133
A geometric approach to cranial sexual dimorphism in fossil skulls
from Předmostí (Upper Palaeolithic, Czech Republic)
Stanislav Katina1, Alena Šefčáková2, Jana Velemínská3, Jaroslav Brůžek4,5
and Petr Velemínský6
1Department of Probability and Mathematical Statistics, Faculty of Mathematics, Physics and CS, Comenius
University, Bratislava, Slovakia, e-mail: katina@fmph.uniba.sk
2Department of Anthropology, Slovak National Museum, Bratislava, Slovakia
3Department of Anthropology and Human Genetics, Faculty of Natural Sciences, Charles University, Praha,
Czech Republic
4UMR 5199 de CNRS, PACEA – De la Préhistoire à l'Actuel : Culture, Environnement, Anthropologie,
Université Bordeaux I, Talence, France
5Department of Social and Cultural Anthropology, Faculty of Humanities, University of West Bohemia, Plzeň,
Czech Republic
6Department of Anthropology, National Museum, Praha, Czech Republic
Received March 8, 2004 Accepted June15, 2004
Abstract. The recently rediscovered glass plate negatives of unique skeletal fossil material from Předmostí
– their estimated ranges from 25–27 000 years but they were destroyed during World War II – were
studied by means of geometric morphometrics. The aim of this study was to determine the sexual shape
variability of the best preserved skulls of adult individuals, using the methods of statistical shape analysis.
While the results roughly support Matiegka’s sex estimation, skull 1 (Matiegka’s female) is more similar
to males. The differences between the skulls of the Předmostí specimens are “more striking” in the
lateral, frontal and vertical views, whereas the inter-sexual differences affect the non-affine parts of
changes in vertical, lateral and basal views in particular. The areas of the metopion, glabella, auriculare,
porion, asterion, gnathion and opisthocranion landmarks present more striking variations, i.e. those with
more discriminatory power. According to the computations made of the various skull distances, the male
skulls are more similar to each other and the female skulls less similar. This suggests possible patrilocal
behaviour on the part of the males.
s Geometric morphometrics, shape analysis, 2D landmarks, thin-plate splines, 3D order penalty, partial
warps, sexual dimorphism, Upper Palaeolithic
INTRODUCTION
The Middle and Upper Palaeolithic site of Předmostí is located at the southern entrance
of the Moravian Gate, an important Central European pass. This site, one of the richest
and most spectacular Palaeolithic sites anywhere in Central Europe, is traditionally cited
as a more or less tragic case. Archaeological excavations started early in the 19th century
but covered only limited areas, and were sometimes undertaken in the unfavourable atmosphere
of industrial limestone and loess exploitation and personal competition
between the archaeologists concerned. The site yielded the largest assemblage of Early
Modern skeletons known hitherto. At the end of World War II, however, most of the accumulated
material, including the human fossils, was damaged by a fire (Svoboda 2000).
Časopis Národního muzea, Řada přírodovědná
(J. Nat. Mus., Nat. Hist. Ser.)
Vol. 173 (1–4): 133–144, 2004 Anthropology
134
An important part of the biological variability of one skeletal sample or a species is determined
by gender. Sexual dimorphism is a common and sometimes prominent feature
of the living world. Males and females differ in this way in a variety of traits. The magnitude
of sexual dimorphism is used in phylogenetic analysis, as well as to infer the social
structure of extinct species and their reproductive behaviour. In human populations,
a reduction in the degree of dimorphism is also indicative of stress and relatively poor living
conditions. In palaeoanthropology, a difference in the pattern and degree of sexual dimorphism
may provide a means for discriminating between populations and explaining a
component of intra-specific variation (Frayer et Wolpoff 1985). The ability to ascertain
sex in skeletal and fossil remains is, however, critical for the reconstruction of the reproductive
behaviour and biology.
An inevitable result of biological processes such as growth, development and evolution
represents a change in the form of the object under study. The form of an object consists
of both its size and shape. Most biological forms contain identifiable loci, which are referred
to as biological landmarks. To be of analytical use, these must be present on all of
the specimens under consideration. Landmark data are the co-ordinates of these biological
loci (Lele et Richtsmeier 1992).
