News and Views Hominid status of the Orce cranial fragment

reasserted
J. Gibert*, *Institut de Paleontología, Escola Industrial 23, 08201 Sabadell, Spain
D. Campillo†, †Museu Nacional d’Arqueología de Catalunya, S. Madrona 39,
08031 Barcelona, Spain
J. M. Arqués§, ‡Universidad de Granada, Facultad de Medicina, Av. de Madrid s.n.
E. Garcia-Olivares‡, 18012 Granada, Spain
C. Borja‡ & §Departament D’Informàtica, Fac. Ciències, Universitat Autònoma de
J. Lowenstein¶ Barcelona, Spain
¶University of California, School of Medicine, Department of Medicine,
San Francisco, U.S.A.

Journal of Human Evolution (1998) 34, 203–217

Introduction Anatomical data

The cranial fragment VM-0 was found in
the Plio-Pleistocene deposits of Venta
Micena site (Orce, Spain), which has an
estimated age of about 1·65 Ma. It was
described by Gibert et al. in 1983 and
attributed to Homo sp.
The first publication to mention the equid
hypothesis was that of Agustí & Moyà
(1987), which was modified slightly in Moyà
& Agustí (1989) and Moyà & Köhler
(1997). In 1989 a complete description of
the cranial fragment (Campillo, 1989) and
a detailed anatomical study that separated
VM-0 from the equids (Gibert et al.,
1989a–c) were published. Later, Gibert &
Palmqvist (1995) studied the complexity of
the sagittal and lambdoid cranial sutures,
seen from the exterior, using geometry frac-
tal analysis, which discriminated the pattern
in VM-0 from that in young equids.
Palmqvist (1996, 1997) reanalysed the frac-
tal dimension of the sutures with new data
provided by Moyà, and concluded that
the cranial fragment belonged to an equid.
Furthermore, Palmqvist (1997) criticized
the immunological techniques used by
Garcia-Olivares et al. (1989), Borja &
Garcia-Olivares (1995), Borja et al. (1992,
1997) and Lowenstein (1995), which show
that the affinities of the proteins of VM-0 lie
within hominids, not within equids.
The following anatomical features were used
by Moyà & Köhler (1997) and Palmqvist
(1996, 1997) to argue for the equine affinity
of VM-0: (1) the presence on the internal
surface of the occipital of a small crest; (2)
the presence of a presumed coronal suture
and the complexity of the sagittal and lamb-
doid sutures; and (3) the presence of digitate
impressions.
Our human sample is composed of 311
X-rays (61 infants, 73 juveniles, 177 adults)
and a further 735 human crania (Campillo,
1989) and 72 horse crania (two, aged 0–2
months; six, aged 3–4 months; 15, aged 7–9
months; 49 adults). Eight of these samples
came from the Institute of Zoology, Uni-
versity of Hamburg, six from the Laboratory
of Anatomy, Paris (LAP), 26 from the
Veterinary School of the Universidad
Autonoma de Barcelona (VUAB), 22 from
the Zoology Museum of Barcelona (MZB)
and 10 from the Institute of Paleontology.
The example also includes the right parietal
of an infant fossil equid, Equus granatensis,
from Venta Micena. The equid species in
the comparative sample are: one E. grana-
tensis, 10 Equus chapmani, 12 Equus burchelli,
15 Equus hemionus onager, nine Equus grevy,
five Equus przewalskii and 23 Equus caballus.
We observed the following anatomical dif-
ferences between Equus and Homo. In Equus

