CLASSIFICATION. 53 



be of suificient value in classifying. In no case can an artificial 

 system fail to be natural in some degree. Groups of plants 

 whicb will resemble each other in one character likely to be used 

 in classifying will surely agree in other characters. 



A good example of an artificial classification, which is natural 

 in many particulars, is Dr. Warder's system, for apples given 

 on page 79, which is the best, natural or artificial, that we now 

 have. 



JSTatueal Classification. — ^A natural classification aims to 

 assemble all known forms into groups, and these into more 

 comprehensive groups, according to their degrees of likeness in 

 all characters so that each group shall stand next to the one 

 which it resembles closest in its whole structure. Such a 

 classification should be an epitome of our knowledge of the 

 plants under consideration. With this definition in mind it 

 must occur to all that the ideal horticultural classification must 

 be an extension of the now universally accepted botanical 

 classification. Wote should be made of the fact and value of 

 indefinite extension in a natural system of classification. In 

 pomology, and horticulture, the number of forms to be arranged 

 is constantly increasing, and new degrees of relationship are 

 . always coming in view and thus extension becomes a prime 

 requisite. 



It must be borne in mind that every natural classification 

 must be artificial to some degree; nor is it thus possible to 

 express all of the relationships of forms even if they were known 

 and had been placed in lineal order. The task is, to make the 

 classification as natural as possible. 



PoMOLOGicAL CLASSIFICATION. — It must be said that so far 

 as pomology is concerned not inuch has been done, though little 



