Tell El Obeid1 is a small isolated mound lying some four miles W.N.W. of Ur on the line of an old canal. Its discovery is due to Dr. H. R. Hall, who, working on behalf of the Trustees of the British Museum, partially excavated it in the spring of 1919;2 he found the building concealed by the main hillock, traced its NE. and NW. walls and a part of the other two sides, and against the SE. face hit upon a hoard of metal objects, lions and other animals in copper and fragments of a great copper relief, which amply proved the importance and antiquity of the site. Dr. Hall has published preliminary accounts of his results in the Proceedings of the Society of Antiquaries (xxxii, p. 22, 1919), in the Journal of the Central Asian Society (ix, 3, 1922), and in the Journal of Egpytian Archaeology (viii, 3 and 4, 1922). It was in consequence of the discoveries made by him that, in the autumn of 1923, the Joint Expedition of the British Museum and of the Museum of the University of Pennsylvania undertook the complete excavation of the mound. In what follows I shall deal with this last season’s work alone; the complete account, embodying the results obtained by the two expeditions, will form the first volume of the Ur publication series.
In treating of the little mound where Dr. Hall was the first to work, it is simplest to begin with a general description of the building as found. The site was occupied at different periods by three different structures, of which the earliest is today the best preserved, and is our principal subject. Set upon a little natural hillock, an “island” rising above the alluvial plain, was a solid platform whose foundations were of stone, its walls of burnt bricks laid in mud mortar for the lower part and of sundried bricks for the upper, and its core of crude brick ; it was approached by a flight of stone steps, and from the SW. side there projected a smaller platform of crude brick throughout, containing a second flight of stone steps. On the main platform stood a temple, now completely ruined.
A fortunate discovery enables us to name and date the building. About eight metres away from the façade, near the front of the stairway, there was found the foundation inscription, thrown out here when the wall in which it had been imbedded was destroyed. It is a tablet of white marble, shaped as a plano-convex brick, measuring nine centimetres by six, and it bears the following text:— “Nin-khursag: A-an-ni-pad-da, king of Ur, son of Mes-an-ni-pad-da king of Ur, has built a temple for Nin-khursag” (pl. XLV, c). The name A-an-ni-pad-da is new to us; his father is known as the first king of the First Dynasty of Ur.
The Sumerian king lists, drawn up about 2000 B.C., place immediately after the Flood the First Dynasty of Kish, then a Dynasty of Erech, and third from the Flood the First Dynasty of Ur; the names of the kings are given, and the number of years of their reigns.
Now the first two dynasties are obviously fabulous, or if they have a historic background it has been largely swamped by legend; for the shortest reign attributed to any king is one of a hundred years, and the longest are of twelve hundred each ! But when we come to the third dynasty there is no such wild chronology; the other three kings have the reasonable reigns of 30, 25, and 36 years each; Mes-an-ni-pad-da is allowed eighty years, which seems improbable for the founder of a dynasty, but the improbability disappears when we find that a son, with a name so like his father’s, reigned as king of Ur but is not mentioned in the lists; there has clearly been a confusion, and the two reigns have been lumped together and attributed to the more famous name.
At least the Tell el Obeid tablet makes the First Dynasty of Ur historical by confirming the name of its founder, and it makes it probable that from this time on the king lists are based upon contemporary written documents. The actual date of A-an-ni-pad-da must remain rather vague. There are at present no means of determining which of the various dynasties given in the Sumerian lists were really consecutive, as they are there represented, and which of them overlap, as contemporary kings disputed the hegemony of Sumer (an overlap in some cases is known to have existed) ; a simple dead-reckoning based on the king lists will therefore not give a correct result. In the Cambridge Ancient History, Professor Langdon brought the date of the First Dynasty of Ur down from 4650 B.C. to 4216 B.C. ; subsequent discoveries have made him modify this further, and in the Oxford Series of Cuneiform Texts, vol. i, he gives c. 4,000 for the start of the dynasty.
