Early India and Pakistan to Ashoka c6: Chapter II. Time
Early India and Pakistan to Ashoka c6
- Language
- English
Contents
- c6.s1 Further Note on Three Wares of Special Chronological Importance — 31 parts
- c6.s2 Northern India — 20 parts
- c6.s3 South India — 40 parts
Chapter II. Time
c6.2
LIF E ON THIS PLANET originated, we are told, over two thousand million years ago. Somewhere within the last of all those millions, creatures in whom we can recognize a simulacrum of ourselves, with emergent ideas and capacities, came in their turn upon the scene and hesitantly assumed command of the world about them: so hesitantly that, as it seems, they became economically creative—and to that extent independent—only a dozen millennia ago, and civilized in any comprehensive sense only some five or six millennia ago. Advancing science will move these ‘dates’ a little this way or that but is unlikely now to revolutionize them. The general framework of human endeavour is set.
c6.3
But within that framework there is infinite scope for subtlety and differential change. The significance and the circumstance of man's major—not to say minor—triumphs are of the essence of the story, and these involve a close reading and re-reading of the evidence as it slowly accumulates. Today in India this accumulation is proceeding at a faster pace than in many other parts of the world, and precise interpretation of the new evidence in relation to the old is proportionately necessary. Without an exact timetable, the tremendous achievement which our evidence reflects cannot be brought into sharp and significant perspective; its phases cannot be interrelated, nor can its position and meaning in the wider context of human effort be established with assurance. Relative dates and sequences are useful, but only *actual* dates, in terms of years from the present, are of substantive value. Since therefore most of our subject-matter relates to prehistory, an introductory note on chronological method is inevitable.
c6.4
The simplest means of determining an archaeological date is naturally by the direct or indirect application of historical criteria. In India, it may be claimed that the earliest literature, the hymnal of the *Rigveda*, has something of the quality of history, since it portrays incidentally the earlier struggles and rivalries of the Aryan-speaking invaders of the north-western territories of the subcontinent, and gives a vivid picture of their social condition at the time. What precisely that time was, is less explicit. An Indian writer, enlarging a remark of Megasthenes by means of supposititious king-lists, would ascribe the Aryan Invasion to 6777 B.C., a fantasy which may be consigned to limbo with Archbishop Ussher's date for the Creation. More rational inference assigns the episode to the centuries following 1500 B.C., when, if not earlier, Aryan gods and men were on the move in western Asia. Of this movement it is reasonable to regard the invasions reflected in the *Rigveda* as an eastward extension; and their possible bearing upon the date and character of the latter end of the great Indus Civilization will be considered later.
c6.5
Further back than that, the literature of India fails to carry us. But Mesopotamia is happier in its early chronology; there a widely agreed central date, that of Sargon of Akkad, who, it has been calculated, was reigning about 2350 B.C., is of special use to us. In and about his time there were recognizable contacts between the cities of Mesopotamia and those of the Indus Civilization, to which the Mesopotamian dating can to that extent be transferred.
c6.6
Behind the latter half of the 3rd millennium there is at present no absolute moment or epoch upon which Indian archaeology can depend. Comparisons have been suggested between certain of the ceramic industries of the Indian borderland and those of Iran and Iraq, but without any pretence of certainty in a chronological sense. Differences are as plentiful as resemblances, and the duration of a local or even a regional industry is in any case incomputable without more collateral evidence than is at present forthcoming. For example, a distinctive ware to which the name of 'Quetta' has been given may, on typological analysis, be regarded as early but on stratigraphical analysis—such little as is available—may equally well be relatively late. For the time being, typology is indeed a fickle guide to the time-sequence of the border cultures of the 4th and 3rd millennia.
c6.7
Nor, in relation to the earliest stone industries of India, has geology yet produced that measure of precision which, in Europe, is gradually emerging from modern geochronological method: for example, by equating ancient climatic fluctuations, deduced from the examination of loess-deposits and the analysis of river-gravels, with fluctuations of the intensity of solar radiation calculated on an astronomical basis; or by computing the rates of weathering, denudation and sedimentation in the light of such evidence as that provided by varved clays or radiocarbon analysis.³ A cross-reference of some of the European to some of the Indian evidence might be attempted if we could with any confidence equate the Himālayan glaciations with those of the Alps. But such equations as have in fact been postulated, however likely, remain guesswork; and even within the subcontinent itself the relation of the pluvials of the tropics to the glaciations of the north is still full of speculation. We are still adrift in time.
c6.8
Unfortunately, one potential test which can, and will, help to regulate conjecture has not yet, except for occasional pioneer attempts,⁴ been extended to the subcontinent. That is the now well-known method, introduced in 1949, of calculating the residual radioactive content of ancient organic material and thereby deducing the time-interval since the 'death' of that material, a method applicable under suitable conditions and controls to periods up to forty thousand years from the present.⁵ This new technique, though still in the experimental stage, is in process of revolutionizing our understanding of the tempo and interrelationship of human achievement elsewhere, and its further application to Indian material by the National Physical Laboratory of India, which is understood to have the matter in hand, is an outstanding need.
