Analysis of Hair Samples of Mummies from Semna South

(Sudanese Nubia)
DANIEL B. HRDY '
Department o f Anthropology, Haruard Universrty, Cambridge, Massachusetts 02138

KEY WORDS Hair analysis . Hair form

Nubia

- Meroitics

Mummy hair

ABSTRACT

Hair samples from 76 burials at Semna South (Sudanese

Nubia) were examined using a variety of techniques. Electrophoresis and fluorescence microscopy indicated some oxidation of the cuticule and keratin protein had taken place. However, the cuticular structure and the lack of fluorescence of the cortex indicate that the low humidity and non-alkaline conditions
preserved the physical and chemical properties of the hair well. Pigmentation,
even allowing for oxidation of melanin, showed a higher proportion of lighter
samples than is currently associated with the Nubian area. Hair form analysis
showed medium diameter and scale count; the curling variables were intermediate between European and African samples. There was a high ratio of
maximum to minimum curvature (a measure of irregularity), approached only
by Melanesian samples. Meroitic and X-group burial types were not statistically significantly different (largely due t o sample sizes), but the X-group, especially males, showed more African elements than the Meroitic in the curling
variables. Principal components analysis showed the Semna sample to be significantly different from seven populations examined earlier.

Though several studies have been conducted and Titlbach ('77) studied Egyptian mummies
on ancient hair, because of small sample sizes, in Czechoslovakian collections; they found
few have allowed adequate statistical quan- generally good preservation, with the samples
tification, and none has dealt with Nubian resembling modern European populations
material. Egyptian mummy samples have with significant African admixture.
been examined in the past for color and strucThis study analyzes hair samples from Semture by Pruner-Bey (18771, Virchow (18981, na South in Sudanese Nubia using several bioand reportedly by Minakow (18993. Woodbury chemical and metric techniques. The samples
and Woodbury ('32) and Trotter ('431 have contain Meroitic (First Century A.D. t o Fourth
examined ancient Peruvian material using Century A.D.), X-group (Fourth Century to
metric techniques; they found the ancient Sixth Century A.D.), and Christian period
hairs to generally fall in the range of modern (Seventh Century to Tenth Century A.D.) mavariation. Brothwell and Spearman ('63) stud- terial. Strouhal has pointed out ('77) that the
ied North African and other material using a physical relationship of Meroitic and Postvariety of techniques, including microscopic meroitic populations is not clear. It is still not
examination, fluorescence microscopy, and re- known whether X-group burials represent a
flectance spectrophotometry; they found the migration of an ethnically distinct people or
state of preservation of the samples closely re- change in situ of the Meroitics. It is more genlated to environmental factors of the burial erally accepted that Christian period inhabisites. More recently, Chiarelli e t al. ('70/'71) tants were the descendants of the X-group.
studied ancient Egyptian samples with scan- Hence this study adds perspective to the
ning electron microscopy, finding significant physical anthropology of the area.
loss of cuticular scale edges. Using microscop' Current address: Department of Microbiology and Molecular Geic and macroscopic techniques, Titlbachova netics, Harvard Medical SchwI, Boston, Massachusetts 021 15.
AM. J. PHYS. ANTHROP. (1978)49: 277-262.

277

278

DANIEL B. HRDY

The hair samples were analyzed by quantitative hair form analysis (Hrdy, '73), electrophoresis of hair keratins (Hrdy and Baden,
'731, qualitative hair pigmentation analysis
(Martin and Saller, '62), and fluorescence microscopy (Brothwell and Spearman. '63). The
findings of the quantitative hair form analysis were compared to four populations examined by Hrdy ('73).
MATERIALS AND METHODS

The sample consisted of 56 Meroitic, 15 Xgroup, and 5 Christian individuals from Semna South collected between 1966 and 1968 in
the course of the excavations of the Oriental
Institute of the University of Chicago directed
by L. V. Zabkar (Zabkar, '73/'74, "78). Specific
information on individual burials is located in
Zabkar ('78). There was no embalming; mummification resulted from burial conditions
alone. Burials were either of a simple pit grave
type, or of more complex types, including separate burial chambers, ramps, and vaults. The
hair was either attached to the skull or associated with the remains in the fill. Hair from
infants under six months, and samples of insufficient size for measurement were excluded
from the analysis. Age and sex determinations
and burial type were according t o the criteria
of Zabkar ('78).
Electrophoretic studies were carried out as
outlined in Hrdy and Baden ('731, with the
addition of soaking the samples overnight in
0.05 M EDTA and 0.05 M Tris buffer at pH 9.6
to chelate heavy metals that interfere with
chemical extraction of keratin. Fluorescence
microscopy was done using the method of
Brothwell and Spearman ('631, using 0.1%
Acridine Orange dye at pH 4.9. Qualitative
hair color analysis was performed with a
Fischer-Saller hair color standard (Peabody
Museum, Cambridge, Massachusetts) (Martin
and Saller, '62).
Quantitative hair form analysis was carried
out by the method of Hrdy ('731, using the
principal components analysis variables: diameter (in microns, an average of several
determinations); scale count (the number of
curicular scale ridges per 0.52 mm); average
curvature (the inverse of the radius of curvature); ratio of maximum t o minimum curvature (a measure of regularity of hair curling); crimp (number of reverse twists along
the hair shaft per unit distance); and ratio of
natural t o straight length (a measure of func-

