SYSTEM. 143 



that two sister tribes may have been able, at some past time, 

 to create on each side of a mountain two different idioms, 

 which may produce in their turn two families of languages 

 absolutely irreducible one from the other. This is what would 

 take place, according to M. Renan, when the sons of the same 

 parents, separating on the sides of the Imaus, became the 

 double branch from whence have sprung the Semites on one 

 side, and the Aryas on the other. This would be the explana- 

 tion of the fact so embarrassing for anthropologists, that 

 physical characteristics are sensibly identical among the Se- 

 mites and Europeans, whilst these races are as distinct as 

 possible in the matter of language. Now, we may even go 

 further, and infer from these facts that the common species 

 from whence the Semites, on the one hand, and the Aryas, 

 on the other, are descended, did not yet know how to 



Inversely to Latham, some anthropologists have given, in 

 our opinion, too little importance to language : we speak espe- 

 cially of Edwards and M. Omalius d'Halloy.* The .truth lies, 

 doubtless, between these two extremes. It must be acknow- 

 ledged that language can very often furnish excellent evidence, 

 but it must not be forgotten that it shows at the same time a 

 more rapid liability to change than moral characteristics and 

 corporal form. Niebuhr seems to us to be right when he 

 insists upon the precautions to be taken in order to apply 

 philology in a useful manner to the determination of races, and 

 he concludes that we must give the greatest attention to physi- 



framed by Mochus, the Phoenician, and afterwards improved by Democritus 

 and Epicurus ; and though it is part of a system in which the first men are 

 represented as having grown out of the earth, like trees and other vegetables, 

 it has been adopted by several modern writers of high rank in the republic 

 of letters, and is certainly in itself worthy of examination." Encyclop. Brit., 

 vol. ix, p. 530, 1797. EDITOR.] 



* I do not here mention thS opinions of the Swede (see Latham, Celtic 

 Nations, p: 2), who thinks that important changes can be introduced into a 

 language by certain customs of a people, who change, for instance, the lips 

 for the nostrils, and thus substitute nasal for labial consonants. These facts 

 are, perhaps, true in the detail, but they ought not to have much importance, 

 as they do not alter the specific and personal character of the language, 

 which is far from consisting in the relative number of one or two kinds' of 

 letters. 



