Volume II

Among the 1,850 graves unearthed in the cemetery there are sixteen which stand out from all the rest not simply because of their richness — indeed, most of them have been plundered and their wealth must be taken on credit, and few have produced such treasures as marked the grave of Mes-kalam-dug, itself not one of the sixteen — but for peculiarities of structure and ritual. All these graves belong to the Early Cemetery, and of most of them at least it can safely be affirmed that they belong to the earlier part of the period which that cemetery represents. In the majority of cases they underlie and are necessarily earlier than graves whose contents assign them to Cemetery A; where this is not the case the contents of the graves themselves are characteristically of the Cemetery A type; it is not a question of positive depth reckoned from the modern surface, though as a matter of fact most of them do lie at much more than the average depth, it is a relation which admits of no dispute. To the connexion between the sixteen graves and the rest I shall return later ; it is necessary first to explain in what way they form a class apart.

The whole of the upper part of the queen's body was covered with beads of gold, silver, lapis-lazuli, carnelian, and agate, U. 10975-8, U. 10980-1 ; they were astonishingly numerous and of exceptionally fine quality, as can be seen from the selection illustrated in colour on PI. 131. The actual strings on which they had been threaded had of course disintegrated and with the collapse of the body the order of the beads had been disturbed, but they were not altogether in disorder and in most cases the individual chains could be followed quite well. Across the body at about the waist ran a broad belt of beads, tubes of gold, carnelian, and lapis, U. 10879, in ten rows of alternating colours, from which had been suspended rings of thick gold wire, U. 10867; they had certainly been sewn on to a background of cloth or leather.

Against the right upper arm of the queen were found three gold pins, U. 10870, 10940-1, which probably secured the cloak, made open on this side, as was the usual fashion. With the pins were four amulets, two fish of gold and one of lapis (U. 10944-5, PL 142) and one, of gold, in the form of two seated gazelles (U. 10943, PI. 142); with them were three lapis-lazuli cylinder seals, U. 10871, 10872, 10939 (Pl. 193), of which the last bears the name of Shub-ad. The cylinders were perhaps tied to the pins to make toggles for securing the cloak. By the right shoulder was a lapis-lazuli amulet in the form of a reclining calf (U. 10946, PI. 143) strung on a row of three large beads of lapis and agate; a pair to it in the form of a bearded bull (U. 10985, Pl. 143) was by the left shoulder; probably they hung from the hair.

On the fingers of the queen's hands were ten gold rings, U. 10877, 10878, 10949, 10950, eight of them of plain gold with cable-pattern decoration, two inlaid with lapis-lazuli, PI. 138. On the top of the thigh-bones were found a number of large faceted date-shaped beads of gold, lapis, and carnelian, U. 10880, which may have been connected with the belt but were of a different character from the other beads composing it. Round the right knee was a garter made of flat rectangular beads of gold and lapis with one carnelian ball, U. 10979. To the left of the waist but clear of the body, resting on the woodwork of the bier, was a gold pin of Type 5, U. 10938, PI. 165. This completed the personal ornaments worn by the queen.

PG/1237. (Pls. 3, 8, 69-77) 'The Great Death-pit'. The name 'The Great Death-pit' was given to PG/1237 for the good reason that it consisted of a death-pit and little else. It might at first sight seem incorrect to class as a royal tomb what was not really a tomb at all, and for a time I was even inclined to think that the character of PG/1236 was in itself an argument against any of the 'royal tombs' being rightly identified as such, for whereas here we had an exact equivalent of the death-pits associated with those other tombs this pit contained no tomb-chamber in which ex hypothesi the king or queen should have been buried ; if a death-pit was not necessarily associated with a burial, did its association with a burial imply that that was a royal one ?

