194 Annals of iho South African Mnsciim. 



constricted at the middle. In front it measures approximately 

 100 mm. across and 90 mm. vertically. Posteriorly the arti- 

 cular surface measures 90 mm. across and 90 mm. vertically. 

 The antero-posterior length of the body is 60 mm. The arch 

 has a well-developed spine and a fair-sized transverse process 

 which is directed outw^ards and downwards ; it articulates with 

 the centrum as in the Dinosaurs and the suture remains dis- 

 tinct. The transverse process has a flattened end for the articu- 

 lation of the rib. The main part of this articular surface is of 

 a broad oval shape, and anterioi'ly is continued downwards to 

 join one nearly as large on the anterior edge of the centrum. 

 From the appearance of this surface it is manifest that the rib was 

 single headed, but with the head broad and the middle portion very 

 slender. In Belodon and StacjonolcpiH the ribs are double headed. 

 The greatest width across the transverse processes is 130 mm. The 

 zygapophyses are situated close together, and are directed upwards 

 and inwards or downwards and outwards. 



There are also a series of four larger dorsal vertebras belonging 

 probably to the lower dorsal region. These resemble each other 

 closely, but ditfer from the vertebra above described in a number of 

 points. In the first and second the body measures antero-posteriorly 

 70 mm., and in the third G7 mm. The width of three of them 

 averages 73 mm., and the depth 90 mm. The arches differ from 

 that of the upper vertebra chiefly in being longer antero-posteriorly, 

 in having a larger spine, and in having the transverse process 

 slightly different. In the second of the series the greatest antero- 

 posterior measurement is about 120 mm. The width at the zygapo- 

 physis is 50 mm., and across the transverse processes 130 mm. 

 The end of the transverse processes differs from that of the anterior 

 vertebra in having the facet for the head proper of the rib shifted 

 from the centrum to the arch. The process has thus a large 

 rounded end for the upper part of the end of the rib, and an equally 

 large lower facet for the head proper. The two articular surfaces 

 are connected by a narrow bridge. None of the figured vertebrye of 

 Belodon and Stagonolepis agree very closely with the known verte- 

 brae of Erythrosuclnis, but this may be in part due to the vertebrae 

 being from different regions. In both of the European genera the 

 transverse processes seem to be, as a rule, longer. A good many 

 vertebrae of Stagonolepis were figured by Huxley, but not well 

 enough to be of much value for purposes of comparison. 



Of the sacral vertebrae the only parts satisfactorily preserved are 

 the ribs. Of these the anterior is the larger, its end measuring 