0047–2484/98/020203+15 $25.00/0/hu970181 ? 1998 Academic Press Limited

204 . 
(Sisson-Grossman, 1982) an interparietal
bone is present and fuses with the occipital
and the parietals at variable ages, but always
before the adult stage. Its form changes
from quadrilateral in equid newborns, to
triangular in colts. On its internal face there
appears, according to Sisson-Grossman,
‘‘an apophysis in the form of a triangular
pyramid, which has three concave surfaces and
three sharp edges. The lateral edges form part
of the tentorial bony apophysis and the remain-
ing edge provides the insertion for falx cerebri’’
[Figure 1(a)]. VM-0 comprises two parietals
and an occipital or interparietal bone. These
bones are incomplete and are united by the
sagittal and lambdoid sutures. On the inter-
nal face of the occipital there is a small
slender crest [Figure 2(a)]. Lambda is very
well defined at the union of the sagittal and
lambdoid sutures.
If we just consider the size and thickness
of VM-0, it is comparable to equids between
3 and 9 months of age [Figure 1(b)]. The
small size of the skull of 0–2 month-old
equids (Gibert et al., 1989a; Figures 4 and 5
in that paper, pp. 48, 49), and the sagittal
crest in equids older than 7–9 months make
it unlikely that VM-0 could belong to either
of these age groups.
Sagittal endocranial crest
In VM-0, as has already been mentioned,
there is a small sagittal crest with laminate
morphology on the external surface of the
occipital. It shows a constant thickness
along its length and crosses the occipital
in the anteroposterior direction passing
lambda, and curves towards the right. It
Figure 1. (a) Internal cranial surface of Equus sp. (Laboratory of Anatomy, Paris), 0–2 months. *Lambda.
T, Tentorial bony apophysis; SI, unobliterated internal lambdoid suture; C, unobliterated internal coronal
suture; SA, obliterated internal sagittal suture; B, obliterated internal interfrontal suture. (b) Frontal and
obelic region of a 4–5 month old equid. 1–1, sagittal suture. * Position of lambda comparable to that
on VM-0. A–A, lambdoid suture; B–B, interfrontal suture; C–C, coronal suture; SI, lambdoid suture;
OC, occipital; P, parietal; F, frontal.
ET AL.
has a maximum length of 15 mm and a
maximum height of 5·5 mm [Figures 2(a),
3, 4(a, c), 5(a, b)]. The crest is concave on
the right face and convex on the left. It is
probable that during the preparation of the
fossil, the crest was broken a little and,
therefore, it could have been somewhat
higher originally (Campillo, 1989). This
crest falls within the human range of vari-
ation, as demonstrated by Campillo &
Barceló (1989) in their comprehensive study
of the variability of the human occipital. It
could be a venous sinus-related left crest,
delimiting the sagittal groove, which can be
very noticeable at lambda (Campillo, 1989;
Campillo & Gibert, 1996). It corresponds to
the variant VC3 defined by Campillo &
Barceló in 1989 (Figure 4 in that paper:
p. 145).
This crest should be compared with the
anterior edge of the tentorial bony apophysis
described by Sisson-Gossman in equids
[Figure 1(a)]. Previous comparisons (Gibert
et al., 1989a–c) and new observations
confirm that the crest of the tentorial
bony apophysis in newborn horses is
laminate and passes beyond lambda (at 2
months of age), but immediately (at 3–4
months of age) the interparietal thickness
increases. Consequently, a sinus appears in
its interior (at 7 months) and the tentorial
bone apophysis grows considerably. For this
reason the crest changes its form and makes
comparison with the VM-0 crest difficult
(Figure 4).
Moyà & Agustí (1989) and Moyà &
Köhler (1997: p. 4) argue that ‘‘the lack of
the protuberance is due to the fact that
the posterior part of occipital, where this
protuberance is located is not preserved’’. In
(b)
 206
. 
ET AL.

Figure 2. (a) Internal surface of VM-0. 1–1, sagittal suture; 2–2, 3–3, 4, 5, 6, posthumous cranial fissures. *, Cortical scale.
OC, apical fragment of the occipital. C, occipital crest. *Lambda. It is possible to observe that the cortical scale 2(a)
corresponds to the presumed indentation of the presumed coronal suture of Agustí & Moyà (1987). Also the presumed
coronal suture coincides with part of the trace of the postmortem fracture. (b) Occipital bone from a current human specimen
showing a sagittal crest which reaches lambda. 1, superior sagittal crest; 2, right transversal groove; 3, left transversal groove;
4, inferior transversal groove; 5, Endinion.
         207

Figure 3. Photograph of the VM-0 X-ray. The bone shows rounded zones with great density, that
correspond to the ruga cerebralia, whereas those that have less density correspond to the digitate markings.
The dotted area that is apparent in greater or lesser clarity, over the bony surface represents substantia
spongiosa indicating diploe of the cranial bone. 1–1, sagittal suture, in which the indentations show zones
of greater density formed by the substantia compacta that covers it; at the level of the obelionic region, a
small wormian bone (W) is present. Continuing anteriorly, the suture is straighter than in the obelic region
and anterior to the fissure 2–2, it is deformed posthumously and is inclined slightly toward the left side.
There is a greater density here on account of carbonatic concretions. It is important not to confuse the
small points of the parasagittal concretions (K) with the suture. OC corresponds to the small apical
fragment of the occipital that shapes lambda. The straight occipital crest (C) is clearly present and is
diverted towards the right. 2–2, a posthumous cranial fissure, that should not be confused with a supposed
coronal suture, because its direction is oblique and it does not meet the sagittal suture at a right angle.
Moreover, it is easy to verify that the substantia spongiosa reaches its edges and in no point is covered by
substantia compacta, as is apparent for the sutures, 3–3, 4, 5 and 6, also are postmortem fissures.

our opinion this interpretation is not correct.
In VM-0 15 mm of crest on the occipital is
preserved; it is small in relation to the large
equid tentorial process in this position. This
is sufficient to prove that the VM-0 crest is
not an equid crest, as we have previously
argued (Figure 4). Furthermore, Gibert
et al. (1993) show that the kenyan hominid
KNM-ER 3733 has an occipital crest which
reaches lambda. This means that VM-0
cannot be excluded from Homo by this
character.
208 .  ET AL.

Figure 4. (a) Comparison of transverse and sagittal sections from the obelionic region of two equid crania
(0–1 month old and 7 months old) and of VM-0. 1, Transverse section at lambda; 2, transverse section
10 mm posterior to lambda. (b) VM-0 superimposed on a sagittal section of an equid skull (4–5 months
old). The sections are superimposed with lambda aligned. (c) Comparison of the size and sagittal
curvature of VM-0 and an equid (7 months old). 1, Sagittal sections at lambda; 2, sagittal sections 10 mm
posterior to lambda.

