι by ἐπιμαρτύρομαι probably! because he thought that
this meaning was suggested by the context of the passage and especially by
the μαρτυρεῖν in 1317. It is a well-known habit of ancient commentaries to
explain an obscure or difficult word with the help of a neighbouring word in
the text. The interpretation μαρτύρομαι or ἐπιμαρτύρομαι is obviously not true
to the proper sense of ἐπιξενοῦσθαι, since the chief element in the word, £evo-,
plays no part in that interpretation at all. Blomfield (followed by later
scholars) added a reference to the information provided by the lexicographers
1 Cf., however, Hesychius προξενεῖ" μαρτυρεῖ and Klaffenbach’s note on JG IX. 12, 1,
no. 138. 9.
615
line 1320 COMMENTARY

(Etym. M. 610. 44, cf. Apoll. Soph., Lex. Hom. p. 117. 25; Simonides fr. 51 D.,
Pind. fr. 311 Schr) on ξεινοδόκος — μάρτυς, although this has probably
nothing whatever to do with ἐπιξενοῦσθαι. It is due to the use made of
Hesychius by Blomfield that in most translations of our passage ἐπιξ. is
rendered either by protestor (Dindorf, Thes. and Lex. Aesch.) or a similar
expression, or else by 'I crave this witness of you' (Headlam), etc. In the
rendering in L-S, too, 'I appeal to thee in these matters', no attempt is made
to take the £evo- element into account. What is suggested by the context
here, especially after the strong exclamation iw ξένοι, has been from time to
time expressed by scholars who did not allow their judgement to be fettered
by Hesychius, e.g. le Pére Brumoy, 'c'est le présent d'hospitalité que je
demande en mourant', Passow's lexicon, 'sich als Gastgeschenk nehmen',
Nägelsbach, ‘for this guest-present a bride of death implores you’, cf. Weck-
lein, Verrall, Wilamowitz. (Schneidewin's note ends in a compromise:
‘ ἐπιξενοῦμαι, μαρτύρομαι, I claim this witness as the duty of the ξένος to the
ξένη.) However, derivation from ξένια is improbable; and, while a general
connexion with £évos may be recognized, the details are not sufficiently clear.
There is also some difficulty as to the connexion of this passage with later
attested meanings! of ἐπιξενοῦμαι (besides the lexica cf. Wilamowitz on E.
Her. 965). Thus we must content ourselves with guessing the sense approxi-
mately.
δέ in the third place: cf. on 653.
1321. Here alone does Aeschylus, who often has θέσφατον, employ θέσφατος
as an adjective, a use which could easily be developed from that found in
Homer, even if we disregard the singular θέσφατος ἀήρ ( 143, cf. on it Schwyzer,
Glotta, xii, 1923, 10). With the utmost brevity it is made perfectly clear that
the old men's pity is roused not only by the doom of death in itself, but by
the fact that Cassandra herself must foresee and foretell it. There is a clear
reminiscence of the θέσφατα of which the Chorus repeatedly spoke in the first
part of this scene (1113, 1130, 1132).
1322. Hermann's criticism of the MS reading (‘tam languidam ac plane
stultam verborum consociationem') is unsympathetic, his alteration, οὐ
θρῆνον (adopted by Mazon among many others), bad. I pass over the wild
proposals of other editors. The text has been adequately explained by
Wecklein, and recently by G. Thomson. Scholars have objected to ῥῆσιν
without sufficient grounds, e.g. Sidgwick: ' ῥῆσιν is an unlikely word’, more
bitingly Housman, J. Phil. xvi, 1888, 278: ‘the totally inappropriate pijow ’,
taken up by Platt, C.R. 1897, 96, and Gow, C. Q. viii, 1914, 5. Cassandra says
‘I will utter one more speech’. Before, from 1299 on, she merely replied to the
questions and admonitions of the Chorus; then (1313) she prepared to depart,
but halted once again with the passionate outcry (1315) which she justified
and explained in the following five verses. All this is no speech. But now she
is quite composed, she can and will speak once again, for the last time. It is
well known that gous need not mean a lengthy speech; cf. Soph. fr. 61 N.

1 In the case of those later instances the details of the derivation of the word are again
far from clear. For the meaning ‘dwell abroad’ one might have recourse to ‘hypostasis’
(cf. on 105 ἐντελέων) from ἐπὶ ξένης (Xen. Lac. 14. 4, cf. S. Trach. 299 f. ἐπὶ ξένης χώρας,
Oed. C. 1256 ξένης ἐπὶ χθονός, 1705 γᾶς ἐπὶ ξένας) and possibly also to the influence of ἐπιδημεῖν,
but that is no help with émé. τινι, ‘to be someone's guest, guest-friend’.
616
 COMMENTARY line 1323

(= 64 P.). The arrangement of the sentence gives us the impression that

Cassandra at first intended to say only ἅπαξ ἔτ᾽ εἰπεῖν ῥῆσιν θέλω and then
becomes aware that the ῥῆσις could in this case be regarded as a θρῆνος;
accordingly she inserts the supplementary idea. Her death is to her quite
certain, it is as good as accomplished ; all that she still has to say applies to
one dead. Therefore her words could also be reckoned a θρῆνος, the more so
as they culminate not indeed in an individual but in a general lament (see
below). The prayer for vengeance (1323 ff.) also justifies the application of
the term θρῆνος to this utterance of hers. This is well pointed out by Eugen
Reiner, ‘Die rituelle Totenklage der Griechen’, Tübinger Beitr. z. Altertumsw.,
30. Heft, Stuttgart 1938, 21, where he stresses the point that from very ancient
times ‘announcements of vengeance formed part of the laments for the dead’.
That this motif of vengeance is common in the dirges of many peoples is
shown by Hedwig Jahnow, ‘Das hebraische Leichenlied im Rahmen der
Volkerdichtung’, Bethefte z. Zeitschr. f. d. alitestam. Wissensch. xxxvi, Giessen
1923, 54 f. By adding ἐμὸν τὸν αὐτῆς Cassandra points to the peculiarity of her
situation, for it is in itself contrary to the nature of things that anyone
should sing or speak his own dirge.! A similar paradox is stressed in the
passages compared by Wordsworth Z 500 ai μὲν ἔτι ζωὸν γόον “Ἕκτορα and
A. Suppl. 113 ff. ἰὴ in, ἰηλέμοισιν ἐμπρεπῇ ζῶσα γόοις με τιμῶ, on which Wila-
mowitz remarks: ‘in modum qui naeniae conveniat (ἰὴ ^j) viva me condecoro
lamentis' (on γόος, which approaches very close to θρῆνος, cf. on 57). Paley
cites Isocr. το. 27 (the Athenian youths and maidens sent to the Minotaur)
πενθουμένους ἔτι ζῶντας. In our passage there is perhaps also a suggestion of
the idea that no one will take on himself ἡδέα facere for the unfortunate
stranger. ἢ θρῆνον is not to be taken as a correction, in the way natural to
modern readers (‘seu potius threnum’ Bothe, ‘oder vielmehr ein Klagelied’
Wecklein, ‘or rather my own dirge’ Headlam), but offers the choice of another
possible designation (‘or, if that be more appropriate... .”). We find the same
thought-pattern formulated rather more fully in Cho. 1073 f. νῦν δ᾽ αὖ τρίτος
ἦλθέ ποθεν---οωτῆρ᾽ 1) μόρον εἴπω;
1323. It is wrong to substitute the genitive for ἡλίωι (so, after others, Weck-
lein, Wilamowitz, A. Y. Campbell, Murray, G. Thomson). The expression
᾿Ηλίωι δ᾽ ἐπεύχομαι is vigorous and natural, whereas the thought ‘by [?] the
last light of the sun I implore my future avengers’ (Wilamowitz) is lame and
laboured. Nor should we be deaf to the emphasis expressed by the initial
position of ᾿Ηλίωι. I will not use it as an argument that in the last section of
his dying monologue (857 f.) Ajax says καὶ τὸν διφρευτὴν “Ηλιον προσεννέπω,
πανύστατον δὴ κοὔποτ᾽ αὖθις ὕστερον, although with others I believe that
Sophocles here, too, had Cassandra’s words in mind (cf. on 1293). If editors
altered an expression which is unexceptionable in itself they did so because
they hoped that this would help them to diminish, if not to remove altogether,
the difficulties in the subsequent words. There we are faced with the puzzle
of three datives; it seemed to be an advantage if at least one of them, rois
ἐμοῖς τιμαόροις, could be given a construction by making it depend upon
ἐπεύχομαι. Here, as so often, instead of defining the limits of the corruption,
1 In different circumstances this idea can be used to provide a comic point: Plaut. Truc.
495 f. sine virtule argutum civem mihi habeam pro praefica, quae alios conlaudat, eapse sese
vero non polest.
617
line 1323 COMMENTARY

editors have tampered with the sound part of a passage, in the delusive hope
of being able thus to do something for the irremediable neighbouring parts.
“Ἥλιος ᾿ (as Nagelsbach rightly says) ‘invocatur ὁ πάντα λεύσσων, but
probably also because he, who is present, is the being most certainly in a
position to receive the prayer of the friendless foreigner whom her own gods
have deserted (1269-72).
1324. πρὸς ὕστατον φῶς. It is not quite easy to determine the function of
πρός. Pauw: ‘hac postrema luce’. But there are no instances in Aeschylus of
this use of πρός (L-S s.v. C. II), which occurs from Homer on. On the other
hand, the widespread πρός with the accusative ‘with all verbs of speaking,
when the speaker turns towards anyone’ (Kühner-Gerth, i. 519) is often
found in Aeschylus (Dindorf, Lex. Aesch. 310). Thus we must on the whole
agree with Verrall, ‘ ἡλίωι, . . . πρὸς ὕστατον φῶς are cumulative, one repeating
the other’, though probably the local function (cf. 1180 ἡλίου πρὸς ἀντολάς,
1182 πρὸς αὐγάς and elsewhere) contributes even more: ‘turning towards the
(for me) last light’.
1324 f. τοῖς ἐμοῖς τιμαόροις . . . ὁμοῦ. In the year 1872 Wecklein, Stud. z.
Aesch. 143 f., said of this passage: ‘there can be no idea of restoring the
incurably corrupt passage; up to the present all attempts at emendation
have been worthless.’ This diagnosis still holds good. The proposals pub-
lished in the interval, including those of Wilamowitz (1885), Housman,
Headlam, do not invite discussion of their details; they are too violent, even
when they do not operate with φόνευσις, invented ad hoc.! We cannot deny
the possibility that besides the alteration of some of the words some loss has
occurred; but Hermann has not succeeded in making that assumption
probable and has certainly not supported it by his attempt to introduce here
the gloss from Hesychius ἀσκεύοις" ψιλοῖς, ἀπαρασκεύοις. Αἰσχύλος Ἀγαμέμνονι;
cf. Nauck on the fragments of the Memnon of Aeschylus,? 127-9 (later, in his
edition, p. 535 on fr. 127, Wecklein made substantial reservations to his
unqualified agreement with Hermann’s insertion). Leaving aside the difh-
culties of the grammatical construction, the thing one misses most of all in
this sentence—which clearly culminates in 1326 δούλης davovons—is a refer-
ence to the vengeance for Agamemnon; this is why in 1324 Hermann read
βασιλέως τιμαόρους, Davies δεσποτῶν τιμαόρους, in 1325 Wilamowitz φονεῦσι
δεσποτῶν; but in the uncertainty of the surroundings such attempts have too
little warrant. We can only say with the greatest reserve that something like
the following suggests itself as a probable thought : ‘I pray that the murderers
when they give satisfaction to my avengers for the master may give it for the
slave as well.’
1326. εὐμαροῦς χειρώματος. Wilamowitz (on E. Her. 938), basing his view
on his conjecture χρέος (τίνειν), classes the genitive with μιᾶς χειρός, τῆς αὐτῆς
6800 and the like, and sees in it a use inherited from the ablative: it qualifies
the action denoted by the verb, viz., according to his view, the calling in of
the debt ; his translation, which is very free in detail, agrees with this. But

ı ‘Es konnte gezeigt werden, dass ein Wort wie φόνευσις im Munde des Aischylos geradezu
undenkbar ware’, wrote Wackernagel to me on 27 May, 1937, referring to C.R. li, 1937, 62.
Headlam, in a footnote to his translation, remarked that ‘ φόνευσις would be a strange
word for tragedy to use’. As a fresh expedient φόνευτρα was coined by C. Bonner, Class.
Philol. xxxvii, 1942, 267. 2 On fr. 130 cf, Wilamowitz (‘Testim.’) on Cho. 1068.
618
 COMMENTARY | line 1326
the extension of the use of such a genitive beyond a small group of special
cases is in itself questionable ; here, at any rate, where less remote possibilities
offer, it should be rejected. The interpretations advanced by other scholars
depend to a certain extent upon an arbitrary assumption about the sense of
χείρωμα. Apart from this passage, the word occurs only in 5. Oed. R. 560 and
[A.] Sept. 1022. Oed. R. 560 is perfectly clear: Adios... ἄφαντος ἔρρει θανασίμωι
χειρώματι, ‘disappeared through deadly overpowering’. That agrees exactly
with χειροῦσθαι (rarely χειροῦν) = ‘overpower’ (properly, ‘inferiorem reddere’,
derived from χείρων ; so Wackernagel, Kuhns Zettschr. xxx, 1890, 300, cf. also
Ernst Fraenkel, Griech. Denominativa, 89). In the present passage, too,
‘overpowering’ is the meaning of the word, as will be shown. On the other
hand, the author of the finale of the Septem used it with a shifting of the
meaning (‘durch Assoziation mit χείρ bewirkten Bedeutungsverschiebung’
Wackernagel), since in τυμβοχόα χειρώματα the noun can in fact mean only
what the paraphrase of the scholion says it does, crassly though this mis-
understands the rest of the passage: διὰ χειρῶν ἐργαζόμενα, ‘Hantierungen’
(Wecklein). There, then, the rare word χείρωμα has been misused to denote,
by means of a compressed nominal phrase reminiscent of Aeschylus’ manner,
the piling up of earth on the grave with the hand.’ In the present passage,
however, the commentators were not satisfied with the sense ‘an overpower-
ing’, but arbitrarily introduced ad hoc a different one, ‘res subacta’; so
Wellauer, Passow, Linwood (he sets alongside this, as do L-S, the more
correct rendering ‘a conquest’), Dindorf (Thes. and Lex. Aesch.), L-S. What
then is the construction? The doubtfulness of the preceding words makes it
impossible to say anything with certainty about the connexion with what
goes before; this entails the impossibility of coming to a final decision on the
syntactical function of edpapots χειρώματος. If the words are considered by
themselves, δούλης might be dependent on χειρώματος as an objective genitive,
or else edu. yep. might stand in apposition to δούλης θανούσης. If we adopt
the latter alternative, which I prefer, we should, of course, not take it as
implying that δούλης is picked up by χειρώματος, since that would necessitate
the wrong interpretation of χείρωμα and would further make εὐμαροῦς difficult
to understand. For we must not be taken in by such translations as ‘an easy
prey’ (L-S), ‘an easy victim’ (Headlam) etc. The predominant meaning of
εὐμαρής is ‘easy to do’. Pindar (three times) uses only εὐμαρές (with or with-
out ἦν) with a dependent infinitive, and the same construction occurs in
Sappho fr. 27. 5 D.; so, e.g., E. Alc. 492, fr. 176.2 N.; similarly Iph. A. 519 οὔκ,
ἣν θάνηι ye πρόσθε" τοῦτο δ᾽ εὐμαρές ; an equivalent construction is ἐν εὐμαρεῖ
with a dependent infinitive: Mel. 1227, Iph. A. 969. In the only other place
where the adjective (the adverb occurs in fr. 366) is found in Aeschylus, it is
combined, as in Ag. 1326, with a noun of action: Suppl. 338 καὶ δυστυχούντων
y’ εὐμαρὴς ἀπαλλαγή. For the very different sense in which Sophocles uses
the adjective (which occurs nowhere else in his works) in Ei. 179 χρόνος
γὰρ εὐμαρὴς θεός see Kaibel ad loc.” To return to εὐμαροῦς χειρώματος : an
! The diaskeuast was perhaps anxious to bring out this detail because in S. Ant. 429 he
read xai χερσὶν εὐθὺς διψίαν φέρει κόνιν,
2 I leave undecided the question whether Aeschylus, when he wrote εὐμαροῦς χειρώματος,
was aware of the derivation of the adjective from μάρη = χείρ, as Kaibel supposes (he
compares E. Bacch. 1128 εὐμάρειαν ἐπεδίδον χεροῖν).
619
line 1326 COMMENTARY
excellent sense is obtained by taking εὐμ. xeıp. as in apposition to the phrase
δούλης θανούσης as a whole: ‘the killing of the slave, a conquest easy to effect’.
In elevated language there would be no difficulty about saying δούλη ἔθανε
εὐμαρὲς χείρωμα, the accusative being in apposition to the actto verbi. Since
the place of the finite verb is taken by the genitive of the participle the
appositional phrase is also put into the genitive without any alteration of its
function. The connexion with what precedes remains, as has been said,
uncertain ; something like ‘requital for the killing of the slave’ is conceivable.
1328. The ἀντρέψειεν of the MSS (cf. Sept. 706) was interpreted by Porson as
av τρέψειεν. This is unexceptionable. The simple translation ‘turn, change’ is
completely adequate to the sense ; a fuller rendering would be the paraphrase
‘bring to pass a turn, a change, an alteration’. When we force attacking troops
which have been advancing upon us to turn about, that is τρέπειν ; the wind
which shifts is τροπαία (219 above) ; in the coward τρέπεται χρώς, and when
Agamemnon (Z 61) ἔτρεψεν ἀδελφειοῦ φρένας, he brings about a change of
heart in Menelaus. It is necessary to remind the reader of things so familiar
because τρέψειεν has been repeatedly doubted here or even declared im-
possible.’ But if we keep τρέψειεν, we are not at liberty to render it ‘overturn’
(so L-S s.v. V): that would be ἀνατρέπειν,2 and the only other instance
adduced for that meaning of τρέπειν, Aesch. fr. 311. 3 N. (the sow) δονοῦσα kai
τρέπουσα τύρβ᾽ ἄνω κάτω, proves nothing, since there the particular shade of
meaning is conveyed not by the verb but by the adverbs. It seems to me that
a use of τρέπειν such as we have in Ag. 1328 must have been natural to every
Greek at every period; but if the reader prefers it he may see in it a Homeric
or poetic use. The idea of ‘turning’ events and fortune to good or to bad was
entirely familiar to Aeschylus and his contemporaries: Cho. 775 ἀλλ᾽ εἰ
τροπαΐίαν (schol. μετατροπήν) Ζεὺς κακῶν θήσει ποτέ, Pers. 941 f. δαίμων yap 68°
! In this point Miss Macurdy, C.R. lii, 1938, 4 f., has recently associated herself with many
predecessors. She prepares the way for her conjecture ῥέψειεν by a categorical condemna-
tion of τρέψειεν, In this she is wrong. But she is right in protesting against the trifling with
σκιαγραφία, which, in consequence of an idea unluckily thrown out by Blomfield (he was
misled by Iambl. Protr. 8, p. 47. 8 Pist.), has for over a hundred years prejudiced interpreta-
tions of this passage. There is nothing to suggest it in Aeschylus, since γραφήν is merely an
element in the metaphor of the obliterating sponge and the first half of the sentence
(εὐτυχοῦντα... τρέψειεν) contains no hint of this conception. G. Thomson’s comments on
σκιαγραφία are concerned with a fable convenue, not with the text of Aeschylus. We might
also be spared a revival of the unhappy conjecture πρέψειεν,
2 A. Roemer, Sitzgsber. Bayer. Akad., Phil.-hist. Kl., 1888, 211 f., believes that ‘ τρέπω is
too weak and avarperw much better and more forceful as a designation of the fact’. He
compares Pers. 163 μὴ μέγας πλοῦτος κονίσας οὖδας ἀντρέψηι moût ὄλβον. Obviously this
expression is much more forcible and appropriate for the violent upsetting with a kick (very
similar is S. Ant. 1275 λακπάτητον ἀντρέπων χαράν), but for that very reason it is not appro-
priate to Ag. 1328, since what Cassandra has in mind is a process which is indeed decisive
in its effect but which takes place quite gently, almost imperceptibly. This idea is required
by the parallelism with the wiping out by a sponge 1329, which is just as strict as is the
parallelism of the subjects σκιά and σπόγγος. Incidentally, we may note the complete
failure of Roemer’s attempt to prove the occurrence in Aeschylus of the potential optative
without ἄν in independent statement-clauses. In questions, as is well known, this is legiti-
mate within certain narrow limits; cf. Kuhner-Gerth, i. 230 (Herodas 5. 75 f. καὶ τίς οὐκ
ἀπαντῶσα ἔς uev δικαίως τὸ πρόσωπον ἐμπτύοι should have been cited), Sidgwick, in his
annotated editions of A. Ag. and Cho., Append. I, Wackernagel, Syntax, i. 236 f.; on Cho.
595 Blass with unnecessary artificiality says of τίς λέγοι that ‘ ἄν can be supplied from what
precedes’ (regardless of the full stop and the ending of a stanza at 593).
620
 COMMENTARY line 1329

αὖ uerdrpomos én’ ἐμοί, Pind. P. 1o. 20 f. μὴ φθονεραῖς ἐκ θεῶν μετατροπίαις

ἐπικύρσαιεν, fr. 177. 1 Schr. πεπρωμέναν ἔθηκε μοῖραν μετατραπεῖν.
The choice of σκιά as subject of the reversing action has the effect of
intensifying the pessimism of Cassandra’s thought. Some mere nothing,
something that has no real substance, is enough to bring about the change in
man’s fortune. No exertion of a god or a daemon is necessary, no instrument
is employed. The continuation is of a piece with this thought ; a sponge, that
soft, yielding thing, wipes the picture out in an instant, and the whole, light
and shade together, has vanished for ever.
It is not certain how the δυστυχῆ of the MSS should be taken. Wellauer:
‘ ἐστὶ subaudiendum est’; so Hermann. Hartung, on the other hand, says:
‘not ἐστί but ἦι would have to be supplied. For εἰ δυστυχεῖ would mean: if it
is unfortunate as it is actually unfortunate. But the subjunctive of εἶναι is
not normally omitted.’ This assertion is incorrect in two points. First, this
conception of the function of εἰ with the indicative is much too narrow; Ag.
338 may suffice as an example to the contrary. Secondly, even in purely
hypothetical conditions the verbum substantivum can be omitted; cf., e.g.,
A. Suppl. 960 f. εἰ δέ τις μείζων χάρις (‘if that should be more pleasing to you’),
πάρεστιν οἰκεῖν καὶ μονορρύθμους δόμους. Thus grammar has nothing to urge
against εἰ δὲ δυστυχῆ. Perhaps this was what Aeschylus wrote. For the idea
of πράγματα being called δυστυχῆ we may compare Sept. 339 πολλὰ... δυστυχῆ
τε πράσσει. But it is equally possible that we have here a form of the verb
δυστυχεῖν, and perhaps the remarkably strict general symmetry (see above) of
the two clauses from εὐτυχοῦντα μὲν tells in favour of this. Since Blomfield,
who gives no reasons for his choice, the usual reading has been δυστυχοῖ.ἷ
Doubtless the optative, even in so general a sentence, would be linguistically
correct, cf., e.g., Ag. 1042 (F); and the ‘itacistic’ confusion of οἱ and ἡ is a
matter of small consequence. But nevertheless δυστυχῆς (Victorius) has the
first claim to consideration. On εἰ with the subjunctive in poetic language cf.
Kühner-Gerth, ii. 474 n. x. In Aeschylus it is certain in Suppl. 92, Pers. 791,
Eum. 234 (on the text of this passage see above, p. 45 ἢ. 1), for a further
possible instance see below on 1338-40. In Suppl. 92 and Eum. 234 the con-
struction is found in the context of a γνώμη, as in the passage before us.
Thus, if δυστυχῇ is not to be retained, I agree with those who have read
δυστυχῆι, e.g. Conington, Hartung, Roemer, op. cit. (see p. 620 n. 2) p. 210,
A. Y. Campbell.
1329. βολαῖς excludes any idea of a slow wiping-out : it suggests the notion
of a movement swift as lightning. The plural as in σφαγαί, φυγαί, and other
nouns denoting an activity (cf. E. Bruhn, Anhang zu Sophokles, § 3. ii); cf.
P. 379 n. I on ἁρπαγάς.
Eur. fr. 618 N. τὸν ὄλβον οὐδὲν οὐδαμοῦ κρίνω βροτοῖς, ὅν γ᾽ ἐξαλείφει ῥᾶιον ἢ
γραφὴν (Meineke, γράφει codd.) θεός has been compared by Blomfield (follow-
ing Jacobs).

1 There is probably no need to doubt the possibility of this form, instead of -oin, in
Aeschylus. Admittedly, if Lautensach, Glotta, vii, 1916, 108, has not overlooked any, there
is no instance in him of the shorter form of the third person singular (the longer form is
found in Suppl. 1064 ἀποστεροίη, Ag. 349 κρατοίη, cf. Lautensach, 105), nor yet of the second
person (long forms Ag. 1049 ἀπειθοίης, Cho. 1064 εὐτυχοίης), but in Prom. 978 voaoiy dv is
certain.
621
line 1329 COMMENTARY

γραφή can, of course, mean ‘writing’, and the removal of writing could be
denoted by an expression corresponding to the metaphor of the sponge,
S. Trach. 683 χαλκῆς ὅπως δύσνιπτον ἐκ δέλτου γραφήν, but in all probability
the idea here is that of a painting or drawing. On the interest shown by
Aeschylus in works of art cf. W. Nestle, Neue Jahrb. f. d. klass. Altert. 1907,
329, and W. Kranz, Stasimon, 74.
Many commentators have introduced into the sentence ei de δυστυχῆι....
γραφήν an idea foreign to it. Schutz explains: ‘adversitatis adeo facile
oblivisci solent homines, ut eius memoria, tanquam scriptura spongia deleta,
prorsus evanescat.’ Itisin vain that we seek in the words of the poet any
foundation for the first part of this paraphrase. Nevertheless, this idea has
remained popular: cf., e.g., Hermann: 'infortunii subito exstinguitur memoria’,
Nagelsbach (commentary), and even Wilamowitz (transl:): ‘Des Menschen
Ungluck wird vergessen, schnell, so schnell, wie nasser Schwamm die Zeich-
nung auf der Tafel loscht’, and Kranz, Hermes, Ixv, 1930, 366: ‘Misfortune is
wiped out of the memory as easily as a sketch is wiped out by a sponge.’
Kranz adds this comment: ‘It is precisely the fact that misfortune and misery
vanish so readily from the recollection of men that she finds far more painful.
Cassandra is right.’ In reality the structure of the sentence leaves no doubt
that what appears as γραφή in the metaphor of the sponge has as its counter-
part in real life the sum total of βρότεια πράγματα. The true thought of the
poet, and in particular its development and sharper definition in the second
half of the sentence, has been understood by Peile and by Conington, who
followed him : ‘easy as is his [man’s] fall from prosperous to adverse circum-
stances, a yet more fatal change hangs over him . . . from a state of adversity
the work of a moment is sufficient to reduce him to a state of absolute nothing-
ness—and this last change, the speaker adds, I deplore much more than that.’
ὥὦλεσεν γραφήν : this wiping-out annihilates the whole existence, fortune and
misfortune, present, past, and everything whatsoever ; along with the picture
vanishes also every recollection that there has once been there a man’s life
and fate; without a trace the unfortunate goes hence ‘as if he had never been’.
This lapse into nothingness is in Cassandra's view far worse! than the state
of misfortune which men lament most.
1330. The explanation of 1328 f. which has been outlined here proves, what
is also necessary in itself, that ταῦτα must be referred to the second and
ἐκεῖνα to the first part of the immediately preceding sentence (1327 ff.). The
view that ἐκεῖνα referred to the utterance of the coryphaeus in 1321 is rightly
controverted by Kranz among others (loc. cit.); he well points out that in
1322 the Pois begins with a completely fresh start and therefore cannot
contain an answer to what precedes it. ταῦτα (or τοῦτο)... ἐκεῖνα is used in
exactly the same way in, e.g., Sept. 264 τοῦτ᾽ (your last word) ἀντ᾽ ἐκείνων
(what you said before) τοὖπος αἱροῦμαι σέθεν and Ag. 1105 f. τούτων (the crime
to which the prophetess had last alluded) ἀιδρίς εἰμι τῶν μαντευμάτων" ἐκεῖνα
(the fate of Thyestes' children which was revealed before that) δ᾽ ἔγνων ; and
the same thought takes in the second part of the scene (1242 ff.) the form:
1 The conception that oblivion, λήθη, is a κακόν has been investigated in its connexion
with Cassandra’s words by G. Rudberg in the Festschrift Gäva och krav, Skrifter tillagnade
M. Bjorkquist, 1934, 285 ff.; he refers to Hesiod, Theog. 227 (where the reading 71:05» has
been challenged by many editors), etc.
622
 THE SECOND PART OF THE CASSANDRA SCENE
τὴν μὲν Θυέστου δαῖτα παιδείων κρεῶν ξυνῆκα... τὰ δ᾽ ἄλλ᾽ ἀκούσας ἐκ δρόμου
πεσὼν τρέχω, i.e. articulation by μέν... de. In 1330 ἐκεῖνα picks up the pre-
ceding (1327 f.) név-clause, ταῦτα the ôé-clause (1328 f.).
H. Weil’s epigram, ‘vaticinatur Cassandra, non philosophatur’, is con-
stantly quoted ; his extraordinary mistake of assigning 1327 ff. to the Chorus!
has been adopted, e.g., by Headlam, and also by MacNeice, capable though he
often is of following the poet’s thought with the sympathetic understanding
of a poet.
Cassandra sees things in their true form; she sees them, too, in their true
setting, freed of the isolation in which they appear to dimmer eyes. μένει τὸ
θεῖον δουλίαι περ ἐν φρενί. In her own lot and that of her nearest and dearest
she has experienced the vicissitudes of fate, the sudden change from fortune
to misery, from misery to utter annihilation. But her thought pierces deeper:
what is happening to her is no other than what is happening time after time
to the generality of men, although most of them lament only their fall from
fortune to misfortune, unconscious of their ultimate nothingness.? It would
be entirely wrong to take the words (à βρότεια πράγματα κτλ. merely as a
veiled pronouncement about Cassandra herself. Here, on the threshold of
death, her glance travels far beyond her own fate to the whole life of mankind.
Not that she finds there any exaltation or consolation; she is no Stoic, and
the last outburst of her sorrow is permeated by an all-embracing feeling of
sadness and utter despair. Shall we then be rash enough to deprive this
great figure in the greatest Athenian’s masterpiece of her share in the old
heritage of the Hellenic mind, the urge that forces her to leave the particular
behind and rise upwards into the sphere of the universal?

Song and speech are the two roots out of which the rich variety of Greek
dramatic forms grew. Aeschylus uses more than once the difference in the
character of lyrics and ῥῆσις to present one and the same dramatic motif
first in song and then in recitation (or vice versa), and thus, by means of
repetition as well as modification, to bring out its various aspects (cf.
W. Kranz, Stasimon, 166). But nowhere else has the combination of the
two elements produced such a marvellous effect as in the Cassandra scene.
All that is characteristic of either of the two forms seems here to have found
its most powerful expression, and where the limits of the one are reached, the
other assists and supplements it so as to provide the whole scene with an
immense scope and depth.
λόγος includes ratio as well as oratio. When Cassandra's ecstasy is over and
her excited songs and those of the Chorus have faded away, we enter through
the gateway of her ῥῆσις (1178 ff.) into a region of lucid rational thoughts,
ἡλίου πρὸς ἀντολάς. Everything calms down, the temper, the words, the ges-
tures of the prophetess. The whole direction of her utterances is altered. During
the spell of her ecstasy she addressed herself to Apollo, to Clytemnestra,
1 That, however, is not the worst. No one can expect me to discuss a treatment of a
tragic text which tears loose 1330 καὶ ταῦτ᾽ ἐκείνων κτλ, from what precedes it and assigns
this trimeter alone to the Chorus.
2 The wisdom of Artabanus (Hdt. 7. 46. 2 ff.) is superior to the melancholy of Xerxes,
who in the common fashion complains only of the shortness of our life, while his uncle
knows: €repa τούτου παρὰ τὴν ζόην πεπόνθαμεν οἰκτρότερα (καὶ ταῦτ᾽ ἐκείνων μᾶλλον οἰκτίρω
πολύν) : our life is so full of misery that, short as it is, everyone often desires an early end of it.
623
lines 1178-1330 COMMENTARY

and perhaps (cf. on 1138) to Agamemnon, but never to the Elders. It is

true that in a vague manner she takes some notice of their presence and
their words (cf. on 1095), but we should exaggerate if we said that she answers
them. It is as if she were catching almost unconsciously at bits of strange
tunes that the wind carries to her from somewhere. What is really and fully
present to her mind is the figures of her visions, not the old men who are
standing near by. This attitude completely changes with the beginning of the
ῥῆσις. The Elders, though not explicitly named, are the object of the con-
cluding clause of the first period (1183 φρενώσω δ᾽ οὐκέτ᾽ ἐξ αἰνιγμάτων), and
with the next words (καὶ μαρτυρεῖτε κτλ.) Cassandra not only turns to the
choreutae, but assigns to them an active part in the διαλέγεσθαι by bidding
them bear her witness. To emphasize the importance of this point, she once
more asks at the end of her speech (1196) ἐκμαρτύρησον κτλ. This detail is
significant : nothing could illustrate more clearly the change by which a pro-
cess of reasoning, supported by proof and evidence, has taken the place of the
former ecstasy. At the present stage Cassandra is determined not merely to
voice her knowledge of the imminent catastrophe but to force the Elders to
recognize its truth. This she cannot achieve without first clearing herself of
the suspicion that she may be but one of the many ψευδομάντεις. In accord-
ance with the generally accepted test of a true prophet (cf. on 1185), she
proves that she is perfectly acquainted with facts which could not have come
to her knowledge in the ordinary way. By revealing what took place at
Argos in the past she guarantees the genuineness of her seership. This
device makes it possible for the poet to display the crimes of the house of the
Atridae again in chronological order as he has done in the lyrical part of the
scene. The two parts correspond also in that each time Cassandra begins with
a general picture of incessant bloodshed (1090 ff., 1186 ff.) and only afterwards
passes on to a particular misdeed, the slaughter of the children of Thyestes
(1096 f., 1217 ff). But although there is a strict parallelism in form, the
nature of the sequence is different. In the visions of which Cassandra sings
one picture after another presents itself to her mind, each arising suddenly
and without any apparent connexion as it emerges from the impenetrable mist
of her trance. In her speeches, on the other hand, she carefully connects the
various events. One link is formed by the idea that her knowledge of the
former horrors is to serve as evidence for the authenticity of her prophecies.
A still more important link is added when she explains that the misdeeds of
an earlier generation are the cause and origin of fresh disasters (1223) and
that these in turn will procreate revenge in years to come (1280 ff.). Where
Cassandra traces the stream of murder to its fountain-head, she goes beyond
what she has seen in her visions. There the remotest stage she reached was
the slaughter of the children (1096) ; here it is the adultery of Thyestes that
appears as πρώταρχος ἄτη (1192 f.). This sharply pointed phrase is significant:
in it the causal nexus between guilt and atonement is as much stressed as in
the words of the Chorus (223) about the παρακοπὰ πρωτοπήμων. It should be
noted that nowhere in the play before the Cassandra scene is there any
mention of Thyestes and his children or indeed the slightest allusion to them.
If Aeschylus wanted to suggest, as many of his commentators suppose him
to do, that the crime of Atreus, visited upon his son, was the primary, or at
any rate a fundamental, cause of Agamemnon’s downfall, he would have had
624
 THE SECOND PART OF THE CASSANDRA SCENE
ample opportunity of mentioning the old misdeed in one of the retrospective
sections of the choric songs. The connexion into which the murder of the
children of Thyestes is brought by Cassandra is unmistakable (1223 ff.): ex
τῶνδε ποινάς φημι βουλεύειν τινὰ... οἰκουρόν κτλ. The line drawn from the
old crime to that which is imminent does not in the first place link the suffer-
ing of Agamemnon with the doing of Atreus, but the doing of Aegisthus with
the suffering of Thyestes and his children. To make such a differentiation
might seem a mere sophism if it were not confirmed by the play as a whole.
Throughout the tragedy the indications regarding Agamemnon’s guilt are
consistent. His main crime is the sacrifice of Iphigeneia. Moreover, he has
done an evil thing by sending so many men to their death for the sake of one
woman. (The question whether he could have avoided acting as he did
without becoming guilty in a different way need not be discussed here again ;
cf. on 212.) It is only as an additional cause that the slaughter of the children
of Thyestes contributes to the overthrow of the son of Atreus. It would be
absurd to attempt an exact calculation as to the degree of efficacy in each of
the different elements that work together towards Agamemnon’s fatal end.
But it isimportant not to overlook the fact that Aeschylus lays far greater stress
on what the king has done himself than on the crime committed by his father.’
The latter motive mainly serves to account for the share of Aegisthus in plot-
ting the murder. In this connexion a more general warning may not be amiss.
Modern readers are apt to read into the Orvesteta a strong influence of the
hereditary curse in the race of Tantalus. This tendency is partly due to an
unconscious recollection of the story of the Labdacidae, and probably still
more to the manner in which in famous plays of Sophocles and Euripides the
fate of Agamemnon and his children is set against the background of the sins
committed by Tantalus, Pelops, and the sons of Pelops. If we want to
understand the Oresteia, we should be careful to respect the limits which
Aeschylus has firmly drawn and should beware of any confusion with elements
familiar to us from elsewhere (cf. on 1193).
It must have seemed perfectly natural to an ancient audience that Cas-
sandra discloses her knowledge of the past to prove the trustworthiness of her
prophecies (see above). There is then nothing artificial in the manner in
which she expatiates on the earlier crimes of the house. Thus the customary
idea about the nature of true seership provided the dramatic poet with an
excellent means of widening his story beyond the time of the actual plot and
strengthening the links between the past and the future. Any reader of
Aeschylus knows that the device in the Cassandra scene has its exact parallel
in the Prometheus, where the story of Io is spread backward and forward for

1 Cf. the well-balanced judgement of Henri Weil, Etudes sur le drame antique, 39:
‘Egisthe . . . n'est qu'un personnage secondaire chez Eschyle ; Clytemnestre tue son époux
pour venger le sacrifice d’Iphigenie; ce sacrifice, et non la querelle d’Atrée et de Thyeste,
est la cause première du parricide qui fait le sujet de la trilogie”, and (p. 43 f.): ‘Les causes
de la catastrophe sont donc multiples, le crime d’un père a fatalement voué la tête d’Aga-
memnon à une mort lamentable, mais cette tête n'est pas innocente: le sang d'une fille
immolée, le sang d'innombrables victimes de son ambition [I do not think the charge of
ambition justified, see the commentary on the parodos], crie contre le rejeton d'une race
grande dans ses desseins et violente dans ses passions. Le fils d'Atrée est trés coupable: le
poéte est trop sincére pour dissimuler ou pallier ses fautes; en revanche, il lui a prété la
noblesse, la hauteur des sentiments', etc.
4872.3 L 62 5
lines 1178-1330 COMMENTARY

the benefit of the audience. Prometheus gives the reason for drawing out
his account in these words (824 ff.) : ὅπως δ᾽ ἂν εἰδῆι μὴ μάτην κλύουσά μου, ἃ
πρὶν μολεῖν δεῦρ᾽ ἐκμεμόχθηκεν φράσω, τεκμήριον τοῦτ᾽ αὐτὸ δοὺς μύθων ἐμῶν,
and (842 f.): σημεῖά σοι τάδ᾽ ἐστὶ τῆς ἐμῆς φρενός, ὡς δέρκεται πλέον τι τοῦ
πεφασμένου.
We have seen that the speeches of Cassandra add a strong element of
causality to the visions and prophecies of her songs. This applies not only to
the matter of her utterances but also to the source and nature of her inspira-
tion. In the lyrical part of the scene Cassandra’s seership had to be accepted
as an astounding fact ; no reason was given why it had been conferred upon
her. In the stichomythia 1202 ff. Cassandra herself explains the mystery.
This stichomythia occupies the middle of the whole scene (1072-1330) ; it is
indeed central. Overwhelmed by her fate that is now approaching fulfilment,
and yielding to the keen questions of the old Argive, the prophetess reveals
step by step the deepest secret of her life. The way in which the very old
device of stichomythia is here employed (in later tragedy it was to decay into
sad mannerism) is characteristic of the art of Aeschylus at its maturest stage.
Full advantage is taken of the process of a most intense διαλέγεσθαι. Several
of these perfect stichomythiae give us the impression that this is the only
means of bringing to the surface experiences, anxieties, fears, and hopes which
in normal and coherent speech must needs remain hidden. The pressure of a
determined interrogation seems to overcome the natural reserve of the person
on whom it is brought to bear: he gradually gives way until a depth has been
reached beyond which words are not allowed to probe. We have a typical
instance of such a stichomythia in Ag. 538 ff. There the questioning of the
herald drives the coryphaeus to a point (548) where he finds it necessary to
shelter behind resolute silence. Exactly the same turn is given to the most
momentous of all Aeschylean stichomythiae, Prom. 515 ff., where the limit
is marked by the hero’s words (520) τοῦτ᾽ οὐκέτ᾽ ἂν πύθοιο, μηδὲ λιπάρει. Other
remarkable instances of such intensified and, as it were, deepened sticho-
mythiae ‘are Cho. 166 ff., 908 ff., and, in a different sense, Eum. 892 ff. (here
Athena, by driving home her points one by one, brings about a result which
her previous speeches have not been fully able to achieve; the stichomythia
is the climax of the whole scene).
Ecstasy and visions, which, when the lyrics ceased, receded into the back-
ground, are not, however, completely absent from the second part of the
Cassandra scene. No timid formalism hampers the flow of this mighty poetry ;
it rushes downwards, monte decurrens velut amnis, imbres quem super notas
aluere ripas. Twice, when the hearer’s mind is wholly engrossed in the
reasoning of a stichomythia, the dialogue is violently interrupted by an
outburst of θεία μανία (1214, 1256). For a moment the quiet trimeters are
swept aside, and wild ejaculations take their place. With frightful sudden-
ness we are thrown out of the calmer region of arguing, proving, and per-
suading, and hurled back into the sphere of horrible visions. Cassandra's
attitude to the visions is not exactly the same as before, when, in complete
aloofness, she was solely concerned with what her inner eye perceived. Now
she addresses herself to the Elders in an effort to make them see the appari-
tions: ὁρᾶτε τούσδε τοὺς δόμοις ἐφημένους νέους κτλ.
After the fierce action of 1264 ff. the tide of emotion ebbs in the later part
626
 COMMENTARY line 1334

of the scene. There are no eruptions of frenzy any longer. Unfathomable

misery speaks in a low voice. The resigned wisdom of Cassandra’s last words
is perhaps more deeply moving than all her ecstasies.

1331 f. Stanley compares Hdt. 7. 49. 4 (speech of Artabanus) εὐπρηξίης δὲ οὐκ

ἔστι ἀνθρώποισι οὐδεμία πληθώρη.
ἔφυ: such is the nature of εὖ πράσσειν.
1332. The MSS have δακτυλοδεικτῶν perispomenon and the scholion (ZyoA.
wax.) interprets τῶι δακτύλωι τις παραδεικνύς. Whether the false explanation
caused the false accent or vice versa cannot be told, nor does it matter.
It has long been recognized that δακτυλοδείκτων (van Heusde compares
S. Oed. R. gor χειρόδεικτα) μελάθρων must be understood (‘‘‘finger-pointed
palaces” that is, "admired and gorgeous palaces'"' Headlam). This should
have been regarded as final, but Farnell, C. Q. iv, 1910, 182, took up the cudgels
for δακτυλοδεικτῶν again ; he was followed by Mazon (cf. also his introductory
‘Notice générale’, p. xxviii n. 2) and G. Thomson, and hesitatingly by Murray
(fortasse recte’). Farnell understands: ‘keeping the Nemesis- of excessive
prosperity away, averting the evil eye by pointing with his middle finger and
uttering the words ‘‘Come no longer to enter here". The gesture which
Farnell describes is coarse and highly abusive,’ and no one could possibly
think of using it against Nemesis: est vemens dea: laedere hanc cavelo. Even
in a negative form, such an idea is wholly impossible, to say nothing of its
incompatibility with the style of tragedy. Against the usual interpretation
of δακτυλοδείκτων Farnell says: ‘if the house is already all-envied, the danger
which the wise owner ought to avert is already incurred: we rather want
some such phrase as ‘‘no one tries to keep off excessive prosperity from the
house of moderate means μελάθρων perpiwv’’.’ He seems to have forgotten
not only ἀκόρεστον but also μηκέτι (1334). The poet’s idea is, of course, that
εὖ πράσσειν continues to visit the house of the fortunate. Stanley compares
Ayal} Τύχη who, in a play of Menander (Plaut. Aw. 100), wishes to pay the
house a call; cf. also Plaut. Rud. sor Malam Fortunam in aedis te adduxi meas.
1333. The aorist ἀπειπών does not describe the manner in which the action of
εἴργειν takes place or indicate anything that is contemporaneous with it, but
something that precedes or is causal. The master of the house makes a
prohibition (whether by word of mouth or by a notice on the door) the result
of which is εἴργειν. On the other hand, 1334 φωνῶν indicates an utterance
which is contemporaneous with εἴργειν.
1334. Hermann's restoration is generally accepted. Hermann (on 754, 757,
1300 in his numbering) pointed out the tendency of Triclinius to fill out the
paroemiac to a full dimeter; cf. 789, 1340, 1540 where Triclinius for the same
reason has inserted νῦν, Pers. 152 where he adds δή. Triclinius' various stop-
gaps (τῶν, εὖ, τις) make monstrous lines of the paroemiacs Ag. 356, 804, 806.
The present passage, the corruption of which goes back to the hyparchetype
of FG Tr, shows that a certain dislike for the paroemiac might influence
copyists earlier than Triclinius.?
1 Since Otto Jahn’s classic discussion, Berichte d. Sachs. Ges. d. Wiss., Phil.-hist. KI.
1855, 81 ff., much has been written on the subject. . . |
2 Had this tendency, at a much earlier stage, anything to do with the distortion of the
verb in 1l. 87 to ϑυοσκινεῖς ?
627
line 1334 COMMENTARY

The correct punctuation before τάδε is found as early as Victorius and

Canter, but many later editors neglected it. For τάδε φωνῶν coming after
the quoted words Blomfield compares Cho. 313 f. δράσαντι παθεῖν, rpıyepwv
μῦθος τάδε φωνεῖ; for τάδε cf. also Sept. 532 τόδ᾽ αὐδᾶι after the indirect speech.
Lobeck, Aglaophamus, 1239 (quoted by Hermann), connects this passage
with the remark of the Emperor Julian about the Cynic Crates (Or. 6,
p. 200 b): ἐπὶ τούτου φασὶ τοὺς “Ἕλληνας ἐπιγράφειν τοῖς ἑαυτῶν οἴκοις ἐπὶ τῶν
προπυλαίων" Εἴσοδος Κράτητι Ayadaı Δαίμονι (on the religious idea behind
this cf. Ganschinietz, RE, Suppl. iii. 39). Farnell, loc. cit., compares the’
well-known μηδὲν εἰσίτω κακόν (Diog. Laert. 6. 2. 39, 50), which has been
found in Greek and Latin inscriptions in various places (cf. Bücheler, Kl. Schr.
i, 88; Wilamowitz, Griech. Lesebuch, Erläuterungen, 18f.). Martial 1. 25. 5 ante
fores stantem dubitas admittere Famam is adduced by Dobree, Advers. ii. 26.
1335. καὶ τῶιδε: ‘and so to the man here . . .’, cf. on 8 kai νῦν and on 184 καὶ
τότε. The application of the general maxim to the special case does not end
with the μέν clause but comprises also the continuation (1338 viv δ᾽ εἰ x7A.),
i.e. the whole of the following anapaests.
1335 f. πόλιν... Πριάμου; a hyperbaton of the emphasizing type such as
267 Πριάμου... πόλιν.
1337. θεοτίμητος (predicative) δ᾽ οἴκαδ᾽ ἱκάνει; cf. on 530.
1338-40. ἐπικρανεῖ cannot stand because the penultimate of xpavet is short (cf.
Cho. 1075). Dindorf, though with some doubt, refers to Eum. 950, wrongly,
for there ἐπικραίνει is the MS reading and is demanded by the sense (cf. on
Ag. 369); Mazon’s treatment of both passages (ἐπικρᾶνεῖ) is unfortunate.'
Hermann refused to admit the abnormal prosody and recognized both the
metrical interpolation in Triclinius (on ἄγαν scanned as two shorts cf. on
1241) and its cause (dislike of the paroemiac except at the end of the complete
anapaestic system, cf. on 1334). Hermann reads ἐπικραίνει (in Eum. 950 the
corrector of the Mediceus altered ἐπικραίνει to ἐπικρανεῖ) ; this might be
defended on the ground that Cassandra’s revelation impresses the fact of
Agamemnon’s death and its after-effects so strongly on the Chorus that they
speak of it as present. But neither Hermann nor Hartung, Paley, and the
rest of his succession have made any attempt to explain the juxtaposition of
ἀποτείσει and ἐπικραίνει. Wecklein’s explanation depends on his wrong con-
jecture ἄλλος. Wilamowitz, who likewise adopts Hermann’s ἐπικραίνει, says
‘praesens pro futuro’ and refers to his note on Suppl. 406; but an examination
of the passages there adduced shows an arbitrary interpretation of Pers. 585,
Eum. 506, Prom. 513 (cf. also 525) ; the same is true of the passages Cho. 509,
S. Phil. 117, E. El. 974 quoted by Headlam in the footnote to his translation
of Suppl. 406. None of these passages? shows a ‘ "prospective" use of the

1 The evidence at our disposal is certainly not strong enough to support O. A. Danielsson’s
assumption (Grammat. und etymol. Studien, i, Upsala 1888, 36) that the future ἐπεκρᾶνεϊ is
derived from *-«pàavet,
2 J have not been able to verify Headlam's reference to E. Iph. T. 977-80 and Rhes. 581,
596. The starting-point of this discussion, the much-vexed μεταλγεῖς A. Suppl. 405f.,
baffles me: Headlam's and Wilamowitz's interpretation of the present tense, which is
bound up with the conjecture &p£as, is not satisfactory (v. supra), nor is Wecklein's defence
of the MS text (explained in his 1902 edition) : ‘was hast du peinliche Bedenken das Rechte
zu tun?’. The natural construction seems to be to connect μεταλγεῖν with the anticipated
(399 ff.) criticism on the part of the people; ‘we require "posthac dolebis" ', Tucker says.
628
 COMMENTARY lines 1339 f.
present tense' (Headlam) ; some of them may perhaps be regarded as 'timeless'
presents, on which see Wackernagel, Syntax, i. 157 f.! Therefore the only
solution, unwelcome though it is, seems to be to alter not only the undoubtedly
corrupt ἐπικρανεῖ but also ἀποτείσει, which is in itself unexceptionable, and to
make the two verbs which denote elements of the same action syntactically
parallel. This can be done either by making both aorist optative (ἀποτείσαι
and ἐπικράναι Keck, ‘perhaps ἐπικράνειεν ' Headlam ; on the distribution of
the forms in -ewv and -aı in the dramatic poets cf. Lautensach, Glotia, viii,
1917, 171 ff., 178 ff.) or by making both subjunctive, as Sidgwick did (on εἰ
with the subjunctive in poetic language generally and in Aeschylus see on
1328).
1338. προτέρων : the interpretation ἤτοι τῆς ᾿Ιφιγενείας (Tricl.) is only partly
right. I agree with Wecklein: 'indefinite, so that it is possible to think of
Iphigeneia as well as of the children of Thyestes'. The vagueness is in-
tentional. The children of Thyestes are suggested by the visions of Cassandra
(1096 f., 1217 ff.) and by the confirmatory words of the coryphaeus (1242 ff.) ;
but it would be wrong to exclude Iphigeneia on the ground that she is not
mentioned by Cassandra (so, e.g., Th. Plüss, Die Tragödie Ag., Basel 1896,
19 £). Why should the Chorus have forgotten their own account of Calchas'
speech, prophesying the avenging act of the Μῆνις τεκνόποινος ?
1339. τοῖσι θανοῦσι refers to the same as προτέρων.
1339 f. F and Tr have a comma after θανὼν. This punctuation (and the
construction of the parts of the sentence which it implies) has long remained
the vulgate, e.g. in Victorius, Canter, Stanley (‘si priorum sanguinem luet,
et morientibus moriens, aliarum mortium poenas nimis adimplebit’), Schütz.
Heath felt the necessity of defending the use of the dative (τοῖσι θανοῦσι)
which has to be assumed here and which, to say the least, is remarkable:
"propler iam mortuos moriens, id est, ut statim subiungit Chorus, in eorum
ultionem. Similis constructionis exempla notavimus ad Prom. Vinct. 899,
974’, but his attempt is unsuccessful.^ Later, too, good Greek scholars have
succumbed to the temptation of making τοῖσι θανοῦσι dependent on θανών,
e.g. Schneidewin (‘den Gemordeten: gemordet, indem er ihnen zum Opfer
fällt’), Nägelsbach (‘um der Geschlachteten willen geschlachtet’), Sidgwick
(‘dying for the dead’), Wilamowitz (‘und für die Toten sterbend Tod den
künftigen vererbt’). So far as I can see, grammar does not allow of such a
construction; it has rightly been remarked 'insolentius illud “moriendo

11 should like here to give a warning against extending to Ag. 1338 ff. ἀποτείσει...
ἐπικραίνει what Wackernagel, 162, says with reference to Hdt. 1. 109. 4, and particularly
Thuc. 6. 91. 3, about the use of the present tense which marks the action of the apodosis as
coincident with a future tense in the conditional clause. In the first place the mixing of the
two tenses within the conditional clause and connected by καί is much harsher, and in the
second place I am dubious about Wackernagel's whole explanation on this point, wherein
he has followed Mahlow. K. W. Krüger on Thuc. 6. 91. 2 (also on 1. 121. 3) seems to me more
on the right lines; this use of the present has nothing to do with ‘coincidence’,
2 In Prom. 899 the dative is the product of an impossible conjecture ; in 974 the dative is
as normal as it is in the other passage here quoted for comparison by Heath, Ag. 136
ἐπίφθονος. . . πτανοῖσιν «vol. It should be noted that the case for the construction τοῖσι
θανοῦσι θανών cannot be supported by the well-known passages in Sophocles where a not
easily definable dative is used with the perfect τέθνηκα, 4j. 970 θεοῖς τέθνηκεν οὗτος, οὐ
κείνοισιν, οὔ, and still more strikingly El. 1152 τέθνηκ᾽ ἐγώ σοι, where the context leaves no
doubt as to the sense; Jebb's ‘I am dead in relation to thee' is not adequate.
629
lines 1339 ἔ, COMMENTARY

mortuis" conciliare mihi non potui’ (Davies) and ' τοῖσε θανοῦσι can only
depend on ποινὰς... émtxp.’ (Wecklein). Hermann’s interpretation satisfies
both grammar (except for a single offence against normal usage which will be
noticed later) and sense: ‘nunc st ille priorum sanguinem. luit, et mortuis
(Iphigeniae) mortendo aliarum mortium (suae ipsius) poenas adducit, quis haec
audiens se innoxio genio maium censebit?' At first sight this seems to be
painfully lame and tautological. The 'nimia orationis prolixitas', which Weil
notes, has frightened reader after reader from following Hermann's inter-
pretation or renderings that move on the same lines. Nor is this stylistic
objection the only one. From the sixteenth century many editors have
wished to find in these concluding anapaests of the Cassandra scene a refer-
ence to the vengeance for Agamemnon’s murder which was several times
mentioned by the seeress; this would also remove the 'prolixitas orationis'.
We should then have three stages: murder of Iphigeneia (προτέρων αἷμα),
murder of Agamemnon (θανών), murder of Clytemnestra and Aegisthus
(ἄλλων θανάτων). ' ἄλλων θανάτων mihi videtur referendum esse ad futuram
Clytaemnestrae et Aegisthi mortem, quos Agamemnonis manes sibi expos-
cent’, says Weil. Canter evidently felt this need when he wrote κτανοῦσι (it
must of course be κανοῦσι) instead of θανοῦσι, which gives the meaning: ‘if he
ordains [cf. on 369] on the slayers a vengeance of further deaths, because he
is (or ‘has been’) slain’. κανοῦσι has been accepted by Davies and more
recently by A. Y. Campbell; Musgrave's θενοῦσι (rots θείνουσι Lawson) has
exactly the same effect. Several editors besides Weil, e.g. Sidgwick, Wila-
mowitz (‘wenn er... für die Toten sterbend Tod den künftigen vererbt’),
Mazon, relate ἄλλων θανάτων to the death of Clytemnestra and her lover
without making any alteration to Üavoóo: And indeed one would not easily
bring oneself to destroy the threefold hammer-blow θανοῦσι θανὼν... . θανάτων
by putting in the place of the first participle a different verb. But whether
we alter θανοῦσι or retain it, it is extremely unlikely that by ἄλλων θανάτων the
death of Clytemnestra and Aegisthus should be meant. From Hermann's
translation sí . . . mortuis (Iphigeniae) moriendo aliarum. mortium (suae
ipsius) poenas adducit it appears that he takes ἄλλων θανάτων as an epexegetic
genitive. The same applies not only to those critics who follow Hermann but
also to those who with Canter read κανοῦσι or a similar verb, and likewise to,
e.g., Weil (see above), Nägelsbach (he joins τοῖσι θανοῦσι θανών, understands
ἄλλων θανάτων, as he expressly states, as genetivus explicativus and refers
ἄλλοι θάνατοι to the death of Clytemnestra and Aegisthus), Verrall, Wilamo-
witz. Sidgwick arrives at a strange compromise: ' "and dying for the dead
brings to pass requital for other deaths", 1.6. he (Ag.) . . . causes a new
requital in the deaths of others (Klyt. and Aegisth.)’; ie. his translation
makes the genitive objective and his paraphrase makes it epexegetic.
Klausen, Conington, Hartung, Wecklein, L. Campbell, and Headlam decide
for the objective genitive, and this is right. A glance at the lexicons shows
that the genitive which is so commonly used with ποινή from Homer's time
onward (seven further cases in Aeschylus) is always objective.! We’need not
1 After I had established this simple point, I found that Kennedy (at the end of his
translation) remarks against Hermann : "This view seems to us very harsh and questionable,
when we see that a gen. dependent on ποινήν or ποινάς universally expresses that of which
the penalty zs paid, and not that of which ft consists.’
630
 COMMENTARY line 1341

go so far as to deny that a Greek could have said ποινὰς θανάτων for a ven-
geance consisting in death, but it would be overbold to force this unexampled
construction on a passage where it is in no way necessary.
Another reason besides grammar makes a reference of ἄλλων θανάτων to the
death of Clytemnestra and Aegisthus impossible. The thought in these lines
is simple and consistent. “The striving of man for success is insatiable. If one
views the fate of Agamemnon, seemingly so highly blessed, how can anyone
allow himself to say: ἀσινεῖ δαίμονι ἔφυν (‘I was born under a lucky star’, as
other periods would phrase it)’? On the very similar sequence of thought in
the chorus 5. Oed. R. 1186 ff. see on 1341. In this connexion the fate of the
king must be painted as darkly as possible ; there must be no redeeming feature
in his fall from the pinnacle of success. A reference to the vengeance which
his death is bound to bring after it would for the feeling of the ancient
audience bring in an element of comfort and relief which would have a
disastrously weakening effect here. It is possible, of course, to think along
these lines: ‘what a curse on Agamemnon that his murder makes his son the
murderer of his nearest kin’, but there is not the smallest hint of this.
Therefore the modern idea that, like Cassandra’s earlier speeches, the words
of the Chorus refer to the future vengeance must be abandoned.
This does not end the difficulties, although the ‘nimia orationis prolixitas’
need trouble us no more. Like the threefold θανοῦσι, θανών, θανάτων, the
sentence itself emphasizes the pitiless working of the law of retribution, eye
for eye, death for death. The repetition of murder of kinsmen is painful;
accordingly the sentence seems to be heavy with painful repetition. But what
is the meaning of ἄλλων" To say ‘if he brings to completion for the dead . .
the vengeance for other deaths’ seems absurd, when sense demands that the
θανόντες are those whose death is immediately afterwards described as ἄλλων
θανάτων. Moreover, it is disturbing to find ἄλλων related to the past, because
in a temporal series of events ἄλλος normally indicates not something earlier
but something additional and later. Examples are to be found everywhere ;
the following are similar in subject-matter to our passage: Ag. 1535 f. (cf.
there) and Cho. 400 ff. ἀλλὰ νόμος μὲν dovias σταγόνας χυμένας ἐς πέδον ἄλλο
προσαιτεῖν αἷμα. βοᾶι yap λοιγὸς ᾿Ερινὺν παρὰ τῶν πρότερον; φθιμένων ἄτην
ἑτέραν ἐπάγουσαν én’ ἄτηι. Hermann and the commentators who related the
phrase to Clytemnestra’s death were probably influenced by a true apprecia-
tion of the normal meaning of ἄλλων θανάτων in such a context. I cannot rid
myself of the suspicion that ἄλλων is corrupt. In this I agree with Wecklein,
but his ἄλλος is wrong. He explains ‘if it is the case that ever another by
his death gives to the dead vengeance for their murder’. But it is incredible
that Aeschylus should have inserted a general subject when everything
points to Agamemnon remaining the subject to the end of the apodosis. I
know of no satisfactory solution.
1341. A syllable is lacking at the beginning. Canter wrote τίς ἂν ζοὐκὸ eit.
Stanley, who remarked against this *mire allucinatus est vir doctus', retained
the MS reading, unmetrical though it is, and translated: ‘quis mortalium
gloriari poterit innoxio genio se natum esse'. And surely this is the sense
required by the context. But Canter's perversion of the thought has every
now and then found adherents, e.g. Blomfield, Conington (‘who would not
! The προτέρων of the MS was unsuccessfully defended by Wilamowitz.
631
line 1341 COMMENTARY
pray for unharmed estate?’), Murray (‘quis non precetur?’). It is question-
able whether a Greek would ever have prayed for something irrevocably
settled in the past, such as the φῦναι of man. Moreover, Hermann has rightly
objected to Canter: ‘omnes, etiam qui nihil umquam de Agamemnonia domo
inaudiverunt, innocua fortuna nati esse optant.’ The reaction of the Argive
Elders to the account of Agamemnon’s fate is parallel to the reaction of the
Theban Elders to the revelation of Oedipus’ destiny (S. Oed. R. 1186 ff.).
There, with the παράδειγμα (1193) of Oedipus before their eyes, the Chorus
asks τίς γάρ, τίς ἀνὴρ πλέον τᾶς εὐδαιμονίας φέρει 7] τοσοῦτον ὅσον δοκεῖν Kai
δόξαντ᾽ ἀποκλῖναι; The thought structure of Ag. 1341 also demands “Who
could boast of lasting happiness?’ Headlam wrongly maintains that ‘ τίς ἂν
εὔξαιτο should naturally mean “who would wish?” ', because εὔχομαι would
yield the required sense ; but he had the merit of calling attention to Schneide-
win's ἐξεύξαιτο. This is perhaps the simplest alteration,! better than the
rather empty τίς wor’ ἂν (E. A. J. Ahrens) and than οὔτις dv (Wilamowitz in
his apparatus), which destroys the emphatic question (cf. Oed. R. 1189 quoted
above). Another possibility is πῶς τις ἂν εὔξαιτο. For the rather unusual
word-order cf. Cratinus fr. 187 K. πῶς τις αὐτόν, πῶς τις dv ἀπὸ τοῦ πότου
παύσεις, Ar. Frogs 1458 πῶς οὖν τις ἂν σώσειε; For πῶς ἄν in the apodosis
after a conditional sentence introduced with ei, cf., e.g., S. Oed. C. 911, 977,
Eur. fr. 832 N., Ar. Frogs 1449 f., Plut. 583, Xen. Hieron τ. 9. For a word
extending from one anapaestic metron into the beginning of the next cf.
1339 and the note on 52.
1341 f. ἀσινεῖ δαίμονι. The poet might well have been perplexed if asked
abcut the classification of the dative. But he was presumably conscious of
using this construction on the analogy of the Homeric τῶ oe κακῆι αἴσηι
τέκον (A 418, compared by Plüss) and similar constructions. He did not
trouble (nor need we) whether in this and similar cases atoyı is to be under-
stood as a dative of circumstance (so Monro, Hom. Grammar, znd ed., $ 144,
p. 137; Kühner-Gerth, i. 435) or otherwise (cf., e.g., Ebeling, Lex. Hom. s.v.
aloa 3a, Leaf on A 418). It is clear that in this phrase δαίμων has lost a good
deal of its personal character and is on the way to what we should call an
abstract idea. The same e.g. Pers. 824 ff. μηδέ τις ὑπερφρονήσας τὸν παρόντα
δαίμονα ἄλλων ἐρασθεὶς ὄλβον ἐκχέηι μέγαν, Ag. 1663 and Cho. 513 δαίμονος
πειρώμενος, Pind. P. 3. 108 τὸν δ᾽ ἀμφέποντ᾽ αἰεὶ φρασὶν δαίμον᾽ ἀσκήσω κτλ.,
5. 122 f. Διός τοι νόος μέγας κυβερνᾶι δαίμον᾽ ἀνδρῶν φίλων, S. El. 999, 1305 f.
οὐ γὰρ ἂν καλῶς ὑπηρετοίην τῶι παρόντι δαίμονι, Soph. fr. 592 N. (= 653 P.)
μὴ σπεῖρε πολλοῖς τὸν παρόντα δαίμονα, E. Alc. 561 πῶς οὖν ἔκρυπτες τὸν
παρόντα δαίμονα and elsewhere in tragedy, Antiphon Soph. Β 49 (Diels-
Kranz, Vorsokr. ii, 5th ed., 357) αὕτη ἡ ἡμέρα, αὕτη ἡ νὺξ καινοῦ δαίμονος ἄρχει,
καινοῦ πότμον. In this meaning δαίμων is very close to πότμος, as is shown by
the passage from Antiphon, and to τύχη, which the scholiasts often use in
their paraphrases. The use is already found in Hesiod, Erga 314 δαίμονι δ᾽
οἷος ἔησθα (cf. Wilamowitz ad loc.) and in the poet of ®, if in © 166 the text
of the MSS (πάρος τοι δαίμονα δώσω) and not that of Zenodotus (πότμον
ἐφήσω) is the genuine version. Cf. E. Rohde, Psyche, ii, 5th ed., 205 n. 5 and
! Readings such as Klausen’s τίς ἂν εὔξαιτο θνητῶν, Franz’s ris ἂν εὔξαιτο βροτὸς ὧν
(accepted by Paley) and the like are faulty because based on the assumption that initial
6v or Bp may lengthen the preceding syllable, cf. Appendix E.
632
 COMMENTARY line 1343
particularly 316 n. ı, and Headlam-Thomson on A. Ag. 1663; cf. also Wila-
mowitz’s note (Berl. Sttzgsber. 1908, 332 n. 1) on the similar use of πότμος.
For δαίμων as almost equivalent to τύχη in the poetic language of the fifth
century see Wilamowitz, Glaube d. Hell. i. 364 n. 1; also Gomperz, Hermes,
lviii, 1923, 44 (partly not relevant) and the survey of Andres, RE, Suppl. iii.
285 f. Add also Pind. N. 4. 84 ὕμνος δὲ τῶν ἀγαθῶν Epyudrwv βασιλεῦσιν
ἰσοδαίμονα τεύχει φῶτα.
In regard to the form of 1331-42 it is worth noticing that towards the end
of a tragedy a brief series of anapaests sometimes serves as a shorter substi-
tute for a stasimon. Cf. W. Kranz, Stasimon, x62.
1343. Blomfield writes ἔχω, saying 'miserabiliter friget istud &ow’. His
criticism has found followers. Weil: ‘ ἔσω, sic nude positum, esset intus (in
domo), non penitus’; he suggested writing πλευρῶν for πληγήν; Wilamowitz
accepted this in his text. Several other conjectures have been made, some of
them grotesque. Paley's remarks suffice to justify the MS reading. Φ 116 ff.
᾿Αχιλεὺς δὲ ἐρυσσάμενος ξίφος ὀξὺ τύψε κατὰ κληῖδα παρ᾽ αὐχένα, πᾶν δέ οἱ εἴσω δῦ
ξίφος ἄμφηκες is indeed much simpler ; however, it seems that the tragedians
ventured farther along the same path and inserted &ow where they wanted to
emphasize that a wound went deep and penetrated the vitals. In E. Ion
766 f. διανταῖος ἔτυπεν ὀδύνα με πλευμόνων τῶνδ᾽ ἔσω the expression is quite
simple because πλευμόνων is added. The passage compared by Hermann is
nearer the use of ἔσω in Ag. 1343: E. Hel. 354 ff. 7) ξιφοκτόνον δίωγμα λαιμορ-
ρύτου odayäs αὐτοσίδαρον ἔσω πελάσω διὰ σαρκὸς ἅμιλλαν; cf. Rhes. 750f. οἵα
μ᾽ ὀδύνη τείρει φονίου τραύματος εἴσω. ἔσω in Ag. 1343 is strengthened by the
idea already expressed in καιρίαν: the thrust or blow has hit the right place
and gone deep.
In the cries of Clytemnestra S. El. 1415 f. ὦμοι πέπληγμαι and ὦμοι μάλ᾽
αὖθις Boeckh, Graecae tragoediae principum . . . num ea quae supersunt genuina
sint (1808), 244, Wilamowitz, Hermes, xviii, 1883, 236 n. 1, and others see a
conscious reminiscence of A. Ag. 1343-5.

The Distribution of the Lines of the Chorus from 1344 to 137I

There has been a great deal of dispute about the number of choreutae
involved in this scene. For the earlier discussion see R. Arnoldt, Der Chor
im Ag. des Aesch. (1881), 65 ff. (he himself wrongly follows Hermann), and
Reisch, RE iii. 2391 (his scepticism is unjustified). Fortunately unanimity
has been reached on this point, a rare phenomenon in these studies. The
view of F. Bamberger (his article was published in 1832), Opuscula philol.
28 ff., and of Otfried Müller, Aesch. Eum., p. 76 £., that the words of this
scene must be assigned to twelve choreutae is now generally accepted, cf., e.g.,
the editions of Weil, Wecklein, Sidgwick, Verrall, and the more recent ones,
and moreover Wilamowitz, Interpret. 175; A. W. Pickard-Cambridge,
Dithyramb, Tragedy and Comedy, 88; Wilhelm Schmid, Geschichte der griech.
Lit. i. 2 (1934), 56 n. 3, 71 n. 5, etc.
Aeschylus has made the distribution of the short speeches as distinct as
possible. 1344 and 1346 f. are obviously recited by the coryphaeus as spokesman
of the undivided Chorus. As Otfried Müller (loc. laud.) observed, the trochees
1 Paley is wrong in taking ἔσω σφαγᾶς λαιμορρύτον together; ἔσω is adverbial, and the
genitive depends on δίωγμα, as A. C. Pearson ad loc. takes it.

633
lines 1344-71 COMMENTARY

mark the fright into which the Elders are thrown by the cries of the dying
king; the excited character of the lines is in keeping with the limitation in
the use of trochaic tetrameters which is noticeable in tragedies of this period
(cf. on 1649 ff.). The subsequent consultation, where a quieter mood prevails,
is suitably delivered in trimeters. From 1348 onwards each of the twelve
choreutae speaks as an individual in his own person. To infer this we need
no stage direction ; the poet has made it unambiguous by placing an emphatic
ἐγὼ μέν, ἐμοὶ δέ, κἀγώ at the beginning of the three first distichs. We do not
know, nor does.it matter, which of the twelve speakers is the coryphaeus.
Modern editors generally assume that it is the last (1370 f.); in this they
follow Bamberger and O. Müller. But those scholars drew their conclusion
from a wrong interpretation of 1370, where they thought that adherence to a
majority of votes was expressed. The splitting of the voice of the Chorus (or
its spokesman) into the voices of its individual members is a bold artifice.
In the extant plays it is unique,! and it is not likely that it was employed
often.
The internal evidence in favour of the distribution of Ag. 1348-71 among
twelve speakers is conclusive. Against it there has been set what Wilamo-
witz, Interpr. 175, terms ‘the ancient tradition’. That tradition is rather
flimsy. The supposed locus classicus which is quoted by Hermann and his
successors, and in the books of reference, is Schol. Ar. Knights 589. The
relevant passage runs thus: συνειστήκει δὲ ὁ χορὸς [6 μὲν κωμικὸς] ἐξ ἀνδρῶν
ἤδη καὶ γυναικῶν, ὁμοῦ δὲ καὶ ἐκ παίδων, [k5', ὡς καὶ οὗτος ἀπηρίθμησεν ἐν
"Opviow, ἄρρενας μὲν ὄρνις ιβ΄, θηλείας δὲ τοσαύτας. ὁ δὲ τραγικὸς te’, ὡς
Αἰσχύλος ᾿ἀγαμέμνονι.] ἔστι δ᾽ ὅτε καὶ ἡμιχόρια ἵσταντο ἤτοι ἐξ ἀνδρῶν καὶ
γυναικῶν κτλ. Ihe bracketed words are missing in the Venetus (in the
Ravennas the scholia on the Knights cease at 214) and in the Laur. conv.
soppr. 140, formerly Badia 2779 (9).^ As we have no critical edition of the
Aristophanes scholia, I cannot say on what MS authority the text of the
words in question rests, nor can I determine with certainty whether or no
they are pre-Byzantine; it seems, however, probable that they are to be
ascribed to Triclinius, [cf. now, after K. Holzinger, Mervyn Jones, C.Q.
N.S. V, 1955, 44]. But be this as it may, Dindorf was obviously right when
he regarded these words as a later addition, for they clearly disrupt the
context. The earlier commentator had no reason to mention any figures
before he came to the proportion of men and women in mixed Choruses:
ἐπλεονέκτει τὸ τῶν ἀνδρῶν μέρος Kai ἦσαν vy’ κτλ. He presupposed as com-
mon knowledge the ordinary number of twenty-four members of the comic
Chorus. It is easyto see where the interpolator, whom Hermann (Opusc.
ii. 131) praised as ‘doctissimum grammaticum’, got his material. As for
his reference to the Birds, Dindorf rightl compared | Schol. Birds 297
ἐντεῦθεν ἀριθμήσας εὑρήσεις τὰ Kd πρόσωπα ἐξ ὧν ὁ κωμικὸς χορὸς συνίσταται.
I suppose that the extra matter in schol. Knights 589, viz. the clause ἄρρενας
1 The unfortunate idea of Weil that the utterances of the coryphaeus Eum. 585-608
should be distributed among the twelve choreutae was adopted by Wecklein and Blass; it
has been fully disproved by R. Arnoldt, Der Chor im Ag. des Aesch. 67 n. x, and Wilamo-
witz, Interpr. 183.
2 On the relation between the scholia of this MS and those of V cf. K. Zacher, ‘Die
Handschriften und Classen der Aristophanesscholien’, Jahrbücher f. class. Phildl., Suppl.
xvi, 1888, p. 548, and, with special reference to the Knights, pp. 710 ff;

634
 THE DISTRIBUTION OF THE LINES 1344-71

μὲν ὄρνις ıB’, θηλείας δὲ τοσαύτας, is derived from a text of schol. Birds 297
which was more complete than that preserved in our MSS. The source of the
remark on the Agamemnon is no less obvious. The long scholion on Ag. 1348
(ZxoA. wad. in Tr) begins thus: πεντεκαίδεκα εἰσὶν οὗ τοῦ τραγικοῦ χοροῦ ὗπο-
κριταὶ καὶ ἕκαστος αὐτῶν δίστιχον γνώμην λέγει" εἰπόντων δὲ τῶν 1B’, πρὶν καὶ
τοὺς πεντεκαίδεκα εἰπεῖν, προλαβοῦσα ἐξῆλθεν ἡ Κλυταιμνήστρα. ἀμίμητον γὰρ
μετὰ τὸ πάντας εἰπεῖν τὰς οἰκείας γνώμας ὥσπερ ἀπὸ συνθήματός τινος τότε
ἐξελθεῖν τὴν γυναῖκα. It is because he remembered this scholion that the
man who is responsible for the enlarged version of schol. Ar. Knights 589
inserted the words ὁ δὲ τραγικὸς ιε΄, ws Αἰσχύλος Ayapéuvore.’ So far, then,
from representing an ‘ancient tradition’, the sentence is a second-hand state-
ment, based on a wholly arbitrary interpretation of the scene in the Agamem-
non. Besides, the way in which the argument is handled by the scholiast on
Ag. 1348 is the strongest possible confirmation of the correctness of Miiller’s
view, as opposed to Hermann’s. The commentator sees perfectly well that
the arrangement of the dialogue requires twelve speakers and twelve only.
But since his authorities (they did not trouble to distinguish between the
different periods of Attic tragedy, cf., e.g., Pollux 4. 108, schol. A. Eum. 585)
told him that the tragic Chorus was composed of fifteen members, he feels
obliged to find an excuse for the recalcitrant fact, and his ingenuity comes out
fairly well from the test. [For schol. Knights 589 see the Addenda.)

1344. On the use of σῖγα as imperative cf. Schwyzer, Glotta, xii, 1923, 28.
1345. πεπληγμένος could possibly be understood in the sense of πεπλ. εἰμίξ
(see on 806), but this is unnecessary, as the nominative can follow ὦμοι in
Tragedy (cf. S. Aj. 340 ὥμοι τάλαινα and elsewhere) just as in Homer (e.g.
Σ 54 ὦ μοι ἐγὼ δειλή, ὦ μοι δυσαριστοτόκεια).
1347. κοινωσώμεθ᾽ ἄν πως cannot be right, as Porson saw. He wrote κοινω-
σαίμεθ᾽ dv πως, but this weakening of the common use of the hortative mood
after ἀλλά, which is here entirely suitable, is intolerable. dv πως (E. A. J.
Ahrens, Hermann) is an artificial expression ; other suggestions need not be
mentioned. Paley’s original conjecture ἦν ws, which he subsequently
abandoned, is probably right. ἦν, which is common in Sophocles and Euri-
pides, is attested in Aeschylus as well: Pers. 708 (cf. [A.] Sept. 1027). In the
subordinate sentence ἀσφαλῆ could perhaps be regarded as marking the
1 It may be as well to compare at least one place where, dealing with a similar matter,
one of the late ‘redactors’ of the scholia on Aristophanes clearly depends on the Aeschylus
scholia: Schol. Wasps 270 (not in V and R) τῶν γὰρ χορικῶν μελῶν τὰ μέν ἐστι mapodına . . .
τὰ δὲ στάσιμα, ὡς τὸ παρόν. καὶ παρ᾽ Αἰσχύλωι ‘ στένω σε τᾶς οὐλομένας τύχας Ilpoundeö’.
Why does the scholiast refer to Prom. 397 of allstasima? Either because there he found the
scholion : τὸ στάσιμον dıder ὁ χορὸς ἐπὶ τῆς γῆς κατεληλυθώς (so in M), or, more probably,
because this commentator was none other than Triclinius (see above), who referred in this
sentence to his own scholion on Prom. 397 (see H. W. Smyth, Harvard Studies in Class.
Philol. xxxii, 1921, 31) : ὅτι τὰ μὲν τῶν χορῶν καλεῖται mapodırd ... ra δὲ στάσιμα ὅσα ἵσταται
. ὡς ἔχει ἄλλα τε πολλὰ καὶ τὰ παρόντα. διὸ καὶ Προμηθεὺς τὸ στάσιμον doa μέλος ταύτας
προτρέπων πρότερον εἴρηκε ‘ πέδοι δὲ βᾶσαι τὰς προσερπούσας τύχας ἀκούσατε᾽.
2 Wilamowitz ad loc.: ‘part. perf. pro verbo finito, cf. Sept. 809’. But in Sept. 809 there
is probably no full stop after κατεσποδημένοι, but we should take it as the breaking off of an
unfinished sentence; so Sidgwick, Tucker, and Pohlenz, Griech. Trag. ii. 27 and Gnomon ix,
1933, 626 n. 1, Sept. 810, deleted by Wilamowitz, is defended by Pohlenz ; Murray, supporting
the deletion, says ‘ δ᾽ οὖν vix sanum’, a view which is at least not shared by Denniston,
Particles, 465.
635
line 1347 COMMENTARY

predicate’ (so, e.g., Paley); but as the ‘ellipse’ of the subjunctive of the
verb εἶναι is uncommon, it would be safer even with this interpretation to
accept Enger’s βουλεύματ᾽ ἦι. Closer inspection shows that the thought
required here is not ‘whether perhaps plans are sound’ but ‘whether perhaps
there are sound plans’; # therefore is necessary. For the way in which this
ἦι may have dropped out see on 1375. The use of ἤν ws is common in Homer
(in X 418 f. λίσσωμ᾽ ἀνέρα τοῦτον... ἣν πως ἡλικίην αἰδέσσεται ἠδ᾽ ἐλεήσηι γῆρας
the ἦν πὼς clause stands after a hortative subjunctive exactly as in Ag.
1347), Sophocles uses ἐάν πως Trach, 584, Oed. C. 1770 (after an imperative).
Keck and Wilamowitz altered dv πως into ἄνδρες. This is farther from the
MS reading; moreover it is unlikely that the members of an Aeschylean
Chorus would have addressed each other as ἄνδρες like the Choruses of
Aristophanes (e.g. Ach. 238). Sophocles uses ἄνδρες (by itself) when an actor
addresses the Chorus (Ant. 162, Aj. 1093, 1318, Phil. 974, Oed. C. 1348); in
Aeschylus, perhaps by chance, the only instance of the plain vocative ἄνδρες
is in the address to the escort, Suppl. 500.
κοινωσώμεθα. Aeschylus uses this verb several times, the middle is found
for the first time here; its exact meaning must be deduced from the context.
With the text advocated above the verb must either have the more general
sense ‘go together to work’ (this would suffice because the more special
nuance ‘take counsel together’, comes out clearly in the édv-clause which
indicates the purpose of κοινώσασθαι) or the more specific sense of ‘meditate
together, take counsel together’, which agrees well with the use of κοινοῦσθαι
elsewhere (cf. L-S s.v. ii. 2, cf. also ἐπικοινοῦσθαι and ἐπικοινᾶσθα!).
1348. τὴν ἐμὴν γνώμην λέγω. γνώμη is not simply 'opinion' here but 'pro-
posal, motion’ (rightly L. Campbell: ‘what I propose’; Wilamowitz : ‘meinen
Vorschlag’). Instances for this use in L-S s.v. III. 2; so γνώμην εἰπεῖν or
λέγειν commonly in Thucydides and Aristophanes (cf. Neil on Knights 267),
especially of a motion made in the Assembly, etc. This use brings out the
essential meaning of the word, for ‘ γνώμη bedeutet nicht nur die Erkenntnis,
sondern eben so gut die Folge der Erkenntnis, nämlich den überlegten,
rationalen Willen' (Ed. Schwartz, Gnomon, ii, 1926, 68), 'das was als erkannt
und demgemäss als beschlossen gilt’ (Wilamowitz, Pindaros, 281 n. 2).
1349. Many editors have correctly understood this. W. Schulze (loc. cit.
ad 1317) says with regard to the present passage: 'The old men do not want to
raise the cry for help themselves but to send messengers to raise it in the town
(E. Or. 1539 ἀγγέλλωμεν ἐς πόλιν τάδε;). The cry for help includes naturally
also the appeal to come πρὸς δῶμα δεῦρο [similarly Schneidewin ad loc.].'?
Usener's treatment of this and similar passages (Kl. Schr. iv. 221 n. 56) has
been superseded by Schulze's article.
1351. σὺν νεορρύτωι ξίφει. It is probably impossible to determine with
complete certainty in what sense veóppvros is used here. In S. El. 894 f.
veoppÜrous πηγὰς γάλακτος the derivation from few is clear; it has been

1 Murray adopts ἦν but construes quite differently : he encloses ἦν πως between commas,
which seems to me stylistically indefensible.
2 A phrase compressed in a similar manner is found in Pind. P. 9. 29 ἐκ μεγάρων Xipwva
προσήνεπε dwväı. On this O. Schroeder observes that ‘we miss here a verb of motion for ἐκ
(= ἔξω) and he describes the passage as ‘a bold example of brachylogy, a much abbreviated
variant of Χίρωνα φωνῶ, στεῖχε δωμάτων πάρος",
636
 COMMENTARY line 1351
assumed for this passage by Stanley and many others, including Wecklein,
Sidgwick, Verrall, Mazon, and G. Thomson. Aeschylus may in fact have
advanced from uses like ῥέε δ᾽ αἵματι γαῖα (Θ 65, O 71 5) to the bold abbrevia-
tion: γεόρρυτον ξίφος, with the result that the expression here has in general
the Same meaning as the Sophoclean (Aj. 30) σὺν νεορράντωι ξίφει. This,
however, is not very likely. Wellauer, C. G. Haupt, and Hermann have put
forward a plea for veóppurov (from ἐρύω 'draw') ; Humboldt translated 'kühn
das Schwerdt gezückt'; so also, e.g., Dindorf, Wilamowitz, Headlam, L-S.
The passage which has often been compared, Eum. 42 αἵματι oralovra χεῖρας
καὶ νεοσπαδὲς ξίφος ἔχοντα, supports this interpretation. Phrases like the
Homeric ῥυτῆρα βιοῦ would help the.recognition in νεόρρῦτος of the connexion
with the verb; cf. also Pers. 147 τόξου ῥῦμα (on this see W. Schulze, Quaest.
ep. 318). More important is the question what is meant here by the expression
σὺν veopp. €. Not only Humboldt, Klausen, and Droysen, but also Wilamo-
witz ('und schaffen uns Gewissheit mit gezücktem Schwert', and in his
edition, Actio 1343 ff.: 'denique strictis gladiis in portam irrumpunt’) and
Headlam (‘put the matter to the proof with the sword drawn’) refer the
word to the swords in the hands of the old men (on whose weapons see on
1650). But this interpretation loses the force of veo- in veóppvrov completely
(the translations quoted above cautiously omit it). On the other hand veo-
has its full sense if it is (as in Eum. 42 νεοσπαδὲς ξίφος) ἃ sign of an act that
has just been perpetrated, and here in particular that would be very much to
the point since what matters most is to catch the doer én’ αὐτοφώρωι. There-
fore the explanation in the interlinear gloss in Tr ἤγουν ἰδόντας τὸ veóppvrov
ξίφος is correct ; similarly Heath, Scholefield, Hermann, and many others.
Several editors regard σὺν νεορρύτωι ξίφει as instrumental, e.g. Kennedy:
‘lit. with (the help of) the fresh-streaming sword’ and Wecklein, similarly
Tycho Mommsen, Beiträge zu der Lehre von den griech. Praepositionen, 608:
‘dadurch dass das Schwert frischblutig ist.’ Such a construction is of course
grammatically possible, see L-S σύν Α 7. But then the sword ought to be the
instrument of the subject of the verb, i.e. the ἐλέγχοντες. On the other band,
it is both natural and effective to take τὸ πρᾶγμα and σὺν v. ξίφει in close
conjunction. Peile's paraphrase, though cumbrous, is helpful: ‘coincidently
with the sword's being newly-bedewed with blood'; Mazon's neat rendering
'et surprendre le crime l'épée sanglante encore' comes to the same. The old
men expect (or fear) to convince themselves of the πρᾶγμα (which may mean
the deed as well as the matter) by seeing the body of the murdered king and
together with it the veóppvrov ξίφος. In order to understand the expression
grammatically we need not resort to Paley's expedient, 'the more full con-
struction would have been ἐλέγχειν τὸν φονέα ξὺν v. ξίφει εἰλημμένον ᾿, It is
quite common to use ovv of things accompanying, or associated with, one
another. Tycho Mommsen (cf. above) failed to see the point here although
it was he who quoted from Pindar and Aeschylus striking examples of this
use, e.g. Ag. 456, 776.
Even nowadays it may still be useful to rule out the insipid objection that
the Chorus could not have known whether Clytemnestra used a sword: ‘as if
their thoughts would not fly at once to the usual instrument of death, which
they would name without hesitation, at the risk of a mistake, in case they
wished to speak of a weapon' (Conington).
637
line 1352 COMMENTARY

1352. On the prosody of τοιοῦτος here and 1360 cf. p. 579.

γνῶμα usually means ὧι τι γιγνώσκεται, ‘token, test’; it is technically used
(cf. γνώμων) of the tooth that marks a child's age (cf. Aristophanis Byzantit
fragmenta, coll. A. Nauck, p. του). Here, on the other hand, and E. Heraclid.
407 the word is equivalent to γνώμη (cf. 1348). Possibly it was Aeschylus who
gave the word this twist, cf. Schuursma, 109 f.
1353. τι δρᾶν: Wilamowitz's translation (‘ich stimme für entschlossene Tat’)
gives the correct colour of δρᾶν. B. Snell, Aischylos, 15, although he over-
emphasizes the point, rightly remarks that in this whole passage ‘dpav denotes
the act by which hovering and indecision are overcome, the beginning . . . of
action' (cf. also Snell, Philol. Ixxxv, 1930, 154). On the other hand, πρᾶγμα
1351 and πράσσειν 1355 are used of the action of those who have now completed
their plan. Cf. 1635 and 1669.
τὸ μὴ μέλλειν δ᾽ ἀκμή: ‘nunc vero adest id temporis punctum, quando
minime cunctari licet’ (Heath). Blomfield compares Pers. 407 κοὐκέτ᾽ ἦν
μέλλειν ἀκμή. Cf. also 5. El. 1338 ἀπηλλάχθαι δ᾽ ἀκμή, E. El. 684 στείχειν δ᾽
ἀκμή and other passages in E. Bruhn, Anhang zu Soph. 146.
1354. Schütz, ‘ ὥς, ut opinor, cum acuto scribendum est, i.q. οὕτως ', so also
Bothe, Wellauer, and even Wilamowitz, who adds: 'procliticum in extremo
verso immane in Aeschylo vitium'. This is wrong. In particular: for ὥς —
οὕτως there is no certain example in Aeschylus, cf. on 93o. In general: Wila-
mowitz has here too (for Eum. 238 f. cf. above p. 588 n. 5) yielded to his un-
willingness to recognize the existence of 'Sophoclean' enjambement in
Aeschylus. The phenomenon is rightly judged, e.g., by E. Martin, De respon-
sionibus diverbit ap. Aesch. (diss. Berlin 1862), 10, 71; Kussmahly, Beob. z.
Prom. des Aesch. (Progr. des Sophien-Realgymnasiums, Berlin 1888), 7;
Blass on Eum. 800f. (though his handling of that passage is mistaken);
P. Maas, Die neuen Responsionsfreiheiten, ii (1921), 7 n. 3, and Griech. Metrik,
appendix to § 136 f. ; cf. E. C. Yorke, C. Q. xxx, 1936, 153 f. ws here cannot be
separated from the accepted instances of ὅτι (also Eum. 98), tva, ὅπως at the
end of a line in the Prometheus ; add the isolated οὗ Pers. 486. For Aeschylean
examples of οὐ or μή at the end of the line in enjambement cf. on 557 (where
again Wilamowitz has removed the enjambement).
1355. rup. σημ. πράσσοντες: 'a singular instance of brachylogy, for πράσ-
σοντες πράγματα (or rather πράξεις) ἃ σημεῖα τυραννίδος cori’ (Paley, similarly
Conington).
τυραννίς is here and 1365 not used, as τύραννος and its derivatives so often
(cf. on 828) are, in a neutral sense, but clearly implies hostility as rupavvis and
τυραννέυω in Solon (fr. 23 D.) ; cf. τύραννος in the Athenian eulogy of the two
heroes who τὸν τύραννον κτανέτην. ‘Aeschylus transfers situations and ideas of
the historical period to the heroic age, just as the dictum in 1365 betrays the
Athenian of the post-Persian-war period’ (Schneidewin).
1356. Blomfield, who chanced to be the first editor of the treatise Περὶ
τρόπων ascribed to Tryphon, saw that the line is quoted there (Rhetores
Graeci, viii. 741. 10 Walz = iii. 196. 24 Spengel) in the form χρονίζομεν ὧδε
τῆς μελλοῦς xapıv! and rightly accepted the varia lectio μελλοῦς as the genuine
reading. Verrall could not bring himself to agree; he saw, however, that τῆς
τ Of χάριν, the last word of the quotation, Weil says, rightly : ‘quod, ut in loco mutilo, ex
ultimo versus sequentis verbo χερί ortum esse videtur’; so too Wilamowitz.
638
 COMMENTARY line 1359

μελλοῦς seems to imply a personification. Headlam elaborated Verrall’s hint ;

cf. also Schwyzer, Griech. Gramm. i. 478: ‘the nomina actionis in - . . . were
probably originally personifications'; lists of such names are collected in
Headlam-Knox on Herodas 4. 6. Against H. L. Ahrens, who conjectured
(De Graecae linguae dial. ii. 389) that Aeschylus took over μελλώ from Sicilian
Doric, W. Aly, De Aesch. cop. verb. 111 f., has rightly pointed out Ionic-Attic
instances of corresponding formations. Meineke’s assertion (Philol. xix,
1863, 211) that ' κλέος here is the same as ὄνειδος ’ and his attempt to relate
the gloss in Hesychius κλέος" ὄνειδος to this passage rest on a misunder-
standing, for κλέος is not in itself disparaging: it means “πὸ lauded name' as
Headlam translates it, who very plausibly infers from the passage a proverb
in praise of cunciari and notes the quotation of such a proverb, the ‘word of
a sage’, in an epigram of Antiphilos of Byzantium, Append. Plan. 4. 136
(Karl Müller, D. Epigramme des Ant. v. Byz., 1935, p. 94 f.) 1. 7 ἀρκεῖ δ᾽ à
μέλλησις, ἔφα σοφός (σοφός, of course, does not mean 'artist' here, as Müller
thinks).
1357. Hermann, who here and Cho. 642 (λὰξ πέδον marospevov) wrote πέδοι,
was soon attacked by, e.g., Blomfield and Conington ; the latter observed (on
Cho. 642): ' πέδον πατεῖν doubtless came to be regarded as a single notion,
as it were a compound verb' (anticipated by Heath: ' πέδον πατοῦντες pro
una voce accipiendum est a qua pendet accusativus κλέος '). More recently
P. Maas, Textkritik (in Gercke-Norden, Einl. 1. d. Altertumsw. i. 2) 14,
showed that the same use is found in Callim. hymn. 4. 227, where Crönert had
earlier restored ἐφετμήν instead of éberuñi as the certainly right reading :! ot
σεῖο πέδον πατέουσιν ἐφετμήν. Maas expressed the optimistic hope that now
the passages in Aeschylus would suffer no further alteration.
οὐ καθεύδουσιν χερί. Sept. 554 χεὶρ δ᾽ ὁρᾶι τὸ Öpdoıuov? has rightly been
compared ; from the context there and here (cf. on 1353) it is obvious that the
hand is regarded as the characteristic organ of δρᾶν. Cf. the passage in the
Sophist Antiphon, fr. 44, col. 3. 8 ff. (Diels-Kranz, Vorsokr. ii, sth ed., 348),
where in a list of the ‘laws’ established for the different organs of the body the
following item is found: καὶ ἐπὶ ταῖς χερσίν, d re δεῖ αὐτὰς δρᾶν καὶ ἃ où dei. In
the antithesis of δρᾶν and βουλεύειν expressly mentioned in 1359 (so also
1634 f.) χείρ might be substituted for δρᾶν, cf. Prom. 619 βούλευμα μὲν τὸ Atov,
᾿Ηφαίστου δὲ χείρ.
1358. βουλῆς: cf. on 884.
1359. πέρι is unintelligible, as Schütz saw, followed by many editors from
Hermann to Wilamowitz. περὶ τοῦ δρῶντος gives no possible sense here nor
does language allow us to understand περὶ τοῦ δρᾶν or δρωμένου with ‘ellipse’
or ‘ad intellectum"; this course was taken by Scholefield and after him by, e.g.,
Verrall (‘to a doer it belongs to advise [?] about the doing’) and Headlam
(it is to the doer that belongs the counsel about the doing’); Hermann
rightly rejected this ('insolentior videtur haec ratio loquendi et quae vix
defendi possit"). The defenders of this view (πέρι = ‘de re gerenda") endeavour
! I cannot see that Wilamowitz's paraphrase (Hellenist. Dichtung, ii. 72 n. 1) of of σεῖο
πέδον πατέουσιν ἐφετμῆι, ‘die in deinem Dienste wandeln’, is warranted by the Greek.
2 Du sublime au ridicule ce n'est qu'un pas: Plaut. Asin. 202 semper oculaiae manus sunt
nostrae: credunt quod vident (for this type of personification see my Plautinisches im
Plautus, 107, and 101 for the same kind of thing in Aristophanes).

639
line 1359 COMMENTARY

to justify it by referring to Cho. 850, where πέρι has ousted the genuine πάρα
(among more recent editors cf. Wecklein, Sidgwick, Wilamowitz, Blass) ;
thus Conington, Headlam, Tucker (all these on Cho. 850), Murray (who refers
also to Sepi. 248, where τῶνδε βουλεύειν πέρι is quite simple). Leaving aside
the difficulty of language, the thought itself is unsatisfying, because what is
required, particularly after βουλῆς, is the antithesis of δρᾶν and βουλεύειν pure
and simple (as 1634 f. and elsewhere) and not the antithesis of δρᾶν and
βουλεύειν περὶ τοῦ δρᾶν. None of the conjectures detailed in Wecklein’s
appendices is really convincing, nor is Platt’s πρέπον (J. Phil. xxxii, 1913,
68). Wilamowitz suggests τέ μή; (or μήν; ? cf. on 672) in place of πέρι:
this has the great advantage that no unnecessary padding is added to the
rounded and complete thought τοῦ 8p. é. x. 7. βουλεῦσαι, but it cannot be
regarded as a certain emendation. There can, however, be no doubt about
the general sense of 1359, for 1360 κἀγὼ τοιοῦτός εἶμι shows that these two
speakers in the main agree (cf. the connecting phrase in 1352). The remark of
the seventh speaker (1360 f.), ‘by mere words the dead cannot be raised to
life’, makes it clear what must be the special point of the opposition between
δρᾶν and βουλεύειν in 1359. On the one side stands δρᾶν, on the other βουλεύειν
and λόγους λέγειν (derogatory when opposed to action). The contrast between
the action of the murderers and the hesitating inaction of the Chorus had
been clearly formulated by the fifth speaker (cf. also the words of the third) ;
but he wanted to stir them to action. The sixth speaker takes up the anti-
thesis in a defeatist spirit: ‘I do not know how to arrive at a useful plan and
decision (βουλή) : to him who employs δρᾶν belongs βουλεῦσαι also (therefore
not to us; we have just been surprised by the δρᾶσαι of the murderers and,
completely powerless as we are, can only talk, not act).’ Wecklein follows
Auratus in reading πάρος for πέρι (Hermann rejected this by reference to the
words of the following speaker) and translates: ‘Who will act, must first take
counsel’ ; this thought has no relation to the context. A good remark on 1359
is given by R. Arnoldt, Der Chor 1m Ag. des Aesch. (1881), 69 n. 1.
1360. κἀγὼ τοιοῦτός εἰμι: in a similar sense κἀγὼ τοιοῦτος E. Heraclid. 266,
Or. 1680 (Peile and Paley).
δυσμηχανῶ only here in place of ἀμηχανῶ, which is common in Aeschylus
and elsewhere; cf. on 1571.
1362. τείνοντες : a fine emendation. Blomfield compares Pers. 708, Prom. 537,
E. Med. 670 ἄπαις yàp δεῦρ᾽ ἀεὶ reivers βίον; the passage of the Medea is
particularly similar in feeling. The thought here is: leading a life which is
not really worth living. Cf. Eur. fr. 1052. 8 f. N. ἡ δ᾽ εὐλάβεια σκότον ἔχει καθ᾽
“Ελλάδα, τὸ διαβιῶναι μόνον det θηρωμένη.
ὧδε: as suggested by the words last spoken.
1363. καταισχυντήρ like αἰσχυντήρ (of Aegisthus, Cho. 990) is one of the many
words formed with this suffix that are peculiar to Tragedy, cf. Ernst Fraenkel,
Nomina agentis, ii. 13 f.
1365. πεπαιτέρα. It is an easy transition from the ripe fruit, which is no
longer hard or sour, to anything mild. Aesch. fr. 264 N. ἀνὴρ δ᾽ ἐκεῖνος
(Hector) ἦν πεπαίτερος μόρων (than mulberries) was adduced by Stanley, cf.
Schadewaldt, Hermes, 1xxi, 1936, 65. Blomfield refers to Eum. 66 and other
passages.
1366 f. ‘Cf. τεκμαίρεσθαι ἐκ τῶν γεγενημένων τὰ μέλλοντα ; τεκμήριον has verbal
640
 COMMENTARY line 1370

force’ (Plüss). The same interpretation of the construction can be found, e.g.,
in Schneidewin and Verrall. Thus: ‘shall we then by inference from his cries
be led to divine . . .?'
1367. τἀνδρὸς ws ὀλωλότος. For this construction in which the genitive is
formally absolute but in connexion with the main verb of knowing, recog-
nizing, saying, etc., fulfils the function of a subordinate clause with örı or ws,
see Lobeck on S. 47. 281, Kühner-Gerth, ii. 93 f.
1368. In place of the impossible μυθοῦσθαιϊ E. A. J. Ahrens's θυμοῦσθαι has
exercised an astonishing fascination on editors, probably because it was
backed by the authority of Hermann, who rejected J. G. Schneider's μυθεῖ-
᾿ σθαι. But θυμοῦσθαι περί τινος is, if possible at all, at least so unusual that to
invent it by a conjecture seems unduly bold. Moreover, the idea of connecting
χρή with θυμοῦσθαι, even when a definite condition is attached to it, seems
very strange. μυθεῖσθαι, on the other hand, in unexceptionable ; it was adopted
by, e.g., Paley and Karsten, and more recently by Ubaldi. This speaker, who,
like his predecessor, wants to avoid any decisive action, maintains that they
have not even exact knowledge of the facts as a basis for preliminary dis-
cussion (he is far from suggesting the possibility of an immediate decision).
He emphasizes the lack of precise information by repeating the phrase σάφ᾽
εἰδέναι. μυθεῖσθαι means here simply 'speak'. Aeschylus and Sophocles prob-
ably borrowed the word from the language of epic, as it appears to be com-
pletely foreign to living Attic (cf. E. Hofmann, Qua ratione ἔπος, μῦθος...
adhibita sint, diss. Göttingen 1922, 40).
1370. ταύτην. The scholiast and many commentators supply τὴν γνώμην.
This is a possible but by no means a necessary supplement, cf. on 916.
πληθύνομαι. We have unfortunately no means of determining precisely in
what sense the verb is used here. The commonly cited passage Suppl. 604 is
an untrustworthy support as it is by no means certain whether there the MS
reading πληθύεται should not be kept, cf., besides the editions, e.g. Ernst
Fraenkel, Griech. Denominativa 65, Pearson on Soph. fr. 718; on the possible
lengthening of the v of πληθύω cf. W. Schulze, Quaest. ep. 344. As regards
the meaning in our passage Blomfield says: ‘Numero augeor ; sed in hoc loco
videtur significare, mullis argumentis urgeor’. Similarly Passow s.v., Hermann
"undique conveniunt mihi argumenta, ut hanc sententiam probem’, and of later
commentators e.g. Wilamowitz 'dem stimm' ich bei, denn alles spricht dafür'.
That is possible. ἃ slightly different, and perhaps more accurate, rendering
is given by Conington: ' "I am multiplied on all sides', a sort of exaggeration
to express the readiness with which the speaker throws all his weight into this
particular scale... “This is the side I choose with all my votes” .. . hanc in
sententiam omnibus pedibus transeo'. The meaning, then, would be that which
the scholiast expresses: ἐπαινοῦμεν διαφόρως ταύτην τὴν γνώμην. Enger (in the
Glossarium of his edition of 1855) compares this passage with Hdt. 7. 220. 2
τὴν γνώμην πλεῖστός εἰμι (the same 1. 120. 4; also 5. 126. 1 is similar), cf. H. Stein
on the passage: Ἵπ a case of doubt the contemplating mind seems to divide
itself equally into two or more parts, of which one finally becomes the pars
maior or maxima ; this, transferred to the thinking subject itself, gave rise to
this turn of speech which is only found in Herodotus'. Accordingly I have
tried in the translation, by using the verb 'prevail, to imply that the
τ Verrall: ‘in such a case the strangeness of the formation is its merit’.
4872.3 M 641
line 1370 . COMMENTARY

πληθύνομαι ἐπαινεῖν is the outcome of a struggle between conflicting views in

the man's own mind.
It is difficult to imagine how πληθύνομαι could have come to mean 'I follow
the majority' (Paley, Wecklein, L-S, and many others) ; nor is it true that
the majority agree with the proposal of the last speaker. The remarks on this
point by R. Arnoldt, Der Chor im Ag. des Aesch. 7o, are pertinent.
1371. κυροῦνθ᾽ ὅπως : ‘dictum esse pro ὅπως κυρεῖ [so already Blomfield]
patet, licet usus probari non possit' (van Heusde). A parallel would be useful.
The expression cannot be regarded as an abbreviation of κυροῦνθ᾽ ὅπως κυρεῖ,
as Plüss, following Doederlein, Peile, and others, still explains: first, the
context requires that ὅπως should be taken as interrogative (Conington is
right), secondly this kind of euphemistic formula, which ‘by its silence hints
at the worst' (cf. on 1171), is unsuitable here where it is implied that perhaps
Agamemnon may be still alive. Of course, the phrase κυροῦνθ' ὅπως expresses
in veiled terms what the speaker refuses to say in so many words ‘we must
know whether he is still alive'. Murray's view (cf. his book Aeschylus, 227)
that 'the sentence appears to be unfinished' seems to me highly improbable
in view of the fact that, however suddenly the door opens, 1. 1371 formally
marks the end of the scene.

Arthur Platt added a footnote to his translation in which, after speaking of

the lines 1347-71 in the most contemptuous terms,! he declared that the only
charitable thing he could do was συγκαλύψαι τὴν γύμνωσιν τοῦ πατρός by not
translating the passage at all. But a scholar who has set himself the task of
writing a plain commentary is not at liberty to indulge in such fastidiousness
even if he wished to do so; he must endeavour to understand and explain the
words of his text as best he may.
The first thing, then, which we have to bear in mind is that the dispute of
the Elders has no result whatsoever. No summing up takes place, nor is any
attempt made to count the votes (if votes they are) or estimate which way the
majority tends. The idea of an orderly procedure with ἐπιψηφίζειν and
ἐπικυροῦν (so Klausen and Otfried Müller, Aesch. Ewm., p. 77, who were
followed, e.g., by Wecklein) is without foundation in the text. Neither is it
true that the dispute of the old men leads up to a violent rush towards the
door of the palace. That was the opinion of Otfried Müller (Aesch. Eum.,
p. 76 f.), of Hermann ('coryphaeus . . . se ducem praebet choro in aedes
irrupturo’), and of many of their successors, down to Wilamowitz (strictis
gladiis in portam irrumpunt, sed horrore perculsi statim recedunt', cf. also
his Interpr. 175), Murray (who, in his translation of 1370, interpolates the
words: ‘Break in, then!) and A. Y. Campbell (translation). This misrepre-
sentation is harmful, for it distorts a vital point of the scene. If the poet
wanted the Chorus to rush towards the house, he would certainly have
inserted in 1370 f. some imperative or subjunctive to that effect, i.e. something
that would correspond to Murray's translation. In ancient dramatic litera-
! He was given to flippant remarks (see the posthumous edition of his essays) and prob-
ably did not expect them to be taken seriously.
2 A warning not to take 1370 πληθύνομαι as ‘I follow the majority’ is given above.
3 Of course no inference as to the action required at r371 can be drawn from the fact that
at 1350 one of the Elders holds that a quick ἐμπεσεῖν would be the best course.
642
 THE DISPUTE OF THE ELDERS 1346-71
ture it is never allowable to invent stage directions which are not related to
some definite utterance in the dialogue. As the text stands, the plan of the
scene is unmistakable: the deliberations of the Elders are abortive. Nothing
comes from them, neither a decision nor any action or movement. Attheend
of 1371 the old men behave exactly as they did at 1348: they stand about in
excited groups and—talk. While they are still busily engaged in giving the
matter their earnest consideration,! the door is thrown open, and we see
Clytemnestra, the bath in which the body of Agamemnon lies, and near by
the body of Cassandra.
In this scene, then, Aeschylus has preferred complete inactivity on the part
of the Chorus to the show of an initiative which would anyway have come too
late. Effectual interference was of course impossible; the groundwork of the
story had long ago been fixed. We may, however, ask: Why did the poet at
this fatal moment present the Elders in helpless torpor instead of allowing
them at least the modest gesture of determination which many of his modern
readers have interpolated? The question cannot be answered with confidence.
It may be assumed that Aeschylus and his audience did not care for the
belated display of an energy which could not have the faintest influence on the
course of events. It is also conceivable that the inactivity of the Elders in
this scene is meant to set off the surprising vigour with which they rise against
the insolence of Aegisthus at the end of the play.
Although the dispute of the choreutae does not result in any action, it
serves a definite purpose. Somehow the complete inactivity of this body of
loyal old men had to be accounted for. A different course might have been
open to the poet if indeed, as has been asserted, the Chorus in the Agamemnon
were nothing but an accompanying instrument. In that case the cries of the
king (1343 ff.) might have been answered by a few words expressing stupefied
horror without suggesting the possibility of an intervention on the part of
the Elders. But that view of the character of the Chorus has proved quite
indefensible (see p. 248). The problem with which Aeschylus is faced in the
scene after the murder arises directly from the part which he intends the
Elders to play in the action. According to the poet’s design the old Argives
have long known that evil things were going on in the house of Agamemnon
and that the king on his return would meet with enmity. They warn him as
far as they can through the herald (615 f.), and, when he arrives himself, they
endeavour in unambiguous terms to rouse his suspicion and put him on his
guard (788 ff). There can be no doubt about their devotion to him. Of the
extent of the peril they are not aware until Cassandra brings it home to them.
At last their incredulity gives way. As their words at 1338 ff. show, they are
now practically convinced that Agamemnon is going to be murdered. After
all these antecedents it would have been impossible to confine the Chorus at
1343 ff. to a few lamentations. It had to be made understandable how it comes
that, despite their loyalty and all the information they possess, the Elders are
not doing anything. To cope with the difficulty, the poet invented the dis-
pute in which he made the twelve choreutae give their views one by one, a
bold and uncommon device (see p. 634). In this dispute the attitude of the -
old Argives towards Agamemnon remains unaltered. Not one of them makes
the faintest attempt to find an excuse for the assault. The majority (the
! On this point I fully agree with Verrall, Introd., p. lvi.
643
lines 1346-71 COMMENTARY

speakers 1-5, 8, and 9) are in favour of immediate action. Of the five who are
not, two (6 and 7) show themselves wholly despondent while the three who
speak last insist on the need of first finding out the truth of the matter. But
not one of them goes beyond mere words. While they are tarrying, the great
active woman has time to achieve her purpose. It is conceivable that in
certain phrases of the more cautious speakers we hear an echo of that
bitterness or at any rate disillusionment which Aeschylus probably felt like
any other man who during a long life has had to listen to political debates at
moments of crisis. If that be so, the poet has certainly not allowed those
recollections to carry him farther than was required by his subject.

From 1372 to the end of the play some part of the interior of the house is
visible. Clytemnestra, in one of her first sentences (1379), points to the fact
that she is standing indoors; the presence of Agamemnon’s body is men-
tioned again and again (1404, 1438 ff.—1440 ff. the body of Cassandra lying
near by—etc., and later on, e.g. 1581, 1603, 1611, 1613, etc.), nor is the presence
of the fatal garment forgotten (1492 = 1516, 1580), and special attention is
drawn to the bath (1539 f). So there can be no doubt about the setting
intended by the poet. What we do not know is what machinery was used in
the Athenian theatre for turning the dramatist’s intention into reality.
Otfried Miiller (Aesch. Eum. p. 77, 103) assumed that it was achieved by
means of the ἐκκύκλημα ; and his view, accepted, e.g., by Peile, Franz, Her-
mann (Opusc. viii. 164 = Aesch. ii, and ed., 652), became more or less com-
munis opinio. The books of reference followed suit, cf., e.g., Albert Miller,
Lehrbuch der griech. Bühnenaltertümer 144; A. E. Haigh, The Attic Theaire,
3rd ed., 203. There is, however, a dissenting minority. We need not dwell on
the odd hypothesis of Reisch, RE v. 2205, who thinks that the whole ap-
paratus, bath, bodies and all, is carried in front of the house from within,!
and that then Clytemnestra takes her stand by the bath. This subterfuge is
barred by the clear words 1379 ἕστηκα δ᾽ ἔνθ᾽ ἔπαισα. But it must be admitted
that the hypothesis which supposes the ἐκκύκλημα to have been employed
in the last scenes of the Agamemnon seems to entail considerable incon-
veniences, as was pointed out by Wilamowitz, Interpr. 175 (cf. also the stage
directions in his translation). Later on Wilamowitz, ‘Die griech. Tragoedie
und ihre drei Dichter’, Griech. Tragoedien, xiv, 1923, 33, again took into
account the possibility that the ἐκκύκλημα was used here, but did not venture
a definite statement. The limit of scepticism was reached by Bethe, Rhein.
Mus. lxxxiii, 1934, 21 ff., who, in a complete recantation of his former views
(Prolegomena zur Geschichte des Theaters, 100 ff.), denied that in the fifth
century the ἐκκύκλημα existed at all. I do not feel qualified to deal properly
with this complex problem. Fortunately, while the technical details of the
staging escape us, we have the most reliable evidence for the horritying
picture which the poet spreads before our eyes.
ı This was Bothe's view too, but he wrote before Otfried Müller.
2 The high-handed rejection of Wilamowitz's opinion by Eugen Petersen, Die att.
Tragódie, 599, cannot be accepted.
3 Much has been written on it. Of the earlier discussions, E. Bodensteiner's article, Neue
Jahrbücher f. class. Philol., Suppl. XIX, 659 ff., is still worth reading. As regards the em-
ployment of the ἐκκύκλημα in the Agamemnon, he carefully balances the pros and cons
(660 f.) without reaching a decision.

644
 COMMENTARY lines 1375 f.
1372. καιρίως : 'so as to be suitable to the moment or to the issue in hand',
cf. λέγειν τὰ καίρια. [For the beginning of the speech see Glotia, xxxix, 1960, 2.]
1373. οὐκ ἐπαισχυνθήσομαι perhaps deliberately echoes 856; cf. on 614.
1374 f. ἐχθροῖς... φίλοις δοκοῦσιν εἶναι: this recalls 788, 797 f., 840.
1375. πημονῆς is a necessary alteration, ἀρκύστατ᾽ av probable. Schütz
demanded dv and therefore wrote φράξει᾽ dv; Elmsley (on E. Med. 417 note e) .
moved ἄν to the end of 1375. The potential optative without dv in a question .
(πῶς ydp τις κτλ.) would not be impossible (cf. p. 620 n. 2), but the number of
instances is so small that it does not seem wise to assume another where
emendation is easy." A further gain is the plural ἀρκύστατα as in Eum. 112,
S. El. 1476; for the singular of the noun, perhaps by chance, is unexampled in
classical Greek (on the use in Pollux see below). It is not uncommon that at
the end of a line after an elision a monosyllable is supplanted by the full
ending of the preceding word: so dv has been lost Cho. 591 (and probably Ag.
1347 qe); Ar. Ach. 709 the correct form of the line ending ἠνέσχετ᾽ àv is pre-
served only in Etym. M. 180. 35 (rightly adopted by Elmsley), while the MSS
of Aristophanes give ἠνέσχετο. At the end of E. Hec. 742, where the two best
MSS have προσθείμεθ᾽ ἄν, in other MSS only προσθείμεθα is left.
Hermann accepted the alterations πημονῆς and ἀρκύστατ᾽ àv, but thought
a further change necessary, 'quia sententia non integra esset, sed referretur
ad id quod non est dictum, 1st faceret quod ego feci, ut alia dicere. quam
sentiret. In point of fact the idea ‘else, otherwise’ is easily ‘understood’. It
is true, the 'parallel' adduced by several commentators (e.g. Schneidewin,
Enger in his re-edition of Klausen) Eum. 607 πῶς γάρ σ᾽ ἔθρεψεν ἐντὸς κτλ. is
unsuitable for comparison (so rightly Karsten) because it is not a hypo-
thetical sentence, but in general Schoemann, Ofusc. iii. 151, says correctly:
'reticetur haec condicio ("nisi sic, ut ego, in tempore dissimulet"), utpote
facile e contextu subintelligenda, usitato Graecorum more in huiusmodi
enuntiationibus post ydp et ἐπεί, quum superius dictorum ratio redditur'.
The handling of this use in Kühner-Gerth, ii. 483 f., is too brief; it may
suffice here to refer to the examples quoted above on 1022-4, Bacchyl. 5. 94 ff.
καὶ yàp av... παῦσεν, Pind. Ol. 9. 29 ff. ἐπεὶ. . . πῶς dv (cf. Ag. 1374 1) . . .
τίναξε; for further instances see Headlam on Herodas z. 72.
For ἄν at the end of a line when it precedes the verb cf., e.g., Prom. 441,
S. Phil. 46, Oed. C. 1579.
τὸ ἀρκύστατον means in Pollux 5. 36 (the section begins with στήσασθαι τὰς
ἄρκυς, cf. [Xen.] Cyn. 6. 5 ff.) the place in which the nets are set, cf. Hesychius
ἀρκύστατα' οἱ τόποι, ἔνθα ai ἄρκυες πήγνυνται. But here and in Eum. 112 and
S. El. 1476 ἀρκύστατα means the net (or nets). The word shows more clearly
than the analogous expressions 358 ff. that here, as in Eum. 112, the poet
has in mind hunting and not fishing. 1382 belongs to a new complex of
thought which begins with 1379.
πημονῆς epexegetic: the nets represent the damage done to the enemy
(on the meaning of πημονή cf. Wilamowitz, Verskunst, 504 n. 1).
1375 f. “ὕψος is in apposition to πημ. ἀρκ. φράξαι, the sense being : through the
setting of the net a high fence is formed over which no jumping is possible'
(Schneidewin). For this type of apposition cf. on 47.
1 Sidgwick, who had especially studied the potential optative without dv, accepted
Elmsley's emendation.

