Advances in Anthropology
2012. Vol.2, No.4, 169-180
Published Online November 2012 in SciRes (http://www.SciRP.org/journal/aa) http://dx.doi.org/10.4236/aa.2012.24020
Copyright © 2012 SciRes. 169
Were There Commercial Communications between Prehistoric
Harappans and African Populations?
Kenneth A. R. Kennedy1,2*, Gregory L. Possehl3
1Department of Ecology and Evolutionary Biology, Cornell University, New York, USA
2Department of Anthropology, and Asian Studies, Cornell University, New York, USA
3Department of Anthropology, University of Pennsylvania, Philadelphia, USA
Received July 4th, 2012; revised August 20th, 2012; accepted September 4th, 2012
This paper is an historical and scientific evaluation of Western archaeologists’ theories concerning ancient
population movements and commercial contacts between the prehistoric Harappans and African popula-
tions during the Indus Age (2500-1900 BC). In this context the human skeletal remains and artifacts from
Harappa and Mohenjodaro are relevant. An urn-burial from the Indus river site of Chanhu-daro has an
important bearing upon this subject. The scientific aspect of this study is the provision of hitherto unas-
certained data to palaeoanthropologists anaylsing the skeletal and dental biology of prehistoric popula-
tions of South Asia.
Keywords: Prehistoric Harappans
Chanhu-daro is an archaeological site of the Indus Civiliza-
tion (2500-1900 BC) situated in the lower Indus valley of Paki-
stan near the east bank of the river and some 643.6 km from
Mohenjo-daro. The circumstances of the burial of a human
cranium in a pottery jar are unique in the record of Indus Age
disposal of the dead. Excavated in the field session of
1935-1936 under the supervision of E. J. H. Mackay (1932), the
cranium was sent to the American biological anthropologists,
Drs. T. Wingate Todd (1943), W. M. Krogman, and W. H.
Sassaman (1943). Their analysis of the specimen at laboratories
at Case Western Reserve University and the University of Chi-
cago led them to conclude that it possessed some morphometric
features suggestive of East African rather than South Asian
ancestral origins for the Chanhu-daro skull. Further implica-
tions involving the skull and its archaeological associations
pertain to economic trade contacts between the Indus valley
within the “Middle Asian Interaction Sphere” and Africa, e.g.
millets and sorghum, crops encountered in prehistoric archaeo-
logical contexts in these parts of the world. The apparent East
African physical features portrayed in the “dancing girl” statu-
ette from the Indus urban site of Mohenjodaro are discussed.
The Chanhu-Daro Cranium
A human cranium was discovered during the 1935-1936
season of excavation at Chanhu-daro, a site in the Lower Indus
Valley of modern Pakistan. This cranium was associated with a
structure of the Indus Age, or Harappan Civilization (Figure 1)
and is assessed by the present authors with respect to two hy-
potheses hitherto unproposed by archaeologists and palaeoan-
thropologists: 1) The Chanhu-daro burial is a unique funerary
practice of the Indus Civilization, unencountered at other Indus
sites; 2) the cranium’s phenotypic features are outside the
range of morphological variation present in the macropopula-
tions of the Indus Civilization, but fit closely with those of in-
dividuals of East African ancestry, possibly Nubian or Ethio-
pian. Our interpretations of these data support these hypotheses
and others relevant to broader questions: How did the people of
this culture dispose of their dead? What was the degree of bio-
logical and demographic diversity of the Indus peoples? And
what was the nature and extent of contacts of the ancient in-
habitants of the Indus Civilization, and Chanhu-daro specifi-
cally, with populations living beyond the Indus Valley? Fol-
lowing the first excavations at the site of Harappa in the Punjab
in 1920 and at Mohenjo-daro in Sind in 1922 under the general
supervision of Sir John Marshall (1876-1958) (1931), Gen-
eral-Director of the Archaeological Survey of India, there was
keen interest in investigating other Indus sites. Nani Gopal
Majumdar (1897-1938) (1934) explored sites in Sind. Through
the efforts of W. Norman Brown (1892-1975), Professor of
Sanskrit at the University of Pennsylvania, the first American
archaeological expedition to British India was organized. Su-
pervision was under the Field Director, Ernest John Henry
Mackay (1880-1943) (1930-1934, 1936, 1943). This excava-
tion was funded by the Museum of Fine Arts in Boston
Chanhu-daro is a settlement marked by three low mounds
covering approximately five hectares in Nawabshah District of
Sind, Pakistan (Figure 2). It lies some 19.3 km from the eastern
bank of the Indus River, 640 km south-southeast of Mo-
henjo-daro (Figure 3). Alluvium covers the lower parts of the
Mackay (1943) established four cultural levels at Chanhu-
Occupation 4 Jhangar (Iron Age?)
Occupation 3 Trihni (Bronze Age or Iron Age?)
Occupation 2 Jhukar (Post-urban Harappan)
Occupation 1 Harappa
Stuwart Piggott (1910-1996) (1950) revised this scheme as
K. A. R. KENNEDY, G. L. POSSEHL
Settlements of the Indus Civilization c, 2500-1900 BC.
Period III Jhangar
Period II Jhukar
Period I a-c Mature Harappan
The occupation that Mackay exposed most fully is his
Harappa II, of Stuwart Piggott’s (1910-1996). (1946, 1950)
Period Ic, best preserved on Mound II. Here he found a “typi-
cal” Mature Sindhi Harappan settlement. Chanhu-daro’s citi-
zens used baked bricks for many buildings, had paved floors
and a drainage system. Some buildings were grouped along a
wide street that ran north-west to south-east (different from
Mohenjo-daro’s city plan) that appears to have been cut by at
least one thoroughfare coming in at a right angle. This attention
to town planning was not seen in the uppermost Mature Harap-
pan levels (Mackay, 1943). Figurines and toys of copper-bronze
and terra cotta were recovered along with an abundance of
chalcedony and carnelian beads, shell ornaments, ceramics and
faience work. These manufactured goods suggest that Chanhu-
daro was an important craft center with trade connections with
other Harappan cities and villages (Possehl, 2002).
The bead factory is one of the most interesting features of the
Piggott’s Period Ic occupation, and has been taken to indicate
Copyright © 2012 SciRes.
K. A. R. KENNEDY, G. L. POSSEHL
Plan of Chanhu-daro. After Mackay 1943: Plate 1.