Immediately above the grave there were found loose in the soil the group of stone vases U.8936-U. 8951, and U. 8978 (cf. PI. 13). It is very probable — although of course it cannot be proved — that they came from the tomb chamber of PG/337 and were dropped by the robbers (the soil was too disturbed for any evidence from stratification to be available) or even that they were offerings placed in the filling of the shaft at the time of the interment ; certainly it is the only grave in the neighbourhood with which objects of such good quality were likely to have been associated, and seldom do we find anywhere except in royal graves such fine examples of stone vessels. Moreover, they were not alone, for close by, not indeed immediately against them, but at about the same level and directly above the plundered grave, there was a collection of small objects of jewellery which again was of the sort generally found only in royal tombs. The most conspicuous of them was a gold amulet in the form of a reclining bull, its chin ornamented with an elaborately curled beard; though only 0.015 m. long it is a beautiful piece of goldsmith's work and compares favourably with the amulets from Queen Shub-ad's tomb (v. PI. 142, U. 8269, and fig. 2). The bull had been worn on a chain, for a single gold bead was attached to its body where a hole passed through it for suspension ; the string had been broken and the rest of the beads scattered ; a number of such, of different sizes, in gold, lapis-lazuli, and carnelian, lay close by (U. 8271, 8274), and a single leaf pendant from a wreath (U. 8270), a gold toilet-set as well as a gold finger-ring and a limestone bowl (U. 8273). These were certainly not objects deliberately placed in the filling of the shaft but stray pieces dropped by robbers and lost in the disturbed soil. I have little doubt that the tomb so robbed was PG/337.

Lying across the body was a silver lamp of the usual sort, Type 115, but of unusual size, its width being no less than 0.32 m.; it was terribly decayed and the spout end had perished altogether. Under it, on the edge of the bier near the queen's right hip, were four brim head-dresses (U. 10873-6) placed together. On the corner of the bier by the right side of the head was a golden bowl, U. 1085 1, of Type 7 and by it a silver drinking-tube.

The nobler metals provide a fruitful source of evidence for technical methods of manufacture. This, unfortunately, is often denied to us in the case of copper and its alloys, which are ravaged by corrosion to such an extent as to hide all trace of tool marks unless in exceptional cases.

Sheet metal-working in copper, silver, and gold was frequently resorted to for the strengthening and decoration of wooden objects. Columns were lined with copper and the structural members of harps with gold and silver, the latter metals being invariably employed with taste and discrimination. As silver and above all gold lends itself to the fabrication of very thin sheets these metals were employed in this form to enrich the heads of copper scarf-pins and to decorate beads. Beads of bituminous shale, wood, &c., have been found covered with gold which is so thin that when removed and held by one end it hangs down, possessing insufficient rigidity to remain horizontal. Could gold have been beaten so fine without the aid of goldbeater's skin ? Whether we are justified in crediting the Ur craftsmen with knowledge of a technique which even to-day inspires the admiration of the layman is open to question, but it may be mentioned incidentally that the properties of gut were well known and utilized in the manufacture of strings for the harp and this substance might just as well have found practical use in the workshop for gold-beating. Such a supposition is entirely unsupported by direct evidence, however, and the balance of expert opinion favours the more probable explanation that the gold was beaten on flat hard stones. 'In view of the great malleability of gold small rents that might appear in the initial stages of the lamination would be welded up later on' (Mr. E. A. Smith, private communication). Mr. Ogden, indeed, has identified what is almost certainly a small goldsmith's hammer or burnisher which might have been used for preparing such thin leaf. In appearance it is like a tiny pestle, slightly bulbous at one end and measuring only 19 mm. in length. The material is haematite. The bulbous end shows unmistakable signs of having been used as a rubber.

Gold sheet was moulded into shape as a rule by rubbing the metal over a model of the object to be made, constructed from wood or bitumen, and the results are seen in the various gold heads from the harps and in the goat figures U. 12357 A and B.

