¹ See Proclus, , EGF, p. 19Google Scholar. Burnett, Anne P., Catastrophe Survived (Oxford, 1971), p. 73Google Scholar, distrusts this attribution to the Cypria of a reference to the Taurians; but in any case by the time of Herodotus (4.103) these had already come to identify their goddess as Iphigenia, the daughter of Agamemnon. In both [Hesiod] (Merkelbach–West 23a, 23b) and Stesichorus, (PMG 215)Google Scholar Iphigenia (or Iphimede) is saved by Artemis and deified as Hecate. A summary of views expressed before 1939 about the originality of the plot of I.T. can be found in Lesky, A., s.v. ‘Orestes’, RE xviii. 1, 997ffGoogle Scholar. Discussion has centred on the content of Sophocles' Chryses and its date relative to that of I.T. Some would use as evidence for Sophocles' plot the story in Hyginus 120 (ad fin.) and 121, which takes the form of a sequel to the rescue of Iphigenia by Orestes. That assumption, if combined with the likelihood that Chryses antedated I.T., constitutes an argument that Sophocles knew a lost earlier version of the rescue. But the derivation of Hyginus' story from Sophocles' plot is very questionable, and even the relative dating of Chryses and I.T. is not quite beyond doubt. On the first issue contrast von Wilamowitz-Moellendorf, U., ‘Die beiden Elektren’, Hermes 18 (1883), 257–8Google Scholar, with Sutton, Dana F., The Lost Sophocles (Lanham, New York and London, 1984), pp. 29–31Google Scholar; on the second, see schol. Aristophanes', Birds 1240Google Scholar (which implies that Chryses antedated 414) and Dale, A. M., Eur., Helen (Oxford, 1967), p. xxviiiGoogle Scholar (where 414 is chosen on metrical grounds as ‘the most likely date’ for I.T.). For further evidence on the date of I.T. see n. 20, infra. H. Grégoire, who believes that Chryses was earlier than I.T. and that its plot followed the lines of Hyginus 120–1, also maintains that the rescue of Iphigenia was originally a local legend of Halae and Brauron which both Sophocles, and Euripides, adapted to the stage (ed. I.T. [Paris, 1948], pp. 96–9)Google Scholar. His argument arises out of I.T. 1449ff. but seems otherwise largely speculative.

² See I.T. 176–7, which must imply a divinely inspired delusion. A similar version of the sacrifice is put in more explicit terms in Euripides, fr. 857 (TGF ²), from the lost genuine end of l.A. See also Hulton, A. O., ‘Euripides and the Iphigenia Legend’, Mnem. 15 (1962), 364–8CrossRefGoogle Scholar. [Hesiod] fr. 23a.21 (Merkelbach–West) speaks of an εἴδωλον.

³ Cf. Webster, T. B. L., The Tragedies of Euripides (London, 1967), pp. 136–43, 192–9, 186Google Scholar; Burnett (supra, n. 1), p. 74. Other examples of the first type are Euripides' Ion and Sophocles' Mysians (if, as commonly assumed, the plot of the latter is reflected in Hyginus, Fab. 100); of the second, Choephori. Though Euripides' Andromeda (a play of 412 B.C.) is probably later than I.T., the story is found earlier, as on the three vase-paintings of 450–440 B.C. reproduced in Trendall, A. D., Webster, T. B. L., Illustrations of Greek Drama (London, 1971), pp. 63–4Google Scholar, and attributed there to Sophocles' Andromeda. Cf. Webster, , Monuments Illustrating Tragedy and Satyr Play², BICS Suppl. 20 (London, 1967), p. 147Google Scholar. The plot of Helen offers very close analogies to that of I.T., but Helen is most probably the later play. See infra, n. 43.

⁴ Lattimore, Richmond, Story Patterns in Greek Tragedy (Ann Arbor, 1964)Google Scholar; Webster (supra, n. 3), p. 12; Burnett (supra, n. 1), pp. 1–17; Strohm, Hans, Euripides: lnterpretationen zur dramatischen Form (Munich, 1957)Google Scholar. Burnett's, study of Bacchae (‘Pentheus and Dionysus: Host and Guest’, CP 65 [1970], 15–29)Google Scholar is also based on a typology of plots of divine punishment. von Wilamowitz-Moellendorf, Tycho, Die dramatische Technik des Sophokles (Berlin, 1917)Google Scholar, had done much to turn attention to plot construction. His influence is acknowledged by Friedrich, W. H. (Euripides und Diphilos [Munich, 1953], p. 6)Google Scholar, who believes that primary and secondary treatments of an action can be distinguished and the former sometimes deduced from the latter. But in treating the latter as ‘counterpoints to lost melodies’ (6) he is still in each case comparing dramatized versions of a single myth. See, however, his comments on Euripides' repeated use of plots based on catastrophe averted (59–60), a theme later re-examined by Burnett. The focus turns more clearly to general patterns in Tübingen, Walther Ludwig's dissertation, Sapheneia, Ein Beitrag zur Formkunst im Spātwerk des Euripides (1954)Google Scholar, and in Hans Strohm's book. Though Strohm was critical of Ludwig's, methods (Gnomon 29 [1957], 494–8)Google Scholar, the two have much in common. Both, for example, argue that Euripides varies his patterns and adapts them skilfully to the needs of each play. Earlier, F. Solmsen had combined the analysis of plot elements with reflections on Euripidean themes in his studies of ναγνώρισις and μηχνημα in several of that author's plays (‘Zur Gestaltung des Intriguenmotivs in den Tragödien des Sophokles und Euripides’, Philologus 87 = N.F. 41 [1931–1932], 1–17Google Scholar; ‘Euripides' Ion im Vergleich mit anderen Tragödien’, Hermes 69 [1934], 390–419Google Scholar).

