46 
Fig. 34 Baryonyx walkeri, holotype, BMNH R9951; left ulna. A, lateral view; B, posterior; C, medial; D, anterior; E, proximal; F, distal. x 0.25. 
with each other, of which the most distal is a much smaller ungual; 
and another four, less complete phalanges. Also preserved are a few 
more small phalanges which are too incomplete and indeterminate 
to merit description. 
We assume that Baryonyx possessed the typical theropod phalan- 
geal formula of 2: 3: 4: (1 vestigial or 0) : 0. 
Before describing these phalanges it was desirable to determine: 
1. Whether they belonged to the manus or the pes. Non-ungual 
phalanges from the theropod manus are generally smaller than 
those of the pes; this comparison, however, requires manual and 
pedal phalanges from the same individual. Unguals from the 
manus are more strongly curved than those on the pes. When non- 
unguals and unguals are found together in articulation, these two 
indicators should reinforce each other. The precise location of the 
phalanges within the excavation is another source of helpful 
information; for example, the only articulated digit found was in 
Block 27b, which, lying at the head end of the ‘dig’ (see Fig. 49) 
and containing also part of the left scapula, was more likely to 
contain parts of the manus than the pes. 
2. Which side they belonged to, left or right. Recognition of which 
side of the phalanx is medial and which is lateral enables us to 
determine whether the element in question is from the left side or 
the right. A study of theropod phalanges illustrated in the litera- 
ture (e.g. Allosaurus figured by Madsen 1976, pls 43 and 44) 
shows that the lateral profile of each non-ungual phalanx is 
almost parallel to the sagittal plane but the medial profile is 
A.J. CHARIG AND A.C. MILNER 
inclined obliquely, in such a way that the distal breadth is |) 
significantly broader than the proximal. Further, a longitudinal |) 
groove divides the distal end into two condyles in such a way that |)! 
the lateral condyle is slightly wider than the medial. As for the | 
qu 
i 
rated by a ridge which begins in a dorsal swelling and then runs 
ventrally; the swelling and the ridge are a little closer to the lateral | 
towards the lateral side. 
3. Which digit they belonged to and their position within the digit. | 
Evidence of this nature suggests that all the well-preserved phalanges | 
are of the left manus. 
considerably longer.) The slightly eccentric position of the dorsal 
swelling between the twin facets of the phalangeal articulation 
suggests that this ungual may be attributed to the left side. The arc of | 
curvature is larger than in an Allosaurus of comparable size, i.e. itis 
less strongly curved. Otherwise this element is a fairly typical} * 
theropod ungual of average proportions, almost perfectly bilaterally | ' 