The skill of the Sumerian goldsmith in hammering up metal is well shown by the silver jug, U. 9334, PI. 171, which has been beaten up in one piece, and by various silver dagger handles only the remains of which have been preserved. The best example in gold, or rather electrum, is the helmet, U. 10000, PI. 1, which also has been beaten from sheet metal.

The joining of metals was done either by sweating or by soldering the seams, depending on the type of work. (See E. A. Smith, 'Solders used by Goldsmiths of Ur', Discovery, Jan. 1930.) Such processes were common; the bowls of Type 7, for instance, were made from a sheet of gold or silver cut to shape and with a triangular piece cut away from either end; the hammering up of the vessel brought the edges of the segmental cuts together and they were sweated and strengthened with a rib disguising the join; similar types in copper may have been raised entirely from the sheet by beating on a bitumen core ; they were finished with end ribs which gave the objects shape and strength. A good cherry-red heat was necessary both for sweating (which required incipient fusion of the metal) and for soldering proper. The flame from an oil lamp would be directed on the spot by means of a mouth blowpipe, probably a reed covered with clay to prevent it from burning. The highly decorated beads like U. 9779, PI. 138, are generally sweated, i.e. the filigree or granulations are attached to the background by carefully controlled heat without the extraneous medium of solder ; but the flat disk beads made from two thicknesses of metal joined together so as to leave a hole for the thread through the centre were apparently soldered, Type 16, Fig. 70. Apparently the gold beads of tubular form, either barrel-shaped or in the form of two cones, base to base, all have a longitudinal soldered seam; the solder used would be either a thin strip of the same gold (autogenous soldering) or of an alloy richer in silver and consequently of lower melting-point than the gold.

Soldering was also employed for the finger-rings made of coiled gold wire. This 'wire' was not drawn but was evidently cut from thin plates of metal (Exodus xxxix. 3) and was square in section. A length sufficient for the purpose required was taken and the middle part of it was twisted on itself, leaving at either end enough plain untwisted wire to form a hoop ; the whole was then wound in a coil with the strands touching, and gold solder was applied to the inside of the ring to hold the strands together. Further twisting, winding, and soldering might be done, so that eventually a wide ring resulted, having smooth borders and cable-pattern centre (such as U. 10943, PI. 138), and the method of manufacture was difficult to detect. One example which had become unsoldered (U. 10599) served to explain the technique. A curious instance of this love of securing an effect by what might be called illegitimate means is given by the head-ornament U. 8693 ; this is a brîm of ordinary type apparently, consisting of two lengths of gold chain and three large beads of gold and lapis divided by carnelian rings ; but whereas the chains seem to have interlocking links giving a square section of traditional form (loop-in-loop), they are not really chains at all, nor even flexible, but lengths of cut wire twisted on themselves and soldered together into an imitation which without careful examination would pass muster as the genuine thing.

The finer decoration of a metal object seems generally to have been done by chasing, seldom by engraving. The most elaborate example of the former is the diadem U. 8173, PI. 139, where all the outlines of the figures are impressed by hammering the flat edge of the chasing tool or die on the gold ; the decoration of the silver vessel U. 9364, PI. 170, is similarly done, and so are the minutely fine lines of the hair on the gold helmet. One of the very few instances of real engraving is the gold bowl U. 10003, PI. 160, from the grave of Mes-kalam-dug, where the fluting has been touched up by the engraver whereas the pattern round rim and foot is chased. Ornamental repousse work in gold is best illustrated by the helmet and in silver by the bowl U. 11794, PI. 167; strictly speaking, it is rare in the cemetery, for the small gold figures of animals (e.g. PI. 141 b) and the details of the goat statuettes are not hammered up from behind but are made by pressing and hammering the gold on to a wooden core already carved to the required form.

