ON SOME FAMILIAR COMPOSITES 



spurs that are ever offered by seedsmen or florists. 

 As already stated, my chief experiments began 

 with the use of the hybrid larkspur as a seed par- 

 ent, but of course the hybridizing experiments 

 soon blended the strains of many of the other spe- 

 cies, until the larkspur colony, like so many others 

 of flower groups, is of such conglomerate ancestry 

 that the precise proportions of the different strains 

 in any given race are not traceable. 



Needless to say, selection has been carried for- 

 ward along with the hybridizing experiments, 

 these two methods always being complementary. 

 Particular attention has been given to size of 

 flower, vigor of plants, and resistance to insects 

 and disease, as well as that of multiplication by 

 division, at the same time that compactness of 

 growth and brilliancy of color of flower have been 

 carefully regarded. 



One of the worst faults of the larkspur is that it 

 tends to grow too tall, with a stalk that does not 

 support it, so that it requires to be staked. But 

 my hybrid larkspurs have been so selected that 

 they are compact in growth, and able to support 

 themselves even in a moderate gale. 



All the characteristic larkspur colors are rep- 

 resented among the new varieties, and in addition 

 there are combinations of color that have never 

 before been seen, I think, in the larkspur. Some 



[171] 



