BARYONYX WALKERI 
mxX.S 
Portion bearing four prominent longitudinal ridges. Just below the 
orsal margin of the medial face lie two large, deep fossae, one 
Above the septum between the 3” and 4" alveoli and the other above 
he 6" alveolus. Comparison with other theropod skulls suggests that 
these may have housed maxillary sinuses. 
The most anterior part of the medial face is in fact an anteriorly 
jlirected flange (on to the lower half of which run the longitudinal 
tidges already mentioned). The anterior tip of this flange articulates 
with a notch on the lateral surface of the anterior process of the 
palatine (q.v.). The anterodorsal border of the flange forms the 
edial wall of the internal portion of the subnarial canal. 
Seen from below, the anterior/ventral surface of the maxilla is of 
ore or less uniform width. The lateral half of this is occupied by the 
ental alveoli; the medial half consists posteriorly of the stout 
unded ridge referred to above, tapering forwards slightly; anterior 
the level of the 4" maxillary tooth this ridge tapers further into the 
in medial flange and a less prominent shelf on the inner wall of the 
ange immediately bordering the medial walls of the alveoli. Also 
n the medial wall of the flange, behind the shelf, a bony wall 
Jurrounds what may well be another maxillary sinus. 
‘| On the anterior surface of the rectangular peg is a vertical depres- 
“ion, behind which is the first maxillary alveolus. Altogether there 
€ seven alveoli in the preserved portion of the maxilla; the first four 
e almost contiguous, separated from each other by only a thick 
ny wall, but from no. 4 backwards they are more widely spaced. 
‘\lveoli nos. 2, 3 and 4 are the largest; nos. 1, 5 and 6 are smaller, no. 
7 smaller still. The alveoli are subcircular; no. 4 contains a com- 
pletely erupted, slightly worn tooth, no. | contains only the base of 
a smaller, broken-off tooth only partly erupted, and no. 6 shows the 
tip of a newly erupting crown. 
We suggested (Charig & Milner 1986, 1990) that there was a 
mobile articulation on each side between the premaxilla in front and 
the maxilla behind. We now believe that that suggestion was prob- 
ably wrong. The anterior end of the maxilla can be fitted neatly into 
the inverted trough formed laterally by the premaxilla and, medially, 
by the thin sagittal lamina which appears to be an anterior prolonga- 
tion of the paired vomers; the rectangular peg on the maxilla then 
locks into the notch on the premaxilla (Charig & Milner 1986, fig. 1; 
1990, fig. 9.2). This complex articulation between premaxilla and 
maxilla would not permit any movement between those elements, 
for which, in any case, there is no obvious functional requirement. 
NASAL (Fig. 4). The paired nasals are fused together into a single 
element, although it is possible to detect intermittent traces of the 
median suture. Three fragments were recovered: the major, poste- 
rior, fragment (Fig. 4) was found lodged against the left lacrimal. As 
preserved, the posterior fragment in dorsal view resembles an 
arrowhead: it is a narrow triangle, with its apex directed forwards, 
and from the base (hind end) of this triangle a stubby median process 
— the ‘shaft’ of the ‘arrowhead’ — projects farther posteriorly and 
must have articulated with the frontals. 
The dorsal (external) surface of the conjoined nasals is raised in 
