READING HALL

THE DOORS OF WISDOM

A HISTORY OF SUMER AND AKKAD

CHAPTER III

THE AGE AND PRINCIPAL ACHIEVEMENTS OF SUMERIAN CIVILIZATION

CONSIDERABLE changes have recently taken place in our estimate of the age of Sumerian civilization, and the length of time which elapsed between the earliest remains that have been recovered and the foundation of the Babylonian monarchy. It was formerly the custom to assign very remote dates to the earlier rulers of Sumer and Akkad, and although the chronological systems in vogue necessitated enormous gaps in our knowledge of history, it was confidently assumed that these would be filled as a result of future excavation. Blank periods of a thousand years or more were treated as of little account by many writers. The hoary antiquity ascribed to the earliest rulers had in itself an attraction which outweighed the inconvenience of spreading the historical material to cover so immense a space in time. But excavation, so far from filling the gaps, has tended distinctly to reduce them, and the chronological systems of the later Assyrian and Babylonian scribes, which were formerly regarded as of primary importance, have been brought into discredit by the scribes themselves. From their own discrepancies it has been shown that the native chronologists could make mistakes in their reckoning, and a possible source of error has been disclosed in the fact that some of the early dynasties, which were formerly regarded as consecutive, were, actually, contemporaneous. Recent research on this subject has thus resulted in a considerable reduction of the early dates, and the different epochs in the history of Sumer and Akkad, which were at one time treated as isolated phenomena, have been articulated to form a consistent whole. But the tendency now is to carry the reaction rather too far, and to compress certain periods beyond the limits of the evidence. It will be well to summarize the problems at issue, and to indicate the point at which evidence gives place to conjecture.

In attempting to set limits to the earlier periods of Sumerian history, it is still impossible to do more than form a rough and approximate estimate of their duration. For in dealing with the chronology of the remoter ages, we are, to a great extent, groping in the dark. The material that has been employed for settling the order of the early kings, and for determining their periods, falls naturally into three main classes. The most important of our sources of information consists of the contemporary inscriptions of the early kings themselves, which have been recovered upon the sites of the ancient cities in Babylonia. The inscriptions frequently give genealogies of the rulers whose achievements they record, and they thus enable us to ascertain the sequence of the kings and the relative dates at which they reigned. This class of evidence also makes it possible to fix certain points of contact between the separate lines of rulers who maintained an independent authority within the borders of their city-states.

A second class of material, which is of even greater importance for settling the chronology of the later Sumerian epoch, comprises the chronological documents drawn up by early scribes, who incorporated in the form of lists and tables the history of their own time and that of their predecessors. The system of dating documents which was in vogue was not a very convenient one from the point of view of those who used it, but it has furnished us with an invaluable
summary of the principal events which took place for long periods at a time. The early dwellers in Babylonia did not reckon dates by the years of the reigning king, as did the later Babylonians, but they cited each year by the event of greatest importance which took place in it. Such events consisted in the main of the building of temples, the performance of religious ceremonies, and the conquest of neighbouring cities and states. Thus the dates upon private and official documents often furnish us with historical information of considerable importance.

Mesopotamian Early Dynastic-Marble Cylinder

But the disadvantages of the system are obvious, for an event might appear of great importance in one city and might be of no interest to another situated at some distance from it. Thus it happened that the same event was not employed throughout the whole country for designating a particular year, and we have evidence that different systems of dating were employed in different cities. Moreover, it would have required an unusually good memory to fix the exact period of a document by a single reference to an event which took place in the year when it was drawn up, more especially after the system had been in use for a considerable time. Thus, in order to fix the relative dates of documents without delay, the scribes compiled lists of the titles of the years, arranged in order under the reigns of the successive kings, and these were doubtless stored in some archive-chamber, where they were easily accessible in the case of any dispute arising with regard to the date of a particular year. It is fortunate that some of these early Sumerian date-lists have been recovered, and we are furnished by them with an outline of Sumerian history, which has the value of a contemporary record. They have thrown light upon a period of which at one time we knew little, and they have served to remove more than one erroneous supposition. Thus the so-called Second Dynasty of Ur was proved by them to have been non-existent, and the consequent reduplication of kings bearing the names of Ur-Engur and Dungi was shown to have had no foundation in fact.

