Prehistoric India to 1000 B.C. c13: Chapter VI: The Time of Troubles and the End of the Cities

Prehistoric India to 1000 B.C. c13

Language: English

c13.s1 Notes to Chapter VI — 5 parts

Chapter VI: The Time of Troubles and the End of the Cities

c13.1

The Time of Troubles and the End of the Cities

c13.2

A host whose onslaught was like a hurricane, a people who had never known a city. (*Third Dynasty Year-date at Ur*)

THE dual pattern of prehistoric Western India at the beginning of the second millennium B.C., with the great urban civilization of Harappā in the plains and the simpler peasant communities in the Baluchistan hills, had endured almost unaltered for seven or eight hundred years at least; the Baluchi settlements themselves are likely to have been established as early as 3000 B.C., and some of them perhaps earlier. There is no evidence of violent change in the archaeological record of this time; villages are deserted, towns rebuilt after a flood, or new settlements are made, but the small, self-sufficient societies in the remote valleys of the hills or the vast organization of the Harappā empire in the plains continue their way of life unaffected by the outside world.

c13.4

But an intimation of what occurred to break this long seclusion of peaceful, if stagnant, communities is to be found in the stratification of the Rana Ghundai mound. Here there had been continuous occupation of the site accompanied by an evolving pottery style from a time likely to be late in the fourth millennium up to one more or less contemporary with some part of the Harappā civilization, the RG IIIc phase, perhaps about 2000 B.C. At this point there is evidence that the settlement was sacked and burnt – ‘everywhere overlying the foundation level ... [of the RG IIIc phase] there are pockets of ashes, as though some great conflagration had taken place’, wrote Brigadier Ross in his description of the stratification of the site. Above this burnt layer, in the RG IV phase, the pottery types show a complete break from those below - coarse bowls appear with painted ornament, ‘bold and not altogether unpleasing in a somewhat tawdry style’. And this settlement was once more destroyed by fire, and again in RG V a change in pottery shows new folk settling on the ancient, flame-scorched Rana Ghundai tell; the pottery is now unpainted, with relief patterns encrusted on to its surface.

c13.5

This evidence of violence and of rapidly-changing populations can mean only that the RG IV and RG V phases represent a time of troubles, of insecurity and raiding, arson and pillaging in North Baluchistan. To the Rana Ghundai evidence can be added that of Nal, where the last phase of the Zhob-ware settlement on the site was also burnt down, to such a degree that the tell is still called Sohr Damb, the Red Mound, from its fire-reddened soil. At Dabar Kot the upper 6 feet of the tell, exposed in section, showed no less than four thick ash layers, implying repeated destructive conflagrations of the later settlements, and here there occurred abundant fragments of the encrusted ware of the RG V type on the surface. At this site too one of the later settlements was that of the Harappā folk, so the sacking of the settlement is approximately dated.

c13.6

In South Baluchistan there seems no recorded evidence of similar burning in the later settlements, but it must be remembered that here little excavation has been done. At one site, however, a small tell at Shahi-tump, near Turbat, a cemetery had been made in the ruins of an abandoned village of the Kulli Culture, which can also be equated with the presence of Harappā folk in the region by means of a fragment of a clay toy cart of characteristic type. So the Shahi-tump cemetery is later than the Kulli Culture and some phase of Harappā. In the area excavated by Stein twelve inhumation burials were found, and an additional seven groups of grave-goods with no recognizable human remains.

c13.7

The inhumations where identified all appeared to be of complete bodies lying on their backs or sides with the legs flexed. Grave-goods in most instances consisted of groups of pots, but copper or bronze tools and ornaments, stone beads and alabaster cups also occurred. Two burials, Stein’s A and B, had richer grave-goods than the others. The skeleton discovered in burial A seems to have been that of a man buried crouched on his right side; in the grave there were more than fifty pottery bowls, cups and beakers, a copper stamp-seal, a stone pestle, a bead of ruby-like stone, a stone blade and a group of twelve marine shells (*Natica asellus*) against the skull. Burial B (certainly male) had again a wealth of pottery vessels stacked in ‘nests’ of up to five bowls, a necklace of lapis, agate and onyx beads, a copper spear-head and a heavy, shaft-hole copper axe. The presence of animal bones, some charred, in certain of the vessels accompanying these and other burials suggests that funeral bakemeats had been deposited with the corpse.

c13.8

The pottery from all the burials is consistent and characteristic (Fig. 25). The ware is very good, hard and thin, varying from grey to pinkish and sometimes having a clear, yellowish-buff colour. The commonest form is a shallow bowl, usually with a rather sophisticated foot-ring, and allied forms are deeper bowls which may attain a considerable size, or be represented by small cups of similar profile. There are two types of beakers, globular or conical, and characteristic ‘squat pots’ which merge into small bottles with small necks and feet and a bulbous body. Exceptional forms are footed bowls and a bowl on four legs, and a couple of triple vessels each of three conjoined units.