645
line 1376 COMMENTARY

1376. ὕψος does not occur before Aeschylus (only here) and Herodotus. The
noun seems to have been formed on the basis of ὕψιστος, ὑψίων, οἱ. Wacker-
nagel, Unters. z. Homer, 214.
1377. ἐμοὶ δέ at the beginning marks the transition from the generalization
to what a person wants to say about himself, as 5. Ant. 1196, A7. 487, 678
(‘he first speaks in general, "we", i.e. weak men, in contrast to the δεινά.
Only with ἐγὼ δέ is the application made to himself’, Wilamowitz, Hermes,
lix, 1924, 249 f.), E. Med. 225, and elsewhere.
πάλαι goes with οὐκ ἀφρόντιστος, which has coalesced into a single phrase of
strong positive meaning; for the combination of a negative particle with a
compound formed with a privative prefix cf. Wackernagel, Syntax, il. 297 f.
1378. νίκης is unmistakably corrupt. The majority of editors accept Heath’s
conjecture νείκης. Whether ἡ νείκη can be ascribed to the Greek of the fifth
century is uncertain. The existence of the noun, which seems to be attested
by the lexicographers (Et. M. 276. 4 νείκη" ἡ φιλονεικία, ἐκ τοῦ νεῖκος, Suid.
N 286 νείκη: ἡ φιλονεικία) was questioned by Blomfield and vehemently
denied by Wilamowitz. There are traces of its existence in later times.
Timon (fr. 14 Wachsmuth, 21 Diels) seems to have invented a daemon
Νείκη as sister of Eris (cf. Wachsmuth ad loc.; Diels does not follow him),
but from this it does not necessarily follow that the word ἡ νείκη then
existed. A stronger argument is this: in the parody of Homer quoted by
Dion of Prusa, 32. 82 1. 4, the reading ὑπὸ vetxns (Wachsmuth), in the sense of
φιλονεικίας, is much more satisfactory than the MS reading ὑπὸ νίκης, which
v. Arnim prints without comment.? Our judgement about the existence of
νείκη in the fifth century depends entirely on the correctness or otherwise of
the reading of the Marcianus veikas at E. Or. 1679; the conjecture 19h. T.
813 must be left out of consideration (if it is adopted, vecxn may be plural as,
e.g., Andr. 1165). These things being so, the possibility of Aeschylus having
written veiens cannot be definitely denied; the word would here mean not
φιλονεικία but νεῖκος. ἀγὼν νείκης παλαιᾶς would be ἃ good expression, cf.
S. 41. 1163 ἔσται μεγάλης ἔριδός τις ἀγών, with which Lobeck compares E.
Herachd. 198, Andr. 725 μάχης ἀγών (the phrase occurs also S. Trach. 20).
Perhaps we should rest content with this. I must, however, confess that
I am tempted by Pauw’s δίκης. Clytemnestra, who said (911), with an
ominous double sense, ‘may Dike bring Agamemnon to his home’, now,
when the murder has been achieved, repeatedly lays the greatest stress on
the element of retributive justice in her deed: 1396, 1406, but particularly
1432 f. μὰ τὴν τέλειον τῆς ἐμῆς παιδὸς Δίκην Ἄτην "Epwiv θ᾽ αἷσι τόνδ᾽ éada£
ἐγώ. What is directly expressed there would be, if δίκης is correct, unmis-
takably hinted in the present passage: the Dike of Iphigenia and the δίκη
παλαιά would be one and the same. δίκη would here as in 1432 and often
elsewhere (cf. on 813) signify the just claim as well as the justice of the claim.?
ἀγὼν 60e . . . δίκης παλαιᾶς ἦλθε: now at last the battle over and for the old
δίκη has come about. The meaning of ἀγών ‘decisive battle, decision’ (cf.
1 A, B. Cook, Zeus iii, p. 811 n. 1, discussing Hesiod Theog. 384, considers the possibility
of reading Ζῆλον καὶ Νείκην κτλ.
2 On the other hand, in the epigram in Paus. 5. 2. 5 quoted by Passow and L-S νείκη is
probably plural.
3 Pl. Crito 45 e ὁ ἀγὼν τῆς δίκης is quite different.
646
 COMMENTARY line 1382

Passow-Crönert, 78. 10 ff.) also enters in. But in spite of these considerations
νείκης, which is practically the same as the MS reading, must be accepted if
the use of νείκη = νεῖκος may be assumed for the classical period. So the
first word of the line remains uncertain, but this only, and it is quite unjusti-
fiable to make further alterations in the sentence. The sequence of πάλαι...
παλαιᾶς ought never to have been tampered with. More helpful than a
comparison with E. Or. 811 πάλαι παλαιᾶς ἀπὸ συμφορᾶς δόμων, where the style
is completely different, is a true appreciation of Clytemnestra’s main concern.
Now that she has at last reached her goal, she strongly emphasizes that
nothing sudden, nothing insufficiently grounded or prepared, has occurred.
For her the decisive battle, long thought out and planned by her, has come
in a legal process (or feud?) which reaches far back. The emphatic πάλαι...
παλαιᾶς provides the background from which the expression ‘at last’, σὺν
χρόνωι, stands out, the opposition being marked by the adversative ye μήν
(cf. Denniston, Particles, 348). The antithesis is blurred if in the place of
παλαιᾶς a form of τέλειος (Karsten, and also Wilamowitz) is introduced by
conjecture. The various attempts to split up the sentence 1377 f. are wholly
unjustified: Weil, who was followed by Kirchhoff and Murray, begau a
new sentence with ἦλθε, Wilamowitz (and after him Mazon and A. Y. Camp-
bell) put a stop after πάλαι. The sentence is perfect in construction; and to
sever ἦλθε from the beginning of the clause is wrong also because Aeschylus
has a marked fondness for the connexion of ἐλθεῖν with the dative of the
person affected by the event: Dindorf, Lex. Aesch. 134, gives nine examples.
1379. ἐπ᾽ ἐξειργασμένοις: also Pers. 525; Hdt. 4. 164. 3; 8. 94. 4; 9. 77. τὶ
Soph. Aj. 377. These parallels show that the phrase is a fixed formula. It
would therefore be unwise to follow those translators who, for the sake of
picturesqueness, take ἐπί in a local sense (‘over the finished work’).
1380. οὕτω δ᾽ ἔπραξα: ‘Thus have I carried it out; thus have I succeeded’
(Snell, Aischylos, 13). πράσσειν often indicates reaching the goal (Snell 12),
cf. 363, 369; a classic instance is E. Her. 1305 ἔπραξε γὰρ βούλησιν ἣν
ἐβούλετο.
It is difficult to understand why Wilamowitz transposed 1381 to after 1383.
Weil too (1858 ; later he corrected this), Dindorf (editio quinta), and Sidgwick,
who put a colon after 1380 and a comma after 1381 so that 1381 is taken
together with what comes after it, destroy the train of thought; they appar-
ently take 1381 in the same way as Wilamowitz (‘indicat retis usum’). And
yet the arrangement given by the poet is perfectly clear. First the speaker
anticipates the climax of the whole action (as far as uópov)—this manner of
introducing an account is common in messenger speeches and elsewhere—
then (1382) the account proper starts from the beginning. This fresh start is
appended without conjunction because ἄπειρον ἀμφίβληστρον κτλ. elaborates
and explains the preceding sentence.
1381. ὡς «rA.: οὕτως followed by ws with the infinitive as, e.g., Pl. Rep. s.
477 a εἰ δὲ δή τι οὕτως ἔχει ὡς εἶναί τε καὶ μὴ εἶναι; More commonly wore, cf.,
e.g., S. Trach. 1126 ἔχει γὰρ οὕτως ὥστε μὴ σιγᾶν πρέπειν.
1382. On ἀμφίβληστρον cf. in addition to the lexica W. Aly, De Aeschyli copia
verborum, 37 f. (the confused remarks of Schuursma 20 are not helpful). In
the language of everyday life the word seems to have meant only ‘fishing-net’,
originally and probably predominantly a ‘casting-net’ (cf. A. W. Mair, edition
647
 line 1382 COMMENTARY

[Loeb Library] of Oppian, Introd. x1 £). There is no evidence that it could

ever, by itself, have indicated a garment. As the festal robe becomes the
instrument with which Agamemnon is caught, it is called here and Cho. 492
ἀμφίβληστρον ; the same is true of 5. Trach. 1052 (cf. below on 1580). Again in
Prom. 81 the idea that Prometheus is now ‘caught’ probably enters in; but
Aeschylus, who was in the habit of obtaining a special meaning through
etymological reinterpretation of a word (cf. on 149 ἐχενῆιδας and elsewhere),
probably also intended to stress the ἀμφιβάλλειν (cf. Pr. 52 τῶιδε δεσμὰ
περιβαλεῖν) of the fetters. In this artificial use of ἀμφέβληστρον Aeschylus was
followed by Euripides in Ih. T. 96, perhaps also Hel. 1079. In Ag. 1382 and
Cho. 492 the word was undoubtedly chosen because of the connotations of
ἀμφιβάλλειν ; it is obvious that another name for a fishing-net would not have
served the poet’s purpose. We have already noticed in general (on rrog) that
the whole conception of Agamemnon’s murder in the Oresieia rests on
premises that are characteristically ‘Homeric’. In the world of Epic, though
not in Athens, women in high society do occasionally themselves attend a
man in the bath. This has a particular relevance for the part which the
festal robe (its richness is repeatedly stressed in the trilogy, cf. on 1116 and
p- 514 n. 2) plavs in the murder. It may be useful to recall the typical Epic
account of the successive stages in the bathing of a noble guest: y 464 ff.
τόφρα δὲ Τηλέμαχον λοῦσεν καλὴ Πολυκάστη, Νέστορος ὁπλοτάτη θυγάτηρ
Νηληϊάδαο. αὐτὰρ ἐπεὶ λοῦσέν τε καὶ ἔχρισεν Aim’ ἐλαίωι, ἀμφὶ δέ μιν φᾶρος καλὸν
βάλεν ἠδὲ χιτῶνα, ἔκ ῥ᾽ ἀσαμίνθου βῆ; cf. κ 364 f., 449 ff. where Circe washes,
anoints and clothes first Odysseus and then his companions, cf. also the
passages where this is done by handmaids, ὃ 49 ff., 0 454 ff., p 88 ff., J 154 ff.
Sometimes χλαῖναν καλήν and the like instead of φᾶρος καλόν is combined with
χιτῶνα. Therefore during the bath not only a chiton but also a fine outer
garment (the φᾶρος mentioned in this connexion clearly serves as a himation)!
is kept ready. When the bather has been dried and anointed in the bath,
these garments are put round him and then finally he gets out of the bath.
The fine garment which plays such a prominent part in the bath scenes is to
be worn by the guest at the subsequent feast : in all the Homeric descriptions
the bath is followed by a meal. According to the Odyssey (A 409 ff.) Agamem-
non was murdered during the feast. In the play of Aeschylus the preparations
for the festal meal remain in the background but are clearly hinted at (1056 f.,
1310). The ceremonial robe is called φᾶρος Cho. 1011, Eum. 634; for other
names see on 1116. Clytemnestra naturally does not wait until her husband
has been anointed and stands upright, but throws the robe over him while he
is still crouching in the short and narrow 8poiry, which leaves as little space
for free movement as do the ἀσάμινθοι found in Knossos and elsewhere.”
What Clytemnestra does with the φᾶρος is the normally harmless but here
fatal ἀμφιβάλλειν (mentioned in all the Homeric passages) ; therefore the poet
! On the nature of the Homeric φᾶρος, an expensive and usually very large cloak, see
W. Helbig, Das homerische Epos, 2nd ed., 193 f.
2 C. Robert, D. griech. Heldensage, 1297 f., discussing the content of the story as presented
by Aeschylus, says: ‘thus when refreshed he steps out of the tub, she throws a large robe
like a fish-net over him... . Then Clytemnestra strikes him two deadly blows on the head,
and a third as he lies dying on the ground’. Robert has here failed to get a true picture of
what happened, and has in particular disregarded 1540, which makes it perfectly clear that
Agamemnon was struck while in the tub.
648
 COMMENTARY line 1382

has a very good reason for calling the garment ἀμφίβληστρον ; this special
meaning and the everyday meaning of a net for catching fish are brought
into an uncanny connexion.
ἄπειρον. Stanley translated inextricabile, Blomfield preferred the non-
committal interminabile. Stanley supported his rendering by producing
evidence (Lycophron 1099 ff.,! Schol. Hom. A 7, Schol. E. Hec. 1277 Dindorf,
which should be replaced by the older? Schol. E. Or. 25) for the version of the
story according to which the fatal garment employed by Clytemnestra was a
χιτὼν οὔτε ταῖς χερσὶν οὔτε τῆι κεφαλῆι ἔκδυσιν ἔχων. For that version cf. also
[Apollod.] epit. 6. 23, Seneca, Ag. 888 f.; for the representations on Etruscan
urns see C. Robert, Heldensage 1297, note 7. The idea of the συνερραμμένος
πανταχῇ χιτών (Porphyr. loc. cit.) or χιτὼν ἀτράχηλος is certainly alien to the
Oresteia,3 but it was almost inevitable that later readers should apply to the
words of the earlier play what in their own time was the recognized tradition.
So at Eum. 634 ἀτέρμονι (πέπλωι) we read in the Mediceus the crude inter-
linear gloss ἀτραχήλωι (cf. [Apollod.] epit. 6. 23 δίδωσι yap αὐτῶν χιτῶνα
ἄχειρα Kal ἀτράχηλον), and at Ag. 1382 ἄπειρον we find the gloss azpyrov in
Tr. In the same manner Stanley harmonized Ag. 1382 with the later version ;
he was followed by Passow s.v. ἄπειρος and several commentators, e.g. Paley:
‘giving no exit to the head and hands’. Faithful to the principle contaminart
non decere fabulas, we may feel inclined to reject any special meaning of
ἄπειρον and be content with ‘boundless, huge’. An uncommonly large size of
the treacherous garment seems to be indicated by Eum. 634 f. dréppow. ..
πέπλωι and is clearly presupposed by the action which takes place at Cho.
983 ff. (cf. Appendix C). We might, then, simply assume that ἄπειρον in a
hyperbolic way serves to convey the idea of enormous width. But another
Aeschylean passage makes us pause, Prom. 1078 eis ἀπέραντον" δίκτυον ἄτης.
This phrase strongly suggests that, besides extension, some notion more
peculiar to a net was implied. Our suspicion increases when we turn to
Ibycus fr. 7. 3 D. ἐς ἄπειρα δίκτυα. There the idea of mere size would be
wholly inadequate: the context requires something like inextricabile, and
that is how it was rendered (‘dem man sich nicht entwindet’) by Wilamowitz,
Sappho und Simonides, 125, who compared the passage with Ag. 1382 and
adopted for ἄπειρον the paraphrase of the ancient lexicographers 6 οὐκ ἔχει
πέρας. Taking a step farther back, I venture to suggest that Ibycus and
Aeschylus may, at any rate to a certain degree, have been prompted to their
bold reinterpretation of ἄπειρον, ἀπέραντον by the model of the famous
1 Lycophron, in his usual enigmatic way, makes no definite statement about the neck
and arms, but his expressions leave hardly any doubt that he has in mind the χιτὼν συνερ-
ραμμένος.
2 To the passages in the Etymologica quoted by Ed. Schwartz ad loc. add Porphyr. ad
Hom. & 200, p. 191. 20 ff. Schrader.
3 The line quoted by the scholiast on E. Or. 25, ἀμήχανον τέχνημα καὶ δυσέκδυτον, which
probably had its place somewhere in the Oresteia (cf. above p. 513 ἢ. 3), contains no hint
of a χιτὼν drpayndos. The words would suit any wide robe, e.g. ἃ ἱμάτιον. On the other
hand, Soph. fr. 483 N. = 526 P. (from the Polyxena) χιτών σ᾽ ἄπειρος ἐνδυτήριος κακῶν
seems to imply some sort of χιτὼν drpaxnAos, while E. Or. 25 ἀπείρωι περιβαλοῦσ᾽' ὑφάσματι
provides no clue as to the type of garment.
4 Neither the testimony of the MSS nor A. Suppl. 1049 ἀπέρατος, where the meaning is
quite different, seems in the least to justify the adoption by Wilamowitz of ἀπέρατον in
Prom. 153 and 1078.

649
line 1382 COMMENTARY

δεσμοὶ μὲν τρὶς τόσσοι drreipoves (0 340).' That this areipoves caused considerable
trouble to later interpreters is obvious from the scholia on the passage and
especially from Porphyrius on & 200 (p. 191 f. Schrader). It is a probable
guess that, although ἀπείρονες (‘countless’) as a supplement of τρὶς τόσσοι is
not really difficult (cf. r 174 = Hesiod fr. 134. 4 Rz. πολλοὶ ἀπειρέσιοι), at a
very early period readers of Homer may have been puzzled by it. The
apparent ambiguity of Seopol . . . ἀπείρονες may have led to the assumption
that the epithet here (and perhaps in similar passages of epic poetry lost to
us) meant ‘where there is no end, i.e. no exit or outlet’.” Such an interpreta-
tion might easily have the consequence that some poets, in connexion with
words like bonds, net, key, etc., used the adjective ἄπειρον, ἀπέραντον with
the connotation or even the primary meaning of ‘inextricable, without
outlet’. That would account for Ibycus fr. 7. 3, A. Prom. 1078, Ag. 1382,
Soph. fr. 483 N. = 526 P., E. Or. 25, and also for E. Med. 212 πόντου κλῆιδ'
ἀπέραντον.
1383. περιστιχίζω: if it were not guaranteed by metre, this by-form of ἃ
verb, otherwise always read as περιστοιχίζω, would scarcely win credence.
The indicative in Tr (presumably based on conjecture) is necessary, although
attempts have been made to defend the participle.
The many instances of the present tense from here to 1390 βάλλει are
noticeable ; they probably serve to enliven the recital (so B. Huebner, Diss.
Phil. Hal. iv. 2, 1880, 140). Clytemnestra lives and acts the whole story again
while she tells it. Miming gestures are likely to have supported the vivid
description of each stage. If we want to appreciate the character of this use
of the present in Aeschylus, we may contrast, e.g., Io’s narrative (Prom.
645 ff.), which is entirely in the past tense, or the much longer account of the
battle of Salamis (Pers. 353 ff.), in which a single present (363 προφωνεῖ)
stands among the many past tenses; the storm speech Ag. 650-70, too, is
entirely in past time except for 659 ὁρῶμεν. On the other hand, in Pers. 181 ff.
the queen begins her story of the dream with a series of past tenses; 191 f.
two presents come in and shortly after (195~9), when she reaches the most
exciting part of the dream, the description is wholly in the present (seven
present tenses).? Clytemnestra's beacon speech is in the main given in past
tenses; in the first part a single present occurs (293), but near the end (305,
307—here I have accepted a conjecture—310), where the narrative rises to
ever livelier triumph, the presents multiply. Cf. also below on 1597. Clearly
this use of the present differs entirely from that in archaic prose, where ‘the
story is told in the present and subsidiary facts in the imperfect’ as Wilamo-
witz, Geschichte der griech. Sprache (1928), 24f., has demonstrated with
specimen pieces from Pherekydes of Syros and Akusilaos.
1384. Although ἐν requires some explanation, it is passed over in silence by
the commentators; only Plüss remarks: ‘ev: the cries are regarded as a
! ‘Locutioni Homericae δεσμοὲ ἀπείρονες 0 340, Aeschyleis πέπλος ἀτέρμων Eum. 634—
sunt autem πεῖραρ et τέρμα synonyma—, ἀπέραντον δίκτυον Prom. 1078 accuratissime
respondent ἄπειρον ἀμφίβληστρον A. Ag. 1382, χιτὼν ἄπειρος Soph. fr, 483, ἀπείρωι ὑφάσματι
E. Or. 25’ (W. Schulze, Quaest. ep. 116).
2 Cf. Wilamowitz, Berliner Klassikertexle, v. 2, p. 59 n. x.
3 The reports of the scout in the Seven Against Thebes must of course not be cited as
parallel, since what is reported there is to be imagined as actually proceeding while the
account of it is heard.
650
 COMMENTARY line 1385

duration, cf. "Augenblick, Atemzug’ ’, thus comparing something entirely

different. I am inclined to see here a use of ἐν which has often been discussed
and has been particularly well illustrated by Wilamowitz on E. Her. 932 (ἐν
στροφαῖσιν ὀμμάτων ἐφθαρμένος). Wilamowitz says: ‘where a person or action
can be thought of as being in something which surrounds it like a sphere, ἐν
is in place. . . but the usage goes further’; for this more emancipated use,
which is therefore the more difficult to understand, he quotes among other
passages E. Bacch. 1165 ff. εἰσορῶ... dpuwyernv . . . Ἀγαύην... . ἐν διαστρόφοις
ὄσσοις (cf. Dodds ad loc.). ἐν οἰμώγμασιν (and ἐν στροφαῖσιν ὀμμάτων) is less
hard than ἐν διαστρόφοις ὄσσοις, but the function of the preposition is the
same.
οἰμώγμασιν. It is wiser to keep this here and in Eum. 600 δυοῖν... μιασ-
μάτων, although the possibility cannot be denied that Elmsley’s conjectures
are right, i.e. that Aeschylus in both places has put the noun as well as the
numeral in the dual. However, S. Phi. 117 9o . . . δωρήματα is a warning
against normalization. It is not quite safe to draw conclusions for the
language of the tragedians from Plato’s usage:! he has δύο ὀνόματα (Crat.
410 d, Soph. 244 c, 255 c) and δυοῖν ὀνόμασιν (Prot. 355 b, Soph. 244 c) ; Gorg.
466 ἃ δύο ταῦτ᾽ ἐστιν τὰ ἐρωτήματα is different; but on the other hand there
are instances such as τὼ δύο κύματε (Rep. 472 a), νοσήματε δύο (Rep. 391 c),
μιμήματε (Rep. 395 a). It is also doubtful whether, in this detail, tragic
language corresponded exactly with the language of documents (Meisterhans,
199 f.). The arguments produced by some scholars in defence of the plural
noun with δυοῖν here and in Eum. 600 are not conclusive. Wecklein’s assump-
tion (Curae epigraphicae, 1869, 16 f., adopted by Rutherford, New Phrynichus,
290) that this usage is legitimate where the noun is abstract is too artificial
and too modern to satisfy : would Sophocles regard δώρημα as an abstract?
A. Cuny, Le nombre duel en Grec, 96, justifies his retention of the MS reading
in Ag. 1384 and Eum. 600 by saying that dual forms of the neuters in -μα, i.e.
-ματε, -udrow, which are common in Aristophanes, never occur in Aeschylus,
Sophocles, or Euripides. He was misled by E. Hasse, on whose collection of
material he depends, for Hasse did not realize that in A. Pers. 720 the variant
στρατευμάτοιν is right ;? moreover Cuny has overlooked an instance correctly
noted by Hasse,? Eur. fr. 362. 9 N. δυοῖν mapóvrow mpaypárow.
1385. With Is. Vossius many editors read aóroó (Kirchhoff, Wilamowitz,
and Murray without stating that the MSS give αὐτοῦ, and although the
breathing signs do not form part of the ancient παράδοσις, in this case an
express statement should have been made). It is doubtful whether μεθῆκεν
αὑτοῦ κῶλα is correct Greek of the classical period ; normal phrases are (Ar.
Wasps 1212) rà γόνατ᾽ ἔκτεινε, (Birds 623) dvareívovres τὼ χεῖρε, (Frogs 2ox)
οὔκουν προβαλεῖ τὼ χεῖρε kákrevets ; (Peace 35 f.) τὴν κεφαλήν τε kai τῶ χεῖρε...
περιάγων, (Knights 571) ei dé που πέσοιεν ἐς τὸν ὦμον, (A. Prom. 32, E. Hec.
1150 and elsewhere) κάμπτειν γόνυ, (Pl Phaed. 60 Ὁ, Xen. Anab. s. 8. xo)
1 The following examples are taken from A. Cuny, Le nombre duel en grec (1906), 249 ff.
2 E. Hasse, Ueber den Dual bei den att. Dramatikern (Programm Bartenstein 1891) 8,
Der Dualis im Attischen (1893), 34, accepts the reading of the Mediceus orparnAdrow—an
impossible form—and places it under the examples of the 2nd declension. Wilamowitz, and
following him Murray, wrongly give M’s reading as στρατηλάταιν (for a full account of the
readings of the MSS here see A. Turyn, The Manuscript Tradition of... Aesch. 91).
3 Der Dualis im Altischen, 46.
651
line 1385 COMMENTARY

συνέκαμψςε τὸ σκέλος etc.! On Men. Epitr. 505 (569 Korte, 3rd ed.) dvékpaye
τὴν κεφαλήν 7 ἀνεπάταξε σφόδρα αὑτοῦ Wilamowitz remarks: ' αὐτοῦ is quite
unnecessary. Menander has extended the use of these pronouns which was
avoided earlier.’ Hermann was right therefore to advocate keeping αὐτοῦ
(locative) ; for μεθῆκεν he compares E. Hipp. 356 ῥίψω μεθήσω σῶμα, cf. also
(Butler ap. Peile) Ih. A. 648 μέθες νυν ὀφρὺν ὄμμα τ᾽ ἔκτεινον φίλον. αὐτοῦ,
‘on the spot’, is in no way an unnecessary addition ; it means that Agamemnon
sank down there after the two blows without being able to move from the
spot (cf. Y 691 αὐτοῦ yap ὑπήριπε φαίδιμα γυῖα, Pind. P. 6. 37 f.) ; there was no
question of resistance or flight.
1386. τρίτην : no noun should be supplied, cf. on 916 and (quoted by Wila-
mowitz on E. Her. 681) Ar. Knights 121 ἑτέραν ἔγχεον, Wasps 1231 ἑτέραν ἄισομαι.
ἐπενδίδωμι only here.
1387. Enger’s Διὸς is clearly right. Sidgwick’s note is a pretty example of
the method of a ‘conservative’ scholar who yet cannot get away from his
knowledge of Greek and his common sense: ‘Enger reads dids, which would
improve τοῦ κατὰ χθονός, rather needless with “Aiôov ; improve the irony ; and
is probable, as it would be easily supplanted by the gloss "Aıdov. But it is
safer to follow the MSS.’ There is no use in trying to save “Aidouv by a
punctuation which is only possible on paper.
On Zeus Soter cf. in addition to the books of reference (e.g. Höfer in
Roscher’s Lexikon iv. 1263 ff.) A. B. Cook, Zeus, ii. 1123 ff.; H. Sjövall, Zeus
im aligriech. Hauskult (Lund 1931), 85 ff.; Martin P. Nilsson in Symbolae
philol. Danielsson dicatae (Uppsala 1932), 227 ff. Ζεὺς Σωτήρ, to whom the
third libation is offered, plays a great part in Aeschylus,” cf. on 245 f., 284 f.,
Eum. 759 f. and the commentators there. The wording of the present passage
is very similar to that in fr. 55 N. (first the two other libations, then) τρίτον
Διὸς σωτῆρος εὐκταίαν λίβα (the scholion in which the fragment is quoted,
on Pind. Isthm. 6. 10 a, gives general information on Zeus as τρίτος σωτήρ).
For the Aeschylean idea of the Zeus of the dead and the underworld the locus
classicus is Suppl. 155 ff. τὸν ydiov (certain restoration) τὸν πολυξενώτατον
Ζῆνα τῶν κεκμηκότων (cf. on this R. Pfeiffer, Sitzgsb. Bayer. Ak., Phil.-hist.
Abt., 1938, Heft ii. 8f., 49).
εὐκταίαν χάριν : Heath explains: ‘gratiam votivam, i.e. ut vota persolverem,
quae ei olim nuncupaveram’, so also Blomfield, Humboldt (‘gelobt Geschenk’),
Wilamowitz (‘Dankeszoll’), and others. But the very similar passage (quoted
above) in the Epigons of Aeschylus (fr. 55), where εὐκταίαν λίβα simply means
the libation which accompanies the prayer, makes it more likely that here,
too, only the welcome gift that accompanies a prayer is meant, ‘a prayerful
offering’ (L. Campbell) or ‘a prayer-offering’ (Headlam).
Here and in Cho. 577 f. (the passage is closely allied also to Ag. 1397 f.) the
pure and healthful third libation changes into something baneful and hor-
rible. τρίτος σωτήρ is used Cho. 1073 in an uncanny connexion, which never-
. 1 This usage is not confined to the Attic dialect. From the fragments of the Lesbian
poets Lobel, 4Axaíov μέλη Ixxxi f., has collected a list of words for parts of the person
which show that *when no emphasis is laid on the person possessing,the sign of possession 15
not expressed’. With the examples quoted above may be compared, e.g., Sappho fr. 1. 11 D.
δίννεντες πτέρα, Alc. fr. 70. 21 D. ὀνάρταις xéppa.
2 Otfried Müller, Aesch. Eum. p. 188, points out that Plato, too, is fond of referring to the
τρίτος σωτήρ.
652
 COMMENTARY line 1388
theless leaves some hope of a happy ending.! Ag. 1386 f. is perhaps the most
powerful example of grandiose blasphemy in Aeschylus (cf. on 645, 1092,
and on 1144 ἀμφιθαλῆ). The harsh impiety must not be interpreted away ;?
the sharp paradox (no paradox for a Christian) of νεκρῶν σωτήρ speaks
clearly enough. Travesty of ritual language to enhance a gruesome effect is
found in other poets too. The Hecuba of Ennius (176 Ribbeck) prays in
blasphemous despair: Iuppiter tibi summe tandem male re gesta gratulor ; she
only departs in a single word from the gratulatio which the triumphator, γέ
bene gesta, addresses on the Capitol to most high Jupiter (cf. my Plautinisches
im Plautus 238). In power of effect everything is surpassed by the beginning
of the last canto of Dante's Inferno, where the opening words of the hymn
of Venantius Fortunatus (2. 6), which had been honoured for centuries in
the Church and accepted into the liturgy of the Passion, are changed by the
addition of a single word so as to describe the approach of the king of hell:
vexilla regis prodeunt inferni.
1388. Schütz, Butler, and others saw that the reflexive αὐτοῦ is required.
öpuyäveı is well restored by Hermann with the help of the gloss in
Hesychius ὀρυγάνει- ἐρεύγεται. Hermann’s doubts about the occurrence of
ὁρμαίνειν in Tragedy are unjustified; in the very difficult passage Sept. 394
(cf. also 393) Hermann resorted to a conjecture to get rid of this verb, but it
has now been found in a satyr play of Aeschylus, the AıkrvovAroi, Pap. Oxy.
2161, col. 2. 24, where ὅπως γάμον ὁρμαΐνωμεν is similar to the Homeric use
(K 28 πόλεμον θρασὺν ὅρμαίνοντες and the like) ; moreover ἐφορμαίνοντα occurs
Pers. 208. But Hermann says rightly that the meaning of ὁρμαίνειν [whether
we look at Homer or Pindar and Bacchylides] does not suit the context ; the
translation 'gasp out one's life' (L-S, cf. Paley) is unwarranted. Pauw and
Butler doubted óppatvew. Stanley, who did not doubt it, naively translated
it by evomit and compared Virg. Aen. 9. 349 purpuream vomit ille animam.
The expression θυμὸν ὀρυγάνει is perhaps an intensification of the Homeric
phrases θυμὸν ἀποπνείων (A 524, N 654), θυμὸν ἀΐσθων (II 468, cf. Y 403). On
the other hand, we might consider the possibility that in the passage Y403 f.
αὐτὰρ ὁ θυμὸν ἀϊΐσθε καὶ jpuyer, ὡς ὅτε ταῦρος ἤρυγεν κτλ. Aeschylus took ἤρυγεν
in the same way as it is taken by Schol. BT on Y 404 (cf. Eust. ad loc., p. 1214.
20 ff.), where we find the paraphrase dvepvyyávov τὸ αἷμα and ἐρεύγεται, and
that he regarded θυμόν as the object not only of ἀΐσθε but of ἤρυγεν as well.
However this may be, the coarseness of the expression fits in with Clytem-
nestra's ferocity.

t The melodramatic accounts of the death scenes of Stoic martyrs in Nero's time are both
in their general spirit and in the particular feature of the σπονδή to Zeus the Deliverer or
Liberator clear testimony to the philhellenism of the chief actors and their public. Tac.
Ann. 15. 64 (the death of Seneca) libare se liquorem illum Iovi liberatori, Ann. 16. 35 (death
of Thrasea, where with a cruder touch the dying man's own blood is the σπονδή) postquam
cruorem effudit, humum super spargens . . . Libamus, inquit, lovi liberatori (= Cassius Dio 62.
26. 4 ἐντεμὼν οὖν τὴν φλέβα dvérewe τὴν χεῖρα, καὶ ἔφη" ‘ σοὶ τοῦτο τὸ αἷμα, ὦ Ζεῦ ᾿Ελευθέριε,
σπένδω ’). See Wissowa, Religion und Kultus, 2nd ed., 120 n. 9: ‘The name Iuppiter Liberator
... was only invented on the analogy of the Greek Ζεὺς ᾿Ελευθέριος or Zwrijp'. There is no
blasphemy in this, even as an undertone, since to the Stoic death, especially when it saves
from shame, is a deliverance and a setting free.
2 Otfr. Müller, Aesch. Eum. p. 187, sadly misses the sense and tone of the passage;
Neustadt too, Hermes, lxiv, 1929, 259, tries unsuccessfully to ignore the blasphemous tone,
of which he is well aware.
653
line 1389 COMMENTARY

1389. If αἵματος σφαγήν is right, no one as yet, as far as I can see, has ex-
plained it satisfactorily. Stanley (‘efflansque celeri sanguinem fusum caede’)
is translating something that is not there. Later editors attempted to clear
their consciences by excuses ; so Pauw: ‘locutionem nota, quae audacior est:
σφαγὴν αἵματος pro αἷμα σφαττόμενον,͵ αἷμα σφαγῆι προκαλούμενον ᾿ (a step
farther away from the Greek). This was subsequently covered by a pseudo-
grammatical cloak: ‘ αἵματος σφαγήν dicitur poetice καθ’ ὑπαλλαγήν, pro αἷμα
ἀπὸ τῆς σφαγῆς: ut in Pers. 95 coi πηδήματος pro πηδήματι modos’? (Blom-
field), and thus won wide recognition, cf., e.g., Dindorf, Lex. Aesch., ‘caede
effusum sanguinem ebulliens’; L-S, ‘the blood gushing from the wound’ ;
Wecklein, ‘das durch Schlachten fliessende Blut’; Sidgwick, ‘a bold stretch
of language; we should say “his lifeblood" by an opposite metaphor’ (he
speaks as though σφαγῆς αἷμα were the MS reading); Mazon, ‘le sang qu’il
rejette avec violence sous le fer qui l’a percé”. That this violent interpreta-
tion is not justified by the use of σφαγή elsewhere is obvious from the un-
successful attempts of the dictionaries to fit in Ag. 1389 under σφαγή; neither
S. Trach. 573 (cf. 717) nor E. El, 1228 helps in the least, nor Antiphon 5. 69, to
say nothing of ‘parallels’ (e.g. van Heusde) like Eum. 450. Plüss takes the
case syntax more seriously than most: 'Genetiv des Inhalts; "eine ganze
Schlachtung Blut", soviel als beim Aufschneiden der Kehle eines Opfertiers
herausschiesst',* but that is grotesque. Nor would it help to have recourse to
the 'genitive of description’, examples of which are found in Jebb on S. Ant.
114 and in A. C. Pearson on E. Hel. 1156 ἅμιλλα αἵματος and E. Phoen. 801.
The difficulty is not only the addition of a genitive to σφαγήν : the expression
ἐκφυσιῶν σφαγήν is very strange, too. It is well known that σφαγή can
signify not only the act of σφάττειν but also its result, the wound. But this
use does not make ἐκφυσιᾶν σφαγήν easier. It seems, then, that Wilamowitz's
verdict 'nihili omnino ὀξεῖα αἵματος σφαγή ' is right (to refute his conjecture
σφυγήν one has only to realize the nature of σφύζειν). In principle Wilamo-
1 So again Wecklein: ‘like αἷμα σφάττειν '.
2 ‘It is needless to refute those who think that ποδὶ πηδήματος can stand for ποδὸς
πηδήματι ᾽ (Paley). For the genitive πηδήματος εὐπετοῦς cf. Wilamowitz, Isyllos von Epi-
dauros 167 n. 20, and on E. Her. 938.— The case would not be strengthened for the 'explana-
tion’ αἵματος σφαγή = αἷμα σφαγῆς by an appeal to other passages where difficult turns of
phrase have been ‘explained’ by a similar tour de force, e.g. Schol. Hom. IT 162 φόνον αἵματος"
ἀντὶ τοῦ * αἷμα pôvou '.
3 Otfried Müller, Aesch. Eum. p. 146, says: ‘To purify from blood guilt a sucking-pig . . .
was slaughtered in such a way that the gush of blood (the σφαγὴ αἵματος) splashed the
hands of the murderer’, etc., and quotes in support of his σφαγὴ αἵματος Eum. 449 f. This
sentence, ἔστ᾽ dv πρὸς ἀνδρὸς αἵματος καθαρσίον σφαγαὶ καθαιμάξωσι νεοθήλου Boro), was
misunderstood by Müller, as his translation shows; Wilamowitz also (‘bis seiner sich ein
Mensch erbarmt und sühnend ihm mit frischem Opferblute seine Hände wäscht’) and
Headlam (‘until at purifying hands he has been blooded by the slaughter of a sucking
swine’) render it incorrectly or at least inexactly. Since ἀνδρὸς καθαρσίου must go together
(there is no question here of πρὸς ἀνδρός without an attributive, as Müller takes it), the
linking of the intervening αἵματος with odayal gives an impossible word-order, on grounds
similar to those discussed on 1127 and 1511. αἵματος must depend on καθαρσίου, as Paley
takes it : ‘by the ministration of a man who is a purifier from murder’. Sidgwick adopts this,
and Blass similarly ad loc. ‘construe ἀνδρὸς καθαρσίον αἵματος".
4 Daube, 162 n. 45, moves on similar lines: ‘Wilamowitz’s conjecture σφυγήν for σφαγήν
does not take into account the fact that often it is the actual gushing out of the blood that
constitutes the real sacrificial act, since it is the giving up of life itself, He makes no
attempt to deal with the grammatical difficulty.
654
 COMMENTARY line 1301

witz agrees with Hartung, Schneidewin, and others who regarded σφαγήν as
corrupt. I think it is worth considering whether possibly pay» should be
read. Where the word occurs in the Hippocratic Corpus, Nat. puer. x2 (vol.
vii. 486 ff. Littré), it does not indicate the act of ῥηγνύναι but the fissure which
has been made. But the possibility of another meaning, which unfortunately
cannot be verified, is indicated by a scholion (possibly to be attributed to
Erotian, cf. fr. 31 Nachmanson) on the chapter quoted above from περὶ φύσιος
παιδίου; it says: τί ἔστι payy: ἀκμή, ὁρμή, Bla. This looks as if the com-
mentator knew passages where pay; denoted the act of ῥηγνύναι or ῥαγῆναι
and was thus the equivalent of ῥῆξις in the sense in which this is found in, e.g.,
Hippocr. Prognost. 21 (i, p. 102. 1 Kuehlewein) αἵματος ῥῆξιν διὰ ῥινῶν and
elsewhere. ὁρμή would be a reasonable rendering of this. The assumption
that Aeschylus knew ῥαγή from medical language is not unlikely, but is not
necessary since any Greek could form this noun from the verb. The possible
objection that such an expression is too technical for Tragedy is answered by
reference to S. Phil. 824 f. μέλαινά 7’ ἄκρου τις παρέρρωγεν ποδὸς αἱμορραγὴς
φλέψ. My suggested ῥαγήν can be understood as internal object of, or cognate
accusative after, ἐκφυσιῶν : ‘and breathing forth a mighty eruption of blood’.
Although the poet does not go into unnecessary anatomical details, his
picture is presumably based on the general idea that at least one of the three
πληγαί has damaged Agamemnon's lungs (a typical wound, cf. Cho. 639 f.):
in his desperate struggle for breath he spits blood.
The likeness of 5. Ant. 1238 f. kai φυσιῶν ὀξεῖαν ἐκβάλλει ῥοὴν λευκῆι παρειᾶι
φοινίου σταλάγματος to Ag. 1389 f. is obvious ; undoubtedly Sophocles imitated
the passage (rightly Dobree, Advers. ii. 26). Borrowing born of admiration
can more than once be seen in the earliest preserved plays of Sophocles (cf.
on 1293 and 1323).
1390. In the expressions ψακάς (cf. on 1534) and δρόσος the imagery of 1391 f.
is already foreshadowed ; with the former in this connexion cf. Ar. Peace
1140 ff. τυχεῖν μὲν ἤδη ᾽σπαρμένα, τὸν θεὸν δ᾽ ἐπιψακάζειν.
1391. οὐδὲν ἧσσον: cf. Schol. E. Or. 465 (τιμῶντέ μ᾽ οὐδὲν ἧσσον ἢ Διοσκόρω)
᾿Ἀἀττικὴ δὲ ἡ σύνταξις, ἀντὶ τοῦ οὐκ ἔλαττον τῶν Διοσκόρων.
διοσδότωι (restored by Porson as also the following γάνει ;' not in his edition
1 The vorw: of the MSS is a corruption of the common type in which two elements are
combined, a mechanical error arising from the literal similarity of two words, and a mental
error, the writer's thought straying to some word suggested by the context: the rainy wind
seemed to fit here. Cf. 1291, where AETW) was of course easily turned into AETW, but the
following προσεννέπω had probably something to do with the change. Similarly in A. Suppl.
308 INAXOY wasturned into OINEIAOY, partly because the Nile plays such a part in the
drama; in A. Pers. 632 several MSS have πέρσαις instead of πέρας ; 5. Aj. 360 πημονάν becomes
ποιμένων, there being in the first part of the play several references to ποῖμναι and ποιμνία.
Polybius 3. 88. 8 the genuine περὶ τὴν NAPNIAN isin the MSS replaced by AAYNIAN
(in the preceding paragraphs Δαυνίων occurs once, and τὴν Aavvlav twice). In the pseudo-
Virgilian Dirae 82 Scaliger’s raptorum should be read for the pratorum of the MSS, prata
being several times mentioned in the Dirae-Lydia, which in our MSS form one continuous
poem ; in the same way the silvis which appears in Lydia 44 was, as Baehrens saw, originally
vilis. In Ciris 470 Salaminia, preserved in B, has in all the other MSS been turned into alia
minos, Minos being a principal character in the poem. Propert. 2. 4. 18 immediately after
Cytaeis (= Medea) the MSS give per mede(a)e for Perimedeae, while in 2. 12. 18 st puer est
has displaced si pudor est; obviously puer Amor is the key-note of this poem. In Manilius
2. 3 Housman has restored pectoraque for Hectoreumque (. . . sub Hectore Troiam). The cod.
Etruscus of Sen. Thy. 68 gives manes (the speaker being the ghost of Tantalus) for amnes.
655
line 1391 COMMENTARY

but Tracts 210) : on this and on Pind. Isthm. 5. 49, Wilamowitz, Berl. Sitzgsb.
1909, 824 n. 4 remarks: ‘ Aids ὄμβρος is for the Greeks a most precious blessing.’
Stanley quoted the famous lines from the Danaids, fr. 44 N.
1392. γάνος, usually of some liquid which brings blessings (a more general
meaning in Ag. 579 and a few other instances), belongs exclusively to the
language of poetry, but in this particular usage (of water, wine, honey)
perhaps goes back to the language of ritual and cult (cf. Schadewaldt,
Monolog und Selbsigespräch 44 n. 1; other Euripidean passages could be
added). épeuvós and διόσδοτος, too, are poetical words.
1391 f. Homer played his part in shaping the thought here, as has long been
seen (Paley, Schneidewin, and others): % 597 ff. τοῖο δὲ θυμὸς ἰάνθη ὡς εἴ τε
περὶ σταχύεσσιν ἐέρσηι ληΐου ἀλδήσκοντος, ὅτε φρίσσουσιν ἄρουραι.2 But Aeschylus
has greatly strengthened the links between image and thing compared because
the idea of the nourishing moisture sent by heaven grows immediately and
naturally from the picture of the ψακὰς δρόσου. The image, in which the
thought is continued, is an organic part of the whole, while Homer appears
to have chosen one of many possible similes to illustrate joy. There are few
equally powerful lines in the whole of Greek tragedy. The horror is un-
escapable when the sweet miracle of the carefully tended sprouting and
growth of crops becomes a symbol of inhuman gloating over murder. Nothing
can bring out the fury of hate more strongly than the loving detail of κάλυκος
ἐν λοχεύμασιν, in which, as in the words of Aphrodite in the Danaids, the birth
of all created life is seen as a homogeneous process. Still pure blessing in the
image contrasts with wild polluting destruction wrought by Clytemnestra’s
deed. In other passages, too, Aeschylus communicates horror by recalling
something friendly and harmless (cf. on 979, on 1186 ff., and in general on
437 ff). Here the outrage done to the ‘heaven-sent’ nourishing moisture
has the same quality as that done to the τρίτη σπονδή just before.
1393. ὡς; ὧδ᾽ ἐχόντων as in S. Ant. 1179, 47.981, Hdt. 1. 126.6 (cf. Hdt. 8.144. 4
ὡς οὕτω ἐχόντων). In the two passages from Herodotus the phrase comes at
the end of a speech and marks the transition toan order given to the audience:
the likeness to Ag. 1393 is obvious. Further confirmation is found in Pers.
170 f., where the queen passes to the request in the last sentence of her speech
with the words πρὸς τάδ᾽, ὡς οὕτως ἐχόντων τῶνδε, σύμβουλοι λόγου τοῦδέ μοι
γένεσθε Πέρσαι. We can hardly doubt that this formula of transition was used in
the same way in public speeches of the time of Aeschylus. So this is another in-
stance illustrating the point made with regard to the formal transitions in Aga-
memnon’s speeches 830 and 950 and Clytemnestra’s concluding words in 1046.
1393 f. mpeoßos ᾿Αργείων τόδε: the same ceremonious address that she used
in 855. But now burning scorn lies under the words ‘my honoured lords of
Argos’. The scorn is still more outspoken in the next line: there the thought
is provocative in the extreme and the polite formula xaipoır’ av, ei χαίροιτε
(cf. on 1049) only enhances the expression of her disdain.
1394. ἐπεύχομαι (glorior, cf. 1262) corresponds to the expression em...
In Leo’s article on the Ciris, Hermes, xxxvii (1902), where he has often to mention Scylla, on
p- 41 |. 10 from the bottom 'Scylla' is printed for ‘Sibylle’. [Men. Dysc. 715 dexomov: aoxa-
mrov pap. (there is a good deal of σκάπτειν in the play).]
1 Relevant passages are collected in A. B. Cook, Zeus, iil. 320 n. 1.
2 The difficulties of this passage need not be discussed here.
3 It is hard to understand why any editor should want to eliminate the idiomatic ws.
656
 COMMENTARY lines 1395 f.

εὐχετάασθαι in the passage of Homer quoted below. This final word of her
sentence not only mocks the old men but also flouts an ancient rule of
Hellenic morality. The words of Odysseus x 411 f. ἐν θυμῶι, γρηῦ, χαῖρε καὶ
ἴσχεο μηδ᾽ ὀλόλυζε" οὐχ doin κταμένοισιν én’ ἀνδράσιν εὐχετάασθαι have been
quoted by Schneidewin (on 1399 = 1358 Schneid.) and others (cf. also
Sheppard, C.R. xxxvi, 1922, 5). Some ancient commentators assumed that a
passage in the Nereids of Aeschylus (fr. 151 N.) was modelled on the same line
of the Odyssey. This line was also regarded by Aristobulos or some other
writer περὶ κλοπῶν, whose lists are used by Clemens Alexandrinus, Strom. 6. 2.
4. 10 f. (ii. 425 Stählin), as the original plagiarized by Archilochus (fr. 65 D.)
and Cratinus (fr. 95 K.). Odysseus voices a moral conviction which had
equally deep roots in the minds of later generations.'
1395 f. The majority of commentators connect πρεπόντων (or whatever they
read instead) with ἦν. Many understand ἦν πρεπόντων on the analogy of
constructions like Demosth. 2. 2 ἔστι τῶν αἰσχρῶν, 20. 2 τῶν ἀδίκων ἐστίν and
the like (cf. Kühner-Gerth, i. 372). On the other hand, Blomfield, Wellauer,
and others (cf. Hermann) maintained that in that case the article would be
necessary ; and they may be right. The so-called poetic (or archaic) licence
of ‘omitting’ the article does not apparently extend to cases where the article
(or should one rather say ‘the deictic pronoun’ δ) has a clearly determinative
function. In any case, none of the passages adduced by Wecklein for the
omission of the article has even a remote resemblance to this: Ag. 39, Pers.
245 (248 Weckl.), E. Ih. T. 1301 (nor has, e.g., any of the passages collected in
Sidgwick’s note on Pers. 245). Verrall takes πρεπόντων as genitive absolute
(‘like ὧδ᾽ ἐχόντων in 1. 1393’), ‘under fit circumstances’ ; this needs no refuta-
tion. On the other hand, it would perhaps in itself be possible to accept the
conjecture of Is. Vossius, take ἦν πρεπόντως together (cf. p. 659 on δικαίως
ἦν), and let the infinitive depend on this. But all these interpretations are
open to two objections, one of language and one of content. (1) ὥστε is, if not
impossible, at least unlikely with an infinitive dependent on an expression
like πρέπον ἐστί. (2) How can a Greek conceive in terms of the unreal (condicio
irrealis) a thought like this: ‘und wäre es geziemend eine Spende dem Toten
nachzugiessen’ (Nägelsbach), ‘had it been a thing accounted seemly to pour
out oblation of wine upon the dead’ (Platt)? The drink offering for the dead
either at the cremation or as a later funeral offering is quite usual (evidence,
e.g.,in P. Stengel, D. griech. Kultusaltertümer, 3rd ed., 149) ; Cho. 149 itis named
ἐπισπένδειν (partly because it takes place ἐπ᾽ εὐχαῖς). The precariousnature of the
case may be seen from the manner in which some scholars resort to makeshifts :
‘wenn es tiberhaupt schicklich wire bei einem Toten Trankopfer (als Ausdruck
des Dankes und der Freude) darzubringen, so...’ (Wecklein), ‘und wenn an
einer Leiche je ein Freudentrunk geziemend war’ ...’ (Wilamowitz). The only
difference between these two versions is that Wecklein puts the thought which
he interpolates in brackets and Wilamowitz incorporates it in his translation.
! The words ody doin κταμένοισιν κτλ. form, however, a striking contrast with the typical
behaviour of Homeric heroes. On this point cf. the books and articles quoted by Von der
Mühll, RE Suppl. vii. 759. 55 ff., and Bury’s observation, Cambr. Anc. Hist. iv. 484.
2 I do not feel able to pronounce more definitely on this point. I am conscious here, as
often elsewhere, of the regrettable lack of any thorough investigation of the use of the
article in Attic poetry and prose; but to be useful, it would have to be based on a sound
and discriminating interpretation of all available texts.
4872.3 N 657
lines 1395 f. COMMENTARY

We get rid of these difficulties if we take ἦν by itself (= ἐξῆν) and connect

πρεπ. with ἐπισπένδειν ; Hermann considered this, and though he preferred a
conjecture, his tentative suggestion was adopted by Weil (edition of 1858),
van Heusde, Headlam. With this (1) we have the idiomatic construction ἦν
ὥστε ἐπισπένδειν, with which Paley, who did not, however, draw the full
conclusion for his interpretation, rightly compared! E. Hipp. 705 ἀλλ᾽ ἔστι
κἀκ τῶνδ᾽ ὥστε σωθῆναι, τέκνον (the order of words—dore being ‘postponed’—
also corresponds exactly with that which I assume for the present passage),
S. Phil. 656 dp’ ἔστιν ὥστε κἀγγύθεν θέαν λαβεῖν κτλ., Dem. 19. 114 ei μὲν μηδένα
. . εἴχομεν δεῖξαι... μηδ᾽ ἦν dor’ ἰδεῖν ἅπαντας κτλ. ; (2) (and this is the chief
point) without any forcing we get a sensible meaning. Before considering
this, I will say that I cannot follow van Heusde and Headlam when they
keep πρεπόντων and understand it as a partitive genitive dependent on ém-
σπένδειν (‘to pour libations of the fitting nature on the corpse’, Headlam).
The partitive genitive seems an unnatural way of expressing, not what the
liquid is (οἴνου or the like), but its suitability; I am also doubtful whether
unqualified πρέποντα (or τὰ πρέποντα) in the language of Aeschylus will bear the
interpretation these scholars put upon it. Contrast the tragic examples for
the plural of this participle, S. Trach. 384 τὰ δὲ λαθραῖ᾽ ds ἀσκεῖ μὴ πρέπονθ᾽
αὑτῶι κακά, Eur. fr. 313 N. δούλωι yap οὐχ οἷόν τε τἀληθῇ λέγειν, εἰ δεσπόταισι
μὴ πρέποντα τυγχάνοι, Iph. A. so4f. yevvai’ ἔλεξας Ταντάλωι τε τῶι Διὸς
πρέποντα. But parallels for πρεπόντως can be found in the fifth century:
Pind. Ol. 3. 8f. φόρμιγγα. . . ἐπέων τε θέσιν Αἰνησιδάμου παιδὲ συμμεῖξαι
πρεπόντως, À. Ag. 687 (cf. [Eur.] Rhes. 202), Ar. Birds 563 f. προσνείμασθαι δὲ
πρεπόντως τοῖσι θεοῖσιν τῶν ὀρνίθων ὃς ἂν ἁρμόττηι καθ᾽ ἕκαστον. I accept
therefore the emendation πρεπόντως, but connect it with what follows; in
both these respects I agree with one of the suggestions made by Pauw.
Hermann, too, thought εἰ δ᾽ ἦν, πρεπόντως ὥστ᾽ ἐπισπένδειν to be worthy of
consideration ; in Weil's first edition (1858) it is in the text. The sense, then,
is: ‘were it possible to give a drink offering to the corpse in a way suitable
(to the circumstances) . . .'.
The reading here recommended has the further advantage that with it
τάδε in 1396 is demonstrably sound. I explain it quite differently from Head-
lam, who says: ‘By τάδε, which is explained in τοσῶνδε... ἀραίων, Clytem-
nestra means that the proper libations for Agamemnon would be taken from
the apata κακά (that is βλαβερά) which he has himself inflicted on his own
house.’ But the sentence which begins with τοσῶνδε clearly starts afresh: it
adds a justification of her deed. Therefore τάδε, ‘this here’, must refer to
what Clytemnestra is doing and cannot point forward to τοσῶνδε κτλ.
She says '. . . this would be the right way of doing it’. It will be easier to
grasp the implication if we follow the thought as it evolves.
When Clytemnestra says ἐπεύχομαι, ‘I boast, I triumph’, the audience is
given to understand that this is a case of κταμένοισιν én’ ἀνδράσιν (or ἐπὶ
vexp@t) εὐχετάασθαι. As ἐπεύχομαι also means ‘pray’, the thought leads on by
an easily intelligible association to ἐπισπένδειν, a transition which is assisted
by Clytemnestra’s reinterpretation of the fatal blow as a libation for the
τρίτος σωτήρ of the underworld, a εὐκταία χάρις (1386 £.). The blasphemous
analogy persists in her thought. σπένδειν or ἐπισπένδειν, at the cremation or
1 Cf, also Kühner-Gerth, ii. 12.
658
 COMMENTARY lines 1405 f.

otherwise, is a normal element in the cult of the dead. But here, over the
body (or ‘after the murder’) of Agamemnon, the kind of σπένδειν must, if
possible, be suited to the particular occasion. ‘Then what I am doing here
would be justified'. τάδε, i.e. in the first place the triumphant cry of ἐπεύ-
χεσθαι, but also further the temper and feeling expressed in every word of
this speech. Joy and libation often go together, cf., e.g., Ar. Clouds 622 f.
ἡνίκ᾽ ἂν πενθῶμεν 7) τὸν Mépvov' ἢ Σαρπηδόνα, σπένδεθ᾽ ὑμεῖς καὶ γελᾶτε, Thesm.
793 obs χρῆν σπένδειν καὶ χαίρειν. In the passion of her vengeance she wishes to
reverse the normal habit, and the ritual homage which she would like to see
paid to the body of the murdered man is not πενθεῖν but a joyful libation.
1396. δικαίως ἦν. Good parallels for this construction have been given by
K. W. Krüger, Griech. Sprachlehre, ii, $ 62. 2 τι. 2, whom Wecklein follows:
cf., e.g., H 424 ἔνθα διαγνῶναι χαλεπῶς ἦν ἄνδρα ἕκαστον, probably also (though
discredited by many) A. Cho. 197 ἀλλ᾽ εὖ σάφ᾽ ἦν ἢ τόνδ᾽ ἀποπτύσαι πλόκον, in
addition (quoted in Matthiä, Griech. Gramm. ὃ 607. 3) E. Heraclid. 369 ποῦ
ταῦτα καλῶς ἂν ein παρὰ γ᾽ εὖ φρονοῦσιν; The present instance is in itself
simple, as τάδε stands for ‘this my manner of ἐπισπένδειν᾽, so that the expres-
sion means the same as τάδε δικαίως ἂν ἐπεσπένδετο. Clytemnestra intensifies
the thought of the protasis (ei δ᾽ ἦν «rA.) by pointing out that such a mocking
of the funeral rites would not only be suitable but perfectly just (cf. on 1378).
She is not content to say δικαίως, but improves on it (‘immo’) with ὑπερ.
μὲν οὖν.
1397 f. Clytemnestra’s mind is moving in the same sphere from which
σπένδειν was taken. Something like the motif of the bowl filled with crimes
was found in 1261 (cf. 1137). Passow compared Ar. Ach. 937 κρατὴρ κακῶν.
Agamemnon has in his house (and for his house) filled a mixing bowl of evil
and woe because he slaughtered his daughter and ruined his wife’s life ; now
he must drain the draught himself, must atone for what he did in the past.
He only suffers (in Clytemnestra’s view) his due (cf. 1527 ἄξια δράσας, ἄξια
πάσχων), therefore the deed and even the rejoicing over it are justified and
more than justified. There is nothing more to be sought in this sentence; we
should be on our guard against Paley's attempt with the help of an over-
subtle association of ideas to link 1397 f. more closely with what has preceded
and with customs of daily life.
1397. τοσῶνδε: it is difficult to say whether it is necessary to follow Blom-
field in writing τοσόνδε (Hermann's references to Prom. 112, which he mis-
interprets, and to Cho. 42 do not help).
“πλήσας ἀραίων per se constant. Male iungunt κακῶν πλήσας dpaiwy’ says
Hermann, whom others have followed. That is rather artificial; κακῶν and
ἀραίων go well together.
1398. μολών: cf. on 675.
1401. Expression and tone are like those of 277. She rejects the application
to her of accepted ideas as to what is womanly ; cf. 348 and elsewhere.
1402 f. πρὸς εἰδότας λέγω: cf. Suppl. 742 καὶ λέγω πρὸς εἰδότας, Prom. 441 f.
καὶ yàp εἰδυίαισιν dv ὑμῖν λέγοιμι. This is a fixed turn of expression found from
Homer onwards (X 250, Y 787 [from which Pindar, P. 4. 142 is derived]). Cf.
Wecklein on Prom. 441, Classen-Steup on Thuc. 4. 17. 3, also, e.g., Ar. Lys. 993.
1404. ὁμοῖον: for meaning and construction cf. 1239.
1405 f. 'Interpungebatur post χερός. Correxit Abreschius Hermann, who
659
lines 1405 f. COMMENTARY

was rightly followed by most (not by Mazon and Murray). It is wrong to put
a comma also after νεκρὸς δέ (so, e.g., Dindorf, van Heusde, Sidgwick, G. Thom-
son). Not only would the statement νεκρὸς δέ standing by itself be feeble after
all that has been said, but the thought clearly is that he as a corpse, his being
a corpse, is the work of her hand. This is a gruesome χειρουργία. Possibly
Euripides took this passage as a model for Her. 1139 μιᾶς ἅπαντα χειρὸς ἔργα
σῆς τάδε. Parallels to the expression δικαίας τέκτονος are quoted by Headlam-
Knox on Herodas 7. 38. Clytemnestra’s choice of the epithet δικαίας rather
than ‘skilful’ or ‘reliable’ shows particularly clearly the idea by which she is
dominated (cf. on 1378). On the occurrence of a stop after the first word in
four successive lines (1403-6; the last break is weaker than the others) cf.
Kranz, Hermes, liv, 1919, 313 and below on 1591-3 (p. 746 f.).

The section from 1407 to 1576 forms one long 'epirrhematic' composition,
the structure of which has been well elucidated by Kranz, Hermes, liv, 1919,
312 ff.
The introductory part or ‘prelude’ (Arnoldt, Kranz), 1407-47, consists of a
pair of stanzas sung by the Chorus and the replies of Clytemnestra in tri-
meters: after the strophe she recites fourteen lines, after the antistrophe
seventeen. The metre of the stanzas is simple:
dochm.
iambic dimeter
2 dochm.
2 dochm.
2 dochm.
pherecr.

When we pass from the prelude to the main part, 1448-1576, we find a more
elaborate form of epirrhematic structure. The stanzas of the Chorus are
answered no longer by trimeters but by anapaests. Of these anapaestic
periods those which follow the first stanza and its antistrophe are exactly
equal in length (11 metra each), while those following the second and third
stanzas correspond only approximately (for the modern attempts to foist
exact symmetry upon the text of the actor’s anapaests see the notes on
1521 ff. and 1554 ff.), so that in this respect they resemble the trimeters of the
prelude.
The whole piece 1448-1576 clearly falls into three parts:! I. 1448-80, II.
1481-1529, III. 1530-76. The arrangement of the central part (II) is simple:
a stanza of the Chorus, followed by an ephymnium, is answered by anapaests
of Clytemnestra ; then the whole is repeated with strict antistrophic respon-
sion save that Clytemnestra’s anapaests are of unequal length, as said above.
Parts I and III, on the other hand, have a different and indeed surprising
structure. It differs from that of the central part in that the ‘ephymnia’
which succeed the stanzas of the Chorus (1455 ἰὼ παράνους ᾿Ελένα krÀ., 1537
ἰὼ γᾶ γᾶ κτλ.) are not repeated after the ἀντίστροφοι, i.e. after 1474 and 1566.
The attempts to introduce them there against the MSS prove destructive to
the consistency of thought (see the commentary) and must therefore be
1 The analysis given by Kranz, loc. cit., was anticipated by R. Arnoldt, Der Chor im
Ag. des Aesch. 77 ff. (in the numeration of the lines he follows Hermann),
660
 STRUCTURE AND METRE OF 1407-1576
rejected. It would be absurd to abandon the most reliable of all criteria in
order to obtain conformity with the limited information which we possess
about possible strophic arrangements in the choric songs of Greek Tragedy.
It is true that the parallelism between the ‘ephymnia’ of I (1455 ff.) and III
(1537 ff.) and those of II (1489 ff. = 1513 ff.) is marked not only by the un-
common (see below) metrical form, i.e. an anapaestic period followed by
purely lyrical lines, but also by the similarity of the opening exclamations:
iw παράνους “Ελένα, ἰὼ ἰὼ βασιλεῦ βασιλεῦ, iw ya γᾶ. But while that clearly
points to the symmetry of the whole structure, it does not follow that the
‘ephymnium’ in I and III should be repeated as that in II certainly is. It is
conceivable that the poet wanted to give additional weight to the central
part, which is essentially a θρῆνος for the king, by repeating the lamentation
iw ἰὼ βασιλεῦ βασιλεῦ κτλ. in contrast with the treatment of the analogous
sections in the part which precedes and in that which follows.
It seems impossible to make out with any degree of certainty whether the
choral parts are all sung by the undivided Chorus. Hermann asserted ‘aperte
nihil a toto choro cani’; consequently he left only the anapaests (1455 ff.,
1489 ff., etc.) to the undivided Chorus and distributed all the rest over five
ζυγά. Blomfield followed Butler in introducing npıxöpıa,' but diverged from
him in assigning the anapaests to the whole body of the Chorus. All this is
of course quite arbitrary. Wilamowitz marks a change of speaker (or singer)
before each ‘ephymnium’ (for his fallacious argument cf. p. 689, n. 1); he
does not, however, make it clear whether he has ἡμιχόρια in mind. A cautious
suggestion is put forward by Kranz, Hermes, liv, 1919, 320, who thinks it
possible that the ‘ephymnia’, i.e. the choral sections composed of anapaests
and subsequent lyrical metres, are to be assigned to the coryphaeus. The
possibility of such an arrangement must be admitted.

Metre of the main part:

I. 1448-54 = 1468-74. 2 dochm.-+pherecr., 3 iamb. (catalect.), decasyll.
. alcaicus, 5 iamb., 3 iamb.?
1455-61. interjection, 5 anap. (catalect.), 2 dochm., 3 iamb., 3 iamb.
(catalect.).
II. 1481-8 = 1505-12. hemiepes, decasyll. alcaicus, then either reizianum
(φεῦ φεῦ, κακὸν alvov) and a variation of the hipponacteum (so Wilamowitz,
Verskunst, 249 n. x), or enhoplius (φεῦ φεῦ, κακὸν alvov arn-) and pherecr. (so
Wilamowitz in his edition and Verskunst, 390; O. Schroeder), 7 (or 4 and 3)
iamb. (catalect.), hipponacteum.
1489-96 (repeated 1513-20). 5 anap. (catalect.), 4 anap. (catal.), molossus,
glyc., then, at 1495, the text is uncertain: as it stands in the MSS, it seems a
variation of the telesilleion (cf. on 1495 f.); accepting Enger's attractive
supplement, we should have the same-|-bacch. (for parallels cf. on 1495 f.),
decasyll. alcaicus.
III. 1530-6 = 1560-6. 3 iamb. (catal.), choriambic-iambic dimeter (catal.),
3 iamb. (catal), 6 iamb. (catal), 6 iamb. (catal.).
! The same had been done, at any rate partially, by Byzantine scholars, see app. crit.
on 1489, 1513, 1537, 1547.
* Cf. note on 1453.
661
line 1407 COMMENTARY

1537-50. 6 anap. (catal), ro anap. (catal), praxilleion! (cf. Hephaestion

7.8, p. 24. 8 Consbruch), choriamb.-iambic dim. (catal.), 3 iamb. (catal.).
The metrical pattern of the three ‘ephymnia’, i.e. a period of anapaests
followed by various lyrical metres, has no parallel in the extant plays of
either Aeschylus? or Sophocles. There are, however, analogies in Euripides.
For the parados of the Medea cf: Wilamowitz, Verskunst, 538. There we have
first a stanza of the Chorus (without strophic corresponsion), 131 ff.: 3 anap.,
lyrical dactyls (cf. Rhein. Mus. Ixxii, 1918, 177), 3 iamb. (catal.). The stanza
148 ff. — 173 ff. begins with 5 anapaests ; these are succeeded by enhoplia and
other lyrical metres. In the Phaethon the second stanza (35 ff. = 43 ff.) of the
parodos (Berliner Klassikertexte, v. 2, Ὁ. 81 f. ; O. Schroeder, Euripidis cantica,
172) begins with 8 anap. (catal), then, after another paroemiac, there follow
different lyrical metres. Finally, a slightly more remote analogy may be
quoted from a Euripidean monody : Hipp. 1370-88, where a long anapaestic
period precedes the part composed of iambics and dochmiacs.?

1407. χθονοτρεφές occurs only here, perhaps inspired by A 741 ἣ τόσα φάρ-
paka ἤιδη ὅσα τρέφει εὐρεῖα χθών.
ἐδανός in extant literature only here (for the group of adjectives so formed
see Schwyzer, Griech. Gramm. i. 490, and Buck-Petersen, A Reverse Index,
621). Hesychius ἐδανά' ἐδώδιμα, βρώσιμα. Edavois" βρωσίμοις, Αἰσχύλος (Blom-
field’s suggestion of restoring the singular is rash) suggests that the word
occurred elsewhere in Aeschylus.
1408. πασαμένα: the scholia misrepresent the meaning of the epic word,
which they render by κτησαμένη ; accordingly the first syllable is marked as
long (räcauéva) in F and Tr. But Triclinius either understood or guessed the
true sense, for he added the gloss πιοῦσα.
putas: it seems strange that a critic of Canter’s acumen was satisfied with
pvoäs.*
‘ ava and ποτά, χθονοτρεφῇ and ἐξ ἁλὸς ὄρμενα correspond to each other’
(Schneidewin).
Wilamowitz rightly (in spite of P. Maas, Sokrates, iii, 1915, 313) keeps (with
Canter) ὀρόμενον, to which the MS reading points. He should, however, have
been consistent and left ópouévav in Suppl. 422 (in the cretics there ὀρομέναν
corresponds with 427 — U — exactly as 423 πολυθέων with 418 πανδέκως). Else-
where in Aeschylus we find ópópevov in Sept. 86, 115, but συνορμένοις in Ag.
429. The responsion here ἐξ ἁλὸς ὀρόμενον = 1427 φρὴν ἐπιμαίνεται is com-
pletely regular, cf. 1x21 ἔδραμε κροκοβαφής = 1132 τίς ἀγαθὰ φάτις,
The Chorus regard Clytemnestra’s attitude as madness and therefore con-
1 As regards l. 1547, I disapprove of Wilamowitz's later suggestion, Verskunst, 292; cf.
O. Schroeder, Grundriss der griech. Versgeschichte, 61.
2 Pers. 986 = 10017 must not be adduced, for there the period of (partly ‘threnodic’)
anapaests is followed by nothing more than a brief clausula (the colon μηδὲν φοβηθῆις+
dochm.).
3 At E. El. 112 f. the anapaests cannot be regarded as certain, cf, H. L. Ahrens, KI. Schr.
55; Wilamowitz, Verskunst, 558.
+ So were some Byzantine scholars, for in Fand Tr γηραιᾶς is written above it. Wecklein
omits this p. 339, and also the gloss βρωτὸν over edavöv. I have not been able to discover on
what principle he (or Vitelli, cf. Wecklein, praef., p. x) gives or omits the glosses in F; cf. on
1264, p. 584, n. 2.
662
 COMMENTARY line 1409

clude that it must have been caused by taking φάρμακα (cf. the Homeric
passage quoted on 1407). So also Ar. Thesm. 533 f. οὔ τοι... ed φρονεῖτε, ἀλλ'
ἢ πεφάρμαχθ᾽ ἣ κακόν τι μέγα πεπόνθατ᾽ ἄλλο, E. Bacch. 326 f. (the end of the
sentence is relevant) μαίνηι γὰρ ws dAyıora, κοὔτε φαρμάκοις ἄκη λάβοις ἂν οὔτ᾽
ἄνευ τούτων νοσεῖς. For the subdivision οὗ φάρμακα (χθονοτρεφές and ῥυτᾶς ἐξ
ἁλὸς dpouevor) van Heusde compares Apoll. Rhod. 3. 529 f. τὴν ‘Ekdrn περίαλλα
θεὰ Sde τεχνήσασθαι φάρμαχ᾽ ὅσ᾽ ἤπειρός τε φύει καὶ νήχυτον ὕδωρ. The first
group naturally suggests herbs, roots, etc. ; pur. ἐ. à. dp. is less simple. Klausen:
‘mari dicto pro aqua omni’; Paley: ‘the sea being mentioned not as a source
of poison, but as descriptive of the sort, liquid opposed to solid.” That does
not bring out the obvious meaning of the words. One of two things seems to
be meant here. We have either to think of a drink of sea-water (or at least
partly composed of sea-water), but I have not been able to discover any
ancient evidence for sea-water being supposed to produce madness;! or the
passage may allude to some drug composed of the juice of sea plants, though
this seems less probable.
The Chorus puts forward a suggestion in interrogative form including
various possibilities (7): the same is found in, e.g., S. 47. 172 ff. (the whole
stanza) and E. Hipp. 141 ff.
1409. τόδ᾽ ἐπέθου θύος. I have no real solution for these words. If they are
right as they stand, no one yet has provided a satisfactory explanation. The
scholiast (ZyoÀ. aA.) paraphrases ἔμαθες τοιαῦτα ποιεῖν θύματα. In this para-
phrase ἔμαθες seems to be based on the reading ἐπεύθου, which, though found
in Tr only, may well be older. But the chief question is the meaning of θύος.
Hesychius agrees with the scholiasts' explanation θύματα in his gloss θύος"
iepov, θῦμα (quoted in Stanley’s posthumous notes). We know that θύεα in
Homer in several passages does not mean θυμιάματα but any burnt offering?
(cf. Lehrs, De Aristarchi stud. Hom. 3rd ed., 83 f.). Therefore Aeschylus could
perfectly well have used @vos == ‘sacrifice’, θῦμα (twice in Ag.). Some com-
mentators who start from the meaning ‘incense’ have reached fantastic
interpretations : imposuisti tibi (ut victimae mox ob crimina mactandae) hoc tus
publicarum execrationum? (Scholefield) ; ‘ "placed on yourself this incense’’,
sc. the incense of the people’s wrath on her devoted head’ (Paley) ; similarly
Conington, Kennedy, and others. This we need not discuss, nor the com-
pletely misleading comparison of ἐπέθου θύος with the common λιβανωτὸν
ἐπιθεῖναι. A number of commentators take θύος quite differently, namely as
‘wrath, fury, etc.’. Thus, e.g., Schütz, Humboldt, Schneidewin, Nagelsbach,
Wecklein, Verrall, and apparently, according to his translation (‘Hast du mit
Hexenkunst Mut dir zur That gemacht? Assest ein Zauberkraut’ etc.),
Wilamowitz. Headlam prefers the meaning = θῦμα but regards ‘frenzy’ as
worthy of discussion. The linguistic foundation of this interpretation does
not seem to have been examined. It is difficult enough that we have not a
shadow of evidence for @vos in this sense; nor is the shortening of the
ı For the effects of certain inland springs in producing madness cf. Ctesias in Antigonus,
Hist. mir. 145(160)—and in Plin. Nat. hist. 31. 9—rijv ἐν Αἰθιοπίαι (scil. κρήνην) τὸ μὲν ὕδωρ
ἔχειν ἐρυθρόν... τοὺς δ᾽ ἀπ᾽ αὐτῆς πιόντας παράφρονας γίγνεσθαι, Pausanias, 9. 8. 2 δείκνυται δὲ
ἐν Ποτνιαῖς καὶ φρέαρ' τὰς δὲ ἵππους τὰς ἐπιχωρίους τοῦ ὕδατος πιούσας τούτου μανῆναι λέγουσιν,
cf. also Diodorus 2. 14. 4, Ovid, Met. 15. 320 f.
2 ‘The notion of @vos has expanded along with the notion of θύειν (properly incense-
offering)’ Blass on A. Eum. 835.
663
line 1409 COMMENTARY

root-syllable an easy matter.’ θύειν, Bview, ‘rage’, has like θύνω and θυ(ι)άς
alwaysalong v (cf. W. Schulze, Quaest. ep. 313 ff.) ; this is not countered by the
fact that an ancient etymology, the worth of which is completely unknown,
derives Θυώνη (v — —) from this verb (Schol. Pind. P. 3. 177 a οὕτω δὲ ὀνομά-
ζεται, viz. Semele, ἀπὸ τοῦ περὶ τὸν Διόνυσον πάθους, ὅτι θύει καὶ ἐνθουσιᾶι κατὰ
τοὺς χορούς). So it seems wiser to discard the meaning furor, although fresh
evidence may some day prove it possible.
But there is an additional difficulty. What the commentators say on
ἐπέθου, in so far as they go into it at all, cannot be regarded as explaining it ;
the same is true of the translations and the entries in the lexica. The
commonest rendering is ‘du hast über dich gebracht’ (Wecklein), ‘du hast dir
zugelegt’ (Schneidewin)—taking @vos as frenzy—or ‘to put upon thy head
this sacrifice and clamour of the people’s curse’ (Headlam) and the like.
While it is quite natural to say κρατὶ δ᾽ em’ ἀμφίφαλον κυνέην θέτο etc., one
hesitates to believe in the possibility of θῦμα ἐπιθέσθαι or ἀρὰς ἐπιθέσθαι. Ag.
1409 ἐπέθου in Dindorf's Thesaurus, iii. 1845 c, is quoted between Apoll.
Rhod. 4. 1536 χυτὴν ἐπὶ γαῖαν ἔθεντο and E. Bacch. 702 f. ἐπὶ δ᾽ ἔθεντο κισσίνους
στεφάνους ; from this and L-S s.v. B. IV. x (with the translation ‘bring on
yourself’) it is clear that the present passage is unique. Were the only problem
to find a verb for @vos ‘sacrifice’, ἔθου (Karsten) would not be bad: ‘you have
made (arranged) this sacrifice’. Cf. Ag. 1570 ὅρκους θεμένη, S. Oed. R. 1447 f.
τῆς μὲν κατ᾽ οἴκους αὐτὸς ὃν θέλεις τάφον θοῦ, Oed. C. 466 θοῦ viv καθαρμὸν
τῶνδε δαιμόνων, 542 f. ἔθου φόνον... πατρός; E. Suppl. 950 f. κατ᾽ ἀλλήλων
φόνους τίθεσθε. The form ἔθου is found Sept. 108. As Clytemnestra herself
spoke of her act in images from the sphere of ritual, it would not be un-
thinkable that this imagery should be picked up by τόδε θύος ἔθου. But
Karsten's text and explanation ' τόδε τ᾽ ἔθου θύος δημοθρόους τ᾽ ἀράς, h.e. et
caedem hanc perpetrast: et populi diras excitasti ; auctor exstitisti et caedis et
dirarum quae inde sequentur' is too violent to commend itself; perhaps such
an interpretation would be more tolerable without the first τε, when θύος and
dpat would describe the same thing and the dpa be contained in the θύος. Nor
is the attempt to save Karsten's text by not punctuating until after ἀπέταμες
entirely satisfactory, as the linking of the ἔθον clause with the ἀπέδικες clause
by re... re does not seem suitable to the excited language of this stanza.
Perhaps the most unfortunate result of this obscurity is that a definitive
judgement on the punctuation and therefore the interpretation of the follow-
ing clause (to ἀπέταμες) is made more difficult if not actually impossible.
For, as long as we do not know what lies behind τόδ᾽ ἐπέθου θύος, we cannot
give a reliable answer to the question whether δημοθρόους τ᾽ ἀράς is the
object of the previous verb (ἐπέθου) or the following (ἀπέδικες). My feeling?
suggests that the punctuation should be after ἀράς. Three arguments can be
adduced, though unfortunately none of them is wholly conclusive. (1) It
seems to be more forceful if the curses of the people are aroused immediately
by Clytemnestra's act (whatever be the meaning of ἐπέθου) than if the curses,
! It is true that metrically two long syllables after ἐπέθου would not be altogether im-
possible (cf. on 1128), but it seems hazardous to add to the few examples of this free respon-
sion by an instance based on a doubtful reading.
21 regret that in discussing this passage I have had to be content with such vague
remarks. It is not due, I think, to any lack of persistence on my part. I hope that others
will be more successful,
664
 COMMENTARY line 1412
which can only take voice when the deed has become known, are assumed to
have been already uttered and are now described as having been cast on one
side by Clytemnestra. (2) ἀπέδικες ἀπέταμες, without a definite object, seems
very attractive (unless we are misled by a modern feeling) and would form a
fine background to the following words. Conington, Paley, Kennedy, Sidg-
wick, Verrall, Headlam, and Platt clearly felt in the same way: witness their
punctuation and translation. (3) This structure would produce exact corre-
spondence with the antistrophe where 1428 πρέπει ends a sentence and at 1429
a new sentence starts in asyndeton. The other structure was championed by
Bothe and then strongly by Hermann: ‘et populi diras contempsisti praefracte’
(he read ἀποτόμως). When Hermann then adds 'respicit ad v. 1403 (1363
Herm.), he is certainly wrong; Clytemnestra's σὺ δ᾽ αἰνεῖν εἴτε με ψέγειν
θέλεις ὁμοῖον cannot mean a casting aside of the people's ill wishes. Wilamo-
witz follows Hermann's punctuation. The construction of ἀράς with ἀπέδικες
is a point in favour of this explanation ; the Byzantine who glossed ἀπέδικες
with ἀπέρριψας knew his tragic language, cf., e.g., 5. El. 1017 f. καλῶς δ᾽ ἤιδη σ᾽
ἀπορρίψουσαν ἁπηγγελλόμην and also Eum. 215, Hdt. 1. 32. 1 ἡ δ᾽ ἡμετέρη
εὐδαιμονίη οὕτω τοι ἀπέρριπται ἐς τὸ μηδὲν ὥστε krÀ. Result: non liquet.
1410. With the asyndeton ἀπέδικες ἀπέταμες (held together by number of
syllables as well as by initial and final assonance)! van Heusde well compares
1553 κάππεσε κάτθανε. Metre forbids the addition of r’ (Kennedy), which
would be bad on stylistic grounds also; see on 1429. For possible meanings of
the words see on 1409. To understand Agamemnon as object of the verbs (so
Schütz and, e.g., Nägelsbach and Kennedy) seems absurd.
ἀπόπολις is a certain restoration. The form ἀπόπολις (not ámómr.) is
demanded by metre S. Oed. C. 207 (cf. Wilamowitz, Verskunst, 342). The
threat? of banishment (1412 ἐκ πόλεως φυγήν) corresponds to the stoning in
the clash of the Chorus with Aegisthus (1615 f.), see ad loc.
1411. On ὄβριμον (for the misspelling ὄμβριμον cf. F. Jacoby in the preface to
his edition of Hesiod's Theog. p. 104) Wilamowitz says, Interpr. 86 n. x (on
Sept. 794): ‘an epic word which tragedy almost entirely eschews [it is used
E. Or. 1454 in the Phrygian's monody] . . . ὄβριμον μῖσος Agam. 1411 follows
the analogy of ὄβριμον ἄχθος ı 233, where the scholia tell us nothing. I am not
sure with what nuance Aeschylus used it.' À survey of the attempts to make
out the etymology is given by C. Arbenz, Die Adjektive auf -wos (diss. Zürich
1933), 24 f. Schwyzer, Griech. Gramm. i. 494. 8, adopts the old combination
with βρί-θω βριαρός.
1412 f. For the exact repetition of the words of the Chorus in Clytemnestra’s
answer here, as soon after in the kommos, cf. p. 694 (top), also on 1137.
1412. ‘ νῦν μέν, as if τότε δὲ οὐδὲν. . ἔφερες (1414) followed’, Wecklein (and
before him Nägelsbach). This is clearly the function of the particle here; its
t A small selection of examples of the combination of these stylistic devices in tragic
lyric can be found in my Plautinisches im Plautus, 365.
2 Blomfield, Wecklein, etc., credit Casaubon only with the conjecture ἀπόπτολις, But he
wrote in the margin of his copy of Victorius’ edition (now in Cambridge) : ‘lego ἀπόπολις vel
ἀπόπτολις ex contrario versu ἀντιστροφ.ἢ
3 There is nothing more here than a threat such as anyone without special legal com-
petence could pronounce. That 'the Gerusia thinks of banishing her, but finally gives way
before her superior power' (Daube, 179) is an unfounded assumption ; Daube wants to find
some sort of legal activity for that body, but there is no 'Gerusia' in this play (cf. on 884).
665
line 1412 COMMENTARY
use in this answer is not, properly speaking, an instance of a speech beginning
with μέν (cf. on 1).
δικάζεις krÀ. : "Now you are acting as a judge and sentencing me to banish-
ment’, etc. 1421 δικαστὴς τραχὺς el is on the same lines. That the old men
should sit as a court and pass judgement on her is for Clytemnestra a piece
of provocation, more intolerable than anything else, because she has insisted
again and again on her incontrovertible claim to justice and on the justifica-
tion of her deed.
‘ φυγήν is directly dependent on δικάζεις, μῖσος and ἀράς on ἔχειν ' (Weck-
lein). This is possible, but the following rendering is perhaps better: 'banish-
ment for me and public execration and suffering of the general curse'
(Headlam), so that ἔχειν is added to strengthen the last phrase. Then the three
objects are constructed in accordance with 'the law of increasing parts' (cf.
on 1243 f.). For further examples of δικάζειν with internal object (acc. cogn.)
see L-S s.v. I. 2.
1414. τότ᾽ is clearly necessary. Blomfield gives examples from Tragedy for
this τότε after an opening νῦν. Hermann's defence of τόδε is lame; it reads
as if he really preferred τότε. Naturally he does not resort to the over-subtle
explanations by which others have defended τόδ᾽. In φέρων after τότε we may
see an 'imperfecti participium' (Blaydes), cf. on 17o.
ἐναντίον φέρων looks like an exact equivalent of ἀντιφερίζειν (primarily
epic, then in Pindar and in anapaests in Aristophanes) with the dative (cf.
also avrıdepeoda:).
1415. οὐ προτιμῶν : cf. (Blomfield) Hesychius οὐ προτιμᾶιϊ' οὐκ ἔχει λόγον. Cf.
1672, E. Hipp. 48 f. τὸ γὰρ τῆσδ᾽ οὐ προτιμήσω κακὸν τὸ μὴ οὐ παρασχεῖν κτλ.
1416. Translations such as Verrall’s ‘against him who, caring no more than
for the death of a beast, though his fleecy herds had sheep enough, . . .' do
not bring out the meaning of the passage ; Schneidewin, Kennedy, L. Camp-
bell, Wilamowitz etc. translate correctly. Clytemnestra cannot, unless she
wants to make herself ridiculous, say that ‘Agamemnon ought to have chosen
an offering from his rich herds’—for Artemis had demanded his own daughter
(Schneidewin). It does credit to the scholiast that he saw that 1416 belongs
to the comparison : κατ᾽ οὐδὲν μεμφάμενος τῶι Ayapéuvor ὅτι τὴν ἐμὴν θυγατέρα
᾿Ιφιγένειαν ἔκτεινεν ὡς πλούσιος ποιμὴν πολλὰ ἔχων θρέμματα θύσειεν' ἐξ αὐτῶν ἕν.
This being the sense, editors should not puzzle the reader by meaningless
punctuation as, e.g., a comma after μόρον (the correct punctuation is given
by C. G. Haupt, Hermann,’ Schneidewin, etc.). In F there is no stop from
1415 ὃς où to 1417 παῖδα.
1418. ἐπωιδόν in this context and on Clytemnestra’s lips has a slightly
depreciatory flavour ; perhaps the same applies to Θρηικίων (barbarian winds,
which had to be soothed by the sacrifice of the Greek king’s daughter). On
these winds cf. on 192.
1 Conington at the end of his edition of the Choephoroe, p. 161, gives examples from its
scholia (M) of the use of the potential optative without ἄν, For the same habit in the scholia
of Triclinius cf. H. W. Smyth, Harvard Studies in Class. Philol. xxxii, 1921, 84. This diver-
gence from classical syntax is, however, much older than our scholia. M. J. Higgins,
Byzantion, xv, 1941, 443 ff., shows that in the wiitings of Gregory Nazianzen the potential
optative, under certain conditions, is frequently used without ἄν,
2 I do not know whether Hermann himself or Moriz Haupt is responsible for the punctua-
tion of the text.
666
 COMMENTARY lines 1421-4
Stanley compared the similar accusation in Clytemnestra’s speech S. El.
530 ff. with 1415-18; Kaibel (ad loc.) thinks that Sophocles was consciously
borrowing.
The first half of this speech, 1412-18, is couched in a single rather elaborate
period. Clytemnestra pleads her cause like an advocate in court.
1419. avöpnAareiv first here, 1586, and Eum. 221. That verbs in -nAareiv are
first found in Aeschylus is observed by Wackernagel, Dehnungsgesetz, 43.
1420. ἐπήκοος : Conington says on Cho. 980: ‘ ἐπήκοος here and in Ag. 1420
seems to have a semi-judicial sense, ‘‘taking cognizance of” ’. Wilamowitz in
his commentary on Cho. 980 translates the word by ‘witness’, referring to in-
scriptions for the Laconian use of ἐπάκοος (cf. L-S s.v. IV) and to Hesychius
émákoov μάρτυρες. But in the present passage at any rate we should not go
beyond the ‘semi-judicial sense’, since it is improbable that Clytemnestra
would call one and the same person both witness and judge.
1421. The final sentence of her answer, which contains a new idea (her power,
whereas up to now she has spoken of her rights), is introduced by λέγω δέ σοι.
This may in this context have a hard ring about it— ‘and I tell you this’.
After the long excited dispute with Oedipus (S. Oed. R. 447) Teiresias says:
‘I have said what I had to say and will now go’, then he adds (449) λέγω δέ σου;
there follows the gruesome revelation of the murderer of Laius, then the seer
departs. E. Hec. 1232 f. (at the conclusion of a long speech addressed to
Agamemnon) σοὶ δ᾽ ἐγὼ λέγω, Aydpeuvov, ei τῶιδ᾽ ἀρκέσεις, κακὸς paris.
Dikaiopolis (Ar. Ach. 169 ff.) : ἀλλ’ ἀπαγορεύω μὴ ποεῖν ἐκκλησίαν τοῖς Θραιξὶ
περὶ μισθοῦ; then a last sentence: λέγω δ᾽ ὑμῖν ὅτι διοσημία ᾽στι καὶ pavis
βέβληκέ pe; this results in the postponement of the assembly. Eum. 98
προὐννέπω δ᾽ ὑμῖν rings equally hard. E. Med. 351 προὐννέπω δέ σοι introduces
a powerful threat. [Cf. Thuc. 2. 71. 4 (conclusion of the speech),5. 5]
1421-4. λέγω δέ cor... ἄρχειν. We must be careful not to tamper with
the blameless parts of this sentence in order to mend the difficult and
unintelligible pieces. ὡς παρεσκευασμένης shculd not be altered, although
attempts have been made from Casaubon to A. Y. Campbell." This construc-
tion is idiomatic, especially when connected with an order or a demand as
here, cf. (in addition to 1393 and the instances from Herodotus quoted there),
e.g., E. Med. 1311, S. Oed. R. 145, 241, El. 316, Thuc. 7. 15. 1; for Xenophon
see Kühner-Gerth, ii. 93. The real difficulty lies in fixing the end of the
clause beginning with ὡς παρεσκ. If an infinitive follows παρεσκεύασμαι, it is
natural to make it dependent on this verb. Accordingly Hermann translates:
'jubeo te talia minari, ut me parata imperare mibi, qui vicissim me vi vicerit'.
So also, e.g., Sidgwick, Verrall, Headlam, who agree with each other and with
Hermann on the main issue and only diverge on the meaning and syntax of ἐκ
τῶν ὁμοίων. But ‘parata sum illum mihi imperare', 'I am prepared . . . that
you should be my master' (Headlam), is not only an unparalleled expression
(as the examples of παρεσκεύασμαι with the infinitive in Aristophanes and
elsewhere show), but does not express the thought: Schoemann, Ofusc. iii.
152, rightly says ‘perverse dicta; quis enim umquam dixit: Parata sum mihi
imperare aliquem pro eo quod est farata sum alicuius imperium ferre aut
alicui parere?! Sidgwick's excuse ‘the construction of the acc. c. inf. after
1 Enger, followed by Wecklein, obelizes 1422, but the reference in τοιαῦτ᾽ ἀπειλεῖν to the
preceding stanza of the Chorus is the most important element of the whole sentence.
667
lines 1421-4 COMMENTARY

παρεσκευασμένης is not unnatural considering the freedom of poetry and the

great flexibility of Greek’ betrays the embarrassment of a good Greek scholar.
Therefore it seems clear that νικήσαντα... ἄρχειν cannot be dependent on
mapeok.; à stop must necessarily be put after ὁμοίων (so Pauw, and later,
e.g., Wellauer, Scholefield, Peile, Conington, Weil, Kirchhoff, Wilamowitz).
παρεσκευάσθαι, ‘to be equipped, ready’ is good and well grounded in examples,
and there is no difficulty in adding to it the adverbial expression ἐκ τῶν
ὁμοίων (this will be discussed below so as not here to divert attention from
the main issue). But what becomes then of the clause χειρὶ... dpyew? One
possible way out is to make the infinitive dependent on τοιαῦτα. So Hartung
and recently G. Thomson: ‘the presence of τοιαῦτα suggests that the construc-
tion is τοιαῦτ᾽ ἀπειλεῖν ὥστε... ἄρχειν, "I bid you hurl such threats as will
make you master after you have beaten in fair fight one who is prepared to
meet you’.’ This subordination of ἄρχειν seems to me quite unsuitable here,
since Ciytemnestra’s words unmistakably culminate in the challenge that
anyone who would rule over her should do it in virtue of deeds, not of threats.
It should also be noticed that if ὡς (= were) is taken with the infinitive, the
construction of the genitive absolute with ὡς (see above) is destroyed. But
Thomson’s remark on the function of τοιαῦτα in this sentence demands careful
examination. It is well known that τοιοῦτος can refer to something which
follows and that it often serves as the basis for an ensuing relative clause (ös,
οἷος etc.). τοιοῦτος may also prepare for a detailed exposition (e.g. Hdt. 1.
178. 1); occasionally, though seldom, it introduces direct speech (e.g. Hdt. 7.
5. 1; Ar. Knights 49; Thuc. 4. 58). τοιοῦτος dore, which is practically the
construction assumed by Thomson, is not uncommon, e.g. À. Ag. 1075; Hdt.
8. 132. 3; Ar. Knights 844; Thuc. 3. 49. 3, 8. 66. 2; Pl. Symp. 175 ἃ; Xen. Mem.
3. 4. 1. I have not come across an example of τοιοῦτος ws . . ., though such
may exist. Against all this, however, it must be remembered that the use of
τοιοῦτος to refer back to something already mentioned is far more common.
As regards, in particular, the present passage, it would need a good deal of
persuasion to convince us that the phrase τοιαῦτ᾽ ἀπειλεῖν does not allude to
the threats which have just been heard from the Chorus (1410 f.). But if at
this point we follow our natural instinct and take the beginning of the sen-
tence accordingly (‘I tell you: threats of this kind you should utter in the
certainty that I am armed in a similar way’), then the following clause
χειρὶ... ἄρχειν is completely in the air. To see in it an instance of explanatory
asyndeton seems inadvisable, for the clause, ‘he whose strong hand has been
victorious shall rule over me’ has no point of connexion in τοιαῦτ᾽ ἀπειλεῖν.
To bridge the gulf one might consider the possibility of taking λέγω out of its
context in the preceding sentence and making the accusative and infinitive
dependent on it. But this, especially in the absence of any connecting par-
ticle, seems so harsh that I do not dare to propose it. Wilamowitz’s χερί (re)
gives a sort of connexion,’ but he does not arrive at a satisfying construction ;
to have τοιαῦτ᾽ ἀπειλεῖν and yepi . . . ἄρχειν on the same level seems extremely
forced. The result is most unsatisfactory; I can only suggest that either
something has been lost after ὁμοίων or (less likely) that some corruption
lurks in the words χειρὶ... ἄρχειν.
1 The change from oot 1421 to νικήσαντα 1423 could be justified by reference to the usage
discussed on 378 ff. and 10s; f.
668
 COMMENTARY line 1423

1423. ἐκ τῶν ὁμοίων. In Passow’s lexicon, s.v. ὅμοιος, the phrase ἐκ τῶν
ὁμοίων παρεσκευάσθαι is quoted from Ag. 1423 and coupled with Thuc. 6. 21. 2
οὐκ ἐν τῶι ὁμοίωι στρατευσόμενοι, and for both passages the explanation is
given : ‘to be armed alike, to fight alike, so that no party has any preliminary
advantage over the other, on ground which favours both equally.’ L-S have
modified this by leaving out παρεσκ. and paraphrasing ‘with equal advantages,
in fair fight’ (9th edition: ‘on equal terms’, ‘in fair fight’). Accordingly
Verrall: ‘he who conquers me in fair fight’; Headlam: ‘conditions equal’. This
restricts the sense of the phrase unduly and gives it a meaning which it need
not necessarily have. ἐκ τοῦ ὁμοίου, which occurs several times in the speeches
of Thucydides! (1. 143. 4; 2. 44. 3; 3. 12. 3; 4. 10. 4; 6. 78. 4; 6. 87. 5), sometimes
with a clear reference to an ἐκ τοῦ ἴσου,2 ἀπὸ τοῦ ἴσου, ἐν τῶι ἴσωι in the vicinity,
means there ‘arising from like circumstances or a similar situation; in a
similar way’ and shows always that something happens or exists in a manner
which corresponds to the action (or situation) of the other side or party. The
point always is that par pari respondet. The meaning is obvious 3. 12. 3;
6. 78. 4; 6. 87. 5. Nor is 4. το. 4 difficult: it does not mean ceteris paribus, as
it is explained by K. W. Krüger and earlier editors (cf. below on Pl. Phaedr.
243 d). What the general says is: οὐκ ἐν γῆι στρατός ἐστιν ἐκ τοῦ ὁμοίου μείζων,
ἀλλ᾽ ἀπὸ νεῶν κτλ., i.e. the Spartans cannot use their numerical superiority on
land under the same conditions, in the same way, and in the same strength
(as we use our troops) but must rely on the fleet (and perhaps successive
landings of smaller units). In 1. 143. 4, where ex τοῦ ὁμοίου ἔσται is predicate,
the meaning is not different: ‘it will not be the same for the Athenians to
ravage part of the Peloponnese as for the Spartans to ravage the whole of
Attica; the damage caused to either side (because the Athenians have sea
power and the empire at their disposal) will be quite disproportionate’.
2. 44. 3: the point is that the possession of children benefits the State (a) ἐκ
τοῦ μὴ ἐρημοῦσθαι, (b) dopadreiar—this is explained by the observation that
only those can ἔσον βουλεύεσθαι who ex τοῦ ὁμοίου παραβαλλόμενοι κινδυνεύουσιν
(ἴσον could be perfectly well substituted for ἐκ τοῦ ὁμοίου). The comparison
here is not between two parties but between certain individuals and the
totality oftherest. Finally Pl. Phaedr. 243 d, where as in Ag. 1423 the plural
ἐκ τῶν ὁμοίων is used: here Heindorf, Stallbaum, and others have again (see
above) brought in ‘ceteris paribus’, which is wrong. Here, too, the meaning
is ‘under the same conditions as on the other side; like or corresponding to
the action of the other side’, so that ἐκ τῶν ὁμοίων χαρίζεσθαι comes very near
to an dvrepäv, ἀντιχαρίζεσθαι (cf. Thuc. 3. 12. 3 ἐκ τοῦ ἴσου... ἀντεπιβουλεῦσαι,
… ἀντιμελλῆσαι.... € τοῦ ὁμοίου ἐπ᾽ ἐκείνους lévai—this restoration is Krüger's).*
1 Other instances of ‘adverbial periphrasis’ with ἐκ and the adjective used as substantive,
such as ἐκ τοῦ εὐθέος, ἐκ τοῦ ἀφανοῦς etc., have been adduced from Thucydides by Classen—
Steup on 1. 34. 3. On ἀπὸ τῶν ὁμοίων in Thucydides cf. Schwyzer, Preuss, Akad. Phil.-hist.
Abhdl. 1942, No. 1o, 41 f.
2 CL, e.g., S. Oed. R. 563, where ὁμοίως and ἐξ ἴσου are strictly parallel. For the original
difference in meaning of ὅμοιος and ἴσος cf. R. Hirzel, Themis, Dike, etc. 421 ff.
3 Formulated by Felix Jacoby, whom I have consulted on my discussion of ἐκ τῶν ὁμοίων.
4 Jacoby, who agrees with my interpretation of Pl. Phaedr. 243 d, remarks: ‘I take éx
τῶν ὁμοίων to be a kind of pleonastic addition which gives a concluding touch of emphasis to
the clause expressing result: it is a fuller and more forcible ἀντιχαρίζεσθαι, so expressed
owing to the way the whole thesis is formulated, since one cannot ἀντιχαρίζεσθαι τῶι μὴ
ἐρῶντι,
Y ^ ,

669
line 1423 COMMENTARY

In the present passage παρεσκευασμένης is placed at the beginning because it

is on this notion that the emphasis is laid. ‘If you old men attack me now,
if you resort to accusations and threats, know this: you find me not unarmed ;
in truth I am armed for battle and prepared, ἐκ τῶν ὁμοίων, in a similar
situation as you, in a manner which corresponds to my adversary’s readiness
to fight.’
1424. Herwerden’s κράνηι has been accepted by Wecklein. The alteration
was probably caused by the desire to satisfy the rule that ‘when referring to
the future, ἐάν with the subjunctive of the present corresponds on the whole
to Latin s? c. fut. and ἐάν with the subjunctive of the aorist to Latin sz c. fut.
exact. (Kühner-Gerth, ii. 474). It almost seems as if the champions of
κράνηι have been determined by the translation of Schütz, 'sin autem con-
traria perfecerit deus'. That is, however, an incorrect translation because it
misses the particular sense of «patvew (cf. on 369). Stanley rightly: ‘sin aliter
decernat deus'. The present is therefore not only possible but necessary.!
1425. Therefore an ὀψιμαθής. For the thought cf. on 71o.
1426. μεγαλόμητις only here (αἰπυμῆτα only Prom. 18). The Homeric
πολύμητις, δολόμητις invited parallel formations (Hesiod's αἰολόμητις was taken
over by Aeschylus, Suppl. 1037). Cho. 626, with reference to Clytemnestra,
γυναικοβούλους μήτιδας. ‘Outside Ionic pris seems only to exist as a borrow-
ing from Homer’ (Wilamowitz, Heimkehr des Odysseus, 190 n. 1). It is not
easy to determine the implications of μεγαλόμητις here. Some commentators
want to discover in it a purely favourable judgement, others a purely un-
favourable one; thus unfavourable, e.g., Wecklein: ‘von verwegenem und
rücksichtslosem Sinn’, Headlam: ‘monstrous are thy devices’; on the other
hand, favourable Wilamowitz : 'Weib, du bist gross und kühn', and similarly
Kranz, Hermes, liv, 1919, 314: 'in the antistrophe the Chorus cannot suppress
a certain admiration for the traitress (μεγαλόμητις el). In actual fact the
expression seems to have been chosen to convey the idea not only of the
greatness of Clytemnestra’s spirit but also of her nature being μεῖζον 7 κατ᾽
ἄνθρωπον (or κατὰ γυναῖκα). Thus the powerful word forms not so much a
contrast with as a preparation for the following clause. In the Iliad μέγα
φρονέων, of men or animals, means praise ; strong praise it means also in the
words of Protagoras (B 9) μεγαλόφρονά τε xai ἀνδρεῖον... καὶ ἑαυτοῦ κρείσσω;
Plato uses it in the same sense (with ἀνδρεῖος Rep. 8. 567 b), cf. on that use
Wilamowitz, Platon, i, 3rd ed., 339 n. 1 (‘equivalent to the later μεγαλόψυχος ’).
In Herodotus μεγαλοφροσύνη, though in 7. 136. 2 it indicates the μεγαλοψυχία
of Xerxes, in 7. 24 means the desire for fame, which causes him to carry out
the unnecessary labour of digging through Mt. Athos.
περίφρονα: glossed in Tr with $mépópova, as on Suppl. 757 περίφρονες the
scholia say dvri τοῦ ómépópoves. That is undoubtedly right for both passages.
Wilamowitz remarks on Suppl. 757: ‘ita [ie. περί = ὑπέρ] certe Lesbii
loquebantur'. The analogy of Lesbian is valuable but should not be regarded
as pointing to Lesbian influence. Parallels in Attic are not lacking: Thuc. x.

ı Ag. 1667 shows a similar conditional clause; to the future ἐγώ oe... μέτειμ᾽ ἔτι
the retort is οὐκ, ἐὰν δαίμων ' Opéarqv Seip’ ἀπευθύνηι μολεῖν. The form ἀπευθύνηι does not
indicate its tense; the sense, as in 1424, supports the present, and by way of confirmation it
may be noticed that the aorist active of εὐθύνω (ameud., παρευθ.) hardly seems to occur in
pre-Hellenistic Greek, whereas the present stem is common.
670
 COMMENTARY line 1428
25. 4 mepibpovoüvres αὐτούς is, both in sense and construction, exactly like
Örrepppoveiv as used by Thucydides elsewhere. Ar. Clouds 225 f. contains a
play on the double meaning of περιφρονεῖν. The use οὗ περιφρονεῖν = ὑπερ-
φρονεῖν probably already existed in Aeschylus’ time.? In other cases, e.g.
περιεῖναι περιγίγνεσθαι, this meaning of the preposition is usual; cf. also the
Thucydidean use of περιέχειν. That περί = ὑπέρ can be found, though less
commonly, in other dialects than Lesbian, has been noticed by H.L. Ahrens,
De Graec. ling. dial. i. 15x (cf. Bechtel, Griech. Dial. i. 111). W. Schulze,
Kuhns Zeitschr. xliv, 1911, 359 = Kl. Schr. 396, points out that ‘from the
fourth century B.c. in Attic and the κοινή the boundaries between περί and
ὑπέρ become obscure’ ;3 cf. P. T. Stevens, C. Q. xxx, 1936, 208 f.
1427. Hermann remarks: ᾿ ὥσπερ οὖν particulae sunt quibus confirmantur
quae adiuncta sunt’; he compares 1171, Cho. 96, 888( elsewhere ὥσπερ οὖν
does not occur in Aeschylus), S. 47. 991 (for allied matter cf. Denniston,
Particles, 421) and rightly draws the conclusion: ‘ex his intelligitur conexas
esse istis particulis sententias’. Therefore Schütz is wrong: ‘haec particula
perspicue prodit in τύχαι latere vocabulum, quod metaphoram contineat’.
The only question is whether the comparative clause ὥσπερ... ἐπιμαίνεται
belongs to what precedes or to what follows. Hermann took it with what
follows (we need not discuss his interpretation in detail because the infinitive
ἐμπρέπειν Which he introduces [πρέπειν still in Headlam] ruins the sense) ; so
also Wilamowitz. On the other hand, the comparative clause ὥσπερ οὖν...
ἐπιμαίνεται is construed with the preceding μεγαλόμητις εἶ κτλ. by Stanley,
Paley (‘magna meditaris et superba loqueris sicut animus caede delirat’),
Conington (‘thy words breathe haughtiness even as thy soul is raving’),
Sidgwick, Verrall, and others. This interpretation seems to me the only
natural one: 'Your words and your thoughts are overbold, in tune with
your mind's raving because of the bloody deed.' The function of ὥσπερ οὖν
here is in strict conformity with the other Aeschylean passages.
φονολιβής, possibly coined for this passage, otherwise only in Ewm. 164.
‘Because of the murder-dripping event your mind raves.' The dative (φονολ.
τύχαι) does not, as is usual with ἐπιμαίνεσθαι, depend upon ém- but is causal,
the verb being used ‘absolutely’ as Sept. 155 (rightly L-S). Although the
Elders in these stanzas use very outspoken language, they refer to Clytem-
nestra's deed by the euphemistic τύχη. Perhaps we are justified in seeing in
this remarkable discretion the effect of deep-rooted αἰδώς.
1428. λίπος I leave in the text though not without qualms. λίπος, like λίπα,
λιπαρός, etc., is normally used of a fatty substance. The comparison (Abresch
! The later (?) scholia (there is no trace of this note in RV) are right here: 76 περιφρονῶ δι-
πλοσήμαντόν ἐστι, καὶ ἀντὶ τοῦ περισκοπῶ καὶ ἀντὶ τοῦ ὑπερφρονῶ.... à Στρεψιάδης δὲ ἀντὶ τοῦ
καταφρονῶ μόνον ἐνόησεν. The joke is ruined if we adopt from the Venetus περιφρονεῖς
(Blaydes, Starkie) in 226 and add the stopgap σὺ.
2 It is surprising that Schuursma, who has observed the use of περιφρονεῖν, still thinks it
probable (De . . . abusione ap. Aesch. 136) that περίφρων ‘per abusionem novo sensu usur-
patum esse' by Aeschylus.
3 For this reason I have not taken into account later instances of περιφρονεῖν = 'to
despise'.
* For the differentiation of ἀτύχημα, ἁμάρτημα, and ἀδίκημα see Arist. Eth. Nic. 5. το,
p. 1135611 ff. Cf. Menander fr. 426 Kock ἀτύχημα κἀδίκημα διαφορὰν ἔχει" τὸ μὲν διὰ τύχην
γίνεται, τὸ δ᾽ αἱρέσει and the euphemistic ἀτυχήματα and συμφοραί in Demosth. 23. 7o. In
general cf. the remarks of Latte, Archiv f. Religionswiss. xx, 1921, 278.
671
line 1428 COMMENTARY
and his followers) with 5. Ani. 1022 ἀνδροφθόρου βεβρῶτες αἵματος λίπος does
not help because that is a brief indication that the birds have fed on the
corpse, not only on the blood but on blood and the fatty flesh together.
Blood is not in itself an oily or fatty liquid. Casaubon wrote Aißos which is at
least a very good conjecture: the word is used of tears in Cho. 448. The
corruption would not, however, be easy to understand here (‘the copyists with
φονολιβεῖ τύχαι before them would hardly have introduced another word had
that been in the original text’ Conington).
ἐπ᾿ ὀμμάτων; taken by the majority of commentators since Schütz as
‘super oculis i.e. in fronte’; thus, e.g., Wilamowitz (translation and in his
edition, ‘Actio’ to 1372): ‘Clytaemestra, cuius frons sanguinis guttis adspersa
est’, Mazon, MacNeice: ‘your crazy heart fancies your forehead with a smear
of blood’, Murray (cf. his book Aeschylus, 229): ‘the blood on your brow’.
This is based on the belief that we have here an allusion to the blood-stain
of which Clytemnestra has spoken (1389 f.). This view was put forward by
Abresch. His misinterpretation of ἐπ᾿ ὀμμάτων was anticipated by the
Byzantines: in Tr we find above it the gloss ἐπὲ τοῦ προσώπου. That én’
ὀμμάτων could mean ‘over the eyes’ is difficult to believe.! Stanley’s transla-
tion (otherwise he completely misunderstood the passage) in oculis is in this
respect quite correct. A Greek sometimes uses ἐπὲ τῶι ὀφθαλμῶι or ἐπὶ τοῦ
ὀφθαλμοῦ where we say ‘in the eye’, i.e. in certain cases a Greek, speaking of
the eye, takes primarily the eyeball into account. This is true not only when
speaking of the bloodshot eye, e.g. Hippocr. Eid. 7. 11 (vol. v, p. 382 Littré)
καὶ ἐπὶ ὀφθαλμοῦ τοῦ δεξιοῦ τὸ ὕφαιμον ἦν (similarly 7. 5, p. 374 f. L. καὶ χρῶμα
ἐπ᾽ ὀφθαλμοῦ δεξιοῦ, οἷον εἴρηται τὸ ὕφαιμον), but also where something, as we
say, gets in the eye: Ar. Lys. 1025 f. τόδε τὸ θηρίον τοὐπὶ τὠφθαλμῶι,32 i.e. the
gnat which the woman with the help of a ring éxoxaAever. Verrall says nothing
of this particular way of looking at things that happen to the eye and of the
use of ἐπί which expresses it,? but it is very much to his credit that he under-
stood the passage (G. Thomson did well in following him): 'referring not to
a stain of blood from the murdered man (which does not suit er’ ὀμμάτων)
but rather to the bloodshot eye, which they see, or suppose themselves to see,
in the furious face of the murderess’. The suffusion of the white of the eyeball
with blood is for the old men a certain symptom of Clytemnestra's madness.
The madness of Euripides' Herakles begins in the same way ; he is described
(932 f.) as rolling his eyes wildly and ῥίζας 7’ ἐν ὄσσοις αἱματῶπας ἐκβαλών.
This agrees with medical texts (quoted by H. Harries, Tragici Graeci qua arte
usi sint in describenda insania, diss. Kiel 1891, 19 £.). Not only is it regarded
in general as a bad symptom ἣν (scil. οἱ ὀφθαλμοῦ διαστρέφωνται 7 ὁ ἕτερος τοῦ

1 It may be noted in passing that as regards the use of the singular ὄμμα, in one or two
places in tragedy the meaning ‘face’ given by L-S s.v. IV and by Jebb is incorrect; the
right meaning, ‘Anblick’, is found in Kaibel on S. El. 903.
2 There is no difference in meaning between ἐπί with the dative here and ἐπί with the
genitive in the passages of the Ilippocratean writings and in Ag. 1428, cf. L-S s.v. B. i,
beginning.
3 I do not know whether the exceptional expression ἔτ᾽ εἰσορᾶν oe, πάτερ, ἐπ᾽ ὀμμάτων
δοκῶ of E. Suppl. 1153 in place of the usual ἐν ὄμμασιν (ἐν ὀφθαλμοῖς) owes anything to the
influence of this use. At Theocr. 4. 7 H. L. Ahrens in his editio minor and Wilamowitz
(without any note in the app. crit.) print ἐπ᾽ ὀφθαλμοῖσιν ὀπώπει, though as may be seen
from Ahrens’s editio maior én’ is only in Parisinus 2884 (Q), while the other MSS and the
lemma of the scholia have ev. [No preposition in K; Gallavotti’s W has ἐπ᾽.)
672
 COMMENTARY lines 1429 f.
ἑτέρου ἐλάσσων γίνηται ἣ τὰ λευκὰ ἐρυθρὰ ἴσχωσιν (Hippocr. Prognost. 2, vol. i,
p. 80. 11 ff. Kühlewein), but π. ἱερῆς νούσου 15 (vol. vi, p. 390 Littré) diseases
of the brain are described in these words: ὥσπερ οὖν καὶ ἐγρηγορότι τότε
μάλιστα τὸ πρόσωπον φλογιᾶι καὶ οὗ ὀφθαλμοὶ ἐρεύθονται, ὁκόταν φοβῆται καὶ ἡ
γνώμη ἐπινοέηι τι κακὸν ἐργάσασθαι, οὕτω καὶ ἐν τῶι ὕπνωι πάσχει. Schol. E. Or.
256 mentions as generally known τοὺς μαινομένους ὕφαιμον βλέπειν καὶ ταρα-
χῶδες. It is scarcely necessary to recall Virgil, Aen. 4. 643 ff. sanguineam
volvens aciem maculisque trementis interfusa genas . . . furibunda, 7. 399
(Amata when driven mad by the Fury) sanguineam torquens aciem and
the like.’ I regard it as a great gain for the true appreciation of the end of
the tragedy that we are now quit of the picture of Clytemnestra treading the
stage with ‘blood-bespattered forehead’. The horror told in 1389 f. has lost
nothing in strength ; the poet may trust the power of his description and need
not give us a visible reminder. The time at which migravit ab aure voluptas
omnis ad incertos oculos et gaudia vana had not yet come when Aeschylus
wrote his work.
εὖ πρέπει ‘stands well out, is set off clearly’ (cf. on 242). Exactly parallel
Suppl. 722 eümpenros (quoted by Klausen). Editors of Aeschylus should note
such parallels rather than print ἐμπρέπει over and over again. The verb
indicates as sharply as possible the conspicuousness of the bloodshot eye-
balls; a similar colour effect is described in Suppl. 719 f. πρέπουσι δ᾽ ἄνδρες
νήϊοι μελαγχίμοις γυίοισι λευκῶν ἐκ πεπλωμάτων ἰδεῖν.
1428 follows on with an explanatory or causal asyndeton: ‘you are mad:
your eyes, as can be seen, are bloodshot’. The statement is completed by
mentioning this proof.
1429 f. ἄντιτον ἔτι σὲ χρὴ κτλ. It is doubtful whether the asyndeton here has
a particular function (adversative? consequential?); the immediate juxta-
position suits the emotion of these violent sentences.
ävrirov. Looking at most editions one would think that the παράδοσις of the
text was arierov only. Of course it is possible that Tr has preserved the
original and avrierov? in FG is a copyist’s mistake. But it is equally possible,
as Weil (1858) saw, that the reading of FG conceals an original ἄντιτον. As
soon as we grasp this point (after Weil, who later took a wrong turning again,
only Wilamowitz seems to have achieved this ;? Murray rightly followed him),
we cannot fail to see that ἄντιτον is not only much better than ἀτέετον but
singularly suitable and therefore genuine. No doubt Aeschylus followed the
Homeric ἄντιτα ἔργα. ἄντιτα means ‘done in requital or revenge’ (L-S
Addenda, p. 2050). The ancient explanation (though rejected by Lobeck,
Pathol. i. 360), ἄντιτα = ἀντίτιτα (Schol. 2 215, cf. Eust.; Schol. p 51), ‘hits the
nail on the head’ (Bechtel, Lexilogus, 47). This is a case of ‘dissimilatorischer
τ For another change in the colour of the eyes as a result of madness see Plaut. Men.
828 f. viden tu illt oculos virere? ut viridis exoritur colos ex temporibus atque fronte, ut oculi
scintillant, vide; this too is undoubtedly based on medical views, and so is 923, where the
physician asks the supposed madman solent. tibi umquam oculi duri fieri? (cf. E. Or. 389
δεινὸν δὲ λεύσσεις ὀμμάτων ξηραῖς κόραις).
2 The wrong word-division is of course οὗ no significance. FG give πρέπειαν τίετον. The
reading in F is inexactly recorded in Hermann, Wecklein, Wilamowitz, etc., but correctly
in Franz, van Heusde, Weil (Teubner edition), Mazon.
3 Keck, who also prints dvrırov in the text, has completely upset the construction of the
sentence; Verrall stops half-way.
4872-3 o 673
lines 1429 f. COMMENTARY
Silbenschwund’, i.e. haplology caused by a dislike for the repetition of
the same or a very similar syllable (cf. on 1438 about δατήριοι from *darn-
THptor) ; ἀμφορεύς = ἀμφιφορεύς is a universally known example of this
development in the case of a preposition ending in «; some further examples in
Kretschmer, Glotta, i, 1909, 43 f. It is obvious that ἀντίτιτον (cf. also 1263
ἀντιτείσεσθαι), ‘done in revenge’, is exactly right here; its predicative position
strengthens τύμμα τύμματι τεῖσαι. “Now you are arrogant in the frenzy of
madness ; vengeance is certain’ : ἄντιτον ‘in vengeance’ is put first for emphasis.
The reading established by sense is confirmed by metre. ἄντιτον ἔτι σε χρή
—- UV U U v- corresponds to 1411 ἀπέδικες ἀπέταμες vvv vuvvuvu;lfor the
free responsion of the first syllable (Ὁ) cf., e.g., Suppl. 393 κράτεσιν ἀρσένων =
403 Ζεὺς ἑτερορρεπής, 394 μῆχαρ ὁρίζομαι = 404 ἄδικα μὲν κακοῖς ; for the free
responsion at the end (Ov) cf. Ag. 1165 μινυρὰ θρεομένας = γοερὰ θανατηφόρα.
ἔτι: common in threats, as here and 1666 ; cf. van Leeuwen on Ar. Ach. 1156,
Denniston on E. El. 485; cf., e.g., Prom. 907, Eupolis δῆμοι fr. 3 b. 9 (p. 48
Demiañczuk, Suppl. Com.).
1430. The ‘lex talionis’ is here very exactly formulated. Everyone will be
reminded of Cho. 309 f. ἀντὶ μὲν ἐχθρᾶς γλώσσης éx0pà γλῶσσα τελείσθω, 312 f.
ἀντί δὲ πληγῆς φονίας φονίαν πληγὴν τινέτω. W. Kranz, Siasimon, 133, regards
these as ‘doublets of archaic legal language’.
1431. καὶ τήνδε: preparing for what follows. Cho. 500 is very similar: καὶ
τῆσδ᾽ ἄκουσον λοισθίου βοῆς, πάτερ. But there the verb is ἄκουσον. We expect
at first something similar here; certainly Triclinius did, since he says:
ἀκούεις ἀντὶ τοῦ ἄκουσον. Casaubon considered the conjecture ἄκουσον. Stan-
ley: ‘malim ἀκούοις vel ἀκούσεις᾽.2 Several other conjectures have been put
forward (cf. also Verrall’s note, who, however, regards the MS reading as not
impossible) ; the weak ἀκούσηι γ᾽, considered by Headlam, has been put into
the text by G. Thomson. It is even worse to make ἀκούεις into a question
(Pliiss and Mazon). To understand the tone of this sentence we have to
consider the structure of the whole section. What Clytemnestra says 1431 ff.
is the immediate continuation of 1421 ff. λέγω δέ σοι κτλ. Therefore the effect
of καὶ τήνδ᾽ ἀκούεις... θέμιν is almost as if she had said: ‘and there is some-
thing else that I have to tell you’; she might perhaps have said καὶ τήνδε
σοι... Aéyw.... Only a deaf ear could fail to hear how much stronger her
ἀκούεις is than if she said ἄκουσον or the like (in the passage Cho. 500, which is
similar in form, ἄκουσον follows a series of requests and prayers). Some
scholars were not blind to the merits of the MS reading. After Pauw had
pointed out the ‘emphasis’ on ἀκούεις, Hermann said of any critic who objected
to the present here that he was ‘gravem a debili orationem discernere ne-
sciens’. This use of ἀκούεις in adding a final statement to the preceding speech
is parallel to κλύεις in similar passages, cf. on 348.
θέμιν : ‘Clytemnestra is not giving them the formula according to which she
usually swears her oaths—that would be absurd—but she points to the
1 Several editors are concerned neither about responsion nor about the possible forms of
Aeschylean dochmiacs. Not so A. Platt, J. Phil. xxxv, 1920, 92. He is judicious enough to
defend 1411 against alteration (‘to insert τ᾽ after ἀπέταμες is simply wicked’), but when in
his desire to keep ἀτίετον and avoid a conflict with metrical canons he asserts that ‘ σὲ is
unnecessary’ and writes drierov ἔτι χρή, few will be willing to follow him.
2 Even Blomfield put ἀκούσεις in the text, though Henr. Stephanus knew and said long
ago that the future ἀκούσω was foreign to the ancient language.

