| articles

Articles by Dr. Pratapaditya Pal

Download the PDF version of this article

The Thoughtgem and Wheel Avalokiteshvara and Lokeśvara of Wish-fulfilling Gems Revisited
by Pratapaditya Pal

Foreword (2021)

The article reprinted directly below was published some 53 years ago in The Journal of the Indian Society of Oriental Art. Recently a colleague from Thailand requested for an offprint and I realized that not only did I not have one, but I had completely forgotten about the article’s existence. I did not even possess a copy of the journal in my personal library and got hold of a scan, courtesy of the Getty Research Institute. I mentioned the fact conversationally to my friend Ian Alsop, the founder/editor of who also had never read it and requested a copy of the scan and after reading it he felt that it was worth republishing.

I agreed with him as it was a fairly comprehensive study of the subject covering several different Asian countries where Mahayana Buddhism continues to flourish. Comparative studies of Buddhist iconography across the various countries where the faith traveled from India in ancient times appear to have gone out of fashion since the days of Alice Getty (d. 1946) and so many others of her generation and so we reprint it in to make it more accessible.

Having the opportunity to reread the article after more than half a century, I was not surprised to note some inelegant linguistic constructions and printing and other errors due to editorial negligence. I have taken the opportunity therefore to make a few changes and corrections but have not altered the substance in any way. Rather, in an Afterword following the article (see Part II below), I have added new information and discussion to bring it up to date, as far as is possible for a mid-octogenarian scholar without access to libraries and other sources in these days of the pandemic, along with new acknowledgments. Substantial changes have been made, however, in the illustrations, which have been updated with new color photographs not available in 1968; the captions of the illustrations have also been enlarged to include more complete catalogue information, wherever possible. In the Afterword, transliteration of certain Sanskrit words that have now entered the English lexicon also follow contemporary practice and are without diacritics such as for instance Avalokiteshvara, stupa, etc. and are no longer italicized.

text and images © the author except as where otherwise noted

September 20, 2021

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

Part I
The Iconography of Cintāmaṇi Cakra Avalokiteśvara

This article was originally published in The Journal of the Indian Society of Oriental Art, (Calcutta) New Series, Vol. II, 1967-1968.

Fig. 1
The iconography of Mahāyāna Buddhist pantheon had already become complex by the time it spread outside the Indian subcontinent in the early centuries of the Common era. Despite the great sanctity attached to forms of icons imported from the country of their origin, it was inevitable that local concepts would modify or expand the pantheon. Thus, in the countries where Mahāyāna Buddhism prevailed we often encounter alterations from the original form conceived in India. Furthermore, many of the icons that have survived in these countries are no longer known in India either in art or in texts. In such instances the Chinese texts are of considerable help for purposes of identification. A prominent example is that of Cintāmaṇi Cakra Avalokiteśvara or Nyoirin Kannon as he is called in Japan and Ju-i-lun Kuan-yin in China.

Fig. 2
Nyoirin Kannon or Avalokiteśvara of the Wishing Gem Wheel is one of the most popular forms of Kannon in Japanese art [1]. He is one of the six Kannons especially venerated by the Tendai sect, and by the Kamakura period numerous temples had sprung up in his honor. In his most common form, where he can be indubitably recognized as Nyoirin, he is represented, as in the painting in the Museum of Fine Arts, Boston, as seated in mahārājalīlā, or royal ease,and with six arms (fig.1). Usually attired and adorned regally as befitting a bodhisattva, his hands display the following gestures and attributes. One of the right hands invariably touches the right side of the face in the gesture of thoughtfulness and two others carry the jewel (cintāmaṇi) and a rosary. Of the three left hands, one rests on the lotus seat, and the others hold the lotus and the wheel. That he is a Kannon or Avalokiteśvara is evident, apart from the lotus in his hand, from the tiny effigy of the Tathāgata Amitābha adorning the crown.

Fig. 3
As was usually the case with later Buddhist iconography, several other variants of the form were conceived, perhaps both by the theologian and the artist. By the Kamakura period at least six different forms of Nyoirin were familiar in Japan [2]. At times he is represented as seated on a rock beside the water, the rock signifying Mount Potalaka, the favorite abode of Avalokiteśvara, as in Figure 1. Alternatively, he may sit in lalitāsana, padmāsana or in mahārājalīlā; or he may be shown as standing upon a lotus or on the back of a dwarfish figure. He may have two, six, eight-or twelve arms, although the eight and the twelve-armed manifestations are altogether different from the six-armed variety. He is also seen presiding over mandalas as well as accompanying others, as in another painting in the Boston Museum or in that from Dunhuang in the British Museum (Figs. 2 and 3).

In his simplest, and perhaps the earliest form, which may be considered non-tantric, he is represented in Japanese art as seated either in lalitāsana or padmāsana, the right hand being raised in abhaya- or vyākhyāṇa-mudrā and the left hanging down. Alternatively, the right hand may hold a jewel near the breast and the left a lotus on which is placed a flaming object, perhaps a wheel. Alice Getty described a two-armed form of Nyoirin where he is ‘seated in the European fashion with the right foot supported by the left knee. The right elbow rests on the right knee, while the head leans on the index and second fingers of the right hand.’ [3] This is obviously a description of the so-called ‘Meditating or Pensive Bodhisattva’ and will be discussed below.


There are several texts in the Chinese Buddhist tradition concerning Cintāmaṇi Cakra Avalokiteśvara. The Indian monk Bodhiruci translated a dhāraṇi of his six-armed manifestation as early as 709 CE [4]. The Taishō Tripitaka lists several others translated by Gijo, Śikṣānanda, Ratnacinta, Amoghavajra and Vajrabodhi [5]. Of these scholars Śikṣānanda was a Khotanese and Ratnacinta a Kashmiri, while the others were associated with central and eastern India. All these monks seem to have been more or less contemporaneous. The fact that they were from as far apart as Khotan and Nalanda (where Vajrabodhi was ordained) would indicate the popularity of Cintāmaṇi Cakra in India by the beginning of the eighth century. Yet neither his representation nor a sutra concerning him is known prior to that date.

According to the sutra translated by Vajrabodhi, Cintāmaṇi Cakra should be of a golden complexion and should have six arms. [6] He should be seated on a lotus on Mount Potalaka and on his crown should be depicted the Tathāgata Amitābha. His first right hand displays the pensive gesture as he is contemplating on the causes of all human suffering. The second holds the wishing gem (cintāmaṇi) which can fulfill all desires, while the third grasps the pearls (or beads) of meditation, evidently a literal Chinese rendering of the word japamālā. Of his left hands, the first rests on Mount Potalaka, for which he is said to provide support, the second carries the lotus which can purify all non-believers, and the third holds the wheel which represents the Dharma without compare. The sutra further states that this six-armed bodhisattva can roam the six roads of transmigration (rikudo) in order that the means (upāya) of great compassion (mahākaruṇā) can end all suffering.

The popularity of Nyoirin in Japan in the Heian and later periods is attested not only by the large number of images of the times but also by literary traditions. These traditions usually recount how a miraculous event occurred by the grace of the kannon and a grateful patron founded a temple to show his appreciation. A typical example is the story associated with the foundation of the celebrated Ishiyama-dera [7].

It is related that Rōben, who built the temple in A.D. 749 by order of the emperor Shōmu, had to obtain 900 r y o of gold for the Daibutsu or Tōdaiji. The oracle of Usa Hachiman had warned however that the gold must be obtained in Japan and not brought from China. The emperor dispatched a messenger to Kimbusen to pray to Kongō Zaō Bosatsu, but the messenger was informed by another oracle that the gold of that mountain could be removed only with the appearance of Maitreya, whereas the gold for the Daibutsu would come if prayers were said to a Kannon image ‘to be made upon the stone seat of an old man on the bank of the Seta river in Shiga district, ÓMI province,’ [8] The spot was discovered and there an image of Nyoirin Kannon dedicated; this was the honzon of the famous Ishiyama-dera. The story is of course no different from many such that were current in India about the origin of an image or a temple.

The Ishiyama-dera with its honzon was destroyed by fire in 1048 A.D., but because of its celebrity a later description gives us a fair idea of the image. As one might expect—considering that the earliest sutra for the six-armed variety was translated into Chinese only forty years previously—the image was only two-armed. We learn that ‘the majority of the images as they are now, and, as in the restored Ishiyama-dera, show the left hand holding a lotus and the right in teaching gesture (vyākhyāṇa-mudrā) [9]. This is really the description of the more common variety of Avalokiteśvara. Two other early images of the two-armed variety are known; one in the Honsumisakiji in Tojushima going back to the Heian period and their other in the Kannōji in Hyōgo attributed to the late Heian period. To the late Heian period also belong the two six-armed images in the Hōryūji in Nara and the Daigoji in Kyōto. The well-known icons in the Kanshinji in Osaka and in the Muroji in Nara, both of which are in painted wood, are of the six-armed variety, and are dated in the ninth century. Obviously, by this time the sutra had been translated from Chinese into Japanese.


