¹ Ag. 748/761 is exceptional. For this form, see table on p. 63, Ia.

² cf. Wilamowitz-Moellendorff, U. v., Griechische Verskunst (Berlin, 1921), 210: ‘…eine Reihe euripideischer Chorlieder, deren Mass bei den beiden anderen Tragikern in dieser Ausdehnung und Reinheit nicht vorkommt’.Google Scholar

³ Maas, P., Greek Metre (translated by Lloyd-Jones, H, Oxford, 1962), §54.Google Scholar

⁴ In their commentaries E. R. Dodds (Bacchae) and W. S. Barrett (Hippolytus) use glyc³ and glyc.c respectively, while C. Collard (Supplices) gives glyc the name ‘cho dim B’ and our colon ‘cho dim C’.

⁵ Hephaestion (ed. Consbruch 56 ff.) treated this metre, with others, in the chapter on ‘polyschematista’. Wilamowitz (SB Berlin 1902, 783 ff., now Verskunst 210 ff.), writing before the appearance of papyri of Corinna, discussed below, was first to name and define the colon. Maas (op. cit. §54 addendum 1961) suggests, without explicit reason, the name ‘wilamowitzianus’. See also Snell, B., Griechische Metrik³ (Göttingen, 1962), p. 29Google Scholar, Dale, A. M., The Lyric Metres of Greek Drama² (Cambridge, 1968), p. 133 (using the name ‘aeolic dim B’ and ‘chor. dim. B’ ) and R. Kannicht, commentary on Helen (following Maas’ nomenclature). The name ‘chor dim’ is doubly inappropriate: firstly, the colon is not actually a dimeter; and secondly the implied metre chor-chor is, as I shall argue below, completely different.Google Scholar

⁶ op. cit. p. 29. He follows the description of Wilamowitz (e.g. Verskunst 212, where he writes that the ‘former metron’ may be either iamb, or troch.). Dale also wrote in the first edition of Lyric Metres of Greek Drama (Cambridge, 1948), p. 129.Google ScholarPubMed J. D. Denniston rightly criticizes this scheme in his review (CR62, 1948,118 ff.):‘…of which [=‘polyschematist’ dimeter] I believe the third syllable to be invariably long, being an essentially different rhythm’ (120–1). This review seems to have been neglected by many authors of metrical studies and commentaries including Raven, D. S., Greek Metre (London, 1962)Google Scholar and Korzeniewski, D., Griechische Metrik (Darmstadt, 1968). But Snell himself clearly says in Lexikon d. alten Welt 3273 (‘Wilamowitzianus’) that the form is and that chor-chor and ia-chor are avoided. For the second edition of Dale, see note 10 below.Google Scholar

⁷ ‘Indo-European metrics and archaic Irish verse’, Celtica 6 (1963), 194ff. Watkins is rash to find in Greek metres traces of initial unfixed syllables supposedly of Indo-European origin. One of his examples is evidently wrong. The metre of Alcaeus 70LP is not but regular repetition of ia + glyc and lesser asclepiad . Watkins should not have ignored Maas’ criticism (DLZ 1924, 517) of this ‘freedom’, which Meillet, A., Origines indo-européennes des mètres grecs (Paris, 1923), first suggested. Another alleged example is illusory. Watkins cites ζηνòς ύψηρεϕ⋯ς ρόμoς (sic) of Anacreon ‘10D’ (sic) to prove the existence of , a variation of which covers iamb. dim. and also glyc. But if one actually consults Herodian (ed. Lenz, 2² p. 642, 23 = Anacr. 7D = (10 Hiller-Crusius) = Adesp 978P) a normal glyc line line ζηνòς ύψερεϕ⋯ς†δόμoις† (ρόμoς Bergk) is found. Misspelling appears already in Meillet op. cit. in another context. In fact, there is no evidence of free initial syllables in Greek metres except in the aeolic base.Google Scholar

⁸ Comparative Studies in Greek and Indie Metre (Cambridge Mass., 1974). The theory of a four-or eight-syllabled ‘Urvers’ from which all actual metres develop as ‘Gestalt’ is fascinating (Wilamowitz, Verskunst 234, etc.), but entirely speculative.Google Scholar

⁹ G. Nagy, op. cit. p. 37.