674
 COMMENTARY line 1434
power which lies in the oath that she swears now and which must urge her to
speak the truth’. (R. Hirzel, Themis, etc. 47 n. 3.) With old and complex
ideas like θέμις a translation is unsatisfactory because, unless it is left quite
vague, it of necessity renders only one of the special applications of the word
and so fails to include anything like the whole of what was implied for the
ancient audience.
1432. τέλειον : cf. on 997 τελεσφόροις.
1432 f. μὰ τὴν... Δίκην "Army Ἐρινύν te. On Sept. 45 Wilamowitz says in
the ‘Corrigenda’ of his edition (p. 382): ‘omissam in medio trium substanti-
vorum copulam spero me aliquando defensurum esse.” But the particular
difficulty of the verse Ἄρη τ᾽ ᾿Ενυὼ καὶ φιλαίματον Φόβον does not lie in the
lack of connexion for the middle member but in the re after Ἄρη (to accept
Ἄρην introduces an error of language) since re cannot join the main verb to
the preceding participle (cf. above on 98). For the particular arrangement of
a triad described by Wilamowitz (I and II unconnected, III with a connecting
particle) the present passage is a good example and there are others: Pers.
312 f. Ἀρκτεὺς ᾿Αδεύης καὶ φερεσσάκης τρίτος Dapvoüxos,! S. Aj. 296 f. συνδέτους
ἄγων ὁμοῦ ταύρους, κύνας βοτῆρας, εὔερόν τ᾽ ἄγραν, E. Phoen. 1147 γυμνῆτες
ἱππῆς ἁρμάτων T’ ἐπιστάται.2
‘The tendency to ensure the sacrosanctity and effectiveness of an oath by
appeal to a trinity of divine witnesses and assistants is very old and almost
universal’, Usener, Rh. Mus. lviii, 1903, 17 (with rich collection of instances),?
cf. also G. Glotz, Etudes sur l'antiquité grecque (1906), 104 ff. Naturally
Clytemnestra swears her very special oath by three 'Sondergótter' suited to
her particular situation.
1433. αἷσι τόνδ᾽ ἔσφαξ᾽ ἐγώ. ‘Sauf dans les cas insignifiants, le serment ne
va pas sans un sacrifice: c'est l’épxwpdotor, où les victimes offertes sont des
ὅρκια ᾿ (Glotz, op. cit. 110).
1434. On this passage Abresch: ‘ φόβου ἐλπίς pro φόβος, et μέλαθρον accipio
simpliciter pro aedibus et palatio Clytaemnestrae, ut ἐμπατεῖν μέλαθρον sit
inhabitare aedes’. Paley likewise took φόβον ἐλπίς together, quoting for com-
parison Thuc. 7. 61. 2 τὴν ἐλπίδα τοῦ φόβου ὁμοίαν ταῖς ξυμφοραῖς ἔχουσιν, and
adopting the conjecture ἐμπατεῖν. Some later editors, however, keep ἐμπατεῖ
and adhere to Abresch’s explanation; so Nägelsbach, Enger in the second
edition of Klausen's commentary (‘ φόβου ἐλπίς timoris exspectatio’, albeit
with some doubt about the text), Wilamowitz (‘Ich hoffe, meinem Hause
1 It is very doubtful whether Aesch. fr. 205 N. Awa δὲ πίσσα κὠμολίνου μακροὶ τόνοι should
be quoted in this connexion, since in the first place we do not know what went before it and
what followed, and in the second πίσσα is contested (cf. Wilamowitz’s editio maior of
Aeschylus, p. 180, and his Interpr. 129 n. 2: his suggestion is ingenious but by no means
certain; ὠμολίνου μακροὶ τόνοι might for instance mean ship's cordage, as Blümner, Tech-
nologie, i, 2nd ed., 197 n. 4 takes it, and fr. 205 might belong to the same speech as fr. 194,
cf. Prom. 462-8; or it might conceivably mean something quite different again). Focke,
Hermes, \xv, 1930, 267, regards λινᾶ δὲ as corrupt.
2 Two of these examples are quoted in Denniston, Particles, 501 c, where, however, no
distinction is drawn between the triads and the quite different longer series. I had included
in my brief list 5. Ichn. (fr. 314 P.) 73 θεοὶ Τύχη καὶ δαῖμον ἰθυντήριε, but now (1946) I learn
from E. Siegmann, Untersuchungen zu Soph. Ichneutai (diss. Hamburg 1941), 11 and 39, that
the papyrus has θεὸς,
3 He quotes p. 21 A. Sept. 42 ff. (swearing by Ares, Enyo, Phobos) but does not mention
Ag. 1432 f.
675
line 1434 COMMENTARY

wird kein Schrecken nah’n’), Platt, J. Phil. xxxv, 1920, 92 (‘the expectation
of terror walks not in my halls’), though his translation, published 1911, gives
‘my expectation walks not in the hall of terror’, Ubaldi and G. Thomson
(‘no thought of fear shall walk beneath this roof’). Kennedy objected to
the connexion of φόβον ἐλπίς, which he called harsh; he might have said
‘incredible’. I have gone through all the dialogue of Aeschylus and Sophocles
without finding a passage where the position of a genitive, separated from its
governing noun which follows later in the sentence, and put immediately in
front of another noun, results in an ambiguity such as we should have here if
we followed the editors who take φόβου ἐλπίς together.! Moreover it would,
in the language of Aeschylean dialogue, be very irregular if the group φόβου
ἐλπίς were separated by μέλαθρον: see Appendix F. We must therefore con-
clude that as the line stands φόβου goes with μέλαθρον. If the text is sound,
it can only mean 'for me hope does not walk the house of fear'. It has been
understood thus by most commentators except where conjecture has been
called in. Hermann protested vigorously against the idea supposed to be
expressed by φόβου μέλαθρον; he thought the text to be corrupt. Paley is
milder: ' $. μ., "the Hall of Fear", is a phrase almost too figurative for
Aeschylus, though it might perhaps be compared with the personification of
Wealth, v. 1334, μηκέτ᾽ ἐσέλθηις.᾽ Sidgwick, however, thinks ' “Hope doth
not tread for me in the halls of Fear” a fine picturesque phrase, surely not too
imaginative or metaphorical for Aeschylus'. This may be right. It is con-
ceivable that the expression did not sound too strange to Aeschylus' audience
—if, that is, we may assume a reference to something they knew. We might
perhaps indulge in a flight of fancy and picture some older (‘Hesiodic’?) poem
in which two houses figured, the house of Θάρσος and the house of Φόβος, into
which ’EAris might enter,? ’EAris representing expectation in general. This
well-known sense of ἐλπίς (somewhat different from the one-sided use in
malam partem as, e.g., in A. Sept. 367) is neatly defined by Plato, Laws 644 c
δόξας μελλόντων, οἷν κοινὸν μὲν ὄνομα ἐλπίς, ἴδιον δέ, φόβος μὲν ἡ πρὸ λύπης
ἐλπίς, θάρρος δὲ ἡ πρὸ τοῦ ἐναντίου (cf., e.g., E. Or. 859 1. προσῆλθεν ἐλπίς, ἣν
φοβουμένη πάλαι τὸ μέλλον ἐξετηκόμην γόοις). The same polar ideas are con-
trasted here: the sentence begins with φόβου and ends with θράσους. Clytem-
nestra seems to say ‘for me expectation walks no house of fear, so long as
Aegisthus watches over my hearth, since he is to us an ἀσπὶς οὐ σμικρὰ
θράσους '. It would hardly be in Aeschylus’ manner to let Clytemnestra first
say où uot . . . μέλαθρον ἐμπατεῖ and immediately afterwards ἐφ᾽ ἑστίας ἐμῆς
without meaning the same house in both cases. Therefore ‘For me Elpis
ı That a similar position of the genitive need not produce the slightest ambiguity may be
seen e.g. from O 690 f. ἀλλ᾽ ὥς τ᾽ ὀρνίθων πετεηνῶν αἰετὸς αἴθων ἔθνος ἐφορμᾶται κτλ.
2 A motive which recalls that of the two gates through which 7 562 ff. dreams pass, those
which come in by the one being false and the others true. ‘ ὑγέεια and νόσος are γείτονες
ὁμότοιχοι, live next door to one another, Ag. 1003 ’ (Beazley).
3 A different relation between φόβου μέλαθρον and éd’ ἑστίας ἐμῆς is assumed by H. L.
Ahrens, De causis quibusdam Aeschyli nondum satis emendati (Schulprogramm Ilfeld 1832),
32: ' "Spes," inquit, "quae nunc in mea domo versatur non abibit in Timoris aedes (i.e.
non mutabitur in timorem), dum Aegisthus mihi auxilio est, quasi ignis in ara [thus like,
e g., Sidgwick after him he takes αἴθειν as intransitive, as it appears in Pindar and Sophocles;
it is not probable here, but this does not affect the main issue]," quo exstincto spem
hospitem in alienam domum abire consentaneum erat.' 1 do not find this convincing, and
in particular in the Greek with its ἐμπατεῖ there is nothing to bear out the contrast between
676
 COMMENTARY line 1436

treads not the house of Fear’ must mean ‘treading my house, E. treads not the
house of Fear’. It is a conception akin to that which is expressed in easier
and more natural words by Euripides, Heraclid. 996 τὸ λοιπὸν μὴ συνοικοίην
φόβωι (compared with the present passage by Eimsley and Bothe). μοι,
dative of the person concerned, qualifies the whole statement : ‘for me, in my
case’ (‘in the case of my house’). This seems a fairly satisfactory explanation.
If the reader objects to it as too harsh and obscure, it is not easy to refute his
contention. In that case nothing seems to be left but to assume a deep-
seated corruption in 1434, which is unlikely. There is no reason to query the
number of μέλαθρον although the plural is more common, nor is there any-
thing against the unusual ἐμπατεῖν, not elsewhere found with this sense. For
the meaning of πατεῖν cf. 1298, and L-S s.v. at the beginning ; cf. also ἐμβαίνειν.
1435. ‘The master of the house performs the family sacrifices at the hearth-
stone; so long as he does so he is master and protector of the family’
(Schneidewin). ‘Every pater familias conducts the religious ceremonies him-
self in his own courtyard’ (Wilamowitz in a general account, Glaube d. Hell.
i. 39). van Heusde ad loc. recalls the Aeolic and Dorian (Pollux 1. 74; 10. 20)
ἑστιοπάμων. In ascribing to Aegisthus the aifew πῦρ ἐφ᾽ ἑστίας ἐμῆς
Clytemnestra assigns to him the position of the legitimate lord of the house.
Conversely, those who regard Aegisthus as a usurper call the hearth in
Clytemnestra’s house ἀθέρμαντον ἑστίαν δόμων (Cho. 629).
1436. Αἴγισθος (‘nuric primum diserte memoratus', Klausen) emphatically?
placed at the end of the clause? (for the words ós . . . ἐμοί form a self-contained
kolon) and at the beginning of the line.? Αἴγισθον similarly S. El. 957 (cf.
Vahlen, Opusc. i. 320, and Kaibel ad loc.) and E. El. 764.

nunc and abibil; but the point is so ingeniously made that I could not withhold from the
reader this example of the youthful work of a great scholar.
! Headlam would hardly have accepted this. The lengths to which one can go if like
Goodell (Transact. and Proceed. Am. Phil. Ass. xxi, 1890, 5 ff.) or like Headlam and G. Thom-
son one categorically denies that the final position in a sentence or clause is quite often
emphatic (though the initial position is of course much more commonly so used) may be
seen from Thomson's treatment (C.Q. xxxiii, 1939, 148) of passages such as E. Hel. 604
ἀγγέλλεις δὲ τί; Ar. Frogs 1439 νοῦν δ᾽ ἔχει τίνα; (cf.Clouds 239 ἦλθες δὲ κατὰ τί; contrasted with
Birds 916 κατὰ τί δεῦρ᾽ ἀνεφθάρης; and much else of the same kind, cf. Thomson, 149 f.).
Sentences like these Thomson lumps together with those where the interrogative stands at
the beginning of a kolon, often assuming an antithesis in the thought where in the text
there is no indication of it. Even Ar. Knights 1078 τούτοις 6 μισθὸς τοῖς ἀλωπεκίοισι ποῦ; he
manages to twist round by elaborate argument.
2 For the proper name pushed to the end of the sentence cf. on 523 and 681 ff.
3 The effect is different in Cho. 134, where Αἴγισθον at the beginning of the line has peculiar
force. If any punctuation is necessary, I think there should be a comma after ἀντηλλάξατο
at the end of 133. I do not believe that the audience of Aeschylus was in danger of taking
ἄνδρα here as a substantival predicate as most modern readers appear to do, e.g. H. Voss:
‘und zum Mann ertauschte sie sich den Aigisthos’; so Franz, Droysen, Paley (‘she has
received in payment Aegisthus for a husband’), so too in general Verrall and Headlam (‘and
in exchange she hath purchased for a mate Aegisthus"). avrnAAd£aro, which continues the
idea of πεπραμένοι in 132, has its technical sense, just as in the epigram on those fallen at
Potidaea (Tod, Greek Hist. Inscr. no. 59) ψυχὰς δ᾽ ἀντίρροπα (note the image of the scales)
θέντες ἠλλάξαντ᾽ ἀρετήν. Thereforé: ‘and in exchange (for her children) she has purchased
a man’( or ‘a husband’)—but what a man: Atyıodov—then, worst of all: ὅσπερ σοῦ φόνου
μεταίτιος. With bitter mockery Αἴγισθον κτλ. is kept to the end after the construction of the
sentence has been completed. After writing this I found that Tucker had interpreted the
passage in the same way.

677
line 1436 COMMENTARY
Clytemnestra’s assurance, when stripped of the decoration of tragic
phraseology, stands pretty near to everyday speech: cf. Ar. Knights 395 οὐ
δέδοιχ᾽ ὑμᾶς, ἕως av ζῆι τὸ βουλευτήριον.
1437. The shield is the symbol of protection and safety as Suppl. 190 κρεῖσσον
δὲ πύργου βωμός, ἄρρηκτον σάκος.
The arrangement of 1431-7 was mistaken by Wecklein and Wilamowitz:
Wecklein brackets 1434-7 and comments ‘the asyndeton in 1438 shows that
the θέμις announced in 1431 follows there’; Wilamowitz places 1431-3 after
1437. Both editors therefore make the oath introduced by μά belong to an
affirmative assertion, which does not occur in classical Greek."
Clytemnestra swears by the Ain of her daughter and the gods of vengeance
that her thoughts have no dealings with fear, at the moment when for the
first time she is conscious of a whisper of fear in her heart. ‘Aegisthus will
protect me still as he has done hitherto.” These words do not come from her
ἀνδρόβουλον κέαρ. She who speaks them is not quite the same woman who a
little while ago was so proud of her ability to take decisions and to act alone
like a man, and who could so recently (1402) declare without any reservations
that she spoke ἀτρέστωι καρδίαι. She is not yet broken; she perseveres for
some time in fighting and using blasphemous language, but the descent from
her summit has begun (cf. p. 694, top). She feels herself in need of some
powerful encouragement, hence her elaborate appeal to the gods whom she
wants to support her oath. After she has thus thrust from her the latest
threats of the Chorus and at the same time silenced her own secret fears, she
returns, with a new unconnected opening (1438), to the subject of her triumph
and to renewed assurances that her deed was just.
1438. κεῖται κτλ, 'Graviter hoc verbo κεῖται hanc. sententiam incipit Cly-
taemnestra’, Hermann.” The emphatic position of the verb recalls the
Homeric «etra: ἀνήρ (E 467, II 558), κεῖται Σαρπηδών (IT 541) etc. So too, e.g.,
E. Or. 366 κεῖται σὸς κασίγνητος θανών.
Hermann, refusing to admit a substantival use of λυμαντήριος, inserted after
1438 a line beginning with ἀνήρ (on the model of Cho. 764 f. ἄνδρα τῶνδε
λυμαντήριον οἴκων). His objection has won much support, but is unfounded.
Cf., e.g., Sept. 710 f. ἐνυπνίων φαντασμάτων ὄψεις, πατρώιων χρημάτων δατήριοι,
where δατήριοι (from Ἐδατητήριοι, cf. Ernst Fraenkel, Glotia, i. 1909, 272, and
Schwyzer, Griech. Gramm. i. 467) is in the Mediceus well glossed by μερισταΐ:
the dream visions, here closely connected with the curses of Oedipus, are
represented as ‘distributors’, as in 944 f. πικρὸς δὲ χρημάτων... δατητὰς Ἄρης,
ἀρὰν πατρώιαν τιθεὶς ἀλαθῆ. In S. Oed. R. 327 ἱκτήριος is used substantivally ;
so too ἀλιτήριος, e.g. Ar. Knights 445 f. éx τῶν ἀλιτηρίων od φημι γεγονέναι τῶν
τῆς deod,? Men. Epiir. 510 (574 Körte?) and elsewhere, a particularly good
parallel since ἀλιτήριος and λυμαντήριος express notions belonging to the same
sphere. The same applies to λαστήριος in the pseudo-Epicharmian lines
1 Protests were made simultaneously by P. Maas, Sokrates, iii, 1915, 237, and E. Petersen,
Die atl. Tragödie, 643. Kranz, Hermes, liv, 1919, 314 n. 1, objects to the changed order of
the lines purely on grounds of sense.
2 Hermann, after considering a lacuna before 1438, eventually decided in favour of a
lacuna after 1438. The words here quoted show that Hermann would have rejected the
punctuation κεῖται, γυναικὸς κτλ.
3 Whether βουλευτήριον ini Sept. 575 is masculine or neuter (as Wecklein, Tucker, Wilamo-
witz, Groeneboom take it; cf. also L-S, Addenda, p. 2057) is doubtful.
678
 COMMENTARY lines 1440-3

(Epich. fr. 44a, Diels-Kranz, Vorsokr. i, 5th ed., 205) on the Berlin ostrakon
of the third century B.c., on which Wilamowitz, Berl. Sitzgsb. 1918, 743,
remarks: ‘dieser Epicharm hat λαιστήριοι für λαισταί gesetzt”.
Χρυσηΐδων. Blomfield: ‘Pluralis numerus vim sententiae auget, sive
sarcasmus sit, seu laudatio. [Longin.] v. ὕψους 23. 3 [commenting on S.
Oed. R. 1406 ff.) πάντα γὰρ ταῦτα ἕν ὄνομά ἐστιν, Οἰδίπους, ἐπὶ δὲ θατέρου
᾿Ιοκάστη, GAN ὅμως χυθεὶς εἰς τὰ πληθυντικὰ ὁ ἀριθμὸς συνεπλήθυσε καὶ τὰς
ἀτυχίας" καὶ ὡς ἐκεῖνα πεπλεόνασται '' ἐξῆλθον "Exropés τε καὶ Σαρπηδόνες ” ”
Blomfield quotes further examples, cf. also Wilamowitz, on E. Her. 455. For
the use of the ‘general’ plural of persons’ names, or words indicating relation-
ship or class, to express contempt and indignation cf. P. Maas, Archiv. f. lat.
Lexikogr. xii, 1902, 499. Here the plural holds a peculiar venom; quite pos-
sibly Clytemnestra means to imply that Agamemnon had other amours of
the kind, such as Thersites casts in his teeth (B 226 ff.) πλεῖαί τοι χαλκοῦ
κλισίαι, πολλαὶ δὲ γυναῖκες εἰσὶν ἐνὶ κλισίηισ᾽ ἐξαίρετοι, à.
as TOL Axauoi πρωτίστωι

δίδομεν, εὖτ᾽ ἂν πτολίεθρον ἕλωμεν. ἷ

μείλιγμα undoubtediy in the sense οἱ lenimentum. In Eum. 886 the word
goes closely with θελκτήριον, and here clearly the notion is of θέλγειν (delenit
amore, Lucilius 1004) in the erotic sphere. The addition? to Hesiod’s Theogony
(206) which describes the τιμαί of Aphrodite (whom Bacchylides 5. 175 calls
θελξίμβροτος) names last φιλότητά τε μειλιχίην re. The word is awkward to
translate. The traditional deliciae (Stanley), ‘darling’ (Linwood, L-S), is
not happy; it would better render μέλημα. The latter word is of course at
home in the sermo amatorius, and I think it conceivable that Aeschylus had
it in mind when he substituted for it the more recherche μείλιγμα. The two
may be almost interchangeable as lovers’ endearments, but while (épo:)
μέλημα is that which is dear to me, μείλιγμα refers to the behaviour of the
lover who μειλίσσει or μειλίσσεται (Middle).
Clytemnestra exaggerates here, for the purpose of her self-justification,
the insult done to her by Agamemnon’s relations with Chryseis and Cas-
sandra. Actually this, as the prophetess declared (1260 f.), merely adds a
flavour to the draught of her revenge (cf. 1447). The words γυναικὸς τῆσδε
λυμαντήριος as spoken by her here appear to refer only to the wrong done her
as a wife, but perhaps the audience is intended to think also of what she
suffered as a mother.
1440-3. Before going into points of detail we must get the construction clear.
Weil: ' πιστὴ ξύνευνος, quod male cum antecedentibus coniungebatur, verbi
locum tenet’; so Wilamowitz: ' πιστὴ σύνευνος praedicati locum obtinet
(dormit mortua cum mortuo)'. It is an attractive suggestion, though in that
case in 1440 ἡ 8’ might seem more appropriate than 77’. But πιστὴ ξύνευνος
taken as equivalent to a verb makes it hard to fit on the following ναυτέλων
κτλ. (see below) and breaks the effective uniformity of the row of insults.
No decisive argument has been put forward against the traditional inter-
pretation giving a series of nominatives from ἥ 7’ αἰχμάλωτος ἦδε, all on the
same level. The word-order of the whole sentence beginning from 1438
! On this passage Aristotle made his interesting comment (fr. 144 R., preserved in Athen.
I3. 556 e) contrasting the alleged behaviour of Agamemnon with the attitude of Menelaus.
2 Cf. besides F. Jacoby's edition Wilamowitz, Glaube d. Hell. i. 95 n. 5, who agrees with
Jacoby in obelizing ll. 201-6.

679
lines 1440-3 COMMENTARY

κεῖται shows this arrangement : first the singular verb, then the first subject,
then a second subject connected by re. So, e.g., Eum. 15 f. μολόντα δ᾽ αὐτὸν
κάρτα τιμαλφεῖ λεὼς Δελῴός τε, χώρης τῆσδε πρυμνήτης ἄναξ, E. Med. 1125 f.
ὄλωλεν ἡ τύραννος ἀρτίως κόρη Κρέων θ᾽ ὁ φύσας φαρμάκων τῶν σῶν ὕπο. The
present passage is remarkable for the absence of the article in πιστὴ ξύνευνος
κτλ. continuing the series which began with the definite 7 7’ αἰχμάλωτος ἧδε.
One expects at first some such uniform series as S. El. 299 ff. σὺν δ᾽ ἐποτρύνει
πέλας ὁ κλεινὸς αὐτῆι ταῦτα νυμφίος παρών, ὅ πάντ᾽ ἄναλκις οὗτος, ἡ πᾶσα βλάβη,
ὁ σὺν γυναιξὶ τὰς μάχας ποιούμενος. But Aeschylus has an instructive parallel
for such a change in Sept. 571 ff. κακοῖσι βάζει πολλὰ Τυδέως βίαν, τὸν ἀνδρο-
φόντην, τὸν πόλεως ταράκτορα, μέγιστον Ἄργει τῶν κακῶν διδάσκαλον,; ᾿Ερινύος
κλητῆρα, πρόσπολον φόνου, κακῶν δ᾽2 Adpdorwı τῶνδε βουλεύτηριον. This pas-
sage, in which the article is used with the first two epithets of the invective,
might have been quoted by Karsten in support of his ἡ (for καὶ) κοινόλεκτρος,
accepted by Weil, Dindorf, Hense, Wecklein, Blaydes, Mazon, G. Thomson.
To me this emendation seems self-evident, restoring as it does the symmetry
of the whole series. The structure is clear (commas after τερασκόπος, θεσφατ.3
and ξύνευνος) ; in 1440 and 1441 the sense is strictly parallel but not the form,
since in the first line the two epithets are joined by καί, in the second the one
is attributive to the other (Headlam: 'the bond-servant and auguress, the
divining concubine’). In each member the paradox of Cassandra's situation
is summed up with merciless scorn: (1) seer of portents (for τερασκόπος cf.
Ῥ. 452) and at the same time captive slave; (2) Agamemnon's bedfellow who
utters prophecies (θεσφατηλόγος only here; for the form of the word cf.
Lobeck, Phrynichus, pp. 651 ff., Wackernagel, Dehnungsgesetz, xo f.). In the
second member the mockery is particularly vitriolic: the dedicated pro-
phetess of the god should give herself to no man.
This explanation of the sentence-structure makes superfluous any detailed
polemic against proposed deletions. Schütz originally wished to cut out καὶ
τερασκόπος.... τοῦδε (1440 f.), though later he withdrew the suggestion ; Nauck
and Wilamowitz, however, took it up;* Bothe bracketed 1441.
1442 f. The real difficulty of this passage lies in the words ναυτίλων δὲ σελμά-
Tov ἱστοτρίβης. For the ἱστοτρίβης" of the MSS (including the wretched codex
Romanus E) most editions give Pauw's ἰσοτριβής. The usual translation of
this is ‘die mit ihm auf des Schiffes Ruderbünken gelegen war’ (Nägelsbach),

ı Hermann thought of deleting this line (573, for a point of detail cf. above on 896), ‘ut-
pote qui factus sit ex v. 575’, just as in the series Ag. 1440 ff. critics have rejected one line
or two half-lines as *doublets! (v. infra). The two series give each other support, and the
liberal measure of vituperation, intelligible in itself, is thereby fully attested.
2 If Wilamowitz was right in adopting the δ᾽ of one group of MSS against the τ᾽ of the
remainder (including the Mediceus), this parallel with the series Ag. 1440 ff. is complete,
even down to the mode of connexion (ναυτίλων δὲ c. ior.). But this is a minor point.
3 The comma before θεσφατηλόγος (so F) was adopted by Victorius and many subsequent
editors.
4 Jachmann, Nachr. Gött. Ges., Phil.-hist. Kl., Fachgruppe 1, 1936, 134, does not hesitate
to add the passage to his instances of 'Binneninterpolation'. Wilamowitz's argument
'quoniam πιστὴ σύνευνος praedicati locum obtinet, concubitus antea non poterat com-
memorari' has been refuted above.
5 Wilamowitz's erroneous account of the accent in the MSS has been accepted by A. Y.
Campbell, Murray, G. Thomson. Mazon's note on the reading of G (Mazon's V) is incorrect :
vp is there, written in ligature.
680
 COMMENTARY lines 14421.
‘that beside him pressed the planks on ship-board' (L. Campbell), ‘die auf
dem Schiffe die Koje mit ihm teilte' (Wilamowitz), 'elle avait déjà partagé
son banc de mer' (Mazon), etc. Yet Hermann had rightly noted against such
an interpretation: 'poeta si illud in mente habuisset, non ἰσοτρίβης, sed
ὁμοτρίβης dixisset’." But even Hermann did not go far enough in his examina-
tion of the linguistic facts. Older Greek has a whole series of compounds with
ico-, but as far as I can see only quite exceptionally? any in which the second
element is unambiguously verbal. Aeschylus himself* has iedpyvpos, ἰσοδαί-
μων (also in Pindar, but with a different meaning), ἰσόθεος (as early as Homer,
and also Bacchylides, Soph., Eur., etc.) ἰσόμοιρος (Sophocles and elsewhere,
also in the laws of Gortyn ; Solon has ἰσομοιρία, and Homer once ἐἰσόμορος),
ἰσόνειρος, ἰσόπαις," ἰσόπρεσβυς, ἰσόρροπος (from ῥοπή, as L-S), which occurs
also in Eur. and in fifth-century prose, ἐσόψηφος (also Eur., Thuc., etc.),
ἰσόψυχος. Sophocles has, besides the words already noted, ἰσαμέριος, ἐσοθά-
varos, ἰσόσπριος (for ἰσοτέλεστος see below). Euripides has further ioddeAdos,
ἰσάνεμος (before him Bacchylides, zo (19). 9, not cited in L-S ; the supplement
of the last two syllables is certain), ἰσήρετμος, ἰσόνεκυς. In Pindar we find
further ἰσόδενδρος, ἰσώνυμος (earlier Ibycus, fr. 2. 3 D., ἐσοκέφαλος); 5 Herodotus
igoxparys, ἰσοπαλής (from maAn—also in Parmenides, Thucydides, and else-
where), ἰσόπεδος, ἰσόρροπος (v. supra), iomyopin, ἰσονομίη ; Democritus ἐσο-
σθενής ; Thucydides ἰσοδίαιτος, ἰσοκίνδυνος, ἰσόνομος7 (the Attic σκόλιον in which
the word occurs is in all probability? considerably older ; cf. ἰσονομίη in Hero-
dotus), ἰσοπλατής, ἰσοπληθής. All these multifarious instances show ἰσο-
compounded with a substantival element.? The same is true in Plato without
! Schütz, whose free paraphrase Hermann attacks, does not seem to have made the
mistake with which he is charged, since he says in his apparatus: 'v. o. (corp. nihil aliud
significat, quam quae aeque ac rex transtra nautica cubando trivit'.
2 A really exceptional case is perhaps ἐσοφόροι (o 373), ἃ ἅπαξ λεγόμενον in extant Greek
literature down to Xenophon, who picks it up (Symp. 2. 20). The Homeric ἐσοφαρίζειν, on
the other hand, is only an apparent exception, since it was artificially coined after the
model of ἀντιφερίζειν.
3 The following lists from various authors should not be regarded as a substitute for a
complete collection of the ἐσο- compounds in earlier Greek literature; they are primarily
meant to provide a background for the practice of the tragedians. 1 have not taken into
account the proper names since that would require full knowledge of the epigraphic evidence;
for a selection cf. Fick-Bechtel, Die griech. Personennamen, 130, and Bechtel, Die histor.
Personennamen des Griechischen, 227 f.
+ Since it is irrelevant for my present purpose to distinguish the various functions to
which the ἐσο- element may be turned, I have ignored this aspect here; it is, however, taken
into account in my remarks on 1470.
5 When on Sept. 784 κρεισσοτέκνων B. Todt, De Aesch. vocab. inventore (Progr. Halle 1855),
44, remarks ‘non potuit formationis certe causa reprehendi . . . nam si ἐσότεκνος laudatur,
cur κρεισσότεκνος non toleratur?' and Wilamowitz ad loc. ‘quamvis invitis concedendum
nobis esse videtur Aeschylum ad ἐσότεκνος sim. hoc κέβδηλον cudisse', they are inventing for
themselves an ἰσότεκνος for which there is no evidence anywhere.
© Unfortunately still labelled ‘dub.’ in L-S9, in spite of Wilamowitz, Textgeschichte der
Lyriker, 46 n. τ, and Pindaros 516, cf. also Rhein. Mus. lxxii, 1918, 178.
7 The present line of argument is not favourable to the assumption of R. Hirzel, Themis,
Dike, etc. 242 ff., that the word ἐσονομέα should be derived not from νόμος but from νέμειν.
Hirzel was followed by Busolt, Griech. Staatskunde, 418. 8 Cf. Hirzel, Themis, 248.
? The conjecture (σόφορτος (cf. L-S, p. 2079) by which Buschor attempts to make sense
of a curious word on the Laconian Arkesilas-cup in Paris of the first half of the sixth century
(Furtwängler-Reichhold, plate 151) is therefore linguistically correct; see, however,
Beazley, Hesperia, xii, 1943, 88.
681
lines 1442 f. COMMENTARY
exception: ἐσάριθμος,; ἰσόδρομος, ἰσομήκης, ἰσόπλευρος, ἰσοσκελής (for ἰσομέτ-
pros v. infra), ἰσογονία ; and also in Aristotle, who uses, besides the words he
has in common with Plato and others, ionpepwös, ἰσοβαρής, looywvıos, ἰσόκωλος,
ἰσοπαχής, ἰσοταχής, ἰσοτύραννος, ἰσοχειλής (used by Xenophon before him),
while characteristically in the treatise [Arist.] 7. κόσμου (written later than
Poseidonios, probably in early imperial times) 6, p. 440°28 there occurs the
phrase νόμος μὲν γὰρ ἡμῖν ἰσοκλινὴς 6 θεός, ‘the scales of God’s law swing in
perfect equilibrium’ (Wilamowitz, Lesebuch, Erläuterungen p. 132), where
ἰσοκλινής derives, of course, not from κλίνη, but from κλίνομαι. Here in fact is
exemplified the general rule that ioo- compounds admit a verbal element in
the Hellenistic age and not before (late Greek shows a continuation of the
practice). Nicander, Ther. 41, has iooeAkei from ἕλκειν (Schol. ἴσην ὁλκὴν
ἔχοντι, ἤτοι ioooraduwı), Nonnus ἰσοφανής, Timotheus of Gaza (turn of sth
and 6th centuries A.D.) and Paulus Silentiarius ἰσοτενής etc. Two words have
been omitted from this synopsis: ionpns (E. Iph. T. 1472), which is not
regarded as a compound at all,? since ‘-npns has gradually assumed the nature
of a suffix’ (Wackernagel, Dehnungsgesetz, 41, with a reference to Lobeck),
and S. Oed. C. 1220 ἰσοτέλεστος, said of θάνατος, ‘which has the same τέλος
(for all)'.3 This word is only in appearance derived from τελέω ; in function it
is 8 possessive compound, just like ἰσοτελής, 'having the same τέλος '. The
suffix element which makes ἰσοτέλεστος look like a verbal adjective fulfils
here only a stylistic purpose, just as in ἀμαρτύρητος, ἀκάρπιστος, καλλι-
πύργωτος, βαρυάλγητος and all the many compounds of this type in fifth-
century language, which are not to be thought of as verbal derivatives, but
as ornamental variants of dudprvpos, ἄκαρπος etc., as Wilamowitz showed on
E. Her. 290 (cf. also Kaibel on S. El. 186 and Pearson on Soph. fr. 249).
Sophocles presumably wished to avoid here the ordinary ἰσοτελής (which
probably by the time of the Oed. C. was in regular use as a political term), the
more so as he based his new compound on a different sense of τέλος. Plato's
ἰσομέτρητος I should regard in the same way, as simply a stylistic variant of
leóperpos.* The conclusion to be drawn is that since in ἰσοτριβής or ἰσοτρίβης
the second element cannot reasonably be derived from τριβή or τρίβος, such ἃ
compound cannot be ascribed to fifth-century speech, and therefore the con-
jecture is wrong.’
Is it possible then to keep the MS reading iororpißns? Stanley, who explains
‘slave born and bred in the house’, on the analogy of οἰκότριψ (not, as he says,
oikorpißns), is certainly on the wrong track. Heath’s suggestion has won
more approval: ‘ fororpiBys . . . proprie est 4j πρὸς τὸν ἱστὸν rpiBerail ?].
t To ascribe such an idea and mode of expression as ἐσάριθμα τῶι χρόνωι to Sappho
(fr. 48. 3 D.), as do Blass, Wilamowitz, and others, seems to me impossible; I believe it to
be a phrase of the letter-writer of late date to whom we owe the quotation. Lobel οὐδὲ
γράφει τὸν στέχον.
2 Cf. also Wilamowitz on E. Jon 1156 διχήρης : ‘the τήρης is simply a suffix'.
3 There is no need to go into the errors of commentators.
4 Even if we assumed that leoréAeoros should be directly derived from τελέω, such a
verbal adjective would be in no way parallel to a supposed ioorpıßys.
5 My note on iororpißns was long completed when Blaydes’s Addenda (p. 369 of his
edition) to Ag. 1443 caught my eye. He gives there a list of some ἐσο- compounds picked up
at random (putting words which do not occur until the Byzantine period side by side with
those of the sth cent. B.C.) and adds the correct observation: ‘in his tamen fere omnibus
nomen aliquod substantivum cum ἴσος compositum est’.
682
 COMMENTARY line 1443

Nauticum videtur fuisse convicium, cuius ratio, quod in talibus saepe accidit,
nobis hodie non satis est perspecta.' This has been adopted by Wellauer,
Peile, Linwood, Conington,? and others, and among later editors by Verrall,
who remarks on iororp. : ‘It is best to leave this, even if we cannot explain it.
We have not the knowledge of sailor's language in Aeschylus’ time which
would enable us to say what terms a woman like Clytemnestra might borrow
from it to apply to a woman like Cassandra, or what those terms might mean.’
But even if we cannot dismiss as absurd the idea of Clytemnestra's borrowing
ἃ coarse expression from sailors' language, we need stronger evidence than
the desperate attempts of modern scholars to puzzle out an unintelligible and
perhaps corrupt? word. Mention may be made, for the benefit of those who
enjoy such speculations, of Donaldson's idea, quoted by Paley, that 'this
proves that the captain's quarters were amidships in the ancient trireme’,
and of the fantastic comment of Plüss, who thinks of a pun on fords, which
means the beam of a loom as well as the mast.
Since ἱστοτρίβης is to me completely obscure, I have no comment to make
about the construction of ναυτίλων σελμάτων, or at least only the negative
one that it is impossible to follow Verrall in making it depend on πιστὴ
ξύνευνος. Heath says wrongly : 'subintelligi debet praepositio ἀπό, ad ναυτίλων
σελμάτων, quae nautica super transtra huc advecta est.’ Conington also is
arbitrary : ᾿σελμάτων ἱστοτρίβης is equivalent to τρίβουσα ἱστὸν καὶ σέλματα, in
the same way as ἄπεπλος φαρέων λεύκων stands for ἄμοιρος πέπλων kal φαρέων,
arid ἄχαλκος ἀσπίδων for ἄνευ χαλκείων ἀσπίδων '; but this familiar type of
adjective in which the second part of the compound is related in meaning to
the dependent genitive is in no way parallel to ἱστοτρίβης. Plüss is less
drastic : "The genitive σελμάτων denotes the sphere within which the attribute
holds good', but no use can be made of this guess so long as ἱστοτρίβης is
unintelligible. |
The adjectival use of vavriAwv is unobjectionable; Hermann, who altered
it, himself referred to 5. Phil. 220 and Eur. fr. 846. 2 N. ναυτίλωι πλάτηι.
So far as can be judged where the details are so uncertain, vaur. δὲ σελμ. ior.
is simply the last item in the series of abusive epithets, so that δέ gives no
trouble. Weil’s ye was the product of his interpretation of πιστὴ ξύνευνος as
predicate (see above) ; Wilamowitz takes πιστὴ ξύνευνος in the same way, and
translates ‘(sie) ruht treu an seiner Seite, wie sie auf dem Schiff die Koje mit
ihm teilte’, but nevertheless keeps δέ, which I cannot understand.
1443. ἄτιμα δ᾽ οὐκ ἐπραξάτην. The verb cannot here mean ‘do, carry out’
(‘poena vero eorum facinoribus haud defuit’ Heath ; so also Schütz and others;
Daube 198 n. 69: ‘sie haben nichts ‘‘unbezahlt” getan’). Wellauer is right in
comparing the passage with Sept. 337 BeArepa τῶνδε πράσσειν, 339 δυστυχῆ
1 The substance of this interpretation was anticipated by Casaubon: ‘ iororpißns:
publicum navis prostibulum ; melius cum gravi in fine’.
2 His rendering ‘ τρίβουσα ἱστόν ’ would be possible linguistically, cf. παιδοτρίβης, where
the short ı (an attempt at explanation was made by Ph. Buttmann, Griech. Sprachlehre,
i, 2nd ed., 37) according to Debrunner, Griech. Wortbildungslehre 49 (§ 99), is accounted for
by a false analogy with otxérpuf etc.
3 The ioro- element cannot be defended by arguing that we have ναυτίλων σελμάτων in the
context. It may have been precisely the mention of a ship in the preceding words which
helped towards a misreading of the original letters of the text; one knows how often this
kind of thing occurs (cf. on 1391, p. 655, n. 1.).
683
line 1443 COMMENTARY

πράσσει, and Peile translates the verb correctly: ‘nor have they fared amiss,
i.e. unworthily of themselves’. For the use οἱ πράσσειν in this sense (cf. on
1289) with a neuter adjective cf. besides the passages quoted above L-Ss.v. II,
Denniston on E. El. 1359.
ἄτιμα (cf. on 354): ‘without their due (reward or punishment) being given
them’.
1444. ὁ μὲν γὰρ οὕτως (scil. ἔπραξε) : οὕτως has been variously understood.
Triclinius : ἀπέθανεν ὡς ἤκουσας, SO, e.g., Schütz, Paley (‘for he died as I have
shown you’), Lewis Campbell, Platt. Wecklein, on the other hand, ‘ οὕτως
(ἔπραξε) pointing out the bath’, so Verrall, ‘For he lies as ye see’, and Plüss.
I believe the first of these interpretations, making οὕτως refer to what Clytem-
nestra said earlier, to be better than the other, which takes it as deictic,
with accompanying gesture. A glance through the comparable instances in
Aeschylus and Sophocles confirms the impression that where something per-
ceptible to the senses is directly indicated ὧδε is to be expected rather than
οὕτως. There is no difficulty in making οὕτως refer to something which does
not immediately precede it; it is appropriate for Clytemnestra here to hark
back to the account she gave soon.after her reappearance on the stage (1372):
as far as the content is concerned all she says down to the conclusion in 1447
is a single speech, amplified to include the retorts evoked by the remarks of
the Chorus.
ἡ δέ τοι, ‘roi videtur esse excitantis auditores ad eius quod nunc cum
maxime dicitur attentionem’ (Nägelsbach), cf. Denniston, Particles, 542,
where under the heading 'directing a person's eye or ear to a sight or sound'
this passage is quoted among others.
1444 f. This is the oldest testimony to the belief that swans sing shortly
before they die, cf. Wilamowitz! on E. Her. 110; the next? is Plato, Phaed.
84 e, where Socrates says τῶν κύκνων δοκῶ φαυλότερος ὑμῖν εἶναι τὴν μαντικήν
(this is important for the association of ideas in the Ag. passage), ot ἐπειδὰν
αἴσθωνται ὅτι δεῖ αὐτοὺς ἀποθανεῖν, ἄιδοντες καὶ ἐν τῶι πρόσθεν χρόνωι, τότε δὴ
πλεῖστα καὶ κάλλιστα ἄιδουσι κτλ. See further D’Arcy Thompson, A Glossary
of Greek Birds, and ed., 181 f.; Gossen, RE ii. A. 785. Although it is obvious
that Cassandra is here compared with the swan principally because the swan,
too, has foreknowledge of its end and proclaims it in song, we can assume that
the link was clearly perceived between the prophetess of Apollo and the
sacred bird which so often serves the god and glorifies him in song (Ar.
Birds 772).
The words ἡ δέ τοι κύκνου δίκην . . . γόον breathe a lovely, tender melancholy,
which for ἃ moment makes it seem that it is the poet himself who speaks and
not Cassandra’s enemy. But for Clytemnestra this moving touch is only a
preparation and a foil for the taunt with which she ends (1446 f.) and which
thereby sounds the harsher. Aeschylus has here achieved a double effect:
without impairing the consistency of the queen’s attitude he once more
Σ The etymology which he accepted for κύκνος (‘the singer’) and for ciconia is not sanc-
tioned by the more recent philologists, cf, Boisacq, Dictionn. étymol. s.v. κύκνος ; Chantraine,
La Formation des noms en Grec, 194; Walde-Hofmann, Lat. etymol. Wörterbuch, 212 f.
2 G. Thomson on Ag. 1445-8 manages to quote E. Her. 691 ff. and even E. El. 151 ff. as
evidence ‘for the swan’s death-song'. In Allen-Halliday's Homeric Hymns, and ed. (1936),
p. 412, Pratinas fr. 1. 5 figures as the oldest evidence though nothing is there said about
death.
684
 COMMENTARY lines 14461.
directs the sympathy of the audience towards the pathetic figure of the
prophetess. He thus rounds off the picture of a former scene: when the
cries of the stricken king burst forth (1343, 1345), nothing at all was heard of
Cassandra. This play is the tragedy of Agamemnon ; the εὐμαρὲς χείρωμα of
the seeress is not important enough to be called to mind at the moment of the
king's assassination. But Cassandra and her sorrows are not forgotten; a
soft and moving echo of her dying voice reaches us through the malicious
exultation of her murderess.
1446. Stanley referred to Ephoros (F Gr Hist 70 F 149 Jacoby) ap. Strabo
10. 4. 21 (end) about the Cretans: τὸν μὲν γὰρ ἐρώμενον καλοῦσι kAewóv, τὸν
δ᾽ ἐραστὴν φιλήτορα, and Hesychius φιλήτωρ' ἐραστής. On the odd gloss
in Tr ἡ ἐκ ψυχῆς φιλουμένη Blomfield justly comments: 'absurde derivat a
φίλος et ἦτορ᾽. It is a pity that Hermann defended this nonsense against
Blomfield.! ¢iAjrwp? in the tragedians as in Cretan everyday speech is of
course the agent noun of φιλεῖν. For φιλήτωρ feminine cf. Ernst Fraenkel,
Nomina agentis, ii. go, in the context indicated above on 664: the large
number of parallel cases there quoted support the retention of the -τωρ suffix
even where the agent is feminine. But it may be a really subtle touch that
a word which properly denotes the ἐραστής is here used of the woman.
Headlam's conjecture, ‘perhaps by the active word she wishes to imply that
the woman was the seducer' is worth considering.
1446 f. ἐμοὶ δ᾽ ἐπήγαγεν . .. χλιδῆς, In this difficult sentence one thing can
be determined at the outset: the subject of émjyayev. Stanley (in his post-
humous notes) made Agamemnon the subject; so Heath and others, e.g.
Hartung, Weil (1858), Schoemann, Opusc. iii. 153, Lewis Campbell, Wilamo-
witz. Wilamowitz originally (1885, and again in the notes at the end of his
translation) read avnp? in the place of εὐνῆς, clearly because he thought the
change of subject to be too harsh ; in his 1914 edition he substitutes αὐτήν.
Schütz, arguing against Heath, demands that the sentence beginning with
κεῖται shall be continued to the end of the speech so that Cassandra remains
the subject. Most editors have followed Schütz; Schneidewin is undecided
(‘the subject of ἐπήγ. is probably Ag. rather than Cass.’). It is strange that
there should have been any hesitation on this point, since everything in the
last lines suggests that from 1444 ἡ δέ to the end the subject is Cassandra.
The correspondence, pointed by the pronouns, between κεῖται φιλήτωρ τοῦδε
and ἐμοὶ δ᾽ ἐπήγαγεν κτλ. is enough by itself to exclude any thought of a
change of subject, moreover ὁ μὲν γὰρ οὕτως (1444) indicates that Clytemnestra
has nothing more to say about Agamemnon in this connexion.
The next point, equally certain, is that παροψώνημα τῆς ἐμῆς χλιδῆς must
be taken together, and that the genitive must on no account be altered, as it
has been by so many from Auratus down to Headlam, A. Y. Campbell,
ı Hermann's treatment of this passage in general is not happy and yet he has succeeded
in imposing upon L-S; even the latest edition follows him in coupling φιλήτωρ with peya-
λήτωρ, and in quoting (with Passow as intermediary) ‘the scholiast’ (what Hermann refers
to is merely an interlinear gloss in Tr) as evidence for φιλήτωρ τῶιδε. Vet it has long been
known that the dative τῶι ᾿ἀγαμέμνονι, on which H. based his arguments, is not to be found
in the MS but is a supplement inserted by Victorius.
2 F, but not G, has φιλήτως, a slight slip; Verrall champions it in the kindness of his
heart.
3 In Mazon's edition we find the note ' ἀνὴρ scripsi'.
685
lines σ446 1. COMMENTARY

Murray, G. Thomson. The interlinear gloss on παροψόνημα in F is right: τὴν

ἐκ περιουσίας τρυφήν ; Similarly Heath: ‘a παροψώνημα pendet τῆς ἐμῆς χλιδῆς,
quasi scripsisset poeta . . . ὀφώνημα παρὰ τὴν ἐμὴν χλιδήν᾽, and Conington: ‘in
παροψώνημα χλιδῆς the preposition exerts its force of addition’. Thus ‘a
savoury relish to my luxury’. The ém- in ἐπήγαγεν, clear in itself, has an
added point if we bear in mind that one says ὄψον ἐπί τινι, e.g. Ar. Peace 123.
This interpretation provides internal evidence for the genuineness of the
expression παροψώνημα τῆς ἐμὴς χλιδῆς ; external evidence, as old as could well
be desired, is afforded by S. Trach. 538, where Deianeira, continuing her
preceding simile (κόρην... παρεσδέδεγμαι, φόρτον ὥστε ναυτίλος), says ‘I have
taken my rival into the house λωβητὸν ἐμπόλημα τῆς ἐμῆς dpevds’. Anyone
who has studied in detail the manner in which Sophocles draws on Aeschylean
material can have no doubt that here the younger poet, dealing with a
similar situation, had in mind the crowning line of Clytemnestra’s speech:
not only is παροψώνημα echoed in ἐμπόλημα (arare word), but, more important,
the rhythmical character of the two lines in their entirety is very similar and
so are their endings; cf. on 943 and on 1293.
Finally εὐνῆς. There is no difficulty in the double genitive depending on
παροψώνημα, since there is nothing unusual in such a construction, especially
where the genitives stand on different levels as here (‘ εὐνῆς παροψώνημα,
additamentum adulterii, unam notionem efficit, unde pendeat τῆς ἐμῆς χλιδῆς᾽
Paley) ; perhaps the best commentary on the subject is that of Wilamowitz
on E. Her. 17o. The trouble lies in the meaning. εὐνῆς has had many de-
fenders, from Heath downwards. Conington translates ‘she . . . hath brought
to me a nuptial dainty-dish of new delight', and explains: 'The additional
zest given to Clytemnestra's pleasure by the fate of Cassandra lies in the
feeling that by her commerce with Aegisthus and murder of her victims she
has not only avenged her daughter but repaid her husband's infidelities: and
though this was not the motive cause of her adultery in the first instance, yet
it adds poignancy to the continued enjoyment.' Similarly, e.g., L-S s.v.
παροψώνημα ‘a new relish to the pleasures of my bed’. Against these attempts
to vindicate the text as it stands we should insist on two considerations which
have a bearing upon the appreciation of the scene as a whole. First: it is
difficult to imagine Clytemnestra speaking of the joys of her bed in a tone
such as Hamlet uses to his mother. Even in her taunts, even where she sets
herself above the dictates of custom and morality, she remains a queen.
She nowhere makes explicit mention of her own adulterous behaviour.
Secondly: we know from Clytemnestra's own lips in what consists the χλιδή
which she here ascribes to herself. The scene from the queen's entry (1372)
onwards must be taken as an organic whole. The words (1391 f.) χαίρουσαν
οὐδὲν ἦσσον à διοσδότωι γάνει σπορητὸς κάλυκος ἐν λοχεύμασιν are the dreadful
expression of her χλιδᾶν. The lust which wholly possesses the soul of this
daemonic woman at the great climax of her life is not sexual, but lust of
revenge—revenge for years the centre of her desires and now accomplished
in this treacherous deed.
My conclusion therefore is that those critics (Casaubon, Hermann, Wila-
mowitz, etc.) who held εὐνῆς to be corrupt were right. I do not know what
Aeschylus wrote. It is by no means certain whether we are justified in
looking for anything palaeographically similar, since εὐνῆς may have crept
686
 COMMENTARY line 1450
in as an undiscerning gloss on χλιδῆς and displaced the original word. Perhaps
this was an adjectival attribute to παροψώνημα; but a word of some quite
different kind is also conceivable.
1447. παροψώνημα. The kindred words which serve to illustrate this ἅπαξ
λεγόμενον are conveniently collected in Blomfield’s Glossarium. The verb
παροψωνεῖν occurs in Old Comedy. It is worth noting how Clytemnestra in
this fierce outburst allows her imagery to sink below the normal level of
tragic dignity.
1448 ff. τίς... μοῖρα: the wide separation (‘Sperrung’) is essentially no
artifice of ‘poetic’ or ‘rhetorical’ style, but on the contrary at home in every-
day speech—not in its ordinary conversational tones, but where it is coloured
by emotion, the ‘Affektsprache’ of modern philologists. Examples of such
separation of the interrogative adjective from its noun, taken at random,!
are: A. Pers. 438 καὶ tis γένοιτ᾽ ἂν τῆσδ᾽ ἔτ᾽ ἐχθίων τύχη ; (corresponding
almost exactly to Lysias 1o. 26 τίς yàp ἂν ἐμοὶ μείζων ταύτης γένοιτο συμφορά ;),
fr. 181 N. τίνος κατέκτας ἕνεκα παῖδ᾽ ἐμὸν βλάβης ; Hdt. 3. 34. 2 κοῖόν μέ τινα
νομίζουσι Πέρσαι εἶναι ἄνδρα ; 7. 21. 1 τί γὰρ οὐκ ἤγαγε ἐκ τῆς Aoins ἔθνος ἐπὶ
τὴν "Ελλάδα Ξέρξης ; S. El. 328 f. τίν᾽ αὖ σὺ τήνδε πρὸς θυρῶνος ἐξόδοις ἐλθοῦσα
φωνεῖς, ὦ κασιγνήτη, φάτιν ; Phil. 598 f. τίνος δ᾽ Ἀτρεῖδαι τοῦδ᾽ ἄγαν οὕτω χρόνωι
τοσῶιδ᾽ ἐπεστρέφοντο πράγματος χάριν ; Pl. Prot. 309 © καὶ τί ἂν γεγονὸς εἴη περὶ
σὲ κἀκεῖνον τοσοῦτον πρᾶγμα ; Cf. ON 1232.
For τίς àv . . . μόλοι expressing a wish cf. on 622.
1448. περιώδυνος : the available evidence suggests that the word may pos-
sibly have been borrowed from the language of physicians.
1449. δεμνιοτήρης : cf. on 53.
1450. ἐν ἡμῖν was recognized as impossible by Porson, who deleted ἐν, since
he did not greatly concern himself with metrical considerations in his dealings
with the lyrical parts.^ Many attempts have been made to keep the MS
reading, e.g. by Conington: ' ἐν ἡμῖν must be taken with φέρουσα, which has
thus the force of φέρουσα εἰς ἡμᾶς kai τιθεῖσα ἐν ἡμῖν, a combination which it is
needless to illustrate by examples’, and Sidgwick : φέρουσ᾽ ἐν ἡμῖν, "bringing
amongst us”: the use of ἐν... with verb of motion is epic’. But a glance at
the examples of this use of ἐν, e.g. in Stallbaum on Pl. Euihyd. 292 e, K. W.
Krüger, Griech. Sprachlehre (vol. i and ii), $68. 12, n. 2, Kühner-Gerth, i.
541, L-S s.v. ἐν A. I. 8, shows that it is limited to a narrow group of verbs
(such as τιθέναι, ἱστάναι, πίπτειν, and the like) or to ‘perfects of result’ (such as
ἐν τούτωι τῶι τόπωι καταπεφευγέναι, with nothing analogous to the φέρουσ᾽ ἐν
ἡμῖν assumed here by some critics. Hermann recognized the force of Porson’s
objection and wrote ἐφ᾽ ἡμῖν, which has won much approval (cf., e.g. Vahlen,
Opusc. acad. ii. 506). But in the cases where Aeschylus uses ἐπί and the dative
with verbs of motion (conveniently seen in Dindorf, Lex. Aesch. 125) a hostile
encounter is always indicated, and this holds for the only instance of ἐπιφέρειν
in Aeschylus (ἐποίσειν, Eum. 766). Any such idea would here spoil the sense,

1 One or two more will be found in my book Iktus und Akzent, 166 and (Latin examples)
164.
2 His countrymen seem to have felt the lack of information on the subject : G. Hermann’s
De metris poetarum Graec. et Rom. libri tres, published earlier than Porson’s Aeschylus, sold
extremely well in England, as the publisher Fleischer told Goethe in May 1800 (Goethe,
Tagebücher, ii. 292).
687
line 1450 COMMENTARY
since the emphasis is on the wish that death may come as a friendly deliverer
without pain or sickness. It is entirely wrong to cut out, together with the
dubious ev, the blameless and indeed indispensable ἡμῖν, as Wecklein and
others have done. When Headlam substitutes ὁμιλεῖν for ἐν ἡμῖν and then
nevertheless translates ‘would quickly come to bring me’ he is contradicting
himself. Emperius, Opusc. 133, appears to have thought of the right solution:
φέρουσ᾽ ἂν ἡμῖν. The addition of a second dv to the participle in a potential
clause is not uncommon. It is natural that the participle more often precedes
the main verb, but the reverse order also occurs, e.g. Ar. Eccl. 118 οὐκ ἂν
φθάνοις τὸ γένειον ἂν περιδουμένη. For the corruption cf. Pers. 706 ἀνθρώπεια
δ᾽ ἄν τοι πήματ᾽ ἂν τύχοι, where several MSS have ἐντύχοι.
1451. ἀτέλευτον is found only here, being a variant of the Homeric ἀτελεύ-
τητος (there is no verb τελεύω). It appears to be a case of 'dissimilatorischer
Silbenschwund’ (suppression of a syllable caused by dissimilation; for the
phenomenon in general see Schwyzer, Griech. Gramm. i. 262 ff.), as, e.g., Sept.
711 δατήριοι (cf. above, pp. 673-4 on 1429, also p. 678 on 1438).
1450 f. Cf. hymn. Orph. 87. 5 (of Θάνατος) τὸν μακρὸν ζώοισι φέρων αἰώνιον
UTTVOV.

1452. φύλακος, W. Schulze might have quoted this passage when he

referred (Berl. Sitzgsb. 1918, 493 = Kl. Schr. 172) to the habit among Indian,
Germanic, and other peoples of regarding the king as watcher and protector
(of peace), as φύλαξ ; Schulze notes ‘so dvaé, too, in the original meaning of the
word, is the equivalent of φύλαξ '. The Romans shared this conception, which
won a new lease of life in the time of Augustus, notably in the lyrics of
Horace, where it was enriched by the admixture of the Hellenistic notion of
σωτήρ (cf. Oxé, Wiener Stud. xlviii, 1930, 40f.; the account in Norden,
Aus altröm. Priesterbüchern, 156 f., is inadequate).
Many editors seem to accept xai, but it is not easy to discover how they
take it, since they either leave it uncommented, or else translate as if καί had
been removed from the text, e.g. Nägelsbach (‘nachdem gemordet ist der
Hirte, der mildgesinnte, der so viel geduldet um ein Weib’), see also Wilamo-
witz and Verrall (‘now that our kindest protector is laid low, who, having
much endured for a woman's sin, hath by a woman lost his life’). καί is in fact
here intolerable, whether one takes it as linking τλάντος with δαμέντος or with
φύλακος εὐμενεστάτον. The sense requires that the phrase πολλὰ τλάντος
γυναικὸς διαί should be subordinate to δαμέντος φύλακος εὐμενεστάτον, not
co-ordinate with it. Franz was therefore right in deleting καί (for the anti-
strophe see on 1472), and many editors have followed him. The insertion of
καί between the apparently coordinate genitives is precisely the same kind of
error as that discussed on 124.!
1453. wodéa (cf. on 723), suggested by C. G. Haupt and others, was accepted
by many (cf. also P. Maas, ‘Epidaurische Hymnen’, Schriften d. Kónigsb.
Gel. Ges., 9. Jahr, Geisteswiss. Kl., 1933, 145). The emendation is easy, and
necessary ; the antistrophe has κόρακος in 1473. The correspondence of a long
to two short syllables is indeed found in Aeschylean lyric iambics—for these
are iambics*—but, as was shown p. 351, not in the Orestera.
1 Cf, also Cobet, Variae Lectiones, 2nd ed., 191: in the passage Pl. Apol, 21 e restored by
him, a meaningless καί had crept into the text between two present participles.
2 Wilamowitz erroneously takes 1452-4 as trochaic (O. Schroeder, Aeschyli cantica, gives
688
 COMMENTARY line 1455

γυναικὸς διαί: ‘by the deed (and in consequence of the deed) of a woman’,
cf. on 448. That so much suffering should be caused and visited upon the
king, the protector of his people, through a woman, Helen, is regarded by the
old men as ignominious, like his death at the hand of a woman.
1453 f., in expression and more particularly in the way the two daughters of
Tyndareus are coupled together as of like baleful effect, is clearly modelled
on the passage of the Nekyia À 436 ff., which is further echoed in 1455 ff.: ὦ
πόποι, ἦ μάλα δὴ γόνον Arpeos εὐρύοπα Ζεὺς ἐκπάγλως ἔχθαιρε γυναικείας διὰ
βουλὰς ἐξ ἀρχῆς" ᾿Ελένης μὲν ἀπωλόμεθ᾽ εἵνεκα πολλοί, σοὶ δὲ Κλυταιμήστρη δόλον
ἤρτυς τηλόθ᾽ ἐόντι. Cf. Hesiod fr. 93 Rz.
1454. The final clause is made independent, cf. Schneidewin: ‘instead of
πρὸς γυναικὸς δ᾽ ἀποφθιμένου the finite verb gives a more vigorous turn to the
end of the sentence.’
1455ff. Here once more, as in the parodos and the three stasima, we are
made to remember the ἀρχαί and αἰτίαι of the present calamities, though the
reference is only brief; here once more, as in the first and second stasima,
there appears, heralded by r453,' the shape of Helen, this time only for a
moment and only in her dreadful aspect.
1455. Editors who note here: ‘ à semel codd.: bis Blomfield’ do less than
justice to Blomfield’s circumspection. He says: ‘Notent tirones ἰὼ extra
metrum esse; nisi legendum ἰώ, is. The reference to 1489 and 1513 is
irrelevant, since those ephymnia are not in responsion to this. The dropping
out of a second iw is possible but cannot be proved.
mapávous. The large majority of editors have fortunately adopted Her-
mann's emendation (there are alternatives, some of them fantastic enough,
in, e.g., Keck, Wecklein, Housman, J. Phil. xvi, 1888, 281). For the corrup-
tion cf. Ar. Clouds 148o, where R!V have παρανομήσαντος instead of παρανοή-
σαντος. To any reader of Aeschylus this line in its emended form recalls the
passage where there occurs, not indeed the word mapdvous, but its derivative,
Sept. 756 f. παράνοια συνᾶγε νυμφίους φρενώλης. The two passages have this
in common, that both imply the idea of an aberration of the mind, a deviation
from its sound and natural course (cf. also Eum. 330 παραφορὰ φρενοδαλής)
regarded as the root (cf. Ag. 223 τάλαινα παρακοπὰ πρωτοπήμων) of a series of
horrible events. παράνοια is closely related to one aspect of ἄτη, cf., e.g., Pers.
98 f. παράγει βροτὸν εἰς ἄρκνας ἄτα.
μία τὰς πολλάς : an example of the old type of word-order described on
320. This particular juxtaposition of ‘one’ and ‘many’ is frequent because of
the sharpness of the antithesis; it is one of the quasi-rhetorical effects, many
of them pre-rhetorical, sought in elevated style. Cf. Pers. 327 f. εἷς ἀνὴρ

a correct metrical analysis). All through this kommos there are many iambic kola, but
there is not one that can with certainty be called trochaic. The scansion of 1547 in Wilamo-
witz’s edition (where incidentally ‘2 troch.’ is a slip for ‘troch.”) is corrected in his Verskunst,
292. A series of 5 cretics such as is left when καὶ is taken out is unusual in these iambics,
though 3 successive cretics are frequently found. Ag. 437 f. = 456 f. gives 4 cretics in
succession, and in 1453 = 1473 the initial o v ὦ — has the effect of preventing monotony.
1 The connexion between 1453 and 1455 ff. is very close: the preliminary brief mention
is elaborated in the passionate apostrophe. I fail to see how Wilamowitz, Interpr. 198,
could make this comment: ‘There is no transition to the thoughts which begin with the
anapaestic piece (1455 ff.); it is impossible that the same person who delivered the strophe
also went on to the anapaests.” For the latter point cf. p. 661.
4872-3 Ρ 689
 line 1455 COMMENTARY
πλεῖστον πόνον ἐχθροῖς παρασχών, Sept. 6 f. ᾽Ετεοκλέης ἂν εἷς πολὺς κατὰ πτόλιν
ὑμνοῖτο, S. Trach. 459 f. χἀτέρας πλείστας ἀνὴρ els ᾿Ηρακλῆς ἔγημε, Oed. C.
563 f. εἷς πλεῖστ᾽ ἀνὴρ... ἤθλησα κινδυνεύματα, E. Heraclid. 7 f. πόνων πλείστων
μετέσχον εἷς ἀνὴρ ᾿Ηρακλέει, 327 f. ἕνα γὰρ ἐν πολλοῖς ἴσως εὗροις ἄν, Thuc. τ.
83. 1 πολλοὺς μιᾶι πόλει... ἐπελθεῖν, 2. 35. 1 ἐν ἑνὶ ἀνδρὶ πολλῶν ἀρετὰς κινδυ-
νεύεσθαι, 8. 68. 1 πλεῖστα εἷς ἀνὴρ... δυνάμενος ὠφελεῖν, Anon. Iamblichi 7. 15
(Diels-Kranz, Vorsokr. ii, 5th ed., 404) εἰ μέλλει συλήσειν ταῦτα παρὰ τοῦ
πλήθους τῶν ἀνθρώπων εἷς ὧν παρὰ πολλῶν, Pl. Rep. 2. 374 a ἀδύνατον ἕνα πολλὰς
καλῶς ἐργάζεσθαι τέχνας, Elogium of L. Scipio, CIL 1.2 9. 1 f. honc oino ploirume
cosentiont Romane duonoro optumo fuise viro.
In E. El. 1026 ἔκτεινε πολλῶν μίαν ὕπερ (Agamemnon's sacrifice of Iphi-
geneia) the antithesis of the present passage is reversed.
1455 ff. τὰς πολλὰς... ψυχὰς ὀλέσασ᾽ ὑπὸ Τροίαι: the clause as a whole
is again formed on the words of Odysseus cited on 1453f. (cf. also £
68Í. ὡς ὥφελλ᾽ 'EMvqs ἀπὸ φῦλον ὀλέσθαι πρόχνυ, ἐπεὶ πολλῶν ἀνδρῶν ὑπὸ
γούνατ᾽ ἔλυσε), while the words πολλὰς ψυχὰς recall the beginning of the
Iliad.
1458ff. On this passage conjectures have taken every possible direction,
partly because the text has been misunderstood, partly because of the super-
stition (it can be traced back to A. Seidler, De versib. dochm., 1811, 408 n., and
G. Hermann) that 1455-61 corresponds to 1537-50, a responsion which can
only be forced by assuming a large lacuna and making the most violent
changes in the text.” The beginning of the lyrics which follow the anapaests
has been restored by Wilamowitz with as little alteration and as much
probability as can be expected in this difficult section, where no responsion
. comes to our aid and metrical alternatives are conceivable.? Wilamowitz
deletes δὲ and regards τελείαν as an instance of the common* misspelling for
τελέαν. On the construction he observes: “τελέαν ἐπανθίζειν [he probably
meant ἐπανθίζεσθαι] est τῶι τὴν στεφάνωσιν τελειοῦντι στεφάνωι στεφανοῦσθαι᾽᾽.
It is no refutation to say, as G. Thomson does without taking into account
Wilamowitz’s grammatical comment, ‘this seems impossible’. For the
feminine reAdav used without a noun Wilamowitz refers to his detailed ex-
position on E. Her. 681 (cf. above on 219 and on 916) ; Cho. 640 διανταίαν is a
particularly good parallel.
1 A different meaning is conveyed by the stereotyped juxtaposition in the’ following
instances: Demosth. 21. 96 τὸ τῶν πολλῶν εἷς εἶναι, Philippides com. fr. 18. 4 (111 306 Kock)
εἶναι δ᾽ ὑπόλαβε καὶ σὲ τῶν πολλῶν ἕνα, epigram on Theocritus (Scholia, Proleg. G, p. 6
Wendel (= Anthol. Pal. ix, 434) εἷς ἀπὸ τῶν πολλῶν εἰμὶ Συρακοσίων, Cic. Brut. 274 qui non
Juit orator unus e multis, Off. 1. 109 ut unus de multis esse videatur, Mor. Sat. 1. 9. 71 f. unus
mullorum and the like (others in A. Otto, Sprichwörter der Römer, 358). 1t is a subtle touch
in Horace's Hypermestra ode (3. 11. 33) that with the words una de multis he begins the
praise of the woman who, so far from being una de multis in the ordinary sense of the
phrase, is just the opposite, μονόψαφον ἐν κολεῶι κατασχοῖσα ξίφος.
2 This method was justly attacked by Blomfield, and by others after him (cf. R. Arnoldt,
Der Chor im Ag. des Aesch. 76 f.), but the prejudice would not die. Even F. Blass (Mélanges
Henri Weil, 1898, 15) substituted τρίδματος for ἐρίδματος (though “Epis ἐρίδματος has the
unmistakable Aeschylean ring) in 1461 solely in order to obtain exact syllabic corre-
spondence with 1550.
3 The same conditions confront us also in what follows down to the end of the ephymnium
(1461). Full certainty is unattainable here.
4 Cf., eg., Sept. 850, In Pindar fr. 122 (Bergk, Schroeder) I. 15 our source, Athen. 13. 573 f,
gives τελείαις, which Boeckh corrected to τελέαις,
690
 COMMENTARY lines 1460f.
πολύμναστον, passive, cf. on 821. For the extension of two syllables beyond
the end of the dochmiac cf., e.g., Sept. 98 f. and 219, in general cf. on 1173.
ἐπηνθίσω : the active ἐπανθίζειν, also in a metaphorical sense and also with
reference to something baleful, Sepi. 951 and Cho. 150 (Schol. στέφειν ὡς
avOeot).
1460. δι᾽ αἷμ᾽ ἄνιπτον. Headlam translates: ‘but now in blood unpurgeable
thou hast crowned . . .', Wilamowitz (Interpr. 198) : ‘jetzt hast du dich mit der
letzten blutigen Blüte gekrönt’. This is a mistake in syntax:' διά with the
accusative could never have this sense, which is just possible for διά with the
genitive although real parallels for such a use of δι αἵματος are lacking:
the meaning of it in Demades' well-known remark in Plutarch's Solon (17. 3) is
completely different, δι᾽ αἵματος, où διὰ μέλανος, τοὺς νόμους ὁ Δράκων ἔγραψεν.
The expression δι᾽ αἷμα would mean ‘because of, as a result of, blood’. So
Schütz, ‘ob inexpiabilem sanguinem (scil. mactatae Iphigeniae)', Conington:
‘owing to the blood yet uncleansed (that of Iphigenia, though probably with
a reference to the children of Thyestes . . .)' and Plüss: ‘ dı@: in consequence
of the offering of the child, which was made for Helen’. This does not fit
the thought : the context shows, particularly after the opening νῦν, that what
is called here the uncleansable deed of blood is the murder of Agamemnon
which has just happened. Moreover, after τελέαν πολύμναστον ἐπηνθίσω an
explanation is needed in accordance with the general habit of Aeschylus,
who nearly always follows a ypt$os with its solution (cf. on 7, p. 9). Verrall's
construction πολύμναστον & alu’ ἄνιπτον is intolerable. Wellauer (followed by
Hermann and others) was probably right in cutting out 41 as dittography for
AI. Wilamowitz's metrical analysis remains substantially unaltered (1460 f.
pure iambics). I find ἄνιπτον αἷμ᾽ (ie. v —o— τῶ τον που -- rhythmically
more attractive, but that may be purely subjective; and in view of the
general uncertainty of this passage it is better to refrain from the trans-
position.
1460 f. A ris . . . oifüs. Whatever be the text and interpretation of these
final words in detail, they, like the preceding words, must refer to Helen, at
least so far as to imply, even if "Epıs does not directly signify Helen, a close
relation between the two.? To limit "Epis, as Schütz does, to the 'rixa inter
Agamemnonem et Clytaemnestram' destroys the unity of the ephymnium ;
Clytemnestra’s answer (1464-7), too, shows that she herself has not been
accused immediately before.
ἥτις ἦν gives no possible connexion, as was realized by Heath; Schiitz,
following him, wrote 7j τις, which most editors have accepted. Headlam took
Karsten's εἴ ris ποτ᾽ and read ef τις ἦν ποτ᾽, ἐν δόμοις ἔρις eptdparos: ‘if e'er
there was one, a dissension firmly planted in a house’. That is wrong, first
because of the manner in which ἦν is torn from ἐν δόμοις, secondly because of
the form given to the idiomatic expression εἴ τίς ποτε, which Headlam finds
particularly suitable here.3 This conditional clause is common in the elliptical
1 The same mistake was committed by many commentators on Cho. 474 δι’ ὠμὰν ἔριν
αἱματηράν, where Blass deprecated the misunderstanding, cf. above p. 334 on 698.
2 On this point Wilamowitz in his edition of Aeschylus (and Interpr. 198) has considerably
improved upon his former view, followed in his translation: he then read in 1459 σὺ δὲ
τελέαν and referred to Clytemnestra the whole from there to the end of the ephymnium.
3 G. Thomson has cut out the whole of Headlam’s comment and substituted one of his
own. What Headlam said was: ‘This use of εἴ τις is not so well recognised as it should be’,
691
lines 1460 f. COMMENTARY
form εἴ τίς ποτε, but not with the addition of a verb. Headlam’s text has the
further disadvantage that it loses the 71s . . . "Epis, which fits the context so
exceedingly well. In so far as a reasoned judgement is possible here at all
(see above), 7 ris appears to be right. Although affirmative # goes ‘mostly
with adjectives and adverbs’ (Denniston, Particles, 280), it goes with a verb
in Ag. 1064 4 patverai ye. What is strengthened here by ἦ is not τις, which,
in the usual way, has come forward to the second place in the sentence, but
the verb (ἦν here is not ‘verbum substantivum’ but has its full strength ‘was
there’) or the whole sentence.
τότε should never have been tampered with; it points, like the mention of
Helen itself, to the past, here to the time which lay before the beginning of
the war.
1461. We cannot say in what sense Aeschylus used the unique ἐρίδματος (cf.
Schuursma 32 f.). On the prefix ἐρι- cf. Debrunner, Griech. Wortbildungslehre,
31; Schwyzer, Griech. Gramm. i. 434. 6. Dindorf translates firmiter fundatus,
so also L-S: ‘strongly built, i.e. immovable, unconquerable’. This is morpho-
logically correct, presupposing as it does derivation from the root δεμ- (in
δέμω, δέμας, δόμος etc.); it is supported by the analogy of the Homeric
(0 519) θεοδμήτων ἐπὶ πύργων (Pindar, Bacchylides, Sophocles, Euripides use
the word in the same sense), moreover by ἐύδμητος (Homeric) and νεόδματος
(Pindar). For the possible meaning of -ὅματος cf. Wilamowitz’s remarks (in
his commentary on Cho. 617) as to the boldness of Aeschylus in describing a
ὅρμος as xpucedduntos, ‘goldgebaut’. An alternative possibility is that
Aeschylus meant ‘strongly subduing’ (‘sehr bándigend' Passow), formed from
Sap- like ἄδμητος (cf. also E. Med. 623 τῆς νεοδμήτου κόρης); an active meaning
is always possible with verbal adjectives in -ros (cf. on 12). But whatever was
the particular meaning adopted by Aeschylus, it is clear that he had in mind
an etymology of "Epis, as shortly afterwards (1485) an etymology of Zeus.
The connexion of the last sentence with what has gone before (always on
the assumption that the text adopted is substantially correct, however
uncertain in details) is loose without being obscure. “Truly there was then
an Eris in the house.’ Helen is not directly equated with Eris (as in 749
she is with Erinys) nor is Eris said to have taken on the shape of Helen, as
Clytemnestra later (1500) maintains that the avenging spirit appeared dis-
guised as Agamemnon’s wife. But the final assertion ἦ τις 7v . . . "ἔρις, after
what has gone before, makes it clear that the power of this undefined Eris
and the power of Helen are closely connected. The audience are left to fill in
the details of the picture: they may or may not connect this Eris with the
Eris made responsible, in an earlier song (698), for the flight of Helen and her
being pursued by the Greeks. ἀνδρὸς οἰζύς also is left purposely vague. The
mention of Helen suggests Menelaus but, because of the whole context here,
Agamemnon may be thought of as well (on this cf. Wilamowitz, Interpr. 212).
Both men have to suffer γυναικείας διὰ βουλάς. This twofold implication of

etc. But Karsten had already called attention to Lobeck on S. 47. 885 (where special
attention is paid to εἴ wore and the like) and Matthia, Griech. Gramm. 1474, g (‘Thus εἴ τις,
εἴ που etc. can stand alone elliptically without a verb’, etc.) ; cf. also Kühner-Gerth, ii. 573,
Classen-Steup on Thuc. 4. 20. 1. In Kühner-Gerth the insertion of the verb in Demosth.
24. 4 ἐγὼ 8’, εἴπερ τινὶ τοῦτο Kal ἄλλωι προσηκόντως εἴρηται, νομίζω κἀμοὶ viv ἁρμόττειν εἰπεῖν
is noticed as a rarity in comparison with the numerous instances of ellipse.
692
 COMMENTARY line 1467
ἀνδρός is perhaps also suggested by the Chorus’ mention (1468 f.) of the
δίφυιοι Τανταλίδαι.
The end of the ephymnium (not only the last word) seems to have been in
Euripides’ mind! when he made the Chorus of Trojan captives sing of Helen’s
marriage (Hec. 946 ff.): ἐπεί με γᾶς ἐκ πατρώιας ἀπώλεσεν ἐξώικισέν τ᾽ οἴκων
γάμος, οὐ γάμος ἀλλ᾽ ἀλάστορός τις οἰζύς (for the idea cf. Andr. 103 f. ᾿Ιλίωι
αἰπεινᾶι Πάρις οὐ γάμον ἀλλά τιν᾽ ἄταν ἠγάγετ᾽ εὐναίαν εἰς θαλάμους ᾿Ελένων).
1465. ävôpohérapa: also Sept. 314 (as an adjective). The word shows that
the vocabulary of Aeschylus in its bolder coinages is sometimes indebted to
the language of post-Homeric Epic—a fact for which there is every now and
then direct evidence, though far more often we have to infer it from what we
find in later poets. In an Epic fragment preserved in Schol. Pind. Nem. 3.
64 Ὁ (= Hesiod fr. [dub.] 278 Rz.) the Amazon Melanippe is called dvöpo-
Aéreipa.? The lines, which ‘Herakles spoke in some Epos’ (Wilamowitz on
E. Her. 408), are not dated; but, even if they are later than Pindar and
Aeschylus, the expression ἀνδρολέτειραν ἀμώμητον Μελανίππην is more likely
to be derived from Epic than from Tragedy. That avöpoAereıpa was originally
coined for an Amazon is extremely probable (cf. the Homeric Ἀμαζόνες
ἀντιάνειραι and the Scythian name for the Amazons, which according to
Hdt. 4. 110. 1 means davöpoxrövoı).
1467. ἀξύστατον ἄλγος. The interpretation of Schütz has been accepted by
most commentators, with minor variations in details: ‘dolorem, qui haud
facile sisti potest, insanabile malum’; it is taken over word for word by
Dindorf Lex. Aesch., so also L-S: ‘that cannot be composed, incurable’.
Conington, while substantially in agreement with this rendering, introduced
into it the image of the wound, quoting Horace Est. 1. 3. 32 nequiquam coit
εἰ rescinditur. This embellishment became popular; see, e.g., Weil (1858)
‘ ἀξύστ. ἄλγος videtur esse malum insanabile, quasi volnus quod non coit' ;
Wecklein ‘a painful wound, which does not close, does not heal' ; Mazon ‘une
blessure qui ne se ferme pas’. Headlam explains the phrase as volnus in-
compositum and translates ‘unhealable’ ; then G. Thomson compared Hippo-
crates Aphor. 3. 17, where, however, ra σώματα ξυνιστᾶσι (describing the
effect of the North wind on bodies) means something quite different (‘con-
densent les corps' Littré) and is no evidence for συνιστάναι — to close a
wound. There is in fact no reason at all to look for this metaphor here: ἄλγος
does not mean ‘wound’. Klausen's incomparabilis (taken up by Verrall) and
other arbitrary interpretations need not be considered. Aeschylus uses
συνίσταμαι (the active does not occur in what we have of him) in a single
sense (Sept. 435, 509, 672), which is also common in other authors (L-S s.v.
B. II. 2), ‘meet in fight, be engaged with’. Therefore I regard ἀξύστατον ἄλγος
as a rarer equivalent of ἄμαχον ἄλγος and find my view confirmed by the fact
that in the parable of the lion cub, Ag. 733, the damage caused by the
animal is called ἄμαχον ἄλγος οἰκέταις. That is a simile for the effect brought
about by Helen; here it is denied that Helen ἀξύστατον ἄλγος ἔπραξεν. Cf.
also Cho. 692 ὦ δυσπάλαιστε τῶνδε δωμάτων ἀρά.
Clytemnestra’s reply 1462-7 answers the two ideas of the preceding stanza,
t I do not know if attention has already been called to the fact that in the same ode,
Hec. 930, the war-cry d παῖδες ᾿Ελλάνων echoes A. Pers. 402.
2 This instance, cited in Dindorf's Thesaurus, is omitted in L-S.

693
line 1467 COMMENTARY

the desire for immediate death and the outburst against Helen. Some of the
phrases of the Chorus are taken up word for word: μοῖρα (1451: 1462), μία τὰς
πολλὰς ψυχὰς ὀλέσασ᾽ ὑπὸ Τροίαι: μία πολλῶν ἀνδρῶν ψυχὰς Δαναῶν ὀλέσασα. On
this cross-connexion (cf. above on 1137 and 1412 f.) see especially W. Kranz,
Hermes, liv, 1919, 315: there and in his book Stasimon, 22 f., he explains
this as a survival from the epirrhematic composition of drama in its
early stages."
The queen’s answer contains nothing provocative or scornful, nothing
sharp at all; her words sound more like a request than an order, and are far
milder than anything she has said since her reappearance (1372). The xara-
στροφή in Clytemnestra’s own soul has already begun (cf. on 1437).
1468. ἐμπίπτεις : for the meaning cf. on 1175. It is, of course, possible that
Aeschylus wrote ἐμπίτνεις (cf., e.g., the discrepancy of the MSS in S. 47. 58,
where the metre settles the issue), but the fact that in the strophe (1448) the
dochmiac has the form — vu — v — is no sufficient reason for altering the MS
reading, cf. on 1128.
1468 f. διφυεῖσι is ἃ false form, moreover the metre requires a long second
syllable. Hermann's διφυίοισι (first published in his appendix to Humboldt's
translation, and then Elem. doctr. meir. 704) is clearly right. Lobeck, Phry-
nichus, 494, discussed σῶμα δίφυιον in the hexameters? of the early Hellenistic
poet Antagoras preserved in Diog. Laert. 4. 26 and compared Callimachus
fr. 162 Schn. δεκάφυια and the evidence of the lexicographers for δίφυιον,
τρίφυιον, etc. Since then Elean inscriptions have produced examples of
δίφυιον (Schwyzer, Exempla, 411. 5 and 419. 8) and ζίφυιον (op. cit. 409. 6—
this earliest example is missing in L-S ; cf. the restoration 410. x). “In place
of διφυής which is used in Ionic and Attic δέφυιος appears in Elis’ says Bechtel,
Griech. Dial. ii. 846, who quotes the literary evidence mentioned above and
concludes : ‘the home of the form cannot be determined.’ It is impossible to
decide in what sense Aeschylus used the word. The Aeschylus lexica (Wellauer,
Linwood, Dindorf), as well as Passow and L-S, regard it as a mere synonym
for ‘two’, so also the majority of the translators. This may be right : ‘double,
twofold' corresponds to the use of the word elsewhere. But the attractive
conjecture of Plüss should also be considered: ‘ διφυίοισι: cf. δύο λήμασι
δισσοὺς Arpeidas 122; in spite of their difference a like fate!', and, clearly
independently, Wilamowitz: 'videtur Aeschylus in voce peregrina διφυεῖς,
λήμασι δισσοὺς, 122, sensisse'. It is certainly right to understand the δίφνιοι
Τανταλίδαι as Agamemnon and Menelaus, as is suggested by the context of
the whole passage. The reference to the two sons of Atreus here corresponds
to the description of their two wives in 1453 f. and 147o. As an alternative to
this interpretation the scholiast (XyoA. wad. in Tr) gives a different one,
Atreus and Thyestes ; some commentators have adopted this, others make a
1 To supplement Kranz's observations attention may be called to Prom. 128 ff., where the
Chorus of Oceanides answers the anapaests of Prometheus and each time picks up and
echoes the words of his preceding utterance. 127 (Prom.) πᾶν μοι φοβερὸν τὸ προσέρπον rw 128
μηδὲν φοβηθῆις, 140 (Prom.) δέρχθητ᾽ ἐσίδεσθε = 144 λεύσσω Προμηθεῦ, 159 (Prom.) ἐχθροῖς
ἐπίχαρτα πέπονθα ~ 160 f. τίς ὧδε τλησικάρδιος θεῶν ὅτωι τάδ᾽ ἐπιχαρῆ ; 175 f. (Prom.) πρὶν
ἂν ἐξ ἀγρίων δεσμῶν χαλάσηι (sc. Zeus) ~ 178 ff. σὺ μὲν θρασύς τε καὶ πικραῖς δύαισιν οὐδὲν
ἐπιχαλᾶις (where a special point lies in the charge of relentlessness thus applied to Pro-
metheus himself).
2 On these cf. Wilamowitz, Antigonos von Karystos, 69 f.; Hellenist. Dichtung, ii. 2 n. 2.

694
 COMMENTARY lines 1470 f.
compromise, e.g. Lewis Campbell (note to his prose translation) : 'this phrase
recalls the pairs of brothers—the rivalry of Atreus and Thyestes, the conjoint
sovereignty of Agamemnon and Menelaus.' Wecklein is correct: ' δώμασι καὶ
dup. Τανταλίδαισιν (Agamemnon and Menelaus), a ἕν διὰ δυοῖν '. δώματα (cf. on
377): the house itself as well as the community and life of those who belong
toit. All through this play special emphasis is placed on the fact that the two
brothers live together in the same house; cf. on 400. Whether the use of the
name Τανταλίδαι is intended to recall the idea of a family curse going back to
the founder of the family (as Kranz, Hermes, liv, 1919, 315 n. 1, and several
commentators assume) is doubtful: Aeschylus nowhere in the Oresteia goes
back beyond the adultery of Atreus, the πρώταρχος ἄτη ; nothing is said of the
guilt of Pelops or the ὕβρις of Tantalus. Cf. on 1193. But it is possible that the
name has an ominous sound here.
1470 f. κράτος ἐκ γυναικῶν . . . κρατύνεις is clear. The work of the daimon
originates ἐκ Κλυταιμήστρας καὶ “Ἑλένης. This thought foreshadows what in a
far more startling manner Clytemnestra says, 1500 ff., of the ἀλάστωρ. The
phrase καρδιόδηκτον ἐμοΐ is no less clear: it is in apposition to κράτος... ἐκ
γυναικῶν κρατύνεις (so Wecklein ad loc. and Wilamowitz in his apparatus).
But ἰσόψυχον is obscure; at any rate I have not arrived at a certain inter-
pretation. The usual way of taking it is determined by the scholion in Tr:
Κλυταιμνήστραν kal “Ἑλένην λέγει, ai κατὰ φαυλότητα ἴσας τὰς ψυχὰς ἔχουσι.
The normal translation is: ‘pari animo praeditus’ (Blomfield), ‘parilis animi’
(Hermann), ‘of equal spirit’ (L-S). This interpretation leads to the question:
equal to what? The context does not provide an obviously satisfactory
answer. The majority of commentators follow the scholiast and connect
ἰσόψυχον with γυναικῶν; Daube, 25 ἢ. 62, actually speaks of ‘enallage’. The
commentators differ sharply on the point of comparison in ioo-: many believe
with the scholiast that the ψυχαί of the two women are compared with each
other ;’ others: ‘the power of two women equally imperious with men’
(Sidgwick), similarly Wilamowitz in his translation: ‘der du dem Weibe
gabest des Mannesmutes Starke’. We need not discuss this in detail because
the whole basis of this explanation is dubious. I do not believe that it is
justifiable to treat the phrase ἐκ γυναικῶν as if it were an attribute of κράτος
(‘enallage’ would, of course, be possible with κράτος γυναικῶν ἰσόψυχον), when
in fact it performs the function of an adverbial determination of the verb
(see above). Moreover the interpretation of ἐσόψυχος as ‘pari animo praeditus’
or the like is in no way certain. Where ἐσο- is the first element of a compound,
either the second element produces ‘kasuelle Determination’ (Debrunner,
Griech. Wortbildungslehre, 41 ff.) as in ἰσόθεος (ἴσος θεοῖς) and the like, or ἐσο-
produces ‘attributive Determination’ (Debrunner, 43 f.) as in ἰσόψηφος ‘with
like voting-pebbles’ and the like. When we go through the lists put together
on 1442 f. we recognize first that in early poetic language the ἰσόθεος type is
far more common,’ and secondly, that in the examples of the ἰσόψηφος type
1 This view is shared by the theologian A. Fridrichsen, who, Symbol. Osl. fasc. 18, 1938,
42 ff., commenting on the Epistle of St. Paul ad Philipp. 2. 20, examines the use of ἐσόψυχος
and in that connexion deals with A. Ag. 1470.
2 Our lexica ascribe a wrong meaning to ἐσόνεκυς in E. Or. 200: ‘in gleicher Weise gestor-
ben’ (Passow), ‘pari fato defunctus’ (Dindorf’s Thesaurus), ‘dying equally or alike’ (L-S,
with an entirely unfounded reference to the scholion). The word clearly means ‘like one of
the dead’. The point of Electra’s words lies in her saying : ‘we are both like the dead: you,
695
lines 1470 f. COMMENTARY

in pre-Hellenistic Greek the second element always belongs to a strictly

limited group of concepts: it denotes number, length, etc., or at any rate
something which can be measured, counted, or weighed, and can therefore be
compared with other things in measure, number, volume, or weight, so that
the equality indicated by ioo- is quantitative. This is particularly clear in the
following type, which is very common in Plato and Aristotle and technical
language, ἰσάριθμος, ἰσομήκης, ἰσόπλευρος, ἰσοβαρής, etc.; to this also belong
ἰσόπεδος (it occurs as early as Herodotus) and ἰσοπλατής, ἰσοπληθής (both in
Thucydides). Numerical equality is indicated also in ἐσόψηφος and ἐσήρετμος ;
the number of pebbles is essential in voting, the number of oars determines
the quality of the ship; cf. also ioaudpıos (ἰσημερινός measures the length of
day and night against each other) and ἰσῆλιξ (Xenophon). The reference in
ἐσοτελής is to number and measure, in ἰσόρροπος to weight. In ἐσόμοιρος (cf.
the Homeric ἐσόμορος) the quantitative element is as clear as in ἰσόνομος and
iooxparns (where it is not the quality of the νόμος and κράτος that is in
question but the greater or less amount in comparison with the other citizens).
Thucydides was probably thinking of social-political concepts like the two
last named when he formed (1. 6. 4) iooëiaros (taken over by Lucian and
Cassius Dio) for a context in which the keynote is δίαιτα (1. 6. 1 τὴν δίαιταν,
I. 6. 2 διαιτημάτων, 1. 6. 3 ἁβροδίαιτον) ; it is quite clear that the point is a
greater or less display in relation to the rest: μετρίαι δ᾽ αὖ ἐσθῆτι... πρῶτοι
“Λακεδαιμόνιοι ἐχρήσαντο καὶ ἐς τὰ ἄλλα πρὸς τοὺς πολλοὺς οὗ τὰ μείζω κεκτημένοι
ἰσοδίαιτοι μάλιστα κατέστησαν. Measuring is also the main element in agonistic
concepts like ἰσοπαλής, icoxivduvos,’ ἰσόδρομος. A single instance falls outside
this fairly well-defined group, Pind. Ol. 9. 64, where Pindar has for once
allowed himself to use ἰσώνυμος, which is bold for the language of his time,
instead of the normal ὁμώνυμος, which he uses elsewhere. The result of all
this examination is that, if Aeschylus used ἰσόψυχος in the sense of part
animo praeditus, he broke loose in this single case from the restrictions almost
invariably observed not only by himself—and he used a great many ioo-
compounds (cf. on 1442 f.)—but by all Greeks of the pre-Hellenistic period.
ψυχή is completely different from the concepts quoted above, all of which
admitted quantitative determination. Nevertheless, in view of the great
boldness which Aeschylus shows in the formation of compounds, I should not
dare to maintain that he could not have used ἐσόψυχος (possibly coined for
this passage) in this sense. But it must be admitted that this is unlikely.”
If ἰσόψυχος was to mean pari animo praeditus, its coinage would have been
considerably bolder than Pindar's unusual ἐσώνυμος, which is supported by
the common ὁμώνυμος. Moreover, in Pindar the clause μάτρωος δ᾽ ἐκάλεσσέ νιν
ἰσώνυμον ἔμμεν leaves no doubt of the relation of the ἐσο- element, whereas on
the meaning of ἐσο- in κράτος ἰσόψυχον no light is thrown by the context.
After reaching this somewhat unsatisfactory result, at least an attempt
Orestes, who lie here now scarcely breathing (cf. 84), and I whose life all these long years
has been more death than life (zo1 ff.)’.
1 Thuc. 6. 34. 7 (imitated by Cassius Dio, found nowhere else), obviously coined by him,
as also is ἐσοδέαιτος. It is rightly explained by Portus, K. W. Kriiger, and others as ‘in pari
discrimine futuros’, while K. L. Bauer and Classen-Steup take it as ‘pares periculo’.
2 There can be no appeal to ἐσόθυμος, translated in L-S like ἰσόψυχος ‘equal in spirit’,
since it is not only quite late but also dubiously attested, Schol. Hom. H 295, where the
B-Scholia have πολλοὺς yap εἶχεν ἰσοθύμους but the T-Scholia m. y. e. ἀντιζήλους,
696
 COMMENTARY lines 1470 f.

must be made to treat ἐσόψυχος as an example of 'kasuelle Determination',

like ἰσόθεος, etc. We have seen that this type of ioo- compound is much
commoner in early poetic language than the other (ioójmóos). Besides, a
Greek could at any time form (or use) ἰσόψυχον in the sense of ἴσον τῆι ψυχῆι.
The sentence E. Andr. 418 f. πᾶσι δ᾽ ἀνθρώποις ἄρ᾽ ἦν ψυχὴ τέκνα is in a
scholion paraphrased by the words ἐσόψυχα τοῖς ἀνθρώποις... τὰ τέκνα (quoted
by van Heusde), where ἐσόψυχα, 'as precious as life', is used in exactly the
same way as, e.g., iodpyvpos by Aeschylus. If then here κράτος ἴσον ψυχῆι (or
ψυχαῖς) is to be understood, it may perhaps mean that the domination which
the daimon exerts 'from the women' (see above) is like a soul, or of the same
value as a soul, i.e. operates as if it came from the souls (of Clytemnestra and
Helen?). But this expedient seems to me very forced, and I am not even
certain whether such a thought accords with the conception of ψυχή held by
Aeschylus and his contemporaries. À further attempt : could ψυχή here mean
‘strength of soul, courage'? This meaning is found in Pers. 28 and 442 (‘inner
strength, courage' Wilamowitz in his commentary, Griech. Leseb., Erläut. 36)
and in the negative Sept. 192 ἄψυχον κάκην, ‘without courage’ (cf. Wilamowitz,
Glaube d. Hell. i. 373 n. 3, who also quotes Hippocr. z. dép. 24, i, p. 70. 14 Kühle-
wein, τὴν ψυχὴν κακοί, Hdt. 5. 124. 1 ψυχὴν οὐκ ἄκρος, etc.) and Ag. 1643 ἀπὸ
ψυχῆς κακῆς. So the expression might possibly mean 'a domination which is
like strength of soul (a courageous heart)', recalling γυναικὸς ἀνδρόβουλον
kéap—but that seems to me an obscure expression, whereas in all the passages
quoted for ‘courage’, etc., the context leaves no doubt of the meaning. Enger
(in Klausen?), too, attempted to take ἰσόψυχος in the sense of ἴσος τῆι ψυχῆι,
but when he renders it by ‘convenientem animo mulieris’, he fails to bring out
the meaning of ioo-.!
Finally I must give a warning against attempts to suppress the second
element of ἰσόψυχον in order to achieve a suitable meaning; so, e.g., S. Butler:
“ἰσόψυχον h. 1. nihil aliud quam simpl. ἔσον, par vel simile valet’, and by
implication Keck : ‘O Dämon, der du... eine gleiche Macht in diesen beiden
Weibern offenbarst' and also Wilamowitz, Interpr. 199: ‘Dämon, der du in
den Häusern der beiden Atreiden gleichmächtig in den Frauen lebst’.
The fact that we do not understand ἰσόψυχον does not justify the conclusion
that it must be considered corrupt. Conjectures have been made, e.g. by
Hartung and Weil ; the latter's κακόψυχον (adopted by Wecklein in his edition
of 1888 and A. Y. Campbell) is palaeographically elegant but contrary to
sense, for in the language of Aeschylus and the Ionians (see above) it would
mean 'coward'.^ However, the possibility of a corruption cannot be com-
pletely excluded so long as the word is not satisfactorily explained.
In order to commend his κακόψυχον Weil says that Hermann inserted τ’
after κράτος ‘metri causa’. This is not fair: Hermann's alteration, like every
good conjecture, restores the thought as well as the metre. The two clauses
ἐμπίπτεις δώμασι καὶ διφυίοισι Τανταλίδαισι and kpdros . . . ἐκ γυναικῶν κρατύνεις

ı Tam not clear what linguistic interpretation of the word underlies Verrall’s explanation.
He gives as a literal translation ‘and winnest a victory, equal in lives on the part of the
(respective) wives’, and expounds ‘a victory (κράτος) equal in lives as between them; i.e.
one in which they may share the destruction equally’.
2 The possibility of reading κακόψυχον is not ruled out by the observation that this word
is not found anywhere. Plato (Laws 791 c) coins xaxojvxía to balance εὐψυχία.

697
lines 1470 f. COMMENTARY

are parallel in thought and complete each other; they should therefore
be co-ordinated (this was probably felt by Pearson, who inserted δ᾽. If
Kpáros . . . kparüveis is made into an independent main sentence, the close
connexion is severed. Hermann suggests a possible reason for 7’ having
been omitted (though it is often dropped through mere oversight) and points
to the usage by which an invocation is often not shaped into a full sentence:
‘Omiserant 7’ librarii, ne deesset quod genio diceret chorus, non memores
huiusmodi compellationes saepissime nudas poni, quoniam intelligitur £e
appello." This form of apostrophe is common: the vocative is followed by one
or more attributes in the form of a relative clause (cf. Norden, Agnostos
Theos, 168 ff.), to which further attributes may be added loosely (by δέ or the
like). It is characteristic that the invocation is not followed by either a
statement (e.g. ‘I summon thee’ or the like) or a prayer (‘Come’, ‘Help’,
‘Bring about’—or ‘Leave undone'— this or that’, etc.), so that the vocative
with its relative (or other) attributes seems to hang in the air. This form is
not confined to prayers and addresses in the form of prayer; the poets used
it extensively for purely poetical apostrophes. It may be well to illustrate the
type in question by a few examples; I have confined my selection to passages
in which the attribute following the apostrophe has the form of a relative
clause. Theognis x. 15 ff. Μοῦσαι καὶ Χάριτες, κοῦραι Διός, at ποτε Κάδμον és
γάμον ἐλθοῦσαι καλὸν ἀείσατ᾽ ἔπος" ‘ ὅττι καλόν, φίλον ἐστί" ᾿ krÀ.,! Pindar, P.
I. τ, N. 8. x f£, S. Ant. 781 ff. (the hymn to Dionysos, Ant. 1115 ff., is
essentially of the same type, although after more than two complete stanzas
(1140 ff.) a genuine prayer, a request for the presence of the god, follows),
Phil. 391 ff. (not a prayer ; the statement 395 ff. oe κἀκεῖ μᾶτερ πότνι᾽ ἐπηυδώ-
pav, ὅτε κτλ. in relation to the apostrophe is, as it were, parenthetic or at any
rate subordinate to it: ‘as then, so I summon you now’). In Euripides this
form is often found as a conventional literary device. Though he sometimes,
like the older poets, uses it for addressing divine or semi-divine beings, e.g.
Iph. T. 156 ff. (δαῖμον as Ag. 1468), El. 54 ff. (νύξ), he also applies it to all
kinds of things; thus Tyo. 122 ff. and El. 432 ff. in addressing ships, Hipp.
752 ff. à... Κρησία πορθμίς, à . . . ἐπόρευσας krÀ.,? Ion 492 ff. ὦ Πανὸς θακήματα
καὶ... πέτρα. .. ἵνα xrd.; Iph. T. 1106 ff. ὦ πολλαὶ δακρύων λιβάδες, al...
ἔπεσον krÀ. The secularization of this originally hieratic form of address is
strikingly shown by the use Euripides makes of it at the beginning of certain
prologue speeches (Alc., Andr., El.).
1471. καρδιόδηκτον (certain restoration; for the type of corruption see on
814) active in sense, cf. on 12. For the idea see on 743.
1472. por might in itself be right in this sentence, where we cannot determine

ı P. Friedlander, Hermes, xlviii, 1913, 575, comments on this: ‘here one is puzzled to note
that after the apostrophe and the relative predication there follows no prayer properly
speaking. One might expect a Sedre’, etc. But his appreciation of the passage is still
valuable (cf. on the subject F. Jacoby, ‘Theognis’, Berl. Sitzgsb. 1931, 102).
2 The scholiast here feels the lack of a principal clause and supplies : ov μοι, φησί, προοίμιον
τῶν θρήνων, ὦ πορθμίς, ἐπεὶ καὶ τῶν συμφορῶν ὑπόθεσις γεγένησαι. A similar apostrophe,
though not followed by a relative clause, opens the choral ode Hel. 1451 fl. Φοίνισσα Σιδωνιὰς
ὦ ταχεῖα κώπα krÀ. Here the typical form is strangely misrepresented by A. C. Pearson
ad loc. : ‘The vocative is not followed up by any verb, so that vv. 1451-1464 serve merely to
direct the attention of the reader to the general subject of the ode which they introduce’
(from so fine a scholar ‘reader’ comes as a shock).
698
 COMMENTARY lines 1473 f.
with certainty the main verb and the precise construction, but its position
inside the phrase δίκαν κόρακος ἐχθροῦ awakes mistrust. Dindorf (Adnota-
tiones, Oxford 1841) probably rightly suggested deletion. It was possibly
inserted to make the metre agree with the strophe (1452), when καὶ had been
wrongly inserted there.
1473 f. Despite the mutilation of the text at the end there seems to me to be
no reason for a pessimistic diagnosis as far as the sense of this sentence is
concerned. Wilamowitz, it is true, with other editors, regards ἐπεύχεται as
corrupt, and it happens in fact often enough that the words beside a lacuna
are tampered with. But ἐπεύχεται seems so fitting here that it is difficult to
see why it should be suspected. The verb in the sense of gloriari was used in
1262 to describe Clytemnestra’s behaviour; then she used it herself in the
first great speech which she made beside Agamemnon’s body (1394). It
should be realized that the sentence ἐπὲ δὲ σώματος... σταθεὶς... ἐπεύχεται
bears a relation to the description that Clytemnestra gave of her own
demeanour. Must we therefore follow Stanley (in his posthumous notes:
‘Vide an non potius σταθεῖσ᾽ legendum sit’), Schütz, Porson, Blomfield,
Hermann, and others! in interpreting σταθεὶς as σταθεῖσ᾽ and making Clytem-
nestra the subject? It would be possible to take a γυνή from the preceding
γυναικῶν, and if 1472 ff. followed what goes before in asyndeton, this inter-
pretation might be considered. But as the connexion is marked by δέ, the
transition to a new and undefined subject after the appeal to the daimon
would be incredibly abrupt. Moreover, the immediately following answer of
the queen 1475 ff. makes it unlikely that the Chorus in this stanza has spoken
of Clytemnestra as well as of the daimon ; in this whole section Clytemnestra’s
answers are related very closely to the preceding words of the Chorus (parti-
cularly significant is the division of the answer 1462 ff. corresponding to the
two separate ideas voiced by the Chorus). Therefore the spirit of evil must
be the subject of the sentence 1472 ff. After their impassioned invocation of
the δαίμων the old men do not address him any further but go on to speak of
him : the appeal falls back into the normal form of utterance.? The transition
from an appeal in the second person to a statement in the third person is even
found where the strict form of prayer has been kept at the beginning, Pind.
Ol. 14. 1-11; but the transition there is not so abrupt as it is here. When we
now turn to the thought expressed in 1472 ff., we see that the action ascribed
to the δαέμων reflects Clytemnestra's words and attitude. This is in harmony
with one of the basic ideas of this scene. It is ἐκ γυναικῶν that the spirit
of evil shows forth his power; soon afterwards Clytemnestra will say that
the old, grim ἀλάστωρ of the house has taken her shape. The suggestive

1 So, e.g., W. Kranz, Stasimon, 141.

? Such a falling back, only without a verb in the third person to pick up the verb in the
second person, is to be found in A. Cho. 363 ff. μηδὲ... πάτερ... τεθάφθαι, πάρος δ᾽ of
kravóvres ww οὕτως δαμῆναι, where, like Hermann, Weil (‘uv ferri nequit, quum Electra
patrem alloquatur’), and others, Blass has removed the νιν, Conington rightly defends it:
*Electra thinks less of addressing her father than of framing her wish, and so passes in
speaking of him from the second to the third person.’ For the uncommon way in which the
subject of an infinitive equivalent to the third person of the imperative is put in the
nominative (οὗ xravóvres) see Wecklein ad loc., cf. also Leaf on Z 92. In S. Trach. 1112 f.
ὦ τλῆμον ‘EAAds, πένθος olov εἰσορῶ ἔξουσαν, ἀνδρὸς τοῦδέ γ᾽ εἰ σφαλήσεται the σφαλήσεται
has been objected to by some critics.

699
lines 1473 f. COMMENTARY

comparison δέκαν κόρακος ἐχθροῦ seems much more appropriate for the δαίμων
than for Clytemnestra.
Now the detail. δίκαν κόρακος: the raven in the ancient world (e.g.
Theognis 1. 833 πάντα τάδ᾽ ev κοράκεσσι καὶ ev φθόρωι, Catullus 108. 5, Horace,
Epist. 1. 16. 48 non pasces in cruce corvos, Petronius 58. 2 cructs offla, corvorum
cibaria, Plutarch fr. 28 Bernardakis ὥσπερ of κόρακες παρεδρεύοντες ἐξορύττουσι
τοὺς τῶν νεκρῶν ὀφθαλμούς), as elsewhere (‘Rabenstein’ etc.), is the bird which
feeds on corpses. That ἐννόμως could make sense here is improbable ; ἐκνόμως,
on the other hand, fits ὕμνον ὑμνεῖν admirably; everyone will think of 1142
θροεῖς νόμον ἄνομον.
The image of the raven belongs to the phrase within which it stands: ei
σώματος orabeis, In what follows, ὕμνον ὑμνεῖν, the raven (for which κρώζειν
or some such word would be suitable ; Wilamowitz, in his translation, almost
rewrites the text ‘singen dein Lied mit Rabengekrächze”) is to be forgotten in
favour of the daimon; thus in 1191 the Erinyes ὑμνοῦσι δ᾽ ὕμνον and in r117
Zrdois (taken up by ’Epwvs in 1119) is given an oAoAuyr. To try and recover
the two syllables lost at the end would be useless guesswork ; an epithet for
ὕμνον would be welcome here.!
R. Arnoldt, Der Chor. im Ag. des Aesch. (1881), 79, Paul Schwarz, De
ephymniorum ap. Aesch. usu (diss. Halle 1897), 36 ff., and Kranz, Hermes, liv,
1919, 315, have brought definitive proof that the mention of the daimon must
be immediately followed by the answer in 1475 ff. and that the whole con-
nexion of thought in the passage is destroyed by the repetition of the
ephymnium? after 1474, for which there is no foundation in the MSS.
1475. viv δ᾽, Headlam deletes δ᾽, perhaps correctly. Hermann’s defence "δέ
refertur ad suppressum πρόσθεν μὲν οὐκ ὀρθῶς ἔλεγες ᾿ is attractive, but a
parallel would be welcome.
γνώμην here is not ‘motion, proposal’ as 1348 and elsewhere, but in a
general sense ‘expression of opinion’.
1476. τριπάχυντον (the emendation of Bamberger, Opusc. 155) I regard as
certain in spite of the scepticism in, e.g., L-S. Bamberger quoted Suppl. 616 ff.
ἱκεσίου Ζηνὸς κότον μέγαν προφωνῶν μήποτ᾽ εἰς ὄπιν χρόνου πόλιν παχῦναι,
‘vetans ne civitas Iovis magnam iram alat’, where the image is immediately
taken up (619 f.): λέγων διπλοῦν μίασμα πρὸς πόλεως φανὲν ἀμήχανον βόσκημα
πημονῆς πέλειν. Hermann compared Sept. 770 ὄλβος ἄγαν παχυνθείς. The τρι-
element is simply strengthening, as in τριγέρων μῦθος and the like. Cf. also
phrases such as the popular (Ar. Knights 1153) τρίπαλαι κάθημαι. The idea of
the raging of the daimon through three generations (Stanley) must not be
read into this passage; Wecklein was right on this, but the wrong view
(Tantalos, Thyestes, Agamemnon) is found again in G. Thomson. Nor does
1 Headlam's suggestion ἐκνόμοις . . . ἐπεύχεται νόμοις was anticipated by Kennedy, who
put it into the text.
2 Now that the unwarranted repetition of the ephymnium here and after 1566 is given up,
the last vestiges of C. Burney’s fanciful book Tentamen de metris ab Aeschylo in chor. cant.
adhibitis (Cambridge 1809) have disappeared from our editions of the Agamemnon.
3 Wilamowitz in his commentary (1896) on Cho. 552 takes 8' οὕτως there as δὴ οὕτως
(in his edition of Aeschylus he reads like Weil θ᾽ οὕτως) and says: ‘this crasis . . . is also
found in δὴ οὕτω [he is probably thinking of δὴ &p8woas] Ag. 1475’. I mention this simply
because it seems to show that at that time he was not happy about νῦν δέ. As regards his
suggestion, it may be noticed that νῦν δή does not occur in Aeschylus.
700
 COMMENTARY line 1479

the text suggest Wilamowitz's interpretation (Interpr. 199) ‘by rpımaxuvrov

she pushes the beginning of the series of crimes back to Atreus’.
1477. It is not clear what induced Hermann to put γέννας into the text in
place of the well-attested! yévvgs. In the anapaestic interludes of this
kommos, as also in those of the kommos of the Choephoroe (306 ff.), there is no
trace of the dialect used in true lyrics (with which are to be classed, e.g., the
anapaests Pers. 922-47, which are almost 'threnodic anapaests’, cf. Wilamo-
witz, Verskunst, 368: they have the dative γένναι and the genitive γέννας).
In the kommos of the Choephoroe Hermann never thought of adopting the
variants (cf. Blass, Aisch. Cho. 22) 377 γᾶς, 403 drav. Cf. on 1569. The forms
γέννης or yévrm do not seem to occur elsewhere, but it is probably pure
accident that in our texts the genitive and dative of this word are not found
elsewhere outside purely lyrical passages.
κικλήσκων, ‘appealing to’, as in a prayer Ζῆνα... μέγαν κικλήσκω etc., cf.
also Suppl. 212 καὶ Ζηνὸς ὄρνιν τόνδε viv κικλήσκετε, corresponding to 217; an
appeal to gods is also implied Zum. 508 ff. μηδέ τις κικλησκέτω . . . ‘ ὦ Δίκα,
ὦ θρόνοι τ᾽ ᾿Ερινύων ’. Clytemnestra interprets the cry wrung from the Chorus
in their grief (1468 δαῖμον ὃς ἐμπίπτεις κτλ., cf. E. Iph. T. 156 ff.) as a prayer.
“Your appeal is justified because it was the daimon who caused the ever-
renewed desire for blood.’ ἐκ τοῦ yap κτλ. must have recalled to a Greek
audience the many prayers and hymns in which the appeal to a god is fol-
lowed by a γάρ clause giving an account of his deeds or his power: cf., e.g.,
Pind. Ol. 4. 1 ff. ᾿Ελατὴρ ὑπέρτατε βροντᾶς ἀκαμαντόποδος Ζεῦ" reat yap ζῶραι
κτλ., E. Bacch. 519 ff. ᾿ἀχελώιου θύγατερ, πότνι᾽ εὐπάρθενε Aipka, σὺ yap ἐν σαῖς
ποτε παγαῖς τὸ Atos βρέφος ἔλαβες κτλ., Ar. Eccl. 3 (for the details see my book
Plautinisches im Plautus, 102 n. ı). Clytemnestra in her turn continues the
address to the daimon which the Chorus had begun and by her added lines
converts it into a true recognition of the power of that divine evil-doer. As
she does not speak to the daimon himself, she turns the prayer formula ἐκ σοῦ
γὰρ κτλ. into the third person. The Chorus completely understands the quasi-
religious, eulogistic tone of her words, cf. on 1482 aiveis.
1478. aiparoAoıxös (only here): the o-grade of the verb, as in other com-
pounds χρυσαμοιβός, ἀλφιταμοιβός (ἀμείβω). The image of the gluttonous
daimon, the τριπάχυντος, is continued here and in τρέφεται.
1479. veipär. Most modern editors have accepted this restoration. Complete
certainty is impossible. Wilamowitz in his app. crit. seems to hint that the
information on the word contained in the scanty remnants of the lexico-
graphers cannot be fully relied on. If we are prepared to follow Hesychius
νεῖραι: κατώταται" of δὲ κοιλίας τὰ κατώτατα and the paraphrase in the scholion
on Lycophron 896 (χθονὸς νειροῖς μυχοῖς) τοῖς ÉGWTATOIS . . . τόποις τῆς γῆς, We
achieve a reasonable sense. It is well known (cf. E. Lófstedt, Syntactica, ii.
345 n. 2) that the Greeks and the Romans preferred in such cases to speak of
‘lowest, deepest’, i.e. imis medullis and the like (Wilamowitz here: ‘libido
sanguinis imis visceribus insita’), where we speak of ‘the innermost’ (‘the
inward maw’, Headlam). Against veipa: G. Thomson (ii. 379) says: ‘it seems
1 F (also G) reads γέννης with the correction as above the end of the word. This has been
correctly noted by Hermann, Wecklein, and others, while the erroneous statement in
Wilamowitz’s edition has been adopted by A. Y. Campbell and G. Thomson ; Murray’s note
is not sufficient.
701
line 1479 COMMENTARY

objectionable both in grammar (it should be ἐν vetpm) and in sense (“is

nourished in his belly")'. The second objection arises from a mistake: it is
not the daimon of whose belly Aeschylus is speaking, and the commentators
have rightly understood νείραι τῆι τῆς yévvgs. As for Thomson's first objection,
it is difficult to see why Aeschylus should not have used the non-Attic νείρηι
or νείραι (always supposing that this is the genuine reading), which came to
him as a γλῶττα probably from earlier poetry, in the sense of a locative just
as Homer used eoxarını which is allied in meaning (Y 242, £ 104; with the
addition of a genitive A 524). The attempt to take νείραι as a true dative
seems less natural; thus Schneidewin 'adnutritur ventri posterorum', and
van Heusde, who compares for the dative 735 f. ἱερεύς τις ἄτας δόμοις προσε-
θρέφθη (but there the addition of the preposition makes the construction
quite clear).
1480. Headlam substitutes fyap for ἐχώρ and interprets νέος ἶχαρ (νέος τὴν
ἐπιθυμίαν) as ‘fresh in appetite’. tyap stands in a hopelessly corrupt sentence
Suppl. 850 with the scholion τὴν ἐπιθυμίαν ἴχαρ εἶπε. From the point of view
of etymology this explanation of ixap (or fyap) seems unobjectionable:
Wackernagel, Vermischte Beiträge, 17, regards tyap as belonging to ἐχανᾶν.
Whether Headlam's ingenious alteration is necessary I do not know. He is
certainly right in establishing that tywp never means simply blood (as here
almost all commentators assume, including Blass on Zum. 359, cf. also L-S).
A similar warning was voiced by Wilamowitz, Litteris, i (Lund 1924), 4:
* ixop cannot be a foreign word for blood ; discussion must not start with the
well-known Homeric passage but with Hippocrates and the later physicians’.
Evidence from the Corpus Hippocraticum is to be found in, e.g., Bechtel,
Griech. Dial. iii. 306. The word means, if we disregard its special use in
Homer, ‘serum’, ‘a sero-purulent discharge' (L-S). This meaning is brought
out in Wilamowitz's interpretation of νέος ἐχώρ : 'hoc si integrum est, dicitur
libido sanguinis imis visceribus insita "nova sanies, novum virus’, quod
morbum novum excitat vixdum sanato priore. But Wilamowitz was fully
aware of the uncertainty which necessarily remains, as his 'hoc si integrum
est' shows.
The craving caused by the δαίμων, a craving for murder (αἱματολοιχός),
manifests itself in fresh disease (before the old one has ceased): therefore
νέος ἰχώρ is put in apposition to what precedes it.
1481. ἢ μέγαν οἴκοις τοῖσδε: Dindorf and Kirchhoff felt as little doubt as the
critics before Hermann that this line could correspond to 1505 ὡς μὲν avairios
εἶ (Verrall's private theory of responsion, Append. II, $ 1 of his edition, does
not call for comment). Nor did Schütz show himself a true metrician when
he thought it sufficient to add a σὺ to 1505, 'ut metrum expleretur', and to
leave 1481 unaltered (so also Murray) : in the dactylic hemiepes, μῆνιν ἄειδε θεά
or ἄνδρα μοι ἔννεπε Μοῦσα, which he often uses, Aeschylus never allows a
spondee except in the first place. To add o? in the antistrophe and alter
1481 so as to produce — v u — u -- Ὁ is metrically unobjectionable ; it was on
these lines that Weil proceeded (Addenda, p. 139, to his edition of Eum.
1 I do not know why Wilarnowitz encloses in brackets this scholion as one of those which
he believes to be late (cf. the foot of p. 366 of his ed. mai.) and remarks 'scholia videntur
recentia, certe inutilia", and on the other hand takes ἐόν at the end of the line to be an old
gloss on ἰχώρ corrupted to ἔχαρ.
702
 COMMENTARY line 1484

[1861]), whose reading ἦ μέγαν, 7) μέγαν οἴκοις was commended with reserve by
Headlam and adopted by Mazon and G. Thomson ; similarly Wecklein (his
conjecture 4 μέγαν ἔγκασι τόνδε is fantastic). But as the wording of 1505 ὡς
μὲν ἀναίτιος el is unimpeachable in sense and metre,’ it is better to add nothing
there and to conclude that the metre of 1481 must also be - v s — oo —. This
was Hermann's view, but his alterations are unnecessarily violent; of his
followers Schneidewin's conjecture, inspired by 155, should be mentioned,
À μέγαν οἰκονόμον δαίμονα xrÀ., but the most elegant is Wilamowitz's οἰκοσινῆ.
The word is not attested but is as correct a formation as the unique πολυσινής
of Cho. 446. Wilamowitz surmised that first corruption to οἰκοσι took place
and then wanton padding. The objection that his correction (or, for that
matter, Schneidewin's) introduces a disturbing piling up of epithets will not
hold water: μέγαν is not on the same level as the adjectives which describe
the effect of the daimon οἰκοσινῆ and βαρύμηνιν, cf. 1102 f., Cho. 358 f. πρόπολός
τε τῶν μεγίστων χθονίων ἐκεῖ τυράννων, Eum. 1032 f. μεγάλαι φιλότιμοι Νυκτὸς
παῖδες ἄπαιδες (Wilamowitz's alteration φιλοτίμωι, to go with the preceding
νόμωι, is not acceptable), also Pers. 852 f. μεγάλας ἀγαθᾶς re πολισσονόμον
Bıoräs, where not only μεγάλας but pey. dy. re is added to the single idea
πολισσονόμου Bıoräs. The order οἰκοσινῆ δαίμονα καὶ BapÜüumrw is uncommon but
legitimate, as has been shown p. 212.
1482. aiveis: ‘you speak of’ (Paley, also L-S) does not render the word
correctly;? Wecklein (‘preisest du’), Wilamowitz, Interpr. 163 n. x (‘prae-
dicas"), are right; cf. E. Hofmann, Qua ratione ἔπος, μῦθος... adhibita sini
(diss. Göttingen 1922), 57 f.: ‘Clyt. daemonem praedicavit, non solum de eo
verba fecit, nam chorus eum iam antea commemoraverat’. In the following
line atvov does not mean ‘tale, story’ or the like but ‘eulogy’. Cf. on 1547.
The Chorus has noticed and is incensed by the tone of Clytemnestra’s words
(prayer and praise of the daimon).
1483. κακὸν alvov is in apposition to the sentence δαίμονα... αἰνεῖς (cf. 1494).
1484. The thought is clearly that the daimon with insatiable greed seeks
disastrous success. Wecklein (edition with commentary) and Wilamowitz

1 The hemiepes — vu — u v — stands at the beginning of the strophe here as in A. Suppl.

42 — 49 (uncertain 843 — 854), Cho. 380 — 394; an alcaic decasyllable follows as in Suppl.
845 f. = 856 f., Ag. 1005 f. (the loss of one line here does not create uncertainty) = 1023 f.
2 This rendering is inadequate also for Ag. 98, cited in L-S in the same section, cf. ad loc.
If one looks through the evidence for the use of αἰνεῖν in Aeschylus and in older Greek in
general, one begins to doubt whether in Cho. 279 it is allowable to follow Hermann, who
took the MS reading raodevaw to mean τάσδ᾽ aivàv, which involves practically no change (for
the confusion of e with ac see on 1653). However that may be, Wilamowitz's defence (his
final view is given Znlerprel. 204) of τάσδε νῶιν does not seem to me successful ; it is the more
surprising since he himself (/nterpr. 205) emphatically states that this χρησμός is not
limited to the special case of Orestes but has a wider application, as Dobree, Adversaria,
ii. 27, and Conington (edition of Cho., Appendix II, p. 164) had pointed out with fine
observation of detail. Conington's rendering of Cho. 1014 νῦν αὐτὸν αἰνῶ is wrong ; he refers
αὐτὸν to φόνος and adds: “αἰνῶ then is to mention, as in Ag. 98’; so Paley (It is only now
that I speak of it’). Nor does Wilamowitz’s interpretation (he puts a lacuna before 1014):
‘ich constatiere die Tat (aio ita factum esse)’ find any support in the use of aiveiv. For
Hermann’s mistaken rendering v. infra p. 800 n.2. I believe Verrall’s interpretation of Cho.
1014 to be clearly right : ‘At last, at last he (Ag.) himself is before me; I utter his praises;
I make his lament.’ Headlam's prose translation is very close to this; so also Tucker;
Sidgwick is right in the main, but the verb is incorrectly translated in ‘Now I speak of
himself".
703
line 1484 COMMENTARY
have rightly adopted Todt’s ἀκόρεστον. Verrall tries to achieve the same
result, keeping the MS reading dxopeorov and explaining: ‘literally “a fatal
praise of him as never tiring of deadly stroke". The genitive τύχας depends
on ἀκορέστου (masculine).’ That is not only complicated, but obscure. If my
metrical analysis is correct, the kolon φεῦ φεῦ κακὸν αἶνον, which has no
syntactic connexion with what follows, is also marked off as a separate
metrical unit.
1485. Dindorf's alteration of ἐὼ ij into ἰὴ ἰή (followed by Sidgwick, A. Y.
Campbell, and Murray without their saying that they diverge from the
MSS) is unwarranted ; ἐὼ iw follows in &j in the θρῆνος of the Persae, 1004.
1485 f. This passage provides a welcome confirmation of what has been said
on 448 about διά with the genitive in a small group of Aeschylean passages:
again the use is causal-instrumental, marking at once both the agent and the
tool. The first is clearly expressed in παναιτίου, the second in zavepyéra. This
use of the preposition, so far from being common, is perhaps peculiar to
Aeschylus. Not only is the agency of a divinity generally indicated by διά and
the accusative (cf. on 698), but this is particularly true in the case of Zeus;
E. Norden, Agnostos Theos 347, emphasizes the noteworthy exception of this
one passage. Oui Aids has an appeal to the ear. Blass on Cho. 948 ff. and
Norden (op. cit. 164 n. 4) have suggested very plausibly that Aeschylus has
in mind here the etymology of the god’s name which is found later in Plato
(Crat. 396 a, b) and particularly in the Stoa' (Chrysippus ap. Stob. Ec. i.
p. 31. 11 W. = fr. 1062 v. Arnim, Sfoic. vet. fragm. ii. 312 Δία δὲ αὐτὸν λέγουσιν
ὅτι πάντων ἐστὶν αἴτιος kai δι᾿ αὐτὸν πάντα, cf. Aristides or. 43. 23, vol. ii,
p. 345 Keil) ; if Norden's surmise (op. cit. 259 n. 1) is correct, Hesiod, Erga 3,
already alludes to this. Aeschylus develops also another widespread ancient
etymology of the name of Zeus (cf. O. Gruppe, Mythologie, 1100 n. 2; A. B.
Cook, Zeus, i. 11 n. 5; O. Gigon, Untersuchungen zu Heraklıt, 1935, 139),
deriving it from ζῆν (Suppl. 584 ff.), cf. R. Pfeiffer, Sitzgsb. Bayer. Akad.,
Phil.-hist. Abt., 1938, Heft ii. 9 n. 2. The complement to the positive διαὶ
Διός is contained in the sentence τί γὰρ βροτοῖς ἄνευ Διὸς τελεῖται, which (cf.
Norden, op. cit. 157 n. 3) is connected with the formula very common in
prayer dvev σοῦ (‘nothing is accomplished’ or ‘this or that is not accomplished’,
so, e.g., in the prayer to Zeus A. Suppl. 823 f. τί δ᾽ ἄνευ σέθεν θνατοῖσι τέλειόν
ἐστιν ;).
παναιτίου comes spontaneously from the poet's particular train of thought ;
the word was never an ἐπίκλησις of Zeus.” Its use as an adjective appears
to be an innovation of Aeschylus, which was not generally adopted until
Byzantine times (the proper name Παναίτιος existed in Athens when
Aeschylus was a boy). It is noteworthy that there is a 'figura sermonis'

1 Cf, Wilamowitz, Reden und Vorträge, i., 4th ed., 320.

2 Pape-Benseler, Wörterbuch d. griech. Eigennamen s.v. is wrong here (the reference to
Cornutus 22 should be deleted). O. Gruppe, A. B. Cook, Fehrle (in his article ‘Zeus’ in
Roscher's Mythol. Lex.) are right in not mentioning the supposed 'epiklesis'. Bechtel,
D. histor. Personennamen des Griechischen 532, was not justified in citing the proper name
Παναίτιος as a name ‘indicating that the person belongs to a god’. I have no idea how this
name is to be explained. May it perhaps have been originally a nickname of derision,
as conceivably Παναίσχης was (IG i.2 9gr. 13, cf. Solmsen-Fraenkel, Indogermanische
Eigennamen [1922] 128), something like ‘guilty of all possible (mischief)’? For Πανοῦργος
as a name I can find no evidence earlier than Cic. p. Rosc. com. 27 ff. ; cf. CIL 1.2 476. 8.

