36 MR. W. H. FLOWER ON THE (Jan. 14, 
the other families of. the group, or of speaking of them as specially 
** subursine.”’ 
Ailurus (an unfortunate name for an animal so essentially Arctoid) 
appears to me to be an isolated form ; but until more is known of its 
anatomy, a very definite position cannot be assigned to it. Its den- 
tition, though remarkably modified in character, is numerically that 
_ of the Procyonide ; but certain cranial peculiarities already pointed 
out, and its Asiatic habitat, lead me to concur with Mr. Turner in 
placing it in a distinct family. 
The Mustelide constitute a large, widely diffused, and somewhat 
disjointed group, but exceedingly difficult to reduce into natural sub- 
fainilies. The most aberrant or specialized are the Otters, which, end- 
ing with Enhydris, run parallel to the Bears towards the Pinnipedia. 
In order to exhibit at one glance the general result of this exami- 
nation, I have arranged the various groups of the Carnivora in a dia- 
grammatic plan (see p. 37), which has obviously a great advantage 
over a linear series 1n showing cross relationships, especially as it 
attempts to indicate, by the distances the groups are placed apart, 
the amount of affinity between them*. 
A tree has long been a favourite image by which to illustrate 
genealogical descent ; and we are generally accustomed to have side- 
views of such trees presented to us, with the stem, main boughs, and 
smaller branches growing from them, all in perfect order. But the 
ancestral records of our existing fauna are so imperfect that it would 
be hopeless, from our scattered fragments of knowledge of them, to 
attempt at present to construct such a view of the descent of any 
zoological group. What we may, however, do with tolerable cer- 
tainty is to take a careful survey of the top of the tree (to keep up 
the simile) as far as it has now grown, the only part that is completely 
exposed to our view, and make out the mode in which the compo- 
nent branches are now arranged. The diagram is intended to indi- 
cate the general outline of what may be called the Carnivora tree at 
its present stage of growth, seen from above; or it may be said to 
represent a transverse section of all the diverging genetic lines, showing 
the amount of differentiation of the groups, and the directions they 
have respectively taken at this particular epoch. Similar sections, 
taken at different geological periods, would probably present very dif- 
ferent appearances. Groups now sharply separated might in other 
times have been united by intermediate forms; and other highly spe- 
cialized groups would be seen which have now entirely disappeared. 
The value of this plan all depends upon whether that great zoolo- 
gical problem, interpretation of true affinity, has been rightly solved. 
It is probable that a longer and more minute study of the details of 
the organization of different members of the order than has yet 
been given will introduce many modifications in this rough sketch ; 
il is not too presumptuous, however, to hope that ultimately it may 
be so perfected that every genus and even species will have its ap- 
propriate place assigned to it. 
* Prof. Milne-Edwards in 1844 (Annales des Sciences Naturelles), and on se- 
veral subsequent occasions, has made use of similar plans to illustrate his views 
of classification. 