Fig. 4
The earliest known Chinese image of the six-armed type is a gilt bronze of the late T’ang period [10]. Four of the arms are damaged, but substantially the figure is very similar to the Japanese representations. The uppermost right hand is placed near the ear and of the remaining left hands one is placed on the lotus seat and the other holds the stem of a lotus. It is possible that he was not as popular in China as in Japan; for he is not represented in the Peking pantheon published by Clark. Alice Getty [11] suggests the identification of some two-armed images in Yun Kang and Lungmen with Nyoirin Kannon, but the identification is doubtful.

Fig. 5
The different forms of Cintāmaṇi Cakra Avalokiteśvara occur in paintings from Dunhuang [12]. In a scroll dated 864 C.E., he is portrayed in the top tier along with three other forms of the bodhisattva (Fig. 3). Richly attired and bejeweled, he stands like the others firmly and frontally. His left hand carries a lotus and the right possibly a jewel, while the figure of Amitābha adorns his crown. The accompanying, cartouche bears an inscription in Chinese which reads 'the great noble Bodhisattva of the Wishing Gem Wheel.’ [13] But the more common form portrayed in Dunhuang paintings is of the six-armed variety, similar to that seen in Japan with only minor variations. For instance, in a mandala of the Sahasrabhūja Avalokiteśvara, he is included as a subsidiary divinity and one of his hands is in varadamudrā, apparently without the rosary. Waley mistakenly described this as a four-armed manifestation, but there is no doubt that he has six arms [14]. However, from the number of his representations recovered from Dunhuang, it is apparent that he enjoyed considerable favor in Central Asia and the form traveled from India via the Silk Road courtesy of the monks mentioned above.

Fig. 6
Among the other countries, where Mahāyāna prevailed and the cult of Avalokiteśvara was popular, neither Java nor Nepal seems to have been familiar with Cintāmaṇi Cakra Avalokiteśvara. I know of only one small figure that shows some compositional affinity with the six-armed form. In a silver icon in the Museum van Aziatische Kunst Amsterdam, we see a four-armed Avalokiteśvara seated in mahārājalīlā on a lotus (Fig. 5). His upper right hand touches his forehead and the lower displays the varadamudra, while the left hands hold a manuscript and a lotus. Although both the essential attributes—the jewel and the wheel—are absent, the gestures of meditation and varada are common. In Nepal, a form that is widely worshipped is known simply as Cintāmani Lokeśvara. In the example illustrated here (Fig 6), he is two-armed and stands in tribhaṅga with crossed ankles below an elaborate kalpavṛkṣa. His right hand holds a ring and the left a sheaf of jewels. He is accompanied by two female attendants, probably two Tārās, and two yaksas who kneel as they fill bags with gems, presumably from the tree [15]. Although this iconic form is altogether different, once again there is a basic conceptual relation between Cintāmaṇi Cakra Avalokiteśvara and Cintāmaṇi Lokeśvara.

In the Tibetan tradition, no sūtra in the Kanjur is devoted to Cintāmaṇi Cakra; nor is there a descriptive sādhana in the otherwise rather exhaustive list in the Tanjur. He is included however in the Rin-lhan of the Pancen Lama Batan-pahi-in-ma-phyoga-lasrnam-gyal [16]. His Tibetan name is Spyan-ras-gzigs-yid-bzhin-hkor-lo and there are two illustrations of him by a Mongol artist in a manuscript of the Rin-lhan.

In one of these he appears seated with four other forms of Avalokiteśvara and is four-armed [17]. His attributes are the rosary, the jewel, the water-pot and the lotus. In the other manifestation he is given three heads but is only two-armed [18]. The heads are arranged one above the other, the first face being benign, the second terrifying and the third that of a Tathāgata. The right hand is raised and carries a jewel and the left a lotus, similar to the two-handed form in Dunhuang. It is significant that in neither of the forms is he given the distinctive pensive gesture or the wheel. It may be of interest to mention here that among the ādikalyāṇa (primordially auspicious) deities in the same text are four Tārās who are named as Cintāmaṇi Cakra Sita Tārā and are the epiphanies of the four eminent masters Atiśa, Ba-ri, Gnan and Vanaratna [19].


It may appear rather curious that a form of Avalokiteśvara that was so popular in Central Asia and Japan should hardly have been known in India. He is certainly not included in the Sādhanamālā or in the Niṣpaṇṇayogāvalī or in any of the later dhāraṇi texts that I have had the opportunity to examine.

Fig. 7
Some years ago, the Asutosh Museum (University of Calcutta) acquired a clay plaque from Nalanda, which certainly represents Cintāmaṇi Cakra Avalokiteśvara and is probably the only example that has come to light so far in India (Fig. 6). It shows a figure seated in mahāirāijalīlā on a lotus and with six arms. His hair is tied in a chignon and a pearl necklace is the only ornament adorning his otherwise bare torso. His head is emphatically inclined to the right and rests against the palm of the uppermost right hand. The second right hand seems to carry the rosary, while the third hangs down from his knee and displays the varadamudrā. It may also have held a jewel which is now not discernible. Of the three left hands only one is clearly distinguishable and rests on the lotus seat. Two others seem to have been damaged or the mold from which the plaque was cast may have been defective, and hence the imperfect impression. Below the lotus is an illegible inscription in two lines. Stylistically, the plaque may be assigned to about the ninth century.

Thus, although the known texts are silent about Cintāmaṇi Cakra Avalokiteśvara, his six-armed form was not altogether unknown. For some reason he may not have captured the imagination of the Mahāyānists in India and may have been rather exclusively worshipped at Nalanda. It is interesting to note that Vajrabodhi, who translated one of the sūtras into Chinese, was for a considerable time associated with Nalanda. What is of significance art-historically is that such plaques were clearly the sorts of object that were easily transported by scholars and pilgrims across the frontiers and served as the artist's models. For one cannot but be struck by the remarkable similarity in the disposition, the composition and the iconographic arrangement of the representations from as far apart as Nalanda (Fig. 7), Dunhuang (Fig. 4) and Japan (Fig. 1). Moreover, often the effigy in a plaque may not always have been very distinct, as in the Nalanda example, and this may account partly for the differences that occur in representations of the same divinity in India and elsewhere.


The name Cintāmaṇi Cakra is composed of two words, cintāmaṇi meaning ‘wishing gem’ and cakra meaning ‘wheel’, which are his two distinguishing attributes. The lotus is of course the most common cognizant of Avalokiteśvara, while the rosary became a distinctive emblem in his tantric manifestations. We have already alluded to the textual explanation of the function of these symbols. The hand resting upon the lotus is probably an artistic necessity in the sense that it serves the physical function of supporting the weight of the body because of the posture. It helps to keep the figure in equilibrium, necessitated in most images for the mahārājalīlā posture. Possibly the symbolical interpretation in the sutra was a later imposition.

Fig. 8
There is a further belief in Japan that the six arms of Nyoirin ward off the six evils. This is probably a modified or earlier version of the concept of eight evils from which Avalokiteśvara protects his devotees, and hence he is known in some forms as Aṣṭamahābhayatāraṇa Avalokiteśvara. A similar function is performed by his female consort Tārā and so she too is known as Aṣṭamahābhayatāraṇī Tārā. The concepts of Avalokiteśvara and Tārā seem to have evolved in a complementary fashion, which is also evident from the fact that in the Tibetan pantheon she is given a form known as Cintāmaṇi Cakra Sita Tārā.

The pensive gesture could have been inspired by the word cintā, meaning 'thought'. At any rate, it is not improbable that because this form of Avalokiteśvara and that of the meditating bodhisattva, as in the Chinese example in Boston (Fig. 8), had this gesture in common, Alice Getty identified the latter type as Nyoirin. There can be little doubt however that iconographically the form of Nyoirin Kannon may well have evolved out of the more elemental type of the 'meditating Boddhisatva', which was so popular in Gandhāra and subsequently in China, Korea and Japan.

A princely figure seated with his right leg across the left and his slightly lowered head poised on the index and the second fingers of his right hand (Pl. XIII, Fig. 8) has variously been identified by scholars as representing (a) Maitreya, (b) Śākyamuni, (c) Nyoirin, (d) Meditating Bodhisattva and (e) Bodhisattva Mahāsattva [20].

It is generally admitted that the iconographic type of the pensive figure was derived from Gandhāra, at least by the Central Asian and Chinese artists, although, as Alexander Soper suggested, the Gandhāra artists may have adopted it from the repertoire of their Roman counterparts [21]. In the art of Gandhāra the type seems to have been employed in several different contexts to represent a variety of divinities.

