^{page 182 note 1} Ran. 939–42; Rhet. III ii 5.

^{page 182 note 2} Numerous examples are collected in Burkhardt, H., Die Archaismen des Euripides (Diss. Erlang. 1906)Google Scholar; see also Krausse, O., De Eur. Aeschyl: instauratore (Diss. Jena 1905)Google Scholar and Kapff, , Die poetische Sprache der griech. Tragiker (Cannstadt 1893)Google Scholar.

The epic-ionic form πεπλὠκαμεν put in mouth of E. in Thesm. 878 is perhaps an allusion to this tendency; cf. E. Hel. 532.

Compare the contrast between the colloquial expressions often used by E. in stichomythia, and the artificial, non-realistic balance and regularity of this conventional form of dialogue, used by E. to a greater extent than by Aeschylus or Sophocles.

^{page 182 note 3} Such as the use of diminutives and of various obscene or abusive expressions confined to Comedy, and the irregularities of form which appear in vase inscriptions.

^{page 182 note 4} See below (17) and (19).

^{page 182 note 5} See Lottich, O., De Sermone Vulgari Atticorum maxime ex Aristophanis Fabulis cognoscendo (Halle, 1881)Google Scholar; Selvers, F.De Mediae Comoediae Sermone (1909)Google Scholar; Mancini, A., Il Dramma satirico greco (Annali della R. Scuola Normale di Pisa, Filosof. e Filolog., XI pp. 72–9)Google Scholar; Guarini, G., La Lingua degli Ichneutae di Sofocle (Aegyptus VI pp. 313–29)Google Scholar.

^{page 183 note 1} It may be so with e.g. ἦ γ⋯ρ and οὔτοι: see Denniston, , The Greek Particles lxxvii n. 1Google Scholar. For some examples of similar changes in English see Smith, L. Pearsall, Words and Idioms, pp. 150–2Google Scholar.

^{page 183 note 2} It is sometimes possible to observe the process of transmuting a common phrase into more poetic language; e.g. compare τ⋯ νεώτερον; with S. OC 1507 τ⋯ δ' ἔστιν, ὦ παῖ ∧αιου, ν⋯ορτον, αὖ and the colloquial use of ϕθε⋯ρεσθαι εἰς with E. Hip. 506 ⋯ς το⋯θ' δ ϕε⋯γω ν⋯ν ⋯ναλωθ⋯σομαι.

^{page 183 note 3} This has some bearing on the style to be adopted in translating tragic dialogue. There are times when Jebb's diction appears to be rather too poetic in tone: see on μ⋯λιστ⋯ γε below (14).

^{page 183 note 4} Wilamowitz, (on E. Her. 575)Google Scholar holds that the literary standard was different in Ionic, and refers to Hdt. IV 127 κλαἰειν λ⋯γω as an example of a phrase colloquial in Attic but not in Ionic. This actually occurs in conversation, where it may well be intended as a colloquialism; but there are certainly words used by Hdt. in serious narrative, but in Attic confined to colloquial contexts; e.g. παχ⋯ς ‘man of substance’ Hdt. v. 30, 77 etc., but in Attic only Ar. Eq. 1139, Pax 839, Vesp. 287.

^{page 183 note 5} Except for Eur., however, I have relied mainly on indexes.

^{page 183 note 6} The use with verbs is an intermediate stage; e.g. PI. Crat. 398E π⋯θεν, ὠγαθ⋯, ἔχω; ‘Of course I can't’.

^{page 184 note 1} Codd. τιμᾷ τιμᾷ · ὡς μ⋯ποτ'. … In C.R. Sept. 1936 p 116 J. D. Denniston gives fresh reason for supposing that ὡς is corrupt, and was inserted at some stage to mend the metre. His suggestion τιμᾷ γ' ἂ τιμᾷ is perhaps less likely than H.'s; the explanation of the corruption is a little less simple and the outspoken nature of the following wish suits something stronger than the equivocal expression. After a sudden out-burst, roused by the irony of τιμᾷ, Creusa recovers her calm in υ 288.

^{page 184 note 2} But the distinction is not in practice invariable; nor is it easy, here as elsewhere, to distinguish the uses of the moods on general principles.

^{page 184 note 3} In this play S. approaches nearest to E. in more realistic treatment of his subject and more conversational language. See Jebb, Intr. xli, and for colloquialisms see his notes on 327 (εὖ γε, the only instance in Tragedy), 577, 1049, and add 234 (τ⋯ κα⋯ γαβεῖν), 291, 205, 442 (iter. ἄν c. past ind.), 762 (βο⋯λει λ⋯βωμαι), 1006 (μηδ⋯ν ὑγι⋯ς), 1233 (οὔ τ⋯ που), 1229 (τ⋯ν ποῖον).