Plan of Mound, with street and bead factory, After Mackay 1943:
Figure 4. jar. After Mackay 1943: Plate XCIV, Figure 1.
at there was a significant amount of craft activity at the site
the northwest of the bead factory and furnace
inus its mandible) was given the
(Mackay, 1943). The centerpiece of this installation was a fur-
nace which would have been used in several ways, including
the glazing of steatite, providing heat to bring out the red color
of carnelian and to prepare stone for better chipping, also a part
of the bead making process. Interest in the craft industry at
Chanhu-daro has not been overlooked. G. M. Shar and M. Vi-
dale (1985). Their findings and observations cannot be assigned
to any particular period, but it is most likely to be the latter
Mature Harappan. They confirmed the presence of massive
amounts of craft activity here, especially noting the presence of
kiln wasters, associated with the production of pottery. Finds
associated with lapidary works further strengthen the sense that
the site was a major bead and seal production center. The scale
and diversity of this craft activity is important for our interpret-
tation of Chanhu-daro and the cranium found there since it
appears that its inhabitants were producing more of the same
products than they were consuming. This leads to the inference
that they were involved in trade, with the possibility that they
were, either directly or indirectly, in contact with distant peo-
ples and places.
About 40 m to
ackay discovered the pot within which the cranium we are
discussing was found (Figure 4). He described the find spot as
a well conditioned pavement of two courses of brick (locus 302;
sq. 7/E). This seems to have been the floor of a bathroom in the
house beyond to the west. This measured 7 feet 10 inches long
by 5 feet wide, but the walls around it had been removed for the
sake of their bricks and the room may actually have been larger.
The storage jar on the eastern side served to drain this floor,
and in it was found a human cranium a large shell, a copper or
bronze ring, and a few other implements of copper or bronze,
some of which were broken very rarely is something of value
found in a drainage jar, save objects that have been washed in
accidentally. This notable exception has proved of the greatest
interest, as the report of this cranium, as described by Drs. W.
M. Krogman and W. H. Sassamain (1943) will show. In their
study of the Chanhu-daro specimen, they include in their report
three line drawings of the frontal left lateral and superior as-
pects. Why this cranium was buried here without the rest of the
skeleton, and whether it had any bones as yet undiscovered at
the site remains unknown.
The cranium (a skull m
Copyright © 2012 SciRes. 171
K. A. R. KENNEDY, G. L. POSSEHL
Under Mackay’s sup er portions of the cra-
anhu-daro is not the practice of jar (urn,
rcumstance of secondary burial
The cranium e postmortem
Morphometric Analysis Figures
c measurements taken by Krogman and Sas-
mber 3324 k. It was found at a level of 2.80 m above the
datum point at Locus 324 in Square 7/E at Mound II. Above
and around the jar were the brickwork structures already noted,
their weight and the pressure of the earth having cracked the jar
to a considerable degree. This is the only human skeletal mate-
rial that Mackay notes in his report of excavations at Chanhu-
daro. However, there are photographs in the Archives of the
University of Pennsylvania Museum of at least three other in-
terments. These were found relatively close to the surface of the
site, and Mackay seems to have decided that they were not
associated with the Bronze Age occupation. No grave goods are
visible in the photographs, but further study might lead to in-
ervision, the upp
al vault and face were cleared of dirt and the cranial base and
palate remained in the matrix. Once lifted from the jar the
specimen was coated with paraffin was, a preservation method
in common use at that time. After the excavation it was a sent
to Professor Wilton Krogman, who, at the Laboratory of
Anatomy at Western Reserve University (now Case Western
Reserve University) in Cleveland, cleared the specimen of its
soil matrix, noting the absence of a mandible and vertebrae. He
concluded that these parts had been separated from the cranium
prior to its deposition in the jar, and that “a sufficient period of
time between death and interment (of the skull) had elapsed to
permit the complete disassociation of the skull from a vertebral
column and upper jaw from lower jaw” (Figure 3) (Krogman &
What is unique at Ch
t) containing portions of human skeletal remains, as this oc-
curs at other pre-, post, and Mature Harappan sites. But it is the
disposition of a single cranium in a jar that is unexpected. Nor
is the Chanhu-daro specimen associated with a cemetery, as are
a majority of the skeletons from Harappan sites. These ceme-
teries are usually positioned away from habitation areas. The
diversity of methods for disposing and complete interments as
well as fractional interments. There are 17 distinct archaeo-
logical contexts of Harappan customs for disposal of the deal
(Possehl, 2002; Singh, 1970).
This may be an unusual ci
nce the body could have decomposed at some location away
from the area of burial for the cranium. Whatever happend3ed
to the rest of the body is not known, but it could be that the
cranium was collected from this area and eventually (a day, a
week, a month, a year?) deposited at the find site.
We have Mackay’s testimony that the Chanh
ntained valuable, high-end objects of bronze or copper, along
with a large sea shell. This was unusual. The cranium is also
unusual in and of itself. As will be seen below, there is reason
to believe it was that of a woman who could be considered an
“exotic” foreigner in the contexts of the Indus Civilization. This
gives us reason to believe that the artifacts, the shell and the
cranium, should be associated together, possibly deposited in
the jar at the same time as a kind of cache. It could even be that
the cranium was disassociated from the rest of the woman’s
body immediately or shortly following her death. The soft tis-
sues could have been macerated by placing the relatively fresh
cranium in boiling water, or otherwise processed, and this ves-
tige of this individual was kept by its “owner” along with the
bronze/copper artifacts and sea shell (Figure 4).
is well preserved except for som
mage to the squamous portion of the left temporal bone. The
bones of the inner ear are fragmentary, and missing are the right
stapes and left stapes and incus. A number of anthropometric
measurements and indices were taken by Krogman and his
associate. Unfortunately, they did not measure or describe the
maxillary teeth still embedded in their alveoli: left central and
lateral incisors, second premolar and first molar; right first and
second molars. The rest of the erupted permanent teeth were
lost postmortem, as indicated by the sharp edges of their alveoli
and nonexistence of any bone resorption in the upper gnathic
region. The absence of erupted third molars, usually an indica-
tor of the late juvenile/young adult age status of an individual,
may in this specimen be a case of third molar agenesis. Radio-
graphic examination did not show any trace of these teeth in the
region posterior to the second molars. This is an anomaly, not a
pathological condition. The left first molar shows a small cari-
ous lesion. Since the soil matrix did not contain the missing
teeth, it is likely that they had been lost antimortum (Figures
5-10). These photographs attributed to Mackay (1943) do not
include metric scales.
man (1943) were influenced by the procedures described by
Rudolph Martin (1864-1923) (1928) and Geoffrey Miles Morant
(1899-1964) (1922-1923). However, the codes represented in
Table 1 are no longer used in modern morphometric studies.
The instruments included spreading and sliding calipers, steel
metric tape, a Molleson craniostat, a dioptrographic drawing
Figure 5. ptographic drawings of the Chanhu-daro cranium by Three diao
Krogman and Sassaman 1943. Plat XCVI.