tional hair shortening due t o curling). Principal components analysis was performed
using scores standardized on the seven population sample (Hrdy, '73) and the Semna sample.
RESULTS

Electrophoresis of alpha SCM-keratin protein from three samples (identification numbers: Meroitic N224-B, N455; X-group M107)
showed similar patterns for all samples. There
was a large band at the origin and a large band
at the buffer from which represented SCMKB.
This aberrant pattern indicates that the
fibrous protein had aggregated a t the origin,
probably from cross linking of the protein
chains.
Fluorescence microscopy on modern controls showed a greenish fluorescence throughout the cortex and cuticle, with areas of bright
orange associated with fractures in the shaft,
as reported by Brothwell and Spearman ('63).
These fractures and areas of orange were more
pronounced on hairs that had been bleached.
Of nine Semna samples, all had a completely
orange cuticle, with brighter orange highlighting the cuticular structure, which was intact on all samples. Dehris clinging to the
shaft was also bright orange. The cortex on all
samples was greenish, except where the shaft
was broken, which was orange. Hair which
was blond or "bleached" appearing (M048,
M061, M205, M228) fluoresced identically to
the brown samples (M069, M098, M107, M188,
M246). Macroscopically the hair was in generally good condition, with approximately one
percent of the shafts damaged. Eight of the 76
samples had debris clinging to the shafts; the
remainder were relatively clean. Two of the
samples were braided.
Qualitative grading of the samples on the
Fischer-Saller scale is shown in table 1. Samples that were graded on the red scale (I-VI)
for degree of red pigmentation were also
graded on the blond-brown-black scale (A-Y)
for degree of black pigmentation. Twenty-six
percent (29% of the Meroitic, 13% of the Xgroup) of the total sample had some red pigmentation, and 10.5% (8.9 Meroitic, 13% Xgroup) had "blond" pigmentation (FischerSaller category G or less).
The crude variables of the quantitative hair
form analysis are presented in table 2. The results are also broken down for subpopulations
of Meroitic, X-group, and Christian; male and
female; and simple burial type and more com-

NUBIAN MUMMY HAIR

plex. Results from Hrdy ('73) for Northwest

European, East African, Bougainville (Melanesian), and Japanese populations are presented for comparison. In no variable was the
Meroitic significantly different from the Xgroup, male from female or simple burial type
from non-simple. However, the X-group sample showed higher curling variables than the
Meroitic, especially in males (the Christian
group is too small to make valid comparisons).
The sample as a whole was significantly different from the other populations in average
curvature, ratio of maximum to minimum curvature, crimp, and ratio of natural to straight
length. Diameter was significantly different
from Japanese and Bougainville, and scale
count significantly different from the European, Bougainville, and African populations.
Principal components analysis (Hrdy, '73)
results on the first three components (accounting for 80% of the variance) are shown
in table 3 for the total population, with
comparative populations from Hrdy ('73). In
component I, which is heavily loaded on general curling variables and scale count, the
total sample centroid was significantly different from European and African samples,
though it was definitely more European than
African. Component 11, loaded on diameter,
was not significantly different from the comparison populations. Due to the large amount
of irregularity (high ratio of maximum to
minimum curvature values), the Semna sample had a higher score on component 111, which
was heavily loaded on that variable, than the
African and European samples. Only Melanesian samples had a higher score on this
component.
DISCUSSION

Hair keratin is remarkably stable due to

cross-chain disulfide linkages. However, prolonged exposure to harsh conditions will alter
the keratin. The Semna samples were in contact with sand for over a thousand years, and
hence were a t risk for oxidation of the protein
molecules. There undoubtedly was some oxidation, as shown by t h e aggregation of the protein on electrophoresis and the orange fluorescence of the cuticle by fluorescence microscopy. However, the cortex did not have this
oxidized pattern, unlike samples from Egypt
examined by Brothwell and Spearman ('531,
which fluoresced orange throughout. Since
hair form is probably determined by physical

279

>
3

c
(0

cb

.a
0

z
E
cl

280

DANIEL B. HRDY
TABLE 2

Uniuariate statistics for Semna South, subpopulations, and comparison groups (Hrdy, '73).
Simple and complex refer to burial type

