¹ See Taplin, O., The Stagecraft of Aeschylus (Oxford, 1977), p. 453Google Scholar.

² Pickard-Cambridge, , The Theatre of Dionysus in Athens (Oxford, 1946), p. 43Google Scholar; Hammond, , ‘Dramatic Production to the Death of Aeschylus’, GRBS 13 (1972), 449Google Scholar; Melchinger, , Das Theater der Tragbdie (Munich, 1974), 155Google Scholar.

³ Pickard-Cambridge, pp. 43, 52, notes that a paraskenion with an entrance into guest chambers may be implied by Alcestis 546–9, and that paraskenia might be a hiding place for Orestes and Pylades, at Choe. 20Google Scholar. The walls of paraskenia, making corners with the skene, could be regarded as internal walls and so make the interior of the temple, which I believe the audience looks into at Eum. 1–235, more realistic.

⁴ The Ancient Theatre, trans. Vafopoulou-Richardson, C. E. (London, 1982), pp. 6, 25Google Scholar. There is an unused mosque (made for symmetry only, and not facing Mecca) opposite the one in use, beside the Taj Mahal.

⁶ Dale, A. M., Collected Papers (Cambridge, 1969), p. 269Google Scholar, spoke for a single door. She assumed the audience would expect a contemporary, not an ante-bellum house (one built before the Trojan War), and thought a change of locale could be as easily made in Choe. as in Pers. Alcestis 546 she understood (p. 268) as matter for the audience to hear, rather than for the personae to enact. Taplin, , Stagecraft, 350Google Scholar, assesses her work and, though with independence, is in general agreement. He would have the servant shout for the opening, and then for the forcing, of a single door only, one that changes its bearings.

⁶ Bain, D., Masters, Servants and Orders in Greek Tragedy (Manchester, 1981), p. 59Google Scholar, holds that the locale has already been shifted into the courtyard when the servant appears, and that, having come through one door, he shouts for two others to be opened, one of them for him to go in by, the other from the women's quarters into the courtyard. Garvie, A. F. (ed.), Choephori (Oxford, 1986), pp. l–lii and n. on 877–9Google Scholar, would have the servant shout only for the women to open the door from their quarters. (Bain discusses Blass and Diels, as well as Dale and Taplin; Garvie discusses Reinhard as well as Bain, and justly reports the views of many others besides. The Homeric parallels are misleading as always.)

⁷ ‘Inconsistency of Plot and Character in Aeschylus’, Proceedings of the Cambridge Philological Society 9 (1963), 21–62Google Scholar. See also Garvie, on Choe. on 479–584Google Scholar, and Taplin, 94–6. In arguing for anti-parallelism, so that there will be no presentiment (other than the dark prophecy of Cassandra, , Ag. 1318–19)Google Scholar that the second tableau should exactly resemble the first, I may be in disagreement with Garvie (p. li, citing Ag. 1291) and Bain (p. 60), who impute sinister moment to the doors of death. The use of a fourth (speaking) actor in Choe. is another surprise (see Garvie p. xlix, note 118). I would suggest that Pylades, with the voice of Cassandra and Apollo, is the third actor, and that the servant, here a fourth actor except that he is the third to speak, is elsewhere the spokesman for the group of armed men, if they have a speaking role in the satyr play. After the servant has delivered his ten lines in Choe., the audience, from the three-actor ‘rule’, would be assuming, though casually rather than with a scholar's attention, that Pylades was a mute; that he should speak at all would then be a surprise, a special effect giving special force to his words. The use of a fourth actor in comedy—see Dover, K. J. (ed.), Ar. Clouds (Oxford, 1968), pp. lxxvii–lxxixGoogle Scholar, and Dearden, C. W., The Stage in Aristophanes (London, 1976), pp. 86–94Google Scholar; is not of much relevance to Aeschylus.