Given that the cranium is a complex, three-dimensional structure, and that the growth
of the cranium involves the co-ordinated growth of many skeletal and soft-tissue elements,
the “new morphometry” of geometric shape analysis should provide a more comprehensive,
detailed understanding of sexual dimorphism (Hens 2002).
In recent years there has been an increasing interest in the use of geometric morphometric
methods as opposed to the traditional multivariate analysis of selected distance
measurements, angles, and ratios in studying variation in shape (Rohlf 2003). Traditionally,
morphometric studies have relied on the statistical analysis of distances, angles or
ratios to investigate morphometric variation among taxa. Recently, geometric techniques
have been developed for the direct analysis of landmark data. Shape co-ordinates, thinplate
splines and relative warp analyses are the essence of these kinds of newer geometric
techniques (Lynch et al. 1996, Dryden et Mardia 1999, Ross et al. 1999, Hennessy et
Stringer 2002, Katina 2002).
Geometric morphometric methods usually begin with the digitised co-ordinates of
numbers of landmark locations. The effects of variation in location, orientation, and scale
of the specimens are eliminated and the differences that remain represent shape variation,
expressed with respect to a suite of shape variables. Statistically, the advantage of this approach
is that with sufficient sample sizes one may expect a much higher statistical power
to detect shape differences, because landmark co-ordinates capture more information
about shape than can be obtained from traditional morphometric measurements. Another
advantage is that these methods usually provide better visualisations of the results (Bookstein
1991, O’Higgins 2000, Rohlf 2003, Katina 2003). Perhaps the greatest advantage of
geometric morphometrics, however, is that it provides a means of quantifying the shape
differences, and therefore differences in character states, of variable traits that cannot be
directly linearly measured (Harvati 2003, Kováč et Katina 2003).
More than 100 years after its discovery, the Předmostí hominid samples remain the
best documentation available for knowledge of individual variation among European Upper
Palaeolithic hominids (Brůžek, Tillier 1996). They are still among the most extensive
samples from the Upper Palaeolithic, and moreover are very homogeneous, and thus are
quite rightly considered a representative sample of the whole population
The aim of the following study was to evaluate the sex assessments and consider the
sexual differences among the best preserved skulls of adult individuals from the Předmostí
collection through geometric morphometrics, with the aid of other kinds of penalty
calculation. The classic sex determination proposed by Matiegka was based on the 20th
century paradigm in which emphasis was placed on cranium morphology. Today it is
known that this approach may be loaded by an error of as much as 20% in a sufficiently
large data file (Masset 1987). The study presented here is a continuation of the first analysis
of the sexual dimorphism of the Předmostí skulls by the way of geometric morphometrics
(Šefčáková et al 2003).
MATERIALS
The discovery of Matiegka’s original, contemporary photodocumentation on glass plates
in the Department of Anthropology and Human Genetics of the Faculty of Sciences at
Charles University in Prague (Velemínská et al. 2003a, 2003b) has made it possible to
achieve new insights into the skeletal sample of 27 subjects of different ages and sexes.
Photographs from these negatives were used to illustrate the results of the anthropological
investigation, which were published in two monographs by Prof. Jindřich Matiegka
himself (Matiegka 1934, 1938). The estimated age of the skeletons is estimated at
25–27,000 years (Svoboda 2001).
The material used in this study consists of the professionally digitised glass plate negatives
of fossil skulls (Předmostí 1 – P1, Předmostí 3 – P3, Předmostí 4 – P4, Předmostí 9
– P9, Předmostí 10 – P10) in the accessible norms: frontal, lateral sin., occipital, basal
and vertical views. The skulls in question are those determined by Matiegka to have been
females (P1, P4, P10) and males (P3, P9).
METHODS
a) Assessment of the morphologically characteristic and defined landmarks (craniometric
points) by direct marking on the scans of negatives. In all, 40 landmarks were employed
in five views (Appendix 1). In the frontal view (Fig. 1a): 35 landmarks, 13 of
them in pairs. In left lateral view (Fig. 1b): 22 landmarks. In the occipital view, 7
landmarks, of which 3 in pairs. In the basal view: 11 landmarks, of which 5 paired. In
vertical view – 7 landmarks, of which 2 in pairs.
b) Assessment of the baseline (reference line) – in the frontal, occipital and basal views
the euryon-euryon distance, in the lateral and vertical views the glabella-opisthocranion
distance. All were used for the standardisation of skull comparisons using Bookstein’s
co-ordinates.
c) Shape analysis was implemented using the S-PLUS 4.5 statistical package, employing
specially constructed program routines by S. Katina (Katina 2002, Katina 2003)
and I. L. Dryden’s routines (Dryden et Mardia 1999).
