History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c17.s4: D. On Some Metalliferous ores in Ancient India
History of Technology in India, Vol. I: From Antiquity to c. 1200 A.D. v1.c17.s4
- Language
- English
Contents
- v1.c17.s4.t1 The Issue of Ore-Artifact Correlation — 9 parts
- v1.c17.s4.t2 Some Observations on the Metallic Slags in Ancient India — 22 parts
D. On Some Metalliferous ores in Ancient India
v1.c17.s4.1
A large number of metallic artifacts have been unearthed from the archaeologically excavated sites in India. Some of them might have been imported; circumstantial evidences suggest however that most of them were made indigenously from their respective ores in the sub-continent.
v1.c17.s4.2
Table 1 lists some of the C-14 dated ancient mines in India starting from the earliest (so far recorded) 1260 B.C. Rajpura- Dariba in Rajasthan, a mining area which yielded complex ores containing, lead silver, zinc and possibly copper also. The table lists other dated mines which produced metalliferous ores containing Cu, Zn, Pb, Ag, Au etc. Elsewhere, we have listed 38 other mines of copper, definitely ancient though undated (Biswas, A.K., 1991). The chalcolithic cultures of the Mature-and Post- Harappan era produced bronze objects containing tin, the source of which has not been precisely determined. Some tin ore could have been imported from the well-known mines of Afghanistan and Iran during the Harappan era, and again later during the prosperity of Taxila (Chakrabarti, 1979).
v1.c17.s4.3
Some tin ore could have been mined indigenously from Rajasthan, Karnataka and Bihar (Brown and Dey, 1955: 167-68). Cassiterite, SnO₂ is often found in the form of water- concentrated placer deposits, referred to as ‘stream tin’. Tylecote (1962: 63 quoted by Hegde) suggested that the ancient metallurgists might have exploited (and exhausted) placer deposits of ‘stream tin’ rather than vein deposits which are hard to mine. Such surface accumulations in the proximity of the Aravalli and Chota Nagpur Hills might have been amongst the sources of tin in ancient India.
v1.c17.s4.4
Similarly, there are several possible sources of lead and silver. The Harappans were acquainted with lead ores such as galena (sulphide) and cerussite (carbonate), and their smelting process which is simple on account of easy reducibility and low melting point of the metal.
v1.c17.s4.5
At Mohenjo-daro, six percent of the tools contained lead; a fraction of the tools contained arsenic which was evidently derived from lollingite, samples of which have been found in the Harappan sites.
v1.c17.s4.6
In view of the fact that silver artifacts from the Mature-and Post-Harappan sites as well as Taxila contained lead and trace of sulphur, it may be surmised that the source was argentiferous galena ore. A few samples of silver in the Harappan sites and the silver plates of Gungeria, showing some gold but no lead, were probably derived from the South Indian electrum deposits.
v1.c17.s4.7
We have provided the C-14 date of 760 B.C. for the Hatti gold field in Table 1. There is no earlier date for gold-mining in India than this. Probably the bulk of gold during the Harappan and even the Historical Periods in India was derived from alluvial washings, a practice widely observed as late as the nineteenth century (Ball, 1880). Table 1 contains several C-14 dates for the zinc-lead ore mines in Rajasthan such as Rajpura- Dariba, Rampura-Agucha, Zawarmala etc.
The Issue of Ore-Artifact Correlation
v1.c17.s4.t1.1
In pursuing the laudable objective of establishing that the metals were produced indigenously in ancient India, our archaeologists have sought correlation between the artifacts and ores in the subcontinent in terms of trace metal impurities. Most of the enthusiastic investigators ignored the cautionary remarks of Thomson (1958) pointing out the hazards in ore-correlations, and occasionally reported 'positive correlation'.
v1.c17.s4.t1.2
We have pointed out (Biswas, A.K. et al, 1990: 50-51) that the correlation depends upon: (a) the type of ore, oxide or sulphide or silicate, the nature of impurities, (b) the precise conditions of smelting, such as temperature, nature of fuel, flux etc., and several other factors.
v1.c17.s4.t1.3
The ore types may very even within a region of small area. Hegde (1968) failed to detect the impurity elements of Cr, Co, Ni, Bi and Mo (present in the Devnimori silver coin) in the nearby Zawar argentiferous galena ore and concluded that the silver was 'imported'. Shekar (1983) however found them in the Ambamata deposit only a hundred kilometres away!
