BRACKETS — TEMPORAL CLAUSES IN HERODOTUS. 191 



L. Meyer * and Gildersleeve f, the preterite rja and its forms are 

 always aoristic in Homer, but Delhriick shows conclusively that they 

 are sometimes extensive as well. He shows further that the present 

 stem of ei/Lu can express both extensive and aoristic action. $ As re- 

 gards the meaning of these preterite forms in Herodotus, in this passage 

 at least it is probable that eo-rpe is used aoristically.§ In !), 119: ol fi<? 



dp<pl tov 'ApravKTrjv vcrrepoi opprjdfvres (f>evyeiv, a>s KaTf\apj3dvovro oXiyov 

 eovres vnep Alyos TloTapmi/, d\t£dp(i>oi \p6vov ini o-v\i>6v ol pev dneBavov, ol Be 



fcovres (\dp(p6r]aav, it is certain that the action of KaTeKapfiavouTo was ante- 

 cedent to that of dnidavov and *\dp.<pdr]o-av. Here the action of d\e£6pevoi 

 was prominent in the author's mind ; and since several persons were 

 engaged in the flight, Herodotus represents the action of KareXapPdvovro 

 not as taking place at once but as continuing ; " Artayktes and his 

 men, who had set out to flee later than the others, after they were 

 being (had begun to be) overtaken," etc. The participle KaTaXapfiavo- 

 pevoi is used in exactly the same sense, 7, 211: al 8' av KaraX-ipliavopevoi 



l'7reo~Tp((pov, K.T.A. 



Let US next examine 8, 70 : eVei 8e naprjyyeWov dvan\eeiv, avr/yov rds 

 veas (n\ ttjv "2a\apiva, k.t.A. TrapfjyyeWov, A"BRP ; TrapfjyyeXov, A, V] 



TTaptiyyetXov, C. The temporal relation here demands the aorist, and 



* Griechische Aoriste, Berlin, 1870, p. 98. 



t Am. Jour. Phil., 4 (1883), p. 161, fn. 



1 Grundriss der vergl. Gram, der indog. Spraehen, 4, 69 ff. ; and cf. Brugmann, 

 Gr. Gram. 3 , 536. 



§ Cf. Brugmann, op. cit., 544, 1 : " Das Imperfekt versetzte die Aktion des 

 Praesensstamms in die Vergangenlieit. War diese Aktion punktuell, so besagte 

 dieses Augmentprateritum dasselbe wie der Aorist." This statement, if correct, 

 is somewhat startling. For if this be true, in the case of verbs whose present stem 

 expresses aoristic ("punktuell ") action, there is no appreciable distinction between 

 the aorist and imperfect, and hence no ground for discriminating between them in 

 the narration of past actions. Brugmann's words seem to imply that a verb of 

 aoristic meaning is just as likely to appear in the imperfect as in the aorist. But 

 he himself virtually says elsewhere (op. cit , 533, 541) that a purely aoristic stem 

 does not as a rule appear in the imperfect, nor a purely non-aoristic stem ("nicht- 

 punktuell") in the aorist. Certainly we do not expect to see the stem of opdw in 

 the aorist, nor that of tlSou in the imperfect, e^rjc and 1tt)v, taav the examples 

 by which Brugmann supports this theory, are not entirely normal in their use; 

 cf. Gildersleeve, Synt., 212; L. Meyer, Gr. Aor., pp. 95 ff. As regards the further 

 example v*6pi)v, in <>d., 4, 585, I regard it as by no means certain that the verb is 

 aoristic : ravra TeA(vT7)aas veopriv, StSoffay Si fj.oi oZpov dOdvaroi, k t.A. Why not 

 " I proceeded on my way " ? Observe too the imperfect Sidoaav. These examples 

 then are certainly insufficient to establish such a general law as Brugmann here 

 lays down. 



