BRACKETT. — TEMPORAL CLAUSES IN HERODOTUS. IT-) 



and of all editors except Stein in his 2d ed. (1884) and van Herwer- 

 den. Madvig * would change ore to on for two reasons: (1) because he 

 thinks it strange that in parallel clauses on is used in one case and ore 

 in the other ; (2) because he believes that in a clause with ore a sec- 

 ondary tense of the indicative could not be changed to the optative. It 

 does seem improbable that Herodotus used different words in introducing 

 two clauses which are so exactly parallel. But should ore be changed? 

 Undoubtedly the reason why the secondary tenses of the indicative are 

 in general not changed in indirect discourse is to avoid obscurity. But 

 here, as Goodwin has said,t no obscurity whatever is caused by using the 

 optative, and further, it is to be noticed that an optative (e'1'7 i^epyacrpevos) 

 has preceded. Should we not rather read ore for the following 6V1? otl 

 KiiTaKvafie can hardly mean " because I took down," as Goodwin says 

 (1. s. c.)j but rather " in that, " or " therein that." $ But this use of on 

 seems not to be common in Herodotus ; Madvig does not cite another 

 instance. § In 9, 75, moreover, in a passage exactly parallel to this, ore 

 is found in MSS. R and S. and was conjectured by Kriiger. In 7, 20, 

 where all MSS. give ore, it is used in the same sense as here, ore should, 

 I think, be retained in our passage. If any change is to be made I would 

 change the following 6V1 to ore. 



A primary tense of the indicative in a temporal clause is changed to 

 the optative only once, 3, 27, where irapeir) stands for napeo-Ti. The sub- 

 junctive occurs twice in a clause with ore after a secondary tense (5, 30 ; 

 6, 78) and in both cases the subjunctive is retained. In clauses of con- 

 temporaneity introduced by relative expressions a secondary tense of the 

 indicative (2, 137 ; 5, 84), a primary tense of the indicative (1, 164), and 

 a subjunctive (1, 21 ; 4, 201) are found in a secondary sequence in indirect 

 discourse. In 5, 63 the optative with okm stands in indirect discourse 

 for the subjunctive with civ. 



In two passages the conclusion of the temporal clause is lost sight of 

 in a long rambling parenthesis. In 1, 137, the clause eWre . . . egefiXrj- 

 drjo-av finds its conclusion only at the end of that section in the words 

 tovs be ovtw 8f} . . . Arjpvov. Of the passage in 9, 84, eneire . . . rjcpdviaro, 

 k.tX., Stein says, "locus et lacunosus et corruptus." But it seems 



* Advers. Crit , III, 26, note. t Moods and Tenses, 693. 



J Cf. Madvig, Synt. of the Greek Lang-., trans, by Browne, London, 1880,3d ed. ; 

 p. 113, n. 1, b. 



§ I have found on used in this sense in Herodotus but once, apart from 

 the present passage, i. e., 7, 137, but my search for such a use was by no means 

 exhaustive. 



