RTHNOLOOy OF TliK iNDO-PACIFIC ISLANDS. 



181 



allow not only tliat'tio cllinic Uwtory of tlir parlicr rnnes am! hiii- 

 miages of Imlia and AsonnsiJi is iiiiiniately cotinecled wiih tlmt of 

 other Asiatic formation!?^ but that hr^rcr iinil morn exhmnfit^ve ftn\i\o~ 

 rm'iom of the affinitive:^ of roots nm\ vocabloe will oertaiuly leatl to 

 many [lositive hiijioricHi rosulia. But ifie com|mnitive glossology 

 of th« ottifir Iiinguaj^ifs of Asin ftml of ihe world mui*t be prosHCU- 

 led &imiiliantoualy, for ii ia cloar tluit ilm history of every separate 

 vocabulary becoinfa more ami more im]>lictiletl in that of ol hers, 

 and emhi-aetis a wider and wider circle of retalionahip the rurtlier 

 onr researclses penetrate into antiquity. 



As each succoi^sive furniation of I rania becomes better flefined, 

 a clearer light will he thrown on the hiter stages of the Drnvirlan. 

 But much of its more fundamenfjtl hiijitory will eonrinue lo liepend 

 on Ihe progress of iiiiiverssit romptiraiive glossology. Although 

 in phonology, ideolotry and glossary it h distinctly connected with 

 the Scydiic, and nho in a lesd tiegree with the Caucasinn and 

 Afrieo-Semitic alliances, it has so large a mass of pecniiaritiea aa 

 to prove that, since the eras in wJach that connection arose, tlic 

 Inngnages of Western Asia and probahly of all Asia have under- 

 gone great changes* At one time Dravirian or AustratisiU may 

 have closely re^jcmbled languages of ihe Panjah, of Pej-sia or of 

 Upper Asia, but no ethnologist would expect to finil snch a resem- 

 blance now. From all the preceding imtications we are u'arranted 

 in concluding that ethnic movements similar to the historical one?^ 

 aometimes rapid, and at other times gradual, have in all eras been 

 going on from S. W. Asia to India and from India to Uitraindia 

 and Asonesia. These movements have alwavs left "lossarial 

 trac^ of greater or lesa importance, and we may therefore hojie 

 that in the progress of cllniology each wilt be more or luiia clearly 

 €edned. 



