Gravrøser på Umm en-Nar

Resumé

Burial Cairns on Umm an-Nar

In the sheikhdom of Abu Dhabi, on the Persian Gulf coast of the Oman peninsula, lies the little, naked, rocky island of Umm an-Nar. With its central limestone plateau and its wide sandy beaches, it resembles all the other small islands off the harsh desert of the Trucial Coast.

The Island is separated from the mainland by a strait only 10-20 yards wide, and is now entirely uninhabited, though it is known to the bedouin of the desert and to the inhabitants of the nearest town, the capital Abu Dhabi, because upon its rocky plateau there is a water-cistern, built in recent years, in which rainwater collects during the all too rare showers.

It was P. V. Glob and T. G. Bibby who, during a reconnaissance in the Abu Dhabi area in 1958 in connection with the Fifth Danish Archeological Bahrain-Expedition, were told of the presence of prehistoric remains on Umm an-Nar. A visit to the island revealed a large number of tumuli, all consisting of cairns of stones, and a settlement "tell" 1). Ancient monuments on so isolated an island were a tempting object for closer investigation, and this was begun the following year. The practical difficulties were overcome with help from the Ruler, His Excellency Sheikh Shakbut bin Sultan, and from the oil company, Abu Dhabi Marine Areas, which had first brought the site to our attention. The archeologists were installed in a palm hut on the wide beach of the island, and supplies were driven by car the 12 miles of desert between the island and the little town of Abu Dhabi, whence also came the dozen Arab workers who assisted in the excavation.

The first excavation campaign took place in the months of February and March 1959. In the course of a little over a month Harald Andersen, Mogens Ørsnes and Knud Riisgård investigated one of the largest cairns on the island (Cairn I), worked out the complicated plan of the large mortuary building, and excavated the greater part of its burials. In addition they carried out a short investigation of the settlement site.

The following year the investigations were continued, on this occasion by Arne Thorsteinsson and the author of this preliminary report. During this second campaign, which lasted from 25th January to 25th March 1960, the excavation of Cairn I was completed, and that of Cairn II begun, though the actual burial chambers were not reached. The investigation concentrated upon a series of smaller mortuary buildings (Cairns IV-VIII), which were completely excavated. The campaign ended with a complete registration and description of the 50 tumuli on the island.

The investigation of these burial mounds was continued in November and December 1961, and the excavation of Cairn II was completed. In addition to the previous year's participants, Elise Thorvildsen and Vagn Kolstrup also took part in this third campaign.

Cairn I.

This complex, excavated in 1959, consisted of a circular mortuary building, 11 meters in diameter, with its outer wall resting on a plinth of regularly cut stone blocks. The building possessed a double ring-wall, about 1 meter in width. The outer wall is built of large, carefully shaped and fitted, limestone blocks, built up without the use of mortar in several courses to a height of at least 1.5 meters, and sloping slightly inwards. Behind this ashlar wall an inner ring-wall of flat unshaped stones was built up.

The area within the ring-wall was, as appears in figs. 2-3 occupied by a number of grave­chambers, separated from each other by a complicated wall system. A flagged cross-passage, flanked by walls running north-south, divided the interior into two halves, though the division was not complete as the walls of the passage stopped short at both ends about 1 meter from the ring-wall. The two halves thus formed by the passage were each divided into two equal parts by a wall stretching from the ring-wall to the passage wall and at rightangles to the latter. These four sections were each again divided into two, this time by curved walls, in such a way that in each quarter there were two flagged burial chambers with a common forechamber without flagging.

The discovery of flat stone slabs on the walls or on the floor where they had fallen from the walls showed that passages and chambers had been stone-roofed. It was observed that opposite each end of the north-south central passage there was a break in the inner ring-wall (shown on the plan fig. 3), and as, outside the wall among the fallen stone blocks immediately opposite each of these breaks, a carefully shaped stone block was found with a carved "handle" these stones were interpreted as "door- stones", and the breaks in the wall as the traces of openings giving entrance to the burial-places.