c13.9

The painted ornament is executed with a soft brush and in paint which varies from black through warm sepia to a reddish-brown, and a very characteristic feature is the soft blurred edge of the painted line, where the thin paint has ‘run’ into the porous background. The technique is slack,

c13.14

FIG. 25. Painted bowls and goblet (twice scale of bowls), Shahi-tump Cemetery

c13.15

often sloppy, and we look in vain for the crisp vigour of the Iranian wares which, as we shall see, are its nearest relatives. The ornament on the outside of the vessels is normally in zones, but sometimes in panels, and in the bowls the centre is either occupied by a swastika or allied motif, above which zonal patterns run, or the whole interior is divided into quarters within which circular elements are placed, or the so-called ‘Maltese square’ is employed. In the exterior zonal and panelled ornament swastikas and angular spirals are used in addition to the more frequent chevrons, lozenges and triangles. One vessel has two-colour chevrons in red and black.

c13.16

The nearest parallels to this pottery come from the cemetery of Khurab near Bampur, 150 miles to the west over the Persian frontier, the date of which appears to approximate to that of Kulli and Harappā. At Khurab the vessels were very similar to pots from the Mehi cemetery (belonging to the latest phase on that site, even possibly dating from after its abandonment). As a whole, the Khurab and other cemeteries near Bampur are likely to belong to the centuries around, or probably just after, 2000 B.C. But ultimately, the Shahi-tump and similar wares are descended from the pottery of the Buff-ware group best represented in Susa I and other sites in Fars, and, at greater remove, from that of Samarra. The shape of the vessels, the arrangement of the patterns, and much of the actual decorative repertoire remains relatively unaltered, but the degenerate patterns, the flaccid brush-work and the thin paint betray the fact that Shahi-tump ware is archaistic, not archaic, and enable us to accept with confidence the late date implied by its associated finds. It represents in fact the final expression of a tradition persisting in Southern Persia from perhaps the fifth millennium B.C. until the beginning of the second; to this tradition also belong the little conical alabaster vessels from the cemetery, unchanged in form and substance from their earlier representatives.

c13.17

Of the other objects from the cemetery, the copper spear and shaft-hole battle-axe from burial B indicate that this is the grave of a warrior (Fig. 26). The spear-head, 9 inches long, has no mid-rib and recalls Harappā types, but the stoutly-made axe, with a shaft-hole 1·2 inches in diameter, seems to derive from Sumerian and Akkadian prototypes, while its nearest parallels, significantly enough, come from the graves of chieftains of barbarian tribes beyond the fringe of the oriental urban civilizations at Maikop and Tsarskaya in South Russia. In the loosely-knit structure of semi-nomadic barbarian society, spread over the steppe and along the frontiers of the more settled communities in Mesopotamia or Persia, such types, ultimately originated by Sumerian metal-smiths, would have a wide and often rapid distribution.

c13.19

The copper stamp-seals from Shahi-tump, five of which were found, are even more explicit in their western relationships (Fig. 26). They are all of one distinctive type, and it is significant that the only other example from Baluchistan comes from the final phase of the Sohr Damb at Nal, and is perhaps to be associated with its destruction by burning. They are all circular, with a simple pattern on their faces built up in compartments by projecting strips of metal, and seals of generally similar type have been found in Hissar IIb and IIIb, in Anau III and at Susa, in a context which suggests a date about 2000 B.C. or a little later. And, as we shall see shortly, at least one similar seal is known from Sind, where it is later than the Harappā period. The evidence is therefore strongly in favour of regarding these metal objects in the Shahi-tump cemetery as indicating a date after 2000 B.C., and associated with movements from the west.

c13.20

Do these objects represent trade or migration? To anticipate the further evidence discussed below, the latter seems the more likely, and, since the pottery represents a local tradition, it seems likely too that migration was one of persons travelling light, bringing with them only the portable property of seals, spears, and battle-axes. Can these be war-bands, enslaving or marrying the local women, who would make the pots, and is the Shahi-tump cemetery in some way related to the sacking of the Baluchistan villages? Skulls are slender evidence and deductions from their measurements must be used with caution, but the anatomist who examined the skull of the Shahi-tump man buried with the spear and axe reported that it showed ‘traces of mixed origin and in certain respects tends to approximate to the Caspian or Nordic type of skull’. Unfortunate though the second adjective is in a modern context, it does at least seem probable that the Shahi-tump cemetery was that of newcomers to the region, rather than that of the old population, with a few novel types of seals and weapons acquired by trade.

c13.21

And in this connexion I would draw attention to the evidence mentioned in the last chapter; as we have seen, in the final phases of Mohenjo-daro there was a sudden strengthening of contact between the Harappā Culture and that of Kulli, and pottery and stone vessels of South Baluchistan types appeared in the later levels of the city’s occupation. Can this be wholly unconnected with the evidence for disturbed conditions in Baluchistan, and the establishment in the south of that region of newcomers from the west, suspect at least of being more than peaceful immigrant farmers? Were there refugees streaming down the passes of the Bolan, Lak Phusi and the Gaj Valley into the Indus plain, seeking shelter in the towns with which they long had traded? If this were indeed so, it is to be feared that they did not escape for long, for from Sind comes clear evidence that the mountains made no more than a temporary halting-place for the men from the west. Once over the Kirthar Range they could look down, upon prosperous communities, ripe for plunder, beside which the Baluchi villages were as nothing; they were on the frontier of civilization, but it was a kingdom unskilled in defence and unprepared for attack. Archaeology has shown something of the end of the Harappā towns at Chanhu-daro, Jhukar, and Lohumjo-daro in Sind, at each of which the invaders settled where they had plundered.

