COMPARATIVE rillLOLOGY. 263 



says : — ' I possess a collection made by the well-known natu- 

 ralist, J. Nattercr, during his residence of many years in 

 Brazil, of more than a hundred languages, lexically com- 

 pletely distinct, from the interior of Brazil.' And he adds : 

 — 'The number of so-called isolated languages — that is, 

 of such as, according to our present information, show no 

 relationship to any other, and which therefore form distinct 

 stocks of greater or less extent — is in South America very 

 large, and must, on an approximate estimate, amount to 

 many hundreds. It will perhaps be possible hereafter to 

 include many of them in larger families, but there must still 

 remain a considerable number for which this will not be 

 possible.' " 



I have quoted this hypothesis, as previously remarked, 

 because it appears to me philologically interesting ; but what- 

 ever may be thought of it by professional authorities, the 

 evidence which the American continent furnishes of a poly- 

 genetic and polytypic origin of the native languages remains 

 the same. And if there is good reason for concluding in 

 favour of polygenetic origins of different types as regards the 

 languages on that continent, of course the probability arises 

 that radical differences of structure among languages of the 

 Old World admit of being explained by their having been 

 derived from similarly independent sources.* 



• I may add that the hypothesis admits of corroboration from sources not men- 

 tioned by its author. For Archdeacon Farrar wrote in 1865 : — "The neglected 

 children in some of the Canadian and Indian villages, who are left alone for days, 

 can and do invent for themselves a sort of lingua franca, partially or wholly 

 unintelligible to all except themselves ; " and he quotes Mr. R. Moffat as 

 *' testifying to a similar phenomenon in the villages of South Africa (Mission 

 Trav(ls)" lie also alludes to the fact th.at "deaf-mutes have an instinctive 

 power to develop for themselves a language of signs," which, as we have seen in 

 an earlier chapter, embraces the use of arbitrary articulations, even though in this 

 case the speakers cannot themselves hear the sounds which they make. 



While this work is paising through the press an additional pnpcr has been 

 published by Dr. Hale, entitled, The Development of Lan^ua^i. It supplies 

 further evidence in support of this hypothesis. 



18 



