4l6 THE THIRD DISCOURSE 



Ancient World, and the population of this habitable 

 globe, that of Mr. Jacob Bryant, whom I name with 

 reverence and affection, has the beft claim to the praife 

 of deep erudition ingenioufly applied; and new theories, 

 happily illuftrated by an afTemblage of numberlefs con- 

 verging rays from a molt extenfive circumference : it 

 falls, neverthelefs,as every human work muft fall, fhort 

 of perfection; and the leaft fatisfaclory part of it feems 

 to be that which relates to the derivation of words from 

 Afiatick languages. Etymology has, no doubt, fome 

 ufe in hiftorical refearches ; but it is a medium of proof 

 fo very fallacious, that, where it elucidates one fact, it 

 obfeures a thoufand; and more frequently borders on 

 the ridiculous, than leads to any folid conclufion. It 

 rarely carries with it any internal power of conviction, 

 from a refemblance of founds or fimilarity of letters ; 

 vet often, where it is wholly unaffiited by thofe advan- 

 tages, it may be indifputably proved by extrinfick evi- 

 dence. We know d pofcriori, that both Jitz and hijo, 

 bv the nature of two feveral dialects, are derived from 

 ji lius ; that uncle comes from aims, and fir anger from 

 extra; that jour is deducible, through the Italian, from 

 dies : and rcffignol from lufcinia, or the finger in groves ; 

 thztfeiuro, ecurcuil, ar\& fquirre /,are compounded of two 

 Greek words defcriptive of the animal ; which etymo- 

 logies, though they could not have been demonstrated 

 a priori, might ferve to confirm, if any fuch confirma- 

 tion were necelTary, the proofs of a connection between 

 the members of one great empire; but, when we derive 

 our hanger, or Jhort pendant [word, from the Perfian,be~ 

 caufe ignoranttravellersthusmif-fpelltheword khanjar, 

 which, in truth, means a different weapon, or fandal- 

 wood from the Greek, becaufe we fuppofe that fandals 

 were fometimes made of it, we gain no ground in prov- 

 ing the affinity of nations, and only weaken arguments 

 which might otherwife be firmly fupported. That Cus, 

 then, or, as it certainly is written in one ancient dialeft, 

 Cut, and in others, probably, Cds, enters into the rom- 

 2 1 ofuio.i 



