STAGING AND SCENERY IN THE ANCIENT

GREEK THEATRE *
BY T. B. L. WEBSTER, M.A., Hon. Litt.D.
PROFESSOR OF GREEK, UNIVERSITY COLLEGE, LONDON

T HE publication of Menander's Dyskolos, dated by an almost

certain emendation in 316 B.C., gives us, with much else,
information about the staging and scenery of an early New
Comedy, produced only some fourteen years after the rebuilding
of the Athenian theatre by the statesman Lycurgus. 2 Discussion
of this will be the end of our survey ; the later Hellenistic theatre,
in which the actors performed on a high stage, is a logical develop-
ment from the Lycurgan theatre : the picture frame in which the
actors appear is completed by the provision of a bottom edge.
But that only happened when the creative period of drama was
over. The interesting questions are these : when and under
what compulsions was the picture frame created, what preceded
it and what different compulsions were then operative. The
reason for reopening this question is that other evidence besides
the Dyskolos has been discovered, and some of the old evidence
can with profit be reconsidered, but I shall allude to points
already rightly established only in so far as they are necessary to
complete the story.
Before trying to interpret, I will state the agreed excavational
facts about the Athenian theatre :
1. The old (and probably original) temple of Dionysos
Eleuthereus,3 in whose honour the City Dionysia with its dramatic
performances was held, lies just below the theatre and from the
material and method of building must be pre-PersJan, i.e. earlier
than 480 B.C., and is probably much earlier. An approximate

1 A lecture delivered in the Library series of public lectures.

2 On the date 330-326, cf. H. A. Thompson, Hesfxria, xii (1943), 300;
G.T.P. (= Greek Theatre Production), pp. 43 ff.
3 The old Dionysos temple: Pickard-Cambridge, Theatre of Dionysos
(= P.C.T.), p. 3. Cf. also my Greek Art and Literature, 730-330 B.C.. pp. 58 f.
493
494 THE JOHN RYLANDS LIBRARY
date of 560/40 B.C., is given by the limestone pediment of
which a fragment was found near the temple.1 It represents a
maenad between two satyrs, who are playing the flute, and the
figures closely approximate to the satyrs and maenads on black-
figure vases by Lydos and the Amasis painter. Presumably the
god himself occupied the centre of the pediment. Whether the
subject was Dionysos with satyrs and maenads or Dionysos and
the Return of Hephaistos we cannot say. Nor can we be certain
whether the pediment belongs to the temple of which the founda-
tions remain or to a yet earlier temple. But we know that ten to
twenty years before the introduction of tragedy the temple of the
god was adorned with a representation of the god accompanied
by satyrs and maenads.
2. (a) In the theatre itself the oldest remains are three short
pieces of wall of the same construction as the old temple; 2 all
are within the later stage-building, and they are numbered from
East to West SM 1, J 3, SM 3. All are situated higher than the
foundations and bottom course of the old temple, SM 1 by 86
cm., J 3 by 1.35 m., SM 3 by 37 cm. There is general agreement
that SM 3 belongs to a way up from the temple to the orchestra,
but SM 1 has been attributed both to the boundary of the
orchestra itself and to the terrace wall supporting the orchestra.
These portions of wall mustbepre-Persian, but more we cannot say.
2. (b) In the auditorium clear signs were found that the hill
was hollowed out in the very early fifth century. 3 It is perfectly
possible that this forming of the auditorium coincided with the
making of the orchestra represented by SM 1, J 3, and SM 3.
3. The earliest surviving trace of stage buildings is the
rectangular foundation T, which projects forward from a long
wall H carrying slots for wooden timbers. This wall is itself
backed by the back wall of the Stoa below it, which was evidently
part of the same reconstruction plan.4 Breccia was used for
Athens, National Museum 3131. Studniczka, AM xi (1886), 78, PI. 2 ;
Frickenhaus, J.d.L xxxii (1917), 2, Fig. 1 ; Brommer, Satyroi, p. 27 ; Bieber,
History of the Greek Theater (= H.T.), p. 99, Fig. 144 ; P.C.T. p. 4.
2 P.C.T. pp. 5ff. ; Fiechter, Das Dionysostheater in Athen, i 39 f.; iii. 48,
66 f.; Dinsmoor, Studies presented to D. M. Robinson, i. 309 f.
3 P.C.T. p. 14; Dorpfeld-Reisch, Das griechische Theater, pp. 30-1.
4 P.C.T. p. 16, 21 ; Bulle, Untersuchungen, pp. 54, 64; Dinsmoor, p. 317.
 ANCIENT GREEK THEATRE 495
these foundations, and breccia was not used much before 430 B.C.
in Athenian architecture. On the other hand this reconstruction
cannot have been much later since the pottery found under T
and the wall was all of the fifth century or earlier. The highest
course of T is 95 cm. above J 3. Therefore the building of T
and its wall H effectively obliterated any earlier stage-buildings.
