434 Proceedings of the Royal Irish Academy. 



Before proceeding to the special topic, attention may be 

 called to a few points in the above Summary, where doubts 

 may be felt. 



It is not easy to determine what the timbre of the n was in 

 the present tense, [Z] indie, and subj. The forms with r [de?-na) 

 have undoubtedly the hard n, the broad n ; and the siibj. forms 

 without r are almost always written with following a, i. e. with 

 the hard n, dena, denat. But in the indie, we find even dein, 

 which seems to justify the assumption that these forms, with 

 soft and hard n respectively, are the real representatives of the 

 indie, and subj. moods. We should thus have the parallel 



denj (= dein) : dena = Lat. regit : regat. 



But the practice of scribes is not uniform, and the analysis 

 given above exhibits a few cases where the two moods have 

 the incongruous vowel. Thus, i. 244, g, 14 we have noco dein, 

 ni dein ; 242, 6 ni dene, but 242, 4 ni dena ; in 11. 186, 20, 21 

 we have na dein, co dein ; iii. 112, g na deine ; iv. 386, 4 na dein 

 [where the parallel passage i. 256, 14 has in the Commentary, 

 na denand~\. But though we have thus generally the soft timbre 

 w in the singular, the plural shows the hard vowel : we get denat, 

 denait, deniiit, which can only denote the hard vowel, cf 11. 374, x; 

 314, 32, &c. On the other hand, a very rare case of the soft 

 timbre in the subj. would be the passage iv. 176, 23 ci2)e 

 inruiret ... do neoch na dene olcc friu, if we are to press the 

 rule ; but it should probably be dena with hard timbre, as almost 

 invariably. 



But it is evident that the language preferred as its subj. 

 the more definite stem dern-, which certainly occurs more than 

 four times as often as den-. 



And here the identity of the stem dern- with that of the 

 perfect stem [Z] has brought about a difficulty in deciding 

 whether derna was felt as a pres. subj. or as a perfect by the 

 writer; probably there was no very pronounced decision, sifaciat 

 or sifecerit may have been felt to come to the same thing. 



