370 ON THE LANGUAGE, METRES AND PROSODY metre requires it. ^schylus does not hesitate to substitute a for v in the 3 pers. pi. of the optative middle, as in iKo-io^oLaro for €K(tw^olvto (Pers. 449). We have also occasional lonisms like vr]6<s for vews (JEsch. Fers. 424), rjixrjv (Soph. Track. 24), kclvov (ibid. 495), kUl<s (^sch. Choeph. 678), tVevos (Soph. Phil. 494), KovX^i^iv (Soph. Aj. 730), yv6ov (Eurip. Electr. 593). The pronoun generally used as the article appears in the oblique cases as a substitute for the relative (^sch. Agam. 628, 642; Choeph. 596; Eumen. 322, 878, 919; Su2-)pl. 262, 301, 516, 579; Soph. Phil. 1112; (Ed. Col. 35-, (Ed. R. 1379), and in the demonstrative use we have even rot 8e for ot Se (^sch. Pers. 424). The use of vlv for avrov is common enough, and we even find ^xiv (Soph. Trach. 388). The reflexive cr<^€ is a perfectly general pronoun of reference in ^schylus (e. g. it is = au- Tov, Sept. c. Theh. 451; avrov, Suppl. 502; avras, Sept. c. Tlieh. 846). It is extremely doubtful if a^iv can be used for oi. In -^sch. Pers. 759, Soph. (Ed. C. 1490, it may be understood as for cr(}>LaLv. It is also an open question whether such a form as eXeetvos is allowable in the Greek dramatists (Pors. Prcef. Rec. p. 7; Lobeck ad Soph. Aj. 421). The rare forms Tjo-v^torcpos (Soph. Antig. 1089) and <f)iLa-To<: (Soph. Aj. 842) may perhaps be regarded as Ionic. Also Kpv^ets for K^vySet? (Aj. 1124). There can be little doubt that an epic tradition suggested the occasional omission of the augment in the speeches of the messengers (Matthia, Gr. Gr. § 160, Obs., see below, iv. 1). Uncontracted forms such as evpoosy v6o<5, pUOpov, are sometimes though very rarely found in the dramatists. Valckenaer rejects the particle T^Se for Kai (ad Phcen. 1683), but it occurs more than ten times in -^schylus, in two fragments of Sophocles (345, 493, Dind.), and in Euripides, Hec. 323, Here. Fur. 30. II. jEolic Forms in the Dramatists. The most common holism is the substitution of TreSa for /xera in compounds, such as TrcSapcrio?, TreSaopos, TreSat'xjW'io?, and this occurs even in dialogue (^sch. Prom. 711; Choeph. 843; see Yalcken. ad Eurip. Phoeniss. 1034). We have also fidaaoiv (^sch. Pers. 432, 694; Agam. 584), yXvcraro)v (Aristoph. ap. Etym. M. p. 235), and similar forms, if these are to be regarded as ^olisms. A more decided instance is sup- plied by opavLav, which the metre requires in the Suppl. 788; cf Alcseus : vei fjikv 6 Zev?, ck 8' opavw /xeya? ')(€L}X(av. And see Buttmann, Lexil. p. 200, Engl. Tr. III. Doric Forms in the Dramatists. In the choruses, for the reasons already given, a certain amount of Dorism is invariably found, such as the substitution of a for -q, e.g.