T and its wall H may be called for simplicity the Periclean theatre.
4. The Lycurgan stage building, which was built about 330
B.C., had a new marble facade with projecting wings some 5 m.
in front of T. The length of the facade between the wings was
20m.
The certain dates then are (1) pediment of old Dionysos
temple (whether in its original form or not) 560/40, (2(a)) earliest
surviving orchestra before 480, (2(b)) forming of the auditorium
soon after 500 B.C., (3) earliest surviving stage buildings with
stone foundations about 430 B.C., (4) Lycurgan theatre 330 B.C.
The first question is whether the earliest surviving orchestra
belongs to the sixth century Dionysos temple or to the early fifty-
century auditorium. Or to put it in another way, is it the theatre
of Thespis or the theatre of Aeschylus ? If the remains do not
answer this problem, we can at least consider other arguments.
The strongest seems to me the complete disassociation of the old
temple from this orchestra. If they were built at the same time,
why was the temple not placed on the same level and facing on to
the orchestra ? If, however, the orchestra was constructed half
a century later in response to a crisis, then the disassociation
becomes intelligible.
The crisis which caused the forming or remodelling of the
auditorium in the early fifth century is generally believed to be
the collapse of the iJyria, the wooden stands on which the audience
had hitherto sat. 1 The notices about the ikria are both late and
confusing, and the word itself can be used both for the stands
(which collapsed) and for wooden seats in the later auditorium.
The Suda lexikon gives under Pratinas : "he competed with
Aeschylus and Choerilus in 499-496 B.C. . . . when he was
giving a performance, the ikria on which the spectators were
standing collapsed and after this the auditorium (theatron) was
1 ACT. pp. I Iff.
496 THE JOHN RYLANDS LIBRARY
built at Athens ", and under Aeschylus : "he went into exile in
Sicily because the ikria collapsed when he was giving a perform-
ance." Aeschylus went to Sicily between 472 and 467 and again
after 458; Pratinas died shortly before 467 ; therefore the
collapse of the ikria must have been before Aeschylus' first visit.
The story of his exile is nonsense ; some scholar has converted,
as often, a chronological sequence into a causal connection. The
names of Aeschylus and Pratinas, who wrote tragedies and satyr
plays, tie these ikria to the City Dionysia and the precinct of
Dionysos Eleuthereus.
These are the only references to the collapse of the ikria.
Another group speaks of " seeing from the poplar tree".
According to Hesychios " the poplar tree was near the temple
where before the auditorium (theatron) was made they fixed the
ikria ". If the temple was the old temple of Dionysos (which is
the natural interpretation), we have here precious evidence that
before the building of the auditorium the ikria were put up near
the old temple. This then was the theatre of Thespis, and it
lasted until the ikria collapsed in the days of Pratinas and
Aeschylus ; then the new auditorium and the new orchestra, our
earliest surviving orchestra, was built.
But we are told that " various passages speak of a single poplar
tree (aigeiros) in the agora which must surely be identical with the
aigeiros close to the ikria 'V Reference to the various passages
shows that only two refer to the Agora : neither of these mention
ikria ; one of them (Andocides, 1, 133) refers to a white poplar
(leuke), not to a black poplar (aigeiros) ; the other in its early form
refers to aigeiroi in the plural not in the singular. Knowing,
therefore, that Aeschylus and Pratinas performed at the City
Dionysia, we can firmly separate the single black poplar near the
temple of Dionysos from the group of black poplars in the Agora.
I do not, of course, deny that there were early dramatic per-
formances in the Agora or that their audiences sat on z'^ria. But
the ikria did not collapse and the performances had nothing to do
with the City Dionysia : there is a tenuous connection with the
Lenaion and the Lenaia, and the dramatic performances at the
Lenaia were transferred to the theatre of Dionysos about the
1 P.C.T. p. 12 ; cf. Wycherley, Athenian Agora, III, Testimonia, nos. 722 ff.
 ANCIENT GREEK THEATRE 497
middle of the fifth century.1 The notices are quite consistent
with this interpretation : according to Photius " the ikria were
the seats in the agora from which they saw the Dionysiac contests
before the theatre in the precinct of Dionysos was built " ; " the
orchestra was a name used in the agora first and later for the lower
semi-circle of the theatre. There the choruses sang and danced."