Copyright © 2012 SciRes.
K. A. R. KENNEDY, G. L. POSSEHL
Figure 6. u-daro cranium, frontal aspect. After Mackay
1943: Plate XCV, Figure 6.
Figure 7. hu-daro cranium, left lateral aspect. After
Mackay 1943: Plate XCV, Figure 6.
Figure 8. u-daro cranium, superor aspect. After Mackay The Chanh
1943, Plate XCV, Figure 7.
Figure 9. u-daro cranium, occipital aspect. After Mackay
1943, Plate XCV, Figure 8.
Figure 10. -daro cranium, basalar aspect. After Mackay
strument and a Broadbent-Bolton roentgenographic cepha-
1943, Plate XCV, Figure 9.
lometer. Measurements were taken in mm units. For the cranial
vault, 14 measurements and 5 indices were recorded; for the
analysis. One of the present authors calculated from their diop-
trographic drawings and angle of 78 degrees and the Nasospi-
nale-Prosthion Frankfort Plan an angle of 80 degrees. These
angle measurements follow the procedures defined as cranial
code numbers 72 b and 73 respectively by Rainer Knussmann
(1988). From the metrical data reported in Table 1 of the pre-
sent study, it was possible to estimate the cranial capacity using
the Lee-Pearson (1901) formulae for adult “White” females for
Basion-Bregma Height (1244.05 cm3) and Auricular Height
(1181.27 cm3). These volume game a mean cranial capacity of
1212.66 cm3 based upon an average of ten calculations of an-
thropometric data. In their summary of the Chanhu-daro cra-
nium, Krogmann and his associate conclude that it had be-
longed to a young adult female, given the relatively small size
of the cranial vault, small mastoid processes, low supraorbital
ridges, smoothly rounded occiput, low nuchal crests and small
inion. Age at time of death was estimated to be between 22 and
Copyright © 2012 SciRes. 173
K. A. R. KENNEDY, G. L. POSSEHL
Table 1. easurements and indices of the Chanhu-daro skull. Cranial m
Maximum length (1) 178.0
Minimum breadth (b) 126.5
Minimum l breadth (b’) 97.
Basi-nasal ngth (lb) 94.
Transverse arc (q’) 2
arc (s1) 12
(gl. × 100/lb)
Basio-bregmatic height (h’) 3.0
Auricular-vertical height (oh 108.0
Basi-prosthion length (gl 92.0
Sagittal arc (s) 353.0
Bregma-lambda arc (s ) 116.0
Lambda-opisthion arc (s 1.0
Horizontal circumference (u) 492.0
Bi-asterionic breadth 98.0
Cranial index (b × 100/ 07
Height-length index (h’ × 69.10
Height-breadth index (h’ × 100/b) 97.23
Transverse fronto-parietal index
(b’ × 100/b)
5 years. The basi-occipital suture was closed and sutures of
ial vault is long and narrow (dolicho-
able 2. asurements and indicies of the Chanhu-daro skull.
Total facial height (na-gn gn)
the cranial vault and face are patent. Muscular attachments are
not prominent. Endocranial margins of venous sinus grooves and
meningeal sulci are shallow and have rounded edges. With the
assistance of the anatomist T. Wingate Todd (1884-1938) (1932)
of (Case) Western Reserve University, Krogman and Sassaman
learned that “The skull is that of an individual with a good
health history”. However, vascular pitting on the right and left
parietal bones was interpreted as “evidence of a slight nutria-
tional disturbance in childhood”, a disease marker we would
attribute today to an anemic condition brought on by iron defi-
ciency, malaria, or abnormal hemoglobins among other patho-
The shape of the cran
anic) and its elevation from the cranial base is low (chamae-
cranic), but it is somewhat broader in its cranial length/height
ration (metriocranic). The ratio of the breadth of the frontal
bone to the breadth across the two parietal bones brings the
specimen within the indicial category of possessing a relatively
broad forehead (eurymetroic). Indicial categories of the upper
facial region indicate it is broad and low (euryenic), and the
orbits have a low height/breadth ratio (chamaeconchic). The
nasal aperture is broad (chamaerrhinic). In the mid-facial al-
veolar tooth-bearing region of the maxilla there is considerable
anterior projection, and the indicial values fall between the
extremes of alveolar projection (prognathic) and a more moder-
ate degree of alveolar projection (mesognathic), as determined
by measurements of the Nasion-Prosthion/Frankfort Plan and
Nasospinale-Frankfort Plan respectively. The latter degree of
facial projection indicates the commencement of prognathy
from the nasal root to the landmark Prosthion at the ante-
rior-inferior margin of the alveolar region along the mid-line
projection of the face. Given these measurements, taken by one
of the present authors but not by the earlier investigators, it is
curious that Krogman and Sassaman described the face as
“broad, low, flat and only slightly projecting”, thereby assign-
ing their facial and vault measurements to “contrasting-dis-
harmonic features”. This conclusion demonstrates that the
Chanhu-daro lady was raising some unanticipated problems
given their traditional racial classificatory scheme of how the
denizens of the Indus Civilization cities and town should appear!
Upper facial height, na-pros (g’h) 61.0 mm
Mid-face breadth (gb) 101.0
Bizygomatic breadth (j) 124.0
Nasal height (nhl) 48.5
Nasal breadth (nb) 25.0
Interorbital breadth (dc)
Orbital breadth (01l) 43.0
Orbital height (0sl) 32.5
Palatal length (gl’) 44.5
Palatal breadth (g2) 39.0
Foramen magnum length (fml) 37.0
Foramen magnum breadth (Fmb) 30.0
Total facial index (gh × 100/j)
Upper facial index (g’h × 100/j) 49.03
Total mid-facial index (gh × 100/gb)
Upper mid-facial index (g’h × 100/gb) 60.39
Nasal index (nb × 100/gb) 51.55
Orbital index (02l × l × 100/ol’l) 75.58
Palatal index (g ’) 89.88
Foraminal index (fmb × 100fml) 81.18
Transverse cranio-facial index (j × 100/b) 98.02
2 × 100’gl
Copyright © 2012 SciRes.