Sutures in VM-0 normal in many fossils of Venta Micena

In VM-0 the sagittal and lambdoid sutures
are easily seen externally and internally, and
the line of the suture can easily be traced.
On the external face the cortical scales pro-
duced by diagenesis are visible, which is
(Figure 6).
The sutures have been described by
Campillo (1989), Gibert et al. (1989a–c)
and Gibert & Palmqvist (1995). The suture
is simple both on the external and on the
internal surfaces. It has a length of 59 mm
Figure 5. (a) X-ray of VM-0. 1, 2, 3 sagittal suture; (b) internal face VM-0, drawing by Bregante (in
Campillo, 1989). (c) Scheme of the external surface of the Orce skull. PS, sagittal plane; PT, tranverse
plane; SS, sagittal suture in the anterior fractured area, showing its deviation towards the left; á, angle
between the anterior fracture and the transverse plane; â, angle between the sagittal suture of the fractured
area (FT) and the sagittal plane: ], indicates the direction of rotation necessary to realign the anterior
portion; 0, fragment of occipital bone; P, P*, fragments of the left parietal bone; q, q*, q+-fragments of the
right parietal bone. (d) Scheme of the internal surface of the VM-0 after Moyà & Köhler (1997). â, angle
of deviation towards the left; – – –, direction of the presumed coronal suture.
Figure 6. (a) External face of VM-0. 1–1, Sagittal suture; 2–2, 3–3, 4, posthumous cranial fissues. OC,
apical fragment of the occipital; *cortical scale; *Lambda. (b) Drawing of the sagittal suture (c) wide line
shows the correct trace of the suture. Fine lines show the boundaries of the main cortical scale. (d) Partial
external face of VM-0. A–B, sagittal suture; *Lambda. 1–1, area partially dissolved; 2–2, eroded areas
which present cortical scale; ], presumed coronal suture according to Figure 1(b) in Moyà & Köhler
(1997). (e) Drawing of the sagittal external on the external surface according to Palmqvist (1996, 1997).
        
