• T.G. Bibby


Bahrain, Glob, tyrebrønden, well of bulls


The Well of the Bulls

In the northwest corner of Bahrain lies the considerable village of Diraz. From the flat roofs of its crowded white houses the waters of the Persian Gulf can be seen over the palmtrees about half a mile away. In clear weather the coast of Arabia can be glimpsed across the narrow strait which separates Bahrain from the mainland, where at night the waste-oil flares from Dhahran light up the western sky.

South and east of the town lies a broad hillocky stretch of country, in summer baked dry and parched by the sun, but after the winter's rain green and luxurient, covered by the swiftly growing desert plants. All this area is covered by millions of potsherds, visible proof that here has once been a big city. At some points on the higher hillocks the winter vegetation forms rectangular patterns in the sand, showing that underground must lie remains of walls and foundations, preventing growth where they lie.

Less than half a mile east of the present village, just on the edge of the palmgroves, there is a large and remarkable hollow, surrounded by high sandbanks, which give the impression of having at some time been dug up from the hollow and piled around it. The hollow is oval in shape and measures 70 x 40 meters. The bottom of the hollow lies about 3 meters below, and the top of the sand­banks about 2 meters above the surrounding desert. The banks are bare, but the bottom of the hollow is clothed with sparse grass and thorn bushes.

Our attention was first directed to this site by the circumstance that, around the southern and western sides of the hollow and more scattered along the north side, there lay on the top of the banks about a hundred very large stone blocks, some over a meter long and all of them carefully shaped to a rectangular form. Some showed in addition revets and "door-post" holes, showing clearly that they had once formed part of a building, or at least had been cut for use as building stone.

Squared building stone is very rare in Bahrain. Only two buildings still standing in Bahrain are of squared stone, the "Sud-el-Khamis" mosque lying some miles south of the capital and dated to the 15th century or earlier, and the Portuguese fort on the northern coast, dated to the 16th century, in the construction of which a large quantity of stone from previous buildings was used. During the last few centuries buildings have been constructed entirely of coral and gypsum cement. Squared stones are thus an indication of respectable age.

The hollow is called by the Arabs "Ain Umm-es Sujur", and they are in no doubt that it once was a spring. The local legend concerning the site is well known among the Arabs and in particular detail among the Dirazis. It is mentioned for the first time in Western writings by Captain Durand who made a survey of the antiquities of the island in 1878 1). As it is not without relevance to an interpretation of the results of the excavations its substance is given briefly here.

Ain Umm-es Sujur was at the commencement of the Islamic era (in the 7th century) the largest of the three most important springs on Bahrain, and from it water was led to the most distant parts of the island, to the south and to the east. By the spring lay at that time the capital city of the island. But the caliph, Abd-el Malik ibn Marwan 1685-705 A. D.), came to Bahrain with an army. After a long and indecisive war he succeeded in assassinating the leaders of the Bahrainis and won mastery over the island. Before he left he punished the inhabitants by destroying the spring of Umm-es Sujur, which has never since been reopened.

The various versions of the legend disagree on the reason for the caliph's invasion. One version relates that the inhabitants of Diraz, after Bahrain's conversion to Islam, returned to the worship of their former idols and gave shelter to non-Islamic fugitives from lraq; another version tells that the caliph desired the daughter of the king of Bahrain.

Our investigations at the spring centred first on the stone blocks. When they were dug free it was found that they all rested on virgin sand, without the slightest sign of foundations or building in association with them. Nor did they lie in any sort of logical association with each other, apart from a group by the western end of the hollow which lay roughly in rows with an interval between each block.

A sondage from the top of the bank immediately besides this group descended through clean sand until it reached, at a depth of 3.27 ms., a layer of loose stones which clearly marked the original surface, after which it descended for a further 1.66 ms. through undisturbed waterlaid sand to the level of the watertable, the water welling up pure and warm.

A trial trench was then driven down the slope of the hollow in the south­western corner. The upper part of this trench revealed nothing but unstratified sand; but at the foot of the slope, where the bottom of the hollow commenced, four stone blocks of the same type as those on the top of the slope lay in a tumbled heap about 30 cms. under the surface. They too rested on plain sand, and immediately under their level the watertable was reached, preventing deeper excavation.

The investigations of the stone blocks proved therefore fruitless. It is possible that the solution to the riddle lies below the water surface. It is not impossible that the blocks originally formed a wall around the spring, a wall the foundations of which lie now below the water. This wall was perhaps pushed down into the well by the army of Abd-el Malik or by some other enemy.

