CHAPTER 4

FROM VILLAGE TO CITY

The story of the passage from Neolithic to History, from the humble villages of the Zagros foothills to the relatively large and highly civilized Sumerian cities of the lower Tigris–Euphrates valley cannot be told in full detail because our information, though rapidly progressing, remains imprecise and patchy. Yet each new prehistoric tell excavated, each buried city dug down to the virgin soil, confirms what forty years of archaeological research in Iraq already suggested: the Sumerian civilization was never imported ready-made into Mesopotamia from some unknown country at some ill-defined date. Like all civilizations – including ours – it was a mixed product shaped by the mould into which its components were poured over many years. Each of these components can now be traced back to one stage or another of Iraqi prehistory, and while some were undoubtedly brought in by foreign invasion or influence, others had roots so deep in the past that we may call them indigenous. In addition, excavations conducted at an ever increasing pace in Iran, Syria, Palestine and Turkey at the same time as in Iraq have thrown considerable light on the interplay of Neolithic and Chalcolithic cultures in the Near East and have supplied enough comparative material and radiocarbon dates to draw up a rough, tentative chronological scale along the six divisions of Mesopotamian proto-history:

Each of these periods is characterized by a distinct cultural assemblage and has been named after the site, not necessarily the largest nor even the most representative, where this assemblage was first identified.

As will be seen, the areas covered by these cultures vary from one period to the other; moreover, cultures long thought to be successive are in fact contemporaneous or at least overlapping, and within each period there is room for a variety of regional and interesting subcultures. The above divisions therefore, are somewhat artificial, but they provide a convenient framework into which can be fitted the changes that occurred during those three millennia when Mesopotamia was pregnant, so to speak, with Sumer.¹

The Hassuna Period

The site type for this period is Tell Hassuna, a low mound thirty-five kilometres south of Mosul, excavated in 1943 – 4 by the Iraqi Directorate of Antiquities under the direction of Seton Lloyd and Fuad Safar.² There, resting on the virgin soil, were coarse pottery and stone implements suggestive of a Neolithic farming community living in huts or tents, for no trace of building was found. Overlying this primitive settlement, however, were six layers of houses, progressively larger and better built. In size, plan and building material these houses were very similar to those of present-day northern Iraqi villages. Six or seven rooms were arranged in two blocks around a courtyard, one block serving as living quarters, the other as kitchen and stores. The walls were made of pressed mud, the floors paved with a mixture of clay and straw. Grain was kept in huge bins of unbaked clay sunk into the ground up to their mouths, and bread was baked in domed ovens resembling the modern tanur. Mortars, flint sickle-blades, stone hoes, clay spindle-whorls and crude clay figurines of naked and apparently seated women were present. Large jars found inside the houses contained the bones of deceased children accompanied by tiny cups and pots for after-life refreshment while, strangely enough, much liberty seems to have been taken with the disposal of adult skeletons piled up in the corner of a room, thrown into clay bins ‘without ceremony’ or buried in cist graves without the usual funerary gifts. The few skulls that have been studied belong, like those from Byblos and Jericho, to a ‘large-toothed variety of the long-headed Mediterranean race’, which suggests a unity of population throughout the Fertile Crescent in late Neolithic times.³

The pottery discovered at Hassuna has been divided into two categories called ‘archaic’ and ‘standard’. The archaic ware ranges from level Ia, at the bottom of the tell, to level III and is represented by: (1) tall, round or pear-shaped jars of undecorated coarse clay; (2) bowls of finer fabric varying in colour from buff to black according to the method of firing and ‘burnished’ with a stone or bone, and (3) bowls and globular jars with a short, straight neck, sparingly decorated with simple designs (lines, triangles, cross-hatchings) in fragile red paint and also burnished. The Hassuna standard ware, predominant in levels IV to VI, is made up of the same painted bowls and jars and the designs are very similar, but the paint is matt brown and thicker, the decoration more extensive and executed with greater skill. A number of vessels are almost entirely covered by shallow incisions, and some are both painted and incised.

