40 
Fig. 27 Baryonyx walkeri, holotype, BMNH R9951; sacral/proximal caudal vertebrae. A, neural spine in left lateral view, B, proximal caudal centrum 
(aB) in left lateral and dorsal views; C, fragment of neural arch of proximal caudal vertebra, in left lateral and posterior views. x 0.25. 
CAUDAL REGION (Fig. 27). This region is represented only by the 
remains of six centra and one neural arch, together with two incom- 
plete neural spines that almost certainly pertain to proximal caudals 
(or possibly sacrals). 
Three isolated centra, though large, show no signs of any 
parapophyses, diapophyses, buttressing laminae, sacral ribs or 
pleurocoels. Further, the neural canal running along the dorsal 
surface of each is very narrow. Taken together, these characteristics 
suggest that the centra represent proximal caudals. On one of them 
(referred to below as CaB) the ventral rim of the posterior face is not 
curved but flattened, as is characteristic of proximal caudals. 
There is no way in which the order of these three proximal caudal 
centra could be determined; indeed, there were probably more than 
three altogether. We refer to them here simply as Caudal A (CaA), 
Caudal B (CaB) and the partial Caudal C (CaC), the lettering having 
no significance as to their place in the vertebral succession. 
These centra are of similar diameter to that of the last dorsal, D14, 
but are much longer; indeed, they are very much longer (134, 144 
and 140 mm respectively) than any of the presacral centra (longest 
preserved is the eighth cervical at 120 mm). Their available 
measurements are listed in Table 1 (p. 30). They are very strongly 
‘waisted’, with moderately amphicoelous ends; the anterior face is 
vertically oval (contrast the dorsal centra) and the posterior face is 
more or less circular. Two of them are reasonably complete up to the 
virtually straight neurocentral suture and still possess much of the 
rugose sutural surfaces for the neural arch; the third is only partial, 
lacking most of the posterior end. 
Fragments of three other caudal centra are preserved. The largest 
consists of the posterior part of a centrum, showing an almost 
complete articular face measuring 65 mm transversely and an esti- 
mated 76 mm vertically. The middle of the face is somewhat 
hollowed out and its ventral portion bears a large and slightly eroded 
haemapophysial facet. Enough of the body of the centrum remains 
to show that it was hollow and strongly waisted and had a somewhat 
flattened ventral surface. 
A.J. CHARIG AND A.C. MILNER 
ipoz.fos 
The other two fragments are too poorly preserved to merit | 
description, but are recognisable as caudals by their small size and |) 
flattened ventral surfaces. 
An isolated, incomplete, but well-preserved neural arch (Fig. 
27C) was found, which can only represent a proximal caudal 
vertebra, but it does not fit on to any of the proximal caudal centra 
described above. The posterior edge of its transverse process is well | 
separated from the anterior end of the postzygapophysial facet; in 
this it resembles the anterior dorsals (for example, D3) and is quite 
unlike the posterior dorsals, in which those two structures are 
contiguous. On the other hand, in other characters it is not unlike the 
posterior dorsals. These are the length of the fragment, which, at 
approximately 100 mm, is much longer than the total length of the | 
centrum in D3-D8; the infradiapophysial laminae, which are well 
developed and enclose a deep infradiapophysial fossa; and the | 
powerful hyposphene, which projects farther posteriorly than the 
postzygapophyses themselves. In any case, the total width across the 
postzygapophysial facets is only 52 mm, much less than in any 
dorsal vertebra and they lie in a more or less horizontal plane. Also 
to be noted is the fact that the ventral margin of the arch, which 
appears to be complete and to represent the neurocentral suture, 
curls upwards at the back like the prow of a boat. (For nearest} 
comparison see Welles on the proximal caudals of Dilophosaurus, 
1984: 124 ff). 
A very large, flattened plate of bone, undoubtedly a neural spine, ; 
was found (Fig. 27A). It is much larger than any of the neural spines | 
of the dorsal vertebrae as preserved. Another, smaller and less; 
complete fragment of similar form was also found. The anteroposte-} 
rior length of the larger spine is 142 mm, which is much the samey, 
length as the two longer proximal caudal centra (CaB 144 mm, CaC 
140 mm) and is therefore very much longer than any other centrum 
preserved. Both spines were found in Block 43, from the pelvic 
region of the skeleton, and both possess a characteristic projection) 
(see below) from the posterior or posterodorsal border, not unlike) 
that described above for vertebra D14 (the last dorsal). These) 
h 
| 
