About the 18 stupas and other treasures once at the Densatil monastery by Jean-Luc Estournel
text and images © the author except as where otherwise noted

September 29, 2020

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

Since the end of the 1960s, public and private collections around the world have housed an important group of Tibetan objects with a very strong typological, iconographic and stylistic unity, constituting a particular group among all those recorded to come from the “land of snows”.

Fig. 1 In 1993, I was given for the first time the opportunity to lay the foundations for the association of this style with the photographs taken at Densatil (gdan sa mthil) by P.F. Mele during the expedition led by Professor Giuseppe Tucci of the Italian ISMEO in 1948. Rapidly written for the needs of an auction catalogue from the few sources then available, this text quickly proved to be incomplete and erroneous on certain points, but it was the starting point of the "Densatil Project" which aims to collect all the photographs, publications and information available since then.

Densatil monastery, the origins of which can be traced to ca. 1198, was perhaps first founded by the monk Dorje Gyalpo (rdo rje rgyal po) who, nearly half a decade ealier in 1158, had settled the area of Phagmodru (phag mo gru).

It is famous for its eighteen stupas housing the relics of the lineage of eighteen successive abbots once built in the main hall (gtsug lag khang), and more particularly the eight tashi gomang stupas (bkhra shis sgo mang) partially visible on the few 1948 photographs. (Fig. 1)

After the fire and the complete blasting of the monastery in the 1960s, many fragments of Densatil's treasures were scattered in various public and private collections around the world, as well as in various other Tibetan monasteries, at Lhasa (Potala and Jokhang), in Drigung ('bri gung) , Samye (bsam yas) or Reting (rwa sgreng) among others. This last point demonstrates that it is necessary today to avoid taking into account the present location of an object in Tibet as an indication of its provenance from that same place.

Among the objects that escaped being sent to smelters to recover the copper, some reached us after being burned to remove their original gilding. The semi-precious stones with which they were inlaid have also often been removed. The numerous fractures to the various parts of their bodies are the result of the dynamiting that caused the collapse of the tashi gomang stupas on themselves and under the walls and upper floors of the shrine.

The details of the history of Dorje Gyalpo Phagmo Drupa and the Lang (rlangs) clan who ruled Tibet from the palace of Neudongtse (sne'u-gdong-rtse) during the 14th and 15th centuries as the Phagmo Drupa dynasty and whose members succeeded one another on the abbot’s throne of Densatil are well known today and have been the subject of an important and remarkable study by Olaf Czaja [1].

The exhibition at the Asia Society in New York in 2014 [2] revealed to the general public the history, and the artistic richness of this high place of history and spirituality, without however really attempting a real approach from an art history perspective.

The development of the "Densatil Project" now makes it possible to draw up a chronology of the famous eighteen stupas, in the light of history, photographs from 1948, and hundreds of listed objects coming from the eight tashi gomang stupas ornamentation.

The classification of the numerous objects from the eight tashi gomang stupas according to stylistic and iconographic elements and the analysis of the treatment of details such as jewellery and lotus pedestals has made it possible to highlight eight globally homogenous groups that should therefore in principle correspond to the eight tashi gomang stupas and to make it possible today to try to propose attributions of a certain number of these objects to their original stupas.

Fig. 2For ease of reading by non-specialists, as the names of tibetan characters and places are complex, all those mentioned several times in the text have been transcribed here according to a simplified phonetic system. Those appearing only once have been left in classical transliteration. Moreover, in the majority of cases, only the Sanskrit names of the deities have been retained and also transcribed according to the most common simplified phonetic system.

Of Densatil's eighteen stupas, we can so far only be certain of the form of thirteen.

A first group is composed of five large stupas, belonging to the ancient typology known as the “mahāparinirvāṇa stupa” in memory of the mahaparinirvaṇa of the Buddha in Kuśīnagar which is known in Tibet as bka ‘gdams mchod rten by association with the Kadampa school and the person of its founder, the Indian master Atisha (982-1054). (Fig. 2)

Fig. 3 This archetype was very common in Tibet between the 12th and 15th centuries, a period during which many specimens were made in sizes ranging from a few centimeters to more than three meters high. For example, the one that would have housed the relics of Naropa (1016-1100) (na ro pa gdung rten) at Nyethang (snye thang) monastery would have measured 3 meters and 22 centimeters high [3]. Shakabpa evokes this tradition which would have been introduced in Tibet by Atisha. [4] The omnipresence of this stupa associated with the multiple portraits of Atisha clearly indicates the importance of this symbol in his tradition and spiritual lineage. (Fig.3)

According to Shakabpa [4], Spyan snga tshul khrims 'bar (1033/8-1103) would have built a stupa decorated with jewels on this model. It could be a stupa that he had built for his master 'Brom ston (1004-1064), himself a direct disciple of Atisha, in his monastery of Lo at Taktse (stag rtse) in northeast Lhasa.[5] Shakabpa also indicates that many nepalese and tibetan craftsmen are said to have made a very large number of them out of metal alloys.

Fig. 4 A second larger group consists of the eight most remarkable and monumental ones, falling under the category of “Bahudvara stūpa” commemorating the first sermon of the Buddha at Sarnath, which is known in Tibet as gomang (sgo mang) or tashi gomang (bkhra shis sgo mang).

In the Tibetan world, the gomang generally takes the form of a stepped base most often with stepped edges on which are signified symbolic doors or chapels housing deities, supporting the classical stupa dome surmounted by its parasols. There is a great variety of them, of all sizes and in all possible materials.

The very large stupas with multiple interior chapels that flourished in the 14th and 15th centuries throughout Tibet such as those of Gyantse (rgyal rtse), Narthang (snar thang), Jonang (jo nang), Jampaling (byams pa gling) or Gyang, more commonly known under the generic name of Kubum (sku 'bum) are the most monumental examples of existing tashi gomang. The two terms can be considered synonymous.

The building of a tashi gomang to house the relics of a great master is attested very early in the Tibetan tradition. One of the oldest of which it is possible to find the trace was probably the earthen one having sheltered part of the relics of Atisha built shortly after 1054 in the sku 'bum lha khang of Nyetang (snye thang) photographed by Charles Bell in the 1920s. [6] (Fig.4)

Fig. 5 The oldest richly ornamented tashi gomang of which we can be aware is undoubtedly the one most probably erected during the last quarter of the 11th century for the relics of Rkyang pa chos blo of rgyang ro spe'u dmar studied and photographed by Professor Tucci in the 1930s in Samada (khyang phu). Rgyang ro spe'u dmar would be the ancient name of the shrine [7] and Rkyang pa chos blo its founder. The photographs show us beside a huge Kadampa type stupa, a several meters high tashi gomang structured in several decreasing levels supporting a form of mahaparinirvaṇa stupa. (Fig. 5) The rich decoration developed on the whole surface in "gilded bronze" with figures under arks supported by small pillars. It restores a Vajradhatu mandala with its thirty-seven main deities. It is known that Rkyang pa chos blo received the initiation to this cycle directly from Rinchen Zangpo (rin chen bzang po) (958-1055) [8] who was himself a disciple of Atisha.

The Blue Annals tells us that at the end of the eleventh century, part of Marpa's relics (mar pa lo tsa' ba chos kyi blo gros – 1012-1097) were placed inside a tashi gomang made of acacia wood that his son lost in gambling. The relics were then removed and placed in a square box [9]. A tashi gomang is also thought to have been erected in Ralung (rva lung) between 1216 and 1225 to house the relics of Tsangpa Gyare (gtsang pa rgya ras) (1161-1211) [10] but apart from the fact that artists may have been involved in its ornamentation, no precise description of its iconographic structure has come down to us. In 1153, a golden tashi gomang was erected by Sgom pa tshul khrims snying po (1116–1169) to house the heart of Gampopa (sgam po pa) (1079-1153) [11]. Once again, no description is known, and it is therefore impossible to extrapolate on its precise structure.

Shakabpa speaks at the same time of tashi gomang stupas and of multi storied Kadampa tombs. He also evokes gold and silver tashi gomang in Narthang, Drigung and Densatil. [12]

It is therefore clear that the term tashi gomang can refer to a number of variations of the same archetype, with different sizes, materials or exterior finishes and iconographic program. Consequently, the mention of the creation of a tashi gomang without further details in a historical account does not systematically induce, even in the Drigungpa or Phagmodrupa specific context which will be the one that will interest us here, that it is an example of the same typology as those photographed in 1948 in the main hall of Densatil.

What remains is a group of five stupas for which little information has reached us, but which have had to take on different forms.

If this group of eighteen stupas recorded at the beginning of the 20th century [13] corresponds to a lineage of the first eighteen abbots of the monastery, its chronological sequence takes place over a period of four hundred years, between the death of Dorje Gyalpo Phagmo Drupa in 1170 and that of Dragpa Jungne (Grags pa ’byung gnas -1508 –1570).

We will therefore try here to restore this chronology by attributing to each abbot the stupa that must have been his, and to each stupa, its shape, its potential decoration and, when possible, the sculptures that must have been placed in it.

Twelfth century stupas

Abbot 1 / Stupa 1. Dorje Gyalpo Phagmo Drupa (rdo rje rgyal po phag mo gru pa) (1110-1170)

Dorje Gyalpo Phagmo Drupa was the founder of the future monastery, which during his lifetime consisted only of a group of huts. He managed to make the community prosper by receiving gifts and offerings, sometimes from very far away.

When he died in 1170, his heart which had remained intact after the cremation was placed in a stupa called tashi wobar (bkra shis 'od 'bar), which means "Radiating Light of Auspiciousness"[14]. It was a large mahaparinirvaṇa stupa of the Kadampa type mentioned above, as was common practice at the time. The representations of the sun and the moon at the top would have been made of gold, or at least gold-plated.[15] According to multiple concordant sources, Taklung Tangpa Tashi Pel (Stag lung thang pa bkra shis dpal) (1142-1209/10) paid for this stupa while Jigten Gonpo ('Jig rten mgon po) (1143-1217) sent Mgar dam-pa cho sdings pa to Nepal to commission a golden copper parasol to shelter it.[16]

Fig. 6 Despite of this installation in a hut, the disciples did not agree to build a real sanctuary on the spot. It was only in 1198, after a long debate between Jigten Gonpo and Taklung Tangpa Tashi Pel, that it was finally decided to build a shrine around the original hut of their master. From then on, the tashi wobar, the first of the series of eighteen commemorative stupas, was housed inside the tsuglagkhang (gtsug lag khang), along with the hut and the first treasures of the community. In 1948, it stood with four other stupas of the same type and size on a tiered altar in the center of the east wall of the tsuglagkhang. Unfortunately, only the third, fourth and fifth are visible, textile wrapped on the famous photograph by P.F. Mele. (Fig.6)

An attempt to estimate the height of these five stupas suggests that each may have been about 1.8 to 2 meters high. It is interesting and unusual to note in the photographs that all five of them, as well as those that made up the final piece of the tashi gomang stupas, are wrapped in protective textiles and that only their tops decorated with the scarf, lotus button, lunar and solar symbols and a triple jewel remain visible and appear extremely shiny and well-kept. This could confirm that these final parts may indeed have been gilded or made of gold. The recurrent mention by Densatil's visitors of the existence of gold and silver stupas could be a clue to the fact that these Kadampa-type stupas may have been made of silver. This could explain why, while many of objects from Densatil are known today, none of these stupas have been found, even in a fragmentary state.

In the years following the construction of the monastery, this tashi wobar, was to be at the heart of an incredible political and artistic adventure due to Jigten Gonpo.

Thirteenth Century Stupas

Abbot 2 / Stupa 2. Jigten Gonpo (’jig rten mgon po) (1143-1217)

Fig. 7 Jigten Gonpo served as second abbot of Densatil from 1177 to 1179 before leaving to develop his own monastery in Drigung (‘bri gung mthil).

Watching over the maintenance of their master's monastery from afar, Jigten Gonpo and Taklung Tangpa Tashi Pel, the main disciples of Dorje Gyalpo Phagmo Drupa, left the monastery without an abbot for twenty-nine years, until 1208.

After the construction of the main hall of Densatil in 1198, Jigten Gonpo went to meditate at Gampo (sgam-po) in Dagpo (dvags po), and there had a vision of the "pure crystal mountain" of Tsāri with Chakrasamvara ('khor lo bde mchog) in his palace, the whole surrounded by 2800 deities, organized like a monumental tashi gomang. This vision of Chakrasamvara in a tashi gomang was perpetuated in a simplified manner on numerous thangkas illustrating the Tsāri pilgrimage on which the stupa takes the more classical Tibetan form, simply covered with small painted doors. (Fig.7).

On the other hand, the large number of paintings associated with the monasteries of Drigung and Taklung depicting Chakrasamvara, sometimes between the Jigten Gompo’s or Taklung Tangpa Tashi Pel footprints [17] highlights the importance of this deity for the disciples of Dorje Gyalpo Phagmo Drupa and illuminates the deep meaning of Jigten Gonpo's vision.

Back in Drigung, he decided to build the monument he had visualized, in memory of his master Dorje Gyalpo Phagmo Drupa. The construction of this tashi gomang seems to be the result of the work of two main protagonists, Tshul khrims rin chen who would have been in charge of the structure, and a Nepalese artist named Manibhadra specialized in decoration and ornamentation in precious material.[18] Despite a 1202, provisional consecration made by Jigten Gonpo, who thought to die soon, the construction work seems to have lasted until 1208.[19] We do not know much about this construction except that it would have been erected first of all in wood, then in precious materials. [20] This last information confirms the intervention of the two personalities mentioned above.

We have seen above that the design of a tashi gomang structured in tiers, decorated with chapels housing deities and surmounted by a mahaparinirvaṇa stupa is attested in Tibet at least since the eleventh century. Jigten Gonpo cannot be regarded as a true innovator in this respect. The particularity of his monument was to be found not in its form, but in the iconographic program designed to cover it.

The iconographic scheme of the monument resulting from Jigten Gonpo's vision has come down to us in some texts that have been widely studied by Olaf Czaja and Christian Luczanits ([21].

These works present some variants among themselves, thus demonstrating that they do not all describe exactly the same monument or reproduce exactly the same original text. Relying on serious iconographic arguments linked to the appearance and diffusion of the initiatory cycles, which it would take too long to repeat here, Christian Luczanits comes to the conclusion that these texts can in no way be contemporary with Jigten Gonpo as Olaf Czaja thinks, but at least a century later. He hypothesizes that these descriptions may in fact be inspired by one of Densatil's tashi gomang. In the latter case, the concordance of an announced number of 2,800 figures could suggest that this model could be the first Densatil tashi gomang erected from 1267 for Dragpa Tsondru. (Grags pa brtson 'grus)

Fig. 8 It is therefore clear that this literary description of the tashi gomang visualized by Jigten Gonpo is most likely slightly different from the original. We must also take into account the fact that Densatil's monuments counted a very variable number of deities and agree that the text considered as initial should only be considered as a basic grid that may have known slight variations on each monument. These variations are probably related to the initiatory cycles associated with each of the abbots for whom the stupas were intended.

Our purpose here will not be to go once more into all these structural and iconographic details, but simply to use them to locate the places where the sculptures we will be talking about should have been.

We will here simply refer to the 6-tiers structure (Fig.8) corresponding respectively from top to bottom:

- At the top, Tier 1, around a huge Kadampa style mahaparinirvaṇa stupa, are arranged in masters lineages.  The text chosen by Olaf Czaja associates a representation of Vajradhara in the centre of each face of the stupa. The one studied by Christian Luczanits favours, for clearly stated iconographic reasons, two representations of Vajradhara to the east and south, Chakrasamavara to the west and Hevajra to the north.  The simple examination of P.F. Mele's photographs showing the presence of a representation of Chakrasamavara at the top of a tashi gomang leads us to consider the text chosen by Christian Luczanits as probably more precise. Three aspects of Vajrayogini/Vajravarahi are indicated to be placed in front of each of these four main deities. It is interesting to note that on the south side of this Level 1, we find at the foot of the Kadampa type stupa, a representation of Atisha founder of the Kadampa school among Tilopa, Naropa, Marpa, Milarepa, and Gampopa. This probably should be seen as an expression of the old and strong connection which will be maintained throughout history between the Phagmo Drupa abbots and those of the Kadampa school even after the reform of the latter by Tsongkhapa and until the reunification of Tibet by the 5th Dalai Lama. 

- Tier 2 presents various cycles of Anuttarayoga Tantra including the 72 deities mandala of Guhyasamaja in the east, the assembly of Vajrakila in the south, that of Hevajra in the west, and the 62 deities mandala of Chakrasamvara in the north. An aspect of Avalokiteshvara is arranged at each of the four corners.

- Tier 3 presents on the east face the Vajradhatumandala in 47 deities, on the south, the assembly of Buddhakapala, on the west, the assembly of Prajnaparamita and on the north, the mandala in 14 deities of Chakrasamvara. An aspect of Acala is arranged at each of the four corners.

- Tier 4 presents in the center of each side a Buddha surrounded by two of the eight Bodhisattvas of the bhadrakalpa and two wrathful deities on each side. They are surrounded on each side by 250 representations of the Tathagata of their respective direction. A dvarapala is arranged at each of the four corners.

- Tier 5 is dedicated to goddesses. On each side are three central goddesses flanked by the group of the sixteen goddesses of sensual enjoyment bearing offerings. The four main goddesses occupying the center of each side are Parnashavari in the east, Eight armed Tara in the south, Dhvajagrakeyura in the west and Vasudhara in the north.

- Tier 6 is the lowest level. On each face, on either side of a central lotus stem, we find an aspect of Mahakala and one of the goddess Lhamo, themselves flanked by two deities of wealth and two nagarajas closing the section.  They are all arranged in the foliage of scrolls supporting the lotus flower from which the rest of the monument seems to emerge.

To sum up, the tashi gomang model created by Jigten Gonpo must be considered, as described by Christian Luczanits in the title of his article, as a "mandala of mandalas", a veritable tangle of universes of deities (2800 in total), referring perfectly to Professor Tucci's description during his visit to Densatil in 1948, "The whole Olympus of Mahayana seemed to have been assembled on those monuments". [22]

This particular type of tashi gomang which must be considered as the "Reliquary Stupa of Many Auspicious Doors for Phagmo Drupa" will then be declined to commemorate some deceased abbots of Drigung and Densatil since they were the depositories of the direct lineage of the master. While very little is known about the possible tashi gomang erected in Drigung after that of Dorje Gyalpo Phagmo Drupa, the analysis of those of Densatil allows us to note that if the general iconographic plan has globally remained the same over time, variations have been made on some of them. The numbers of divinities indicated by the texts in fact imply that some mandalas constituting certain tiers have been modified or replaced by more complex ones, while perhaps retaining the same basic divinities. These variations are probably due to the particular cycles studied by the concerned abbots.

Being large mandalas, the convention is that the face of the tashi gomang stupas exposed to the viewer should always be the eastern face, even if this does not correspond to the actual geographical orientation. As a result, P.F. Mele's photographs essentially show us the eastern face of the monuments. The south side should therefore be considered to be on the visitor's left, the west on the back, generally facing the wall, and the north on the viewer's right.

The fact that Jigten Gonpo chose to build this memorial stupa to his master in his own monastery rather than in Densatil, which would have been his logical location, is historically interesting. Indeed, it may suggest that the abbot of Drigung may have attempted to shift the power of Densatil's spiritual center from its abandoned location to his own monastery, to gain political advantage over other powerful contemporary buddhist schools.

This became obvious when, in 1208, he tried to empty Densatil of his sacred substance by sending the library to Gampo, supposedly to fulfil his master's wish, and above all by transferring the commemorative stupa tashi wobar containing the heart of Dorje Gyalpo Phagmo Drupa to Drigung in order to place it at the top of the complex structure he had built for this purpose and thus complete his Tashi gomang. [23]

What appeared to be a looting of the monastery of Dorje Gyalpo Phagmo Drupa did not fail to provoke a local revolt. To calm the spirits, Jigten Gonpo decided to return the tashi wobar stupa to Densatil with Dragpa Jungne (grags-pa 'byung-gnas -1175-1255), one of his main disciples as abbot. The latter belonged to the Lang (rlangs) clan, a powerful local family. He had to offer many gifts to the local nobility to calm the spirits.

Once the tashi wobar stupa returned to Densatil, Jigten Gonpo made a replica filled with other relics of his master to restore its greatness to his tashi gomang in Drigung.

Fig. 9 Apart from the iconographic scheme of 2,800 deities that covered it, we know nothing specific about this commemorative tashi gomang for Dorje Gyalpo Phagmo Drupa in Drigung. This is mainly because the monastery was completely destroyed by the Mongols and the Sakyapa in 1290.

In 1948, P.F. Mele photographed in Tsetang (rtse thang) a gilt copper plate depicting offering goddesses made in a 13th century characteristic style. (Fig.9) The shape and iconography of this object is characteristic of a tashi gomang for Dorje Gyalpo Phagmo Drupa decoration.

Considering first that Tsetang was only built in 1351, preventing de facto the object from being considered as originating in it, and secondly that only the Drigung or Densatil monasteries were able to build a tashi gomang of this particular typology during the thirteenth century, it is conceivable that this plate could be a survivor from the Dorje Gyalpo Phagmo Drupa's Tashigomang in Drigung preserved there after de complete destruction of the monastery as a sufficiently important relic to be exhibited and shown to Professor Tucci, who unfortunately has not published any note on the subject. [24]

When Jigten Gonpo died in 1217, a tashi gomang, possibly close to the one he had created for Dorje Gyalpo Phagmo Drupa, is reported to have been built for him in Drigung and a large commemorative stupa of the same shape and size as the tashi wobar called bkra-shis kun-tu 'od, was erected at Densatil, also with the sun and moon in gold or gold plated.[25] In 1948, it stood textile wrapped with four other stupas of the same type and size on a tiered altar in the center of the east wall of the tsuglagkhang. Only the third, fourth and fifth are unfortunately visible in the famous photograph by P.F. Mele.

