2. Greek Tragedy in Action (London, 1978), 77ff.Google Scholar Striking examples are the purple tapestries in Agamemnon, the bow of Heracles in Philoctetes, and the head of Pentheus in Bacchae.

3. One of many possible examples is the costumes and props which Dicaeopolis begs off Euripides himself in Achamians 407–79.

4. For an excellent bibliography of the various approaches to the play see Porter, J. R. ‘Tiptoeing through the Corpses’, GRBS 31 (1990), 255–81Google Scholar.

5. Sophocles can be added to Aeschylus, if one accepts the line of interpretation to S. El. advanced by Winnington-Ingram, R. P., Sophocles (Cambridge, 1980), 217–47CrossRefGoogle Scholar.

6. ‘Loss of royal status, a humiliating marriage, prolonged virginity and childlessness, a life of laborious poverty and physical deprivation’ (Cropp, n. 1 above, xxxvi).

7. ‘Euripidean Comedy’ in Word and Action (Baltimore, 1979), 250–74Google Scholar.

8. ‘Euripides and the Unexpected’ in Greek Tragedy, ed. McAuslan, I. and Walcot, P. (Oxford, 1993), 138–52Google Scholar (reprinted from G&R n.s. 20 (1973)).

9. ‘Realism and Character in Euripides’ Electra', Phoenix 40 (1986), 1–19CrossRefGoogle Scholar.

10. For the most recent discussion of Aegisthus' murder, see Burnett, A. P., Revenge in Attic & Later Tragedy (Berkeley, 1998), 233–5Google Scholar.

11. We need not go as far as Hammond, N. G. C., ‘Spectacle and Parody in Euripides’ Electra', GRBS 25 (1984), 373–87Google Scholar, who sees Euripides indulging in open, quite unsubtle, mockery of Aeschylus.

12. There is no certain evidence for a revival of the Oresteia in the 420's or 410's, but Ar. Ach. 9ff., Nub. 534–6 and Ran. 868ff., 1126ff. taken with Vita Aesch. 12f. make one highly probable. See Marshall, C. W., ‘Literary Awareness in Euripides and his Audience’ in Voice into Text, ed. Worthington, I. (Linden, 1996)Google Scholar.

13. See Cropp's admirably balanced discussion in his commentary (n. 1 above, xlviii–1).

14. The juxtaposition of the two motifs in 309–10 is a strong argument in favour of this punctuation, as in Murray's rather than Diggle's OCT. 308 is linguistically problematic, but the exaggeration, taken parenthetically, would be typical of Electra.

15. I am inclined to agree with Denniston and Cropp that the pitcher is put down by a servant. The brief entrance of an extra here (as at 500) reinforces the point that Electra does not need to fetch water.

16. Electra can easily take it off on her next exit during 422–5.

17. In a stage performance these breaks in the dialogue require pauses which have the effect of underlining Electra's protests still more heavily.

18. Cf. 360, 394 and also 766.

19. 360 is athetized on rather slender grounds by bothDiggle (O.C.T.) andDonzelli (Teubner). See Cropp, comm. ad loc, 123, also quoting Mastronarde's defence of the line. Attendants recur in the undisputed 394 and it is hard to see why Euripides should have introduced them if not as baggage-carriers.

20. τɛ⋯χη 360 must refer to baggage, but it is a curious word to use for the more prosaic σκɛ⋯η. In tragedy τɛ⋯χος is normally used to mean ‘vessel’ in the singular and ‘arms’ in the plural. Perhaps the baggage is called τɛ⋯χη to link it with τɛ⋯χος as used of the water-pitcher at 140. Cf. also the use of τɛυχ⋯ων in 496 in association with yet another prop.

21 πράσσονθ' ἅ πράσσω δείν΄ ύπ΄ Αἰγίσθου παθών,

ὃς μου κατέκτα πατέρα χἠ πανώλεθρος,

μήτηρ.

(85–7)

The dots after πατ⋯ρα in Murray's O.C.T. usefully suggested a momentary aposiopesis betraying Orestes' reluctance. See also Grube, G. M. A., The Drama of Euripides (London, 1941), 299–300Google Scholar. The effect is reinforced by the enjambement and heavy stop after μ⋯τηρ.