⁵ See the review of Strohm's book (supra, n. 4) by Stevens, P. T. in CR 9 (1959), 115–17Google Scholar and Knox's, B. M. W. criticisms of Burnett's book (supra, n.l) in his otherwise generally favourable appraisal in ‘New Perspectives in Euripidean Criticism’, CP 67 (1972), 277Google Scholar.

⁶ Lucas, D. W., Aristotle, Poetics (Oxford, 1968), p. 294Google Scholar; cf. 186. See Poetics 1452a12ff., 1452b30–32, 1454b19–1455a21.

⁷ See Burnett (supra, n. 1), p. 16, where the poet is said to ‘arouse [in the audience] the combined memories of all the other tragedies of his chosen type and then transcend them with variations…’. It is likely enough, as she claims, that spectators of tragedy drew from their general experience of plots some expectations which helped to guide their reactions and laid them open to surprise. She has aptly compared such regularities of plot to rules of metre (pp. vii–viii). But the spectator she envisages on pp. 15–16 of her book seems at the least implausibly attentive to the mere prosody of tragic plots (to use her own figure) and unduly preoccupied with their formal irregularities. Observation of the latter, it appears, can be ‘deeply disturbing’ to him (16), and the disappointment of his trained expectations will teach him that he is not ‘omniscient’ (15). There may be an overstated point in this, but it is as if one were to describe the audience's reaction to a tirade of Pentheus as a series of responses to its deviations from metrical norms. The difference is that in metre the standard rhythms, against which deviant examples can be measured, are not in short supply, whereas in Burnett's theory of tragic action every real play is a variant and the norms merely deductions. Moreover, those humdrum plots that followed these norms most closely (and so might have provided us with the best evidence of their grip upon the imaginations of spectators) were the least likely to survive (16). All this seems an insecure basis on which to build a theory of audience expectations. Suspense and surprise were undoubtedly features of tragedy, but there is clearer evidence of another way of producing these, viz. the invention ad hoc of misleading detail. Friedrich (supra, n. 4), pp. 150–3, has shown how Sophocles toys with the audience in his Electra, but not with memories of wellworn patterns. More to the point here, Hamilton, R. does something similar for I.T. itself (along with Ion, Hel., Ale.) in ‘Prologue, Prophecy, and Plot in Four Plays of Euripides’, AJP 99 (1978), 283–8Google Scholar. (But I do not find his paraphrase of the prologue speeches in I.T. [p. 283] altogether accurate.)

⁸ Antigone's story is like Niobe's (Soph, . Ant. 824–33Google Scholar); Cassandra's like Procne's (Aesch, . Ag. 1140–5Google Scholar); Medea's like Ino's (Eur, . Med. 1282–92Google Scholar); Leda's like Callisto's, but – alas for Leda – not nearly enough like it (Eur, . Hel. 375–80Google Scholar). Agamemnon's story begins like that of Amphiaraus, but Electra doubts (wrongly) that it will follow the same course (Soph, . El. 837ff.Google Scholar). Clusters, of myths, each similar in some way to the story being dramatized, occur at Aesch. Cho. 585–651Google Scholar and Soph, . Ant. 944–87Google Scholar. Analogies may also exist between the parts of a complex legend; a trilogy will offer opportunities for exploiting these. See Lesky's, A. discussion of Oresteia in Die tragische Dichtung der Hellenen ³ (Göttingen, 1972), pp. 123–4Google Scholar. A precedent for the sustained presence of a background myth exists in the Odyssey, where the story unfolds against the analogies and contrasts of the Orestes legend (see esp. 1.298ff., 3.193ff. and 303ff., 11.405ff., 24.191ff.; also 1.35ff., 3.247ff., 4.521ff.). Snell, Bruno speculates on the roots of the mythical paradigm in Die Entdeckung des Geistes ⁴ (Göttingen, 1975), pp. 188–91Google Scholar, and offers other Homeric examples. Eight Iliadic examples are discussed by Willcock, M. M. in ‘Mythological Paradeigma in the Iliad’, CQ 14 (1964), 141–54CrossRefGoogle Scholar. His intention is to show that in each case the paradigm is either wholly invented or freely adapted to its context. An inventory of examples from many authors, with analyses, can be found in Oehler, R., Mythologische Exempla in der älteren griechischen Dichtung (Aarau, 1925)Google Scholar. Johansen, H. Friis, General Reflection in Tragic Rhesis (Copenhagen, 1959), pp. 50–3Google Scholar, discusses the connection in tragedy between mythological exempla and the expression of general rules and notes some important distinctions.

⁹ Oehler (supra, n. 8), pp. 87–9 cites Eur, . Hipp. 525–64Google Scholar as an example of an implicitly paradigmatic passage. Another, not cited, is Eur, . El. 458–60Google Scholar, a reference to Perseus λαιμοτμαν linked by no expression of comparison to the cutting of Clytemnestra's throat, which the chorus foresee in the last lines of this ode. But the paradigm was already familiar from Aesch, . Cho. 831–7Google Scholar, a play much on Euripides' mind in composing his El. (see El. 520–46) and one which Newiger, H.-J. thinks may have been revived a few years earlier (‘Elektra in Aristophanes' Wolken’, Hermes 89 [1961], 422–30)Google Scholar. Its presence is also consistent with other passages in El. itself, e.g. 1218–23. See my discussion in AJP 85 (1964), 13–39. Bowra, C. M., Pindar (Oxford, 1964), pp. 278–316, esp. 298, 304–7Google Scholar, discusses the uses of myth in that author, including that of serving as implicit example.

