592 FOSSILS ATTRIDUTED TO GENUS HOMO: SOME GENEK.IL NOTES
are raised and arise far medially, behind the midpoint
of the orbit. In all of these features, this cranium
differs very distinctly from the type material of Indo-
nesian Homo erectus. However, such morphological
disjunction does not stop here. Also distinct from
Indonesian Homo erectus, including the facially reason-
ably complete (if distorted) Sangiran 17, is the skull of
KNM-WT 15000-which, in its own way, is equally
different from ER 3733 (although it bears some
compelling similarities to the KNM-ER 1813
cranium from East Turkana, as noted above). WT
15000 is adolescent, but subsequent growth would
not have made it any more similar to 3733; if any-
thing, the differences already clearly present would
presumably have been exaggerated in maturity. The
15000 cranium has a short braincase that is well
rounded in profile; and, to the extent that they are
preserved, its supraorbital surfaces apparently con-
sisted of little more than substantial thickenings of
the superior orbital margins, lacking the aggressive
projection seen in 3733, in either direction. The facial
architectures of the two specimens are also quite dis-
tinct: 15000 has a longer, narrower face with much
more alveolar prognathism; the nasal aperture is both
higher and narrower; and the wide and relatively long
nasal bones of 15000 are gently concave in profile,
and would not have projected significantly above the
nasal aperture. In turn, both the 3733 and 15000 cra-
nia differ significantly from the ER 3883 cranium.
This latter is, however, more or less identical in com-
parable features to the ER 3732 partial cranium,
which suggests the existence of a consistent morph
represented by both of these fossils. In contrast to the
Indonesian specimens and 3733, the 3883/3732
morph shows thickened supraorbital margins that
protrude forward and even slightly down, overhang-
ing the nasion and the rest of the face; the nasal
bones are flat transversely and descend quite verti-
cally from the nasion, probably having possessed little
original outward curvature; and the frontal slopes
strongly up and back, with the highest point of the
cranial profile reached quite far posteriorly. Unlike
both 3733 and 15000, 3883 has a large, downwardly
protrusive mastoid process; and in this morph, the
zygoma flares out from top to bottom, whereas in
3733 and 15000 it is roughly vertical.
Dental comparisons highlight this variety even
further. No teeth are directly associated with 3883
or 3732, but 15000 has a full dentition and 3733
preserves an upper second molar. This latter toothboasts distinct trigon cusps with a much larger para-
cone than metacone; its hypocone is offset distally; the
cristae are sharp; the precingulum is thick; the talon
and trigon basins are well excavated; and the enamel is
smooth. In contrast, the same tooth in 15000 is high
crowned and quite flat occlusally, with wrinkled
enamel and filled-in basins and subequal-size trigon
cusps; there is a mesially arcing, thick preprotocrista,
and a thick postcingulum that broadens into a mesi-
olingually extended hypocone. In both specimens, this
tooth contrasts with the almost ovoid upper second
molar of Sangiran 4, with its low trigon cusps, its
massive, distally rounded hypocone, and its heavily ac-
centuated postprotocrista. Clearly, the Indonesian and
the two African tooth-bearing specimens each repre-
sent a distinctly different morph (although, as already
noted, the cheek teeth of ER 1813 and WT 15000
comfortably occupy the same one).
This, of course, raises another question. While the
Sangiran 4 specimen clearly must be regarded as
Homo erectus, is either of the others Homo ergaster?
The type specimen of Homo ergaster is the mandible
KNM-ER 992, which obviously bears only lower
teeth. The only other east African specimen of this
group that possesses comparable teeth is WT 15000,
in which a more or less complete lower jaw is pre-
served. In 992, the lower canines are quite tall and
buccolingually compressed; the protoconid is by far
the largest cusp of the anterior premolar, which bears
mesial and distal foveae; the posterior premolar and
the molars evidently had not had very excavated
basins; the posterior premolar protoconid and meta-
conid are subequal in size; all molars are elongated,
with rounded and protruding hypoconulids; and the
molar enamel is somewhat wrinkled. In contrast,
the 15000 lower canines are short crowned, with quite
large, deep mesial and distal foveae bounding a strong
lingual pillar that descends to swell out the base; the
premolars, especially the posterior, have deep mesial
and distal basins; the anterior molar is distended
mesially; the two erupted molars have deep but com-
pressed talonid basins ringed by bulbous but well-
defined cusps; and on both erupted molars, the
hypoconid is large and lingually offset. The distinctly
different lower dentitions of both specimens contrast
equally with comparable Indonesian specimens such
as Sangiran 9. Finally, it is worth mentioning that the
OH9 partial calvaria from Tanzania continues to pose
a puzzle. Long regarded, rather vaguely, as an “Afri-
can Homo erectus,” this specimen does not make a