194 PROCEEDINGS OP THE AMERICAN ACADEMY. 



3, 36, 78, 131; 4, 151; 5, 30, 44, 62; 6, 91; 8, 121, 125; 9, 118; 

 i>s — 1, 10, 67, 77,84, 112, 163, 205; 2, 2, 118, 121 S; 3,56; 4, 124, 

 126 ; 5, 92 d, 104 ; 6, 87, 90, 107 ; 7, 107, 210, 212 ; 8, 4 ; 9, 20, 38, 

 49, 108. In every one of these cases the verb is in the imperfect. Is 

 this merely accidental ? 



Clearly, the function of a clause of antecedence is, by naming some 

 event or describing some condition, to give either a point or a period of 

 time after which, either immediately or with some interval, another 

 action took place. This being the case, it is necessary that the time- 

 defining action should be of a nature sufficiently definite and recognizable 

 to furnish such a point of departure for reckoning. Let us now apply 

 this to negative clauses. 



An action may be negatived in either of two ways : first, simply and 

 unemphatically, and secondly, with more stress on the negation, as in the 

 case of the failure of an attempt, a wish, or a hope. Thus a man who 

 was not present at a certain public address may say, " I did not hear that 

 address ; " whereas a man who was present but was unable to hear the 

 speaker says, " I did not hear that address," but with a different connota- 

 tion, i. e., " I could not hear it." It. will be evident that a negation of the 

 former type is not sufficiently strong, it does not, so to speak, partake 

 sufficiently of the nature of an event, to give a point of time from which 

 to reckon. But, as has been recognized for some time,* we have in 

 Greek two different forms of expression corresponding to these two 

 kinds of negation, i. e., the aorist with the negative for the simple nega- 

 tion, and the imperfect for the other. According to this principle we 

 shall expect the imperfect tense to appear in negative clauses of ante- 

 cedence. The resistance to the action attempted, desired, or hoped for, 

 causes, so to speak, a ripple on the surface of time, while the simple 

 absence of an event, or its failure without resistance, leaves that surface 

 unruffled. 



The facts of Herodotus's usage seem amply to support this principle. 

 As we have seen, the imperfect is used in every case. In 36 cases the 

 idea of disappointment or failure is unmistakable. A clear sign of this 

 meaning is given by the frequent use of verbs which refer specifically, 

 when negatived, to failing in something, as e. g., olos re yiveadat. (1, 67 ; 

 2, 181 ; 6,91 ; 8, 121) ; dvuaadm (1, 10; 2,2; 3, 131 ; 7,211; 9, 108); 

 e\eiu (3, 36) ; irpn^pieiv { 1 , 84, 205; 5, 44, 62) ; "npoKOTTTto-Bai (3, 5C). Cf. 



* This principle was first distinctly stated, so far as I have seen, by Gildersleeve, 

 Am. Jour. Phil., 2 (1881), pp. 466 ff. 