v1.c17.s4.t1.4
So far as the occurrence of metallic artifacts in the ancient sites are concerned, the existence or absence of mines in the neighbourhood areas in the modern age hardly establishes any truth. On the one hand, the currently known mines might have been unknown to the ancient people, and on the other, some small near-surface reserves might have been thoroughly exploited by the ancient miners, so that we do not know them at present. Certain impurities such as arsenic might have been eliminated through certain simple pre-smelting operations such as roasting. Thus we are not sure whether arsenic-free copper ingots of Lothal were made at Oman as claimed by S.R. Rao (1979 & 1985) or made out of Indian arsenic-bearing ore after a pre-smelting roast.
v1.c17.s4.t1.5
Having stated the hazards of the ore-metal correlation, let us report some of the 'positive correlations' obtained by our archaeo-metallurgists. Friedman et al (1966) claimed that with regard to copper artifacts, the elements of Ag, As, Bi, Fe, Sb and Pb were the most important metallic impurities in relating the metal back to the original type of ore. Following this argument, Agrawal (1971, 1984) suggested that Chanhu-daro celt and Mohenjo-daro spearhead were made from Khetri chalcopyrite ore, since all the three materials contained trace impurities of Sb, Pb, but not Fe and As.
v1.c17.s4.t1.6
Hegde (1969) compared the artifacts from Ahar with the chalcopyrite ore from Khetri, and the similarities led him to conclude that 'copper metal of the Ahar artifacts was probably smelted from the ore of the Aravalli region'. The slag dumps and neighbouring mines indeed suggest that copper was made at Ahar indigenously, but Hegde's argument is unsound. 'The Aravalli region' alluded to by Hegde corresponds to a vast area and ores of more than one type; it is hard to believe that chalcopyrite ore was brought all the way from Khetri to Ahar over a distance of hundreds of kilometres while there were ores in the vicinity.
v1.c17.s4.t1.7
Hegde (1985, 1989) has himself modified his earlier view and suggested that the Ahar artifacts were possibly made from the nearby region of the Aravalli which constitutes of not-too-hard tremolite-chlorite schists rather than the far-off Khetri ore embedded in quartzite rocks, hard and less easily extractable. While impurity patterns of the ores in different parts of the Aravalli 'may be similar' (the previously stated Ahar-Khetri correlation now loses much of its significance), the lead isotope ratios markedly differ in the north-eastern, central, and south- western parts of Aravalli (Hegde, 1985). Efforts are in progress to correlate metal artifacts and specific ore deposits through lead isotope ratio analysis.
v1.c17.s4.t1.8
Further cautioning is necessary in this regard. The ancient smelting process, which is not definitely known at present, would certainly have considerable effect of the trace element profile as proven by smelting experiments (Tylecote et al, 1977). Lead isotope correlation between ore and artifact has been successful in one case (Gale et al, 1982). Reporting failure in another case, Craddock (1985: 59) has rightly questioned that lead in the artifact could come not only from the copper ore (as presumed) but also from the flux and deliberate addition of lead during smelting.
v1.c17.s4.t1.9
We do not discourage attempts for ore-artifact correlation, but these must be supplemented with additional evidences for (a) mines of antiquity, (b) outfits for smelting, furnaces and crucible, (c) heaps of slag and slag-ore correlation regarding the non-metallic phases, and (d) a reasonably large number of artifacts at different stages of manufacture and of composition and properties varying within narrow limits (Biswas, A.K. et al, 1990: 51). Fortunately, several such evidences for indigenous metallurgy in India have been obtained from Mehargarh, Ganeshwar-Jodhpura, Lothal, Ahar, Khetri (for copper), Chanhu-daro and Lothal (for beads and seals), Atranjikhera, Ujjain (for iron), Zawar (zinc, lead and silver) etc.
