190 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Helen back and thus stopped the slaughter" the action of tjrel aTroiWvvro 

 is wholly antecedent. Bat here there is an unmistakable demand for 

 continued action in the temporal clause ; the imperfect here gives a 

 strong causal coloring with indication of the spectator (i. e., when Priam 

 saiv that his sons were perishing), and there is at the same time the idea 

 of the interruption of a state of affairs previously existing.* 



To pass now to clauses with cos, we read in 1,47: lv hk AeX0oun ws 



icrrjXdov rd^crra ts to peyapov . . . Ka\ titeip lTg>v, t) TIv0ir] . . . Xeyet TaSe. 



Surely we must suppose that the Lydiaus had ceased to speak, when the 

 priestess began to give the answer of the oracle. As to the reason for 

 this imperfect, no aorist of the simple verb ei'pcoruco (Att. epcordco) seems 

 to occur in Ionic Greek, f This is a case where we should expect an 

 aorist, if an aorist of the desired meaning were in common use. The 

 fact that t'ipovTo is not used here makes somewhat against the view of 

 Kallenberg, who believes that tlpopevos and tlp'todai (for so he would write 

 it) are not presents, as they have usually been regarded, but aorists.J 



A noteworthy case is found in 7, 14G: cos Se Tiepuovras avroiis KaTt\a$ov 

 Kin >)ytw es o\f/LU ttjv /3acrtXeos, to evdevTev ttv66^jl(vos • . • excXeutTf, k t.X. It is 



clear that the action of rjyov must have been completed before that of 

 either nvdopfvos or eVeXfuo-e began. The readings of the MSS. and 

 the editors here is as follows : rjyov, ABCP, edd. Kail., Stein, Kri'tg.; 

 rjyuyov, RS, edd. v. Hervv., Hold. The aorist is surely needed here ; 

 cf. Miller, § on Polybius : " The aorist . . . expresses the successful 

 conduct to a goal, or the successful transference from one state to 

 another." rjyayov is certainly to be preferred. In the passage 2, 

 121 (i : cos &e tiurco koi 8is koi rp\s avoi^avrt ale\ eXuVcrco (paiveaOai ra 

 Xpi'ipara, iwir)a>u piv rdSe, the action of the temporal clause is shown 

 by the context to be wholly antecedent. The present tense (= im- 

 perfect) is due to the idea of repetition in the protasis, to the pres- 

 ence of cue/, and also to the distinct causal coloring: "after he saw 

 that his money all the time was growing less." 1, 65: cos eo-rpe es t6 

 peyapov, Wvs f) Uv6irj Xeyet rdSf. Here again we must suppose that the 

 action of tcnjie was finished before that of Xeyei began. Cf. 1, 47, 

 cbs eafj\6ov (above), in an exactly similar sentence. Does the im- 

 perfect ecnjte express here extensive or aoristic action ? According to 



* Cf. Gildersleeve, Synt., 223 and 208. 



t Cf. Veitcli, Greek Verbs Irregular and Defective 2 , Oxford, 1878, s. v. 

 t Jahresb. d. Pliilolog. Vereins zu Berlin, 28 (1902), p 96. Cf. Smyth, Ionic 

 Dial., 224, 3; Hoffmann, Der ion. Dial., 167, 12. 

 § Am. Jour. Phil., 16 (1895), p. 166. 