¹⁰ Bloody violence recurring in a family is the commonest form of this theme. See Aesch, . Cho. 1065–76Google Scholar, Soph, . Ant. 594–8Google Scholar, Eur, . Or. 816–18Google Scholar. Cf. Ismene's, chronicle of the Labdacids at Ant. 49–60Google Scholar, where the flow of repetitive circumstance is expressed in a verbal pattern (αὐτοϕώρων…αὐτουργῷ…αὐτοκτονοντε…διπλς…διπλον…δο). For examples from Euripides see n. 48 infra.

¹¹ Other paradigms for the plot of I.T. are sometimes proposed. Burnett (supra, n. 1), in discussing the genesis of the story of Iphigenia's rescue, speaks of the innovating poet (who may well be Euripides, she says) as having seen ‘the relevance of the Andromeda pattern’ (74). But Andromeda is here a starting point for composition rather than, it appears, an element in the finished play; in any case, her name may be intended simply as the label for a plot type rather than as an image in the ruminating poet's mind. On the other hand, Caldwell, R., ‘Tragedy Romanticized: the Iphigenia Taurica’, CJ 70.2 (1974), 23–40Google Scholar, protests against Burnett's derivation of the play from a repertory of ‘abstract hypothetical plots’ (25) and argues for a single, concrete model, the Oresteia (25ff., 32). But his scheme of correspondences is too complex to be credible: plots must be truncated in order to match each other, and characters must undergo multiple cross-identifications (e.g. Iphigenia, of I.T. has four counterparts in Oresteia [30]Google Scholar). He also thinks that one cannot and need not prove that Euripides or his audience was conscious of all this (24, 34).

¹² Forthcoming in RhM. In Or. what is foreshadowed does not include the final rescue.

¹³ Quotations from I.T. conform to the text of Diggle, J., Euripidis Fabulae, vol. 2 (Oxford, 1981)Google Scholar.

¹⁴ One view is that it is a mere genealogical ‘hors d'oeuvre’ (Méridier, L., Le Prologue dans la tragédie d'Euripide [Bordeaux, 1911], p. 58Google Scholar). Euripides often begins a play with a genealogical excursus, as in the seven extant and four lost plays cited by Méridier on pp. 54–5. These passages, like other elements of his prologue technique, have often been criticized. See the history of scholarly opinion in Erbse, H., Studien zum Prolog der euripideischen Tragödie (Berlin and New York, 1984), pp. 1–19Google Scholar and the censorious views assembled by Treu, Max in ‘Wohl dem, der seiner Väter gem gedenkt’, Gymnasium 75 (1968), 437–8Google Scholar. A limited relevance is sometimes conceded to I.T. 1–2; but it is not enough to say, as Grégoire (supra, n. 1) does in his note to line 1, that the later reference to Pelops during the recognition (823ff.) justifies the earlier one in the prologue. Why is the recognition made to hinge upon this reference? Cf. Erbse, p. 194, where the connection of 1–2 with 823ff. is also made. The question is treated at some length by Treu. He thinks Iphigenia's thoughts go out to Pelops as the founder of the house in Greece; lines 1–2 express her longing for home and family (pp. 443–5). But this way of looking at Pelops is never brought out in the text. That being so, his foreign origin gives him a very uncertain value in the thoughts of an Iphigenia supposedly concentrating on her ties to Greece. In the diatribe of Teucer, at Soph. Ajax 1292Google Scholar Pelops is called a barbarian, and in art he wears a Phrygian costume (see Séchan, L., Études sur la tragédie grecque dans ses rapports avec la céramique [Paris, 1926], pp. 316, 460Google Scholar). For Sansone's view see n. 19 infra.

¹⁵ The legend is narrated at length in Apollodorus, Epit. 2.3–9. The early stages of its development are far from clear; the main question is whether Myrtilus' betrayal of Oenomaus is an accretion to the original myth. On the Chest of Cypselus, as described by Pausanias (5.17.7), Pelops drives winged horses, but Myrtilus is not said to be present. Säflund, M.-L., The East Pediment of the Temple of Zeus at Olympia (Göteborg, 1970), p. 120Google Scholar, says that even on the Chest Oenomaus ‘can hardly do without a charioteer’. Even if this is granted, the latter is presumably one of the unnamed figures (Paus. 5.17.6 implies that there are some) and so will not have been thought of by the artist as playing the important role of the betrayer. Pherecydes mentions both Myrtilus and winged horses (FGrHist 3 F 37). On the Olympia pediment the horses are not winged, and it is unclear whether Myrtilus is there. Säflund, pp. 118–21, believes she has identified him as the third figure from the left (C), but her book also documents the lack of consensus on this point (pp. 11–59). There are various ways of sorting out this evidence. Kakridis, J., ‘Des Pelops und Iamos Gebet bei Pindar’, Hermes 63 (1928), 415–29Google Scholar = Calder, W. M. III, Stern, J., Pindaros und Bakchylides (Darmstadt, 1970), pp. 159–74Google Scholar, argues that the winged horses belong to the original version, perhaps as a gift of some other god than Poseidon; Myrtilus belongs to a later version, where his trickery displaces the winged horses as the means of Pelops' victory; Pindar introduces the motif of Poseidon's love and excludes Myrtilus (159–67). See also his ‘Die Pelopssage bei Pindar’, Philologus 85 (1930), 463–77Google Scholar = Calder, /Stern, , Pindaros, pp. 175–90Google Scholar. More recent discussions are those of Calder, W. M. III, ‘Sophocles, Oinomaos and the East Pediment at Olympia’, Philologus 118 (1974), 203–14, esp. 205, 211CrossRefGoogle Scholar; Köhnken, A., ‘Pindar as Innovator: Poseidon Hippios and the Relevance of the Pelops Story in Olympian 1’, CQ 24 (1974), 199–206CrossRefGoogle Scholar; Gerber, D. E., Pindar's Olympian One: a Commentary (Toronto, 1982), pp. 134–6Google Scholar; and Willink, C. W., Orestes (Oxford, 1986), pp. 248–50Google Scholar. For the vase-paintings see L. Séchan (supra, n. 14), pp. 447–66; and Webster, T. B. L., Monuments (supra, n. 3), pp. 150, 161–2Google Scholar. In dealing with allusive poetic versions of the myth it is sometimes forgotten that missing elements may be absent because they have been suppressed (see the article by Stinton cited infra, n. 16). In this way Myrtilus is often wrongly introduced into discussions of I.T. E.g. for Burnett (supra, n. 1), p. 64, the spear of Pelops, for all its heroic connotations, is a ‘symbol of the family's first crime’, viz. the ‘removal of a lower class accomplice’. Cf. Whitman, C., Euripides and the Full Circle of Myth (Cambridge, MA, 1974), p. 23CrossRefGoogle Scholar: ‘Pelops won by cheating and thus brought down the original curse on the family’. See also Vickers, B., Towards Greek Tragedy (London, 1973), pp. 298–300Google Scholar.

