^{page 343 note 2} For the fullest treatment, see now Mellert-Hoffmann, Gudrun, Untersuchungen zur ‘Iphigen`e in Aulis’ des Euripides (Heidelberg, 1969), 91–155Google Scholar See Additional Note on p. 364.

^{page 343 note 3} Bremi (1819); so G. Murray (Oxford Classical Text), and Page, D. L. in Actors' Interpolations in Greek Tragedy (Oxford, 1934)Google Scholar, to whose convenient discussion frequent reference will be made (as AIGT).

^{page 343 note 4} Studi U. E. Paoli (1955), 293–304.Google Scholar

^{page 343 note 5} See Murray's app. crit. and AIGT, 128.

^{page 343 note 6} Seh. Ar. Ran. 67; England (Ed. 1891), xviii ff., and AIGT, 9.

^{page 343 note 7} The First Episode (303–542) opens with an abrupt confrontation which is unintelligible unless we have seen the Retainer before, and been told not only about this letter (307), but also about the previous one of which it is a cancellation; also about the seeds of dispute between Agamemnon and Menelaus. It could scarcely have been written thus by one who had not devised his play-opening in detail. Cf. p. 349, n. 2 below.

^{page 343 note 8} I hope to publish something about these other problems at a future date. In general I shall argue that Euripides had all but completed I.A. before his death, and that almost all spurious passages are inorganic additions, not stopgaps or revisions. For two examples (413–14, 518–21) see below on pp. 36a and 363; a longer one is 231–302 in the Parodos (p. 361, n. 4). So also the Finale, Euripides having intended his play to end with a lyric envoi (cf. Troades), approximately as 1531, though he may not have got quite so far. There may have been two ancient attempts to provide the play with an iambic Exodus (Angelns-narratrve and/or Dea ex Machina), both of which can be shown to be inconsistent with the Euri-pidean plot.

^{page 344 note 1} Cf. Zuntz, G., An Inquiry into the Transmission of the Plays of Euripides (Cambridge, 1965), 102, 249 ff.Google Scholar

^{page 344 note 2} Ritchie, W., The Authenticity of the Rhesus of Euripides (Cambridge, 1964)Google Scholar, 101 ff. It is not very clear what he means by suspecting interpolation ‘in 106–14’, since this passage is the linch-pin of the structure he is defending.

^{page 344 note 3} Cf. Ar. Ran. 946 f.; Arist. Rhet. 3. p. 1415^a18; Sch. E. Hec. 1. These passages explicitly confirm the Euripidean norm. They cannot, however, be held to exclude the possibility of an exception anciently attributed to Euripides. This qualification has a bearing on whether our prologue can have had its extant structure in antiquity. This is not the same question as whether the extant structure can be attributed to Euripides.

^{page 344 note 4} Ritchie argues compellingly that Rhesus has lost its original prologue, for which there is both internal and external evidence (op. cit. 101–13, 29 ff.). The manner and date of its loss, if known, might be relevant to the present problem (below, pp. 363 f.). Ritchie (pp. 42 f.) assumes, on the basis of Hyp. ii, loss before Aristophanes of Byzantium. But we can surely identify the play-opening indexed by Dicaearchus (Nυνεὐc⋯ληνον…) with one of the Πρόλογοιδιττί described in Hyp. i as extant (‘øἘρονται’) in the third century b.c. (or later), at which date there were numerous copies with the variant openings . One possibility is that Ar. Byz. himself excised it under a misapprehension (was the ‘actors’ πρόλογοc’ of 11 lines simply a clarifying ‘pre-prologue’ in some texts ?). But Ritchie neglects another obvious loophole: why should not Hyp. ii have been modified later (it was certainly abbreviated) to suit a beheaded text? The Aristophanic subject of προλογίζονcι, instead of οι Kαι…, may have been something like δύο θεαί (cf. Ritchie, p. no). On the problems of Rhesus, see also Fraenkel's review of Ritchie in Gnomon, , 37 (1965), 228–41.Google Scholar

^{page 344 note 5} Sch. Ar. Th. 1015 ff. (cf. Ritchie, 103 f.). We might visualize Andromeda preceding her trimeters with an anapaestic θρηνοc occasioned by her visibly uncomfortable situation; but this (even if proved) would afford no support for a plunge in medias res in dialogue form.