Usually in Gandhāra such figures appear in pairs amidst a host of others, and without any attributes. In the steles portraying, according to Foucher, the Great Miracle of Sravasti [22], and according, to John Rosenfield [23] - (perhaps more plausibly) - the Theophany of the Buddha, a pair of such figures occurs usually at the top within separate niches. What does lend plausibility to Rosenfield’s suggestion is that the entire composition, with the majority of the figures centered around the seated Buddha and engaged either in listening or discoursing among themselves, shows a narrative intent, which is quite different from the formulaic representations of the Great Miracle in later art. The literary parallels that the Gandhāra compositions immediately evoke are the stereotyped scenes described at the beginning of the Mahāyāna sūtras. In such instances it may be safer to identify the pensive figures simply as 'Bodhisattva Mahāsattva', as suggested by Rosenfield.

Among the so-called Bodhisattva figures in Gandhāra, perhaps the most definitely identifiable is that of Maitreya, because of the kamaṇḍalu or water-pot in his hand. In later iconography, the nāgakeśara flower was added as the water-pot became a common attribute of many of the later forms of Avalokiteśvara. As a further distinguishing feature later iconographers gave Maitreya the additional emblem of the stūpa depicted on his crown. The stūpa was the principal symbol of the Buddha after his mahāparinirvaṇa, and probably it came to signify the very essence of 'buddhahood'. Hence it was an apposite symbol for Maitreya, the future Buddha.

In at least one relief portraying the theophany of the Buddha [24], Maitreya is represented unmistakably along the top in his Tusita paradise, while a pensive figure is seated on Śākyamuni's left. In this instance, therefore, the meditative figure cannot be identified as Maitreya with any justification. What is of greater interest is that the left hand of the meditating figure clasps the stalk of a lotus. There are several other examples in Gandhāra of a lotus-bearing pensive Bodhisattva. In a stele from Loriya Tongai [25], a preaching Buddha is flanked by the two pensive Bodhisattvas, the one on his right carrying a manuscript in his right hand and the other on his left holding a lotus.

Despite the objections raised about identifying these figures as bodhisattvas [26], I see no reason why the manuscript-bearing figure in this instance cannot be identified with Mañjuśrī and the lotus-bearing figure with Padmapāṇi. Foucher has published at least two other independent images [27] - and such images could well have served as the prototypes of the Chinese Meditating Bodhisattvas of the Northern Ch'i period where the right leg of the figure is placed laterally across the left, the right hand resting on the left knee and apparently raised to the chin, but with a lotus in the left hand. Here also they would seem to represent the Bodhisattva Avalokiteśvara, whose duty it is to meditate upon the causes of suffering.

Fig. 9
It is somewhat curious that despite the popularity of this type of figure with the Gandhāra artists, in no other Indian center of Buddhism, whether Mahāyāna or Hīnāyana, was he known to have been represented. In view of this, a small metal image in the collection of Nasli and Alice Heeramaneck in New York (Fig. 9) assumes considerable importance. It is cast in copper with lingering traces of gilt and is only three inches (or 7.8 cm.) high.

Wearing a dhoti, he is seated on a lotus in mahārājalīlā. A scarf is tied diagonally around his bare torso and a deerskin is thrown across the left shoulder. His hair is arranged in a jaṭāmukuṭa on which is a tiny effigy of the Tathagata Amitābha. His right arm rests on the right knee and the index finger is raised almost to the forehead. The left hand holds a full-blown lotus. Behind his head is a flame-shaped halo, somewhat reminiscent of the haloes of Chinese Wei Dynasty steles.

The lotus together with the effigy of the Tathāgata Amitābha on the jaṭāmukuṭa leaves no room for doubting the identification of the figure with the Bodhisattva Avalokiteśvara. The piece was probably cast in Kashmir although stylistically the artist responsible for it seems to have been aware of the T'ang style, judging from the treatment of the physiognomy as well as the shape of the halo. Such an awareness in Kashmir in the seventh or the eighth century is not at all impossible. Iconographically, however, this image confirms that at least those pensive Bodhisattvas holding a lotus in the art of Gandhāra may represent Padmapāṇi Avalokiteśvara.


Thus, in Gandhāra the meditating figure may portray (a) Mañjuśrī, (b) Avalokiteśvara in specific contexts, (c) Śākyamuni or (d) Bodhisattva Mahāsattva. In no instance, however, has this type of figure been used to represent Maitreya and this is a relevant point to bear in mind [28]. With the popularity of the cult of the Bodhisattva in and around Gandhāra by the second century, it was natural for the artists to conceive a prototypical figure in a pensive attitude to convey to the public the basic ideals and ideas embodied in the concept of the 'Bodhisattva'. They could hardly have chosen a more expressive type. Just as the image type of the Buddha was employed generally by the artists with a mere difference of a gesture to represent the historical Buddha as well as the past Buddhas or Tathāgatas, similarly the iconic type of the pensive Bodhisattva may have been employed as a stock template to portray different Bodhisattvas in different contexts.

In view of the above it seems reasonable not to be dogmatic about our identification of the figures of the ‘pensive Bodhisattva’ in the Buddhist art of Asia. In some instances, where he is seated under a tree or has a horse near his feet, he certainly represents the Bodhisattva Siddhārtha before his enlightenment, which is also stated explicitly in the accompanying inscriptions. In many other steles, however, he has been referred to simply as 'the heir apparent’, which probably indicates Prince Siddhārtha or Maitreya. As I have mentioned, at least in Gandhāra, there seems no reason to identify all these figures as Maitreya, and this may caution us to so recognize all such pensive figures in China, Korea or Japan. It is not improbable that artists of the Northern Ch’i period felt it appropriate to adopt the meditating type for Maitreya. As the future Buddha, Maitreya must in the present kalpa be a Bodhisattva contemplating upon the way of the Dharma. At any rate, where the context is not clear, such images may best be identified as Bodhisattva Mahāsattva or the archetypal Bodhisattva.

Thus, at least by the beginning of the eighth century, when Bodhiruci translated a dhāraṇi of the six-armed Cintāmaṇi Cakra into Chinese, the form must have been current in India. The solitary plaque from Nalanda corroborates this, although the form does not seem to have survived in later Buddhist art of the subcontinent or Himalayas or Southeast Asia. Yet in Central Asia and Japan it proved to be immensely popular. Another such example seems to be that of Amoghapāśa Lokeśvara, whose image also has not come to light in India with certainty [29].

There can be little doubt that the basic idea as well as the elementary form of the six-armed Cintāmaṇi was derived from the two-armed ‘Pensive Bodhisattva' figures of Gandhāra. Conceptually there could be no more expressive a gesture for a Bodhisattva than that of ‘contemplation'. The multiplication of arms as a manifestation of the omnipotence of a divinity was a distinctive development of Indian iconography by the Gupta period, possibly due to tantric influence. The lotus of course is the most favorite emblem of Avalokiteśvara and both the cintāmaṇi as well as the rosary were already familiar elements of his iconography by the eighth century [30]. The compassionate Bodhisattva dedicates himself to finding ways for the liberation of the sentient beings from the suffering of the phenomenal world, and the only proper way is the way of the Dharma, symbolized appositely by the wheel. *

* I am thankful to Mr. J. Fontein and Mr. C. Horioka for their generous help with the Chinese and Japanese sources.

Part II
The Thoughtgem and Wheel Avalokiteshvara and Lokeśvara of Wish-fulfilling Gems Revisited:
An Afterword to Part I


Since the above article was published, the scholarship on Buddhism has increased prodigiously including the literature on Buddhist iconography. However, no other discussion on the iconography of Cintāmaṇi Cakra Avalokiteśvara has come to my attention. In fact, the study of comparative Buddhist iconography among the various nation states of Asia remains rather limited.

The standard and almost the sole work in English in comparative iconography is still The Gods of Northern Buddhism by Alice Getty. It was first published by Oxford University Press in 1914 following the pioneering works of several continental European scholars, though geographically less extensive than Getty’s book. Getty herself received help from several scholars, most generously from M.J. Deniker without whom, it is clear, her book would have remained unpublished. [1] Getty did not know all of the languages involved, and, undoubtedly, this is the principal reason why such comparative studies are so rare a century after her publication: it can be achieved only by institutions bringing together scholars with knowledge of several languages.

