Arabiens arkæologi

Resumé

Arabian Gulf Archeology

The Eleventh Campaign of the Danish Archeological Expedition 1965

On the 25th of January 1965 the advance party of the 11th Danish expedition to the Arabian Gulf arrived in Bahrain to reopen the camp within the ruins of the Portuguese fort, before the main party, following nine days later, arrived on the 3rd of February. This comparatively late start to the season's work was made in order to avoid the month of Ramadhan, with its fasting, shortened working hours and subsequent public holidays. To compensate for the late start the expedition continued working throughout April and only terminated on May 5th, considerably later in the year than it had ever ventured to work before. The weather, however, was unusually mild, the customary April heat-wave with its foretaste of summer failed to materialise, and only on one day did the temperature exceed that of the human body.

The expedition, under the direction of Professor P. V. Glob, assisted by T. G. Bibby, was reduced this season to twelve persons, as, with the cessation of field work in Qatar, excavation was to be confined to Bahrain and Abu Dhabi. Generous financial support was provided, as usual, by oil companies in these two lands, by the Government of Bahrain, the Danish Carlsberg Foundation, and other well-wishers. With the assistance of the Arabian American Oil Company a reconnaissance was once more made in the Eastern Province of Saudi Arabia. The supervision of our commissariat in Bahrain was once more in the capable bands of Lillian Bue-Madsen; the remaining members of the expedition are recorded in the reports of the individual excavations.

Bahrain

The season's excavations in Bahrain were planned as a continuation of the two-year projects begun in 1964, and were in the main confined to the tell on the north coast of the island, known as Qala'at al-Bahrain, on which the expedition has dug every year since 1954. The only other excavation of significance was that of an undisturbed burial mound brought to our attention by amateur archeologists of the Bahrain Petroleum Company.

Qala'at al-Bahrain. In the central area of the tell, Karen Frifelt and Henning Nielsen continued the area excavation begun the previous year, taking it down now through the "Kassite" and "Barbar" levels.

First, however, a careful investigation was made of the better preserved of the two clay benches with built-in pottery vessels (Kuml 1965). The theory that they might be multiple bellows was strengthened when it was again confirmed that there was no stoking system within the structure, while lumps of bronze and considerable traces of firing were found everywhere around the benches. During their removal it was clearly determined that both benches had been built up over earlier structures of the same type, even the impress left by the pottery vessels being visible. The third and earliest structure had been considerably smaller, with only two vessels.

In the western area of the excavation the "corridor" and the adjacent earth-closet of the "palace" building was removed, exposing the easternmost room and the east wall of the large Kassite building (Kuml 1964 p. 102). Experience had shown that at the floor level of this building a burnt layer several cms. thick could be expected, with possibilities for important discoveries; picks and shovels were accordingly kept so far as possible in the background, even though the layer was hard and difficult to dig by band. Our expectations were not disappointed. In the northeast corner of the room, opposite the door in the south wall, stood a cut limestone block, cubic in shape but split in two places from top to bottom, so that it now consists of three pieces of approximately equal size. It is uncertain whether this splitting, and some irregular chipping off of pieces from the top, was the result of deliberate destruction. Around this block, and firmly embedded in a hard layer of bitumen, there were found several fist-sized stones of a reddish-purple colour, some lumps of a red colouring matter, a quantity of potsherds of Kassite vessels, and two complete and several fragmentary tablets of unfired clay with cuneiform inscriptions; in addition there were two damaged seal-impressions in clay.

In the same room, to the left of the doorway, there was a circular "patch" in the cement floor. This proved to be a cement lid; below it, and further sealed below a layer of coarse gravelly cement and a layer of fine hard cement, lay a pottery vessel firmly cemented into the floor. But here expectations were disappointed; the vessel contained only fragments of cement and gypsum, a one cm. thick layer of sandy clay, and at the bottom a very thin layer of light sand containing two minute potsherds and two scraps of bone. The vessel was of hard-fired red clay with a slightly blackened red surface; sherds of the same type have previously been found together with the characteristic tall beakers of the Kassite period.

In the remainder of the excavation area the Kassite level was disturbed by several large pits, possibly dug in Islamic times and doubtless with the object of acquiring the imposing blocks of stone of the type of which the surviving Kassite building is constructed. An abandoned block of this character was occasionally encountered in the "fill", and it was always a Herculean task to remove it from the depths of the excavation without undue damage to section walls and fingers.

