| articles

Download the PDF version of this article

Halji 2012 update

A four-fold Vairocana in the Rinchen Zangpo tradition at Halji in Nepal

by Mimi Church and Mariette Wiebenga

October 21, 2008

Halji and the Rinchenling Gompa are threatened by glacier lake flooding caused by climate change. For an update, and for details on how to help preserver this ancient gompa please see Nepal's Oldest Himalayan Buddhist temple and monastery threatened by Floods.

Fig. 1
Halji is a village of some 85 households in the Limi (Tib. Sle-mi) [1] valley of northwestern Nepal. It lies directly south of Mount Kailash and Lake Manasarovar, separated from them by the Gurla Mandata massif. Figure 1 shows the location of the Limi valley in the context of the western Himalayan region. Figure 2 shows Halji village with a snow peak to the north. The village lies between steep rocky mountainsides and small terraced fields that descend to the river. The southeastern entrance to the village, in the lower right corner, suggests the village’s importance as a religious center, with walls and cairns of inscribed stones between two stupas that each arch over the trail (Tib. mchod-rten kyang-gnyis). While Limi falls just within the modern boundary of Nepal, in Humla (Tib. Wom-glo) district, access to the rest of Humla and Nepal is constrained by high passes with steep, treacherous trails. Until very recent times, Limi’s connection with the outside world was almost exclusively through the Tibetan Kailash area, where Limi enjoyed grazing rights until the 1960s, when they were revoked by agreement between the governments of China and Nepal. [2]

Given this setting, it is perhaps not so surprising to find a Tibetan Buddhist monastery in Halji, whose foundation has been attributed to the 13th-14th century, based on available Tibetan documents. [3] This monastery, called Rinchen Ling Gompa (Tib. Rin-chen gLing dGon-pa), belongs to the Drigung Kagyud (Tib. ‘Bri-gung bKa' brgyud), a sect that was active during this period around Kailash and throughout the western Himal, as far west as Ladakh (Tib. La-dwags). [4] The monastery, seen in Figure 3, sits towards the middle of the village. Its golden spires distinguish it from the surrounding houses, which are constructed of dressed stone and wood and are about the same height as the monastery.

Fig. 2

Fig. 3

The three-storied monastery structure, some 36x32 meters in extent, surrounds a rectangular open courtyard. The upper two stories of the northern side, which is opposite the main entrance to the courtyard, house the main rooms used for religious functions – assembly halls, protectors’ room, library, head lama’s quarters, and store rooms for ceremonial objects such as musical instruments and masks. The paintings and sculptures in these rooms follow an iconographic program consistent with a Drigung Kagyud establishment. The monks say that a number of these works of art were created at the time of the founding of the monastery, and the refined artistic style of some is generally considered consistent with a 13th-14th century date. The sculptures described in this note suggest that the monastery also houses pieces that may have been made several centuries earlier.

Fig. 4

With one exception, the ground floor on all four sides of the courtyard comprises monks’ quarters, kitchens, and storage rooms for mundane objects. The exception is a single room, approximately 6 meters on each side and 6 meters in height, underneath the assembly hall on the north side. This room houses a large four-fold image of Vairocana (Tib. rNam-par sNang mdzad), surrounded on three sides by a 1 meter high ledge with smaller images of seated attendant Bodhisattvas. [5] Figure 4 shows the central portion of the courtyard’s north side. The entrance to the Vairocana room is just to the left of the stairs. The remainder of this note on Halji monastery concerns itself with the contents of this room and their relation to what little is known of the early history of Limi valley.


Fig. 5
The Vairocana ensemble, which is made of clay and appears to have been recently repainted, dominates the room with its diameter of some 3.5 meters and height of 5 meters (from floor to top of spire above the figure). The images themselves are seated in lotus position and are slightly larger than life-size, with the hands of each forming the dhyana mudra. This position accords with one iconography of the Vajradhatu Mandala. [6] A flame nimbus surrounds the head of each image. Above each image a garuda holds a naga in its beak. The ends of the nagas descend to armatures that separate each image from the next, and where the nagas become makaras that adorn the tops of the armatures. Atop the central axis of the whole ensemble is a stupa with its white dome surmounted by a gold colored spire. Figure 5 shows the ensemble from the door on the room’s eastern wall; a simple wooden altar sits in front.