c13.22

The site of Jhukar has given its name to this characteristic intrusive culture in Sind, which appears at all three sites stratified above and in the ruins of deserted Harappā settlements, but it is at Chanhu-daro that the largest body of evidence occurs. In the ensuing description, therefore, attention will be directed mainly to that site, with amplification where necessary from the other finds. The stratification at Chanhu-daro can be expressed as follows, following normal practice in these matters:

c13.23

Chanhu-daro I*a* Chanhu-daro I*b* Chanhu-daro I*c* Chanhu-daro II Chanhu-daro III } Harappā Culture Jhukar Culture Jhangar Culture

c13.24

The final phase of occupation of this site will be described later, and for the present we are concerned with the Chanhu-daro II phase, a settlement of people of the Jhukar Culture.

c13.25

This settlement was one of ‘squatters’ among the Harappā ruins of the I*c* phase, and, while there seems to have been some rough rebuilding of house walls with brickbats robbed from other parts of the site, other huts seem to have been made of some perishable walling, such as matting, roughly set round a rectangular brick floor, which in one instance was some 17 by 8 feet overall. These brick floors might also be interpreted as stances for tents, and at all events they suggest rather squalid and not very permanent settlement on the site. Fireplaces are an interesting innovation, sometimes roughly built into doorways or niches of ruinous house walls or made of small rectangular or curved settings of bricks on edge.

c13.26

The material culture represented among these primitive habitations includes abundant and characteristic pottery; stamp-seals; amulets; beads; metal tools and pins; bone awls and a pottery head-rest (Fig. 27). It is altogether a curious and distinctive assemblage, contrasting strongly with that of the Harappā Culture it supplants, and yet not without clear contact with it. The pottery is a buff ware, with a varied repertoire of ornament painted on it in black and red. The forms include small-footed jars and bottles of types which have general parallels in Baluchistan, but the offering-stands and saucers with a looped pattern on the edge must derive either from Harappā or from the Kulli Culture in its Harappā-influenced phase. The range of ornament is quite varied, and employs geometric and stylized plant forms which in general have a strong resemblance to the trees on Kulli ware. Other simple geometric patterns accompanied with broad horizontal bands of plum red recall the Amri style; close-set chevrons in black and bright red can be paralleled precisely on a sherd from Zayak in South Baluchistan, where other patterns too can be matched - the multiple loops and ball-and-stem motifs of Jhukar, for instance, occur at Sar-Parom and Jai-damb. And one of the two animal representations at Chanhu-daro is really not unlike one from Periano-ghundai in North Baluchistan.

c13.27

In general, then, the Jhukar pottery from the Sind sites seems to combine a variety of elements, in which Kulli and probably Harappā motifs predominate, with an underlying Amri strain. It is, in fact, just what might be expected from a native, non-Harappā substratum in the local population, with a certain infusion of novel strains from South Baluchistan. Isolated motifs could be traced farther west - for example the plant form with double-spirals occurs in Sialk III and in Hissar Ib, but it then turns up again in the late cemetery of Jiwanri in the Makran referred to below. On the whole there seems no reason to regard Jhukar pottery as anything but a native product arising out of the disturbed conditions and folk-movements after the fall of the Harappā empire, when refugee tribes were leaving Baluchistan and settling in Sind.

c13.28

A most interesting contrast and comparison with Harappā is afforded by the Jhukar seals and allied amulets found abundantly in Chanhu-daro II (Fig. 27). The seals are made of pottery, faience, stone, and metal, and are totally different in their ornamental motifs from the well-known Harappā series; they also lack any inscriptions. But they are stamp-seals (usually circular, exceptionally square) within the great Western Asiatic family of such objects, and certain parallels can be noted. One of the faience seals has a pattern precisely comparable with that on the copper seals from the Shahi-tump cemetery; the treatment of animals on two others recalls Sumer and Elam rather than anything Indian; an interlaced coil pattern on another seal is best paralleled in Hittite contexts in Asia Minor, and this region produces many other less precise parallels for the types of seals and double-sided amulets in Chanhu-daro II. Whatever may be said of the pottery, the Jhukar seals are foreign to India, and imply newcomers from the west, not unconnected with those who buried their dead at Shahi-tump.

c13.30

The beads point in the same direction. One ornamented stone bead from Lohumjo-daro is paralleled at Hissar IIIc and Anau III, and the long, barrel-shaped faience beads from Chanhu-daro II, roughly ornamented to imitate segmented forms, recall similar segmented barrel-beads in Jemdet Nasr and Early Dynastic contexts in Sumer. The very large segmented beads from the same site are less easy to parallel, though in Egypt at least such clumsy forms come very late in the bead sequence. Again, we may infer the introduction of new types from the west, from the fringes of the kingdom of Sumer and Akkad.

c13.31

A copper shaft-hole axe provides another link with the Shahi-tump cemetery, and the series of pins offers a contrast to the Harappā customs of dress, which did not include garments held with this type of fastening (Fig. 27). There are seven pins with the head rolled over in a manner known in Sumer since Early Dynastic times at least, but this type significantly makes its appearance first in the IIIb phase of Hissar. Of the other types of pins, it is difficult to say much except that one is vaguely reminiscent of a type found in Sialk IV.