704
 COMMENTARY line 1494
both here (mavarriou mavepyéra) and in the only other place where Aeschylus
uses the word, Eum. 199 f. αὐτὸς σὺ τούτων οὐ μεταίτιος meAnı, ἀλλ᾽ εἷς τὸ πᾶν
ἔπραξας ὧν παναίτιος.
πανεργέτης (only here) presumably formed on the model of the older and
common εὐεργέτης (on the formation cf. Ernst Fraenkel, Nomina agentis,
i. 150).
In the iambics 1485-7 (from i& ij to τελεῖται) the end of a metron always
coincides with the end of a word (so also in the antistrophe except 1510
ἐπιρροαῖσιν). Cf. Wilamowitz (Verskunst, 120), who adduces more examples
and suggests that this feature may have been borrowed from folk poetry.
1488. θεόκραντον : not ‘accomplished or wrought by the gods’, but 'estab-
lished, ordained as valid by God’, like μοιρόκραντος, πυθόκραντος, δημόκραντος,
cf, on 369.
For the thought cf. 5. Trach. 1278 κοὐδὲν τούτων 6 τι μὴ Ζεύς.
1489. Direct address to the dead in Jewish and other ritual is a normal part
of laments of an early date, cf. Hedwig Jahnow (op. cit. on 1322), pp. 50 fff.
and 100,
1490 ff. πῶς σε δακρύσω;. . τί ποτ᾽ εἴπω; On such questions as a normal
part of ritual laments cf. on 1541.
1491. φρενὸς ἐκ φιλίας : cf. on 805.
τί ποτ᾽ εἴπω; "When two questions are asked in succession, the second
amplifying the first, the interrogative introducing the second question . . . is
liable to be postponed' (G. Thomson, C. Q. xxxiii, 1939, 148). Of the instances
quoted by Thomson, Cho. 858 is very similar to the present passage (there
two initial questions are followed by a third: ὑπὸ δ᾽ εὐνοίας πῶς toov εἰποῦσ᾽
ἀνύσωμαι ;).
1493. ἐκπνέων : Tragedy seems to provide no exact parallel for the synizesis
demanded by the metre here;! reference to noun endings -ews, -ewv, etc., is
useless. But the Homeric type of participial forms, written -éwv but pro-
nounced as a monosyllable, was perhaps sufficient justification for Aeschylus.
Hartung and A. Y. Campbell read ékmvesoas, thereby treating the metre
exactly as it is treated by Triclinius (ἐκπνείων), whose dislike for the paroemiac
is well known (cf. on 1334). Itisno uncommon phenomenon that an anapaestic
period which is followed by lines in a different metre? ends without catalexis
(cf., e.g., Pers. 985, E. Med. 132, Hipp. 1378), but the paroemiac is, of course,
unobjectionable. Exception has been taken to ἐκπνέων also on grounds of
syntax: Karsten maintained 'aoristus hic aptior est quam praesens', and
Nägelsbach, too, finds the present tense difficult and resorts to the desperate
expedient of taking ἐκπνέων closely with δαμείς. Presumably the tense of
ἐκπνέων is accounted for by the fact that κεῖμαι serves as passive perfect to
τίθημι (which also explains the singular construction E. Iph. T. 620 eis
ἀνάγκην κείμεθα). Thus: ‘you are in (and have been brought into) this lying
position by being murdered in a wicked manner'.
1494. ' κοίταν ex antecedente κεῖσαι explicandum est.’ This statement of
1 Cf. the summary of the evidence in Kühner-Blass, i. 227. If G. Hermann had known
a second example, he would probably have quoted it, when on S. Ant. 1146 (1132 Hermann)
he quoted A. Ag. 1493 to support πνεόντων as a disyllable.
2 I do not know how Hartung and A. Y. Campbell understand the metre of 1494 (they
delete one μοι).
4872.3 Q | 705
line 1494 COMMENTARY
Schütz has rightly been accepted by most commentators. Passow well com-
pares the passage with Hdt. 1. 1o. 1, 5. 20. 2, ὥρη τῆς κοίτης. ᾿ κοίτη is not
merely the place of lying but also the state of lying’ (Wilamowitz, Platon, ii.
358); cf. on 565 f. “ὦμοι μοι κοίταν τάνδ᾽ ἀνελεύθερον quasi per parenthesim
interjectum est’ (5, Butler), in apposition to the sentence. It is therefore
wrong to follow the many editors (e.g. Hermann and Wilamowitz) who place
a full stop after ἐκπνέων ; what is required is a comma (so in F after 1493, not
after 1517 ; in Tr at both places). The idea of the κοίτα ἀνελεύθερος is elaborated
in 1539 f.
1495 f. Weil (1858) with a fine feeling for language: ‘ é« χερός, sic nude
positum, sive exponis manu, sive comminus [this is hardly possible here,
though it was recently assumed by Lawson, who maintains ‘ ἐκ χερός alone,
“at close quarters", affords a good sense'], valde piget’. Such a translation
as ‘struck down with a two-edged weapon wielded in the hand of treachery’
(Headlam) is not close enough to the text, nor can SoAiw: μόρωι be so con-
nected with ἐκ χερός. A closer definition of χερός is needed as in, e.g., Sept. 805
ἄνδρες τεθνᾶσιν ἐκ χερῶν αὐτοκτόνων, Eum. 102 κατασφαγείσης πρὸς χερῶν
μητροκτόνων, Suppl. 66 αὐτοφόνως ὥλετο πρὸς χειρὸς ἔθεν, S. Trach. 1133 πρὶν ὡς
χρῆν σφ᾽ ἐξ ἐμῆς θανεῖν χερός, Ar. Frogs 1142 (see below). S. 47. 26 f. κατηναρισ-
μένας ἐκ χειρός is unsuitable for comparison since there the emphasis is on the
strange fact that the herds have been killed by a human hand (not by wild
beasts). However, δολίων μόρωι δαμεὶς ἐκ χερός would perhaps be possible, if
ἀμφιτόμωι βελέμνωι did not follow ; but with this the expression is so lame that
it can hardly be considered Aeschylean. Therefore δάμαρτος, which Enger
inserts, is at least very attractive. The identity of the first syllable would
explain its omission after δαμείς (otherwise γυναικός might be an equally good
supplement). This reading has the further advantage of giving the last
sentence of the Chorus a word with which the first sentence of Clytemnestra’s
answer directly connects; that agrees with the arrangement noticeable all
through this ‘kommos’. Perhaps the restoration may also be supported by
adducing a passage which refers to the Oresteia and in particular has close
affinities with Ag. 1495 f. (δολίων μόρωι «rA.): Ar. Frogs 1141 ff. ὡς ὁ πατὴρ
ἀπώλετο αὐτοῦ βιαίως ἐκ γυναικείας χερὸς δόλοις λαθραίοις.2 The fact that in the
MSS the words of 1495 f. recur in exactly the same form in 1510 f. constitutes
no objection against the assumption of a lacuna. It is a habit of copyists to
harmonize antistrophic stanzas: e.g. in Eum. 808 ff. all the errors of 778 ff.
are repeated, including the omission of the second δίκα in 785 = 815. Never-
theless I cannot reach a final decision, particularly as the metre gives no help.
δολίωι μόρωι δαμείς needs no addition to yield a satisfactory metre. It is a
dimeter which may perhaps justifiably be regarded as a variant of the
telesilleion ; cf. Wilamowitz, Verskunst 305 (on Pind. Ol. το. 102 κεῖνον κατὰ
χρόνον after a regular telesilleion) and 317 (on Pind. Isthm. 7. 5 τὸν φέρτατον
θεῶν, after a glyconic, a regular telesilleion and a choriamb. dimeter, i.e. all
dimeters). I take the same view of 5. Oed. R. 1209 θαλαμηπόλωι πεσεῖν (text of
! In the same context Passow quotes E. Hipp. 131 τειρομέναν νοσερᾶι κοίται, where
‘lying, lying there’ seems to make better sense than ‘bed’ (‘in a sick-bed’ L-5). 180 νοσερᾶς
δέμνια κοίτης the former meaning is equally suitable, although not absolutely necessary.
2 Hermann found the style of these words so Aeschylean that he tried to recover from
them a fragment of the lost beginning of the Choephoroe.
706
 COMMENTARY line 1497
the antistrophe uncertain), which is syllable for syllable identical with δολίωι
μόρωι δαμείς. Such a dimeter would be very suitable after the glyconic 1494
(cf. my remark on Pind. Isthm. 7. 5 above). But a trimeter doAiwı μόρωι δαμεὶς
δάμαρτος would apparently be equally good (perhaps in the preceding line too
ὦμοι μοι Should be combined with the following glyconic into a trimeter). A
telesilleion which with a bacchius following makes a trimeter occurs Pind.
N. 4. 32 (= 8, 16 etc.) pélovra τι καὶ παθεῖν ἔοικεν, E. Alc. 437 τὸν ἀνάλιον οἶκον
οἰκετεύοις, 442 πολὺ δὴ πολὺ δὴ γυναῖκ᾽ apiorav, 460 σὺ γὰρ ὦ μόνα ὦ φίλα
γυναικῶν, Med. 648 θανάτωι θανάτωι πάρος δαμείην, Hec. 92] ἐπιδέμνιος ὡς
πέσοιμ᾽ ἐς εὐνάν (cf. Wilamowitz, Verskunst, 548), Iph. T. 1250 ἐπὶ ματέρος
ἀγκάλαισι θρώισκων, [Eur.] Rhesus goo f. ἀπομεμφομένας ἐμοῦ πορευθείς, ἀπὸ δ᾽
ἀντομένου πατρὸς βιαίως, Ar. Birds 1411 τανυσίπτερε ποικίλα yeôot. A trimeter
composed of teles.+spond. is E. Alc. 576 βοσκήμασι σοῖσι συρίζων (cf. Rhein.
Mus. 1xxii, 1918, 341).[teles.-+bacch. also Pind..Wem. 2.2 and fr. 106. t Schr.]
1496. ἀμφιτόμωι βελέμνωι (with the gloss ξίφει in Tr): on the intentional
ambiguity of this expression cf. on 1149 and p. 808.
1497. αὐχεῖς. Some commentators have tried to cling to a seemingly literal
rendering (e.g. Humboldt: 'Vollführt von mir sey, rühmst du, die That’,
Conington: ‘So thou boastest of this as a deed of mine’); but who would
speak of boasting when the boaster and the doer are different persons? If
we substitute gloriari for the somewhat ambiguous modern words, the im-
possibility of such a thought becomes at once clear. Stanley tried to avoid
the awkwardness by translating ‘dicis meum esse hoc opus’, so also Schiitz,
and with minor variations this has remained the vulgate, cf., e.g., Nagelsbach,
‘Laut nennst du diese Tat die meinige’; Verrall, ‘darest thou say this deed
was mine?’; Lewis Campbell, ‘thou declarest this to be my deed’ ; Wilamo-
witz, “wagst du dies Verbrechen mein zu nennen?'; Headlam, 'avouch you
this deed to be mine?' But these attempts at camouflage avail nothing, for
αὐχεῖς cannot mean 'declare loudly' that something has happened or has been
done by somebody else, but only ‘declare loudly’ (‘boast’) that one has
achieved etc. something oneself. In point of fact αὐχεῖν here is not a verbum
dicendi at all. Wellauer, Lex. Aesch., gives opinari as the only meaning for
Aeschylus (7 passages). Dindorf followed him in the Thesaurus, while in his
Lex. Aesch. he gives 'glorior, opinor', i.e. he makes a compromise with the
vulgate as Linwood did before him (‘to say or think confidently’). Except in
the abbreviated expression Eum. 561 an infinitive depends on αὐχεῖν in every
Aeschylean passage, including the papyrus fragment from the Carians
(fr. 99 N.; on the text cf. Wilamowitz, Interpr. 234 f.), where 1. 19 was αὐχεῖξ
δὲ Τρώων ἄστυ πορθήσειν βίαι. The meaning 'boast' is possible here, but the
other is at least as good : 'expects confidently, believes confidently, imagines,

1 Is perhaps E. Hipp. 533 ἴησιν ἐκ χερῶν (between an enhoplion and a reizianum) another
variation of the telesilleion? O. Schroeder's and Wilamowitz's analysis of the end of the
stanza, dim. choriamb. + lecyth. + reiz., is not satisfactory, as Wilamowitz himself admits,
Verskunst, 543. The stanza of Pind, P. 10 (1.6, 12 etc.) ends witho o — vu —o— + HV:
I am inclined to regard this as 2 teles., the second being of the type τὸν φέρτατον θρῶν (see
above) or ἴησιν ἐκ χερῶν (with the analysis of Wilamowitz, Verskunst, 320 f., I disagree on
several points).
2 Subject, as it seems (the words are known to be very corrupt), ‘the flower of Hellas’.
To alter αὐχεῖ into αὐχεῖν with Wilamowitz and make it depend on 17 κλέος (dar) is perhaps
unnecessary, but in 17 ἥκειν seems probable.

797
line 1497 COMMENTARY
fancies’. Only this meaning suits A. Suppl. 329 (326 Wil.), Pers. 741, Prom.
338, 688, Ag. 506, Eum. 561 (‘vermeinen’ Blass ad loc.), E. Alc. 675, Med. 582,
Heraclid. 333, 832 πόσον τιν᾽ αὐχεῖς πάταγον ἀσπίδων βρέμειν (very instructive,
because πόσον αὐχεῖς is clearly the elevated poetic equivalent of the colloquial
πόσον Soxeis,' cf., e.g., Ar. Eccl. 399, also in a messenger speech,? «dme ὁ
δῆμος ἀναβοᾶι πόσον δοκεῖς), Heraclid. 931, Andr. 311, Tro. 770 (‘glauben’
Wilamowitz), El. 939, Hel. 1619 (‘ ηὔχουν = expected’ A. C. Pearson), Bacch.
310 (on which Elmsley remarks: μὴ adxeı mihi μὴ νόμιζε significare videtur"),
always with dependent infinitive. So also in the undoubtedly paratragic use
in Cratinus fr. 1 K. There is no instance of the uncompounded verb in
Sophocles, but he uses éfauyei with dependent infinitive in an exactly cor-
responding meaning Ant. 390 (Jebb wrongly ‘I could have vowed’) and Phil.
869, ἐπαυχεῖν El. 65 (‘almost πέποιθα ' Kaibel ad loc.). The scholiasts had
recognized that in these tragic passages the common meaning of αὐχεῖν
(= κομπάζειν or the like) is inadequate, cf. Schol. A. Eum. 561 (on τὸν οὔποτ᾽
αὐχοῦντα) τὸν μηδέποτε προσδοκήσαντα, Schol. S. Phil. 869 (on ἐξηύχησα)
ἐνόμισα, Schol. E. Med. 582 αὐχῶν' κομπάζων θαρρῶν σεμνυνόμενος κτλ., Schol.
Andr. 311 (onnöxeis) ἔλεγες ἐκόμπαξες ἐνόμιζες. It appears, then, that Aeschylus
used αὐχεῖν with dependent infinitive in a special sense (wecannot say whether
he was the first to do so). The link which connects this usage with the pre-
valent meaning of the word, viz. the notion of θαρρεῖν, can still be clearly
discerned, but the function of the verb has somewhat changed: in several of
these passages αὐχεῖν seems to serve as a verbum sentiendi rather than as a
verbum declarandi. This, however, is only a rough-and-ready description, for
it is not always possible to determine whether the meaning in a particular
context is ‘uttering a confident belief’ or ‘holding a confident belief’ or
whether both may be implied. Euripides adopted and developed the special
Aeschylean usage to a large extent, whereas Sophocles restricted it to com-
pounds. Most of the tragic passages examined above were mustered with
sound judgement by Dindorf in the Thesaurus at the beginning of the article
αὐχέω, which, however, has not had the influence which it deserved (Weck-
lein’s commentaries on plays of Aeschylus and Euripides deal more correctly
than most others with this usage, but he, like the rest, has no word on Ag.
1497). Of the translations known to me J. F. Davies (1868) is sensible: ‘You
fancy this deed to be mine.’ On the punctuation of 1497 cf. on 1498.
1498. The notorious difficulty of ἐπιλεχθῆις has attracted the attention of
most commentators so strongly that they do not express any view on μηδ᾽.
Sidgwick mentions it, but only touches on a minor point: μηδέ, “but not”,
is Epic; in Attic it was always ἀλλὰ μὴ. See Od. 5. 177; 10. 342. It is well
known that Homer, Herodotus, Tragedy, and (to a smaller degree) Aristo-
phanes differ from Attic prose writers in their use of οὐδέ and μηδέ without a
preceding negative clause, and this has been noticed in the grammars, com-
mentaries, and special lexica ; cf. Denniston, Particles, 190 ff. But the passage
before us is unique. As it stands, a connecting μηδέ follows a positive state-
1 Valckenaer observes on Hipp. 446 that Euripides occasionally (Hipp. 446, Hec. 1160, and,
supposing this to be Euripides, [ph. A. 1590) adopted πῶς δοκεῖς ; from colloquial language.
2 This whole speech is strongly reminiscent of Euripides, cf. my dissertation De media et
nova comoedia quaest. sel., 1912, 34 f.
3 Several editors print in Ag. 1498 μὴ δ᾽, The fact that this mode of accenting is often
708
 COMMENTARY line 1498

ment. This seems ungrammatical; at least I have found nothing at all

comparable in Homer, Herodotus,’ Aeschylus, Sophocles, Aristophanes. To
continue with μηδέ after a negative statement is legitimate, cf. Ag. 879, 8.
Trach. 367, Ar. Lys. 990. The passage Hdt. 8. 142. 4 μηδὲ ὑμέας Ἀλέξανδρος ὁ
Μακεδὼν ἀναγνώσηι is only an apparent exception because μηδὲ refers back
beyond what has intervened to the beginning of the speech (142. 1): "Huéas
δὲ ἔπεμψαν “Λακεδαιμόνιοι δεησομένους ὑμέων μήτε νεώτερον ποιέειν μηδὲν κατὰ
τὴν “Ελλάδα μήτε λόγους ἐνδέκεσθαι παρὰ τοῦ βαρβάρου. The μηδέ in our passage
would of course be simple if it were not connective but meant ne . . . quidem.
Verrall explained: ‘do not even suppose that this is I at all’ (so also Mazon:
‘ne crois méme pas que je sois la femme d’Agamemnon’). But that puts a
completely false emphasis on the verb (whatever be the exact meaning of the
rather uncertain ἐπιλεχθῆις, see below). As ἐπιλεχθῆις may be corrupt, we
cannot be certain that the corruption does not extend to μηδ᾽ (this was
assumed by Hermann and apparently also by Wilamowitz, who puts a
dagger before μηδ᾽ and, as his apparatus shows, regards ἐπιλεχθῆις as corrupt).
If, however, μηδ᾽ is sound, Keck’s assumption that something is lost before it
may be right. The uncertainty about the beginning of the next sentence has
the unwelcome consequence that it is impossible to decide whether 1497
αὐχεῖς... ἐμόν is a statement or (as Verrall, Wilamowitz in his translation,
Headlam, Platt make it) a question.
The scholiast’s paraphrase (ZyoÀ. wad. in Tr) reads: μὴ νόμιζε ἐμὲ εἶναι,
φησίν, ἄλοχον τοῦ Ἀγαμέμνονος κτλ. The assumption is clearly that the use
of ἐπιλέγομαι agrees with that known from Herodotus (19 passages in
Powell’s Lex. s.v. 5). Hermann protested against this interpretation, which
was adopted by some modern scholars (e.g. Bothe’s ‘putes’): ‘ ἐπιλεχθῆις non
potest significare cogitaveris: nam hoc ἐπιλέξασθαι dicitur’. Influenced by
found in the MSS (as is the case here) would be no sufficient reason for retaining it in
modern editions. It seems, however, to be held by many Greek scholars that differentiation
between μὴ δέ and μηδέ is desirable. ‘Ea ratio ineunda esse videtur, ubi utriusque voculae
notio manet propria, ita ut “sed non” valeat, ut μὴ δὲ scribamus, in unam autem vocem
coniungamus si idem est ac lat. "neque" vel “ne... quidem" ’, says Ebeling, Lex. Hom. i.
1090 f., where further literature can be found. The carrying out in detail of this rule would
be difficult; editors who, in certain cases, print μὴ δέ are entirely inconsistent in their
procedure, Nor can this attempt at differentiation be based upon the practice of the MSS,
For the medieval scribes have a preference for writing μὴ δὲ without taking into considera-
tion the difference between the two usages. Thus, e.g., (I quote in this brief selection only
instances which I have checked with the help of facsimiles) Ven. A of the Iliad: on one
side Γ 407 μὴ δ᾽ ἔτι, V 574 μὴ δ᾽ ex’ ἀρωγῆι, on the other side (ne . . . quidem) Z 59 μὴ δ᾽ ὃς
φύγοι, K 239 μὴ δ᾽ εἰ B. ἐς Mediceus of Aeschylus: on one side Prom, 327 μὴ 8° ἄγαν, 783 μὴ
δ᾽ ἀτιμάσηις, Sept. 34 μὴ δ᾽ ἐπηλύδων, 238 μὴ δ᾽ ἄγαν, on the other side (ne . . . quidem) Sept.
615 μὴ δὲ προσβαλεῖν; the Sophocles section of the same MS (L): S. 4j. 1011 μὴ δ᾽ εὐτυχοῦντι,
Trach. 1275 μὴ δὲ σὺ. Ravennas (R) and Venetus (V) of Aristophanes: on one side Knights
580 μὴ δ᾽ ἀπεστλ,, 860 μὴ δ᾽ οἰηθῆις, on the other side (ne . . . quidem) Wasps 178 μὴ δε
mapax., Peace 121 μὴ δε ψακὰς (in both δὲ without accent). So also the much later MSS, F
and Tr, of Aeschylus, e.g. on one side Ag. 921 μὴ δ᾽, 1464 μὴ δ᾽ εἰς ἑλένην, 1498 μὴ 8’ ἐπιλ.,
on the other side (ne . . . quidem) Ag. 569 μὴ δ᾽ ἀναστῆναι.
1 The instances of οὐδέ (μηδέ) instead of καὶ (ἀλλὰ) οὐ (μή) in Herodotus have been
examined by Kallenberg, Jahresber. des Philol. Vereins xxiii (Anhang zur Zeitschr. f. d.
Gymnasialwesen, li, 1897), 201 ff.
2 The details of Keck’s conjecture are not worth discussing, but his insertion of a clause
with μή satisfies syntax. For the same reason Lawson writes μὴ δῆτ᾽ αὔχει for αὐχεῖς εἶναι,
without any warrant, but at least he has not slurred over the grammatical difficulty.

709
line 1498 COMMENTARY

this the defenders of ἐπιλεχθῆις have been anxious above all to justify the
voice of the verb, e.g. Paley, whose discussion has been taken over in its
entirety by Verrall and in an abbreviated form by Sidgwick. Paley says:
‘there are several instances of passive aorists used in a deponent sense. So
προσδερχθῆι Prom. 53, διελέχθη Hdt. 3. 51. τ, διαλεχθέντας PL Symp. 174 d,
φρασθείς Hdt. 7. 46. τ, ὑποδεχθείς E. Heraclid. 757, ἐφράσθη E. Hec. 546', etc.
This is impressive, especially when it is remembered that the phenomena
described here are not isolated. Wackernagel, Syntax, i. 139, has illustrated
some of the factors in the early history of Greek which led to this use. He
shows that the aorist in -@yv has its roots in the middle voice and that in the
oldest preserved texts it has often a clearly middle meaning. “The passive
meaning of forms in τ-θην is not of primary importance.’ About the later
development Wackernagel says: ‘the use of -θην forms in the middle and from
deponents increases with every generation; διελέξατο was superseded by
διελέχθη, ἀπεκρίνατο by ἀπεκρίθη. The language of the Hellenistic period
increäsed the use of -On» still further.’ Nevertheless ἐπιλεχθῆις must be viewed
with great scepticism because of the meaning ascribed to it by its defenders,
The sense of ἐπιλέγομαι in the numerous comparable passages in Herodotus
(put together in L-S s.v. III. ı with Ag. 1498) is unambiguous: ‘bei sich
bedenken, überlegen’ (Passow), ‘consider’, ‘take into account’ (this nuance
is found in Hdt. 7. 49. 5 and at the beginning of the next chapter; also, with
a different construction, in 3. 65. 3 and 7. 149. 1). Cf. Hesychius: ἐπιλεγόμενος"
ἐπιλογιζόμενος amd ἐπιλεξάμενος " διαλεγείς (9), ἐνθυμηθείς. This clearly defined
usage gives no support to the μὴ νόμιζε of the scholion on Ag. 1498 or to
translations like ‘do not assume’ (Paley), ‘but do not think’ (Sidgwick), ‘deem
me not to be... .' (L-S, where, however, in the ‘Addenda’ we find the well-
founded warning ‘si vera lectio’), ‘imagine not’ (Verrall), ‘never conceive me
... (Headlam). If the verb has here this meaning, it is an entirely isolated
use (this comes out in L-S in spite of the juxtaposition of the Herodotean
passages). Such a recoining cannot be regarded as impossible but is unlikely
when the abnormal use does not serve to express something bold and highly
imaginative but merely replaces a simple ἡγεῖσθαι, νομίζειν, δοκεῖν, or the
like. I can but subscribe to the verdict of Wilamowitz: ‘nihil commemora-
tione dignum propositum est.’ G. Vossius’ ἐπιλέξηις (adopted by Wecklein
in his edition with commentary) bears no relation to any of the known usages
of ἐπιλέγειν, and as for Hermann's μηκέτι λεχθῆι δ᾽ (adopted recently by A. Y.
Campbell), it looks like a straw clutched by a despairing emender but not
like a living Greek expression. I pass over other suggestions. The worst
difficulty seems to me that I cannot even discover a train of thought into
which the words Ayaueuvoviav εἶναί μ᾽ ἄλοχον will fit and give a satisfactory
sense. What we expect is ‘that Agamempon's wife has done it’.
1499. ’Ayapepvoviav . . , ἄλοχον after the model of y 264 Ayaueuvovénv
ἄλοχον (Wecklein). The form ‘Ayaueuvémos occurs first Pind. P. 11. 20; it is
found, as here, in anapaests Cho. 861. Euripides always uses this form in
lyric metres: Andr. 1034, Iph. T. 1115, Or. 179, 838, [E.] Rhes. 44, 258, and
it occurs also in anapaests, Tro. 139, Iph. T. 170, but 'Ayapeuvóvevs in the
trimeter Ih. T. 1290. For the formation of the suffix see Lobeck on S. Aj.
108 (p. 98 of the 3rd edition) and Wackernagel, Sprachl. Unters. z. Hom. 69.
1500. φανταζόμενος. ‘In early writers only in Pass, . . . become visible,
710
 COMMENTARY line 1501

appear’ (L-S). The meaning and construction in this passage are unique
(rightly Peile) ; ‘ “likening himself to", as Homer uses εἰδόμενος and εἰσάμενος '
(Paley). The comparison with the Homeric εἰδόμενος is helpful, particularly
when it is remembered that the φάντασμα in Aeschylus corresponds to the
Homeric εἴδωλον.
1501. ἀλάστωρ, A critical survey of the evidence from the ancient lexico-
graphers is given by R. Reitzenstein, Der Anfang des Lexikons des Photios,
on p. 70. ı5. On the history and use of the word see, besides the earlier
discussions (e.g. Nägelsbach, Die nachhomerische Theologie, 482 f.) and the
summaries in G. Glotz, La Solidarité de la famille, 61 n. 4, and A. B. Cook,
Zeus, ii. 1098, also the good summary in A. C. Pearson on E. Phoen. 1556 and
Ernst Fraenkel, Nomina agentis, i. 216 f. The meaning ‘spirit of destruction,
spirit of vengeance, evil spirit’! seems to be the older;? it occurs in four
passages of Aeschylus, twice in Sophocles, commonly in Euripides, also in
Hippocrates #. ipfjs νούσου τ (closely related to S. Trach. 1235). In A. Eum.
236 a human slayer is called ἀλάστωρ, and this looks like a bold transference?
(for this and for fr. 92 and fr. 294 cf. Wilamowitz, Interpr. 220 n. 1). Later
the application to a human being occurs in Demosthenes (as a term of dis-
praise), in Menander, etc. In Ag. 1501 and ı509 the ἀλάστωρ shows the
unmitigated features of a daimon. “When the messenger in Persae 353 has
to tell of Themistocles’ trick, he begins ἦρξε. . . τοῦ παντὸς κακοῦ φανεὶς
ἀλάστωρ ἣ κακὸς δαίμων ποθέν. Themistocles is not the daimon but at most
his instrument’ (Wilamowitz, Glaube d. Hell. i. 368). So the Chorus says in
1470 that the daimon of the house exercises its power ἐκ γυναικῶν ; Clytem-
nestra goes considerably farther here: the ἀλάστωρ in her form and shape
killed Agamemnon. Speaking thus, she is not making excuses for herself :*
she means what she says. The deed now appears to her so frightful that, at
least at this moment, she is convinced that only the spirit of vengeance,
Alastor, can have done it: he has maliciously borrowed her shape. In
Euripides (El. 979) Orestes, tortured with doubt whether Apollo could really
have commanded him to kill his mother, asks: dp’ αὔτ᾽ ἀλάστωρ εἶπ᾽ ἀπεικασ-
θεὶς θεῶι ;—five years later, apparently recalling this passage of the Electra,
Euripides makes Orestes say to Apollo (Or. 1668 f.): καίτοι μ᾽ ἐσήιει δεῖμα,
μή τινος κλύων ἀλαστόρων δόξαιμι σὴν κλύειν Gra, Clytemnestra has moved far
t E, Rohde's contention (Kl. Schr. ii. 235 n. 2) ‘the ἀλάστωρ is not essentially different
from the ᾿Ερινύς ᾿ does not reach the root of the matter. Though naturally points of contact
between the two exist, in such passages as À. Pers. 354 (also elsewhere) the difference is clear.
2 But what is the meaning of the Homeric proper name ἀλάστωρ In the picture of a
scene of the ἄθλα ἐπὶ Πελίαι on the Corinthian Amphiaraus krater in Berlin an ‘AAdorwp
takes part in the chariot-race: ‘we have no idea how he came into this saga’ (Hauser in
Furtwängler-Reichhold, Text, iii. 7).
3 But this is not certain. We could judge with greater confidence if we had reached a
better understanding of the various meanings of μιάστωρ in Tragedy and knew something
about the development of those meanings. The brief formulae for ἀλάστωρ, μιάστωρ, and
parallel concepts, which are found in Rohde, Psyche, i, 5th ed., 277 n. 2 and Blass on A. Eum.
177, whet our appetite but are far too general. On the formation of μιάστωρ (which does not
come directly from μιαίνειν, though it is related to μιαένειν and μιαρός) cf. Ernst Fraenkel,
Nomina agentis, ii. 24 and Glotta, xx, 1932, 92.
4 Daube 190 (add his discussion pp. 186 ff.) believes that what Clytemnestra says here
‘is meant to exculpate her and protect her from punishment' ; he also speaks of the ‘sophis-
tic character of the argument’. It seems to me that he has wrongly introduced a legal
point into the passage (particularly 1497) and missed the pathos of tragic suffering.
7II
line 1501 COMMENTARY

away from her deed, which does not seem to belong to her own self any
longer ; this change in her attitude is well illustrated by the fact that she does
not mention, as so often elsewhere, vengeance for Iphigeneia but atonement
for the deed perpetrated by Atreus against the children of Thyestes. That is
really Aegisthus’ justification, not her own. For the moment she feels herself
to be a mere tool of something not herself which has overpowered her.
1502. θοινατῆρος : for the ἃ cf. on 128r.
1503. ἀπέτεισεν. The scholiast has gone completely wrong: ἀπήντησε τὸν
Ἀγαμέμνονα τὸν φόνον Tod Ἀτρέως. Stanley is also wrong: punivit, Hermann
counters: ‘quae non est activi potestas’ and correctly translates: hunc
adultum persolvit ut vindiclam pueris, propter eos mactatum. In all probability
Hermann was also right in taking the construction to be τόνδε ἀπέτεισεν τέλεον
veapots, ἐπιθύσας αὐτόν, whereas editors before and after him put a comma
after ἀπέτεισεν. That the dative need not always denote the immediate
recipient of the payment (as I' 286 τιμὴν δ᾽ Apyeioıs ἀποτινέμεν), but may also
denote the dead man to whom satisfaction is made by it, is clear, e.g., from Hdt.
3. 109. 2 ἡ δὲ θήλεα τίσιν τοιήνδε ἀποτίνει τῶι ἔρσενι (i.e. to the male she has
killed before). The construction by which to a complete clause an explana-
tory participle is added (ἐπιθύσας) has many parallels in Aeschylus, cf., e.g.,
on 893. The punctuation after ἀπέτεισεν does not, however, seem to me im-
possible ('that hath made this man the price, crowning ['having crowned'
would be better] the sacrifice of young tender victims with a victim full-
grown and complete' Headlam).
1507. At first sight it seems difficult to believe that Aeschylus could have
written πῶ here. 'These ablative adverbs occur only in Doric', says Bechtel,
Griech. Dial. ii. 271, with a reference to Apollon. Dysk. De adverb. (Gramm.
Graec. i. 1), p. 190. 17 ff., where it is stated that forms like αὐτῶ for αὐτόθεν, πῶ
for πόθεν, τούτω for τουτόθεν can be found παρὰ Δωριεῦσι. The well-known
statement (Athen. 9, p. 4o2 b) ὅτι «Αἰσχύλος διατρίψας ἐν Σικελίαι πολλαῖς
κέχρηται φωναῖς Σικελικαῖς οὐδὲν θαυμαστόν is scarcely sufficient to justify an
Aeschylean πῶ. The only certain borrowing from Sicilian dialect is ἀσχέδωρος
(fr. 261 N.), where probably Aristophanes of Byzantium established the
Sicilian origin of the word; Aeschylus may have first met the wild boar in
Sicily, in which case he naturally took over the local name for it. Of the
other instances of Sicilian words in Aeschylus collected by W. Aly, De
Aeschyli copia verb. 100 ff., not one is certain,? and except for τί μήν ;—the use
1 Cf. Cretan and Locrian (both sth cent. B.C.) ὦ = ὅθεν, ὅπω = ὁπόθεν ; moreover in the
inscription of the Labyadai at Delphi (about 400 B.c., Schwyzer, Exempla, no. 323; Solmsen-
Fraenkel, Inscr. Gr. sel. no. 49) C 23 ξοίκω = οἴκοθεν (ascertained by Solmsen, Rh. Mus. li,
1896, 303 f., cf. also Wackernagel, Syntax, ii. 222). M. Lejeune, Les adverbes grecs en -Bev, 250,
gives a list of the instances of πῶ and similar adverbs but omits A. Ag. 1507.
2 Lobel, in the Introduction to Pap. Oxy. 2161, notes the unusual proportion of Doric
words and forms in the remains of the Δικτυουλκοί (N.B. a satyr-play) and suggests the
possibility that they are Sicilian. Cf. also on 366. W. B. Stanford (Proc. Royal Irish
Academy, xliv, 1938, 229 ff.) believes himself to be in a position considerably to increase the
number of probable ‘Sicelisms’ in Aeschylus, since he regards as an indication of Sicilian
influence the occurrence in Stesichorus, Ibycus, and Empedocles of words used by Aeschylus.
So we find in his lists such words as ἄμεικτος, πίστωμα, μακραίων, ἄπειρος (in the sense in
which it is used in Ag. 1382), ἁλώσιμος. When Stanford discusses the connexion between
Aeschylus and his contemporary Epicharmus (p. 238), he adduces as one piece of evidence
the tetrameters quoted by Stobaeus 69. 17 Mein. (vol. iv. 529 Hense) τὸ δὲ γαμεῖν xrA,, which,
712
 COMMENTARY line 1507
of which in the tragedians! and Plato was certainly not derived from Sicily
as Dittenberger* believed—Aly's lists contain only nouns and verbs and
nothing comparable with πῶ. Although πώμαλα, hardened as it is into a
formula, occasionally appears in Comedy and the orators, no inference can
be drawn from this for Tragedy. But despite such general considerations it
must be regarded as probable that πῶ πῶ ; is right here because of the mean-
ing. πῶ corresponds to the Attic πόθεν.2 πόθεν is used idiomatically in a
meaning which is admirably suitable here. Apollonios Dyskolos remarks
(op. cit., p. 191. 1, explaining πώμαλα) : τοῦ δὲ πόθεν avveyéorepov καὶ οὕτως
λεγομένου" ‘ ἔγραψας ; πόθεν ; ᾿. Cf. Schol. E. Hec. 613 πόθεν: ἀντὶ τοῦ οὐδαμῶς
καί ἐστιν ἐπίρρημα ἀρνήσεως, Suidas πόθεν: ἀρνητικόν. ἴσον τῶι οὐδαμῶς. In
this sense πόθεν without further addition‘ is used by Aristophanes (Wasps
1145, Frogs 1455, Eccl. 389, 976, fr. 655 K. σὺ δ᾽ ὁμέστιος θεοῖς ; πόθεν ;) and often
in Plato and Demosthenes (on 19. 30 H. Weil comments: ‘Cette formule, un
peu familiére, ne se trouve pas dans les harangues’), also in Euripides (Andr.
83 with the Schol. dvri rod οὐδαμῶς, Hec. 613—for the Schol. see above—El.
657, Phoen. 1620 with the Schol. ἀντὶ τοῦ οὐ Suvardv),5 once (Alc. 781) in the
form πόθεν γάρ; In all these passages the sense is clear: ‘woher denn?’ (to
use a German colloquialism), ‘how can that be?’, ‘impossible’ or the like. It
should also be noticed that with the exception of E. El. 657, in all the instances
from drama πόθεν follows a question, whether one asked by another or, like
that which precedes πῶ in the present passage, the speaker’s own. The sense
‘that cannot be’ perfectly suits the context here. That makes one shy of
alterations. Wilamowitz,® Mazon, A. Y. Campbell follow Auratus in putting
πῶς in the text. This would perhaps be justified on the ground that καὶ πῶς;
without a verb or other addition is used by Aeschylus and others not only in
the sense of a true question (cf. πῶς δή ;) but also to express strong doubt so
that it comes near to an οὐδαμῶς or οὐ δυνατόν, cf. on 549. But it does not
follow that a similar use of πῶς without a particle may be assumed. ποῦ
(Bothe) is not bad in itself, for it is known that this interrogative ‘saepe apud
tragicos non sine indignatione negat’ (Elmsley on E. Heraclid. 371 with ex-
amples, cf. L-S s.v. II). But there appears to be no example of the ‘elliptic’
use of ποῦ like the elliptic use of πόθεν. It seems therefore that in the very
as has long been seen (cf. A. O. F. Lorenz, Epicharmos, 265), have nothing whatever to do
with Epicharmus or any sort of Dorian comedy ; 'they seem to belong to a very late period'
(Wilamowitz, Berliner Klassikertexie, v. 2, p. 124).
1 Aeschylus used it in Suppl. 999 (wrongly altered by Wilamowitz, cf. also above on 672),
i.e. probably before the Sicilian journey.
2 Hermes, xvi, 1881, 334 f. ; Dittenberger's conclusion is accepted by Aly, loc. cit., and
Blass (on A. Eum. 203), but Denniston, Particles, 334, withholds judgement.
3 Only this meaning is well attested; cf., in addition to the discussion in Apollonios
Dyskolos (see above) and the lexica, Kaibel's apparatus on Sophron fr. 125 and A. Adler's
apparatus on Suidas, vol. iv, p. 186 s.v. πώμαλα. The weight of this evidence should dis-
courage any attempt to accept identity of meaning with ποῦ, whether or no the Hesychius
gloss mà ποῦ, ὅθεν, ὁπόθεν. Δωριεῖς is to be emended in accordance with the very attractive
suggestion of H. L. Ahrens, De Graec. ling. dial. ii. 374, n. 12 (mà: πόθεν, önödev).— The
equation of πῶ with ποῦ stands unfortunately at the beginning of the article in L-S, and
Sidgwick on Ag. 1507 gives only this meaning.
4 Cf. P. T. Stevens, C.Q. xxxi, 1937, 183 f.
5 Alc. 95 does not belong here: the scholiast and Paley ad loc. are right.
$ His reference to Prom. 576 has lost its point in Murray's apparatus, since the MSS which
in Prom. 576 have πῶ are altogether left out of account in Murray's edition.

713
line 1507 COMMENTARY
striking πῶ πῶ we possess the genuine reading. It can hardly be doubted that
in Athenian everyday language πόθεν ; as a strong denial was as common in
458 as it was twenty years later (Alcestis), and we may surmise that Aeschylus
had here the colloquial πόθεν in mind, but by using the foreign form and the
duplication wanted to give to the expression the ring of choral language and
also to intensify its effect. He had no need to fear that his hearers might be
mystified. Enough people must have been about in the Athenian market
who said πῶ; for πόθεν; perhaps πώμαλα had already been received into the
speech of the lower classes in Attica.
συλλήπτωρ. For the conception of the god (or daimon) who ‘joins in the
action, lends a hand’, see on 811; a συλλήπτωρ in a murder is mentioned in
Antiphon Tetral. 2. 3. 1o.
Sense makes it necessary to connect πατρόθεν with ἀλάστωρ ; the construc-
tion is supported by Sept. 841 πατρόθεν εὐκταία φάτις and Pind. Ol. 3. 28
ἀνάγκα πατρόθεν. ἷ
πατρόθεν : ‘resulting from the crime of Atreus’ (Paley). The clear expression
must not be blurred (Sidgwick speaks of the ‘ancestral avenger’, Headlam
translates: ‘an avenging ghost of ancestry’); but any idea of an ancestral
curse including Pelops, let alone Tantalus, is foreign to this passage and to
the Oresteia in general (cf. on 1469). Wilamowitz translates: ‘nur Mithelfer
der Untat sind die Sünden der Väter’; the plural in this version is clearly
meant to convey that the sentence πατρόθεν... ἀλάστωρ, while of course
relevant to the particular case, is also formulated purposely as a general
truth: this is unmistakable in the following βιάζεται κτλ,
1509. βιάζεται: Schütz, Nägelsbach, Kennedy, Paley (he translates ‘is forced
onwards’), and others take this as passive, so does Wilamowitz ' "cogitur",
nempe ut pro natura sua agat, sanguinem effundat'. This seems artificial
and the idea not really appropriate to the daimon of bloodshed. Earlier
Wilamowitz himself: 'In Strómen des Verwandtenblutes stürmt einher der
schwarze Mord', more correctly. For the absolute use of βιάζεσθαι cf. the
dictionaries ; besides passages like Prom. roro, S. Aj. 1160, Thucydides should
be especially considered. The meaning in Prom. 1oro 'behave violently, in
unruly fashion’ fits perfectly here. βιᾶται is absolute in Ag. 385, where,
however, the object can be supplied from the context.
ὁμοσπόροις ἐπιρροαῖσιν αἱμάτων : there seems to be better reason to speak
of an 'enallage' of the adjective here than in the instances quoted on 504, in
which the adjective qualifies a single concept expressed by noun plus genitive.
As here, e.g., Eum. 327 f. ματρῶιον ἅγνισμα . . . φόνου (Blass ad loc. is over-
subtle).
émippoatow: there is no evidence for the word before Aeschylus (who uses
only the plural). ém- here clearly indicates the addition of the flowing to
something already present (in this case new murder to the earlier murders) ;
ı In connexions of this kind the sense ‘from . . .' remains clear, nevertheless it must be
remembered that ‘in Homer and the tragedians the personal pronouns ἐμέθεν, σέθεν, &8ev
have a specifically genitive meaning as well as an ablative meaning' (Wackernagel, Syntax,
i. 300). On Cho. 941 θεόθεν εὖ φραδαῖσιν ὡρμημένος Conington remarks: ‘ θεόθεν, prob. with
φραδαῖσιν, being nearly equivalent to θεῶν, as πατρόθεν, Ag. 1507, to πατρός", That goes
perhaps a little too far: it would be more correct here as there to assume a kind of inter-
mediate function between ablative and genitive. Cf. also Lobeck, Pathol. 11, 246, and my
note on 469 f, (Διόθεν).
714
 COMMENTARY line 1512

so also Eum. 694 and fr. 143 N. ἰὼ Kdixe Μύσιαί τ᾽ emippoai.! The poet may
have in his mind the picture of a swelling brook or stream (here of a stream of
blood). That would be very suitable for Ares and would carry on βιάζεται
well. In E 87 ff. Homer speaks of a warrior storming forward in battle: θῦνε
γὰρ ἂμ πεδίον ποταμῶι πλήθοντι ἐοικὼς χειμάρρωι, which no dams hold back,
ὅτ᾽ ἐπιβρίσηι Διὸς ὄμβρος ; the scholia (BT on 88) emphasize τοῦ ποταμοῦ τὴν
βίαν, as (also) Horace (Carm. τ. 2. 14) uses violenter in connexion with the Tiber
in flood.
1511. "Apns. Cf. Cho. 461 (when Agamemnon is avenged) Apns Ἄρει ξυμ-
βαλεῖ, δίκαι δίκα, Eum. 355 δωμάτων... avarpomds: (cf. the context of the Ag.
passage) ὅταν Ἄρης τιθασὸς ὧν φίλον EAqı κτλ, In a passage resembling the
present, S. El, 1384 f., the Chorus sings immediately before the punishing of
Agamemnon’s murderers ἴδεθ᾽ ὅπου προνέμεται τὸ δυσέριστον αἷμα φυσῶν
Ἄρης.
’ ὅποι δὲ καὶ προβαίνων idem est atque ὅποι δ᾽ ἂν καὶ προβῆι ' says Hermann
(also ad Vigerum, p. 770, n. 215; he quotes there Ag. 556 f., but that rests on
a wrong interpretation). It is a pity that neither he nor any of those who
follow him, e.g. Schneidewin? or Nägelsbach, has produced a parallel? One
would like to refer to Ag. 423* εὖτ᾽ ἂν ἐσθλά τις δοκῶν ὁρᾶν, for the relative use
of the conjunction εὖτε might perhaps be compared syntactically with ὅποι;
there is, however, the difference that there the thought provides an excellent
reason for breaking off the sentence. There remains, then, some uncertainty
and Wilamowitz's suggestions, προβαίνει (1885) and προβαίνοι (1914), seem well
worth considering. But this uncertainty should not have tempted anyone
to unearth δίκαν (or δίκας), already rejected by Hermann (‘nihil lucri ex
languidiore sententia: it could not stand between ὅποι and προβαίνων
(cf. the remarks p. 512 concerning the unity of a participial clause, which
must not be disrupted by the intrusion of elements belonging to the
primary clause).
On xaí in a relative clause cf. Denniston, Particles, 295.
1512. πάχνα of the MSS cannot be defended as a nominative. Neither is it
satisfactory to take it with Hermann as πάχνᾳ." Hermann understood
‘nativoro cruori eas exhibebit, scilicet ὁμοσπόρους ἐπιρροὰς αἱμάτων ', which is
impossible. Klausen, who accepted the same reading, took παρέξει as used
without an object absolutely or reflexively ; Peile and others (e.g. Nägelsbach)
have followed this. But Conington rightly objects that this use would be
extremely harsh here and is not supported by the passages quoted; an
examination of the examples of the reflexive use in L-S s.v. A. II. 2 shows,
ı Passow and following him L-S, it is true, separate this passage from the other instances
in Aeschylus under the heading 'stream of a river' (similarly H. Weir Smyth translates
‘ye streams of Mysia") but ‘Mysian tributaries’ seems to me better. This need not mean
the particular tributary Motos, to which the rives mentioned by Strabo 13. 1. 70 referred
the expression, but may mean the tributaries of the Caicus in general. Except for the
Caicus and the minor streams flowing into it there is no waterway of any importance in
Teuthrania, the scene of the drama Μυσοί.
? He refers to 1371, which is completely different.
3 Nor have Kühner-Gerth, ii. 105 ff. ($ 493), anything comparable.
4 I see now that J. Franz (in the notes at the end of his edition of the Oresteia) has done
this.
5 Probably nothing else was meant by πάχνα kovpofiópw (so F). The scribe of F regularly
omits the « subscript of datives in -«x and often of those in -qı and αι,

715
line 1512 COMMENTARY

first, that the aim or purpose of παρέχειν is always clearly indicated (which
would not be the case here) and, secondly, that the meaning which the verb
has in those passages would not suit here. Auratus’ πάχναν kovpofópov! is
much better and probably right. πάχνην παρέχειν as 564 χειμῶνα παρέχειν, and
the use of the verb elsewhere in Aeschylus agrees. van Heusde compared for
πάχνα kovpoßöpos Pind. Isthm. 5. 49 f. ἐν πολυφθόρωι Σαλαμὶς Διὸς ὄμβρωι
ἀναρίθμων ἀνδρῶν xadaldevrı φόνωι and 7. 27 χάλαζαν αἵματος. Wilamowitz,
Berl. Sitzgsb. 1909, 824 n. 4, put Ag. 1534 (1234 is ἃ misprint) beside the
same two passages from Pindar; on Ag. 1512 he remarks ‘pruinam componas
cum imbre 1533’. As Eirene? is κουροτρόφος (Hesiod, Erga 228, E. Bacch. 420),
so is Zrdoıs an ἐχθρὰ κουροτρόφος (Pind. fr. 109 Schr.), and in a very similar
sense Aeschylus calls the god of murder ‘bringer of πάχνη κουροβόρος '. It is
wrong to refer this phrase simply to the children of Thyestes, as was done by
the eighteenth-century commentators and by Schneidewin, Wecklein, Sidg-
wick, Verrall, and others. The thought is rightly expressed by Wilamowitz:
‘und überall, wo er erscheint, wird Wetters Wut die Kinder fressen.” We have
seen that the sentence 1507 f. πατρόθεν δὲ συλλήπτωρ κτλ, is general in form
and that this is continued in the following βιάζεται κτλ. In the clause ὅποι δὲ
κτλ. it is not a particular instance that is considered but the typical result of
Ares’ running amok. When murder rules instead of peace, the children, born
and still unborn, pay with their lives, in a quarrel between the members of a
family as much as in war against external enemies (Suppl. 663 ff., Sept.
348 ff). So the wrath of the Erinyes will manifest itself as λειχὴν ἄφυλλος
ἄτεκνος (Eum. 785). In the mind of the Chorus the curse which the ἀλάστωρ
brings is closely connected with the curse in the train of Ares. Clearly
the particular expression κουροβόρος is influenced by a recollection of the
Thyestes meal: if in future, as the quarrel rages on, the young are destroyed,
the law of exact retaliation will produce effects of an especially horrible
kind in the house of the Atridae, to which the general statement must
be applied.
1521 ff. A. Seidler’s conjecture, De vers. dochm. (1811) 408, made in passing,
that the words οὔτ᾽ ἀνελεύθερον... γενέσθαι were interpolated, has had great
success even with the most recent editors in spite of Wilamowitz. The super-
stition, backed by the authority of Hermann, that the anapaests 1497—1504
should correspond to 1523-9,? has helped. As usual, once the words were
branded as spurious, there was no lack of severe criticism of them: 'frigent
maxime hi versus produntque manum interpretis, rationem reddentis eorum
quae sequuntur' (Hermann), 'versus pedestres, subinepti' (Weil) and the
like. Except for the difficulty of οὔτε (see below) nothing is wrong with the
1 Pauw's πάχναν κουροβόρωι (accepted by Schütz and, e.g., by Plüss) is apparently an
easier alteration (only apparently, for once πάχναν became πάχνα(ι), it easily attracted its
epithet to it), but it gives an impossible sense.
2 Of course not she alone. For 1% κουροτρόφος cf. A. B. Cook, Zeus, iii. 177 n. 1.
3 An early protest against the theory that the 'epirrhematic' anapaests of the actor in a
κομμός were subject to the same strict correspondence as the lyrics of the Chorus was raised
by Klausen, whom Nágelsbach (on 1521) followed. H. Weil, too, in his description of early
tragic κομμοί, very properly said (Études sur le drame antique, 274) : 'tout en conservant au
chant la forme antistrophique, les poétes se permirent d'en affranchir l'épirrhéme, surtout
quand il était en anapestes’. More recently Wilamowitz, Interpr. 74, and Kranz, Hermes,
liv, 1919, 319, showed the fallacy of the conventional doctrine. Nevertheless, it continues
to appeal to editors of Aeschylus.
716
 COMMENTARY lines 1521 ff.
words. The metre is unexceptionable. Some editors, it is true, have rejected
Seidler’s deletion, but unavailingly because they either have not faced up to
οὔτε (e.g. Conington) or have adopted an artificial solution’ (Verrall is fan-
tastic). However, a few others (e.g. Klausen, Enger, Keck) have rightly
stressed that when Clytemnestra, in 1521, takes up the charge made by the
Chorus in 1518 (ἀνελεύθερον), this is in keeping with the structure of this
whole section of the kommos. Against the traditional view that the words
οὔτ᾽... γενέσθαι were added because the interpolator ‘non intellexit ab οὐδὲ
incipere posse orationem’ (Weil), Keck remarks pointedly: ‘an interpolator
would certainly not have begun his botching with an οὔτε to which apparently
no second clause corresponded’. But no one before Wilamowitz gave a
satisfying diagnosis of the text. So far from believing in an interpolation
here (‘quasi ullum interpolatorum anapaestorum exemplum extaret’),? Wila-
mowitz assumes a lacuna after γενέσθαι, in which a second clause beginning
with οὔτε has been lost. This part of Wilamowitz’s hypothesis has great
probability. The loss of several words in anapaests (causing as in 1522 ἃ false
hiatus) has been seen in 794, another loss in 57; for lacunae in anapaests cf.,
e.g., also Sept. 824, 830, Eum. 313,35. Ant. 156. But it is more difficult to form
an idea of the content of the lost words. Wilamowitz restored exempli gratia:
τῶιδε γενέσθαι {τἀνδρὶ πρὸς ἡμῶν οὔτ᾽ οὖν ἀδίκους ἔρραψα τέχνας") οὐδὲ yap
οὗτος κτλ. Apart from minor flaws* this does not express clearly what must
be expected as an answer to the last words of the Chorus. As may be seen
from the detail of ἀνελεύθερον and generally speaking from the structure
of this whole kommos and of similar sections in Aeschylus (cf. 694, top),
1 Lewis Campbell (at the end of his prose translation of the Oresteia, p. 147 n. 46) wants to
write οὐδ᾽ for οὔτ᾽ : but what does it mean? Although it may seem unnecessary, I will give
a special warning that there can be no idea here of the well-known use of οὔτε... οὐδὲ, . .
(cf. Boeckh, Not. crit. on Pind. P. 5. 48 f. [vol. i, p. 476, cf. Schroeder on Isthm. 2. 45],
Matthiä, Griech. Gramm. 1446, Kühner-Gerth, ii. 290, Denniston, Particles, 193, who rightly
retains S. Oed. C. 1141 οὐδ᾽, cf. Hermann ad loc., whereas Jebb, Radermacher, Pearson adopt
Elmsley's οὔτ᾽) : a glance at the examples in poetry and prose shows a relation between the
two clauses quite different from any relation that can be assumed here between 1521 f.
and 1523 f.
2 Wilamowitz is, of course, thinking of the kind of explanatory interpolation which
Hermann, Weil, and others have assumed here, not of anapaestic passages of considerable
length which diasceuasts have put in for later production, like the insertion Sept, 861-73,
detected by Th. Bergk, Griech. Literaturgesch. iii. 304 f. (this insertion is connected with the
extensive interpolation at the end of the play) and the insertions both dactylic and ana-
paestic which Wilamowitz (Hermes, xliv, 1909, 446 f.) has tried to prove in the anapaests of
the first scene of Euripides' Hecuba (73-8, 90-7, 211-15). The intrusion of a gloss such as
S. Ant. 628 is of course quite different.
3 To put in the critical apparatus ‘ τὸν . . , προνέμοντ᾽ Hermann’ is not fair to Hermann,
who afterwards explicitly recanted the error of his youth, defended the plural of the MSS,
and established a lacuna. For the transition from the general plural which indicates a
class of men to the singular see, in addition to the notes of Wecklein and others on this
passage and on Eum. 338 f., Wilamowitz on E. Her. 195 and on E. Jon 1317, and further
the more recent discussion (e.g. P. Maas, 'Textkritik', in Gercke-Norden's Einl. i. d.
Altertumsw. I. Band, 2. Heft, 1927, ὃ 36) of the words of Virgil which are now taken cor-
rectly by most editors: Ecl. 4. 62 f., qui non risere parenti, nec deus hunc mensa, dea nec
dignata cubili est.
* Only because οὔτε cannot follow immediately on γενέσθαι has Wilamowitz inserted as a
stopgap τἀνδρὶ πρὸς ἡμῶν (Clytemnestra uses ὅδε by itself of her husband in this scene 1397,
1446, 1503, οὗτος 1419, 1523). It also seems to me (but this is perhaps subjective) that in the
second οὔτε clause an infinitive dependent on οἶμαι would rather be expected.

717
lines 1521 ff. COMMENTARY

Clytemnestra in her reply should either take up word for word, or should at
least allude to, the utterances of the Chorus by using phrases of closely similar
meaning. Wilamowitz quite rightly demands this! when he says: 'deest
membrum quo δόλιον μόρον defendebat uxor’, so also Kranz, Hermes, liv,
1919, 316 ἢ, 1. Moreover, the continuation οὐδὲ γὰρ οὗτος δολίαν ἄτην οἴκοισιν
ἔθηκε strongly suggests that δόλος or δόλιον must have appeared in the
preceding words, i.e. the second οὔτε clause. Yet Wilamowitz instead of the
obvious δολίας ἔρραψα τέχνας suggests ἀδίκους ἔρρ. τ. (his earlier suggestion was
ἀσεβεῖς Epp. r.). This deviation from the apparently natural thought certainly
did not arise from an oversight but from unwillingness to make Clytemnestra
deny the δόλος of her deed. His concern is easy to understand. Undoubtedly
the murder as planned and carried out by Clytemnestra is in the highest
degree an act of δόλος. Aegisthus, as well as the Chorus, treats this as an
obvious fact (1636): τὸ yap δολῶσαι πρὸς γυναικὸς ἦν σαφῶς, and accordingly
Clytemnestra says herself (Cho. 888): δόλοις ὀλούμεθ᾽, ὥσπερ οὖν ἐκτείναμεν.
Nevertheless, we should not regard it as impossible that the poet here made
Clytemnestra deny the δόλος; but whether he did it, we do not know: in the
text as it stands nothing is said about it. It is also uncertain whether between
the first oùre- clause and the second (the existence of which is probable,
though not certain) there was not yet another short sentence, in which the
queen stated why or in what sense she did not regard Agamemnon’s death as
ἀνελεύθερον. I find no scorn in the surviving first words of her answer;
ironical or scornful οἶμαι would be in parenthesis (as always, from Homer's
time — oio —) ; so also Prom. 968.
1523 f. οὐδὲ yap οὗτος δολίαν ἄτην οἴκοισιν Enke: ‘and he too did not...’
(‘for he too did not . . .'). Parallels for this use of οὐδὲ yap are not very com-
mon but there are enough examples not only in Homer, but also later. Cf.,
e.g., K 251. ds δ᾽ αὕτως Μενέλαον ἔχε τρόμος, οὐδὲ yàp αὐτῶι ὕπνος ἐπὶ βλεφάροισιν
ἐφίζανε (before, 3 f., it was said ἀλλ᾽ οὐκ Arpelönv Ayapéuvova . . . ὕπνος ἔχε
γλυκερός), & 503f. (‘the parents of the dead Ilioneus should mourn him’)
οὐδὲ γὰρ ἡ Προμάχοιο (his death is narrated 476 ff.) δάμαρ ᾿Αλεγηνορίδαο ἀνδρὶ
φίλωι ἐλθόντι γανύσσεται, T 411 f. (οὐδέ τοι ἡμεῖς αἴτιοι... .) οὐδὲ γὰρ ἡμετέρηι
βραδύτητι. . . Tpdes . . . τεύχε᾽ ἕλοντο. Kühner-Gerth, ii. 330 f., quote Pl.
Euthyphr. 13 c καὶ σὺ τοῦτο συγχωρήσαις ἄν, ds... ἀπεργάζηι ;---μὰ AC οὐκ
ἔγωγε---οὐδὲ γὰρ ἐγώ, ὦ Εὐθύφρων, οἶμαί σε τοῦτο λέγειν, Denniston, Particles
195, refers to Xen. Memor. τ. 4. 9 (οὐ γὰρ ὁρῶ τοὺς κυρίους... .)---οὐδὲ γὰρ τὴν
σαυτοῦ σύ γε ψυχὴν ὁρᾶις. Wecklein compared with the use of oëôè. . . οὗτος
here 5, Oed. R. 325 ὡς οὖν μηδ᾽ ἐγὼ ταὐτὸν πάθω.
Since Schütz, with whom Hermann agrees, nearly all editors make the

1 It is remarkable that his translation has the exact verbal correspondence which we miss
in his suggestion for the text : (Chor.) ‘Wie liegst du so schmählich gefällt durch Verrat... .'
(Klyt.) ‘Was ist schmählich denn an seinem Tode? was verräterisch an meinen Künsten ?
Auch das Leid, das er den Seinen antat, war ja nicht Verrat, noch war es Tücke.'
2 Conington (on 1465 in his numbering) speculated how we should understand Clytem-
nestra’s statement that Agamemnon’s death was not ἀνελεύθερον; but there is nothing in
the text to bear out Conington’s linking up of ἀνελευθερία with a special type of murder.
Lewis Campbell (note on his prose translation) : ‘She regards the death of Agamemnon as
the deliberate judicial act of a freeborn woman—therefore not ἀνελεύθερον ' : quite possible
but incapable of proof. It seems to me that xoírav τάνδ᾽ ἀνελεύθερον is immediately intel-
ligible in view of what the Chorus and the audience have before their eyes (cf. 1538 fl.).
718
 COMMENTARY lines 1523 f.

sentence which begins οὐδὲ γὰρ οὗτος into a question ;! Wilamowitz,? however,
does not follow them (on C. G. Haupt and Plüss see below). That οὐδὲ yap...
ἔθηκε is not a question but a statement is clear from the continuation ἀλλ᾽ ἐμὸν
ἐκ τοῦδ᾽ ἔρνος üepÜév,? however one understands the details of the text and the
construction of this clause (see below). Schütz felt only too clearly how in-
appropriate is this aAAd-clause after the question οὐδὲ yàp . . . ἔθηκε; and
therefore translated ἀλλά by enim vero.* Others, instead of being warned by
ἀλλά against the arbitrary punctuation, do not translate the particle at all,
thus, e.g., Paley ('My own dear sapling', etc.) and Mazon. The reason why
Schütz and his followers decided for the interrogative form is that, as Schütz
says, the negation is disturbing, 'cum sensus ac veritas aientem sententiam
postulet'. For the same reason Stanley suggested ὧδε for οὐδὲ, and this has
been adopted more recently by Lawson and A. Y. Campbell. C. G. Haupt and
Plüss tried another method of making the negative harmless without altering
the text or taking the sentence as a question: they say that Clytemnestra
speaks here 'acerbissima ironia' (C. G. Haupt), 'the denial shows sarcastic
irony ; that killing was no treachery nor unworthy of a hero, accordingly this
killing is no treachery but entirely worthy of a hero’. Nowadays there may
not be many readers willing to believe in the magic wand of irony, by which
the commentator converts the sense of a sentence into the exact opposite of
what, to the ordinary man, it seems to say.
All these scholars assume as self-evident that ‘in truth’ (to quote Schütz's
expression) Agamemnon used guile in the sacrifice of his daughter. That is
certainly so in the current form of the story (δόλωι expressly E. Iph. T. 371).
But reference to this form of the tradition is not enough. We should rather
ask whether it is likely that Aeschylus here had in mind the one feature in
the story which is decisive for the king's δόλος, the bringing of Iphigeneia to
the camp at Aulis on the pretext of her intended marriage with Achilles.
Unfortunately this question cannot be answered with certainty. The tradi-
tion on which Aeschylus drew probably included that feature. In Proclus'
excerpt from the Cypria we read this: ὡς ἐπὶ γάμον αὐτὴν AA μεταπεμ-
ψάμενοι θύειν ἐπιχειροῦσιν (cf. [Apollod.] Epit. 3. 22). Modern criticism does
not doubt that this detail of the story belongs to the Κύπρια, cf., e.g., Wilamo-
witz, Hermes, xviii, 1883, 250; Kjellberg, RE ix. 2603 f.; C. Robert, Griech.
Heldensage, 1100, and (with careful hesitation) Bethe, Homer, ii. 2, 2nd ed.,
241; for the reliability (in general) of Proclus’ ὑπόθεσις of the Κύπρια see
Wilamowitz, Die Ilias und Homer 278 n. 2, and F. Jacoby, Berl. Sitzgsb.
1932, 611 n. 4. Full certainty could only be provided by a further piece of
evidence. But even if we ascribe the motif of the pretended marriage to the
Cypria, as in the state of our knowledge we must, it does not in the least
follow that Aeschylus need have made any use of it in the Oresíeia. As has
! Cf. above, p. 461, where Schütz's wrong assumption that the sentence beginning with
οὐδὲ should be taken as a question has found agreement, though not so generally.
2 The misprinting of a full stop instead of a comma after ἔθηκ᾽ is noted at the end of
Wilamowitz's edition.
3 On this point Blomfield's observation is correct: Hermann brushes it away without
argument.
* Similarly the ‘however’ in Headlam's prose translation appears to point to a conscience
not entirely clear.
5 It is unfortunate that we know nothing of the plot of his ᾿Ιφιγένεια, Welcker’s

719
lines 1523 f. COMMENTARY

been pointed out in the commentary on the parodos (p. 99), Aeschylus deliber-
ately omitted the motivation given in the Cypria from his account of the
events that led up to the sacrifice of Iphigenia. It seems to me most unlikely
that, if in this play the treacherous deceit on the part of Agamemnon was
presupposed at all, Clytemnestra would not have exploited it in quite a
different way and to a much larger extent. She might have been expected to
allude to it in unambiguous words and to pillory it in all its baseness. What
we actually read here can be construed as an allusion to that particular
δόλος only if we either punctuate the sentence in a manner incompatible with
its structure, or turn the sense upside down by means of ‘irony’. Either
expedient may be regarded as more audacious than an attempt to understand
the words in their obvious sense and accept the consequences, however strange
they may seem at first sight. Clytemnestra says: Agamemnon did not use
treachery in committing his crime against the house; rather he murdered
his wife’s and his own child (and in this, not in any δόλος, lay his guilt). This
is not an improbable thought for Clytemnestra. If the reader insists on trying
to reconstruct the events which lie behind the actual play (such an attempt
is at least superfluous, if not inadmissible, as far as certain details are con-
cerned which the poet passes over in silence), he may draw the following
conclusions. It is in no way unlikely that, when Calchas had made his demand
for the sacrifice of Iphigenia openly before the assembled host,’ Agamemnon,
as soon as he had taken his decision, carried it out quite openly. How
Iphigenia was brought to Aulis was no concern of the great poet, whose
theme was Agamemnon’s tragedy. So far as anything can be inferred from
what survives in ll. 1521 ff., Clytemnestra denies δόλος on Agamemnon's part.
Here, as elsewhere, she insists on retaliation in its most precise form (cf.
especially 1527-9). Agamemnon must pay for his sins and must pay in
exactly the way in which he sinned, not otherwise and not more. The
negative, ‘he committed no δόλος himself’, paves the way to the positive
formulation of the lex talionis at the end of the anapaests. I regard this
interpretation as highly probable; but it cannot be regarded as certain, first,
because of the lacuna before 1523, and, secondly, because of the elements of
uncertainty in the clause beginning with ἀλλ᾽ ἐμὸν.
1525. ἐμὸν ἐκ τοῦδ᾽ ἔρνος ἀερθέν, Schiitz is correct: ‘progeniem meam ex eo
susceptam’. Later editors have mostly returned to the explanation of Pauw,

reconstruction (D. Aesch. Trilogie 408 ff.) is entirely in the air ; he went so far as to find in the
parodos of the Agamemnon references to the (pretended) project of marrying Iphigenia to
Achilles. That this motif formed part of the action in the ᾿Ιφιγένεια of Sophocles as in
Euripides is clear from fr. 284 N. (= 305 P.). The hypothesis of W. H. Friedrich, Hermes,
Ixx, 1935, 79 ff., 96 ff., that in contrast to the older form of the saga Sophocles introduced the
plan of a marriage which was intended in earnest, need not be discussed here, because even
if it could be proved, it would have no bearing on the Oresteta.
_ 1 In Euripides’ Iphigenia in Aulis the army at first knows nothing of the command of the
goddess which was given to Calchas and by him passed on to Agamemnon and a few trusted
friends. For a poet who believed in realism of detail this was an almost unavoidable con-
sequence of the motif of the pretended project of marriage. If many had known that the
sacrifice of Iphigenia was demanded, the danger was far too great that Clytemnestra too
would hear of it in advance. From the short account in the prologue of Iph. T. (16 ff.) it
does not appear whether there the idea is that Calchas informed Agamemnon in the presence
of other witnesses. How this was arranged in the Cypria is not clear from the summary
(Κάλχαντος δὲ εἰπόντος... καὶ ᾿Ιφιγένειαν κελεύσαντος θύειν τῆι ᾿Αρτέμιδι).
720
 COMMENTARY line 1526
who (rejecting the interpretation ‘sublatum de liberis quos pater pro suis
agnoscit) comments ‘figuratius dictum de ἔρνει quod in altum crescit ex
stipite et ima arbore’. With the addition of Blomfield's reference to Z 56 ὁ
δ᾽ ἀνέδραμεν ἔρνεϊ ἶσος this became communis opinio, e.g. Schneidewin, ‘the
image is continued in ἀερθέν ' ; Wecklein, ‘ ἀερθὲν (ἐκ τοῦδε), which comes to
the same as dvo βλαστόν, agrees with the proper meaning of ἔρνος '. Against
this A. C. Pearson, C.R. xliv, 1930, 55, rightly raised two objections. He first
showed that it was unnecessary to find an image which would fit the original :
meaning of ἔρνος, because ἔρνος — 'child’! is a faded metaphor (cf. 'scion',
German ‘Sprössling’) in which an image need no longer be felt (he might have
referred also to Eum. 661, 666). Secondly he demands that the precise meaning
of ἀείρειν should be considered, and in fact no one has made out a good case
for believing that ἀείρεσθαι could indicate the growth of a tree or a shoot
(Keck had pointed out the difficulty of ἀερθέν and the unsuitability of ἀνέ-
Spapev ἔρνεϊ ἶσος for comparison, but he then turned to one of his wild con-
jectures). Pearson's own suggestion is unfortunate. He translates: 'My
branch begotten of him, her whom I upreared' (without remembering that in
Linwood's lexicon to Aesch. the passage is translated with 'to bring up,
educate’). For the meaning of ἀείρειν which he assumes he has recourse to
A. D. Knox's probable restoration of Herodas 9. 13 {τοῖς τολκεῦσί σ᾽ ἤειρα,
where, however, the continuation is unknown. But the chief objection to
Pearson's interpretation is that he is compelled to ignore the close connexion
of ἀερθέν with ἐκ τοῦδε which most commentators have rightly understood
(on the order of words according to which an intervening noun separates the
participle from the prepositional phrase which belongs to it see H. Schóne,
Hermes, 1x, 1925, 159). depÜév is not ‘upreared’ here but ‘conceived’. It is the
passive of the middle ἀείρεσθαι (cf., e.g., Sept. 505 ἡιρέθη, ‘was chosen’ from
αἱρεῖσθαι; in general cf. Kühner-Gerth, i. 126). The verb is used of a woman
(in exactly the meaning we require here) in Nicander fr. 108 (quoted in
Parthenios 34), where it is said of Korythos, the son of Helen and Paris,
ὅν τε kai ἁρπακτοῖσιν ὑποδμηθεῖσ᾽ ὑμεναίοις Τυνδαρὶς atv’ ἀχέουσα κακὸν (cf.
Ο. Schneider ad loc.) γόνον ἤρατο βούτεω. Ο. Schneider, Nicandrea, 131, has
discussed the meaning οἱ ἤρατο there? and shown ‘ αἴρεσθαι idem esse quod
φέρειν, sive hoc sit φέρειν ἐν γαστέρι sive, quod malim, proferre et procreare’.
Possibly Nicander has carried on the tradition of a late epic use, and presum-
ably Aeschylus, like the Hellenistic epic poet, has in these anapaests borrowed
a word from post-Homeric Epic (cf. on 1465). ἐκ τοῦδε is used here as Cho.
992 (ἐπ᾽ ἀνδρὶ) ἐξ οὗ τέκνων ἤνεγχ᾽ ὑπὸ ζώνην βάρος and elsewhere.
1526. The reading τὴν πολύκλαντόν τ᾽ ᾿Ιφιγένειαν raises two grave objections.
As for the first, Stanley demanded deletion of 7’, then Porson (on E. Med.
1 Cf. also Keyssner, Philol. xcii, 1937, 272 τι. 1.
2 Rejecting the conjecture ὃν τέκεν for ov re καὶ, which nevertheless appears again in
Martini’s apparatus to Parthenios.
3 Schneider was mistaken in quoting as a parallel Aratus 1057 ὅν τινα γὰρ κάλλιστα λοχαίη
σχῖνος ἄρηται and then regarding ἄρηται as coming from αἴρεσθαι, when, as the prosody
shows, it belongs to ἄρνυμαι like the corresponding forms in Homer (on the possibility of
misinterpreting ἄροιτο as belonging to deipay even in the time of Homeric poetry cf. Leaf on
Y 247, Karl Meister, D. homerische Kunstsprache, zo n. 1). The error has been included in
Passow-Crönert’s lexicon s.v. deipw B 1, while Dindorf’s Thesaurus i. 1, p. 1050 D rightly
quotes only the fragment of Nicander (not in L-S).
4872.3 R 721
line 1526 COMMENTARY

822 [826]) remarked in passing: ‘dele inutilem copulam et lege πολυκλαύτην.᾽

Klausen defended re by the assumption that here with the help of the particle
an apposition was connected to the preceding clause. He compares Ag. 124
(where he wrongly keeps 7’), 215, and Suppl. 43 and 62. On Ag. 215, where
two aspects of one and the same thing are put side by side and connected by
τε, see the note on 214 f. ; instances of this kind are quite unsuitable to support
the assumption that ‘and Iphigenia’ could be added in apposition to ‘my
child’. Suppl. 43 and 62 will be treated more fully in a moment. Klausen’s
view was adopted not only by Peile, Paley, Conington, but also by Denniston,
Particles, 502'e, and Murray.! In Denniston’s discussion it looks as if Ag.
1526 were corroborated by numerous parallels. This is not the case. As
regards Ag. 1585 (see the note there) and similar passages, Elmsley’s criticism
has not been sufficiently appreciated by Denniston. Moreover, such juxta-
positions as ‘my father, his brother’ (Ag. 1584 f. πατέρα Θυέστην τὸν ἐμὸν, ὡς
τορῶς φράσαι, αὑτοῦ T’ ἀδελφόν, E. Iph. A. 1153 ‘the sons of Zeus, my brothers’,
and the like) are wholly different from the present passage, because here we
have not an indication of a double relationship but the designation ‘my
child’ is followed by the name. The two remaining instances which Denniston,
like others before him, adduces in support of the MS reading, are A. Suppl.
42 and 62. In Suppl. 42 f. re was struck out by Hermann ;? he rightly says
‘rationi repugnat copula’? (in a prayer re would naturally serve to add the
name of another protective deity, cf. on 509). The interpolation of this τε is
of the type mentioned on 124.4 Finally as regards Suppl. 62, the difficult
! Murray’s reference to Suppl. 42 and 62 hangs in the air; for in 42 he rightly follows
Hermann in deleting re and in 62 he reads with Wilamowitz not only ήτιδος (rightly) but
also κιρκηλάτου 7’ ’Andorns (on which see below) ; in other words he, like Wilamowitz, assumes
that Maris and ’Andovn are two different persons; and if this is so, the one name cannot be
in apposition to the other, consequently the passage provides no parallel for Ag. 1526.
2 Hermann, however, owing to his belief in the necessity of exact syllabic responsion,
was reluctant to admit the correspondence of πόντιον τιμάορ᾽ Inv, — -- τ — o — v with 51
ματρὸς ἀρχαίας τόποις τῶν, -- ὦ -- -- τυ το -- , He surmised therefore that in tw» the final
syllable could be lengthened, which cannot be proved impossible, but is improbable.
Hermann's principle is too rigid. Cf. Prom. 415 f. Κολχίδος re γᾶς ἔνοικοι παρθένοι μάχας
ἄτρεστοι = 420 f. ‘Apaßias τ᾿ Ἄρειον ἄνθος ὑψίκρημνον οἱ πόλισμα (Καυκάσου), 531 βουφόνοις
παρ᾽ = 54I μυρίοις μόχ[θοις, 534 ἀλλά μοι τόδ᾽ ἐμμένοι = 543 ἰδίαι γνώμαι aéBm, Pers. 957 ποῦ
δέ σοι παραστάται = 969 ποῦ δὲ Σευάλκης dvaf, Sept. 833 γένεος Οἰδίπου τ᾽ apa = 841 πατρόθεν
εὐκταία φάτις, Cho. 606 πυρδαῆ τινα πρόνοιαν = 617 χρυσεοδμήτοισιν ὅρμοις. As regards this
last passage, Hermann’s violent alteration (πυρδαῆτιν πρόνοιαν = χρυσοκμήτοισιν ὅρμοις) has
had undeserved success; it stands in the text, e.g., of Weil, Kirchhoff, Wecklein, Sidgwick,
Blass, G. Thomson (without a word in the commentary). Wilamowitz has made the neces-
sary comment on χρυσεοδμήτοισιν in his special edition of this play (cf. also W. Schulze,
Quaest. ep. 39 f.). The note of Blass ad loc. shows the strength of a metrical prejudice when
it is adhered to without fresh examination of the evidence. He says quite rightly that
Hermann’s πυρδαῆτιν is ‘a rather odd form whereas the πυρδαής of the MS is like θεσπιδαής,
and πυρδαῇ rıva φρενομανής τις Ag. 1140, μεγασθενής τις Sept. 977 etc.’. Nevertheless Blass
accepts Hermann’s alteration and gives as his reason that ‘ πυρδαῇ τινα πρόνοιαν = χρυσεο-
δμήτοισιν ὅρμοις is simply wrong because no spondee is admissible in these trochees’. Scholars
who wanted to discredit the MS reading in this passage could at least have been expected to
point out its real singularity, viz. the lengthening of a before the initial mp in τινα πρόνοιαν
(Paley vaguely touches upon it). For a fuller treatment of this point see Appendix E.
3 A very artificial attempt at justification has been made by Bücheler, Rhein. Mus. xli,
1886, 6 (= Kl. Schr. iii. 87).
4 In the passage quoted as an example by Denniston, loc. cit., E. Phoen. 1028 f. ἄλυρον
ἀμφὶ μοῦσαν ὀλομέναν [τ ᾿Ερινύν, the τ᾽ has been rightly deleted by Hartung.
722
 COMMENTARY line 1526

passage must be interpreted before it can be used as evidence. After what

has been written about it,’ I think it very likely that MHTIAOZ is a proper
name. Welcker, D. Aesch. Trilogie, 503 n., thought of this as a possibility,
then Wilamowitz (first on E. Her. 1022) and after him Wecklein (in his com-
mentaries, Athens 1896 and Leipzig 1902) put Myjridos in the text. The order
of words in Typetas Myrıdos . . . ἀλόχου is correct; a selection of similar in-
stances is given by H. Schöne, Hermes, 1x, 1925, 155. In the following words
Wellauer and Hermann? were possibly right in reading κιρκηλάτας for κιρκη-
λάτου τ᾽, so that κιρκηλάτας ἀηδόνος (‘Anddvns I dare not keep; a possible
explanation for the intrusion of the ending -ns into the text is given by
Hermann) is in apposition to Myjridos.* Result : the defence of τ᾿ is unsuccessful.
Secondly we are told by Dindorf and others to scan the name *J¢tyevetav5
although it is elsewhere always "Idıyeveia and although Aeschylus himself
shortly afterwards, 1555 (again in anapaests), has the short a. Meineke, Philol.
xix, 1863, 211, rightly asks: “Where is there an example of such a quantity for
a name in -γένειαϑ᾽ It will be sufficient to recall ἠριγένεια, Τριτογένεια in
Homer or the post-Homeric Kurpoyeveıa. We should therefore be careful not
to take this badly damaged passage as evidence for a linguistic abnormality.
On the other hand, Porson's πολυκλαύτην cannot be called ungrammatical.
The collections in Lobeck, Paralip. 455 ff., show how often the feminine in -n
is found in compound adjectives which have normally two terminations, cf.
also A. C. Pearson on Soph. fr. 394. Of adjectives with πολυ- the feminine of
which sometimes ends in -n, the following have a noun in the second part:
I 568 πολυφόρβην (on the other hand, & 200, 301 πολυφόρβου. . . γαίης),
Anacreon fr. 8o. 2 D. πολυκρότηι, Pind. N. 3. 2 πολυξέναν, fr. 122. 1 Schr.
πολύξεναι; others have a verbal second part (like πολύκλαντος) : Homer's
πολυμνήστη ‘much pursued by suitors' (the same word means ‘much remem-
bering’ in Empedocles fr. 3. 3 Diels-Kranz, cf. above on 821 f.), Pind. fr.
194. 4 Schr. πολυκλείταν, Ar. Peace 978 πολυτιμήτη. In the fourth century
Erinna (Anth. Pal. 7. 712, cf. Wilamowitz, Sappho und Sim. 229) uses πολυ-
kAavrav® without metrical necessity. But none of these instances of the
1 See especially Wecklein ad loc. and O. Schroeder, Hermes, lxi, 1926, 430 ff. Cf. also
W. Buchwald, Studien zur Chronologie d. ait. Tragödie, diss. Königsberg 1939, 35. The treat-
ment of the problem in Vürtheim's commentary (1928) is a step backwards: he takes
Τηρεΐας μήτιδος with the scholiast as said κατὰ mepíópaow τοῦ Τηρέως.
2 Stanley (γ᾽) and Schütz (δ᾽) had already objected to 7’, also Elmsley, who (on E. Med.
807 ff.) says of the passage: ‘Nemo non videt copulam sententiae nocere, quippe quae
significat diversam a Terei uxore esse κιρκήλατον ἀηδόνα ’, Elmsley's κιρκηλάτα was rightly
rejected by Hermann. For the feminine form κιρκηλάτη cf. below on πολυκλαύτην.
3 The inscription (Kretschmer, Vaseninschriften, 129) on the cup of the Magnoncourt
painter, Munich 2638 (Beazley, A.R.V. p. 300, top, no. 1), cannot justify this; the form of
the name there is not clear, above all in Suppl. 62 we do not require a proper name but the
designation of the bird as such.
* Wilamowitz (besides his note ad loc. and his Interpr. 28, cf. also his commentary on
E. Her. 1022) keeps the MS reading and assumes that Metis the wife of Tereus (who is sup-
posed not to be the mother of Itys), and Aedona who was raped by Tereus and became the
mother and murderess of Itys, raise a common lament. This is an intolerable story, and the
structure of the ode is against it; it has been pretty generally rejected.
5 In L-S this quantity is given as though it were established for this passage.
6 This should have been quoted in L-S, where the conjectural reading in Ag. 1526 is
mentioned. (The earliest instance of πολύκλαυτος, Archil. fr. 51 D., col, i. A. 56, is missing
in L-S.)
723
line 1526 COMMENTARY
termination -7 in πολυ compounds belongs to Tragedy.! With the corruption
assumed by Porson, πολύκλαυτόν τ᾽ instead of the unusual πολυκλαύτην, we
may compare the change from κιρκηλάτας to κιρκηλάτου τ᾽, which took place
in Suppl. 62 if the assumption made in the earlier part of this note is correct.
But all the same there is no point in introducing the doubtful conjecture
πολυκλαύτην here, because this would not help the passage in general. It
seems far more likely that down to πολύκλαυτον inclusive the text is sound.
Elmsley (on E. Med. 807 ff. n. b) assumed that ᾿Ιφιγένειαν came in as a gloss,
and then 7’ was added to restore the metre. That is possible, but τὴν πολύ-
KAavrov suggests that the article prepares for the name (which need not
follow immediately after roux). Wilamowitz made the valuable observa-
tion that the sentence which begins with 1525 ἀλλ᾽ ἐμὸν lacks a finite verb
(he suggested the supplement ἔθυσε πατήρ) : the train of thought shows that
in 1527 the words ἄξια (on the reading see below) δράσας «rA., closely con-
nected as they are with what comes after, begin a new sentence, which in the
well-known asyndetic form draws the conclusion from what has gone before.
ἄξια δράσας is as complete as ἄξια πάσχων which it balances; it is therefore not
possible that the accusative épvos . . . ᾿Ιφιγένειαν is governed as an object by
δράσας, as almost all commentators assume. Moreover, it is obvious that
᾿Ιφιγένειαν ἄξια δράσας makes nonsense. How big the gap between πολύ-
κλαυτον and ἄξια is cannot be determined (these anapaests are not in respon-
sion, cf. on 1521 ff.), nor is it certain whether ’Idıyeverav belongs to the original
text.
1527. ἄξια δράσας : Hermann’s correction, which is demanded by the sense
(cf. 533), is confirmed by the metre: the words as they stand produce a
monstrous form of the anapaestic metron, and the same is true of Elmsley’s
suggestion (on E. Med. 807 ff.) τὴν πολύκλαυτον ἀνάξια δράσας, which Verrall
still thought worthy of consideration. The phrase ἄξια δράσας, ἄξια πάσχων is
explained rightly by Hermann: ‘i.e. ἀξια ἀξίων δραμάτων πάσχων ᾿. Wecklein
compares Eum. 435 and in addition Plaut. Poen. 1270 eveniunt digna dignis,
cf. also the beginning of the dedicatory inscription in the temple of Ardea of
the Republican period (Plin. Nat. hist. 35. 115 = Fragm. poet. Lat. ed. Morel,
Ῥ. 32) dignis digna, and the lampoon against Sarmentus (time of Augustus),
schol. Juvenal 5. 3 (Morel, op. cit. p. 112) digna dignis. There is perhaps a
proverb behind this; Meleager of Gadara, Anth. Pal. 12. 132. 13 f. ἄξια πάσχεις
ὧν pas (quoted by van Heusde) may allude to it or may have preserved a
common version of the proverb in the typical form of the paroemiac (cf. on
this Meineke at the end of the 3rd edition of his Theocritus, pp. 454 ff. ; Usener,
Altgriechischer Versbau, 44 ff.). Dicola bound together by the equal length of
the limbs and by anaphora, like ἄξια δράσας, ἄξια πάσχων, are common in the
lyrical parts and anapaests of tragedy from its origins, cf. my Plautinisches
im Plautus, 365. Here as often elsewhere certain stylistic devices of exalted
poetry served as prototypes for what became favourite patterns in the
‘Kunstprosa’ of Gorgias, etc. (cf. on 1541).
1528. μεγαλαυχείτω : the verb in the active only here before Hellenistic times
(the middle in Plato), but peydAavyos in Pers. 533 (and in the line of the
diasceuast Sept. 1054), then Pind. P. 8. 15.
1529. ξιφοδήλητος only here and Cho. 729 (also anap.). On this unambiguous
1 Aeschylus, as is well known, used ἀδμήτα, εὐφιλήταν, περικλύστα, περιρρύτας πόλεις,

724
 COMMENTARY line 1532

mention of the sword as the weapon by which Agamemnon was killed, and on
the particular reason for its mention here (close parallelism between the
crime and the retribution for it), see Appendix B.
ἔρξεν is a necessary correction. That the indicative of the aorist of ἔρδειν
in Aeschylus, as in Homer and Herodotus, is épéa, is clear from Set. 923.
Peile’s, van Heusde’s, Housman’s (J. Phil. xvi, 1888, 284), and Verrall’s
derivation of ἦρξεν from dpyw cannot be considered; what the context here
demands is something completely different from the Euripidean passages
quoted by van Heusde and Housman; moreover, 1564 παθεῖν τὸν ἔρξαντα is
reminiscent of this passage, cf. also 1658 πρὶν παθεῖν Epfavres. The juxta-
position of the two verbs is found in θ 490 ὅσσ᾽ ἔρξαν τ᾽ ἔπαθόν re (the line, it is
true, does not belong to the original stratum—see the discussion referred to
in Ameis-Hentze’s ‘Anhang’ and cf. P. Von der Mühll, RE, Suppl. vii. 718—
but is older than Aeschylus).
1531. It is possible at a pinch to keep the accusative εὐπάλαμον μέριμναν and
make it depend on στερηθείς (to join ἀμηχανῶ μέριμναν, which has also been
tried, is artificial), as, after Herrhann, e.g. Paley (with doubts), Schneidewin,
Sidgwick, and Headlam. The accusative of the thing taken away after the
simple verb orepéw has been duly defended by Karsten, although he does not
believe in this construction here: he quotes 5. El. 959 f. (already in Hermann)
ἦι πάρεστι μὲν στένειν πλούτου πατρώιου κτῆσιν ἐστερημένηιϊ and E. Hel. 95
βίον στερείς. But as the use of στερέω and στέρομαι with the genitive is much
more common, it seems wise with Karsten and Enger to make the slight
alteration (if it can be called alteration) to εὐπαλάμων μεριμνᾶν. This being
accepted, μεριμνᾶν is dependent on φροντίδος. ' φροντίς ᾿ says Wilamowitz
(commentary on A. Cho. 603 ff., p. 214), ‘is "Berechnung, Überlegung” (cal-
culation, consideration)’. μέριμνα here, as elsewhere, stands very near to
φροντίς in meaning ; both denote 'thought' and cover both the act and its
result ; in Pindar both may mean 'poetry' or ‘writing poetry'.
εὐπαλάμων takes up the idea of ἀμηχανῶ ; for παλάμη, παλάμαι have often
the sense of μηχανὴ καὶ τέχνη (Schol. Pind. OL. 9. 26), and παλαμᾶσθαι is
translated by the lexicographers by τεχνάζεσθαι, ἀπάλαμος by ἀμήχανος. The
word εὐπάλαμος occurs in the language of later epic (Phoronis fr. 2 Kinkel) ;
Cratinus fr. 7o K. τέκτονες εὐπαλάμων ὕμνων parodies probably a lyric poet.
In the Ziad, then in Hesiod, then later in the lyric and elegiac poets, we find
dardAap(v)os, but δυσπάλαμος is only found in Aeschylus (cf. on 1571 δύστλητα).
As here εὐπάλαμνον is the MS reading, so ἀπάλαμν. instead of ἀπάλαμ. stands
incorrectly in the older MSS of Hesiod, Erga 2o, Pind. Ol. τ. 59.
1532. ὅπαι τράπωμαι depends on ἀμηχανῶ, cf. (Elmsley on E. Heraclid. 595)
Pers. 458 f. ὥστ᾽ ἀμηχανεῖν ὅποι rpámowro, E. Or. 635 ὅπηι τράπωμαι τῆς τύχης
ἀμηχανῶ, 723 ὅπηι τραπόμενος κτλ. (in both these latter passages the evidence
of the MSS is strongly in favour of ὅπηι. For the confusion of ὅπηι and ὅποι
cf. A. Sept. 659, Prom. 183, Blass and Wilamowitz on Cho. 1021). The same
expression also occurs in Cho. 409 πᾶι τις τράποιτ᾽ ἄν, ὦ Ζεῦ ;
The Chorus, having been reminded of the sacrifice of Iphigenia and the
requital here exacted for it, admit that in regard to this conflict they feel
! It seems impossible to make κτῆσιν depend on στένειν (considered by L-S s.v. στερέω
and by Jebb), if only because of the strict parallelism of πάρεστι μὲν στένειν and πάρεστι δ᾽
ἀλγεῖν.

725
line 1532 COMMENTARY

unable to have recourse to nimble reflection and that all their thought and
trouble of spirit ends in ἀμηχανία. The parodos had drawn the picture of
Agamemnon in a similar conflict, able to weigh only evil against evil, guilt
against guilt, without discerning any way out: βαρεῖα μὲν κὴρ TO μὴ πιθέσθαι,
βαρεῖα δ᾽ ei τέκνον Saigw . . . τί τῶνδ᾽ ἄνευ κακῶν ; and in the same context the
Chorus had professed its belief that man will have to rely on Zeus alone εἰ τὸ
μάταν ἀπὸ φροντίδος ἄχθος χρὴ βαλεῖν ἐτητύμως. There the Elders were medita-
ting on the terrible dilemma to which Agamemnon had been reduced. Here
and in the related passage 1561 (δύσμαχα δ᾽ ἐστὶ κρῖναι) the Chorus is a prey to
doubts over the strait into which, they now feel, Clytemnestra has been
brought: she has indeed committed a crime in murdering her husband, yet
in so doing she has taken revenge for a grievous crime committed against her.
Thus here towards the end of the drama the underlying religious themes of its
beginning emerge once more, here as there clear, profound, and uncom-
promising. The common man may content himself with the wish μὴ εἴην
πτολιπόρθης and hope thus to escape irresolvable conflicts, but the man on
whom God has laid the burden of a great undertaking, seeing in front of him
two ways open, may feel that whichever way he chooses there is no way out
of evil. As the trilogy unfolds we see Orestes confronted by an analogous
but even more terrible antinomy. The predominant role of this ἀμηχανία in
the thought of Aeschylus is well assessed by W. Kranz, Stasimon, 66 f.
1533 f. ὄμβρου κτύπον. .. τὸν αἱματηρόν : cf. on 1512, and also Pind. P. 5.
το, where the poet refers in the words χειμέριον ὄμβρον (taken up again in
120 f. φθινοπωρὶς ἀνέμων χειμερία mvod) to the deep political unrest. δομο-
σφαλῇ (here only) is active. ὄμβρου κτύπον is used like a compound (cf. on
504), hence the accusative in the epithets.
1534. On the authority of the Atticistic doctrine (Moeris, p. 214 Bekker
ψακάς Arrıkot, ψεκάς “Ἕλληνες, Phrynichus praep. soph. p. 128. 9 de Borries
ψακάζειν" διὰ τοῦ à, οὐ διὰ τοῦ €) Blomfield restored here ψακάς, cf. 561. At
1390 the MSS have ψακάδι. The MSS of Aristophanes seem on the whole to
be in favour of regarding ψακ- as the Attic form, although in Clouds 580 V U
have ψεκ-, and in Ach. 115othisis the reading of Suidas. Following J. Schmidt,
Kuhns Zeitschr. xxxii. 357, both Gustav Meyer, Griech. Gramm. 393, and
Bechtel, Griech. Dial. iii. 89, assume that ψακάς was due to assimilation of the
earlier ψεκάς. Kretschmer, ANTIAQPON, Festschrift für J. Wackernagel
(1923), however, thinks the reverse process (i.e. earlier ψακάς became ψεκάς
through dissimilation) equally possible.
ψακὰς δὲ λήγει. The old explanation ‘ guttatim enim sanguine pluere desut
[or rather desinit] (Schütz) is correct. Verrall thinks this impossible: 'In no
language could “the shower is ceasing” stand for “there is falling more than
a shower".' A. Platt, C.R. xi, 1897, 96, subscribes to this and calls the
traditional explanation *downright nonsense'. Verrall's assertion that ψακὰς
λήγει stands for something else does not meet the case. The sentence begins
quite straightforwardly with δέδοικα δ᾽ ὄμβρου κτύπον κτλ. and the added
clause ψακὰς δὲ λήγει develops the thought ; it is therefore obvious that the
ceasing of the ψακάς is simultaneous with the setting in of the heavier and
noisier downpour (κτύπος as in a cloudburst). It is scarcely necessary to
refute Verrall's own arrangement of the text, which wrenches ψακὰς δὲ λήγει
away from the beginning of the sentence and thus from its place in the
726
 COMMENTARY lines 1535 f.

ὄμβρος picture, to link it with what follows in the form of an antithesis, as

if Aeschylus conceived of ψακάζειν and θηγάνειν as belonging to the same
sphere, Like Platt with his ψακὰς δ᾽ ἐπείγει Wilamowitz, too, rejects λήγει;
his conjecture δηλοῖ, as appears from his explanation Interpr. 201 n. 3, is the
product of his view, refuted above on 1428, that there were drops of blood,
visible to the spectators, on Clytemnestra’s brow. He argues: ‘ ads is the
name given to the drop, not the act of dripping, and the thought is intolerable.’
But on opening the Thesaurus we find not only Arist. Meteor. 1. 9, p. 347"1
ὅταν μὲν γὰρ κατὰ μικρὰ φέρηται, ψακάδες, ὅταν δὲ κατὰ μείζω μόρια, ὑετὸς
καλεῖται, but also Hdt. 3. 10. 3 φάσμα Αἰγυπτίοισι μέγιστον δὴ ἐγένετο" ὕσθησαν
γὰρ Θῆβαι αἱ Αἰγύπτιαι (which had never before happened there), οὐ γὰρ δὴ
ὕεται τὰ ἄνω τῆς Αἰγύπτου τὸ παράπαν" ἀλλὰ καὶ τότε ὕσθησαν ai Θῆβαι ψακάδι,
on which Stein’s comment, referring to Ag. 1534, is ‘mit nur tropfenweis
fallendem Regen’.' The successive degrees are common experience; a nice
differentiation, adapted to the English climate, is given by James Beresford
(quoted O.E.D. s.v. ‘drizzle’): ‘A mist which successively becomes a mizzle,
a drizzle, a shower, a rain, a torrent’. Paley recalls the use of ἀστακτί etc.,
E. Iph. T. 1242 τὰν ἀστάκτων ὑδάτων... Παρνάσιον κορυφάν, S. Oed. C. 1250 f.
δι’ ὄμματος ἀστακτὶ λείβων δάκρνον, Apoll. Rh. 3. 805 rà (sc. δάκρυα) δ᾽ ἔρρεεν
ἀσταγὲς οὕτως. Cf. also Hesychius doraxrov οὐ καταστάζον, ἀλλὰ ῥύδην.
In an approaching storm the ceasing of the shower coincides with the onset
of the heavy downpour. On this analogy the Chorus can find no hope for the
future in the fact that for the moment the shedding of blood appears to have
ceased ; quite the contrary. If bloodshed must indeed be avenged by further
bloodshed, then what has happened hitherto cannot be the end. The sacrifice
of Iphigeneia, the murder of Agamemnon, horrible as they are, are but the
prelude to deeds even more dreadful. Hitherto single victims have fallen,
but the Chorus now expects murder answering murder until the whole house
is wiped out; the emphasis is here all laid on the thought twice uttered in
close succession πίτνοντος otkov . . . δομοσφαλῆ.
1535 f. There is no warrant for the scepticism which gives up the end of the
strophe as 'unintelligible' (Kranz, Hermes, liv, 1919, 316; he includes ψακὰς
δὲ Anyeı). The thought is clear, and the uncertainty is confined to two points
where the reading of the MSS is not intact. In the first place the nominative?
δίκη or δίκαϑ is impossible in conjunction with μοῖρα, and the verb has no
object ; in the second θήγει is a syllable short. The subject of the sentence is
quite certainly μοῖρα (or better Moipa). This, as will presently be shown, is in
1 Perhaps it is a reminiscence of this passage of Herodotus when Cassius Dio 51. 17. 4
says, likewise about Egypt, dod τε yàp οὐχ ὅπως ὕδατι, ἔνθα μηδὲ ἐψέκασέ ποτε, ἀλλὰ καὶ
αἰματι.
2 In codices such as F and its fellow MSS the ending -a, -7 may anywhere be taken as
= -e, +9, Cf. on 1512. It should, however, be noted that the reading of Tr is δικᾷ (the error
in Wilamowitz’s apparatus has found its way into the editions of A. Y. Campbell and
Murray). Apparently Triclinius tried to obtain a possible construction by putting in the
dative.
3 In lyric verse our material does not normally allow us to decide with certainty between
the Attic form of the ending and the non-Attic. On this and on the haphazard methods of
bringing this and other similar details of spelling into uniformity see Wilamowitz, Com-
mentary to Aesch. Cho. (1896), p. 147; cf. also his Verskunst, 509 n. 1, where he remarks on
S. Aj. 597 ἁλίπλακτος : ‘I am not able to determine whether the vocalism is correct here or
in 632 χερόπληκτοι,᾽"
727
lines 1535 f. COMMENTARY

keeping with a conception of Moipa which is dominant in Aeschylus; a

further illustration of the kind of imagery used in this connexion is to be
found in Cho. 646 f., compared by Peile to this passage: Aikas δ᾽ ἐρείδεται
πυθμήν προχαλκεύει δ᾽ Alva! baoyavovpyds. Any conjecture which involves
altering μοῖρα, such as Musgrave’s μάχαιραν and the like, should therefore be
disregarded. (Weil in 1858 had pronounced correctly upon this, with the
comment: ' Moípa verum esse etiam Choeph. v. 647 docet’, but in the Teubner
edition he nevertheless read μάχαιραν. A. S. F. Gow, C.Q. viii, 1914, 5 f., also
states the case for keeping μοῖρα, but goes on to propose violent and un-
necessary alterations in words of the text which are indubitably sound.)
Hermann, adducing Hesychius θηγάνει: ὀξύνει ('fortasse ex hoc ipso loco’
Hermann), substituted θηγάνει for the θήγει of the MSS. This restoration has
been widely approved, and with justice ; since the verb is elsewhere always
θήγει, the corruption is easily accounted for, much more easily than if one
reads with Emperius (Opusc. 133) δίκη δ᾽ ἐπ᾽ ἄλλο πρᾶγμα θήγεται βλάβης πρὸς
ἄλλαις θηγάναισι Μοίρας, which is in itself a possible thought. Headlam, while
accepting θήγεται, gives the sentence a different turn: dixa δ᾽ en’ à. mp. θήγεται
βλάβης mp. ἄλλ. θηγ. μοῖρα, taking βλάβης μοῖρα together and translating ‘yet
there are other whetstones whereon destined hurt is being whetted for the
hand of Justice to another end’. This is unsatisfactory.? πρᾶγμα is to be
taken with βλάβης, as Stanley and most of his successors saw. ἐπ᾿ ἄλλο
πρᾶγμα βλάβης is ‘for another deed (or ‘affair’) of harm’ (Conington and Paley).
βλάβη means not ‘revenge’ but ‘harm, damage, outrage’; see Schadewaldt,
Hermes, Ixvii, 1932, 322. The genitive βλάβης is perhaps best taken not as
objective but as qualitative, as in S. Aj. 1004 ὦ δυσθέατον ὄμμα Kal τόλμης
πικρᾶς,3 Oed. R. 532 4 τοσόνδ᾽ ἔχεις τόλμης πρόσωπον κτλ. There τόλμης is
nearly the equivalent οὗ τολμηρόν, as here βλάβης οἱ βλαβερόν. The sentence
cannot be properly understood without a grasp of the complex notions
embraced in Moipa. In a number of passages in which the basic ideas of
Aeschylus’ religious and moral convictions find their clearest expression
Μοῖρα (or Μοῖρα) denotes not Destiny in general but the particular fate
which causes the appropriate penalty to follow inevitably upon every sin.
Moira is the goddess who sees to it that this connexion between cause and
effect, i.e., in the sphere of moral or legal obligations, between debt and
payment, or between guilt and atonement, is safeguarded against any dis-
turbance; in this capacity Moira is nearly related to Erinys, and it is no
accident that Dike, too, is her companion. Cf. Sept. 975 ff. ἰὼ Μοῖρα, Bapv-
δότειρα μογερά, πότνιά τ᾽ Οἰδίπου oxida: μέλαιν᾽ ᾿Ερινύς, À neyaoderns τις el, Cho.
306 ff. ἀλλ᾽ ὦ μεγάλαι Μοῖραι, Διόθεν τῆιδε τελευτᾶν, He τὸ δίκαιον μεταβαΐφει.
ἀντὲ μὲν ἐχθρᾶς γλώσσης ἐχθρὰ γλῶσσα τελείσθω τοὐφειλόμενον πράσσουσα Δίκη
μέγ᾽ ἀυτεῖ κτλ. (to the end of the stanza). There is also much more than a
colourless ‘Fate’ in the Moira of retribution who appears in the last dialogue
between mother and son in Cho. 909 ff.: πατροκτονοῦσα yap ξυνοικήσεις ἐμοΐ ;—
ἡ Μοῖρα τούτων, ὦ τέκνον, παραιτία.----καὶ τόνδε τοίνυν Motp’ ἐπόρσυνεν μόρον.
1 ‘In the tragic poets αἶσα is synonymous with μοῖρα᾽ (Wilamowitz, Glaube d. Hell. i. 358
n. 2, with a special reference to Cho. 647).
2 Headlam had doubts himself: ‘I do not know that this is right’ (note to his prose
translation).
3 Modern commentators, rightly as it seems to me, have rejected the view of Eustathius,
followed by Hermann, according to which we should have to understand ὦ τόλμης πικρᾶς.
728
 COMMENTARY lines 1535 f.

The particular character of this conception emerges very clearly from the
well-known utterance of Prometheus (516) about the Μοῖραι. But in that
passage their peculiar function has been often overlooked and a complete
theological hierarchy worked out instead,! a matter with which Aeschylus
is not concerned at all. The passage must be examined in its context. Pro-
metheus says (514) τέχνη δ᾽ ἀνάγκης ἀσθενεστέρα μακρῶι. This comes to a
simple admission of his own situation: ‘all my ingenuity and inventiveness
are powerless in face of the compulsion to which I am subject (ἀνάγκης here
exactly as in Prometheus’ words 108 ἀνάγκαις ταῖσδ᾽ ἐνέζευγμαι). But as the
sentence has the form of ἃ general statement, it is meant to be valid far
beyond the individual case which the speaker has primarily in mind.? Its
general application is taken up by the leader of the Chorus; in her question
(515) τίς οὖν ἀνάγκης ἐστὶν οἰακοστρόφος; in addition to the meaning ‘com-
pulsion’ (cf. above on 218) ἀνάγκη assumes, as this word always may assume,
the profounder and more comprehensive sense of ‘necessity’. Thus the dis-
cussion penetrates to a deeper level. The answer Μοῖραι τρίμορφοι μνήμονές T’
'Epwies marks a further advance towards the central problem. The next
question and its answer point to a still greater mystery. But when the
Oceanid seeks to lift the last veil and asks ri γὰρ πέπρωται Ζηνὶ πλὴν ἀεὶ
κρατεῖν ; Prometheus breaks off. When dealing with the crucial sentence of
this sequence Μοῖραι τρίμορφοι μνήμονές 7’ ᾿Ερινύες we are not at liberty to
isolate the Μοῖραι, look at them as controllers of a universal destiny,’ and
ignore their connexion with the μνήμονες (cf. on Ag. 155) ᾿Ερινύες, who are
ranked equal with them. The two groups of goddesses have here the same
function. Zeus is ἀσθενέστερος than Moirai and Erinyes in that no more than
any other being can he escape the consequences of his actions (this is one
sense, while not yet the full sense, of οὔκουν ἂν ἐκφύγοι ye τὴν πεπρωμένην).
If Zeus were to beget a son by Thetis, nothing could avert from him the fate
that must ensue; the same law is applicable everywhere. Every deed of
guilt, whatever its cause may be, brings the doer, man or god, to the pass
where he ἀνάγκας ἔδυ λέπαδνον. Aeschylus was not the first to link Moira‘

! Such experiments are rightly rejected by W. Kranz, Siasimon, 46 (see also 278).
G. Thomson, Aeschylus and Athens, 1941, 52, asserts with reference to Prom. 516: ‘Aeschylus
says that in the beginning the world was ruled [italics are mine] by “the threefold Moirai and
the unforgetting Erinys".' Aeschylus says nothing of the kind; for him the whole point is
that the function of the Moirai and Erinyes as asserted by him is valid now and will always
be valid.
2 It seems not to have been noticed that in τέχνη δ᾽ ἀνάγκης ἀσθενεστέρα μακρῶι there
is a polemic against ‘Musaios’, fr, 4 Diels, Vorsokr., ws αἰεὶ τέχνη μέγ᾽ ἀμείνων ἰσχύος ἐστίν.
3 So even Kranz, Siasimon, 46: ‘that the (still young) gods, as everything that exists,
are held in bounds by some higher power’. Wilamowitz, on the other hand, insisted on the
essential point, Interpr. 123: ‘Prometheus says the final word, that the Moirai and Erinyes
stand above even this personal god : the world’s eternal law and retribution, i.e. the moral
order of the world. Zeus, too, will have to submit himself to this order. “Learning by
suffering” is valid also for divine beings’, and 149 f. ‘it is entirely wrong to speak of fatalism
when he puts Moirai and Erinyes above Zeus .... The meaning of the Erinyes is that
guilt and punishment are interconnected as necessarily as cause and effect, they therefore
signify the invincible domination of the moral law (δράσαντι παθεῖν). W. C. Greene, Moira
(Harvard Univ. Press, 1944) 124, too, in discussing Prom. 515 ff., does justice to the con-
junction of the Μοῖραι and the ’Epıvies.
4 There is no need for our purpose to discuss the very difficult problems of text and
interpretation in A. Eum. 961.

729
lines 1535 f. COMMENTARY

with Erinyes and Dike. In the I/liad Agamemnon says when confessing his
ἄτη (T 86 £.) : ἐγὼ δ᾽ οὐκ αἴτιός εἶμι, ἀλλὰ Ζεὺς καὶ Μοῖρα καὶ ἠεροφοῖτις ᾿Ερινύς,
and Hesiod makes the Moirai, like the Horai Eunomia Dike and Eirene,
daughters of Zeus and Themis (Theog. 901 ff., cf. on the subject Pohlenz,
Griech. Tragödie, ii. 30; Wilamowitz, Glaube d. Hell. i. 360 n. 3). But it is
characteristic of Aeschylus that he not merely says in general that Zeus dare
not undo the work of destiny (cf. Pindar, Paean 6. 92 ff. νέφεσσι δὲ χρυσέοις
᾿Ολύμποιο καὶ κορυφαῖσιν ἵζων μόρσιμ᾽ ἀναλύεν Ζεὺς 6 θεῶν σκοπὸς οὐ roAu&)
but links the conception of retribution, embodied in the Erinyes or Dike,
closely with that of destiny, in the case of Zeus as in all else. At the very
end of the Oresteia the song that begins by calling upon the Erinyes, now
appeased and incorporated in a new and just order of things, ends with the
sentence Ζεὺς ὁ πανόπτας οὕτω Μοῖρά re συγκατέβα. Here, then, there appears
once more, in the most significant position, the trinity Ζεὺς καὶ Μοῖρα καὶ
ἠεροφοῖτις ᾿Ερινύς ; but all Homeric thought is far transcended in this notion
of a unified and all-prevailing justice which in the wisdom of reconciliation
triumphs in the end over the tangled fates and frightful misdeeds of human
kind.
This is the background, here only broadly sketched, against which Ag.
1535 f., [Μοῖρα sharpening the weapon of retribution, has to be understood.
The reading of Auratus, which gives δίκην (or δίκαν), cannot, with absolute
certainty, be pronounced right, but everything seems to favour it. δέκην as
object of θηγάνει indicates the underlying idea much more clearly than
μάχαιραν or ἄορ, which some have conjectured. Here, as so often, it is not
easy to give an exact translation without making it too narrow: ‘just cause’,
‘just claim’ (cf. on 813) is perhaps included in the meaning, but so is ‘justice’
and its realization in a just sentence. δίκη is here the weapon of retributive
execution ; Δέκα, personified, is called ξιφηφόρος in E. Bacch. 991 (the parallel
with Ag. 1535 was pointed out by van Heusde; for the further history of this
conception cf. R. Hirzel, Themis Dike und V erwandles, 8o, n. 1, which includes
an account of its representations in art, cf. also C. Robert, Hermes, xxxviii,
1903, 629 f.). For the Oresteia it is particularly important to note that the
same idea underlies Cho. 639 ff. τὸ δ᾽ ἄγχι πλευμόνων ξίφος διανταίαν ὀξυπευκὲς
οὐτᾶι διαὶ Δίκας. What follows (646), Δέκας δ᾽ ἐρείδεται πυθμήν" προχαλκεύει δ᾽
Alcoa φασγανουργός, has already been quoted above. The allusion in Ag.
1535 f., as in the preceding δέδοικα xrÀ., is of course to the inevitable punish-
ment of Clytemnestra and Aegisthus (as Heath pointed out). The repeated
ἄλλο in ἐπ᾽ ἄλλο πρᾶγμα βλάβης and πρὸς ἄλλαις θηγάναισι drives home the
notion of a horror which grows greater as it proceeds on its course, just as in
Cho. 400 ff. (adduced above on 1339 1.) the continuation of bloodshed is empha-
sized in the expressions ἄλλο αἷμα and ᾿Ερινὺν.... ἄτην ἑτέραν ἐπάγουσαν én’ arnt.
1537 ff. For such wishes in Aeschylus as the expression of strong emotions
(cf. Pers. 915 ff., Prom. 152 ff.) see Schadewaldt, Monolog und Selbstgespräch, 48.
1537. ἰὼ γᾶ ya is treated as an interjection, so that the hiatus is legitimate
(Hermann, Elem. doctr. meir. 372 f.) as elsewhere after interjections and
vocatives of an interjectional character, cf. Kühner-Blass, i. 196 f., P. Maas,
Griech. Metrik, § 141. The a, too, is probably due to the fact that this is ‘the
exclamatory ya’ (H. W. Smyth, Harvard Studies in Class. Phil. vii, 1896, 144).
1539. Here is put into words what the spectator has had before his eyes since
730
 COMMENTARY line 1545

1372. The öpoirn is the bath which in Cassandra’s vision 1129 was called by
the more general designation λέβης. Like the whole device by which the
murder is brought about in this play (cf. on 1382), the bath-tub is Homeric,
and the epithet ἀργυρότοιχος recalls the δύ᾽ ἀργυρέας ἀσαμίνθους of ὃ 128. For
a possible etymology of the word ôpoirn, which is not known to occur earlier
than the Oresteta (here and in Cho. 999 and Eum. 633), cf. E. Schwyzer,
Kuhns Zeitschr. \xii, 1935, 199 ff.; for the post-Aeschylean evidence cf.
R. Pfeiffer, C.Q. xxxvii, 1943, 29 n. 6.
1540. The a of δροίτας is not probable in these anapaests (cf. on 1477); H. W.
Smyth, however, defends it (op. cit. [on 1537], p. 159). The form of the word
which in the MSS here appears as χαμεύναν' is uncertain. Most editors read
χαμεύνην, but Wilamowitz following Solmsen gives χάμευναν, on the strength
of the entry in the lists of the Athenian πωληταί (414/13 B.C.) IG i*. 330. 5,
XAMEYNA? παράκολλος (cf. Pollux το. 36). It cannot be determined with any
certainty which form of the word Aeschylus used. More important is the
particular colour which this word gives to the whole expression. χάμευνα,
denoting a poor and mean couch, vividly marks the contrast between the
usual solemn πρόθεσις on a high «Aivn and the unseemly position in which
the body of the great king is seen, huddled in the low and narrow tub. So the
passage brings out again the idea expressed in the phrase xoirav τάνδ᾽
ἀνελεύϑερον (1494).
1541. tis ὁ θάψων viv; κτλ. Questions like these about the correct or desirable
manner of conducting the mourning rites are common in Aeschylean laments ;
so in what follows here 1547 ff., 1490 f. above, Cho. 315 ff., Sept. 737 f. (with
anaphora, as here and in passionate questions of different content, e.g.
Pers. 956 f., 966 ff., Sept. 91, 156, Cho. 338) ; probably we should see in these an
element of ritual mourning. Cf. R. Hölzle, Zum Aufbau der lyr. Partien des
Aisch. (diss. Freiburg i. Br., 1934), 18 n. 27, 42; Kumaniecki, 71. The piling
up of questions (originally, as it seems, addressed to the congregation of
mourners) is a characteristic ingredient in the laments of the ancient Hebrews
and many other peoples, cf. Hedwig Jahnow (op. cit. above on 1322), ox and
130 f. The particular form used here (with predicate consisting of participle
and article) illustrates once more (cf. on 412 and on 1527) the stylistic affinity
between the language of ceremony in the pre-rhetorical period—hieratic
utterance, for instance—and later Greek ‘Kunstprosa’. Cf., e.g., Demosth. 18.
158 ris οὖν 6 ταῦτα συμπαρασκευάσας αὐτῶι ; Tis 6 Tas προφάσεις ταύτας ἐνδούς ;
1544. αὑτῆς (so F Tr correctly ; it is of no consequence, but Wecklein and G.
Thomson should not have printed αὐτῆς without adding a note): cf. on 1141.
ἀποκωκύειν only here. It is the intensifying ἀπο- of ἀποδύρομαι, ἀποικτί-
ζομαι, ἀποιμώζω, ἀποκλαίω, ἀπολοφύρομαι etc., all verbs of a similar com-
plexion.
1545. The correction ψυχῆι τ᾽ is of course necessary.
ἄχαριν χάριν: Cho. 42 χάριν ἀχάριτον (an obvious emendation) to describe

1 The ἡ which in Tr is written above the a of the ending is probably, as usual in this MS,
meant as an explanation and not as a correction of the form in the text.
2 The epigraphists (e.g. A. Wilhelm, Oesterr. Jahresh. vi, 1903, 237; Kirchner in Ditten-
berger, Syll. 102; Hiller v. Gaertringen, JG i2. 330; Tod, Greek Hist. Inscr. 80) and Meister-
hans 119 accent here χαμεῦνα, but χάμευνα (with Solmsen and Wilamowitz) e.g. L-S,
Chantraine, La Formation des noms, 102; E. Schwyzer, Griech. Gramm. i. 476.