^{page 345 note 1} The case for ‘Conflated Prologues’ is argued in AIGT, 138–40. Page himself favours b, assigning the anapaests to ‘a good early writer’—not Euripides the Younger (as Bremi suggested, approved by Murray), for the reason given. ‘One thinks of the romantic atmosphere of Chairemon’ (p. 139).

^{page 345 note 2} Continuous popularity in antiquity: cf. AIGT, 9 f., 128.

^{page 345 note 3} Cf. 133–5.

^{page 345 note 4} Cf. 124–6.

^{page 345 note 5} Cf. 118, 139.

^{page 346 note 1} 115 ff—or rather 117–18–15–16? (see below on p. 356).

^{page 346 note 2} Cf. Platnauer, M., Iph. Taur. (Oxford, 1938), x-xii.Google Scholar See also p. 348, nn. 5–6 below.

^{page 346 note 3} The public nature of Calchas' oracle is here taken as certain, though some critics seem to have misunderstood, or missed the significance of, this cardinal feature of the plot. See further on pp. 361 f.

^{page 347 note 1} Arnott, P., Greek Scenic Conventions in the 5th century B.C. (Oxford, 1962), 106.Google Scholar

^{page 347 note 2} Cf. 34—a passing reference as to something that has long been in view. Agamemnon himself will take it within at 163.

^{page 347 note 3} A new suggestion, I think, but cf. Arnott, op. cit. 43 if. Agamemnon does not, presumably, brandish his δ⋯λτοι throughout 49–96.

^{page 347 note 4} So Helen, Andromache, and others (Held., Suppl.) are discovered as suppliants at a . Cf. also Rh., Hec, Tro., Or.

^{page 347 note 5} Minor textual points: in 53 f. I should read … øθόνοc(Markland) , as the positive question at issue. In 77 I should prefer : cf. 382–7. In 80 Aristotle's is the better variant (‘at a run’ —see LSJ); δερί tiresomely anticipates and this dative would here be happier with a supporting epithet.

^{page 347 note 6} Certainly 531–5 is melodramatic (AIGT, 157 f.), but not for that reason un-Euri-pidean. We are meant, I think, to hear the fears there as exaggerated; pessimism is a characteristic of Euripides' Orestes also.

^{page 348 note 1} Cf. below, nn. 3, 6; p. 350, n. 2; p. 358,n. 5; P. 359 n. 4; P. 362, nn. 2, 4, 6; also p. 351, n. 6; p. 352, n. 7; p. 357, n. 2; p. 358, n. 4.

^{page 348 note 2} Kατ<ιc>α: cf. 130. A copyist was no doubt satisfied by someone else then desperately made it K⋯ταra for the metre, kα’ ἲcαtea: cf. Il. 11. 336 and LSJ s.v. ἲcοc iv. 2 (adverbial phrases).

^{page 348 note 3} Il. I. 16 A.Ag. 43f

^{page 348 note 4} Cf. 173–8, 818, 842, 928 f. (contra 414, but see p. 362, n. 5).

^{page 348 note 5} Not (pace Platnauer) alluded to in I.A. 530, which refers to I.A. 360. Editors strangely infer (despite the absence of πάλαι or the like in I.T. 21 or 23) that the vow had been taken before Iphigenia's birth. The absurdity of such a story (vow in abeyance for fifteen years? Baby prize adjudicated fifteen years too late?) makes it vain to postulate allusiveness to a non-extant version of the myth in the imagined sense. Cicero (De Off. 3. 25. 95) worded his version clumsily, but apparently placed the vow in the current year (‘eo quidem anno’). If so, he was right. We are to understand that Agamemnon had made the vow recently at Aulis, in a dark hour (cf.øωcøόρωι 21);he was then trapped by Calchas' perversely mantie interpretation of the vow's wording. A degree of perversity in the interpretation is appropriate (cf. n. 6 below), as long as the perversity is that of the hated Calchas (I.T. 531 ff., etc.). It was as a maiden, not as a baby, that Iphigenia was adjudged Kαλλίcτη. I.T. 22 ff. ═ ‘Well now (ο⋯) your wife has (τίkει) a beautiful daughter … whom (as this year's fairest τόkοc, even if not born within the year) you must sacrifice (before the fleet can sail from Aulis).’ Payment had to precede the desired benefit (18 if.); and the year's worn D does not only bring forth babies.