From the compilation of lists of the separate years it was but a step to the classification of the reigns of the kings themselves and their arrangement in the form of dynasties. Among the mass of tablets recovered from Niffer has been found a fragment of one of these early dynastic tablets, which supplements the date-lists and is of the greatest value for settling the chronology of the later period. The reverse of the tablet gives complete lists of the names of the kings who formed the Dynasties of Ur and Isin, together with notes as to the length of their respective reigns, and it further states that the Dynasty of Isin directly succeeded that of Ur. This document fixes once for all the length of the period to which it refers, and it is much to be regretted that so little of the text has been recovered. Our information is at present confined to what is legible on part of one column of the tablet. But the text in its complete form must have contained no less than six columns of writing, and it probably gave a list of various dynasties which ruled in Babylonia from the very earliest times down to the date of its compilation, though many of the dynasties enumerated were doubtless contemporaneous. It was on the base of such documents as this dynastic list that the famous dynastic tablet was compiled for the library of Ashurbanipal at Nineveh, and the existence of such lengthy dynastic records must have contributed to the exaggerated estimate for the beginnings of Babylonian history which have come down to us from the work of Berossus.

A third class of material for settling the chronology has been found in the external evidence afforded by the early historical and votive inscriptions to which reference has already been made, and by tablets of accounts, deeds of sale, and numerous documents of a commercial and agricultural character. From a study of their form and material, the general style of the writing, and the nature of the characters employed, a rough estimate may sometimes be made as to the time at which a particular record was inscribed, or the length of a period covered by documents of different reigns. Further, in the course of the excavations undertaken at any site, careful note may be made of the relative depths of the strata in which inscriptions have been found. Thus, if texts of certain kings occur in a mound at a greater depth than those of other rulers, and it appears from an examination of the earth that the mound has not been disturbed by subsequent building operations or by natural causes, it may be inferred that the deeper the stratum in which a text is found the earlier must be the date to be assigned to it. But this class of evidence, whether obtained from palaeographical study or from systematic excavation, is sometimes uncertain and liable to more than one interpretation. In such cases it may only be safely employed when it agrees with other and independent considerations, and where additional support is not forthcoming, it is wiser to regard conclusions based upon it as provisional.

The three classes of evidence that have been referred to in the preceding paragraphs enable us to settle the relative order of many of the early rulers of Babylonia, but they do not supply us with any definite date by means of which the chronology of these earlier ages may be brought into relation with that of the later periods of Babylonian history. In order to secure such a point of connection, reliance has in the past been placed upon a notice of one of the early rulers of Babylonia, which occurs in an inscription of the last king of the Neo-Babylonian empire. On a clay cylinder of Nabonidus, which is preserved in the British Museum, it is stated that 3200 years elapsed between the burial of Naram-Sin's foundation-memorial in the temple of the Sun-god at Sippar, and the finding of the memorial by Nabonidus himself when digging in the temple's foundations. Now Naram-Sin was an early king of Akkad, and, according to later tradition, was the son of the still more famous Sargon I. On the strength of the figure given by Nabonidus, the approximate date of 3750 B.C. has been assigned to Naram-Sin, and that of 3800 B.C. to his father Sargon; and mainly on the basis of these early dates the beginning of Sumerian history has been set back as far as 5000 and even 6000 BC.

The improbably high estimate of Nabonidus for the date of Naram-Sin has long been the subject of criticism. It is an entirely isolated statement, unsupported by any other reference in early or late texts; and the scribes who were responsible for it were clearly not anxious to diminish the antiquity of the foundation-record, which had been found at such a depth below the later temple's foundations, and after so prolonged a search. To accept it as accurate entailed the leaving of enormous gaps in the chronological schemes, even when postulating the highest possible dates for the dynasties of Ur and Babylon. An alternative device of partially filling the gaps by the invention of kings and even dynasties was not a success, as their existence has since been definitely disproved. Moreover, the recent reduction in the date of the First Dynasty of Babylon, necessitated by the proof that the first three dynasties of the Kings' List were partly contemporaneous, made its discrepancy with Nabonidus's figures still more glaring, while at the same time it furnished a possible explanation of so high a figure resulting from his calculations. For his scribes in all good faith may have reckoned as consecutive a number of early dynasties which had been contemporaneous. The final disproof of the figure is furnished by evidence of an archaeological and epigraphic character. No such long interval as twelve or thirteen hundred years can have separated the art of Gudea's period from that of Naram-Sin; and the clay tablets of the two epochs differ so little in shape, and in the forms of the characters with which they are inscribed, that we must regard the two ages as immediately following one another without any considerable break.