¹⁰ Dale, op. cit., has cf. Barrett, Hippolytos, Appendix l. Since Dale has the qualification that - x may be changed to x -, her configuration agrees with Maas‘, so long as is excluded. However, in her commentary on Helen p. 150, is treated as'chordim’.

¹¹ It must be emphasized that such terms as ‘metron’, ‘the former metron’ or ‘dimeter’ are used only for convenience. No implication, structural or historical, is intended.

¹² Including 38 τ’ Ώαρί ωυ α-.

¹³ For the usage of other comic poets, cf. p. 70 below.

¹⁴ This explanation of Dale (op. cit. p. 147 n. 1) seems plausible. Snell (Lexikon d. alten Welt 918, ‘Eupolidēus’) adopts wil/wil ٨. Or one might speculate that the initial two syllables of lecythion () are occasionally inverted or ‘dragged’ through the influence of the first colon.

¹⁵ cf. Denniston, op. cit. p. 121. Apart from our current concern, it is noteworthy that aeolic base is hardly ever occupied by either in Corinna or in Ar. Nub. (parabasis). We shall treat the case of Euripides later.

¹⁶ 3 chor + ba: Med. 643 f./652 f., Heracl. 353 f./362 f., HF 637 f./658 f. 3 chor + sp: Alc. 984f./995f., ia+chor+ba: Med. 431/439 etc.

¹⁷ Thus, rightly, D. M. MacDowell, against Wilamowitz, who treats all lines as ‘chor dim’. Dale analyses Ar. Nub. 563–74 similarly (op. cit. pp. 135–6).

¹⁸ The single syllable is the prevailing type of word-overlap in aeolics. Cf. Dale, op. cit. p. 146. There is another more positive reason for moving this syllable. The colon has pendant ending, and T. C. W. Stinton proves that’ pendant close is always a mark of period end, if it is followed by a short or anceps’ (CQ, n.s. 27 (1977), 39). Diaeresis (word-end) is not a safe guide to colometry in aeolic metres’ (p. 45). ‘The safest guides are (i) the analogy of other verse-forms or sequences of verse-forms elsewhere, (ii) a repeated metrical pattern’ (p. 43).

¹⁹ Though the avoidance of split resolution is not an absolute standard, it is reasonable to deduce from word-division Cf. L. P. E. Parker, CQ, n.s. 18 (1968), 241.

²⁰ can be regarded as, but Euripides may have treated it actually as . This shape is occupied by wordsτρυχόμευoς (Hel. 521) оἰκτρότατα (Or. 814) τύυδαρις (Or. 826) εὐμενίσc (Or 836). They are familiar types of words used in iambics - lyric dimeter or trimeter, and spoken trimeter of later plays - with ‘dactylic opening’.

²¹ Supp. 971 f., El. 126 f., IA 579 f.

²² cf. examples in Appendix.

²³ Stinton suspects this colometry of Wilamowitz because it ‘gives no sense-pause at ⋯ρϕναν (BICS 22 (1975), 92). Cf. n. 18 above. But at the same time he hesitates to approve Murray's colometry ⋯ς ⋯ρϕυαυ μoλόυτα πα***δ (this is counted as a doubtful case of (p. 94)).

²⁴ Wilamowitz, Verskunst, p. 244.

²⁵ As far as I notice, parallels are: Hipp. 553/563 (with ⋯ξέδωκευ *** and μέλισσα δ'o*** (Barrett's colometry. Cf. Stinton, CQ, n.s. 27 (1977), 42)) and IA 793.

²⁶ cf. OT 465/475 and Ar. Nub. 572. All of the three may be regarded as ‘chor dim’ used as a variation of ‘iambo-chor’ with long syllables constituting the former metron’. See note 41 below.

²⁷ εὐ⋯νεμoν (Musgrave) is attractive (cf. Dale, ad loc), but this line is excluded from the statistics. Cf. Kannicht, ad loc.

²⁸ Except in IT 1096/1113 (and, if we accept Murray's emendation with different colometry, HF 795/812 νε χαλκασπίδων λόχoς, ὃ ς/-νεξιϕηϕόρων ⋯ς ⋯γω This phenomenon deserves further examination. If the same result is acquired from other aeolic cola and in other poets too, we shall have to find a new explanation of ‘aeolic base’ of tragic cola.