^{page 348 note 6} Critics of the Agamemnon Parodos commonly fall into the error of seeking to explain why Artemis demanded a victim (so, despite some useful insights, Hammond, N. G. L. in JHS lxxxv [1965], 42–55)CrossRefGoogle Scholar. The Chorus nowhere say that she did. Nor do they say that either the eagles or the winds were sent by any god. The whole point of the elaborate ‘portent’ is to make it clear that the ‘prescription’ (μηχαρ:99) came from a human augur who could have been challenged ( 186). Calchas is wrily allowed to be a proficient seer (122, 249): but nothing guarantees his divine spokesmanship (cf. 201 f.); and Agamemnon himself in 206–17 does not claim to be obeying the gods, in performing an explicitly ‘impious’ sacrifice (219 ff.). Aeschylus' treatment was no doubt an innovation calculated to absolve Artemis from the savagery attributed to her in cult and myth; it was surely so understood by Euripides (see also I.T. 385–91).

^{page 349 note 1} Murray's ‘cf. 359’ is obscure, if intended as a ground for deletion. He was surely not (as implied in AIGT, 137) thinking 93 to be derived from 359; but neither does it conflict with it.

^{page 349 note 2} The invention was dramatically fruitful: it leads naturally to a dispute with Menelaus which would otherwise have had to precede the decision to send for Iphigenia; and a structurally important role is created for the Retainer. Both of these (and the very novelty f the idea) look like Euripidean features.

^{page 349 note 3} A legitimate veil is drawn over what ‘in fact’ the Atridae said to keep the army at Aulis (cf. 8i8 μελίκατα) and over the ‘actual’ time-lapse since the oracle (travelling-time was freely distorted or disregarded by the dramatists, as in A. Ag.). The narrative-ellipse, here explained as necessary for a special dramatic effect, is also appropriately in character: Agamemnon stresses his first, creditable reaction and stops short of revealing his subsequent tergiversations (Menelaus' account at 359 f. is rhetorically distorted in the opposite direction). For Euripidean misdirection in prologues cf. W. S. Barrett on Hp. 41–50.

^{page 349 note 4} Cf. 1475, Andr. 103.

^{page 350 note 1} The parallel is overlooked in AIGT, 131. Ritchie indeed (op. cit. 290 ff.) would remove Rh. 16 and insert after Rh. 12; but his reasons for so bold a step seem inadequate. The asyndeton and 3rd person verb in 13 f. are consistent with the question in 12 being unanswered by these undisciplined guards; so also the repeated ‘speak up!’ in line 14. Further, 16 is satisfactory in situ: Hector's θαρcω can be taken as a (rebuking?) acknowledgement of the Chorus's λάρcει, and I see no reason to insist that his following questions are ‘agitated’— he is rather denying (μων …;) the need for the guards' agitation. It is hardly conceivable that the medieval copyists of Rh. were familiar with the text of I.A., and it is sounder to regard I.A. 2–3 and Rh. 16 as providing mutual support.

^{page 350 note 2} Cf. A. Ag. 151 and Eu. 360; also Med. 153 and 183 (codd.).

^{page 350 note 3} See further below on p. 359.

^{page 350 note 4} Cf. S. Aj. 905. τί+άρα recurs in I.A. 1228.

^{page 350 note 5} Contrast Ba. 661 f. (CQ n.s. xvi [1966], 227Google Scholar): were these companion dramas consciously set in opposite seasons of the year ?

^{page 350 note 6} Arat. 148; Pliny 24. 17. 12.

^{page 351 note 1} Not ‘still on the meridian’. The Retainer has neither compass nor sextant; nor with that rendering is ‘still’ intelligible (Sirius is on the meridian before dawn in late September).

^{page 351 note 2} Il. 22. 29; Od. 11. 572 fr.; LSJ s.v.

