Minerals and Metals in Ancient India v1.c20: 12. Iron in Ancient India

Minerals and Metals in Ancient India v1.c20

Language: English

v1.c20.s1 The Science of the Ancient Iron Technology — 43 parts
v1.c20.s2 The Earliest Indian Iron in 1300 BC Chalcolithic Ahar? — 3 parts
v1.c20.s3 The Five Iron Age Cultures in India — 11 parts
v1.c20.s4 Iron Metallurgy at Dhatwa — 6 parts
v1.c20.s5 Iron Metallurgy at Rajghat, Varanasi — 7 parts
v1.c20.s6 Conclusion — 88 parts

12. Iron in Ancient India

v1.c20.1

In this chapter we would examine several questions which invariably crop up in the discussions related to Indian archaeo-metallurgy. What are the earliest dates for the use of iron in the world and in the Indian sub-continent? Was the technology imported to India or indigenously developed? How soon were carburised iron and steel invented and used in India?

v1.c20.2

Before we try to answer the above questions, and attend to the Indian scenario, let us probe into the scientific reasons underlying the late emergence of the iron age relative to the chalcolithic age.

The Science of the Ancient Iron Technology

v1.c20.s1.1

Different metallurgical traditions were possible on account of easy availability of the ores and the ease in thier reducibility to the respective metals having not too high melting points. Table 12.1 provides the melting and smelting temperatures of some metals of archaeological importance.

v1.c20.s1.2

It is evident that tin and lead ores were the easiest to smelt. In a furnace providing a temperature of 1100-1200°C all the five metals except iron, mentioned in the table, could be easily smelted and the molten metals could be alloyed and cast into various shapes.

v1.c20.s1.3

At a temperature of 1200°C, iron oxide could be reduced to the metallic and solid state (not molten), provided the partial pressure of oxygen was low and CO/CO₂ ratio in the carbothermic environment high. There are archaeological finds of ‘magnetic copper’ and evidences of accidental production of iron in the copper smelting of sulphide ores. (Ashenbrenner and Cooke, 1975) It is further claimed that copper slags containing iron oxide phases might have been smelted to produce iron. However, it should be noted that iron silicate phases in copper

v1.c20.s1.4

Table 12. 1

v1.c20.s1.5

Melting and Smelting Temperatures of Some Metals of Archaeological Importance

v1.c20.s1.6

| Metal | Melting Point in °C | Minium Reductivce Smelting Temperature in °C |
| Tin | 232 | 600 |
| Lead | 327 | 800 |
| Silver | 960 | 800 |
| Gold | 1063 | 1063 |
| Copper | 1083 | 400 |
| Iron | 1540 | 800 |

v1.c20.s1.7

slag are not easily reducible to the metallic state, and the hypothesis of discovery of iron in a chalcolithic smelter may not be universally true.

v1.c20.s1.8

Man-made iron can be easily distinguished from meteoritic iron which almost always contains high (more than four) percentage of nickel. In the ancient Near East and eastern Mediterranean, iron ore used to be smelted in small furnaces resembling copper smelting furnaces. (Tylecote, 1962; Maddin et al., 1977; Maddin, 1982). Furnaces about a meter deep and about two-thirds of a meter in diameter were filled with iron ore and charcoal in a number of layers. A clay tuyere was led in from the side or from the top and attached to a bellow. The temperature in the hottest part was around 1200°C, and although reduction took place due to low Po₂ and high CO/CO₂ ratio, the metallic product was not liquid but a spongy mass.

v1.c20.s1.9

Caught up in the spongy solid mass was slag, mostly fusible ferrous silicate or fayalite (melts at 1170°C). The slag had to be squeezed out of the sponge iron by hammering at or above the temperature (1170°C) — a process known as forging. Hammering produced a continuous network of iron grains, sometimes interrupted by stringers of remaining slag elongated in the direction of the deformation of the iron bloom.

v1.c20.s1.10

Unlike copper, iron objects had to be forged stage by stage from a spongy mass or bloom to a bar, then thinning, pointing, folding and largely forge-welding at white heat. This knowledge did not come to the copper-smiths automatically, and thus it took centuries of a chalcolithic age to obtain the first few pieces of forge-welded iron.

v1.c20.s1.11

Even at this stage, iron was hardly useful because wrought iron, containing very little carbon, is softer than ten per cent tin bronze which was well-known to the copper-smiths. Despite the early successes to make iron (Coghlan, 1956) in Anatolia, Mesopotamia; Egypt and Afghanistan (3000-2000 BC), also later in Crete, Cyprus (2000-1600 BC) and still later in Greece and the islands of the Eastern Mediterranean, (1600-1200 BC), the spurt and diffusion of iron technology did not take place till 1200 BC, when it became known that iron can be carburized and hardened. The discovery must have been accidental.

v1.c20.s1.12

The solubility of carbon in iron is small at room temperature, but it increases greatly above 910°C. The hot spongy mass of iron when kept in prolonged contact with charcoal, absorbed carbon into the surface layers and thus became surface-hardened. The iron with the absorbed carbon is arranged in a two-phase structure called pearlite, which consists of alternate layers of soft and ductile iron (ferrite) and hard and brittle iron carbide (cementite). The hardness of the carburized sample depends upon the relative proportions of the two phases. Careful microscopic studies permit identification of pearlite structure relics even in a totally corroded iron object. (Knox, 1963)

v1.c20.s1.13

Incontrovertible proof of deliberate carburization can be derived from objects made by laminating layers of iron of varying carbon contents, for there is no reason to make such a structure unless the different properties of the layers were understood. Maddin et al. (1977b: 14) reported that “the earliest known laminated object is an Egyptian knife from the Ramesseum which probably dates to 900-800 BC”.

v1.c20.s1.14

Maddin further reported (1982) that quenching of carburized iron was practised by people in Cyprus and perhaps in Palestine by the eleventh century BC. A knife excavated from Idalion in Cyprus and belonging to a twelfth-eleventh century BC context showed surviving martensite and a form of bainite. An excavated pick from Mt. Adir in Northern Gallilee, Palestine, of the same age exhibited high hardness (545 DPH) corresponding to martensite, obtained by quenching from an austenitic condition, further tempered at a lower temperature. Homer described in his Odyssey (8th to 7th century BC) how a “black-smith plunges a screaming great axe into cold water treating it for temper and to make the steel strong”.