¹⁶ Stinton, T. C. W., ‘The Scope and Limits of Allusion in Greek Tragedy’, Greek Tragedy and its Legacy, Essays Presented to D. J. Conacher (Calgary, 1986), 67–102Google Scholar. See especially p. 74, with its enlightening distinction between the basic schema of a legend, the current standard version, and mere trivial detail. Stinton's argument, mainly concerned with Sophocles' treatment of the central myths in Ant., O.T., El. and Trach., extends as well to the use of exempla in tragedy. Of these he says: ‘An exemplum is geared to its specific paradigmatic function’ (81), a point relevant to the use of the Pelops story in I.T. Taplin, O., Greek Tragedy in Action (London, 1978), pp. 162–4CrossRefGoogle Scholar, also insists on the dramatist's freedom to select, invent, and suppress details of myth. He concludes that ‘what is not alluded to does not, within the play, exist’ (p. 164).

¹⁷ δινευοσαις ἴπποισι 〈ῥιϕα Πλοπος 〉 πταναῖς (Murray, G., Eurip. Fab. ii ³ [Oxford, 1913]Google Scholar). The horses, however, probably belong not to Pelops but to the sun, a view taken by several editors, including E. B. England (London, 1886), M. Platnauer (Oxford, 1938), and N. Wecklein, 3rd edn (Leipzig, 1904), ad loc. Both the participle and the adjective are suitable to the sun's team (cf. ἴπποισιν εἱλσσων ϕλγα at Eur. Phoen. 3, ἴπποις ἂμ πτερσσαις at Eur. El. 466, and πτερωτν…ἄρμα at Eur. Or. 1001–2), and the words that follow πταναῖς in this corrupt and lacunose passage certainly refer to the change in the sun's course. (Murray's ῥιϕα is a term found in Lycophron.)

¹⁸ Apollodorus, , Epit. 2.7Google Scholar, mentions as alternative causes of Oenomaus' death: (1) that he is dragged to death while entangled in the reins; (2) that Pelops kills him. The second is unlikely to allude to a version like that of Pindar and I.T., since it continues: ὃς ν τῷ ποθνσκεν κατηρσατο τῷ Μυρτλ1FF3;. Probably, then, it refers to one in which Pelops kills him after Myrtilus has made his chariot crash. For examples of this in art see Körte, G., I Rilievi delle Urne Etrusche, ii. 1 (Rome and Berlin, 1890), pp. 109ffGoogle Scholar. and the corresponding plates, particularly V.I and XLII.4. One cannot therefore say that line 825 by itself demonstrates the exclusion of Myrtilus from I.T. The sum of the play's references to Atreid history, however, leaves a heavy burden of proof on anyone who claims that he is implicitly present.

¹⁹ Cf. Sansone, D., ‘The Sacrifice-Motif in Euripides' IT’, TAPA 105 (1975), 290Google Scholar: ‘Pelops seems to have been the only member of the family…to have prospered…’ as Iphigenia will in the end prosper. Sansone, however, has in mind something other than the contest with Oenomaus. He means that Pelops, like Iphigenia, was ‘innocent of wrongdoing’ and ‘saved by the gods’; i.e. the gods (except Demeter) forbore to eat him when Tantalus served his flesh at a banquet. But in order to say this he must take 386–8 as an acceptance of this (the usual) version of the myth rather than as a denial, in the manner of Pindar, O. 1.52, that Pelops was ever offered as food to the gods. I see two difficulties with this theory: (1) Iphigenia clearly dismisses some version as unworthy of belief, and the existence of one more discreditable to the gods than the usual version is very uncertain (see Gerber [supra, n. 15], p. 87, and Roscher, , Lexikon iii.2 s.v. ‘Pelops’, cols. 1870–1Google Scholar); (2) her words give no hint that she is willing to accept any version of the banquet or even knows of more than one. It is, therefore, more natural to read 386–8 as saying, with Pindar, that there was no cannibal feast from which Pelops had to be rescued. The relevant part of his legend is the marriage-contest.

