Prof. W. B. Scott — Osteology of Ancodus {Ryopotamus). 493 



In 1884 I showed that Oreodon possessed a pollex, the first 

 artiodactyl in which this structure had been observed ; later I 

 found it in the Uinta (Upper Eocene) form Protoreodon. Ancodus 

 is thus the third genus known to possess it, and is doubtless an 

 indication that the Eocene artiodactyls will be generally found to 

 have five digits in the manus. In this genus the pollex is larger 

 relatively than in either of the others, and must have been of some 

 functional value. 



The femur and tibia differ from those of Oreodon only in their 

 much greater size and massiveness ; they are longer and heavier 

 in proportion to the fore-limb than in that genus, but the shape 

 and character of these bones are similar. The fibula is complete, 

 but the shaft is much reduced and the distal portion is shifted more 

 beneath the tibia than is common among artiodactyls of such 

 antiquity. 



The tarsus is oreodont in type, with some curious specializations, 

 which are more elaborated in the American species than in those 

 figured by Kowalevsky and Filhol. The astragalus and calcaneum 

 are interlocked in a remarkably complex and perfect way, which, . 

 however, can hardly be rendered intelligible without the aid of 

 figures. As in the oreodont, and indeed most artiodactyls, the ecto- 

 and mesocuneiforms are ankylosed into a single piece. The ento- 

 cuneiform is very unusually large and heavy, and to its distal end 

 is attached a good-sized nodule representing a rudiment of the first 

 metatarsal. The artiodactyl hallux was first demonstrated by 

 Osborn and Wortman, who found it in the foot upon which they 

 established the genus Artionyx, but which I have shown to be 

 referable to Agriochoerus, Leidy, a most remarkable artiodactyl 

 with claws instead of hoofs. Marsh has lately published the state- 

 ment that a rudiment of the hallux is preserved in Protoreodon 

 (Fjomeryx). 



The phalanges of Ancodus hracJiyrJiynchiis, and especially the 

 unguals, are quite different from those of the European species and 

 approximate more to the oreodont type. 



The general resemblance of Oreodon and Ancodus has been com- 

 mented on by many writers, but we are now for the first time in a 

 position to determine just what that resemblance is, and we find it to 

 extend to all parts of the structure. When the Upper Eocene genus 

 Protoreodon is taken into account the similarity is found to be all 

 the greater, for the latter has the oreodont peculiarities less markedly 

 developed and still retains the pentaselenodont type of dentition. 

 Agriochoerus is also clearly related to the oreodonts, but represents 

 a very aberrant line. We may infer with considerable confidence 

 that Oreodon, Ancodus, and Agriochoerus belong to three divergent 

 branches of the same stock. It is true that the former is an ex- 

 clusively American genus, while Aricodns very probably originated 

 in the Old World, reaching North America by migration in early 

 Oligocene times. Doubtless, however, we shall eventually find a 

 Middle Eocene form, common to both hemispheres, which will prove 

 to be the ancestral type required. 



