A COMMENTARY ON

VIRGIL:
AENEID VIII
MNEMOSYNE
BIBLIOTHECA CLASSICA BATAVA

COLLEGERUNT
W. DEN BOER • W. J. VERDENIUS • R. E. H. WESTENDORP BOERMA
BIBLIOTHECAE FASCICULOS EDENDOS CURAVIT
W. J. VERDENIUS, HOMERUSLAAN 53, ZEIST

SUPPLEMENTUM TRICESIMUM QUINTUM

P.T.EDEN

A COMMENTARY ON
VIRGIL: AENEID VIII

LUGDUNI BATAVORUM E. J. BRILL MCMLXXV

 A COMMENTARY ON

VIRGIL:
AENEID VIII
BY

P. T.EDEN

LUGDUNI BATAVORUM E. J. BRILL MCMLXXV

 ISBN 90 04 04225 3

Copyright 1975 by E.]. Brill, Leiden, Netherlands

All rights reserved. No part of this book may be reproduced or

translated in any form, by print, photoprint, microfilm, microfiche or
any other means without written permission from the publisher

Printed in Great Britain

at the University Printing House, Cambridge
(Euan Phillips, University Printer)
 DIS MANIBUS
RHE
PATRIS BENEVOLENTISSIMI
SACRUM
 CONTENTS
Preface IX
Works cited and abbreviations used . XI
Introduction. xv
Aeneid VI I I: the material XVII
Aeneid VI I I: structure and coherence xx

Commentary I
Appendix. Metre and Verse 1 93
Indexes 202
 PREFACE

This commentary may be used with the text and apparatus of any
edition of Aeneid VIII. Textual problems of significance are fully
discussed, and an ample number of specimens provided of the treat-
ment to be given to the aberrant and trivial.
'A great deal of the commentary on Virgil has become tralatician ',
says Mr John Sparrow (Half-lines and Repetitions in Virgil 138), but
to ignore it is folly. My debt to earlier commentators and critics,
Servius in primis, Heyne and Wagner, Conington and Nettleship,
Page, Mackail, Norden, and Austin, will be evident, if only from dis-
agreements with them. But of course there are new questions to be
raised, and more often new solutions to be offered to some very old
problems.
Virgil's many-sided achievement is unique in the literature of the
classical world. He stood at the convergence of diverse cultural forces
and literary traditions and harmonised them. He imposed on suc-
ceeding generations a linguistic revolution of his own making (so that
frequently, like Heyne (on Aeneid VIII, 30), monemus hoc, quod
plerique verba tenent, causas sermonis ignorant). He achieved the
definitive perfection of his chosen verse-medium. The individual
factors contributing to these results acquire significance only if they
are seen in their contexts of development and association. No
apology, therefore, is offered for the detailed scrutiny to which some
of them are subjected; on the contrary, I am very conscious of how
inadequate some of the discussion may appear to experts in indi-
vidual fields.
Entirely subjective reactions to Virgil's poetry, especially if based
on standards quite alien to him, as is fashionably the case, may
inspire enthusiasm, but of the wrong kind and for the wrong reason:
they will not be found. Observations on such topics as allegory,
symbolism and structure attempt to reflect the circumspect modesty
of the scholars who first drew attention to these aspects of Virgil's
composition (connoisseurs of the Virgilian literature of the last two
decades will be aware of bibliographical omissions: the reason can
usually be found in reviews, by myself and others).
I am deeply in debt to the help of friends. Professor R. D. Williams
gave me the benefit of his ripe wisdom at an early stage. Dr N. M.
X PREFACE

Horsfall checked the details of a later draft with painstaking

thoroughness. Miss Susan French prepared the final typescript with
exemplary accuracy. Mr P. R. Home gave unstinting care to proof-
reading. The unfailing courtesy of the publishers and the skill and
efficiency of the printers relieved me of many anxieties in the final
stages. I offer them all my sincere thanks.
London
New Year's Day 1975 P. T. E.
 LIST OF CHIEF WORKS CITED AND
ABBREVIATIONS USED

The following is a list of authors and works referred to in the commentary

in an abbreviated form; it may therefore serve as a highly select biblio-
graphy. An author's name alone in the commentary refers to his work listed
below. Abbreviations for some individual works are shown in brackets.
References to periodicals follow in general the system of L'Annee Philo-
logique. ALL = Archiv fur lateinische Lexicographie; CAH = Cambridge
Ancient History; CIL = Corpus Inscriptionum Latinarum; LS= Lewis
and Short's Latin Dictionary; OCT = Oxford Classical Text; PW = Pauly-
Wissowa (-Kroll), Real-Encyclopiidie; TLL = Thesaurus Linguae Latinae.
Alfoldi, A. Early Rome and the Latins, University of Michigan, 1965.
Austin, R. G. Aeneid I, Oxford, 1971.
Austin, R. G. Aeneid II, Oxford, 1964.
Austin, R. G. Aeneid IV, Oxford, 1955.
Axelson, B. Unpoetische Worter, Lund, 1945.
Bailey, C. Religion in Virgil, Oxford, 1935. (RV)
Binder, Gerhard. Aeneas und Augustus, Meisenheim am Gian, 1971.
Camps, W. A. An Introduction to Virgil's Aeneid, Oxford, 1969.
Carcopino, J. Virgile et les Origines d'Ostie, Paris, 1919.
Cartault, A. L'Art de Virgile dans l'Eneide, Paris, 1926.
Conington, J. and Nettleship, H. P. Vergili Maronis Opera, ed. J.
Conington, vol. III, 3rd ed., revised by H. Nettleship, London, 1883.
Conway, R. S. Aeneid I, Cambridge, 1935.
Daremberg, C. and Saglio, E. Dictionnaire des antiquites grecques et
romaines, 9 vols., Paris, 1877-1919.
Donatus, Tiberius Claudius. Interpretationes Vergilianae, ed. H. Georgii,
2 vols., Leipzig, 1906.
Ernout, A. and Meillet, M. Dictionnaire etymologique de la langue latine,
4th ed., Paris, 1959. (E-M)
Ernout, A. and Thomas, F. Syntaxe Latine, 2nd ed., Paris, 1953.
Fordyce, C. J. Catullus, Oxford, 1961.
Fowler, W. Warde. Aeneas at the Site of Rome, Oxford, 1917. (ASR)
Fowler, W. Warde. The Religious Experience of the Roman People, London,
1922. (RE)
Gjerstad, E. Early Rome, 3 vols., Lund, 1953-60.
Grassmann-Fischer, B. Die Prodigien in Vergils Aeneis, Munich, 1966.
Heinze, R. Virgils epische Technik, 3rd ed., Leipzig, 1928. (VeT)
Henry, J. Aeneidea, 5 vols., Leipzig-Dublin-Meissen, 1873-1892.
Heyne, C. G. and Wagner, G. P. E. Virgilii Opera, 4th ed., 5 vols.,
Leipzig and London, 1830-41.
Hirtzel, F. A. P. Vergili Maronis Opera, Oxford, 1900.
Knauer, G. N. Die Aeneis und Homer, Gottingen, 1964.
XII WORKS CITED

Kuhner, R. and Stegmann, C. Ausfuhrliche Grammatik der lateinischen

Sprache: Satzlehre, 3rd ed., 2 vols., Leverkusen, 1955. (KS)
Latte, K. Romische Religionsgeschichte, Munich, 1960. (RRG)
Leumann, M. Kleine Schriften, Zurich, 1959.
Leumann, M., Hofmann, J.B. and Szantyr, A. Lateinische Grammatik:
Syntax und Stilistik, Munich, 1965. (LHSz)
Lofstedt, E. Syntactica, Lund: vol. I, 2nd ed., 1942; vol. II, 1st ed.,
1933.
Mackail, J. W. The Aeneid of Virgil, Oxford, 1930.
Maguinness, W. S. Aeneid XII, 2nd ed., London, 1960.
Munzer, F. Cacus der Rinderdieb (Programm), Basel, 19n.
Mynors, R. A. B. Vergili Opera, Oxford, 1969.
Nash, E. Pictorial Dictionary of Ancient Rome, z vols., London, 1961.
Neue, F. and Wagener, C. Formenlehre der lateinischen Sprache, 3rd ed.,
4 vols., Leipzig, 1902-5. (N-W)
Nisbet, R. G. M. and Hubbard, M. A Commentary on Horace: Odes Book 1,
Oxford, 1970. (N-H)
Norden, E. Aeneis, Buch VI, 3rd ed., Leipzig, 1926, reprinted 1957.
Ogilvie, R. M. A Commentary on Livy 1-5, Oxford, 1965.
Otis, Brooks. Virgil: A Study in Civilized Poetry, Oxford, 1963.
Page, T. E. Virgil, 3 vols., London, 1894-1900.
Pallottino, M. The Etruscans, 3rd Italian ed., translated by J. A. Cre-
mona, Penguin Books, 1955.
Palmer, L. R. The Latin Language, London, 1954. (LL)
Pease, A. S. Aeneid IV, Harvard University Press, 1935.
Peerlkamp, P. Hofman. Virgilii Aeneidos Libri, Leiden, 1843.
Peter, H. Historicorwm Romanorum Fragmenta, Leipzig, 1883. (HRF)
Platnauer, M. Latin Elegiac Verse, Cambridge, 1951.
Platner, S. B. and Ashby, T. A Topographical Dictionary of Ancient
Rome, Oxford, 1929.
Posch!, V. Die Dichtkunst Virgils, Bild und Symbol in der .Aneis, Inns-
bruck, 1950. (English translation by Gerda Seligson: The Art of Vergil:
Image and Symbolin the Aeneid, University of Michigan Press, 1962.)
Raven, D.S. Latin Metre, London, 1965.
Rehm, B. Das geographische Bild des alten Italiens in Vergils Aeneis,
Philologus Suppl. 24, 2; 1932.
Ribbeck, 0. Vergili Opera, 2nd ed., 4 vols., Leipzig, 1895.
Roscher, W. H. Ausfuhrliches Lexicon der griechischen und romischen
Mythologie, Leipzig, 1884-1937.
Rose, H. J. Ancient Roman Religion, London, 1948. (ARR)
Sabbadini, R. Vergili Opera, z vols., Rome, 1930.
Saunders, C. Vergil's Primitive Italy, Oxford, 1930.
Scollard, H. H. The Etruscan Cities and Rome, London and Southampton,
1967.
Servius and Servius Auctus: ed. G. Thilo and H. Hagen, Leipzig, 1878-
1902.
Sparrow, J. Half-lines and Repetitions in Virgil, Oxford, 1931.
 WORKS CITED XIII

Stanford, W. B. The Odyssey of Homer, 2 vols., London, 1947-8.

Tilly, B. Vergil's Latium, Oxford, 1947.
Vahlen, J. V. Ennianae poesis reliquiae, 3rd ed., Leipzig, 1928. (V)
Wagner, G. P. E. Quaestiones Virgilianae, vol. 1v,part2ofHeyne-Wagner,
q.v. (QV)
Walde, A. Lateinisches etymologisches Worterbuch, 2nd ed., Heidelberg,
1910. (W)
Weinstock, S. Divus Julius, Oxford, 1971.
Wetmore, M. N. Index Verborum Vergilianus, Yale-London-Oxford,
19n.
Wilkinson, L. P. Golden Latin Artistry, Cambridge, 1963.
Williams, R. D. Aeneid III, Oxford, 1962.
Williams, R. D. Aeneid V, Oxford, 1960.
Winbolt, S. E. Latin HexameterVerse, London, 1903.
Wissowa, G. Religion und Kultus der Romer, 2nd ed., Munich, 1912.
(RuK 2 )
Woodcock, E. C. A New Latin Syntax, London, 1959, reprinted 1960 and
1962.
INTRODUCTION
 AENEID VIII: THE MATERIAL

In skeletal terms of the main narrative of the Aeneid rather less

than ten lines of Book VIII (494-6, 603-7) actually forward the story,
those that describe Aeneas' departure to take command of the leader-
less Etrucans; and from this very limited and superficial viewpoint
the action of the book could be entitled 'Aeneas' Departure'. To
inaugurate the 'Iliadic' half of the Aeneid Virgil had to provide
Aeneas with troops, and in sufficient number to withstand the com-
bined Latin allies; there was no plausible alternative to the Etruscans,
and Virgil chose them, even though his choice involved him in a total
recasting of the traditional account of the early relations between
Latium and Etruria.
Other incidents of the book, however, are so closely connected
with the main narrative that they must have been part of Virgil's
original conception and appeared in his prose-sketch of the Aeneid.
Aeneas' arrival at Pallanteum and the prospect of assistance is pre-
dicted by the Sibyl of Cumae at A. 6.96 f. via prima salutis, / quod
minime reris, Graia pandetur ab urbe, and the death of Pallas, which
bulks large in A. IO (479-509) and II (1-181), is foreshadowed in this
book (520 ff., 578 ff.). Virgil clearly had Evander, 'The Benefactor', in
mind at an early stage of composition as occupant of the site of
Rome. Legend had placed him there at least by the third century B.c.
(Fabius Pictor, frg. 5 b, Peter HRF); by Augustan times the Arcadian
foundation was an important rival to the story of Romulus (it may
possibly preserve the dimmest memory of Greek Bronze Age migra-
tions to Italy). Evander was available to act as mediator between
Aeneas and the Etruscan forces, and to offer his only son to Aeneas'
tutelage.
The simple relationship Pallanteum-Evander and Pallas-the
Etruscans was the nucleus for the first stage in selecting material for
the book. The next stage had two problems to solve: how to bring
Aeneas from the Tiber-mouth to the site of Rome, and how to
despatch him from there to the Etruscan host. Dry annalistic
reporting could not have made much of either, and the challenge to
Virgil's imagination and grasp of epic technique was one of the most
demanding in the whole poem. Two Odyssean situations suggested
themselves as germ-ideas for the expansion of the journey to Pal-
XVIII THE MATERIAL

lanteum, and (more surprisingly, for there was no necessary or natural

connection with the main narrative) the themes of two Iliadic pas-
sages, much longer and better-known, for incorporation after the
arrival at Pallanteum.
Odysseus, having swum after shipwreck to a river-mouth on
Phaeacia (?Corfu) (Od. 5.441 ff.), prayed to the river-god, who checked
his current and allowed him to land safely: on this slender basis
(helped perhaps by a legend that Aeneas was warned by a god of his
country and by the example of river-gods actually speaking in
Homer and Hellenistic poetry), Virgil devised the splendid vision of
the Tiber-god, 'the ruler of the rivers of Italy', to confirm that Aeneas
had at last reached the promised land. Aeneas glides up the stilled
river and reaches a party engaged on sacrifice and feasting on the
river-bank-just as Telemachus had arrived by boat when Nestor and
his people were similarly engaged on the sea-shore (Od. 3.1 ff.).
Nestor was (appropriately) sacrificing to Poseidon :1 Virgil's recasting
of this suggests that he already had in mind the epic possibilities, to
be exploited more fully later, of the oldest and most venerable
relics of the very centre of the ancient city and the customs associated
with them. A natural place for Aeneas to land would be in the
neighbourhood of the forum boarium, the cattle-market, where river-
transport from the interior had moored in the oldest times: nearby
was a group of religious buildings, but one pre-eminently was suitable
for the scene of sacrifice by the Greek Evander and his Arcadian
settlers-the Ara Maxima with its cult of Hercules more Graeco. The
sacrifice is in progress when the Trojans arrive and after diplomatic
exchanges, recognition, and renewal of friendship, they are invited to
participate. Virgil describes the sacrifice in Homeric detail: it is the
longest account of any religious ceremony in the Aeneid, and for some
of its peculiarities Virgil is our sole source. But Hellenistic-Roman
interest centred as much on the origin of religious customs as on their
ritual, and in one of his most skilful compositions, Virgil dramatised
the myth which had grown out of the associations of monumental
relics near the Ara Maxima, and created the story of Cacus the fire-
breathing monster whose inhuman atrocities were ended by the arrival
of the superman Hercules. The story is told by Evander in the interval
between the morning and evening sacrifice, and becomes a narrative

1 For details of the extensive debt of the first half of Aeneid VIII to Odyssey
3 see the note on 97 ff.
 THE MATERIAL XIX

inset, a 'little epic' in form, in the middle of the description of ritual

which ends with a hymn composed very much in the Greek style.
Evander next conducts Aeneas back to his humble dwelling on the
Palatine, discoursing first on the prehistory of Italy and then on the
antiquities which they pass. Evander's excursus on prehistory is a
synthesis of ideas from Hesiod, Lucretius and folk-myth, some of
which Virgil had already worked out in other contexts; but by trans-
ferring the topics of antiquarian research to epic, Virgil promoted the
commonplace of Augustan poetry which dwelt on the primitive features
of Rome (many of which were saved from extinction by Augustus'
conservative policies), and the contrast between the virtuous
insignificance of the past and the degenerate grandeur of the
present.
Night falls and the real business of the skeleton narrative must be
deferred to the following day, when Evander discloses the source of
help for Aeneas. The long account of how this comes to be available is
indispensable: Virgil had to establish his own novel legend of the
Etruscans rising in revolt against their king Mezentius and seeking a
foreign leader in their crusade against him. But for no necessary
reason, probably from sheer emulation of the most celebrated and
elaborate descriptive inset in Homer, Virgil had also decided that
Aeneas should have a decorated shield, which must of course be
delivered to him before the serious Iliadic fighting. Achilles' mother
Thetis had asked Hephaestus to make her son a shield after he had
lost his armour (Il. 18.457 ff.); with no such motivation Venus asks
Vulcan to make her son one as well. But Thetis was a minor marine
deity with no special relationship to Hephaestus; Venus on the
contrary was a major Olympian goddess, a chief protagonist in the
story, and the wife of Vulcan. The new relationship required more
than a perfunctory request, and the appropriate time for divine
colloquy was the night, when the human action of epic-except in
very unusual circumstances-was suspended. The meeting between
Venus and Vulcan owes its erotic overtones to that between Hera
and Zeus (in Il. 14.153 ff.), but Venus' set speech, a rigorously per-
suasive suasoria, is a faithful echo of the Roman schools of rhetoric.
Vulcan is won over and departs, not to make the shield himself, like
Hephaestus, but to supervise his servants the Cyclopes at work in
their smithy under the Vulcanian islands: here Virgil adapts an idea
from Hellenistic poetry and elaborates it (rather mechanically in
426 ff.). When the shield is made and delivered by Venus, after
xx THE MATERIAL

ominous thunder and lightning, the pictures on it, res Italas Romano-
rumque triumphos-Roman victories over moments of peril and crisis,
contrast with the quiet idyllic themes of the earlier part of the book in
exactly the reverse manner to that in which the peaceful activities on
Achilles' shield contrast with the scenes of fighting which surround it
in the Iliad; the pictures, the narrative of the whole book, and, in its
contemporary relevance, of the whole Aeneid, culminate in a centre-
piece representing the battle of Actium and Augustus' victorious
return to Rome. Virgil had already introduced contemporary refer-
ences into the Eclogues and Georgics, but to incorporate such themes
into epic poetry, as here and at A. 6.789 ff. and 875 ff., was a very
bold innovation.
This, in brief, seems to be how Virgil evolved material to fit in with
the basic design of his first stage of composition. We do not know in
what order or at what intervals of time the various sections of the
book were written; some of them, the Cacus-narration, the Shield,
Evander's excursus on Italian prehistory and the archaeology of the
site of Rome, can in some respects be regarded as self-contained
episodes.

AENEID VIII: STRUCTURE AND COHERENCE

The apparent lack of any overriding theme and of any unity of

composition in the final form of the book has proved a headache to
interpreters. It stands in the sharpest contrast to Aeneid VII whose
material is arranged according to a very clear design: in the first
main section of the book the Trojans and Latinus alternate as the
subjects of attention until peace is concluded between them (A. 7.285);
Juno, the symbol of furor, immediately intervenes, and in the second
section her instrument Allecto makes three attempts to disrupt the
peace: they are all successful, and their practical results follow in
reverse order to culminate in the third section of the book, the
mobilisation of Italy and the Catalogue of the Italian Allies. 1
Aeneid VIII on the other hand is episodic; it is sometimes possible (as
at 268 ff.) to see where one section has been stitched to another; the
subject-matter of the episodes is quite different and does not sustain
any one clearly directed theme like the episodes of Aeneas' wander-

1 An excellent detailed analysis is given by Eduard Fraenkel, 'Some Aspects

of the Structure of Aeneid VII', ]RS 35 (1945). 1-14.
 STRUCTURE AND COHERENCE XXI

ings in Aeneid III. In short, no clear structural scheme, comparable

with that of Aeneid III or VII, can be found in the more ambitious
Aeneid VIII. But this is not to say that in the extraordinary variety
of material in the book there is no coherence, no progress in the
narrative, no dominant themes which reflect each other. All these
are present, and more too.
Two vital sequences of ideas, mutually indispensable and inter-
woven, give momentum to a narrative which begins with its hero in
a state of solitary despair, friendless and futureless, and ends with
the prospect of countless thousands to command, and a vision of
world-empire. The crucial point where these two sequences coincide
(520-40) has all the force of a dramatic m,pml-ma, 'a reversal of
fortune'.
Aeneas in despair (19, 29, 35)
I
Tiberinus points to Evander
as the source of help (51 ff.) Venus persuades Vulcan to
make invincible arms for
I
Evander in turn points to the
Aeneas (400 ff.)
I
Etruscans(470--519) ........................ Venus sends her sign: arms
I are seen to flash in the
sky (520 ff.)

I
Aeneas joins the Etruscans
(606 f.) 1
Venus brings the divine
armour (612 ff.), which
depicts the future great-
ness of Rome '

Evander's long speech (470-519) promised Etruscan support, but it

had filled Aeneas with gloom: could he really win over a nation in
revolt, and could he face the effort? ... the issue in any case would
still be doubtful. Immediately afterwards Venus sends her sign, and
Aeneas comes to the triumphant realisation of heaven's blessing, and
a change of heart comparable only with that of Anchises in A. 2. 701 ff. :
iam iamnulla mora est: sequor et qua ducitis adsum, / di patrii. In
other circumstances the sign might have been ominous of disaster,
portending the bloodshed and desolation of war, but Aeneas, now in
fully awakened consciousnesss of his mission, and psychologically

1 Cf. R. Heinze, Virgils epische Technik,~449.

2 EAC
XXII STRUCTURE AND COHERENCE

ready to accept it, himself interprets the sign as the most favourable
possible. Hence his exultant cry (533) ego poscor Olympo :1 he and no
one else really is the hero quem numina poscunt, the foreign leader
chosen by fate, the true object of the theme which develops out of
Aeneid VII and drums insistently through Aeneid VIII (12,477,503,
5n ff.). 2 The reversal of fortune is not, as it might have been, a tragic
change for the worse, but a serious joyful and self-confident change
for the better.
But long before Aeneas can avow his change of heart to himself,
there must be some catalyst to transform his outlook; without this
the heavenly sign could have no effect: this is, in fact, the symbol of
that change, even if the art of story-telling presents it as the cause,
just as the love-potion drunk by Tristan and Isolde is a striking
dramatic externalisation of their final awareness of the love which
has slowly and unconsciously grown between them. The ground for
this change is prepared above all by Aeneas' unsuspecting introduc-
tion, with his emotions stirred by a hallowed religious ceremony, to
the Arcadian bucolic atmosphere of the Arcadian site of Rome.
Nestor's discursive reminiscences of the Trojan War and the home-
coming of the heroes (Od. 3.102 ff.) gave Evander precedent to
reminisce himself. But the Hercules-Cacus episode and the excursus
on the archaeology of the site of Rome appear at first sight to be
entirely detachable from the main narrative traced above; in fact,
the depth and reflection they give to it are indispensable.
A comparison with the versions of Livy (1.7.5-15) and Dionysius of
Halicarnassus (1.39) clearly shows that Virgil chose his material for
the Hercules-Cacus episode specifically to enhance the monstrosity of
Cacus, and hence the achievement of Hercules the deliverer, the
worthy object of deification. There are two other monsters in the
book, one of primitive bestiality, Mezentius, the other, in official
eyes, a tainted immoralist and instigator of civil war, Mark Antony.
Aeneas accepts the commission to defeat the first, Augustus triumphs
over the second: both rank as divine saviours and benefactors of
mankind. This major theme organises the structure of the book in
three layers or panels: it is clearly exampled in the mythical past with
Hercules, projected on to the narrative present with Aeneas, and on to

1 The significance of this short phrase has unfortunately been obscured by

the perverse punctuation of editors like Hirtzel; see the note on 533.
1 Emphasised by Franz Bomer, 'Studien zum VIII. Buche der Aeneis',
RhM 92 (1944), 319-69.
 STRUCTURE AND COHERENCE

the historical present with Augustus. 1 And it is this theme which

binds together sections which have no formal similarity to each
other in length or literary genre, and which appear most to disrupt
the unity of the book if considered purely from the point of view of
their literary antecedents: 'little epic' (the Cacus-narration), nar-
rative sketch (the site of Rome), and descriptive inset (the Shield).
There are other symbolic links of the same kind, but not so pervasive:
Saturn, Evander and Aeneas, all foreign immigrants, are the past,
present and future colonisers and civilisers of Latium (a link fully
discussed by Binder 84-9); and Evander and Augustus reflect on
each other through their humble quarters on the Palatine, and the
institution and preservation of the priestly colleges of the Salii and
Luperci.
Aeneas is made to feel at home on the site of Rome, but this is only
half of his introduction to res Romanae in the book; after his change
of heart he takes on himself the future destiny of the city and its
glorious heroes and history shown on his Shield. The link between the
encomium urbis antiquae and the laudes virorum is as close and sig-
nificant here, although they are separated by quite different material,
as it is in the famous praise of Italy in G. 2.136 ff. where the pane-
gyric of the land is followed immediately by supreme examples of the
genus acre virum. 2 In other words, the Site of Rome and the Shield
complement each other materially: 'by showing us first of all the

1 The importance of the Hercules-Augustus equation was first suggested by

Hermann Frankel, Miscellanea G. Galbiati = Fontes Ambrosiani XXV, vol. 1
(1951), 127 f.,laterdeveloped by Hermann Schnepf,' Das Herculesabenteuer in
Virgils Aeneis', Gymnasium 66 (1959), 250 ff., and amplified by Brooks Otis,
Virgil: A Study in Civilized Poetry (1963), 330 ff. Vinzenz Buchheit, Vergil
uber die Sendung Roms (Heidelberg, 1963), u6 ff., concentrates attention on
the Hercules-Cacus episode not as the key to internal coherence within
Aeneid VIII, but to the whole book's connection with the rest of the Aeneid:
the clearly intended parallelism between Hercules and Aeneas is convincingly
demonstrated, but the Cacus-Turnus equation involves considerable difficul-
ties. That Cacus and Mezentius are so characterised as to reflect on each other,
and that this reflection supports the structural coherence of Aeneid VIII,
seems to me demonstrably true. That images and phrases used in connection
with Cacus hint unambiguously at the character of Turnus, and that this cross-
reference integrates this book with the rest of the second half of the Aeneid,
seems much less solidly based, although this view is gaining acceptance (see
F. J. Worstbrock, Elemente einer Poetik der Aeneis. Untersuchungen zum
Gattungsstil Vergilianischer Epik (Miinster Westf., 1963), 103 ff., and G. Karl
Galinsky, A]Ph 87 (1966). 18-51, esp. 35 ff.).
2 See the most brilliant overall appreciation of Aeneid VIII to date, the

essay by Friedrich Klingner in Virgil: Bucolica, Georgica, Aeneis (Ziirich,

1967), 527 ff.
2·2
XXIV STRUCTURE AND COHERENCE

empty theatre Virgil gives us an inkling that great things are to

happen there later, and has filled us with an expectation which is only
waiting to be transformed into admiration: at the sight of the shield
this admiration comes to flower' .1 And it is in these two sections that
particular stress is placed on Aeneas' similar reactions: joyful
wonder as he drinks in the scenery of Rome (310 f.), fascinated
wonder as he sees the divine armour (617 ff.). 2
Finally, throughout the book the listener's interest is continually en-
gaged by the ebb andflowoffeelingswhichcannot be reduced to a simple
schema, but which none the less secure unity and coherence: mood,
emotional intensity, and the light and shade of the colour-painting.
The mood of the hero is never the same for long. The gloom and
despondency which envelop him when he falls asleep on the river-
bank are lifted by the marvellous apparition of the river-god and his
paternal reassurance. The joy which Aeneas feels at Evander's
friendly welcome and the ceremony which confirms it gives way to
relaxed surrender to the pastoral charm of the site of the settlement.
The doubt and bafflement at the unexpected hazard of joining the
Etruscans is banished by a surge of triumphant self-confidence
through Venus' sign, and finally succeeded by astonished half-
awareness of the significance of the Shield.
The range of emotional intensity is no less varied: the flat, prosaic
resume of the opening paragraph and the stilted diplomacy of
Aeneas' credentials soar to Vulcan's incoherent exhaustion and-the
real emotional climax of the book in human terms-Evander's
passionate farewell to his son.
Light and brightness of colour pervade the book from the tran-
quillity of the Tiber valley to the triumph of Actium, and win their
alternating battle with the sombreness evident in Turnus' prepara-
tions for war, the monstrous atrocities of Cacus and Mezentius, and
the premonition of Pallas' fate. 3 All but one of the similes in the book
contribute light and colour, and artfully offset moods of depression
in the narrative: Aeneas' thoughts may be gloomy, but they are
restless too like shimmering light; Pallas may be doomed, but he sets
off in the pure beauty of the Morning Star.
1 A. Cartault, L'Art de Virgile dans l'Eneide (Paris, 1926), 634.
2 See J. R. Bacon, •Aeneas in Wonderland: A Study of Aeneid VIII', CR 53
(1939), 97-104.
3 See Viktor Poschl, The Art of Vergil: Image and Symbol in the Aeneid,
(Michigan, 1962), 169 ff., some of whose observations, however, will not stand
up to pragmatic tests.
 STRUCTURE AND COHERENCE XXV

The deeper one looks into Aeneid VIII the more one sees. Enough
has perhaps been said not only to vindicate it against any superficial
charge of disunity, but to encourage the reader to look for himself
and penetrate multiple layers of meaning, the surest criterion of a
great work of literature.
COMMENTARY
 1-17 Turnus gives the signals for war, and the whole of Latium swears
allegiance to him inspired by the same uncontrollable frenzy. Allied chieftains
raise levies and an envoy is sent to the Greek Diomedes to win his support
against the Trojan newcomers.
1-17 This opening paragraph is at once a summary, a transition, and
a setting. Book VII ended with the Catalogue of Latin Allies (641-end),
a lengthy interruption of the narrative; the last part of the narration
proper (601-40) had described what Augustan poetry represented as the
time-honoured procedure for a formal declaration of war, the opening of
the gates of the temple (actually a double archway) of Janus by the
consul: ipse Quirinali trabea cinctuque Gabino / insignis res erat stridentia
limina consul, / ipse vocat pugnas; sequitur tum cetera pubes, / aereaque
adsensu conspirant cornua rauco (6r2-5). Latinus himself had refused to
perform this regular office. War therefore was declared in the panic-
stricken atmosphere of an emergency: the gates had been pushed open by
Juno and immediately there was feverish mobilisation and armament,
cf. esp. signaque ferre iuvat sonitusque audire tubarum (628); classica
iamque sonant, it bello tessera signum (637). The opening lines of A. 8 (1-6)
draw together these threads from the preceding narrative, adding Turnus•
own summons to repel the invaders-hasty and unceremonious in accord-
ance with his character. Three of the heroes already characterised in the
Catalogue of Latin Allies raise levies (6-8)-a grim trio: Mezentius,
arrogant and cruel (A. 7.647 ff.), Messapus, descendant of Neptune and
supernaturally protected (A. 7.691 ff.), and Ufens, victorious chieftain of
brigands (A. 7.744 ff.). Finally (9-17)-a new addition to the narrative-an
ambassador Venulus is sent to try by a subtly persuasive speech to enlist
the aid of Aeneas' old enemy Diomedes now settled in Apulia: note that
A. 8.12 echoes A. 7.272 f. where Latinus says hunc illum poscere fata /
.. . rear.
The first two sections of this opening paragraph (1-6, 6-8) are resumptive,
and the last (9-17) gives in formal terms the substance of a diplomatic
dispatch (cf. A. ro.149 ff., Aeneas' audience with Tarcho-a similar
situation with very similar phraseology): the style is inevitably un-
impassioned and almost prosaic. But it is an indispensable part of the
narrative: the reader needs to be impressed at the beginning of the book
that war is everywhere being prepared (giving Aeneas full cause for
sleepless anxiety, 18 ff.), and to see the contrast with the tranquil and
unwarlike themes of the rest of the book.
1-6 Narrative accelerando and dramatic tension in this opening sentence
are achieved by a skilful fitting of balanced words and clauses on to the
hexameter framework. The three parallel clauses of decreasing length
introduced by ut speed the account of Turnus' actions to an almost tele-
grammatic expression of their results: words required by a full prose
expression (sunt with turbati and et before simul) are omitted, the balancing
4 COMMENTARY

word-order in omne tumultu coniurat trepido Latium (adj. 1, noun 2,

verb, adj. 2, noun 1) spans two hexameters with no breath-pause, and the
whole sentence ends with a heavy stop after a first dactylic foot (effera),
a favourite means of conveying energetic haste (extulit (2) followed by a
lighter pause has a less forceful effect of the same kind).
On stopped first foot dactyls see R. D. Williams on A. 5.480.
The opening sentence is exceptional (and the sentence in 9-17 even
more so) in exceeding the four-hexameter unit which Virgil regarded as the
normally desirable maximum. As pointed out above, the first sentence is
a resume of a series of events; the other, 9-17, recalls by its structure the
long periods of historical narrative in prose (such a long sequence of
Oratio Obliqua is rare in Virgil)-and the tone of it too is unpoetic and in
keeping.
1 Tumus, Aeneas' principal antagonist, makes his appearance in the
first line, although his name is not the seal of the whole book as it is of
A. 12 (cf. regina A. 4.1). Latinus had refused to make a formal declaration
of war and Tumus himself takes measures designed to meet an emergency
situation. Virgil is evidently thinking of a custom of Republican Rome.
Servius (ad loc.) tells us that in the case of a tumultus (as here, cf. 4), a
sudden uprising in Italy or Cisalpine Gaul (so defined by Cic. Phil. 8.3),
there was no time to administer individually the military oath (sacra-
mentum), so the man who was going to command the army went to the
Capitol (cf. ab arce 1) and raised one or possibly two vexilla, uttering the
formula qui rempublicam salvam esse vult, me sequatur. The oath was then
taken together (coniuratio cf. coniurat 5) by all those of military age,
i.e. those who were technically iuvenes and constituted the iuventus (5).
Messengers were also despatched to various places in the neighbourhood to
raise levies (evocatio), a task which is here performed by the three chieftains
Messapus, Ufens and Mezentius. See further Th. Mommsen, Romische
Forschungen II, 245 note 27.
1 Laurenti: noun or adjective? Only this form of the word is am-
biguous: in all its other forms Laurent- in Virgil is invariably an adjective.
At A. 8.38, solo Laurenti, it is most naturally taken as an adjective because
of the balancing arvisque Latinis, and because of solo Laurente at A. 12.547;
the only other occurrence is A. 12.769 Laurenti divo, where, as here, there
is no good ground for not taking it as an adjective. Virgil lends no support
to the view that there ever was a primitive town called Laurentum, and
there is no clear evidence for its existence either in inscriptions (see Dessau,
CIL 14, pp. 186 ff.) or in literary sources (see Carcopino 198 ff.). It is now
generally believed that the capital of the ager Laurens and the populus
Laurens was Lavinium, the modem Pratica di Mare, about 16 miles south-
east of Ostia and three miles back from the coast (evidence for the identifi-
cation is summarised in Saunders 55 f.). The phraseology of Tibullus
2.5.49 points to the same conclusion: ante oculos Laurens castrum murusque
Lavini est, 'a camp in Laurentine territory and (there also) the wall of
Lavinium '. But it is by no means certain that Virgil identified Lavinium
with the city of Latinus; A. 12.193 f., on the contrary, imply that it was a
 COMMENTARY 5

later Trojan foundation, and a plausible case for believing that Virgil lo-
cated Latinos' city near the marshy sea-coast by the mouth of theNumicus
(Rio Torto or, more probably, Fosso di Pratica) is made by Tilly 83 ff.
2 The alliteration of c and r here and in the similar A. 7.615 may derive
from Lucretius 2.619 ... raucisonoque minantur cornua cantu. Onomato-
poeia, the adaptation of sound to sense, is native to early Latin poetry;
Ennius and Lucretius achieve obvious and unmistakable effects with it.
In Virgil, although sibilant snakes (A. 2.209 ff.) and spuming seas (A.
6.174) invite the obvious treatment, sound-patterns are more subtle;
where they exist, one should be cautious in estimating the intended
'effect'-some Roman poets would be astonished at the discoveries of
their hypersensitive modern critics. In particular, one should bear in mind
that the same sound may appropriately be associated with contexts of
different feeling and content, and that sound-patterns are often not
strictly onomatopoeic-they may accompany the sense without reinforcing
it. To take the example of v alone: it reinforces the feeling of violence at
259 vana vomentem and perhaps at 644 f.; it supports the oracular tone of
500 fios veterum virtusque virum because alliteration in general recalls
early poetry; but it is purely ornamental at 155 f., 356 veterumque vides
monimenta virorum, and 576 si visurus eum vivo et venturus in unum.
The consensus of MSS supports strepuerunt; P alone, before correction,
has sonuerunt, a recollection of A. 7.637 classica iamque sonant or the like.
3 utque acris concussit equos utque impulit arma: a startlingly
novel expression (too novel for Ribbeck, who wished to remove it from the
text). The situation seems to require that Turnus drove forward the horses
of his chariot and shook his weapons.
In poetic language the usual phrase for a charioteer 'driving forward'
his yoked horses is impellere equos (Statius Th. 7.83 and cf. Th. 12.733 f.;
the same phrase with planta or calce for a rider 'spurring on' his single
horse at Silius 2.71 f., 7.696 f.); 'to clash' or 'shake one's weapons' is
concutere arma (Ovid Met. 1.14~ f., 7.130, 12.468; Seneca Troades 683).
Virgil nowhere uses either of these obvious and natural expressions
(although following the archaic preference for simple to compound verbs
he uses quatere (arma) at A. 10.762 and 12.442). Here, on the contrary, he
seems to have interchanged the position of the two verbs. The resultant
phrases, typical of many which make Virgil so difficult to translate, cannot
mean what they appear to say: Turnus does not now hurl any of his
weapons forward, as he does later (A. 9.52) actually to begin an attack,
nor does he 'clash his horses together' !
The phrases in the meaning they have here show Virgil's general
tendency to avoid the commonplace (metrical considerations may also
have played some small part), and can bear this meaning because they are
extensions of normal, literal expressions: concutere frena or Zora (A.
5.146 f., 6.100 f.) is inverted to concutere equos (frenis) (and cf. equos . ..
quatit, A. 12.337 f.). Through exploiting the fringe area of the meaning of
the words there is a gain in expressiveness and emotive violence hardly
conveyed by 'he whipped up his horses and brandished his weapons'.
6 COMMENTARY

Turnus is filled with the spirit of war; in A. 12.331 ff. he is explicitly

compared in a simile with Mavors, god of war. In the proem to Argonautica
6, which has enough similarities of phrasing and structure with the
opening paragraph of A. 8 to suggest that Valerius Flaccus may have had
the latter in his mind when writing, Gradivus (Mars) is thus described
(6 f.): impulit hinc currus, monstrum irrevocabile belli ( = hastam) / con-
cutiens. His actions and those of Turnus here are the same, but the
language is different: the lesser poet has used the expected phraseology.
Lucan's phrase (7.16) et quaecumque jugax Sertorius impulit arma
(= armatos concitavit) may rest on an individual interpretation of this
line of Virgil.
F. Bomer, Hermes 93 (1965), 130 f. compares the bold metathesis
verborum here-a figure quite outside the recognised categories of
rhetoric-with A. 6.847 f. excudent alii spirantia mollius aera / (credo
equidem), vivas ducent de marmore vultus, where the usual expressions
would be ducent aera and excudent de marmore vultus. This 'interchange'
is less surprising. (The label metathesis must not be taken to imply a purely
mechanical decision; it describes the result of an interplay of words hardly
to be analysed.)
This was also no doubt one of the aspects of the nova cacozelia which was
levelled against Virgil as a criticism. Donatus Vita 44: M. V ipsanius a
M aecenate eum suppositum appellabat novae cacozeliae repertorem, non
tumidae nee exilis, sed ex communibus verbis, atque idea latentis, 'M.
Vipsanius used to call him the inventor, foisted on them by Maecenas, of a
new form of perverse originality, neither bombastic nor jejune, but based
on everyday words and for that reason escaping notice'. (Editors unani-
mously print Vipsanius, for the otherwise unattested name Vipranius of
the MSS, identifying the critic with M. Vipsanius Agrippa, in which case
this is the only attributable piece of contemporary criticism of Virgil which
has come down to us. But it is strange (in spite of F. Marx, RhM 74 (1925),
185) that Agrippa should be referred to only by his gentile name, especially
as he himself had suppressed it quasi argumentum paternae humilitatis
(Seneca Contr. 2-4-13); it is even more surprising that Augustus' bluff
generalissimo should have expressed himself in technical terms of criticism
worthy of Quintilian, and with a penetration greater than his. If Agrippa
did actually make the remark he most probably got it from Quintus
Caecilius Epirota, the first man actually to lecture on Virgil. Caecilius was
the freedman of Atticus and tutor to Atticus' daughter, Agrippa's wife.
A more than academic interest in his pupil led to his expulsion from
Agrippa's house-but not perhaps without leaving a dictum well worth
remembering: see Suetonius De Grammaticis 16.)
The chief ways in which Virgil gives his everyday words an unfamiliar
shade of meaning are:
(1) by varying the normal prose idiom: examples abound (see the
Index s.v.), from simple changes affecting one word only (see the note on
56 foedera iunge), to an interchange in the position and construction of two
words as at, for example, A. r.660 ossibus implicet ignem (' entwine fire
 COMMENTARY 7

in her bones' for ossa igne implicet, 'entwine her bones with fire'),
A. 10.268 f. donec versas ad litora puppis / respiciunt totumque adlabi
classibus aequor (' the whole ship-dotted sea was gliding towards them',
for toto adlabi classes aequore, 'ships were gliding towards them over the
whole sea'). This tendency sometimes leads to extreme cases of hypallage
(see the note on 542);
(2) by giving a word a sense no longer current (see the note on 84 f. enim),
or current only as a technical term (of religion: see the notes on 85 mactat,
and cf. 106 tura dabant);
(3) by giving a word a sense it could etymologically have had but did
not (see the notes on 260 f. angit and 263 abiuratae). For Virgil's interest in
the etymology of proper names, see the note on 425.
Some instances of Greek-imitated constructions are 'new' (cf. the note
on 127), but the general tendency is of course not. Some of the ways in
which Virgil exploits the ablative-making it serve as an adjective, or
attach itself to more than one part of the surrounding context (see the
Index s.v. ablative)-may count as an individual innovation.
On the whole topic see also L. P. Wilkinson, CQ 53 (1959), 181-92.
3 equos: the horses of Turn us' chariot, for they and the arms belong
to him alone, in spite of both Servius and Henry, whose opposite view is
due to a misunderstanding of both situation and language. The violence
of the expression of 1. 3 (discussed above) eminently suits Turnus: (1) vio-
lentia is Virgil's key-word for his character and used exclusively of him
(although the adjective violentus is used also of the uncontrollable forces
of the winds and waves): in Drances' taunting speech (A. 1r.354) violentia
points to Turnus as clearly as if he had been named. (2) He is still the
victim of the Fury Allecto sent by Juno to incite him against her old
enemy Aeneas, and still in the same state: A. 7.461 f. saevit amor Jerri et
scelerata insania belli, / ira super.
Turnus has the limited tribal ideals of a Homeric hero: his preoccupation
with his own fulfilment and his own honour assigns him clearly to the
company of Achilles and Hector. The inadequacy of his standards and
attitudes in the face of the new order of things represented by Aeneas leads
to his downfall. His tragic dilemma evokes sympathy, but not admiration:
for Virgil and his Roman readers his character was almost a compendium
of social sins, audacia, superbia and furor, not to mention violentia and
insania.
3 utque impulit arma: Servius, puzzled to know the meaning of the
phrase here, and thinking (wrongly) that extemplo in the next line might be
read ex templo, comments: est autem sacrorum: nam is qui belli susceperat
curam, sacrarium Martis ingressus primo ancilia commovebat, post hastam
simulacri ipsius, dicens 'Mars vigila '. Henry dismisses this with charac-
teristic impatience and, for once, justifiably; we are not intended to
'conceive Turnus to have gone into the sacrarium and actually given a
push with his hand to certain ancilia there'. But the information Servius
gives, although quite irrelevant here, is correct and valuable in under-
standing early religious thought: before statues of gods came into
8 COMMENTARY

existence, the spear itself was the god of war and his spirit was imman-
ent in it (see H. J. Rose, ARR 21); hence shaking it and the ancilia
was an appropriate preliminary to beginning hostilities (Servius on A.
7.603).
4 extemplo, 'immediately', derives from the language of augury:
ex templo meaning 'immediately on leaving the templum', the area de-
marcated on the ground or in the sky by the augur, within which omens
were significant. But the original associations soon faded out and are not
to be read into the word here.
4 The phrase extemplo turbati animi at A. 11.451 introduces a similarly
terse description of excited mobilisation (in Latinus' city at Aeneas'
sudden advance).
4 simul here as often in Virgil (e.g. A. 8.80) is an adverb; the connecting
particle et is omitted and the asyndeton helps to give an impression of
speed: cf., for example, A. 2.755 horror ubique animo, simul ipsa silentia
terrent.
5 trepido: the adjective is very rare in prose before the Augustan
period (but Cicero has the noun trepidatio (Deiot. 20), in close connection,
as here, with tumultus). It means 'agitated', usually with fear (perhaps
etymologically connected with tremo and terreo), but not always; here, and
clearly at A. 9.233, it= 'excited'.
6 ductores (cf. 496) has a grand heroic ring about it compared with the
prosaic duces; it does not occur in the Eclogues, and only once in the
Georgics (4.88), in the mock-heroic battle of the 'king' bees: verum ubi
ductores acie revocaveris ambo. The word is rare before Virgil. ductores
Danaum as verse-beginning at both Lucretius r.86 and A. 2.14 suggests an
origin in Ennius, who probably used it not simply as a dignified equivalent
of dux (which for him seems to have meant 'pathfinder', cf. Ann.441 V),
but as a metrically usable equivalent of implrator (induperator-too
pseudo-archaic for Virgil-was another way out of the difficulty). See
Leumann 147, n. I.
6 primi: out of the notion of 'first', 'at the front' of a series in space
(cf. 281 sacerdotes primusque Potitius, 'the priests and Potitius at their
head') developed that of 'first' in precedence or rank, as here and at 105
iuvenum primi; the transition is clear at 586 f. Aeneas inter primos et fidus
Achates, / inde alii Troiae proceres. For other meanings of primus see the
Index s.v.
6 Servi us ad loc.: M essapus et Ufens: bona electio personarum ad
dilectum habendum: unus eques bonus, id est Messapus, mittitur, alter pedes
egregius-a reasonable deduction from A. 7.691 at M essapus, equum
domitor, Neptunia proles (Neptunus = Ilocmowv 't1T1nos-, cf. G. 1.12 ff.) and
A. 7.745 ff. where Ufens is said to belong to the gens ... ditris Aequicula
glaebis. / armati terram exercent. . . Messapus is recast by Virgil as an
Etruscan prince; elsewhere he was the eponymous ruler of Messapia (the
whole of Southern Italy, including both Apulia and Calabria); Ennius, who
was born in ancient Calabria (modem Puglia), claimed descent from
Messapus (see Servius on A. 7.691 and Silius 12.393).
 COMMENTARY 9

7 contemptor ... deum (gen. pl.) is almost a 'tag' for Mezentius

(cf. A. 7.648), who is marked out by it as the destined antagonist (impius)
for pius Aeneas.
Macrobius (3.5.10) explains the title by citing Cato's first book of
Origines: ait enim M ezentium Rutulis imperasse, ut sibi offerrent quas dis
primitias offerebant, et Latinos omnes similis imperii metit ita vovisse:
'Iuppiter, si tibi magis cordi est nos ea tibi dare potius quam Mezentio, uti
nos victores facias '. Later versions substitute for primitiae, the first-fruits
of all crops, the whole of the wine produce only (see Ovid F asti 4.879 ff.
and Frazer's note there). The demand for this offering, which later genera-
tions would regard simply as arrogant blasphemy, may originate in
Mezentius' status as a sacred chieftain or king-god: see the note on 485 ff.
The political implication of the account-the power of the Etruscans of
Caere and their king Mezentius to attempt coercion over a subject
Latium-is undoubtedly historical, although Cato's statement implies
neither that the Latins did in fact pay tithes to Mezentius (as Alfoldi
209 ff.) nor that they did not (as A. Momigliano, ]RS 57 (1967), 213). (For
the spelling Mezzentius, which occurs in the Romanus here and elsewhere
in the M ediceus, and which is due to the influence of the grammarian
Verrius Flaccus, for whom z represented one consonantal sound only, see
Roscher, 2.2.2952 f.)
7 f. undique cogunt / auxilia et latos vastant cultoribus agros:
a composite idea analysed into simpler components. Simplicity of syntax
is one of the hall-marks of Virgil's style. He prefers parataxis, a succession
of coordinate clauses each with an indicative verb, to hypotaxis, the
subordination of some clauses to others. So where in Ciceronian prose we
should normally find a complex idea expressed in such a way that the
grammatical structure reflects the logical or chronological connection
between its component parts, Virgil takes apart the various aspects of a
complex notion and links them together with coordinate verbs, each
reflecting a different side of the same situation. One clause therefore is
often a restatement in different terms of the one which preceded: the
dicolon (or tricolon) abundans of the Latin grammarians. The 'padded'
style of Ciceronian oratory with an accumulation of (near-) synonymous
clauses is exactly similar. This feature of Virgil's style has conveniently
been labelled' theme and variation' by Henry (see his note on A. r.546 ff.,
and on A. 8.505 f. where he hails himself as the discoverer of the secret of
Virgil's style). Examples of the technique are very numerous; cf., for
example, 241 f., 374 f., 505 f. See further the note on 171.
8 latos . .. agros: single-syllable rhyme is fairly common before the
strong caesura in the third foot and the end of the line; usually the rhyming
words are noun and adjective (cf. also, for example, rr8, 682), sometimes
two adverbs (149) or participles (649). Earlier poetry shows the same
feature: see W. W. Ewbank, The Poems of Cicero 68 and Bailey, Lucretius
vol. l, 154. Although Roman poets were probably sensitive to the effect
produced, it is unlikely that the rhyme was primarily devised for its own
sake: it was the natural result of word-symmetry and the structure of the
IO COMMENTARY

verse. The same probably holds good for the more frequent internal rhyme
of the pentameter in elegy (but see Platnauer 49).
8 vasto, 'lay waste', takes an accusative of direct object, and its sense
can be further completed by an ablative of instrument (as at 374), or an
ablative of separation (as here) showing of what the object is laid waste;
this second construction is the usual complement of verbs of depriving,
like viduo at 57I. Statius imitates the line at Th. 3.576 f. agrosque viris
annosaque vastant / oppida, but the phrase and construction are not
exclusive to poetry, cf. Aulus Hirtius E.G. 8.24.4.fines eius vastare civibus,
aedificiis, pecore.
9 mittitur et: et = etiam here, at 200 and 630, and probably at 161;
notice how often in the surrounding context (2, 8, 12, 14) a normal con-
nective et occurs in the same position in the verse.
9 Diomedes, the Greek hero, Aeneas' old opponent in the Trojan War,
could expect to be attacked in southern Italy if ever Aeneas gained
possession of Latium, and so was an obvious potential ally. In addition he
had a close personal link with Turnus, for he had been king of Argos, and
Virgil attributes a Greek origin to Turnus (A. 7.371 f. et Turno, si prima
domus repetatur origo, / I nachus Acrisiusque (kings of Argos) patres mediaeque
Mycenae). According to the post-Homeric legend preserved by Servius,
Venus had punished him for the injury inflicted on her (Il. 5.334 ff.) by
causing his wife to be unfaithful during his absence, and he could not
return to Argos because her lover was in possession there. Ultimately he
sailed to the Gargano promontory in southern Italy, founded the city here
referred to, Argyrip(p)a or Arpi (A. n.246-50), near the modern Foggia,
and was credited with many other foundations in Apulia, Samnium and
elsewhere (see Saunders 34 ff.).
The embassy was in fact unsuccessful (cf. A. n.225-95), Diomedes being
unwilling to fight the unconquerable Aeneas again. At the great debate in
the Council of Gods Venus represents him as actively hostile to the Trojans
(A. rn.28 f. atque iterum in Teucros Aetolis surgit ab Arpis / Tydides), but
this is a rhetorical lie (see the note on 41 concessere).
Venulus, as emerges from A. n.742 and 757, was a Tiburtine and an
appropriate ambassador to Diomedes, as Tibur too was credited with
Argive foundation. He appears at A. lI.242 to report the unsuccessful
results of his mission, and at A. n.741 ff. is swooped upon by Tarcho and
disappears from the narrative.
IO ff. The messenger's speech is designed to play on Diomedes' feelings:
the Trojans are already established in Latium (Servius ad lac.: consistere,
iam esse conditos, id est fundasse civitatem); Aeneas is importing the house-
hold gods of Troy which Diomedes had helped to conquer (victosque penates
is no doubt a taunt emanating from Turnus, Juno's instrument, and
recalls her words to Aeolus at A. 1.68); he is claiming to be the fated king
of Latium (according to the oracle of Faunus, A. 7.254 ff.); many tribes
are joining him (this assertion is made plausible by describing him as vi,
Dardanius, implying that as Dardanus had come from Italy (cf. A. 7.240)
Aeneas could claim kinship with some of the Italian tribes); finally,
 COMMENTARY II

Diomedes is in the best position to guess what Aeneas is plotting (because

he knew him of old and might surmise that if Aeneas won Latium he would
proceed to take vengeance on Diomedes, not least for violating his father
Anchises' tomb (a legend known to Varro, cited by Servius on A. 4.427)).
The whole speech is a disguised suasoria in which exaggeration and mis-
representation are regular and intentional features: 13 f. are not literally
true, but this is no ground to suppose that Virgil is temporarily nodding
(as Wagner, QV 40, Virgilius dormitans), or that the lines are interpolated
(as Peerlkamp).
IO f. Latio consistere Teucros . .. victosque penatis / inferre
represent as already accomplished the objects of Aeneas' mission stated
(in the usual terms of colonisation) at A. r.5 f. dum conderet urbem /
inferretque deos Latio.
IO victos: to represent the enemy one is about to meet as already
conquered is a commonplace of eve-of-battle harangues, cf. Caesar B.C.
2.5.2; and Livy 21.40.5 ff. and 21.41.IO ff.
IO consistere: the right word for the 'settling' of aliens (as the
Trojans were), cf. Caesar E.G. 7.3.r cives Romanos qui negotiandi causa ibi
constiterant. considere can also be used of permanent settlement, as at
Val. FL 5.234; it is a variant here, most probably deriving from A. 6.67.
I I classi: ablative; a small group of masculine and feminine nouns with
-i stems, although usually having an ablative singular in the regular -e,
sometimes keep the older alternative in -i in the classical language (amni
473, igni e.g. A. 6.742).
13 s(e) adiungere: for the rather harsh elision of the monosyllable see
the note on 503 tum.
15 struat: a word with rather sinister undertones for Virgil, cf.
A. 2.60 (Sinon) hoc ipsum ut strueret Troiamque aperiret Achivis, and
Austin's note. The sense of 'plotting' is common in prose, e.g. Cicero
De Orat. 2.208 odium in alios; Tacitus Ann. 4.IO mortem.
15 his coeptis: 'sequential rhyme' already in Homer, e.g. Od. r.83
ovSt: S6µ,ovSt:. Ancient critics including Servius were much concerned with
the propriety of rhyming endings in two consecutive words (homoeo-
teleuton): it is of course pronounced where both bear the ictus (as here,
51 his oris, IO0 res inopes, 261 elisos oculos, 389 solitam flammam, 715 dis
Italis). This feature is far less common in Virgil than in Ennius and
Lucretius, but there is no evidence that he avoided it per se and used it
only for descriptive effect (pace Norden 405 ff.). On A. 8.545 Servius
auctus says Euandrus pro Euander, sed vitavit oµ,oio-rt>.t:v-rov sequentis verbi
(sc. pariter): but as Euandrus is the form always used (five times) in A. 8,
whereas Euander occurs only once at A. I0.515, his argument has little
weight. Norden (407) maintains that the hypallage at A. 8.526 Tyrrhenus-
que tubae . .. clangor is due to deliberate avoidance of sequential rhyme; it is
more likely to be primarily caused by word-symmetry, and as this is far
more frequent in Virgil than in Ennius and Lucretius, the absence of
homoeoteleuton which is its natural consequence is also far greater in him
than them.
3 EAC
12 COMMENTARY

Occasionally the rhyme is emphasised by -que . .. -que (94 noctemque

diemque; cf. 431, 490), or linked by a single -que or -ve (os oculosque 152;
cf. 38, 402, 663, 670, 690, 698); often one of the rhyming syllables is
suppressed by elision (248 cavo saxo atque; cf. 361, 580, 646).
16 ipsi: most probably Diomedes rather than Aeneas. To end a plea
for help by telling someone that he himself has a shrewder notion of a
third party's intentions than others have is flatteringly persuasive; to
suggest that the very people requesting help are uncertain about the actual
need for it because they cannot guess a third party's intentions is very
uncompelling.
17 regi . .. regi: the formal repetition of diplomatic language, such as
might be used in the set speech of an ambassador laying stress on his
credentials: cf. A. 10.149 regem adit et regi memorat . .. and A. 11.294 f. et
responsa simul quae sint, rex optime, regis / audisti.

18-35 The impending war fills Aeneas with a turmoil of unrest: his
thoughts dart in all directions like rays of light reflected from troubled water,
and he does not rest until long after night has fallen and all other creatures are
asleep. Then the god of the Tiber appears to him in his sleep with a message
of comfort.
18 talia per Latium: 'gerebantur' subaudis: et est formosa ellipsis,
Servius. The verb is easily 'supplied'; if actually expressed, it would
attract too much attention to an unimportant idea, and cause the style to
become diffuse, against Virgil's whole tendency. Such ellipses are not
common in earlier poetry; in Virgil cf., for example, G. 4.528; A. 9.636;
A. 12. 154 and 195.
18 Laomedontius heros: Laomedon succeeded his father Ilos as king of
Troy (see the stemma 134 ff.), and was the father of Priam and Hesione
(cf. 157 f.). He fortified the citadel (Pergama 37) with the help of Apollo
and Poseidon (Il. 7.425 f.), but refused to pay them the promised reward.
To punish him, Apollo sent a plague and Poseidon a sea-monster. Lao-
medon's crime was advertised by the Augustan poets as the sin of the
Trojan fathers continually visited on their Roman descendants: cf.
G. 1.501 f. satis iam pridem sanguine nostro / Laomedonteae luimus periuria
Troiae and Horace Odes 3.3.21 ff.
18 Laomedontius heros: in marked contrast to Homer's repetition of
stylised formulae with proper names, like 1r6oas- wKv, 'AxiUEv, etc., Virgil
varies his periphrases as much as possible: in A. 1 Jupiter is referred to as
divum pater atque hominum rex (65), o qui res hominumquedeumque / aeternis
regis imperiis (229 f.), rex magne (241), hominum sator atque deorum (254);
in A. 6 Aeneas is referred to as sate sanguine divum, / Tros A nchisiade
(125 f.), Anchisa generate, deum certissima proles (322), dux Anchisiade
(348), Troius heros (451). This is one symptom of the great difference
between poetry which is orally transmitted (lays, ballads, etc.), where the
reciter's memory needs support from repeated formulae, and written
poetry, with time and scope for artistic variation. For an excellent dis-
 COMMENTARY 13

cussion of this and other differences between primary and secondary epic,
see C. S. Lewis, Preface to Paradise Lost, esp. ch. 4.
19 magno . .. fluctuat aestu: this metaphor for emotional disturbance,
'billowing on a mighty surge', is common in Greek and Latin poetry: see
Pease's collection of instances on A. 4.532, and cf. esp. Luer. 6.34 volvere
curarum tristis in pectore fluctus, and the very similar Catullus 64.62
prospicit et magnis curarum fluctuat undis.
20 f. These lines are exactly the same as A. 4.285 f. Here they are un-
doubtedly genuine and appropriate, as they are almost certainly also in
A. 4 (so Pease and Austin ad loc., against Sparrow 143). Nor is it surprising
that Virgil should repeat lines modelled on the repeated Homeric formula
SuivSixa fL£PfL~p,g£v (fl. r.189, 8.167, 13.455); he was following epic
precedent, although he does not and does not need to repeat set-phrases as
often as Homer, and in the case of formulae with proper names clearly
preferred variation (see the note on 18 Laomedontius heros).
20 celerem: the adjective here probably represents an adverbial
notion (see the note on seram 30) and should be so translated. If it is a true
epithet it is purely ornamental: it is not specially appropriate to Aeneas
here, and the speed of the human mind in general was a commonplace,
cf., for example, Soph. Ant. 353 ff. Kat ¢,01yµa Kat av£µo£v / ef,poVT)µa . .. iS,Sa-
taTO (d~p) and Cic. Tusc. 1.43 nihil est animo velocius.
21 in partisque: not of course a true case of hyperbaton, for -que
joined to the noun and not the preposition is common practice in prose;
see KS r.583 f.
22 ff. This simile is inspired by Apollonius of Rhodes, Argonautica
3.755 ff.: 'the heart within her (Medea's) breast gave violent starts, just
as a sunbeam, rising out of water which has just been poured (To S~ vlov . ..
KlxvTai) in a bowl or pail, darts about the house (Soµo,,): hither and thither
it dashes shaken by the swift eddy'.
Servius (on 19) is both wrong and superficial in saying comparatio(-nem)
quae est Apollonii verbum ad verbum. There are three important differences:
(1) To S~ vlov ... Klxvmi, which explains why the light reflected by the
water is not still, is compressed to tremulum; (2) Soµoi, is expanded into
the luxurious dignity of summi . .. laquearia tecti; (3) as Posch! notes
(239 f. = 146 f.), moonlight is added to the simile to harmonise with the
peaceful night-setting of the wider context.
Virgil's adaptation is commended by Warde Fowler (ASR 35 f.),
Posch! (Zoe. cit. above), condemned by Conington, Tyrrell (Latin Poetry
(London, 1895), 141) and Heinze (VeT 250). More interesting than indi-
vidual feelings about its appropriateness is the imaginative process by
which Virgil came to use it. Mental agitation in Homer is regularly
illustrated by a simile; at the beginning of Od. 20 a principal character,
Odysseus himself, turns his troubles over in his mind before sleeping: this
may have been Virgil's point of departure for the present passage (see
M. Coffey, 'The Subject Matter of Virgil's Similes', BICS 8 (1961), 71). But
Homer chose to illustrate the situation by similes of great realism:
Odysseus' angry heart growls like a protective bitch confronted with a
3-2
14 COMMENTARY

stranger (14 ff.), and his body turns from side to side like a black pudding
twisted round by a man eager to have it roasted (25 ff.). This type of
simile was too unheroic for the canons of epic poetry as Virgil conceived
them. The problem of what to put in its place was solved by the aftermath
of associations provoked by the image of troubled water in 19, and this
led Virgil to recall a brilliant example of one of Apollonius' chief charac-
teristics, his treatment of the play of colour and light. (It is noteworthy
that four out of the five similes in this book (here, 391 f., 589 ff., 622 f.)
depend wholly or chiefly on light images and contribute to the unsombre
atmosphere of the book.)
But the simile may have suggested itself also because of the similar
setting: just before it (Arg. 3.751) Medea could find no share in the universal
sleep, like Aeneas here.
22 aquae . .. lumen: = 'the light from the water' is an unusual phrase
and like radiantis (23) may have been inspired by the diction of Lucretius
4.2n ff ... . simul ac primum sub diu splendor aquai / ponitur, extemplo
caelo stellante serena / sidera respondent in aqua radiantia mundi.
22 labris here = 'basins' (Apollonius' Mf3rrri ¥ TTov Ev yav.\cp), labrum
being a shortened form of lavabrum (at, for example, Luer. 6.799), cf.
l(av)atrina. The quite different labra = 'lips' (connected with Zambo) can
however also refer to the lip or brim of a vessel.
22 obi: postponed conjunctions usually occupy the second place in
their clause (ubi 589; ut 88 and 89,412; donec 326; si 140, 147, 396; cum
622; quin etiam 485; namque 497; neu 582). More rarely they are found
in the third or even later place (ubi here; ut 58, 191; si 560; cum
276, 391; quia 650; dum (in fifth place) 454). In earlier Roman prose and
poetry (including Lucretius), conjunctions are postponed frequently and
freely (see KS 2.614 f.), but an analysis of the above cases suggests that
Virgil's practice is by no means haphazard:
(1) the removal of a conjunction from the first place in a clause (or
verse) enables the emphasis natural to that position to fall on the word(s)
coming before the conjunction, and usually they can bear such emphasis
appropriately (see further Marouzeau, L'Ordre des Mots 3 (Paris, 1949),
122 ff.);
(2) conjunctions delayed beyond second place draw attention to them-
selves, and are almost invariably found in the middle of a symmetrically
arranged group of nouns and adjectives: the artistic arrangement of the
words and the delayed conjunction point to neoteric (Catullan) practice;
(3) postponed conjunctions are particularly noticeable (with or without
word-symmetry) at the beginning of similes: ubi here could hardly follow
directly after sicut without being rather clumsy, and the same applies to
391 non secus atque . .. cum; qualis ubi (589), and qualis cum (622) are
quite inoffensive. A good example with word-symmetry is A. 2.379
improvisum aspris veluti qui sentibus anguem . ..
23 ff. What is actually happening is unclear at first sight, partly because
of poetic language, partly because of Virgil's tendency to say multum in
parvo et saepe in silentio.
 COMMENTARY

The reflected light (22) is tremulum, darting and mobile like Aeneas'
thoughts, because the surface of the water is also tremulous, for it has
just been poured into the bowl (Virgil omits what Apollonius expressly
states).
The sunlight is reflected. lumen ... sole repercussum, 'light reflected by
(the action of) the sun' is a characteristically Virgilian variation of the
normal prosaic expression sole repercusso, 'when the sun is reflected'
(proposed as an emendation by Hoffmann, but unnecessary): an example
of one aspect of nova cacozelia (see the note on 3).
imagine lunae is simply a periphrasis for tuna (conclusive parallels are
cited by Henry); cf. the note on Caci facies 194. It does not concentrate
attention on the shape of the moon as seen in the water (as supposed by
Long, cited by Conington): no distinct image can be seen on the surface
of choppy water.
iamque (24) marks a transition to a different situation (cf. Wagner, QV
24.9). The difference is conveyed by erigitur, 'rises straight up' (cf. rigidus,
and erigit at A. 3.422 f., 3.575 ff.). The diffused flickering light is now
concentrated into a single beam, as would naturally happen after the
water-surface had become still and smooth, as it was bound to do. Virgil
follows out the simile to its pictorial conclusion, even though this last idea
has no parallel in Aeneas' situation: his mind had not achieved a beam-like
singleness and directness when he went to sleep, for his problems were still
unsolved.
sub auras = 'skywards'; in view of laquearia (25) and Apollonius'
ooµoLs- it cannot be pressed to mean 'up to the open air'. As at 104 ante
urbem, Virgil may have put in a new context where it does not strictly fit a
phrase remembered from a previous one where it did, A. 3.422 f. sub
auras / erigit.
24 Note the bucolic diaeresis (the fourth foot, almost always a dactyl,
ends with the end of a word) and the pause at it. It is called 'bucolic'
because of its frequent use in Greek pastoral poetry, and (therefore) by the
Virgil of the Eclogues: compare A. 8.352 (quis deus incertum est) habitat
deus; Arcades ipsum with Eel. 7.4 ambo florentes aetatibus, Arcades ambo
(and note that the diaeresis and pause here are closely associated with the
Greek loan-word Arcades; similarly in A. 8 with scyphus 278, chorus 718).
But it was by no means the exclusive property of pastoral poetry: it occurs
in Homer at the rate of about one line in ten. It does not occur nearly so
frequently in the Aeneid, but Virgil clearly recognised its usefulness to
secure variety of pause. Usually a pyrrhic word (of two short syllables)
precedes the pause, as here and 198, 249, 278, 352, 388, 598, 718, a longer
word at 205, 238, 666. Originally this pause seems to have carried with it
overtones of pathos (see Norden 122 f.); in the Aeneid no clearly defined
descriptive use can be discovered, although in recitation appropriate
feeling could no doubt be voiced in lines as different as 249 desuper Alcides
telis premit, omniaque arma and 388 cunctantem amplexu molli fovet. ille
repente. Virgil does not often follow Lucretius in having a molossus before
a pause at the end of the fourth foot (A. 8.613, 675, 729), and only excep-
16 COMMENTARY

tionally allows a spondaic word to stand there (A. 8.518; contrast G. 2.43,
A. 6.43). The diaeresis at A. 8.86 and 5n is secondary and unimportant.
late loca occurs at the same position in the verse at G. 4.515, A. 2.495
and 9.190, differently arranged at A. 2.698, 6.265: it looks like a formula
of older poetry, probably Ennius.
In an interesting article in REL 43 (1965), 261 ff., R. Lucot examines
the 6I heavy bucolic punctuations in the Aeneid and establishes that a high
proportion of bucolic clausulae of the type si bona norint occur in speech-
passages (14 out of 22), only a low proportion in narrative and descriptive
passages (9 out of 39). The three-word clausula produces greater conflict of
ictus and accent than the two-word, and hence also a greater liveliness in
recitation, which Virgil appropriately associates with direct speech. The
same cause and effect operate when the bucolic diaeresis is preceded by a
pyrrhic, and this too occurs more often in speech-passages than elsewhere.
25 laquearia here, as often elsewhere, = 'the panels of the ceiling';
some grammarians, ancient and modern (see TLL and Nettleship, Contr.
Lat. Lex. respectively, s.v. laquear), believe that the correct spelling of the
word in this sense is lacuaria, making clear the connection with the basic
sense of lacus (see the note on 74); but the consensus of MS authority
forbids introducing any such 'orthography' here. See further Austin on
A. 1.726.
26-35 In composing this passage Virgil seems to have been reminded of
two other sections of the Aeneid with partly similar situations:
(1) night-time was the appropriate time for visions of the gods (cf.
A. 2.268 ff. and 5.835 ff.). In A. 3.147 ff. the effigies sacraedivumPhrygiique
penates appear to Aeneas, sent by Apollo to explain his ambiguous oracle;
their words reassure him, like those of Tiberinus here (tum sic adfari et
curas his demere dictis A. 3.153 = 8.35); they specified the goal of his
wanderings, Tiberinus confirms that he has reached it;
(2) the situation which opens the episode in A. 3 (147 nox erat et terris
animalia somnus habebat, where animalia as here = living creatures of all
kinds, including mankind) is here expanded to two lines, while in A.
4.522 ff. it is amplified to six lines. Here Aeneas' mind is in a turmoil of
anxiety (19 magno curarum fluctuat aestu); there (A. 4.504 ff.) Dido
prepares her magic charms: nox erat (522) and everything is asleep, at non
infelix animi Phoenissa, neque umquam / solvitur in somnos ... (529 ff.):
Dido never gets to sleep, because she is in a turmoil of anger (532 magnoque
irarum fluctuat aestu). And there is a further variation at A. 9. 224 f.
(everything is asleep ... except the Trojan leaders holding an extraordinary
council of war). This type of EK<ppaa,, xpovov, with its expected contrast,
had been a locus communis of Greek poetry (see Pease on A. 4.522): the
first example involving a comparison with nature is Euripides I.A. 9 ff.;
it is tempting to believe that the famous fragment of Aleman describing a
night scene, Poetae Melici Graeci (D. L. Page, Oxford, 1962) Aleman
frg. 89, must have ended in accordance with the same convention (but sec
R. Pfeiffer, Hermes 87 (1959), l ff., last footnote).
26ff. nox erat . .. cum . .. Aeneas . .. procubuit: a construction known
 COMMENTARY 17

as cum inversum; the position of the cum is 'changed around' so that it

introduces not, as usual, the subsidiary idea but the main statement; in
such cases the verb of the main clause, now introduced by cum, remains in
the indicative. In archaic Latin cum (quom) can be used with an indicative
to express any of its wide range of meanings, and serves as an all-purpose
subordinating conjunction (in some cases the sense of subordination is very
weak). With this history of development behind it, the indicative appro-
priate to a main statement of fact naturally held its own in the cum
inversum construction. See R. D. Williams on A. 5.84 f., and cf. in A.
8 97 ff. conscenderat . .. cum . .. vident . .. , and 276 f. dixerat . .. cum . ..
velavit. Virgil's fondness for the cum inversum construction is to be ex-
plained partly because it rests basically on simple parataxis (cum = little
more than et tum), and so coheres with the general tendency of his style,
but more because it was one of the ways (postquam, ubi, etc., with a
perfect indicative were others) in which he could avoid narrative cum with
the pluperfect subjunctive. This latter construction, so frequent in narra-
tive prose, is extremely rare in narrative poetry. It does not occur in any
poem certainly written by Virgil (see Wetmore's Index s.v.), and there seem
to be only about ten instances in the whole of Ovid's Metamorphoses,
Lucan, Silius Italicus and Valerius Flaccus (see Axelson 87 f.). This
suggests that it was under a stylistic ban from epic diction, quite apart
from the metrical inconvenience of the cumbrous verb-form, perhaps
because of the cacophony of the -ss- sounds, perhaps because of the remote-
ness and lack of vividness of the tense. (The pluperfect subj. does of course
occur in other constructions, notably at A. 4.603-6, where the repeated
-ss- may echo the violent hatred of the sense: see Austin on A. 2.585.)
27 alituum: the regular gen. pl. alitum being a cretic in form could
only be used in hexameters by a harsh elision of the last syllable; Virgil
uses this practice only 'with clear limitations' (analysed by Austin on
A. 4.684). The 'heteroclite' form alituum first appears in Lucretius
(frequently: 2.928; 5.801, ro39, ro78; 6.1216), here only in Virgil, sporadic-
ally in later poets as a useful hexameter form (e.g. Manilius 5.369; Statius
Silv. r.2.184). Ennius may well have originated it, perhaps on the analogy
of uncontracted Homeric gen. pls. like ldwv, Movaawv. Leumann 150, n. 3,
noting that alititum genus is a regular phrase, suggests that alituum was
originally an adjective formed on the analogy of perpetuus, and sub-
sequently interpreted as a gen. pl.
The reverse process of contracting a gen. pl. (of the 4th decl.) is a bolder
licence, extended by Virgil only once to a word for which the licence is
not already established, at A. 6.653, currum for citrruum (on the model of
deum, equum, etc.).
alites originally had sacral connotations in augury: 'prophetic birds'.
28 f. pater . .. Aeneas: note the wide separation of the two words
closely connected in sense: a minor instance of the interlacing word-order
which helps the flow of the verse (see the Appendix, p. 199).
28 sub aetheris axe also ends A. 2.512; as usual with such small
repeated phrases there is no cross-reference between their contexts. Virgil
18 COMMENTARY

inherited a store of adaptable hexameter phrases from earlier poetry, but

many he invented for himself. From the total stock the same phrase is
often used, quite naturally, to describe something which is the same-at a
literal superficial level, just as in the Homeric poems. Occasionally other
associations in the contexts of the repeated phrase clearly relate to each
other; particularly in the case of images and symbols the facet they present
in any one context is sometimes complemented by those evident in
another (the different aspects of Golden Age symbolism are an obvious
example), and remembered associations enrich the new appearance of the
image for a reader who has digested their material with the same reflective
care as the poet himself. This is of course entirely different from supposing
that Virgil used his repeated phrases as Wagner used his leitmotivs-
deliberately and consciously to echo the emotional context of their first or
most significant appearance; for an imaginative fantasy based on the
improbable assumption that he did, see M. J.C. Putnam, The Poetry of the
Aeneid, Four Studies in Imaginative Unity and Design (Harvard and
Oxford, 1965).
29 turbatus pectora. The middle (or semi-reflexive) voice, which is
common in Greek of all periods, was a mere vestige in Latin by the time
of the earliest literary records, and confined to the perfect participles of
verbs of dressing and undressing, induor, (ac)cingor and the like. For
example, Plautus M enaechmi 5n f. non ego te indutum foras / exire vidi
pallam?, 'did I not see you going out of doors dressed in (having put on
yourself) a cloak?', where indutum is a middle participle (though passive
in form) and pallam is its accusative direct object. This may be regarded as
a native Latin construction on the point of extinction: Greek influence
soon gave it a new lease of life:
(1) Ennius, whose syntax is often grecising, seems to have launched the
extension whereby not the item of dress but the part of the body was made
the direct object of the middle participle, with the dress or equipment in
the instrumental ablative: Ann. 400 V succincti corda machaeris;
(2) Augustan poetry in particular extended the usage from participles
to finite verbs, e.g. A. 7.640 loricam induitur (see the note on 457);
(3) the type succincti corda, where the participle is middle and the
accusative one of direct object, could hardly avoid being reinterpreted as a
truly passive participle with a 'Greek' accusative of respect (see the note
on 425 nudus membra): hence expressions of the type percussi membra
timore, Luer. 5.1223, 'shattered by fear with respect to their limbs'.
When a perfect participle is used in this way with an accusative, only the
context and meaning can decide whether it is more appropriate to analyse
it as (1) a middle participle with an accusative direct object, or (2) a
passive participle with an accusative of respect; sometimes, where the
fusion is complete, it is impossible to say which of these parent construc-
tions contributed the more. In the present passage, where Aeneas is
clearly the passive object of external forces, (2) is more appropriate, 'his
heart confused by grim war'; but 286 evincti tempora, and 662 protecti
corpora allow either interpretation, although the fact that both phrases are
 COMMENTARY

connected with the basic idea of 'dressing' gives a slight bias in favour of
explanation (1). For Virgil's fondness for this type of construction see
Norden on A. 6.281, and for its development see the references in R. D.
Williams' note on A. 5.135.
29 pectora is poetic plural, like the Homeric aT7J81:a, used with reference
to one individual only. Many parts of the body were naturally and properly
referred to as pairs (manus, oculi, tempora 'temples' (684)), or in the plural
(digiti, capilli), and by analogy, singular parts came to be pluralised,
pectora perhaps through the influence of papillae, mamillae. Neuter plurals
in -a of concrete and abstract nouns form by far the largest class of poetic
plurals; usually their form is very suited to the needs of hexameter verse,
and they became a recognised part of epic diction (but not of course
confined to epic). Some contexts allow one to suppose that by this use of
the plural poetry is aiming to be more emphatic or expressive; others make
it quite clear that if this is occasionally true, the usage rapidly degenerated
into a cliche. See further Li:ifstedt, Syntactica II 2 , 27 ff .. Marouzeau, Traite
de Stylistique Latine 2 (1946), 222 f., Norden 408 f.
(For the history of pectora see 534 ff. of the fundamental discussion of
poetic plurals by Paul Maas, ALL 12 (1902), 479 ff.)
Occasionally the metrical convenience of the plural form of an accom-
panying adjective seems to be more decisive than that of the noun itself,
e.g. at A. 6.809 f. nosco crinis incanaque menta / regis Romani.
(The variant pectore, offered by the Medicean MS, though grammatically
possible as an ablative of respect, is intolerable in juxtaposition with bello
ablative, through assimilation to which it probably owes its existence.)
30 seramque dedit per membra quietem: it is common in Latin
generally to find an adverbial notion, one which is logically most inti-
mately connected with the verb of a sentence, expressed by an adjective
attached to a noun; this is particularly the case when the notion is one of
direction (like transversus), or of order in space or time (like extremus,
proximus). In this construction the adjective is not attributive but pre-
dicative: the notion it expresses really belongs to the verb, but the inflected
form of the adjective necessitates agreement with some noun. As Housman
points out on Manilius 5.226, the adjective so used for an adverb may be
attached to either subject or object (this follows from the fact that it does
not directly qualify either): vesperi forum pererro may have as an equi-
valent either vespertinus forum pererro or vespertinum(que) pererro J
(saepe) forttm (Hor. Sat. r.6.n3 £.).
The existence of this idiom may partly explain why in Latin adverbs
either simply do not exist to express some adverbial notions, or are only
very rarely found. Even when adverbs were available for use, Virgil seems
to have avoided all but the most common and colourless, probably because
of their prosaic overtones; often of course they were simply intractable in
hexameter verse (cf. celerem 20: celeriter is impossible in hexameters).
Virgil developed the idiom even further, especially with adjectives of
time in imitation of Greek usage (with crastina 170, matutini 456, matu-
tinus 465, cf., for example, Od. 4.655 f. ioov lv8ao€ Mlvropa ofov / xB,tov
20 COMMENTARY

V1MJoi:ov); and his authority launched it into post-classical prose, as did that
of Livy, e.g. praef. II serae avaritia luxuriaque immigraverint; 3.56.7
seras ... venire poenas.
See further LHSz 2, 171 f., KS 1, 234 ff., R. D. Williams on A. 5.868.
30 dedit = 'let it spread'; cf. the similar sense of verbs like demitto =
'to let fall', etc.; see the note on 707 f.
31-96 The situation of Aeneas, uncertain of his exact whereabouts, with
neither formed plans nor outside help to combat growing opposition recalls
the situation of Odysseus on wakening in Ithaca (Od. 13.187 ff.). Virgil's
structural scheme for this part of the book was no doubt shaped to some
extent by Odysseus' situation there and, to a greater extent, by his
reaching of the river-mouth on Scheria (Od. 5.441 ff.) (see the note on
86-9). Minutiae of detail should not be pressed (Knauer 239 ff. goes much
too far in the search for similarities), but there is a correspondence in some
salient features: Athene (at first in disguise) gives Odysseus confirmation
that he has reached his goal, encourages him, promises him help, and tells
him where he can find allies. Tiberinus does the same for Aeneas. (The
germ-idea of giving such a role to Tiberinus probably derives from the
friendly river-god of Homer's Scheria (Od. 5.451 ff.); the importance of
the role he plays reflects Virgil's patriotic pride in the river of the capital
city; the fact that he speaks directly and in person may point to Hellenistic
precedent: the Nile is represented as speaking in Callimachus' poem on the
victory of Sosibius (frg. 384.27 f. Pfeiffer); but the first precedent is again
in Homer, the Scamander dvlpi £laa.µ£vos speaking in Il. 21 (212 ff.,
229 ff., 308 ff., 357 ff.).) After the divine appearance Odysseus prays to the
local nymphs (Od. 13.356 ff.), as does Aeneas (71), and goes to meet a
hospitable reception from the predicted ally Eumaeus the swineherd
(Od. 14.1 ff.), just as Aeneas makes his way to Evander. But at this point
Virgil seems to have been reminded of different Odyssean material (see
the note on 97 ff.).
31 deus . .. Tiberinus: the most striking testimony to the superstitious
respect accorded to the Tiber is an incident at the beginning of the reign
of Tiberius; when it was proposed to divert some of the tributaries of the
Tiber to prevent flooding, the envoys from Reate urged (Tac. Ann. 1.79)
spectandas etiam religiones sociorum, qui sacra et lucos et aras patriis amnibus
dicaverint: quin ipsum Tiberim nolle prorsus accolis fluviis orbatum minore
gloria fluere, and eventually no change was made.
See also the note on tumentem 86.
31 deus ipse loci: virtually equivalent to genius loci. There is an
etymological connection between genius and ingenium: the genius loci was
the character of the place, which became more and more personalised as
religious thought became more and more sophisticated.
See R. D. Williams on A. 5.95, and R. B. Onians, The Origins of European
Thought 2 , Index, s.v. genius.
31 fluvio (Tiberinus) amoeno: the same phrase at A. 7.30 is a
descriptive ablative, and it is best so taken here also because of the word-
order, the phrase enclosing the proper noun: Virgil may well have been
 COMMENTARY 21

trying to devise an equivalent to a Greek compound adjective (ioppoo, or

Ka>J.tppoo,, both Homeric). But it may simultaneously function as an
ablative of 'place from which'.
Tiberinus: Servius here comments bene Tiberinus, quia supra dixerat
deus: nam in sacris Tiberinus, in coenolexia Tiberis, in poemate Thybris
vocatur. Virgil only twice uses the' ordinary language' form Tiberis, in the
topological description at G. 1.499, and in the geographical catalogue at
A. 7.715. Tiberinus is certainly most appropriate to the deity; it is the form
which occurs in the prayers of the pontifices and augurs (Cic. Nat. Deorum
3.52 in augurum precatione Tiberinum . .. alia propinquorum ftuminum
nomina videmus), and in the formula of the ancient prayer, uttered no
doubt during summer drought, adesto, Tiberine, cum tuis undis (Servius on
A. 8.72); forinscriptional evidence, cf. GIL n.1.3057 on a stone discovered
not far from the Tiber channel, Sex. Atusius ... primus omnium aram
Tiberino posuit quam caligatus voverat. The termination -inus was popular
in plebeian Latin and gave rise to numerous names of divinities, like
Libitina (see F. T. Cooper, Word Formation in the Roman Sermo Plebeius
(New York, 1895), 139 ff.).
32 populeas ... frondes: poplars naturally grow on river-banks, but
the leaves also suit the river-god pictorially, aqueously grey-green (cf.
glauco 33), and hoary, like his hair (senior). Poplar sprays are worn by a
priest of a river, the Phasis, at Val. Fl. 6.296 f.
32 populeas ... frondes: lines 'framed' by an adjective at the
beginning and the corresponding noun at the end are fairly common
(cf. also 58, n6, 235, 280, 286, 300, 526 (causing hypallage), 704; and
cf. 207, 690). The reverse order, with noun at the beginning and adjective
at the end, is much rarer (only ten examples in the Aeneid-in A. 8 only
one, at 315). Virgil was fond but not over-fond of this kind of word-
symmetry. Ennius and Lucretius had stumbled into it, but rarely and
apparently unconsciously (although there was Homeric precedent for it,
e.g. Od. 22.396 Sµ,cpawv UK01T6, laa, KaTtt µ,!yap' ~Jl,€T€pa.wv), whereas the
Roman poets who imitated the Alexandrians had overworked it (in
Catullus' epyllion (64) it occurs 21 times in 408 verses). In this as in other
matters, Virgil steered a middle course between the crudeness of the older
poets and the self-conscious artistry of the moderns, using a design fre-
quently enough to be effective without becoming tediously repetitive (in
A. 6, for example, it occurs 14 times in 901 verses). See the Appendix,
p. 200, and Norden 391 f.
32 inter: it is common enough in poetry to find instances of anastrophe
(' turning round') with disyllabic prepositions, so that they follow the noun
or pronoun which they govern (cf. 3ro omnia circum; 418 quam subter;
631 ubera circum; 671 haec inter); similarly the preposition may be inter-
posed between noun and adjective (as at 297 ossa super . .. semesa with
intervening words), or between adjective and noun (as here, and 608 At
Venus aetherios inter dea candida nimbos). Marouzeau discusses the whole
topic (L'Ordre des Mots 3, 44 ff.), and observes (p. 58) that instances are
particularly frequent with inter, as if the very sense of this preposition
 22 COMMENTARY

predisposed it to being interposed. Virgil has a simple example at G. 1.33

Erigonen inter Chelasque sequentis, but in this line and at 608 he has gone
further: not merely the preposition, but the senior literally appears placed
between populeas . . .frondes, and the goddess likewise between aetherios .. .
nimbos: word-symmetry is accommodated to sense with almost neoteric
playfulness; cf. also A. 12.583 exoritur trepidos inter discordia civis.
Similar too is G. 1.259/rigidus agricolam si quando continet imber where the
agricola is literally 'hemmed in' by the frigid us . .. imber: here the word-
symmetry is virtually made possible by the delayed conjunction si quando
(see the note on 22 ubi). See further Wilkinson 65 f.
32 senior: 'Old Man River' as always. Rivers and seas naturally
impressed the primitive mind as primeval, elemental forces, and their pre-
siding deities were worshipped with the respect and veneration accorded
(especially in primitive societies like early Rome, as Aulus Gellius points
out, N.A. 2.15.1) to old age (cf. A. 5.823 et senior Glauci chorus . .. ). See
also the note on genitor 72.
33 A pause after a trochaic word at the beginning of the line (as here,
visus) is not nearly so common as after an opening dactylic word (as in the
next line, carbasus). Both pauses can be, and often are, used simply to
secure variety; but the light tripping rhythm of a trochee, if halted by a
pause, naturally suggests quick light easy action, so visus here, surgit, et
68 (where there is a light pause whether marked or not), and in the formula
dixit, et 366, 615 (compare the dactylic formula dixerat 276, 387 which also
despatches rapidly an unimportant idea).
33 eum: the acc. masc. sing. of is occurs twice in this book (also at
line 576), only in four other places in the Aeneid; the acc. masc. pl. eos
only once at A. 1.413. The oblique cases of is are weak and colourless words
in a style which aims at economy and density of expression, and Virgil
generally prefers the oblique cases of vir as more forceful equivalents; prose
authors had done the same before him, using also the oblique cases of homo
(see Holden on Cic. Sest. 22). The infrequency of is and its forms in poetry
after the prosaic Lucretius is examined by Axelson 70 ff. Further statistics
for the elegists, who show a like aversion, are given by Platnauer,
Appendix A.
33 f. Tiberinus appears with the conventional trappings of a Greco-
Roman river god. In art (coins, statues, bas-reliefs) he is represented as a
bearded old man reclining on his elbow and letting water flow from an um;
he is scantily clad and crowned with reeds (see J. A. Hild in Daremberg-
Saglio 5, 298 ff. s.v. and esp. figg. 6938-40). Virgil clothes him in carbasus,
linen (of the kind also used for sails) made from a type of fine flax which,
according to Pliny N.H. 19.(2).ro grows well by river-irrigation and for
which the Tiber valley was famed (in the Augustan period) according to
Grattius, Cynegetica 36 ff. aprico Tuscorum stuppea campo / messis conti-
guum sorbens de flumine rorem, / qua cultor Latii per opaca silentia Thybris /
labitur inque sinus magno venit ore marinos. His headgear is naturally made
of grey-green reeds, cf. A. ro.205 f. velatus harundine glauca (of the river
Mincius), and Ovid A.A. 1.223 hie est Euphrates, praecinctus harundine
 COMMENTARY 23

frontem (of a picture carried in a triumphal procession). The river-god's

dress was Greek in origin before the Romans adopted it; but at A. 10.423
Thybris pater has an Italian oak on which the spoils taken from the enemy
are dedicated: another example of the syncretism of Greek and Roman
myth and religion which Virgil promoted.
33 ~lauco, Gr. yAavK6s, is one of anumberof colour-adjectives (fulvus and
fiavus are others) which are emotive and suggestive rather than precisely
descriptive; glaucus refers to some shade in a grey-green-blue complex.
34 harundo: collective singular, see the note on abiete 599.
35 = A. 2.775 and 3.153: a repeatable formula for consoling apparitions
(Creusa, Penates, Tiberinus).
adfari. .. demere: may be regarded as historic infinitives (as Page
ad loc.); or, like attollere, after visus (as Servius auctus here, and R. D.
Williams on A. 3.153): in strict logic auditus adfari would be required, but
videor was weakened idiomatically = 'seem', and even when not so
weakened video is used alone in synaesthetic expressions where sight- and
sound-impressions are closely linked, cf. A. 2.732 visus adesse pedum
sonitus, A. 6.257 visaeque canes ululare per urbem, and in prose Cic. Or. 168
contiones saepe exclamare vidi. See also the note on 526 mugire. For historic
infinitives see the note on 493.

