VI.] TESTING OF VARIETIES. 171 



winters or long dry seasons (see Essay I.); and the 

 development of thickened parts, as tubers and bulbs, in 

 regions of long -enforced rest. 



In short, the theories of evolution teach that the 

 keynote of progression, either in untamed nature or in 

 the garden, is adaptation to environment. The selection 

 of varieties to suit one's soil and climate and other con- 

 ditions, is really a fundamental requisite to success in 

 horticulture; and, if this is true, there must be a con- 

 stantly-increasing tendency for every locality and every 

 commercial demand to develop a variety of its own. So, 

 instead of coming nearer to the perfect all-round variety 

 in any fruit, we are continually getting farther away 

 from it, for what is perfection for one place may be 

 imperfection, or even failure, for another place. Vari- 

 eties are not distinct entities, which can be recom- 

 mended to growers like so many machines or imple- 

 ments, but they are complex combinations of various 

 attributes, so nicely adjusted that every change of con- 

 ditions is likely to disengage the composition, and often 

 so intangible, in comparison with others, that the nicest 

 description cannot distinguish them. 



I must now make an application of these remarks to 

 the testing of varieties by experiment stations, for this 

 is a subject in which every horticulturist is vitally inter- 

 ested. What varieties shall I plant ? This and similar 

 questions are always asked of the experimenter, and 

 people seem to think that it is one of the simplest ques- 

 tions to answer. At all events, it is the universal 

 impression that the experiment station officer, of all 

 others, should be able to answer it definitely. He has 

 the facilities and the time for making tests, and it seems, 

 upon the face of it, that he should have exact knowledge 



