

144 THE THEORIES OF EVOLUTION 



selection. How can we, however, account for varia- 

 tions, when ancestral plasms are transmitted to the 

 offspring without being ever affected by any outward 

 influence? According to Weismann, variations re- 

 sult from sexual generation which combines the germ 

 plasms of the parents and the ancestral plasms they 

 contain in a way which varies for every product ; it is 

 upon this variety that natural selection exercises itself. 



When two germ plasms, slightly different from 

 each other, are united in one cell by fertilisation, their 

 differences are preserved in the sexual cell thus pro- 

 duced. This cell, when mature, undergoes a process 

 which changes the constitution of its plasm; it sends 

 forth polar particles, liberating certain chromosomes 

 and also a certain number of ids, while some other ids 

 are allowed to remain in the cell. Here we have a 

 first cause of variation. Then there arises a compe- 

 tition for food and space between the various ids, de- 

 terminants and biophors, and as some of them emerge 

 victoriously from the struggle, certain ancestral char- 

 acters reappear to the exclusion of certain others. 



Let us suppose that character A can assume four 

 different aspects and that the four corresponding de- 

 terminants are a', a", a"', a"" ; let us suppose also 

 that for some reason which we will not discuss, the 

 determinants of type a' predominate in the germ 

 plasm of the father, constituting 80% of the sum 

 total of its determinants; the balance, or 20% being 



