^{page 241 note 1} See, most recently, Dale, A. M., ‘Resolutions in the Trochaic Tetrameter’, Glotta xxxvii (1958), 102–5Google Scholar, and Irigoin, J., ‘Lois et règies dans Ie trimetre iambique et le tetramètre trochaique’, R.E.G. lxxii (1959), 67–80.Google Scholar

^{page 243 note 1} References in this paper to Kraus are to Strophengestaltimg in der griechischen Tragödie and to Wilamowitz to Griechische Verskunst, unless otherwise stated.

^{page 244 note 1} See my article, ‘Porson's Law Extended’, C.Q. xvi (1966), 12.Google Scholar

^{page 244 note 2} Many would hold that there is no choice here because Euripides does not allow medial caesura without elision. Did an Athenian actor, then, when deciding how to articulate such a verse, take into consideration whether it was written by Sophocles or Euripides?

^{page 245 note 1} On the status of Maas is less illuminating than usual. Writing on the ‘Wilamowitz-Knox Bridge’ (Greek Metre, p. 95), he says ‘Semonides fr. 7. 79 and Hipponax fr. 29. 1 would be exceptions if there had to be word-end after words like and but there need not be (see 135).’ The reference intended must be to 137, where Maas notes the ambiguity in the treatment of not to 135, where he presents it simply as postpositive. More important, to say that ‘there need not be’ word-end after is to imply that it can be either prepositive or postpositive or both at once, which is incredible, yap must be either postpositive or an independent monosyllable. If it is a monosyllable, then, of course, the word-end after it is insignificant where regulations for the incidence of word-end within the verse are concerned.Google Scholar

^{page 246 note 1} On Or. 171 see below,p.248, and on Phoen. 1295 = 1306 see p. 261. Denniston (‘Lyric Iambics in Greek Drama’, Greek Poetry and Life. Essays presented to Gilbert Murray, Oxford, 1936, 141) lists a few other examples which I should be less inclined to accept as iambic dimeters, although some of them are certainly akin to the ‘impure iambics’ of Or., El., and Phoen. Ion 676 = 695 I should analyse as δ cr. = δ mol. (as an ia. dim. 695 would have word-end after long anceps, in violation of the extended Porson's law). I. T. 645 and 649 are astrophic, and their scansion, therefore, highly problematic. Ran. 1293, also astrophic and, presumably, quoted from Aeschylus, may be no more than a snatch. As it stands, it is hard to account for as an Aeschylean colon, and it seems hardly worth while to speculate on how Aristophanes might have analysed it.Google Scholar

^{page 247 note 1} See C.Q. xvi (1966).Google Scholar

^{page 247 note 2} Evidence for the tendency of prepositions (especially monosyllabic ones) to adhere to the following word, whether it is substantive or attribute, can also be gathered by observing caesurae and diaereses, e.g.: Ag. 546: or, in anapaests, Ag. 62:

I have not, however, found any complete and reliable study of this subject. The examples of ‘enclosed phrases’ (gescklossenen Ausdrücken) at verse-end in iambic trimeters where the division falls after the preposition have been collected by Tycho Mommsen (Beiträge zu der Lehre von den griechischen Präpositionen, Frankfurt, 1886).Google Scholar

The accentuation of enclosed phrases seems to have been in dispute since accentuation was invented. The view adopted by most modern authorities (e.g. Chandler, Greek Accentuation, 2nd ed., 915, Kuhner-Blass Bd. i, Teil i, 86. 3Google Scholar) is that when the substantive stands first and the attribute follows there is anastrophe, but not the other way about: while a proper name takes precedence over a common noun: This is the view attributed to Aristarchus in the Homeric scholia (Schol, . Ven. ad Il. 2. 839, 877Google Scholar). From the same source we learn that Ptolemaeus Ascalonites maintained exactly the opposite, while ‘Apollonius and some other grammarians’ introduced anastrophe in all cases (ibid. and on Il. 23. 718). In the examples collected by Tycho Mommsen of enclosed phrases divided between verses in tragedy, the MS. tradition is generally in favour of anastrophe. I have adopted anastrophe in the examples. quoted above, but without conviction.Google Scholar

^{page 249 note 1} I use the term ‘cretic’ to cover not only (Hephaestion's first and fourth but its resolved forms and paeons) and

^{page 250 note 1} Compare the metrical joke at Pax 180 pointed out by Hardie, W. R. (Res Metrica, Chap. V). There, Hermes, opening the door to Trygaeus, begins to speak, as becomes a god, in tragic diction. But the sight of the dung-beetle startles him into a prosaic exclamation and a violation of Porson's law.Google Scholar

^{page 251 note 1} On which see below, p. 254.