K. A. R. KENNEDY, G. L. POSSEHL
In additional to the cranial size and shape variables noted
above, morphological analysis reveals a rounded glabella, a
, a rounded and rather bulbous frontal
contour with a median
nts. Two of the In-
The third millen
political con an be called
en the Indus and the Mediter-
and Central Asia
tion, a place called “Meluhha”, through loca-
he Red Sea (Wright, 2011).
plain frontonasal junction
frontal eminence, a slight sagittal plateau
with a trace of postbregmatic depression, a rounded occipital
contour forming an occiput en chignon (prominent bulging of
the occipital when the cranium is viewed from its posterior
aspect (norma occipitalis). From this point of observation, the
cranium assumes a pentagonal form. This shape is due to
prominent left and right parietal eminences. The vertical aspect
reveals that the cranium has an elongated ovoid form. There is
an os epiptericum on left and right sides of the vault. The pte-
rionic regions are H-shaped. There is moderate occipital bun-
Facial morphology is characterized by a broad nasal aperture
with “troughed” or “guttered” sills at the inferior margins of the
cavity. The interorbital width, while not recorded in the Krog-
man and Sassaman table of measurements, is described in their
text as “rather narrow, with the upper half of the nasal bones
excessively narrow”. They have described the orbits as
“slightly angular”, but the photograph of the cranium’s facial
aspect (norma facialis) indicates it has more rectangular form,
particularly for the left orbit. There is neither postorbital con-
striction nor prominent temporal fullness. The inferior margins
of the nasal cavity are dull (orygmocraspedotic) and do not
exhibit sharp edges. A striking feature of the face, when viewed
from the left lateral aspect (norma lateralis sinistrum) is its
prognatheous alveolar region. This feature is confirmed by the
indicial values and angles discussed above. But the cranial vault
and face are not elevated. The form of the palatal arcade is
parabolic and it is very wide across the alveolar area containing
the incisors and canines. The palatal suture is relatively even
with a slight posteriorly directed rippling at the mid-line. There
is no anterior “bulging” of the suture at this region. Palatal
depth was not recorded (Figures 6-10).
Given these anthropometric and morphological data should
we ascribe the Chanhu-daro female as falling within the range
of phenotypic diversity of the widely scattered populations of
the Indus Civilization, as determined by a comparative study of
human skeletal remains recovered and scientifically examined
from other sites within its cultural domain, or can her ancestry
be traced to populations living well outside its borders, i.e. was
she of “foreign” birth?
Absence of dental morphometric for the Chanhu-daro speci-
men precludes the use of dental train analysis. The maxillary
teeth include LI1, LI2, LM2, LPM2.RM1, RM2 (I = incisor,
PM = premolar, M = molar). Earlier studies of the dental fea-
tures of Indus Civilization crania support the hypothesis that
ancient populations settled along the Indus and Ravi Rivers
were indigenous and not derived from some invading outside
populations (Hawkey, 1998, 2004). That is the Harappan popu-
lations showed revealed close genetic similarities to earlier
hunting-gathering populations or the Veddas of Sri Lanka this
value is 0.086. Dental phenotypic variables also indicate that
gene flow between the Indus region and Egypt (MMD = 0.124)
and values for the ancient Kingdom of Nubia (MMD = 0.088)
suggests that these values were applicable at the times of the
Mature and post Harappan phases (Hawkey, 1998, 2004).
The biological anthropologist S. O. Y. Keita (1988) lists 13
craniofacial components with significant heritability that are
useful in population affinity assessment. Sufficient data for nine
of these measurements (Maximum Cranial Length, Maxi
ranial Breadth, Basio-Bregma Height, Bizygomatic Breadth,
Upper Facial Height, Minimum Frontal Breadth, Minimum
Frontal Breadth, Nasal Height, Nasal Breadth) were recorded
for females of ten other Harappan females.
The Chanhu-daro female and other females from the burials
at the Mature Phase of Harappa cemetery R37 and the Late
Phase jar burials are from the adjacent Harappa cemetery H are
compared for both size and shape compone
s Civilization burial areas at Harappa have skeletons that
reveal closest similarities with Chanhu-daro in size, but only a
moderate similarity in shape (Harappa R37 size = 0.063, shape
= 0.426; Harappa cemetery H size = 0.043, shape = 0.529).
Absence of dental morphometric data for the Chanhu-daro
specimen precludes use of dental trait analysis. The maxillary
teeth present are LPM2, LM1, LPM2.LM1, LI1, LM2, M1 and
LM2 (L = left side of the palate, I = incisor; PM = premolar
molar). Earlier studies of the dental features of Indus Civili-
zation specimens support the hypothesis that populations along
the Indus and Ravi Rivers were indigenous and not derived
from some invading foreign population (Hawkey 1998, 2004).
That is, the Indus population showed close genetic similarities
to earlier hunting-gathering populations of the Indian subconti-
nent (Mean Measure of Divergence = 0.022; for historic Ved-
das of Sri Lanka this value is 0.086). Dental phenotypic simi-
larities also suggest that there had been gene flow between the
Indus region and Egypt (MMD = 0.124) and New Kingdom
Nubia (MMD = 0.088) was probable at the time of the Mature
and Late Harappan Phases (Hawkey, 1998, 2004).
Economic Trade Contacts
nium BC was a time of unique economic and
figurations in a part of the world that c
“Middle Asia”, the region betwe
ranean, bounded on the north by Afghanistan
d on the south by the Arabian Gulf. By the middle of the
third measureable upsurge in economic activity and “intercom-
nectedness” in Middle Asia. This economic development linked
the Indus Civilization with Mesopotamia by sea and across the
Iranian plateau. It also strengthened the ties between the Indus
and Central Asia. Taken as a whole, we have what archaeolo-
gists call the “Middle Asian Interaction Sphere” (MAIS)
Archaeologists first learned of this interaction sphere through
Indus stamp seals that were found in Mesopotamia. Cuneiform
texts also mentioned maritime trade between Mesopotamia and
the Indus Civiliza
ns called “Dilmun” (modern Bahrain) and “Magan” (Oman)
were also cited (Oppenheim, 1954; Possehl, 1996). The pres-
ence of Indus pottery in the Gulf and Oman, at some sites in
considerable quantity (Mery, 2000; Cleuziou & Tosi, 2000;
Carter, 2001), as well as Indus seals, weights, beads, combs and
the like document the presence of Indus commerence and sail-
ors in these areas (Possehl, 1997).
The maritime trade linking ancient India to Mesopotamia is
documented by historical records, and by archaeology. There is
also evidence that this intercourse extended down the eastern
coast of Arabia to the mouth of t
Copyright © 2012 SciRes. 175
K. A. R. KENNEDY, G. L. POSSEHL
The clearest of these indicators is the presence of African
millets in the subcontinent in prehistoric times, which has been
discussed for a number of years (Mehra, 1963; Vishnu-Mittre,
1974) (Figure 11). The principal large seed summer grasses in
lan (1992) has produced two maps
ted that this coastal route “may be a great lost
uth Asia are sorghum or jowar (Sorghum bicolor), pearl mil-
let or bajra (Pennisetum typhoides) and finger millet (Elleusine
coracana). All are endemic across a broad band of sub-Saharan
Africa, that has its eastern terminus in Ethiopia and the mouth
of the Red Sea (Figure 11).