c6.9
In the latter half of the 1st millennium B.C., however, history and archaeology at last become fully effective partners in this matter of timing, without the intervention of physical science. A datum-line is provided by the partially historic life of the Buddha, who, for our purpose, represents that miraculous upsurge of reflective and creative thought which swept across the world from Greece to China about 500 B.C. Sharing in this upsurge, Achaemenid Persia, with its systematized imperialism and its rich artistry, reached out to the Indus region about 518 B.C.* At this time, too, in the Ganges valley native kingdoms were already laying the foundations of the great literate Indian empire that was to emerge two centuries later. To them, Persia transmitted, directly or indirectly, not merely a pattern of empire but also important new skills and utilities: above all the use of iron and of coinage. It can be no mere coincidence that in northern India (as indeed in distant Britain) 500 B.C. or a little earlier marks the nominal beginning of the Iron Age, always with the proviso that cultural backwaters are a familiar feature of the Indian landscape. Nor is it likely to be chance that then or shortly afterwards the first Indian coinage appears in the Ganges valley as a new aid and incentive to organized commerce. And home-enterprise was in no way lacking. Following the arrival of iron—perhaps, indeed, literally reflecting the aspect of the new and fashionable metal—the Ganges craftsmen invented a distinctive and remarkable ceramic which is a godsend to the archaeologist: the so-called 'Northern Black Polished' or 'N.B.P.' Ware, of steel-like quality, which will constantly recur in later chapters. This pottery spread outwards from the Ganges Basin to the North-West Frontier, Bengal and the Deccan; its nuclear date seems to be 5th-2nd centuries B.C. (see p. 30). It followed an equally distinctive ‘Painted Grey’ or ‘P.G.’ Ware, often of comparable quality but less securely dated within the earlier half of the 1st millennium; 8th-5th centuries B.C. is a provisional bracket (p. 26). Of these and other crafts and industries more will be said; here they are mentioned as examples of an accumulation of data to which an approximate timetable can be attached.
c6.10
∗ Gandhara (the Peshawar plain and its environs) was already in Persian hands when the Behistun inscription was cut for Darius, c. 520-518 B.C. It had either been acquired by Darius at the beginning of his reign or, more probably, by his grandfather Cyrus. Darius now extended his realm to the Indus valley, but his eastern frontier is not defined.
c6.11
And then came Alexander the Great. To describe his advent in the Punjab in 326 B.C.—with the loot of Asia in his haversack and his entourage of historians and philosophers—as an ‘archaeological datum’ is scarcely to add to the stature of the liveliest of the Nine Worthies. But the fact remains that he opened the way to an inrush of refugee craftsmanship from prostrate Persia and was thus indirectly responsible for lasting elements in Indian sculpture and architecture. The Lions of Sārnāth, the badge of the modern Republic, are in this sense his legacy no less than Ashoka’s. And these lions and their kin are a part of the grammar of Mauryan and Shunga archaeology, as are the gems and ornaments which can be related to the same cultural tradition.
c6.12
It was however in the first centuries B.C. and A.D. that the classical West contributed most emphatically to the repertory of the Indian archaeologist. The consolidation of the Roman Empire by Augustus and his successors brought Roman coinage and Graeco-Roman wares in great quantities to the Indian peninsula, where spices, gems and a middleman-trade with China were the main objectives. Backed by a considerable and circumstantial literature—Greek, Roman, Indian, Chinese—these imports of known date and origin have been identified archaeologically and have lent their precision to associated Indian cultures. It is no exaggeration to say that in recent years they have provided a fresh starting-point for the study of central and southern Indian materials of which the context was otherwise unknown. In particular the widespread 'Rouletted Ware' (p. 31) is due to them, and its association with Italic (Arretine) pottery and coins of the Roman principate gives it a valuable initial date in the first half of the 1st century A.D.
c6.13
Such are some of the means whereby a great and growing mass of material of the prehistoric and protohistoric ages is now being sorted out and arranged in something approaching an absolute time-sequence. Under the stimulus of the Indian and Pakistani State Departments of Archaeology, aided by a few of the universities (notably Calcutta, Poona, Allahābād and Baroda) the work proceeds apace. There is far to go.
Further Note on Three Wares of Special Chronological Importance
c6.s1.1
Fig. 1
c6.s1.2
1. *Painted Grey Ware (P.G. Ware).* This distinctive Bronze Age ware is of fine well-levigated clay, wheel-turned, thin and grey in section, well-fired, and grey to greyish brown, sometimes dark, in surface colour. On the surface are painted linear and dotted patterns, concentric circles, spirals, *signas* and *swastikas*, generally in matt black but occasionally red.
c6.s1.3
The commonest forms are straight-sided bowls, and dishes with upright convex sides and sagger-bases.
c6.s1.4
The ware is characteristic of the Jumna-Ganges *doāb* but occurs as far west as Bikaner and as far south as Ujjain.
c6.s1.5
At Ahichchhatrā in U.P., where the ware was first recognized, it occurred mainly below the Northern Black Polished or N.B.P. Ware (see below), though there was some likelihood of an overlap between the two. At the more carefully excavated Hastināpura on the Ganges, however, there was a stratigraphical gap between the P.G. Ware and the overlying N.B.P. Ware, and there is a hint of a similar gap at Kaushāmbi on the Jumna, though the evidence there is less abundant. Other sites confirm the general sequence of the two wares. Moreover, whilst the N.B.P. Ware is associated with iron implements, the P.G. Ware is found exclusively with copper or bronze: even the stone industries which often accompany the use of copper in India seem to be absent from P.G. Ware levels.
c6.s1.6
Fig. 1 Painted grey ware: 1-4, from Ahichchhatrā; 5-9, from Pānipat; 10-12, from Hastināpura. ‡
c6.s1.7
If the initial date of the N.B.P. Ware in the Ganges region be placed somewhere in the 5th century B.C. (see below), then the bulk of the P.G. Ware is certainly earlier than that date. At Hastināpura this ware extended downwards through 4-7 feet of occupational strata; what that represents in time can only be conjectured, but two or three centuries are a feasible guess. On this basis, and at this representative site, the beginning of the P.G. Ware might be ascribed to the 8th century B.C. The skilful excavator of Hastināpura would prefer to take it back to 'circa 1100 B.C., with a probable margin on the earlier side'. This estimate is influenced by the fact that the P.G. Ware is the earliest reputable pottery known to us from a number of town-sites—Hastināpura amongst them—mentioned as already established before the Mahābhārata war, for which F. E. Pargiter suggested a date of c. 950 B.C. The equation may be right, but the available archaeological evidence scarcely carries so long a chronology. In any case, the date of the historical elements of the Mahābhārata is highly speculative.
