SO far as I know, this is the first attempt at a systematic classification of the whole American race on the basis of language. I do not overlook Dr. Latham's meritorious effort nearly forty years ago; but the deficiency of material at that time obliged him to depart from the linguistic scheme and accept other guides.

While not depreciating the value of physical data, of culture and traditional history, I have constantly placed these subordinate to relationship as indicated by grammar and lexicography. There are wellknown examples in the ethnography of other races, where reliance on language alone would lead the investigator astray; but all serious students of the native American tribes are united in the opinion that with them no other clue can compare to it in general results. Consequently the Bureau of Ethnology of the United States and the similar departments in the governments of Canada and Mexico have agreed in adopting officially the linguistic classification for the aboriginal population within their several territories.

Wherever the material permitted it, I have ranked the grammatic structure of a language superior to its lexical elements in deciding upon relationship. In this I follow the precepts and example of students of the Aryan and Semitic stocks ; although their methods have been rejected by some who have written on American tongues. As for myself, I am abidingly convinced that the morphology of any language whatever is its most permanent and characteristic feature.