²⁰ As assumed by Calder, , ‘Sophocles…’ (supra, n. 15), 206Google Scholar, and Sutton (supra, n. 1), p. 96. The fr. is no. 473 (TrGF). The evidence favouring an early date for Sophocles' play is summarized by Sutton, p. 97. It amounts to very little. Calder's article constructs a speculative case for 468. The essential point is that it antedated 414 (schol. Aristophanes, , Birds 1337Google Scholar), which is Dale's preferred date for I.T. (supra, n. 1). Moreover, Cropp, Martin and Fick, Gordon, Resolutions and Chronology in Euripides: the Fragmentary Tragedies, BICS Suppl. 43 (London, 1985), p. 61Google Scholar, record three metrical criteria that favour placing I.T. after Tro., hence after 415. According to Webster, , Monuments (supra, n. 3), p. 150Google Scholar, Sophocles' Oenomaus begins to be reflected in vase-paintings c. 420. For the display of severed heads in Sophocles see fr. *473a (TrGF). Cf. Apollodorus, , Epit. 2.5Google Scholar. On a vase illustrated by Séchan (supra, n. 14), p. 453, the severed heads of two previous suitors are affixed to the background. Séchan, p. 464, associates this painting with Euripides' play; Webster, , Monuments, p. 150Google Scholar (cf. pp. 161–2), would evidently prefer Sophocles, since it portrays an Oenomaus in his prime and ‘Euripides' Oinomaos seems to have been an old man’. There is, however, no real proof that he was, though fr. 575 (TGF ²), with context and speaker unknown, may suggest it. See also Webster's, Euripides (supra, n. 3), p. 115Google Scholar.

²¹ E.g. by Scherling, in RE xvi.1, col. 1156, s.v. ‘Myrtilos’Google Scholar. Some maintain that Euripides' Oenomaus was produced with Phoenissae c. 409. The evidence is a mutilated sentence in Aristophanes' hypothesis to the latter play. But see Lesky (supra, n. 8), p. 444. Cropp/Fick (supra, n. 20), p. 86, find the metrical evidence for dating inconclusive.

²² These include Evenus, Antaeus, the Thracian Diomedes, and others. For the references see Pearson, A. C., The Fragments of Sophocles (Cambridge, 1917), fr. 473, n.Google Scholar; and TrGF iv fr. *473a, n.

²³ Calder, , ‘Sophocles…’ (supra, n. 15), 208, 206Google Scholar; D. F. Sutton (supra, n. 1), p. 96.

²⁴ See also Arnmian. Marcell. 22.8.34. Editors of I.T. differ on the meaning of line 74. Platnauer regards the display of heads as beyond doubt; England is less certain; and Wecklein, p. 19, thinks the σκλα referred to are weapons (all refs. supra, n. 17). Strachan, J. C. G., ‘Iphigenia and Human Sacrifice in Euripides' Iphigenia Taurica’, CP 71 (1976), 132Google Scholar, argues that armour is a more natural reference for the words κροθνια and σκλα. But at I.T. 459 κρθνια refers to the new victims themselves, and at Eur. El. 897 σκλον is used of a dead body and perhaps of a severed head. Eur. Ba. 1168ff. shows a similar taste for the horrific. Orestes' fright before these σκλα (I.T. 102–3) is identical to Pelops' reaction before a spectacle of severed heads in the unknown source of Hyginus 84.3, which may be Euripides' Oenomaus. See Scherling (loc. cit., supra, n. 21).

²⁵ See lines 1429–30, and again cf. Hdt. 4.103.

²⁶ (Berlin, 1875), p. 190. General discussions of etymologizing in tragedy can be found in Kannicht, R., Euripides, Helena (Heidelberg, 1969), ii.21Google Scholar; in Wilson, J. R., ‘The Etymology in Euripides Troades 13–14’, AJP 89 (1968), 66–71Google Scholar; and in an article influenced by Wilson, , Van Looy, H., ‘Παρετυμολγει Εὐριπδης’, Zelesis [de Strycker, Festschrift E.] (Antwerp and Utrecht, 1973), 345–66Google Scholar.

²⁷ See n. 1, supra.

²⁸ Hyginus, Fab. 15, 120; Flaccus, Valerius, Argon. 2.300–2Google Scholar. Aristophanes, , parodying I.T. 32–3Google Scholar in fr. 373 (PCG = fr. 357 CAF, from the Lemnians), forges a comical link between one Thoas and the other. For the date of Lemnians see Geissler, P., Chronologie der Altattischen Komödie, ed. with Nachträge (Dublin and Zürich, 1969). pp. xvi, 55Google Scholar.

²⁹ S.v. ‘Thoas’ in Roscher, , Lexikon v. col. 815Google Scholar.

³⁰ In ‘Die beiden Elektren’, Hermes 18 (1883), 254 n. 3Google Scholar. Having explained why the lines could not be deleted, he added: ‘Aber wie in aller Welt kam Euripides dazu, den überflüssigen Namen zu erfinden, oder vielmehr aus der Sage von Hypsipyle…zu übertragen?’ He had no answer.

³¹ Hel. 13–14; Ba. 1–2, 27ff. and passim (see n. 32, infra).

