ON BRITISH FOSSIL REPTILES. 85 



corresponds with the second or third of the first series of caudal vertebraa of 

 the Caen PoUnlopleuron figured by M. Deslongchanips. There is one cha- 

 racter in the Wealden vertebra whicli is not mentioned in M. Deslong- 

 chanips' description of the Caen species, viz. a longitudinal sulcus at the 

 middle of the under surface of the bodj^ of the vertebra, at least at its ante- 

 rior half: the sulcus is not deep, and is 1 centimeter or 4 lines in breadtli. 

 The extremity of the under surface seems to be obliquely leveled off to form 

 a single haemapophysial surface. From this structure it might have been in- 

 ferred that the haemapophyses, which by their union form the chevron bone, 

 were closely approximated, if not confluent at their bases, as in the Iguano- 

 don ; the Caen specimen, in which many of the chevron bones are preserved, 

 proves this to have been the case. The fortunate fracture Avhich demonstrates 

 the peculiarly large medullary cavity in the centre of the vertebral body, 

 gives the best proof that could be required of the generic identity of the 

 Wealden vertebra with tl-.e Caen Poihilopleuron ; and the absence of that 

 cavity in the vertebrae of the Megalosaurus, which I have determined by a 

 section of one of the caudal vertebrae, establishes the distinction between that 

 genus and the Poihilopleuron. 



In the form of its sub-biconcave vertebrae, and the simplicity of their 

 neural arch as compared with the Streptospondylus and the Dinosauricms, 

 the Poihilopleuron manifests its closer affinity to the Ccelospondylian Croco- 

 diles. It agrees with the Teleosaurus in the comparative shortness of the 

 fore legs ; the mode of articulation of the vertebral ribs appears to be the 

 same, and there is no evidence that it differs in the structure of the ab- 

 dominal ribs. 



The number of caudal vertebrae would appear to be greater ; but I know 

 not in what material respect the Poihilopleuron resembles the Lizard tribe 

 more closely than does the Teleosaurus, unless it should be proved to have 

 five toes on the hind foot, and to Avant the dermal armour. Subsequent dis- 

 coveries may prove it to belong, like the Megalosaurus, to the Dinosaurian 

 order ; but as the Poihilopleuron is at present known, it seems to have most 

 claim to be received into the Codospondylian family of the Crocodilian 

 order, and perhaps has the closest affinity in that family to the Crocodilus 

 Bollensis, Jaeger (^Macrospondylits, H. v. Meyer). 



To the genus Poihilopleuron it is most probable that the specimen 



No. ^^ in the Mantellian Collection belongs, as it agrees in size, in texture, 

 and especially in the character of the external surface with the caudal ver- 

 tebra last described. As it consists of the annular part or neural arch only, 

 the test of the medullary cavity of the body cannot be applied. It belongs to 

 one of the anterior dorsal vertebrae, and is distinguished by well-marked and 

 peculiar characters from the corresponding vertebrae of the Iguanodon, Me- 

 galosaurus, Hylaosaurus, Cetiosaurus, and Streptospondylus ; and in the chief 

 of these differences it approximates to the sub-biconcave Crocodilian type of 

 vertebrae. As only the caudal vertebrae of the Poihilopleuron appear to have 

 been hitherto discovered, I cannot avail myself of the aid, in the determina- 

 tion of the present fossil, which the able descriptions of M. Deslongchanips 

 afforded in reference to the preceding one ; there remains, therefore, to record 

 the characters of the present fragment of a dorsal vertebra, in order that they 

 may be compared with more perfect ones from the oolite of 3Ialadrerie near 

 Caen, in the event of the remainder of the vertebral column of the Poihilo- 

 pleuron ever falling into the hands of the Palaeontologist. 



The present fossil is imbedded in the ferruginous sand of the Tilgate strata ; 

 its antero-posterior diameter from the extremity of the anterior to that of the 

 posterior oblique process, is 5 inches 4 lines. 



