PLUM AND PRUNE PROFITS 169 



it is that the plant developer who experiments as 

 I have done cannot hope for a quick financial 

 return for his efforts. 



One chief reason why experimentation of this 

 order does not pay is that it was done so compre- 

 hensively, thoroughly, and on so large a scale. 



Where a man conducts plum improvement, for 

 example, as an adjunct to a nursery business, 

 there is no reason why he might not eventually 

 secure even a single improvement that could 

 directly pay him for his care and expense in 

 producing it. There would be no certainty as 

 to this, to be sure, as the chance of securing a 

 really good new variety is not better than about 

 one in ten thousand. That is to say, in handling 

 ten thousand seedlings, there would be no prob- 

 ability of securing more than a single good new 

 variety. 



But, on the other hand, sometimes even a small 

 lot of seedlings may give more than one good 

 variety, as has been the case several times. 



In any event, the nurseryman can carry out a 

 line of experiment on a moderate scale without 

 considerable monetary outlay. So at worst he 

 will lose very little. 



But where innumerable crosses are made and 

 thousands of seedlings are raised each year only 

 to be destroyed; and where all needed improve- 