and it is not obliterated (degree of obliter-
ation of Broca=0). On both the internal
and external surface, the suture is similar
to that of humans in both morphology
and fractal dimension. The fractal dimen-
sion values are: sagittal external (infant
human)=1·138–1·201, VM-0=1·106; sagit-
tal internal (infant humans)=1·167–1·051,
VM-0=1·128 (Arqués & Gibert, 1998). The
lambdoid suture on both the external and
internal surfaces is also simple and not
obliterated. The right segment is 22 mm
long and the left 23 mm [Figures 2(a), 5(a,
b) and 6(a, b)]. The pattern of these sutures
on the internal face is different from the
external pattern, as it is normal in humans
(Testut & Latarjet, 1975; Masset, 1982).
False coronal suture
Agustí & Moyà (1987: p. 536), Moyà &
Agustí (1989: p. 448, lam. 1) and Moyà &
Köhler [1997; Figure 1(a)] suggest the
existence of a presumed coronal suture.
They indicate the trace of the presumed
internal coronal suture on photographs of
the internal face of the right parietal. Only
Moyà & Köhler (1997) indicate presumed
indentations of the presumed coronal
suture, on the external face of the right
parietal. All of the points indicated by these
authors on both faces of the right parietal
correspond to cortical scales or fractures
(Gibert et al., 1989a–c; Figure 6).
In reply, we argue that: (1) the coronal
suture in hominids, equids and other
mammals is always symmetric and perpen-
dicular to the sagittal suture; (2) in VM-0
it is essential before drawing the coronal
suture to study the X-ray data, which permit
sutures to be distinguished from fractures
(Figure 3); (3) it is important to note that on
the internal and external faces, cortical
scales are present [Figures 2(a), 6]; and (4)
the anterior portion of the parietals has not
been correctly restored. This is the reason
why the sagittal suture in this zone has a
211
deviation of 13) (Campillo, 1989). This is
clearly visible in the internal face of VM-0
and in Figure 5.
Agustí & Moyà (1989, Figure 1 in that
paper: p. 449), Moyà & Köhler [1997,
Figure 1(a)] drew the sagittal and coronal
suture over the original drawing of Bergante
[in Campillo (1989), Figures 7 and 8:
p. 194–195]. The drawing of the internal
sagittal suture made by these authors is not
precise and it is not possible to detect the
deviation of 13). As a result, the presumed
coronal suture, following partly posthumous
fractures [Figures 2(a) and 3], points 2–2;
Gibert et al., 1989a–c), seems to be sym-
metric with and perpendicular to the sagittal
suture [Figure 5(d)].
In the correct drawing of Bergante and
also in the X-ray it is possible to observe how
the sagittal suture deviates [Figure 5(b)].
If we consider that the correct trace of
the sagittal suture on the external surface
followed the points 1–2 in Figure 5(a, b),
the coronal suture would have to move
anteriorly [as is indicated in Figure 5(a)] in
order to be perpendicular and symmetric
with the correct interpretation of the sagittal
suture. Neither direct observation, nor the
X-ray, or the drawing of Bergante shows any
indication of coronal suture in this area.
Agustí & Moyà (1987) and Moyà &
Agustí (1989) invoke obliteration to justify
the existence of a coronal suture. In our
opinion they make serious errors in obser-
vation and method because they have not
studied the sequence of obliteration in
equids. In the equids, we have observed the
following. (1) The endocranial aspect of
the sagittal suture and the internal inter-
frontal suture are the first sutures that are
obliterated in equids, and this occurs at
2–3 months of age [Figure 1(a)]. (2) The
external sagittal suture is the second suture
that is obliterated. This process begins
distally and continues progressively towards
the coronal suture (three examples at the
VUAB, four examples at the MZB, 951160,
