part 8] FBOM THE WEALD CLAY OF BERWICK ( SUSSEX). 291 



the anterior angle of the pedicle to the posterior zygapophysis, a 

 ridge which is wanting in our specimen. 



Comparison with the vertebrae of Plesiosaurios ca pens is 

 Andrews,^ from probably the nearly contemporary beds of Uiten- 

 hage (Cape Province), shows very considerable similarity in several 

 respects. The form of the neural spines in the posterior cervical 

 and thoracic vertebrae is nearly the same ; the articular face of the 

 centrum has the same sudden depression of its middle portion ; 

 the zygapophyses in both are large, and project considerably in 

 front of and behind the articular surfaces. The proportions of 

 the centra are ahnost the same : thus if, as already mentioned 

 above, the length of the centrum in the new specimen be taken as 

 100, the width will be represented roughly as 155, the height 

 as 146. In P. capensis, under the same conditions, the width 

 would be 156, the height 138. Differences in the skull, apart 

 from other considerations, make it very unlikely that the specimen 

 here described is specifically identical with the South African 

 species. 



JBircncasaurus hrancai Wegner is by far the best known of the 

 Wealden Plesiosaurs, most of the skeleton having been described 

 and figured.2 The vertebrae resemble to a considerable degree 

 those described above, the neural spines and the articular surfaces 

 of the centra being very similar. On the other hand, the length 

 of the cervical centra is greater in proportion to their width and 

 especially to their height; the proportions are: — length = 100, 

 width = 140, height =112. Other differences in the skull and 

 shoulder- girdle are mentioned elsewhere (pp. 288, 296). 



The shoulder- girdle (PI. XV, figs. 1-3) is in a remarkable 

 state of preservation, the bones, even the thin clavicular arch, being 

 quite undistorted. Some parts of the bones are wanting ; but the 

 remaining portions being retained in their original positions in 

 the matrix, it has been possible to fill up the gaps, and to restore 

 accurately the original form of the various elements. 



The general form of the shoulder-girdle will be best understood 

 from the figures. It will be seen that the coracoids {cor.) are 

 long and narrow. Opposite the glenoid and scapular surfaces they 

 are thickened, the thickening extending to their median border, 

 where it ends in a symphysial surface, convex longitudinally above 

 and concave in the same direction below. The glenoid {gl.) sur- 

 face, which forms about two-thirds of the articular surface for the 

 humerus, is nearly flat, and makes an angle of about 120° with the 

 facet for the scapula. The post-symphysial portion of the coracoids 

 is thin, and in this region they must have been separated by a short 

 interval in the naiddle line. The thin inner border of the left 

 coracoid just behind the symphysis is cut into by a notch with a 

 thin sharply-defined border : this notch seems to represent the 



1 Ann. S. Afr. Mus. vol. vii (1911) p. 309. 

 ^ Branca-Festschrift, Leipzig, 1914, p. 235. 