³² In fact, irrelevance of this order is hard to match in Euripides. Van Looy (supra, n. 26), who discusses 69 Euripidean etymologies, arranged by location, allows a wide range of comparison. Of those occurring near the end of a play (e.g. in a god's speech), most, so far as one can generalize from examples in extant plays, have no bearing on the action itself; all but one, however, have an aetiological purpose (pp. 354–8). The exception, Or. 1635, makes a dramatic point (by explaining why Helen has been rescued) through a play of etymology. A high proportion of those in the episodes and choral odes (pp. 358–65), including all explicit references in extant plays to the meaning or aptness of proper names, have at least minor and local dramatic relevance (e.g. Su. 496 [Capaneus], Tro. 891 [Helen], Tro. 989–90 [Aphrodite], I.A. 321 [Atreus], [?Eur.] Rhes. 158–9 [Dolon]) and sometimes have major relevance to the play's themes (e.g. Phoen. 636, 1493 [Polynices], Ba. 367, 507–8 [Pentheus], Tro. 1217 [Astyanax], fr. 781.12 [TGF] [Apollo], Ion 661, 802 [Ion]). Within this group, the derivation of Ion's name, in which Van Looy sees no serious intention (pp. 359–60), is actually a way of illustrating the important theme of τχη (cf. 661–3 an d 1512–15, and see Owen, A. S., Ion [Oxford, 1939], ad 1514Google Scholar). Among examples from the prologues, Van Looy concedes to three a clear dramatic justification (p. 354). These are Dionysus at Ba. 1–2 (see Dodds, E. R., Ba. 2nd edn [Oxford, 1960], ad loc.Google Scholar, and lines 27, 84, 466, 550, 859–60, 1340–2), Theonoe, at Hel. 13Google Scholar (see 317, 530, and esp. 823), and Thetideion, at Andr. 20Google Scholar (see Stevens, P. T., Andr. [Oxford, 1971], ad 16–21Google Scholar and line 1231). His conclusion, however, that most prologue etymologies lack dramatic relevance (p. 354) is overstated. If one leaves out of account the above three and Thoas (in his view a mere παγνιον [350, 354]), 12 of his remaining 17 examples from prologues (his own tally on p. 353 is slightly inaccurate) come from the fragments, where dramatic relevance cannot easily be assessed. (Of these, 5 are aetiological, which gives them at least a non-dramatic point.) In 3 others either the text or the etymologizing intention is open to question (El. 1, Phoen. 3 and 57–8). Tha t leaves the possible play on ΚκλωΨ at Cy. 21–2, which has certainly acquired some point by 462–3, and the derivation of Οἰπους at Phoen. 27. This last line is deleted by Page, D., Actors' Interpolations in Greek Tragedy (Oxford, 1934), pp. 25–6Google Scholar, and by Fraenkel, E., Zu den Phoenissen des Euripides, Bayer. Ak. d. Wiss., Philos.-Hist. K I., Sitzungsb. 1963. i.8–11Google Scholar. Genuine or not, it has some relevance to the sufferings of Oedipus, a theme which recurs a t 801ff. Among passages not considered, the interpolations at Hel. 9–10 and Tro. 13–14 can be left out of account. Van Looy also regards Hipp. 29–33 as interpolated (p. 349), though few editors delete these. But the lines, if genuine, have an aetiological motive. In short, the wholly irrelevant etymology is at least exceptional in Euripides. Within the class of etymologies to which ‘Thoas’ belongs, viz. explicit references to the meaning or aptness of proper names found in extant plays (where relevance can best be judged), all the other examples in Van Looy either provide an aition or have dramatic point.

³³ No-one at least can object that the notion is too naive for Euripides. ‘Thoas, the pursuer’ is no more naive tha n ‘Ion, who comes out’ and ‘Theonoe, wh o knows like the gods’. But why a swift runner rather than a swift driver or sailor (32–3)?

³⁴ Pindar, O. 1.76–9; Diod. 4.73.4–5; Apoll. Rhod. 1.756–8. The horses of Oenomaus, which were sometimes identified by name (Schol. Apoll. Rhod. 1.752; see FGrHist 3 F 37), were the gift of Ares (Apollod. Epit. 2.5) and were faster than the north wind (Hyginus, Fab. 84.2; cf. Diod. loc. cit.). Unlike Oenomaus, Thoa s does not mount his chariot or raise his spear; instead, he orders all the citizens to pursue the fugitives (1422ff.). But he speaks later as if he had intended to take part himself (πασω σ λγχην ἣν παρομαι 1484). This expression, although it may be figurative, is reminiscent of Oenomaus' role in the Pelops legend. For the construction at 1484 cf. Eur. Ba. 789 and see LSJ s.v. παἱρω 1.2. Line 1326 may anticipate this image of the raised spear, but some commentators prefer to translate σρυ there as ‘ship’ (e.g. England [supra, n. 17], ad loc.).

³⁵ So on the Chest of Cypselus (Pausanias 5.17.7) and on the five vase-paintings (H, I, J, K, L) discussed by Séchan (supra, n. 14), 454–8. At Apollod. Epit. 2.5 one of the terms of the contest is that Hippodameia must ride with Pelops. Cf. Apoll, . Rhod, . Argon. 1.752ffGoogle Scholar. and Philostr, . Imag. 1.17Google Scholar.

³⁶ Burnett (supra, n. 1), pp. 49, 16. ‘Vengeance’ in her scheme has a technical sense (as on p. 49) and refers broadly to actions ‘in which destruction [is] wrought by the principal upon another figure’ (p. 16).

