128 Annals of the South African Museum. 



low down on the centrum, the upper in the 3rd and 4th between the 

 lower and the pre-zygapophyses, and in the 5th considerably nearer 

 the zygapophysis. The zygapophyses are comparatively small, so that 

 these cervical vertebrae are nearly twice as high as they are broad. 

 Intercentra have lain between each pair of vertebrae . 



In the 6th vertebra the two articulations for the head of the Y\h 

 become united, and in all the succeeding presacral vertebrae there 

 is only a single, though usually very extensive, articular surface 

 for the head of the rib. On the articular surface becoming 

 single the upper part becomes very much more strongly developed 

 and projects well outwards from a point a little below and behind 

 the prezygapophysis. 



On passing backwards the zygapophyses become much more 

 strongly developed. The width between the points of the prezyga- 

 pophyses of the anterior vertebrae is as follows approximately : — 

 2nd -OTOm. ; 3rd -OSOm. ; 4th -OgOm. ; 5th -OOOm. ; 6th -lOOm. ; 

 7th •130m.; 8th •135m. 



Seeley figures (I. plate 17) a restoration of a cervical vertebra of 

 P. bombidens, but he does not say which of the cervical vertebrae it 

 is intended to represent. It does not bear much resemblance to any 

 of the vertebrae in the present skeleton. Whereas Seeley represents 

 the body as considerably more than half the depth of the whole 

 vertebra, in all the upper vertebrae, except of course the atlas and 

 possibly the axis, of which the spine is lost, the arch is very con- 

 siderably higher than the body is deep. Thus in the 5th vertebra 

 the whole height is •200m., while the body only is •075m. ; and in 

 the 6th vertebra the height is •210m., while the body is •090m. 



The vertebrae from the 9th to the 18th inclusive, resemble one 

 another very closely, only varying a little in the relative size of the 

 parts. 



Seeley's restoration of a dorsal vertebra (I. plate 18) is very unlike 

 any of the dorsal vertebrae of Pareiasaurus serridens either in struc- 

 ture or proportions. In his later paper (II. p. 320) he admits his 

 error in making two articulations for the head of the rib instead of 

 only one. 



The following are the measurements of the 11th vertebra which 

 may be taken as typical of the series. Height •193m. ; height of 

 centrum •075m. ; width at transverse process •215m. ; width at 

 postzygapophysis •165m. The very deep lateral process which gives 

 articulation to the rib has an articular surface measuring •113m. in 

 length. Its lower end arises from a little below the level of the 

 middle of the centrum. The postzygapophyseal surfaces look down- 



