MARSUPIAL QUADRUPEDS OF NEW HOLLAND. 261 



however, I consider the generic distribution of marsupial ani- 

 mals as approaching more nearly to a perfect and natural 

 division, than that of almost any other order of mammals, as 

 at present constituted. Indeed I know of none against which 

 fewer reasonable or valid objections can be urged, upon this 

 head; but whilst I concede this just praise to the generic 

 distribution of the marsupials, I am bound, at the same time, 

 to acknowledge, that I know of no other order of which the 

 families, or intermediate groups, are at once so arbitrary, so 

 illogical, and so entirely inconsistent both with the structure 

 and economy of the animals. Yet this higher step in gene- 

 ralization is a most important link in that chain of affinities 

 which constitutes a good classification ; since upon it depend 

 not only the scientific character of the system, but even its 

 practical value, so far as relates to the knowledge of the su- 

 perior group, and the nature of its composition ; and it be- 

 comes, therefore, a matter of some consequence, to point out 

 the errors of the present division, and to endeavour, if possi- 

 ble, to substitute one less objectionable to logical criticism, 

 and more conformable to the natural affinities of the animals. 

 Baron Cuvier, and after him M. Desmarest, divide the order 

 Marsvpialia into six families ; to which, however, they have 

 given no distinctive names ; probably because the illustrious 

 naturalist first named, considered the primary group, at the 

 period of publishing the first edition of his ' Regne Animal,' 

 only as a subordinate family of his order Carnivora ; and it 

 is well known that he was not in the bad habit, which has 

 since become so prevalent among a very inferior class of imi- 

 tators, of burdening the memory by assigning useless and 

 high-sounding appellations to the mere subdivisions of fami- 

 lies and genera. M. Desmarest, who scrupulously follows the 

 footsteps of Baron Cuvier in every department of Mammalo- 

 gy, has not departed from his guide in the present instance ; 

 nor does the second edition of the ' Regne Animal ' make any 

 change in this respect, though the primary group is there ele- 

 vated to the rank of an order, and the subordinate divisions 

 consequently assume the station of families ; groups which it 

 was M. Cuvier' s invariable practice to distinguish by appro- 

 priate names. This, however, had been done some years pre- 

 viously by the late M. Latreille, who, in his ' Families Natu- 

 relles,' published in the year 1825, had moreover introduced 

 some modifications into the divisions of Baron Cuvier, which 

 have but tended to make it still more confused and unnatural 

 than it was before ; and only show how imperfectly this emi- 

 nent entomologist was acquainted with the habits and structure 



2e3 



