| articles

Back to main article

Observations on The Metal Analysis of the Gilt Copper Throne and Mandorla Associated with the British Museum’s Yongle Period Buddha
by Keith Mitchell

A technical analysis of the British Museum’s gilt brass Yongle Buddha and associated gilt copper throne form part of a paper published in the BM technical Research Bulletin.[1] The report contained previously published metal analysis of the Buddha that the authors considered to be consistent with a Yongle date. However, analysis of the metal used to construct the throne was undertaken for the first time, with the results showing copper of between 98.1% and 98.6% purity, with measurable trace elements of tin and lead: minor traces of iron and silver were observed but not quantified. The results of this analysis, including the very low iron content (>0.1%), partly informed the authors’ conclusion that the gilt copper throne was not of the Yongle period but of a later, possibly eighteenth century conception.

It is rare that an analysis of the composition of a metal will on its own define the age of the object. As Craddock cautions, “… usually it cannot be said that a particular trace element composition is impossible for a culture or at the date concerned, merely that it has not been previously encountered”.[2] While a number of studies of Buddhist images have been undertaken,[3] analyses of Ming period objects made of copper (<97%) are rarely encountered,[4] making comparison by grouping impossible at present.

When metal analysis is used in questions of authenticity it is usual for a sample to be cut or drilled from the object to obtain a clean section free from contamination. The sample is then tested by Atomic Absorption Spectroscopy as described in most of the published papers,[5] or in the case of a cut sample it is mounted for examination and analysis using scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM/EDX). Both these techniques give a high degree of accuracy in detecting trace elements, while a lower level of trace element detection is achieved when using surface X-ray fluorescence (XRF).[6] When surface XRF is used it is usually considered necessary to abrade the surface to remove patina or possible surface enrichment to improve accuracy,[7] and when considering the results possible errors in detection of trace elements need to be taken into account.[8] The limitations of surface XRF analysis are acknowledged within the BM report. The authors state that no preparation was carried out to remove possible surface corrosion, but this reduced level of accuracy was considered sufficient since the analysis was to be used only to allow comparison of the metal composition between the components of the throne.[9] The surface XRF analysis, therefore, can only be seen as giving a general indication of the composition of the copper used in the making of the stepped base and mandorla.

In the BM report the result of quantitative analysis of copper used in the construction of the throne and mandorla is seen as possible evidence for eighteenth century manufacture because the copper was found to be well refined and with low iron content. When considering the content of trace elements in metal it is generally thought that their levels will be reduced over time by technical improvements in metallurgy, and therefore well-refined metals will be evidence of more recent manufacture. However, when modern metals used for craftwork have been analyzed this has not been found to be the case.[10] The possible identification of well-refined copper with eighteenth century metalwork is based on the study of the reintroduction of imports of Japanese copper into China and analyses of Chinese coins.[11] Cowell, La Niece and Rawson discuss this work and its implications in late Chinese metalwork. The uniformly low traces of antimony, nickel and silver are seen as indicative of Japanese copper but very few examples of its possible use in Chinese metalwork are cited. These elements combined with traces of cadmium are further discussed as indicators of later eighteenth century metalwork. The authors, however, caution that the volatile nature of cadmium and the possible use of recycled materials could make such data difficult to use.[12] Since the levels of these elements are not quantified by the BM analysis of the throne and mandorla, and low iron content on its own is generally not cited in wider surveys as a significant indicator of age, the identification of the copper as being later than the Yongle period and possibly of eighteenth century manufacture appears premature.

It is usual for reports on individual objects, or surveys of wider groups, to include cautionary declarations on conclusions drawn from analysis of metal composition. The statements can seem unremarkable and their implications are seldom explained, but the issues behind the declarations do have a bearing, and include the method of analysis, metal heterogeneity, provenance and date contexts, number of analyses performed, workshop practice and method of manufacture, the use of recycled metals, historical maintenance or repair, modern restoration or conservation, changes in opinions based on art history and connoisseurship. These topics are well covered in Craddock’s Scientific Investigation of Copies, Fakes and Forgeries. In Cowell, La Niece and Rawson some of these issues are highlighted by their discussion of a group of Chinese fifteenth century inscribed metal images that include the BM Yongle Buddha.[13] Based on metal analyses the authors discuss the possibility that the sculptures are of a later date and that the inscriptions are thus not of the period. Conclusions based on interpretation of metal analyses and connoisseurship can change as illustrated by the fact that the BM report under current discussion acknowledges that the Yongle Buddha is indeed of the period, citing metal composition, method of manufacture, style and workmanship as all being indicators of Yongle period manufacture.[14]

Thus, the results of the XRF analysis of the BM gilt copper throne sections cannot be seen as beyond question in defining the date or provenance of the metal used in their construction.