Abbot 3 / Stupa 3.  Dragpa Jungne (Grags pa ‘byung gnas) (1175-1255)

Sent to Densatil by Jigten Gonpo, Dragpa Jungne occupied the abbot throne from 1208 to 1235 before leaving to ascend the temporarily vacant throne of Drigung from 1235 to 1255. This period is an important turning point in the history of Tibet which then had to face the Mongol invasions. The main Buddhist schools competed for the graces of the invaders and for local political advantage. The advantage was taken from 1244 onwards by the Khön clan who held the great monastery of Sakya.

Before leaving for Drigung in 1235, Dragpa Jungne took care to place his brother Dragpa Tsondru (Grags pa brtson 'grus - 1203-1267) on Densatil's throne, thus initiating a shift of dependence, which took Densatil from the tutelage of Drigung to that of his clan, the Lang.

Fig. 10 As abbot of Drigung, he would have been behind the building of two tashi gomang (probably for his two predecessors and direct successors of Jigten Gompo) and would have asked that another be built for him as well. [26] The mention of gemstone decoration unfortunately does not inform us about the exact structure of these tashi gomang stupas, namely whether they followed the archetype of the monument for Dorje Gyalpo Phagmo Drupa created by Jigten Gonpo with many cast figures or had a less developed votive character, perhaps more in line with the cost of such monuments in a complicated and financially less prosperous period than a few years earlier. Again, we will never know anything precise about these three Tashi Gomang of Drigung mainly because the monastery was destroyed by the Mongols and the Sakyapa in 1290.

At his death in 1255, his brother Dragpa Tsondru (grags pa brtson ‘grus) ordered for him a large commemorative stupa placed in Densatil in the same shape and size as the Dorje Gyalpo Phagmo Drupa's tashi wobar, called bkra shis 'od dpag.[27]

In 1948, it was placed, textile wrapped, with four other stupas of identical type and size on a tiered altar in the center of the east wall of the tsuglagkhang. It can be seen partially on the famous photograph by P.F Mele. (Fig.10)

Abbot 4 / Stupa 4 / tashi Gomang 1- Dragpa Tsondru (Grags pa brtson ’grus) (1203-1267)

Brother of Dragpa Jungne, Dragpa Tsondru ascended the throne of Densatil in 1255. The extent of his fame brought him many offerings from as far away as the Khasa Malla kingdom or present-day Sri Lanka. The Mongolian takeover of Tibet then placed Densatil and his region under the protection of Prince Hülegü Khan (1217-1265), grandson of Genghis Khan and brother of Kublai, who governed the areas that make up present-day Iran and Iraq.[28] The prince is said to have made important offerings to the monastery on three occasions, among them the famous six golden pillar ornaments that would have adorned the surroundings of the first tashi gomang erected in Densatil from 1267 for Dragpa Tsondru.[29]

Fig. 11 Apart from differing information on the dimensions of these six pillar ornaments we know nothing about their shape or decoration. Olaf Czaja suggests that one of them can be seen in a photograph by P.F. Mele hanging from a large pillar separating a tashi gomang from the altar supporting the five large kadampa-style stupas and the sculptures of the spiritual lineage of the abbots of Densatil. This proposal is unfortunately impossible to hold for two reasons. The first is that if these ornaments were indeed arranged around the first tashi gomang erected in Densatil in 1267 for Dragpa Tsondru, the monument visible in the photograph can in no way be identified as that of the fourth abbot of the place since it is stylistically attributable to the fourteenth century and not to the thirteenth. The second reason is that one only has to look at the photograph carefully to see that it is in no way a pillar ornament, but simply the Prabhāmaṇḍala to which the sculpture hanging from this same pillar is leaning. (Fig.11)

The realization of this first tashi gomang, probably following the model of the one erected in Drigung for Dorje Gyalpo Phagmo Drupa since it would also have been composed of 2,800 deities, is situated at a key moment in the history of the Lang / Phagmo Drupa clan.

In fact, with the definitive seizure of power over Tibet by the Sakyapa through the nomination of Phagpa (‘Phags pa - 1234-1280) by Kublai as viceroy of Tibet, the Sakyapa administration divided the territory into 13 myriarchies, one of which fell to the Lang, already strong with the prestige of their monastery of Densatil. To further assert their power, they erected the fortress of Neudongtse, which became the political seat of their dynasty.
It is conceivable that at that time the Lang clan, thanks to the wealth poured into the monastery by Hülegü Khan and other generous donors, was in a prosperous financial position to decide to undertake this exceptional construction in order to affirm the greatness of its temporal and spiritual power over the myriarchy and beyond.

Before considering what this first tashi gomang to Densatil might have looked like, it is essential to set out a number of points relating to the methodology used in the attempt to analyse the eight tashi gomang stupas that will follow.

In order to propose a reconstruction of the decoration of each of Densatil's eight tashi gomang stupas, we will use all the objects that can be considered stylistically and iconographically coming from them. At this stage, it is interesting to note that this large group of objects can precisely be divided into eight globally coherent sets highlighting a logical stylistic chronology.

Within each of these eight groups, it is interesting to note that beyond an overall stylistic unity essentially marked by the treatment of costumes, jewellery and ornaments, it is clear that for each of the eight tashi gomang stupas, several artists or workshops, each bringing its own particularities, had to work either simultaneously or successively on the construction of these monuments. It is conceivable that these various workshops have in fact worked for different sponsors, all anxious to acquire merits by offering sculptures for the ornamentation of tashi gomang stupas.

It also appears that the proximity of a tashi gomang with each other allowed artists to look at what their predecessors had achieved and to be inspired by it. This last remark explains why it is sometimes delicate, without having an object in hand, to attribute it to one tashi gomang rather than another, especially if they were built at very short intervals, making it possible to envisage that the same artists worked on them.

Another difficulty encountered for a precise positioning of each deity in the place that should originally have been his, is the fact that we are often faced with ancient iconography and little known aspects of certain deities, and that in the chaos that must have presided over the destruction of Densatil their attributes were often broken, making their exact identification even more delicate.

Last point, Densatil having housed thousands of sacred objects in addition to those that adorned the eight Tashi gomang stupas, objects probably often made by the same artists, it must be integrated that many of these works may not come from these eight great stupas. Thus, for example, the excessive number of Nagarajas, long considered as all coming from the tashi gomang stupas, is probably explained by the fact that they must have come from the bases of certain sculptures or altars that they were supposed to give the impression of supporting. It is also possible that many of them may have come from other shrines.

For each tashi gomang, we will approach the study, first of all by the figures of the four Guardian Kings (Lokapalas), which constitute a very interesting and significant group in the reconstruction of the stylistic and chronological evolution of the eight monuments. Each tashi gomang will then be considered level by level.

Fig. 12 Tibetan texts describe the tashi gomang visualized by Jigten Gonpo for Dorje Gyalpo Phagmo Drupa from the top (tier 1) to the bottom (tier 6). This is due to the descriptive logic of a huge mandala of mandalas that must start from the center which is necessarily at the top in a three-dimensional projection.

We prefer here to consider the study from the bottom to top, following the logical order of construction. Considering that the construction of a tashi gomang may have lasted many months, or even a few years, this will sometimes allow us to better understand the chronology of each one and the stylistic transitions from one to another.

It should be pointed out here that comparisons with P.F. Mele's photographs must take into account the fact that the majority of them were taken with wide-angle lenses, which strech the figures and monuments, and sometimes gives us a distorted view of them.

Apart from the fact that it should have been adorned with 2,800 figures like the one erected in Drigung for Dorje Gyalpo Phagmo Drupa, no more precise information has been received about this monument.  We will therefore try to reconstruct what it is possible to envisage from the elements which, for logical and stylistic reasons, cannot integrate any of the fourteenth and fifteenth centuries tashi gomang stupas.

Fig. 13 Among P.F. Mele's photographs, one showing us the Nagaraja Ananta, Vaishravana and an aspect of Mahakala could for various reasons be considered as the only photograph of this first tashi gomang in Densatil. (Fig.12). However, it cannot be totally excluded that this may be a detail of the eighth and final of these tashi gomang stupas. Nevertheless, one of the overall photographs of the monument, which for technical and stylistic reasons can only be considered as the eighth of these tashi gomang stupas, shows a wooden balustrade positioned very close to the large double lotus supporting the monument. (Fig.13) The positioning of this barrier, whose shadows can be seen on the petals of the great lotus, makes it difficult to imagine that the photograph showing us the three deities could have been taken behind it with so much space and light. On the other hand, another slightly blurred photograph taken from above partially shows the tops of the lotus petals, which appear to be more rounded and less pointed than what can be seen in the upper right corner of Figure 12.

In any case, we will see further on that these first and eighth tashi gomang stupas were probably arranged side by side and that the first seems to have in some ways considerably influenced the artists who worked on the last one.

This first representation of a Nagaraja supporting a lotus pedestal placed next to protective deities in loops of lotus scrolls following Jigten Gonpo's vision for the tashi gomang of Dorje Gyalpo Phagmo Drupa directly inspired by Indian archetypes of the Indian Pala period, refers to certain thangkas of the same time.

Fig. 14 Many of these paintings are clearly associated with the lineage of Dorje Gyalpo Phagmo Drupa and Jigten Gonpo. They are a good demonstration of artistic and aesthetic unity between the pictorial and sculptural works.

One of the best example is undoubtedly a remarkable portrait of a lama preserved in the Pritzker collection (Fig.14). In the lower register we see protective deities surrounded by lotus scrolls among which we find Vaishravana, Mahakala, Lhamo and Nagarajas supporting the throne alike they are found around the bases of tashi gomang stupas of the type we are interested in here.

It is therefore obvious that in the artistic movement developed around the visions of Jigten Gonpo, the same spectacular and extremely precious character was used to magnify the memory of the succession of masters of the spiritual lineage. We will again note the presence of Atisha in this lineage.

The precise identity of the abbot at the center of this remarkable painting, which can only come from a major sanctuary in view of its quality of execution, is not specified by any inscription.

Insofar as the Marpa, Milarepa, Gampopa, Phagmo Drupa and Jigten Gonpo lineage is perfectly identifiable in the upper register, the latter being represented in a larger size in the center, it is highly probable that this is one of his direct successors. Thus, if we consider this painting as coming from Drigung, only three abbots after him were his direct disciples, Tshul-khrims rdo-rje (1154-1221), Bsod-nams Grags-pa (1187-1235) and Dragpa Jungne (1175-1255). In a logic associated with Densatil, it could only once again be Dragpa Jungne (1175-1255).

Returning to the tashi gomang for Dragpa Tsondru, it should have housed all the sculptures attributable to a tashi gomang from an iconographic point of view and stylistically datable from the 13th century since, as we shall see later, there has been no other tashi gomang of the Dorje Gyalpo Phagmo Drupa type, built in Densatil before 1360.

Another hypothesis could be that this important group could come from one of Drigung's tashi gomang stupas. However, as Drigung was destroyed and burnt down in 1290, it is unlikely that such a group survived the sack of the monastery and remained grouped in such a relatively good state of conservation for seven centuries. The very high probability that these works come from Densatil itself is reinforced by the fact that Chinese scholars have found some of them in the ruins of Densatil.

The analysis of these sculptures reveals that different sculptors' workshops must have worked on the construction of this monument. This multiplicity of artists having worked successively or concomitantly seems logical if we consider that, as the texts report, each tashi gomang was made over periods of several months or even several years. An attempt to place each of the objects in the place normally assigned to it reveals that the two most active workshop seem to have generally worked one on the lower part and the other on the upper part. This is probably a confirmation of the chronology of the work. The consistency of this set is confirmed by the fact that none of the piece attributable to each workshop overlap with those of the others.

To try to illustrate the particularities of this first tashi gomang in Densatil, we will retain the following few works, knowing that we will limit ourselves here to evoking the dominant styles and that the secondary styles will be the object tashi gomang by tashi gomang of future publications.

The Four Guardian Kings

Fig. 15 One of the first pieces that can be attributed to this first tashi gomang is a Dhritarashṭra figure photographed by Michael Henss in the Jokhang of Lhasa together with two others from the fourteenth and fifteenth centuries. As the latter two are recognizable in the photographs taken by P.F. Mele at Densatil in 1948, the probability that this one also comes from Densatil is therefore very high.(Fig.15)

Fig. 16 The representations of these lokapalas will be very interesting in the course of this demonstration, since we have identified eight different types that can be attributed to each of the eight tashi gomang stupas and guide us in the chronology. His costume consists of a kind of armour combining textile, mail shirt, straps and ornamental plates following an iconography imported from China. Such representations can be seen in some paintings of the same period associated with Indo-Nepalese style deities such as we will meet mostly on the tashi gomang stupas (Fig.16). We note, however, that on this Dhritarashṭra, if the Chinese influence is present on the garment, the head and tiara can be judged as belonging to the same Indo-Nepalese style as the majority of the deities that will be associated with this monument.

In the best-known photographs of Densatil's tashi gomang stupas, the four figures of guardians appear arranged in front of the monuments, at the level of the protective deities and the great lotus. In the photograph that we can consider as representing this first tashi gomang for Dragpa Tsondru, the clear space in front of the three deities suggests that the lokapalas were not here arranged in this location. They were more likely directly rest on the great lotus. It is this layout that can be seen on the eighth tashi gomang, creating a new similarity between the two monuments, which were probably neighbors.

It is also possible to propose to attribute to this monument, level by level, many sculptures including the following few.

Tier six
The three deities visible on the P.F. Mele's photograph (Fig.12) are cast in a global Nepalese style. They are leaning against and fixed to gilded copper plates decorated in repoussé with secondary deities surrounded by lotus scrolls. One of them is known. (Fig.17)

Fig. 17 The left side shows us four deities holding a kapala and brandishing a kartrika, which according to Jigten Gonpo's description of Tashi Gomang for Dorje Gyalpo Phagmo Drupa, indicate that the deity who was to stand in front was undoubtedly an aspect of the goddess Lhamo (Prānasādhanā Śrīdevī). On the other part of the plaque was to be fixed a representation of the goddess Prithivi (Pṛthivī). This tells us that this plaque must have been on the east side of the monument, just like the three deities in the 1948 photograph, but at the other end of this side.

This back plate, executed in the exact style visible in P.F. Mele's photograph, also highlights the mounting system with a central joint and a hole for the deity's fixing hook, which is also seen in the photograph. (Fig.18)

Sculptures potentially attributable to this level were probably executed by several different artists. Many of them show a clear stylistic unity, essentially marked in the treatment of the bodies, tiaras with central anchor-shaped elements, armbands, bracelets and anklets, and garb elements and scarves.

We can attribute a number of figures to this level, including at least five Nagarajas now in the Capital Museum in Beijing. Two of them are illustrated here (Fig. 19). Note that each of them has a different costume and ornaments, which will also be the case on all the following tashi gomang stupas. A study of the set would perhaps make it possible to identify each of them by these costumes and jewels. The number of hooded snake heads above the nagarajas' bun will vary from tashi gomang to tashi gomang. That of the photograph taken by P.F Mele seems to have nine of them, like the sculptures in the Capital Museum in Beijing.

Fig. 18

Fig. 19

Fig. 20

Most of these sculptures were cast separately and then fixed to lotus pedestals of more or less different shapes as can be seen in the photograph by P.F. Mele.

A representation of Rahu kept at the Capital Museum in Beijing, (Fig.20) which can only come from this first tashi gomang, and therefore from its southern side, presents a stylistic variant which is expressed in the treatment of the ornaments with a motif in the form of a blooming flower forming a jewel as the main mark in the diadem. This particularity, which was probably applied to other deities at this level, at least on the south side, is also found on the tiaras of certain deities that we will find on tier 5 and inspired certain artists who worked on the 15th century tashi gomang stupas. His armbands seem to be closer to the model worn by the representation of Mahakala in P.F. Mele's photograph.

Tier 5

Fig. 21At this level were panels supporting on each side of the tashi gomang stupas the group of the sixteen goddesses of sensual enjoyment bearing offerings. These offering goddesses are linked to the Chakrasamvara suite, as can be seen in the lower register of a famous thangka from the Pritzker collection. (Fig.21)

These panels supporting the offering goddesses are present on each of Densatil's eight tashi gomang stupas and are easily reunited into eight distinct groups that can be associated with each of these eight stupas. The plates attributable to this first tashi gomang once again illustrate the diversity of the artists who worked on its construction. What they have in common is the lotus pedestal composed of 23 to 25 petals, depending on damages and what can be seen in the photographs. This is the only series to have so many lotus petals in such large numbers, which confirms their common origin.

Their overall Nepalese style with tiaras showing a central anchor-shaped element makes it possible to link them to the artists who worked on tier 6. (Figs. 22 & 23) One of them presents a slightly different style with goddesses wearing tiaras in the form of blooming flowers of the same type as what we have observed on the south side of tier 6 on the representation of Rahu. (Fig. 24) One might be tempted to see in this an indication that a particular workshop may have worked at least on the southern side of this tashi gomang. Another plaque closer to the first two (Figs. 22 and 23) seems to synthesize the two styles by also presenting the diadem with jewels in the form of blooming flowers. (Fig. 25)

Fig. 22

Fig. 23

Fig. 24

Fig. 25

Taking into consideration that on the majority of tashi gomang stupas the eastern face which is most exposed to public view seems to be executed with the greatest care, the treatment of the first of these plates (Fig.21) could lead us to think that it comes from this eastern side. It should be noted that as on the plate photographed at Tsetang in 1948 mentioned above, (Fig.9) they all hold a khatvanga in the hollow of their arm.

Fig. 26Only one sculpture that has come down to us could be considered to have been part of the twelve that were grouped three by three at this level on each side of the monument. Considering the very small number of one-faced and four-armed goddesses mentioned in the description of the tashi gomang for Dorje Gyalpo Phagmo Drupa, it is conceivable to consider as an aspect of Chunda this four-armed deity holding a vajra, a blossoming lotus, and a triple jewel. (Fig.26) Chunda was a very popular goddess in ancient India, which seems to have quickly become confused with aspects of Tara when it spread to Tibet. Although it is difficult to identify, probably because it is described according to ancient texts, its iconography is recurrent in most of the eight groups of sculptures, and thus constitutes a good chronological indicator of the evolutionary sequence of the tashi gomang stupas. It is also executed in a mixture of Nepalese and remains of Pala styles and rests on a lotus base supported by a stepped plinth. This characteristic type of base associating lotus and stepped plinth will be assigned without absolute logic to certain deities throughout the 168 years during which the tashi gomang stupas were built.

Tier 4
At this level are normally the representations of Buddhas. As the number of Buddhas that could come from Densatil's tashi gomang stupas is extremely small, perhaps because those who extracted them from the ruins have piously preserved them, it is delicate in the present state of the documentation to attribute precise images to them.

On the other hand, we can identify at this level some very interesting representations. The central Buddha figures were flanked by two of the eight Bodhisattvas of the bhadrakalpa and two wrathful deities on each side. Coming from the eastern side, we can mention Maitreya and Aparajita (Fig.27) which were cast on the left of the background plate in front of which an image of Shakyamuni was to be placed. From the south side, Vajrapani and Acala (Fig.28) had to be placed also on the right side of the central Buddha. These two groups are executed in a style that is certainly Nepalese, but also betrays a classical Pala influence.

Fig. 27

Fig. 28

Fig. 29

At each corner, to the left of the spectator, were four guardian deities, Vajrankusha, Vajrapasa, Vajragantha and Vajrasphota. A representation of the latter attributable to this first tashi gomang is known. (Fig.29) It was placed on the west face at the northwest corner of the tier. The most characteristic elements of this style appearing on this object are the treatment of the lotus pedestal, the scarf, the earrings and the tiara with circular elements that may also evoke blooming flowers.

This sculpture is the first one really attributable to a particular workshop which seems to have mainly worked on the upper half of the monument.

David Weldon recently published an article [30] isolating this style by analyzing 14 objects attributable to this workshop. He has clearly shown for all these objects that they must have been executed by a workshop of artists most probably of Nepalese origin but working in a style strongly influenced by the aesthetics of India Pala.

Proposing a dating of the 13th century for this set he refutes a possible provenance of Densatil by forgetting the existence of this first tashi gomang of 1267. To further isolate this group from those generally attributed to Densatil, he argues that in Densatil the tangs that fixed the sculptures to the monuments emerge from the backs of the deities whereas in this group they emerge from the backs of the lotus pedestals. However, the examination of the whole corpus of objects coming from Densatil reveals that the tangs can, depending on the case, be placed either on the backs of the deities or on the backs of the lotus pedestals. Some objects sometimes do not show tangs.
He suggests that many other shrines may have housed such monuments.

However, since all fourteen of the objects he studied, totally fit into the very specific iconographic scheme of the Dorje Gyalpo Phagmo Drupa memorial stupa as visualized by Jigten Gonpo in 1198 with 2,800 deities represented on its surface, it is unlikely that they could have adorned a monument anywhere else than in Drigung or Densatil. For the reasons already mentioned above, and in particular the destruction of Drigung in 1290, the most likely hypothesis seems to be that these sculptures may have come from the upper part of Dragpa Tsondru's Tashigomang in Densatil.

His remark on the strong Pala influence present on this group of objects is very interesting, as it could be an additional clue to add to other points to measure the contribution of the Khasa Malla rulers who made numerous and important offerings to Densatil and other Tibetan monasteries throughout the thirteenth century. This is probably related to the fact that many monasteries dependent on Drigung were established in western Tibet by Jigten Gonpo.

