392 NOTES ON THE CANIDEH OF THE WHITE RIVER OLIGOCENE. 
bral column, whence it becomes evident that all the limb bones of Cynodictis are propor- 
tionately shorter than those of Canis, and that the bones of the forearm are especially 
short. The tibia of Cynodictis differs from that of the modern canines in several par- 
ticulars. The proximal condyles are of nearly equal size, but the external one projects 
much farther behind the plane of the shaft than in Canis, and on the distal face of the 
overhanging shelf thus formed is a facet for the head of the fibula, which is much larger 
and more distinct than in the recent genus. The tibial spine is bifid and very low, but 
the two parts are closely approximated, the condyles being less widely separated than in 
Canis. The cnemial crest, though stout and prominent, is much less so than in the mod- 
ern forms, and the sulcus for the extensor longus digitorum is much less deeply incised. 
In its proximal portion the shaft is stout and trihedral, but for most of its length it is 
slender and subcylindrical, expanding moderately at the distal end; it has a double cur- 
yature, arching forward and outward. ‘The various ridges which serve for the attach- 
ment of muscles are much the same as in Canis and are, consequently, better developed 
than those of the femur. The distal articular surfaces of the tibia are intermediate in 
character between those of Daphenus and those of Canis. The grooves for the astraga- 
lar condyles are deeper and the intercondylar ridge higher than in the former, less so 
than in the latter, and the suleus which in Canis invades the articular surface has not 
yet been developed. The internal malleolus is somewhat smaller than in Daphenus, 
but, as in that genus, it forms a heavy, prominent ridge, which extends across the whole 
dorso-plantar diameter of the bone, while in Canis the process has not half this exten- 
sion. The groove for the tendon of the long flexor muscle is very distinctly marked and 
has more elevated borders than in the modern dogs. The distal fibular facet is some- 
what larger than that of Canis and differs from it in haying its principal diameter trans- 
verse instead of longitudinal. The resemblance in the structure of the distal end of the 
tibia between Cynodictis and Daphenus, on the one hand, and the primitive sabre- 
tooth Dinictis, on the other, is very marked and very suggestive, though Cynodictis has 
already begun to change in the direction of the modern Canidw. Among living forms 
the tibia of Herpestes offers a close analogy to that of the White River genera which 
have been mentioned. 
The fibula is relatively much less reduced than in the existing Canidae, and both the 
shaft and the terminations are larger. The proximal end of the fibula is much larger 
and heavier proportionately than in Canis, and though smaller than in Dinictis, it has 
a very similar shape; its principal diameter is the antero-posterior one, while trans- 
versely it is narrow and compressed ; the thickening of the anterior and posterior border 
is present, as in Dinictis, but much less conspicuous. The facet for the head of the 
tibia is large, subcircular in shape and proximo-lateral in position. The shaft, though 