^{page 184 note 4} It is less likely that the first τ⋯ merely anticipates the second.

^{page 185 note 1} For the linguistic peculiarities which distinguish this play from the other extant plays of A., and are often reminiscent of Sophoclean and later usage, see Schmid, W., Untersuchungen zum Gefesselten Prometheus, pp. 68–77Google Scholar. These peculiarities, however, might well prove to be less numerous if more evidence for Aeschylean usage were available, and in the circumstances hardly justify the ascription of the Pr. to some other writer.

^{page 185 note 2} Said ‘ad eum quem stulte aut temere aut perperam agentem aliquid viderent’, Cobet, , Misc. Crit. p. 150Google Scholar.

^{page 185 note 3} Or possibly, keeping the traditional divisions, emend to κα⋯ σῷζ⋯ νυν … (κα⋯ and να⋯ are not infrequently confused).

⁴ We should not expect δ⋯ with a repeated question word, and there is some doubt about the MS reading; see Murray's crit. note.

^{page 186 note 1} So Stahl, , Syntax dir griech. Verb. I p. 305Google Scholar, ‘Es scheint also diese Gebrauch zuerst in der Umgangssprache aufgekommen zu sein’.

^{page 186 note 2} Several passages in Hdt., e.g. i 2 εἴησαν δ' ἂν οὗτοι Κρ⋯τες, are probably to be explained on similar lines, but generally seem to suggest rather than assert; cf. K.-G. I pp. 232–3.

^{page 186 note 3} E.g. on hearing a description, ‘That will (would) be so-and-so’.

^{page 186 note 4} Suppl. 764 ϕα⋯ης ἄν, εἰ παρ⋯σθ' ⋯τ' ἠγ⋯πα νεκρο⋯ς is different and may be an echo of the Homeric ϕαἰης ἂν.

^{page 187 note 1} Eq. 414 is doubtful, and Richards', H.⋯κτραϕε⋯ς ἦν (C.R. xvii p. 143)Google Scholar may well be right.

^{page 187 note 2} In Ph. 99B, cited by Stahl, (op. cit. II 673)Google Scholar, τ⋯ μ⋯ διελ⋯σθαι should probably be taken as an instance of anacoluthon.

^{page 187 note 3} See Mommsen, T., Beiträge zu der Lehre von den griech. Präpp., II pp. 76 ffGoogle Scholar.

^{page 187 note 4} On the periphrastic uses of περ⋯ and κατ⋯ see Rudberg, G., Ad usum circumscribentem praep. Graecarum adnotationes, Eranos xix pp. 173 ffGoogle Scholar. (mainly on post-classical usage), and C.Q. xxx (1936) pp. 212–15.

^{page 188 note 1} On ⋯παινεῖν so used in E. see Amati p. 142, and Wilam. on Her. 1235.

^{page 188 note 2} Esp. with μολεῖν, which is poetic except for 16 exx. in Ar., all in lyrics, parody, or dialect speech; for evidence of its Doric origin see Gautier, , La Langue de Xénophon, p. 30Google Scholar.

^{page 188 note 3} Cf. the almost purely adverbial use in Dem. xxiii 143 το⋯το … εὖ πο⋯ουν, οὐ συν⋯βη ‘fortunately’.

^{page 188 note 4} See Mayser, , Grammatik d. griech. Pap. aus der Ptolemäerzeit, II p. 174Google Scholar.

^{page 189 note 1} It may be assumed that the combination of ⋯ς καλ⋯ν and στεἰχειν produced no impression of incongruity. καλ⋯ς, though here and elsewhere used in colloquial phrases, is also, of course, common in quite different contexts, and does not in itself strike a note of triviality; and on the other hand στε⋯χειν (like μολεῖν), though not used in prose, was so common in tragic dialogue that in that context it would seem to be the normal word and not especially poetic in tone.

^{page 189 note 2} For the form see L. and S. s.v.; and on form and meaning, esp. as discussed in ancient grammarians, see Uhlig in Rhein. Mus. xix pp. 33 ff.

^{page 190 note 1} συγκατ⋯θεσις μ⋯ν τ⋯ν ε⋯ρημ⋯νων, συναϕ⋯ δ⋯ πρ⋯ς τ⋯ μ⋯λλοντα, Timaeos, , Lex. Plat. p. 94Google Scholar.

^{page 190 note 2} For a purely poetic variation cf. E. Alc. 229 (lyr.) κα⋯ πλ⋯ον ἢ βρ⋯χψ δ⋯ρην οὐραν⋯ψ πελ⋯σσαι. For other references to suicide by hanging see E. Andr. 811, Tro. 1012.

^{page 190 note 3} To these Her. 714 must be added, if Murray's reading is accepted: τ⋯ χρ⋯μα δ⋯ξης; το⋯ δ' ἔχεις τεκμ⋯ριον;