c13.32

The numerous bone awls were also an innovation, since none was found in the Harappā levels; Mackay suggested that they had been used for mat-weaving. These unsophisticated tools stress the barbaric character of the Jhukar Culture, which is already suggested by the hovels built among the ruins of the deserted town. The gaily-decorated pottery head-rest is a remarkable find, and it is difficult to suggest parallels, but once more the contrast with anything we know of the Harappā or the Baluchi Cultures is emphasized.

c13.33

The picture given by the Jhukar material is not dissimilar from that of the less abundant relics from the Shahi-tump cemetery. In both cultures the pottery appears to represent a more or less indigenous type in a late stage of evolution or devolution, but the smaller portable objects, such as seals, beads, metal implements, and weapons, and habits of dress implied by the use of pins, all suggest the arrival of new blood in the region. At Chanhu-daro one of the most interesting features is the use of faience by the Jhukar people, and it can hardly be explained in any way except by the supposition that the invaders conscripted, or at least were prepared to acquire the products of, local craftsmen skilled in this technique who survived the Harappā disasters and were set to work imitating as best they could the type of beads demanded by their new rulers. Although the newcomers may have known faience by sight, they are unlikely to have brought with them craftsmen who could make it, though as we have seen, some faience beads may have been imported into India from the Eastern Mediterranean region. At these three sites in Sind where the Jhukar Culture has been found, we can feel only that we are witnessing the arrival of barbarians, either as the destroyers of the Harappā civilization or following hard in the wake of the first raiders. It is their portable objects that are significant and which give the clue to their origins: here for once we can relegate pottery to a secondary place in estimating the content of these allied cultures.

c13.34

It is convenient here to say a few words about the Jhangar Culture, which at Chanhu-daro III appears as a final occupation of the site after the Jhukar settlement. Little enough is known of the culture, recognized only here and at the type-site, and also in Sind, except that it marks a break with the old painted pottery traditions, and has instead a grey or black ware with incised ornament. The pottery looks primitive, but its date is wholly unknown, and it may represent a purely local culture restricted to a small tribe or two, and may be not earlier than the first few centuries B.C. At Jhangar itself, apparently associated with the primitive ware, were sherds which look suspiciously like pottery of Sunga date, about the second or third century B.C.

c13.35

When we turn to Mohenjo-daro, the evidence for the final phases of the city's history is less abundant than at Chanhu-daro, and there does not seem to have been any actual settlement made on the ruins after the collapse of the Harappā Culture - at least, not in the areas excavated. But there are nevertheless several significant pointers to a disturbed state of affairs followed by disruption and decay, at this site as at others. I have already drawn attention to the hoards of jewellery and other precious objects found in the latest strata at Mohenjo-daro, and pointed out that the hiding of such hoards is a typical feature of disturbed and insecure conditions - there is a good early Indian parallel from the town of Rairh in Jaipur state, where numerous hoards of coins could be attributed to the troubled times following the collapse of the Mauryan dynasty early in the second century B.C. Many of the largest hoards of copper tools at Mohenjo-daro, too, belong to the later phases of the city, and emphasize the feeling of insecurity that must have been in people's minds.

c13.36

The actual buildings and their construction show a falling-off in these latter-day levels as well: street frontages are not so strictly observed, brickwork may be shoddy, and large houses are divided up into smaller rooms to take more persons with a lowered standard of amenity. Pottery kilns are allowed to be built within the city boundaries in the DK area, one even in the middle of a street - a shocking piece of decadence which could never have been condoned by the City Fathers of an earlier generation. Insecurity, the threat of raids, the big houses split up into flats, the slackening of civic pride and the decline of residential areas of the city into industrial slums, are all familiar enough signs of decadence and impending collapse. And that raids, if nothing more, were taking place at this time is shown by the huddled skeletons of persons murdered in the street or on a staircase - murdered under circumstances where no one was concerned with giving them burial, but rather anxious to save his own life by flight.

c13.37

There are a few pieces of significant archaeological evidence which can be related to this 'time of troubles' at

c13.38

FIG. 28. Copper shaft-hole axe-adze, Mohenjo-daro (after Mackay)

c13.39

Mohenjo-daro. A copper axe-adze with a shaft-tube for hafting was found in a late stratum, and is a completely exotic object among the metal types of the Harappā Culture (Fig. 28). But it is characteristic of a number of sites in North Persia - Hissar III, Shah Tepe, Turang Tepe - likely to be approximately contemporary with the Akkadian period in Mesopotamia, and surviving in other sites (e.g. at Assur and in the B cemetery at Sialk) probably as late as the ninth century B.C. It is an import from the west, and a weapon of war, likely on the whole to date from the second millennium B.C., and to be taken into consideration with the shaft-hole axes from Shahi-tump and Chanhu-daro II. There is also a group of what can only be called swords, up to 1½ feet long and with a strengthening mid-rib in a manner foreign to the Harappā tradition, found again in the later strata only, and which present something of a problem when one comes to look for parallels. The swords or daggers nearest in type to those from Mohenjo-daro, with a tang and rivets to hold the handle, come from Palestine and date between 1800 and 1500, where they are often associated with the Hyksos period, but Palestine is remote enough from India, and one would expect intermediate examples to have been recorded. However, they certainly seem intrusive at Mohenjo-daro, and likely to have been brought by invaders from somewhere outside the western frontiers of the Harappā kingdom.

