¹ The fullest and most consistent treatment of catalexis that I have found in any current handbook is that of A. Dain (Traité de métrique grecque, Paris 1965, 34–5). He, however, uses the term ‘syllable’ for ‘element’, and does not explore the rhythmic function of the phenomenon. And here, as elsewhere, he becomes over-systematic and theoretical.Google Scholar

² Cf. Dale, A. M., The Lyric Metres of Greek Drama ² (Cambridge 1968), 19: ‘The ear is conscious not merely of the curtailment but even more of the contrasted close.’Google Scholar

³ Chapters on Greek Metric (New Haven 1901), 22–3.Google Scholar

⁴ Griechiscbe Verskunst, 249: ‘… die Bildung einer Klausel durch Zusatz ebenso nahe lag wie durch Abstrich; beides ergab den erwunschten klingenden Schluss.’

⁵ Aj. 348 ff. is not comparable. There, appears as clausula to a stanza which includes dochmiacs, but it does not follow directly on them, and the dochmiacs themselves are of the forms / , and, once, . This is merely an example of the colon acting as an all-purpose clausula.

⁶ BICS 5 (1958), 13–24.Google Scholar

⁷ Snell, (Griechische Metrik, 1962) still calls the dactylic hexameter ‘catalectic’, although he does not appear to subscribe to the ancient doctrine which would make it so. This is a notable (and surprising) example of the confused treatment of catalexis in current metrical handbooks.Google Scholar

⁸ WS 77 (1964), 15 ff. and Collected Papers, 185 ff. As there is a confusing misprint on p. 18 of the original publication in WS, references to this article will be given with the pagination of the reprint in Collected Papers.

⁹ 15. 8, Consbruch, p. 50.

¹⁰ Tarditi's suggestion (ad loc.) that the corruption was introduced in the process of abridging Hephaestion is ingenious, but, as I have argued in Lustrum (15, 1970, 50), hard to account for. It is easier to assume corruption in the text of Archilochus, although, in the absence of the context, impossible to emend with any degree of confidence. See West's apparatus for possibilities.

¹¹ De metris poetarum graecorum et romanorum, I. 13. 41–2.Google Scholar

¹² Pap. Colon. inv. 7511. See Merkelbach, R. and West, M.L., ‘Ein Archilochos-Papyrus’, ZPE 14 (1974), 97–113. The presence of three-verse stanzas in Archilochus is also interesting in that it brings him a little closer technically to the Lesbian poets.Google Scholar

¹³ I do not see the force of Dale's objectioi (‘Observations’, 187) that the last element cannot be ‘contracted long’ (i.e. single-long biceps) because that ‘by definition may resume its two shorts, so that the sixth metron is unlike the remaining five’. This is altogether too abstract. It is perfectly normal for rules of internal responsion to be modified at verse-end. Thus, in the iambic trimeter the final long is not resolvable. We discussed this point soon after the publication of her article. As I remember, she was prepared to allow that the idea of a regularized cadence was plausible, but maintained that her objection was valid. The few dactylic cola in tragic lyric ending in double-short biceps and hiatus are listed by Korzeniewski, Griechische Metrik, 74. His treatment of verse-end in dactyls, however, differs considerably from mine, because he includes as dactylic ‘prosodiac’ and ‘enoplian’ cola which I should regard as essentially aeolo-choriambic.

¹⁴ Dain and Raven regard… / as the catalectic form for dactyls. Raven's list of ‘catalectic’ dactylic cola (G.M.² 49) is interesting, but not all the examples are beyond dispute. Pers. 882 (= 891): is probably not a separate colon, since the passage continues …, so that the proposed division is actually in the middle of a dactylic metron (). Dale (L.M.², 40) while making the point that the dactylic sequence is uninterrupted, prefers Murray's layout on the page, following word end in preference to metron-end. But that is not, strictly speaking, a colometry. Eum. 1042 (= 1046), however, is a genuine example for it coincides with verse-end, shown by brevis in longo in the strophe and hiatus in the antistrophe. Eum. 533 = 545, not one of Raven's examples, is also unambiguous, since the next colon begins ….

¹⁵ Cf. Dale, L.M.², 51.

¹⁶ L.M.². 72.