While the overall framing of each image is similar, and their upper torsos have broad shoulders tapering to a thin waist, there are differences between the number and details of their necklaces, the modeling of the torsos, the waist fastenings of the lower garments, the colors of the garments, and the modeling of the faces. In the current pigmentation, all of the bodies are white. The eastern and western images have red top garments, and the northern and southern have blue top garments; each of the lower garments has a different color scheme, using red, yellow, green, and light and dark blue. Whether this pigmentation reflects earlier coloration is not known. In all images the pectorals are well-defined, but the lower halves of the upper torsos are treated differently. The eastern and western images have slightly distended upper bellies that outline the rib cage; a belt covers the navel and holds the top of the lower garment in place. The northern and southern images have flat upper bellies, with slightly distended lower bellies, which are divided into four sections by horizontal and vertical scores that cross at the exposed navels; the tops of the lower garment were not visible. The faces on the northern and southern images seem a bit more rounded, with more prominent chins, than the eastern and western images.

Fig. 6

Figure 6, the Vairocana on the western side of the ensemble, and Figure 7, the Vairocana on the southern side, illustrate the differences in upper torso and face modeling.


Fig. 7

On three sides of the room that houses the Vairocana composition – the northern, eastern, and southern – a ledge about 1 m from the floor and 50 cm deep holds 25 clay images. The western side has a recess some 2 m long and 50 cm deep that contains storage shelves for Tibetan xylograph books (Tib. dPe cha). With the exception of a standing Avalokitesvara and 3 Buddha figures, the clay images on the ledge are slightly smaller than life size, some 75 cm high. These 21 images appear with the attributes of Boddhisattvas – on lotus thrones, seated with legs in lotus position, adorned with five pointed jewel crowns and wearing similar necklaces, arm and wrist bracelets, earrings, and garments. Their hands form very specific, and mostly different, mudras. Any objects they may have held have disappeared. No attempt was made to identify each image or the mandala to which they might belong. The consistency of the modeling of these images suggests that they were created at the same time, by the same artists; all have been recently repainted.

Fig. 8

No pattern could be discerned in the arrangement of the images, which are simply placed on top of the ledge. As seen in Figure 8, some small standing images of protectors have been placed in the corner next to one of the peaceful seated Bodhisattvas, apparently because of a visual similarity – both the Bodhisattva and protector figures have a raised and extended right arm.

The upper torsos of the Bodhisattva images display modeling characteristics similar to the northern and southern Vairocanas: molded pectoral muscles, flat upper belly, and a slightly protruding lower belly. Figures 9-13 show the diversity of mudras, along with variations in the details of modeling the navel and its surrounding flesh. Figure 9 shows one of the Bodhisattva images. Figure 10 shows the hands and belly of Figure 9; the spiral-inscribed navel is surrounded by a fleshy protuberance. Figure 11 shows another of the Bodhisattva images. Figures 12 and 13 show details of two other Bodhisattvas’ belly moldings. Both navels are surrounded by pronounced rhomboid protuberances; the navel in Figure 12 is inscribed with a spiral, while that in Figure 13 simply has a horizontal crease across a circular navel.

Fig. 9

Fig. 10

Fig. 11

Fig. 12

Fig. 13

The increasing popularity of Vairocana images, and translations of related tantras, beginning in the 8th-9th centuries is well known. This popularity spread from northern India into Tibet; through Central Asia via the Silk Road, to China and Japan; and even appeared at Borobodur in Java. Tantric texts describe several forms of Vairocana, and his images typically show him seated in a lotus posture, with hands in teaching or meditation mudras. These images may have one or four heads. The image with four bodies back to back, like that at Halji, is less frequently seen. A recent study of clay sculpture in the western Himalaya identified three examples of the four-fold Vairocana in this region. [7] All three examples – Tholing (Tib. Tho-ling), Tabo (Tib. Ta-po), and Lalung (Tib. Lha-lung) – are closely associated with the 11th century translator, Rinchen Zangpo (Tib. Rin-chen bZang-po), who made several trips from his home in Guge-Purang (Tib. Gu-ge Pu-hrang) westwards to Kashmir (Tib. Kha-che) and returned with artists, scholars, and texts. Many of his translation and temple construction projects were undertaken with the patronage of the Guge-Purang rulers. Rinchen Zangpo’s own translations from Sanskrit into Tibetan of root texts and commentaries related to Vairocana form a part of the Tibetan canon to this day.