731
line 1545 COMMENTARY

something exactly analogous, to wit, Clytemnestra’s offerings to the dead

Agamemnon.
1545 f. ἀντ᾽ ἔργων μεγάλων. Most scholars refer this to the murder of the
king, but it means Agamemnon’s great deeds. See the Addenda.
1547. The easy emendation ἐπιτύμβιον alvov was accepted as obvious by
Stanley, Schütz, Blomfield, and others, before Wellauer had the unfortunate
idea of practising ‘conservative’ textual criticism here. He was followed not
only by Klausen but by Hermann himself, who thought to get over the diffi-
culty by reference to the ‘neutral’ (or ‘reflexive’) use of idrrew A. Suppl. 547.
More recently the MS reading was championed again by Headlam. The
personification of alvos which this requires offended Conington’s fine sense of
the πρέπον, and Platt, too, (J. Phil. xxxii, 1913, 68) declared alvos πονεῖ, as here
used, to be impossible. The real objection, however, (pointed out by S. Butler,
cited in Peile) is that τίς δὲ «rÀ., as Clytemnestra's answer (1552 ff.) shows, is
merely a continuation and development of the preceding question (1541) : τίς
ὁ θάψων νιν ; τίς ὁ θρηνήσων ; Thus in the last question again the subject must
be a person, therefore Vossius’s conjecture is necessary. Compromise solutions
are of no avail, such as Verrall's overbold assumption that αἶνος might be neuter
accusative and ἐπιτύμβιος kept as ‘participial adjective agreeing with ris’ ; in
this point he comes close to Weil who (in the Teubner edition) reads ἐπιτύμβιος
alvov. This is altogether too artificial; moreover, Cho. 335 ἐπιτύμβιος (the
emendation is certain) θρῆνος and 342 θρήνων ἐπιτυμβιδίων show what was for
Aeschylus as for any Greek the natural connexion of ἐπιτύμβιος.
It may look at first sight as though in 1547 ἐπιτύμβιος alvos denoted the
speech at the grave, the oratio funebris as distinct from the sung θρῆνος (1541).
So Paley in his later editions, following Schneidewin, who says: ‘besides
θάψαι and θρηνῆσαι Aesch. mentions also the ἐπιτάφιος λόγος, carrying back
into the heroic age the custom introduced after the time of the Persian wars’.
The most important innovation in the ritual of Athenian funerary celebrations
for those killed in war was the custom of ‘substituting the speech of a citizen
to the assembled people for the sacral ceremony, viz. songs of the mourning
women or θρῆνοι such as those composed by Pindar’ (Wilamowitz, Gr.
Lesebuch, i. 136, cf. Eduard Schwartz, Geschichtswerk des Thukydides, 144 n. 2).
‘It is clear from Thucydides’ account that this practice had been introduced
at any rate some considerable time before 431’ (Classen-Steup on Thuc. 2.
35. 1)—the precise date of the introduction is not known.! The passage in the
Agamemnon does not give a terminus ante quem, since it is highly improbable
that here any spoken address is referred to. S. Butler in Peile is right when
he says: 'Lessum funebrern intellige, qui iam tum heroicis temporibus fre-
quentabatur. Cf. [in connexion with a θρῆνος] Hom. 2 725 sqq. [in a footnote
Peile refers to Cho. 335, 342]. Hinc postea apud Graecos λόγος ἐπιτάφιος, apud
Romanos oratio vel laudatio funebris'. The expression ἐπιτύμβιος θρῆνος in
X We need not discuss here Leo Weber's thesis, set forth in Solon und die Schöpfung der
altischen Grabrede (Frankfurt a. M. 1935), 66 and passim, that ‘the creator of the Attic
funeral speech was Solon'. Hauvette in a very learned and discriminating article (Mélanges
Henri Weil, 1898, 159 ff.) endeavoured to prove that the custom of the ἐπιτάφιος λόγος was
presupposed in the ᾿Βλευσίνιοι of Aeschylus, but he was not able to support his thesis by
any evidence. Quite recently F. Jacoby, Journ. Hell. Stud. lxiv, 1944, 55, has made it
probable that the first Adyos ἐπιτάφιος in Athens was delivered after the catastrophe of
Drabeskos, in 464 B.C.; cf. on 435.

732
 COMMENTARY line 1553

the passage quoted from the Choephoroe is itself an indication that here
ἐπιτύμβιος alvos forms a part of the θρῆνος without being distinguished from
the rest by the mode of its delivery. A further confirmation is the meaning
of the word αἶνος (cf. on 780 and 1482), which in Aeschylus is never ‘speech’
but always ‘praise’, cf. E. Hofmann’s dissertation (cited on 1482), p. 65.
Wilamowitz (‘Wer wird . .. das preisende Grablied singen?’) renders the
sense correctly; the true significance of emır. alvos was also recognized by
M. P. Nilsson, Neue Jahrb. f. d. klass. Altert. xxvii, 1911, 622, and by Fr.
Pfister, Der Reliquienkult im Altertum (Religionsgesch. Vers. u. Vorarb. v. 2,
1912), 554. Leo Weber, op. cit. (see previous footnote) 2, is right when he says
that here ‘ ἐπιτύμβιος αἶνος is a distinct part of the θρῆνος, though integral to
it’; I do not, however, follow him in the use which he makes of the Orestera
for his reconstruction of the history of funeral rites in Athens. Praise of the
dead man is not unnaturally an important element in the funeral laments
of many peoples, cf. Hedwig Jahnow, p. 58 1. of the article cited above
on 1322, and for the Greeks in particular see E. Reiner (cf. above on
1322), 62 f.
ἐπ᾽ ἀνδρὶ θείωι. Usener, Kl. Schr. iv. 203 n. 7, is arbitrary in seeing herein
a ‘reminder of the former divinity of Agamemnon’. θεῖος is here as in Cho.
867 (θεῖος "Opearns) a reminiscence of Homer. Blass on Cho. 867 points this
out, adding: ‘cf. Suppl. 967 δῖε Πελασγῶν. A clearer consciousness of the
distinction between gods and men prevented the tragedians from applying
more often such Homeric epithets as "godlike" to human beings.’
1549. ἰάπτων. Cf. Wilamowitz, Berliner Klassikeriexte, v. 1, p. 35 (on Hesiod
fr. 96. 8o Rz., 3rd ed.) : ' ἰάπτειν had become a γλῶττα as early as the time of
the tragedians, who used the word for a variety of different things (Euripides
avoids it), cf. Hesychius. ‘Quod αἶνος ἰάπτεσθαι idque σὺν δακρύοις dicitur,
habet hoc imaginem ac speciem donorum in sepulcri monumento imposi-
torum' (Nägelsbach). Cf. on 1616 δημορριφεῖς ἀράς. Leo Weber reads too
much into αἶνον ἰάπτων, op. cit. (p. 732, n. 1), 11.
1550. The fine πονήσει has been ousted by a bad conjecture (Platt, J. Phil.
XXXii, 1913, 69). πονεῖν cannot, of course, mean 'to grieve', as Paley and
others hold; it is rather 'to give oneself trouble'. W. Sewell: ‘who the laud
sepulchral . . . in truth of heart shall toil to pour ?’; so, too, Kennedy (‘shall
labour’) and L. Campbell. For the properly conducted θρῆνος (cf. on 1547) it
is essential thatit shall not merely express genuine sorrow (σὺν δακρύοις,
ἀληθείαι φρενῶν) but be carefully elaborated and given a dignified form as
befits the memory of a great king. πόμος indicates careful, toilsome applica-
tion and its result; so in Aesch. fr. 357 N. a building is ὑψηλὸς τεκτόνων πόνος,
in Pind. P. 6. 54 the honeycomb is μελισσᾶν τρητὸς πόνος, and later 'So-and-so's
poetry’ is πόνος τοῦ δεῖνα (Asklepiades, Anth. Pal. 7. 11. x, Callim. Epigr.
6. 1 Wil).
1551. The epic dAéyew, which finds its way into lyric, is so far as we know
confined to Aeschylus (cf. Suppl. 752) among the tragedians.
1552f. The plural in ἡμῶν and καταθάψομεν of course refers to Clytemnestra
only, in accordance with the familiar usage (cf. on 1279 f.). Blass on Cho. 428
compares Cho. 673 and 716 ff., which are also spoken by Clytemnestra.
1553. No one will now be impressed by Porson's and Elmsley's objections to
this magnificent line. The introduction of what looks like a dactylic sequence
733
line 1553 COMMENTARY

into an anapaestic system is quite legitimate, though remarkable (cf. G. Her-

mann, Elem. doctr. metr. 377, who cites also the yet longer sequence in E. Hipp.
1361 f.). It is possibly not a mere accident that in this line which could be
read as a dactylic tetrameter there occur the Homeric word-forms κάππεσε
κάτθανε, in which the absence of augment, unusual in anapaests, may be
regarded as an epicism. For the form of the dicolon κάππεσε κάτθανε cf. on
1527.
1554. The epic word «Aavduos occurs in Tragedy only here.
On τῶν ἐξ οἴκων Schneidewin's gloss is ‘ τῶν οἰκετῶν, οἰκείων ᾿. Wilamowitz
translates ‘das Gesinde soll ihn nicht begleiten’ (‘the household servants shall
not accompany him forth’), by which he seems to narrow the sense unduly:
obviously all members of the household, including kindred, are meant. Some
idea of the large numbers often participating in these funeral ceremonies can
be gathered from vase-paintings and also from the sumptuary laws to restrict
funeral display (cf. Dittenberger, Sy. 1218 and the evidence collected there
from literary and epigraphical sources.).
1555. In , Tragedy the Homeric adverb ἀσπασίως occurs only here, the adjec-
tive ἀσπάσιος nowhere. It seems likely that the word is meant to intensity
the idea of ἀσπάζεσθαι as expressed i in 1559.
1554 ff. As far as the sense is concerned, nothing points to a lacuna here (for
further detail see below). On the fallacy of believing that there should be
strophic responsion in these anapaests and the consequent assumption of a
lacuna here see on 1521 ff.
1557 f. πρὸς... πόρθμευμ᾽ ἀχέων. Schütz aptly cites the long fragment of
Apollodorus of Athens περὶ θεῶν (F Gr Hist 244 F 102 Jac.) preserved in Stob.
Anthol. 1. 49. 5o, p. 418. 8 ff. W., where the great scholar quotes in support of
the etymology Ἀχέρων διὰ rà ἄχη some lines from the dithyrambic poets
Melanippides and Licymnius; one of the latter's expressions (fr. 2) Axepwv
ἄχεα πορθμεύει βροτοῖσιν is strongly reminiscent of this passage.
πόρθμευμα, which occurs only here, scarcely differs in meaning from πορθμός.
For Aeschylus' tendency to substitute for the root-word a rare form of verbal
noun in -μα cf. Kumaniecki, 18 f. ; he quotes in addition to this passage: Pers.
745 δεσμώμασιν (-- δεσμοῖς), Eum. 753 τἀρίθμημα, fr. 184 N. σικυφώματα
(= σκύφοι), Pers. 71 ὅδισμα, 232 φθινασμάτων. To these could be added (I have
not aimed at completeness), e.g., Sept. 413 ῥίζωμα, Ag. 873 μόρφωμα, fr. 315 N.
τέκνωμα, used in the place of simple nouns, whereas e.g. ὁρκώματα could
perhaps in Eum. 768, but not in 486, be replaced by ὅρκοι, nor could in Cho.
1003 δόλος take the place of 5éAwya.! Sophocles uses, e.g., Ant. 1220 τύμβευμα
in a sense not different from that of τύμβος, and he has ῥοίβδημα for poißdos
and ψόφημα for ψόφος in satyr-plays, cf. R. Pfeiffer, Sitzgsb. Bayer. Akad.
Phil.-hist. Abt. 1938, Heft ii, 38. On Empedocles Wilamowitz says, Berl.
Sitzgsb. 1929, 657 n. 1: ‘Characteristic of this author are coinages such as
κρούνωμα ῥίζωμα [cf. Aesch.] τράνωμα, where the corresponding verb either
does not exist or is not relevant to the meaning of the noun.’
1559. Is Porson’s emendation χεῖρε justified? No objection can be raised on
the grounds that χεῖρε, as was shown by E. Hasse, Der Dualis im Attischen
The differences in the meaning of the artificial coinages of the μόρφωμα type are not
always fully appreciated in the very useful lists which Chantraine, La Formation des noms,
184 ff., gives for the whole of Tragedy.

734
 COMMENTARY line 1559

(Hanover 1893) 47, is an extremely rare form in the tragedians.! Hasse

himself says correctly of E. Andr. 115 that ‘the hexameter ending in περὶ
χεῖρε βαλοῦσα is clearly a reminiscence of περὲ χεῖρε BaAóvre Od. A 211’, The
same passage of the Nekyia could equally well be claimed as the model for
the words περὶ χεῖρε βαλοῦσα φιλήσει with their parallel rhythm, especially
since in A 211 the sentence ὄφρα καὶ εἰν Aidao φίλας περὶ χεῖρε βαλόντε κτλ. gives
a situation very like that of Ag. ı559. There exists, however, no conclusive
argument in favour of Porson’s alteration. Blomfield does indeed say, with
reference to E. Suppl. 272 Bá0. καὶ ἀντίασον γονάτων ἔπι χεῖρα βαλοῦσα, ‘quod
recte quidem dicitur cum praepositione ἐπί, sed περί pluralem vel dualem
postulat’, citing in evidence E. Andry. 115 (v. supra). But Schneidewin
quotes in favour of χεῖρα Simonides fr. 13. 6 D. ἀμφί τε Περσεῖ βάλεν φίλαν xép'
εἶπέ re and E. Iph. T. 799 ἀθίκτοις περιβαλὼν πέπλοις yépa.? The safest course
is therefore to keep «epi χεῖρα βαλοῦσα here (so also Kranz, Hermes, liv, 1919,
316 n. 2) without ruling out the possibility that what Aeschylus wrote
was χεῖρε.
The sense of Clytemnestra's words (1553-9) is obvious. Agamemnon is to
have no mourning ceremony ; contrary to all tradition the members of his
household will absent themselves from his funeral, which is certainly to
include no πρόθεσις and no solemn ἐκῴφορά, but will be limited to mere inter-
ment. But, she continues, the marks of honour and affection which normally
find expression in the funeral rites will not after all be denied to Agamemnon:
in the realm of the dead his daughter will meet him and embrace him
lovingly. The mockery is horrible just because nothing is said of how and
through whom Iphigenia lost her life. It is mockery directed against a father
who is guilty of his daughter's death, but the belief which is the point of the
jest is real enough: in normal circumstances a daughter who died first would
have a loving greeting for her father when he reached the underworld. For
the belief in such a greeting in Hades by friends or relations we possess a piece
of evidence from Athens which is only fifteen to twenty years later than
the Oresteia.? On the ‘Nekyia krater' in New York (Richter and Hall, Red-
figured Athenian Vases in the Metropolitan Museum, pls. 135-7), which
P. Jacobsthal in his editio princeps of the vase (Metrop. Mus. Studies, v. x,
1934, 117 ff.) has admirably interpreted as an important record ‘of what
Periclean Athens believed about the future life’, we see, side by side with
heroic scenes from Hades depicting Ajax, Palamedes, Theseus, Peirithous,
and others, situations taken from the life (or rather death) of the common
people. One of these latter groups concerns us particularly here: a young
woman who has just reached the kingdom of the dead* is here greeted with

N I E. Bacch. 615 χεῖρε (accepted in most of the more recent editions) is a conjecture by
auck.
2 Wilamowitz's conjecture E. Bacch. 1163, adopted into the text by Bruhn and Murray,
must be excluded from consideration, as being completely uncertain.
3 Instances from Hellenistic and Imperial times are given by C. B. Welles, Harvard
Theol. Review, xxxiv, 1941, 86.
4 Jacobsthal (p. 130) out of excessive caution gives too indefinite an interpretation of
this scene, That the young woman has just arrived in Hades can be deduced not only from
the fact that she wears round her chin the band with which the lower jaw of a dead person
is tied for the πρόθεσις (so rightly P. Friedlander, Arch. Anz. 1, 1935, 21), but from the general
consideration that here, as in the heroic scenes (Herakles led by Hermes to the place where

735
line 1559 COMMENTARY

solemn δεξίωσις by a youth. This δεξιοῦσθαι indicates, as it were, the positive

degree; in the case of the first greeting of a loved father by a daughter who
died before him the superlative is in place: περὶ χεῖρα βαλοῦσα φιλήσει. Thus
would Agamemnon be welcomed in Hades— were he not the murderer of his
child.!
1560. ὄνειδος . . . ἀντ᾽ ὀνείδους. The Chorus draw the parallel between the
last sentences of Clytemnestra and their own grave charges (1542 ff.) which
have been answered by the queen. There is nothing in this passage to justify
the interpretation of Peile, Conington, and Schneidewin. Conington says:
“ὄνειδος refers not to the scoff just uttered by Clytaemnestra, but to the
murder of Agamemnon, as the context shows. It will then mean “res quae
exprobrari potest", a thing to be mentioned for blame, a sense which it bears
in many other places', etc. Conington's assertion about the meaning of
ὄνειδος is not borne out by any evidence ; his interpretation of S. Oed. R. 523
is wrong. The opening remark of the Chorus, ὄνειδος ἥκει xrÀ., provides no
more than a formal link ; only from the following sentence does it appear that
now the old men do to a certain extent regard Agamemnon as also guilty.
The words of the indignant mother have not been spoken in vain: the Elders
are forced to admit in veiled, yet unmistakable, terms that it is not possible
for them any longer to throw all the blame on Clytemnestra and that they
find themselves faced with an insoluble dilemma. This, in a way, corresponds
to the confession made in the stanza 1530 ff., whereas in the subsequent
anapaests, 1537 fL, compassion for the murdered king sweeps aside all
antagonistic feeling. Such changes are natural in a conflict of such gravity.
1561. δύσμαχα δ᾽ ἐστὶ κρῖναι, since there is clear guilt on both sides, and
both have acted under a fatal compulsion. The Aeschylean character, when
faced with such a conflict, cannot emerge from the state of ἀμηχανία, cf. on
1532.
1562. φέροντα, 'him who carries away, plunders,' without the addition of an
object, as, for example, in the ancient formula dyew xai (or ἢ, Law of Gortyn,
Inscr. Graec. ad inlustr. dial. sel. ed. Solmsen-Fraenkel, 4o. 5. 37, Demosth.
23. 60) φέρειν. Peile compares Sept. 352 (ἁρπαγαί precedes) ξυμβολεῖ φέρων
φέροντι.
1563. There are no grounds for altering μίμνει to μένει with Heimsoeth,
Wecklein (in his commentary), Wilamowitz, and others: for the responsion
cf. on 210; moreover, it is clear from Pers. 805 and Sept. 34 that Aeschylus
uses μέμνειν without metrical necessity.?
θρόνωι, The same corruption as here is found in Suppl. 374, Eum. 18.
μίμνει δὲ κτλ. has often been compared with Eum. 381, where at the begin-
Theseus and Peirithous are sitting; the drowned Palamedes arriving utterly weary at
Persephone’s palace), a significant moment must be represented, therefore, as in fact the
δεξίωσις suggests, the arrival of the dead woman, not a chance meeting at some later time.
An additional argument may be seen in the statement in the Epitaphios of Hyperides (6. 35,
p. 110 Jensen), which Jacobsthal rightly compares, that the heroes will greet Leosthenes in
Hades with öe£ioöodaı—clearly on his arrival.
ı I cannot find the faintest support in the text, or the trend of thought here, for what
G. Thomson says on Ag. 1554-8 (with reference to Pl. Phaedo 114 a, b) about the things that
happen in Hades when murderers meet their victims and about ‘a transference to the other
world of the primitive conception of αἴδεσις᾽.
2 For a discussion of the supposed difference in meaning between μένω and μέμνω see
Schwyzer, Griech. Gramm. i. 690.

736
 COMMENTARY line 1567

ning of the stanza the sentence opens in the same way and the thought is
similar: μένει γάρ" εὐμήχανοι δὲ καὶ τέλειοι κακῶν TE μνήμονες Zeuvai. The
notion of an unending perspective of retributive justice is there compressed
into a phrase of lapidary brevity. μένειν (μίμνειν) is used by Aeschylus several
times also in the similar context of the perpetual and abiding manifestation
of guilt and its consequences, οὗ, 154 (where μέμνει again stands emphatically
at the beginning of the sentence) and the passages there quoted, and also
Suppl. 385 f. μένει τοι Ζηνὸς ἱκταίου κότος, δυσπαράθελκτος (cf. above on 68 1.)
παθόντος otkrow. [Cf. the Addenda.]
The rule of Zeus is for Aeschylus synonymous with the rule of justice and
just retribution. The agreement with the Zeus-hymn in the parodos is here
particularly noticeable (παθεῖν τὸν ép£avra).
1564. τὸν épéavra: cf. on 1529. For the thought cf. Hesiod fr. 174 Rz. εἴ κε
πάθοι τά τ᾽ ἔρεξε, δίκη κ᾿ ἰθεῖα γένοιτο, A. Cho. 313 and similar passages.
θέσμιον: here we can see clearly that θεσμός means in the first place ‘an
abiding order’, ‘a basic institution’ (R. Hirzel, Themis, Dike etc. 326, 335,
338). The θέντα in 178, referring to the same thing, corresponds to θέσμιον
here (Kranz, Hermes, liv, 1919, 317).
1565 f. ‘dpatov . . . and πρὸς dra . .. are both so indisputable that it is
hardly requisite to particularize them as departures from the reading of the
MSS' (Conington). No time need be wasted in rebutting the attempt of
Emperius, Opusc. 133, to justify προσάψαι as an epexegetic infinitive (so
Nägelsbach, Verrall, Plüss).
κεκόλληται... πρὸς ἄται is ἃ drastic intensification of the Homeric
expression in B 111 (= 7 18) Ζεύς pe . . . ἄτηι ἐνέδησε βαρείηι, which (as Reisig
saw) Sophocles takes over word for word in Oed. C. 525 f. Beazley surmises
that in κεκόλληται Aeschylus had in mind bird-lime (cf. on 1316).
The words of Clytemnestra 1567 ff. are a direct reply to 1563-6; it is there-
fore quite wrong to interpolate after 1566 a repetition of 1537-50. This was
pointed out by Hermann (on 1411 in his edition), and afterwards by, e.g.,
R. Arnoldt, Der Chor im Ag. des Aesch. 79, and Kranz, Hermes, liv, 1919, 317 ;
cf. also Pohlenz, D. griech. Tragödie, ii. 32 f. Even in the year 1935 Mazon
perpetuated the old error.
1567. évéBns: hardly anyone except Sidgwick and Verrall has doubted that
the addition of-the single letter lost by haplography restores the correct
form of the original text; Hermann's carefully weighed pronouncement is
particularly good. The expression is in general parallel to that of 1475. There
Clytemnestra, at the beginning of her reply to the charges of the Elders,
recognizes the justice of what they have just said, whereas in her previous
rejoinder (1462) she had sharply rejected their comments (μηδὲν... ἐπεύχου....
unôè . . . ἐκτρέψηις). In just the same way here, after a reply (1551) which
began with the sharp rejection οὐ σὲ προσήκει, she now assumes a more
lenient tone (ξὺν ἀληθείαι here as in 1475 ὥρθωσας). This is the answer to

1 Misinterpreted by Wilamowitz : ‘he [the guilty man] waits for us’, cf. his Interpr. 224.
Most modern commentators rightly understand ‘it remains’, ‘it abides' and a general sub-
ject; it is explained in what follows (on the easy change of many verbs from the personal
to the impersonal use, cf. Wackernagel, Syntax, i. 116 ff., particularly 118). There is no
need, then, to alter μένει. In the following sentence the three predicates, εὐμήχανοι, τέλειοι,
κακῶν μνήμονες, stand in front, the subject Zeuval follows.
4872.3 S 737
line 1567 COMMENTARY

Sidgwick’s objection that ἐς τόνδ᾽ eveßns ξὺν ἀληθείαι χρησμόν ‘gives a poor
sense’. It is also clear that τόνδε χρησμόν must refer to the preceding utterance
of the Chorus (Kranz, loc. cit., compares 1560 ὄνειδος. . . τόδε, 1551 f. τὸ
μέλημα . . . τοῦτο). Several German scholars have tried to modify or water
down the meaning of χρησμόν, so, e.g., Humboldt, Nägelsbach, Wecklein (‘Auf
diesen Spruch bist du im Einklang mit der Wahrheit gekommen’), Kranz
and Pohlenz in the passages quoted above; cf. also Wilamowitz, Interpr.
201 : ‘dieser Spruch’. A similar kind of nondescript rendering is produced by
several English commentators, e.g. Paley, ‘You have rightly touched upon
that divine law of retribution’; Sidgwick, ‘this maxim’; Headlam, ‘On that
pronouncement you hz.ve entered with full truth’. Conington (Appendix II
to his edition of the Choephoroe, p. 164) claims that ‘the χρησμός is not a
prediction [the remarl: is equally incorrect in regard to the passage from
which he starts there, C ho. 297],! but the announcement of a general law’. But
some others have kept strictly to the right meaning, e.g. Klausen, ‘In hanc
vaticinationem cum veritate pervenisti’ ; W. Sewell, ‘Upon this say oracular
Thou didst embark with truth’; A. Platt, ‘Thou hast spoken this oracle with
truth upon thy lips’. T 16 general remark of the Chorus (1563 ff.), μίμνει δὲ KrA.,
with its asseveration ‘so long as Zeus abides on his throne’, has itself a con-
siderable element of prediction. Clytemnestra emphasizes still more strongly
the future aspect wher she calls the utterance a prophecy. The words of the
Chorus applied to past and present as well, but what above all stings the
heart of the guilty woman is their relevance to the future, to the fulfilment of
her destiny which she constantly dreads and knows to be inescapable. As
in the past, so also in the future there can be no exception to the law that
ordains ἐκτίνει ὁ καίνων. Though she says at the same time that she will
strike a bargain with the spirit of evil she is not deceived for a moment ; this
is another case of γένοιτο δ᾽ εὖ παρὰ γνώμην ἐμήν (cf. on 255).
1568. δ᾽ οὖν breaks off and passes to a new point, cf. on 34. The thought
which is of general applicability leads up to a point which concerns the
speaker.in particular. For the function of οὖν in such cases cf. on 255.
1569. ‘In anapaesticis carminibus ἐθέλειν usitatissimum' (Lobeck on S. 47.
24). For the distribution of θέλειν and ἐθέλειν in general cf. Rutherford,
New Phrynichus 415 f., and Wilamowitz on E. Her. 18.
It is rash to alter tle form Πλεισθενιδᾶν of the MSS? into -δῶν,᾽ as Dindorf
demanded in his 'Annotationes' (1841). Cf., e.g., 44 (anap.) Ἀτρειδᾶν. ‘Dori-
cisms admitted by Aeschylus into his anapaests occur chiefly in proper names
or in words that were not used in the ordinary Attic speech of his time’
(H. W. Smyth, Harvard Studies in Class. Philol. vii, 1896, 144).
1569 ff. In order to understand 1569-73 the first point to get clear is the
construction and the subject of the infinitive orépyew and implicitly the
relation between στέργειν and ἰόντα... τρίβειν. The latter clause refers of
course to the δαίμων. A5 for the subject of στέργειν, older commentators from

! There the χρησμός is indeed general, i.e. not confined to the special case of Orestes (cf.
above on 1482, p. 703 n. 2), but none the less a real χρησμός, a 'prediction'.
2 It is quite possible that the small «v above -ἂν in Tr is here, as similar additions else-
where, meant, not as a correction or variant, but as an explanation. Cf. p. 731 n. 1.
3 Still less justified is Kirchhoff's and Wilamowitz's silent substitution of ᾿Ατρειδῶν for
᾿Ατρειδᾶν of the Mediceus in lyric iambics, Cho. 407.

738
 COMMENTARY lines 1569 ff.

Stanley downwards and most of the later ones have taken it to be Clytem-
nestra. That is the natural solution and should never have been doubted.
Wilamowitz has caused confusion by punctuating after ἐθέλω and explaining:
* ἐθέλω sc. θάψαν καὶ θρηνῆσαι᾽, ie. he takes this as an answer to the question
(1541) in the ephymnium which he, like many others, wrongly inserts before
1567, and then makes the δαίμων the subject of στέργειν; (cf. his translation:
‘Wie schwer die Schuld sei, lass’ er’s beim Geschehenen bewenden’, and also
his Interpr. 201 f.); thus the infinitives would have to be solely dependent
upon ὅρκους θεμένη. There is no need to point out the artificiality of this
rendering once the repetition of the ephymnium is disallowed ; I would merely
call attention to a decisive point, which seems to have been overlooked by
Wilamowitz and by his followers. The concessive element in the clause rdöe
. στέργειν δύστλητά περ ὄντα would lose all point if the person to whom falls
the τλῆναι were not the same who has to be prepared to στέργειν. Who is it
that finds the fatal events in the house of the Atridae so hard to bear, the
evil daimon whose proper element they are? or the mortal who is smitten by
them? For some time now Clytemnestra has felt the series of ever-renewed
misdeeds δύστλητα, and soon (1654 ff.) she is once more to give moving expres-
sion to this feeling. Wilamowitz, in referring τάδε μὲν στέργειν to the daimon,
persuaded not only Luise Reinhardt? but surprisingly enough even such
critical scholars as Pasquali (Stud. Ital., N.s. vii, 1929, 231 ff.)* and Pohlenz
(Griech. Trag. ii. 33).5 Kranz (Hermes, liv, 1919, 317) has rightly returned to
the older explanation.
There is no difficulty in the change of construction in the infinitive clauses
dependent upon ἐθέλω. As Nägelsbach says: ‘ ἐθέλω duobus cum infinitivis
diversa vi iunctum est: ἐθέλω τάδε μὲν στέργειν, ὃ δὲ λοιπόν, ἰόντα... τρίβειν,
quorum illud volentis est aliquid facere, hoc iubentis aliquid fieri. Sed
ad hunc alterum infinitivum etiam ὅρκους θεμένη non habuit nullam vim’ (so
also Wecklein). The last statement is correct, but holds good also for the
first infinitive, στέργειν. Since not only ἐθέλω but also δαίμονι... ὅρκους
θεμένη precedes, the hearer naturally takes the following τάδε μὲν στέργειν as
the content of both—of Clytemnestra’s intention and readiness (ἐθέλω) and
also of the agreement to be sworn with the daimon (ὅρκους θεμένη). Thus the
obligations to which both parties would have to pledge themselves in the
covenant are clearly indicated.$ The introduction of a new subject in ἐόντα
* Karsten preceded him: "Equidem parata sum Plisthenidarum daemonem obsecrare, ut
quae acta sunt, quamvis gravia, aequo animo ferat'. van Heusde agrees.
2 I do not know what van Heusde meant when he explained ‘ad δύστληταὄντα intell.
δαίμονι ἅτε ἀκορέστωι φόνου'.
3 ‘Zu den Liedern des Aischylos’, Beilage zum Bericht für 1927 der Staail. Augusta-Schule,
Berlin 1928, 30. Her treatment, which contains a few good observations, is far too much
controlled by the desire to prove that Wilamowitz is almost always right. Her 'psycho-
logical' considerations on Ag. 1568 ff. are taken too seriously by Pohlenz, loc. cit.
+ He goes so far as to say ‘Clitennestra non potrebbe mai adoptar στέργειν di se mede-
sima'. The infinitives he, like Wilamowitz, makes dependent on ὅρκους θεμένη and, in order
to give ἐθέλω ἃ content, he assumes an anacoluthon, in which his pupil M. Berti (cf. on 12),
243 f., follows him.
s He omits in his paraphrase the decisive words δύστλητά περ ὄντα : ‘that he be satisfied
with what has up to now come to pass and leave the house’ (so also 1. 103). Pohlenz’s treat-
ment is accepted without qualification by R. Hölzle, Aufbau der lyr. Partien des Aisch.
(diss. Freiburg i. Br. 1934), 40 n. 85.
. $ Provided it is remembered that the infinitives are formally dependent on ἐθέλω, we
739
lines 1560 ff. COMMENTARY

is the less harsh in that the clause ὃ δὲ λοιπόν (which is parallel to τάδε μέν)
has already broken the construction.
1569. Πλεισθενιδᾶν, and 1602 τὸ Πλεισθένους γένος. We cannot tell at what
point in the family tree of Atreus Aeschylus fixed Pleisthenes; it is very
doubtful if he intended to fix him at all. Pleisthenes, probably a somewhat
shadowy figure even to fifth-century Greeks, has ‘not acquired a settled place
in the pedigree, and of his deeds we know nothing’ (Wilamowitz, Pindaros,
510, see his examination of the evidence there n. 4, and cf. also C. Robert,
Heldensage, 301; Bowra, Greek Lyric Poetry, 128; W. Ferrari, ‘Studi Stesi-
corei', Athenaeum, 1938, 15 f). It seems as though Aeschylus had brought
in the sonorous name chiefly for its ornamental effect. In this he may well
have been influenced by some lyric poets such as Stesichorus, fr. 15. 2 D.,
and Ibycus, fr. 3. 21 D. In Ibycus, as in Stesichorus (in spite of what Wilamo-
witz says Arist. u. Ath. ii. 183 n. 32, commentary on the Choephoroe, p. 248
n. 3, Aisch. Interpr. 19; cf. more recently Ferrari, loc. cit., and Lesky, RE
xvili. 976 f), Agamemnon is called Πλεισθενίδας βασιλεύς (and in the same
sentence Arpeos . . . mais), cf. Bacchylides 15 (14). 48 Πλεισθενίδας Μενέλαος.
1570. For ὅρκους θέσθαι, which comes to mean ‘to make a covenant’, because
the form of sanction by oath was the ordinary custom, cf. K. Latte, Heiliges
Recht (1920), 98.
1571. The reading of G and Tr δύστλητα is obviously right.‘ The word
δύστλητος, known to Empedocles, occurs only here in Tragedy (in E. Phoen.
1438 the δύστλητον which is found in all MSS except M cannot stand), while
ärAnros is not uncommon from Homer onwards, cf. on 1360 δυσμηχανῶ and
on 1531 εὐπάλαμος. Aeschylus uses δύσοιστος once in the Prometheus and
twice in the Oresteia ; Sophocles also has it. [Cf. P. Oxy. 2362, fr. 1, col. 2. 10.]
1573. αὐθένταισιν : ‘brought about by blood-relations’, ‘carried out on blood-
relations’; the word has the same meaning as αὐτοφόνα in 1og1 (v. ad loc.).
Cf. Blass on Eum. 211.
The principal item in the proposed covenant with the δαίμων (1571-3)
reminds us of a type of prayer such as is often made in the hope of ridding
oneself of a harmful god or spirit. On this type of prayer, the ἀποπομπή, see
the brief remarks of Jane Harrison, Proleg. to the Study of Greek Rel. 8 f.,
and the very full account (prompted by an essay of Wünsch) in O. Wein-
reich's book Gebet und Wunder (Tübinger Beitr. z. Altertumsw., Heft. v, 1929),
175 (9) ff. A god or δαίμων in whom the power or the desire to harm is strong
(and in what god or δαίμων in his original form is it not ?), once he is pouncing
on his prey, will not be induced by a mere prayer for mercy to give up his
purpose: if he is to spare his intended victim, another (or others) must be
shown him on which he can wreak his will. So if the prayer is to be effective
it will not suffice to cry parce, recor, precor; something positive has to be
added such as: abi quo blandae iuvenum te revocant preces ; tempestivius in
domum Pauli... comissabere Maximi.” This naive belief can be seen in its
need not criticize the translators who are primarily concerned to emphasize that in these
infinitives the stipulations of the contract and oath are indicated, so, e.g., Humboldt : ‘Aber
ich will gern Plisthenes Stamms Rachdämon mit Schwur zusagen, nun dies zu erdulden'
etc., Lewis Campbell: ‘I would compound with the Genius of the race of Pleisthenes, here
making oath that I am willing to acquiesce in what has been, however hard’, others likewise.
1 Miswriting of T as Π iscommon. Cf., e.g., Eum. 356 πίθασος, Pers. 904 θεόπρεπτα (MF).
2 Horace, Odes. 4. τ. 2 ff. In this passage abt corresponds exactly to tovr’ ἐκ τῶνδε δόμων,

740
 COMMENTARY line 1575
unmitigated force in the invocation still common to-day in south Germany
(Weinreich, 191 [25] f.): ‘O heiliger Sankt Florian, verschon dies Haus (cf.
ἐκ τῶνδε δόμων), zünd andre an!’ The diversion to ‘others’ (by no means
necessarily referring to the person or property of an enemy of the man who
makes the prayer) we find in Ag. 1571 ff., where the traces of such a prayer
to a maleficent daimon are still clearly marked (ἄλλην γενεὰν τρίβειν), exactly
as, e.g., in Theognis 351 ff. d δειλὴ wevin, τί μένεις προλιποῦσα παρ᾽ ἄλλον ἄνδρ᾽
iévav ; ... ἀλλ᾽ ἔθι καὶ δόμον ἄλλον ἐποίχεο κτλ., E. Hel. 360 f. ἄλλοσ᾽ ἀποτροπὰ
κακῶν γένοιτο, τὸ δὲ σὸν εὐτυχές,; Catullus 63. 91 ff. dea magna, dea Cybebe...
procul a mea tuos sit furor omnis, era, domo: alios age incitatos, alios age
rabidos, Propertius 4. 6. 9 tte procul fraudes, alio sint aere noxae and elsewhere.
1574. The alteration of τε to δέ is here, as often elsewhere, perhaps the
simplest solution. An asyndeton would in itself be quite reasonable, both as
summing up and as rounding off? But the order of the words is against
Karsten’s τι μέρος (one would expect τι βαιὸν μέρος), moreover ‘a small part’
is better than ‘some small part’.
1575. It is true that ἀπόχρη does not occur elsewhere in Tragedy or even in
exalted poetical language at all. But it is a mechanical way of treating
stylistic evidence to reject on these grounds a form of expression which is
common in Herodotus and found also in Aristophanes, Plato, and the Attic
orators of the fourth century (Erfurdt on S. Phil. 475 conjectures in our
passage πᾶν ἀρκεῖ μοι, Headlam mavemapkés ἔμοιγ᾽; Platt, too, in J. Phil.
XXXV, 1920, 93 discards ἀπόχρη). Can we really doubt that ἀπόχρη μοι and the
like was a common Attic idiom at the time of the Oresteia, and if we admit
this are we to forbid Aeschylus to borrow once in a way from colloquial
language a word which in general was regarded as below the stylistic level of
Tragedy? Cf. p. 608 n. 1.
πᾶν aroused the displeasure of Housman, J. Phil. xvi, 1888, 277; Verrall
(in his note),’ Headlam, and A. Y. Campbell have followed him here. In
defence of πᾶν Platt, loc. cit., compares the expression in the comic poet
Ephippus fr. 15. x f. K. (Athen. 8. 359 a) ἀλλ᾽ ἀγόρασον εὐτελῶς ἅπαν γὰρ
ἱκανόν ἐστι, but this is not really comparable: leaving aside the lack of a
dative, dzav has here a narrowly limited sense, 'any piece of fish or venison
or the like', cf. in the following lines ἀρκεῖ τευθίδια σηπίδια κτλ. But a true
analogy to the construction and the meaning of πᾶν in the present passage is
provided by the parallel quoted by Blaydes, E. Hec. 317 f. καὶ μὴν ἔμοιγε
* Weinreich has overlooked not only these passages (there are probably more, but I have
not looked further), but also the important example Ag. 1571 ff. (In Mélanges F. Cumont,
1936, 493 n. 3, where he returns to the subject of ἀποπομπή, Weinreich does mention Ag.
1571 ff., but asserts that it is an entirely untypical instance.) This oversight makes it
possible for him, pp. 177 (11), 194 (28), 198 (32), to posit the general rule that, in opposition to
the corresponding Roman prayers, Greek prayers of this type were always directed against
enemies, not against other men in general. For ἀποτροπή against enemies cf. also (not in
Weinreich) Sept. 255 ὦ παγκρατὲς Ζεῦ, τρέψον εἰς ἐχθροὺς βέλος, 628, Apoll. Rh. 4. 448, Nicander,
Ther. 186, Virgil, Georg. 3. 513, Hor. Odes, 3. 27. 21 ff., Tacitus, Hist. 3. 10 (near the end),
Ann. 1. 43, Àusonius 417. 58 ff. (p. 278 Peiper), in connexion with a prayer to Nemesis.
? Housman, J.Phil. xvi, 1888, 277 remarks: 'Auratus' δὲ is sufferable but still an en-
cumbrance: the connecting particle should be γὰρ or there should be no connecting particle.'
3 In his translation, obviously meant to serve faute de mieux, Verrall renders the passage :
‘A part of the wealth is not much to me who have it all’, very artificial, if not entirely
impossible.

741
line 1575 COMMENTARY

ζῶντι μέν, καθ᾽ ἡμέραν Kei σμίκρ᾽ ἔχοιμι, πάντ᾽ ἂν ἀρκούντως ἔχοι. Cf. also
Alexis fr. 167 K. γυναιξὶ δ᾽ ἀρκεῖ πάντ᾽, ἐὰν οἶνος παρῆι πίνειν διαρκής.
For the remaining problems of the text of 1575, Erfurdt’s change in the
order of the words and the removal of the obviously otiose δ᾽ (put in to avoid
hiatus) are, if not completely certain, at least very probable; most editors
have accepted them. A closely similar displacement in anapaests can be
seen in Eum. 314: there it has been corrected by Porson with general approval.
The tone of Clytemnestra’s last sentence and the conditional, not purely
temporal, character of ἀφελούσηι were strangely misunderstood by Sidgwick.
(‘The cold irony is at its height here. ‘AI I want is a quiet life and a humble
competence, having cured the house of its blood-feud." ') The queen is filled
with deep anxiety about the future and prepared to give up much, indeed
the greater part, of her possessions if she could only hope thereby to buy
herself free from the vengeance of the spirits of retribution and bring to an
end the madness of murders of kindred, murders of which she herself must
be the next victim.

1577. For the greeting of a day which has brought about a decisive change
cf. S. Ant. 100 ff., and (Blaydes) Ar. Peace 556, ὦ ποθεινὴ τοῖς δικαίοις καὶ
γεωργοῖς ἡμέρα κτλ. For the meaning of eódpov and the erroneous rendering
εὐφραντικόν, which has been put forward here (Tricl.) and elsewhere, cf. on 806.
As Aegisthus feels it, the long-desired day of vengeance, which has dawned at
last after so many bardships, is 'kindly disposed' : ἄλλοτε μητρυιὴ πέλει ἡμέρη,
ἄλλοτε μήτηρ.
δικηφόρου: cf. on 525. Here again one of the fundamental motifs of the
Oresteia is stressed. It is significant that in the very first sentence he utters
Aegisthus insists upon his δίκη. His legal claim (cf. on 813) against the son of
the man who had committed such frightful atrocities against Aegisthus’
father and brothers and himself is fully valid not only from the speaker’s
point of view but from the poet’s as well. Aegisthus, it is true, is a loathsome
coward, but that cannot annul or diminish his legal claim. On the other hand,
the fact that he has a right to claim for himself dixn is of course no proof that
he was justified in using the means he has used against Agamemnon. As Paley
observes, the speech both begins and ends (1604, 1607, 1611) with the δίκη motif.
1578. νῦν: for the word-order cf. on 1036.
It is generally recognized that βροτῶν depends on τιμαόρους, but there are
doubts about the connexion of γῆς. Most commentators follow Stanley in
taking γῆς ἄχη together. On the other hand, Karsten, Housman (J. Phil.
Xvi, 1888, 286), and Verrall maintain that γῆς depends upon ἄνωθεν (Verrall
translates ‘from above earth’), and these have been followed by Pliiss and
Platt (‘from high upon the earth’). This is very unlikely, for it presupposes
an artificial order and connexion of words, whether with some of the trans-
lators we separate βροτῶν entirely from τιμαόρους and connect it only with
dyn, or whether we regard βροτῶν as being used ἀπὸ κοινοῦ (Pliiss). Nor would
ἄνωθεν γῆς be correct. It might perhaps be possible for Aeschylus to join
ἄνωθεν to a genitive, though so far no instance of it has been found in Tragedy,
and it only rarely occurs in older Greek at all.‘ But as far as I can see,
1 The rarity of this usage is duly mentioned in L-S, ἄνωθεν I. 2 b, but in addition to the
instances in Hdt. 1. 75 and Hippocr. s. ἄρθρ. 80 reference should have been made at least
742
 COMMENTARY line 1581

ἄνωθεν with the genitive is never used with the meaning ‘from above’ but
is always ‘above’. ‘Above the earth’, however, would be meaningless here.
On the other hand, γῆς ἄχη is appropriate: the gods from their cosmic view-
point, ἄνωθεν ἐποπτεύοντες (cf. on 1270), observe everything that happens on
earth; so the clouds in the play of Aristophanes invite each other (289 f.)
ἐπιδώμεθα τηλεσκόπωι ὄμματι γαῖαν. With θεοὺς ἄνωθεν ἐποπτεύειν cf. Ar.
Birds 1509 ἄνωθεν ὡς ἂν μή μ᾽ ὁρῶσιν οἱ θεοί (the traditional punctuation, as in
R and V, after ἄνωθεν has been corrected in the editions of O. Schroeder and
Coulon ; for the postponement of ὡς dv cf., e.g., A. Ag. gıı) and 1550 f. ἵνα με
κἂν ὁ Ζεὺς ἴδηι ἄνωθεν.
ἄχη must not be altered, cf. on 1251.
The thought expressed in the first three lines is of a familiar pattern.
Stanley refers to the proverb (Diogenianus 6. 88) νῦν θεοὶ μάκαρες" emi τῶν
ἀξίως τιμωρουμένων ἐφ᾽ οἷς ἔπραξαν. The same idea appears in the exclamation
of Laertes, w 351 f., Ζεῦ πάτερ, ἦ pa ἔτ᾽ ἔστε θεοὶ κατὰ μακρὸν "Ολυμπον, ei
ἐτεὸν μνηστῆρες ἀτάσθαλον ὕβριν ἔτεισαν. Cf. also (Paley) E. Suppl. 731 f. νῦν
τήνδ᾽ ἄελπτον ἡμέραν ἰδοῦσ᾽ ἐγὼ θεοὺς νομίζω. The way in which Aegisthus
starts marks him as one who has hitherto adhered to the opinions attacked
in 369 ff.—6 δ᾽ οὐκ εὐσεβής.
1580. Nauck’s πάγαις for πέπλοις, brilliant like most of the conjectures of
that eminent critic, has been adopted by Wecklein (commentary), A. Y.
Campbell, and G. Thomson. It is not right, however. πέπλοι is the word used
to denote the fatal robe in 1126! (cf. Eum. 635 δαιδάλωι merAwı), and ὕφασμα in
1492 (where there is a special reason for the use of this noun) and Cho. 1015;
here we have the two words combined. Anyone who feels inclined to take
exception to the phrase (Wecklein asserts that ‘ ὑφαντοῖς is pointless with
πέπλοις ") may regard ὑφαντοῖς πέπλοις as a Homeric expression as in v 218
ὑῴφαντά τε εἵματα καλά, v 136 (= π 231) ἐσθῆτά θ᾽ ὑφαντήν.
“ πέπλοι ᾿Ερινύων dicitur tunica perniciosa, quia hac arte exsecrationes
Thyestis ratas faciunt Furiae’ (Klausen).
This passage was at the back of Sophocles’ mind when he blended it with
the Aeschylean use of the word ἀμφίβληστρον (cf. on 1382): Trach. τοῦτ f.
᾿Ερινύων ὑφαντὸν ἀμφίβληστρον.
1581. Τί we admit modern punctuation into our texts at all, it is very appro-
priate to put a comma after κείμενον (as Karsten and Murray do, while Plüss
puts a dash). Here we have an arrangement which is a favourite one with
Aeschylus (cf. on 616, 893, 1272 [οὐ διχορρόπως μάτην] and elsewhere) : after the
clause ἰδὼν... κείμενον, which is really complete in thought and syntactical
construction, a further important idea is added with the words φίλως ἐμοΐ,
and this is followed by a still further supplement in the explanatory participial
clause of 1582.?
to Ar. Ach. 433, Birds 1526 and Xen. de re equ. 7. 7, while Plutarch, Them. ı2. 1 should have
been omitted, for there the deletion of ἄνωθεν by Cobet (adopted by Lindskog) seems clearly
rıght.
* Cf. also Cho. 1000, but I am convinced that Schütz and Wilamowitz (in his separate
edition of the play) were right in deleting that line.
2 [n Eum. 1025 Weil (1861) was right in putting a comma before the word δικαίως at the
end of the sentence, and recognizing that this δικαέως is explained in the sentence which
follows, ὄμμα γὰρ πάσης χθονὸς κτλ. He has been followed by Davies, Wecklein, Headlam
(cf. his translation), Wilamowitz (cf. also his Interpr. 228), and Mazon.

743
line 1582 COMMENTARY

1582. The tus talionis extends to the manner in which vengeance is executed:
Atreus treacherously lured Thyestes into a trap (1590 ff.), and now Atreus'
son lies tangled in the μηχάνημα (cf. on 1127) of the deadly robe.
It seems likely that up to the end of 1582 Aegisthus takes no notice of the
presence of the Chorus. He first gives vent to his excitement, addressing
himself to something above the level of men, a semi-divine being (φέγγος
ἡμέρας δικηφόρου). This trait is characteristic of the quasi-monologues in
Aeschylus, cf. p. 25 and on 1648 ἀμφοῖν τοῖνδε. The following part of the speech
is delivered in a different tone: he gives a carefully balanced narrative with a
view to justifying himself before the Elders. He probably turns to them at
1583 ; in 1597 they are unmistakably addressed. As far as the dramatic form
is concerned, the first sentences (1577-82) may be compared with the first
speech of the Herald (cf. on 522). For a different view see F. Leo, Monolog
im Drama, 8, 3o n. 4; Schadewaldt, Monolog und Selbsigespräch, $3 n. a.
1583-5 has been discussed with as much forcefulness as lack of sympathy by
Platt (J. Phil. xxxii, 1913, 69), who, treading in the footsteps of former
critics, speaks of ‘the senseless rubbish ὡς τορῶς φράσαι᾽. The proper ex-
planation was advanced as long ago as 1856 in Schneidewin's edition: 'he
adds the words because he is giving the most exact account possible of the
persons and the circumstances as if in a court of law.’ And how can the
addition of the words αὐτοῦ δ᾽ ἀδελφόν be called ‘irrelevant’ in this context?
1584. πατέρα Θυέστην τὸν ἐμόν: the same type of word-order as in the
examples collected by H. Schöne (Hermes, lx, 1925, 155), who on p. 144 quotes
E. Hel. 390 τὸν ἐμὸν ‘Arpéa πατέρα.
The line contains two resolved ‘longa’. E. B. Ceadel, C. Q. xxxv, 1941,
84 n. 1, quotes three more trimeters from Aeschylus with two resolved longa
(I am not now taking into account lines with one resolved longum and one
resolved anceps), viz. Suppl. 341 (= 338 Wil, 342 Murray), Sept. 593, Cho. 89.
In these lines it is always the third and, except in Sept. 593, the first longum
that we find resolved. [ὡς τορῶς φράσαι: Wackernagel, Kl. Schrift. 817.]
1585. αὑτοῦ δ᾽ ἀδελφόν. If in classical Greek two relations of a person are
indicated, the two indications are sometimes connected by re . . . re, some-
times by μὲν... δέ, far more often by simple δέ, but not, on the other hand,
by single re. So the normal form is: A. Pers. 151 f. μήτηρ βασιλέως, βασίλεια
δ᾽ ἐμή, Hdt. 7. 10 a 2 πατρὶ τῶι σῶι, ἀδελφεῶι δὲ ἐμῶι, Δαρείωι, S. Oed. C. 322 f.
παῖδα σήν, ἐμὴν δὲ... ὅμαιμον, E. Hec. 534 ὦ παῖ Πηλέως, πατὴρ δ᾽ ἐμός, El.
1243 σφαγὰς ἀδελφῆς τῆσδε, μητέρος δὲ σῆς, Pl. Prot. 310 a “Immorparns, ὃ
AroAAodwpov ὑός, Φάσωνος δὲ ἀδελφός. Elmsley’s note on E. Mea. 970 (940
Elmsi.) illustrates this usage with abundant material from poetry and prose.
Anyone reviewing the evidence afresh can hardly doubt that Elmsley was
right to correct the τε of the MSS in the few passages which seem to deviate
from the rule, as, e.g., Ag. 1585 and E. Med. 970. Elmsley's results, with which
Hermann, despite his profound dislike (see, e.g., his Opusc. vi, pars i, 92 ff.)
for the rigid canones of the English Hellenists, agreed in the main (supplement
to Elmsley’s commentary on the Medea), have been deservedly taken over
by grammars, cf. Matthiä, Griech. Grammatik § 616. 3, Kühner-Gerth, ii.
243 n. 1. The sceptics (e.g. Peile and Conington defend τ᾽ in Ag. 1585) have
now been joined by Denniston, Particles, 502 e (cf. above on 1526). The
passage from which Elmsley started, E. Med. 970, is not mentioned by
744
 COMMENTARY line 1587
Denniston ;! what is more regrettable is that he does not make it clear which
usage is the normal one (even if we add the discussion p. 163).
Most editors agree with Elmsley in writing αὐτοῦ. That αὐτοῦ should be
kept (‘ut ex Aegisthi persona dictum’ Hermann) may be admitted as a
possibility, but it is unlikely.
ἀμφίλεκτος ὧν κράτει. For the meaning of ἀμφίλεκτος in this passage,
‘asserted on both sides, challenged, disputable’, cf. on 88x. Thus ‘chal-
lenged with regard to his claim to sovereignty’ or ‘challenged by a claim to
sovereignty’. As for the dative κράτει, we may perhaps class it among the
comparatively rare instances of the ‘dative of relation’ (Matthiä, Griech.
Gramm. § 400; Kühner-Gerth, i. 440. 12), or simply take it as instrumental,
especially as the verbal element in ἀμφίλεκτος is still strongly operative; cf.
the instrumental (or local?) dative in phrases like κινδυνεύειν τῶι σώματι, τῆι
ψυχῆι, ἁπάσηι τῆι ᾿Ελλάδι and the like.?
The allusion here to the quarrel between Atreus and Thyestes need not be
taken, as it is by Schütz,? to show that Aeschylus had in mind a version of
the story differing from that given elsewhere. It is rather the case that a
veil is deliberately drawn here over all details We can hardly expect
Aegisthus to accuse his own father: so this account avoids the subject of
Thyestes’ adultery, which, however, the poet assumes to be known to his
audience (1193). The quarrel for power between the two brothers was part
of the established legend: there was no occasion to go into the tale of the
golden lamb at this point. The time at which the banishment took place is
left as obscure as the particular incident which ultimately caused it.
1586. ἠνδρηλάτησεν : cf. on 1419.
1587. The true meaning of προστρόπαιος was elucidated by Otfr. Müller,
Aesch. Eumeniden, p. 135: ‘a προστρόπαιος, according to the primary meaning
of the word, is, like a ixerns, one who turns towards another, beseeching him
to take him in [for this Müller quotes Ag. 1587 among other passages] . . .;
the action of προστροπή then denotes humble supplication in general . ...
But it is usually associated with the murderer who is a fugitive with his
crime unexpiated, and προστρόπαιος assumes the meaning of a man under a
a curse, homo piacularis’, etc. See further Wilamowitz on E. Her. 1161 and
E. Rohde, Psyche, i, sth ed., 264 n. 2, 275 n. 2. G. Murray (The Rise of the
Greek Epic, 4th ed., 88 n. 2), with his objection to the statement in L-S as
to the origin of the word, does less than justice to the results of Müller’s
investigation.*
1 What Denniston says about E. El. 1243 is due to his having inadvertently read the
critical apparatus on 1245 as referring to 1243.
2 These expressions, it is true, are entered by Kühner-Gerth, i. 420 under the heading of
‘dat. incommodi’, but this is less convincing than the ‘instrumental’ interpretation (thus
K. W. Krüger on Hdt. 2. 120. 2 and on Thuc. 2. 65. 4 [7]).
3 The parallel which he draws with the case of Eteocles and Polynices has been taken up,
though given a rather different turn, by C. Robert, Oidipus, i. 406, Heldensage, 294 n. 1.
Yet as far as Pelops' curse upon his sons and their descendants is concerned, we possess no
evidence for it apart from the ἱστορία (it represents the vulgate tradition as these ἱστορίαι
generally do, cf. Ed. Schwartz, Fleckeisens Jahrb. Suppl. xii, 1881, 441 ff.) in the Scholion A
on Hom. B τος, for it is very doubtful whether Hellanikos recorded the curse, cf. Jacoby in
his commentary on F Gr Hist 4 F 157.
4 In L-S, s.v., though not in the earlier editions, Ag. 1587 is wrongly classified and
explained.
745
line 1587 COMMENTARY

‘ ἑστίας : objective genitive to προστρόπαιος ' (Plüss), cf. Kühner-Gerth, i.

371 n. 19.
1588. μοῖραν is not taken into account by many translators. Those who do
pay attention to it differ in regard to its meaning. On the one hand, e.g.,
Stanley: 'sortem invenit securam’, and Klausen: ‘ μοῖρα: vitae condicio' ;
on the other hand, Verrall : ‘found a partial safety so far as that he did nol . . .,
literally “obtained the saving of part” '. The former interpretation is
obviously correct. The unqualified assertion of success with which the
sentence begins is extremely effective here as a set-off to the horror which is
to follow: ‘Thyestes hoped that his life would be spared. It was.—His own
life (adrôs).—But . . .' (I owe this paraphrase to Beazley).
1589. For the infinitive with consecutive force introduced by τὸ μή cf. on 568.
1590. Schütz took exception to αὐτοῦ, and Paley's attempt to understand it
in a local sense ('there on the spot', accepted by Headlam and Mazon) is un-
convincing. Hermann's violent rewriting of the text calls for no refutation ;
Blomfield's emendation is clearly right. 'Haec erat μοῖρα ἀσφαλής, quod supplex
ad focum non occisus est ipse, ξένιον vero accepit liberorum carnem’ (Enger).
ξένια. The only other passage in Aeschylus where we find a tribrach con-
sisting of a single word in this position in the line is Cho. x (χθόνιε) (cf. C. F.
Müller, De edibus solutis, diss. Berlin 1866, 16 f.).
1591-3. Schütz remarked: 'verba Arpeds προθύμως μᾶλλον ἢ φίλως manifesto
glossam produnt' and in his later edition he entirely omitted the words which
he found objectionable, while he rearranged 1590 by deleting αὐτοῦ and
inserting the word πατρί from the end of the next line: ξένια δὲ τοῦδε δύσθεος
πατὴρ πατρί. This was not actually adopted by Hermann, but he did call it
an 'ingeniosa coniectura', with the result that an undue importance has been
attached to this unfortunate experiment. How could the words προθύμως
μᾶλλον 3) φίλως give the impression of being a gloss? Wecklein (1888) rightly
compared the words with E. Med. 485 πρόθυμος μᾶλλον ἢ σοφωτέρα, but
nevertheless spoke of Schütz's deletion as worth consideration; Weil finds
room for it in the meagre apparatus of the Teubner edition, Sidgwick goes so
far as to bracket in his text the words to which Schütz objected, and Platt
omits them in his translation (1911). Finally it may be a consequence of tbe
uneasiness aroused by Schütz that Wilamowitz (as early as 1885 and again in
1914) transposed the initial words of 1591-3 (τὠμῶι προθύμως... δοκῶν
Kpeoupyov . . . ‘Arpeds παρέσχε) ; his transposition has been adopted more
recently by A. Y. Campbell and Murray. I have never been able to under-
stand what we are supposed to gain by it. If anything is certain in this
passage, it is the fact that the monotonous repetition of the syntactical
pause which in four successive lines follows a two-syllable word at the
beginning of the line (αὐτός Arpevs, τὠμῶι, δοκῶν.) is intentional; Plüss
aptly compares 1403-6 (in 1404 the word at the end of the clause is one
syllable longer).! A strong emotional element is common to both passages,
and we may surmise, with all reserve, that in the words of Clytemnestra as in
those of Aegisthus the extraordinary passion, in the one case of triumph, in
the other of indignation and horror, finds its expression not only in the words
but also in the structure and sound of sentence and line: no smooth flow,
1 We find the same device on a smaller scale in the two lines Eum. 7 f., where the purpose
of hammering in the relation between Φοίβη and Φοῖβος is obvious.

746
 COMMENTARY line 1592

but stabbing and tearing, and a harsh and jarring rhythm. The ‘needless’
mention of the name Atreus after τοῦδε... πατήρ is presumably to be ex-
plained in part by the effort to secure the utmost clearness and precision (cf.
on 1583-5); besides ‘the name focuses the hatred’ (Beazley). προθύμως
μᾶλλον ἢ φίλως is most appropriately put after the clause which contains to
begin with only one single element of the action, ξένια : in normal circum-
stances it would be a matter of course that the host is a ξεῖνος φίλος. Sidgwick
missed the point: ‘ ‘‘More zealous than friendly” is only possible as a joke,
when applied to a man who, under cover of a banquet, murders his brother’s
children.’ Of this murder no mention has yet been made. The long sentence
with its many clauses is so carefully built up that the revelation of what
really happened comes only at the very end, after δαῖτα, with a horrible παρ᾽
ὑπόνοιαν. Up to this point everything seems on the whole to conform to the
procedure customary on such occasions: even xpeovpyóv ἦμαρ does not fully
disclose its ghastly meaning until later. So the expression προθύμως μᾶλλον
ἢ φίλως accords with the general scheme: the words, it is true, do hint that
the host was not behaving quite in the way one would wish and expect
among φίλοι, but still the expression is colourless enough not to weaken by
anticipation the gruesome conclusion. The effectiveness of this arrangement
with its element of surprise is not impaired by the fact that the audience
knows the story of the Thyestean banquet and has a specially vivid recol-
lection of it since the Cassandra scene. Aegisthus' speech is not merely an
integral element of the drama as a whole but also plays a part of its own,
characterizing the man and his temperament.
πρόθυμος also in E. El. 395 expresses the warmth of a host's feeling towards
his guests.
1592. xpeoupyóv ἦμαρ, which has been compared (Klausen, Paley, van
Heusde) with Cho. 261 βουθύτοις ἐν ἤμασιν, E. Hel. 1474 βούθυτον ἁμέραν,
S. Trach. 609 ἡμέραι ταυροσφάγωι, etc., approximates to the meaning of
'festival day' (apart of course from the sinister note which is introduced into
it here). It is well known that the Greeks of the classical period could not
allow themselves the luxury of eating any considerable amount of meat
except on festive occasions. On this point Wilamowitz, "Staat und Gesell-
schaft der Griechen’, 2nd ed., Die Kultur der Gegenwart, ii. 4. x. 28, remarks:
"Their gods found it harder to settle down to the new way of life than the
men... the latter only got a taste of meat, anyhow of beef, on holy days
when cattle were sacrificed to the gods; but for the gods too the ‘hecatomb’
which they received once upon a time soon became nothing more than a
rather more sonorous name without any numerical significance, and the
masses of the poor could only offer a model in clay or lead for their sacrificial
victim, since the new dwelling places only seldom allowed of the keeping of
cattle on a large scale', etc. Cf. also Martin P. Nilsson, Greek Popular
Religion (New York 1940), 22 f. In the present passage the festal banquet is
one of the regular features of the ceremonies of reconciliation, cf. G. Glotz,
La Solidarité de la famille, 159. As for the expression «peoupyér, it is obviously
chosen because it lends itself to connoting inhuman savagery,’ cf., e.g., Hdt.
1 This ἀπροσδόκητον was noted by Keck, who also rightly emphasized the structure of
lines 1590-3 and the effect of the pause after the first word. .
2 Bücheler in a famous article, Rhein. Mus. xxxviii, 1883, 479 = Kl. Schr. ii. 500, after
747
line 1592 COMMENTARY

3. 13. 2 τοὺς ἄνδρας κρεοργηδὸν διασπάσαντες, 7. 181. 2 μαχόμενος ἐς ὃ κατεκρεορ-

γήθη ἅπας. It may, however, be permissible to see a still more particular
point in the word here. There is a widespread tradition associating the word
κρεουργεῖν OT kpeovpyia With Tantalus’ slaughtering of Pelops ; I do not know
how it originated, but I am inclined to think that the use of this particular
expression is fairly old in spite of its late attestation. The following references
are from the Thesaurus: Lucian, Salt. 54 ἡ τοῦ Ταντάλου φλυαρία καὶ ἡ map’
αὐτῶι θεῶν ἑστίασις καὶ ἡ Πέλοπος xpeovpyia, Etym. M. 279. 15 (= Schol.
Lycophron 156) φησὶν οὖν ὅτι μετὰ τὴν κρεουργίαν τοῦ Πέλοπος οἱ θεοὶ ἀνεζωο-
ποίησαν αὐτόν, Greg. Naz. Or. 4. 70 (= Patrol. Gr. xxxv. 591 M.) and
Or. 39. 5 (= P. Gr. xxxvi. 340) τὴν Πέλοπος kpeoupyiav. To these passages
should be added Schol. Pind. Ol. 1. 43 a ὅτι ἐκρεουργήθη ὁ Πέλοψ (similarly
Schol. 75 b), Schol. Aristides vol. iii. 564. ro Dind. of δὲ ὑπομνηματισταὶ
λέγουσιν ὅτι τὰ ἐπὶ (ὅτι ἐν codd.: corr. Wilamowitz, Pindaros 271 n. τὴ τῆι
κρεουργίαι τοῦ Πέλοπος ὠρχήσατο 6 Πάν. R. Pfeiffer has drawn my attention
to ‘Nonnus’, Histor. 4 (= Pair. Gr. xxxvi. 989 M.) περὶ τῆς Πέλοπος kpeoupyias
... TOv Πέλοπα... .. 6 Τάνταλος κατασφάττει καὶ κρεουργεῖ καὶ ἔψει. When, with
these passages in our mind, we read in the schol. on A. Cho. 744 ἡ κρεουργία
τῶν Θυέστου παίδων, we shall not be inclined to assume that the phrase had
been simply evolved from Ag. 1592, but rather consider the possibility that a
commentator used the word which was commonly associated with the crime
of Tantalus to describe that of Atreus. Alternatively (I suggest this with
some hesitation) in the legend of the murder of Thyestes’ children, too, it
may have become traditional to speak of xpeovpyetv. If that was the case, it is
likely that Aeschylus alludes to the stock expression in Ag. 1592. Otherwise
he may have been transferring to the deed of Atreus a word which in his own
time had already its fixed place in the tale of the crime of Tantalus.
εὐθύμως should be left undisturbed. I do not know whether ‘the assonance
of προθύμως... εὐθύμως is intentional and pointed’ (Verrall) : it is well known
that ancient writers, even in more polished styles than that of Aeschylus,
are less sensitive than modern ones about such repetitions. In any case,
εὐθύμως is a very suitable word to indicate the mood of one participating in a
festal banquet, cf., e.g., Ion of Chios fr. ı (= 26 v. Blumenthal). 14 D. εὐθύμων
συμποσίων πρύτανι, Where εὔθυμος is used as εὔφρων often is in similar cases
(cf. on 806). Similarly Pindar in a prayer at a symposium (fr. 155 Schr.,
cf. Wilamowitz, Pindaros, 141 n. 2) turns to Zeus, the Muses, and Εὐθυμία.
Cf. also E. Cycl. 530 πῖνε κεὐθύμει.
1594. ‘The asyndeton is explanatory’ (Wecklein).
1594 ff. This sentence, which has been a source of trouble for centuries, has
not come down to us uninjured; thus much is almost universally agreed ;
but whether the original text has been upset in one part or in several is a
matter of controversy. The difficulties caused by the form in which we read
the sentence in the MSS have been unnecessarily increased by the com-
mentators, many of whom look for an explicitness and accuracy of detail
stating that ‘die echte Bedeutung des italischen Wortes [caro], welche bei Umbrern und
Samnitern stets blieb, ist Teil, Stück, pars’, goes on to say that ‘ohne den Stückbegriff war
die Composition carnifex unmöglich’. But what about xpeovpyós? [Only now (1961) do
I see that E. Benveniste, ‘Festschrift für Prof. W. Havers’, Die Sprache i, 1949, 121,
observes: ‘carnufex calque exactement gr. xpeoupyds’. That seems to me convincing. ]
748
 COMMENTARY lines 1594 f.
such as is not to be forced into a narrative which only dwells upon a few
selected features. We should be particularly careful to steer clear of remini-
scences from the account in Herodotus (1. 119) of the meal arranged for
Harpagus by Astyages. That story and the tale of the cena Thyestae (the two
are combined by Ovid, Ibis 545 f) have indeed a good many points of
resemblance, and have therefore been compared by the commentators time
and again. But there are also important points of difference, and as far as
the general character of this Aeschylean narrative is concerned Karsten was
right in saying ‘poetam non describere factum, ut Herodotus facit, sed memo-
rare tantum' : it would be still more correct to say that here Aegisthus dwells
only upon those features of the story by which the inhuman crime of Atreus
is brought out in the strongest relief, while he passes in haste over other
points which he would be bound to explain in greater detail if he were
engaged in telling a. tale for its own sake. As regards Seneca's version of the
cena Thyestea, it should be obvious that we can make no inferences about
Aeschylus from the gloating relish with which the horrors of the story are
depicted by that lurid writer. Probably we shall never arrive at any certain
restoration of the text in every particular, unless some fresh MS evidence
comes to the rescue. On several important points the following observations
can do no more than raise questions and weigh up possibilities. But since
continual discussion and desperate attempts to iron out the difficulties of the
text by means of compromise have had the effect that nearly every detail has
become disputable, it will be as well to start with what I regard as being, if
not certain, at least highly probable.
First: ἔθρυπτε must not be altered. Casaubon's ἔκρυπτε, adopted by
Hermann and others, and still to be found in some of the most recent editions,
is nothing more than an interpolation based on Hdt. x. 119. 4 ff. (κατακεκα-
λυμμένα etc.) and the more common form of the Thyestes story ; the altera-
tion of this detail is typical of the treatment to which the whole passage has
been subjected. ἔθρυπτε is so entirely appropriate in the context that we may
regard it as certainly genuine. However, it does not mean 'he carved up
(like a scissor)’ as Wecklein says! (Wilamowitz's ‘Zehen und Finger hackt' er
ab’ is also incorrect), but ‘he minced up’ (Platt, C.R. xi, 1897, 97, cf. also
Verrall’s translation) or 'broke up small’. Among the instances of the word
there is a particularly instructive one in Pl. Parm. 165 b θρύπτεσθαι δὴ οἶμαι
κερματιζόμενον ἀνάγκη πᾶν τὸ ὄν. For the parts of the body which are the
objects of ἔθρυπτε here, see below. These preparations obviously took place
before the beginning of the banquet: παρέσχε δαῖτα also includes the pre-
liminaries. The sentence τὰ μὲν ποδήρη κτλ. contains the more detailed
exposition of the subject introduced by παρέσχε δαῖτα : this accounts for the
asyndeton and also for the imperfect ἔθρυπτε (cf. Wilamowitz on E. Her. gor,
Wackernagel, Syntax, i. 183). It cannot be gathered with certainty from
ἔθρυπτε whether Atreus undertook the execution of the act with his own
hand? (which would be quite understandable owing to the risk of his secret

! The scissor is the man in charge of the skilled carving of roast meat, poultry, etc., cf.
Friedlaender on Petronius 36 and on Juvenal 5. 120 ff.
2 Horace says in Ars poet. 186, where he may perhaps be thinking of some tragedy of
Sophocles or Euripides, ne . . . humana palam coquat exta nefarius Atreus. In the sentence
from a messenger’s speech in the Aireus (after an original of Sophocles? or of Euripides?

749
lines 1594 ff. COMMENTARY

being betrayed) or left it to some trusted servant (there is no objection to the

latter interpretation from the point of view of language, cf. A. C. Pearson on
Soph. fr. 620) ; in any case the deed is his.
Secondly: ἄσημα δ᾽ αὐτῶν is sound. Paley rightly stresses ‘the antithesis
between ra μὲν ποδήρη and ἄσημα δ᾽ aurav,—the articulated extremities
which would have at once revealed the contents of the dish, and the fleshy
parts which bore no such distinguishing mark’. In many of the more recent
editions one finds Dindorf’s ἄσημ᾽ " ὁ δ᾽ αὐτῶν. This conjecture has appealed
to critics, from Hermann onwards, because it gets rid of the harsh change of
subject. But a very high price had to be paid for the smoothing of the
construction. For quite apart from the pointless addition of the predicative
ἄσημα to ἔθρυτττε (or ἔκρυπτε, which was the reading accepted, not indeed
by Dindorf himself, but by many who adopted his conjecture) and the
consequent ambiguity of αὐτῶν (partitive depending on Aaßwv? or joined to
ayvoiaı ?)—what is now to be made of the μέν in ra μὲν ποδήρη, to which 6 δ᾽
αὐτῶν cannot possibly provide the balance? Few will be inclined to speak of
‘ μέν solitarium’ here. In ἄσημα αὐτῶν the genitive is probably partitive.’ I
am not sure whether it is necessary to stress the absence of the article (so
Klausen: ‘nonnulla ex iis quae nullo signo agnosci poterant’), or whether
ἄσημα αὐτῶν is used here as τὰ ἄσημα αὐτῶν would be in ordinary language.”
Thirdly: dv8paxds καθήμενος seems to be right, difficult though it is to
settle the exact place occupied by this phrase in the original sentence (see
below). The Homeric word ἀνδρακάς and the Homeric custom denoted by
καθήμ. (for the details see p. 754 f.) belong together. Is it too much to hope
that future editors will leave avöpaxas alone and no longer play with a
charcoal fire as Abresch did in the eighteenth century and Housman in the
nineteenth ?
Now we turn from imaginary difficulties to real ones. As the text stands,
the change of subject between ἔθρυπτε (Atreus) and ἔσθει (Thyestes) is beyond
belief, whatever Paley may say to comfort us. Then there is ἄνωθεν." It is
not clear to what it can refer ;* it might be thought that from the context we
have to supply ‘on top (of the dish of meat)’, but this seems to me extremely
doubtful. Lastly, dvSpaxds καθήμενος, placed where it is, can only be under-
cf. F. Leo, Gesch. d. róm. Lit. i. 396) of Accius, 220 f. Ribbeck, concoquit partem. vapore
flammae, veribus in foco lacerta tribuit (text at the end uncertain), the subject is probably
Atreus. In Seneca's Thyestes 760 ff. Atreus does the dissecting and cooking himself.
1 Lewis Campbell says (in a note on his translation) : ‘I take αὐτῶν as the genitive after
the privative adjective: “without the tokens of them," namely of the hands and feet,
which would have betrayed the nature of the meat’. I once considered this way of con-
struing the phrase but eventually rejected it because the well-known instances of ἀλαμπὲς
ἡλίου, ἄλυπος ἄτης, and the like provide a parallel for, let us say, ἄσημα τεκμηρίων, but not
for ἄσημα αὐτῶν.
2 Here again (cf. on 1395 f., p. 657 n. 2) we are hampered by insufficient information about
the details of the use of the article in Attic literature.
3 Blomfield’s ἄνευθεν (it was revived by Wecklein in 1888) no longer figures in our texts
as a conjecture to take the place of ἄνωθεν; when Headlam introduced it into the lacuna
which, following Platt, he assumed, he should at least have pointed out that this epic
word (which found its way into lyric too) is nowhere found in Drama.
4 It has been suggested that ἄνωθεν may have referred to the place in which Atreus sits
in the banqueting hall. But this would have needed to be made clear by some sort of addi-
tion. In his prose translation Headlam quite rightly puts a question-mark after the words
‘seated in a high place’, which clearly can only be regarded as a makeshift.

750
 COMMENTARY lines 1594 ff.

stood of Atreus. Even if we do assume provisionally that avöpaxas could be

used with the meaning of καθ᾽ ἑαυτόν (this we shall examine presently), still
the statement that Atreus, the host, is sitting at a separate table would be
pointless: what really does matter is that Thyestes is served with his ghastly
meal apart? from the rest of the company.? In view of these difficulties it is
understandable, in fact almost inevitable, that a lacuna should have been
assumed in the text. Hermann (in Humboldt’s translation) supposed some-
thing (at least a line) to have dropped out after 1594: later he changed his
mind in the belief that the text could be patched up by the conjecture dou":
ὁ δ᾽ and other means. Hermann’s earlier view was revived by Platt (loc. cit.),
whose main points were taken up by Headlam. He refers to the story of
Jezebel (2 Kings ix. 35), of whom the dogs left only ‘the skull and the feet
and the palms of her hands’, and, while appealing to the statement of experts
(‘the hands and feet are exceedingly bitter’), he makes an ingenious endeavour
to show that the process of ‘mincing up’ cannot have been applied to the
extremities, but only to the other parts.” Platt gets rid of the extremities
(together with the heads, which he adds on his own responsibility to the
Aeschylean account)* by inserting ἔκρυψε into the lacuna. To this my answer
is: Aegisthus says 'he minced up the feet and the hands’; and I accept his
statement without committing myself to any speculations about the palatable
qualities of those parts.5 It is not difficult to guess why Platt (and anyone
else who changed ἔθρυπτε to ἔκρυπτε) was so anxious to produce a text
according to which Atreus 'kept back head, hands and feet'. If Aegisthus
said anything to that effect, his account would be in almost complete har-
mony with the story in Herodotus and with the traditional version of the
meal of Thyestes as recorded by later writers, e.g. [Apollod.] Ef. 2. 13 and
the paroemiographi (Zenob. 2. 34, etc., hence Suidas) s.v. Arpews ὄμματα,
cf. O. Crusius, Ad Plutarchi de proverb. . . . lib. comment. (Tübingen Univ.-
Verzeichnis 1895), 26 f. Readers who take a keen interest in irrelevant
details would probably feel gratified if the dramatist told them by what
means Thyestes was made to realize what he had eaten. It does not appear,
however, that Aeschylus was bent upon satisfying this kind of curiosity. For
1 Cf. Accius, Atreus, 217 f. Ribb. ne cum tyranno quisquam epulandi gratia accumbat
mensam aut eandem vescatur dapem.
2 Platt, C.R. xi, 1897, 97 asserts that at the Thyestean banquet there was no one present
except Thyestes, and that this is the only point in which this banquet differs from that of
Harpagus in Herodotus. Aeschylus makes it impossible to think of one guest only by using
the phrase ἀνδρακὰς καθήμ., of which Platt gets rid by means of a conjecture. But even if
these words were not there, it would have to be regarded as most improbable that the poet
would have formed such an idea of the arrangements for the banquet. On the occasion of
the great feast of reconciliation on the xpeovpyóv ἦμαρ it is natural and fitting that a cere-
monial banquet should be given, in which some members of the household and the family
take part and presumably some prominent citizens of Argos as well.
3 Platt's idea that the dish served up to Thyestes consisted of a kind of 'haggis' is highly
unlikely in itself : the guest of honour at a festival banquet had a right to expect substantial
portions of meat for his helping. Perhaps we should not take Platt too seriously when he
indulges in remarks of this kind.
4 Th. Heyse, too, made a conjecture to supply the apparently missing detail. Aeschylus
does not indicate in any way what became of the heads, which are especially mentioned by,
e.g., Seneca, Thyest. 764, and Persius 5. 18.
5 As to pigs’ feet and heads cf. Athen. 3. 96 b-e, also the 4eimvov of Philoxenus (Ath. 4.
147 d) b 31 Diehl.
751
lines 1594 ff. COMMENTARY

him it was enough to say that Thyestes did at last discover the truth, no
matter how. In this connexion we should notice the swift transition from
ἀγνοίαι λαβὼν ἔσθει to κἄπειτ᾽ ἐπιγνούς. Possibly the fact that the heads are
not mentioned is partly due to the deliberate intention of the dramatist that
the audience should not stop to think about the details of the ἐπιγνῶναι.
Aeschylus mentions mincing up the extremities, i.e. making them unrecog-
nizable, in order to show clearly how the deception was at all possible to
start with: this idea, as we have seen, is taken up again in the words ἄσημα δ᾽
αὐτῶν. But it is difficult to see why it is not simply stated that ‘he got rid of
the feet and hands first’, and what is the real point of the extraordinary
detail in θρύπτειν. So I have to admit that there is an important element in
the sentence which I do not understand.
A lacuna after ἄνωθεν has been assumed by Hense (Appendix to Schneide-
win’s edition), and independently of him by Wilamowitz. The latter is
strongly influenced by the version of the story in Herodotus, as can be seen
in general from the supplement which he gives exempli gratia! ἔθρυπτ᾽,
ἄνωθεν (δ᾽ ὀπτὰ περιβαλὼν κρέα κανοῦν ἐπλήρου" ταῦτα δ᾽ ἐν δαιτὸς νομῆι πατὴρ
ἔδεξατ᾽) ἀνδρακὰς καθήμενος, and in particular from the detail of the κανοῦν
(cf. Hdt. 1. 119. 4). So here again we see the tendency to explain precisely
how it was possible for Thyestes to identify the flesh afterwards. Wilamowitz
does not seem to have noticed that his reconstruction of the text does not
produce the result at which he aimed, for if Atreus had had the θρύπτειν (for
the way Wilamowitz translates it see above) carried out, the extremities were
rendered unrecognizable. But in spite of this and other doubts raised by
Wilamowitz’s supplements I think he was probably (I cannot say more) right
in assuming that something has dropped out after ἄνωθεν. Before we can
consider what may have been the gist of the lost words we must examine
more Closely the expression ἀνδρακὰς καθήμενος.
The only other passage where the adverb dvdpaxds is found in extant
literature is v 14: ἀλλ᾽ dye of δῶμεν τρίποδα μέγαν ἠδὲ λέβητα avdpaxds. The
context of the passage (with which has been rightly compared the parallel
list of presents in @ 392 ff.) shows the correctness of the explanation κατ᾽
ἄνδρα (i.e. viritim), which is presupposed in Plutarch, Sept. sap. 6 (Mor.
P. 151 e) and is actually given in the lexicographers (Hesychius, Etym. M.
s.v.; Anecd. Bachm. p. 86. 27 = Phot. Berol. and Suidas s.v.) and in the
varia lectio ἄνδρα καθ᾽ in v 14.2 As the lexicographers tell us that Cratinus
used ἀνδρακάς in the sense of κατ᾽ ἄνδρα or χωρίς in the Βονκόλοι (fr. 19 K.),
though they do not quote the comic poet's actual words, it seems likely that
there the meaning was the same as in the passage of the Odyssey? It there-
fore seems doubtful whether we may credit Aeschylus with having used
ἀνδρακάς — καθ᾽ ἑαυτόν. Triclinius (or one of his Byzantine forerunners), who
gives this as a gloss on the word here, had possibly been trying to make the
best of a bad job, as he did elicit from the text he had in front of him a

1 Wilamowitz's idea of the manner in which the dish was arranged is apparently the same
as that of Enger (in his revision of Klausen) : "Videtur Atreus pedes et manus texisse super-
positis (ἄνωθεν) carnibus, quibus comesis ipse [Thyestes] rem agnovit’. |
2 The explanation of Phrynichus, fr. 20, p. 135 de Borries, κατὰ δέκα ἄνδρας obviously
cannot be entertained.
3 Or does xwpis cover the meaning καθ᾽ ἑαυτόν ἢ

752
 COMMENTARY lines 1594 ff.

meaning which was at least possible in a general way. Wilamowitz thinks

this interpretation may go back to an ancient tradition. In order to reach
agreement with the only sense of ἀνδρακάς that is found elsewhere, Casaubon
wrote καθημένοις ; the necessity for this emendation or something similar has
often been maintained, with great vehemence by Schoemann, Ofusc. iii. 156:
' ἀνδρακὰς καθῆσθαι de uno graece non magis quam viritim sedere latine dici
potest'. On the other hand, Wilamowitz with Paley and others has adopted
the interpretation of Triclinius ; similarly, e.g., Schwyzer, Griech. Gramm. i.
630. 4. I do not feel certain about this, but I do not venture to deny the
possibility of its being right, particularly as we cannot be sure that some
later epic poet, between the Odyssey and Aeschylus, may not have twisted
this extraordinarily rare word into a meaning different from that which it
bears in the Homeric passage.
Assuming, then, that ἀνδρακὰς καθήμενος is the genuine reading and that it
is said of Thyestes, we proceed tentatively to outline what is lost in the lacuna
between ἄνωθεν (which, as it stands, has nothing to which it can be referred)
and ἀνδρακὰς καθήμενος (which requires a previous mention of Thyestes, as
does the following clause ἀγνοίαι λαβὼν ἔσθει «rÀ.). With the greatest reserve
I would suggest some such thought as: “The extremities he broke up small on
top (of the dish, and with this Thyestes was served, Thyestes> who was
sitting apart by himself . . .”. This reconstruction is based on the assumption
that joints, or portions, were sometimes served up 'garnished' with a top
layer of minced meat made up of the less substantial portions, or of a ragout
of poultry, hare, or something of the kind. I have, however, been unable to
discover anything bearing on this in Athenaeus or Pollux, though they have
plenty to tell us about stuffing and forcemeat. There is therefore no solid
foundation for my suggestion. From every point of view ἄνωθεν appears to
be the most obscure part of this obscure sentence: the reference of the word
was perhaps made clear in a way which we cannot guess by something that
came after it in the lacuna. However, we cannot exclude the possibility of
a corruption. If we may suppose that the lacuna contained some such phrase
as "Thyestes was served with it', there would be no difficulty in understanding
the object to be the whole dish, and not just the minced portions expressly
mentioned. In any case the phrase must have been put in such a shape that
the edge of the antithesis of rà μὲν ποδήρη κτλ, and ἄσημα δ᾽ αὐτῶν was not
completely blunted. To go back to ἀνδρακὰς καθήμενος : 11 we make the con-
trary assumption that ἀνδρακάς cannot be applied at all to a single individual,
we are obliged to evolve some more complicated explanation such as, e.g.:
‘Atreus . .. had it served up to Thyestes, to him alone, for the guests were
sitting each by himself (ἀνδρακὰς καθημένων). All this is entirely hypothetical
and very far from satisfying.
Now for one or two details. In 1594 the meaning of ποδήρη is completely at
variance with its ordinary use, see besides the dictionaries Schuursma 161 f.,
cf. also above on 898, where, however, the deviation from the ordinary use is
not so violent. Here we should bear in mind that ‘-npns gradually acquired
the nature of a suffix', cf. on 1442 f. (p. 682) and Blass on Cho. 1029 ff. with
regard to Eum. 763 πλειστήρη χρόνον. The commentators cannot make up
their minds as to whether the whole feet are meant or just the toes. The
former is the more probable: the breaking up of the toes alone would have
4872.3 T 753
lines 1594 fl. COMMENTARY

been of no use, for a human foot would be easily recognizable even without
them. As regards χερῶν ἄκρους krévas, it is possible that the gloss in Hesychius
κτένας" τοὺς τῶν χειρῶν καρποὺς Kal τῶν ποδῶν goes back to an explanation of
this passage, as was assumed long ago. But the explanation of the lexico-
grapher is not quite correct, for the καρποί of the hands cannot possibly be
described as ‘combs’.’ There is, however, an informative passage in Pollux
2. 144 ἕνιοι δὲ τὸ μὲν πρόσθιον τῆς δρακὸς πᾶν θέναρ οἴονται καλεῖσθαι... τὸ δὲ
ἀντικείμενον πᾶν ὀπισθέναρ (cf. 143) ἢ κτένας, where κτένες very appropriately
denotes the lower part of the back of the hand with the base of the fingers.
Thus Ag. 1594 evidently refers to the part of the hand next to the fingers
together with the fingers, and not the fingers by themselves.
Leaving the reconstruction of the story aside, the most significant detail
in this sentence is ἀνδρακὰς καθήμενος, no matter what ending of the participle
we regard as the most likely. In καθήμενος the contrast with the established
practice of the fifth century (and a somewhat earlier period as well) is un-
mistakable, the same contrast which we find contained in a nutshell if we
compare the invitation in Euripides (fr. 691 N.) κλίθητι καὶ πίωμεν with the
Homeric account (v 136) οἶνον μὲν γὰρ πῖνε καθήμενος. This marked difference
of practice must have struck any Athenian who read his Odyssey with the
slightest degree of attention, just as it struck the later grammarians and
antiquarians.? It is characteristic of Aeschylus that he makes deliberate use

! Inthe Etymol. Gudianum the following gloss of Seleucus is preserved (cf. Reitzenstein,
Gesch. d. griech. Etymologika, 162, no. 43): «reis (xéris codd.): ἡ τοῦ ποδὸς 4 τῆς χειρὸς...
ἃς κτένας φαμέν, Unfortunately the keyword to the explanation is missing, but it is obvious
that it was not καρπός.
2 Athen. (Epitome) 1. 11 f εὐωχοῦνται δὲ map! ᾿Ομήρωι καθήμενοι. οἴονται δέ τινες καὶ
ἑκάστωι τῶν δαιτυμόνων κατ᾽ ἄνδρα (the ἀνδρακάς of Aeschylus) παρακεῖσθαι τράπεζαν, and
again (Epit.) 1. 17 f καθέζονται δ᾽ ἐν τοῖς συνδείπνοις οἱ ἥρωες, οὐ κατακέκλινται, 5. 192 €
ἐκαθέζοντο δὲ καὶ δειπνοῦντες οἱ τότε. πολλαχοῦ γοῦν à " Ounpós φησιν κτλ., 8, 363 f ὅθεν οὔτε
κατακλίνεσθαι παρὰ τοῖς ἀρχαίοις ἔθος, ἀλλὰ ‘ δαίνυνθ᾽ ἑζόμενοι ᾽ (y 471), οὔτ᾽ εἰς μέθην πίνειν
(cf. 1. 10 e sqq., 17 f, immediately before the passage just quoted, also 2. 40 c) κτλ. I suggest
that all these passages may be traced back to Seleucus. ‘We know the keen interest taken
in the Homeric banquets by Seleucus, on whose materials Athenaeus often draws’ (Reitzen-
stein, Gesch. d. griech. Etymologika, 171). There are certain details which support the
assumption of Seleucus as authority. The expression in 8. 363 f οὔτε κατακλίνεσθαι παρὰ τοῖς
ἀρχαίοις ἔθος corresponds exactly to that in (Epit.) 2. 40 c Σέλευκος δέ φησι τὸ παλαιὸν οὐκ
εἶναι ἔθος οὔτ᾽ οἶνον ἐπὶ πλεῖον, . . προσφέρεσθαι κτλ. (= fr. 78 in Maximilian Mueller, De
Seleuco Homerico, diss. Göttingen 1891). Further, Reitzenstein (loc. cit.) traces to Seleucus
the origin of the preceding section 362 e to 363 b (the Ephippos fragment which follows may
have been added by Athenaeus from his own reading). The passage quoted above, 5. 192 e
ἐκαθέζοντο δὲ καὶ δειπνοῦντες οἱ τότε, 15 the continuation of 192 Ὁ πᾶσα δὲ συμποσίου συναγωγὴ
παρὰ τοῖς ἀρχαίοις τὴν αἰτίαν εἰς θεὸν ἀνέφερε κτλ. Now this passage, which is very closely
connected in subject-matter with the section 2. 40 c, known from Athenaeus to have come
from Seleucus, was traced back to Seleucus by Bapp, Lezpz. Studien z. class. Philol, viii.
265, and by H. v. Arnim, ‘Quellenstudien zu Philo v. Alex.’ (Philol. Unters. xi. 1888) 123,
and this was certainly right, although v. Arnim goes too far in his conjectural attribution
to Seleucus of the whole comparison of the Homeric banquet with that in Xenophon etc.
(cf. M. Mueller, op. cit. p. ron. 1), Robert Weber, ‘De Dioscuridis . . . libello' (Leipz. Stud.
2. class. Philol, xi, 1889), 110, 179, has traced the passages quoted here from Athen., 1. 11 f
and 5. 192 e, to that book on the subject of Homeric customs which was so often used by
Athenaeus (‘eine Compilation von Homererklärungen’, Ed. Schwartz), a book to which it
would be better not to affix the label ‘Dioskurides’ (cf. Schwartz, RE v, 1128 f.). Yet Weber
himself (p. 180) derives the closely connected passage Athen. 8. 363 f from Seleucus and
admits (p. 171 f. and 183) in the case of 5. 192 b-e the possibility of its being a miscellany of

754
 COMMENTARY line 1600

of the discrepancy between the customs of Homeric society and those of his
own time. We have here another instance of the ‘Homerizing’ of the cultural
background which we have noticed first in Iphigenia’s presence at the men’s
banquet and then in the motif of the fateful bath, the pivot of the plot. Just
as the Homeric word dvépaxds here points to a Homeric situation, so the
participle καθήμενος reminds us of the passages in the Odyssey in which it is
used in connexion with drinking and eating, cf. 8 238, ἡ 203, p 478, v 136, ¢ 89.
1597. ἔσθει. Here and again in 1599 ἀμπίπτει we find the present tense,
whereas elsewhere in this narrative a past tense is used. Cf. Pers. 191 f. and
above on 1383.
ἄσωτον γένει: ‘making a good outcome impossible for the race’ (for the
‘active’ function of ἄσωτος cf. on 238).
ὡς ὁρᾶις, not only because Agamemnon’s dead body is lying before their
eyes, but also because of the whole situation that has now developed. For
the fact that Aegisthus is addressing the Chorus, see on 1582 and on 1603.
1598. The rare adjective καταίσιον could at any time be formed by ‘hypo-
stasis’ (cf. p. 62) from Homer’s κατ᾽ αἶσαν (Blomfield), as, e.g., καταθύμιον
from κατὰ θυμόν. I cannot see that this is a case of ‘intensifying xara-’ (as
E. Schwyzer says in Mélanges de linguistique offerts à F. de Saussure, 1908,
254).
1599. ἀπεράω (here used in ‘tmesis’) is not found elsewhere till later, ἐξεράω
being common in the fifth century. Cf. Hesychius ἐρᾶσαι: κενῶσαι.
1600. As Wilamowitz’s treatment of this line and the two following (except
for his alteration of ξυνδίκως) seems to me to be final, it will be best to quote
his note in full: ‘dittographiam ad v. 1601. o2 delevi; 1602 deleverat Schütz,
nam eandem gentem duobus nominibus eodem loco appellare vesanum est.
Thyestes agnito fratris scelere ac suo errore reicit se in lectum [‘sellam’
would be more correct ; καθήμενος here recalls the κλισμοί τε θρόνοι re of the
Odyssey], nempe ut caput obnubat [there is nothing about this in the text],
pede autem proculcat mensam, vociferatus "ita ut haec mensa et fratris et
mea familia pereant". ergo recte dicitur diris proculcasse mensam, versus
autem 1602 necessarius est.' It matters little that J. C. Schmitt, as can be
seen from Wecklein's Appendix, had already deleted 16oo (he is wrong in his
interpretation of the following words). A far more disturbing feature than the
juxtaposition of ‘Pelopidae’ and ‘the race of Pleisthenes' is the weakening
anticipation of the curse (in the form in which we read the curse in 1602 it is
obviously provoked by the λάκτισμα δείπνου mentioned in 16or). Another
extracts partly from ‘Dioskurides’ and partly from Seleucus. If we add the indications
noted here, it seems more probable that 41} these passages come originally from Seleucus
and were in some cases put into the context of the ‘Dioskurides’ by Athenaeus.
1 I find it quite impossible to accept the representation of the action given, e.g., in Sidg-
wick's note ('spurning the banquet to aid his curse") and in Mazon's translation : 'Sur les
Pélopides il appelle alors un destin d'horreur, et, pour accompagner son imprécation,
renversant la table du pied... .’ It seems obvious rather that the upsetting of the table is
not done with deliberate purpose (‘pour accompagner), but involuntarily, as the result of
the violent movement with which the horror-stricken man flings himself back on his chair
so that he loses his balance and kicks over the table standing just in front of him. (In the
same way when Zeus has human flesh served up to him in Lycaon's house, μυσαχθεὶς τὴν...
τράπεζαν ἀνέτρεψεν, [Apollod.] Bibl. 3. 99.) This then becomes for him a token of doom:
when uttering the curse that fills his soul, he uses this external symbol, as is rightly under-
stood by Plüss ad loc. I know of course that in many cases of magic, according to the 'law

755
line 1600 COMMENTARY

awkwardness is that in the text as it stands the description of closely con-

nected actions ἀμπίπτει. . . λάκτισμα τιθείς is interrupted by the clause
μόρον... ἐπεύχεται. What induced somebody, whether the producer of a
later performance or a reader, to insert 1600 next to 1601, or instead of 1601 f.
(as Wilamowitz supposes), it is impossible to tell. Did the man perhaps find it
difficult (in spite of 1569) to accept the obscure Pleisthenes-genealogy? Or,
as I think is more likely, did someone feel the need of an explicit verbum
dicendi! to lead up to 1602, though the implicit indication in 1601 should have
been enough for an intelligent audience? If the latter supposition is right,
1600 was not intended to be a substitute for the passage 1601-2 but to be
added toit. After Wilamowitz’s successful attack upon line 1600 I should not
like to do what the poet calls τὸν πεσόντα λακτίσαι πλέον. But the question
must be asked whether μόρον ἄφερτον is really an admissible expression here.
The only possible meaning of the words (cf. note on 1146 μόρον), if Aeschylus
had written them, would be ‘mortem intolerabilem Pelopidis imprecatur’
(Stanley), not ‘an intolerable fate’ (Paley), ‘un destin d’horreur’ (Mazon).
and the like. But is it appropriate to use ‘mortem intolerabilem’, in itself a
strange phrase, in this passage where, as the following words show, the speaker
is concerned with the extinction of the race and not with the kind of death its
individual members are to suffer? In an authentic passage of Aeschylus,
Cho. 441, where we find the word ädepros connected, though in a different
way, with μόρος, the expression has its full force. The meaning there of
μόρον κτίσαι μωμένα ἄφερτον αἰῶνι σῶι is of course not as paraphrased by the
scholiast (who misunderstood the Aeschylean use of μόρος), δυστυχίαν μεγίσ-
τὴν κατασκευάζουσα τῶι ad βίωι (similarly Schadewaldt, Hermes, Ixvii, 1932,
340: ‘da sie strebte ein untragbar Los deinem Dasein zu schaffen’), but
‘Klyt. wollte dadurch den Tod des Ag... . zu einer dauernden Last für den
Sohn machen’ (Blass; Conington and Tucker, too, interpret it correctly, as
does Headlam : ‘to make the manner of his end more dreadful than thy life
could bear’). It is possible that the interpolator of Ag. 1600 used μόρος in
the sense of fatum. As to ἐπεύχεται, it is perhaps only a coincidence that the
interpolator? of Sept. 276 also ends his line with ἐπεύχομαι: he may have
taken the word from 1. 279, as Ritschl suggested.
1601. ξυνδίκως τιθεὶς ἀρᾶι: ‘as he makes the upsetting of the meal (and the
table) of equal right with the curse’. To our way of thinking the words
should be ‘as he makes the curse of equal right with the upsetting’, but the
of homoeopathy’, the analogous act is deliberately performed (Juppiter populum Romanum
sic ferito ut ego hunc porcum hodie feriam and the like).
1 This, as is well known, is a frequent cause of interpolation : a classic example is afforded
by E. Phoen. 1225 (with the scholia), see also my remark (above on 527, p. 266 n. 3) about
Sept. 426. Cf., e.g., K 191, P 585, 0 27, « 265, Virgil, Aen. 2. 76 and 10. 278. That the complete
expression was felt to be the normal thing is illustrated, e.g., by the Schol. on A. Prom. 674
Weckl. (647): λείπει τάδε λέγουσαι, Schol. E. Or. 1447 λείπει τὸ καὶ τοιαῦτα ἔλεγεν.----ἰ am
inclined to believe that Verrall was right in bracketing Sept. 579 καλεῖ" λέγει δὲ τοῦτ᾽ ἔπος
διὰ στόμα (cf. my note on Ag. 605), although some of his arguments are so flimsy as to be
fair game for Headlam (On Editing Aeschylus, 74 ff.). It seems worth while to compare the
arrangement of the whole passage with the similar one in Sophocles’ Eurypylus, fr. 210.
0-83 P.
’ 2 Wilamowitz did not notice that Sept. 276 had already been deleted in 1857 by Ritschl
(Opusc. i. 367 ff.), who was followed by Wecklein, Studien zu Aeschylus, 53 f., and in his
commentary. See further Wilamowitz, Interpr. 107.

756
 COMMENTARY line 1602

result comes to the same thing, and the expression as it stands is in keeping
with a widespread habit of Greek and Latin.’ With all its terseness the sense
seems to me to be perfectly clear. ‘What is happening now to the table and
the meal is to happen in the same way and on the ground of just such a
lawful right to the whole race of Pleisthenes.' The monstrous meal contrived
by crime is dashed to the ground, and in this action δέκη is taking effect ; an
equal or an associated (συν-) δίκη is to bring about the overthrow of the
accursed race. To alter ξυνδίκως to ξυνδίκωι, as Jacob and Wilamowitz have
done (while others read ξύνδικον), is objectionable as spoiling a construction
which, though by no means of common occurrence, is actually found several
times in Tragedy and particularly in Aeschylus (cf. on 913 θήσει δικαίως).
The objection that if ξυνδίκως is retained the noun ‘ dpa has no attribute
such as we should like to find with it’? is not conclusive, for in the following
line the content of the curse is made clear. Others do not take ξυνδίκως
τιθείς together, but take λάκτισμα τιθείς as a periphrastic expression (as, e.g.,
66 παλαίσματα θήσων, 1059 μὴ σχολὴν τίθει) and ξυνδίκως ἀρᾶι = σὺν ἀρᾶι,
thus Headlam (who offers besides another rendering for consideration):
‘giving a kick to the laden board together with his imprecation’. This seems
to me not impossible, but I doubt whether it brings out the full force of
ξυνδίκως.
1602. οὕτως emphasizes the operating of the ‘law of homoeopathy’ (in this
particular case? there is of course no question of a magical ceremony), cf., e.g.,
T 300 f. ὧδέ σφ᾽ ἐγκέφαλος χαμάδις ῥέοι, ὡς ὅδε οἶνος, αὐτῶν καὶ τεκέων, Theocr.
2. 26 οὕτω τοι καὶ Δέλφις ἐνὶ φλογὶ σάρκ᾽ ἀμαθύνοι and often elsewhere. Some
commentators object that Thyestes includes himself and his descendants in
the curse, but Wecklein says rightly : ‘this indicates excess of grievous excite-
ment.’ Such imprecations, involving the speaker himself and his race, are, as
a rule, pronounced by a man who is sure, or who feels sure, that the essential
! In the ancient languages, as is well known, we not only find ‘to exchange A with B’ in
many cases where we expect ‘to exchange B with A’ (Velox amoenum saepe Lucretilem
mulat Lycaeo Faunus and the like), but also the same process applied in expressions of
comparison. Thus, e.g., Caesar, B. Gall. 1. 31. 11 neque enim conferendum esse Gallicum cum
Germanorum agro, where Kraner-Meusel’s commentary remarks ‘as the German territory
is the less valuable, one would expect it to be put the other way round "Germanorum agrum
cum Gallico", and quotes parallels from Cicero. Here, as usually happens, it is the thing
nearest to the speaker which is put first (Horace's estate lies in the neighbourhood of
Lucretilis, and not in Arcadia). Commenting on the sentence in Pericles’ Funeral Oration,
"huc. 2. 42. 2 καὶ οὐκ ἂν πολλοῖς τῶν ᾿Ελλήνων ἰσόρροπος ὥσπερ τῶνδε ὁ λόγος τῶν ἔργων
φανείη, Wilamowitz remarks (Griech. Lesebuch, ii. 98 f.) : ‘in the case of all others deeds are
inferior to words. In comparisons of this kind we are accustomed to put the two values in
the reverse order from that of the ancient languages.’ Pericles gives first place to ὁ λόγος,
for he is delivering a speech. Ag. 1601 also starts from what lies before the speaker’s eyes;
the λάκτισμα δείπνου takes first place in the course of the events, the curse is secondary.
2 Thus Daube, 6 n. 20. He rightly rejects Wilamowitz’s artificial interpretation, ‘ “cum
communi devotione" utriusque scilicet familiae’, but he defends the conjecture ξυνδέίκωε.
'There is no reason to try to find here, as some commentators do, an allusion to the technical
use of ξύνδικος.
3 It very naturally happens elsewhere too that in an imprecation motifs to be found in
*homoeopathic magic’ are employed in cases where there is no idea of a magical ceremony.
Cf., e.g., S. Aj. 839f. and 1177 ff. : in the latter passage the offering of the lock to the dead man
is made for its own sake, but at the same time it suggests to the speaker the idea of a
‘homoeopathic’ curse, and, as in the words of Thyestes, the object of the imprecation is
*he and all his race’.
757
line 1602 COMMENTARY

conditions for the fulfilment of the curse are not applicable to his own case:
εἰ δ᾽ ἐπιορκοίην, ἐξώλης εἴην αὐτὸς καὶ τὸ γένος and the like.
Πλεισθένους : cf. on 1569, p. 740.
Later Tragedy made much of the Thyestea execratio (Cic. in Pis. 43), the
Thyesteae preces (Horace. Epod. 5. 86). Some sentences of these curses, from
the Thyestes of Ennius, are preserved in Cicero, Tusc. ı. 107 ; the original was.
probably, though not certainly, the Thyestes of Euripides.
1603. ἐκ τῶνδε: cf. on 1223.
1603. σοι: ‘Dubitari potest scribendumne sit ἐκ τῶνδέ roi’ (Hermann). This
is possible (cf., e.g., Pers. 287 μεμνῆσθαί τοι πάρα), but not necessary. It may
be preferable to follow Hense, who remarks on 1597 ὡς ὁρᾶις : ‘Aegisthus is
addressing the coryphaeus as in 1603, while pointing to the corpse of Agamem-
non. Such an address . . . imparts a more personal character to his account
and prepares in a natural manner for the intervention of the coryphaeus.’
1604. δίκαιος : cf. on 1577.
pabeus: H. Frankel, Glotia, xiv, 1925, 3 f., points out that ‘crafty plotting
is often denoted by the word ῥάπτειν ' ; cf., e.g., y 118 f., m 379 οὕνεκα of φόνον
αἰπὺν ἐράπτομεν, A. Cho. 221, where μηχανορραφῶ picks up the δόλον πλέκεις of
the preceding verse, Eum. 26 λαγὼ δίκην Πενθεῖ καταρράψας μόρον. Cf. Festus,
p. 310. 27 M. sutelae dolosae astutiae, a similitudine suentium dictae sunt, and
Lindsay’s commentary on Plaut. Capt. 692, F. Marx on Lucilius 747. One
would like to know whether ῥαφεύς was already in Aeschylus’ day used to
denote a stitcher or patcher (Pollux 7. 42) ; if so, we have here one of his many
recoinings.
1605. For more than a hundred years it has been thought necessary here to
correct not an error of a copyist, but the thought of the poet. The same
tendency still prevails with the majority of the editors.’ There is no lack of
ferocity in the attacks (‘ungeschickt und unsinnig’, Wecklein; ‘absurd’,
Verrall; ‘ridiculous’, Headlam); the chief objections are summarized by
Sidgwick thus: ‘the only traditions we have give at most three children to
Thyestes, and the mention of the number seems so needless and inappro-
priate.’ This last is a strange assertion ; does it not heighten the horror of the
bloody deed to say that twelve children were butchered and the thirteenth
alone, a babe still in swaddling-clothes, escaped with its bare life? As for the
former argument, it would be a hazardous undertaking to attempt to bring
every Aeschylean version of a legend into accord with the later vulgate
tradition. All that has to be said here has been said by Wilamowitz:
1 Quite recently (1940) in his obituary of D. S. Margoliouth (Proceedings of the Brit.
Acad. xxvi), p. 4, Gilbert Murray went out of his way to call the conjecture μ᾽ ἔλιπε κἀθλίωι
‘almost certainly right’. The extensive discussion of the passage by I. E. Harry, Rev. de
phil. lviii, 1932, 361 ff. does not make any helpful contribution.
2 We have no evidence earlier than imperial times for the number or the names of
Thyestes’ children. The late schol. E. Or. 812 (vol. ii, p. 210 f. Dindorf), which led Furt-
wangler in Roscher’s Lexik. d. Mythol. i. 714, along with earlier scholars, to ascribe to
Sophocles the mention of Aglaos, Orchomenos, and Kaleos, has since ingloriously dis-
appeared; Ed. Schwartz did not print it at all (A. C. Pearson, The Fragments of Soph. i,
P. 92 f., is unduly lenient) ; C. Robert, Heldensage, 296 n. 4, of course takes no account of it.
Robert himself, however, quotes schol. Eur. Or. 4 as evidence that Hellanikos mentioned the
three sons of Thyestes, viz. Orchomenos, Aglaos, and Kaleos. But ‘the ascription to
Hellanikos of schol. Eur. Or. 4 with its detailed genealogy is entirely problematical’ (F.
Jacoby in his commentary on F Gr Hist 4 F 157, vol. i. 470, with full discussion).

758
 COMMENTARY line 1605

‘tredecim filios Thyestae invident critici, quia ceteroquin duo vel tres nume-
rantur, quasi nostrum esset fabularum probabilitati prospicere,! aut larga
suppeteret mythographorum memoria. immo qui fabellis popularibus student,
bene norunt haud raro unum qui servatur aut duodecimum esse aut decimum
tertium." For the view expressed in the last sentence cf. Welcker, D. Aeschyl.
Trilogie, 21, where he terms thirteen ‘a poetic number’, instancing Ag. 1605.?
There is no need to go into details about the part played by the number
‘thirteen’ in the legends and fairy tales of many peoples.® So far as concerns
ancient Greek ways of thought, what is essential is the fact stressed by Otto
Weinreich in his article ‘Lykische Zwölfgötter-Reliefs’, Sttzgsb. Heidelb.
Akad., Phil.-hist. Kl. 1913, 5. Abh., 35 ἢν, and his book 'Triskaidekadische
Studien’, Religionsgesch. Versuche u. Vorarbeiten, xvi, 1916, τό n. 1, in opposi-
tion to other scholars,* that among the Greeks the view that thirteen was an
unlucky number by no means prevailed ;? it was rather an ‘Uberschusszahl’
(‘excess-number’), in itself neither lucky nor unlucky, but denoting simply
the accession of one extra to a dozen. Accordingly in one case it may be
related that twelve men were killed and the thirteenth with them, in another
that the thirteenth alone escaped the fate of the twelve. In the Doloneza it is
told of Rhesus that he was the thirteenth man slain (K 495), the thirteenth in
addition to twelve others (K 560 f.). We do not know the context of Pindar
fr. 171 Schr. κατὰ μὲν φίλα τέκνα ἔπεφνεν θάλλοντας HBat δώδεκ᾽, αὐτὸν δὲ τρίτον.
According to Hesiod (fr. 147 Rz.), followed by Pindar, Ol. τ. 79 f. and in the
Threnoi (fr. 135 Schr.), Oinomaos killed thirteen suitors of his daughter
Hippodameia (cf. C. Robert, D. griech. Heldensage, zıı n. 4); according to
(Apollod.] Epit. 2. 5 τινὲς λέγουσι δώδεκα : this version made the lucky Pelops
the thirteenth. That thirteen is not in itself an unlucky number is clear from,
e.g., θ. 390 f. (Alcinous) δώδεκα γὰρ κατὰ δῆμον ἀριπρεπέες βασιλῆες ἀρχοὶ
κραίνουσι, τρισκαιδέκατος δ᾽ ἐγὼ αὐτός and r 202: for twelve days the north
wind kept the Achaeans from setting sail, τῆι τρισκαιδεκάτηι δ᾽ ἄνεμος πέσε,
τοὶ δ᾽ avayovro. Herakles, the liberator of Prometheus, is to be τρίτος ye
γένναν πρὸς δέκ᾽ ἄλλαισιν γοναῖς (A. Prom. 774). Thirteen children played a
part in the story, unknown to us, which was told by Antimachus of Colophon,
fr. 178 Wyss: τρισκαιδεκάτην τέκνων καὶ iv λελοχυῖα.7
Furthermore, the present passage ought to be protected from wanton
1 The reference is clearly to G. Hermann, who says: ‘Sed habuerit ille tredecim filios, etsi
nihil de isto numero proditum est, certe ad summum tres mactari potuerunt, quod reliquos
iam adolevisse necesse erat’, etc.
2 Later, however, Welcker changed his mind and regarded ἐπὶ δέκ᾽ as corrupt (Rhein.
Mus. ix, 1854, 203).
3 E. Böklen’s book, ‘Die “Unglückszahl” Dreizehn und ihre mythische Bedeutung’
(Mythol. Bibliothek, v. 2, 1913) is inaccessible to me; for its thesis cf. Latte, Deutsche
Literaturzeitung, 1914, 1939 f. In Hoffmann-Krayer’s Handwörterbuch des deutschen Aber-
glaubens s.v. ‘Dreizehn’ the reader is referred to ‘Zahlen’, s.v. ‘Zahlen’ (in the fascicle
published in 1938, the last available in the Bodleian Library) to the supplements.
+ E.g. Postgate in his very sketchy observations C.R. xix, 1905, 437 f. and xx, 1906, 443.
5 In the article ‘Magic’ in the Encyclopaedia Britannica (11th ed., xvii. 309) it is stated
that the idea of ‘unlucky thirteen’ is ‘of recent date’.
6 Unfortunately, Weinreich did not go into the question of A. Ag. 1605 and other instances
of the number thirteen in Greek heroic legends.
E. Petersen, D. attische Tragödie, 643, asserts in the course of a polemic against Wilamo-
witz that ‘the legends of thirteen sons, etc., have no relevance to Aegisthus’, but he gives
no grounds for his assertion.