c13.40

There is one burial at Mohenjo-daro, made in the ruins of a deserted courtyard of a house in the HR area, which seems to belong to the invaders rather than the invaded. Among a confused mass of fragments of a large pot, and other objects, such as small vases, beads, a shell spoon, etc., were a few bones and a human skull. One pot found with this group (the only one published) is of a type which we shall find recurring at Harappā in a cemetery on that site which is demonstrably later than the main Harappā occupation of the city, though its ornament rather recalls Kulli or Jhukar motifs. Surprisingly enough, the skull from this burial is that which clearly belongs to the Mongolian group – closely comparable, for instance, to the modern Naga type. If this is really a burial of one of the attackers of the Harappā civilization in its last days, it suggests a very mixed invading force, possibly including mercenaries. Can he have been a Gurkha?

c13.41

It had already been noted by Vats in his excavations at Harappā that the superficial deposits of the mound, which represent the ruins of the Citadel, contained sherds of pottery of a distinctive type, dissimilar from that of the Harappā occupation of the city, but identical with that which he found in an important cemetery lying beyond the limits of the city on the south - the ‘cemetery H’, demonstrably subsequent to the Harappā period. Stone has noted that Wheeler’s excavations produced several segmented faience beads of types closely comparable to that equated with the Minoan faience bead of *c.* 1600 B.C., in these superficial layers of the citadel. The summit of the Citadel Mound is much disturbed by brick-robbing, and Vats did not establish an occupation layer to which this pottery could be assigned, but Wheeler in 1946 was able to show that his fourth (and final) period of construction at the West Gate and its associated terraces was in fact characterized by just this peculiar pottery. This last phase ‘followed an interval during which the structures relating to the terraces had fallen into decay and had been covered by debris. It is represented by fragments of poorly constructed buildings, presumably dwellings, with walls sometimes only one brick in thickness’.

c13.42

It seems likely that some of the structural remains found in the earlier excavations on the Citadel Mound in the uppermost strata belong partly to a decadent version of the Harappā occupation, with some mingling of the intrusive cemetery H elements. The typical cemetery H pot from Stratum II, recorded by Vats, would suggest that to this phase should be attributed the badly-built walls, made of re-used brick-bats, which he notes as characteristic of this stratum. Among these structures may be noted a circular bin or granary, still standing 6 feet high and 4 feet in diameter, and areas of brick paving, roughly square, which recall those of the Jhukar Culture in Chanhu-daro II. The similar pavements associated with the charred remains of pine and bamboo in what is alleged to be a Harappā stratum elsewhere on the Citadel Mound are suspiciously suggestive of the huts of barbarian squatters rather than of any structures likely to be within the Citadel walls in the days of its splendour, but the excavation records do not contain enough evidence for us to express a definite opinion. It should be noted in passing that what appears to be a ‘fractional’ burial was found in this same area, and was assigned to the Harappā Culture. If it is not a more or less modern grave, it is much more likely to be subsequent to the Harappā occupation than contemporary with it; the brachycephalic skulls of Alpine type with high-bridged noses in the grave at all events imply an intrusive ethnic element, and may conceivably be connected with invaders in the last phase of the site’s prehistory.

c13.43

Unquestionable evidence of the arrival of newcomers at Harappā after the main phase of occupation of the city is, however, afforded by the large cemetery H, itself of two periods, both of which have been shown by the recent excavations to be later than the R 37 cemetery, of Harappā age. This latter cemetery was situated in a natural hollow, which, at some time after the disuse of the cemetery, was filled up level with debris which contained a mass of Harappā potsherds, and it was into this debris, at a point some 120 yards to the north-east of cemetery R 37, that the graves of cemetery H were dug. It is characteristic of the conservatism of the Harappā Culture that no significant change in the types of pottery could be detected between the layers underlying the R 37 cemetery, the cemetery itself, or either of the two debris-layers overlying it. But with cemetery H the change is abrupt and decisive.

c13.44

The earlier burials in cemetery H (Stratum II) are extended inhumations, normally lying north-east and south-west, with the legs slightly flexed, and at an average depth of six feet beneath the present surface. About twenty-four such burials were found in the area excavated: some were claimed as dismembered or fractional, but it is not clear whether these were not sometimes the result of ancient disturbance. The burials are sometimes accompanied by the remains of food-offerings - in one grave an entire dismembered goat had been laid with the dead man - and large numbers of pottery vessels were always present. In only two graves were any other objects found: a woman with a gold bangle on her wrist, and the three remaining teeth in the skull of another looped round with gold wire - 'for security or decoration', as Wheeler rather sardonically remarks.

c13.45

The later burials in cemetery H (Stratum I) were only some two to three feet below the present surface, and consisted of true fractional burials in large pots, without any accompanying grave-goods or offerings. The bones in many instances showed that the bodies must have been exposed after death for some time before burial, when only the skull and a few long bones were selected for deposition in the cemetery. About 140 such burials were found, and a dozen of these were of babies, buried whole and crouched up in the funeral jars. These jars were closed at the mouth by lids of broken pots. The pottery from both strata of the cemetery is dissimilar from that of the Harappā Culture, and although certain characteristic forms are confined to the upper and lower strata respectively, this distinction is functional rather than cultural, and the technique of ware and painting is essentially the same in both phases of the cemetery. We can then regard it as a whole, and as the product of one culture despite the change in burial customs.