Only two of the four-fold Vairocanas identified to date remain – Tabo and Lalung. Both of these sites are in the Spiti (Tib. sPi-ti) River valley, with Tabo on the west bank of the Spiti and Lalung on a nearby upstream tributary to the Spiti. The third four-fold Vairocana, at Tholing, the capital of Guge on the upper Sutlej, was destroyed during the Cultural Revolution; only the base of its throne remains. By the time the Vairocana temples at Tholing and Tabo were constructed, the larger Guge state comprised all of the areas traditionally associated with Rinchen Zangpo’s sphere of activities: the Spiti valley, Zanskar (Tib. Zangs-dkar), and Ladakh to the southwest of Guge; Kinnaur (Tib. Khu-nu) to the south; and Purang, the areas around Mount Kailash, including Khojarnath (Tib. Kha-char) and likely the Limi valley to the southeast. [8]

Construction of the temple at Tholing (the gtsug lag khang, or Yeshe ‘Od temple), which housed the Vairocana, is dated to 996, but it is not known whether the four-fold Vairocana was introduced at that time or in a later renovation. [9] At Tabo, the Vairocana ensemble is integrated stylistically and architecturally with a three dimensional sculpted and painted Vajradhatumahamandala, which can be dated by inscription to 1042. [10] At Lalung the much larger Vairocana composition has suffered considerable repair to the images and damage to its superstructure; now it occupies nearly all the space in an unornamented room, which is attached to a private house. Nearby, in the Serkhang temple, smaller figures are mounted on the wall to form several mandalas. Recent scholarship dates the smaller Serkhang figures to the last half of the 12th century, while the Vairocana figures are attributed to a somewhat earlier period. [11]


Fig. 14
The Halji Vairocana and Bodhisattva figures bear the most resemblance to the Tabo figures. [12] The overall size and composition of the two four-fold Vairocanas are similar, from the lotus throne to the top of the central spire, with some differences in detail. At Halji, each Vairocana has a flame nimbus, surmounted by a garuda holding a naga whose extremities become makaras when they reach the clay armature that separates the Vairocanas; at the top of the central spire, above the four garudas, is a stupa topped by a spire. At Tabo, the Vairocanas do not have a nimbus; they are surmounted by a cylindrical relief of four stupas that face the intermediate directions between the Vairocanas. Above this relief are four kirtimukhas, facing the same direction as the Vairocanas. The wooden armature that separates the Vairocanas once supported makaras that originated from the kirtimukhas; only their tails remain visible. [13] The elements in the composition are strikingly similar: separating armatures, makaras, garudas/kirtimukhas, and stupas, with a slight variation in the order of these elements on the spire. It has been speculated that the Tabo armatures once sheltered four goddesses, which are the only figures missing from the otherwise complete Tabo Vajradhatumahamandala. [14] In the Halji composition these positions are occupied by leaping horses, as shown in Figure 14.

In terms of modeling the Vairocana and Bodhisattva torsos at both Halji and Tabo, the overall proportions follow those attributed to “the Kashmiri aesthetic of about the 10th century…[:] broad shoulders with a slightly muscular chest flowing uninterruptedly to a narrow waist”, [15] even though the clay torsos lack the clear tripartite division found in Kashmiri bronzes of the period. [16] The Tabo and Halji Vairocana images are approximately the same size; the Tabo figures show the teaching mudra, while the Halji figures show the meditation mudra. The figures on the north and south in Halji, with their flat upper bellies and rounded faces with prominent chins appear more like the Tabo style than the eastern and western figures.

The bodies of the 21 surrounding Bodhisattva figures at Halji resemble the bodies of the 32 attendant figures mounted on the wall of the Tabo assembly hall. A cursory examination of the walls and figures at Halji did not suggest that they had ever been mounted, nor were there any indications of wall paintings in the Vairocana room. (Closer examination, particularly of the repainted figures themselves, might reveal otherwise.) The thrones of the mounted Tabo figures have a small radius, usually extending about half way down the thigh – a design that reduces the weight that must be supported by the wall mounting. The thrones of the Halji Bodhisattvas extend a bit beyond the knees, resulting in a radius approximately double that of the Tabo thrones and a proportional increase in the weight of the piece. No indication was observed that these might be replacement thrones, but closer inspection might reveal otherwise.

