430 EXPERIMENTS WITH PLANTS 



possible combinations is so great, and the results so 

 bewilderingly varied, that short-cut methods must be 

 used. In other words, the plant- breeder must be able 

 to iadge beforehand what the result of crossing certain 

 plants will be, in order not to waste time on profitless 

 experiments. It is just here that the opportunity 

 comes for the highest skill, based not only on 

 empirical knowledge but on a profound insight into 

 the laws of heredity and variation and a sound 

 philosophy of nature. Moreover, there must be a 

 clear-cut ideal present in the mind of the worker, 

 toward which he persistently strives and from which 

 he refuses to diverge even for the most promising 

 side issues. He wields forces which at best are but 

 partly understood; they manifest themselves in be- 

 wildering variety: to hold persistently to a definite 

 ideal is the surest path to success. 



There is no mystery in the method by which these 

 crosses are made. Mr. Burbank's methods of pollina- 

 ting Plums will serve to illustrate the matter. The 

 flowers which are to furnish the pollen are carefully 

 gathered a daj^ or so beforehand, -the pollen sifted 

 out and kept in a cool place. The tree to which the 

 pollen is to be applied is deprived of most of its 

 blossoms, in order that the remainder may be sure 

 to develop and that there may not be too many to 

 look after properly. The blossoms which remain on 

 the tree are prepared by cutting away the petals 