Some Observations on the Metallic Slags in Ancient India
v1.c17.s4.t2.1
There are heaps of ancient slag in the excavated sites which conclusively prove indigenous smelting in India. A detailed examination of the slags and slag-ore correlation as suggested above (c) should help us in ascertaining the kind of ores and fluxes which might have been used during the process of smelting.
v1.c17.s4.t2.2
During the excavation of Period I levels at Ahar, heaps of slag-like material were recovered from more than one trench at three different depths. Evidently the copper smelting industry at Ahar was extensive. The analysis of the slag is:
v1.c17.s4.t2.3
| SiO2 | 35.2-38.2 p.c | MnO2 | 1.2-2.3 |
| Fe2O3 + FeO | 43.9-48.3 | CuO | 0.7-0.9 |
| MgO | 2.4-3.6 | SO3 | 3.6-6.1 |
| Al2O3 | 5.3-7.8 | | |
v1.c17.s4.t2.4
Evidently, complete removal of sulphur and recovery of copper were not possible, and the ore was sulphidic. Copper ore from the neighbouring area contains less than 20 p.c. silica and therefore it may be presumed that the copper ore was fluxed with silica during smelting (Hegde, 1969).
v1.c17.s4.t2.5
It is unfortunate that phase studies were not conducted on the Ahar slag. The modern Khetri Copper Plant produces slag which contains phases such as magnetite Fe3O4 and copper ferrites such as CuFeO2, Cu6Fe3O7 and some free copper oxide (Sharma and Shekhawat, 1984). Similar phase studies on the ancient slag would have shown whether (a) such a slag could have been derived from the neighbouring ore containing specific gangue meterials and (b) if so, what temperature and other physico-chemical conditions might have been employed.
v1.c17.s4.t2.6
Atranjikhera showed not only iron objects but also furnace and iron slags assaying 64-67 p.c. Fe2O3, 25-30 p.c. SiO2, 3-5 p.c. Al2O3, 1-3 p.c. CaO, 0.6-0.9 p.c. MgO and 0.7 p.c. P2O5 (Gaur, 1983: 496).
v1.c17.s4.t2.7
It has been proposed that iron could have been made from the iron-bearing shales occurring in the neighbouring Agra-Gwalior range which assays 21.7 p.c. Fe, 41.9 p.c. SiO2, 14 p.c. Al2O3, 6.3 CaO. 2.3 MgO, 1.4 Na2O and 6.8 p.c K2O. This suggestion (Gaur, 1983: 487-500) is plainly unacceptable, since the proportions of SiO2, Al2O3, CaO and MgO are lower in the slags, and iron percentage much higher, than in the iron shale. The raw material must have been rich iron ore (from eastern Uttar Pradesh or Bihar) out of which the iron recovery was incomplete and inefficient on account of solid-state reduction.
v1.c17.s4.t2.8
Excavation by Willies et al (1987, 1989) has revealed ancient mining of the Rampura-Agucha deposit (25° 50' N, 74° 44'15" E) in Rajasthan. There is the clear base of a gossan at about 15 m. depth from which marmatite, pyrite and sphalerite (ZnS) were selectively oxidised and leached by the natural agencies leaving galena-cerussite-anglesite ore-body enriched in terms of silver.
v1.c17.s4.t2.9
The ancient mining timbers have been C-14 dated as 370 and 250 B.C. Tiwari et al (1984:85) have proved that here mining done at the 80-125m. depth was selective following lead mineralisation and pure galena veins in preference to the zinc ore veins.
v1.c17.s4.t2.10
There are huge slag dumps in the Rampura-Agucha area. The petrographic thin sections of slag sample indicate the presence of quartz and feldspar grains which have not undergone much change. Evidently, low-temperature smelting was done which could have yielded argentiferous metallic lead; the main target was possibly silver.
v1.c17.s4.t2.11
The slag dump area also showed several cylindrical retort-like pieces with inside coating of dirty white ZnSO4 and slag assaying 0.01 p.c. zinc and as high as 9.30 p.c. lead. Considering the fact that the ore-body assays 12p.c. Zn and 2-3 p.c.Pb, Paliwal et al (1986) rightly inferred that an appreciable amount of zinc must have been recovered by retort distillation process which was perfected in the Zawar Mines.
v1.c17.s4.t2.12
At Agnigundala, Andhra Pradesh, extensive mining, mineral processing (crushing, grinding and beneficiation) and smelting (of copper and lead) took place in ancient India which is evident from the rock dumps, slag heaps, cupolas and crucibles (Ziauddin, 1961: 120). The ore must have been complex and multi-metallic. Similarly, selective mining of the Ambamata deposit might have resulted in recoveries of copper, lead and silver, and zinc which was strikingly rich in cadmium (Shekar, 1983).