The torso modeling of the Halji Bodhisattvas resembles the Tabo figures, with broad shoulders, muscular chest, and tapering waist. The spiral inscribed in the navel of the Tabo figures [17] can also be seen in the Halji figures. The main difference between the Halji and Tabo torsos is the variety of modeling around the navels of the Halji figures. The area around the navels of the Tabo figures appears flat or slightly convex. The faces of the Halji figures appear more angular than the rounded Tabo figures.

A similar comparison with the figures at Lalung is less useful, in part because the four-fold Vairocana is less well preserved, and because the style of the Serkhang Bodhisattvas is known to be unusual and not comparable to clay sculptures of the region during the same time period. [18] This comparison of Tabo and Halji is not intended to suggest that the one influenced the other directly. Both could have been inspired by a third – the now destroyed Tholing, for example – or both could have been created by the same artists or related ateliers.

Evidence from the hagiographies (Tib. rnam-thar) of Rinchen Zangpo, the royal chronologies (Tib. rgyal-rabs) of the Guge-Purang kingdom, and catalogues (Tib. dkar-chag) of Khojarnath, the main Rinchen Zangpo temple in Purang, suggests that a temple associated with Rinchen Zangpo was constructed in Limi during the 11th, or possibly 12th, century. Local tradition in Halji provides similar evidence. The hagiographies of Rinchen Zangpo agree in associating him directly with the simultaneous foundation of three major temples – one in each of the three main regions of greater western Tibet (Tib. mnga’-ris skor-gsum): Khojarnath in Purang; Tholing in Guge; and Nyarma (Tib. Nyar-ma) in Maryul (Tib. Mar-yul) (Ladakh). These hagiographies, along with other histories, chronologies, and catalogues, usually list another 21 smaller places where Rinchen Zangpo founded temples. While the sources are far from complete agreement, Tabo and Purang Zher (Tib. Pu-hrang bZher) appear in several lists. [19] Most sources also note that Rinchen Zangpo constructed a total of 108 edifices, but do not provide a complete list.

Zher, located south of Khojarnath, and quite possibly near Sher, the modern Chinese immigration point from Limi and Humla, was Rinchen Zangpo’s main residence in Purang. [20] This area is only a two day walk from Halji. Two of the catalogues associated with Khojarnath include Limi as a site of one of Rinchen Zangpo’s 21 minor temples, although Limi is not mentioned in this context in other lists of these temples. [21] Zher was also the site where the master from Nepal, Ashwa dharma (Tib. Bal-po A-shwa-dharma), and the master from Kashmir, Wang kula (Tib. Kha-che Wang-ku-la), met to cast the famous silver Jowo (Tib. Jo-bo) for Khojarnath, probably in the first quarter of the 11th century. [22] It is certainly possible that they, or members of their entourage, traveled to Halji to sculpt the compositions that show traces of integrating two sculptural traditions.

During the 12th century, construction and renovation of religious establishments slowed considerably, at least partly because Guge was devastated by an invasion of Garlog from Central Asia. In the beginning of the 13th century, the influence of the Drigung sect began in western Tibet and in Halji, where it has continued to the present day, perhaps with some interruptions. “A pilgrim's guide to the Tise region reports that the Dri-gung Lama Spyan-Snga was given Limi by one of the Malla Kings (circa thirteenth century).” [23] The pilgrim’s guide is presumably the ‘Bri-gung Ti-se lo-rgyus, which records that the spyan-snga met the Yatse (Malla) king Grags-pa-lde near Manasarovar in the first quarter of the 13th century. [24] This spyan snga is Drigung lingpa Sherab Jungne (Tib. ‘Bri-gung-gling-pa Shes-rab ‘byung-gnas), who was active throughout western Tibet, including the furthest areas in Ladakh that had been influenced by Rinchen Zangpo. [25] By the third quarter of the 13th century, Drigung yogins were active throughout Purang, including Limi. The ‘Bri-gung Ti-se lo-rgyus contains a list of temples in which they were active; this list includes two in Limi: Sle-mi Til chen and Sle-mi Til kun-‘dzom. [26] The first is in Jang, the village to the east of Halji; its founding is attributed to Drigung lingpa. The second is in Til, the village to the west of Halji; its founding is attributed to Senge Yeshe (Tib. Seng-ge Ye-shes), a Drigung yogin who lived in various caves near Mount Kailash shortly before the arrival of Drigung lingpa. The Halji Tulku is considered to be a reincarnation of Senge Yeshe. [27]