By rejecting the figures of Nabonidus we cut away our only external connection with the chronology of the later periods, and, in order to evolve a scheme for earlier times we have to fall back on a process of reckoning from below. Without discussing in detail the later chronology, it will be well to indicate briefly the foundations on which we can begin to build. By the aid of the Ptolemaic Canon, whose accuracy is confirmed by the larger List of Kings and the principal Babylonian Chronicle, the later chronology of Babylon is definitely fixed back to the year 747 BC; by means of the eponym lists that for Assyria is fixed back to the year 911 BC. Each scheme controls and confirms the other, and the solar eclipse of June 15th, 763 BC, which is recorded in the eponymy of Pur-Sagale, places the dead reckoning for these later periods upon an absolutely certain basis. For the earlier periods of Babylonian history, as far back as the foundation of the Babylonian monarchy, a chronological framework has been supplied by the principal List of Kings. In spite of gaps in the text which render the lengths of Dynasties IV and VIII uncertain, it is possible, mainly by the help of synchronisms between Assyrian and Babylonian kings, to fix approximately the date of Dynasty III. Some difference of opinion exists with regard to this date, but the beginning of the dynasty may be placed at about the middle of the eighteenth century BC.

With regard to Dynasty II of the King's List it is now known that it ruled in the Sea-country in the region of the Persian Gulf, its earlier kings being contemporary with the close of Dynasty I and its later ones with the early part of Dynasty III. Here we come to the first of two points on which there is a considerable difference of opinion. The available evidence suggests that the kings of the Sea-country never ruled in Babylon, and that the Third, or Kassite, Dynasty followed the First Dynasty of Babylon without any considerable break. But the date 2232 BC, which probably represents the beginning of the non-mythical dynasties of Berossus, has hitherto played a considerable part in modern schemes of chronology, and, in spite of the fact that no amount of ingenuity can reconcile his dynasties with those of history, there is still a strong temptation to retain the date for the beginning of Dynasty I of the Kings' List as affording a fixed and certain point from which to start calculations. But this can only be done by assuming that some of the kings of the Sea-country ruled over the whole of Babylonia, an assumption that is negatived by such historical and archaeological evidence as we possess. It is safer to treat the date 2232 BC as without significance, and to follow the evidence in confining the kings of the Sea-country to their own land. If we do this we obtain a date for the foundation of the Babylonian monarchy about the middle of the twenty-first century BC.

The second important point on which opinion is not agreed, concerns the relation of the First Dynasty of Babylon to that of Isin. From the Nippur dynastic list we know the duration of the dynasties of Ur and Isin, and if we could connect the latter with the First Dynasty of Babylon, we should be able to carry a fixed chronology at least as far back as the age of Gudea. Such a point of connection has been suggested in the date-formula for the seventeenth year of Sin-muballit's reign, which records a capture of Isin; and by identifying this event with the fall of the dynasty, it is assumed that the kings of Isin and of Babylon overlapped for a period of about ninety-nine years. In a later chapter the evidence is discussed on which this theory rests, and it is shown that the capture of Isin in Sin-muballit's seventeenth year had nothing to do with the dynasty of that name, but was an episode in the later struggle between Babylon and Larsa. We thus have no means of deciding what interval, if any, separated the two dynasties from one another, and consequently all the earlier dates remain only approximate.

The contract-tablets dating from the period of the Dynasty of Isin, which have been found at Nippur, are said to resemble closely those of the First Babylonian Dynasty in form, material, writing, and terminology. It would thus appear that no long interval separated the two dynasties from one another. We have seen that the foundation of the Babylonian monarchy may be set in about the middle of the twenty-first century BC, and by placing the end of the Dynasty of Isin within the first half of that same century we obtain the approximate dates of 2300 BC for the Dynasty of Isin, and 2400 BC for the Dynasty of Ur. It is true that we know that the Dynasty of Ur lasted for exactly one hundred and seventeen years, and that of Isin for two hundred and twenty-five years and a half, but until we can definitely connect the Dynasty of Isin with that of Babylon, any attempt to work out the dates in detail would be misleading. We must be content to await the recovery of new material, and meanwhile to think in periods.

There is evidence that Ur-Engur established his rule in Ur, and founded his dynasty in the time of Ur-Ningirsu, the son of Gudea of Lagash. We may therefore place Gudea's accession at about 2450 BC. This date is some thirteen hundred years later than that assigned to Naram-Sin by Nabonidus. But the latter, we have already seen, must be reduced, in accordance with evidence furnished by Tello tablets, which are dated in the reigns of the intermediate patesis of Lagash. If we set this interval at one hundred and fifty years, we obtain for Naram-Sin a date of 2600 BC, and for Shar-Gani-Sharri one of 2650 BC. For the later Semitic kings of Kish, headed by Sharru-Gi, one hundred years is not too much to allow; we thus obtain for Sharru-Gi the approximate date of 2750 B.C.