While the archaic pottery has several traits in common with that found in the deepest layers of Turkish (Sakçe Gözü, Mersin), Syrian (Kerkemish, ‘Amuq plain) and Palestinian (Megiddo, Jericho) sites, the standard pottery seems to have developed locally⁴ and is distributed over a relatively small area of northern Iraq. Sherds of Hassuna ware can be picked up on the surface of many unexcavated mounds east and west of the Tigris down to Jabal Hamrin, and complete specimens have been found in the lowest levels of Nineveh, opposite Mosul, at Matarrah,⁵ south of Kirkuk and at Shimshara⁶ in the Lower Zab valley. They were also present throughout the thirteen levels of mound 1 at Yarim Tepe,⁷ near Tell ‘Afar, associated with the remains of square or round houses, with tools and weapons of flint and obsidian, with pieces of copper ore and a few copper and lead ornaments, with small seated clay figurines and with minute stone or clay discs with a loop at the back, engraved with straight lines or criss-cross patterns. These objects, probably worn on a string around the neck, may have been impressed as a mark of ownership on lumps of clay fastened to baskets or to jar stoppers, in which case they would represent the earliest examples of the stamp-seal, and the stamp-seal is the forerunner of the cylinder-seal, a significant element of the Mesopotamian civilization. Some authors, however, regard them, at least in this period, as mere amulets or ornaments.

Forty-eight kilometres due south of Yarim Tepe, at the limit of the rain-fed plain and the desert of Jazirah, lies Umm Dabaghiya, excavated by Diana Kirkbride between 1971 and 1973.⁸ Umm Dabaghiya was a small settlement, a simple trading post where nomads from the desert brought the onagers and gazelles they had hunted to be skinned, the raw hides being later sent elsewhere to be tanned. Related by its coarse and painted pottery to the archaic levels of Hassuna but most probably older, the site has many distinctive and strangely sophisticated features. For instance, the floors of the houses are often made of large clay slabs which announce the moulded bricks of later periods; floors and walls are carefully plastered with gypsum and frequently painted red, and in one building were found fragments of frescoes representing an onager hunt, a spider with its eggs and perhaps flying vultures. Several houses contained alabaster bowls beautifully carved and polished. Predominant among the clay vessels are bowls and jars with ‘applied decoration’, i.e. small figurines of animals and human beings stuck on the vessels before firing. Other sites representative of this Hassunan subculture are Tell Sotto and Kül Tepe,⁹ near Yarim Tepe, and mound 2 at Tulul ath-Thalathat,¹⁰ in the same Tell ‘Afar area. Not unexpectedly for places lying on the trade routes to the west and north-west, some

Caption
Buildings, potteries, figurines, seals and tools characteristic of the Hassuna, Halaf and Ubaid periods.

elements of the ‘Umm Dab culture’, such as plastered floors and arrow-heads, point to Syria (Buqras on the Euphrates and even Ras Shamra and Byblos), whilst the red and frescoed walls are reminiscent of contemporary Çatal Hüyük in remote Anatolia.

The Samarra Period

In the upper levels of Hassuna, Matarrah, Shimshara and Yarim Tepe the Hassuna ware is mixed with, and gradually replaced by, a much more attractive pottery known as Samarra ware because it was first discovered, in 1912 – 14, in a prehistoric cemetery underneath the houses of the medieval city of that name, famous for its spiralled minaret.¹¹ On the pale, slightly rough surface of large plates, around the rim of carinate bowls, on the neck and shoulder of round-bellied pots, painted in red, dark-brown or purple, are geometric designs arranged in neat, horizontal bands or representations of human beings, birds, fish, antelopes, scorpions and other animals. The motifs are conventionalized, but their distribution is perfectly well balanced and they are treated in such a way as to give an extraordinary impression of movement. The people who modelled and painted such vessels were undoubtedly great artists, and it was long thought that they had come from Iran, but we now know that the Samarra ware was indigenous to Mesopotamia and belonged to a hitherto unsuspected culture which flourished in the middle Tigris valley during the second half of the sixth millennium B.C.