Even this shorter chronology presents difficulties in view of the close resemblance of some of the objects found, and of the epigraphy of our inscriptions to specimens of the art and texts of Ur-Nina (c. 3100 B.C.) and it may yet be proved that the date of the First Dynasty of Ur comes well within the second half of the Fourth Millennium; but at present we can only say that the foundation tablet of A-an-ni-pad-da is probably the oldest historical record yet deciphered and his temple the oldest whose authorship and relative date are known.
This building perished, violently destroyed by some enemy it would appear, for not only the shrine itself but also the upper part of its supporting platform was overthrown, so that by the time the deserted site attracted the notice of a new builder the original structure was represented by a gently sloping mound whose top was no more than three and a half metres above pavement level.
Who the new builder was we cannot say, for the large square burnt bricks which he employed bear no written stamps, only the impressed finger marks characteristic of the old plano-convex bricks, and no objects from his temple were found. We can only say that he worked on a far more ambitious scale than did his predecessor. At this time a small canal seems to have run between the temple site and the rising ground of the cemetery to the south of it. The old temple mound was made to form the core of a large brick platform which, with stepped foundations resting on the brick debris, descended in terraces to the edge of the canal, whose bank was roughly revetted with burnt brick and, along the limits of the terrace, faced with a brick water wall still standing over two metres high. The terraces were of grey mud brick, now weathered to a uniform slope so that the steps can no longer be distinguished, and thinning down to nothing at its edges so that the outline of the building, except where it is given by the water wall, could not be determined; the highest platform was virtually on the level to which the old ruins survived, but its area did not coincide with that of the original, extending well to the southeast of the First Dynasty platform. Of the temple of this second period only a scrap of wall foundation in burnt brick remained.
Very little more survived from the third period. This time the builder seems to have found his predecessor’s terraces at least in tolerable condition, for he used their upper platform as the base for his new temple, whose foundation courses rest at practically the level of those of the second period. Only the foundations of a small corner of the building remain, but these are invaluable for the history of the site, for they are made of burnt bricks stamped with the name of Shulgi [Dungi], the second king of the Third Dynasty of Ur (circ. 2250 B.C.). With Shulgi as the builder of the third temple, and A-an-ni-pad-da as builder of the first, it is tempting to assign the intermediate temple to a ruler of the Second Dynasty of Ur; according to the king lists the Second Dynasty should be nearer in date to the First than to the Third Dynasty, and this would be quite consistent with the use in the intermediate building of flat bricks (and flat bricks were already employed by the close of the First Dynasty) distinguished by finger prints, a survival from the earliest days of brickmaking.
King Shulgi was the last man to build at Tell el Obeid; when his temple fell in its turn, the site was deserted, and for four thousand years the sun’s heat has crumbled the bricks, the rain has swept the mud of them down to the gradually rising plain, and the wind has carried off their dust, until the ruin dwindled to a little mound whose top was capped by a few bricks of the Third Dynasty king, and its slope carved from the terraces of his nameless forerunner. But this same terrace, a solid casing of well laid mud brick impervious to rain and air, has preserved exactly as it fell the underlying debris of Nin-khursag’s oldest shrine. When we came to excavate the site we were obliged, in order to find the older walls, to cut down through the brickwork of the second period terrace, which, in front of the southeast wall, was over two metres thick; so hard was this that the entrenching tools ordinarily employed by our men were useless, and recourse was had to heavy railway picks. Under this grey brickwork lay as thick a layer of red crude brick, if possible harder still, representing the walls of the first temple, fallen in great masses and still bound together by its grey mud mortar; in and under it were the objects which had adorned the façade.