It is possible that Manishtusu, King of Kish, was the contemporary of Urukagina of Lagash, but the evidence in favour of the synchronism is not sufficiently strong to justify its acceptance. By placing Urukagina at 2800 BC, we obtain for Ur-Nina an approximate date of 3000 BC, and for still earlier rulers such as Mesilim, a date rather earlier than this. It is difficult to estimate the age of the early graves, cylinder-seals and tablets found at Fara, but they cannot be placed at a much later period than 3400 BC. Thus the age of Sumerian civilization can be traced in Babylonia back to about the middle of the fourth millennium BC, but not beyond.

It must be confessed that this is a reduction in the date usually assigned to the earliest relics that have been recovered of the Sumerian civilization, but its achievements are by no means belittled by the compression of its period of development. It is not suggested that this date marks the beginning of Sumerian culture, for, as we have noted, it is probable that the race was already possessed of a high standard of civilization on their arrival in Babylonia. The invention of cuneiform writing, which was one of their most noteworthy achievements, had already taken place, for the characters in the earliest inscriptions recovered have lost their pictorial form. Assuming the genuineness of the "Blau Monuments", it must be admitted that even on them the characters are in a comparatively advanced stage of development. We may thus put back into a more remote age the origin and early growth of Sumerian culture, which took place at a time when it was not Sumerian.

In the concluding chapter of this volume an estimate is given to the extent to which Sumerian culture influenced, either directly or indirectly, other races in Asia, Egypt, and the West. In such matters the interest attaching to the Sumerian original is largely derived from its effects, and its study may be undertaken mainly with the view of elucidating a later development. But one department of Sumerian activity forms a striking exception to this rule. The arts of sculpture and engraving, as practised by the Sumerians, are well worthy of study on their own account, for while their work in all periods is marked by spirit and originality, that of the later time reaches a remarkable standard of excellence. The improvement in technique observable in the later period may largely be due to the influence of Semitic work, which was derived from Sumer and reacted in its turn on the parent stem. But the original impulse to artistic production was of purely Sumerian origin, and it is possible to trace the gradual development of its products from the rudest reliefs of the archaic period to the finished sculpture of Gudea's reign. The character of the Semitic art of Akkad was secondary and derivative, though the Semites certainly improved on what they borrowed; in that of the Sumerians the seeds of its later excellence may be detected from the beginning. The most ancient of the sculptured reliefs of the Sumerians are very rudely cut, and their age is attested not only by their primitive character, but also by the linear form of the writing which is found upon them. These, owing to their smaller size, are the best preserved, for the later reliefs, which belong to the period when Sumerian art reached its fullest development, are unfortunately represented only by fragments. But they suffice to show the spirit which animated these ancient craftsmen, and enabled them successfully to overcome difficulties of technique which were carefully avoided by the later sculptors of Assyria. To take a single instance, we may note the manner in which they represented the heads of the principal figures of a composition in full-face, and did not seek to avoid the difficulty of foreshortening the features by a monotonous arrangement in profile. A good example of their bolder method of composition is afforded by the relief of a god, generally identified with Ningirsu, which dates from the epoch of Gudea; he is seated upon a throne, and while the torso and bearded head are sculptured full-face, the legs are in profile. On another fragment of a relief of the same period, beautifully cut in alabaster but much damaged by fire, a goddess is represented seated on the knees of a god. The rendering of the group is very spirited, for while the god gazes in profile at his wife, she looks out from the sculpture curving her body from the hips.

In neither instance can it be said that the sculptor has completely succeeded in portraying a natural attitude, for the head in each case should be only in three-quarter profile, but such attempts at an unconventional treatment afford striking evidence of the originality which characterized the work of the Sumerians. Both the sculptures referred to date from the later Sumerian period, and, if they were the only instances recovered, it might be urged that the innovation should be traced to the influence of North Babylonian art under the patronage of the kings of Akkad. Fortunately, however, we possess an interesting example of the same class of treatment, which undoubtedly dates from a period anterior to the Semitic domination. This is afforded by a perforated plaque, somewhat similar to the more primitive ones of Ur-Nina, engraved in shallow relief with a libation scene. The figure of a man, completely nude and with shaven head and face, raises a libation-vase with a long spout, from which he is about to pour water into a vase holding two palm leaves and a flowering branch. The goddess in whose honour the rite is being performed is seated in the mountains, represented as in later times by a number of small lozenges or half circles. While her feet and knees are in profile, the head is represented full-face, and the sculptor's want of skill in this novel treatment has led him to assign the head a size out of all proportion to the rest of the body. The effect is almost grotesque, but the work is of considerable interest as one of the earliest attempts on the part of the Sumerian sculptors to break away from the stiff and formal traditions of the archaic period. From the general style of the work the relief may probably be dated about the period of Eannatum's reign.

