Zs) Ceratosauria 
Fig. 45 Cladogram of the Theropoda, based on our modified version of Holtz’s data-matrix and including also Baryonyx. 
In all our cladograms (summarized in Fig. 45) Baryonyx is placed 
as the sister-group of the Neotetanurae Sereno ef al., 1994 
(=Avetheropoda Paul, 1988), arising from node 5 (numbering of 
nodes and characters are as used in this work, not Holtz 1994a; for 
details seeAppendix C).The nearest outgroup of the combined clade 
is Megalosaurus, arising from node 4, and the second nearest 
Torvosaurus, arising from node 3. All these taxa (1.e., Zorvosaurus, 
Megalosaurus, Baryonyx and Neotetanurae) together comprise the 
Tetanurae Gauthier, 1986. If Baryonyx were inserted into Holtz’s 
cladogram (his fig. 4) it would split off from the main stem of 
his Hennigian comb, leading to the Coelurosauria sensu 
Gauthier, between his nodes 6 and 7. 
The specific evidence for the position of Baryonyx on our 
cladogram (Fig. 45) is assessed critically below. 
At node 3 (Jorvosaurus dichotomy, Holtz’s node 5, Tetanurae) 
there are fourteen synapomorphies. Baryonyx provides positive 
evidence of only three of those, namely: 
50. Dorsal vertebrae pleurocoelous (“unambiguous synapomorphy’ 
of Tetanurae, fide Holtz loc. cit.). 
4. Dorsal vertebrae with transverse processes that are not strongly 
backturned or triangular in dorsal view (supposedly reversed 
character). 
110. All three pelvic elements not fused together in adults (suppos- 
edly reversed character). 
At node 4 (Megalosaurus dichotomy, Holtz’s node 6, unnamed) 
« 
a) . w = y 
Ses et ne) ees Bin wots 
ey Cees: ow « $ $ s | 
S f ¢ F Seat Se S & 
S S S cy 9 S 5 & $ § S 
g ss RS 2 S S S 9 y G gy 
9 gS 8s o ~~ x 4 y 9 
Oe ee Re ee ee ee 
Sine exQhol C initio; aaeQagean Bie ett mite, Sa 
A.J. CHARIG AND A.C. MILNER 
there are eight synapomorphies, but unfortunately Baryonyx affords © 
evidence of none of them. 
At node 5 (Baryonyx dichotomy, new node, unnamed) there are | 
five synapomorphies, on only one of which does Baryonyx provide 
data, namely: 
96. Coracoid tapers posteriorly (‘unambiguous synapomorphy’ of | 
Holtz). We thought it unwise to propose a new name for the 
taxon based on this node; the phylogeny is not yet sufficiently 
well established. 
This evidence is not very convincing. More helpful is the list of | 
synapomorphies for node 6 (dichotomy between Allosaurus/ 
Acrocanthosaurus and Coelurosauria sensu Gauthier; Holtz’s node 
7; Neotetanurae orAvetheropoda); there are four of them upon which ! 
diagnosis of the Neotetanurae is based. In this case, fortunately, 
Baryonyx provides clear evidence to show that it has none of them: 
105. Axis with spine table. 
117. Manual digit I reduced in length. 
88. Obturator process present. 
92. Pubic boot pronounced (‘unambiguous synapomorphy’ of | 
Neotetanurae, fide Holtz loc. cit.). 
Certain characters of Baryonyx, though used in Holtz’s and our 
analyses and seeming to be apomorphous for that O.T.U., also occur; 
elsewhere in the cladogram. If our suggested phylogeny is correct, | 
