78 
since they seem to be merely instances of non-development of the 
bony lamina which should enclose the canal, I think we need not 
deprive ourselves of the assistance the character affords in the discri- 
mination of groups; and further, I believe it will be found that by 
taking the aggregate of the characters I am here attempting to de- 
scribe, the true affinities of any member of the order may without 
much difficulty be ascertained. The foramen glenoideum, when 
existing in this family, is of very small dimensions ; the auditory bulla 
has very distinctly the appearance of being divided into two portions, 
of which the posterior is much the larger, and elongated in form : 
the more anterior division, which encloses the meatus auditorius ex- 
ternus, is much smaller, and partly overlapped by the other. In 
Herpestes and the genera allied to it the separation is rather less 
distinct, and the general form of the bulla partakes a little of that of 
the Weasels. The canalis caroticus is most frequently represented 
simply by a groove in the inner side of the auditory bulla, to give 
protection to the artery before it enters the cranium by the foramen 
lacerum anterius; but in most of the Herpestine genera it is pro- 
tected by a closed canal, as in the Weasels. These genera, however, 
have a slight peculiarity of their own in respect to the entrance of 
the internal carotid artery, and that is, that after emerging from its 
canal it runs exposed for a short distance before finally entering the 
cranium. One of the most striking of the essential characters in this 
family is the structure of the paroccipital process; it is spread out, 
widened, and closely applied to the posterior surface of the auditory 
bulla, and the foramen condyloideum is by this means more or less 
concealed within the aperture of the foramen jugulare: these cha- 
racters are very distinct in the Civets and Paradoxuri; in the Her- 
pestine genera they are manifested a little less in degree, and the 
mastoid process is a little more extended, also spread over the audi- 
tory bulla, and blended with the paroccipital, so that the bony plate 
clothing, as it were, the posterior part of the bulla, has the appear- 
ance of being pushed a little towards the side. 
The characters of the lower jaw, I before remarked, although not 
sufficiently constant in all cases to separate the groups, sometimes 
show interesting marks of affinity. In most of the genera of this 
group the coronoid process curves gently backwards as it rises, which 
is also its character in the Dogs, the Cats, the subursine group, and 
even in a few of the Weasels; but in the Herpestes, of whose ap- 
proximation to the Weasels I have already mentioned so many indi- 
cations, it presents that form of the corunoid process which charac- 
terizes most of the members of that group; and the angular process, 
although it is a salient process, as usual among the Civets, instead 
of being pushed up towards the condyle, yet shows that flatness on 
its lower margin which is more distinctly manifested in the Weasels 
and Bears than in any other sections of the order. The lower out- 
line of the jaw has considerable curvature, both in Viverra and Her- 
pestes. 
Being of opinion that of the two remaining groups, the Cats ap- 
proach the more nearly to the Civets, I will point out their characters 
a 
