Marsh Collection^ Peabody Museum. 197 



raw. 



Length of centrum of axis including odontoid. 16 



Length of centrum of third cervical _ — 8* 



Length of humerus 68*5 



Transverse diameter of humerus (distal end) . . 16-5 



Length of ulna (estimated) 71* 



Length of radius 58* 



Length of carpus and metacarpus (estimated). 24- 



Length of phalanx, second row 10* 



Length of pelvis . _ 74* 



Transverse diameter of ilium 12* 



Length of femur 92 • 



Transverse diameter of proximal extremity 16*5 



Transverse diameter of caput femor is ... 8-3 



Transverse diameter at lesser trochanter 15- 



Transverse diameter of distal extremity 16* 



Anteroposterior diameter of condyles 12* 



Length of patella 6" 



Transverse diameter of patella 6*8 



Length of tibia (estimated) . „ 81*5 



Transverse diameter of head of tibia 14-5 



Antero posterior diameter of head of tibia 11' 



Discussion. — In the foregoing description of this family, 

 frequent reference has been made to their relationship with the 

 Viverridse, the characters in which the two families resemble 

 as well as differ from each other have been pointed out as far 

 as they are known, and the position has been generally 

 assumed that the one stands in direct ancestral relation to the 

 other. I will now proceed to summarize the arguments upon 

 which this assumption is based, but it is first necessary to 

 observe that the idea is by no means a new one. Professor 

 Marsh, in giving the name Viverravus to the species first 

 described, expressed this relationship, although he did not 

 follow it up with the necessary evidence to establish it upon a 

 firm foundation. In my work on the " Comparative Anatomy 

 of the Teeth" (1886), I gave utterance to the same view, but did 

 not attempt a demonstration of the same. Flower and Lydek- 

 ker, in " Mammals Living and Extinct," 1891, p. 539, state : 

 " The North American genera Miacis and Didymictis 

 ( Uintacyon and Viverravus) are generally regarded as 

 representing a separate family — Miacidse — with affinities to 

 both the Yiverridas and the Canidae." In our paper on " The 

 Ancestry of Certain Members of the Canidae, the Viverridse, 

 and Procyonidee" (Bull. Amer. JVIus., 1899), Matthew and 

 myself reiterated this opinion, and from a study of the skeleton 

 of a species of Viverravus brought forward some evidence 

 in its support. 



From a consideration of the materials at present before us, 

 it may be regarded as clearly established that there were at 