212 .  ET AL.
950410, 941229, 941159, one example at
the IPS). The formation of the external
sagittal crest coincides in equids with this
obliteration. (3) The third suture that is
obliterated is the external interfrontal
suture. This process begins in equids before
the external sagittal suture is totally obliter-
ated (an example at the VUAB, two at the
MZB, 941229, 950410, one example at
the IPS). (4) The last sutures that are
obliterated in equids are the coronal and the
lambdoid.
Therefore, if VM-0 was an equid, it could
not have an obliterated coronal suture at
the same time as having an unobliterated
external and internal sagittal suture. This
finding contradicts Agustí & Moyà (1987)
and Moyà & Agustí (1989), and invalidates
the claim that a coronal suture is present
in VM-0. Furthermore this characteristic
permits VM-0 to be included in Homo.
The X-ray (Figure 3) shows the following
features. (1) The sagittal and lambdoid
sutures do not show any sign of synostosis.
(2) Several posthumous fractures are
apparent. The most important of these is the
fracture 2–2, which Agustí & Moyà (1987),
Moyà & Agustí (1989) and Moyà & Köhler
(1997) confuse with a suture without notic-
ing the following that the cancellous bone
reaches the fracture line, which annuls the
suture interpretation because the edges of
the sutures are always covered by compact
tissue (it is well known that the cancellous
bone is never visible in normal bone); and
that at the edges of sagittal and lambdoid
sutures, compact tissue can be seen. (3) The
digitate impressions are apparent and are
normal for the age of the subject.
Complexity of the VM-0 sagittal and
lambdoid sutures according to Moyà &
Köhler and Palmqvist
The endocranial sagittal suture described by
Agustí & Moyà (1987), Moyà & Agustí
(1989) and Moyà & Köhler (1997) is simple
and similar to that of modern human
infants, and it is also similar to the suture
drawn by us. The values of the fractal
dimensions [Agustí & Moyà (1987), Df=
1·128; Moyà & Köhler (1997), Df=1·162]
can be included within the modern human
infant range of variation [1·078–1·167;
Arqués & Gibert (1998)].
In equids the internal sagittal suture is
very simple from 0 to 2 months of age
(Gibert et al., 1989; Figure 5) and after this
age it is obliterated, as can be observed in
the example from the LAP (Figure 1).
These arguments imply that the suture
drawn by these authors can not belong to an
equid older than 2 months.
Palmqvist (1996: p. 19) described and
drew the external sagittal and lambdoid
suture of VM-0 based on Moyà’s interpret-
ation of the sutures drawn for detailed
photographs. The internal sagittal suture is
not very difficult to follow, but it is necessary
to take care when the external sagittal suture
is drawn, because it is eroded (as is normal
in the Venta Micena fossils) and the cortical
scales, small fissures and carbonatic precipi-
tations must be discriminated from the trace
of the real suture. This makes direct obser-
vation of the fossil necessary, in order to
interpret correctly anatomical and tapho-
nomical factors. It is not possible to trace the
correct suture over a photograph (Figure 6).
Because we feel that the suture trace of
Palmqvist (1996) is not correct, his fractal
dimension value (Df=1·335) is not relevant
to the taxonomic interpretation of VM-0.
Palmqvist’s (1996) fractal dimension value
(Df=1·335) suggests to him that VM-0
belonged to an equid of 4–5 months of age
[Figure 1(b)]. At this age, however, the
equids have an obliterated internal sagittal
suture [Figure 1(a)], which is not the case in
VM-0 (Agustí & Moyà, 1987; Moyà &
Agustí, 1989; Moyà & Köhler, 1997). In
relation to the detail of the posterior part of
the sagittal suture and the parieto-occipital
suture of VM-0, Moyà & Köhler (1997) say
        