759
line 1605 COMMENTARY

tampering not only by its content but also by its form. τρίτον ἐπὶ δέκα is an
instance of an old idiom, of which it is hard to believe that the occurrence
here is due to a corruption. This use was briefly referred to and set against
the background of linguistic history by Wackernagel (Festschrift Gustav Binz,
Basel 1935, p. 35 n. 2): ‘Praepositioneller Anschluss der Zehn findet sich
vereinzelt auch in den klassischen Sprachen, hauptsächlich bei den zu den
Additivzahlen gehörigen Ordinalien (mit Einschluss der kalendarischen
Tagbezeichnungen,wie attisch τετράδι ἐπὶ δέκα, ‘am vierzehnten Monatstage’’).’
Meisterhans, 163, gives instances from inscriptions of this way of designating
the day of the month, ἄχρι τῆς τρίτης ἐπὶ δέκα, πέμπτηι ἐπὶ δέκα, etc. ; add, e.g.,
Demosth. 19. 57 ἐλαφηβολιῶνος ἐνάτηι ἐπὶ δέκα, 19. 58 τρίτηι ἐπὶ δέκα τοῦ
σκιροφοριῶνος μηνός, etc. Apart from its use in dating this type of numeration
may be illustrated by the following passages, which I have picked up at
random: Kleostratos fr. ı. ı (Diels-Kranz, Vorsokr. i, sth ed., 42) τρίτον ἦμαρ
ἐπ᾿ ὀγδώκοντα, Pind. N. 6. 58 πέμπτον ἐπὶ εἴκοσι τοῦτο γαρύων εὖχος ἀγώνων
ἄπο, Eupolis fr. 276. 2 K. 6 στιγματίας τέταρτός ἐστιν ἐπὶ δέκα.ἷ
1606. There are only three other instances in Drama of the epic word τυτθός,
two in Aeschylus (Pers. 564, fr. 337 N.) and one in the Eurypylos of Sophocles
(fr. 210. 5r P.). Here there is a clear connexion with the Homeric formula
(ert) τυτθὸν ἐόντα, as there is in Pers. 564 (notice the context) with O 628
τυτθὸν γὰρ ὑπὲκ θανάτοιο φέρονται (scil. ναῦται).
1607. For the thought, see note on 911. On κατάγειν cf. on 1283.
1608. ‘Mire interpretantur θυραῖος wy de Aegisthi exilio pristino [so also
Verrall] aut de Agamemnonis absentia, cum hoc dicat Aegisthus, se foris
versantem caedi quidem non adfuisse, sed tamen quasi manus intulisse
Agamemnoni insidias fatales excogitando struendoque' (Weil).
1609. ‘ δυσβουλία dici solet de malo consilio, quod ei qui hoc cepit perniciosum
est, ut Sept. 802, S. Ant. 95, 1269. Hoc loco de consilio alii pernicioso'
(Klausen).
1610. οὕτω: ‘that being so’; a conclusion is drawn and a final point made;
so, e.g., Suppl. 771, Eum. 739. For the thought cf. 539, 550.
1611. ἰδόντα after ἐμοί: on this type of construction, illustrated by Elmsley,
cf. on 378-80. In the present passage the idiomatic construction was des-
troyed by Triclinius as it is in Cho. 411 in Murray's edition.
τῆς δίκης ἐν €pkeow: here again we find the motif of the net. Note, too,
that the end of the speech takes up the beginning almost word for word (1580
ἰδὼν ὑφαντοῖς ἐν πέπλοις "Epwiwv)—a favourite device of Aeschylus (cf. on
1196), here used in a particularly effective way.
1612. ἐν κακοῖσιν has been misunderstood by many translators; e.g. Hum-
boldt, ‘bei Freveltat’; W. Sewell, ‘in deeds of guilt’; Wilamowitz, ‘nach dem
Frevel’; Mazon, ‘dans le crime’; another interpretation, also wrong, is
Verrall’s ‘to insult a wretch’ (cf. on 765). ἐν κακοῖσιν must obviously mean ‘in
the midst of evils’, ‘in misfortune’. The sense of the expression is always the
same, cf., e.g., Sept. 187, 227, Ag. 765 (see note), Cho. 222, Eum. 276, S. Aj. 532,
1118, Oed. R. 127, El. 335, 1287, 1329, Phil. 471, Oed. C. 592, etc. S. Aj. 1151
ὃς ἐν κακοῖς ὕβριζε τοῖσι τῶν πέλας is closely akin to our passage in thought as
1 A remarkable early instance of the corresponding use in the case of cardinal numbers is
provided by the Corinthian inscription of the early sixth century on the back of a sima-
block from Calydon : μία ἐπὶ ρίκατι, cf. H. Payne, Necrocorinthia, 258 n. 1.
760
 COMMENTARY line 1612
well as expression. Here the κακά are the misfortunes of the royal house and
the miserable death of Agamemnon.
οὐ σέβω has been rendered by most editors in a manner which certainly
tallies with the general sense of the passage, but leaves it at least doubtful
whether the peculiarity of the expression has been adequately grasped; e.g.
Stanley 'haud laudo', Blomfield and Hermann non probo,! Peile and Paley ‘I
do not approve of’, Kennedy ‘I respect not’, L-S ‘I do not respect, approve
it’. These translations depart substantially from the normal use of σέβειν.
Sidgwick’s linguistic instinct was, as so often, on the alert, but finally he
acquiesced in the conventional expedient ‘ σέβω, "honour", understatement
for "approve".' This interpretation is rendered still more improbable by the
assumption that the action denoted by the infinitive which is the object of
σέβειν is performed by others and not by the σέβων himself. This seems to be
the view of the translators quoted above and of most others; it appears in
so many words in the following and other versions: S. Butler (quoted by
Peile), ‘eos non amo qui in aliorum rebus adversis insolenter se gerunt' ; Lewis
Campbell, ‘I have no respect for the man who insults the unfortunate’ ;
Platt, ‘I suffer not him that insulteth the dead’; Mazon, ‘l'insolence . . . me
révolte'. In every other case where σέβειν (and σέβεσθαι) governs an infinitive,
the infinitive denotes an action or a state of the σέβων, e.g. Eum. 749 τὸ μὴ
ἀδικεῖν σέβοντες, Pers. 694 f. σέβομαι μὲν προσιδέσθαι, σέβομαι δ᾽ ἀντία λέξαι,
E. Iph. A. 824 αἰνῶ δ᾽ ὅτι σέβεις (for the meaning see below) τὸ σωφρονεῖν,
Pl. Tim. 69 ἃ σεβόμενοι μιαίνειν τὸ θεῖον, Laws 798 Ὁ σέβεται καὶ φοβεῖται πᾶσα
ἡ ψυχὴ τό τι κινεῖν τῶν τότε καθεστώτων. Thus the translations quoted above
are unsatisfactory ; they probably derive from the fact that the lexicons give
as the primary meaning of the verb veneror, adoro (Stephanus, though he does
later give colo; see below), ‘in Ehrfurcht scheuen . . . verehren' (Passow), ‘to
feel awe or fear; to worship, honour’ (L-S). This has been modified more or
less drastically until something which seemed to fit the context here was
obtained. In Linwood’s lexicon to Aeschylus, on the other hand, we find s.v.
σέβειν the heading ‘to use, to exercise as an office, etc.’, and under this heading,
besides Ag. 1612, Eum. 715 ἀλλ᾽ aiparnpa πράγματ᾽ οὐ λαχὼν σέβεις and Eum,
749 TO μὴ ἀδικεῖν σέβοντες ἐν διαιρέσει. Linwood’s paraphrase may at first
sight seem a rough-and-ready one, but what really matters is the fact that
he kept clear of veneror and the like and gave a substantially correct inter-
pretation of the meaning that is peculiar to the verb as used in those passages.
A particularly clear case is Eum. 715, aptly translated by Otfried Müller ‘du
übst ein Blutamt, das dir nicht zum Loose fiel’. We have to acknowledge that
σέβειν has not only the narrower sense of venerari, but also the wider one of
colere? in its various shades of meaning. The use of σέβειν provides close
analogies with colere officium, colere amicitiam (cf., e.g., E. Or. 1079 q σοι
κατηγγύησ᾽ ἑταιρίαν σέβων), etc. Sometimes the transition from 'reverence' to
‘practise reverently' is barely perceptible, e.g. E. [ph. T. 1189 τὸν νόμον
ἀνάγκη τὸν προκείμενον σέβειν, Phoen. 293 f. γονυπετεῖς ἕδρας προσπίτνω σ᾽ ἄναξ

1 This interpretation of σέβω Ag. 1612, which agrees with that of Blomfield, was given by
Hermann in his note on Suppl. 597 (579 Hermann); Wilamowitz, Interpr. 36, expressly
commends it. Since I can make nothing of the text of Suppl. 596 (cf. my remark p. 110),
I am unable to say what σέβει there means.
2 σέβω is used as a gloss to colo ; Corp. Gloss. Lat. ed. Loewe-Goetz, ii. 430. 24.
761
line 1612 COMMENTARY

(i.e. proskynesis in the oriental style), τὸν οἴκοθεν νόμον σέβουσα ; perhaps also
Hel, 1270 τόδ᾽ “Ἑλλὰς νόμιμον Er τίνος σέβει; (though here, I think, ‘practise’
predominates). In other cases the mental attitude is less prominent and the
emphasis falls on the ‘practising’. The verb denotes on the whole activities
within a sphere which man approaches with awe, that is to say σέβειν is not
extended at random to what might be called purely secular actions and
attitudes of mind, but has for its object a σεμνόν or its direct opposite (so Ag.
1612 and Eum. 715). The special meaning which concerns us here is confined
to the active. The earliest example of this active form is significant: Archi-
lochus fr. 119 D. Δήμητρος ἁγνῆς καὶ Κόρης τὴν πανήγυριν σέβων, not ‘reveren-
cing’ but ‘celebrating (reverently)’, colens (as one says sacra, ferias, caeri-
montas etc. colere). I give some examples from Tragedy in addition to those
already quoted: 5. Ant. 744 ἁμαρτάνω γὰρ τὰς ἐμὰς ἀρχὰς σέβων; there is now
no need for us to distort the sense with the aid of artificial explanations
(Jebb, ‘when I respect mine own prerogatives’ ; Bruhn, ‘The king is permeated
by the consciousness of his own royal dignity to such an extent that he believes
he is "respecting" his authority as ruler when he is trying to enforce his
will’). Instead, we recognize exactly the same use as in Eum. 715 ἀλλ᾽
αἱματηρὰ πράγματ᾽ οὐ λαχὼν σέβεις. Creon is convinced that he is simply
exercising his royal authority in a lawful way. The point of Haemon’s retort
lies in his insisting more strongly on the usual sense of σέβειν: οὐ yàp σέβεις,
τιμάς ye τὰς θεῶν πατῶν (‘in the case of conduct which is clearly ἀσεβές there
can be no question of σέβειν ἢ. E. Suppl. 74 f. (summons to raise the dirge)
ir” ὦ Evvadynddves χορὸν τὸν Ἅιδας σέβει, not ‘which Hades reverences’, but
‘which Hades celebrates’ (like τὴν πανήγυριν σέβων). Iph. A. 824: Achilles has
approached the royal lady whom he has never seen before with shy deference
(ὦ πότνι᾽ αἰδώς) ; at this Clytemnestra says: αἰνῶ δ᾽ ὅτι σέβεις τὸ σωφρονεῖν, 'I
commend you for practising (reverently) σωφρονεῖν '. Closest to the expres-
sion ὑβρίζειν ἐν κακοῖσιν οὐ σέβω is the fine fragment of a chorus from Euri-
pides’ Phoenix (fr. 814 N.) φθόνον οὐ σέβω, φθονεῖσθαι δὲ θέλοιμ᾽ ἂν ἐπ᾽ ἐσθλοῖς,
‘I do not practise envy’, ‘indulge in envy’. Here it is obvious that the sharply
antithetical φθονεῖσθαι δέ demands something preceding it which is the
equivalent of an active infinitive. To return to Ag. 1612, what the coryphaeus
says is something of this sort (the pregnant brevity of the original is in-
evitably lost in the paraphrase): ‘I do not practise (as if it were an act of
piety) insolence towards misfortunes (of others) ; I neither have nor ever will
have anything to do with it." Using a common form of polemic he does not
explicitly attack the conduct of his opponent, but seemingly formulates a
maxim for his own conduct in contrast to that of the other man. This
renders the rebuke still sharper. Here at the beginning of his reply the
coryphaeus repudiates the tone and the content of Aegisthus’ speech on the
ground of a universal religious and moral sentiment. Aegisthus, like Clytem-
nestra before him, has offended against that recognized rule of conduct
which has been discussed on 1394. After the generality of the opening
sentence the speaker passes straight to the charge proper with its strong
legal colouring.
1613. τόνδε dais: Pauw's division of the words must be accepted, not only
1 It is thus clear that Verrall with his ‘I care not’ comes pretty close to the true force of
the expression.
762
 COMMENTARY line 1614

because ἔφης is foreign to older Attic (cf. Kühner-Blass, ii. 211 ἢ. 1), but also
because the thought demands ‘you admit’ or ‘do you admit?'.
Several editors query or alter ἑκών. This is an instance of the obtuseness
with which legal terms are commonly ignored.! To connect dis ἑκών
(Verrall) is perhaps worse. Right at the outset the accusation is put in the
sharpest conceivable form: Aegisthus is reproached with having, by his own
admission, committed wilful murder of the man whose corpse is lying there,
i.e. he is charged with the crime which in legal language, at any rate in
Athens, was originally termed φόνος ἐκ mpovolas: to its perpetrator the word
ἑκών could probably be applied at any period,? since even in legal language
ἄκων ἔκτεινε was used of a man who killed anyone without deliberate intent;
cf., e.g., IG i? 115. 17, 34, ie. ‘in the Draconian code or rather what passed as
such in fifth-century Athens’ (Latte, RE xvi. 281), Hdt. 1. 35. 3 φονεύσας δὲ
ἀδελφεὸν ἐμεωυτοῦ ἀέκων, Antiphon 3 a 1 ἑκόντα μὲν οὖν οὐκ ἐπικαλῶ ἀποκτεῖναι,
ἄκοντα δέ. Every Athenian was naturally conversant with the distinction
between the most heinous sort of homicide, which alone was tried before the
court of the Areopagus, and other sorts of homicide.
1614. This sentence forms the kernel of the accusation. The poet has left
his audience in no uncertainty as to the decisive importance of this point,
since in his two next speeches (1627, 1634) the coryphaeus repeats the charge
that Aegisthus (again by his own admission) planned, ἐβούλευσε, the murder.
The three passages show a deliberate monotony in the choice of words and
the sound of the second part of the verse after the caesura (on the repetition
of the deictic pronoun in this passage cf. on 1635) ; it would be impossible to
hammer in more forcibly than this the assertion made in them. This scene
provides our oldest evidence for the characteristic concept of βούλευσις in
connexion with criminal law. What is involved is ‘the legal principle which
Andocides 1 (7. puor.). 94 terms νόμος : οὗτος ὁ νόμος καὶ πρότερον ἦν (kat) ὡς
καλῶς ἔχων καὶ νῦν ἔστι, καὶ χρῆσθε αὐτῶι, τὸν βουλεύσαντα ἐν τῶι αὐτῶι ἐνέχεσθαι
καὶ τὸν τῆι χειρὶ ἐργασάμενον, 1.6. it was the practice to follow this procedure
in trials falling under that category, and naturally this practice ultimately
found its expression in legal language’, to quote R. Maschke, Die Willenslehre
im griech. Recht (Berlin 1926), 84 ; this book contains a thorough investigation
of βούλευσις (this scene of the Agamemnon is discussed on p. 86). Cf. also
Latte, RE xvi. 285: ‘It is certainly a later extension [of old Attic law] when
the party intellectually responsible who has used another person as his tool
could be brought into court as βουλεύσας '. It would certainly be wrong to
impute to Aeschylus the idea of making his coryphaeus threaten Aegisthus
with a γραφὴ βουλεύσεως, which probably did not yet exist; but it is clear
that in the argumentation of this scene the same value is set on βουλεῦσαι as in
the current practice mentioned by Andocides. The obvious antithesis of
6 βουλεύσας and ὁ τῆι χειρὶ ἐργασάμενος (cf. also Antiphon 6. 16 ἐγὼ 8e—scil.
διωμοσάμην---μὴ ἀποκτεῖναι μήτε χειρὶ ἐργασάμενος μήτε βουλεύσας) is found
1 For the same reason editors continue to despise in Cho. 620 Porson's division of words
νοσφίσασα προβούλως, despite the comments of Wilamowitz and Blass (who refers to Anti-
phon 1. 3), and accept the nonsensical ampoßovAws πνέονθ᾽ ὕπνωι,
2 Cf.,e.g., the passage quoted below from the tetralogies of Antiphon and (for a later
period) the phrase in the law of Priene (2nd cent. B.c., discussed by Latte, Hermes, Ixvi,
1931, 136), Inschriften von Priene 84: ὃς ἂν τὸν ἐλεύθερον ἑκὼν ἀπο[κτείνηι.
3 C£. Bruno Keil, Anonymus Argentinensis, 227.

763
line 1614 COMMENTARY

also in Pindar, P. 4. 72 deodarov ἦν Πελίαν ἐξ ἀγαυῶν Αἰολιδᾶν θανέμεν χείρεσσιν

ἢ βουλαῖς ἀκάμπτοις, A. Prom. 619 βούλευμα μὲν τὸ Aiov, “Hoaiorov δὲ χείρ, and,
e.g., Pl. Laws 9. 872 a ἐὰν δὲ αὐτόχειρ μὲν μή, βουλεύσηι δὲ θάνατόν τις ἄλλος
ἑτέρωι κτλ., Demosth. 19. 21 οὐδὲν ἧττον ἠσεβήκασιν οἱ βεβουλευκότες τῶν ταῖς
χερσὶ πραξάντων. Since the hand is regarded as the proper organ of δρᾶσαι
(cf. on 1357), the passages just quoted contain exactly the same contrast as
does Ag. 1634 f. with its opposition of ἐβούλευσας to δρᾶσαι; cf. also S. Ant.
266 f. τὸ μήτε δρᾶσαι μήτε τωι ξυνειδέναι τὸ πρᾶγμα βουλεύσαντι μήτ᾽ εἰργασμένωι.
μόνος. The coryphaeus is justified in imputing this admission too to
Aegisthus, since in 1609 he boasted that he was himself πᾶσαν (the point
stressed here) ξυνάψας μηχανὴν δυσβουλίας. The emphatic μόνος at the head of
the sentence imparts additional force to the accusation.
There is in the long and carefully arranged planning, the βουλεῦσαι, of
Aegisthus one detail which is not mentioned at all in this scene, but revealed
at the close of the following tragedy (Cho. ıo11): Aegisthus lent his sword
to Clytemnestra for the murder (cf. Appendix B).
After 1614 the scribes of G and Tr and, more recently, Bothe (only in his
first edition) put a question-mark, which was accepted by Blomfield, Her-
mann, and many others. It may, perhaps, not be indispensable, but it is
certainly an improvement. Above all, it renders the asyndeton of 1615 more
forceful: ‘you avow, do you ...? Well then, the consequence will be... .'.
In 1619 we find an asyndeton of precisely the same character. The general
parallelism of 1613 ff. and 1617 ff. (first the questions od... dys... and σὺ
φωνεῖς... respectively, each consisting of two lines, then the asyndetic
continuations, which contain threats for the future) is greatly in favour of
taking 1613 f. as a question. To begin the retort 1613 ff. with a question (how-
ever ‘rhetorical’), rather than a mere statement, would also be in agreement
with the whole conduct of the coryphaeus, who in this scene performs some
kind of ἀνάκρισις. For où δέ introducing an interrogative sentence cf., e.g.,
Prom. 67, S. Ant. 510, Oed. R. 1025, El. 402.
1615. ἐν δίκηι is variously interpreted. Many take it as ‘justly’, ‘in strict
justice’ (so Conington ; similarly,e.g., Kennedy, ‘righteously’ ; L. Campbell, ‘I
warn thee, thou shalt have thy due. Thou shalt not save thy head’, etc.).
This is in keeping with the use of ἐν δίκηι = δικαίως in Pindar, OL. 6. 12, S. Tr.
1069, etc. On the other hand, Stanley has ‘in iudicio’, so, e.g., Humboldt, ‘im
Volksgericht’ ; Schneidewin, Paley, ‘in the hour of justice’. That tallies with
the use in Cho. 987 ὡς ἂν παρῆι μοι μάρτυς ἐν δίκηι ποτέ (Schol. ἐν τῆι κρίσει τῶν
᾿Ερινύων, where ἐν τῆι κρίσει is correct while τ. 'Ep. intrudes something which
does not belong here); cf. also Solon fr. 24. 3 D. ev δίκην χρόνον, ‘vor dem
Richterstuhle der Zeit’ (Wilamowitz, Lesebuch, ü. 24). The latter view is
doubtless correct. The reference to a future trial is of a piece with the whole
attitude of the coryphaeus here, which strongly recalls forms of legal pro-
cedure. ἐν δίκηι points to the final stage, when the culprit, after being con-
demned by the people’s sentence, will be stoned amid the curses of the
multitude.
1616. δημορριφεῖς . . . λευσίμους ἀράς. It is characteristic of stoning, in con-
1 For the opposition in general cf. also the argumentation in Arist. Rhet. 1. 7 p. 1364218 ff.
τὸ γὰρ τέλος μεῖζον καὶ οὐκ ἀρχή, ὥσπερ ὁ Λεωδάμας κατηγορῶν ἔφη Καλλιστράτου τὸν βουλεύ-
σαντα τοῦ πράξαντος μᾶλλον ἀδικεῖν... πάλιν δὲ καὶ Χαβρίου, τὸν πράξαντα τοῦ βουλεύσαντος.

764
 COMMENTARY lines 1617f.
trast to other forms of the death penalty, that 'the participation of everyone,
of the whole community, in the execution is the natural and original practice
... When they speak of it, the Greeks tend to give prominence to the people
as the agent of execution' (R. Hirzel, ‘Die Strafe der Steinigung', Abhandl. d.
Sächs. Gesellsch. d. Wiss., phil.-hist. Kl. xxvii, 1909, 238; as instances he
quotes, besides Ag. 1616, A. Sept. 199 λευστῆρα δήμου... μόρον, S. Ant. 36
φόνον... δημόλευστον, E. Or. 442 θανεῖν ὑπ᾽ ἀστῶν λευσίμωι πετρώματι). ‘The
deed which has a stoning for its consequence must always be one which
provokes the anger of all’, ‘the murder, too, which is to lead to a stoning,
must be of such a character that it strikes at the existence or feelings of the
community, the murder of a king like Agamemnon, or of Philopoemen [Plut.
Philop. 21. 9], or, again, of the Boeotian Phocus, whose death, according to
Plutarch’s narrative [Mor. 775 a], was rendered by the circumstances a
matter of common concern to all Boeotia’ (Hirzel, op. cit. 247 f.). The Greeks,
like other peoples, combined cursing with the stoning (cf. Hirzel, 236 f.). The
parallelism between the threats of the Chorus against Clytemnestra 1410 f.
δημοθρόους ἀρὰς... ἀπόπολις δ᾽ ἔσηι (cf. her rejoinder 1412 f. ἐκ πόλεως
φυγὴν... δημόθρους τ᾽ ἀράς) and the threats against Aegisthus here makes
evident the close relation between expulsion into exile by the community
and stoning. ‘L’expulsion figure la lapidation, dont elle est horrible adou-
cissement’, writes Glotz, La Solidarité de la famille, 24; cf. on this affinity
Hirzel, op. cit. 242, with references to the modern discussion of the subject.
δημορριφεῖς (here alone), a happy coinage, since λόγους ῥίπτειν, etc. (e.g.
Prom. 312, cf. 932 τοιάδ᾽ ἐκρίπτων ἔπη) were current phrases, as was ῥίπτειν
ἀράς (E. Tro. 734), while at the same time there is clearly perceptible in the
word the picture of the stoning; volleys of mingled curses and stones are
hurled at the head of the condemned criminal. Cf. Propertius 4. 5. 78 (the
stoning is directed against a tomb) mixtaque cum saxis addite verba mala.
1617 f. ZxóA. παλ.: ἀντὶ τοῦ ὑποδεεστέραι καθέδραι ὦν: of yap ζύγιοιϊ τῶν
θαλαμιῶν ἄνωθέν εἰσιν. This scholion has done much harm. The misunder-
standing of the image was probably promoted by the ambiguity of the plural
κρατούντων. From Stanley on most commentaries have been dominated by
the thought of the different lines of rowers, together with the discussion
(which this assumption makes all but unavoidable) of whether the ship in
question had two or three lines of rowers.” What manifest nonsense results
from this is shown, e.g., by Schneidewin’s note, ‘The phrase is derived from
the rowers of biremes where . . . the Zuyiraı occupied the higher position’,
with the translation (in itself correct) which he adds ‘since these are masters
of the vessel and hold the command’. Does that apply to the &uyira—or to
any of the rowers? Are they not bound like all the rest to obey orders?
General considerations, as well as those adduced above on 182 f., make it
plain that the command must lie with the helmsman. Typical of the use of
κρατεῖν in conjunction with the image of the helmsman is Prom. 149 νέοι γὰρ
οἰακονόμοι κρατοῦσ᾽ ᾽Ολύμπου.
Several scholars have ventured to follow common sense, instead of the
1 This word should not be altered with Victorius, since Pollux, 1. 87, 120, attests the use of
ζύγιοι in the sense of ζυγῖται.
2 Tarn’s treatment of the passage (Journ. Hell. Stud. xxv, 1905, 205 n. 80), which is not
very happy in other respects, proceeds from this same wrong starting-point.

765
lines 1617 f. COMMENTARY

established interpretation ; so, e.g., Humboldt, ‘da das Schiff regieren, die am
Steuer sind’, J. H. Voss, Franz, J. G. Droysen, Paley,’ ‘this was the position
of the steersman . . .' (he rightly quotes E. Phoen. 74, Ion 595, and refers to
Ag. 183), C. Torr, Ancient Ships, 57 n. 131, A. C. Pearson (on E. Phoen. 74
and on Headlam's verse translation of Ag., p. 260), Mazon, L-S s.v. ζυγόν,
G. Thomson, A. Y. Campbell. It has already been remarked on 182 f., with
reference to the Homeric ὑψίζυγος, that the common word ζυγόν (— σέλμα,
transtrum) was also used for the little deck at the stern of the ship (in ancient
times there was no deck which ran the length of the ship), on which the
helmsman had his seat (cf. Kóster in Kromayer-Veith, Heerwesen der
Griechen und Römer [1928], 177 ; Miltner, RE, Suppl. v. 933. 25 ff.). In addition
to Ag. 1618 see E. Ion 595 és τὸ πρῶτον πόλεος ὁρμηθεὶς Cuydv,* Phoen. 74 f.
ἐπεὶ δ᾽ ἐπὶ ζυγοῖς καθέζετ᾽ ἀρχῆς. The contrast with the rowers, νερτέραι
προσήμενοι κώπηι, is clearly brought put. Given the construction of the hull,
it is obvious that the helmsman on the stern sits at a considerably higher
level than the rowers sitting more amidships. Any thought of ‘biremes or
‘triremes’ should be banished from the mind. Even the simplest ship with
only one line of rowers shows the difference of level between the seat of the
helmsman and the benches of the rowers: this appears very plainly, e.g., on
the famous red-figure stamnos from Vulci with the picture of Odysseus’
adventure with the Sirens (Brit. Mus., Beazley A.R.V. 177; Furtwängler-
Reichhold iii, pl. 124, reproduced also in A. Késter, Das antike Seewesen 97,
and often elsewhere). Notice also the epithet ἄκραι in E. Cyclops 14 f. ἐν
πρύμνηι δ᾽ ἄκραι αὐτὸς λαβὼν ηὔθυνον ἀμφῆρες δόρυ.
κρατούντων... δορός belong together; Stanley translated rightly: ‘im-
perium navis obtinent’.
The plural number of κρατούντων and τῶν ἐπὶ ζυγῶι corresponds exactly
with the actual situation. In no circumstances can Aegisthus allude to him-
self as ruler, excluding Clytemnestra. But it is possible that it is at the same
time a generalizing plural of the sort sometimes termed ‘allusive plural’ (cf.,
e.g., Jebb on S. Oed. R. 366, A. C. Pearson on E. Phoen. 40 τυράννοις). In
Prom. 149 the clause νέοι yap οἰακονόμοι κρατοῦσ᾽ ᾽Ολύμπου differs only in form
and not in content from the phrase, which follows immediately, νεοχμοῖς δὲ
δὴ νόμοις Ζεὺς ἀθέτως κρατύνει. On the generalizing plural Ag. 1625 τοὺς
ἥκοντας κτλ. v. ad loc.
1617 £. is rightly punctuated as a question in the MSS. The beginning of a
question is framed in exactly the same way Eum. 896 σὺ τοῦτο πράξεις ὥστε
με σθένειν τόσον ;.3 Cf. 1625 ff. and, e.g., Ar. Thesm. 918 f. σὺ τὴν ἐμὴν γυναῖκα
κωλύεις ἐμὲ... ἄγειν; so also Alcaeus fr. 45. xo ff. D. σὺ δὴ τεαύτας ἐκγεγόνων
ἔχηις τὰν δόξαν κτλ., where Lobel has recognized the question.
1619. For the thought cf. on 710. Even in details the sentence is an echo of
1 At any rate in the fourth edition (1879) of his commentary, where it is only at the end of
his note that he cites the traditional interpretation, to which he himself adhered in the
first edition.
2 Curiously misunderstood by Wilamowitz : ‘As the son of a prince, he, in contrast to the
common people, the rowers, stands [N.B.] upon the deck. Aesch. Ag. τότ]. A. S. Owen
comes to no decision on the passage. Paley gives the correct explanation : ‘A metaphor from
the high seat on the stern where the pilot sat and worked the oar-paddle, πηδάλιον [the
plural would be better].’
3 Rightly punctuated by the editors as a question ; M has no punctuation-mark.
766
 COMMENTARY line 1623

Clytemnestra’s threat to the Chorus in 1425 γνώσηι διδαχθεὶς ὀψὲ γοῦν τὸ

σωφρονεῖν. The moralizing advice is unpleasant enough, but it leads up to
something that is worse, the undisguised brutality of the subsequent threats.
1620. σωφρονεῖν εἰρημένον. On the construction cf., besides the grammars
(Kühner-Gerth, ii. 88 f.), K. W. Krüger on Thuc. 1. 125. 2 and Cobet, Misc.
crit. 143 f. Among the examples of this type of accusative absolute Thuc. 7.
77. 6 προπέπεμπται δ᾽ ws αὐτούς, Kal ἀπαντᾶν εἰρημένον καὶ σιτία ἅμα κομίζειν
shows exactly the meaning of εἰρημένον which we have here (cf. above on 301),
1621 f. It is difficult to decide whether the δεσμὸς of Tr or Karsten’s δεσμοὶ
should be preferred as the emendation of δεσμὸν.
The word-order is very lively. One gets the impression that at the start
Aegisthus intends to say only δεσμὸς de καὶ τὸ γῆρας διδάσκειν KrA., 'vincula
etiam ad senectutem docendam aptissima sunt’, but then, by way of after-
thought, he seizes upon, and relishes, the detailed picture (cf. 1641 f.) of the
tortures which the old men will have to endure in prison; thus the supple-
ment ai re νήστιδες δύαι is added. Finally νήστιδες δύαι gives rise to an element
of personification ἰατρομάντεις ; the regulation of a rigorous dieting, a 'hunger-
cure’, is a matter for the physician.
1623. ἰατρόμαντις: ‘a seer who is at the same time a physician’; cf. on this
type of copulative compound Debrunner, Griech. Wortbildungslehre, 46;
Buck, Comparative Grammar of Greek and Latin, 354; Schwyzer, Griech.
Gramm. i. 453 1. Presumably the word was coined (we do not know whether
by Aeschylus or by another poet) to denote Apollo in his double capacity as
divine physician and seer (so Eum. 62; in Suppl. 263 it is used of Apollo’s
son). In its original sphere the noun expresses dignified praise; here it has
become an instrument of cruel sarcasm. The sufferings are called ἰατροί since
they will cure an obstinate presumption by means of a starvation diet ; they
are μάντεις because they will reveal to the old men what so far has been
hidden from them. Perhaps this application of the sacral word, coming from
Aegisthus, would have the effect of blasphemy.
οὐχ ὁρᾶις ὁρῶν τάδε: the commentators are not agreed about the meaning.
Conington: ‘may be taken either as ὁρῶν, ody ὁρᾶις τάδε; as I have taken it
[he translates, as Peile had done, ‘Canst see and seest not this?’], or as ody
ὅρᾶις, ὁρῶν τάδε ; "Dost thou not see, with this before thy eyes?" i.e. “Canst
thou not see thine own interest, now that thou seest from this dead body how
matters stand?” ' The latter interpretation is found, e.g., in Schneidewin,
Wecklein, Lewis Campbell (‘Are thine eyes not opened as thou lookest on
this?’), Verrall, Pliiss, the former, e.g., in Sidgwick, Wilamowitz (‘Seht ihr das
nicht mit sehenden Augen ?’), Headlam, Platt. The way of taking it as ὁρῶν,
οὐχ ὁρᾶις τάδε ; (‘do you not see that, although you have eyes to see’) seems
definitely preferable. With the parenthetic concessive ὁρῶν without an
object the antithesis gains its full force, and the reproach of unnecessary and
unnatural spiritual blindness receives the sharpness which we expect from
the lips of Aegisthus. Furthermore, the expression belongs to a widespread
popular ‘thought-pattern’, as has long been recognized. . Peile compared
Prom. 447 f., ev. Marc. 8. 18 (= ev. Matth. 13. 13) ; Schneidewin (on 5. Oed. R.
413) adduced Soph. fr. 837 N. (= 923 P.) ἀλλ᾽ οὐδ᾽ ὁρῶντες εἰσορῶσι τἀμφανῆ.
G. Thomson aptly refers to [Demosth.] 25. 89 ὥστε, τὸ τῆς παροιμίας, ὁρῶντας
μὴ ὁρᾶν καὶ ἀκούοντας μὴ ἀκούειν. Cf. also O 128 ff. ἦ νύ τοι αὕτως οὔατ᾽ ἀκουέμεν
767
line 1623 COMMENTARY

ἔστι, νόος δ᾽ ἀπόλωλε καὶ αἰδώς. οὐκ ἀΐεις κτλ. Very likely the tragedians are
alluding to ἃ proverb.
τάδε: probably not ‘the corpses here’ nor ‘the situation in the house’, but
‘what I am now trying to make clear to you’.
1624. That the reading of the MSS, πήσας, goes back to παίσας is clear from
Schol. Pind. Pyth. 2. 173. It is surprising that Butler's πταίσας was adopted
by Wilamowitz.! But Hermann rightly observes 'Mirum, quum παίσας,
offendens [cf. L-S s.v. II 'dash against or upon'] aptissime dictum sit, prae-
latum esse quod languet'. Meineke's observation (Philol. xix, 1863, 212), ‘the ox
kicks, not stumbles, against the pricks', should be sufficient to deal the death-
blow to the conjecture.
There is no doubt that the sentence πρὸς κέντρα μὴ λάκτιζε κτλ, Was pro-
verbial as early as the time of Aeschylus. Aegisthus, throughout inclined to
vulgar turns of speech, shows his customary brutality in using the churlish
phrase to the old man. Blomfield and others have pointed out the other
instances of the proverb, in Aeschylus himself (Prom. 323), in Pindar, in
Euripides (first in the Peliades),* and elsewhere.
1625 ff. From the text of the MSS (1626 aioxvvovoa) we should have to infer
that these three lines are addressed to Clytemnestra; they are, however,
addressed to Aegisthus. Wellauer was the first to catch a glimpse of the
truth, but he started from a wrong presupposition (‘Clytaemnestra non est in
scena’) and did not pause to restore the genuine text and secure the inter-
pretation of the sentence. Consequently Hermann found no difficulty in
rebutting Wellauer’s hypothesis. But it is astonishing that long afterwards,
when all the decisive points had been made, Wilamowitz still clung so ob-
stinately to the wrong view (Interpr. 176). The crucial argument against the
words being addressed to Clytemnestra is not to be derived from the fact,
strange though it is, that the queen does not react in any way to this weighty
charge,? but in the specific character of the accusation levelled at her (ἐβού-
t Also by G. Thomson and by A. C. Pearson in Headlam’s posthumous edition. I do not
know what Headlam finally determined to read, but his prose translation (‘for fear thou
strike on it and suffer’) presupposes παίσας.
2 W. Buchwald, Studien z. Chronol. d. ait. Tragödie (diss. Königsberg 1939), 11, is inclined
to see in this ‘an instance of the influence of the old Aeschylus on the young Euripides’,
although he admits that the phrase goes back to an old proverb. On the relation of Acts
26. 14 to E. Bacch. 795 cf. O. Weinreich, ‘Gebet und Wunder’, Tübinger Stud. 2. Alter-
tumsw., Heft 5, p. 335 (169) f.
3 On this point Murray attempts to set his reader’s mind at ease: ‘Clytaemestram allo-
quitur, illa silet: cf. S. Phil. 1066 sqq.’ But the momentary silence of Neoptolemus is pre-
pared for in the most emphatic way and, as a motif full of violent tension, arises from the
situation itself (ᾧ σπέρμ᾽ ᾿Αχιλλέως, οὐδὲ σοῦ φωνῆς ἔτι γενήσομαι προσφθεγκτός, ἀλλ᾽ οὕτως
ἄπει;); then, before the young man has time to answer, Odysseus presses him to go;
immediately after Neoptolemus does as a matter of fact speak. Not for a moment is the
audience left in any doubt as to his reaction. Does that provide anything like a parallel to
the scene in the Agamemnon? Eum. 303 shows well how in the case of a character not
reacting to an address the poet draws attention to it. Wilamowitz (Interpr. 176) finds ‘the
long silence of Clytemnestra significant’. As regards the principle which we have to apply
in dealing with phenomena such as this in Greek dramatic poetry, I may be permitted to
quote what, elaborating ideas of F. Leo, I have formulated in a different context (Plauti-
nisches im Plautus 253): ‘There is nothing to be gained by arbitrarily-devised psychological
explanations, foisted into the text; what we are entitled to demand is an explanation of
certain glaring absurdities, an explanation in harmony with the technique of Greek drama
so well known to us.’
768
 COMMENTARY lines 1625 ff.

Aevoas) and, above all, in the clear layout of this whole section. On this point
the right view was finally stated by R. Enger, Emendationes Aeschyleae (16.
Jahresbericht des Gymnasiums zu Ostrowo, 1861), 7: ‘Nam quid tandem
secutus ita rem instituerit poeta, ut Chorus cum Aegistho colloquens hoc
colloquio interrupto repente ad Clytaemnestram se converteret nihil respon-
dentem, quum praesertim quid Chorus de Clytaemnestrae facinore cogitaret,
huic in eo quod modo cum illa habuit colloquio significasset? Neque ea ipsa
quae dicit Chorus, recte Clytaemnestrae dicerentur, nam Aegisthus, non
Clytaemnestra, ἐβούλευσε μόρον, Clytaemnestra caedem perfecit, machinatus
est Aegisthus, ut ipse gloriatur vss. 1608 54. καὶ τοῦδε τἀνδρὸς ἡψάμην θυραῖος
ὦν, πᾶσαν συνάψας μηχανὴν δυσβουλίας, Chorus saepius repetit, ut v. 1614 μόνος
δ᾽ ἔποικτον τόνδε βουλεῦσαι φόνον, v. 1634 ὃς οὐκ, ἐπειδὴ τῶιδ᾽ ἐβούλευσας μόρον
eodemque modo hic dvdpi στρατηγῶι τόνδ᾽ ἐβούλευσας μόρον. Wilamowitz,
Interpr. 176, paraphrases 1625 ff.: ‘hast du zur Schande für das Heer [this is
apparently supposed to render τοὺς ἥκοντας ἐκ μάχης νέον] und! zugleich deine
Ehe schändend dem Feldherrn den Tod BovAescaca gegeben, d.h. nimmst du
dem Buhlen die βούλευσις ab?’ Juristically, this is utterly inconceivable (on
the sharp distinction between the βουλεύσας and the χειρὶ ἐργασάμενος cf. on
1614) ; and from the point of view of dramatic structure it would be intolerable
if the coryphaeus broke away thus from the consistent line of arraignment of
Aegisthus which extends from 1613 to 1648. The conclusion that ll. 1625-7
are directed at Aegisthus is confirmed by the exact correspondence (pointed
out by Schneidewin) with Cassandra's words 1223 ff.: ἐκ τῶνδε ποινὰς φημὶ
βουλεύειν τινὰ... oikovpóv, οἴμοι, τῶι μολόντι δεσπότηι. Pohlenz, D. griech.
Tragödie, Erläuterungen 34 f., gives a good summary of the reasons which
make it necessary to understand 1625 ff. as addressed to Aegisthus.
From what has been said it is clear that 1626 αἰσχύνουσ᾽ is corrupt. The
corruption is very slight, since it was almost inevitable that γύναι would be
misunderstood. The true reading was restored by Keck with αἰσχύνων
(adopted by Wecklein and others) ;? that fits the context better than Wieseler's
αἰσχύνας, and it would be senseless to allow any importance to resemblance
of letters here where the corruption did not arise from an error of the eye, but
from the desire to obtain a feminine form. Apart from this simple emenda-
tion, there is, so far as I can see, not a single alteration necessary in these
three lines.” We must now examine the details.
γύναι. Hermann alleged that if the address were to Aegisthus it would
have to be γυνὴ σύ, and was followed by Wecklein (annotated edition) and
others. Meineke conjectured γύννις (the word is attested as Aeschylean by
fr. 61 and now also by the satyr play Pap. Oxy. 2162, fr. 1 col. 2. 32, where it
actually has the epithet ἄναλκις, used of Aegisthus in y 310, Ag. 1224, 5. El.
301). I admit the possibility of a slight corruption here, but can see no valid
ground for doubting the correctness of γύναι, σὺ κτλ. In this connexion
‘woman’ seems opprobrious enough. The commentators who assume that
Aegisthus is here addressed have cited corresponding passages, beginning
1 Wilamowitz inserts 7’ after εὐνήν.
2 Klausen had already said ‘si hoc designaturus erat poeta [i.e. that Aegisthus was being
addressed, which Klausen considered impossible], dicendum erat αἰσχύνων᾿. Cho. 990
Aegisthus is termed αἰσχυντήρ.
3 My text thus agrees with that in Wecklein’s edition of the whole of Aeschylus (1885),
not with that in his annotated edition of the Oresteza.
4872.3 U 769
lines 1625 ff. COMMENTARY

with the famous (B 235) Ἀχαιΐδες, οὐκέτ᾽ Ἀχαιοί; cf. besides, e.g., S. El. 302,
where Aegisthus is termed ὁ σὺν γυναιξὶ τὰς μάχας ποιούμενος. In the Oresteia
the treatment of Aegisthus as ‘a woman’ is clearest? in Cho. 304 f. δυοῖν
γυναικοῖν ὧδ᾽ ὑπηκόους πέλειν. θήλεια γὰρ φρήν᾽ εἰ δὲ μή, τάχ᾽ εἴσεται. For other
instances of words like ‘woman’, ‘female’, etc., being applied to a man in
order to characterize him as unmanly, effeminate, see the passages collected
by Wecklein (annotated edition) and Blaydes; especially Soph. fr. 136 N.
(140 P.) θῆλυς μὲν αὐτός, ἄρσενας δ᾽ ἐχθροὺς ἔχων (from the play cited here as
Μυκηναῖαι, in Hesychius as Arpeds ἢ Mukqv(at»av; on the words just quoted
Pearson remarks ‘The parallel to Aegisthus is so close, that one may suspect
that his father Thyestes is referred to.’). Without aiming at completeness one
might add N 623, where Menelaus reviles the Trojans as κακαὶ κύνες, with the
BT scholion on it: ἄκρως δὲ καὶ τῶι θηλυκῶι ἐχρήσατο eis τὴν ἀνανδρίαν τῶν
βαρβάρων, Hdt. 2. 102. 5 ὅτεων δὲ ἀμαχητὶ καὶ εὐπετέως παρέλαβε τὰς πόλιας,
τούτοισι ἐνέγραφε ἐν τῆισι στήληισι.... καὶ δὴ καὶ αἰδοῖα γυναικὸς προσενέγραφε,
δῆλα βουλόμενος ποιέειν ὡς εἴησαν ἀνάλκιδες, Cic. Nat. deor. 1. 93 Zeno...
Chrysippum numquam nisi Chrysibpam vocabat, Sallust Or. Lepidi 21 Fufidius,
ancilla turpis, honorum omnium dehonestamenium. Perseus, King Philip's son,
appears in the same light as Aegisthus in Livy 40. 5. 3 cum se ne ad id quidem,
quod muliebri cogitabat animo, satis ber se validum crederet εἴς. On the trans-
ference of words ‘die Weibern als Ausüberinnen ... despektierlicher .. .
Handlungen zukommen' to males by way of an expression of scorn, cf.
Ernst Fraenkel, Nomina agentis, ii. 120 f. ; on femina as a term of abuse of
effeminate men (e.g. Lucilius 732; [Virgil] Catal. 13. 17) and on the abusive
application to men of nouns denoting female animals or women's occupations
see J. B. Hofmann, Latein. Umgangssprache, and ed. (1936), 194.
oikoupós accords excellently both in sense and tone with γύναι as addressed
to the man. That is clear from 1225 (the whole of 1223 ff. should be taken
together with 1625 ff.). In addition there is the passage compared by Enger
and others E. Heraclid. 700 f. αἰσχρὸν yàp οἰκούρημα γίγνεται τόδε, τοὺς μὲν
μάχεσθαι, τοὺς δὲ δειλίαι μένειν. oikoupetv, ‘keeping the house’, is primarily ‘the
special task of women' (Wilamowitz on E. Her. 45); cf., e.g., Pind. P. 9. 19
(on the meaning and the difficulty of recovering the authentic form of the
word here cf. the commentary of O. Schroeder), S. Trach. 542 οἰκούρια,
'reward (given by the husband to his wife) for keeping the house', Oed. C.
343 (the sons κατ᾽ οἶκον oikovpoücw ὥστε παρθένοι), Soph. fr. 447 N. (= 487 P.;
see Pearson's note on oikovpós μόνη), E. Hipp. 787, Hec. 1277, Or. 928. In the
passages quoted at the beginning of this paragraph the word bears a secondary
sense, being used in disparagement of the man who, whether from cowardice or
feebleness or old age, stays at home with the women, the ‘stay-at-home’ (L-S).
In E. Her. 44 f. Amphitryon says λείπει (Herakles) γάρ με τοῖσδ᾽ ἐν δώμασι
τροφὸν τέκνων oikovpóv (‘Amphitryon is no longer ἐν àvópdow' Wilamowitz).?
1 Kaibel ad loc. rightly takes Ag. 1625 as an address to Aegisthus.
2 I do not hesitate in this connexion to refer to Cho. 630 γυναικείαν ἄτολμον αἰχμάν, since I
have no doubt that Verrall and Wilamowitz (in his commentary [1896] p. 215 and /nterpr.
254) were right in making this expression apply to Aegisthus. Cf. also Sidgwick : ‘ ἄτολμον
is the last word that Aesch. would use of Klyt.’
3 Plutarch, Cic. 41. 6, shows an accurate perception of the tone of the old word : ‘Avramos
. ἐν ταῖς πρὸς τοὺς Φιλιππικοὺς ἀντιγραφαῖς ἐκβαλεῖν φησιν αὐτὸν γυναῖκα παρ᾽ ἣν ἐγήρασε,
χαριέντως ἅμα τὴν οἰκουρίαν ὡς ἀπράκτου καὶ ἀστρατεύτου παρασκώπτων τοῦ Κικέρωνος.

770
 COMMENTARY lines 1625 ff.

In order to obtain a smooth construction (see below) Auratus conjectured

τοῦ γ᾽ ἥκοντος, which was slightly modified to τοῦδ᾽ ἥκοντος by Casaubon,
whom Stanley, Schütz, Porson, and Blomfield followed. τοὺς ἥκοντας, which
in reality means Agamemnon alone, but only in so far as he is in the same
situation as others, is rightly explained by Klausen as ‘generalizing plural’
(see, besides the grammars, Wackernagel, Syntax, i. 94 f., cf. also on 1618):
‘pluralis pro singulari ponitur, ubicumque designandum est esse universi
aliquid in condicione eius de quo sermo fit’. τοὺς ἥκοντας corresponds to the
phrase τῶι μολόντι δεσπότηι (1225) in the sentence of Cassandra which has so
much in common with this passage and is obviously meant to serve as a
preparation for it.
νέον is excellent and full of significance: ‘those who are only just back from
the battle’. That makes us feel the villainy of the murderous plot twice as
keenly. The adverbial use of veov, after the example of Homer, is found in
three other passages of Aeschylus, all in the Prometheus, but in Sophocles
only in Oed. C. ı775. For this particular phrase, ἥκοντας νέον, cf. y 318, v 191,
360 νέον... εἰλήλουθεν or εἰληλουθώς, p 112 ἐλθόντα... νέον; E. El. 1070 νέον
τ᾽ dm’ οἴκων ἀνδρὸς ἐξωρμημένου. Clearly, no one would have tampered with
this unimpeachable word but for a desire to substitute something which
would restore what appeared a defective construction, i.e. lend support to
the accusative τοὺς ἥκοντας. The attempt of Wellauer and Conington to make
τοὺς ἥκοντας depend on οἰκουρός was conclusively proved by Hermann to be
impossible. But to some scholars it seemed perfectly easy to achieve by
means of an alteration of the text the meaning which Wellauer elicited from
the words of the text, i.e. ‘domi exspectans eos qui ex pugna rediissent’. This
was why Wieseler for νέον read μένων, in support of which Enger adduced the
deceptive parallel E. Heraclid. 7o1 (there μένειν is absolute and has a different
sense), and this ill-considered conjecture still appears in the texts, e.g., of
Wecklein (annotated edition), Mazon, A. Y. Campbell, G. Thomson.’ This is
the more vicious since it absolutely misses the sense of ἥκοντας ('those who
have come from the battle and are now here, the returned soldiers’); the
fault is evident in the translation given by Wecklein : ‘der du... die aus der
Schlacht Kommenden abwartest'. Platt felt this, and therefore conjectured
(J. Phil. xxxii, 1913, 70) 9éovras. But a textual criticism which alters a
faultless expression in order to adapt it to a conjecture (μένων) confutes itself.
If, then, we keep νέον, as we are bound to do, what becomes of the construc-
tion or, in other words, on what does τοὺς ἥκοντας depend? Wilamowitz tried
to obviate the difficulty by inserting τ᾿ after εὐνήν in 1626, thus making τοὺς
ἥκοντας as well as εὐνήν the object of αἰσχύνουσα ('dem Heer zum Hohn, das
frisch vom Feld kam, und zum Hohr? der Gattenpflicht'). The result is a far-
fetched idea (‘an insult to the army’) and an involved expression ; moreover
ἅμα is robbed of its force. dua added to a participle connects closely the
action denoted by the participle and that expressed by the main verb (in
this case éfloVAeveas), as was noticed by Blomfield, who quoted parallels; the
use is quite common (cf. Passow-Crönert, 326. 11 ff.). In the Oresieia, Cho.
897 f. is similar: τόνδε... μαστόν, πρὸς ὧι σὺ πολλὰ δὴ βρίζων ἅμα οὔλοισιν
t According to his own principle Thomson ought to have said that Headlam kept νέον;
his prose translation runs ‘soldiers newly come from the war’, his verse ‘soldiers thus come
newly from the field’.
771
lines 1625 ff. COMMENTARY

ἐξήμελξας εὐτραφὲς γάλα. In the present passage the speaker stigmatizes the
fact that the adultery and the plotting of the assassination took place ἅμα,
as in the case of Hamlet’s uncle: the two affairs went, so to speak, hand-in-
hand. So far as concerns the construction of τοὺς ἥκοντας, there is an anaco-
luthon,' not nearly as harsh as others in Aeschylus. First comes τοὺς ἥκοντας.
ἐκ μάχης νέον, in the case natural for an object ;2 there is in prospect some verb
or other (‘you have treated thus’, ‘have made over to destruction’, or what-
ever one chooses to imagine), but the verb is delayed because the thought
τοὺς ἐκ μάχης ἥκοντας immediately evokes the contrasting thought of the
οἰκουρός and his adultery. Once the interruption has taken place the former
construction shifts a little, as so often in Aeschylus. At the same time there
is operative in the passage the desire to intensify the reproach of conduct
hostile to the men returned from the war (opened in general terms in 1625)
by adding to it the particular sting of βουλεῦσαι φόνον, as already in 1614.
Consequently, ἀνδρὶ orparnyaı takes the place of the preceding τοὺς
ἥκοντας.
1627. ἀνδρὶ στρατηγῶι: not ‘a chieftain and a soldier’ (Conington) or ‘a man
and captain of an army’ (Headlam), but simply ‘the chief of the army’. Cf,
e.g., Sept. 717 οὐκ ἄνδρ᾽ ὁπλίτην τοῦτο χρὴ στέργειν ἔπος, Ag. 259 φωτὸς ἀρχηγοῦ,
cf. also Leaf on À 194.
τόνδ᾽ ἐβούλευσας μόρον: the pronoun is as apposite here as in 1614: the
old men see the body lying in front of them ; cf. on 1635. In advocating τῶιδ᾽,
under the impression that it is the reading of FG, Pohlenz, Griech. Tragödie,
Erläut. 34, was, like A. Y. Campbell and G. Thomson, the victim of a mistake
in Wilamowitz’s critical apparatus (the note ' τῶδε βουλεύσας FG’ belongs
to 1634) ; exactly the same mistake was made by Blomfield and corrected by
Hermann.?
As has long been observed, the accusation of 1625-7 picks up the motif of
y 262 ff.: ἡμεῖς μὲν γὰρ κεῖθι πολέας τελέοντες ἀέθλους quel’ " 6 δ᾽ εὔκηλος μυχῶι
Ἄργεος ἱπποβότοιο πόλλ᾽ "Ayapeuvovégy ἄλοχον θέλγεσκεν ἔπεσσιν. The Electra
of Euripides stresses the same reproach (916 f.) κάγημας αἰσχρῶς (Ag. 1626
αἰσχύνων) μητέρ᾽ ἄνδρα τ᾽ ἔκτανες στρατηλατοῦνθ᾽ "EAAnow, οὐκ ἐλθὼν Φρύγας,
while the Sophoclean Electra only alludes to it with spiteful brevity (301 f.)
ὁ πάντ᾽ ἄναλκις oÙros . . . ὁ σὺν γυναιξὶ τὰς μάχας ποιούμενος.
1628. This is the tragic equivalent of the colloquial κλαύσει μακρά; cf. on
1148. In his fury the boor Aegisthus becomes more and more vulgar.
ἀρχηγενής here alone. It is in consonance with what has already been
observed (cf. on 556 and 562) that such a word, at once bold and with a ring
1 After I had convinced myself of the necessity of taking the sentence thus, I discovered
that Scholefield and Verrall had already done the same, and that Casaubon, though
hesitatingly, had preceded them; he says: ‘potest tolli insolentia constructionis si hic
legamus τοῦδ᾽ ἥκοντος postquam hic venit e bello. sed Aeschyli audaciae in sermone non
male convenit quod est editum. mira est autem locutio, ponit enim dativum ἀνδρὲ στρατηγῶι
appositive cum accusativo τοὺς ἥκοντας ᾿, Murray, too, seems to incline to this view, as may
be inferred from the dash which he puts after 1625; but it is not easy to see why he adopts
Wilamowitz’s εὐνήν (r’), despite his altogether different conception of the structure of the.
sentence, or how he takes dua.
2 Cf. Conington on the anacoluthon Cho. 749 ff. : "The accusative is put first, as frequently,
as the object of the sentence, as though a transitive verb were to follow’, etc.
3 Murray’s apparatus contains a different error, the assertion that in 1627 FTr have
τόνδε βουλεύσας.

772
 COMMENTARY lines 1629 ff.
of tragic grandeur, should appear in a sentence of humble content, thus
ennobling its form.
1629 ff. Here we have an echo of a popular pastime akin to the favourite
εἰκάζειν. While the εἰκάζειν! is done by one's saying to someone: ‘you look
very like this or that (a person or an animal or whatever it may be)', in the
type to be observed here the witty and unexpected coupling of a man and
something else refers not to his appearance but to his behaviour or actions.
The comparison is brought about by saying: 'you're doing (or 'are going to
do' or something of the sort) exactly the opposite of what X does; for X
acts (or ‘behaves himself’, etc.) in such and such a way, but you .. .'. A good
example of this type, in its original sphere of popular wit, occurs in Ar. Birds
30 ff. (the speaker applies the formula to himself, but that is irrelevant from
the point of view of the customary form): ἡμεῖς yàp . . . νόσον νοσοῦμεν τὴν
ἐναντίαν Σάκαι" ὃ μὲν yàp ὧν οὐκ ἀστὸς εἰσβιάζεται, ἡμεῖς δὲ φυλῆι καὶ γένει
τιμώμενοι, ἀστοὶ μετ᾽ ἀστῶν, οὐ σοβοῦντος οὐδενὸς ἀνεπτόμεθ᾽ ἐκ τῆς πατρίδος κτλ.
Similarly Plutus 1204 ff. καὶ μὴν πολὺ τῶν ἄλλων χυτρῶν τἀναντία αὗται ποοῦσι"
ταῖς μὲν ἄλλαις γὰρ χύτραις ἡ γραῦς ἔπεστ᾽ ἀνωτάτω, ταύτης δὲ νῦν τῆς γραὸς
ἐπιπολῆς ἔπεισιν ai χύτραι. Cf. also Xen. Anab. 5. 8. 24 ἣν οὖν σωφρονῆτε, τοῦτον
τἀναντία πονήσετε 7) τοὺς κύνας ποιοῦσι" τοὺς μὲν γὰρ κύνας τοὺς χαλεποὺς τὰς μὲν
ἡμέρας διδέασι, τὰς δὲ νύκτας ἀφιᾶσι, τοῦτον δέ, ἣν σωφρονῆτε, τὴν νύκτα μὲν
δήσετε, τὴν δὲ ἡμέραν ἀφήσετε. Here the general background should be taken
into consideration. Xenophon is in a pretty critical position in his relations
with his mutinous soldiers and is doing everything in his power not only to
appeal to their reason but to keep them in a good humour by means of all sorts
of witticisms and a consistently humorous tone. The comparison of himself
with a donkey (5. 8. 3) and other things of the same kind in this chapter are
parts of this proceeding, and so are the instructions just quoted on the sensible
way of dealing with the boxer Boiskos. The general arrangement of the
homely comparison is the same in Aristophanes and Xenophon’ as here in
Aeschylus: first the evavriov-statement, then its justification in two parts,
6 μὲν yap... σὺ δὲ (or ἡμεῖς δὲ or something corresponding) κτλ, The form
employed here thus leads us unmistakably into the sphere of those witticisms
with which the lower orders of society and people of modest intellectual pre-
tensions are accustomed to amuse themselves. Of the same kind as the form
is the content, the vulgarity of dragging in Orpheus and the forced βωμολοχέα
of ἦγε and ἄξηι. That Aegisthus has the face to employ such devices at such
a juncture characterizes the man who, though in origin a Pelopid, is through
and through made of common stuff. The effect must have been very startling
to the Athenians. Even without taking into account the traditional dignity
of tragic princes (of course messengers, heralds, servants, and their like stand
on a lower level), we can be fairly certain that in consequence of the strict
rules of Attic εὐσχημοσύνη no one who was the Athenian counterpart of what
is called in England ‘a gentleman’ and in Basle ‘ein wirklicher Herr’ would, in
similar circumstances, have behaved in such a way as this. Sophocles uses
an exactly corresponding means to characterize his Menelaus in the dispute
1 The custom is very old and not confined to Athens; cf. my book Plautinisches im
Plautus, 171 f.
2 And also in the rather frigid ἐναντίωσις in Isocrates 18, 66 : τοὐναντίον δ᾽ ἡμῖν συμβέβηκεν
1j rots ἄλλοις" οἱ μὲν γὰρ ἄλλοι... ἡμεῖς δὲ «rA.

773
lines 1629 ff. COMMENTARY
scene of the Ajax. There Menelaus, when driven into a corner, insults his
opponent under the transparent cover of an αἶνος or ἀπόλογος, and in so
doing he keeps in every detail the primitive form of a coarse popular custom,
whereupon Teucer pays him back in the same coin.! So far as concerns the
Aegisthus of this scene, we may surmise (though this kind of thing does not
admit of proof) that the immoderate use which he makes of proverbs and
proverbial turns of speech is intended by the poet to contribute to his
characterization. One proverbial expression we have met in 1623 and an
unmistakable proverb in 1624, another will be found in 1639 ff., and finally in
1668 (q.v.) yet another proverb is hinted at by Aegisthus. The accumulated
effect is that of a certain vulgarity. In the ironical words of Socrates in
Plato (9) Hipp. mai. 301 c the phrase φασὶν ἄνθρωποι ἑκάστοτε παροιμιαζόμενοι
(in Theaet. 162 b the tone is different) is by no means complimentary to that
class of persons.
1629. For the traditions about the singer Orpheus cf. O. Kern, Orphicorum
fragmenta, p. 14f.; Ziegler, RE xviii. 1247 ff.; I. M. Linforth, The Arts of
Orpheus, 33. The earliest literary evidence is Simonides fr. 27 D. (cf. Wila-
mowitz, Pindaros, 393), and this comes next. In Athen. Mitt. lxii[lxiv,
1938/9, 107 ff., O. Kern pictures and describes a small Attic? black-figured
plate of about 500 B.c., whose interior shows a bearded man, perhaps Orpheus,
seated on a folding chair, holding a cithara with four strings; behind and
beside him are five birds on branches and a doe.
For the Orpheus tragedy of Aeschylus, the Bassarids, cf. W. K. C. Guthrie,
Orpheus and Greek Religion, 32 ἴ., 232 f.; for representations of Orpheus on
Attic vases contemporary with Aeschylus see A. Furtwängler, 50. Berliner
Winckelmannsprogramm (1890) 163 (= Kl. Schr. ii. 531 £.).
δὲ γλῶσσαν. It is seldom that yA does not make position; cf. in general
P. Maas, Griech. Metrik, § 124, on yA in particular Denniston on E. El. 1014.
Other Aeschylean instances of a short syllable in front of γλώσσης or γλῶσσα
are fr. 169 N. (trimeter), Pers. 591 (dactyls).
1630. ἦγε: Virgil, Georg. 4. 510 (Orpheus) mulcentem tigris et agentem carmine
quercus.
Stanley’s translation attraxit omnia vocis suavitate, correct though it is in
its general sense, has been harmful to the understanding of the construction
and the more delicate nuances of the expression. Schneidewin: ‘Connect
χαρᾶι ἀπὸ φθογγῆς ᾿. This has almost become communis opinio, cf.,e.g., Paley
(‘he led everything through delight at his voice’), Wecklein, Verrall, Plüss,
Headlam (‘by rapture of his voice’), Platt (‘had all things like a flock rejoicing
in his voice’). Even from the purely grammatical point of view this is not
without difficulty, since the alleged parallels, Ag. 1366 τεκμηρίοισιν ἐξ οἰμωγ-
μάτων, Ag. 1412 ἐκ πόλεως φυγήν, are completely different, as one habitually
says τεκμαίρεσθαι Ex τινος, φεύγειν ἐκ τῆς πόλεως, but not xatpew ἀπό τινος. In
contrast, Blomfield’s explanation ‘dro φθογγῆς : vocis ope’ is linguistically
beyond reproach. ἀπό is not uncommonly used in such a way that the local
starting-point of the action can still be clearly discerned, while at the same
1 Cf. my remarks Rhein. Mus. xxiii, 1920, 366 ff., with the additional notes in Plauti-
nisches im Plautus, 346 n. 3.
2 Kern regards it as Boeotian, but Beazley informs me that it is Attic; he adds a list of
19 platelets (several of them from Athens) belonging to the same class.

774
 COMMENTARY line 1637
time the prepositional phrase performs an ‘instrumental’ function (cf. L-S
ἀπό III. 3; Kühner-Gerth, i. 458), as in 22 605 roùs . . . medvev am’ ἀργυρέοιο
βιοῖο and elsewhere. Here the ἄγειν, which draws all sorts of creatures after
it, proceeds from the voice of Orpheus. The sentence up to φθογγῆς could be
regarded as complete, but to it is added, as an additional feature, χαρᾶι, ‘in
delight’. This descriptive detail leads directly to its opposite é£opivas, while
vqm. ὑλάγμ. corresponds with ἀπὸ φθογγῆς. Correct translations are given by,
e.g., L. Campbell, ‘his utterance drew after him with delight'; Mazon, ‘lui,
par ses accents, enchainait la nature charmée'.
1631. é£opivew occurs here alone; in meaning it seems hardly to be different
from the simple öpivew (cf. on 1033 ἐκτολυπεύσειν). The epic ὀρίνειν is as yet
not attested in Tragedy; Archilochus fr. sx D. col. i. A. 57, has ópwev in
trochaic tetrameters (not noticed in L-S), Alcaeus fr. 45. 8 D. ὀννώρινε.
ἐξορίνας is presumably used absolutely here, or with an indefinite object
(anyone who hears you; above all, the speaker) ; that πάντα should be sup-
plied from 163o (Schneidewin) is unlikely.
1632. ἄξηι passive (on the form cf. on 170). Here as so often (e.g. Sept. 340)
ἄγειν ‚means ‘take into custody, to prison’.
1633. On the use of this ironic ὡς δή (quasi vero, quasi scilicet) cf. Blomfield’s
glossary ad loc., Wilamowitz on E. Her. 1407, Denniston, Particles, 229. ‘For
practical purposes, the comparative clause forms an independent sentence.’
1634 f. On οὐκ placed at the beginning of the sentence cf. on 1312, on the
repetition of οὐκ Kühner-Gerth, ii. 205 n. 2. Here and in similar instances
(cf. 167 ff.) the two inherited possibilities of placing the negative particle,
viz. either at the beginning of the sentence (or syntactical kolon) or before
the verb (cf. Wackernagel, Syntax ii. 259 ff.), are combined.
1635. δρᾶσαι, ‘putting into effect’ (B. Snell, Azschylos, 12), in contrast with
βουλεῦσαι; cf. on 1353.
τόδ᾽ ἔργον: τόδε should not be tampered with. The constant repetition of
the deictic pronoun (‘this is your murder, this is your deed of blood’) 1613 f.,
1627, 1634 f., is part of the deliberate monotony (see on 1614) by means of
which the force of the denunciation is increased. τόδ᾽ ἔργον here follows τῶιδ᾽
ἐβούλευσας μόρον just as in 1614 τόνδε φόνον follows τόνδε κατακτανεῖν.
1636. The sorry wretch, who owes all that he now is to the enterprise of
Clytemnestra and to her ἀνδρόβουλον κέαρ, is unchivalrous enough to insinuate
that he, the man, was too good for the δολῶσαι. In so doing he avails himself
of the prevalent prejudice against women (Hesiod, Erga 67 f., in the ‘Creation
of Woman’, ἐν δὲ θέμεν κύνεόν τε νόον καὶ ἐπίκλοπον ἦθος ‘Epueinv ἤνωγε, 375 ὃς
δὲ γυναικὶ πέποιθε πέποιθ᾽ 6 γε φιλήτηισιν, Eur. fr. 321 N. ἦν γάρ τις αἶνος, ὡς
γυναιξὶ μὲν τέχναι μέλουσι, λόγχηι δ᾽ ἄνδρες εὐστοχώτεροι. εἰ yap δόλοισιν ἦν τὸ
νικητήριον, ἡμεῖς ἂν ἀνδρῶν εἴχομεν τυραννίδα, and often elsewhere). In the
following clause he adds the assertion that his conduct conformed to the
dictates of political shrewdness. In a later speech of Aegisthus also, Cho.
845 f., ‘we are intended to note the contemptuous reflection, so unseemly in
his mouth, on women as the weaker sex’ (Conington, Introd. to Choeph.,
p. xxiii).
1637. Schütz: 'Ego vero ut antiquus hostis suspectus eram'; Hermann:
“ὕποπτος ἦ ἐχθρὸς wy’; Schneidewin: ' dre ἐχθρὸς ὧν παλαιγενής ' ; so most
commentators and translators. But nothing in the text points to such
775
line 1637 COMMENTARY

subordination ; on the contrary, the position of ἦ (Porson’s correct interpreta-

tion of 7; for the form cf. on 2) between ἐχθρός and παλαιγ. seems to tell against
it (Karsten). Others have sought to give expression to the subordination by
means of an indefensible alteration of 7 (ἦι Schütz, ὧν Bothe, ὡς Karsten).
It would be better to recognize that the desire for such subordination (‘as an
old enemy, I was suspect’) arises from the usages of our logic, whereas
Aeschylus felt no hesitation in putting the whole as a single complex idea:
‘Agamemnon knew that I, Thyestes’s son, had been from of old his enemy to
be regarded with suspicion and mistrust.’ Stanley’s simple translation, ‘ego
suspectus hostis antiquus fui’, hits the mark.
1638. The MSS have ἐκ τῶνδε with a stop after it. The separation into
ἐκ τῶν δὲ was taken over by Stanley from Casaubon and has been almost
universally accepted (not, e.g., by Hermann) ; cf.on 57. The word-order ἐκ τῶν
δὲ κτλ. seems to be rather unusual. Denniston, Particles, 186, cites Ar. Lys.
593 περὶ τῶν δὲ κορῶν and a few other examples,’ but does not mention Ag.
1638. Hermann’s objection, ‘nimis abrupta opum Agamemnonis mentio,
quibus sic demum recte se fretum dicere poterat Aegisthus, si eas nunc suas
esse significasset’, is met by Klausen’s correct understanding of the under-
lying idea: ‘ante omnia enim locatis satellitibus opus erat tyranno neque
adhuc habuerat Aegisthus, quo his stipendium solveret’. Cf. (van Heusde)
5. Oed. R. 540 ff. dp’ οὐχὶ μῶρόν ἐστι τοὐγχείρημά σου, ἄνευ τε πλήθους καὶ
φίλων τυραννίδα θηρᾶν, ὃ πλήθει χρήμασίν θ᾽ ἁλίσκεται ;
1639. πειθάνορα. Here alone. On the formation of the word cf. Specht,
Kuhns Zeitschr. lix, 1931, 76 n. 1: ‘A. Ag. 1639 πειθάνωρ, Sept. 224 πειθαρχία,
Pers. 374 πείθαρχος, where πειθ- cannot but come from πείθομαι᾿. [Pind. P.
2. II πεισιχάλινα must not be compared: see D. S. Robertson, Studi Castigliont
(1960), 803 ff.)
1640. After all kinds of fantastic suggestions from earlier commentators as to
what should be supplied with βαρείαις, Blomfield gave the terse and apposite
explanation 'subaudi ζεύγλαις ', referring to Prom. 462 f. Cf. also Wilamowitz
on E. Her. 681.
μὴ cannot be right. The scholiast (SoA. wad. in Tr), it is true, somehow or
other made do with it: οὐχ ὑποζεύξω τὸν un πειθαρχοῦντα ὡς κριθῶντα πῶλον,
ἀλλὰ συγκλείσας καὶ λιμώττειν ἐάσας πραῦνῶ. Sosomemoderneditors,e.g. Verrall,
‘The appearance of irregularity in this sentence (où μή or οὔτι μή being pro-
perly constructed with finite verbs, subj. or future indicative) is an appear-
ance only. In reality the negative applies by relation to the verb ζεύξω.᾽ So
tangled a construction and word-order is incredible. No less unconvincing is
Nägelsbach’s explanation, 'enuntiatio in participio contracta est ex οὔτε μὴ
ἔσται σειραφόρος κριθῶν πῶλος ᾿. In every other case where οὔτι μή occurs in
Aeschylus and Sophocles it is followed by a finite verb: Sept. 38, 199 (cf. also
281 οὐ γάρ τι μᾶλλον μὴ φύγηις τὸ μόρσιμον), Cho. 895, 5. Trach. 621, Oed. C. 450.
Many editors, including Wilamowitz, have accepted Pauw’s οὔτι μοι as a way
out of the difficulty (Hermann argued forcibly against it, but without saying
ı Hdt. 2. 159. 3 ἐν τῆι δὲ ἐσθῆτι should be removed from the list, since there τῆι is a relative
pronoun. The following should be added: Ar. Ecel. 625 ἐπὶ τοὺς δὲ καλοὺς βαδιοῦνται,
Pherecrates fr. 130. 7 K. ἀπὸ τῶν δὲ τεγῶν ὀχετοὶ βοτρύων... ὀχετεύσονται, Eupolis fr. 276.
5 K. (trimeter) ἐς τὸν δὲ φαλακρὸν ἑπτακαίδεκα (the MS reads ἐς δὲ τὸν; this substitution of
the normal for the abnormal position of δέ and γάρ often occurs, cf. Headlam, On Editing
Aeschylus, 114 f.).
776
 COMMENTARY lines 1641 f.

how he himself took the passage). To me it seems that the contrast between
the horse harnessed to the yoke and the trace-horse would forfeit some of its
sharpness if the second half ran 'by no means a trace-horse for me'. More-
over, it is not easy to see why μοι should have given place to μή, though one
might, of course, resort to that ever-present help in trouble, itacism.! On the
other hand, Wieseler's and Karsten's μήν, which was adopted, e.g., by Weck-
lein and A. Y. Campbell, seems excellent in every respect. Cf. S. El. 817 ἀλλ᾽
οὔ τι μὴν ἔγωγε τοῦ λοιποῦ χρόνου κτλ. (after this the text is uncertain, but it
is clear that Electra is setting her plan of action for the future in sharp
contrast with her former conduct), Phil. 1273 ἀλλ᾽ où τι μὴν νῦν (in sharp
contrast with his former conduct, which Philoctetes has described in the
words immediately before this). où rı μήν has here either the same adversative
force as single οὐ μήν often has (cf. Denniston, Particles, 335) or it is assevera-
tive (this I have tried to imply in the translation). The corruption is
easy. In 1068 the first hand in M reads o? μὴ instead of o? μήν, in S. Phil.
1273 some of the MSS have οὔ τι μὴ (followed, it is true, by νῦν); on the cor-
ruption of τί μήν to τί μή, of which there are several instances, cf. on 672.
On the way in which the primary notion of ζεύξω gave rise to the image of
the σειραφόρος cf. on 842.
1641. κριθῶντα, The locus classicus is Pollux 7. 23 f. τὸ μέντοι ὑπερεμπε-
πλῆσθαι kai ὑπερκεκορέσθαι καὶ ὑπερμαζᾶν ἀπὸ τῆς μάζης ἔλεγον οἱ παλαιοί, οἱ
δὲ νέοι κριθιᾶν ἀπὸ τῶν ὑποζυγίων: Αἰσχύλος μὲν γὰρ εἴρηκε ‘ceipaddpov κριθῶντα
πῶλον ',? Σοφοκλῆς (fr. 792 N. = 876 P.) δὲ ‘ews ὅτε (?) κριθώσης ὄνου ᾿. There
is a close parallel in Z 506 oraros ἵππος, ἀκοστήσας ἐπὶ φάτνηι, which 1s illus-
trated by the scholion κυρίως δὲ πᾶσαι αἱ τροφαὶ ἀκοσταὶ καλοῦνται παρὰ
Θεσσαλοῖς and Hesychius ἀκοστή" κριθὴ παρὰ Κυπρίοις. ‘For trace-horses
(σειραφόροι) particularly spirited and strong animals were chosen, because in
wheeling round the νύσσα the outside horse on the right had to make the
widest turn’ (Wecklein), cf. S. El. 721 f. Naturally they were particularly
well fed. As in 1621 f. and again in the very next sentence to this, the idea of
the prisoners suffering from hunger exercises a lively attraction on Aegisthus
so that he insists on οὔτι... κριθῶντα. In the speech of his Aegisthus (El.
1462) Sophocles transmutes this passage into the words ws . . . στόμια δέχηται
τάμα.
1641 f. ὁ δυσφιλὴς σκότωι λιμὸς ξύνοικος. H. Schöne, Hermes, |x, 1925,
160, has discussed the position of λιμός between σκότωι and ξύνοικος along
with other instances of this type of word-order. He has also protested against
the old alteration δυσφιλεῖ, which was rightly passed over in silence by
Wilamowitz, but which appears again in some of the most recent editions.
Repeatedly in this passage Aegisthus lays stress on hunger; and δυσφιλής is
at least as appropriate to hunger as to its companion σκότος. It is indeed not
hard to feel that this δυσφιλὴς σύνοικος (the second word is, of course, a
substantive) is contrasted with a φίλος σύνοικος (cf. the passage of Semonides
quoted below). An exactly similar perversion of what usually is a friendly
relation, along with an exactly similar personification, occurs in 155 οἰκονόμος
! In another corrupt passage, Prom. 606 (restored by Elmsley), some of the MSS have
τί μὴ, Others τί μοι.
2 So, as it appears from Bethe’s edition, the trustworthy MSS FS; former editors
accepted the interpolated reading of A.
777
lines 1641 f. COMMENTARY

δολία, and again in the characterization of the fatal iron sword Sept. 729 f.
κτεάνων xpnuarodairas πικρός, Which is taken up 944 f. πικρὸς δὲ χρημάτων ἴσος
δατητὰς Ἄρης. Accordingly, δυσφιλής must be retained.’ But one would have
to be a Verrall to reject σκότωι, which has lost its o through the simplest
possible haplography. The connexion with σκότωι lends a heightened vividness
to ξύνοικος ; darkness is the permanent inhabitant of prison. For the sub-
stance cf., e.g., S. El. 379 ff. μέλλουσι γάρ ae . . . ἐνταῦθα πέμψειν ἔνθα μή ποθ᾽
ἡλίου φέγγος προσόψηι κτλ., and (quoted by Scaliger) Pacuvius 158 f. Ribb.
nam te in tenebrica saepe lacerabo fame clausam et fatigans artus torto dis-
traham. In a little compass Aegisthus unites the grimmest features of a
terrible imprisonment. In general the conception of λιμός as a fellow-inmate
of the house is older: Schneidewin cites Semonides of Amorgos fr. 7. ror f. D.
οὐδ᾽ αἶψα λιμὸν οἰκίης ἀπώσεται, ἐχθρὸν (here, too, the contrast with the normal
state of things is stressed) συνοικιστῆρα δυσμενέα 0eóv,—clearly a popular
idea, cf. Sappho fr. 92 D. ὁ πλοῦτος dvev] dpéras οὐκ ἀσίνης πάροικος and other
similar passages. On the Alastor as βαρὺς ξύνοικος cf. p. 93.
1643. ἀπὸ ψυχῆς κακῆς. On the use of ἀπό cf. L-S s.v. III. 6 and above
p. 606 f., on the sense of ψυχὴ κακή cf. on 1470.
1644. The defence of σύν was undertaken by Butler, Peile, and Paley in their
day, and recently by A. Y. Campbell and Murray. σύν destroys the antithesis.?
1645. The syntactical connexion, and consequently the real meaning, of the
words ywpas . . . ἐγχωρίων is not free from ambiguity. Most of the commen-
tators take μίασμα to be simply in apposition to γυνή. This seems to be the
obvious interpretation: moreover Klausen compared Cho. 1028 μητέρα. ..
πατροκτόνον μίασμα καὶ θεῶν στύγος (but this line is possibly interpolated, see
Appendix C). On the other hand, Wecklein regards μίασμα as ‘in apposition
to the content of the sentence’ (or to the ‘actio verbi’, cf. on 47), and so do,
e.g., Verrall, Lewis Campbell (‘slew him, to the pollution of Argolis and the
desecration of the Argive deities’), Wilamowitz (‘Aber dieses Weib erschlug
ihn. Auf dem Land und seinen Göttern ruht Entweihung’), Platt (‘didst
suffer his wife to kill him, whereby shall come a plague upon all the land and
all her gods’). Plüss wavers between the two possibilities. I incline towards
the second, because the apposition to the sentence seems to me to strengthen
the whole, and also because in the present scene (in contrast to the preceding
one) we expect not so much a violent attack upon Clytemnestra herself as
some expression stressing the atrocity of the crime. But I have not arrived
at a definite conclusion. Against taking the words as in apposition to the
sentence it might perhaps be argued that they are not preceded by the main
verb (ἔκτεινε), as is commonly the case with sentence apposition (cf. the
examples collected by Wilamowitz on E. Her. 59, and also, e.g., 4 27 f., A.
Eum. 376, E. Hipp. 755 .). S. Oed. R. 603 ff. is not a real instance of the con-
1 [f it may be supposed that Sophocles had Ag. 1641 f. in mind when, in Oed. C. 1258 f.,
he wrote τῆς ὁ δυσφιλὴς γέρων γέροντι συγκατώικηκεν (cf. ξύνοικος) πίνος, we should have a
confirmation of the necessity οὗ accepting δυσφιλής. The rarity of δυσφιλής (outside the
Oresteia it is found only in this passage of the Oed. C.; cf. on 34) makes the reminiscence
probable. Peile cites the passage, but decides for δυσφιλεῖ.
2 It is hard to understand how Paley, following the example of Bothe (‘ αὐτός, solus. σύν,
simul’), could say : αὐτός, “alone”, requires the antithesis of ody’. Does αὐτοκτόνως in
1634 f. (where exactly the same contrast appears as in 1643 ff.) mean ‘killing alone’ and not
‘killing with your own hand’?

778
 COMMENTARY line 1647

trary, for here τῶνδ᾽ ἔλεγχον is placed in front because it provides the link
with the preceding section of the speech; moreover, it is followed by τοῦτο
μὲν... τοῦτ᾽ ἄλλο. A special reason for the position at the beginning is
obvious in enumerations or headings of a fresh paragraph as, e.g., E. Her. 196
ἕν μὲν τὸ λῶιστον κτλ. or Hec. 1168 f. τὸ λοίσθιον δέ, πῆμα πήματος πλέον,
ἐξειργάσαντο δεινά, where πῆμα πήματος πλέον is necessarily placed immediately
after τὸ λοίσθιον δέ. Nor should S. El. 563 f. be quoted as evidence for the
sentence apposition coming first, for the question τίνος ποινάς could not be
placed otherwise. But an instance such as E. Phoen. 1219 f. rw παῖδε τὼ σὼ
μέλλετον, τολμήματα αἴσχιστα, χωρὶς μονομαχεῖν παντὸς στρατοῦ shows perhaps
that the order is sometimes freer, although it should be noticed that here one
element of the verbal expression, μέλλετον, comes first (in E. Hel. 77 the text
is not absolutely certain). [Cf. Wilamowitz on E. Her. 992.)
The additional phrase χώρας μίασμα κτλ. is not superfluous decoration or a
mere outburst of emotion: it is in the extreme sinfulness of the murder (or
the murderess) that the warrant for its expiation is to be found; so this
thought provides a transition to the retribution at the hands of Orestes.
1647. κατελθών: cf. on 1283.
1648. ἀμφοῖν τοῖνδε shows how this sentence (from 1646 "Opeorns onwards)
ıs to be spoken, and how the end of this part of the scene should be
played (cf. on 965). The coryphaeus has turned away from Aegisthus, and is
speaking (in the language of the modern stage) ‘aside’, or (in terms of
Aeschylean ideas) to the gods, to whom alone such a prayerlike wish can be
addressed.’ For this sentence has something of the nature of prayer: not
only do the words πρευμενεῖ τύχηι point to this, but also the whole content of
the sentence. It is true, no particular gods are addressed here, but we have
just be:n reminded of the pollution inflicted on the ἐγχώριοι θεοί. Like so
many scenes in the Oresteta, the iambic part of this scene closes with a prayer
or at least something very like it. There is a stirring effect in the parallelism,
and the contrast, to the prayerlike wish of the faithful servant at the end of
the prologue : γένοιτο δ᾽ οὖν μολόντος εὐφιλῆ χέρα ἄνακτος οἴκων τῆιδε βαστάσαι
χερί. Between the wish for the return of Agamemnon and the wish for the
return of Orestes the gigantic tragedy has unrolled itself.
παγκρατής (cf., in addition to the dictionaries, Schuursma 122 f.) is used
here in a rather different sense from that which it has elsewhere in lyric
poets (hence Ar. Thesm. 317 and 368 in lyrics) and the tragedians,? where as
an epithet of gods and divine or quasi-divine beings (and also of their im-
plements) it indicates their power.? But since κράτος from Homer onwards
also means ‘victory’ and κρατεῖν often (though not yet in Homer) means ‘to

ı W. Schadewaldt, Monolog und Selbstgespräch, 53, in summing up his observations on

Aeschylus, says: ‘It is of course his emotion that causes the actor to turn away from the
others on the stage, but all the same the man is not absorbed in his own thoughts, self-
centred and solitary: it is the presence of the divine power that confronts him now to
which he pours out his feelings.’ He might have referred also to Ag. 1646 ff. Cf. on 1582.
2 παντοκράτης in the Ξάντριαι(), P. Oxy. 2164, fr. 1. 14 [τῶι] mavro«pd[rm] . . . Ζηνέ
(anapaests), is perhaps a fresh coinage of Aeschylus, cf. on 221 παντότολμον.
3 The Latin omnipotens is a ‘Bedeutungslehnwort’, formed after παγκρατής. As little as
its Greek prototype does it primarily belong to the language of ritual: it is confined to
poetry (cf. my Plautinisches im Plautus, 208), whence it found its way into the vocabulary
of Christianity.

779
line 1648 COMMENTARY
win the mastery’, ‘to be victorious’, it seems natural that it means here
‘all-victorious’, ‘triumphant’.
1649 ff. Even if neither Aristotle nor anyone else had told us anything about
the difference in character between iambic trimeters and trochaic tetra-
meters, we should probably be immediately conscious here of the jerk with
which the rhythm changes to greater excitement, just as in some of Mozart’s
operas we are affected by the change of measure in the finale of an act. In all
that remains to us of Greek drama the only parallel to be found—it is a very
close one—is in the final scene of the Oedipus Rex (cf. Wilamowitz, Hermes,
xxxiv, 1899, 67) : in the closing scene of Euripides’ Jon there is also, it is true,
a change over from trimeters to tetrameters, but without any excitement in
the action: ‘the finale is devoid of interest, as is usual when it gives merely the
execution of the orders of the deus ex machina’ (Wilamowitz on Ion 1606), At
the time of the Oresteia the prevalence of trochaic scenes in Tragedy had long
ceased ; where Aeschylus made use of them, ‘a change to a higher key is the
indispensable condition’ (Wilamowitz, Verskunst, 265 ; for a slightly different
view see P. Maas, Griech. Metrik, § 75). Cf. on 1344 ff. (p. 633 f.).
1649. δοκεῖς : cf. on τό.
ἔρδειν. Though as yet nothing but words has passed between them,
Aegisthus sees active resistance in the attitude of the Elders and a first step
towards revolutionary activity ; so he concludes that they have further steps
in mind (δοκεῖς), He means to anticipate any more energetic action on their
part by the orders he gives to his men-at-arms.
γνώσηι τάχα: Hermann, on the same lines as Butler before him, observes:
‘statim cognosces—. Non eloquitur sententiam, gestu se gladium stringere
velle significans.' This shows that like others he misunderstood the nature of
this form of speech. Paley says rightly: ‘a formula of threatening'. The
parallels he quotes (in his note on Cho. 305 τάχ᾽ εἴσεται)" show that in this
expression in ordinary speech it was quite usual to omit the object: E.
Heraclid. 65 γνώσηι σύ, Suppl. 580 γνώσηι σὺ πάσχων, Theocr. 22. 63 γνώσεαι
and, exactly like Ag. 1649, [Theocr.] 26. 19 Αὐτονόα τόδ᾽ ἔειπε ' τάχα γνώσηι
πρὶν ἀκοῦσαι᾽.2 Like Hermann, but with worse consequences, Wilamowitz
mistook the function of γνώσηι τάχα :* after 1649 he inserted 1664, in which he
wrote σώφρονος γνώμης θ᾽ ἁμαρτὼν and a supplementary θ᾽ ὑβρίσας to end the
line, and so (as his translation shows) used the participles to supply the con-
tent of γνώσηι. His violent transposition* needs no further refutation. In-
cidentally, there is no reliable example in the text of the Agamemnon of a line
having been transposed to any considerable distance (for 1290 see note): in
1203 f. two consecutive lines in stichomythia have exchanged places.
1 He might have quoted as further instances of the unmistakably colloquial τάχ᾽ εἴσεται
and the omission of the object Ar. Clouds 1144 τάχα δ᾽ εἴσομαι, Birds 1390 σὺ δὲ κλύων εἴσει
τάχα, Lys. 1114 τάχα δ᾽ εἴσομαι ᾽γώ, Plut. 647 εἴσει τάχα (in Wasps 1224 τάχ᾽ εἴσομαι is re-
stored with a high degree of probability). Cf. also E. Hel. 811 εἴσηι, Iph. A. 675 εἴσηι σύ.
2 Here even Ahrens misunderstood the character of the expression, for he deliberately
changed the traditional punctuation to Aëürovda * τόδ᾽ ? ἔειπε ‘ τάχα γνώσηι... à.
3 In the main his treatment of this passage follows the line of the proposals made by
Keck, R. Arnoldt (D. Chor im Ag. des Aesch. 85 f .), and Lehrs (cf. Arnoldt, loc. cit. =
K. Lehrs, Kl. Schr. 212), who assume that a line has dropped out after 1649 in "which there
had been an object for γνώσηι, while Weil (1858) proposed to introduce this object into
l. 1649 by altering the text.
4 It is precisely in this detail that A. Y. Campbell has followed Wilamowitz.
780
 COMMENTARY line 1650
1650. In the MSS this line is assigned to the Chorus (i.e. the coryphaeus).
The grave objections to this arrangement were recognized by Stanley. He
gave 1649-51 (in his commentary, not in his text) to Aegisthus, 1652 to the
Chorus, and 1653 to Aegisthus. In assigning τόσο to Aegisthus he has been
followed by the great majority of editors. As this involves very serious
consequences, we cannot adopt it without asking ourselves once more whether
it is really impossible to accept the assignment of the lines to the different
speakers given in the MSS. The few critics who wish to keep the coryphaeus
as the speaker in 1650 find their view supported by an assertion of Otfried
Müller’s. He says (Aesch. Eum. p. 82): ‘Now in the treatment of the Chorus
there is a strangely persistent analogy drawn between them and a group or
band of warriors (lochos) armed for battle; thus Aeschylus likes to call the
Chorus itself a lochos, and actually in the Agamemnon he makes the Elders
advance against Aegisthus as λοχῖται gripping their sword-hilts.’ The state-
ment that Aeschylus ‘likes to call the Chorus itself a lochos’ is not exactly
correct, and put in this form it is misleading. In point of fact the poet uses
λόχος not only of a military formation but also as meaning ‘crowd, company’
in general, as in Eum. 1026 f. εὐκλεὴς λόχος παίδων γυναικῶν (N.B. not with
reference to the Chorus), and so too when the company happens to be that
which constitutes the Chorus, Sept. 112, Eum. 46 (presumably also fr. 379 N.).
This cannot, however, justify Müller’s reinterpretation of λοχῖται. In the
only other two passages where this word occurs in the fifth century, it clearly
means the armed retainers, the bodyguard, the δορυφόροι, of a ruler: Cho. 768
(actually of Aegisthus’ bodyguard) and S. Oed. R. 751. This makes it utterly
improbable that the line should, in accordance with the MSS, be assigned to
the coryphaeus, as was done by, e.g., Conington, Hartung (he introduces semi-
choruses at this point), Nägelsbach, van Heusde, R. Arnoldt (Der Chor im
Ag. des Aesch. 84 f.), Plüss (who also has semi-choruses ; his explanation of
λοχῖται, ‘the elders have at times been on campaigns together’, is fanciful) and
Tucker (see below). van Heusde’s compromise: ‘sunt satellites (δορυφόροι),
choro praesidii causa adiuncti’ cannot be taken seriously.
Apart from the assignment of 1650 to Aegisthus, Stanley’s arrangement of
the lines is not worth considering. We should not need to discuss it at all,
had not Verrall and recently G. Thomson followed Stanley’s lead in one
important point, that is to say in giving 1652 not to Aegisthus, but to the
Chorus (it is of minor importance that Thomson supposes 1653 to be spoken
by Aegisthus, as Stanley proposed, while Verrall gives it to a Aoxírgs of
Aegisthus). The impossibility of this distribution of the characters was con-
clusively established by Hermann :“Absurdum est Argivis mortem portendi,
quod omen in Aegisthum expetere, qui Agamemnonis necem capite luiturus
erat, oportebat.' That the Elders should die before long is natural enough;
Aegisthus, however, is now in the prime of life: those who are hoping for his
death are bound to welcome the unlooked-for encouragement given them by
an oracular word, a φήμη. Little depends on the death of the Elders, while
everything depends on that of Aegisthus. It is also evident that this veiled
prophecy of Aegisthus' death forecasts the play which is to follow, as do so
many passages in the last scenes of the Agamemnon. 1652 must therefore be
left to Aegisthus.
G. Thomson indicates quite clearly his reasons for following Stanley and
781
line 1650 COMMENTARY

Verrall in the distribution of the speakers. He is impressed by ‘the absurdity,

against which the whole play cries out, of supposing that these old men,
ἐσχὺν ἰσόπαιδα νέμοντες ἐπὶ σκήπτροις, have been carrying arms unknown to us,
ὑφ᾽ εἵμασι ξίφη λαθραϊῖ᾽ ἔχοντες ’. It is not only the right of an interpreter but
his duty to express surprise at a thing which is indeed astonishing and
unusual. But it is arbitrary to make violent alterations in the text in an
effort to get rid of what is open to objection. For Thomson cannot attain his
object (which is to represent the Elders of the Chorus as unarmed) simply by
taking away from the coryphaeus line 1651, in which the ξίφος is mentioned.
He is further obliged to change πρόκωπος in 1652, although, as will be shown,
this word is ineradicably established in the context of this closely knit
stichomythia. Yet even that is not the end of the sacrifices Thomson has to
make. Clytemnestra requests the Elders to go home πρὶν παθεῖν ἔρξαντες
(1658). This unobjectionable phrase Thomson is forced to alter (as others
before him have done). ép£avres implies the same idea which is conveyed by
this tumultuous finale as a whole, namely that until Clytemnestra intervenes
at the last moment there is a dangerous threat of bloody conflict between two
sets of armed antagonists. Now if the Elders are to be thought of as armed,
the most natural assumption is that they, too, carry swords.! Obviously not
hidden swords which they only produce in the closing scene to the surprise of
the audience. If we assume the existence of the swords at all, we must picture
the appearance of the Chorus from the very beginning in the manner indi-
cated by Wilamowitz, who in his translation (at the beginning of the parodos,
l. 40) gives this stage-direction : ‘The Elders carry swords at their sides and
long staves in their hands.’ This is very remarkable. There is no parallel
instance, either in Tragedy (for its characteristic treatment of old men with
staves cf. the footnote below) or (as Beazley stated in answer to a ques-
tion) in vase-painting, where we never find that typical old men, furnished
! This assumption has been almost universally adopted ; it was, however, rejected before
G. Thomson by T. G. Tucker. In an article C.R. xi, 1897, 404 f., and in his translation of the
Agamemnon (Melbourne 1935) Tucker gives 1650 to the Chorus (following the MSS), 1651 to
Aegisthus, 1652 to the Chorus, and 1653 to Aegisthus. He translates 1650 by: ‘Come,
friends of my company!’ Against this (and against the assignment of the line to the Chorus)
what was said above on λοχῖται should be borne in mind; to this we may add Hermann’s
pertinent remark about the mistake of assigning the ominous words about death not to
Aegisthus, but to the Chorus. In order to get rid of the Elders’ swords completely, Tucker
translates 1652 (which he puts in the mouth of the Chorus) : ‘Nay, I too shrink not from meet-
ing death, weapon in hand’, and then adds the stage direction: ‘They brandish their
staves.’ This leads us directly to the question: what becomes then of πρόκωπος ἢ do the
staves of the Elders possess κῶπαι Still more disastrous is the grotesque, indeed farcical,
effect produced when we picture the old men defending themselves with their wooden
sticks against the swords of Aegisthus’ bodyguard. It is of course an altogether different
matter if helpless old men, by way of threat or wish, say that they will (or would like to)
make their opponents’ heads bleed with their staves. Thus E. Andr. 588 (Peleus) σκήπτρωι
δὲ τῶιδε σὸν καθαιμάξω κάρα (there is no reason for making this a question, as Murray does),
Her. 254 ff. (the aged coryphaeus is addressing his companions) οὐ σκῆπτρα, χειρὸς δεξιᾶς
ἐρείσματα (cf, the beginning of the parodos, 107 f., ἀμφὶ Bdxrpots ἔρεισμα θέμενος), ἀρεῖτε καὶ
τοῦδ᾽ ἀνδρὸς ἀνόσιον κάρα καθαιματώσετε; This scene of the Herakles, by the way, well shows
the gulf dividing the old man, with only a staff, his γεροντικὸν ὅπλον, at his disposal, from
the swordsman in the prime of life: Amphitryon says 232 ff. εἰ δ᾽ ἦ νέος re «dre σώματος
κρατῶν, λαβὼν ἂν ἔγχος τοῦδε τοὺς ξανθοὺς πλόκους καθηιμάτωσ᾽ ἄν, This contrast helps
us to realize how unusual a feature are the old men’s swords in the closing scene of the
Agamemnon.
782
 COMMENTARY line 1650

with sticks, at the same time carry swords.! Yet the language and action of
the last scene seem to necessitate? the conclusion that Aeschylus has done the
unusual thing and for the sake of a stirring effect at the end? has let the Chorus
carry swords, despite the fact that such an arming of ‘civilians’ probably
involved a departure from ordinary Athenian custom at the time of Aeschylus
(cf. Thuc. 1. 5 £). The question why the Elders, who could not possibly
anticipate such a collision, should, at the beginning of the play, appear with
arms would never enter the heads of the audience. For although the swords
are there from the start, they are practically non-existent for the minds of the
spectators, who are fully occupied in taking in what is being offered to their
eyes, their ears, and their minds. Here, as elsewhere in fifth-century drama,
it has to be taken as a matter of course that for the spectator any object on
the stage is practically non-existent until its presence is expressly mentioned.
G. Thomson has not, of course, been alone in noticing that the final scene,
in which the Elders brandish their swords, seems inconsistent with the
parodos where they are represented as feeble and quite incapable of any
warlike action. Wilamowitz, Znterpr. 166, endeavoured to suggest an excuse,
though without much confidence: ‘Der Chor [in the parodos] . . . schildert
seine Gebrechlichkeit mit so starken Farben wie es . . . euripideische Manier
ist; sein Verhalten gegen Aigisthos wird dazu wenig stimmen. Das [i.e. the
attitude in the parodos] war also wohl [this qualification is important] schon
konventionelle Stilisierung des Chores von Greisen, der so háufig auftrat.' It
is not very satisfactory to assume that a conventional device of later tragedy
is anticipated in the Agamemnon* and that the carefully elaborated Chorus of
this play is but another example of the typical representation of old age. As

1 With younger men, the combination of sword and stick, though rare, is sufficiently
attested in works of Attic art. 1. M. Cook, Annual Brit, School at Athens, 35, 1938, 184 f.,
pl. 50, pictures and describes an amphora in New York, belonging to the “Early Proto-
attic’ period (c. 710-680 B.C.); on its neck we see the lone figure of ‘a gentleman of very
smart appearance; in spite of his sword he is not a soldier, but one of the processional
dignitaries with long staves who later become popular in Middle Protoattic’. Beazley
points out to me the following instances: the relief on a tomb-stele from the Kerameikos,
first half of the 6th century, published by G. Karo, An Attic Cemetery (Philadelphia 1943),
plate 18, which shows a naked man with a sword at his left side and a stick in his right
hand; the picture of the mission to Achilles by the Eucharides painter (Beazley, Attic
Red-fig. Vase-Painters, 154 no. 11), figured in K. Friis Johansen, Iliaden 1 tidlıg gresk
kunst, Copenhagen 1934, fig. 28, where the man at the right-hand side wears over his right
shoulder a sword-belt (partly covered by the himation) and carriesa stick ; the picture, in all
probability representing the Greek chieftains at Troy, on the back of the skyphos by the
Brygos painter in Vienna (ARV, p. 253 no. 129), figured in Furtwängler-Reichhold, text II,
p. 122, where the third man from the left again shows the combination of sword-belt and
stick.
21 use this strong expression deliberately. Impressed by G. Thomson's arguments I
made a persistent effort, first by myself and then in repeated discussions with Rudolf
Pfeiffer, to try every possible means of getting away from the assumption that the Chorus
carried swords. But every attempt led to absurd results.
3 [t would seem that the poet had already mentioned the swords of the Chorus in an
earlier passage of the play if we could refer the expression in 1351 νεορρύτωι ξίφει to the
swords of the old men. But there it is necessary to take the ξίφος as the weapon with which
the Elders suppose Agamemnon to have been murdered.
* On very similar grounds it has been necessary to reject the view of W. Kranz that the
Chorus of the Agamemnon should be regarded as nothing more than an accompanying
instrument (cf. p. 248).
783
line 1650 COMMENTARY
for the parodos, the emphasis which is laid on the extreme feebleness of the
Chorus serves a special purpose there. It has to be made perfectly clear why
the Elders, though eyewitnesses of the events in Aulis, could not possibly
have joined in the expedition. Ten years ago, at the outbreak of war, they
were already too old to be of any military use: now they seem to be tottering
to their graves. Besides, this detail helps to bring out the magnitude of the
war and the devastation wrought by it: every man save the hopelessly unfit
had to fight before the walls of Troy. In the last scene of the play the position
is different. What matters here is that the Elders, the only representatives of
the Argive people who can be brought on the stage, offer bold resistance to
the threats of the usurper. As was stated on a former occasion (cf. p. 249),
the reason why the members of the Chorus at a particular moment behave in
a particular way must be found to a large extent in their function of leading
up to, or reacting to, what is said and done by the main actors. This sometimes
involves changes which, when judged by ordinary standards of probability,
may look like inconsistencies. But here it should again be noticed that the
behaviour of the Elders is in keeping with their general outlook. They
represent the honest and loyal citizens of Argos. In their hearts they side
with the legitimate king and his son (1646, 1667). They gave Agamemnon
what warnings they could and would gladly have saved his life had they not
been utterly powerless. Now they are strongly provoked by the insolence of
Aegisthus. Their physical weakness, which had been over-emphasized for
the purpose of the parodos, is in the last scene somewhat neglected to make
the clash appear more serious. Aeschylus was indeed doing a bold and
unusual thing when he supplied these old men with swords. It is unlikely
that he could have done the same in the case of an individual actor. But with
the Chorus he had greater freedom : without destroying the general outline
of their character, he could at one time stress their old age and at another
time their loyalty and determination to resist indignities. In either case the
choice of the greater or lesser emphasis depended up to a point upon con-
siderations of dramatic contrasts. The poet had no need to be afraid lest
κομψός τις ἔροιτο θεατής: Why have you not been consistent? Why do the
Elders in the last scene behave in a manner that belies their own statement
in the parodos?
It follows from what has been said that unless we want to create insuper-
able difficulties, we must retain the arrangement of speakers which has been
followed by most editors since Bothe and Hermann (his unnecessary assump-
tion of a lacuna after 1649 need not be taken into account). That means that
the only alteration we make ‘against the MSS’ is to assign 1650 to Aegisthus,
i.e. the speaker of the preceding line. Here, as so often (cf. on sor), it was
probably a wrongly added παράγραφος that caused the mistake. Perhaps we
may also consider Karsten’s hypothesis: ‘Error ex eo facile natus, quod
librarius putaret λοχέτας intelligendos esse choreutas.’
1650. It is possible that for Aeschylus and his audience εἶα δή had the ring of
colloquial speech about it. We cannot judge of this with certainty, but it is
a fact that apart from this passage «la δή appears in drama, so far as we know
at present, only in Comedy (Ar. Thesm. 659 similarly at the beginning of a
trochaic tetrameter) and in satyr plays, 1.6. in Aeschylus’ Θεωροί (Pap.
1 Cf, p. 248 f. above.
784
 COMMENTARY line 1650

Oxy. 2162, fr. 1, col. ı. 18, again at the beginning of a trochaic tetrameter), in
Sophocles’ ᾿Ϊχνευταί (fr. 314. 87 P.) and in Euripides’ Συλεύς (fr. 693 N.).
This might be only coincidence. But if εἶα δή really sounded to the Athenian
like a word ‘from the street’ or, what is more to the point here, ‘from the
parade-ground’, it would be in keeping with the kind of expressions used by
Aegisthus elsewhere (cf. on 1628 and 1629 ff.) ; the Chorus adopts the same
provocative word of command in the next line. Euripides has a partiality for
ἀλλ᾽ εἶα, which is also used by Sophocles in the Eurypyius (fr. 221 P.)' and is
common in Aristophanes. εἶα without ἀλλά is used in Euripides IPA. T. 1423,
Hel. 1597, perhaps also in the Phaethon (fr. 781.8 N.), and often in Aristophanes.
φίλοι λοχῖται. For λοχῖται see above. R. Arnoldt, Der Chor im Ag. des
Aesch. 85, has objected to φίλοι in the mouth of Aegisthus: ‘it is not in keeping
with the relation between the commander and his retainers, nor with the
harsh and imperious temper of Aegisthus. Might we not use exactly the
reverse argument, and say that it is in keeping with Aegisthus’ method that
he adopts a studiously friendly attitude towards the underlings on whose
help he now depends, while, bully that he is, he attempts to intimidate the
Elders?
τοὖργον οὐχ ἑκὰς τόδε, The phrase οὐχ ἑκάς serves as predicate to τοὖργον
τόδε, ‘this work of yours here’, ‘what you have now got to do’. 8. Phil. 26
τοὖργον οὐ μακρὰν λέγεις was quoted by Stanley for comparison. Aegisthus
probably accompanies his words with an expressive gesture. In Ar. Peace
426 Hermes,’ before giving his orders in detail, says: ὑμέτερον ἐντεῦθεν ἔργον,
ὦνδρες.
Those who persist in ascribing the two lines 1650 f. to the Chorus (see above)
might have produced in support of their view an argument which the use of
ela apparently affords. For elsewhere we almost invariably find in a sentence
containing an ela either an imperative or some other hortatory expression
(E. Iph. T. 1423 ff. οὐκ ela . . . δραμεῖσθε καὶ... δέξεσθε....; Hel. 1561 ff. οὐκ
ela... ἐμβαλεῖτε... .; 1597 ff. οὐκ el’ ὁ μέν τις... ἀρεῖται... .; Or. 1622 οὐκ εἶα
. . . βοηδρομήσετε; Or. 1060 f. ἀλλ᾽ el’ ὅπως . . . κατθανούμεθα, Ar. Thesm. 659
εἶα δὴ πρώτιστα μὲν χρὴ κοῦφον ἐξορμᾶν πόδα). Far the commonest type, with
an imperative, is in fact also present in E. Hel. 1429 ff. : ἀλλ᾽ εἶα" τοὺς μὲν Πελο-
πιδῶν ἐῶ νόμους" καθαρὰ γὰρ ἡμῖν δώματ᾽ - οὐ γὰρ ἐνθάδε ψυχὴν ἀφῆκε Μενέλεως"
ἴτω δέ τις φράσων ὑπάρχοις τοῖς ἐμοῖς κτλ., for here the trw τις is clearly in-
tended from the start, but owing to the intrusion of the parenthesis it is
postponed. An exception which is only a seeming one is the long section in
Aristophanes’ Peace (459 ff.) where, after the request ὑπότεινε δὴ πᾶς καὶ
kdraye τοῖσιν κάλωις the real work is accompanied by the exclamations ὦ εἶα,
εἶα ὦ etc. Despite this restricted use of εἶα we should not jump to the con-
clusion that both the lines 1650 f. must either be left to the Chorus or in any
* The final A of ἀλ]λ᾽ can be clearly recognized in the photograph, Pap. Oxy. ix, plate IV,
fr. gr. This instance of ἀλλ᾽ εἶα has been overlooked by Denniston, Particles, 14, and by
Page on E. Med. 401. The restoration [ἀλλ᾽] ela is probable in the Δικτυουλκοί of Aeschylus,
Pap. Oxy. 2161, col. 2. 23. In Aeschylus fr. 78 N. ela is only a doubtful supplement, moreover
the attribution of the fragment to Aeschylus has been disputed by Wilamowitz, Interpr.
I9 n. 2.
* For some inexplicable reason Coulon (following Lenting) gives these words to Trygaios
without so much as mentioning that the reading of the MSS is not ἡμέτερον, which he prints,
but ὑμέτερον, [In his second edition Coulon prints ὑμέτερον.
4872.3 X 785
line 1650 COMMENTARY

case given to a single speaker as an undivided sentence in order to provide

the first εἶα δή as well as the second with its imperative in εὐτρεπιζέτω. Nor
is there any need to suppose that Aegisthus is interrupted by the coryphaeus
before he has time to give a specific order. For the whole situation and in
particular the strongly deictic roëpyov τόδε, which was probably accompanied
by an appropriate gesture, make it perfectly clear what kind of action is
demanded from the λοχῖται. The assignment of 1650 f. to one and the same
speaker cannot be considered at all. Quite apart from the devastating conse-
quences for this whole section of the scene (see above), we may be guided by
the telling way in which 1651 and the two following lines are clamped together
so that at each change a word of the first speaker is combatively taken up by
his opponent in the next line (1651 πρόκωπον: 1652 πρόκωπος; 1652 θανεῖν:
1653 θανεῖν) ; from this observation it follows with great probability that there
is the same relation between the εἶα δή of 1650 and the ela δή of 1651.
1651. εἶα δή... πᾶς. This is reminiscent of the formula of encouragement
obviously borrowed from everyday life in Ar. Peace 517, 519 εἶα πᾶς, similarly
S. Ichn. 87 dy’ ela δὴ was... .
TpókwTov. I cannot speak with confidence about the formation of the
word and in particular the function of προ-. It is only as an hypothesis that
I suggest the possibility that πρόκωπος was coined (by Aeschylus?) on the
analogy of πρόχειρος. The use of πρόκωπος, which is altogether rare, seems
originally to have been confined to the phrase τὸ ξίφος πρόκωπον ἔχειν (or
with some similar verb). Now we find an exactly corresponding use of 7ó
ξίφος (or φάσγανον and the like) πρόχειρον ἔχειν or something similar in, e.g.,
S. Phil. 747 f. πρόχειρον εἴ τί aot, τέκνον, πάρα ξίφος xepoiv,! πάταξον krÀ., E.
El. 696 πρόχειρον ἔγχος χειρὶ βαστάζουσ᾽ ἐμῆι, Hel. 1563 f. φάσγανόν θ᾽ ἅμα
πρόχειρον ὠθεῖ. It is true that the particular combination of πρόχειρον and
ξίφος is not attested till after Aeschylus, but that is probably only accidental,
cf., e.g. the similar phrase in Prom. 54 Kal δὴ πρόχειρα ψάλια δέρκεσθαι πάρα.
The origin of πρόχειρος thus used is clear: it is formed, as was recognized by
Usener (Kl. Schr. i. 252), by way of ‘hypostasis’ (cf. on 105 evreAewv) from πρὸ
χειρῶν (cf. also Wilamowitz on E. Her. 161) ; πρὸ χειρῶν (S. Ant. 1279, etc.) is
used exactly like prae manu (Plaut. Bacch. 623 etc., cf. Solmsen, Glotta, iii,
1912, 249). ξίφος πρόκωπον was possibly formed directly after the model of
ξίφος πρόχειρον (ἔχειν or the like), with complete disregard for the meaning of
mpo-. However that may be, there is no doubt about the meaning of πρόκωπον:
in Ag. 1651 it is the same as in the passages cited by Stanley, viz. E. Or. 1477
ὁ δὲ ξίφος πρόκωπον Ev χεροῖν ἔχων and the two passages which Suidas s.v.
quotes from unknown historians (derived from the Excerpta Constantiniana),
(1) ὁ δὲ ὀργισθεὶς κατὰ τοῦ βασιλέως πρόκωπον ἔχων τὸ ξίφος, (2) ἀλλὰ γεννικὸν
ὁρῶν εἶχε πρόκωπον τὸ ξίφος. Similarly, to all appearance after the model of
earlier writers, the historian Herodian 7. 6. 8 ἔχοντές re αὐτὰ (scil. ra ξιφίδια)
πρόκωπα προπηδῶσιν (and, with a slightly different use, ibid. 7. 5. 4 ἔχων
πρόκωπον τὴν δεξιάν). So τὸ ξίφος πρόκωπον ἔχειν means 'to hold the sword by
the hilt (ready to strike)’, cf. S. Phil. 1254 f. χεῖρα δεξιὰν ὁρᾶις κώπης ém-
ψαύουσαν ; In the present passage πρόκωπον stands predicative to εὐτρεπίζειν
as it does elsewhere to ἔχειν.
ı If Denniston had noted this passage, he might not have remarked on E. El. 696:
“πρόχειρον... χειρί, An awkward repetition.’

786
 COMMENTARY line 1652

1652. ἀλλὰ κἀγὼ μὴν κτλ. It is to the credit of Ahrens (637) and Headlam
that they rejected Porson’s transposition ἀλλὰ μὴν κἀγώ, which is usually
adopted. As regards the separation of ἀλλὰ... μήν, the ‘parallels’ quoted by
Headlam are indeed irrelevant, for two of them, Pl. Laws 644 d and Theocr.
5. 122 (quoted by Ahrens too), contain no ἀλλὰ μήν at all, and in the others
we find ἀλλ᾽ οὐδ᾽ ἐγὼ μήν and the like, i.e. the quite usual insertion of the
negative between ἀλλὰ and μήν. But Denniston’s collection of passages,
Particles, 341 ff., is helpful. There we find the following examples of aa...
μήν used in a positive statement:! S. Ichn. 109 ἀλλ᾽ αὐτὰ μὴν ἴχίνη Te] χὠ
στίβος, Oed. C. 28 ἀλλ᾽ ἐστὶ μὴν οἰκητός, then (in the 4th century?) “Epichar-
mus’? fr. 170. 3 Kaibel ἀλλὰ λέγεται μὰν χάος πρᾶτον γενέσθαι τῶν θεῶν and
from prose PI. Clit. 407 a ἀλλ᾽ αἰσχρὸν μὴν σοῦ γε ὠφελεῖν με κτλ. In the present
passage it is obvious that the purpose of the unusual, but admissible, separa-
tion of the particles is to bring the strongly emphatic κἀγώ as near as possible
to the beginning of the sentence. The emphasis laid on the words ‘I too’ is
accounted for by the nature of the clash between the Elders and Aegisthus.
A very similar passage (also in regard to the action and the frame of mind of
the persons) is S. Phil. 1254 f., where after Odysseus’ challenging words χεῖρα
δεξιὰν ὁρᾶις κώπης ἐπιψαύουσαν;; Neoptolemus replies ἀλλὰ κἀμέ τοι ταὐτὸν τόδ᾽
ὄψηι δρῶντα κοὐ μέλλοντ᾽ ἔτι. Aegisthus, in his quarrel with the Chorus,
several times begins his remarks with an ἐγώ (standing second in the sen-
tence), cf. 1666, 1668.
Now for the metrical point. There is a familiar rule which, though it does
not apply to the homely trochaic tetrameters which are below the level of
literature proper (cf. Hermes, lxii, 1927, 366 f.) or to those of comedy, holds .
good for the tetrameters of the old iambographers? and those of Tragedy.
According to that rule, at the end of a μέτρον, apart from the place before the
diaeresis in the middle of the line, there must be no word ending in a long
syllable* or one lengthened by position (cf. Wilamowitz, Verskunst 264, and
P. Maas, Griech. Meirik, $ 48). What is meant by ‘word’ in this connexion
is best described in the words of Maas (op. cit. $ 125, for the details cf. his
subsequent sections): ‘nicht jeder in unserer Schrift abgesetzte Redeteil,
sondern nur das Gesamtbild eines bedeutenden Redeteils (Nomen, Verbum
usw.) zusammen mit den zugehörigen Praepositiva . . . und Postpositiva.’
If a monosyllable stands before the end of a μέτρον or after it, it is only its
syntactical connexion that can decide whether it is more closely associated
with the preceding or the following part of the sentence. Thus on the one
hand it is clear that in lines like Archilochus fr. 62 D. ’Ep£in, πῆι önör’ ἄνολβος
τ Denniston quotes Ag. 1652 in the form ἀλλὰ μὴν κἀγώ, but does not mention, any more
than Sidgwick does, that this is not the reading of the MSS.
^ Wilamowitz has repeatedly (e.g. Gótt. gel. Anz. 1906, 622 = Kl. Schr. v. τ. 387, n. 1,
Platon, ii. 28 τι. 2) expressed the view that the four fragments 170-3, the only source of which
is Alcimus, show the influence of Plato. It seems to me that the style of these fragments also
points so plainly to the fourth century that I regard this date as probable, in spite of the
objections raised by A. Körte (Bursians Jahresberichte, clii, 1911, 231 and RE xi. 1224). For
a discussion of the problem see also Diels-Kranz, Vorsokr. i, 5th ed., 193 ff.
3 It is true that we find in Hephaestion 6. 2 (p. 18. 18 Consbr.) the tetrameter Μητροτίμωι
δηῦτέ με χρὴ τῶι σκότωι δικάζεσθαι, which may perhaps be traced to Hipponax (fr. 72 D.),
but Μητροτέμωι is suspected with good reason by A. D. Knox (see his edition of Herodas
etc. in the Loeb Library, Introduction to Hipponax, p. 4 n. 1).
+ Mazon, however, reads without any qualms in Ag. 1657: στεῖχε δ᾽ ἤδη xoi γέροντες.

787
line 1652 COMMENTARY

ἀθροΐζεται στρατός; and A. Pers. 231 ὦ φίλοι, ποῦ τὰς ᾿Αθήνας φασὶν ἱδρῦσθαι
χθονός; the break comes after the vocative and before the interrogative
adverb, and that in E. Or. 1509 πανταχοῦ ζῆν ἡδὺ μᾶλλον ἢ θανεῖν τοῖς σώφροσιν
the infinitive ζῆν belongs intimately to what follows and not to πανταχοῦ;
there is therefore in such cases no “Wortschluss’, word-ending (taking ‘word’
in the sense defined above), after the first μέτρον (nor is there any, e.g., in the
end of the line E. Jon. 518 εὖ πράξομεν after εὖ). On the other hand, in Ag.
1652 ἀλλὰ κἀγὼ μὴν κτλ. the connexion of ἀλλὰ... μήν is close enough to pre-
vent κἀγὼ before μὴν being felt as a real “Wortschluss’. There is a similar
case in E. Iph. A. 908 ἀλλ᾽ ἐκλήθης γοῦν ταλαίνης παρθένου φίλος πόσις (for
ἀλλὰ... γοῦν cf. Denniston, Particles, 458 f.), except that apparently γοῦν is
more definitely than μήν a ‘postpositivum’.
πρόκωπος : it is only in comparatively recent times that exception has been
taken to this: Herwerden conjectured ξιφήρης, A. 5. F. Gow (C.Q. viii,
1914, 6) ξιφουλκός, which has been adopted by A. Y. Campbell, while G. Thom-
son, following a suggestion of F. E. Adcock, replaces it by πρόχειρος. Against
this I would say that «ai ἐγώ points to a close parallelism in the actions and
utterances of the two adversaries. The repetition of the identical word also
15 characteristic of this dispute (cf. on 1650), helping to express the bitterness
of the clash. The earlier commentators showed far more insight, e.g. Karsten:
‘In priore horum versuum ξίφος πρόκωπον proprie dictum est ; in altero ἐγὼ
πρόκωπος breviter pro ἐγὼ πρόκωπον ξίφος ἔχων. We may suppose that this
bold contraction is made easier by the analogy of zpdyetpos,' which in the sense
of ‘ready’ may be used of a person. If the conjecture put forward in thenoteon
1651 is correct, that the word πρόκωπος was formed entirely on the pattern
of πρόχειρος, then there was a particular reason for the analogous use.
The sentence as written in the MSS must mean: ‘But I too, be assured, do
not object to die with my hand on my sword’ (Paley). But what we really
expect in this context is what, e.g., Wilamowitz and Headlam give: ‘Schlag-
bereit ist auch das meine [i.e. my sword], und ich scheue nicht den Tod’,
“Well, I stand ready too with sword advanced, and do not shrink from death’.
This provides an accurate reply to the challenge of the preceding line, but it
does not render the Greek words as they stand. E. Lobel suggested to me
οὐδ᾽ ἀναίνομαι. I have adopted the substance of his suggestion but should
prefer κοὐκ ἀναίνομαι. My ear recalls Ar. Birds 313 οὑτοσὶ πάλαι πάρειμι κοὐκ
ἀποστατῷ φίλων. It is not unusual for k(ai) to drop out before οὐ, cf., e.g.,
E. Hipp. 326 where it is only the ancient parchment book from Egypt
(Berliner Klasstkertexte, v. 2. 88 ff.) that has KOY, and the Marcianus καὶ o2,
while the other MSS give ov by itself, and A. Prom. 1077 κοὐκ ἐξαίφνης, where
the «(ai) is missing in some of the MSS. If we accept the correction {κοὐκ
in Ag. 1652, it follows that there is an ‘ellipse’ of eut in the first half of the
sentence, but actually this is quite normal in cases where ἐγώ is expressly
mentioned, cf. on 806. With the expression in general cf. E. El. 796 ἑτοῖμοι
! We should, however, beware of quoting here the alleged gloss in Photius’ Lexicon,
πρόκωπος" ἕτοιμος, πρόχειρος, as was done by Dindorf (in the Thesaurus and in his Lex.
Aesch.) and Blaydes. For in the passage of Photius the reading of the MS is πρὸ κώπης,
which was convincingly emended to προνωπής long ago (Cobet was not the first to do so,
as Naber asserts; see Alberti in his Hesychius, on προνόπως, vol, ii, p. 1037 n. 5); cf. Schol.
E. Andr. 729 προνωπής" προπετὴς kai eis τοὔμπροσθεν φερόμενος, Hesychius s.v. προνωπής
(thus Soping for προνόπως) . . . οἱ δὲ προπετής, ἔτοιμος, πρόχειρος.
788
 COMMENTARY line 1653

(without ἐσμέν) κοὐκ ἀπαρνούμεσθ᾽, ἄναξ, The construction of οὐκ ἀναίνομαι

θανεῖν is the same as in A. Suppl. 800 f. (where the content is in essential
points similar) κυσὶν δ᾽ ἔπειθ᾽ ἕλωρα κἀπιχωρίοις ὄρνισι δεῖπνον οὐκ ἀναίνομαι
πέλειν (taken up immediately in the following words by θανεῖν).
1653. δεχομένοις λέγεις θανεῖν σε. It occurred to Platt, J. Phil. xxxü, 1913,
70, that σε might be replaced by σύ. He did not really like his conjecture
himself," but Mazon and G. Thomson thought it good enough to put in their
texts. The objection to oe is in itself unjustified, quite apart from the special
point in this passage, for the use of the pronoun in the accusative is possible
even in cases where there is no idea of antithesis: it is perhaps only a slight
nuance which differentiates this construction from that in which the infinitive
is used by itself, cf., e.g., Pl. Charm. 173 a οἶμαι μέν, ἦν δ᾽ ἐγώ, ληρεῖν pe ὅμως
τό γε προφαινόμενον ἀναγκαῖον σκοπεῖν «rÀ., and in general see Kühner-Gerth,
ii. 32. Here, however, the special point, i.e. the interpretation of Aegisthus’
words as a κληδών (see below), urgently requires that the person whose
identity matters so much here must be clearly indicated : ‘you speak of your
death, you speak of your dying’ (to give a somewhat indefinite rendering for
the moment). The thought was explained with understanding by S. Butler
(quoted in Peile): “Verbum θανεῖν accipiunt ex ore dicentis, omenque acci-
piunt eum revera moriturum esse, quod dixisset οὐκ ἀναίνομαι θανεῖν; quod
genus augurandi ad κληδόνας pertinebat. Cuiusmodi erat dictum illud τὴν
κατὰ σαυτὸν ἔλα iuveni Mitylenaeo de ducenda uxore, a pueris in trivüs
trochos versantibus, oblatum Callim. epigr. x [Wil]. Et illud apud Cic. de
div. 1, 103 L. Paulo a filiola sua dictum, “Persa erit." Tum ile artius
puellam complexus, “Acctpio (dexonau)”, inquit, “mea fia, omen." Erat
autem mortuus catellus eo nomine. Sic Cic. de div. 2. 84 Cauneas interpretatur
pro cave ne eas.’ It is essential for all such κληδόνες ‘that (1) the spoken word
may produce an effect, not indeed irrespective of its meaning, but other than
the meaning or intention of the person who carelessly uttered them ; (2) the
act of acceptance makes them irrevocable, and that in the sense which best
accords with the interest of the person who accepts them. Indeed, it is
almost not too much to say that he forces his own meaning on the omen'
(W. R. Halliday, Greek Divination, 47; examples are to be found, e.g., in
A. S. Pease's commentary on Cic. De div. x. 103 and in E. Riess's article
‘Omen’, RE xviii. 374 ff.). The perversion of the meaning which was intended
by the original speaker is part of the regular procedure for making use of
κληδόνες of this kind. So I venture to put the question, at the risk of seeming
fanciful, whether in this case it 15 simply the word θανεῖν in the utterance of
Aegisthus which affords the chance of conversion into an omen (as most
commentators seem to assume), or. rather the whole expression οὐκ ἀναίνομαι
θανεῖν. For though it is true that the expression as spoken by Aegisthus has
an unmistakable meaning, yet it is in itself ambiguous (cf. on 300). To
illustrate the point it will suffice to quote two passages of the Z/jad which we
find in close proximity to each other: 2 450 ἔνθ᾽ αὐτὸς μὲν ἔπειτ᾽ ἠναίνετο λοιγὸν

1 ‘It does not altogether please me, but I see nothing better to be done. Had I been
writing the line myself I would have said δεχομένοις θανεῖν &Ae£as.! This confession suggests
wonderful possibilities of a verse-composition match of English schools against Aeschylus
of Athens, and it is at the same time so delightful in its self-mockery that I cannot forbear
to quote 1t.

