A FEW OLD FAVORITES 127 



members, and you cannot be at all sure as to 

 what results you may attain by sowing seed from 

 any individual plant. 



But these complications result in part from 

 the fact that the different columbines are so 

 easily crossed by the bees. This is a case where 

 there is no difficulty in effecting hybridization; 

 the difficulty is to prevent crosses that are not 

 desired. If the plants are shielded from the 

 visits of the bees, and careful hand-pollenizing 

 is effected, there is no great difficulty in combin- 

 ing the different forms in such a way as to get 

 very definite results, and the hybrid forms may 

 be fixed by careful selective breeding. 



Of course, when you deal with a spurless form, 

 if the individuals that you use are themselves 

 hybrids of the first generation of a cross between 

 a spurred and a spurless variety, their progeny, 

 when they are crossed with a spurred variety, 

 will be in effect second-generation hybrids and 

 only half of them will be spurless. But this, 

 again, merely illustrates the familiar segregation 

 of characters and the reappearance of the reces- 

 sive trait — in this case the spurred condition — 

 in a rather definite proportion of the second- 

 generation progeny. 



Another anomaly among the columbines that 

 offers good opportunity for experimental tests 