Using these program routines each skull was compared to the others in a particular view
(for example, P1 with P3, P1 with P4, P1 with P9, P1 with P10, etc.); a matrix of Bookstein
penalties was then constructed for each view, as were global penalties for all the views together.
Global penalties were counted as order-means of penalty-values for particular views.
Next, the skull-set was divided into two groups (male and female) on the basis of penaltysimilarities,
and here group shape differences were analysed using Bookstein’s mean
135
136
shapes. The same approach was used in the analysis of male and female differences as for
global penalties. The inter-sexual order-means were counted on the basis of the order of
penalty-differences from the reference object (P1) in particular views. Another criterion
used for inter-sexual singularities is the percentage of affine and non-affine changes.
Not only the correctness of Matiegka’s sex determinations were assessed, but also the
power (measure) of the skull differences.
The algorithm of order-means counting may be understood as a specific approach to
Fig. 1a. Frontal and lateral view sin.: Předmostí Cranium 3 with landmarks in Bookstein co-ordinates,
Frontal view.
a
137
shape analysis, where the researcher has only 2D and digital photographs from particular
views (frontal, lateral, occipital, vertical and basal) available, but not the 3D material, as
is the case with the skeletal material from Předmostí, which is now preserved only in the
form of glass negatives. This paper uses an unweighted algorithm, but in general, particular
views may be weighted by their anthropological significance. Other algorithms relating
to the 2D/3D problem and their applications can be found in Šefčáková et
al.(2003), Katina (2004) and Katina et al (submitted).
Fig. 1b. Frontal and lateral view sin.: Předmostí Cranium 9 with landmarks in Bookstein co-ordinates,
Lateral view sin.
b
138
RESULTS AND DISCUSSION
1) In general Matiegka’s sex estimation may be supported. Only skull 1 (according to
Matiegka a female) seems more similar to skulls 3 and 9 as well (Tab. 1), both of the
latter being considered male. Is skull 1 also male? Such a conclusion might be supported
by the similarities to male skulls P3 and P9 (Tab. 2). Finally, skull P4 is more
similar to skull P10 (female) (Tab. 3).
2) Sex group assignments (excluding P1) enabled the determination of sex differences
between the average male (shape average of skulls P3 and P9) and average female (the
Table 1. Penalty table for skull P1 (in brackets – Bookstein penalty, order; global penalty – order-mean of
particular views).
Order 1 Order 2 Order 3 Order 4
lateral view P9 (0.0111, 3) P3 (0.0111, 2) P10 (0.0074, 1) P4 (0.0126, 4)
frontal view P9 (0.0077, 1) P3 (0.0090, 2) P10 (0.0109, 3) P4 (0.0242, 4)
vertical view P9 (0.0023, 1) P3 (0.0026, 2) P10 (0.0095, 4) P4 (0.0062, 3)
basal view P9 (0.0042, 4) P3 (0.0018, 2) P10 (0.0012, 1) P4 (0.0019, 3)
occipital view P9 (0.0331, 2) P3 (0.0374, 4) P10 (0.0354, 3) P4 (0.0239, 1)
GLOBAL P9 (2.2) P3 (2.4) P10 (2.4) P4 (3)
Table 4. Percentage /100 of AP and PW deformation between mean (P3, P9) and mean (P4, P10) (AP
(ZOW) – affine part (zero order warp), in brackets the x co-ordinate and y co-ordinate changes, PW – partial
warp, GLOBAL – mean percentage/100).
AP (ZOW) PW
lateral view 0.1290 (0.1109; 0.0181) 0.8710
frontal view 0.4553 (0.0159; 0.4393) 0.5447
vertical view 0.0208 (0.0008; 0.0199) 0.9792
basal view 0.1522 (0.0407; 0.1115) 0.8478
occipital view 0.4101 (0.0153; 0.3948) 0.5899
GLOBAL 0.2335 (0.0367; 0.1967) 0.7665
Table 3. Penalty table for skull P4 (in brackets – Bookstein penalty, order; global penalty – order-mean of
particular views).