Total Semna

X
9

n-

Meroitic

x-group

n-

Christian

n-

Meroitic male

X
8

n-

Meroitic female

X
8

n-

Meroitic simple

X
8

Meroitic complex

nX
S

n
X-groupmale

X
8

n-

X-group female

n
X-group simple

n-

X-group complex

X
S

Northwest Europe ('73)

nX
8

East Africa ('73)

n-

Japan ('73)

X
8

Bougainville ('73)

nX
S

Diameter

Scale
count

Average
curvature

Ratio
curvature

78.92
14.40
76
80.93
14.98
56
72.20
13.59
15
76.60
9.24
5
85.14
11.58
21
80.57
17.40
21
75.90
8.61
10
82.14
16.03
44
75.20
11.30
5
71.13
15.37
8
69.00
25.16
3
73.00
10.80
12
79.17
11.27
30
76.10
15.83
20
95.53
10.35
30
84.06
13.15
30

16.32
3.36
58
16.51
3.21
45
15.00
4.34
9
17.10
2.32
4
16.20
3.72
14
16.60
4.55
18
15.60
2.49
8
16.40
4.53
36
18.60
3.28
2
14.00
3.22
5
16.40
1.62
3
14.30
5.11
6
15.07
1.80
30
17.95
2.04
20
15.47
1.46
30
18.33
2.00
30

1.56
1.44
75
1.47
1.47
55
2.08
1.55
15
1.00
0.44
5
1.43
1.12
21
1.68
2.09
20
1.90
1.66
10
1.37
1.47
43
2.33
2.01
5
2.14
1.51
8
1.29
0.16
3
2.28
1.69
12
0.19
0.13
30
7.55
1.78
20
0.07
0.04
30
4.72
0.85
30

6.66
9.51
53
6.53
10.42
37
6.76
7.51
13
7.85
3.65
3
4.80
7.61
12
10.25
14.36
15
11.07
16.89
9
5.36
7.54
25
10.87
10.43
5
4.58
3.97

7
2.46
1.89
2
7.54
7.94
11
1.10
0.28
30
1.46
0.43
20
1.00
0.00
30
2.11
0.82
30

Crimp

0.337
0.265
47
0.329
0.278
35
0.368
0.245
10
0.325
0.021
2
0.296
0.226
12
0.294
0.278
15
0.368
0.404
9
0.306
0.242
25
0.340
0.300
3
0.380
0.270
6
0.270
1
0.382
0.261
9
0.028
0.056
30
0.949
0.569
20
0.000
0.000
30
1.117
0.806
30

Ratio
length

1.56
0.59
73
1.49
0.52
53
1.84
0.80
15
1.40
0.56
5
1.46
0.55
19
1.54
0.57
20
1.76
0.64
10
1.42
0.48
41
1.79
0.95
5
2.04
0.77

1.66
0.63
3
1.89
0.85
12
1.21
0.31
30
2.97
1.39
20
1.01
0.04
30
2.67
0.75
30

TABLE 3

Scaled component scores for Semna South and comparison populations

Component I

Semna
N.W. Europe ('73)
East Africa ('73)
Japan ('73)
Bougainville ('73)

Component I1

Component 111

Mean

S.D.

Mean

S.D.

Mean

S.D.

-0.344
-0.761
1.640
-0.654
1.002

0.862
0.274
0.697
0.173
0.704

-0.108
-0.674
-0.636
0.345
-0.075

0.985
0.756
1.064
0.635
0.953

0.258
-0.349
-0.329
-0.499
1.015

1.187
0.497
0.708
0.129
1.429

281

NUBIAN MUMMY HAIR

arrangements of the alpha helical proteins

within the cortex (Hrdy and Baden, 731, the
apparent limitation of oxidation to the cuticle
in the Semna sample argues for the maintenance of hair form in the samples in spite of
their age. In line with this is the large variability in hair form (rather than t h e uniformity that one would expect if a uniform
environmental force was acting on t h e sample), and the lack of macroscopic cuticular and
shaft damage. Also arguing for intact keratin
is the large number of samples with intact
cuticle, as opposed to the ancient Egyptian
sample analyzed using scanning electron microscopy by Chiarelli et al. (70/71). In general, low humidity and non-alkaline conditions are optimal for preservation of keratin; both conditions were met in the Semna
samples.
As Brothwell and Spearman (63) point out,
reddish-brown ancient hair is usually the result of partial oxidation of the melanin pigment. This color was seen in a large proportion
of the Semna sample, and also noted by
Titlbachova and Titlbach (77) on Egyptian
material, where it also may have resulted
from the mummification process. However,
t h e large number of blond hairs that are not
associated with the cuticular damage that
bleaching produces, probably points to a significantly lighter-haired population than is
now present in the Nubian region. Brothwell
and Spearman (63) noted genuinely blond
ancient Egyptian samples using reflectance
spectrophotometry. Blondism, especially in
young children, is common in many darkhaired populations (e.g., Australian, Melanesian), and is still found in some Nubian
villages (J.Zabkar, personal communication).
Only one sample (M197) showed cuticular
damage and irregularities definitely consistent with bleaching, although bleaching could
not be ruled out in some of the blond samples.
The average diameter of the Semna sample
was close to both the N.W. European and East
African samples, which are of medium thickness. Of the variables that best distinguish
European and African samples, the total Semn a sample was closer to the European on average curvature, crimp, and ratio of length. The
ratio of curvature, however, was higher than
either, indicating a degree of irregularity approached only by Melanesian samples. Obviously the sample has a greater degree of
African admixture than the Egyptian hair