The investigation in 1959 ended with the excavation of the floor-level layers of the central passage and the eastern half of the building. This showed that in the forechambers to the 4 flagged burial chambers there were piles of human bones, while in the central passage and in the actual chambers there were only few scattered bones. Among the skeletal remains lay 3 broken pottery vessels, a number of scattered potsherds and a few small stone beads. During the campaign of 1960 the floor level in the western half of the building was excavated and proved to be considerably more rewarding than the eastern half. Here too only a very small number of pottery vessels and bones were found in the 4 flagged chambers. By far the greater part lay heaped up in the forechambers in a pile against the ring-wall (fig. 4). This circumstance is doubtless best explained on the assumption that the burial chambers could only hold a single burial at a time, and that at each new burial the earlier ones were swept out into the forechamber. To judge by the number of skulls and jawbones at least 15 individuals had been buried in the western half of the building, and of these at least one was a child. Among the bones lay 23 more or less complete pottery vessels (fig. 21), one vessel of green steatite (fig. 21, bottom right) and the base of a very badly preserved alabaster vessel. In addition about 650 beads were found, the majority short and tubular of a grey stone, though a number were of coloured stone of various shapes and sizes. Finally two small bronze or copper fragments were found, one appearing to be part of a pin, while the nature of the other cannot be determined.

Cairn II.

80 meters southeast of the burial complex above described lay Cairn II. Though a little larger, with a diameter of 12 meters, this mortuary building corresponds exactly in form and appearance to the first investigated (fig. 5). The division of the interior differed from Cairn I only in the circumstance that the central passage was here blocked by a cross-wall, so that there was no connection between the northern and southern halves of the building.

The outer ring-wall had consisted in its lowest course of 60 finely cut blocks, of somewhat varying length but all 45-50 cms. high. 42 of them were still in situ. Of the second course only 4 stones were in place, and these were, some few cms. less in height than those of the lowest course. Hidden beneath the sand in a ring around the whole mortuary building lay 310 fallen blocks. If one assumes that no blocks have been removed from the site it is possible to work out from the total length of the blocks that the ashlar wall was originally of 5 courses with a total height of 2.25-2.5 meters. This calculation agrees very well with the height of the best preserved portion of the wall of the central passage, which still stood to a height of 2.24 meters.

Against the north side of the ring-wall just opposite the central passage a stone block was found with a square hole cut on its face large enough for an arm to be inserted. As an exactly similar stone was found opposite the central passage on the opposite side of the building it would be natural to compare these with the two "door-stones" found at Cairn I (fig. 6).

These specially shaped stones had, however, not been the only variation in the smooth surface of the outer wall. Close to the "door-stone" by the south wall 2 stone blocks were uncovered ornamented with animal figures in raised relief. In the centre of the one could be seen a boldly portrayed bull 2). On the other (fig. 7) were carved a camel and another animal which can only with some uncertainty, on account of surface weathering, be identified as a gazelle. Two further stones with reliefs were found fallen from the ring-wall in the southwest. The one (fig. 9) was ornamented with a figure reminiscent of an idol, while the other (fig. 8) showed once more a camel, though this time carved in rather a different way.

While these four relief blocks must have adorned the wall, the placing of yet another relief stone, which was found among the fallen blocks in the southeast, is uncertain. This stone is of the same local limestone as the ashlar blocks, and is of the same length as the other relief stones, but it is of a completely different shape, being half-cylindrical in form. The inner surface of the half-cylinder is smooth, but on the outside it is ornamented with two figures of snakes in boldly raised relief (fig. 10).

Behind the outer ashlar ring-wall stood the inner ring-wall, built, as in the case of Cairn I, of piled, unshaped stone, and joined to the walls bounding the various chambers of the building. All these walls stood vertical up to a height of 70-90 cms., and thereafter project outwards over the chambers they bound, at the top -to judge by the best preserved- approaching each other closely enough to form a corbelled vaulting, doubtless closed at the top with a flat slab. As in the case of Cairn I, there are in each quarter of the building two low cistlike chambers, originally roofed with large flat slabs resting on projecting stones which in most places can still be seen in about 80 cms. height. At one point, above the eastern cist in the southwestern quarter, one of these roofing slabs was still in situ, but elsewhere they had all fallen in.

On excavation all the chambers of the building were found to be full of drift-sand and fallen stones. Below these lay the burial layer with skeletal remains and grave furniture. At some points this layer was clearly marked as a dark-brown layer of decomposed organic material, at other points it was only marked by single bones or scattered potsherds. Also in this mortuary building the traces of burial were very unevenly distributed. While objects were fairly numerous in the southeasterly quarter of the building and in the southern half of the central passage, they were quite rare in the northeastern portion and were almost entirely absent in the north­western part. The southwestern quarter yielded the largest number of objects, and there both the two burial chambers and in particular the area in front of them were full of skeletal remains and artifacts. No order in the remains of the burials could be traced, and the piling up of skeletal remains in the forechamber supported the conclusion that new burials were accompanied by a clearing out of old burials from the cists which the conditions found in Cairn I had suggested.

