| articles
Download the PDF version of this article

British Museum technical report | Observations, by Keith Mitchell

Yongle Period Metalwork:
The British Museum Sakyamuni

by David Weldon

text and photos © the author except as where otherwise noted

January 19, 2017

(click on the small image for full screen image with captions.)

Fig. 1

Fig. 2
In 1957 Douglas Barrett published two gilt brass statues in the British Museum, a Shakyamuni Buddha and a Manjushri: both are inscribed with a Chinese reign mark, da Ming Yongle nian shi (bestowed in the Yongle era of the great Ming) (1403-1424).[1] (Figs. 1, 2) Barrett found little reason to doubt the evidence of the inscriptions, but others questioned the attribution of an early Ming (1368-1644) date.[2] The statues seemed to have more in common with Himalayan style and iconography than other early Ming Buddhist works that conform to classical Chinese sculptural traditions, such as a Hongwu period (1368-1398) Buddha in the British Museum dated 1396.[3] This stylistic anomaly contributed to the doubts that some had about the age of the two Yongle marked statues. In 1975 Heather Karmay (Stoddard) published research on early Ming annals that detailed imperial gifts of Buddhist statues to Tibetan monasteries during the Yongle period,[4] and concluded that the reign marks engraved on the relatively small number of Yongle statues known at the time were of the period: she was aware of eleven examples including the two statues in the British Museum.[5] In 2001 Ulrich von Schroeder recorded fifty-four gilt brass statues in Tibetan monastery collections bearing the same da Ming Yongle nian shi mark, confirming the evidence found in the early Ming archives.[6]

Fig. 3
While the authenticity of the da Ming Yongle nian shi mark on the BM gilt brass statue of Shakyamuni Buddha is now generally accepted, doubts have been raised over the age of its separately made throne back (torana or prabhamandala) and the throne base into which the figure is secured. The authors of a technical and stylistic study published in a British Museum Technical Research Bulletin concluded that the sculpture of Buddha was of the Yongle period but the throne and mandorla (sic) were probably of a later date.[7] As a result of metal analysis by X-ray fluorescence it was established that the throne sections were constructed from relatively pure copper with a very low iron content (<0.1%). The low iron content suggested to the authors that the gilt copper sections were probably made at a later date than the Yongle period gilt brass Buddha. No comparison with other Yongle period or later copper samples was cited, and no elucidation was offered as to why copper with a low iron content is indicative of a post Yongle date. While acknowledging stylistic links to bas-relief depictions of Buddhas and thrones at the Yuan period (1271-1368) Juyongguan Cloud Platform, the paper argues that the gilt copper throne sections are “nearly identical” to thrones depicted on Qing period (1644-1911) embroideries that pay homage to an early Ming ‘prototype’. (Figs. 3, 4) The authors thus conclude that the gilt copper throne sections of the British Museum’s Yongle period Buddha are later than the statue and were possibly made in the Qing dynasty.

Fig. 4
It is important to consider the reliability of the evidence used in the BM report. The paper concedes that the use of copper itself cannot be used as an indicator of date: the low iron content of the copper was the factor believed to suggest the later date. Attribution of date by metal analysis might be cautiously considered if there were a large number of comparative copper samples that could be reliably dated to the Yongle through the Qing period, but this data does not exist and in any case would not be conclusive, as Keith Mitchell has demonstrated.[8] Rigorous stylistic analysis, however, reveals that the throne of the BM Buddha is far from “nearly identical” to the thrones of the Qing embroideries, which are in fact merely superficially comparable to the gilt copper throne. As this paper will argue, the BM throne and torana are composed entirely of quintessential Yuan and early Ming period design elements, with some detail unique to the Yongle period, while the embroideries depict Qing period motifs in a composition that imitates a Ming style. The British Museum gilt copper throne sections can therefore be shown beyond doubt to be Yongle period and not later copies as proposed in the BM Bulletin.

While the Qing embroidery recreates the general shape of an early Ming throne it does not include the following highly distinctive Yuan and early Ming stylistic details that make up the essential structure of the BM gilt copper throne and torana.

I. The Yuan/early Ming style pillars supporting the upper tiers of the BM gilt copper throne base

The pillar at each corner of the recessed central panel corresponds to those on Yongle stupas. (Figs. 5, 6) A similar design is seen in Yongle sutra illustrations and on the thrones at Juyongguan. (Figs. 7, 3) There is no reference to this ubiquitous Yuan/early Ming throne pillar on the Qing embroidery.