Tier 3
From this level, the majority of the sculptures seem to come from the workshop we have just mentioned above with the Vajrasphota. Careful examination of these works allows us to be certain that this workshop must have been composed of several artists who did not all showed the same talent.

Among these objects on this tier 3, we can mention an Amitabha from the east side (Fig.30), and a Vajrasattva from the south side (Fig.31). We will focus on a one-faced four arms female deity. His lotus pedestal lotus is spread out over two lions. (Fig.32) Two of his hands are in Dharmachakramudra and the other two are in meditative attitude. Two lotus flowers bloom above his shoulders, respectively supporting a conch and a book. Considering the very small number of one-faced and four-armed goddesses mentioned in the description of the tashi gomang for Dorje Gyalpo Phagmo Drupa, it is conceivable to consider this one as a rare aspect of Prajnaparamita. This could be confirmed by the presence of a book over his left shoulder.

Fig. 30

Fig. 31

Fig. 32

Since this deity is present with small variations on other monuments in Densatil, it may serve as a thread in the chronological stylistic evolution of the tashi gomang stupas.

These deities, like many others from the tashi gomang stupas, rest on lotus pedestals, themselves arranged either on stepped bases on which are represented the animals constituting their vehicles (vahana), or the stem and the lotus scrolls supporting the blooming flower, sheltering in their foliage the same vehicle animals. Once again, this particular form evokes certain aspects of Khasa Malla art.

A Hayagriva, still in the style of this workshop, was to be found on the north face (Fig.33). With the exception of the armbands which here resemble those of the peaceful deities mentioned above, he wears the same type of scarf, earrings and tiara with circular elements that may also evoke blooming flowers than those of the Vajrasphota in Tier 4. Still in the same style, a standing Acala now preserved at the Potala was originally intended to be positioned on the west side (Fig.34).

On each left corner (from the viewer's point of view) of each side of this floor were representations of Acala. One of them (Fig. 35) is currently preserved in the Jokhang of Lhasa.

Fig. 33

Fig. 34

Fig. 35

Fig. 36

In his article, David Weldon focuses on this Acala by comparing it with the one inscribed as coming from Sangphu (gsang-phu ne’u thog) monastery formerly in the Bertie Aschmann collection now held at the Rietberg Museum in Zurich. (Fig. 36) The two seem almost cloned one on the other.  The one in Zurich bears an inscription added long after its creation designating it as the personal deity of an abbot of Sangphu named Namkha Tsenpa from Tsetang (Rtses thang pa Nam mkha' brtan pa).

Founded in 1351 by Changchub Gyaltsen (Byang chub rgyal mtshan), close to the Neudongtse palace, Tsetang was in fact the second great spiritual place of the Lang/Phagmodrupa clan along with Densatil. More than a simple Phagmo Drupa monastery, it seems to have been a large open center for Buddhist studies where masters of all schools were widely welcomed.

If Namkha Tsenpa was originally from Tsetang and was able to follow studies that led him to an abbatial throne in Sangphu, it is likely that he came from a family of a certain rank, possibly associated with the Lang/Phagmo Drupa clan. The path of a sculpture possibly originating from Densatil to Sangphu could this way find a form of explanation.

In his brilliant contribution to David Weldon's article, Yannick Laurent explains that Namkha Tenpa occupied the abbatial seat of gLing stod, the superior college of Sangphu during the last quarter of the fifteenth century. [31] He says he hasn't been able to track down to which collegiate school Namkha Tenpa belonged. In his study on Sangphu [32] Karl-Heinz Everding refers to his various colleges mainly of the Kadamapa and Sakyapa traditions. He mentions two of these colleges, sGros rnying and sGros gsar, which are said to have been linked to the Kagyupa traditions. [33] Perhaps this is how Namkha Tsenpa from Tsetang was introduced to Sangphu which was predominantly linked to the Kadampa then Gelugpa school, with whom the Phagmo Drupa always maintained privileged relationship.

It is absolutely obvious that the Acala once kept in Sangphu was made in the same workshop as the one now preserved in the Jokhang. It is also clear that this workshop is the same one from which the majority of the objects we attribute here to the upper part of the tashi gomang for Dragpa Tsondru come from.

We will never know exactly where this sculpture was cast in the thirteenth century, nor by which meanders in history it may have found its way to Sangphu two centuries later, but the existence of two identical Acala on a tashi gomang is not technically impossible, since one was originally placed at each corner of the third tier.

Careful examination of P.F Mele's photographs clearly shows that in 1948 some of the tashi gomang stupas' sculptures were no longer exactly arranged according to the original plan of Jigten Gonpo's vision, that some were missing or replaced by deities unrelated to those that should or could have adorned a tashi gomang of the Dorje Gyalpo Phagmo Drupa commemorative type. This implies that many of these sculptures may on certain occasions, perhaps for the maintenance of the monuments, have been removed to be replaced on subsequent tashi gomang stupas or possibly offered as relics to certain abbots.

Tier 2
On this level, we can mention a Vajrayogini today kept at the Capital Museum in Beijing. It is characteristic of this style with the typical scarf, and must have been located on the west side of the tashi gomang. (Fig. 37) It is almost identical in all respects to those we will find on the upper level with the only difference that it is resting on a classical lotus pedestal.

Fig. 37

Fig. 38

Fig. 39

According to the descriptive texts of the tashi gomang for Dorje Gyalpo Phagmo Drupa, an aspect of Avalokiteshvara was positioned at each corner of this level.  Two representations attributable to this style which seems to characterize the upper part of the tashi gomang and at this level 2 have come down to us. (Fig.38 et Fig.39)

The two are iconographically almost identical and should therefore represent Sadakshari and Chintamani who respectively occupied the south-eastern and south-western corners of the monument.

Tier 1
On this last level, the objects are arranged around the large Kadampa-type stupa.
Most of the figures clearly attributable to this level are in the style of workshop we are speaking about from tier 4.
The most significant group consists of Vajravarahi figures (Fig.40) that were originally arranged around this last level, thus strongly marking the association of the entire monument with Chakrasamvara. The descriptive texts of the "Reliquary Stupa of Many Auspicious Doors for Phagmo Drupa" respectively translated by Christian Luczanits and Olaf Czaja mention the presence of three of them on each side.

Fig. 40

Fig. 41

As mentioned above, they are almost identical to the one on level 2, the major difference being the lotus pedestal, which is different here, with a double row of petals. This double row of lotus petals seems to have been in use only on this first tashi gomang and on the eighth, indicating once again their probable proximity. It is supported by a stem allowing them to be hung from the edge of the monument. With their specific ornaments and scarves ending like fish tails, they can in a way be considered as the stylistic archetype of this workshop. Careful examination of the details within this small corpus makes it possible to envisage that they could be the work of at least two distinct artists trained in the same tradition. The Capital Museum in Beijing preserves a Vajravarahi dancing on a blooming lotus with double row of petals, executed in a more classical Nepalese style, but being strongly influenced by the previous group with regard to the ornaments and the scarf with two ends in fish-tail style. (Fig.41) At the base of his bun, behind the skulls diadem, is visible, here encrusted with coloured beads, the circular element evoking a blooming flower that we have already encountered in other works. The existence of this sculpture is a new demonstration of the multiplicity of artists and workshops having worked more or less concomitantly on a same tashi gomang.

Still associated with this workshop, among the lineage of masters, one must mention a representation of Mahasiddha Kukkuripa and his dog from the Capital Museum Beijing, (Fig.42) as well as another figure kept at the Potala of Lhasa brandishing a kartrika above a kapala (Fig.43) This last figure was identified as Nairatmya by David Weldon,[34] but this seems impossible for many reasons. The first is that the character is resolutely masculine, which explains why Ulrich Von Schroeder identified it as an aspect of Padmasambhava. The second reason is that no representation of Nairatmya exists in the iconographic scheme of the Dorje Gyalpo Phagmo Drupa memorial tashi gomang stupa. It must therefore be a mahasiddha who remains to be identified more precisely, or as Ulrich Von Schroeder proposed, an aspect of Padmasambhava since a representation of this character is part of the list of lineage members to be present on this tier one.[35] However, Christian Luczanits considers difficult to imagine that a Padmasambhava figure could be encountered in 1267. If, however, Ulrich Von Schroeder's intuition was right, it could be one of the first representations of this master. [36] It should be noted that these representations are not resting on lotus pedestals.

Fig. 42

Fig. 43

Fig. 44

Fig. 45

From the same level must come a representation of Chakrasamvara (Fig.44) that is stylistically totally associated with the group of Vajravarahi figures, same jewels, same scarf. It should have been positioned on the west side.

The comparison of this work with a drawing now kept at the Rubin Museum in New York (Fig.45) is interesting at more than one point in reaching the top of this first tashi gomang. This textile drawing shows traces of regular folds, suggesting that it was originally part of the consecration charge of a sculpture or reliquary. He presents Chakrasamavara embracing Vajradakini between the footprints of Jigten Gonpo. As on the thangka depicting a lama mentioned above (Fig.14), the lower register shows us a row of lotus scrolls surrounding various protective deities as well as Nagarajas supporting the throne of the deity as on the tier 6 of a tashi gomang. These scrolls go up the sides of the composition to surround a line of masters and Mahasiddhas, again including Atisha, as on the first tier of a tashi gomang. Leaning against a background of flames, Chakrasamvara and Vajrayogini are executed in a graphic style totally echoing that of sculpture, as much in the ornaments as in the scarf ending in fish tails style, or in the circular floral element at the base of the buns behind their skulls crowns. This similarity must be seen as the expression of a strong artistic tradition, both pictorial and sculptural, attached to the monasteries of Drigung and Densatil during the thirteenth century.

Fig. 46 Pillars
On all the tashi gomang stupas, from levels two to five, the floors are connected by pillars, each decorated with two leaning four arms bodhisattvas forming caryatids resting on lotuses emerging from a kalasha (immortality liquor vase). The two leaning bodhisattvas are generally male and female.
Of the eight tashi gomang stupas, it should be noted that these pillars do not always appear to have been executed with great care. Examination of the sets clearly shows that some, probably the most visible ones, could have been executed with great care, while the majority, although each time following a pre-established prototype, gives the impression of having been entrusted to poorly gifted apprentices. The general treatment of those attributable to this first tashi gomang is rather sketchy and does not show great homogeneity from one to another. Their general common point seems to be the rather small number and shape of the petals of the lotus on which they rest, and the fact that the capitals only present in their center a simple horizontal line where on the following monuments a row of pearls will appear. (fig.46)

Background plates
On all the tashi gomang stupas, from floors 2 to 5, the main deities are leaning against plates decorated in relief with numerous smaller deities. It is these very many deities distributed over the eighty plates arranged on these four levels that make it possible to arrive at sets of more than 2,800 figures on each tashi gomang. It would seem that in the majority of cases, at least for the first seven tashi gomang stupas, these plates supporting the offering goddeses on the fifth tier and the Buddhas on the fourth one have been cast.  For the two upper tiers (2 and 3) on which the deities appear surrounded by lotus scrolls illustrating the mandalas of those who were positioned in front of them, it seems that they were treated using the repoussé technique. By stylistic association with the plate from tier 6 mentioned above, and those visible in the photograph by P.F Mele, we can propose the attribution of two plates to this first tashi gomang. Both are today at the Museo d’Arte Orientale of Torino (Italy). The first is a fragment of the Vajradhatu mandala occupying the east face of the third tier. (Fig.47) The second is associated with Avalokiteshvara and must have been on the west face of this same third level. (Fig.48)

Fig. 47

Fig. 48

At the end of the examination of this first tashi gomang, we can get an idea of the size of these monuments visible on P.F. Mele's photographs. If we retain the average heights of each of the floors and the various constituent elements, which seem to have varied little over time, we can imagine that Densatil's tashi gomang stupas could have been a little more than four meters wide at the level of the large lotus, about three meters high up to tier one and four and a half to five meters high, including the kadampa-type stupas at the top. They also appear not to have been placed directly on the ground, but slightly raised.

Abbot 5 / Stupa 5. Rinchen Dorje (Rin chen rd -rje) (1218-1280)

Rinchen Dorje acceded to the throne of Densatil in 1267 and is considered a reincarnation of Dorje Gyalpo Phagmo Drupa. This is undoubtedly related to the Lang/Phagmo Drupa clan's desire to increase its prestige by better linking Dorje Gyalpo Phagmo Drupa's direct spiritual lineage to Densatil and thus to their clan rather than to Drigung. At the same time, it should perhaps be seen as an attempt to fight against the ascendant prestige of the Karmapa who in 1216 established a new system of succession to the throne of their school by recognizing Karma Pakshi (Karma pak shi - 1204-1283) as the reincarnation of Dusum Khyenpa (Dus gsum mkhyen pa) who died in 1193.

Olaf Czaja believes that a tashi gomang of the Dorje Gyalpo Phagmo Drupa type in Drigung was erected for Rinchen Dorje in Densatil by one of his disciples named Spu-rtogs. He relies on two points for this. The first is a reading of the rgya bod yig tshang and the deb ther dmar po[37], but in both cases, and mainly in the deb ther dmar po, the ambiguous turnings may also suggest a misinterpretation evoking the construction by Rinchen Dorje of Densatil's first tashi gomang for Dragpa Tsondru, the famous Spu rtogs having then been able to participate in the work.

It should be noted that the Blue Annals, whose author was close to the court of Lang/Phagmo Drupa and therefore well informed of everything that may have affected Densatil, only evokes for him his status as the reincarnation of Dorje Gyalpo Phagmo Drupa and the fact that he was at the origin of the tashi gomang of 1267 [38]. Nowhere else in the same book does it mention that one of his successors built one for him.  The deb ther dmar po gsar ma mentions that Changchub Gyaltsen (1302-1364) is said to have had redevelopment work carried out in the Densatil mchod khang where were the tashi gomang of 1267 and the sku 'bum (sgo mang) built by him in 1360 [39]. This description seems to indicate that in the time of Changchub Gyaltsen, there were only two tashi gomang stupas of the Dorje Gyalpo Phagmo Drupa commemorative type in the main temple of Densatil.

Fig. 49 Olaf Czaja's approach also seems to rest on the fact that no stupa is attributed to Rinchen Dorje in the gnas yig of Chos kyi rgya mtsho next to those of Dorje Gyalpo Phagmo Drupa, Jigten Gompo and Dragpa Jungne in the center of the east wall of the temple. [40].

Another obstacle to accepting that a Dorje Gyalpo Phagmo Drupa commemorative type tashi gomang was erected in Densatil for Rinchen Dorje is that it is possible to establish eight groups of works that may have belonged to the eight historical tashi gomang stupas, but their stylistic chronology does not allow us to envisage that one of them could have been made in 1280.

If no commemorative stupa for an abbot considered as the reincarnation of the founder of the lineage is mentioned among the five arranged in the middle of the east wall, and no tashi gomang of the Phagmo Drupa type could be erected for him, it is because his commemorative stupa must have had another form.

Fig. 50 In the northeast corner, not far from the five Kadampa stupas in the center of the east wall and probably next to the site that must have been occupied by the tashi gomang of 1267 for Dragpa Tsondru, we can still see the remains of a huge stone-structured stupa that could be attributed to Rinchen Dorje, (Fig. 49) first of all because it is located in one of the oldest parts of the temple, and is in a way a counterpart to Dorje Gyalpo Phagmo Drupa's hut, of which he was considered the reincarnation and which was located in the southeast corner. Assuming that this classical stupa structure could have been covered with a rich decoration of repoussé copper sheets in accordance with common usage, it is not impossible to think that the iconography may have allowed it to be included, in a more modest way than the 1267 stupa, in the tashi gomang stupas category.

It is hardly conceivable that a great tashi gomang like the one in 1267 could have been built for Rinchen Dorje without having been commissioned by Dragpa Yeshe (Grags pa ye shes) who acceded after him to the throne of the monastery or of Mkhanpo rin rgyal or Byang gzhon who succeeded each other in Neudongtse in 1281/1282. Indeed, only the latter could logically have decided to erect such an important and costly monument in the monastery. This stupa, although restored, is the only original element still in place today in the reconstructed temple (Fig. 50).

Fig. 51 Abbot 6 / Stupa 6. Dragpa Yeshe (Grags pa ye shes) (1240-1288)

Having studied in Sakya with Phagpa (‘Phags pa), Dragpa Yeshe ascended the throne of Densatil in 1280. At his death in 1288, a large commemorative stupa of the same shape and size as Dorje Gyalpo Phagmo Drupa's tashi wobar was placed next to the three previous ones. In 1948, it stood with four other stupas of the same type and size, textile wrapped, on a tiered altar in the middle of the east wall of the main hall. It can be partially seen on the famous photographs by P.F Mele. (Fig.51)

The fourteenth-century stupas

Abbot 7 / Stupa 7. Dragpa Rinchen (Grags pa rin chen) (1250-1310)

Dragpa Rinchen ascended the throne of Densatil in 1288. During his twenty-two years at the monastery, he received a title from the Mongols giving him temporal power over the region. It is also during this period that the Sakyapa and the Mongols razed the monastery of Drigung and the tashi gomang stupas that it was supposed to shelter. From then on, the tashi gomang erected in 1267 at Densatil became the only surviving testimony of this type of achievement. 

It is also known that under his ministry a king of Yatse (most probably a Khasa Malla sovereign) covered with gold the two stupas East and West of Densatil [41]. This description should most probably be interpreted as the making of two golden copper ceiling parasols or mandalas, placed above the tashi gomang of 1267 and the stupa of 1288 for Rinchen Dorje, which were to be placed side by side on the north wall, thus necessarily one to the east and the other to the west.

This event marks once again the strong links that seem to have been woven between the Khasa Malla kingdom and the monastery of Densatil, which are probably not alien to certain artistic developments in this monastery and would need to be the subject of a specific study.

This tradition of placing umbrellas including gilt copper mandalas above stupas intended to receive the relics of great figures seems to have been relatively common in Tibet, and in Densatil, since P.F. Mele's photographs show at least one above the tashi gomang commissioned by Changchub Gyaltsen in 1360. (Fig. 52)

Fig. 52

Fig. 53

Fig. 54

Numerous fragments of such mandalas originally plated on the ceilings are preserved in the collections, and many of them probably come from Densatil given their high quality of craftmanship. (Fig. 53)

Upon his death in 1310, a large commemorative stupa of the same shape and size as Dorje Gyalpo Phagmo Drupa's tashi wobar was created. In 1948, it stood with four other stupas of the same type and size on a stepped altar in the center of the east wall of the main hall. It can be seen partially wrapped in cloth in the famous photograph by P.F. Mele. (Fig. 54)

Fig. 55 Abbot 8 / Stupa 8 / Tashi Gomang 2 - Dragpa Gyaltsen (Grags pa rgyal mtshan) (1293-1360)

Dragpa Gyaltsen occupied the throne of Densatil from 1310 to 1360. All sources unanimously confirm that his brother Changchub Gyaltsen (Byang-chub rgyal-mtshan) (1302-1364), who succeeded in unifying Tibet and regaining temporal power over the country from the Sakyapas, had a tashi gomang of the Dorje Gyalpo Phagmo Drupa commemorative type erected for him. The Lang / Phagmo Drupa clan now reigning over Tibet and being at the head of one of the most prestigious and richest monasteries, Changchub Gyaltsen had no choice but to revive this prestigious tradition to affirm the temporal and spiritual power of his family.

The stylistic analysis of the tashi gomang stupas visible in P.F. Mele's photographs and of the stylistic chronology of the groups of objects attributable to all of them makes it possible to propose that this second tashi gomang may be the one for which we have several photographs and is therefore the best documented. (Fig. 55)

We therefore have here more elements and therefore almost certainties about the objects that must have adorned it in the past.

To try to illustrate the particularities of this second tashi gomang, we will retain the following few works.

The Four Guardians Kings
The overall photograph gives a good view of the structure of the monument with the four lokapalas arranged in front. They were not seen at this location in the photograph of the 1267 tashi gomang, which could indicate that in this first example these Guardian Kings must have been resting above the lotus circle. Although they are still depicted dressed in some sort of armour, the whole is executed in a more Nepalese style than a century earlier with rich inlays of semi precious colored stones. Note the lotus pedestals whose lower petals are smaller than the upright ones.

Of the four kings in the photograph, a similar burnt bust of Virupaksha remains today in private collection. (Fig. 56 & 57)

Fig. 56

Fig. 57

Fig. 58

Tier 6
This tier of protector deities encircled by lotus scrolls is particularly visible in the photographs of P.F. Mele. (Fig. 58) The deities of the eastern and south faces appear on a background of lotus scrolls encircling secondary deities in relief.

It would seem that here as on the tashi gomang of 1267 for Dragpa Tsondru the deities were cast, and then fixed on the bakground plates. Even if none of these plates have come down to us, careful examination of P.F. Mele's photographs allows us to be almost certain that they are no longer as we saw on the 1267 tashi gomang attribution proposal, executed in repoussé copper sheets joined by their centre.

 The fractures visible on all the edges of the cast plates that we will encounter on the following tashi gomang stupas might make us wonder if the complete circle on which the twenty-four deities were leaning could not have been cast in one time. Beyond the unlikely technical tour de force that this would imply, even if we can't doubt the talent of the Newar masters, it hardly seems conceivable. Chinese scholars have brought to light that the workshops for making cast objects intended for the Densatil monastery would have been located in the plain, on the opposite bank of the Yarlung Tsangpo. [42] It is difficult to imagine that such a heavy metal circle, three and a half to four metres in diameter, could have been loaded onto a ferry to cross the Yarlung Tsangpo and then transported up the mountain in one piece.