789
line 1653 COMMENTARY

ἀμῦναι, ‘he refused ... to avert (in the future)’, and in contrast Z 500 6 δ᾽
ἀναίνετο μηδὲν ἑλέσθαι, ‘he denied that he had received anything’ (thus cor-
rectly paraphrased in the scholia τοῦ δὲ ἀρνουμένου εἰληφέναι, cf. for the inter-
pretation of this highly disputed passage Wackernagel, Syntax, i. 175;
Ridgeway, The Early Age of Greece, ii. 365; R. J. Bonner and G. Smith, The
Administration of Justice etc. i. 3x ff., and especially! Latte, Hermes, Ixvi,
1931, 129 n. 1; in L-S ἀναίνομαι II, in contrast to the older editions of L-S, we
are given only the wrong translation 'he declined to take anything. I
regard it as not only possible but probable that the coryphaeus, in accord-
ance with the practice of people who want to find a κληδών favourable to
themselves, pretends to understand the words οὐκ ἀναίνομαι θανεῖν in the
sense which is grammatically quite correct 'I do not deny that I have been
killed’. If this is the case, λέγεις θανεῖν σε does not simply mean ‘you speak of
your death (of the fact that you are going to be killed)', which would be all
that is needed, but actually ‘you say you have been killed’.
τὴν τύχην δ᾽ αἱρούμεθα. It should never have been questioned that
Auratus was right in taking EPOYMEOA as αἱρούμεθα. Confusion of a and €
is quite common from the second century after Christ onwards, cf. Meister-
hans, 34. For instances of this confusion in the MSS of Aeschylus cf. on 1657
and, e.g., Prom. 710 véovo’ M, Sept. 292 λεχέων instead of λεχαίων, 474 αιγαρεύς
P, Pers. 817 πλατέων M, 897 εὐκταιάνους M, Ag. 198 karé£evov M!, 297 παιδίον
MV, Cho. 409 πετις, 410 πεπάλατε', 474 αἴρειν instead of ἔριν, Eum. 559 δυσπα-
Acirat,? 645 παῖδας corrected to πέδας M, 656 mpoode£aıre M, Suppl. 107 μενόλιν.
Here αἱρούμεθα is not the exact equivalent of the customary δεχόμεθα (accipi-
mus omen), but parallel to it, something like ‘we decide in favour of it’. In
phrases expressing agreement it is quite usual to intensify the effect by the
addition of parallel words, as has been remarked on 830; in this case, where
the intention 15 to pin down Aegisthus to the evil omen he has just uttered,
the emphatic expression is particularly apposite. For the use of αἱρεῖσθαι
in Aeschylus 'of choosing or closing with a piece of fortune' cf. on 35o. Henr.
Stephanus quoted for comparison Suppl. 380 τύχην ἑλεῖν, ‘durch eine Entschei-
dung das μέλλον τυχεῖν bestimmen’ (Wilamowitz on E. Jon 692). In Ag. 1653,
too, the meaning of τὴν τύχην is, to quote the paraphrase of Wilamowitz (loc.
cit.), τὸ κατὰ τύχην avvrvyóv, or rather 7.«.7. μέλλον συντυχεῖν. We find ἐν
τύχαι in 685 (see note) used in relation to the predestination conveyed by an
ominous name. Translations like ‘we take our good fortune' (Sidgwick)
therefore fail to bring out the special meaning. The sense is rather: ‘what,
! Latte's observations are completely ignored by Pflüger, Hermes, Ixxvii, 1942, 140 ff.,
whose treatment of the linguistic evidence is prolix and at the same time superficial,
whereas Latte's brief remarks are discriminating and pertinent. A correct translation of
the passage is given by Bonner and Smith, loc. cit., and by G. M. Calhoun, Introduction to
Greek Legal Science (1944), 76.
2 The generally received text (ev) μέσαι δυσπαλεῖ τε δίναι is quite correct: there is no
reason why we should introduce an unknown word δυσπαλεῖν here instead of the known
δυσπαλής, The objection to re does not hold good, either in the crude form in which it was
expressed by Bothe (‘neque enim copulari solent epitheta ornantia’) and Paley (mere
epithets are not properly coupled by re’) (cf. to the contrary, e.g., A. Pers. 852, S. Ant. 960,
Oed. R. 561), or in Denniston’s more careful remark, Particles, 501, where he regards ‘the
presumed coupling of two disparate qualitative epithets by τε in A. Eum. 559' as question-
able. I cannot agree that in this case μέσαι and δυσπαλεῖ are ‘disparate’: the whirlpool is at
its most violent in the middle, and that is where the struggle is most hopeless for the sailor.

790
 COMMENTARY line 1655

in accordance with your utterance, is bound to happen’. So the gloss τὸν

οἰωνὸν δηλονότι, which in Tr is written above τὴν τύχην, is substantially
correct, and Hermann was wrong in his assertion that the gloss does not
refer to τὴν τύχην but to δεχομένοις.
1655. Schütz's brilliant emendation O€POC for DEPWC! has been univer-
sally accepted ; he himself referred to Pers. 822 ὅθεν πάγκλαυτον ἐξαμᾶι θέρος,
while others have cited Eur. fr. 419. 4 N., Bacch. 1315. But opinions are
divided about the construction of the sentence and consequently about its
exact meaning. Schütz translates: ‘Sed iam haec (mala) messuisse valde
infausta messis est’; he evidently connects πολλά (adverbial) with övornvov,?
as do Klausen, Dindorf (Lex. Aesch. 298, right hand column), Wilamowitz,
who notes this expressly in his app. crit., and, in order to avoid τάδε...
πολλά being taken together, substitutes τόδε for τάδε. This interpretation of
πολλά as valde is in this context, though not impossible, yet highly improbable.
For it is desirable to take πολλά here in the quantitative sense as referring to
the harvest of evil, and this for two reasons. In the first place, when Clytem-
nestra says τάδε (cf. her exactly similar τάδε in 1570) she is probably not
thinking only of the murder of Agamemnon, but also of κακά in the remoter
past, the harvest of which is to be reaped by those who have now been left
as masters in the house. Cassandra had already described the house as πολλὰ
συνίστορα.... κακά (1090 f.). Later in Cho. 932 f. the fate of Orestes is described
thus: ἐπεὶ δὲ πολλῶν αἱμάτων ἐπήκρισεν τλήμων ᾿Ορέστης κτλ. It appears that
πολλά has a particular significance in this connexion. Moreover there seems
to be a point of sinister meaning in the fact that in this case it is precisely the
rich harvest that turns out to be the δύστηνον θέρος, while elsewhere the
harvest-yield—often described as moAv—is a longed-for gift of the gods:
Sept. 360 f. πολλὰ... γᾶς δόσις, Ag. 1014 ff. πολλά τοι δόσις ἐκ Διὸς ἀμφιλαφής τε
καὶ ἐξ ἀλόκων ἐπετειᾶν νῆστιν ἤλασεν νόσον. If therefore we abandon the
adverbial interpretation of πολλά, it is perhaps possible to take it, with a
number of commentators, to mean: ‘Nein, auch dieses viele einzuheimsen ist
eine unglückliche Ernte’ (Wecklein). This, however, seems rather lame,
particularly in the way in which it connects with the preceding words.
Hermann felt quite rightly that the context suggests something like ‘sed haec
quoque satis multa sunt’. But his interpretation ‘ ἀλλὰ καὶ τάδε πολλά ἐστιν
ἐξαμῆσαι δύστηνον θέρος, sed haec quoque satis multa sunt, ut inde tristem
messem metamus’, which has met with considerable approval, is rather dis-
credited by the manner in which the order of words is twisted. The most
attractive solution seems to be that of Weil, who (1858) put a comma after
πολλά and took πολλά = πολλά ἐστι (as Hermann did), so that ἐξαμῆσαι is
attached to πολλά as explanatory or illustrative infinitive: ‘but even these
are many to reap, a harvest of misfortune’.? Weil was followed by Sidgwick,
Verrall, Headlam (verse translation; in his prose translation the line is
missing). However I am not fully satisfied with this, for I find no parallel for
this kind of predicative use of πολλά, nor am I quite sure that the epexegetic
1 O for @ is found also, e.g., Suppl. 3 (ἀρόεντ᾽ instead of ἀρθέντἢ).
2 Hermann’s criticism that Schütz and his followers had misunderstood the sentence
‘prorsus neglecto πολλά ' is therefore inapplicable to Schütz.
3 Enger (in his re-edition of Klausen’s commentary, 1863) raises the objection that the
phrase ἐξαμῆσαι θέρος should not be interfered with. This does not strike me as cogent,
for even if θέρος follows as apposition, its intimate connexion with ἐξαμῆσαι is felt.
791
line 1655 COMMENTARY

addition of ἐξαμῆσαι to πολλά is admissible. Thus there remains an element

of uncertainty. [See the Addenda.]
1656. We may regard as certain the slight alterations ὑπάρχει (Scaliger) and
αἱματώμεθα (Jacob). The way in which Wilamowitz has chopped the sentence
up is intolerable ; moreover he does not tell us what he really supposes to be
the meaning of ὕπαρχε undev! (nor does he make good this omission in
Interpr. 202 n. 1 : in his translation he has “brauche nicht Gewalt’). As regards
μηδὲν αἱματώμεθα, we should beware of following Auratus (as A. Y. Campbell
does) in substituting for it the apparently obvious μηκέθ᾽ αἷμ. For if she said
that, ‘Clytemnestra would be doing something not quite in keeping with the
conciliatory purpose of her words, making a direct reference to the murder
of Agamemnon. With μηδὲν αἱματώμεθα this is avoided’ (Hense).
The only doubtful point in this line is whether the particles can be kept as
they stand in 8’ ἅλις y’. Hermann, without giving a reason, wrote πημονῆς
ἅλις δ᾽ ὑπάρχει. Αἱ the root of this conjecture there lies an observation
which is well worth noticing. ἅλις by itself, not strengthened by ye, is used
in a number of common colloquial phrases, e.g. S. Oed. C. 1016 ἅλις λόγων,
A. Eum. 675 ἅλις λελεγμένων, Hdt. 9. 27. 5 παλαιῶν μέν vvv ἔργων ἅλις ἔστω,
E. Suppl. 1148 f. ἅλις τύχας, ἅλις δ᾽ ἀλγέων éuoi πάρεστι (this instance also
shows that a verb such as πάρεστι or, as in Ag. 1656, ὑπάρχει, may be added to
the expression, which is more commonly used without a verb), Pl. Politicus
287 a καὶ τούτων μὲν ἅλις, and so on; for the meaning of ἅλις in these and
similar phrases cf. on 1659. The information provided by R. A. Neil, The
Knights of Aristophanes, p. 191, and with still more plentiful illustration by
Denniston, Particles, 155 (unfortunately without mentioning Ag. 1656)
about the rarity of δέ ye or δὲ... ye in Tragedy—except in cases where there
is a change of speaker or where an antithesis is intended—is not calculated
to inspire much confidence in the 8' . . γ᾽ of the present passage.? Hermann’s
πημονῆς ἅλις δέ would be quite in keeping with Aeschylus’ tendency to relegate
δέ to the third place in the sentence (cf. on 653) ; and it can easily be under-
stood that a copyist may have objected to this order and transferred δέ to
the second place, so that y’ had to be added from metrical necessity. But it is
possible that a weakened force of ye may be admissible here: so I retain the
MS reading though with some hesitation.
πημονῆς: cf. on 1375.
1657. στείχετ᾽ αἰδοῖοι is practically preserved in the MSS: only an o has
dropped out by haplography (for e instead of a cf. on 1653). For αἰδοῖος
cf. on 600. The apparently unique? form of address αἰδοῖοι γέροντες brings out
very clearly the altered attitude of Clytemnestra.
1 Mazon, who in this passage follows Wilamowitz, translates ὕπαρχε under: 'n'entre-
prends plus rien’. I cannot see how this rendering, particularly the ‘plus’, can be justified
by the ordinary use of ὑπάρχειν.
2 In the single passage of Aeschylus quoted by Denniston in this section, Sept. 283 ἐγὼ
δέ γ᾽ (a certain emendation of δ᾽ em’) ἄνδρας &£ . . . τάξω, the observation of Kaibel on 5.
El. 1367 holds good : in the phrase ἐγὼ δέ ye the ye is similar to that in ἔγωγε.
3 In Epic αἰδοῖος is used as a form of address only in the regular phrase (p 152 and 4 times
in τὴ ὦ γύναι αἰδοίη Δαερτιάδεω ' Obvafjos. Thilde Wendel, ‘Die Gesprächsanrede’, Tübinger
Beitr. z. Altertumsw., 6. Heft, 1929, omits Ag. 1657 from her list p. 31, 2 y; on p. 84 she
asserts that 'the first time yépovres is found as a form of address to the Chorus is in Euri-
pides'. Since she says on p. 5 n. 2 that her quotations are based on the text of Wilamowitz,
she ought at least to have discussed this passage if she adopted a different reading.
792
 COMMENTARY line 1657

Editors are agreed that τούσδε at the end of the line must be deleted. It is
probably a gloss on δόμους and rests of course on a misunderstanding of the
context. πεπρωμένους is more of a problem. One would have wished that
Wilamowitz and others who adopted this reading had let us know what they
thought it meant. ‘It must be admitted that πεπρωμένους is suspicious’
(Sidgwick) puts the matter very mildly; I myself agree with those scholars
who use stronger language, e.g. Karsten: ' δόμους quos quis habitet verpw-
μένους, fato destinatos, dici ridiculum est’, and Headlam: ' zp. 8. πεπρ., “to
your predestined houses” is absurd.’ Mazon translates literally ‘chacun dans
la demeure que le Destin lui donne'. Tt is true, in the speech of modern people,
who rather glibly speak of 'Fate' or 'Destiny', an attribute equivalent
to πεπρωμένους does not sound too bad; but the expression would be felt as
absolutely incongruous by a Greek like Aeschylus, who (as the instances of
its use suffice to show) invariably adheres to the full force of the word
πέπρωται Or πεπρωμένον. But the objection is not merely to this particular
epithet : we should rather say that any epithet here would be not only super-
fluous, but probably quite out of place. The idea we expect in this context is
not ‘go to such and such a kind of house’, but ‘go home’. 'Home' in tragic
diction is expressed by πρὸς (or és) δόμους (ἰέναι etc.), e.g. A. Pers. 1038 πρὸς
δόμους δ᾽ ἴθι (request to the Chorus by way of preparation for their exit at the
end of the play, exactly as in Ag. 1657), Eum. 764 νῦν ἄπειμι πρὸς δόμους (from
Athens home to Argos), S. Ant. 1087 (Tiresias) ὦ παῖ, σὺ δ᾽ ἡμᾶς ἄπαγε πρὸς
δόμους, Aj. 509 ζῶντα πρὸς δόμους μολεῖν, Trach. 34 f. τοιοῦτος αἰὼν ἐς δόμους
τε κἀκ δόμων ἀεὶ τὸν ἄνδρ᾽ ἔπεμπε, Phil. 517 f. πορεύσαιμ᾽ ἂν ἐς δόμους, Oed. C.
784 οὐχ ἵν᾽ ἐς δόμους dynes, 1408 f. éàv . . τις ὑμὶν ἐς δόμους νόστος γένηται. This
usage would be paralleled in our passage by oreiyere . . . πρὸς δόμους without
any additional word. This seems to have been recognized as desirable by
Heath: he put a comma after δόμους, and substituted for πεπρωμένους a
conjecture (τετρωμένους ---- which need not be discussed). There have been
unsuccessful attempts to retain the stem of πεπρωμένους and only to alter its
termination. The first to resort to this expedient was Bothe, who read
πρὸς δόμους, πεπρωμένας (πημονάς Vel ποινάς) πρὶν κτλ. Next Ahrens (quoted
in the edition of Franz) proposed πρὸς δόμους" πεπρωμένως κτλ., Orelli πεπρω-
μένον, Madvig πεπρωμένοις, πρὶν παθεῖν, εἴξαντες. Madvig’s conjecture was
put into the text by Headlam, A. Y. Campbell, and G. Thomson. Housman
was more cautious (J. Phil. xvi, 1888, 288). Although he praises Madvig’s
suggestion far more than it deserves, he does not adopt it, and emphasizes
that we should hesitate about ‘parting with the familiar juxtaposition of
παθεῖν and épfac’. However we deal with the corruptions before and after
πρὶν παθεῖν ἔρξαντες, we must not tamper with this expression itself (cf. on
1529). Hermann is excellent on this point. Presumably πεπρωμένους has
taken the place of something entirely different. In all probability there was
here an expression in the text that was neither an attribute of δόμους nor a
component part of the entirely separate clause πρὶν παθεῖν ἔρξαντες. What it
was I do not know. Only by way of indicating a possible line of thought, and
not as a conjecture, I offer for consideration something like πρὸς δόμους"
κρεῖσσον φρονεῖν πρὶν παθεῖν ἔρξαντες, or a participle in the nominative, e.g.
1 Lawson’s wildly fanciful Πεπρωμένης, πρὶν παθεῖν, στέρξαντες οὖρον is not likely to
recommend itself to anyone.

793
line 1657 COMMENTARY

πρὸς δόμους, φρονοῦντες εὖ, πρὶν κτλ. How πεπρωμένους got into the text is not
clear. I cannot help suspecting that it was a marginal variant of 1663 πειρω-
μένους (which is of course what Aeschylus wrote). κἀκβαλεῖν ἔπη τοιαῦτα
δαίμονας (this is the MS reading) πεπρωμένους would yield very poor sense,
but still some sort of sense: ‘and to utter such words against the fate that is
ordained’ ; in this case the periphrastic expression ἐκβαλεῖν ἔπη τοιαῦτα would be
constructed like κακῶς λέγειν, and δαίμονας πεπρωμένους would have to be taken
in the same way as τὴν πεπρωμένην αἶσαν (Prom. τοῦ f.) and the like. It goes
without saying that surmises of this kind are bound to be extremely uncertain.
1658. With the reading ἔρξαντες we are not on very firm ground, since the
agreement of GTr seems to point to &p£avra as the reading of the hyparche-
type, although it is not impossible that the hyparchetype exhibited a variant
here as elsewhere (see vol. i, p.29f.). In any case ép£avres recommends itself
(as Hermann observes) because the subject is the same as that of the pre-
ceding verb areiyere. Moreover, the expression which originally came after
δόμους in 1657, whether it was something like κρεῖσσον φρονεῖν or a participle in
the nominative or whatever it may have been, must have referred to yepovres
either as its grammatical or its ‘logical’ subject. Consequently, what we
require is ἔρξαντες and not ép£avras! (not to mention ἔρξαντα). Hermann is
right in saying that any words added to πρὶν παθεῖν ἔρξαντες would only spoil
this expression. We shall arrive at the same conclusion, viz. that it is neces-
sary to take épéavres as the end of the clause, when we examine the remainder
of the line. Several editors, it is true, both of earlier and of more recent times,
take χρῆν τάδ᾽ ὡς ἐπράξαμεν as an independent clause by itself, Conington’s
comment being ‘it is equivalent to ἐπράξαμεν τάδε ὡς χρῆν ', while Wecklein
translates ‘dies musste stattfinden, wie wir es abgemacht haben’, and Headlam
‘it was fated we should act herein as we have acted’. But I have the gravest
doubts whether χρῆν τάδ᾽ ὡς ἐπράξαμεν with nothing added to it is intelligible
Greek. Nor have I been able to find any comparable construction in the
dictionaries s.v. χρή. It is difficult to see what help is to be got from Head-
lam’s parallels, Ar. Ach. 540 and E. Her. 311. The upshot, then, is this: πρὶν
παθεῖν ἔρξαντες is a complete expression, χρῆν τάδ᾽ ws ἐπράξαμεν is incomplete ;
it is therefore most likely that the meaningless καιρόν, the initial consonant of
which upsets the metre, conceals an infinitive which would provide a normal
construction for the χρή clause which follows. We can hardly hope for a
certain restoration where the text is in as bad a condition as it is in these
lines. But the sense which one would expect can be obtained with αἰνεῖν,
read by Heath (though he does assume a different construction for the
sentence) and independently of him by Lobeck, Pathologiae Proleg. 248. For
instances of αἰνεῖν = ‘to be content with, acquiesce in’ see L-S s.v. II. 2 (cf.
also above on 98). The sentence αἰνεῖν χρὴ τάδ᾽ ὡς ἐπράξαμεν, ‘you and 1
ought (instead of letting it come to παθεῖν ἔρξαντας) to content ourselves with
things as they have turned out for us’, has a close parallel in Ewm. 469 πράξας
yap ἐν σοὶ πανταχῆι τάδ᾽ αἰνέσω, ‘wie ich auch immer durch dich fahre, ich
werde damit zufrieden sein’ (Wecklein) ‘however I fare because of you, I shall
be content’ :? the similarity goes so far that in both cases the phrase express-
1 I cannot understand why Wilamowitz writes ἔρξαντας, in spite of his quite correct
translation ‘eh’ ihr Ubel tut und leidet’. uu
? Verrall's translation 'before thee, however I fare, I shall be content' (similarly Head-

794
 COMMENTARY line 1659

ing the experience (ὡς ἐπράξαμεν and mpd£as . . . πανταχῆ) has the demon-
strative τάδε added to it by way of reinforcing the object. The euphemistic
passing over the worst in ὡς ἐπράξαμεν is analogous to that found in the
passages discussed on 1171 ; in the sense of Clytemnestra’s present mood there
has been a κακῶς πρᾶξαι, and the whole of this speech leaves no doubt about it.
1659. Hermann’s' emendation δεχοίμεθ᾽ has been almost universally? ac-
cepted. And the more recent editors, at least, are unanimous about the
necessity of putting a comma before δεχοίμεθ᾽ ἄν (not, as has often been done,
after γένοιτο). It is, however, the word ἅλις that has given rise to serious
objections. Many before and since Housman have had the feeling which he
expresses in J. Phil. xvi, 1888, 289: ‘To say εἰ μόχθων γένοιτο ἅλις the moment
after you have said πημονῆς ἅλις ὑπάρχει is so obviously inconsistent that
there is a general consent [this was an exaggeration at the time when Hous-
man wrote and would be one to-day] against the genuineness of ἅλις. But
quite apart from the fact of its being preceded by 1656, the phrase e...
μόχθων γένοιτο τῶνδ᾽ ἅλις is not quite intelligible on the face of it. It is
dangerous to translate here in a way which lessens or glosses over the sense,
as, e.g., Lewis Campbell does ‘If only this might prove enough of misery’ and
Wilamowitz ‘Wenn es nur damit genug unsrer Mühn und Sorgen wire’; both
of these make μόχθων ἅλις a predicate of τῶνδε in a way for which there is no
indication in the Greek, instead of taking τῶνδε μόχθων together. Such
attempts as I have seen (and they are many) to substitute some other word
for ἅλις result in spoiling the sense. What we expect here (cf. the entirely
parallel utterance of Clytemnestra in 1570 ff.) is something like: ‘if no further
sufferings followed on these, we would accept it’. And this is precisely what
we have in the text as it stands, expressed like this: ‘if "enough of these
sufferings"? were brought about ...’. It is fairly common to find phrases like
τούτων (or some such word) ἅλις — 'enough (and more than enough) of that,
no more of that',* ἅλις λόγων (S. Oed. C. 1016) = 'now no more words’, etc.
lam’s: ‘before thee, whatever may be my fate, I will accept it’) wrenches ἐν coi out of the
setting where it clearly belongs on the evidence of the word-order. For this type of mistake
(breaking up the unity of the participial kolon) cf. above on 1127 and 1511.
τ It is of no consequence, but if we attach any value at all to priority, the credit for the
emendation should be given to Hermann, not to Martin, to whom it 15 attributed by
Wecklein, Sidgwick, and several other editors. Fr. Martin's Observ. crit. in Aesch. Orest.
appeared in 1837. Hermann says he communicated his emendation to Humboldt for his
translation (published in 1816), and Humboldt translates: ‘wollen wir...sie nehmen’.
? Not of course by Verrall. |
3 By paraphrasing the words like this and putting them between inverted commas, I
have slightly over-emphasized the point: what is meant will become clear from my subse-
quent remarks.
+ The usual rendering of ἅλις by ‘sufficiently’, ‘enough’ is not always adequate. By a
natural transition ἅλις often comes to mean ‘enough and more than enough’, ‘more than
can be borne’, “too much of it’. Especially the exclamatory expression ἅλις or ἅλις τούτων
and the like may become almost equivalent to ‘no more of it’, but the intensified meaning
is not confined to this usage. The same is true of satis. The idiom is not always sufhiciently
appreciated. On Horace, Odes, 1. 2. 1 Jam satis terris nivis etc. R. Heinze remarks: ‘iam
satis : for the object is by now attained, the wrath of the god is clearly recognized by every-
one’. What the phrase does express is not satisfaction with the fulfilment of the divine will,
but something like a groan, an utterance of pain caused by the unbearable sequence of cata-
strophesand dire omens. tam satis at the beginning of Horace’s ode and satis am in the kindred
passage of Virgil (Georg. 1. sor) is but a more dignified version of the impatient ejaculation
ohe, tam satis (Plaut. Cas. 249, Stich. 734, Hor. Sat. 1. 5. 12 f.): ‘for Heaven's sake, stop it!’

795
line 1659 COMMENTARY
These phrases thus emphatically indicate the urgent desire to bring the fore-
going action or conversation to a definite end, just as ‘basta’ and ‘assez’ in
meaning and tone of voice often amount to ‘But now stop!’ It is in keeping
with this usage that the possibility of such a final stop being made is indi-
cated by εἰ τούτων ἅλις γένοιτο, for which the colloquial forms of expression
which we have quoted provide the starting-point.! The reader who finds it
hard to accept this explanation will have to assume that the text is corrupt.
I admit the possibility of a corruption, but do not think it likely.
For ei τοι and ei δέ τοι cf. Denniston, Particles, 546.
1660. ‘The χηλή of the daimon in A. Ag. 1660 belongs, it is true, to the
horse... But that does not prove that the daimon, who has no definite
individuality, is thought of in the shape of a horse, but only that the poet,
familiar with the horse-shaped beings of the underworld, assigns him a hoof
—a more violent expression than that in Pers. 515 f. δαῖμον ὡς ἄγαν βαρὺς
ποδοῖν ἐνήλου᾽2 (Wilamowitz, Glaube d. Hell. i. 152 n. 3; perhaps he draws too
definite a line between this and Eum. 779 καθιππάσασθε, which he regards as
‘almost purely metaphorical’). Cf. on 1175.
1661. ὧδ᾽ ἔχει λόγος : a formula of conclusion as in Sept. 225, Cho. 521; cf. on
582. The additional word γυναικός is full of meaning, and makes the con-
cluding phrase here echo the one which closes Clytemnestra's speech in 348.
In that passage γυναικός was the expression of a proud claim, of protest
against generally prevailing prejudice (and it was strongly stressed by the
coryphaeus in 35r in the sense intended by the queen); here the feeling is
quite different: ‘I, the woman, am trying, while the men quarrel, to prevent
bloodshed and further diaster.' She is still full of royal dignity, but now her
words show plainly how deeply she has been smitten by the blow of the
daimon: this js in harmony with many of her utterances in the second part
of the preceding kommos.
1662f. For the exclamatory infinitive cf. besides Goodwin, $ 787, and
Kühner-Gerth, ii. 23, P. T. Stevens, C. Q. xxxi, 1937, 187, who says: “This use
of the infinitive without the article appears first in A. Zum. 837 (lyr.), Ag.
1662. It is generally used to express indignation, and is perhaps emotional
and dramatic rather than purely colloquial.’ It is worth noticing that in the
five examples which are quoted in the grammars for this use of the exclama-
tory infinitive without the article,? there is always a pronoun or pronoun-
adjective at the head of the sentence: Ag. 1662, Eum. 837 ἐμὲ παθεῖν τάδε
(the deictic τάδε is similar to the ὧδε of Ag. 1662), S. Aj. 410 f. τοιάδ᾽
ἄνδρα χρήσιμον φωνεῖν, Ar. Wasps 835 τοιουτονὶ τρέφειν κύνα, Dem. 21. 209
τοῦτον δ᾽ ὑβρίζειν, ἀναπνεῖν δέ (add [Dem.] 25. 9x τοῦτον δὲ ταῦτα ποιεῖν). This
corresponds exactly to the group of Latin exclamatory infinitives which
shows -ne attached to a pronoun at the head of the clause (cf. 1. B. Hofmann,
Lat. Umgangssprache, 50; C. Giarratano on Horace, Epodes 8. 1), as, e.g., Plaut.
Pseud. 202 huncine hic hominem pati colere iuventutem Atticam, Ter. Eun. 644
1 Peile seems to have taken a similar view of the passage, for he gives this paraphrase:
‘but assuredly if there might be found a point at which to cry “Hold! Enough of these troubles”,
we would . . .'.
2 This passage had been cited by Hermann on Ag. 1660, but he did not avail himself of
the parallel because he was misinformed about the reading of F (χηλῇ).
3 Stevens, loc. cit., wrongly includes Soph. Ichn. (fr. 314 P.) 74, which is an instance of
the infinitive common in prayers: Pearson ad loc. is right.

796
 COMMENTARY line 1664
hoccin tam audax facinus facere esse ausum, Cic. S. Rosc. 95 tene . . . potissimum
libi partis istas depoposcisse . . ., Hor. Sat. 1. 9. 72 f. huncine solem tam (here
we find in tam a second demonstrative element as in Ter. Eun. 644 and in both
the passages of the Oresteia—apparently a very ancient mode of expression)
nigrum surrexe mihi, 2. 4. 83 ten lapides . . . radere, 2. 8. 67 lene . . . torquerier,
Virgil, Aen. 1. 37 mene incepto desistere victam and the like.
It seems difficult to make out the meaning of ἀπανθίσαι. Perhaps this is
what Aeschylus wrote, but I have not found a satisfactory explanation of it.
At the end of his not very lucid discussion Schütz gives the rendering linguae
florem decerpere, similarly Blomfield and Hermann, who says ‘idem fere ac
δρέψασθαι ' (his note shows that to some extent he doubts the authenticity of
ἀπανθίσαι) ; L-S ‘cull the flowers of idle talk’, and so, e.g., Sidgwick, with the
remark ‘a characteristic Aeschylean figure’. If the verb here is to mean
‘pluck’, it is difficult to understand ἐμοί. Others give the verb a different
meaning: thus Butler (quoted in Peile) compares Cicero's (de orat. τ. 20) ex
rerum cognitione efflorescat et redundet oportet oratio and the English ‘flowers
of speech'; similarly Verrall: ‘make a foolish tongue break out in bloom’,
which is in itself quite improbable, and as far as ἐμοΐ is concerned, is no less
harsh than the meaning given by Schütz. The dative is taken into account
by van Heusde's interpretation, which otherwise seems fanciful: 'sunt λόγοι
μάταιοι, quos sibi chorum tamquam corollam obtulisse ac vovisse Aeg.
queritur. The word itself does not provide any clue to its meaning, for we
have no other example of it which can safely be said to be earlier than the
period of the empire! (the supplement in Aesch. fr. 100 N. is quite uncertain) ;
and from the fact that in later times it usually means 'pluck off' and the like
it does not necessarily follow that it had the same meaning in the fifth century.
For the multifarious use made of the suffix -/£ew see the notes on 286 and on
96o. None of the conjectures offered (Wakefield's ἀκοντίσαι, which was called
by Hermann ‘non inepta coniectura’, Keck's ἀκανθέσαι etc.) is convincing.
So we are thrown back upon guessing.
1663. The emendation δαίμονος has been generally accepted; cf. Cho. 513
δαίμονος πειρώμενος. For the meaning of δαίμων here cf. on 1341 f.
1664. I do not venture to hope for a certain restoration of this line, which has
a corruption in the middle and a lacuna at the end of it; the following
observations are merely meant to point the way to a possibility. lt is gener-
ally admitted that in ἁμαρτῆτον (the writer of G saw that this was unintelli-
gible and therefore left a blank between γνώμης and κρατοῦντα) there must be
some form of ἁμαρτεῖν concealed, for on no account can there be any question
of ἁμαρτῆι or ὁμαρτῆι here. This is corroborated by the parallel expressions
cited by Blaydes, Hdt. 3. 81. 1 γνώμης τῆς ἀρίστης ἡμάρτηκε, 9. 79. 1 γνώμης...
ἡμάρτηκας χρηστῆς (cf. also Thuc. 1. 33. 3 γνώμης ἁμαρτάνει, 6. 92. I οὐχ
ἁμαρτήσεσθαι οἶμαι γνώμης). For the underlying idea cf. on 175. Whether
ἁμαρτεῖν (Casaubon), which most editors have adopted, or some other form is
! The unknown writer Πωλίων, whose words davÜ(Lew ἐπεχείρει τοὺς Φρύγας Axıddeis are
quoted in Photius Berol. p. 156. 1x (the source being probably Phrynichus) is regarded by
Wilamowitz, Berl. Sitzgsb. 1907, 12, as a contemporary of Phrynichus, ‘whichever of the
many Pollios he may have been'. As for the citation of the active ἀπανθίσαι in Phrynichus,
Praep. soph. p. 9. 5 de B., it is of course possible that it refers to A. Ag. 1662, as de Borries
(who makes a supple'aent in the passage accordingly) assumes, but it is by no means
certain. [Wilamowitz, Ki. Schr. ii. 88 n. 3: ‘Der Name Polio ist vulgár'.]

797
line 1664 COMMENTARY
probable, can only be decided when we have arrived at a clear idea of the
purpose which the line as a whole fulfils. And not till then can we profitably
consider possibilities for supplements at the end of the line.
Hermann transposed the line to follow 1655. His rearrangement of the group
is: 1654, 1656, 1655, 1664,1657. Thisisentirely arbitrary (even with Hermann’s
regrouping of the three preceding lines 1664 does not continue their sense
satisfactorily, nor can it be said that this intruded line improves the transi-
tion to 1657), so that the more recent editors were justified in not adopting
these alterations. We may, however, admire Hermann’s sound instinct,
which prompted his feeling that 1664 was ‘alienissima in sede positum’. If
1664 is joined on by its ἁμαρτεῖν to the infinitive clauses in 1662, 1663, then,
whatever supplement is suggested for the end of 1664,! the result is that
σώφρονος γνώμης κτλ, seems lame and feeble after the forceful κἀκβαλεῖν
ἔπη τοιαῦτα. Moreover, it is difficult to justify the δ᾽. Stanley proposed to
change it to θ᾽, and he has been followed by editors who place the line else-
where, e.g. Wecklein (annotated edition of 1888) and Wilamowitz. Now the
position of re in the phrase σώφρονος γνώμης τε 15 extremely suspicious, and
should certainly not be ‘restored’ by way of conjecture, cf. the warning given
in the note on 229. The principal objection, however, lies in the poverty of
the thought of 1664 if it is meant to be a continuation of the exclamations
contained in 1662, 1663. One can well understand therefore that numerous
attempts have been made after Hermann to find another place for 1664
(Gilbert put it after 1661, Wecklein after 1662, Wilamowitz after 1649).
What was said above on 1649 about the infrequency of transposition of lines
in this play need not prevent us from assuming that it has happened here if
there were cogent reasons for it. However, that does not seem to be the case.
Weak though 1664 is if it is merely a continuation of 1662 f., it may serve
quite well as an introduction to a fresh bout of abusive language. Besides, I
think that the reply of the coryphaeus in 1665 would have an additional
point if his words were the answer to a direct attack on the part of Aegisthus.
I therefore suggest tentatively that after his exclamations (1662 f.) addressed
to Clytemnestra, Aegisthus turns angrily back to the Chorus once more with
the words σώφρονος γνώμης [δ᾽ ἁμαρτὼν τὸν κρατοῦντα (λοιδορεῖς». In the
supplement of the verb at the end my suggestion coincides with that of
Vossius, as I learn from Wecklein's Appendix. I cannot do more than indicate
a possibility of this kind.
1666. For ἔτι in threats cf. on 1429.
1667. Headlam's addition of γ᾽ restores a nice point of idiom. G. Thomson
compares 1249 οὔκ, εἴπερ ἔσται ye, Sept. 1048 (by the diaskeuast) où, πρίν γε
χώραν τήνδε κινδύνωι βαλεῖν, S. Oed. R. 583 οὔκ, εἰ διδοίης ye . . ., Phil. 109 οὔκ,
εἰ τὸ σωθῆναί ye τὸ ψεῦδος φέρει (all in stichomythia as here), Pl. Phaed. 89 b
ἔοικεν, ἦν δ᾽ ἐγώ, d Σώκρατες.---οὔκ, ἄν γε ἐμοὶ πείθηι.
ἀπευθύνηι: probably present, cf. p. 670, n. 1.
1668. Both Stanley and Schütz saw that Aegisthus is derisively using a

1 Blomfield’s supplement θ᾽ ὑβρίσαι, which many have adopted, leads to the undesirable
result that there are then two co-ordinated infinitives in this line, whereas in each of the
two preceding lines there is only one. This will be avoided if with Headlam we introduce a
participle at the end of the line, but his ἀρνουμένους is not commended by the way in which
Aeschylus uses ἀρνεῖσθαι elsewhere.

798
 COMMENTARY line 1670

proverbial expression. old’ ἐγώ (‘it’s a familiar story’, cf., e.g., S. El. 837 ff.
οἶδα γὰρ ἄνακτ᾽ ᾿Αμφιάρεων κτλ.) accentuates the scornful tone. Paley and
others understand ‘I know of my own experience’. But tlie hopes which
Aegisthus nourished while an exile have come true: so it is not likely that he
should recall them in connexion with the hope nourished by Orestes. The
proverbial nature of the phrase φεύγοντας... ἐλπίδας σιτουμένους is expressly
(ὡς λόγος) brought out in E. Phoen. 396 ai δ᾽ ἐλπίδες βόσκουσι φυγάδας, ὡς
λόγος, where the comment of the scholia is: ἐντεῦθεν ἡ παροιμία" ᾿ αἱ δ᾽ ἐλπίδες
βόσκουσι τοὺς κενοὺς βροτῶν ᾿. Cf. (Peile, van Heusde, Blaydes, Headlam)
S. Ant. 1246 ἐλπίσιν δὲ βόσκομαι, Soph. fr. 862 N. (= 948 P.) ἐλπὶς γὰρ ἡ
βόσκουσα τοὺς πολλοὺς βροτῶν, E. Bacch. 617 ἐλπίσιν δ᾽ ἐβόσκετο, Eur, fr. 826
δι᾿ ἐλπίδος ζῆ καὶ δι᾽ ἐλπίδος τρέφου, Antiphanes fr. 123 K. ἀεὶ πεινῶμεν ἐπὶ ταῖς
ἐλπίσιν, Eubulus fr. 1o. 6 f. K. οὗ paar’ ἀεὶ πεινῶσι Κεκροπιδῶν κόροι κάπ-
rovres αὖρας, ἐλπίδας σιτούμενοι. The commentators do not mention the
earliest instance of all, Semonides of Amorgos fr. x. 6 f. ἐλπὶς δὲ πάντας kamı-
πειθείη τρέφει ἄπρηκτον öpuaivovras. For Aegisthus' excessive use of proverbial
expressions cf. on 1629 ff.
οἶδ᾽ ἐγώ : for the addition of ἐγώ in this and similar phrases cf. F. Sommer
in ANTI42PON, Festschrift für ]. Wackernagel (1924) 26. This usage appar-
ently belongs to colloquial language, as may be seen from the many instances
in comedy, e.g. Ar. Ach. 309, Knights 314, Clouds 660, 683, 1100, Wasps 310,
Thesm. 769. οἶδ᾽ ἐγώ (also A. Suppl. 740) and ἐγώιδα are common enough in
Sophocles and Euripides.
1669. πρᾶσσε has been suspected (Hense: ' πρᾶσσε allows of no satisfactory
connexion with the preceding words ... πρᾶσσε therefore seems to be a false
reading of the MSS’; two horrible conjectures are quoted in Wecklein’s
Addenda [1893]). The verb does indeed require some explanation. The right
comment is provided by Wilamowitz on E. Her. 323 ff.: ‘ πρᾶσσε, especially
as a concluding clause and at the beginning of a line, is much more than
“do it’’, it is τελείωσον, διάπραξον.᾿ For this sense of πράσσειν and the differ-
ence between it and δρᾶν cf. on 1353.
miatvou: the abusive word is here applied to the material enjoyment of riches
(cf. 1638) as well as to the lack of sober judgement born of surfeit : cf. on 276.
μιαίνων τὴν δίκην, δίκη, if anything, should belong to the ἀμίαντα. But
Aegisthus is one of those ὅσοις ἀθίκτων χάρις πατοῖθ᾽- ὁ δ᾽ οὐκ εὐσεβής. His
boastful insistence on his δέκη (1577-1611) has not for a moment impressed the
Chorus: there is no sign here of any wavering in their judgement as when
they deliberated upon the rights and wrongs of Clytemnestra's case (1530 ff.
and 1561).
For the z- alliteration at the beginning and end of the line cf. on 268.
1670. With the substitution of χρόνωι for χάριν Wecklein (Stud. z. Aesch.
148) has hit the mark. His arguments are first that in accordance with the
common usage there should be a plain genitive depending on ἄποινα, and
secondly that we expect a similar indication of time here as in 1666. In
support of his correction Wecklein quotes Suppl. 732 f. χρόνωι τοι κυρίωι τ᾽
ἐν ἡμέραι θεοὺς ἀτίζων τις βροτῶν δώσει δίκην. It may be assumed with some
probability that, to begin with, the end of the column was here, as at 1664,
1672, 1673, mutilated by some injury, and that next someone filled in the
two missing syllables by writing χάριν without giving much thought to it.
799
 line 1671 COMMENTARY

1671. Stanley compared Pindar, Ol. 12. 14 ἐνδομάχας ἅτ᾽ ἀλέκτωρ συγγόνωι
παρ᾽ ἑστίαι, to which Boeckh ad loc. adds A. Eum. 861 ff. Hermann (on Pers.
756) supposes that in the passage of the Eumenides and similarly in Pers. 756
(ἔνδον αἰχμάζειν) and in Ag. 1671 the same proverb is alluded to. For the
particular colour given to the expression here cf. Vet. Test. Proverbia 30. 31
ἀλέκτωρ ἐμπεριπατῶν θηλείαις εὔψυχος (note the context there also).
For the history of the word ἀλέκτωρ, which is foreign to pure Attic, see
Ernst Fraenkel, Nomina agentis, i. 154 f. ‘The post-Homeric word ἀλέκτωρ
means “‘fighter’’, and provides evidence that in those times the Greeks were
primarily interested in the fighting-cock, and not the laying hen’ (Kretschmer,
Glotta, xxvii, 1939, 36).
The confusion of ὥστε (here required by the metre) and ὥσπερ is common,
cf., e.g., Bast in G. H. Schaefer's edition of Gregorius Corinthius, p. 888.
1672 f. EyoA. wad. in Tr: ᾿Εγώ, φησί, kai σὺ κρατοῦντες τῶνδε τῶν δωμάτων
διαθησόμεθα Ta! καθ᾽ αὑτοὺς καλῶς. On the authority of this scholion ἐγώ was
rightly put into the text at the end of 1672 by Canter, and καλῶς by Auratus
at the end of 1673. But if we are to be on secure ground here, we must first
guard against misinterpretation of the scholion. A. Kirchhoff (Berl. Sitzgsb.
1894, 1052), who was also wrong in doubting that the scholion is one of the
Σχόλια παλαιά, spoke of ‘the ra καθ᾽ αὐτοὺς for which there is no place’, and
Wecklein (at the end of his commentary on the Ag. [1888], and in the Addenda
[1893] to his edition of Aeschylus, p. 368) stated: ‘legendum: κρατοῦντες
τῶνδε τῶν δόμων διαθησόμεθα τὰ κατ᾽ αὐτοὺς καλῶς ', which is fantastic.
διαθησόμεθα τὰ καθ᾽ αὑτοὺς καλῶς means ‘we will put our affairs properly in
order’. The use of the reflexive pronoun ἑαυτοῦ, etc., for the second person
(cf. on 1141, 1297, 1544) and the first person? occurs as early as the time of
Aeschylus; there are examples of it in Herodotus (ἑωυτοῦ = σεωντοῦ) and in
the Attic orators, and it is found with increasing frequency in the Hellenistic
period. It is the only form used, e.g., by Polybius (cf. Kühner-Blass, i. 599
n. 2 and also Kühner-Gerth, i. 572) and in the New Testament (Blass-
Debrunner, Gramm. des neutest. Griech., sth ed., 38: ‘In the plural, ἑαυτῶν
alone is used for all three persons, as is the general rule in Hellenistic Greek.)
This usage is normal also in the language of the scholiasts. Schol. Ar. Birds
807 f. may perhaps not be regarded as cogent evidence, for the paraphrase
' οὐχ ὑπ᾽ ἄλλων πάσχομεν ταῦτα ἀλλὰ τῆι ἑαυτῶν γνώμηι and ἀντὲ τοῦ ᾿ ἑαυτοῖς
ταῦτα πεποιήκαμεν᾽ might possibly have been influenced by the words of
Aeschylus (fr. 139 N.) quoted in the text of Aristophanes, τοῖς αὑτῶν πτεροῖς
ἁλισκόμεσθα. But there is no doubt about Schol. Apoll. Rhod. τ. 866 f. (on
ἦε γάμων ἐπιδευέες ἐνθάδ᾽ ἔβημεν κεῖθεν, ὀνοσσάμενοι πολιήτιδας ;), where we
read the paraphrase ἢ γάμων ἐνδεεῖς ὄντες ἐνταῦθα ἥκομεν, μεμψάμενοι τὰς
ἑαυτῶν πολίτιδας κτλ. The pronoun in our scholion is therefore correct.
! ra, which is required by usage (see the instances quoted below of ra κατ᾽ ἐμέ etc.),
is a correction made by Victorius, whereas Triclinius wrote τὸς There is a confusion in the
notes of van Heusde and Wecklein. Dindorf's account (Philol. xx, 1863, 29) of the reading
in Tr is incorrect too.
2 Cf. Aesch. fr. 139. 4 N. Hermann took the pronoun in Cho. 1014 νῦν αὐτὸν αἰνῶ as =
αὐτόν (me ipsum), and he has been followed by many commentators (including Blass) and
by Kühner-Gerth, 1572. But this rests on a wrong interpretation of the passage (cf. above
on 1482, p. 703 n. 2). I leave aside the fixed combination αὐτὸς κατ᾽ αὐτοῦ and the like (cf,
above on 836), which sometimes refers to the speaker himself (e.g. Cho. 221).
800
 COMMENTARY lines 1672f.

Perhaps we should restore καθ᾽ <é)aurows, cf. both the scholia just quoted and
the remark of Blass-Debrunner, loc. cit. (‘always ἑαυτοῦ, not αὐτοῦ, in the
N.T.). The phrase διαθησόμεθα τὰ καθ᾽ ἑαυτούς, too, is in keeping with the
habitual language of the scholia. This periphrastic expression (τὰ κατ᾽ ἐμέ
or with some corresponding pronoun) is often ernployed when the com-
mentator wants to indicate the grammatical object more fully. Cf., e.g.,
schol. A. Prom. 527 f. Wecklein (== 511) on οὐ ταῦτα ταύτηι μοῖρά πω τελεσ-
φόρος κρᾶναι πέπρωται: od ταῦτα οὕτως πέπρωται, ἵν᾽ ἡ τελεσφόρος Μοῖρα ταχέως
τὰ κατ᾽ ἐμὲ κράνηι καὶ πληρώσηι, schol. A. Suppl. 250 Weckl. (== 244) on καὶ
τἄλλα πόλλ᾽ ἐπεικάσαι δίκαιον ἦν: ἔμελλον ἂν στοχασμῶι τὰ καθ᾽ ὑμᾶς λέγειν,
schol. S. Oed. R. 377 on τάδ᾽ exmpäfaı! ἀντὶ τοῦ τὰ κατ᾽ ἐμὲ ἐκδικῆσαι, schol.
S. El. 1366 on at ταῦτά σοι δείξοισιν: ταῦτα, τὰ κατ᾽ ἐμέ φησιν, schol. E. Or.
451 ON καὶ μὴ μόνος τὸ χρηστὸν ἀπολαβὼν ἔχε: ἀντὶ τοῦ ἀπόλαβε τὰ κατὰ σαυτὸν
φυλάσσων, schol. Apoll. Rhod. 1. 660 f. on μὴ δ᾽ ἄμμε κατὰ χρειὼ μεθέποντες
ἀτρεκέως γνώωσι: μήπως διὰ τὴν τῶν ἐπιτηδείων ἀνάγκην καὶ χρείαν... ἀκριβῶς
τὰ καθ᾽ ἡμᾶς μάθωσιν. To return to the scholion on Ag. 1672, there is nothing
in it to make it probable that, apart from the ἐγώ and the καλῶς (both words
were obviously still extant at the time when the scholion was written), the
commentator read the endings of the two lines in a form different from that of
our MSS. The τῶνδε in τῶνδε δωμάτων, which, as will be shown, is suspicious,
recurs in the scholion. The words τὰ καθ᾽ αὐτούς, ‘our affairs’, apparently
represent an attempt on the part of the paraphrast to supply an object for
θήσομεν from the context.
Now we can at last turn our attention to the two lines themselves. The
point that strikes us at once is that θήσομεν καλῶς requires an object. Any
unprejudiced reader will feel the want of it, and this first impression is con-
firmed by an examination of the usage. Instances of the use of τιθέναι καλῶς
and τιθέναι εὖ in Tragedy have been collected in the note on 913. It appears
that there is always an object attached to these expressions and that the
object precedes the verb? in the great majority of cases (exceptions are, c.g.,
E. Her. 938, Iph. A.672). The difficulty about the missing object here was felt
long ago; consequently Musgrave wanted to write dAaypa: πάντ᾽ ἐγώ, Weil πᾶν
ἐγώ, Wecklein (with some hesitation) ταῦτ᾽ ἐγώ. But this would definitely make
things worse, for in the first place ματαίων ὑλαγμάτων belong together (cf. 1631),
and in the second place the idiomatic construction of μὴ προτιμήσηις with the
genitive must not be disturbed. An objection has been raised to this construc-
tion by E. Harrison, Cambridge University Reporter, 11 November 1941 (he
1 Cf.,e.g., (where the accusative after «ard is not a pronoun) schol, A. Cho. 84 on δωμάτων
εὐθήμονες : εὖ τιθεῖσαι τὰ κατὰ τὸν οἶκον, Cho. 510 f. Weckl. (= 512 f.) on τὰ δ᾽ ἄλλα... ἔρδοις
ἂν ἤδη: τὰ κατὰ τὴν σφαγὴν Αἰγίσθου, schol. 9. El. 1434 on τὰ πρὶν εὖ θέμενοι: τὰ κατὰ τὴν
Κλυταιμήστραν, schol. Ar. Birds 1536 δοκεῖ τὰ κατὰ τὴν ἀθανασίαν αὕτη οἰκονομεῖν.
2 In Eur. Antiope B 10 (D. L. Page, Greek Lit. Pap. 64) we must accept v. Arnim’s supple-
ment for the lost beginning of the line ra δ᾽ ἔνδον ἡμ]εῖς καὶ σὺ θήσομεν καλῶς, or supply some
similar object. In A. Pers. 283 the text is very uncertain, for in this passage both strophe
and antistrophe (289) have suffered damage (Wilamowitz’s supposition that the middle
syllable of εὔνιδας is long here is extremely improbable and is not sufficiently supported by
the reference to the δορκαλῖδες of Herodas), and Hermann’s text, which is doubtful in other
respects as well, ὡς πάνται παγκάκως θεοὶ ἔθεσαν (without an object for the verb), appears to
be even more questionable after what has been said above. The same construction and
order as in Tragedy are to be found, e.g., Hdt. 7. 236. 3 τὰ σεωυτοῦ δὲ τιθέμενος εὖ γνώμην
ἔχε κτλ.
4872.3 Y 801
lines 1672 f. COMMENTARY

proposes with Rauchenstein and others to read ὑλάγματα) : 'προτιμῶ elsewhere

in Aeschylus governs the accusative.’ This is not really a sufficient argument.
‘Elsewhere’ in this case means ‘in two passages’ (Eum. 640, 739). As against
this we have to consider the fact that a negatived προτιμᾶν with the genitive
is a peculiarly Attic construction, the idiomatic character of which was
pointed out by ancient grammarians: schol. Ar. Plutus 883 ‘ οὐδὲν προτιμῶ
cov ' où φροντίζω σου" ἡ φράσις Arrwy. εἴρηται δὲ τό προτιμᾶν ἐπὶ τοῦ λόγον
ἔχειν (cf. Phrynichus, Praep. soph. p. 103. 8 de B., Photius, Suidas s.v. προτι-
μᾶν). Passow and the dictionaries deriving from his work quote for this
construction Ag. 1672, E. Alc. 761 f., Ar. Plutus 883, [Demosth.] 7. 16, but
do not include Thuc. 8. 64. 5 τῆς ἀπὸ τῶν ᾿Αθηναίων ὑπούλου εὐνομίας οὐ mpo-
τιμήσαντες (this, against the majority of MSS, is the reading of B, confirmed
by the scholia and Dion. Hal). This being the case, it would hardly be
reasonable to challenge in the text of Aeschylus a characteristic Attic
construction." Lawson rightly agrees with those who feel the need of an
object to θήσομεν, but he proposes τῶνδε δωμάτων σέβας and thus rejects
καλῶς, which as we have seen rests on good authority. The apologetic
explanation of Schneidewin and the similar one of Wecklein ‘ θήσομεν, scil.
τὰ περὶ ra δώματα [cf. his misinterpretation of the scholion pointed out
above], which is to be understood from xparoüvre τῶνδε δωμάτων ᾿ are un-
satisfying.
But the absence of an object for θήσομεν καλῶς seems not to be the only
oddity here; I have difficulties also with the pronoun in τῶνδε δωμάτων. In
the only two passages in Aeschylus where δώματα has a deictic pronoun
attached to it the distinction made is an indispensable one (the singular
δῶμα with such a pronoun does not occur in Aeschylus, which may be merely
an accident) : Zum. 179 (hardly relevant to our passage, for there it does not
mean ‘household, family’, etc., but the actual building, the temple) e&w...
τῶνδε δωμάτων (‘out of this house of mine’; similarly Ewm. 60 the priestess,
referring to Apollo, says τῶνδε δεσπότηι δόμων, cf. above on 1225 δεσπότηι)
and Cho. 692 ὦ δυσπάλαιστε τῶνδε δωμάτων Apd. What is needed in Ag. 1673
is not ‘since we are masters in this house’ or ‘in the house here’ (there is no
question at all of any other house) but ‘since we are masters in the house’.
Fortunately we know how Aeschylus denotes the exercising of authority in a
house: Cho. 716 τοῖς κρατοῦσι δωμάτων, Cho. 658 τοῖσι κυρίοισι δωμάτων (cf.
also Eum. 740 ἄνδρα κτανούσης δωμάτων ἐπίσκοπον). Wilamowitz, it is true,
detected a particular nuance in the τῶνδε. He says, Hermes, xxxiv, 1899, 67:
‘The queen closes the whole play with an utterance of which the exact
wording is uncertain,? but the sense unmistakable “We shall arrange it, this

! Harrison raises a further objection to the ending of 1672 in the form which is commonly
printed : ‘a strong stop before ἐγὼ gives a Sophoclean detachment of a short word from its
own line’. The observation that this punctuation with a stop before the last two syllables
of a line and this type of enjambement has no parallel in the tetrameters of the Persae
and the little we possess of the trochaic tetrameters of the old iambographers, though there
is one in 8. Oed. R. 1529, is in itself valuable, only it is not enough to discredit the reading
ὑλαγμάτων" ἐγὼ, on the contrary, it rather supports it. What the consequence would be if
we tried to remove the ‘Sophoclean’ enjambements from the Oresteia (and the Prometheus)
has been shown in the notes on 1271 and 1354.
2 This sound scepticism has disappeared in Wilamowitz’s treatment of the passage in his
edition of Aeschylus.
802
 COMMENTARY lines 1672.

house is ours”. If only we do full justice to the indication of the stage action
contained in these words, if we picture Aegisthus being led with apparent
reluctance by his wife to the door, and, when she says τῶνδε δωμάτων, turning
towards the house which he is to enter, while the Chorus, in spite of their
scornful utterances, prepare to depart .. . we shall need nothing more.’ It
seems almost past belief that Aeschylus should have combined the idea ‘we
are lords and masters (and therefore need be afraid of nobody)’ with the
indication, which for Aegisthus is quite unnecessary, ‘this is the house whose
masters we are’. If Aeschylus had meant to imply a stage-direction such as
Wilamowitz supposes, he would probably have made Clytemnestra say:
‘Let us go into this house (the house here).' Actually her significant closing
words are entirely devoted to a wider issue; so she speaks of what she
claims to have won, and what she hopes for (and also what she is secretly
afraid of).
To sum up. I believe that in 1673 Aeschylus wrote καὶ σὺ δωμάτων kparoûvre
— u θήσομεν καλῶς, and that the object of θήσομεν must have been contained
in the two missing syllables. How τῶνδε got into the text I donot know. Itis
perhaps something more than an accident that in 1657 a certainly spurious
τούσδε has intruded itself after πρὸς δόμους. Perhaps in 1673 too someone to
begin with inserted an explanatory τῶνδε before δωμάτων, and then someone
finding the line without a diaeresis proceeded to transpose the words. As
regards the gap which I have left unfilled I can only offer a guess. A possi-
bility would seem to be πάντα θήσομεν καλῶς (or ταῦτα or τἄλλα). Perhaps too,
θάμά (= ra dua) might be considered. duds with the meaning of ἡμέτερος is
found in Homer (cf. Wackernagel, Unters. z. Homer, 52) and its use is also
established in Aeschylus: Eum. 311; the pronoun in this form is found in
dialogue too: Suppl. 322 (319 Wil.), Sept. 654. This would imply the assump-
tion that in τὰ καθ᾽ αὑτούς the scholiast was not intending to supply an object
from the context, but was actually paraphrasing ra dud (and on this assump-
tion we should have to regard the τῶνδε of the scholion, not as a repetition
of a word from the text of the poet, but as explanatory of δωμάτων). How-
ever, it seems rather hazardous to assume that duds = ἡμέτερος could
be used in dialogue, moreover the expression τὰ ἡμέτερα (καλῶς θήσομεν)
sounds clumsy. I therefore prefer πάντα or some similar object of a more
general kind.
θήσομεν καλῶς, as the concluding phrase of this drama, has a powerful
tragic effect. There can be no doubt about Clytemnestra’s true feelings: her
deep anxiety and her fear that the daimon of the house may indulge in
fresh horrors has found expression first in the kommos and now in this very
scene, shortly before the end (1654-60). The audience will not be deluded by
the apparent self-assurance in the queen’s last words ; and her lover’s tyranny,
based on brutal force, may be triumphant for the moment, but the next play
will soon reveal a complete and terrible retribution.
A Kirchhoff, Berl. Sitzgsb. 1894, 1050, assumed that 1673 was originally
followed by a series of tetrameters and then anapaests by the departing
Chorus. This has been refuted by Th. Pliiss, Die Tragödie Agamemnon
und das Tragische, Basel 1896, 22, and by Wilamowitz, Hermes, xxxiv,
1899, 67 f. Wilamowitz’s comparison of this finale with the end of the
Oedipus Rex is helpful, since there the play closes in the same way with the
803
lines 1672 f. COMMENTARY
actor's! trochees. In the case of the Agamemnon there is a special reason
for the unusual form of the conclusion: 'facit Clytaemnestra finem
tragoediae, quoniam chorus, cuius alias hoc officium esse solet, susceptis
hac in scena actoris partibus non recte potuit ad perorandum adhiberi'
(Hermann).
11 cannot make any compromise on this much-disputed point.

804
 APPENDIX A
On the Postponement of certain Important Details in Archaic
Narrative
THE particular mode of archaic narrative alluded to above (p. 39) is well
illustrated by the story of the infant Cyrus’ deliverance from death in Hdt.
x. 110-ı2. To begin with we have a detailed description of how the herdsman
Mitradates lives out on the mountainside with his wife Spako, then how
Harpagos sends for him to come to the city, and charges him by the king’s
order to expose in the wilderness the infant who is handed to him. The
herdsman takes the child home with him. τῶι δ᾽ dpa καὶ αὐτῶι ἡ γυνὴ ἐπίτεξ
ἐοῦσα πᾶσαν ἡμέρην, τότε κως κατὰ δαίμονα τίκτει οἰχομένου τοῦ βουκόλου ἐς
πόλιν. ἦσαν δὲ ἐν φροντίδι ἀμφότεροι ἀλλήλων πέρι, ὃ μὲν τοῦ τόκου τῆς γυναικὸς
ἀρρωδέων, ἡ δὲ γυνὴ ὅ τι οὐκ ἐωθὼς à Ἅρπαγος μεταπέμψαιτο αὐτῆς τὸν ἄνδρα
(111. x). When the husband comes home, his wife asks him why Harpagos has
sent for him. A long and detailed narrative by the herdsman follows (to the
end of 111). Then he shows her the child whom he has brought with him.
Full of pity for the child she begs her husband to spare his life, but in vain;
the husband insists that he is bound to carry out his orders. So in her
desperation she proposes a solution: τέτοκαϊ γὰρ καὶ ἐγώ, τέτοκα δὲ τεθνεός"
τοῦτο μὲν φέρων πρόθες, τὸν δὲ τῆς ‘Aorudyeos θυγατρὸς παῖδα ὡς ἐξ ἡμέων
ἐόντα τρέφωμεν (112. 2 f). And so it is done. In this narrative—here merely
sketched in outline—we not only have the feelings of the characters concerned
tenderly and vividly developed, but the whole picture is worked up with
considerable artistry. The baby Cyrus keeps his central position throughout ;
Spako’s infant is only mentioned as a means to an end. This is all far from
primitive.” All the same it must be recognized that the statement τέτοκα...
τεθνεός is very strangely delayed, judging from the point of view of later
narrative technique. H. Stein says in his commentary on 111: ‘Anxiety
about her husband and the reason for his being sent for have put all other
things out of the woman’s mind.’ That may be so. It is all the same astonish-
ing that the husband, who has been described as τοῦ τόκου τῆς γυναικὸς
ἀρρωδέων, never asks a question about the birth. But above all the author
might just as well have said in 111. 1, if he had chosen, τίκτει παῖδα τεθνεῶτα
instead of the laconic τέκτει. This he did not do, because he was postponing
this touch until the point in the story at which it was most necessary and
important, that is to say the plan for the child’s substitution.
For the postponement of ‘pieces of information which are necessary and
help to elucidate the situation’ in a narrative of Pindar, cf. L. Illig, Zur Form
der pindarischen Erzählung (1932), 25 n. 3.
ı For the function of the perfect cf. J. Wackernagel, Studien zum griechischen Perfehtum
(Univ. Programm Göttingen 1904), 19 f.
2 Geffcken (Hermes, lxii, 1927, 13), it is true, passes this judgement on the narrative: ‘a
story which shows, not without a certain unintentional comicality, that Herodotus is
occasionally incapable of handling his material in a convincing and impressive way’.

805
 APPENDIX B
On the Weapon with which, according to the Orestera, Agamemnon
was murdered
In regard to this question there are three different views which have found
supporters in recent times. (1) Wilamowitz, following Stanley and others,
repeatedly and vehemently declared in favour of an axe (or a hatchet),! e.g.
Interbr. 173 n. 1, where he says in conclusion: ‘the assertion that Clytem-
nestra is represented anywhere in Aeschylus as using a sword is given the lie
direct in the text’; but by this he means the text (at any rate as regards Ag.
1116) in the form which results from his own conjecture. The earlier history
of this ‘hatchet’, elaborated in imaginary colours, can be found in Wilamo-
witz’s introduction to his translation of the Agamemnon, Griech. Tragödien,
ii. 40 (the main point had already been put forward doubtfully by Abresch
on Cho. 889, and rejected by Schiitz). The view that the implement was an
axe is shared by, e.g., Bethe, RE xi. 892, H. W. Smyth, Aeschylean Tragedy,
157; Porzig, Aischylos, 72; Pohlenz, Griech. Tragüdie, i. 101; Latte, Hermes,
Ixvi, 1931, 132 n. 2; Lesky, ibid. 193 ; Schadewaldt, Hermes, lxxi, 1936, 64 n. 1;
Deubner, Abhandl. Preuss. Akad., Phil.-hist. Kl. 1941, no. x. 22; Snell, Die
Antike, xx, 1944, 132 (against his own better judgement in Gnomon, x, 1934,
416).
(2) The view that the weapon was a sword has been maintained by Klausen
(on Cho. 1011), Lewis Campbell (Am. Journ. of Philol. i, 1880, 438 f.), Wecklein
(on Ag. 1497, Cho. 1009 in his numbering, Sitzgsb. Bayer. Akad., Phil.-hist. Kl.
1911, 3. Abhdl, 7f.), Tucker (edition of the Choephoroe [1901], p. 265 f.),
among others, and particularly by Blass (edition of the Choeph. on 886 ff.),
and so, too, E. Petersen, Rhein. Mus. Ixvi, 1911, 31 f., Bruhn, Introd. to
Soph. El. (Schneidewin-Nauck, roth ed., 19:2), p. 8 n. 1, Mazon on Ag.
1149 (translation), Denniston on E. El. 164, Tierney, C. Q. xxx, 1936, 103.
Schuursma, 77 n. 3, regards it as more probable that the weapon is a sword ; he
touches briefly on one or two arguments, but makes no claim to have gone
into the matter thoroughly.
(3) A verdict of ‘non liquet' was reached by Carl Robert in his last dis-
cussion of the question, Griech. Heldensage, x298, n. x, whereas previously
(Bild und Lied, 176 1.) he had decided in favour of the hatchet.
Of these three views? it is actually the first mentioned which ‘is given
1 This view seems to be influenced by the form in which the story is presented by Sophocles
and Euripides and which is common in later antiquity. As far as the Oresteia is concerned,
the identification of the weapon with an axe is found in the scholion (ZxoA. mai. in Tr)
on Ag. 483 (489 Weckl.) τὸν πέλεκυν λέγων, δι’ οὗ καθεῖλε τὸν ‘Ayauéuvova and in the quite
arbitrary interlinear gloss of the Mediceus on 1149, where τῶ() πελέκει stands above the
words ἀμφήκει δορί. On the other hand, both here and on 1496 βελέμνωι Tr has the gloss
ξίφει. It appears that Triclinius (or his authority), starting obviously from the unambiguous
passages (vide infra), had arrived at a clear idea of the facts, though he gives a more definite
meaning to both expressions than was intended by tbe poet.
2 A compromise between (1) and (2) is recommended by Warr, C.R. xii, 1898, 348 ff. : ‘she
first cut him down with two blows of the axe falling on the head . . . then, when he was
down and at her mercy, she finished him with one thrust of the sword of Aegisthus.” But
this view is not borne out by the text of Aeschylus. Verrall, however, The Ag. of Aesch.,
and ed. (1904), Introd. xlvi n. 2, seems to incline to this solution.
806
 APPENDIX B
the lie direct in the text’; to realize this, one need only look at Wilamowitz’s
explanatory note on rorr in his commentary on the Choephoroe, where
something is introduced into the passage of which there is not the slightest
indication in the context. On the other hand, of the solutions (2) and (3), each
has a certain degree of justification. As far as (2) is concerned (‘sword’), Cho.
torr should leave no doubt on the point here in question : ἔδρασεν ἢ οὐκ ἔδρασε ;
μαρτυρεῖ δέ μοι φᾶρος τόδ᾽, ὡς ἔβαψεν Αἰγίσθου ξίφος. φόνου δὲ κηκὶς ξὺν χρόνωι
ξυμβάλλεται. The bloodstain on the robe testifies to the murder. Whose sword
was responsible for inflicting the wound? Orestes cannot tell just from the
robe; if, nevertheless, he is quite sure about it, we find here a psychological
process which takes a short cut, in just the same way as previously in the
anagnorisis. He knows about Agamemnon’s murder: so the bloodstain and
the rents in the robe become fully valid for him as evidence. The really sur-
prising thing is that just at this juncture where everything depends for him
on the proof of his mother’s guilt (ἔδρασεν ἢ οὐκ ἔδρασε ;) Orestes speaks of the
sword of Aegisthus. Yet there is good reason for this. Orestes is standing
beside the dead bodies of Clytemnestra and Aegisthus ; he wants to establish
his own justification fully, and accordingly he refers expressly once again
(though only incidentally, as the man's share in the action is only a sub-
ordinate one) to the part played by Aegisthus, not indeed as accomplice in
the act, but as accomplice in the planning and preparation for Agamemnon’s
murder. What he is saying now is all of a piece with his statement just before
(978) ξυνώμοσαν μὲν θάνατον ἀθλίωι πατρί, but especially with the altercation
between Aegisthus and the Chorus at the close of the Agamemnon (1612 ff.),
where strong emphasis is laid on the charge that Aegisthus was jointly
responsible for the murder as βουλεύσας. This share in the responsibility is
most clearly indicated by the fact that Aegisthus lent his sword to the
murderess for the deed which the two of them had planned long before. The
case is quite different when Clytemnestra, suddenly taken by surprise (Cho.
889), cries out δοίη τις ἀνδροκμῆτα πέλεκυν ws ráxos,! ‘since there are no actual
weapons in the γυναικών ' (Blass, loc. cit.). Latte, Hermes, lxvi, 1931, 132.
developing this hint of Blass's, says ‘while real weapons are regularly kept in
the great hall or in the entry of the house, the axe is always ready to hand in
the inner rooms of the house. . . . It is the nearest available weapon, the
weapon which is seized in the moment of excitement [cf., e.g., Merope in
Hyginus Fab. 137], in contrast to sword and spear, which first have to be
fetched.’ Though this distinction cannot be maintained in all cases without
exception, it is nevertheless clear how much less suitable an axe would be
than a sword for the carefully premeditated act of Clytemnestra and her
lover. At all events in Cho. ıorı, where the share of Aegisthus in the planning
of the deed is to be indicated, it was for the sake of this that Aeschylus made
mention of his sword. Similarly there is a special point in the mention of the
sword in Ag. 1528. ‘An eye for an eye and a tooth for a tooth’: the lex talionis
demands the most exact correspondence of execution as between the guilty
deed and its expiation. This is why Clytemnestra uses these words here of
Agamemnon, after mentioning the sacrifice of Iphigeneia : ἄξια δράσας, ἄξια
πάσχων μηδὲν ἐν Audov μεγαλαυχείτω, ξιφοδηλήτωι θανάτωι τείσας ἅπερ ἔρξεν.
ı For a corresponding representation of Clytemnestra with an axe in vase paintings
earlier than the Oresteia see below, p. 809 ἢ. 2.
807
 APPENDIX B

The daughter was slain with the sword (cf., e.g., Eur. [ph. T. 27 ἐκαινόμην ξίφει,
785 φάσγανον) : for this sin the murderer, her father, has atoned by death by
the sword. To these passages a third must be added which points in the
same direction. In Ag. 1262 f. Cassandra says of Clytemnestra: ἐπεύχεται
θήγουσα φωτὶ φάσγανον ἐμῆς ἀγωγῆς ἀντιτείσεσθαι φόνον. Here of course the
emphasis is not on ihe wav in which the murder is carried out, but on the
connexion between the fates of Cassandra and Agamemnon, but here again
the weapon is the sword, and the fact that the prophetess has a clear vision
of the details of her coming doom gives especial value to her statement. The
three passages discussed here leave no doubt that, in agreement with the
narrative in the Nekyia (A 424), Aeschylus thought of a sword as the instru-
ment of Agamemnon’s murder.
But the unambiguous explicitness of the three passages in which the sword
is mentioned—not primarily to depict a detail of the murder but to stress
certain circumstances connected with the deed (the participation of Aegisthus
in the plot, the retribution for Iphigenia’s death, the destiny shared by
Cassandra and Agamemnon)—in some measure is counterbalanced by the
strange indefiniteness in the mention of the weapon precisely where we might
expect special attention to be called to it. It is this vagueness which induced
some scholars to answer the question about the nature of the weapon with
a non liquet (see above). In the words in which Cassandra's vision of the
king’s murder culminates (1126 ff., for the details see the commentary) all the
emphasis is laid, as elsewhere throughout the play, on the fateful robe: as to
the kind of weapon not a word is said; only a thrust or a blow is mentioned.
But Aeschylus might have found further opportunity of giving a clear indi-
cation of the weapon if he had been so minded. The fact that he was not can
be seen above all from Clytemnestra’s own account after the deed (1380 ff.).
She gives many details fully and with the greatest explicitness, she devotes
three whole lines to the description of the entangling robe, but then (1384) all
she says is: παίω δέ wv Sis. Not a word about the weapon. The same holds
good for the narration of the deed in Apollo's speech at the trial, Hum. 633 ff.
Finally there is the refrain of the stanzas of the Chorus (Ag. 1495 f. =
1519 f) in which the dead king is addressed: δολίωι μόρωι δαμεὶς (δάμαρτος)
ἐκ χερὸς ἀμφιτόμωι βελέμνωι. The wording of this passage almost gives the
impression of being devised to elude inquisitiveness. The noun and the
adjective are so chosen that both of them (the noun only at a pinch) might
possibly refer to a sword, though they are by no means bound to do so; the
same is true of the words with which Cassandra foretells her own death,
σχισμὸς ἀμφήκει δορί (1149, vide ad loc.). Now the objection should not be
made that the poet has spoken unambiguously of the sword in other passages
(i.e. those mentioned above). To this the answer is that it is not allowable to
cite evidence from different scenes of a play—or from different plays of a
trilogy—in order to settle the facts of a given case in the way in which one
would quite properly, indeed inevitably, compare the evidence in widely
separated chapters of an historical text, where one detail naturally supple-
ments the other. If, during the scenes (or parts of scenes) which centre in the
king's murder, the spectator is left in doubt as to the nature of the weapon
used, it does not help him much to be enabled by casual statements made
later on to say to himself by way of afterthought: ‘ah, then it was a sword
808
 APPENDIX B
after all.” But even if it did help him—the poet deliberately shaping his
drama must have had some definite intention in being secretive about the
nature of the weapon for so long and exactly where the audience was entitled
to hope for some elucidation. This intention on the part of the poet is the
factor with which we are primarily concerned here: we can recognize it quite
plainly. In order to heighten the significance with which Aeschylus invests
the unique and characteristic instrument of the woman’s treachery, the
splendid festal robe which turns into a net of death, he will not allow the
weapon which actually deals the fatal blow to obtrude itself in any way
upon the consciousness of the audience. As to its being a sword or some other
weapon—we should give as little thought as possible to that; it is the robe,
the net, which claims the whole and undivided power of our imagination.
These considerations, however, as we have seen, did not prevent the poet
from letting the sword be mentioned more than once outside the immediate
sphere of the act of murder.
It is, of course, well known that in Sophocles’ Electra (99) Agamemnon is
said to have been slain with the axe, and similarly in the Hecuba (1279, here
by way of prophecy), the Troades (361 f.) and the Electra (279) of Euripides.
It is impossible to say whether this goes back to an older! tradition, which
differs from that of the Nekyta in this detail (this was the view of Robert,
Bild und Lied, 177 [though no longer held in Griech. Heldensage, 1298], Wila-
mowitz, Griech. Tragödien, ii. 40, Blass, Introd. to Choephoroe, p. 5), or
whether the axe was not introduced till the time of the younger tragedians.
There is no information on the point in Schol. E. Hec. 1279, nor have we any
reliable monuments of decorative art of sufficient antiquity.? It seems to
me likely that Pindar, Pyth. 11. 20 by the words πολιῶν χαλκῶν means a
sword, as the paraphrase in the scholia understands it. [See the Addenda.]

APPENDIX C
Cho. 991-1006
THE speech of Orestes, Cho. 973-1006, contains a difficulty of the first order.
It will be expedient to leave aside, at any rate for the moment, some minor
points and concentrate upon the crucial issue. What is the meaning of 997
τί νιν προσείπω! As the text stands in the MS, any reader must needs refer

1 It is arbitrary to conclude from the well-known fragment of Stesichorus (fr. 15 D., cf,
p. 740) Tae δὲ δράκων ἐδόκησε κτλ. that he represented Agamemnon as having been slain
with the axe, as is maintained by Robert, Wecklein, and others.
2 Jt cannot be proved that the woman running with a double axe to a great wooden door,
on the Berlin cup by the Brygos painter (Furtwängler, Beschreib. d. Vasensammlg., no. 2301,
Beazley, Aitic Red-Figure Vase-Painters, p. 252 no. 104), is Clytemnestra about to murder
Agamemnon. (On the analogy of other vase-paintings, as Beazley points out to me, one
may perhaps see here, too, a representation of Clytemnestra after the surprise attack of
Orestes.) Representations of the death of Aegisthus of an earlier date than the Oresteia,
depicting Clytemnestra with a double axe in her hand hurrying to the rescue of her lover
who is being attacked by Orestes (Furtwängler-Reichhold, plate 72, with other examples
given in the text), do not, of course, prove anything with regard to the way in which
Agamemnon’s murder was represented at that time. For the considerably later picture on
the cup from Spina (the slaying of Cassandra) cf. on 239, p. 138 n. 2.

809
 APPENDIX C
νιν to the person who is the subject of the whole preceding period 991-6,
Clytemnestra. Only when we go on do we realize with a shock that Orestes
seems to be talking no longer about his mother but about the fatal garment,
the ‘net’. The difficulty was of course noticed long ago,' and for more than a
hundred years many critics have resorted to transpositions or other altera-
tions of the text. Before, however, considering any such changes, we must
give the arrangement in the MS the benefit of the doubt. Our first task,
therefore, will be to examine the views of some of those scholars who thought
they could understand the unaltered text.
In Gilbert Murray’s edition we read the following note on 997: ‘997-1004
post 982 ξυνωρίδα traiecit Scholefield, sed vide Verrall ad loc.’ It might have
been better to refer to Conington instead of Verrall, for the latter confines
himself to quoting Conington’s note and expressing complete agreement
with it. This is what Conington says: ‘Or. wishes to find a name for his
mother, without saying anything that ought not to be said. νιν is Clyt., as is
evident from what goes before. He proceeds to identify her with the net, the
instrument of her crime, enlarging on its villainous uses, as if he had nothing
else in his mind, till in v. 1005 he at last returns to her. This identification is
doubtless a symptom of the frenzy which is beginning to work on him, at the
same time that it has its own imaginative truth. Precisely the same identifi-
cation is made by Cassandra, Ag. 1114 foll. € £, παπαῖ παπαῖ, ri τόδε φαίνεται ;
*H δίικτυόν τί γ᾽ Ἅιδου ; AAN ἄρκυς ἡ ξύνευνος, ἡ £vvavría φόνου. Thus it would
be worse than useless to follow Meineke (anticipated by Scholefield) in trans-
posing this and the seven following lines so as to insert them after v. 982’ etc.
After what I have said already in the commentary on the Cassandra scene,
there is no need to reject once more the identification which Conington
assumes there. The subject of the phrase ἀλλ᾽ ἄρκυς ἡ ξύνευνος κτλ. cannot be
anything but the garment, the net, as it appears to Cassandra in her vision.
The whole character of the scene with its clear distinctness of each separate
vision makes it impossible to imagine that even for a moment Cassandra
could confuse the apparition of the garment with that of the woman. The
rejection of the parallel from the first play of the trilogy would not, however,
be fatal to Conington’s main thesis if his interpretation of the speech of
Orestes carried conviction in itself. But I think it does not: Conington,
whose sympathetic understanding of Aeschylus often proves most admirable,
has erred here. It is fundamentally wrong to believe that at any moment
during Orestes’ first speech in the last scene, 973-1006, ‘the frenzy is beginning
to work on him’. The poet clearly marks the transition from a perfectly sane,
if highly excited, mood to mental disturbance. In his third speech, at 1. 1026,
Orestes says ἕως δ᾽ ἔτ᾽ ἔμφρων εἰμί. He is obviously aware of the imminent
danger of madness, but not yet in its grip. At 1048 a fit of frenzy sets
in:* he bursts into the wild shout d, ἃ (cf. the corresponding interruption
of the trimeters Ag. 1214); now he sees the Erinyes, though no one else
can see them (1061), they are unmistakable and real, no deceptive illusions
(1053 f). The difference between the stage at which a tormented man is
afraid lest he may go mad, while the working of his mind is still perfectly
1 See, e.g., Butler: ‘non premam suspiciones meas versiculum in quo diserte vestem
nominaverat Orestes hic [i.e. after 1. 996] intercidisse.'
2 For the text see p. 318 n. 1.
810
 APPENDIX C
clear, and the later stage at which frenzy has actually taken hold of him
was as well known and as important to Aeschylus as it was to Shakespeare.!
We are not entitled to correct the poet’s conception by confusing what he
kept separate. Neither is it permissible to resort to psychological subtleties
and argue that Orestes is mistaken about his own state of mind and that,
in fact, he has already been raving a considerable time before he protests
that he is still sane but fears madness. Such a presentation of a decisive
turn would not be in the manner of Aeschylus. Moreover, according to
Conington, Orestes would ‘identify Clytemnestra with the net’ in that very
speech whose central theme (980 ff.) is the real net, the garment, and, what
seems to me quite intolerable, would do so a moment after that garment has
been unfolded with elaborate care before the eyes of the people on the stage
and of the audience. No one who is aware of the dominating position heid by
the fatal cloak throughout the Oresteia, no one who appreciates the con-
centrated energy with which the poet in every speech of this culminating
scene of the Choephoroe dwells on one primary issue, will be prepared to
accept Conington’s ‘identification’ and the substitution implied in his
hypothesis.
Another attempt to interpret 997 ff. without altering the order of the lines
in the MS was made by Wilamowitz. His view is fully expressed in his com-
mentary (published 1896) on the Choephoroe and in his German translation
with its stage-directions ; a brief summary is given under the heading ‘Actio’
in his editio maior of Aeschylus.” Wilainowitz assumes that in 983 f. Orestes
bids his servants spread out the garment and carry it round in a circle
(‘entfaltet das und tragt es rings im Kreis herum’), ‘from παρασταδόν we learn
that the servants approach each man in the crowd separately; the horror of
the onlookers as they one by one inspect the murderous cloak must have
greatly enlivened the scene’ (commentary on 983). Then, according to Wila-
mowitz, the garment, after being carried round the whole circle, would be
returned to Orestes at 997. ‘Orestes uses the mere pronoun vw, which was
quite intelligible to the spectators although the reader here and at 1004
requires a παρεπιγραφή᾽ (ibid.). “vw was quite intelligible to the spectators.’
Was it indeed? The light, purely anaphoric pronoun, after the long diatribe
against Clytemnestra? In his translation Wilamowitz goes round the obstacle
by rendering τί νιν προσείπω : ‘das da —. . . wie bezeichn' ich's recht?’ That
would be τόδε rather than νιν. But we need not elaborate this point. For
Wilamowitz’s whole construction is based on a misunderstanding of 983
κύκλωι παρασταδόν. Whether we accept with Wilamowitz the reading of M
παρασταδόν [cf. S. Trach. 194 f. κύκλωι... παραστάς) or write περισταδόν, the
meaning of the clause is clearly : ‘spread it out and stand near by (or “round’’)

! T am referring to the third act of King Lear, where Lear says (iii. 2. 67): ‘My wits begin
to turn’, and again (iii. 4. 21 £.): ‘O, that way madness lies; let me shun that; No more of
that’. Then, iii. 4. 28 ff., he speaks of the ‘poor naked wretches’ in deeply moving lines full
of humanity and wisdom, without any tinge of mental disturbance. A short time afterwards
the words ‘Hast thou given all to thy two daughters?’ etc., and later on ‘What, have his
daughters brought him to this pass? ...’ (iii. 4. 48 f. and 62 ff.) and his following outbursts,
reveal the first fit of unmistakable madness.
2 The scanty later remarks of Wilamowitz, Interpr. 215, can hardly be regarded as a
substantial contribution. It may, however, be noticed that the line of psychological
approach which he followed there had been anticipated, even in the details, by Paley.
8ir
 APPENDIX C
inaring...to make it possible for Helios to see it.” The ἄπειρον ἀμφίβληστρον
is a garment of uncommon size (cf. on Ag. 1382); to unfold the huge thing
properly and show it to the Sun, the attendants have to stand round in a
circle,! each of them holding a piece of the edge. There is nothing in the text
of Aeschylus to suggest that the cloak should be carried round.
Other critics? have resorted to transpositions. Scholefield,’ and, inde-
pendently, Meineke (see Hermann’s note), placed 997-1004 after 982. This was
accepted by Hermann and many others. At first sight it seems very attrac-
tive, for it rids us of the two main difficulties in this speech by giving the νιν
of 997 its proper relation to the garment and removing the awkward abrupt-
ness with which, according to the MS text, Orestes at 1005 returns to Clytem-
nestra, Still we should hesitate to acquiesce in this alteration. For, as Weil
observed, ‘ita Orestes alloqueretur velum nondum explicatum’. In the text
as it stands in the MS the large cloak is first unfolded in an action which takes
a considerable time (see below) and only then does Orestes turn to it with his
excited questions τέ νιν προσείπω κτλ. We shall presently be in a position fully
to appreciate this arrangement, a master-stroke of Aeschylean stage-effect.
But even now we see enough to reject an alteration in consequence of which
Orestes would first expatiate on the enormous size of the voluminous garment
(997-1004) and afterwards have it unfolded so that the spectators may have
the full view of it. It is obvious that it is the actual sight of the evil thing
when it is widely spread out that provokes the outburst τέ νιν προσείπω κτλ.
Weil himself inserted 997-1004 in the next speech after 1013, Blass and
A. Y. Campbell, C. Q. xxix, 1935, 31 n. 4, followed him. The prejudice that it
was desirable to make the three speeches of Orestes about equal in length had
a share in Weil’s transposition. The change implies grave consequences. In
the first place, it is obviously much more appropriate that the questions τί νιν
προσείπω κτλ. should belong to that part of the dramatic action which has as
its centre the spreading out of the cloak. Secondly (and this I regard as even
more vital) Weil seems completely to miss the keynote of the intense excite-
ment in Orestes’ second speech (roro ff. ἔδρασεν ἢ οὐκ ἔδρασε ; krÀ.). In this
speech it is one issue and one alone that matters : the matricide, weighed down
by the consciousness of his pollution (1017), asks from the depth of a tortured
mind ‘was she really guilty?’ Only in so far as it provides an answer to this
horrible doubt does the garment play a part in this brief speech. Any diver-
sion from the burning problem would be intolerable here. It should also be
1 Cf. P 391 f. δεξάμενοι (the ox-hide) δ᾽ dpa τοί ye διαστάντες τανύουσι κυκλόσε (κύκλωι
Zenodotus).
2 Tucker maintains the arrangement of the MS, but gives no substantial explanation.
From his comment on 997 τέ νιν προσείπω, together with his critical note (‘Weil strangely
transposes’ etc.), it appears that he is not even aware of the real difficulty. His com-
mentary on the Choephoroe, published in 1901, shows no sign of any acquaintance with the
commentary of Wilamowitz published in 1896.
3 In his second edition of Aeschylus, 1830, Scholefield speaks with remarkable reserve:
‘Neque tamen diffiteor, olim me in suspicionem incidisse, versus hos a recto ordine detrusos
fuisse. Certe facilius procederet oratio, si post 982 legerentur vv. 997-1004. Deinde v. 1005
ad Clytaemnestram optime referretur. Sed huiusmodi suspiciones omnino ipsae non sine
suspicione sunt excipiendae, ne forte pro ipso Aeschylo exhibeamus “disiecti membra
oetae”’.’
P + He also transposed 1014: in his first edition he placed it after 1015, later on (Teubner
text) after 1016. But that is a point of minor importance.
812
 APPENDIX C
noticed that the whole thought of 997-1004 shows that when Orestes speaks
these sentences he feels no doubt whatsoever as to his mother’s guilt. It may
perhaps be tempting that, if we followed Weil, the προσφωνῶν in 1015 would
seem to hark back to 997 τί vw προσείπω. But this apparent harmony is
deceptive, for προσφωνῶν points in fact to the following words ἀλγῶ μὲν ἔργα
KrA.: the προσφωνεῖν of the murderous cloak finds its expression in the ἀλγεῖν.
After finding ourselves forced to discard several ingenious attempts, we
may now be disposed to do justice to a suggestion of a different type. That
ruthless obelizer W. Dindorf applied his formidable scalpel to the speech of
Orestes. But he went much too far, cut out twenty lines (987-1006), and thus
killed the patient straight away. Of his excellent observations on certain
passages of the speech something must be said later on. More cautious than
he, Wecklein contented himself with bracketing 991-6 and 1005 f. The result
of this operation seems to be very satisfactory. Before we go into details, it
will be useful to examine the general consequences. Let us then provisionally
accept Wecklein’s deletion as a working hypothesis. What is left after the
removal of 991-6 proves to be a consistent and highly effective piece of
dramatic construction. At 983 Orestes gives the order to unfold the garment.
Several attendants are wanted to hold the large drapery: they step round ina
circle, grasp the border, and then slowly lift the thing and spread it out until
it is fully displayed. This action, performed not in a hurry, but in the digni-
fied manner suitable for a tragic climax, takes time:! only at the end of
1. 99o is it completed. It is worth noting that the whole period from 983 to 990
is governed by only one primary clause: ἐκτείνατ᾽ αὐτὸ kai . . . δείξατε. This
command is the pillar that supports the whole structure; all that follows is,
both in syntactical form and in matter, subservient to the central thought
and action. Therefore, if Orestes after 99o, ie at the moment when the
spectators have obtained a full view of the instrument of crime, continues
(997) τί νιν προσείπω κτλ., there is no break or abruptness whatsoever.
It is possible that at least some of the obscurities and the clumsy phrases
which in 993-5 have puzzled the critics are to be accounted for by the assump-
tion that this section was not written by Aeschylus but patched together (see
below) by a botcher. To begin with 993, who or what is the subject of
gpaiver? ‘Nobody knows’ (A. Y. Campbell, C. Q. xxix, 1935, 32). Many editors,
e.g. Paley, Tucker, Headlam, Blass, regard τέκνον or τέκνων βάρος as the
subject ; Sidgwick ‘as she shows’; Conington ' φαίνει seems to be impersonal
.. as things show" '. Wilamowitz in his first edition connected ws φαίνει
καικόν, which seems unbearable; afterwards he gave this up, without, how-
ever, indicating what he thought to be the subject. In 994 f. the construction
is utterly baffling. As it stands (with θιγοῦσαν), its disentanglement seems
hopeless, as may be seen from the determined, but unsuccessful, attempt of
1 It can often be observed that a dramatic poet calculates the length of a speech or a
section of a speech in accordance with the time required for a particular action on the stage
(for S. Aj. 579-95 see T. v. Wilamowitz, D. dramat. Technik des Sophokles, 60). The speech
of Clytemnestra Ag. 958 ff. affords a very striking example. At 957 the king begins slowly
to walk from the place where the chariot stands (probably in the centre of the orchestra)
towards the door of the palace, at 972 he is still outside, for the words ἀνδρὸς τελείου κτλ, are
obviously meant for his ear, whereas the following outburst Zed Ζεῦ τέλειε κτλ, would be
inconceivable if he could hear it. It is therefore clear that between 972 and 973 he enters
the house.
813
 APPENDIX C
Wilamowitz (p. 242 f. of his commentary). The majority of the editors
accepted Robortello's reading of θιγοῦσαν as θιγοῦσ᾽ dv, but then we should
certainly expect μὴ δεδηγμένον (which here is metrically impossible) instead of
οὐ öed. (cf. Wilamowitz) ;! the scholiast, with disarming naiveté, paraphrases:
καὶ τὸν μὴ δηχθέντα. Moreover, many editors have rightly objected to ἄλλον
(Blass’s excuse is not helpful, his treatment of the whole sentence is violent
and unconvincing). I doubt whether any amount of critical ingenuity would
be sufficient to rid these lines of their muddle, for this seems to be a case where
incerta haec si tu postules ratione certa facere, nihilo plus agas quam si des
operam ut cum ratione insanias. As for the content, I gladly subscribe to the
verdict of Sidgwick : ‘these violent and almost grotesque words’. In 996 there
is nothing confused, but the sense, especially in connexion with the monstrous
snake, strikes me as extremely weak and the phrasing as commonplace.
The lines 991-2 are of better quality ; it would not be surprising if this clause,
or part of it, had been taken over from another tragedy.
But it is the general turn of thought in 991-6 and 1005 f. that, more than
any detail, ought finally to shake our belief in the genuineness of these lines.
Here W. Dindorf, despite his rashness, deserves high praise for his thorough
insight into the character of the Aeschylean Orestes. He says (in the preface
to his sth edition of Aeschylus, 1870, p. xcviif.): “Perpetrata matris caede
Orestes orationem suam his finit verbis [980-6], quae nullum produnt
vehementiorem animi motum, sed solam recte convenienterque oraculo facti
conscientiam, quod ut omnibus patefiat, Solem . . . invocat, ab conviciis vero
in matrem iaciendis prorsus abstinet. Nam Orestes, licet apud Aeschylum
pariter atque Sophoclem et Euripidem matris facinus aversetur, tamen cum
ea et viva et post caedem, quam solo Apollinis iussu ab se invito et haesitante
patrari identidem dicit, satis moderate pro criminis ab ea commissi gravitate
agit, ut vel ex versibus huius fabulae 892-930 videre licet.' The last remark is
particularly valuable.? It may be added that the same noble restraint can be
perceived in the words with which Orestes, Eum. 458-61, describes the murder
of Agamemnon, and also in his moving allusion to his own crime, Eum. 611
(on the meaning of these words cf. my note on Ag. 1171, p. 534 n. 1).
From these observations the following conclusions may be drawn with a
fair degree of probability (there can be no certainty in such a matter). The
noble simplicity of Orestes' speech would not satisfy the coarser taste of a
later generation ; some people might complain that ‘there were no sallets in
the lines to make the matter savoury'. À clever producer, sailing with the
wind of the aura popularis, and perhaps working in the interest of an actor
who wanted a piece of rant to display the power of his voice, thought he
could improve on Aeschylus by putting in a few lines of his own. In doing so,
he freely stole feathers from the old master's plumage. As Dindorf saw,
μύραινα (994) is ἃ variation of the ἀμφίσβαινα in Ag. 1233 (there the expression
is as suitable to the ecstatic horror of Cassandra as μύραινα is revolting in the
mouth of the son), and ἔχιδνα comes probably from the very fine passage
Cho. 249. The question 994 τί σοι δοκεῖ; seems to have been inspired? by 997
1 Tucker speaks of ‘the emphatic οὐ (instead of the possible μὴ), “though not bitten".
2 This aspect of the scene had already been emphasized by Otfried Müller (Aesch. Eum.
8 98, p. 195), who gave a fine general appreciation of the figure of Orestes in the Choephoroe.
3 The author of the ridiculous lines E. Or. 588-90 (the thought is even worse than the
814
 APPENDIX C
Tí vw mpooeinw...;1regard this very repetition as a strong argument against
the genuineness of 994 ff., for Aeschylus would hardly, within a few lines, have
capped the question *what is my mother (or, as others understand the text,
‘her deed’) like?’ with the question ‘what name shall I give to the garment?’
So anxious was the amplifier to have Orestes abuse his mother more than
once that he did not even refrain from spoiling the conclusion of the original
speech by adding 1005 f. As long ago as Scholefield's edition it was realized
that 1005 does hot attach to the preceding lines. But the fact that there is no
connexion in thought is not the only objection to 1005 f. It should also be
noticed that the sentence zoor ff. τοιοῦτον. . . φρένα (with most modern
editors I regard Lobeck's correction as necessary) constitutes the conclusion
of the whole speech: for the opening of the tail-end of a fous by τοιοῦτον cf,
on Ag. 613 f. Itisintolerable that the formal ending τοιοῦτον κτλ. should be
duplicated’ by the addition of 1005 f. τοιάδε κτλ. As regards the thought of
1005 f., it is not only vulgar, it is revolting.? Fancy this Orestes at this
juncture considering possibilities of marriage!
So much about the main issue. I want to add a brief remark on a point of
minor importance. It seems to me almost certain that 1028 πατροκτόνον
μίασμα kai θεῶν στύγος was not written by Aeschylus, but concocted by the
same man who added 991-6 and 1005 f. It is the only other passage in which
Orestes applies abusive terms to his mother. It destroys the fine structure of
the preceding sentence, which culminates in the momentous οὐκ ἄνευ δίκης,
exactly as the verdict on Aegisthus at 99o culminates in the concluding ἔχει
... δίκην. 1028 is patched together from 1015 πατροκτόνον θ᾽ ὕφασμα and Ag.
1645 χώρας μίασμα kai θεῶν ἐγχωρίων.

ΑΡΡΕΝΌΙΧ D
The Footprints in the Choephoroe
OF those critics who are not conservative at all costs, very few seem to be
inclined to accept ll. 201-11 and 225-30 as they stand in the MS. In many
modern editions we find the lines printed with transpositions, deletions,
lacunae, of a more or less violent character. Some of the main difficulties of
the first passage have been so clearly pointed out by Schütz (in his first
edition) that we cannot do better than quote from his note the following
remarks: ‘. . . ipsa verborum collocatio laborat. Ut enim nihil dicam de
pleonasmo satis inficeto [205 f.] orißoı . . . ποδῶν ὅμοιοι τοῖς 7” ἐμοῖσιν ἐμφερεῖς,
quorsum quaeso pertinet illud [207] καὶ yap? quod facit ut exspectemus
grammatical blunder which shows that the man was as unfamiliar with the difference
between γαμεῖν and γαμεῖσθαι as the interpolator of E. Med. 262) got the cue for his insertion
ὁρᾶις κτλ, from 591 ὁρᾶις. [Cf. Dem. 9. 6-8 el μὲν οὖν... εἰ μὲν odv....]
1 There is no such duplication in the apparently similar ending of Clytemnestra's speech
Ag. 312-16, for there rouoiôe . . . νόμοι concludes the description of the beacon-post, whereas
the following τέκμαρ τοιοῦτον links the whole speech up with, and makes it explicitly an
answer to, the question of the coryphaeus (272) ἦ γάρ τι πιστόν ἐστι τῶνδέ σοι τέκμαρ;
2 For the manner in which the speaker passes on to a wish that concerns him personally
cf., e.g., E. El. 948 f. (Electra facing the body of Aegisthus) ἀλλ᾽ ἔμοιγ᾽ εἴη πόσις μὴ παρθεν-
ὠπός, ἀλλὰ τἀνδρείου τρόπου.
815
 APPENDIX D
causam, cur haec pedum similitudo δεύτερον τεκμήριον sit. . . . omninoque
versus 209.210 inanem habent eorum quae v. 205.206 dicta erant repeti-
tionem. . . . illa [212] εὔχου rà λοιπά, quibus verbis Orestes infit, statim finitis
Electrae precibus v. 201-4 pronuntiari debebant. Ut nunc est, inepta arguta-
tione! de indicio ex vestigiorum similitudine capiendo nimis distrahuntur.
Quicquid fit, si ab Aeschylo reperta fuerit haec ἀναγνώρισις, ingenue fateor me
eius invento minime delectari, et facile patior eum ab Euripide hisce versibus
perstrictum, El. 532 sqq.' The solution suggested by Schütz will be discussed
later on. Many of his successors tried to blunt the edge of his criticism by
resorting to the only too familiar excuse that what seems to be muddle is in
fact a product of refined psychology. It may suffice to quote, as an early
specimen of that mode of interpretation, Wellauer's (1824) attempt to refute
Schütz with arguments like these: 'Electram prae desperatione et subita spe
perturbatam animo fingit. . .. Quod autem idem saepius dicit, ignoscendum
est animo perturbato, qui quae spem faciunt non satis habet semel dixisse.'
Such soothing considerations will always appeal to readers who are apt to
forget that we know something about the structure and lucidus ordo of
Aeschylean speeches, whether they be delivered in a normal frame of mind or
in great excitement.
Among the questions raised by Schütz, there is one which fortunately
cannot be answered by vague generalities: quorsum pertinet illud [207] kai
ydp?' Several translators? skip the disconcerting γάρ altogether. Others
ascribe to «ai γάρ a meaning which, though in itself possible, cannot be
thought of here, e.g. Lewis Campbell: 'Moreover, there are two of them,
outlines of different feet.’ On this usage of καὶ γάρ where ' καί is the connec-
tive, and καὶ ydp means “yes, and", or "and further" ’ cf. Denniston, Par-
ticles, 109 f., who shows that it is sometimes found ir. answers and adds:
"Whether xai ydp (as distinct from καὶ yàp οὖν, καὶ γάρ τοι) is ever so used in
continuous speech, may be questioned.' Therefore ydp here must denote a
reason or point to an explanatory argument. In regard to this detail Her-
mann, while rejecting the general criticism of Schütz, took advantage of his
predecessor's observation. This is what Hermann says: ‘Nam quum duorum
vestigia cerni dicat Electra, altera Orestis, altera comitis cuiuspiam, quis
dubitabit, praesertim quum etiam ad orationis integritatem aliquid desi-
deretur [since καὶ yàp . . . seems to indicate the reason for something which is
not mentioned in the text], quin alterius istorum vestigia suis dissimilia,
alterius similia esse dixerit? Itaque ante v. 209 [πτέρναι κτλ. lacunae signa
posui. Hermann's assumption of a lacuna was accepted by some of his
successors, e.g. by Weil and, in principle, by Wilamowitz,* who thought,
however, that something was lost not after 208 but after 206. Others, more
conservative, thought they could follow Hermann in his reconstruction of the
thought without assuming a gap. Conington: ' xai ydp, confirmatory. “The
correspondence is easily seen: for there are two different sets of footprints,
1 Probably a misprint for argumentatione.
2 eg. Verrall, Introduction, p. liii, and Wilamowitz.
3 Murray's note ‘post 206 [it ought to be 208] lacunam Herm.’ is based on a wrong
inference from the app. crit. of Wilamowitz.
^ His arrangement of the text and his note in the edition of 1914 imply a recantation of
his earlier (1896) treatment, which was partly based on the impossible view that 'if she sees
two footprints, they are one of the right foot and one of the left’.
816
 APPENDIX D

his own, and those of some fellow traveller :” z.e. the footprints reduce them-
selves to two types, one of them doubtless his, while the other merely shows
that some one has been with him. The condensation of the expression is as
natural as the simplicity and precipitancy of the reasoning. . . . Had she pro-
ceeded, as Herm. thinks she ought to have done, to enlarge on the character-
istics of the various footprints, remarking that some are like, and others
unlike her own, the explicitness would only have injured her case, and done
injustice to the vehemence of her feelings.’ This explanation was adopted by
Sidgwick,' Verrall, Tucker. Conington’s remark on the ‘condensation of the
expression’ in a yap clause is in itself sound, cf. Denniston, Particles, 61, 76,
and my notes on Ag. 214 f. and 217. But the line of argumentation which he
finds in the sentence seems extremely odd, and the same must be said of the
interpretation of Hermann and his followers. While it may be natural for a
sister who is longing for her brother’s return to say: ‘Here are footprints for
a second piece of evidence (that he has come back), very much like my own’,
it is absurd (particularly from the point of view of a Greek, who takes the
individuality of a foot for granted) that, by way of confirmation, she should
add: ‘Yes, they are like, for there are others here not like, probably a com-
panion’s’ (Sidgwick’s summary of Conington’s reconstruction). To sum up:
the attempts to vindicate the transition καὶ γὰρ δύ᾽ ἐστὸν κτλ. from the criti-
cism of Schütz have been unsuccessful.
Neither has it been possible to invalidate Schütz’s observation that the
first words of Orestes (212) εὔχου rà λοιπά would be much more effective if they
immediately followed the conclusion of Electra’s prayer 201-4, in other words
if Electra’s speech ended with σπέρματος μέγας πυθμήν. To this we may add
an observation of a different character. That 204 should form the end of
Electra’s ῥῆσις is desirable not only because thus her concluding sentence
would be closely linked up with the following utterance of Orestes but also
because, generally speaking, Aeschylus is very much given to crowning an
important speech by a prayer or prayer-like invocation at its end.” In regard
to the conclusion of Electra's ῥῆσις, Schütz's criticism left a sting in the minds
of some of his successors. Butler, though afraid of the boldness of the
deletions advocated by Schütz (see below), admitted that 'illa εὔχου τὰ
λοιπά... in vulgatis, ut vere obiecit Schutzius, ab Electrae precibus nimis
longo intervallo separata erant'. He therefore placed 201-4 after 210. This
Blomfield accepted with some reserve (‘et haec quidem correctio verisimilis
videtur. Weil improved on Butler's suggestion by placing 201-4 after 211.
But, as Blass rightly remarks, 2or does not fit on well to 2x1 (nor, for that
matter, to 210). Several other transpositions have been suggested ; but with
all their violent and improbable alterations (see, e.g., the arrangement of
Wecklein, adopted by Headlam in his translation) they do not yield a
satisfactory result.
We now turn to the other passage in this scene where a grave disturbance is
unmistakable, 226-30. Here the difficulty was, at any rate partly, recognized

1 His alternative suggestion, ‘ “Yes, it is he, for here are his companion’s footsteps”, i.e.
he is likely to have come with a companion', need not be taken seriously.
2 The fact is well known, instances can be easily collected. Cf. also the general remark of
W. Kranz, Stasimon, 39: ‘a short prayer may almost be regarded as the regular form for
concluding a long speech or choral ode in Aeschylus'.
4872.3 2 817
 APPENDIX D
centuries ago by Robortello. He transposed 228 (ἰχνοσκοποῦσα κτλ.) to the
place after 226 (κουρὰν δ᾽ ἰδοῦσα «rA.). We need not discuss this alteration
or the transpositions of Pauw, Heath, Bothe, and others. Neither need we
dwell on the many changes which, apart from transpositions, have been
recommended. While some of these devices are ingenious, all are utterly
arbitrary. The discussion would probably have taken a more reasonable
course had it been realized that the trouble in the footprint-passage, 205-11,
and the trouble round the footprint-line 228 ixvookoroüca κτλ. should not be
treated as two separate problems but as one. The one scholar who did see
this simple truth was Schütz.
Schütz, in his first edition, came to the conclusion that 205-10 and 228
were probably! interpolated. It would have been better if he had obelized
211 and 229 as well (see below). But even as it is, his verdict does credit to
the fine Greek scholar of whom G. Hermann said (Opusc. vi. ı, p. 97) that of
the editors of Aeschylus ‘der Veteran Schütz ist der einzige, dem poetischer
Sinn nicht abzusprechen ist’. Schütz’s deletion of the obnoxious lines was
unsuccessful: no editor followed him, and unfortunately Schütz himself lost
confidence. In his second edition he spoilt his discovery by a weak com-
promise. In a case like this no half-hearted treatment can be of any avail.
The text of the speech of Orestes which was the subject of Appendix C had to
be cured in accordance with the old aphorism ὁκόσα φάρμακα οὐκ ἰῆται,
σίδηρος ἰῆται. We shall presently see that the same maxim holds good for
the criticism of the footprint-lines. But before we fully explain the salutary
effect of our amputation, we must return for a moment to 228 and its neigh-
bourhood.
When Wilamowitz, in his first edition, bracketed 228, 229 (from ixvooxo-
ποῦσα to «apaı)? and left the order of the surrounding lines unchanged, he
might have been told: ὦ Αἰσχύλου ἐξηγητά, εἰ kai τὰ δέοντα ποιεῖς, ἀλλ᾽ οὐκ
εἰδώς γε ποιεῖς. He removed the ἰχνοσκοπία in 228 without allowing this
deletion to influence his judgement on 205 ff. Moreover, he did not under-
stand the meaning of 229. In his note, he translated the passage thus: ‘und
auf meinen Tritten spürend, die deinem Bruder gehörten, der dir gleicht’,*
although he was fully aware of the absurdity of using in this context ‘the
periphrasis τῶι oc κάραι instead of coi’. There is no reason for assuming
such an artificiality. 229 is indeed miserable and clumsy verbiage, but not
really obscure, provided that we read it in the place where it stands in the
MS and that we do not wantonly alter συμμέτρου. It is obvious that the words
σαυτῆς... κάραι are meant to go with what follows; therefore we must
punctuate after 228. I reluctantly give a translation of 229 f. as the lines
stand: ‘Put the lock of hair to the place from which it was cut (on the head)

τ *Parum mihi tempero a suspicione . . .' he says in his cautious way.

2 R, Böhme, too, who in Hermes, Ixxili, 1938, 195 ff., renewed the ἀθέτησις of Schütz,
confined himself to bracketing 205-10 and 228.
3 What matters most is that he deleted both these lines while leaving the rest untouched.
Before him, l. 228 was regarded as an interpolation by Schütz, and on 229 Wecklein remarked
in his edition of 1885 'eiciendus videtur' (afterwards, in his commentary of 1888, he did not
even mention his former suggestion).
4 There is even less excuse for Headlam's translating in the same way (‘and when thou
wast track-reading in my footprints, thine own brother's, proportioned like thyself’), for
he believed the lines to be Aeschylean.
818
 APPENDIX D

of your own brother, who corresponds to your head,' and consider it.’ This
is far from good, but it is what the botching writer of 229 wanted to express.
After interpolating a line (228) about the ixvooxoria he had to work his way
back to 230 and the theme of the lock. He managed to do that, snvifissima
Minerva, by adding a perfectly superfluous explanation to τομῆι. In the
edition of 1914 Wilamowitz recanted his misinterpretation of 229 (‘factus est
ut post 226 legeretur’) and rightly maintained the deletion of the line, but
failed to see that it was written for the place where it stands. 228 he now
regarded as genuine (accepting Robortello’s transposition).
As 229 was intended to serve as a link between the primary subject of the
interpolation (the reference to the footprints, 228) and the original continua-
tion in the text of Aeschylus, so was 211. With 210 the speech of Electra
would have ended on a note of unqualified hope so that the reply of Orestes
would have been hanging in the air. To avoid that, the interpolator inserted
the words πάρεστι δ᾽ ὠδὶς καὶ φρενῶν καταφθορά.
Now at last we are in a position to look back and appreciate the outlines of
this section of the play as they present themselves after being cleared of the
ugly accretions (205-11 and 228-9). Throughout the scene which begins after
the short song of the Chorus (164) and ends with the entry of Orestes (212), the
lock is the dominating factor. Not for a moment is our attention diverted
from it. Electra's fervent desire to communicate her discovery of the lock to
the women of the Chorus and let them participate in her profound emotion
sets the dialogue afoot. After the end of the stichomythia the eyes of the
audience as well as their thoughts are entirely concentrated on the little
ringlet of hair which Electra holds in her uplifted hand: 187 πλόκαμον τόνδε,
188 τῆσδε φόβης, 193 τόδε, 197 τόνδε πλόκον. After what is perhaps the most
beautiful of the many ἀποσιωπήσεις in Aeschylus (194), Electra's excitement
reaches its culmination in the outburst φεῦ. εἴθ᾽ εἶχε φωνὴν «rA. At this point
the actor's movements and gestures, too, must have come to a climax, when
Electra, holding the dear treasure in front of her eyes, gazes at it passionately
as if it were the face of her brother, until, worn out by her longing and the
alternating pangs of hope and despair, she at last drops it in resignation and
addresses herself to the gods, who alone know the truth. No sooner is her
prayer concluded than the bringer of its fulfilment, Orestes, steps forward.
His first words εὔχου τὰ λοιπὰ rots θεοῖς κτλ. are a perfect echo of the beginning
of her prayer (dAAa . . . τοὺς θεοὺς καλούμεθα) and with the following τυγχάνειν
καλῶς he takes up her phrase εἰ δὲ χρὴ τυχεῖν σωτηρίας. The complete unity of
purpose in this admirable scene and its concentration on a single deeply
moving theme affords as strong an argument against the genuineness of
205—11 as the difficulties of detail which have caused so much trouble to the
critics.
After the sostenuto of Electra's speech, the dialogue between brother and
sister carries the action on in a powerful accelerando until the final solution of
the tension is achieved when (233) Electra's last doubts are overcome and,
without a word, she passionately embraces him ; her heart betokens her rapture,
προφθάσασα καρδία γλῶσσαν. It is only to remove those last faint doubts that
a transient use is made of the motif of the embroidery as an ἀναγνώρισμα
ı Probably a legitimate compression of a phrase such as ‘whose head and hair correspond
to yours'.
819
 APPENDIX Ὁ

(231: f). The two lines, coming at the very end and quickly passing by, serve
their purpose well without unduly engaging our attention or impairing the
importance of the lock whose identity Electra has first to recognize (230)
if she is to believe in her brother’s return at all. If, on the other hand, 228
were genuine, it would split the fine counterpart to the clause αὐτὸν μὲν οὖν
ὁρῶσα κτλ. into two rivalling conceptions and grossly spoil its intimate con-
notations. For the lock is something infinitely more delicate and precious
than footprints: it seems, in its tender shape, to embody the beloved head
itself. In this respect the feeling of the Greeks was not different from our own.
Leaving out of account religious ideas (¥ 146 ff., A. Cho. 6 f., etc.), it may
suffice to recall A. Sept. 49 f. and similar representations in figurative art and
poetry, cf. Schadewaldt, Röm. Mitteil. xxxviii/xxxix, 1923-4, 490, C. H. E.
Haspels, Attic black-figured Lekythoi (1936), 71. Thus the balance between 225
and 226 is perfect, and in 227 the emotions that dominated the preceding
scene are beautifully summed up.
What was the motive of the interpolator? The answer is not so plain as it
was in the case which we examined in Appendix C. This time nothing more
than a general guess seems to be possible. From the sixteenth chapter of
Aristotle's Poetics it is apparent that in the fourth century the tragedians set
great store by the use of the ἀναγνώρισις and that each tried to outdo the other
in inventing fresh patterns and modifying and refining the old ones.! It is
highly probable that the theory of dramatic art kept pace with, and partly
stimulated in its turn, the progress made by the playwrights. The very
manner in which Aristotle introduces his treatment of the problem makes it
unlikely that the subdivision of the mass of ἀναγνωρίσεις into different species
and their arrangement in order of merit (ἀτεχνοτάτη . . . drexvou. . . πασῶν δὲ
βελτίστη ἀναγνώρισις) was an invention of his own; it rather looks as though
he were reminding his reader of a well-known classification. We have good
reasons for believing that from the later part of the fifth century onwards
there existed an ever increasing literature on the τέχνη and theory of tragedy ;?
in those books the discussion of the fashionable topic of the ἀναγνώρισις must
have played a considerable part. We may then surmise that at the time when
so much interest was taken in the refinement of a recognition scene some
ambitious producer of the Choephoroe thought that the manner in which
Aeschylus had handled the subject was really too poor. He trusted perhaps
that it would make the speech of Electra more interesting if in addition to
what seemed to him her one and only 'piece of evidence' (from the point of
view of Aeschylus the lock serves a purpose very different from that of a mere
τεκμήριον) another, a δεύτερον τεκμήριον, were inserted. It is conceivable that
the interpolator saw a further advantage in the opportunity of dragging in
Pylades (207 f). In doing so he sinned sadly? against the plan and spirit of
the play of Aeschylus. It has long been observed that Pylades, who had his
1 Cf. Pohlenz, Die griech. Tragödie, i. 519: ‘the later writers [i.e. the tragedians of the
4th century] were content to vary the forms used by their predecessors or to improve on
them by artistic means. In particular they were constantly trying fresh methods of treating
the motif of “recognition”.’
2 Cf. W. Kranz, Neue Jahrbücher f. d. klass. Altert. xliv, 1919, 148 ff. ; Pohlenz, Nachr.
Gölt. Ges. 1920, 142 ff.
3 Not so sadly, however, as the scholar who in the Annual of the Brit. School at Athens,
xxxvii, 1936-7, 201 ff., assigned to Pylades a fine series of long speeches from line 653 on.
820
 APPENDIX D
fixed place in the tradition of the story (cf. Pind. P. 11. 15), is but a dim figure
in the Choephoroe, and is not permitted to grow into a substantive character.
In the whole drama he speaks only three lines (900 ff.), momentous words
indeed, but he does not utter them on his own account but as a mouthpiece of
Apollo (cf. Otfried Müller, Die Eumeniden, p. 132; Wilamowitz, Introd. to his
commentary on the Choephoroe, p. 41). Throughout the recognition scene,
throughout the following speeches of Orestes and the long kommos, Pylades
is never mentioned or alluded to in a single word. So careful is the poet to
make it clear that in this tragedy it is not Orestes the friend who matters but
Orestes the son, Orestes the brother. Only when it proves unavoidable, in the
exposition of the plan for the murder (561 ff.), is Pylades taken into account,
for the first time after the prologue. Up to that point, he remains somewhere
in the background, a shade rather than a person, unmentioned and un-
thought of.! Whoever endeavoured to draw attention to Pylades at a most
inappropriate moment (207 f.) did so in complete disregard of the intentions
of Aeschylus.

