A FEW OLD FAVORITES 131 



crossed, and you will have opportunity to test 

 the color variation in heredity in some of their 

 simplest relations. There are, to be sure, many 

 species of Campanula, and it is true that the 

 garden varieties are likely to have been hybrid* 

 ized. I have, for example, raised seedlings from 

 C, rotundiflora, white variety, without securing 

 any white ones, all returning to the original blue 

 color. It will be necessary, therefore, for you to 

 test your varieties by raising plants of uncrossed 

 seeds at the same time that you are making the 

 cross-pollinations. But this complication will 

 only add interest to the experiment. 



The many tribes of Coreopsis give opportunity 

 for experiments of equal interest. These plants 

 are composites, and in hybridizing them it will 

 be necessary to use the method detailed in our 

 story of the dahlia, washing away the pollen 

 before applying pollen from the other flower. 



The different members of the family vary in 

 color from deepest purplish crimson to light 

 yellow and white. There are numerous species 

 under cultivation, and there are wild ones grow- 

 ing as roadside weeds that are readily accessible 

 The variability of the different races makes them 

 an interesting race with which to work. 



My own work with the tribe has included a 

 good many species, the most important of which 



