foundation 6 (Fig. 19.4). The massive occurrence in
the east-west running foundation of light brown
plaster is a prominent feature of Temple III walls (26).
The platform fill of Temple III, layer 4, reaches up
to level 90 in the North-South Section and the
corresponding layer 4 in the old Main Section to
level 80 (27). This would indicate a minimum top
level for the surface of the oval platform, which is
supported by the fact that the largest plinth-stone is
completely evenly finished on the outer sides, with
the exception of a rough and irregular footing-
portion in level 90, which was presumably set into
the ground (28). This gives a height of the terrace
wall, including the foundation, of 1.90 m. The
distance to the central platform is c. 13 m (Fig. 19).
About 12 m east of the North-South Section a new
trench was dug in 2004 in order to trace the
continuation of the Temple III oval terrace wall. In
the nearby old section 31 the wall itself has
disappeared, leaving a well-defined robbers’ pit
with the centre in 58.2 m N (29), and the same
appeared in the new section, this time with the
centre of the robbers’ pit in 59.4 m N. The bottom of
the pit is at level 2.70, corresponding to the base of
the foundation further west. On both sides of the
robbers’ pit were thick layers of oolitic limestone
chippings, residue from the final dressing of the
ashlars of the terrace wall. A similar occurrence of
oolitic chippings was noted in 2004 at the base of
Temple IIb’s ashlar wall (Fig. 3.11), and it is inter-
esting to note that there, the lowest-lying chippings
measured up to 10 cm across, whereas the upper-
most chippings only measured up to 5 cm, indica-
ting a first rough dressing followed by a fine one.
Another stretch of wall of similar dimensions,
levels and building technique, excavated in 1959
west of the pool (30), can now be interpreted as the
foundation for the western terrace wall of Temple
III. The distance to the central platform is in this case
c. 14 m. Whether the lower platform of Temple III
was oval or round is not possible to ascertain before
further excavations are made east of the temple, and
this actually also applies for the lower platforms of
Temples I and II!
Temple IIb oval platform
A short stretch of the northern oval terrace wall of
Temple IIb had already been excavated in 1960 in
section 31, but without being properly identified
(31). In 2004 this piece of wall was re-excavated and
shown to be part of the Temple IIb terrace wall. In
section 31 the wall (A) delimits the fill of the oval
platform of Temple IIb. Onto this piece of wall a
semicircular structure was attached (Fig. 20), similar
to the oval structure attached to the terrace wall of
Temple IIa towards the west (32). The building
technique and dimensions correspond to the oval
walls of Temples IIa and b in the west: use of uncut
stones varying from hand-size to 30 cm across and
laid in clay. The oval structure to the west of the
temple had a floor at least 0.5 m under the assumed
surface of the oval platform, and likewise the clayey
floor of the semicircular structure was sunk into
subsoil sand to level 3.99, more than 1 m under the
Fig. 19.
Interpretation of North-South Section north of the temple (Fig. 18). 1: northern terrace wall of Temple III. 2: plinth stone north of the
central platform of Temple III. 3: minimum height of oval terrace of Temple III. 4: oval terrace wall of Temple III. 5: round stone block
with central hole. 6: oval terrace of Temple IIb. 7: oval terrace wall of Temple IIb. 8: north wall in semicircular structure attached to 7.
9
: subsoil sand. 10: layers dating to Temple I/IIa. 11: layers dating to Temple IIa. 12: layers dating to Temple IIb. 13: huge pit cutting
10–12, along the North-South Section. 14: bottom of huge pit east of the North-South Section.
BARBAR TEMPLE, BAHRAIN
117
platform surface. The continuation of the terrace
wall both east and west was disturbed and could
therefore not be followed in the sections.
The semicircular structure contained clayey sandy
fill with animal bones and pottery dating to Temple
IIb, e.g. an unusual import from the Indus region, a
rough-surfaced mould-shaped base (Fig. 21), as well
as a fragment of an altar in oolitic limestone (Fig. 22)
and a rim sherd from a finely polished stone jar
(Fig. 23).
Seal-stamped tokens
On the surface of the Temple IIb oval platform at
level 2.00–2.30 two parallel walls were excavated in
1960, surrounded by ashy layers (33). From this area
Fig. 21.
Pottery found in semicircular structure.
Fig. 22.
Fragment of stone altar.
Fig. 20.
Semicircular structure attached to Temple IIb oval wall, from the north (2004).
Fig. 23.
Fragment of stone jar.
F. HØJLUND ET AL.
118
came six seal-stamped tokens and from a neighbour-
ing trench to the west stems a further seal-impressed
token (34). The same ashy layer occurs as layer 7 in
the North-South Section (Fig. 18), and c. 25 cm above
this layer, in layer 4, a seal-impressed token was
found (x1). Another token (x2) was found in the
surface layer of the north-south sounding at c. 63 N.
Token no. 692x1 (Fig. 24).
Diameter of the obverse 17.5 mm, maximum thick-
ness 7.5 mm, edge thickness 4 mm. Light brown,
baked clay. Reverse convex, surface plain, vertical
edge.
Obverse: Two standing nude men with legs crossed
above the knees. Both of them are looking outward
and are grasping a crescent enclosing a big bird
between and behind them. Their free hands, with
three prominent, pointed fingers, are raised. The
space between their elbows and legs is filled with
two prominent sun rings.
A very similar motif with two crossed, nude men
grasping ritual emblems is present on a bifacial seal
from Failaka (35). The design of the hand is
exceptional, but known from a seal from Saar
Settlement (36). Sun rings are known from more
motifs of ritual character in style 1A2–3 from Failaka
(37). Birds in crescents are not known from other
seal motifs than here, but often crescents or crescent
staffs are flanked by birds, which is, however, often
the case with other emblems (38).
Token no. 692x2 (Fig. 25).
Diameter of the obverse 17.5 mm, maximum thick-
ness 7.5 mm. Hemispherical token, surface plain.
Light brown, baked clay.
Obverse: The details of the motif have flaked off,
leaving only the traces of one horizontal and four
vertical, shallow depressions, which resembles the
impression on three tokens found earlier at Barbar
and one token found at Saar Settlement (39).
Token 692x1 differs from the previously found
tokens from Barbar in its vertical edge, while 692x2
has the same hemispherical form as the other tokens
from Barbar. Both tokens differ, however, from the
other tokens from Barbar in their size of only
17.5 mm in diameter against 20–21 mm and a height
of 7.5 mm against 10 mm, which is the norm for the
others. In view of the great uniformity in size of all
the other tokens, it is possible that not only the
stamp, but also the size of the tokens bears witness
to a different origin (40).
The Temple III pool chamber and the northern foreland
The main objective for the sounding north of the
temple was to search for a pool chamber related to
Temple III. As suggested earlier, the presence of two
plinth-stones north of the Temple III platform (41)
points to a likely position of a staircase leading
down to a pool related to Temple III, in analogy with
the plinth-stones along the pool staircase of Temple
II (42).
A further analogy which became clear during the
2004 campaign is the presence of a round stone block
with a central hole placed just outside the Temple III
oval terrace wall and on a line right between the two
plinth-stones (Fig. 19.5). This stone can be compared
to a structure of similar shape and dimensions
placed under the staircase leading down to the
Temple II pool, right where the staircase is met by
the Temple IIa oval terrace wall (43). The function of
these structures is unclear, but Andersen’s sugges-
Fig. 25.
Seal-stamped token, eroded.
Fig. 24.
Seal-stamped token.
BARBAR TEMPLE, BAHRAIN
119