PRACTICAL PLANT-BREEDING. 425 
and semi-double flowers I had selected one of proved value as a useful 
breeder, after some years of experimental trial. The influence of this 
type not only carried the tendency of petal multiplication, but the seeds 
produced twin corms from time to time, an effect not manifested in the 
offspring of normal types. Experiences like this prove the contention 
advanced in the early part of this paper, that the hybridist does not need 
to spend time in special analysis for the purpose of securing examples of 
an interesting character, but that the great secret of success and satis- 
faction is large production from high-class composite parents bred for the 
purpose, and that by these means many new, valuable, and interesting 
types can yearly be produced, in addition to the highest ratio of useful 
and beautiful varieties, developed by practical plant-breeding. . 