From the early 13th century onwards, Limi was clearly in the sphere of Drigung influence, and it is unlikely that the Vairocana ensemble, whose roots lie in a different tradition, would have been created during or after this time. For activity in Limi before that period, there are a few hints in the literary tradition, outlined above; there is the obvious visual evidence; and there is the local tradition. The authors know of two publicly available sources of local tradition: one is the website of the KailashZone Charitable Foundation, [28] which was initiated with the active collaboration of Halji Tulku, and which contains considerable useful information on the history and culture of Limi; the second is a contemporary pilgrim’s guide to Kailash by a Humli scholar and spokesman for indigenous peoples. [29]

Both of these sources place the origins of the Halji temple in the 10th century and associate the first temple with Rinchen Zangpo; however, neither mentions the four-fold Vairocana. Indeed, this unusual piece is curiously neglected these days. It is in a dark room beneath all the other rooms that house fine sculptures and paintings; it is not shown to visitors unless one specifically asks. Yet the lay and monk attendants recognize its antiquity. The authors were told that it is the oldest part of the temple. A visitor who recently documented the monastery’s architectural characteristics was told that the Vairocana is as old as the temple itself, but recorded no characteristics of its room that would suggest its differentiation from the rest of the temple. This observer was also told that it “was built at the same time as Thuling Gompa in Ghuki Village in Tibet.” [30] This name could well be a phonetic transcription of Tholing in Guge.

KailashZone traces the temple’s name, Rinchen Ling, to the association with the Great Translator, but focuses its description of the temple on influences of the Drigung Kagyud order. [31] The pilgrim’s guide relates a tradition that the earliest temple was above the village; after it was destroyed by fire it was moved down to the village. [32] This tradition suggests an explanation for the current odd position of the Vairocana in the temple. The Vairocana and attendant Bodhisattvas could have been saved from the earlier structure and placed in the new temple, but not integrated into the overall iconographic composition of the new Drigung temple.

Fig. 15

Both local sources refer to Halji by its local name, Weltse / Waltse. KailashZone glosses this as “wel” meaning side, and “tse” meaning top. The authors have been unable to find other attestations where “wel” means side, although this could be a local meaning. Wel / wal is a homonym for the archaic Tibetan term dBal, meaning sharp, pointed. It forms the root of the name of a class of Bonpo deities, and is associated with Bonpo terms denoting a sacred structure or place, as in dBal gsas. Limi’s most sacred mountain and yul lha dominates the opposite side of the river from Halji (Weltse / Waltse / dBal tse ?). At the peak of the mountain is a sharp, pointed sword-like formation, and the peak itself is surrounded by groups of sharply pointed formations. Figure 15 shows the mountain with its sharp pointed forms as viewed from Limi’s Talung pasture ground.

Whatever the derivation of the name Weltse / Waltse may be, Limi is included as part of the ancient kingdom of Zhang Zhung, whose territory also included Ladakh, Zanskar, Spiti, and Kinnaur, as well as Guge and Purang. [33] And this brings us around again to the Rinchen Zangpo tradition. The sphere of activity of the Guge-Purang state and of Rinchen Zangpo in the 10th-11th centuries falls in these areas, where Zhang Zhung rulers supported Bonpo until they were conquered by the Tibetan empire of the 7th-9th centuries, and where Bonpo traditions remained after the disintegration of central Tibetan imperial power.

Beginning in the 20th century both Tibetan and western scholarship showed renewed interest in the relation between the Bonpo tradition and the religious and cultural reforms represented by the Rinchen Zangpo tradition. However, the evidence is scanty and scattered, and interpreted differently by various scholars. A strong local and oral tradition remains in Limi and in its immediate environs in Humla, although it is rapidly declining. There is said to be a catalogue (Tib. dkar chag) for Halji in the possession of Halji Tulku. And there are a number of historical and literary texts in Tibetan that document Bonpo and Buddhist traditions in the Kailash area. Halji may offer a unique opportunity to document cultural evolution from the earliest recorded time to the present.


Bidari, Keshav. 2004. “Halji Monastery – A Hidden Heritage in North-West Nepal”, Ancient Nepal, No. 155, June 2004, pp. 1-5.

Frederic, Louis. 1995. Buddhism: Flammarion Iconographic Guides, Flammarion, Paris.