The plan of the early temple platform is, in intention, a rectangle with its corners orientated to the cardinal points of the compass, having on its southeast side a staircase projection and on its southwest a square platform projection also containing a staircase; the main building is (for its lower courses) of burnt brick, the two projections are of crude brick. At first sight it might appear as if the projections were accretions to the original plan, perhaps even of later date; but they are in fact strictly contemporary, or rather, they are equally essential parts of one original plan, while in the process of construction the “accretions” precede the main element of the design. The wall of the principal platform is of burnt brick, but it is not carried the whole way round the rectangle; at each projection it comes to a stop with a clean end abutting on the face of the mud brick, which must therefore have been laid first. It was probably not without reason that this was done. The main wall was the containing wall of a platform which had to be filled in solid with brick earth and mud brick; this would be easy enough at first, but as the wall rose (and it rose, as we shall see, to a considerable height) it would be impossible to bring in the filling material over the top, and ramps for the basket men would be necessary. Now each of the planned projections was, or contained, a stairway, and it was an obvious economy to build these first, so as to have ready-made a ramp which would not need to be removed after the completion of the platform. There are other features which show not only that this course was followed, but that it proved not altogether a blessing. The primitive builders, starting with the southwest projection, laid out the two long sides of their platform as nearly parallel as could be expected, and they joined these up at the northwest end with a straight wall which, if it was not quite at right angles, was not discreditably far from being such. But on the southwest side the long gap in the burnt brick work seems to have put them out in their measurements, for the section southwest of it is much too long; and on the southeast face everything went wrong. In the first place the staircase is all awry with the axis of the platform, perhaps because the builders had to set it out in relation to a wall which was still imaginary. Anyhow it was built, and it effectually prevented the bricklayers engaged on the burnt brick work from sighting through from the south to the east corner, with the result that the two sections of the southeast wall are neither at right angles to the side walls nor in line with each other: as they would, if produced, meet just in the centre of the stair ramp, it looks as if a man had got up on this to give the direction for the two gangs and, as almost invariably happens in such a case, had stood not on the line joining the two points but slightly behind it. Certainly on paper the ground plan looks oddly inefficient, but its very oddity, if the explanations given above are in any way correct, may help us to understand the simple methods of construction employed on this early temple.
A little way east of the temple, down the slope, we found two rectangular blocks of brickwork set parallel to each other, in the tops of which were shallow troughs running from the inner edge of the brickwork back to about half its width; the insides of the troughs were blackened by fire, and they contained remains of burnt wood or charcoal. The whole thing is the ordinary kitchen range that can be seen in any native cook shop of the Near East; Tell el Obeid lies sufficiently far out from Ur for a visit to the temple to have been a regular excursion, and I can only suppose that some enterprising caterer set up a restaurant in the temple grounds to supply lunch for the pious excursionists.
Nearly all the objects found in the ruin were of an architectural character, and a careful study of the positions in which they lay and of the manner in which the walls of the building had collapsed enables us to restore the façade and to assign the several classes of objects to their original places in the scheme of decoration with tolerable certainty; the full arguments for such a reconstruction of the temple must be held over for a final publication, but here it will be simplest to deal with the objects in the order which they may be presumed to have occupied.
Columns
Scattered both under and over the debris of the fallen walls were numerous fragments of timber, always circular in section, sheathed with plates of copper held together and made fast to the wood by large-headed copper rivets. The fragments differed considerably in diameter, and therefore do not seem all to have fulfilled the same function; some, the most solid, were almost certainly column shafts and can be assigned to a kind of porch which stood in front of the main door and stretched out to the stair head; others are probably roofing beams from the same porch and from the shrine itself : the dedication texts of later kings give authority enough for the assumption that the roof of a temple might have been of timber overlaid with metal. Lying together on the white floor in front of the platform were two columns of another sort: these were of wood which had been thickly covered with bitumen and encrusted all over with square and triangular tesserae of light red sandstone, black paste, and mother of pearl; each tessera had at the back a loop of copper wire which was driven into the bitumen and so made the piece fast in its place. These columns were 2.30 in. long and 0.90 in. in circumference; they came from the main door of the shrine. Fragments of columns similar but with smaller tesserae, found both by us and by Dr. Hall, appear to have belonged to a second door in the northeast side of the shrine.