36-65 The river-god greets Aeneas and assures him that he has at last
reached the promised land. In token of this he will see a white sow with a
litter of thirty young under the oak-trees on the river-bank (an omen of the
founding of Alba Longa by Ascanius). He then advises Aeneas to solve his
most pressing problem (the lack of troops) by seeking an alliance with Evander
and the Arcadians, to whom he will himself lead him. Finally the river-god
reveals himself as Tiberinus.
Notice how Tiberinus' speech is designed to give Aeneas assurance and
encouragement. This is no deceptive dream; the truth of what is said will
be confirmed by an omen, and the revelation of the god's identity confirms
its truthfulness beyond doubt. Both of the immediate obstacles can be
overcome (superes 58 .. . supera 61), the first, the down-stream current, by
Tiberinus' own help, the second, Juno's anger, by Aeneas' prayers.
37 qui: relative pronouns are more commonly postponed from their
position in the clause than any other connective word; usually they occupy
the second place (e.g. 62, no, 141, 226, 239, 272, 340, 564, 680), less
frequently an even later one (as here, 194, 427). By vacating the leading
position in the clause, they allow greater emphasis to fall on the word(s)
before them, but this should not always be assumed: earlier usage in prose
and poetry had become fairly fluid, and the demands of metre played a
part. See also the note on postponed conjunctions, 22 ubi.
37 revehis: 'you bring back' the civil traditions which had been
founded among the Trojan race by Dardanus, after his emigration to
Phrygia from Corythus in Etruria: see the note on 134.
37 aeterna: here a 'proleptic' adjective whose sense anticipates what
24 COMMENTARY

is to be the case rather than stating what is at the moment true; cf.
Servius auctus 'pro servas et aeterna f acis '. In this idiom, which is rare in
prose but more frequent in poetry, the adjective is predicative, and quite
unambiguously describes a state which results from the action of the
verb, cf., for example, A. 10.103 premit placida aequora pontus.
When the adjective is a perfect participle the situation is not so clear
because a number of factors are at work. The idiom urbem captam incendit
was well established by the time of Caesar to mean 'he captured the city
and (then) burnt it'; in other words the past time-reference of the perfect
participle could be taken for granted. But originally this participle had no
specific time-reference (see the note on 636). Virgil therefore was moving
(as usual) in the direction of archaism in devising an idiom with the same
formal structure as the Caesarian idiom mentioned above, but with an
opposite meaning: at, for example, A. 1.69 submersasque obrue puppes, the
time-reference of the participle must be either contemporaneous or sub-
sequent to that of the main verb, either 'overwhelm the ships and sink
them' or 'overwhelm the ships so that they sink'. In examples of this kind
the participle can be labelled 'proleptic' because it has obvious links with
pure adjectives which are proleptic, but only the context can decide
whether the participle refers to the same time as that of the main verb or
a subsequent time, and it is often difficult to pin this down precisely:
cf. G. 4.547 placatam Eurydicen vitula venerabere caesa; A. 8.227 jultosque
emuniit obice pastis; A. 8.260 f. angit inhaerens / elisos oculos. See further
Conway on A. 1.69; Palmer, LL 327; KS r, 239 f.
38 exspectate ... : cf. the oracle of Faunus, A. 7.254 ff.
38 solo Laurenti. .. arvisque Latinis: either ablative of place, or
possibly of agent. In the latter case Virgil is employing the pathetic fallacy
(i.e. attributing personal, human feelings to nature) as in the Eclogues
(e.g. 1.38 f. ipsae te Tityre, pinus, / ipsi te fontes, ipsa haec arbusta vocabant),
and especially in the T61ros of the pastoral lament (e.g. E. 10.13 ff. illum
etiam lauri, etiam flevere myricae, / pinifer illum etiam sola sub rupe
iacentem / M aenalus, et gelidi fleverunt saxa L ycaei).
39 certa . .. certi: when a disyllabic word is repeated in the same line
(for rhetorical emphasis or to enhance pathos) it is Virgil's practice to vary
the position of the metrical stress by making it coincide with the word-
accent on its first occurrence and clash with it on its second: cf. 45 alba . ..
albi; 71 nymphae . .. nymphae; 76 semper . .. semper. The same thing
happens with a word-group, iam tum, in successive lines at 349 f.
39 ne absiste: ne with a present imperative in second person prohibi-
tions is common in Old Latin but was rejected by the classical prose
writers. It is one of many archaic features of style revived by Virgil to give
dignity to his epic, and is used once by Livy, 3.2.9 (a soldier of the Aequi
shouts to the Romans) crastino die oriente sole redite in aciem; erit copia
pugnandi, ne timete, where Livy is trying to suit to the speaker the tone
of old-fashioned downrightness which this construction had come to
convey. ne absiste requires either ab incepto or a loco to complete its sense.
Virgil consciously exploited the wide area of meaning which certain words
 COMMENTARY 25

have when not further qualified-another instance of means (economy of

expression) going hand in hand with intention.
40 tumor . .. et irae: cf. A. 6.407 tumida ex ira titm corda residunt. The
•swelling rage' (tumor . .. et irae are best taken as a hendiadys) is perhaps
another metaphor from the sea (cf.the note on 19), but more probably from
the swelling of the liver to produce, or other organs to absorb, bile (cf. fl.
9.646 ol8av€Tat KpaSl-r1 x6.\cp and R. B. Onians, The Origins of European
Thought 2 (Cambridge, 1954), 84 ff.). When two substantives are linked by
a simple copula to express two complementary aspects of the same thing,
it is usual for the one with more general application to be placed first, as
here (tumor, a swelling of any kind; irae, a specific emotion).
41 concessere: this word has presented a problem of interpretation
ever since antiquity. Some scholars (cf. Servius auctus) took it as equi-
valent todiscessere, 'have vanished') (cf. A. 2.91 superis concessit ab oris;
A. I0.819 f. tum vita per auras/ concessit maesta ad manis). The objection
to this, of course, is that it is not literally true: Juno was still as hostile as
ever. Servius suggests that concedo is here used in the sense of permitto or
indulgeo (cf. E. 2.57, IO.I), 'the angry ferment of the gods has allowed it'
(sc. the founding of the certa domus), and to answer the same objection he
points out that even Juno's anger did not alter the decree of fate (at
A. 7.313 f. she says non dabitur regnis, esto, prohibere Latinis, / atque
immota manet Jatis Lavinia coniunx). But the ellipse of any word or phrase
to indicate what was permitted is extremely harsh, and the ancient
supplement profugis nova moenia Teucris, perhaps inspired by A. I0.158,
can hardly be taken as certainly representing Virgil's intention, since he
could so easily have completed the line and the sense in this way himself.
The solution to the problem is that lack of truth is not a relevant
objection (as was seen in antiquity, cf. Servius auctus: alii ita tradunt:
nondum concesserunt, sed utiliter dissimulat (sc. Tiberinus)). Even though
he alludes to Juno's present hostility in 60 f. he is at the moment concealing
the truth to encourage Aeneas. Epic tradition sanctioned lies in the mouths
of gods and heroes (cf., for example, Athene at Od. r.179 ff. and Venus at
A. 5.794 f.) as well as mere mortals (cf. the note on IO ff.). All the most
notable nocturnal apparitions to Aeneas give him advice and encourage-
ment when he is in doubt and despair (cf. A. 2.270 ff., Hector; 3.147 ff.,
Penates; 4.556ff., Mercury; 5.722ff., Anchises): in such contexts mis-
representation is permitted and expected.
41 One of the three 'half-lines' in this book (the others are 469 and 536).
Some scholars believe that Virgil intended to leave the incomplete lines
as they were and that they would still have remained if Virgil had lived to
carry out a final revision of the Aeneid. If this had been his intention, it is
very strange that of succeeding hexameter poets who imitated his every
mannerism so closely, not one should have imitated him in this also; and
it would be even more surprising if Virgil, who had such great respect for
precedent in matters of form, should have seen fit to make such an inno-
vation (and that only in his epic poem). Some of the 'half-lines' are ad-
mittedly effective in themselves, and it is difficult to see how some of them
 COMMENTARY

could have been completed-but it is no doubt this very difficulty which

caused Virgil to leave them temporarily, because he knew that he would
have to change the surrounding lines as well, and could not interrupt the
flow of inspiration to settle the details of a troublesome passage. This is the
explanation given by Donatus in his Life of Virgil (22 ff.) in a section which
is of cardinal importance for understanding Virgil's methods of composition:
'We are told that every day he used to dictate a very large number of
verses which he had worked out in the morning, and during the whole day
he would work them over and reduce them to a very small number, aptly
saying that he gave birth to a shapeless poem like a mother bear giving
birth to a cub, and that at long last he licked it into shape. The Aeneid was
first sketched in prose and set out into twelve books, and he began to
versify it piece by piece as any part took his fancy and not taking any-
thing systematically (nihil in ordinem arripiens); and, so that nothing
should check his inspiration, he passed over some things which were
unfinished ( quaedam imperfecta transmisit) and others he propped up with
trivial words, which he used to say were placed at intervals through a
passage to serve as props (pro tibicinibus), to hold the work up until solid
columns arrived.'
See Sparrow, Half-Lines and Repetitions in Virgil (Oxford, 1931) for a
cautious and necessarily inconclusive examination of the whole topic, and
F. W. Lenz, The Incomplete Verses in Vergil's Eneid in Vergiliana (ed.
H. Bardon and R. Verdiere, Leiden, 1971), 158 ff.
Inventing supplements to complete the half-lines became a favourite
pastime soon after the Aeneid's publication; most of them fail to impress
even as speculations because the majority of the half-lines themselves
complete the sense of the passage they end (cf. Donatus Vita 41).
42-9 Following a suggestion of Heyne, Ribbeck wished to delete these
lines from iamque to haud incerta cano. (This would solve the problem of
the half-line 41 by making the passage run concessere deum. nunc qua
ratione quod instat .. . ). Hirtzel thought that they had not received their
final form, Mackall that they were a later insertion. The chief problem is
that 43-6 are, with the exception of the one word hie 46, an exact repetition
of Helenus' prophecy in A. 3.390-3. But it is no serious objection that the
same prophecy is made by both Helenus and Tiberinus: the portent of
'eating the tables' is announced by the Fury Celaeno (A. 3.250 ff.), but
attributed to Anchises (A. 7.123 ff.). Helenus' prophecy is more integrally
connected with its context than 42-9 here; but if 42-9 are deleted the
actual appearance of the sow, 81 ff., is left without interpretation of its
significance or back-reference to Helenus. It seems highly probable that
the Aeneid left Virgil's hand with the prophecy repeated. Varius and
Tucca, commissioned by Augustus to remove superfluities (according to
Donatus' Life of Virgil), presumably left both intact. Grassmann-Fischer
54 f. plausibly argues that A. 3.389 cum tibi sollicito secreti ad fluminis
undam is so aptly realised in A. 8 (anxiety: 19, 29; locality: 28, 31) that
the repetition of the prophecy must be intentionally planned.
One can only speculate about what Virgil's final intention might have
 COMMENTARY 27

been. There are a number of repetitions or inconsistencies between A. 3

and other books which would be explained if, as Nettleship suggested
(p. 25 of his introduction to Conington's ed.) A. 3 was the first book to be
sketched, when Virgil intended to follow an annalist sequence rather than
Homer's example of plunging in medias res (see M. M. Crump, The Growth
of the Aeneid (Blackwell, Oxford, 1920), ch. 2). As the work grew, the
poet's ideas also grew and changed: A. 3 would probably have received
more drastic alteration than any other at the final revision.
42 iamque tibi: prophecies regularly begin by addressing the interested
party, cf. cum te, nate ... A. 7.124 ff.; this is here expressed by a dative of
interest with the intransitive iacebit (44).
43 litoreis: this adjective, probably a coinage of Virgil's, occurs in the
Aeneid only here ( = A. 3.390) and at A. 12.248, also in an omen-scene
with rather exalted language. ripa, the usual word for a river-bank, had
produced no adjectival form by Virgil's time.
43 The pattern of assonance and alliteration is evident enough:
litoreis ingens i'nventa sub ilicibus sus, but note that although the first
syllables of ingens and inventa are long metrically, the vowel sound itself
is short: cf. Cic. Or. 159 quid vero hoc elegantius, quod non fit natura sed
quodam instituto: indoctus dicimus brevi prima littera, ins anus producta,
inhumanus brevi, infelix longa et, ne multis, quibus in verbis eae primae
litterae sunt quae in sapiente atque felice, in producte dicitur, in ceteris
omnibus breviter. On the important distinction between vowel-length and
syllabic quantity see W. S. Allen, Vox Latina (Cambridge, 1965), 89 ff.
ingens: the first appearance in this book of the Aeneid's most over-
worked adjective; see the note on 192.
43-5 The legend of the sow. The modifications which the story of the sow
underwent are instructive. The earliest version, that of Lycophron
(1255 ff.), gives these details: a black sow will be carried on board Aeneas'
ships from Troy; having' numbered' her litter of thirty, Aeneas will found
a city near the Latins; he will make a bronze image of the sow and
consecrate it to the city. This account clearly implies that the sow and her
litter belong to the class of signa or monstra sent by a guiding deity to
indicate to the coloniser where he should settle; such monstra may be
capable of further interpretation, like the horse's head in A. 1.443 ff.,
which not only indicates a settling-place to the Poeni, but is also an omen
of military distinction and abundant sustenance.
The early annalist Fabius Pictor (apud Diodorum 7.5 = frg. 4 HRF
Peter) tells of a white sow escaping from the sacrifice and leading the
Trojans to the site of Alba Longa. It looks as if an important detail has
been changed (assuming Lycophron's K€AatV7J not to refer to the bronze of
the later statue, or to be vaguely = horrenda) to form an aetiological
myth to explain the name Alba, and to bring that town into connection
with the Aeneas saga.
Varro (L.L. 5.144), deriving the name Alba from the sus alba in this
way, adds that when the sow had escaped from Aeneas' ship to Lavinium
she gave birth to thirty pigs, and as a result of this portent Alba was
4 EAC
 COMMENTARY

founded thirty years after the foundation of Lavinium. Lavinium had

obviously gained prominence in the saga as Aeneas' city (perhaps to meet
the demands of a crude chronology), and was duly incorporated in the
legend.
Dionysius of Halicarnassus (r.56), probably drawing on works of Varro
no longer extant, supplies two important details not yet explicit in the
tradition: (1) after the sow had been followed to Lavinium, the Trojans
were dismayed at the unsuitable site where she had chosen to stop, but
were cheered by the prediction that after a number of years corresponding
to the number of the sow's litter, a new city would be founded from there;
(2) this prediction was made by a divine voice issuing from a wood, or
(according to another version he records) delivered to Aeneas in a dream
by a god of his country. In this account we can perhaps detect (1) embar-
rassment at Aeneas' connection with two city foundations, and (2) the
idea for Virgil's Tiberinus.
Varro (R.R. 2.4.18) says that this was the most ancient example of the
prodigium of a sow with more young than teats; and that a bronze statue
of the group existed in his day, and the body of the sow, pickled in brine,
was displayed by the priests (an early example of that local Italian pride
in relics which encourages fraud).
The panel of the' Ara Pacis' with the sacrifice of Aeneas (CAH Plates 4,
122 f.) shows him with covered head about to sacrifice the sow. Like
Virgil, the artist regarded the sow primarily as a foundation-symbol: in
the background is a little shrine of the penates (certi, A. 8.39). It is possible
that the artist conceived the sacrifice as being made to the female figure on
the panel on the opposite side of the altar; there is a seated female figure
on the same panel of the Altar of the Belvedere which shows Aeneas with
the sow and, for reasons of space, less than thirty young (see L. R. Taylor,
The Divinity of the Roman Emperor (Middletown, Connecticut, 1931),
188 f.). The female figure on both altars has been identified with Terra
Mater, but whatever the artists' intentions, there is nothing in the narra-
tive of the Aeneid to suggest that Virgil is equating Juno with Terra Mater
or any other deity or personification.
As Alfoldi suggests (271 ff.), the sow, or rather wild boar, may originally
have been a Latian tribal totem which developed into a symbol of leader-
ship; variations in the myth reflect competing claims to the leadership of
Latium.
See further W. Ehlers,' Die Griindungsprodigien von Lavinium undAlba
Longa', MH 6 (1949), 166-75.
43 sub ilicibus sus: the same closing monosyllable at Lucretius 5.25
et horrens Arcadius sus, and G. 3.255 ipse ruit dentesque Sabellicus exacuit
sus. Norden 441 points out that with monosyllabic nouns of this type
ending the verse there is a clear echo of Greek practice. Often the rhythm
they produce has evident pictorial effect, as at A. 5.481 sternitur exani-
misque tremens procumbit humi bas (see Williams' note there for references
to the relevant literature); at A. 8.83 the sus, the long-delayed subject of
the sentence, appears with the effect of surprise; here, with the sound-play
 COMMENTARY 29

-bus sus, it is almost a comic surprise (like the famous exiguus mus of
G. r.181): the sow is an authentic but quaint detail of the tradition.
44 triginta capitum fetus enixa: 'having delivered a litter of thirty
young'. capitum fetus is a dignified periphrasis for porcelli, avoided in-
tentionally by Virgil (cf. nati 45;jetu 82; grege 85). Of animal diminutives
bucula, capella and capreolus occur in the Eclogues (pastoral poetry was
regarded as holding a low place in the hierarchy of literary forms), bucula,
capella and asellus in the Georgics (with stricter stylistic requirements).
In any case capella and asellus seem to have been preferred to capra and
asinus in all varieties of poetry. No animal diminutive occurs in the Aeneid
except catulus, in a simile, at A. 2.357. See Axelson 40 and 44 f., and the
note on 660 sagula.
45 R. D. Williams (on A. 3.392) is right in maintaining that the phrase
alba solo recubans represents one idea only, and there should be no comma
after alba as in some editions.
46 This line is generally regarded as spurious, and is suspect on grounds
of text, style and meaning:
(1) It is not found in either of the oldest most complete MSS, the
M ediceus and Palatinus; it was not paraphrased by Tiberius Donatus, and
it was apparently absent from texts available to Servius because he
understands ex quo to mean qua ratiocinatione (if he had had 46 in front of
him he would surely have taken it to mean ex quo loco).
(2) It involves a clumsy repetition of words in the following line: hie
locus urbis erit . .. ex quo (loco) ... urbem . .. A scanius . .. condet.
(3) As A. 3.393 (with is for the hie here) it is appropriate: Helenus
means that the country in which the omen occurs will provide the site for
the city. Here it is inappropriate: locus can hardly mean 'country'
(= Latium, Grassmann-Fischer 58 f.), and if it did the line would be a
rather pointless repetition of the sense of 39; equally it cannot mean
'precise spot' because Aeneas' city Lavinium, which is presumably
referred to here, was twelve (Roman) miles from the Tiber (see the note on
1 Laurenti), and an allusion to it in the middle of a context exclusively
concerned with Alba Longa is too awkward to be acceptable (see Grass-
mann-Fischer 61 for the contrary view).
Tiberinus is simply making the equation alba sus = Alba Longa,
triginta annos = the time which will elapse before Ascanius founds it (the
traditional interpretation of the legend of the portent, cf. Varro R.R. 2.4,
L.L. 5.144 referred to in the note on 43-5). The portent will establish the
truth of Tiberinus' prediction that the Latin plain is Aeneas' destined
home and the seat of power of his son and descendants; Aeneas' city
Lavinium is not mentioned and remains under the veil of mystery which
Virgil casts over it.
The cause of the interpolation is either that a knowledgeable scribe let
his memory run on too far, or that a marginal note of A. 3.390-3 (for
comparison) was incorporated in toto into the text here in spite of the
inappropriateness of the last line (the same thing has happened in a ninth
century MS, Bernensis 255/239, at A. 8.60 f. where a recollection of
4-2
30 COMMENTARY

A. 3.438 f. has ousted the true reading, and may have happened much
earlier in the tradition at A. 6.702 which inappropriately repeats A.
2.794).
V. Buchheit in Philologus 109 (1965), 104 ff. argues convincingly that
Tibullus' poem 2.5 shows his knowledge of the whole of the Aeneid and in
particular of Tiberinus' speech in A. 8; but it is doubtful whether the very
common words of the phrase hie magnae iam locus urbis erit (Tib. 2.5.56,
and cf. Ovid F asti 2.280 hie, ubi nunc urbs est, tum locus urbis erat and
Met. 15.18 hie locus urbis erit) prove the authenticity of A. 8.46, or even that
this line was in Tibullus' copy of the Aeneid.
46 The demonstrative pronoun is, as usual, attracted into the gender
of the predicate: the strictly logical hoc erit locus, id erit requies becomes
hie erit locus, ea erit requies (cf. A. 6.129 hoc opus, hie labor est; A. 1.17 hoc
(for hanc = Karthaginem) regnum dea gentibus esse . .. ) ; and in prose, Cic.
Phil. 7.14 quamquam illa (for illud) legatio non est. Sporadic examples of
the more logical construction appear in prose from the early Empire
onwards, e.g. Seneca Tranq. 2.4 id tranquillitas erit.
47 f. •Consequently, in the course of thrice ten revolving years, Ascanius
will found the city of Alba with its brilliant name.' ex quo = •after and
because of this'; cf. ex illo at A. 2.169 and 12.32. The archaic tone of
Tiberinus' oracle is sustained by the poeticism ter denis (the force of the
distributive numeral being lost), an echo of Lucretius (1.3n multis solis
redeimtibus annis, itself an echo of the Homeric TTEpmAoµl1,1w1,1 (TTEPLTEAA0-
µl1,1w1,1) J1,1iavTwJJ at, for example, Od. 1.16; fl. 2.551), the forceful pattern of
assonance and alliteration in Ascanius dari condet cognominis Albam, and
by the ambiguity of clari = •bright shining' and• distinguished', a punning
gloss on Alba.
Prophecies cannot be pressed for the precision they generally aim to
avoid, but if Tiberinus means that Ascanius will found Alba thirty years
after Aeneas sees the sow, this is not inconsistent with A. 1.267 ff. where
it is prophesied that Ascanius will reign for thirty years and transfer the
kingdom from Lavinium to Alba, unless this is implausibly pressed to
imply that the transfer took place at the end of Ascanius' reign.
The descendants of Aeneas as the kings of Alba are a fiction, introduced
into the annalist tradition by Fabius Pictor, designed to fill the gap between
conjectural dates for the Trojan War and the founding of Rome by
Romulus. Virgil implies a date for the Trojan War, much later than that
accepted as traditional by Roman historians, in Jupiter's prophecy,
A. 1.26r ff. There, as here, he exploits the associations of the number
three and its multiples in folklore and ritual. For other evidence of his
preoccupation with significant numbers see 0. A. W. Dilke, CQ N.S. 17, 2
(1967), 322-6.
49 baud negatives the adjective incerta; it is rarely found with verbs
outside Old Latin, except in formulae like haud scio (an).
49 cano: as always of prophetic utterance (cf. 340, 499, 534), because
it was metrical and so nearer to •singing' or •chanting' than speech.
49 ff. It is noticeable that when Virgil has a sense pause at the strong
 COMMENTARY 31

caesura in the third foot, he often repeats it in the next line as well, or
even as here and 580-2 in the next two lines; cf. 64 f., 185 f., 223 f., 298 f.,
362 f., 4n f., 442 f., 514 f., 535 f., 572 f., 594 f., 621 f. But this does not
alter the fact that the majority of such pauses are isolated, and in any
verse-paragraph are not frequent enough to become monotonous. One of
Ovid's failings as a hexameter poet is that he could not unlearn elegiac
technique: in the Metamorphoses pauses at this position are far more
frequent than in the Aeneid, no doubt because the metrical unit which
precedes it is identical with that of half a pentameter.
50 adverte: the full expression anim(um)adverto (cf. the variation at
A. 8.440 advertite mentem) is the older and original formula; advertere alone
in the same sense is first found in poetry with Virgil (A. 4.n6, the
same phrase as here), and appears only sporadically in later prose and
poetry.
Another variation of the same formula, also with ellipse of verbis, occurs
at A. 11.314 f. nunc adeo quae sit dubiae sententia menti, / expediam et paucis
(animos adhibete) docebo (whence, and perhaps also from A. 6.759, expediam
has crept into the MS tradition here). It carries the didactic tone of self-
conscious authority (the speakers are Juno, Tiberinus and Latinus).
51 'The Arcadians, a race descended from Pallas, ... (chose a spot) on
these shores.' proficiscor here = 'originate from' (see LS s.v. II B 2). The
wide separation of his oris from delegere locum (53) and the more common
literal sense of profectum caused Servius to insist wrongly that the line
meant profectum ad has oras genus a Pallante, non a Pallante profectum
(although genus a Pallante could by itself = 'a race descended from
Pallas', cf. 628 f. genus omne futurae / stirpis ab A scanio).
51 f. Evander was the grandson of Pallas, king of Arcadia. After killing
his father, Echenus, he was driven into exile and persuaded by his mother
Nicostrate, alias Carmentis (quia carminibus vaticinabatur, Servius), to
make for Italy. With his followers he occupied part of the future site of
Rome, founding the city of Pallanteum on the mans Palatinus (montibus 53
is a poetic plural); both city and hill derived their names from the grand-
father. Such is the substance of the legend; variant versions report that
Evander's mother lived to be a hundred and was then killed by her son
(Servius auctus on this line), and that Palatinus was derived from Pallantia,
Evander's daughter, who had been raped by Hercules and buried on that
hill (Varro L.L. 5.53). Virgil, however, alludes neither to Evander's past
career as patri- and/or matri-cide, nor to Hercules' typical misbehaviour.
52 When a monosyllabic word (as qui here and ut 3) begins a line and so
necessarily bears the metrical stress, a repetition of it after a strong caesura
in the third or, as here, fourth foot places it in an unstressed position, and
makes for rhythmic variety; cf. the note on 39 certa . .. certi.
52 This line exemplifies one of the simplest types of 'theme and varia-
tion' (see the note on 7 f.), produced by splitting up the prose complex
qui regis Euandri signa sunt secuti.
52 secuti, sc. sunt: the third person of esse in compound tenses and
when used as a simple copulative verb is more frequently omitted than the
32 COMMENTARY

first or second, singular or plural; cf. effatus (est) 443 and see Wagner, QV 15
for a list. Note that here and in 362 it is omitted in a relative (i.e. sub-
ordinate) clause (cf. the omission of est in 184): see the note on 175.
53 delegere . .. posuere: poetry preferred the 3rd pl. perfect termi-
nation -ere to -erunt both for metrical convenience and per se: it occurs at
the end of a verse where -erunt would have been metrically possible,
e.g. A. 2.243 substitit atque utero sonitum quater arma dedere. Cicero (Or. 157)
found -ere flattering to the ear but felt that -erunt was more correct. The
latter is the standard form of classical prose; Livy admits -ere, especially
in Books r and 21, to give epic Ennian colouring. See Lofstedt, Syntactica
II, 295; Leumann 144 f.
54 proavus is here loosely used to mean 'their ancestor'; strictly it
means great-grandfather, and Pallas was Evander's grandfather. Virgil is
not meticulously precise about such details of ancestry: Pilumnus is said
to be Turn us' parent at A. 9.3, his grandfather at A. IO. 76, his great-great-
grandfather at A. ro.6I9.
54 Pallanteum: a spondaic fifth foot, making a versus spondeiazon.
Such verses are much more frequent in the poetry of the Alexandrians
than in Homer (roughly three times as frequent in Aratus' Phaenomena as
in the Iliad), and the neoterics favoured them to excess (in three con-
secutive lines at Catullus 64.78-80): Cicero makes a joke of the foible,
Att. 7.2.r Brundisium venimus ... usi tuafelicitate navigandi; ita belle nobis
'flavit ab Epiro lenissimus Onchesmites' (hunc a1ro110Ha,oV'Ta si cui voles
Twv 11£wTlpw11 pro tuo vendito); and his dislike was not affected-there is
only one instance (Orionis) in his translation of Aratus. On the other hand,
Ennius and Lucretius had by no means deliberately avoided them. In the
whole of Virgil (12,085 verses) there are only 33 instances (a severe
reaction against neoteric practice), and the majority of these are clear
imitations of Greek verse technique, with a Greek proper name (Pallanteum
here and 341) or loan-word (potest electro 402); some are echoes of earlier
Roman poetry, intervallo A. 5.320 of Lucretius, and A. 8.679 (= A. 3.12)
et magnis dis of Ennius (see the note there; this is the only example in
Virgil of a spondeiazon which does not end with a word of three or more
syllables). See further the notes on Argileti 345 and intertextam 167, and
Norden 438-46 for detailed statistics.
55 bell um assidue ducunt: 'continuously prolong war'; ducere and
trahere are common in this sense in the historians, cf. TLL 2, 1834 ff. s.v.
bellum.
55 Evander might have been thought to be an ally of Turnus, since
both were of Greek descent, and it is important that Tiberinus should let
Aeneas know that he is on the contrary a potential ally also at war with
the Latins, now led by Turnus (see note on 1). assidue is presumably a
rhetorical exaggeration, for according to A. 7.45 f. Latinus had had a long,
peaceful reign finally broken by the arrival of Aeneas and Turnus'
hostilities, and Virgil can hardly intend us to understand a reference to
Mezentius, an old enemy of the Arcadians (cf. 569 ff.), carrying on raids
against Pallanteum under Rutulian and Latin protection. According to
 COMMENTARY 33

Virgil, the tyrannical Mezentius had been expelled by his subject Etruscans
who were all prepared for war against Turnus and the Rutuli who were
sheltering him. But an oracle (cf. 502 f.) had ordered them to choose a
foreign leader: they had approached Evander, who had declined the
position but made an alliance with them, as a result of which he was being
attacked by the Rutuli (cf. 474).
On Virgil's own method of using the synthesis of legend and history
behind this situation, see note on 492 ff.
56 foedera iunge: the plainest and most literal prose expression is
sibi (con)iungere socios (dextram) foedere, and this is the expression used, in
its passive form, at A. 8.169 and A. ro.105. When the verb is active,
however, Virgil prefers the variation of the usual prose idiom, iungere
joedera (as at A. 7.546, 8.641, 12.822): this supports the reading foedera
iunge here, although the presence already of one possible object to the
verb, socios, emphasises the peculiarity of the variation and suggests the
'facilitation' foedere (a variant reading here). A similar lectio facilior appears
at A. 4.n2 miscerive probet populos aut foedera iungi, for the same reason
that an accusative subject of the infinitive occurs already in the line.
57 ripis et recto flumine: 'between my banks and straight up the
river'. recto describes the manner of going, not the river, which was not in
fact straight, cf. 95 tangos superant flexus. Similarly at A. 6.900 recto
litore = 'straight along the shore'; Terence Ad. 574 praeterito hac, recta
platea sursum, 'pass by this way, straight up the street'. This is another
case of an adverbial notion being expressed by an adjective (see note
on 30), in this case an adjective of direction (see Housman's note on Lucan
r.220 for other instances, obliquus, transversus, etc.).
57 et recto: in the Aeneid the fourth foot is predominantly spondaic
(there are only 2,873 dactylic fourth feet to 6,940 spondaic according to
Winbolt n5), and is said to be so in all Latin hexameter poets (although in
the Metamorphoses dactylic fourth feet come almost to equal spondaic). It
was clearly felt to be a built-in feature of hexameter rhythm that the verse
should be slowed down before the expected dactyl of the fifth foot. It is
not common, however, for the fourth foot to be a self-contained spondaic
word; when this occurs, the preceding word must be a monosyllable (or,
exceptionally, a pyrrhic disyllable as at A. 8.132) so that the verse can
have a main caesura, and this monosyllable is most often proclitic,
intimately connected with the spondaic word. Of the twenty-four self-
contained spondaic fourth feet in A. 8, thirteen are preceded by a con-
junction (et, ut, sed, ni), two (82, 697) by a preposition (cum, a). The
comparatively small number of times that a monosyllable or pyrrhic
disyllable was naturally available to precede a spondaic word in the
fourth foot is enough to account for the latter's rarity; when such a word
was available, a self-contained spondee could be admitted with no extra
special effect.
58 A remarkable line, for its rhythm (predominantly spondaic and
'heterodyned ', i.e. with clash of word-accent and metrical stress until the
last two feet), its alliteration (of sand m), the enclosing pattern of adjective
34 COMMENTARY

and noun (adversum ... amnem) (see note on 32), and the conjunction ut
postponed to the fifth place (see note on 22).
58 adversum ... amnem: 'the down-stream current'.
subvectus: subvehi is the regular word for sailing or rowing against the
stream, cf. esp. Livy 24.40.2 legati . .. venerunt . .. nuntiantes Philippum
primum Apolloniam temptasse, lembis biremibus centum viginti ftumine
adverso subvectum, and Caesar E.G. r.16.3.
59 sur~(e) a~e,: the energy and decisiveness of an imperative, perhaps
more than its metrical form, associate it with a stopped first-foot dactyl
(see the note on 87 leniit): in A. 8 compare egreder(e), o ... adloquer(e), ac
(very light pauses) 122-3, quare agit(e), o 273, ingreder(e), o 513.
59 primisque cadentibus astris: the general use of primus to convey
an adverbial notion is discussed in the note on 30 seram . .. quietem. But
when it is so used, it sometimes has a special nuance of meaning, pointed
out by Housman on Manilius 5.226: 'adiectivum primus pro adverbio
actionem circumscribente pauci interdum sic ponunt ut primus rem aliquam
f acere pative dicatur qui eam vel f acere vel pati incipit '. He paraphrases this
present phrase by 'astris cadere incipientibus ', and cites as other illustra-
tions of this idiom A. 11.573 f. utque pedum primis in/ans vestigia plantis /
institerat (which clearly cannot mean 'on her first feet', but 'on her feet
which she was using for the first time'); A. 6.8ro f. primam qui legibus
urbem / fundabit (sc. Numa) (which must mean 'he will establish the city
with laws which it did not have before his time'). So the phrase here means
'as soon as the stars begin to set'.
But this raises another problem: cadere used of heavenly bodies regularly
means 'to set, sink below the horizon', cf., for example, A. 4.480 Oceani
finem iuxta solemque cadentem, and is quite intelligible in connection with
the sun, moon, or a specific star or constellation; but with 'stars' in general
it is not clear, because some stars can be seen to sink below the horizon as
soon as they can be seen at all, i.e. at dusk. Poetic language cannot of
course be pressed too far for prosaic precision, but Virgil has transferred
a precise term to an imprecise context, and used cadere as if stars col-
lectively rose, passed their zenith and set. Night also is regarded in the
same way at A. 2.8 f. et iam nox umida caelo / praecipitat suadentque
cadentia sidera somnos, which means that the night and the stars are falling
from their metaphorical zenith towards setting, i.e. it is well past midnight.
With the phrase here cf. A. 4.81, G. r.439 f. solem certissima signa
sequentur, / et quae mane refert et quae surgentibus astris, and Propertius
4.4.63 f. et iam quarta canit venturam bucina lucem, / ipsaque in Oceanum
sidera lapsa cadunt.
59 Aeneas does not in fact include Juno in his prayer' at the first sign
of dawn' (71 ff.) as Tiberinus instructs, but no doubt prays to her during
the sacrifice of the white sow (84 f.), which seems to have taken place well
after dawn.
Virgil is not at pains to make explicit the chronology of events between
the Tiber-vision and Aeneas' arrival at Pallanteum. The probable sequence
seems to be: night (26 ff.); dawn (68 ff.); prayer (71 ff.); equipping of the
 COMMENTARY 35

two biremes (79 f.); beginning of journey (not explicitly mentioned but
implied by inceptum 90); appearance of the sow and sacrifice (81 ff.);
night (86, ea ... nocte); resumption of journey (90 ff.) (presumably during
the aforementioned night); continuation of journey during that same night
and part of the following day (94, noctemque diemque); arrival at the site
of Rome after midday (97 ff.). Total length of journey: one and a half
days.
This has been the generally accepted view of the time-sequence from
Servius (on 97) to Heinze (342 and 386 ff.) and Cartault (640). It rests on
the assumption that noctemque diemque 94 do actually refer to the night
and day following the sacrifice, and are not a vaguely emphatic cliche for
'a long time'; and that nocte 86 refers to the same night as 94, or if not, to
the immediately preceding one when the Tiber-vision appeared. (If 26, 86,
94 refer to three successive nights, the journey must have taken two and a
half days, which is rather excessive for rowing 30 odd kilometres, even
allowing for interruptions.)
Mackail's attempt (CR 32 (1918), 103 f.) to accommodate the chronology
of the passage to his personal experience of rowing over the same course
downstream in about five hours is not acceptable, because it involves
ignoring inceptum 90, taking spectans orientia solis / lumina 68 f. purely as
an indication of direction and not of time, and referring noctem 94 to the
actual night of the Tiber-vision. But this view is astonishingly repeated by
Knauer 249 n. 3 and Binder 39.
It is important to remember what Virgil has passed over in silence: the
sacrifice may well have taken the major part of a day (cf., for example,
Homer, Od. 9.556 ff.), not only to comply with ritual but also to allow time
for feasting (an important matter if, like the Trojans here, one is living off
the land). The equipping of the biremes may also have taken some time.
See G. E. Duckworth, A]Ph 49 (1938), 135 n. 2, for references to other
dissentients from the one-and-a-half days view, and add W. A. Camps,
PCPhS 185 (1959), 22.
No disagreement is possible about the chronology of the rest of the book:
103-368 afternoon and evening; 369-453 night; 454-731 morning colloquy
and departure to Caere later the same day.
60 f. The same advice was previously given by Helen us at A. 3.437 ff.;
the Trojans must repeatedly try to win over the implacable Juno. Virgil
makes of her a very important and complex symbol. For although she has
familiar associations in terms of Roman religion (function as pronuba
A. 4.166; cult-centres at Ardea A. 7.419, and Gabii A. 7.682 f.), and in
terms of Greco-Roman mythology (the judgment of Paris, etc., A. r.27 f.),
she appears first and foremost as the opponent of the destined future (and
so always at odds with Jupiter her husband), and the disrupter of rational
progress (Turnus and her other instruments are dominated by furor).
These most important aspects were built on comparatively slender
foundations: folk-mythology probably credited the queen-consort of the
gods with obstructive conservatism (Fricka in Nordic myth is a close
parallel); the stories of Io and Hercules Furens already suggested that
 COMMENTARY

madness was her weapon against her victims. Virgil also innovated in
equating Juno with the Carthaginian goddess Tanit and making her the
champion of Carthage (cf. A. 1.15 f.), for the statements of Macrobius
Sat. 3.9.7 f., that the Romans used the formula of evocatio to call out the
protecting deity of Carthage, and of Servius on A. 12.841, that Juno was
so called out in the Second Punic War and subsequently transferred to
Rome by Scipio, cannot be regarded as historical, nor can any reason for
inventing the fiction be found until long after the Augustan period.
60 The variant irasque, though unobjectionable both in meaning,
'bursts of anger' (cf. A. 8-40) and in the homoeoteleuton it produces with
minasque, has too little MS authority to be acceptable.
61 Tiberinus' intention was demere curas (35) (cf. the note on 41
concessere); the prophecy of ultimate success (victor here), and the fol-
lowing assurance that it is divinely inspired crowns his purpose.
61 honorem: sacrifice and worship, as generally in this book (76, 102,
189, 268, but not 339, 617).
62 ff. ego sum . .. Thybris: a river-god and therefore gifted with
prophetic powers, cf. the Clitumnus in Pliny Epp. 8.8 (quoted in the note
on 74 ff.). The same is true of sea-gods, e.g. Proteus, G. 4.391 ff., Nereus in
Hor. Odes 1.15. Water competed with wood in the primitive mind as the
elemental abode of divine oracular power; see further the notes on 348 ff.
and 336 ff.
62 pleno . .. flumine: a dignified description and a claim to honour,
irrigation being a benefit; but see the note on 86 for the other side of the
picture-the danger of flooding.
63 For the 'framing' of the line by two participles, see the Appendix,
p. 200.
64 caeruleus is surprising, as Page points out, because jlavus is a
constant epithet of the Tiber (cf. A. 7.31, 9.816; Ovid Met. 14.448; Hor.
Odes 1.2.13); Dryden actually translates this line 'The God am I, whose
yellow water flows'. caeruleus is the standing epithet for sea-gods (e.g.
Proteus, G. 4.388; and cf. A. 3.432 Scyllam et caeruleis canibus resonantia
saxa), and it is possible that Virgil chose it here because Aeneas is in fact
near the sea (cf. A. 7.35 f.); the same adjective is used at 713 of the Nile
delta, where caeruleum gremium and latebrosa flumina are 'theme and
variation'. But it is more likely that Virgil was attracted by the sound-
echo between caelo and its derivative caerul(e)us (originally caelulus);
word-play is the obvious point of Ennius' caeli caerula templa (Ann. 49 V),
imitated by Lucretius r.1090 per caeli caerula.
caerul(e)us originally= 'sky-like', hence the wide range of actual
colours it could connote; the only certain quality of water which reflects
the changing colours of the sky is clearness.
For the adjective applied to rivers, cf. Tibullus r.7.12-14, of the Loire
and the Cydnus.
64 caelo gratissimus amnis recalls Sdcf,,>.os (or 9,rncf,,>.~s); cf. A.
12. 142 (Juno to Juturna) nympha, decus fluviorum, animo gratissima nostro.
Tiberinus hints at the power he has through Jupiter's favour (caelum =
 COMMENTARY 37

Diespiter), implying that his words deserve extra credence because his
authority is backed by a higher power, and perhaps that the majesty of
his river depends on constant rain-water supplied from, and by the favour
of, heaven. Cf. Homer's description of the Nile (Od. 4.477): Alyv1TToio
Su1TET£of 1T0Taµofo, where, as Stanford points out, 'falling by ordinance of
Zeus' properly refers to the rain not the river.
65 hie mihi magna domus celsis caput urbibus exit: so it would
appear, with no punctuation, in the earliest and best MSS. At first sight
a puzzling line because of its ambiguities: what is the case and function of
celsis . .. urbibus, and is it a true or a poetic plural? Does caput mean 'a
capital city', or 'the source or upper part of a stream' (cf. Digest 43.20
caput est unde aqua nascitur), or 'the lower part of a stream' ( = os, as at,
for example, Caesar E.G. 4.ro.5)?
Servius and Tiberius Donatus instinctively saw that the natural inter-
pretation of the line is 'here is my great home; the upper part of my
stream flows from among lofty cities' (the two halves of the line are anti-
thetical, and should be separated by a semi-colon after domus). This is
certainly correct: (1) exit is echoed by exis 75, where the source of the river
is specifically mentioned; (2) caput = 'source' at G. 4.368 in a context with
many similar river-associations to those in A. 8; (3) celsis . .. urbibus refers
to the Etruscan hill-fortresses on the Apennines, like Arezzo and Perugia,
whose sites suit exactly with a literal meaning of celsis; (4) the Tiber is
constantly 'Etruscan' (Tuscus G. 1.499, A. 8.473; Tyrrhenus A. 7.242;
Lydius A. 2.781), and Virgil is here attempting a periphrasis for the idea
commonly expressed by advena Thybris (at, for example, Ovid Fasti
3.524), 'the Tiber which comes to us Romans as a stranger from a foreign
land'; cf. also Ovid Met. 15.432 (Roma) Appenninigenae quae proxima
Thybridis undis.
Scholars have ignored or undervalued Servius' note to their cost. Some,
failing to see the antithesis in the line, have supposed that caput is in
apposition with domus, and have consequently been faced with the unreal
problem of making sense of domus exit. So Heyne tried to extract the
meaning 'here will come into being my great home, the capital of towering
cities', and Warde Fowler (CR 30 (1916), 219 ff.; ASR 37 ff.): 'here my
great home (the whole Tiber basin), which is the source (of life and power)
to lofty cities, flows to the sea'. Bentley's reputation can hardly profit
from his suggestion Tuscis caput amnibus, which is as violent as it is un-
necessary. Havet, RPh 35 (19n), 5 ff., favouring the apposition of domus
and caput as the interpretatio difficilior, and changing exit to escit = erit
(already proposed by Faber on Luer. 2.n45, but an improbable archaism
for Virgil), took the line to mean that the temple of the Tiber-god (at
Ostia) will be the centre of a confederation (urbibus; but celsis is unadapt-
able). Mackail, CR 29 (1915), 227, finding it impossible to make sense of
the verse as it stands, applied his favourite remedy: 'rhythm, no less than
sense, calls for another line to complete the period'.
The MS variant certa for magna derives from a recollection of line 39.
65 hie mihi magna domus: hie may have a quite specific reference if
 COMMENTARY

(as suggested by W. H. Semple, CR 50 (1936), n2) Virgil is thinking of the

'Halls of Tiber' mentioned by Ovid Fasti 4.329 f. fluminis ad ftexum
veniunt (Tiberina priores / atria dixerunt), unde sinister abit. This was at
the first bend of any significance which would be met by vessels coming
upstream from the sea (it was almost a hairpin bend and some distance
beyond Ostia before the Tiber changed its course in the floods of 1557: see
the map in Tilly 6). Aeneas' ships had already gone some way up the river
(A. 7.36), and the site of the first camp (A. 7.157 ff.) may possibly have
been at this bend (but not for the reasons given by Carcopino 523 ff.;
T. Frank locates it at Ostia, AJP 45 (1924), 64 f.; see further R. Meiggs,
Roman Ostia (Oxford, 1960), 483 ff.).
There is then a strong antithesis between 'my home here at this
particular spot' and the 'source' far away (not far from the coast on the
other side of the peninsula), which is preferable to taking hie as vaguely =
'this land of Hesperia'.
Ovid (Amores 3.6.89 ff.) remembered the passage to curse the unTiber-
like torrent which barred his way to his mistress: quid si legitimum flueres,
si no bile flu men, / si tibi per terras maxima f ama for et? / nomen ha bes nullum,
rivis collecte caducis, / nee tibi sunt fontes nee tibi certa domus.
Ovid too (Met. 1.568 ff.) locates the domus of the Peneus in the wooded
ravine of Tempe-not so far from its mouth: 574 ff. haec domus, haec
sedes, haec sunt penetralia magni / amnis; in his, residens facto de cautibus
antro, / undis iura dabat nymphisque colentibus undas; contrast Statius
Th. 4.831 ff. cited on 74 f.
65 At A. 10.198 ff. Virgil links Mantua, the nearest large town to his
birthplace Andes, and the Tiber in a typical piece of invented 'heroic'
genealogy: Ille etiam patriis agmen ciet Ocnus ab oris, / fatidicae Mantus et
Tusci filius amnis, / qui muros matrisque dedit tibi, Mantua, nomen. It
must be remembered that Etruscan power not only extended on both sides
of the Apennines (cf. Livy 5.33.9 in utrumque mare vergentes incolttere
urbibus duodenis terras), but was also paramount in Cisalpine Gaul before
the Etruscans were driven out by the Gauls (cf. Polybius 2.17): many
traces of them still remained there, especially at Mantua, of which Virgil
himself says (A. 10.203) ipsa caput populis, Tusco de sanguine vires;
(cf. Pliny N.H. 3.(19).130 Mantua Tuscorum trans Padum sola reliqua).
Mary L. Gordon, 'The Family of Virgil', (]RS 24 (1934), 1 ff.), arguing
from the probable etymology and derivation of the poet's names and their
recorded occurrences over the Italian peninsula-Vergilitts is considerably
more common in Etruria than elsewhere-has made a good case for
believing that Virgil was actually of Etruscan origin. On a question of this
kind certainty is of course impossible (Etruscan origin is accepted tenta-
tively by, for example, Chilver, Cisalpine Gaul (Oxford, 1941), 215 f.,
denied by, for example, Tenney Frank, Vergil (Oxford, 1922), 4 f.). There
can however be no question of Virgil's keen interest in Etruscan history
and culture (cf. J. W. Mackail, 'Virgil and Roman Studies', ]RS 3 (1913),
5 ff.), or of his appreciation of the wide area of their influence in northern
Italy (A. 10.198 ff.), or even of their transmission of Greek culture (cf. A.
 COMMENTARY 39
10.719 f. venerat antiquis Corythi (Cortona?) de finibus Acron, / Graius
homo.). He quite probably knew that it was the Etruscans who had trans-
mitted what was, from his point of view, the most important legacy of
Greek culture, the Aeneas-legend itself. This was established in Etruria by
the end of the sixth century B.c.: cf. especially the terracotta statuette of
Aeneas carrying Anchises, from Veii (Scullard, Plate II7), and cf. Alfoldi
278 ff.

66-80 The river-god disappears into the depths of the stream. Dawn breaks,
Aeneas awakes and prays to Tiberinus, promising sacrifice and asking for
confirmation of the vision. He then equips two biremes and arms his com-
panions for the journey.
66 dixit: for the initial spondaic word and pause, see the note on 71.
deinde: scanned as a trochee here and at 481 by synizesis (deinde >
deinde); see the notes on 292 Eurystheo and 372 aureo.
lacu: see the note on 74.
67 nox Aenean somnusque reliquit: Heyne compares It. 2.71 lµ.J SJ
jlAVKVS' V71"VOS' avijK£V, and Conington Ap. Rh. 3.632 TTJV s· V11"VOS' • •• µ.£8tTJK€V.
In these expressions the sleeper is thought of as passively in the grip of an
external force: •yrrvos- and Somnus are often personified as supernatural
powers.
68 f. aetherii spectans orientia solis / lumina: Aeneas faces east
where the sun is just rising, a usual Greco-Roman practice in making
prayers (at A. 12.172 ff. Aeneas, Latinus and their company face the rising
sun, which is the first deity Aeneas invokes in prayer, at the making of the
treaty, and the Argonauts are ordered Phoebi surgentis ad orbem / ferre
manus in prayer when being purified from the plague in Valerius Flaccus
3.437 f.; and cf. Soph. O.C. 477 (xp~) xoas- xlaa8at a-rav-ra 7Tp0S' rrpW77JV lw,
because the rising sun was regarded as an omen of success.)
69 f. Before praying, Aeneas takes up the river water(' live' and flowing)
in the palms of his hands. It was of course common practice before prayer
and/or sacrifice to wash, or at least to touch water, and so to purify oneself
symbolically, merely as a precaution against the possibility of being
'unclean' (cf. esp. Hesiod W.D. 724 ff.). Aeneas has already formed the
intention of journeying by river; washing the hands and praying to the
river were a preliminary obligation (Hesiod W.D. 737-41). In a wider
context, it was usual practice for travellers arriving in a strange land to
try and win over the gods of the locality by prayer and sacrifice; arrival
by boat at a river estuary naturally directed the traveller's attention to the
river(-god), as here and at Ap. Rh. 2.1271 ff. (Jason after anchoring in
the Phasis).
The situation here is made more complicated by following a dream. Sleep
itself was thought to cause pollution, dreams even more; and as dreams
usually necessitated prayer on waking, the sequence dream: purification
by water: prayer was very common. (One or more links in the chain find
illustration at Homer Od. 12.335 ff.; Aesch. Pers. 200 ff.; Soph. El. 424 f.;
 COMMENTARY

Ar. Ra. 1338 ff.; Plautus Aul. 612; Virgil A. 4.635 ff.; Tibullus 2.1.14 ff.;
Propertius 3.10.13; Persius 2.15 f.)
The particular import of a dream and its psychological effect on the
dreamer are irrelevant to the sequence of events; whether the dream bodes
good or ill, the purpose of the water is to wash away the contact, conceived
as physical, with the supernatural, all dreams and visions being uncanny
and taboo.
The situation here, where Aeneas dreams of, purifies himself with water
from, and then prays to the Tiber, is full of coincidence (which has misled
some interpreters, see below), but its basic ideas are the same as those at
the beginning of A. 9, where after a day-vision of Iris, Turnus purifies
himself with water from a river (unnamed), and then prays to the gods
(unspecified) (A. 9.22 ff.).
Other views which have been taken of the passage are quite un-
convincing. The taking up of the water is certainly not an act of hygiene
(in any case the Romans did not habitually wash after rising, see Carcopino,
Daily Life in Ancient Rome (1956), 159). It is most unlikely that we have
here an unparalleled example of water as the material of sacrifice (so
Warde Fowler, ASR 42 f.); the water which was offered to the spring of
the Camenae (Varro cited by Servius auctus on E. 7.21) and of Juturna
(Servius auctus on A. 12.139) was not a sacrifice in the usual sense but a
survival of an act of homoeopathic magic. The view that Aeneas' action is
simply an instance of a racial habit, very evident in ritual, to touch what-
ever was in question (R. B. Onians, The Origins of European Thought 2
(1954), 129 and note 7), is superficially plausible, but does not explain the
clearly parallel situation in A. 9.
70 sustinet: the MS variant sustulit derives from A. 9.17 or a similar
passage.
71 nymphae: Virgil was not fond of beginning a verse with a self-
contained spondaic word (for reasons given in the note on nulli 502), and
much less fond of entrenching it in front of a sense-pause; there are only
four or five examples of this latter phenomenon in A. 8: dixit 66 (Tiberinus
has finished his speech with an awe-inspiring declaration of his godhead),
nymphae here (in a solemn invocation), salve 301 (another solemn invo-
cation-to Hercules), stravi 562 (with a very slight pause: Evander's
complete flattening of the battle-line). A very slight pause may also
follow stabant 641.
71 nymphae . .. nymphae: repetition of a word is common in contexts
of heightened feeling (here in a fervent prayer); cf. the pathos in hie tibi
mortis erant metae, domus alta sub Ida, / Lyrnesi domus alta, solo Laurente
sepulcrum (A. 12.546f.). But here there is also an echo of the ritual
repetition of prayer (cf. 84 f. and note).
Repetitions of a word-group, as in the example from A. 12 above (there
is no instance in A. 8), are a self-conscious elegance of poetic style, fostered
by the Alexandrians; there are very few instances indeed in the Iliad (see
Leaf and Bayfield on Il. 22.128).
71 nymphae, ~enus amnibus unde est: i.e. nymphae unde genus est
 COMMENTARY 41

amnibus, 'nymphs from whom the streams take their birth' (unde = e
quibus; amnibus, dative); not' whose birth is of the rivers', although that is
an equally common idea.
71 unde: as in Greek, an adverbial relative, ubi, imde, quo, is often used
instead of an adjectival relative with a preposition; this occurs most
usually in a connection indicating place; unde is used of persons by Virgil
at A. 5.123 and 568, 6.766.
71 unde est: colourless monosyllables rarely end a verse in Virgil as
they do in Lucretius, because they cannot bear the emphasis of that
position in the Virgilian hexameter. But here est is virtually enclitic on
unde, which is itself emphasised by being delayed in word-order.
72 Ennius Ann. 54 V Teque pater Tiberine tuo cum flumine sancto: the
line may have been spoken by Ilia (Vahlen), Aeneas (Norden), or even
Horatius Codes (E. H. Warmington), but it is at least clear from the -que
that the speaker had just named (an)other god(s), as Aeneas does here.
Virgil substitutes:
(1) for Tiberine, the correct ritual form (see the note on 3r), Thybri, the
Greco-Etruscan form (8v(µ,){3pis-, Thebris) which makes its first Latinised
appearance in Virgil's poetry: perhaps he already had in mind the epony-
mous hero Thybris of 330 f., or the Greekness of the preceding line
(nymphae ~ vvµ,,f,ai, genus ~ yivos-), or considerations of metre, or the
general preference of poetry for Thybris because of the pleasant-sounding
Greek vowel v (see Wilkinson II f.);
(2) for pater, genitor: a change probably suggested by genus in the line
before, and here therefore including the notion of parenthood. But pater
(genitor) and mater do not necessarily contain any such notion: a bachelor
could be a pater(Jamilias) and a virgin goddess addressed as mater;
originally the words referred to status, authority and dignity and they
never lost some of this meaning; a full discussion can be found in Fustel de
Coulanges, La Cite Antique Book 2, ch. 8, and his view is supported (against
Frazer) by Warde Fowler, RE 155 ff. Applied to gods, pater occurs in the
earliest prayer-formulae (although curiously not in Servius' citation of
adesto, Tiberine, cum tuis undis on this line): cf. Livy 2.ro.n Tum Cocles
'Tiberine pater' inquit, 'te sancte precor'; 5.52.7 quid de ancilibus vestris,
Mars Gradive tuque, Quirine pater?; 8.9.6 (the formula of Decius' devotio)
'lane, luppiter, Mars pater, Quirine ... '. Similarly, patriarchal dignity, not
parenthood, is implied by pater Aeneas.
72 tuqu(e), o Thybri: Virgil's recasting of Ennius' line (see the
preceding note), entails a pause after the first long syllable of the line:
pauses at this position in the Aeneid are not very common, and are always
light: in A. 8, cf. 236, 364, 480, and perhaps 579.
73 periclis: syncope, the shortening of a word through the loss of a
middle vowel (always e, i or u), is regularly found in the oblique cases of
a few words which would not otherwise be usable in hexameter verse:
dext(e)ra, peric(u)lum, vinc(u)lum, saec(u)lum are all so used in this book.
(The nom. and acc. plurals pericula, vincula, saecula are of course suitable
for dactylic verse and so predominate markedly over their shorter alter-
42 COMMENTARY

natives: see Wetmore's Index ss.vv.) The syncopated forms were taken from
Ennius by the Augustan poets and used too widely for any specifically
archaic flavour to remain.
74 ff. 'At whatever spring the well-pool houses you as you pity our
misfortunes, from whatever piece of ground you flow out in all your
beauty, you shall always be celebrated with my worship and offerings ... '
There was a shrine to the Tiber-god on the insula Tiberina,but the refer-
ence here is clearly to the river's most important source or Jons (cf. caput,
65) (in the Apennines, visible from San Marino). All springs and sources
were sacred and could be worshipped, especially at the festival of the
Fontinalia on October 13th, with garlands of flowers, libations and animal
sacrifice; occasionally shrines were built at or near them.
The best commentary on the situation here is the poetically written
letter of Pliny the Younger (Epp. 8.8) on the source-spring of the Clitumnus
(which rises near Spoleto and flows eventually into the Tiber itself):
(5 ff.) adiacet templum priscum et religiosum. stat Clitumnus ipse amictus
ornatusque praetexta (but the upright stance and scarlet-bordered toga must
have made him look more like a magistrate or priest than a river-god-
perhaps some unconventional local influence was at work); praesens numen
atque etiam f atidicum indicant sortes (strips of some material, each inscribed
with a message, would be shaken in an urn by the priest, and whichever
fell out would be the god's answer to the worshipper). sparsa sunt circa
sacella complura, totidemque di. sua cuique veneratio suum nomen, quibusdam
vero etiam /antes.nam praeter illum quasiparentem ceterorum sunt minores
capite discreti . .. leges multa multorum omnibus columnis, omnibus parieti-
bus inscripta, quibus Jons ille deusque celebratur (records of ex voto dedi-
cations).
74 f. quo . .. cumque . .. fonte . .. quocumque solo: the phraseology is
strongly reminiscent of the ritual quicumque es used in addressing an
unknown god; Aeneas here does know the identity of Tiberinus but not his
Jons et origo; compare, in their contexts, quaecumque A. r.330, quisquis es
A. 4.577, quisquis A. 9.22; these are literary echoes of the formula which
showed the very real anxiety of primitive Romans to address the divinity
worshipped correctly: even Jupiter was addressed by the pontifices as
'Iuppiter optime maxime, sive quo alio nomine te appellari volueris' (Servius
on A. 2.351). See further Warde Fowler, RE 120 and C. J. Fordyce,
Catullus 34.21 f. The same scruple and formula is Greek, cf. ZdJs ouns
1r6-r' JunP Aesch. Ag. 160 f., and especially ons Juul Homer Od. 5.445,
Odysseus' prayer to the river-god of Phaeacia, the passage Virgil had in
mind for the miracle of stopping the flow of the stream (see the note on
86-9).
74 f. are' theme and variation', suitable to ritual repetition, and referring
to the source of the Tiber as yet unknown to Aeneas, where the spring
(Jonte) bursts through the ground (solo) from the subterranean pool
(lacus). Madvig's suggestion qui ... locus (Adversaria Critica 2 (1873), 39)
misunderstands lacus, and misses the point. The MS reading is further
supported by the closely parallel passage at Statius Th. 4.831 ff. tuque o
 COMMENTARY 43

cunctis insuete domari / solibus, aeternae largitor corniger undae, / laetus

eas, quacumque domo gelida ora resolvis / immortale tumens, and at Ausonius
Masella 469 ff. Corniger externas celebrande Masella per oras, / nee solis
celebrande locis, ubi Jonte supremo/ exeris auratum taurinae frontis honorem.
(lacus might be plural, but in view of lacu in 66, the singular is preferable
here too, and the variant tenent in 75 probably arose from a simple ditto-
graphy ten-en-t.)
74 lacus is here correctly explained by Servius as the subterranean
pool from which a spring of water rises; by extension it means simply
'pool', in 66 one which is part of a larger stretch of water. Originally the
word meant any deep cavity or trough, and so came to be applied to wine
and oil vats, and water-troughs (as at A. 8.451).
77 Hesperidum fluvius regnator aquarum: Virgil seems to have
had Ennius in mind, Ann. 67 V Postquam consistit (? constituit sese)
fluvius qui est omnibus princeps, which has plausibly been referred to the
Tiber causing its floods to subside at Jupiter's command in order to
deposit the twins Romulus and Remus on dry land (comparable with the
situation here, where the Tiber checks its own down-current). This frag-
ment is cited by Fronto Orat. 160 Naber = 153 van den Hout, in connec-
tion with what may be a saying of Marcus Aurelius 'Tiber amnis et
dominus et fluentium circa regnator undarum' (cf. CIL 6.r.773, a dedication
at Rome of Diocletian and Maximian, after cleansing the water-channels,
Tiberino patri aquarum omnium). Virgil is echoing national pride in the
river of the ruling city: elsewhere, thinking primarily of physical size, he
calls the Po fluviorum rex Eridanus (G. r.482), Eridanus, quo non alius per
pinguia culta / in mare purpureum violentior effluit amnis (G. 4.372 f.)
(note the echo of pinguia culta, referred to the Tiber, at A. 8.63).
77 corniger: rivers likened to bulls are a commonplace of Greek and
Latin literature, cf., for example, Eur. I on 1261 cI, -ravp6µ,opcf,ov 5µ,µ,a K'Y}cptaoii
1ra-rp6,, and the citations in 74 f. above. On Horace Odes 4.14.25 sic
tauriformis volvitur A ufidus Porphyrion remarks: omnium fluminum genii
taurino vultu etiam cum cornibus pinguntur propter impetus et fremitus
ipsarum aquarum (cf. G. 4.371 f. et gemina auratus taurino cornua vultu /
Eridanus). This is also Servius' first explanation in his note here, and no
doubt represents the basic idea of the image: at Il. 21.237 the Xanthus is
described as p,Ep,vKw, ¥-rE -raiipo,. The Achelous is conceived as horned at
Ovid Amores 3.6.35 f. cornua si tua nunc ubi sint, Acheloe, requiram, /
Herculis irata fracta querere manu, and the Inachus so represented at
Statius Th. 2.217 f. pater ipse bicornis / in laevum prona nixus sedet Inachus
urna where the images of the ancestors of the royal house of Argos, with
Inachus first, are carried in procession.
The image was extended to liken the channels or delta of a river to
the horns of a bull: this is the point of Rhenusque bicornis at 727, and of
Ovid Met. 9.774 septem digestum in cornua Nilum. And from, for example,
Valerius Flaccus 8.185 ff. haud procul hinc ingens Scythici ruit exitus
H istri, /fundere non uno tantum quem flumina cornu / accipimus it is clear how
this interpretation led to the picturing in art of river-gods holding a horn.
5 EAC
44 COMMENTARY

77 fluvius: Latin has a separate form for the vocative (i.e. different
from the nominative) only for the singular of second declension nouns in
-us and -ius; but this form, common and standardised in the classical
language, seems to have come into existence late and to have had a rather
precarious life afterwards. The nominative itself, whose basic function is to
concentrate attention on the temporary centre of interest, can be (and in
the other declensions is) quite appropriately used to convey exclamations,
often in the form of an address to someone or something. Moreover some
second declension nouns in -us either never acquired a vocative (like deus),
or only acquired one late (the form popule does not occur before Quintilian
Deel. 302 (Ritter) p. 192, 3), or could not acquire one which sounded
euphonic (e.g. fluvius). The tendency of grammarians from Servius (on
this passage) to Page, with the standardised vocative in -e and -i as their
model, has been to regard forms in -us which are exclamatory as 'archaic
vocatives'; it is better to regard them as nominatives of exclamation,
becoming rare and old-fashioned by the late Republic. (The classic
example of an' archaic vocative', in Livy's admittedly archaising formula
(1.24.7) audi, inquit, Iuppiter, audi pater patrate populi Albani, audi tu,
populus Albanus, is now explained as an appositional nominative; see
LHSz 2, 74).
78 'Only be near and confirm your divine will closer at hand.' Virgil
was alive to the wide area of meaning covered by numen, from its primary
etymological sense (from nuo, 'nod'), meaning 'nod (of assent), decision,
declaration of will', to its secondary concrete sense (perhaps influenced by
late Greek TTvEvµ,a, 'spirit, superhuman being') meaning 'a deity (invested
with authority)'; here and at, for example, 574 the primary sense prevails.
numina firmes here means virtually the same as omina firma at A. 2 .691 :
in both passages an apparently significant omen or portent has occurred
(the flame on Ascanius' head in A. 2) spontaneously: in the technical
language of augury an augurium or auspicium oblativum. A further sign is
then requested (augurium impetrativum) to confirm the first, if the witness
regards what has 'cropped up' as significant-he is at liberty to ignore it
if he wishes (Servius on A. 12.260). The same meaning lies behind the
typically varied phraseology of A. 10.254 f. tu rite propinques / augurium.
The genuineness of the Tiber-vision is immediately confirmed by the sow,
the predicted portent (mirabile monstrum 81).
prop-ius and propr-ius are frequently confused in the MSS, as here; the
context usually makes the choice easy (cf. 280, 556); proprius here,
although recognised by Servius, would be strained in the sense ' as our
own individual protector' compared with the natural sense of propius.
As at A. 1.526 the concept of nearness in propius is passing into one of
participation and help: see N-H onpraesens, Hor. Odes 1.35.2; as Wolfflin
points out, ALL 12 (1902), 421, propior was made to do duty for the non-
existent comparative of propitius, on the analogy of ferus / ferocior, fidus /
fidelior.
78 adsis ... firmes: note the 'framed' line (see Appendix, p. 200).
The number of instances of two finite verbs so 'framing' a static line whose
 COMMENTARY 45

sense-unit is complete in itself is remarkable, as also is the number of

times when such a line concludes a speech (as here) or a section of the
narrative (as at 165, 218, 250, 491, 507, 559, 591, 607). This may be as
important a guide to paragraphing as homodyned fourth feet (see
Appendix, p. 197).
79 biremis: ships with two (or three) banks of oars were not known
to the Homeric Age of Aeneas and his contemporaries. This is one of many
material details which, strictly speaking, are anachronistic. Virgil's
intention, it seems, is to minimise unfamiliar material details which might
hamper his reader's sense of involvement in the contemporary significance
of the Aeneid. See F. H. Sandbach PVS 5 (1965-6), 26 ff.
Bo = A. 3.471 remigium supplet, socios simul instruit armis.

