326 ANTIENT METAPHYSICS. Book IIL 



of the other, nor both derived from the fame original. But, in or- 

 der to prove that the languages were originally the fame, there muft 

 be many words in both agreeing in fenfe as well as lound, and 

 thefe fuch as muft have been original words in every language, de- 

 noting things which muft lirft have got names in every language ; 

 fo that when thefe names are the fame in two languages, we are 

 fure that the two muft have been originally the fame. 



My laft obfervation is, that, as the art of a language is lefs arbi- 

 trary and more determined by rule than either the found or fenfe of 

 the words, it is one of the principal things, by which the connedion 

 of languages with one another is to be difcovered : And, there- 

 fore, when we find that two languages pra£life the three great arts 

 of language, derivation, compofition, and fledlion, in the fame way, 

 we may conclude, I think, with great certainty, that the one lan- 

 guage is the original of the other, or that they are both dialedls of 

 the fame language. 



To apply thefe obfervations to the fimilarities which Mr Wilklns 

 has difcovered betwixt the Shanfcrit and the Greek ; — I will begin 

 with thefe words, which, as I have faid, muft have been original 

 words in all languages, as the things denoted by them muft have 

 been known in the firft ages of civility, and have got names; fo 

 that it is impoffible, that one language could have borrowed them 

 from another, unlefs it was a dervative or dialed of that language. 

 Of this kind are the names of numbers, of the members of the hu- 

 man body, and of relations, fuch as that di father^ mother^ and bro- 

 ther. 



And, firjl-t ^s to numbers, the ufe of which muft have been coe- 

 val with civil fociety. The words in the Shanfcrit, for the numbers 

 from one to ten, are ek, dwec, tree, chatoor, punch, JJoat, fopt, aght, 



nava. 