Copper Bulls
Remains were found of four copper statues of bulls, made in the round; two of these were in such a condition that they could not be removed (indeed, one could only with difficulty be recognised), and two were brought away, of which one was headless.
The animals, which stood 0.60 m. high and were 0.70 m. long, are represented as walking along slowly with the head turned sharply outwards over the left shoulder; they clearly were meant to be seen from the left side only, and it is probable that they stood in a row along a low step or ledge between the top of the platform and the wall of the shrine.
The bull was carved first in wood, the body, legs, and head in separate pieces which were morticed together and secured by copper bolts; then the legs and head, and last the body, were covered with thin plates of copper whose edges overlapped and were held down by copper nails; the tail, horns, and ears were attached afterwards.
Artificial Flowers
The stem and calyx of the flower is of baked clay, the petals and corolla of white limestone, red sandstone, and black paste: the corolla was always either red or white; of the eight petals four were white, two red, and two black, these being arranged crosswise. A large lump of bitumen was pressed round the tall corolla, sloping down to the scalloped edge of the calyx, and the petals were set in this also sloping downwards and outwards so as to make the blossom sharply convex. Each petal had behind it a loop of copper wire, the ends of which passed through holes in the calyx and were twisted together underneath.
The total length of a flower varies from 0.18 m. to 0.37 m. Generally, though not always, there are near the pointed base of the stem two small budlike projections; low down in the stem there is a small hole pierced right through, as if for a string, and high up near the calyx bears a cut made horizontally in the wet clay.
When Dr. Hall found specimens of such flowers, he suggested that they were rosettes for wall decoration, the long stem being inserted in the crude brickwork and the circular top resting flush against the wall face. Now that we have a number of specimens to judge from (over fifty entire examples were found), this view, which had seemed to be justified by the precedent of the small slender cones which Loftus found at Warka driven into the mud-brick wall so that their round tops formed a pattern on the surface, proves to be untenable. The size and length of the stems, and their tapering shape, are against the theory ; the flowers were always found loose, never embedded in the wall, though there were plenty of great masses of brickwork fallen intact wherein the flowers ought to have retained their places if they had been fixed there; the fact that the stems were almost always broken, which would not have been the case if they had had the protection of the brick mass, shows that they were free; and the hole through the stem and the nick in it could not be explained if that stem had merely served as a peg. The flowers must have stood upright in the open, the pointed ends of the stems resting in shallow sockets such as the finger print holes in the upper surface of the plano-convex bricks, a thread or wire passed through the hole low in the stem, thus stringing the flowers in line, and another thread or wire running from flower to flower and twisted once round each stalk just below the calyx, where the nick in the clay prevented it from slipping, kept the row upright; the two strings would be stretched taut and made fast at the ends to posts or attachments in the wall; there might be just enough play to allow of the flowers swaying in the wind! Details of the positions wherein the flowers most often occurred showed that they were closely connected with the standing figures of bulls described above, probably occupying a slightly lower shelf than they; the bulls would thus seem to be walking in a meadow full of daisies.
Frieze of Copper Bulls
Higher up on the facade of the shrine there ran a continuous frieze, of which the greater part was found fallen down below; it consisted of a series of reliefs in copper representing young bulls. Each is lying down with three of its legs doubled up under it in repose, but the far front leg is raised with the hoof firmly planted on the ground as if the beast were just in the act to rise; the bodies are shown in full length profile, but the heads are turned towards the spectator; whereas the bodies are in low relief, the heads are modelled in the round, and project boldly from the general plane, giving an extraordinary vitality to the figures.
Technically these figures are most interesting. The body of the bull was carved on the surface of a stout board, and to this a thin wash of bitumen was applied. The head of the animal was cast hollow in copper, the hollow was filled up with bitumen, and a wooden peg was let into this and made fast with a copper bolt put through the back of the head, and the end of the peg let into the neck of the wooden relief. Then a thin plate of copper was laid over the body and hammered down on to the wood so as to reproduce all the carved detail, the neck being brought up so as to overlap the casting, and the edges of the sheet bent over the edges of the board and nailed down to it.
