¹ My previous notes in CQ n.s. 18 (1968), 11–43, related mainly to W S. Barrett's edition (Oxford, 1964). Since then we have had the new texts of Diggle, J. (Euripides Fabulae i, Oxford, 1984)Google Scholar, Kovacs, D. (Euripides ii, Loeb 1995)Google Scholar, and Stockert, W. (Euripides Hippolytus, Teubner 1994)Google Scholar, all with supporting critical studies: respectively Studies on the text of Euripides (Oxford, 1981) and Euripidea (Oxford, 1994); The Heroic Muse (Baltimore, 1987), ‘Conjectanea Euripidea’, GRBS 29 (1988), 125, and Euripidea altera (Leiden, 1996); ‘Zum Text des Euripideischen Hippolytos’, Prometheus 20 (1994), 211–33. Of other studies, more than one reference will be made to: Fitton's, J. W. review of Barrett in Pegasus 8 (1967), 17–43Google Scholar; Broadhead, H. D., Tragica (Christchurch, 1968)Google Scholar; Parker, L. P. E., ‘Split resolution in Greek dramatic lyric’, CQ 18 (1968), 241–69CrossRefGoogle Scholar; Buijs, J. A. J. M., ‘Studies in the lyric metres of Greek tragedy’, Mnemosyne 38 (1985), 62–92, and 39 (1986), 42–73CrossRefGoogle Scholar; Sommerstein, A., ‘Notes on Euripides’ Hippolytus’, BfCS 35 (1988), 23–41Google Scholar; Stinton, T. C. W., Collected Papers on Greek Tragedy (Oxford, 1990)Google Scholar; M. R. Halleran (trans, with comm.) (Warminster, 1995). I have benefited greatly from correspondence or direct contact with Professor Diggle, Professor Kovacs, and Dr Stockert; Professor C. Collard also has made many helpful comments. My review of Stockert's edition (to which these notes may be read as complementary) is in JHS 118 (1998), 213. Some observations on metrical issues anticipate fuller treatment in my projected Cantica Tragica.

² ἂσημον would suggest that only the object of the ἓρως was unknown. Hartung's excision of 29–33 has been revived by H. Van Looy (Sludi… De Falco [Naples, 1971], 135–40); remedial indeed, but the five lines are not convincingly accounted for as an interpolation (ex hypothesi, with the text unemended), whether as from the earlier Hippolytus or as composed ad hoc. G. Danek (WSt 105 [1992], 19ff.) contributes little of value.

³ LSJ τίθημι A.vii, cf. Diggle, Euripidea 262–5.

⁴ False elision of -αι is frequent in the tradition (always false in tragedy, cf. Diggle, Euripidea 313).

⁵ Professor Diggle has drawn my attention to Cresphontes fr. 453.9 and Tekphus fr. 149A (727c K.).5. In the former I prefer Bergk's ἴθι μοι, πōντ[ι]α, πόλιν to Diggle's ‹ἴθ’› ἴθι μοι, πότνια, πόλιν. An ionic verse there associates well with the preceding ΦιλοστєΦάνους τε κώμους, akin to ˇˇ̶ˇ̶ˇ̶̶. In the latter, εὐθὺς Ίλίου πόρον Ἀτρєΐδας ἰδέσθαι should be taken together as e ˇ D: ba, akin ˇ e ˇ D ┆ ba at Ale. 231–2 γυναῖκα κατθανοῦσαν ἐν ἄματι τῶιδ᾽ ἐπόψηι (as I shall argue further elsewhere).

⁶ M. L. West, BICS21 (1980), 9.

⁷ Diggle, Euripidea 325.

⁸ ἀṿ᾽ αὐλάν, cf. ἀṿ᾽ ὒλην 17,χρνσέαν κατ᾽ αὐλάν Held. 916, etc. The authority for such relative clauses is E. Norden, Agnostos Theos (Leipzig, 1913), 168–76, but cf. also Kühner–Gerth 2.419., cf.єὐπατ‐ also єὐπάτєιρα (of Artemis?) at Cretans fr. 472bc K.8.

