ON BRITISH FOSSIL REPTILES. 87 



(presumed) vertebra of the Poikilopleuron is 2 inches 2 lines. The length 

 of the transverse process is 4 inches. The vertical diameter, or thickness of 

 the transverse process, where unsupported, is from 2 to 3 lines. It is obvious, 

 therefore, that this long, thin, lamelliform plate of bone must need further 

 support, in order to sustain the rib which is appended by its tubercle to the 

 extremity ; and the requisite strength is here given precisely as the carpenter 

 supports a shelf by a bracket. The bracket-like process is a vertical tri- 

 angular plate of bone, the breadth or depth of which, at its origin, is 1 inch 

 4 lines, and which gradually diminishes in depth and increases in thickness 

 as it extends along the middle of the under part of the transverse process, 

 until it is finally lost near the extremity of the process, which here has ex- 

 changed its lamellar for a prismatic form, terminating in the obtuse extremity 

 against which the tubercle of the rib abutted. The supporting bracket is 

 not quite vertical, but inclines a little forwards, and behind it there is a deep 

 angular fossa. The posterior oblique processes diverge from each other and 

 from the neural arch immediately above the posterior extremity of the spinal 

 canal : each articular surface, which is directed downwards and outwards, 

 forms, as it were, the base of a posterior root of the spinous process, which is 

 convex externally, diminishing in breadth as it converges to meet its fellow 

 at a very acute angle above a deep fissure extending forwards into the sub- 

 stance of the base of the spine, similarly to the fissure before described as 

 extending backwards from the opposite part of the spine into its substance. 

 As far as I could detach the matrix, these fissures extended so that they seem 

 to communicate, and the neural arch to be perforated by two longitudinal 

 passages, one for the spinal chord, and the other, running above and parallel 

 with the former, througli the base of the spinous process. This process is 

 thus partially separated at its base into two laminte, and presents a structure 

 which almost realizes Prof. Geoffroy's theoretical idea of the essential nature 

 of a spinous process, viz. that it consists of two elementary laminae, which in 

 fishes are superimposed one on the other, but in other Vertebrates are placed 

 in juxtaposition. 



The anterior parts of each spinal plate are thickened and rounded, like 

 those behind, and extend to the origins of the anterior oblique processes. The 

 diameter of this remarkable spinal fissure is from 4 to 3 lines. It is present in 

 an inferior degree in the Teleosaurus, but not in the vertebrae of the Igua- 

 nodon, Megalosaurus, or other Dinosauria. 



The base of the spinous process in this (presumed) Poikilopleuron's ver- 

 tebra, instead of descending from behind forwards in a graceful curve, as in 

 the Dinosaurs, forms a straight and almost horizontal line, 3 inches in extent : 

 the spine maintains the same breadth to its summit, which is truncated rather 

 obliquely; its height is 4 inches 9 lines, measured from the upper end of the 

 posterior oblique processes ; it is thickened and rounded at its truncate sum- 

 mit. The heiglit of the spine of a corresponding vertebra of the Iguanodon, 

 with a centrum of the same length, is 9 inches. Thus the present vertebra 

 more resembles, in the form and proportions of its spinous process, as in other 

 characters, the vertebrae of the Crocodilians. 



The posterior part of the neural arch, with the spinous process of the ver- 

 tebra here described, is figured in Dr. Mantell's ' Illustrations of the Geology 

 of Sussex,' pi. xii. fig. 1, as the ' Lumbar Vertebra of the Iguanodon.' It is 

 unquestionably not a lumbar vertebra ; and if it does not belong to the Poi- 

 kilopleuron, it indicates an unknown genus of Crocodilians. 



In a collection of fossils belonging to S. H. Christie, Esq., from the sub- 

 merged Wealden beds near the Isle of Wight, there is one half of the cen- 

 trum of a dorsal vertebra from Brook Bay, which agrees in size, in the form 