The pottery found in Cairn II resembles closely that found in Cairn I. Here, as there, the small vase-shaped vessels of red ware with painted ornamentation dominate, but several specimens of the grey ware with black painted ornamentation were also found. The two vessels decorated with incised lines shown in fig. 20 were of a type not previously found. The number of beads was not as large as in Cairn I, but the shapes were the same.

Of particular interest are the 3 daggers of copper or bronze 3) which were found among the skeletal remains (fig. 18). They consist of one dagger, 25 cms. lang, with a solid hilt of square section, and two thin, fragmentary dagger-blades. The one with two rivet-holes at the upper end of the blade has been 18 cms. long, while the other is a complete blade with a narrow centre rib, 19 cms. long.

As Cairns I and II, on account of their size and refined construction, were not considered to be fully representative of the fifty or so gravemounds on Umm an-Nar, the great majority of which were less than half so large and appeared to be built of simple unshaped stones, it was decided to investigate a group consisting of 5 "normal" cairns of slightly varying sizes which lay about 100 yards south of Cairn Il. The mounds were given the numbers IV-VIII (figs. 11-12).

Cairn IV.

This cairn was the largest of the five, being 8.5 meters in outer diameter and having a double ring-wall with both walls built of unshaped quarried stone. The building's division into chambers corresponded closely to that found in Cairn I; it had a central passage, the walls of which did not extend all the way to the ring-wall, and two cross-walls. Here, however, no further sub­division of the 4 sectors could be observed, though this may be due to the difficulty of identifying fallen walls in the tightly packed stone filling of the chambers. No paving was found in the chambers, nor could any observations be made concerning the roofing of the chambers. However, at each end of the central passage a break could be seen in the build-up of the ring­walls, and here too this was interpreted as traces of entrance openings. The investigation of the floor layer gave poor results; only at one point, close to the ring-wall to the east were a few bones found and sherds of 2-3 vessels. Of interest was the fragment of a relief block, used secondarily in one of the inner walls, cf. fig. 14.

Cairn V.

Before excavation commenced the three mounds V, VI and VII could only with difficulty be distinguished, as the collapsed buildings formed one large oval cairn. Cairn V lay southernmost in this complex. The building was only 6.5 meters in exterior diameter, and was, like the others, circular and with a double ring-wall, built upon a plinth of larger quarried stones. The outer wall was built up of carefully shaped blocks of a grey stone laid without mortar. The stones were curved on the outer face and wedge-shaped in order to fit into the sharp curvature of the wall. At one point the outer wall still stood to a height of 80 cms. Here 5 courses were preserved, and it could be seen that the shaped stones were largest in the lower courses, deminishing in size upwards. As it is unlikely that cut stones have been removed from this building the original height of the wall could be calculated on a basis of the fallen stones to have been, with 10 courses, about 1.4 meters, above the 20 cm. high plinth (fig. 13).

Within the ring-wall the burial space was divided into two halves without connection with each other, as a cross-wall spanned the area from ENE to WSW. At rightangles to the cross­wall walls were built out to form the arms of a cross, but without reaching the ring-wall. Each of the two halves of the mortuary building was thus divided into two rooms of equal size, connected by the passage around the end of the arm of the cross. Opposite these wall-ends there was on each side a break in the ring-wall, in fact to the north, where the wall still stood to some height, there was an actual hole large enough for a grown man to creep through. Each of these two separate half-chambers had therefore possessed its own entrance. Large flat stone slabs which lay where they had fallen on the floor had served as roofing for the chambers, and to the north, where the walls still stood tall, it could be seen how the upper courses of the cross-walls and the inner ring-wall projected towards each other to make the interval which the roof-slabs should span as narrow as possible (fig. 15).