81-101 Suddenly the white sow is sighted on the river-bank, and Aeneas
sacrifices her and her litter to Juno. The Tiber-god stays the current of the
stream, and Aeneas' ships glide effortlessly upstream along the wooded banks
until, just after mid-day, they reach Evander's scattered settlement and put in
to shore.
81-3 Notice how the excitement and suspense build up until the sense
is completed by the main statement conspicitur sus right at the end. See
the note on 43 for the monosyllabic ending.
82 candida per silvam: 'gleaming white through the wood'; the
distinction between candidus 'shining white' and albus, 'mat white' (so
Servius on G. 3.82), is not excluded here by concolor.
82 silvam: nemus 92; arboribus . .. silvas 96; luco 104, 125: Virgil
envisages the Tiber banks (and the whole of Latium, see Rehm 70) as
covered with dense primeval forest; much of it had been cleared for
pasture-land when Pliny had his Laurentian villa (Epp. 2.17.3), and very
little remains today.
82 concolor: the majority of neologisms in Virgil are compound
adjectives (for nouns see on 89 luctamen, for verbs 404 indubitare and
677 effulgere), and these are not conspicuously numerous. Plautine comedy
and the mime (Pomponius, Laberius, Novius) had revelled in inventing
new compounds, often designed for comic effect; Pacuvius and Accius in
their translations of Greek tragedy had tried to emulate also the natural
Greek facility for compounds. But it was soon evident that Latin could
not be treated in the same way with serious purpose, and Horace's taste
disliked the resultant bombast and Schlangenworter (sesquipedalia verba
A.P. 97). The neoterics, notably Catullus in his Attis, also bent on natural-
ising Greek forms, continued the practice, and lent themselves to parody
(Petronius Sat. 55 (ciconia) pietaticultrix gracilipes crocalistria). Horace
A.P. 48 ff. legislated generously for neologisms: they are traditional and
natural, he says, and will win acceptance if used sparingly and based on a
Greek model; but his own poetic practice in this respect is extremely
conservative and unadventurous, and even if he momentarily forgot
(A.P. 53 ff.) the difference in genre between Plautus and Caecilius on the
5-2
 COMMENTARY

one hand and Virgil on the other, Virgil did not, and avoided ostentatious
novelties as carefully as Caesar.
In view of the very fragmentary nature of much pre-Virgilian literature,
it is very hazardous to say positively that Virgil actually coined any
specific word: Servius, with more material at his disposal, made an error
in so doing at A. 3.221 caprigenum.
In A. 8 compound adjectives fall into the following categories (those
found only once in Eclogues, Georgics and Aeneid are marked with an
asterisk):
(I) suffixes:
(1) -ger: corniger* 77, already in Cicero's poetry (Nat. Deorum 2.no)
and Lucretius (2.368; 3.751); laniger 664, cited from Accius by Cicero
Div. r.44.
(2) -fer: horrifer* 435, cited by Cicero from Pacuvius (Or. 155) and
Accius (T.D. 1.68), in his own Aratea (Nat. Deorum 2.n1) and in Lucretius,
3.1012; sagittifer* 725 (Catullus in a lyric, n.6). These compounds in -fer
were quite in the spirit of older poetry; hence, first attested in Virgil,
pacifer* n6, fumifer 255 (and A. 9.522 only), fatifer 621 (and A. 9.631
only).
(3) -dicus: fatidicus 340 (and twice elsewhere in A.).
(4) -ficus: terrificus 431 (and twice elsewhere in A.), already in Lucretius,
2.632; 5.1315; hence vulnificus* 446, perhaps not before Virgil.
(5) -gena: Troiugena n7, Graiugena 127, omnigenus* 698 (anomalous,
see the note there): all these and many other formations of the same type
occur in Lucretius: hence Virgil's analogous coinage nubigena 293 (and
A. 7.674 only}, and indigena 314 (and A. 12.823 only).
(6) -potens: ignipotens 414 and often, unattested before Virgil but
perhaps taken from Ennius or other archaic poetry (omnipotens 334
certainly was).
(7) -aevus: longaevus 498 is a favourite word of Virgil's (it occurs 14
times in A.) but is not traceable before him. As Norden 177 points out, at
A. 8.498 it is in a very Ennian context, and so was probably already to be
found in archaic poetry (grandaevus certainly was).
(8) -color: decolor* 326 was perhaps coined by Cicero, T.D. 2.20 de-
colorem sanguinem, to translate Soph. Trach. 1055 x>..wpov alµa. Virgil
himself seems to have coined bicolor 276 (and A. 5.566 only), and concolor*
82 (which became a great favourite with Ovid).
(9) -abilis: not uncommon in Ciceronian prose and earlier poetry: cf. the
'labyrinth' group inobservabilis error Catullus 64.n5, inremeabilis error
A. 5.591, inextricabilis error A. 6.27; hence, apparently first used by Virgil,
ineluctabilis 334 (and A. 2.324 only), and enarrabilis* 625.
(10) -ilis: indocilis* 321, already in Cicero's philosophical prose (Nat.
Deorum r.12; Ac. 2.2), promoted by Virgil to epic.
(II) prefixes (multipliers are common in earlier poetry):
(n) semi-: semi/er 267 (and A. 10.212 only), already in Cic. Aratea 59
and often in Lucretius; hence, for the first time in epic, semesus 297 (and
A. 3.244 only) and semihomo* 194.
 COMMENTARY 47

(12) bi-: some of these compounds were taken over from earlier poetry,
e.g. bipatens from Ennius (Servius on A. 10.5); and even if they are not
attested before Virgil it was an easy extension to coin bicolor 276 (and
A. 5.566 only), and bicornis 727 (and G. 1.264 only). biremis 79 (and
A. 1.182 only) and bimembris* 293 (see the note there) seem to have been
promoted to epic by Virgil.
(13) ter-: tergeminus 202 derives from earlier poetry, cf. especially
Luer. 5.28 tripectora tergemini vis Geryonai.
(14) quadri-: quadripedans 596 (and A. n.614 only) occurs already in
Ennius' tragic diction (Sc. 184 V), and probably also at Accius 603 R 2 •
(15) e(x)-: egelidus* 610 and exsors 552 are both special cases of
adjectives with more than one meaning: see the notes ad loc. See (9)
above for enarrabilis.
(16) prae-: praefulgens* 553 was first promoted to high poetry by
Virgil, see the note on effitlgere 677.
(17) in- as a negative prefix: many of these forms naturally occur in
prose or verse or both before Virgil: intractatus* 206, insperatus 247,
imperfectus* 428. inausus 205 (and A. 7.308 only) is a formation along the
same lines, which commended itself to Valerius Flaccus, Seneca and
Tacitus. inaccessus 195 (and A. 7.n only) and inexpletus* 559 are more
startling: see the notes ad loc. For ineluctabilis and indocilis see (9) and
(10) above.
Virgil's compound adjectives therefore are homogeneous; those he
invented are formed in a quite accepted manner and so take an incon-
spicuous place among those he inherited; he would evidently have agreed
with Quintilian's verdict (1.5.70) res tota magis Graecos decet, nobis minus
succedit. See further Leumann 150 ff. for a useful survey of composita in
poetic diction.
84 f. Two lines full of the solemn language of ritual: the repetition of
tibi is not purely rhetorical, but represents the repetition of the actual name
Juno in the prayer which accompanied the sacrifice (cf. the repeated
nymphae in 71, and the hymnic repetition of tu, te at 293 ff.).
enim (e+nam) here has its archaic sense of 'indeed, certainly, to be sure';
cf. A. 1.19 f. progeniem sed enim Troiano a sanguine duci / audierat, where
Quintilian (9.3.14) comments: alia commendatio vetustatis, cuius amator
unice V ergilius fuit . .. Quorum similia apud veteres tragicos comicosque
sunt plurima. Illud et in consuetudine remansit 'enimvero '. In Plautus enim
almost always has this meaning; in Terence the explanatory sense 'for'
is gaining the upper hand, and this established itself in classical prose to
the virtual exclusion of the earlier meaning. This may explain why
Servius failed to appreciate the force of enim here (vacat enim et tantum ad
ornatum pertinet).
Ovid also uses the phrase, Met. 15.581 f. 'Rex' ait 'o salve; tibi enim, tibi,
Cipe, tuisque / hie locus et Latiae parebunt cornibus arces ', as does Silius 13.
135 ff.
For other instances of archaic enim in Virgil see Norden on A. 6.28;
R. D. Williams on A. 5.395, and cf. namque A. 10.614.
 COMMENTARY

84 pius: pietas is a very wide-embracing word (see the discussion in

Bailey RV 79 ff.); basically it means a sense of the obligation of right
conduct to gods and men (without the puritan overtones of 'duty'), and
this central meaning is quite clear in contexts of religious function like
sacrifice (as here) or burial (as, for example, A. 6.176). But Virgil's use of
pius is more puzzling in some human contexts: at A. 4.393 pius Aeneas
leaves the impassioned Dido, not without compassion, but in answer to
'Stern daughter of the voice of God! 0 Duty', and Virgil's contemporaries
would not have been surprised to find pietas applied to loyalty to one's
sense of mission and self-fulfilment in the face of personal tragedy; but
compassion and mercy are ruled out at A. 10.591 (Aeneas has just killed
Lucagus) quem pius Aeneas dictis adfatur amaris: Aeneas may be thinking
of vengeance for Pallas, and pietas might conceivably be extended to
cover personal vendetta, but what of reviling the dead? There is a curious
combination of sober Roman virtue and typically Greek' heroic' behaviour
(cf. ll. 16.742 ff. esp. TOV s· £1TtKEpTOJ.LEWV).
84 f. tibi, maxima Iuno, / mactat: this sudden apostrophe to the god
seems much more artificial than the usual invocation by the worshipper
(as at 71 f., 293 f. and cf. A. 7.389 f.); but it occurs elsewhere, always in
connection with a word of sacrificing, and is obviously a formula: cf.
A. 3.118 f. sic fatus meritos aris mactavit honores, / taurum Neptuno,
taurum tibi, pulcher Apollo; A. 6.250 f. Aeneas matri Eumenidum magnaeque
sorori / ense ferit (sc. agnam), sterilemque tibi, Proserpina, vaccam; Ovid
Met. 4.755 f. mactatur vacca Minervae, / alipedi vitulus, taurus tibi, summe
deorum. Very similar is A. 6.18 f. tibi Phoebe sacravit / remigium alarum,
which Norden (ad lac.) derives from the style of Greek ex voto dedications
like A .P. 5.9 aot • .• <l>o'i{JE Swpa Ta.DE KPEJ.LaTm; if this is so, the natural
invocation by the dedicator is echoed in an apostrophe by the poet in the
context of the narrative, and this will hold good for the sacrificial formulae
above. The metrical convenience of the vocative, and the desire for variatio
in expression no doubt played some part in some examples but should not
be pressed. For apostrophe in general and its degeneration into a mannerism
see the note on 643 Albane.
This kind of apostrophe is foreign to Homer: Aulus Gellius N.A.
13.27.3, comparing Homer Il. 11.728 Tafipov s· 'A>.qm{jJ, Tafipov SJ IIoaEL-
Sawvi with A. 3.119 (above), regarded Homer's verse as simplicior et
sincerior, Virgil's as vEwTEptKWTEpos et quodam quasi ferumine inmisso
fucatior, 'rather daubed with a kind of added paste' (the word tibi).
A white sow would turn out to be an entirely suitable sacrifice to Juno:
the quotations above show the general tendency (but not fixed rule) to
suit the sex of the victim to that of the divinity, and white or shining
victims were chosen for gods of the upper world whom one wished to
encourage, black for those of the underworld whom one wished to dis-
courage. A sow was the standard sacrifice, in certain rituals, to Ceres and
to Maia, but Virgil is not suggesting that Juno was, or should be, assimi-
lated to any other deity.
85 mactat . .. sistit: Page and Conington follow Servius in regarding
 COMMENTARY 49

this as a case of vcrr£pov 1rpo-r£pov (he must have stationed the sow at the
altar before he slaughtered her), but this is not necessarily so, for although
mactare came regularly to mean 'to slay a victim', Virgil may well be
restoring its older sense here, meaning approximately 'to hallow, conse-
crate' (see Warde Fowler, ASR ad loc.). For the controversy as to whether
macto derives from the unattested macio, 'I sprinkle' (L. R. Palmer), or
the equally unattested mago, 'I increase' (0. Skutsch and H.J. Rose) see
CQ 32 (1938), 57 f. and 220 ff.; 35 (r94r), 52 ff.; 36 (1942), rs ff.
86-9 Suggested in substance by Od. 5.451 ff., but with a difference in
details. Odysseus, after taking the initiative in praying to the unknown
river-god of the land of the Phaeacians (445 K.\vB,, a.vat, o-ri<; iaal), has his
prayer granted immediately and is rescued into the river-mouth, 451 ff.:
·ns ¢,6.8' · o S' a?rrlKa 1rava£v €0V p6ov, fox€ St Kvµa, / 1rp6a8£ s,
, 1ro-raµov- 1rpoxoa<;.
o,1TOl'Y}a€
yal\'Y}V'Y}V, Tov o £aawa£v I £<;
\ I \ ~' , I I

86 ea . .. quam lon~a est nocte may be an attempt to render the Greek

compound adjective 1ravvvxws. Indeed the idea of Tiberinus giving noc-
turnal assistance may possibly owe something to Athene's assistance
(Od. 2.420 f.), which sped Telemachus' voyage during the night (434,
1ravvvxt'Y/ µiv p' ij Y£ Kat ~w 7T€tp€ KEArnBov).
The phrase = 'that whole livelong night'; cf. A. 4.193 nunc hiemem
inter se luxu, quam longa, fovere and Pease's note there for other instances
of this poetic idiom.
86 tumentem hints at the constant danger of flooding, cf. Horace Odes
r.2.13 ff. with N-H ad Zoe. and Tacitus Ann. r.76 and 79 (quoted in the
note on 31). During Virgil's lifetime five floodings of the Tiber are recorded
(see Le Gall, Le Tibre (Paris, 1953), 29). They were naturally interpreted
as portents and connected with the most striking event of the time; so
when the Tiber flooded on the very night after Octavian received the title
Augustus (r6 January, 27 B.c.), it was taken to portend his future greatness
(Dio Cassius 53.20).
87 leniit: a light or heavy pause after a dactylic word in the first foot
of a hexameter is frequent in Virgil (occurring in about six per cent of his
lines, according to Winbolt 14 f.). It can have an obviously descriptive
value: the initial stress and falling rhythm of the dactyl brought to a halt
suggest forceful energy spending itself rapidly, and so this rhythm seems
naturally to carry with it a feeling of speed, suddenness or decisiveness.
This is particularly true when the first dactylic word is a third person
singular verb, either historic present or, more often, a perfect of instan-
taneous action: with leniit here, compare extulit 2, mugiit 218, impulit 239
(a good example), advocat 250, evomit 253, excitat 543, spargitur 695.
87 refluens: a wonder; rivers flowing backwards as a sign thatthe natural
order of things is in chaos were a venerable literary trope, avw 1r0Taµwv
i£pwv xwpovai 1rayal (Eur.Medea 4ro); Turnus uses it for rhetorical sarcasm
at A. rr.405 amnis et Hadriacas retro fugit Aufidus undas. The trope often
appears as an 'impossibility' (aSvva-rov), e.g. Ovid Her. 5.29 f. cum Paris
Oenone poterit spirare relicta, / ad fontem Xanthi versa recurret aqua. The
actual occurrence was of course attributed to some supernatural inter-
50 COMMENTARY

ference, usually witchcraft (A. 4.489; Tibullus 1.2.43 f.; Lucan 6.472 ff.).
88 mitis with stagni, the 'theme' of which placidae paludis is the varia-
tion. For ut postponed to second place in both this line and the next, see
the note on 22 ubi.
89 sterneret aequor aquis: '(so as to) smooth the surface of the
waters'; cf. A. 5.821 sternitur aequor aquis (by Neptune stilling the waves).
aequor basically means a flat, level surface, and is often used alone to mean
the sea (cf. Varro L.L. 7.23 aequor mare appellatum, quod aequatum, cum
commotum vento non est), or the land (cf. G. 1.50, G. 1.97, A. 5.456). The
addition of aquis in the above phrases suggests that aequor is being used
in its basic sense.
89 aquis itself is another dative/ablative ambiguity; it could be a
possessive dative, or a local ablative, and there are other possibilities;
Virgil is exploiting this ambiguity, giving the word virtually the force of an
adjective, 'the watery floor' (see Mackail, App. A, p. 513).
89 luctamen: Virgil ventures two new noun coinages in this book,
luctamen here and (with related suffix) the plural temptamenta in 144. Noun
formations in -men had Ennian authority and Lucretian precedent;
Virgil's formations here are from first conjugation verbs (contrast Ennius'
momine Ann. 595 V and Lucretius' molimine 4.902). Ovid is much freer
than Virgil in coinages of this type: alongside temptamenta (Met. 7.728,
15.629) appears temptamina (Met. 3.341, 7.734); luctamen did not have the
same vogue, but was taken up by Valerius Flaccus (6.510), Claudian
(B.Goth. 138), and Ausonius (Masella 34). More novelties of the same type
were later devised by Apuleius, for example, Met. 11.9.2 mulieres . .. vario
laetantes gestamine, verno florentes coronamine. For a' sprachgeschichtliche
Untersuchung' of the whole topic, see Norden's note, Ennius und Vergilius
27 ff.
90 iter ... celerant: celerare is always transitive in Virgil, adcelerare
intransitive. The variant peragunt derives from a misrecollection or
marginal note of A. 6.384.
90 rumore secundo: on balance it is best to punctuate heavily after
these words. They mean 'with shouts of applause', as at Ennius Ann.
255 V, Cicero Div. 1.29, Horace Epp. 1.10.9, and Sueius (a contemporary
of Catullus) redeunt referunt rumore petita secundo (the source of Virgil's
phrase here, according to Macrobius 6.1.37). Cf. also the similar fremituque
secundo A. 5.338, and clamore secundo A. 10.266. The shouts came either
from the crews themselves (Cerda, Conington, Sabbadini), delighted at the
ease of their task and encouraging each other, or more probably from
their comrades on the bank cheering them off (Dryden, Mackail; and cf.
Ap. Rh. 1 .556: when the Argo set out, Chiron v6«nov £7TEV<p~µ,T)aEv dKrJSla
viaaoµ,lvo,ow). Henry objects that an exploring party in unknown terri-
tory would not wish to attract attention by making any kind of noise, but
Virgil's view is not likely to have been so ruthlessly practical. The very
similar A. 6.384 may further suggest that here also we have a whole-line
sense-unit.
Tiberius Donatus (followed by Henry) took rumore to mean the ripple
 COMMENTARY 51

of the water. This meaning is accepted by Heyne and Hirtzel who connect
r. secundo with what follows. But this sense is difficult to parallel except by
Ausonius M osella 21 f. amoena fluenta / subterlabentis tacito rumore M osellae:
clearly a poetic extension of the basic use, perhaps inspired by an individual
interpretation of Virgil's phrase here.
The original reading of the codex M ediceits was Rumone, which Servius
on A. 8.63 explains as an ancient name of the Tiber quasi ripas ruminans et
exedens; so here it = favente fluvio. The name has an authentic ring about
it, and its rarity would expose it to corruption at an early stage; but that
same rarity suggests that it was a late antiquarian fiction. The sense of the
phrase = amne secundo would fit well with refluens 87, and is not contra-
dicted by adversum 58 (the usual state of the river) or f atigant 94.
91 This line owes its first two words and their rhythms, but nothing
more, to a reminiscence of Ennius Ann. 386 V Labitur uncta carina, volat
super impetus undas, and Ann. 478 V Labitur uncta carina per aequora cana
celocis. In both lines Ennius chooses the obvious device to convey speed, a
purely dactylic sequence (the exact opposite of his most frequent pattern
in the Annales, with spondees in all feet except the fifth), and with complete
coincidence of ictus and accent (except for volat). Virgil's rhythmical
pattern is very much more subtle: a dactylic sweep is maintained until the
strong caesura in the fourth foot, but whereas labitur uncta (three and two
syllables respectively) has a falling rhythm with coincidence of ictus and
accent, vadis abies (two and three syllables respectively) has a rising
rhythm with clash of ictus and accent; at this point the swift movement is
arrested by the slow wonderment of the waves, and the completion of the
fourth foot as a spondee; and at the same time the preceding cla;;h of
stresses is resolved (as usual) in the last two feet of the line, only to be
resumed at the beginning of the next. The two lines are bound together by
anaphora of mirantur, miratur (of which Virgil is extremely fond, cf. A.
1.420 f.; A. 1.709; A. 8.161), so placed that the ictus falls on exactly
different parts of the word.
uncta = 'smeared', i.e. 'well-tarred'.
mirantur et undae: note the postponement of the normal connective
et to second place in its clause. This is not so common as one might expect:
there is only one example in A. 6 (840), three in A. 4 (124, 418, 513); in
A. 8 there are two more besides this one (329, 517). nee (neque) and sed
also occur in second place in Virgil (although not in A. 8; see Norden 404),
as they had before him in Catullus who provides the earliest evidence for
the postponement of connectives: hence the strong supposition that this
is a neoteric mannerism, inherited from Hellenistic poetry (see Austin on
A. 4.33); et too was probably influenced by postponed Kal, although there
seems curiously to be no example in Catullus (LHSz 2, 484); see Fordyce
on Catullus 23.7.
92 f. 'and the woodland, unused to such a sight, marvels at the heroes'
shields gleaming afar on the river, and the painted hulls floating in it.'
insuetum nemus is another instance of pathetic fallacy, cf. E. 6.40
(cumque) rara per ignaros errent animalia mantes, 'and when (for the first
 52 COMMENTARY

time) animals wander here and there over the hills which do not recognise
them'.
The hulls of the ships were decorated, as usual in the Heroic Age, and
emblazoned shields were hung on their sides while their owners rowed.
For the variation between fulgentia (participle) and innare (infinitive)
see the note on 630 f.
(The MS variant mirantur in line 92 makes no sense and is a mindless
repetition of the same word in line 91.)
93 virum gen. pl. (as also at 197, 312, 315, 500, 539); the -um termi-
nation (Indogermanic -om, Greek -wv) is the original one in the gen. pl. of
the second declension, and survived in some of the formulae of law and
religion like praejectus f abrum and pro deum fidem; the 'expanded' form in
-orum was taken over from the pronoun inflexion, illorum, etc., and
established through the similarity of -arum of the first declension. See
LHSz 1, 278 f.
94 olli = illi (cf. 594; 659 ollis, dat. pl., and olli, dat. sing., A. 5.10, the
only three cases used by Virgil): the root ol- is the same as in olim. olle,
etc., had already disappeared from living speech by the time of Plautus
(LHSz l, 286), and in Lucretius and Virgil is a deliberate archaism.
Quintilian (8.3.24 f.) approved of Virgil's perfect taste in using it and other
archaisms like quianam and pone which, not being quaintly obsolete, gave
a venerable dignity to style.
noctemque diemque fatigant: cf. Propertius 4.11.81 sat tibi sint
noctes, quas de me, Paulle,jatiges; another variety of pathetic fallacy, where
a period of time is personalised and regarded as affected by or sharing in
what happens to someone during it; it reflects a primitive view of time as a
material object, cf. Aesch. Ag. 894 opwaa 1r>..,dw TOV [vvEvl>OVTOS' xp6vov; in
Virgil cf. A. 2.647 f. annos / demoror; A. 5.766 noctemque diemque morantur;
A. IO. 808 exercere diem; bolder extensions are E. 9.52 condere soles and
A. 4.193 hiemem ... fovere.
noctemque diemque: the use of -que . .. -que (strictly 'both ... and',
but the first -que adds little or nothing to the meaning and can be ignored
in translation) to link closely together words closely related in meaning or
reference is a formula modelled on the Greek TE ••• TE (e.g. in Homer It.
5.426 cJis- ¢,a.To, µEL8riaEv Si 1raT~P dv8pwv TE 0Ewv TE). In early Latin literature
it seems to have been reserved for the high style (Plautus confines it to
passages of elevated tone, see Fraenkel, Plautinisches im Plautus 209 f.);
Ennius had promoted it because, like its Greek model, it was very service-
able in hexameter composition, especially in the last two feet (cf., for
example, Ann. 405 V semper abundantes hastas frangitque quatitque).
Virgil continued its use, and latter epic poets followed his example. It is
however completely avoided by classical prose (the sole exception, Cic.
Fin. 1.51 noctesque diesque-the same formulaic commonplace as here-is
probably a reminiscence of Ennius (see Ann. 334 V)); in post-classical
prose its use is severely restricted (see LHSz 2, 515). See further Palmer,
LL n3 f.; Austin, A. 4.83; R. D. Williams, A. 5.92. In A. 8 this calque
couples the last two words of the hexameter at 490, 550, 601, the first two
 COMMENTARY 53

at 291, 294, 312, 425; it occupies some other part of the verse at 94 (here),
277, 433. It should be carefully distinguished from instances where the
first -que is truly connective (joining the whole clause to what precedes),
and the second connects yet a further clause or phrase; cases of this
legitimate use are more frequent than those of the calque, cf., for example,
60 f. Iunoni fer rite preces, iramque minasque ( = et iram minasque) f sup-
plicibus supera votis; 431 sonitumque metumque.
95 f. A good instance of a repeated word-pattern (adj., verb, noun)
whose second occurrence overflows into the next line, helping the mobility
of the verse; see the Appendix, p. 200.
96 viridisque secant placido aequore silvas: 'and cut through the
green woods on the calm surface of the water', i.e. they row between small
islets of clumps of trees, and 'cut through them' just as the Tiber itself
'cuts through' the rich crops, pinguia culta secantem (63). This is the
interpretation of Heyne, Henry and Page. It fits in well with Virgil's
picture of the densely wooded banks (A. 7.29 ff.) of the Tiber, which is
now in flood (86 f.}, isolating groups of trees.
Servius, followed by Mackail and Conington, took the line to refer to
the reflection of the trees in the water. But there is no word in the Latin to
convey the notion of 'reflected' as there is in, for example, Pliny Epp.
8.8.4 ripae fraxino multa, multa populo vestiuntur, quas perspicuus amnis
velut mersas viridi imagine adnumerat, or Petronius 119 Bellum Civile 28
(a highly polished table) citrea mensa greges servorum ostrumque renidens,
or Ausonius Masella 189 ff., 218 f., 222 ff., 225-9.
97 The line recalls Homer's -ryµos o' ¥Aios µlaov ovpavov aµcpif3£/3~K€L
(' was bestriding') (fl. 8.68); its sense simply = post meridiem. Circum-
locutions for the time of day, the season, etc., were traditional apparatus
of the high style of epic and tragedy (cf. A. 8.280, 369,407 f.); they became
a hackneyed ornament (Quintilian 8.6.59 ff.), the delight of poetasters
(Seneca Epp. 122.11 ff.), and a fit object for mockery (Seneca Apocol. 2).
97 ff. Aeneas' arrival and reception at Pallanteum are modelled closely
on Telemachus' visit to Nestor at Pylos (Od. 3.1 ff.), and, to a lesser extent,
his visit to Menelaus at Sparta (Od. 4.1 ff.). Up to now (see the note on
31-96) the Homeric subtexture of the book has implied an equation
between Aeneas and Odysseus; at first sight it is puzzling that Aeneas'
situation should recall so unambiguously that of Odysseus' son Telemachus
in search of his father. Virgil's motive in making the change was no doubt
the necessity of giving plausibility to Aeneas' friendly reception. In the
world of Homeric heroes the sure guarantee of a warm welcome was
already existing family friendship (which could even prevent a duel,
Il. 6.119 ff. and esp. 215 Diomedes and Glaucus): Evander had once been
host to Anchises (154 ff.) and recognises Aeneas from his likeness to his
father, just as Nestor detected a similarity in speech (alone) between
Telemachus and his father (Od. 3.122 ff.), and Helen noticed the physical
resemblance (Od. 4.141 ff.), which Nestor and Menelaus had missed.
The chief points of similarity in the two passages: when Telemachus
arrives Nestor and the Pylians are sacrificing to Poseidon (Od. 3.4 ff.), when
54 COMMENTARY

Aeneas arrives Evander and the Arcadians are sacrificing to Hercules

(102 ff.); Nestor's son Peisistratus greets Telemachus (Od. 3.36 ff.),
Evander's son Pallas welcomes the distinguished Trojan newcomer (121 ff.);
Telemachus explains who he is and what he wants (Od. 3.79 ff.), as does
Aeneas (117-20, 127 ff.); Nestor's first narration (Od. 3.103-200) is compar-
able in length with Evander's story of Cacus (185-275). Nestor reminisces
discursively on the Trojan War and the homecoming of the heroes (Od.
3.102 ff.), Evander reminisces on the past history of the site of Rome
(185-272, 314-61); Nestor refers his guest for further help to Menelaus at
Sparta (Od. 3.317 ff.), Evander refers Aeneas to Tarcho and the Etruscans
(475 ff.). At the end of Nestor's talk with Telemachus there is a sign from
heaven (Od. 3.371 f.), which Nestor answers with a sacrifice: Evander's
talk with Aeneas is also ended by a heavenly sign (523 ff.) also answered by
sacrifice. Finally, Nestor sends his son, the first to have greeted the
stranger, to accompany him on the next stage of his journey: so does
Evander.
There are however significant differences of detail: Nestor's large host
(over 4,500 strong) naturally shows no alarm at the arrival of Telemachus-'
ship (even though Nestor suspects they may be pirates, Od. 3.72 ff.), and
Telemachus and his companion Mentor (the disguised Athene) are
immediately invited to pray and pour libations to Poseidon (Od. 3.40 ff.),
then take part in the sacrificial meal (Od. 3.65 ff.), and only then are asked
their identity and purpose (Od. 3.69 f., Nvv o~ Killi6v iun fl,£Ta.AAfjuai
Kat iplu8m / [£lvov<;, oi nvt<; £luiv, €7T€t TO.P'TT'TJCTav iowofj<;). (This is
the sequence demanded by heroic etiquette, observed even by the
swineherd Eumaeus (Od. 14.80 ff., 187 ff.), ignored only by the brute
Polyphemus (Od. 9.252). Amyclus, on the arrival of the Argonauts (Val. Fl.
4.204 f.), is almost as savage as Cacus, devolat inde furens, nee quo via curve
profecti / nee genus ante rogat, sed tali protonat ira.) But the small company
of the poor Arcadian settlement are terrified by the sudden sight of
Aeneas' reconnaissance party (109 f.), whose members are immediately
asked 'friend or foe?' (112 ff.), and are invited to the feast (172 ff.) only
after inquiries and diplomatic interchanges. Virgil has managed to convey
unobtrusively the fear and insecurity of a comparatively weak settlement
in a hostile country.
Two important consequences for characterisation followed from Virgil's
decision to model Aeneas' arrival at Pallanteum on Telemachus' arrival at
Pylos: Evander became endowed with some of the characteristics of
Nestor (see the note on 560 ff.), and Pallas was modelled after Peisistratus.
97 ff. conscenderat . .. / cum . .. / ... vident: for the cum inversum
see the note on 26 ff.
98 f. muros . .. ac rara domorum / tecta: the 'walls ... and scattered
roofs of houses' are pure imagination on Virgil's part; in fact, however, the
earliest huts of the Palatine settlement were 'scattered' and not organised
into anything like streets or vici, or the village communities of Teutonic
countries (see Einar Gjerstadt 3, 46 f.).
98 procul, 'at a distance', often, but not always, a great distance.
 COMMENTARY 55

prociil: note the scansion. Virgil occasionally lengthens final syllables

which Augustan and later poetry regularly treats as short, but only under
strict conditions relating to both verse and prosody: (r) the syllable
invariably bears the ictus (the stress of the verse-rhythm), and usually
either precedes a main caesura in the third or fourth foot (which in turn
usually coincides with a sense-pause however slight), or occurs in the last
two feet of the line in close association with a Greek-derived word or a
Greek rhythm (spondaic fifth foot). (2) As regards prosody, Virgil often
follows Ennius (and other early poets) in conserving what was originally
a genuinely long syllable, although it is a moot point whether Virgil
actually appreciated this: masculine nouns in -or (-os) almost invariably
have this syllable long in Ennius (sudor Ann. 406 Vis an exception), and
so in Virgil with amor (E. ro.69; A. n.323), dolor (A. 12.422) and others
(note melior G. 4.92). But since all Latin words have their natural stress on
some syllable before the last, some originally long final syllables tended to
become short: the imperfect termination -bat already fluctuates between
long and short in Ennius (hence too in Virgil it is long at A. 5.853, ro.383,
12.772), and the same is true of present terminations in -et (cf. videt
A. r.308), and -it (cf. facit E. 7.23), and the imperfect subjunctive in
-eret (cf. peteret A. r.651). This explanation however will hardly hold good
for the lengthening of the last syllable of populus at Ann. 90 Iamque
expectabat populus atque ora tenebat: here Ennius is lengthening a naturally
short syllable in imitation of similar 'irrational' lengthenings in Homer,
and Virgil ventured even further in this strictly epic tradition (it is note-
worthy that in Cicero's poetry, Lucretius and Catullus short final syllables
lengthened contrary to contemporary usage occur so rarely in the MS
traditions that they have always been called into question and emended).
Noun and verb terminations in -ur and -us form by far the largest part of
his innovations (cf. datur A. 5.284 and Williams' note; pectoribus A. 4.64
and Austin's note), and the consonants -r and -s are easily capable of
being prolonged (if they were double consonants, -rr, -ss they would
necessarily lengthen the preceding syllable); -l too falls in this category
and may account for procul here, if it is a complete innovation and not
precedented in early poetry now lost. caput at A. ro.394 nam tibi, Thymbre,
capiit Euandrius abstulit ensis is an unparalleled innovation suggested
perhaps by the Greek flavour of the proper names.
See further Nettleship's excursus to A. 12 (in his revision of Conington's
edition, vol. 3); Norden 450 ff. For the special cases of subiit 363, and -que
425, see the notes there.
In epic poetry after Virgil the use of this licence is rare and unexperi-
mental compared with his: lengthened final syllables of 3rd sing. verbs
other than eo and its compounds do not occur in Lucan (Housman's note
on 2.565) and are very rare in Statius (Th. r.403 agit; 6.351 erat (erant
Slater)); in Valerius Flaccus I have noted only 8.259 impediit, although
he is fond of lengthening final -us, 2.225, 6.152 and 570, 8.149.
ror ocius: comparative adverbs are often used not with reference to
a specific point of comparison, but in a more independent way = 'more ..
 COMMENTARY

than not', •fairly ... '. So saepius = •more often than not', •fairly often',
and in this way the comparative comes to represent a weakening of the
force of the straightforward positive •often'. Lively colloquial speech took
a great liking to some of these comparative forms: ociter was soon eclipsed
(in Plautus) by ocius, and citius, celerius, tardius, serius are favourite forms
of the sermo cotidianus, with a meaning virtually indistinguishable from
that of the positive. See further LHSz 2, 168 f. ocius has the same force
at 278 and 555; contrast the true comparative adjective at 223 ocior Euro.
urbique propinquant hardly fits well with either time or place: they
do not disembark until later, and there was in any case a grove between
the river-bank and the city. L. Delaruelle, RPhil 36 (1912), 309 f., proposes
ripae for urbi (cf. A. 6.410), but it is more plausible to assume that Virgil
wishes to allude, even prematurely, to Evander's settlement.

102-125 The Arcadians happen to be engaged in sacrifice to Hercules when

they are disturbed by the sudden appearance of strangers. Evander's son Pallas
runs forward to inquire who they are and what they want, and when he learns
from Aeneas' reply that they are Trojans, he welcomes him hospitably and
leads him to his father.
102 forte die ... illo: £iluvµf36>.w,, id est bono omine victori deo (sc.
Herculi Invicto) Troianis venientibus sacra celebrantur, Servius. The annual
festival of Hercules Invictus at the Ara Maxima was probably celebrated
on the same day as at his temple nearby at the Circus Maximus (12 August)
or at the rather more distant Porta Trigemina (13 August). P. Grimal,
REA 53 (1951), 51-61, sees another designed coincidence in the situation:
Augustus' triple triumph in 29 B.C. (see 714 and note) began on 13 August
(Acta Triumphorum CIL 1.2.180). In view of the importance of the link
between Hercules and Augustus as divine saviours (see the Introduction,
p. xxn f.) in this book, Virgil may well be reminding contemporaries of the
association of dates, perhaps even arranged by Augustus to stress the
connection. See also Camps 100.
102 die ... illo: Austin (on A. 4.169) gives a useful summary (based on
E. Fraenkel, •Das Geschlechtvondies ', Glotta8 (1917), 24 ff.) of the fluctua-
tions in gender. The rule of classical prose, that dies masculine denotes a
single day as a definite date, whereas the feminine marks a day as the
expiry of a longer period (exemplified by Caesar E.G. 1.6.4), was contra-
vened by the poets for metrical reasons. Virgil makes dies masculine
wherever metrically possible-in all the oblique cases, and in the nomi-
native with ille; ordinary adjectives in the nominative are invariably
feminine (necessarily so in formulae like orta dies, etc.).
102 sollemnem . .. honorem: the derivation of sollemnis is uncertain
(perhaps soll- = totus, and -emnis from annus), but the phrase here means
the same as sacra . .. annua 172 f.
103 Amphitryoniadae: first in Greek in Hesiod (Theogony 317) (not
in Homer); first in Latin at Catullus 68.rr2 falsiparens A mph., • falsely
called Amphitryon's son'. Hercules' real father was Jupiter, but Alcmene's
 COMMENTARY 57

husband was Amphitruo: hence the patronymic (cf. Alcides 203 and often,
from Alcaeus father of Amphitruo). Only the gen. and abl. of Hercules
itself are suitable for hexameters.
103 divisque: which? It is strange that any other gods should be
involved in the sacrifice to Hercules. The ninetieth section of Plutarch's
Quaestiones Romanae is entitled 'Why is no other god mentioned during a
sacrifice to Hercules, and why is a dog never seen within his enclosures, as
Varro records?' The answer is that this and many other features of the
Hercules cult preserved the memory that he was a foreign (Greek) god, in
whose rites the usual invocation of other (Roman) deities had no part.
On A. 6.830 Servius explains Hercules' cult-title Monoecus (Movo,Kos,
' Lone-dweller'), which gave its name to the principality of Monaco, 'because
in his temple another god is never worshipped at the same time, as
Minerva and Juno are in the temple of Jupiter (sc. on the Capitol), and
Cupid in the temple of Venus'. But if this exclusiveness was a feature of
the organised state worship, it was not the case in the popular religion of
the countryside, which often invoked him with other protectors of home
and farm like Liber and Silvanus (see Wissowa RuK2 281 f.). It is
possible that Virgil is thinking of this type of private cult, and that the
inclusive formula divisque ferebat (cf. divosque precantur 279), which pro-
tected the worshippers against offending any deity by neglect, here refers
to any spirits there might be in the lucus, including Silvanus (cf. 600 f.).
104 ante urbem: in front of Evander's settlement on the Palatine.
Tacitus Ann. 12.24 expressly says that in historical times the Ara Maxima
lay within the boundary of the old Palatine settlement, which was still
(Varro L.L. 5.143) marked by boundary-stones (a fact which suggests that
the cult was not considered alien enough to be kept extra pomerium).
Virgil was probably not equating Pallanteum with the old historical
Palatine settlement so closely, nor is it likely, as Conington and others
have suggested, that he was thinking of the hallowed Athenian custom of
sacrificing to Heracles outside the city walls (Dem. Fals. Leg. p. 368).
ante urbem in luco occurs also at A. 3.302, and divisque ferebam at A. 3.19:
Virgil remembered nobody's poetry better than his own, and his remini-
scences do not always chime in with the material background of the new
context. Binder 43, n. 13, sees an intentional cross-reference between
A. 3.302 whose setting is Buthrotum, a copy of Troy, and A. 8.104 whose
setting is the site of Rome, the new Troy-to-be.
D. L. Drew's suggestion (The Allegory of the Aeneid, Oxford, 1927), that
Virgil moved the position of the altar deliberately (as his readers would
have appreciated immediately), because on Hercules' Day (12 August)
29 B.C. Augustus himself was ante urbem, outside the walls, waiting to
enter next day for his triumph, is one of his more attractive ideas, but,
like his general thesis, founders on the improbability of assuming that
Virgil was more preoccupied with oblique sous-entendus than with the
surface level of the narrative. A perceptive contemporary could no doubt
have read allusions to recent history into the narrative here, and 188-201
(see H. Bellen, RhM 106 (1963), 23 ff.).
58 COMMENTARY

104 Pallas huic filius una: 'Pallas his son was with him. ' It is best to
take huic (cf. Woodcock 45 f.) as a possessive dative, and una as an adverb
with its usual meaning of 'alongside, in company'. The normal way of
expressing 'along with him, in his company' is una cum eo. LS s.v. cite
this passage as a 'poetical' use of una with the dative, and Conington (ad
lac.) says' una with dat. like similis, pariter '. In the absence of any parallel
this explanation is virtually impossible, even though the dative is regularly
used with adjectives and adverbs expressing proximity. At A. 7.649filius
huic iuxta Lausus, huic is certainly a possessive dative and iuxta an adverb,
and at A. 8.631 huic ubera circum, huic is once again a possessive dative
and circum a preposition governing ubera.
105 pauper: a distinction is usually drawn (e.g. by Servius auctus on
G. 1.146) betweenpaupertas, respectable 'humble circumstances' (a much
advertised virtue of the early Republic), and egestas, squalid 'poverty'.
pauper is unlikely to mean few in number, as Servius suggests, although
this would fit well with the traditional account (see the note on 333).
106 tura dabant: Servius auctus sees a reminiscence of the technical
language of ritual: da quad debes de manu dextra aris. Note that the people
in general offer incense as a preliminary purification (cf. in its context
Horace Odes 4.2.51 f. dabimusque divis tura benignis); only magistrates and
priests could actually make prayer and sacrifice pro populo (in 179 f. the
roasted offal is served by the priest of the altar of Hercules and by lecti
iuvenes).
107 f. 'When they saw the towering ships and their crews gliding
towards them, parting the shady grove, and bending over the oars, rowing
silently ... '
tacitos: with nobody shouting the time to the rowers (so Servius) and
with as little noise as possible. The ill-supported variant tacitis, accepted by
Ribbeck, Hirtzel and Sabbadini, is not a case of a more difficult reading
being preferable, but of a less preferable reading being impossible. It is
superficially tempting to take tacitis incumbere remis = 'resting on their
silent oars', at an 'easy' position with the blades out of the water, but
Virgilian idiom (see the note on 30) requires tacitis to be taken as an
adverb, not an attributive adjective, as similarly with validisque incumbere
remis A. 5.15. But the chief objection to tacitis is that the subject of
incumbere is then rates, and to speak of 'ships bending over the oars' is
surely impossible. In the course of the sentence attention evidently moves
from the ships to their crews: this needs to be made explicit, as it is by the
consensus of the most reliable MSS, which offer tacitos.
no rumpere here = interrumpere, simple for compound verb. Com-
pound verbs, often intensive, are a feature of vigorous speech, popular and
educated; to use a simple verb in the sense of one of its compounds has an
archaic flavour, and often also a metrical advantage. In Virgil cf. A. 1.20
verteret for everteret (with Conway's note), A. 6.620 temnere for contemnere
(with Norden's note), and in A. 8 the particular case of ponere in the notes
on 329, 639.
If a sacrifice or festival was interrupted, it had to be begun again
 COMMENTARY 59

from the very beginning (instaurare), and in addition a propitiatory

offering (piaculum) might have to be made to win back the god's favour;
hence Pallas' anxiety. 175 f. seem to imply that there was some kind of
interruption, but 'religious' continuity could be maintained so long as one
worshipper remained to carry on the ceremony.
II2 procul: see the note on 98; here evidently within earshot, cf. Ter.
Hee. 607 quem cum istoc sermonem habueris, procul hinc stans accepi, uxor.
n2 f. subegit ... temptare: 'has driven (you) to try'. subigo has the
meaning of both (1) 'compel', (2) 'drive' in a literal sense implying motion.
(1) cogo (itself also a compound of ago) is followed by a prolative
infinitive in all periods and styles: subigo with the same meaning is the
easiest of extensions (cf., for example, Plaut. True. 783).
(2) After verbs of motion it seems entirely natural that an infinitive
should be used to express purpose, and it is commonly so used in the
colloquial Latin of Plautus, but it was unacceptable to the purist Latinity
of Caesar and Cicero. Nevertheless it continued to exist through the late
Republic in prose which had little claim to literary style (e.g. Varro's Res
Rusticae); and it was maintained, as a convenient verse form, by the
Augustan poets. In popular speech it never died, flourished in vulgar and
late Latin, and passed into the Romance languages (e.g. Fr. je vais
chercher).
n4 qui genus, unde domo is Virgil's compressed recasting of the
Homeric Tls 1r68Ev Els dvlipwv; 1r68, Toi 1r6A,s ~lit ToKfjEs; (e.g. at Od. r.170).
genus (as also at A. 5.285 Cressa genus) is an accusative of respect, of the
type frequently used by the Augustan poets in frank imitation of the
normal Greek construction (the native Latin basis for the construction is a
slender one, see Palmer, LL 288 f.); it does not occur in Roman comedy or
in classical prose, which prefers an ablative of respect with adjectives
(e.g. natione Medus). Tacitus is the first prose author to use this idiom with
adjectives other than participles (e.g. manum aeger, Hist. 4.81). Most
Greek-imitated constructions are extensions of vigorous Latin usages
already in existence, and find their way into prose by the time of Livy:
this is a notable exception. See the note on 425 nudus membra. unde domo
(as also at Plaut. Poen. 1376; Horace Epp. 1.7.53), literally 'from where,
from home', is best regarded as an idiomatic conflation, the second idea
added (in apposition) to make the first more precise; this seems more in
keeping with language development than to suppose with Conington that
domo is an ablative of respect. See LHSz 2, 105 and cf. Servius on Caerete
domo (A. 10.183): ac si diceret, qui sunt domo Carthagine.
An arma is a convenient metrical equivalent for the usual antithesis
pacem, an bellum at, for example, Sallust lug. 46.8.
IIS puppi. .. ab alta: Aeneas was perhaps the helmsman of the ship
he was in, and is therefore at the stern; cf. A. 5.12 ipse gubernator puppi
Palinurus ab alta; the ship itself is presumably alongside the river-bank for
which it was heading (101), but no details of the actual mooring manreuvre
are given. (Claude's famous painting of The Arrival of Aeneas at Pallanteum
shows Aeneas speaking from the prow of a ship with one bank of oars only;
6 EAC
60 COMMENTARY

and other details too do not suggest the careful reading of the text which
has sometimes been maintained.)
n6 paciferaeque manu ramum praetendit olivae: the olive branch,
often bound with wool (at 128 with a 'b'itta, a ceremonial chaplet), was carried
by suppliants (supplex 145) according to Greek practice (cf. Soph. O.T. 3
lKTTJplo,, KAalia,a,v JgE<TTEJLJdvo,) which associated the olive with Pallas
Athene as guardian of civil rights and the arts of peace. Virgil's mention
of Minerva as oleae inventrix (G. r.18 f.) is poetic assimilation, not an echo
of popular belief (see the note on 409), and at Lucan 3.306 orant Cecropiae
praelata fro.nde M inervae, the adjective Cecropiae is not merely picturesque.
However, to judge from Livy 30.36.4 velata infulis ramisque oleae
Carthaginiensium occurrit navis, the emblems had become as inter-
nationally recognised as the white flag of truce.
n7 Troiugenas (like Graiugenum 127) has an exalted epic ring about
it, as in Lucretius (r.465) and Catullus (64.355); it is ascribed to a carmen
of the vates Marcius by Livy (25.12.5). Juvenal uses it to mock the 'blue-
blooded Romans' (r.roo: see Mayor's note there for the Roman search for
Trojan ancestry, and cf. the note on 134-41).
n8 superbo: not an ornamental epithet of bello; the Latins had shown
their 'arrogance' by not recognising the rights of suppliants and guests,
cf. A. r.523 gentis . .. superbas . .. 540 hospitio prohibemur harenae. So
profugos egere does not = 'have driven to flight' but 'have driven out,
though they were refugees'.
n9 ferte . .. dicite: the plural is puzzling after no f. Pallas / ... volat
... obvius ipse; in the meantime presumably some of Evander's men had
come up to join Pallas. A similar (and equally understandable) change
from singular to plural occurs at Ennius Ann. 198 V accipe (Fabricius as
leader of the legation) ... 201 ducite (Fabricius and his retinue). See also
the note on 546 graditur.
124 This line is Virgil's recasting of the Homeric formulae X"palv 7'
~a1ra{ov70 (Od. 3.35, cf. Il. ro.542 liENi ~a1ra{ov-ro), and lv 7' a.pa al <f,v
X"'P{ (at, for example, Il. 6.253).
125 subeunt luco: with compound verbs of motion poetry and post-
classical prose could use a number of alternative constructions: (1) the
simple accusative in its basic use to express goal of motion, common in
Old Latin with simple verbs of motion; with compound verbs the accusa-
tive is also partly felt as 'governed' by the preverb sub-, etc.; cf. A. 6.13
iam subeunt Triviae lucos; (2) the accusative with a preposition which
repeats or varies the preverb, cf. A. 8.359 ad tecta subibant; (3) a dative, as
here, related partly to the poetical dative of motion after simple verbs
(it clamor caelo A. 5.451), but more closely to the dative of interest which
classical prose fittingly combines with compound verbs of motion used
figuratively (invaserat furor improbis, Cic. Att. 16.12.2): see Woodcock 39 f.
and 44 f.
125 This line is hardly a case of v<TTEpov 1rp67Epov, for relinquo is nearer in
meaning to 'I leave behind, abandon' than to 'I depart from'. (MS
authority supports the older spelling relincunt.)
 COMMENTARY

126-151 Aeneas addresses Evander, laying stress on their close kinship and
common enmity with the Latins, and suggests mutual pledges of support.
127 ff. Notice the carefully measured way in which Aeneas' ambassa-
dorial speech proceeds: the long exordium (127-33) (see the note on 374-80),
whose cola swell in length (two + two + three verses), to reach the balancing
statements of the two family trees (each four lines long, 134-7, 138-41)
and the terse conclusion (142); artful diplomacy abandoned for direct
personal intervention (143-5: note the change to lively emphasis); state-
ment of compelling reason for alliance (146-9); proposal and guarantee
(150-1). Aeneas exaggerates to gain his point (esp. 147 ff.) as much as his
enemies (13 ff.). The tone also changes: what started as a highly formal
request ends as a confident offer.
127 Servius auctus on this line says ignotus a demonstratione personae
suae incipere debebat: he may have been thinking of ambassadorial
protocol; he could hardly have been thinking of epic tradition. Neither
Odysseus at the court of Alcinous (Od. 7 and 8), nor Aeneas on meeting the
disguised Venus (A. 1.314 ff.) announces his identity in his first words,
though neither is inhibited by modesty when he at last does (Od. 9.19 f.;
A. 1.378 f.). Telemachus, in a situation which may have been at the back
of Virgil's mind here (see the note on 97 ff.), first addresses Nestor (Od.
3.79 ff.), then names his home in Ithaca, then tells his business-the
search for his father Odysseus: he does not name himself.
127 Graiu~enum: the older gen. pl. termination in -um was adopted
as a convenient alternative by the hexameter poets for first declension
masculine compounds in -cola and -gena, like agricola, caelicola: contrast
Lucretius' heavy Graiugenarum (1.477).
127 cui: a strictly anomalous dative with precari, which regularly
takes an accusative of the person prayed to, or a(b) with the ablative when
the object of the prayer is itself mentioned. In the isolated instances of
the dative in Plautus (Asin. 477 pergin precari pessumo?, Amph. frg. 13
(Lindsay) noli pessumae precari) it appears = pro with the ablative. The
abnormality here is due to the fact that the dative is entirely appropriate
with praetendere ramos (which is simply the 'variation' on the 'theme'
precari), that quem me . .. precari would be clumsy and ambiguous, and also
perhaps that Greek £vxm8ai is regularly combined with the dative in this
sense.
129 For equidem see the note on 471.
129 f. quod . .. fores: the indicative is used in a causal clause when the
speaker or writer vouches for the factual truth of the reason given, the
subjunctive when he does not. But the subjunctive does not necessarily
throw doubt on the truth of the causal statement (it certainly does not in
the present case); often the nuance of meaning which differentiates it from
the indicative is very slight. If fores is not simply being used metri gratia
for the awkward fuisti, its shade of meaning must be 'I was not alarmed
by the thought that you were ... ' as opposed to 'I was not alarmed
because you were .. .'.
6-2
62 COMMENTARY

130 a stirpe: 'by lineage', both being of Greek stock; but geminis . ..
coniunctus Atridis implies a closer relationship, and ancient mytho-
graphers quoted by Servius devised a number of possible connections: the
most interesting is that the sons of Atreus could, if they so wished, trace
a connection with Evander through a Pleiad daughter of Atlas, just as
Aeneas does in 134 ff. (see the note there). Accius' tragedy Atreus seems
to have expounded this Atlas-connection of its hero and may have in-
fluenced Virgil here (see S. Stabryla, Latin Tragedy in Virgil's Poetry
(Wroclaw, 1970), 62 f. and 122).
131 sancta oracula divom: those of Tiberinus, which had elucidated
the prophecy of Apollo through the Sibyl at A. 6.96 f. (MS authority
generally, as here, supports the older spelling divom).
132 Jortin's suggestion, mentioned by Conington, to transpose the
second halves of this and the following line, avoids the slight harshness of
the lack of a connective with tua t.d.f.; but the Ovidian point in sed mea
me virtus, tua terris didita Jama is questionable style for Virgil, and the
MS consensus forbids change.
terris: a self-contained spondaic fourth foot which can here suitably
gain emphasis through the slowness of the rhythm, but see the note on
57 recto.
133 coniunxere: a deliberate echo of coniunctus 130 for rhetorical
point; but such close repetitions of a word are by no means always
designed, see Mackail, Introduction Bo f.
133 fatis egere volentem: there is an echo here of Stoic dogma and
phraseology, cf. Seneca's sententia (Epp. 107.II) in his 'translation' of the
Stoic Cleanthes, ducunt volentem fata, nolentem trahunt. The same is true of
praecipere = 'to anticipate' in Aeneas' words at A. 6.105 omnia praecepi
atque animo mecum ante peregi(see Norden ad loc.), and of exercitus = 'hard-
tested' in Anchises' words about him at A. 5.725 Iliacis exercite fatis (see
Williams ad loc.). This does not make Aeneas a Stoic model by any means
(the Stoic traits in his character are discussed by Heinze, VeT 271 ff.,
276 ff. and C. M. Bowra, G & R 3 (1933), 8 ff.), but explains to some extent
his moments of unHomeric passivity and resignation.
volentem here superficially contradicts invitus, regina, tuo de litore cessi
at A. 6,460, the almost verbatim quotation of Catullus 66.39; if Virgil
thought about it at all, he is to be taken as making Aeneas equally sincere
in two totally different contexts.
134-141 This whole passage, devoted to establishing a mythological
family-tree of doubtful authenticity, and simply an expansion of the
phrase cognatique patres (132), may strike a modem reader as frigid and
unreal. But one must remember, first, that there was an epic tradition of
such genealogies: Homeric precedent ranges from the short notice of
Polyphemus' antecedents (Od. r.71 ff.), to Aeneas' boastful detail of his
own pedigree (Il. 20.213 ff.). Genealogy is here used in the service of
diplomacy, and such appeals to mythology had actually been used to
lend weight to various claims in the course of Roman history. In 262 B.C.
the Romans received the voluntary submission of the Segestans because
 COMMENTARY

they claimed to be related through common descent from Aeneas, and the
claim was probably made actually in the negotiations (see J. Perret,
Les Origines de la Legende Troyenne de Rome (Paris, 1942), 452 f. and
G. Karl Galinsky, Aeneas, Sicily and Rome (Princeton, 1969), 173); at the
end of the First Punic War the Acarnanians made diplomatic overtures to
Rome appealing to the fact that they had not fought with the other Greeks
against Troy (because they are not mentioned in the Catalogue of Ships?)
(cf. Justin 28.1.6). These early claims are interesting as showing that the
legend of the Trojan foundation of Rome must have been already widely
accepted by the period of the First Punic War. Later, in a much more
sophisticated and critical age Claudius (a confessed antiquarian) dwelt at
length on the mythological origins of the island of Cos as a just cause to
relieve it of tribute (Tac. Ann. 12.61); and the philhellene Nero, also in
53 A.D., obtained exemption from all public burdens for the citizens of
Ilium after discoursing on the foundation of Rome from Troy and the
Julian house from Aeneas (Tac. Ann. 12. 58), reasons which had previously
influenced Julius Caesar himself to grant them favours (Strabo 13.1.27).
Secondly, there was considerable interest in antiquarian genealogy
current in late Republican and Augustan Rome. Varro and Julius Hyginus
(cf. Servius on A. 5.389 and 704) both wrote books entitled De Familiis
Troianis: the quest of undistinguished families for Trojan ancestors was as
eager as the quest for Norman blood once was in England, and Virgil
himself had ministered to the need (see A. 5.n6-23 and 568).
134 ff. The stemma of the family of Atlas (cf. Ovid Fasti 4.31 ff.):
Atlas

Corythus =Electra= Jupiter = Maia Dione= Tantalus

I I I I
lasius Dardanus Mercury Pelops

Ericht~Ilos
I I
I
Atreus

----------
Tros Laomedon
I
Assaracus
~
Priam Hesione
Agamemnon Menelaus

I
Capys
I
Anchises
I
Aeneas

134 ff. The crucial point in Aeneas' version of the genealogy is that it
was one and the same Atlas (caelifer) to whom both families could be
traced: hence the virtual repetition of 137 in 141. Roman mythographers
quoted here by Servius might have contested this; they distinguished three
 COMMENTARY

different Atlases, one who was a Moor, the African caelifer; a second who
was Italian and the father of Electra; and a third who was Arcadian and
the father of Maia.
134 According to Homer (Il. 20.215 ff.), Dardanus, son of Zeus, settled
Dardania before the actual foundation of the city of Troy, being himself,
apparently, indigenous to the Troad. Virgil however follows a recasting
of the myth, probably of Etruscan origin, to the effect that Dardanus
emigrated to the Troad from Corythus (Cortona or Tarquinii) in Etruria
(A. 3.170, 7.240), leaving behind some of his race in Italy: hence Aeneas
says (A. 1.380) ltaliam quaero patriam et genus ab love summo (there, and
at A. 6.123 et mi genus ab love summo, Aeneas is appealing to his descent
from Jupiter via Dardanus, a point insisted on elsewhere (cf. A. 7.219 ff.),
not via Venus, daughter of Jupiter by Dione). See further N. M. Horsfall,
]RS 63 (1973), 68 ff.
advehitur Teucros (136) is confusing: it appears to suggest the alternative
story, that Teucer was Troy's' first founder' and already established when
Dardanus arrived. This version had been preferred by Anchises (A.
3.104 ff.) and proved wrong by the event.
134 primus (pater . .. et auctor): here and at 269 primusque Potitius
auctor, primus is used with an adverbial meaning, 'originally', reinforcing
the idea of initiation in auctor. Pleonastic expressions with primies are
common, cf. A. 4.284 prima exordia and Austin's note.
135 Electra . .. Atlantide: cf. Ap. Rh. 1.916 vfjaov Js 'HMKTP7JS 'AT.\av-
Tl8os.
ut Grai perhibent: a source which Evander would hardly question.
138 ff. Mercury's antecedents were established beyond dispute: Hesiod,
Theogony 938 Z71v, 8' ap' 'AT.\avTtc; Mal71 TEKE KV8iµov 'Epµfjv. Maia was a
mountain-nymph of Cyllene before becoming the brightest (candida,
'gleaming') star of the Pleiades cluster, which included also her sisters
Electra and Dione. Cf. G. 1.221 ff. where she represents the whole group of
Atlantides = Pleiades.
138 vobis (sc. Arcadibus) Mercurius pater est: the cult of Hermes at
Pheneus (165), in a valley west-north-west of Mt Cyllene, was famous in
antiquity, cf. Cicero Nat. Deorum 3.56 Mercurius, quem colunt Pheneates
(distinguished by him from Mercurius love ... natus et Maia). Hermes
frequently appears on coins of the city: see especially Greek Coins, C. M.
Kraay and M. Hirmer (London, 1966), Plate 160, no. 515, Hermes carrying
Areas. pater, here, cannot mean 'male parent'. Dion. Hal. 1.31.1 and
Pausanias 8.43.2 regard Evander as son of Hermes-Mercury; Virgil is not
so explicit about the relationship, probably for stemmatic reasons, see
above on 134 ff.
139 Cyllene, in north-eastern Arcadia, rises about 2,375 metres above
the valley in which Pheneus lies. Because of its altitude it is cold even in
summer, when snow still sometimes stays on the peak (gelido . .. vertice,
cf. Priapea 75.10 Cyllene ... nivosa): the same is true of the other moun-
tainous regions of Arcadia (159).
(The present tense of the variant fundit is not objectionable (see the
 COMMENTARY

note on generat 141), but not preferable to fudit which has greater MS
support.)
140 f. Atlas: note the repetition to secure emphasis; cf. me, me 144,
ne . .. ne 532, nunc, nunc 579, contra . .. contraque 699, omnibus . .. omnibus
718, and see the note on anaphora 564 ff.
141 generat: the present indicative is often used (where English uses
a simple past) in cataloguing contexts. This use has obvious affinities with
the historic present but is important enough to constitute a separate
category. It is found in lists of items, e.g. A. 9.266 cratera antiquum quem
dat Sidonia Dido; and more particularly in lists of ancestors and genea-
logical contexts (as here and A. 10.517 ff. Sulmone creatos / quattuor hie
iuvenes, totidem quos educat Ufens, / viventis rapit; Livy 1.3.6 Silvius deinde
regnat; ... is Aeneam Silvium creat). As this idiom only comes into vogue
with the Augustan poets, it is certainly an adaptation of the exactly
corresponding Greek usage with verbs connected with birth, e.g. Eur.
Bacchae 2 .1,6vvao,, Jv -rlKT€t TTo(J' ~ Ka.8µ.ov K6pTJ. See LHSz 2, 306.
142 scindit se: · divides itself' iuto streams; the Romans thought of
'family rivers' rather than' family trees', cf. Ovid Met. 15.739 with Statius
Th. 1.225 f.
143 f. 'Relying on these facts, I decided not to put you to a preliminary
test through ambassadors or diplomacy.' Note the interwoven word-order.
per governs both legatos and artem, but is placed before the second only;
cf. A. 5.512 illa Notos atque alta volans in nubila fugit, A. 6.692, Horace
Odes 3.25.2; further examples are assembled by A. E. Housman, CQ IO
(1916), 149 = The Collected Classical Papers 3, 938. This is better than
supposing, with Conington and Page, that legatos is the object of pepigi by
a kind of zeugma (for pango/paciscor rarely have a personal object-except
the woman who is the 'object' of a marriage covenant), and better than
following Henry's suggestion to punctuate after legatos and 'supply' with
it some such verb as misi. H. W. Garrod, CR 24 (1910), n9 f. ingeniously
removed the difficulty by suggesting a parenthesis: his fretus non legatos
(neque prima per artem / temptamenta tui pepigi) me me ipse meumque /
obieci caput. Such a parenthesis may be reconcilable with the formally
prepared style of the speech.
After presenting his credentials, Aeneas is speaking in a rather high-
flown diplomatic tone: the abstract plural temptamenta is a coinage of
Virgil's own (see the note on 89 luctamen), and pango in its various forms
has a ring of legal formality.
143 prima is here virtually equivalent in sense to the adverb prius,
'beforehand'; cf. A. 6.152 f. sedibus hunc refer ante suis et conde sepulcro. /
due nigras pecudes; ea prima piacula sunto, and see Wagner, QV 28,
3 (d).
144 f. me, me ipse meumque / obied caput: 'I personally put myself
and my individual freedom in jeopardy'; Conington compares Soph. O.C.
750 a.€{ a€ KTJl>€1fovaa Ka~ To aov Kapa. Primitive Roman belief held that the
head was the source of procreative power; and, more important, that it was
the seat of a man's essential personality and of his power of self-determi-
66 COMMENTARY

nation: this latter is very evident in legal phrases like deminutio capitis and
capitis damnare, and underlies Virgil's use of the word here, and at 484
and 570. See further R. B. Onians, The Origins of European Thought 2 ,
Index s.v. caput. So at A. 2.751 caput obiectare periclis the dangers include
not only death but capture and enslavement.
144 me ipse: me is elided more than any other monosyllable in the
Aeneid, but then only 28 times in the whole poem; cf. in this book me
excidiumque (386), and see the note on tum 503.
146 Daunia here = Rutula. The Daunus who in Virgil's account of
the myth is father of Turnus and Jutuma (A. ro.615 ff., 688, etc.), is to be
distinguished from Daunus the eponymous king of Daunia, a name of
prehistoric Apulia (cf. Horace Odes 3.30.rr f., 4.14.25 f.), who welcomed
the immigrant Diomedes and gave his daughter in marriage to him (Ovid
Met. 14.456 ff., 510 f.).
147 insequitur: a sense pause within the line itself occurs most
frequently in Virgil at the strong caesura in the third foot, and next most
frequently at the strong caesura in the fourth foot; next (longo sed . ..
intervallo) in frequency is the pause after the first long syllable (i.e. at the
strong caesura) in the second foot. Before this last pause a single word
may be either a choriambus or a molossus (there are sixteen and fifteen
instances respectively in A. 8), and is very often a verb-form, whose sense
may be reinforced by the rhythm (but obviously a pause of such frequent
occurrence does not always have a descriptive effect); with insequitur here
before a pause, with meaning and rhythm supporting each other, compare
protrahitur 265, distulerant 643.
149 I.e. et teneant mare quad supra adluit quodque infra (adluit); here
teneant has 'overstepped' into the relative clause (hyperbaton). mare
superum = the Adriatic, mare inferum = the Tyrrhene sea. The line is
substantially repeated from G. 2.158; its genuineness is above suspicion
in both places.
150 Ennius Ann. 32 V Accipe daque fidem foedusque Jeri bene firmum
(perhaps Aeneas speaking to the king of Alba Longa). We do not need to
follow Norden (Ennius und Vergilius 162, n. 3) either in his certainty that
the line of Ennius necessarily ended Aeneas' speech, or his view that Virgil
has spoiled its powerful effect by adding to it. Virgil could hardly have
taken the line over wholesale: neither its forceful (but unsubtle) alliteration,
nor its ending with three disyllables could have been blended with his
usual manner: he takes Ennius' 'theme', but for his' variation' substitutes
a recommendation rather than a restatement. In any case, the already
existing ties of family guest-friendship, stressed by Evander (169), quite
apart from the relationship established by the family-tree, make the second
half of Ennius' line inappropriate.
151 pectora: Virgil is quite fond of neuter plural nouns as dactylic-
word beginnings of a hexameter; often they are followed by a pause of
some kind, as here and pocula 176 (very slight), saecula 325 (heavy),
funera 571 (light), munera 613 (heavy), moenia 715 (light); such additional
emphasis as they sometimes have when followed by a pause is due to their
 COMMENTARY

position as the last word in a clause and first in the verse, not to their
dactylic rhythm as such (contrast the note on 87 leniit).
151 spectata = 'tried and tested', = probata, as commonly in Cicero.

152-174 Evander recognises Aeneas from his likeness to his father Anchises
whom he had met and befriended in Arcadia. He promises to arrange help for
Aeneas on the following day, and invites the Trojans in the meantime to join
in the sacrifice as allies.
152 dixerat: this 'absolute' use of the pluperfect as the only verb of
the sentence in which it stands is best explained with reference to the
normal use of the tense: the perfect stem denotes completed action,
the pluperfect tense completed action prior to the time of some part of the
surrounding context. The idiom is well illustrated by Prop. 2.8.10 et Thebae
steterant altaque Troia fuit which strictly means 'and Thebes had stood
( = has ceased to stand) and towering Troy has been ( = exists no longer,
is a thing of the past)'. So here dixerat = 'finished speaking'. See KS
1, 139 f., and on dixerat as a formula, Norden on A. 6.633 and Anhang
1.3 (p. 374).
153 iamdudum . .. lustrabat: Latin and the Romance languages stress
the nearer aspect of a continuous action, English the remoter; so to
describe what had been and was still going on Latin uses an imperfect,
English a pluperfect (the pluperfect and the perfect in Latin could not be
used to express the idea because they denote completed action, see the
previous note). The continuity is often stressed by the addition of
iam + dud um, pridem, or diu.
lustrabat: see the note on 229.
153 lumine, 'with his eyes', an exceptionally bold use of the collective
singular. At A. 3.677 f. cernimus astantis nequiquam lumine torvo / Aetnaeos
fratres, lumineshould probably be taken strictly: Polyphemus' brothers had
only one eye like himself. But Evander was not luscus. No doubt the metre
and the sound-echo in lumine lustrabat encouraged Virgil to be bold
(cf. A. 2.754, and to Austin's references there add Ovid Met. 10.293 f.
(Pygmalion's statue charmed into life) timidumque ad lumina lumen /
attollens pariter cum caelo vidit amantem). But lumen does not belong to
any of the usual categories of collective singular (two are mentioned in the
notes to 599 and 705 f.).
157 ff. 'For I remember that Priam the son of Laomedon, while on a
visit to the kingdom of his sister Hesione (Salamis was his objective),
passed on to visit the chilly territory of Arcadia.' This makes geographical
sense of a passage whose expression is not entirely clear (and note the
pointless repetition of visentem . .. invisere). protinus = 'onward, con-
tinuing on the journey'.
The time-reference of the present participles here is as vague as is
usually the case with their English equivalents. In Latin they can have
future reference (cf. in its context Catullus II.IO Caesaris visens ( = vi-
surus) monimenta magni), but not here: nobody who had sailed from Troy
68 COMMENTARY

would 'go on' (protinus) to pass through Arcadia 'intending to visit' the
island Salamis. Sometimes they do duty for the past participle active which
Latin does not have: this would suit both protinus and the geography, but
to use this idiom in conjunction with the verb petere is at the best con-
fusing. Virgil may subconsciously have remembered the structure and
phrasing of A. 1.619 ff., which deals with the banishment of Hesione's son
Teucer to found Cypriot Salamis, where the context gives petentem a
different and unambiguous sense.
157 ff. Servius auctus (on A. 3.80 ff., where Anchises is recognised by
Anius king of Delos) says that Anchises had visited Anius before the
Trojan War to ask whether he should accompany Priam on the journey
here referred to. This may be an interesting glimpse into the ramifications
of earlier myth which surrounded even the secondary characters, or (more
probably) a commentator's attempt to devise such ramifications on the
slender basis of the two episodes of recognition (although such recognition
from previous hospitality is not uncommon in Homer; see the note on
97 ff. and cf. also Il. 3.205 ff.).
157 f. Part of Laomedon's punishment (see the note on 18) was to
expose his daughter Hesione to a sea-monster sent by Poseidon; but she
was rescued by Hercules and given in marriage to Telamon king of Salamis,
to whom she bore Teucer and Telamonian Ajax.
158 Four-word lines (103, 214, 263) are rare enough to be arresting.
160 prima (iuventas): 'my earliest youth'; cf. 517 primis ... ab annis.
For the form of the imperfect vestibat see the note on 436 polibant.
160 Idea and phraseology recall Lucretius 5.888 f. tum demum puerili
aevo florente iuventas / occipit et molli vestit lanugine malas; the metaphor
of 'blooming' goes back to Homer Od. 11.319 f. aµ,<poTlpw, 1rplv a<{,wrv v1ro
Kponuf,o,aiv lovAovs- / dv071aai 1TVKO.Uat T€ ylvvs- dmv0lr Aaxvr,,
162 cunctis altior: Homeric heroes regularly tower head and shoulders
above the rest, like Menelaus and Ajax at It. 3.210 and 227. This may
reflect an ethnic difference (Northern origin?) of the conquering Achaean
invaders (as maintained by Sir W. Ridgeway, The Early Age of Greece,
passim). Hence in Homer they and whatever belongs to them tend to be
over life-size, and Virgil continues the tradition, also, for example, at
A. 7.783 f. See the note on 250 molaribus.
163 Anchises, like other words which end a sentence and begin a line,
sustains heavy emphasis.
163 f. ardebat amore / compellare: the basic verbs of wishing, volo
and cupio, have their sense completed by an infinitive at all periods of the
language. Verbs of the same basic meaning but stronger emotive force
were introduced at various stages (usually by Augustan or later poets) and
by analogy constructed with an infinitive: ardeo was established by Virgil
(although there is an isolated example in Sallust lug. 39.5) ;furo was intro-
duced by Horace, caleo and flagro by Statius. In the present passage the
analogical extension is particularly easy, for ardebat amore is equivalent
in sense to ardentissime cupiebat (or amabat). See further LHSz 2, 346 f.
165 Phenei sub moenia: the inhabitants and other inland Arcadians
 COMMENTARY 69

tmo Kv.:U~V1), t5po, al1rv were provided with sixty ships by Agamemnon to
sail to Troy (Jl. 2.603 ff.). According to Strabo 8.388 (who should not be
taken too seriously) little remained of the city in Augustan times.
166 f. pharetram . .. chlamydemque: Greek loan-words fall with
natural ease from the Greek Evander. But even without such special
appropriateness they had a traditional place in epic: Ennius introduced
them, usually without further explanation: e.g. lychnorum, campsant,
bradys (Ann. 323, 328, 423 V). Note also scyphus, A. 8.278.
166 Lycias: with pharetram as well as sagittas-a feasible gift for
someone to bring from the neighbouring Troad; but the adjective is partly
ornamental-Camilla too has a 'Lycian quiver', A. 7.816.
167 intertextam: according to Norden 446, the seven instances of
spondeiazons in Virgil which are not echoes of either Greek or earlier
Roman poetry (see the note on 54 Pallanteum) achieve some special effect
by their rhythm. This is a plausible view, for the only example of a
spondeiazon of any kind in the most finished books of the Aeneid (2, 4, 6)
clearly aims at such an effect: A. 2.68 constitit atque oculis Phrygia agmina
circumspexit (Sinon slowly surveys the enemy ranks). But what special
impression is conveyed here? wonder at the splendour of the cloak (so
Norden 446) or the effort of weaving it (so Win bolt 131)? More probably the
word is simply treated, through the nearness of pharetra and chlamys, as
another Greek loan-word.
There is nothing to choose in sense or style between intertextam and the
variant intertexto: MS authority favours the former, which would be
liable to assimilation to the nearer aura.
168 The simple language of the line (although the word-order is inter-
laced) and its almost entirely homodyned dactylic rhythm are unusual,
since no descriptive effect is aimed at; but it contrasts strongly with the
preceding line and may be an extreme case of variety for its own sake. Note
also the number of words in 166 and 167 on the one hand, and 168 and 169
on the other.
168 bina: distributive numerals, 'two each', etc., are regularly used in
Latin instead of the corresponding cardinal numerals (r) with nouns
which have only a plural form (bina castra); (2) with the plural of nouns
which in the plural have a different meaning from the singular (binae
litterae, 'two letters'; Servius on this line tells us that Cicero once repri-
manded his son for saying direxi litteras duas, which in strict usage could
only mean 'I despatched two letters of the alphabet'); (3) with nouns of
objects forming a pair (Cic. Verr. 4.32 binos (scyphos) habebam). bina frena,
'a pair of bits' falls into this last class, but poetry exploited the metrical
convenience of the distributive form far beyond regular usage, cf. 47 ter
denis, and note.
169 'And so the pledge (of alliance) which you seek is (already)
established through the bond (of friendship).' ergo must refer to what
immediately precedes: the guest-friendship, confirmed as usual by exchange
of gifts, between Evander and Anchises, entails mutual support between
Evander and Aeneas; hospitium and amicitia constitute the foedus. This is
70 COMMENTARY

the natural interpretation and that of Servius; for others see Conington and
Page. The social and political implications of hospitium and amicitia are
discussed by E. Badian, Foreign Clientelae (Oxford, 1958), II ff. mihi is
dative of interest, 'as far as I'm concerned', rather than dative of agent.
171 auxilio laetos dimittam opibusque iuvabo: 'Clever Aeschylus
has said the same thing twice' is the taunt of Euripides at Aristophanes
Frogs ro24. The history of repeating oneself is long and venerable. The
Hebrew Psalms are full of doublets like 'Fret not thyself because of evil-
doers, neither be thou envious against the workers of iniquity. For they
shall soon be cut down like the grass, and wither as the green herb'
(Ps. 37, 1 f.), and their characteristic style depends on the ritual repetitions
of the chant (carmen). Early carmina are often religious, and the solemn
repetitions of the same idea ensure that the deity will hear, and not mis-
understand. The same scruple operates also when the audience is human,
and may not hear or understand the first statement of an idea; and so we
find the same mannerism in recited poetry (and all ancient poetry was
expected to be read aloud, whether it had a history of oral transmission
behind it or not). In Latin poetry it was Virgil who first exploited doublets
and triplets of this kind to such an extent that they became the chief
feature of his style. The mannerism of connecting two or more statements
(verbs) referring to the same idea is obviously closely allied to the same
connection with nouns (see the note on 436); but the chief impetus to
develop this technique probably came from Virgil's study of Homer, who
not only employs tautologous doublets like T€A€VTTJcrrJS' T£ Kal lpfns- (Od.
r.293), but also coordinates two or more verbs when the relationship is not
purely connective (see Stanford, Odyssey pref. p. lxxxi parataxis and his
note on Od. ro.3ro-20).
Lucretius' usual manner had been hypotactic, to string over a number
of hexameters the subordinated periods of prose (see Norden Anhang 11.2):
perhaps necessary to the subtlety of didactic, but cumbersome, inelegant,
and straining to the reader (listener). Virgil's practice was continued by
Ovid in descriptive passages (see A.G. Lee, Metamorphoses 1, intro. 25 f.),
but rhetorical elaboration for its own sake begins to make itself felt, and
in Lucan 'theme and variation' often assume the appearance of two
complementary riddles (e.g. 5.383 f., 7.512 f.). Of numerous examples in
this book, cf., for example, 439, 598 f.
Dido also fulfils the claims of hospitium regally, and in almost the same
words, A. r.571 auxilio tutos dimittam opibusque iuvabo; there the ship-
wrecked Trojans needed protection (tutos), here Aeneas and his followers
need the increase in morale which extra reinforcements would bring (laetos).
172 f. A clear reminiscence of Peisistratus' words to Mentor (the disguised
Athene and Telemachus' companion) in Od. 3.43 ff.; cf. especially Od. 3.44
TOVAyap
\ (
SC.
n OU€WaWVOS'
\:,I )
Kai\ \Oat'T'T/S'
:,/
7JV'T'T/UaT€
t /
o€vpo
\:,A \/
µ,0110VT€S',
172 f. Strangers were not welcome to the Romans at religious functions;
they were unfamiliar and so dangerous. But the Trojans were not strangers,
as Virgil emphasises (amici 172, sociorum 174), and Hercules was in any
case a very hospitable god.
 COMMENTARY 71

173 faventes: Servius auctus is no doubt right to see an echo of the

religious formula favete linguis (repeated by Horace in Odes 3.1.2),
EVc/>TJfLE'i-rE, which amounted to an order for silence, the effective way of
preventing any words of ill omen.
celebrate faventes is another echo of Dido's welcome (A. 1.735).

175-183 Aeneas is given the place and portion of honour at the sacrificial
banquet.
175 haec ubi dicta: cf. 337 vix ea dicta; the omission of est or sunt as
here in a compound tense and in a subordinate clause is not common before
Virgil (see Austin's note on A. 2.634 and his reference to Leo, Senecae
Tragoediae 1.188 f.). Virgil exploited it to devise formulae of rapid transi-
tion; compare the omission of sunt in prose expressions like sed haec
vetera; illud vero recens (Cic. Phil. 2.25); haec admirabilia, sed prodigii simile
est quod dicam (Lig. n). In A. 8, cf. the similar omissions after postquam
(184), and ut (191, 362).
The pause after a trochaic word in the second foot which these formulae
carry with them is something of a rarity; but it occurs also at 396 quo tibi,
diva, mei (which recalls Lucretius 1.38 hunc tu, diva).
175 dapes (and 186): Servius on A. 3.224 says dapes deorum sunt,
epulae hominum, but the original distinction was blurred in course of
time. Daps originally meant simply a meal offered to an unseen spirit, but
when the Romans adopted the practice of animal sacrifice, in which a
portion of the animal was offered to the god and the flesh was allocated to
the worshippers, dapes and epulae came to be applied inclusively to both
(see further Bailey RV 51 ff.).
176 sedili . .. solio (178): Servius suggests that the choice of' seats' is
a good one, because a lectisternium (the ritual spreading of a couch with
the image of the god on it) was not allowed 'in the temple of Hercules'
(he must mean at the Ara Maxima itself and the nearby shrines, for at
other cult-centres in the city it did take place); the same information is
given from Cornelius Balbus by Macrobius 3.6.16 who adds propria
observatio est in Herculis sacris epulari sedentes. Virgil may have had this
custom in mind (toro 177 in context hardly means a piece of furniture).
177 f. 'and he welcomes Aeneas to the place of honour on a cushion and
shaggy lion-skin, and entertains him on a maple-wood chair.' invitat (cf. se
invitare, 'to enjoy oneself') is a near-synonym of accipit (cf. A. 7.210 f.
solio . .. accipit); the ablative with both simultaneously expresses means
and place.
The lion-skin hints strongly at Hercules' customary garb.
177 torus (perhaps connected with sterno) was first used to mean 'a
strand of rope' (Cato R.R. 135.4), then 'a bulge' of a plaited rope (so
E-M), then 'a bulge' of any kind: a bank of earth, a muscle of the body,
a cushion (and hence a couch). Here toro is the cushion of the chair over
which the lion-skin is spread: the structure of the similar 367 f. suggests
that it is not a case of hendiadys.
72 COMMENTARY

179 lecti iuvenes: 'chosen', presumably, according to the taboo-

qualifications still in force in Virgil's day for participating in certain
religious functions, notably aristocratic birth (the point of et tollens apicem
generoso vertice flamen, Lucan 1.604) and both parents still alive (see the
note on the Salii, 663 f.).
180 viscera: non intestina tantum, sed quidquid sub corio est, Servius
on A. 1.2u.
181 dona laboratae Cereris Bacchumque ministrant: it is almost
impossible to translate this into English accurately without being grotesque.
The phrase means that they serve bread and wine. The metonymic use of
Ceres and Bacchus to mean bread and wine respectively had long since been
established in Latin poetry (since Naevius): Lucretius 2.655 ff., thinking
of this and similar usages, allowed such 'abuses' as poetic diction, but of
course warned against believing in the reality of the gods concerned; the
Stoics had actually come to interpret the names of gods as personifications
of things (cf. Cic. Nat. Deorum 1.40). The diction of Virgil's phrase is
made more stilted by the addition of laboratae, which refers not, as Servius
thought, to the labour of growing corn, but to the labour of grinding and
kneading it into bread. A similar periphrasis occurs at A. 1.177 f. tum
Cererem corruptam undis Cerealiaque arma / expediunt . .. , where 'Ceres
spoiled by the waves and the tools of Ceres' means in plain English soggy
corn and grinding-mills. On this line Servius remarks: fugiens vilia ad
generalitatem transiit propter carminis dignitatem et rem vilemauxit honestate
sermonis, and his explanation is undoubtedly correct and of wide applica-
tion. Virgil took no greater pains to avoid anything than undignified
commonplaces and vulgarities of expression, sordida vocabula, in the Aeneid.
In fact the ordinary word for bread, panis, occurs in no poem certainly
written by Virgil (it is however found at I. u7 of the Moretum, which was
once thought to have been written by him).
Homer seems to have been quite unselfconscious about what he
described and how he described it; he is direct and unselective. But
although there is dung in the courtyard of Odysseus' palace (Od. 17.296 ff.),
and the dog which is lying in it is infested with fleas, Homer declines to
call a flea a flea, but speaks instead of Kvvopat<rrlwv 'dog-ravagers'. If this
is not deliberately mock-heroic, it is an exceptional instance of that
tendency to dignify the undignified which Virgil exalted to a cardinal
principle of epic. As a sensitive member of a mannered society he was
probably led to do this by his own natural delicacy; but his practice
degenerated into affected mannerism in his own Roman successors, and
paved the way for the dislocated values of the Baroque age (see Gilbert
Highet, The Classical Tradition (Oxford, 1951), 272 ff. and notes).
183 f. perpetui tergo bovis is Virgil's attempt to convey in Latin the
Greek S,71vEKij vwra /306, (cf. in an applied critical sense, carmen perpetuum
Horace Odes 1.7.6 and Ovid Met. 1.4 with arnrµa S,71vEK£, Callimachus
frg. 1.3 Pfeiffer), but he has given a typical twist to the expression by intro-
ducing hypallage (transferred epithet), for the adjective perpetui properly
belongs to tergo. Even so, Virgil's phrase is hardly so quaintly artificial as
 COMMENTARY 73

Dryden's translation 'Beefs continu'd Chine'. Translate: 'the undivided

back of an ox'.
In Il. 7.313 ff. Agamemnon sacrifices a five-year-old ox to Zeus and gives
this portion of the animal to Ajax alone as a mark of honour (after his duel
with Hector): 321 v6JToiaw 8' AZavm St'Y)VEKlmai ylpaipEv. But the situa-
tion which Virgil is likely to have had in mind here is the sacrifice by
Eumaeus of the best swine in his herd (Od. 14.414 ff.) and his similar
honouring of the unrecognised Odysseus (437 f. v6JToiaw S' 'OSvaija S,'Y)vE-
Klma, ylpatp€V / dpyi68oVTOS v6s, KViJatV€ s~ 8vµ,ov ci'.vaKTos). Other features
too of Evander's hospitality recall that of Eumaeus. Both hosts are in
humble circumstances, both go out of their way to honour their guests;
Odysseus is offered a seat spread with a shaggy wild goat's skin (Od.
14.50), and a bed of sheepskins and goatskins (Od. 14.519): Aeneas is offered
a seat spread with lion-skin (177), and a bed of leaves and Libyan bear-
skin (367 f.). Virgil has lent the details a certain picturesque dignity, for
a king (however humble) is not a swineherd (however noble).
Line 184 is Virgil's recasting of the Homeric formula avTd.p l1rd 1r6aws
Kat JS'YJTVos eg lpov iVTo which occurs after both the feasts mentioned in
the preceding note (Il. 7.323; Od. 14.454), and often elsewhere. Similarly,
Virgil had already used the first half of his line at A. r.216.
183 ' ... on the undivided back and entrails of a fattened, five-year-old
ox.' Here the exta are eaten by the worshippers, after the Greek heroic
custom (in both Greek and Latin epic; cf., for example, Homer Od. 3.9
and Val. Fl. 2.347 f.). But this is also an authentic and traditional part of
the Graecus ritus of the Ara Maxima cult (Livy r.7.13). (In Roman sacri-
fice, the exta were first cooked and then placed on the altar and burnt: this
was the god's share of the meal; man's share was the viscera tosta, all the
rest of the animal roasted.)
lustralibus suggests at first sight a connection with lustratio, a ceremony
of purification; it has this sense of 'expiating' or 'cleansing' at Lucan
6.786 and Val. Fl. 3.414 f. But such a meaning would be pointless here.
Warde Fowler (ASR ad lac.) takes the adjective to indicate that 'the exta
were holy, as having been on the altar', but it is hardly likely that any-
thing which had been so consecrated for divine consumption should then
be eaten by humans.
lustralibus here should be taken = 'five-years-old': (1) the preceding
part of this, and the whole of the next line are inspired by Homer: pre-
cedent for lustralibus is suggested by passages like Il. 2.402 f. ailTd.p &{3oiJv
t€p€Va€v
r,
ava~
,, i:. , ""'
avopwv
~ ,
'A yaµ,Eµ,vwv / 1rwva
' , .. K
1T€VTaET'Y)pov v1rEpfJ,EV€t
I r ,
poviwvi;
(2) Latin has no word for 'five-year-old' in the numeral series usually
applied to animals (bimus, trimus, quadrimus); Virgil gives that sense to
lustralibus here-nova cacozelia again-on the basis of its connection not
with the primary sense of lustrum, 'purification', but its secondary meaning,
'a period of five years'; (3) if we can believe Servius' note here and the
clearer version of it by Lactantius Placidus on Statius Th. r.507, lustralia
pecora were animals reared during the course of a lustrum specifically to be
sacrificed at its end, when the censors 'buried' it; they were therefore
74 COMMENTARY

fatten_ed-a secondary notion, but probably present in the context here

as well.