¹⁷ Ag. 197–8 = 210–11, Eum. 919–20 = 941–2 (accepted by Dale, L.M.², 95), Pers. 136 (there is word-end after the bacchiac in the strophe, 131), Sept. 735 = 744, Supp. 95 (or 90, if Westphal's transposition is accepted, as by Page. The corresponding verse has word-end after the bacchiac), Supp. 136 (where the text of the antistrophe at 146 is in doubt), Soph. El. 482–5 = 498–500 (a singular passage. There is a bacchiac at 485, while in the other three repetitions of the verse the syncopated metron is a molossus), Hec. 946–7, Phoen. 686, Tro. 1295–6 (all three Euripidean passages are astrophic). The examples with elision are: P.V. 183 (word-end without elision in the strophe, 165), Hel. 637 (astrophic), Med. 646 (word-end without elision in the antistrophe, 656), Phoen. 1027 (word-end without elision in the antistrophe, 1051), Tro. 1322 (word-end without elision in the antistrophe, 1307).

¹⁸ The comparative figures for clausulae to stanzas in tragedy are: Aeschylus: 111 pendant, 60 blunt; Sophocles: 74 pendant, 10 blunt; Euripides 153 pendant, 66 blunt. Most of the blunt clausulae are dochmiac. There are, in addition, a few ‘dragged’ clausulae (… ). Pindar's preference is not so strong: he has 48 pendant clausulae in the epinicians to 33 blunt. In the dactyloepitrite poems alone, however, the proportion is higher: 29 to 15. Dactylo-epitrite has no catalectic form, so the preference for pendant clausulae manifests itself independently of catalexis.

¹⁹ Page's colometry follows word-end more closely in the first half of the stanza, producing two trimeters: which, formally, has the effect of bringing one to the beginning of a colon. The other two, however, remain embedded in mid-verse. To place all the syncopated metra at colon-end, as in my colometry, probably makes the verse easier to read rhythmically. The problem of layout here is a modern one. In performance, one supposes, the passage must have run: Or, avoiding any implication about musical realization, one can notate the rhythmic pattern a b a b a a.

²⁰ ‘Complex’ (where applies to the subordinate clause and to the main clause, as here) is, according to Denniston (Particles, 98–9), ‘exceedingly rare’. Bentley's transposition ates the rarity and produces a pair of identical cola of the form (cf. Thesm. 1034–5, with Dale's comment, quoted above, p. 20).

²¹ The traditional colometry of Pindar includes a number of demonstrable blunders produced in the pursuit of neat patterns of uniform cola. At the opening of Pyth. 5 any of us might be seduced by: had not Boeckh's method of locating verse-ends shown that the true division is: .

²² Also, possibly, Vesp. 318–19. See above n. 20.

²³ This is awkwardly put, for a syllable can hardly be made ‘prosodically long’ by a following pause. Short syllable+pause could add up to the time-value of a long, as crotchet note+crotchet rest add up to a minim.

²⁴ ‘Anceps: vocale, sillaba, elemento’, RFIC 91 (1963), 52 ff.Google Scholar

²⁵ BICS Suppl. 21. 1 (1971), in progress.Google Scholar

²⁶ It is, I think, possible to make out an aesthetic case for pause after : unlike the other epithets, its function here is to complete the meaning of the verb (‘abides re-arising’). That it should be rhythmically marked out in some way is not inappropriate. In any case, Dale is subscribing to the highly dubious interpretation of as adjectival. Fraenkel (ad loc.) puts with overwhelming force the case for the view shared by most modern editors (Wilamowitz, Verrall, Denniston and Page) that it is substantival in apposition to : ‘There abides, terrible, re-arising, a treacherous housekeeper: Wrath…’

²⁷ J. Irigoin (in W.J.W. Koster, , 1967, 65–73) suggests an intermediate situation: obligatory word-end without metrical pause. A similar concept seems to underlie Dale's ‘obligatory colon-end’. I do not see how this hypothesis can be either proved or disproved (cf. Lustrum 15, 1970, 52). The practice of Pindar, at least, suggests straight alternatives; either the possibility of synartesis, or metrical pause.

²⁸ Parker, , CQ 16 (1966), 20–4.Google Scholar

²⁹ The formulation proposed by W.S. Barrett in his edition of the Hippolytus (Appendix 1, p. 422) does introduce anceps into the post-choriambic part of the colon, but in fact the ‘dragged’ close () is so exceptional that it seems to me mislead ing to notate . See also Dale, L.M.² 153.

³⁰ ‘Notes on metrical analysis’, 89. Professor Dover introduces some other refine ments which it is not relevant to discuss here. He in fact uses to mark vowel-sounds shortened by epic correption. He also uses Maas's finale mark, , for final elements in abstract metrical schemes, as distinct from scansions. This is particularly useful for the description of stichic metres and ready-made stanzas, such as the sapphic and alcaic, to show points where synaphea is generally broken.