v1.c17.s4.t2.13
One of the most spectacular discoveries during the recent eighties related to mining archaeology and archae-metallurgy has been the reconstruction of the glory of zinc ore mining and smelting in ancient India (Biswas et al, 1984; Craddock et al, 1985; Willies, 1987 & 1989; Hegde, 1989 etc.). Limited space does not permit us to present the full report which is available elsewhere (Biswas, A.K., 1991). Only a few brief comments may be made here with reference to the 13th century A.D. ancient slags in Zawar mine, Rajasthan.
v1.c17.s4.t2.14
There are at least sixteen deposits of ancient slag around the Zawar Mine, which are basically of two types; (a) glassy and (b) crystalline. The first variety resulted on account of smelting and recovery of lead and silver but not zinc.
v1.c17.s4.t2.15
At a later stage, the value of zinc was realised, sphalerite removed carefully from the lead ores and reduced/distilled in retorts leaving some residual slag. Pyrite used as flux precipitated in the crystalline slag as fayalite (Freestone et al, 1987).
v1.c17.s4.t2.16
Detailed X-ray and electron diffraction studies were conducted on the Zawar retort residue (content), wall and (lead) slag (Biswas, et, al, 1984). The phases identified are reported in Table 6.
v1.c17.s4.t2.17
It is evident from the data that a small part of sphalerite was not converted into oxide and remained in the retort as ZnS and ZnSO4. The presence of goslarite or ZnSO4, 7H2O (hydrated at later stage) was confirmed by the endothermic DTA peaks apart from X-ray and electron diffraction studies. Some ZnO was converted to complex silicates during roasting and some underwent atmospheric conversion at a later stage to the basic carbonates (vide Table 6). Trace element detection in the zinc residue/slag by the Energy Dispersive X-ray Micro-Analysis or EDAX in the Scanning Electron Microscope showed the existence of elements such as Na, K, S, Ti etc., which must have originated from the ore-body charged into the retort.
v1.c17.s4.t2.18
The above-mentioned work on zinc-lead slags illustrates that valuable information may be obtained by sophisticated techniques regarding the nature of
v1.c17.s4.t2.19
Table 6 — Phases identified in Zawar Retort Residue (Content), Wall and (Lead) Slag by X-Ray and Electron Diffraction Studies (Biswas et al 1984: 236).
v1.c17.s4.t2.20
| | Phases identified | |
| *Sample* | *X-ray diffraction* | *Electron diffraction* |
| Retort content | Zn2P2O7. 5H2O goslarite ZnSO4, 7H2O hemimorphite, Zn4 (OH)2 Si2O7, H2O esperite Ca3Pb (ZnSiO4)4 Sphalerite ZnS, chalcopyrite CuFeS2 larsenite, Pb ZnSiO4 | Zn2P2O7. 5H2O goslarite hemimorphite, esperite, quartz, Mg2 Al4Si5O18, hardistonite Ca2ZnSi2O7, hydrozincite or aurichalcite, Zn5(CO3)2 (OH)6 |
| Retort wall | Goslarite, hardistonite, hemimorphite Zn2P2O7. 5H2O, aurichalcite, quartz, sphalerite | Goslarite, hemimorphite, Willemite Zn2SiO4, hardistonite |
| Slag | Quartz, goslarite, Zn2P2O7. 5H2O hemimorphite, aurichalcite, sphalerite, chalcopyrite, hardistonite. | Hemimorphite, goslarite, hydrozincite, quartz, sphalerite |
v1.c17.s4.t2.21
the ore and reduction/smelting techniques, the details of which are outside the purview of this paper.
v1.c17.s4.t2.22
Electron diffraction studies have further shown that the complex Zawar ore contains not only sphalerite (α- and β ZnS) and galena (PbS) as the major phases but also red silver, pyrite, chalcopyrite, anglesite (PbSO₄), PbCO₃, calamine or Zn₄(OH)₂Si₂O₇, H₂O etc. (Biswas, A.K., 1981). The gangue impurities are dolomite, quartz and white mica.