Thus the evidence is good for ikria and orchestra in the Agora,
and the orchestra was near the statues of Harmodios and Aristo-
geiton. 2 By the time of Socrates' Apology in 399 B.C. it had
become the book-market, which agrees perfectly with the trans-
ference of the dramatic performances to the theatre in the middle
of the fifth century. The further problem is the identification
of the Dionysiac contests in the Agora with the performances at
the Lenaia festival. The Lenaian festival was held at the Lenaion
" a shrine of Dionysos, where they had contests before the
theatre was built". This is the familiar phrasing, but the
connection with the Agora is tenuous : the Lenaion is said to be
near the Kalamites /zeros ; Aischines' mother held her rites in the
shed near the Kalamites heros ; on this passage of Demosthenes
one ancient commentator remarks : " shed. The building with
big doors in the Agora." This is the only evidence for putting
the Lenaion in the Agora. On the other hand what were the
Dionysiac contests in the Agora if they were not the contests at
the Lenaia which we know were later transferred to the theatre ?
It is therefore tempting to suppose that the orchestra in the Agora
was in the enclosure of the Lenaion. For the history of Greek
staging this tells us very little. The performances in the Agora
took place on a dancing floor (orchestra) and the audience sat on
benches. The performances at the Lenaia (whether identified
with the performances in the Agora or not) took place in a large
precinct which included the shrine of Dionysos. It must also
have had ikria and an orchestra.
Similarly the theatre of Thespis had ikria near the old temple
of Dionysos Eleuthereus and, we must assume, a dancing floor
(orchestra). We cannot say how much levelling had to be done
1 Pickard-Cambridge, Festivals (= P.C.F.), P- 38, and p. 25, nos. 22-7 ;
cf. also Wycherley, Testimonia, nos. 524-8.
t Quoted by Wycherley, no. 276.
498 THE JOHN RYLANDS LIBRARY
to make the dancing floor, but if the stories about the collapse of
the ikria are true we must suppose that the hill was only used to
support some of the seats. If, as seems natural, the orchestra was
in front of the temple, the ikria on the southern side must have
been stands and it was presumably these that collapsed. If this
is accepted, one further and very important point follows. The
theatre of Thespis had from the beginning the front of the temple
of Dionysos as a background. Thespis and his successors down
to and including Aeschylus in his early period acted in front of
its columns and under its pediment. Thus a doorway with
columns and pediment is from the beginning the natural back-
ground of Greek tragedy. 1 It may not be chance that the
assumed width of this background is very close to the preserved
width of T, the projecting foundation of the Periclean stage
building.
When the ikria collapsed, the decision was taken to use the
steepened hillside as a foundation for the seats and to have an
orchestra well above the temple supported by a retaining wall.
The temple thus fell out of the picture, and the audience would
see little but its roof below and to the right of the orchestra. As
a background it must have been replaced by a building of wood
or stone all traces of which have been obliterated by the Periclean
rebuilding. The chief reason for assuming such a building is
the evidence of the plays themselves. 2 This is the theatre in
which all the surviving plays of Aeschylus, the Ajax and Antigone
of Sophocles, the Alcestis of Euripides were produced, even if we
exclude the Trachiniae, Oedipus Tyrannus, Medea, and Hippolytus
as being possibly among the earliest productions in the Periclean
theatre. When every allowance is made for the dramatist's skill
in substituting verbal description for visual communication, the
evidence of the plays establishes (1) that communication between
the actors and the chorus was easy, (2) that a single practicable
door faced the audience, (3) that a high platform existed, (4) that

1 The theatre at Thorikos in Attica has no visible remains of a background

(cf. Dilke, B.S.A. xlv (1950), 25), but this does not exclude a temporary back-
ground. The arrangement assumed is found in the Cabeiran theatre.
2 Cf. G.T.P. pp. 7ff. On the ekhklema see now A. M. Dale, W.S. Ixix
(1956), 98. South Italian vase, New York 16. 140, G.T.P. no. A 26.
 ANCIENT GREEK THEATRE 499
a scene inside the building could be represented by a platform
(ekkyklema) which was rolled out through the doors. We need
not assume scene-painting if, as I have argued, Agatharchus is
better dated to the time of the Periclean theatre and painted
scenery for a reproduction of Aeschylus. Nor perhaps is the
crane (geranos or mechane) a necessary assumption for Aeschylus.
