2. Aeschylus'Theatre to stage scenes in which the pagos okhthos, with its altar or tomb, is entirely surrounded by a chorus; such a staging is explicitly attested for an unknown play of Aeschylus (fr. 379)and is highly probable for the binding song of Eum. 307-96 And it enables one, as Rehm shows, to posit a continuity between the staging of suppliant scenes in Aeschylus'time and in the later fifth century, when choruses are still found supplicating at altars(as in Euripides Children of heracles and Suppliant Women However, a problem for any such assumption, at least in the ' early Aeschylean theatre, is posed by another feature of theatrical topography which is called for by the script of Aeschylus'oldest surviving play In Th Persians the ghost of Darius, called up from the underworld, emerges, so the text invites us to suppose from the top(659)of his burial-mound, and his own words make it clear that he has come from under the earth(683 685, 697)and returns thither at the end of the scene( 839). It is of course possible in principle that the actor simply enters from the side and ascends the mound, but this is unlikely: the heroic dead were felt to be present in their tombs - that indeed is why Darius is summoned, and offerings poured to him, at this particular place -and he ought not therefore to enter as if arriving from somewhere else. One might then be tempted to suppose that the ghost rose through a hole in the soil of the mound, the performer having got there by an underground passage whose other entrance lay somewhere at or beyond the edge of the orkhestra. Such passages, built in stone, existed in some later theatres under the name of"Charon's steps but there is no evidence at all that one existed in the athenian theatre at any time in its history, let alone in the fifth century. Hence the pagos/ okhthos from which Darius seemed to emerge cannot have been in the centre of the orkhestra The only solution that I can see is one proposed by bieber 1961 and revived by Scullion 1994: 70n. 9: that the tomb-mound was built up against the back wall of the orkhestra terrace(which must, for safety reasons, have stood a metre or so above the level of the orkhestra itself)and that the ghost climbed a ladder placed behind it. This then was doubtless the pagos/okhthos in Seven and Supp as well(cf Sandin 2003: 15-19). Once the back of the orkhestra was occupied by the scene-building, the pagos/okhthos must have been moved somewhere else-presumably then to the centre of the orkhestra. It is then necessary to assume that Aesch fr. 379. where a chorus of women is instructed to 'stand all round this alta in] an endless i.e. circular band, and pray, must come from a play produced after the creation of the scene- building. To the environment thus far described the oresteia adds a scene build- ing. I say 'adds because nothing seems to have been subtracted: all the features of the earlier theatre remain available. The pagos / okhthos is shifted to the centre of the orkhestra, around the thymele, but in its new position it serves much the same functions. The elevation represents a burial-mound in Cho. as it did in Pers., and Orestes and Electra beg
to stage scenes in which the pagos/okhthos, with its altar or tomb, is entirely surrounded by a chorus; such a staging is explicitly attested for an unknown play of Aeschylus (fr. 379) and is highly probable for the ‘binding song’ of Eum. 307-96. And it enables one, as Rehm shows, to posit a continuity between the staging of suppliant scenes in Aeschylus’ time and in the later fifth century, when choruses are still found supplicating at altars (as in Euripides’ Children of Heracles and Suppliant Women). However, a problem for any such assumption, at least in the ‘early’ Aeschylean theatre, is posed by another feature of theatrical topography which is called for by the script of Aeschylus’ oldest surviving play. In The Persians the ghost of Darius, called up from the underworld, emerges, so the text invites us to suppose, from the top (659) of his burial-mound, and his own words make it clear that he has come from under the earth (683, 685, 697) and returns thither at the end of the scene (839). It is of course possible in principle that the actor simply enters from the side and ascends the mound, but this is unlikely: the heroic dead were felt to be present in their tombs – that indeed is why Darius is summoned, and offerings poured to him, at this particular place – and he ought not therefore to enter as if arriving from somewhere else. One might then be tempted to suppose that the ghost rose through a hole in the soil of the mound, the performer having got there by an underground passage whose other entrance lay somewhere at or beyond the edge of the orkhêstra. Such passages, built in stone, existed in some later theatres under the name of ‘Charon’s steps’; but there is no evidence at all that one existed in the Athenian theatre at any time in its history, let alone in the fifth century. Hence the pagos/ okhthos from which Darius seemed to emerge cannot have been in the centre of the orkhêstra. The only solution that I can see is one proposed by Bieber 1961 and revived by Scullion 1994: 70n.9: that the tomb-mound was built up against the back wall of the orkhêstra terrace (which must, for safety reasons, have stood a metre or so above the level of the orkhêstra itself) and that the ghost climbed a ladder placed behind it. This then was doubtless the pagos/okhthos in Seven and Supp. as well (cf. Sandin 2003: 15-19). Once the back of the orkhêstra was occupied by the scene-building, the pagos/okhthos must have been moved somewhere else – presumably then to the centre of the orkhêstra. 