All the chambers were paved with closeset flags. On this paving lay a layer 10-20 cms. thick of skeletal remains and artifacts. Everything was heaped together with no semblance of order. In the 4 burial chambers were the remains of at least 36 individuals, of whom at least 3 were children. The number of objects found corresponded to the number of burials. No less than 22 whole pottery vessels were taken up, and sherds of at least as many again. A very weathered alabaster vessel and a hemispherical spindle-whorl formed of the end of a thighbone were also found. In all the chambers there were large numbers of beads, most in the northwest chamber, where it was possible to clear a large group of small tubular beads which lay undisturbed in several rows close besides each other. They were undoubtedly the remains of bead­embroidered clothing (fig. 16). Other important finds were, in addition to some fragments of bronze pins, three simple daggers of copper or bronze 3) (cf. fig. 19 for the two best preserved).

The discoveries made outside the mortuary building were unexpected. Here on the western side, under the fallen ring-wall, lay 4 human skeletons, two with their heads to the south and legs contracted, one outstretched with its head to the north, the position of the fourth skeleton being uncertain. The skeletons lay close up to the wall, the three of them so close that they rested against the plinth (fig. 17). Together with the skeletons, crushed under the fallen stones, lay a number of potsherds and a single unidentifiable fragment of bronze. Also beneath the fallen stones by the plinth lay two quernstones, one of them oval, about 30 cms. long with a smooth-polished surface. Beneath it lay the remains of a woven palmleaf mat or basket.

Cairn VI.

With an exterior diameter of 5.5 meters and a ring-wall where both the inner and outer walls were constructed of unshaped dry-stone walling Cairn VI was far from impressive besides the well-constructed facade of cut stone of Cairn V. Its interior, too, was simpler, being merely divided into two rooms by an east-west cross-wall, which ended at the eastern end 1 meter from the ringwall to leave a passage between the two chambers. As in the case of Cairn V traces were found of an entrance to the grave-appearing as an 80cm. wide break in the ring­wall-immediately opposite this passage. The ring-wall was preserved to a height of 75 cms., and the cross-wall reached a height of 1.3 meters, allowing observations to be made of the roofing of the chambers. It could be clearly seen that both the inner ring-wall and the cross­wall sloped outward, so that the grave-chambers narrowed towards the top. Thus in the southern chamber the gap between the walls was at the top only 1.2 meters while at the floor level it was 1.9 meters. With the walls in their full height the gap at the top would have been even smaller.

The floor of the mortuary chamber was covered with large flat flags except in the fore­chamber and the eastern end of the southern room. On the floor lay skeletal remains pushed together into heaps. The bones were very badly preserved, but it could be seen that at least three individauls had been buried in the southern room and at least two in the northern. Among the bones lay sherds of at least 8 different pottery vessels, and one small vessel lay completely unbroken between two of the flags. Scattered among the bones lay in each chamber over 100 beads, the majority grey and tubular like the majority of those found in Cairns I and V, though a number were yellow, red or white and of various shapes. A copper or bronze dagger was also found in this grave, but unfortunately so badly preserved that its shape was not clear.

Traces of skeletons by the southern foot of the cairn show that here too bodies had been placed outside the building. And close to the outer side of the ring-wall to the east lay a quernstone and a hammer stone.

Cairn VII.

This grave was the last to be investigated within the oval cairn, and lay only 1 meter from grave VI. It was a little mortuary building only about 3 meters in exterior diameter, and, like its neigbour, was built with a double ring-wall of uncut stone. The grave-chamber was only 1.6 meters wide and therefore provided no room for division by cross-walls. In the little chamber, which was entirely unpaved, the bones of 3-4 individuals were found in no discernible order. Among the bones lay 4 unbroken pottery vessels, and against the ring-wall stood a vessel 52 cms. high, the largest found in any of the graves. Also by this little grave discoveries were made close to the outer side of the ring wall. To the south the greater part of one or two skeletons was observed, and to the northeast, upon a quernstone lay two pottery vessels, broken by the fall of the ring-wall's stones.

Cairn Vlll.

About 4 meters northeast of grave VI lay the little grave VIII, which is only mentioned to complete the record. The few stones remaining permitted no certain conclusions concerning the construction of the grave, while only a single potsherd was found in the sand between its "walls".

*

Observations made during the investigations here described show that all the mortuary buildings are constructed on the same principle, though they may differ in details, dependent on the size of the building and the care devoted to its construction and adornment.