Electrum was employed in decoration not merely as a substitute for true gold ; it had its value in adding to the range of colours at the disposal of the goldsmith. Thus in the case of a silver bowl, U. 10891 , PI. 173 , the bowl itself is of silver, there are very slight vertical grooves in its sides, and in them strips of electrum have been fixed either by sweating or by lapping the silver over the edge of the inlaid metal. The tubular handle-rings of such bowls are not infrequently of electrum (e.g. U. 11871, PI. 173). A decorative use of the alloy is also made in the case of the brooch, U. 8007, PI. 219, where the filigree background of silver wire contrasts strongly with the smooth electrum bosses in the centre. In scarcely any object found in the old cemetery are gold and silver brought into direct juxtaposition ; they are used together, but there is always some other material to separate them. This particular contrast seems not to have been to Sumerian taste, the employment of the paler metal being apparently the result of deliberate preference.

The Sumerians are the first people we know to have shown skill in filigree and granulated work. A fine example of minute craftsmanship in the latter technique is afforded by a gold ring measuring only 2 mm. in diameter and composed of six gold balls sweated together. The filigree was occasionally enlivened by inlay à jour, as in the case of Shub-ad's pendant (PI. 128), or the roundels U. 11806, PI. 138. Only one piece of this inlay is preserved in the latter case, but it must have been general. True cloisonné work was also practised, the wires for the cloisons being soldered or sweated on to the background and the stone inlaid, e.g. the finger-ring U. 9778, PI. 138, and the two roundels shown on PI. 133.

The more or less uniform size of certain objects such as ear-rings raises the question whether they could not have been used as a commercial medium with a recognized value (cf. Genesis xxiv. 22) ; in PG/789 with body 10 there were found all together eleven short pieces of gold cut from a hair ribbon, each folded into a tight wad, and their combined weights came to exactly one shekel (v. Catalogue, U. 10801); the complete ribbons might have served in the same way, or a definite length of ribbon might have had its known value. Two similar discoveries lend probability to the idea that the little gold wads were a recognized means of exchange; in PG/1413 a set (U. 12712) of seven haematite weights, two pebble weights, and some of copper (decayed) were found, and these had clearly been together in a bag, while apart from these were also some fragments of lead which may have been weights ; in another grave, PG/1609, were a number of calcite pebbles not carefully shaped but almost certainly weights. It seems likely, therefore, that the gold wads fall in the same category ; if used as currency the varying quality of the alloy is admittedly difficult to understand.

All the gold is apparently alluvial, and a natural admixture of a certain proportion of silver is to be expected, but the range of proportions is astonishing — from almost pure gold to an alloy in which gold forms less than 30 per cent. As it was not possible for the Sumerian goldsmiths to remove silver from gold when they were once alloyed together it is perhaps natural that a fair quantity of metal of the electrum type should be found, but what is strange is that the difference in colour to-day at all events is not nearly so obvious to the eye as might be imagined. When the gold objects were being cleaned in the laboratory there were several instances where hair-ribbons which seemed to be of good 'red' gold turned to a yellowish white or brass colour under the influence of heat. Reference has already been made to the electrum spearhead, U. 9122, which showed internally an almost white crystalline formation whilst the surface (though it faded later) was found of a deep colour which promised pure metal. This has been explained as due to saline action over a long period of time. It should be added, in conclusion, that such a surface amelioration of gold alloys has been practised deliberately by the modern goldsmith from the beginning of the nineteenth century; he dissolves the silver from the immediate surface, leaving a spongy gold which can be burnished down so that soldered joints, &c., are concealed beneath a casing of pure metal. We have no evidence that the Sumerian goldsmith was unacquainted with a similar technique and in the circumstances it may be that the uniform richness of his work is not entirely attributable to natural causes.