Concurrently with Getty, Ananda K. Coomaraswamy made some important contributions to the iconography of all three religions and set the standard for new methodology that was then followed by one of my teachers Jitendra Nath Banerjea. Although his book on Hindu iconography is better known, he contributed significantly to the study of Buddhist iconography as well, both by teaching and writing. [2]

For those of us who studied Indian art for the M.A. degree in 1958, we had the good fortune that Benoytosh Bhattacharyya’s expanded edition of The Indian Buddhist Iconography came out that very year—a copy of that edition is still with me. Only one other colleague of mine—Dipak Chandra Bhattacharyya—at the University decided to work on both Hindu and Buddhist iconography and revisited the manuscripts that Benoytosh had consulted, especially in the Asiatic Society, Calcutta, and made many improvements on the elder scholar’s readings (see bibliography). Having read many of them as a doctorate student under the tutelage of another of my professors, Sarasi Kumar Saraswati, who was both a great Sanskritist and an epigraphist, I can testify how difficult the old scripts are to decipher.

Benoytosh Bhattacharyya’s compilations were culled from different texts of sādhana (invocations) which he titled Sādhanamālā or Garland of Evocations. He published the Guhyasamāja tantra as well as the Niṣpannayogāvalī by the 12th century Indian paṇḍita (pundit) Abhyakāragupta based on manuscripts that had survived in Nepal but not in India.

Interestingly, Benoytosh was the son of Mahāmohapādhyāya Haraprasad Shastri (1853–1931) from an illustrious family of scholars in Bengal. In fact, he was probably one of the Sanskrit scholars in Calcutta whom Getty and Deniken had contacted for her book. Haraprasad was also responsible for publishing in 1917 the first of a series of catalogues of the manuscripts that he and his no less illustrious predecessor, Rajendra Lal Mitra, another giant in the field of Indology (1824–1891), had collected. Haraprasad had as a model the 1883 catalogue of the Sanskrit manuscripts in Cambridge University Library by Cecil Bendall. [3] Another author whose works are a must for all those interested in Buddhist iconography is, of course, the Italian scholar, the eminent Giuseppe Tucci (1894–1984), particularly the magisterial opus, Tibetan Painted Scrolls (1949).

Several scholars have continued to mine the Sādhana and Dhāraṇī texts which are the principal sources in Sanskrit for Buddhist iconography in the second half of the 20th century. I have mentioned Prof. Saraswati from whom both I and my fellow student Dipak Bhattacharya benefited immensely in learning to recognize the various scripts necessary to read the texts. The German Sanskrit scholar Gudrun Bühnemann has by herself, or along with Musashi Tachikawa, reproduced some original Sanskrit manuscripts making them more easily available. She has also published a few Nepali sketchbooks with labeled drawings of hundreds of deities which should be of considerable interest to scholars (see bibliography). [4] Although they contain numerous forms of Avalokiteshvara/Lokeśvara, I am afraid Cintāmaṇi Cakra is not one of them. In fact, just as that particular form remained popular in E. Asia, in Nepal, for instance, a favorite was Amoghapāśa (“one with the unfailing noose”) Lokeśvara. Why one particular form was popular in one region and not another remains a mystery.


In her Preface to the first edition Getty wrote, “the most accurate source of information in regard to the Northern Buddhist divinities has been found in the sādhana, or texts of invocations of the gods, in which they are described with much detail. Unfortunately, sādhana of all the gods of the Mahāyāna Pantheon have not as yet been discovered, and there remain a number of deities about whom very little is known.” [5] This is still true today.

This was partially remedied by Benoytosh Bhattacharya by publishing his two volumes of texts of these sādhana (or dhyāna) culled from several manuscripts emanating from Nepal. (No similar collection has yet been found on Indian soil). He also published another important text called the Niṣpaṇṇayogāvalī of Abhyakāragupta (12th c.), which too contains descriptions of evocations of mandalas and individual deities. It may be emphasized that the evocations that Bhattacharya culled from several manuscripts in different libraries were copied at least as early as Abhyakāragupta’s time. Some provide their authors’ names but most are anonymous. It must be stated that in this regard it is imperative to consult Chinese sources which have more historical information about the authors as well as dates when they were composed as has been discussed in my article. However, unfortunately, it is still unknown why the forms were composed, nor why Bodhiruci’s translation of Cintāmaṇi Cakra was considered insufficient and five other translations were made. Moreover, why certain iconographic forms were popular in certain regions and not others still remains unexplained. After a lifetime of dedication to the study of Buddhist iconology it is still a mystery why, for instance, Cintāmaṇi Cakra captured the imagination of the East Asian Buddhists but not of the Indian subcontinent, the Himalayan cultures or Southeast Asia.

Since writing that article I still have not found a Sanskrit sādhana of Cintāmaṇi Cakra among the dhāraṇī or sādhana texts that I have had access to. Although a great deal of Tibetan sources both textual and visual have been published in several European and Asian languages, not much can be added to the discussion about this particular form of the bodhisatva, nor have I discovered more visual representations except a couple more in Korea and Tibet in sketchbooks rather than complete artistic representations, either painted or sculpted.

Cintāmaṇi Cakra does not carry a generic gem (maṇi), but a specific one which is called the cintāmaṇi; the prefix cintā literally means thought but is generally translated as “wish fulfilling.” Getty further suggested a “magic jewel,” but Bhattacharyya characterizes it as a “gem that satisfies all desires.” This is the meaning behind the creation of the totally different form of Cintāmaṇi Lokeśvara that prevailed in Nepal (see fig. 6 above and fig. 12 below), to be discussed below.

In passing it should be mentioned here that the cintāmaṇi is the principal symbol of Ratnasambhava, one of the five transcendental Buddhas. Noteworthy is that his name is translated by Getty variously as “Buddha of Precious Birth” or as “the source of precious (or holy) things.” Literally, it should be “jewel-born,” as is the case with Padmasambhava or the “lotus-born.” [6] Interestingly, the Tibetans translated Ratnasambhava as rin byuṅ and Padmasambhava as padma byuṅ-gnas. The jewel is the principal symbol of the deities belonging to the family of Ratnasambhava, especially the cintāmaṇi or “thought-gem,” and not of Amitābha, whose principal bodhisatva is Avalokiteshvara.

The other component of the name of the particular form of Avalokiteshvara under discussion is, of course, cakra or wheel, a multivalent word. Primarily it is a direct referent to the first sermon of Buddha Śākyamuni after his enlightenment at the deer park in Sarnath near Banaras. It has remained a very visible and potent symbol of the faith at least from the time of the first cakravartin or universal emperor Ashoka of the third century BCE as it occurs frequently in his stone monuments. It also appears on early Jain monuments going back ultimately to the chariot wheel of Sūrya, the sun god of Vedic origin. In the Buddhist context the wheel is called dharmacakra, which is best translated as the “wheel of Dharma” as the Sanskrit word dharma (or dhamma in Pali) has no English equivalent. [7] Subsequently, it came to denote a specific hand gesture of the Buddha formed with both hands that became known as dharma-cakra pravartana (turning or rotating the wheel of Dharma).

In earlier Gandhāra art the Buddha literally turns the wheel with his right hand. It should be pointed out, however, that the wheel is also an important symbol of Viṣṇu who holds it both as a symbol of his original solar character and as a weapon. It is sometimes translated as a “discus” which it is not, but does act like a boomerang, returning always to the god’s hand after its service. The wheel is also the symbol of Mahāvīra, the 24th and most important Jina, and a historical figure like the Buddha.


A few words should be added here about the rich Tibetan sources on Buddhist iconography published since the article above was written in 1968. This is possible because of several books published since then by Lokesh Chandra, which contain extensive material from Tibetan but also from Chinese, Korean and Mongolian sources (see bibliography). Not only are they indispensable for the study of comparative Buddhist iconography both for textual resources and illustrations, but some, such as the volume on Thousand-armed Avalokiteśvara (1988), is a substantial iconological study on the concept of the compassionate deity, significantly expanding the earlier exemplary study of de Mallman.

Frederick Bunce (1994) in An Encyclopedia of Buddhist Deities, Demigods, Godlings, Saints and Demons has given us, as the title reveals, an encyclopedic compendium of names and descriptors as well as bibliographic references which makes it a convenient handbook and easy reference. However, his sources are mostly secondary and there are few illustrations. For instance, for Cintāmaṇicakra Avalokiteshvara, he has three different entries: a) the first is “Cintamani-Avalokitesvara” where he provides the six-armed form as in Nyoirin kannon but standing instead of seated and cites Walter E. Clarke (1965) as his source; b) Cintamani-chakra where he refers the reader to “Nyōirīn Kwan-non and Getty”; and c) Cintamani Cakra Avalokiteśvara for which he sites Manfred Lurker (1987) but not Helen Chapin’s pioneering article. [8] Incidentally, in this third entry with his full name, Cintāmaṇi Cakra is standing and is two armed, the right holding the wheel (cakra (skt) or khor-lo (tib)) and the left against the chest supporting the jewel (ratna or nor-bu). This is a variation of the two-armed version in the Machendrabahal in Kathmandu whose drawing was published by Benoytosh Bhattacharya (1958) referred to in my original article above. It should be noted, however, if the sketch in Bhattacharya is correct, then the outstretched right hand supports a small stupa shaped object which may well symbolize the jewel, but the empty left arm hangs down to the waist. Are these due to textual variations or the artists’ whims?

