318 



ANNALS NEW YORK ACADEMY OF SCIENCES 



\\ hen the second member of a consonantal combination. This lack 

 of aspiration gives to the ear accustomed to English an intermediate 

 surd-sonant effect, not very different from that of the sonant stop 

 when truly intermediate in character. "When final or the first 

 member of a consonantal combination, the surd quality is more 

 evident and in the former case a slight aspirate release is occasion- 

 ally heard. In the latter case there is generally no release, the 

 surd blending with the ensuing sound. Here it has frequently 

 been written incorrectly as of doubled length. 



The glottal stop, ', is frequent but not harsh or abrupt and has 

 been occasionally written in place of a mere cessation of voice. 

 Frequently it is synchronous with a stop, forming what is known 

 as a globalized stop. In such cases the glottal stop has generally 

 been written before the stop thus globalized, less frequently after it. 



The affricative tc is very rarely met, ts practically never. Neither 

 occur in final position. 



Consonantal T</hlc 



Characters in parentheses are of secondary derivation or of doubt- 

 ful nature. 



only l\ the former admits that p and b, k and g are confused. In Lumholtz's Tubar 

 vocabulary neither <j nor p are found. 



The evidence is almost incontrovertible that the sonant stops possessed by the Pima- 

 PupsgO Tepchuane-Tepecano group have developed from original Uto-Aztekan spirants, h 

 from kw, d from y and g from w. (Cf. Edward SapiR: American Anthropologist, (n 3.), 

 Vol. 17, ]). 30b). The original spirants appear to remain in the other Sonoran languages. 



