Peterson : Mountkd Skki.kton ov Strnomvlus uncHcocKi. 271 



much in the same proportion but shorter than in the recent form. 

 This indicates that the hea\\' muscles of the shoulder and thigh in 

 Stenomylus did not extend as h)\v down from the body, thus giving the 

 limbs less resistance and conseciucntly much freer action when in 

 motion. 



Measirements. Cm. 

 Length of skeleton, .skull to imkI of isrhiuiii aloiii; ciir\es of vertebral 



column 114 



Length of skull 20 



Length of cervical region 34 



Length of dorsal region 25 



Length of lumbar region 18 



Length of sacrum 7 



Height of skeleton at si.xth dorsal 70 



Height of skeleton at anterior part of pelvis 66 



The writer is pleased to see that his original views in placing Steno- 

 mylus in a separate phylum of the Camelidcc is accepted. I do not, 

 however, think that it is necessarily a direct derivative of Poebrothe- 

 rium as Loomis thinks. On the other hand it is more probably an 

 offshoot of an earlier Tertiary form — a heterochthonic rather than an 

 autochthonic type. Thus the peculiar structure of the palate at the 

 sphenoid bones, the backward sloped coronoid process of the lower 

 jaw, caused by the higher position of the orbit, and also the develop- 

 ment and position of the dentition of Stenomylus are such deep-seated 

 characters and so different from those in Poebrotherium that one must 

 hesitate before seriously regarding the former genus as directly derived 

 from the latter. It requires no stretch of the imagination to regard 

 some Eocene form such as Protylopus, or one cotemporaneous with it, 

 as a possible progenitor. The transition from Poebrotherium to other 

 Miocene forms and to the recent Tylopoda is within a reasonable 

 probabilit}', while such rapid modification, as would be required from 

 the middle Oligocene to the superimposed beds of this locality regarded 

 as lower Miocene, in order to transform Poebrotherium into a form like 

 Stenomylns seems, to say the least, to be most anomalous. 



In confining our attention to the region of the palatines, the vomer, 

 the pterygoids, and the presphenoid of the cranium, it is obvious that 

 the variation in form between the two genera in question is quite out 

 of the ordinary. In Stenomylus the posterior palatine processes, the 

 posterior portion of the vomer, and the presphenoid unite to form a 



