Life reconstructions are a useful means of providing a package of informati
on about morphology, functionality, behaviour, biology and ecological chara
cteristics of an extinct organism. These reconstructions are of interest no
r only to researchers but also to a wider public. Reconstructions of sabret
oothed cats in particular are widely published and exhibited, despite the a
bsence of a general consensus on either how their canines were used or the
prey sought. Cat-like restorations of the American Pleistocene sabretooth g
enus Smilodon prepared by Charles Knight under the direction of J.C. Merria
m were accepted as valid for over three decades until G.J. Miller criticize
d them, claiming that Smilodon should have looked very different from moder
n felids. In particular, he argued for a longer mouth opening and lip line
to provide a wider gape, a retracted nose and ears set relatively lower on
a head with a straighter dorsal profile. These arguments were accepted by m
any authors employing reconstructions, and have lead to depictions of bizar
re appearance and interpretations of rather specialized feeding behaviour.
We believe thar phylogenetic, anatomical and functional considerations poin
t to substantial flaws in the basis for such depictions, and argue for a re
turn to more felid-like morphology and to interpretations of broadly car-li
ke eating patterns. (C) 1998 The Linnean Society of London.