This culture was revealed in the 1960s by the Iraqi excavations at Tell es-Sawwan, a low but large mound on the left bank of the Tigris, only eleven kilometres to the south of Samarra.¹² The inhabitants of Tell es-Sawwan were peasants like their Hassunan ancestors and used similar stone and flint tools, but in an area where rain is scarce they were the first to practise a primitive form of irrigation agriculture, using the Tigris floods to water their fields and grow wheat, barley and linseed.¹³ The yield must have been substantial if the large and empty buildings found at various levels were really ‘granaries’ as has been suggested. The central part of the village was protected from invaders by a 3-metre-deep ditch doubled by a thick, buttressed mud wall. The houses were large, very regular in plan, with multiple rooms and courtyards, and it must be noted that they were no longer built of pressed mud, but of large, cigar-shaped mud bricks plastered over with clay or gypsum. A thin coat of plaster covered the floors and walls. Apart from numerous pots and plates of coarse or fine Samarra ware, these houses contained exquisite, translucent marble vessels. The bodies of adults, in a contracted position and wrapped in matting coated with bitumen, and of children, placed in large jars or deep bowls, were buried under the floors, and it is from these graves that have come the most exciting finds in the form of alabaster or terracotta statuettes of women (or occasionally men) squatting or standing. Some of the clay statuettes have ‘coffee-bean’ eyes and pointed heads that are very similar to those of the Ubaid period figurines, whilst other clay or stone statuettes have large, wide-open eyes inlaid with shell and bitumen and surmounted by black eyebrows, that are ‘astonishingly reminiscent of much later Sumerian technique’.¹⁴ Could the Samarran folk be the ancestors of the ‘Ubaidians’ and even perhaps of the Sumerians?

So far, no other settlement comparable to Tell es-Sawwan has been excavated,¹⁵ but apart from copies or imports in Baghuz, on the middle Euphrates, and Chagar Bazar, in central Jazirah, the Samarra pottery has been found in a limited but fairly wide area along the Tigris valley, from Nineveh to Choga Mami near Mandali, on the Iraqi–Iranian border.¹⁶ In the latter site, where canal-irrigation was practised, not only do we find statuettes resembling the ‘coffee-bean’-eyed statuettes of Sawwan, but the Samarra ware seems to have developed locally into new ceramic types (called ‘Choga Mami Transitional’) similar to the Eridu and Hajji Muhammad wares of southern Iraq, themselves considered as early forms of the Ubaid pottery.¹⁷ This unexpected discovery might provide the beginning of an answer to our question.

The Halaf Period

The third period of proto-historic Mesopotamia takes its name from Tell Halaf, a large mound overlooking the Khabur river near the village of Ras el-‘Ain, on the Turkish–Syrian border. There, just before the First World War, a German archaeologist, Max Freiherr von Oppenheim, came upon a thick layer of beautifully painted pottery immediately beneath the palace of an Aramaean ruler of the tenth century B.C. The discovery was not published until 1931.¹⁷ At that time little was known of Near Eastern prehistory and the date of von Oppenheim's ‘Buntkeramik’ was the subject of much controversy. But in the following years British excavations a Nineveh,¹⁸ Tell Arpachiyah near Mosul¹⁹ and Tell Chagar Bazar,²⁰ as well as American excavations at Tepe Gawra, put the Halaf period into its proper chronological place and supplied a complete assortment of its cultural assemblage. The Russian excavations of mound 2 at Yarim Tepe and, more recently, the stratigraphic exploration of Arpachiyah by the Iraqi Ismail Hijara,²¹ as well as soundings and partial excavations of several sites in the Hamrin basin and the upper Tigris valley, have considerably added to our knowledge.²²

