82 
to the Cats; in all the characters common to the Cats and Civets 
the Hyznas also agree. However, if these three groups be united, 
as I think would be justified by the characters, the difficulty in a 
great measure will be overcome, ‘Three subfamilies will then con- 
stitute the family Felide ; the Felina are characterized by the absence 
of the ali-sphenoid canal, the very minute size of the canalis caro- 
ticus, the absence of the foramen glenoideum, the auditory bulla 
having but little or no trace of division, and the molar series consisting 
of only three premolars and one true molar on each side above, and 
two premolars, with one true molar, below. The next subfamily, 
Hyenina, presents, as I have just before pointed out, the same cranial 
characters as the Cats, but it differs, as is well known, in the den- 
tition. The largest subfamily of the Felide, the Viverrina, possess 
in general the ali-sphenoid canal, and sometimes a minute foramen 
glenoideum ; the auditory bulla is more or less distinctly divided 
into two portions; the canalis caroticus is of average dimensions, 
although not always completely enclosed, and is placed towards the 
anterior part of the bulla, and the artery, after having passed through 
the canal, shows itself externally before finally entering the cranium : 
the dentition is on each side usually four premolars and two true 
molars, both above and below. The very singular genus Prote/es has 
the cranial characters common to the Cats and Hyznas: from the 
dentition, so singularly modified by arrest of development, but little 
evidence of zoological affinity can be adduced; I should therefore be 
rather inclined to consider it a modified Hyzena, since in the external 
characters it so closely resembles the animals of that genus. 
The Dogs, or the subfamily Canina of Mr. Gray, remain, and I 
think must constitute a separate family Canid@ ; they all agree pre- 
cisely in those cranial characters which I have pointed out, and not- 
withstanding the variation in the number of the true molars, do not 
seem to admit of division into subfamilies. It is rather remarkable 
that in the different genera of this family we find precisely the same 
variation in the number of the true molars as in the subfamilies of 
the Urside ; as, setting aside the genus Ofocyon, in which they are 
developed, we may almost say, beyond their normal number, there 
are two above and three below in the genera Canis, Vulpes, Nycte- 
reutes, and Lycaon, as in the Ursina, two above and two below in the 
genus Cyon, as in the subfamilies Ailurina and Procyonina, and 
only one above and two below in Cynalicus, which is the number found 
in the Musteline group. This being the case, on first looking at the 
imperfect skull of the Cynalicus in the British Museum, which un- 
fortunately exhibits only the anterior portion, I was led for a moment 
to suspect that the true place of this remarkable genus might be 
among the Weasels, and was the more led to this idea from the cir- 
cumstance that the animal, in proportions and in colouring, bears a 
remarkable resemblance to the Galera barbara, also an inhabitant of 
South America ; but on more attentive examination I perceived, that 
not only did the last molar resemble precisely the penultimate in the 
Dog, but sufficient remained of the skull to show, in the form of the 
pterygoid appendage, and the presence of the ali-sphenoid canal, 
. 
. 
: 