v1.c20.s1.15

To summarise, it is clear that iron known before 1200 BC was mostly wrought iron and ‘used sparingly, mostly for votive, amuletic or prized items of exchange’. (Maddin, 1982) Examples of deliberately carburized and quenched samples are available only from twelfth-eleventh century BC context onwards. The spurt in iron technology throughout the world is believed to be associated with the break of the Hittite empire and diffusion of technical know-how across the national borders.

v1.c20.s1.16

How Did the Iron Technology Arise in India?

v1.c20.s1.17

Gordon (1958) and Wheeler (1959) attributed the appearance of iron in India to the Graeco-Roman influences in the fifth and fourth centuries BC. Their theories have been discarded in view of the definite evidences regarding the presence of iron in India as early as 1200 BC, if not earlier. Some of the Iron Age Culture Sites in South Asia are shown in fig. 12.1. Prakash and Tripathi (1986) have documented the early iron age cultures, which could have been independent of each other, arising in five different zones of India. The sites corresponding to the five cultures (A to E) are shown in fig. 12.2. Some radio-carbon dates are indicated in fig. 12.3.

v1.c20.s1.18

Even if we ignore the early date of Ahar (1300 BC) which is disputed, it is evident that iron was discovered in India during or before the twelfth century BC. Of the five cultures:

v1.c20.s1.19

(A) Cairn Burial Culture in north-western India

v1.c20.s1.20

(B) Painted Grey Ware (PGW) Culture in north India

v1.c20.s1.21

v1.c20.s1.22

(D) Megalithic Culture of central India

v1.c20.s1.23

(E) Megalithic Culture of Peninsular India

v1.c20.s1.24

at least three cultures (B), (C) and (E) show dates before 1000 BC for the introduction of iron in India. Each one of these had developed distinct ceramic and chalcolithic traditions and therefore there is a strong case for the hypothesis of indigenous and independent development of iron technology in more than one Indian nuclear zone. N.R. Banerjee (1965) one of the celebrated early authorities on the iron age in India did not accept hypothesis as stated above.

v1.c20.s1.25

The Diffusionist Paradigm

v1.c20.s1.26

Banerjee reviewed the data available till 1965. Further excavations and discussions have taken place for more than a quarter of this century; therefore, we should not rule out the possibility of Banerjee substantially modifying his earlier conclusions (1965) which are present below.

v1.c20.s1.27

In the summary of his monumental treatise, Banerjee concluded (1965: 238-39):

v1.c20.s1.28

It was the Aryan settlers in India, who also used the PGW ceramic, that were the first to use iron in ancient India around 1000 BC, if not earlier. Around 800 BC or slightly earlier, the use of iron is testified among a people who buried fragmentary remains of their dead in

v1.c20.s1.29

Fig. 12.1: Map of the Iron Age Culture Sites in South Asia.

v1.c20.s1.30

Key : 1. Pirak, 2. Charsada, 3. Timargarha, 4. Taxila, 5. Rupar, 6. Bhagawanpura, 7. Panipat, 8. Sonepat, 9. Indraprastha, 10. Hastināpura, 11. Lalqila, 12. Jakhera, 13. Atrañjikhera, 14. Sravasti, 15. Jodhpura, 16. Noh, 17. Kauśāmbi, 18. Rājghāt, 19. Prahladpur, 20. Buxar, 21. Chirand, 22. Sonpur, 23. Rajgir, 24. Somnath (Prabhas Patan), 25. Nagara, 26. Ahar, 27. Nāgdā, 28. Ujjain, 29. Eran, 30. Nagal, 31. Ranjala, 32. Prakash, 33. Takalghat, 34. Mahurjhari, 35. Naikund, 36. Sisupalgarh, 37. Pochampad, 38. Bhongir, 39 Raigir, 40. Terdal, 41. Jevargi, 42. Piklihal, 43. Maski, 44. Yelleshwaram, 45. Nāgārjunakoṇḍa, 46. Tekkalakota, 47. Sanganakallu, 48. Hallur, 49. Brahmagiri, 50. Jadigenahalli, 51. Savandrug, 52. Kunnattur, 53. Pallavaram, 54. Sanur, 55. Perumbair, 56. Arikamedu, 57. Souttoukeny, 58. Mouttrapaleon, 59. Porkalam, 60. Alagarai, 61. T. Kalluppatti, 62. Adichanallur.

v1.c20.s1.31

Fig. 12.2: Map of the Indian Sub-continent showing some of the Archaeological Sites where Iron Age Articles have been found.

v1.c20.s1.33

Fig. 12.3: Radio-carbon Dates of Some Iron Age Sites in India.

v1.c20.s1.34

cairns in the borders of Iran and Pakistan. It is possible that they obtained their knowledge of iron from Sialk VI in Iran even as the Aryans of the Ganga Plains apparently did (emphasis supplied). A little later iron came into its own in West-Central India in the post-chalcolithic phase, as also in the megaliths of South India. In both cases, they (the inhabitants) owe not a little of their metallic inspiration to the northern plains. The possibility of a cultural influence direct from Iran (Sialk VI) being transmitted through the cairn burials to the megaliths of South India, cannot be ruled out.

v1.c20.s1.35

It is evident that Banerjee put lot of emphasis on the ‘diffusion from outside’ theory, the immigration of the so-called Aryans into India and the primacy of the iron-making traditions in Sialk VI. He referred (1965: 237) to the Hittites being defeated in circa 1190 BC, which set “in motion a colossal exodus of an Indo-European people in all directions with all their arts and sciences”, and resulted in “a new people occupying Iranian plateau, i.e., Sialk VI and introducing iron in the new habitat”.

v1.c20.s1.36

The iron-making technology could be consolidated in Sialk VI only long after the stated date of 1190 BC and we do not see how the knowledge of iron-making could diffuse from Sialk VI into India, if it ever did, before the eleventh century BC. Cemetery B of Sialk, exhibiting an abundance of iron, has been dated to c. 900 BC by Ghirshman (1954). Chakrabarty (1979: 355-58) explained how Gordon’s (1950) argument to establish a Sialk analogy for the cairn-burials in India was untenable, and yet accepted uncritically by Banerjee (1965) to build up his hypothesis. Prakash and Tripathi (1986) have quoted the C-14 date (of TF-191 sample) 1155 ± 110 BC for the PGW culture in Atrañjikhera and the twelfth century BC date of Tadakanhalli — both corresponding to the introduction of iron in India. The above-mentioned dates just do not seem to support Banerjee’s hypothesis and ‘eagerness to explain anything in terms of a migration from the West’.

