BRACKETT. — TEMPORAL CLAUSES IN HERODOTUS. 183 



editors, is found in only one MS., the Vaticanus, being omitted in A, C, 

 E, F, G, and M.* 



I am aware that the doctrine lias been advanced that av is sometimes 

 omitted owing to the presence of the same sound in another word in the 

 sentence. | But until this theory is more firmly established for prose 

 than it is at present, it cannot have much weight in determining the 

 usage of prose writers. % For the repetition of the same sound in 



Herodotus, cf. e. g., 3, 117: 8ia 8iacr(payos dyoptvos (KaaTr/s (icdcTTOicn- 



cos occurs with the subjunctive in Herodotus twice : 1, 132, and 4, 172 ; 

 in both cases the clause expresses generic action and av is omitted. The 



examples follow. 1, 132: tcov 8e cos endo-rcp Bvnv dtXrj, . . . KaXcti tqv 

 deov, k.t.X. ; 4, 172 : tcov 8e cos eKacnos oi pixdrj, 8c8oi 8a>pov to av i'^rj (pepopevos 

 ££ o'ikov. 



According to Herodotus's usage elsewhere and according to the 

 general law of Greek grammar we should expect av in both of these 

 clauses with cos. Cf. with 4, 172 a passage in 1, 199, encdv 8£ pt\G^ k-tX 

 Bahr § attempts to explain the passage in 4, 172, by saying that he would 

 take cos closely with eKuo-ros, but he does not make it plain how he 

 would translate cos. It is clear that cos is here a temporal conjunction 

 meaning "after,'' Latin simulae, which is the opinion of Sauppe, 

 Hartung, and Hermann. If 



As regards the usage of other Greek authors, as has already been 

 pointed out,** cos is almost never used in classical Greek with the 

 subjunctive or optative. The only other examples seem to be Sappho, 

 2, 7 : ws o-e ydp FiSw, (Spoxtcos pe epebvas ov8ei> eV e ik«, and Cebes, Tab., 4, 3, 

 (ed. of Drosihn, 18/1): TrpocrTarret 8e roif elcnrupevopevois, rl Set avrovs 

 iroifiv, ws av elo-e\8a>o-iv els rbv fiiov. The authorship of the Tablet of Cebes 

 is not fully determined, but the work probably does not belong to the 

 classical period. 



* For the loss of liv by haplography cf. also Ant., 1, 10: avayKcifyi ra yeyovora 

 KCLTriyopuv, where Giklersleeve would, correctly, I think, insert av before avayKcifyi. 

 Cf. Giklersleeve, Syntax, 450. 



t Wright, Harv. Stud, in Class. Phil., 12 (1001), 145 ff. ; Gildersleeve, Syntax, 

 450, Am. Jour. Phil, 12 (1801), p. 387. 



t Hermann, Opusc, 4, 103, attempts to justify the omission of av in our passage, 

 but his argument is entirely refuted by Baumlein, Unters. iiber die gr. Modi, p. 237. 



§ Herodoti Musae, ed. alt., J. C. F. Bahr, Lipsiae, 1857, note ad loc. 



IT Sauppe, Epist. Crit. ad Herm., p. 89; Hartung, Partikeln der gr. Sprache, vol. 

 2,296; Hermann, Annot. in Vigerum, Lipsiae, 1834, p. 941. 



** Cf. Klotz, Devarius, Lipsiae, 1840, vol. 2, p. 759; Gildersleeve, Am. Jour. 

 Phil., 7 (1886), pp. 1G7 and 543. 



