508 



is less concave than in Phascolomys, and is much less so than in Macropus. The ulnar 

 half, indeed, of the marginal surface is slightly convex. 



The triangular facet has its outer angle continued as a narrow groove, also as in 

 Phascolomys, dividing the radial (a) from the trapezoidal (e) surfaces. The outer ball 

 {</) for the trapezium is rather larger relatively and more prominent, more hemispheric, 

 than in I'hascolomyz, and a fortiori differs from that in Macropus. 



Thus the scapholunar carpal bone in Diprotodon exhibits, as in some other parts of 

 the skeleton, combined phascolomydian and macropodal characters. It may be said to 

 exhibit a more generalized type. But I am led to remark, considering that the huge 

 quadruped, with a dentition essentially macropodal, required a development of its fore 

 limbs to enable them to take such a share in habitual support and progression as do 

 the fore limbs of the Wombat, the predominance of similarity in certain carpal bones 

 to those in Pliascolomys may not indicate affinity, or derivative relation, to that genus, 

 but may be simply an adaptive character, such as a Kangaroo so huge as to be obliged 

 to walk on all fours, and to part with the power of leaping, may be expected to 

 show. 



§ 2. Metacarpal. — In the bones of the fore foot of the Kangaroo one of the meta- 

 carpals, the fifth (Plate LXX. fig. 10, v), has a process extending outward from its base 

 just beyond the proximal articular surface. Not any of the metacarpals in the Wombat 

 (Plate XCIX. fig. 5) shows an answerable prominence, but it appears in the fifth meta- 

 tarsal (ib. fig. 8). In neither genus does any phalanx of fore or hind foot exhibit such 

 lateral prominence or process. But this is a marked feature in both bones of the fore 

 foot of Diprotodon, one of which (Plate CXXII. figs. 7, 8, 9, & 10) I suspect may 

 prove to be a metacarpal of an aborted fifth toe; the other (Plate CXXIII.) is an 

 ungual phalanx of one of the normally developed toes, probably the second, i i. 



The proximal articular surface (Plate CXXII. fig. 10, a) of the metacarpal is slightly 

 concave transversely, almost flat from before backwards, and was adapted to the outer 

 metacarpal facet of the " os unciforme." That the present metacarpal was a terminal 

 or marginal one of the series is indicated by the absence of any surface for a contiguous 

 metacarpal at the smaller end of the carpal surface. At the greater end there is the 

 narrow fore-and-aft extended surface for articulation with the base of the contiguous 

 '• fourth " metacarpal. 



The fifth (?) metacarpal is rather broader than it is long, it is flattened and even on 

 the dorsal (anconal) surface (fig. 8), is irregular on the palmar surface (fig. 7), and is 

 there deeply grooved between the outer process (c) and the outer condyle (d). The 

 distal articulation is reduced to two distinct and somewhat remote convexities or 

 condyles (b, d, figs. 7 & 9), which I conjecture to have supported a clawless phalanx, 

 the sole rudiment of a fifth digit. 



§ 3. Ungual phalanx. — The proximal surface (a) of the subject of Plate CXXIII. 

 shows a simple uniform concavity, indicative of its articulation with the distal convexity 

 of an antecedent metacarpal ; yet the distal surface (fig. 1, d, and fig. 5) proves the phalanx 



