HAPLODONTIDAI—DISCUSSION OF AFFINITIES. 553 
coming between Castoride and Sciuride, with close relationship on the one 
hand with Castor, and on the other with Arctomys. 
My views of the position in the Rodent series, and relative raik in the 
scale, of the Haplodont type rest upon an examination of the whole structure 
of the animal. I do not find that any one has hitherto examined—at any rate, 
given an account of—the viscera, or even the skeleton, excepting the skull and 
leg-bones; our knowledge having thus far rested upon these portions of the 
bony frame-work, the teeth, and the external characters. Jam, consequently, 
enabled to add many new particulars to substantiate the position here taken. 
Detailed descriptions are offered beyond under head of the species; here I 
shall simply advert to some of the leading points involved. 
The skull of Haplodontide is strongly and unmistakably Sciuromorphic, 
not only in its general structure, but in many ultimate details. In fact, it 
resembles in superficial aspect the skull of certain typical Sciurines more 
closely than some of these resemble each other. For example, no one who 
compares the skull of Haplodon with that of Arctomys can fail to be struck, 
as Dr. Peters was, with their close general resemblance. In comparison with 
Sciurus, or even with such a Marmot-like form as Cynomys, the skull of Arc- 
tomys is seen to be much more massive, much more depressed, broader behind, 
and with a straightness and mutual perpendicularity of various planes, all of 
which features would require little exaggeration to match those of Haplodon. 
Even the shape of the angle of the mandible, peculiar to Hap/lodon, is really 
approached in Arctomys, where further twisting of the already oblique plate 
of the descending ramus would bring it into the nearly horizontal plane 
which it occupies in Haplodon. The most prominent difference is the total 
absence of postorbital processes in Haplodon, and their full development in 
Arctomys as in other Sciuride. It may be fairly questioned, however, whether 
the presence of these processes is more than a character of the family Sczu- 
ride itself; for they are lacking* in the three other families referable to 
Scturomorpha. The preponderance of recent genera and species of Sciuride 
may have unconsciously led us to attach too great importance to this feature. 
Obviously, the fact that the family Sciuride at present contains many genera, 
while the other three Sciuromorph families have but one genus apiece, is no 
argument for the making of postorbital processes a requisite for any series of 
Sciurines of higher value than a family. The argument is the other way, in 
* Mr. Alsten calls them ‘ obsolete” in dnomalurida. 