184-199 Evander explains the origin of the ceremony. He points to a

shattered cliff face where there was once the cave of Cacus, a fire-breathing
monster who used to terrorise Evander's settlement.
184 ff. After the first (meat) course (mensae primae) of a banquet it was
customary for silence to be ordered (so Servius auctus on A. 1.730), no
doubt so that the magister epularum could make a prayer (which he
might have to begin withfavete linguis); he then made a libation in which
the company joined, and this was followed by general drinking and possibly
also by entertainment. This is the sequence at Dido's banquet in A.
1.723 ff., and is also the basic scheme here although much extended and
elaborated: Evander makes prayer and libation (273 ff.) after the Cacus
narration, and the entertainment, accommodated to the ritual, is the
hymn to Hercules (285 ff.). For the problem of mensa secunda 283 see the
note there.
187 superstitio means more than 'superstition'; it refers to anything
outside established religious observance (see Warde Fowler, RE 106).
Cicero considered superstitio an evident threat to the solidity of the State:
ut religio propaganda etiam est . .. sic superstitionis stirpes omnes eligendae
(Div. 2.149). vana is explained by veterumque ignara deorum: Virgil makes
Evander speak like an orthodox supporter of the Augustan religious
programme. Some foreign cults were introduced under the auspices of the
State (the worship of Cybele, Magna Mater, was so introduced from Asia
Minor during the Second Punic War), those that were not (like the cult of
Isis, see the note on 698) were regarded as a threat to traditional forms of
worship, and often also as morally subversive (quo cuncta undique atrocia
aut pudenda confluunt celebranturque, says Tacitus on the arrival of
Christianity in Rome (Ann. 15.44)). The spiritual and emotional pov-
erty of the Roman state religion itself was the cause of the spread of
new cults, but the Romans hardly thought of orthodox religion in those
terms.
189 honores may be the common object of bothfacimus (cf. A. 5.762 f.
et aris / factus honos) and novamus; but facio was used intransitively from
the earliest times= 'I sacrifice' (as by Virgil at E. 3.77).
novamus cannot mean' we institute' here: sollemnem (honorem) 102 and
ex illo 268 rule out the idea that this was the inaugural ceremony; the
simple verb is being used for the compound (renovamus), 'we begin anew',
after Virgil's manner.
190 iam primum: iam is often used in conversational narrative to
mark the transition to a new point; Virgilian examples are collected by
Wagner, QV 24, 9 (he however curiously assigns a temporal meaning to
the present passage, 'antequam ad alia pergamus '-but what could the
alia be?). iam primum here means something like 'now then, to begin
with ... ', and introduces an EKtppacn, (picked up by hie in 193) in a more
 COMMENTARY 75

colloquial way than the usual est rupes and the like (see the note on 597).
The MS variant iam pridem is acceptable syntactically with the imperative
aspice (cf. A. 2.103 iamdudum sumite poenas), but the sense 'do look, as
you should have long since' is not compellingly appropriate.
19off. aspice ... disiectae ... ut moles (sunt) ... stat ... traxere: indic-
atives in an indirect question; contrast Horace Odes 1.9.1 f. vides ut alta stet
nive candidum / Soracte ... ? Virgil's paratactic manner made it easy for him
to adopt a form of expression current in the older language, where the ut
clause was felt to be neither indirect nor a question, but a direct exclama-
tion (cf. A. 8.154-6); cf. G. 1.56 f., A. 6.771 f., 779 f., 855 f.
(This construction, frequent in comedy, is absent, probably by accident,
in the surviving fragments of Ennius' Annales (see E. M. Steuart, The
Annals of Quintus Ennius 116); it appears intermittently in the old-
fashioned diction of Catullus' marriage-poems (see Fordyce on Catullus
61.77 f.), and persisted in colloquial speech even through the classical
period, e.g. in Cicero's letters. The influence on it of the regular Greek
use of the indicative in indirect questions is likely to have been negligible:
disregard Norden on A. 6.615.)
190 ff. Other authorities (Diodorus 4.21.2; Plutarch Romulus 20, 4)
suggest that Cacus' cave and the scalae Caci were on the Palatine. What
authority, if any, Virgil had for locating the spelunca on the Aventine
(cf. 230 ff.) we do not know (see Miinzer 89 ff.); his motive was to coordi-
nate the position of the cave with the fact that in the myth as organised
by him the Palatine had already been settled by Evander, and that the
Aventine had a special connection with Hercules and took its name from
his son who was born there (A. 7.655 ff.). It is possible that the cave had
not been precisely localised even by Augustan antiquarians and more than
one claimed the association (as was apparently the case with the casa
Romuli, see the note on 654). The statement of Solinus (1.7 f.) (third
century A.o.), that Cacus' dwelling was at the Salinae, the later site of the
Porta Trigemina (at the north corner of the Aventine) may go back to
Varro. No visible traces of the' steps' remain, but the path associated with
them which leads from the Palatine down to the valley of the Circus
Maximus, and whose top is very near the casa Romuli, has been identified,
Nash 2.299 f.
191 f. deserta . .. montis . .. domus: 'a derelict mountain home',
Cacus' cavern.
192 in~entem: ingens (which has already occurred in this book at 43)
is Virgil's favourite adjective: it occurs 168 times in the Aeneid, an
average of once every 59 lines, a constant reminder of the heroic stature
of everything in the epic poem (it does not occur in the Eclogues). Henry's
notorious note on A. 5.118 suggests that Virgil uses it quite indiscrimi-
nately, but Conway (on A. 1.453) more perceptively points out that its
concentration in certain passages can hardly be accidental: he might have
added to his examples the fact that five of the fourteen occurrences in this
book are in the Cacus-narration 190-267. Similarly immanis (also absent
from the Eclogues) occurs three times in this book, twice (225, 245) in thf'
7 EAC
 COMMENTARY

Cacus-narration. And one should not forget the undertone of strangeness

in ingens and monstrosity in immanis.
193 vasto: the primary sense of vastus, connected with English' waste'
and Latin vacuus and vanus, is 'desolate' (this is evident in the verb
vastare, at A. 8.8 and 374). Perhaps by association with tracts of barren
land or the sea, it came to mean 'huge, vast', the usual meaning in Virgil;
but he is fond of applying it to caves (here, 217, 424), perhaps with under-
tones of its basic sense.
194 semihominis: adapted to hexameters by treating the i as semi-
consonantal (see the note on 599 abiete); the vowel of the preceding
syllable is in any case long by nature (semi- ~ Greek~µ,,-), cf., for example,
Catullus 61.213 semihiante (semjiante) labello, and Aulus Gellius' friend
(N.A. 19.11.3 f.), ovK aµ,ovaos adulescens, who wrote cum semihiulco
(semjHilco) savio / meum puellum savior.
194 Caci facies: this kind of periphrasis, not very common in Latin
poetry (outside the Ilias Latina) is a cliche of Greek tragedy and epic, or
at least tended to become one; appropriately used, the periphrasis draws
attention to the aspect of the individual most relevant to the context, cf.,
for example, Il. 5. 78r f. fo-raaav, d.µ,</,1 /3i71v Llwµ~OEOS l1T1rooa.µ,ow / €lA6fL€VOt;
Soph. Ant. I "Q KOWOV avTCJ.0€A</,ov 'Iaµ,~V'TJ• Ka.pa; Soph. o.c. 1568 f. awµa.
T, O.VtKCJ.TOV 871p6,; but at Od. 2.409 TOtat (sc. J-raipois) oi Kat /L€Tfo</,' frp~
is T71>.£µa.xow 'the formidable might of Telemachus' is hardly appropriate
to describe the callowness of his still unformed character, which Homer
seems at pains to portray. Ennius introduced this formula into Latin epic:
he is quoted by Varro (L.L. 7.42) as using the phrase suavis sonus Egeriai
(Ann. n9 V).
Caci mens (205) is a formula of exactly the same type, and cf. A. 2.601
Tyndaridis facies, A. 9.272 f. lectissima matrum / corpora.
194 tenebat, 'used to occupy', see the note on 657. The variant tegebat
has overwhelming MS support and is kept by Sabbadini, who explains
thus: tegebat inaccessam: i.e. ita tectam servabat, custodiebat, ut inaccessa
esset, sc. solis radiis subtrahebat; deinde ab Hercule detecta est, v. 241. This
is unconvincing: Cacus might conceivably be said to protect the cave which
more naturally and appropriately would be said to shelter him, but hardly
to keep it out of the sunlight. But the conception of Cacus protecting his
cave is itself strained and implausible. We must suppose that a copyist's
error (of one letter only) was made very early in the tradition. It is no
objection to tenebat that tenebant occurs very shortly afterwards at 204:
there is an even closer repetition of exactly the same words at 653 and 657.
195 inaccessam is a new coinage of Virgil's, used only here and A. 7.n
inaccessos . .. lucos (where Servius comments inaccessibiles, inaccedendos,
non ad quos nullus accessit). Virgil, unlike Servius, favoured only forma-
tions in -bilis from first conjugation verbs; but the meaning they bear
could also be borne by the past participle (see the note on 588 conspectus,
and cf., for example, the instructive line A. 5.59r indeprensus et inreme-
abilis error 'undetectable and unretraceable '). Virgil has many new forma-
tions with negative in- and past participle (see the note on 82 concolor
 COMMENTARY 77
(17)), where the participle is regularly formed from a transitive verb;
but accessus from the intransitive accedo does not exist, and is comparable
only with errata (A. 3.690), regnata (A. 3.14, 6.793) and the like (see Austin
on ululata A. 4.609). After Virgil inaccessus is found in poetry in Silius
(3.516) and Statius (Ach. 1.599) and passed into later unliterary prose.
195 (spelunca) ... solis inaccessam radiis: the ultimate parent of
these sunless caverns may be the lair of the boar which scarred Odysseus,
Od. 19.441 OUT€ µ,iv (sc. TrJV >.6xµ,YJv) 'Hi>.w, <f,a/Bwv O.KTlaLV l/3aAA€V.
196 foribus ... superbis: the adjective (from super) may here keep
part of its basic physical sense, 'towering' (as at A. 7.630 Tiburque
superbum), although this is rare compared with the moral sense 'over-
bearing'.
196 f. This practice of Cacus is transferred to the Thracian Diomedes by
Ovid (Her. 9.89 Threiciis adfixa penatibus ora), for any atrocities could be
ascribed to the heavy-drinking Thracians (cf. Horace Odes 1.18.9 ff. and
1.27.1 f.).
197 pallida: Bentley, in his note on Lucan 2.166 proposed squalida
(itself the original reading of the Medicean MS here), on the ground that
tabum is atrum (cf. A. 9.471 f. simul ora virum praefixa movebant / nota
nimis miseris atroque fluentia tabo), and faces cannot be 'pale' with some-
thing that is black. This is sophistical preciseness; Virgil's compressed
pallida tabo means that the faces were sallow with decay and partly covered
with gore, cf. Ovid Met. 15.627 pallidaque exsanguisqualebant corpora tabo.
pallidus = 'yellowish', often used of unhealthy discoloration: see
Fordyce on Catullus 81.4. ·
198 The original nature and attributes of Cacus and his feminine doublet
Caca are unknown beyond what can be inferred from Servius' notes. Their
worship seems to have disappeared before the fixing of the earliest calendar
of festivals, and only the scalae Caci leading from the neighbourhood of
the Ara Maxima to the Palatine (and perhaps also the Caci atrium, see
Hiilsen in PW 3.n65 s.v.) preserved any memory of them. On A. 8.190
Servius says Cacus secundum fabulam Vulcani filius fuit, ore ignem ac
fumum vomens, qui vicina omnia populabatur, but there is no reason to
believe that this derives from a source earlier than Virgil himself; in the
same note Servius mentions a sacellum of Caca, in quo ei pervigili igne sicut
V estae sacrificabatur (so the best MS; the others give in quo ei per virgines
Vestae sacrificabatur), but Vesta was not the only deity in Greco-Roman
religion worshipped with an everlasting fire, and one cannot infer with
certainty that there was a closer parallel. It is at best a possibility that
Cacus and Caca were the male and female fire-spirits of the early Palatine
settlement: it has also been inferred from the fact that he was a cave-
dweller that Cacus was a chthonic power of the dead and the underworld.
The myth of Hercules and Cacus. The legend which connects Hercules
and Cacus is first found in Livy 1.7 and here, and is unlikely to be older
than the first century B.C. It is an aetiological myth, designed to explain
why certain old features of the city were near together, the forum boarium
(cattle market), a shrine of Hercules Invictus, an altar of Jupiter In-
7-2
 COMMENTARY

ventor at the Porta Trigemina and perhaps also a cult of Evander not far
from there (cf. Dion. Hal. r.32.2 Eva.v'Bpcp Ouata, iµ.a0ov v1To 'Pwµ.atwv
E7T£7'€Aouµ.lva, ... 7Tpo, ... AvEv-rlvlf) ... rrj, Tp,Soµ.ou 7TOA71, oti 7Tpoaw),
and the scalae Caci. To draw these elements together into one myth was
not a difficult task. Hercules in possession of cattle was already supplied
by the Greek myth of Geryon (201 ff.); his reputation as a monster-killer
required a monster to kill (Cacus) and a sufferer to be relieved (Evander);
but he clearly had to be wronged-obviously by Cacus' stealing of his
cattle, a motif supplied from the Greek myth of Hermes and Apollo; and
being uncertain where to find the cattle naturally prayed to Jupiter
Inventor (this last detail is not mentioned by Virgil but was an obvious
extension: an altar to Jupiter erected at Tibur in the time of Augustus or
Tiberius bears the inscription Iovi Praestiti Hercules Victor dicavit, CIL
14.3555).
Of other authors who deal with the Cacus-Hercules myth, Propertius
(4.9.1 ff.) and Ovid (Fasti r.543 ff.) are clearly dependent on Virgil;
Dionysius of Halicarnassus (r.39 and 42.2 f.) is independent, as also is
Livy (r.7.3 ff.), who may even have been Virgil's source (see A. Santoro,
I problemi delta composizione dell' Eneide: Livia Jonte di Virgilio (Altamura,
1938), 8 ff.).
Attempts were also made to rationalise the myth on euhemerist lines,
by reducing Cacus to a local baron (Cn. Gellius in Solinus 1.8), or Hercules
to a shepherd of overpowering strength named Garanus (Verrius Flaccus,
an Augustan antiquarian, cited by Servius on A. 8.203).
See further Wissowa RuK2 283 f.; Latte RRG 221; Munzer, Cacus der
Rinderdieb (Basel, 19rr).
198 f. atros . .. ignis: flames with black smoke, but in many contexts
ater means as much 'deadly' as 'black' (see Pease's note on A. 4.384 for
its constant connection with the flames of the funeral pyre).
198 The Volcanal, at the foot of the Capitoline hill at the north-west
corner of the Forum Romanum, was no doubt inaugurated as a cult-
centre of the god in very early times, when it lay outside the boundary of
the primitive Palatine settlement. But interest in it, as in other ancient
relics, seems to have revived in Augustan times, and Augustus himself
dedicated a pedestal to Vulcan in 9 B.C. (CIL 6.457). So the fathering of
Vulcan on Cacus could also be related to 'pre-urban' monuments. See
further Platner-Ashby s.v., and the pictures of what may be its remains in
Nash 2, 517 ff.
Note the bucolic diaeresis, which here stresses the antithesis between
huic and illius.
199 magna se mole ferebat, 'moved with massive bulk'; cf. A.
3.656 f. vasta se mole moventem / pastorem Polyphemum. magna (vasta) mole
are ablatives of manner making adverbial phrases which qualify the verb.
Comparable perhaps is the Homeric µ.EyaAw<17't at, for example, Od. 24.40
KEfoo µ.lya, µ.EyaAw<17'L, AEAaaµ.lvo, l.1T1Toauva.wv.
 COMMENTARY 79

200-216 Hercules arrives driving the cattle of Geryon. Four bulls and four
heifers are stolen by Cacus who drags them backwards into his cave, hoping
that the hoof-prints facing in the opposite direction will prevent discovery.
After a vain search Hercules prepares to move on, and the cattle fill the
countryside with their lowing.
200 aliquando: 'at length', a prose usage, with or without reinforce-
ment from tandem.
201 auxilium adventumque dei: noted by Servius as an instance of
v<TT£pov 7rpoT£pov (Hercules must have arrived before he could give help).
Before Virgil this mannerism was a rarity in Latin poetry; it was tradi-
tionally associated with Homer (cf. Cic. Att. 1.16.1 Respondebo tibi
v<TT£pov 7rpoT£pov 'Oµ:YJpiKws), and it was no doubt from Homer that Virgil
took and developed it (cf., for example, Od. 4.50 aµ,cf,i S' a.pa x>ialvas ov>ias
/3cU1ov ~a€ X LTwvas and see the Indexs.v. 7Tpw8v<TT£povin Stanford's edition);
but it is also of course entirely in keeping with his favourite manner of
joining two aspects of a complex idea with a simple copula (see the note
on 436): sometimes it may be more convenient not to join them in chrono-
logical order (here metrical convenience plays a part). See also the note on
6n, and contrast the apparent cases of v<TT£pov 7rpoT£pov in the notes on
85 and 125.
202 tergemini . .. Geryonae: the earliest and fullest notice of I'-YJpv-
ov£vs, I'77pvol"TJS, or I'77pvwv ( < yrJpvw) 'The Roarer', is in Hesiod Theogony
287-94: he was slaughtered by Hercules and robbed of his cattle (the
tenth labour in most versions) on the island of Erytheia, later identified
with Cadiz. In Hesiod Theogony 287 he is TpiKicf,a>ios, but in Aeschylus Ag.
870 he is Tpiawµ,aTos, and this is followed by Lucretius 5.28 tripectora
tergemini vis Geryonai, and Virgil, A. 6.289 forma tricorporis umbrae;
in art the three-bodied concept is regular from the mid-seventh century
B.C. onwards, cf. M. Robertson in CQ N.S. 19 (1969), 207 ff.
The gen. Geryonae is formed from nom. Geryones, like Anchisae from
Anchises (cf. Servius on A. 7.662).
203 Alcides aderat: Hercules belonged to the generation before the
heroes of the Trojan War (cf., for example, Od. 8.223 f.), like Anchises and
Priam (155 ff.).
203 hac: through the area of the forum boarium, the cattle-market, which
of course came to have its name associated with the cattle of Hercules.
204 tenebant, 'were reaching', cf. 657.
205 furiis is preferable to the variant Juris on grounds of sense and
style.
(1) It has been argued that furiis = 'furious anger', and that this is
entirely appropriate to describe Hercules after the theft of the cattle
(219 and cf. 228), but is inept to describe Cacus in the present situation.
But the word can also mean 'madness' (it is so used of Ajax at A. 1.41),
like Greek a77J: Cacus was out of his mind to attempt such a thing.
(2) It is no objection to furiis here that it is repeated at 219, for Virgil
is capable of even closer repetitions with no significant point.
80 COMMENTARY

(3) Propertius 4.9.n ff. is clearly influenced by the present passage:

hie, ne certa forent manifestae signa rapinae, / aversos cauda traxit in antra
boves, / nee sine teste deo: furem sonuere iuvenci, / Juris et implacidas diruit
ira fores. Juris may be a direct echo. But Virgil is at pains to represent
Cacus as a terrifying monster of evil, not as anything so human and
commonplace as a thief.
(4) Heyne objected to Juris as vocabulum epica gravitate indignum. The
associations of the word in Roman comedy, where it is often simply a
strong term of abuse, lend weight to this objection, which is not sufficiently
met, as Mackail thought, by the word's occurrence at G. 3.407 (or E. 3.16):
the epic Aeneid has more stringent rules of word-choice.
These reasons outweigh those which support Juris, viz. (a) that
although the relative merits of the MS groups which preserve each variant
count for little in cases like this where a scribe could so easily have mistaken
either for the other, nevertheless Juris was the only reading known to
Servius, and much more likely to have been corrupted into the common
juriis than vice versa; (b) that Juris Caci mens effera is a periphrasis,
identical in form with semihominis Caci jacies . .. dira (194), and this may
not be purely coincidental.
205 ff. The theft of Hercules' cattle. Virgil says little about Hercules
himself, and what he does say is carefully selected from the morass of
myth which surrounded him to emphasise his function as saviour and
civiliser. In Homer (Od. 21.22 ff.) Heracles is himself a sacrilegious cattle-
rustler, and Livy (r.7.5) adds the unsaintly detail that Cacus stole the
cattle when Hercules was in a drunken sleep (cum eum cibo vinoque
gravatum sopor oppressisset).
205 ff. Caci mens . .. tauros / avertit, but it was Cacus himself as a
person that did this, not his mind. This rather bold personification of
mens or animus derives from the high style of epic and tragedy. It occurs
1n w,
. H omer, Il . 17. 68 f . ,,. 'TWV- ovw 'TLVL (J vµo, ' ' G'TTJ'(J EGGLV E'TOl\µa
' EVL ' '' / avnov
' '
iA01µEVat MEvEAcfov 1<v'8aAlµow, and elsewhere in Virgil (A. 6.467 f. talibus
Aeneas ardentem et torva tuentem / lenibat dictis animum). At Livy
r.15.1 inritati Veientium animi .. . in fines Romanos excucurrerunt, a
subordinate clause intervenes (as here) to make the transition easier.
205 f. ne quid inausum / .. . fuisset: a pluperfect subjunctive in a
final clause is something of a rarity; its shade of meaning here can perhaps
be conveyed by 'so that nothing might be left undared '. The perfect stem
lays stress on completed action; Cacus' thought was nihil juerit inausum,
'nothing will have been undared ', his intention was ne quid juerit inausum,
'let nothing have been undared ', and this latter, subordinated and moved
back in time, becomes the phrase we have here.
208 avertit: 'carries off, steals', with praeda as object, A. ro.78,
Caesar B.C. 3.59.3; the vox propria for both embezzlement and cattle-
rustling (cf. TLL 2, 1321 f. s.v.); abducere is also often used of cattle-
lifting, cf. Ovid Met. r.677 and Lee's note.
207 f. Repeated from G. 4.550 f., which is itself a repetition of G. 4.538
and 540 (no wonder that Virgil remembered it). Note the adjustments
 COMMENTARY 81

needed to fit the lines to the new context; the most important is that
intacta . .. cervice iuvencas appropriate to a context of sacrifice is changed to
forma superante iuvencas because here the point is their beauty and hence
value. There is no cross-reference or similarity between the two contexts,
although each occurs in an epyllion-type narrative.
209 ff. Dragging cattle into a cave backwards so that the footprints face
in the opposite direction and baffle any search for them was a famous
trick of Hermes in the story of how he stole Apollo's cattle. This episode
is treated at length in the Homeric Hymn to Hermes: cf. esp. 75 ff.
7/'Aavo8{a, 8' ij>.av11€ Sui ,f,aµ,a8c!J8rn xwpov I Zxvi' a7/'0aTpl,f,a,. 80>.t'YJ, 8' ov
>.18€TO T€X"'YJ>, I avTla 7/'oi1aa, 07/'Aa,, Ta, 71'p6a8€v oma8€11, I Ta, 8' om8€11
71'p6a8€v, KaTa 8' Ef:1,7/'aAiv aVTO!, Ef3at11€, and ro3 d8µ,fjT€S 8' i'Kavov £<; av>.wv
v,fnµ,l>.a8pov.
Allusion was probably made to it also by Alcaeus in a hymn of which
only one stanza is extant (see Denys Page, Sappho and Alcaeus (Oxford,
1955), 252 ff.) (imitated by Horace Odes I.IO), and by many Alexandrian
poets and mythographers (see Allen and Sikes, The Homeric Hymns
ro3 f.).
209 hos: masculine, in accordance with the general rule that a predica-
tive adjective (or resumptive pronoun as here) is masculine if it refers to a
group including both masculine and feminine nouns, even if (as here) the
feminine noun is the nearer.
209 ff. Cf. Propertius 4.9.9 ff. incola Cacus erat, metuendo raptor ab
antro, / per tria partitos qui dabat ora sonos. / hie, ne certa forent manifestae
signa rapinae, / aversos cauda traxit in antra boves. The imitation is close,
and on the basis of it Wakefield made the attractive suggestion to read
raptor in 2II in place of the otiose raptos, which could easily have arisen
as an error through tractos in the line before or through rapto(r)saxo in
undivided script.
209 ff. 'And so that there should be no marks of feet going forward, he
dragged them into his cave by the tail, and having thus reversed the indi-
cations of their tracks, he busied himself with hiding his plunder in the
shadowy rock. There were no signs to lead to the cave if anyone searched.'
2ro Livy's similar phrasing (1.7.5) caudis in speluncam traxit (and cf.
the note on 215 f.) has been thought to point to a common source, possibly
Ennius; but the words are too commonplace and inevitable to prove any
borrowing in any direction.
212 The variants quaerentes and quaerentem look like normalisations
made on the assumption that ferebant is transitive; but with via, etc., as
subject the verb has a virtually absolute use, cf. A. 6.295 hinc via Tartarei
quae fert Acherontis ad undas, and this is perfectly compatible with a
dative (here a 'dative of local standpoint'), as at Caesar B.C. 3.80.1
oppidum primum Thessaliae venientibus ab Epiro.
Wakefield considered this line superfluous and spurious; his view is
elaborated by M. D. Reeve (CR N.S. 20 (1970), 134 f.) who objects (1) that
if quaerenti refers to Hercules, it is awkward to have another reference to
him in 214 after interea which marks a change of scene; (2) that 212 is the
 COMMENTARY

only self-contained line in the episode 201-67 and likely therefore to be an

interpolation. But (1) quaerenti most naturally = si quis quaerat, cf. KS
1, 776; (2) the second objection cannot be compelling in a poem which did
not receive the finishing touches: it may be a tibicen •interpolated' by the
author.
213 stabulis here (and just before in 207) means 'pastures' (Servius on
215 says quia avellebantur a pascuis). It should probably be taken therefore
dm, Kotvoii to complete the sense of both saturata and moveret, the ablative
doing double duty in different functions.
215 f. •as they departed the cattle lowed, and the whole woodland was
filled with their moans, and they bellowed as they left the hills behind'.
A good example, as Henry points out, of Virgil's 'theme and variation'
technique: contrast the subordinated prose of Livy describing the same
incident, 1.7.7: inde cum actae boves quaedam ad desiderium, ut fit, relic-
torum mugissent, reddita inclusorum ex spelunca boum vox Herculem
convertit (and notice the similarity with Virgil's phrasing of reddita . ..
boum vox). The clauses are bound together by an overlapping of some of
the ideas, discessu . .. relinqui, mugire . .. querelis . .. clamore; Virgil's
fondness for the simple ablative, discessu 'time when', querelis instru-
mental, clamore modal; and the usual series of historic infinitives (see the
note on 493).
Servius insists that the cattle bellowed because they were being taken
away from good grazing, 'not, as many say, because they had lost their
companions'; but the phraseology strongly recalls Lucretius' description
of the cow which has lost her calf (2.358 f.), completque querelis / frondi-
ferum nemus.
colles clamore relinqui = colles a bubus relinquebantur cum clamore,
admittedly compressed, but not un Virgilian. Difficulty has been felt
about the topography: if 'hills' were left behind, the cattle must have
moved some way up the valley, but, it is objected, we are not told this, or
that Hercules had to turn back to regain the Aventine. This led G. Watson
(CR 68 (1954), 99 f.) to suggest that the phrase means that the sound
echoes from the hills (impossible), and A. Y. Campbell (CR 69 (1955),
137 ff.) to propose propinqui for relinqui (Peerlkamp, unpersuasively
radical, had already suggested clamare propinqui, see CR 72 (1958), 15 f.).
It is probably misguided to press for precision; those who cannot dispense
with it would do least harm to the text by interpreting relinqui as equi-
valent to an inceptive imperfect.

217-232 One of the stolen cattle answers the lowing and betrays the theft.
Hercules blazes with anger and pursues Cacus, who takes refuge in his cave,
blocking the entrance with a boulder. Hercules is temporarily baffled.
219-24 A clear example of the use of descriptive dactyls to reinforce the
notion of speed, offset by the slow spondaic wonderment of 222 and the
beginning of 223.
219 f. 'At this point Hercules' hot-tempered indignation really blazed
 COMMENTARY 83

out in fury ... ' and syntax and word-order become almost dithyrambic:
the sentence is welded together by the grammatical ambiguity of all the
nouns in it (except dolor): Alcidae may be genitive or dative, and the
ablative furiis can legitimately be interpreted in more than one way (for
the similar ambiguity of Jelle see the note below). The translation suggested
above assumes that Alcidae is possessive genitive with dolor (but observe
the wide separation), and that/uriis is ablative of manner; but a strong
case could be made for taking Alcidae as dative of interest and furiis as
ablative of cause.
219 furiis: furiae not personified, and meaning 'fury, frenzy' is found
only in poetry and only in the plural. Poetry from Plautus onwards was
very fond of the plural of abstract nouns. Basically the plural seems to have
represented a repetition or continuation of the state denoted by the
singular (ira, anger: irae, outbursts of anger), then to have become simply
a more emotive equivalent, and finally to have degenerated into a cliche.
Only the context and what we know of the author's practice can allocate
any particular example to one of the above three stages. See further
KS 1, 77 ff., LHSz 2, 18.
219 furiis (and 205): Virgil uses the oblique cases of furiae plural for
the less manageable forms of the plural of furor (which occurs in the plural
only in the accusative at A. 4.501, 5.801, 7.406): there is no difference of
meaning.
219 f. exarserat atro / felle dolor: the ablative phrase, closely
following another ablative furiis, cannot be allocated to any one exclusive
grammatical category, because its function in the sentence is not narrowly
precise. In fact, in view of many other similar examples, it looks as if
Virgil's intention was to let its meaning pervade the surrounding context.
It is possible to take it as an ablative of manner' with black gall', or as an
ablative of 'place from which' governed by the preverb in exarserat, the
gall being thought of as the source of the indignation. Mackail (App. A,
p. 515) takes the view that Virgil intends the ablative phrase to take the
place of the kind of compound adjective which had come to be felt as
'inappropriate to the quality of a pure Latin idiom' (like Pacuvius'
repandirostrum, 'back-curving-snouted', or Varro's stellimicantibus 'starry-
glittering' (M enippeae frg. 92 Biicheler)); so atro Jelle dolor renders the
Greek a.xos µE"Aayxo'Aov. This is a very attractive view, supported by other
examples (see the notes on 31, 693). 'Black bile' was thought to be the
cause not of melancholy in a modern sense, but of passionate emotions
like anger.
219 f. exarserat . .. rapit: the dactylic lightness of the pluperfect no
doubt encouraged its use in epic poetry, and the use of the historic present
is common in all literary forms to secure a vividness which the perfect does
not have. Even so the tenses here were no doubt chosen for their sharp
time-contrast; the pluperfect draws attention to the extreme rapidity of
the action and the speed with which it gave way to the second. An exactly
similar instance occurs at A. 12.430 f. ille avidus pugnae suras incluserat
auro / hinc atque hinc oditque moras hastamque coruscat. The nuance of
 COMMENTARY

meaning implied by the pluperfect cannot be reproduced neatly in English.

220 arma . .. nodisque gravatum / robur: is the -que connective, 'his
weapons (bow and arrows according to Servius, and Hercules as archer is
familiar enough) and his club', or 'epexegetic ', 'his weapons, viz. his
club'? Stylistically either is possible; the sense of 249 f. supports the first.
221 ardua montis: a substantivised neuter plural adjective with a
partitive genitive is a rarity in Caesar and Cicero's speeches, and its
authenticity there is usually open to doubt (see KS I, 433): this suggests
deliberate exclusion of a construction which might at first sight be un-
clear (the adjective ardua rouses the expectation of a following noun to be
construed as normally with the possessive genitive: petit ardua mantis
(e.g.) iuga). But the construction was adopted by poetry from Ennius
(Ann. 89 Vin infera noctis) onwards, by Sallust and post-classical prose;
in some cases it is possible to detect a different nuance of meaning between
this construction and the straightforward agreement of adjective and noun,
e.g. between viarum angusta (Tac. Hist. 4.35), 'the narrow parts of the
streets', and angustae viae, 'the narrow streets' (implying that they were
all consistently narrow); but the metrical convenience of the periphrasis
no doubt promoted its use where the difference in meaning is not parti-
cularly significant.
223 oculis has overwhelming MS authority and is not open to serious
objection if taken with turbatum = 'with panic in his eyes' (cf. Livy
7.26.5 oculisque simul et mente turbatum); the preceding timentem is
narrowed to one of its most obvious symptoms (Conington compares Soph.
Ajax 139 f. 1-d>.av OKVOV ixw Kat 1mf,6f3TJp,ai I 1TT'Y}vijS' WS' oµ,µ,a 1T€A1das), which
h2.:5point in contrast with Cacus" terrifying' eyesastheywereusually (266).
Admittedly turbatits = simply 'confused' does not need further qualifi-
cation (cf. 435), and when this is expressed Virgil elsewhere uses an
accusative of respect (cf. 29); the ill-supported variant oculos, although
preferred by Heinsius and Heyne, looks like an attempt to produce this
usual construction, which Virgil may deliberately have avoided so as not
to have two consecutive accusatives. oculi, which is also very poorly
supported, is preferred by Henry (and Hirtzel and Sabbadini), who objects
that nostri used absolutely is unparalleled in Virgil: but so is nostrorum
A. 2.4rr, and the emphasis which nostri ... videre ... oculi puts on the fact
of being eye-witnesses is not nearly so apt to the context as the fact of
Cacus' confusion (which also precludes taking oculis with videre = Homer's
,Sov d</,9a>.µ,o'ia,, etc.).
223 ilicet, derived from ire licet (cf. sci(re)licet and vide(re)licet),
originally meant 'you may go, the business is over' and was used as a
formula of dismissal in law-courts, at funeral ceremonies, etc. The secon-
dary meaning 'immediately' was probably helped by the similarity to ilico
(in loco, 'on the spot') which is not tractable in hexameters. The history of
the word is fully examined by S. Timpanaro, RFIC gr (1963), 323 ff.
Leumann 153 regards Virgil's use of the word as artificial and pseudo-
archaic. See also Austin's note on A. 2.424.
225-30 This long period is constructed hypotactically, with subordinate
 COMMENTARY 85

clauses, more in the manner of Lucretius than of Virgil: contrast the

period in 251-5, in Virgil's normal paratactic manner.
225 ff. Cacus has something of Homer's Cyclops Polyphemus in him: ...
a1ravw8€v £Ci>V a8€fLLO'TLa if871 (Od. 9.189); he too blocked the entrance to his
cave with a huge boulder-which twenty-two four-wheeled waggons could
not have moved (Od. 9.240 ff.).
226 f. ferro quod et arte paterna / pendebat: literally 'which was
hanging by an iron chain and by his father's craft': a slight instance of
zeugma, the concrete noun ferro and the abstract arte (both ablatives of
instrument) being coordinated by et with no intervening words to mitigate
the harshness. In English there is always an element of intentional
burlesque in this construction: 'Mr Weller then took his hat and his
leave.' In Latin when the humour is not made quite obvious by puns or
word-play (as at Cic. Phil. 13.24 cum in gremiis mimarum mentum men-
temque deponeres), the Romans do not seem to have found the construction
funny (or at any rate not to have avoided it in contexts where burlesque
was out of place), cf. Cic. Cat. 2.n et in urbe et in eadem mente permanent;
and in poetry, where it is more frequent, A. 2.654 inceptoque et sedibus
haeret in isdem; Ovid A .A. 1.551 et color et Theseus et vox abiere puellae.
See further LHSz 2, 832 f.
227 pendebat: a molossus at the beginning of the verse and preceding
a pause has a slow massive ponderous rhythm, which is often a counter-
support to the meaning of the word, as here, and with instabant 434,
fundebat 584. See also the note on 147 insequitur.
227 fultosque emuniit obice postis: fultos may not be strictly
proleptic but refer to a contemporaneous action (see the note on 37
aeterna); translate 'and (after) he had fortified the gate-posts by shoring
them up with the barricade' (of the boulder, saxum 226).
227 obice is the only form of the noun used by Virgil in the Aeneid:
at A. 10.377 and n.890 it also occupies the dactylic fifth foot; it is derived
from the verb obicere which was originally ob-iacere, then passed through
obiecere to obiicere, and finally was regularly spelt obicere. All of Virgil's
uses of both noun and verb have the first syllable long, as it correctly was,
because of the combination b and semi-consonantal i in iacere (whether
this latter was actually written or not). Later poets (less sensitive to
etymology?) allow the first syllable to be short, cf. (noun) Silius Italicus
4.24 et fidos certant obices accersere silva; (verb) Lucan 8.796 cur obicis
Magno tumulum manesque vagantis? The vowel-sound o in ob- is by nature
short and remains so whether the syllable is long or not: Romans who
pronounced the o long are corrected by Aulus Gellius N.A. 4.17.1 f.,
7 f., IO f. (see C. G. Cooper, An Introduction to the Latin Hexameter 17).
Only at A. 10.377 is the gender of obice made explicit, by the addition
of magna (but Servius knows of a variant magno); on the other hand at
Tac. Ann. 13.39 obices portarum subversi, and Hist. 3.30 ferrati portarum
obices, the word is certainly masculine. Similar fluctuations in gender
occur with pumex and cortex, and may be due to the influence of the
gender of competing synonyms in either Latin or Greek. On the whole
86 COMMENTARY

question see Klotz, RhM Bo (1931), 342 ff., and compare the note on
silex 233 below.
228 f. 228 is the one hypermetric line in this book (see Appendix, p. 193).
The metrical artistry of 227 ff. is noteworthy: Cacus blocked the entrance
with ease (in the 'falling' rhythm of homodyned dactyls, emuniit obice
pastis); Hercules arrived with lightning speed and pounding energy (in
'rising' anapaestic rhythm with clash of ictus and accent, eccl fiirens
dnimis dderiit), but was abruptly halted by the blockade and slowly surveyed
the means of getting in (omnemq(ue) iiccessum lustriins, in heterodyned
spondees). The hypermetre solders the two lines together, and produces the
sudden halt in the rhythm after Tirynthius instead of at the end of the
line.
228 animis: here and at 256 animi = 'passion, anger'.
228 Tirynthius: Hercules; he was born in Argos, but Servi us (on
A. 7.662) says he was brought up (nutritus) in Tiryns. The accepted
version, in so far as there is one, knows of his association with Tiryns only
after the Pythian priestess advised him to settle there while performing the
labours for Eurystheus, king of Argos (a localising of the story which may
reflect the early history of Tiryns' subjection to Argos).
229 lustrans: lustrare originally meant 'to purify'; because purifi-
cation was often effected by going round a place or person in procession,
completing the magic circle to keep good influences in and bad ones out,
it came to mean 'to go round' (as in 231), or 'to go over, traverse';
associated with oculis, etc., it then acquired the meaning' to go over with
one's eyes, to scan' (as in 153 lustrabat lumine), and eventually it came to
have this sense without any qualification (as here).
Translate: 'and scanning every way of approach kept turning his face
this way and that, gnashing his teeth'.
230-2 A skilfully constructed period: three clauses of diminishing
length, linked by the initial anaphora of ter, are 'pegged' on to the hexa-
meter frame by the word-pattern formed by placing the verbs in the
emphatic positions of the verse-beginning and end (lustrat II temptat,
temptat II resedit). See Appendix, p. 200, for further illustration.
231 A ventini montem: in geographical descriptions classical Latin
prose allows only the apposition in the same case of the common noun and
proper name, ad urbem Romam not ad urbem Romae. The 'appositional'
genitive (as in the second example) did not come into vogue until
Augustan poetry and prose, the chief impetus being given by Virgil,
e.g. A. 1.244 fontem ... Timavi; 247 urbem Patavi. This usage has clear
affinities with the genitive of definition of the type nomen servitutis, 'the
name of slavery', i.e. 'the name "slavery'". But the possessive genitive
may have contributed something if, as seems possible, the proper name
was thought of as personalised (lacus Averni, 'the pool of the god Aver-
nus'; see Servius on G. 2.164). Here it is quite likely that Virgil means
'the hill of (the hero) Aventinus ', the son of Hercules (prominently men-
tioned in the Catalogue of Latin Allies, A. 7.655 ff.).
 COMMENTARY 87

233-246 Hercules wrenches away the ridge of rock over the cave, and the
cave's hellish depths are revealed.
233 stabat acuta silex . .. 236 bane: for this simple variation on est
locus opening an lKef,paai, T61rov see the note on 597 est ingens ... lucus.
233 acuta silex: like obex (see the note on 227), silex fluctuates in
gender between masculine and feminine: where it is qualified in Virgil it is
always feminine (E. r.15; A. 6.471 and 602, as well as here), and Servius
here is probably right in regarding this as a Virgilian innovation: paene
omnes hunc silicem dixerunt: nam et Varro et Lucretius ita dicunt. tanta
tamen est Vergilii auctoritas, ut persuadeat nobis etiam hanc silicem dici. In
the examples in the Aeneid cited above it is only the nominative silex
which is qualified, by an immediately preceding adjective which has to be
feminine because of the metre. Virgil no doubt thought that such a
variation in gender was a traditional epic licence: in Homer >..l8o, is
masculine and feminine; lapis which is generally masculine was made
feminine by Ennius (Ann. 553 V = Nonius 211.9 tanto sublatae sunt
a(u)gmine tune lapides); silex is virtually synonymous with lapis, and is
often found joined with it (e.g. Plautus Poenulus 291, tu es lapide silice
stultior) in Old Latin. The same chain of influence ultimately from Greek
practice may account for the exceptional arida modo pumice at Catullus 1.2.
235 nidis: 'nests' (cf. G. 4.307 for the singular in this sense) or
'nestlings' (cf. G. 4.17), or both.
236 f. laevum ... (ad amnem) ... dexter: from whose point of view?
Clearly not, it is argued, that of the narrator Evander at the Ara Maxima,
who, looking south to the Aventine, would have the river on his right.
Commentators ancient and modem have been exercised to make topo-
graphical sense of the passage. If Virgil had a precise picture in his mind,
which is not certain, the viewpoint is apparently that of someone between
the southern part of the Aventine and the river, looking roughly north-east
(Hercules himself, according to Warde Fowler, ASR ad loc.-but then only
if his lustratio of the Aventine was clockwise, and dexter applied to
Hercules from the point of view of his own previous position is possible).
Further, in adversum nitens would naturally = 'against the direction of
the slope', which makes nonsense of dexter: hence Bentley's conjecture
in aversam, palaeographically very attractive-aversi in tergum Sulmonis
is certainly right at A. 9.412 against adversi of all the principal MSS.
Adjectives of direction are peculiarly liable to hypallage (cf. the note
on 57). It is just conceivable that the viewpoint is Evander's, that laev-
really belongs to silex or iugo: the silex rose from a ridge on the left (east)
of the Aventine, sloping towards the river: Hercules did heave against
the direction of the slope, from the right (west) side.
The form of the sentence hanc, ut . .. incumbebat . .. , concussit which
repeats that of 209 ff. is not common in Virgil; but cf. A. rz.488 ff. and
623 ff.
237-9 Notice how the heterodyned spondees in adversum nitens ...
avulsam solvit suggest the slow painful effort required to make the rock
88 COMMENTARY

budge; then suddenly it gives way quickly and easily in the homodyned
dactyls of -dicibus inde repente, and Hercules' force spends itself rapidly
and effectively in the stopped first foot dactyl impulit (see the note on
87 leniit). The situation, rhythm and phraseology recall A. 2.464 ff.
(turrim) convellimus altis / sedibus impulimusque; ea lapsa repente ruinam /
cum sonitu trahit et Danaum super agmina late / incidit.
239 f. Critics are divided on the question of whether or not the rock
torn off by Hercules actually fell into the river. If it did, 240 is a very
forceful way of representing the natural consequence; if it did not, 240 is
an exaggerated piece of pathetic fallacy, with personalised reactions of the
banks and the stream. The second view gains support from A. 9.124 f.
where in terror at the portentous transformation of ships into nymphs,
and with no physical cause, cunctatur et amnis / rauca sonans revocatque
pedem Tiberinus ab alto.
241-58 Another excellent illustration of the 'theme and variation'
technique. The narrative proper hangs on the themes alone; the elaborated
variations give it depth and colour.
242 regia, with undertones of despotism, like arx and aula (cf. A. r.140).
penitus . .. 243 penitus at exactly the same position in the line;
perhaps, as Page maintains, an awkward oversight. Roman poets were on
the whole less sensitive to such repetitions at short intervals than English
poets; Housman (Lucan, pref. 33) remarks 'Each author has his own
principles and practice. Horace was as sensitive to iteration as any
modern ... Virgil was less sensitive, Ovid much less; Lucan was almost
insensible.' But the repetition here is almost too blatant to be unintended,
and it has rhetorical point: the subject and the simile are provided with a
common element and so united in sharper comparison. Cf. the line-endings
at 271 f. and 396 f. and the notes there. (It is remotely possible that
penitus 242 derives from that word in the line below and was substituted
for the true reading; in this case Schrader's late from Ennius Ann. 440 V
tum cava sub monte late specus intus patebat is as good a guess as any.)
243 (si) qua ... vi: better taken together, 'if by some (unknown)
force', than = £Z 1Tov (1Tw,) . .. f3lq., 'if somewhere (-how) ... by force'.
244 f. Virgil has taken a hint but little more from Homer Il. 20 ( Brnµ,axla)
61 ff.: in the battle of the gods Poseidon shook the earth so that the king
of the dead was afraid he might tear the earth apart and reveal the lower
world: i8nu£V •.. / .. . µ,~ o[ V1T£p8E / ya'iav dvapp~gHE IloanSawv lvoulx8wv, /
• ' 0£
o,K,a "' 8VTJTO£U£ Kat a'8 ava-ro,u,
A ' ' .,,avn71
,I. " \ ,, EvpwEv-ra,
' I uµ,EpoaAE , ' -ra' TE UTvyEovu,
I

8Eol 1TEp.
Macrobius 5.16.12 ff. notes that Virgil has transferred material from
Homer's direct narrative to a simile, dissimulanter imitatur.
245 super: generally taken = desuper, but 'the vast pit descried from
above', i.e. from the upper world, by mortal men, introduces an extra and
unnecessary element into the simile; super = 'above' (them), as seen by
the underworld and its occupants underneath, coheres better with the
following Manes, and keeps attention on Cacus' similar situation.
246 trepident: the asyndeton is not harsh enough to be suspect, and
 COMMENTARY 89

Wagner's trepidentque is not necessary. It is generally assumed that

trepident is coordinate with the preceding subjunctives; it could however
be in apodosis to them, non secus ac . .. trepident . .. Manes si ... terra . ..
infernas reseret sedes. The order of ideas in 241 ff. is chiastic: Cacus' cave
uncovered: the Underworld : : the ghosts of the Underworld: Cacus.
246 Manes means the indiscriminate mass of the spirits of the dead;
primitive Roman thought regarded them collectively, usually as a potential
menace; the word only exists in the plural and this rather illogically came
to be applied to individuals in funerary inscriptions from the last decades
B.c. onwards, e.g. GIL I. 22 .761 Dis Manibus sacrum L. Caecilii Rufi.

247-267 Hercules hurls missiles into the cave, and when thetrapped Cacus
fills it with smoke and flame Hercules leaps on to him and strangles him. The
stolen cattle are brought to light, and Cacus' monstrous corpse is exposed to
view.
248 rudere, regularly used of loud animal noises (especially the
braying of asses), suits the bestial half of semihomo Cacus.
248 insueta rudentem: 'bellowing as never before'; cf. 489 infanda
jurentem: accusative neuter adjectives used adverbially are an offshoot of
the use of the internal (limiting) accusative. Native Latin idiom confined
itself in this usage to adjectives of quantity only (multum, nimium, ali-
quantum, pauca, cetera, etc.), and only these are found in good prose.
Poetry extended the usage to other adjectives, clearly under Greek
influence: the singular adjective so used adverbially was established by
Catullus (51.5 dulce ridentem, a translation of Sappho's y€Aalaa, lµlpo€v);
the plural was launched by Ennius (Ann. 342 ululat ... acuta, if the reading
is sound), and continued by Cicero (in his translation of Aratus, frg. 26
truculenta tuetur) and Lucretius (e.g. 5.33 acerba tuens). See further
I<S 1, 279 ff., LHSz 2, 40.
250 molaribus, 'stones the size of mill-stones'. In the duel between
Ajax and Hector in Iliad 7, Ajax smashed Hector's shield with a' stone the
size of a mill-stone' (270 µ,v>.onUi 1rlrp<tJ), and in Il. 12.160 f. the helmets
and shields of the combatants clattered as they were pelted with such
stones (µ,vM-Kma,). Homer's heroes are endowed with superhuman strength
(as befits their superhuman size, see the note on 162): Diomedes (Il.
5.302 ff.) found it easy to hurl a stone which two men of Homer's genera-
tion could not have carried; and Pandarus (Il. 4.109 ff.) drew a bow which
no mortal man could have drawn (see Leaf and Bayfield's note ad Zoe.).
Virgil continued the exaggeration: Mezentius (A. ro.698) strikes his
victim ingenti fragmine mantis. In the present passage Hercules, the son of
Jupiter, possesses superhuman strength in his own right, and the epic
convention of exaggeration is not inappropriate. The same is not true of
Acmon, an ordinary Trojan warrior at A. ro.127 f.: Jeri ingens toto conixus
corpore saxum, / haud partem exiguam mantis, Lyrnesius Acmon. The process
by which he comes to hurl 'no small part of a mountain' is a fascinating
illustration of Virgil's association of ideas. At Od. 9.481 the giant Cyclops
90 COMMENTARY

'9K€ a· a1roppiJga, Kopvcf,~v op€o<; 1uya.AOLO; in A. 8 (see the note on 425)

Virgil mentions a number of Cyclopes, to three of whom he gives significant
names, one of which is Pyracmon ('Fire-anvil'). He remembered this
coinage of his, and decided that the Trojan warrior who performs the
Homeric Cyclops' feat should bear the Cyclopean name Acmon ('Anvil'),
while the adjective Lyrnesius gives a feasible birthplace (Lyrnesus in
Mysia) to the fictitious hero.
251 The omission of est in the subordinate (or at least parenthetic)
clause introduced by neque enim is nothing unusual. It has to be 'under-
stood' and coupled with super (superest, 'is left, remains'); there is no
need to supply it materially as a section of the MS tradition has by
adding it to the end of the line.
253 caligine caeca, 'in blinding darkness'; caecus = literally 'blind'
does not occur in Virgil (although it does occur metaphorically in phrases
like caeci furore).
256 'In his fury Hercules could not endure this ... ' non tulit is a Virgi-
lian formula, see Austin on A. 2.407.
260 corripit in nodum com plexus: 'seized and knotted him in his
embrace', and crushed him, in the same way in which he had eventually
killed Antaeus, who could only be killed by breaking his contact with his
power-giving mother Earth: cf. Lucan 4.632 Herculeosque novo laxavit
corpore nodos, after a detailed description of Hercules' 'grip' of Antaeus.
nodus is a technical term of wrestling, 'a firm hold': see Dilke on Statius
Ach. 2.155. There is also an allusion to the nodus Herculis or Herwlaneus,
the knot invented by him and difficult to untie, good for tourniquets
(Pliny N.H. 28.(17).63), particularly associated with the costume of a bride,
but also applied, for example, to the closed circle formed by the snakes on
Mercury's caduceus (Macrobius 1.19.16).
Different accounts of Cacus' manner of death are given by Propertius
4.9.15 and Ovid Fasti 1.575 f.
260 In this line, and at line 595 armati tendunt; it clamor, et agmine facto,
there is a light trochaic pause in the fourth foot; this in itself is a great
rarity (see Williams' note on A. 5.166-7). But 260 is also one of the com-
paratively few examples in Virgil (there are only about 100 in the whole
Aeneid, see Austin on A. 4.58, and on A. 2.380), where a line contains a
trochaic caesura in both the fourth and fifth feet: the effect of this is to
produce coincidence of ictus and accent on the penultimate and ante-
penultimate words of the line, whose rhythm then suggests that the hexa-
meter has already come to a close: corripit in nod um complexus et
angit / inhaerens. This 'premature ending' seems to be the chief reason
for the general avoidance of lines with this rhythmic pattern (although a
similar impression can be given also by lines of a different pattern like,
for example, 275 communemque vocate deum et date vina / volentes;
or 653 stabat pro templo et Capitolia celsa / tenebat). There are other
instances in A. 8 at 34, 94, 189, 206, 226, 390, 433, 656. For further
discussion, see the references cited above.
260 f. angit . .. guttur: tr. 'strangled him until his eyes started from his
 COMMENTARY gr

head and his throat was drained of blood'. Literally = 'clinging to him he
throttles his dashed-out eyes and throat dry of blood'. The phraseology is
compressed and violent through the combination of two common figures:
(1) angit is strictly connected with guttur, only by zeugma with oculos
which however comes first; in many cases of this figure the 'surprise'
noun is less startling through following an expected one, cf., for example,
Val. Fl. 8.68 iamque manus Colchis crinemque intenderat astris; (2) elisos
and siccum are both proleptic (see the note on 37). Difficulty was felt
about elisos in antiquity: Servius defends it as a soloecophanes, an apparent
solecism; elidens was preferred by many scholars and has found its way
into the MS tradition, but after inhaerens another present participle is
of course stylistically impossible (even if separated by a comma).
ango and angor (noun) seem to be used exclusively of mental distress
before the Augustan period; here, in its only occurrence in the Aeneid,
Virgil imposes the literal physical sense (cf. Greek a'.yxw). For elidere = 'to
strangle', cf. the similar phrases in Ovid Her. 9.85 f., also of Hercules,
scilicet immanes elisis f aucibus hydros / infantem caudis involuisse manum;
Lucan 2.154 hie laqueo fauces elisaque guttura fregit.
263 Four-word lines of deliberate artistry occur occasionally in Virgil
in highly finished passages, cf. in its context G. 1.470 obscenaeque canes
importunaeque volucres; identical with the line here in rhyme, rhythm,
caesura, is A. 11.870 disiectique duces desolatique manipli.
263 abiuratae strictly means 'denied (or disclaimed) on oath'. Servius
comments: abiurare est rem creditam negare periurio. sed hoc isti loco non
congruit (for Hercules never entrusted the cattle to Cacus, nor has the
narrative given Cacus any opportunity of denying that he did possess
them). There are two possible lines of interpretation:
(1) Virgil is using the word loosely in its legal sense to mean' disclaimed
on oath' without any implication that the cattle had been entrusted to
Cacus (this is in any case ruled out by rapinae). In this case we must
follow Heyne and Conington in believing that it 'refers to a disclaimer of
Cacus not mentioned, but easily understood'. In fact Virgil will be ex-
pecting the reader to supply the narrative with a quite important detail
which he has not made explicit, but which is specifically stated in his
underlying source, for Hermes does deny on oath that he has stolen
Apollo's cattle, Hymn to Hermes 274 ff. £i 8£ 0lAHs TTarpos K£cpaA~v 1.dyav
opKOV oµ.ofJµ.a, · I µ.~ /J,€V iytil µ.~r' avros tl7Tlaxoµ.m airios Elva,, I µ.~T€ riv'
aAAov oTTwTTa {3owv KAoTTov vµ.£r£pdwv. This would be quite in Virgil's manner
(see the note on 531).
(2) But it is not in Virgil's manner to use words loosely. The abundant
evidence of his interest in etymology suggests that Servius auctus is right
in paraphrasing abiuratae by alieni iuris factae: the 'rights' (ius) of the
cattle belonged to Hercules, but these had been taken away from him
(ab-) and usurped by Cacus. In this case Virgil is using the word according
to the basic meaning of its compounds, and in so doing has changed the
accepted sense. This is no doubt part of the nova cacozelia he was charged
with.
8 EAC
92 COMMENTARY

Translate: 'the plunder he had appropriated'.

265 nequeunt expleri corda tuendo: theoretically corda could be the
subject of nequeunt, 'their hearts cannot be filled with gazing'; but in view
of the extremely similar phrase at A. 1.713 expleri mentem nequit ardescit-
que tuendo, it is better to regard corda as the accusative direct object of
expleri which is being used in a middle (or semi-reflexive) sense: 'they (the
onlookers) cannot fill their hearts with gazing'. The imposing of this
middle sense on passive non-participial forms of verbs was much favoured
by Augustan poetry under Greek influence: it is an extension of the type
of construction found at 457 tunicaque inducitur artus (see the note there).
(It is also possible to regard expleri as truly passive, and corda as an
accusative of respect, for this type of expression undoubtedly contributed
to the variety of forms exploited by Virgil: see the note on 29 turbatus
pectora.)

268-279 From that time there has been an annual sacrifice to Hercules at
the Ara Maxima. Having.finished his account, Evander leads the company in
pouring a libation to Hercules.
268-72 After the smooth well-ordered Cacus-narration, this section is
abrupt and incoherent. Apart from style, it contains three serious diffi-
culties:
(1) minores (268) = 'descendants, posterity' (infrequent and poetic as a
substantive in this sense) is obviously anachronistic in the mouth of
Evander; Servius was puzzled, and Servius auctus' stress on Evander's
longevity is not a convincing explanation, even though Virgil did model
him after Nestor. Nor can Virgil be speaking in propria persona in the
middle of Evander's speech. (Contrast 314-36 with 337 ff., the absence of
anachronism in Evander's own words against the anachronistic comments
of the poet when merely reporting Evander's words.)
(2) statuit (271) lacks a clearly stated subject if there is heavy punctua-
tion at the end of 270 (correctly; to make domus Pinaria, which is merely
the guardian of the rite, into the subject is impossible); see the note on 271.
(3) The repetition of the same phrase, quae maxima semper (271 f.) at
the same position of two successive verses is an unparalleled anomaly in
Virgil, different both from the repetition of a single word at the end of a
line (see the note on 396 f.), and from the artistically contrived repetitions
of phrases in Ovid (e.g. Met. 1.325 f., and 361 f.) and in the Virgil of the
Eclogues (8.48 ff.) where there is rhetorical point: here, in spite of Servius
auctus and Mackail (Aeneid, Intro. 80), it is difficult to see anything but
clumsy emphasis. The lines as we have them are almost certainly a sign
of the unfinished state of the Aeneid, most probably a tibicen, a temporary
prop, or a versified marginal note, which Virgil would have changed or
expanded later; they were (characteristically) deleted by Peerlkamp.
Bomer's attempt to solve some of the problems by excision and transpo-
sition (RhM 92 (1944), 363) is quite unconvincing.
269 diem appears to refer to the day of Hercules' victory over Cacus;
 COMMENTARY 93

in fact, celebration took place at a shrine or temple on the anniversary of

the day of its dedication; for the Ara Maxima this was probably 12 August.
Such a date does not admittedly :fit in well with the overall chronology of
the Aeneid, for Aeneas spent only a short time in Sicily and Italy after his
departure from Carthage at the end of winter, but if Virgil attached precise
significance to this date he may have remembered it primarily as the eve
of Augustus' triumph in 29 B.C. (see the note on ro2).
269 f. 'Potitius founded the rite of Hercules in the :first place, and the
family of the Pinarii preserved it.'
The traditional story, told by Servius and Servius auctus here, and
Livy r.7.12 ff., was that after slaying Cacus Hercules, recognised as a
divinity, instituted sacrifice to himself and taught the procedure (that
there should be a morning and an evening sacrifice) to two old men
Pinarius and Potitius. On the :first occasion of the evening sacrifice,
however, Potitius arrived in time to officiate, Pinarius only after the exta
had been distributed, and Hercules in anger confined Pinarius and his
family to the function of ministri. The two families continued in charge of
the rite until Appius Claudius bribed the Potitii to surrender it to servi
publici, as a result of which he himself became blind (Caecus) and the
Potitii soon became extinct.
The historical facts behind this story are very difficult to recover beyond
the basic one that the cult, like many others in early Rome, was originally
very much a family concern of patrician genies, and was eventually taken
over by the State, probably in the censorship of Appius Claudius (312 B.c.),
'one of a family of pestilent innovators' (H. J. Rose, ARR 92). There is
no trace of the gens Potitia by the end of the Republic; the gens Pinaria
still survived, and so did the memory of its association with the rite (Cic.
Dom. 134 on L. Pinarius Natta, Clodius' brother-in-law), but hardly with
any high standing or prestige. So it would be quite natural for the State to
adopt the cult and conserve it (Appius Claudius looks like the victim of
party propaganda). In historical times the praetor urbanus was the offici-
ating magistrate, taking the place of the gens Potitia, and servi publici
(who could not of course hold a priesthood) were the ministri looking after
the material needs of the cult: this humbler function seems to have been
projected back on to the Pinarii. The two families, as co-patrons of the
cult, are likely to have shared all responsibilities equally; the difference of
function, mentioned by Virgil, was suggested partly by folk-etymology,
Potitius--potitus, 'in possession of' (auctor), Pinarius-1T£tvav, 'to go
hungry'. See further Wissowa, RuK2 274 f., Latte, RRG 213 f., H. J. Rose,
ARR 92.
271 statuit: the subject appears to be Potitius auctor (269); other
authorities, however, are divided between ascribing the foundation either
to Evander or to Hercules himself (see Wissowa, RuK 2 273, n. 6, who
believes that in the original myth Evander was named, and Hercules
later adopted because of his connection with the founding of the altar of
Jupiter Inventor). The subject here could be Hercules (for abrupt changes
of subject see the note on 546 graditur). Mackail held that the passage here
8-2
94 COMMENTARY

is incomplete and unrevised, and was attracted by Pomponius Sabinus'

proposal (Commentarii, Basel (1547), 1307) to read statui (but this attempt
to regularise the story will not hold up against the unanimous reading of
the MSS).
A careful examination of the whole tradition can be found in J. G.
Winter, The Myth of Hercules at Rome, University of Michigan Studies,
Humanist Series IV (1910), 171 ff.
273 tantarum in munere laudum: 'in celebration of such great
exploits'. munus basically refers to the discharge of a duty, here and
A. 5.537, 6.637 a religious duty; munus can also mean a public show
which it was a magistrate's duty to provide. laudes often = 'deeds which
merit praise' as here.
274 porgite: by syncope (see 73 periclis) for por(ri)gite: the only other
occurrence of a finite form of this verb in Virgil is porrigitur A. 6.597. The
syncopated form occurs in Old Latin (e.g. Plautus Pseudolus 708), and
porgite here is specifically said to have been taken from Ennius by Servius
(on A. 1.26 alta mente repostum: repostos et porgite de Ennio transtulit); it
may have had an old-fashioned ring about it, but was no doubt maintained
in popular speech (cf. Italian porgere): perrigo had given way to pergo before
the time of the earliest written documents. porg- forms were taken up by
the Silver Latin epic poets (see the r~ferences in LS s.v.).
On porgite dextris Servius comments et est Homeri: possibly a recasting
of Il. 9.670 f. TO v, fl,£11 a.pa xpvalo,a, I( V 7TO,.>,o' > vlE, 'Axaiwv I OHO' xa-r'
a),>..oBEv a>..>..o, dvaa-rao6v.
275 communem . .. deum: ' the god of us both ', Greeks and Trojans, now
in alliance; similarly at A. 12.n8 Rutulians and Trojans prepare altars
for the di communes to ratify the general truce during the duel of Aeneas
and Turnus. In both contexts Virgil is thinking solely of the necessity of
having a commonly recognised god for the solemnity to be available and
binding for both parties (cf. Propertius 1.n.16 perfida communis nee
meminisse deos): what gods the Trojans and a primitive Italian tribe
could in fact have had in common is difficult to see; and here there is no
reason to believe that the Trojans were particularly devoted to Hercules
who is commemorated in the Salian hymn (290 f.) as the destroyer of
Troy. To interpret communemque vacate deum = 'and summon the god to
share in the feast' is prima facie very attractive, primitive belief holding
that gods were actually present at sacrifices in their honour; but the idea
of 'contributing to', often present in communis (as in Greek avµ,/30>..-; and
cf. immunis), would not be appropriate here.
275 volentes, 'with a will', because reluctance or constraint hindered
the act of worship-a surprising sidelight on the psychological demands on
the worshipper in Roman religion, whose outward forms are generally cold
and contractual; cf. Turnus' prayer at A. 10.676 ff. vos o potius miserescite,
venti; / in rupes, in saxa (volens vos Turnus adoro) / ferte ratem saevisque
vadis immittite syrtis. Similarly libens often occurs in inscriptions recording
the discharge of a vow, in the formula V(otum) S(olvit) L(ibens). laetus is
used in the same sense at A. 5.236, 8.279, 8.544. 'Volition' was also a
 COMMENTARY 95

prerequisite for the action of the deity answering a prayer, cf. Livy r.16.3
pacem precibus exposcunt uti (Romulus) volens propitius suam semper
sospitet progeniem.
276 Herculea bicolor . .. populus umbra: note the schematism; this
tendency to organise balanced pairs of nouns and adjectives is distinctive
of Latin poetry; it is not noticeable in Greek, was inaugurated, as far as
we can tell, by Cicero in his translation of Aratus, and much developed by
Catullus in the Epyllion (poem 64). See T. E. V. Pearce, 'The Enclosing
Word Order in the Latin Hexameter', CQ N.S. r6 (1966), 298 ff.
The poplar was often associated with Hercules, cf., for example,
E. 7.61; G. 2.66; Ovid Her. 9.63 f., and became one of his symbols: on
Horace Odes r.7.23 where Teucer, fleeing from Telamon and Salamis,
tempora populea fertur vinxisse corona, Kiessling-Heinze 8 rightly point out
that he does this because he intends to sacrifice to Hercules ~yEµ.wv, whose
guidance he needed in his uncertainty about his objective. Servius (here
and on E. 7.61) says that Hercules on his way to Hades, weary with his
labours, made himself a poplar garland and returned to earth still wearing
it, the lower side of the leaves made white by his sweat, the upper dark from
contact with the underworld.
In Roman sacrifice the priest or magistrate in charge covered his head
from a curious ostrich-like belief that if he could not see any bad omens
there could not be any (at A. 3.403 ff. the custom is attributed to Helenus'
advice to Aeneas); one of the Greek features of the cult at the Ara Maxima
in historical times was that the presiding praetor sacrificed with un-
covered head (Servius auctus on A. 8.288; Macrobius Sat. 3.6.17). velavit
is ambiguous; it is the word used in connection with covering the head at
A. 3.405, but foliis innexa suggests a garland, not a covering and veto is
elsewhere also (A. 7.154, A. n.ror) weakened in sense to mean 'deck,
adorn'. Servius (on A. 8.276) tells us that sacrificers at the Ara Maxima
wore garlands of laurel, and whether or not Virgil has here made a
mistake is the topic of erudite discussion in Macrobius Sat. 3.12.r ff.,
where he is defended by the authority of a work of Varro now lost, to
the effect that laurel was only used at the Ara Maxima after the laurel
grove had been planted on the nearby Aventine, which was long after the
founding of the city, and so would have been anachronistic in Virgil's
account of Evander. Virgil may conceivably have read this in the second
book of Varro's Res Humanae, but the association of the poplar with
Hercules in myth is enough to explain his choice.
276 f. dixerat . .. cum . .. / velavit: for the cum inversum construction
in general see the note on 26 ff. Here and at A. 5.84 ff. dixerat haec,
adytis cum lubricus anguis ab imis / ... volumina traxit, the sense of
dixerat is that of its' absolute' use (see the note on 152) sharpened by the
immediacy of the following context; translate 'he had just finished
speaking, and then .. .'.
278 et sacer implevit dextram scyphus: the goblet (scyphus is a
Greek loan-word = Latin poculum) was holy and huge. Servius had read
in libris antiquis that Hercules had brought a huge wooden goblet to
96 COMMENTARY

Italy, which was smeared with pitch to prevent it rotting and used in
sacrifices: some such vessel was no doubt part of the religious furniture at
the Ara Maxima. Macrobius Sat. 6.21.16 also mentions scyphus as peculiar
to the rites of Hercules. A scyphus is a votive offering to Hercules in CIL
52 .6952. The 'goblet of Hercules' seems also to have become proverbial
in connection with heavy drinking: cf. Plutarch Alexander 75 aKv</,ov
'HpaKMovs eKm£'iv and Seneca Epp. 83.23 intemperantia bibendi et ille
Herculaneus ac fatalis scyphus condidit (sc. Alexandrum).
279 Why is the libation poured in mensam and not in aram? The
sacra supellex of every temple included tables for sacrifice (cf. A. 2.764),
braziers, vessels, etc., all of which were of course consecrated. Such tables
could be used instead of the altar (Macrobius Sat. 3.n.5 ff. and Festus
p. 149-Lindsay (1913)). There are two reasons which make the substitution
appropriate here: in primitive times and later the gods were believed to be
actually present at the meal in their honour (cf. Ovid Fasti 6.307 ff. (of
the cult of Vacuna) ante focos olim scamnis considere longis / mos erat, et
mensae credere adesse deos); secondly, the worshippers at the Ara Maxima
sat but did not recline, and Hercules there had a mensa, not, like other
gods, a lectus (see Wissowa, RuK 2 281, n. 3). The Ara Maxima cult of
Hercules, without lectus or lectisternium, probably dates from the fifth
century B.C.; the first attested lectisternium for Hercules is dated 399 B.C.,
and this later cult was observed in temples near the Circus Flaminius.

280-305 Evening comes on, and while the feasting continues a double
chorus sings a hymn to Hercules, commemorating his victories over a series of
monsters which culminates with Cacus.
280 'Meanwhile the Evening Star draws nearer down the sloping sky.'
devexo, 'sloping downwards' is equally appropriate to a star or constella-
tion (Orion, Hor. Odes r.28.21) sinking in the western sky, or to that part
of the sky to which it was regarded as permanently fixed sinking to the
horizon. The Evening Star was actually the planet Venus and so not fixed,
but that is not in point here.
282 pellibus in morem cincti: in historical times the only religious
ceremony at which animal skins were worn was the Lupercalia, and the
Luperci were not officiating priests. Virgil was probably thinking of
Hercules' lion-skin, and suggesting that the rite was a very ancient one;
what historical justification he had for dressing the priests in skins we do
not know.
283 f. instaurant here (and also most probably at A. 7.146) does not
have its usual meaning that they begin all over again a rite which has been
nullified by bad omens, for here that is not the case. They 'make a fresh
start' and resume the feasting after some lapse of time: one of the pecu-
liarities of the cult at the Ara Maxima was that sacrifice was offered both
in the morning and in the evening (so Servi us on A. 8.269). But looked at
in terms of a usual banquet (see the note on 184 ff.) the second sacrifice
and feast could be regarded as the second course (of a very protracted
 COMMENTARY 97

meal), hence mensa secunda (Roman practice was to remove the small
individual tables after they had been soiled by the first course, eaten
without knives and forks, and to replace them by fresh ones for the dessert
and wine of the second course). In 276-9 there is a ceremonial libation to
Hercules to end Evander's story; evening comes on (280), and the regular
evening meal then follows (283 f.): this all suits the present situation;
A. 7.146 f. is not parallel.
There were other unusual and unRoman features of the cult: the
general public was invited to participate (for Hercules was regarded as a
protector in a very personal, individual way), but only the male public,
women were excluded-and so are never represented by correct authors as
using the oath me hercle; anything edible could be offered for the sacrifice
and the feast, but it was prohibited to take home away from the sacred
enclosure anything that remained over (a part of the Graecus ritus), and
this was burnt. dona ferunt therefore refers to the offerings of all who were
present, not exclusively to the Salii and the dapes Saliares of Horace Odes
r.37.2 ff.; and the lances oneratae might have contained almost anything
(Servius quite wrongly says exta reddunt, as though no more time had
elapsed than at an entirely Roman sacrifice between the slaying of the
victim and the burning of the exta on the altar).
285 ff. Singing to honour and invoke the god to whom sacrifice has been
made is heroic practice, cf. fl. r.472 ff., where Apollo is hymned, as he is
also by the Argonauts at Ap. Rh. 2.701 ff. Virgil may have had this
passage at the back of his mind when he wrote this hymn to Hercules:
they have one important structural feature in common. Virgil (293)
introduces an abrupt change from third person reporting to the more vivid
and engaging plane of second person address. Similarly in Apollonius,
after Orpheus has sung of a youthful victory of Apollo over a monster, the
god is addressed directly (708 f.) ;;>,~KOL!,. al€{, TOL, ava~, aTµ'T}TOL lfh,pai, I
aUv dS~A7JTOL · Tw, yap 81µ,,. Servius, not in his commentary, but as the
speaker in Macro bi us Sat. 6.6, admired the change as a mutatio elegantissima,
but was wrong to classify it with figurata non a veteribusaccepta.
285 Virgil associates Salii with the cult of Hercules at the Ara Maxima,
and it is quite clear from his mention of the cult title (invicte 293) and the
general character of the hymn, that he is thinking of Hercules as a god of
war. There is no other evidence of the' Dancing Priests' being involved in
this cult at Rome, where they were at various times in tutela I ovis, Martis,
Quirini (Servius on A. 8.663). There is evidence, however, of their con-
nection with the cult of Hercules at Tibur: did Virgil model his account
of the Roman cult on this without historical justification, or is he echoing
(or guessing) a true account not preserved anywhere else?
In the evolution of Roman religion, rituals of various forms came into
existence long before the concept of anthropomorphic gods; and in
historical times the ritual of the Salii was a symbolic copy of the primitive
war-dance, performed in March and October (the beginning and end of the
campaigning season). How they were selected is unknown, but the evidence
of inscriptions shows that they were young men (see Latte, RRG n5, n. 1)
98 COMMENTARY

(so Virgil's semi-chorus of elders (287) is suspect). They were obviously

available to be attached to any warlike god.
Knowledge of Hercules infiltrated gradually into Latium, directly
through the Greek settlements of southern Italy and indirectly through the
Etruscans. Unlike the cults which were introduced into Rome under
State auspices and so formalised from the start, Hercules brought with
him many associations of myth. The chief feature of the Ara Maxima cult
was the ex voto dedication to him of tithes of profit from commercial and
military enterprises; by the second century B.C. military dedications were
by far the more important, but there is no reason to believe that they were
so previously. In fact, the localisation of his cult near the forum boarium,
the trading centre of the earliest settlement, strongly suggests that it was
Hercules' reputation as a traveller and benefactor that attracted most
attention in early times. (Current theories suggest that Hercules was
assimilated to a preexisting cult, instituted by Phoenician traders, of their
god Melqart; see J. Heurgon, ]RS 56 (1966), 2 f.)
At Tibur, where Hercules was patron god of the settlement, his cult was
also served by Salii and also included dedication of tithes (decumae);
the interplay of influence between the two cults cannot be reconstructed;
it has been suggested (see Jean Bayet, Les Origines de l'Hercule Romain
(Paris, 1926), 322 ff.) that after 338 B.C., the date of the 'peace' between
Rome and Tibur, the cult of Jupiter Invictus was transferred from
the latter to the former, and with it the associated cult of Hercules In-
victus, with the emphasis on his military character and his Salii. This
is not improbable but cannot be proved. The complete lack of any cor-
roboratory evidence for Salii being in the service of Hercules at Rome
after he had established his military character there argues against its
historicity.
The truth of Virgil's account was called into question as early as
Macrobius (Sat. 3.12); he is defended there by the remark that Hercules
Invictus and Mars were often held to be the same (as is true), but the
identification was of course made on the basis of secondary developments
and throws no light on the origins of their cults.
See further Wissowa, RuK2 272 f., Latte, RRG 215 ff. and the references
there. Allegorists (Drew and Bellen, see the references in the note on 104)
explain the connection of Salii with the Hercules of the Ara Maxima as an
allusion to the inclusion in 29 B.C. ex senatus consulto of Octavian's name
among the gods to be invoked in the carmen Saliare, the nearest conces-
sion he made to allowing himself to be worshipped in Rome: an attractive
explanation, because as yet the only one. W. A. Camps, 98-104, and Ap-
pendix V, draws attention to another parallel between Octavian and
Hercules: the consul who officiated at the thanksgiving in 29 B.C. was
Valerius Potitus (cf. Potitius 269). Virgil seems consciously to point to the
equation.
285 altaria: usually synonymous with arae, but here, in contrast to
aras 284, it must have its primary meaning and refer to the combustible
material (incensa) placed on the altar, the offering itself and the material
 COMMENTARY 99
to burn it: cf. [Tibullus] 3.12.17 celeres urunt altaria flammae; Lucan 3.404
structae diris altaribus arae; A. 5.93 (anguis) depasta altaria liquit. This
was why the priests brought torches (282), not primarily to give light,
although evening does advance during the ceremony.
286 populeis . .. ramis: for the poplar garlands, see the note on 276;
for adjective and noun 'framing' the line, see Appendix, p. 200.
286 evincti tempora: 'their brows bound round'; for the construc-
tion see the note on 29 turbatus pectora.
287 hie iuvenum chorus, ille senum: if Salii were associated with
Hercules at Rome, it is unlikely that old men would have been thought
appropriate to a warlike ritual. Conington suggests that Virgil is alluding
to the older-established college of the Salii Palatini who served Mars, and
the younger college of the Salii Collini (or Agonenses) who served Quirinus
(although as regards the actual performance of the rite in historical times
they were always amalgamated). This is possible but hardly likely. He
may simply be reproducing the situation and division of 105 una omnes
iuvenum primi pauperque senatus, recasting it into two ~µ,,xopia, to sing the
very Greek hymn to Hercules.
287 carmine: there is of course (see the note on 285) no evidence of
a carmen Saliare to Hercules at Rome. Fragments of a Salian hymn (to
Mars) still exist, but the hymn itself was unintelligible to Augustan Romans
(cf. Horace Epp. 2.r.86 ff.) and even to the priests themselves (Quintilian
r.6.40), and may well have been a pious fraud of religious archivists.
288 prima: not only the first of all the monstra Juno was going to
inflict on him, but also = primum, 'at the very beginning of his life';
both meanings are appropriate to the context; Servius opts for the first-
and then goes on to describe the circumstances of Hercules' birth.
Juno as the wife of Hercules' father but not herself his mother might in
this sense be called noverca; the proverbial harshness of stepmothers is at
least as old as Hesiod W.D. 825 llioTE JLTJTpv,~ 7r/).i;i ~µ,lpTJ, a'.MoTE JL~TTJP·
289 monstra . .. geminosque . .. anguis: hendiadys, the simplest
form of' theme and variation'; -que is epexegetic (explanatory).
manu often (as at 294) indicates physical prowess and means simply
'by force', {3tq.. Here it should be taken literally: 'how, to start with,
he crushed and throttled with his bare hands the monstrous pair of
snakes sent by his stepmother.'
eliserit: cf. elisos 261 and note.
The earliest authority for the detail that Hercules had only just been
born when the snakes appeared is Pindar Nemeans r.33 f. (Theocritus 24.r
makes him ten months old). The feat mentioned here is associated with the
circumstances of his birth, which was the traditional way of beginning
hymns to gods and heroes (cf. Horace Odes I.IO.I Mercuri, facunde nepos
Atlantis).
290 hello with disiecerit, and perhaps also with egregias, although
Oechalia (in Euboea or Thessaly or Messenia: there was doubt in antiquity
about which of these equally undistinguished alternatives was the Oechalia
of Eurytus) and even Troy at the time of Hercules' expedition against
IOO COMMENTARY

Laomedon could hardly be said to have a warlike reputation like the future
Carthage of A. r.444 f. bello / egregiam . . . gentem.
Military exploits were an expected item of hymns; how the Trojan
guests reacted to the mention of Hercules' sack of Troy, we are not told,
but the hymn was of course composed before they arrived.
V. Buchheit, Vergil iiber die Sendung Roms (Heidelberg, 1963), 123,
discussing similarities which suggest that Virgil wished to draw a close
parallel between Hercules and Aeneas, points out that just as Eurytus had
refused to give his daughter Iole to Hercules, so Latinus was unable to
make good his promise to give his daughter Lavinia to Aeneas.
291 duros mille labores: Homer mentions specifically only one
'labour', the bringing of Cerberus from Hades (It. 8.366 ff.); Hesiod
names others, but the selection of the canonical twelve from the great
mass of adventures connected with him was probably the work of the
cataloguing Alexandrians (and twelve chosen because of the assimilation
of the Tyrian Hercules to Gilgamesh). No kind of chronological coherence
is to be looked for in the confused morass of legend.
291 Notice how the obvious Greekness of the Hymn to Hercules is
maintained by the Greek flavour of the idiom -que . .. -que (here and in 294
at the beginning of the line: contrast the note on 94), connecting Greek
proper names, as often in Homer; compare the exclusively Greek sea-
nymphs at G. 4.336 Drymoque Xanthoque Ligeaque Phyllodoceque.
292 Eurystheo: note the scansion: the two adjoining vowels coalesce
into one sound (synizesis). Latin poetry from the neoterics onwards
exercised considerable freedom in the declension of Greek proper names,
many of which would have been intractable in verse without taking some
liberty with them. Greek poetry from Homer offered many instances of
synizesis in the terminations of proper names (e.g. fl. I.I Mijv,v aHO€, 8€a,
II71>.71,ao€W 'Ax,>.ijos), and this was gratefully imitated when Latin trans-
ferred Greek third declension names in -ms to its second declension: so
Eurystheo became Eurystheo, Nereo Nereo (Prop. 3.7.67), Nere'i Nerei
(A. 8.383).
292 fatis Iunonis iniquae: 'through the decrees of harsh Juno', who
was hostile to Hercules because of her jealousy of Alcmene, just as she was
hostile to the Trojans because of her jealousy of Ganymede and resentment
at the judgment of Paris (A. r.26 ff.). But Hercules had triumphed over
her obstacles: an encouraging example for the Trojans.
293 ff. Such of the canonical labours as Virgil mentions all fall within
the 'quoted' part of the hymn:
(1) The Centaurs 'born of a cloud' (which Jupiter fashioned in the
likeness of Juno when Ixion tried to rape her) included Hylaeus and
Pholus who were slain in the battle between Centaurs and Lapithae, the
most celebrated drunken brawl of Greek myth (cf. G. 2.455 ff.). Hercules
participated, en route to the fourth labour, bringing back alive the wild
boar of Mount Erymanthus.
(2) The Cretan bull, which according to some accounts fathered the
Minotaur on Pasiphae, was captured by Hercules (as his seventh labour)
 COMMENTARY IOI

and taken alive to Eurystheus (mactas 294 is applicable to Nemeae ...

leonem but not strictly to Cresia ... prodigia).
(3) The Nemean lion was killed by Hercules and its skin taken to
Eurystheus (the first labour).
(4) The last labour was to descend to Hades and bring back Cerberus
(296 f.); cf. A. 6.395 f.
(5) Typhoeus (or Typhon), hundred-headed and serpent-limbed, was
the protagonist of the Giants in their battle against the Gods. Hercules was
involved in some versions of a gigantomachy, but not as one of the canon-
ical labours. Before Virgil, Euripides alone (Hercules 1271 f.) mentions a
combat between Hercules and Typhon. Virgil probably regarded Typhoeus
as one of the monsters Hercules saw in Hades, because (1) he was hurled
down to Tartarus by a thunderbolt from the victorious Jupiter (cf. Hesiod
Theogony 853 ff., 868 and A. 1.665); (2) he was a natural link between
Cerberus (296) and the Lernaean Hydra, being the father of both (and the
Chimaera's as well, Hesiod Theogony 304 ff.), and Virgil groups the giant
Briareus with the Hydra and the Chimaera as terrors of the Underworld
at A. 6.287 ff.
(6) The Hydra of Lerna was eventually killed by Hercules in his second
labour by burning eight of the nine heads and burying the ninth and
immortal one under a rock.
The tenth labour, the fetching of the cattle of Geryon, was amplified in
Roman myth by the story of Cacus (largely modelled on it, see the note
on 198).
293 ff. tu (293) ... tu (294) ... te (296 bis) . .. te (298) ... te (299): the
repeated pronouns (in the usual manner of invocations and hymns)
punctuate the list of the apETat Toii 8wii (laudes / Herculeas et facta 287 f.);
cf. the proem to Lucretius De Rerum Natura (the invocation to Venus);
Horace Odes 2.19.17 ff. (and Fraenkel, Horace 200); Catullus 34.13 ff. (and
Fordyce ad loc.); and see Norden, Agnostos Theos 149 ff.
293 nubi~enas, invicte, bimembris: three compound adjectives in a
row is unparalleled in Virgil. For their formation see the note on 82 con-
color. invictus is common enough even in prose before Virgil: nubigena (only
again in the Aeneid at 7.674, also in reference to the Centaurs) was
apparently coined by him, and bimembris appropriated by him for high
poetry from Cornificius (who had first used it according to Macrobius
6.5.13). Together they give a rhapsodic Greek flavour to the opening of
this hymn of praise: the Homeric Hymns are full of them, cf., for example,
19 (to Pan) I f. 'Aµcf,{ µo, 'EpµElao cf,D.ov yovov lvvrn€ Moiiaa, I aly,'TT0871v
SiKepwrn cpiAoKpoTov; Hymn 8 (to Ares) is a positive tour de force in stringing
them together. Three compound adjectives also occur (but not together)
in the famous encomium of Augustus in A. 6.791 ff. where Anchises
predicts that his empire will stretch beyond the limits reached by Hercules
and Bacchus: Norden (on 6.796 ff.) detects a 'dithyrambic' character in
this passage, but there is no need to press this: Virgil was well within the
Aristotelian canon (Poet. 1459a) that epic could make use of compound
words, although these were primarily appropriate to dithyramb.
102 COMMENTARY

(In invicte there is certainly an allusion to the Roman cult-title of

Hercules Invictus, which Haupt wished to make explicit by printing an
initial capital letter here (Hermes 3 (1869), 153).)
294 mactas: for the present indicative in the context of a catalogue
see the note on generat 141. mactare here simply = 'slaughter', with no
religious overtones (see the note on 85).
294 The pause, slight here, after a trochee in the third foot is not very
common in Virgil. It is the characteristic pause of Homeric poetry, but
Latin's comparative lack of words ending with a short syllable made it
impossible to naturalise it. When it does occur in Virgil it is often con-
nected with a Greek proper name, as here and 440 Aetnaei Cyclopes; often
also, as here, there is an enclitic -que before the pause, cf. 679 populoque,
688 sequiturque.
297 ossa super recubans: 'Where did he get his bones in that region
of shadows?' asks Henry, hot, as usual, in pursuit of material realism.
'He devoured with great gusto the cake the Sibyl was so thoughtful as to
bring with her. That cake, and unfortunate Pirithous, are the only food
I ever heard from any reliable authority of his getting ... Is the omission
Virgil's or of Virgil's religion? I am inclined to think, of the latter.' But
mythology is not completely silent on the point. The germ of the idea,
which reappears in Statius Th. 2.29 sparsa solo turbaverat ossa (sc.
Cerberus), is probably Hesiod's WfLTJ<TT'T/• (Theog. 3n, cf. 772 f.) (so Maguire,
Hermathena II (1885), 344).
297 ossa super: for the anastrophe see the note on 32 inter. semesa:
for semi-esa, but the i which becomes semi-consonantal before a following
short vowel (as in semihominis, see the note on 194), disappears completely
before a long vowel, as here.
299 arduus arma tenens: cf. A. 3.70 lenis crepitans vocat Auster in
altum, where Servius says duo epitheta posuit vitiose . . .fecit autem hoc prope
in decem versibus (elsewhere in the Aeneid at 5.278, 567, 763 f., 838;
8.559, 683; n.755; 12.902). Servius felt that the first adjective really had
the force of an adverb, and should have been one. But Latin was not
nearly so rich in adverbs as a modern language: some adjectives, like
arduus, simply did not form them. In poetry, although they are not so
hard to find in Lucretius, the growing refinement of poetic taste developed
a distinct dislike for all but the commonest like longe, alte, late, which were
virtually unavoidable (see Axelson 62 f.). So Virgil and other Augustan
poets developed the two-adjective idiom on the basis of usages like
Lucretius r.258 f. candens lacteus umor / uberibus manat distentis, where
the first adjective can be felt either as a true adjective in asyndeton with
the second, or as qualifying it adverbially.
299 rationis egentem perhaps from Lucretius 4.502 f. praestat rationis
egentem / reddere mendose causas but hardly in the same precise sense of
'one who cannot give the true explanation'; the sense is nearer to aµ.~xavo,,
'resourceless'. According to one account (in, for example, Val. Fl. 7.623 f.)
it was Pallas Athene who suggested to Hercules the use of fire to kill off
the hydra; this was his second labour, and he was still rudis (Lucan 4.634).
 COMMENTARY rn3

301 addite agrees with tu implied by salve; decus is in apposition to tu.

