² R. D. Dawe (ed.), Sophocles Tragoediae i (Teubn., 1975, 1984). The relevant apparatus, after Dawe, is: LFHGR, CZc, NPV, O, ADXrXsZr, T; Hermann, Erfurdt. The corresponding verses (the opening of the Parodos) are: (121–123). An alternative colometry wil wilsp (uncommon, but cf. n. 18 below) is theoretically possible; but wil wil J... is plainly superior, with the longer third verse presumably completing a period. For the two-syllable overlap, cf. Ant. 332f. , etc.

³ H-Jones Lloyd and Wilson N. G.(edd.). Sophoclis Fabulae (Oxford, 1990). They also accept Schwerdt′s unappealing in 123, a verb which seems nowhere to be used of lamenting utterance. In support of (codd.), cf. Jackson, J., Marginalia Scaenica (Oxford, 1955), 206f. (on Phil. 190).Google Scholar

⁴ Sophoclea (Oxford, 1990), 46.

⁵ Stinton, T. C. W., JHS 97 (1977) 128–129, 132 = Collected Papers on Greek Tragedy (Oxford, 1990), pp. 273, 278–279.Google Scholar

⁶ J. Diggle, CR 33 (1983), 346–348 = Euripidea(Oxford, 1994), pp. 258–260.

⁷ Phil. 208–209 Dindorf) 217–218 (Wunder, for see further below.

⁸ Med. 159 ; see further below.

⁹ Itsumi, K., ‘The glyconic in tragedy’, CQ 34 (1984), 66–82.CrossRefGoogle Scholar

¹⁰ Their partly new triple emendation (after Dindorf and Hartung) is advocated in Sophoclea, 183–184, now against Stinton and Diggle, and with no mention of El. 123/139. The double excision of is no improvement on Wunder′s easy (regarded by Stinton as ‘certain’). As to , I should prefer (corruptible, cf. Phil. 1401), since the ‘weary’ utterance heard as Philoctetes approaches is scarcely ‘threnody’.

¹¹ Art. cit., 81C: p. 75 is concerned only with ‘dragged close’ in the glyconic itself.

¹² Art. cit., 78–9. Itsumi does not identify his ‘twenty-three occurrences’, and there is sometimes colometric doubt, e.g. whether Hp. 130/140 has ten syllables (with Stinton′s overlap) or nine (as Barrett, etc.). In my colometry (* = not as in the latest Oxford Texts) the following are gl, not counting El. 123/139 and Phil. 209/218: Aj. 603/615, Ant. 816/833, 846/865, *947/958, A. Su. 46/56, E. Med 159/183 (with prefixed cr), Su. 957/965, IT. 1093/1110, Ion 1060/1073. wil occurs at Hec 925/935 (with prefixed ia), El. 174/197,434/444. For the related t1(gl), as Hp. 130/140, see n. 17 below.

¹³ gl + ba is the ‘phalaecian’, occurring at Aj. 634/645, 697/710, E. Held. 758/769, Su. 962/979, Or. 833; the similar wil ba occurs at Ant. *105/122, 1145/1154, Whil. 139f./154f. (wil: ba: gl), E. Hp. 547/557. El. 432/442, 736/746. Ion 1052/1065, Hel. 1464/1477, Or. 810/822. Other suffixes (ch, ia, mot) are uncommon; of the prefixes, ia and cr are quite frequent, ba only in Soph. (Itsumi, art. cit., 79–80: add Phil. 140/155 as a possible ba gl).

¹⁴ The changed position is recognized by Diggle (as to Med. 159) in Eikasmos 6 (1995), 41 n. 8.

¹⁵ Phil. 209/218 remains variously uncertain (n. 10 above, but there is more to be said also about the colometry ot 206– 209/215–218; I hope to discuss this further in another place), should be read in Med. 159 (Prof. Diggle gives me his assent). Conjectured by Tyrwhitt after Brunck, it has some manuscript support of which Stinton was unaware (). The other three occurrences of in tragic lyric (all Eur.) are similarly in places where is at least as likely as : at El. 1171 and Hyps. 64.78 terminal in dochmiacs (in both resting on the evidence of a single witness); at Or. 1392 terminal in the verse (where again is a thinly attested variant). A good case can be made for reading or restoring in all four places.

¹⁶ ch ia : Rh. 466/831 (cf. ch ia ba at 457/823). iach : Hp. 147/157. ia is common in Soph.; this and 2ia appear first in Alcman, cf. 3ia in Hipponax and Ik in Stesichorus (West, Greek Metre, 52–53). There was evidently ancient precedent for terminating a verse with a spondee.

¹⁷ occurs also (in my colometry, cf. n. 12 above) at Aj. *196, O.C. *520/533, E. Ale. 576/586, Hp. 130/140, *552/562. In this case there is indeed an apparent correlate with short penult, at Aj. 399/416 and E. Ale. 443/453: but this latter nine-syllable verse is essentially different, to be analysed as × d × e, analogous to × e × d at O. T. 870/880 and E. Ale. 573/583.

¹⁸ The cadence may sometimes be as well or better regarded as generated by the addition of one syllable (‘hypercatalectic’) to a verse ending with The short strophe 2ia at Tra. 947–949/950–952 is a case in point, where the third verse is more naturally taken as a prolongation of ar (ch ba) than as dod + sp. Similarly in ionic, terminal is naturally taken as a prolongation of. . But even if hypercatalexis is the preferred explanation, there can be no reasonable objection to the convenient notations gl sp and tl sp as equivalent respectively to hi– and hag –.

¹⁹ My attention has been drawn to Dawe′s latest thought (Teubner Einzelausgaben, 1995), which is open to the same (not in itself fatal) objection.

²⁰ for is technically plausible, but ) lacks sufficient attestation: only Hsch. codd. also is known only from Hesychius.

²¹ See my commentary (1989, with Addendis Addenda), pp. 141–142, 365; I am now more inclined to read (with Diggle).

²² A date for S. El. shortly before E. Tro. and I. T. would be consistent with other indications (cf. my commentary on Or. p. lvi n. 91), reinforced by further metrical studies which I hope to publish in due course.

²³ A reader objects that ‘no one sought to raise dead fathers by ’, and another suggests ‘if one is to go down this avenue’. The objection is misconceived: the argument is simply ‘No amount of lamenting will bring your dead father back from Hades’, as a self-evident a rhetorical point similar to in the Parodos of Euripides′ Electra (193–195). No one sought to gain the upper hand over their foes by . As to the suggested , ‘loud utterance(s)’ is relatively weak sense, and typically corrupts to . More seriously, like is a word used only in the singular.