Compared with the previous cultures, the Halaf culture offers a number of new and highly distinctive features. The settlements are still of village type and size, but cobbled streets, at least at Arpachiyah, indicate some municipal caretaking. Pressed mud or mud bricks remain the standard building materials, but rectangular houses tend to be smaller than before while round houses called tholoi (plural of tholos) by analogy with the Mycenaean tombs of much later date become predominant. The tholoi of Yarim Tepe are usually small; some are divided into two rooms, others are surrounded by rectangular rooms or concentric walls of pressed mud. Those of Arpachiyah, however, are much larger structures, up to 10 metres in diameter; they rest on stone foundations and to some of them is appended a long antechamber which further increases the resemblance with the Mycenaean tombs. Since they had been built and rebuilt with great care and since they were found empty, it was long thought that they were shrines or temples, but the finds at Yarim Tepe clearly show that most tholoi were simple, beehive-shaped houses such as can still be seen around Aleppo, in northern Syria. In fact, the only building of that period that might be considered a sanctuary is a small, square structure with mud pedestals and an ox skull on the threshold of a doorway, excavated by Mallowan at Tell Aswad, on the Balikh river. At Arpachiyah the dead were buried in pits beneath the floors or around tholoi, but there are examples of collective burials of dismembered bodies there as at Tepe Gawra and of cremation, perhaps for ritual purposes, at Yarim Tepe.

No less interesting than the tholoi are some of the small objects found at Arpachiyah and elsewhere. We allude, in particular, to amulets in the form of a house with gabled roof, a bull's head or a double-axe, and to terracotta figurines of doves and women. The latter are not new in Mesopotamia, but they now differ from previous models. The woman is usually squatting or sitting on a round stool, her arms supporting her heavy breasts. The head is reduced to a shapeless lump, but the body is realistic and covered with painted strips and dots which may stand for tattoo marks, jewels or clothes. It is probable that these figurines were talismans against sterility or the hazards of childbirth rather than ‘Mother Goddesses’, as too often assumed.

Last but not least comes a very remarkable painted pottery, the most beautiful ever used in Mesopotamia.²³ The Halaf ware is made by hand in a fine, ferruginous clay slightly glazed in the process of firing. The walls of the vessels are often very thin, the shapes varied and daring: round pots with large, flaring necks, squat jars with rolled-out rims, footed chalices, large and deep ‘cream bowls’ with an angular profile. The decoration perhaps

Examples of decorated pottery in proto-historic Mesopotamia: i. Neolithic (Jarmo); 2 – 3, Hassuna culture (3 is an incised jar); 4 – 6, Samarra culture; 7, Eridu (Ubaid 1) culture; 8, Hajji Muhammad (Ubaid 2) culture; 9–10, Halaf culture; 11–13, Ubaid 3 and 4 cultures; 14–15, Nineveh V culture; 16 Jemdat Nasr culture. All drawings are not on the same scale.

lacks the bold movement of the Samarra ware, but it is perfectly adapted to the shapes, minutely executed and pleasant to the eye in the manner of Persian rugs. On a cream or peach ‘slip’ is laid, originally in black and red, later in black, red and white, a closely woven pattern covering most of the vessel. Triangles, squares, checks, crosses, scallops and small circles are among the favourite designs, though flowers, sitting birds, crouching gazelles and even a leaping cheetah are also encountered. Most characteristic of all and perhaps loaded with religious symbolism are the double-axe, the ‘Maltese square’ (a square with a triangle on each corner) and the bucranum, or stylized bull's head.

It has recently been proven by neutron activation analysis²⁴ that this attractive pottery was manufactured in large quantities in certain specialized centres, such as Arpachiyah, Tell Brak, Chagar Bazar and Tell Halaf, and exported to specific settlements from which it gradually reached more distant places. The people who transported this ware (perhaps on the back of cattle or on ox-drawn sledges) presumably returned loaded with such ‘luxury’ goods as marine shells, gem stones and particularly obsidian, which is predominant in most Halafian sites. It has also been suggested that the Halafian formed a ‘ranked society’ (i.e. with social, but not economic, classes) and that the pottery-producing centres were the residences of local chieftains. The inhabitants of these relatively small villages were farmers and pastoralists. They grew emmer, wheat, einkorn, barley, lentils, flax and other vegetables and bred sheep, goats, pigs, cattle and domestic dogs.

