306 THE DOCTRINE OF DESCENT. 
The rest is incidental and a matter of detail. The ‘ 
question so often ventilated, and now thoroughly worn 
out, whether mankind is descended: from one or more 
pairs, is solved by the inference that the stock in which 
language first arose, separated itself gradually from its 
animal progenitors, and that the selection which led to 
language and reason necessarily took place among large 
communities of individuals. The scriptural conception 
of the unity of the human race would be more nearly 
approached if all linguistic families pointed to a single. 
source. But if it could be shown that certain linguistic 
families lead to utterly discordant roots, the investiga- 
tion of nature might furnish the inevitable corollary that 
language originated in various parts of the world,— 
in other words, that a separation into species. took 
place before selection had reached the point of forming 
language. The latter case is by far the most probable, 
and is, in fact, received as the only one possible by most 
of the linguists occupied with this question, and is most 
especially defended by Friedrich Miiller.° “At the 
time,” he says, “when there were races and no nations, 
man was a speechless animal, as yet, entirely destitute 
of the mental development which rests upon the agency 
of language. Independently of the premisses unfolded 
by natural history, this hypothesis is forced upon us by 
the contemplation of the languages themselves. The 
various families of languages, which linguistic science 
is able to discriminate, not only presuppose, by their 
diversity of form and material, several independent 
origins, but, within one and the same race, they point to 
several mutually independent points of origin.” 
In order to afford the reader some notion of the con- 