^{page 351 note 3} Cf. Licymn. 2

^{page 351 note 4} There is no error in Rh. 529 ff. (as stated in AIGT, 132). The error was the failure of Crates, according to the scholiast, to take αỉθ⋯ιαι as predicative, so that he thought that Euripides was describing the Pleiades as setting. [For πρωταI should prefer πρωταc ]

^{page 351 note 5} Il. 8. 247; 12. 201.

^{page 351 note 6} In Ag. 826 cannot simply (if at all) have a calendar significance, and the phrase is regarded by Denniston– Page as inexplicable. Aeschylus wanted, I think, a vivid indication of night-time suitable for a context full of images drawn from hunting (for this reason I should retain παγάc in 822).

^{page 351 note 7} Cf. Rh. Hyp..

^{page 352 note 1} Enn. ap. Varr. L.L. 5. 19.

^{page 352 note 2} Bremi; so Kirchhoff, England, el al. Their motive has usually been to make άcτίρcείιοc ctipioc refer to an unnamed planet or star, to obviate the ‘astronomical error’ in ⋯γγύc. Despite Theon's similar view in antiquity (n. 4 below), there can be no question that Cείροc is here the Dogstar, in the light of the parallels given above.

^{page 352 note 3} Cf. Sch. Med. 148. See further on p. 363 below.

^{page 352 note 4} Theon Smyrn. 16. p. 202 Martin: Cf. n. 2 above.

^{page 352 note 5} Paley assigned 9–11 (but not 6) to the Retainer.

^{page 352 note 6} There is good reason to suppose that there was such a rival version of the story (contra A. Ag. 147 f., 192 ff.). England, arguing otherwise on I.T. 15, overlooked Callim. Dian. 230. S. El. 564 is ambiguous and textually uncertain, but the best Greek is Froehlich's (with Artemis as the subject). In l.T. also the text is controversial (see Platnauer, ad loc), but there is no good reason for altering is quite impossible Greek for 'obtaining adverse winds’). I.A. 813 (λεπταcπνοαιc) is decisive, if the text is sound. It is vain to dispute whether adverse winds or doldrums make more rational sense of the irrational Aulis episode. The pre-Aeschylean legend was very likely noncommittal (‘without favourable winds’), following, e.g., the pattern of Od. 4. 351–63. The preparation and rich detail in A. Ag. is consistent with ad hoc invention, in this as in other features (p. 348, n. 6; p. 358, n. 5); Aeschylus' defiance of climatic plausibility was not binding on his immediate successors, but his positive, colourful explanation of the άπλοια naturally prevailed in later antīquity.

^{page 352 note 7} Cf. A. Ag. 412 (Fraenkel).

^{page 353 note 1} A comma suffices before : cf. I.T. 64 f. (Platnauer).

^{page 353 note 2} C.Q.N.S. xviii (1968)Google Scholar, 14. Cf. also Ba. 881 (CQ.N.s. xvi [1966], 229 fr.Google Scholar).

^{page 353 note 3} There is practically nothing to be said for Nauck's ερότιμον (non-tragic, and not extant at all in quite the required sense).

^{page 354 note 1} π⋯πειν in i oo (so obtrusive after 99) is more likely to be a false variant than ‘opens non perfecti vestigium’ (Murray). Conceivably Euripides preferred π⋯πειν here (as against cτ?⋯λλειν in 119) to express an ambiguity in the wording of the summons, thus preparing the ground for Clytemnes-tra' uninvited arrival (cf. on 154 below); but such a point here would be a rather esoteric subtlety, not worth the stylistic blemish. If a pointer be sought to the authenticity of 97–103, is a strong one: cf. Ba. 1144..

^{page 354 note 2} Cf. Hp. IO (Barrett); also Hdt. and PI. Smp. 177 a

^{page 354 note 3} Cf. PI. R. 350 c , and Denniston, Particles, 238.

^{page 355 note 1} For the corrected apparatus of 109, see now ckιάν, op. cit. 97 f. cKiáv is a post-Triclinian stopgap, of sufficient but not commanding merit.

^{page 355 note 2} Med. 777 ff. includes a strikingly similar interpolation: . Deletion of 778 obviates other emendation, which still leaves the repeated ἗χει as a blemish. Someone thought to clarify ταυτα, but he mistook the point, since τυτα should (without 778) be taken as referring to Medea's exile.

