192 Bekker's Digammated Text of Homer, 



that from F(5? came tFo? by a prefixed e, and then, by transposition of 

 digamma, F£<5j. But such a transposition is a more hazardous assump- 

 tion than he seems to think; and the form fccJ? has little support 

 either in the Homeric text, or in the suggestions of comparative phi- 

 lology. 



It remains to speak of the pronoun txaarog each, and the kindred 

 adverb txureQOe on each side: the pronoun IxdneQog itself is incapable 

 of appearing in the heroic hexameter. The derivation of ix&Tsoog and 

 txaaxog is as yet far from certain. It is probable, however, that the 

 -xaieqog and -xaaiog are a comparative and superlative form from the 

 interrogative stem xa, that they are in fact identical with the inter- 

 rogatives ndxegog and noazog, which in their Ionic forms are xozeQog and 

 xoajog. It is probable also that the first syllable ^ is the same as in 

 the numeral ^xaT(5»', Lat. centum, Sk. fatam (i. e. katam). If so, it is 

 probably for ev, the root of the numeral elg one : thus ^xaTov = one 

 hundred, axdiegog =. one which-more, one which of tioo. Here now 

 we stumble again upon an uncertainty : but of all the explanations 

 proposed for the numeral sTg, eV, the most probable is that which con- 

 nects it with Sk. sama, our Eng. same, Lat. semel, simplex, singidi. 

 It thus appears that or may probably have been the primitive initial 

 of hx&ieqog, Ixaurog. We have already observed that, according to 

 Ahrens' enumei-ation, sxauiog occurs 110 times in the Iliad. Now 66 

 of these ai'e cases of hiatus, some of them easily admissible, but 

 many othei'S giving strong indication of a consonant initial. It is 

 not therefore surprising that Bekker should have written vexuarog and 

 TpsxdTE^de wherever the verse allows it. Out of 28 unconformable 

 cases he makes 17 conformable by various conjectures, several of 

 which belong to the most hazardous that he has ventured. In 11 

 cases he has left the initial vowel untouched. Here the proportion of 

 unconformable cases, 25 per cent, throws suspicion on the digamma, 

 which is much increased by the fact that comparative philology has 

 no plausible explanation for the forms vexdreoog, Fixaawg. Such forms 

 as asxiieoog and aixaazog are much more probable on grounds of 

 comparative philology. Practically then the case as to txaarog stands 

 in this way. It cannot well be doubted that the word, sometimes at 

 least, began with a consonant in the Homeric language. If we as- 

 sume that digamma is the only initial consonant of the Homeric lan- 

 guage which has failed to appear in our text, then we must recognize 

 a Homeric vixaarog : such, doubtless, was the reasoning of Bekker. 

 But the assumption is an unsafe one : there is reason for suspecting 

 that other initial consonants of the Homeric language have had, 

 though to a far less extent, the same fortune as digamma ; and in 



