_ BARYONYX WALKERI 
| 
A 
47 
‘Fig 35 Baryonyx walkeri, holotype, BMNH R9951; ungual attributed to left digit I (pollex). A, lateral view; B, proximal. x 0.5. 
“symmetrical, very little compressed, smoothly rounded along its 
j length, and sharply pointed. Grooves for the horny sheath run right 
to the tip, and the flexor tubercle is pinched off more sharply on the 
“medial side than on the lateral. 
A distal end of another large phalanx, broken on one side, is 
Probably part of the first phalanx of the same digit. If so, it is with 
this element that the large ungual articulated. It appears to be a left 
ohalanx, and it compares well with the articulated phalanx described 
‘mmediately below as the first phalanx of left digit II or the second 
of left digit III. It has a deep groove between the two condyles, the 
nedial condyle projects obliquely and has a very distinct ligament 
sit, and the lateral condyle is broken off almost entirely. 
| 
_EFT DIGIT II OR III (Fig. 36). The three phalanges that make up 
his digit were found in articulation; they might be either an entire 
igit II, or a digit III lacking its basal phalanx. We can try to resolve 
his dilemma by comparing the bones with the corresponding ele- 
ments of Allosaurus and other theropods. The greater length of the 
phalanges suggests digit II. On the other hand, the proximal end- 
surface of the most proximal phalanx of the three (II, or Il =) does 
10t have the somewhat irregular, single-facetted form characteristic 
>f a basal phalanx (which would articulate with the distal end of 
etacarpal II) but has instead the double-facetted form typical of a 
Ton. basal phalanx (which would articulate with the pulley-like 
listal end of phalanx III,). These elements, in most respects, are 
: tandard for a theropod; the ungual, unlike its neighbour on digit I, 
not enlarged, its length being estimated (it lacks its tip) at 165 
. The middle phalanx of the three is appreciably shorter (91 
m) than the most proximal (132 mm). This contrasts with the 
Ondition in digit II of Allosaurus, where the second phalanx is 
yery slightly longer than the first (Madsen 1976, pl. 45), and in 
leinonychus, where it is considerably longer (Ostrom 1969: 104). 
gain, the elements are almost perfectly symmetrical, but a few 
inor asymmetries indicate that we are here dealing with a finger 
f the left hand. The most pronounced asymmetry is seen in distal 
fiew, where the lateral profile is almost parallel to the sagittal plane 
ut the medial profile is inclined obliquely. 
GHT UNGUAL ITORIII. This element is poorly preserved, it lacks 
{s distal half and its proximal end is badly damaged. Nevertheless, 
} is apparent that it is a damaged mirror-image of the ungual 
attributed to left digit II or III, and is therefore identified as the 
corresponding right ungual. 
RIGHT DIGIT III ORIV. One of the smaller phalanges (length 65 
mm) is presumably from a digit III or IV, and the obliquity of its 
medial profile in dorsal aspect indicates that it comes from the right 
side. 
Pelvic girdle 
ILIUM (Figs 37, 38). Only the right ilium is preserved. The remains 
consist of four substantial unconnected fragments of bone and one 
natural mould: 
1. The pubic peduncle, with the anterior third of the acetabulum. 
2 The ischiadic peduncle, with the posterior two-thirds of the 
acetabulum (there can be no more than a narrow gap between this 
and fragment no. 1). 
3.A large part of the anterior portion of the iliac blade, albeit with no 
natural edges. 
4. Part of the posterior process, with the brevis shelf and brevis fossa, 
and a large part of the iliac blade dorsal to it. 
5. A mould of the medial surface of a large part of the iliac blade, 
lacking the anterior end; the original natural mould, formed on 
the rock which underlay the blade in the field, has been repro- 
duced in plaster of Paris via a silicone pull. 
The broken end of fragment no. 4 certainly joins on to fragment 
no. 2, but the loss of what appears to be a few millimetres of 
material prevents a firm connexion between them. The information 
afforded by all of the above permits us to describe the ilium as a 
whole. 
The estimated length of the ilium, as preserved, is 835 mm. The 
material available suggests that the dorsal outline of the blade was a 
smooth convex curve. The length and precise form of both the 
anterior and posterior processes are unclear, but the posterior proc- 
ess seems to have tapered posteriad to a slightly pointed, spatulate 
termination. This process bears three diverging ridges separated by 
three concavities, i.e. it is triradiate in posterior view. Thus, we have 
a concave lateral surface, a concave medial surface, and a deeply 
