GENERAL COI^^SIDERATIONS UPON CLASSIFICATION. 83 



an idea of its diflFerences and community with other objects of the same 

 category, so far as those differences and community are understood. 



The differences and affinities (that is to say, community of attributes 

 and relations) between the members of a category are ordinarily not few, 

 much less single, but numerous and complex ; and the value and utility 

 of a system of classification is about proportional to the number of differ- 

 ences and affinities which it truthfully expresses. Systems of classification 

 are spoken of as "artificial" and "natural." My understanding is that an 

 artificial system is one which takes account of the agreements and disagree- 

 ments of the classified objects with respect to only one characteristic or 

 one very limited set of characteristics. The meaning of the expression 

 "natural system of classification" is much more difficult to assign. Most 

 probably different authors would entertain widely differing conceptions as 

 to its meaning, none of which would be very definite or precise. They 

 might, however, agree that a natural system as contradistinguished from an 

 artificial one takes cognizance of all the characteristics and relations of the 

 members to each other ; the difference and affinity in any case being rated 

 and valued, therefore, in accordance with the totality of characters and not 

 dependent upon merely one of them. But it is far easier to say this much 

 about a system of classification than it is to comprehend it! The truth is, 

 that a natural system in any such length and breadth is impossible for any 

 category, unless we know all the members of it and the totality of their 

 relations ; and there is no reason to believe that human knowledge has ever 

 reached to that perfection. But as knowledge is ever increasing, we may 

 at least hope for the time when it shall be sufficient to enable us to find 

 and designate the greater and more important relations with absolute verity; 

 and if the systema naturce is fitted and keyed together in order and harmony, 

 as we are fain to believe, the outstanding facts will fall readily into their 

 places ; just as the final parts of a puzzle are quickly placed when the true 

 arrangement of the other parts is discovered. A purely artificial system 

 marks the initial stage of generalization of knowledge ; a perfect natural 

 system is for the time being unattainable. The growth of knowledge and 

 philosophy, however, is marked by a transition, long, laborious and very 

 gradual, from one to the other ; a transition, which is marked by an indefi- 



