186 Bekker's Digammated Text of Homer. 



of them once, we must presume that the first e is not part of a redu- 

 plication, but the same common prefix which we find, for instance, in 

 iaXdofiai (i. e. sTe^doyai^j for -pi^dofiat to icish, and the aorist participle 

 iBiad/uspog (i. e. eFsiaafisvog) having likened oneself. We shall not then 

 be surprised to see that iTegyf^ipai' and tFegx'^To do not admit initial 

 digamma. 



We come now to consider the question, whether our editor does 

 right in recognizing only one lost consonant, the digamma, or whether 

 he should not have recognized others as producing similar appear- 

 ances in the Homeric verse. Curtius, in the concluding part of his 

 Principles of Greek Etymology, maintains in the case of several 

 Avords, that the epic hiatus was occasioned by a consonant y-sound. 

 He holds this to be true in reference to I'otxa, iaxetv, which we have 

 just considered. He remarks that dialects and inscriptions give no 

 evidence of digamma in this word ; that no root vik in the sense of 

 likeness is to be found in the cognate languages ; and that it is there- 

 fore very hazardous to write revoixa, FfFiyxetv, in the text of Homer, 

 He observes that there are clear traces in Herodotus and elsewhere 

 of a word Selxrjlog or dslxeXog having the sense of (e)l'xsXog like, simi- 

 lar. He is therefore led to adopt the conjecture of Bopp, that the 

 root of soixa is formed from that of delxvvfii,, Lat. dico, Sk. dig (i. e. 

 dik), to show. He conceives that the S assumed a parasitic y, and 

 then dropped away itself, thus dik, dgik, yik, and that from yik thus 

 formed came by reduplication ysyoixa, yeyaxsip. I cannot think that 

 there is much plausibility in this explanation. If the transition from 

 dik to yik was made in the formative Indo-European period, we might 

 expect to find a root yik having the sense of likeness somewhere in 

 the cognate languages, which Curtius does not pretend is the fact. 

 It is evident indeed that he regards the evolution of yik from dik as 

 taking place in the Greek after the Indo-European time. We must 

 think then of the root dik as already provided with inflection, making 

 a reduplicated preterite dsdoixa, from which would come first, by add- 

 ing y to both ^'s, dysdyutxoc, and then, by dropping both the 5's, 

 yeyoixa. But the change from d to y is confessedly a rare one in the 

 Greek language : how hazardous then to assume that it has occui-red 

 twice in the same form. It might perhaps be said that the change 

 occurred first in some such form as dixsXog, meaning like, Avhich passed 

 into yixeXo; ; that this gave the suggestion of a root yix^ meaning to 

 be like, and that yeyoixa was formed independently from this sug- 

 gested root, and not by phonetic change from a pre-existing dsdoixa. 

 This is indeed possible : but we should scarcely expect that a root 

 arising at this comparatively late stage of linguistic development 



