SINCLAIR! MARSUPIALIA OF THE SANTA CRUZ BEDS. 355 
its articular surface on the trapezium, it was larger than in Thylacynus. 
The second metacarpal overlaps proximally the third, and the fourth 
on the fifth. The proximal end of the third has not been preserved. The 
fourth is the longest element in the metacarpal series. Its proximal end 
is irregularly quadrangular in outline, wider dorsally than at the palmar 
margin. The head of metacarpal V is separated by a sharp keel into two 
convex facets for the overlapping metacarpal IV and the unciform respec- 
tively. Distally, the metacarpals are flattened, with well-developed keels 
confined to the palmar surfaces. The arrangement of the phalanges 
shown in the figure is arbitrary, as there was no possible way to determine 
the original association of these elements in the matrix. The phalanges 
of the proximal row are much shorter and heavier than in Thylacynus , 
with straighter shafts. Those of the second row are less flattened than in 
the recent genus. The distal trochleae of the phalanges of the first and 
second rows have no greater dorso-palmar extension than in Thylacynus 
and there is no reason to believe that the angulation of the digital elements 
was any greater. The unguals are stout, resembling the claws of Thy- 
lacynus, but differing from that genus in the broad Assuring of the tips. 
The subungual processes are large. Hoods are developed to about the 
same extent as in Thylacynus. An ungual foramen is present in all the 
claws. 
In proportion to the size of the skull, the femur (PI. XLVI, figs. 3, 3 a) 
is remarkably short. The head is of about the same size as in Thylacynus , 
but the neck is considerably longer. The great trochanter projects slightly 
above the level of the head, from which it is widely separated. The lesser 
trochanter is incompletely preserved, but was probably as large propor- 
tionately as in Prothylacynus. The shaft is straight. The condyles are 
narrower antero-posteriorly than in Thylacynus and are plane or slightly 
convex transversely. The inner condyle is considerably wider than the 
outer. 
Restoration (PI. LXI, fig. 3). — The disproportionately large size of the 
head and great length of the neck are at once apparent in the drawing of 
the restored skeleton. The length of the back has been determined by 
comparison of the lengths of the few dorsals preserved with the corre- 
sponding parts in Thylacynus. The lumbar vertebrae and pelvis are sup- 
plied from Prothylacynus. The length of the tail is largely hypothetical, 
but the size of the proximal caudals indicates that it was greatly thickened 