Fig. 10
In Lokesh Chandra’s manuscript of the Narthang Pantheon, however, Cintāmaṇi Cakra Avalokiteshvara is standing and with two arms, the right outstretched and displaying the jewels in crystalline shape, as one sees in the rocky mountain motif in early Central Tibetan paintings such as the famous Tārā thangka in the Ford collection. [9] Another major deviation in the Narthang sketch is that the Avalokiteshvara is given two additional heads, placed vertically above one another: The second is of a ferocious deity and the topmost that of a Buddha. This Buddha head certainly represents Amitābha, the parental transcendental Buddha of the bodhisatva, while the ferocious head may depict either of the two forms of Mahākāla who are described as Cintāmaṇi-Rakta (red), and Cintāmaṇi-Sīta (white) Mahākāla. It should be noted that a second form of Avalokiteshvara this time is represented in the peaceful form with four arms, two hands in the namaskāra (salutation) gesture against the chest, the second right grasping the crystalline jewel and the second left, the lotus. This particular jewel bearing form is simply called Maṇidhārin (jewel-holder) in Sanskrit and Nor-bu-hdzin-pa in Tibetan. It may be recalled that in one of the Dunhuang paintings (fig. 3 above) Cintāmaṇi Cakra is portrayed in standing posture.

None of these variations, it should be stressed, are yet to be recognized either in textual or in artistic forms in the Indian subcontinent. Even more than the wheel, the jewel seems to have remained the preferred attribute associated with the bodhisatva in most other Buddhist countries though the origin can be traced back to the Indian teachers who ventured forth to China as early as the 8th century. The standing Avalokiteshvara with the jewel is preserved only in Tibet but we cannot ascertain whether they were created there or in India.

What can be said with certainty is that the form began its life in the generic type of “pensive” bodhisatva in the early centuries of the Common Era in the Gandhāra region as exemplified in a little known example in LACMA (fig. 10). This remained popular in Kashmir but nowhere else in the subcontinent, or in Nepal and Tibet (fig. 9 above).

Fig. 11
What is even more interesting is that at least one relief in Gandhāra style is known where we encounter the Buddha Śākyamuni himself as a pensive figure in the most realistic manner, not seen in Buddhist art so far anywhere else (fig. 11). Published in 2008, Christian Luczanits rightly identified the scene as Brahmā and Indra, along with the first five human disciples, each entreating the master to teach the Dharma. [10] The relief was found in Saidu Sharif and is now in the Swat Museum. A popular subject, in all other compositions the Buddha is an impassive, still figure, but here he displays strong emotion as he turns away his head as if unable to face the audience or the suffering of humanity. By comparison the bodhisatva’s gestures in the LACMA sculpture and in all later representations of the figure are more formulaic and less intense. Indeed, this Swat relief by an unknown but sensitive sculptor expresses with remarkable acuity the thoughtfulness (cintā) that will be required, even by the Enlightened One, to show the path leading to the gem (maṇi) that is the Dharma. As the devout Indian Master Matṛicetā (6th c. CE) in his famous ode to the Buddha wrote,

There is only one thing which resembles you, O kind one
The jewel of the Dharma, by gaining which you won preeminence

Which shall I praise first, you or the great compassion, by which
For so long you were held in Samsara, although you well knew its faults? [11]


Finally, I return to the rich world of Buddhist iconography of Nepal and the form of Cintāmaṇi Lokeśvara, which we will further explore with figures 14 - 18. I first encountered the Nepali form of Cintāmaṇi Lokeśvara in the small example with the collectors Nasli and Alice Heeramaneck collection in New York in the late 1960s and published it in the 1968 article (above, fig. 6, click here to view). The Eastern Asian versions of the Cintāmaṇi Cakra Avalokiteshvara that I explored in my earlier article seemed to be entirely limited to East Asian Buddhist traditions, and, as I wrote then, it certainly appeared “rather curious that a form of Avalokiteshvara that was so popular in Central Asia and Japan should hardly have been known in India”. The little Nepalese image was so different from the Central Asian and Japanese forms that I illustrated and described, that I thought it must be a Nepali invention, which I can no longer maintain for reasons below.

Before writing this afterword, I decided to look up the multivolume publications on Buddhist iconography that have been produced by an intrepid researcher of our time, Ulrich von Schroeder (see bibliography). Anyone who dares to explore the world of the complex and seemingly infinite Buddhist iconography or iconology today does so at his/her own peril ignoring von Schroeder. Not only has he revealed to the world what treasuries of Buddhist art the monasteries of Tibet are, but he has provided enough material in his monumental and luxuriously illustrated, and might I add, literally and figuratively, weighty volumes--enough material for scores of PhD aspirants. I hereby raise my right hand in the gesture of Tathāgatavandanā (praise be to the Buddha) to him. He certainly has earned the merit deserving of a bodhisatva.

Sure enough, in volume 1 of his Buddhist Sculpture in Tibet devoted to India and Nepal I encountered two Indian brass sculptures that now become possible earlier models for Nepal’s Cintāmaṇi Lokeśvara. [12] Schroeder attributes both to 12th century Northeastern India, but I would state that they likely originated in West Bengal.

Fig. 12
The simpler of the two bronzes (fig. 12) shows the bodhisatva standing in the usual posture with a slight contrapposto (ābhaṅga) beside a slender vine rising from his left like a partial frame of an aureole over his head and terminating at his right shoulder. At about the height of his left thigh is clearly a bud (which Schroeder identifies as a water pot). Higher up, next to the bent left arm a second shoot from the stem opens into a flower, thereby indicating that the rising plant is the lotus vine. More buds are depicted along the stem as it bends over his head. What is most unusual, however, is that his left hand is raised near his face and displays “the pensive gesture.” This is clearly a deviation on the part of the artist for usually in most instances since Gandhāra it is the right hand that makes this peculiar nameless gesture. It may be stressed that so far, no textually approved name of this gesture, either in Sanskrit or in any of the other languages, has ever been suggested. All English versions, such as a pensive, contemplative, thoughtful, etc. are modern constructs. The identification of the oval object in his extended open palm (varada or gesture of charity) likely symbolizes the jewel and here his identification as Cintāmaṇi Lokeśvara is not incorrect. This is the standing type that we met with in the Tibetan texts of Chandra though the right hand there is raised above the shoulder level, as if the jewel is being offered to the invisible Buddha. Although visually differing from the Nepali representations (figs. 6 and 14-18), conceptually they are related. This pensive gesture, however, relates it to the Cintāmaṇi Cakra form in its essence while the jewel with the charity gestures foreshadows symbolically the Nepali Lokeśvara of the wish-fulfilling gem tree form.

Fig. 13
The second example (fig. 13), however, is much more elaborate and vibrant both in the posture of the figures and the representation of the plant, both of which make it a possible model for Nepal’s Cintāmaṇi Lokeśvara (figs. 6, 14-18) unless we discover a Nepali prototype that is earlier than the 12th century Indian example. But here too we find some extraordinary variations that make it, as far as I know, a remarkably original composition in the Buddhist world.

The posture of Cintāmaṇi here makes use of the jaunty models that go back indeed to the tree dryads or śālabhanjikā motif that was adopted early in Buddhist art for yakshis in general, but also for Māyādevī giving birth to the Buddha. It should be noted also, as I have suggested in the article, the posture was adopted for the Hindu god Krishna with his flute standing below a Kadamba tree. The manner with which the bodhisatva grasps the wavy stem or trunk of the rising plant is clearly related to Māyā, who by the way, would be regarded in Indian society by the euphemism ratnagarbhā (jewel-wombed) for bearing such a son.

In any event, the plant in this Bengal bronze is not a specific one but an imaginary vine rising sinuously and shooting tendrils that form exuberant volutes or whorls; each encloses one of the seven jewels that are venerated by the Buddhists. This kind of exuberant treatment of scrolling plants rising from the base (or around the base) was a popular element of design in Pala art and was popular with artists in Tibet in their thankas and murals. Noteworthy also is the use of similarly designed trees to represent lineages of teachers hanging like fruits in Tibet and other East Asian countries.

Fig. 14
In the Indian tableau, each outward curling shoot has become a support or step on which rests animals, objects, and humans, altogether seven of them, each a symbol of an universal monarch or cakravartin. The first three portrayed here from the bottom up, above the energetic and acrobatic monkey shaking the tree or adoring the bodhisatva are clear: an elephant, a horse, and a sword (symbolizing the military commander). As to the others, the fourth is likely the queen (strīratna), generally denoting wealth or the treasury. The two others are the wish fulfilling jewel or (maṇiratna) and the wheel (cakraratna). It will not be possible here to go into a detailed discussion of this symbolism but it seems to be a Buddhist attempt to combine politics (rājanīti) with right conduct (dharmaratna). [13]

In any event, there seems no doubt that this unique image served as a model for the Nepali creation of their form as Cintāmaṇi Lokeśvara as seen in a lovely bronze image recently featured in a Bonham’s auction in Hong Kong (fig. 14). This remains one of the finest sculptural realizations of the form known to date from Nepal, and is strikingly close to the Indian images we have just examined.