^{page 355 note 3} is sound, though the rendering may suggest ehov can sufficiently imply utterance (cf. Y etc.), and would be open to misunderstanding before .

^{page 355 note 4} AIGT, 138.

^{page 356 note 1} e.g. I.T. 308, Hel. 745, etc. Was England misled by the artificial sundering of and in Beck's Index?

^{page 356 note 2} ‘One might say then that these were phrases which Euripides carried in his mind’ (Ritchie, op. cit. 224, discussing similar repetitions within Rh. and between Rh. and other plays).

^{page 356 note 3} Cf. I.T. 123–235, Tro. 122–229; cf. also Ion 144–83.

^{page 359 note 4} Cf. I.T. 252, 256.

^{page 357 note 1} On the characteristic opposition, cf. Conacher, D. J., Euripidean Drama (Oxford, 1967), 290 f.CrossRefGoogle Scholar

^{page 357 note 2} The four occurrences point to unity of authorship, at least of their contexts, much as occurs thrice in A. Ag.

^{page 357 note 3} Adversely criticised in AIGT, 135, and recalled on p. 138 as one of the cornerstones of the anti-Euripidean case. Probably it is in itself defensible: cf. Ale. 18.

^{page 357 note 4} Perhaps the original co-ordination was cf. Denniston, 193, and p. 360, n. 4. There may be (appropriately enough) a further overtone from the ⋯πι prefix: ‘nor to him additionally did I utter …’

^{page 357 note 5} Elsewhere in Euripides only in Cyc. 510.

^{page 357 note 6} Il. 14. 213 (Also Hes., A.R., AP.)

^{page 357 note 7} Cf. Ritchie, 156, 178 f., 181.

^{page 358 note 1} 137

^{page 358 note 2} AIGT, 204 f.

^{page 358 note 3} But could be right in 100 also (p. 354, n. 1).

^{page 358 note 4} ‘The tradition was variable’ (AIGT, 205). But the only variation in this aspect of the tradition seems to be Zielinski's view (Trag. Libr. Tres, 257) that Aeschylus had brought Clytemnestra to Aulis (against the unanimity of the Cypria, Sophocles, and Euripides' earlier plays). There is no ques- tion that Clytemnestra had been an in the Oresteia: she would hardly have failed to mention the fact if she had personally witnessed her daughter' death. Only one enigmatic line survives of Aeschylus’ Iphigenia (fr. 94, Nauck); there is no means of knowing what ladies are there referred to (a protesting Chorus?).

^{page 358 note 5} Cf. Denniston on El. 1020 ff. (Oxford, 1939) In one form or another ‘marriage with Achille’ seems to have been the usual pretext (if any) for getting her there, and away from her mother. As in other features, the Agamemnon Parodos strikes a different note: at Ag. 243–7 Iphigenia is simply and effectively (if strangely) pictured as having been regularly present at her father's warrior parties.

^{page 359 note 1} AIGT, 131–6. I omit some features which Page himself discounts; (false conj. in 22) and 130, which have already been dealt with (pp. 353 and 357).

^{page 359 note 2} Ritchie (pp. 141–92) gives valuable statistical tables for Rh. and other plays.

^{page 359 note 3} CQ N.s. xvi (1966), 45.Google Scholar

^{page 359 note 4} Cf. also A. Ag. 183

^{page 359 note 5} CQ.N.S. xvi (1966), 228.Google Scholar

^{page 360 note 1} 16 and 140 were wrongly criticized by Wilamowitz as further examples of irregular (cf. AIGT, 131). The division there falls between metra, and only the lineation is in question.

^{page 360 note 2} S. Tr. 977; Rh. 561 is doubtful (Ritchie, 292).

^{page 360 note 3} Ritchie, 290. For the apparatus of 149, see now Zuntz, op. cit. 98.

^{page 360 note 4} This leaves 141 f. as the only place where the heavy paroemiacs do not come in pairs. 141 becomes a paroemiac (cf. n. 5 below) if we delete the superfluous , leaving cf. 978, 1319 fr. (codd.), P- 357" n- 4 and Denniston, Particles, 509.

