BOAS] HANDBOOK OF INDIAN LANGUAGES TAKELMA 215 



as the aorist naga- is formed from the verb-stem nd"'g- say to some 

 ONE. Similarly, the noun yele^x burden-basket is phonetically 

 related to a hypothetical base *yelx-, as is the aorist leme-kl- to the 

 non-aorist lem-k!-. A small number of nouns appear in two forms, 

 one corresponding to the aorist stem, the other to the verb-stem of a 

 verb: gulu^m oak, but with characteristic - (a) n-:gulm-an- (the non- 

 aorist gula'm with inorganic -a- also occurs). Similarly, yulu'm and 

 yula^m eagle. In such variable nouns we have a complete morpho- 

 logic analogy to Type 2 (or 3)) verbs like aorist xudum- whistle, 

 Yeih-stem. xufm- (with inorganic -a-: xudam-) . In both gulu^m and 

 xudum- the -m- is almost certainly a suffixed element. It must be 

 carefully noted, however, that, while in the verb we very often have 

 both the aorist stem and the base (as verb-stem) in actual existence, 

 in the case of nouns we rarely can go beyond the stem as revealed in 

 an absolute or incorporated form. It is true that sometimes a 

 hypothetical noun-base phonetically coincides with a verbal base, but 

 only in the minority of cases can the two be satisfactorily connected. 

 Thus, yut!-, abstracted from yutlu^n duck, is very probably identical 

 with the yut!- of aorist yutluyad- swallow GREift)iLY like hog or 

 DUCK. On the other hand, little is gained, by comparing the yul- of 

 yulu\n EAGLE with the yul- of aorist yuluydl- rub; the 'p!iy- of 

 'pli'yin deer and pli'yax fawn with the aorist -p!iyin-(¥v)a-) lie 

 ON pillow (cf. gwen-f!%xap pillow), unless the deer was so called, 

 for reasons of name-taboo, because its skin was used for the making 

 of pillows (or, more naturally, the reverse) ;^ the way- of way a" knife 

 with way- sleep ; or the noun-stem yaw- rib (occurring as ya-u- when 

 incorporated) with the verb-stem yaw- (yiw-) talk. It is not justi- 

 fiable to say that noun-stems of apparently non-primitive form are 

 necessarily amplified from the bases that seem to lie back of them 

 (e. g., will- from wil-; yulu-m from yul-), but merely that there is a 

 strong tendency in Takelma for the formation in the noun of certain 

 typical sound-groups analogous to those found in the verb. 



§ 86. TYPES OF STEM FORMATION 



Though it is probably impossible to duplicate all the various types 

 of aorist and verb stem found in the verb, most of those that are at 

 all frequent occur also in the noun. 



'Improbable, however, if aorist p'.eyen- lie and p.'iyin-k'wa- lie on pillow are radically connected (see § 31). 



§ 86 



