126 LUTHER BURBANK 



acteristic feature of the flowers of the wild col- 

 umbines, acts as a dominant factor in heredity. 

 This, of course, is what should be expected if it 

 be true that the newly developed characteristics 

 of a plant are dominant over the older ones. 

 But the case of the columbines furnishes another 

 interesting corroboration of this interpretation 

 of Mendelian heredity. 



In the course of other experiments with the 

 columbines numerous other species were brought 

 into the combination through successive hybridi- 

 zations, until my columbine colony carried the 

 strains of more than a dozen recorded species. 

 A most beautiful lot of hybrids resulted. Their 

 various members revealed nearly all the colors of 

 the rainbow. These were introduced to the 

 trade as mixed varieties, as it did not seem to be 

 worth while to fix the different types. On the 

 contrary, the variety of blossoms seemed to be 

 considered an advantage. 



But even if it had been desired to fix these 

 beautiful new types, it would have proved ex- 

 ceedingly difficult to do so. When you have 

 two or more species of columbine in combina- 

 tion, the hereditary complications are compa- 

 rable to those in the gourd family, to which we 

 have had occasion to refer. It seems as if every 

 member of a fraternity differs from all other 



