140 LUTHER BURBANK 



desirable to have a number of species, and it is 

 obviously worth while to make careful selection 

 in deciding which ones to transplant to your gar- 

 den. I have spent many days on a few acres of 

 ground, searching among the multitudes of 

 goldenrods and asters for the most beautiful 

 individual specimens. From these selected seed 

 was collected, or the roots themselves dug, to 

 furnish the basis for further experiment. 



Some of the wild forms seem almost perfect, 

 yet when taken under cultivation and care- 

 fully selected they even prove susceptible of 

 betterment. 



The hybrids, in my experience, are not as vari- 

 able as might be expected. One can seldom be 

 sure, in working with the goldenrods, that one 

 is working with pure species. 



But such complications, of course, give added 

 interest to the work of the plant developer after 

 he has the fundamentals of the method fairly in 

 hand. And I think of few problems that would 

 be more interesting than to attempt to untangle 

 some of the hereditary complications amdng the 

 goldenrods. The fixing of types by selection; 

 the improving of the best existing ones ; and the 

 development of new types by crossing — these are 

 all methods that offer opportunity for fascinat- 

 ing experiments. 