But it is not sheer folly to obelize lines the authenticity of which is at-
tested by Euripides? Anyone with even a slight knowledge of Greek tragedy
remembers what was stated in Stanley’s commentary on Cho. 231 (229 Stanley)
and has been said repeatedly since, that Euripides in his Electra (520-44) went
out of his way to ridicule the improbabilities of the recognition in the
Choephoroe and among them the use made of the footprints. It is not to the
credit of our classical studies that this wretched problem keeps on troubling
us. The way to its solution was shown many decades ago ; all that we have to
do is to follow it up. Two factors seem to have co-operated to prevent the
acknowledgement of the simple truth. There was first the general dislike for
accurate analysis of a dramatic structure. Moreover, in this case, particular
circumstances helped to divert attention from the fundamental problem.
Most readers of the recognition scene in the Electra have been so entirely
engrossed in the question whether or no the ‘criticism’ of Euripides was fair
to Aeschylus that they did not pause to ask what was the relation of that
part of the scene to its surroundings and to the play as a whole? The
proneness to narrow the field of observation was further encouraged by the
great and lasting influence? which A. W. Schlegel’s rash invective exerted on
the whole subsequent discussion. One scholar, however, managed to avoid
the fallacies of an isolating interpretation and went a long way towards the
understanding of the true contents of the scene. In 1877 August Mau pub-
lished an article ‘Zu Euripides Elektra’, Commentationes philol. in honorem
Th. Mommseni, 291 ff., in which he conclusively proved that El. 518-44 is an

1 The conventions of sth-century dramatic art do not allow the spectator to take an
interest in an object (or a person) on the stage until the dialogue draws attention to it.
Cf. the commentary on 950 and on 1650 (p. 783).
2 It may suffice to quote two examples of this typical attitude, Verrall's Introduction to
his edition of the Choephoroe, pp. xxxiii-Ixx (he dwells at great length on the Euripidean
scene but is in no way concerned with its amazing inconsistencies), and G. Murray’s note,
at the end of his translation of Eur. El., p. 89 f.
3 Cf., e.g., the Introductions of Verrall to the Choephoroe, pp. xxxvi ff., Ixii ff., and of
C. H. Keene to the Electra of Euripides, pp. iv ff., where much space is given to a discussion
of Schlegel’s views.
821
 APPENDIX D
interpolation and made it highly probable that 545 f., attaching to 517, must
be assigned to the Old Man. Mau did not carry the day. Being suspicious,
as well he might be, of a method which then was in the hey-day of its wanton-
ness, Wilamowitz dismissed Mau's suggestion with a brief general remark,!
which shows how completely he failed to see the objective of the genuine
Euripides, and with a wrong interpretation of 577. Sailing in the wake of the
great man, minor scholars endeavoured to rebut Mau’s arguments with
futile sophisms.? Others tried to blunt the edge of the deletion by only partly
accepting it. The latest commentator on the play, Denniston, has not re-
examined the arguments of Mau, whose article he mentions only as used by
Tucker and Keene. It will therefore be necessary first briefly to recall the
most relevant points of Mau’s analysis and then to carry the investigation a
step further. In dealing with Mau’s arguments, I am taking advantage of
the careful summary given by C. H. Keene, The Electra of Eur. (1893), 146 f.,
without, however, confining myself to it. ‘The passage occurs at a critical
point in the action when the spectators know that the ἀναγνώρισις is impen-
ding, for they have been told at lines 285-287 that this old man alone is in a
position to recognize Orestes. It is hard to suppose that at the moment of the
1 Hermes, xviii, 1883, 236 n. 2 (cf. also p. 224 n. 2): ‘Electra’s would-be-wise criticism of
the suggestions made by the Old Man (representing the 'recognition signs' of Aeschylus'
version) is not only not superfluous, but it is precisely for the sake of this criticism that
Euripides made the Old Man go to the tomb. There is also an indication in what follows.
In 577 Electra says, after recognizing the scar (which Euripides has borrowed from the
Odyssey by way of substitute for the ἀπίθανα of Aeschylus' story), συμβόλοισι yàp τοῖς σοῖς
πέπεισμαι θυμόν. Do the σύμβολα consist only of the scar? If so, the word would be τοῖσδε,
The fact is that she is now admitting the truth of the conclusions of the Paidagogos which
she had rejected before. So we cannot think of deleting the lines, much though we should
wish for Euripides’ own sake that he had not been playing the role of Zoilo-Thersites.'
(With the last remark cf. Wilamowitz's commentary on A. Cho., p. 169: “The sophist
[Euripides] has only given himself away by his criticism of Aeschylus' devices : his behaviour
is like that of Voltaire with regard to the Merope of Maffei, or rather like that of his own
caricature in the Frogs.) Of the reason why ‘Euripides made the Old Man go to the tomb’
we shall speak presently. In 577 f. the words συμβόλοισι rois σοῖς refer, of course, to the
scar and nothing else. To say that τοῖσδε would be required is quite arbitrary. The words
συμβόλοισι yàp τοῖς σοῖς πέπεισμαι θυμόν unmistakably hark back to Electra's question (572)
ποῖον χαρακτῆρ᾽ εἰσιδών, ὧι πείσομαι; to which the answer is: οὐλὴν παρ᾽ ὀφρύν κτλ. There is
no difference of meaning between χαρακτήρ in the former line and σύμβολα in the later.
Wilamowitz never changed his mind about the genuineness of the lines rejected by Mau.
In his Erinnerungen, published in 1928, he alluded to the old controversy with a facetious
remark (p. 145 f.): ‘August Mau ... A pupil of Ribbeck’s, he obelized quite innocent lines in
the texts of several poets (as late as the Commentationes Mommsenianae); a shy man’ etc.
‘Innocent lines’ (unschuldige Verse) indeed!
2 Cf., e.g., R. Wolterstorff, Sophoclis et Euripidis Electrae quo ordine sint. compositae,
diss. Jena 1891, p. 31 n. 1.
3 I have not been able to see the book which Radermacher (see below) quotes: F. W.
Schmidt, Kritische Studien zu den griechischen Dramatikern, where ll. 532-44 are bracketed.
Radermacher, Rhein. Mus. lviii, 1903, 546 ff., agrees with Schmidt as to the extent of the
interpolation while differing from him concerning the explanation. This attempt is far
from convincing, but may at any rate be taken seriously. The same could not be said of the
amusing παλινωιδία which Radermacher, inspired by St. Barbara, afterwards sang (Zeitschr.
f d. österr. Gymnasien, Ixvi, 1915, x f£). T. G. Tucker (Introduction to his edition of the
Choephoroe, p. lxxi f.) accepted Mau's result in the main but spoilt it by confining the inter-
polation to 520-3 and 525-44. Pohlenz, D. griech. Tragódie, i. 328, Solmsen, Hermes, lxix,
1934, 395, H. D. F. Kitto, Greek Tragedy, 342 f., and others deal with the scene as though
everything were in perfect order.
822
 APPENDIX Ὁ
greatest suspense so silly a discussion would be allowed to interrupt the course
of a scene otherwise skilfully composed’ (Mau, 297). While there are possible
reasons for the first two suggestions of the Old Man (for a lock has been found
on the tomb, and footprints might conceivably be discovered there), the
following question about the robe is in this context quite objectless (ibid.).
It may be added that the contents of the question (538-40), anticipating a
mere eventuality and discussing possible means of meeting it, seem childish.
Then there is the absurdity of 532 f. (cf. Mau, ibid.). No one who hears the
words σὺ δ᾽ eis ἔχνος βᾶσα κτλ. can help asking: ‘But how? Where and when?’
The spectators (and the readers of the play) have been told (210) that the
locality where the action takes place is at a considerable distance from the
city of Argos; there can be no idea of the house of Electra and her husband
being near the tomb of Agamemnon. The suggestion made in 532 f. does not,
in the context of this scene, make sense at all.“ Mau also points to the
strangeness of the Old Man’s suggesting several doubtful experiments instead
of inquiring without delay after the news which, as he knows, the messengers
have brought from Orestes. On 526 Patin, quoted by H. Weil (Sept iragedies
d’Euripide) in his note on the passage, remarked: ‘Electre dit qu’Oreste a
trop de coeur pour cacher son retour dans sa patrie par crainte d’un Egisthe.
Or, cette timidité qui l'indigne, Euripide l'a précisément attribuée à Oreste,
qui, chez lui, ne visite que de nuit le tombeau de son pére, ne se fait pas
connaître, méme à sa sœur, et a bien soin de se tenir, en cas de besoin, à
portée de la frontiére.' The same criticism was made by Mau (298). If it was
necessary for Euripides to make Orestes act in this clandestine way, he would
certainly not have made Electra say that secrecy suggested failure of courage.
Finally Mau (298) draws attention to 'the dry and pedantic manner in which
the matter is discussed point by point'. In regard to this, it should also be
noticed that the lack of inventiveness shown by using twice within a few
lines one and the same argument ('difference between male and female',
528 f. and 537) is very different from what we are entitled to expect of the
mature art of Euripides, a past-master in skilful variation. 520 σκέψαι δὲ
χαίτην προστιθεῖσα σῆι κόμηε: when we try to work it out, the idea proves as
grotesque as the conception of the model, Cho. 230 σκέψαι τομῆι (i.e. on the
head of Orestes) προσθεῖσα βόστρυχον τριχός, is natural and beautiful. Electra's
head is κούριμον (148) ; is she perhaps expected to use a mirror to perform the
act suggested in 520? In the preceding line the flatness of the thought μολὼν
δ᾽ ἐθαύμασ᾽ ἄθλιον τύμβον πατρός is most distressing. But there is no need to
dwell any longer on the details of a passage which, even in his weakest
moments, Euripides could not have written. Enough of this; paulo maiora
canamus.
There is indeed in the Electra a severe criticism of the Aeschylean ἀναγνώ-
piois, but it is not contained in the gibes of the interpolator. In the long
stichomythia between Orestes and his sister (220 ff.) the lines 282-7, or, at
any rate, 283-7, look almost like a digression. Their dramatic purpose, how-
ever, is obvious: they serve to prepare the hearer's mind for the recognition
scene. HA. ἀλλ᾽, ὦ ξέν᾽, οὐ γνοίην dv εἰσιδοῦσά vw. Op. νέα γάρ, οὐδὲν au”,
ἀπεζεύχθης νέου. HA. εἷς ἂν μόνος νιν τῶν ἐμῶν γνοίη φίλων. Op. dp’ ὃν λέγουσιν
1 Denniston's expedient ‘the sense is "go to the tomb and try the test", though, as
Tucker says, one misses “go to the tomb” ’ does not relieve my perplexity.
823
 APPENDIX D

αὐτὸν ἐκκλέψαι φόνου; HA. πατρός γε παιδαγωγὸς ἀρχαῖος γέρων. The most
remarkable detail in these lines is the emphasis which is laid on the impossi-
bility of Electra’s recognizing her brother since she was too young at the time
when he was taken away from his home.' This leads to Electra’s assertion
that no one but the old servant would be able to recognize Orestes. The
emphatic manner in which the statement οὐ γνοίην ἂν εἰσιδοῦσά νιν is supple-
mented by the words εἷς ἂν μόνος ww . . . γνοίη should not be underrated. The
implication must have been immediately obvious to an Athenian audience.
The ἀναγνώρισις in the Choephoroe, however excellent in detail, is, in the
opinion of Euripides, based on an assumption which is not in agreement with
the probabilities of ordinary life. But Euripides was not satisfied with merely
passing a negative judgement on the device of his great predecessor. His
criticism was meant to be constructive. In the passage of the play where it is
first uttered, it does not stand for its own sake, but points in advance to the
later scene where the recognition is brought about, not by Electra as in the
Choephoroe, but by the Old Man. In other words: it was not enough to
criticize Aeschylus; Euripides wanted to show by his own example how
the thing ought to have been done, Ina play written a few years after the
Electra, the Phoenissae, Euripides proceeded on strictly parallel lines. The
remark of Eteocles (751 f.) ὄνομα δ᾽ ἑκάστου (viz. of the Theban commanders
who are in charge of the seven gates and their opponents) διατριβὴ πολλὴ
Aéyew,? ἐχθρῶν ὑπ᾽ αὐτοῖς τείχεσιν καθημένων is doubtless a reflexion on the
famous scene in The Seven against Thebes. It is not to the enumeration of the
names, at any rate not to that of the names of the aggressors, that Euripides
objects, but to the inappropriate moment at which it is inserted in the
Aeschylean tragedy. He himself, while forbearing to give a list of the de-
fenders at all, provides a survey of the Argive commanders in the lively scene
after the prologue, the τειχοσκοπία, where the description is delivered by a
non-combatant, and (if the section 1104-40 was really written by Euripides)
another one, after victory has been won, in the account of the messenger.
To return to the Electra. The implicit promise of 285 ff. is fulfilled in the
anagnorisis scene. In this the Old Man's part is not a subordinate one. He,
and he alone, brings about the recognition. This solution does not come as
a surprise or appear as a mere accident. It is preceded by a psychological
preparation which is as well designed, if not as original, rich, and touching,
as the movement that carries the Aeschylean Electra up to the climax where
she will be in the right mood to recognize her brother. The discovery of
the offerings on the tomb has filled the heart of the faithful old servant with
a sudden hope which he hardly dares own to himself (515 f.). So strong is his
emotion that he weeps (sor ff.). In this detail, too, it is manifest that Euri-
1 As far as this argument, which is consistent in itself, is concerned, it does not matter
what was the poet’s idea as to the time when Electra was born (on this problem cf. Dennis-
ton, Introd. p. xxvi n. τ). Lines τό f. do not make it clear when, according to Euripides, the
rescuing of Orestes took place. For the problem in general and the earlier traditions cf. my
note on Ag. 886.
2 Wecklein, A. C. Pearson, and J. U. Powell prefer the reading of the Marcianus διατριβὴν
πολλὴν ἔχει, but this is a harsh construction since it seems doubtful whether ὄνομα can be
regarded as equivalent to an infinitive.
3 This, I think, is the reason why he is mentioned in the prologue (16) among persons of
much greater distinction.
824
 APPENDIX D
pides, for the purposes of the recognition scene, deliberately assigned to the
Old Man the part played by Electra in the Choephoroe and that he planned
his own scene after the model of the scene of Aeschylus: cf. Cho. 185 ff. But
while the dramatic function of the Old Man is analogous to, and dependent
on, that of the Aeschylean Electra, his character as παιδαγωγός is closely
related to another figure of the Choephoroe. The nurse of Orestes, who saved
his life, had a place in the story as told by Stesichorus. Aeschylus, though he
severed her connexion with the escape of Orestes, took advantage of her
existence to create one of his most accomplished scenes. We are not con-
cerned here with the problem whether or no Euripides when he made the
γεραιὸς τροφεύς of Orestes (and of his father) the rescuer of the boy was in-
fluenced by Sophocles.' What matters for the present investigation is the
undoubted fact that Euripides borrowed some details from the quite different
scene of the nurse in the Choephoroe and thus intensified the pathos of the
Old Man’s attitude in the recognition scene. It is sufficient to compare El.
507 ἀνόνητ᾽ ἔθρεψας and the responding ἀνόνητα (508) with Cho. 752 f. ἀνωφέλητ'
ἐμοὶ τλάσηι, and El. 508 ὅμως δ᾽ οὖν τοῦτό γ᾽ οὐκ ἠνεσχόμην with Cho. 747 ἀλλ᾽
οὔτι πω τοιόνδε πῆμ᾽ aveoxöumv.?
As we see, Euripides has done everything to represent the Old Man in a
state of mental readiness for the acknowledgement of the truth as soon as it
becomes apparent. The visible sign of Orestes’ identity (572 ff.), derived from
the Odyssey, is but the ultimate confirmation of a hope which, though with
doubts and hesitations, the Old Man has been nourishing for some time. That
token is done with as quickly as the corresponding token in the Choephoroe
(231 1). There is no gap between the psychological preparation and the
execution of the ἀναγνώρισις ; neither is there room for any delaying dis-
cussion. This observation tells very strongly against the genuineness of
518-44. Still more important is the conclusion which we are at last in a
position to draw from our interpretation of the scene and the place held by
it in the plan of the whole play.
Euripides emphatically insists on the impossibility of Orestes being recog-
nized by Electra. He employs all his skill as a dramatist to render the Old
Man a central figure in the recognition and lead him through deep emotions
to the final climax. It seems unbelievable that the same Euripides should
have been mad enough not only to spoil, but cornpletely to destroy, the fine
texture of his own creation. For it would be nothing less than the destruction
of acareful dramatic plan if, in the midst of a process which tends towards the
recognition by the Old Man, three suggestions should be made which are
based on the assumption that the obvious person to effect the recognition is
Electra. In 518-44 the very thing that Euripides is most anxious to reject is
regarded as the natural solution. The criticism in those lines is directed
against the usage of certain τεκμήρια, the criticism of Euripides is directed
against the employment of a person unsuitable for the task of recognizing
Orestes.
! Tam convinced that the Electra of Sophocles is the earlier play.
2 None of the commentators on Euripides’ Electra (not even Keene, who on 1. 506 f.
quotes A. Cho. 750 ff.) nor E. Bruhn, who in the introduction of his edition of Sophocles’
Elecira (1912), p. 27, collects what he regards as reminiscences from the Oresteia in the
Electra of Euripides, has noticed this striking similarity, which, incidentally, is fatal to the
conjectures put forward for the end of El. 508.
825
 APPENDIX Ὁ
The puzzle of 545 f. cannot be completely solved. This is disappointing, but
the decision of the major issue does not depend upon it. With Mau (300),
Wilamowitz (Hermes, xviii, 1883, 236. 2), Murray, Denniston, and others, I
regard it as certain that the lines must neither be deleted nor removed from
their place. Mau, who rightly took 545 f. to be the continuation of the Old
Man’s speech (attaching to 5ı7), surmised with a high degree of probability
that after 546 a line containing a main verb has dropped out (cf. also Dennis-
ton’s note). His suggestion that the lost line contained something like ἔμολεν
"Opéargs κτλ. is attractive though not provable. We have to admit the
possibility that the Old Man may in a less direct form have hinted at what he
hopes to be the true alternative.' After uttering his guesses as to the person
who put the lock on the tomb he breaks off: οἱ δὲ ξένοι ποῦ ;
We cannot say by what motives the man who inserted 518-44 was actuated.
Was he a later producer of the play, who hoped to amuse his audience by
poking fun at Aeschylus and performing a mountebank’s tricks with the
cheap juggling of a vulgarized πιθανότης at a time when many serious poets
and critics were engaged in refining the system of ἀναγνωρίσεις and estab-
lishing the relative values of the various types? Whatever his motives were,
he had before him the scene of the Choephoroe in its interpolated form.

APPENDIX E
Short Syllables before Initial Mute and Liquid in the Lyrics of
Aeschylus
Cho. 606 πυρδαῆ τινα πρόνοιαν (cf. my remarks p. 722 n. 2)is, unless I have missed
an instance, the only case of the lengthening of a final short vowel before
mute and liquid in the lyrics of Aeschylus, with the one exception of Eum.
378 ἐπὶ κνέφας in dactyls (in the dactyls of this chorus there are also a few
other prosodic Homerisms: 349 μιν or ἁμὶν, 387 δύσοδοπαίπαλα) ; Pers. 664 f.
is corrupt, Prom. 582 emended by Elmsley. Aeschylus does not admıt the
lengthening caused by initial mute and liquid even in the case of ‘prae-
positiva’; in this he differs from the two other tragedians, e.g. S. Ant. 612
τὸ πρίν, E. Hipp. 837 6 τλάμων, 5. Aj. 1220 ὑπὸ πλάκα, Trach, 1012 (hexameter)
κατά τε δρία, E. Alc. τοι ἐπὶ προθύροις, Heraciid. 753 παρὰ θρόνον. Neither do
the anapaests of Aeschylus provide an example: in fr. 96 Nauck wrongly
writes μήτε κρωσσοὺς, whereas Bethe (in his edition of Pollux, 6. 23) rightly
adopted μήποτε from one section of the MSS; in the verse of the diaskeuast
Sept. 1056 γένος ὠλέσατε πρυμνόθεν οὕτως there is a choice between a prosodic
abnormality (judging by the standard of Aeschylean prosody) and writing
with Elmsley #Adooare (which is not completely justified by the ὀλέσσας
required in S. Aj. 390—cf. Wilamowitz, Verskunst, 506—, for oAeoo- is
Homeric, ὠλεσσ- is not, cf. Lobel, Ἀλκαίου μέλη liii ; ὠλέσσατε would give the
form of the anapaestic metron found in A. Suppl. 6, Cho. 401, Eum. 949). But
although this lengthening in τινα πρόν. has no parallel in what is preserved
1 The mutilation of the sentence beginning with ἢ τῆσδε σκοπούς may perhaps not be due
to mechanical destruction or accidental loss but to a deliberate attempt of the interpolator
to cut out what proved untenable after the addition of 518 f.
826
 APPENDIX E
of Aeschylus, it would be unwise to tamper with words which are irreproach-
able in metre, language, and content. Perhaps πυρδαῆ τινα πρόνοιαν was felt
as an expression which was closely bound together. E. Tro. 833 (P. Maas gave
me this reference) is in some degree comparable: ra δὲ σὰ δροσόεντα λουτρά
(vuv-uu-u-u) There may be other similar instances in Tragedy ; I have
not searched the lyrical parts of Sophocles and Euripides. We lack a presenta-
tion of the whole material founded on careful textual criticism. The mono-
graphs ‘De correptione Attica’ (the older ones are mentioned in Joh. Schade,
De corr. Ait., diss. Greifswald 1908) limit themselves almost entirely to the
dialogue (Schade himself gives only selected specimens from lyrics, p. 48 f.,
in A. Suppl. and Eum.). A collection of passages which needs both completion
and revision was given in 1811 by A. Seidler, De vers. dochm. 21 f. At the
moment it is impossible to get beyond embarrassing questions. Is, e.g., E.
Hipp. 759 the reading of one group of the MSS ἔπτατο κλεινὰς credible? Or
S. El. 853 ἃ θροεῖς, on which so much has been written? (Phil. 209 need not be
discussed here. It is true that Wilamowitz, Verskunst, 533, puts with Tric-
linius διάσημα θροεῖ ydp in the text to correspond to προβοᾶι γάρ τι δεινόν, so
also A. C. Pearson; but A. M. Dale in Greek Poetry and Life, 1936, 198 ff. has
shown that it need not be considered.) Are we to assume that in S. Oed. C.
180 ἔτι ;---προβίβαζε κούρα syllaba anceps comes in where the person changes?
On the MS reading στολίδα κροκόεσσαν E. Phoen. 1491 (dactyls) see p. 138 n. 1
above. On E. Iph. T. 230 (‘threnodic’ anapaests) τὸν δ᾽ Apyeı δμαθέντα
κλαίω Hermann remarks ‘verbi δμαθέντα ultima syllaba ante κλαίω... in
medio versu vix ac ne vix quidem produci potuit’, Weil suggested ἀγκλαίω.
Another doubtful instance is E. Bacch. 1021 γελῶντι προσώπωι (Murray sug-
gests προσώπωι γελῶντι, which seems to deteriorate the word order): the
metre of this epode is altogether far from clear.

APPENDIX F
The Word-order in Ag. 1434 οὔ μοι φόβου μέλαθρον ἐλπὶς ἐμπατεῖ
Ir we wanted to take φόβου ἐλπίς together, we should (in addition to the diffi-
culty discussed ad loc.) have to face also a more general objection. Aeschylus
would hardly allow a hyperbaton (‘Sperrung’) such as φόβου... ἐλπίς to stand
with its parts divided only by an accusative depending on the main verb.
Where a word-group consisting of a genitive and (placed after it) its governing
noun is divided, the intervening element belongs to one of the following
syntactical types (my examination is here limited to the dialogue of Aeschylus).
Much the most common is the simple verb or verb and predicate, as in πόνων
ἐμῶν ἥκεις ἐπόπτης or ἔτ᾽ dp’ ᾿Αθηνῶν ἔστ᾽ ἀπόρθητος πόλις ; etc. ; examples need
not be multiplied. To this type belong also the simple substantival predicate
(‘without copula’) as in Prom. 34 Διὸς yap δυσπαραίτητοι φρένες and the
similar, but slightly more complicated, form (Ag. 1239) καὶ τῶνδ᾽ ὁμοῖον εἴ τι
μὴ πείθω. Several times the intervening element consists of the verb with an
adverb or adverbial phrase : Suppl. 325 πόνου δ᾽ ἴδοις ἂν οὐδαμοῦ ταὐτὸν πτερόν,
Sept. 34 f. μηδ᾽ ἐπηλύδων ταρβεῖτ᾽ ἄγαν ὅμιλον, Prom. 632 τὴν τῆσδε πρῶτον
ἱστορήσωμεν νόσον, Cho. 240 f. τὸ μητρὸς ἐς σέ μοι ῥέπει στέργηθρον, 992 τέκνων
827
 APPENDIX F
ἤνεγχ᾽ ὑπὸ ζώνην βάρος, 1047 δυοῖν δρακόντοιν εὐπετῶς τεμὼν κάρα. So also verb
with subject: Ag. 267 Πριάμου γὰρ ἡιρήκασιν ᾿Αργεῖοι πόλιν, 1246 Ayauéuvovés
σέ φημ᾽ ἐπόψεσθαι μόρον (with oe in the second place in accordance with
Wackernagel's law), Zum. 640 πατρὸς προτιμᾶι Ζεὺς μόρον, fr. 266. 5 N. καὶ τοῦ
θανόντος ἡ Δίκη πράσσει κότον. Verb with pronominal object: fr. 243 νέας
γυναικὸς où με μὴ λάθηι φλέγων ὀφθαλμός. Often a simple adverb: Sept. 655
πατρὸς δὴ νῦν ἀραὶ τελεσφόροι, Co. 177 μῶν οὖν 'Opéarov κρύβδα δῶρον ἦν τόδε
(‘ κρύβδα ἐδωρήσατο, ist gedacht: wäre die blosse Copula gemeint, so würde
κρύφιον δῶρόν ἐστιν stehn’ Wilamowitz), 502 οἴκτιρε θῆλυν dpoevós θ᾽ ὁμοῦ
γόνον, probably also 883 f. (details of the text uncertain), Eum. 624 μητρὸς
μηδαμοῦ τιμὰς νέμειν. The simple subject can come between if it is a pronoun:
Pers. 598 κακῶν μὲν ὅστις ἔμπειρος κυρεῖ (where, however, the quasi-verbal
character of ἔμπειρος puts the example in a slightly different category), Prom.
823 τὸ πᾶν πορείας ἥδε τέρμ᾽ ἀκήκοεν (where the initial position of τὸ πᾶν, the
attribute to τέρμα, makes the order more natural), Cho. 21 γυναικῶν ἥτις Ne
προστροπή. (Pronoun object in, e.g., S. Ant. 908 τίνος νόμου δὴ ταῦτα πρὸς χάριν
λέγω ;). Eum. 636 ἀνδρὸς μὲν ὑμῖν οὗτος εἴρηται μόρος is easily intelligible : here
the intervening dative of the person interested has the form of ἃ pronoun
which in accordance with Wackernagel’s observations tends to come as near
as possible to the beginning of the sentence (cf., e.g., the usual formula in
law-court speeches kai τούτων ὑμῖν μάρτυρας παρέξομαι). In Prom. 612 πυρὸς
βροτοῖς dornpa there is no hyperbaton, since βροτοῖς δοτῆρα form a single
grammatical unit. In two instances closely related to one another, the
intervening word looks like an adjectival attribute: Prom. 448 ff. ὀνειράτων
ἀλίγκιοι pophaîor . . . ἔφυρον, Ag. 1218 (ὁρᾶτε τούσδε roûs . . . νέους,) ὀνείρων
προσφερεῖς μορφώμασιν, where the adjectives are inserted just as if they were
actual participles (ἐοικότες or ἐοικότας respectively). There is only one in-
stance in Aeschylus which defies classification: Eum. 50 f. εἶδόν wor’ ἤδη
Φινέως γεγραμμένας δεῖπνον φερούσας. The position οἱ γεγραμμένας here is
highly unusual, whether γεγραμμένας itself is regarded as the object (so, e.g.,
Wecklein: ‘painted female forms’) or whether (as seems more natural) the
object is ‘female monsters’. Perhaps the order Φινέως γεγραμμ., as Miss Alford
suggests, is due to the fact that the idea ‘in a painting representing the story
of Phineus’ is uppermost in the speaker’s mind. However, I do not feel
absolutely certain that this is what Aeschylus wrote. Hermann (cf. Opusc.
vi. 2. 20), and others before him, presumed a lacuna in front of this line,
without, however, objecting to the order ®. yeyp. ὃ. φ.; W. Dindorf, πολυ-
πράγμων and over-hasty worker, but a man who knew his Greek, assumed a
twofold omission : εἶδόν ποτ᾽ ἤδη... γεγραμμένας... Φινέως δεῖπνον φερούσας,
with the comment ‘verbo γεγραμμένας, quod nunc loco minus apto collocatum
est, in suam sedem revocato’.

828
 ADDENDA ET CORRIGENDA TO COMMENTARY
P. 43, bottom. Cf. Bacchyl. 5. 94 ff. χαλεπὸν θεῶν παρατρέψαι νόον ἄνδρεσσιν
ἐπιχθονίοις.
P. 58 n. x. According to Latte, Philol. xcvii, 1948, 47 ff., the play of Pap.
Oxy. 2164 is not the Zavrpuaı.
P.75n.2. Among the compounds the first element of which shows 'productio
epica’ I ought to have mentioned Pers. 80, 856 ioödeos, Prom. 548 ἰσόνειρον.
P. go n. 2. That ἀλέκτωρ in Bacchyl. 4. 8, whatever it means, cannot mean
‘husband’ is now made certain by the scrap of papyrus which was recently
discovered and published by Medea Norsa (for a discussion of the details see
Gallavotti, Revista di filol. class. xxii-xxiii, 1944-5, 1 f.).
P. 149. Without assuming in 265 εὐφρόνης an allusion to 263 εὔφρων, I might
have admitted the possibility that Aeschylus chose the word εὐφρόνη here
because its beginning with ev- made it boni ominis.
P. 174, L 12 from the bottom: ἀδικέαν εὐδαίμονα. The passage is E. Phoen. 549.
P. 185 n. 1. For the metrical analysis of the parodos of the Prometheus see
G. Hermann, Elem. doctr. metr. 492 f., with whom I agree.
P. 201, bottom (on I. 387). For the assimilation of a final v before an initial
labial and its treatment in the doctrine of the grammarians and in the
practice of the inscriptions and papyri cf. Wendel, RE xviii, 1455.
P. 241 n. x. For this Hellenistic poem (Pap. Tebt. 1. 5-ır, vol. i, p. 3) see also
J. U. Powell, Collectanea Alexandrina, p. 185 f.
P. 258 (on l. 505). As W. S. Barrett observes, I ought not to have referred to
Pearson on E. Hel. 277. Pearson’s main piece of evidence, Ar. Knights 1244,
has nothing to do with an anchor: see the commentaries of Casaubon (in
Küster’s edition), Kock, and Neil (not invalidated by Pearson on Soph. fr.
685 or by England on Pl. Laws 699 b).
Pp. 275 f. (on A. Suppl. 288 ἦτε). During a short stay at Florence in September
1948 I examined, together with Signorina Teresa Lodi and Giorgio Pasquali,
the passage in the Mediceus. The τ of ἦτε is written over an erasure; it is
certain that the first scribe wrote or (in ligature) : the circle of the a is clearly
visible.
P. 280, end of the first paragraph. Cf. Servius on Virgil, Aen. i. 177 rem vilem
auxit honestate sermonis.
Pp. 283 ff. (on 1. 562 τιθέντες). R. Pfeiffer remarks that I might have men-
tioned Theocr. 15. 119 although it is later than my other instances. There the
medieval MSS have oxidSes . . . βρίθοντες, whereas the Antinoë papyrus has
βριθουσαι, ‘eine fast zu einfache Lösung, um Glauben zu finden’ (Pohlenz,
Gott. gel. Anz. 1931, 367).
P. 284 n. 2. To the instances of postponed & Pfeiffer adds Callim. hymn. Del.
325, Aet. iv (from Dieg. v. 3) ἥρως ὦ κατὰ πρύμναν.
P. 295 (on 1. 590). ‘The accent φρυκτωρός is attested by Theognostos, Anecd.
Oxon. ii. 72. 4’ (Pfeiffer).
829
 ADDENDA ET CORRIGENDA
Pp. 361 f. The idea of the king as shepherd is common in Babylonian,
Assyrian, etc., texts; see C. J. Gadd, ‘Ideas of Divine Rule in the Ancient
East’, The Schweich Lectures of the Brit. Acad., 1945 (1948), 38 f.
P. 387, top. To the instances of λέγοιμ᾽ ἂν ἤδη Barrett adds E. Or. 640 (I.
640 f. are genuine, cf. Wilamowitz, Hermes, lix, 1924, 250); cf. Iph. T. 939
λέγοιμ᾽ ἄν.
P. 404 (on Il. 892 f.). A mistake in my interpretation has been corrected by
Barrett. He writes: ‘I am quite convinced by your main point: that θωύσσειν
can only be used of a loud noise, and that the apparent contradiction with
λεπταῖς ῥιπαῖσι is due to a shift in Clytemnestra’s thought. But I find it
difficult to believe in two different noises, the whir of the wings (A. p.) and
the buzz (8.).... May it not all be one and the same noise, the buzz, described
in two different ways? Properly speaking it is a trivial noise, that has no
business to wake a healthy sleeper, and so at first she describes it : λεπταῖς...
ῥιπαῖσι. But once she has said ἐξεγειρόμην she thinks of the noise as it sounds
magnified when one is restless and cannot sleep; and so θωύσσοντος.
peat I should take not as “wing-beats’’ (can it mean that without πτερύγων?)
but in the more general sense of "rush", “quick movement”: i.e. the zig-
zagging of the gnat to and fro about the room.’
P. 4x5 n. r. I take the opportunity of correcting the misplacement of another
Aeschylean fragment, viz. the line which is universally printed in the form
βριθὺς ὁπλιτοπάλας, δάϊος ἀντιπάλοις and supposed to come from an elegy or an
epigram (cf. Bergk, Poet. lyr. ii, p. 242, Diehl, Anthol. lyr. 12. 1. 80). But in
two of the three places where Plutarch quotes the complete line (the third
place, Cic. 51, shows a slight adaptation) the MSS have ἀντιπάλοισι; moreover
Eustathius quotes the last word as ἀντιμάχοισι. The idea of turning the
fragment into a pentameter goes back to H. Stephanus’ edition (1572) of
Plutarch, and this change has been accepted in all subsequent texts of
Plutarch and in the editions of Aeschylus from Stanley to the present day.
But neither the manner in which Plutarch introduces the quotation nor the
form of the ending of ὁπλιτοπάλας (the first vowel of δάιος provides no argu-
ment, cf: on Ag. 559) are in favour of Stephanus’s assumption. The fragment
is obviously a piece of a choral ode. For the metre cf. Eum. 963 f. δαίμονες
ὀρθονόμοι, παντὶ δόμωι μετάκοινοι. [My observation had been anticipated by
Wilamowitz, Textgesch. d. Lyr. 6o n. 1.]
P. 418 (on 1. 926). I ought to have said that this scholion is not an old one,
but Triclinian (juér.) In F there is the simple gloss 8¢ ὧν of πόδες δηλονότι
καλλωπίζονται.
Pp. 430 f. For the interpretation of Agamemnon’s behaviour which I have
attempted to refute cf. also Headlam in Praelections delivered before the
Senate of the Univ. of Cambridge, 1906, 129 ff.
Pp. 443 f. Is the possibility of a syllable before uv being treated as short
perhaps due to the same phonetic tendency which has often led to the simpli-
fication of py to u? Many epigraphic instances of this simplification are
collected by Nachmanson, Giofía, iv, 1913, 246 ff.
P. 497 n. 1. Pfeiffer refers to Pap. Oxy. 2080 (saec. II). 39, 86, where we find
the same practice of putting dots before and after the variants above the line.
830
 ADDENDA ET CORRIGENDA

P. 541 n. 1 (cf. also p. 297 n. 3 and p. 779 n. 3). 'Bedeutungslehnwort', semantic

loan-word, is the result of what the French philologists call ‘un emprunt
de sens’ or ‘un calque sémantique’.
P. 549 n. 2. My criticism of Birklein’s lists is wrong ; it is the result of a con-
fusion in the rewriting of the footnote.
P. 581 (on I. 1256). For fire appearing to a medium in trance see also Dodds,
Journ. Rom. Stud. xxxvii, 1947, 68.
P. 583 (on l. 1263). W. S. Barrett observes: ‘I feel that everything between
κτενεῖ με τὴν τάλαιναν and ri δῆτα (why, then, since I am about to be killed”)
must be referring in some sort to Cassandra’s death. The μισθός (which I
think is a payment made by, not io, Clytemnestra) seems, I agree, to be
ambiguous, between wages paid to Agamemnon for his adultery and wages
paid to Cassandra for her part in it. May there not be a similar ambiguity,
or indefiniteness, in 1263? Cassandra deliberately omits after ἀντιτείσεσθαι
the accusative of the person from whom the penalty is exacted, so leaving the
victim inexplicit ; ἐμῆς ἀγωγῆς takes the emphasis, and means not simply
“his bringing me” (as it would if only Agamemnon’s φόνος were in question)
but more generally “the fact that I have been brought”. So that the words
admit (without making explicit) the inclusion of Cassandra with Agamemnon
as a victim of the φόνος (just as she was involved with him in the aywyn).’
I have tried to translate accordingly.
P. 634, second paragraph. Mr. D. Mervyn Jones, who for some time has been
working on the Aristophanes scholia, kindly informed me (October 1948) that
the two passages in schol. Knights 589 which, following Dindorf, I regard as a
late insertion and therefore have bracketed ‘are interpolated by Musurus:
they do not appear in any MS containing old scholia’.
P. 636, top. Cf. Wilamowitz, Griech. Tragoedien, i. 188, on E. Hipp. 659:
‘Hermann hat ἔκδημος in ἐκδημῆι verändert, weil ihm die Auslassung der
Copula im Konjunktiv anstössig war. Er wird recht behalten’, etc.
P. 724 (on 1. 1527). For digna dignis see F. Bücheler, Rh. Mus. xlvi, 1891, 243
(Kl. Schriften, iii. 207).
P. 732 (on li. 1545 f.). I owe it to W. 5. Barrett that I now understand ἀντ᾽
ἔργων μεγάλων correctly. A right translation was given by, e.g., H. Voss,
F. Bellotti (Milan 1821), Kennedy, Browning, Wilamowitz, G. Thomson,
A. Y. Campbell; cf. also Karsten (p. 308) and Plüss ad loc. The same point is
made in the archaic epitaph from Troezen, IG iv. 800 (Geficken, Griech.
Epigramme, no. 58; P. Friedlander and Hoffleit, Epigrammata, no. 29), τοῦτο
δ᾽ ἑταῖροι σᾶμα xéav βαρέα areváyovret ρέργων ἀντ᾽ ἀγαθῶν. Cf. the well-known
epigram on the victors of Eion ἡγεμόνεσσι δὲ μισθὸν ᾿Αθηναῖοι τάδ᾽ ἔδωκαν ἀντ᾽
εὐεργεσίης καὶ μεγάλης ἀρετῆς (or μεγάλων dyadav?—see F. Jacoby, Hesperia,
xiv, 1945, 197). For the later history of this topic see Friedlander, op. cit.,
Pp. 79 (on no. 71).
Pp. 736 f. (on 1. 1563). For μένειν as an essential feature of the highest god
see W. Jaeger, The Theology of the Early Greek Philosophers (1947), 141 and
especially 237 n. 59.
P. 783 n. x. The tomb-stele from the Kerameikos is figured also in Gisela
M. A. Richter, Archaic Attic Gravestones (1944), fig. 55.
831
 ADDENDA ET CORRIGENDA

Pp. 791 f. (epexegetic infinitive added to πολλά). Barrett adduces Thuc.

6. 34. 4 TO δὲ πέλαγος αὐτοῖς πολὺ περαιοῦσθαι κτλ.
P. 809 (end of Appendix B). E. Kunze, Neue Meisterwerke griechischer Kunst
aus Olymbia (Munich, 1948), 22 f., has seen that the very impressive bronze
relief (second quarter of the 6th century B.c.) which he figures on pl. 46
represents the murder of Agamemnon. There the weapon with which Clyte-
mnestra kills the king is clearly a sword.
Pp. 809 ff. (Appendices C and D). I have not been able to see M. Valgimigli's
edition of the Choephoroe (Bari 1926).

P. 11 f. (on 1. 12). I ought to have mentioned, and perhaps accepted, Weil's

conjecture (in his edition of 1858) ἔχων.
P. 277 (on 1. 549). In Eur. fr. 337 N. (Stob. Flor. 44. 2, vol. iv, p. 115 Hense)
the MSS have τυράννοις instead of κοιράνοις.
P. 490 f. (on 1. 1072 δᾶ). At Cho. 405 Bamberger's emendation ποποῖ δᾶ (ποῖ
ποῖ δὴ cod.) νερτέρων τυραννίδες is almost certainly right. Blass, it is true,
endeavours to justify δή: 'das δή knüpft die Worte an die des Chores an, u.
vgl. Pers. 1070 ἰωὰ δὴ κατ᾽ ἄστυ. But Pers. 1070 is preceded by 1069 io ἰώ
(cf. Denniston, Particles, 227), whereas the stanza Cho. 400-4 contains nothing
to which zozot δή could be ‘attached’. Accepting, then, the reading dä, we
may régard it as very unlikely that in front of νερτέρων τυραννίδες the meaning
of δᾶ should be 'O Earth!', for in that case we should expect a connective
particle.
P. 611, bottom (the bird afraid of an animal under the bush). I might have
referred to the poem Megara (in the collection of the Greek Bucolics), ll. 2x ff.

ADDENDUM TO VOL. I
Pp. 3of. (on the codex Venetus G). Either in the Prolegomena or in the
critical apparatus I ought to have said that at 12, 1153, and 1261, where in F
something is written above a letter or a word, G has only what in F is on
the line of the text, whereas at 1252 the ἡ above the « is found in G as
well. G has at 1226 ζυγὸν, 1324 τοῖς, 1348 ὑμῖν, 1379 ἐξειργασμένοις, 1414 οὐδὲν,
1451 ὕπνον, 1549 δακρύοις, 1599 ὥμωξεν: in none of these instances is there in
G a correction or a trace of a different reading.
[I use the empty space for a suggestion made by Beazley: 'I have been
thinking about Agamemnon's purple path. How can walking on the πετά-
σματα amount to δωματοφθορεῖν ἡ They are surely δευσοποιά, and will wash.
Perhaps one is to imagine such delicate fabrics that a foot may make a rent.
If so the word 'carpet', which suggests a fairly substantial fabric, should be
avoided.']

832
 Ι
GENERAL INDEX
Throughout the indexes numbers in heavy type indicate the pages of Vol. I.
Abresch: 44 f. [ἀγαθὴ τύχη : 348.
Accusative governed by the verbal element Agonarchs: 260.
of a noun: 494. ἀγώνιοι θεοί: 260 ff,
— of exclamation: 523 f. ἀγοραῖοι θεοί: 54.
Acheron, ancient etymology of: 734. Ahrens, H. L.: 54 ff.
Actor's reply echoing phrases of the Chorus: αἶνος, beginning of: 339.
4. αἰνόθεν αἰνῶς and the like: 125.
Adjectives, two weighty ones coupled : 376. Alcidamas : 386.
Adverb (or adverbial phrase) placed am- Alliteration : 149.
biguously : 132 f., 468 f., 542 f. ἀμηχανία in the thought of Aeschylus : 726,
Adverbial expression added as afterthought : 136.
590, 743» 775- ἀμοιβαῖα of Chorus and actor: 487 f.
Anddvy? : 723 n. 3. dv, position of: 422 f.
Aerope : 494 n. 1. — repeated : 176.
Aeschylus: see Antinomies, Archilochus, — with participle : 475 f.
Etymologies, Εὐμενίδες, Everyday life, Anacoluthon: 11, 221, 321, 445, 456, 587, 772.
Excess, Forensic formulae, Framing, ἀνάγκη leading to σωφρονεῖν: 108.
ypidos, Homer, Homeric customs, Irrele- Anapaestic metron, its end not coinciding
vant circumstances, Arai, Monologue, with the end of a word: 33.
‘Oresteia-words’, Prayer, Productio epica, Anapaestic piece sometimes substitute for
Repetition, Sicilian, Sophocles, Sticho- stasimon : 633.
mythia, Tenses, Tragic decision, Zeus. Anapaests: see Doricisms.
— Agamemnon: see Bath, Cassandra, — dialect: 701.
Chorus, Clytemnestra, Erinyes, Garment, — (‘epirrhematic’) without exact strophic
γερουσία, Herald, Hereditary curse, ὑπό- correspondence : 660, 716 n. 3.
θεσις, Net. — lacunae in text of: 38, 717.
— — Agamemnon, figure of: 119, 294, 372 f., ‘Anastrophe’ of ἀντι: 593.
425, 429 ff., 441 f. ἄνδρες used as address in Drama: 636.
— — choric odes: 356 f. Animals representing human beings in
— — guilt of Agamemnon: 98 f., 625, 726. prophecies and oracles: 69, 510 f.
— — Helen, figure of: 357. Antinomies in Aeschylus : 235, 725 f.
— — οἰκονομία: 50, 293 f., 685. Antistrophic stanzas, text of, harmonized
— — Paris, Agamemnon wronged by : 210. by copyists: 706.
— — weapon used to murder the king: Aorist in -@nv used with middle sense: 710.
806 ff. — participle : 207.
— contradicting his earlier religious belief? : — participles subordinated to verbs of
349. saying : 358 f.
— dramatic fragment not recognized as Aphaeresis of ἐν: 225, 553.
such : 830. ἀπὸ κοινοῦ : 205.
— influenced by Homer: 324, 499, 656. Apollo Ayvıevs : 491.
— 'Iduyévew. : 141 n. 3, 719 n. 5. — connected with Zeus: 259.
— knowledge of Oriental habits: 412. — interceding with Artemis: 87.
— lyrics, having syntactical breaks corre- — Avkeos: 581.
sponding in strophe and antistrophe : 75, ἀποπομπή Of a god or a daimon: 740 f.
122, 354 f. ἀποστρέψ- and ἀποτρέψ- confused : 388 f.
— Παλαμήδης : 387 n. 1. Apposition to the sentence : 29, 129, 706, 778 f.
— Prometheus, language of: 164, 608 n. 1. ἄπυρα ἱερά: 42 f.
— Πρωτεύς: 294, 313, 387. ἄρα, position of: 578.
— Ψυχοστασία : 229. Archilochus and Aeschylus : 50.
— Septem, large interpolations towards the Aristophanes of Byzantium : 83, 85.
end: 321 n. I, 610. Artemis ἀϊγροτέρα, sacrifice to: 133.
Afterthought added : 172, 309, 743, 775. — her wrath against Agamemnon: 97 f.
4872,3
833 Aa
 GENERAL INDEX
Artemisium, battle of: 116. Capture of Troy, month of: 280.
Article absent : 579 [cf. S. Ant. 1282]. Carpets, Oriental: 413.
— — from some items in a series of parallel Casaubon: 36 ff., 62 ff.
nouns : 680. Cassandra and Agamemnon : 433.
— added to only one of a pair: 168. — and Apollo: 554.
— ἀπὸ κοινοῦ: 419 f. — at what moment identified by the audi-
— deictic : 323. ence? : 432.
— force of: 128. — costume of: 584.
— in Attic literature : 657 n. 2, 750 n. 2. — death of: 685.
— not necessary with ἀνὴρ ὅδε etc. : 392. Casualty list, Athenian : 227.
— possessive use of: 106, 157. χιτὼν ἀτράχηλος : 649.
— separated from substantive: 481. Choirs singing alternating songs: 546 ἢ. 1.
ἀσάμινθοι : 648. Choreutae, number of: 633 ff.
Ashes of warriors brought home: 227. Chorus in Ag., an ‘accompanying instru-
Asklepios, story of, in Aeschylus: 463. ment’? : 643.
Assimilation in MSS of final consonant to — — divided ? : 661.
initial consonant : 201, 829. — — their share in the dramatic action:
Astrology, unknown to Aeschylus: 5. 247 ff., 784.
Asyndeton, explanatory : 433, 582, 597, 673. clamor armisonus : 30.
— in prayers: Lor. Clausula vu—-u-0 (où δὲ oóxd μ᾽ αἰτεῖς)
— of second item of a triad where the third concluding ionics: 59.
has connecting particle: 675. Clytemnestra, change in behaviour of: 678.
Athena [Tpovdia: 330 n. 2. — her position similar to that of a regent:
Athenaeus : 415 n. I, 754 n. 2. 145.
Athenian campaigns in the year 459: 227. Coins of Akragas: 70 f., 96.
Athos, Mount: 154. Colloquialisms sparingly admitted in
Atlas metope (Olympia): 373. Tragedy : 608.
Atridae living in the same house: 210. Colon (metrical), Ὁ — v — — between ionici a
— their dwelling-places in the earlier minore : 185 n. 1.
Stories: 210 n. 1. — coinciding with syntactical unit: 185,
Attic and non-Attic forms of word-ends in 192, 333, 704.
lyrics : 727 n. 3. Colon (syntactical), initial, serving as head-
Attic manners: 119, 225, 386 f., 476, 572 n. 3. line: 298, 431.
Auratus: 35. — self-contained : 512.
Colouring of metal: 304 f.
Bacchiac tetrameter : 488. Commentary in the modern sense: 38, 49.
Barbarians speak like birds: 477. comparalio paralaclica: 47, 172.
Bath, 'Homeric' details : 648. Comparison and thing compared inter-
Bath-tub, Homeric: 731. woven: 438, 456, 542.
'Bedeutungslehnwort' : 831. — in Aeschylus not self-contained : 39, 208,
Birds, liming and trapping of: 611 f.
Blasphemy : 653. 341, 456.
Compound adjectives suggested by per-
Blass, F.: 6 ff., 57. sonal names in Homer: 230.
Blomfield, C. 7. : 47. Compounds in dithyramb etc. : 75.
Bloodshed irreparable : 459. — voluminous, two side by side: 495.
bona verba indicating something disastrous : Confusion in MSS, a: and €: 790.
41, 523, 652 f. — — ém- and ὑπο-: 42.
Bosker, H. N.: 4oo n. 3. — — μέγα and μετα: 589.
βούλευσις : 763 f. — — ὥσπερ and ὥστε: 800.
βοῦς ἐπὶ γλώσσηι: 23.
Conington, J.: 51 f.
Bread of slavery : 472. Copulative inclusion : 262 f.
Brennan, C.: 6. Copyists trying to improve the construc-
Bronze coinage: 204. tion : 170.
Bronze, lead-bearing : 204 f. Corrections, metrical, before Triclinius: 10,
‘Brother of . . .', ‘sister of . . .', etc., said of I4 n. I.
inanimate objects : 251. correplio Attica in the lyrics of Tragedy:
Caesura (in iamb. trim.) cutting into syn- 826 f.
tactic unit : 605 n. 3. Correspondence between strophe and anti-
Campbell, Lewis: 87. strophe, words of equal length in corre-
Canter, W.: 35 f. sponding places: 192.

834
 GENERAL INDEX
Corruption in MSS: see Confusion, Copy- Dramatic technique: 305, 768 n. 3.
ists, δέ, ἐλθεῖν, Ending of verse, Gloss, ‘Drinking the water of .. .', to denote one's
γνώμη, lambic trimeter, Interpolation, homeland: 531.
Lines, Marginal entry, Monosyllable, δρόμος, περὶ ψυχῆς : 73.
Paragraphos, rıva, Transposition. Droysen, J. G.: 227, 602.
— — compounds split (καρδίᾳ δηκτόν for Dual (for the second person) of historic
καρδιόδηκτον): 376 n. 1. tenses: 556.
— — connecting particle wrongly inserted : — forms of neuters in -ka: 651.
76, 362 f., 688 n. 1. — of participles : 70.
— — xai dropped out before od: 788. — of the participle in Aeschylus: 171.
— — misreading combines with thought sug-
gested by context: 655 n. 1, 683 n. 3. Eagles feeding on hare: 96.
Cothurnus : 429. ἔχεις in a concluding sentence: 473.
Council of regency: see γερουσία. Egyptian hymns: 410, 412.
Crasis : 550 f. εἰ with subjunctive: 621.
Cry, double: 73. εἴη, ἔστω, at the beginning of a clause: 198.
Curiosity of modern reader which the poet εἰκάζειν, custom of: ror, 576, 773.
declines to satisfy: 141 n. 4, 210 n. 2, εἶπον at the end of a speech: 421.
255 f., 433, 720, 748 f., 751 f. εἴσθεσις διπλῆς : 20 f.
Curses : 254, 272. Ejaculations : 488 ff.
Cursing expressions used by tragic char- éx-, é£-, verb amplified by: 464 f.
acters : 585. ἐκδικεῖν, ἐκδίκησις etc., in scholia : 448 n. 2.
Cypria : 97 f., 332, 334, 554, 719 f. ékküxAmua: 644.
ἑκὼν ἐπιλήθομαι: 24.
Dactylic hemiepes in Aeschylus: 702 1. ‘Ellipse’ of εἰμί, ἐστί: 367 f.
Dactylic word in first foot of trimeter: 7 f. Elmsley : 47.
Dactyls, lyric: 57 f., 89. ἐλθεῖν, forms of, intruded in MSS : 259.
δαίμονες ἀντήλιοι: 264 f. Enallage: 257 f., 714.
δαίμων used like πότμος or τύχη : 632 f. Ending of verse altered by intrusion from
Dative, instrumental, of persons: 307. neighbourhood : 559.
δέ, in the third place : 301, 323. Enjambement (in lyrics), overlapping of
— placed after preposition and article : 776. sentence into new stanza: 135 f.
— position of, normalized by scribes: 515. — ‘Sophoclean’: 588, 638, 802 n. x.
Dead persons greeted in Hades by relations : Epexegetic infinitive added to πολύς: 832.
735 f. Ephymnium rhythmicum: 186, 333.
Deity, idea of, purified : 350 n. 1, 375. — wrongly repeated against the MSS: 700.
Demeter, ancient etymology of: 49o. Epic, post-Homeric, language of: 4, 693,
δέος and αἰδώς : 92. 721, 725, 753-
Dewdrops, metaphor for young animals : 83. Epicharmus: 712 n. 2.
διὰ μέσου (position) : 169, 303 f. [Epicharmus] frs. 170-3 Kaibel, date of:
di consenles: 261. 787 n. 2.
digna dignis: 724, 831. Epirrhematic structure : 660.
Dike with sword : 730. Erinyes, choir of, in Ag. : 543.
Dindorf, W.: 53 f., 410 n. 3, 814. ἐρινύες, ἐπίσκοποι τῶν παρὰ φύσιν : 462 n. 4.
Diphilos : 44. Erinys related to Moira: 728 ff.
Disjunctive form of a general notion (οὗ 7’ ᾿Εσχατιῶτις λίμνη : 160 f.
ὄντες οἵ τ᾽ ἀπόντες) : 189. Estienne: see Stephanus. |
‘Dissimilatorischer Silbenschwund’: 673 f., ἔτλη and the like used in paraenetic παρα-
688. δείγματα: 471.
Dobree : 47. Etymologies in Aeschylus: 331, 692, 704,
Dochmiacs, correspondence in: 531 f., 538, 134.
662, 674, εὖ εἰδώς said of ἃ seer: 422.
— words overrunning the end of: 537. Eumenides, text, mock variants in F: 29
Dogs on the roof: 4 f. n. 5.
domi habeo ne quaeram foris and the like: Εὐμενίδες in Athens not an old name for the
435 f. Erinyes: 547 n. 2.
Doric (?) à in Attic words: 596. — Aeschylus’ own (?) title of the play:
Doricisms in anapaests : 738. 547 n. 2.
Dosiadas : 376. εὐφημία: 23.
Double-adoneus : 58. Euphemism : 252.
4872.3 Aaz 835
 GENERAL INDEX
Euripides influenced by Aeschylus: 63 f., Geryon : 393.
569, 585, 660, 693. Ghosts haunting threshold : 559.
— his use of new adjectives: 435. yÀ sometimes not making position : 774.
— Andromeda, prologue: 187 n. 1. Gloss, expanded to form complete line: 564.
— Electra: 821 ff. — has expelled the original word : 44.
— Iph. A., prologue: 187 n. 1. — slipped into the text : 66.
— — un-Euripidean features: 580 n; 4. γνώμη interpolated : 287, 385, 565.
Euripus, currents in: 116. God assisting man: 373 f.
Eustathius and the Epitome of Athenaeus: Gods, addressed in speech : 437 f.
412. — appeased by sacrifices? : 43 f.
Euthanasia : 603. — having regard for men (ἐπισκοπεῖν, etc.) :
Everyday life features used by Aeschylus 235, 587.
to enhance horror: 544. Goethe: 50, 175 n. 3.
Evil eye and the lightning-stroke : 237. Greek spoken in Tragedy by non-Greeks:
Excess, the idea of, in Aeschylus: 197. 484 f.
‘Exchanging (or comparing) A with B’ ypidos in Aeschylus followed by explana-
instead of‘... B with A’: 757 n. 1. tion: 37, 9, 8o ff., 379, 691.
'guttatim' (arrangement of clauses) : 2, 202,
Fables imported to Greece from the East: 395-
342. ἃ (or rà) δ᾽ és... as formula of transition:
Family sacrific:: performed by master of
383.
the house: 133 n. 2, 677.
Hand, organ of 8páv: 639.
Feminine of compound adjectives ending
in -n: 723. Haplology within a word : 673 f., 688.
Hartung: 54.
Feminine adjective without a noun: 415.
Headlam, W.: 58 f.
Sigura sermonis : see ἀπὸ κοινοῦ, τρίκωλον.
Fire felt by person in ecstasy : 581, 821, Heart, its neighbourhood seat of emotions:
108.
Flying (of fame) over the land : 289.
Heath: 45.
Foot of the victor: 412.
Heimsoeth, F.: 6.
Formulae of public speeches: 383, 656.
Helen raised above human level: 346.
Framing of section of speech: 15, 378, 382,
Helios, bringer of a message : 289.
418, 551, 760. Helmsman's seat : 11o, 766.
Franz, J.: 55.
Hense, O.: 54.
Frenzy, different stages of: 810 f.
Herald in Ag., purpose of his speeches:
Funeral speech: 732 f.
293 f.
Hereditary curse in Oresteza limited: 546 f.,
y, intervocalic, softened in later Greek: 625, 695.
525 n. 4. Hermann, G. : 47 ff., 744.
Gaming-board : 21. Heroes invoked in prayers: 263.
γάρ, unusual position of : 128 f. Hesychius to be used with circumspection :
Garland sign of victory : 251. ΣΙ.
Garment used in Ágamemnon's murder: Heusde, van, J. A. C.: 54.
504, 514, 648 f. Hiatus after indefinite 74? : 502 n. 1.
Gates of Hades: 602. — in dochmiacs : 197 n. I.
Gender, real, of a person, taking up a — in lyric jambics : 137, 197.
neutral metonymy : 345. — in trimeter? : 579.
Genealogical connexion of powers of evil: Hipponacteum : 58, 185, 344.
201. Hipponax, style of : 251.
generalis invocalio : 262. Homer and Aeschylus: 29, 119, 19o f., 212.
Genitive, absolute, when the subject of the Homeric customs in Aeschylus: 41, 320,
main verb is the same as that of the 500, 754 f.
participle: 439. Homeric god, features of, combined with
— of respect : 431 f. features of Attic cult: 259.
— of separation : 73, 240, 462. Homerism (linguistic) : 429.
— separated from its governing noun: 676. ὅμως in scholia : 448 n. 2.
Geraneia : 160. Horace, re-etymologizing : go n. 1.
γερουσία in Ag, a modern invention : 398 ff., Horror brought out by similes taken from
665 n. 2. innocent things: 228 f., 544, 656.
gerüfte, gerüchte: 614. Horse-bits: 485 f.
836
 GENERAL INDEX
“s-clause, pleonastic addition of (uófos . . . Ionic dimeters, their end not coinciding
ὡς ἀγορεύεις) : 170. with word-end: 337.
ὡς λόγος etc. without verb: 148. Ionici a minore: 328.
Hospitality, bond of: 395 f. — — with the clausula v v τῷ — —: 185.
Household, self-contained economy of: ‘Irony’ last resource of commentators: 719.
435 f. Irrelevant circumstances disregarded by
οὕτως introducing a story : 338 f. Aeschylus: 255 f.
Humble matter ennobled by language : 280, ἴσα πρὸς ἴσα: 269.
283, 772 f., 829. ico-compounds : 681 f., 695 ff.
Humboldt, W. v.: 50, 254 f. Ityn Ityn: 522.
ὕβρις : 349 f. Iuppiter Liberator : 653 n. 1.
Hymenaeus recalled in dirge: 336.
Hyperbaton (‘Sperrung’): see τίς. Jealousy of the gods : see φθόνος θεῶν.
— 12 f, 76, 94, 301 n. I, 376, 404 f, 568 -
Jettison (ἐκβολή) : 454 f.
628, 676, 687, 827 f. Juridical technicalities: 294.
Hypostasis: 62, 755, 786. Juxtaposition of the type θυσίας παρθενίου
ὑπόθεσις of the Agamemnon : 370. θ᾽ αἵματος : 123 f.

καὶ γάρ introducing παράδειγμα: 470.

Iambic metron (in lyrics) ending with end καὶ ταῦτα μὲν τοιαῦτα and the like: 432.
of word: 705. , ἸΚαλλίστη (Artemis): 83.
Iambic trimeter (in lyrics) produced by & κάλλιστον τὸ δεῖνα : 407 f.
copyist : 238. Kennedy, Β. H.: 53.
Iambics (lyric), correspondence : 211, 351. Kennings: 251.
— syllaba anceps: 132, 355. Kerkopes : 68 f.
Images in Aeschylus taken from everyday King as shepherd : 820,
life: 228 f. — as watcher and protector (φύλαξ) : 688,
imis medullis and the like: 701. Kingship, constitutional, in Aeschylus : 388.
Imperfect, after a preceding aorist : 296. — ideal of: 362.
— and aorist : 329. κιθαρωιδικοὶ νόμοι : 50.
— denoting detail subordinate to main κληδών, nature of : 789.
action : 138. κλύεις at the end of a speech : 178.
κοινὴ συλλαβή: 18 ff., 579.
Imprecations : 217.
Indefinite pronoun: see ris. Korai of the Acropolis : 219.
κρείσσων = θεός : 39.
Indirect speech continued by direct speech :
303. 'Kunstprosa' : see Stylistic devices.
Infinitive, consecutive, with article : 286, Kußeia: 20.
— epexegetic, after οὕτω ..., ὧδε... .: 238 f. Labbé, Charles: 67 f.
— — to adjective : 311. Lecythion : 59, 499 f.
— exclamatory : 796 f. Legal metaphors: 27.
— with τὸ μή: 549. λέγοιμ᾽ ἄν, polite : 386 f., 830.
— without article, subject of a verb: 292. λείπει τὸ ὡς in Comparisons : 206.
Insatiate daimon : 505. Lines assigned in MSS to wrong speaker:
Inspiration of the poet : 64 f. 253 f., 306.
Internal object, μέγα etc. not added to a Lion: 561 f.
verb that has another internal object Arai: Homer's view repudiated by Aeschy-
(therefore not πόλεμον μέγα πνεῖν etc.): lus: 209.
τοῦ. Litotes : 365.
Interpolation: 9, 266f., 287, 385, 390 f., — by means of negatived adverb: 172.
392 f., 408 n. 4, 482, 562 ff., 575 n. 1., 580 Luca bos: 23 n. 1.
n. 4, 602 n. I, 743 n. I.
— by a producer: 321 n. 1, 814 f., 820. Madness, symptoms of: 672 f.
-— caused by absence of verbum dicendi: Magical ceremony according to 'law of
756 n. 1. homoeopathy' : 755 n. 1, 757.
— in anapaests? : 717. μακαρισμός : 420,
— of y”: 574. Makistos : 157.
Interrogative postponed : 705. Makron, vase-painter : 345.
Invocation without a following prayer : 698. Marginal entry cause of interpolation:
Iohannes Siculus: ro f. 266 f., 385, 602.
Ionians μακρολόγοι: 415 n. 1. Marriage-formula : 555.

837
 GENERAL INDEX
Marrow, Greek idea of: 47 f. Νόστοι: 176 f., 313, 322.
Masculine of aorist participle used as fem.?: *Nothilfe' : 614.
73: Nouns, abstract, in «ov, in legal phrases:
Massinger : 508. 273.
Mau, August : 821 ff. — of action: see nomina aclionis.
Meat rare in classical Greece: 747. — verbal, in -pa: 734.
Medical language: see Surgery. νυκτὶ βουλήν : 399.
— — 502, 687, 702.
μελάμπυγος : 68 f. Object : sec Internal.
μέν... öe-clauses interrupted by another —- attracted to predicate noun: 427.
clause : 287. — κείρειν φόνον and the like: 14.
μέν shifted to the second place in the sen- Oblivion : 622.
tence instead of following the word after obscura diligentia of commentators: 104 n. I.
which it might be expected : 348. Odysseus : 387.
Menelaus, depreciation of: 70. ot shortened in οἷος, ποῖος, τοιοῦτος : 579.
Μῆνις : 92 ff. ὀλολυγμός : 18, 297, 572 f.
Metre: see Anapaestic metron, Bacchiac Omen: 789.
tetrameter, Caesura, Clausula, Colon, ‘One’ and ‘many’ in juxtaposition : 689 f.
Dactylic, Dactyls, Dochmiacs, Double- Optative, in indirect speech : 300 f.
adoneus, Ephymnium, Hiatus, Hippo- — of contracted verbs, forms of: 621 n. 1.
nacteum, Iambic(s), Ionic, Ionici, Lecy- —- with dv in protasis: 420.
thion, Paroemiac, Paroxytone word, — without dv, after οὐκ ἔστιν ὅστις etc. : 310.
Praxilleion, Reizianum, Syllabic corre- — — — potential: 620 n. 2, 645, 666 n. 1.
spondence, Telesilleion, Tribrach, Tro- Oracles : 79.
chaic tetrameter, Trochees. Oreithyia : 116 ἢ. 1.
Metrical units coinciding with syntactical ‘Oresteia-words’ : 200 ff., 266.
units: 118. Orestes sent to Strophius : 402.
Middle of the future serving as passive: 104. Orpheus, singer : 774.
misere miser sum and the like: 125. ‘Others’ to be the victim of a god: 741.
pv sometimes not lengthening the preceding οὐ (μή) at the end of a trimeter: 281.
syllable: 443 f., 830. Oxymoron of the type νᾶες dvaes : 519 f.
Moira : 728 ff.
Monologue in Aeschylus: 25.
Monologue-like speech in Aeschylus: 383, Paean : 140 f., 577.
744, 779. Paley, F. À.: 52 f.
Monosyllable dropped out at the end of a Pan in Athens: 36.
line: 645. mav-compounds: 434 f.
Mooring-cable (or anchor) image of hope: πάντ᾽ ἔχεις λόγον and the like: 291.
258, 829. πάντες θεοί: 262.
Mountain tops, gods on: 35. παράδειγμα (illustrating a maxim) in the
Müller, Otfried: 50 f. form of a hypothetical sentence: 461.
Mykenai: 17 f. — introductory formulae of: 470 f.
Paragraphos in MSS: 147, 169, 252 f., 784.
Nägelsbach: 54. Parenthesis: 60, 363.
Name of a daimon, gives power over him: Parodos, oldest form of: 27.
100. Paroemiac ending in molossus: 192.
Narrative, archaic: 39, 116, 805. — in proverbs : 724.
Negative, initial, not going with next — levelled down by Triclinius : 627.
word : 609. — sentence continued after: 42.
— repeated : 103, 604, 775. Paroxytone at end of trimeter: 5.
νεόπλουτοι disliked : 473. Participle, accusative of, after dative: 198.
‘Net’ in which Agamemnon is caught : 512 f. — added to complete picture: 394.
Night, mother of day : 148 f. — in -vres etc. feminine? : 283 ff.
— personified : 187. — nominative, in apposition to a clause:
Nightingale, its colour: 520 f. 166.
nomen alque omen: 331. — with verbs of saying: 150.
nomina aclionis in -w denoting personifica- — — — showing: 158.
tions : 639. παύομαι at the end of a speech : 475.
Nomina sacra, ‘contracted’: 531. Pauw: 44.
Nominative used as vocative : 489, 492. Pearson, John: 40 ff., 78 fi.

838
 GENERAL INDEX
Pelops, not mentioned in Oresteta: 546 f. Privative compound in connexion with
— his curse upon his sons? : 745 n. 3. negative particle : 235.
Perachora : 160 f. ‘Productio epica’ in Aeschylus: 75 n. 2, 829.
περὶ δὲ τοῦ δεῖνος serving as a colon : 431. Proper name postponed : 328, 331, 543.
Periphrasis, τὰ κατ᾽ ἐμέ etc.: 801. Prophets, their knowledge of hidden facts
Periphrastic phrase governing accusative: of the past and the present: 543.
315. προσκυνεῖν τὴν γῆν : 257 n. I.
περιφραστικῶς, in scholia : 243. Proverbs: 767 f., 799, 800.
Personification : 13, 91. — use of: 774.
merteia: 20. Proverhial expressions of the type ὄνος
$, sound-value of: 149 n. 2. λύρας : 612.
Φόβος : 676. Psychology used to explain nonsense: 393,
φόνος ἐκ προνοίας : 763. 811, 816.
@0dvos-motif: 411. Punctuation, wrong, used in MSS to ease
φθόνος θεῶν : 349 f., 463. word-order : 128.
Plato’s gods, contrast with Aeschylus: 111. Punishment of sinners belated : 351.
Platt, Arthur: 204, 642, 751. πύγαργος : 67.
Πλειάδων δύσις : 380 f. Pylades in Choephoroe : 820 f.
Pleisthenes : 740.
Pleonasm in phrases like εἶπε φωνῶν : 120. Question, double, broken off: 151 f.
Pleonastic description of anything exces- — with πῶς and subjunctive of aorist : 358.
sive: 125, 197, 457, 458 n. 2. — without interrogative particle: 275.
πληροῦται ὁ χρόνος εἴς. : 604 n. 2. Quotations from A. Ag. in lexicographers
Pluperfect, in Aeschylus: 213, 323 f. etc.: 10 f.
— rare in early Greek: 140, 194.
Plural, general or allusive : 216, 679, 771. Rationalistic criticism of Aeschylus in anti-
— taken up by singular: 717 n. 3. quity : 254.
Plüss, Theodor: 97 n. 2, 536 n. 1. Raven feeding on corpses : 700.
πνεῦμα: 65. Recognition, types of : 820.
ποῖος used jeeringly: 507. Re-etymologizing of words: 9o.
πόλις, disease of: 389. Reflexive pronoun : 385, 8oo f.
Politeness, fotmulae of: 386 f., 476. Refrain: 73 f., 95 f.
πόλλ᾽ ἀγαθὰ γένοιτό σοι: 179. Reizianum : 489 n. 7.
Polysyndeton in invocations of the gods: Repetition of the same word in Aeschylus:
259. 499.
Pomp at later performances of tragedies: Responsibility for human actions: 374.
370 f. Return from Troy, epic version: see Νόστοι.
Porson: 46 ff. Rhythmical refrain : see Ephymnium rhyth-
Positive subject evolved from οὐδείς : 209. micum.
Possessive not used in classical Greek for Riddle: 328, 581.
‘stretch out your arms’ etc. : 651 f. Ring-composition : 119.
Postponement of important details: 805. Robbery distinguished from theft : 27o.
postpositivum : 787 f. Romanticism, endangering the understand-
‘Praesens pro futuro’? : 628 f. ing of Aeschylean thought: 451 f.
Praxilleion : 662 n. 1.
Prayer, ἀποπομπή: 740 f. Scales of justice : 142.
— at the end of a speech in Aeschylus : 817. Scaliger: 36, 67 f.
— concluded by cursing an ill-wisher : 254. Scaligeri praeceptum : 292.
— language of: see θέλων. Schneidewin: 54.
— — 74, 99 f., 101, 126, 179, 236, 259, 262, Scholia, beginning with τὸ X.. πρὸς...
325, 359. 525 n. 2.
— recited by the king, etc.: 133. — defending the poet : 254 n. 1.
— οὗ returning traveller : 257. — language of: see ἐκδικεῖν, ὅμως.
Prayers, Latin: 179 n. 2. — — 666 n. 1, 801.
Preposition at the end of a trimeter: 587 f. — misused in textual criticism: 190 n. 1,
Prepositions, ἀναστροφή of : 469 f. 448 n. 2, 8οο f.
Present tense in narrative: 324, 650, 755. — (on Aeschylus) compiled by Victorius:
Priamel: 407 f. 341.
Privative adjectives without verbal notion: — quotations in, adapted to the text of a
137. particular MS: 274 n. 2.

830
 GENERAL INDEX
Scholiasts discussing divergent readings: Superlative of praise used with discretion:
Sir. 269.
School edition (14th century) of Aeschylus: Surgery : 388.
17 ff., 26f. | Swan’s death-song : 684.
Schütz, C. G.: 45 f. Sword and stick carried by an old man?:
scriptio plena of καί: 379. 782 f.
Sealed rooms, treasures etc. : 302 f. Syllaba anceps: see Iambics.
Sea-water producing madness? : 663. Syllabic correspondence in Aeschylus’ lyric
Seeress παλλακή of the god : 554. | iambics not over-strict: 121 f.
Seers in the houses of the great : 214. — — in dochmiacs: 515, 523.
Seleukos ὁ ‘Opnpixds : 754 n. 2. Symmons, John: 431 n. 2.
Senate : see γερουσία. Syntactical pause after first word of tri-
Seneca's Thyestes : 749. meter : 660, 746.
Sentence apposition : see Apposition.
Sententiousness of the uneducated : 312 n. 1. lalio : 615, 674, 744, 807 f.
Separation: see Hyperbaton. Tantalus: 695.
Sewell, W. : 140, 203, 372 n. 4. — his slaughtering of Pelops: 748.
Sicilian words in Aeschylus: 712 f. re after noun and adjectival attribute : t30 f.
Simile : see Comparison. — appositional? : 722.
Singer's reward : 444. — position of: 211 ff.
Sitting at meals : 754 f. τε καί in threefold expression: 249.
Slaves members of the household : 469. Telesilleion : 706 f.
Smoky house: 354. Tenses in Aeschylean narratives : 213, 650.
coi λέγει and the like: 475. θεία μανία: 65.
Sophocles, influenced by Aeschylus: 158, θέλων in prayers : 325.
266, 428, 603, 617, 633, 655, 686, 743, 777, θεοί referring to Zeus: 194.
778 n. 1. Thirteen, the number: 759.
— Oedipus Rex, finale of: 803 f. Thousand, number of the Greek fleet: 28.
Spear-fight, first stage in battle: 41. Thyesteae preces: 758.
speculum consuetudinis : 386 n. 1. Thyestes’ adultery: 546 f., 745.
Speech beginning in the middle of the line: — his children: number? names? : 758 n. 2.
295. τί δεῖ μακρηγορεῖν; and the like : 298.
— — with fresh line: 77. Time (imaginary), of dramatic action: 26.
‘Sperrung’: see Hyperbaton. — man's lifetime born with him etc. : 63.
Sphinx, riddle of: 50, 581. — required for what is happening on the
Stage-convention by which persons, or ob- stage: 813 n. I.
jects, are disregarded unless expressly τινα corrupted to πνεῦμα: 115 n. I.
mentioned: 432, 783, 821 n. I. τίς separated from its noun : 687.
Stage directions in relation to dialogue: Tmesis : 293, 348, 557.
643. τοιόσδε (τοιοῦτος) in concluding statement :
Stanley, Thomas: 38 ff., 78 ff. 305 f., 394, 410, 815.
Stasimon, preceded by anapaests: 184. τῶν νῦν (τῶν τότε) added to a superlative:
— sometimes replaced by anapaestic piece: 267 ff.
633. Torch-races : 166 ff.
— its stanzas not divided among sections Tragic decision in Aeschylus: 126 f., 726.
of Chorus: 246. Transition, formula of: 383.
Stephanus, Henricus: 34. Transposition in MSS to conclude trimeter
Stesichorus: 218. with paroxytone: 5.
Stichomythia in Aeschylus : 626. — — — to produce normal word-order:
Stoic martyrs in Nero's time : 653 n. 1. 207.
Stoning: 506, 764 f. — of lines in text of Ag.: 780.
Stylistic devices common to early poetry — — — sometimes less probable than
and ‘Kunstprosa’: 217, 689 f., 724, 731. deletion: 602 n. 1.
Subject of a verb to be inferred from con- Tribrach : 746.
text: 44 f., 205, 457. Triclinius alters the text: 12 ff.
Subjunctive with μή in independent clause — autographs of: 3 n. 3.
expressing fear: 176. — dislikes paroemiacs : 627.
Substantive expressing in a picture the idea — metrical conjecture of: 197, 214.
of a preceding verb: 202, 382. — restores metrical symmetry by deletion:
Suicide attempted by women: 394. 142 n. 1.
840
 GENERAL INDEX
Triclinius, his scholia on the dramatists: 17 ‘War!’ a cry: 29 f.
n. 4. War, experience of: 117, 226 ff., 278, 282 f.
— hissympathy with the religion of Aeschy- 285, 294, 390 f.
lus: 102 n. 1. Wecklein: 56 f.
τρίκωλον, crescendo (c longer than Ὁ, Ὁ Weil, H.: 56.
longer than a): 574, 666. Welcker: 51.
— in expressions of assent : 384. Wilamowitz : 59 ff.
— privative (e.g. ἀφρήτωρ, ἀθέμιστος, ἀνέ- Wind in metaphors: 115, 127 f.
aris) : 217. Wings, of the sun, of stars : 289.
Trimeter : see Iambic. Wives, Athenian, position of: 303.
.— unusual type of: 428. ‘Woman’, etc., applied to a man: 770.
— with two resolved longa : 744. Women, prejudice against : 775.
Trimeters, iambic, interrupted by ex- Word-order: see Adverb, dv, dpa, δέ, διὰ
clamations : 558. μέσου, γάρ, Hyperbaton, μέν, Negative,
Trinity of divine witnesses : 675. τε, Tmesis.
τρίτος σωτήρ: 154, 652 f. — εἴ τι μή and the like: 608,
Trochaic tetrameter, in Tragedy : 634, 780. — emphasis sometimes on the last word:
— punctuation before last two syllables: 677 n. 1.
8o2 n. 1. — ‘extended’ object forming colon of its
— regulation of word-ending : 787 f. own: 459 f., 461.
Trochees, Aeschylean, correspondence of: — γυνὴ γυναικός etc. : 615.
722 n. 2. — juxtaposition of words denoting con-
Trojan booty distributed over the shrines trasted, or complementary, ideas : 171.
of Greece: 290. — noun with two attributes: 212 f., 703.
Trojan war regarded as lawsuit : 27, 270 ff., — πῶς τις dv: 632.
374 ff., 599. — proper name put between the elements
Tyndareus, daughters of: 689. of an 'attributive group' : 314.
— sequence 'noun, verb (or some other part
Ubaldi, Paolo: 553, 571 n. 3. of speech), preposition, adjectival attri-
unilas temporis: 256. bute' : 436 f.
Urbanity : see Politeness, — type ἡ τιμὴ θεῶν: 315 ff.
— type inier densas, umbrosa cacumina,
Vase-paintings: 138 n. 2, 139, 175 f., 345, Jagos absent from early Greek poetry?:
681 n. 9, 735 f., 774, 783 n. 1, 809 n. 2, 820. 71 f.
Vengeance mentioned in laments for the — verb emphatically at the beginning : 678.
dead : 617. Words occurring only in Oresteia : 200 f., 266.
Verbal adjectives in -ros: 12, 164, 191.
— — — active as well as passive: 137 n. I.
— — — stylistic variants of simple adjec- Yoke of compulsion : 127.
tives (ἀκάρπιστος for ἄκαρπος etc.): 322,
682.
— — — with only two endings: 155. Zephyros: 332.
Verbal adjectives of the type wav... τος : ugma : 362.
4341. Zeus: see θεοί.
Verbs in -dlew, -ἰζειν with £ in aorist: 358. — as helmsman : 109 f,
— of sending with infinitive: 162. — βασιλεύς : 186 f.
Verrall: 57 f. — etymology (ancient) of the name: 704.
Victim to acquiesce in sacrifice: 604. — Herkeios, altar of: 482 ἢ. 1.
Victorius, Petrus: 34 f. — Ὕπατος: 35.
Vocative and imperative: 16. — in Aeschylus: 111.
volens in prayers : 325. — Ktesios: 470.
Voting of the jury : 376 f. — Soter: see τρέτος σωτήρ.
Vouliagmeni: 16o f. — τέλειος : 440.

841
 II
WORDS
GREEK
ἀβλάβεια: 462 f. ἀμφι- in compounds: 331. &réAevros : 688.
dydAaxros : 338. ἀμφίβληστρον : 647 ff. ἀτελὴς στρατείας : 46 f.
ἄγαλμα: 121. ἀμφιθαλής : 522 f. ἄτη: 545.
Ayapeuvônos : 710. ἀμφιλαφής : 458. ατιμος : 595.
ἄγαν, prosody of: 574 n. 1. ἀμφίλεκτος : 396 f. ἀτίτης : 45 f.
dyavôs : 55 f. ἀμφινεικής : 331. Ἀτρεῖδαι, not Ärpeida: 28,
dyyaprıov: 153 f. ἀμφίσβαινα : 569. ἀτύχημα: 671 n. 4.
dykaDev : 4. ἀνάγκη : 729. αὐγαί = ‘eyes’: 509.
ἁγνός : 128. ἀναίνεσθαι: 159, 291 f., 789 f. αὐτοσαυτοῦ etc. : 385.
ἀγρηνόν : 584 n. 3. ἀναρχία: 397. αὐτοφόνος : 494 f.
ἀγύρτης : 591. ἄναυδος: 136 f. αὐτόχθονος : 272.
ἄγχιστον: 144 f. ἀνδρακάς : 752 f. αὐχεῖν : 707 f.
ἀγχόνη : 455. ἀνδρολέτειρα : 693. ἀφεῖσθαι passive: 216.
ἀγών : 388. -ανδρος, words in: 40. ἄφερτος : 200.
ἄδην : 382. ἄνευθεν : 750 n. 3. ἄφθονος : 237.
ἀδίκιον : 273. ἀνήρ with a noun added: &xqvía : 219 f.
ἄδολος : 54. 772. ἄχος : 577.
deípew, ἀείρεσθαι : 721. ἀντήλιος : 264. ἄψυχος : 697.
ἄεπτος : 83 f. ἀντι, ‘anastrophe’ of: 593.
ἀηδοῦς, genitive of ἀηδών: ἀντίδικος : 27. βάλλειν intransitively?: 534 f.
523. ἄντιτος : 673 f. βαστάζειν: 22 f.
ἄθεος : 493. ἄνωθεν : 742 Í. Ba?tew : 231.
αἰγίπλαγκτον : 160. -dvwp (-ἠνωρ), words ending βέβηκε: 213.
αἰδοῖος : 208, 792. in : 40, 92. βία: 111.
αἰνεῖν : 55, 87 f., 702. ἀξυνήμων : 483. βιάζεσθαι : 714.
almos : 154. ἀξύστατος : 693. βλάπτειν with genitive : 240.
αἱρεῖν : 556. dolos : 132. βοᾶν τινα: 29.
αἱρεῖσθαι: 178 f., 790. ἀπανθίζειν : 797. βοή: 29 f.
αἴσιμος : 355. ἀπαρκεῖν : 198. βοηθόος : 614.
αἰτεῖν : 85 f. ἀπάρχειν, ἄπαρχος : 565 f. BovkoAetv : 325.
αἰχμή : 240. ἄπειρος : 649 f. βουλεύειν : 561, 640.
αἰῷ accus. of aids : 131 f. ἀπένθεια : 225. βουλή: 398 f.
αἰών: 62 f., 278. ἀπέχειν : 510. βούπρωιρος: 135.
ἀκόλουθος gloss of ὀπαδός: ἀπό quasi-instrumental : βραβεύς : 132.
223. 455, 774 f.
ἄκρος : 313. ἀπο- : preverb: 95. γάγγαμον: 190 f.
ἀλαπάζειν : 70. — privative use of: 363. γαμβροί: 336.
ἀλάστωρ: 711. ἀπόξενος : 596. ydvos : 656.
ἀλέκτωρ : 90, 800, 829. ἀπόπολις : 665. γάρ: 'for there is no need to
ἀληθῶς: 576. ἀπορραντήριον: 495. mention . . .': 348.
ἅλις : 792, 795 f. ἀπόχρη : 741. — introducing a paren-
ἀλιτήριος : 678. ἄπτερος : 1521. thesis : 363.
ἀλκή : 64. ἀρβύλη : 420. —- where thought is com-
ἀλλὰ... μήν: 787. ἀργᾶς : 67. pressed : 123.
ἀλλ᾽ εἶα : 785 n. 1. ἀρείων : 50. ye, in sentences of the type
ἅμα added to participle: ἁρπαγή : 270. οὔκ, εἰ... ye: 798.
111|. ἀρχαιόπλουτος : 473. — not used in answer to
ἁμαρτάς : 273 N. I. ἀρχαῖος : 291. imperative : 274.
ἁμάρτια (ra): 273. dpwyy: 28 f. γενναῖος : 552.
ἀμαυρός : 236. ἄσκοπος: 235. γῆν πρὸ γῆς : 456 n. 1.
ἀμήνιτος : 322. ἀσφάδαιστος : 603. γνῶμα: 638.
ἁμός : 803. ἀταύμωτος : 140. γνώμη : 636.

842
 WORDS
γνώσηι without object: 780. ela, δή: 784 f. éare (conjunction), its con-
γόος : 36. εἶναι : see ἧστε. struction in early Greek:
γύννις : 769. εἶτα very rare in Aesch. : 164. 163 ἢ. 2.
εἴτε... εἴτε: 250. ἐς φθόρον and the like: 585.
ἐκ τοῦ ὁμοίου, ἐκ τῶν ὁμοίων :, ἔσω : 633.
δᾶ: 490 f., 832. ἕτειος : 2.
$afjva: : 76. ἐκ τῶνδε : 560, ér in threats : 674.
ddios (δήϊος) : 282. ἐκμαρτυρεῖν : 549 f. εὖ, substantival : 74.
δάκος : 379. ἐκπάτιος : 30 f. εὐδαίμων : 174.
δάπεδον : 490 n. 3. ékmpáaaew ; 592. εὐεστώ: 321.
δατήριος : 678. ἐκτελής, ἐντελής : 61. εὐθύδικος : 348.
δέ (σὺ δέ in apodosis) : 484. ἐκτολυπεύειν : 464 f. εὐθυμεῖν, εὔθυμος : 748.
δέ ye, δὲ... ye: 792. exdarus : 336. εὐθύνειν : 670 n. I.
δεμνιοτήρης : 33. ἑκών: 24, 486 f., 763. εὐκαμία: 23.
δεσπότης : 564. ἐλπίς : 676, εὐκλεῶς : 607.
δή preparatory to a quota- ἐμβάς : 429. εὐκταῖος : 652.
tion (τὸ σὸν δή etc.) : 277 f. ἐμπίπτειν : 176. εὐμαθεῖν : 292.
δῆγμα: 531 f.
ἐναίσιμος : 355. εὐμαρής : 610.
δημόθρους : 397 f. ἔναλλος : 228 n. I. εὔμορφος : 233.
δημόκραντος : 233 f. ἔνδικος : 448. εὐπειθής and εὐπιθής : 150 f.
διά; with genitive: 23of., ἔνθηρος : 283. εὐπρεπής : 307 Î.
295, 691, 704. ἐνί, Evi: 48. εὑρίσκειν: 301.
— with accusative : 691. ἐν τέλει: 61 f. εὐσεβεῖν transitive: 175.
— — denoting action of a ἐντέμνειν : 14. εὐσεβής : 195.
deity : 333 f. ἐξ ἐλπίδος : 450. εὖτε with aor. indic.: 163.
διαί: 230, ἐξαγίζειν : 319. εὐφρόνη : 149, 829.
διανεκής : 169 f. ἐξαυχεῖν : 708.
διατρίβειν : 479 n. 3. εὔφρων : 148, 365 ff.
ἐξεικάζειν : 575 f. εὔχεσθαι, 'to vow’, construc-
δικαιοῦν : 202 f.
ἐξωμοσία : 550. tion οὗ: 422.
din: 375, 142, 164. ἐπαντέλλειν : 18, ἐφέστιος : 389.
δίλογχος : 320. ἐπάργεμος : 502.
Διόθεν : 236 f. ἔχειν intransitive : 116.
διπλοῦν = διπλάσιον, of
ἐπαρκεῖν : 532 f. ἔχειν, 'to hold a conquered
ἔπειτα after participle : 240. city’: 171.
damages: 273. ἐπεύχεσθαι: 583.
δίπους : 581 f. ἔχειν with infinitive: 192.
ἐπήβολος : 105. ἐχενηΐς : 9o.
dtoooi: 75 f.
ἐπήκοος : 667.
δίφυιος : 694. ἐπί with dative, with verbs
διχορρόπως : 589. ζῆλος : 426.
of motion : 687. ζύγιοι, ζυγῖται: 765 n. 1.
διχοστατεῖν: 171.
— — numerals: 760, ζυγόν: 110, 766.
δμοιός : 318 n. 1. ἐπὶ τῶι ὀφθαλμῶι: 672.
δορίγαμβρος : 330.
ἐπικραίνειν : 194, 628.
δορύξενος : 305 f. 4 affirmative : 692.
ἐπιλάμπειν : 258.
δρᾶμα: 270, ἤ apparently omitted after
ἐπιξενοῦσθαι: 615 f.
δρᾶν: 638, 640. comparative: 299.
ἐπίξηνον : 593. ἤδη : 440.
δροίτη : 731. see Index I,
δυσαυλία: 279.
€TLOKOTIELV :
ἠλίθιος : 192.
‘Gods’. ἡμεῖς etc., ‘ego’: 595.
δυσοίζειν : 609 ff. ἐπίσσυτος : 527.
δύστλητος : 740. ἐπίστροφος : 209. ἣν πὼς : 636.
δύσφατος : 528. ἐπίφοβος : 528. τήρης, words in: 682.
δυσφημεῖν: 401. ἐποπτεύειν : 587. fore (not ἦτε) early Attic:
ἐποπτήρ: 587.
275 f., 829.
€a : 580 n. 4. ἐράσμιος : 200.
ἑαυτοῦ etc. used of first and ἔργα γάμοιο etc. : 555. θᾶκος : 75 n. I, 261 f.
second person: 519, 8oo. ἐρείδειν : 453. θάρσος, θράσος: 364.
ἔγγονος, ἔκγονος : 195. ἐρείκη, ἐρίκη : 158. θεῖον, τὸ : 403.
ἐδανός : 662. ἐρίδματος : 602. θεόθεν : 64 f., 79.
ἔδεθλον, ἔσθλον : 355. ἽἝρμαιον, accent of: 154. θεόκραντος : 705.
εἰδότας : See πρός. ἔρνος : Ἴ21. θερμόνους : 536.
εἰ δ᾽ οὖν: 326. ἔρρειν : 220. θέσμιος, θεσμός : 737.
els : see Index I, ‘One’. ἐσθλός : 302. θεσπέσιος : 530.
εἴ ris ποτε: 691 f. ἔσκε: 340. θέσφατος : 616.
843
 WORDS
θηρᾶν: 547. κοιμᾶσθαι: 3. pn in statement depending
θιγγάνειν : 226, 412. κοινοῦσθαι : 636. on verb of saying: 347 f.
θραῦμα, θραῦσμα: 532 n. 1. κοίτη: 286, 705 f. μηδέ: 708 f.
θρεῦμαι: 532. κολοσσός : 218 f,
θρύπτειν : 749. κομίζειν : 363 f. μὴ εἰδέναι in crasis: 550 f.
θύος : 663 f. κουροτρόφος: 716. μή vvv: 425.
8vooxóos : 53. κραίνειν : 86, 193 f. μὴ οὐ with infinitive: 533.
θωύσσειν: 404. -xpavros, compounds with: μῆτις : 670.
I94. μηχάνημα: 512 ff,
κρατεῖν not = iubere: 10. μιάστωρ: 711 n. 3.
ἰάπτειν : 733.
larpópavris : 767.
κράτος : 60, 147. μεσήτη : 567.
κρεουργεῖν, kpeoupyós : 747 f. μισόθεος : 493.
-ilew, verbs in: 155, 434.
᾿Ιλιοφθόρος : 376.
κροκωτός: 137 f. μνάμων (μνήμων) : 94.
κτεάτειρα: 187 f. μοῖρα: 463.
ἵνις : 338. μολεῖν : 326.
Kreis: 754.
ἰσήρης : 682.
loo-compounds : 681 f., 695 f. κτήνη : 778. μόρος : 523 f., 756.
ἱἰσοκλινής : 682.
κυκλεῖν, κυκλοῦν: 447. μόρσιμος : 95.
κύνες = ‘servants’: 82. μυθεῖσθαι: 641.
icovopía : 681 n. 7.
ἰσόψυχος : 695 ff. λαβεῖν, its meaning in οὐκ νεάζειν : 350,
ἱστοτρίβης : 680 ff. νεαρός : 180.
ἂν λάβοιμι etc.: 151.
ἰσχάς : 90. νείκη, 1: 646.
λαγός, λαγώς : 72.
*Ipeyeveta? : 723. νέμειν: 47, 54 n. I.
λαμπρός: 540 f.
-ıxos, words ending in: 34. veoyvös: 531.
λαμπτήρ : 16 f.
ἐχώρ : 702. veóppvros : 636 f,
ἰώ in Aeschylus : 256, λαμπτηροῦχος : 403.
λαστήριος : 678 f. vewpns: 352.
λέβης: 515 f. νόμος : 296.
καί going with the whole λέγω σοι almost threatening : νοσηλεύειν : 192.
following clause: 118, 667. νοῦς and its compounds:
281 f. λιπόναυς : 122 f. 567 f.
καὶ μήν: 540. λίπος : 671 f. νυκτίπλαγκτος : 12.
καὶ νῦν: 9. λοχῖται, λόχος : 781. vvudóxAavros : 346 f.
καὶ πῶς;: 277. λόχος, 'childbirth' : 82. γωτίζειν : 155.
καὶ τότε: 114.
καινίζειν : 434. μάθος: 9. ξουθός : 520.
καιρός : I9I. μάκιστος in Tragedy : 157.
κακός : 760 f. ὄβριμος : 665.
κακοφρονῶν : 538. μακρὰν érewas et sim. : 414 Í. ὀβρίχοιοι: 85.
καλλίπαις : 348 f. μαρτύρεσθαι: 614. öde: 37, 100 f.
καρατόμος : 404. μάσσων: 298. οἶδ᾽ ἐγώ : 799.
κάρτα: 578. μαστεύειν, ματεύειν : 498. οἴζεσθαι (?) : 610 n, 1.
Κασσάνδρα, spelling of : 4671. μάταιος, μάτην : 102 f. οἰκοδεσπότης : 564 n. I.
καταρρίπτειν : 400. μεγαλόμητις : 670. οἴκοθεν: 435 f.
κατειδέναι: 5. μείλιγμα: 679. οἰκονόμος : 93 f.
κατεύχεσθαι : 577. μέλαθρον : 70, oikovpety, olkovpós : 369, 770.
κάτοπτρον: 385 f. μελαμπαγής: 205. olun, oluos: 530.
κέ) - "dar: 507. μέλλω : 638 f. οἰμώζειν : 143.
κεναγγής : 118. μέν 'solitarium' : I. ὀλβίζειν : 420.
κηκίς, κηκίω : 434. μὲν οὖν : 403. ὀλολύζειν : 573.
xnpós : 136. μένειν : 831. ὄμμα: 672 n. 1.
κικλήσκειν : TOI. μένειν and μίμνειν : 156. ὁμοῖος : 302.
κλάζειν : 104. μένειν with dative: 235. ὁμοῦ: 529.
κλαύματα: 526. μέντοι: 320. ὄνησις : 170.
κλέος : 244 f. μέριμνα: 725. ὁπαδός : 222 f.
κληδών : 120, 789. μεσόμφαλος : 482. ὅπως with the first person
κλοπή : 210. μετά with the genitive: 470. of future indicative: 298.
Κλυταιμήστρα, spelling of: μεταγνῶναι: 128. ὅπως dv with optative: 101.
521. perairios : 371 f. ὀργᾶν: 124.
κλυών:: 327. μεταμανθάνειν : 336. ὀργή: 125.
κοιμᾶν : 297. μετέχειν μέρος τινός : 258. opivew : 775.

844
 WORDS
ὅρμαίνειν : 653. πῆγμα: 551. npooßolal: 203.
ὅρος : 243 f., 529 f. TGOAVELV : 152. προσεικάζειν : or f.
ὀρτάλιχος : 34. πιθανός : 241 f. προσεννέπειν ; IOI, 140, 172.
-opros, Words ending in: 93. πικρός : 301 n. I. προσήκων = ὧι προσήκει:
ὅσιος : 355. πλᾶτις : 2I5. 491.
οὐ μόλις : 492. πλέω, neut. pl, in Drama: πρόσθε, ‘prepositional’: 78.
οὔ φημι: 194. 392. προστατήριος : 443.
οὐδὲ ydp: 718. “πληθϑής, second element of πρόύόστριμμα: 207.
οὐδὲν ἧσσον: 655. compounds: 78. προστρόπαιος : 745.
οὐδέπω, temporal meaning πληθύνεσθαι: 641. πρόσφαγμα: 594.
essential: 158. πληροῦσθαι: 166. προτέλεια : 40 f., 129, 339.
-oüv, verbs in: 79 f. πνεύει Aeolic ? : 62. προτιμᾶν : 802.
οὐρανίζειν : 155 n. I. ποδήρης : 753. προφέρειν: 118.
οὔτ᾽ οὖν : 189. ποδόψηστρον: 418. προφήτης : 497 f.
οὔτις not = οὐ; IIS. πόθεν = οὐδαμῶς : 713. πρόχειρος : 786.
οὗτος : 100 f. πόθος : 226. “πρωιρος, words ending in:
οὕτως : 684. ποινή τινος, ‘penalty of 134 f.
οὕτως εἶπε: 308 f. something’ : 630. πρῶτον καὶ μάλιστα: 326.
ὀφθαλμός : see ἐπί, ποίου χρόνου. 153. πτολίπορθος : 357.
ὄχθη, ὄχθος : 531. πόλεμος gloss on ἄτη: 341. πυρωθέν: 229 f.
πολλά, πολύ, adverbial, πῶ: 712 f.
παγκαίνιστον : 434. Aeschylean use: 214. πῶς av... introducing a
παγκρατής : 779 f. πολλὰ χαίρειν : 288. wish: 312 f.
παῖς not used of the young πολύανδρος, πολυάνωρ: 40. — — in rhetorical ques-
of beasts: 32. πολύκλαντος : 723. tion: 551.
παιών (not παιάν) Attic: 140. πολύμνηστος : 278. πωτᾶσθαι : 444.
παλαμᾶσθαι, παλάμη : 725. πόνος : 33 f. pay : 655.
παλίγκοτος : 304. πορθμός : 164 f.
ῥάπτειν, ῥαφεύς : 758.
πάμμαχος : 103 f. ποτε in dedications : 290. ῥῆσις : 415 n. 1, 616 f.
παναΐτιος : 704. mov: 112.
ῥίμφα : 214.
πάνδημος : 78. πράκτωρ : 66.
ῥυσιάζειν: 271.
πανεργέτης : 705. πράσσειν : 193, 600, 638, 647,
ῥύσιον : 270 ff,
πάντα λόγον (πυθέσθαι) : 208. 683 f.
παρὰ γνώμην εἰπεῖν : 424. πράσσειν with double accusa- σέβειν: 761 f.
παρ᾽ οὐδὲν τίθεσθαι: 130. tive, 'reddere aliquem ali- σέλμα: 109 f.
παραθέλγειν : 43. quid': 592. σκήπτειν : 161.
παραίτιος : 372. πρέπειν : 139, 171, 426, 673. σκῆψις : 402.
παράσημος: 355 f. πρέποντα: 658. σποδεῖν: 325.
πάραυτα: 342 f. πρέσβος : 390. στάθμη : 474.
παρέχειν : 715 f. πρό not adverb in A.: 456. στάσις: 505. -
πάρηξις : 279 f. προ-, prefix to verbs, tem- στεγανός : 188 f.
παροψώνημα: 686 f. poral: 81. στένειν : 143.
πατεῖν: 195, 433, 547. — meaning in προειπεῖν orißos: 215.
πατρόθεν : 714. etc.: 118. στορνύναι (στρωννύναι) : 412.
πατρῶιος : 129. προβατογνώμων: 361 f. στρουθός : 80.
πέδον πατεῖν: 639. πρόβουλος : 200. Στροφίος, accent of: 396.
πείθου, πιθοῦ : 418. πρόδικος : 232. σύγγονος : 544 f.
πειθώ : 64. πρόκωπος: 786. συλλαμβάνειν : 373.
πελανός : 54. πρόμος: 118. σύμβολον : 88.
πένθος : 225. πρόνοια : 330. συμμαχία: 123.
πέπλοι: 134. προνωπής : 788 n. 1. σύμφυτος : 91 f.
πέραν accus. of πέρα: 116. προξενεῖν: 615 n, I. σύν: 66 f., 324.
περί = ὑπέρ: 670 f. προοίμιον : 558 n. I. — denoting an accompany-
περιπίπτειν : 134. πρόπαρ: 459. ing fact or thing : 233, 637.
περιρραντήριον : 495. πρόπυργος : 532. συνάπτεσθαι: 373.
περιφρονεῖν: 671. πρὸς βίαν, πρὸς ἡδονήν etc.: συνιέναι : SOI.
περιών (not περιιών) Attic: 199. συντελής : 260.
525 n. I. πρὸς εἰδότας λέγω : 659. σφαγή: 654.
πηι not found in Tragedy: πρὸς (τῶν) θεῶν: τό. σφενδόνη : 454 f.
238. προσαυδᾶν (προσειπεῖν) : 263. σωφρονεῖν: 105.

845
 WORDS
ταγά, ταγός : 66. τοιοῦτος first syllable short: φίλος : 173. 445, 570 ff., 589.
ταῦτα denoting object οὗ ἃ 579. φλέδων : 548.
dedication : 290. τοιοῦτος referring back : 668. φοιτάς : 591.
ταῦτα (τοῦτο) introducing a τοισίδε: 265. φρένες : 447 f.
speech? : 299 f. τόλμη not in Tragedy : 196. φρενοῦν: 542.
re, adding epexegetical τρέπειν : 620. φρονεῖν: 105.
clause : 10. τρίβειν: 236, 479. φρόνημα: 343 f.
— (single) not used to de- τριβή: 117, 236. ppuxtwpta: 249 f.
note a person as ‘my τρίβος : 117. φρύκτωρος (dpuxrwpds?):
father and his brother’: τρίπους : 00, 295, 829.
744. τροπαία: 127. φύλαξ: 688.
— supposed to connect par- τυγχάνειν : 104 Í., 313. $vráAuos : 173.
ticiple and main verb: 55. τύπτειν: 81. φωνὴν λαβεῖν: 24.
τεθναίην, τεθνάναι etc. : 275. τυραννίς, τύραννος : 382, 638. φῶς: τῇ.
τέκτων : ΟἹ, τυτθός : 760. φῶς -- σωτηρία: 266.
τέλεος, τέλειος : 690 n. 4. τύχη: 516, 671.
τελεσφόρος : 440. τυχηρός : 236. χαίρειν: 142 f.
τελευτὴ γάμου: 346. τώς = ds: 139 f. χαλινοί : 136.
τέλος : 61 f., 554. χαλκεύς : 305.
τερπνός : ὃς, 40]. ὑπ᾽ αὖ pe and ὑπό μ᾽ αὖ: 557.
τεύχη : 226 f. ὕπατος: 34 f. χάμευνα: 731. [555.
χάρις : III, 181, 195, 221, 360,
«τήριον, nouns in: 495. ὑπεραίρειν : 359 f. χειροῦσθαι, χείρωμα : 619.
«τήριος, words in, substanti- ὑπερβάλλειν : 359. χίμαιρα: 133.
val: 678. ὑπεργήρως, accent of: so. χρέος : 53.
τί δ᾽ où.,.;:280. ὑπερτελής : 154 f.
τί δῆτα... drawing a con- ὑπέρφευ : 197. χρησμός : 738.
χρόνος : 604 f.
clusion : 598. ὑπερφρονεῖν : 470. χώρα: 48 f.
τί μή OF τί perv? : 325 f. ὑποκάμπτειν ; 359 f. χωρίς as predicate : 419.
τί χρῆμα; : 607 f. ὑπτίασμα: 597.
τιθέναι = facere: 207. ὑφέρπειν: 232.
τιθέναι (τίθεσθαι) with ad- ὑψίζυγος : 109 f. ψακάς, ψεκάς : 726,
verb (καλῶς etc.): 413 f., ὕψος : 646. ψευδής : 310.
Bor. ψῆγμα: 230.
τίθεσθαι τὴν ψῆφον: 20 f. φαιδρόνους : 567 f. ψυχή, ‘courage’ : 697.
τιμαλφεῖν: 417. φαιδρός : 265.
τιμάορος : 263, φαιδρύνειν, φαιδύνειν: δοο, ὦ with imperative: 15 f,
τιμή: 181, 216f., 318. 507. ὦ postponed : 284 n. 2, 829.
τις: = nonnulli: 194. φάσμα: 218. ὥμοι followed by nomina-
— in one part of ἃ τρίκωλον : φάτις : 303. tive: 635.
36. φέγγος ἔτους : 258, ὠνόμαζε: 329.
— not needed to denote φέρμα: 72. ὡς with substantival ex-
‘the person concerned’: φεῦ: 521. pression, limiting a state-
44 f., 205, 457. φεύζειν : 608. ment : 306.
τίτης : 45 f., 353. φθόνος : 81 f., 349 f. ὥς by itself for οὕτως not in
τλήμων : 606 f. φιλάνωρ : 215, 390. Aeschylus: 420, 638.
τλησικάρδιος : 225. -φιλής, adjectives in: 22. ὡσπερεί: 559 f.
τὸ μὲν πρῶτον : 200. φιλήτωρ: 685. damep οὖν: 671.
τόθεν : 120. φίλοικτος : 521. Gare causal: 401.

LATIN
annuus : 2. eminere, of details in paint- prae manu: 786.
carnifex : 747 τι. 2. ing: 139. salis : 795 n. 4.
clarus said of a wind: 541 omnipolens : 779 n. 3. soporare : 297 n. 3.
n. I. plorare: 526 n. 2,

846
 III
PASSAGES DISCUSSED
The numbers of the fragments of the elegists, iambists, and lyric poets are those of
E. Diehl, Anthologia Lyrica Graeca (vol. 1, and ed., vol. ii, 1st ed.) ; the numbers of the
tragic fragments those of Nauck, 2nd ed. ; the numbers of the comic fragments those
of Kock. The early philosophers and Gorgias are quoted after Diels-Kranz,
Fragmente der Vorsokratiker, 5th ed.
ACHAEUS fr. 29 (p. 753 N.): AESCHYLUS, Choephoroe AESCHYLUS, Eumentdes
67. (cont.) (cont.)
AESCHYLUS 842: 106f. 585-608: 634 n. 1.
Agamemnon: 2: 37. 844: 358. 611: 534 n. I.
254: 8. 845 f.: 245 634: 525.
258-354: 37. 850: 640 636: 828.
297: 6. 883: 593n.1 638: 415.
312-16: 815 n. 1. gor: 600 n.1 715: 761.
931: 37. 910: 374 784: 197 n. 1.
1304: SIN. 4. 927: 244. 785: 137n.3.
1496: 808. 983 ff.: 614 n. 2. 789: 488 n. 1.
1527 fl.: 807. 991-1006: 809 ff. 9o5f.: 211 n. 2.
Choephoroe: 63f.: 481 993-5: 813f. gızf.: 431 n. 4.
n. 2. 997: 313 n. 2. 052 f.: 590.
67: 45, 172, 205. 1000: 743 N. I. 96c ff.: 284 n. 2.
133 f.: 677 n. 3. 1011: 807. 1025: 743n. 2.
201-11: 815 fl. 1014: 703 n. 2, 800 Persae: 95: 654 n. 2.
225-30: 815 fl. 97 ff.: 129.
279: 703 n. 2. 1028: "815, 98 ff.: 189 ἢ,
330: 4481. 1048: 318 n. 1. 133 ff.: 226.
363 ff.: 699 n. 2. 1052: 214 n. I. 152: 192.
394 f.: 523. 1073: 568 n. 3. 170 f.: 656.
396: 207 n. 3. Eumenides: 7f.: 746n. 1. 181-99: 650.
405: 832. sof.: 828. 214: 448 n. 2.
411: 760. 119: 572n. I. 275: 351.
441f.: 756. 163 ff.: 515 n. I. 282 ff.: 801 n. 2.
454: 124.1. 176: 515. 309: 160n.r.
474: 691n.1. 222: 2gn. 5. 325: 172.
481f.: 239f. 238 f.: 588. 565: 420 Nn. 3.
490: 233 n. I. 245: 29n. 5. 595 f.: 233.
492: 486n. 5. 259: 197 ἢ, I. 604: 316 n. 3.
505: 228n.1,406n.1. 286: 63. 659 ff. (metre): 185
510: 556n. r1. 299: 29n.5. n. 2.
562: 564n. 302: 720.1. 720: 651 π. 2.
MH CT
€

583: 491 n. 313: 717n. 3. 749: 323n. I.

NN
+

595: 620n. 365: 58. 794: 197.

8

606: 722 n. 381f.: 737 n.r. 827: 197.

9

620: 763 n. 405: 37: 902: 57.

me
-

621: 62 n. 4. 449 f.: 654 n. 3. gızf.: 198.

628: 351. 472: 322 n. 1. Prometheus: 29: 129.
630: 240, 770 n. 2. . 476: 29 n. 5. 128ff.: 694 n. 1;
642: 639. 530: 105, 112 n. I. (metre) 185 n. 1, 829.
645: 172. 550: 560n.I. 153: 649 n. 4.
658: 317. 553: 518. 172: 33
691 fl.: 305 n. 5. 559: 790 n. 2. 181: 122.
774: 366 n. 1. 560: 8n. 3. 246: 368.
814: 325. 565: 256n.2. 514 ff.: 729.

847
 PASSAGES DISCUSSED
AESCHYLUS, Prometheus | AESCHYLUS,Supplices (cont.)| ARISTOPHANES (cont.)
(cont.) 679 ff.: 30. Frogs: 168: 564 n. 4.
520: 626. 698: 603 n. 1. 1141 fl.: 706.
526 ff.: 442 n. 2. 781 f.: 446. 1261 ff.: 58.
741: 226. 485: 234. 1359: 83n. 1.
944: 124. 791: 323. Knights: 1244: 829.
1024: 544. 846: 538 n. 1. Lysistrata: 857: 16.
1031: 159 n. 3, 307 n. 1. 850: 702. 1025 f.: 672.
1078: 649 n. 4. 908: 83f. 1287 ff.: 89 n. 2.
Seplem: 276: 756 n. 2. 925: 526. . Wasps:43: 315.
285: 130. 1042: 131. 1369: 579 Ὁ. 4.
359: 538 n.r. 1057 ff.: 105n.2, 112. 1507, 1514: 559 8.1.
363 fi: 174 f. 'epigram', fr. 4 Diehl, fr. 619: 612 n. 4.
385: Anthol, lyr.12. x, p. 8o:
426: 266 n. 3. 830. BACCHYLIDES
435: 568 n. 2. ᾿Ισθμιασταί, Pap. Oxy. 4.8: gon. 2, 829.
446: 575 n. 1. 2162, fr. 1, col. 1. 25: 4.12: 142.
498: 202. 172 n. 2. 5. τό ff.: 520 n. 1.
549: 266 n. 3. Μέμνων, frs. 127-30: 618. 5. 94 ff.: 829.
571 ff.: 680. Νιόβη, Pap. Soc. It. 1208 17 (16). 14 ff.: 29.
573: 405. (fr. 116 Mette): 129 fr. 11 Snell: 520 ἔ,
575: 678 n. 3. n. I.
579: 756 n. 1. fr.44: 372. CALLIMACHUS: hymn. 4.
10: 195. fr. 48: 85. 227: 639.
619: 80 n. 4. fr. 14: 303. CATULLUS
699: 515. fr. 78: 785 n. 1. 31. 5 fl.: 257 n. 2.
720 ff. (metre): fr. 96: 826. 64. 269 ff: 542.
727 ff.: 331. fr. 132: 58 ἢ. 4. 64. 327: 96n. 1.
729: 229. fr. 143: 1, 256, 715 n. x. | CHAEREMON: fr. 1. 7 (p.
131: 205. fr. 160: 389. 781 f. N.): 520 π. 2.
809 f.: 635 n. 2. fr. 192: 33. Ciris 470: 655n.1.
830: 80 n. 3. fr. 200: 35. CORINNA: fr. 11: 329.
854 fl: 491. fr. 205: 675 n. 1.
871 ff. : 321 n. 1. fr. 213: 82. Democritus: fr. 182: 106.
1022: 619. fr. 273: 58 n. 4. DEMOSTHENES
1056: 826. (?) fr. 304: 314.n.1. 9. 54: 334 f.
Supplices: Y: 572 n. 1. fr. 355: 186 n. 1, 21. 166: 46.
gf.: 130. fr. 362: 63. Dio CHRYSOSTOMUS: 32. 82
42f.: 722. fr. 375: 513 n. 3. v. Arnim: 646.
62: 7221. fr. ap. Athen. 13. 573 b: | Dirae 82: 655 n. 1.
64: 523 Nn. 2. 415 n. I.
86: 177n. 1. ALCMAN fr. 44: 201. EMPEDOCLES: fr. 3. 3: 378.
92: 193. ALEXIS fr. 230. 4: 573. ENNIUS
110: 128. ANDOCIDES 1.61: 398 ἢ. 2. Trag. 1. 56 Ribbeck (Scen.
118: 485. Anthol. Pal. 6. 274. 3: 378. 71 Vahlen): 347.
198: 376 n. 1, 436 n. 1. | ARCHILOCHUS 1. 6of. R. (76f. V):
282: 130. fr. 9: 425n. 1. 379.
288: 275 f., 829. fr. 30: 492. | 1. 176 R. (209 V.): 653.
308: 655 n.1. fr. 93: 68f. Epigram IG 1v. 800 (Fried-
378: 366. ARISTOPHANES lànder and Hoffleit,
386: 43. Acharnians: 318: 593. Epigrammata, no.
396: 237 0.1. 376: 376n. 1. 29): 831.
405f.: 628 n. 2. 709 : 645. EURIPIDES
472: 592. 1051 fl.: 395 n. 1. Alcestis: 213 ff.: 246n.2.
481: 55. Birds: 30 fl. : 773. 420: 274.
542: 213. 255 ff. : 131. 818f.: r30f.
621 f.: 193. 868 f.: 262 n. 4. 1072 ff.: 13.
622: 420. 1247 f. : 389. Andromacha: 361f.: 431.
638: 228n.1,437 n. 2. 1509 : 743. 467: 427 n. I.
646 f.: 76n.2. Clouds : 225 f.: 671 n. 1. 588: 782 n. i.
848
 PASSAGES DISCUSSED
EURIPIDES (cont.) EURIPIDES (cont.) HOMER (cont.)
Bacchae: 29: 317 n. 1. Supplices: 252: 393 n. 1. ® 252: 68.
571: 131 Π.1. 211 ff.: 58. a177: 209.
1021: 827. 655: 531.1. 8 340: 650.
1028: 565 ἢ. 2. 1153: 672 n. 3. (122: 228 n. 1.
Cyclops: 507: 367. Troades: 884 ff.: 100. A4231.: 534f.
Electra: 283 ff.: 823 f. 1069: 121. v 229: 568.
368: 317 nN. 1. Μελανίππη δεσμ., no. 13. 8 p57: 152n. I.
507 f.: 825. Page, Greek Lit. Pap. σ 373: 681 n.a.
518-44: 821 ff. i p. 112: 383 n. 2. xarıf.: 657.
545 f.: 826. fr. 420: 450. w351f.: 743.
945: 383 n. 2. fr. 472. 4 fl.: 180. Hymn. Hom. Apoll. οἵ:
1351: 410 ἢ. 2. fr. 696: 257. 83 n. 2.
Hecuba: | 167 f., 200 f.: fr. 791: 609. HORACE
56 n. 3. fr. 814: 762. Epodes: 10. 17: 225 n. 2.
282: 10. Odes: 1.2.1: 795 n. 4.
930: 693n. 1. 1.31.17: 179 ἢ. 2.
GORGIAS: fr. 11 a. 37: 474.
946 ff.: 693. 3. 11. 32: 690 n. 1.
Helena: 935: 179 n. 2. 4. 1. 2 ff.: 740.
1118-21: 337. Hellenistic poem, ‚Diehl, Satires: 1. 9. 72 ἴ.: 797.
1451 ff.: 698 n. 2. Anthol. lyr. ài. 297 2. 3. 48 fl.: 591 n. 2.
Hercules: 677: 185 n. 4. (Powell, Coll. "Alex.
678: 63f. p.185): 241 n. 1.
Hippolytus: 141 ff.: 52. HERACLITUS Isvcus
219: 16 n. 3. fr. 1: 133m. 1. fr. 2. 3: 681 n. 6.
759: 827. fr. 11: rir. fr. 7.3: 649.
Ion: 595: 766 n. 2. HERODOTUS Inscriptions: see Epigram.
695 ff.: 246 n. 2. 1. 63. 2: 390. CIL τῶ. 9: 690.
782: 528. . I10-12: 805. IG 122.975: 192.
μι

929: 535. . 207. 1: 112. IV. i, 616: 532.

Ὁ. μὶ
OI

1367 f.: 474. . 22.2: 95.

Iphig. A.: 578: 62n.4. . I0. €: 237.
824: 762. .140B3: 131 ἢ. 2. LUCAN 9. 719: 569.
Iphig. T.: 230: 827. 9. 79. 2: 427 n. 2. LUCRETIUS 2. 1 ff.: 408.
813: 646. HESIOD
Medea: 262: 814 n. 3. Erga: 203: 520.
752: 130. 328: 547. MENANDER
970: 7448. 355: 46. fr. 292: 179.
Orestes: 44 f.: 613 n. 1. 483 f.: ıı2n.1. fr. 550: 63.
200: 695 n. 2. 568: 522 n.2. MIMNERMUS fr. 11: 461n.2.
478: 580 ἢ. 4. Theog.: 31: 65.
497: 423 n. 2. 206: 679.
588-90: 814 n. 3. Hıpponax (?): fr. 72: 787 PHILEMON fr. 10: 63.
640 f.: 830, PINDAR
1024: 565. Isthm. 6.17: 172 Ὁ 2.
1679: 646. A106: 115. Nem.: 3. Ἰ4 1.: 335.
Phoenissae: 428: 391. B 329: 77. 6.60: 533.
630: 559 n. J. F246: 366 f. OL.: 6. 15: 284.
636 f.: 330. E 487 f.: 1gof. 9. 64: 696.
751 f.: 824. 0 183: 564. - Pyth.: 2.44: 329 n. 3.
1028 f,: 722 n. 4. Θ 548, 550-2: 532. 4. 78f.: 320 n. 1.
1104-40: 824. 163: 217. 4. 268: 228 n. 1.
1317 f: 588 n. 2. I 497 fl.: 43. 5. 15 ff.: 287.
1491: 138. K96f.: 483n. 1. 9. 29: 636 n. 2.
1724: 284 n. 4. A 674 fl.: 271. 11.9: I31 n. 3.
[Euripides] Rhesus: 219: 274. N 198 ff.: 96 n. 4. 11.17fl.: 402 n. 2.
527 ff.: 246 n.2. P 244: 564 n. 3. I2. 13: 530 n. 3.
529f.: 381 n. 2. Z 500: 790. fr. 87 Schroeder (78
724: 610f. T 86f.: 374. Bowra, 156 Turyn):
805: 6rof. Y 403 f.: 653. 192.

849
 PASSAGES DISCUSSED
PLATO SoPHOCLES, Antigone (cont.) SOPHOCLES (cont.)
Charmides 157 d: 268. 744 f.: 762. fr. 775 (859 P.): 1317.1.
Critias: 109 b-c: 111. 766: 213. fr. 780 (864 P.): 202.
109C: 78n.1. _ 1124: 130 n. 2. fr. 864 (950 P.): 63 n. 1.
Gorgias: 483 e: 342. 1146: 62 n. 4. SOPHRON: Pap. Soc. It.
493a: 2411. 1238 f.: 655. 1214 (Page, Greek
Phaedrus 243 ἃ: 669. 1272 ff.: 455 n. 4. Lit. Pap. i, no. 73):
PLAUTUS, Bacchides 368 ff.: Electra: 317: 43x n. 3. 284 f.
602 n. 3. 428 fl.: 253 f. STATIUS, Thebais 6. 235 f.:
PLUTARCH, Themistocles 29. 566 ff.: 98. 297 n. 3.
4: 4I3n. I. 686 f.: 315f. STESICHORUS fr. 15: 740,
POLYBIUS 3. 88.8: 655n. r. 853: 827. 809 n. r.
959Í.: 725n. 1.
SAPPHO fr. 48. 3: 682 n. 1. 1058 ff. (metre): 185 THEOCRITUS
Scholia n. I. 1.134: 228 n. 1.
Aeschylus: Choephoroe 1374 f.: 264f. I5. 119: 829.
74f.: 341 n. 1. Oedipus Coloneus: 315: THEOGNIS 1. 15 ff.: 698.
Aristophanes: Knights 558. THUCYDIDES
589: 634 f., 831. 1220: 682. 1. 6. 4: 696.
Wasps 270: 635 n. 1. 1258f.: 778 n.r. . 143. 4: 669.

Ne
Homer: H 295: 696 n. 2. 1622: 324. .45. 1: 426.
(21: 260. 1632: 291. .10.4: 669.
Pindar: Nem. 11. 12: 1721: 316. 48.6: 479n. 3.
SSP
544 n. I. Oedipus Rex: 449: 667. -49. 2: 479 n. 3.
Sophocles: Oed. Col. 934: 507: 142 Ὦ. I. Tragicorum fragmenta ades-
f. 616: 309. pota
‘Soon’ fr. 28: 348 n. 1. 1394 f.: 548. 90: 412.
SOPHOCLES Philocletes: 787: 558. 458: 308 n. 2.
Ajax: 289: 489 n. 4. 1066 ff.: 768 n. 3.
477f.: 151 n. 2. 1099 f.: 178 f. VIRGIL: see Ciris, Dirae.
941] ἴ.: 137 n. 2. Trachiniae: 441 f.: 613. Aeneid: 4.625: 596.
951: 107. 538: 686. Eclogues: 4. 62 f.: 717
1142 ff.: 774. 1051f.: 743. n, 3.
1353: 428. i112Í.: 699 n. 2.
Antigone: 113 f.: 189. 1233 f.: 371. XENOPHANES fr. 1. 13:
233 f.: 292. ’Ixvevrai: fr. 314. 39 366 f.
423 fl.: 32. Pearson: 580 ἢ. 4. XENOPHON
446: 424. fr. 314.73 P.: 675n.2. Agesilaus: 11. 10: 419
536 f.: 384n. 1. fr. 314. 137 P.: 319. n. 3.
668-71: 397 n. 1. fr. 694 (761 Pearson): 90. Anabasis: 5. 8. 24: 773.
692 fl.: 235. fr. 767 (851 P.): go. 6.1.23: 71.

850