376 LUTHER BURBANK 



would be measured not in mere years but in 

 millenniums. For evolution is a slow process, 

 and the history of the development of natural 

 species is measured in geological eras. 



Selection and Mendelism 



Perhaps it may be worth while to illustrate 

 this matter a little more in detail, that we may 

 make clear precisely what manner of thing the 

 plant developer is doing when he produces a new 

 race by selection. 



We have stated over and over that the process 

 of hybridizing and the process of selection are 

 complementarJ^ One supplements the other. 

 In hybridizing we make possible new combina- 

 tions of the hereditary factors, and in selecting 

 through successive generations we isolate certain 

 definite combinations, and thus produce what 

 we call new varieties. Now it is frequently 

 stated by the experimenters who have paid 

 attention only to a few conspicuous characters 

 that Mendelize, that all possible combinations of 

 characters will occur among the second genera- 

 tion hybrids, provided only enough of these are 

 produced. 



Possibly this statement is correct. But it is 

 not susceptible of demonstration because it 

 would not be feasible to produce enough indi- 



