A.J. CHARIG AND A.C. MILNER 
Fig. 43. Baryonyx walkeri, holotype, BMNH R9951; right calcaneum. A, proximal view; B, lateral; C, medial; D, distal. x 0.5. 
OTHER MATERIAL REFERRED TO THE 
SAME FAMILY 
Taquet (1984) described two fragments from the Elrhaz Formation 
(Aptian) of Gadoufaoua, Niger, as mandibular symphyses of a 
spinosaurid (Musée National d’Histoire Naturelle, Paris, MNHN 
GDF 365 and 366); they were refigured more clearly in Kellner & 
Campos 1996 (fig. 7). Each is virtually identical to the conjoined 
premaxillae of the holotype of Baryonyx (Charig & Milner 1986; 
1990: 139) except in that they possess seven alveoli on each side, not 
six on the left and seven on the right as in R9951. We consider that 
these snouts, despite their much younger age, are referable to 
Baryonyx sp. indet. 
More recently, Viera & Torres (1995: figs 1-2) referred a left 
maxilla fragment (Museo de San Telmo, San Sebastian, GA-2065) 
from the Enciso Group, Barremian of Igea, La Rioja, Spain, to 
Baryonyx walkeri. It is less complete and about 75% of the size of 
the holotype maxilla but otherwise indistinguishable. The tip, ante- 
rior to the 3rd alveolus is missing but complete 8th, 9th and partial 
10th alveoli are present confirming that the pattern of evenly spaced 
subcircular alveoli, gradually decreasing in size, continues further 
posteriorly than is preserved on the holotype maxilla. 
The following isolated tooth crowns are referred to cf. Baryonyx; 
their fragmentary nature and the consequent lack of further informa- 
tion precludes a more precise identification. 
Wessex Formation (=Wealden Shales, Barremian), Isle of Wight 
(Isle ofWightCounty Museum, SandownIWCMS 3642 and5120from 
Hanover Point; IWCMS 5122, IWCMS 1995 207-209, University of 
Portsmouth, UOP.97, all unlocalised) described as possible 
baryonychid by Martill & Hutt (1996). All these crowns are virtually 
identical with the teeth of the holotype R995 | intheirshape, ornamental 
pattern, enamel texture, and possession of finely serrated anterior and 
posterior carinae with seven or eight denticles per millimetre. } 
Upper Weald Clay (Barremian), former Ewhurst Brickworks, | 
Surrey (National Grid ReferenceTQ 108379), a single crown (Maid- 
stone Museum MNEMG 1996.133) collected by Dr E A 
Jarzembowski in the mid 1980’s and reported recently from just 
below BGS Bed Sc, equivalent to the top of the Smokejacks beds. 
Ashdown Sand, (Hauterivian) at Redlands Bricks, Ashdown 
Works, Scallets Wood, Turkey Road, Bexhill-on-Sea, East Sussex, 
National Grid reference TQ60/70 721097, a single crown (Bexhill | 
Museum BEXHM:1993.485), collected by Mr D. Brockhurst in | 
1993. This crown differs from the Barremian material only in that i 
the carinae do not extend the full distance to the base of the crown. | \ 
PHYLOGENETIC RELATIONSHIPS AND 
SYSTEMATIC POSITION 
Our first publication on Baryonyx (Charig & Milner 1986) made no / 
suggestions as to the relationships of the genus, beyond the claim thatit | - 
‘was a typical large theropod in certain respects, resembling, for i 
example, Allosaurus’. We nevertheless considered it to be sufficiently | 
distinctive to merit designation as the type-genus of a new family, } 
Baryonychidae. Our second, more detailed publication (1990) con- | 
firmed our view of its separate family status; our only other conclusions 
were that it was not a spinosaurid and that it could not be fitted into 
Gauthier’s (1984, 1986) cladistic classification of the theropods. 
Other workers have referred to Baryonyx in various contexts. | 
Their assessments of its phylogenetic position were made in the light | 
of one or both of our preliminary descriptions, which were incom- 
plete and contained some errors. Some of those assessments, because 