Heller, Amy. 2003. “The Three Silver Brothers”, Orientations, Vol. 34, No. 4, April, 2003, pp. 28-34.

Jest, Corneille. 1981. Monuments of Northern Nepal, UNESCO, Paris, 1981.

KailashZone Charitable Foundation (

Klimburg-Salter, Deborah E. 1998. Tabo, a Lamp for the Kingdom: Early Indo-Tibetan Art in the Western Himalyas, Thames and Hudson, New York.

Lama, Tsewang, 2002. Kailash Mandala: A Pilgrim’s Trekking Guide, Humla Conservation and Development Association, Simikot.

Luczanits, Christian. 1998. “The Clay Sculptures” in Klimburg-Salter 1998.

____________. 2004. Buddhist Sculpture in Clay: Early Himalayan Art, late 10th to early 13th centuries. Serindia Publications, Chicago.

Namdak, Tenzin. 1997. sNga-rabs Bod kyi byung ba brjod-pa’i ‘bel-gtam lung gi snying-po. Paljor Publications, New Delhi.

Petech, Luciano. 1998. “Western Tibet: Historical Introduction” in Klimburg-Salter 1998.

Vitali, Roberto. 1996. The Kingdoms of Pu.hrang: According ti tmNga’.ris rgyal.rabs by mkhan.chen Ngag.dbang Tho.ling gtsug.lag.khang lo.gcig.stong ‘’i rjes.dran.mdzad sgo’i go.sgrig tshogs.chung; Dharamsala, India.

____________. 1999. Records of Tho.ling: A Literary and Visual Reconstruction of the “Mother” Monastery in Guge. High Asia (an imprint of the Amnye Machen Institute), Dharamshala, India.


1. Tibetan spellings, particularly of names, follow the conventions used by Roberto Vitali.  See Roberto Vitali.  Records of Tho.ling: A Literary and Visual Reconstruction of the “Mother” Monastery in Guge, (Dharamshala, India: High Asia (an imprint of the Amnye Machen Institute), 1999) and Roberto Vitali, The Kingdoms of Pu.hrang: According to mNga’.ris rgyal.rabs by mkhan.chen Ngag.dbang, (Dharamsala, India: Tho.ling gtsug.lag.khang lo.gcig.stong ‘’i rjes.dran.mdzad sgo’i go.sgrig tshogs.chung, 1996).

2. A recent guidebook to the Kailash region by a Humli scholar provides considerable information on the history and cultural traditions in Humla and Limi – Tsewang Lama, Kailash Mandala: A Pilgrim’s Trekking Guide (Simikot: Humla Conservation and Development Association, 2002).  Pages 77-81 summarize Limi’s trading and agricultural ties to the Kailash area.

3. Halji and the Limi valley were opened to tourists in the late 1990s, after many years of restricted access. However, visitors are rarely shown the Vairocana room. During the authors’ first visit to Halji, in 2002, they did not see it. During their second visit, in 2006, they requested to see it. The authors would like to thank the late Agu Lama and his family for their many kindnesses during trips to Humla and Limi, and specifically for sharing their extensive knowledge of local traditions. Thanks also go to Ian Alsop for pointing out the importance of this Vairocana and his ongoing stimulation of this investigation into its origins.

4. A member of an architectural conservation team recently documented the structural characteristics of the monastery – Keshav Bidari, “Halji Monastery – A Hidden Heritage in North-West Nepal,” in Ancient Nepal (No. 155, June 2004), pp. 1-5.  On page 1 of this article, Bidari includes general information about the monastery available from contemporary sources in English and from local informants.  He refers to page 48 of Corneille Jest’s catalogue of monuments in northern Nepal (Corneille Jest, Monuments of Northern Nepal (Paris: UNESCO, 1981) for this historical information.  Jest’s catalogue is unavailable to the authors at present.  Presumably, his dating is based on the ‘Bri-gung Ti-se lo-rgyus, which details the activities of the Drigung Kagyud in the Kailash area.

5. During the authors’ brief visit to this room in May 2006, time did not permit detailed recording of the sculptures or measurement.

6. See Louis Frederic, Buddhism: Flammarion Iconographic Guides (Paris: Flammarion, 1995), p. 129, for a summary of the several iconographies of the Vajradhatu Mandala.