⁹ Cobet deleted οΐκον, but he also got rid of αὐλάν (proposing αἰγλ⋯εντα with οὐρανόν). The status of οἶκον here is similar to that of ἐστίν at Held. 376 (del. Blomfield).

¹⁰ ὦ καλλἴ‐⁄ στα also gives word-end after overlapping long anceps, in conflict with a colo-metric rule which it must suffice merely to adumbrate here.

¹¹ A conjecture χαῖρέ μοι, ὦ καλ‐⃒λίτα… is wrongly attributed to me by Kovacs. The idea (no longer favoured) which I had communicated to him was in fact ‹χαῖρє› χαῖρέ μοι ὦ ⃒καλλίστα… (cf. 64, where only A has the anadiplosis); ⃒̲̬̲̬̬̲… as 545, etc.

¹² Corr. Harry; cf. Diggle, Euripidea 232–3.

¹³ The sentiments expressed by the προσπόλων ὀπισθόπους κῶμος(54–5) cannot be other those of their royal ἔξαρχος. Editors differ as to the speaker-indications at 58 and 61. My inclination would be simply to print ΙΠΠΛΑϒΤΟΣ ΚΑΙ ΚΩΜΟΣ ΠΡΟΣΠΟΛΩΝ at 58, with no further indication of speaker. In general, song designated as ‘choral’ may often in practice have been initiated by a single vocalist as ‘Leader’.

¹⁴ For the split resolution after long anceps in str. (Parker 252), cf. S. El.212, Phil. 201. Eur. has both Φᾱρ‐ (more often Φᾰρ‐). On ῐ before ῥ‐, see Diggle, Euripidea 456–8.

¹⁵ ‘Enoplian’ as defined (so as to include D/e) in my commentary on Orestes, p. xx.

¹⁶ I should now (with LSJ s.v. ἦ II. l.a) writer ἦ also in Or. 1425 σὺ δ᾽ ἦσθα ποῦ τότ᾽; ἦ πάλαι Φεύγεις Φόβωι; (so also Biehl, ed. Teubner, but perhaps merely by accident since he does not comment). ‘Where were you at that time? Can it be that (you were not present to give aid because) you had already run away?’ ‘Or’ does indeed make sense, but the many parallels cited forvour ἦ. A reader proints out that in such sequences the ἦ…ἔ question is nearly always epexgetic. So here: ‘Is she mentally afflicted (as the cause of this apparently suicidal fasting)? Can she really be trying to die? (which would be evidence of παρακοπ⋯)’.

¹⁷ Editorial practice varies as to ὤ and ὦ (see also n. 26 below). Mine, after some consultation, is to write ὦ everywhere in second-person address (with or without a vocative, indifferently before or after an imperative, whether or not ‘exclamatory’), and ὤ otherwise, thus clarifying self-pitying exclamations. I should similarly write ὤ τάλας at 822, and ὤ τλάμων at 837 (where, pace Barrett, the exclamatory self-pity is unexceptionable between the wish for death and the explanation τῆς σῆς στερηθεὶς ὁμιλίας).

¹⁸ Mnemosyne 49 (1996), 53–5.

¹⁹ Cf. Stinton ap. Lloyd-Jones, JHS 85 (1965), 168.

²⁰ D. Korzeniewski's defence of Φονίοις ὑμεναίοις in responsion with Φονίωι κατηύνασεν did not merit a place in Stockert's appendix and hence a place in his bibliography.

²¹ TAPhA (1960), 162–77.

²² GRBS (1970), 307–20.

²³ The responsion of solo strophe with choral antistrophe in the Parodos of Helen is likely to have been a novelty; see my discussion in CQ 40 (1990), 77ff.

²⁴ Cf. my commentary (p. 302) on Or. 1535–65, where I may have inclined too far towards unison utterance. Weil gave the whole of 362–72 to the Coryphaeus, but a shocked reaction there from all fifteen choristers, as also from the Nurse (353–61), is what the context calls for.

²⁵ The chorus in Supplices takes such ‘vicarious’ song a stage further; CQ 40 (1990), 347–8.