Below the Kassite level came thick occupation levels of the Barbar period, full of animal bones and sherds of Barbar ware, together with the remains of fairly light walls of fist-sized stenes. A single stamp-seal (fig. 2)-the first of the otherwise so well-known Bahrain/Failaka type to be found in tbis central area of the tell-was the only interesting small-find made. These Jevels, too, had been disturbed at many points by Jater pits. A broad band, over 1,5 metres wide, of hard compacted green sand, such as is employed as foundation under walls of the Barbar period, ran in a curve across the excavation area, and one of the later "robber" excavations could be followed very exactly along this "lost wall"; close in to the wall of the Kassite building the broken-off stump of this same wall appeared. It was of limestone and gypsum conglomerate set in a very hard gypsum-like cement, and ran in under the Kassite building, its foot being on a level with the bottom of the foundation of the Kassite wall.

(The above based on a report by Karen Frifelt)

At the northern city wall Svend Bue-Madsen and Else Roesdahl continued their investigations of the previous year (Kuml 1965), in the narrow area between the inner face of the wall and the deep Islamic excavation further south.

It will be recalled that the year before the area had been divided into an eastern and a western section, with a two-metre baulk between, and that this baulk was found to coincide almost exactly with a well-constructed road of City V (Seleucid) date running north to a blocked gate in the city walI. The baulk had been excavated down to this road level, while in the sections on either side excavation below the Seleucid level had begun to reveal walls and occupation debris of City II ("Barbar") date.

In the course of the 1965 season the baulk was reestablished, involving the removal of the walls edging the Seleucid road and cutting through the solid foundation of the road itself. Excavation was then continued in the east and west sections, the sides of the baulk being drawn as they were revealed and the baulk itself excavated layer by layer, all potsherds being kept. During the period of excavation it proved possible to excavate the whole of the east section, the baulk and the 2½ metres of the west section nearest the baulk down to bedrock, while in the remainder of the west section some 1-1 ½ metres of soil were left unexcavated.

Where the Seleucid road, with its drain-system, met the city wall there was a pair of limestone hinge-holes for a gate (fig. 3, port 3A) similar to those discovered in the bathroom in 1964, and this system proved to have a long ancestry, as beneath it lay the city gates of the earlier periods. In the uppermost "Barbar" levels not very much deeper appeared two bluish-black diorite hingepost-stones (fig. 3, stolpesten-port 2B) against the inner (southern) face of the inner wall, and beneath these came further gateways (Port 1, port 2A), the various phases of which are not yet worked out with certainty.

In the western section a well, partly stone-plundered, was excavated and emptied. Beside it was a trough of the cement-hard local gypsum, with another similar trough immediately below it in the previous building phase. Together the well and trough have occupied a little roadside halt where animals could be watered before or after passing the gate.

Several building phases could be distinguished in excavating the houses on either side of the road (which persisted through all phases). In some cases walls had been in use through several phases and were correspondingly high, but the phases were revealed by floor levels and by the bases of other less persistent walls.

A definite break in the sequence appeared -most obviously to the east of the baulk- about 1½ metres above bedrock. Here was a thick levelling layer, and below it the house-walls were built of somewhat larger stones than those above the layer though the material was the same, the sea-bed limestone conglomerate known locally as farush. Where house-walls above the layer coincided with those below they were almost always separated by a thin clay layer, while at four points the city wall was built above the surviving courses of house-walls of this period. In association with these walls four pottery vessels without bottoms were found in the east section, three of them halfway under the city wall; they may be bread-ovens. This evidence, still somewhat slender, suggests that the stone fortification wall of the city was not built at the same time as the first stone buildings, but relatively shortly thereafter.

Below the levelling layer were characteristic dark occupation levels, with numerous "chain-ridge" sherds (Kuml 1957). The lower levels were completely without Stone walling, and the last half metre above bedrock consisted of beach sand, also containing potsherds and scraps of copper. As the bedrock, here composed of horizontal layers of the farush conglomerate, was seen to contain embedded potsherds, a portion of it, along the line of the city wall in the east section, was broken up, and the potsherds were found to persist, between and embedded in the layers, to a further depth of 30 cms., below which the farush ended in a thin layer of coarse gravel followed by sterile green sand. The potsherds from the farush levels, while containing a small admixture of "chain-ridge" ware, were in the main of a completely new type, a fairly thick straw­tempered ware with massive outturned rims.

The season was rich in "small-finds", including 9 stamp-seals, one with the remains of the cord in its suspension-hole, a seal impression on a typical Barbar sherd, and a clay bulla with an impression of the same type as that found on the seals (fig. 4). There was also a cubical steatite weight of the type known from the Indus Civilization, and a very richly varied potsherd material, including sherds of beakers with elegantly painted gazelles (fig. 5).

No seals were found below the levelling layer recorded above, whereas in the occupation levels below there were recovered a 31 cm. long spearhead of copper (fig. 6) and a little terracotta female figure (fig. 7).