U. 8173, Pl. 139. This gold fillet, which was tied round the head by strings passed through the holes at either end, was unique in having an elaborate design covering its whole surface. The figures are not beaten up from behind, but their outline has been impressed from above in the thin soft metal. Unfortunately the gold contained a proportion of silver sufficient to make it very brittle when the latter had been transformed into crystals of chloride by the action of time ; it was broken into a number of fragments and some of these had disintegrated to powder or minute flakes which it was impossible to reconstruct or even to collect; there are therefore gaps in the metal of which one seriously affects the design. The word 'design' is really not quite appropriate, for the different subjects are all independent, arranged at haphazard, and do not illustrate any consistent theme. The eight-petalled star-like flowers at either end are made, apparently, with a chisel-edge pressed on the gold, but the other figures would seem to have been done each in one piece by means of a matrix of steatite or metal (matrices of this sort, but of later date, have been found at Ur), and it certainly looks as if the maker had simply drawn upon his stock of patterns and used them at random; their different sizes and the varying level of the base-lines by which he has tried to counteract the difference support this. The impressions were often very weak and the crushing of the metal has gone far to confuse or obliterate them ; on the photograph the lines had to be strengthened in Indian ink to make reproduction possible.

The animal figures are for the most part conventional, those of men are much more crudely drawn but much more interesting. From left to right we have a bull cropping the leaves of a tall bush, a goat, a headless animal which may be a bison, judging by its hooves and shaggy mane, and then a man who is either controlling the bison with a rope or attacking it with a spear while he kneels on one knee ; then comes a second bearded bull or bison, a calf, and then a goat, turned the other way and rearing up with his front feet on the stem of a flowering shrub ; the change of direction marks the centre of the fillet, and the antlered stag which comes next faces to the right — in spite of the absence of spots he is probably the fallow-deer. The next subject is a group of two men who face each other and hold between them what may be a rope or lasso ; they are hunters back from the chase and the man on the left carries on his back a goat tied up in a net. Then there is a naked man riding on a donkey and in front of him a curious animal which I take to be a mastiff, similar in breed to the modern Kurdish sheep-dog. All the men have long straight beards and on their heads are lines which represent either long hair or a (feather ?) head-dress ; they should be akin to the bearded hunters on the 'Standard' who bring in the wild animals from the hill country.

The grave, PG/153, in which the fillet was found was unplundered; it contained no weapons, but on the body were three necklaces and a pin, together with a cylinder seal (U. 8169, Pl. 203) having a design of cattle and birds: it is practically certain that the body was that of a woman, and this point is of interest in view of the fact that the only objects remotely resembling the gold fillet with its frieze of animals were the second diadem of Queen Shub-ad (U. 10948, PI. 141) on which again we find figures of bulls, domestic rams, gazelles, and fallow-deer, and with them rosettes of gold, golden plants, as well as the bunches of pomegranates and ears of corn which have no counterpart on the gold fillet, and the fragmentary figures from similar diadems found in two other graves (Pl. 142, U. 11776, and 143, U. 8269). Having regard to this resemblance we might argue that the haphazard nature of the ornament on the gold fillet is confined to the arrangement, but that the selection of subjects was determined by some probably religious tradition. The combination of domestic and wild animals and the fruits of the earth (it is even possible that the pendants of twisted gold wire which hang from the edge of Shub-ad's diadem are highly stylized date-clusters, and in that case the staple products of Mesopotamia would all be represented) does suggest a 'fertility charm'. If so, the animals are not chosen because of their reproductive power as amulets suitable for a man, but the result of fecundity in all its branches is associated rather with the woman : but it must be remembered that amulets tend to lose their original significance and to become mere ornaments, and what was a charm to promote fertility may change to a symbol for the multitude of possessions.

