Collection online

painting / 繪畫

  • Object type

  • Museum number


  • Description

    Painting of the Paradise of Bhaiṣajyaguru, Buddha of Healing, set in a landscape with Mañjuśrī, Samantabhadra and attendants. Bodhisattvas beside the Buddha are painted in Tibetan style. Esoteric Buddhist figures in the lower section (partly destroyed). Tibetan and Chinese inscriptions. Ink and colours on silk.


  • Culture/period

  • Date

    • 836
  • Findspot

  • Materials

  • Technique

  • Dimensions

    • Height: 152.3 centimetres
    • Width: 177.8 centimetres
    • Height: 180 centimetres (Framed)
    • Width: 201 centimetres (Framed)
    • Depth: 3 centimetres (Framed)
  • Inscriptions

      • Inscription Type

      • Inscription Position

        in central catrouche
      • Inscription Language

        Tibetan, Chinese
      • Inscription Translation

        [Tibetan:] In the year of the dragon, I, the monk dPal-dPyangs, painted in a group the images of Bhaisajyaguru; Samantabhadra; Manjusri-Kumara; a thousand-armed thousand-eyed Avalokitesvara; Cintamanicakra; Parinatacakra etc. for the benefit of my health and to transfer the merit (?) [The Chinese inscription is similar and gives the date: the year bingchen, ninth moon guimao, the day dingsi (A.D. 836)].
  • Curator's comments

    In spite of extensive damage to the lower half, this painting is one of the largest silk paintings from Dunhuang. It is also one of the most important because the central inscription in Tibetan and Chinese, although almost completely faded when first seen, can be read and does provide a date. This inscription (see below) had not been read when Waley catalogued the Stein collection, but has recently been revealed by infra-red photography (Fig.43) and published by Heather Karmay in her book on Early Sino-Tibetan Art. Her translation of the Tibetan part reads as follows:
    In the year of the dragon, I, the monk dPal-dbyangs, painted in a group the images of Bhaisajyaguru; Samantabhadra; Mañjusri-kumara; a thousand-armed, thousand-eyed Avalokitesvara; Chintāmanicakra; Parina-tacakra etc., for the benefit of my health and to transfer the merit (?) (created by the act of painting, to all beings).
    The Chinese inscription, in nine columns from left to right, repeats the substance of the Tibetan part, but is more specific as to date. More of the text can be made out than when it was published by Karmay:
    Respectfully painted one group of Bhaisajyaguru on his dharma seat; Mañjusrī, Samantabhadra and their assemblies, one group; thousand-armed, thousand-eyed (Avalokitesvara), one group; Cintāmanicakra, one group; Amoghapāsa, one group; through this to acquire merit on behalf of his deceased (father and) be reborn in the Dharmadhatu ...together ascend the way of enlightenment. The year bingchen, ninth moon guimao, the first fifteen (days), the day dingsi, executed and completed.
    This inscription is set in the very centre of the painting as it survives today, rather than below as with most of the silk paintings. Both the Chinese and the Tibetan texts agree in giving precedence to Bhaisajyaguru (Pl.16-3), who presides over the whole in the top centre, accompanied by two Bodhisattvas and a numerous retinue. For the rest, the text confirms what is already fairly clear from the painting. Below, on either side of the inscription, Samantabhadra and Mañjusrī are clearly recognizable by their vehicles (Pls.16-4,16-5),while at the bottom Avalokitesvara (Fig.46) in the centre can be seen to be accompanied on the left by one of his transformations, Cintamanicakra (Fig.45),with the missing figure on the right identified by the Chinese text as yet another form of Avalokitesvara, Amoghapāsa. The popularity of Mañjusrī and Samantabhadra in the period of Tibetan domination at Dunhuang is attested to by many wall paintings, and the great size of Avalokitesvara here demonstrates his popularity also. The painting is indeed identified as a mandala of Avalokitesvara by Waley, supported by Karmay, but I prefer the precedence which the dual inscription seems to give to Bhaisajyaguru. Perhaps this is not so much the paradise of the Medicine Buddha, however, as a composite reflecting the personal predilections of the Tibetan monk who commissioned the painting.
    Not only the dual inscription in Tibetan and Chinese, but the painting itself reveals its Tibetan connections, for here in the same composition we find, as Heather Karmay has pointed out,"two quite different styles that existed side by side in Dunhuang. The first, in which the main body of the painting is painted, is in Chinese style, and the second, confined to the portrayal of the Bodhisattvas on either side of Bhaisajyaguru, may be directly associated with the group of Tibetan paintings [from Dunhuang]." In fact, of course, it is not so much two different styles as the integration of two different modes of depiction that we find here, since the actual means-line and colouring-is exactly the same for both types of figure. The initial drawing is in pale ink, still quite easily discernible in details such as the eyebrows. Over this is the white of the skin colour, with slight shading in pink. Finally, outlines such as those of the eyebrows, mouth and major drapery lines are drawn again in a darker ink, or, in the case of the hands and fingers, ears and nose, in red. These represent common stylistic elements which are, I think, sufficient to show that the same artists were at work here. Where the figures differ is in their attitudes, in the configuration of their facial features, and in the clothes that they wear. The two chief Bodhisattvas above, lightly clothed in the Indian manner, are seated with one leg pendent, the body swaying to one side, the head not only inclined in the opposite direction but also lengthened on the one side and foreshortened on the other, while the hair falls in many small ringlets over the shoulders. This is in contrast to the demurely clothed figures of Manjusri and Samantabhadra below, although they also are seated with one leg pendent. They have almost symmetrical faces and long smooth locks of dark hair framing the shoulders and arms.
    In some ways the painting appears to be midway between a mandala and the usual type of paradise painting with an architectural setting. There is a landscape background, clearly visible in the upper left (Pl.16-7)and originally in the right corner also. This can just be seen to have extended behind the central Buddha group also, and the latter is flanked by now-faded plantain trees. The space around is sparsely filled with flowering plants, but there are dividing lines of tiles also, suggesting a rectangular platform for the main group and dividing the middle assemblies of Manjusri and Samantabhadra from the triad at the bottom. Some of the most charming details are the small figures of flying apsarasas around the canopies of these two Bodhisattvas; in addition one attendant on either side holds a staff from the top of which flies a panelled banner with streamers. These two great assemblies, like the apsarasas and the smaller groups of sixteen and eighteen seated figures above them (Fig.44), are essentially supported on clouds which issue out of the landscapes above. All are shown advancing towards the centre. The triad below is static, dominated by the geometric forms of the multiple arms of Avalokitesvara and the great white red-bordered circles of the haloes of Cintamanicakra and Amoghapasa.
    As explained by Karmay, the only bingchen year to fall within the period of Tibetan domination at Dunhuang corresponds to A.D.836. Among the sutra manuscripts also, those of this period which bear dates commonly lack any indication of the Chinese reign period, and are dated, as one might expect, by the cyclical combination alone. As the style, too, is manifestly of the ninth century, and the depiction of Manjusri and Samantabhadra not far removed from such a securely dated example as Pl.23, this date can be regarded as reliable and of considerable value in understanding the development of painting at this time.儘管下邊大半部分已缺失,但該繪畫仍屬敦煌絹繪中超大型作品之一。說其重要是因爲在中央部位有吐蕃和漢文對照的題記,因為褪色,肉眼幾乎無法看清,但仍可識讀,並提供年代。Waley編寫斯坦因收集的敦煌繪畫目錄时,尚不能解讀此題記,而近年通過紅外線拍照才達到解讀的可能(參見Fig.43)。Heather Karmay在其著作《早期中原與西藏藝術》(Early Sino-Tibetan Art)中有介紹,她将吐蕃文解讀如下:


