¹ So e. g. Kaffenberger, H., Das Dreischauspielergesetz in der griechischen Tragödie (diss. Giessen, 1911), 44 Google Scholar; Pickard-Cambridge, A., The Dramatic Festivals of Athens, second edn, rev. Gould, J. and Lewis, D. M. (Oxford, 1968 [with addenda 1988]), 148 Google Scholar; Ritchie, W., The Authenticity of the ‘Rhesus’ of Euripides (Cambridge, 1964), 126–9Google Scholar; Taplin, O., The Stagecraft of Aeschylus (Oxford, 1977), 366 Google Scholar, n. 1; Burlando, A. L., Reso: i problemi, la scena (Genova, 1997), 81–2.Google Scholar Fuller doxography in Ritchie, 127, n. 1. Ritchie also lists scholars who believe a fourth actor was present; add Rees, K., The Rule of Three Actors in the Classical Drama (diss. Chicago, 1908), 41 Google Scholar, 43; K. Schneider, s.v. ὑπoκριΤς in RE Suppl. VIII (Stuttgart, 1956), 187–232, at col. 191–2. Rees rejected the ‘three actors rule’ altogether: see Sifakis, G. M., ‘The one-actor rule in Greek tragedy’, in Griffiths, A. (ed.), Stage Directions: Essays in Ancient Drama in Honour of E. W. Handley (BICS Supplement 66) (London, 1995), 13–24 Google Scholar, at 13 and n. 2. Di Benedetto, V. and Medda, E., La tragedia sulla scena (Torino, 1997), 230 Google Scholar consider the presence of a fourth actor as possible but not certain.

² So e.g. Fraenkel, E., Gnomon 37 (1965), 228–1Google Scholar, at 235; Bryce, T. R., ‘Lycian Apollon and the authorship of the Rhesus ’, CJ 86 (1990–1991), 144–9Google Scholar, at 149; Pseudo-Euripide, Reso, introduzione e note di G. Paduano (Milano, 1991), 15; Burlando (n. 1), 127.

³ The fullest defence of the traditional attribution is given by Ritchie, whose conclusions are accepted by G. Zanetto (Euripides, Rhesus [Stutgardiae et Lipsiae, 1993], vi). See also Burnett, A. Pippin, ‘Rhesus: are smiles allowed?’ in Burian, P. (ed.), Directions in Euripidean Criticism (Durham, NC, 1985), 13–51 Google Scholar and 176–88 at 50–1.

⁴ This is also J. Diggle's view on the matter: Euripidis fabulae, tomus III (Oxonii, 1994), vi.

⁵ Ritchie (n. 1), 58 argues convincingly that, even if spurious, the play is probably not later than the middle of the fourth century.

⁶ See the problems discussed by Ritchie (n. 1), 101–40.

⁷ Ibid., 126–7.

⁸ Cf. below, notes 41 and 42.

⁹ Lines 138–9 imply that the Trojans and the allies (before the arrival of Rhesus) are all in one location. The position of the allies is explicitly clear if we keep σνμμχων at line 847: the Thracian charioteer knows that the Trojans and their allies are all in the same location. The word σνμμχων here needs to mean ‘all who fight on your side’, i.e. Trojans and the allies other than the Thracians—who have just arrived, and whose allegiance has been repeatedly questioned. Diggle (n. 4) obelizes σνμμχων, Zanetto (n. 3) keeps it.

¹⁰ Cf. Wiles, D., Tragedy in Athens: Performance Space and Theatrical Meaning (Cambridge, 1997), 133–60.CrossRefGoogle Scholar Wiles, p. 158 rightly argues that the Greek camp lies to the west in the play as in the Iliad, in correspondence to the audience's right.

¹¹ However, it seems that in the Iliad ‘the Trojan dispositions… lie roughly on a north-south line’ ( Hainsworth, B., The Iliad: A Commentary, vol. III: Books 9–12 [Cambridge, 1993 CrossRefGoogle Scholar], on II. 10.428–31). The text of the Iliad only says that ‘the Trojans’ order of battle extends from a point ‘on the side towards the sea’ … presumably the Hellespont, to a point “on the side towards Thumbre”’ (Hainsworth, ibid.). The poet of Rhesus could have taken the sea as the Mediterranean, or simply have left this detail out. A general overview of the problems in Homeric topography is given by Cook, J. M., ‘The topography of the plain of Troy’, in Emlyn-Jones, C. and Hardwick, L. Purkis, J. (edd.), Homer Readings and Images (London, 1992), 167–74.Google Scholar