c13.46

Before proceeding to discuss this pottery, however, it will be convenient to summarize the very scanty knowledge available of the human types represented. No adequate report has yet appeared, but the dominant skull-form in the earlier phase seems to have been the proto-Australoid type, and there is at least one such skull among the second phase burials. But in these burials there can be traced a 'small, low-headed race, such as is seen among the present aboriginal population of India', and probably (though not certainly) from the second phase of the cemetery are two skulls considered by Friederichs and Muller to be Armenoid. On the whole, however, there seems little to suggest an intrusive ethnic type which could plausibly be connected with migration from the west.

c13.47

The pottery from both strata is very well made, with a hard red fabric and a brilliant red slip, and the frequent ornament painted on in black has often the same ‘running’ at the edges already commented on in the otherwise totally dissimilar ware from Shahi-tump. Apart from Harappā, this ware is known only from two sites in Bahawalpur State, 150 miles away to the south, on the other side of the Sutlej River. Although its technique in detail is not that of Harappā ware, the use of a black-on-red decorative scheme may imply some sort of relationship, at least within the Red-ware province as a whole, while one of the characteristic forms of the earlier phase of the cemetery is a squat ‘offering-dish’, which itself might be compared with the higher Harappā types. But the decoration takes us into a world entirely removed from that of Harappā.

c13.48

The technique of painting is as assured and sophisticated as the forms of the vessels, and implies the use of a kiln at least as advanced as those known from the late period of Mohenjo-daro. Characteristic motifs are stars of various kinds, stylized plant forms, ring-and-dot patterns, groupings of straight and crinkled lines, and short lengths of zig-zag lines used to fill up backgrounds. And in addition to these more formal elements of design, there occur frequent representations of cattle, goats, peacocks and fishes, treated in a brilliantly decorative convention; sometimes in continuous scenes which encircle the vase and sometimes in panels or in the roundels afforded by the circular undersides of pot-covers (Fig. 29). These representations are not naturalistic, but are instinct with symbolism – a series of peacocks, between groups of star or sun motifs, have within themselves a circle containing the tiny figure of a man; on another vessel peacocks alternate with a sort of centaur with the horns of a goat or deer. One remarkable vase from the upper stratum has a continuous frieze of design beginning with two peacocks and followed by a group in which a man with wild, flying hair holds two cows, behind one of which is a dog. Then comes a band of pattern, to be followed by an extraordinary version of a humped bull whose enormous horns bear seven enigmatic objects like little standards, and then another man between two cows (or bulls), each with a single standard between its horns, for all the world recalling a Minoan bull with double-axe. The whole scene is filled in with stars, leaves, and zig-zags, and similar zig-zags crowd the background in another vessel, where humped cattle stand between stylized trees.

c13.49

FIG. 29. Designs on pots from Cemetery H, Harappā (black on red). (After Vats)

c13.50

It is very difficult to suggest even partial analogues to this remarkable series of pottery. As a whole, it is quite without parallel in any of the known cultures of the Ancient Orient, and one can only say that the groupings of animals and trees does recall the ‘animals and landscape’ designs of Kulli ware in rather more than general terms and the star and bird motifs in association recall similar patterns on pots from the Giyan II cemetery of c. 1550–1200 B.C. The treatment of the animals and men is quite different from that known in any of the prehistoric wares of Baluchistan or India, but although remote in time and space, it is worth mentioning that the treatment of the human figures, and to some extent that of some of the animals is not altogether dissimilar to that on Samarra ware and its analogues in Persia. It is just possible that these similarities may represent at least an ancient Iranian or North Mesopotamian strain in the culture responsible for the burials in cemetery H. Beyond this, we can say nothing, though again the possible relationships with the Kulli style may point to Baluchistan elements as well. The cemetery H people are certainly newcomers to the Harappā world, but where they came from or what was their later history only further excavation can tell us.

c13.51

There remain a few scattered pieces of information to be brought together, not directly connected with the Harappā civilization and its end, but not without importance. A stray find from Kurram, in the North-West Frontier Province, is a curious type of flat copper axe with lateral lugs – the so-called ‘trunnion-celt’, that has a wide distribution in time and space in Europe, but which in Western Asia appears to be represented only by a specimen from Turang Tepe in Northern Persia and that from Kurram. The Turang Tèpe axe would be contemporary with Hissar III-Anau III, probably early in the second millennium B.C., and the Kurram find may again imply arrivals from the west at a somewhat later date.

c13.52

FIG. 30. Bronze Sword from Rajanpur, Punjab

c13.53

Another find suggesting western contacts of rather later date is that of a bronze sword from Rajanpur in the Punjab, a remarkable weapon with an elaborate hilt of a type well known in the Luristan graves of Persia, which are likely to be about 1400-1200 B.C., and similar to others from the Caucasus of approximately the same date (Fig. 30). This sword stands alone in India, though there are two hoards of swords of different types, one from Fathgarh in the Upper Ganges Valley and the other from Kullur in Hyderabad State, Deccan. The Fathgarh hoard contained no less than thirteen swords, said to be of copper, and with slight mid-ribs; all but one had characteristic hilts cast in one with the blade and spreading out into ‘antennae,’ and this was the type of the four swords from Kullur. Swords with hilts of the Rajanpur type occur in the Late Bronze Age Talish graves of the Caucasus, again about 1350-1200 B.C. and representing a school of metal-work closely in touch with that of Luristan. It is then just possible that all the swords in question may be related to Caucaso-Persian styles of the late second millennium B.C.

