120 ON SYSTEMATIC ZOOLOGY. 



where the other begins. We make use of an artificial 

 system to become acquainted with the name of a species ; 

 and to learn all that has been written upon its peculiar 

 structure. We turn to the natural system^ to know the 

 probable station of this species in the scale of beings 

 the affinities it possesses to others, and the analogies 

 by which it is related and represented. Hence the per- 

 fection of an artificial system, as we have frequently in- 

 timated, consists in the clearness and precision of its 

 subdivisions, and the facilities which it affi^rds to de- 

 termine the name of the object we are in search of. In 

 this respect, a good artificial system is to be judged by 

 the same rules as those by which we should decide on 

 the merits of a copious index to a voluminous publica- 

 tion, for the purposes of both are the same : both are 

 equally useful, and the merit of both lies in clearly 

 directing the reader to the precise point upon which he 

 desires information. A good artificial system is, there- 

 fore, not only a useful, but even, in some respects, a 

 valuable, invention, requiring^, much more skill than is 

 generally supposed ; and it is, perhaps, much more 

 adapted for general use than any other. The most 

 admirable classification of this sort ever invented, is that 

 denominated the Sexual System of Plants, by Linnaeus. 

 Many natural assemblages are preserved, without any 

 great violation of the principles on which he set out. 

 This is always a great recommendation to an arti- 

 ficial system, yet it is by no means necessary to its 

 formation. Natural affinities may be overlooked, wher- 

 ever they interfere with precision of arrangement : 

 the first are secondary, the latter primary. We open 

 an artificial system to come to the knowledge of a mat- 

 ter of fact ; but if we wish to proceed farther, and to 

 know how this fact bears upon other facts, we turn to 

 the natural system. Such are the uses of the two methods 

 of classification upon which we have been speaking, and 

 such the theoretic distinctions by which they are sepa- 

 rated. Between them, however, is a third sort of 

 system, which, from combining artificial division with 