Order 1 Order 2
lateral view P10 (0.0146, 2) P9 (0.0103, 1)
frontal view P10 (0.0544, 1) P9 (0.0592, 2)
vertical view P10 (0.0017, 1) P9 (0.0024, 2)
basal view P10 (0.0014, 1) P9 (0.0093, 2)
occipital view P10 (0.0122, 2) P9 (0.0071, 1)
GLOBAL P10 (1.4) P9 (1.6)
Table 2. Penalty table for skull P3 (in brackets – Bookstein penalty, order; global penalty – order-mean of
particular views).
Order 1 Order 2 Order 3
lateral view P9 (0.0093, 1) P10 (0.0114, 3) P4 (0.0107, 2)
frontal view P9 (0.0090, 1) P10 (0.0091, 2) P4 (0.0116, 3)
vertical view P9 (0.0028, 1) P10 (0.0104, 3) P4 (0.0064, 2)
basal view P9 (0.0126, 3) P10 (0.0012, 1) P4 (0.0024, 2)
occipital view P9 (0.0187, 1) P10 (0.0220, 2) P4 (0.0688, 3)
GLOBAL P9 (1.4) P10 (2.2) P4 (2.4)
139
shape average of skulls P4 and P10). These could then be used to determine the landmarks
with the highest variability and to depict the overall variability in TPS nets.
3) Inter-skull differences: Judging from the global penalties (allocation to orders), the
skulls differ mostly in the lateral, frontal and vertical views – but adding the other
views (occipital and basal) resulted in no changes to the ordering.
4) Inter-sexual differences: Skulls differ mostly in the non-affine part of changes in the
vertical, lateral and basal views (Tab. 4). The frontal and occipital views have approximately
the same percentages of affine and non-affine parts of changes.
5) Globally (Tab. 4) the affine (stretchable) component of the deformation of the average
male skull to the average female skull presented 23.4% variability, of which the variability
on the x-axis orientation was 3.7% and the variability on the y-axis orientation
19.7 %. Partial warps (PWs) of the non-affine (deformational) component presented
76.6% variability.
6) In the frontal view (Fig. 2), the affine component part (AP) of PWs, by the deformation
of the average male skull to the average female skull, presented a variability of
45.53%, 43.93% of which is attributed to the y-axis orientation, while the variability
on the x-axis orientation is 1.59 %. Furthermore, the non-affine component part of
PWs, by the deformation of the average male skull to the average female skull, presents
a variability of 54.47%.
The sex differences analysed in the frontal view revolve around the irregular movement
in height (softly compression) with a mild inclination to the right side for the affine component
of PWs as well as the movement in the forehead, nasal and jawline areas for nonaffine
parts of PWs. The areas of the bregma, euryon, coronale, frontotemporale,
prosthion, infradentale, gnathion, mentale and gonion are the most variable (with the
greater discriminatory power).
7) In the lateral view, which is particularly important (Tab. 4, Fig. 3), the affine component
of PWs, by the deformation of the average male skull to the average female skull,
presented a variability of 12.9%, of which 11.0% is attributable to the x-axis orientaa
b
Fig. 2. Affine and TPS grids of the deformation of the average male skull (P3, P9) compared to the average
female skull (P4, P10), frontal view sin.: a) AP component of PW grids, b) TPS grid
140
tion while the variability in the y-axis orientation is 1.8 %. Furthermore, the non-affine
component of PWs, by the deformation of the average male skull to the average female
skull, presents a variability of 87.1 %.
The sex differences analysed in the lateral view revolve around movement in the forehead
and occipital areas for the affine component of PWs as well as the movement in the
forehead, jawline, ear and occipital areas for non-affine parts of PWs. The metopion,
glabella, auriculare, porion, asterion, gnathion and opisthocranion areas are the most variable
(with greater discriminatory power).
8) According to the computation of various skull distances (Tab. 5), male skulls are more
homogeneous than female skulls, which are less mutually similar. It follows that females
were probably more heterogeneous and their morphological characteristics were impacted
by a higher gene pool, more specifically by the patrilocal behaviour of males.
A priori these results accord with those of the authors’ first study (Šefčáková et al.