    不僅題記是吐蕃文和漢文合璧,繪畫表現上也有吐蕃化的成分。有關兩種形式的融合,我們也從畫面上可以看到。這一點正如Heather Karmay已經指出的,“……在敦煌,有兩種完全不同的形式共存。首先是繪畫的主要图像體現了完全中國化的形式,其次是在[敦煌]的一系列藥師的兩脅侍菩薩與吐蕃有緊密联係。”當然,實際上我們在這裏看到的,與其說兩種不同形式的共存,還不如說兩種不同圖像表現的統一。在兩種圖像中,線條和顔色均採用了完全一致的手法。比如可清楚辨认的眉的部分,用薄墨描下線;皮膚塗白色,用粉色薄薄晕染;施彩後,眉、口以及衣褶等均用濃墨描出,而手和指、耳朵等卻用紅色線。因爲兩种圖像都存在這種共同的表現方式,所以,可以充分肯定在該畫是由同一組畫家製作的。圖像表現不同的地方,體現在那些姿態上,即容貌和身上的衣裳的區別上。上邊的二脅侍菩薩著印度風格的輕衣,結半跏趺姿勢而坐。不僅身體向一邊扭,頭側向反面,頭的一側也顯得稍長,反側略短,有深度感。一方面,披在肩上的頭髮,髮梢形成小卷。這與下面同樣結半跏趺,姿勢端正的文殊和普賢的形成對比。文殊和普賢的面部幾乎是左右對稱的,頭髮長而平滑,可見多處結,從肩上懸到手腕。


  • Bibliography

    • Whitfield 1982 pl.16, figs 43-46 bibliographic details
    • Stein 1921b pl.III bibliographic details
    • Stein 1921a pp.1041-1042; vol.IV, pl.LIX bibliographic details
  • Location

    Not on display

  • Exhibition history

    2014 Mar-Sept, Rubin Museum of Art, New York, 'The Art of Tibetan Medicine'

  • Subjects

  • Associated names

  • Associated places

  • Acquisition name

  • Acquisition date


  • Acquisition notes

    For full acquisition history, see 1919,0101,0.1.

  • Department


  • Registration number


  • Additional IDs

    • Ch.xxxvii.004 (Stein no.)


If you’ve noticed a mistake or have any further information about this object, please email: 

View open data for this object with SPARQL endpoint

Object reference number: RFC638

British Museum collection data is also available in the W3C open data standard, RDF, allowing it to join and relate to a growing body of linked data published by organisations around the world.

View this object

Support the Museum:
donate online

The Museum makes its collection database available to be used by scholars around the world. Donations will help support curatorial, documentation and digitisation projects.

About the database

The British Museum collection database is a work in progress. New records, updates and images are added every week.

More about the database 


Work on this database is supported by a range of sponsors, donors and volunteers.

More about supporters and how you
can help