It is true that Pollux's example of the use of the geranos is drawn
from Aeschylus' Psychostasia, but the geranos may have been used
in a later reproduction like that of the Europe figured on a South
Italian vase of the early fourth century. The crane was certainly
used in the Periclean theatre. A temple-like wooden building
with a usable roof and a practicable door through which the
ekkyklema could be rolled is all that we need assume for the
pre-Periclean theatre. We know that the retaining wall was
curved and therefore this stage building, which stood between
the orchestra circle and the retaining wall, is unlikely to have had
any great lateral extension like the wall H behind T in the
Periclean theatre. And because the stage-building stood between
the retaining wall and the orchestra circle, it seems to me necessary
to accept Fiechter's reconstruction with J 3 as the orchestra and
SM 1 as the retaining wall rather than Dorpfeld's larger orchestra
circle which includes J 3 and SM 1 and coincides on the south
with the retaining wall.1
Of the contemporary vases which illustrate Aeschylean and
Sophoclean tragedy only one shows indirectly something of stage
production. Sir John Beazley 2 has published fragments of an
Attic hydria painted between 480 and 450, which shows a chorus
of Persians evincing astonishment as a king rises above a lighted
pyre. The fluteplayer in long robes and the chorus point to
tragedy and specifically to Aeschylus' Persians. But no one
would build a pyre on the stage and light it for a single play in
the long day's contest. The burning pyre therefore must be the
painter's imagination, his way of saying that the king was well
and truly buried. Why he chose this way of saying it we cannot
tell. Or did the producer perhaps use a pyre-like structure as a
shorthand for a tomb and was this rolled out on the
1 Cf. Schleif, A.A. liJ (1937), 26 f.; Dinsmoor, p. 312.
2 Corinth T 1144, Hesperia, xxiv (1955), 305, PI. 45.
32
 500 THE JOHN RYLANDS LIBRARY
A raised structure on the front of the e^fe/yema is the obvious
way of introducing Dareius' tomb and the only way of making
his rising above it convincing. A fragment of approximately the
same date in Boston l shows an armed hero rising above his tomb
mound (Achilles demanding the sacrifice of Polyxena ?) but
unfortunately there is nothing to connect it with tragedy.
To the Periclean stage-building belong the foundation T
(7.53 m. wide, 3.26 m. deep) and the slotted wall H (28 m. long)
behind it. In this theatre provision was undoubtedly made for
the crane and for scenery. 2 All that we can say about the crane
is that two holes at the back of T, the Western one measuring
70 X 70 cm. and the Eastern one 1 -25 m. X 70 cm., may have held
the mast of the crane and the winch which worked it; the jib of
the crane must have been attached to the mast above the roof of
the stage building since the actor could be raised from the interior
of the stage building, swung over, and landed on the stage and
vice versa. The scenery which Agatharchos painted was affixed
to the stage-building when required. But what shape was the
stage-building and what was the real function of T ?
Pickard-Cambridge's plan shows a flight of steps leading down
from T through the wall H into the long Stoa behind and in front
of T a wooden facade with wings touching the orchestra circle.
Essentially the same plan with more elaborate wings is given later
by Dinsmoor and with the front moved back level with T in
Fiechter's earlier " classical theatre ". 3 There is no evidence
either for the staircase or for the facade. Fiechter noted a tooling
of the stone of the wall of the Stoa where it runs along the back
of T and took this to be the threshold of a doorway which was
0.79 m. above the floor of the Stoa. Schleif 4 accepted this as a
doorway but saw that a further rise of 1 -38 m. was needed to reach
the level of the orchestra ; he therefore assumes a further set of
1 Museum of Fine Arts, 13. 169. Beazley, A.R.V., p. 188, no. 59.
2 Cf. G.T.P. pp. 12ff. for instances. Medea, Bellerophon, Perseus, and
Apollo in the Orestes are certain. I am now inclined to add the plays where a
supernatural appearance over the roof is announced by chorus or actor : Thetis
in Andromache, Iris and Lyssa in Hercules Furens, Athena in Ion, Dioscuri in
Electro, and the Muse in the post-Euripidean Rhesus.
3 P.C.T. p. 16, Fig. 7 ; Dinsmoor, pp. 326 f. Fiechter, iii, PI. 18.
4 Schleif, pp. 30 f.
 ANCIENT GREEK THEATRE 501
seven steps into T from the doorway. In fact we can probably
assume that a further three steps would be needed since the door
on to the stage (which is in any reconstruction the Northern end
of T or T's extension) is above the orchestra. Thus Schleif s
staircase would take up about two-thirds of the depth of T and
make T almost useless for anything else. Pickard-Cambridge's
plan conceals the awkward fact that the alleged threshold is not
symmetrical with T but starts well over to the East side and ends
shortly past the middle of T, and Schleif 's staircase only uses the
Western half of the threshold so as not to foul the big Eastern
hole in T. In fact the threshold is not a threshold at all; the
stones have been tooled to take some large object, picture,
relief, or inscription, on the back wall of the Stoa and have
nothing to do with T. T was backed by the unpierced Stoa
wall.
Seen from the audience T (or rather its superstructure)
occupies the position of the old temple in the theatre of Thespis
and the assumed scene building of the Aeschylean theatre. In
Attic and South Italian art 1 of the late fifth and first half of the
fourth century a doorway between columns and under a pediment
is shorthand for the theatre ; and this must have been a main
feature of T's superstructure, so that it looked not so very unlike
the old temple. The doors had to open, and the depth available
for their opening is limited by the arrangements for the mechane
to a maximum of 2'30 m. Thus a 4'60 m. opening is an absolute
maximum, if the doorposts were as near the front as possible.
