[760] Am. Anth. XIV. 1912, p. 55.
[761] Such disintegration is clearly seen in the Carib still surviving in Dominica, of which J. Numa Rat contributed a somewhat full account to the Journ. Anthr. Inst. for Nov. 1897, p. 293 sq. Here the broken form arametakuahátina buka appears to represent the polysynthetic arametakuanientibubuka (root arameta, to hide), as in Père Breton's Grammaire Caraibe, p. 45, where we have also the form arametakualubatibubasubutuiruni = know that he will conceal thee (p. 48). It may at the same time be allowed that great inroads have been made on the principle of polysynthesis even in the continental (South American) Carib, as well as in the Colombian Chibcha, the Mexican Otomi and Pima, and no doubt in some other linguistic groups. But that the system must have formerly been continuous over the whole of America seems proved by the persistence of extremely polysynthetic tongues in such widely separated regions as Greenland (Eskimo), Mexico (Aztec), Peru (Quichuan), and Chili (Araucanian).
[762] R. de la Grasserie and N. Léon, Langue Tarasque, Paris, 1896.
[763] J. E. R. Polak, Ipurina Grammar, etc., London, 1894.
[764] The Eskimo Tribes, their Distribution and Characteristics, Copenhagen, 1887, I. p. 62 sq.
[765] In fact this very word was first given "as an ordinary example" by Kleinschmidt, Gram. d. Grönlandischen Sprache, Sect. 99, and is also quoted by Byrne, who translates: "They disapproved of him, because he did not give to him, when he heard that he would go off, because he had nothing" (Principles, etc., I. p. 140).
[766] "Indian Linguistic Families of America north of Mexico," Seventh Ann. Rept. Bureau of Ethnology, 1885-6 (1891). See also the "Handbook of American Indian Languages," Part I by Franz Boas and others, Bureau of American Ethnology, Bulletin 40, 1911. The Introduction by F. Boas gives a good general idea of the characteristics of these languages and deals shortly with related problems.
[767] Following this ethnologist's convenient precedent, I use both in Ethnology and here the final syllable an to indicate stock races and languages in America. Thus Algonquin = the particular tribe and language of that name; Algonquian = the whole family; Iroquois, Iroquoian, Carib, Cariban, etc.
[768] Forum, Feb. 1898, p. 683.
[769] Studies of these languages by Kroeber and others will be found in University of California Publications; American Archaeology and Ethnology, L. 1903 onwards. Cf. also A. L. Kroeber, "The Languages of the American Indians," Pop. Sci. Monthly, LXXVIII. 1911.