The springs of Bahrain occur where a breach in the underlying limestone strata allows the artesian water to make its way up towards the surface. There is always a danger of such springs being choked by sand, and they are therefore without exception surrounded by walls, sometimes enclosing a considerable area, which hold back the sand, often lying several yards higher than the level of the water.

On first hearing the story of Abd-el Malik's destruction of the Sujur spring one is apt to ask how it is possible to destroy a large spring; but a knowledge of the walled springs of the desert provides the answer. If the immense holding walls of such a spring be broken down the sand will close in swiftly from all sides and the spring disappear as though it had never been. Such may have happened at Sujur.

If the wall around the spring was in fact formed of these meter-long stone blocks, then the wall was a finer piece of work than is found at any of the present-day springs in Bahrain. But only a continued investigation will be able to give a final answer to the question of the original appearance of Umm-es Sujur.

The excavation here did, however, give one positive result. In the slope of the southeast corner of the hollow the ragged corners of walls of stone set in plaster could be seen. On investigation they proved to be two parallel walls with a plastered surface on both sides, about 80 cms. wide and distant from each other about 1 meter. They ran for 4 meters due east and then made a rightangle turn southwards (Fig. 1) and continued parallel at the same distance apart for a distance of 5.10 ms., before debouching into a little room, with interior dimensions 1.48 X 1.4 ms. Between the walls lay a sloping ramp which in the southgoing passage developed into a staircase going down to the chamber (Figs. 2-3). On the staircase lay two large fallen roof-slabs, and at the mouth of the chamber one of the roofstones was still in position (Fig. 2 A). In the eastern wall of the chamber was a little niche, measuring 33 x 36 cms. and 40 cms. deep (Fig. 2 B). At the same level as the niche was found the first sign that the structure had a religious significance. The chamber was here full of fallen building-stones, and among them, by the western wall, lay a limestone statue, a kneeling bull, the head of which had been broken off (Fig. 4). It was 31 cms. long and 21 cms. high. Among the stones was also found a little oblong stone block, hollowed at the top. It resembles the small incense braziers of wood covered with iron sheeting in which the Arabs today burn sandelwood in honour of their guests. It is probable that this object was also used as an incense burner and perhaps stood in the niche.

The floor of the chamber proved to consist of a single large squared stone, measuring 1.28 X 1.02 ms. In its centre was a circular hole, 72 cms. in diameter, around which ran a flange 12 cms. wide and 2 cms. high (Fig. 2 C and Fig. 3). The stone proved to be a wellhead, and 53 cms. deep in the central hole lay the surface of a gushing spring of sweet water. The wellhead stone was 33 cms. thick and below it the wellshaft was continued by another 79 cms. deep (Fig. 2 D) which rested upon sand.

Between the edges of the wellhead stone and the sides of the chamber was a gap of some cms., and from the sand filling this gap came a number of pot­sherds and half of an alabaster bowl. And on the fourth step up from the chamber stood another limestone bull identical with the one first found, and similarly decapitated.

So much of this building complex is preserved that it is possible to reconstruct its original appearance. The first "leg" of the passage must, on account of the lowness of its sides, have been open to the sky. The two bulls presumably flanked the entrance to the roofed southgoing "leg", as is usual with animal statues in the ancient Middle East. The inner well-chamber probably had a domed roof and was lighted perhaps by nothing more than the little incense lamp.

This tiny well-chamber cannot have been built as part of an ordinary system of water-supply. There was water enough and to spare in the large spring immediately outside the entrance. The bull statues, the incense burner and the niche all suggest that here was a holy well, a little water-temple, associated with the religion current on the island before the coming of Islam. The destruction of the well-temple and the decapitation of the bulls complete the argument, and throw a new light on the legend connected with the spring: "The Dirazis began again to worship their ancient idols ... and Abd-el Malik in retribution destroyed their well ... "

Archeologically, however, the building cannot be dated with certainty. The buIIs could well belong to any Pre-islamic period. The potsherds are of a type which appears to have had a considerable extension; the same material is, for example, found in the pottery from the graves connected with the Sin-temple at Hureidha in the Hadhramaut, which are dated to the 4-5th centuries B. C.2).

T. G. Bibby.





Bibby, T. (1954). Tyrebrønden. Kuml, 4(4), 154–163. Hentet fra https://tidsskrift.dk/kuml/article/view/97164