by rudiments at the upper and at the lower end only; they are deficient in the middle. The middle large metacarpals, though closely applied, are not fused. The metatarsals, on the other hand, are, or are not fused, according to the species. A later form is the genus Leptomeryx from the Miocene of North America. This genus departs from the typical Traguline structure in more than one point. The tympanic bulla is hollow instead of being filled with cancellated bone; the cuneiform is not fused with the cuboid and navicular, though the latter are with each other; the lateral digits of the hind-feet are rudimentary. The magnum and trapezoid, however, are fused. In the fore-feet the middle metacarpals are separate, and the lateral less perfect metacarpals have toes. The metatarsals are fused.
Not definitely referable to the Tragulidae, but coming near to them, are the Protoceratidae. Of this family there is but one well-known genus, Protoceras,[1] from the Miocene of North America.
The skull is singularly reminiscent of Dinoceras, with which this quite Artiodactyle genus has, of course, nothing to do. It merely exemplifies the phenomenon of "parallelism." In general form it is peculiarly long and low. There are three pairs of bony protuberances: one, the largest, pair are situated on the maxillae rising up just behind the implantation of the canine teeth; the parietals have a second pair; and a third much more diminutive pair of bosses are upon the frontals, near their junction with the nasals. This description refers to the male; the female has only traces of the parietal bosses. These were all possibly tipped or sheathed with horn or roughened skin. The dentition of this genus is precisely that of the Tragulidae, i.e. I 0/3 C 1/1 Pm 4/4 M 3/3. The orbit is completely encircled by bone; the auditory bulla is not swollen; the premaxillae are small.
The nasal cavity is very large and open, the end of the nasal bones anteriorly being situated at about the middle of the skull; this would seem to indicate at least a flexible and long nose like that of the Saiga Antelope, if not a trunk.
The brain was of good size, and quite well convoluted.
The limbs are constituted on the Traguline plan; in the fore-limbs the middle metacarpals are quite free from each other, and the more diminutive lateral digits are complete. The meta-
- ↑ Marsh, Amer. Journ. Sci. 1897, p. 165.
