NOTES ON THE CANID#® OF THE WHITE RIVER OLIGOCENE. 389 
the phalanges of the two genera. As compared with the ungual of Daphenus, it has a 
somewhat less concave proximal trochlea, a smaller subungual process, and a much less 
extensive bony hood reflected over the base of the claw. Indeed, this hood is rudi- 
mentary and can hardly be said to exist at all. The phalanx is also slightly thicker and 
has more conyex faces. Comparing this ungual with that of Canis, we find it to be 
decidedly sharper, narrower and more compressed and to haye a more deeply concave 
trochlea. In the modern genus the bony hood is almost as well developed as in Daphenus. 
? VII. Tur Hinp Live. 
The pelvis approximates more nearly to the modern canine type than does that of 
Daphenus, though still retaining a number of primitive characters. A conspicuous 
difference from the recent members of the family consists in the elongation of the post- 
acetabular portion of the pelvis, which in Canis is short, and in the consequent change 
of shape of the obturator foramina. The ilium is fairly elongate and in shape is rather 
more yiverrine than canine ; the peduncle is short and laterally compressed, but of con- 
siderable dorso-ventral breadth. The anterior expansion of the ilium is less extensive 
than in Canis, in which genus the ilium widens gradually to the free end, or erista, while 
in Cynodictis it attains nearly its full width immediately in front of the peduncle, and 
from this point forward the dorsal and ventral (or ischial and acetabular) borders pursue 
an almost parallel course. The widening is almost confined to the ischial border, being 
very feebly marked on the acetabular border, and owing to this the shape of the ilium is 
much as in the modern Herpestes. The gluteal surface does not display the wide and 
simple concavity which is seen in Cunis, but, as in Daphenus and Dinictis, there is a 
narrow dorsal depression and beneath this a convex ridge, but this ridge is not so 
prominent as in the other White River genera which haye been mentioned. The iliac 
surface is short and narrow, and the sacral surface is small and placed far back, so that 
the ium projects well in front of the sacrum. When viewed from above, the two ilia 
are seen to curve outward less, and to diverge less anteriorly than in the modern dogs. 
The acetabular border ends in a well-marked tubercle and the ilio-pectineal process is 
also quite prominent. | 
The ischium is relatively long and its anterior portion is slender, but posteriorly it 
expands into a broad plate. This posterior portion is much less decidedly everted and 
depressed and oceupies a more vertical position than in Canis, and the ischial tuberosity, 
just as in Daphenus, is much more feebly developed than in the existing Canide. On 
the other hand, the spine of the ischium and the ischiadic notch are much more distinetly 
shown and are placed farther behind the acetabulum than in the latter, though not so far 
back as in Herpestes, The obturator foramen is narrower and more elongate than in 