²⁹ cf. Appendix.

³⁰ Raven notices that does not correspond with (op. cit. 78 n. 22). But he regrettably treats ‘chor dim’ with the configuration in main text.

³¹ Certain periods of'iambo-chor’ often appear at the initial position. See my note 16 above.

³² Wilamowitz already warns against treating ‘antispast’ as a metron (Verskunst, p. 236). Dale strongly and rightly criticized the old theory which analysed indivisible cola, in ‘Metrical units of Greek lyric verse, l’ (Collected Papers, pp. 42–3).

³³ This hypothesis is expressed in other words thus: ‘The choriamb which they all [= ‘chor dim’ and glyc] contain may occupy the first four, the middle four or the last four syllables’ (Dale, op. cit. (first ed.), p. 128). This is rightly denied by Denniston, op. cit. p. 121.

³⁴ Hel. 1481/1498

³⁵ The second str. pair of the first stasimon of Hec. ends with unambiguous ‘chor nucleus’ (474/483). This line may be ‘wil’ though I follow Murray's colometry . In any case ‘his ending is unusual. There is a number of blunt clausulae in tragic stanzas, but ‘most of them are dochmiac’ (Parker, CQ n.s. 26(1976), 17).

³⁶ Ant. 100 ff./117 ff., 806 ff./823 ff. Judging by metre only, Ant. 100 ff. is quite similar to Euripidean stanzas. Each line is glyc or equivalent arranged κατά στίχoυ Except for one line 806/823, the initial line, ‘ia-chor’ and ‘chor-chor’ are not mixed.

³⁷ Phil. 203 ff./212 ff., 1081/1101 f, 1123 ff./1146 ff. In the last example two ‘ia-chor’ cola appear (1136–7/1159–60). But the metre changes after 1130/1153. All other cola in this play are regular in type.

³⁸ Glyc/wil, Phil. 1082/1102, 1124/1147. ∧ glyc/∧ wil, Tr. 960/969. Cf. Pohlsander, H. A., Metrical Studies in the Lyrics of Sophocles (Leiden, 1964), p. 204, where, however, some uncertain cases are included and ‘ia-chor’ is named ‘chor. dim. B’.Google Scholar

³⁹ Dale, op. cit. p. 149.

⁴⁰ cf. Aj. 1185–91/1192–8. Though this stanza seems mixed with ‘iamb-chor’, it may be completely aeolic. Both ‘expansion of nucleus with choriamb’ and ‘longer compound’ (e.g. ia+glyc) are already found in the Lesbian poets.

⁴¹ OT 463–6/473–6

⁴² cf. Dale, ‘Metrical units of Greek lyric verse I, II, III’ (Collected Papers, pp. 41–97). Her basic idea about Pindaric metre may be seen at 53 ff.

⁴³ ol. 9 str. 6, Nem. 4.1 ff., Islhm. 8.1 ff.

⁴⁴ Snell doubts whether Sappho and Anacreon actually knew this free responsion and supports the change of texts (Metrik, p. 38). And if we accept Sappho 141 LP, is in responsion with (see my note 29 above).

⁴⁵ Corinna (London, 1953), p. 62.Google ScholarPubMed

⁴⁶ 29 χάριυ παλαι***υ χαρίτωυ (Powell). Cf. Eur. Or. 811 πάλαι παλαι***ς ⋯πò συμϕoρ***ς δόμωυ

⁴⁷ Grundriss der griechischen Versgeschichte (Heidelberg, 1930), §244.Google Scholar

⁴⁸ Page, op. cit. pp. 63 f.

⁴⁹ Nagy, op. cit. p. 37.

⁵⁰ Evadne's monody (Eur. Supp. 990 ff./1012 ff.) and Electra's parts of parodos (Eur. El. 140 ff./157 ff. and 175 ff./198 ff.).

⁵¹ I am very grateful to Mrs E. M. Craik for guidance and encouragement in my studies; and in particular for considerable assistance in drafting and revising this paper. I am grateful also to Dr C. Carey for helpful comments, to the anonymous CQ reader for constructive criticisms and suggestions for improvement, and to Professor R. M. Ogilvie, the editor, for sympathetic assistance.