2003), and with the studies by Ahlström (1996), and Rosas et Bastir (2002). According
to these, sexual dimorphism is clearly expressed in the mandible, and the chin is more
prominent in males. The occipital bone is strongly affected by the sex factor. Likewise the
mastoids and the inion project more downward in males. Furthermore, most of the nonuniform
shape changes are located in the posterior part of neurocranium.
In palaeoanthropological studies, the variability patterns of a fossil population are
mostly unknown, making it difficult to determine to what extent sexual dimorphism is responsible
for the variation detected in fossil samples. An important component of this uncertainty
is the effect of size and sex on shape (Rosas et Bastir 2002).
The applications of geometrical morphometrics have been extended over the last few
years (e.g. Bruner et al. 2003, Detroit 2000, O’Higgins 2000, Yaroch 1996, Zollikofer et
Ponce de León 2001) with the aim of better describing morphological differences in the
size and shape of fossil hominids in the genus Homo, and more pertinently to document
the differences in particular taxa or chronospecies.
In palaeoanthropology, inter-population variability is hard to determine and intra-population
variability is another, even more difficult problem. From the Upper Palaeolithic
only isolated finds or burials with few individuals (e.g. Dolní Věstonice) are available.
From this point of view, the Předmostí assemblage is unique. Only after understanding
the variability of recent populations through geometrical morphometry, relevant information
can be found.
Studies of inter-population variability and the sexual dimorphism of the craniofacial area
in recent populations using geometrical morphology are not plentiful (Ahlström,1996,
Rosas and Bastir 2002, Ross et al. 1999, Lynch et al. 1996). The outcomes of these studies
are often inconsistent, e.g. the “finding of no sexual dimorphism in Romano-British crania
and therein no clear confirmation” (Hennessy et Stringer, 2002). According to Rosas et BaTable
5. Differences between male and female groups on the base of differences in Bookstein penalties
(P1 – reference object), in brackets – order; global difference – order-mean of particular views.
distance (P3, P9) distance (P4, P10)
lateral view 0.0000 (1) 0.0052 (2)
frontal view 0.0013 (1) 0.0134 (2)
vertical view 0.0003 (1) 0.0033 (2)
basal view 0.0024 (2) 0.0007 (1)
occipital view 0.0043 (1) 0.0115 (2)
GLOBAL 1.2 1.8
141
stir (2002), “no difference in the influence of size on shape was detected in females and
males and that both size (explained 54%) and sex (explained 37% of the total variance) had
significant influence on shape”. It is possible to allow that in the range of one population
there exists shape affinity of individuals with the same sex.
Because of Matiegka’s classic sex determination, it could be admitted that – from the
results of this study – Předmostí 1 is a male cranium.
In conclusion it must be said that the aim of this study was achieved by providing the
first analysis of adult skulls representing all of the Předmostí fossils. This study is seen
as a preliminary work, while further investigations will require the collection of a larger
comparative sample of recent and fossil specimens, as made, for example, by Rosas et
Bastir (2002).
CONCLUSIONS
The unweighted order-mean approach for the global penalty and finding differences between
male and female groups on the basis of differences from a reference object lead to
the same results as the using of a special form of Euclidean norm in Šefčáková et. al. (2003).
The Předmostí 1 skull is more similar to the skulls in the male group. The skulls
analysed differ mostly in the lateral, frontal and vertical views. The inter-sexual differences
between the skulls are mostly in the non-affine part of changes in the vertical, lateral
and basal views. This deformational component represents 76.6 % of the total. As
expected, the influence of centroid size on shape (allometry) revealed a shift in the proportions
of the neurocranium and the viscerocranium, with marked allometric variation
of the lower face.
Overall, in the frontal view the affine component presents the same changing part as
the non-affine component part (cca 50 %) of PWs. In the lateral view the non-affine (deformation)
component of PWs represents a variability of 87.1 %.
According to the computation of the various skull distances, females were probably
more heterogeneous and males more homogeneous.
a b
Fig. 3. Affine and TPS grids of the deformation of the average male skull (P3, P9) compared to the average
female skull (P4, P10), lateral view sin.: a) AP component of PW grids, b) TPS grid
142
ACKNOWLEDGEMENTS
The authors are grateful to Mr. Andy Bezák for preparing figure 1. This project was supported by grants from
the Grant Agency of the Czech Republic (GAČR, project 206/04/1498) and the Slovak Grant Agency (VEGA,
projects 1/7154/20 and 1/0272/03).