Judging from the distribution of true Halaf pottery, at the peak of its expansion the core of the Halaf culture occupied a wide, crescent-shaped area entirely located in the dry-farming zone. It extended from the region of Aleppo to the Diyala valley, covering the whole of Jazirah and of future Assyria, and it was surrounded by a halo of peripheral areas where this pottery was copied or merely imported; these included the heart of eastern Anatolia, Cilicia and northern Syria up to the Mediterranean coast, the Harim basin and parts of western Iran and Transcaucasia.

While the Samarra culture may be regarded as a derivative of the Hassuna culture, the Halaf culture has no ancestor in prehistoric Mesopotamia. It is strikingly intrusive and clearly has some connections with Anatolia (virtually all the symbolic designs painted on the Halaf ware and many of the artefacts already described are reminiscent of those found in Anatolian Neolithic sites), but it is not possible to be more precise at present.²⁵ Whatever the origin of the ‘Halafians’, there is no evidence of brutal invasion; in fact all we know of them points to a slow infiltration of peaceful people who came to settle in regions that might have then been sparsely populated.

The Ubaid Period

Between 4500 and 4300 B.C. several Halafian settlements in northern Mesopotamia were abandoned, while in many others the tholoi and the painted pottery typical of the Halaf culture were gradually replaced by square houses and by another type of pottery which bears the name of Ubaid because it was first found in the 1920s during excavations of a small mound called al-‘Ubaid, in the vicinity of the celebrated Sumerian city of Ur.²⁶ This name is significant as it implies that for the first time in proto-history one single culture extended from the Jazirah (and even beyond) to the Tigris–Euphrates delta. The lack of rupture between the Halaf and Ubaid cultures excludes a conquest of northern and central Iraq by ‘Ubaidians’ coming from the south, and the most plausible hypotheses are a peaceful infiltration or the adoption by the ‘Halafians' of the culture of another population after a long period of contact.

That southern Iraq had been inhabited long before the middle of the fifth millennium was demonstrated in 1946 – 9 by the excavations conducted at Eridu (Abu Shahrain, nineteen kilometres to the south-west of Ur).²⁷ The ruins of Eridu are now marked by low mounds and sand dunes surrounding a much dilapidated ‘ziqqurat’, or stage-tower, erected by Amar-Sin, king of the Third Dynasty of Ur (2046– 2038 B.C.), but under one corner of the ziqqurat Seton Lloyd and Fuad Safar unearthed an impressive series of seventeen temples²⁸ built one above the other in proto-historic times. The lowest and earliest of these temples (levels XVII – XV) were small, one-roomed buildings which contained altars, offering tables and a fine quality pottery (Eridu ware) decorated with elaborate, often elegant geometric designs in dark-brown colour and presenting affinities with the Choga Mami transitional ware. The poorly preserved remains of temples XIV – XII yielded a slightly different ceramic characterized by its crowded designs and ‘reserve slip' decoration, which was identical with the pottery found in 1937 – 9 by German archaeologists at Qal‘at Hajji Muhammad, near Uruk.²⁹ This Hajji Muhammad ware, as it is called, is also present on other sites of southern Iraq, notably Ras el ‘Amiya, eight kilometres north of Kish,³⁰ where, it must be noted, fragments of walls, clay vessels and other objects lay buried (as indeed at Qal‘at Hajji Muhammad itself) under a few metres of alluvium and were discovered by chance. Finally, temples XI to VI, generally well preserved, contained numerous specimens of standard Ubaid ware, whilst temples VI – I could be dated to the early stages of the Uruk period. Since the Eridu, and Hajji Muhammad wares are closely related to the early and late Ubaid ware, these four types of pottery are now commonly called Ubaid 1, 2, 3 and 4 respectively.

More recently, a startling discovery was made at a site called Tell el-’Oueili by the French archaeologists who were digging at the nearby city of Larsa. Oueili is a relatively small mound partly above and partly below the present level of the surrounding plain, and it has the advantage of being entirely Ubaidian. Two deep soundings conducted in 1981 and 1983 respectively enabled the explorers to divide it into twenty levels of occupation.³¹ The uppermost levels (1 to 8) contained Ubaid 4, 3 and 2 pottery, and samples of Ubaid 1 (Eridu) ware were recovered from levels 8 to 11. But this was not the end, as would have been expected, for below these were no less than eight additional levels (12 to 19) which yielded a pottery (tentatively classified as Pre-Ubaid or Ubaid Zero) that was hitherto unknown but had affinities with the Samarra ware, while the cigar-shaped mud-bricks of a wall in level 12 were reminiscent of the bricks found at Tell es-Sawwan. Furthermore, below level 20 (in the water table and therefore unexplorable) other layers of occupation could vaguely be seen, and no one knows how far back into the sixth millennium the roots of this modest South Mesopotamian village go.