c13.54

In the Ganges Basin and on the adjacent uplands to the south-east a number of finds of copper tools have been made, either singly or in hoards of up to 400 objects (as at Gungeria, Balaghat). The finds fall into two groups, one in the valleys of the Ganges and the Jumna rivers above their point of junction and the other centred round the Ranchi uplands, and the characteristic tool present throughout is the flat copper or bronze axe, of simple forms all similar in general terms to the flat axes of the Harappā Culture. But the very simplicity of the forms make close affiliation difficult, and we can turn with more confidence to another type recurring in these finds, the elongated axe or chisel which has been called a ‘bar-celt’. Although precise parallels to these tools do not occur in the Harappā metal series, nevertheless at Chanhu-daro in particular very narrow, elongated axes appear, which could well be ancestral, and in the important series of copper tools from the Nal cemetery in Baluchistan a small version of the ‘bar-celt’ does in fact occur. As we have seen, the date of the Nal cemetery is likely to be approximately contemporary with the Harappā Culture, but it may in part survive later. It is a long way from Nal to the nearest ‘bar-celt’ find (Rajpur on the Upper Ganges), but in general the whole Gangetic axe and ‘bar-celt’ series is unlikely to be a separate and independent evolution from that of Harappā metallurgy.

c13.55

Associated with these axe types in several hoards, confined to the valleys of the Jumna and Ganges, is a remarkable series of copper harpoons, which can be arranged in a typological series evolving from primitive types which clearly copy prototypes in bone or horn. These harpoons must represent a local development among river-side communities whose economy centred round fish-spearing and whose primitive equipment was transformed into metal when the knowledge of metallurgy spread among them. The implied existence of harpoons of forms similar to those known in many mesolithic cultures of Europe and Western Asia is of some interest, and has been commented on in Chapter II.

c13.56

The date of these hoards of axes and harpoons cannot be fixed precisely. The axes and ‘bar-celts’ probably owe their origin to types within the Harappā or allied cultures in Western India, unless we assume an independent invention of metallurgy in another centre or centres of prehistoric India, and one’s instinct is therefore to date the finds within the period of that culture or in one shortly after its end. The concentration round Ranchi suggests the exploitation of the local copper deposits there, and the distribution of the finds westwards to Rajpur, 300 miles east of Harappā, does strengthen the idea that there was a spread of metallurgical techniques, and of the people responsible for them, from the confines of the Harappā kingdom eastwards. It would be tempting to associate this movement with something more than trade, and to see in it the colonization of the Ganges Basin by refugees and displaced persons from the Punjab and the Indus Valley during the time of the break-up of the Harappā empire and the coming of the raiders from the west. The deposition of hoards itself suggests a time of insecurity and economic instability, and may mean that the refugees were not left undisturbed for long, as the invasions gathered momentum and pressed on, beyond the old frontiers of the Harappā kingdom and down into the Ganges Valley. But here we leave archaeology for the ambiguous hints of legend and tradition.

c13.57

In sum, however, the evidence from Baluchistan and from Sind and the Punjab is reasonably consistent in implying that at some period likely to have been before 1500 B.C. (to use a convenient round figure) the long-established cultural traditions of North-Western India were rudely and ruthlessly interrupted by the arrival of new people from the west. The burning of the Baluchi villages and the equipment of the graves at Shahi-tump suggest that these new arrivals were predominantly conquerors who travelled light, and adopted the pottery traditions of the regions in which they established themselves. In Sind, at Chanhu-daro, a barbarian settlement appears in the deserted ruins of the Harappā town, and here some local craftsmen may have remained to work for alien masters, while the pottery suggests a resurgence of local, non-Harappā elements. At Mohenjo-daro it seems clear that the civilization that had survived so long was already effete and on the wane when the raiders came, and at Harappā we know from the evidence of the rebuilding of the Citadel walls that the inhabitants were on the defensive in the last days of the city, though these precautionary measures did not suffice to keep away the intruders, wherever they came from, who afterwards settled on the ruins and buried their dead in cemetery H for generations. From the Ganges Valley comes evidence suggestive of a spread of techniques and peoples eastward from the old Harappā kingdom as danger threatened from the west.

c13.58

This movement of peoples, this sack of the ancient cities by the outer barbarians, is something not confined to India in the centuries immediately after 2000 B.C. In Mesopotamia the kingdom of Sargon of Akkad rapidly disintegrated after the death of his son Naram-sin, when the Guti and other tribes broke in upon the land. Though there was recovery and considerable prosperity, especially in the south, the barbarian attacks were renewed a couple of centuries later, when the contemptuous if frightened scribe wrote in a year-date of the Third Dynasty of Ur of the Amurru, ‘a host whose onslaught was like a hurricane, a people who had never known a city’. It is the period that sees the rise of the Hittite Empire in Asia Minor, while archaeological evidence of folk-movements is clear in Syria and again in North Persia, where in the Hissar III phase the elements alien to the old painted pottery cultures, already beginning to show themselves in Hissar II, dominate the town. In this phase, as in the contemporary sites at Turang Tepe and Shah Tepe, we see the appearance of novel types of tools and weapons and ornaments based on those of Early Dynastic and Akkadian times in Mesopotamia, and now for the first time spreading extensively among the barbarian tribes who had broken down the frontiers of the civilized communities, carrying off loot, a knowledge of new techniques, and doubtless actual craftsmen to their own chieftains' courts.

