322 ANNALS NEW YORK ACADEMY OF SCIENCES 



Languages being apparently identical. The rules for such retention 

 or disappearance can be worked out only by a comparative study 

 of kindred languages, but it may be tentatively suggested that an 

 original short final vowel was dropped in Tepecano while an original 

 long final vowel was shortened, both forms being retained in Nor- 

 thern Tepehuane as originally. 



X<» vocalic timbre seems to replace the lost final vowel in Tepe- 

 cano, but that its influence has not entirely gone is proven by the 

 fact that, at least in the ease of nominal stems, the original stem 

 vowel reappears in more extended forms. 



hut 



The final stem vowel i appears, for some unexplained reason, to 

 be more tenacious than the other vowels. Certain stems will thus 

 appear normally without, but at other times with stereotyped 

 final vowel. 



'I 'he latter group of words are used with ceremonial context which 

 may possibly explain the persistence of the final stem vowel, through 

 religious conservatism. 



I" 11 the case of verbal stems no certain instances of the reap- 

 pearance of the lost final vowel have I n noted. 



\m .\ I Mi I I l:i: -77 ' i/> 



Noun and verb steins are generally dissimilar in Tepecano but 



ia Rinai.iiinm, op. cit. 



