278 THE THEORIES OF EVOLUTION 



selection gradually eliminates the individuals in which 

 an adaptive variation does not coincide with the cor- 

 respondent innate variation. Germinal variations are 

 really the ones which are transmitted hereditarily but 

 as they facilitate the unavoidable ontogenetic adapta- 

 tions it is the latter which seem to be directly trans- 

 mitted. 



The theory of coincident variations also disposes of 

 another objection to natural selection, an objection 

 based upon parallel adaptations, neither of which 

 taken separately, would serve any purpose; for in- 

 stance, antlers would be useless to a deer, unless the 

 muscles of neck and shoulders were developed in pro- 

 portion. This argument was used by Spencer in the 

 controversy with Weismann in order to demonstrate 

 the heredity of acquired characters. Here, again, 

 however, heredity is only apparent; it is, in reality, 

 an innate variation which causes the antlers to develop 

 in some individuals more fullv than in others; onto- 

 genetic adaptation brings about, afterwards, the cor- 

 responding enlargement of the neck and shoulder 

 muscles which renders the modification of the antlers 

 useful. On the other hand, it may happen in one of 

 the following generations that an innate variation 

 strengthens the muscles. This additional strength 

 would be useless and disappear if the previous devel- 

 opment of the antlers (an innate variation preserved 

 thanks to an acquired variation), had not assigned it 