Fig. 5

Fig. 6

Fig. 7

II. The Yuan/early Ming style corner projections of the BM gilt copper throne base

Fig. 8
The corners of the gilt copper throne have subtly upturned ends. This feature originates in medieval eastern India and in Newar art and architecture of Nepal, as seen in a thirteenth century Tibetan thangka of Tara by a Newar painter. (Fig. 8) Newar artists, including Anige (1245-1306), the Controller of Yuan Imperial Workshops, were responsible for innovation in Chinese Buddhist art during the Yuan period. Indeed the thrones of the Buddhas at the Yuan period monument at Juyongguan, which so closely resemble the throne of the BM Buddha, depict the same upturned corners seen in the thirteenth century Tara painting. Upturned corner projections are prevalent throughout early Ming sutra illustrations and textile thangkas. Each corner of the throne on the Qing embroidery, however, is marked by a post and finial, a virtually ubiquitous design element of late Ming and Qing period shrines and pedestals.[9] This universal late Ming/Qing period architectural device is clearly not the inspiration for the classic Yuan/early Ming upturned corner design of the gilt copper throne.

III. The quintessential Yongle style of the lotus petals on the BM gilt copper throne base

Fig. 9
The design of the lotus petals above and below the central recessed panel is similar to that on the pedestal of the BM Yongle Manjushri. (Fig. 2) The slim, elongated petal with a simple three-point curl at the tip is used exclusively in Buddhist sculpture of the Yongle period. This distinctive design is not encountered in the Xuande period (1425-1435) or later, and is indeed a defining feature of Yongle period Buddhist art. In contrast, the lotus petal design on the embroidery incorporates an ornamental flourish on either side, and an elaborate upturned tip. This baroque petal style is used throughout the Qing period on porcelain, enamel, cloisonné, and sculpture such as the BM Kangxi period (1661-1722) lacquered wood Buddha consecrated in 1692.[10] (Fig. 9) The style of the lotus petals depicted on the embroidery is not the inspiration for the quintessential Yongle period lotus petal design of the gilt copper throne.

IV. The early Ming style of the lotus flowers on the BM gilt copper torana

The torana depicts the classic early Ming lotus design of compact multi-petal flower heads, as seen throughout fifteenth century Chinese painting, ceramics and cloisonné.[11] The torana of the Qing embroidery depicts the opened eight-petal lotus seen throughout the Qing period.[12]

Thus, the thrones of the embroideries selectively interpret the form of early Ming thrones using Qing period detail, while the throne and torana of the BM Buddha embody the very essence of Yongle style in both its details and overall structure. These gilt copper sections, displaying quintessential stylistic elements of the Yongle period that are absent in the Qing embroideries, are indeed rare and exquisite examples of Yongle period metalwork. Important stylistic anomalies demonstrate that Qing period works such as the embroideries could not have been the inspiration for the gilt copper throne sections, as proposed in the BM Bulletin.[13] Conversely, it is the embroideries that take their inspiration from early Ming works such as the British Museum’s Yongle Buddha, throne and torana.

Fig. 10

Fig. 11

Fig. 12

To date, only one other imperial Yongle period gilt brass Buddhist statue with original throne is recorded. The Buddha formerly in the Speelman collection is the same style but larger than the British Museum example and has a similar gilt copper throne and torana with overall height of 72.5 cm. (Fig. 10) The overall height of the BM Buddha and throne is 59 cm. Similarities extend to the way in which both statues are prepared for consecration, when sacred materials were placed in the cavities. (Fig. 11, 12) Evenly spaced chisel cuts are usually made on the base rims of Yongle statues to retain a metal plate that seals the consecration material within. (Fig. 13) The BM and Speelman examples do not show these marks. (Fig. 14) Instead, holes are drilled horizontally through the rims into which dowels would have been fitted to secure the sealing plates. (Figs. 15, 16) Once the dowels were inserted the outline of the holes would still be visible, and a blemish on the smooth gilt surface,[14] but once statue and throne are united the lower rim is concealed within the well of the throne, hiding the imperfection by design. (Fig. 17) The holes in the lower tier of the separate gilt copper thrones, which the British Museum paper was unable to explain,[15] are also for the insertion of dowels to retain metal sealing plates.[16] The dowels would have been considerably easier to conceal within the elaborate scroll motif of the thrones than on the smooth gilt surfaces of the Buddhas. This is demonstrated on the Yongle stupa that is sealed in the same manner: gilt-headed dowels are inserted into holes made in the scroll design on the lower edge. (Figs. 5, 18)

Fig. 13

Fig. 14

Fig. 15

Fig. 16

Chased and repoussé copper thrones and torana are common in fifteenth century and earlier Himalayan art. Newar artists perfected the metalworking techniques by at least the early eleventh century.[17] From at least the thirteenth century, Nepalese metal statues of deities were often cast and then attached to separate wrought copper lotus bases.[18] Cast metal statues in Tibetan monastery collections often have Nepalese style gilt copper repoussé torana, dating from at least the twelfth century.[19] Nepalese craftsmen were employed in the Yongle imperial workshops,[20] and it is not inconceivable that their renowned skill with wrought copper informed the construction of the thrones for the British Museum and Speelman Buddhas.

