' BARYONYX WALKERI 
/ of this inadequate basis and because of their brevity, scarcely merit 
discussion. Nevertheless the authors concerned (including ourselves) 
‘have generally considered Baryonyx to be a close or not-so-close 
Telative of Spinosaurus, so that the two genera have often been 
placed together as a single family (Spinosauridae Stromer, 1915). 
This has led certain authors (notably Sereno ef al. 1994) to employ 
that family as a single unit for the purpose of character distribution 
analysis. 
Bonaparte (1991: 18) suggested a relationship between the 
Spinosauridae, later fragmentary theropods from Africa 
(Carcharodontosaurus and Bahariasaurus), and the South Ameri- 
can Abelisauridae; he claimed that the femur of all these forms 
possessed an anteromedially directed head and a plesiomorphically 
low lesser trochanter and, if this were true, then they would together 
represent a major diverse clade (named Neoceratosauria by Novas in 
1991), distributed mainly in Gondwana. Holtz (1994a: 1105) cited 
Bonaparte’s work, but misleadingly stated that the Spinosauridae 
comprised Spinosaurus and Baryonyx; the latter genus, in fact, was 
never mentioned by Bonaparte. Holtz also questioned Bonaparte’s 
suggestion, quoting our claim (1990) that Baryonyx possessed the 
tetanuran synapomorphy of an obturator process on the ischium; we 
now know that to be incorrect — the structure is only a flange. 
Nevertheless, we agree with Holtz because Baryonyx has the femo- 
ral head directed medially, not anteromedially, and its other characters 
are tetanuran rather than neoceratosaurian (see below). Meanwhile 
Sereno ef al. (1996), on the basis of new material from the 
Cenomanian of Morocco, have interpreted Carcharodontosaurus as 
a member of the Allosauroidea. 
Elzanowski & Wellnhofer (1992, 1993) proposed a monophyletic 
group of theropods consisting of Baryonyx, Spinosaurus, the 
Troodontidae, Lisboasaurus (tentatively identified as a manuraptoran 
by Milner & Evans in 1991 but reidentified more recently as a 
crocodilomorph by Buscalioni et al. 1996), and Archaeornithoides 
(a tiny fragmentary skull from the Late Cretaceous of Mongolia, 
described by Elzanowski & Wellnhofer in 1992). This suggestion 
was based on the following shared characters: 
1. Interdental plates absent. 
2. Paradental groove present (separating interdental septa from 
lingual wall of dentary). 
3. Lingual wall of dentary lower than labial wall (Baryonyx, 
Spinosaurus and Archaeornithoides only). 
55 
Fig. 44 Baryonyx walkeri, holotype, BMNH R9951; reconstruction of entire skeleton, x 0.020. 
4. Details of articulation between premaxilla and maxilla (Baryonyx 
and Archaeornithoides only). 
5. Ridge in midline of anterior part of palate (Baryonyx and 
Archaeornithoides only). 
Elzanowski & Wellnhofer suggested that Archaeornithoides was the 
closest known theropod relative of birds, and that its consistent 
similarities to Baryonyx, Spinosaurus and the troodontids narrowed 
down the ancestry of birds to those theropods that lacked interdental 
plates and possessed paradental grooves. However, interdental plates 
are unequivocally present in the dentary of Baryonyx (Figs 13, 14), 
forming a barrier between the alveoli and the paradental groove. 
Further, since Baryonyx does not possess any apomorphous charac- 
ters of the Manuraptora' (in which taxon is placed the Troodontidae), 
both it and Spinosaurus must be far removed from the origin of 
birds. Another argument against the hypothesis of Elzanowski & 
Wellnhofer is that Archaeopteryx too had interdental plates (as 
shown by the dentary of the seventh specimen in lingual view, 
Elzanowski & Wellnhofer 1995). The same is true (Currie 1995) of 
Dromaeosaurus, type-genus of the family Dromaeosauridae, which 
many authors (most recently Holtz 1994a) regard as the sister-group 
of Archaeopteryx. 
Claims of relationship at family level 
A minor misunderstanding has arisen over the systematic relation- 
ship between (a) Baryonyx, (b) Spinosaurus, and (c) a partial maxilla 
from the Upper Cretaceous of Morocco, described and figured by 
Buffetaut (1989) as Spinosaurus. The maxilla of the Spinosaurus 
holotype is known only from a poorly preserved fragment now lost 
(see below p. 56). Buffetaut was therefore obliged to refer the 
Moroccan maxilla to Spinosaurus on the basis of the only structures 
common to both specimens, namely the teeth and their alveoli 
(using, for Spinosaurus, dentary teeth and alveoli and isolated 
teeth). Buffetaut claimed also that Baryonyx must be closely related 
to Spinosaurus (basing his evidence both on the original Spinosaurus 
'The name ‘Maniraptora’, proposed by Gauthier (1986: 30) and used by several 
workers since then, is etymologically wrong. Manus (which is Latin, not Greek as 
stated by Gauthier), meaning hand, is a fourth-declension feminine noun with the root 
manu- (see International Code of Zoological Nomenclature, 3rd edition, 1985: 215), 
not man- as Gauthier obviously supposed. His replacement of the English adjective 
manual with ‘manal’, op. cit., is presumably based on the same incorrect supposition. 