³⁷ The broad similarity in their fates (both are the intended victims of sacrifice) impressed Polyidus the Sophist, who thought Euripides might well have used it to bring about the recognition, or who perhaps actually wrote a play of his own in which it was so used. See Aristotle, Poet. 1455a7, where the text is problematic; and cf. b10. In contrast, modern criticism makes little of the coincidence that Orestes' fate is like his sister's. D. J. Conacher goes further than most in observing that in the first part of I.T. each thinks the other dead and that this matching of delusions is artfully brought out in prologue and kommatic parodos (Euripidean Drama [Toronto, 1967], pp. 305–6, 310Google Scholar). In Burnett's analysis (supra, n. 1, pp. 47–8), the play involves (by my count) eleven rescues, past, present, and to come, in seven of which Orestes and Iphigenia share; to that extent their fates are parallel. The most elaborate theory is that of D. Sansone (supra, n. 19). In his view Orestes and Iphigenia are the last two intended victims of a chain of human sacrifices in the family, of which the first was the banquet of Tantalus (288–9, 293); and Orestes' ordeal in the play is a ritual re-enactment of these crimes (286, 292–3) which finally purifies him and ends her bitterness about Aulis (286–7). There are several objections to this theory, (i) Iphigenia denies the truth of the banquet of Tantalus (supra, n. 19). (ii) The supposed references to the ritual re-enactment are a very mixed bag of the literal and the symbolic. Most of those cited on p. 287 have to do either with the routine preparation for Orestes' sacrifice (e.g. line 705) or are part of the ruse played on Thoas (e.g. lines 1191ff.); some actions mentioned as evidence are merely like purificatory acts (e.g. lines 255, 296–8). All function as parts of the alleged symbolic pattern. Lines 92 and 981, also cited, admittedly bear on Orestes' present madness and its prescribed cure, and here indeed is evidence that he still needs purifying after the Areopagus trial (cf. Parker, R., Miasma [Oxford, 1983], pp. 107, 129Google Scholar). But the cure for this madness is denned by Apollo as renewal of exile and the completion of a task (lines 85–92). There are parallels for such supplementary purification by exile and πνοι in the legends of Heracles and Jason (Parker, pp. 370, 382; for exile as a component of purification see further pp. 114, 387). In short, the action of the play taken literally, not symbolically, is sufficient to account for the end of Orestes' madness. Finally, (iii) if Iphigenia forgets her bitterness about Aulis, it is reason enough that she has found her brother and that rescue is in sight; no curative ritual is needed. I agree with Sansone, however, on the importance of some passages, e.g. lines 338 and 358 (see text infra).

³⁸ Cf. Sansone (supra, n. 19), 284. The herdsman speaks in general terms of killing ‘strangers like these’, and he uses the present subjunctive of repeated action (ναλσκῃς 337). But the form of expression can hardly be said to stress the quantity of victims required for payment, as argued by J. C. G. Strachan (supra, n. 24), 136. The emphasis is on their quality; τοισε and τοιοσδε call attention to what the herdsman has already said about the noble looks and bravery of the captives (267, 321–2).

³⁹ Lines 36, 389–90, 585–6, 1021.

⁴⁰ Iphigenia's words at 843–4, δδοικα δ' κ χεπν με μ ππóς αìθπα μπτμενος φὑγῃ, may remind us that after her own escape from death at Aulis she was sent δι…λαμππòν αíθπα (29). The coincidence is striking, but it may be deceptive. References to flight through the air involving one or more of the terms ναπτομαι, ππòς αíθπα (or αìθπιος etc.), and πτεποῖς (or πτεπωτóς etc.), are commonplace in Euripides (and not unknown in other authors) to express one of the following: (1) a disappearance or loss, usually deplored; (2) an escape desperately desired; (3) an escape ironically dismissed as impossible. Examples of each include: (1) Andr. 1219, Hec. 334, HF 69 and 653, Med. 440, and the parody of Euripides at Aristophanes, Frogs 1352 (cf. Empedocles, 31 B 2.4 [Vorsokr.]); (2) Hec. 1100ff., HF 1158, Hipp. 1290–3, Ion 1238, Eur. fr. 781.6Iff. (TGF ²); (3) Hec. 1264, Hel. 1516, Med. 1297, Or. 1593, Phoen. 1216 (cf. Herodotus, 4.132.3). I.T. 843–4 falls easily into the first category, and more than that perhaps should not be said about it. It may be, however, that these well-worn expressions occasionally come to life in surroundings where their literal meaning is particularly apt. So interpreted here, Iphigenia's way of expressing her fear would fall easily into the pattern of parallels linking brother and sister.

⁴¹ There is a further reference to the Aulis story at 1418–19, where the messenger ends by saying that Iphigenia has forgotten Aulis and betrayed the goddess. But there are difficulties in 1414–19 that have brought these final lines under suspicion, and 1413 would be an effective conclusion to the speech. See England (supra, n. 17), ad loc, and Page (supra, n. 32), p. 78.

⁴² At 1439–41, spoken by Athena, Orestes' rescue of his sister is now included in the details of his mission from Apollo and made syntactically parallel with his return of the statue. This implies that the rescue was part of Apollo's intention. The lacuna after 1014 may have contained matter relevant to this question. For the meaning of 1440 see Platnauer (supra, n. 17), ad loc.

⁴³ It is perhaps no coincidence that in the ‘second prologue’ of Helen with its genealogical proem (386ff.) Menelaus, , like Iphigenia at I.T. 1ff.Google Scholar, begins with Pelops' victory over Oenomaus. (The Binneninterpolation at Hel. 388–9, which would have complicated this allusion with a reference to the banquet of Tantalus, is sufficiently discredited by the discussion of Kannicht [supra, n. 26] ad loc.) Pelops is Menelaus' ancestor too, and his success in escaping with Helen from a barbarian king is closely modelled on the escape from Thoas in I.T. This relation between the plays will be reversed if we take I.T. to be later than Helen, as argued by Perrotta, G., ‘Studi Euripidei I’, SIFC N.S. 6 (1928), 28–9Google Scholar, and Pohlenz, Max, Die Griechische Tragödie ² (Göttingen, 1954), p. 391Google Scholar. But metrical evidence is against this dating. See Dale, loc. cit., supra, n. 1. Perrotta's argument, in any case, requires assuming a date of 413 for Euripides' Electra; against this see the arguments of Zuntz, G., The Political Plays of Euripides (Manchester, 1955), pp. 64–71Google Scholar.

