224 BUREAU OF AMEEICAlSr ETHNOLOGY [bull. 40 



noun-characteristic -n- (-m) lies in the fact that the former is a neces- 

 sary part of the absolute form of the word, while the latter appears 

 only with grammatical increments. Thus the -am of TieHa'm board 

 can not be identified with the -am of ha-gwd^la^m in the road, as 

 gwaHa'm has no independent existence. The exact morphologic cor- 

 respondent of gwaH-am- is Tie^lam-a- (e. g., JieHam-a^-fV my board). 

 A doubt as to the character of the -n- can be had only in words that 

 never, or at least not normally, occur without possessive sufiix: 



lege^m-fV my kidneys 



wo^'plu'n-t'V my eyebrows ^ 



7. -a. There are a rather large number of dissyllabic nouns or 

 noun-stems with final -a, in which this element is to outward ap- 

 pearance an integral part of the radical portion of the word. 

 The number of instances in which it occurs, however, is considerable 

 enough to lead one to suspect its derivational character, though it 

 can be analyzed out in an even smaller number of cases than the 

 sufiix -n above discussed. The most convincing proof of the exist- 

 ence of a suffix -a is given by the word xu'ma food, dry food, 54.4; 

 188.1, a derivative of the adjective xu'm dry 168.15 (e. g., fim xu^m 

 DRIED salmon; cf. also xumii'Fde^ i am sated [132.1]). Other pos- 

 sible examples of its occurrence are: 



yoW fox (? cf. verb-stem yul- rub) 70.1, 4, 5; 78.2, 3, 9 



mena.' bear 72.3; 73.2, 3, 4, 5; 106.7, 10 



p.'eUsi'' slug 105.25 



noxwa,'^ small pestle 



fe'Zma small pestle 62.1; 116.18, 19; 118.2 



ma'xlsi dust 172.3; 184.5, 9 



Jcledsb" grass for string (sp. ?) 



tlelsi' shinny-stick (? cf. verb-stem t!eu- play shinny) 



tlela,' louse (? cf. verb base t.'el- lick) 116.3, 6, 7, 8, 11 



tlihsi- pancreas 46.1, 9; 49.7 



eZa- tongue (characteristic -a-?) 



dola,'' old tree 24.1 



^aTia' oak 22.11; 168.1, 2,3, 6, 7 (cf. yangwa's oak sp.; with 

 -gwas cf . perhaps al-gwa's-i- yellow) 

 It is of course possible that some of the dissyllabic nouns in -a 

 listed above (§ 86, 2) as showing a repeated vowel (e. g., ya'pla) really 

 belong here. 



1 These seem to be parallel to gwit.H^n-t'k' my wrist, in .which -«-, inasmuch as it acts as the equivalent 

 of the characteristic -u- (cf. gwit.'iuxde'k' my weist with luxde'k' my hand), is itself best considered 

 characteristic elejjjent. 



§ 87 



