144 BRITISH FOSSIL REPTILES. 



five lines, the length of the two being nine lines. In each, the sides of the centrani 

 are nearly plane, and converge at an acute angle to a ridge, which forms the under 

 surface ; a very small hypapophysis was continued from the back end of the ridge. 

 This process is broken away from each vertebra, as are also the diapophyses, which 

 are indicated by their rough fractured base ; they are situated near the lower part of 

 the side of the centrum, like the long diapophyses of the posterior caudal vertebrae of 

 the Pythons ; had they been preserved, their proportions would have determined 

 whether the anchylosed vertebrae were caudal or not. 



In the skeleton of a Tiger-boa {Python tigris) in the museum of the Royal College 

 of Surgeons, anchylosis of the 148th to the 149th vertebra has taken place; and the 

 166th and 167th vertebrae have been more completely and abnormally fused together, 

 so as to appear like a single vertebra on the left side, and a double one on the right 

 side, where there are two diapophyses and two ribs. The compressed form, however, 

 and diminutive size of the two anchylosed vertebrae of Palaojjhis, strongly indicate 

 them to be from near the end of the tail, in which case it must be concluded that that 

 part was compressed, as in the smaller modern liydroplddes, and that the present 

 extinct Ophidian was a Sea-serpent of at least twenty feet in length. 



All the vertebrae with the characters specified in the description of the large 

 specimens from the trunk, and referable to the PaloiopMs Ti/phmis, have been obtained 

 from the Eocene clay at Bracklesham, Sussex : they form part of the collections of 

 the late Frederic Dixon, Esq., F.G.S., of Worthing; of James S. Bowerbank, Esq., 

 F.R.S. ; and of George Augustus Combe, Esq., of Preston, near Arundel, to whom I 

 have been indebted for some beautiful examples, including the two vertebrae in natural 

 conjunction (PI. 3, fig. 28), and the vertebra with the best preserved neural spine 

 (PI. 3, fig. 27.) 



Pal^ophis porcatus, Oioen. Plate 3, figs. 13-15, 18, 20, 21. 



On comparing together eighteen Palaeophidian vertebras of different sizes from 

 Bracklesham, the smallest of the dimensions represented in figs. 14, 15, and thence 

 gradually increasing to the size of the specimen fig. 20, 1 find the following difi"erences : 

 in fio-. 14, e. g. the articular cup and ball at the ends of the centrum are larger in 

 proportion to the length of the centrum, as compared with the next-sized vertebra, 

 fio-. 5 : the under surface of fig. 1 5 is convex transversely between the diapophyses 

 and sends down a short median ridge ; in fig. 6 it is concave at the same part, and 

 withovit the median ridge ; but both vertebrae have the median process or ' hypapo- 

 physis' at the back part of the under surface. In fig. 14 the fore part of the zygosphene 

 is concave, in fig. 5 it is flat ; in fig. 5 the upper border is straight, in fig. 14 it forms 

 an open angle ; the space between the zygosphene and zygapophysis is greater in 

 fig. 5 than in fig. 14. 



