Bekker's Dlgammated Text of Homer. \11 



has in some instances given rise to tlie same appearances in the 

 Homeric verse as those occasioned by the v- or ?<'-souncl (the digara- 

 ma). But the words which are thus supposed to show traces of 

 initial y, Bekker either Avrites with digamma, or leaves them with 

 a vowel-initial. We shall refer again to this point before closing. 



We have said that in determining what words had initial digamma 

 in the Homeric language, our editor relies mainly on the indications 

 of the Homeric verse. Having satisfied himself in the case of any 

 particular word that it did have the digamma, he proceeds to write 

 the word, as also its derivatives and compounds, with that character: 

 and this he does, not only where metrical reasons favor or require 

 the introduction of a consonant, but Avherever metrical reasons do 

 not absolutely forbid the introduction of a consonant. In very many 

 instances where the verse as we have it in our traditional text will 

 not alloAV the digamma to come in, the difficulty can be removed by 

 changes of the text Avhich are more or less obvious. Bekker's princi- 

 ple, it is evident, has been to write the word with digamma, whenever 

 this is consistent with the verse as it stands in the ordinary text, or 

 can be made so consistent by some slight and easy change of reading. 

 He shows his judgment and moderation as a critic by refusing (at 

 least, in general, with only rare exceptions) to make any considerable 

 or arbitrary change of reading for the sake of getting in his digamma. 

 Rather than do this, he will allow the word to appear in a particular 

 case Avithout the initial consonant which usually belongs to it. I 

 may illustrate his mode of procedure by a more particular statement 

 of what he has done in the first book of the Iliad, In the 611 lines of 

 which it consists, there are found, if I have counted right, 162 which 

 show the digamma. But some of these contain it more than once, 

 so that 184 words are written with this character. In 36 of these, it 

 is found, not at the absolute beginning of the word, but after a jjvej)- 

 osition or other prefix. Of the 184 words there are only 31, or about 

 a sixth, in which the introduction of the digamma has required any 

 further change of text; and in 18 of these 31, the only change re- 

 quired has been the omission of a movable v from the end of a pre- 

 ceding word. Thus in verse 



14. orr^,u,uaT' I'jf'aiv Iv xsoalv Ixrj^ohiv ^AnoXltDvog. B. '/^Q*^^ F^''- 



96. Toiii^ex' ao' SA'^e' edutxev kxrj^dlog ^J' en d(ha&i. 6do)xs fsx. 



294, el 8i] uol nuv eqyov vnel^ofiai, ott* xsv e\nr^c. o 7Xt xe feItt'^S- 



In 8 instances, a slight change has been made in the grammatical 

 form of the preceding word, though in 2 out of the 8 the change was 

 23 



