0. C. Marsh — Discovery of Cretaceous Mammalia. 257 



first, or inner one, being much the largest. A very stout canine 

 (fig. a, c) was present, and behind it, in close contact, follow 

 two premolars, the second being the larger. These are both 

 much worn, but the surface of the crowns preserved is coarsely 

 striate, or rugose, thus permitting a comparison with various 

 isolated teeth of similar size and structure found in the same 

 deposits. Two of these teeth are shown in the same plate, 

 figures 4: and 5, while a separate canine referred to this species 

 is represented in figure 6. 



These specimens taken together indicate a carnivorous 

 animal, which apparently has its nearest living ally in the 

 genns Sarcophilus, the only representative of which is the 

 " Tasmanian devil " (S. ursinus, Cuvier). The resemblance of 

 the present type specimen to the corresponding part of this 

 living marsupial is close. 



The teeth of the smaller species of Stagodon shown on the 

 same plate appear to be all upper premolars, and each has two 

 roots. The crowns of those best preserved show indications of 

 the same rugose markings seen on the larger lower teeth, and 

 all other features prove their close affinity with them.* 



The Cimolestidw. 



The remaining specimens figured on the accompanying 

 plates (IX-XI) include various fossils belonging to the fam- 

 ily Ctmolestidce, which appear to be certainly marsupials more 

 or less related to the modern opossums. Beside these, how- 

 ever, there are quite a number of remains which show features 

 of the Insectivora, but until more perfect specimens are 

 obtained, their true affinities must remain in doubt. Under 

 these circumstances, some of these fossils have been referred 

 to forms previously made known in Parts I and II, and only 

 the most distinct and best preserved specimens are described 

 as new. It is believed that the material already secured, when 

 properly worked out of the matrix and investigated, will 

 remove most of the uncertainty now remaining in regard to 

 several typical forms already in part made known. 



The larger forms most nearly' allied to the genus Didelphys 

 are not abundant, but their size has kept some of the remains 

 in good preservation. There are several distinct species of 

 these, but at present they are known only from imperfect 

 material, mostly isolated teeth. 



* Prof. Osborn in the paper already quoted lias questioned the mammalian 

 nature of the type of this genus, without giving any evidence for such an 

 opinion. In a later paper (Am. Nat., p. 780, 1891), he has even placed in the 

 Stagodontidce, as my types, figures of quite different fossils. No reply is necessary, 

 as the specimens shown on Plate VIII tell their own story to every anatomist. 