(The above based on a report by Else Roesdahl)

In an attempt to locate the eastern wall of the Barbar-period city, Peter Crabb and Malene Møller-Jensen cut two trenches about 80 yards south and a little to the east of the excavation of the northern city wall. In the eastern trench traces of Barbar-period occupation were entirely absent, and the tell consisted here mainly of City V (Seleucid) remains. Besides the Iocal green-glazed ware imported Greek black-varnished sherds were encountered.

In the western trench a well-preserved industrial establishment was partially cleared, consisting of a series of regular elevations in the cement floor of a narrow room between ashlar walls preserved to a height of over two metres. These cross-ribs measured 25 cms. in both directions in cross-section and occurred at intervals of about 50 cms. (fig. 8). Apart from the thin-walled sherds of the City IV (snake-burial) period, the space between the ribs was occupied by decayed plant remains. The nearest parallel is to the modem floors for the extraction of date juice, with the important difference that the ancient floor lacked any kind of outlet for juice.

Well above the floor level of this building a large green-glazed vessel with tall neck and single handle was found standing beside a large shallow dish showing a lighter glaze (fig. 9).

(The above based on a report by Peter Crabb)

Burial mound-southermost moundfield. A mile and a quarter due west of the southern tip of Jebel Dukhan, the central hill of Bahrain, on the summit and westward-facing slopes of the rim-rock which rings the central depression of the island, lies an isolated group of burial mounds, an outlier of the moundfields which so thickly cover the northern and northwestern parts of Bahrain. To this moundfield we were taken, one Friday towards the end of the expedition, by Cliff Bernett and Harvey Blackmore of the Bahrain Petroleum Company, who brought the news that they had found an undisturbed burial in the central chamber of one of the mounds. This was indeed news, because in the course of excavating some fifty of the Bronze-Age burial mounds of Bahrain (Kuml 1954, 1964) we had in no case found a completely undisturbed grave, and only two (Kuml 1954, p. 119 and an unpublished "rescue" dig at Buri in 1959) in which the position of the skeleton could be even partially determined. In the course of the following week the chamber was exposed and excavated by Jørgen Lund and Malene Møller-Jensen (fig. 10).

The stone chamber in the centre of the mound proved to lie between NE and ENE (about 60° from N); it measured 183 cms. in length and 65 cms. in width at the western end, perhaps 70 cms. at the eastern end. These were dimensions at floor level, but the walls, of dry-stone coursing and 78 cms. high, sloped inwards to a width of not much more than 30 cms. at the top where the chamber was covered by five capstones. At the eastern end there was a single recess (not two, as is most commonly found) in the northern wall, 25 cms. wide and 32 cms. deep, and Jess than the full height of the chamber.

The skeleton, the long bones and skull of which were well preserved, lay in the centre of the chamber with its head towards the eastern end. It lay on its right side facing north, with its legs sharply drawn up so that the bones of the lower legs were practically at right angles to the axis of the chamber, and its hands in front of the face, the fingers seemingly covering the features. This posture, rather a "kneeling" position than either "flexed" or "contracted", agrees precisely with that of the surviving portions of the skeleton shown in Kuml 1954 (p. 119). (The Buri skeleton, on the other hand, Jay on its left side, with the upper leg bones at right angles to the axis of the chamber and the backbone and lower leg bones somewhat oblique -a much more conventional "flexed" position; it was accompanied by a red ridged Barbar-type jar).

The present body bore no surviving ornaments or other objects, and the only grave goods were a round-bottomed vase of red clay, 23 cms. high, which had been removed for safe keeping by the discoverers, but which had lain, in large fragments, in the alcove above the head of the skeleton.

Abu Dhabi

Umm an-Nar. The Abu Dhabi team, consisting this year of Jørgen Lund, Claus Malmros and Niels Bech, the last-named as architect, spent the period from 25th February to 20th March on Umm an-Nar. Work, carried out with the aid of up to 17 local workmen, was concentrated on the settlement on the island, the aim being to obtain clearly stratified material from the thickest deposits in the settlement and thus determine whether the settlement had been occupied during more than one cultural phase. A north-south trench was accordingly dug clean across the settlement in a width of about 21/4 metres at the point where the depth of deposit seemed likely to be greatest. Half the width of the trench was dug all the way down to bedrock, walls, objects of stone and copper, and characteristic potsherds being plotted in. The sides of the trench were then drawn, and the remainder of the trench dug as accurately as possible by layer, all potsherds being kept. It is not possible at this stage to present the results of the excavation, but the indications point to a comparatively short occupation period, the sherd material showing considerable uniformity from top to bottom of the section.