The exquisitely delicate animals of Shub-ad's diadem are made of thin gold worked over a core apparently of wood covered with bitumen; in the same technique are other gold amulets shown on Plate 142, the two gazelles whose bodies are ingeniously combined in one, U. 10943, the fish U. 10944 (also from Shub-ad's tomb), the fine bearded bull U. 8269, and the birds U. 9078 and U. 8005, though in the latter case while the fruit which the bird pecks is hollow the minute figure of the bird itself is solid ; as an instance of solid casting we have the magnificent little monkey, U. 10010, Pl. 165, from the grave of Mes-kalam-dug. From the Sargonid period onwards monkey amulets are fairly common ; they are usually made of glazed frit and always show the beast seated in the same attitude as in this early gold example; in the Larsa period (2000 B.C.) we find terra-cotta reliefs of a man carrying one monkey and leading another by a string which anticipate the late Assyrian reliefs in stone; the monkey was therefore always known, but it would be difficult to assign to it any religious significance — probably the people liked monkeys and were amused by them. Mes-kalam-dug's is the oldest yet found, and it is the best; working on a miniature scale no one could better render the character of the beast, and the details of eyes and ears and hair are extraordinarily faithful.

Casting, even by the cire-perdue process, was obviously too wasteful a method where work of any size had to be carried out in gold, and therefore the bulls' and lions' heads on Shub-ad's chariot and the gold heads on her harp and on the 'gold lyre' from PG/1237, as well as the goat figures from that death-pit (U. 12357) are all of metal hammered over a core; in the case of the bull's head from the gold lyre U. 12353, Pl. 115, the gold is heavy and thick, but in all the rest it is very thin. The same process was employed with silver for the statue of a stag on the lyre, U. 12355, PI. 112, and for copper on the double statue, U. 12356, PI. 113; but generally where these metals were concerned the animal heads which adorned musical instruments or furniture were cast, and how admirably the process was utilized is shown by the figures on the Plates 116, 117, 119, 120. The silver cow's head, U. 10916, PI. 120, has been hailed as the finest piece of animal sculpture preserved to us from antiquity; it is perfectly true to nature but at the same time an abstract from nature in which the individual is altogether lost. There is something of the same high art in the heads of lionesses, U. 10465, PI. 127; other heads are more conventional (U. 10577, PI. 120, &c.) but they preserve plenty of life and fire and are quite unlike the stereotyped creatures belonging to the First Dynasty. It is curious that the gold figures are generally much more conventional than those in silver or copper, and nowhere does convention go farther than in the heads of bearded bulls. It has been suggested that whereas the cattle of the country were of the aurochs breed such as is splendidly represented by the silver figure from the rein-ring (U. 10551, PI. 167) of the king buried in PG/789, and that therefore the artist had his models constantly before his eyes, the heavy and shaggy bison had by this time become a creature of legend and tradition. In the little figure U. 8269, PI. 143, the beard is no part of the animal at all but is tied under the chin by a string which can clearly be seen crossing over the top of the nose, and I put forward the theory that this postiche was a symbol dedicating to the god the beast which was to be sacrificed to him and really transforming it by the added emblem of divinity into the god himself, the 'great bull of heaven': the other bearded bull heads do not show any actual string, but the beard none the less looks like an addition and a rather awkward one, though this may be merely because the artist meaning to represent a bison which he had not seen could not conceive how a beard would naturally grow on a bull. The possibility that an artificial beard was intended is perhaps strengthened by the evolution of the human-headed bull which can be traced in the cylinder seals ; there it is evident that from the beginning something more than the natural bull was meant, although the human element becomes prominent only later : the same amalgamation of human and bestial elements differently combined is given by the unique casting U. 11798 on PI. 121. The head was found loose in disturbed soil and there is nothing to show to what it had belonged ; it is a god or demon, but instead of being purely human and wearing the horned head-dress which is the regular attribute of divinity it has real horns which sprout from the human skull.

Other Tools, Pls. 229, 230. The Adzes, as one would expect of well-made socketed instruments, are cast in bronze and all belong to the Early Cemetery; the finest of them, Type 5, which approximates closely to the axe types, is represented by the gold example from PG/580, U. 9339, PI. 165, and Type 2 with its upper and lower collar is also axe-like. Type 4 has only one representative, and the common Type 3 is very heavy and utilitarian.