15 



^\Mdtll of premolar 4. greatest 0133 



Leugtli of true molar series 066 



Length of molar 1 0208 



Width of molar 1 018 



Length of molar 2 025 



AVidth of molar 2, greatest; antiTior 0215 



Width of molar 3, posterior ' 017 



Length of molar o 021 



Width of molar 3, auterioi- 022 



Width of molar 3, posterior ok;.') 



PROCAMELUS MADISONIUS. N. SP. 



Plate II. 



This species is represented by a nearly complete cranium found by 

 Mr. Dunbar of Gallatin City several years ago, at the foot of the bluffs 

 bordering the Madison valley. The molars on one side had been broken 

 away. I purchased the specimen and on clearing away the hard ma- 

 trix found most of the cheek teeth on the other side in a good state 

 of preservation. The skull is considerably broken and crushed an- 

 teriorly, and the roof of the brain case including the sagitfcil crest 

 is gone. 



The skull differs in many respects from that of F. ocddentalis 

 Leidy, found by Cope in the Loup Fork beds of New Mexico. 



The skull is about one-fifth larger than that of Camelus dromedar- 

 ius and is proportionally broader. The width between the second pre- 

 molar and third incisor is more nearly uniform; but the palate widens 

 much back of premolar 4. The malar ridge of the maxillary is almost 

 obsolete; being represented by only a slight narrow prominence at the 

 anterior-inferior border of the molar. The posterior part of the malar 

 rises to the middle of the orbit; the anterior part does not rise 

 as high as the middle of the orbit— only about one-third the 

 height. The zygomatic process of the squamosal is higher relatively 

 to the orbit than in Cope's specimen or in the modern camel and is 

 arched upward being inserted obliquely into the notch of the malar 

 just back of the lower half of the orbit. Though the skull is broken 

 in this region I judge that the lachrymal sinus was small. 



The anterior-inferior border of the basioccipital is not angulate. 

 but is smoothly rounded, as is also the basisphenoid. In front of the 

 occipital condyle there is a quite high, broadly sloping, transverse 

 ridge, divided in the middle by an oblong fossa. Posteriorly this fossa 

 is continued in the form of a suture which is evidently the division 

 line between the iinpcrfectly coossified occipitals. Above the foramen 



