50 THE OEIGIN OF ANEMAL-WOKSIIIP. 



tions. The moment we try it, the key unlocks for us with 

 ease what seems a quite inexplicable fact, which the cur 

 rent hypothesis takes as one of its postulates. Speaking 

 of such words as sky and earth, dew and rain, rivers and 

 mountains, as well as of the abstract nouns above named, 

 Prof. Max Miiller says : &quot; Now, in ancient languages every 

 one of tnese words had necessarily a termination expres 

 sive of gender, and this naturally produced in the mind 

 the corresponding idea of sex, so that these names received 

 not only an individual but a sexual character. There 

 was no substantive which was not either masculine or 

 feminine ; neuters being of later growth, and distinguish 

 able chiefly in the nominative.&quot; (&quot; Chips,&quot; etc., vol. ii., 

 p. 55.) And this alleged necessity for a masculine or 

 feminine implication is assigned as a part of the reason 

 why these abstract nouns and collective nouns became 

 personalized. But should not a true theory of these first 

 steps in the evolution of thought and language show us 

 how it happened that men acquired the seemingly-strange 

 habit of so framing their words for sky, earth, dew, rain, 

 etc., as to make them indicative of sex? Or, at any rate, 

 must it not be admitted that an interpretation which, in 

 stead of assuming this habit to be &quot; necessary,&quot; shows us 

 how it results, thereby acquires an additional claim to 

 acceptance ? The interpretation I have indicated does 

 this. If men and women are habitually nicknamed, and 

 if defects of language lead their descendants to regard 

 themselves as descendants of the things from which the 

 names were taken, then masculine or feminine genders 

 will be ascribed to these things according as the ancestors 

 named after them were men or women. If a beautiful 

 maiden known metaphorically as &quot; the Dawn,&quot; afterward 

 becomes the mother of some distinguished chief called 

 u the North &quot;Wind,&quot; it will result that when, in course of 



