Xvi PRELIMINARY SKETCH OF THE 



arrange all known animals under certain divi- 

 sions and sub-divisions, according to their de- 

 grees of affinity or resemblance. The more 

 that such arrangement approaches to the or- 

 der observed by nature herself, the better. In 

 other words, the more that systems are con- 

 structed on the similarity, not only of external, 

 but of internal conformation the better. Yet, 

 perhaps, too much refinement in this way, is 

 sometimes insisted on. It should never be for- 

 gotten, that the only, or at all events, by far 

 the principal use of systems is to assist the 

 memory. They are, after all, in a great mea- 

 sure, the arbitrary creations of man, and their 

 object of facilitating retention will not be 

 gained, if simplicity be much lost sight of. 

 The undue multiplication of divisions, must, 

 therefore, prove of most especial detriment to 

 the progress of science. Much has been said 

 about natural methods ; but it may be ques- 

 tioned whether an artificial method, constructed 

 on simple principles, and not deviating too 

 widely from the general order observable in 

 nature, be not better adapted to answer the 

 purposes of scientific arrangement, than a more 

 complicated, though, perhaps, a more strictly 

 philosophical system. But when we come to 

 examine the methods of celebrated Naturalists, 