REFERENCES
Ahlström, T. (1996): Sexual dimorphism in medieval human crania studied by three-dimensional thin-plate
spline analysis. – In: Marcus, L. F., Corti, M., Loy, A., Naylor, G. J. P., Slice, D. E. (eds.): Advances in
Morphometrics, NATO ASI Series, Serie A: Live Sciences Vol. 284. Plenum Press, New York and London,
pp. 415–422.
Bräuer, G. (1988): 2. Osteometrie. – In: Knussmann, R. et al. (eds.): Anthropologie. Handbuch des vergleichenden
Biologie des menschen. Band I, Teil 1, G. Fischer Verlag, Stuttgart, New York, pp.160–232.
Bookstein, F. L. (1991): Morphometric Tools for Landmark Data. Geometry and Biology (Orange Book). New
York, Cambridge, University Press, 435 pp.
Bruner, E., Ricci, F., Manzi, G. (2003): Faces from ancient Fezzan: a geometric morphometric approach. – In:
Di Lernia S., Manzi, G. (eds.): Sand, stones, and bones. The archaeology of death in the Wadi Tanezzuft
Valley. Firenze, All’Insegna del Giglio, pp. 251–260.
Brůžek, J., Tillier, A.-M. (1996): A reconsideration of the Předmostí pelves (Gravettian, Czech Republic). – Anthropologie,
34(3): 231–238.
Detroit,F.(2000): The period of transition between Homo erectus and Homo sapiens in East and Southeast Asia:
New perspectives by the way of geometric morphometrics. Acta Anthropologica Sinica, Suppl. to Vol.19:
62–68.
Dryden, I. L., Mardia, K. V. (1999): Statistical shape analysis. – Ed. Wiley, New York, 348 pp.
Frayer, D. W., Wolpoff, M. H. (1985): Sexual Dimorpohism. – Ann. Review of Anthropology, 14: 429–473.
Harvati, K. (2003): Quantitative Analysis of Neanderthal Temporal Bone Morphology Using Three-Dimensional
Geometric Morphometrics. – Am. J. Phys. Anthrop., 120: 323–338.
Hennessy, R. J., Stringer, Ch., B. (2002): Geometric Morphometric Study of the Regional Variation of Modern
Human Craniofacial Form. – Am. J. Phys. Anthrop., 117: 37–48.
Hens, S. M. (2002): A Geometric Approach to Cranial Sexual Dimorphism in the Orang-Utan. – Folia Primatol.,
73: 165–174.
Katina, S. (2002): Multivariate Shape Analysis: Implementation of some methods to S-PLUS, Application to
Biological Sciences. PhD. minimal thesis, Comenius University, Bratislava.
Katina, S. (2003): Application of allometric shape analysis to fish growth with geometrical notes. – Tatra Mountains,
Mathematical Publication, 26: 323–336.
Katina, S. (2004): Total variation penalty in image warping and selected multivariate methods in Shape Analysis.
PhD. thesis, Comenius University, Bratislava.
Katina, S., Šefčáková, A., Thurzo, M., Brůžek, J., Velemínský, P., Velemínská, J.: Affinity and global 3D penalty
in image warping. – Journal of the Royal Statistical Society, Series C (Applied Statistics) – submitted.
Kováč, V., Katina, S. (2003): Ontogenetic patterns and interspecific variability in external morphology of three
sympatric Gymnocephalus species. Proceedings from the Percis III Conference, Medison, Wisconsin,
USA: 25–26.
Lele, S., Richtsmeier, J. T.(1992): On comparing Biological Shapes: Detection of Influential Landmarks. – Am.
J. Phys. Anthrop., 87: 49–65.
Lynch, J. M., Wood, CH. G., Luboga, S. A. (1996): Geometric Morphometrics in Primatology: Craniofacial
Variation in Homo sapiens and Pan troglodytes. – Folia Primatol., 67: 15–39.
Malinowski, A., Božilow, W. (1997): Podstawy antropometrii. Metody, techniki, normy, Wydawnictwo
Naukowe PWN, Warszawa – Lodž, 512 pp.
Masset, Cl. (1987): Le recrutement d'un ensemble funéraire. – In: H. Duday and Cl. Masset, (eds.): Anthropologie
Physique et Archéologie: Méthodes d'études des sépultures. Paris, Editions du CNRS, pp.
