ENGLISH COMPSOGNATHIDS

[darren.naish@port.ac.uk]

Recently there has been some discussion of what characters might 
unite a monophyletic Compsognathidae. This is something I've been 
looking into for my thesis as I am fairly confident that the holotype 
of _Aristosuchus pusillus_ (Owen 1876) (BMNH R178, a partial sacrum 
and pubes) is a compsognathid. Chen et al. (1998) diagnosed 
compsognathids on the basis of 'fan shaped' neural spines and some 
other characters, all of which are missing from BMNH R178. 

However, the well-preserved pubic boot of the specimen is pretty much 
identical to that of MNHN CNJ 79, the _Compsognathus corallestris_ 
holotype, and also recalls that seen in _Sinosauropteryx_ and SMNK 
2349 PAL, a new specimen from Brazil (Sues et al., in press). I 
suggest that a unique kind of pubic boot morphology might be a 
compsognathid synapomorphy: namely, one that has a reduced or absent 
cranial process, extends caudally to a marked degree, and has an 
almost horizontal ventral margin. Chen et al. (1998) I think 
described _S. prima_ as having a cranial process, but in their 
diagrams and photos the process is tiny. Also, the reconstructions of 
the BSP AS I 563 pubic boot (this is the _C. longipes_ holotype) vary 
widely according to which author you consult. Having examined a good 
cast (BMNH R49159) and high quality photos, I think that Ostrom 
(1978) made the _C. longipes_ boot too short craniocaudally, and 
added a cranial process where there isn't one. Instead, the _C. 
longipes_ boot apparently lacks a cranial process and has a very long 
caudal process that, in some photos (there is a good one in Charig's 
book 'A New Look at the Dinosaurs', as well as many of the Solnhofen 
texts), is visible as a tiny splint just caudal to the femur. Funnily 
enough, von Huene (1923, 1926) came to the same conclusion as me and 
gives his reconstructed _C. longipes_ a very elongate pubic boot with 
no cranial process. 

The deep 'boat shaped' pubic boot of _Coelurus_ is different, and in 
fact preserves a small cranial process that is apparently the remnant 
of a larger one that is now broken off (O. Rauhut pers. comm.). Some 
dromaeosaurids and other maniraptorans also lack a cranial process, 
but they have a completely different boot morphology. 

That the BMNH R178 holotype is so similar to the _C. corallestris_ 
holotype, in fact it's hardly distinguishable (but I say that going 
only from Bidar et al's unhelpful monograph), could I suppose mean 
that they are the same species. Stratigraphical and size differences, 
however, indicate that this is very unlikely, so _Aristosuchus_ 
remains distinct but presently without autapomorphies. Am I correct 
in thinking that this therefore makes it a metataxon?

Dorsal and caudal vertebrae, claws, a proximal femur, a complete left
tibia and an ischium have all been referred to _A. pusillus_. There 
are also Wessex Formation teeth in private collections that probably 
belong to compsognathids (in fact, to go from Zinke's discoveries in 
late Jurassic Portugal, it's worth noting that compsognathids may 
have been fairly numerous in late J-early K times). The dorsal verts 
(BMNH R178a)  lack 'fan shaped' neural spines.. so, does this mean 
that _A. pusillus_ is not a compognathid, or is this character not 
seen in all members of this group? Or do the vertebrae not belong to 
the same animal as BMNH R178? The claws (BMNH R179 and R899) were 
suggested by Seeley (1887) to be of pterosaurian origin, but I think 
they are certainly from a theropod (pterosaur manual unguals tend to 
have square-shaped flexor tubercles (pers. obs.)); the femur (BMNH 
R5194) was described by Galton (1973) and it may well be from a 
compsognathid; the ischium (BMNH R6426) also looks like it could be 
from a compsognathid but lacks precise provenance data while the 
tibia (MIWG 5137) may also be referable to the Compsognathidae. I say 
the latter three 'look like they could belong to a compsognathid' 
because they are pretty much identical to what's seen in SMNK 2349 
PAL.. can't say more about this until the latter is published 
however.

Incidentally, there was once apparently a specimen very similar to 
BMNH R178, again in a private collection. Seeley (1887) wrote that 
'Henry Woodward submitted to me the pubis, imperfect distally, of a 
type very similar to _Coelurus_, from Tilgate' (p. 223). This 
specimen is of unknown whereabouts but if it was from Tilgate 
(Sussex), it could indicate the presence of compsognathids in the 
Weald Clay: all the other records are from the Wessex Formation, a 
Wealden Group horizon exclusive to the Isle of Wight. In fact, 
following the use of iguanodonts and ankylosaurs as crude marker 
fossils by Norman and Pereda-Suberbiola, I have suggested (in press) 
that compsognathids could one day prove to be stratigraphical markers 
in the Wealden (in that they so far appear restricted to the Wessex 
Fm.). This is speculative and based on absence of specimens.

FINALLY, _Aristosuchus_ was reported from Lower K Romania by Jurcsak 
(1982) and Jurcsak and Popa (1983, 1984). However, this ID is based 
only on two isolated caudal vertebrae. The referral "therefore cannot 
be confirmed and should be regarded as unsubstantiated" (to quote 
myself). Benton et al. (1997) said pretty much the same.

"Kill them all. Let god sort it out"

DARREN NAISH 
PALAEOBIOLOGY RESEARCH GROUP
School of Earth, Environmental & Physical Sciences
UNIVERSITY OF PORTSMOUTH
Burnaby Building
Burnaby Road                           email: darren.naish@port.ac.uk
Portsmouth UK                          tel: 01703 446718
P01 3QL                               [COMING SOON: 
http://www.naish-zoology.com]