c13.59

From the Caspian this movement of war-bands and migrating people can be traced as far eastwards as Anau in Russian Turkestan, where the third phase of the site's occupation shows clear traces of contact both with Hissar III and, less strongly, with Harappā. It is in this context of folk migrations around 2000 B.C., and the subsequent few centuries, that we can set the end of the Baluchi villages and the Harappā cities, but there is evidence that a second wave of conquest or colonization from the west left traces in Baluchistan nearly a thousand years later.

c13.60

At both Giyan and Sialk the latest settlements and cemeteries on the sites represent a semi-barbarian culture that numbered amongst its achievements, however, the domestication of the horse for riding and probably chariotry as well. Typical pottery and bronze arrowheads from the tell of Nad-i-Ali in Afghan Sistan show that these people had reached the borders of India, and at Moghul-ghundai in North Baluchistan a small cemetery of cairn-burials produced a whole array of bronze objects - rings, bracelets, horse-bells, a bangle, and a tripod jar - typical of the B cemetery at Sialk. In South Baluchistan the cemeteries of Jiwanri and Zangian have pots of distinctive shapes related to those from the same Persian cemetery, and from Londo not far from Nal come painted sherds in a provincial version of the Sialk B style with animal ornament, including a frieze of horses. The Sialk B cemetery probably dates from about 1250-1100 B.C., so that a date around 1100-1000 B.C. would be likely for the Baluchistan cemeteries just described.

c13.61

All these migrations of peoples come at a time when there is a breaking-up of the ancient Western Asiatic civilizations and the introduction into the old strongholds of civility, of new blood, new ideas, and even perhaps new techniques, such as the use of the horse-drawn chariot in warfare. From the mixture of old and new civilizations of an altered texture and fabric were in many places to arise, with the dissonant characteristics toned down in a mutual blending over generations. The barbarians imperceptibly assimilated the traditions of the civilization they had despised yet prized; they had conquered only to be themselves conquered at last by its persuasive and subtle power. The attempt in a Dark Age to re-create what the age itself has destroyed is not confined to early medieval Europe.

c13.62

This chapter has been restricted to the evidence of prehistoric archaeology for the movements affecting India in the middle of the second millennium B.C. But there is another part of the story to be told, based on the evidence of philology and on scraps of literary evidence: dangerous ground, full of quicksands and pitfalls which have too often trapped the unwary, and not infrequently the would-be wary too. The question of the dispersal of the speakers of languages within the Indo-European group, and of the arrival of the Aryan immigrants in Northern India, must be considered in connexion with the archaeological evidence.

Notes to Chapter VI

c13.s1.1

The Rana-ghundai sequence is described in *Journ. Near Eastern Studies* v (1946), 284-316, the burning of the Nal settlement by Hargreaves, *Excavations in Baluchistan 1925* (1929), and that of Dabar Kot by Stein in *Arch. Tour in Waziristan and North Baluchistan* (1929), 57. The Shahi-tump cemetery is also described by Stein, *Arch. Tour in Gedrosia*, 88–105, and that of Khurab (Bampur) in *Arch. Reconnaissances* 106–110. I discussed the Shahi-tump stamp seals in *Antiquity* XVII (1943), 179–181; the skull from the cemetery was described by Sewell in *Arch. Tour in Gedrosia*, 191–200.

c13.s1.2

The Jhukar Culture occupation at Chanhu-daro is described by Mackay, *Chanhu-daro Excavations* (1943); for the pins, see my note in *Ancient India*, no. 3, 145. Jhukar and Lohumjo-daro were described by Majumdar in *Explorations in Sind* (1934), where Janghar is also published, though the late-looking pottery (in the Central Asian Museum) is not commented on. The late phases of Mohenjo-daro and Harappā are described in the relevant excavation reports quoted in Chapter V, and the final occupation of the Harappā citadel by Wheeler in *Ancient India*, no. 3, 59–130. In the same paper (pp. 84–86) the H cemetery is discussed, supplementing the original report by Vats in *Excavations at Harappa*.

c13.s1.3

The copper and bronze implements from the Ganges Basin are discussed by me in *Antiquity* XVIII (1944), 173–182, where the Kurram axe is wrongly assigned to the Ranchi district instead of to the N.W.F.P. The Rajanpur sword is in the National Museum of Antiquities, Edinburgh. For the Moghul-ghundai cairns see Gordon in *Man in India* XXVII (1947), 234; the Londo site was discovered by Miss B. de Cardi (see p. 131, above).

c13.s1.4

c13.s1.5

Chronological Table II

Prehistoric India to 1000 B.C. c13: Chapter VI: The Time of Troubles and the End of the Cities

Contents

Chapter VI: The Time of Troubles and the End of the Cities

Notes to Chapter VI

Linked texts (19)