v1.c20.s1.37

Like many other scholars, Shaffer (1984a) and Biswas (1990) have exposed the Aryan myth. Several eminent Indian archaeologists have fallen into this linguists’ trap. Banerjee himself (1965: 224, 232) has criticised H.D. Sankalia and D.P. Agrawal for equating Aryans with the Chalcolithic people and the Banasians respectively, and has approvingly quoted A. Ghosh who was “not so sure of the correctness of the attempted equation, and the old theory of the linguists”. Yet, Banerjee imagined that the ‘Aryans’ (whoever they were) came to India and introduced iron technology and PGW culture. Chakrabarti (1968) has given cogent arguments in favour of an eastern, rather than western, origin for the PGW, thereby negating the PGW-Aryan correlation.

v1.c20.s1.38

Banerjee (1965: 65-67) proposed that the megalith builders of Peninsular India might have picked up iron age culture around 800 BC either from the migrant Aryans or the cairn burial people of the North-Western India. The iron age of the cairn burial people in Pirak of Baluchistan, Swat Valley and Gufkral at Kashmir are all dated around 900-800 BC which is much later than the Tadakanhalli date of twelfth century BC. Thus, there was little possibility of the megalith builders ‘picking up’ iron age culture from the cairn burial people of the North-West. It could have been the other way round. There are four distinct possibilities regarding the origin of iron-age culture in the North-Western India: (a) indigenous development, (b) influence of the migrant megalith people from the South who shared the burial traditions, (c) influence of the Atrañjikhera/ Ahar PGW culture, or (d) influence from the Iranian culture of Sialk VI after ninth century BC. The possibility of Sialk VI iron-making tradition influencing the whole of India through the immigrant ‘Aryans’ or the cairn burial people may be safely ruled out.

v1.c20.s1.39

In this connection, we may refer to the bronze age iron from Afghanistan. (Shaffer, 1984b) Decorative iron ore nodules found at Said Qala Tepe and decorative iron balls and buttons, found at Said Qala Tepe and nearby Mundigak IV, have been dated to be around 2300-1900 BC. These early Harappan iron balls and buttons may be compared to the decorative pieces from the Mediterranean area dated third millennium BC, and indicate that wrought iron was not completely unknown in the ancient India of the Harappan era. However, these soft decorative materials could not replace bronze, and the real breakthrough or the advent of harder carburized iron took place in India, as in the rest of the world, only around 1200 BC.

v1.c20.s1.40

We do not deny the possibility of diffusion of knowledge in the ancient world. What we question is the romantic notion of the Central Asian ‘Aryans’ coming all the way to India with horses, conquering the land and spreading PGW and iron technology. Of course, the ancient chalcolithic cultures within and outside India were linked by trade, and therefore interactions must have taken place.

v1.c20.s1.41

We propose that diffusion of knowledge related to iron technology was not uni-directional. The technology grew in the ancient world in several phases, each covering centuries of time. We may identify at least five of these phases: (1) discovery of wrought iron and its ornamental use (3000-1200 BC), (2) discovery and use of surface-carburised iron, (3) use of forge-welded laminated sheets of carburised iron, (4) practices of quenching, hardening and annealing, (5) making of high carbon crucible steel.

v1.c20.s1.42

The first phase might have started in Anatolia, and yet in the bronze iron of Afghanistan (2300 BC) we have an example of early iron-awareness in the Harappan India. The second phase might have been indigenously developed in India in three nuclear zones viz., Gaṅgā-Yamunā doab, Karnataka Megalithic Culture and Bengal-Bihar area. Early iron in these areas, as we will record, are dated 1200-1000 BC. We do not have many dates outside India, earlier than the above, for iron deliberately carburised. We have mentioned (Maddin, 1977 b) ‘the earliest known laminated object’: an Egyptian knife dated 900-800 BC, which came centuries after Komaranhalli spear and Tadakanhalli axe in India clearly demonstrating lamination techniques in iron artefacts. (Agrawal, O.P. et al., 1990)

v1.c20.s1.43

We do not have any incontrovertible evidence of quench-hardening practice in ancient India as early as twelfth-eleventh century BC, the date corresponding to Cyprus knife showing martensite and bainite. (Maddin, 1982) But the credit of first producing crucible steel or ‘wootz’ certainly goes to India. These facts justify the notion of India entering the iron age substantially through indigenous efforts, occasionally receiving as well as transmitting technological details across its borders.

v1.c20.s1.44

The above ideas were accepted by most of the participants in the 1983 Poona Seminar. (Chakrabarty, 1985)

The Earliest Indian Iron in 1300 BC Chalcolithic Ahar?

v1.c20.s2.1

The participants in the 1983 Poona Seminar debated on the above question and the answer appears to be negative. Sankalia et al. (1969) had excavated the chalcolithic site of Ahar in Rajasthan, and their publication containing an erroneous stratigraphical chart created confusion. The iron artefact-containing layer was wrongly marked as IC, which led Sahi (1979) to propose that the earliest iron in India was evolved in chalcolithic Ahar during the thirteenth century BC. Although this view was endorsed by Chakrabarti (1982: 83) as well as Allchin and Allchin (1982: 325-26 & 345), the excavators themselves disagreed with this interpretation in view of “the great chance for contamination (between layers I and II), and the later objects getting mixed up with the earlier”. This point was subsequently clarified by Lal (1987-88).

v1.c20.s2.2

Charkrabarti (1982: 83) endorsed Sahi’s conclusion that “iron metallurgy was invented independently as an offshoot of copper metallurgy in India, particularly in the regions of Malwa and Banas cultures” on the basis of the works of Charles (1979) and Hegde (1981). We agree about the concept of independent origin of iron in India but feel uncertain whether this was ‘an offshoot of copper metallurgy’. Hegde (1981) was quoted but probably misunderstood; what he wrote was that it was not easy for copper-smiths to invent iron. They had the furnace which could provide a temperature of 1250°C for solid state reduction of iron ore, but they had to wait for centuries before the conflicting demands of high rate of oxygen supply and high CO/CO₂ ratio or a reducing atmosphere within the charcoal bed could be met substantially. The evidence of small quantities of poor quality iron occurring at the earliest levels of the iron age sites shows intensive experimentation, and proves that “we did not receive the technical know-how of iron smelting and smithery from an outside source”. (Hegde, 1981: 196) The argument may be further extended. Atrañjikhera and Jakhera PGW sites exhibit a large number of iron tools of varied nature; therefore, we may propose that iron was possibly discovered in the earlier periods of BRW era which shows ill-defined iron objects.

