Arjuna Thantilage Phd

Postgraduate Institute of Archaeology

The history of use of metals in Sri Lanka is an area yet to be studied thoroughly and hence have very vague picture on the subject. But it has recently been shown that the copper metallurgy may have been played a very important role in our cultural formation process against the common belief that we have started directly from an Iron Age (Thantilage 2008). However starting from a period after the 500 AC there is a rich collection of metal sculptures and objects made of copper alloys. So far even very few attempts have made to understand the resource utilization aspect of the Sri Lankan historical metallurgy.

Historical records and chronicles had indicated that Sri Lanka used its own source of copper during historical periods (Seneviratne 1995:125). Because of the high nickel content (Maliyasena 1987) and consistent copper/nickel ratio the copper artifacts found form the Anuradhapura Gedige excavation are thought to have been made from ores from the Seruwila deposit. It is also suggested that Seruwila is the place described in Mahavamsa as Tambapitta on the basis that the location given for Tambapitta perfectly matches the present Seruwila deposit (Seneviratne 1995:125). It has been suggested that the occurrence in bronze artifacts of a particular element at an unusually high concentration may be characteristic of a particular geographic or mineralogical origin, for instance iron content in bronzes (Reedy 1997: 83-85; Craddock and Meeks 1987: 132). There were evidences visible for ancient mining work within this deposit area. In fact charcoal had been found 9 –12 m below in the drilled core samples confirming some smelting activity in the past. Further observation of strewn magnetite close to the old trenches suggests that the earlier workers had gone for some other mineral apart from magnetite (Jayawardhana and Padmasiri 1977: 3).

It is also apparent apart form the Seneviratne’s findings that no much attention has been paid so far to understand the utilization of this major metal source in historical times. This paper intends to address the resource utilization of the ancient copper metallurgy during the Anuradhapura time period in Sri Lanka. The trace element and stable lead isotope results of the twenty bronze icons (See Appendix A) belong to the Anuradhapura period were used for this study.

The Geological Survey Department has been carried out a scientific examination of the deposit by the drilling method and has found that it is a copper iron pyrites deposit. Also the trace element characteristic for this particular deposit has been obtained an extremely valuable factor in identifying the copper obtained from this particular deposit (Jayawardhana and Padmasiri 1977: 15-20). The extremely useful, cobalt/nickel elemental ratio which is used by geologists to discriminate ore sources and as reliable indicator in ore genesis (Barlia et el. 2004: 353-374; Loftus Hills and Solomon 1967: 228-242) has been calculated for the Seruwila deposit as between 0.43 – 0.55 (Jayawardhana and Padmasiri 1977: 19). The same Co/Ni ratio value has been successfully utilized to identify the possible copper ore source for some South Indian bronze images (Sirinivasan 1999(a)). The Co/Ni ratio value is also used in this study too to identify the provenance of the copper deposit used to make the bronze icons analysed.

Since copper magnetite is the main metal bearing mineral of the Seruwila deposit, iron would be the main impurity in copper metal obtained using metal ores from this source. So it was thought that the iron content would also be a good indicator of the deposit and also gave excellent results as anticipated. It was a known fact that relative concentrations of elements (i.e. elemental ratios) maintain consistency during metallurgical processes from ore to metal than direct element concentrations (Reedy 1997: 85). By plotting the absolute ratio values of elements Co/Ni Vs Fe/Ni in a scatter plot for the studied icons, it is possible to make clearly distinguishable and interesting clusters of icons (Figure 1). The analysed icons, which are having Co/Ni value within the Seruwila deposit’s value of 0.43 – 0.55 also make a separate cluster with linear trend in the scatter plot indicating that the copper may be from a separate source i.e. Seruwila deposit. If an element group has linear trend of Co/Ni ratio, which may be a characteristic of the copper ore (Sirinivasan 1999(b): 203). This study has successfully shown that Co/Ni and Fe/Ni scatter plot clearly identified icons made from the Seruwila copper.

The already identified Seruwila deposit icons where Co/Ni value ≥ 0.43 have variable Ni content of 0.1% – 0.33 wt%. It was also seen that high Ni content is also an indicator of the deposit. But exclusively Ni content it self of the images studied may not give clear identification of the source. But by comparing the Ni content of Sri Lankan bronze images having Co/Ni ³ 0.43 with the images having Co/Ni £ 0.43 (Figure 2) clearly indicate that the nickel content grater than 0.3% implies the use of only the copper from the Seruwila deposit.

The graph above shows the correlation of nickel in composition of the images made from Seruwila deposit in the studied images.

Even though the absolute Ni content in the composition is less than 0.3%, images made from the Seruwila deposit’s copper had almost maintained the Co/Ni ratio value greater than or equal to 0.43.

It can also be seen that the high iron content has some correlation with the copper from the Seruwila deposit. It is a known fact that the iron present in the copper as an impurity gives negative properties to the copper metal and also alloys made from that metal. So the refining of copper had been done in antiquity to remove the iron content in copper by remelting it in a crucible (Craddok and Meeks 1987:195). The results of the present study also indicate that some icons with low iron content have been made with Seruwila copper. Hence the iron contents itself in the composition of the copper alloys of Sri Lankan historic objects may not always be a good indicator of the Seruwila deposit.