111–134.
Matiegka, J. (l934): Homo Předmostensis. Fosilní člověk z Předmostí na Moravě. I. Lebky. ČAVU, třída II, Praha.,
145 pp.
Matiegka, J. (1938): Homo Předmostensis. Fosilní člověk z Předmostí na Moravě. II. Ostatní části kostrové.
ČAVU, třída II, Praha, 91 pp.
O’Higgins, P. (2000): The study of morphological variation in the hominid fossil record: biology, landmarks
and geometry. – J. Anat., 197: 103–120.
143
Rohlf, F. J. (2003): Bias and error in estimates of mean shape in geometric morphometrics. – J. Hum. Evol., 44:
665–683.
Rosas, A., Bastir, M. (2002): Thin-Plate Spline Analysis of Allometry and Sexual Dimorphism in the Human
Craniofacial Complex. – Am. J. Phys. Anthrop., 117: 236–245.
Ross, A., H., McKeown, A., H., Konigsberg, L., W. (1999): Allocation of crania to groups via the ”new morphometry”.
– J. Forensic. Sci., 44 (3): 584–587.
Svoboda, J. (2000): Předmostí. – Archeologické památky střední Moravy. Svazek 1. Vlastivědné muzeum v
Olomouci, 41 pp.
Svoboda, J. (2001): K analýze velkých loveckých sídlišť... výzkum v Předmostí v roce 1992. – Archeologické
rozhledy, LIII.: 431–443.
Šefčáková, A., Katina, S., Velemínská, J., Brůžek, J., Velemínský, P.(2003): Geometric Analysis of Sexual Dimorphism
in Upper Palaeolithic Skulls from Predmosti (Czech Republic. Slovenska antropológia 6
(n.s.1): 141–146.
Velemínská, J., Šefčáková, A., Brůžek, J. (2003a): Matiegkova fototéka Katedry antropologie a genetiky člověka
PřF UK. Sborník Integrální antropologie dneška. Memoriál prof. J. Matiegky. HMČ UK, Praha, pp.
20–26.
Velemínská, J., Šefčáková, A., Brůžek, J. (2003b): Discovery of Photo Negatives of Destroyed Upper Paleolithic
Fossil Material from Předmostí (Czech Republic). Abstracts of Annual Meeting of Paleoanthropology
Society (April 22 and 22), University Arizona, Tempe, pp.34–35.
Yaroch, L. A. (1996): Shape analysis using the thin-plate spline: Neanderthal cranial shape as an exemple. –
Yearbook Phys. Anthrop., 39: 43–89.
Zollicofer, Ch. P. E., Ponce de León, M.S.(2001): Visualising pattern of craniofacial shape variation in Homo
sapiens. – Proc. R. Soc. Lond. B, 269: 801–807.
APPENDIX 1.
List of the landmarks in five different views, after Martin (Bräuer 1988, Malinowski et
Božilow 1997).
Landmarks lateral frontal occipital basal vertical
apt – apertion Y *
ast – asterion Y Y * Y *
au – auriculare Y
b – bregma Y Y Y
ba – basion Y
co – coronale Y * Y *
ek – ektokonchion Y *
enm – endomolare Y *
eu – euryon Y * Y * Y * Y *
fmo – frontomalare orbitale Y *
ft – frontotemporale Y *
g – glabella Y Y
gn – gnathion Y Y
go – gonion Y Y *
i – inion Y Y Y
id – infradentale Y Y
ju – jugale Y
l – lambda Y Y Y
m – metopion Y Y
mf – maxillofrontale Y *
ml – mentale Y Y *
ms – mastoidale Y Y *
n – nasion Y Y
* pairs landmarks
144
Landmarks lateral frontal occipital basal vertical
ns – nasospinale Y Y
o – opistion Y Y
ob – obelion Y Y
ol – orale Y
on – ophryon Y
op – opistocranion Y Y Y
pg – pogonion Y
po – porion Y
pr – prostion Y Y
sbk – subkonchion Y *
sg – supraglabellare Y
so – supraorbitale Y Y
spa – suprakonchion Y *
ss – subspinale Y
sta – staphylion Y
zm – zygomaxillare inferior Y * Y *
zy – zygion Y * Y *
* pairs landmarks