ON THE COMPARISON OF ASIATIC LANGUAGES. 227 
of the reduplication of a root representing continuous action. 
In many grammars, such as the Akkadian, Egyptian, or 
Sanskrit, the reduplication has such a force, and it appears to 
have been the very oldest way of expressing the plural. 
Many animals appear to us to utter cries, expressed by such 
sounds as Mu and Mau, Ba, &e., and the names for crows 
and similar birds are taken from their caw. <A parrot can 
utter such sounds, and some we hear from a dog. But the 
great dividing line between human speech and animal cries 
seems to lie in the power, which no known animal has been 
proved to possess, of putting together, with an intelligible 
object, two distinct sounds, uttered with different parts of 
the mouth, and conventionally received as expressing a 
definite sense. And these double sounds we encounter in 
human speech in all the earliest languag ges to which we have 
access. ‘Thus from DU, “to go,” we obtain DUK, “to lead” 
from BHA, “to shine,” are formed BHAS, BHAK, BHAN; fer 
KA, “to call,” we obtain KAR, KAK, KAL,and KAN; from RA, “ to 
roar,” RAG, RAS, RABH; from PA, to “ go,” PAD, PAR, and BHAG. 
In some cases we can still trace the origin of the secondary 
root, as in KAK, to “ cackle,” which is a simple reduplication 
of KA, “to call.” The Chinese method of joining two roots in 
what is called a “ clamshell” word, for the greater distinction 
of the sense intended, seems to cast light on the formation 
of the secondary roots, so that RAG, for instance, might have 
been originally made up of RA, “to roar,” and KA, to or 
Whatever be the truth as to such speculation, it can, I think, 
hardly be doubted that the evidence will be found strongly 
in favour of an original community of true speech for 
Asiatic man. 
We are often reminded that questions of race and of 
language must be separately treated, since changes of 
language have occurred in various parts of the world. But 
it cannot be forgotten that in Asia, as far as we are able to 
speak of either a pure language or a pure race, even in the 
earliest ages, the great families of speech are found to be 
co-extensive with the ereat races which have used them 
pe oe the course of history. When languages change 
r die, it is usually because the old stock also changes or 
ae When conquerors hold a country they do not succeed 
in imposing their speech on their more numerous subjects, 
but themselves absorb into that speech words from the 
vocabulary of the native. Thus English has grown out of 
the mingling of the Latin and Teutonic and Celtic races, 