Botanists, interested in the history of domestication, have
discussed this in a number of places (Harlan & Stemler, 1976;
Harlen, deWet, & Stemler, 1976; Harlen, 1992) and it appears
to be an established fact. Har
at are important to understanding the domestication and
spread of sorghum and pearl millet. On the first map, circles
indicate location of wild pearl millet, Pennisetum violaceum;
dark shading indicates the northern pearl millet belt; light
shading indicates areas in which pearl millet is grown, but in
which sorghum is the dominent crop. Sorghum is present at the
site of Hili 8 in Oman. It comes from Phase IIe, which can be
dated to 2330-2250 BC (Cleuziou & Costantini, 1980) who
write that sorghum was probably cultivated around Hili. Co-
stantini (1990) has also documented the presence of these mil-
lets in Yemen at sites dated to the second half of the third mil-
lennium BC. These same three millets are all documented in
India and Pakistan at sites dated from 2400 to 1800 BC
(Possehl, 1997). These include Hulas, Rohira, Rojdi B and C,
Rangpur III, Babar Kot, Surkotada and Daimmabad (Possehl,
Sailors, and probably sailing families, seeking a livelihood,
pioneered the route from the Subcontinent to the Gulf region
and on to Africa (Gogte, 2000). In 1952 Carl O. Sauer (1952,
rridor of mankind”. While these sailors were in African wa-
ters they would have come into contact with the foods of the
local population. They may have brought food of their own, but
Distribution of pearl millet and sorgum in Africa. After Harlan 1992.
boats were small and drawing on local grain and meat would
have made good sense. It is in this context that the African
millets were taken on board boats and brought to places like
Oman, Pakistan and India. When the millets were domesticated
is a question that remains to be settled, but the probability ex-
ists that one or more species were not domesticated but wild, or
in one of those states of limbo in food production, between wild
and fully domesticated (Sauer, 1969).
These millets are drought resistant and amenable to dry
cropping. They are hearty and require little tending as they
mature. They have large seeds on commodious heads covered
with nourishing grain and can be productive and efficiently
harvested. They also grow in hot weather, and were adapted to
the summer monsoon growing of the Subcontinent. These are
the principal reasons for their adoption by the ancient Indian
The millets tell us in a remarkable way that sailors reached
the mouth of the Red Sea. These may have been Indus people
themselves, although less direct contact is also possible, and
complementary to this scenario. In any event, the millets were
taken back to the Subcontinent and integrated into the lives of
the Indus peoples. Given this historical context, it is reasonable
to believe that humans migrated from East Africa to “Magan”
ha”. Their social status (“slave” or “free” or some-
njo-daro is a representa-
1926 and 1927 field season. The house is a small structure deep
where along the multidimensional social continuum between
these categories) is not known, but the Chanhu-daro cranium
would seem to offer strong support of this suggestion.
The Bronze “Dancing Girl” (Figures 12 and 13)
We would be remiss not to mention the fact that the E. C. L.
During Caspers (1987) has suggested that the so-called “Bronze
Dancing Girl” from HR area of Mohe
n of a Nubian. A second “Bronze Dancing Girl” was recov-
ered from the DK area of Mohenjo-daro but it is too corroded to
be a serious part of this discussion.
Depictions of persons with obvious “racial” physical and
dress styles appear in ancient Mesopotamian cultures, namely
carved figurines and paintings, but the only sculptures from the
Harappans is the figure of the “dancing girl” at Mohenjodaro
The statuette of the young woman was found by Sahni dur-
ing the 1926-1927 field session within a house structure of
Block 7 in the area Hr5721. The figure is 10.8 cm in height.
Her trunk is relatively short and the extremities long in her
linear build from the HRr 5721. This is a splendid piece of
copper-bronze casting She was cast by the cire perdue tech-
nique and may well be a portrait of an individual living at that
time. This piece of Harappan art has been discussed elsewhere
(Possehl, 2002). She is 10.8 cm high and was cast using the lost
wax process. The statuette was found in a Late Level house at
Block 7 by D. R. Sahni (1879-1939) (1926, 1927) du
ithin the southwestern quarter of the city.
The figure is a very thin young woman, standing upright,
with her head tilted slightly back, her left leg bent at the knee.
There is little sense of flesh on the body, and upper and lower
limbs are gracile and do not exhibit prominent muscular anat-
omy. Her right arm is bent, with her hand placed provocatively
on the back of the hip, the thumb beside a clenched fist. The
left arm rests slightly bent on her left thigh. The thumb and
refinger of this hand form a circle, and it is apparent that she
Copyright © 2012 SciRes.
K. A. R. KENNEDY, G. L. POSSEHL
Figure of the “dancing girl” from Mohenjo-daro. After
Harlen 1992. Anterior aspect.
Figure of the “dancing girl” from Mohenjo-daro. After
Harlen Posterior aspect.
once held a small object, possibly a baton of some kind. She is
naked except for some ornaments. Around her neck is a small
necklace with three large pendant beads. On the left arm she
wears 24 or 25 bangles. The right arm has four bangles, two at
the wrist and two above the elbow. Her hair is coifed into a
loose bun, held in place along the back of the head, in much the
same way as some Indian women wear their hair today. The
artist has rendered this feature in detail, and it is worth noting
that the hair could be called wavy. It is not a spiral hair form
such as occurs in some modern populations from Nubia. How-
ever, we cannot outrule the possibility that this is a headpiece,
such as a wig, or that the naturally tight curley hair of many
Africans had been artificially straightened. Whether this small
statuette actually portrays a dancer is open to question. Only the
pose seems to convey this.
Some scholars have proposed that the broad nose, large
membranous lips, the anterior projection of the tooth-bearing
portion of the face, and the linearity and proportions of the
upper and lower extremities suggest a Nubian (Caspers, 1987);
a Proto-Australoid (Piggott, 1950). Piggott wrote in a reference
to her that: “when we are describing the Harappa culture, I
think I recognize a Kulli girl in a “foreign city”. He thought that
her hair style was like that found on terra cotta figurines of the
Kulli peoples, a view that the present authors do not share. Sir
John Marshall (1931) caught something special in the figure: It
give him a “vivid impression of a young aboriginal nau girl,
her hand nt pose with legs slightly for-
ward, a ·· the
on hip in a half-impude
s she beats time to the music with her legs and fee ·
ing of the back, hips and buttocks is quite effective”.
The statuette was a favorite of Sir Mortimer Wheeler
(Hawkes, 1982). He described her thusly: “There is her little
Baluchi-style face with pouting lips and insolent look in the eye.