The total height of the frieze was 0.22 m., the average length of an animal 0.60 m. To attach the frieze to the wall, copper hold-fasts were inserted between each pair of animals. These consisted of bars of metal, rectangular in section, the ends of which went through two holes set vertically in the plain field of the frieze and were carried through the wooden background; then they were twisted to form two circular rings, the first vertical, the second horizontal, which were laid between the crude bricks of the wall and secured by wooden pegs passed through them.
The frieze, which ran along the whole of the southeast façade of the shrine, is represented by twelve more or less complete figures found in the ruins.
Inlay Frieze
Above the frieze of copper heifers ran a second, of the same size but of very different character. The frieze was framed above and below by a narrow border of copper nailed over wooden battens; the background was a wooden board (now wholly perished) which was secured to the wall by copper holdfasts exactly like those of the lower frieze except that in this case the holdfasts did not come to the face of the panel but were fastened to the board only. Over the wood was laid a thick layer of bitumen, and on this the design was worked out in mosaic. The figures were cut in white limestone or in shell; if in the latter they were always made up of a number of small pieces carved separately, if in stone they were sometimes composite, sometimes in a single piece; the background was made up of tesserae in black paste of varying sizes, cut to fit into their places: as usual with Sumerian inlay, each piece was secured by a loop of copper wire fixed into it from behind and forced into the bitumen backing.
Large fragments of this frieze were found and removed intact. The most interesting gives us a genre scene of pastoral life. At one end are two men milking cows; the men are squatting awkwardly under the cows’ tails (the same position for milking is adopted today by some of the Lower Mesopotamian tribes) and hold long slender milking vessels; in front of the cows stand the calves, duly muzzled so that they cannot get milk for themselves. In the centre is a byre. It is built of big reeds (?) bound with ropes; it has a kind of entrance tower with a window above the door; the door itself is flanked with spears and adorned with the peculiar side loops familiar to us from later pictures of Sumerian buildings; above is a sort of crescent which may be derived from the sacred horns. From the door of the byre issue two heifers. On the other side of this building are four men, dressed in the usual sheepskin skirt, engaged in straining and storing a liquid which we may guess to be the clarified butter resulting from the milking operations conducted at the opposite end of the scene. The man on the extreme left has plunged his hand into a great jar, presumably to draw out the liquor from it; the next man is pouring the liquid from a small jug into a a strainer held by his fellow, from which it runs down into a big spouted jar set on the floor. The fourth man has between his knees a great store jar destined to receive the strained liquor. From every point of view the panel is of the utmost importance—for the light it throws on the domestic life of the Sumerians at this early period, as an illustration of the art of the time, and for the possibility it gives of dating other objects; its value in this last respect will be seen when we come to deal with the tombs of Tell el Obeid.
In the milking panel the figures are cut in limestone. The stone is not of very good quality, and the fact that it was selected for a scene so important might be taken as evidence for the use of colour—the stone being smoothed over with plaster and painted. Certainly in finished workmanship it does not compare with the panels wherein the figures are carved in shell. The complete examples of shell inlay that were found represent a procession of bulls, each animal made up of six or seven pieces, all carved with the utmost delicacy of relief; the general type is the same throughout the whole frieze, but in the drawing of each animal there are slight differences which relieve what might otherwise have been monotonous. It is probably due to the mere accident of preservation rather than to any sameness in design that the bull figures seem to form so large a proportion of the frieze; connected with one panel, though no longer actually attached to it, was found a small relief plaque of a human headed bull on whose back is a lionheaded bird, a subject obviously mythological. In this case the plaque is formed of a single square of limestone, and the design is cut in true relief ; but traces of black colour applied to the background prove that the effect was identical with those parts of the frieze where the figures were in silhouette inlaid against a black field. Besides this, isolated fragments representing human figures, the goat, the ibex, etc., both in shell and in limestone, are evidence of variety in subject matter. Perhaps belonging to the same frieze, perhaps to another set higher up in the façade, are a number of birds, silhouetted in limestone; only one of these was found with its background and copper holdfast more or less complete (the dimensions of the panel are the same as those of the bull sections), but nine or ten birds are represented by whole figures or by fragments, so that there must have been a considerable length of frieze of this sort. There can be little doubt that they were coloured, for the modelling is of the most summary description, and the surface of the stone poor and rough; the evidence of the exact positions in which they were found tends to show that they formed a separate frieze.