c6.s1.8
At the upper limit of the dating, a clear archaeological gap has been established between the end of the Indus Civilization (c. 1500 B.C.?) and the beginning of the P.G. Ware at Rupar and other sites in N.W. India. Nevertheless, if as archaeologists we must be for ever tracing the steam-engine back to the tea-kettle, we may postulate a link, however tenuous, with Baluchistan, where the bowls from the secondary burials at Shāhi Tump seem to have a certain ancestral affinity with the Ganges material; and the Shāhi Tump burials post-dated the Indus Civilization at that spot. Certainly the sudden appearance of the high-class P.G. fabric in the Ganges basin suggests, or even proves, that its technique had already been perfected elsewhere. If Aryans must be dragged into this picture, it is possible to suppose that the P.G. Ware may represent the second phase of their invasion of India, when, from the Punjab, they entered and Aryanized the Middle Country of the Ganges-Jumna doāb, after picking up ideas and doubtless craftsmen in the Indus valley and the Baluch borderland. But see p. 126.
c6.s1.9
Northern black polished ware: 1, 3, 5, 8, 10, 17, from Rupar; 2, 16, from Rājghāt; 4, 9, 12, 18, from Tripuri; 6, from Bahal; 7, from Taxila (Bhir Mound); 11, 13, from Hastināpura; 14, 15, 19, from Ahichchhatrā
c6.s1.10
Briefly, in the following pages the schematic dates for the important Painted Grey Ware will be the 8th–5th centuries B.C.
c6.s1.11
Fig. 3
c6.s1.12
2. *Northern Black Polished Ware (N.B.P. Ware).* Many sites in northern and central India, occasionally as far south as Amaravati, have produced this Iron Age ware, which is as distinctive in the subcontinent as is *terra sigillata* on European sites. It is wheel-made and normally thin, with a highly lustrous surface ranging from grey or brown to black, and steel-like in quality. The paste is well levigated and is white to reddish. The bright gloss is not a glaze or lacquer. The process of manufacture is doubtful: it has been thought that, after being turned on the wheel, the pots were subjected to elaborate burnishing, and then coated with a finely levigated, highly ferruginous clay, and again burnished; and that they were then fired under reducing conditions to a temperature producing an incipient fusion of the slip, this accounting for their exceptional hardness and lustre. Recently, the laboratory of the British Museum has questioned the burnishing without, as yet, providing any complete alternative explanation. Its present verdict is that 'the unfired pots were dipped in a suspension of a ferruginous inorganic material, probably resembling a red earth; and that, after firing to a temperature of c. 800° C., the kiln was sealed so that the pots cooled in a reducing atmosphere. The mineralogical identity of the "red earth" has not been discovered, and the main problem, namely the precise nature of the surface layer, still remains unsolved.'
c6.s1.13
Fig. 3
c6.s1.14
The shapes are commonly bowls or convex-sided dishes, with a certain analogy—as, indeed, has the high quality of the fabric—to the P.G. pottery. The two may well represent the same tradition, with the addition of an improved or variant technique in the case of the N.B.P. Ware. The discovery of the new technique may be ascribed to the potters of the Ganges Basin, where the ware is particularly abundant. It may be suggested as an idle guess that the great popularity of the Ware was not altogether unconnected with its superficial resemblance to forged iron, which first appears in northern India shortly before N.B.P. Ware. Alternatively, it may reflect the liking of the Persians for a high metallic polish, as represented very notably in some of their masonry (p. 174).
c6.s1.15
The fabric has been dated on the basis of the prolonged excavations at the somewhat peripheral site of Taxila in the Punjab (West Pakistan), which yielded about twenty sherds. Of these, eighteen came from the earliest of the successive sites of the town, the Bhir Mound, which was founded late in the 6th or early in the 5th century B.C. and is thought to have ended about 180 B.C. It was visited by Alexander the Great in 326 B.C., when the average level was about 6 or 7 feet below the present surface. It is noteworthy that sixteen of the Bhīr Mound sherds occurred at depths between 7 and 13 feet, only two being higher than 7 feet (one at 4 feet 10 inches, and one at 6 feet 2 inches). Since Taxila was not dug stratigraphically in the modern usage of the term, these depths are an unreliable guide, but, consistently with them, one of the two sherds found on the succeeding site of Taxila at Sirkap occurred at 18 feet below the surface, and should therefore be early in the history of that site, that is to say, in the first half of the 2nd century B.C. On this showing, the N.B.P. Ware may be ascribed to the 5th-2nd centuries B.C. The occasional survival of this high-grade fabric into later levels is to be expected; thus at Shishupalgarh in Orissa three N.B.P. shreds occurred above Rouletted Ware, which is not likely to be earlier than the beginning of the 1st century A.D. But much more evidence is required as to the initial date of the ware in the Ganges Basin, and the Taxila dating may not in this respect be representative. On recent observation I am indeed inclined to suspect that in the far North-west, at Chārsada near Peshāwar (where a dozen stratified sherds of this fabric were found in 1958), at Udegram in Swat (where Professor G. Tucci's expedition has found a sherd in a mileu attributed to the 3rd century B.C.), and even at Taxila itself, its arrival should mostly be equated with the spread of the Mauryan dominion from the Ganges to these parts after 323 B.C. In other words, I would provisionally ascribe the N.B.P. Ware of the north-westerly regions of the subcontinent to the period 320-150 B.C., without prejudice to the possibility of an appreciably earlier beginning in the Ganges Basin itself.
