THE ETTWOLOOY OF TDK INDIAN ARCH IPO. AOO. 33 



To other nations which have struck me, I will endeavour to deter- 

 mine the geographical extension of the more important ethnic trait* 

 and their various forms, and to trace each to \\i source. This 

 enquiry connects itself closely with the history of the civilisation of 

 the Archipelago. At the end of this enquiry we shall be better 

 able to understand the respective values of the different kinds of 

 ethnic evidence. There is a great difference of opinion on this 

 Bubject. Some writers exhibit a strong tendency to find in every 

 similarity or coincidence in custom or kinguugc, between remote 

 and mutually isolated tribes, a convincing proof of their descent 

 from a common centre, and of the primitive unity of the human 

 raco. Others again only view such resemblances as proofs of the 

 tendency of the formative or artistic activity of human nature to 

 work in certain moulds or follow certain types, which are funda- 

 mentally implanted in it, or necessarily result from its intellectual 

 development. As in al! such cases the reaction of the one tendency 

 against the other widens the breach, and removes both to a greater 

 distance from the medium in which I believe the truth is to be 

 found. I shall also say something on the art of comparing lang- 

 uages, for the real value of such comparisons, and the conditions 

 necessary to preserve them from degenerating into mere specula- 

 tions, are far from being generally understood. 1 need do no more 

 here than allude to the ^reat assistance given to all our most archaic 

 researches, by that primitive chronology of mankind preserved 

 in the structure of languages. The true place of the As&ncsian 

 languages has never been determined. Duponceau omits them 

 altogether in his ideologic classification, and, as we have seen, 

 Pricnard and Bunsen incluJe them in their Turanian and Japhetic 

 alliances somewhat distrustfully. I shall endeavour to shew what 

 their place is, and, at the same lime, consider what value the 

 linguistic chronology can claim. At present I shall only odd thai 

 no real progress can be made in ethnology w ithout resting our 

 conclusions on a combination of every available kind of ethnic 

 evidence. Connections and relations can be discovered by pursuing 

 one branch of the subiect by itself. But no approach can be made 

 to a historical ethnology without an accumulation of evidence 

 respecting all the traits we have mentioned. 



The next division of the subject consists in a brief description of 

 each race and its country, its history and traditions, and its relations 

 to other tribes al the present and during historical times, — following 

 it as far as we can to all its migration*, till we lose it in another 

 tribe, or find that all further traces of its earlier life are obliterated. 

 As the histories of particular tribes are sometime* found to discon- 

 nect themselves from the district which thev now occupy, it b 

 necessary not only to follow each into its older locations, but to 

 trace back ^ the history of the country itself. 1. .; t: 1 . ; , .,. 

 has a twofold history, one special," and the other common to it 

 with several other parts of the Archipelago, or with the Archipe- 