They seem in fact here to have been made by grouping two or three together in a single cast.  A detailed photograph of the Vaishravana of this monument reveals a lotus stem, partially masked by the costume of a lokapala, attached to the plate by a tenon that seems to act as a cover for a joint.
According to Chunhue Huang, the rich local copper mines known in the 13th century are still in operation today.

The analysis of P.F. Mele's photographs provides us with a precise vision of at least five sculptures of this level 6 and allows us once again to see that beyond a globally common aesthetic, it is obvious that various artists worked on them at the same time. The Nagarajas and Vaishravana, like the lokapalas, wear tiaras with a large flower-shaped element in the middle and large circular earrings.

The Lhamo on the east face and Rahu on the south face wear tiaras without the central floral element, and simpler armbands. Rahu's earring is not a large disc either but is in the shape of a blooming flower with inlaid stones organized around another more in relief. This type of earring is found on a Druma (Fig.59) and an aspect of Mahakala (Fig. 60) buried in the ruins of Densatil, kept at the Capital Museum in Beijing, and logically coming from the north face of this tashi gomang. An aspect of Aparajita holding his vase surmounted by a triple jewel and wearing large circular earrings and a tiara with a circular flower in the centre could potentially come from the north face of this monument. (Fig. 61) A probable representation of the wealthy god Atavaka with a richly adorned body and probably holding two jewels whose inclusions have now disappeared should come from the southern side. (Fig. 62)

Fig. 59

Fig. 60

Fig. 61

Fig. 62

A nagaraja from the Kronos collection (Fig. 63) and another formerly published by Marcel Nies (Fig.64) show strong similarities with those visible in the photograph by P.F. Mele, essentially diadems, large circular earrings and a five snake heads cap rising above their buns.

Fig. 63

Fig. 64

The clothing of the Nagarajas attributed to the tashi gomang of 1267 was already enriched with inlays of stones of various colours. With this second monument, this propensity to overload sculptures with stone inlays will experience a strong development that will continue on the following tashi gomang stupas and the entire artistic production of the monastery. The tashi gomang that Jigten Gonpo wanted in Drigung to glorify his master Dorje Gyalpo Phagmo Drupa had a religious basis, but it soon became clear that the erection of this monument was also politically motivated. The first tashi gomang erected in Densatil for Dragpa Tsondru in 1267 coincided with the accession of the Lang/Phagmodrupa Clan to the head of one of the thirteen myriarchies imposed by the Sakyapas and the Mongols. After long nationalist efforts and struggles to try to return Tibet to the greatness it had in the imperial period (7th-9th centuries), Changchub Gyaltsen succeeded in unifying the country around his person in 1354. Governing from the family castle of Neudongtse, he occupies the de facto position of "King of Tibet" even if, in an outburst of false modesty, he only claims the title of Regent (sde srid). In 1360, Changchub Gyaltsen had this second tashi gomang erected in Densatil for his brother Dragpa Gyaltsen who had been its abbot since 1311.

There is no doubt that this first opportunity for the new monarchy to make its mark and to strengthen its sacred image has favoured the blossoming of this ornamental wealth which for three quarters of a century followed the rise and fall of the Lang/Phagmo Drupa dynasty. From the erection of this monument in 1360, many deities saw the flowers usually engraved on their dhotis replaced by floral representations in relief inlaid with coloured stones. This unique treatment of dhotis is undoubtedly one of the few main markers that will help define the Densatil style.

Tier 5
We can attribute to this rank a series of offering goddesses detached from their background plates. Careful examination of P.F. Mele's photographs seems to reveal a gap between the back and the deities that must have been attached to it. This would seem to be the only tashi gomang to present this technical particularity. Mention of the use of silver on Densatil's tashi gomang stupas could perhaps shed some light here if we consider that the plates to which these goddesses were attached may have been made of silver. The variations in the treatment of these goddesses once again reveal the variety of artists who worked on the different sides of the monuments. (Fig. 65) Also to be noted here is a change in the iconography of these offering goddesses whose arms are more developed in movement to better expose what they present. Therefore, unlike those attributed to the tashi gomang of 1267, they no longer support any Khatvanga in the hollow of their arm.

Fig. 65

Among the goddesses attributable to this level, a Parnashavari with a richly inlaid garment to represent foliage now kept in the Capital Museum in Beijing is undoubtedly the one photographed by P.F. Mele in 1948. (Fig. 66) It is identifiable with near certainty by the inverted armband on its upper right arm. (Fig.67) As several representations of Parnashavari have come to light, they can also serve as a complementary thread to understand the chronological and stylistic evolution of Densatil's tashi gomang stupas.

Fig. 66

Fig. 67

Still in the collections of the Capital Museum in Beijing, we have a deity catalogued as Marichi (Fig. 68) who must have been next to Parnashavari in the middle of the east side. A four-armed goddess holding a vajra, a blooming lotus and a triple jewel takes up the exact iconography of the one we had considered as a probable aspect of Chunda on the tashi gomang of 1267. (Fig. 69) It is treated here in a more rounded style and does not have a lotus pedestal supported by a stepped plinth. It should logically have been located on the south side of the monument and is now kept at the Capital Museum in Beijing. A fierce-looking deity with eight arms whose identifications have been extremely variable could in fact be an aspect of Ekajati that should have been on the west side of this fifth tier. (Fig. 70) One will note on these two last figures the shape of the armbands which will mark many sculptures of this second tashi gomang.

Fig. 68

Fig. 69

Fig. 70

Fig. 71

Still attributable to this tier 5, it is worth mentioning a rare small representation of Agastya that must have been in the northeast corner. (Fig. 71)

Tier 4

No background plaques with rows of buddhas attributable to level 4 of this second tashi gomang have come down to us. As for the Buddha statues that were to be placed in front, we stated earlier that they are paradoxically rare, whereas about twenty of them were to be placed on each of the monuments. It must be said that in so far as the Buddhas are poorly adorned, there are few stylistic elements apart perhaps from the lotus pedestals to link them to the production of the workshops that worked at Densatil and even less to a tashi gomang in particular.

A Buddha resting on an open lotus raised by a stepped plinth (Fig. 72) could be attributed to this second tashi gomang. A group representing Maitreya and Aparajita (Fig. 73) is most likely the one that would have been to the right of the central Buddha on the east side, but P.F. Mele's photograph is unfortunately not sharp enough to clearly identify it.

Fig. 72

Fig. 73

Fig. 74

A guardian deity whose lost attributes do not allow precise identification, but who must have been at one of the corners of this tier 4 is most likely the work of one of the artists who worked on tier 6. (Fig. 74) The pendants of his belt and jewels are indeed to be compared with those visible on the Nagarajas, (Fig. 64) and the jewels are to be compared with those of the Druma and Mahakala from the north side. (Fig. 59, 60)

Tier 3
The Amoghasiddhi formerly in the B. Aschmann collection now kept at the Rietberg Museum in Zurich should find its place on the east side of this tier (Fig. 75). On the south side, there must have been a Vajrasattva now preserved in Drigung (Fig. 76), and the Manjushri from the former Zimmermann collection (Fig. 77). According to the texts it should be the so-called “Lion of Debaters” aspect which usually presents him sitting on a lion. Lions are, however, represented in the scrolls under his lotus pedestal.  We meet for the first time on the dhoti of this sculpture the realization of the typical floral decoration in inlays of coloured gems which will develop even more in the two following tashi gomang stupas.

Fig. 75

Fig. 76

Fig. 77

Fig. 78

Fig. 79

We had on the stupa of 1267 placed in the middle of the west face a female deity with four arms considering that it could only be an aspect of the goddess Prajnaparamita.  Although the book has been broken here, we find exactly the same deity with two lions on the lower part of his throne. The great care taken in making it with a great wealth of stone inlays in its ornaments and its dhoti pleads in favor of an important deity of the pantheon, and thus of Prajnaparamita (Fig.78). Also on this west side, we can place an aspect of Avalokiteshvara with armbands having the same characteristics as many other objects attributable to this monument, and the dhoti with richly inlaid stones floral decoration. (fig.79).

The north side was to be occupied in its center by a mandala of Chakrasamvara in 14 deities. A sculpture of this deity in union with Vajrayogini can be associated with this location. The god is dancing in Ardhaparyaṅkāsana (Fig.80). This sculpture is interesting because it seems to be a transition between the first and third tashi gomang stupas. The armbands recall those of the Nagarajas attributed to the stupa of 1267 and the treatment of the bone apron of Vajrayogini already announces what will be generalized in 1370. Still from this north face at this third level, must come a Vajrapani which unfortunately lost its gilding (Fig.81). At each corner of this tier 3 was a representation of Acala. At least one seems to be attributable to this second tashi gomang (Fig.82).

Fig. 80

Fig. 81

Fig. 82

Tier 2

Fig. 83 The eastern side of this second level was logically intended to illustrate various cycles of the Anuttarayoga Tantra including the 72 deities mandala of Guhyasamaja. P.F. Mele's photographs allow us to see that the iconographic scheme of this tier has lost all coherence and the statues in it don't seem to be in their logical place. In its middle, in place of Guhyasamaja, is a representation of Vajrabhairava, that should have been placed on the right side of the South side (Fig. 83). Examination of the background plates at this level, however, shows that they were originally associated with the Guhyasamaja mandala. Once again, this detail allows us to raise the question of a possible displacement of certain sculptures to later tashi gomang stupas, or to be offered as relics to certain dignitaries.

Although we cannot see them in the partial photograph on this side two sculptures must have originally been on this side. It is a probable aspect of Ratnasambhava (Fig. 84) and a probable aspect of Amitabha (Fig. 85) in their characteristic form of the Guhyasamaja retinue. Although executed in slightly different styles, like all the sculptures on the lower floors, the unit is maintained by the treatment of the armbands, which will be radically different thereafter.

The Mahacakra Vajrapani from the Ford Collection (Fig. 86) was to be positioned to the south, and the 11-headed Lokeshvara (Fig. 87) was to be located at the northeast corner of the tier. Once again, the characteristic shape of the armbands already observed on many of the sculptures in the lower rows confirms the attribution to this tashi gomang.

Fig. 84

Fig. 85

Fig. 86

Fig. 87

Tier 1
This first tier was partially photographed by P.F. Mele. We can see the east side with seven representations of Vajravarahi fixed on the rim (Fig. 88). The descriptive texts of the tashi gomang visualized by Jigten Gonpo mention only three of these goddesses on each side of the monument. This therefore implies that a variant by positioning twenty-four dakinis on the periphery has been introduced for this stupa or that the decision has been taken to group the twelve of the initial program towards the front of the tashi gomang to give it an even more sumptuous appearance.

Fig. 88

Fig. 89

Fig. 90

Fig. 91

The style of these figures has slightly evolved in a century and the main character of this series seems to be a circular winding of scarves arranged on one or both sides, as can be seen on the Vajravarahi of the Seattle Museum (Fig. 89). It should also be noted that the bone apron of the goddess has a more structured shape and is like the ends of the scarf enriched with inclusions of semi-precious stones creating a form of harmony with the rich decoration placed on many of the sculptures on the lower tiers. The lotus pedestals have only one row of petals. The comparison of this Seattle Museum Dakini with another very similar one (Fig. 90) confirms that even if they come from the same workshop, they are obviously the work of different artists.

We have already mentioned on this tier 1 the presence of a Chakrasamvara figure occupying the center of the west side. A superb representation of this deity whose scarf has the same windings as on the Vajravarahi figures should find its place on this second tashi gomang. (Fig. 91). Each of the five skulls on her tiara is surmounted by a floral disc. Note its lotus pedestal whose lower petals are smaller than the upright ones, as on the Lokapalas.

P.F. Mele's photograph partially shows the figure of Vajradhara occupying the middle of this east side, to the right of the Dakini supporting a long white scarf. (Fig. 88)

Pillars

Fig. 92_1 As seen on the first tashi gomang, the caryatid pillars decorated with two leaning bodhisattvas are not treated with great attention, although we can note a greater stylistic homogeneity. The lotus emerging from the vase always has erect petals that seem more or less numerous depending on the floors and sides of the monument. The essential common point is the appearance of a horizontal row of small pearls in the centre of the capitals. (Fig. 92_1)

On the photographs of this tashi gomang, we can observe at the corners of each tier, pillars of another type. The base is always composed of a vase from which emerges loops of lotus scrolls encircling deities or auspicious symbols (Fig. 92_2). The capital is structured in the shape of a lotus with erect petals. Looking at the stylistic details, none of these corner pillars known to date can be attributed with certainty to this second tashi gomang.

Fig. 92_2 The photograph also allows us to observe above the corner pillar a tri-lobed lotus petal adorned with deities. This type of ornament that we will find omnipresent at all the corners of tiers 2, 3, and 4 of the east side of the third tashi gomang only existed in two examples on this second monument in 1948.

In his monumental historical study of the Lang/Phagmodrupa clan [43] Olaf Czaja devotes a small chapter to an attempt at a typological classification of tashi gomang stupas. In this attempt to form four groups based more on the examination of ornamental details than on real stylistic considerations, he dissociates the photograph of the upper part of this tashi gomang from those of the lower part by classifying them into two different categories. His argument based on the counting of vajras, antelopes and wheels of the law on horizontal bands separating floors does not take into account the narrowing of the width of these floors, and cannot be considered conclusive. On the other hand, it is obvious that among all the photographs of P.F. Mele, only one tashi gomang presents two scarves hanging from the dakinis of row 1, one long in the center and a smaller one on the northeast corner of the monument. Beyond the perfect stylistic coherence between the objects appearing in the two photographs (Fig. 55 and Fig. 88), the presence of these two scarves confirms that the two pictures represent the same monument.

Abbot 9 / Stupa 9 / Tashi gomang 3. Dragpa Sherab (Grags pa shes rab) (1310-1370)

Second brother of Changchub Gyaltsen, Dragpa Sherab occupied the throne of Densatil from 1360 to 1370. Densatil's third tashi gomang had been built for him by his successor.
The stylistic analysis of the stupas photographed by P.F. Mele in 1948 and the groups of objects that can be attributed to them makes it possible to propose that this third tashi gomang may be the one that appears on our left when we look at the overall photograph of the east wall of the main hall (Fig. 1). It is the only one of these monuments whose photographs allow us to apprehend all the tiers of the east side. (Fig. 93 & Fig. 94)

According to Chos kyi rgya mtsho, it would have housed 3,900 deities, indicating an iconographic evolution compared with the archetype created by Jigten Gompo [44]

Fig. 93

Fig. 94

To try to illustrate the particularities of this third tashi gomang, we will retain the following few works.

The Four Guardian Kings
The four Lokapalas are clearly identifiable in P.F. Mele's photographs, arranged in front of tier 6 and the large lotus. All four escaped destruction and are now shared between the Jokhang in Lhasa and the Capital Museum in Beijing. They present a great stylistic unity and are most probably the work of great Newar masters. Their general shape continues the aesthetics seen on the two previous tashi gomang stupas but with even more inlays of semi-precious stones on their headdresses and armors.

Fig. 95

Fig. 96

Fig. 97

Fig. 98

Note that the Viruḍhaka had already lost the blade of his sword in the 1948 photograph (Fig.95). Like many of Densatil's objects exhumed from the ruins of the monastery by Chinese scholars, he was photographed without having been cleaned of all the soil and dirt that covered him. The Vaishravana's face has been repainted in Jokhang and has gained a pair of moustaches that it did not have before. (Fig.96). Virupaksha holds his stupa with his left hand (Fig.97). On the posterior specimens, he will wield it with his right one. The Dhritarashtra has lost its lute (Fig.98).

Tier 6
The tier 6 is partially hidden behind the five large jinas in repoussé which we will discuss later. This clutter shows only three lotus scrolls surrounding Vaishravana, Mahakala and Lhamo in its Prānasādhanā Śrīdevī aspect of which we can see the head, two arms and the head of the mule (Fig.99). The three lotus scrolls all have a particular ornamental pattern that will be characteristic of the works from row 6 of this tashi gomang. This pattern is found on the Rahula from the Los Angeles County Museum which seems to be the only fragment of this Row 6 to have been published to date (Fig. 100).

Fig. 99

Fig. 100

With rank 6 of this third tashi gomang, we enter a new artistic era which with the next tashi gomang to be built sixteen years later in 1386 will mark the peak of Densatil's art. The deities here are no longer cast apart and fixed on background plates. They are now solidary cast with this background. As evoked for the background plates of the tashi gomang of 1360, it is highly probable that several of these deities were cast at once, further accentuating the artists' technical feat. As on the two previous monuments, the secondary deities adorning the background of the main deities are surrounded by lotus scrolls.  One will note the richness of the inlays of semi-precious stones on the ornaments of Rahu. Once again this come to accentuate the sumptuary character wished by the "royal" sponsor of this monument since Dragpa Changchub who succeeded Dragpa Sherab on the Densatil’s throne in 1370 occupied too the Neudongtse’s one from 1374 to 1381.

Tier 5

Fig. 101 At this level five, as on all the tashi gomang of the Phagmo Drupa type were the sixteen goddesses of sensual enjoyment bearing offerings. We can attribute to this tier a plate kept at the Dallas Museum of Art with such goddesses attached to their background, whose typology is recognizable in the photographs by P.F. Mele (Fig.101). As with the deities of rank six, the majority of these offering goddesses seem to be lost. The lotus pedestal is absent, but a triple lotus scroll is present on the side of the plate. 

The Parnashavari which was in the middle of the east side is partially hidden behind a tri-lobed lotus petal ornament. Careful examination of P.F. Mele's photographs makes it possible to be certain that this is a known object today preserved in a private collection. (Fig.102) In spite of the disappearance of the upper part of her scarf visible in the blurred photograph, with her garment richly inlaid with stones to represent foliage, this goddess is one of the masterpieces of Densatil's art. A Tara with four faces and eight arms was to occupy the center of the south side (Fig.103). Her dhoti is decorated with flowers inlaid with coloured stones. The treatment of the armbands shows an evolution compared to the objects of the previous tashi gomang. A four-armed goddess holding a vajra, the probable stem of a broken blooming lotus and a triple jewel takes up the exact iconography of the one we had considered as a probable aspect of Chunda on the tashi gomang stupas of 1267 and 1360. (Fig.104) It should have been located at the center of the south side of the monument. Her dhoti is decorated with flowers inlaid with semi-precious stones and her armbands are also of a new shape. An eight-armed Mahapratisara now preserved in the Jokhang of Lhasa must have been positioned to the right of a Vasudhara on the north face (Fig.105). Her armbands are also characteristic of the new style appearing on this tashi gomang.

Fig. 102

Fig. 103

Fig. 104

Fig. 105

Tier 4
The iconography of row 4 consists mainly of representations of Buddhas. In the middle of each side an image of Shakyamuni is surrounded by two bodhisattvas of the Bhadrakalpa accompanied by a wrathful deity. They are backed by 250 miniatures representations of the Tathagata of their respective direction spread over five plates.
A Buddha figure should come from one side of this level (Fig.106). The background of the east row was to be decorated with plates of a particular typology identifiable in P.F. Mele's photographs (Fig.107). It is structured by 44 miniature representations of Aksobhya arranged around an arch bordered by motifs that may be flames or foliage. In the center of the lower part, there is a hole to pass the tang which is usually placed at the back of the lotus pedestal of the statues. Once blocked from behind, this system made it possible to fix the Buddha on the monument. In comparison with the photographs by P.F. Mele, it should be possible to consider that the presence of only ten Aksobhya divided into two groups in the upper row could be an indicator that this plate was on one of the two outer registers on this eastern side.

Fig. 106

Fig. 107

Fig. 108

Fig. 109

Fig. 110

A group representing Manjushri and Yamantaka (Fig.108) was to be positioned next to the Shakyamuni occupying the center of the south side. A background plate structured by 48 miniature representations of Ratnasambhava arranged around an arch bordered by motifs that may be flames or foliage should come from this level (Fig.109). Two offering goddesses are placed one above the other on the side (Fig.110). The presence of two groups of seven Ratnasabhava in the upper row and the arrangement of the offering goddesses on the side make it possible to situate this plate to the left of the one on which the central figures of this row are leaning, from the point of view of the observer.

Tier 3
The east side of this third level is clearly identifiable in P.F. Mele's photographs, which allow us to identify three sculptures of the Vajradhatumandala in 47 deities logically represented there. The lotus pedestals of the three deities rest on stepped plinth. Although partially hidden behind a white scarf, the Vairochana is recognizable by its lotus on the a stepped plinth, the stone inlays of the dhoti and the two wheels supported by lotus above its shoulders (Fig.111). The Aksobhya that stood next to it is also identifiable at least by the elephant on its base, since it is hidden behind a white scarf (Fig.112). The Amoghasiddi is much more easily recognizable (Fig.113). The three deities wear dhotis with floral motifs inlaid with stones, and armbands of the new type that we have seen on the lower floors.

Fig. 111

Fig. 112

Fig. 113

At each corner of this floor was to be a representation of Acala. One of them found in the ruins of Densatil by the chinese scholars and today in the Capital Museum of Beijing presents a treatment of its almost unique lasso that allows us to consider that it is the one visible at the south-east angle in the photograph by P.F. Mele (Fig.114). Like many of Densatil's objects exhumed from the ruins of the monastery by chinese scholars, he was photographed without having been cleaned of all the soil and dirt that covered him. Another sculptures representing Acala must also have come from another corner on this level and show the multiplicity of artists who worked there (Fig.115). Note that they both present scarves with a buckle, like those that adorned the Vajravarahi figures and some other deities on the previous tashi gomang.

