Belcher's Bigammated Text of Homer. 179 



in the epic verse, so tliat hiatus occurring in such cases furnishes little, 

 if *ny, presumption for an initial digamma. In my Grammar (67 d) 

 are mentioned three cases of this kind: — 1. When the lirst of two 

 words ends in a close vowel {h v) and seldom or never suffers elision : 

 this applies especially to the dative singular of the third declension ; 

 as nttiSl (tTtacrce. — 2. When the two words are separated by a clearly 

 required mark of punctuation; as x&drjao, t>(j d' inmaldeo /n-uda. — 3 

 (the most important case). When the vowels which make hiatus are 

 the two short syllables of the third foot, or, in other words, at the 

 feminine caesura of the third foot ; as tu*' ol sS iyivovio ivl fieydiQoiai 

 yBvWlrj. In this place it has been proved that hiatus is allowed with 

 much the same freedom as at the end of a verse. There is another 

 ease which ought to be added to these three — a case in which hiatus 

 is easily excused, if not freely allowed — and that is, after a long vowel 

 or diphthong in arsis, and particularly the arsis of the third or tlie 

 fifth foot. The first line of the Iliad is an instance in point : 



1, fi^viv cceids, 6e6., ni]h]i'udea) 'A'/i^^log. 

 which shows hiatus after the arsis of the fifth foot : after the arsis 

 of the third, we find it in ^, 



24. (i^^' oix ^uirgsidri ^Ayanifxvovt r^vdavB dvudj. 

 42. Tloeiav ^davaol ifi& d(xxQva ooloi, ^iXsaaiv. 

 Here also, after the long arsis of the third foot, as well as in the 

 feminine caesura of that foot, we find something of the same freedom 

 as at the end of a verse. This appears in such lines as A, 



153. dsvQO (.mxriao^isvog, inel ov tl /j^ot al'iiol slaiv. 

 where og, the last syllable of (laxrjaousvog, stands in the third arsis 

 before Insl^ which certainly did not begin with digamma. 



The other remark relates to the negative evidence, that which goes 

 to disprove a digamma. It is well known that for every digammated 

 word, even the best ascertained, there is some evidence of this char- 

 acter : there are some passages in the poems as we have them, in 

 which the digamma cannot be written without violating the metre. 

 It is obviously desirable that we should have some idea as to the 

 range of these exceptional cases, their numerical ratio to the Avhole 

 number of passages in which the word occurs. On this subject there 

 are some good remarks in Kuhn's Zeitschrift, X, 60 ff, in an article by 

 H. L. Ahrens on txaoroc, one of the words which Bekker has digam- 

 mated. Ahrens enumerates all the instances, 110 in number, where 

 this word occurs in the Iliad, and states that there are 44 of these in 

 which the digamma could not be written into the traditional text 

 without a violation of metre. But as 16 of the 44 can be made to 



