michelson: American Indian languages 231 



of comparative Indo-European phonology. Another example 

 of a case where words of two different stocks superficially re- 

 semble each other is Sanskrit nasta- "dead," Fox ne'tow" 

 "he kills." It will be recalled that the combination of a sibilant 

 followed by t appears as H in Fox. As soon as we note that 

 Sanskrit nasta- is composed of the elements nas-\-ta- and Fox 

 ne'tow 0, of the elements ne+ 'to-\-w a , the comparison ceases to 

 interest us. [The Fox word can, as I think, be reduced still 

 further in analysis, but this only still further emphasizes the 

 point at issue.] We should exercise the same prudence in com- 

 paring morphological elements. For example the verbal termina- 

 tion of Latin in the third person singular is t which superficiary 

 resembles Tsimshian t, to say nothing of similar terminations in 

 various Algonquian dialects. Again Greek nev, the verbal 

 termination of the first person plural, has an entirely fortuitous 

 resemblance to Ojibwa min in ni — min, ki — min of the inde- 

 pendent mode. 



Turning now to the classification of the languages belonging 

 to single stocks, — it must be said that very little work has been 

 done on this important topic in American linguistics. The 

 three stocks of which we have the best knowledge in this respect 

 are Salishan, Siouan, and Algonquian. To a certain extent 

 the classification is arbitrary. We select a number of salient 

 features and base our classification on it. In most cases we have 

 overlappings which are indicative of more than one association. 

 For example, Peoria fundamentally belongs with the Ojibwa 

 division of Central Algonquian languages ; at the same time there 

 are certain traits which clearly prove that it has also had an 

 association with the Sauk, Fox, Kickapoo group, also one with 

 Cree; and there are some indications of contact with Delaware- 

 Munsee. Nevertheless in spite of such short comings, we can 

 make classifications which are entirely satisfactory even to the 

 Indo-European philologist. The object of our classifications 

 is to determine the prehistory of the tribes of any given stock. 

 For example, the Abnaki dialects exhibit so many .special traits 

 in common with Shawnee, as wel as Sauk, Fox, Kickapoo, 

 that it is absolutely certain that in prehistoric times the tribes 



