324 LUTHER BURBANK 



a certain number of the second generation of the 

 walnut hybrids. 



This wide diversity of form and vigor in the 

 first-generation hybrids is a rather unusual phe- 

 nomenon. As a rule, we have observed that first 

 generation hybrids are somewhat uniform in 

 character, and that the tendency to wide diversity 

 appears in the second generation. Attention has 

 more than once been called to the fact that the 

 discovery that such is the tendency among 

 hybrids was the one that put me on the track 

 of most of my successful plant developments. 



At the time when these experiments in hybrid- 

 izing the Japanese plum and the almond were 

 commenced, there were few, if any, other plant 

 experimenters anywhere in the world who had 

 fully grasped the principle that variation occurs 

 in the second generation, and that it is by raising 

 large numbers of second-generation hybrids from 

 which to make selection that the development of 

 new and useful varieties of plants may best and 

 most rapidly be carried out. 



This principle is so familiar to-day that horti- 

 culturists and botanists who refer to it very com- 

 monly overlook the fact that the recognition of 

 the principle is very recent. 



Thirty-five years ago I found it impossible to 

 convince most well-known horticulturists, bot- 



