186 PROCEEDINGS OF THE AMERICAN ACADEMY. 



220, does not count, as the sentence is part of an oracle in hexameter 

 verse; and in two places (1, 32, 82) the MS. authority is divided. 



As to the usage of other authors, the three tragic poets omit av with 

 Trplv and the subjunctive ten times, three times in generic action, seven 

 times in particular.* Thucydides omits av with -rrpiv five times. f av is 

 omitted with rrpiv in the lyric poets once (Simouid. Amorg., 1, 12). 

 Excluding, then, the three places where the MSS. disagree, and 7, 220, 

 av is found in all MSS. eight times, and omitted in all MSS. twice. So 

 far, then, this furnishes a supposition in favor of the use of av. In 1, 32 

 (■rrpiv av, A' 2 Rbdz ; -rrpiv, ABCP), and 1, 82 (nplv av, A 2 Rdz ; -rrpiv. ABCP) 

 I would read av with Sturm. It is harder to account for the insertion of 

 av than its omission. Moreover, not infrequently in Herodotus the cor- 

 rect reading is found in only a small number of MSS. J 



It now remains to consider the passages where av is omitted in all 



MSS., 4, 157: ov yap 8r) crcpeas drrUi 6 8eos . . ., nplv 8r) uiriKoovTai is avrrjv 

 Aifivrjv ; and 6, 82 : 7rpos civ raiira ov 8lkciiovv -rrtipav tt)s ttoXios, npiv ye 

 8q Ipdio-i xpwrjrai, k.t.X. Sturm would read np\v av in both of these places. 

 His statement, however, "die Partikeln 8r] und ye 8r) . . . finden sich 

 in Verbindung mit diesem Modus bei keinem Schriftsteller vor," is 

 not accurate. 8r) is found with the subjunctive in II., 5, 684 (pr) 8r) 

 pe . . . edo~T]s) and II., 7, 24 (Iva 8r) . . . 8wt), and in Herodotus 1, 29 

 (tva dr] prj . . . dv.iyKao-drf), and ill 1, 32 (Iva 8r] al a>pai o-vpfiaivaso-i). For 



the repetition of 8rj within a short space, which Sturm calls " storend," 

 cf. Her., 1, 43 : i'vBa 8fj 6 £elvos, ovros 8rj 6 Ka$apde\s t6v (povov, k.t.A., and 



G, 83 : e« tovtov 8r/ 7rdXf/xdj cr^)t fjv e7rt xpovov cru^wi', es o 8f] poyis. K.r.X. It 



is worth noting that both in 4, 157, and G, <S2. the npiv clause follows a 

 past tense, and the subjunctive might, on the principle of oratio obliqua, 

 have been changed to the optative. The meaning of 8r] in 4, 157, seems 

 to me to be entirely appropriate : " for in fact the god would not let 

 them off till they actually had come to Libya itself." Although the 

 change of 8r) to av may be right and is tempting, the evidence is not suffi- 

 cient to warrant us in making- the a. Oration. 



* Of these seven cases, five are in Euripides. Cf. Clapp, Trans. Am. Phil. Ass., 

 22 (1891), p. 01. 



t The use of &v in Thucydides is discussed by Warren, Temp. Clauses in Thuc, 

 p. 19, who concludes that the omission of &v with npiv in Thucydides must for the 

 present be admitted. 



t It is a little difficult to understand upon what principle Holder, in 3, 109, 

 writes " &v 0, om. a," and reads &v, while in 1, 32, he writes " irp\v &v /3, irpiv a," but 

 omits &v. 



