1889.] OF A THERIODONT REPTILE. 573 



somewhat compressed from side to side ; and the neural spines are 

 likewise flattened, and of moderate length and height. Their 

 terminal faces are but sh'ghtly cupped, and siiovv a notochordal 

 canal penetrating towards the middle of the vertebra. The transverse 

 processes {t.p) are of moderate length, and directed backwardly as 

 well as outwardly. These transverse processes are strengthened by 

 three plate-like buttresses arising from the sides of the arch, and re- 

 calling the structure observed in the dorsal vertebrae of Megalosaurus. 

 A trace of a rib-facet is observable on the anterior border of the 

 centra of these vertebrae, which probably indicates that they belong 

 to the earlier part of the series. The hajmal aspect has a sliarp 

 carina. Intercentra wei'e totally wanting in this part of the spinal 

 column. The total height of these vertebrae is ti"7 inches, the 

 vertical diameter of the centrum being 2'4 inches. 



Somewhat larger trunk-vertebrae probably belong to the lumbar 

 region, an imperfect specimen being represented from the anterior 

 aspect in fig. 3 of the same Plate. In this specimen the centrum is 

 very much shorter than in the dorsals, but the haemal carina is still 

 sharper. In the caudals, of which there are three in ap[)Osition, the 

 centrum again lengthens, and the haemal carina becomes less sharp. 

 Chevrons were certainly present, but whether intercentra occurred 

 in this region cannot be satisfactorily determined. 



The apparently notochordal character of these vertebrae indicates 

 that this series of specimens does not belong to the Dicynodont sub- 

 order of the Anomodonts. Notochordal vertebras are met with both 

 in the Pariasaurian and Theriodont subdivisions of that order; but 

 since the associated humerus to be immediately noticed accords with 

 that of the Theriodouts and is quite different from the type 

 apparently referable to the Pariasauria ', there seems every reason 

 for regarding these specimens as referable to the Theriodontia. 



Turning to the bones of the appendicular skeleton, we have first 

 to consider the scapula, of which tbe proximal portion of that of the 

 right side is preserved. This specimen is represented from tbe dorsal 

 aspect in Plate LV. fig. 1, on a scale of one third, with a restoration of 

 the missing half from the scapula described by Sir R. Owen as Platy- 

 podosaurus. The proximal portion of this bone agrees so closely in 

 general characters with the latter specimen, as figured by its describer 

 (Quart. Journ. Geol. Soc. vol. xxxvi. pi. xvii. fig. I), that there is no 

 necessity for a detailed description. It will be seen that the process 

 marked a in the figure corresponds with that marked f in Owen's 

 plate, while the one here marked b represents e of the latter. 



In regard, however, to the homology of these two processes with 

 those of other Dicynodont scapulae, it is necessary to make a few 

 remarks, since some confusion has arisen between the two. In 

 his ' Catalogue of the Fossil Reptilia of South Africa,' Sir B.. Owen 

 figured, in plate Ixx. fig. 1, the ventral aspect of the right scapula 

 of a species of Dicynodon, with a portion of the precoracoid attached 



1 Described by Prof. Seeley (Proo. Eoy. Soc. vol. xliv. p. 142, 18SS) as 

 Propaj)]nis. The vrriter will take a later opportunity of giving his reasons for 

 this reference. 