v1.c20.s2.3

Since chalcolithic tradition preceded the iron age, the latter must have evolved from the efforts of copper-smiths. This generality need not be interpreted specifically to justify an early advent of iron in the chalcolithic Ahar. It is true that the Ahar slag contained 43-48 per cent iron oxide (and 35-38% silica) (Hegde, 1969), but this was most probably in the form of a silicate such as fayalite, which originated from the use of iron-containing chalcopyrite and silica as flux. Copper artefacts of Ahar contain 1-6 per cent iron which probably exists as slag inclusion rather than in the metallic state. Unfortunately, phase studies of these materials were not undertaken by Hegde, but we know that reduction of iron to the metallic state is possible from the oxide phases but not the silicate phases. Therefore, the examples of iron being accidentally produced in copper smelter (Ashenbrenner and Cooke, 1975) need not be indiscriminately invoked.

The Five Iron Age Cultures in India

v1.c20.s3.1

Following Young’s division of the Iron Age in Iran into three stages, Allchins (1983: 345) divided the Iron Age in India into three periods: (I) spanning 1300-1000 BC — ‘the earliest report from south Rajasthan (Ahar), then Pirak (north-west), Hallur (south), and still later from Brahmāvarta (Noh)’, (II) second period 1000-800 BC, and (III) 800-500 BC period.

v1.c20.s3.2

Such a periodisation of the iron age in India is at best tentative and certainly open to question. Most people do not accept the Ahar date as we have discussed before. Pirak’s date comes after that of Atrañjikhera, which has been strangely ignored by Allchins. It may be useful to have separate discussions on five distinct iron age cultures (Prakash and Tripathi, 1986) A to E mentioned before in this chapter.

A. North-Western Culture

v1.c20.s3.t1.1

In north-east Baluchistan, in the Zhob-Loralai region, a chain of cairn-burials containing iron was discovered by Stein and Mockler. There are no C-14 dates and the chronology is quite controversial. Chakrabarty (1979: 355-58) has questioned whether these can be dated much earlier than the Historical Period of only a few centuries BC. Eight consistent C-14 dates show that the Iron Age levels of Pirak, in the Kachchi plains of Baluchistan, are datable to 800 BC. Dani and Stacul differ in the periodization of the Swat sites e.g. Loebarn, Timargarh, Ghalighai etc. Dani puts the iron-bearing Period III to c. 900-600 BC. Stacul suggested an earlier date. Following Stacul’s scheme, D.P. Agrawal (1984: 251) suggested that the first appearance of iron in the Swat area is datable to c. 1000 BC. Chakrabarty (1979: 360-62) does not find this acceptable, and points out that even Stacul admits that Period V of Katelai (showing iron for the first time in Swat Valley) continued up to 900-800 BC.

v1.c20.s3.t1.2

Thus, north-western India can hardly claim any antecedence over Atrañjikhera in the western Uttar Pradesh and the sites of Noh and Jodhpura in Rajasthan regarding the first advent of iron in India. The anxiety to emphasize ‘early’ dates in the north-west may be traced to the subjective preference for the diffusionist theory (intrusion of the ‘Aryans’ from the west).

B. Iron Age Culture of North India

v1.c20.s3.t2.1

Having discounted the claims of Ahar, Loebarn, Pirak, Timargarh etc. (vide fig. 12.2 taken from Prakash and Tripathi, 1986) for being the earliest Iron Age Sites in India, we are left with the dates of Atrañjikhera (1155 BC) and Tadakanhalli (12th century BC). Until new data force us to review our stand, we may propose that the advent of iron took place in India indigenously and independently in at least two nuclear zones: Rajasthan-Uttar Pradesh area in the north and Karnataka area in the south. We do not preclude independent contribution from the West Bengal-Bihar area in the east.

v1.c20.s3.t2.2

Regarding iron age in north India, one of the most crucial sites is that of Atrañjikhera which we have discussed in chapter 11. The iron-bearing layer, C-14 dated 1155 ± 110 BC (sample TF-191) is associated with the PGW phase which was preceded at Atrañjikhera by OCP phase (up to 1500 BC) and B&RW (black and red ware) phase (1450-1200 BC). The B&RW potteries resembled those of the Ahar and Gilund complexes. It is quite probable that around 1400-1300 BC, a branch of the BRW people left Ahar region, which is not far away, crossed Chambal and settled at Atrañjikhera and then Noh. (Gaur, 1983: 20) The copper-smiths in their group might have stumbled on to the discovery of illdefined wrought iron as indicated by the presence of a few indeterminate objects which have not been so far analysed. In the PGW phase also, fourteen indeterminate objects were obtained along with iron slag pieces and 135 iron objects of 14 varieties. The indeterminate objects could be the results of first few smelting experiments, and if so, the credit for discovering iron at Atrañjikhera area should go to the B&RW people, rather than the PGW culture.

v1.c20.s3.t2.3

We have similar situations at Noh and Jodhpura. Noh provided a C-14 date of 900 ± 225 BC. (UCLA-703 B sample) for its PGW phase and two bits of iron were obtained in the preceding B&RW phase. At Jodhpura, iron slag was found from the black-and-red ware level, and crucibles and hearth in the PGW level. Thus we may say that the rudiments of iron technology were evolved in different sites around Rajasthan and western U.P. during the B&RW era. The early experiments for producing iron might have been performed by the copper-smiths from the Rajasthan sites of Ahar, Noh and Jodhpura.

v1.c20.s3.t2.4

A remarkable evidence of iron smelting has come from Jodhpura in Rajasthan. (Agrawal, 1984: 255) A crucible-shaped furnace used for direct reduction of ore was discovered. Perhaps the bloom was heated in an open furnace and was forged on an adjacent platform. Some lumps of ore were also discovered. Such high grade ore is known from the nearby mines of Bania Ka Bas, Morija, Rampura etc. in Jaipur district. For Atrañjikhera iron, ore could have been obtained from the Jaipur area or the rich haematite deposits around Mirzapur with which the cave drawings were executed.