Fig. 15
A more elaborate composition can be found in a spectacular painting or paubhā of the late 17th or early 18th century from Itum Bahal that I included in an exhibition of the arts of the Himalayas in Chicago in 2003 (fig. 15). [14] Unlike the necessarily reductive and simpler composition of the sculpted example in metal (fig. 14), the more complex composition consists of two figures: the strikingly larger white male with a more demure and dusky lady in Newari attire turning in the other direction and distributing jewels from the tree to a solitary supplicant. Both are luxuriously attired and bejeweled, as if competing with the spectacularly rendered tree (a veritable kalpavṛikṣa indeed) adorned with bright green leaves and an abundance of floral gems and ornaments, as sumptuous as a Christmas tree. Cintāmaṇi has a large red halo with a golden border, while his spouse Cintāmaṇi-Sita Tārā has a green halo; both could also be considered the universal ruler and his queen.

Fig. 16
The recipients of their largesse, however, hardly can be identified with either courtly or common people but are two ascetics (Hindu?) facing Lokeśvara and a third plump and well fed bearded middle-aged man who wears a cap (as Newars and Muslims do), but a short dhoti and not pajamas, with a waterpot in his left hand. Indeed, this composition reminds me of the much earlier scene in a fragmentary painting in the Heeramaneck Collection in the Los Angeles County Museum of Art, which I last published in 1985 (fig. 16). [15] There we see the last great Buddhist pandit/pundit of ancient Bengal Vanaratna’s (d. 1469) Newar wife (like Tārā but without a halo) distributing alms to a group of holy men. In my opinion, the Itum Bahal paubā of Cintāmaṇi and Tārā could have been done by a descendant of the artist responsible for the LACMA fragment of 1469 CE.

What is interesting for me today is how adroitly the unknown artist has combined reality with imagination and an infusion of spirituality to provide a visually compelling composition. The two deities are gracefully comfortable as they distribute their jewels plucked from the tree, even if their postures give the impression of two dancers posing on a stage. Above in the sky sits the Buddha Amitābha of red complexion watching the altruistic act as angels throw flowers from floating clouds. In this highly charged spiritual world a mood of levity is introduced in the recipients of divine largesse on earth in the form of three obvious caricatures of “kusali yogis” as they are referred to in Newar communities. Why shower mendicants with wealth and not the general pietistic and poor ordinary devotees?

Fig. 17
It was while researching for this “afterword” that I noted two other representations of Cintāmaṇi Lokeśvara from Nepal that I had acquired for LACMA. [16] One is a very beautiful example in wood revealing the natural talent of Newar woodcarvers in that medium (fig. 17). It was acquired in 1984 and clearly, I had misdated the piece in the catalogue (Pal, 1985: S36 pp 114-115, 15th c.). A comparison with the painting discussed above with the “Radha-Krishna” like images (fig. 14), and particularly of the wish fulfilling tree, it is clear that stylistically they are coeval and hence a 17th century date for the image is more likely.

Fig. 18
The second is a brass object that once was a temple lamp and represents a distinctive Newari design with two separately cast components: a crowning image above the wider lamp (fig. 18). How different from the two painted and carved representations this modeled metal ensemble is. Not only in terms of its visual complexity with wild animals frolicking in the canopy of the tree which is more like a woven spiderweb of an arch (toraṇa) providing entertainment for the tiny effigy of Amitābha near the apex, but noteworthy is how Cintāmaṇi is now provided with two spouses flanking him and how the so-called mendicants are almost undistinguishable and sit with open sacks like a group of musicians on the base of reliefs both in Indian and Nepali tableaus. It is interesting to note that fig. 6 of the original article depicts the same group and was originally also part of a lamp. Amazing how inventive the unknown Newar artists were!

I wish I had a contemporary example of a Cintāmaṇi Lokeśvara in any media to see how ingenious today’s artists are. It is clear no matter how stringent the invocations (sādhana, dhyāna or dhāraṇī) are, spiritual needs did not restrict the imaginations of the artists as we see with the several examples of Cintāmaṇi Lokeśvara of India and Nepal, or with Cintāmaṇicakra Avalokiteśvra in East Asia. As is the case with the latter, I am yet to find a Sanskrit description of the former.


To return to the Boston painting (fig. 1) which began this journey for me, my initial curiosity was aroused for what might be considered as an aesthetic response rather than an iconographic puzzle. As I stood before the painting in the gallery at the MFA in Boston, I still remember how mesmerized I was by the head of the Kannon and the gesture of the right hand against the cheek as is seen in the close-up (fig. 19). Indeed, it was a perfect expression of empathy and thoughtfulness, I thought as I gazed. Not until I found, while writing this afterword, the relief from Gandhāra of another representation–––this time of the Buddha Śākyamuni (fig. 11)–––did I have the same reaction though the two gestures are so markedly different.

Fig. 19
The Gandhāra Buddha’s gesture seems to express a sense of despondency at the enormity of the task ahead to save the world. Was the unknown artist familiar with the poem titled the “Buddha’s Pity” that was translated as early as 428 CE into Chinese? It begins,
            My children,
            The Enlightened One, because He saw Mankind drowning in the Great Sea of Birth, Death and Sorrow, and longed to save them,
            For this He was moved to pity.
            Because He saw the men of the world straying in false paths,
            and none to guide them,
            For this was He moved to pity. [17]

The unknown Japanese master, on the other hand, has given us the more gentle and thoughtful being that Avalokiteshvara is. He has attained all the qualities to become a Buddha but he must remain in the world to save it. As the hymn to him in the Lotus Sutra (Saddharmapuṇḍarika), one of the earliest to be translated into Chinese, reads:
            He who is now so compassionate to the world,
            He will be a Buddha in future ages
            Humbly I bow to Avalokiteśvara
            Who destroys all sorrow, fear and suffering.
The unknown Japanese master well understood the enormity of his task! [18]

In all other representations of the so-called “pensive gesture” transformed in visual language by artists from Gandhāra to Japan and Southeast Asia, the gesture has never been expressed with such pathos, though with differing vocabulary, as in the Butkara relief or the Boston painting. However, as I have said above, I am yet to find what this gesture is called in Sanskrit. Would cintā-mudrā be appropriate?

More than half a century after I first encountered Cintāmaṇi Cakra Avalokiteshvara as Nyoirin Kannon in the MFA Boston (fig. 1), my search has produced no Sanskrit sādhana of the deity and we owe our gratitude for diligently preserving the Chinese translations to the devotion of the Buddhists of Tang China.

The only evidence that the bodhisatva was ever represented in the Indian subcontinent remains the modest and miniature terracotta medallion found in Nalanda, Bihar (fig. 7), where one of the 8th century pundits/translators Vajrabodhi was ordained. Although another of these travelling monks was Ratnacinta of Kashmir, no image of the bodhisatva from that region has yet emerged. Incidentally, noteworthy is the similarity of the names Cintāmaṇi and Ratnacinta.

Except for the use of the word cintāmaṇi in the names of the two descriptive epithets, the two forms of that bodhisatva Cintamani Chakra and Cintamani Lokeshvara have little in common conceptually. While the older form is conceptually a more spiritual incarnation, the latter is a late creation in Nepal. While the two Bengal bronzes could have served as formal models, conceptually even this seems not to reflect any loftier goal than to seek material wealth. One wonders why he was created even though the Buddhists of Nepal already had a much older deity for wealth and prosperity in the goddess Vasudhārā which literally means stream of vasu, or wealth. It may be recalled that the word vasudhārā is of Rigvedic antiquity. There is also Jambhala Kubera whose attribute is a sack of gems. The Sanskrit adage adhikantu na doṣāya (excess is never a fault) may be applied to both materiality and piety.


While the earliest literary evidence shows the arrival of Cintāmaṇi Cakra Avalokiteśvara in China from India in the 8th century C.E., it was not long thereafter that he moved to Japan. Visual evidence demonstrates that he was brought to Japan in the early Heian period by the famous Japanese Buddhist monk pilgrim Kōbō-daishi (774–835 CE).

Fig. 20
Kōbō-daishi went to China in 804 C.E. and returned some years later laden with Buddhist icons and sutras. There he visited many monasteries and studied with the Esoteric Buddhist patriarch Hui-Kuo at Quin-long Si Temple at Chang’an. Three years later, he received a pair of mandalas which he brought back and are among the national treasures of Japan in the Toji Temple in Kyoto. The temple was given to him by the Emperor Saga (see Washio 1998).