22. Lloyd (n. 9 above), 11 regards Orestes' behaviour as ‘sensible’ and Cropp, comm. on 96, compares his caution with that of Odysseus on his return to Ithaca. But this is not how Euripides’ audience would have expected Orestes to behave. Aeschylus’ (and Sophocles') Orestes also uses guile, but is nothing like so tentative. Burnett (n. 10 above), 230 n. 18, treats this timidity as the behaviour of a youth in the process of becoming a young man, but Euripides lays no special emphasis on Orestes' immaturity elsewhere.

23. Attempts to explain Orestes' delay purely in terms of dramatic suspense seem to me entirely unconvincing. Even if, as Lloyd, , op. cit. 11Google Scholar, argues, he needs to hear the rest of Electra's story, this only takes him to 337. Similarly, it is unnecessary to find Euripides guilty of lax economy for the sake of the digression on ɛὐαν⋯ρ⋯α (367–90), if Orestes' delay can be shown to make good sense in its own right.

24. Lloyd, , op. cit. 8, n. 20Google Scholar, points out that ξιπ⋯ρης in 225 simply means ‘armedwith a sword“, not necessarily ‘sword in hand”.

25. Arnott (n. 8 above), 141–2.

26. Cropp, comm. on 215–27, 114 calls this contrast ‘pathetic and humorous’.

27. The statue and altar of Apollo Aguieus before the house-entrance is not treated as a ‘prop’, though it might well have been. At this point, the irony of Electra appealing to the god who has ordered Orestes to kill his mother cannot have been missed; and later on, Orestes' reproaches of Apollo (971, 1190ff.) and Castor's judicious criticism of the oracle (1245–6) have added point in the light of the god's symbolic presence as part of the ‘set’.

28. The tricolon has an amusingly melodramatic ring.

29. With 507 cf. Cho. 752–3; with 508 cf. Cho. 1M (both passages spoken by the Nurse).

30. The vexed issue of the authenticity of 518–44 cannot be discussed further here. The latest contribution to the debate by Davies, M., ‘Euripides' Electra: the Recognition Scene again’, CQ 48 (1998), 389–403CrossRefGoogle Scholar, establishes useful analogies for Euripides' ‘criticism’ of Aeschylus in the tokens scene and sees the intertextuality as a means whereby the poet draws attention to his own novel treatment. Davies adds in a footnote (402, n. 58) that ‘Electra's criticisms of the three tokens make dramatic sense in their own right (as an expression of character, vehicle for irony etc.)’, but he does not enlarge further on this essential point.

31. This interpretation, based on 525–6, goes back to Adams, S. M., CR 49 (1935), 120–2Google Scholar. But 525–6 is only one argument of Electra's among many (cf. ἔπɛιτα 527) and it is not echoed elsewhere. By skilful irony Euripides is putting into Electra's mouth the normal, but already falsified, audience expectation of how Orestes should behave (see n. 22).

32. Note especially the uncomprehending tone of 564, 566, 568; the ironical ⋯ν⋯λπιστον in 570; and the use of μ⋯ν solitarium in 575, implying that for Electra to see the scar is not necessarily to believe.

33. The intertextual point is surely that Eurycleia in Odyssey XIX immediately recognizes the scar on Odysseus' thigh for herself, although the hero has turned away from the light to avoid her doing so (388–91). Electra is similarly confronted with a scar on Orestes' brow which she is bound to recognize, but even then hesitates (n. 32) to draw the inevitable conclusion. The detail is one of several Odyssean features in the scene. There may be a deliberate bathos in its introduction to clinch the recognition after the failure of the other tokens, but this does not seem to me the essential point.

34. Compare and contrast 580–4 with Soph, . El. 1220ff. or /827ff.Google Scholar

35. That is, if all the tokens are Aeschylean. The authenticity of A. Cho. 205–11, 228, 231–2 seems to me more problematic than that of El. 518–44. See especially Bowen, A., Aeschylus, Choephori (Bristol, 1986), Appendix A, 77–81.Google Scholar A position which would excise the footprint and woven garment in Cho. but accepts their authenticity in El. raises difficult questions; but it is not untenable if Electra's captious resistance to the three tokens is interpreted in its own right, as I have suggested.

36. See Denniston, comm. 124 and Cropp 142–3 ad loc.

37. For the significance of gestures and ‘mirror’ scenes, see Taplin (n. 2 above), chapters 5 and 8.

38. Note especially the adversarial quality of the long passage of stichomyfhia (220–89), during which Orestes and Electra express their egocentric preoccupations, and their almost complete lack of direct communication, very obvious in performance, when it comes to planning the two murders (596–667). These both anticipate the grand confrontation between brother and sister after Orestes sights his mother on the way (962–87).