c6.s1.16
It is relevant to add that in the careful trial-trenches dug at Hastināpura in the Ganges valley over a hundred sherds of N.B.P. Ware were found in Period III, represented by 8 vertical feet of occupation-material. In the succeeding Period IV, which produced Mathurā coins ascribed to the 2nd century B.C., the ware was absent.
c6.s1.17
3. Rouletted Ware. This ware was first recognized in 1945 at Arikamedu, near Pondicherry in South India, where it was found in association with Arretine ware imported from Italy in the early decades of the 1st century A.D. In a more general sense this dating is confirmed at Fig. 3 Distribution of painted grey and northern black polished wares
c6.s1.18
CHARSADA OTAXILA O R. Jhelum R. Ravi R. Sutlej RUPAR KOTLANIHANG PANIPAT BARNAWA BAGHPAT HASTINAPURA (DELHI) INDRAPAT TILPAT AHICHCHHATRA BAIRAT OMATHURA KAMPIL R. Chambal R. Jumna R. Ganges AYODHYA VAISALI PATNA BANGARH OKAUSAMBI RAJGHAT RAJGIR OCALCUTTA TAMLUK O OKARACHI R. Luni R. Narbada R. Tapti OBAHAL ONASIK R. Godavari OUJJAIN MAHESWAR SANCHI TRIPURI ONAGPUR SISUPALGARH R. Mahanadi BAY OF BENGAL ARABIAN SEA HYDERABAD OAMARAVATI R. Kistna Miles 0 100 200 300 • PAINTED GREY WARE ONORTHERN BLACK POLISHED WARE ⊗ SITE WITH BOTH TYPES □ MODERN CITY
c6.s1.19
Chandravalli in northern Mysore, where a denarius of Tiberius, dated A.D. 26-37, was found with closely derivative pottery and near actual rouletted sherds. The characteristic shape is a dish, sometimes more than 12 inches in diameter, with an incurved and beaked rim and two or three concentric rings of rouletted pattern on the flat interior base. This pattern is not an Indian feature and is certainly derived, like the associated Arretine, from the Mediterranean. The fabric is grey or black, has often a remarkably smooth surface, is usually thin, brittle and well-burnt, with almost a metallic ring. The better examples are doubtless imported;
c6.s1.20
cruder varieties seem to be local imitations. The ware is particularly characteristic of the South Indian areas reached by Roman trade in the 1st and 2nd centuries A.D. with an appropriate extension in the north-east to Tamluk, the ancient Tamralipti or Ptolemy's Tamalites, in Bengal,⁹ and a stray sherd in the far north-west at Jhukar in Sind.¹⁰
c6.s1.21
This abundant and distinctive ware lies itself outside the range of the present book, but as the first clear datum in the protohistoric chronology of South India it serves as a terminus for some of the dating in our later chapters.
c6.s1.22
CHAPTER III
c6.s1.23
Stones
c6.s1.24
INDIAN STONE AGE TOOLS have been provisionally classified in three broad categories, chronological as well as typological: Series I and II representing the Lower and Upper Palaeolithic, and Series III marked by microlithic industries. In spite of its crude simplicity, this basic classification is useful in the present stage of research. Series II still lacks definition, but I and III, however interpreted, are readily recognizable brackets. This chapter is concerned mainly with Series I.
c6.s1.25
In India it is more than ordinarily difficult to set Man of the Old Stone Age, Palaeolithic Man, squarely upon his feet. That he abounded for a great many thousands of years is sufficiently evident from the unnumbered lumps of stone which he split and shaped and left for us in a variety of geological environments. But of the bones and branches which he may also have hewn in the fulfilment of his daily needs, practical or spiritual, of the skins and matting which may have adorned his person and his precinct, nothing has survived. Of his physical aspect we know nothing. His solitary memorial is an infinitude of stones.
c6.s1.26
His setting lies within the Pleistocene epoch of geology—the last prior to the Recent or Holocene in which we live—as defined by the presence of a distinctive fauna: Elephas, Equus and Bos. More particularly, the stone implements encountered in this chapter may be equated with the Middle Pleistocene and a schematic time-bracket of 400,000 to 100,000 years ago, though the terminal date for some of the industries may well be later. Broadly, they fall into three main categories.
c6.s1.27
1. *Pebble-implements*. These are more or less crudely sharpened pebbles, characteristically of quartzite, and frequently retaining a part of the cortex or crust of the pebble. Sometimes flakes are struck upwards from the edges of the flat platform of a pebble which has been cloven transversely by man or nature; the steep cutting-edge thus formed would enable the implement to be used as a chopper or scraper. Alternatively, a rounded or oval pebble may be shaped by the removal of flakes, normally on one side only to produce a scalloped cutting-edge or even a point. In the latter case, the sharpened pebble sometimes vaguely resembles a handaxe, and a genetic connection between pebble-handaxes and true handaxes in India has been postulated. Equally, the resemblance may be fortuitous or at most the result of interaction. Pebble industries in India are often referred to as 'Soan' (or 'Sohan') industries, from the Punjab river where they were first recorded.
c6.s1.28
2. *Flake-implements*. Two main series (with many varieties) have been recognized. The first is comparable with the Clactonian of Essex, in that the flake is struck off an unprepared block by means of a hammer or by striking a fixed anvil ('block-on-block' technique); the striking-platform of the flake shows no preparatory working and is set at a wide angle (110-125 degrees) from the ventral side of the flake, and there is a large semi-cone of percussion where the formative blow was struck. The second series approaches the French Levallois technique whereby a domical or 'tortoise' core is first trimmed by the removal of a series of small flakes from the stone block; then a large flake is detached sometimes by a blow on a small flat surface prepared at right-angles to the surface of the core. The detached flake shows converging flaking or fluting on the upper side and a flat primary surface on the under side, and is generally usable as a knife or scraper or spearhead without further trimming, though it may on occasion be submitted to some secondary working.