The Sumerians did not attain the decorative effect of the Assyrian bas-reliefs with which the later kings lined the walls of their palaces. In fact, the small size of the figures rendered them suitable for the enrichment of stelae, plaques, basins and stone vases, rather than for elaborate wall sculptures, for which in any case they had not the material. The largest fragment of an early bas-relief that has been recovered appears to have formed the angle of a stone pedestal, and is decorated with figures in several registers representing ceremonies of Sumerian worship. In the upper register on the side that is best preserved is a priest leading worshippers into the presence of a god, while below is a crouching figure, probably that of a woman who plays on a great lyre or harp of eleven cords, furnished with two uprights and decorated with a horned head and the figure of a bull. On the side in the upper row is a heavily bearded figure on a larger scale than the rest, and the mixture of Sumerian and Semitic types in the figures preceding him suggests that the monument is to be assigned to the period of Semitic domination, under the rule of the kings of Kish or Akkad. But it is obviously Sumerian in character, resembling the work of Gudea's period rather than that of Naram-Sin.

The perfection of detail which characterized the best work of the Sumerian sculptors is well illustrated by two fragments of reliefs, parts of which are drawn in outline in the accompanying blocks. The one on the left is from a bas-relief representing a line of humped cattle and horned sheep defiling past the spectator. It is badly broken, but enough is preserved to show the surprising fidelity with which the sculptor has reproduced the animal's form and attitude. Though the subject recalls the lines of domestic animals upon the Assyrian bas-reliefs, the Sumerian treatment is infinitely superior. The same high qualities of design and workmanship are visible in the little fragment on the right. Of the main sculpture only a human foot remains; but it is beautifully modelled. The decorative border below the foot represents the spouting vase with its two streams of water and two fish swimming against the stream. A plant rises from the vase between the streams, the symbol of vegetation nourished by the waters. The extreme delicacy of the original shows to what degree of perfection Sumerian work attained during the best period.

The use of sculpture in relief was also most happily employed for the decoration of basins or fountains. The most elaborate of those recovered, unhappily represented by mutilated fragments only, was decorated on the outside with a chain of female figures passing from hand to hand vases of spouting water. Better preserved are the remains of another basin, which was set up by Gudea in Ningirsu's temple at Lagash. Rectangular in shape, each corner was decorated with a lion. The head, drawn in the accompanying block, is a fine piece of sculpture, and almost stands out from the corner, while the body, carved in profile on the side of the basin, is in low relief. In this portrayal of a lion turning its head, the designer has formed a bold but decorative combination of relief with sculpture in the round.

The most famous examples of Sumerian sculpture are the statues of Gudea, and the rather earlier one of Ur-Bau, which, however, lose much of their character by the absence of their heads. It is true that a head has been fitted to a smaller and more recently found figure of Gudea; but this proves to be out of all proportion to the body—a defect that was probably absent from the larger statues. The traditional attitude of devotion, symbolized by the clasping of the hands over the breast, gives them a certain monotony; but their modelling is superior to anything achieved by the Babylonians and Assyrians of a later time. Thus there is a complete absence of exaggeration in the rendering of the muscles; the sculptor has not attempted by such crude and conventional methods to ascribe to his model a supernatural strength and vigour, but has worked direct from nature. They are carved in diorite, varying in colour from dark green to black, and that so hard a material should have been worked in the large masses required, is in itself an achievement of no small importance, and argues great technical skill on the part of the sculptors of the later period.

For smaller figures and statuettes a softer stone, such as white limestone, alabaster, or onyx, was usually employed, but a few in the harder stone have been recovered. The most remarkable of these is a diorite statuette of a woman, the upper part of which has been preserved. The head and the torso were found separately, but thanks to their hard material they join without leaving a trace of any break. Here, as usual, the hands are crossed upon the breast, and the folds of the garment are only indicated under the arms by a few plain grooves as in the statues of Gudea. But the woman's form is visible beneath the stuff of her garment, and the curves of the back are wonderfully true. Her hair, undulating on the temples, is bound in a head-cloth and falls in the form of a chignon on the neck, the whole being secured by a stiff band, or fillet, around which the cloth is folded with its fringe tucked in.

