BOAS] HANDBOOK OF INDIAN LANGUAGES — TAKELMA 281 



TcH'xinlii round dance (said by Frog) 102.18 

 ^o'cu ^o'cu roiuid dance (said by Frog) 102.23 

 gwa'tca gwatca round dance (said by Bluejay) 104,7 

 tda'itdia round dance (play on tcla'^'c bluejay) 104.7 

 he'lyehinibl' a round dance (said by Mouse; play on hehe^n rushes) 



104.10 

 heleldo round dance (play on help' swan) 104.15 

 hi'gi hi'gi hi'gi+ Skunk's medicine-man's dance ([?J play on 



iiF"" skunk) 164.18, 22; 166.5 

 hd'^gwatci ha'^gwatci said by s"omloho'lxa^s in doctoring 



§ 116. CONCLUSION 



The salient morphologic characteristics of Takelma may be summed 

 up in the words inflective and incorporating, the chief stress 

 being laid on either epithet according as one attaches greater impor- 

 tance to the general method employed in the formation of words and 

 forms and their resulting inner coherence and unity, or to the par- 

 ticular grammatical treatment of a special, though for many Ameri- 

 can languages important, syntactic relation, the object. Outside of 

 most prefixed elements and a small number of the post-nominal 

 suffixes, neither of which enter organically into the inner structure 

 of the word-form, the Takelma word is a firmly knit morphologic 

 unit built up of a radical base or stem and one or more affixed (gen- 

 erally suffixed) elements of almost entirely formal, not material, 

 signification. 



It would be interesting to compare the structure of Takelma with 

 that of the neighboring languages; but a lack, at the time of writing, 

 of published material on the Kalapuya, Coos, Shasta, Achomawi, 

 and Karok makes it necessary to dispense with such comparison. 

 With the Athapascan dialects of southwest Oregon, the speakers of 

 which were in close cultural contact with the Takelmas, practically 

 no agreements of detail are traceable. Both Takelma and Atha- 

 pascan make a very extended idiomatic use of a rather large num- 

 ber of verbal prefixes, but the resemblance is probably not a far- 

 reaching one. While the Athapascan prefixes are etymologically 

 distinct from the main body of lexical material and have reference 

 chiefly to position and modes of motion, a very considerable number 

 of the Takelma prefixes are intimately associated, etymologically 

 and functionally, with parts of the body. In the verb the two lan- 

 guages agree in the incorporation of the pronominal subject and 



§ 116 



