182 PROCEEDINGS OP THE AMERICAN ACADEMY. 



present in both. Conditional relative sentences are found in Homer 

 without av or *<f, but no case is cited by writers on this subject of such an 

 omission with a temporal clause of antecedence which refers to future 

 time.* There is no such case in Thucydides. As regards the use of 

 the tragedians, in the case of relative conditions that have the sub- 

 junctive without av, exclusive of clauses introduced by a word mean- 

 ing " until," " the generic meaning preponderates over the future in 

 the overwhelming proportion of 33 to 2."| Furthermore, out of 52 

 cases in the tragedians of the omission of av with the subjunctive intro- 

 duced by relative words, there is not one case witli eW or ineidt] in a 

 clause of future meaning, and only two cases in clauses of generic 

 meaning (both with eVet). Ilartung X cites besides Od., 15, 453, only 

 Anacr., 29, 8 and 29, and the passage under discussion. Hartung 

 refers apparently to Anacreontica, 16 (Bergk), v. 8: ilfas, wt 6i\u>ai, 

 Keladat,, and v. 42 : Xa/3e fxiaflov ocraov emj)s. These passages are how- 

 ever much too late to have much weight in justifying the omission 

 of av in the passage from Herodotus. In the Greek lyric poets av is 

 omitted in clauses of antecedence but twice, Mimn., 1, 5, and 11,4 (ed. 

 of Hiller-Crusius, 1897) ; in both of these cases the action is generic. 

 The case in Herodotus seems therefore to be the sole case in classical 

 Greek of such an omission of av, and as such is in the highest degree 

 suspicious. 



Kriiger § in a note ad loc. wrote : entire, enedv ? but did not explain 

 how enedv could have become ineire. I believe that Herodotus wrote here 

 enelre av dveveixdj). If av had been written it would necessarily have 

 followed eneire (cf. 1, 200, and 202), and could have easily been lost 

 before dveveixdf) by haplography. A striking parallel to such a loss of 



av is found in TllUC, 7, 77, 5 : firj aXXo re fjyrjo-dnevos eKaGTOs rj e'v <o a v 



dvayKaadfi xcopi'a> /idxea-dai, k.t.X., where civ, which is read by all recent 



* &v is omitted in Homer in clauses with eirei but twice, and in each case the 

 main verb expresses generic action. Cf. Schaub, De Usu Coniunctivi et Optativi 

 in Enuntiatis Lyricorum Graecorum Secundariis, Liestal, 1889, p. 15. 



t Clapp, in Trans. Am. Phil. Assoc, 22 (1891), p. 89. 



X Cf. Hartung, Partikeln der gr. Sprache, Erlangae, 1833, vol. 2, p. 296, who says, 

 speaking of the use of &v in relative clauses which express particular action : " Sie 

 (i. e., &v and ««') sind hier dermassen an ihrem Platze dass sie schon bei Homer nur 

 hochst selten weggelassen werden ; denn es ist mir keine Stelle ausser Od. o, 453, 

 gegenwartig." But in this passage : onrj Trepda-nre, Monro regards mpda-^re as 

 post-Homeric both in form and syntax, and reads irepdaatTe. See Monro, Homer's 

 Odyssey, Books 13-24, Oxford, 1991, crit. note on 15, 453. 



§ Herodotus, 1856. 