7 It is then necessary to assume that Aesch. fr. 379, where a chorus of women is instructed to ‘stand all round this altar [in] an endless [i.e. circular] band, and pray’, must come from a play produced after the creation of the scene-building. To the environment thus far described the Oresteia adds a scene-building. I say ‘adds’ because nothing seems to have been subtracted: all the features of the earlier theatre remain available. The pagos/okhthos is shifted to the centre of the orkhêstra, around the thymelê, but in its new position it serves much the same functions. The elevation represents a burial-mound in Cho. as it did in Pers., and Orestes and Electra beg 2. Aeschylus’ Theatre 21
Aeschylean Tragedy Agamemnons ghost to appear to them there( cho. 489-96); their prayers are not answered, and indeed it may no longer have been possible for a ghost actually to appear from underground in the way darius'ghost did In Eum. to be sure, a ghost does appear that of clytaemestra, but she does not emerge from her tomb(for the action is set at Delphi, not at argos where she was buried) and she may simply have come on stage along one of the eisodoi. Later in Eum.(235-489) the mound may function as a sanctuary and place of supplication, as in Supp when Orestes appeals to Athena and the erinyes sing their binding song around him; and in Ag it may be towards the thymele that the herald turns when he addresses the 'assembled gods of Argos (513) whom Danaus and Pelasgus had designated by the same expression in Supp (189, 242) The features of the scene- building(skene), of which we learn from the resteia, are as follows The building was capable of representing either a palace or a temple. It had one main door, presumably in the centre of its facade; and analysis of Cho. 875ff. strongly suggests that there was also at least one other door(see $7.6.6). If so, there is likely, for symmetry, to have been a third door on the opposite side. In addition, the roof was accessible to the actors(Ag. 1-39 ). All this is consistent with, though it does not of course prove the hypothesis that the scene- building of 458 was similar in all essentials to that of the 420s and later. which is known to us principally from Euripides and aristophanes There has been a long and sterile argument over whether in aeschylus time(or indeed thereafter) there was an elevated platform raised stage ') between the scene-building and the orkhestra. There is, in my opinion, evidence in the play-texts(and also in art) for the existence of such a platform from about 422 onwards - though even then, and a fortiori earlier, there was easy access in both directions between the orkhestra and the house. Earlier than that there is no clear evidence, and all we can say is that if there was a platform it can hardly have been raised by more than a couple of steps. It is perhaps likeliest that there always was a platform As we have seen, there was an elevated place in the acting area before there was a scene- building, so that dramatists and audience were used to the idea of a significant contrast of levels; moreover the scene- building was behind the orkhestra and further from the audience. and a little added height would increase the visibility and prominence of those performers who were near the building. This effect could be put to good use in Agamemnon, where time and again Clytaemestra, coming out of the palace, would visually dominate the elders or the herald or Aegisthus below her in the orkhestra In the later fifth century two'special-effects' devices were used in the theatre the flying-machine(methane or krade) and the so-called ekkyk lema. The flying-machine, essentially a crane, was used for airborne entries, such as those of Bellerophon in Euripides' play of that name and Perseus in his Andromeda (parodied by aristophanes in Peace and Thes-
Agamemnon’s ghost to appear to them there (Cho. 489-96); their prayers are not answered, and indeed it may no longer have been possible for a ghost actually to appear from ‘underground’ in the way Darius’ ghost did. In Eum., to be sure, a ghost does appear, that of Clytaemestra, but she does not emerge from her tomb (for the action is set at Delphi, not at Argos where she was buried) and she may simply have come on stage along one of the eisodoi. 