¹² Aristot. Poet. 1455a27–9 drew attention to what was apparently an accident involving a non-realistic presentation of exits and entrances in the Amphiaraos of Karkinos. The precise reconstruction of what happened on stage is disputed: see Snell in the apparatus to TrGF 70 F lc and Craik, E. M., ‘Arist. Po. 1455 a 27: Karkinos’ Amphiaraos’, Maia 32 (1980), 167–9.Google Scholar

¹³ Wiles (n. 10), 156–8 discusses possible ways for Odysseus and Diomedes to slip past the chorus.

¹⁴ On her appearance see below, notes 34 and 36.

¹⁵ Some manuscripts assign the part of the chorus at 675–91 to a secondary chorus of Lycians (cf. 543–5), but this (unlikely) suggestion would not affect the problems we are discussing here. This hypothetical secondary chorus would have to come from the side of the Trojan camp.

¹⁶ Many scholars suggest that Diomedes does not reappear on stage after his exit at line 637: cf. e.g. Euripides’ Werke verdeutscht von F. H. Bothe, fünfter und letzter Band (Berlin-Stettin, 1803), 265; Burlando (n. 1), 83; Euripides Iphigenia among the Taurians, Bacchae, Iphigenia at Aulis, Rhesus, trans. J. Morwood (Oxford, 1999), 153, 233. I find this very unlikely. It would give the impression that Diomedes never left the tent of Rhesus. It also goes against the explicit words of the chorus, who take more than one prisoner: see Τοσδ∊ at line 681 and cf. perhaps also 678–9.

¹⁷ Diomedes leaves too late: from 637 to 642 he has not enough time to change ( Hermann, G., ‘De Rheso tragoedia dissertatio’ in Opuscula, volumen tertium [Lipsiae, 1821], 262–310 Google Scholar, at 284).

¹⁸ Ritchie (n. 1), 126–9 (with other references); Pickard-Cambridge (n. 1), 148.

¹⁹ Ritchie (n. 1), 127: ‘a little confusion in the entrance of the Chorus would add to the delay’. Ritchie also claims that a slight delay ‘is indeed suggested in 673’, presumably by τ μλλ∊τ∊;. This expression is however ‘equivalent to an urgent imperative’ (Euripides, Phoenissae, ed. D. J. Mastronarde [Cambridge, 1994], on 1146; cf. also Euripides, Orestes, ed. C. W. Willink [Oxford, 1989²], on 275–6) and does not imply a pause in the action.

²⁰ Wiles (n. 10), 158.

²¹ Cf. Taplin, O., ‘Aeschylean silences and silences in Aeschylus’, HSCPh 76 (1972), 57–98,Google Scholar at 57–8. Commenting on Eum. 33–4, Taplin ([n. 1], 185,362) notes that between the exit and re-entry of the Pythia there is a hiatus: ‘nothing happens. This is unique in surviving Greek tragedy, which generally abhors a vacuum, and nearly always preserves continuity.’ In Eumenides, the hiatus is a suspense-building technique.

²² Taplin (n. 1), 353.

²³ See below, n. 41.

²⁴ Ritchie (n. 1), 126–9.

²⁵ Ibid., p. 128.

²⁶ Pickard-Cambridge (n. 1), 145 and Di Benedetto and Medda (n. 1), 222–3 argue for the employment of three actors in the Alcestis, but do not raise the question of the direction of the exits.

²⁷ Cf. Mastronarde (n. 19), 16 and on lines 88–102; Pickard-Cambridge (n. 1), 147; Di Benedetto and Medda (n. 1), 228.

²⁸ The most persuasive reconstruction is in Di Benedetto and Medda (n. 1), 228: actor I: Jocasta, Menoiceus, Messenger B, Oedipus; actor II: Servant (lines 89ff.), Polyneices, Creon, Messenger A; actor III: Antigone, Eteocles, Tiresias. Mastronarde (n. 17), 17 and on lines 88–102 leaves the question open.