that the complexity of the sutures is clearly
visible. As we have mentioned earlier, the
distal part of the sagittal ectocranial suture
of VM-0 is the most complex, but the rest of
the suture becomes simpler, as can be
observed in Figure 6(a–c) and Figure 2 of
Moyà & Köhler (1997). Therefore, in our
opinion, the interpretation of these authors
is not correct.
Angles
In Gibert et al. (1989a–c) the values of the
angles of the lambdoid sutures are discussed
as a character that differentiates equids and
humans. This value is related to the volume
of the brain. The angle of the sutures at
lambda is wider in hominids than in equids.
VM-0 has a value of 105) which falls within
the hominid range (90)–160); Gibert et al.,
1989a–c).
Moyà & Köhler (1997) indicated that
‘‘. . . the wide lambdoid angle is not an exclu-
sive human characteristic and with the same
arguments one can affirm that it is ‘completely
within the horse variability’ [lambdoid angle in
extant Equus between 66) and 143) (Gibert
et al., 1989a)]’’.
It is correct that colts 0–2 months old
present wide angle values that change
quickly (Gibert et al., 1989a–c). However, it
is also true that comparison with these colts
is not possible, as indicated in our publica-
tions, because the curve and surface of
VM-0 exceeds the skull vault of these colts.
Correct comparisons can be made only after
3–4 months of age, and at these ages the
maximum angle between the lambdoid
sutures is 66) [Figure 1(b); Gibert et al.,
1989a–c].
Curvatures and size of VM-0
Moyà & Köhler (1997) argue that according
to size and curvature, it is very difficult
to classify VM-0 and it can be included
213
within the range of variation of juvenile large
mammals. According to them
‘‘the degree of the parietal curvature (sagittally
and transversally), the general size of the speci-
men, and the slenderness of the bone suggest
that it belongs to an infant or juvenile large
mammal. Curvatures of surfaces of the size of
the Orce fragment (75 mm transversal diam-
eter) can easily be found in other mammals
such as young equids, rhinos etc.’’
We have compared the transverse
curvature of VM-0 obtained from com-
puterized tomographs at lambda, at 5 mm
anterior to lambda and at 10 mm posterior
at lambda [Figure 4(c)], with the homo-
logous curvatures in juvenile equids and
VM-0 shows greater curvature. We also
compared (Gibert et al., 1989a,b) VM-0
with juvenile equids from different species
and the same results were obtained. Finally,
we observed (Gibert et al., 1989a) even
greater difference between VM-0 and nu-
merous examples of carnivores, ruminants
and primates (juvenile and adult).
Digitate impressions (impressions for
cerebral gyri)
All the mammal infants, with the exception
of humans, have digital markings that persist
into adulthood. In general digitate impres-
sions in Homo tend to be more circular, less
deep and have less sharp boundaries than in
other mammals. VM-0 presents the charac-
teristic digitate marks of infant humans
(Campillo, 1989).
Comparison with an infant
E. granatensis of Venta Micena
It is very important that a fossil right parietal
of an infant E. granatensis has been found at
Venta Micena. It has sagittal, lambdoid and
coronal sutures, and digitate impressions of
equal dimensions and morphology to those
214 . 
of modern colts. The comparison with
VM-0 does not allow any doubt. VM-0 is
different in sagittal curvature, in the length
of the sagittal arc and in the angle of the
lambdoid sutures. For more details see
Gibert et al. (1989a–c).
Anatomical conclusions
VM-0 presents three important anatomical
characters, which point to a large cranial
capacity, the main characteristic of Homo.
These characters also exclude VM-0 from
the equids and other mammals. These three
characters are as follows: length of the
sagittal arch—longer in VM-0 (60 mm)
than in the young equids (maximum 45,
minimum 15, average 32; Gibert et al.,
1989a–c; large size of the parietal curves of
VM-0 [Figure 4(a, c)]; and large angle
between the lamboid sutures (105)).
Moyá & Köhler (1997) asked ‘‘But where
are the evident features diagnostic (autopo-
morphic) for humans?’’ We have a clear
answer, VM-0 has a large brain size. It is
also important to note that VM-0 is larger
than are the skulls of 0–2 month-old equids;
there is nonobliteration of the internal
sagittal suture in VM-0, which is found in
equids older than 2–3 months; and there is
nonobliteration of the external sagittal
suture in VM-0, which is found in equids
older than 7 months. In our opinion this
means that VM-0 can not be an equid.
We have demonstrated that (1) the inter-
nal occipital crest of VM-0 lies within the
range of human variability. It is not possible
to equate this crest with the tentorial bony
process of equids (Figure 4). (2) The
internal sagittal suture has simple trace
(Campillo, 1989), which is similar to that
described by Agustí & Moyà (1987), Moyà
& Agustí (1989) and Moyà & Köhler
(1997). It can be included within the range
of the human infants by virtue of its fractal
dimension. The external sagittal suture also
has a simple trace, similar to humans. (3)
ET AL.
The drawing of the sagittal suture presented
by Palmqvist (1996, 1997) is not correct.
(4) The coronal suture does not occur in
VM-0, whether the obliterated zones are
considered (Agustí & Moyà, 1987; Moyà &
Agustí, 1989) or not (Moyà & Köhler,
1997). (5) The digitate impressions of
VM-0 are similar to those present in modern
humans. (6) VM-0 is very different from
the infant E. granatensis parietal found at
Venta Micena and other infant equids and
nonhuman mammals.
Biomolecules
Palmqvist (1997) maintains that the
meanders of the sagittal suture of the Orce
skull (VM-0) are a more reliable method of
species identification than the immunology
of macromolecules. DNA encodes proteins
that determine morphology, but molecules
of DNA and proteins have highly species-
specific sequences, whereas morphological
criteria such as skull shape and suture lines
have an unavoidably subjective component,
especially when the specimen is frag-
mentary, as is the case here. It would be
ideal if DNA and proteins persisted in
fossils in their original form, but of course
only a small proportion survive the chemical
vicissitudes of weathering and diagenesis,
and those that do survive are frag-
mentary and to some extent altered by
substitutions, racemization and other such
processes.
Nevertheless, under favourable circum-
stances, it has been possible using radio-
immunoassay (RIA) to obtain useful
phylogenetic information from extinct
species such as mammoths and mastodons,
the South African quagga (a zebra-like
equid), the Tasmanian marsupial wolf,
Steller’s sea cow, 8-Ma-old fossil apes from
Pakistan, and human remains that include
an Egyptian mummy and Homo erectus
(Lowenstein, 1981, 1983; Lowenstein &
Scheuenstuhl, 1991).
        