Fig. 114

Fig. 115

Fig. 116

Fig. 117

In the midle of the south face, a representation of Buddhakapala was to be placed, which should be the one formerly in the Josette Schulmann collection (Fig.116). Note that the diadem of this deity has retained some of its strings of natural pearls ornamentation that most often have been lost but seem to have been more widespread on the next tashi gomang.  On the same level, a sculpture of Vajrasattva was to be placed (Fig.117).

In the middle of the west side, should be placed an aspect of the four-armed goddess that we have already encountered on the two previous tashi gomang stupas and that we had considered to be an aspect of Prajnaparamita (Fig.118). Her attributes that were supposed to rest on the lotuses that flourish above his shoulders have been lost. His tiara, jewellery and dhoti are richly inlaid with coloured stones. The innovative element in this work is the treatment of the scarf that surrounds the top of the head before forming an arc following the curve of the shoulders. This particular form will be widely used on the sculptures of the next tashi gomang. It is also possible to attribute to this western side an aspect of Avalokiteshvara (Fig.119).

Fig. 118

Fig. 119

Fig. 120

As on the previous tashi gomang, the north side was to be occupied in its center by a Chakrasamvara mandala in 14 deities. A sculpture from the Capital Museum in Beijing showing the dancing god in ardhaparyaṅkāsana is most likely the one that was to occupy this place (Fig.120). By certain stylistic and technical aspects, notably the three struts that reinforce the support of the sculpture resting on a single foot at its base, this sculpture can be compared to the Buddhakapala on the west side (Fig.116). It is interesting to note here the treatment of the Vajrayogini’s bone apron with inclusions of floral motifs inlaid with semi-precious stones between the classical swags of bone beads. This particular treatment of the apron will be found on the other two representations of Chakrasamvara attributable to the second and first tiers of this third tashi gomang.

TIER 2

Fig. 121 As on the previous two tashi gomang, the eastern side of this second level was logically intended to illustrate the 72 deities mandala of Guhyasamaja. The only sculpture recorded to date that may have occupied this site is probably the representation of Guhyasamaja photographed by Ulrich Von Schroeder at the Reting Monastery (rwa sgreng).[45] The fact that the photograph shows the deity embracing Sparsavajra holds the vajra in his left hand and the bell in the right one, implies that the direction of the photograph may have been reversed in printing. Once put back in its logical direction (Fig.121), the arms, and knees covered with a dhoti inlaid with coloured stones visible on either side of the white scarf in P.F. Mele's photograph suggest that it is indeed the same sculpture. 

In the middle of the south side we can propose to place a representation of Vajrakilaya (Fig.122), surrounded by an aspect of Mahacakra Vajrapani (Fig.123) and a Vajrabhairava (Fig.124), the latter two being nowadays kept in the Capital Museum of Beijing.

Fig. 122

Fig. 123

Fig. 124

On the southern side, the center was to be occupied by a representation of Hevajra, which should be the one kept at the Capital Museum in Beijing (Fig.125). Next to him must have been Mahamaya. It is probably the one today kept at the Dayton Art Institute (Fig.126). The particular treatment of the faces of these two deities makes it possible to envisage that they may have come from the same workshop, which could confirm their proximity on this tashi gomang.

At each end of this western side were to be representations of Vajrayogini. The one that we propose to attribute to this location (Fig.127) presents very strong stylistic similarities with the Vajravarahi figures that we will find on tier 1. It therefore seems logical to attribute it to this monument.

Fig. 125

Fig. 126

Fig. 127

Fig. 128

The northern side of this second tier is normally decorated with a large Chakrasamvara mandala of 62 deities. This seems to imply that a representation of the god embracing Vajrayogini occupied the middle, leaning against a vast repoussé decoration on which were represented the various deities of the mandala as well as the mahasiddhas and charnel grounds that traditionally surround the mandala. We propose to place here a representation of a Chakrasamvara whose Vajrayogini is dressed in a bone apron with the same floral decoration inlaid with stones as the one seen on the Chakrasamvara of the lower level, and which we will find on the Chakrasamvara of tier 1 (Fig.128).

Tier 1
The eastern side of this upper row organized around the large kadampa-type stupa is largely visible in P.F. Mele's photographs. On the border of this eastern side, we find an alignment of seven representations of Vajravarahi marking once again the distance from reality with the descriptive text of the tashi gomang for Dorje Gyalpo Phagmo Drupa which indicates only three. Among this alignment of dakinis, some still seem to follow the iconography observed on the previous tashi gomang with the characteristic loop of the scarf at the elbow (Fig.129). However, a new model with richer ornamentation is emerging which seems to better match with the general aesthetics of the monument and should be considered as the dominant style (Fig.130). The most famous recorded Vajravarahi of this style is unequivocally the one formerly in the B. Aschmann's collection, now in the Rietberg Museum in Zurich (Fig.131). Note that the floral decoration of his bone apron is identical to that of Vajrayogini on the three Chakrasamvara representations of this tashi gomang.

Fig. 129

Fig. 130

Fig. 131

All the texts describing the tashi gomang visualized by Jigten Gonpo for Dorje Gyalpo Phagmo Drupa are unanimous on the presence of a representation of Vajradhara in the center of this east side. One was partially visible in the photographs of the second tashi gomang. Here, the photograph being very sharp, it is obvious that if there ever was one, no statue representing Vajradhara embracing Bhagavani was there in 1948. Insofar as another statue of Vajradhara should have been in the middle of the southern side, it could be the one photographed by Ulrich Von Schroeder at the Jokhang of Lhasa (Fig.132).

Fig. 132

Fig. 133

Fig. 134

On the left of the photograph, one can see in the southeast corner a deity in yab-yum (Fig. 133) that Olaf Czaja identifies as Vajradhara, probably on the assumption that his reference text mentions one on each side of this first tier. [46] It is however enough to simply observe this photogaphy to note that this deity standing in alidhasana can in no way be identified as Vajradhara since the latter is always seated. The base without lotus pedestal, the double necklace of shrunken heads, the positioning of the dakini and the movement of the scarf make it possible to recognize in this object the Chakrasamvara now kept in the Capital Museum in Beijing (Fig.134). Note again the floral decoration inlaid in the Vajradakini bone apron which is of the same type as what we have already observed on the two other Chakrasamvara and some Vajravarahi figures attributable to this tashi gomang. The existence of this Chakrasamvara sculpture on top of a tashi gomang demonstrates the greater reliability of the text selected by Christian Luczanits than the one chosen by Olaf Czaja for the study of the monuments that interest us here. However, it should be mentioned again that the organization of this first tier is not totally consistent with the texts, since logically Chakrasamavara should have been located in the center of the west side.

Fig. 135

Apart from the representations of Vajravarahi hanging on the edge and that of Chakrasamvara, all we can perceive on this eastern side of tier 1 is a dozen portraits of lamas (Fig. 135). Four are arranged on stepped thrones with halos (Prabhāmaṇḍala). Most of the others are devoid of lotus pedestals. We had already noted this detail with the Chakrasamavara of which we have just spoken and with two figures attributed to the tier 1 of the tashi gomang of 1267. We will find this peculiarity on many sculptures from the tier 1 of the following monuments. It must therefore be a special convention on this floor.

Fig. 136

Fig. 137

Again, we find that this does not match with the description of the texts describing tashi gomang for Dorje Gyalpo Phagmo Drupa, which indicate that on this side should be six representations of Dorje Gyalpo Phagmo Drupa as the five family Tathagata and Vajradhara with the six Paramitas as their partner. This has nothing to do with our 12 portraits of monks which seem more to constitute a spiritual lineage probably associated with the master to whom the stupa is dedicated. As they stand, these portraits, which all look more or less alike, are absolutely unidentifiable. The one photographed by Ulrich Von Schroeder at the Jokhang of Lhasa could be one of them (Fig.136). In the middle of the upper row, balanced on two thrones, a portrait of a bearded lama is of particular interest. It is most probably the one in the Bordier collection at the Tibet Museum in Gruyère (Fig.137). It is identified as "Chöškyi Dorje Dragpa Jungne" by inscription. Given the context, it can only be Dragpa Jungne (1175-1255), third abbot of Densatil. The white pigments that seem to have originally covered his beard and hair of an eighty years old man have been removed, leaving the ungilt copper visible. The realization of this sculpture could be slightly posterior to the construction of this third tashi gomang.

Fig. 138 One of these portraits has the peculiarity of not having the head marked by a hairline. It seems more that this one is wearing a monk's cap, which could indicate that it is a portrait of an Indian master, possibly Atisha, who, we have seen that he had an important place in the lineage to be found at the top of the tashi gomang stupas (Fig.138).

A Kukkuripa from the Tibet Museum in Lhasa (Fig.139) and a Vajraghanta from the Capital Museum in Beijing (Fig.140) should also find their origin at the top of this tashi gomang. Another mahasiddha in a dancing attitude brandishing a damaru should be stylistically classified with the two previous ones (fig.141). The rich inlays of his bone apron are reminiscent of those visible on the representations of Chakrasamvara and Vajravarahi also attributed to this monument. We still have to determine his identity. At first glance, one could think of a Damarupa, but this mahasiddha does not appear in the lineage established by Jigten Gonpo in 1198. Given the iconographic variations we see from tashi gomang to tashi gomang, we cannot rule out the hypothesis that Damarupa was part of the initiatory lineage of Dragpa Sherab. Close examination of this sculpture reveals that it is in fact only fragmentary. Projections appear behind his arms and earrings showing that the whole thing must have been complex.  A tenon under his left foot was supposed to hold him on a kind of pedestal. A representation of Dombi Heruka that we will see at the seventh tashi gomang presents him with his snake, sitting on his tigress and brandishing a damaru. It is therefore possible to consider that we are here facing another representation of this Dombi Heruka who is also sometimes represented dancing on his tigress.

Fig. 139

Fig. 140

Fig. 141

Pillars
Just like the previous tashi gomang stupas, from tiers two to five, the floors are connected by pillars, each decorated with two leaning four arms bodhisattvas forming caryatids resting on lotuses emerging from a kalasha (immortality liquor vase). The photographs of this monument taken by P.F. Mele once again give us a perfect idea of the typology of these pillars. They are now executed with more care, making their quality more uniform. The stone inlays on the vegetal elements emerging from the vases are more consistent with those covering the sculptures resting on the floors. The lotuses are more widely spread with more finely shaped petals and the capital is marked in its center with a row of larger pearls than on the previous tashi gomang (Fig.142). 

Fig. 142

Fig. 143

Fig. 144

On the photographs of this tashi gomang, we can, as on the previous monument, observe at the corners of each tier, pillars the other type with bases composed of a vase from which are emerging loops of lotus scrolls encircling deities or auspicious symbols and capitals structured in the shape of a lotus with erect petals. In comparison with the photographs, several of them are listed in the collections (FIG.143). It should be noted that the two types of pillars are supported by identical vases resting on square bases and that the treatment of the lotuses on which the leaning bodhisattvas rest is very similar to that of those at the heart of which emerge the upright lotuses forming capitals. 

At all the corners of tiers 2, 3, and 4 of the east side of this third tashi gomang we can observe tri-lobed lotus petals adorned with deities, of the same typology as the two mentioned on the previous monument. The photograph shows us several types, at least two of which are listed in several copies in the collections (Fig.144). Again, it would seem conceivable that some of the original elements have been displaced over time and that the monks have grouped together on this monument a large part of those that remained. For reasons of stylistic logic, it would be conceivable that the model with inclusions of red semi-precious stones could originate from this third tashi gomang, and that the one with a vegetal border with turquoise inclusions should be more associated with the fourth.

Abbot 10 / Stupa 10 / Tashi Gomang 4 - Dragpa Changchub (Grags pa byang chub) (1356-1386)

Dragpa Changchub occupied the throne of Densatil from 1370 to 1386.  From 1374 to 1381 he occupied at the same time the Neudongtse’one, thus reuniting temporal and spiritual power in his hands. He can thus, in a way, be considered as the first "king of Tibet" to obtain a tashi gomang in Densatil. With the power of the Lang/Phagmodrupa clan at its peak, the stupa dedicated to such a prestigious member of the family could only be equal to the prestige of its recipient.

Fig. 145 This monument is documented by multiple sources, the most prestigious being his biography written by Tsongkhapa. [47] This is indicative of the strong links that still seem to have existed between the Kadampa/Gelugpa school and the Lang/Phagmo Drupa clan. Although constituting a branch of the great Kagyupa school, they always supported the political ascent of the Gelugpa until the accession to power of the fifth Dalai Lama in 1642. The latter will always, at least on the surface, hold them in high esteem by considering them in his historical writings as "kings of Tibet". [48]

Stylistic analysis of the tashi gomang visible in P.F. Mele's photographs and of the groups of objects attributable to all of them allows us to propose that this fourth tashi gomang may be the one we see in a photograph, located south of the east wall (Fig.145). In the background, to the right of the tashi gomang, we can also see what was probably Dorje Gyalpo Phagmo Drupa's thatched hut.

To try to illustrate the peculiarities of this fourth tashi gomang from the objects that could constitute a fourth coherent stylistic group, we will retain the following few works.

The Four Guardian Kings
A sculpture and two heads coming without any doubt from the same set of four guardians constitute a perfect stylistic transition between the three types of Lokapalas observed on the first three tashi gomang stupas and those we will meet on the next four monuments. The complete sculpture having lost its attributes it is difficult to identify it precisely. It could be a Vaishravana or a Viruḍhaka. It is now kept in the Museum of Turin in Italy. He wears a kind of armour derived from those seen in the previous examples but simplified and less inlaid with coloured stones but already presenting all the structure of those more ornamented with engravings and enriched with stones of the examples that would follow in the fifteenth century (Fig.146).  The head of the statue, like the two others (Figs.147 & 148), definitively abandons the Nepalese style of its predecessors to present all the characteristics of the Chinese style which will be that of the then nascent Ming Dynasty.

Fig. 146

Fig. 147

Fig. 148

One will note on P.F. Mele's photograph that at least one of the Guardian Kings sculptures is clearly visible in front of the tashi gomang as on the two previous ones.

TIER 6
A certain number of plates presenting deities surrounded by large lotus scrolls and leaning against other lotus loops surrounding small secondary deities can be attributed to tier 6 of this fourth tashi gomang. We can be almost certain that these plates belong to the fourth tashi gomang since from the fifth and therefore the fifteenth century, their structure will be different. The lotus pedestals here have only five petals, all treated with great nervousness.

Even when they have been broken or burned, all these plaques are of remarkable artistic interest and are undoubtedly the work of accomplished masters.

Fig. 149

Fig. 150

Fig. 151

The Nagarajas that in pairs occupied the corners of each side are still richly adorned and inlaid, and bear caps of nine snake heads spreading over their buns (Fig.149, 150,151).

A remarkable aspect of Lhamo in its Dhumavati Shridevi form was to be positioned on the south face (Fig.152). Her diadem is adorned with five skulls surmounted by flowers inlaid with semi-precious stones, the central one being framed by two vegetal stems. We will find this detail several times on the objects attributed to this tashi gomang and the next two.

Fig. 152

Fig. 153

Fig. 154

On the western side, we can mention an aspect of Mahakala (Legs ldan nag po chen po) with diadem adorned, as on the lhamo, with five skulls surmounted by flowers inlaid with semi-precious stones, the central one being framed by two vegetal stems. (Fig.153) A Jambhala should also come from this west side. (Fig.154).

On the northern side, should come Druma (Fig.155) and again Lhamo in its Saptadasasirsi Sridevi aspect (Fig.156).

Fig. 155

Fig. 156

TIER 5
At this tier five, as on all the tashi gomang of the Phagmo Drupa type were the sixteen goddesses of sensual enjoyment bearing offerings and playing music. A few plaques each featuring four of these goddesses have survived the destruction. The aesthetic continues that already seen on the tashi gomang of 1370, but here the deities rest on two-row pedestals of fourteen or sixteen lotus petals, depending on where the plate was to be placed (Fig.157). They are also more covered with inlaid jewellery, the semi-precious stones going as far as to decorate the ends of their buns. Some of them show the indication of a scarf forming a halo around their heads (Fig.158). This detail had already been observed on one of the plaques attributed to the tashi gomang of 1276, and will be systematic on all monuments of the fifteenth century. Some are still associated with their corner plates which are always decorated with four lotus scrolls (Fig.159).

Fig. 157

Fig. 158

Fig. 159

While the groups of Lokapalas, tier 6 deities and offering goddesses each have a relative stylistic unity, the same cannot be said for the individual deities.  Examination of the goddesses in this tier 5 shows that the eight known goddesses that should come from there are clearly the work of five or six different artists.

In the middle of the eastern side, we can propose to place two goddesses out of the three that should have been there, Parnashavari and Janguli. As with the two specimens observed on the previous two tashi gomang, the goddess Parnashavari (Fig.160) has her torso covered with a richly stone-encrusted garment that seeks to imitate the foliage associated with her traditional iconography. His armbands correspond to what will be one of the two dominant styles on this fourth tashi gomang. Its dhoti is richly decorated with floral motifs inlaid with semi-precious stones. Many of these have a four-petalled shape, that will recur on many of the sculptures attributable to this monument. A scarf unfolds behind her, surrounding the top of the head before forming an arc following the curve of the shoulders. This pattern is one of the striking stylistic elements of this fourth tashi gomang. As if to further magnify the goddess, inlays were placed on the part of the scarf surrounding her head. The lower sleeves of his garment and the lower part of his diadem have spaces designed to receive threads supporting strings of natural pearls that have now disappeared.

Fig. 160

Fig. 161

Although less richly adorned, the figure of Janguli (Fig.161) clearly comes from the same workshop and bears the same type of armbands and floral motifs inlaid on the dhoti as Parnashavari.  Her scarf has unfortunately largely disappeared.

In the center of the south side, there must have been an aspect of a four-faced, eight-armed Tara whose armbands are characteristic of the second dominant style on this fourth tashi gomang and whose dhoti bears inlaid flowers with four petals as we have seen on Parnashavari and Janguli (Fig.162). To his right was to be a three-faced, six-armed Usnisasitatapatra (Fig.163). On the other side of the Tara we find a four-armed goddess holding the probable stem of a broken blooming lotus, the stem of another supporting a vajra and a triple jewel takes up the exact iconography of the one we had considered as an aspect of Chunda on the tashi gomang stupas of 1267, 1360 and 1370 (Fig.164). Her Dhoti is also richly inlaid with floral motifs, including a number of stones carved like a four-petal flower.

Fig. 162

Fig. 163

Fig. 164

In the middle of the west side, must have been an aspect of Dhvajagrakeyura with four faces and eight arms which despite its broken attributes is most certainly this goddess in prathyalidhasana (Fig.165). To her left must have been this aspect of Mahasahasrapramardani also with four faces and eight arms (Fig.166). Her dhoti is richly inlaid and the scarf unfolds behind her, surrounding the top of the head before forming an arc following the curve of the shoulders. The top of each of the three arches, above the head and shoulders, is embellished with a three-petalled floral element that on other sculptures may sometimes be inlaid with stones.

Fig. 165

Fig. 166

Fig. 167

On the northern side, should come this Mahapratisara which presents all the characteristics of the Mahasahasrapramardani seen on the west side, indicating that it is most probably the work of the same artist (Fig.167).

TIER 4
Few objects have come down to us from this Buddhas floor. P.F. Mele's photographs gave us an idea of what the ornate Buddha plates on the tashi gomang of 1360 and 1370 looked like. A few plates following the same pattern but with the arch to which the deities are leaning rising less high and leaving space for a full row of figures in the upper part seem to mark an evolution with the previous ones. It seems stylistically logical to attribute them to this fourth tashi gomang. One of these plates with 50 diminutive representations of Amitabha should therefore come from the west side (Fig.168). A hole is visible at the bottom of the plate. It was designed to allow the fixing tang of the statue leaning against it to pass through. Two offering goddesses are depicted on the corner plate attached to it. A representation of Amitabha must also come from the same western side (Fig.169). The position of the tang fixed to the back of its lotus pedestal seems totally adapted to the height of the perforation of the plate. (Fig.170).

Fig. 168

Fig. 169

Fig. 170

A plate of the same type decorated with 50 diminutive representations of Amoghasiddhi should theoretically come from the north side (Fig.171). Two goddesses of sensual enjoyment bearing offerings are shown on the side plate attached to it. Unlike the plate from the west side, these two goddesses are surrounded by lotus scrolls (Fig.172).

Fig. 171

Fig. 172

TIER 3
The eastern side of this level is assigned to the Vajradathumandala in 47 deities. We can attribute to the middle of this side the Vairocana from the Yury Khokhlov collection (Fig.173).  The god sits on a lotus pedestal resting on a stepped plinth. Its armbands are of one of two common types attributable to this tashi gomang. As already seen on the lower levels, the scarf is surrounding the top of the head before forming an arc following the curve of the shoulders. This movement of the scarf is found on an Aksobhya whose lotus also rests on a stepped plinth (Fig.174). An Amoghasiddhi should also find its place on this side (Fig.175). As for those of Tier 5, these sculptures reveal the multiplicity of artists who worked on this tashi gomang.

Fig. 173

Fig. 174

Fig. 175

At each corner of this floor was to be a representation of Acala. The treatment of the scarf and armbands, which are to be compared to those of the Parnashavari of the fifth tier, leads us to attribute this one to this fourth tashi gomang (Fig.176).

Fig. 176

Fig. 177

Fig. 178

It is conceivable to attribute to the south side a Vajrasattva (Fig.177), and a Manjushri whose lotus pedestal lies on a stepped base (Fig.178). Remaining fragments of the scarf indicate that it was the classical pattern surrounding the top of the head before forming an arc following the curve of the shoulders. According to the descriptive texts of the tashi gomang visualized by jigten Gonpo for the Dorje Gyalpo Phagmo Drupa monument, the only aspect of Manjushri to adorn a tashi gomang should be the "lion of debaters" aspect. The iconography specific to this aspect should show him sitting on the back of a lion. As for the one already observed on the tashi gomang of 1360, the lion appears here only in the base of the throne. The contents of the eight groups of objects attributable to Densatil's eight tashi gomang allow us to see that the iconography presenting the god sitting on the back of the animal does not seem to appear until the fifteenth century on the last monuments of this type.