²⁹ Di Benedetto and Medda (n. 1), 227 consider sixteen iambic trimeters insufficient to allow the same actor to play both the messenger and Athena at I.T. 1419–35: the actor had to change costume and to gain access to the mechane or to the top of the skene. On the possibility of a very quick change of costume in comedy (nine lines), cf. Macdowell, D., ‘The number of speaking actors in old comedy’, CQ 44 (1994), 325–35,CrossRefGoogle Scholar at 327 and 335; even quicker and more difficult changes are discussed in Aristophanes, Ecclesiazusae, ed. Sommerstein, A. H. (Warminster, 1998), 31 ndn. 118.CrossRefGoogle Scholar

³⁰ The horses are not explicitly mentioned in lines 675–91, but I think they must have been present. Heath, Musgrave, and other earlier scholars had no doubts about their presence: cf. Euripidis opera omnia (Glasguae-Londini, 1821), vol. 5, 378–9. They were right to do so: cf. the chariots on stage in Persians, Agamemnon, Trojan Women, Iphigenia in Aulis and in the Electro of Euripides. See Taplin (n. 1), 75–9; Di Benedetto and Medda (n. 1), 95–6. It is possible that Rhesus entered on his chariot at line 380: Taplin (n. 1), 77.

³¹ The suggestion was originally made by Vater, F. (Euripidis Rhesus [Berolini, 1837],Google Scholar p. LV n. *). If Athena was not actually seen on stage, the scene would be unparalleled. When voices are heard from within in tragedy, it is always clear that the speaker is inside the house, even at Ba. 576–95. The speaker is never a disembodied spirit: see Di Benedetto and Medda (n. 1), 58–65, 69. Unlike Taplin (n. 1), 366–7,1 think Clytemestra was visible at the beginning of the Eumenides (cf. Aesch. Eum. 103).

³² Taplin (n. 1), 366 n. 1; Burlando (n. 1), 81–2.

³³ Burlando (n. 1), 82.

³⁴ See Mastronarde, D. J., ‘Actors on high: the skene roof, the crane, and the gods in Attic drama’, CA 9 (1990), 247–94,Google Scholar at 274–5 and 284.

³⁵ Ibid., p. 275.

³⁶ Athena pretends to be Aphrodite in order to deceive Alexander, but I disagree with the scholars who imagine that she changes her costume on stage. There is no mention of a change of costume in the text. See Burnett (n. 3), 39–40.

³⁷ Wiles (n. 10), 158.

³⁸ Wiles does not discuss the problem of the fourth actor, and does not draw this inference from his staging.

³⁹ Kaffenberger (n. 1), 44.

⁴⁰ Pace Burlando (n. 1), 81, n. 43.

⁴¹ Cf. Hermann, C. F., Disputatio de distributione personarum inter histriones in tragoediis Graecis (Marburgi, 1840), 65,Google Scholar n. 48; Taplin (n. 1), pp. 353–4; Di Benedetto and Medda (n. 1), 218–19.

⁴² Cf. Taplin (n. 1), 185, n. 2; Di Benedetto and Medda (n. 1), 222. The alternative is an elaborate role-splitting: Pickard-Cambridge (n. 1), 142–4 (discussing earlier suggestions) and Kamerbeek, J. C., The Plays of Sophocles. Commentaries. Part VII. The Oedipus Coloneus (Leiden, 1984), 23.Google Scholar Sifakis (n. 1), 19–21 argues that role-splitting was quite common in Greek tragedy, but I am not persuaded by his arguments.

⁴³ Dearden, C. W., The Stage of Aristophanes (London, 1976), 92.Google Scholar See the discussion at 88–94. Similar explanations in Pickard-Cambridge (n. 1), 149–52, Sifakis (n. 1), 18, 24, and in the works quoted by MacDowell (n. 29), 325 n. 3.

⁴⁴ Taplin (n. 1), 186.

⁴⁵ MacDowell (n. 29), 335, arguing against the possibility, accepted by some scholars, that a fifth actor could be occasionally required to play ‘barbarians, children, or small parts’.

⁴⁶ Ibid., p. 326.

⁴⁷ Pickard-Cambridge (n. 1), 236–7; cf. Carriere, J., Le choeur secondaire dans le drame grec (Paris, 1977), 6–9,Google Scholar 50–9.

⁴⁸ The only certain example is the chorus of frogs, which probably appeared on stage: Aristophanes, Frogs, ed. with introduction and commentary by K. J. Dover (Oxford, 1993), 56–7. Carriere (n. 47), 9–12, 59–64 discusses other possible secondary choruses.