sn-ANTox) THE TLINUIT ANU HAIDA LANfiUAGES 485 



it even rephues the regular suffixes of tlie past tense. The Haida 

 forms are, in fact, about what we should expect if the Tlingit particle 

 A,i should become agglutinated to the verbs it so often follows. Pos- 

 sibly wc alst) have here an explanation of the origin of the llaida verl) 

 "to be," which usuall}' appears as e'djt, but contracts to 7.t in compo- 

 sition. 



The consonant s also figures in another connection. Most Tlingit 

 interrogative pronouns and adverbs are accompanied by a syllable 

 sa — da't«i, what? irn'sa how? — and Haida interrogative pronouns and 

 adverbs contain this same consonant .v with such uniformity as to lead 

 one to suspect it was originally an affix, viz, gVsto, who? (/(hvTi, why? 

 what? 7/7.V, what? i/ri'(/i/x, what? f/l'n(/t'f, where? (i7x7'{xf<ili(i<i) whence? 



Both languages also make a very extensive use of a general demon- 

 strative form a to stand for noiuis or pronouns, and in both it is of ten 

 used to form compound demonstratives or connectives. Finally, as 

 savoring of dialectic change ratiicr than borrowing, attention should be 

 Ciilled to the fact that the last six words compared in the list completed 

 at th(> top of page 483 appear to change regularly fi'oni Tlingit to 

 Haida by drojiping a final k sound or replacing it with a or k. 



Concliixidiix. In spite of the very great divergences which these 

 two languages present, the points last considered, along with many of 

 the coincidences noted on pages 481-4S3, can hardly l)e ai'countcc! for 

 on the basis of either boriowing or coincidence. They seem to the 

 writer to be the faint echo of a time wiicn the ancestoi'sof some of the 

 people now represented by the llaida and Tlingit spoke one tongue, and 

 there is certainly nothing in the structure of stems, word^, or sentences 

 to contradict this view. At the same time, in order to l)ring about the 

 diU'erences which now exist, the two peoples nuist have lived long apart 

 and have been subjected to very diti'erent influences. It is quite possi- 

 ble that, instead of being inmiediatrly related. Haida and 'IMingit have 

 developed separately from some third language, and, urdess it were one 

 now entirely extinct, this would naturally be Athapascan, the only lan- 

 guage in this region that eitjier of them resembles at all closely. \Vc 

 shall be iuial)le to come to a delinite conclusion on this point until the 

 neighboring Athapascan dialects have been thoroughly examined, 

 analyzed, and compared with Haida and Tlingit. 



