XLII BUREAU OF AMERICAN ETHNOLOGY 
the other hand there is a spasmodic enrichment and modifica- 
tion, both in terms and in grammatic structure, produced by the 
shock of contact (whether peaceful or inimical) with other 
peoples—the changes consequent on conquest being especially 
important, as has been shown by different philologists. At the 
same time both the lexic and the structural forms—i. e., both 
words and sentences—are simplified through the natural tend- 
ency toward economy in expression. These and other proc- 
esses connected with the growth of language have been 
indicated in some detail in earlier reports. 
Now, on examining the aboriginal languages of America, it 
is found that many of them are interrelated in such manner as 
to indicate specific courses of development, and in all such 
cases the dominant process has been the union or blending of 
more or less diverse elements, both lexic and structural. This 
blending can be explained only as a record of intertribal con- 
tact, and the cases are so numerous—indeed, they are charac-: 
teristic of all of the aboriginal tongues—as to indicate that 
practically all of the native languages have been built up and 
shaped chiefly by the combination and blending of antece- 
dently distinct and presumptively discrete tongues. This con- 
clusion as to the development of oral speech in America is 
corroborated by the simpler history of the development of the 
so-called gesture speech, which was widely used by the Indians 
as a partial substitute for, and convenient supplement to, oral 
speech as an intertribal language. When the course of devel- 
opment ascertained by these comparisons is so extended as to 
apply to the entire assemblage of native American peoples, it 
at once becomes evident that the sixty linguistic stocks and five 
hundred dialects extant at the time of the discovery (themselves 
the product of long-continued combination and blending of 
distinct tongues, as the researches have shown) are indubitable 
records of still more numerous and still more widely distinct 
languages of an earlier time, and the more carefully the record 
is scanned the more numerous and the more distinct do the 
original components appear. It is accordingly a necessary 
inference that a large number of distinct, albeit simple if not 
inchoate, tongues originally existed in North America, and that 