She’s about fifteen years old I should think, not more, but she
stands there with bangles all the way up her arm and nothing
else on. A girl perfectly for the moment, perfectly confident of
herself and the world. There is nothing like her”. The present
authors regard the Bronze Dancing Girl from the HR are of
Mohenjo-daro as “a splendid piece of Indus art, but it cannot be
demonstrated that it is a representation of an African. However,
her body proportions of a relatively short trunk and long upper
and lower extremities fit the essence of Bergmann’s Rule
(Bergmann, 1847; Ruff, 2002) that (African) equatorial popula-
tions are best adapted to heat loss under conditions of high solar
radiation and humidity by this type of body form. Settled north
of the Tropic of Cancer at some 30 degrees north of the Equator
in the southern Temperate Zone, the ancient inhabitants of the
Indus and Ravi Rivers do not exhibit these features, as deter-
mined by analysis of their skeletal remains.
Possible “Ancestry” of the Chanhu-Dara Female
Morphometric of the Chanhu-daro cranium allow us to com-
pare it to skeletal specimens recovered from other sites of the
Indus realm. Previous research indicates that even within the
Indus Valley of the third millennium BC there was considerable
phenotypic diversity (Hemphill et al. 1991; Kennedy, 2000).
The claims that the Harappans constituted a homogeneous
“race” or even a new Bronze Age “sub-species”, as advocated
by P. C. Dutta (1983), has not been vindicated by recent mor-
It is not surprising to encounter in both ancient and modern
panmictic urban populations a wide range of biological diver-
Copyright © 2012 SciRes. 177
K. A. R. KENNEDY, G. L. POSSEHL
sity. But individual variation due to adaptive responses to ge-
netic, cultural and environmental stresses were not viewed in
this way by earlier anthropologists. Their conversions of “racial
classifications” into schemes of “racial palaeontology” were
based upon a concept that ancestral origins of an individual or
population required their categorization into sub-specific natu-
ral entities. The traditional racial concept was based upon lim-
ited and arbitrarily selected frequencies of physical characters
indicative of supposed natural divisions within Homo sapiens
and classification of populations fitting into a hierarchial evolu-
tionary progression from savagery to civilization. Erosion of
this false notion, beginning about fifty years ago among sys-
tematic biologists and biological anthropologists, resulted from
the recognition that classifications of sub-species of any organ-
ism, fossil or living, were invalid when conceived as natural
biological species. Today the formerly designat “races” are
understood to be Homo sapiens and distinctive from other
unexpected that the first describers of the
ical characters in prehis-
ome individuals of
es than males. Orbital forms vary from
hominin taxa because of reproductive isolation, reproductive
and by salient morphological differences. Geographi-
l loci and trinomial appendages to species names for purposes
of identification have replaced racial typology and taxonomy.
Of course popular racial classifications prevail in our societies
at the present time, but there are no valid scientific data to sup-
port “racial categories” arranged in a ladder of superior and
inferior peoples, i.e. “Whites” or “Caucasians”, “Blacks or
Negroids”, “Mongoloids”, etc. (Gill & Rhine, 1990; Kennedy,
Thus it is not
Chanhu-daro cranium sought to fit it into one of the racial clas-
sifications current during the first half of the last century. What
is relevant to Krogman’s and Sassaman’s study is not that they
placed Chanhu-daro with a “Proto-Mediterranean type” and
compared it to other specimens assigned to this racial category,
namely the skeletons from Mohenjo-daro and a Sialkot (a sin-
gle skull collected in 1912 in the western Punjab) as well as
specimens from Iran and the Near East. Rather it is significant
that these American anthropologists recognized that the
Chanhu-daro cranium was not a “perfect fit” within the sorting
criteria of cranial traits used to define the macropopulations of
the Indus Valley. They wrote that the specimen “is Negroid in
its low flattened vault, in its broad nose with its “traughing” or
slight sulcus praenasalis, in its diminished sub-nasal height and
low orbits, and possibly its (cranial) length and occipital full-
ness” (Krogman & Sassaman, 1943).
A notion held by many anthropologists of this period was
that some physical “traits” encountered in modern African
populations appeared in the phenotypic patterns of ancient and
modern peoples of South Asia has a long and complex history.
Precepts were modeled upon European taxonomy of plants and
animals with its beginnings in the mid-eighteenth century.
Carolus Linnaeus’ (1701-1778) (1758) classification of the
genus Homo was adopted and modified by anthropologists of
the nineteenth century (Kennedy, 2000). The carry-over of
racial classifications into British India was in place by 1901
with Herbert Risley’s (1851-1911) (1901) publication of The
Census of India 1901 (Risley & Gait, 1903), and these were
extended to include both skeletal and living populations by the
merican-trained biological anthropologist B. S. Guha (1894-
1961) (1935) in the Census of India 1931.
As other racial classifications for the populations of the In-
dian subcontinent emerged, a persisting them was that a “Ne-
groid” or “Negritoid” (pygmy) racial element was present, and
was especially evident in southern India. Populations assigned
to this type were considered to be “primitive” or of a “lower
stratus”, both biologically and culturally. The small body size,
spiral hair form, dark skin pigmentation and broad noses of
certain tribal groups in Kerala led Guha (1928, 1929) to regard
them as members of a “Negroid strain”. It is possible that their
physical features are not phenotypic parallels with some
sub-Saharan or Nubian African peoples; they may be the result
of genetic contributions of Africans escaping from slave ships
plying the southeastern coast of India and mating with the tribal
hunter-gatherers of the mainland. There are other cases of writ-
ers perceiving African or pygmy phys
ric populations recovered from Indian Iron Age sites (Lapique,
1905; Thurston, 1909).
At this point in this discussion, the reader will have per-
ceived that the Chanhu-daro female has many cranial features
that do not fall within the range of variables encountered in the
macropopulation of the Indus Valley, as this is understood to-
day by anthropologists who have examined directly the skeletal
series from cemetery collections at different Indus burial locali-
ties. These specimens tend to have long and narrow cranial
vaults (dolichocranic) which are moderately elevated when
measured from the base of the craniums to the highest point on
the vault (mesocranic, and more frequently in males). But in
both sexes there are specimens with low cranial vault heights
(chamaecranic, and more frequently in females). As would be
expected in a heterogenous population, s
th sexes had very high cranial vaults (acrocranic). Since there
are various technical methods for measuring the cranial vault
height, it is not surprising that the indicial categories show
some minor differences, but these anthropometric data can be
summarized by noting that this feature ranging from low to
high vaults is found in adults of both sexes.
