368 PROCEEDINGS OF THE AMERICAN ACADEMY 



they have all split.* After the most careful study that I have been 

 able to give to the subject, and especially after a comparison of several 

 thousand classic examples, I am convinced that no such principle can 

 be found. Every example that I have met Avith has only confirmed 

 the opinion, which I can now express with the greatest confidence, that 

 there is no inherent distinction between the present indicative and the 

 present subjunctive in protasis (between ci jSovXerai and iav ^ovXrjrai), 

 except that of iitne. 



But before this distinction can be properly applied, another must be 

 understood and kept in view. A supposition may be either particular 

 or general, that is, it may refer to a definite act (or a definite series of 

 acts) performed at a definite time, or it may refer indefinitely to any 

 one of a class of acts. Thus we may say. If he now has money, he de- 

 sires to give it, and also, If he ever has money, he always desires to give 

 it : we use the present tense has in English in both conditions, but it is 

 obvious that it refers to strictly present time only in the former case, 

 in which we call the supposition particular ; in the other case it refers 

 to any occasion on which we may suppose the person to have money, 

 and we therefore call this supposition general. If now we leave out of 

 account for the present all conditions which express a general supposition, 

 and confine ourselves to those which refer to definite times, we shall 

 find in Greek f the simple rule invariable, that the subjunctive refers to 

 the future. Thus iav tovto ttoi^, (ro(f)bs ea-rai means if he shall do this 

 he will be loise ; whereas el tovto noiel, aocf)6s ia-Tiu means if he is (now) 

 doing this, he is wise. In neither case is anything whatever implied 

 by the construction as to the opinion of the speaker with regai'd to the 

 truth or the possibility of the supposition : the context or the nature of 

 the supposition may or may not indicate these things, but the construc- 

 tion itself simply indicates the time. We have seen that the indicative 

 can express a future condition by its future tense. This is merely a 



* I cannot claim to have been entirely free from the same error myself; for, in 

 attempting, five years ago, to state the relations of the Greek moods in conditional 

 sentences (Syntax of the Moods and Tenses of the Greek Verb, Cambridge, 1860), 

 I could not persuade myself to abandon entirely a principle which nearly all 

 grammarians had accepted in some form, and 1 reluctantly admitted the idea of 

 ■possibditij as distinguishing the subjunctive from the optative. 



t In what follows I shall use the Attic Greek as the standard of comparison, for 

 several obvious reasons. The Latin constructions will thus, I think, be more clearly 

 understood. 