The drawing in Fig. 23 scarcely does justice to the beauty of the face, since it exaggerates the conventional representation of the eyebrows, and reproduces the texture of the stone at the expense of the outline. Moreover, the face is almost more striking in profile. The nose, though perfectly straight, is rather large, but this is clearly a racial characteristic. Even so, the type of female beauty portrayed is singularly striking, and the manner in which the Sumerian sculptor has succeeded in reproducing it was not approached in the work of any later period. Another head from a female statuette, with the hair dressed in a similar fashion, is equally beautiful. The absence of part of the nose tends to give it a rather less marked ethnographic character, and probably increases the resemblance which has been claimed for it to types of classical antiquity.

The art of casting in metal was also practised by the Sumerians, and even in the earliest period, anterior to the reign of Ur-Nina, small foundation-figures have been discovered, which were cast solid in copper. In fact, copper was the metal most commonly employed by the Sumerians, and their stage of culture throughout the long period of their history may be described as a copper age, rather than an age of bronze. It is true that the claim is sometimes put forward, based on very unsatisfactory evidence, that the Sumerian metal-founders used not only tin but also antimony in order to harden copper, and at the same time render it more fusible; and it is difficult to explain the employment of two ideograms for the metal, even in the earlier periods, unless one signified bronze and the other copper. But a careful analysis by M. Berthelot of the numerous metal objects found at Tello, the dates of which can be definitely ascertained, has shown that, even under the later rulers of Lagash and the kings of Ur, not only votive figures, but also tools and weapons of copper, contain no trace of tin employed as an alloy. As at Tello, so at Tell Sifr, the vessels and weapons found by Loftus are of copper, not bronze. The presence of an exceedingly small proportion of elements other than copper in the objects submitted to analysis was probably not intentional, but was due to the necessarily imperfect method of smelting that was employed.

No trace has yet been found of any mould used by the Sumerians in the process of casting metal, but we may assume that clay was employed both for solid and hollow castings. While many figures of the same form have been found, no two are exactly alike nor of quite the same proportions, so that it may be inferred that a mould was never used a second time, but that each was broken in order to remove the casting. The copper foundation-figures usually take the form of nails, terminating with the bust of a female figure, and they were set in a socket beneath stone foundation-inscriptions which they support. Later, votive objects, cast in copper, represent male figures, bearing on their heads the builder's basket, in which is clay for the sacred bricks of the temple's foundation; or they consist of great cones or nails supporting a recumbent bull, or clasped by the kneeling figure of a god. Large figures of wood were sometimes covered with thin plates of copper joined by a series of small nails or rivets, as is proved by the horn of a bull of natural size, which has been discovered at Tello. But hollow castings in copper of a considerable size have also been found. A good example is the bull's head, figured in the accompanying block, which probably dates from a period not later than the close of Ur-Nina's dynasty. Its eyes are inlaid with mother-of pearl and lapis-lazuli, and a very similar method of inlaying is met with in the copper head of a goat which was found at Fara.

A far simpler process of manufacture was employed for the making of votive figures of terra-cotta, which, in order of development, preceded the use of metal for this purpose, though they continued to be manufactured in considerable quantities during the later periods. Here the mould, in a single piece, was cut in stone or some other hard material, and the clay, after being impressed into it, was smoothed down on the back by hand. The flat border of clay left by the upper surface of the mould, was frequently not removed, so that the figures are sometimes found standing out from a flat background in the manner of a sculptured plaque, or bas-relief. In the period of Gudea, the mould was definitely used as a stamp, thus returning to the original use from which its later employment was developed. Interesting examples of such later stamped figures include representations of a god wearing a horned headdress, to which are added the ears of a bull, and of a hero, often identified with Gilgamesh, who holds a vase from which two streams of water flow. The clay employed for the votive figures is extremely fine in quality, and most of them are baked to a degree of hardness resembling stone or metal.

