340 Prof. H. G. Seeley on a Pneumatic Type 



voider tlian deep, though the depth postenorly is reduced by 

 criisli^ which depresses the neural arch. 



The neural arch is rather small, with strong, thick, large 

 posterior zygapophysial processes, which diverge backward 

 and outward, and are angulated superiorly. The facets, which 

 are imperfectly preserved, look downward and outward. 

 The anterior processes, which were small, are lost. The 

 neural spine is lost. Its base was six tenths of an iuch long; 

 it a))pears to have been widest behind, and may have been 

 channelled at the back, in harmony with tiie notch between 

 the posterior zygapophjses. As preserved, the neural arch 

 has the aspect of being inclined backward. 



Taken as a whole, the characters of this vertebra shown on 

 the facets for the intercentra and the articular faces of the 

 centrum appear to me to indicate that the animal from which 

 it was derived was an Anomodont, and not a Saurischian 

 Dinosaur; and it differs from all known members of that 

 group in possessing pneumatic vertebrae. It may therefore 

 indicate a group of Pneumatospondylia, showing some affinity 

 between the Anomodontia and Saurischian Dinosaurs, in 

 which a similar pneumatic condition of the vertebral column 

 is found. In the forms of the pelvic bones, especially the 

 ilium and ischium, there are interesting resemblances between 

 these groups, which extend also to various bones of the linibs, 

 in son)e genera. 



The locality which yielded this bone is chiefly remarkable 

 for the remains of large animals, such as Pareiasaurua and 

 Tajnnoce2)halus ; but at Cypiier I found, in association with 

 limb-bones and skull-fragments of those types, the figured 

 fragment of a skull of the Lycosaurian genus Pristerogmithus. 

 1 have seen no vertebrate remains, except those of Lycosauria 

 from the Lower Karroo rocks of that part of Africa, which 

 approximately correspond in size with the animal indicated 

 by this vertebra. Lycosaurian vertebras are unknown from 

 African specimens. 



There is a general resemblance in type to the vertebra 

 named Arctosaurus, though that genus shows no indication 

 of a pneumatic excavation or of intercentra. I followed the 

 trail of travelled and broken bone-fragments up the slope 

 down which they had been swept by the rains for a con- 

 siderable distance without finding any further evidence of the 

 animal, unless it is a much smaller crushed dorsal vertebra 

 too distorted for description, which has some resemblance to 

 A7-ctosaurus. 



The name suggested for the species commemorates the 



