<)V ACCIDENTAL LINGUISTIC RESEMBLANCES. 



33 



further investigation unnecessary, because it diminishes, in a rapidly increasing ratio, the 

 chances for merely accidental analogies. 



We may, therefore, safely assume that any single coincidence in words of two or more 

 syllables, or any two coincidences in radical syllables, furnish almost irresistible evidence 

 of a. national intercourse to which the coincidence is attributable. The presumption is 

 greatly increased by every additional coincidence of either kind, and if the concurrence is 

 frequent, or if it extends alike to derivatives and primitives, the hypothesis of mere 

 national intercourse, by commerce or conquest, becomes less satisfactory, and it is difficult 

 to imagine any sufficient explanation other than a common genealogy. As there are no 

 two known languages which are destitute of a large number of such coincidences, the 

 d posteriori evidence of a unity of language appears to be stronger than that of a, unity of 

 race. 



Any considerable number of accidental coincidences being, as we have shown, not only 

 improbable but morally impossible, the reasonable effort of a just philosophy, whenever 

 they occur, is to search for some law by which they can be satisfactorily explained. By 

 such a, search, carefully conducted, philology may gradually be placed on a "positive" 

 basis. It is certain that no adequate explanation has yet been offered for the various 

 linguistic phenomena, except that which is based on common descent, and if none other 

 should be found hereafter, the believers in the specific unity of man, whether upon scriptu- 

 ral, historical, psychological, or mere philosophical -rounds, may well be gratified at such 

 strong confirmation of their views. 



voi„ \in. 