The art of inlaying was widely practised by the Sumerians, who not only treated metal in this way, but frequently attempted to give more expression or life to stone statues by inlaying the white of the eye with mother-of-pearl or shell, and representing the pupil and iris by lapis-lazuli or bitumen. A similar method was employed to enrich votive stone figures of animals, and to give a varied and polychrome effect to vases carved in stone. The finest example of this class of work is a libation-vase of Gudea made of dark green steatite, which was dedicated by him to his patron deity Ningishzida. The vase has a short projecting spout running up from the base and grooved, so as to allow only a small stream of liquid to escape during the pouring of a libation. Its scheme of decoration is interesting as it affords an excellent example of the more fantastic side of Sumerian art, inspired by a large and important section of the religious belief. The two intertwined serpents, whose tongues touch the point where the liquid would leave the vase, are modelled from nature, but the winged monsters on each side well illustrate the Sumerian origin of later Babylonian demonology.

It is probable that such composite monsters, with the bodies and heads of serpents and the wings and talons of birds, were originally malevolent in character, but here, like the serpents, they are clearly represented as tamed, and in the service of the god to whom the vase was dedicated. This is sufficiently proved by the ringed staffs they carry, their modified horned headdresses, and their carefully twisted locks of hair. They were peculiarly sacred to Ningishzida and in Fig. 12 they may be seen rising as emblems from his shoulders. The rich effect of the dark green steatite was originally enhanced by inlaying, for the bodies of the dragons are now pitted with deep holes. These were no doubt originally inlaid with some other material, probably shell, which has been found employed for this purpose in a fragment of a vase of a very similar character.

In the same category with the monsters on the vase we may class the human-headed bulls, of which small sculptured figures, in a recumbent attitude, have been found at Tello; these were afterwards adopted by the Assyrian kings, and employed as the colossal guardians of their palace door-ways. The extent to which this particular form of composite monster was employed for religious and decorative purposes may be seen on the cylinder-seals, upon which in the earlier period it represents the favourite device. Examples are frequently found in decorative combinations, together with figures of early bearded heroes, possibly to be identified with Gilgamesh, and with a strange creature, half-man and half-bull, resembling the later descriptions of Eabani, who strive with lions and other animals. Gudea's catalogue of the temple furniture and votive objects, with which he enriched E-ninnu, throws light upon the manner in which Sumerian art reflected this aspect of the Sumerian religion. Some of the legends and beliefs may well have been derived from Semitic sources, but the imagery, which exerted so strong an influence upon the development of their art, may probably be traced to the Sumerians themselves.

The engraving upon cylinder-seals during the Sumerian period appears to have been done generally by hand, without the help of a drill or a revolving tool. Outline engraving with the point was also practised, that on stone having probably preceded the use of the bas-relief, but it continued to be employed in the later periods for the decoration of metal and shell. The finest example of metal engraving is the silver vase of Entemena, around which is incised in outline a decorative band, consisting of variations of the emblem of Lagash, arranged beneath a row of seven calves. But the largest number of designs engraved in outline have been found, not upon stone or metal, but upon shell. It is an interesting fact that among the smaller objects found by M. de Sarzec at Tello, there is not a single fragment of ivory, and it would seem that this material was not known to the earliest inhabitants of Babylonia, a fact which has some bearing on the disputed question of their relations to Egypt, and to the earlier stages of Egyptian culture.

From the earliest period at Lagash fragments of shell were employed in place of ivory, and the effect produced by it is nearly the same. Certain species of great univalves or conch-shells, which are found in the Indian Ocean, have a thick core or centre, and these furnished the material for a large number of the earliest cylinder-seals. Small plaques or lozenges could also be obtained from the core by sectional cutting, while the curved part of the shell was sometimes employed for objects to which its convex form could be adapted. The numerous flat lozenges that have been found are shaped for inlaying furniture, caskets, and the like, and curved pieces were probably fitted to others of a like shape in order to form small cups and vases. Each piece is decorated with fine engraving, and in nearly every instance the outline is accentuated by the employment of a very slight relief. The designs are often spirited, and they prove that even in the earliest periods the Sumerian draughtsman had attained to a high standard of proficiency.

One of the most interesting engraved fragments that have been recovered consists of a slightly curved piece of shell, which probably formed part of a small bowl or cup. The rest of the side seems to have been built up of pieces of similar shape, held together by bitumen, or, more probably, fitted to a metal lining by rivets through holes in the shell. The scene engraved upon the fragment represents a lion seizing a bull in a thicket of shrubs or high flowering plants. Though the group upon the fragment is complete in itself, there are indications that it formed only part of a more elaborate composition. For in the space on the right of the fragment behind the lion's mane are engraved two weapons. The upper one is a hilted dagger with its point towards the lion; this may be compared with the short daggers held by the mythological beings resembling Ea-bani upon one of Lugal-anda's seals, with which they are represented as stabbing lions in the neck. Below is a hand holding a curved mace or throwing stick, formed of three strands bound with leather thongs or bands of metal, like that held by Eannatum upon the Stele of the Vultures. It is, therefore, clear that on the panel to the right of the lion and bull a king, or patesi, was represented in the act of attacking the lion, and we may infer that the whole of the cup was decorated with a continuous band of engraving, though some of the groups in the design may have been arranged symmetrically, with repetitions such as are found upon the earlier cylinder-seals.

