154 



EVOLUTION AND ANIMAL LIFE 



360 



3Z0 



m 



240 



200 



160 



120 



80 



40 



processes involved in species-forming, that is, evolution. The 

 methods and phenomena of evolution are intimately linked 

 with — indeed throughout are based upon — the methods and 

 phenomena of variation. What causes variation is a contrib- 

 utory cause in evolution, 

 and one of the funda- 

 mental and all-important 

 causes. 



Concerning the causes 

 of variation, at least of 

 those of congenital varia- 

 tion, we are almost wholly 

 in the dark. Only such 

 influences as can affect 

 the actual germ cells are 

 presumably potent to ef- 

 fect congenital variation. 

 Such influences are not 

 proved to the satisfac- 

 tion of many biologists 

 to be numerous. In the 

 fusion of the germ cells 

 of two individuals, the 

 phenomenon called by 

 him amphimixis, Weis- 

 mann finds the most ef- 

 fective cause of variation. 

 Now the wider apart the 

 two parents are in struc- 

 tural and functional char- 

 acteristics, the greater is 

 the variation in their off- 

 spring likely to be. Hence hj^bridization, or the mating of 

 unlike parents, even to the degree of race and species un- 

 likeness, is a great resource of the breeder who would have in 

 his hands large variation. But if the parents are too unlike, 

 their mating, even if possible, proves sterile. Llsually parents 

 must be of the same species, although experiment has shown 

 that considerable extraspecific hybridization is possible. Among 

 cultivated plants and animals the artificially selected races 

 differ vei-y mucli, but these races are mostly easily hybridiz- 



Classes::": 

 Variates 3l3 40 



55 



388 



miscfl' 



09 total 905 



Fig. 92. — ^Frequency polygon of variation of 

 elytral pattern in 905 specimens of the Cali- 

 fornia flower beetle, Diabrotica soror, col- 

 lected at Stanford University, October, 1902. 

 (After Kellogg and Bell.) 