The authors of the BM report also suggest that the virtually identical throne of the Yongle Buddha formerly in the Speelman collection is so similar to the BM example that it too is likely to be later than the Yongle period. However the Speelman example has not been systematically examined, and no metal analysis has been undertaken to establish a possible conformity with the BM throne sections. Thus no conclusion on the dating of the Speelman throne sections is possible relative to the metal analysis of the BM throne and mandorla.

A question has arisen as to the difference in colour of the fire gilding on the figure of Buddha and the throne sections. The tone of the gilt surface will depend on the base metal and workshop practice such as surface finishing before gilding, temperature used in the process of fire gilding, the length of time of heating, and final finishing and patination.[15] These factors together with environmental circumstances, surface wear and corrosion would explain the differences in the gilt colour.


1. Quanyu Wang and Sascha Priewe, “Scientific analysis of a Buddha attributed to the Yongle period of the Ming dynasty”, The British Museum Technical Research Bulletin, Volume 7, 2013, pp. 61-8

2. P. Craddock, Scientific Investigation of Copies, Fakes and Forgeries, Butterworth-Heinemann, 2009, p. 139

3. P. T. Craddock, “The Copper Alloys of Tibet and their Background”, W. A. Oddy and W. Zwalf, eds, Aspects of Tibetan Metallurgy, British Museum Occasional Paper No. 15, 1981, pp. 1-33
Jett. P., and J. G. Douglas, “Chinese Buddhist Bronzes in the Freer Gallery of Art: Physical Features and Elemental Composition”, in the Materials Issues in Art and Archaeology 111: Materials Research Society Symposium Proceedings, Volume 267, 1992, p. 205-24
M. Cowell, S. La Niece & J. Rawson, “A Study of Later Chinese Metalwork”, P. Jett, ed, Proceedings of the First Forbes Symposium at the Freer Gallery of Art, 2000, pp. 80-9

4. P. T. Craddock, 1981, op. cit. p. 22

5.ibid. p. 20. Jett and Douglas, 1992, op. cit. p. 210 . Cowell, La Niece, Rawson, op. cit. p. 81

6. Sheridan Bowman, Michael Cowell, Joe Cribb, “Two Thousand years of Coinage in China: an Analytical Survey” in Helen Wang, Michael Cowell, Joe Cribb, Sheridan Bowman, eds, Metallurgical Analysis of Chinese Coins at the British Museum, British Museum Research Publication Number 152, 2005, p. 9 . Cowell, La Niece & Rawson, 2000, op. cit p. 81

7. Quanyu Wang, Sascha Priewe, Kwang-tzuu Chen and Susan La Niece, “A Chinese bronze gui vessel: genuine Western Zhou object or fake?” British Museum Research Publication, Volume 5, 2011, p. 63

8. Bowman, Cowell, Cribb, 2005, op. cit. p. 9

9. Wang, Priewe, 2013, op. cit p. 63

10. P. Craddock, 2009, op. cit p. 139

11. M. Cowell and H. Wang, “Metal Supply for the Metropolitan Coinage of the Kangxi Period”, in Metallurgical Analysis of Chinese Coins at the British Museum, British Museum Research Publication No. 152, 2005, pp. 84-95

12. Cowell, La Niece, Rawson, 2000, op. cit. p. 87

13. ibid p. 87

14. Wang, Priewe 2013 op. cit p. 67

15. W. A. Oddy, M. Bimson and S. La Niece, “Gilding Himalayan images: History, Tradition and Modern Techniques,” in W. A. Oddy and W. Zwalf, eds, Aspects of Tibetan Metallurgy, British Museum Occasional Paper No. 15, 1981, p. 87-101

Back to main article | articles