180 PROCEEDINGS OF THE AMERICAN ACADEMY. 



and the tense is the aorist. Lastly, in 4, 42, contrary to Herodotus's 

 regular usage,* the whole passage in oratio obliqua is very short and 

 no infinitive in oratio obliqua has preceded. 



These reasons are, I think, sufficient to make it doubtful whether the 

 current interpretation of this passage is correct. All these difficulties 

 can be avoided however by a change in the interpretation of one word. 

 Is not eW to be taken closely with ej, in the sense of the Latin usque 

 ad f (iTriKveeo-dai will then be co-ordinate with diennXtfiv, and we may 

 translate : " Necos despatched the Phoenicians in boats, commanding 

 them to sail, on their homeward voyage, through the Pillars of Hercules 

 clear to the Northern Sea, and in that way to get back to Egypt." 

 It should be noted that a few lines later, Herodotus, writing of this 

 same event, says : ware . . . rpirot eret Kapy\ravT(s H/onxXed? <TTrj\as dnl- 

 k%vto is A'lyvTTTov. ewj is used with ds by Polybius (1, 11, 14; 

 34, 4, et al.) and in the Septuagint (Num., 17, 13; Mac, 1, 2, 58). I 

 am aware that no examples are given of ems used in this sense, as early 

 as Herodotus. But Xenophon uses axpi (Anal)., 5, 5, 4), pe^pi (Anab., 

 (!, 4, 26), and eVre (Anab., 4, 5, 6) with this meaning, and I see no a 

 priori reason why Herodotus may not have used Has in the same sense. 

 The difficulties in taking tut = usque are, to my mind, much fewer than 

 those involved in the common interpretation.! 



The only other case where the infinitive may be thought to stand for 

 the subjunctive in oratio obliqua is 1, 165: upocrav pr) np\v h Qoxaiav 

 ij^eiv nplu rj tov pvdpou tovtov dv.i(pivr)vai (MSS. avcKpijvai). The use of the 



infinitive here is very strange. First, if we exclude the passage in 4, 42, 

 it is the only case in Herodotus where the infinitive is used in oratio 

 obliqua in a temporal clause which refers to future time. Exclusive of 

 this passage, after a negative main clause in a clause referring to the 

 future, irp\v rj is used six times (1, 19 ; 4, 9 ; 6, 133 ; 7, 8 (3 ; 9, 93, 1 17). 

 Curiously, in all but one of these cases (7, 8 j3) the clause is in oratio 

 obliqua, but in each case the subjunctive is retained. Secondly, as we 

 saw above (p. 176), this is the only case where npLv, meaning "until," 

 takes the infinitive after a negative main clause. 



Assuming that the infinitive should be read here, why was it used ? 

 No satisfactory explanation has yet been given. Sturm \ regards the 



* See below, p. 181. 



t Helbing, Die Priipositionen bei Herodotos und anrlern Historikern, in Schanz'a 

 Beitrage, Heft 16, p. 44, takes eW here with h. I may state that the above inter- 

 pretation occurred to me before I saw the article of Helbing. 



| Schanz's Beitriige, vol. 1, pp. 293 ff. 