Clearly, then, large parts of southern Mesopotamia had been occupied long before the Ubaid period proper perhaps by people related to those Samarra folk who, it will be remembered, invented irrigation agriculture on the middle Tigris and in the Mandali area. Moreover, while a progressive architectural development can be followed throughout the superimposed temples of Eridu, there is no break in ceramic styles or techniques. The Ubaid ware – so the experts tell us – derives from the Hajji Muhammad ware, which derives from the Eridu ware, which in turn appears to derive from, or at least to share common ancestors with, the Samarra ware. Another, inescapable, conclusion to be drawn from the Eridu temples is that the same religious traditions were handed down from century to century on the same spot from about the middle of the sixth millennium B.C. until historical times, and from the relatively recent finding of two Ubaid shrines near to Anu's ‘White Temple’ at Uruk (see Chapter 5). Thus the more we dig, the more we find that the Sumerian civilization was very deeply rooted in the past.

Even easier to identify than the Halaf ware is the Ubaid ware, the hall-mark of that period, which is less sophisticated and much less attractive. The clay, frequently overfired, varies in colour from buff to green. The paint is matt, dark brown or bluish-black and the decoration restricted as a rule to only parts of the vessels. Although occasional plants, animals and broad sweeping curves are not without charm, the monotony of the common motifs (triangles, striped or cross-hatched bands, broken or wavy lines) betrays a lack of imagination. Yet the fabric is often fine, some specimens seem to have been made on a slow wheel or ‘tournette’, and spouts and loop-handles appear for the first time. Among the most characteristic forms are a bell-shaped bowl, a jar with basket-handle, a cream bowl with pouring lip and a lenticular vessel with a flat base and a long, tubular spout, called ‘tortoiseshell’. With a few exceptions (Kish, for example), this pottery was found on all sites of southern Iraq and on many sites of northern Mesopotamia, but there are marked differences between the north and the south in the other elements of the Ubaid cultural assemblage.

The words ‘clay and water’ would aptly qualify the Ubaid culture in southern Iraq. As stone is rare in that part of the country, its use was limited to heavy tools and a few ornaments. All other objects, including bent ‘nails’ (in fact, probably mullers), sickles, spindle-whorls, loom-weights, net-sinkers, sling pellets and even models of axes, adzes and knives, were made of terracotta. The erroneously called ‘Mother Goddess’ type of clay figurines – a slim, standing woman with a lizard-like head crowned by a coil of hair made of bitumen and whose ‘coffee-bean’ eyes recall those of Tell es-Sawwan and Choga Mami – was very popular, and there were figurines of men as well. A number of houses were frail structures of reed matting supported by wooden poles and sometimes plastered with clay, such as can be seen around Basrah today, but pressed mud or mud bricks were widely used for more comfortable buildings. The Ubaid period temples of Eridu were made of large mud bricks set in clay mortar and consisted of a long, oblong nave, or cella surrounded by small rooms projecting forward at the corners. At one end of the cella, against the wall, was a low podium which had once supported the statue of the god, while at the other end stood –a brick altar. The walls were adorned externally with shallow buttresses and niches that caught the light and broke the monotony of the plastered brickwork. Let us also note that these temples were raised on mud-brick platforms which tended to be increasingly larger and higher, foreshadowing the ziqqurats of later times.