v1.c20.s3.t2.5

While PGW is normally associated with iron in north India (Tripathi, 1976: 99-106), it should be noted that PGW was primarily a de luxe ware arising amidst a larger proportion of red ware along with grey, black slipped and B&RW, traditions of which persisted. The PGW tradition continued from 1200-400 BC while the more sophisticated NBPW was introduced around 700 BC. The sequential developments of the above-mentioned wares were not essentially linked with the developement of iron technology; these traditions varied from site to site. For example, the PGW was found in non-iron Late-Harappan, non-iron late Chalcolithic, early Iron Age and late phase of Iron Age contexts. Similarly the NBP ware era continued from 700 BC to the Kuṣāṇa period and witnessed several phases of the Iron Age.

v1.c20.s3.t2.6

Thus, the dating of sites cannot be done in terms of the appearances of PGW and NBPW which took place at different time in different sites. Tandon (1967-68) believes that at Alamgirpur iron appeared around 800 BC even though the PGW layer in the site has been C-14 dated as 1025 BC. Allahpur in the Meerut district has yielded a rich collection of iron consisting of spearheads, arrow-heads, blades, rods, pins and nails, along with slags of iron. The PGW layers in Śrāvastī, Kauśāmbī, Hastināpur, Bairat, Atrañjikhera etc. have yielded rich varities of iron objects. Tripathi (1976: 101-02) has noticed a gradual evolution of the iron tool-types in the PGW/NBPW layers over centuries. Hunting tools were the earliest types to be evolved: spearheads, arrowheads, points, socketed tangs, blades etc. Then agricultural tools such as sickle (Atrañjikhera) and axe (Noh) were evolved. Later, household objects such as knives or tongs (Atrañjikhera) and lastly building materials such as rods, pins and nails were manufactured and used.

v1.c20.s3.t2.7

There is a general correlation between the evolution of iron tool-types and material prosperity in the ancient north India. (Tripathi, 1990) Forests could be cleared, wild animals tackled and with the tools for digging wells, the congested sites could be moved far away from the rivers. Improved spades and ploughshares brought in varieties of agricultural crops. Surplus production led to the emergence of large sites like Jakhera, more sites in the Kanpur area and cities like Vārāṇasī, Kauśāmbī etc. The second urbanisation in India inevitably led to mahājanapadas and empires, and evolution of more sophisticated iron weapons for warfare.

C. B&RW Culture Iron in the Eastern India

v1.c20.s3.t3.1

The chalcolithic traditions in (a) the Pre-, Mature-, and Post-Harappan (chapters 2-7), (b) Peninsular India (chapter 8) and (c) Eastern India (chapter 9) gradually gave rise to the Iron Age cultures in the respective areas. In chapter 9, we mentioned about the chalcolithic culture of Pāṇḍu Rājār Dhibi and the neighbouring areas of West Bengal and Bihar. This culture C-14 dated 1380 to 855 BC (Allchins, 1982: 261) spanned the period from the early chalcolithic to the advent of iron. The people there were in touch with the Copper Hoard tradition in the Doab and the Peninsular Chalcolithic tradition. In West Bengal and Bihar, endowed with high grade copper as well as iron ores, the Iron Age emerged from the chalcolithic period without any interruption; the well-dated period 1050-950 BC has been termed as ‘Ferro-chalcolithic’ by Chattopadhyaya and De (1989).

v1.c20.s3.t3.2

The Black and Red Ware (B&RW) has very distinct local characteristics in the areas of Uttar Pradesh, Rajasthan and Central India. We are concerned with the B&RW culture of West Bengal and Bihar which clearly shows the advent of Iron Age during the period 1200-800 BC. (Chakrabarti, 1976, 1985)

v1.c20.s3.t3.3

At the West Bengal site of Bahiri (Chakrabarti and Hasan, 1982) there is iron slag right on the natural soil C-14 dated 950 ± 150 BC. Hatikira village in Birbhum district has shown 1000 BC Iron Age culture. At Mahisadal in West Bengal, Iron Age starts right in the chalcolithic column dated around 800 BC. Iron slags, ingots and tools like arrowhead, point, spearhead have been obtained from the lowest chalcolithic strata Period I (1100 BC) of Mangalkot (in the Katwa sub-division of Burdwan district), 30 km south-east of Pāṇḍu Rājār Dhibi. (Ray, 1990) This shows that West Bengal acquired the knowledge of iron technology independently, and almost simultaneously as the PGW people of Atrañjikhera.

v1.c20.s3.t3.4

At Barudih in Singhbhum, iron object has been found which occurs in an otherwise neolithic assemblage, characterised by a profusion of celts. Four radio-carbon dates have been obtained, the earliest being 1200 ± 210 BC (TF 1100). Chattopadhyaya and Ghosh (1982) have analysed a sample of carburized iron clearly showing the presence of pearlite, from coarse granular to lamellar, at places in equiaxed ferrite matrix. The normalized structure clearly shows that the implement was thoroughly forged and cooled slowly in air. We may recall that the iron samples from Atrañjikhera (chapter 11) of the same age also show carburization and pearlite structure.

v1.c20.s3.t3.5

In chapter 9 we discussed the ferro-chalcolithic culture of Pāṇḍu Rājār Dhibi (1000-950 BC), a key site in the Ajoy Valley, district Burdwan, West Bengal. The site was excavated in 1964 and again in 1984-85. Specific C-14 dates have been obtained for the ferro-chalcolithic phase: 1012 ± 120 BC and 1045 ± 55 BC.

v1.c20.s3.t3.6

Chattopadhyaya and De (1989) examined a dish-shaped iron object (length 62 mm, width 18 mm and thickness 2 mm) obtained from the ferro-chalcolithic phase of Pāṇḍu Rājār Dhibi. Observation of the specimen in computer automated electron probe micro-analyser (EPMA) revealed that the dish-shaped object was made of iron with a high amount of fayalite (iron silicate) inclusion which made it brittle. No carbide or pearlitic structure was detected. Although carburization did not take place, the presence of substantial amount of potassium in the inclusion indicated that the firing may have been done with charcoal.