c6.s1.29
3. *Handaxes*. These are pear-shaped or oval implements, sometimes of considerable size, a foot or more in length, worked to an edge on both sides of a core or occasionally a flake. The more primitive examples appear to have been made in a fashion similar to that of the Clactonian flakes; their French category is 'Abbevillian' (formerly 'Chellian'), but in India the term is liable to imply a greater measure of precision than the coarse material used there justifies. On the other hand the best Indian handaxes, showing, it seems, controlled flaking produced by the use of a bone or wooden mallet, may justifiably borrow the name 'Acheulian' from the more evolved French category. With the Acheulian handaxes are associated cleavers, oblong or rhomboidal in outline and with a wide chisel-edge, made by the same techniques. Levallois flakes partially overlapped the Acheulian handaxes but also outlasted them. Indian handaxes, with or without cleavers, are commonly known collectively as the 'Madras industry', from the region where they were first recognized, but the name bears no relation to the distribution of the type.
c6.s1.30
In 1863, R. Bruce Foote, of the Geological Survey of India, picked up a palaeolithic implement in a gravel-pit at Pallavaram, near Madras. From this early start began the intermittent study of the Old Stone Age of India, though it is a fair comment to add that that study is still in a rudimentary stage. Foote's own work was of fundamental value, but only in the last quarter-century has the subject received any further sustained attention from trained enquirers.¹¹ Here no complete survey or even summary can be attempted, but a few of the more notable palaeolithic sites in northern, central and southern India are selected to illustrate the current state of enquiry. If the result displays a somewhat sombre monotony, it is at least a sympathetic reflection of millennia of mental and social stagnation modified by reluctant technical variation and occasional slight advancement. In India, not less than in other parts of the world, the human race got off to a slow start.
c6.s1.31
Fig. 4 Early Soar implements: 1-4, pebble tools; 5-9, discoidal cores; 10-14, 'Clactonian' flakes. 1/2
Northern India
c6.s2.1
In northern India (Kashmir and the Punjab), as in Europe, the Pleistocene was an epoch of intermittent glaciation and, again as in Europe, four main advances of the ice have been recognized. It is likely but unproved that the Himalayan and the Alpine series approximate to one another in time, and may be equated in tropical India with pluvials or phases of extreme rainfall.
c6.s2.2
Towards the middle of the Pleistocene occurred the second of the four Himalayan glaciations, and the Boulder Conglomerate which represents it* has yielded the earliest tools of Stone Age India on sites in the sub-Himalayan valleys of the Indus and its tributary the Soan (or Sohan), in the so-called Potwar Area centred about Rawalpindi.¹ With the tools have been found bones of *Elephas namadicus*, which help incidentally to link these deposits with others on the Narbadā and Godāvarī of central India.
c6.s2.3
¹ The sequence here followed is that of de Terra and Paterson. For a suggested variation, see F. E. Zeuner, *Dating the Past* (4th ed. London, 1958), pp. 274-7. The available evidence is admittedly inconclusive.
c6.s2.4
These early implements, which have been christened 'Pre-Soan', are worthy of their lowly position in the human scale. They are characteristically big flakes or split pebbles of quartzite, with large unfaceted striking-platforms mostly at angles of 100-125 degrees and with large, flat but well-marked cones of percussion. Their upper surface is usually unflaked, and there is no secondary working. Their primitive quality is enhanced by the native intractability of the material; furthermore, they are mostly very worn, and the edges are often battered either by use or by natural attrition. Many of them are recognizable as artifacts only by the trained eye, assisted by their association with more distinctive specimens, and it may be suspected that even the trained eye is not always impeccable.
c6.s2.5
Following the second glaciation was a prolonged interglacial phase—the Second Interglacial—during which the Boulder Conglomerate was partially eroded and redeposited as a covering on Terrace 1. In this Terrace two new industries make their first appearance: (i) the 'Early Soan', a pebble and flake industry which has been subdivided typologically but consists broadly of chopping and scraping tools; and (ii) a handaxe or 'Madras' industry consisting of more or less ovate tools worked on both faces. In both of the Soan groups occur worn and deeply patinated specimens, but their complete absence from the undisturbed Conglomerate suggests that they belong truly to various stages of the succeeding interglacial phase represented by Terrace 1.
c6.s2.6
Fig. 4
c6.s2.7
The flakes of the Early Soan include examples with plain high-angled Clactonian platforms, and others with faceted low-angled platforms recalling the early Levalloisian. Retouch is absent. The pebble choppers and scrapers are made in both the fashions described on p. 35. No fauna has been associated with this phase.
c6.s2.8
During the third of the four glaciations, Terrace 2 was formed in a process of aggradation represented by a basal gravel and a thick overlying deposit of loess or silt, known sometimes as the 'Potwar Loess'. Teeth of horse and dog (or wolf) have been extracted from this loess, but statements that camel and ox have also been found in it require confirmation.
c6.s2.9
The 'Late Soan' industry recovered from this Terrace is of two phases, equating with the two component strata. Phase A, from the basal gravel, is largely an extension of the pebble industry of the Early Soan, and is particularly characterized by rough oval pebbles with untouched butt along one side and flaking along the opposite side, though sometimes from each surface alternately, thus producing a wavy edge, straight or convex. Phase B, from the overlying loess, has produced recognizable workshop-sites with fresh tools, consisting mainly of flakes comparable with the Late Levalloisian of Europe, though including some pebble-choppers and cores as in Phase A. It was observed that there was an evolving tendency towards smaller and more finely made types. ‘In the early stages [of the Soan industry] the flakes are crude. In the late Soan alongside the simple forms there are other flakes showing a development in technique, with much more regular primary flaking and often with faceted platforms.’
