CHAP, xvni A "FAIRY TALE OF SCIENCE"? 223 



must be regarded as the more coherent and original. It 

 is almost as interesting as a fairy tale, if possibly not 

 much truer. To an outside critic, at any rate, Weis- 

 mannism seems to have grown latterly after the manner 

 of a false hypothesis, not after the manner of truth. It 

 has modified itself endlessly by adding on ingenious 

 epicycles. Instead of leading to new generalisations 

 and broad views of things, the changes have made it 

 complex and artificial-looking. True or false, the older 

 Weismannism is at any rate clear, clearer than the 

 new. And Mr. Kidd's sociology seems to appeal to 

 the Weismann of 1893, or of still earlier years, not 

 to the author of the later more hesitating statements. 



At first, then, Weismann had held that germ plasm 

 was never affected by the life of the organism in which 

 it was temporarily lodged. It was perfectly continuous, 

 absolutely stable. 1 Yet varieties occurred ; for evolution 

 occurred ; and there was no cause of evolution except 

 natural selection ; and natural selection could only work 

 upon given materials. Whence then did varieties pro- 

 ceed ? From amphimixis and from that alone ; in 

 other words, from the processes of bisexual parentage. 

 There was "nowhere else" for variations to come from 

 on this early and rigid theory of Weismann's ; and 

 the theory threw a delightfully definite and clear light 

 on the cloudy problem, what is the origin of varia- 

 tions? No doubt there was a difficulty here. If 

 individual variation is due simply to parentage, why are 

 not all the offspring of the same pair facsimiles of each 

 other ? Can science clear up this mystery ? Weismann 

 in his early phase explained it by the extrusion of one of 



1 Apparently the phraseology is Komanes'. To a layman it looks 

 tautological. Romanes himself (pp. 49, 86 of Weismannism) seems unable 

 to keep the two terms distinct in their application. 



