AX AXOXODONT REPTILE. 



207 



part of the first vertebra; and a larger proportion of the arti- 

 cular concavity is there preserved. The entire breadth of this ver- 

 tebra is 4J inches. 



Turning to the dorsal aspect of the sacrum (PI. X. fig. 3). the 

 hinder end of a lofty but narrow crest of bone, n s, due to confluence 

 apparently of the neural srjines of the fourth and fifth sacrals, has 

 been worked out. The entire breadth of the fifth sacral vertebra is 

 5 inches. The condition of the spines of the antecedent sacrums 

 could not be determined. In the portion of a Dicynodont sacrum 

 described in the - Transactions of the Geological Society,' 2nd series, 

 vol. vii. pi. xxxiii. figs. 4 & 5, the short and thick spines of the 

 first and second sacrals are distinct, not confluent. 



The section of the circular area of the neural caual of the fifth 

 sacral of Platypodosaurus gives a diameter of 6 lines ; that of the 

 first sacral has a diameter of 9 lines ; that of the last lumbar ver- 

 tebra one of 8 lines. Thus we have an indication of an expansion 

 in the sacral region for the lodgment of the part of the myelon 

 transmitting the nerves to the pelvic extremities, which accords with 

 the development of the limbs indicated by the femur of Platypodo- 

 saurus *. Comparing the sacrum above described with that of the 

 larger and more entire pelvis, the subject of a paper in the ' Geolo- 

 gical Transactions ' above quoted, I incline to regard " five " as the 

 total number of the sacral or pelvic vertebrae in Platypodosaurus t. 



The degree of coalescence of these vertebrae is such as to sustain 

 the anthropotomical or mammalian consideration of the coalesced mass 

 as one bone or " sacrum but the ventral outlets are relatively 

 larger and the wings consequently less ossified. The general shape, 

 moreover, is quadrate rather than triangular, with deeper lateral 

 concavities between the subcarinate bodies and the iliac bones. The 

 present Reptilian sacrum consequently comes nearer in shape to that 

 of the Megatherioid mammals : but it includes fewer vertebral 

 constituents. 



The length of such "sacrum" in Platypodosaurus is 7| inches; 

 its breadth, greatest at the third vertebra, is 5j inches; it is con- 

 sequently more mammalian in character than is the sacrum of Bicy- 

 nodon J. In this larger example of the South- African reptilian pelvis 

 the sacral centrums are more constricted between the articular ends, 

 which are concomitantly more expanded. The space between these 

 sacral bodies and the "ossa innominata" is relatively greater : the 

 transverse processes are consequently longer, and retain more of the 

 ordinary reptilian rib-like character. 



The iliac bone in Platypodosaurus rapidly expands from its attach- 

 ment to or confluence with the first sacral transverse process ; it ex- 

 tends and curves forward, outward and dorsad, with the convexity 

 mesiad, the concavity lateral. The extreme breadth is 4| inches ; 

 and the anterior margin of the ilium runs almost parallel with the 



* Loc. tit. p. 422, pi. xyii. figs. 6 & 7. 



t Compare also with page 40, plate xxxvi. (pelvis of Dicynodon), in the 

 1 Catalogue of Fossil Eeptilia of South Africa,' 4to, 1870. 

 \ Loc. cit. 



