42 SEXUAL DIMORPHISM 



modifications in the growth of living tissues is enough for 

 my present purpose. I make no attempt in the present work 

 to give an account of our knowledge on this subject. With 

 regard to the question of the inheritance of such effects, it 

 may be true that at present we have no direct or experimental 

 proof of its occurrence, or not sufficient proof. But on 

 the other hand its impossibility has not been established, 

 and my argument is merely that the hypothesis of 

 such inheritance as I have formulated it, affords the only 

 sufficient explanation of the phenomena which I pass in 

 review. 



Weismann in one of his essays maintains that the in- 

 heritance of acquired characters being theoretically impossible, 

 a theory of the selection of variations arising in and from the 

 germ is our only hope of escape from a supernatural explana- 

 tion of evolution. I maintain, on the contrary, that theories 

 of selection being found on application to the facts to be 

 insufficient for their explanation, and the theory of the 

 inheritance of acquired characters being found to harmonise 

 with the facts, we are logically bound to believe that such 

 inheritance does take place, at any rate until some other 

 explanation can be found. I do not concern myself with the 

 question how such inheritance can be produced, it is a fact 

 that the modifications with which I have to deal are hereditary, 

 and my object is to produce inductive evidence that they were 

 determined by special stimulations. 



I proceed, therefore, to examine the most important 

 examples of secondary sexual characters in detail, taking the 

 chief divisions of the animal kingdom in systematic order. 

 In this survey I shall endeavour to show that in every case 

 the development of the special character is associated with 

 special habits or conditions that necessarily involve stimula- 

 tions closely corresponding to the character in question, and 

 absent where the character is absent ; also that there is 