A hundred metres north of the section trench a smaller trial excavation was made on another of the projecting rock ridges of the settlement. Here, in addition to a quantity of potsherds, a house was identified, seemingly of similar type to that excavated in 1962/63. In the western end of the house an almost whole jar was found, bottom upwards and with its mouth stoppered with a lump og clay. In the same end of the house a copper chisel lay beside a threshold stone. The house showed only one building phase.

Buraimi. After 3-4 weeks' work on Umm an-Nar the team left on 21st March for the oasis of Buraimi, pitching its camp a mile to the east of the village of Hili. The aim this season was to complete the excavation of the stone circle commenced the year before. With the help of 12 local workmen the whole structure was now excavated. one quadrant at a time in order to preserve sections through the structure. These sections were not drawn, as they added nothing to those obtained the year before. All potsherds were collected separately from each quadrant.

After the whole structure was excavated it was drawn, photographed and described (fig. 1). There can now no longer be any doubt that the structure is a burial chamber of the same type as that known from Umm an-Nar, as traces of interior walls were found in the much disturbed interior. Sherds of several hundred pottery vessels were found, the majority being of red ware with geometrical patterns in black paint. The shapes correspond closely to those of Umm an-Nar, though some new types are here met with. The grey ware already known from Umm an-Nar was also prominently represented, particularly noteworthy being jars with painted friezes of antelopes or goats in black [1], and tall vessels with incised designs, including the "doorway" design known on steatite vessels from Mesopotamia to the Indus (figs. 11-12). Another reminiscence of the Indus was the presence of fragments of compartmented steatite vessels, while a further short copper sword was also this year found (fig. 13). It proved impossible, without heavy machinery, to lift the large stones of which the chamber had been built. If opportunity arises this should be done, as undoubtedly many potsherds still lie beneath them, while a relief carving of an oryx -with difficulty discernible on one stone overlaid by another stone too massive to be moved- suggests that further reliefs might be found thereby.

We had this year the pleasure of frequent visits from Sheikh Zaid, who showed great interest in our work. This interest was further evidenced when his secretary took us to a large tell a little to the southwest of Hili. On the surface of the tell lay quantities of fragments of steatite vessels, scraps of copper, and potsherds. Many of the potsherds were of the same type as those found the previous year at Dibba on the east coast of the Oman peninsula (Kuml 1965), while there was a complete absence of Islamic sherds ­and, incidentally, of "Umm an-Nar Culture" sherds. There is, of course, no telling what may lie beneath the surface of this tell. It will be interesting to find out.

(The above based on a report by Jørgen Lund)

Saudi Arabia

In the period 9th to 17th March T. G. Bibby visited Dhahran in Saudi Arabia in order, for the third year in succession, to examine sites and objects discovered by the part-time archeologists of the Arabian American Oil Company. He was also enabled to spend two days in Riyadh where he had the pleasure of a long discussion of aims and methods with the Director, Mohamed al-Ibrahim, and the staff of the newly-constituted Department of Antiquities and with the Deputy Minister of Education, His Excellency Dr. Abd al-Wahhab. (The enthusiasm of the new Department is well shown in the fact that its Director shortly afterwards visited Bahrain and spent five days on our digging sites).

In the course of this year's reconnaissance a further visit was paid to the central site on the island of Tarut (cf. Kuml 1965), where among the Barbar sherds lying thickly on the surface of the mound below the Turkish fort one sherd of "chain-ridge" ware was picked up, showing this site to be even earlier in origin than already attested.

A reconnaissance trip to the area of Jubail and Jebel Berri, sixty miles northeast of Dhahran, discovered a limestone outcrop, now half drowned in sand, where large stone blocks had once been quarried in considerable numbers and where a quantity of rock carvings of animals, fish-traps and geometric figures covered the upper slopes.

One day was spent in the area fifteen miles north of Uqair where air photographs showed an expanse of several square miles covered with ancient irrigation ditches and, at one point, the clear outlines of a small fort. On examination on the ground the whole area proved to be sabkha (salt-flats), the irrigation-ditch bottoms being raised above the general level, presumably because the clay deposited in the ditches had been more resistant to wind erosion than the fields they had once served. The fort was difficult to find as it was almost completely drowned in the sabkha, the rampart mounds projecting scarcely two feet above the surface. Nevertheless surface collection among the irrigation ditches and within the fort gave a surprisingly rich result, comprising a barbed and tanged flint arrowhead, two leaf-shaped arrowheads of bronze, 5 cms. long, a large number of beads of carnelian and bloodstone, a circular bead of bone and a large number of potsherds. Objects and potsherds of Islamic type were conspicuously absent, and the site was unanimously labelled "Gerrha" (with firmly inverted commas), the well-known but still undiscovered Greek city on the East Arabian coast, it being agreed that the actual city could well lie beneath the 30-foot sand dunes which sweep down to meet the sabkha not much further inland.

Geoffrey Bibby