HI 



]an[;UiTge3 have always been iubject lo t-hiiugp^ piinilar in k\ml 

 lo those which ar<i now going on in the world, atnl havu becu 

 doing BO thfoujfhout Jiisioricul ^jeriotli. 



In IfKtiu the Dravirian formatiDn hai CL'nseJ io be difTuf^ive 

 and assimilalive. It has lon^ been ex|jo!*e*l to llie iniiiK nc<i of 

 Ihti Sanftkrit and of the norlIn.in In<lian lon^^ues that were early 

 assinahiled in a gicaier or leesi di gree to Saiiskiit In ihu Dm- 

 virinri family wc have therefore to note the main a I notion of rhe 

 difl'crciit langfiiagt?^ and disdt-cis, and ihc anion o\i each of the 

 f aaskrit and oi the Sanskritised or prakrit ton^^nes of the north. 

 The very close degree in which iUp. Dravirian Un^uugea of South- 

 ern India are rehitt'd to each olher and to the loiij^t Uhraindiatitsed 

 languages of the Vindyas, in [ihonology and itlcolo^y, has appeared 

 from the d^ild in chap. IV. They are dialect* of one tongue, 

 atkii they appear to differ less from each other than the Philipitte 

 Iungnagc-8, The dialecrtc diiicordaneei* are exactly the same in kind 

 as those which prevail amonght the Phiiipine and other groups of 

 At^onesian languages, or amongfit the Asiatic memhei-s of the 

 8emilico-Lthyan formation. All the great iumilies that have been 

 recognized ehow mut.h larger mutual deviations in their com- 

 jionent languages, and we must include AustmHan to give tlte 

 loriuation a eomprchcngiveness sindliir to tlie Beylhic or the 

 Sens itico- Libyan. Even the Indo-European and the Malag:i8y- 

 Polynesian are much more diversifiiud than the continental or 

 KoUTamuliaii division. All these widely disseminated families 

 present single iaaguageu or groups that, from long and complete 

 separation, have become alienaled from each other in the greiircr 

 number of their roots, in phonology and even iji many deiaih of 

 ideolo;^y. The transitions are seldom so abrupt aa from ihu 

 Dravirian lo tlic Australian, but this arises from the former bei ig 

 only ihe last continental and the hutcr the last insular remnant 

 of a once continuous and widely expanded family, that was early 

 disjoined, and has evor since been subjected in iti» two diviisions 

 to the influence of formations of opposite ch-iracter, — theScythico- 

 Iranian tending in the conttnejital division to give a more flexional 

 development to the ]>rimary structure which it has in cojnmon 

 with them, — and tJic Niha-Polynesi.ui tending to arrest the natural 

 flejtional development und concretion of llie insular division, and 



