LUTHER BURBANK 



is also a slight variation in the color of the flower. 

 So there is opportunity for selective breeding. 

 Moreover, I judge from physiological characteris- 

 tics that the plant should cross readily with the 

 brodiaea, although I have not attempted to make 

 the cross. 



It is almost certain that improved varieties 

 might be obtained by hybridization. 



There is a bulbous plant called Alstroemeria, 

 that is botanically related to the Amaryllis rather 

 than to the true lilies, which offers possibilities of 

 tuber improvement. The plants are natives of 

 Western South America. I grew seedlings and 

 hybrids by the ten thousand for several years, and 

 became convinced that if the roots and tops could 

 be taught to grow in a more compact form this 

 would become a very popular flower, and perhaps 

 also a very valuable food plant, as the roots are 

 sometimes eaten and are quite palatable and nutri- 

 tious. I have worked on the species known as A. 

 Chilensis, A. pulchella, and A. Brasiliense, and 

 subsequently on a large number of new species 

 from Chile. 



A great variety of colors and combinations 

 occur in the hybrid forms that may be fixed by 

 selection. I am endeavoring to obtain a more 

 hardy strain with improved flowers and more 

 compact growth. 



[260] 



