THE EVOLUTION OF LIVING BEINGS. 89 



sisting of four different species : brown — violet — gol- 

 denyellow — and white-flowering ones and their hybrids. 



With this demonstration of course the real origin of 

 these species is not explained. 



It probably lies, as experiments with Antirrhinums 

 show, in a previous cross between two individuals 

 belonging to different Linneons. 



Such a cross gives us — if the forms originally cros- 

 sed were pure — a uniform F lf each individual of 

 which, after selfing, gives rise to a large number of differ- 

 ently constituted individuals, some of which are hetero- 

 zygotes while others are homozygotes or pure species. 



Every cross between two individuals, differing in 

 many respects, consequently gives rise to the produc- 

 tion of a mass of differently constituted types, far grea- 

 ter than the number of separate types usually found 

 within a wild Linneon. 



This is caused by the fact that of the number of 

 different forms which were born from a cross, only a 

 small number survives. 



The Linneon is consequently the 

 vestige, consisting of a comparatively small num- 

 ber of differently constituted types, of that very 

 large number of types born as the 

 result of a cross. 



The number of types born from crossing in nature, 

 is usually even much larger than that resulting from 

 a cross we effect purposely in our experimentgarden, 

 because in nature crossing is not limited to a single ini- 

 tial cross, but usually a large number of crosses occurs 

 at the same time. This can be esasily demonstrated by 



