85 
process, but that on the anterior rim of the orbit is very strongly 
developed; there is also an ali-sphenoid canal whose outer wall is 
very thick; the mastoid is a large thick process, projecting very 
much as in Arctocephalus, but its surface is for the most part con- 
tinuous with that of the flattened auditory bulla. 
If we adopt all the five subfamilies into which this family is 
divided in the list published by Mr. Gray, the Walrus, and the 
Arctocephaline group, which differ so decidedly from the other 
Seals, would almost seem entitled to the rank of families; but rather 
than so completely dismember such a well-marked group as that of 
the Seals, I should feel disposed to assign to the differences of the 
teeth no more than a generic value, and to restrict the number of 
subfamilies to three,—the Phocina, Trichecina, and Arctocephalina, 
including in the latter the genera Osaria and Arctocephalus, the 
Walrus alone constituting the Trichecina, and all the other Seals 
falling under the first-named section. 
I cannot conclude without offering some apology for the length 
of my communication, more particularly as the number of minute 
details of form alluded to may render it a little tedious, and among 
the facts enumerated the number is so small that possess any claim 
to be considered new; but if I have to any extent succeeded in 
placing in a clearer light the mutual affinities of the different 
genera of Carnivora, a subject of which I think all will admit the 
difficulty, or if I have but brought into its due importance any 
character, however small, which may render the determination of a 
fragment more easy to the palzontologist,—if I have achieved but 
a very small share in the important task of elucidating those real 
affinities existing throughout nature, which must, when completely 
made out, render classification not a mere alphabet of reference for 
the determination of species, but a key to higher generalizations, I 
trust that my labours have not been thrown away, and that my ap- 
parent prolixity may be overlooked. 
In offering the annexed synopsis with a view to render the ar- 
rangement I would propose more readily comprehensible, I must 
observe, first, that the lists of genera include only those whose crania 
I have examined, and therefore I must not be considered as rejecting 
any that I have omitted, nor doI pledge myself to adopt all that 
are inserted. Secondly, that the difficulty of expressing in a man- 
ner sufficiently decided, and at the same time sufficiently brief for a 
synoptical form, the characters I have made use of, has compelled 
me to omit some of them. In order to place the Herpestine genera 
of the Viverrine subfamily in juxtaposition with the Weasel group, 
it is advisable that the series of terrestrial Carnivora should either 
commence with the Bears and terminate with the Dogs, or vice 
versd; and as I have not seen in the Seals anything which, in my 
opinion, warrants their approximation to any of the other families 
more than to another, it matters little which mode be followed. 
