BARYONYX WALKERI 
The foramen magnum is semicircular, with a more or less straight 
lower border and a curved upper border. The lower border lies 
immediately above the robust, backwardly projecting occipital con- 
dyle. Most of the backwardly directed, articulating face of the 
condyle is formed by the basioccipital; the exoccipitals make only a 
comparatively minor contribution to this face, namely the dorso- 
lateral corners, each of which is demarcated from the basioccipital 
by a straight suture running obliquely. (This suture has separated on 
the left-hand side.) The dorsal surface of the occipital condyle 
between the exoccipital sutures on either side is also formed by the 
basioccipital, which bears a shallow longitudinal depression in the 
midline; this means that the condyle, seen from behind, has a 
roughly heart-shaped appearance. 
A vertical apron of bone, also part of the basioccipital, extends 
ventrally from beneath the condyle; it is of uniform width, as wide as 
the condyle itself, and terminates below in a broad median tongue 
with a shorter, much narrower lappet on either side. The basisphe- 
noid projects even farther ventrally; this has parallel sides for some 
distance and then terminates in a pair of processes, the basipterygoid 
processes of the basisphenoid, which splay out ventrolaterally. The 
lower margin of the basisphenoid, between the splayed processes, 
forms a smooth concave curve. The posterior surface of the basi- 
sphenoid, below the median tongue of the basioccipital, is deeply 
furrowed in the midline; the furrow becomes broader and shallower 
ventrally and peters out altogether before reaching the ventral 
margin. 
On either side of the foramen magnum/occipital condyle complex 
lies a horizontally directed lateral wing, the paroccipital process. 
This is more or less triangular in transverse section, with one face 
directed posteriorly, another face dorsally and a little anteriorly, and 
a third anteroventrally. Its medial part is formed by the exoccipital 
and its more lateral part by the opisthotic; in most dinosaurs those 
wo elements are fused together indistinguishably, but in this speci- 
en an ill-defined suture still demarcates the exoccipital as a slender 
riangle with its apex directed laterally. The opisthotic extends also 
entrally, lateral to the whole length of the basioccipital. At about 
id-height of the occipital condyle, on the suture between the 
xoccipital and the opisthotic and a short distance lateral from the 
asioccipital, lies a fairly large foramen which is divided in its depth 
nto two separate canals. Recent Crocodylia possess several 
oramina in the same general region, one of which, with exactly the 
ame topographical relationships as the foramen in Baryonyx, is 
ikewise divided into two and is unequivocally the vagus foramen; 
e therefore give that same identity to the ‘double’ foramen of 
aryonyx. lordansky (1973: 226) observed that in the crocodilian 
oramen vagi ‘The medial canal extends to the cerebral cavity and is 
aversed by the IXth and Xth nerves; the lateral canal extends to the 
iddle ear cavity and contains the Ramus communicans (N. 
ympathicus) connecting the VIIth and [Xth nerves.’ [actually Xth 
nd XIth; see Romer 1956: 66]. In Baryonyx the external aperture of 
e smaller canal is dorsomedial in relation to its larger, ventrolateral 
eighbour. There are two internal apertures within the cavity of the 
rain stem, just anterior to the foramen magnum; the lower one, 
ying at the junction of the floor and wall of the cavity, connects with 
‘he larger, ventrolateral external foramen, while the upper one, lying 
ehind and a little above its partner, probably connects with the 
maller, dorsomedial external foramen. Whether these two are pre- 
isely homologous to their crocodilian counterparts is open to 
uestion. Meanwhile the dorsomedial part of the anteroventral face 
f the paroccipital process is overlapped by the prootic (q.v.) below 
nd the supraoccipital (q.v.) above. 
The left paroccipital process, i.e. exoccipital plus opisthotic, is 
resent but has been cleanly detached from the rest of the skull. It 
21 
lacks the ventral extension and has been distorted to some degree. 
However, laterally it is more complete than the paroccipital process 
of the right side, and therefore gives a better indication of the true 
shape and extent of this structure, of which about 25 mm is missing 
on the right side (see also Charig & Milner 1986, fig. 2; 1990, fig. 
9.4). 
The dorsal part of the occiput is contributed by the supraoccipital, 
which forms (a) the central region of the dorsal margin of the 
foramen magnum, between the medially directed arms of the 
opisthotics; (b) the central part of the back end of the skull roof; and 
(c) arising from the latter, a prominent stout process which, though 
essentially projecting dorsally, is also inclined a little forwards at its 
upper end. This process, somewhat compressed antero-posteriorly 
and is rectangular in shape when viewed from front or rear; its dorsal 
surface is broadly crescentic. On its posterior surface, on either side 
of its base, is the external aperture of a canal that has its other end in 
the upper part of the internal wall of the braincase and presumably 
served for the passage of a large blood vessel. The basal part of the 
anterior side of the supraoccipital process, together with the dorsal 
part of the underlying prootics, forms on each side a strongly 
furrowed surface which must have sutured with the parietal. 
The prootic on either side is a solid, chunky element, forming the 
wall and floor of this posterior part of the braincase; each meets its 
fellow along the midline of the floor. A row of foramina for the 
passage of the cranial nerves runs along the angle between the wall 
and the floor, effectively dividing the internal and external surfaces 
of the prootic into an upper part and a lower part (see below, next 
paragraph). Anteriorly each prootic bears furrowed surfaces for 
articulation with the laterosphenoid and, more dorsally, with the 
parietal. Between those articulating surfaces, below the floor of the 
braincase, is the posterior part of the pituitary fossa; it appears here 
as two deep median concavities (a smaller above and a larger below, 
with a saddle between) which penetrate horizontally backwards into 
the substance of the bone. The posterior wall of the pituitary fossa is 
the dorsum sellae. On either side of the top of the upper concavity is 
a large foramen, through which the VIth (abducent) cranial nerve is 
presumed to have passed forward from the floor of the braincase into 
the pituitary fossa (see Romer 1956: 67) and thence continued in the 
same direction to emerge laterally into the orbit (see Osborn 1912: 
17). 
Externally the upper part of the prootic is applied to the anterior 
surface of the opisthotic, extending about half-way along the length 
of the paroccipital process. Its lower margin is a smooth, slightly 
concave curve, separated from the opisthotic behind by a deep 
furrow which terminates medially in the fenestra ovalis. Just 
anteromedial to the latter is a foramen for the VIIth cranial nerve 
and, farther forward still, a large open notch for the emergence of the 
Vth. Ventral to these nerve exits the lower part of the prootic forms 
a large plate-like posterolateral process which is applied to the 
anterolateral face of the basisphenoid. 
On the left side, the opisthotic has become detached from the rest 
of the braincase; this exposes the cavity of the otic capsule, con- 
tained within the body of the prootic. Inside this capsule are two 
small foramina, both of which lead through into the endocranial 
cavity to emerge posterior to the large foramen for the passage of the 
Vth nerve. One of these could well be the duct for the passage of the 
VIlIth cranial (auditory or acoustic) nerve. These ducts and 
foramina serve for the passage, not only of blood vessels and lymph 
ducts, but also of pneumatic openings. 
The whole of the ventral portion of this occipital fragment, seen in 
anterior view, is formed from the basisphenoid with its splayed, 
downwardly projecting basipterygoid processes. As stated above, 
the more dorsal part of the basisphenoid is hidden by the postero- 