v1.c20.s3.t3.7

The slag found in the chalcolithic level (1250-1050 BC) of the above-mentioned site contained 34 per cent magnetite, 6 per cent wüstite, 48 per cent fayalite, 7 per cent anorthite (calcium aluminium silicate) and 5 per cent glass phase. The microstructure reveals that dendritic wüstite (lighter) phase occurs in a fayalite matrix. This kind of slag is similar to ‘Type A’ as described by Tylecote and Wynne. The substantial amounts of slag inclusion in the metal, and high proportions of unreduced oxides in the slag (a few slag samples were strongly magnetic and had 50 per cent magnetite) suggest that smelting could have been conducted below 1100°C. Traces of alpha-iron and hematite (altered from magnetite) were also noticed. Chattopadhyaya and De (1989) suggested that the original charge could have been magnetite ore; however, this need not have been the case. The sequence Fe₂O₃ (hematite) → Fe₃O₄ (magnetite) → FeO (wüstite)

v1.c20.s3.t3.8

→ α Fe (iron) represents the successive stages of reduction, and the slag under consideration shows the results of the earliest efforts, only partly successful, to obtain iron from hematite, commonly available in the West Bengal-Bihar area.

v1.c20.s3.t3.9

Matured Iron Age had existed around eighth century BC in Pāṇḍu Rājār Dhibi; the hardening and quenching methods were observed (Chattopadhyaya and De, 1989) in a third century BC sickle, which is so far one of the earliest Indian specimens exhibiting the steeling process. The length of the sickle is 255 mm and the maxium width 22 mm. The internal curved region of the agricultural tool has a sharp cutting edge. Chemical analysis revealed that it is a low carbon steel: carbon 0.22 per cent, copper 0.15, nickel 0.13, cobalt 0.11, rest iron. The metallographic structure shows that of tempered martensite. The structure also exhibits retained acicularity at certain places.

v1.c20.s3.t3.10

Electron micrograph obtained at a magnification of 100 X clearly represents the tempered martensitic structure of the sample. Microhardness measurement showed a value of VHN 391 in the ferritic region, and 549 in the martensitic region. It is evident that the smiths of Pāṇḍu Rājār Dhibi, during third century BC, knew how to forge an iron sample at high temperature to exclude slag particles, how to carburize at red heat, then quench (quickly cool possibly in water), and then finally temper or reheat it to an intermediate temperature.

v1.c20.s3.t3.11

Earlier, Chattopadhyaya (1984a) had studied a second century BC sample of iron razor from Ror, district Kangra, Himachal Pradesh which showed, apart from massive ferrite, martensite plates indicating the technique of rapid quenching. The Pāṇḍu Rājār Dhibi evidence of low carbon steel has pushed the Indian antiquity of quenched steel further back to third century BC. The Period III in Pāṇḍu Rājār Dhibi, which yielded the sickle, also exhibited a furnace probably for smelting iron and also one iron-smelting crucible.

v1.c20.s3.t3.12

The Eastern tradition of smelting copper and iron continued well up to the Early Historic Period 450 BC-AD 400 (Chattopadhyaya and De, 1989: 33), and even beyond. The chronology of the traditions is yet to be fully discovered. At several stages, the industrial operations might have been conducted by special tribes.

v1.c20.s3.t3.13

S.C. Ray discovered at Saradkel, 50 km from Ranchi and 9 km from Kunti, an Asura site of Iron Age settlement tentatively dated first-second century AD (Banerjee, 1965: 220-22 and Ghosh, 1989 Vol. 2: 393) A large number of iron objects such as arrowheads, axes with single or double cutting edges, chisels, nails, caltrops, knives and longitudinal ploughshares bespeak the prevalence of a full-fledged iron industry. Also found were hematitic ores, slags and sections of several vertical pits laid into the earth, one metre uniform width, 2-3 metre in depth, described as smelting furnaces. These present a contrast to the cylindrical or cubical clay-built shafts of the Agarias. The presence of punch-marked coin, beads of chalcedony, agate and carnelian, copper objects etc. suggest a complex of the Kuṣāṇa era. It is not clear how and when this Asura culture was replaced by the Muṇḍa tribes. The Muṇḍas continued making wrought iron which the Sanskritists called muṇḍaloha (the variety made by the Muṇḍas). The topic of tribal metallurgical practices, continued right up to the pre-British period, is beyond the purview of our discussion.

v1.c20.s3.t3.14

West Bengal and Bihar definitely constituted one of the nuclear zones for the ancient metallurgical traditions in India. This zone contributed towards the emergence of the second urbanized culture (after the Harappan) and the political power in northern India during the Maurya-Gupta eras. Just as the north-western post-Harappan chalcolithic tradition was followed by the Iron Age in Rajasthan/Western Uttar Pradesh area, and the Copper Hoard and Eastern Chalcolithic traditions gave rise to Iron Age in East India, we witness the third nuclear zone in the Peninsular India where the iron-making tradition was rooted more in the neolithic rather than the chalcolithic era.

D & E. Iron Age in the Megalithic Cultures of the Peninsular and Central India

v1.c20.s3.t4.1

South India evolved a mega (big) lithic (stone) culture in which the burials were surrounded by a circle of stone boulders; this culture extending gradually to the Central India and even trans-Vindhyan areas, exhibited early traditions of the use of iron.

v1.c20.s3.t4.2

In 1823, Babington discovered a rock-cut burial chamber in Malabar. Since then, 1400 megalithic sites have been brought to light — 1116 in the region of Peninsular India. (Deo, 1985) The largest number of megalith monuments have been found in Tamil Nadu (388), followed by Karnataka (300), Kerala (188), Andhra Pradesh (147), Maharashtra (90) and Pondicherry (3). There are some megalithic sites north of Vindhya range such as in the Mirzapur district. Apart from the megalithic tombs and graves, we have the neighbouring habitation sites. The contemporaneity of iron age habitation and the burials have been proved at least in some cases, as in Brahmagiri and Maski. (Nagaraja Rao, 1971 and 1981) .