Hercules had proved that his father really was Jupiter by the feats he had
accomplished; these were beneficent and civilising; his immortalisation
not merely increased the company of the gods, but graced it (decus).
Bentley's deus is unnecessary; cf. Statius Th. 1.22 (Domitian) tuque o,
Latiae decus addite f amae.
302 dexter . .. pede . .. secundo: the language of augury; dexter =
propitius, pede secundo = pede fausto (Hor. Epp. 2.2.37). The greeting
(salve) is followed by the wish for benevolent attention, a regular prayer-
formula (cf. Livy 1.16.3).
303 f. 'to crown it all they add (the story of) Cacus' cave and the
creature himself snorting with flames'.
305 = A. 5.149 f. consonat omne nemus (vocemque inclusa volutant /
litora, pulsati) colles clamore resultant.

306-336 After the ceremony Evander leads Aeneas home and tells him the
history and legend of the landmarks they pass. The original inhabitants were
Fauns and Nymphs and savages; a Golden Age of civilisation was introduced
by Saturn, but gradually deteriorated, and the land changed its name under
successive waves of invaders; Evander himself through divine guidance was
the last to settle there.
306 f. se ... ad urbem ... referunt: The Site of Rome. The Trojans
had their first glimpse of the settlement on the Palatine when, rowing
upstream, they rounded the Aventine. They came to land on the wooded
bank alongside the forum boarium, in the south-east part of which lay the
Ara Maxima, very near the site of the later temple of Hercules Victor
(Nash 1, 472 ff.) at the north end of the Circus Maximus.
After the sacrifice they walk north, roughly parallel with the river,
while Evander gives his account of the prehistory of Italy; with his
concluding reference to his mother Carmentis (336), they have reached the
river-end of the valley between the Palatine and Capitoline.
At the south-west foot of the Capitoline is the ara Carmentalis whose
remains are perhaps identifiable (see Nash 1, 415). Nearby is the porta
(338) which took its name from it, and was chiefly famed because its
right-hand postem was the Porta Scelerata through which the Fabii were
said to have marched to their extinction in 479 B.C. (Livy 2.49.8); some
remains of the Porta Carmentalis were discovered in 1959 (see Gjerstad
3, 459 f.).
Evander then points in the direction of the asylum (342) which lay in
the depression between the two summits of the Capitoline, inter duos lucos,
in Augustan times (Livy 1.8.5); but for Virgil the whole of the hill is
wooded (342). This very early asylum was presumably thought to have
been under the protection of Veiovis ( = Jupiter), the remains of whose
temple inter duos lucos date from the middle of the second century B.C.
(see Nash 2, 490 ff.). The asylum lay outside the boundaries of the primitive
Palatine settlement: its foundation was ascribed to Romulus, as by
 COMMENTARY

Virgil here (the poet speaks in propria persona through Evander's account
in 342 f., 347 f. and 361).
As Evander and Aeneas look up the valley between Palatine and Capi-
toline, the Lupercal (343), a grotto in the base of the Palatine cliff, is on
their right (extreme right, if Platner-Ashby are correct in locating it at the
south-west foot of the hill, but no certain identification has yet been made).
But Evander calls attention not to the Forum-to-be, which now lay
straight ahead of them, but to the wooded slope of the Quirinal beyond it,
where later the street of the Argiletum led down to the Forum (see the
note on 345, and Nash 1, 151 ff.).
Moving on up the valley, they pass the Tarpeia sedes (347), which could
refer to the whole of the Capitoline hill (Varro L.L. 5.41), but here and at
652 (Arx Tarpeia) refers to the precipitous cliffs of the more southern
peak (pictured in Nash 2,409 f.), the Capitolium proper, as opposed to the
more northern peak, the arx Capitolii. See the note on 652 for Tarpeia's
connection with the Sabines, long after Evander.
On the Capitolium proper there stood in Virgil's day the temple of
Jupiter Capitolinus (Nash 1, 530 ff.), splendidly restored (aurea nunc 348)
by Q. Lutatius Catulus after the fire of 83 B.C., and again by Augustus,
very sumptuously, in 26 B.C. (R.G. 20). (For the nearby temple of Iuppiter
Tonans see the note on 352 ff.) In Evander's day the Capitolium was
covered with a numinous grove (like the arx of Latinus' city, A. 7.172).
Evander (pace Mackail) does not climb the Capitoline (he is far too frail
for that!), but with Varronian learning (L.L. 5.41 f.) points out the relics
of two abandoned prehistoric settlements, the hill of the Janiculum across
the river (358), and at hand the Arx Saturnia, the Capitolium proper (see
further the note on 357 f.).
Rounding the north side of the Palatine, Evander and Aeneas then
climb up it (subibant 359) to a spot from which they can see the sites of the
forum Romanum and the Carinae. The forum Romanum (361) did not come
into existence until the spreading settlements on other hills required a
more central market than the riverside mart in the forum boarium which
had served the original Palatine settlement. The Carinae, 'Keels' (361)
was so called, according to Servius, from the keel-shaped appearance of
certain buildings; in late Republican and Augustan times it was a very
fashionable and luxurious quarter on the south-west spur of the Esquiline,
facing the north-east slope of the Palatine (Pompey once had a house
there, appropriated by Mark Antony). Evander's house itself (366) was
obviously very near and very like that of Augustus (Suetonius Augustus
72), 'not conspicuous for its lavishness ... and for more than forty years he
slept in the same bedroom winter and summer' (whereas luxurious houses
had both cubicula aestiva and hiberna).
Evander's promenade follows the topography of the three central and
isolated hills of the Tiber's east bank; the Janiculum owes its mention to
the legendary connection between Saturnus and Janus: see the note on
357 f.
See R. Pichon, REA 16 (1914), 4ro ff. and P. Grimal, REA 50 (1948),
 COMMENTARY 105

348 ff. for discussions of Evander's promenade, very much in the tradition
of Gaston Boissier, and with very questionable archaeological identifi-
cations.
Virgil omits all mention of the tradition, common in descriptions of
early Rome (Tib. 2.5.33 f.; Prop. 4.9.5 f., 4.2.7 f.; Ovid Fasti 6-405 ff.),
that the valley of the Velabrum between Capitoline and Palatine was in
olden times a stagnant arm of the Tiber (and in Augustan times still the
route by which the flooding river could threaten the forum Romanum,
cf. Horace Odes r.2.13 ff.). If Virgil had this in mind, Evander must be
supposed to round the Palatine keeping close to its lower slopes-and there
is nothing in the narrative to contradict this.
307 obsitus aevo: 'overgrown with age'; obsitus (like consitus)
literally = 'thick-sown', an old and typically Latin (agricultural) meta-
phor, cf. Plautus Menaechmi 756 ff. consitus sum / senectute, onustum gero
corpu', vires / reliquere, and Terence Eunuchus 236 video sentum (literally
'thorny', cf. A. 6.462 loca senta situ) squalidum aegrum, pannis annisque
obsitum.
310 facilis either accusative plural with oculos or nominative singular
with Aeneas; in either case translatable as an adverb of manner: see the
note on 30 seram.
310 omnia circum: a simple case of anastrophe; see the note on 32
inter.
3rr capitur: 'is capitivated'; capere = 'to charm' is common in love-
elegy (see TLL 3,337 f. s.v.); cf. also Cic. Cael. 42 .. . si quem forte inve-
neritis qui aspernetur oculis pulchritudinem rerum, non odore ullo, non tactu,
non sapore capiatur, excludat auribus omnem suavitatem . ..
312 monimenta: with monere 'to put someone in mind of something'
are cognate monimentum 'reminder' (hence 'memorial' of the dead, etc.),
and monstrum, that which puts one in mind of the will of the gods.
313 Romanae conditor arcis: that the Palatine and not the Capitoline
should be described as arx Romana is a clear indication of how serious a
rival the legend of the Arcadian foundation of Rome was to the traditional
story of Romulus. Romulus was associated with the Capitol, whose
religious importance had been devalued in Augustan Rome by the temple
of Apollo and Augustus' own house on the Palatine.
314-32 The prehistory of Latium. This excursus, full of Roman associa-
tions, is a perfect balance to the Greek mythology of the hymn to Hercules.
It is a combination of elements from four sources: Lucretius, folk-etymo-
logy, Hesiod, and Roman antiquarianism.
(1) 314-8 is almost a resume of Lucretius' famous account of man in the
earliest times (5.925 ff.); cf. especially at genus humanum multo fuit illud
in arvis / durius, ut decuit, tellus quod dura creasset (925 f.); nee robustus erat
curvi moderator aratri (933); nee commune bonum poterant spectare neque
ullis / moribus inter se scibant nee legibus uti (958 f.). The rami are those of
the acorn-bearing oak (Luer. 5.939), and the asper victu venatusis described
more fully at 5.966 ff.
Whenever Virgil touched on any part of the creation-myth he found it
ro6 COMMENTARY

impossible to forget Lucretius (E. 6.31ff.; G. 2.476-82; G. 4.221-7;

A. r.743; A. 6.724 ff.). But nymphs and fauns are the stuff of poetry, and
Virgil chose not to follow Lucretius (4.572 ff.) who had rationalised them
away as the fantasies of rustics who did not understand echoes and the
science of sound.
(2) 319-25. The commonly held view was that man evolved from
primitive anarchy when organised into communities by the natural
leaders (so, for example, Luer. 5.rro5 ff., and Cic. Sest. 91 f.): again a
'scientific' view, unexciting poetically. The connection of Saturn with a
Golden Age of agriculture depends ultimately on a piece of folk-etymology
(see the note on 319 f.); it is a well-worn theme with Virgil, but worked
out with different associations, which should not be missed here: it is an
age of blissful hope and peace (E. 4.4 ff.), like the age of the simple, sturdy
way of life which made Rome great (G. 2.532 ff.) and which prevailed in
Latium when Aeneas arrived (A. 7.202 ff.); it will be inaugurated again
(A. 6.792 ff.) by Augustus, a second great civiliser, who will put an end to
the aspera saecula of interminable civil war (A. r.291).
(3) 326 f. But the Age of Gold became discoloured (decolor), ending
with the Age of Iron under the reign of Jupiter (G. 2.536 ff.). Here, as at
G. r.121 ff., Virgil pictures two Ages only, associated with Saturn and
Jupiter respectively: Hesiod W.D. 109 ff. had five, Ovid Met. r.89 ff. has
four, Aratus Phaen. rr4 ff. three. In the Golden Age each land was self-
supporting (Hesiod W.D. rr6 ff.); but one symptom of man's degeneracy
was greed (amor habendi) which led him to violate the natural boundaries
of the land and sail the seas for commerce or war. It would have been
quite precedented at this point either to blame Jupiter (G. r.129 ff.) or to
curse the man who first set sail (Horace Odes r.3.21 ff.); instead Virgil
thinks of the implications of this for Italy-invasion by foreign tribes.
(4) 328-32: a brief mention of names to evoke the dim and distant past
(see the note on 328 and 330 f.).
314 Note the typically Virgilian combination of Greek nymphs and
native Latin fauns, although in this case he had been forestalled by
Lucretius (4.580 f.). Here the Fauni are indigenous before the arrival of
Saturn; at A. 7.47-9 Faunus is the father of Latinus and grandson of
Saturn: unlike most mythographers Virgil makes no attempt to impose a
spurious precision on details of this kind.
315 gensque virum truncis et duro robore nata: the belief that
men were born from the ground or from trees seems to have been wide-
spread among very primitive communities, a product of the animistic
outlook which invested everything, living and inert, with a life and soul
of its own. Such a belief could be seriously put forward by Hesiod W.D.
145, describing 'Bronze Age' man, or scoffed at in Homer's time, cf.
Od. 19.163 (Penelope asking about Odysseus' family) ov yap a1ro 8pv6s luu,
1ra>..aicf,&:rov ov8' a1ro 1rfrp71s, 'for you are not from the proverbial oak and
rock' (there may be a reference here to the oracular oak-trees of Dodona,
where Greek tradition located the first dwelling-place of the human race,
cf. G. 1.7 f.). It could also be turned into a taunt: at Il. 16.33 ff. Homer
 COMMENTARY

makes Patroclussay to AchilleSV1)A££S',OVK a.pa uol y1: 7ra-r1Jp ~v lmr6ra II71.\1:vs-

/ oUOE 0fr,s- P,~TTJP • y.\av,o} 0£ <T£ rlKT£ Oa.>..auua / 7rfrpai r' ~.\tf3aroi, on roi
v6os- eurlv d.7TTJ~S', and so launched one of the most popular of literary
commonplaces (see Pease's instructive note on A. 4.366 f.).
The view that men were born of rocks and stones is alluded to by Virgil,
G. 2.340 f. (the spring-time of the world) cum primae lucem pecudes
hausere, virumque / terrea progenies duris caput extulit arvis, and was the
basis of the creation-myth of Deucalion and Pyrrha (Ovid Met. 1.313 ff.).
The concept was perhaps promoted by a belief in the magical connection
(nomen omen) of .\aas- 'stone' and .\a6s- 'people'.
316 quis (dat. pl.): poetic and colloquial equivalent of quibus: see
N-W 2,469.
mos= 'code of behaviour', cf. A. 6.852 pacique imponere morem.
318 rami (alebant) sc. fructibus, cf. G. 2.500 f.
318 aspervictu: cf.A.1.445/acilemvictu ... gentem. victuisanablative
of respect (or limitation), an offshoot chiefly of the ablative of instrument
(but see Woodcock 38).
319 f. Greek myth (from Hesiod Theogony 453 ff.) told of how Cronus
tried to defeat the prophecy that he would be dethroned by one of his sons
(by his sister Rhea) by swallowing them in turn. Zeus however was hidden
away at birth on Crete, and grew up to fulfil the prophecy (in one version
of the story actually killing his father with a thunderbolt).
The original attributes of Saturn as a god of Roman religion are very
obscure; Greek influence certainly modified his cult at a very early date,
if it was not there from the very first. In historical times at his temple in
the forum Romanum at the foot of the Clivus Capitolinus, the oldest in
Rome after Jupiter's temple on the Capitol itself, he was worshipped
with uncovered head (Servius on A. 3.407)-a sign of Graecus ritus (as
also with Hercules, see the note on 276), and his image was that of the
Greek Cronus.
In Roman myth (where his wife-sister is Ops, goddess of plenty) he is
chiefly associated with agricultural prosperity and the Golden Age in
Italy (G. 2.538, A. 6.792 ff.); folk-etymology connected his name with
sowing (ab satu est dictus Saturnus, Varro L.L. 5.64), and this may be
responsible for agricultural connections which he did not originally
possess.
319 primus here should primarily be taken as adverbial = primum,
'first of all', picked up by tum ... tum in 328, 330; but its sense clearly
extends also to is genus indocile ... composuit, he was the first (of a series)
to civilise Latium.
322 legesque dedit: here Saturn 'gave laws', but in A. 7.202-4 we
were told that the •race of Saturn' lived without them. The contradiction
results from Virgil's different intentions in the two passages. Here he
wishes to hint at the affinity, as lawgivers, between Saturn and other
figures: Evander in this book, Aeneas elswhere in the poem (A. 3.132 ff.;
12.190 ff.), and Augustus in contemporary practice and poetry.
322 f. ' ... and preferred the name Latium, because he had lain hid
9 EAC
108 COMMENTARY

safely in these parts'. Latium was preferred to Saturnia tellus (329);

cf. Ovid Fasti 1.237 f. inde diu genti mansit Saturnia nomen; / dicta quoque
est Latium terra latente deo. The reason for the name depends on a piece
of folk-etymology, which Virgil would have liked (and may have been the
first to bring into connection with Satumus, Binder 85 and n. 42) (cf. the
note on 425 Pyracmon), connecting Latium with latere. Examples of this
kind (they abound in Varro's De Lingua Latina) often have the air of a
bad pun, but one should remember that the feeling of mystic power in
names, especially as omens, was deep-seated among the Romans. Varro
(quoted by Servius here) subscribed to a more prosaic derivation: (Latium)
quod latet Italia inter praecipitia Alpium et Apennini.
323 latuisset: for the subjunctive in a causal clause, see the note on
129 f.
324 aurea quae perhibent: 'the age (saecula poetic plural) which
they style "Golden"'; note the idiomatic word-order, as usual in prose
also.
(The misreading of aurea quae as aureaque, easy in undivided capital
script, disrupted the metre, which in the later MS tradition was healed by
inserting ut before perhibent. For the preferability of fuere to the variant
fuerunt see the note on 53.)
328 Dionysius of Halicamassus (1.60 ad fin.) gives the following succes-
sion of immigrants to the site of Rome: (r) Siculi, (2) Aborigines, (3) Pelas-
gians from Thessaly, (4) Evander and followers, (S) Hercules and followers,
(6) Aeneas and followers. Almost all other ancient authorities agree in
giving priority to the Siculi, whose king !talus (Thucydides 6.2.4) was the
eponymous coloniser of Italia, 'calf-land' (lraA6s- being = vitulus ac-
cording to both ancient and modem etymologists); Virgil is the first to
refer to them as gentes Sicanae. A usonius elsewhere = 'Italian' in a wide
sense; the reference here may more narrowly be to the Aurunci (etymo-
logically connected with Ausones), settled in historical times between the
Voltumus and the Liris. But for Virgil the names are evocative of pre-
history, ancient, vague and confused: to list their changes (cf. nomen
posuit 329) was expected in an encomium of a city or territory. See further
Rehm 64 ff. and n. 139, and for the ethnology of the Sicani, CAH 4, 436 ff.
Siciinae here, as always at the end of a verse in Virgil, is an innovation
metri causa; Siciinium (416) is the traditional form (Homer Od. 24.307).
329 Saturnia tellus recalls Ennius' Saturnia terra (Ann. 25 V), and
Evander's excursus on the prehistory of Latium, especially its references
to Saturn, may owe something to Ennius' description of it (at the time of
Aeneas' arrival?) in Ann. 23 ff. V.
329 posuit here for deposuit, cf. A. 11.308 f. spem si quam ascitis
Aetolum habuistis in armis / (de)ponite; but contrast A. 7.63 Laurentisque
ab ea nomen (im)posuisse colonis, and A. 8.639 posito for composito. Other
instances of simple for compound verbs are given in the note on IIO, and
cf. A. II.2II (e)ruebant, 375 (pro)vocat.
330 f. Virgil is silent about the provenance of the eponymous hero
Thybris: ancient speculation was rife (Servius on 72 and 330); one version,
 COMMENTARY 109

followed by Livy 1.3.8, made him a king of Alba; another (in, for example,
Varro L.L. 5.30), more plausibly, an Etruscan.
330 ff. Ogilvie (Livy 1-5, 43) points out that the driving out of the name
Albula by the name Tiberis/Thybris 'represents the victory of the Etruscan
language (Thebris) over the indigenous': he supports the theory that
Albula has nothing to do with Latin albus but, like 'Alps', derives from a
pre-Indo-European word= 'mountain'. Servius on A. 8.332, however,
connects the name with the colour of the water (which has nothing
necessarily to do with mountains). Albula(e) was also the name given to
sulphur springs at Tibur (Tivoli), which were and are used medicinally,
and can still be detected from some distance; the sulphurous waters of the
Nar, which flows into the Tiber, are mentioned by Ennius (Ann. 260 V),
and Virgil, A. 7.517 sulpurea Nar albus aqua. So there is a strong possi-
bility that deposits of sulphur (and calcium) in the Tiber basin did actually
give the water a whitish colour, and this is what the Romans at any rate
would have connected with the name.
331 itali: the first vowel is naturally short, as here; metrical pressure
caused it to be lengthened, first in Latin by Catullus (1.5), after Calli-
machus (see Norden 141). With the adjective !talus Virgil preferred to
observe the natural quantity rather than follow the Greek irrational
lengthening; the noun Italia compelled him to follow it; see Leumann
146, n. 3.
333 Dionysius of Halicarnassus (1.31) knew of a tradition that Evander's
immigration took place about sixty years before the Trojan war, and his
followers were so few that they occupied only two fishing-boats (cf. 473
exiguae vires).
333 pulsum patria: because at the instigation of his mother he had
murdered his father Echenus (other versions say that Echenus was not
his real father-that was Hermes/Mercury).
333 pela~ique extrema sequentem, 'pursuing the ocean's very
limits '. Navigation in the heroic age tended to be timorous and un-
adventurous; while searching for the promised land, Evander (like
Aeneas) found the voyage so long that it seemed that the shores he was
looking for must be receding from him: the same sentiment underlies
A. 3.496 f. arva neque A usoniae semper cedentia retro / quaerenda; A.
5.629 f. Italiam sequimur fugientem. A similar phrase is used of Dido at
A. 6.457 ferroque extrema secutam in reference to her suicide, and it may
have been this and other phrases like it which led Tiberius Donatus (cited
by Servius on this passage) to interpret the present passage to mean
'going to meet death on the high seas': this entails taking pelagi as locative,
which is unparalleled and unlikely (see LHSz 2, 149).
334 Servius here says secundum stoicos locutus est, qui nasci et mori
f atis dant, media omnia fortunae: nam vitae humanae incerta sunt omnia.
A sharp opposition between destiny which is inescapable (cf. A. 10.467)
and chance which is capricious (cf. A. 5.709 f.) is certainly Stoic, and
probably present in this passage. But Bailey's analysis of Virgil's use of
Fortuna (RV 234 ff.) shows that it ranges freely from' chance' to' fate' itself.
'.)·2
IIO COMMENTARY

334 An accumulation of five-word lines (here, 338, 341) is a sign of the

elaboration of the high style in the passage in which they occur: cf.
examples in the proem (1-48) of G. 3, and the Aristaeus episode (315-558)
of G. 4.
336 ff. Carmentis: the original attributes of Carmentis (or Carmenta)
as a Roman deity are not known for certain, and probably were not
remembered by the first century B.C. Etymological guesses were made,
which may throw light on the original situation, or at any rate have been
influential in remoulding the cult of the forgotten goddess:
(1) with carpentum, the covered carriage used by matronae in religious
processions. This connection, coupled with the interpretation of her cult-
titles Postvorta and Antevorta (or Prorsa or Porrima) to refer to the two
positions in which a child may emerge from the womb (an interpretation
current in the Augustan period), suggested that she was a goddess of
childbirth, and worshipped as such at the Carmentalia on II and
15 January;
(2) with carmen, first in its original meaning of 'magical spell': this
supplemented (1) above, and made of Carmenta a weaver of spells for
good or ill at the birth of a child.
Later, stress was placed on another meaning of carmen, 'prophetic
utterance', and a relation seen between Carmen ta and Ca(s)mena; this is
the connection Virgil is concerned with. He makes her a water-nymph
(336, 339), who, like other water-deities (e.g. Egeria and Tiberinus),
would naturally be credited with prophetic powers; this he emphasises
in 340, vatis Jatidicae, cecinit. This view caused her cult-titles to be re-
interpreted to refer to her power of looking into the past and the future.
When the worship of Evander as a demi-god passed into Italy, Car-
mentis was identified with his mother, the local Arcadian nymph Themis
or Nicostrate; as such she was believed by some to have taught the Italian
Aborigines to write (a benefit attributed by other authorities (e.g. Tac.
Ann. II.14.4) to Evander himself); and she was considered to be the
Italian Sibyl. Like the Cumaean Sibyl, she would be inspired by Apollo:
in 336 we are to understand that the 'warnings' of Carmen tis were
prompted by Apollo.
See further Wissowa, RuK2 219 ff., Ogilvie on Livy r.7.8.
336 deus auctor Apollo: Apollo's gifts of prophecy and prediction
were particularly valued by the Greeks in a mapless age when they were
engaged in travelling. His advice as guide and archegete was sought on
voyaging to war (It. r.71 f.), and more often when embarking to found a
new colony: in this sense Apollo was the auctor, 'patron' of Evander's
settlement. The same association of ideas lies behind A. 6.56 ff., Aeneas'
prayer to the Apollo of Cumae, esp. 59 tot maria intravi duce te, and 69 f.
where Aeneas promises a temple and festival to Apollo if he grants him a
settlement in Italy.
 COMMENTARY III

337-358 They pass the Ara and Porta Carmentalis, named after Evander's
prophetic mother, Romulus' Asylum, the Lupercal, the Argiletum, the
Tarpeian Rock and the Capitol, with its awe-inspiring grove, and finally the
J aniculum is pointed out.
337-9 'Scarcely were the words spoken when next he moves on and
points out both the altar and the gate which the Romans call by the name
"Carmentalis ", an ancient tribute to the nymph Carmen tis.'
337 dehinc is here scanned as an iambus, as usually in Virgil; occasion-
ally it is scanned as a monosyllable by synizesis (cf. deinde 66).
339 honorem: for the accusative in apposition, either with portam or
with the whole preceding sentence, see the note on 487.
340 fatidicae: 'soothsaying, prophetic'; fatum (from fari) = the
spoken word, originally that of the prophet or seer.
340 prima: quia postea etiam sibylla dixit, Servius, referring to the
Sibyl of Cumae.
342 f. lucum . .. quern Romulus acer asylum / rettulit: Livy 1.8.5,
recording the same event, says .. . locum qui nunc saeptus escendentibus
inter duos lucos est asylum aperit. rettulit here is probably equivalent to
reddidit, 'a grove ... which Romulus brought into a new use, as a sanc-
tuary', cf. A. 11.425 f. multa dies variique labor mutabilis aevi / rettulit in
melius ( = meliora reddidit). referre often means 'to repeat' in some way
or other, and Servius here (followed by Henry) paraphrases rettulit 'created
in imitation of the Athenian asylum', and in view of the Greco-Roman
syncretism of 343 f., there may be such an additional meaning here.
quem Romulus . .. rettulit is a footnote added by Virgil speaking in
propria persona: it would of course be anachronistic for Evander to say it.
343 f. ' ... the Lupercal, called after the Arcadian tradition of the
Lycean Pan'. Parrhasia was a town of southern Arcadia, and the adjective
is loosely applied after the usual fashion to the whole district: Evander is
called Parrhasius at A. n.31, although his city was Pheneus (A. 8.165)
in the north of Arcadia. The mons Lycaeus was also in Arcadia in the
Parrhasia region.
Virgil seems to be suggesting that the Arcadian Evander introduced the
worship of the Arcadian Pan: it suited his mise-en-scene to follow anti-
quarians who assumed Lupercal to be derived from li,pus and arceo
(' warding-off wolves') and inspired by Pan's title AvKafos- taken to be
derived not from the mountain but from AVKos- = wolf. (Virgil therefore is
probably not following the same tradition as Ovid Fasti 2.423 f. quid vetat
Arcadia dictos a monte Lupercos? / Faunus in Arcadia templa Lycaeus habet,
and the monte of some inferior MSS, preferred by Schrader, is a misleading
facilitation.) The same theory made easy the identification of Pan and
Faunus, who were in any case both rustic deities, and easily conceived as
protectors of flocks.
The favourite explanation of Lupercal (as, for example, in Ovid Fasti
2.381 ff.) made it the cave where the she-wolf, lupa, suckled Romulus and
Remus-long after the time of Evander according to traditional chronology.
II2 COMMENTARY

The phraseology of 344 is slightly contorted: Servius was puzzled, but

his paraphrase is unacceptable (ordo est: monstrat Lupercal Panos Lycaei,
dictum de more Parrhasio, id est, ostendit Lupercal Panos Lycaei, dicatum
ex more Arcadico).
The Lupercal itself had a topical interest: it was restored by Augustus
(R.G. 19, and App. 2): see the note on 663 ff.
345 f. In Augustan times the Argiletum was a street between the
Subura and the foritm Romanum which it entered between the Curia
Julia and the Basilica Aemilia. The name was probably derived, as in one
tradition preserved by Varro (LL. 5.157), from argilla, 'clay', and perhaps
designated the potters' quarter. Folk-etymology saw in it Argi-letum, and
invented the story, told by Servius here, of 'a certain Argus' who plotted
to kill Evander and usurp his position; he was detected and killed by the
followers of Evander, who however built and consecrated a tomb for him
because he had been his guest. 345 f. then mean 'and he also points out
the grove of the consecrated Argiletum, and calls upon the spot to vouch
for his account of the death of his guest Argus' (the fact that he had given
holy burial proved his own innocence in the matter). But another inter-
pretation is hinted at in the Servian commentary: sacer ('taboo') can
mean 'accursed' as well as 'holy', and it is just conceivable that testatur
might in the Virgilian manner stand for the compound detestatur; the
lines would then mean 'and he also points out the grove of the accursed
Argiletum, and execrates the place in his account of the death of his guest
Argus' (who had so violated the sanctity of hospitality).
Attempts to solve the problem of testaturque locum by emendation are too
violent to be convincing: see Bursian 59 (1889), 167 for ostenditque locum
(0. Giithling); testatumque loco or taetratumque locum leto (P. Deuticke).
345 Ar~ileti: elsewhere the polysyllabic proper name ending a
spondeiazon is Greek (see the note on 54 Pallanteum); this is precedent
for the Latin name here (cf. the immediately preceding Pallanteum 341).
Norden 439 suggests that it may already have occurred in a passage of
Ennius dealing with the same subject-matter.
347 ad Tarpeiam sedem: see the note on 652 Tarpeiae ... arcis.
348 ff. Groves and woods on uncleared land were unfamiliar and awe-
inspiring to a primitive population, and so naturally accredited with a
numen: the memory of this lingers on in the Augustan poets. Virgil (A.
7.81 ff.) speaks of the groves beneath' lofty Albunea' which were haunted
by Faunus, Ovid (Fasti 3.295 ff.) of the oak-grove which had stood on the
Aventine 'at the sight of which you might say "There is numen here"'
(and cf. Amores 3.1.1 f., 3.13.7 f.). Seneca (Epp. 41) sermonising Lucilius
on the text sacer intra nos spiritus sedet, was reminded of A. 8.352 which he
quotes, and amplifies Virgil's description of the grove as a proof from
nature (as opposed to man) of the existence of the divine; he goes on to
say that the same conviction comes from deep caverns and the sources of
mighty rivers: was he thinking of Cacus' cave and the source of the Tiber,
also in A. 8?
Numinous groves existed also in primitive Germany (Tac. Germ. 9.3)
 COMMENTARY II3

and in Anglo-Saxon England (R. Branston, The Lost Gods of England

30).
348 aurea ... horrida: the contrast between the past and present state
of the Capitol became a locus tritus in Augustan poetry, cf. especially Pro-
pertius 5.r.5 ff.; Tibullus 2.5.23 ff. The laquearia of the Capitoline temple
had been gilded in the censorship of L. Mummius (142 B.c.); but it was
Catulus who had the bronze roof-tiles gilded (see the note on 306 ff.), and
provoked mixed criticism from his contemporaries (Pliny N.H. 33.(18).57).
horrida is both descriptive, 'bristling', and emotive, 'shudder-making'
(made explicit in the following lines).
The nemus on the Capitol included the specific oak-tree thought to be
the dwelling of the numen. It was on this that Romulus was said to have
hung the spolia opima. Later there was an ara and then a very small
temple, eventually restored by Augustus. This was the cult of Iuppiter
Feretrius, the oldest on the Capitol; later it was overshadowed by the more
splendid temple of Jupiter Capitolinus built by the Etruscans (see Warde
Fowler, RE 129 f.).
349 f. A good example of how anaphora (here the repeated iam tum)
can be used to organise a' theme and variation' instead of a simple copula;
other examples are frequent.
349 f. religio . .. / dira loci: 'the dreadful awesomeness of the place',
a good illustration of the basic meaning of religio, although the emphasis
is more often on the person feeling a 'sense of awe' (as at 598) than on the
object inspiring it. Cic. Nat. Deorum 2.72 derived the word from relegere,
but modem etymologists follow Servius on this line in connecting it with
religare (so the fundamental meaning of religio would be 'scruple',
'inhibition').
349 religio and reliquiae (356) both have a naturally short first syllable,
lengthened purely to suit hexameter poetry: a genuinely 'irrational'
lengthening, but helped by the intermediate stage rell-, with the consonant
doubled after the model of early Latin redduco (reduco), and perhaps also
of perfects like repperi, rettuli.
352 ff. Jupiter as weather-god, sender of thunder and lightning, had
cult-centres on hilltops throughout Latium. This basic character survived
uncontaminated in a very ancient temple in the Campus Martins (Iuppiter
Fulgur). P. Grimal (REA 53 (1951), 51-61) may be right in seeing in this
passage an allusion to the temple of Iuppiter Tonans on the more southerly
of the two peaks of the Capitoline hill, dedicated by Augustus, after a
miraculous escape from being killed by lightning, on l September 22 B.C.
(which according to Grimal provides a terminus post for the composition
not only of this episode but of the whole book).
353 cum saepe: the full meaning here is 'when, as often happens',
cum, ut saepe fit; saepe often has this pregnant sense at the beginning of
similes after qualis or ceu; cf. A. r.148, 5.273, 5.527, 10.723. See Munro on
Lucretius 5.1231.
354 See the note on 435 for Jupiter raising storm-clouds with the aegis.
357 f. It is likely that there was originally a cult of the god Janus on the
 II4 COMMENTARY

J aniculum, but if there was, it had disappeared by historical times, and

euhemerist interpretation of the memory made of Janus an early king of
the district like Saturn (coupled together also at A. 7.180 f.), whom he
welcomed to a share of his kingdom; cf. Augustine De Civ. Dei 7 .4 following
Varro: Saturnum fugientem benignus excepit (I anus); cum hospite partitus
est regnum, ut etiam civitates singulas conderent, iste I anicielum, ille
Saturniam (Servius on A. 8.319 says the same).
I aniculum has generally been taken to refer to the ridge on the west side
of the Tiber whose highest point is almost due east of the Palatine.
P. Grima! (RA 23-4 (1945), 56-87), assuming (correctly) that Evander
does not actually climb the Capitol, and that hanc ... huic ... illi have a
strictly local reference (which is not absolutely certain), argues that this
is topographically impossible, and that Ianiculum means the arx Capitolii
(the north side of the hill), and Saturnia (Arx) the more southerly summit
of the hill; he assembles evidence to show that the name I aniculum was not
given to the Transtiberine region until comparatively late-but that is
surely what a contemporary of Virgil's would have understood by it.
358 fuerat: the pluperfect indicative (especially fueram and habueram)
is sometimes found in contexts where at first sight a perfect or imperfect
might have been expected. The usage seems to originate in a vigorous
idiom of ordinary speech to stress that a given action or state in the past
was well and truly over and done with. It has a continuous history in
colloquial and vernacular styles from Plautus (cf. Lindsay, The Syntax of
Plautus 62 f.) through Cicero's letters to Petronius and later, and occasion-
ally invades more serious literary styles as well. It can be argued in some
cases that the pluperfect is used quite in accordance with its conventional
time-reference, and the difficulty is only an apparent one caused by literal
English translation; e.g. non sum ego quod fueram (Ovid Tr. 3.11.25),
literally rendered 'I am not what I had been', would have conveyed to a
Roman ' I am not what I long ago ceased to be '.
See further Palmer, L.L. 308, KS 1, 140 f., LHSz 2, 320 f., Fordyce on
Catullus 10.28 and 64.158.
The present line is additionally puzzling because to an Augustan Roman
(and Virgil allows Evander here to speak anachronistically like one)
the Ianiculum was still so called, but nobody referred to the Capitolium as
Arx Saturnia; so here fuerat nomen cannot be taken with huic to mean
'long ago ceased to be called'. Translate: 'This was long ago given the
name Janiculum, that Saturnia.'
358 Saturnia (sc. arx): cf. Varro L.L. 5.42 who says of the Capitol
hunc antea montem Saturnium appellatum prodiderunt.

359-369 Evander leads Aeneas through the meadow which is to be the

busiest and most magnificent part of Rome, and welcomes him to his modest
home. Night falls.
360 f. videbant ... mugire: see the note on 35.
361 lautis . .. Carinis: lautus, 'chic', here makes its unique appear-
 COMMENTARY 115

ance not only in Virgil, but in all poetry, epic or lyric, of the high style.
There is a pointed editorial comment: the quartier chic of metropolitan
Rome has lost the true values of Evander's paupertas. The confused
scholia of Servius auctus on this passage are disentangled by S. Timpanaro
RFIC 95 (1967), 428 ff.
362 limina: poetic plural; limen = the threshold (inferum) or lintel
(superum), and then (pars pro toto), a door, doorway or house. Here
limina subiit = 'stooped to pass under this lintel'.
363 subiit: note the scansion. Originally the last syllable of the perfects
iit and petiit (and of course their compounds) was long by nature; in
archaic Latin inscriptions it is found spelt -eit. By Virgil's time the last
syllable was regularly short; here he restores its earlier length, while
keeping to his rule that the syllable must bear the ictus and precede a
caesura (here a main caesura coincident with a sense-pause). See the note
on 98 procul.
· 363 cepit: 'contained, had room for'; cf. A. 9. 644 nee te Troia capit
and Juvenal 10.148 hie (Hannibal) est quem non capit Africa.
364 te quoque di~num: lines ending with two disyllabic words are
not so frequent in the Aeneid as in Lucretius and Virgil's earlier work (see
Conway on A. r.719). If both words kept their full word-accent, the ictus
would clash with accent in the fifth foot, against the whole tendency of the
Virgilian hexameter; but the usual rhythm is not disturbed when the first
disyllable is virtually enclitic and loses its own accent, as with quoque
here, tamen 566, and probably procul 666.
364 f. 'Have the courage, my friend, to scorn wealth; you too (like
Hercules) must model yourself in a way worthy of a god, and come with
no harshness towards our poverty.'
Hercules, when alive, had shown a godlike humility towards his host's
humble circumstances (like, for example, Jupiter and Mercury on their
visit to Philemon and Baucis, Ovid Met. 8.637 ff.), and at his death had
become a god: the lesson for Aeneas is clear.
But deo has caused difficulty:
(1) An ancient view, according to Servius auctus, took it = immortali-
tate; this is implied by the phrase, but not expressed.
(2) Page takes it to refer to Hercules, but 'make yourself worthy of
Hercules quoque, as Hercules did' is hardly possible.
(3) Henry takes the meaning to be 'worthy of the god from whom you,
no less than Alcides, are derived' (viz. Jupiter); but deus unqualified and
unhelped by the context cannot have such a narrow and precise reference.
(4) The emendations of Deuticke (decus for deo) and Cauer (dignam finge
domum) make lucid sense and are what anyone but Virgil would have
written. te quoque dignum / finge deo is guaranteed by Seneca's citation of it
(twice: Epp. 18.12, 31.n) and by the remark which follows it in Epp. 18.12:
nemo alius est deo dignus quam qui opes contempsit. For the idea of' making
oneself like a god' in Greek philosophy (the Stoics and Cynics had Hercules
in mind), see C. Koch, Religio (Niirnberg, 1960), 216 f. deo dignus maybe an
attempt to render 8rn11p€'1rtJS,
II6 COMMENTARY

367 f. Evander placed Aeneas where he was supported (effultum is pro-

leptic, see the note on 37) on a bed of strewn leaves with a bearskin thrown
over them.
368 pelle Libystidis ursae occurs also at A. 5.37 where it is part of
Acestes' dress; Virgil, like Milton after him, uses proper names for the sake
of their associations-Libystidis suggests something remote, exotic and
valuable.
369 The writer of a narrative poem which covers any length of time is
continually faced with the problem of how to effect the transition from the
time of going to sleep, when normal activity ceases, to the time of getting
up, when it recommences. To say simply 'The sun set. They went to sleep.
Dawn appeared' is suitable to recital epic (Homer says this, with only
slight elaboration, at Jl. r.475-7). For literary epic there were two ways
out of the difficulty:
(1) To keep the abrupt formula of transition but to vary its expression.
Ennius Ann. 89 f. V may have been alive to this possibility, and great
ingenuity was expended by later poets in ringing all the possible changes
of phrase. For Virgil's own variations, cf., for example, A. 2.250, 5.721,
11.201 f.
(2) To bridge the night with action, in which either mortal contact is
made with the supernatural (in dreams and visions), or immortals are
engaged among themselves. The first night of Aeneid 8 was taken up with
Aeneas' vision of the Tiber-god (31 ff.); the second here is bridged with the
activity of Venus and Vulcan. The progress of this particular night is care-
fully advanced (280-369-407-455) and suggests that the plan of the book was
more elaborately prearranged than that of some others. Sometimes, as at
A. II.I and 182 f., Virgil simply omits any mention of dusk or night, and
marks the passage of time only with a rather perfunctory allusion to dawn.

370-393 Venus, alarmed at the situation in Latium, tries to persuade

Vulcan to make arms for Aeneas. Her embrace works on him like lightning.
370 ff. In terms of the Homeric substructure of the poem the night
spent here by Aeneas after the first day with Evander corresponds with
the night spent by Telemachus after the first day with Nestor (Od. 3.397 ff.)
(see note on 97 ff.), and to some extent also with the night spent by Odysseus
after the first day with Eumaeus (Od. 14.520 ff.) (see the note on 183 f.).
With these Odyssean nights Virgil synchronises the Iliadic night during
which Thetis visits Hephaestus and he makes the arms for Achilles
(It. 18.369 ff.). But apart from the broad situation, the encounter of Venus
and Vulcan owes little of significance to that of Thetis and Hephaestus.
In fact a crucial difference between the two episodes, the fact that
Hephaestus from a sense of obligation is willing to do whatever Thetis asks
even before she has asked anything, whereas Venus makes her request first
and gets her way through her more-than-wifely charms (387 ff.), suggests
that Virgil took more than a hint from the more-than-wifely blandish-
ments Hera uses to cajole Zeus in It. 14.153 ff. (Hera prepares for the
 COMMENTARY IIJ

encounter with a loan of sex-appeal from Aphrodite (Il. 14.214 ff.), which
of course as Venus (here) she possesses in her own right).
Servius auctus (on 383) is concerned to point out that although Aeneas
was Venus' son, he was not Vulcan's. Vulcan may not care to have been
reminded of the implications of the relationship. But heroic tact (and
perhaps heroic feeling too) was not so fine on these matters: at It. 14.315 ff.
Zeus tells his wife how much more desirable he finds her than any of his
past amours, whom he lists in detail.
372 auroo, cf. 553 auras: according to Norden (on A. 6.280) the
Augustan poets were the first to use synizesis with purely Latin words, the
first example being Horace Sat. r.8.43 cerea; before, it is found only with
Greek proper names, to make them manageable in Latin verse (see the
note on 292 Eurystheo). But it was an easy extension from proper names
ending in -eus to adjectives with the same termination (always adjectives
of material in Virgil, aureus, aereus,ferreus, themselves modelled on Greek
xpvaioi, (for xpvaor,, etc.), and then to nouns in -eus (alveo, A. 6.412;
balta, A. ro.496). Other examples of synizesis in Virgil are dande (A. 8.66
and 481), anteirent (A. 12.84), eodemque (A. 12.847): all cases of short e
followed by a long vowel.
373 ' ... and into the words she breathes divine allurement'. dictis,
dative; cf. A. 5.607 ventosque a(d)spirat eunti.
374 ff. Venus' speech. The Servian commentary, which includes scholia
dealing with Virgil as an exponent of rhetorical principles, makes it clear
that this speech is almost a text-book model of a suasoria (a rather more
elementary exercise in declamatio than a controversia).
374-80 This period, whose main sentence (376-8) is encased by two
subordinate clauses (374 f., 379 f.), extends over seven lines. Cicero
(Or. 222) laid it down that the optimum normal length of a period in prose
was the equivalent of four hexameter verses. By far the largest part of
Virgil's poetry consists of units of four or less hexameters. Periods longer
than this in Aeneid 8 group themselves into four well-defined categories:
(1) The exordia of formal, dignified speeches are rhetorically elaborated
in the same way as the openings of prose orationes: 127-33 (Aeneas'
ambassadorial overtures), 185-9 (Evander launching the Cacus-narra-
tion), 314-8 (Evander's excursus on Roman prehistory), 374-80 here
(Venus begins her carefully prepared suasoria), 470-4 (Evander begins his
royal reply to the request for help), 560 ff. (the prelude to Evander's
prayer). Apart from instances in the Cacus narration (which although
spoken by Evander is not as a whole comparable with a speech), the only
example in a speech outside the exordium is the famous passage where
Vulcan's emotion carries him away (400-4).
(2) The [1<</,paai, -r611ov at 416-22 is elaborated into a topographical
excursus (see the note on 597); the £K<ppa.aH<; xp6vov at 26-30 and 407-15
are extended: the first is woven back into the narrative, the second issues
in a very complex concentration of ideas (see the note there).
(3) All five similes in the book (21 ff., 243 ff., 391 ff., 589 ff., 622 f.),
appended to what they illustrate, build up longer units than usual.
n8 COMMENTARY

(4) In the two long narrative passages of the book the four-line unit
length is not exceeded more often than one might expect to be necessary
to produce variety of presentation: 193-7, 225-30, 251-5 (the Cacus-
narration), 630-4, 666-70, 709-13 (the Shield); the rare sporadic examples
elsewhere are no doubt due to the same cause (31-5, 337-41); the special
case of the opening paragraph of the book is discussed in the note on 1-6.
The period at 288-93 (ut prima ... pertulerit) is an indirect presentation of
the seriatim manner of prayers and hymns (cf. the direct presentation
which follows at 293 ff., and G. r.5 ff.).
374 ff. dum ... vastabant ... non ... rogavi: the rule of classical prose,
that a clause introduced by dum, which describes an action or state during
the course of which something else did or did not happen, has an atem-
poral present indicative verb, broke down through the almost inevitable
confusion with the constructions of cum. The imperfect indicative appears
in Sallust, and in Livy, e.g. 10.36.16, dum haec in Apulia gerebantur,
altero exercitu Samnites I nteramnam . .. occupare conati urbem non tenuerunt.
Sec KS 2, 374 f.
375 debita: Servius' explanation Jataliter ad exiti,um destinata is
undoubtedly correct, but this cannot be the meaning of debita alone.
Usually there is a dative in the vicinity making it clear to whom or what
something or somebody is owed (e.g. A. 3.184, 7.120, 11.759); there is
however an almost absolute use of the participle = 'destined' at A.
9.107 f. tempora Parcae / debita complerant, where to be rigorously explicit
one would need to say debita compleri: so here, the sense of debita is defined
by vastari (or vastationi) implicit in vastabant.
In any case there can be no doubt about the meaning for casuras
inimicis ignibus arces is the explicit variation on the theme Pergama . .. /
debita (Pergama = 'the citadel of Troy'; arces is poetic plural). For
debita used absolutely = Jataliter debita in post-Virgilian poetry, see Heuvel
on Statius Th. r.80.
378 labores . .. laborem (380): note the repetition without significance.
378 H. W. Garrod, CR 33 (1919), 105, ingeniously proposed vetitos for
-ve tuos, which supports the sense of vetabant 398, and gives the unexception-
able construction nee volui te exercere labores. But volui exercere te = 'I
wanted to make you work', is not questionable Latin, as Garrod thought;
cf. Terence Ad. 587 i sane: ego te exercebo hodie, ut dignus es.
379 natis: an evasive plural; her 'debt' was to one only of Priam's
sons, Paris, who had awarded her the prize for beauty in the contest with
Juno and Minerva-an incident best veiled in view of its consequences.
Compare natos at A. 2.579 and 138 which seems to allude respectively to
Helen's single child and Sinon's non-existent children, both passages where
emotion is allowed, intentionally or otherwise, to get the better of
arithmetic.
382 f. 'and so it is I (who asked for nothing before) who come as a
suppliant and ask your divine pleasure, which I hold sacred, for arms, a
mother (begging) for her son'. numen and arma are both objects of
rogo.
 COMMENTARY II9

382 f. ergo eadem supplex venio et . .. arma rogo . .. nato recalls Il.
18.457-60 (Thetis to Hephaestus) TOiJVE:Ka vvv Ta aa. yovva8' lKa.voµ.a,, ai K'
'8 ,, 8 I v l' £µ.Cf)
€ €11r,a a
, wKvµ.op'f' ooµ.E:v
A , ,I> ,
aCT1Twa
, '" ,I._,,
Ka,' Tpv.,,a11Hav I Ka,' Ka11as
\ ' KVYJµ.was,
A\:,

ema<{,vplo,s a.papvlas, I Kat BwpTJx·•

382 sanctum mihi numen: a rhythmical rarity; usually two di-
syllables ending a verse are preceded by an emphatic monosyllable (cf.
tune tamen omnis 566). Norden 447 f. lists 13 instances in the Aeneid of a
preceding word of two or more syllables; there is a remarkable concen-
tration in books IO and II (was Virgil experimenting with a harsher
rhythm in the books mainly concerned with fighting?). Here the harshness
is mitigated by the fact that mihi is virtually enclitic (as also at A. 10.442
and 849). In 'I ask arms of a godhead sacred in my eyes' Venus assumes
the tone of a human suppliant praying to a specially reverenced god.
Ribbeck's (sanctum mihi nomen) (' a holy name to me') makes very poor
sense: it is difficult to see why any of the nouns in the sentence should be
holy to Venus, or if any were, what the point could be in stressing her
feeling. Schrader's sanctum tibi nomen makes sense: Venus could well say
that the word 'suppliant' or 'mother' was, or ought to be, respected by
Vulcan; but there is no justification for any change, since sanctum mihi
numen is supported, against the variant nomen and the conjectures based
on it, by A. r.666 ad te confugio et supplex tua numina posco (also Venus-
to Cupid).
383 arma rogo, genetrix nato: in Macrobius r.24.6 f., Evangelus, the
detractor of Virgil, says that he ordered the Aeneid to be burnt at his
death, because he was afraid of the criticism of posterity-against the
situation here, where Venus asks the only god she had ever lawfully
married to make arms for a son of hers, but not of his. The Servian
commentary tries to meet the objection by punctuating heavily after rogo
and taking genetrix nato with what follows, but so long as these two words
stand together, as they must, they inevitably make one think of Venus'
own situation. But Evangelus' criticism is misdirected: Virgil does not try
to make his gods patterns of human behaviour, and here he is more
concerned with the rhetorical effectiveness of quoting precedents and
exempla.
383 f. The 'daughter of Nereus' was Thetis, mother of Achilles; her
words to Hephaestus (Il. 18.457 ff. quoted above) suggested the preceding
sentence. The 'wife of Tithonus' was the Dawn-goddess (Eos, Aurora),
the mother of Memnon, king of Ethiopia. He went to Troy to fight for his
uncle Priam, killed Antilochus son of Nestor, and was himself killed by
Achilles, after which Zeus made him immortal. The legend is alluded to in
Od. 4.187 f. and 11.522, and was treated at length in the epic poem
Aethiopis by Arctinus of Miletus (8th century B.c.), one of the lost poems
of the Trojan cycle. Memnon's armour seems to have been almost as
celebrated as that of Achilles: Dido asks Aeneas to describe it (A. 1.751),
no doubt to supplement the picture her own craftsmen had made of it on
the portico-frieze of the temple of Juno (A. r.489). There was almost
certainly a description of Memnon's arms, or at least his shield, in the
 120 COMMENTARY

Aethiopis; and there are other features of the Aeneid which suggest that
Virgil had direct acquaintance with the content of that poem, probably
through a mythographer's resume, possibly with the actual text: the
major role given to Camilla in A. 7 and II seems to reflect the pre-
dominance of Penthesilea, and Lausus dying to save Mezentius against
Aeneas seems to repeat the situation of Antilochus saving Nestor from
Memnon: see E. Fraenkel, Philologus 87 (1932), 242 ff. = Kl. Beitrage II
(1964), 173 ff.
383 Nerei: for the scansion as a disyllable see the note on 292
Eurystheo. Servius here read Neri and assumed that the nominative was
Neres (like Achilli, genitive of Achilles, see Conway on A. 1.30), but this
form seems to be unattested, and names in -eus form either a Greek genitive
in -eos, or a Latin one either in -ei or, when the synizesis is represented as a
complete fusion of vowels, in -i. See further Leumann 124.
385 f. moenia . .. acuant: two removes from prosaic expression, 'city
walls' for 'cities' for 'citizens'; cf. A. 7.629 f. urbes / tela novant.
387 f. The situation here and in 388-90 may owe something respectively
to fl. 14.346 ~ pa Kai ayKaS' lµ,ap1TT€ Kp6vou 1Ta,s- ~v 1Tapaxomv (where how-
ever Zeus takes the initiative), and fl. 14.294 f. ws o' ioev, ws µ,iv lpws-
1TVKivas- <pplvas aµ,<peKaAU1/)€V, / olov OT€ 1TpwT6v 1T€p lµ,urylu0TJV <piA6TTJTt (where
Zeus is the subject). The phraseology too may overlie Homeric terms:
niveis . .. lacertis recalls AwKWAevos, the almost 'constant' epithet of Hera,
and laeta dolis (393) was perhaps suggested by oo>..o<f>povlouua (fl. 14.329).
Lines 405-6 which close the episode similarly recall Il. 14.352 f.: i:'>s- oµ,Jv
aTplµ,as- €VO€ 1TartJp dva I'apyap't) aKpCf!, / V7nl't) Kai <piA6TTJTt oaµ,els-, lxe a· ayKaS'
a.Koinv.
387 f. dixerat ... / .. . fovet: for the tenses see the note on 219 f. and
compare the note on 152.
388 The emphatic balance cunctantem . .. repente points a line whose
rhythm so aptly reflects its sense: heterodyned spondees for his reluctance
and the fondling embrace which gradually overcomes it, then his electrified
response in a rapid dactylic unit. The bucolic diaeresis (and pause) (see
the note on 24) does not of itself contribute to the effect: it is there simply
to admit the dactylic unit in the last two feet, cf. inde repente 238.
389 so Ii tam . .. notus: 'familiar' and 'well-known', and so intimate
and pleasurable; contrast novus which often means 'new and distasteful'.
solitdm flammdm: the homoeoteleuton here is remarkable: it ornaments
and stresses the two words; see further the note on 15.
389 f. medullas / ... labefacta per ossa: the Roman view of the
sexual excitability of the cerebro-spinal marrow considered as the source
of procreative power is discussed by R. B. Onians, The Origins of European
Thought 2 151.
390 labefacta, 'through his shattered bones' suits the intensity of the
feeling; cf. Lucretius 4.n13 f. usque adeo cupide in Veneris compagibus
haerent, / membra voluptatis dum vi labefacta liquescunt.
390 intravit calor: a pause at the end of the second foot is not
common: it entails a diaeresis, which, emphasised by the sense pause,
 COMMENTARY 121

produces a jolty effect and disturbs the rhythmic flow of the verse. At
A. I.II5 f. in puppim ferit: excutitur pronusque magister / volvitur in caput,
ast illam ter ftuctus ibidem . .. the rhythm is used intentionally to convey
sudden and violent disturbance; the same effect to a lesser degree is
achieved here and at 525 cum sonitu venit (in both places there is a slight
pause). In Virgil this pause is always preceded by a pyrrhic word (of two
short syllables), but is not always descriptive; it would be perverse to read
disturbance into 287, 357, 487, 538.
Note also that the division between words (diaeresis) at the end of the
second foot is accompanied by a caesura within that foot: this is an almost
invariable rule in Virgil (see Raven 96 f.), whose cause is a desire for clash of
ictus and accent in the second foot.
391 olim: 'generally, from time to time', an archaic sense it often has
in Virgil; see Page's note on A. r.289 and R. D. Williams on A. 5.125.
Like quondam it came generally to refer to past time because that was the
most needed concept in its range.
391 f. corusco should be taken with lumine, 'with flashing light'
rather than with tonitru because (1) lumine is far more limp without an
adjective than tonitru; (2) Virgil may well have had in mind Lucretius
6.281 ff. inde ubi percaluit venti vis et gravis ignis / impetus incessit,
maturum tum quasi fulmen / perscindit subito nubem, f erturque
coruscis / omnia luminibus lustrans loca percitus ardor; (3) Page is wrong
in saying that the rhythm of the line forbids the adjective to be separated
from tonitru, for although the interlacing word-order is intricate (see
Appendix, p. 199), the wide separation of lumine and corusco can be easily
paralleled, by, for example, 405 f. placidumque ... soporem and G. r.331 ff.
ille flagranti / aut Atho(n) aut Rhodopen aut alta Ceraunia telo / deicit;
(4) like al6.:\os- and micans 392, coruscus ranges in meaning from' vibrating'
or 'trembling', in terms of motion, to 'glittering' or 'shining' (cf. Servius
on A. 2.172: coruscum alias fulgens, alias tremulum est): either or both will
suit 'light', neither will suit 'thunderclap'.
In 391 note the contribution of t and r to the sound-painting of the
lightning-burst: cf. A. 5.693 ff. (a violent cloudburst); see R. Maxa, WS
19 (1897), 78 ff. for a reasonable attempt to classify individual letters by
descriptive effect.
393 The rhythm of the line indicates a paragraph closure and a long
pause after it is called for in recitation: there is coincidence of ictus and
accent in every foot except the third, the first and fourth feet are self-
contained spondees, and the line ends with the sound-echo c6nscia c6niunx.

394-406 The enraptured Vulcan promises all the resources of his craft to
meet her request, and returns his wife's embrace.
394 pater: a certain example of the word's use as a status-title, with
no implication of parenthood (see the note on 72).
394 This line is one echo, 633 may be another, of Lucretius' description
of Mars in Venus' embrace (r.33 ff.) .. . in gremiitm qui saepe tuum se /
122 COMMENTARY

reicit aeterno devictus vulnere amoris, / atque ita suspiciens tereti cervice
reposta / pascit amore avidos inhians in te, dea, visus. devinctus amore here,
'enslaved by love' eliminates the military overtones, appropriate to
Mars, of devictus (an ill-supported variant here) vulnere.
395 ex alto: 'from the distant past', but with the overtones of the
idiomatic alte (or ex alto) repetita, 'far-fetched, out-of-the-way' (examples
in TLL 1,1783 s.v.).
396 f. As Mackail remarks (Aeneid, Intro. 80) there is no obvious reason
for the identical word-ending fuisset in the two consecutive verses: other
examples, including those with stylistic point, are comparatively rare
(E. 10.53 f.; G. r.407 f., 4.341 f.; A. 7.653 f., 9.544 f., n.204 f.); and for a
noun repeated in a different case (or number) as at A. 8.330 f. Thybris-
Thybrim, cf. E. 2.63 f., 10.75 f.; A. r.349 f., 5.569 f., 7.553 f., 12.617 f.
There is no example of such a repetition in the most finished books of the
Aeneid (2, 4, 6), but it does not necessarily follow that Virgil would have
removed such apparently pointlessly repeated words at the end of a line
in a final revision, since he admits other end-rhymes and also allows the
repetition of a word at close intervals with no significant point. See further
R. G. Austin, CQ 23 (1929), 46 ff. Heinsius' subisset for fuisset 396 is
therefore not compelling, though very apt.
398 f. The 'almighty father' and the 'fates' are different aspects of the
same thing, as usual in Virgil: the will of the world-god is the world's
destiny. This however can be delayed, at least in epic: Juno decides on
delaying tactics at A. 7.313 ff.; but it cannot be altered (so there is no
inconsistency with debita A. 8.375). This rather curious view, which is to
some extent already implicit in Homer, may have been made into theo-
logical doctrine by the Etruscans (Servius here refers to aruspicinae libros).
Lucan also expresses the view that fate can be delayed (3.392, 7.88, 7.295)
or accelerated (5.41, 7.51 f.), a position adopted by the Stoics to try and
deal with the problem of free will.
400 mens est: a monosyllable ending the verse, preceded by another
monosyllable, or disyllable with elision (71 unde est), recalls the practice
and prosaic phraseology of Ennius and Lucretius (3.647 mens est). Virgil
uses it occasionally, reacting against the studied avoidance of it by the
neoterics (there is no example in Catullus 64).
401 ff. 'Whatever care I can promise in my craft ... what can be made
from molten iron or silver-gold ... all the force of furnace and bellows'
blast. .. do not doubt your powers any longer by making requests.'
quidquid . .. quod . .. quantum are grammatically three parallel
clauses (although quod might be relative to the antecedent quidquid), in
breathless asyndeton, each one the subject or object of a verb which should
follow them, but does not, because Vulcan is carried away by his emotion
and substitutes a quite different main clause. As Henry puts it, ' ... in the
midst of his protestation, and just as he was going to say "All this and
more I promise you" or "All this and more I swear to you" he stops short,
breaks off, and laying down the armourer and husband, and assuming the
lover and bridegroom, throws his arms round her waist and simpers:
 COMMENTARY 123

"Demand no more as a favour what you have the right and the power
to command."'
The anacoluthon is not of an unusually abrupt kind: no difficulty is
raised about the comparable construction at A. n.288 ff. quidquid apud
durae cessatum est moenia Troiae, / Hectoris Aeneaeque manu victoria
Graium / haesit et in decimum vestigia rettulit annum, where the quidquid-
clause stands in the same syntactically semi-detached relation to the main
clause as is often the case with the familiar Lucretian quod superest.
Attempts have been made to regulate the grammar and smooth out
the anacoluthon by suggesting that we must understand promitto as
governing quidquid, or id (in apposition with quidquid, etc.) as the object
of precando. These attempts are misguided and based on the assumption
that if a self-conscious stylist using highly ordered language could not
admit anacoluthon unwittingly, he could not do so deliberately either.
But Cicero in his philosophical dialogues does precisely this (see KS 2,
584 f.). Virgil must be assumed to know what he is doing here, even though
characterisation by style (as in Neptune's almost incoherent indignation
at A. r.132 ff. and his almost incoherent rambling at A. 5.804 ff.) is not
one of his main features. (Homer, who is given to it much more, has also
been a victim of misguided grammarians: at Od. 2.45 f. ill' Eµ.ov aOTov
xpefo,,;; µ.o, KaKov £µ.1reaev o'tK</' / oouf, Aristophanes of Byzantium 'cor-
rected' KaKov to KaK&., which regularises the syntax-and reduces Homer's
portrayal of Telemachus' growing indignation to passionless arithmetic.)
402 A versus spondeiazon: see the note on 54 Pallanteum. Four-syllable
words are commoner at the end of such lines than three-syllable in Homer,
Alexandrian poetry, Catullus 64 and Virgil-probably a question of
availability rather than rhythm.
electrum: some silver occurs naturally in gold ore; when the pro-
portion was one part silver to four gold, the mixture was called electrum
(Pliny N.H. 33.(23).80 f.). The same mixture could be made artificially,
in this or slightly different proportions, producing a range of colours
(evident in coins).
403 animae: 'the breath' (of the bellows) = auras 449.
404 indubitare: 'sane indubitare quis ante hunc?' Servius auctus. No
earlier instance is known to TLL s.v. It is a new coinage of Virgil's and a
surprising one: in- is not the usual negative prefix, and the simple verb
dubitare is not used with the dative in the sense of diffidere, which is what
the compound means here. Perhaps the splendid sweep of 400 ff. made
Virgil bolder than usual. The word occurs only once again: Statius Silvae
3.5.109 f. sed ingratus qui plura adnecto tuisque / moribus indubito.
405 dedit amplexus: = amplexus est; periphrases of this type with
dare are common in poetry from Ennius (e.g. sonitum ... arma dederunt
Ann. 415 V) onwards (see TLL 5, 1687 s.v. do).
optatos: 'that he longed for'.
405 f. Virgil's allusion to the act of sexual intercourse produced
interesting reactions from ancient critics (cited by Servius here and Aulus
Gellius N.A. 9.10). M. Valerius Probus, who flourished in the second half
IO EAC
124 COMMENTARY

of the first century A.D., took exception to the impropriety of infusus

gremio and bowdlerised to infusum (with soporem), which produced poor
sense-dormiit cum coniuge dormiente, and worse style-a second adjective
with soporem, but quickly infiltrated into the MS tradition. Annaeus
Cornutus, the Stoic teacher of Persius and Lucan, approved the technique
of the passage but objected to the word membra-because of a possible
double-meaning. It is noteworthy that on a question of ethos Probus,
unquestionably in the front rank of ancient critics of Virgil, shows himself
to no better advantage than the generally obtuse Cornutus (see Nettleship's
introduction to Conington's edition, pp. 61 f. and 65). By the time of
Hadrian fashions had changed: Annianus and other neoteric poets
enthused about the pure delicacy of the treatment.
petivit ... per membra (sua) soporem: a further variation on the
phrase already varied at 30 dedit per membra quietem.

407-423 Vulcan rises energetically in the early hours of the morning and
descends from heaven to his smithy on the island V ulcano.
407-15 'Then when the first rest had driven out sleep in the mid-course
of the now receding night, at the time when a woman, whose burden it is
to support life by the distaff and the slender help of Minerva, first stirs the
ashes and slumbering embers adding night-time to her task, and works her
servants with a long stint by lamplight, so as to be able to keep her
husband's bed chaste and bring up her small sons: in the same way and
no less energetically than she at that time, the lord of fire rises from his
soft blankets to do the tasks of his craft.'
407 ubi prima quies: prima here is best taken as attributive, 'the
first rest', cf. A. 2.268 f. tempus erat quo prima quies mortalibus aegris /
incipit; but in view of Virgil's constant practice, it is impossible to exclude
the possibility that the predicative (adverbial) sense was also present in his
mind, cf. cum . .. primum in the next line. The 'first rest', in the early
hours of the morning, follows the first sleep; this is clear from A. 2.268 f.
above, which follows some time after 253 sopor fessos complectitur
artus.
407 f. medio iam noctis abactae / curriculo: 'in the mid-course of
the now receding night', i.e. night has just passed its 'zenith' and the
third of the four watches has just begun; medio curricula is a locative
ablative of' time when', and the participle virtually has a present (passive)
sense: see the note on 636magnis circensibus actis. Cf. Ovid Met. 10.174 f.
iamque Jere medius Titan venientis et actae / noctis erat, spatioque pari
distabat utrimque.
407 ff. ubi. . . / ... expulerat. .. cum . .. primum / ... suscitat:
after ubi and cum primum ( = simulac, 'as soon as'), as after postquam
and ut, the narrative style of classical prose is content to use an aorist-
perfect to describe an action prior to that of the main verb, and to leave
the priority to be inferred from the context. English insists on making the
sequence clear by using the pluperfect. This Latin tense and also the
 COMMENTARY 125

historic present occur in Sallust and later historians, and not infrequently
in poetry. See Woodcock 172 ff. for a careful analysis of uses and exceptions.
The present indicative after cum primum sometimes has a generalising
force, cf., for example, G. I.II2 f. luxuriem segetum tenera depascit in
herba, / cum primum sulcos aequant sata, where it means 'as soon as' with
strong undertones of 'whenever' (it is impossible in English to convey
both notions without resorting to a cumbersome paraphrase). This general-
ising sense is also present in this passage. Translate: ' ... at the time when
a woman ... first stirs the ashes ... '.
408 f. Periphrases for times of day and night are very common in epic
poetry (see the note on 97). Sometimes (as here) time is defined with
reference to a predictably regular social activity, cf. Od. 12.439 f. IJµ,os o'
£1T, o6p1rov aV1Jp ayopfj8€V avl<TTT/ / Kplvwv vdKm 1TOAAa OtKatoµ,lvwv alt71wv, /
-rfjµ,os 0~ TU. Y€ ooiJpa Xapv/30,os lt€<f,aav871: these are comparatively elaborate
examples of what could be expressed by a simple adverbial phrase,
e.g. {3ovAv-r6vo€ 'at the time of loosing the oxen', It. 16.779; Od. 9.58.
408 ff. This simile weaves together strands from three earlier epic
similes.
(1) Homer Il. 12-433 ff. had likened the deadlock in the fighting to
the equally balanced scales when wool is being weighed: d,\,\' lxov (' they
held on'), ws T€ Ta.AavTa YVV'tJ X€pv'ins aA7181s, I ij T€ <TTa8µ,ov lxovua Ka,
€tpiov aµ,<f,,s avtAKH / luatovu', iva 1raiu,v anKla µ,iu8ov ap71-rm · / ~s µ,iv TWV
£1T, lua µ,ax71 -rtTaTO 1TT6A€µ,6s T€.
Notice that in this, as in many other similes in Homer, there is only
one point of comparison between the illustration and the situation being
illustrated, and it is not this point which has most artistic elaboration
lavished on it (a good example of this is It. 4.422 ff. where the Greek army
advancing silently in successive ranks is compared with a succession of
waves thundering on the shore). Homeric practice must have influenced
Virgil considerably, and this should be borne in mind when evaluating
the 'appropriateness' of Virgil's similes on the basis of the number of
points of comparison. See further David West, 'Multiple-Correspondence
Similes in the Aeneid', ]RS 59 (1969), 40 ff.
Apollonius of Rhodes has two similes describing the activities of a
spinning-woman, both illustrating Medea's emotional state.
(2) Arg. 3.291 ff. Medea's heart is aflame with love: C:,, SJ yvV1J µ,aA€pijJ
1T€p, Kap<f,m X€VUTO OaA<jJ I x€pvfj-ris, -rfj1r€p TaAa~ia lpya µ,lµ,7JA€V, I ws K€V
IJ1Twp6<f,wv VVKTWP ulAas £VTVVatTO, / ayxi µ,a.A' lypoµ,lV7J · TO o' a8lu<f,a-rov Jg
oAlyoio / OaAoiJ av€yp6µ,€VOV CTVV Kap<pm 1rav-r' aµ,a8vvn • / TOLOS V1TO Kpaolv
€lAvµ,lvos aW€To M8prJ / ovAos #Epws.
(3) Arg. 4.1062 ff. night has put all the earth to sleep, but Medea weeps
in agony: ofov OT€ KAw<TTfjpa YVV'tJ TaAa€pyos iAluun / £vvvxl71 · Tfj o' aµ,<f,,
Ktvvp€-rai op<f,ava TtKVa / X71pouvvr, 1r6uios · <TTUAU.€t o' V1TO Oa.Kpv 1rapnas /
µ,vwoµ,lV7Js. 0 i71 µ,iv £1T' uµ,vy€p~ Aa/3€V alua. I Ws -rfjs lKµ,alvov-ro 1rap71lO€S . £V
a, oZ IJ-rop I ot€lvs €lA€LTO 1T€1Tapµ,lvov aµ,<f,' OOVVTJCTtV.
Virgil's simile obviously owes most to (3), although the motif of children
is already in Homer, and that of fire-kindling may possibly have been
10-2
126 COMMENTARY

suggested by (2), although the phraseology cinerem et sopitos suscitat ignis

is an exact repetition of A. 5.743. It looks as if Virgil, about to describe
supernatural activity at night, was led to think of the favourite locus
communis on human restlessness, and so remembered (3). If this is so, he
has treated it in the same ways as the earlier simile based on Apollonius
(see the note on 22 ff.): he has dignified the status of the yvvi} Ta.AaEpy6s by
providing her with maidservants (and making her virtually a Roman
matrona), and declined to be any more explicit than his source about the
cause of her solitary state. Virgil's language is closely imitated by Valerius
Flaccus 2.136 ff. in his description of Eurynome, whose husband is away
on active service; Virgil may well have had the same situation in mind
(Henry insists that he did).
There is a more important debt too which Virgil owes to Apollonius (and
his Hellenistic contemporaries). The plight of nameless commoners rarely
roused feelings of pathos or sympathy in the author(s) of the Homeric
poems (a notable exception is Od. 20.109 ff., the woman who is still at her
job of grinding corn when dawn breaks because her health had been ruined
by the heavy demands of the suitors). The description of ordinary people
with realism and humanity was a development fostered by Hellenistic
writers and no doubt contributed to Virgil's own deeper and more compre-
hensive humanitas.
409 tolerare ... vitam: 'to support life', cf. Luer. 2.n71 tolerarit ...
aevum.
tenuique Minerva: Minerva = lanificium here, a common metonymy
(see the note on 181). In Roman religion Minerva was the patroness of
handicrafts and of manual workers (including doctors and school-teachers);
Virgil associates her particularly with the feminine activity of spinning
(cf. A. 5.284, 7.805 f.).
tenui is ambiguous, meaning either 'finely spun' or 'meagre'. Servius
favours the latter, parvo pretio lanificii, which fits well with the a.ELKla
µ.tu8ov of Il. 12.435 which Virgil may have had in mind here (see the note
on 408 ff.). Both meanings may be present, as at A. I.I a gracili modulatus
avena, where gracili means both 'slender' and 'unambitious'.
410 impositum: some Roman readers, confronted with an un-
punctuated text, not surprisingly took impositum to agree with cinerem:
the ashes were put over the embers to keep them burning low, cf. Horace
Odes 2.1.7 f. incedis per ignes / suppositos cineri doloso. This, according to
Servius, led them to punctuate at the end of 409 and in that line to
'understand' mos est or placet. Bentley, realising the difficulty of any such
'understanding', decided to mutilate the text at 408 f. (Jemina, cui mos /
est tolerare) to make the same interpretation explicit. But the commonly
accepted interpretation makes good sense without raising any serious
difficulty: the omission of est in the subordinate cui-clause is easily
paralleled (see the note on 52), and that impositum should be detached in
sense from what follows is strongly suggested by A. 5.743 haec memorans
cinerem et sopitos suscitat ignis.
4n f. Spinning as a morally commendable feminine activity had no
 COMMENTARY 127

necessary connection with social status: it is done both by the noble

Lucretia (Ovid Fasti 2.741 ff.) and by the slave-girl who says on her
epitaph gravitatem officio et lanificio praestiti (Biicheler, Carmina Epi-
graphica 63.4).
414 nee segnior (illa understood, comparison) illo tempore (time).
416 ff. insula . .. erigitur . .. 423 hoc: for the elaborated opening and
development of this iK<f,pacns T611'ov see the note on 597 est ingens . .. lucus.
It follows almost immediately after an £K<ppaais xp6vov, for the basic point
of the simile in 408 ff. is to elaborate the time of Vulcan's rising (414 nee
tempore segnior illo). In fact from an austere and prosaic point of view the
whole verse paragraph 407-23 contains only two short sentences which
forward the main narrative (414 f., 423), each at the end of a long poetic
footnote.
416 Aeoliamque: polysyllabic Greek proper names at the end of the
verse and not involving a spondaic fifth foot are common (see Norden
438).
416 f. Lipari is the largest and most important of the group of seven
islands off the north-east coast of Sicily, the Al6.\ov l"ijaoi (Thuc. 3.88),
insulae Aeoliae, the domain of Aeolus god of the winds. The only island
between it and the Sicilian coast (Sicanium ... latus) is 'IEpd. 'Hcf,alGTov,
Volcania ... tellus (422), the modern Vulcano. The volcano, rising straight
out of the sea (erigitur 417), still smokes, but feebly and undramatically
compared with Strongyle (Stromboli), and not enough to discourage
growing tourist traffic to its black sands and sulphur springs. The author
of the Aetna (about the middle of the first century A.D.) says that the
island could safely harbour ships because volcanic activity had largely
died away and would have become quite extinct but for a supply of
combustible material from Etna itselffurtim, underground (444 f.).
Of the two neighbouring islands, Valerius Flaccus r.579 ff. repeats the
traditional allocation of Volcania to Vulcan and Lipari to Aeolus. The
suggestion that Lipare (nom.) should be read here, with Aeoliam referring
to the Italian coast as conjectured at Lucretius 1.721, is misguided.
416 ff. In G. 4.170 ff. Virgil had located the smithies of the Cyclopes
under Mount Etna in Sicily, and there is a vestige of this tradition in
antra Aetnaea (419), and Aetnaei Cyclopes (440). Here he situates the
smithies on the island Vulcano, no doubt following Hellenistic tradition
(cf. Roscher s.v. Hephaistos col. 2070 ff.): Callimachus (Hymn to Artemis
46 ff.) chose Lipari itself for the site, and also seems to have been the first
to have brought the Cyclopes quite clearly into connection with He-
phaestus (cf. Roscher, s.v. Kyklopen, esp. cols. 1678 ff.). The colourful
detail in Callimachus' description of their activities was no doubt in
Virgil's mind when he wrote 424 ff.
Virgil's choice of Vulcano may have been prompted by recent volcanic
activity there, although this was feeble and intermittent by the first half
of the first century A.D. (cf. Aetna 440-2).
417 Vulcano which 'rises sheer' (erigitur) from the sea recalls the floating
island of Aeolus in Od. ro.4: .\iaC17J S' ci.vaSISpoµ.E 11'1Tp7J.
128 COMMENTARY

418 exesa caminis: Servi us auetus asks exesa ad usum eaminorum . ..

an eaminis exesa pro ignibus exesa? The second is more natural and prefer-
able, 'eaten out by the forges'.
419 antra Aetnaea: 'caves like those of Etna', fiery with volcanic
activity. But in spite of Heyne's denial (ad loe.), Virgil was probably also
thinking of them as belonging to Etna in the sense that it was Etna which
supplied them with volcanic material (cf. Aetna 445 ff.).
419 f. validique incudibus ictus / auditi referunt gemitus: 'and
mighty blows on the anvils are heard and re-echo their groans'. The
'groans' are of course inanimate as at A. 2.53, 3.555, 9. 709.
420 f. striduntque cavernis / stricturae: on A. 10.174 Servius quotes
Varro as saying nasei illie (on the island of Elba) ferrum, sed in strieturam
non posse eogi nisi transveetum in Populoniam, and from this it appears that
strietura originally referred to the mould into which the molten metal
flowed from the furnace; here it probably means the ingot or 'pig'
itself. The red-hot ingots hiss (stridunt) as they are formed in the
moulds.
421 The Chalybes were reputed to be a Scythian tribe beyond the river
Thermodon which flows into the Black Sea (Pontus); cf. G. 1.58 f. at
Chalybes nudi ferrum, virosaque Pontus / eastorea (sc. mittit). They were
credited with the discovery and first working of iron, xa.>..vip. For this they
are cursed in Callimachus Aetia frg. no, 48 ff. ( = Catullus Coma Bereniees
48 ff.); Apollonius (2.1005 ff.) represents them as labouring unceasingly in
black sooty smoke.
423 hoc, the older form of hue (cf. illoe, istoe) occurs only here in Virgil
but is guaranteed by Servius ad loc., and by the grammarian Priscian;
the facilitation to hue, a variant here, has occurred also at, for example,
Lucan 2.625. The occurrence of the two forms is a good illustration of the
struggle between classical conformity and the persistence of older forms in
colloquial speech: with verbs of motion Cicero uses only hue in his Letters,
while many of his correspondents, including Caelius, use hoc; Caesar
invariably uses hue, while his continuators, especially the author of the
Bellum Hispaniense, admit or prefer hoc: see Wolffiin, ALL 7 (1892),
332.

