1869. | CLASSIFICATION OF THE CARNIVORA. 35 
I think moreover that there is not the slightest question that 
their cranial characters indicate most strongly their approximation 
to the Arctoid type, as has often been noticed before on other 
grounds*, Indeed their skulls seem to be simply a further modi- 
fication of this type, showing resemblances to the true Bears on 
the one hand, and the Otters on the other ; but I hope to take some 
other opportunity of examining more fully into these relationships. , 
The presence of a ceecum in this group is a cireumstance not easy to 
be accounted for. 
Conclusion—Mr. Turner was strongly impressed with certain re- 
semblances, which appear to me rather superficial or accidental, be- 
tween the Ichneumons and the Weasels; and it was in order to bring 
these groups in juxtaposition, in his synoptical table of the Carni- 
vora, that he commenced with the Bears and ended with the Dogs, 
placing the Felide in the central position; in this arrangement I 
cannot, as I have said before, concur. The Dogs, for reasons given 
above, should be placed in the central position, while the Ailuroids 
occupy one flank and the Arctoids the other. 
Of the former, the Felide are perhaps the most specialized, and 
the Hyenide the least so. The Viverride are closely connected with 
the Felide on the one hand, especially by the intervention of Cryp- 
toprocta, and, though less closely, with the Hyenide on the other, 
the gap being partially closed by the annectent Proteles. The }7- 
verride show a great tendency to break into two groups, of which 
Viverra, Paradoxurus, Arctictis, Cynogale, and Genetta belong to 
one, and Herpestes and its various modifications to the other, Rhy- 
zena being an aberrant member of the last. The distinction between 
the five families of #luroidea is founded mainly on the characters of 
the teeth, too well known to need recapitulation here ; but, as shown 
above, the cranial characters alone would suffice to distinguish them. 
Africa and Southern Asia are the head quarters of the group, all the 
families being restricted absolutely or very nearly (two of the Viver- 
ride alone passing into Southern Europe) to these regions, except 
the Felide, which are almost cosmopolitan. 
The Cynoidea admit of no subdivision into families ; and, although 
there is a considerable tendency to variation in external characters, 
they are remarkably “ true”’ in cranial conformation. They are per- 
haps the most universally diffused of any of the groups. 
Of the Arctoidea the true Bears are the most specialized or aber- 
rant ; they form a very compact group, distinguished by their very 
characteristic dentition and their completely plantigrade mode of 
progression. They have a very wide geographical range. On the 
other hand the Procyonide, though few in numbers and restricted 
to the warmer and temperate parts of the American continent, are 
structurally less closely connected, at least if the singular Cercoleptes 
is truly a member of this group. Except for the increased number 
of the molar teeth, which is the only definite character by which 
they can be separated from the Mustelide, I see no reason for con- 
sidering the Procyonide more nearly allied to the Urside than are 
* De Blainvyille says “ Tes Ours, dont les rapports avec les Phoques ont été 
sentis de tout temps et méme par Aristote”’ (op. cif. tome ii. p. 49). 