The forms of the upper facial region vary from an extremely
broad face, when straight-line measurements are taken across
the cheek bones (mesene and eurene, especially in females) to
narrow faces (leptene) measured in this same way. Measure-
ments of the total face with its mandible attached (which is not
applicable to the Chanhu-daro specimen which is lacking its
lower jaw), reveal both broad (euryprosopic) to narrow (lepto-
prosopic) facial structures among the ancient people of the
Indus valley. Broad noses (chamaerrhinic) and narrow noses
(leptorrhinic) appear across the samples, females having
somewhat narrower nos
gh (hypsiconchic) elevations to moderate elevations (meso-
chonicic). Measurements of the length and breadth of the nasal
aperture yield indices within the medium (mesorrhinic) to nar-
row (leptorrhinic) categories for both sexes. But palatal shape
and size differ significantly between the sexes, females more
often possessing wider palates (brachystaphylinic) than males.
Anterior projection of the upper and total facial regions is rare
(prognathic) since straight profiles are in highest frequency
Morphological features also distinguish the cranial speci-
mens from several Harappan sites from the one at Chanhu-daro.
Harappan skulls are generally robust with vertical frontal ori-
entation, larger nuchal muscular attachments at the back of the
head and temporal lines that are prominent and sweep posteri-
orly from the sides of the frontal bone to the parietal bones.
Nasal sills are sharp (origmocraspedotic). Skulls are generally
large for both males and femailes. These craniofacial properties
are not in high frequencies among ancient and modern popula-
Copyright © 2012 SciRes.
K. A. R. KENNEDY, G. L. POSSEHL
tions of African ancestry (Gill & Rhine, 1990; Byers, 2002),
Dr. Diane E. Hawkey er participation in this
t, as noted above, certain African features appear in the cra-
nium from Chanhu-daro. Her manner of death is unknown. At
the time of this writing the Chanhu-daro cranium is lost, and
efforts to locate it at the University of Chicago, Case Western
Reserve University and the University of Pennsylvania have not
The human cranium from the Mature Harappan site of
Chanhu-daro presents archaeologists with a unique funerary
practice by the Indus peoples. While jar burials have been en-
countered in Early and Mature Harappan sites and continued
into the Post-urban Jhukar culture, the Chanhu-daro cranium
indicates it was a secondary burial. The cranium is not associ-
ated with postcranial bones of the original body. These had
been disposed of elsewhere, possibly at a significant time prior
to the insertion of the cranium into the jar.
Anthropometric and morphological analyses indicate that the
cranium belonged to a young female in relatively good health
as based upon absence of skeletal and dental markers of a
pathological nature. The manner of her death is unknown. The
hypothesis that the Chanhu-daro female was of African ances-
try is supported by the results of comparative data from ten
cranial-bearing from ancient and modern South Asia.
The notion that individuals of African descent were present
in centers of the Indus Civilization is congruent with an emerg-
ing story of maritime activity in the Arabian Sea and the Ara-
bian Gulf. It appears that Indus sailors, or others, reached the
Horn of Africa where they acquired millets, adapted to a hot,
wet summer growing season and brought millets to the Subcon-
tinent. It may well be that the Chanhu-daro cranium informs us
that Africans joined the Indus peoples on their sail back home
and were resident in the Indus Valley during the second half of
the third millennium BC.
is thanked for h
storical study. The senior author was provided financial sup-
port from the National Science Foundation, the American In-
stitute of Indian Studies, the American Institute of Pakistaln
Studies and the Smithsonian Institution. Dr. Possehl’s was
funded by the Anthropology Museum of the University of
Pennsylvania in Philadelphia, the American Institute of Indian
Studies and other funding agencies.
Bergmann, C. (1847). Uber die verhaltnisse der warmeokonomieder
tiere zu ihrer grosse. Gottinger Studien, 3, 595.
Breasted, J. H. (1944). Ancient times: An introduction to the study or
ancient history and the career of early man. Boston: Ginn and Com-
Byers, S. N. (2002). Introduction to forensic anthropology: A textbook.
Boston: Allyn and Bacon.
arter, R. A. (2001). Saar and its external relations: New evidence for
interaction between Bahrain and Gujarat in the early 2nd millennium.
Arabian Archaeology and Epography, 12, 183-201.
Caspers, E. C. L. (1989). Was the dancing girl from Mohenjo-daro a
Nubian? Rome: Note Discussionni.
Cleuziou, S., & Costantini, L. (1980). Premiers elements sur l’agricul-
ture protohistorique de l’Arabie orientale. Paleorient, 6, 245-251.
Cleuziou, S., & Tosi, M. (2000). Ra’s al-Jinz and the prehistoric coastal
cultures of the Ja’alan. Journal of Oman Studies, 11, 19-73.
Costantini, L. (1990). Ecology and farming of the protohistoric com-
munities in the cen gre (Ed.), The
Bronze Age cultu -hada (Republic of
me contacts during the Bronze
Age: Scientific study of pottery from Ras al-Junayz (Oman). Admatu,
Guha, B. S. (1928). Neg e, 121, 793.
tral Yemeni highlands. In A. deMai
re of hawlan at-tiyal and al
Yemen) (pp. 187-204). Rome: Instituto Italiano per il Medio de
Estremo Oriente, Centro Studie Scavi Archeologici.
Crawford, H. E. W. (1998). Dilmun and its gulf neighbors. Cambridge:
Cambridge University Press.
Dutta, P. C. (1983). The Bronze Age harappans. Calcutta: Anthropo-
logical Survey of India.
Gill, G. W., & Rhine, S. (1990). Skeletal attribution of race: Methods
for forensic anthropology. Anthropological Papers of the Maxwell
Museum of Anthropology 4.
Gogte, V. D. (2000). Indo-Arabian mariti
rito strain in India. Natur
Guha, B. S. (1929). Negrito strain in India. Nature, 123, 342-343.
Guha, B. S. (1935). The racial affinities of the peoples of India. Census
of India 1931. Simla: Government of India Press.
Harlen, J. R., de Wet, J. M. J., & Stemler, A. (1976). Plant domestica-
tion and indigenous African agriculture. In J. R. Harlan (Ed.), Origin
of African Plant Domestication (pp. 3-19). The Hague: Mouton.
arlen, J. R. (1992). Indigenous African agriculture. In C. Cowan, C.
Wesley, & P. J. Watson (Eds.). The Origins of Agriculture: An in-
ternational perspective (pp. 59-70). Washington DC: Smithsonian
Hawkes, J. (1982). Adventurer in archaeology: The biography of sir
mortimer wheeler. New York: St. Martin’s Press.
awkey, D. E. (1998). Out of Asia: Dental evidence for affinity and
microevolution of early and recent populations of India and Sri
Lanka. Ph.D. Dissertation, Tempe, AZ: Arizona State University.