The position of the lion upon the fragment, represented with luxuriant mane and with head facing the spectator, and the vigour of the design as a whole combined with certain inequalities of treatment, have suggested a comparison with the lions upon the sculptured mace-head of Mesilim. The piece has, therefore, been assigned to the epoch of the earlier kings of Kish, anterior to the period of Ur-Nina. It may perhaps belong to the rather later period of Ur-Nina's dynasty, but, even so, it suffices to indicate the excellence in design and draughtsmanship attained by the earlier Sumerians. In vigour and originality their representations of animals were unequalled by those of the later inhabitants of Babylonia and Assyria, until shortly before the close of the Assyrian empire. But the Sumerian artists only gradually acquired their skill, and on some of the engraved fragments recovered it is possible to trace an advance on earlier work. The designs in the accompanying blocks have been selected as illustrating, to some extent, the change which gradually took place in the treatment of animal forms by the Sumerians.

Of the three designs, that on the left is engraved upon a convex piece of shell, thin as the shell of an egg; it represents a lion-headed eagle which has swooped down upon the back of a human-headed bull and is attacking him with mouth and claws. The subject resembles that found upon the most primitive Sumerian cylinder-seals, and its rough and angular treatment is sufficient indication of the very archaic character of the work. The central panel resembles in shape that of the lion and the bull. The design represents a leaping ibex with flowering plants in the background, and the drawing is freer and less stiff than that of the animals on the silver vase of Entemena. Some archaic characteristics may still be noted, such as the springing tufts of hair at the joints of the hind legs; but the general treatment of the subject marks a distinct advance upon the archaic conventions of the earlier fragment. The third design is that of a leaping kid, engraved upon a flat piece of shell and cut out for inlaying. Here the drawing is absolutely true to nature, and the artist has even noted the slight swelling of the head caused by the growing horns.

The Sumerians do not appear to have used complete shells for engraving, like those found on Assyrian and Aegean sites. A complete shell has indeed been recovered, but it is in an unworked state and bears a dedicatory formula of Ur-Ningirsu, the son and successor of Gudea. Since it is not a fine specimen of its class, we may suppose that it was selected for dedication merely as representing the finer shells employed by the workmen in the decoration of the temple-furniture. The Sumerians at a later period engraved designs upon mother-of-pearl. When used in plain pieces for inlaying it certainly gave a more brilliant effect than shell, but to the engraver it offered greater difficulties in consequence of its brittle and scaly surface. Pieces have been found, however, on which designs have been cut, and these were most frequently employed for enriching the handles of knives and daggers. The panels in the accompanying blocks will serve to show that the same traditional motives are reproduced which meet us in the earlier designs upon fragments of shell and cylinder-seals. They include a bearded hero, the eagle attacking the bull, a hero in conflict with a lion, the lion-headed eagle of Lagash, a winged lion, a lion attacking an ibex, and a stag. Even when allowance is made for the difficulties presented by the material, it will be seen that the designs themselves rank far below those found upon shell. The employment of mother-of-pearl for engraving may thus be assigned to a period of decadence in Sumerian art when it had lost much of its earlier freshness and vigour.

The above brief sketch of the principal forms and productions of Sumerian art may serve to vindicate the claim of the Sumerians to a place among the more artistic races of antiquity. Much oriental art is merely quaint, or interesting from its history and peculiarities, but that of the Sumerians is considerably more than this. Its sculpture never acquired the dull monotony of the Assyrian bas-reliefs with their over-elaboration of detail, intended doubtless to cloak the poverty of the design. Certain conventions persisted through all periods, but the Sumerian sculptor was never a slave to them. He relied largely on his own taste and intelligence, and even the earliest work is bold and spirited. After centuries of independent development fresh vigour was introduced by the nomad Semitic races who settled in the north, but in the hands of the later Semites the Sumerian ideals were not maintained. For the finest period of Babylonian art we must go back to a time some centuries before the founding of the Babylonian monarchy.

CHAPTER IV

THE EARLIEST SETTLEMENTS IN SUMER

; THE DAWN OF HISTORY AND THE RISE OF LAGASH