34 Transactions of the South African Philosophical Society. 



having long claws. The phalangeal formula is 2, 3, 3, 3, 3, and 

 there is no indication of any approach to the usual reptilian formula, 

 such as appears to be present in the hand of Theriodesmus. Though 

 the phalanges of the three middle toes are about equal in length, the 

 4th toe is longer than the 3rd and the 3rd than the 2nd, owing 

 to the difference in length of the metacarpals. 



The pelvis has only the lower elements preserved in specimen A. 

 The pubis is an irregular quadrangular concave bone, with the 

 concavity directed downwards. The anterior border is nearly 

 straight and in line with the anterior border of the neighbouring 

 bone. Near the posterior border of the bone is a large oval pubic 

 foramen, apparently completely surrounded by the bone. The 

 ischium is of the usual plate-like type seen in most early reptiles. 

 It forms a long symphysis with its neighbour, and makes with it an 

 angle of about 90°. In specimen B there is preserved a very good 

 cast of the left ilium. The acetabular portion is large, and the upper 

 portion of the bone relatively small, and is directed slightly back- 

 wards. The crest is short. 



The femur of the left side is preserved in specimen A, and' shows 

 the outer side, but not sufficient of the bone is showing to permit of 

 a detailed description. The bone is long and slender and very 

 different from the femur of Ictidosuchus. The tibia and fibula are 

 fairly well preserved in both specimens. Both bones are con- 

 siderably curved and a wide gap is left between them in the middle 

 region. 



In neither specimen is the foot well shown. In specimen A a 

 large proximal bone is present, probably the tibiale, and the greater 

 part of three toes. So far as preserved, the toes are very similar to 

 those of the front foot. 



With the exception of the caudals none of the vertebrae are well 

 preserved. The few cervicals showing have only the centra pre- 

 served, but are specially interesting from having well-developed 

 intercentra. The impressions of the dorsals are not very satisfactory, 

 but there are apparently no intercentra. The ribs in the upper 

 dorsal region are single-headed. In both specimens there are to be 

 seen numerous abdominal ribs. From the crushed nature of the 

 specimens it is impossible to say exactly how these have been 

 arranged, but there is probably a median series and two lateral 

 series. The tail is extremely long and there are -probably over 60 

 vertebrae in it. Specimen B has a series of about 30, and the first of 

 these is probably not further forward than the 20th, and there 

 must be about 10 distal ones missing. There are well-developed 

 chevrons, but the spines are small. 