sample described by Titlbachova and Tiltbach

(771, which had three of 14 samples showing
Negroid elements. Although there is not a
consistent statistically significant difference
between the X-group and Meroitic samples, it
is interesting t h a t the X-group sample, especially the males, had higher curling variables,
indicating more of an African element. Although larger sample sizes are needed for statistically significant results, the results here
are consistent with the evidence summarized
by Strouhal (77) for skeletal material, which
shows X-group very similar to Meroitic, but
having increased negroid elements.
The principal components analysis showed
the Semna population in a unique position on
the three component space when compared to
seven other populations (Hrdy, 73). The combination of high ratio of curvature with moderate diameter and curling differentiates the
sample from the Melanesian, European, African, and Mongoloid groups.
The Semna sample had high coefficients of
variation compared to four other populations,
especially in scale count, average curvature,
and ratio of curvature. This high intra-population variability undoubtedly reflects the
heterogeneous nature of the Nubian population during the Meroitic and Post-meroitic
periods.
ACKNOWLEDGMENTS

I would like to thank Doctor H. P. Baden

and Doctor S. Blaffer Hrdy for their kind help
in this project, which was partially supported
by Grant AM 06838 from the National Institutes of Health. This study will be included in
the publication of the Semna South excavations (University of Chicago Press), and is
being published in this Journal by permission
of Doctor L. V. Zabkar.
LITERATURE CITED
Brothwell. D., and R. Spearman 1963 The hair of earlier
peoples. In: Science in Archaeology. D. Brothwell and E.
Higgs, eds. Thames and Hudeon, London.
Chiarelli, B., A. Conti Fuhrman and E. Rabino Mama 1970/
71 Nota preliminare sulla ultrastruttura dei capelli di
mummia egiziana a1 microecopio elettronico a scansione.
Rivista di Antropologia Roma, 67: 275-278.
Hrdy, D. 1973 Quantitative hair form variation in seven
populations. Am. J. Phys. Anthrop., 39: 7-18.
Hrdy, D.,and H. P. Baden 1973 Biochemical variation of
hair keratins in man and non-human primates. Am. J.
Phys. Anthrop., 39: 19-24.
Martin, R., and K. Saller 1962 Lehrbuch der Anthropolopie. Fischer, Stuttgsrt.

282

DANIEL B. HRDY

Minakow, P. A. 1899 Neue Ergebnisse bei Untersuchung von Haaren aus alten Grabstatten und von
Mummien. Quoted by Woodbury and Woodbury (1932).
Pruner-Bey, D. 1877 On human hair as a race character.
J. Roy. Anthrop. Innt., 6: 71-92.
Strouhal, E. 1977 The physical anthropology of the
Meroitic area. Paper given a t Third International
Meroitic Gmference, Toronto.
Titlbachova, S., and Z. Titlbach 1977 Hair of Egyptian
mummies. In: Multidisciplinary Research on Egyptian
Mummies in Czechoslovakia. E. Strouhal, ed. Zeitschrift
fur Agyptische Sprache und Altertumskunde., 104:
79-85.
Trotter, M. 1943 Hair from Paracas Indian mummies.
Am. J. Phys. Anthrop., 1: 69-75.

Virchow, R. 1898 Uber die Ethnologische Stellung der

lrahistorischen und Protohistorischen Agypter nebst
Benerkungen uber Entfrabung und Verfebung der
Haare. Berlin.
Woodbury, G., and E. T. Woodbury 1932 Differences between Certain of the North American Indian Tribes as
shown by a Microscopical Study of their Head Hair. Colorado State Museum. Denver.
Zabkar, L. V. 1973/74 The Oriental Institute expedition
t o Sudanese Nubia. The Oriental Institute Report.
Chicago, pp. 41-46.
1978 Semna South, a preliminary report on the
1966-68 excavations of t h e University of Chicago Oriental Institute Expedition t o Sudanese Nubia. Kush, XVZ,
in press.