²⁶ Diggle has τάλανες ὦ… here, and conversely ἂϊες ὢ, ἒκλυες ὢ… at 362 in secondperson address (cf. n. 17 above). ὤ…πότμοι at 367 are like Sept. 739 and Ch. 466 (ὤ West; contrast Lloyd-Jones/Wilson at Aj. 1197, El. 201, 203, Phil. 177–8).

²⁷ The thinly attested reading τίνας νῦν τέχνας (B2) is negligible, associable with the wrong λόγονς.

²⁸ Kovacs (1996) does not mention it in his discussion, arguing for his conjecture … ἢ τίνας σΦαλεῖσαι κάθαμμα λύειν λογους. On his own evidence the need for the repeated interrogative arises only if there is a difference of number, i.e. if we accept the v.l. λόγυς.

²⁹ Triclinius' ὦ for ἰώ at Aj. 1197 was not simply ‘a favourite device’ of his (Zuntz, Inquiry 23), but an instance of a frequently needed correction: for other accepted (if not universally agreed) instances of this corruption in the cantica of tragedy (not including ὤμοι→ ἰώ μοι, as at Aj. 900, 901, 909, etc.), cf. Ch. 466;; Eum. 511, 512, 785; Ant. 1276, 1286; El. 121 OC 1085; Ale. 92, 435, 568; Hp. 852; Andr. 1204; Hec. 1033, 1074; Here. 115; Tro. 580; Hel. 191, 362; Pho. 226, 1500–1 (IT); Or. 332, ?976; Kannicht on Hel. locc. cit. The corruption may have been encouraged by ‘prevocalic consonantalization in vulgar speech’ (cf. West, GM 14); it is less likely that ἰώ could actually be scanned as a monosyllable in tragedy, in addition to the scansions ̬̲ and ̲̲. The reverse corruption of ἰώ to ὦ, as apparently in part of the tradition at OT 1186, is surprisingly rare (omission of an iota being so easy), and editorial ἰώ is at best doubtful not only here and in 884 (q.v.), but also at IT 143; and further corrections of ἰώ may be needed (as I shall argue elsewhere) at Sept. 97; Aj. 348/356; Ant. 1146, 71261; OT 163 (ὥὢ), 1207/1216; El. 1273; Hec. 1091 (twice); Hel. 335.

³⁰ I write Φῶς, in conjunction with ‘allocutory’ ὦ (cf. n. 17 above, an d exclamations like ὦ Zεῦ). Earth is a goddess, and Light stands for Sun (often appealed to as ‘all-seeing’, etc.).

³¹ Hadley intended his πρόσθ᾽ εἰρῶ as simply introducing a restatement of Phoedra's view of death as κράτιστου (401–2); but 717–21 certainly alludes (with a fine terminal ambiguity in 721) to her discovery of a new plan, productive of compound ὄνησις.

³² Previously I proposed κλύω or ἒκλυον, not ούδέν as stated in Stockert's appendix.

³³ δ ia δ would be unparalleled in Eur., who has single iambic metra in dochmiac context only at the beginning of a sequence and then probably only when the ia is exclamatory; cf. CQ 41 (1991), 527, n. 12. One might consider supplementation to 3δ, e.g. with ‹γ’ ὂν› or ‹περ› after δυσεπέρατον, but scarcely with enthusiasm.

³⁴ My commentary there includes some observations about Eur.'s practice in the doubling of adjectives (much rarer than verbs and nouns). ‘Dittography is a less common fault than haplography, but is commoner than may generally be supposed’ (Diggle, Euripidea 381).

³⁵ El. 1153–4 πατρίδα δεκέτεσι έλθόντ᾽ έμάν ~ 1161–2 πόσις ὂ τι ποτὲ τάν ⃒ τάλαιναν ἒοχεν κακόν is δ ia cr in Diggle's colometry, but should be taken rather as ending with a dochmius, i.e. as ‸ 2ia ƒδ, akin to 2ia ƒδ (n. 36 below); likewise Tro. 1227–8, Phu. 1286ff./1298ff. (twice), Rh. 136/200, A. Ch. 944–5. For extreme rarity of ̺̲̲̺̲̺̲ (ba ia) before late Euripides (and nowhere followed by dochmius), see Stinton, 113ff.