But there is another limiting factor. For the Periclean theatre
the ekkuklema 2 is not merely assumed but is attested by Aristo-
phanes. By definition the e^fe/^/ema shows the inside of the
palace, house, or temple outside the front door. The doors must
therefore be set far enough back on T for the ekkyklema to appear
in front of them. The total available depth of T is 2*30 m., as
we have said, and perhaps 1 m. may be regarded as a minimum
distance for the ekkyklema to project beyond the door when it is
rolled out. If we take this figure, then the available depth behind
the door posts is 1'30 m., which gives a maximum opening of
1 Cf. below, p. 505, n. 3; p. 506, n. 3, 4.
2 Cf. A. M. Dale, W.S. Ixix (1956), 98 on Ach. 408, Thesm. 96.
502 THE JOHN RYLANDS LIBRARY
2*60 m. and a maximum width for the e^r/^/ema of about 2'55 m. 1
The available depth behind the doorposts is also the maximum
depth for the e^fe/^/ema, since, when the ekkykl&na is not in use,
its front must be flush with the doorpost and in fact appears as the
threshold of the doors. Thus starting from the possibilities of
T we arrive at 2'55 x 1 -30 as the maximum size for the e^y^/e/na ;
starting from the other end, the requirements of the plays, I think
we can see that a platform 2'55 m. wide by 1 '3 m. deep would
provide enough room for the required tableaux, for Euripides
and his properties in Aristophanes' Acharnians and for Herakles,
the bodies of his wife and children, and the broken column in
Euripides' Hercules. 2
Together the arrangements for the mechane and the e^r/^/ema
would occupy the whole of T behind the doors. T, as we have
seen, had no access from the back ; but it must have been possible
for an actor who had entered the palace to change his clothes and
come on through the parodoi in another part, and it must have
been possible to reach the roof.3 Changing rooms, access to the
parodoi, and access to the roof must therefore have been contained
in the wooden building erected on the slotted wall H which
stretches on either side of T, and this building must also have
had openings leading on to T.4 Our only direct evidence for this
building is the wall H, which is 28 m. long with slots at 2'56 m.
intervals, five on each side of T. The door posts on T according
to our assumption stand 2*26 m. in front of the slotted wall.
Presumably the front wall of the lateral building was not in front
of the door posts on T; therefore its maximum depth cannot
have been more than 2*26 m. The only building which affords

1 Width of doorways vary greatly : e.g. Oropos 1 m., Priene 1 *6 m., Sikyon
2*5 m.; Syracuse and Pergamon total width between wooden supports, 3'5 and
3'7 m. respectively.
2 G.T.P. p. 9 ; Euripides, H.F. 1081 ff.
3 E.g. Antigone and the old man in the Phoenissae ; gods when they do not
use the crane ; Bdelykleon at the beginning of the Wasps.
4 Demosthenes, Meidias, 17, and his commentators (quoted P.C.T. p. 24)
are perfectly consistent with this. His chorus wanted to change in the changing
rooms and enter the orchestra (whether through the central doors or through
the exits giving on to the parodoi). Meidias nailed up the paros^ema, the lateral
extensions of the skene proper, which was the superstructure of T.
 ANCIENT GREEK THEATRE 503
any sort of parallel is the stage building of the theatre in Corinth. 1
This stage building also belonged to the very late fifth century
and was built of wood. It was 21 m. long and 4 m. deep, and
front and back had eight slots for wooden posts set at 3 m.
intervals. The remains at Corinth give no indication of height.
We have perhaps a pointer in Xenophon.2 He is discussing the
ox teams used by Cyrus to move his fortresses on wheels. These
fortresses were " 3 fathoms high with their wheels " and " their
beams had the thickness (of the beams) of the tragic skene ". If
Xenophon is thinking of the upright beams (or columns) of H
and T, then they may have had a height of about 5 m., which is
the height assumed by Fiechter for the columns of the old temple.
It is also the length of the beam provided for the logeion in the
theatre at Delos. The other theatres of which we know the
height are much later and the Hellenistic proskenion beneath the
stage may well have been lower. But the height of the fourth
century (?) skene at Epidaurus is probably given by the gateways to
the parodoi and this was 3'53 m. for a skene 19'50 m. long ; 5 m.
does not therefore seem unreasonable for the Athenian skene
which was 28 m. long.