A valid criticism of immunological tests
on fossils is that there have not been
enough independent analyses of the same
specimens, which would either validate or
refute the reliability of the various method-
ologies (Downs & Lowenstein, 1995). In the
present instance, however, independent
tests were carried out by two groups, one
working at the University of Granada in
Spain (Borja & Garcia-Olivares, 1995) and
the other at the University of California
in San Francisco (UCSF), U.S.A.
(Lowenstein, 1995). Different methods
were employed (enzyme-linked immuno-
sorbent assay and RIA), yet the results were
nearly identical (Borja et al., 1997).
Both detected human albumin in Venta
Micena fossils VM-0 and VM-1960. The
UCSF team also detected human collagen
and transferrin in VM-0. The Granada
team detected human immunoglobulin in
VM-1960 and in VM-3691. Both found
equine albumin in several fossils un-
equivocally identified as equid. The UCSF
team found bison albumin in two bovine
fossils. Neither team obtained albumin reac-
tions in Cueva Victoria putative hominids
CV-1 or CV-2. The UCSF team also
detected anomalous human-like albumin in
one equid fossil. This has been used by
Palmqvist (1997) as the basis for his
criticism of the rest of the concordant results
of both groups.
Although great care is taken to avoid
contamination from human hands or hair,
we have occasionally detected human
albumin on nonhuman fossils and stone
tools, a well-known potential problem in all
molecular evolutionary research. It seems
extremely unlikely, however, that the suite
of reactions described above are invalidated
by a single discrepancy. The overall findings
of the two teams are concordant with each
other and include appropriate equine and
bovine controls. We also included strict con-
trols, especially with fossils assigned to
hominids, to demonstrate that the protein
215
detected was indeed integrated in the min-
eral structure of the bone, and was not
superficial, as would be expected if this
protein had originated from exogenous
contamination.
Another source Palmqvist (1997) cites in
support of his criticism of our results with
fossil bones is a review article reporting the
discrepant results obtained in immunologi-
cal studies of blood residues of prehistoric
stone tools (Fiedel, 1996). Although
Palmqvist claims and erroneously reported
that Fiedel’s paper deals with ‘‘studies of
proteins recovered from fossil bones and
blood residues from prehistoric artefacts’’,
the paper deals only with prehistoric arte-
facts, and not with fossil bones.
Discrepant results from archaeological
stones cannot be extrapolated to fossil
bones, in contrast to what Palmqvist appears
to be implying. Downs & Lowenstein
(1995) found that archaeological blood
residues generally could not be identified
reliably. Nevertheless, whereas proteins can-
not be clearly identified on artefacts only a
thousand years old, these biomolecules can
survive under the right conditions and be
characterized in much more ancient fossil
bones (Wykoff, 1972; Lowenstein, 1981;
Lowenstein, 1983; Rowley et al., 1986;
Ulrich et al., 1987). These reports clearly
show that proteins from some fossil bones,
although fragmentary, can persist and retain
enough of their immunological properties
to provide usable genetic information for
periods of millions of years.
Preservation in bones, unlike stones, is
favoured by the proteins being embedded in
the mineral phase that provides considerable
protection from the chemical environment.
We are still learning which conditions of
fossilization are most likely to result in
long-term stability of DNA and protein
molecules. Nevertheless, some authors
found that there is no strict relation between
the biomolecule content (proteins or DNA)
and the age of the fossils or ancient bones. In
216 . 
many instances, older bones contain more
biomolecules than younger ones (Ulrich
et al., 1987; Poinar et al., 1996). Fossils
therefore cannot be judged solely by their
age. Other factors such as average tempera-
ture, presence or absence of water and pH
all appear to play more important roles. The
fossils at Venta Micena were apparently
accumulations around hyaena dens, pre-
served in a lake bed, in a chalky white,
probably alkaline soil. Nevertheless, the
‘‘high’’ concentrations of protein recovered
from some of them, referred to by
Palmqvist, are still less than one-hundredth
of 1% of the amount present in a fresh bone.
Moreover, the amounts of protein detected
in the fossils from Venta Micena are within
the range of those proteins detected in even
older fossils (Ulrich et al., 1987).
The last part of Palmqvist’s (1997)
criticism is based on a study by Cattaneo
et al. (1993). These authors reported that
proteins did not survive in a bone buried for
a few months, leading Palmqvist to wonder
how proteins can survive for 1·6 Ma.
A fossil bone is a singular specimen that
becomes a fossil only under very specific
conditions. These conditions are responsible
not only for the existence of the fossil, but
also, probably, for the preservation of bio-
molecules. Fossil bones that are 1·6 Ma old
and bones buried for few months are by no
means comparable. Nevertheless, we have
also studied buried bones (collected 10 years
after burial) and found amounts of albumin
not much higher than those detected in
fossils. This apparent contradiction can be
explained if we consider that after death,
under normal conditions, proteins tend to
disappear. However, under the singular con-
ditions that lead to fossilization, as at Venta
Micena, proteins are probably ‘‘frozen’’ into
the mineral phase of the bone, so that they
are preserved for millions of years.
Biomolecular palaeontology is a new way,
still in its adolescence, of attempting to
extract genetic information from fossils. Its
ET AL.
findings are complementary to those of the
palaeontologist, physical anthropologist and
archaeologist. At Orce, several lithic assem-
blages testify to the presence of hominids in
the lower Pleistocene. Three fragmentary
bones from Venta Micena (out of the 15,000
excavated) are thought by some physical
anthropologists to be hominid, by others to
be equid or indeterminate. Our immuno-
logical findings provide molecular evidence
that these fossils are probably hominid
remains.
Acknowledgements
This paper has been supported by the
project DGICYT PB 94-1222-01 of the
Spanish Ministry of Education and Science.
We thank Prof. Dr Philip V. Tobias and Drs
F. Ribot, F. Sanchez, A. Malgosa and M.
Walker for helping us in the discussion of
the article, and Dr M. Uribe and Dr Valls
from UDIAT for performing the tomogra-
phy. We thank Rosa Campos and K.
Kupczik for assistance of the translation.
Finally, we thank L. Gibert, who drew
the figures, translated and discussed the
text.
References
Agustí, J. & Moyà, S. (1987). Sobre la identidad del
fragmento craneal atribuido a Homo sp. de Venta
Micena (Orce Granada). Estudios Geol. 42, 538–443.
Arqués, J. M. & Gibert, J. (1998). Review of the fractal
analysis techniques used in paleoanthropology (in
press).
Borja, C. & Garcia-Olivares, E. (1995). Detection
and characterization of proteins in fossils from Venta
Micena and Cueva Victoria by immunological
methods. Abstracts International Conference on
Human Palaeontology (Orce, Granada), p. 26.
Borja, C., Garcia Pacheco, J. M., Ramirez Lopez, J. P.
& Garcia-Olivares, E. (1992). Cuantificación y car-
acterización de albúmina fósil en el cráneo de Orce.
In (J. Gibert, Ed.) Presencia humana en el Pleistoceno
inferior de Granada y Murcia. Museo de Prehistoria y
Paleontologia ‘‘J. Gibert’’, Orce, Granada.
Borja, C., Garcia Pacheco, J. M., Garcia-Olivares, E.
Scheuenstuhl, G. & Lowenstein, J. M. (1997).
Immunospecificity of Albumin detected in 1.6
million-year-old fossils from Venta Micena, Granada,
        