8 Later in Eum. (235-489) the mound may function as a sanctuary and place of supplication, as in Supp., when Orestes appeals to Athena and the Erinyes sing their ‘binding song’ around him; and in Ag. it may be towards the thymelê that the Herald turns when he addresses the ‘assembled gods’ of Argos (513) whom Danaus and Pelasgus had designated by the same expression in Supp. (189, 242). The features of the scene-building (skênê), of which we learn from the Oresteia, are as follows. The building was capable of representing either a palace or a temple. It had one main door, presumably in the centre of its façade; and analysis of Cho. 875ff. strongly suggests that there was also at least one other door (see §7.6.6). If so, there is likely, for symmetry, to have been a third door on the opposite side.9 In addition, the roof was accessible to the actors (Ag. 1-39). All this is consistent with, though it does not of course prove, the hypothesis that the scene-building of 458 was similar in all essentials to that of the 420s and later, which is known to us principally from Euripides and Aristophanes. There has been a long and sterile argument over whether in Aeschylus’ time (or indeed thereafter) there was an elevated platform (‘raised stage’) between the scene-building and the orkhêstra. There is, in my opinion, evidence in the play-texts (and also in art) for the existence of such a platform from about 422 onwards10 – though even then, and a fortiori earlier, there was easy access in both directions between the orkhêstra and the house. Earlier than that there is no clear evidence, and all we can say is that if there was a platform it can hardly have been raised by more than a couple of steps. It is perhaps likeliest that there always was a platform. As we have seen, there was an elevated place in the acting area before there was a scene-building, so that dramatists and audience were used to the idea of a significant contrast of levels; moreover the scene-building was behind the orkhêstra and further from the audience, and a little added height would increase the visibility and prominence of those performers who were near the building. This effect could be put to good use in Agamemnon, where time and again Clytaemestra, coming out of the palace, would visually dominate the Elders or the Herald or Aegisthus below her in the orkhêstra. In the later fifth century two ‘special-effects’ devices were used in the theatre, the flying-machine (mêkhanê or kradê) and the so-called ekkyklêma. 11 The flying-machine, essentially a crane, was used for airborne entries, such as those of Bellerophon in Euripides’ play of that name and Perseus in his Andromeda (parodied by Aristophanes in Peace and ThesAeschylean Tragedy 22
2. Aeschylus'Theatre mophoriazusae respectively), and for such scenes as the escape of Medea by flying chariot; later also for the interventions of what, long afterwards, came to be called dei ex machina. The ekkyklema, a platform on wheel which could be rolled out of the central doorof the scene-building, was used n tragedy mainly for the revelation of tableau-like interior scenes, in comedy also for a variety of other purposes. It is disputed whether these devices were already in use at the end of Aeschylus'career, Taplin 1977: 442-7 in particular, though hesitantly denied him both of them. Yet the scenes for which the use of the ekkyklema has generally been posited in the Oresteia- Clytaemestra over the bodies of Agamemnon and Cassandra (Ag. 1372ff ) Orestes over the bodies of Clytaemestra and aegisthus(Cho. 973ff ) Orestes beset by the erinyes at Delphi (Eum. 64ff )-seem precisely parallel to scenes in later tragedy where its use is not now seriously disputed. Taplin argued that in these later scenes the impending use of the device is carefully signalled by what seems to be a well-established convention, and there are no such signals in the Oresteia scenes; but this may indicate merely that in 458 the convention in question was not yet established. The object of the conven tional cues was to make it clear to the audience that the tableau they were about to see was meant to be inside the background building. In the Oresteia this purpose is served by other, less stereotyped methods: in Ag by the bathtub and by clytaemestra's statement that i stand where I struck(1379); in Cho, more subtly, by the almost perfect repetition of the stage-picture from Ag. in Eum. by the fact that the scene portrayed is precisely that which the Pythia, a moment previously, reported having seen inside the temple(see further $7.6.2) Regarding the mekhane there is considerably more room for uncer- tainty. It is suspicious that the two plays in the Aeschylean corpus where its use seems virtually certain- Prometheus Bound (see $8.3) and The Weighing of souls(see pp 7-8 above)-are both, for quite other reasons, of doubtful authenticity. There is some case for giving Athena a flying entrance in Eumenides(397), but it is far from conclusive. While, there ore, it seems highly likely that the ekkyklema was introduced at the same time as the scene- building in the late 460s the mekhane may be of hat later origi 2.2. Performers and properties The performers of Aeschylean tragedy comprised a chorus, two or three speaking actors (hypokritai), an unlimited number of mutes, and (by no means least)a piper to accompany the songs. The chorus consisted of twelve members (khoreutai): Sophocles, according to his ancient life. increased the number to fifteen, but the debate among twelve speakers in Ag. 1348-71 shows that this happened after 458. The khoreutai were ordinary citizens, selected and trained by the khorgos or sponsor of the
