On Some Neiv TJicroccpliaUan Bcpliles. 363 



usual way. The temporal arch is formed by the squamosal and the 

 jugal, the latter being of large size and forming the lower half of 

 the bar. There is evidence of a columella cranii. 



The mandible is fairly well preserved. The dentary is of large 

 size and has a well-developed coronoid process. Behind it there is 

 a large I'adiating angular and above it a surangular, very like those 

 bones in Lycosuchus. ■ There are 3 fairly large incisors and a single 

 large canine. 



The second specimen is not in very good preservation. It is 

 chiefly of value in showing the structure of the squamosal, which is 

 a large bone somewhat similar to that of the Anomodonts, having a 

 descending process which supports the quadrate. 



Alopecodon rugosus, n. sp. 



This species is founded on an imperfect skull from near Zeekoegat. 

 The fragment wants both the front of the snout and the back of the 

 head. There is thus some doubt as to whether it should be referred 

 to the genus Alopecodon. As the molars are of the same flattened 

 type as in A. priscus, and the maxilla, mandible, and other bones of 

 a similar type, I think it advisable to refer this species to the 

 same genus. It is undoubtedly a distinct species, and the above 

 name is proposed for it in allusion to the very rough condition of 

 the maxillary bone. The animal is about two-thirds the size of 

 A . priscus. 



The maxilla is of the same radiating type as in most Thero- 

 cephalians. It is chiefly remarkable for the very marked pitting on 

 the surface of the upper part of the bone. Whether this is due to 

 glands below, a soft skin, or to sense organs homologous with those 

 at the base of the tactile hairs in mammals, or to the close adherence 

 of horny plates to the bone is not manifest. Perhaps the suggestion 

 that the pits are due to sense organs is the most probable, but against 

 it is the fact that in no known Therocephalian has a large infra- 

 orbital foramen been detected, and that hence if the maxillary branch 

 of the 5th nerve passed on to the face as is probable it must have 

 been small. 



On the one side there are evidences of 6 and possibly of 7 

 molars, but on the other side only 5 can be made out. Probably 

 the complete number is 8, as in ^. priscus. The teeth are flattened 

 and relatively shorter than in A. priscus. Three of them show 

 serrations on the posterior margin, but none on the anterior. The 

 average height of the first 3 molars is 5*5 mm., and the average 