Hawkey, D. E. (2004). The peopling of South Asia: Evidence of affini-
ties and microevolution of prehistoric populations of India and Sri
Lanka. Spolia Zeylanica 39. Sri. Lanka: Colombia National Muse-
ums of Colombo.
emphill, B. C, Lukacs, J. R., & Kennedy, K. A. R. (1991). Biological
adaptations and affinities of Bronze Age Harappans. In R. H.
Meadow (Ed.), Harappan excavations 1986-1990: A multidiscipli-
nary approach to third millennium urbanization (pp. 137-192).
Madison, WI: Prehistory Press.
Keita, S. O. Y. (1988). An analysis of crania from tell-duweir using
multiple discrimiant functions. American Journal of Physical An-
thropology, 75, 375-389. doi:10.1002/ajpa.1330750309
Kennedy, K. A. R. (1976). Human variation in space and time. Du-
buque, IA: William C. Brown.
ennedy, K. A. R. (1995). But professor, why teach race identification Knces, 40,
if races don’t exist? American Journal of Forensic Scie
Kennedy, K. A. R. (2000). God-apes and fossil men: Palaeoan
pology of South Asia. Ann Arbor, MI: University of Michigan Press.
Knussman, H. R. von R. (1988). Biologie des menschen, ban 1, wis-
senschafts—Theorie, geschichte, morphologishe methoden. Stuttgart:
Gustav Fisher Verlag.
Krogman, W. M., & Sassaman, W. H. (1943). Skull found at Chanhu-
daro excavations. In E. J. H. Mackay (Ed.), Chanhu-daro excava-
tions 1935-1936 (pp. 252-263). New Haven, CT: American Oriental
Lapique, L. (1905). Note sommaire sur une mission ethnologique dans
le Sud de l’Inde: La race noire predravidienne. Bulletin de Museum
d'Histoire Naturelle, 11, 283-285.
Lee, A., & Pearson, K. (1901). A first study of the correlations of the
human skull. Philosophical Transactions of the Royal Society, 196,
Copyright © 2012 SciRes. 179
K. A. R. KENNEDY, G. L. POSSEHL
Copyright © 2012 SciRes.
Bulletin of the Museum of Fine Arts, Boston, 34,
Mand the Indus civilization, being an
Von Linnaeus, C. (1785). Systema naturae (10th ed.) London: Lack-
ington and Allen.
ackay, E. J. H. (1930-1934). Annual report of the archaeologi
survey of India 1930-1934.
ackay, E. J. H. (1936). Excavations at Chanhu-daro by the American
school of Indic and Iranian studies and museum of fine arts, Boston.
ackay, E. J. H. (1931). DK area. In J. Marshall (Ed.), Mohenjo-daro
and the Indus civilization (pp. 233-261). London: Arthur Probsthain.
ackay, E. J. H. (1943). Chanhu-daro excavations, 1935-1936. New
Haven: American Oriental Socie
Majumdar, N. G. (1934). Explorations in sind: Being a report of the
exploratory survey carried out during the years 1927-1928,
1929-1930, 1930-1931. Memoirs of the Archaeological Survey of
arshall, J. (1931). Mohenjo-daro
official account of archaeological excavations at Mohenjo-daro car-
ried out by the government of India between the years 1922 and 1927.
Martin, R. (1928). Lehrbuch der anthropologie (2nd ed.) Jena: Fischer.
Mehra, K. L. (1963). Considerations of the African origin of Eleusine
coracana (L.). Current Science, 32, 300-301.
ery, W. (2000). Les ceramiques d'oman et l'asie moyenne: Une ar-
cheologii des exchanges a l'age du bronze. CRA Monographies 23.
Paris: CNRS Editions.
Morant, G. M. (1922-1923). A first study of the Tibetan skull. Bio-
metrika, 14, 193-260.
ppenheim, A. L. (1954). Seafaring merchants of Ur. Journal of the
American Oriental Society, 74, 6-17. doi:10.2307/595475
Piggott, S. (1846). The chronology of prehistoric north-west India.
Ancient India, 1, 8-26.
Piggott, S. (1950). Prehistoric India. London: Penguin.
Possehl, G. L. (1986). African millets in South Asian prehistory. In J.
Jacobson (Ed.) Studies in the archaeology of India and Pakistan (pp.
237-256). Delhi: Oxford and IBH and the American Institute of India
Possehl, G. L. (1996). Meluhha. In J. E. Reade (Ed.), The Indian Ocean
in Antiquity (pp. 132-208). London: Kegan Paul International in As-
sociation with the British Museum.
Possehl, G. L. (1997). Seafaring merchants of Meluhha. In B. Allchin
(Ed.), South Asian archaeology (pp. 87-100). Delhi: Oxford and
Possehl, G. L. (1999). Indus age: The beginnings. Philadelphia, PA:
University of Pennsylvania Press.
Possehl, G. L. (2002). The Indus civilization: A contemporary perspec-
tive. Walnut Creek, CA: Altamira Press.
Risley, H. (1908). The people of India. Calcutta: Thyaacker, Spink and
Risley, H., & Gait, E. A. (1903). Census of India 1901
Appendices 1. Calcutta and Simla: Government of India.
uff, C. (2002). Variations in human body size and shape. Annual
Review of Anthropology, 31, 211-232.
Sns and dispersals. New York:
tion of animals and foodstuffs. Cambridge, MA: MIT Press.
Universitario Orientale, 45, 585-
St India: A study of eschato-
n subcontinent (pp.
W : The defences and cemetery R37.
Sahni, D. R. (1926-1927). Mohenjo-daro. Annual Report of the Ar-
chaeological Survey of India 1926-1927:60-88.
auer, C. (1952). Agricultural origi
American Geographical Society.
er, C. O. (1969). Seeds, spades, hearths and herds: The domestica-
Shar, G. M., & Vidale, M. (1985). Surface evidence of craft activity at
Chanhu-daro. Annali dell'Instituto
ingh, P. (1970). Burial practices in ancien
logical beliefs of early man as revealed by archaeological sources.
Thurston, E. (1909). Castes and tribes of central India. Madras: Gov-
Todd, T. W. (1932). Preface. Punjab Gazetteer (Attock District)
XXIX-A, 1930. Lahore: Government of India.
ishnu-Mittre (1974). The beginnings of agricult
evidence in India. In J. Hutchinson (Ed.), Evolutionary studies in
world crops: Diversity and change in the India
3-30). Cambridge: Cambridge University Press.
heeler, M. (1947). Harappa 1946
Ancient India, 3, 58-130.
right, R. P. (2011). The ancient Indus: Urbanism, economy a
ety. Cambridge: University of Cambridge Press.