24 TMH ETHNOLOGY OF THE IXIHAN A HCH M' Kl \ u«.>. 



livery wild or alpine region which it contains is surrounded, and 

 often partially jKHietratod, by clans of the four great ci vilified 

 peoples who are spread over all its more favoured and much of its 

 more mountainous and inhospitable tract*. Some of the Khuming 

 and other wild tribes towards the north and north-oast of the basin 

 of the Irawadi, may be completely mountain-locked and shut in 

 from all access, but it is more probable t hut oven they are every- 

 where directly or indirectly in contact with the Tibetans or Mung- 

 fen on the one fide and the Singfu and Klmnui on the mhr-r. The 

 valley .of 'lie Manipuri contains several tribes speaking distinct 

 dialects, and it is probable that other districts in the lest known 

 parts of Uie mountain lands between the Irawadi and the Brahma- 

 putra, as well as in the tij per basin of the former, present a similar 

 linguistic divergenrt-. Hut we have already learns] enough to be 

 certain that Eolation cannot exist in this region. In the nomadic 

 tribes of Anst alia we have instances of a certain degree of isolation 

 combined witii a considerable geographical range/ but etill retain- 

 ing thecharacter of primary seats, the increased limits of the wan- 

 derings being naturally and perennially prescribed by the character 

 of the region. It is to the eastward that we must look for tribes that 

 have been longest isolated. The seclusion of every Polynesian 

 tribe is evidently modern, and it is only amongst the j^apuanesians 

 that an ancient isolation is possible. * We know too little of them 

 to pronounce that none such will be found, but the character of 

 the regio , both as to lands and winds, tends to counteract the 

 maritime rudeness of the race* and it will !><> extraordinary if the 

 usual consequence of a better acquaintance with an insular 'people, 

 the discovery of an intercourse between it and other islands, does not 

 follow in the case of every tribe of a race which has spread itself 

 over all the limits of the southern monsoon. 



The larger or secondary locations, which result from the posses- 

 sion of canoes, are Btill so common in the Archipelago that we must 

 reserve their enumeration till we describe tho different tribes. It 

 is principally in the upper navigable portions of rivers that ihoy 

 preserve much of an original character, but many Email rivers are 

 still rhidJy in the occupation of distinct tribes. The most fertile 

 river and lute basins have immemorial ly been the Beats of large 

 and united communities, and every principal river basin, even 

 in the less productive regions, has served to assimilate the families 

 by whom it has been occupied, and to transmit foreign in- 

 fluences for into the interior. Save in Java, each considerable 

 river basin in the Archipelago has still so much that is peculiar 

 in its history and present population, as to demand a distinct place 

 for itself in Uie ethnography of the region. 



The next districts are eminently those which have determined 

 ths internal migrations of the region and the diffusion of foreign 

 races or their influence*. These arc the seas of the Archipelago, 

 which are again united into wide regions by winds. For all our 



