Sinclair: marsupialia of the santa cruz beds. 367 
the caudals, the tail of Prothylacynus must have been not only longer 
but much thicker at the base than in Thylacynus. It is possible that too 
much flexibility has been given to it in the drawing of the restored skele- 
ton. The floor of the neural canal in the cervical, dorsal, lumbar and 
anterior caudal vertebrae is keeled and perforated on either side of the 
keel by a large foramen, as in Thylacynus and Borhycena. The epiphyses, 
although more or less annular, as in Borhycena , do not have such promi- 
nent bosses from the centra projecting through the median perforations. 
Frequently the latter are absent. 
Fragments of several posterior ribs are preserved, but are too incom- 
plete to describe. They have about the same dimensions as the posterior 
ribs in Thylacynus . The incomplete presternal segment represented in 
fig. 6, PI. LI, is considerably smaller than that of either Thylacynus or 
Borhycena , differing from both in possessing a strong, inferior, median 
keel. 
Appendicular Skeleton . — The scapula (PI. XLIX, figs. 2, 2 a) is shorter 
than in Thylacynus , with thicker neck and longer coracoid process. The 
surface of the supraspinous fossa is convex, in contrast with the approx- 
imately plane surface of this region in Thylacynus. In part, the con- 
vexity has been accentuated by crushing. The glenoid cavity is almost 
circular in outline. The coracoid process projects greatly below the 
glenoid margin. Its anterior border is strongly inflected. The infra- 
spinous fossa is somewhat wider than in either Borhycena or Thylacynus. 
As in those genera, the axillary border is strongly deflected outwardly. 
The spine has been broken, destroying the acromion. The border of the 
suprascapular notch is perforated by a small foramen. 
The humerus (PI. XLIX, figs. 1-1 b) is shorter than in Thylacynus but 
much heavier. The head is very broad and projects considerably beyond 
the posterior surface of the shaft. The greater tuberosity rises slightly 
above the head. The lesser tuberosity is low, not as sharply separated 
from the head as in Thylacymis. A third tuberosity rising from the inner 
side of the shaft below the lesser tuberosity probably marks the point of 
insertion of the coraco-brachialis muscle. The deltoid ridge is long and 
powerful, extending at least two thirds the length of the shaft. The distal 
end is exceedingly broad, with greatly enlarged supinator ridge and elon- 
gated inner epicondyle. The condylar surface is much narrower antero- 
posteriorly than in Thylacynus , but considerably wider transversely, with 
