THE DAHLIA 339 



plenty of material for selection in any lot of 

 seedlings. 



Still it is well to try to gain some new com- 

 binations by the use of wild strains, and in this 

 my expectations have been realized. 



One of the faults of the dahlia, even in the best 

 varieties, is that there is a tendency to expose the 

 center of the flower, owing to the fact that not all 

 of the stamens have been transformed into florets 

 even in the most developed varieties. The result 

 is that in a dry summer, or toward the end of 

 the season, even good varieties may fail to show 

 the fully rounded head that is prized by the 

 connoisseur. 



Through selection this defect was overcome, 

 causing the heads to fill out altogether, so that 

 they were double to the very center, even at the 

 end of our dry California seasons. A number of 

 varieties were thus perfected. 



As the ideal sought was approximated, the 

 flowers produced less and less seed, and 

 the perfectly double ones produced none 

 at all. 



So the races thus developed must be propa- 

 gated altogether from the tubers. This, indeed, 

 is not an insuperable objection, inasmuch as 

 this is a common way of propagating the dahlia. 

 But, of course, there is always an added merit 



