THE MAMMALIA OF THE DEEP FJVER BEDS. 57 



often seen covering over ridges of trachyte. The line of separation between the 

 Miocene and Pliocene beds is in some places well marked. It consists of about six 

 feet of hard sands, interstratified with layers of very small water-worn pebbles 

 soldered together into a hard mass, but very easily picked out with a knife. Imme- 

 diately above the strata of pebbles the Pliocene fossils were found. In several places 

 fragments of trachyte were noticed in the Pliocene beds" (No. 18, pp. 126-128). 



The next account of this formation was given by Cope, who had sent his assist- 

 ant, Isaacs, to collect in the valley and described a number of new forms from it. 

 His collection embraced specimens from the upper beds only, those called Pliocene 

 by Grinnell and Dana, and these he referred to the Loup Fork. In 1879 (No. 3), 

 Cope divided the Loup Fork into two horizons, which he called the Ticlioleptus and 

 JProcamelus beds respectively, the former being the beds of the Deep River region. 

 Subsequently (Nos. 6, 8, 10), Cope raised the Ticlioleptus beds to a rank coordinate 

 with that of the John Day or the Loup Fork, and gives the following list of species 

 as occurring in the Montana area : Mastodon proavus, ProtoMppus sejunctus, Merxj- 

 codicerus montanus, M&rychyus zygomaticus, M. pariogonus, Cyclopidim simus, C. 

 emydinus, Pithecistes brevifacies, P. decedens, P. lieterodon, Procamelus vel Protolabis 

 sp. Blastomeryx borealis (No. 8, p. 369). In his latest paper on the subject, this 

 writer defines the formation as follows: " Tioiioleptus. Mammalia. Presence of 

 Anchitherium, Proboscidea and Camelidse and the Oreodont genera Merycochoarus, 

 Merychyus, Cyclopidius and Pithecistes. Absence of ? Elotheriidse, ? Poebrotherii- 

 dse, ? Nimravidae and Cosoryx. This horizon requires further exploration, as but 

 twenty species have been thus far determined from it. But it is evidently interme- 

 diate in age between the John Day and Loup Fork epochs, with greater affinities to 

 the latter. It differs from the latter in the presence of Anchitherium, numerous 

 genera and species of Oreodontidse, and in the absence of Cosoryx. The formation 

 is known from three regions : first, from western Nebraska ; second, from the valley 

 of Deep river, Montana ; and third, from Cottonwood creek, Oregon. Its thickness 

 has not yet been stated " (No. 6, pp. 456, 457). 



It should be noted that in these lists the name Ancliitheriam is used for the John 

 Day eqnines, to which, in this paper, I have applied Marsh's name, Mioliippus, for 

 reasons which will appear later. This point is of importance. 



In 1891, I published a brief note upon the subject of this horizon (No. 31). At 

 that time the fossils collected were still in the matrix, and only the hasty examina- 

 tions in the field Avere available for the purposes of comparison. Consequently, a 

 number of errors crept into the work, so as to greatly vitiate its conclusions, which 

 will not be further referred to here. In a second note (American Naturalist, 1893, 