⁴⁴ By Sophocles; possibly also by Euripides (nn. 20, 21, supra). For evidence of sculpture, painting, lyric see nn. 15, 20, supra. Sophocles' Oenomaus evidently achieved some fame. It is parodied at Birds 1337 (schol. ad loc.) and, according to Hesychius s.v. ρονραῖος Οἰνóμαος, was revived in the fourth century by a company that included Aeschines. There is some confusion in the evidence about this production. See O'Connor, J. B., Chapters in the History of Actors and Acting in Ancient Greece (Chicago, 1908), pp. 74–7, 106Google Scholar; TrGF vol. 4, p. 381; Pickard-Cambridge, A. W., The Dramatic Festivals of Athens², rev. by Gould, J. and Lewis, D. M. (Oxford, 1968), p. 50 n. 5Google Scholar. That it was Sophocles' play seems most likely, though O'Connor, who does not cite Hesychius, assumes Euripides. In contrast, not much can be said about possible dramatic versions of the sacrifice at Aulis that may have antedated I.T. The extended lyric account at Aesch. Ag. 104–257 does not mention a rescue of the victim. Aeschylus and Sophocles each wrote an Iphigenia, about which a little is known, and Friedrich (supra, n. 4), pp. 89–109 has attempted some inferences. See also Fraenkel, E., Aesch. Ag., 3 vols. (Oxford, 1950), ad 247Google Scholar and 1523ff. In any case, I.T. 6–30 gives the audience a full account of the event.

⁴⁵ See Pickard-Cambridge (supra, n. 44), pp. 52, 275–8. The sharpest apparent contradiction is between the evidence of Antiphanes, fr. 191 (CAF) and that of Aristotle, , Poet. 1451b25–6Google Scholar. As an example of divergent modern interpretations, contrast Lucas, D. W., Aristotle, Poetics (Oxford, 1968), ad loc., and Pickard-Cambridge, p 276Google Scholar.

⁴⁶ There are well-known comic examples at Frogs 830–1471 and Thesm. 848–1135; Oehler (supra, n. 8), pp. 78–111, gives many examples from tragedy. In comic poets this practice is sometimes thought to imply little more than flattery of the audience, irony, or reliance on general features of tragic style and on the notoriety of certain quotations; otherwise, references may be confined mainly to recent, well-remembered productions. Tragic poets, it is supposed, aim some allusions at the learned few. See Dover, K. J., Aristophanic Comedy (Berkeley and Los Angeles, 1972), pp. 188–9Google Scholar; Pickard-Cambridge (supra, n. 44), pp. 276–7; Taplin, O., The Stagecraft of Aeschylus (Oxford, 1977), p. 17Google Scholar.

⁴⁷ Supra, n. 37.

⁴⁸ At Hipp. 337–43, Phaedra finds that her troubles repeat those of her mother and her sister, also victims of love, and concludes: κεῖθεν μεῖς, οὐ νεωστ, δυστυχεῖς. At Ba. 337–41, Cadmus warns Pentheus to avoid the fate of his cousin Actaeon, who offended a god and died by σπαραγμς, an example that Pentheus has begun to follow. At El. 1147ff., Clytemnestra is killed while the chorus sing about Agamemnon's death; here the earlier event is both model and cause. Later in El., the Dioscuri trace the troubles of Electra and Orestes to ἄτη πατρων (1307). In Or., the punishment of Tantalus provides a distant precedent for Orestes' present situation (supra, n. 12). In Ion the theme of family solidarity among the Erechtheids is overshadowed by that of Athenian nationalism, with which it merges; it is clearly present, however, at 469–70 and 1464–7, which link the fortunes of the γνος and δμος with those of Creusa and Ion. Its relevance to the shape of the plot is suggested in Burnett's remarks (supra, n. 1), p. 105, about the parallel careers of Ion and Erichthonius. The latter proves to have been a model from birth for his descendant Ion, who like him will become king of Athens (lines 20–6, 1427–9).

⁴⁸ The common refusal to attach any importance to genealogical detail in the prologues is one expression of this attitude (supra, n. 14). Another is the belief that notions such as inherited ἄτη cannot coexist with the highly developed personal motivation of Euripidean characters except as empty formulas or marginal themes. Greene, W. C., Moira (Cambridge, MA, 1944), p. 204CrossRefGoogle Scholar, speaks of the pronouncement of the Dioscuri at the end of El. as ‘jumbled’ because of their insistent use of ἄτη πατρων and similar terms (see line 1307). Cf. Lesky (supra, n. 8), p. 404, who uses the word ‘Randmotif’ of the same phrase.Kamerbeek, J. C., ‘Mythe et réalité dans l'oeuvre d'Euripide’, Entretiens sur l'Antiquité Classique vi, ed. Reverdin, O. (Vandoeuvres-Genève, 1960), 1–41Google Scholar, defends at length his theory of a contradiction between Euripides' own realism and a legacy of myth with which he did not feel completely at ease but found charming and ornamental as an Alexandrian poet might have. Genealogical elaboration in the prologues is explained in this light (9–10, 29, 39).

⁵⁰ IT. 50–1, 154–5, 848–9, 992, 995. Orestes is equally concerned about the house: see 697–8, 984–5 (σσον πατρῷον οἶκο∋…).