c6.s2.10
Fig. 5
c6.s2.11
It is noteworthy that the Late Soan seems normally to be devoid of handaxes. This statement, however, needs modification if in fact the Soan site of Chauntra (20 miles south of Rawalpindi), which has produced a mixture of Madras and Soan pebble elements, is of the succeeding third interglacial age from Terrace 3. More evidence is required. From Terraces 4 and 5 no significant material has been recorded; in any event Terrace 5 is of post-Pleistocene age.
c6.s2.12
In recent years (1951-55) other ‘Soan’ distributions have been found in the Himalayan foothills of the Punjab: in the upper valley of the Sirsa, a tributary of the Sutlej, 30 miles west of Simla, and near Daulatpur in Hoshiarpur District 25 miles north of Jullundur. More important are B. B. Lal’s observations in the upper valley of the Beas, with its tributary the Bāngangā, in the Kangra District 230 miles north-northwest of Delhi.¹ Four sites are named: Guler, Dehra, Dhaliara and Kangra itself. At Guler five river-terraces have been identified, though whether they are equivalent to the Kashmir-Potwar series has not been determined. Again, only the three higher (earlier) terraces produced implements, all of quartzite. Terrace 1 yielded unifacial choppers, a bifacial chopping-tool, and Clactonian flakes, but no handaxes. The workmanship is crude; the choppers consist of roughly sharpened pebbles, but the more evolved pebble-tool, approximating to a handaxe, is absent. From Terrace 2, more choppers were recovered, with a pebble-handaxe, Clactonian flakes, and two others which are reasonably described as proto-Levalloisian, with due allowance for the difficulty of the material. Both the pebble-handaxe and the proto-Levalloisian flakes seem to mark an advance upon the industry of Terrace 1. On Terrace 3 were at least seven unifacial choppers and two Clactonian flakes, and either from Terrace 2 or Terrace 3 came another proto-Levalloisian flake, a pebble-handaxe, and two bifacial implements which were, with some hesitation, regarded as genuine handaxes, distinguished perhaps with an excessive nicety as Abbevillian and Acheulian respectively. Both handaxes, if such they be, are rolled and worn. On the whole, the sequence from these three terraces is not in conflict with that from the Potwar area.
c6.s2.13
Fig. 3 Late Soan industry: 1-6, cores; 7, chopper; 8-12, cores; 13-15, Levalloisian flakes.
c6.s2.14
Farther south, in Rajasthan and Mālwā, a number of sites in the valleys of the Chambal river and its tributaries, particularly in Chitorgarh district, have produced Palaeolithic assemblages which seem to fall stratigraphically into two series.¹⁴ The earlier of these, Series I, included handaxes and cleavers of the Madras industry in association with Clactonian flakes and choppers of the Soan industry. A later gravel contained Series II, characterized by Levallois flakes with prepared striking-platforms, and hollow and blade-scrapers, but devoid alike of choppers and handaxes. The occurrence of occasional flake-tools of Series II in deposits of Series I suggested to the explorers a more or less continuous occupation with changing needs and modes. Away to the east, the Jumna valley in the Banda district of Uttar Pradesh is reported to have produced a comparable tool-sequence in so far as handaxes were followed by Levallois flakes, but apparently without the admixture of Soan choppers.
c6.s2.15
The next region towards the south where serious study of the Old Stone Age has been undertaken lies along the three main rivers which empty into the Gulf of Cambay—the Sabarmati, the Mahi and the Narbadā (or Narmadā)—and about the upper reaches of the Godāvarī, which nearly bisects the peninsula.¹⁶ Eastward the area of investigation has been extended to the valleys of the Kaimūr Range south of Banaras. The whole of this transcontinental region shows an interesting overlap between the flake-and-pebble industries on the one hand and the handaxe-cleaver series on the other, with an emergent hint of later Palaeolithic groups of which the full significance is not yet adequately understood. Here too, if anywhere, is it likely that an equation or compromise between the plurials of the tropics and the glaciations of the Himālayas may eventually be postulated;¹⁶ but much further fieldwork, geological and archaeological, must precede this happy result. Indian research is already busy in the matter.
c6.s2.16
The great Narbadā valley¹⁷ must have been a superb hunting-ground for Palaeolithic man, as it is for the modern explorer in his tracks. Its geological structure, so far as we are concerned, consists in the central reaches of two sedimentary deposits, a Lower and an Upper, comprising gravel capped with clay and sand; they overlie a thick deposit of laterite (see below) and are covered by the dark alluvial 'regur' or 'cotton soil' familiar to travellers in central and southern India. The junctions between these successive formations are well marked, and the Lower group, with extension into the gravels of the Upper group, has produced bones of a mid-Pleistocene fauna including *Bos namadicus*, *Elephas namadicus* and *Bubalus palaeindicus*. It may be recalled that *Elephas namadicus* has been found likewise in the Boulder Conglomerate which yielded the earliest (Pre-Soan) flake implements of the sub-Himalayan Potwar area, and thus provides a useful palaeontological link between the two regions. It is appropriate that the earliest vestiges of Narbadā man are likewise contemporary with the formation of the Lower Narbadā group, and that they include an industry comparable with the Pre-Soan.
c6.s2.17
With the large flakes of Pre-Soan type, however, the Lower Narbadā group—unlike the Potwar Boulder Conglomerate—contains an appreciable handaxe industry, represented by heavily rolled Abbevillian implements and both rolled and unworn Acheulian tools. In the Upper group the handaxes seem to be less at home; they are predominantly rolled Acheulian and are thought to be confined to the basal gravel of the group. They may therefore be wholly or largely derivative. In any event, the characteristic industry of the Later group resembles the Late Soan, and consists of small discoidal cores, pebble cores, pebble choppers and beaked tools.