The Corinth building certainly had no projecting wings, and
at Athens the projecting wings which appear on the reconstruc-
tions are a pure assumption, an assumption which has become
less likely since the publication of the Corinthian parallel. The
1 Corinth : R. Stilwell, Corinth, ii. 15 ff. I suspect that the building was
originally 21 m. long and 4 m. deep, set symmetrically with the orchestra. Of
other wooden buildings of which traces survive Syracuse is too doubtful to use
and it is quite unclear to what date the 3*5 m. openings should be referred;
Pergamon presumably from the beginning had the shape of a late Hellenistic
stage-building. The slots for the beams in H were from 0'40 to 0-49 m. wide
and from 0-35 to 0-46 m. deep ; Fiechter, i. 15, compared the 0-40 m. slots in
Pergamon ; the sockets for the front row of beams in Corinth were 0-20 X 0-30 m.
and for the back row 0-50 X 0*60 m.
2 Cyropedia, vi. 1, 52-4. Fiechter, iii. 66. Beam for the logeion in Delos :
Vallois, Architecture Hellenique a Delos, p. 237 ; this may be a horizontal rather
than a vertical beam, cf. G.T.P. p. 147. The cross beams of the naval arsenal
at the Peiraeus were 075 X 075 X 6 m. (Robertson, Greek and Roman Archi-
tecture, p. 182). Ptolemy's Banqueting Tent had two wooden columns 22-5 m.
high. Preserved heights of proskenion in Hellenistic theatres : Oropos, 2-68 m.
(length 12-30 m.); Priene, 2723 m. (length 21 m.) ; Sikyon, 3-10 m. (length
2375 m.).
504 THE JOHN RYLANDS LIBRARY
main functions of the building were to house changing rooms, to
provide communications for actors between the palace entrance,
the outside entrances (parodoi), and, when necessary, the roof,
and lastly to provide a framework for the display of scenery.
Dinsmoor 1 assumes, as I do, a line of wooden vertical beams
parallel to and in front of the slots but then adds at each end
further vertical beams in front of them to provide the wings.
Such projecting wings would of course have no reflection in the
existent slots of the back wall and cannot be refuted on these
grounds. The arguments against them are first the analogy with
the wooden building at Corinth, secondly the use of the central
doorway by itself as a shorthand for the theatre in vase-painting,
and thirdly the demands of the plays themselves. The plays
demand communication between the orchestra and the parodoi
and the door of the stage-building but nowhere suggest the
desirability of a long narrow area framed by wings. Neither
tragedy nor comedy gives evidence for more than one door in the
stage-building itself 2 (and the existence of such doors would
spread the action along the back wall and thereby make the
framing by projecting wings desirable), and there is no reason to
suppose that in fundamentals the theatre of Euripides and Aristo-
phanes looked any different from the theatre of Aeschylus and
the theatre of Thespis : the fundamentals are a circular dancing
floor dominated by a temple (or palace).
In the Periclean theatre (according to the reconstruction pro-
posed above) the central doorway was 2'60 m. wide and gave on
to a platform 7'53 m. wide and 1 m. deep. This area was the
stage in the modern sense, the normal location for the actors
(although at times they remained on the orchestra level), and was
connected with the orchestra by a short flight of steps, which were
usually in the centre but could be placed at the sides when the
ekkykjema was pushed forward for the whole play, as in Sophocles'
Philoctetes ; there is good evidence for such steps in vase-
paintings of comedy, both Attic and South Italian. 3 On either
1 PP.326f.
2 On tragedy cf. G.T.P. p. 10. On comedy I accept the conclusions of
A. M. Dale, J.H.S. Ixxvii (1957), 207.
3 On the Philoctetes cf. A. M. Dale, W.S. Ixix (1956), 104. Steps in comic
scenes on vases : e.g. G.T.P. no. B 1 (Attic), 32, 43 (S. Italian), 67 (Sicilian).
 ANCIENT GREEK THEATRE 505
side of the central door was a panel roughly 2'30 m. wide which
connected the central door with the lateral building H stretching
out beyond T, masked the access to the roof and wings, could
carry scenery and when necessary contain a window. 1 It is
probable that these panels continued the line of the front wall of
the lateral building but whether this whole front line was level
with the central door or was set slightly back I see no means of
determining, but it seems to me more likely that the central door
was set a little forward and that this emphasized both its own
importance and its connection with the orchestra. If the door
columns were thus set forward, the junction with the lateral
panels would be masked by columns and the four columns would
give rise to the stylization as an aedicula which is commonly used
on South Italian vases.2
We have no direct evidence as to whether the columns were
of wood like the rest of the stage building or of stone. It is
probable that the breccia foundation carried a stone floor but
this does not necessarily mean that the columns and entablature
were of stone. The analogy of South Italian theatre buildings
suggests wood and on the Attic Iphigeneia vase 3 the building is
certainly wooden although Euripides imagines an elaborate
temple with high walls, triglyphs, gilded cornice, pillars, bronze
doors, and marble base for the statue, like any fifth-century
marble temple. On the Iphigeneia vase the temple is poised on
two high steps and the painter may have been thinking of the
high threshold formed by the ekkyklema when it was pushed back
and of the stage in front of the door. On the South Italian vases 4
1 On windows in Wasps, Ecclesiazusae, Clouds cf. A. M. Dale, J.H.S., Ixxvii
(1957), 205, 208, 210. Windows on South Italian vases : G.T.P., no. B 32,
37,60.