424-438 In the cavern the Cyclopes are busy forging a thunderbolt for
Jupiter, a chariot for Mars and an aegis for Pallas.
425 Brontesque Steropesque: Virgil is very fond of lengthening the
-e in the first -que of a pair by placing after it a word beginning with two
consonants (as here), a double consonant, or a single consonant capable of
prolongation (l, r, s); the second -que is always either elided or scanned
short. This is frank imitation of Homer's practice with TE ••• TE, e.g. at
It. 4.295 f. aµ,<f,, µ,lyavlIEAayovra 'A>.aa·TOpa 'T°f: Xpoµ,loVTl I AZµ,ova'T°f: Kp€lovra
Btav'Ta 7"€ 1Totµ,lva Aawv. Hesiod Theog. 140 Bp6V'TTJV 'T€ l:T€p67TTJv 7"€ ••• is the
immediate source of the phrase here. As in Greek epic, the lengthened
syllable is always the first in its foot (and therefore necessarily bears the
 COMMENTARY 129

ictus). This is the only example of this licence in this book (for others see
Page's note on A. 12.89). See also R. G. Austin on A. 4.146.
425 nudus membra: an accusative of respect of a noun with an
adjective, as here, is an innovation of Augustan poetry modelled on a
familiar Greek construction (compare Ovid Met. 9.307 ftava comas with
Homer Od. 15.133 Ka.p71 gav8os MEviAaos). The noun almost always designates
a part of the body. This 'Greek' accusative with adjectives (as opposed to
participles) makes its first appearance in literary prose with Tacitus
(e.g. Germ. 17.3/eminae ... nudae bracchia et lacertos) no doubt as a result
of Virgil's influence; but according to Quintilian (9.3.17) Virgilian cliches
like saucius pectus were already a part of Roman journalese. See also the
note on qui genus (114), and R. D. Williams on A. 5.97 and 135.
425 Pyracmon: Virgil's sensitivity to etymology is most evident with
proper names, and this one was invented by him. Hesiod's trio of Cyclopes
(T heog. 140) were Bp6vTTJS ('Thunderer'), l:TEp67rrJs ('Lightner') and u ApY']s
('Flasher') and the same names occur in Callimachus Hymn to Artemis
68, 75. Just after the naming of the Cyclopes in Hesiod, there is explicit
mention of the origin of that name itself (by an interpolator?), Theog.
144 f. KvKAW1T€S o' ovoµ.' IJaav Jmovvµ.ov, OVVEK' a.pa acplwv I KVKAoTEp~s ocf>BaAµ.os
£ELS ivEKELTo J-LETw1rqi. Homer had given invented significant names to his

minor characters: cf., for example, Od. 8.1n ff. and especially n4 'Aµ.cpt-
a>.6s 8', vios IloAvVTJOV TEKTovtoao, 'Sea-girt, son of Many-Ship, son of
Craftsman' (see Stanford, Odyssey 1-xn, p. xxi); and Kretschmer (Glotta
28 (1939-40), 253) has suggested that Homer actually chose the form
'OovaaEvsin preference to 'OAvaaEVs or 'OAvnEVs because of the etymological
connection he could devise with dovaam8ai (evident in Od. 1.60 ff. 'Oova-
aEvs . .. clio-6aao and Od. 19.407 ff.); and see R. Pfeiffer, History of Classical
Scholarship (Oxford, 1968), 4 f. So there was good Greek precedent for
Virgil's coinage of the Greek Pyracmon (Ilfip•a.Kµwv, 'Fire-Anvil'), which
was later taken up by Statius (Theb. 2.599) and Claudian (De Rapt. 1.240;
De Cons. Hon. III 195). Similarly at A. 12.391 f. Phoebo antealiosdilectus
Iapyx / I asides, the physician Iapyx may owe his patronymic as much to
Greek laaBaL, 'to heal', as to any putative connection with Iasius (like
Palinurus, A. 5.843). Compare Ovid Met. 3.206 ff. for dogs with appropriate
Greek names, including acutae vocis Hylactor ('Barker'), and Met. 2.153 f.
for the significant names of the horses of the Sun, Pyrois, Eous, Aethon
and Phlegon.
Rather different and more frequent are cases where Virgil shows that he
is conscious of the etymology of a foreign proper name: A. 12.539 f. nee di
texere Cupencum / Aenea veniente sui, where Servius ad loc. tells us that
cupencus was the Sabine word for 'priest'; A. 7 .684 H ernica saxa, 'the
Rocky rocks', for hernae was the Sabine equivalent of Latin saxa (Servius
ad loc.). Often by a jeu etymologique Virgil recasts in Latin adjectives the
meaning of a Greek name: A. 6.550 f. quae rapidus flammis ambit torrenti-
bus amnis / Tartareus Phlegethon ( ~ cpAEyiBELV, 'to scorch'). See further
R. D. Williams on A. 3.693 and 5.2 (and his very useful references there).
Later epic also has examples: Lucan 1.214 puniceus ( !) Rubicon; 6.369 f.
130 COMMENTARY

quique nee umentes nebulas nee rore madentem / aera nect enues ventos
suspirat Anauros; Val. Fl. 6.roo f. hiberni qui terga Novae gelidumque
securi / eruit et tota non audit Alazona ripa.
For the (again rather different) folk-etymology of Latin names see the
note on 322 f. Latium; and for Acmon as a fragment of Pyracmon the note
on 250 molaribus.
426 f. Virgil elaborates (with numerical precision 429 f.) his description
of the forging of the thunderbolt from Ap. Rh. Arg. 1.730 ff. (one item of
the eip.broidered border of Jason's cloak): 'Ev µJv laav KvK>..wm,<; br' dcf,8lTcp
i/µ€110, lpycp' I Zriv, K€pavvov avaKn 7TOll€Vp.€110, • O<; T6aov rfori I 7raµcf,al11w11
E'TE'TVK'TO, µ,fj<; o' fr, 0€V€'TO µoiJ11011 I 0.K'TtllO<;, TTJII oiy€ a,01Jp€lr,,; eM.aaKOII I
acf,vpfja,11, p.a>..€p0ta 7TVpo<; '€fovaa11 O.V'Tp.~11.
426 his (dative); manibus (ablative).
informare (as at 447) = 'to give shape to', a favourite word with
Cicero in the sense of 'to sketch' (e.g. the ideal orator, Or. 7.33); so
informatum = 'roughly shaped' (as opposed to informe (264), 'shapeless').
426 polita: for the meaning, here perhaps 'burnished', see the note
on 436 polibant.
427 fulmen, 'a thunderbolt' (forked lightning), of which the fulgores
(431), 'flashes of lightning', are one of the ingredients: cf. Seneca N.Q.
2.57 .3 fulgur quod tantum splendet . . .fulmen quod mittitur (v.l. incendit);
Ovid Met. 1.56 et cum fulminibus f acientes julgura ventos; and Lucan
4. 77 f. nee servant fulmina flammas / quamvis crebra micent: exstinguunt
fulgura nimbi.
427 fulmen . .. quae plurima: 'a thunderbolt, of the kind which the
Father hurls down in profusion on to the lands from the whole sky'. For
the construction Heyne compares Od. 5.421 f. KfjTo<; • •• olci 'T€ 7To>..>..a Tplcf,H
KAVTo<; 'Aµcf,,Tpl-rri; cf. also A. 7.199 f. seu tempestatibus acti / qualiamulta
mari nautae patiuntur in alto, and Cicero Mil. 9 si tempus est ullum ... quae
multa sunt.
427 toto . .. caelo: to explain toto Servius cites ancient astrological lore:
ab omni parte caeli: nam dicunt physici de sedecim partibus caeli iaci
fulmina. Pliny N.H. 2.(55).143 attributes this division to the Etruscans (in
sedecim partes caelum in ea spectu divisere Tusci), and this has been
strikingly confirmed by the discovery of the bronze liver at Piacenza the
surface of which is divided up to represent these divisions of the sky
(cf. Pallottino 163). The phenomena which occurred in each of the sixteen
divisions were connected with a particular deity (cf. Martianus Capella
1.45 ff.); Jupiter, Diespiter, the supreme sky-god, might be thought to
manifest his will in any part of the sky, hence, on this interpretation, toto.
Donatus (Vita 15) tells us of Virgil inter cetera studia medicinae quoque
ac maxime mathematicae operam dedit, where (ars) mathematica included
especially astronomy and astrology, and it is quite possible that Virgil had
learned about this division of the sky from Varro or Nigidius Figulus. But
it is hardly plausible to read a specific reference to this into the present
passage; at the most Virgil is stressing the paramount power of Jupiter
as sky-god.
 COMMENTARY 131

The same phrase toto . .. caelo (ablative of place without a preposition

as is usual with totus, universus, cunctus in prose as well) occurs in a
similar connection at G. 1.474 f. armorum sonitum toto Germania caelo /
audiit, where 'the sound of arms' is of course thunder.
429 imbris torti: 'pelting shower'; tortus unqualified suggests the
usual sense of torquere = 'to hurl'.
430 Austri: the rain-bearing South wind is a natural ingredient; cf.
Lucretius 5.745 auster fulmine pollens.
432 flammisque sequacibus iras: 'bursts of anger with pursuing
flames': the wrath of Jupiter pursuing guilty victims.
433 f. currumque rotasque volucris / instabant: 'were busy at
work on a chariot and flying wheels'. instare in this sense is normally
construed with a dative, cf. A. 1.504 instans operi regnisque Juturis. Servius
auctus calls its use here with an accusative nova elocutio: the only other
example known to TLL s.v. is Novius (Atell. 61) instat mercaturam, spero,
rem faciet: frugi est homo (si vera lectio !). The construction is easily under-
standable as modelled on the analogy of, for example, parare, and Virgil
probably ventured it because of the presence of one dative in the line
already (Marti). (At A. 8.537 instant has the quite different sense of
impendent.)
435 aegis is a transliteration of the Greek aly{, which (as at Aesch.
Choeph. 593) means a storm-cloud with its thunder and lightning, the
appropriate weapons of the sky-god Zeus or Jupiter. But by the time of
Homer the alyls had become represented (through the demands of visual
art?) as an item of personal equipment: first, through the supposed etymo-
logical connection with at{, 'a goat-skin', as a shaggy goat-skin (as
apparently at A. 8.354), then as a protective covering (in the present
passage a fully elaborated armoured breastplate). Then, as a piece of
equipment, it came to be regarded as the cause of thunder and lightning
when wielded by the sky-god, and as his possession (made for him by
Hephaestus, Il. 15.308 ff.). But in Greco-Roman, as in Nordic mythology,
the sky-god (Zeus, Jupiter; Wotan) has a warrior daughter (Pallas
Athene; Brunhilde); and Pallas Athene too has the use of the aegis, cf.
Aesch. Eum. 826-8 (Athene speaks) Kayw 7Ti7To,0a Z71vl, Kat -rt 8e, Uyeiv; I
Kat KAfj8as ol8a 8wµ,a-ros µ,6v71 0ewv / EV t[, KEpavv6s EUTW Jacf,pay,aµ,ivos.
She is named as its wielder in the present passage (and cf. A. 1.42 ff.),
as at Jl. 5.738 ff., 2.446 ff., 21.400 f. (It is more exceptionally wielded by
Apollo at Il. 15.229, 15.307 ff. and 24.20 f.)
Here, as often in art, the aegis has the Gorgon's head in the centre:
very appropriately, for the Gorgon was one of the most general symbols of
blinding lightning and deafening thunder (hence probably the myth of the
Gorgon's glance turning men to stone, which represented in the language
of myth the highest degree of paralysed shock caused by lightning and
thunder of great intensity). The same association underlies and explains
the problematic passage at A. 2.615 f.: iam summas arces Tritonia, respice,
Pallas / insedit nimbo effulgens et Gorgone saeva, where in rather com-
pressed language Virgil is surely alluding to the aegis and its associations.
132 COMMENTARY

See further Roscher 1. i. 149 ff. s.v. Aigis, and 1. ii. 1698 f.
435 turbatae: pro turbantis Servius, favoured by Mackail, but un-
likely. It means 'when roused, angered': so Heyne, citing the first Silius
passage quoted on 438.
436 squamis ... auroque: = squamis aureis, by hendiadys: 'with
serpent scales of gold'; in keeping with the general paratactic tendency of
his style, Virgil is extremely fond of linking nouns together with a simple
copula (et, -que, atque), although their relationship to each other may be of
a number of different kinds. It may be helpful to distinguish the following
categories even though it is often impossible to allocate any particular
example to one category exclusively: (1) at one end of the scale there are
cases where neither of the conjoined nouns can be properly or completely
understood without the other (e.g. 177 taro et villosi pelle leonis; 368 Joliis
et pelle Libystidis ursae; 694 stuppea flamma manu telisque volatile ferrum);
(2) at the other end of the scale the nouns are synonymous, and all but one
are strictly dispensable without impairing the sense (e.g. 88 stagni placidae-
que paludis; 487 sanie taboque; 500 flos veterum virtusque virum). In between,
one finds (3) examples where one of two nouns is really descriptive of the
other and performs the function of an adjective (as here), or (4) instances
where one noun is explanatory of the other (e.g. 289 monstra ... geminosque
anguis; 531 sonitum et divae promissa parentis): in this case the copula is
often referred to as epexegetic. Finally (5) there are instances where a
copula joins the name of a class and a conspicuous member of it, e.g. 330 reges
asperque . .. Thybris; 698 omnigenumque deum monstra et latrator Anubis.
Before Virgil the spreading of the same idea over a series of correlated
nouns had been indulged in by Lucretius 'almost ad nauseam' (Conway on
A. 1.54): e.g. 1.147 non radii solis neque lucida tela diei.
436 polibant: the imperfect indicative of fourth conjugation verbs in
-ibam (perhaps influenced by ibam itself from eo) came to look old-fashioned
against the standard form in -iebam. Virgil's general practice is to use the
-ibam form when the alternative would not have been metrically possible
at any part of a hexameter verse, cf. vestibat A. 8.160; nutribant A. 7.485,
nutribat A. 11.572; A. 7. 790 insignibat; A. 10.538 redimibat. poliebant on
the other hand would have been possible in a hexameter verse, although
certainly very cumbersome.
The English derivative 'polish' has a much narrower range of meaning
than the Latin polire, which meant 'to make neat and tidy', hence 'to
smarten, adorn'; noteworthy is its connection with cultivated land, cf.
Varro R.R. 3.2.5 polito cultura (!undo); and Nonius 66.18 citing Ennius:
'Politiones' agrorum cultus diligentes, ut polita omnia dicimus exculta et ad
nitorem deducta . .. Testes sunt / lati campi quos gerit Africa terra politos
(Sat. IO f. V), and cf. Ann. 319 f. V causa poliendi / agri.
438 '(and) the Gorgon (herself) with rolling eyes and severed neck'.
vertentem, 'rolling' for the more usual torquentem: cf. Silius I talicus'
reminiscences, 9.460 ff. tum virgo, ignescens penitus, violenta repente /
subfudit flammis ora atque, obliqua retorquens / lumina, turbato superavit
Gorgona vultu; 4.232 ff. inferias caesis mactat Labarumque Padumque / et
 COMMENTARY 1 33

Caunum, et multo vix fusum vulnere Brencum, / Gorgoneoque Larum

torquentem lumina vultu. The lifelike Gorgon made by the Cyclopes recalls
the animated maidservants created by Hephaestus in Il. 18-417 ff.

439-453 Vulcan orders the Cyclopes to suspend all their tasks, and they turn
their gigantic energy to making a shield for Aeneas.
441 usus (est), 'there is need of'; after the model of utor the noun
usus was also construed with an (instrumental) ablative, and then this
construction attached itself to the phrase of equivalent meaning, opus est.
The two were used as virtually interchangeable in Old Latin, cf. Plautus
Most. 250 f. specielo ei (sc. mulieri) usus est: quid opust speculo tibi?, and
A mph. 505 citius quad non facto est usus fit quam quad facto est opus. The
ablative with these expressions was no doubt reinforced by the analogy of
the ablative (of separation) after verbs of lacking, egeo, careo, etc. opus est
superseded usus est in classical Latin; the latter is entirely confined to the
old language and does not occur in Cicero (Ad Att. 9.6.3, navis quibus usus
non est omnis aut praecidisse aut incendisse dicunt is a citation from an
anonymous despatch); this seems to be the only example of usus bearing
precisely this meaning of 'need' in Virgil, and it has all the air of a
deliberate archaic revival; cf. also Livy 30.41.8 reduceret naves, quibus
consuli usus non esset. See further KS 1, 387 f.
443 Note the sense pauses in the line: speech-endings not infrequently
terminate at the strong caesura in the third foot (e.g. A. 1.370, 9.22, 11.98).
A pause after the trochee in the fifth foot often occurs in Virgil to mark a
halt before a transition to a sharply contrasting or antithetical notion:
with ejfatus, at illi here compare in their contexts at illum G. 1.242 and
G. 4.360, and sed illos G. 1.225. All other instances of this pause in this book
are associated with vocatives (as often elsewhere): 188, 293, 578, 613, 643,
668. The one exception, anomalous and unparalleled, is 533 as printed by
some editors: it is highly questionable (see the note there).
443 ff. 'And he said no more; but they all lay to the work with speed
and an equal division of labour.'
446 f. Descriptive verse at its best: in 446 the hissing and clanking of
the smithy (s and c(h), qu-); in 447 the pounding beat of the hammers
(ingentem . .. informant), the effort and labour (heterodyned feet, mostly
spondees ingEntem cl'ipe(um) inf5rmant, and two harsh elisions clipe(um)
informant, un(um) omnia). vulnificus is perhaps a new coinage, see the
note on 82.
448 f. orbibus orbis / impediunt, transferred from the completely
different context of the lusus Troiae (A. 5.584 f.), here= 'they firmly
fasten the round layers to each other'; there were seven layers, as with
Tumus' shield (A. 12.925). septenos = septem, see the notes on 47, 168.
orbibus (ablative of instrument) and orbis refer to the same thing (pace
Henry and Warde Fowler); a noun is often repeated idiomatically in
expressions of reciprocity, cf. 692 mantis concurrere montibus for mantis
inter se concurrere.
1 34 COMMENTARY

449-53 = G. 4.171-5 with a few changes, notably antrum for Aetna to

suit the new mise-en-scene on Vulcano. Another surprising transference:
Virgil's more frequent custom is to use material which had appeared as
narrative in the Georgics for similes in the Aeneid (cf., for example, G.
3.232-4 with A. 12.103-6); here the reverse is true.
451 lacu, 'water-trough', see the note on 74, and cf. Ovid Met. 9.170
ipse cruor, gelido ceu quondam lamina candens / tincta law stridit. stridentia
450 is of course proleptic.
452 f. The idea of reinforcing the sense of the lines by a rhythm
suggesting mighty labour and controlled effort was no doubt inspired by
Callimachus' description of the same scene (Hymn to Artemis 59-61): £v8'
ory£ paurrijpac; a.npaµ£VOL VTT£P wµwv / ~ Xa.AKOV {dov-ra KaµLv68£v ~£ alOT)pov /
a.µfJo>..ao1c; ( = EK OLaOoxi)s = in numerum (453)) T£T07TOVT£<; ETT1 µiya µvx8ta-
U£Lav. But the rhythmic units themselves are quite different: in each of
Callimachus' first two lines initial spondaic effort eases itself half-way
through the line, and in the last the rapid dactylic clatter is arrested by the
two labouring spondees at the end. Virgil's first line is entirely spondaic
apart from the fifth foot, but the rhythmic effect of the line is not due
simply to the use of spondees, it depends more on their placing in the line.
There is another famous line used in connection with the Cyclops which is
also spondaic except for the fifth foot: A. 3.658 monstrum horrendum,
informe, ingens, cui lumen ademptum. The sound of this line suggests
gigantic, slow, irresistible movement, and this is due to the fact that every
word in the line except ingens is placed in such a position that the natural
word-stress (accent) is reinforced by the metrical stress (ictus). In 452 on
the contrary, inter sese multa vi is so placed as to cause a conflict of accent
and ictus: hence the suggestion of conflict and resistance in the sound,
and a total effect of gigantic effort laboriously overcoming great resistance.
In Virgil's second line (453) of alternating dactyls and spondees the
struggle has been overcome (for the moment at least) and the well-directed
effort proceeds without hindrance, for in versantque tenaci forcipe massam
accent coincides harmoniously with ictus.
For rhythmic effect comparable to that of 452, cf. G. 2.526 inter se
adversis luctantur cornibus haedi; A. ro.146 f. illi inter sese duri certamina
belli / contulerant; A. 12. 720 illi inter sese multa vi vulnera miscent (of two
bulls fighting for mastery).
453 in numerum: 'rhythmically', numerus = 'the measure' in
music; cf. E. 6.27 f. tum vero in numerum Faunosque ferasque videres /
ludere (and Lucretius 2.631 and 637). The author of the Aetna refers to the
material of this passage with scientific disapproval (36 ff.): discrepat a
prima facies haec altera vatum: / illis Cyclopas memorant fornacibus usos, /
cum super incudem numerosa in verbera fortes / horrendum magno quaterent
sub pondere fulmen / armarentque I ovem: turpe et sine pignore carmen.
 COMMENTARY 1 35

454-468 Meanwhile dawn breaks, and Evander and Aeneas rise immedi-
ately and meet for the appointed parley.
454 ff. Knauer (254 ff.) rightly analyses the Homeric subtexture of this
'second day with Evander '. The broad scheme of events follows those of
Od. 3.404 ff.: at dawn guest (Telemachus, Aeneas) and host (Nestor,
Evander) meet to confer; the host offers the escort of his son (Peisistratus,
Pallas) to the place where help may be obtained (Sparta, the Etruscan
camp); the guest, who arrived by ship, departs on a horse (Aeneas) or in a
chariot (Telemachus) he has been given. But important details are woven
in from Odyssey 2 and 4, and from the Iliad. Evander is accompanied by
hounds (461 f.) like Telemachus (Od. 2.n); Evander's dressing (457 ff.) is
described in detail, like that not of Nestor but of Telemachus himself
(Od. 2.1-4) and of Menelaus (Od. 4.306-9); the omen just before departure
(523 ff.) has no parallel in the situation at Pylos, but there is an omen just
before Telemachus leaves Sparta (Od. 15.160 ff.), interpreted (cheeringly)
by Helen. Evander's prayer for his youthful strength recalls the Iliadic
Nestor (see the note on 560 ff.). The description of the Shield of Aeneas
which closes the book is the only extensive structural derivation from the
Iliad (the Shield of Achilles) in Aeneid 8, apart from the visit of Venus
(Thetis) to Vulcan (Hephaestus) (see the note on 370 ff.). The Catalogue of
Latin Allies at the end of Aeneid 7 clearly overlies the Catalogue of Ships
in Iliad 2, and in this respect Aeneid 7 opens the' Iliadic' Aeneid. But the
Homeric themes which underlie Aeneid 8 are predominantly Odyssean,
and become Iliadic only with the emergence of the Venus-Vulcan-armour
motif. As a whole the eighth book depends on a more complex inter-
weaving of Homeric themes than any other book of the poem. In it the
point of transition from 'Odyssey' to 'Iliad' may lie (Knauer 255 ff. is sure
that it does) in an 'identification' of Helen's gifts to the parting Tele-
machus (Od. 15.125 ff.) with Thetis' gifts to Achilles (Il. 19.ro ff.).
454 pater . .. Lemnius: Lem nos cara deo, nee f ama notior A etne / aut
Lipares domus: has epulas, haec templa peracta / aegide et horrifici formatis
fulminis alis / laetus adit, Val. Fl. 2.95 ff.: the cult of Hephaestus (Vulcan)
on Lemnos no doubt originated from its now extinct volcanic mountain,
Moschylus, where Vulcan's smithy is located and described by Valerius
Flaccus 2.332 ff. Like Vulcano (see the note on 416 f.), the island could be
described as 'Hcpalcrrov l€p&. (scholiast on Ap. Rh. 1.851); it was there that
he fell when hurled from heaven by Zeus for taking sides with his mother
Hera (Il. 1.590 ff., and cf. Milton P.L. 1.742 ff.), and there was his favourite
haunt (Od. 8.283 f.).
454 haec ... properat: this transitive use of properare is poetic, cf.,
for example, [Tib.] 4.1.205 seu matura dies celerem properat mihi mortem,
and does not appear in classical prose.
454 Note the very delayed position of dum, in a line with a symmetrical
word-pattern: see the note on 22.
454 dum. . . The conditions of oral composition and delivery made it
almost impossible for Homer to envisage or describe separated events as
136 COMMENTARY

taking place simultaneously (see W. B. Stanford, Odyssey 1-xn, 292). For

Virgil it was quite possible, although this is the only example in this book
(interea at 280 is simply an indication of time and involves no change of
scene). Homer often introduces a change of scene by ai'mJ.p (as, for example,
at Od. 24.472), and Virgil follows this convention (instead of making the
contemporaneity explicit) at 370 and 608 At Venus . ...
455 lux ... alma: almus, a sacral word, basically means 'fostering' (as
opposed to alumnus, 'fosterling', both ~ alere), or 'nurturing' (alma
Ceres G. 1.7), and comes to be applied to any beneficial power or force of
nature, especially daylight (lux here and A. 1.306, 3.3n, 11.182 f.; dies at
E. 8.17, A. 5.64); it is obviously appropriate to mothers (cf., in their
contexts, Venus at A. 1.618, and Cybebe at A. 10.220); alma Phoebe at
A. 10.215 f. means little more than 'the kindly moonlight'.
Wakefield's rewriting of the lines (Euandrum ex humili lecto lux suscitat
alba / et matutinos volucrum sub culmine cantus) is quite uncompelling,
although it meets the problem of the tectum humile (see 461).
456 volucrum: cocks or swallows? Critics since antiquity have solemnly
disputed the pros and cons; Wagner even seems to have done some field
research among German rustics to try and find out. Servius auctus was un-
doubtedly right in saying that Virgil avoided specific names propter
sermonis humilitatem; the everyday material details of Homeric realism
or the ostensibly didactic Georgics were not suitable for the epic Aeneid.
Silius' reminiscence (14.22), the Straits of Messina are reputed to send
across the barking of dogs et matutinos volucrum tramittere cantus, must
presumably refer to the crowing of cocks.
457 tunicaque inducitur artus: inducitur in this context is a syno-
nym of induitur, and this, as a member of the class of verbs of dressing and
its opposite, is Latin's only trace of a native middle voice (the form of
course is indistinguishable from the passive); artus here therefore is the
accusative direct object of the middle (or semi-reflexive) verb inducitur,
and tunica is an ablative of instrument. See the note on 29 turbatus
pectora. English would turn the expression round and translate 'and he
draws his tunic over his limbs'.
458 'and binds the Etruscan straps round the soles of his feet'. 'It
matters little whether we take pedum with vincula or with plantis ',
Conington; word-symmetry, with a possessive genitive taking the place of
the usual adjective, favours taking pedum with plantis.
Tyrrhena . .. vincula: Servius suggests that Virgil was thinking of the
distinctive patrician and senatorial shoe with its four laces and moon-
emblem which, like other official insignia, the Romans had adopted from
the Etruscans. At least his shoes were in the style of his adopted land, his
sword had been brought from home (Tegeaeum 459).
459 f. Evander puts on his baldric so that it passes from his left hip
(lateri), where his sword is hanging, over his right shoulder (umeris is
poetic plural); he then puts the panther-skin round his shoulders, flinging
it back (retorquens) like a cloak over his left shoulder so that it hangs down
on that side (demissa ab laeva) at the back, leaving the sword free and
 COMMENTARY 1 37

accessible. Cf. A. 7.666 (Aventinus) ipse pedes, tegimen torquens ('flinging

round him') immane leonis.
461 limine ab alto: Virgil is not at pains to describe the royal quarters
in realistic detail, but it is baffling to see how a tectum humile (455) can have
a limen altum if this means 'lofty threshold'. Markland on Statius Silv.
1.1.46 conjectured arto, 'narrow', to get over the difficulty, but it seems
very unlikely that this is what Virgil actually wrote. Wagner, QV 40,
thinks that Virgil is nodding (and Mackail ad lac. substantially agrees):
he has fallen unthinkingly into an epic cliche, always associated with
palaces and temples (cf. A. 11.482, alto ... de limine of the templum
Palladis in Latinus' city), but here curiously inept after the stress on the
poverty and humility of Evander's settlement. It should however be noted
that although the adjective altus came to lead a quite independent exist-
ence, it was originally the participle of ala and meant 'nourished' or
'increased' in any way (Virgil the etymologist would have known this);
and physical tallness is so often a symbol, especially in heroic contexts, of
high standing in honour or prestige that altus came to carry with it
naturally the implication regius or the like. Cf. 541 solio . .. ab alto where
'high chair' carries with it the same implications as 'high table'.
There is a further problem: why does Evander go out of his house to
meet Aeneas who was apparently (366 ff.) inside it? Precision is not to be
looked for, but Virgil seems to have visualised Evander's regia as a
collection of hut-chambers surrounding a courtyard (basically the same
plan as a small Pompeian house). At 366 f. Evander leads Aeneas through
the ianua, and then presumably across the courtyard or atrium into a
guest-room; dawn rouses Evander from his own low-roofed room (455),
which he leaves (461 f.) to cross the courtyard towards Aeneas' private
quarter ( = the guest-room) (463); but Aeneas is already up and about, so
they meet in the courtyard itself which is in the centre of the hutments
(467 f.) and hold their parley there. They are still sitting there when it
thunders ominously (520 ff.); after this they all sacrifice to the household
gods (543) in a niche off the courtyard; only then (546) does Aeneas leave
through the main gateway. The same general view is taken by Mackail,
CR 32 (1918), 106, and this interpretation of mediis aedibus (467 f.) is
supported by mediis in sedibus aulae Statius Th. 2.148 in a passage clearly
inspired by this one.
licito tandem sermone (468) is taken by commentators to mean that
they at last have privacy (in the middle of the courtyard!); but what had
not permitted an earlier discussion of state affairs was the religious festival
of the day before, during which public business of any kind would
be suspended.
461 f. An extreme case of the sense of a 'theme' being both incomplete
and misleading without the 'variation'; the custodes are the canes: see the
note on 436, (1).
462 erilem, a word frequent in comedy but rare in epic, gives a colloquial
and homely touch to the scene; cf. A. 7 .490 (also in connection with a dom-
esticated animal, Silvia's stag) ille manum patiens mensaeque adsuetus erili.
 COMMENTARY

463 secreta: poetry was fond of the neuter plurals of abstract nouns
when they were metrically convenient; such 'poetic' plurals depend
ultimately on their being analogous to straightforward legitimate plurals:
secreta may have been helped by the analogy of aedes pl., 'house, rooms',
as well as by the prose secreta loca. Compare the note on 29 pectora.
463 f. Note the stress on heroic etiquette and honour (hospitis ... pro-
missi ... heros).
465 matutinus: see the note on seram 30, and cf. G. 3.538 nee gregibus
nocturnus obambulat (lupus), and Jl. 8.530 f. TTpw, S' V1TTJofo, uvv Tevxt:cn
1 ' t:m
8WPTJX 8EVTE~ I V7JV<nv \ ,I,. -
H yl\a'l'vpr1a,v t:yt:,poµ,ev
• , •t' HA p71a.
05vv
466 huic = 'the former', illi = 'the latter', as also at 358: the opposite
of the usual prose idiom.
467 f. See the note on 461 ad fin.

469-519 Evander explains that the Arcadians themselves have only small
military forces, but an Etruscan army, already mobilised, is waiting to be led
against the tyrant M ezentius, who was expelled by the citizens of Caere and is
now protected by Turnus. An oracle has declared that the leader of the army
must be a foreigner; E vander is too old to accept the commission, but Aeneas is
clearly fated to fulfil it, and shall be accompanied by Evander' sonly son, Pallas.
469 rex prior haec: clearly an insertion between two sections which
were independently composed; a tibicen, 'prop', in the full sense of the
word (Mackail, CR 29 (1915), 226 compares A. 5.653, 9.295, ro.580, 12.631).
It is probably a mere metrical memo, which would have been replaced by a
complete line in the final draft, although the unit it builds (up to the
strong caesura in the second foot) is itself used elsewhere to introduce direct
speech (cf. 499, 532).
470 f. 'Greatest of Trojan leaders, while you are safe and sound I
personally shall never allow that the realm and kingdom of Troy is
conquered ... '
471 equidem: Servius thought this was compounded of ego and
quidem, and some modern philologists agree (see LHSz 2, 174, KS 1, 805)
(others take it as formed from e (of quite uncertain origin) and quidem).
It is most usually (and in Virgil always) found with the first person singular,
like the English 'I personally .. .'.
472 f. Livy (1.7.8) says that Evander ruled auctoritate magis quam
imperio-as might naturally be the case with an exile.
ad belli auxilium: 'to give help for war'; the genitive is objective.
pro no mine tan to: 'to match so great a name as ours'; nomen has
the overtones of 'power, dominion' which it would have in nomen Ar-
cadium, cf., for example, Sallust Cat. 52.24 gentem infestissumam nomini
Romano.
473 Tosco . .. amni: see the note on 65. claudimur recalls the time
when the Tiber was the early boundary between Rome and Etruria, after
the territory on the right bank had been surrendered to Porsena; cf.
Juvenal 8.265 imperiiftnes Tiberinum virgo natavit (Cloelia, A. 8.651).
 COMMENTARY 1 39

474 Rutulus: for the collective singular of national names, see the
note on 705 f.
475 tibi ego: note the marked tendency in verse as well as prose for
personal pronouns to be placed next to each other at the beginning of the
clause; cf. ego te 496.
475 ingentis populos: listed in the Catalogue of the Trojan Allies
A. 10.166 ff. For the extent of Etruscan land-power see the note on 65,
and Livy 5.33.7-9 (where Ogilivie gives an excellent summary of its
history). Their maritime power reached its peak in the late sixth and early
fifth centuries B.C.: Virgil alludes to it in condensae . .. puppes (497).
475 f. opulentaque regnis ... castra: 'camps rich in royal forces',
a slightly unusual way of saying castra opulentorum regnorum; cf. Val. Fl.
5.607 eras acies atque illa ducum eras regna videbis.
476 quam ... salutem: the construction is discussed in the note on
487.
478 ff. For the iKcppacns -r611ov (picked up by 481 hanc) see the note on
597.
478 ff. urbis Agyllinae: Agylla, the ancient Caere, modern Cerveteri,
about 30 miles north-west of Rome (haud procul hinc 478) and three and a
half miles from the coast, on a spur of volcanic tufa-rock, which although it
weathers quickly (saxo . .. vetusto) was extensively used for building.
Through neighbouring ports, particularly Pyrgi (the modern Santa Severa),
it received a great quantity of foreign, especially Greek, imports, and the
discoveries in the extensive necropolis show that at its peak in the seventh
and sixth centuries B.C. it must have been an extraordinarily rich and
populous centre (multos florentem annos 481), perhaps one of the most
spendid of the then-known world. See Pallottino n2 f., and for plates of
Caere and maps of Etruria and the Etruscan empire, Atlas of the Classical
World (A. A. M. van der Heyden and H. H. Scullard, 1959), 94 f. See also
J. Heurgon, ]RS 56 (1966), 3 f. for a reassessment of Caere in the light of
the bilingual inscriptions of Pyrgi, confirming its importance.
(Ogilvie on Livy 1.2.3 distinguishes between Caere and its port Agylla,
but it is difficult to conceive of a harbour on a hilltop, which is where
Virgil places the urbs Agyllina (iugis 480), and his younger contemporary
Strabo (5.220) explicitly identifies Agylla and Caere; cf. also Rutilius
Namatianus De Reditu Suo 1.225 f. iam Caeretanos demonstrat navita
fines:/ aevo deposuit nomen Agylla vetus.)
479 f. Lydia ... gens: Roman poets (e.g. Catullus: see Ellis and
Fordyce on 31.13) followed the majority of ancient historians in the view
that the Etruscans originated from Lydia. The only dissenter was
Dionysius of Halicarnassus (1.30), who claimed that they were indigenous.
The debate continues between modern supporters of the two theories.
The majority view of the ancients derived most plausibility from re-
semblances between Etruscan culture, especially in its 'orientalising'
phase, and the culture of Asia Minor: but these may be due to later trade
and infiltration, rather than to the original immigrants, if such there were.
Pallottino has tried to by-pass profitless speculation about origins by
II EAE
140 COMMENTARY

stressing that the characteristic elements of Etruscan civilisation developed

on Italian soil. Scullard 34 ff., after a judicious examination of the alter-
native views, inclines to a tentative form of the theory of eastern origin.
quondam is to be taken with insedit.
480 iugis . .. Etruscis: most Etruscan cities near the sea were built on a
natural elevation dominating the coastal strip: Caere is typical (see the
note on 478 ff.). Such a siting offset the menace of pirate-raids.
481 deinde: for this delayed adverb, see the note on 22; this is a
special case, for hanc . . .florentem = postquam floruit, and deinde is
delayed to second position only in its own clause.
484 effera: for the descriptive stopped dactyl (impetuous violence
here), see the note on 1-6.
484 capiti: see the note on 144 f.
reservent: sc. eadem or talia.
485 ff. The practice described here was attributed to Etruscan praedones
by Cicero in a passage of the lost Hortensius, cited by Augustine contra Pel.
4.15: qui quondam in praedonum Etruscorum manus incidissent, crudelitate
excogitata necabantur, quorum corpora viva cum mortuis, adversa adversis
accommodata, quam aptissime (v.l. artissime) colligabantur. In PVS 4
(1964-5), 36 ff., I suggested that this form of human sacrifice was originally
intended to placate (and nourish) the spirits of the Etruscan dead cla-
mouring for revenge, and a potential menace to the community itself if
unappeased. Mezentius may have been connected with the custom as
overlord and king-god of the Etruscan Confederation.
The Etruscan custom of sacrificing captives at the burial of their own
warriors (Servius on A. ro.519), the origin of the Roman gladiatorial show,
and Livy's record (7.15.9 f.) that in 358 B.C. the Tarquinienses ritually
slaughtered (immolarunt) three hundred and seven Roman captives,
suggest that Mezentius' torture was simply a more grisly form of the same
thing. The human sacrifice made by Aeneas (A. ro.517 ff.) has not only
literary precedent (Achilles in Il. 23.175 f.), but also reflects intermittent
practice during Virgil's own lifetime (cf. Weinstock 398 f.).
487 tormenti genus: in apposition with the whole idea of the pre-
ceding sentence, cf. A. 6.222 f., 9.53, ro.3n, n.62 f. In all these cases the
noun is neuter, and may be either nominative or accusative, because both
of these cases could express exclamation, out of which the appositional use
arose-the difference is purely one of emphasis. (There is no need to
assume the ellipse of a verb like est or dico.) Highly-ordered classical prose
avoided appositional nouns-they looked too much like untidy after-
thoughts-and preferred to subordinate the idea in a relative clause
(cf. A. 8.475 ff. iungere castra para, quam /ors inopina salutem / ostentat).
See further LHSz 2, 429 f.
487 sanie taboque fluentis: Servius says sanies mortui est. 'tabo'
viventis scilicet sanguine: an entirely sophistical distinction, inspired solely
by 485 mortua . .. vivis; both words refer to the gory putrefying flesh of
those already dead, with which the living bodies dripped (fluentis). They
may well be a hardened formula of the high style, cf. Ennius Thyestes
 COMMENTARY

(Sc. 362 f. V) ipse summis saxis fixus asperis evisceratus / latere pendens
saxa spargens tabo sanie et sanguine atro. Compare also Lucan 6.547 ff.
(the witch Erich tho preying on the crucified dead) ... nigramque per artus /
stillantis tabi saniem (a near-tautology) virusque coactum / sustulit.
489 infanda furentem: 'raving inhumanly'; for the accusative neuter
plural adjective used adverbially see the note on 248 insueta.
489 f. Note the effect of two consecutive heterodyned spondaic lines:
like the slow menacing marking-time of resolute demonstrators.
492 ff. That Mezentius fled to Turnus (Rutulorum ... in agros), thus
leaving the Etruscans without leader or king, is the most substantial
difference in Aeneid 8 between Virgil's account of the story and the
traditional accounts (see the Introduction, p. xix). Cato (cited by Servius
on A. r.267) and Livy (1.2.3) say that Turnus fled to Mezentius, and
Dionysius of Halicarnassus (r.64) differs only in saying that Turnus
(Tvpp71v6s = the Etruscan) was killed before the remaining Rutulians allied
themselves with Mezentius.
But in his overall account of Latian-Etruscan relations in the Aeneid,
Virgil is completely at variance with what must have been one of the
oldest and most respected versions, that of Cato in the first book of his
Origines (the main outlines of which can be reconstructed from scattered
notes of Servius, see Peter HRF 44 ff.); Cato's sequence seems to have told
how Latinus originally granted a piece of land to the Trojans for settle-
ment, but further encroachments by the Trojans led to war against the
Latins and their allies, the Rutulians led by Turnus. In the first engage-
ment in which Aeneas was involved Latinus was killed. Turnus then fled
to Mezentius and renewed hostilities with his support; in the second
engagement Turn us fell and Aeneas 'disappeared' (non comparuit).
Ascaniusinherited the war and eventually killed Mezentiusinsinglecombat.
Virgil most probably originated the version he gives, and if he did not,
must have been in a distinct minority in preferring it to the other. His
object was to make the story of 'Aeneas in Latium' culminate in Aeneas'
marriage with Lavinia, and the founding of a city through the synoecism
of immigrants and natives. This scheme made an early and complete
alliance with the Latins impossible, and the telescoping of the chronology
ranged them and Turnus and Mezentius simultaneously against Aeneas,
who was in any case lacking in troops. See further Heinze, VeT part r,
ch. 5, and P. T. Eden, PVS 4 (1964-5), 33 f.
The traditional story of Turnus perhaps reflects part of the history of
Ardea, the capital of the Rutuli, when Etruscan domination, established
there in the sixth century B.C., was becoming effectively weakened.
(Cato's account is summarised by Servius on A. 9.745 to defend Virgil
against the majority of critics, who thought that Turnus should have met
his death at that point.)
492-7 The succession of homodyned fourth feet in these lines, a rhythm
which tends to suggest a paragraph-ending (see Appendix, p. 197), com-
bined with a high proportion of end-stopped lines, makes Evander sound
as if he is simply giving the chief points of a resume.
11-2
 COMMENTARY

493 confu~ere et ... defendier: in lively impressionistic narrative a

past finite verb (with its meticulous details of person and tense) is often
replaced by a historic infinitive; usually there is more than one such
infinitive, and usually they do not open the narrative passage but follow
after the time-reference has been established as past by appropriate finite
verbs (here the historic presents in 490 f.). Normally the historic infinitive
conveys some shade of meaning covered by the imperfect indicative (a
continuous state, or the beginning or repetition of actions), but this is not
always the case: in some passages, as here, it is difficult to see its equi-
valence to anything but a simple past (aorist perfect).
The subject of the historic infinitive is nominative (here ille, 492), and
the whole idiom is a piece of verbal shorthand: the nominative is used in
its basic function to draw attention to the temporary centre of interest,
and the present infinitive, originally a substantive, expresses the verbal
notion pure and simple without any other consideration. This is also
adequate explanation of the origin of the construction; one need not follow
Quintilian (9.3.58) in believing that it arose from ellipse of coepi.
See further Woodcock 15, Palmer, LL 317 f., LHSz 2, 367 f.
493 defendier: forms of the present infinitive passive in -ier instead of
-i occasionally found favour with Virgil, both because of their archaic
flavour and their metrical convenience: this latter is usually exploited by
placing the termination in the dactylic fifth foot of the hexameter (cf.
G. 1.454 immiscerier igni; A. 4.493 accingier artis; A. 7.70, 9.231, n.242):
its position here in the fourth foot is exceptional. See further Austin on
A. 4.493.
494-6 'And so the whole of Etruria has risen in righteous fury; they
demand their king back for punishment with active war. I will unite you
with these thousands, Aeneas, as their leader.'
Note the abrupt change of subject in 494-5.
497 condensae: first, and most frequently, in Lucretius (e.g. 1.575,
606).
497 litore: the ships are drawn up on the beach on the coastal side of
the open plateau near Caere (campo 504 = latis ... in arvis 605) where the
Etruscan host is camped.
498 si~naque ferre iubent: 'they command the standards to be
advanced'. In this and similar cases English idiom prefers a passive
infinitive; this is possible in Latin too, but more often one finds an active
infinitive after iubeo and veto when the personal object of the verb ( = the
subject of the infinitive) can be easily supplied from the context or is quite
general and unspecific; cf. A. 2.3 infandum, regina, iubes (sc. me) renovare
dolorem; Livy 29.7.6 receptui canere cum iussisset (sc. tubicines); Cicero
Phil. 13.6 F ortitudo dimicare iubet, 'Bravery commands one to fight'.
Compare 646 f. See further KS 1, 717.
Metrical convenience was of course a decisive factor in choosing the
infinitive form, contrast A. 1.647 f. munera ... / ferre iubet with A. 8.175 f.
dapes iubet et sublata reponi / pocula, and cf. Val. Fl. 5.261 f. reddi iubet
exitiale sacerdos / vellus et Haemoniis infaustum mittere terris.
 COMMENTARY r43

498 haruspex: literally 'an inspector of entrails', haru- being cognate

with Greek xopS~, Latin hira, 'gut'. Divination by this means was the
characteristic part of the disciplina Etrusca; augury, on the other hand,
was a widespread Italic custom, practised in, for example, Umbria as well
as Latium and Etruria. The haruspices never acquired the same official
status as the augurs in Republican Rome, but their advice was increasingly
sought after in the desperate crises of the Second Punic War, and eventu-
ally the Roman State requisitioned a constant supply from Etruscan
cities. See further Pease's note on Cicero Div. r.3.
499 fata canens: see the notes on 49 and 340.
Maeonia = Lydia (479) = Etruria.
500 'flower and valour of a historic people '.flos here implies both youth
(see the note on/fore 160) and choiceness (cf. Plautus Casina 18 ea tempes-
tate flos poetarum fuit, and English 'the pick of the bunch').
For the alliteration veterum virtusque virum, significant here, see the
note on 2.
502 nulli: Virgil did not like a self-contained spondaic word at the
beginning of the verse; even so, he did not have Ovid's clear aversion to it:
in A. 8 there are 25 instances, of which four or five are followed by a
sense-pause, and this is an average number (see Austin on A. 2.80 and
4.185; R. D. Williams on A. 3.57 and 5.80). The reasons for its general
avoidance are discussed in the Appendix, p. 196. Most of the instances
which Virgil admits can be explained under one of four heads: (1) the
slow, ponderous rhythm reinforces the meaning (e.g. stabat 653); (2) the
meaning of the word is naturally emphatic and can stand in this emphatic
position appropriately (e.g. toto 497 and nulli here); (3) rhetorical contrast
would have given the word a leading position in prose too (e.g. vobis 138);
(4) adverbs and connectives, closely associated in sense with what follows,
do not really interrupt the flow of the verse (e.g. exim 306, quorum 547).
olli at 94, 594 is an echo of Ennian rhythm (Ann. 555, 624 V). See further
Norden 435 f.; Winbolt 16 ff., 107 ff.
502 tantam subiungere gentem: 'to harness such a great nation'
(the Etruscans).
503 externos optate duces: 'choose leaders from abroad'; for the
'oracular plural' Conington compares A. 7 .98 externi venient generi.
503 tum Etrusca: note the elision; the monosyllables most often
elided in the Aeneid are me, te, se and iam (cf. 557), but such elision is by
no means frequent: it is harsh and colloquial (and so occurs often in
Horace's Satires but never in his Odes) and seems to be admitted by Virgil
only often enough to avoid an over-studied smoothness. See Norden 456 ff.
505 ff. The legend, followed by Virgil, which takes Aeneas to Caere to
meet a friendly army under Tarcho (cf. 603 ff.), is an obscure one; it is
alluded to in riddling words at Lycophron Alexandra 1232 ff. The con-
nection of Tarcho with Caere is however established by the discovery there
of a tomb with inscriptions, which cover two centuries, of the Etruscan
family Tarcna (Latin Tarquinius), of which Tarcho is clearly a repre-
sentative (see Ogilvie on Livy r.60.2).
 r44 COMMENTARY

505 oratores: 'ambassadors'; orator basically = is qui orat, 'a spokes-

man' charged with an oral message; so used by Ennius (Ann. 207 V)
orator sine pace redit regique refert rem (so at Ann. 170 V, spernitur orator
bonus, horridus miles amatur there is not simply a contrast between the
arts of peace and war; the line means that well-intentioned diplomacy has
had to give way to violent hostilities). Cf. also A. 7.153, 11.100, 331, and
Livy 1.15.5 Veientes pacem petitum oratores Romam mittunt.
505 f. coronam / cum sceptro . .. insignia: a golden crown and a
sceptre surmounted by an eagle were two of the emblems of Etruscan
kingship; the others were an ivory throne, a purple tunic decorated with
gold, an embroidered purple toga, and the / asces with axes in them.
Dionysius of Halicarnassus (3.61 f.) was no doubt presenting the general
Roman view in saying that these items first appeared in Rome under
Tarquin the Elder, and were adopted by the succeeding kings; they then
passed to the annual consuls, except for the crown and embroidered purple
toga which were reserved for triumphatores. Virgil is clearly following this
tradition (the actual historical facts may have been different, and are
debatable; see Ogilvie on Livy 2.16.r and L. B. Warren, ]RS 60 (1970),
49 ff.).
506 f. mandat . .. insignia . .. , / succedam: mando means 'I commit
(something or somebody to someone's charge)'; what is committed can be
either an object (a noun, here insignia) or a wish or command (a noun-
clause, here succedam): both together, as here, are unusual although similar
double constructions occur at G. 1.25 f. urbis(ne) invisere, Caesar, / terra-
rumque velis curam; A. n.180 f. (if these lines are not spurious) non vitae
gaudia quaero, / nee fas, sed nato manis perferre sub imos; A. 3.41 f. iam
parce sepulto, / parce pias scelerare manus. The -simple parataxis of
mando, impero with a jussive subjunctive (of will or volition) was the rule
in the earliest stages of the language; the classical language generally
stresses the subordination by introducing the subjunctive with ut, but the
old-fashioned simplicity is often kept by Caesar (e.g. B.G. 3.n.2 huic
mandat, Remos reliquosque Belgas adeat) and Virgil.
misit mandatque: note the difference in tense to mark the precise time-
relationship between the two events.
508 Cf. A. 5.395 f. sed enim gelidus tardante senecta / sanguis hebet,
frigentque effetae in corpore vires. saeclis here may simply = annis as
Heyne suggests, but it is much more likely to mean 'generations' or
'lifetimes'. As elsewhere Homer's Nestor is lurking under Virgil's Evander,
and he was reputed to have outlived two generations of mortal men,
o, o,
Il. 1.250 ff. T(t' 8' 71871 8vo µ.Jv ym,a, f1,Ep61rwv av0pclJ1TWV I e¢,0£a(J'. 1rp6u0Ev
aµ.a Tpa</,Ev ~s· E')'€VOVTO / EV llv>.cp ~ya0lr,, fl,ETO. SJ TPLTUTOL<TLV ava<T<TEV (hence
Laevius' grotesque coinage trisaeclisenex, Aulus Gellius N.A. 19.7.13).
Virgil's condensed phrase no doubt alludes to this; translate: 'But my old
age ... worn out with the passing of generations.'
510 f. 'I would encourage my son (to accept the leadership), but for the
fact that he is of mixed descent through his Sabine mother and derives
part of his native origin from here (Italy).'
 COMMENTARY 1 45

With Italian blood in his veins Pallas was ineligible to be the foreign
leader of the Etruscans (503). But even though born in Italy and of an
Italian mother the more important part of his 'native origin' was the
civil rights and status he inherited from his father and his father's city.
510 mixtus matre Sabella: a compressed phrase; Virgil probably
started thinking of the perfectly normal natus matre Sabella (ablative of
origin), and then avoided this prosy phrase and adapted it to his particular
context by substituting mixtus (which slightly dislocates the ablative into
one of respect). Quite different (pace Conington) is A. 7.661 mixta deo
mulier, where deo is best regarded as dative (cf. Il. 16.176 yvVTJ 0E{jJ EVV7J-
8Etua and Od. 1.73 IlouEiStiwvi 1.uyEtua). Similarly, in the next line, hinc
partem patriae traheret is a bolder variation of the prose hinc partim
originem duceret or the like.
512 fata . .. numina: linked here and at 574 f.; they are obviously
near-synonyms bearing their basic meanings 'the spoken decrees of fate'
and 'the nodded decrees of fate'.
513 Teucrum atque ltalum: for the gen. pls. see the note on 93.
514 spes et solacia: because, as Servius points out, Evander was a
widower and Pallas his only son.
518 f. 'I shall give him two hundred Arcadian horsemen, the picked
strength of our youth, and Pallas will give you the same number as his
own gift.' Evander turns the fact-that Pallas will head a force of 400
cavalry to support Aeneas-into a compliment to his son.
munere is preferable to the variant nomine in the Medicean MS, and read
by Servius, as being slightly less usual and prosaic: in point of sense there
is nothing to choose. The variant may derive from A. 8.121.
robora pubis / lecta: a clear reminiscence of Catullus 64.4 lecti
iuvenes, Argivae robora pubis; see Austin on A. 2.18 for similar formulae
which stress heroic physique.

520-540 Aeneas and Achates ponder the implications of what Evander has
said when, in a clear sky, there is thunder and lightning and the sight of
clashing armour. Aeneas recognises the omen as the fulfilment of Venus'
promise to bring him arms made by Vulcan, and foresees destruction for
Turnus and his forces.
520-3 'He had hardly finished speaking, and Aeneas, Anchises' son, and
the faithful Achates would have gone on keeping their faces fixed on the
ground and trying to clear up many cruel doubts in their gloomy hearts
if the goddess of Cythera had not given a sign in the clear sky.'
520 ff. vix ea fatus erat, defixique ora tenebant . .. putabant, ni
signum . .. Cytherea dedisset: the simple form of the statement at, for
example, A. 2.692 f. vix ea fatus erat senior, subitoque fragore / intonuit
laevum is here elaborated: (1) -que, where one might have expected cum
inversum, introduces not a statement of the next important event, but a
description of how the listeners had been and still were reacting to
Evander's speech; (2) putabant ni dedisset: the 'mixed conditional' form
 COMMENTARY

is due to an ellipse: they were thinking and would have continued to do so

if she had not given a sign; cf. A. 6.358 ff. iam tuta tenebam, / ni gens
crudelis. . . / ferro invasisset, and for examples in Cicero and Livy, and
especially Tacitus, see KS 2, 404.
520 ora is to be taken am> Koivov as the direct object of both defixi (used
in a middle sense, see the note on 29), and tenebant. The same explanation
holds good for A. 2.1 conticuere omnes intentique ora tenebant which is not
an example of• theme and variation' (pace Henry, and Austin who follows
him), but means •they all fell silent and were keeping their faces intent
upon him'; the second phrase, like the phrase here, may imply silence but
does not explicitly state it any more than A. 6.469 illa solo fixos oculos
aversa tenebat. ora teneba(n)t may be an Ennian clausula, cf. Ann. 87 V
Sic expectabat populus atque ora tenebat (but Vahlen keeps the reading of
the codex optimus, ore timebat).
521 f. Aeneas' reaction to Evander's speech is gloomy silence, chiefly
because an offer of 400 cavalry is no immediate solution to his pressing
need for troops, and his gloom is lifted only by the prospect of invincible
arms. But the multa dura also included a premonition of the fate of the
young Pallas: that, as Warde Fowler (ASR 86 ff.) points out, is conveyed
by the patronymic A nchisiades; Aeneas the son thinks of Pallas the son,
spes et solacia of his father. The same connection of ideas is made explicit
at A. 10.821 ff. (but earlier at A. 10.250 the patronymic has no such
implication).
Aeneas takes precedence in the word-order (cf. 586), 'Achates knows his
place' (Wilkinson 177)-as a courtly subordinate; fidus has a more cere-
monious tone than fidelis.
522 multaque dura suo tristi cum corde putabant: a line with rich
overtones of archaic gravity: cum corde instead of a simple locative ablative
(cf. Cato, cited by Gellius 16.1.4, cogitate cum animis vestris: si quid . .. ) ;
the strictly superfluous possessive adjective suo with another adjective
tristi; the sense of putare, • to clear up', all suggest archaic (in fact Ennian)
phraseology. See Norden on A. 6.185 suo tristi cum corde volutat, and 332
multa putans: and cf. Ennius Ann. 482 V tristi cum corde, and Ann. 51 V
aegro cum (tum?) corde meo; and Lucretius 1.413 meo ... diti de pectore.
putare derives from the basic agricultural layer of the Latin vocabulary;
originally 'to prune, clean away impurities', it came to mean 'to clear
away anything superfluous or hindering': so here, Aeneas and Achates
'were trying to clear up' in their minds the difficulties they foresaw, until
the divine omen put an end to their thoughts by showing the solution.
523 ff. Venus' sign, thunder and lightning, and the appearance of the arms.
(1) At Il. 19.13, when Thetis deposits the arms in front of Achilles, they
ring aloud, but their arrival is not heralded by any portents. Virgil trans-
forms the purely physical manifestation of the arms in the sky into a
supernatural omen: the flash of lightning and the thunder in a clear sky,
twice repeated and confirmed, is an infallible sign.
(2) In terms of Roman augury, it is here a sign which gives advice,
cf. Seneca N.Q. 2.39 (fulmen) consiliarium ante rem fit sed post cogitationem,
 COMMENTARY 1 47

cum aliquid in animo versantibus aut suadetur Julminis ictu aut dissuadetur.
Aeneas, dismayed by the prospects of war, and doubting whether the
small reinforcements he has just been offered really mark him out as the
man of destiny, is encouraged by the sign to persevere.
(3) Thunder and lightning and the song and flight of birds as augural
signs belong to a basic stratum of Roman religion which antedates
anthropomorphic gods; later, presumably any one of these relevant to
the situation at the time could be connected with the sign. This meets
Gladstone's objection (Studies on Homer, vol. 3, 523) that for Venus to
use lightning is 'utterly irreconcilable' with her Homeric character: Virgil
is not primarily thinking of her Olympian attributes. By implication,
however, he adds her to the list of those who borrowed the aegis of Zeus
(see the note on 435), and this is later explicitly done by Valerius Flaccus,
probably on the basis of this passage: tonitru pater auget honoro, 2.198 f.
(and cf. 2.n5 f.).
(4) The comparison of thunder with the sound of an Etruscan trumpet
alludes to the Etruscan origin of Roman augury (the supplier of Seneca's
information cited above has the good Etruscan (Volaterran) name of
Caecina). The significance of the trumpet-omen is made clear by Plutarch
Sulla 7.3: various prodigies announced the contest between Marius and
Sulla, 'but most important of all, out of a cloudless and clear sky there
rang out the voice of a trumpet, prolonging a shrill and mournful note, so
that all were amazed and terrified at its loudness. The Etruscan wise men
declared that the prodigy foretokened a change of conditions and the
advent of a new age.' The same omen is commonly associated with Caesar's
crossing of the Rubicon (Lucan 1.530 ff., 569, 578; Petronius Bellum
Civile 134 f.) and his death (G. 1.474 f.; Tibullus 2.5.73 f.; Ovid Met.
15.783 ff.).
(5) The trumpet had widespread fame as an Etruscan invention (cf.
Soph. Ajax 17 Xa.AKocrr6µov KwSwvos ws TvpCTT)v,Kf'Js); and the word tuba, like
lituus, is itself probably of Etruscan origin (see Ernout, RPh 24 (1950),
5 ff.). So the trumpet-omen is particularly appropriate to foreshadow a
military alliance between Aeneas and the Etruscans.
523 caelo . .. aperto, 'in a cloudless sky', = caeli in regione serena 528;
cf. G. 1.393 f. nee minus ex imbri soles et aperta serena / prospicere et certis
poteris cognoscere signis.
Cytherea: the legend that foam-born Aphrodite first stepped on land on
the island of Cythera, south of the easternmost prong of the Peloponnese,
reflects the fact that her cult spread from there to the mainland of Greece.
524 'foralightning-flash, unexpectedly hurled from thehighheavens ... ';
vibratus, 'hurled' (by Venus), cf. Ovid Met. 2.308 unde movet tonitrus (sc.
I uppiter) vibrataque fulmina iactat; literally 'shaken as it is launched', cf.
Ovid Met. 8.374 f. vibrata per auras/ hastarum tremulo quatiebant spicula
motu. vibratus (or vibrans intr.) refers to light, 'flashing', as well as
motion, cf. 391 f. coruscus, micans. See the note on 427 for Julgor.
525 et ruere omnia visa repente: 'and the whole sky suddenly
seemed to come crashing down' vividly describes the impact on the
 COMMENTARY

observers; contrast G. r.324 ruit arduus aether, where there is a literal

downpour.
525 The rhythm of the line is unusual: it suggests speed by means of
the dactyls, and disturbance by means of the elision ruer(e) omnia and the
slight pause after venit (see the note on 390 intravit calor).
526 Tyrrhenus: the epithet is transferred, from tubae to clangor, not
primarily (pace Norden 407) to avoid the homoeoteleuton of Tyrrhendeque
tubae, but in the interests of word-symmetry (see Appendix, p. 200).
526 With mugire one must understand visus from visa in the preceding
line; another example (see 35 and note) of video(r) used of a sound im-
pression, cf. A. 4.490 f. mugire videbis / sub pedibus terram (and Austin's
note there), and 529 below, rutilare vident et pulsa tonare. The idiom is not
confined to Virgil, cf. Hor. Sat. 2.8.77 f. videres / stridere ... susurros;
Propertius 2.16.49 vidistis toto sonitus percurrere caelo.
The line owes something to EnniusAnn. 531 V clamor ... per aethera vagit.
527 'and again and again there roars a mighty crash'; increpare comes
to be used of any loud noise (the variant intonat, more literal and prosaic,
probably derives from A. 9.709); Silius 16.94 f. uses it of a trumpet giving
the battle-signal, dixerat, et raucus stridenti murmure clangor / increpuit;
Lucan 9.288, of clashing brass cymbals, Phrygii sonus increpat aeris.
528 inter nubem is curious in view of the insistence on cloudlessness;
Henry is no doubt right in saying that it is there for the benefit of the
materially-minded, to provide visible support for the arms (like the nimbi
of 608 which support Venus).
529 sudum, in Virgil only here (as noun) and G. 4.77 (as adjective),
= se-udum, 'without moisture, cloudless'; the prefix se- = sine, as in
securus, secerno, etc.
530 animis: the ablative singular and plural of animus are often
found in close connection with verbs and adjectives denoting mood and
emotion (cf. Cic. Brut. 7 angor animo; T.D. r.96 pendemus animis); the
ablative can be regarded as either local or as an ablative of respect; the
latter is perhaps preferable, because in the singular a genitive of respect
animi is often found as an exact equivalent (but not in classical prose),
e.g. Plautus Trin. 454 sanus mentis aut animi. See LHSz 2, 75.
531 'recognised in the sound the promises of his goddess mother';
sonitu divae would have been quite possible, but Virgil preferred a
hendiadys, see the note on 436.
promissa: no previous promise of either a favourable omen or the
delivery of divine armour has been made by Venus in the Aeneid, but
Thetis had made such a promise to Achilles in the Iliad (18.134 ff.). Virgil
can afford to be allusive not only about general Homeric phenomena
(concealment by cloud, A. 12.52 f.), but also about a well-known detail of
the narrative (as here).
532 Hirtzel and Ladewig removed the traditional punctuation after
profecto, and took it as a participle (dative singular masculine). The line
can then be interpreted to mean 'Do not ask what event these portents
forebode for him who has set out: it is I who am summoned'. But (1) to
 COMMENTARY 1 49

make an emphatic contrast, the pronoun ego ought to be balanced by

another pronoun or a noun, making it quite clear that Pallas is referred to:
the contrast can hardly be extracted from the participle alone; (2) Pallas
had not yet 'set out' in any literal sense, and to understand profecto to
mean vaguely 'who has embarked on this enterprise', as Mackail suggests,
is unsatisfactory.
Since Servius profecto has traditionally and correctly been taken as the
particle of personal affirmation (see Nagelsbach, Lateinische Stilistik 432),
Tm sure', which then appropriately links up with ego in the following
line. The word certainly had a lively colloquial flavour which did not
appeal to serious poetry, and it does not occur anywhere else in Virgil.
But this is no serious objection: it occurs once only in Horace too (A.P.
315) (where it cannot possibly be construed as a participle), and is paralleled
by the unique occurrence of the equally colloquial sane at A. 10-48 (see
Axelson 94).
For the emphatically repeated ne cf. A. 11.278 and 12.72 f. Aeneas is at
pains to stop Evander guessing what he half foresees himself-the doom
of Pallas.
532 ne quaere: for ne negativing an imperative, a non-classical usage,
see the note on 39 ne absiste.
532 ff. The passage from 532 to 536, ending with the 'half-line', looks
from the number of obscurities in it (see the individual notes) as if it
probably did not receive the final touches. But even as it stands, the latent
irony of the situation should not be missed. At 520 ff. the company was
pondering gloomily on the future; uncanny thunder and fearful lightning
rend the sky; Evander is terrified by the thought that this is a prediction
of disaster for his son (whose name was the last spoken word, 519). But the
sign is, on the contrary, the most cheering that Aeneas could have received.
Hence Aeneas' emphatic reassurance to Evander that the portents concern
himself (and nobody else}, and mean good not evil.
533 Even in antiquity there was doubt as to whether this line should
have a heavy pause after poscor or after Olympo. Hirtzel opted for the
former; 533 f. then mean 'it is I who am summoned. My goddess mother
prophesied that she would send this sign from heaven', or perhaps
' ... in the sky' (cf. Catullus 62.1 with 20). But a heavy sense-pause at this
position in the line without rhetorical point (see the note on 443) is a
rarity, and throws great stress on the word(s) which simultaneously ends
the line and begins a sentence: Olympo can hardly bear this (we had no
reason to suppose that the sign would come from hell).
A heavy pause at the end of the line produces a more normal balance
between sense and rhythm. ego poscor Olympo means 'it is I who am
summoned by heaven', i.e. by the divine power responsible for the sign.
There is almost a double metonymy, Olympo for caelo for numine divino or
the like: cf. 385 moenia for urbes for urbani.
The phrase as a whole is comparable in syntax and sense to A. 8.12
f atis reg em se dicere posci: both repeat the important theme of Aeneas the
destined leader (see Introduction, p. xxn).
 COMMENTARY

534 cecinit missuram . .. laturam (536): the full form of the standard
expression would be cecinit se missuram esse; the reflexive subject pronoun
in oratio obliqua is however commonly omitted in Roman comedy when it
bears no particular stress, and in Plautus esse is regularly omitted from the
'future infinitive'. The same double omission is not entirely uncommon in
prose, but by Virgil's time it must have had an old-fashioned or at least
'traditional' ring about it (cf. A. 12.654 f., 762).
See further Woodcock 82 f., and for the omission of other than reflexive
subject pronouns in oratio obliqua, and with other tenses of the infinitive,
Austin on A. 4.383.
535 in~rueret: 'if war should make its onslaught' : the etymology is
uncertain, perhaps in(g)ruo. Before Virgil the word occurs only at Plautus
Amph. 236 hastes crebri cadunt; nostri contra ingruunt, in a parody of a
military dispatch. Livy has a similar phrase at 6.6.6 ubi quid bellici terroris
ingruat; and the sinister overtones of the word naturally recommended it
to Tacitus.
536 auxilio: predicative dative, 'as a help'. This use of the dative,
which is common enough with a restricted number of nouns (usually
abstract) (see Woodcock 49), developed when the dative of interest
denoting the person concerned in the action (often as recipient) was
extended to things, which then came to denote the end or purpose of the
action.
538 ff. Line 539 is an exact repetition of A. I.IOI (with valves for
volvit) : what was earlier stated of the Simois is here predicted of the
Tiber, and the link between the two (which shows Virgil's careful methods
of composition) is A. 6.86 ff., where the Sibyl prophesies that Aeneas will
meet his Trojan past all over again in Latium: bella, horrida bella, / et
Thybrim multo spumantem sanguine cerno. / non Simois tibi nee Xanthus nee
Dorica castra / defuerint (and see Norden ad loc.). This is a very special
case of the repetition of a line (see Sparrow 103). The prediction made here
was evidently fulfilled: at A. 12.35 f. Latinus says recalent nostro Tiberina
fluenta / sanguine adhuc campique ingentes ossibus albent.
540 A strong pause at the main caesura in the second foot is often
associated in Virgil with a forceful speech-ending following it, cf. A. 2.49
quidquid id est, timeo Danaos et dona ferentis, and see Winbolt 21 f.

541-553 Aeneas and the company sacrifice to Evander's household gods;

he then selects his most courageous troops to follow him and sends the rest to
Ascanius with the news. The Trojans are provided with horses, with a
specially magnificent one for Aeneas.
541 dicta dedit: = dixit, a periphrasis common in comedy (e.g. Plautus
Cist. 595 perfectum ego hoc dabo negotium), Lucretius, and Virgil (A. r.63
laxas dare = laxare; A. 3.69 f., 9.323, 12.437). This idiom is connected on
the one hand with the vulgar Latin (and subsequently Romance) compound
perfect, id habeo dictum, 'je l'ai dit ', and on the other with (red)do, 'to
make' something different from what it was before (cf. frictum ego illum
 COMMENTARY

reddam, Plautus Bacch. 767). (dare with a noun is to be explained dif-

ferently: see the note on 405, and Page on dare ruinam, A. 2.310.)
solio ... ab alto: see the note on 461, and compare the torus altus of
A. 2.2 with which Aeneas was honoured by another royal host.
542 ff. 'and first of all he stirs up the altar, slumbering with Hercules'
fires, and joyfully approaches the guardian spirit and humble gods of the
household, worshipped the day before.'
542 It is possible but unlikely that H ercuZeis strictly qualifies ignibus,
and refers to a torch, still alight, brought from the previous evening's
sacrifice to light the way for Aeneas and Evander through the growing
dusk; if this were the meaning, ignibus would be instrumental ablative
going closely with excitat. It is more likely that we have here a bold varia-
tion, dictated by word-symmetry, of what would be expressed in normal
prose as et primum ignes sopitos aris Herculeis excitat: cf. sopitos suscitat
ignes at A. 5.743 and 8-410 (hence the variant suscitat here), and Ovid
Met. 8.641 f. et ignes / suscitat hesternos. There is then a double hypallage,
and the result is another example of nova cacozeZia.
But if this is so, where is the altar of Hercules (aras poetic pl.)? It cannot
be the Ara Maxima of the day before, because Aeneas is now somewhere in
Evander's home on the Palatine. The best explanation, due to Heyne, is
that Hercules was one of Evander's household gods, the genius or Zar
domesticus. Zar and penates were originally vague protective powers
without personal characteristics; these they later acquired from specific
demi-gods with whom they came to be identified. The Zar famiZiaris of
Plautus' Aulularia corresponds to a ifpws- in the Greek original (see Leo,
Hermes 43, 127): Hercules' identification with the Zar is proved by his
frequent appearance with the penates in Pompeian wall-paintings, and he
may have been worshipped as domesticus (if the restoration of CIL 14.3542
is correct). He was of course entirely suitable to be identified with the Lar
of Evander's home, where he had once stayed (362 f.).
543 hesternum larem can only mean the Zar to whom sacrifice had
been made the day before, as could be assumed, though not mentioned in
the narrative. If so, this phrase too is a bold one, an extreme case of
adjective for adverb (see the note on 30), hesternum being equivalent not
merely to heri but to heri adoratum. The variant externum (MR) was
disliked by Servius. It can = tlvwv, hospit(al)em, 'of his host', where the
context makes this meaning clear, as at Cic. Off. 2.64, but the overtone of
'alien' seems pointless, if not jarring, here.
Two attempts to solve some of the problems by emendation are worth
mention. Heyne suggested that the correct reading might be not Herculeis
but Herceis, a reference to the altar of Z£vs JpK£1.0s-, the protector of the
enclosure of the household (see Austin on A. 2.513): Evander's hutment
would have been as likely to contain such an altar as Priam's palace, or
rather its ruins, at Lucan 9.979 'H erceas' monstrator ait 'non respicis
aras?'. But once again the emphasis on Evander's Greek customs is not
evidently justified here. Warde Fowler, ASR ad loc., suggested that
Herculeis and hesternum should be transposed to give hesternis (542) and
 COMMENTARY

Herculeum (543). This is an attractively simple remedy, but the result is

straightforward and Ovidian (cf. Met. 8.641 f. above), of the type that
Virgil often consciously rejected.
These and other difficulties-that only Aeneas and Achates had spent
the night with Evander, and the Troiana iuventus of 545 are the socii of 546
whom Aeneas only rejoins later-led Mackail to the probable conclusion
that this passage is confused and unrevised. The one clearly intelligible
feature of it is that sacrifice regularly follows a supernatural appearance,
as at A. 3.176 ff., 5.743 ff. and 7.135 ff.
544 Every word in the line has religious overtones: for laetus see the
note on 275 volentes; adire is regularly used of approaching a consecrated
place (see TLL I, 622, 18 ff. s.v.); for mactat see the note on 85; lectas de
more = 'selected in the traditional way', for victims as well as sacrificers
had to be free of taboo-blemishes (see the note on lecti 179); bidentes means
sheep whose two front teeth (the first of the permanent ones) are notice-
ably longer than the rest (the 'milk' teeth), and so are at a specific stage
of growth and count as hostiae maiores as opposed to lactentes (this correct
explanation of bidentes goes back at least to the Augustan Julius Hyginus,
de Vergilio: his and rival views are discussed by Aulus Gellius N.A. 16.6).
The adj. bidens came to mean 'sheep' because it was the most available
mature animal for sacrifice; others which fulfilled the religious qualifi-
cations could of course be used, cf. Pomponius, cited by Non. Marc. 53 M
Mars, tibi voveo facturum, si umquam redierit, bidenti verre. For the same
group of ideas, cf. A. 4.56 f. principio delubra adeunt pacemque per aras /
exquirunt; mactant lectas de more bidentis.
545 Servius auctus' suggestion (quoted in the note on 15 his coeptis)
that Virgil used the form Euandrus and not Euander in order to avoid the
end-rhyme with pariter may partly account for the form here (and 558
after pater), but Euandrus occurs elsewhere (roo, 185, 313) where no such
reason obtains and is Virgil's preferred form.
546 ~raditur, sc. Aeneas: the context makes this clear; the abrupt
change of subject from the preceding line is surprising only because
Virgil's organised style (like Cicero's) generally takes care to avoid even
small ambiguities of this type. In the comparatively unorganised style of
Ennius and the early annalists abrupt changes of subject are common (see
LHSz 2, 733); but they are natural enough in any case, and there is
hardly any question here of deliberate archaism. Cf. the notes on n9
and 494-6.
548 f. prona . .. aqua: 'on the down-flowing water' of the stream,
prono ... amni (G. r.203); on this they 'float idly (segnis) down', not
needing to row. Whether they arrived or not does not emerge from the later
narrative.
549 The rhythm of the line pictures the effortlessly smooth passage:
accent and ictus coincide in every foot except the second; and the weak
caesura in the third foot (the only one in the line) contributes to the effect,
cf. A. 4.486 spargens umida mella soporiferumque papaver.
552 exsortern: exsors seems to have been first used, if not actually
 COMMENTARY 153

coined, by Virgil, and endowed with two senses (cf. the note on egelido
6ro). Basically it means 'without lot': so in A. 6.428 vitae exsortes it means
'having no lot, share in' ( = expertes, a.KA71poi); here, however it means that
the horse was not drawn by lot, but specially set aside before the division
was made (lfalpETo,) (made explicit at A. 9.270 f. ipsum illum (Turnus'
horse) ... / excipiam sorti,iam nunc tua praemia, Nise). The first and more
natural meaning had a much more flourishing life in Silver Latin than the
second. For the problematic A. 5.534 see R. D. Williams' note.
553 un~uibus aureis: gilding the claws of the lion-skin made it an
especially valuable gift, cf. A. 5.351 f. For the scansion of aureis see the
note on 372.

