1?B0E. H. G. SEELEY ON PELOBATOCHELYS BLAKII ETC. 235 



in the fossil, on the other hand, they are as broad as in Pleuro- 

 sternon, and yet, owing to the great width of the carapace, the free 

 end of the rib, which is flattened above and coarsely striated in length, 

 extends beyond the epipleuron for If- inch — though as this thin element 

 divides, it shows the rib beneath, which is exposed for a length of 3^ 

 inches. The fifth neural plate is the longest ; and from that one the 

 neural plates become shorter to both ends of the carapace ; but behind 

 the sixth neural plate the neural arch ceases to be attached to the cara- 

 pace. But the character which seems to me most typical of the fossil, 

 and best to justify its claim to generic distinction, is the singular 

 sagittal margins, in which the vertebral scutes meet each other. 



The nuchal plate is rather more than an inch long, and joins the 

 adjacent plates in the usual way, though the British-Museum speci- 

 men appears to show that the nuchal and first neural plates had a 

 squamous union. 



The nuchal plate is imperfectly preserved and gives no indication 

 of its original form, its greatest transverse width being only 1-| inch. 

 Its anterior margin is straight, compressed, and sharp, bordered on 

 the external surface by a slight ridge which extends backward for 

 less than a quarter of an inch, marking what is possibly the narrow 

 nuchal scute, which is separated from the first vertebral scute by a 

 series of tubercles. 



The next fragment, comprising the first three neural plates and 

 portions of the six adjacent costal plates, is four inches long. The 

 neural plates increase in thickness from before backward. 



The first is of oblong form, lg inch long, J inch wide in the middle, 

 and rather less at the two ends. It is convex from side to side. 



The second plate is 1^ inch long, with the anterior margin, which 

 meets the first plate, slightly concave transversely ; and the posterior 

 margin, which meets the third neural plate, convex transversely. 

 Owing to short lateral borders, -^ inch long, which meet the first costal 

 plates, this second neural plate attains its greatest transverse width of 

 ±f inch near the anterior end, while the sides converge posteriorly to 

 ^inch. ^ 



The third neural plate is slightly worn at the hinder end ; it is 1| 

 inch long, joins the second pair of costal plates by anterior shoulders 

 | inch long, which widen the bone to more than an inch anteriorly, 

 while the sides for the third pair of costal plates converge posteriorly 

 to a width of about -^ inch. This plate is T 7 ^ inch thick and marked 

 by a sharply angular median ridge. The posterior | inch of this 

 plate is excavated, as I have already remarked, but by the removal 

 of an equal amount of bone from each side of the plate, so that its 

 angular character is preserved. 



Of the left costal plate, only narrow fragments remain ; but of those 

 on the right side the first is nearly 1^ inch long. On the underside 

 of the plate the head of the rib inclines a little backward. The plate 

 is remarkably thin, and only preserved for an inch transversely. Like 

 the adjacent first neural plate, it is crossed transversely below the 

 middle by the posterior border of the first vertebral scute, which is thus 

 shown to be 1 1 inch long; its width is not indicated. This bounds 