All the graves are circular, with double walls of dry-stone construction. Even in the case of the largest of the buildings the outer walls can hardly have been more than 2.25 meters high. Observation of the ashlar-built ring-walls seems to show that the outer walls sloped slightly inward, so that from the outside the buildings must have resembled the lower part of a beehive. Only the larger of the buildings appear to have been constructed with shaped stone.

The interior of the buildings is divided into a number of chambers by means of a varying system of cross-walls. The chambers were roofed with flat stone slabs laid upon outward leaning (corbelled) walls. The number of chambers and the placing of the walls appears dependent on the size of the building, the larger the area to be roofed the greater the number of walls required. Both ring-walls and cross-walls are double, to facilitate the roofing of the grave-chambers. Where the two halves of a double wall lean outwards away from each other the interval is filled with stones which act as ballast. In all the mortuary buildings except the very smallest traces can be seen in the ring-walls of one or two entrances to the building.

The size and free height of the grave chambers has permitted free movement under the roof­stones, though hardly in upright position. It can also be seen that the grave chambers have been used time after time, older burials being pushed to one side to make room for new. At the time of burial the body appears to have been at least partially clothed (cf. the head embroidery in grave V). Bead necklaces have accompanied the burials as ornaments, and the burial furniture has consisted of vessels of pottery or other material (stone or alabaster), as well as daggers and pins of copper or bronze, while in one case a spinning whorl was found.

It is possible that the burials in the buildings have been accompanied by sacrifices outside the tomb. This is at least a natural interpretation of the circumstance that outside three of the mortuary buildings were found human skeletons, laid together with pottery vessels and quernstones at the foot of the tomb and now lying under the fallen ring-wall.

The pottery from Umm an-Nar is wheel-made, and its material is a fine clay which in firing has taken a grey colour, or more often a terracotta red. Traces of a thin grey slip can be seen on some of the vessels. The vessels are hard fired and in most cases very thin-walled. A few of the vessels are, however, made from a sand-tempered clay. While these latter are quite large and coarse, by far the majority of the vessels are elegant small vases, most of them only 10-12 cms. in height, though some (of the type shown in fig. 24 lower right) are about 25 cms., while several specimens (of the type shown in fig. 21 top left) are only 6 cms. high.

The smaller vessels almost all possess a short neck and a more or less outsplayed rim, and they can be divided roughly into two groups. The one, found mainly in grave V, has a body formed as a more or less flattened sphere, resting on a more or less accentuated foot. The other group is vase-shaped with a conical body below more or less sloping shoulders; vessels of this type have a broad base without trace of a foot. This type was found in greatest numbers in grave I. The vessels shown in fig. 24 and the little bowl in fig. 22 top are not the only exceptions to the two groups described, others being larger vessels bearing a short spout on the shoulder.

Apart from the larger and coarser vessels almost all the pottery appears to have borne painted ornamentation. Trace of colour is, however, very unevenly preserved, and on some of the vessels all traces have disappeared. The painted decoration is in most cases on the upper part of the vessels, though it sometimes covers the whole surface. On several of the vessels decoration is also applied to the interior of the outsplayed rim. All the ornamentation consists of lines, either incised with a sliver of wood or painted, with black paint on the grey vessels and with dark brown on the red. Designs found are oblique lines, semicircles, triangles and rows of parallel lines, all regularly repeated in horizontal belts bordered above and below by lines running round the vessels. A single vase (fig. 21 upper right) is decorated with a raised wavy ridge bordered above and below by horizontal ridges. On the top of this vessel a chequer design is painted in black.

Two of the larger vessels (of the type shown in fig. 24, lower right) bear in addition to the normal geometric ornamentation painted animal figures (cf. fig. 23 bottom), in both cases the animal represented being a humped bull.

This rich ceramic material from the graves on this little desert island will be of great assistance in determining the date and cultural associations of the site, a determination which will be undertaken at a later date. But in this connection it is already possible to make reference to the pottery known from the Kulli culture in southern Baluchistan 4). The pottery of this culture, which flourished in the Third Millennium BC in the area west of the lower reaches of the Indus, shows both in form and in style of decoration so great resemblance to the pottery from Umm an-Nar that there appears to be no doubt that this pottery belongs to the same early period. The contents of the graves on this island thus form a new example of the seaward connections between the Kulli culture and the Arabian Gulf for which evidence is claimed to have been found at the head of the Gulf as early as the Early Dynastic Period about 3000 BC5).

Knud Thorvildsen