98 LUTHER BURBANK 



But when selection has been carried to a stage 

 where we have one race of balloon-flowers pre- 

 senting plants that are uniformly of graceful and 

 attractive form, and another race that has the 

 flowers arranged in a satisfactory way on the 

 stalk, and a third race that produces flowers of 

 a brilliant white color, the materials are in hand 

 for an amplification of the experiment along 

 lines with which the reader is already familiar, 

 through which the desired combination of these 

 traits in a single race may be effected with almost 

 absolute certainty. 



THE COMBINATION or QUALITIES 



The method in question consists, of course, in 

 cross-pollenizing the best individuals of the three 

 new races. Of course, one cannot blend three 

 strains in a single cross-pollenizing experiment. 

 But one can cross-pollenize specimens of each 

 one of the three with each of the others, making 

 the cross reciprocal in all cases to make quite 

 sure. Each of the new hybrid races will thus 

 blend, in one way or another, the traits of two 

 of the parent forms. 



Selection being made to find the best types 

 among these two cross-bred races, the ones 

 selected will, of course, be interpollenized, and 

 their offspring, representing the second genera- 



