BRACKETT. — TEMPORAL CLAUSES IN HERODOTUS. 221 



C. Subsequence. 



Clauses of subsequence show abnormal order iu the following pass 

 irpiV, with infin. : 1, 46, 71, 76; 3, 25; 4, 93, 180; 6, 87, 119; 8, 144; 

 nplv, with subj. : 1, 32 ; irpiv rj, with iniin. : 1, 92, 170 ; 2, 1 ; 4, 1, 167 ; 

 ('., 22; 8, 12, 59; 9, 13, 13 2 , 48, 68; nplv rf, with subj.: 1, 136; sore: 



7, 158; €?5: 1, 93; pe'xP* : 4, 119. 



Out of 43 cases of irplv or 7r/)tf 7 with the infinitive, the temporal clause 

 precedes the main clause 21 times.* The reasons which cause the infini- 

 tive clause to be placed before the main clause, or within it, but before 

 the leading verb, are in general the same as those given above. The 

 temporal clause precedes, if it is emphatic (1, 46, 71 ; 4, 93 ; 6, 22, 119 ; 



8, 144; 9, 13, 48, 68), or if it is closely connected in thought with the 

 preceding sentence (1, 76, 78, 92; 3, 127; 4, 1, 167). Sometimes the 

 irpiv clause is put first to avoid ambiguity (2, 2) ; sometimes to bring 

 raSe or roiovSe at the end of a sentence (4, 180; 6, 87). In 8, 12, and 

 6, 119, the order is probably due to the parallel arrangement of clauses. 



Clauses with nporepov 7 always stand after the main clause. In 7, 158 : 

 criTov re anacrrj t[j EWtjvw uTpaTifi ear av SianoXep.rjo'cop.ev, viro8eKo/j.at. nape- 

 £eiv, the clause with eVre is connected closely with &Itov, and is virtually 

 a modifying expression co-ordinate with rrj 'EX. aTpanrj. Clauses with 

 « o follow the main clause except in 1, 93 : tov yap 87 AvSav typov al 

 Buyarepes nopvevovrai nacrai, crvXXeyovcrai <r(f)i<ri (fiepvas, is o av ovvoiKrjaaxri tovto 

 7ToUovaai, but here the participle nWouo-ac is little more than a repetition 

 of TTopvcvovrai. In 4, 119, the clause with fiexP 1 1S emphatic. 



VI. 



Below is given a classified list of the temporal clauses in Herodotus. 

 This list is based upon Kallenberg's revision of Dietsch's Herodotus, 

 Leipzig, imprint of 1898, although iu some cases, as has been stated 

 above, I should prefer a different reading. In some cases it is impossible 

 to draw a sharp line of demarcation between temporal and causal clauses ; 

 not infrequently both ideas are present in the same clause. In making 

 this list I have aimed to include all clauses in which the temporal idea 

 is clearly discernible. 



Since the verb dpi cannot be classed as exclusively either extensive or 



* Sturm's statement (op. cit., pp. 205 f.), " Schliesslich bemerken wir hinsieht- 

 lich der Stellung des Infinitivsatzes nach irpiv und npiv tf, dass derselbe seinem 

 Hauptsatze meistens vorangeht," is clearly wrong. 