c6.s2.18
The interpretation of these resemblances and discrepancies between northern and central India is not easy. Nor is it facilitated by more recent observation in the Sabarmati valley.¹⁸ There, in the main gravel, it is noted by Dr. Sankalia that crude pebble-implements occur alongside more finished types which include Acheulian handaxes of latish appearance. There are also numerous flakes, though rarely with the faceted striking-platform suggestive, as might be expected from the northern evidence, of a Levalloisian technique. Typologically, this complex assemblage could be divided into a sequence of cultures beginning with pre-Abbevillian, but there is no stratigraphical warrant for this and it must be assumed that the various types were approximately contemporary.
c6.s2.19
The problem is again illustrated but not solved in a tributary system of the river Son, the Singrauli basin, of Mirzapur District in the Kaimur Range south of Banaras.¹⁹ Here a bed of redeposited Pleistocene gravel, covered by 5 feet of alluvial silt and humus and overlying a layer of sand upon a deep Boulder Conglomerate, has produced a series of quartzite implements which include about 43 per cent of handaxes and cleavers and about 15 per cent of pebble-and-core chopping tools. There are also numerous flakes of which a fair proportion (some 26 per cent of the collection) may be described as proto-Levallois, or even unqualified Levallois, with faceted platforms. The pebble-tools recall the Early Soan 'chopper' industry and may be so classified. The bifacial handaxes are of Abbevillio-Acheulian or Madras types, some of them exhibiting all-over controlled flaking, with a more or less tongue-shaped point. The cleavers are straight-edged and round-butted, with oblique cutting-edge.
c6.s2.20
It should be emphasized that there was no apparent stratigraphic division between the pebble and the biface industries in this deposit. The pebble-tools were thought to be rather more worn than the bifaces, many of which were fresh in appearance; but this is the only hint of a differential chronology, and for the present the whole series must be regarded as approximately contemporary. No associated fauna is recorded.
South India
c6.s3.1
South of the Narbadā, the pebble industry dwindles but does not completely vanish. Such at least is the evidence as it stands, with the proviso that, where handaxes or cleavers are present, there is a natural tendency to collect them at the expense of cruder pebble artifacts, and so to create a false bias. As things are, the most southerly site at which pebble-tools have been recognized in quantity is in Mayurbhanj in Orissa, where river-sections near Kamarpal have produced a mixed handaxe and pebble industry resembling the Early Soan but with the pebble-tools in a minority.²⁰ Farther south, pebble-tools occur only in small proportion, though other types or industries supplement and extend the Madras handaxes and cleavers which dominate the Peninsula. At first sight, a tenuous link between north and south is sustained by a distinctively mid-Pleistocene fauna. Thus at Nevasa on the Pravara river, tributary of the Godāvarī, between Ahmadnagar and Aurangābād, 150 miles east-north-east of Bombay, a hard cemented basal gravel (I), which is capped successively by two further gravels (II and III) and a thick layer of yellow-brown silt, has produced teeth of *Bos namadicus*;¹ and both *Elephas namadicus* and hippopotamus have been found in comparable gravels of the Godāvarī itself. As between the Soan and the Narbadā, a faunistic equation of this kind seems to make sense (p. 38), but it must be confessed that a similar equation between the Soan and the Pravara is at present less easy. Thus the basal Gravel I of the Pravara, which produced *Bos namadicus*, is said to have yielded a handaxe-cleaver industry of Early-Middle Acheulian type (using trap and dolerite); whereas the Boulder Conglomerate of the Soan, which produced *Elephas namadicus*, yielded only the crude 'Pre-Soan' industry. This apparent unconformity may represent a true technological variation between the two regions, or may be due rather to contemporary differences between the tropical and the submontane fauna; but it needs watching, and illustrates the perils of hasty equation.
c6.s3.2
On the Pravara, the Acheulian industry occurs also, though rarely, in Gravels II-III; in particular, an unrolled handaxe of chalcedony was found in Gravel II, 8 feet above the top of Gravel I, and a large pear-shaped handaxe of trap, flaked on both sides by fine step-technique and with subsequent retouch along the edge, one side weathered and the other side in mint condition, lay as high as the surface of Gravel III, under the final silt. But the normal industry of the upper two gravels was of another kind, consisting of scrapers, blades, cores, burins and points of agate, chert, chalcedony and jasper. Here is an acceptable Upper Palaeolithic craftsmanship of a sort at present inadequately represented in our Indian material, and further reference will be made to it in the next chapter.
