50 BUREAU OF AMERICAN ETHNOLOGY [bull. 40 



partly due to phonetic difficulties, and consist in the adaptation 

 of an unfamiliar group of sounds to the familiar similar sounds 

 of the language by which the word has been adopted. There may 

 be assimilations by which the grammatical form of a word is made 

 similar to more familiar forms. Furthermore, changes in the sig- 

 nificance of the word are common, and new derivations may be 

 formed from the word after it has once become entirely familiar, 

 like other native words. 



In this respect a number of American languages seem to be- 

 have curiously when compared with European languages. Bor- 

 rowing of words in Europe is particularly common when a new 

 object is first introduced. In almost all these cases the foreign 

 designation is taken over with more or less fundamental phonetic 

 modifications. Examples of this kind are the words tobacco, canoe, 

 maize, chocolate — to take as illustration a few words borrowed 

 from American languages. American natives, on the other hand, 

 do not commonly adopt words in this manner, but much more 

 frequently invent descriptive words by which the new object is des- 

 ignated. Thus the Tsimshian of Britisli Columbia designate rice 

 by a term meaning looking lilce maggots. The Kwakiutl call a 

 steamboat Jire on its hack moving on the water. The Eskimo 

 call cut tobacco being blown upon. Words of this type are in 

 wide use; nevertheless, loan words taken from English are not by 

 any means rare. The terms biscuit, dollar, coffee, tea, are found in 

 a great many Indian languages. The probable reason why descrip- 

 tive words are more common in American languages than in Euro- 

 pean languages lies in the frequent occurrence of descriptive nouns. 



We find, therefore, that there are two sets of phenomena which 

 must be considered in the classification of languages: (1) differences 

 which can easily be proved to be derived from modifications of a 

 single ancestral language; and (2) similarities which can not be 

 thus explained, and some of which may be due to the effects of 

 mixture. 



Origin of Similarities; by Dissemination or by Parallel 



Development 



Before we proceed with this consideration, we have to discuss 

 the two logical possibilities for such similarities. Either they may 

 be due to dissemination from a common source, so that they origi- 