^{page 360 note 5} (Bothe) may well not be the right treatment of primitus L). The present suggestion might at once explain the mess at the end of the line and the wrong in 151. The symptoms are of transcriptional muddle, very possibly involving marginalia, accompanied or followed by the common attempt to pad paroemiacs into dimeters.

^{page 360 note 6} Cf. p. 356, n. 3.

^{page 360 note 7} e.g. 2/838; 130/1356; 135/936.

^{page 360 note 8} e.g. 9ff./8i3; 46ff./860; 136 f./i 132–6; 155/307.

^{page 360 note 9} Cf. Med. 96 fr., Hp. 176 fr.; similarly Hec. 59 ff., Tro. 98 ff. (Chorus of Attendants).

^{page 360 note 10} See especially on 4 f., 7 f., 20 ff., 128, 130, 137, 161. I have not thought it necessary to multiply parallels throughout.

^{page 360 note 11} e.g.19/85 f.; 12 + 34/109; 37f./no; 46ff./55+114; 116/49; 133/98.

^{page 361 note 1} See Platnauer on I.T. 1380/1404 and 1346/1394.

^{page 361 note 2} Cf. Zuntz, op. cit. 102.

^{page 361 note 3} Cf. p. 344, n. 3.

^{page 361 note 4} e.g. 231–302 by Allen, T. W. in CR v (1891)Google Scholar, 364 ff. But there is no evidence that Euripides the Younger ever wrote anything, and I should think rather of Cephisophon (Vit. Eurip. and Ar. Ran. 944, 1408, 1452 f.).

^{page 361 note 5} ‘The seer Calchas has revealed to Agamemnon, the chief commander, and his two principal advisers, Menelaus and Odysseus …’ (England, p. x). ‘Chief commander’ is also questionable (above, p. 348).

^{page 362 note 1} Norwood, G., Essays on Euripidean Drama (Cambridge, 1954), 37 ff.CrossRefGoogle Scholar

^{page 362 note 2} Il. I. 68 ff.; A. Ag. 122.

^{page 362 note 3} See p. 349, n. 3.

^{page 362 note 4} Cf. À. Ag. 214–17.

^{page 362 note 5} One objection (AIGT, 153) is quickly dealt with: read for in 423. There is, admittedly, something wrong with the Forerunner's entry in mid line. But excision of 413–14 (perhaps only of 414) leaves no scar. With the simple opening cf. Ba. 434 Note that the words conflict with the view taken above of 84 (p. 348). The interpolation is variously explicable, but lack of space precludes further discussion here.

^{page 362 note 6} very likely suggested by A. Ag. 65 and 227 . All these passages (and cf. I.A. 718 f.) depend on the root meaning ‘(perform) a preliminary rite’— not necessarily preliminary to marriage.

^{page 363 note 1} Few, since Härtung, have not condemned at least 520–1. The casual coldbloodedness of 519 was rightly condemned by Hermann, and Page suspected the whole of 506–42 (AIGT, 158). But most of the context is certainly integral (cf. on p. 347 with n. 6).

^{page 363 note 2} ⋯k⋯ιο refers to the same thing in 516 and 522, viz. the demagogy of Odysseus who knows about the marriage-plot. Agamemnon speaks allusively, and Menelaus has to force him to say what he means. Even then Aga-memnon is inexplicit (τάδε 524: cf. 106) until further pressed. This is a well-characterized fraternal exchange. The interpolator, having seriously mistaken the point, tried to answer the question in 517, and to ‘clarify’ ṳkεινο in 516. Cf. Med. 778 (p. 355, n. 2).

^{page 363 note 3} 520 f. is, I think, a secondary interpolation, of inferior, derivative quality and perhaps never intended as stichomythia. Cf. on Hp. 405–12 and 1045–50 in CQ n.s. xviii (1968), 21 ff., 34 fr.

^{page 363 note 4} Above, pp. 350–2.

^{page 364 note 1} Cf. AIGT, 14.

^{page 364 note 2} Cf. Zuntz, op. cit. 251 f.

^{page 364 note 3} Cf. the possibility that Ar. Byz. arbi trarily published Rhesus without a prologue, and so with an abruptly dramatic opening (p. 344 n. 4).