554-584 The prospect of war looms large in the settlement. Evander regrets
the passing of the strength and valour he had as a young man, and prays
passionately for immediate death if his son is doomed.
555 limina regis, a variant here, looks like a cliche intended to refer to a
regal palace, although in his present situation it can only refer to Tarcho's
tentorium: litora is preferable because it produces a less common phrase
whose meaning however suits the situation: the Etruscan ships were
beached on the shore (497), and their camp was on an inland ridge near
it (cf. 603 ff.). limina and litora are often confused by scribes, and relative
MS authority alone (which in fact here supports litora) is not decisive.
556 f. propiusque periclo / it timor: 'fear goes closer to the danger';
their anxieties go to meet the peril which is not yet at hand, and among
them looms the phantom of War (Martis ... imago is not simply a peri-
phrasis like imagine lunae 23), actual hostilities (Mars praesens, cf. 495)
still being a thing of the future.
557 lam: for the elision see the note on tum 503. See Austin on A. 2.rr2
and 254 for Virgil's surprising fondness for eliding iam, usually in well-
defined word-patterns.
558 dextram complexus (cf. 124) euntis (sc. Pallantos).
559 inexpletus lacrimans: 'weeping insatiably'; for the two-ad-
jective idiom where the first has adverbial force, see the note on 299.
inexpletus, because of its participial form, is an unusually bold instance,
and early attempts were made to regularise it out of existence. Two
variants, both' facilitations', had appeared by the fifth century: inexpletum
(adverbial acc. neuter) lacrimans and inexpletus lacrimis.
The sense of inexpletus is the same as Lucretius' majestic but prosaic
adverb (3.907) insatiabiliter deflevimus aeternumque; for past participle
forms with an -able, -ible sense, see the notes on 195 inaccessam, 588
conspectus.
560 ff. This speech of Evander's is the most high-pitched emotionally
(and therefore rhetorically) in the book: each of its three sections (560-71,
572-7, 578-83) makes up a unit longer than the normal Virgilian maximum
of four lines, a feature which is otherwise exclusive to exordia (see the note
on 374-80). Anaphora plays an obvious part in organising each of the three
154 COMMENTARY

sections. The speech ends with a run-over first foot dactyl (583 vulneret),
like other speeches which trail off in a speechless emotion: A. 6.886
(Anchises' ghost), A. 9.250 (Aletes), A. ro-495 (Tumus). The last word of
this speech is also an omen of the disaster which Evander fears (see Conway
on A. r.165), and with supremo the poet more than hints at what will
happen. 584 brings the episode to an end with all the weight of a para-
graph-closure, a homodyned fourth foot (see Appendix, p. 197) and two
finite verbs 'framing' the line (see the note on 78).
The exemplar of all scenes of leave-taking is the parting of Hector and
Andromache in Iliad 6. But the fierce emotion and pathos here owe much
more to Alexandrian influence, especially Ap. Rh. r.260 ff. where the
heartbreaking effect of separation between Jason and his parents is
exclusively mirrored, as here, in the reactions of those who stay behind:
the young heroes themselves leave, unmoved, in an aura of splendour,
Jason 'like a god', Pallas like the Morning Star (a simile which may
itself have been suggested by Apollonius, see the note on 589 ff.).
560 ff. Evander's sentiment and unrealisable prayer were no doubt
inspired by the words of Nestor at Il. 7.132 ff. at yap, ZEv TE 1ra.TEP Kat
'A871val71 Kat •A1ro..Uov, / ~f3ijJµ,', w, OT' l1r' WKvpoc.p KEAa.Oovn µ,axoVTO / ayp6-
f1,EVOt Ilv>..,ol TE Kat 'ApKO.OE, lyxEalµ,wpoi, I <PEtos 1ra.p TElXEC1C1tV, 'lapoavov aµ,ef,t
ptE8pa. / TOLClt o' 'Epw8a>..lwv 1rp6µ,o, 'taTaTO, lao8Eo, ef,w,, and cf. also
Il. 7.157, n.670 ff. But recalling youthful prowess against local enemies
was obviously common practice among heroes emeriti: Laertes does the
same (with the same prayer-formula) at Od. 24.376 ff., and cf. Aeson in
Val. Fl. r.336 ff. and Lycomedes in Statius Ach. r.776 ff.
The enemies of Nestor and Laertes however were human even if heroic:
Evander's adversary is a Geryon-type monster (for the implications of this
see the note on 563 ff.).
560 o ... referat si luppiter annos: subjunctives of wish came to be
regularly introduced by ut(i)(nam) (or in Old Latin by qui = 'in some way
or other'). (o) si with a present subjunctive to express a wish begins to be
found in Augustan poetry, e.g. Horace Sat. 2.6.8 f. o si angulus ille /
proximus accedat, qui nunc denormat agellum. The tone of o si is exactly the
same as the English 'if only'; the clause is self-subsistent and the present
subjunctive (as in ideal conditions) carries no implication about whether
the wish is or is not capable of being fulfilled; cf. A. 6.187 f. si nunc se nobis
ille aureus arbore ramus / ostendat. In form however we have the protasis
of a conditional, and in the present passage there is a (delayed) apodosis,
at 568 ff. non . .. divellerer . .. neque . .. dedisset . .. , viduasset, where the
tenses, as in prose, convey the impossibility of fulfilment.
561 primam aciem, 'the front rank of the battle-line' ( = 1rpoµ,axov,);
primam here is strictly attributive (cf. A. 9.595), and cum primam here is
not equivalent to cum primum (see the note on 407).
561 Praeneste sub ipsa: the name of the town is feminine here and
at Juvenal 3.190: more often it is neuter as at A. 7.682 where Servius
comments dicimus autem et hoc Praeneste ... et haec Praenestis. Changes of
gender are often caused by the influence of a synonym or related concept:
 COMMENTARY 1 55

here the underlying notion of urbs may account for the feminine; cf. Ovid
Fasti 6.61 ff. inspice Tibur J et Praenestinae moenia sacra deae J .. . nee
Romulus illas (sc. urbes) / condidit; Rutilius Namatianus De Reditu Suo
1.325 f. eminus Igilii (neuter) silvosa cacumina miror, / qitam (sc. insulam)
fraudare nefas laudis honore suae.
562 Evander bums the shields of the enemy dead. This custom, in
honour of Vulcan, was instituted by Tarquinius Priscus after victory over
the Sabines (Servius here and Livy 1.37.5); it was followed by Marcellus
after his defeat of the Carthaginians at Nola (Livy 23.46.5). Warde Fowler
(ASR ad loc.) sees in it an attempt to be rid of the mischief latent in the
possessions of aliens; but the taboo gradually broke down under a more
materialist view of the value of booty. Epics set in the age of heroes
maintained the custom: Val. Fl. 7.28 f. At A. II.193 ff. the Trojans and
Etruscans not only bum valuable spoils (enses decoros); their own dead are
cremated with their non felicia tela (this gives added point to the death of
Turn us: by actually wearing the baldric of his dead foe Pallas he was, in
Roman though not Homeric terms, challenging a taboo-prejudice).
In an interesting article in PVS 5 (1965-6), 26 ff., F. H. Sandbach points
out that when Virgil conceives of men in the mass he describes their
shields as scuta (as opposed to the clipei of individual warriors). These,
the shields of the contemporary Roman army, were made of wood with a
covering of leather and a rim and boss of metal: hence they were com-
bustible (as here) and buoyant on water (as at A. 8.539).
563 ff. Erulus is clearly an invented doublet of Geryon: both had three
bodies and hence three lives (compare Aesch. 869 ff. with Servius here).
By implication, Evander in his prime was as powerful a monster-slayer as
Hercules.
Feronia appears at A. 7.800, in connection with the Jupiter of Anxur
(Tarracina), as deity of the grove to the north of that town (on the border
of the Pomptine marshes), where there was also a spring (Horace Sat.
1.5.24 ora manusque tua lavimus, Feronia, lympha). She was widely
worshipped in central Italy (see Latte, RRG 189), especially at the lucus
Capenatis near Mt Soracte (Livy 1.30.5 and A. 7.697), and in origin was
probably an Italic wood-water spirit, as suitable as any other for mothering
on Erulus. (In historical times she seems to have been regarded primarily as
protectress of slaves; according to Servius auctus here there was a stone
seat in her temple at Tarracina with the engraved verse bene meriti servi
sedeant, surgant liberi.)
564 ff. tris ... terna ... ter: the repetition of the same word at the
beginning of successive clauses (anaphora) is a common method of concen-
trating emphasis in colloquial speech; it also organises the structure of such
a succession of clauses, and so naturally became a figure of rhetoric,
frequent in all literary Latin, prose and verse. Virgil is fond of a group of
three (cf. the repetition of si at 574 ff.); in the present instance note the
variation in the form of the word, and cf. A. 1.448 f. aerea cui gradibus
surgebant limina nexaeque / aere trabes, foribus cardo stridebat aenis.
'at his birth his mother Feronia had given him three lives-a fearful
E>.C
 COMMENTARY

thing to tell-and three sets of arms to wield (three times he had to be laid
low in death; but then this right hand of mine robbed him of all his lives
and stripped him of all three sets of arms).'
terna arma movenda should be connected with what precedes, because
of the obvious parallelism between nascenti . .. animas . .. arma and leto . ..
animas ... armis; if the phrase is joined with what follows it means that
Evander had to wield three sets of arms: simultaneously? impossible; in
succession? implausible.
568 f. umquam at the end of the line echoes usquam at the end of the
line before with rhetorical emphasis, and is comparable with the repetition
at 271 f.: contrast fuisset 396 f. and the note there.
There is no significant difference of meaning between usquam 568 and
umquam (v.l. usquam) 569; the chief function of both is to emphasise the
negative.
569 f. finitimo. . . / huic capiti insultans: 'trampling on my
neighbouring independence'. For the meaning of capiti see the note on
144 f. finitimo is best taken as an adjective with huic capiti, a rhetorical
balance to dulci tuo amplexu (and cf. Soph. Antigone l cl, Koivov
a?na.Sd¢,ov 'Iaµ,~VTJ• Ka.pa); huic = meo, as often in the popular
speech of comedy (e.g. Plautus Stichus 751 fugit hoc libertas caput,
where hoc caput virtually = me).
Ribbeck's proposal to readfinitimos with the corrector of the Palatine
MS produces a feasible Virgilian variation (finitimos . .. saeva dedisset
funera = finitimos saeve reddidisset funera (i.q. cadavera); for dare funera
cf. G. 3.246 f.), but the authority for the reading is weak, and the need for
adoption slight.
571 The reference is not to the citizens of Caere (489 f.) but to those of
the Palatine whom Evander had been too old and weak to protect.
573 ff. Like the strong caesura with a pause in the third foot (see the
note on 49 ff.), a pause at the strong caesura in the fourth also has a
tendency to echo in the following line: cf. 60 f., 180 f., 256 f., 287 f., 293 f.,
376 f., 403 f., 532 f., 565 f., 667 f.; here, in 573-5 it echoes in the following
two lines. All the same, most lines with this caesura occur in isolation from
another. The strong caesura in the fourth foot seems an entirely natural
place to have a sense pause: but it was uncommon in Greek hexameter
poetry, and is a Latin exploitation (see Winbolt 42 ff.).
576 venturus in unum: cf. E. 7.2 compulerantque greges Corydon et
Thyrsis in unum, and A. 2. 716 hanc ex diver so sedem veniemus in unam, 'to
one and the same place'; Greek uses the same idiom, El, lv lpxtau9a,.
Translate: 'If I live with the prospect of seeing and meeting him.'
577 durare laborem: durare originally = 'to harden' (trans.); then
'to be hardened, patient' (intrans.), often coupled with an infinitive on the
analogy of patior; finally, in Augustan poetry and post-Augustan prose,
'to endure' with a direct object, but only here in Virgil.
578 sin aliquem: it is usual to find the unemphatic (enclitic) forms
quis, qui immediately following si(n), nisi, ne, num and relative pronouns;
aliqui(s) in this position has a stronger tone, 'any at all', cf. Livy 39.28. 13
 COMMENTARY 1 57

sin aliquis respectus est mei, 'if you have any regard for me at all'. When
the indefinite pronoun or adjective is separated from the introductory
word and cannot lean on it enclitically, aliqui(s) is the rule, e.g. Livy
1.17.4 timor . .. patres incessit, ne civitatem sine imperio, exercitum sine
duce . .. vis aliqua externa adoriretur.
579 crudelem abrumpere vitam: the same words are used by
Euryalus' mother praying to have her grief for her son's death ended by
death for herself also (A. 9.497). Here however Evander is looking into the
future and means a life which will be cruel if he outlives his son (crudelem
is proleptic). It is tempting to think that the 'later' passage in A. 9 was
composed first and afterwards remembered for the present passage.
580 spes: generally = 'hope'; much less frequently, as here, 'antici-
pation, expectation', like Greek i>..TTfr, cf. LS s.v.
581 mea . .. voluptas recalls the endearments of comedy as at Plautus
True. 353 (and love-poetry), just as artificis scelus (whatever its construc-
tion at A. 11.407) recalls its abuse (in phrases like scelus viri, etc.).
sera, because he was the child of Evander's old age. 'Late' is a compara-
tive notion, 'only' a superlative one: there is a slight gain in rhetorical
point if the latter follows the former, and sera et sola is in fact offered by
some MSS.
582 neu: when, in an independent sentence, a negatived jussive sub-
junctive follows one which is itself negatived (by ne) it is regularly intro-
duced by neve (neu); when as here it follows one which is positive it is
regularly introduced by neque (nee): neu in this second case is a rarity of
Augustan poetry (cf. A. 7.265), and its rarity may account for the variant
ne (but a final clause is a less effective end to the rhetorical sweep of the
passage). See further LHSz 2,338; KS 1, 193.

585-607 The cavalry column starts off, with Aeneas at its head and Pallas,
conspicuously dressed, in its centre, and gallops towards the sacred grove of
Silvanus, near to which the Etruscans are encamped; there they arrive and rest.
Henry well compares the brilliance of 585 ff., and its position in the
narrative, with A. 4.129 ff.: 'To the reader unacquainted with the sequel
both pictures are as gay and exhilarating as they stand to him who reads
the story for the second, or it may be for the hundredth time, in the saddest
contrast to the grim catastrophes by which their sunny morning brightness
is so soon, so very soon, to be overcast, and for ever extinguished.'
585 iamque adeo: iam can be used like the English 'now' without any
temporal sense, simply to mark a transition to a new subject: this is
particularly clear in didactic or 'lecturing' contexts like Cic. Inv. 2.68
iam iura legitima ex legibus cognosci oportebit, Virgil G. 2.57 iam quae
seminibus iactis se sustulit arbos. The use of adeo as here to throw emphasis
on the word it follows, like the Greek enclitic y1:, is common in Roman
comedy and a favourite with Virgil; iamque adeo is used in this same way
at A. 2.567, 5.268, 864, 11.487.
588 Mynors in his Oxford Classical Text accepts Markland's conjecture
12-2
 COMMENTARY

it, for the MSS' in, as the first word of the line, but an expressed main verb
is not indispensably required.
588 chlamyde et pictis conspectus in armis: 'conspicuous in his
cloak and emblazoned arms'; chlamys is a Greek military cloak, suitable for
an Arcadian commander. For the past participle acquiring a sense equi-
valent to an adjective in -bilis (conspectus = 'remarkable') see the note on
inaccessam 195.
589 ff. Pallas is radiantly dazzling like the Morning Star rising from the
Ocean-stream (the Greeks had given the name 'ilKmv6s- to the river which
they imagined to encircle the then-known earth and to be coterminous with
the horizon). The simile owes something to Il. 22.317 ff. (where the armed
Achilles is as dazzling as the Evening Star): ofos- o' O.arTJP Ela, µer' a.UTpaa,
VVKTOS' dµo>..ycj, I Ea1TEpOS', Bs- Killt<n'OS' EV ovpav{j, 'lUTaTai O.CTT7Jp, I WS' alxµijs-
d1rl>..aµ1r' EV'7/KEOS', ~v ap' 'AxiAAEVS' / 1rillEv OEfmpfi, more to Il. 5.4 ff. (whose
context Virgil had in mind at 620 and 680 f.), where Diomedes is compared
with the star of summer (Sirius, the Dog Star) after it has bathed in the
Ocean-stream: Oa'il ol EK K6pv86, TE Ka, aa1rloos- U.KaµaTOV 1rfJp, / O.UTEp'
01rwpiv{j, Eva>..lyKiov, as- TE µa.AiUTa / >..aµ1rpov 1raµcf,alvr,a, AEAovµlvos- 'ilKEavo'io ·
/ To'i6v ol 1rfJp oa'iev a.1ro KpaT6, TE Ka, wµwv (Apollonius too had imitated
this simile, and compared Jason with Sirius rising from Ocean, Arg.
3.956 ff.).
Lucifer . .. Venus: when the planet Venus appears in the eastern sky
before dawn it is called Lucifer, 'Bringer of the light' (of day); in the
western sky after sunset it is called Hesperus, the Evening Star (cf.
Catullus 62.35 Hespere, mutato ... nomine Eous).
It was significantly Lucifer which rose at the end of the night of the sack
of Troy (A. 2.801), and this guiding star was Aeneas' mother's (cf. A. 1.382
matre dea monstrante viam).
ante alios . .. ignis: Servius tells us that Venus also had stars in the
constellations Taurus and Ursa Major: these of course would be fixed
stars; for ignis = 'heavenly body', cf. Catullus 62.20.
594 f. 'They advance armed through the thickets, by the way that brings
the goal of their journey nearest.' Thickets would be as unhelpful to the
cavalry here as they were helpful for the stealthy advance of the Gallic
infantry on the Capitol at 657, but they chose that way because it was the
shortest to their objective (qua proxima meta viarum is a curious para-
phrase for this); the thickets might conceal an enemy ambush, so they
proceed armati (perhaps they have also dismounted); when they emerge
on to the open plain they raise the signal and gallop off.
For olli see the note on 94, and for the pauses of 595 the note on 260.
596 The sound and rhythm of this splendid onomatopoeic line, probably
the most obvious and quoted example of descriptive dactyls in Virgil, owes
a lot to Ennius, as Macrobius (Sat. 6.1.22) points out: cf. Ann. 224 V
explorant N umidae totum: quatit ungula terram; Ann. 277 V consequitur.
summo sonitu quatit ungula terram; Ann. 439 V it eques et plausu cava
concutit ungula terram. Even so, 'four-footed, galloping sound' looks like
Virgil's own invention, and a rather bold one, more reminiscent of the
 COMMENTARY 1 59

language of Greek tragedy than unfanciful Latin: contrast A. II.875

quadrupedumque putrem cursu quatit ungula campum. For the compound
adjective, see the note on 82 concolor.
597 est ingens . .. lucus . .. amnem. . . (603) baud procul hinc . ..
(606) hue: an interruption of the main thread of the story to set the scene
for the next main event, (606) hue pater Aeneas ... This is an entirely
natural formula of narrative; it started in Homer (Il. 6.152 f. is an in-
conspicuous example, Il. 2.8n ff. is more developed), and was taken up
in the messengers' narratives of tragedy (e.g. Aesch. Persae 447 ff.); for
its subsequent history see Austin's notes on A. 4.480 ff. and 483 (it seems
eventually, perhaps through over-elaboration, to have exposed itself to
parodies of the high-flown style). The main narrative is always picked up
after the eK<ppaui, T61rov with a demonstrative pronoun or adverb (hue,
hinc, etc.). The simplest kind of ei<<ppaai, begins est locus . .. (A. 3.163, an
early book which contains a surprising number of fairly short instances,
see R. D. Williams on A. 3.13-16), but there are numerous variations like
A. 8.233 stabat acuta silex. . . 236 hanc. The beginning can be quite
disguised and worked into the main narrative, A. 8.190 .. . hanc aspice
rupem ... 193 hie (cf. A. 6.548 ff. and 703 ff.). A. 8.416 ff.( ... 423 hoc) is
elaborated into a topographical inset; and the example here bifurcates into
603 haud procul hinc and 606 hue (compare the internal divisions of the very
elaborate A. r.159 ff.). The example at 478 ff. shows how natural and
integral to narrative this formula was.
597 f. Another numinous grove (see 348 ff.), again holy and awesome
(see religio, 349 f.).
Caeritis amnem: probably not the Minio (Mignone), as Servius
thought (it is coupled with Caere at A. ro.183 in a regional not local
description, and was in fact the boundary between the territories of Caere
and Tarquinia), but the Vaccina which runs alongside the town, or
perhaps even the stream which feeds it, the Fosso della Mola.
The genitive of the noun Caere occurs only here (cf. TLL Onomasticon
32, 3 s.v.).
599 cavi: the hills are not 'hollow' in themselves, but as a unit
enclosing the grove.
nigra nemus abiete: possibly on the ridge to the south-east of
Cerveteri, whose name, Monte Abetone or Abatone, may well derive from
abies, like Italian abetina, ' a fir-grove'.
599 nigra . .. abiete, 'black fir trees'. Collective singular nouns denoting
the material from which something is or can be made are common, cf.
Caes. E.G. 5.12.5 materia cuiusque generis . .. est praeter fagum atque
abietem, where abietem really means 'fir wood', and the same is true of
A. 2.16 sectaque intexunt abiete costas. This quite understandable idiom was
extended by the poets to cases like the present one where a number of
individual items is referred to. See Lofstedt 12 , 14.
599 abiete: note the scansion; only the nominative singular of abies
(as at A. 8.91), aries, par'ies allow themselves to be fitted naturally into
hexameter verse; the other cases are made tractable by treating the i as
 160 COMMENTARY

semi-consonantal (pronounced like y), and this, with the consonant b there
already, causes the first syllable to be lengthened (but does not alter the
fact that the vowel-sound a is and remains short). Ennius seems to have
been already quite bold in using this device, cf. Ann. 436 V hie insidiantes
vigilant, partim requiescunt; Ann. 94 V avium, praepetibus sese pulchrisque
locis dant. Virgil on the contrary spares his reader any hesitation over the
scansion of the word by placing abiete (and ariete) invariably in the dactylic
fifth foot.
600 f. The woodland in which Silvanus protects the flocks is naturally
dedicated to him. As a rural god he had no organised state worship in the
urbs Roma itself, but there is no evidence of a fixed festival in the country-
side either. Virgil is probably following the common identification of
Silvanus with Faunus, whose country festival at Mandela on the Nones
(5th) of December is fully described by Horace Odes 3.18: hencediemque601.
Silvanus would be at home near Caere: he has an Etruscan equivalent
Selva(ns) (see Pallottino 165, and Lily Ross Taylor, Local Cults in Etruria
(Rome, 1923), 123 f.).
600 Pelasgos: ancient and modern speculation about the ethnic
origins of the Pelasgi is confused and inconclusive. Elsewhere in the
Aeneid (1.624, 2.83, 106, 152, 6.503, 9.154) the word invariably means
'Greek', as it probably does here (see the note on 328; primi in 602 is not
to be taken = 'first in succession'). For Virgil they were pre-Etruscan
settlers (Servius auctus on this line quotes a tradition which actually
identified them with the Etruscans, to whom, however, Virgil (A. 8.478 ff.)
gives a Lydian, non-Greek, origin). See also Saunders 46 ff., Rehm 63 f.,
Scollard 35 ff.
602 primi: 'in the earliest times'; it is most unlikely that Virgil
wished to suggest that the Pelasgian settlement was prior to that of the
immigrants mentioned in 328 ff.
aliquando here = •once long ago' (contrast 200); like olim it basically
means 'at some time other than the present' and so can refer to either
past or future.
603 f. tuta . .. castra locis: apparently = •their camp protected by
the terrain', tutus being the past participle of tu(e)o(r), 'to look at, to look
after'; the construction is the same as secretum fl,umine 610. But how can
a camp which is pitched in broad fields and dominated by a nearby hill
(6o4 f.) be said to be so protected? Servius saw the difficulty and envisaged
a plateau surrounded by hills. Mackail, less probably, suggested that
tuta = 'guarded' (by patrols), and that haud-procul-hinc is a qualifying
phrase with locis equivalent in sense to propinquis.
603 Note the alliteration of t in this line, and of c and l in the next,
forming a background sound-pattern which does not reinforce the sense.
605 tendebat: 'was under canvas'; used absolutely the verb is a
military term= 'to pitch tent' (itself derived from the past participle).
606 f. The Trojans and Arcadians presumably encamp for the night, so
the following apparition of Venus (symbolised in the Evening Star?)
occurs at night(fall) like most supernatural visions in the Aeneid.
 COMMENTARY 161

Aeneas' forces now disappear from the narrative and do not reappear
until A. ro.146: the whole of A. 9 is taken up with an account of Turnus'
attack on the Trojan camp by the Tiber of which Ascanius was in
command; and A. ro.1-n7 with the debates of the gods which end with
Jupiter's oath not to interfere with the natural course of events, and
A. ro.n8-45 with an account of interim fighting at the Trojan camp. The
crucial moment when Aeneas is actually accepted as the Etruscan leader
is delayed: perhaps an intentional case of dramatic suspense otherwise
rare in Virgil.

608-625 Venus appears and encourages Aeneas with the gift of arms. He is
lost in wonder at the helmet, sword, breastplate, greaves, spear and shield.
608 inter: for the position of inter, and the word-symmetry of the
whole line, see the note on 32.
6ro egelido: nimium gelido ... ut eduramque pirum (G. 4.145), id est,
valde duram, Servius. This must certainly be right (the same stream was
referred to as gelidum 'cold' in 597) and is very surprising. Catullus seems
to have been the first to use the word (46.1 iam ver egelidos refert tepores),
to mean 'with the chill taken out', as one might expect from the usual
privative sense of e- (elinguis, edentulus, etc.). Virgil reverses the meaning
of the word, giving e- a very rare intensive force (perhaps suggested by
verbal compounds like emori, emunio or by Greek intensive d-). Manilius
(5.131) later uses the word to mean 'intensely cold'; Ausonius applies it
to both the Moselle (Epp. 20.4) and the Danube (Caes. 21.1), apparently
with the same meaning.
The variant et gelido has formidable MS authority, but egelido was the
only reading known to Servius and is backed by the indisputable eduram
at G. 4.145.
secretum (after reducta): Aeneas is alone and solitary, ready for
communion with the supernatural, like Moses (Exodus 19) and Odysseus
(Od. 12.335 ff.). Achilles, however, was with his companions at the presen-
tation of the arms (fl. 19.3 ff.).
secretum is best taken as a participle, with flumine abl. of instrument:
'cut off by the river'.
6n This line is noted by Servius as a case of vUTEpov 1rp6-rEpov (Venus
presumably 'presented herself' before she spoke); he might have added
615 f. below, where Virgil's order suggests that the goddess embraced her
son before depositing the arms. Instances of this mannerism with two
clauses (see the note on 201 for nouns) are bold: they are a natural develop-
ment of Virgil's paratactic manner, and usually he assembles the 'theme
and variation(s)' in chronological order: when he does not, other consider-
ations (relative emphasis especially) were more important than time-
sequence, as at A. 2.353 moriamur et in media arma ruamus (see Austin's
note). So here, the fact that the goddess spoke is unimportant compared
with her appearance in person (ultra). See further A. S. McDevitt in CQ
N.S. 17 (1967), 316-21.
162 COMMENTARY

612 f. en ... munera: the exclamatory particle en can be followed by

either a nominative or an accusative of exclamation, cf. E. 5.65 en quattuor
aras, A. 7 .452 en ego victa situ; alternation was possible because en was not
felt to have any close syntactical relation with the noun. See KS 1, 272 ff.
for examples, but ignore the explanation given there.
613 ne ... : syntactically this clause could be either independent (a
negative jussive) or dependent (a negative final clause); in speech the
difference would be made clear by emphasis and tone of voice. It is
certainly more lively and dramatic to take the clause as independent, with
a heavy pause after the first foot dactyl munera.
614 poscere here = provocare, 'to challenge'; with Turnum as object
it hints at the duel of A. 12.
616 Note the word-symmetry, approximating to that of a golden line;
see Appendix, p. 199.
617 donis, 'theme': honore (= munere), 'variation'.
618 oculos belongs dm~ Koivoii to both expleri and volvit; see the note
on 265 for the construction expleri nequit oculos.
619 f. Modelled, as Macrobius 5.8.u f. points out, after Homer, ll.
19.18 Ttp1T€TO o' £V X£Lp€UULV lxwv (hoii dy,\ad. Swpa.
619 versat, 'keeps turning over', of fascinated inspection also at
A. 5.408.
620 f. vomentem . .. rigentem: disyllabic rhymes at the end of two
consecutive lines are not so very rare in Virgil (in this book cf. 646 f.
iubebat ... premebat). Deliberate rhyming is native to Latin popular verse,
from proverbs and charms against gout to Christian hymns. Virgil, unlike
Milton in Paradise Lost, obviously did not avoid' the jingling sound of like
endings' as a matter of policy; on the other hand, he does not intentionally
admit them to echo their popular overtones, and they cannot be said to
occur more often than they would by accident. Rhyme also occurs between
present participles which 'frame' a line: 669 pendentem . .. trementem; and
cf. 63 stringentem . .. secantem, 712 pandentemque . .. vocantem. See further
Austin on A. 4.55.
vomentem is probably right against the variant minantem, cf. 680 f.;
translate 'spurting' (Virgil clearly did not regard it as a sordidum voca-
bulum). 'Spurting flames' may refer to the colour and gleam of the metal
(cf. ll. 5.4 f.), but the 'theme and variation' shape of the line strongly
suggests a reference to the shape and colour of the plumes themselves (see
the note on 680 f.).
621 'and the death-dealing sword, and the stiff breastplate of bronze';
for rigentem cf. A. 5.404 f. tantorum ingentia septem / terga boitm plutnbo
insuto ferroque rigebant, and A. u.72 f. tum geminas vestis aitroque ostroque
rigentis / extulit Aeneas. It is unclear whether Virgil visualised the breast-
plate as of solid bronze or of bronze plates threaded together with wire;
see F. H. Sandbach, PVS 5 (1965-6), 31 f.
621 According to Virgil (A. 12.90 f.) Vulcan had also made an invincible
sword for Daunus which had come into the possession of his son Turn us:
this is one sign of an attempt to build up the native hero as a significant
 COMMENTARY r63

opponent of Aeneas; another is Virgil's choice for the mother of Turnus

of Venilia, a deity of extreme obscurity, whose very name however
contrasted her with her opposite number Venus.
622 f. Achilles' breastplate receives only a short, almost perfunctory,
description, Il. 18.6m (Hephaestus) TEvg° a.pa ol BclJpTJKa ef,aEiv6TEpov 7Tvpo,
avyfj,. The simile here owes much more to Apollonius' colourful descrip-
tion of the golden fleece, Arg. 4.125 f.: (Kwa,) . .• vEef,l>..r, ivaMyKwv, ijT'
dv,6VTo, / ~EMov ¢,>..oyEpfia,v ipEvBETa, aKTlvEaa,v. Aeneas' breastplate of
bronze, which can reflect a number of quite different colours, has a cloudy
blue-grey ground shot through with blood-red glints.
624 recocto: saepe purgato (Servius), 'highly refined'.
625 clipei non enarrabile textum: 'the indescribable fabric of the
shield', whose material composition defied analysis; but there is more
than a hint at the 'indescribability' of the pictures on the surface, a
warning that the description of them is not to be pressed for pictorial
completeness.
The decoration on the shield is in one sense a supreme culmination of the
tradition of heraldic emblems: Helenor, parma inglorius alba (A. 9.548),
had not yet won his spurs and so had no device; the proven heroes of
Aeschylus' Seven against Thebes choose monster-totems or propaganda-
symbols for their emblems; Aeneas' personal emblem is to be a vision of
history, totally surpassing any of the traditional categories of device
(analysed for Greek life and literature by G. H. Chase, HSCPh 13 (1902),
61 ff.).

626-666 On the shield Vulcan had fashioned scenes from Roman history:
the she-wolf suckling Romulus and Remus; the rape of the Sabines and the
war between Romulus and Titus Tatius, concluded by a peace-treaty; Tullus
Hostilius' punishment of the treachery of Mettus Fufetius; Porsena's siege
of Rome after the expulsion of Tarquin, and the heroism of Horatius Cocles and
Cloelia; Manlius' defence of the Capitol against the night-attack of the Gauls;
the Salii and Luperci.
627 'in full awareness of the predictions of prophets and knowledge of
the age to come': the theme is made clear by the variation: Vulcan knew
the truth of the responsa given by the mouthpieces of the gods, and their
details of Rome's future history; for the implication of truth in prophecies
cf. A. 6.798 f. huius in adventum iam nunc et Caspia regna / responsis
horrent divum.
Damste's conjecture haud avium ignarus (Mnem. N.S. 38 (19ro), 52)
implausibly makes Vulcan into a skilled auspex like Asilas (A. ro.175 ff.).
630 ff. The idea of the Shield of Aeneas unquestionably derives
from Homer's Shield of Achilles, but the subject-matter of the two has
no detectable connection. Virgil's subject-matter falls into four main
sections:
(1) 630-66: early Roman history down to the Gallic invasion, with an
unexpected addendum on religious institutions, 663 ff;
 COMMENTARY

(2) hinc procul, 666-70, Cato and Catiline, saint and devil of the late
Republic;
(3) 671-713, the sea and Actia bella;
(4) 714-28, the triumphs and ex voto dedications of Augustus.
Critics have been much exercised to discover what principle(s) Virgil
was following in selecting the material (particularly of section (1)). Heyne
looked for a certum iudicium, failed to find it, and blamed Virgil (Excursus
IV to A. 8). Warde Fowler (ASR 103 ff.) thought there was one unifying
theme, that of escape from moments of great peril which threatened the
nomen Romanum with extinction. Brooks Otis (341) also sees one main
theme, 'the constant opposition of virtus, consilium and pietas to the
forces of violence in all Roman history'. D. L. Drew (The Allegory of the
Aeneid (Oxford, 1927), 26-31) had suggested that the material falls into
four groups, each illustrating one of the cardinal virtues of Augustan
Rome, virtus, dementia, iustitia, and pietas, the qualities specified on the
gold shield granted by senate and people to Augustus in 27 B.C. (Res Gestae
34). With goodwill each of these schemes can be made to fit the bulk of
the material, but in each case there is some inexplicable residue: this is well
demonstrated by D. E. Eichholz (PVS 6 (1966-7), 45 ff.), whose own
solution, however, that coherence derives from the dynamism of the
narrative, is unconvincing. The view that the episodes were chosen
because of their visual suitability as pictures, and were primarily inspired by
particular statue groups which Virgil had seen in Rome, can also be
dismissed: some details simply cannot be' envisaged' (see the note on 634).
The subject-matter is announced as res Italas Romanorumque triumphos.
This had already been treated in poetry; every Roman knew by whom, and
none had read it with deeper admiration than Virgil. Servius (on A. 8.631)
remarks sane totus hie locus Ennianus est. Uncertainties about the Annales
themselves make it impossible to say positively that the Shield is Virgil's
most sustained tribute to Ennius in the Aeneid, but such evidence as there
is strongly suggests it. Section (1) contains obvious similarities of subject-
matter with the fragments of the Annales. The She-Wolf and the Twins
(630 ff.) recall Ann. 68 ff. V; the punishment of Mettus Fufetius (642 ff.)
recalls Ann. 126 and 137-9 V (plausibly connected with each other by
Warmington); the Gallic attack on the Capitol (655 ff.) strongly recalls
Ann. 164 f. V. So far Virgil, like Ennius, has followed chronological order.
The episodes chosen are those which stirred his imagination without pro-
ducing a close repetition of material used for the Show of Heroes in
A. 6.756 ff.: hence there is no mention in the Shield of the Alban kings,
Numa, Camillus, Cato the Censor or Hannibal (who lurks behind A. 6.842 f.
geminos, duo Julmina belli, / Scipiadas, cladem Libyae).
Chronology too is the key to the puzzlingly abrupt transition to primi-
tive religious institutions whose origins were lost in the mists of time
(663 ff.). The passage recalls Ennius' reference to the institutions of Numa,
Ann. 120 ff. V: both passages mention ancilia, and Virgil's lanigeros apices
is an intelligible equivalent of Ennius' tutulatos, 'conical-capped'. But the
hero of the crisis which followed the Gallic withdrawal from Rome was not
 COMMENTARY

of course Numa, but M. Furius Camillus, diligentissimus religionum cultor

(Livy 5.50.1) and the clue to the presence of the Salii here is in the last
words which Livy puts into his mouth (5.54-7): hie (sc. in Capitolio) Vestae
ignes, hie ancilia caelo demissa, hie omnes propitiimanentibus vobis di. The
whole stress of Camillus' speech, urging the Romans not to transplant
themselves to Veii in despair at the devastation caused by the Gauls, is on
the sanctity and protective power of the gods and the sacra publica. After
describing the peril of the Gauls' invasion, Virgil thinks of the religio and
pietas which, after their withdrawal, had kept Rome a living city; so he
reflected, with Camillus (Livy 5.52.6 f.) ne omnia generatim sacra omnesque
percenseam deos . .. quid de ancilibus vestris, Mars Gradive? After the Salii
the priestly college of the Luperci naturally suggested itself as a companion.
The matrons in religious procession, riding in two-wheeled carriages, are
mentioned next, because it was they who had raised the gold to ransom
Rome from the Gauls, and for this were honoured by the privilege of riding
in carriages (this tradition, represented by Diodorus 14.rr6, is clearly
what Virgil had in mind, and in this matter he is not following the same
source as Livy (5.25.9), who says that this honour was conferred on matrons
after they had contributed their gold to the building of the temple of
Apollo, an ex voto duty of Camillus after the capture of Veii. How Ennius
treated the topic we shall unfortunately never know). Servius' contri-
bution to the Salii-problem is the remark congrue sacerdotes statim post
templa commemorat: superficial and tangential, but a more direct lead to
the solution than any of the speculations of modern criticism.
For further discussion of the Salii-scene and its bearing on the relation-
ship of Livy's first pentad with Aeneid 8 see P. T. Eden, RhM n6 (1973),
78 ff.
Section (2) contains the well-contrasted pair Catiline and Cato Uticensis.
Which suggested itself to Virgil first? Ennius had paid a glowing tribute to
the man who first recognised his talent, Cato the Censor: in caelum huius
(Cato the Younger) proavus Cato tollitur: magnus honos populi Romani rebus
adiungitur, Cic. Arch. 22 (see Vahlen pp. iv, xii).
Sections (3) and (4), the victory and triumphs of Augustus, are the
centre of attraction of the Shield pictorially, and the culmination and end
of the narrative. There is good evidence (Pliny N.H. 7.(28).101) for
believing that Ennius originally ended the Annales (Book 15) with the
victory and triumph of his patron, M. Fulvius Nobilior, over the Aetolians
at Ambracia (not very far from the promontory of Actium) : victoriam per se
magnificam Q. Ennius amicus eius insigni laude celebravit (Aurelius
Victor, De Vir. Illustr. 52).
Virgil did not, of course, need Ennian precedent to describe Actium, nor
could an unselective reading of the Annales have predisposed him to
highlight Roman virtues in action at or near the city itself in periods of
crisis, as he does. But both in particular detail and in general structure the
Shield as narrative has more points of contact with the fragmentary
Annales than it has as description with any single surviving work of
Greco-Roman art.
166 COMMENTARY

630 f. fecerat. .. procubuisse: literally, 'he had fashioned the she-

wolf as having lain down'; after verbs of 'representing' somebody or
something in art or literature, Jacere, fingere, inducere, an infinitive or
participle can be used with no appreciable difference of meaning (exactly
as with verbs of perception, cf. 92 f.); compare Cic. T.D. 5.115 Poly-
phemitm Homerus . .. cum ariete etiam colloquentem facit, with Cic. Nat.
Deorum 3.41 quem tamen Homerus conveniri apud inferos facit ab Ulixe;
for pictorial representation cf. Claudian De Raptu Pros. 2-44 f. Hie
Hyperionio Solem de semine nasci / fecerat (sc. Prosperpina, in embroidering
her dress). Elsewhere in the description of the Shield Virgil uses present
participles (669 f., 711 ff.); at 710 ff. a present infinitive alternating with
present participles. The time-reference of procubuisse here in contrast with
the following infinitives ludere ... lambere, etc., should be noted: the she-
wolf is represented as having lain down already with the twins playing
around her.
addiderat (637) and addit (666) continue the sense and so naturally the
construction of fecerat.
See further Ernout-Thomas 283; Anton Szantyr, 'Bemerkungen zum
Aufbau der Vergilischen Ekphrasis', MH 27 (1970), 28 ff.
630 f. fetam . .. lupam: fetus = 'with young' which may or may not
have been delivered, according to the context; Columella (7.3.26) draws a
distinction between (oves) agiles et Jetae and tardiores et gravidae.
630 Mavortis in antro: no doubt the Lupercal which became associ-
ated with the legend of the lupa, called 'the cave of Mars' because Ilia
had borne Romulus and Remus to him (as Jupiter predicted A. 1.273 ff.).
Virgil clearly felt at liberty to allude to this version of how the Lupercal
got its name, in spite of his earlier statement of the alternative version at
A. 8.344.
631 ubera circum: another simple case of anastrophe: see the note
on 32 inter. It is on this line that Servius remarks 'to be sure, all this
passage is Ennian ', a remark which presumably applies directly to the
She-Wolf and Twins; for the wider implications which it suggests, see
above, 630 ff. In the very fragmentary state of Annales 1, one can detect
only one similarity with 630 f., and that of very expected words: 68 V
lupus f emina feta repente.
632 R. Rau, MH 22 (1965), 237 f., wishes to delete this line as an inter-
polation on the grounds that: (1) pendentis only makes sense if the she-wolf
is standing upright, whereas in fact she had lain down (631) (captious
precision: the meaning of pendere is not restricted to a vertical plane);
(2) matrem is the subject of lambere and this idea is then needlessly repeated
in 634 corpora fingere lingua (this is simply perverse: pueros is the subject
of lambere = virtually 'suck').
Rau is refuted in detail by J. Delz, MH 23 (1966), 224 ff., who adds that
Virgil's recumbent she-wolf is at variance with the standing she-wolf of
plastic art-not surprisingly, for Virgil was working from literary sources:
cf. Fabius Pictor mediated by Dionysius of Halicarnassus 1.79.6 f.,
especially W!, 3€ KaK1:,vo~ 1TA71alov J),,8611T1:!, £µ.a8ov TtJV P,€V WU1T1:p -rlKva 1T1:pd-
 COMMENTARY 167

1TOUaav, TU. s· tiJS' fl,'TJTPOS' Jt€x6p,€va ( = pendentes), Saiµ,611,611 n xp~µ,a opa.v

v1ro>.af36VT€S' KTA.
633 tereti cervice reflexa(m): this has generally been thought to be
an echo of tereti cervice reposta, Mars 'pillowing his shapely neck' on Venus'
lap, at Lucretius 1.35, where Bailey comments 'Virgil loves to imitate
Lucretius in a slightly less dignified context', of which there is very little
evidence. Virgil may have taken his phrase directly from Cicero's descrip-
tion of the constellation Draco in his translation of Aratus (cited in Nat.
Deorum 2.ro7), obstipum caput, a(t) tereti cervice reflexum, / obtutum in
cauda maioris figere dicas. In spite of apparently strong MS authority for
reflexa, the variant reflexam, preferred by Heyne and Sabbadini, is
supported by Cicero's phrase, and by the un Virgilian succession of three
words in agreement in the alternative. And MSS written in capitals (like
the primary MSS of Virgil) commonly represent an M at the end of the
line merely by a nasal suspension (a short horizontal line), very easily
overlooked by a scribe at an earlier stage of the transmission. A. ro.535 f.
and 11.622 give no support to reflexa here.
634 mulcere alternos: 'was stroking each in turn' -very difficult to
represent pictorially, and an indication that Virgil was not at pains to
describe the shield as a material object. rubescunt 695 and iam iamque 708
are also difficult to visualise in terms of static art.
634 corpora fingere lingua: Conington compares Virgil's remark about
himself licking his verses into shape, like a mother-bear her cubs (lam-
bendo ... conformaret et fingeret, Aulus Genius N.A. 17.ro.3).
635 sine more: 'lawlessly', cf. Ovid A.A. r.n9 sic illae (Sabinae)
timuere viros sine more (v.I. sine lege) ruentes.
636 consessu caveae: 'from the seated audience', cf. Lucretius 4.78
consessum caveai subter and A. 5.340 caveae consessum.
636 magnis circensibus actis: 'during the performance of the great
circus-games'. In origin the perfect passive participle was not exclusively
'perfect' with a past time-reference (nor exclusively passive). Classical
prose uses the participles in -tus, -sus of many deponent verbs with a present
time~reference, e.g. Caesar E.G. 2.7.1 isdem ducibus usus ... Numidas ...
sub.sidio ... mittit; this was of course available to Augustan poetry (e.g.
G. r.293 solata), which extended it to participles with a reflexive sense like
vectus, 'travelling' (A. 1.121, G. 1.206), and even further, as here, to
participles of strictly transitive verbs; and post-Augustan prose welcomed
the new substitute for Latin's lack of a present participle passive (classical
prose through its love of precision had had to use a cumbersome sub-
ordinate clause), cf., for example, Livy 23.r.6 prae se actam praedam ...
ostentantes, 'showing off the booty being driven in front of them'. See
further Woodcock 81 f. ·
636 magnis circensibus actis: it was common tradition that the
Sabine women were seized when they were watching horse-racing. In
historical times the chief festival including this was the ludi magni or
Romani, possibly referred to here, thought to have been instituted by
Tarquinius Priscus (Livy 1.35.9), and celebrated with circus games on
168 COMMENTARY

15-18 September (and on the 19th after Caesar's death). They followed and
were intimately connected with the celebration of the dedication-day
(13 September) of the temple of Jupiter Optimus Maximus on the Capitol,
also founded by the Tarquins. This may have led to the view that the
games instituted by Romulus, and designed for the rape of the Sabine
women, must be some other: Livy (r.9.6) and Servius here name the
Consualia, whose dates (21 August and 15 December) do not coincide with
those of the ludi magni, and the reasons for whose choice are discussed by
Ogilvie on Livy 1.9.
It is best therefore to disengage Virgil's reference from any specific
festival and to read circensibus (without a capital) with Ribbeck and
Sabbadini.
638 Romulidis: coined, like A eneades, by Lucretius, in the 'epic' line
(4.683) Romulidarum arcis servator candidus anser.
638 Curibusque severis: Cures, a short distance from the east bank
of the Tiber above Rome, was the chief settlement of the Sabines; here,
as at Ovid F asti 3.201, 6.216, Cures refers to the inhabitants (strictly
called Curenses). Ancient etymologists believed that the name was derived
from curis the Sabine for 'spear', and was in tum the origin of Quirites
(so that the official title populus Romanus Quiritium reflected the synoecism
of the two settlements). Ethnically the Sabines may have been connected
with northern tribes; they were a byword for virtuous simplicity and
strictness (severi) (which was no doubt why Cato (quoted by Servius on
this line) thought they were of Spartan origin). Their king Titus Tatius is
a very obscure figure. Numa Pompilius also came from Cures, A. 6.8ro f.
639 posito for composito (as Servius) rather than deposito; see the note
on 329.
640 lovis ante aram: the altar of the temple of Jupiter Stator, 'the
Stayer' of battle, at the east end of the forum Romanum near the old
entrance-gate to the Palatine. There may have been an ancient shrine on
the site before then, which suggested the story that Romulus himself had
vowed the temple when his men were being routed by the Sabines (Livy
r.12.6; Ovid Fasti 6.793); but the historical temple (whose remains are
pictured in Nash 2.534) was vowed in 296 B.C. by M. Atilius Regulus and
erected soon after.
641 caesa iungebant foedera porca: treaties were accompanied by a
ceremony conducted by the fetiales who slaughtered a pig to symbolise
the fate of the party which broke the treaty; cf. Livy 1.24.8 (although
hardly the authentic old formula) 'si prior de/exit publico consilio dolo
malo, tum illo die luppiter populumRomanum sicferito ut ego hunc porcum
hie hodie eriam;
f tantoque magis f erito quanta magis pates pollesque.' id ubi
dixit (sc. fetialis) porcum saxo silice percussit. Virgil, unlike Livy, speci-
fically refers to a female pig. Servius here, amongst other unconvincing
explanations, suggests that a female was more efficacious in sacrifice;
Quintilian (8.3.19) that the female was more poetically elegant! Virgil is
more likely to be remembering some antiquarian formula where the
masculine form, which could be common in gender, was qualified by a
 COMMENTARY 169

feminine adjective: cf. Festus (Lindsay, p. 286) s.v. recto fronte: in com-
mentariis sacrorum pontificalium frequenter est hie ovis et haec agnus, haec
porcus, quae non ut vitia, sed ut antiquam consuetudinem testantia debemus
accipere; and cf. Ennius' lupus femina quoted on 631.
642-5. Mettus Fufetius, dictator of Alba, was pledged to support Tullos
Hostilius king of Rome in the war against Fidenae. During the battle
Mettus deserted the Romans and joined the enemy. Tullos saved the day
by shouting to his troops, loud enough for the enemy to hear, that they
were not to panic because Mettus had moved at his orders to attack the
enemy rear. The Romans were victorious and Mettus was made to pay the
penalty of treachery. Cf. Livy r.27 f.
642 citae here has its participial value (from ci(e)o), = citatae, 'sped'
as at Luer. r.997, Tibullus 1.5.3.
642 Mettum: Livy (r.27 f.) always uses the form M ettius: quem autem
dicit V ergilius M ettum M ettius F ufetius dictus est, quad nomen mutilavit
causa metri (Servius). But Mettus is an attested Latin form and at Ann.
126 V Ennius probably wrote Mettoeoque Fufetioeo (as F. Skutsch, Kleine
Schriften 259, note 3) and so provided precedent for Virgil's M ettum, rather
than Metioeo Fufetioeo (as Hermann) which would be the other way out
of the metrical dilemma (M etius). See further W. Heraeus, Kleine Schriften
249 f.
642 f. should be taken as quadrigae citae in diversa, Mettum distulerant:
cf. Livy 1.28.10 in diversum iter equi concitati. This and other similarities of
phraseology in the accounts of this episode in Virgil here and in Livy
(whose first pentad was completed by 27 B.C. (T. ]. Luce, T APhA 96
(1965), 209-40) and who is most unlikely to have revised details of phraseo-
logy through seeing or hearing the unpublished Aeneid), point in this
instance to a common source, perhaps Ennius mediated by some early
chronicler. In the extant fragments of the Annales (137 ff. V), Ennius
seems to allude to Mettus' being dragged over a plain, and subsequently
devoured by a vulture, adding in propria persona, 'Heu quam crudeli
condebat membra sepulcro '. Virgil and Livy know nothing of a vulture, but
comparable with Virgil's outburst (643) is Livy's concluding remark
(r.28.n) that this was the first and last punishment inflicted by Romans
which was regardless of the laws of humanity: in aliis gloriari licet nulli
gentium mitiores placuisse poenas.
643 dictis ... maneres, cf. A. 2.160 tu modo promissis maneas; the
ablative is difficult to classify, both instrumental and locative associations
being involved. English says 'stand by one's promises', and so sometimes
does Latin, Cic. Off. 1.32 illis promissis standum non esse.
643 Albane: examples of apostrophe (an address by the narrator in
person, as opposed to the natural vocatives in the mouths of speaking
characters) in the Aeneid almost always arise from some intenser feeling
in the context: here in an indignant and remonstrating aside, to justify
what Virgil must have regarded as an appallingly savage punishment; at
A. 6.30 f. tit quoque magnam / partem opere in tanto, sineret dolor, I care,
haberes in the pathos of a father's bereavement. But at 668 Catilina and
170 COMMENTARY

his entirely mythical punishment hardly justify the abruptness of the

address, and the metrically convenient form of the vocative (especially at
that position in the line) is an obvious factor in its choice; and at A. 7.685
Amasene pater is used chiefly to vary the formal list of a catalogue. The
very beginnings of a mannerism are detectable in Virgil: he himself had
apostrophised clover (admittedly in the Georgics, r.215); Ovid hails sheep
(Met. 13.927), dice (A.A. 3.354) and the Circus Maximus (Fasti 2.392),
and Lucan finds it as easy to speak to something as about it. See further
Austin on A. 4.27, R. D. Williams on A. 5.840; and for the special case of
address in hymns and prayers the note on 84 f. maxima Juno.
643 maneres: a past jussive subjunctive: the present maneas, 'you
should remain' transposed into past time, 'you should have remained'. Its
meaning is made quite clear by passages where it alternates as an equi-
valent with debeo and an infinitive: Livy 45.37.4 Servius quidem Galba ...
non triumphum impedire debuit ... sed . .. nomen deferret et legibus
interrogaret. In this example and in the present passage the imperfect
subjunctive has its proper and original past time-reference; but by the time
of classical prose the pluperfect subjunctive had established itself as the
tense appropriate to convey past time (the same tendency was at work also
in conditionals, where the prevalence of the pluperfect 'promoted' the
imperfect to present time-reference), cf. Cic. V err. 3. l 95 quid facere de buisti?
... rettulisses, ... solvisses; jrumentum .. . ne emisses (note that the
negative is ne: a clear indication of the jussive nature of the subjunctive).
Virgil himself prefers the pluperfect: A. 4.604 ff., 678 f., ro.854; here of
course mansisses, a molossus, could hardly have been fitted into a hexa-
meter without being pointlessly cumbersome.
See further Woodcock 86 f.; Palmer, LL 310.
644 f. Note the preponderance of v, r and s, creating a feeling of un-
pleasant harshness which suits well with the sense.
646-51 After Tarquinius Superbus had been expelled from Rome by
Brutus and Collatinus, because the latter's wife had been raped by his son
Arruns, he fled to Lars Porsena of Clusium. With united Etruscan forces
Porsena captured the Janiculum and tried to force a passage across the
river by the pons Sublicius. Horatius Codes, Titus Herminius and Spurius
Lartius defended the bridge on the Etruscan side while the Romans
chopped it down behind them; when it was almost down Horatius sent
back Herminius and Lartius, and after it had fallen leapt into the Tiber
and gained the opposite shore. Porsena pressed the siege but eventually,
through the intervention of Mucius Scaevola, peace was made (a stage of
the story omitted by Virgil). The Romans gave ten youths and ten girls
as hostages for the peace; one of these was Cloelia, who escaped and swam
across the Tiber-only to be sent back by the Romans. Porsena, amazed
at both feminine courage and Roman good faith, allowed her to return
home with the other girls. Cf. Livy r.57-9, 2.9-15. Later authorities,
Tacitus Hist. 3.72.1 and Pliny N.H. 34.(39).139, tell the unpatriotic truth:
Porsena did actually capture Rome. He did riot however restore Tarquinius,
and he may not have intended to.
 COMMENTARY 171

646 f. 'And Porsena was commanding the Romans to welcome the

expelled Tarquinius': Porsenna iubebat (Romanos) accipere Tarquinium.
The idiom which omits the object of iubeo when it can be easily supplied
from the context (urbem) (see the note on 498) must be carefully noted
when the infinitive itself has an object as here.
646 Porsenna: metre guarantees the spelling Porsena at Horace
Epodes 16.4; Silius Italicus 8.389 and 478; 10.483 and 501; Martial 1.21.6;
Porsina occurs as a Roman name at CIL 6.4 2 .32919. Servius says that
Virgil added another n metri causa, and he may have had no other justifi-
cation; cf. the 'mutilation' of Mettius, above 642. But the variation may
be due to an original Etruscan sound which was ambiguous to Latin-
speakers.
iubebat: note the end-rhyme with premebat (647) (and ruebant 648):
'doubtless intentional' according to Conington, but see the note on 620 f.
648 Aeneadae is first attested in Lucretius, the first word of the De
Rerum Natura; for coining such patronymics there was the precedent of
Lucilius (see Munro's note on Memmiadae, Luer. 1.26).
649 illum = Porsennam: Servius unaccountably takes it = Coclitem.
indignanti similem: for the turn of phrase cf. A. 5.254 anhelanti
similis, which also emphasises the animation of a picture (Ganymede
hunting, embroidered on a tunic).
The two-syllable rhyme in -anti before the strong caesura in the third
foot and at the end of the line (for other examples see Austin on A. 4.260)
makes it a true leonine-a term whose origin probably derives from the
first animal, Leo, in the widely-read medieval poem Physiologus sive De
Bestiis by Theobaldus, where this type of rhyme is common, e.g. in the
opening line Tres leo naturas et tres habet inde figuras.
650 aspiceres like videres (676) and credas (691) is the indefinite second
person singular of the potential subjunctive. Latin and English idiom are
exactly alike in this use of the indefinite second person singular = 'you'
without reference to any specific individual, where French uses 'on' and
German 'man'.
650 f. auderet quia . .. / ... innaret: a causal clause with a subjunctive
of virtual oratio obliqua. The subjunctives here do not cast doubt on the
historic truth of the actions described, but stress that it was the thought of
these actions that caused Porsena's indignant threats. See the note on
129 f.
For the postponed quia see the note on 22 ubi.
651 fluvium . .. innaret: (in)nare, 'to swim (in)' is most naturally
construed with an ablative of place as at line 691 pelago . .. innare. The
accusative here with innare (as also at A. 6.369 ftumina tanta paras
Stygiamque innare paludem; G. 2.451, 3.142) is hardly to be explained as
due simply to the force of the preverb in-, 'into', because the idea of
motion into the water is clearly not uppermost, if present at all; it has
much closer affinities with the accusative after the simple verb nare (as at
G. 3.260 nocte natat caeca serus freta), which is an accusative of space
traversed, developing out of an internal accusative (ire mille passus
13 EAC
172 COMMENTARY

(accusative of extent) is a development, by substitution, of ire viam

(internal accusative)).fluvium .. . innaret here is an attempt to produce an
unusual variation on the standard expression of classical prose; the
general tendency of the spoken language also was to standardise verbs as
transitive and complements as accusative.
651 vinclis ... ruptis: of the many accounts of Cloelia and her escape
(see Ogilvie, Livy 1-5, 267), Virgil's alone makes any reference to vincla.
Virgil's phrase is a poetic equivalent of Livy's frustrata custodes (2.13.6).
652 f., 655 ff. The Gallic attack on the Capitol.
After the disastrous defeat of the Romans by the Galli Senones led by
Brennus on the banks of the Allia (18 July 390 B.C., a never-forgotten
'black day' of the Roman calendar), most of the citizens evacuated to
Caere, but a small garrison was left on the Capitol under the command of
M. Manlius. The Gauls certainly penetrated the city and may actually
have gained possession of the Capitol after scaling it, according to tradi-
tion, by its steepest side, the Arx Tarpeia; but the widespread story,
accepted by Virgil, that Manlius was warned by the cackling of geese,
implies that it was not (tenebant 657 is conative). Manlius earned the
cognomen Capitolinus for successfully repulsing the invaders.
Servius (on 655) knew of a silver image of a goose on the Capitol;
Virgil's single anser is representative rather than collective.
652 in summo: at the top of the shield)( in media 675.
652 Tarpeiae . .. arcis: the Capitolium proper (see the note on 306 ff.).
Livy (1.11.5-9) and Servius on A. 8.347 tell the story of how Romulus,
waging war against the Sabines (the certamen of 639), gave the defence of
the Capitol to Tarpeius, whose daughter Tarpeia fell into enemy hands
while on her way to fetch water. The Sabines asked her to betray the
citadel, and she agreed, asking as her reward what they carried on their
left arms, meaning their gold bracelets. But when, through her treachery,
they had reached the citadel, they granted her request by taking from
their left arms their shields, throwing them on her and crushing her to
death. (Propertius 4-4 preferred the romantic alternative version of
Tarpeia's motive-love for the Sabine general).
The story is another aetiological myth to explain the name T arpei-
which, like Tarquiti- and Tarquini-, was actually of Etruscan origin
(W s.v.) and no doubt dated from the era of intensive Etruscan domination
of the arx Capitolii and the building of the temple of luppiter Capitolinus
by one or other of the Tarquins (if there were in fact two). Criminals
convicted of particularly heinous offences, e.g. perjury (Aulus Gellius
N.A. 20.1.53), were, in early times, hurled e saxo Tarpeio.
654 The mention of Romulus' 'palatial' hut, the casa Romuli, is an
awkward interruption of the account of Manlius and the Gauls (652-62),
and might more appropriately have been connected with the section on
Romulus (630-41): this is the justification for transposing it (with Ribbeck,
Mackail and others) to follow 641, although the position of the line in the
final but unrevised draft of the Aeneid is probably faithfully reproduced
(unanimously) by the MSS.
 COMMENTARY 1 73

The casa was traditionally supposed to antedate the invasion of the

Gauls by three centuries, but recens does not contradict this, since it need
not and probably (see below) does not refer to the pristine condition of the
casa. Nor does the fact that some authorities (Dionysius of Halicarnassus
r.79.n; Plutarch Romulus 20.4) located a casa Romuli on the south-west
corner of the Palatine prove that there was not another on the Capitoline
(as Vitruvius 2.1.5; Seneca Contr. 2.1.4). Both, presumably, were thatched
huts kept in trim (recens) through religious veneration and restored in the
same style whenever damaged (see Platner-Ashby s.v.). For an illustration
of a straw-thatched hut of the' pre-urban', i.e. pre-Etruscan, epoch on the
Palatine-the casa Romuli was no doubt a maintained survival of one
such-see Gjerstad 3, 52 f.; Nash 2, 164 f.; Scullard, Plates I and 3. These
huts, whose general shape can be reconstructed from post-holes in the
stone flooring and from the imitative shape of the hut-urns used to
contain the ashes of the cremated dead, belonged to the Iron Age culture
of Latium, essentially the same as the widespread Villanovan type
(exemplified at Villanova near Bologna); they are datable to the eighth
century B.C., in the middle of which traditional chronology placed the
founding of Rome by Romulus.
654 Romuleo . .. recens: double hypallage, in the interests of word-
symmetry, for Romulea ... recenti (see the note on 542).
655 auratis: only the sense of the context decides whether it has its
strict sense, 'gilded', or is merely = aureus whose oblique cases are only
available to hexameter poetry by synizesis. The sense here will allow either,
for although auratis and argenteus primarily refer to the metals used by
Vulcan, the tiles of the actual Capitoline temple were 'gilded' at a later
date (see the note on 348 ff.), and the statue of the goose on the Capitol
was of silver, according to Servius.
656 canebat: not a euphemism for 'cackling', but = 'foretold', like
a prophet or oracle.
657 tenebant, 'were reaching': awkward so soon after tenebat 653 =
'was occupying', but not so unVirgilian as to justify Wakefield's tegebat
in 653. The difference of meaning reflects the difference between the
conative and descriptive functions of the imperfect.
659 vestis: hoc est barba, Servius, who may have thought that vestis =
'clothing' was tautologous with sagulis in the next line, and that the
appropriate contrast to caesaries could be extracted from vestis in the sense
which it has in the Late Latin compounds vesticeps 'bearded' and investis
'beardless'. But the context here cannot impose any such metaphor on the
word as it very clearly does at Luer. 5.673 f. (imperat aetas) ... impubem
molli pubescere veste / et pariter mollem malis demittere barbam (for the
metaphorical use of vestire in the same sense cf. A. 8.160). virgatis lucent
sagulis is an elaboration, almost in parenthesis, of vestis, and was so
understood by Silius Italicus 4.155 auro virgatae vestes (of the Celts).
For gold as the traditional material of Gallic chieftains' armour, cf. the
account by the annalist Claudius Quadrigarius of the Gaul armis auro
praefulgentibus defeated by Valerius Corvinus (Aulus Gellius N.A. 9.11.5).
1 74 COMMENTARY

660 virgatis: first apparently used here in the sense 'striped', from
virga 'a stripe of a garment' (purpureae, Ovid A .A. 3.269).
660 sagulis: there are only three cases of true diminutives in the
Aeneid (i.e. of words used in the diminutive form where contemporary
idiom would use the root word with no significant change in meaning).
The particular fitness of parvulus to its context of tenderness and pathos at
A. 4.328 is well discussed in Austin's note; palmula at A. 5.163 (also in
direct speech) occurs in a setting of bluff familiarity; sagulis here is more
difficult to account for (the tiny cloaks of the tiny figures on the shield?).
Diminutives are a frequent feature of 'affective' colloquial speech:
Horace had no reason to exclude them from his hexameter poems, and
Lucretius did not; they became a darling mannerism of the neoterics,
Catullus (see Fordyce on Catullus 3.18) and the Virgil of the Eclogues. But
the mature Virgil banned them from epic, and his word was law: they are
not discoverable in Lucan or Valerius Flaccus, and Silius I talicus shows his
complete enslavement to Virgil by overworking two of the three to which
Virgil had given his blessing (sagulum four times, parvulus twice), and
using no others. See Axelson 38 ff., A. S. F. Gow, CQ 26 (1932), 130 ff.
660 f. tum lactea colla / auro innectuntur: i.e. with a torques, a
twisted golden circlet; the gens Manlia acquired the cognomen Torquatus
when T. Manlius (A. 6.825) took a torques from a gigantic Gaul he had killed
in single combat.
661 coruscant: coruscare (etymologically connected with Greek Kopva-
a£Lv = 'to butt with the horns') comes to denote, transitively or intransi-
tively, any light darting movement; here, 'they brandish'.
quisque strictly requires a singular verb, and coruscat is found in part
of the MS tradition, but this may well be a correction ultimately due to a
regularising scribe, for the idiom whereby quisque stands parenthetically
in apposition to a plural subject, whether actually expressed or not, is
very common: see KS l, 22 ff. and 249 ff. Cf. A. 5.500 f. incurvant . .. / pro
se quisque viri; A. 12.457 f. densi cuneis se quisque coactis / adglomerant;
here the plural subject is made explicit by protecti 662.
662 gaesa: gaesum means a long heavy Gallic javelin: it had been
used in connection with the Gauls by Caesar E.G. 3.4.1 (most Celtic loan-
words in Latin refer to horse-drawn vehicles, cisium, essedum, petorritum,
etc.). Virgil adopted it for epic as a glossa, a rare foreign word, of the kind
which Aristotle (Poetics 1458a-1459a) had prescribed as particularly
suitable for it (a rule based on the curious assumption that Homer wrote
for posterity, and that the words which were bizarre or unintelligible to a
contemporary of Aristotle must have been so to Homer himself and
deliberately chosen for that reason). Just as Virgil had given epic cachet to
the Punic words mapalia (G. 3.340) and magalia (A. 1.420, 4.259) in
appropriate connection with Libyan shepherds and the suburbs of Carthage
respectively, so here gaesum is used in a description of Gauls (and, in a
passage closely modelled on Virgil, in a description of the Gallic chieftain
Viridomarus, by Propertius 4.10.41 ff ... . genus hie Rheno iactabat ab ipso,/
mobilis e rectis jundere gaesa rotis. / illi ut virgatis iaculans it ab agmine
 COMMENTARY 1 75

bracis / torquis ab incisa decidit unca gula). It is later repeated without any
such fittingness by Silius (2.444) in connection with an African, and by
Statius in connection with Greeks (Th. 4.64) and Macedonians (Ach. 2.132).
On glossae see Palmer, LL 97, 99, and note that Ennius had used native
words for spears more than once: Ann. 390 V rumpia (the long Thracian
spear), Ann. 504 V sybinis (the Illyrian hunting-spear).
662 scutis protecti corpora longis: 'protecting their bodies with
long shields'; for the construction see the note on 29 titrbatus pectora.
663 f. The Salii were an ancient ritual brotherhood of Dancing Priests
(Salii from satire), who performed a vestigial war-dance (tripudium: which
suggests that it had a three-beat rhythm and was quite different from a
march). Dancing, beating their shields with staves or daggers, and singing
the carmen Saliare, they went in procession through the city; originally
their function seems to have included the purifying of the arms of the host
before (March) and after (October) the campaigning season. At Rome
they were organised into two colleges (see the note on 287) each of twelve
members. Their extreme antiquity is attested by the social and religious
taboos connected with them: they had to be of patrician birth and have
both father and mother living; their armour was of bronze (a Bronze Age
relic and so antedating their alleged institution by Numa); they carried
figure-of-eight shields, ancilia, said to be copies of the ancile which dropped
from heaven as a sign from Jupiter to Numa, who, warned that supreme
power was conferred by it, had replicas made to baffle theft. They also
wore the trabea, the blood-red war-dress. On their heads they wore bronze
helmets, according to Plutarch Numa 13.4, recalling their origin as Bronze
Age infantrymen. But according to Dionysius of Halicarnassus 2.70.2 they
wore apices, conical felt hats with a pointed piece of olive-wood at the top
and a lock of wool attached to it. These were traditionally worn by the
/!amines, whose primary and primitive connection with sacrifice is un-
questioned (the lock of wool, and perhaps the felt of the hat as well, will
have come from sacrificed animals). The Salii had no such primary func-
tion, and if Dionysius is right, an incongruous innovation must have taken
place. But there is no good reason (the following mention of ancilia is not
one) for believing that Virgil is alluding to anything but the headgear of
the jfamines, the flamen Dialis, whom Augustus restored to prominence,
in particular, but also perhaps the flamen of Mars, whose cult as Mars
Ultor Augustus was also to promote. A denarius of 17 B.C. (BMC Emp. I, 14,
No. 74) shows the head of Augustus on the obverse and two ancilia with
an apex on the reverse: the coupling of two emblems, which had become
as symbolic of the identity of the state as Vesta's fire itself, may have been
inspired by Virgil's example.
663 The Luperci were patrician youths who, organised in two groups
Luperci Fabiani and Quinctiales, celebrated a rite of extreme antiquity,
attached in historical times to the festival of Faunus (Pan) on 15 February.
Originally they ran in a circuit around the Palatine (Latte, RRG 84, n. 4),
naked except for a loin-cloth, striking anyone they met with a strip of goat-
skin (according to Servius on A. 8.343 this was anciently called februitm
r76 COMMENTARY

(hence February), apparently a generic term for any magical object which
could ward off unwholesome influences (so Warde Fowler, RE 210)). The
original purpose(s) of the ceremony can only be guessed at, and were probably
unknown to Augustus when he restored it; his contemporary Ovid (Fasti
2.425 ff.) believed that it was intended to promote the fertility of women
(the striking of men too could well have had the same object). Modem
theories regard it as a purifying and apotropaic ceremony, directed against
wolves and other menaces to the early pastoral community; or against
werewolves reincarnating dead spirits. See further Ogilvie on Livy
r.5.1-2.
663 ff. The main reason for the inclusion of the Salii, Luperci and
matronae at this point is explained in the note on 630 ff. Virgil's associated
thought may have been concerned with the connection between the
traditional founding of the rites and their contemporary revival: he
associates Salii at the Ara Maxima with Evander (285 ff.), and Livy
(r.5.1-2) ascribes the founding of the sacrum Lupercale to Evander; the
latter was restored, together with other antique half-forgotten ceremonies,
by Augustus (cf. Suet. Aug. 3r.4: nonnulla etiam ex antiquis caerimoniis
paulatim abolita restituit, ut Salutis augurium, Diale flaminium, sacrum
Lupercale, ludos saeculares et compitalicios).
665 extuderat: 'had beaten out in relief'; the force of ex- can be
pressed here, although usually the verb simply = 'to fashion' (G. 1.133,
4. 315).
666 matronae had long had the right to ride in two-wheeled carriages
(pilenta, carpenta) on public occasions, so long that different authorities
assigned its origin to different events: see the note on 630 ff. As Binder
points out (202), the castae matres glance at the crystallising policy of
Augustus which issued in the social legislation of 18 B.C., the Lex Julia de
maritandis ordinibus, and the Lex Julia de adulteriis.
mollibus = molliter stratis (Servius auctus).

666-674 The next scene is the Underworld, with Catiline being punished
among the damned, and Cato giving laws to the righteous. Separating these
and the preceding scenes from what follows is a broad band representing the
sea, and dolphins swimming in it.
667 Here and at G. 4.467 Taenarias etiam fauces, alta ostia Ditis, ostia
'gateway' is in apposition: more fittingly with f auces 'entrance-gorge'
than with sedes 'habitation'.
668 f. Catiline's punishment was no doubt inspired chiefly by a re-
collection of A. 6.601-7, where Virgil ascribes to Ixion and Pirithous the
fate earlier associated with Tantalus, the double torture of an overhanging
rock, likely at any moment to fall and crush them, and of being' tantalised'
by an aggressive Fury who keeps them away from a regal banquet. This
and other traditional punishments of mythological malefactors are allo-
cated in Virgil's Hell to human sinners, including those qui . .. arma
secuti / impia (612 f.), like Catiline. There (614 f.) Virgil refused to be
 COMMENTARY 177

pressed about the details of individual punishments, and here in A. 8 is

merely outlining the situation which, given in full, would have been a
tedious repetition of A. 6, and unbalanced the terse catalogue-entries of
the pictures. Even so, pendentem is difficult to visualise as part of the
Tantalean torture: Heyne thought it meant 'hanging fixed', like Prome-
theus, with the Furies instead of the vulture to torment him; Conington
took it to mean 'just in the agony of falling' over a precipitous cliff
(referring to the Tarpeian rock), over which the Furies were hounding him
(Catiline would then have the punishment he might have expected in
life-but the Tarpeian rock is not a landmark of any Underworld).
Virgil may have remembered Lucretius 3.980 f. nee miser impendens
magnum timet aere saxum / Tantalus, and transferred the notion of 'hanging'
from the rock to the victim.
Catiline and Cato are to be regarded as historical personages as much as
those before and after them on the Shield, not (pace Conington) as repre-
sentative types. But the unsuitability of connecting them like the others
with any particular historical event led Virgil to picture them differently
and represent the judgment of history on their whole careers in terms of
the rewards and punishments of the after-life.
The unrest of Cisalpine Gaul (Sallust Cat. 42) caused by Catiline's
sedition was perhaps the first political event to impress itself on the
boyhood of Virgil (about seven at the time). It was on Cato's motion that
the Senate decreed the death-penalty for Catiline and his associates
(Sallust Cat. 52 f.).
For the apostrophe, Catilina (voe.), see the note on 643 Albane.
669 Note the two participles 'framing' the line, and see the Appendix,
p. 200.
670 secretosque pios: 'and the righteous in a place apart'; secretos =
'separated, sequestered' (see the note on secretum 6ro).
dantem iura: a Roman legal phrase: iura (iudicia) dare (reddere),
used of a presiding magistrate (a praetor acting as judge) means 'to give
a ruling' to a plaintiff: hence the broader sense, 'to legislate (by estab-
lishing a precedent)', as at, for example, A. r.293. In Silver Latin dare
came to be used absolutely = decernere, cf., for example, Pliny Epp.
7 .6.9 finem cognitioni quaestio imposuit, quae secundum reos dedit.
Cato's own praetorship in 54 B.C. had been energetic and stringent in
the pursuit of justice, particularly against miscreant senators; but his
incorruptible rectitude in all directions made him suitable to judge the
dead. Roman judicial language is similarly applied to Minos at A.
6.431 ff.
On this line Servius insists that the Cato referred to is Cato the Censor,
not Cato the Younger qui contra Caesarem bella suscepit: he is certainly
wrong, but may be forgiven for not understanding the dilemma which
Cato and Caesar posed for the Augustan poets. Augustus was an autocrat,
but he had taken elaborate care to conceal the fact behind a fac;ade of
constitutional republicanism. Cato, who had committed suicide in 46 B.c.
rather than outlive the republican cause, had become the hero of unshaken
 COMMENTARY

faith in constitutional government; he therefore qualified for a niche in the

Augustan fa-;ade, and hence the encomium here and at Horace Odes
1.12.35 f. and 2.r.23 f.
Julius Caesar, conspicuously absent from the Shield, presented an
even more difficult problem: the undisguised autocrat could not be
reconciled with Augustus' public professions and deserved public con-
demnation; but Augustus, so far from condemning him (which would
have been an act of gross personal impietas) had supported his deification.
Silence was the usual answer to such a paradox.
Peerlkamp objected that Cato the suicide had no place among the pii
and should have been put with the other suicides in the Lugentes Campi
(A. 6.434 ff.). Thiel tried to rescue him for Elysium by classing him among
the ob patriam pugnando vulnera passi (A. 6.660). But it is evident from
the description of Catiline (above) that Virgil is not now working within
the narrow prescriptions of A. 6; nor does he mean Cato to supersede the
judges of the Underworld in any way.
671 haec inter: for the position of inter see the note on 32. But what
does haec refer to? 'In Virgil's shield the attention of the Roman reader
was to be absorbed by successive pictures of Roman history, and that
reader would be the last man to inquire how they fitted in within the
shield's rim; nor need we attempt that unnecessary task': so Warde
Fowler (ASR ror) (and before him to the same effect Heyne (Excursus 4
on A. 8), and Heinze (VeT 394)). This is entirely right-as regards
particulars; but Virgil gives enough hints to make the broad plan
intelligible. From the She-Wolf (630) to Cato (670), two pictures are said
to be near the preceding ones (635, 642), the rest can be assumed to be,
except for the last (666) of Hades, which naturally has to be separated from
pious matrons; of these Manlius holding the Capitol is at the top of the
shield (652 in summo). in medio (675), as the centrepiece, is the victory and
rout of Actium and Augustus' triumph in Rome. The sea and the dolphins
here (671-4) are the only part which is pure ornament, and the only items
which are derivative: on the Shield of Achilles (Il. 18.607 f.) the stream of
Ocean runs round the outer rim, and there are dolphins on Hesiod's
Shield of Heracles (209 ff.). It is plausible to suppose that Virgil too
thought of his sea as a broad (late 671) circular band, on and around which
a repeated motif of dolphins was swimming (in orbem 673, not gyrating in
individual circles), not however on the rim of the shield but parallel to it,
separating all the pictures mentioned before it from the centrepiece
mentioned after it, in between the former (haec inter) and the rest.
672 spumabat (primary MSS) caerula (sc. imago), or spumabant
(secondary MSS) caerula (id est maria, Servius); the second is preferred by
some editors, but its superiority is not evident. Virgil has overloaded the
line with colour: the emphasis in caerula, whether noun or adjective, is on
dark-blueness. Keeping spumabat, translate: 'Separating these there ran
in a broad band the likeness of the swelling sea in gold, but dark-blue
where it foamed with white waves.' The fabric of the shield is indescribable
(625), but a metallurgist might produce this effect with some combination
 COMMENTARY 1 79

of the metals we are told were used: gold (passim), silver (673), bronze
(675), and iron (701).
673 ar~ento clari delphines: 'shining silver dolphins', argento being
used with adjectival force (as Mackail observes, Appendix A, p. 513), and
argento clari together recalls a Greek compound adjective. As at A. 2.765
crateresque auro solidi the ablative is partly one of material (with the noun
delphines), partly one of means (or instrument) (with the adjective clari).
The adjective certainly lends support to the phrase; a simple ablative of
material would be extremely bold (for prose ex or de with the ablative).
674 aestumque secabant: 'and were cutting through the surge',
'breasting' it.