mophoriazusae respectively), and for such scenes as the escape of Medea by flying chariot; later also for the interventions of what, long afterwards, came to be called dei ex machina. The ekkyklêma, a platform on wheels which could be rolled out of the central door of the scene-building, was used in tragedy mainly for the revelation of tableau-like interior scenes, in comedy also for a variety of other purposes. It is disputed whether these devices were already in use at the end of Aeschylus’ career; Taplin 1977: 442-7 in particular, though hesitantly, denied him both of them. Yet the scenes for which the use of the ekkyklêma has generally been posited in the Oresteia – Clytaemestra over the bodies of Agamemnon and Cassandra (Ag. 1372ff.), Orestes over the bodies of Clytaemestra and Aegisthus (Cho. 973ff.), Orestes beset by the Erinyes at Delphi (Eum. 64ff.) – seem precisely parallel to scenes in later tragedy where its use is not now seriously disputed. Taplin argued that in these later scenes the impending use of the device is carefully signalled by what seems to be a well-established convention,12 and there are no such signals in the Oresteia scenes; but this may indicate merely that in 458 the convention in question was not yet established. The object of the conventional cues was to make it clear to the audience that the tableau they were about to see was meant to be inside the background building. In the Oresteia this purpose is served by other, less stereotyped methods: in Ag. by the bathtub and by Clytaemestra’s statement that ‘I stand where I struck’ (1379); in Cho., more subtly, by the almost perfect repetition of the stage-picture from Ag.; in Eum. by the fact that the scene portrayed is precisely that which the Pythia, a moment previously, reported having seen inside the temple (see further §7.6.2). Regarding the mêkhanê there is considerably more room for uncertainty. It is suspicious that the two plays in the Aeschylean corpus where its use seems virtually certain – Prometheus Bound (see §8.3) and The Weighing of Souls (see pp. 7-8 above) – are both, for quite other reasons, of doubtful authenticity. There is some case for giving Athena a flying entrance in Eumenides (397), but it is far from conclusive. While, therefore, it seems highly likely that the ekkyklêma was introduced at the same time as the scene-building, in the late 460s, the mêkhanê may be of somewhat later origin.13 2.2. Performers and properties The performers of Aeschylean tragedy comprised a chorus, two or three speaking actors (hypokritai), an unlimited number of mutes, and (by no means least) a piper to accompany the songs. The chorus consisted of twelve members (khoreutai): Sophocles, according to his ancient Life, increased the number to fifteen, but the debate among twelve speakers in Ag. 1348-71 shows that this happened after 458. The khoreutai were ordinary citizens, selected and trained by the khorêgos or sponsor of the 2. Aeschylus’ Theatre 23
Aeschylean Tragedy production in conjunction with the author-director and possibly with a specialist trainer; they were probably always young men, aged between eighteen and thirty Since music and singing were an important part of traditional greek education, and since every year 500 boys under eighteen performed in dithyrambic contests at the City Dionysia alone and many more at other festivals, there will always have been a substantial pool of candidates with a basic competence in the required skills. We are told that unlike the circular,choruses of dithyramb, tragic choruses danced in a rectangular formation, though this should not be regarded as a rigid rule certain ritualistic scenes focused on the thymele (above, p. 22)cry out for a circular formation, and in others(above all the entry of the chorus in Seven)the dancing may have been designedly chaotic. By a convention which probably went back to the earliest days of drama, the leader of the orus(the hegemon or koryphaios) could represent the whole group in dialogue with the actors; he often has quite an important speaking part (in Ag. he speaks 148 lines, more than anyone else in the play except Clytae mestra), though he only very rarely delivers a long set speech. Since in addition the chorus-leader was a full participant in all the songs and dances of the chorus as a whole, and must in practice have found himself largely responsible for their moment-to-moment direction on stage, he bore arguably a heavier burden than any of his fellow-performers -and seems not to have gained as much honour by it as he deserved: authors, khoregoi, pipers, leading actors, might all win fame, but in the whole classical period we do not know the name of a single tragic koryphaios 6 We may well, though, suspect that a remark on the chorus -leader's impor tance, made by the fourth-century orator(and khorgos) Demosthenes deserves a fair degree of credence You know, I am sure that if the leader(hegemon)is taken away, the rest of the chorus is done for (Demosthenes, Against Meidias 60) In some plays, both by aeschylus and by later dramatists, there is more than one chorus. In such cases one chorus is always clearly the principal one: it enters early in the play, it remains present continuously or almost continuously until the end, it sings repeatedly, and its leader has a speaking part. a subsidiary chorus will normally be present, and sing, in one scene only, and(at least in tragedy) its leader will not be given word to speak. two surviving plays of Aeschylus have subsidiary cho Eum. a chorus of the female temple-staff of Athena take part in the final procession and sing the concluding lyrics. In Supp the concluding lyrics take the form of a dialogue between the main chorus of the daughters of Danaus and a subsidiary chorus whose identity is disputed(see $6. 