³⁶ 2ia ┆ δ is very common. For 2ia ƒδ with short-syllable overlap, cf. Med. 1281/1292, Or. 329/345, already at Eum. 158–9/165–6 (not as West).

³⁷ For the equivalence of singular and plural, cf. Kiihner-Gerth 1.84. For confusions in the tradition, cf. Diggle, Euripidea 263 and 344–5. There is a variant έξομαι at 660, which might suggest καθέξομαι as an alternative here; but LSJ offer no support for that.

³⁸ Broadhead ended by proposing έΦ᾽ ιὦ… πονεΐς without commitment as to ‘what τινι represents’. He should not have regarded Chr. Pal.'s τά νΰν as ‘at least a possibility’.

³⁹ Aesch. and Soph, by contrast have ὂπα only in lyric; once each as things stand, but Blaydes plausibly proposed ὄπα χεΐται (χ‐ Musgrave) for ύπόκειται at Phil. 190.

⁴⁰ ού μ⋯ν … γε is surprisingly not given separate notice by Denniston. One instance (Phil. 811) appears under μ⋯ν (p. 331); others (Ag. 1279, OT 810, OT 810, Hp. 285, PV 268 [not 270], Held. 885, Hp. 914, IT 1004, Held. 885, IA 1004, Rh. 958) under μ⋯ν ‘adversative’ (335), still others (Cretans fr. 472e K.16, Hel. 571, Pho. 1622) under μ⋯ν ‘strong adversative’ (335–6). To these add Sept. 538.

⁴¹ CQ 17 (1967), 181–3.

⁴² BICS 16 (1969), 50–1.

⁴³ Halleran should not have called the change of αθανας to αγαιας ‘paleographically not difficult’. Barrett considered it ‘not obvious but not untinkable’. I think it incredible: a misreading (?) unlikely in itself, and away strainghtforward sense (in line with a scholion).

⁴⁴ Sommerstein's own Άλθηπίας has little appeal, however: another long shot, which frigidly demotes Hippolytus from ‘brightest star of Greece’ to ‘brightest star of Trozen’ and implies exile only ‘to a land other (than Trozen)’. A reader (favouring γαίας) objects that ξένην ἐπ᾽ αΐαν in 898 and 1049 has created the impression that Hippolytus' exile will be in a non-Greek land. No t so: ξένος is not the same as βάρβαρος. At 898ξένην έπ᾽ αΐαν is straightforwardly antithetic to χώρας, and at 1049 (interpolated, as argued in CQ 18 [1968]) antithetic to παρώιας χθονός.

⁴⁵ A hybrid form Άθηναν may be possible (cf. Barrett on 155–8 and Björck, Das Alpha impurum, 368ff.); but Burges was doubtless right at Tro. 803 λιπαρυΐσί τε κόσμον Άθάναις (άθ⋯ναις VP), and similarly Triclinius at Held. 359; cf. attestations of ὰθάν at A/c. 452, Ion 184, Pers. 286 (dub.), 976, Aj. 1222.

⁴⁶ It should not be argued that reference to Hippolytus' mother here is confirmed by the ‘echo’ of Hippolytus' lament ὦ δυστάλαινα μῆτερ, ὢ πικραί γοναί at 1082. Rather, we may cite 1082 as an instance of the same self-pitying idiom, lamenting one's own birth.

⁴⁷ Cf. also the sequence Tith Tith (or Tba?) at Hec. 653–5, another epode.

⁴⁸ There are no other exceptions in Eur. and none in Soph. In Aesch., apart from a very doubtful conjecture in the apparatus at 5M. 829, West's text offers only Sept. 629–30 (∼ 566–7) έπιμόλους πύργων δ᾽ ἒκτυθεν βαλών Ζεύς οΦε κάνοι κεραυνὦι, where a different division gives cr • δ followed by a hipponactean like Ag. 1488/1512, Ch. 469–70,474–5.

⁴⁹ Conomis (Hermes 92 [1964], 23ff.) nos. 24, 27, 29; but this must be pursued elsewhere.

⁵⁰ Cf. C. Segal, GRBS 11 (1970), 101–7. But his defence of the paradosis is less persuasive.