376 PROCEEDINGS OF THE AMERICAN ACADEMY 



2. "O Ti e^oi toir) av, he woiild give whatever he might have ; like ei n 

 e^ot, if he should have anything. AvVatro 70^ av, oTrore ^ovXoiro, tovto 

 TToie'iv, for he would be able to do this whenever he might ylease ; like ei' 

 TTore j3oi;Xoiro, if ever he should please. 



II. 1. "O rt av exj] 8iS&)o-i, whatever he has (at any time) he gives ; 

 like edv rt exj], if he (ever) has anytldng. ^vjifxaxelv tovtois edeXova-iv 

 arravrfs, ovs av opaxri TTapfaKfvaa-fievovs, those ivliom they {at any time) see 

 prepared ; like idv rivas opuxriv. 



Here we sometimes find the indicative, especially after oam, that pro- 

 noun expressing the indefiniteness sufficiently without help from the verb. 

 See Soph. Antig. 178; and compare Od. XIV. 157 with II. IX. 313. 



2. "O TL e'xoi ibiBov, whatever he had (at any time) he (always) gave ; 

 like e'i Ti i'xoi" Ovs '1801 firaKTas 'iovras eirrjvei, he (always) praised those 

 whom he saw marching in good order ; like ei nvas t'Soi.* 



I have attempted, in the preceding remarks, to call attention to the 

 general principles which govern the constructions in question, leaving 

 out of account all real or apparent exceptions which would tend to 

 obscure the principles themselves. It is seldom that any language will 

 allow its principles of construction to be stated so accurately that no 

 exceptions will have to be made ; if such a case is to be found, it is 

 in the Attic Greek, and in the constructions which are here discussed. 

 In the Homeric language, which is much less accurate in its use of the 

 moods and tenses, and in some respects less complete (having for ex- 

 ample no future optative), we find many constructions which would be 

 impossible in the Attic Greek ; and the same is true of the Latin, 

 which never attempted such accuracy in the expression of thought as 

 its sister language. The introduction of these exceptional construc- 

 tions in too prominent a position has often greatly interfered with a 

 right understanding of the regular forms in both languages. 



In conclusion I would add a few remarks on the English subjunctive in 

 conditional sentences. The modern English, especially the spoken lan- 

 guage, makes no distinction between present and future time in protasis, 

 if he does this being used to express all the various meanings which the 

 Greek expresses by el tovto npda-aei, el npd^ei, and eav TTpaaar) (or irpd^rj), 



* Although the Latin agrees with the Greek in several of these constructions, 

 still the peculiarities of the use of the subjunctive in Latin relative sentences are 

 so great, that it would have created confusion to include them in the table. 