Fig. 179 In the center of the west side, should be placed an aspect of the four-armed goddess that we have already encountered on the three previous tashi gomang and that we had considered to be an aspect of Prajnaparamita (Fig.179). Her attributes resting on the lotus spread out above his shoulders are a conch and a vase. Her tiara, jewellery and dhoti are richly inlaid with semi-precious stones. The scarf surrounds the top of the head before forming an arc following the curve of the shoulders following the classic pattern on this stupa. Since the conch is not a classical attribute of the goddess, its identification could be questioned.  We explained above that only two female deities with one face and four arms can appear on a tashi gomang following the model described by Jigten Gonpo. The other being more clearly identified, this one could therefore only be an unusual aspect of Prajnaparamita, unless on this precise point the reality of Densatil's tashi gomang does not coincide with the texts describing the monument visualized by Jigten Gonpo in 1198.

Fig. 180 As on the previous tashi gomang stupas, the north side was to be occupied in its center by a Chakrasamvara mandala in 14 deities.  We have observed on the two preceding tashi gomang stupas apsects of the six pairs of arms god dancing in ardhaparyaṅkāsana with Vajradakini. The objects in the eight groups attributable to the eight tashi gomang stupas lead us to believe that from this fourth monument, it is the more classical form, still with six pairs of arms, but in ālīḍhāsana that will occupy this location. The one we attribute to this side has a decorated diadem with five skulls on discs, the central one being framed by two vegetal stems as seen on some divinities from tier 6 and as we will find on the Chakrasamvara in the upper tier (Fig.180). The Vajrayogini’s bone apron is structured like a garland of a typology very close to those we will be able to observe on the Chakrasamvara and Vajrayogini figures of the two upper floors.

TIER 2
One of the most remarkable representations of Guhyasamaja in Tibetan art is undoubtedly the one that logically should occupy the middle of this side (Fig.181). The couple's ornaments are very finely inlaid with coloured stones and were originally underlined by rows of natural pearls that are still partially preserved. A probable Ratnasambhava must also have been placed on this side (Fig.182).

Fig. 181

Fig. 182

In the middle of the south side was to be positioned a Vajrakilaya (Fig.183). The god embraces his companion while maintaining his eponymous attribute with his two hands. As it appears on other sculptures of this fourth tashi gomang, his scarf is surrounding the top of the head before forming an arc following the curve of the shoulders. The headband of his diadem shows the hanging system of a string of fine pearls that has now disappeared. To his right must have been a Kalachakra (Fig.184). The skulls of his tiara present a treatment close to that observed on the Chakrasamvara figures of the second and third tiers. A band of white pearls is still partially preserved on the top of his forehead.

Fig. 183

Fig. 184

The northern side of this second tier is normally decorated with a large Chakrasamvara mandala of 62 deities. This seems to imply that a representation of the god embracing Vajrayogini occupied the middle, leaning against a vast repoussé decoration on which were represented the various deities of the mandala as well as the mahasiddhas and charnel grounds that traditionally surround the mandala. We propose to place here a representation of a Chakrasamvara whose Vajrayogini is dressed in a bone apron with the same patern as the one seen on the Chakrasamvara of the lower level, and which we will find on the Chakrasamvara of tier 1 (Fig.185). The god has a decorated diadem with five skulls on discs, the one in the middle being framed by two vegetal stems as seen on the one from tier 3. The armbands of Vajrayogini with lozenge-cut stone inlays are similar to those of a Hevajra that we will find in tier 1.

Fig. 185

Fig. 186

A four-armed bodhisattva with only an ewer as attribute does not correspond to any precise iconography in the descriptive texts of the tashi gomang for Dorje Gyalpo Phagmo Drupa (Fig.186). Another version is known from the following tashi gomang who keeps a rosary in his right hand. It might be conceivable that this could be a particular aspect of Avalokiteshvara of which four declinations are arranged at the corners of this level 2. This one is dressed according to the common mode on this tashi gomang with the scarf surrounding the top of the head before forming an arc following the curve of the shoulders.

TIER 1
We have for reasons that we will explain later, considered that this tashi gomang of 1386 for Dragpa Changchub was the one located in the south-eastern corner of the shrine, near the hut of Dorje Gyalpo Phagmo Drupa. In the only photograph taken of it by P.F. Mele, we can see the Kadampa-type stupa wrapped in textiles resting on the tier 1 and Vajravarahi figures arranged on the perimeter of this level. Careful examination of this image reveals that none of these dakinis appear to have been placed on the rear half of the monument, except maybe one that might be seen in place in the middle of the west face (Fig.187).

Fig. 187

Looking at the photographs of the preceding tashi gomang, one can get the impression that seven dakinis are placed on the east side, five in the middle and two on the southeast and northeast corners, but no more than two have been placed at least on the south side. This is the same disposition that seems to have been put in place on the fourth tashi gomang. So, if we consider that seven are placed all along the east side, and that two have been placed at each beginning of the south and north sides, we arrive at a number of eleven.  If, as is very likely, the visible shadow in the center of the back of the first tier of our fourth tashi gomang is indeed a Vajravarahi, we arrive at the number of twelve which corresponds to the number given by the texts describing Jigten Gonpo's vision in 1198. It is thus conceivable that on this point at least, the abbots of Densatil did not modify the number of these dakinis which originally had to be arranged three by three in the middle of each side, but preferred to group them together as much as possible on the front of the monuments for an ostentatious purpose.

Fig. 188

Fig. 189

Fig. 190

For the lower tiers, we had noted the multiplicity of artists who worked at the same time in significantly different substyles that marked their individuality. It is the same on this tier 1 to which we can attribute a first Vajravarahi which shows the archetypes seen previously but with the scarf unfolding behind her, surrounding the top of the head before forming an arc following the curve of the shoulders (Fig.188). The top of each of the three arches, above the head and shoulders, is embellished with a three-petalled floral element inlaid with stones. A second Vajravarahi attributable to this monument is the many times published Vajravarahi preserved in the Los Angeles County Museum (Fig.189). The scarf follows the classical form on this tashi gomang by following the curves of the head and shoulders, but the artist has attributed it a very particular and complex movement in the parts falling down under the arms. This very particular treatment of the scarf makes it possible to attribute this dakini to a sculptor who also executed a Vajradhara, which can only come from the same monument. (Fig.190). The armbands of this Vajradhara inlaid with lozenge-cut stone are similar to those of Vajrayogini embracing Chakrasamvara seen in tier 2 (Fig.185).  We have already noted that many of the deities and portraits of lamas on top of the tashi gomang stupas were missing lotus pedestals. It should be noted here that the base on which the figures rest is thicker than what we have seen so far and that this detail is found on other objects attributable to the first tier of this monument. We can thus mention a portrait of a monk with round forms and a face marked with a beard which is most probably a representation of Dorje Gyalpo Phagmo Drupa (Fig.191). A mahasiddha Luyipa identifiable by its fish also rests on a simple thick base of the same typology (Fig.192). He also wears a scarf around his head and shoulders, and the treatment of his kapala is identical to that of the Vajravarahi by the same artist.

Fig. 191

Fig. 192

Fig. 193

An aspect of the two-armed Chakrasamvara was to be placed in the center of the west side, although we have previously encountered it highlighted on the east side. We can attribute to this location the remarkable two-armed Chakrasamvara embracing Vajrayogini once in the Pan Asian collection and published many times (Fig.193). The scarf encircling the god is remarkably flamboyant and the Vajrayogini bone apron is of the same typology as those of the two Chakrasamvara observed on the lower two tiers. (Fig.193)

PILLARS
A set of columns supported by representations of leaning bodhisattvas resting on lotuses emerging from longevity vases and supporting the capital each with two arms can be attributed to this tashigomang (Fig.194). They are very similar to the examples of the previous tashi gomang, but the capital is structured here with stepped edges reminiscent of those on the face of the monument, and also features a row of large pearls in its center. The bases supporting the vases are also structured here with stepped edges. This detail being found on some corner pillars (Fig.195), it would seem logical to attribute them to this fourth tashi gomang.

Fig. 194

Fig. 195

The pillars of these third and fourth tashi gomang stupas unequivocally mark the apogee of the style. Their quality will experience an obvious decline throughout the fifteenth century.

The stupas of the fifteenth century

At the beginning of the fifteenth century, the succession of abbots was complicated by the abdication of the eleventh and the death of his successor a few months before him. Logic would therefore suggest that the erection of the funeral monument to the twelfth abbot began only a few months before that of the eleventh, and that the two projects could have been temporarily carried out at the same time.

Based on the few historical texts to which he had access and finding no specific mention of a tashi gomang for Sonam Dragpa, Olaf Czaja infers that none was built for him. This shortcut is convenient for his demonstration insofar as we have a certainty that there were eight tashi gomang in Densatil and that he attributes one to Rinchen Dorje in 1280, thus leaving the possibility only for three in the fifteenth century. We have outlined above the various reasons why the existence of this tashi gomang of 1280 is impossible. 

However, Olaf Czaja supports his choice by referring to the biography of Dragpa Gyaltsen (Grags pa rgyal mtsan - 1374-1432) who occupied the family throne of Neudongtse from 1385 to 1432 and is said to have ordered three tashi gomang stupas. The first of them is the one for Dragpa Changchub who had preceded him on the throne of Neudongtse from 1374 to 1381 while being abbot of Densatil. [49] The existence of the one for Paldan Zangpo erected in 1407 being attested, and still considering that only three tashi gomang were erected in Densatil in the fifteenth century, he also attributes to Dragpa Gyaltsen the construction in 1431/1432 of the monument for Sonam Zangpo who died in 1417. However, the Blue Annals report that this tashi gomang would have been ordered by Sonam Gyaltsen (Bsod nams rgyal mtshan – 1386-1434) who would have started the works in 1431 [50]. This information should be considered reliable, since Gos lo tsa ba gzhon nu dpal, (1392-1481) who compiled this historical sum was contemporary to these events and close to the court of the Lang/Phagmo Drupa clan. Olaf Czaja relies on the biography of Sonam Gyaltsen which states that he met Dragpa Gyaltsen in 1431 and discussed plans to erect a tashi gomang. [51] The fact that he was consulted for the plans does not make him a sponsor of the monument.

On the other hand, it would seem surprising that Dragpa Gyaltsan, who did everything during his long reign to assert his power and the greatness of the Lang/Phagmo Drupa clan, had a tashi gomang erected for Dragpa Changchub who occupied the two thrones of Neudongtse and Densatil, one for Paldan Zangpo who was certainly his brother but only occupied the throne of Densatil for two years, but none for Sonam Drakpa who was abbot of Densatil from 1387 to 1405 after having occupied the Neudongtse’s throne from 1381 to 1385. Considering that Densatil's tashi gomang stupas seem to have had both a religious and a political role, it even seems totally inconceivable.

On the other hand, we have evoked since the beginning of this panorama the existence of eight groups of objects that can be linked to Densatil's tashi gomang stupas. We have already observed four that may correspond to the monuments of 1267, 1360, 1370 and 1386. Four remain to be attributed to the fifteenth century. It so happens that we are in the presence of two groups of objects that seem to be in the continuity of the previous ones but have strong similarities between them. This sometimes gives the impression of having two almost twin stupas in many ways and could therefore indicate that these two tashi gomang stupas were built, if not at the same time then at least one after the other in a fairly short period of time. It therefore seems logical to attribute these objects to two tashi gomang stupas built for the eleventh and twelfth abbots around 1407-1409. It seems, therefore, that the objects here can fill in the gaps in the historical texts, which, especially in the field of Tibetology, are often rather succinct and abstruse for everything that does not directly belong to the spiritual lineages.

Here we find ourselves at about the same point as Christian Luczanits in his study on the descriptive texts of Dorje Gyalpo Phagmo Drupa's tashi gomang, who in his conclusion marking his difference of interpretation with Olaf Czaja wrote: « The issue at hand is thus clearly a conceptual and methodological one and reflects the different perspectives we have according to our respective fields and also the different ways of assessing the material at hand.
The crucial question from my perspective is the extent to which the fragmentary art historical evidence can be used as a ‘historical’ source.” [52]

This proximity in the realization of the monuments makes it very difficult to determine what must have belonged to the stupa of 1407 rather than that of 1408. We have therefore chosen to divide the objects according to criteria that may seem subjective, but which for the most part, are based on what seems most logical in relation to a stylistic chronology. Perhaps new information or unpublished objects will one day allow us to refine the proposal for these two tashi gomang stupas.

In order to maintain the chronological logic, we will therefore deal first with the eleventh stupa and fifth tashi gomang erected for the twelfth abbot.

Stupa 11 -Tashi Gomang 5 – Abbot 12  Paldan Zangpo (Dpal ldan bzang po) (1383- 1407)

Fig. 196 Paldan Zangpo occupied the throne of Densatil from 1405 to 1407. As with his predecessor, his untimely death led to difficulties within the Lang/Phagmo Drupa clan for the appointment of his successor. [53] His tashi gomang is mentioned in the dbus gtsang gnas yig of  Chos Kyi rgya mtsho [54] According to the same source, it is said to have housed 3,900 figures, which leads us to suppose that it followed the model of the one in 1370, which would have counted the same number. [55]

According to the logic of the stylistics and the probable arrangement of the eight monuments in the sanctuary, this fifth tashi gomang should be the one of which we have a partial view of the south face on the right side of a photograph by P.F. Mele mostly showing the one of 1360 (Fig.196). Insofar as we have photographs of the eighth and know some important elements of the seventh, it is at least certain that the one visible in this photograph can only be the fifth or sixth tashi gomang of Densatil.

To try to illustrate the peculiarities of this fifth tashi gomang, we will retain the following few works.

The Four Guardian Kings

Fig. 197 The three lokapalas figures attributable to the two tashi gomang stupas of 1407 and 1408 show a number of similarities, but the presence in the group of two Virupaksha gives us confirmation that they do indeed belong to two different monuments. We have chosen to attribute to this fifth tashi gomang the Virupaksha preserved in the Guimet Museum in Paris. (Fig.197) This representation presents a costume generally taking again that observed on the Dhritarashṭra attributed to the tashi gomang of 1267 but treated with more details on the level of the armor and important inlays of turquoises and lapis lazuli. Unlike that of the Dhritarashṭra of 1267, the head of this Virupaksha is like those of the Guardian Kings of the previous monument with Chinese-style hairstyles. This Virupaksha therefore happens to be totally executed in the Chinese Ming style.

TIER 6
A certain number of plates presenting deities surrounded by large lotus scrolls can be attributed to tier 6 of these fifth and sixth tashi gomang stupas.

All these plates are stylistically very close to each other. The most significant difference that can be established between the two groups lies in the rim of the lotus pedestals, which on one is smooth and on the other is marked with beads probably seeking to indicate the stamens of the flower on whose pistils the deities rest.
We have chosen for this fifth tashi gomang the smooth edge group.

As with the previous two monuments, the deities are solidary cast with their background. However, from this fifth tashi gomang, an important change appears on the backgrounds on which are represented secondary deities associated with the main one. The loops of lotus scrolls that encircled these secondary deities have now disappeared, leaving them arranged in space in a less structured manner than before.

Even if all the sculptures coming from this tier 6 are of good quality of execution, it is obvious that more or less gifted artists collaborated on them at the same time. This could perhaps be explained by the development of native artists' workshops freeing themselves from the Nepalese masters who seem to have worked in greater numbers on the previous tashi gomang stupas.

Fig. 198

Fig. 199

Fig. 200

Fig. 201

Fig. 202

Fig. 203

Fig. 204

This is obvious for example if one compares two Nagarajas. The first one wears jewels that still refer to the greatness of fourteenth-century monuments, with finely inlaid ornaments and flowers also inlaid on its dhoti (Fig.198). The second one, although surrounded by a scarf, is for its part treated in a rougher way (Fig.199). Apart from these two Serpent Kings, which it is difficult to attribute to a precise side of the monument, a Rahula that was supposed to be on the south side was burned but survived the destruction. (Fig.200). We can attribute to the west side a Jambhala (Fig.201) and to the north side a Druma (Fig.202), an aspect of Mahakala (Fig.203) and an aspect of Lhamo (Fig.204).

TIER 5
At this level five, as on all the tashi gomang stupas of the Phagmo drupa type were on each side the sixteen goddesses of sensual enjoyment bearing offerings and playing music. The goddesses rest here on a pedestal lotus with two rows of eleven to thirteen petals depending on the width of the site they were to occupy. They all now have their heads encircled by the movement of their scarves and are dancing under a garland draped across the top of the frieze (Fig.205 et Fig.206). Their side plates are decorated with four loops of lotus scrolls as on the previous tashi gomang. The majority of these plaques as well as those that we will attribute to the next monument carry indications intended for their positioning on the monuments by the inscription of a letter of the Tibetan alphabet.

Fig. 205

Fig. 206

Fig. 207

In the middle of each side of this tier 5 were initially three goddesses. A Parnashavari was to occupy the middle of the east side (Fig.207). Even though she is still richly inlaid with semi-precious stones, this goddess has lost much of the splendour of the specimens we saw on the tashi gomang stupas of 1360, 1370 and 1386. The foliage that is supposed to make up his garment is here treated in relief, in a more or less realistic way, and no longer by rich inlays. Two rings at the belt level must have been used in the past to hold the upper part of his scarf.

A Tara with four faces and eight arms must have been in the center of the south side (Fig.208). It too is executed with great refinement, but it has also lost some of the splendour of those observed in the fourteenth century. Its armbands are the first we encounter of a model that will be found with variants on many objects throughout the fifteenth century. The gem is there framed by two vegetal stems. At his side, must have been a four-armed goddess holding a vajra, the probable stem of a broken blooming lotus, and a triple jewel that takes up the exact iconography of the one we had considered as an aspect of Chunda on the tashi gomang stupas of 1267, 1360, 1370 and 1386 (Fig.209). Its dhoti is still decorated with floral motifs signified by inlays of semi-precious stones, as seen on at least one Serpent King from Tier 6 (Fig.198) and on many sculptures from previous monuments. Its ornaments are underlined by inlays of natural pearl threading. However, the white pearls on her necklace and tiara appear to be recent replacements.

Fig. 208

Fig. 209

Fig. 210

On the northern side should come a superb aspect of Kurukulla (Fig.210). The face of this Dakini seems very close to those of the offering goddesses who must have been on the same level. Her bone apron is of a model that abandons the old "garland" style that we have observed on the dakinis and Vajrayoginis associated to Chakrasamvara of the fourteenth century. This new type of apron now makes the successive arches of the old garland shape cross each other, and floral pattern fasteners appear at the points of intersection.

TIER 4
The south side of this floor dedicated to the Buddhas is partially visible in the photograph by P.F. Mele (Fig.211).

Fig. 211

The plates supporting the diminutive images of the Buddhas are of a new type of which no specimen has reached us. Careful examination of the sculptures on this side, however, raises some questions. Logic would have it that we find a Buddha in the center surrounded by Vajrapani and Acala on his right and Akasagarbha and Amrtakundalin on his left. Four representations of Ratnasambhava should have rested in the four free spaces distributed on either side of the central Buddha. On the left of the photograph, through the space left free by the foreground tashi gomang, we can glimpse an image of Buddha, but obviously the representations of Vajrapani and Acala, and Akasagarbha and Amrtakundalin, which should have surrounded him, are absent.

In the locations intended to house images of Ratnasambhava, we can observe, on the left, a two-armed deity with a tiara on its head indicating that it is not a Ratnasambhava, and on the right, another six-armed deity that is difficult to identify. We can therefore see once again that P.F. Mele's photographs clearly show that in 1948 this tashi gomang had seen some of his images moved to be used on later monuments or taken elsewhere. Another hypothesis could be that not all the images initially planned were realized and that sculptures of various origins were put in place to fill in the gaps before the tashi gomang was consecrated.  

TIER 3
The eastern side of this tier is assigned to the Vajradathumandala in 47 deities. We can propose to place on this side an Amoghasiddhi that still generally follows the aesthetics observed on the previous monuments (Fig.212). An Aksobhya with a lotus pedestal resting on a stepped base has a new type of armbands approaching those of the eight-armed Tara on Tier 5 (Fig.213). An Amitabha also synthesizes ancient forms with new stylistic features. (Fig.214)

Fig. 212

Fig. 213

Fig. 214

The south side is partially visible in the photograph by P.F. Mele. As with the lower tier, the normal iconographic plane is not maintained here (Fig.215). On the right, instead of the aspect of Manjushri, we can see a deity in yab-yum, and on the left, a female deity which does not correspond to any of those appearing in the list of those logically adorning a tashi gomang. It is most probably the one now kept in the Shivaji Museum in Mumbai. (Fig.216). The representation of Acala, which should be in the southeast corner, is missing.

As with Tier 4, the question arises here as to whether the original sculptures were removed for another use and replaced by what the monks found to replace them without iconographic discernment, or whether this improbable arrangement goes back to the origin of the monument.

Fig. 215

Fig. 216

Fig. 217

In the middle of the west side, should be placed an aspect of the four-armed goddess that we have already encountered on the four previous tashi gomang and that we had considered to be an aspect of Prajnaparamita. We do not find exactly the same iconography among the pieces attributable to this fifth monument. However, a goddess with one face and four arms treated in a style close to that of the previous tashi gomang should be considered as an evolution of this goddess, two of her hands still performing the teaching gesture and the other two now being raised to support two lotuses having had to support the attributes (Fig.217). Her dhoti is still decorated with floral motifs signified by inlays of semi-precious stones, as seen on at least one Serpent King from Tier 6 (Fig.198) and on many sculptures from previous monuments. The lower part of the lotus pedestal is damaged, indicating that it had to be separated by force from a stepped plinth as all previous versions of this probable aspect of Prajnaparamita had.

