594 FOSSILS I~TTRIRUTED TO GENUS HOMO: SOME GENERAL NOTES
some extent, also in the cranium from Dali. Further-
more, in the juvenile midface of H. antecessor, a curious
depression lateral to the infraorbital foramen is
formed by the eversion of the lateral lip of the inferior
orbital margin. Again, though, this conformation is
shared with Jebel Irhoud and some individuals of
Homo sapiens as well as with Dali and Jinniushan in
China; and the thickened everted inferior margin of
the orbit is lacking in the single comparable adult
specimen from the Gran Dolina. The backward pro-
file tilt of the root of the infraorbital plane resembles
that seen in Homo sapiens, but is not detectable in the
Chinese or Irhoud fossils.
At the present state of our knowledge, it is impos-
sible, moreover, to know whether any of these charac-
ters is invariant in adults of the Gran Dolina and
other populations, a lacuna of a sort that is all too
common in our knowledge of extinct Homo. Similarly,
while the double-arched supraorbital torus is com-
monly cited (Stringer, Hublin and Vandermeersch,
1984) not only as a distinguishing feature of Homo
neandertbalensis but also as a character linking Nean-
derthals with the form represented by the Steinheim
fossil (Schwartz and Tattersall, 1996), it is also seen in
the ATD-6 Gran Dolina juvenile upper face. Such
observations emphasize that while it is possible to
match individual derived characteristics of the Gran
Doha specimens with various other fossil humans, it
is frustratingly difficult to know definitively which of
these characters are genuinely phylogenetically reveal-
ing. In our view, then, it is not possible at present to
determine the exact phylogenetic position of Homo
antecessor, although there can be no question as to its
validity as a distinct species.
The Ceprano partial calvaria is of approximately
the same antiquity as the Gran Dolina material.
However, even though it is both adult and composed
of mostly different cranial components, it appears to
be of very different aspect from the Spanish species.
Like the Gran Dolina fossils, this specimen clearly
demonstrates that hominids had penetrated western
Europe by about 800 Ka. Nonetheless, it seems very
unlikely that the two sites represent the same hominid
species, as already noted by Bermudez de Castro et al.
(1997) and by Manzi et al. (2001). In some ways, the
posterior cranial profile of this specimen recalls that
of OH 9; but a closer comparison would appear to
be with the distinctive supraorbital regions of such
specimens as Arago 21 and Petralona (see below).
HOMO HEIDELBERGENSIS AND
ITS PUTATIVE RELATIVES
In 1908, Otto Schoetensack applied the nomen Homo
beidelbergensis to the virtually complete hominid man-
dible recovered from a gravel pit at Mauer, Germany
(Schoetensack, 1908). In recent years, it has become
increasingly common to refer other fossils to this
species (e.g., Tattersall, 1986; Rightmire, 1990;
Stringer and McKie, 1996). Such fossils are mostly
from the time range of 600-300 Ka, and derive from
sites in Africa and Asia as well as from Europe. Most
specimens in this category are not directly comparable
to the Mauer jaw, consisting as they do of cranial
rather than mandibular components. However, the
hominid remains from the French site of Arago are
particularly valuable in this context, since they
comprise elements from both upper and lower com-
ponents of the skull. The corpora of the two better
preserved Arago mandibles are more gracile than
those of the Mauer specimen, and their rami are not
as elongated a/p as those of the latter. However, we
are particularly impressed by the similarities in pre-
molar and molar morphology between the Mauer and
Arago 13 mandibles. Notably, the anterior teeth in
both were rather large, and the molars long and ovoid;
the Pls are obliquely truncated mesiodistally along
their lingual surfaces, and are thus more more taper-
ing mesiodistally and elongate than the short, buccol-
ingually wide and more ovoid P2s; the protoconids on
the Pls and P2s are centrally placed in both; on P1,
the low lingual swelling lies opposite the protoconid,
whereas on P2 the metaconid is mesially situated rela-
tive to the protoconid; and on both premolars, the
lingual swelling or cusp is bounded by a very small
fovea mesially and a much deeper fovea distally. In the
molars, M2 is larger than both M1 and M3; the
protoconids and metaconids are situated very mesially
on the crowns, and in the same relationship to each
other; M1 shows evidence of a tiny trigonid basin,
whereas this basin is more pronounced on M2 and 3;
and on all molars, the hypoconulid lies just buccal to
the crown midline. In all molars, the talonid basin is
(or was) quite long mesiodistally and very truncated
buccolingually, with some evidence of enamel wrin-
kling. In the mandible itself, both specimens show a
common configuration of the anteroinferior margins
of the jaw, and also share excavated and rounded
gonial regions, a/p long coronoid bases, posteriorly
decreasing corporal heights, and tall but shallow