478 THE PRIMITIVE HOME OF THE ARYANS. 



possessed by the parent speecb, is evidence that the object denoted by 

 it was unknown to the speakers. But the assumption is contradicted 

 by experience. Because the Latin equus has been replaced by caballus 

 in the modern Eomanic languages, we can not conclude that the horse 

 was unknown in Western Europe after the fall of the Roman Empire. 

 The native Basque word for a "knife," haistoa, has been found by Prince 

 L.-L. Bonaparte in a single obscure village; elsewhere it has been re- 

 placed by terms borrowed from French or Spanish. Yet we can not 

 suppose that the Basques were unacquainted with instruments for cut- 

 ting untd they had been furnished with them by their French and 

 Spanish neighbors. Greek and Latin have different words for "fire;" 

 we can not argue from this that the knowledge of fire was ever lost 

 among any of the speakers of the Indo-European tongues. In short, 

 we can not infer from the absence of a word in any particular language 

 that the word never existed in it ; on the contrary, when a language is 

 known to us only in its literary form it is safe to say that it must have 

 employed many words besides those contained in its dictionary. 



A good illustration of the imposibility of arriving at any certain re- 

 sults as long as we confine our attention to words which appear in one 

 but not in another of two cognate languages is afforded by the Indo- 

 European words which denote a sheet of water. There is no word 

 of which it can be positively said that it is found alike in the 

 Asiatic and the European branches of the family. Lake, ocean, even 

 river and stream, go by different names. A doubt hangs over the word 

 for "sea;" it is possible, but only possible, that the Sanskrit jM^/ms is 

 the same word as the Greek novToq, the etymology of which is not yet 

 settled. Nevertheless, we know that the speakers of the parent lan- 

 guage must have been acquainted, if not with the sea, at all events 

 with large rivers. Natis, "a ship," is the common heritage of Sanskrit 

 and Greek, and must thus go back to the days when the speakers of the 

 dialects which afterwards developed into Sanskrit and Greek still lived 

 side by side. It survives, like a fossil in the rocks, to assure us that 

 they were a water- faring people, and that the want of a common Indo- 

 European word for lake or river is no proof that such a word may not 

 have once existed. 



The example I have just given illustrates the second way in which 

 the attempt has been made to solve the riddle of the Indo-European 

 birthplace. It is the only way in which the attempt can succeed. 

 Where precisely the same word, with the same meaning, exists in both 

 the Asiatic and the European membei'S of the Indo-European family — 

 always supposing, of course, that it has not been borrowed by either 

 of them — we may conclude that it also existed in the parent speech. 

 When we find the Sanskrit asivas and the Latin equus, the exact pho- 

 netic equivalents of one another, both alike signifying "horse," we are 

 justilied in believing that the horse was known in the country from 

 which both languages derived their ancestry. Though the argument 