2 Yet another possibility is that doors and side panels were in line and that
this whole front on T was set forward of the line of the front wall of H and was
topped by a single pediment. This would preserve the proportions of the Old
Temple but would not agree either with the Attic Iphigeneia vase (see p. 505,
n. 3) or with the South Italian vases which show either a pedimented doorway
or an aedicula (see p. 505, n. 4; p. 506, n. 3, 4).
3 G. T.P.no. A 8.
4 pediment: G. T.P. nos. A 26, 36, 41, 42, 43, 44 ; B 32, 33, 43. threshold :
G.T.P. nos. A 26, 36, 39; B 34, 37, 61. double step : G.T.P. nos. A 36, 42
43; B41.
506 THE JOHN RYLANDS LIBRARY
which have scenes inspired by tragedy or actual representations
of comedy, the pediment is often shown, the threshold is some-
times shown, and occasionally in the tragic scenes a double step
is shown as on the Iphigeneia vase. Where the comic scenes
seem to show a plain door without a pediment,1 this may be
illusory as in every case the top of the door coincides with the top
of the vase. Three vases deserve special mention here. An
Apulian vase 2 of the early fourth century shows a comic scene in
which an old man is welcomed by an old woman ; the setting is
exactly what I have assumed for the Periclean theatre to the
left of the old man a column, which could cover the junction
between the lateral building and T, and to the right of the old
woman double doors surmounted by a pediment. A rather later
Apulian vase 3 gives a version of the letter scene from the Iphi-
geneia in Tauris in which the aedicula stands on two steps but the
lower step is extended to support the altar: in our terms the
altar is on the stage at the side of the central door. An Apulian
vase 4 of 330/20 B.C. shows Niobe standing in an aedicula with a
high decorated base, which represents the tomb of Niobe's
children. If the aedicula represents the central door then the
high decorated base is inspired by the e^r/^/ema : in the original
production of Aeschylus the tomb of Niobe's children was
probably represented on the e^t/^/ema like the tomb of Dareios
in the Persae.
In the Periclean theatre the fundamentals are a circular
dancing floor dominated by a single door with a stage in front of
it. The last play for which we can say this arrangement was
necessary is Aristophanes' Ecclesiazusae which was produced in
391 B.C. There the whole effect of the scene between the girl,
the hag, and the young man depends on the two women having
separate windows and a common front door.5 The Dyskplos by
Menander is our first and only complete text of a play written for
the theatre of Lycurgus. It was produced in 316 B.C. when the
l GT.P.nos.B34,37,61.
2 Harvard University, McDaniell Collection, Hirsch Sale Catalogue, no. 30,
PI. 16. See now A. D. Trendall, Phlyax Vases, B./.C.S., Supplt. 8, no. 23.
3 Apulian vase, Moscow, P.C.T. Fig. 16.
4 GT.P.no.A44.
5 Cf. A. M. Dale, J.H.S. Ixxvii (1957), 208.
 ANCIENT GREEK THEATRE 507
new theatre was about fourteen years old. The setting is made
clear in the prologue which is spoken by the god Pan : " The
shrine of the Nymphs from which I make my entry belongs to
the people of Phylae . . . Knemon (the hero) occupies the land
on the right ... a little estate nearby here belongs to " Gorgias,
who is Knemon's stepson. The set therefore must represent a
central shrine of the Nymphs with Knemon's house on the left
(for the audience) and Gorgias' house on the right. The audi-
ence sees an elaborate central door (the shrine here, the palace
or temple in tragedy), and a simpler door at each side (the two
houses in comedy). Behind this facade an actor could pass
swiftly between the three doors and the parodoi.
Those needs are made particularly clear by the first act. Let
us call the three actors A, B, C. A speaks the prologue as Pan ;
he makes his entrance and his exit through the Nymphaion
(central door). He goes off when he sees the young lover
Sostratos (B) and his parasite Chaireas (C) appearing up the right
parodos. After 30 lines they are joined by the slave Pyrrhias who
enters up the left parodos ; actor A has therefore changed his
mask and clothes and reappears up the left parodos as Pyrrhias.