c6.s3.3
Fig. 6
c6.s3.4
Further south, in and near the Nallamalais Range of the Kurnool-Cuddapah districts north of Madras, a number of
c6.s3.5
Fig. 6 ‘Series 1’ from Nevasa, Pravara valley: trap handaxe flaked by fine step-technique, with subsequent retouch. 1/2
c6.s3.6
sites have produced a large assemblage of implements which were analysed in 1930 in a classic paper by Cammiade and Burkitt.¹⁸ The geological environment indicates a succession of prolonged pluvials and interpluvials, the basis being a reddish clayey mass of 'laterite', largely of hydrated alumina and iron-oxide, formed by the weathering of crystalline rocks under torrential rainfall alternating with dry seasons over a long stretch of time. It was on the weathered surface and in washed-down deposits of this laterite, when 'the swamps and forests of the laterite period gave place to large open plains' under supervening dry conditions, that human relics first appeared upon the scene. They consist largely of handaxes (classified as 'Series 1'), unrolled and stained by a largely-vanished capping of laterite. The implements are characteristically made from quartzite pebbles; they show the bold flaking and irregular edges which are encouraged by the coarse material and must not therefore be over-emphasized in a typological classification. Amongst the handaxes is an occasional 'rostro-carinate', with flat ventral plane produced by the removal of a large flake, and a keel-like back formed by part of the original surface of the pebble—a type similarly associated with handaxes at Victoria West in South Africa and in the Oldowan of East Africa.²³
c6.s3.7
¹⁸ [Note text not visible on page]
c6.s3.8
Fig. 9.1
c6.s3.9
Above the laterite-stained pebble-bed which produced the handaxes, a red clay laid down under renewed pluviation yielded flakes ('Series 2'), and other sites in the region are marked by flake-industries at an equivalent horizon. Handaxes are now the exception. Some of the flakes have prepared striking-platforms, others have not. Occasional examples may fairly be described as burins; one as 'a double-ended burin of a very Upper Palaeolithic appearance'. Whilst this post-handaxe series must represent (a part of) the later Palaeolithic of peninsular India, it is at the best a crude product, at a lower level of technical achievement than either the best handaxes and cleavers of 'Series 1', or the more slender backed-blades, planing-tools, end-scrapers, core-scrapers, and burins, made of materials of a more flinty nature, classified as 'Series 3'. The latter occurs on the surface of the red clay, and prepares the way for 'Series 4', which marks the topmost deposits and is generally microlithic in character (agate and quartzite). This belongs to the next chapter.
c6.s3.10
Madras industry: 1, handaxe trimmed all over by controlled flaking; 2, similar, with pebble butt; 3, cleaver; 4, handaxe. All from Attirampakkam. 1/2
c6.s3.11
Of the chronology of these industries, all that can be said at present is that not less than three pluvial episodes are seemingly identifiable since the first appearance of mankind in these parts, and the successive climatic phases were certainly of long duration. But it must again be affirmed that any attempt to equate these pluvials with the three glaciations associated with implements in the Potwar area of the north is premature on the evidence available.
c6.s3.12
A parallel series was identified by T. T. Paterson in and above the Red Hills near Madras, where four terrace surfaces were recognized.¹² Here a considerable layer of detrital or redeposited laterite overlay a well-developed Boulder Conglomerate; and the laterite was eroded to produce Terrace 1, on which were minor deposits of sand and gravel. This in turn was eroded to form Terrace 2, where thicker gravels were deposited and then covered by alluvium. Three groups of implements are associated with this sequence at a number of sites.
c6.s3.13
¹² [Note text not visible on page]
c6.s3.14
At Vadamadurai Tank, north-west of Madras, the Boulder Conglomerate at the base of this sequence produced handaxes and cores of the first group, which, on grounds of typology and patination, falls in turn into an earlier and a later series. The earlier series includes very crude Abbevillian handaxes with thick pebble butts and much remaining cortex; the flaking denotes a 'stone' technique producing large irregular flake scars. The cores belonging to this series were large, also with rough irregular flaking. Both handaxes and cores bore a very deep whitish crust. The later series is less heavily patinated and is typologically more advanced. It is marked by handaxes of Early Acheulian type, showing the beginnings of a step-flaking technique, though large free flaking is still commonly used. The cores are mostly discoidal, with fairly regular alternate flaking.
c6.s3.15
The overlying laterite gravel had stained to a characteristic red colour the implements of the second group. This comprises neater and more regularly shaped (ovate and pear-shaped) handaxes of Middle Acheulian type showing considerably more step-flaking, together with flakes which have only primary flaking and unfaceted platforms.
c6.s3.16
The third group has no laterite-staining and but little patination. The handaxes may be classified as Upper Acheulian, and include both ovates with small, fairly flat step-flaking, and
c6.s3.18
Fig. 8 Madras industry: 1, ovate; 2, cordate with secondary step-flaking along margin. Both from Attirampakkam
c6.s3.19
long pointed forms with thick pebble butts and large, neat free-flaking. The cores are discoidal; the flakes still have unfaceted platforms, but a few have been retouched for possible use as side-scrapers. One cleaver is present. This third group belongs to the gravels of Terrace 1.
c6.s3.20
Attirampakkam near Madras is an important site which demands further work. Professor Zeuner informs me that the early series here, consisting of pebble-tools, flakes and handaxes, is pre-laterite. Then came a laterite-weathering phase, [Figure: Figs. 7-8] during which the decomposition of the earlier implements proceeded. This was followed by the denudation of the laterite deposit, and the sediments thus formed in the valleys contain an Acheulian industry with a strong flake-component. The so-called Attirampakkam 5 series (late Acheulian) is fresh and is found in the lowest or latest terrace.
c6.s3.21
In recent years fresh fieldwork has been carried out in the vicinity of Giddalur, District Kurnool.²⁴ The Series 1 and 2 of the Cammiade-Burkitt analysis have again been recognized, and within them a gradation from earlier and rougher to later and more evolved has been partially established on the basis of the degree of rolling present in the individual specimens. Thus the Abbevillio-Acheulian handaxes and associated rostro-carinates of Victoria West type, made out of pebbles which often preserve cortical patches, are liable to exhibit a fair measure of rolling. With them are associated Clactonian flakes and cores, likewise rolled, and flake-made handaxes. The more advanced Acheulian ovates and handaxes, ranging from narrow, elongated specimens to wide, almost discoidal forms, are generally fresher in appearance; but no clear geological stratification is described. Cleavers are abundant, and flakes with prepared platforms, of Levallois type, occur though in a small minority. One or two scrapers, including a double-ended scraper with 'nibbled' retouch, recall Upper Palaeolithic industries of the West, but typologically the two Series may be described generally as of a Mid-Palaeolithic facies. Frequent cross-references to African analogies are convincing, though the significance of these comparisons is matter for discussion. Here the problem in relation to the Indian Palaeolithic as a whole, can only be indicated as a background to further research.
First 83 of 107 passages. Read on