5 o 4 THE POPULAR SCIENCE MONTHLY. 



reduced supply of water, and Balkash, Aral, and Caspian gradu- 

 ally shrank to their present dimensions. In the course of this 

 process the broad plains between the separated inland seas, as 

 soon as they were laid bare, threw open easy routes to the Cau- 

 casus and to Turkistan, which might well be utilized by the 

 blond long-heads moving eastward through the plains contempo- 

 raneously left dry south and east of the Ural chain. The same 

 process of desiccation, however, would render the route from east 

 central Asia westward as easily practicable ; and, in the end, the 

 Aryan stock might easily be cut in two, as we now find it to be, by 

 the movement of the Mongoloid brunet broad-heads to the west. 



Thus we arrive at what is practically Latham's Sarmatian hy- 

 pothesis if the term " Sarmatian " is stretched a little, so as to in- 

 clude the higher parts and a good deal of the northern slopes of 

 Europe between the Ural and the German Ocean ; an immense 

 area of country, at least as large as that now included between 

 the Black Sea, the Atlantic, the Baltic, and the Mediterranean. 



If we imagine the blond long-head race to have been spread 

 over this area, while the primitive Aryan language was in course 

 of formation, its northwestern and its southeastern tribes will 

 have been fifteen hundred or more miles apart. Thus, there will 

 have been ample scope for linguistic differentiation ; and, as ad- 

 jacent tribes were probably influenced by the same causes, it is rea- 

 sonable to suppose that, at any given region of the periphery, the 

 process of differentiation, whether brought about by internal or 

 external agencies, will have been analogous. Hence, it is permis- 

 sible to imagine that, even before primitive Aryan had attained 

 its full development, the course of that development had become 

 , somewhat different in different localities ; and, in this sense, it 

 may be quite true that one uniform primitive Aryan language 

 never existed. The nascent mode of speech may very early have 

 got a twist, so to speak, toward Lithuanian, Slavonian, Teutonic, 

 or Celtic in the north and west; toward Thracian and Greek 

 in the southwest ; toward Armenian in the south ; toward Indo- 

 Iranian in the southeast. With the centrifugal movements of 

 the several fractions of the race, these tendencies of peripheral 

 groups would naturally become more and more intensified in 

 proportion to their isolation. No doubt, in the center and in other 

 parts of the periphery of the Aryan region, other dialectic groups 

 made their appearance ; but whatever development they may 

 have attained, these have failed to maintain themselves in the bat- 

 tle with the Finno-Tataric tribes, or with the stronger among their 

 own kith and kin.* 



Thus I think that the most plausible hypothetical answers 



* See the views of J. Schmidt (stated and discussed in Schrader and Jevons, pp. 63- 

 67), with which those here set forth are substantially identical. 



