430 BUREAU OF AMERICAN ETHNOLOGY [bull. 40 



By certain grammatical processes, consonants may be weakened 

 hardened, or aspirated. These changes take place in accordance 

 with the phonetic table given before. The hardened surd becomes 

 a fortis, and the weakened fortis or surd becomes a sonant. The 

 hardened and softened sonants strengthen their glottal element to 

 an E. Examples of these changes will be given in §4. By aspiration 

 the series of Tc sounds and of l sounds are transformed into their cor- 

 responding spirants, while in the dental and labial series aspira- 

 tion does not occur. The hardening and weakening of the spirants 

 reveals a number of unexpected relations of sounds. We find — 



Spirants Hardened Weakened 



X X^ X 



x{w) ^W W 



X' n % 



s ts! y or dz 



I H I 



Similar relations of consonants appear in cases of reduplication. 

 Thus we have — 



e'qa reduplicated e's^eqa (q and s) 



qlu'lyak^ reduplicated q.'ulsq.'u'lyaJc^ {s and y) 



The change of x' into n suggests that the n may belong rather to the 

 anterior palatal series than to the alveolar series. 



The nasals, I, y, and w, when weakened, become sonant by being 

 preceded by the glottal stop, y and w are clearly related to i and u. 



§ 3. Sound Groupings 



The Kwakiutl language does not admit clusters of consonants at 

 the beginning of words. Extensive clusters of consonants are rare; 

 and even combinations of two consonants are restricted in number, 

 their sequence being governed by rules of euphony. On the whole, 

 a stop (i. e., a sonant, surd, or fortis) can not be followed by another 

 consonant. This is carried through rigidly in the case of the palatals 

 and laterals, while combinations of consonants in which the first is 

 an alveolar or bilabial stop do occur, p followed by consonants is 

 not rare; t followed by consonants is by far less frequent. The cor- 

 responding sonants followed by a consonant do not appear as often, 

 because the intonation of the vocal cords tends to increase in strength, 

 and an e is introduced which separates the sounds. 

 §3 