675-713 The sea off Actium; Augustus with Agrippa confronts Antony
with Cleopatra. The massive ships join battle, the gods of the East fight
against those of Rome, and in the middle of the conflict rage the gods of war.
Apollo routs the hordes of the East, and Cleopatra, with the hand of death
already on her, flees to the Nile.
675-728 There are evident similarities between some of the features of
this passage and the proem to Georgics 3 (10-39) which ends with the promise
(46 ff.) mox tamen ardentis accingar dicere pugnas / Caesaris. This has led
some scholars (Nettleship and Warde Fowler) to believe that the connected
section of the Shield was written about the same time as the Georgics
passage and in fact formed part of the projected poem on Augustus'
victories, from which it was transferred to its present place in A. 8.
675 ff. This passage left a deep impression on Propertius, who had read
it before its publication: me iuvet hesternis positum languere corollis, / quem
tetigit iactu certus ad ossa deus; / Actia Vergilium custodis litora Phoebi, /
Caesaris et f ortis dicere posse ratis / ... cedite Romani scriptores, cedite Grai ! /
nescio quid maius nascitur Iliade (2.34B.59 ff.).
675 in medio: utrum clipeo an mari? (Servius): the former. The
centre of the shield contains all the scenes that follow: there are four
separate ones, (1) the actual conflict of the opposing fleets (675-703),
(2) the rout and retreat of Cleopatra and her forces (704-13), (3) Augustus'
triumphal return to Rome (714-9), (4) his review of captives and conquests
(720-8). They are narrated in chronological order; how they were grouped
pictorially we are not told. The portrayal of successive events in the life
of one person in adjacent panels (as with (3) and (4) above) is common in
medieval art, inherited from Roman practice.
aeratas: bronze is the material of the pictured ships; the actual ones at
Actium could have had bronze rams and prows only.
676 cernere erat: 'it was possible to see'. est = 'it is possible' acquired
this meaning (which was alien to it in Latin) under the influence of the
quite normal Greek idiom (;g}E0'7"i; and the Greek infinitive could also
quite easily be taken over because Latin verbs of 'being able' were regu-
larly followed by an infinitive. Lucretius had previously used this con-
struction, and so does Virgil elsewhere, A. 6.595 f. nee non et Tityon . .. /
180 COMMENTARY

cernere erat; G. 4.447 scis, Proteu, scis ipse; neque est te fallere quicquam;
Augustan and later poets continued its use. It is not found in classical
prose, and only rarely in post-classical prose. See further LHSz 2, 349.
677 fervere . .. effulgere: Virgil uses only the third conjugation
(metrically more suitable) forms of the infinitives fervere (here and in four
other places), effervere (G. r.471, 4.556 only), fulgere (A. 6.826 only),
effulgere (here only), but prefers the second conjugation forms of the
indicative: fervet, fulgent (as at 684 below) and effulgent (A. 5.133). The
third conjugation forms were the original ones, and Virgil derived them
from earlier poetry (Lucretius has all the above infinitives except effulgere,
and this compound is first attested in Virgil); they were losing ground to
the second conjugation forms as early as Lucilius who protests against the
trend (356 f. f ervere, ne longum. vero: hoc lectoribus tradam. / fervit aqua et
fervet (future).fervitnunc.fervetadannum); Quintilian (r.6.7) brands themas
archaic and incorrect: si quis antiquos secutus fervere brevi media syllaba dicat,
deprendatur vitiose loqui. stridere had a similar history but Virgil seems to
have kept to third conjugation forms (e.g. stridunt A. 8.420; there is
however a variant stridet at G. 4.262). See LHSz I, 322 and N-W 33, 267 ff.
fervere Leucaten here may be a conscious reminiscence of Luer. 2.43a
(Bailey) fervere cum 'liideas classem; if so, there is a significant change of
context: Virgil is glorifying the activity which Lucretius despised as no
help to peace of mind.
677 Leucate, the modern Leukas, is an island to the south-east of the
promontory of Actium, and this, with the promontory opposite, encloses
the Ambracian Gulf to a narrow channel.
678 hinc and hinc (685) signal the protagonists, their aides are secon-
dary (parte alia 682; sequitur 688).
678 Augustus agens Italos: Norden notes the view that at A. 6.792,
Horace Odes 2.9.19 and in this line Augustus is still felt as an appellative,
i.e. an adjective rather than a proper name. Warde Fowler, ASR III,
suggests that it should therefore be written with a small initial letter; the
phrase is then exactly parallel to arduus agmen agens in 683 (for the idiom
they represent see the note on 299); moreover, Virgil would have known
that Octavian did not take the official title Augustus until 16 January
27 B.C., more than three years after the Battle of Actium, the dramatic
date of this passage. All this is quite possible: Virgil will be using the
technique of foreshadowing, in the narrative, the event which had already
taken place, historically, as he did extensively in A. 6. But Warde Fowler's
view that this line actually suggested to Octavian the adoption of the title
Augustus is improbable: the delicate relations between Augustus' court
poets and himself and his privy councillors hardly allowed anything so
direct. The etymological connection of the name with augur and augere
(salutem populi Romani/ imperium) is emphasised by the Augustan poets.
hinc . .. ltalos )( hinc ope barbarica 685: the contrast is between the
amorphous Orient and a united Italy, not Rome alone. In 32 B.c. Octavian
had held a national plebiscite: iuravit in mea verba tota Italia sponte sua,
et me belli quo vici ad Actium ducem depoposcit (R.G. 25). This was the first
 COMMENTARY 181

of many acts which encouraged a sense of national unity by stressing the

political power of the whole peninsula, always dominated in the past by
the capital city. More than one blue-blooded chauvinist must have had
an unpleasant shock at hearing the army of the Senate and People of
Rome described as 'Italians'. But Virgil's sympathy could hardly have
been more sincere to any other aspect of Augustus' policies: he was him-
self a provincial, and Cisalpine Gaul had only received the franchise (from
Julius Caesar) when Virgil was already twenty-one, eighteen years before
the battle of Actium. The theme which was perhaps nearest the poet's
heart, merely hinted at here, stated at 714 f. Romana . .. moenia )( dis
Italis, finds its finest expression at A. 12.827 sit Romana potens Itala
virtute propago, 'at once a prayer, a decree, and a benediction' (Mackail,
Intro. 14). In the Actium elegy of Propertius (4.6) the idea is completely
absent (cf. esp. 57, vincit Roma fide Phoebi: datfemina poenas).
agens . .. agens (683): the pictures of Augustus and Agrippa, like those
of the ships in Masefield's Cargoes, contain no finite verbs.
679 A line of forceful dignity: note the alliteration (especially of p and
-bus), the unusual rhythm with no strong caesura in either of the middle
feet, and the Ennian ending (cf. Ann. 201 V Dono, ducite, doque volentibus
cum magnis dis) which brings with it a fifth-foot spondee and a final mono-
syllable. Virgil had already used the second half of the line in A. 3.II f.
feror exsul in altum / cum sociis natoque penatibus et magnis dis, spoken by
Aeneas recounting his leaving of Troy. In both passages there is a situation
of the utmost solemnity as the religious symbols (and so the very identity)
of the state are exposed to grave danger. In this passage however Virgil is
at pains to stress the legality and justice of Octavian's position: he is the
leader of senate, people, gods and fatherland, and his victory will not be
the disgraceful victory of a civil war, but the triumph of the Italian
nation over a foreign enemy (hence the description of Antony's forces,
685 ff.). By the coniuratio totius Italiae Octavian had secured a kind of
technical legality for his proceedings: he claimed (R.G. 25.3) to have had
over 700 senators under his standards in 31 B.C. (when the total senatorial
roll was about 1,000); how many personally took part in the battle at
Actium is unknown.
679 penatibus et magnis dis: are the penates the same as the magni
di or not? The learned Varro (cited by Servius on A. 3.12) knew that the
statue of the Penates publici populi Romani (in the guise of the Dioscuri)
in their temple on the Velia (restored by Augustus, R.G. 19) had on its
base the inscription MAGNIS DIIS. (R. B. Lloyd, A]Ph 77 (1956), 38 ff.,
suggests that this is the source of Virgil's phrase.) Originally the penates
were characterless powers who guarded the penus, larder or storehouse.
The requirements of plastic art, and Greek influence, caused them later to
be identified with established deities, chiefly apparently with minor gods
or demi-gods (heroes), cf. Hercules as lar, 542 ff.: this suggests that they
are contrasted with the' great gods' of the Roman state, who also fight for
Rome at Actium (699). But Roman thought was itself confused on the
subject, some antiquarians holding that the Penates must be Apollo and
182 COMMENTARY

Neptune, the gods who built Troy, A. 5.8n (see Macrobius 3.4.6). Bailey
(RV 91 f.) believes that the two phrases are synonymous. The gods were
literally present because their images were set up in the stern. Ahas carries
such an image of Apollo on his ship at A. ro.171, and Augustus on his at
Actium, as apparently implied by Propertius 4.6.29. There could also have
been images of Neptune, and of the Penates in whatever likeness.
680 f. geminas cui tempora flammas / laeta vomunt patriumque
aperitur vertice sidus: Virgil's language is expertly chosen for inten-
tional ambiguity. (1) On a purely literal level the lines balance the quite
material description of Agrippa's headgear in 683 f. fiammas vomunt
probably refers to the shape and colour of the plumes: see the note on
620 f., and cf. the very similar phraseology of A. ro.270 f. ardet apex capiti
cristisque a vertice fiamma / funditur et vastos umbo vomit aureus ignis,
where Aeneas is specifically compared with a comet-the same complex o_f
associations as here. (The reference may be to the gleaming metal cheek-
o,
pieces of the helmet if Virgil is remembering fl. 5.4 8a'U iK 1<6pv86, 7'€ Ka,
amri8o, a1<aµ,aTOV 1rvp). According to Servius here, Augustus had a star shown
on his helmet to symbolise the famous comet, Iulium sidus, which appeared
when he was celebrating funeral games to honour Caesar, his adoptive
father (patrium). (2) This comet was believed to be Caesar's soul being
received into the company of the immortal gods, the last and supreme
'transformation' of Ovid's Metamorphoses. This belief derived from the
idea that a man's genius, his essential personality, could manifest itself as
fire around the head (see Onians, The Origins of European Thought 2 163 f.):
notable examples of this actual phenomenon portending future greatness
are the infant Servius Tullius (Livy 1.39.1),Ascanius (A. 2.682 ff.), and
Lavinia (A. 7.73 ff.). So here, Augustus' genius is triumphantly aflame
and sure to conquer Antony's: 'under him/ My genius is rebuk'd, as it is
said/ Mark Antony's was by Caesar' (Shakespeare, Macbeth, Act 3, sc. 1).
On the Caesaris astrum see further Weinstock 370 ff.
tempora therefore refers both to Augustus' own temples and to the
plume-sockets of his helmet. Henry was the first to see the deeper meaning
of the passage, but wrong to see that only.
The twin plumes of Augustus, Rome's second founder, seem deliberately
to recall those of Romulus at A. 6.779.
681 aperitur: 'dawns'. aperire = operta detegere = 'to disclose'; cf.
Livy 22.6.9 cum ... dispulsa nebula aperuisset diem, and (of the planets)
Cic. Nat. Deorum 2.51 tum occitltantur (through proximity to the sun),
tum rursus aperiuntur (' become visible ').
682 An unusual amount of rhyme, -(d)is. There is an allusion to the
storms which hampered Antony's fleet at the beginning of the battle;
Agrippa's tactics were successful ventis . .. secundis.
683 arduus agmen agens: 'leading his column from a towering
position'. For the Virgilian idiom which qualifies a participle with an
adjective where English (and Latin too) would normally have an adverb
or an adverbial phrase, see the note on 299 arduus arma tenens. In both of
these places Page wished to punctuate with a comma after arduus, but
 COMMENTARY

comparable instances of adjective and participle in close juxtaposition

suggest that they are very intimately connected (at, for example, A. 3.69 f.
placataque venti / dant maria et lenis crepitans vocat Auster in altum, it is
impossible to separate them); cf. also A. 2.328 f. arduus armatos mediis in
moenibus astans / fundit equus.
Note the jigura etymologica in agmen agens (noun and verb both formed
from the same stem ag-), the kind of word-play which had deep roots in
Latin's love of alliteration.
683 belli insigne superbum: insigne is perhaps better taken in
apposition with the whole of the following sentence than simply with
tempora . . rostrata: see the note on 487 tormenti genus.
684 navali ... corona: according to Aulus Gellius N.A. 5.6.18 f. the
'naval crown', made of gold, was awarded to the armed man who first
leapt on to an enemy ship in a naval battle. But the only known recipient
of the honour before Agrippa was the antiquarian M. Varro, who was
awarded it after the war against the pirates by Pompey the Great (Pliny
N.H. 7.(30).n5 and 16.(3).7 f.). Ironically, Agrippa received his for the
victory over Pompey's son Sextus in 36 B.C. Agrippa appears with his on
coins (see Daremberg-Saglio 1, 2, 1536 and fig. 2012).
Note the word-symmetry of the 'golden line'.
685 ope barbarica: a Roman reader could hardly have failed to be
reminded of the famous passage of Ennius' Andromacha Aechmalotis
(Sc. 92-6 V): o pater, o patria! o Priami domus, / saeptiim altisono cardine
templum; / vidi ego te adstante(m) ope barbarica / tectis caelatis laqueatis /
aura ebore instructam regijice. Parts of this passage, which is cited en-
thusiastically by Cic. T.D. 3.44, had already been drawn on by Virgil
twice in A. 2 (see Austin's notes on 241 and 499). ope barbarica = 'with the
wealth of the orient'; cf. (perhaps also an imitation of Ennius), Milton
P.L. 2.3 f. Or where the gorgeous East with richest hand / Showers on her
kings barbaric pearl and gold. One should perhaps not press the undertone
that associates the fall of Troy with Antony's doomed cause, although the
fear of a resurgent Troy and the transference of the Roman capital to the
East seems to have been a real one from the last years of Caesar's life
onwards: Virgil may be insinuating a protest against such an idea; Horace
explicitly forbids it (Odes 3.3, esp. 57 ff.).
variis . .. armis: Heinsius' ingenious conjecture Phariis seems to be
based on a misunderstanding. After Augustus' helmet and Agrippa's
crown one might have expected some item of Antony's personal equip-
ment. In that case to give him multi-coloured arms, like the Etruscan
Astyr of A. IO. 181 (versicoloribus armis) would have been ludicrous,
whereas to make him wear Egyptian costume would have been a splendid,
if untruthful, touch (although Florus 2.21.3 credits him with sumptuous
Eastern dress, he presumably did not wear it on active service). But it is
Antony's forces which are in question here (variis armis complements ope
barbarica), and they were not all wearing Egyptian costume, as is evident
from 723 (see the note there).
686 victor: in spite of the early successes of his lieutenants against
 COMMENTARY

Parthia, Antony's own expedition in 36 B.c. had been disastrous: he had

lost some 37 per cent of his total force, including 22,000 of his veteran
legionaries (CAH IO, 75). The picturesque peoples that followed him to
Actium were hardly more than a miscellaneous and undisciplined rabble.
Virgil grossly exaggerates the number of his forces to make him appear a
powerful opponent (Markland's ductor (on Statius Silv. 2.6.3) is an inept
concession to history). It is unlikely that victor is an ironic sneer.
686 litore rubro: the mare Rubrum, 'Epv0p~ 0a.>..aaaa, included but was
not confined to the modern Red Sea. The Greek-speaking merchant who
wrote the Periplus maris Erythraei in the 1st century A.D., included in his
account the Indian Ocean, Persian Gulf, and even the Bay of Bengal.
688 Bactra: now a mere village, Balkh, in Afghan Turkestan, about
twenty miles south of the river Oxus, was known to the classical world as
middleman on the important trade-routes from central Asia and India,
purveying exotic Eastern produce. ultima (687), •at the end of the world',
like extremi 727, must not be pressed.
nefas is best taken as a parenthetic exclamation (as also at A. 7.73,
I0.673) because its feeling extends to each of the three words sequitur
Aegyptia coniunx: that Antony should even consort with a foreign woman
was bad enough, that he had actually married her was shocking (cf.
Horace Odes 3.5.5 ff.), and that any woman, let alone 'yon ribaudred nag
of Egypt', should accompany her husband on a military campaign, was
deplorable. This nefas is the faintest echo of the vicious propaganda put out
by both sides before Actium; Cleopatra was then called •a sorceress ... a
worshipper of beast-gods, a queen of eunuchs as foul as herself, a drunkard
and a harlot. .. a poisoner, a traitor and a coward' (CAH IO, 98). Pro-
pertius follows the party line (3.n.29 ff.); Horace (Odes 1.37) corrects it
with a tribute to her noble bearing in defeat and death.
689 ruere . .. spumare are historic infinitives (see the note on 493
confugere et . .. defendier), perhaps best taken here as the equivalents of
scene-setting imperfects, •they were all rushing together ... '.
Note the •interweaving' of the word-order in this and the next line, and
the mobility of the verse (see Appendix, p. 199)
690 The same line at A. 5.143. The rostrum, •beak', was basically a
beam projecting below the prow and above the water-line; this was
usually carved into three points, one above the other, each capped with
iron; looked at sideways it resembled a trident. The warships at Actium
no doubt had three-pronged rams, but three prongs are an anachronis!ll on
Aeneas' own (Homeric) ships at A. 5.143 (see J. S. Morrison and R. T.
Williams, Greek Oared Ships 900-322 B.C. (Cambridge, 1968), 7).
691 pelago ... innare: see the note on 651ftuvium ... innaret.
691 f. revulsas Cycladas: the ships are like floating islands. Virgil is
thinking primarily not of the Symplegades, •Clashing Rocks', at the
entrance to the Black Sea, but of Delos, roughly central in the Cyclades,
which according to legend floated until it was firmly fixed by Apollo;
cf. A. 3.75 ff. and R. D. Williams' notes there. So here, Delos seems to
have become unfixed again, bringing the whole group it dominated with it.
 COMMENTARY 185

693 tanta mole: another ablatival phrase which is not firmly anchored
to any part of the sentence. It is better to take it closely with the verb as an
adverbial phrase (ablative of manner), 'with such great bulk', 'so mas-
sively' (see the note on 199 magna ... mole), than to connect it closely with
the subject viri, or with the adjective turritis. And the same analysis holds
good for A. 12.161 f. ingenti mole Latinus / quadriiugo vehitur curru.
Translate the line: 'the crews press forward so massively in their
turreted ships'. (Mackail's suggestion that tanta-mole-turritis is a compound
epithet attached to puppibus is by no means impossible (see the note on
219 f.), but here the word-order is against it.)
turritis puppibus: the 'turreted ships' belonged to both sides, and
Virgil's imaginative picture 691-3 is historically accurate. The widespread
view that Actium was the victory of Octavian's light Liburnian 'cutters',
which admittedly received most publicity, out-manoeuvering the massive
and unwieldy ships of Antony is false and rests largely on a misunder-
standing of Horace Epodes 1.1-4. See further C. G. Starr, The Roman
Imperial Navy 2 (London, 1960), 7 f. Propertius 4.6.47 ff. is flattering
exaggeration.
For warships and merchant-ships carrying turrets which could easily be
erected and taken down, see C. Torr, Ancient Ships (Cambridge, 1895), 59 f.
694 'Flaxen flame and iron flying by means of shafts' is a typically
Virgilian hendiadys for a single weapon, the malleolus, which was basically
a bunch of esparto grass or flax, woven or tied together and smeared with
pitch (Non. Marc. s.v. 556 M; Livy 42.64); this was fixed to a shaft, which
could be thrown by hand, or from a bow (Amm. Marc. 23.4.14), or from a
ballista (Vegetius De Re Mil. 4.18). The last is not precluded here by manu,
which may simply = vi; see E. W. Marsden, Greek and Roman Artillery
(Oxford, 1969), 172.
Austin on A. 2.236 notes that stuppeus is first found in Virgil, and
frequently applied in later poetry to ships' cables.
695 arva ... Neptunia: an obvious metaphor, made even easier by the
application of aequor to both land and water; cf. A. 10.214 campos salis;
Ennius Ann. 143 V has (pont)i caerula prata and Cicero A rat. 129 Neptunia
prata, metaphors which may well have been inspired by Greek tragic
diction (see Norden 309).
nova ... caede, as at A. 10.515, = caede recenti (A. 8.195f.), 'fresh
slaughter'.
696 patrio . .. sistro: the sistrum, a kind of rattle, was used in the
worship of Isis (see J. Gwyn Griffiths, Plutarch's De !side et Osiride
(University of Wales, 1970), 525 ff.), with whom Cleopatra identified
herself. Virgil envisages its use as a battle signal: a contrast with the
Roman tuba is made explicit in Propertius' imitation of this line (3.11.43)
(ausa) . .. Romanamque tubam crepitanti pellere sistro. patrio . .. sistro, the
symbol of an oriental ruler-cult, is contrasted withpatrium ... sidus (681),
the symbol of the new Augustan ruler-cult.
697 geminos . .. anguis: (1) Henry rightly points out that for Virgil
twin snakes are a recurrent symbol of disaster and death (A. 2.203 f., the
186 COMMENTARY

sea-serpents which kill Laocoon and his sons; A. 7.450, in the hair of
Allecto who brings war and death; A. 8.289, threatening the life of the
infant Hercules): the symbol has the same value here. It embodied a wide-
spread belief that snakes were ill-omened and associated with death: they
appeared from the earth, the abode of the dead, and so were identified with
the spirits of the dead (cf. A. 5.84 ff. and 95 f.); twin snakes were repre-
sented on the wand of Hermes/Mercury as escort of the souls of the dead;
to see a pair of snakes mating was considered particularly unlucky (see
further Pease on Cic. Div. I.36).
(2) The symbolism does not preclude an allusion also to the actual
manner of Cleopatra's death. If there is such an allusion, this line is alone
among ancient sources in numbering two snakes specifically. Horace Odes
1.37.26 f., Propertius 3.11.53 and Floros 2.21.n (Loeb, ed. E. S. Forster)
use the plural; the majority of sources, further from the event in time,
including Plutarch Antony 85-6, Dio Cassius 51.14 and Servius here,
mention a single asp only. J. Gwyn Griffiths, ]EA 47 (1961), n3-8 and 51
(1965), 209-n (criticised by B. Baldwin, ]EA 50 (1964), 181-2), believes
that Cleopatra used two cobras, partly to bring death more effectively, but
also inspired by the symbol of the double uraeus. The uraeus, the hooded
cobra, was a distinctive emblem of the Pharaohs; doubled, it symbolised
territorially the 'Two Lands' of Upper and Lower Egypt. (Cleopatra VII
was also certainly represented as Hathor-Isis, and one or more snakes
were associated with the Isis-cult; but this association seems to have
developed in the Roman cult of the goddess, and to have been rare in the
Egyptian.) See also N-H 409 f., 419.
698 ff. Iliad 20.31 ff. provides precedent for a battle between gods-but
Virgil's Olympians are all on the same side.
698 omnigenumque deum: omnigenum is the reading of all MSS
except the first hand of the codex M ediceus which omits the first two letters.
This adjective does not occur anywhere else in Virgil, but the MSS of
Lucretius unanimously give some form of it at four places; Lachmann
changed all four to omne genus and on 5.440 suggested that the reading
here in Virgil might be Niligenum; Hoffmann got nearer to the MSS with
amnigenum (also referring to the Nile).
Superficially omnigenum looks comparable in formation to Troiugenum
(see the note on 82 (5)), and Priscian (732 P) recognised a nominative
omnigena; but the -gen- does not mean 'born of' as it does in the other
compounds, nor can the adjective here mean 'all-begetting', omniparens,
as it was later used by Prudentius (in Symmachum 1.12). The adjective
omnigenus is formed from the adverbial omne genus, 'of all kinds', like
multimodus from multi(s)modis.
Here in Virgil omnigenum should be kept:
(1) the mistake of the Mediceus is not significant;
(2) it is quite possible that it is the correct reading in all cases in
Lucretius (so Bailey on 1.1026), and is in any case what Virgil is likely to
have read in his copy of Lucretius;
(3) it makes good sense: the Egyptians worshipped a multifarious
 COMMENTARY

medley of gods, not only latrator Anubis represented with a dog's head, the
constant companion of Isis, who was thought to have raised him, but also
the crocodile, ibis, cat, and even some vegetables (see Juvenal 15,
beginning).
For the genitive plural terminations see the notes on 93 and 127.
698 The cult of Isis (and of other Egyptian gods) began to appear in
Rome in the time of Sulla. Between 59 and 48 B.C. repeated official
attempts were made to eradicate it, but altars and chapels always re-
appeared. In 43 B.C., as a political concession to the populace, the Triumvirs
intended to give official recognition to the cult by building a temple to
Isis, but it is unlikely that this ever materialised. Augustus expressly
shunned the worship of Apis (Suet. Aug. 93), and Virgil is faithful to his
conservative religious policy in nowhere mentioning them except here in
the defeat of a treacherous cause; but the Isis cult flourished privately,
especially with the demi-mondaines of Augustan elegy, a public temple was
established under the raving Caligula, and Lucan laments (8.831 ff.) nos in
templa tuam Romana accepimus I sim / semideosque canes et sistra iubentia
luctus / et, quem tu plangens hominem testaris, Osirim.
698 f. Again (cf. 696) closely followed by Propertius, 3.11.41 f.: ausa
Iovi nostro latrantem opponere Anubim, / et Tiberim Nili cogere ferre minas.
699 Why these Roman gods? Neptune was controlling the sea in Augustus'
favour; Venus was the constant protectress of theJulian house. Minerva
may be here as representative of the Capitoline triad whose other members
were not eligible to appear on this occasion, Jupiter being the final arbiter
of the struggle and Juno the arch-enemy of Trojans and Romans; as such
Minerva had the warlike attributes of Pallas Athene: so Bailey, RV 156, 201.
J. Gage, MEFR 53 (1936), 66 ff., points out that Venus had local
associations with Actium (where as 'Aq,po'8l77J AlvEla, she still possessed a
sanctuary), and that both she and Neptune figured on coins struck by
Octavian immediately after Actium; but Minerva cannot be accounted
for in any such connections. His suggestion that on the Shield she personi-
fies western uwq,pouvJ/"f/ fighting against eastern idolatry may contain part
of the truth. But her role, symbolised by the Palladium, as protectress of
Troy and of Rome (cf. Cic. Scaur. 48, Ovid Fasti 6.417 ff. esp. 427 f.),
fully justifies her presence here. For the historical background of 699 f.,
see Binder 242 ff.
701 tristes . .. Dirae: tristis does not often = 'sad', 'melancholy';
usually it means 'repellently gloomy, stem, bitter', so here, 'grim Furies'.
702 f. Discordia and Bellona recall the horrible abstractions in front of
the entrance to Hades in A. 6.273 ff.: there (280) Discordia is demens, and
here too she is insanely joyful at the violent division she will cause,
symbolised by her rent mantle.
In Ennius' lines postquam Discordia taetra / Belli /erratas pastis portasque
refregit (266-7 V = Horace Sat. 1.4.60 f. and cf. A. 7.622), Discordia
apparently breaks down the doors of the temple, to allow the spirit of war,
Bellona, to burst out. Virgil's description of the pair can hardly have
failed to remind his contemporaries of the ecstatic violence and frenzied
14 EAC
r88 COMMENTARY

mutilations of the fanatici in their worship of Bellona-Ma (see H. J. Rose,

ARR 128). Note the word-symmetry of 702.
704 Actius . .. Apollo: there was an ancient temple of Apollo on the
promontory of Actium (at the entrance to the Ambracian Gulf), mentioned
by Thucydides 1.29.3. Augustus ascribed his success to Actian Apollo,
enlarged this temple, set up many other monuments there, and raised the
status of the local festival to that of the Olympian. But desuper (705)
would more appropriately allude to the dominating position of the temple
of Apollo on Leucas. For Augustus' early attachment to Apollo, see the
note on 720. Note the 'framing' of the line with adjective and noun.
705 desuper: an excellent example of the force of a first-foot dactylic
word with a heavy pause after it, in the middle of a passage of descriptive
verse of a high order. One need not be fanciful to feel in 704, predominantly
spondaic and encumbered with an elision (arc(um) intendebat), the slow
and deliberate drawing of the bow; the arrow is let fly with sudden speed
(conveyed by the stopped dactylic rhythm, not by the meaning of de-
super), and causes universal panic and retreat (note the repetition of
omnis . .. omnis . .. omnes).
705 eo terrore: 'in terror at this', for eius terrore: this idiomatic
attraction of the genitive of a demonstrative or relative pronoun to the
noun dependent on it, so that it becomes the corresponding demonstrative
or relative adjective in agreement, is far commoner than the strictly logical
construction; cf., for example, Livy r.58.5 quo terrore (' through terror of
this', sc. his threat) cum vicisset (Tarquinius) obstinatam pudicitiam. See
KS I, 64 ff. and cf. A. 2.170 f. aversa deae mens. / nee dubiis ea signa
( = eius, aversae mentis, signa) dedit Tritonia monstris.
705 f. Notice the alternation of collective singular and plural in
Aegyptus ... Indi ... Arabs ... Sabaei. Collective singulars are frequent
with nouns of military designation like miles, pedes, eques, hostis, remex,
and also with names of nationalities especially when they occur in a
military or political context, cf., for example, Parthus velim quiescat,
Cic. Att. 5.9.r. These collective singulars of national names, which are
archaic in modem English, seem to have had their origin in Latin in lively
colloquial speech. See Lofstedt 12 ,12 ff.; N-H on Horace Odes 1.19.12.
706 Sabaei: the subjects of the kingdom of Saba (Sheba), occupied the
southern half of the Yemen, Arabia Felix (Eudaimon), the richest and most
fabled part of the whole peninsula. In the course of his rule, Augustus had
dealings with all the four nations of 705 f.
707 f. 'The queen herself, summoning fair winds, was seen setting sail
and on the very point of letting the sheets run out loose.' ventis vocatis
might be dative here after dare, but the same phrase is ablative at A. 3.253
and 5.2u. immittere = 'to let go forward'; this basic sense of mitto is
clear at Plautus Rudens 1015 mitte rudentem, 'leave hold of the rope'.
laxos is proleptic.
709-13 Note the identical scansion and rhythm of 709 and 7u: the
heavy spondees of death and grief, contrasting with the entirely dactylic
713 for the flight of the vanquished.
 COMMENTARY 189

709 pallentem morte futura, like Dido at A. 4.644 pallida morte

Jutura; and this recollection is significant: in much of the description of
Dido Virgil invites the reader to think of Cleopatra.
710 undis et lapyge ferri: according to Aulus Gellius N.A. 2.22.21 f.,
who refers to this line, the Iapyx blows from Apulia, the Apulians them-
selves being known as Iapyges, in roughly the same direction as the
caurus, i.e. from west-north-west. See Horace Odes 1.3.4 and N-H ad loc.
7n ff. On the Shield the Nile would of course be pictured as a river-god
(see the note on Tiberinus, 33 f.), opening the folds of his garment to receive
the fugitives.
712 Another line 'framed' by rhyming participles, cf. 669.

714-731 Augustus celebrates his triple triumph and vows temples for the
whole city in the midst of the festivities. Then, sitting at the entrance of the
temple of Apollo, he reviews the long procession of conquered and conquests.
Aeneas marvels at the pictures on the Shield, and shoulders the destiny of his
descendants.
714 triplici . .. triumpho: for the three victories in Dalmatia, at
Actium, and at Alexandria, celebrated on three successive days, 13-
15 August (but then still called Sextilis), 29 B.C., cf. Suet. Aug. 22.
714 f. invectus. . . / moenia: 'driving into the city'. Classical prose
would normally have repeated the preverb in- as a preposition, invectus in
moenia; poetry and post-classical prose can omit the preposition (invehor
with a simple accusative is usual and common in Livy, e.g. 2.31.3, 4.29.4,
etc.). moenia, by the common 'part for whole' figure, stands forurbem, or
the buildings of the city. For the significant coupling Romana ... Italis
see the note on 678.
715 f. In his own account (R.G. 20.4) Augustus says that he restored 82
temples, and names twelve other new temples (R.G. 19) which he built
(ter centum is used for any large number as at A. 4.510). Virgil is not
necessarily telescoping the Augustan (re)building programme into one day
(or even three): the temples are the subject of the vow and it is this which
he ' was solemnising'.
717 f. The day of feasting (13 August) was made an annual event for
the future by decree of the Senate; in 29 B.C. the consul M. Valerius
Potitus (auspiciously named-valere, potiri) offered sacrifice for the return
of the victor on behalf of Senate and People (Dio Cassius 51.20 f.). Just as
717-23 represent a telescoping of historical time (see the note, 720), it is
possible that in 717 ff. Virgil is at least glancing at festivities which
followed the days of the triple triumph: the dedication of the aedes Divi
I uli and celebration of the lusus Troiae on 18 August; the dedication of the
Curia Julia and the Atrium Minervae on 28 August; the celebration of the
second anniversary of the victory at Actium on 2 September; the celebra-
tion of Augustus' own birthday and the cult festival of Mars and Neptune
(on the Campus Martius) and of Apollo (near the theatre of Marcellus) on
23 September.
190 COMMENTARY

718 For the bucolic diaeresis and pause following a Greek loan-word,
chorus, see the note on 24.
It is important to remember that Virgil is describing what was on the
Shield, not stating the truism that all temples have altars; Peerlkamp's
proposal to read omnibus arae / ardentes, shifting the emphasis to a feature
they would have on a festival, is therefore ingenious but unnecessary.
720 niveo candentis limine Phoebi: the temple of Apollo on the
Palatine, next to the imperial palace, was the most splendid in the Rome
of its day. It was built of solid blocks of Luna marble (hence the 'snow-
white threshold'), and in front of the entrance stood the cult-statue of
Apollo, also of marble, dazzling white (candentis, appropriate for the
sun-god).
Augustus vowed the temple in the campaign against Sextus Pompey in
36 B.C., but although work began on it in that year, it was not dedicated
until 9 October, 28 B.C.: by that time the association of Apollo with the
victory of Actium had become paramount, so that Propertius refers to the
temple there as Actia monumenta (4.6.17), and Virgil alludes to its statue as
Actius Apollo (704).
It is no coincidence that Virgil places Augustus in the very spot the
statue of Apollo was later to occupy: he had been very ready to regard
Apollo as his patron, and even to identify himself with him (Suet. Aug. 70
and 94.4, and perhaps implied in tuus iam regnat Apollo, E. 4.10). Through
his support the new temple threatened to eclipse that of the chief state-
cult on the Capitol, from where the Sibylline Books were transferred to it.
The narrative of 714 ff. seems designed to give the impression that
Augustus' triumphal procession ended there and not at the temple of
Jupiter on the Capitol, as it in fact did.
In terms of the plan of the Aeneid, Augustus discharges the vow made
by his ancestor Aeneas to the Apollo of Cumae, A. 6.69 f. tum Phoebo et
Triviae solido de marmore templum / instituam.
No absolutely certain traces of this splendid edifice remain, but a
podium and stairway, previously assigned to the temple of Iuppiter
Victor, are now generally thought to have belonged to it (see Platner-
Ashby s.v., and Nash 1, 31 f.).
About fourteen months of historical time are telescoped in the picture
of Augustus sitting in the apparently completed temple of Apollo
(October 28 B.c.) reviewing the triumphal procession (August 29 B.c.);
see D. L. Drew, CQ 18 (1924), 195 ff.
723 Compare Lucan's description of Pompey's eastern allies (3.288 ff.)
coiere nee unquam / tam variae cultu gentes, tam dissona vulgi / ora; for
other possible reminiscences of Virgil in Lucan, see Heitland's introduction
ro8 ff., to Haskins' Pharsalia (London, 1887).
724 ff. The victae gentes of Virgil are hardly compatible with a sober
historical list; mostly, they are frontier-peoples of the empire, chosen to
suggest the extent of the world-conqueror's power.
Nomadum genus, as at A. 4.320, means the nomads par excellence,
in and around Numidia (in fact, Bogudes, king of Mauretania, had been
 COMMENTARY

an ally of Antony: Dio Cassius 50.6.4). The Leleges cannot be given a local
habitation in the Augustan period; for Herodotus (r.171) they were
Carians, and here too they are coupled with them (as already in Homer,
Il. 10-428 f.). But Caria too had ceased to exist in any geographical or
administrative sense after it became part of the province of Asia in 129 B.C.
The Geloni (cf. G. 2.115) were a Scythian race on the north of the Black
Sea around the Dnieper. Horace also associates them with the Euphrates
in a similar piece of patriotic exaggeration, Odes 2.9.21 ff. Medumque
ftumen gentibus additum / victis minores volvere vertices, / intraque prae-
scriptum Gelonos / exiguis equitare campis. The nomadic Dahae were also
a Scythian tribe, east of the Caspian, who shared the Parthian reputation
for archery on horseback.
Virgil's language about the Euphrates and Araxes might suggest that
Augustus imposed Roman rule on Parthia and Armenia in 30 B.C. Public
opinion may indeed have expected Augustus after Actium Parthos . ..
reposcere signa (A. 7.606); in fact he refrained from any offensive action in
the East and allowed the situations in Parthia and Armenia to remain as
he found them. The four or five legions which he garrisoned in Syria were
not stationed on or near the Euphrates which was that province's natural
line of defence against Parthia (CAH IO, 279 ff.), and although he settled
the princeling Artavasdes on the throne of Armenia Minor, he allowed the
rebel Artaxes to retain that of Armenia Major (through which the Araxes
flows), and even after Artaxes' death he' preferred' (R.G. 27.2) diplomatic
manipulation of it through a client-king rather than making it a province.
Lucan took a more historical view: but for civil war sub iuga . .. iam
barbarus isset Araxes (1.19).
The above places and peoples might very loosely be connected with
Augustus' victory over Antony and the triple triumph of 29 B.C., but the
Morini and the Rhine cannot be and in any case disturb the geographical
order (should 727 and 728 be transposed?).
The Morini, one of the stronger tribes of Gaul (Caesar B.G. 2-4-9),
lived on the coast of Belgium and northern France, approximately between
Boulogne and Zeebrugge. (extremi is surprising after Caesar's invasion of
Britain, but the Britons were cut off by 'the stream of Ocean' from the
mainland continent (E. r.66 penitus toto divisos orbe Britannos), and
Augustus at any rate did not intend to disturb them.) They and the Rhine
estuaries (bicornis 727) simply mark the northern limits of Roman power.
724 Mulciber (Vulcan) is connected with mulcere; Servius (here) takes
it of the power of fire to subdue everything (especially of course metals);
more probably it refers to the power of the god to quell fire itself.
discinctos . .. Afros: 'ungirt ', with no belt to prevent their robes from
flowing loose: to the Roman military mind a symptom of unenergetic
softness; but here the adjective is more picturesque than derogatory.
726 ff. The Shield showed pictures of personified river-gods, Euphrates,
Rhine and Araxes, being carried in Augustus' triumphal procession (see
the note on Tiberinus, 33 f. and especially Ovid A.A. r.223 quoted there).
727 bicornis: see the note on corniger 77; the two horns may allude to
192 COMMENTARY

the two main estuaries of the river, the Rhine proper and the Waal: the
adjective certainly refers to the two branches of the Tiber which form the
Isola Sacra in Rutilius Namatianus De Reditu Suo 1.179 f. fronte bicorni /
dividuus Tiberis dexteriora secat. But here the primary fact is the picture
of the river-god carried in the triumph, on which he could not of course
have had more than two 'tauromorphic' cornua.
728 pontem indi~natus Araxes: according to Servius a bridge was
built over the Araxes by Augustus to replace one erected by Alexander the
Great but destroyed by floods: when and where exactly he does not say,
and such a bridge is nowhere attested. But the reference may be general,
'disdaining to be bridged' at any point because of the violence of its
stream. This personification of the river was regarded by Quintilian
(8.6.n) as producing mira sublimitas.
729 dona (poetic pl.) parentis (Venus).
730 rerum is to be taken a1To Koivov with both ignarus and imagine,
'ignorant of the events, he rejoices in their picture'.
731 famamque et fata: -que et, like-que . .. -que, was already an archaism
in Old Latin (this follows from the fact that both formulae occur mostly
at the end of verses in Plautus and other early poets (e.g. Ennius Ann. 240 V
.. . malaque et bona dictu), and the verse-end was the most conservative
part in structure, see Fraenkel, Plautinisches im Plautus 209 ff.); both
were reserved by the early poets chiefly for passages of a loftier style, and
so readopted by the Augustan poets with a similar aim in view; both were
strictly avoided by Cicero and Caesar. See the note on 94 -que . .. -que and
LHSz 2, 515. Note also that although here -que et is a formula, -que being
quite inessential to the sense, in other cases, e.g. 625 hastamque et clipei,
both -que and et are true connectives.
This last line was decried by critics (cited by Servius auctus) as super-
fluous, undignified and neoteric. The objection is mainly against the
notion of 'lifting' weightless abstractions (Anchises at the end of A. 2
was quite a different matter). But other expressions in Virgil combine
material and abstract with no hint of playfulness or amusement: see the
note on 226 f.; and the ancient critics in question seem to have been obtuse
to the evident symbolism. The echo in famamque et fata hints at an intimate
connection, as in the title of Housman's lecture on 'The Name and Nature
of Poetry': cf.famafatisque canebant A. 7.79.
 APPENDIX: METRE AND VERSE

(Note. The following works are referred to by author only:

Hardie, W.R. Res Metrica, Oxford, 1920, reprinted 1934.
Knight, W. F. Jackson. Accentual Symmetry in Vergil, Oxford, 1950.
Norden, E. Aeneis, BuchVI, 3rd ed., Leipzig, 1926, reprinted 1957.
Winbolt, S. E. Latin Hexameter Verse, London, 1903.
An excellent short summary of the basic principles of prosody and Virgil's
metre can be found in W. S. Maguinness, Aeneid XII (London, 2nd ed.,
1960), 20-8.
Also to be consulted: the Excursus on the hexameter in Sir Frank
Fletcher's edition of Aeneid VI (Oxford, 1941, reprinted 1951); on the
fourth foot, A. M. Woodward, Philological Quarterly 15 (1936), 126 ff.;
T.E. V. Pearce, 'TheEnclosingWordOrderin the Latin Hexameter', CQN.S
16 (1966), 140-71 and 298-320. For further discussion of the Ictus-Accent
theory on which the following sketch is based, and for criticism of rival
theories, see L. P. Wilkinson, Golden Latin Artistry, n8 ff., and Appen-
dix I.)
Basic notions
The dactylic hexameter of Greek and Latin epic poetry is a line contain-
ing six rhythmic units or feet, with a regularly recurring verse-beat (ictus)
on the first syllable, necessarily long, of each foot. The last foot of the line
contains only two syllables and is followed by a slight rhythmic pause,
which severs the line from the one that follows, and enables the last
syllable which precedes it to be indifferently long or short (syllaba anceps).
Each of the remaining five feet may contain either two or three syllables
which are distinctly pronounced and not slurred over or suppressed (by
elision): the first long syllable which carries the ictus is followed either by
another long syllable (producing a spondee), or by two short syllables (pro-
ducing a dactyl).
Homer clearly regarded each hexameter line as a unit marked off by the
pause at the end. Virgil occasionally makes a unit of two lines by giving the
first an extra syllable at the end (making it hypermetric), which must be
elided before the first syllable of the second line if it is not to disrupt the
metre. Some intimate connection in sense is often reflected by the fusion
of the lines, as, for example, at A. 6.602 f.
quo super atra silex iam iam lapsura cadentiqu(e)
imminet adsimilis
(and see the note on A. 8.228 f.). Hypermetre occurs very rarely indeed in
Latin poetry before Virgil (five instances only are cited by Austin on
A. 4.558), and there are no examples in Greek epic poetry.
Latin poetry, like English and German poetry, was originally accentual
and depended for its rhythm on the natural stress (accent) which every word
1 94 APPENDIX

carries with it on one or more of its syllables. Epic poetry in Greek had
always been quantitative, with the rhythm of the verse made clear by the
length or time-value of the separate syllables of each word. Ennius was the
first to impose this quantitative Greek metre on Latin epic, and in so
doing posed the problem of how to reconcile the natural accentual bent of
the language with the alien discipline of quantity.

The Latin hexameter

(I) The middle of the verse.
Ennius is plausibly credited with having written the two following lines:
(1) sparsis hastis longis campus splendet et horret (V aria 14 V),
(2) poste recumbite, vestraque pectora pellite tonsis (Ann. 230 V).
These lines may be experimental, but in one vital respect they run counter
to the whole trend of the epic hexameter in both Greek and Latin: they
fall apart in the middle because the break between the third and fourth
feet of the verse coincides with that between the third and fourth words in
the line. The most stable characteristic of the epic hexameter is the
unbreakable link between the two rhythmic halves of the verse, formed by
a word which spans the end of the third and the beginning of the fourth
foot. The shortest and simplest word which can do this is obviously one of
two syllables, as in these (far more typical) lines of Ennius:
(3) hastati spargunt hastas, fit ferreus imber (Ann. 284 V),
(4) aeternum seritote diem concorditer ambo (Ann. 106 V).
In (3) hastas spans the critical point, and the break between words does not
occur in the middle of the verse, but 'cuts' it (producing a caesura) in the
middle of the third foot, and also in the middle of the fourth foot (or after
five and seven half-feet respectively, a penthemimeral and hephthemimeral
caesura respectively). In (4) the break in the third foot occurs not after the
first, long syllable (which would make it a strong or masculine caesura), but
after the second, short syllable (a weak or feminine caesura); and in (4) too
there is a strong caesura in the fourth foot.
Of the first two hundred complete and genuine lines of A. 8
(1-202, excluding 41 and 46), 67·5 % have the mid-point of the verse
spanned by a single word which has either (a) spondaic form (30 %) , or
(b) anapaestic form (24·5 %), or (c) iambic form (13 %). They are all illus-
trated in A. 8.102-9 (102-5 (a); 106-7 (c); 108-9 (b)). The figures quoted in-
clude words which come to have the metrical form in question through
elision of the last syllable.
The second obvious way of avoiding a break at the crucial point is to
place before it a single long monosyllable whose sense unites it much more
intimately with the word(s) following than with the word(s) preceding: it
is almost always a preposition or conjunction; cf., for example, A. 8.82-3:
candida per silvam cum fetu concolor albo
procubuit viridique in litore conspicitur sus.
 APPENDIX I95
Lines of this type form 13 % of the first two hundred of this book. It
follows that very nearly eight lines out of every ten in this section, which
there is no ground to regard as untypical, avoid the break at mid-verse by
one of the simple and obvious means so far discussed. (Note in passing that
words of two short syllables before the mid-point of the verse are not nearly
so common (3·5 % in the same section), probably because the rhythm they
produce fails to achieve the vital object; there is however a sequence of
two at A. 8.182-3.)
The most usual metrical forms of the link word cause a 'cutting• of the
verse-rhythm both before and after them, but to secure two caesurae on
either side of the middle of every verse, the spanning word would always
have to be of the metrical form of hastas (or its equivalent, e.g. gladios), or
of diem. This would be restrictive and monotonous, and Latin hexameter
verse in its maturity requires as indispensable only (i) a strong caesura in
the third foot, e.g.
(S) aspice, ventosi ceciderunt murmuris aurae (Eel. 9.58);
or (ii) a strong caesura in the fourth foot, e.g.
(6) munera praetere(a) Iliacis erepta minis (A. 1.647);
or (iii) a weak caesura in the third foot, e.g.
(7) spargens umida mella soporiferumque papaver (A. 4.486).
According to Hardie, p. 8, about 64 % of the lines in Ennius, Lucretius
and Virgil have a type (i) caesura (with or without (ii) or (iii) as well).
A line which lacks a break between words either before the last syllable
of the third foot or after the first syllable of the fourth came to be considered
unrhythmic or 'untuned'. This is the point of Horace's line in the Ars
Poetica (263):
(8) non quivis videt immodulata poemata iudex,
'not every critic can discern verses which are untuned • (like this one).
Note that the weak caesura after the second syllable of the fourth foot
does not qualify as a principal caesura because it is too far from the mid-
point of the verse. Virgil himself twice came very near to writing such an
'untuned • verse:
(9) portat ovans. ducis exempl(um) eventumque secuti (A. n.758),
(ro) magnanimi Iovis ingrat(um) ascendere cubile (A. 12.144).
Both of these are however helped out by the quasi-caesura, or break
between words in the fourth foot which would have been effective if the
last syllable in -um had not been elided. (In (8) and (ro) a slight break may
also have been felt between the prefix and root of the compound word,
im-modulata and in-gratum. There are seven lines in Lucretius whose only
possible 'cutting' is of this kind: see C. Bailey, Litcretius vol. 1, p. n2.)
(II) The beginning and end of the verse
It is obvious that if words of the same metrical form as the most common
link-words (spondee and anapaest) were built around them (two on each
side), a caesura would be produced in every foot. In fact however the link-
196 APPENDIX

word is often of a different (longer) metrical form than those discussed so

far (it is in very nearly 20 % of the first two hundred lines of A. 8): frequent
is a molossus (of three long syllables) or its metrical equivalents. The
surrounding words too can be of many different shapes. Great variety is
possible, and a mathematical computation of all the possibilities is not
needed to show that chance or necessity alone will generate caesurae in
feet other than the two central ones-unless a deliberate attempt is made
to avoid them. These secondary caesurae are important in echoing the
principal one(s), and are a regular and normal feature of the first two feet
of the line; in the last two, however, a significant restriction was imposed.
It is rare in Virgil to find a caesura after the first long syllable of the
sixth foot (which gives a monosyllabic ending to the line; see the notes on
43, 400, 679), or of the fifth foot (see the note on 382). On the other hand,
it is common to find a caesura after the second (short) syllable of the dactylic
fifth foot, as it is also to find a break between words (diaeresis) at the end
of the fifth foot. In other words Virgil followed Cicero, who standardised what
was already a marked tendency in Ennius, to end the verse with a word of
two or three syllables (of the metrical shape -~ and v-~ respectively).
The effect of this restriction on the rhythm of the verse-end is evident if
one considers the place of the natural stress on Latin words. According to
the 'law of the (syllaba) penultima', which holds good for them all, a word
of two syllables only bears the stress on the penultimate; a word of more
than two syllables bears the stress on the penultimate if that is long, but
on the antepenultimate if the penultimate is short. (Note in passing (1) that
the elision of the last syllable of a word has no effect on the position of the
accent; but (2) that when the enclitic -que is joined to a word it seems that
there is a change in the position of the stress(es), and -que is regarded as an
integral part of the new word, which is re-accented according to the general
'law'.) It necessarily follows that a word of two or three syllables which
ends a hexameter must have its accent in the same position as the ictus of
the sixth foot; and that any word of two or more syllables which precedes
(other than a word of two short syllables (a pyrrhic)) must have its accent
in the same position as the ictus of the fifth foot.
Strong rhythmic coincidence in the last two feet goes hand-in-hand with
the restriction on the metrical shape and length of the last word of the
verse: in all probability it was the desire for the first which caused the
second, just as it must have been the desire to establish the dactylic
character of the line which caused the virtual exclusion of spondees from
the fifth foot (see the note on 54 for verses which contain a spondee in this
position, versus spondeiazontes). The result is the characteristic rhythmic
unit which in recitation serves to identify the metre and punctuate the
flow of the verse with a familiar cadence. In the Virgilian hexameter the
first foot also is much more often a dactyl than a spondee (but not so
'regularly' as the fifth). It is rare to find a self-contained spondaic word in
the first foot, especially one followed by a sense-pause (see the note on 71),
no doubt because the inevitable coincidence of ictus and accent on it
echoes the same pattern in the last foot of the verse before, and prevents
 APPENDIX 1 97

the new line from establishing its dactylic character. As Latin, unlike
Greek, has a preponderant number of long syllables, this dactylic identity
of the metre was in danger of being lost, and needed to be preserved by the
'rule' of the dactylic fifth and the encouragement of the dactylic first foot
-features of hexameter poetry already evident before Virgil.
It has already been observed that in the first half of the verse caesurae
are naturally produced by words which span and link two consecutive feet:
this entails a conflict of ictus and accent and a consequent feeling of tension
in the rhythm of the line. It is clear that Virgil deliberately aimed at this
except in cases where its absence has particular point. After the conflict of
ictus and accent, their coincidence in the last two feet is even further
emphasised by contrast, and felt as a resolution of the tension in the line.
(Jackson Knight, op. cit. above, gives the convenient labels 'heterodyne'
to the conflict of accent and ictus, 'homodyne' to their coincidence.)

(III) The fourth foot

In the Aeneid as a whole there is coincidence of ictus and accent in the
fourth foot in 35·95 per cent of the lines (this and other figures in this section
are those of Jackson Knight, p. 108). In other words, about seven lines out
of twenty have rhythmic coincidence in the whole of the last half of the
verse. This is an appreciable drop on the estimate for Catullus 64 (the
Epyllion) (61·25 %), and for Lucretius 1 (51·49 %). This is unlikely to be
mere accident, and it can safely be deduced that Virgil deliberately tried to
prolong the conflict and tension of the line further than his predecessors.
Jackson Knight also notes that there is a tendency for Virgil to use
rhythmic coincidence in all the last three feet of lines which conclude
paragraphs: this can be illustrated by a number of cases (for a discussion of
this feature and further references, see R. D. Williams' note on A. 5.5 f.).
But in respect of the fourth foot, A. 8 has a higher percentage of coincidence
of ictus and accent (39·97) than any other book of the Aeneid, and also a
higher percentage of such lines which do not end a paragraph. Jackson
Knight, pp. 40 and 53, concludes from this that A. 8 or much of it is early
work, incorporated later into the bulk of the poem. A safer inference might
be that the nature of A. 8, lacking for its greater part any high dramatic or
emotional tension, is reflected to some extent in this feature of the verse. It
is remarkable that in A. 8, 184-267, where the drama of the Cacus episode
could have been reflected in a high proportion of heterodyned fourth feet,
there is in fact just the reverse: three lines in seven (42·86 %) have a
homodyned fourth foot. This may be an indication that Virgil in this
respect as in others is treating the episode primarily according to the
conventions of the epyllion, quite distinct from his own battle-narrative in,
for example, A. IO (where fourth foot homodyne averages 32·59%).
The rhythm of the Virgilian hexameter is marked by the regularly re-
peated position of the ictus in each verse, varied by divergence from the
I 2 3 4
'average' pattern of dactyl/spondee arrangement (D s s S; see Win bolt,
p. n5), and given tension by the clash of ictus and accent.
198 APPENDIX

The movement of the verse

Verse-movement depends on how a sense-unit (sentence, clause or
phrase) is fitted on to the hexameter framework. Two opposite tendencies
had shown themselves in pre-Virgilian verse, according to whether the
poet was more preoccupied with content or form.
(A) This passage is taken at random from Lucretius (r.387 ff.):
is porro quamvis circum celerantibus auris
confluat, haud poterit tamen uno tempore totum
compleri spatium; nam prim um quemque necessest
occupet ille locum, deinde omnia possideantur.
quod si forte aliquis, cum corpora dissiluere,
tum putat id fieri, quia se condenseat aer,
errat; nam vacuum tum fit quod non fuit ante
et repletur item vacuum quod constitit ante.
Here a prose passage has been draped over the hexameter framework
rather than fitted on to it; word-order differs from prose only through
necessary concessions to the metre and lacks any formal arrangement. But
the verse has great mobility; accident alone produces variety in the internal
pauses and no preponderant number of final pauses.
(B) The practice of Catullus (64) and the neoterics is quite different.
This example is taken from the post-Virgilian Ciris (29 ff.):
ergo Palladiae texuntur in ordine pugnae,
magna Giganteis ornantur pepla tropaeis,
horrida sanguineo pinguntur proelia cocco,
additur aurata deiectus cuspide Typhon,
qui prius Ossaeis consternens aethera saxis
Emathio celsum duplicabat vertice Olympum.
Here a 'poetic' passage has been nailed on to the hexameter framework;
word-order shows a highly formalised and artificial arrangement (by the
same pattern of the 'golden line', adjective I, adjective 2, verb, noun I,
noun 2). The verse is an aggregate of static units; the only sense-pauses are
final ones, and they are frequent.
Vocabulary and subject-matter apart, Virgil's epic verse leans more
towards (A) than (B), but is most often (C), a refined combination of both.
Statistics are of limited value in the appreciation of verse-technique, and
it would be unhelpful to compile them for each of the factors which can
determine verse-movement, which in Virgil usually depends on a combi-
nation of them which defies neat dissection. The chief factors are best
illustrated by examples.
(A) Mobility is achieved when different parts of the same sense-unit
(subject-verl:r-object, or noun-attribute) fall in different lines: e.g.
(a) dixerat et niveis hinc atque hinc diva lacertis
cunctantem amplexu mollifovet. ille repente
 APPENDIX 199

accepit solitam ftammam, notusque medullas

intravit calor et labefacta per ossa cucurrit. (387 ff.)
(b) ter leto stemendus erat; cui tune tamen omnis
abstulit haec animas dextra et totidem exuit armis:
non ego nunc dulci amplexu divellerer usquam,
nate, tuo, neque finitimo Mezentius umquam
huic capiti insultans tot ferro saeva dedisset
funera, tam multis viduasset civibus urbem. (566 ff.)
Noun and attribute are sometimes widely separated:
(c) miratur facilisque oculos fert omnia circum
Aeneas capiturque locis et singula laetus
exquiritque auditque virum monimenta priorum. (3ro ff.)
And such wide separation may accompany the interweaving or 'over-
stepping' (hyperbaton) of a part of one clause into another:
(d) (an unusually intricate example):
non secus atque olim tonitru cum rupta corusco
ignea rima micans percurrit lumine nimbos. (391 f.)
(In A. 8 cf. also 149, 168, 689 f.; and see R. D. Williams' note on A. 5.26-7.)
Mobility can be achieved, and frequently is by Lucretius, by ignoring
the emphasis which the end of the verse carries with it, and placing there
a word which cannot appropriately bear it (generally a part of speech
other than a noun, adjective or verb). Virgil curtailed this practice, more
drastically in A. 8 than elsewhere in his poetry: in this book prepositions
at the end of the line are always anastrophic, and adverbs and pronouns
intentionally given the emphasis of their position (see further Norden,
pp. 400 ff.).
(B) Static lines, where sense-unit and verse-unit coincide, are much less
frequent in Virgil than in neoteric hexameters. Although word-order
differs little from that of prose in
(e) hi bellum assidue ducunt cum gente Latina;
hos castris adhibe socios et foedera iunge, (55 f.)
static lines generally show the formalised arrangement of a pair of nouns
and adjectives, as in 'golden lines' like
(f) tempora navali fulgent rostrata corona.
(684; cf. 319, 616 and with some variation 355, 458, 497)
or with different patterns like
(g) haec pater Aeoliis properat dum Lemnius oris (454)
(h) et scissa gaudens vadit Discordia palla (702)
(i) Fortuna omnipotens et ineluctabile fatum. (334; cf. 341, 521)
In neoteric poetry one of the adjectives required by the pattern is some-
15 EAC
200 APPENDIX

times otiose and purely ornamental; in Virgil preoccupation with the

pattern sometimes leads to hypallage (see the notes on 542 and 654).
Aggregates of static lines, as in the example from the Ciris quoted above,
do not appear in the Aeneid.
(C) Virgil achieves his refined combination of the two techniques in two
chief ways: either (1) by imposing some kind of symmetry on the verse
without thereby interrupting its flow, or (2) by spreading a formalised
pattern itself over the last part of one and the first part of the next line.
(1) Two coordinated instances of the same part of speech 'framing' the
line at its beginning and end, or beginning and ending two successive lines
('linking' the lines), e.g.
(j) interea, cum iam stabulis saturata moveret
Amphitryoniades armenta abitumque pararet,
discessu mugire boves atque omne querelis
impleri nemus et colles clamore relinqui.
reddidit una boum vocem vastoque sub antro
mugiit et Caci spem custodita fefellit. (213 ff.)
(k) ego sum pleno quern flumine cernis
stringentem ripas et pinguia culta secantem. (62 f.)
A line may also be 'framed' by an adjective at the beginning and its noun
at the end, e.g.
(l) populeis adsunt evincti tempora ramis.
(286; see the note on 32)
'Framing' and 'linking' patterns do not of course of themselves produce
mobility of verse, but Virgil modified the neoteric tendency to associate
them with static lines (see however the note on 78).
(2) Examples of a complete pattern of a pair of nouns and adjectives
spread over two lines occur sporadically, e.g.
(m) ut sese inclusit ruptisque immane catenis
deiecit saxum, ferro quod et arte paterna
pendebat,. . . (225 ff.)
But the most common, adaptable and subtle form of pattern in Virgil
consists of one noun and its attribute (it deserves the name pattern
because both words occupy one of the emphatic positions of the verse-
at its beginning or end, or before or after one of the main caesurae;
contrast example (c) above). Great variety is possible; the following three
examples are taken at random:
(n) surgit et aetherii spectans C?r!~I}~!~ solis
!?~!!:}?- rite cavis ?I}~?-IP- de flumine palmis
........ ac talis effundit ad aethera voces.
sustinet (68 ff.)
(Notice that although this pattern is the same in three successive lines, the
verse is kept mobile by the relation of the dotted words: see (A) above).
 APPENDIX 201

(o) insula Sicanium II iuxta 11 latus Aeoliamque II

erigitur Liparen 11/umantibusII ardua saxis. II (416 f.)
{Note the position of words relative to beginning and end of verse and
caesurae.)
(P) fecerat et viridi 11/etam II Mavortis in antro II
procubuisse lupam, II geminos huic ubera circum
ludere pendentis II pueros et lambere matrem
impavidos. (630 ff.)
(Here, on balance, the pattern helps mobility.)
It will be obvious from the above account that there is another factor
which goes hand-in-hand with word-arrangement and also contributes to
variety: a range of different ways of relating the naturally emphatic
positions of the prose sentence to the emphatic positions of the hexameter
verse. In other words, there can be great variety in the position of sense-
pauses within the line; some of these positions have descriptive effect, all
of them help to avoid monotony (they are discussed in individual notes in
the commentary).

15-2
 INDEX VERBORUM
(References are to the numbered notes of the commentary)
abducere, 208 effulgere, 677
abiurare, 263 egelidus, 610
adire, 544 elidere, 260 f.
advertere (animum), 50 en, 612 f.
aequor, 89 enim, 84 f.
ales, 27 equidem, 471
aliquando, 602 erilis, 461 f.
(ali)quis, 578 est (with infinitive}, 676
almus, 455 et ( = etiam}, 9
altaria, 285 exercere, 378
alte (ex alto}, 395 exsors, 552
altus, 461 extundere, 665
ango(r), 260 f.
aperire, 681 facere (sacral term}, 189; (= fingere},
=
at ( a.th-ap}, 454 630 f.
ater, 198 f. fatidicus, 340
augustus, 678 fatum, 340, 512
auratus, 655 favere (sacral term), 173
avertere, 208 fervere, 677
fetus, 630 f.
bidens, 544 :fidus, 521 f.
bimembris, 293 flos, 500
foedus, 169
cadere (of stars, etc.), 59 fulgor, 427
caecus, 253 fulmen, 427
Caere, 597 f. Iuriae, 205, 219
caerul(e)us, 64 furor, 219
candidus, 82
canere, 49, 656 gaesum, 662
capere, 311, 363 gemitus, 419 f.
caput (of river) 65; 144 f., 569 f. glaucus, 33 f.
cavus, 599 gracilis, 409
celerare, go
citus ( = citatus}, 642 haruspex, 498
communis (deus}, 275 hoc ( = hue), 423
condensus, 497 honos, 61
consistere, 10 horridus, 348 ff.
coruscare, 661
coruscus, 391 f. iam, 190
iamque adeo, 585
daps, 175 ilicet, 223
dare(= immittere}, 30; (sacral term}, immanis, 192
106; (e.g. sonitum = sonare}, immittere, 707 f.
405; (e.g. dicta= dicere}, 541; inaccessus, 195
(iura}, 670 increpare, 527
debitus, 375 indubitare, 404
devexus, 280 induci ( = indui), 457
dies, 102 informare, 426
domus (of river), 65 ingens, 43, 192
ducere, 55 ingruere, 535
ductor, 6 (in)nare, 651
durare, 577 instare, 433 f.
 INDEX 203

instaurare, 283 f. proficisci, 51

in unum, 576 properare, 454
invehi, 714 f. prop(r)ius, 78
invitare, 177 f. putare, 522

labrum, 22 -que ... -que, 94

lacus, 74, 451
laetus (sacral term), 275 recoctus, 624
laquear, 25 referre, 342 f.
laudes, 273 religio, 349 f.
lautus, 361 relinquere, 125
libens (sacral term), 275 rostrum, 690
limen, 362 rudere, 248
litoreus, 43 rumor, 90
lustralis, 183
lustrare, 229 saeclum, 508
saepe, 353
mactare, 85 sagulum, 660
Manes, 246 sanies, 487
manus, 289, 694 secretus, 610, 670
micare, 391 f. silex, 233
monimentum, 312 simul, 4
monstrum, 312 sistrum, 696
mos, 316 solitus, 389
munus, 273 sollemnis, 102
spectatus, 151
neu (for nee), 582 spes, 580
nidus, 235 strictura, 420 f.
nod us (wrestling term), 26o; (Herculis), struere, 15
260 stuppeus, 694
nomen, 472 f. subvehi, 58
notus, 389 sudum, 529
novus, 389 superbus, 196
nubigena, 293 superstitio, 187
numen, 78, 512
numerus, 453 tabum, 487
tendere, 605
obex, 227 tenuis, 409
obsitus, 307 tortus, 429
olle, 94 torus, 177
omnigenus, 698 trepidus, 5
orator, 505 tristis, 701
Troiugena, 117
pallidus, 197 tuba, 523 fl.
pater, 72, 394 tutus, 603 f.
pauper, 105
pelagi (? locative), 333 unde domo, 114
perpetuus, 183 f. =
usus est ( opus est), 441
pius, 84
polire, 436 vastare, 8
porgere, 274 vastus, 193
Praeneste, 561 vestis, 659
primus, 6, 59, 134, 143, 288, 319, 407, vibrare, 524
561,602 virgatus, 660
proavus, 54 viscera, 180
procul, 98, II2 volens (sacral term), 275
profecto, 532
 INDEX NOMINUM ET RERUM
(References are to the numbered notes of the commentary)
Abetone (Monte), 599 Alba Longa, 43-5, 46, 47 f.
ablative Albula(e), 330 ff.
ambiguous or pervasive, 3, 31, 38, 89, Alcaeus, 209 ff.
213, 219 f., 510, 693 Aleman, 26-35
descriptive, 31, 219 f., 693 allegorical interpretation, 104, 285
locative: (of time), 407 f.; with alliteration, 2, 43, 47 f., 58, 391 f.,
totus, 427 500, 603, 644 f., 679
modal, 199 ambiguity
of material, 673 grammatical, 219 f.
of respect, 223, 318, 510, 530 intentional in meaning, 680 f.
(animis) Amphitryoniades, 103
simple abl. exploited, 215 f. anachronism, in material details, 79,
Accius, 130 690
accusative anacoluthon, deliberate, 401 ff.
adverbial use of neut. pl., 248, 489 anastrophe, of disyllabic preposi-
appositional or exclamatory, 339, tions, 32
487 (main note). 683 ancilia, 663 f.
direct object of• middle' verbs, 265, animals, sacrificial, 84 f.
457, 618 Anius, 157 ff.
of extent, 651 Annaeus Cornutus, 405 f.
of respect with adjectives, 114, 425 Antonius, Marcus, 679, 686
with perfect participles, 29 (main apices, 663 f.
note), 286, 520, 662 a,ro Ko,voii constructions, 213, 290,
Achates, 521 f. 520, 618, 730
Actium, battle of, 682, 693 Apollo
address, in hymn, 285 ff. Actius, 704
adjectives as archegete, 336
compound: as neologisms, 82; favoured by Augustus, 720
Greek flavour, 293, 596 Palatine temple, 720
Latin equivalent of Greek com- Apollonius of Rhodes, 22 ff., 285 ff.,
pound: 31, 64, 86, 219 f., 364 f., 408 ff., 421, 426 f., 589 ff., 622 f.
387 f., 673 apostrophe
neut. pl., as adverb, 248, 489; as in prayer, 84 f.
noun, 221 literary mannerism, 643
used adverbially, 20, 30 (main Ara Carmentalis, 306 ff.
note), 57, 107 f., 310, 465, 543; Ara Maxima
with participle, 299, 559, 683 cult, 279, 285
d8uV4TOV, 87 festival day, 102
adverbs foundation day, 269
comparative force weakened, 101 founder, 271
disliked by poetry, 299 Graecus ritus, 183, 276, 283 f.
relative, for adjective, 71 history of cult, 269 f.
shortage in Latin, 30, 299 locality, 104, 306 ff.
aegis: evolution, 435; users, 523 ff. • Ara Pacis ', 43-5
Aeoliae insulae, 416 f. Araxes, 724 ff., 728
Aetna, 416 f., 416 ff., 419, 453 Arcadian foundation of Rome, 313
agricultural layer of Latin vocab- archaism, 73, 84 f., 94, uo, 189, 441,
ulary, 307,436,522 522, 534, 639, 677, et saepe
Agrippa, M. Vipsanius, 3, 682, 684 Argiletum, 345 f.
Agylla, 478 ff. Argus quidam, 345 f.
A ethiopis, known to Virgil, 383 f. Armenia, 724 ff.
 INDEX 205

Arx Saturnia, 306 ff., 357 f., 358 cleansing ritual

Arx Tarpeia, 306 ff., 652 with incense, rn6
assonance, 43, 47 f. with water, 69 f.
asylum Romuli, 306 ff., 342 f. Cleopatra, 688, 696, 697, 709
Atlas, and his descendants, 130, 134 ff. cliche, inept, 461
augural terms, 4, 27, 78, 302, 523 ff. Cloelia, 646-51, 651
Augustus, I02, 269,666,678,714,715 f. colonisation, IO f., 336
Ausonius, 74 f., 90, 96 '.:onditional, 'mixed', 52off.
Ausonius = Italian, 328 coniuratio totius Italiae, 679
auspicium, oblativum and impetra- conjugation, varying, 677
tivum, 78 conjunctions, postponed, 22, 32, 454,
(481), 650 f.
Bactra, 688 connective et in second place
banquet, order of events, 184 ff., 283 f. in clause, 91
Bellona, 702 f. in verse, 9
Bentley, as critic of Virgil, 65, 197, copula, epexegetic, 436
236 f., 301, 4rn corona navalis, 684
bones creation myth, 314-32, 315
scarce in hell, 297 Cretan bull, 293 ff.
sexually excitable marrow, 389 f. Cronus, 319 f.
bowdlerisation, 405 f. cum inversum, 26 ff. (main note), 97 ff.,
276 f.
Cacus cum primum, generalising, 407 ff.
death, 260 Cures, 638
genesis of myth, 198, 225 ff. Cyclades, 691 f.
locality of cave and scalae, 190 ff. Cyclopes, 416 ff.
original nature, 198 Cyllene, 139
Caecilius Epirota (Q.), 3 Cythera, 523
Caecina, .523 ff.
Caere, 478 ff., 597 f. Dahae, 724 ff.
Callimachus, 31-96, 416 ff., 421, 452 f. Dardanus, 37, 134
Camillus, 630 ff. dative
Capitoline temple, 306 ff., 348 ff., 655 anomalous, with precari, 127
Capitolium, 306 ff., 357 f. of goal of motion, with compound
Gallic attack, 652 f. verbs, 125
Carinae, 306 ff. of local standpoint, 212
carmen Saliare, 285, 287, 663 f. possessive, 104
Carmentis, 51 f., 336 ff. predicative, 536
carpentum, 336 ff., 666 Daunus, 146
casa Romuli, 190 ff., 654 demonstrative pronoun
Catiline, 630 ff., 668 f. attracted to gender of predicate, 46
Cato the Elder, 630 ff. genitive becomes dem. adj., 705
Origines, 7, 492 ff. oblique cases of is avoided, 33
Cato the Younger, 630 ff., 668 f., 670 diminutives, 44, 660
Catullus, 19, 32, 54, 82, 91, 157 ff., Diomedes, 9
233, 518 f., 6IO, 660 Discordia, 702 f.
Centaurs, 293 ff. Donatus (Aelius), Life of Virgil, 3, 41,
Cerberus, 293 ff., 297 42-9, 427
Chalybes, 421 dreams, polluting, 69 f.
characterisation by style, 401 ff. dum, confused with cum, 374 ff.
chronology of Tiber journey, 59
Cicero (poetry), 54, 82, 633, 695 e-, privative and intensive, 6rn
circenses (ludi), 636 Echenus, 333
Claude, The Arrival of Aeneas at Egyptian gods, 698
Pallanteum, 115 £Ktf,paa,s T01TOII, 190, 233, 374-80,
Claudian, 425, 630 f. 416 ff., 478 ff., 597 (main note)
Claudius Caecus (Appius), 269 f. lK,f,paa,s xp6vo11, 26-35, 374-80, 416 ff.
206 INDEX

electrum, 402 Gellius (Aulus), 32, 84 f., 194, 227,

ellipse, of main verb, 18, 588, 678 405 f., 544, 634, 652, 659, 684,
endearment, expressions of, 581 710
Ennius, 6, 24, 29, 64, 72, 77, 82, 91, Geloni, 724 ff.
94, 98, Il9, 150, 166 f., 194, 221, gender
233,242,274,329,369,405,436, common, of animals, 641
487,505,520,522,526,596,599, varying, 227,233,561
630 ff., 631, 642, 662, 679, 685, genealogy, far-fetched, 134-41
695, 702 f., 731 genitive
Erulus, 563 ff. 'appositional', 231
esse, parts of, omitted in subordinate partitive, with neut. pl. adj., 221
clause, 52, 175, 251, 410 pl. in -um, 93, 127, 513; in -uum, 27
omitted in 'future infinitive', 534 possessive? (Aventini mans), 231
Etna, 416 f., 416 ff. genius, incandescent, 680 f.
Etruscans Geryon, 202, 563 ff.
coastal sites, 479 f. glossae, in epic, 662
culture, and Virgil, 65 Golden Age, and degeneration, 314-32
developers of astrology, 427; of Gorgon, symbol of shock, 435
augury, 498, 523 ff. Graecus ritus
eastern origin?, 479 f. Hercules, 183, 276, 283 f.
Etrusca disciplina, 498 Saturn, 319 f.
human sacrifice, 485 ff. Grattius, 33 f.
land-power, 65 groves, numinous, 348 ff.
official insignia, 458, 505 f.
relations with Latium, 55, 492 ff. half-lines, 41, 469
sea-power, 475 haruspices, 498
etymology head, (un)covered in worship, 276,
folk, 319 f., 322 f., 345 f., 425 319 f.
of proper names, 425 Hellenistic influence, 408 ff., 560 ff.
Evander, 51 f., 97 ff., 333, 336 ff., 508, hendiadys, 40, 289, 436, 531, 694
560 ff. Hercules
Evening Star, 280, 589 ff. birth, 103, 289
Invictus, 102, 198, 293
family streams, 142 labours, 209, 291, 293 ff.
fate nodus, 260
delayed, 398 f. poplar associated with, 276
relation to chance, 334 portrayal by Virgil, 205 ff.
Fauni, 314 scyphus, 278
Faunus Tirynthius, 228
as Pan, 343 f. worship: as lar domesticus, 542; at
as Silvanus, 600 f. Ara Maxima, 102, 103, 285; by
festival, 663 Salii?, 285; in countryside, 103;
Feronia, 563 ff. with tithes, 285; without lecti-
fetiales, at treaty-making, 641 sternium, 176, 279
fiction, historical, 47 f. Hermes, 138, 209 ff.
figura etymologica, 683 Heroic Age
flamines, 663 f. etiquette, 97 ff., 127, 157 ff., 463 f.
Fontinalia, 74 ff. stature, 162, 19i, 250, 518 f.
foreign cults, 187, 698 tact, 370 ff.
formulae Hesiod, influence on Virgil, 314-32
ritual, 74 f. Hesione, 157 f.
stylised with proper name, 18 Homer, 31-96, 86-9, 86, 97 ff., 124,
forum boarium, 198, 203, 285, 306 ff. 127, 162, 172 f., 183 f., 195,
forum Romanum, 306 ff. 225 ff., 244 f., 274, 370 ff., 382 f.,
foundation legends and symbols, 43-5 387 f., 408 ff., 417, 438, 454 ff.,
fulmen consiliarium, 523 ff. 508, 523 ff., 560 ff., 589 ff., 619 f.,
Fulvius Nobilior (M.), 630 ff. 622 f., et passim
 INDEX 207

Homeric Hymn Juno

to Hermes, 209 ff., 263 as symbol, 60 f.
to Pan and Ares, 293 not identified with any other deity,
homoeoteleuton, 15 (main note), 389, 43-5, 84 f.
5 26, 545 noverca, 288, 292
Horace, 77, 82, 106, 209 ff., 276, 410, Jupiter
526, 563 ft., 688, Q93, 724 ff., et Capitolinus, 306 ff., 348 ff., 636
alibi Feretrius, 348 ff.
Horatius Codes, 646-51 Fulgur, 352 ff.
hut, pre-urban, 654 Stator, 640
Hydra, Lemaean, 293 ff. Tonans, 306 ff., 352 ff.
Hylaeus, 293 ff. Victor, 720
hymn, form and structure, 285 ff.,
289, 290, 293 ff., 374-80 Laomedon, 18, 157 f.
hypallage, 3, 183 f., 236 f., 526, 542, lar, 542
654 Latium
hyperbaton, 21, 149 etymology of name, 322 f.
Va-rEpov 1rp6TEpov prehistoric, 82, 314-32
apparent, 85, 125 Laurens, Lauren tum?, 1
real, 201, 6II (main note) Lavinium, 1, 46
Leleges, 724 ff.
Iapyx, 710 Lemnos, 454
imperative, after ne, 39, 532 Leucate, 677
indicative Lipari, 416 f.
imperfect: conative and descrip- Livy and the Aeneid, 198, 210, 215 f ..
tive, 657; of continuing past 630 ff., 642 f.
action, 153; in -ibam, 436; with loan-words: Celtic and Punic, 662;
dum, 374 ff. Greek, 166 f., 278, 588
in indirect questions, 190 ff. Lucan, 3, II6, 171, 179, 227, 260 f.,
perfect in -ere, 53, 324 398 f., 425, 427, 487, 527, 543,
pluperfect: absolute, 152, 219 f., 698, 723, 724 ff.
276 f.; equivalent to (im)perfect, Lucifer, 589 ff.
358; with ubi, etc., 407 ff. Lucilius, 677
present: in catalogues, 141, 294; Lucretius, 2, 6, 19, 22, 27, 43, 64, 82,
generalising, with cum primum, 160, 171, 215 f., 299, 314-32,
407 ff. 389 f., 391 f., 394, 409, 430, 436,
infinitive 522, 633, 638, 659, 668 f., 677,
after verb of perceiving, 92 f. 698
after verb of representing, 630 f. ludi magni (Romani), 636
future, 534 Lupercal, 306 ff., 343 f. (main note),
historic (descriptive), 35, 215 f., 630, 663 ff.
493 (main note), 689 Luperci
of purpose, II2 f. on shield, 630 ff.
present: passive, in -ier, 493; with origin and ritual, 282, 663
iubere, 498, 646 f.; with ardere, Lycaeus mans, 343 f.
163 f.; with est, 676
insignia Macrobius, 7, 176, 244 f., 278, 279,
of kings and consuls, 505 f. 285 ff., 285, 383, 596, 679
of senators, 458 Maia, 138 ff.
interpolated line, 46 malleolus, 694
Isis, cult in Rome, 698 Manlius Capitolinus (M.), 652 f.
Italy, sense of unity, 678 Manlius Torquatus (T.), 660 f.
mare Rubrum, 686
Janiculum, 306 ff. masculine precedence, in syntax,
locality, 357 f. 209
Janus, 357 f. matronae, 630 ff., 666
Julius Caesar, 670 Memnon, 383 f.
208 INDEX

Messapus, 6 spondaic disyllable

metaphor, 19, 40, 307, 695 in first foot, 502
metathesis verborum, 3 in fourth foot, 57, 132
metonymy, 181, 385 f., 409, 533 spondeiazon, 54, 167, 345, 402,
metre and verse (see also Appendix) 679
bucolic diaeresis, 24 (main note), Mett(i)us Fufetius, 642, 642-5
198, 388, 718 Mez(z)entius, 7, 485 ff., 492 ff.
caesura, trochaic, in both fourth Minerva
and fifth foot, 260 and aegis, 435
descriptive verse, 91, 219-24, 228 f., and manual workers, 409
237-9, 388, 446 f., 452 f., 489 f., and Pallas Athene, II6, 699
525, 549, 596, 705, 709-13 Minio (Mignone), 597 f.
different verse-position of repeated misrepresentation, 10 ff., 41, 55
word, 39, 52, 91 Morini, 724 ff.
elision Morning Star, 589 ff.
descriptive effect, 446 f., 525, 705 Moschylus, 454
of monosyllable, 13, 144, 503, 557 Mulciber, 724
five-word line, 334 multum in silentio, 23 ff., 59, 263,
four-word line, 158, 263 531
' framing' of line myth, aetiological
by noun and adjective, 32, 286 Alba Longa, 43-5
by participles, 63, 669, 712 Hercules and Cacus, 198
by two finite verbs, 78, 230-2, of birth of mankind, 315
560 ff. Tarpeia, 652
hypermetre, 228 f.
ictus and accent, clash of, 58, 452 f., Namatianus, Rutilius, De Reditu Suo,
489 f. 478 ff., 561, 727
line-ending Nemean lion, 293 ff.
with monosyllable, 43, 71, 81-3, neologisms
400,679 adjectives, 82 (main note), 195, 552,
with two disyllables, 364, 382 559
with polysyllable, 416 nouns, 89, 143 f.
(and see spondeiazon) verbs, 404, 677
paragraph-closure, features, 393, Neptunus, 699
560 ff. Nomades, 724 ff.
pauses, after verse-units: nomen omen, 315, 322 f.
-, 72 nominative
-v, 33 appositional, 487
-vv, I-6, 33, 59, 87, 151, 237-9, exclamatory, 77, cf. 487
484, 560 ff., 705 nouns
- -, 71,502 in -men(tum). 89
-\,,&!-, 147,227,469,540 repeated to express reciprocity,
-l,,&!-v, 175 448 f.
-l,,&!-vv, 390, 525 subtleties of copulation, 436
-l,,&!-VV-, 49 ff., 443 nova cacozelia, 3 (main note), 23 ff.,
- V V - ~ - v , 294,549 183, 263, 542
-\,,&!-~-\,,&!-, 573 ff. numerals, distributive, 47, 168
-\,,&!-l,,&!-~-v, 260
-\,,&!-'=C::.1.-~-vv, 24, 198 Oceanus, 589 ff.
-'=C::,/.-'=C::,1.-\,,&!-VV-V, 443, 533 Oechalia, 290
rhyme olive branch, u6
leonine, 649 onomatopoeia, 2, 596
medial and final, 8, 682 Ovid, 33 f., 49 ff., 57, 65, 77, 84 f., 87,
of successive line-endings, 268- 153, 171, 197, 198, 260, 260 f.,
72, 396 f., 568 f., 620 f., 646 279, 315, 322 f., 343 f., 348 ff.,
sequential (see homoeoteleuton) 364 f., 4o7 f., 425, 427, 451, 524,
rhythmic pattern, 91 680 f.
 INDEX

Palatine personal, mutually attracted, 475

rival of Capitoline, 313 relative, postponed, 37
settlement, 98 f. proper names, ornamental, 166, 364 f.
Pallanteum, 51 f. Propertius, 94, 152, 198, 205, 209 ff.,
paragraphing, guides to, 78,393, 56off. 260, 275, 526, 662, 675 ff., 678,
parataxis, 7 f., 171, 190 ff., 225-30, 688, 696, 698 f.
506 f. prosody ~
Parrhasia, 343 f. dehinc, 337; !talus, 331; -que ...
Parthia, 724 ff. -qu(e), 425; religio, reliquiae, 349;
participles Sicanius, Siciinus, 328
present: after verb of perce1vmg, semi-consonantal i, 194 (semi-
92 f.; after verb of representing, hominis), 227 (obice), 297 (seme-
630 f.; with future or past sa), 599 (abiete)
reference, 157 ff. syllable, 'irrationally' lengthened
perfect: with '-able' sense, 195, under ictus, 98 (procitl), 363
559, 588; with present reference, (subiit), 425 (-que ... -que")
37, 407 f., 636 syncope, 73 (periclis), 274 (porgite)
pathetic fallacy, 38, 92 f., 94, 239 f. synizesis, 66 (deinde), 292 (Eurys-
patronymic theo), 372 (aureo), 383 (Nerei)
coinages of, 638, 648 vowel-length, distinct from syllabic
implications of, 521 f. quantity, 43, 227, 599
Pelasgi, 600 purification, by water, 69 f.
penates, 542 Pyrgi, 478 ff.
Penales publici populi Romani, 679
periphrasis -que, relation of nouns so joined, 220,
dignified, 181 289, 436
with proper name (Cacifacies) 194, -que ... -que, 94, 291
(Caci mens) 205 ff. -que et, 731
times of day and night, 97, 408 f. Quintilian, 82, 94, 641, 677, 728
Pheneus, 138, 165, 343 f. quisque with plural verb, 661
pilentum, 666
Pinarius, 269 f. religious continuity, interrupted, 1 ro
Pliny the Elder, 33 f., 348 ff., 402, repetition
427, 630 ff., 646-51, 684 intentional: diplomatic, 17; pa-
Pliny the Younger, 74 ff., 82, 96 thetic, 71; rhetorical, 133, 140 f.;
plural ritual, 71, 74 f., 84 f., 293 ff.
evasive, 379 of line(s), 20 f., 35, 46, 80, 149, 171,
' oracular', 503 207 f., 305, 538 ff., 667, 690
poetic, l9 (main note), 219, 324, of passage, 42-9, 449-53
362, 375, 463, 729 of phrase, 23 ff., 28, ro4, 173, 183 f.,
poplar 268-72, 448 f., 579, 679, 709
and Hercules, 276, 286 of word in close context, 194, 205,
and rivers, 32 242, 378, 396 f., 568 f., 657
Porse(n)na, 646, 646-51 revision, lack of final, 268-72, 532 ff.,
Porta Carmentalis, 306 ff. 543, 654
Potitius, 269 f. rhetoric
Potitus, Valerius, 285 anaphora, 349 f., 560 ff., 564 ff.
prayers exordia, 374-80
at sunrise, 68 f. four-hexameter unit, 1-6, 374-80,
pro populo, ro6 560 ff.
preposition, delayed, 143 f. organisation of speech, 127 ff.,
proleptic adjectives and participles, 560 ff.
37 (main note), 227, 260 f., 364 f. tricolon, of diminishing length,
pronouns 230-2
omitted: object of iubere, 498, river-gods
646 f.; reflexive subject in 0.0. dress of, 33 f., 7n f.
534 'home' of, 65
210 INDEX

river-gods (cont.) subject, abrupt change of, 119, 271,

horned, 77, 727 494-6, 546
prophetic, 62 ff., 74 ff., 336 ff. subjunctive
shrines of, 74 ff. in causal clause, 129 f., 323,
rivers 650 f.
flowing backwards, 87 of wish, 560
venerated, 31, 32, 74 ff. past jussive, 643
Rumon, 90 pluperfect, in final clause, 205 f.
potential (second person singular),
Sabaei, 706 650
Salii Sueius, 90
on shield, 630 ff. Suetonius, 3, 306 ff., 663 ff., 724
origin and ritual, 287, 663 f. synaesthetic expressions, with vi-
Palatini and Collini, 287 deo(r}, 35, 526
worshipping Hercules?, 285 syncope, 73, 274
Saturnus, 314, 319 f., 322 syncretism, 314, 342 f.
scalae Caci, 190 ff., 198
Seneca the Younger, 348 ff., 364 f., table, used for altar, 279
427, 523 ff. taboo
separation, of connected words, 28 f., as proof of antiquity, 663 f.
391 f. on enemy weapons, 562
Shield of Aeneas, 625, (subject- on ministrants, 179
matter) 630 ff., (arrangement) on victims, 544
671, 672, 675 Tarcho, 9, 505 ff.
shields, 562, 625 Tarpeia, 652
Sicani, 328 Tarquinius Superbus, 646-51, 652
Siculi, 328 Teucer, 134
sidus Iulium, 680 f. 'theme and variation', 7 f. (form), 52,
significant names, 250, 425 74 f., 88, 171 (history}, 215 f.,
Silius Italicus, 227,438,456, 527, 659, 241-58, 289, 349 f., 375, 461 f.,
660,662 611, 620 f., 627
Silvanus, 600 f. Thybris, 72, 330 f.
similes Tiber, 31, 62, 65, 77, 86, 96, 473
length, 374-80 Tiberinus, 31, 72
points of comparison, 408 ff. tibicen, 212, 268-72, 469
treatment of sources, 22 ff., 408 ff. Tibullus, 1, 46
simple verb, for compound, 110, 189, time
329, 345 f., 639 as material object, 94
singular, collective, 34, 153, 474, 599, circumlocutions for, 97, 369, 408 f.
705 f. historical time telescoped, 715 f.,
sky, Etruscan division of, 427 717 f., 720
snake symbolism, 697 topography, confused, 215 f., 236 f.,
soloecophanes, 260 f. 461
somnus, half-personified, 67 of site of Rome, 306 ff.
sordida vocabula, avoided, 181, 456, transition
620 f.; cf. 44, 205 dusk to dawn, 369
sound-patterns, 2, 391 f., 446 f., 603, marked by iam, 190
644f. and iamque adeo, 585
sow, legend of, 43-5 trees, men-bearing, 315
spear-worship, 3 trumpet-omen, 523 ff.
spinning, classless activity, 411 f. Tullus Hostilius, 642-5
Statius, 8, 74 f., 77,297,301, 404, 461, tumultus, procedure for, 1
560 ff., 662 Turnus, 1, 3, 9, 492 ff., 562, 621
Stoic dogma, 133, 334, 398 f. Typhoeus, 293 ff.
strangers, unwelcome at festival, Typhon, 293 ff.
172 f.
suasoria, 10 ff., 374 ff. Ufens, 6
 INDEX 2II

Vipranius?, 3
Vaccina, 597 f. Virgil
Valerius Flaccus, 3, 32, 68 f., 77, 97 ff., ancient criticism of, 3, 41, 276, 285,
98, 260 f., 408 ff., 416 f., 425, 454, 383, 405 f., 492 ff., 731
475 f., 523 ff., 560 ff., 562 and Cisalpine Gaul, 668 f., 678
Valerius Probus, M., 405 f. Etruscan connections?, 65
variation of prose idiom, 3, 56, 385 f., vocative, •archaic', 77
405 f., 475 f., 510, 569 f., 651 Volcanal, 198
Varro, M. Terentius, 43-5, 89, 276, Vulcan, 198, 416 f., 454
306 ff., 319 f., 322 f., 330 f., 345 f., Vulcano, 416 f., 416 ff.
357 f., 358, 420 f., 436, 679, 684
Velabrum, 306 ff. word-order (and see Appendix)
Venulus, 9 interwoven, 28 ff., 168, 391 f., 689
Venus, at Actium, 699 symmetrical, 32, 95 f., 276, 454,
verse, see metre 458, 526, 616, 654, 684, 702
victims, sacrificial
age, 544 zeugma, 260 f., 731, (of abstract and
sex and colour, 84 f. concrete) 226 f., cf. 731