24 BUREAU OF AMERICAN ETHNOLOGY [bull.gs 



Button, after the lapse of two years, was unable to converse mth his 

 parents in his own tongue, although he understood them (Hyades, 

 g, 271), while one of the native women taken in Crooked Reach in 

 1876 had entirely forgotten by 1883 her own language and spoke only 

 Spanish (ibid., 278). 



The Alacaluf are or were in contact with the Onas and Tehuel- 

 ches throughout a large section of their territory. In recent years, 

 moreover, the Onas and Alacaluf have been brought into close 

 association at the Dawson Island missions. These facts explain 

 sufficiently the minor Ona-Tehuelche element in the Bo-Sk lists, an 

 element entirely or almost entirely absent from H and Fi. The 

 present writer has made no exhaustive comparison, but has utihzed 

 only the more readily accessible Ona-Tehuelche material. 



If we eliminate the Yahgan element from the H-Fi group and the 

 Ona-Tehuelche element from the Bo-Sk group, the two groups are 

 brought into much closer harmony and their differences largely 

 accomited for. The remaining differences are probably due to the 

 various causes to be mentioned below. 



First, inaccuracies of transcription. The ^Uacalufan lajiguage is, 

 as observers agree, extremely guttural^ or "buccale et comme 

 muqueuse" as Dr. Topinard put it, which makes the task of catching 

 the sounds correctly and transcribing them an exceedingly difficult 

 one (cf., e. g., Hyades, q, 12, quoting Mr. Bridges; Scitz, a, 185; 

 Skottsberg, d, 580). A glance at the comparative glossary given 

 above will show numerous instances where words evidently the same 

 have been caught and transcribed very differently. Then, too, there 

 are many individual and local differences in pronunciation and dic- 

 tion (Skottsberg, d, 605; Hyades, loc. cit.). The addition or 

 omission of s, sh, I, etc., whatever be the explanation, has been noted 

 already. We may recall, too, that the observers themselves who 

 gathered the various vocabularies represented six or seven different 

 European languages, and naturally have caught and transcribed the 

 native words somewhat differently. For instance, H usually ex- 

 pressed by e what Fi expressed by d or a; H and Fi frequently omit 

 the final r where the others give it; H in several instances inserts an/ 

 or m where Fi omits it, etc. Or compare Lu and Se, both gathered 

 from the same Hagenbeck group of natives in Europe: eye — ^Lu, 

 te'leh-kwa, Se, decorliqua; teeth — Lu, che'riFtil-Tcwa, Se, tscMligiqua; 

 tongue — Lu, le'Jcel-Jcwa, Se, lecorqua, lekkersqua, etc. The above 

 causes largely account for many of the minor differences between the 

 various vocabularies and between the two groups, H-Fi and Bo-Sk. 



The more radical differences are probably due first of all to misunder- 

 standing on the part of the native informants. Admiral Fitz-Roy 

 obtained his words largely by signs, although his natives learned to 

 speak a little English. ''I found great difficulty in obtaining words, 