39. See Taplin (n. 2 above), 83–4.

40. Arnott, W. G., ‘Double the Vision: a Reading of Euripides’ Electra(n. 2 above), in Greek Tragedy, ed. McAuslan, I. and Walcot, P. (Oxford, 1993), 212–14Google Scholar (reprinted from G&R n.s. 28 (1981), 179–91).

41. Though this interpretation has been challenged in some recent discussion (above, pp. 150–1, with nn. 9 and 10), I need hardly do more than refer to the excellent and exhaustive exploration of Aegisthus' killing by Easterling, P. E., ‘Tragedy and Ritual’, Metis 3. 1–2 (1988), 101–9Google Scholar. Also helpful and interesting are Porter, , op. cit. (n. 4), 260ff.Google Scholar and Morwood, J. M. W., ‘The Pattern of the Euripides Electra’, AJP 102 (1981), 362–70Google Scholar. Parker, R., Miasma (Oxford, 1983), 159–60Google Scholar has been quoted (e.g. by Cropp, comm. 154) as evidence that historical murders at sacrifices or festivals were not necessarily condemned by those who approved of the killing for political or other reasons. But in such cases the impiety was still there; and the point is irrelevant to the horrific character of the Messenger's narrative. No less horrific is Orestes' suggested treatment of Aegisthus' corpse (896–5).

42. See Cropp's full bibliography, comm. on 855–7, 157.

43. Kovacs, , CP 82 (1987), 139–41Google Scholar, followed by Cropp, removes the comma after ⋯πιδɛ⋯ξων and construes the negative with κ⋯ρα: ‘He comes bringing to display to you not the head of the Gorgon [a sight to fill you with horror] but instead Aegisthus, the object of your hatred [a sight to make you glad]’. Surely, though, the position of κ⋯ρα and οὐ in relation to the two participles in 856 requires the translation I have given. Violation of word-order seems more objectionable than the easy substitution of ἂλλ’ οὐ στuγɛ⋯ς Aἲγισθον for the more grammatically precise, but non-metrical, ἂλλ’ οὐ στυγɛ⋯ς Aἲγισθον.

44. Cropp. comm. ad loc, 159, usefully cites Od. 22.412 and A. Agam. 1393–1406 against boasting over the dead.

45. Cropp, comm. on 907–56, 160 draws attention to 22–35, 1109–17, 1138.

46. Denniston, comm. on 907–56, 159 comments on the dramatic flatness of 921–A, 932–7, 940–4, 948–51.

47. A second vehicle is apparently implied by the language of 998–9 and is confirmed by the plural το⋯σξ’ ὂχους in 1135, though this could be plural for singular on metrical grounds. The plural ὂχους in 966, where ὃχῳwould fit metrically, seems to clinch the matter. See Hammond (n. 11 above), 375.

48. Taplin, Pace, The Stagecraft of Aeschylus (Oxford, 1977), 304Google Scholar, 1 believe the second chariot for Cassandra, mentioned in the hypothesis of Agam., to be necessary for staging reasons.

49. If this interpretation is correct, it seems to me a peculiarly strong pointer to the priority of the Sophoclean Electro. See Cropp, comm. on 998–1096, 167–8.

50. Burnett (n. 10, above), 241–2, comments on the uneasy mixture between the agon and the scene which frames it.

51. See Cropp, comm. on 1227, 181. For the language of 1227 and 1231 cf. A. Cho. 1000 and 1011.

52. The complex characterization of Orestes and Electra is not so much there for its own sake, as it would be in a modern play, but rather to reinforce a point, viz: the total unacceptability of matricide. Resistance to the ‘psychological’ line of interpretation springs, I suspect, from an unwarranted sense that subtle and elaborate human portrayal is fundamentally alien to the style and conventions of masked theatre in 5th-century Athens when drama was in its infancy. Characters in Greek plays are not supposed to behave like that. But there can be no disputing that Euripides was a very sophisticated playwright, well capable (like Sophocles and indeed Homer) of profound human insights. Ingenious characterization was as much open to him as ingenious rhetoric. We do his originality an injustice if our interpretations restrict him to conventional forms and modes of expression when a text strongly suggests otherwise.