v1.c20.s3.t4.3

The megalithic burials may be broadly classified into three groups: the chamber tombs, the unchambered graves and monuments not actually associated with burials. There are many sub-groups such as (a) urn burial, (b) sarchophagus burial, (c) pit burial, (d) dolmen or table-stone-chamber raised from above the ground, (e) passage chambers of north Karnataka, (f) port-hole chambers (g) topikal or hat-stone in the lateritic area of Kerala, (h) kudaikal or umbrella-stone at Kolappuram, Koduvayur etc., (i) menhirs or long-stone in Kerala and Andhra Pradesh, (j) rock-cut chambers of Kerala (k) cists or round receptacles stones in the quartzite-sandstone region of Andhra-Karnataka, (l) stone circle with cairn filling in Vidarbha region and so on.

v1.c20.s3.t4.4

We are not sure whether the varieties as stated above are due to differences in traditions of the sub-cultures or geological diversity regarding the availability of stones or both. We should not overlook the wide range of the dates of these megalith sites. The MASCA corrected date of the Hallur site is 1200 BC. On the other end of the spectrum is the Coimbatore megalith in which Roman coins (Christian era) have been found.

v1.c20.s3.t4.5

Amidst the diversities in time and place, one commonality emerges in the megalith sites; as early as 1921, L. Munn pointed out that the megalithic sites were close to the ancient mine sites in the region. Sundara (1975) emphasized that the Kappatgudda range in South India, where there is a number of megalithic sites, “is rich in gold as well as in iron and has many gold-working pits and sites strewn with iron slags”. Very useful comments have been made by Vibha Tripathi and V.D. Misra with reference to Deo’s (1985) paper on megaliths. Tripathi commented that “the megalith-builders basically appear to have been smelters and makers of a wide variety of metallic objects, iron being the most important among them”. She felt that they were nomads and marginal agriculturists. Misra partially disagreed and pointed out that ‘the very existence of the professional groups — potters, copper-smiths, black-smiths, bead-makers etc. — presupposes surplus production, however marginal it might have been”.

v1.c20.s3.t4.6

The earliest phase of the Iron Age in the South has been revealed in the excavation at Piklihal and Hallur (1150-1030 BC) and the Karnataka sites of Tadakanhalli, Komaranhalli etc. In many of these sites, the Iron Age started almost from the neolithic stage. Further north in Central India, megalithic sites are dated somewhat later 800-700 BC, and it is possible that the Chalcolithic sites imbibed Iron Age Culture from the megalithic people.

v1.c20.s3.t4.7

Among the most common tools are flat iron axes, often with crossed iron bands for hafting. The shaft-hole axe of North India (Noh and other sites) is unknown in the Iron Age graves of South India. There are flanged spade, hoc, spud or pick axe, sickles or bill-hooks made or iron. Iron wedges or crowbars or perhaps spears of great length; knives, chisels, tripods, daggers and swords, some time with ornamented bronze hilts were meant for military purposes. Iron tridents, in one case with a wrought iron buffalo attached to the shaft, are reminiscent of (Śiva-Durgā) ritual tradition. Around Nagpur graves were found horse furniture, iron snaffle bits, cheek pieces found in the mouth of a buried horse.

v1.c20.s3.t4.8

The traditions of the Iron Age as stated above were evolved over centuries and across a large section of the sub-continent; these definitely started in southernmost India, but the origin of the culture is shrouded in mystery. There are unfounded and wild speculations that the culture had diffused from Egypt, Arab countries or the North-West, people coming to South India by boat (Tripathi, 1973: 274; Allchins, 1982: 340-43) and so on. However, we would support the notion of indigenous and independent development of Iron Age in South India unless new facts prove the contrary.

Early Iron Age in the Hallur Area

v1.c20.s3.t5.1

Hallur is in the district of Dharwar on the left bank of Tungabhadra, anciently known as Hallavaru. In this area, the Neolithic Period I (1710-1030-100 BC) overlapped with Iron Age starting around 1200-1100 BC with a very short iron-free chalcolithic period. The megalithic iron age was characterised by typical B&RW, varieties of iron tools, megalithic cairns and cists, bone, carnelian, terracotta, gold, green stone and antler beads. (Nagaraja Rao, 1971 and 1981)

v1.c20.s3.t5.2

The Hallur area provides one of the three earliest dates for Iron Age Culture in India (Atrañjikhera and Pāṇḍu Rājār Dhibi in North India providing the other two). In chapter 10, we referred to some of the ancient mines of copper in the Karnataka area which must have witnessed very early open cast iron ore mining also. We would like to record the technical details of a few iron objects found from Komaranhalli and Tadakanhalli. Komaranhalli is one of the important megalithic burial sites 12 km east of Hallur dated 1200-1100 BC by thermoluminescence determination. (Indian Archaeology — A Review 1980-81, Archaeological Survey of India, New Delhi, 1983, p. 29) Tadakanhalli is the megalithic site dated 1000 BC, near (and contemporary to) Hallur site. (Nagaraja Rao, 1981: 25) The National Research Laboratory for Conservation of Cultural Property (NRLCCP), Lucknow, India received 31 objects from Komaranhalli (arrowheads, axes, knives, daggers, swords, spearheads, bars and rods etc.) and 20 from Tadakanhalli (above types and also chisels) for metallographic examination. Several samples were found to be heavily corroded; investigations were carried out on the rest which had some residual metal. (Agrawal, O.P., 1980-81 and 1990) A Komaranhalli spear and an axe from Tadakanhalli were examined with great detail, the results of which are presented below.

v1.c20.s3.t5.3

Both the specimens revealed laminated structure. Two sections were drawn from the blade and stem portion of the Komaranhalli spear measuring 178 cm in length and 3 cm in breadth. The blade showed a small metal core (the rest corroded) in three layers, two carburized and one uncarburized iron. While the light-coloured layer was ferritic and had low carbon (0.05%) and low hardness (HV 105), the darker carburized layer was of hypoeutectoid steel with dispersed pearlitic structure and had higher carbon content (0.2 to 0.24%) and higher microhardness. (HV 125-35) The section from the stem also revealed an air cooled dispersed pearlitic structure. Agrawal *et al.* (1990) concluded that the spear must have been prepared by hammering and welding together the sheets of carburized iron. The blade of the spear after receiving the final shaping at high temperature was also worked below the critical temperature 730°C and cooled in air. There is no evidence of hardening. The examination of other artefacts from Komaranhalli (1200-1100 BC) such as spearheads, daggers and swords also showed the evidence of lamination technique.