By comparing the iron content of Sri Lankan bronze images having Co/Ni > 0.43 with the images having Co/Ni < 0.43 (Figure 3) clearly indicate that the images having Co/Ni > 0.43 contain more iron in their composition.

The above graph indicates the images having Co/Ni ≥ 0.43 often has high iron content, which implies the local Seruwila deposit’s copper. This is not true always because of the possibility of subsequent metal refining or the use of weathered secondary ores for the extraction of copper. Over 340 statues examined belong to the Himalayan region, that only 39 of them have an iron content of greater than 1%. Out of this 39, twenty-six were found in the Kashmir group. So the iron content feature may carry some technological style or use of a particular ore type (Reedy 1997: 84).

The scatter plot of stable lead isotope results of the studied Anuradhapura period bronze icons clearly separate the two clusters A and B as in the case of trace element results (Figure 4). This indicates that not only the distinct copper sources but also lead from two distinct sources were used to produce these two series of images A and B.

The high number of images in image cluster A (Seruwila copper) than the image cluster B implies the intensity of the use of Seruwila deposit is considerably higher than the use of unknown copper source used to produce images in cluster B during the Anuradhapura period.

Interestingly, when we look at the studied images which make separate clusters A and B in both trace element and the lead isotope graphs, there are identifiable differences in their technological styles as indicated below.

Due to the fact that the images belonging to both groups A and B makes separate clusters in both lead isotope and trace element scatter plots, the above technological differences clearly indicate that these two image groups must belong to two separate schools of image production that existed during the Anuradhapura period.

Acknowledgment

I deeply indebted to Prof. Senake Bandaranayake who encouraged me in this study. The lead isotope analysis was carried out under the financial support I received through the SAREC/SIDA supported Environmental Change and Human Response program conducted by Prof. Paul J.J Sinclair, Department of Archaeology and Ancient History, Uppsala University, Sweden. I wish to express my deepest gratitude to Prof. Paul J.J Sinclair for all his assistance and finding me a place to conduct the lead isotope analysis of the studied icons. I wish to thank Prof. K. Tennakoon, the Director, Institute of Fundamental Studies (IFS), Kandy , Sri Lanka for his assistance and also allowing me to use the facilities at the IFS for elemental analysis of this study. I wish also to thank Dr. Kjell Billstrome , Natural History Museum, Stockholm, Sweden who conducted the lead isotope analysis of the metal samples.

References:

Bralia, A., Sabatini, G., Troja, F.  2004. A Revaluation of the Co/Ni ratio in pyrite as Geochemical tool in Ore Genesis Problems, Mineralium Deposita, Vol. 14, No. 3, Springer Berlin/ Heidelberg,

Craddock, P.T., Meeks, N.D. 1987.  Iron in Ancient Copper, Archaeometry, Vol 29, Part 2, Research Laboratory for Archaeology and the History of Art, Oxford University

Jayawardhana, D.E. de S., Padmasiri S. 1977. Report on the Copper Magnetite at Seruwila Arippu Prospec” Ceylon Geological Survey Department

Loftus-Hills, G., Solomon, M. 1967.  Cobalt, Nickel and Selenium in Sulphides as Indicators of Ore Genesis, Mineralium Deposita, Vol. 2, No. 3,Springer Berlin/ Heidelberg

Maliyasena, H.B. 1987. Metallurgical and Archaeological Studies of some iron objects from the 1969 Anuradhapura Citadel (Gedige area) excavation in Seminar on the Archaeometallurgy of Sri Lanka, September 1987, IFS Kandy

Reedy, C.L. 1997. Himalayan Bronzes Technology, Style and Choices, University of Delaware Press

Seneviratne, S. 1995. The Ecology and Archaeology of the Seruwila copper magnetite         prospect northeast Sri Lanka, The Sri Lanka Journal of the Humanities, Vol. XXI, No 1&2, University of Peradeniya, Sri Lanka

Sirinivasan, S. 1999a. Lead Isotope and Trace Element Analysis in the study of over Hundred South Indian metal icons, Archaeometry, Vol 41, Part 1, Research Laboratory for Archaeology and the History of Art, Oxford University

Sirinivasan, S. 1999b. Priliminary Insights into the Provenence of South Indian Copper Alloys and Images Using a Holistic Approach of Comparisons of their Lead Isotope and Chemical Composition with Slag and Ores In S.M.M Young et. el (eds), Metals in Antiquity, BAR International Series, Archaeopress, No 792, 1999, p 208

Thantilage, A. 2008 Protohistoric Copper Metallurgy in Sri Lanka: an overview In Dr. Roland Felicitation Volume, Postgraduate Institute of Archaeology, Sri Lanka

List of Images Studied:

Images belong to group A

Related posts:

Posted in Articles