No. 2. J CAMPBELL AMERICAN INDIAN TRIBES. 79 



their analogues in the Ceram babuf, the Gilolo <t/od and the 

 Ceram nifnn. Two peculiar cases of inversion are presented in 

 the AVilliamct fshita-pimia, girl, and yrian-f.sal, river, which in 

 Wahai, one of the Ceram dialects, are pina-hieti and tolo-rmaina. 

 For further examples I refer to my comparative vocabularies in 

 the Canndian Journal^ as those stated are amply sufficient to 

 establish a connection of the Sahaptin family with the Malays of 

 the Moluccas group. 



In this sketch I have not been able to fulfil all the conditions 

 of perfect induction, but sufficient evidence, has, I think, been 

 adduced to make a case worthy of fuller investigation. The 

 theory of a Polynesian migration to America is one that has been 

 long and generally held, but held in so loose and indefinite a way 

 that it has been barren of ethnoloirical results. The Rev. 

 Richard Garuett has indicated the presence of Polynesian gram- 

 matical forms in South America, but has vitiated his comparison 

 by dragging into it the Dravidian languages, which are Turanian 

 as distinguished from Malay. Now, if an American language 

 can be proved Malay in origin, it is thereby cut ofi from all 

 Turanian connection, for the two systems are radically different. 

 Dr. Edkins, of Pekin, recognizes the distinction, and finds that 

 "on the American continent Turanian and Polynesian linguistic 

 principles meet in the various Indian languages ; but so far is 

 he from allowing the relationship of Turanian with Polynesian 

 that he maintains " a Polynesian immigration from the Ocean 

 and a Turanian immigration by the Aleutian Islands and by Ice- 

 land and Greenland, which united to form the population of the 

 American continent." Still, like Mr. Garuett, we find Dr. Ed- 

 kins looking for his Polynesians in South America and in Mexico. 

 If the Mexicans and Peruvians be Polynesians, they are Poly- 

 nesians with Turanian grammar, vocabulary, religion and arts, 

 and consequently ethnology as a science, on American ground at 

 least, is impossible. Again Dr. Edkins seems to saj^ that in 

 individual American languages Turanian and Polynesian prin- 

 ciples meet. This is true to a certain extent, as we have seen in 

 the case of the languages of Malay-Polynesian origin, which, 

 intruding into a Turanian region, could not fail to be partially, 

 but very partially, influenced by their .^surroundings. Some Poly- 

 nesian principles may perhaps be detected in the Iroquois 

 dialects, though of this I am not sure, t can find none in the 

 Tinnch, the Dacotah, the Choctaw or the Peruvian ; they are 

 essentially and exclusively Turanian. 



