ORDER CARNASSIER. 57 



must either exclude from the order amphibia animals that 

 are amphibious, or arrange animals together that differ in 

 the most important and essential points. 



With the Marsupial tribes, indeed, the case is somewhat 

 different. They differ materially, not only from the fore- 

 going families, but from all the rest of the animal world. 

 The pouch in most, the Marsupial bones, probably in all *, 

 and the double matrix, are high leading characters. But yet 

 these animals exhibit such very various modifications, that, 

 as the Baron observes, if we are to separate them, it should 

 be into a distinct class, rather than an order. They furnish 

 in themselves analogous modifications to all or most of the 

 other orders of Mammalia. But before any such arrange- 

 ment can take place, or any other arrangement indeed, that 

 shall pretend to permanence and precision, it will be neces- 

 sary that our knowledge of these animals be considerably 

 enlarged. At present, our acquaintance with them is far 

 too imperfect for this purpose. 



Let travellers and naturalists, then, proceed in the grand 

 business of discovery and investigation, and let us mean- 

 while content ourselves with a classification, which, under 

 existing circumstances, is the least objectionable that can be 

 given. After all, as we have had already frequent occasion 

 to remark, we must recollect that classification is the 

 work of man, not of nature; that general systems are entirely 

 arbitrary, and that however we may pique ourselves in our 

 ingenuity in their formation, they involve a confession of 

 the weakness and imperfection of the human faculties. 

 Classification is intended to assist our apprehension and 

 our memory, which would otherwise be puzzled and over- 

 whelmed by the infinite variety and stupendous extent of 

 the visible creation. The little eye of human intellect 



* It does not appear that the Dog- of New Holland, the Dingo, has 

 ever heen examined with reference to the Marsupial bones more or less 

 developed. 



Vol. III. P 