Spain. American Journal of Physical Anthropology. 103,
433–441.
Campillo, D. (1989). Study of the Orce man. In (J.
Gibert, D. Campillo & E. Garcia-Olivares, Eds) Los
Restos Humanos de Orce y Cueva Victoria, pp. 187–
220. Sabadell: Inst. Paleontología ‘‘M. Crusafont’’.
Diputació de Barcelona.
Campillo, D. & Barceló, J. A. (1989). Morphometric
study of the internal surface of the squama occipi-
talis. In (J. Gibert, D. Campillo & E. Garcia-
Olivares, Eds) Los Restos Humanos de Orce y Cueva
Victoria, pp. 109–186. Sabadell: Inst. Paleontologia
‘‘M. Crusafont’’. Diputació de Barcelona.
Campillo, D. & Gibert, J. (1996). El hombre de Orce.
Investigacion y Ciencia 234, 64–69.
Cattaneo, C., Gelsthorpe, K., Phillips, P. & Sokol,
R. J. (1993). Blood residues on stone tools: indoor
outdoor experiments. World Archaol. 25, 20–45.
Downs, E. F. & Lowenstein, J. M. (1995). Identifica-
tion of archaeological blood proteins: a cautionary
note. J. Archaeol. Sci. 22, 11–16.
Fiedel, S. J. (1996). Blood from stones? Some
methodological and interpretative problems in blood
residue analysis. J. Archaeol. Sci. 23, 139–147.
Garcia-Olivares, E., Gallardo Galera, J. M., Martinez
Lopez, F., Borja, C. & García-Olivares, D. (1989).
Detección y caracterización de proteinas fósiles en el
craneo de Orce (resultados preliminares). In (J.
Gibert, D. Campillo & E. Garcia-Olivares, Eds) Los
Restos Humanos de Orce y Cueva Victoria. Inst. Pale-
ontologia ‘‘M. Crusafont’’. Diputació de Barcelona.
Gibert, J., Agusti, J. & Moya, S. (1983). Fragmento
craneal atribuido a Homo sp. de Venta Micena (Orce,
Granada). Paleontologia i Evolucio, Publicacion
especial. Inst. Pal. M. Crusafont.
Gibert, J. (1986). El yacimiento de Venta Micena
(Orce, Granada): su importancia, acción antrópica y
caracteristicas paleoantropológicas del fragmento de
craneo de Homo sp. In (O. Arteaga, Ed.) Homenaje a
Luis Siret (1934–1984). Actas del congreso ‘‘Homenaje
a Luis Siret. Cuevas de Almanzora, Junio 1984,
pp. 37–49. Madrid: Dirección General de Bellas
Artes de la Consejeria de Cultura de la Junta de
Andalucía.
Gibert, J., Ribot, F., Ferrandez, C., Martinez, B. &
Ruz, C. (1989a). Diagnosis diferencial del fragmento
de craneo de Homo sp. del yacimiento de Venta
Micena (Orce, Granada). In (J. Gibert, D. Campillo
& E. Garcia Olivares) Los restos humanos de Orce
y Cueva Victoria, pp. 31–108. Sabadell: Inst.
Paleontologia. Diputación de Barcelona.
Gibert, J., Ribot, F., Ferrandez, C., Martinez, B. &
Caporicci, R. (1989b). Caracteristicas diferenciales
entre el fragmento de craneo de Homo sp. de Venta
Micena (Orce, Granada) y los equidos. Estudios Geol.
45, 121–128.
Gibert, J., Campillo, D., Caporicci, R., Ribot, F.,
Ferrandez, C. & Martinez, B. (1989c). Anatomical
217
study: comparison of the hominid cranial fragment
from Venta Micena (Orce, Spain) with fossil and
extant mammals. Hum. Evol. 4, 283–305.
Gibert, J. & Palmqvist, P. (1995). Fractal analysis of
the Orce skull sutures. J. hum. Evol. 28, 561–575.
Gibert, J., Campillo, D., Martinez, B., Caporicci, R.,
Ferrandez, C., Ribot, F. & Canals, J. (1993). Nuevos
restos de hominidos en los yacimientos de Orce y
Cueva Victoria (España). In El Cuaternario en España
y Portugal. Madrid: Actas de la II Reunion del
Cuaternario Iberico. 345–355.
Masset, C. (1982). Estimation de l’âge au décès par les
sutures crâniennes. Ph.D. Dissertation. Université
Paris VI.
Moyà, S. & Agustí, J. (1989). Una reinterpretación del
fragmento craneal de Orce: Equus stenonis COCHI.
In (J. Gibert, D. Campillo & E. Garcia Olivares,
Eds) Los restos humanos de Orce y Cueva Victoria.
Sabadell: Inst. Pal. Dr M. Crusafont. Diputacion de
Barcelona.
Moyà, S. & Köhler, M. (1997). Anatomy of a mistake.
J. hum. Evol. 33, 911–97.
Lowenstein, J. M. (1981). Immunological reactions
from fossil material. Phil. Trans. R. Soc. Lond. B 292,
143–149.
Lowenstein, J. M. (1983). Fossil proteins and
evolutionary time. Pontif. Acad. Sci. Scr. Var. 50,
151–162.
Lowenstein, J. M. (1995). Immunological reactions on
fossil bones from Orce. International Conference on
Human Palaeontology (Orce, Granada), p. 27.
Lowenstein, J. M. & Scheuenstuhl, G. (1991).
Immunological methods in molecular palaentology.
Phil. Trans. Roy. Soc. Lond. B 333, 375–380.
Palmqvist, P. (1996). La verdadera dimensión del
património de Orce. Alcazaba 103, 15–22.
Palmqvist, P. (1997). A critical re-evaluation of the
evidence for the presence of hominids in lower
pleistocene times at Venta Micena, southern Spain.
J. Hum. Evol. 33, 83–89.
Poinar, H. N., Höss, M., Bada, J. L. & Pääbo (1996).
Amino acid racemization and the preservation of
ancient DNA. Science 272, 864–866.
Rowley, M. J., Rich, P. V., Rich, T. H. & Mackay, I. R.
(1986). Immunoreactive collagen in avian and
mammalian fossils. Naturwissenschaften 73, 620–623.
Sisson, S. & Grossman, J. D. (1982). Anatomia de los
animales domesticos. In (R. Getty, Ed.). Salvat,
Barcelona.
Testud, L. & Latarjet, A. (1975). Anatomia Humana.
Ed. Salvat. Barcelona.
Ulrich, M. M., Perizonius, W. R., Spoor, C. F.,
Sandberg, P. & Vermeer, C. (1987). Extraction of
osteocalcin from fossil bones and teeth. Biochem.
Biophys. Res. Commun. 149, 712–719.
Wyckoff, R. W. G. (1972). The Biochemistry of Animal
Fossils. Bristol: Scientechnica.