RECORDING EXPERIMENTS 51 



shows a very considerable increase of good quali- 

 ties over the first. 



After a record has been kept for four or five 

 seasons, a fair estimate may be made of the gen- 

 eral value of this particular fruit. If in addition 

 we know the characteristics of the parent forms 

 whether the ancestors were hardy or tender, 

 and the like we are now in position to form a 

 clear judgment as to the probable value of the 

 fruit. 



Such a record as this is absolutely essential to 

 actual progress. It is important, if for no other 

 reason, to prevent the experimenter from deceiv- 

 ing himself. It is very easy to imagine that a 

 certain product that has caused one much trouble 

 is better than some other; or that a fruit of a 

 given tree is larger than some rival variety. But 

 the record book enables one to put the matter to 

 a precise and definite test; it makes self-decep- 

 tion impossible; and it affords an invaluable 

 guide to further experimentation. 



There are thousands of graphic records such 

 as these on the shelves of my extensive library. 



I would not think of attempting to conduct 

 an intricate series of experiments for the develop- 

 ment of a new fruit without these plan books. 



When the experiment is finally completed, a 

 series of these loose leaves, properly collated, 