Virtually unknown for many years, the secular architecture of the Ubaid period is now well illustrated in Lower Mesopotamia and in the Hamrin basin. In the upper, late Ubaid levels of Tell el'Oueili, for example, the French archaeologists have unearthed the remnants of several large and carefully constructed mud-brick houses separated from each other by open spaces. One of these houses was remarkable in that within and around it were dozens of small, square and shallow cavities between thin walls of tauf, which is rather puzzling but can probably be interpreted as infrastructures of granaries. The village was in a very flat region criss-crossed by streams and partly marshy. Its inhabitants grew barley, date-palms and other edible plants; they bred almost exclusively zebus and pigs which fed on aquatic plants, and they cut reeds for wickerwork with their baked clay sickles. The presence of obsidian and bitumen testifies to a certain amount of long-distance trade.³²

Half-way between the extreme south and the extreme north of Iraq lies the Hamrin basin where about twelve Ubaid settlements have been explored. Among them, Tell Madhhur, excavated by a British expedition in 1977 – 80, is of particular interest as it contained ‘one of the best preserved prehistoric buildings ever to have been found in Mesopotamia’.³³ This was a relatively small house built on the ‘tripartite plan’ characteristic of all the main buildings of the Ubaid period (temples included), with a central cruciform hall and smaller rooms on two sides. The walls were still more than six feet high, and the doors and windows remained perfectly visible. A ramp in one of the rooms suggested an upper floor, but this could not be proven. The house had been destroyed by fire, which preserved most of its contents, including pottery in situ and basically the same agricultural tools and household implements as elsewhere in those days, with thousands of clay sling-bullets, but no baked clay sickle.

If we now turn to the north, we are confronted by a somewhat different picture. Reed habitations are unknown and all buildings are made of bricks. Stone is commonly used and stone stamp-seals, very rare in the south, are here quite numerous; they bear linear designs, but also representations of animals and human beings arranged in what may perhaps be considered as mythical scenes or ritual dances. At Tepe Gawra,³⁴ the most important site of that period in northern Iraq, the three large temples with painted walls that form a grandiose ‘acropolis’ in level XIII closely match the Eridu temples, but two tholoi betray the persistence of regional traditions, as do the Halaf-style sitting and painted female figurines. More important, perhaps, the burial customs are very different from those of the south. At Eridu, in a large cemetery outside the settlement, adults and children alike are lying supine on a bed of potsherds in cist-graves lined and covered with mud bricks. At Gawra there is only one burial of that type; the other graves are simple pits grouped around the houses, and the bodies lie, flexed, on one side; children are buried in urns. This would suggest that the bearers of the Ubaid culture were in the minority in the north. Outnumbered but not eliminated, the descendants of the ‘Halafians’ still probably formed a large part of the population, whereas the south was entirely ‘Ubaidian’. In the next chapter we shall see how the gap between the north and the south gradually widened and how the south took the lead in the march towards civilization.

These differences, however striking, do not fundamentally alter the unity of the Ubaid culture. Whether imported from south-western Iran or, as is increasingly probable, developed locally, this culture – which lasted at least a thousand years – spread all over the cultivable areas of the Mesopotamian plain with the notable exception of the middle Euphrates and lower Tigris valleys. The highest chains of the Taurus and the Zagros mark its limits, but neither these mountains, nor the rivers, nor even the sea offered an insuperable barrier to commercial intercourse. The reality and extent of this trade is attested by the presence of obsidian on many sites of southern Iraq and of gold and amazonite (a semi-precious stone obtainable only from India) at Ur, as well as by the presence of unmistakable Ubaid pottery at Ras Shamra, on the Syrian coast, and on the Arabian shores of the Gulf.34 In southern Mesopotamia the Ubaid settlements were situated along the Euphrates and its branches and communicated with each other by water, as illustrated by the clay models of boats found at Eridu and Ur. Most of them were villages, but there were larger centres from which were to spring later on all the main cities of ancient Sumer. Another fact has yet wider implications: of all the buildings of the Ubaid period, the temple was always the largest and best constructed. It therefore looks as though the future Sumerian cities grew not around a palace or a castle, but around a shrine, and it is perhaps not unreasonable to think that the temple was already the hub around which most social and economic activities revolved. It would be bold at this early stage to speak of ‘Sumerians’, but there is strong reason to believe that the Ubaid period represents the first stage in the development of the Sumerian civilization.