v1.c20.s3.t5.4

Similarly the cutting edge of the Tadakanhalli axe revealed a banded structure caused by three carburized and pearlitic layers (hardness HV 120-130 and carbon 0.2-0.3 per cent) and four uncarburized and ferritic layers (HV 98-109 and carbon 0.04-0.05 per cent). A close examination of the ferritic microstructure in the uncarburized layer indicated that the diffusion of carbon was only along the grain boundaries. This probably took place around 700-800°C. The deformed state of the lamellar pearlitic structure suggests that the axe was forged below the critical temperature (around 650°C) and cooled in air. In their first communication, Agrawal *et al.* (1980-81: 99) claimed the presence of bainite and martensite in the Tadakanhalli axe which was suggestive of quench-hardening. But Chattopadhyaya (1984b: 364) pointed out that “from the published photomicrograph nothing can be ascertained about the presence of martensite or bainite”. Subsequently, Agrawal *et al.* also agreed (1990: 15) that ‘there is no indication of the Tadakanhalli axe having been hardened and tempered’.

v1.c20.s3.t5.5

Nevertheless, Agrawal *et al.* have proven beyond doubt that the metalsmiths of the megalithic period (1200-1000 BC) introduced in India the lamination (layering or faggoting) technique in which two sheets of iron were welded and forged together. One sheet was carburized and the other was not. In this way several sheets of iron and carburized iron could be piled alternately and forged into a flat slab having requisite ductility (of wrought iron) as well as hardness (of carburized iron). Later on these piles could be heated, hammered and folded during the fabrication of the artefacts. Agrawal *et al.* showed that the laminated samples examined had (a) clearly visible welding joints, (b) continuous chain of slag inclusions, (c) sudden changes in concentration of carbon, and (d) sudden changes in the orientation of elongated slag inclusions. These evidences confirm that lamination had indeed been done.

v1.c20.s3.t5.6

Earlier in this chapter, we have quoted Maddin *et al.* (1977) reporting ‘the earliest known laminated iron object’ which is an Egyptian knife dated 900-800

v1.c20.s3.t5.7

BC. This date is much later than that of the Hallur site laminated weapons and we believe that it is quite possible that the technology diffused westward from India.

v1.c20.s3.t5.8

Agrawal et al. (1990) further showed that the lamination techniques was in vogue in India 1200 BC-AD 600. This is evident from the diversity of samples from Jājamau (600 BC-AD 600), Hulaskhera (700 BC-AD 700), Sringverpura (250 BC-AD 200), Kauśāmbī, PGW AD 500 etc. which were examined by them.

v1.c20.s3.t5.9

Jājamau (near Kanpur city) yielded 210 iron artefacts. Many of them were laminated structures. The cutting edge of an iron axe had a low carbon layer (0.2-0.24% carbon) sandwiched by two high carbon layers having carbon content as high as 0.6-0.65 per cent in some places. The low carbon layer had hardness HV 125 and the high carbon layer HV 267. The latter had a ‘tempered structure’ near the cutting edge having hardness HV 280. This suggests that the axe in the final stages was quenched, tempered and allowed to cool in air. It is not clear whether the axe had martensite in the cutting edges and whether the date of the axe could be 600 BC or later, since the site is dated 600 BC to AD 600.

Megalithic Culture of Central India

v1.c20.s3.t6.1

Iron appears in the megalithic burials of the Vidarbha region in Central India, with B&RW which is quite different from that found in the north-east. The region of Vidarbha has 72 of the 83 Megalithic sites in Maharashtra exhibiting only one kind of megalithic monument viz., the stone circle with cairn filling.

v1.c20.s3.t6.2

A large majority of the Vidarbha sites are in the north and south-west of Nagpur and within a circle of 50-60 km radius around the modern city. Thus what we witness here is a highly compact and homogeneous megalithic culture dated around 800-700 BC. Takalghat, Khapa and Gangapur are in south-west, Mahurjhari, Junapani, Borgaon in north-west, and Naikund in north-east direction relative to Nagpur. Kaundinyapur site is near Amaravati.

v1.c20.s3.t6.3

The megalithic people around Mahurjhari practised a mixed subsistence economy, and thier diet included agricultural products as well as hunted game animals. In the entire area, real agricultural tools like sickle and ploughshare were rare. At Borgaon, the tools of offence like spikes, axes, trident and other objects account for nearly 77 per cent of the total collection. At Takalghat, 28 out of 76 iron implements (37%) are offence tools like axe, arrowhead, dagger, blade, sword, spear, knife etc. We have the same pattern in the peninsular megaliths: Brahmagiri shows 28 out of 44 (63%) as offence tools. Many items were ‘socio-technic’ — related to the status or rank of a person in a given social set-up. These indicated specifc status; for example only one out of four burials at Borgaon showed horse-bit of iron and trident. At Takalghat and Mahurjhari were found hose-trappings, ornaments, dagger with iron blade and copper hilt etc. The culture reflected nomadic pastoralism, emphasis on warfare using horses, ritual tradition for sacrificing animals, and a strong tradition for metallurgy.

v1.c20.s3.t6.4

Whereas the Vidarbha Megalith culture shows less use of tin in bronze items, the peninsular counterparts show more lavish use of the metal: Brahmagiri sample 16 per cent tin in bronze, Raigir, Andhra Pradesh sample 21 per cent, Nilgiri area bronze 30 per cent etc. At Mahurjhari and Takalghat 15-30 per cent of the metallic items in the burials were made of copper. The number of copper bronze artefacts used by the Megalithic people is much higher than that associated with the full-fledged chalcolithic Jorwe culture in Western Maharashtra. The Megalith traditions in gold are equally impressive. Takalghat and Mahurjhari show a high percentage of silver (51.2%) in gold items. The gold diadems of Adichanallur are the most classic specimens of skilled technology. Paiyampalli provides the evidence of a goldsmith’s stone mould. The Megalithic people of Maski must have obtained gold from the nearby Hutti mine.

Iron in the Vidarbha Megalithic Culture

v1.c20.s3.t7.1

Thus it is evident that the Iron Age in the Vidarbha area emerged in the framework of a well-established chalcolithic culture and strong metallurgical (tin, gold, silver etc.) traditions. We do not know whether the metallurgical traditions in the Vidarbha area were inherited from the earlier Megalithic culture of the Peninsular region in the south.

First 95 of 244 passages. Read on