Fig. 17

Fig. 18

David Weldon is an independent researcher and consultant living in Brussels, Belgium. He has published numerous books and articles, including:
The Perfect Image: The Speelman Collection of Yongle and Xuande Buddhist Icons, Arts of Asia, May-June, 1996, Vol. 26, no. 3
Early Tibetan Manuscript Covers: 12th-15th century, exh. cat, Rossi & Rossi Ltd, London, 1996
Sacred Symbols: The Ritual Art of Tibet (co-authored with Robert Thurman) exh. cat., Sotheby’s and Rossi & Rossi Ltd, New York, 1999
The Sculptural Heritage of Tibet: Buddhist Art in the Nyingjei Lam Collection (co-authored with Jane Casey), London, 1999
Tibetan Sculpture Inspired by Earlier Foreign Sculptural Styles, Oriental Art, Vol. XLVI, no. 2, 2000, reprinted in The Tibet Journal, Vol, XXVII, Spring and Summer 2002
Divine Presence: Arts of India and the Himalayas (co-authored with Jane Casey and Naman Ahuja), Barcelona, 2003
Cast for Eternity: Bronze Masterworks from India and the Himalayas in Belgian and Dutch Collections (co-authored with Jan Van Alphen and Ian Alsop), Antwerp, 2004
On Recent Attributions to Aniko 2010,


1. Douglas Barrett, “The Buddhist Art of Tibet and Nepal”, Oriental Art, 1957, Vol. III, No. 1, pp. 90-5

2. Heather Karmay, Early Sino-Tibetan Art, Warminster, 1975, p. 72

3. W. Zwalf, ed, Buddhism: Art and Faith, London, 1985, p. 208, cat. no. 300

4. Karmay, op. cit. pp. 76-83

5. ibid, p. 73

6. Ulrich von Schroeder, Buddhist Sculptures in Tibet, Hong Kong, 2001, Vol. II, pp. 1237-91

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

8. Keith Mitchell, “Observations on the Metal Analysis of the Gilt Copper Throne and Mandorla Associated with the British Museum’s Yongle Period Buddha”,

9. cf. the post and finial design of Qing period Buddhist shrines in the Summer Palace Collection, Vicky Hsu, Jack Cheng, Lin Lin Chang, eds, Buddhist Art from Rehol, Taipei, 1999, p. 189-208, cat. nos. 85-92

10. The use of this elaborate lotus petal type originates in the early Ming period but is exclusive to large-scale works. A 57 cm Xuande Amitayus appears to be the smallest early Ming imperial statue with this elaborate petal style, Sotheby’s New York, March 25, 1999, lot 121. The rounded petal with an elaborate flourish at the tip (but with no ornamental flourish on the sides) as seen on the integral lotus pedestal of the 37 cm high British Museum Buddha, is used only in bronzes over ca 25 cm and occur in both the Yongle and Xuande periods. The slim petal type with the three-point curl at the tip is used in Yongle cast images below ca 22 cm. Conventions for lotus petal design according to size are maintained throughout imperial Yongle and Xuande Buddhist metal sculpture.

11. cf. the early 15th century Chinese cloisonné chilug water vessel in the British Museum, W. Zwalf, ed, Buddhism: Art and Faith, London, 1985, p. 208, cat. no. 301

12. cf. the classic Qing lotus design on an imperial Qianlong gold and cloisonné ewer, Evelyn S. Rawski and Jessica Rawson, eds, China: The Three Emperors, 1662-1795, London, 2005, p. 150, cat. no. 54

13. The Buddha and throne have since been published in the British Museum’s Ming exhibition catalogue, where the throne and torana are again erroneously attributed to the Qianlong period, Craig Clunas, Jessica Harrison-Hall, eds, Ming: 50 years that changed China, London, 2014, p. 227, fig. 195

14. As the statues would already be gilded at the time of their consecration no further metalwork could be done to conceal the dowels without damaging the gilding.

15. Quanyu Wang and Sascha Priewe, op. cit. p. 65

16. Separately made parts of Yongle and Xuande imperial Buddhist statues are individually consecrated, Sotheby’s Hong Kong, “Visions of Enlightenment: The Speelman Collection of Important Early Ming Bronzes”, October 7, 2006, p. 12

17. A Nepalese gilt copper repoussé Vishnu from the Zimmerman Family Collection, now in the Metropolitan Museum of Art, is dated 1004, Pratapaditya Pal, Nepal: Where the Gods are Young, New York, 1975, p. 109, cat. no. 79

18. Three Nepalese ca thirteenth century cast copper images of deities with separate copper repoussé lotus pedestals, see Ulrich von Schroeder, Indo-Tibetan Bronzes, Hong Kong, 1981, p. 354, pls. 93A,B,C. The practise probably predates the thirteenth century, but it seems that no separate repoussé base of an earlier date survives before those cited.

19. Giuseppe Tucci, Transhimalaya, Geneva, 1973, p. 136, pl. 141; Michael Henss, “Early Tibetan Sculpture”, in Pratapaditya Pal, ed, On the Path to Void, Mumbai, 1996, pp. 101-41, pls. 15, 16, 23

20. Karmay, op. cit., p. 72

British Museum technical report | Observations, by Keith Mitchell

Download the PDF version of this article
| articles