1); it is also possible that earlier in the play(836ff )the Egyptian herald is accompanied by a chorus of Egyptian soldiers We have no firm evidence for any conventions regarding the size of such subsidiary choruses 24
production in conjunction with the author-director and possibly with a specialist trainer; they were probably always young men, aged between eighteen and thirty.14 Since music and singing were an important part of traditional Greek education, and since every year 500 boys under eighteen performed in dithyrambic contests at the City Dionysia alone and many more at other festivals, there will always have been a substantial pool of candidates with a basic competence in the required skills. We are told that unlike the ‘circular’ choruses of dithyramb, tragic choruses danced in a rectangular formation, though this should not be regarded as a rigid rule: certain ritualistic scenes focused on the thymelê (above, p. 22) cry out for a circular formation, and in others (above all the entry of the chorus in Seven) the dancing may have been designedly chaotic. By a convention which probably went back to the earliest days of drama, the leader of the chorus (the hêgemôn or koryphaios15) could represent the whole group in dialogue with the actors; he often has quite an important speaking part (in Ag. he speaks 148 lines, more than anyone else in the play except Clytaemestra), though he only very rarely delivers a long set speech. Since in addition the chorus-leader was a full participant in all the songs and dances of the chorus as a whole, and must in practice have found himself largely responsible for their moment-to-moment direction on stage, he bore arguably a heavier burden than any of his fellow-performers – and seems not to have gained as much honour by it as he deserved: authors, khorêgoi, pipers, leading actors, might all win fame, but in the whole classical period we do not know the name of a single tragic koryphaios. 16 We may well, though, suspect that a remark on the chorus-leader’s importance, made by the fourth-century orator (and khorêgos) Demosthenes, deserves a fair degree of credence: You know, I am sure, that if the leader (hêgemôn) is taken away, the rest of the chorus is done for (Demosthenes, Against Meidias 60). In some plays, both by Aeschylus and by later dramatists, there is more than one chorus. In such cases one chorus is always clearly the principal one: it enters early in the play, it remains present continuously or almost continuously until the end, it sings repeatedly, and its leader has a speaking part. A subsidiary chorus will normally be present, and sing, in one scene only, and (at least in tragedy) its leader will not be given words to speak. Two surviving plays of Aeschylus have subsidiary choruses. In Eum. a chorus of the female temple-staff of Athena take part in the final procession and sing the concluding lyrics. In Supp. the concluding lyrics take the form of a dialogue between the main chorus of the daughters of Danaus and a subsidiary chorus whose identity is disputed (see §6.1); it is also possible that earlier in the play (836ff.) the Egyptian herald is accompanied by a chorus of Egyptian soldiers. We have no firm evidence for any conventions regarding the size of such subsidiary choruses. Aeschylean Tragedy 24
2. Aeschylus'Theatre The tragedies of the sixth and early fifth centuries had only one speak ing actor other than the chorus - leader, and Aeschylus himself is said to have added a second; a third actor was introduced, either by aeschylus or by Sophocles, in the late 460s, and is employed in the Oresteia. There are moments in that trilogy when one feels Aeschylus would have liked to go even further: many have suggested that he actually uses a fourth actor at one moment of Choephoroi (but see $7.6.6), and in Eumenides the chorus- leader virtually functions as a fourth actor in the trial scene Yet at the same time he remains capable of writing brilliant drama for a single actor and chorus, as in the greater part of Agamemnon. It had always been possible for the same actor to take different parts in successive scenes, but in Pers, two or three in the genuine portion of Seven, three in Supp. d. in the two-actor plays the number of speaking parts remains small (for has six(seven if 1651 is spoken by the captain of Aegisthus bodyguard) Cho. seven or eight, Eum. five. For comparison, the number of speaking parts in Sophocles' surviving