348 



Wortman — /Studies of Eocene Mammalia. 



the fox, but the coracoid is larger and more elongate as in 

 Cynodictis. The neck is short, the spine 

 6 rises abruptly, close to the glenoid border, 



and terminates in an overhanging acromion 

 and metacromion. There is not enough of 

 the bone preserved to indicate the relations 

 of the fossae with certainty, but they were 

 apparently like those of Cynodictis as 

 described by Scott.* 



The humerus, figure 9, is broken in the 

 region of its proximal third and does not 

 give the character of the head of the bone. 

 The most conspicuous feature of the shaft 

 is the development on its anterior surface of 

 the very large and prominent deltoid crest, 

 which extends more than half way to the 

 distal end. Associated with this are seen 

 the unusually prominent supinator ridge and 

 internal condyle, all of which serve to give 

 to the bone a combination of characters 

 hitherto completely unknown among the 

 Canidse. The entepicondylar foramen is 

 present but there is no anconeal perforation. 

 Figure 6. — Left The anconeal and anticubital fossae are dis- 



wortman; luS"' tillot and the articular surface is of unusual 

 transverse extent, as it is in the cats and 

 viverrines, and very different from the modern dogs. 



The radius, figure 6, resembles that of the civets much more 

 than of the modern dogs. The shaft is slightly curved and 

 moderately flattened. The head is cup-shaped and subcircular 

 in outline, indicating power of complete pronation and supina- 

 tion ; the tubercle for the insertion of the biceps is strong and 

 that for the principal pronator of the fore arm is unusually 

 well developed. The distal end is broken so as to destroy 

 a considerable part of the articular surface for the carpus. 

 Judging from what is preserved of this surface, however, 

 there was apparently no division into separate scaphoid and 

 lunar facets. This would seem to indicate a consolidated 

 scapho-lunar, but this evidence is not altogether reliable. 



* Trans. Philos. Soc. Phila., vol. ix, p. 381, 1898. 

 [To be continued.] 



