194 NATURAL SCIENCE [September 



Oplithahnosaurus icenieus (B. 2137), described and figured by Seeley 

 (Proc. Roy. Soc, vol. liv., 1893, fig. 1, p. 151). No such difficulty 

 beset the restoration of the pelvis, since the parts are hardly at all 

 displaced in specimen (41849). 



The diagram of the Plesiosaur is mainly based on the splendid 

 specimen of Plcsiosaurus rostratus from the Lower Lias of Char- 

 mouth, Dorsetshire, exhibited in the Geological Gallery of the 

 Museum, and bearing the register number (38525). This specimen 

 was described and figured by Owen in his " Liassic Eeptilia " 

 (Sauroptcrygia, 1865, pi. 9), but it did not form the basis of his 

 well-known text-book restoration of Plcsiosaurus (" Anat. of Vert.," 

 vol. i., 1866, p. 52), the species of which, according to Lydekker 

 (Brit. Mus. Cat. Foss. Rcpt., part ii., 1889, p. 121), is macroccphalus. 

 The number of cervical vertebrae in PUsiosaurus rostratus is not 

 definitely known. Owen put it down as twenty-four, but there 

 were probably more, since in the specimen (38525) there are 

 evidently some vertebrae missing after the seventeenth (see 

 Lydekker, loc. tit., p. 272). Judging from the shape and relations 

 of the cervical ribs flexion of the neck must have been as difficult 

 of achievement in Plcsiosaurus as in our modern crocodiles, and so 

 the vertebral column in the cervical region has been drawn nearly 

 straight (Fig. 3), instead of being allowed the graceful sinuous curve 

 which characterises Owen's figure. The outline of the body has 

 been introduced from the figure given by Dames (Abkandl. konigl. 

 Almcl. Wiss., Berlin, 1895, ii., p. 79) ; and special attention may be 

 called to the shape of the tail fin, and to the presence of an integu- 

 mentary extension of the paddle behind the part supported by the 

 internal skeleton. The transverse temporal ridge at the back of 

 the skull would probably not have influenced the general contour of 

 the body to the extent suggested by the diagram. This improba- 

 bility should have been avoided by making the vertebral column 

 articulate a little higher up the occiput, and by putting the cranial 

 axis more in a line with the cervical vertebrae. 



The cranium of the specimen above mentioned is considerably 

 crushed ; therefore, while preserving the proportions of the cranial 

 bones of this species, the actual details were added from the more 

 perfect skull (49202) of the allied species P. macroccphalus, de- 

 scribed and figured by Andrews (Quart. Journ. Geol. Soc, vol. lii., 

 1896, pp. 246-253, pi. 9). The skeleton of the paddles in the 

 specimen of Plcsiosaurus rostratus is extremely well preserved, and 

 nothing more was necessary than to copy the outlines of the 

 constituent bones ; but as the bones of the pectoral and pelvic 

 girdles are disturbed, a certain amount of restoration was here 

 inevitable, and the assistance derived from the perfect girdles of 

 Muracnosaurus plicatus (R. 2428) and Crypt oclldus oxoniensis 



