BOAS] HANDBOOK OF INDIAN LANGUAGES TAKELMA 117 



Verb-stem Aorist stem 



da-sg&'Mfd'^ it will lie scattered da-sgaXi it lies scattered about 



about 

 p'ildi'^'a" fiat thing will lie p'ildi flat thing lies 



t'ge'its'!i(Za« round thing will lie t'geitsMi round thing lies 



(13824) 

 s'emi'fd" it will lie with open- s'eini it lies with opening on 



ing on top (like box) top 



s'u'k'didd'^ it will lie curled up s'ugwidi it lies curled up 

 wl'^k'didd'^ it will lie heaped wik!idi.it lies heaped about 

 about 

 Of similar appearance, though the aorist (not the future) is transi- 

 tive in form, is : 



Verb-stem Aorist stem 



dd°--sge'k\iie^ I shall listen (Za°-sgek!iya'^n I listened (third 



person fZa'^-sgek !i 102.8) 



In speaking of verbs of Types 15 and 16, the terms verb-stem and 

 aorist stem are used in a purely relative sense, the portions of the 

 listed forms printed in Roman characters not being really on a par 

 with those similarly marked in the first fourteen classes. These last 

 two types have significance as such only m so far as certain elements 

 of an essentially derivative character {-P-, -i-, -as-) are at the same 

 time formal means of distinguishing aorist from non-aorist forms. 

 it is not difficult to show that in several cases these elements are 

 themselves preceded by non-radical elements. 



One or two aorists have been found in the material obtained that 

 can not be well classified under any of the sixteen types illustrated 

 above. They are: 



gwen- xoxog[\\']a'^n I string (salmon) together ( = fully redupli- 

 cated xogxog- ; otherwise to be analyzed as xoxo-g- of Type 

 10 a) 74.14 

 saZ-s'a'^xs'ix he slid 

 This latter verb with its mysterious l'^ in the repeated syllable is 

 absolutely without known parallel. Irregular is also the defective 

 verb ei- be (see §60, fourth footnote). 



3. Verbal Suffixes of Derivation (1141-38) 



§41. GENERAL REMARKS 



Although the absolute number of non-pronominal suffixes in the 



verb is considerable (almost or quite thirty), the number of those 



that have a well-defined, more or less transparent signification is not 



large (hardly more than a dozen or so) when compared with what 



§ 41 