One of the pair of mandalas is filled with hundreds of esoteric Buddhist deities including our Nyoirin Kannon (fig. 20). The six-armed bodhisatva is exactly similar to that painted in the Boston painting with identical iconographic features (see fig. 1 above) but differs significantly in style. Particularly different is the gesture of the uppermost right hand and the expressiveness of the face. Both display far more emotion in the Boston painting than in the Chinese mandala. Clearly, the earlier rendering in the mandala was meant to be a simple iconographical sketch rather than the eloquent, soulful and pensive bodhisatva of compassion, as rendered by the masters of the Boston scroll.

The unknown master who painted the superb and elegant Boston Nyoirin Kannon may well have been aware of Kōbō-daishi’s lamentation that “The teachings of Esoteric Buddhism described in the sutras are extremely mysterious; therefore, the only means to transmit them to the people is by using visual images” (Washio, p. 109). The detailed iconographic notes of the mandalas in the Toji catalogue are an excellent source for the identification of the vast pantheon of Esoteric deities as well as the earliest.

Dr. Pratapaditya Pal
Dr. Pratapaditya Pal is a world-renowned Asian art scholar. He was born in Bangladesh and grew up in Kolkata. He was educated at the universities of Calcutta and Cambridge (U.K.). In 1967, Dr. Pal moved to the U.S. and took a curatorial position as the 'Keeper of the Indian Collection' at the Museum of Fine Arts, Boston. He has lived in the United States ever since. In 1970, he joined the Los Angeles County Museum of Art and worked there as the Senior Curator of Indian and Southeast Asian Art until retirement in 1995. He has also been Visiting Curator of Indian, Himalayan and Southeast Asian Art at the Art Institute of Chicago (1995–2003) and Fellow for Research at the Norton Simon Museum in Pasadena (1995–2005). Dr. Pal was General Editor of Marg from 1993 to 2012. He has written over 70 books on Asian art, whose titles include, Art of the Himalayas: Treasures from Nepal and Tibet (1992), The Peaceful Liberators: Jain Art from India (1994) and The Arts of Kashmir (2008). A regular contributor to among other journals at the age of 87+, Dr. Pal has just published a biography of Coomaraswamy titled: Quest for Coomaraswamy: A Life in the Arts (2020).

References: endnotes to Part I

1. For a detailed study of Cintāmaṇi Cakra Avalokiteśvara, with special reference to Japan, see Helen B. Chapin ‘A Study in Buddhist Iconography’, Ostasiatische Zeitschrift, NO. 8 (1932), p. 29f and p. 111f; NO. 11 (1935), p. 125f and p. 195f.

2. For several variation in Japan, see Sawa Ryūken, Butsuzō Zuten, Tokyo, 1963, pp. 68-70.

3. A. Getty, The Gods of Northern Buddhism, Oxford, 1928, p. 95.

4. A. Waley, Catalogue of Paintings Recovered from Tub Huang by Sir Aurel Stein, London, 1931, p. xxxiv; Nanjio, 324; Taishō Tripitaka, 1080.

5. Taishō Tripitaka, 1081-1088.

6. Ibid., no. 1087; vol. 20, p. 213b.

7. For this and other stories and tradition regarding shrines of Nyoirin, see M. W. De Visser, Ancient Buddhism in Japan, Leiden-Paris, 1935, p. 636 f.

8. ibid., p. 644.

9. Ryuken, op. cit., p. 69.

10. S. Matsubara, Chinese Buddhist Sculpture, Tokyo, 1966, pl. 295.

11. Getty, op. cit., p. 96.

12. Waley, op. cit., pp. 8, 15 and 99.

13. ibid., p. 10.

14. ibid., p. 55.

15. The Macchendra Bahal painting of Cintāmaṇi Lokeśvara is, however, quite different. Cf B. Bhattacharyya, The Indian Buddhist Iconography, 2nd ed. Calcutta, 1958, p. 424 Fig. 94 (A).

16. Raghu Vira and Lokesh Chandra, A New Tibeto-Mongol Pantheon, pt. 8, New Delhi, 1963, p. 18.

17. Ibid., p. 39.

18. ibid., p. 40.

19. ibid., pp.2 and 19.

20. The most common appellation for such figures is Maitreya, for instance cf. Guide Book, National Museum of Korea, Seoul, 1964, Figs. 115-116. Both W. P. Yetts (The George Eumorfopoulos Collection Catalogue, vol. III, London, 1932, pp. 24-25) and A.C. Soper (Literary Evidence of Early Buddhist Art in China, Ascona, 1959, pp. 225-26) have quite convincingly shown that at least in some instances the figure must be identified with Śākyamuni.

For the identification with Nyoirin see. n. 2 above. S. Matsubara (ibid., pl. 118) labels such figures as Hanka-Shiyuzō, which is almost a literal description of the iconic features.

The identification with Bodhisattva Mahāsattva has been suggested by J.M. Rosenfeld, The Dynastic Arts of the Kushans, Berkeley and Los Angeles, 1967, p. 238ff.

21. Soper, op. cit., p. 226. It seems, however, that the motif of the ‘pensive bodhisattva’ was already known to the artists of Bharhut. In at least two instances there we see two seated figures with their left hand touching their left cheeks. One of these scenes represents the defeat of Māra (B.M. Barua, Barhut, Bk. III, Calcutta, 1937, pl. XXXVIII, fig. 33), and the other the Uraga-Jātaka (ibid., pl. LXXXI, fig. 116), where the seated pensive figure definitely portrays the Bodhisattva. Thus, the motif may well have had an Indian origin.

22. Foucher, A., L’art Gréco-Boudhique du Gandhâra, vol. II, Paris, 1918, Fig. 459.; The Beginnings of Buddhist Art, Paris-London, 1917, p. 147ff.

23. Rosenfield, op. cit., Figs. 90-91; p. 236.

24. ibid., Fig. 91.

25. Foucher, L’art Gréco-Boudhique, II, Fig. 408.

26. Rowland B., ‘Bodhisattva or Deified Kings: A Note on Gandhara Sculpture’, Archives of the Chinese Art Society of America, XV, 1961, p. 6ff.

Although Prof. Rowland rightly suggests that all such figures in Gandhāra cannot be identified as Bodhisattva—some probably representing lay donors and worshippers of other cults and sects—the reason that the Bodhisattva Avalokiteśvara is described apparently for the first time (p. 12) in the fifth century version of the Amitāyus Sūtra is not sufficient ground to preclude the possibility of his earlier artistic representation. Many of the Mahāyana or Vajrayāna deities are described in texts which are certainly posterior to their representation in art. Besides, that Avalokiteśvara had already gained eminence is evident from the fact that a whole chapter is devoted to him in Saddharma Puṇḍarikā, which was translated into Chinese in the 3rd century (cf. H. Kern, The Saddharma Puṇḍarikā, Oxford, 1884, pp. xxff, for a discussion of the date).

Cf. also Y. Krishnan, ‘Was Gandhara Art a Product of Mahayana Buddhism?’, The Journal of the Royal Asiatic Society of Great Britain and Ireland, Parts 3 and 4, 1964, p. 104ff.

27. Foucher, L’art Gréco-boudhique… Figs. 410-28.

28. For an explanation of the emblems of Avalokiteśvara is to be identified with Maitreya, it would appear that there were two different traditions of portraying him in these countries. It is of interest to note that no textual description seems to have been brought to light in China which indubitably identifies such figures with Maitreya. On the other hand, where Maitreya is definitely identified by inscriptions (cf. Wai-Kam Ho, ‘Three seated Stone Buddhas’, The Bulletin of the Cleveland Museum of Art, vol. LIII, 4 p. 98a), he is portrayed as a Buddha—despite the fact that in Indian texts he is always referred to as a Bodhisattva—seated in the European fashion (pralambapāda), his right hand displaying the abhayamudrā and the left resting on his thigh or lap. This again may have been a common iconic type, probably created in Gandhāra, which was employed for various Tathāgatas, and, consequently, unless the context is clear, all such figures should not indiscriminately be identified with Maitreya.

It is curious that no literary tradition in India is known to describe Maitreya, either as a Buddha or a Bodhisattva, seated in the European fashion and displaying the dharmacakra-pravartanamudrā, although the gesture is, indeed, apposite. In late Lamaist iconography, however, such a Bodhisattva holds both the wheel and the stūpa (cf. F.D. Lessing, Yung-Ho-Kung, Stockholm, 1942, pp. 87-89), and must certainly be identified with Maitreya. Here also the cakra is an unusual attribute for Maitreya, although not inappropriate, since, as the future Buddha, he is expected to preach the Dharma. It seems clear that the posture of Maitreya, viz. pralambapāda was of extraneous derivation—probably introduced by foreign artists in Gandhāra since it is not characteristically Indian—as is evident from the rather labored explanation that it indicates Maitreya’s desire to rise to preach the Law.

