THE OSTEOLOGY OF ELOTHERIUM. 303 



facets on both bones are more or less prominent, they are elsewhere separated by con- 

 siderable interspaces. The distal side of the scaphoid is much narrower than the 

 proximal and is occupied by facets for the trapezoid and magnum, no articular surface 

 for the trapezium being apparent. The trapezoidal facet is considerably the smaller of 

 the two, and is simply concave. The magnum facet is in two parts, a very slightly 

 concave distal portion, and a somewhat smaller lateral portion on -the ulnar face of the 

 scaphoid. 



In the European species figured by Kowalevsky ('76, Taf. XXVI ) the scaphoid i> 

 somewhat broader than in the American forms. In both groups a remarkable resem- 

 blance to the scaphoid of Anthraeotherium is observable, which extends to even the details 

 of structure (see Kowalevsky, '73, Taf. XI, Fig. 38). As Anthraeotherium is, however, a 

 tetradactyl form, the scaphoid is somewhat broader in proportion to its height than that 

 of Elotherium, though hardly so much so as would be expected. In Hippopotamus and 

 Sits the scaphoid is of quite a different shape from that of the fossils, being distinctly 

 shorter and wider. 



The lunar is a very large and complex carpal, which exceeds the scaphoid in all of 

 its dimensions, and especiall} r in breadth. The radial facet is in two parts, continuing 

 those which occur on the scaphoid ; the anterior or dorsal part extends across the width 

 of the bone and is very convex antero-posteriorly, while the palmar portion is very much 

 larger and is concave in the same direction. The dorsal border rises steeply toward the 

 ulnar side, where the lunar is drawn out into a blunt, projecting, hook-like process, which 

 extends over the pyramidal, as the scaphoid does over the lunar. On the radial side are 

 three facets for the scaphoid, corresponding to those on the latter, which have already 

 been described. The palmar face is greatly extended transversely, and, though lower, is 

 much broader than the dorsal surface. On the ulnar side are two facets for the 

 pyramidal, which constitute an interlocking joint of unusual firmness and strength. One 

 of these facets is proximal and dorsal and overlaps the pyramidal ; the second, which is 

 very much larger, is palmar and distal in position, and has a saddle-like shape ; it interlocks 

 closely with a similar facet upon the pyramidal. When seen from the front, the contact 

 between the lunar and the magnum appears to be entirely lateral, but as it jiasses toward 

 the palmar side, the magnum facet broadens, becomes veiy concave, and assumes a distal 

 position. The unciform facet is also oblique and the beak between the two is not in the 

 median, but shifted far toward the radial side. Dorsally the unciform facet is consider- 

 ably wider than that for the magnum, but on the palmar side these proportions are 

 reversed. 



The lunar of E. magnum figured by Kowalevsky resembles that of E. ingens, except 

 that its proximal surface does not rise so steeply toward the ulnar side ami does not 



A. P. S. — VOL. XIX. 2 M. 



