THE MAMMALIA OF THE DEEP RIVER BEDS. 59 



in the typical Oregon localities, and while, owing to the very small area of the Mon- 

 tana beds, we cannot lay much stress upon the absence of certain characteristic John 

 Day forms, yet the presence of such relatively modernized genera as Cynodesmus 

 and Mesoreodon indicates that these beds should be referred to the summit of the 

 John Day formation. This is of interest as being the first identification of this hori- 

 zon east of the Rocky mountains. 



The upper beds, which Grinnell and Dana called Pliocene, present a very differ- 

 ent assemblage of species. Cope's collection, so far as I can judge, was gathered 

 entirely from these beds and contains nothing from the lower horizon. His collection 

 and that made by the Princeton party are, as would naturally be expected, not quite 

 coextensive, each containing some forms which the other does not. Combining the 

 two, we obtain the following list: Canisf anceps Scott, Chalicotherium? sp., Aphe- 

 lops sp., Miohippus sp., AncMthermm equinum Scott, Desmatippus crenidens Scott, 

 Protohippus sejunclus Cope, Protoliippus {Merychippus) insignis Leidy, Merychyus 

 (Ticlioleptus) zygomaticus Cope, M. pariogonus Cope, Merycochoerus montanus 

 Cope, Cyclopidius simus Cope, C. emydinus Cope, C. incisivus Scott, Pitliecistes 

 Irevif acres Cope, P. decedens Cope, P. heterodon Cope, Protolabis sp., Procamelus 

 sp., Blastomeryx lorealis Cope, B. antelopinus Scott, Mastodon proavus Cope. In 

 addition to this list should be mentioned a considerable number of equine animals, 

 which cannot be well identified, as the specimens are scattered vertebrae, limb and 

 foot bones, not accompanied by teeth, but which, from the variations in size and 

 details of construction, point to several species not enumerated above. 



The resemblance of this fauna to that of the Loup Fork has been obvious from 

 the first, for it was doubtless the latter formation to which Grinnell and Dana referred 

 under the name " Pliocene." For the same reason of very limited exposures, as in 

 the case of the lower beds, when compared with the John Day of Oregon, we cannot 

 insist very strongly upon the absence of typical Loup Fork genera from the upper 

 series of Deep Eiver strata. Of much greater significance is the occurrence in the 

 latter of five genera and fourteen species of mammals which have not been found in 

 the vastly more extensive and carefully examined Loup Fork deposits. This fact, 

 having regard to the character of the species involved, points to the conclusion, 

 already drawn by Cope, that these beds are older than the typical Loup Fork hori- 

 zon, but their faunal connection with that horizon is so close that there seems little 

 ground for considering the Deep Eiver as an " epoch " of coordinate rank with the 

 three other Miocene epochs. The relation between the Deep River and Loup Fork 

 beds is more intimate than that between the Wind River and the Bridger proper of 

 the Eocene. My own preference is, therefore, to refer both series to the Loup Fork, 



