32 A REVISION OF THE COTYLOSAURIA OF NORTH AMERICA 



absence of abdominal armor may well remain in doubt. Williston and Moody 

 have shown it to be absent in this specimen, but it has also been absent in many of" 

 the specimens from the Permian and finally has been found to have been present. 

 In fact, the abdominal ribs would be one of the first things to disappear, as the soft 

 abdominal wall inclosing them would be loosened by the decay of the intestines. 

 There is no certainty that the abdominal ribs were absent and much probability 

 that they were present. 



Genus BATHYGLYPTUS nov, 

 Bathyglyptus theodori sp. nov. 



Type: Fragments of two lower jaws. Field No. 250 Am. Mus. Nat. Hist., 

 Expedition of 1908. From Willbarger County, Texas. 



The jaws belonged to an animal considerably larger than any of the well-known 

 forms of the family DtadectidiB. The teeth are broken at the base, and with the 

 exception of one or two are poorly preserved. The roots of the teeth indicate that 

 they were of equal size, except a single tusk at the anterior end, and that all were 

 round in section. The jaw might be taken for that of a large Pelycosaur, but it is 

 much heavier and more rugose than that of any known form. It resembles that of 

 the Diadectidce much more closely, but it is distinguished from that family by the 

 round teeth. There are nineteen teeth, which was probably the complete number. 



Measurements. 



nun 



Length of longest fragment of jaw jn 



Probable length of jaw 375-+00 



Anteroposterior diameter of largest tooth at base 21 



Breadth of same I^ 



FOREIGN FORMS. 



Genus SAURAVUS. 



Sauravus costd Thevinin. 



Annales de Paleontologie, Paris I, Heft. 3, 1906. 



This specimen was found in the upper part of the Stephanien of France, near 

 Blanzy (Saone-et-Loire). It most unfortunately lacks the skull, but a good part 

 of the skeleton of the body is preserved. The vertebrae are notochordal; the 

 neural spine is low and coossified with the centrum. Intercentra were not observed, 

 but large spaces between the lower edges of the centra indicate their presence, pos- 

 sibly in a cartilaginous state. There were twenty-three or twenty-four dorsal ver- 

 tebrae, two sacrals, and at least seventeen caudals. The ends of neural arches and 

 haemapophyses are striated as in Keraterpeton from the Permian of Nyran in Bohe- 

 mia. All presacral vertebrae bear ribs and there are abdominal ribs present. The 

 interclavicle is T-shaped. The humerus has the proximal and distal ends almost 

 at right angles and there is no entepicondylar foramen (?). The tarsus has fibulare, 

 tibiale, and five distal tarsalia. The phalangeal formula of the posterior foot was 

 probably 2, 3, 4, 5, 4. 



