482 GEOLOGICAL SURVEY OF THE TERRITOEIES. 



lateral walls of the long nasal fossa, and inclose a large cbamber on 

 each side with the lateral cranial walls. This sinus appears to be sepa- 

 rated by a thin osseous septum from the brain-case proper; the septum 

 extending from above obliquely forward aud downward, altogether 

 behind the arch-like projection of the frontal bone above descriDed. As 

 it is not perforated, it gave no exit for the olfactory nerves. I cannot 

 detect the proper boundary separating the hemispheres from the olfac- 

 tory lobes, as their usual position, both in front of aud behind the frontal 

 arch, is included in the large siuas above described. The nas£! fossa is 

 divided by the usual septum, and each half communicates with the" 

 large fossse above mentioned by a large longitudinal oval foramen. In 

 one specimen of iSymhorodon acer, in which this lateral septum is almost 

 entirely preserved, there is no indication of attachment of turbinal 

 bones, but the surface is smooth. On these and the median septum, the 

 olfactory nerve was, no doubt, distributed. The olfactory lobes were 

 contracted in dimensions, and the large lateral sinuses in front of them 

 are doubtless the frontal sinuses of the mammalian skull. Their length 

 is double that of the brain-case, - and they extend far posterior to the 

 orbits and above and beliind the olfactory lobes. They do not ai)pear 

 to have been divided by septa, excepting a small one springing from the 

 outer wall near the posterior fourth in S. acer, aud near the same place 

 from the inner wall in >S^. trigonoceras.- This huge cavity was doubtless 

 an air-chamber, which gave a lightness to the skull not otherwise 

 attainable consistently with its* great length, aud which has rendered 

 the use of the nasal horns entirely practicable. They explain the elonga- 

 tion, of the skull in the Eocene genera Eobasiieus and Uintatherium, 

 and prove that the sinus at the base of the horns of the middle pair is 

 the frontal sinus aud not the alveolus of the canine tooth, as supposed 

 by Marsh. 



The cervical vertebrae in all the species are concavo-convex, and much 

 deeper than long ; they are longer than in Uobasileus. The odontoid 

 process is a solid cone ; the coracoid process is a tubercle -, there is no 

 acromion ; and the spine of the scapula rises gradually from its base. 

 The ilium is strongly pedunculate. 



In S. hucco, the femur has a third trochanter, and is relatively longer 

 than in Mhinocerus. There is a fossa for the round ligament, and the 

 condyles are expanded. The fibula is enlarged distally, and is distinct. 

 The phalanges, including the ungueal, are very short. The carpals 

 interlock.; and the ulna is much reduced, giving the carpal articulation 

 to the radius. 



The S. acer is the smallest species described, and has the longest 

 Iiorns. Its astragalus resembles that of Ehinocerus, having a deeply- 

 grooved trochlea and well-defined head and neck. The cuboid facet is 

 Tather larger than in that genus, but is considerably smaller than the 

 navicular, aud extends with an acuminate outline behind it, as in Eoha- 

 sileus ; otherwise it has no resemblance to that element in that genus. 

 The metatarsals have much the form of those of Ehinocerus. 



The palate is deeply incised, as in BJdnocerus, and other cranial pecu- 

 liarities resemble those of that genus. In no case are any traces of 

 inferior incisors present in the numerous under -jaws at my disposal. 



In estimating the ordinal affinities of the genus, the greater number 

 at once assign it to the Perissodactyla. The teeth,* the incised palate, 

 the distribution of the cranial foramina, including the perforation of the 

 pterygoid, thepostgleuoid and paramastoid processes, are all characters 



* See the Structure and Homologies of Molar Teeth of Mammalia, by E, D. Cope, 

 -Journ. Acad. Sci., Philad., 1874. 