Fig. 218 As on the previous tashi gomang stupas, the north side was to be occupied in its middle by a Chakrasamvara mandala in 14 deities.  Insofar as two six-armed aspects of this deity were to be arranged on tiers 3 and 2, it is tricky to determine which one comes from which level. We will therefore select one arbitrarily and assign the other to the next level up. (Fig.218).  The armbands of this couple are inlaid with lozenge-cut stones similar to those of Vajrayogini embracing Chakrasamvara seen in tier 2 and the Vajradhara seen in tier 1 of the fourth tashi gomang (Fig.185 & 190).  On the other hand, the Vajrayogini’s bone apron is of the new type that now makes the successive arches of the old garland shape cross each other, with floral pattern fasteners at the points of intersection. 

TIER 2
The south side of this tier 1 appears partially in the photograph by P.F. Mele (Fig.219). We can notice the same disorganization there that we had already seen on the two lower levels. On the right, in the place that should have been occupied by Vajrabhairava, is placed a four-armed bodhisattva who is most likely the Sadaksari Lokeshvara who is usually in the southeast corner. In the middle, a portrait of a lama occupies the place logically assigned to Manjushri Yamari with three faces and six arms, joined to his female counterpart. On the left, in front of the central part of this row which usually houses Vajrakilaya, we see partially hidden by a pillar of the tashi gomang in the foreground the representation of a deity with his left knee resting on the ground and his foot raised. This iconography is characteristic of Acala's representations, which should be placed at each corner of the third tier.

Fig. 219

Fig. 220

Fig. 221

At each end of the south side was to be placed a figure of Vajrayogini (Fig.220). It is surrounded by a scarf reminiscent of the patterns seen on the previous tashi gomang, but the bone apron follows the new type that now makes the successive arches of the old garland shape cross each other, with inset fasteners at the points of intersection.

Fig. 222 The northern side of this second tier is normally decorated with a large Chakrasamvara mandala of 62 deities. As in the lower tier, an aspect of the god with six pairs of arms must have been leaning against it. (Fig.221). On this example again, the Vajrayogini’s bone apron is of the new type that now makes the successive arches of the old garland shape cross each other, with floral pattern fasteners at the points of intersection. The jewels of this couple are inlaid with lozenge-cut stones.

A four-armed bodhisattva with only an ewer and a rosary as attributes does not correspond to any precise iconography in the descriptive texts of the tashi gomang for Dorje Gyalpo Phagmo Drupa (Fig.222). We had observed a very similar one attributed to the previous tashi gomang. It might be conceivable that this could be a particular aspect of Avalokiteshvara of which four declinations are arranged at the corners of this level 2. Its armbands are characteristic of the style appearing on this monument.

TIER 1
P.F. Mele's photograph provides us with a partial view of the south side of this tier (Fig.223). The first observation that can be made is that this side is totally devoid of the representations of Vajravarahi that should logically be hung on the edge. Two portraits of lamas, one of which leaning against a halo are facing east. The general impression is that as on the lower floors, sculptures have been moved or have never been put in place.

Fig. 223

While no Vajravarahi appears on the south side, one is nevertheless visible on the southeast corner, indicating that at least seven must have been placed on the east side of the monument. Once again, the sculptures attributable to these fifth and sixth tashi gomang stupas are clearly the work of different artists. (Fig.224, 225, 226). The treatment of their scarves indicates that the artists still seek inspiration from the models of 1386, but their bone aprons are characteristic of the fashion of this early fifteenth century.

Fig. 224

Fig. 225

Fig. 226

Fig. 227

Fig. 228

Two representations of Vajradhara embracing Bhagavani were to be positioned on the east and south sides. It's most likely these two nearly identical sculptures (Fig.227 et Fig.228). Their difference lies mainly in the hairstyle, one having a high bun and the other a tiara. The other differences are minimal and are mainly in the treatment of Vajradhara’s anklets and of Bhagavani’s bone aprons, which also follow the new fashion of the early fifteenth century.

PILLARS
The photograph of the south face taken by P.F. Mele shows us four pillars which are not very sharp but whose typology is now familiar to us. Bodhisattvas standing on lotuses emerging from longevity vases are still supporting capitals (Fig.229).

Fig. 229

Fig. 230

These capitals are like the previous ones narrowed in their center, but we can't see if there are more or less large pearls in them. There also seem to be differences between them in the proportions of the vases and the capitals. They seem quite similar to those we have noted on the first two monuments, and perhaps the artists were inspired by the pillars of the second neighbouring tashi gomang. Given the multiple variants and potential subgroups observed in the pillars attributed to the first and second tashi gomang stupas, it is conceivable that some of them may in fact have come from this southern side of the fifth monument. With the exception of the plates in tiers 6 and 5, we have noticed a general lack of stylistic unity between the majority of the works attributable to this fifth tashi gomang. It is therefore not impossible that on the other sides of this monument there may have been pillars of a new style which seems to have been dominant over the sixth monument. This new style is essentially characterized by the envelopment of the capital by a large open lotus that is somewhat reminiscent of the capitals of the corner pillars. (Fig.230). The three examples presented here demonstrate the variety of styles and qualities of execution. This style will know a variant with the foot of the vase modeled in the shape of an open lotus that technically we will attribute to the following tashi gomang. However, it is not impossible that the two variations could have coexisted on the two monuments.

The photograph shows corner pillars following roughly the same pattern as those observed on the previous tashi gomang but does not allow analysis of the details.

As we have already seen on this fifth tashi gomang and this will be confirmed with the following ones, the talents of the artists who made the various sculptures were very unequal, which leads to a great disparity in the quality of these Pillars.

Stupa 12. / Tashigomang 6 / Abbot 11 - Sonam Dragpa (Bsod nams grags pa) (1359-1408)

As explained above, this twelfth stupa and sixth tashi gomang has for historical reasons been dedicated to the eleventh abbot of Densatil.

On the death of Dragpa Changchub in 1386, as dissension had been brewing among the Lang/Phagmo Drupa clan for some time, the appointment of a family member as the new abbot for Densatil seems to have been problematic. Sonam Dragpa who had occupied the throne of Neudongtse from 1381 to 1385 before abdicating in favour of Gragpa gyaltsen (Grags pa rgyal mtshan / 1374-1432) seems to have been sent to Densatil largely against his will. [56] In 1405, after showing little interest in his position, he abdicated and withdrew from the world until his death in 1408. [57]

We have also explained above why although the erection of a tashi gomang for Sonam Dragpa is not mentioned in any source known to date, it is impossible to imagine that from a political point of view this was not the case for the prestige of the dynasty and the monastery since in addition to having occupied the abbatial throne of Densatil, he had also been in his time "king of Tibet".

We have already explained above how difficult it is to determine what must have belonged to the stupa of 1407 rather than that of 1408 and why we have chosen to divide the objects according to criteria that may seem subjective, but which for the most part, are based on what seems most logical in relation to a stylistic chronology.

To try to illustrate the peculiarities of this sixth tashi gomang, we will retain the following few works.

The Four Guardian Kings
Of the three guardian figures attributable to the two tashi gomang stupas of 1407 and 1408, two have such similarities that they can only come from the same set. A Virupaksha is very close to the one we attributed to the previous tashi gomang, the essential differences being a slightly different treatment of the face and the inclusion on it of a large lapis lazuli rectangle in the centre of the strap of his armour passing over his belly (Fig.231). A lapis lazuli of the same shape is placed in the same place on a Virudhaka, confirming their association within the same group of four guardians (Fig.232). These figurestherefore happens to be totally executed in the Chinese Ming style. Insofar as the lokapalas of the last two tashi gomang stupas are relatively well known to us, we can be almost certain that these three come from the two monuments of 1407 and 1408.

Fig. 231

Fig. 232

TIER 6
The plates supporting the deities of this tier 6 are as mentioned above always encircled by lotus scrolls and very close to those of the previous tashi gomang. The small secondary deities in the background of the plates are no longer encircled by loops of lotus scrolls either.  The almost only difference separating the two groups is in the rim of the lotus pedestals which here is marked with beads, probably seeking to indicate the stamens of the flower on whose pistils the deities rest.
The comparison between three Serpent Kings allows us to quickly see the differences in the skills of the sculptors who worked on the monument. (Fig.233, Fig.234, Fig.235)

Fig. 233

Fig. 234

Fig. 235

Fig. 236

On the east side, we can place an aspect of Mahakala. (Fig.236)

On the south side, another aspect of Mahakala and one of Lhamo. (Fig.237 et Fig.238)

Fig. 237

Fig. 238

Fig. 239

Fig. 240

On the west side, Mahaganapati and possibly another unusual aspect of Lhamo. (Fig.239 et Fig.240)

This group of sculptures seems to be more inspired by the works observed on the five previous tashi gomang stupas and thus integrate some stylistic archaisms.

TIER 5
At this level five, as on all the tashi gomang stupas of the Phagmo Drupa type were on each side the sixteen goddesses of sensual enjoyment bearing offerings and playing music. The goddesses rest here again on lotus pedestals with two rows of eleven petals. They all now have their heads encircled by the movement of their scarves and are dancing under a garland hanging across the top of the frieze (Fig.241 et Fig.242). This garland no longer follows a realistic draped pattern as on the previous tashi gomang but is signified in a more schematic way with inclusions of semi-precious stones.  Their side plates are decorated with four loops of lotus scrolls as on the previous tashi gomang. The majority of these plates carry indications intended for their positioning on the monuments by the inscription of a letter of the Tibetan alphabet.

Fig. 241

Fig. 242

TIER 4
Few objects are attributable to this floor. We can nevertheless propose to place a Ratnasambhava which was to be placed on the south side. (Fig.243)

Fig. 243

TIER 3
The eastern side of this tier is usually assigned to the Vajradathumandala in 47 deities. As for the lower tiers, we have few sculptures to propose to place there. A Ratnasambhava whose lotus pedestal lies on a stepped plinth evokes by certain points works observed on the previous tashi gomang stupas. (Fig.244) It is certainly coming from a workshop that sought inspiration from what it could observe on the older monuments and to which other sculptures can be attributed. However, the dexterity of this (these) artist(s) is clearly less than that of his (their) predecessors, whether in the treatment of bodies or of jewellery and ornaments which are now only summarily inlaid with coloured stones.

Fig. 244

TIER 2
As on the previous tashi gomang stupas, the eastern side of this second tier was logically intended to illustrate the 72 deities mandala of Guhyasamaja. An aspect of Amitabha undoubtedly coming from the same workshop as the Ratnasambhava of the lower tier can be attributed to it. (Fig.245) It clearly gives the impression of being more or less copied from the one we attributed to the tashi gomang of 1360 with its lotus pedestal supported by lotus scrolls. (Fig.85).

Fig. 245

Fig. 246

Fig. 247

Fig. 248

We can attribute two sculptures to the south side. An aspect of Avalokiteshvara which was to be located at the southwest corner (Fig.246) and a Krishna Yamari (Fig.247) which was to be placed on the right side, between Vajrakilaya and Vajrabhairava. The aspect of Avalokiteshvara presents the same archaisms as the previous Ratnasambhava and Amitabha and should therefore be attributed to the same workshop. The Krishnayamari is executed in a style prolonging that of the two Vajradhara attributed to the previous tashi gomang (Fig.227 & Fig.228).

The Chakrasamvara that was to occupy the middle of the north side (Fig.248) undoubtedly comes from the same workshop as the Krishnayamari. He is very similar in appearance to that attributed to the same location on the previous tashi gomang (Fig.211), but there is nevertheless a lesser overall refinement of execution, which is particularly noticeable in the treatment of the ornaments and the Vajrayogini’s bone apron.

TIER 1
No significant pieces can be attributed to the tier 1 of this monument. A Vajravarahi synthesizing some stylistic elements of the previous tashi gomang stupas and being still different from those that will be attributed to the last two monuments could however find its place there (Fig.249).

Fig. 249

PILLARS
The pillars attributable to this twelfth tashi gomang should in principle be those which, as on the previous monument, present a capital wrapped in a large lotus, with the foot of the vase modelled in the shape of an unfolded lotus. (Fig.250) It is possible, however, that the two variations were spread over these two monuments of 1407 and 1408.

Unlike the first four tashi gomang stupas, these fifth and sixth monuments can only be attributed few sculptures, with the exception of the plates of the fifth and sixth tiers. In a photograph by P.F. Mele we could see that already in 1948 the arrangement of the sculptures seemed random and almost unrelated to the iconographic scheme of Jigten Gonpo's vision for the tashi gomang in memory of Dorje Gyalpo Phagmo Drupa. This can be explained in two different ways. Either the original images were moved and replaced by objects of chance, or the floors of the monument were furnished with various sculptures from the outset for simple decorative purposes, on the understanding that the uninitiated would see no difference. In both cases, these phenomena can be explained by a lack of time to be able to complete a construction at an auspicious date, or by a lack of money. It could be a mixture of the two.

Fig. 250

Fig. 251

Only the guardian kings, and the plates in tiers 6 and 5, which are the most visible objects since they are within direct view, as well as a few other deities, seem to have been entrusted to experienced artists. Many of the sculptures attributable to these two monuments reveal a drop in quality compared to those observed on previous tashi gomang stupas. It can be seen that the artists working on these tashi gomang stupas are no longer inspired innovators, but seek their inspiration in the works adorning the previous monuments. This artistic decline is undoubtedly related to the wavering of the Lang/Phagmo Drupa dynasty.

Another factor could also explain the small number of objects attributable to these two monuments. Indeed, we will see later that logic would dictate that they were positioned in the southwest corner of the sanctuary, whose pre-reconstruction photographs clearly show that the large walls were completely destroyed and therefore must have caused more damage to the objects still in place (Fig. 251).

Abbot 13 / Stupa 13 / Tashi Gomang 7 - Sonam Zangpo (Bsod nams bzang po) (1380-1417)

13th Abbot, Paldan Zangpo occupied the throne of Densatil from 1408 to 1417. He is established in 1408 although not yet fully ordained. Like his predecessors, he maintains close relations with Tsongkhapa and the nascent Gelugpa school. [58]  One would probably have to look into the growing dissension within the Lang/Phagmo Drupa clan to find out why none of his two brothers, one occupying the throne of Neudongtse and the other succeeding him on the throne of Densatil, considered commissioning a tashi gomang for him, despite the political and religious prestige of these monuments.

It was not until 1431 that Sonam Gyaltsen pel zangpo began the construction of this stupa which had been completed in 1432. He would have been decorated with more than 5,000 figures [59]. However, as this number is not precise, is not based on any historical text relating to the construction of the monument, and is simply the result of a visitor's impression, it is likely to be exaggerated.

If this tashi gomang was built in 1431-1432, we also know that the eighth was built two years later with the death of Sonam Gyaltsen pel zangpo. This extreme proximity renews the effect of twin stupas that we observed for the fifth and sixth of 1407-1408, and once again makes it delicate to attribute certain groups of objects to one tashi gomang rather than the other. Once again we have therefore chosen to divide the objects according to criteria that may seem subjective, but which for the most part are based on what seems most logical in relation to a stylistic chronology and pictures. Perhaps new information or unpublished objects will one day allow us to refine the proposal for these two tashi gomang stupas.

To try to illustrate the peculiarities of this seventh tashigomang, we will retain the following few works.

The Four Guardian Kings
A Dhritarashṭra kept at the Capital Museum in Beijing has strong similarities with the one we will see on the next tashi gomang, so it should undoubtedly be associated with this seventh monument (Fig.252). Beyond the classic armor, in addition to the shape of the tiara, both have a representation of a dragon on their rounded belly. Insofar as the four lokapalas of the eighth tashi gomang were photographed by P.F. Mele, a Vaishravana can also only find its place on this monument of 1431-1432 (Fig.253).

Fig. 252

Fig. 253

TIER 6

Fig. 254 From this seventh tashi gomang, the ornamentation of the sixth tier undergoes a radical change. While the third, fourth, fifth and sixth monuments were adorned with plaques on which the deities were solidly cast with the backgrounds, the artists, probably in order to face economic and technical constraints, chose to return to the old traditions. On the last two tashi gomang stupas, the deities are cast separately and fixed on ornamental plates, repoussé for the first and cast for the second. At least one of these plates has reached us (Fig.254). As observed on the plate attributed to the first tashi gomang of 1267 (Fig.17), we again have plates that joined behind the deities leaving the ornamental richness of the lotus scrolls in evidence. However, as on the two previous monuments, the secondary deities are no longer surrounded by loops of lotus scrolls either.

An examination of the deities attributable to this tier 6 reveals once again that several workshops collaborated at the same time and that not all the artists were equally talented. This will be evident on all floors of the monument. On the other hand, even more than on the two previous tashi gomang, the artists seem to have been strongly inspired by their predecessors, and essentially by the 1267 model.  The multiplicity of artists is probably linked to a desire to build the monument quickly, since we saw above that it would have lasted less than a year.

In any case, even if many objects attributable to these last two stupas tashi gomang still echo those of the preceding monuments, it would seem that this mixture of diverse influences can be, associated with the development of local artists' ateliers, at the origin of the emergence of a new Tibetan art style.

Fig. 255

Fig. 256

Fig. 257

Fig. 258

A group of four Serpent Kings, all from the same workshop, could come from the ornamentation on one side of the monument. (Fig.255, 256, 257, 258) Their buns are surmounted by nine snake heads, a number that we have so far only seen on the sculptures attributed to the first tashi gomang. Another reference to this first tashi gomang can be seen in the shape of the armbands. The tiara has a central finial in the shape of an anchor with an inlaid flower in the center, but with the central arch wider than previously. This detail will develop on many objects from then on. Likewise, the lotus petals become simpler, rounder and wider in a new style that will also develop from then on.

Fig. 259 Another Nagaraja from another workshop, but still with nine hooded snake heads above the bun, bracelets reminiscent of those of 1267, and a tiara similar to those of the four previous figures must have originated from this seventh tashi gomang (Fig.259). This same workshop seems to have provided a number of other approaching figures for the eighth monument.

An aspect of Lhamo (Prānasādhanā Śrīdevī) should come from the east side (Fig.260). Its high quality of execution certainly makes it one of the most accomplished pieces of this tier 6 and the particular treatment of the lotus pedestal allows us to suppose that the artist was strongly inspired by the one that must have once adorned the same spot on the fourth tashi gomang of 1386. However, the jewels and ornaments are different from those of the 1386 sculptures, and above all, the central final of the diadem and the earrings of this goddess are characteristic of the new style that will be seen on many of the objects of this seventh tashi gomang and some of the eighth.

We can attribute to the southern side, an aspect of Mahakala (Fig.261), a Rahu (Fig.262) and a probable aspect of Atavaka, god of wealth brandishing jewels in his two hands (Fig.263). Their large circular earrings are adorned with eight coloured beads arranged around a central one. The colours of the eight beads are generally alternated on the majority of the works.  Rahu and Atavaka wear armbands with small scrolls at the base of the fleuron, following the tradition that has appeared on certain deities since the fifth and sixth tashi gomang stupas.

Fig. 260

Fig. 261

Fig. 262

Fig. 263

A Jambhala holding a jewel and his mongoose should come from the west side (Fig.264). The treatment of its diadem and lotus pedestal make it possible to attribute it to the same workshop as the four Serpent Kings (Fig.255, 256, 257, 258). An aspect of Mahakala wearing a large mantle must also have originated from this side (Fig.265). He wears large earrings with beads of alternating colors of the same type as those seen on the sculptures on the east and south sides.

Fig. 264

Fig. 265

Fig. 266

Fig. 267

From the north side should come an aspect of Lhamo (Fig.266) and a Druma (Fig.267). These two sculptures clearly come from the same workshop as the Serpent Kings (Fig.255, 256, 257, 258) and the Jambhala (Fig.264). The lower part of Druma's lotus pedestal has an arrangement indicating that the lotus scroll to which it was leaning came to fit underneath to support it.

TIER 5
Two large fragments of this level have come down to us. They are now both in the Guimet Museum in Paris. They each consist of two plates of four offering goddesses with part of the cornice above them still in place (Fig.260 & Fig.261). The corners of this cornice are decorated with triangular extensions with large inlays of coloured beads of a pattern that does not appear in any photograph taken by P.F. Mele. Since the offering goddesses can only come from this seventh tashi gomang, we can consider that this detail is probably specific to this monument. Careful examination of these plaques reveals a significant drop in artistic quality compared to the previous monuments. The goddesses are represented dancing on lotus pedestals of 14 or 16 petals, but their bodies are extremely simplified and their arms are less projected into space. Their jewellery is treated in a simplified manner, without inlays, and their complex buns are no longer represented either. The side plates are simplified and no longer support the usual lotus scrolls. Some of the offering Goddeses carry again a Khatvanga in the hollow of their arm as was the case on the plates attributed to the tashi gomang of 1267 (Figs.22,23,24,25) and on the plate photographed at Tsetang in 1948 (Fig.9).  However, this scepter is misunderstood here since only the upper parts are represented.

Fig. 268

Fig. 269

Some sculptures of female deities should be attributed to this tier. They are all the work of different artists but all present stylistic archaisms at the level of the ornaments associated with the large circular earrings of the type already observed on many of the sculptures of this seventh tashi gomang.