Limestone Wellhead
This was found in fragments near the foot of the main flight of stairs. In contrast to the objects hitherto described, it is curiously primitive in style and execution, and might well have belonged to an entirely different period, but must certainly be assigned to the same date as the temple itself; indeed, it is tempting to connect it with a fragment of a stone vase dedicated in the temple whereon is an inscription recording that one Ur-Nannar made here a well for the service of the goddess and for the life of King A-an-ni-pad-da. The well head is decorated with figure in two zones. In the upper register there are small figures bearing palm leaves, etc., about what seems to be a statue of a lion on a raised base, the scene being twice repeated. In the lower register, of which only the upper part is preserved, the figures are on a much larger scale; a god, facing right, receives the worship of four mortals, two large and two small. The men are beardless, and wear a skirt and a cloak folded over the left shoulder; the faces are grotesque, and the drawing of the bodies is clumsy in the extreme
Inscriptions
The foundation tablet has already been mentioned, and so has the vase fragment with the well inscription. The only other object inscribed, apart from two or three very small pieces of clay tablets whereon no more than a character or so was preserved, was a gold bead of scaraboid form on the rounded top of which was the name of A-an-ni-pad-da, king of Ur. The bead was found in the debris of the fallen wall, and most probably had formed part of the foundation deposit; it is difficult to see what else it can have been doing in the building. It is, I suppose, the oldest piece of royal jewellery known.
The Cemetery
The Cemetery lay on the second low mound, another “island” site, just south of the temple. A deep trench cut along its highest ridge showed that the greater part of the rise was natural, formed of river silt, and that at a very early period, long antecedent to the building of the first temple, it had been occupied by a settlement. There were remains of huts built with daub and wattle walls, with stone hinge sockets for the doors and floors of trodden earth; the objects found in these were all of a primitive type, rough stone querns and rubbing stones, painted handmade pottery, incised wares, and plain rough cooking pots, together with flakes of flint and obsidian, clay sickles, etc. Graves of the same period had occupied other parts of the mound.
Later, the settlements were abandoned, but the use of the mound for burial purposes continued, probably after a considerable interval, for the older graves, if they had not been forgotten, at least no longer commanded any respect, seeing that they had been ruthlessly destroyed to make way for the later interments. And in these the objects found were quite different from the contents of the first graves; the pottery was wheel made, painted wares were wholly lacking, and together with flint implements there were tools and weapons in copper. Only one grave of the early type was found by us intact, so thickly were the later burials set, side by side and one above the other; often it was impossible to say to which particular interment the tomb furniture belonged, so confused were bones and objects alike; anything like a sequence was therefore difficult to obtain. But it was clear that these later interments represented a long period, and at least the limits of this could be fixed. It starts after the close of the painted pottery time, whenever that was, and it ends before the beginning of the Third Dynasty of Ur, i. e. before 2300 B.C., for none of the objects found could by any possibility be attributed to that date, the archaeology of which is by now becoming fairly familiar to us. Within these limits, we find varying forms of interment. There are plain interments with the body in the contracted “embryonic” position, others with the body extended at full length, burials in narrow trenches lined and covered with bricks (and the bricks are those of the second building period of the temple site), and burials in circular or oval clay “baths.” A few of the graves with contracted burials, which, for reasons of depth, position, etc., must fall early in the period covered by the cemetery, contained pottery vessels identical with those represented on the inlay panel of the A-an-ni-pad-da temple; the depth of the bath burials and their position on the outskirts of the cemetery make it almost certain that they fall late in the period and link up with the bath and pot burials familiar to us at Ur, where they continue in use right down to Persian times. In a preliminary report such as this, it is impossible to attempt anything like a chronological sequence of the tombs, which could only be established by a detailed examination of them all; indeed, the material at our disposal may well prove insufficient to establish such at all except in the roughest outlines; here it is enough to say that from the hundred graves dug we have obtained a great mass of objects belonging to a period more or less defined which, archaeologically speaking, was hitherto altogether unknown to us.