29. P. Pal, ‘The Iconography of Amoghapāśa Lokeśvara’, Oriental Art, n.s. XII, 4, pp. 234ff and XIII, 1, pp. 1ff.

30. For an explanation of the emblems of Avalokiteśvara, see M.T. De Mallmann, Introduction à l’étude d’Avalokiteśvara, Paris, 1948, p. 270.

Endnotes to Part II

1. It is amazing today how both Deniker and Getty were able to communicate with scholars in Europe and the Asiatic Society of Bengal (Kolkata) to do their research in so many languages without the computer. In any event, the second edition was brought out in 1928 but in limited numbers. I feel grateful that the second edition was reprinted by Charles E. Tuttle in Tokyo in 1962 which is the edition that has remained my bible since. Getty’s book has a very good bibliography citing works of earlier scholars, one of the most important being A. Foucher. Between 1905 and 1909 he published seven noteworthy books on Buddhist art and iconography. Another was A. Grüwedel with half a dozen between 1900 and 1912.

2. Although Banerjea’s book is mostly about Hindu iconography, he copiously mentions the Buddha (see 1956 index) and his studies of the iconometry of Buddha images in several appendices remain among the classics on the subject.

3. Bendall’s catalogue was followed by that of Foucher’s catalogue of paintings of Nepal and Tibet in the B.H. Hodgson Collection in the library of the Bibliothèque de l’Institute de France in Paris in 1897. All scholars of Buddhist iconography should remain grateful to Brian Hodgson, the medical officer in the British Residency in Kathmandu for saving Indian, Nepali, and Tibetan manuscripts without which Western scholars such as Bendall and others could not have begun the study of the subject.

4. Nepali sketchbooks are among the best sources for identifying Buddhist deities. See Pal 1985 for an extensive collection of such sketchbooks at Los Angeles County Museum of Art.

5. Getty 1962 (1928), p. VIII.

6. Getty 1962, p. 37.

7. To cite one or two instances, here are two versions of the very first word of the Hindu text Bhagavadgitā, dharmakshetre, according to Annie Besant “Holy Plain” (1924), and “on the field of truth.” In the Buddha’s sermons Dhammapada, Dhamma is translated as “righteousness” (1896), whereas in the introduction (John Ross Carter and Mahinda Palihawadana), the translators quote Frederick Max-Muller (1881) as stating “Dhamma has many meanings. Under one aspect it means religion … Under another aspect ‘dhamma’ is virtue, or the realization of the law.” Go figure!

8. In fact, he also missed my 1968 study above which justifies the reprint here.

9. Chandra 1986 v. 1. p. 244, no. 630. See p. 238, no. 610 for Maṇidhārin mentioned below.

10. Luczanits 2008, fig. 4, p. 48.

11. Conze 1964, p. 191 and 192. The translation is by D.R. Shackleton-Bailey.

12. Schroeder 2011, pp. 276–277. Pls. 90A and B. Both objects are now in the Li Ma Lakhang of the Potala in Lhasa, Tibet, according to Schroeder 2001, vol. 1, pp. 276–277, Pls. 90A and B. Schroeder attributes both to Eastern India and the late Pāla style (12th century). He is right about their date, but I think both were created in what was once Bengal in British India. Likely they were taken to Tibet either by Indian or Tibetan pilgrims for personal use and then deposited in a monastery and never duplicated in that country. It is worth noting that Tucci also published a photograph from the Bonardi collection of a  similar bronze from Sakya monastery in Transhimalaya, 1973, pl 151, which he identified as “The Bodhisatva Avalokiteśvara, 11th c. Pala school”. This sculpture is strikingly similar to our Fig. 13, one of the small bronzes in the Lima Lhakhang of the Potala.

13. See Bronkhorst 2011 for an extensive discussion of Buddhism and politics.

14. Pal 2003, p. 82, cat. no. 47. At that time it was not known that the painting was from Itum Bahal in Kathmandu. It was recently reported to have been stolen from Itum Bahal in 1979:  see Kantipur (daily newspaper in Nepali) : Iṭumbahālakā paubhācitra nyuyorkamai (Itumbahal's paubha painting in New York) Śrāvaṇa 25, 2078 (Aug. 9 2021) by Devendra Bhaṭṭarai.

15. Pal 1985, pp. 66 and 211, cat. no. P15.

16. Pal 1985, pp. 114–115, cat. no. S36 and p. 141, cat. no. S71

17.Conze 1964, p. 285. The name of the sutra is Upāsaka Sīla Sūtra translated in 428 CE. English translation by J. Takakusu.

18. Conze 1964, p. 196.

Bibliography (2021)

Banerjea, J.N. 1956. The Development of Hindu Iconography. Calcutta: University of Calcutta. Especially Appendices: 565–625.

Beer, R. 1999. The Encyclopedia of Tibetan Buddhist Symbols. Boston: Shambala.

Bendall, C. 1883. Catalogue of the Buddhist Sanskrit Manuscripts in the University Library, Cambridge, Cambridge University Press.

Bhattacharyya D.C. 1974. Tantric Buddhist Iconographic Sources. Delhi: Munshiram Manoharlal.
––––––. 1978. Studies in Buddhist Iconography. Delhi: Munshiram Manoharlal.

Bronkhorst, 2011. Buddhism in the Shadow of Brahmanism. London: Boston: Brill.

Bühnemann, G. 1994. Sādhanaśataka and Sādhanaśatapañacāśikā. Wien: Arbeitskreis Für Tibetische und Buddhistische Studien Universität Wien.
––––––. 2003. Buddhist Deities of Nepal: Iconography in Two Sketchbooks. Lumbini: Lumbini International Research Institute, Occasional Papers, 4.
–––––– and M. Tachikawa. 1991. Nipaṇṇayogāvalī: Two Sanskrit Manuscripts from Nepal. Tokyo: The Center for East Asian Cultural Studies.

Bunce, F. W. 1994. An Encyclopedia of Buddhist Deities, Demigods, Godlings, Saints and Demons. 2 vols. New Delhi: D.K. Printworld (P) Ltd.

Chandra, L. 1986. Buddhist Iconography of Tibet. 2 vols. Kyoto: Rinsen Book Co.
––––––. 1981. The 108 Forms of Lokeśvara in Hymns and Sculptures. New Delhi: International Academy of Indian Culture.
––––––. 1988. The Thousand-Armed Avalokiteśvara. New Delhi: IGNCA/Abhinav.

Fontein, J. And Pal P. 1969. Museum of Fine Arts Boston: Oriental Art. Boston M.F.A. and Greenwich, Connecticut: New York Graphic Society. 

Lowry, J. 1977. “A Fifteenth-Century Sketchbook (Preliminary Study),” in Essais sur l’art du Tibet, edited by A. Macdonald and Y. Imaeda. Paris: Librairie d'Amérique et d’Orient.

Luczanits, C. 2008. “Buddhism in a Cosmopolitan Environment,” Orientations, v. 39, no. 7, 46–52.

Mallmann, M. T. de. 1964. Étude iconographique sur Mañjuśrī. (Publications de l’École Français d'Extrême-Orient, 55). Paris: l’École Français d'Extrême-Orient.
 ––––––. 1975. Introduction à l’iconographie du tāntrisme boudhique. Paris: Bibliothèque de Centre de Recherches sur l’Asie Centrale et la Haute Asie, I.

Pal, P. 1985. Art of Nepal. Catalogue of the Los Angeles County Museum of Art Collection, LACMA and University of California Press.
––––––. 2003. Himalayas: An Aesthetic Adventure. Chicago: The Art Institute of Chicago in association with the University of California and Mapin Publishing.

Schroeder, U. von. 2001. Buddhist Sculptures in Tibet Vol. One: India and Nepal. Hong Kong: Visual Dharma Publications Ltd.

Tucci, Giuseppe, 1973, Transhimalaya. Series Achaeologia Mundi, series prepared under the direction of Jean Marcade, Professor of Archaeology at the University of Bordeaux, translated from the French by James Hogarth. Delhi: Vikas Publishing House.

Washio, R. 1998. Universe of Maṇḍalas: Buddhist Divinities in Shingon Esoteric Buddhism. Kyoto: Museum of Toji Temple.

Acknowledgement (2021)
I am indebted to the following persons for their generous help in preparing the above discussion: Rob Aft, Ian Alsop, Vajra Alsop, Christian Luczanits, Ulrich von Schroeder, Hope Flores, Nancy Rivera, and Sameer Tuladhar.

Download the PDF version of this article

Articles by Dr. Pratapaditya Pal | articles