^{page 251 note 2} Lyric Iambics, p. 129.Google Scholar

^{page 253 note 1} ‘Sophoclea’, Journal of Philology xx (1892), 47.Google Scholar

^{page 253 note 2} Jebb's note on 866 remains the best defence of the manuscript text, although attempts to defend it have been made recently by Henry, A. S. (C.Q. xv [1965], 203CrossRefGoogle Scholar) and Kamerbeek, J. C. (W.S. Ixxix [1966], 84–6). Kamerbeek does not seem conscious of the verbal incongruity of and Henry quotes as parallels for the split resolution O.T. 202 (see above, p. 245) and 215. In both of these, however, the resolution is followed by short anceps, and at 215 there is not even a split, as counts as one word.Google Scholar

^{page 254 note 1} C.Q. xvi (1966), 16.Google Scholar

^{page 254 note 2} See above, p. 251.

^{page 254 note 3} Lyric Iambics, p. 129.Google Scholar

^{page 254 note 4} The Lyric Metres of Greek Drama, 2nd. ed. 1968, p. 82 n.Google Scholar

^{page 255 note 1} Lyric Iambics, p. 128.Google Scholar

^{page 255 note 2} The passage is discussed and other arrangements disposed of by Dale, A. M., ‘Lyrical Clausulae in Sophocles’, Greek Poetry and Life, pp. 187–8.Google Scholar

^{page 256 note 1} Elementa Doctrinae Metricae, Leipzig, 1846Google Scholar. lib. iii, cap. xxv. Hermann removes the split in 1307 and turns the colon into δ cr. by reading for The most likely parallel for the dochmiac is at Ant. 1273

^{page 256 note 2} This is a combination drawn from the conjectures of several scholars, including Heath, Dindorf, Schneidewin, Brunck, and Hermann. So many permutations have been produced that it is unprofitable to try to disentangle individual contributions.

^{page 257 note 1} Such correption is not, of course, impossible in iambo-trochaic, cf. Phil. 851Google Scholar and Track. 846–7, quoted on P. 258.Google Scholar

^{page 257 note 2} Cf. Dale, A. M. ad loc.Google Scholar

^{page 259 note 1} For a discussion of this colon and its affinities, see Dale, , Lyric Metres ², p. 102. In conversation, Miss Dale told me that she preferred to start from the view that it is a self-contained colon invented by Sophocles, and not further analysable. I should accept this, but if it must be verbally described, then ‘iambic tripody’ is truer to its nature than ‘ia. sp.’ (‘spondee’ being properly used for a doubly syncopated iambic metron: or cr. mol. (which, in any case, will not do at all for the version with short initial anceps). If it is permissible to hazard a guess at Sophocles' creative process, it might be suggested that appealed to him as a rhythmic phrase, and that, when he chose to vary it, he did so in accordance with the normal iambo-trochaic sequence: anceps, long, short, long, anceps.Google Scholar

^{page 261 note 1} For a complete rewriting, see Wilamowitz, , Griechische Verskunst, pp. 163 ff.Google Scholar

^{page 261 note 2} See above, p. 246.

^{page 261 note 3} See Wilamowitz, , Griechische Verskunst, p. 405. For other possible examples of such dochmiacs in Euripides, see Tro. 311= 328,. Her. 1057 (astrophic), also Rhes. 456 = 822. For a dochmiac of the form corresponding with see Bacch. 978 = 998.Google Scholar

^{page 264 note 1} This point is not noticed by Platnauer (ad loc), who considers that a catalectic trochaic dimeter ‘suits the system very well, even if it does not correspond’. If 389 is to stand, the next colon should begin with a consonant.

^{page 264 note 2} See above, p. 249.

^{page 265 note 1} On relative frequencies of types of dochmiac, see Conomis, N. C., ‘The Dochmiacs of Greek Drama’, Hermes xcii, Heft I, 1964.Google Scholar

^{page 265 note 2} See below, p. 266.

^{page 266 note 1} But this is a problematic passage. See Barrett, , ad loc.Google Scholar

^{page 266 note 2} Murray.

^{page 266 note 3} Hermann.

^{page 266 note 4} Dindorf.

^{page 266 note 5} codd.

^{page 266 note 6} On the problem of responsion in this passage, see Dale, A. M., Lyric Metres ², pp. 56–7.Google Scholar

^{page 266 note 7} To the examples listed, more or less doubtfully, by Conomis, (op. cit.)Google Scholar I add Eum. 844, O.C. 1561, Her. 758, Phoen. 1538. These seem to me no more improbable than some of his.

^{page 266 note 8} See above, p. 265.

^{page 267 note 1} Cf. Dale, A. M., ad loc.Google Scholar

^{page 267 note 2} So Schroeder at Phoen. 1538, but Murray's sequence of bacchiacs with as clausula is certainly plausible.

^{page 267 note 3} The only possible example is Rhes. 676 But Murray's is surely right. must then be a cretic with brevis in longo

^{page 268 note 1} See above, p. 250.

^{page 268 note 2} See, for example, Dain, A., Traite de metrique, § 46.Google Scholar