plays ranges from five to nine: Euripides Phoenician Maidens has eleven, as does the pseudo-Euripidean rhesus (and Aristophanes' Acharnians and Birds each have twenty-two In addition to the speaking and singing performers there could be an unlimited number of silent performers. A ruler, for example, would normally be accompanied by servants, of whom no notice would be taken by the characters unless they become relevant to the action. Thus we know that Clytaemestra has maids with her in Ag. 855ff because they are needed to spread the crimson cloth before Agamemnon(908ff ) nowhere else in the play is any reference made to her being attended, but neverthe- less it is likely that she was attended in the earlier scenes as a noble woman appearing in public normally would be (cf. Cho. 712ff ) In the murder-tableau(Ag. 1372ff ) on the other hand, Clytaemestra must cer tainly be alone with the two corpses. The Oresteia in particular also contains other groups of non-speaking performers who make a vital con tribution to the structure and effect of the drama(see $7.6.5 ) Sometimes, contrariwise, the absence of expected attendants can be put to dramatic use,as when in Pers. 598ff the queen enters without her suite to bring alone from the expedition on which all Asia had gone forth erxes returns All the performers were male. There may have been some tendency for certain actors to specialize in female roles: in the four plays certainly by Aeschylus that have female actor-parts it is each time possible to distrib ute the parts in such a way that no actor plays both male and female characters. s No such principle can be applied, in general, to Sophocles or Euripides(or to Prometheus bound Only one of Aeschylus performers can be individually identified with certainty, namely Aeschylus himself, who we can safely assume was a principal actor in his own plays. This had always been one of the respon sibilities of the author-composer-choreographer-director, as stories about
The tragedies of the sixth and early fifth centuries had only one speaking actor other than the chorus-leader, and Aeschylus himself is said to have added a second; a third actor was introduced, either by Aeschylus or by Sophocles, in the late 460s, and is employed in the Oresteia. There are moments in that trilogy when one feels Aeschylus would have liked to go even further: many have suggested that he actually uses a fourth actor at one moment of Choephoroi (but see §7.6.6), and in Eumenides the chorusleader virtually functions as a fourth actor in the trial scene. Yet at the same time he remains capable of writing brilliant drama for a single actor and chorus, as in the greater part of Agamemnon. It had always been possible for the same actor to take different parts in successive scenes, but in the two-actor plays the number of speaking parts remains small (four in Pers., two or three in the genuine portion of Seven, three in Supp.); Ag. has six (seven if 1651 is spoken by the captain of Aegisthus’ bodyguard), Cho. seven or eight, Eum. five. For comparison, the number of speaking parts in Sophocles’ surviving plays ranges from five to nine; Euripides’ Phoenician Maidens has eleven, as does the pseudo-Euripidean Rhesus (and Aristophanes’ Acharnians and Birds each have twenty-two!) In addition to the speaking and singing performers there could be an unlimited number of silent performers.17 A ruler, for example, would normally be accompanied by servants, of whom no notice would be taken by the characters unless they become relevant to the action. Thus we know that Clytaemestra has maids with her in Ag. 855ff., because they are needed to spread the crimson cloth before Agamemnon (908ff.): nowhere else in the play is any reference made to her being attended, but nevertheless it is likely that she was attended in the earlier scenes as a noble woman appearing in public normally would be (cf. Cho. 712ff.). In the murder-tableau (Ag. 1372ff.), on the other hand, Clytaemestra must certainly be alone with the two corpses. The Oresteia in particular also contains other groups of non-speaking performers who make a vital contribution to the structure and effect of the drama (see §7.6.5). Sometimes, contrariwise, the absence of expected attendants can be put to dramatic use, as when in Pers. 598ff. the Queen enters without her suite to bring offerings to Darius’ grave, and when later the shattered Xerxes returns alone from the expedition on which all Asia had gone forth. All the performers were male. There may have been some tendency for certain actors to specialize in female roles: in the four plays certainly by Aeschylus that have female actor-parts it is each time possible to distribute the parts in such a way that no actor plays both male and female characters.18 No such principle can be applied, in general, to Sophocles or Euripides (or to Prometheus Bound). Only one of Aeschylus’ performers can be individually identified with certainty, namely Aeschylus himself, who we can safely assume was a principal actor in his own plays. This had always been one of the responsibilities of the author-composer-choreographer-director, as stories about 2. Aeschylus’ Theatre 25