The pottery is all wheel-made and unpainted; the forms are very varied, decoration is confined to occasional rope-mouldings in relief, incised ornament (rare), and sometimes the employment of reserved slip ornament; whereby a slip is applied to the surface of the pot and then partly wiped off so as to expose the body clay. Stone vessels are nearly always of bowl form; decoration does not extend beyond a simple notching or line engraving of the rim; the materials are limestone, aragonite, and greenish grey stone. Copper vessels are generally of bowl form, though one large cooking-pot of curiously modern shape was found. Copper tools include axes, celts, knives or daggers, pins (one has a head of lapis lazuli set in gold). Stone implements are most often coups de poing with a rounded and flattened head, almost spoon-shaped; shorter pear-shaped and nearly circular coups de poing, knife edged and saw edged flakes of flint and obsidian; rough rubbers; a few miniature polished celts. Occasionally the copper and stone implements were imitated in clay; these imitations seem to belong for the most part to the destroyed graves of the earlier period, as several of them are painted. Beads are of lapis lazuli and carnelian, and as a rule very few were worn, a set of not more than a dozen beads on a long string being more common than an entire necklace. In some of the graves there were shells used as palettes and containing either soft red haematite paste, presumably for rouge, or green malachite paste, which was probably employed, as by the Egyptians of the pre-dynastic and early dynastic times, as an eye paint. A curious object in copper was perhaps the head of a ceremonial staff.
As long then as the Nin-Khursag temple existed, the neighbouring mound was used as a graveyard; even after its destruction though the old cemetery fell into disuse, the tradition seems to have continued, for graves of the Kassite period were found in another mound about a mile to the west and a few hundred yards to the east there was a ruined cemetery of a later date. It would certainly appear that the worship of the goddess was in some way associated with the idea of burial. Nin-Khursag is known as a goddess who took part in the work of creation. In the decoration of her temple a very prominent place is given to representations of cattle, and we have the domestic scene of the milking of cows. The bull is a regular Sumerian symbol for divinity, but this would not explain the cow element ; but the cow is elsewhere known as a symbol of fertility, of the preservation of life, and for a primitive pastoral people with whom cow’s milk is a staple diet, this is a very natural conception. In the graves, the presence of tomb furniture shows a belief in a future world, and the embryonic attitude of the dead in the earlier graves connects death with new born life; may not then the creator-goddess be thought of as safeguarding the continuity of the life that she has given and bringing to fresh birth those whom life has outworn?
1 Mr. Woolley’s preliminary Report on the work of the Joint Expedition at this small site in 1923-24 was published by arrangement in the Antiquaries Journal, from which the abstract in these pages is taken and to which the reader is referred for the complete report with plans.—EDITOR.↪
2 In the article by Dr. Leon Legrain in the September JOURNAL, page 151, there occurs a statement that the temple of Ninharsag at Tell El Obeid was discovered by the Joint Expedition. This is a mistake. As Mr. Woolley states, the building was discovered and partly excavated in 1919 by Dr. Hall while working for the British Museum alone. The remarkable lions’ heads and other objects in bronze now in the British Museum were recovered by Dr. Hall at that time.—EDITOR.↪