7. Christian Luczanits, Buddhist Sculpture in Clay: Early Himalayan Art, late 10th to early 13th centuries (Chicago: Serindia Publications, 2004), pp. 207-8.

8. A number of sources discuss Rinchen Zangpo’s activities in relation to the history of Guge during the 11th century.  Two recent sources present this discussion in the context of Rinchen Zangpo temples known to have been under construction or renovation in the 11th century.  Luciano Petech, “Western Tibet: Historical Introduction” in Deborah E. Klimburg-Salter,  Tabo, a Lamp for the Kingdom: Early Indo-Tibetan Art in the Western Himalyas, (New York, Thames and Hudson, 1998).  See pp. 232-233 for a summary of the extent of the Guge state.  Vitali, Records of Tho.ling.  See pp. 14-16 for the phases of temple building in Guge.

9. Vitali, Records of Tho.ling, p. 14.

10. Luczanits.  2004.  Buddhist Sculpture in Clay: Early Himalayan Art, late 10th to early 13th centuries, p. 44.

11. Luczanits.  2004.  Buddhist Sculpture in Clay: Early Himalayan Art, late 10th to early 13th centuries, p. 93 and p. 106.

12. Klimburg-Salter.  1998.  Tabo, a Lamp for the Kingdom: Early Indo-Tibetan Art in the Western Himalyas, see Fig. 61 for illustration of the Tabo four-fold Vairocana and Figs. 66-97 for attendant Bodhisattvas.

13. Luczanits.  2004.  Buddhist Sculpture in Clay: Early Himalayan Art, late 10th to early 13th centuries, Figs. 32 and 296.

14. Klimburg-Salter.  1998.  Tabo, a Lamp for the Kingdom: Early Indo-Tibetan Art in the Western Himalyas, p. 104; and Luczanits.  2004.  Buddhist Sculpture in Clay: Early Himalayan Art, late 10th to early 13th centuries, p. 208.

15. Klimburg-Salter.  1998.  Tabo, a Lamp for the Kingdom: Early Indo-Tibetan Art in the Western Himalyas, p. 98.

16. Christian Luczanits.  1998.  “The Clay Sculptures” in Klimburg-Salter, Tabo, a Lamp for the Kingdom, p. 201.

17. Luczanits, Buddhist Sculpture in Clay, p. 50.

18. Luczanits. Buddhist Sculpture in Clay, p. 106.

19. Luczanits. Buddhist Sculpture in Clay, p. 26, and Vitali, Records of Tho.ling, p. 20, compare various sources.

20. Vitali, The Kingdoms of Pu.hrang, p. 362.

21. Vitali, The Kingdoms of Pu.hrang, n. 405.

22. Vitali, The Kingdoms of Pu.hrang, pp. 263-264, and Amy Heller.  2003.  “The Three Silver Brothers” in Orientations (Vol. 34, No. 4, April, 2003), pp. 28-34.

23. Bidari, “Halji Monastery – A Hidden Heritage in North-West Nepal”, quoting Jest, Monuments of Northern Nepal, p. 48.

24. Vitali, The Kingdoms of Pu.hrang, pp. 416-417.

25. Vitali, The Kingdoms of Pu.hrang, p. 389.

26. Vitali, The Kingdoms of Pu.hrang, n. 683.

27. The information on Jang ( and Til ( temples comes from the website of the KailashZone Charitable Foundation (  It is not clear from the website whether the information comes from literary or from oral traditions.


29. Lama, Kailash Mandala: A Pilgrim’s Trekking Guide.

30. Bidari, “Halji Monastery – A Hidden Heritage in North-West Nepal”, pp. 1-2.


32. Lama, Kailash Mandala: A Pilgrim’s Trekking Guide, p. 81.

33. Tenzin Namdak, sNga-rabs Bod kyi byung ba brjod-pa’i ‘bel-gtam lung gi snying-po, (New Delhi: Paljor Publications, 1997), p. 24.  The complete list is sBal-te rKang-phran, La-dwag, Zang-mkhar Gar-zha Nyung-ti sPi-ti Khu-nu tshang-sa Drug-nyi Nyi-ti kyo-nam Sha-khog mGar-yang tShang-ro Ti-dkar Sle-mi dang Wom-glo Se-rib Dol-po Krug-skyes sogs Kashor-mtshams Mang-yul Yan gyi lho-rgyud dang Ru-thog Gu-ge sPu-hreng

Download the PDF version of this article | articles