After another 60 lines Chaireas (C) goes off by the right parodos,
changes mask and clothes, and reappears up the left parodos as
Knemon. As he appears, Pyrrhias (A) retires into the Nym-
phaion ; later Knemon (C) goes into his house on the left, leaving
Sostratos (B) on the stage. Actor A comes out of Knemon's
house as Knemon's daughter, having changed his mask and
clothes. The daughter wants to fetch water from the Nym-
phaion, and Sostratos goes to get it for her. She goes back into
her father's house (left). As Sostratos takes her the water, actor
C, who has changed mask and clothes and crossed over, comes
out of Gorgias' house (right) as Daos, Gorgias' slave, and com-
ments on the proceedings. After his last words from Knemon's
door as Knemon's daughter, actor A does an extremely quick
change and reappears as Pyrrhias from the Nymphaion.
The remains of the theatre of Lycurgus show the foundations
of a front wall with projecting wings and an open hall behind.
It is always assumed that the front wall was pierced by three
doors of which the central and more elaborate door was used for
508 THE JOHN RYLANDS LIBRARY
tragedy; comedy could use all three doors when necessary or
only the two side doors when two houses were all that was needed.1
The Dyskolos shows the Tightness of this assumption. All three
doors are needed, and the actors have to be able to change masks
and clothes and pass from door to door or into the parodoi in the
minimum of time. The projecting wings perform the double
function of framing the long stage and of covering a short cut
from the hall behind the facade to the parodoi. This complete
change in the character of the theatre corresponds to two essential
changes in the nature of comedy. The first is that comedy now
requires a realistic scene with either two houses or two houses
with a third building between them. In our scene Knemon's
house, the Nymphaion, and Gorgias' house must have their
separate doors : Daos must be able to enter from Gorgias' house
and watch Sostratos come out of the Nymphaion, walk across to
Knemon's house and give the girl her pot of water as she stands
in the doorway. Later the old woman Simike must be able to
enter from Knemon's house to announce that he has fallen down
the well at the same moment as the cook Sikon enters from the
Nymphaion to ask for silence for the sacrifice.
The other change is the exclusion of the chorus. In the
Dyskolos the chorus enter after the first act and Daos says : "I
had better get out of the way ; some drunks are coming." The
manuscripts then note XOPOY, and we assume that the interval
between the acts was marked by a song which was not specially
composed for the play. There is no reason to suppose that the
chorus remained in the orchestra during the action of the play;
they may only have come on for the intervals. Daos' preliminary
remark is the only link between actors and chorus. Dialogue
sung or spoken between actors and chorus and commentary by
the chorus leader on the action have alike vanished. In Old
Comedy and classical tragedy the action on the stage was watched
1 Only the foundations of the front wall of the Lycurgan stage-building exist ;
so no trace of doors remains. No other surviving theatre helps. Dorpfeld (in
Dorpfeld-Reisch, p. 124) attributes the three doors at Epidaurus to the later
Hellenistic rebuilding. The first stage-building of Eretria (which to judge from
its masonry is not earlier than the fourth century) gives no evidence for three
doors : the openings which remain belong to three rooms at the back of the
stage buildings.
 ANCIENT GREEK THEATRE 509
by the chorus and annotated by its leader, and this relationship
was expressed in the design of the Periclean theatre. Now the
action takes place on a long, framed stage and the orchestra is
empty. Some idea of the appearance of the Lycurgan stage-
building is given by a South Italian terracotta l of the late fourth
or very early third century B.C.; it evidently represents a South
Italian theatre remodelled on the lines of the Lycurgan theatre.
It is only a logical step further to remove the action to the roof
and give the picture frame a bottom as well as sides and top ; and
this step was taken in the second century.
The theatre of Lycurgus had been in existence about ten years
when Menander started to write. He was therefore exploiting
existing conditions. The new design of the theatre of Lycurgus
must have met an existing demand and, of course, we do not know
that it did not have a forerunner in wood. There is perhaps a
slight indication of this on a Lucanian aryballos 2 of 340/30 B.C.,
on which the mask of Electra is posed above a long stage supported
by crossbeams. This stage is quite unlike the stages on the
rather earlier comic vases which are supported on columns and
approached by flights of steps : it might, however, very well be
the long low stage of a three-door theatre with wings. If my
conjecture 3 that the Menaechmi of Plautus is an adaptation of
Alexis' Adelphoi, which was produced about 340 B.C., is right,
then plays which demanded two doors were being written about
340 ; confusion of characters in such plays is quite sufficient
without confusion of localities as well. In general it seems to me
likely that the demand grew with the comedies of intrigue, which
were already being produced before the middle of the fourth
century, and that it may have been first met by the provision of
side doors in the lateral building H of the Periclean theatre. But
this is conjecture, and we only reach comparative certainty with
the theatre of Lycurgus and Menander's Dyskolos.
1 G.T.P. no. A 49.
2 British Museum Quarterly, 29 (1959), p. 100.
3 Cf. Studies in Later Greek Comedy, p. 72.