A Parnashavari could be the one that occupied the center of the eastern side (Fig.270). A Marici from another the same workshop was to be at his side (Fig.271). Their style fits in with the new aesthetics of the time with wider and rounded petals on the pedestals. This style is closer to that of the four Nagarajas in tier 6.

Since these figures from the east side are not identifiable in the photos of the eighth tashi gomang, the probability that they may come from the seventh is very high.

Fig. 270

Fig. 271

Fig. 272

Fig. 273

An Usnisasitatapatra could be the one that was supposed to be on the southern side (Fig.272), and a Vasudhara on the northern side (Fig.273).  Both are based on lotus pedestals with rounded and domed petals with ornate borders that fit in with the new trend. They wear jewellery and ornaments richly inlaid with turquoise and other coloured stones.

TIER 4
An important fragment of this tier is known (Fig.274). The presence of numerous Aksobhya on the bottom of the plates makes it possible to consider that this fragment comes from the eastern side of the monument. Like the fragment attributed to the lower tier, this one has retained part of its cornice with a triangular extension on one of the corners. As this detail has not been observed on any other tashi gomang, it seems logical to consider that this fragment comes from this seventh monument. In the center of the largest plate, we can observe in detail the prabhamandala to which a figure of the Buddha must have been leaning. No Buddha sculptures are attributable to this tier. The presence of two horizontal rectangular perforations on the plates makes it possible to consider that if buddhas with tangs arranged horizontally were to appear, whereas they usually seem to have been more vertical, there would be a good chance that they could be those from this seventh tashi gomang.

Fig. 274

Fig. 275

Fig. 276

Fig. 277

At each corner of this tier was a figure of a guardian deity. A Vajrankusa must have been located in the southeast (Fig.275) and a Vajrapasa in the southwest (Fig.276).  Both clearly come from the same workshop and the lotus pedestal, jewelry, necklaces, tiaras and especially earrings are characteristic of what we have seen on other objects attributed to this seventh tashi gomang and the Densatil final style. A bodhisattva holding a lotus, associated with a representation of Hayagriva identifiable by the horse's head at the top of his skull, is most probably the Lokeshvara, which must have stood next to the Buddha in the middle of the west side (Fig.277). The simple row of large single petals of its lotus pedestal fits in with the new trend.

TIER 3
The eastern side of this tier is usually assigned to the Vajradathumandala in 47 deities. An Amoghasiddhi presented today leaning against a prabhamandala that is not his should come from this tier (Fig.278). His earrings are encrusted with large turquoise beads. His lotus pedestal shows rounded and domed petals with ornate borders that fit in with the Densatil final style. We can also attribute to this tier a representation of Buddhakapala which was to occupy the center of the south side (Fig.279). This sculpture is the creation of a workshop that once again seems to have been inspired by ancient models, which is particularly visible in the treatment of the goddess' bone apron. The treatment of the lotus pedestal is to be compared to that of the four serpent kings of tier 6. A representation of Manjushri sitting on a lion, perfectly corresponding to his Vadisimha aspect which is the one appearing in the texts describing the monument visualized by Jigten Gonpo in 1198, must have come from the south side of this level (Fig.280). Its ornaments and lotus pedestal are fully integrated into the stylistics of the period, of which it can be considered one of the most representative objects.

Fig. 278

Fig. 279

Fig. 280

A standing Acala had to be positioned on the west side (Fig.281). It presents strong similarities with the two guardians observed in tier 4 and another kneeling Acala that must have been at one of the four corners of this floor (Fig.282), which allow us to attribute them all to the Densatil final style. Another Acala, with a slightly different lotus pedestal but still in the new style of the period, should also come from this level. (Fig.283). The unusual shape of its armbands is similar to those of the Jnanadakini that we will find in tier 2.

Fig. 281

Fig. 282

Fig. 283

TIER 2
As on the previous tashi gomang stupas, the eastern side of this second tier was logically intended to illustrate the 72 deities mandala of Guhyasamaja. Three sculptures representing Guhyasamaja (Fig.284), Amitabha (Fig.285) and another bodhisattva from the retinue, difficult to identify on photographs (Fig.286) can be attributed to this side. All three of them rest on lotus pedestals with rounded petals with ornate edges arranged on stepped plinths reminiscent of older images. All three wear the circular earrings encrusted with large beads of the type that seems most common on this tashi gomang. They are a good expression of the final Densatil style. A Jnanadakini on a lion should come from the west side (Fig.287). The unusual treatment of his armbands is similar to that of an Acala observed in the lower tier (Fig.283). Its lotus pedestal is characteristic of the new emerging style.

Fig. 284

Fig. 285

Fig. 286

Fig. 287

TIER 1
At least three Vajravarahi figures should come from the rim of this tier 1 and illustrate the variety of workshops that worked on the monument. The first two are of a type never before encountered on a tashi gomang (Fig.288). They should come from the same workshop as the Buddhakapala of Tier 2 and the 4 Serpent Kings of Tier 6. The second (Fig.289), of a more common type, is more in line with the previous models but with a treatment of the scarf that is more summary and less well interpreted than before.

Fig. 288

Fig. 289

Fig. 290

In the center of the east and south sides were to be placed representations of Vajradhara. One, derived from the previous models, was originally intended to be placed on one of these sides (fig.290).

Five representations of mahasiddhas from the western and northern sides have come down to us. From the west side, a Dombi Heruka on his tiger (Fig.291), a Kukkuripa and his dog (Fig.292) and a Saraha (Fig.293).  From the North side, a Jalandhara (Fig.294) and a Vajragantha (Fig.295). They all come from the same workshop, four rest on lotus pedestals with rounded ornate-edged petals, and wear armbands of the ancient type seen on the Serpent Kings attributed to the Tashi Gomang of 1267.

Fig. 291

Fig. 292

Fig. 293

Fig. 294

Fig. 295

PILLARS
The pillars attributable to this seventh tashi gomang seem to want to present an innovative aspect by cumulating various elements of the previous ones. The bases of the vases are treated in the shape of a blooming lotus flower resting on a square base, echoing the one that emerges to support the two leaning bodhisattvas. The deities support a lotus with a double row of petals from which the capital of square section seems to emerge (Fig.296).

Fig. 296

We cannot, to this day, attribute any corner pillars to this monument.

Abbot 14 / Stupa 14/ Tashi Gomang 8 - Sonam Gyaltsen pel zangpo (1386-1434)

14th Abbot, Sonam Gyaltsen palzangpo occupied the throne of the monastery from 1417 to 1434. Previously he was recognized at a very young age as a great scholar who was able to exhibit the Hevajra Tantra at assemblies from the age of ten. [60] His fame would have made him very influential at the court of Neudongtse. [61] It seems he was considered a living Buddha. [62] After the death of Dragpa Gyaltsen in 1432, his authority was important in convincing his clan not to place his own brother Sangye Gyaltsen on the Neudongtse’s throne, but rather his nephew Dragpa Jungne. As the last great unifying authority of the Lang/Phagmo Drupa clan, his death in 1434 was to lead to a conflict between the various factions that had been maneuvering for power for some time. After this date, the monastery of Densatil followed the slow fall of the Lang/Phagmo Drupa clan and its artistic production continued to decline.

Fig. 297 A tashi gomang for Sonam Gyaltsen palzangpo is attested in various sources. It would have been ordered in 1434 by his nephew who was then occupying the throne of Neudongtse thanks to him. The Blue Annals report that Dragpa Jungne would have organized a pompous funeral for him even if the period was not favourable, and that he would have erected a silver stupa in the shape of a tashi gomang. [63] This description reported by Gö Lotsawa (1392-1481), born in Tsetang and a contemporary of the event, raises the question of whether this tashi gomang could have been plated with silver, or whether, as mentioned above, only the kadampa-type stupa at the top of the monument could have been cast in this precious material. According to Chos kyi rgya mtsho, this tashi gomang would have housed 3,962 figures, again indicating an iconographic variation from the others. [64] According to the same source, a golden copper Kalachakra mandala would have been placed above, as we have already noted for other tashi gomang stupas. [65]

The stylistic analysis of the tashi gomang stupas visible in P.F. Mele's photographs and of the groups of objects attributable to all of them makes it possible to propose that this eighth tashi gomang may be the one from which we possess several views essentially depicting the four guardian kings and portions of the east sides of the fifth, fourth and third tiers (Fig.297).

The upper corner of the photograph shows a white area that may correspond to a kind of drapery that obscures the top of the stupa. In the lower part of the picture, a wooden barrier veneered against the big lotus separating thirds 6 and 5 prevents the details from being seen.

To try to illustrate the peculiarities of this eighth tashigomang, we will retain the following few works.

Fig. 298 GUARDIAN KINGS
The four guardians visible in P.F Mele's photographs are as they should have been on the tashi gomang of 1267 placed on the circle constituting the open lotus supporting the monument. In a photograph by P.F. Mele, one can see that, at least in front of the lokapalas, he married a shape evolving in arcs of circles which results in an identical movement of the frieze of lotus petals. (Fig. 298).

The Dhritarashtra and Virudhaka were arranged on the left side of the great lotus from the observer's point of view. (Fig.299). Examination of Virudhaka shows that it must have been executed by a different artist than the one who had to work on the other three. The Virupaksha and Vaishravana on the right were recovered. The first was photographed by Michael Henss at the Jokhang in Lhasa (Fig.230) and the second is today at the Capital Museum in Beijing (Fig.301).

Fig. 299

Fig. 300

Fig. 301

TIER 6
As mentioned above, this tier does not appear in P.F. Mele's photographs because it was hidden behind a wooden railing. The shadows of this railing on the petals of the great lotus should be taken as a good indication of their extreme proximity. If we consider in addition that this large lotus was a priori not totally circular, but instead was structured in a succession of arcs of circles (Fig.298), it is very unlikely that the photograph that we had considered as representing a quarter of the eastern side of tier 6 of the first tashi gomang (Fig.12) could in fact have been taken on the eighth.

Once again, an examination of all the objects that could be attributed to this sixth tier shows that several artists or workshops worked there at the same time.

Fig. 302

Fig. 303

Fig. 304

Fig. 305

Fig. 306

Even though their overall design has a more modern character, two Serpent Kings are executed in a style that seeks to resemble those attributed to the tashi gomang of 1267 (Fig.302 and Fig.303). This is particularly noticeable on ornaments such as diadems, armbands and bracelets. However, on this eighth monument, there are only five snake heads on the top of their buns. Three others (Fig.304, Fig.305, Fig.306) are executed in a different style and must be the work of artists from the same workshop as one attributed to the previous tashi gomang (Fig.259). The unity of this last group is clearly visible in the decoration of the dhotis, and the jewels. The caps of snake heads usually placed above their buns have all disappeared. Only the holes intended to fix them are still visible, proving that they are indeed Nagarajas.

Fig. 307

Fig. 308

Fig. 309

Fig. 310

From the east side, should come an aspect of Lhamo (Fig.307). A Rahu should come from the south side (Fig.308). A group of sculptures undoubtedly produced by the same workshop comes from the west and north sides.  From the west, a Mahakala (Fig.309), and from the north, a Druma (Fig.310), a Mahakala (Fig.311), a Lhamo (Fig.312) and an Aparajita (Fig.313). Their ornaments, diadems, earrings, armbands and bracelets combine patterns observed on previous tashi gomang stupas. The earrings with large colored beads seem less numerous than on the previous tashi gomang, giving way to the classic model of the blooming flower with a colored stone in the middle.

Fig. 311

Fig. 312

Fig. 313

TIER 5
This tier of female deities is partially visible in P.F. Mele's photographs. The offering goddess plates continue to decline. They seem to be made in the technique of embossed metal, making almost all relief disappear (Fig.314).  The center of the eastern side which should have sheltered Parnashavari, Janguli and Marici shows here only a Vairocana which should have been in tier 3 and a bodhisattva which is difficult to identify precisely but which also has no reason to be at this location (Fig.315). These two sculptures seem to belong to styles prior to the construction of this eighth tashi gomang and could demonstrate that the works adorning this monument may have been taken from older ones or from altars in the sanctuary.

Fig. 314

Fig. 315

TIER 4
What can be observed of this third tier in the 1948 photographs confirms the lack of iconographic coherence of the objects placed on this tashi gomang. Indeed, the side visible in the photograph should show the Buddha in Bhumisparça mudra in the center surrounded by four representations of Aksobhya. Instead, we see in the center a Buddha without pedestal lotus and of a format clearly unsuitable for this location making the gesture of teaching (Fig.316) surrounded by representations of Amitabha (Fig.317 and Fig.318) that should have been placed on the west side, i.e. the opposite side of the monument. An Aksobhya kept in a private collection and executed in exactly the same style as these Amitabha should logically have been on this east side. (Fig.318_B). This series of representations of Amitabha and this Aksobhya are of fine quality, and their lotus pedestals are in a style found on a number of fifteenth-century sculptures often attributed to an artist named Sonam Gyaltsen, who will be mentioned later.

Fig. 316

Fig. 317

Fig. 318

Fig. 318_B

It is impossible to tell from the photographs whether the background plates with miniature buddhas are embossed or cast. However, if the offering goddesses in Tier 5 were made in repoussé, there is a very good chance that this would also be the case in Tier 4.

A Vajrankusha (Fig.319) should have been placed at the southeast corner. Its lotus pedestal is in line with those observed on the previous tashi gomang and its ornaments are also characteristic of the superimposition of archaisms that marked Densatil's objects in the fifteenth century. Note that on the guardian figures attributed to the seventh tashi gomang (Figs. 281, 282), the earrings were richly decorated with coloured beads. On this Vajrankusha, as on many other sculptures attributable to this eighth monument, the earrings are in the simple shape of a blooming lotus with a single stone in its center. Another guardian deity having lost the attributes that would allow to identify him with certainty would have to come from this tier and rests on a base quite characteristic of the new emerging style (Fig.320).

Fig. 319

Fig. 320

TIER 3
P.F.Mele's photographs allow us to see that the iconographic disorder observed on the lower floors is repeated on the eastern face of this third tier. Instead of the Vairocana that should have occupied the center of the Vajradathu mandala, there is an Amoghasiddhi, and on the left, at the logical location of an Aksobhya, a deity with a raised arm could be a Manjushri (Fig.321). On either side of the Amoghasiddhi, we can observe two standing bodhisattvas, who are probably two of the eight normally accompanied by a fierce deity seated near a vase that should frame the buddhas in the center of each face of the fourth tier.

Fig. 321

Three of these groups could be from this tashi gomang. A Vajrapani accompanied by Acala must have been located on the south side (Fig.322). The other two bodhisattvas are more difficult to identify precisely (Fig.323 & Fig.324).

Fig. 322

Fig. 323

Fig. 324

The major and innovative stylistic element of this group of sculptures is the treatment of the lotus pedestals, which are composed of two circles joined together. In addition to stylistic evidence, as this succession of circles is found on the large lotus separating the sixth and fifth levels of the eighth tashi gomang, it seems logical that these three bodhisattvas should be attributed to it.

A Buddhakapala should have been placed centrally on the southern side of this third tier (Fig.325). It is very similar to the one attributed to the same location on the previous tashi gomang. The Citrasena bone apron is once again of the "garland" type as on the first tashi gomang stupas, but unlike the previous one, an ornament with a coloured stone hangs in the center of the curves.  The pedestal lotus seems to seek a synthesis between the Densatil final style and the new one. A Manjushri Vadisimha should come from the south side of this tier (Fig.326) and is resting on a lotus pedestal of the new style of the period.

Fig. 325

Fig. 326

Three Acala by three different artists but with characteristic lotus pedestals and jewels could come from three of the four corners of this tier (Fig.327, 328 & 329). The lotus pedestal of the last one has a particular shape with an angle at the back, probably for a better support at one of the corners of the tashi gomang.

Fig. 327

Fig. 328

Fig. 329

Fig. 330

A Chakrasamvara of very similar construction to that of the Buddhakapala on the south side should be the one that was in the center of the north side (Fig.330). Their lotus pedestals and ornaments are almost identical, with the same ornament decorated with stones on the bone apron of the goddess. The lotus pedestal with bulging petals with ornate edges is typical of the Densatil final style.

TIER 2
The eastern side of this second tier was logically intended to illustrate the 72 deities mandala of Guhyasamaja. Two of the five main deities have come down to us (Fig.331 - Fig.332). They obviously come from the same workshop as those attributed to the seventh tashi gomang (Fig 284, 285, 286). The essential differences are ornaments encrusted with turquoise beads placed at the corners of the stepped plinth supporting the lotus pedestal, the armbands which here take up the initial model of the serpent kings from the tashi gomang of 1267, earrings simply adorned with a single bead in the heart of a blooming lotus. A scarf also surrounds these deities. They mark a less richly ornamented evolution of the final Densatil style

Fig. 331

Fig. 332

Fig. 333

A Mahavarabhairava should come from the southern side (Fig.333).

TIER 1
On the periphery of this tier, there must have been Vajravarahi figures whose quality of execution continues to degenerate. The scarves are once again misunderstood, demonstrating that the artists devoted to these representations were not the most gifted. A few, however, stand out from the rest and illustrate once again the variety of artists who collaborated on this set (Fig.334 and Fig.335). They clearly seek to approximate the archetypes attributed to the tashi gomang of 1267 and as these are based on pedestal lotus with two rows of petals. This detail could be a clue that the two monuments were placed side by side.

Fig. 334

Fig. 335

Fig. 336

A Chakrasamvara with one pair of arms with all the stylistic characteristics of Buddhakapala and a Chakrasamvara with six pairs of arms of the third level should therefore be the one formerly placed at the top of this tashi gomang (Fig.336).

From the west side must come a Kukkuripa and his dog (Fig.337) and a Saraha (Fig.338).  Both are almost identical to the same ones attributed to the previous tashi gomang but nevertheless show a weakening of the quality of execution.

Fig. 337

Fig. 338

PILLARS

Fig. 339 The shape of the pillars of this tashi gomang is known to us from the photographs of P.F. Mele (Fig.339).

It can be seen that the pillars still follow the iconography of the bodhisattva standing on a vase supporting the capital. However, here they are only represented by an embossed sheet of metal, plated on the wooden structure of the monument. 

The cornices of the fragments of the seventh tashi gomang (Figs.260, 261 and 266), were cast in copper alloy. On this eighth monument, these cornices are like pillars, simply made up of wooden structures on which embossed metal sheets are nailed. There is also a lack of consistency in tracking the decoration of overlapping plates.

An examination of the works related to this Densatil's last great tashi gomang has enabled us to observe a form of decline, whether in terms of the iconographic rigour associated with this type of monument, or more simply on the artistic level. This is undoubtedly linked to the fact that the death of Sonam Gyaltsen in 1434 led to a conflict over the throne of Neudongtse, between Dragpa jungne which had been imposed there in 1432 by Sonam Gyaltsen, and his father who had been chosen for it and then aspired to take it.

This "anarchy" of 1434 led to the decline of the hegemony of the Lang/Phagmo Drupa clan and internal struggles did not allow the appointment of a new abbot for Densatil until 1454. We can therefore consider that this eighth tashi gomang of Densatil was ordered by a sovereign in turmoil, for a monastery without a religious authority powerful enough to ensure its perfect realization.

It is therefore conceivable that the construction of this eighth tashi gomang could have been spread over several years. The difficulties of the time may also explain the concept of filling the tiers with works unrelated to the original iconographic plan in order to give an impression of faster completion.

Taking advantage of the chaos, the powerful servant of the Neudongtse court, Norbu Zangpo (Nor bu bzang po) (1403-1466) of the Rinpung (Rin spungs) clan linked to the Lang /Phagmo Drupa through numerous cross marriages seized the opportunity to increase his political influence.

From then on, the Lang /Phagmo Drupa try to maintain their power in the face of the Rinpungs who, while officially respecting them, try to make them their puppets, sure of the military force that gives them the advantage.

At the death of Dragpa jungne in 1444/1445 the political crisis is such that the throne of Neudong remains vacant until the summer of 1448 and the the establishment of Kunga Legpa (Kun dga' legs pa - 1433-1482/83) who is forced to marry a woman of the Rinpung clan.

In 1454, Ngagi Wangpo (Ngag gi dbang po - 1439-1491) was sent to occupy the throne of Densatil. However, Kunga Legpa who already occupied the throne of Neudongtse, wanted to strengthen the power of the Lang/Phagmo Drupa clan by relying on the aura of the monastery. He pushes Ngagi Wangpo to resign in order to occupy both thrones at the same time.

This is unfortunately not enough, and slowly but surely, thanks to alliances and successions, the Rinpung manage to capture power, that of the Lang/Phagmo Drupa being almost only nominal or symbolic.

In 1473, Ngagi Wangpo who is the favourite of the Rinpung Clan is put back on Densatil's throne. In 1480-81, the Rinpung also made him accede to the throne of Neudongtse.

Kunga Legpa died in 1483. It is thus for him, the sixteenth abbot, that the fifteenth commemorative stupa of Densatil must have been erected around 1483.

Abbot 16 / Stupa 15. Gong-ma Kung Legpa (kun dga’ legs pa) (1433-1482-83)

Kunga Legpa will have occupied the throne of Densatil from 1458 to 1473.  At his death, having also occupied the throne of Neudongtse from 1448 to 1481 Ngagi Wangpo (who succeeded him on both thrones, certainly did not fail to have a memorial stupa erected for him in Densatil.

We do not have any photographs that could correspond to such a monument, nor does any description seem to have ever been made of it.

A small group of objects could, however, give us some idea of this by again using the concept of the evidence of a fragment of art history as a historical source.

A group of three Guardian Kings photographed by Ulrich Von Schroeder at the Tsuglakang of Lhasa's Jokhang presents strong similarities with those observed on Densatil's seventh and eighth tashi gomang stupas (Fig.340 Fig.341 Fig.342). Moreover, the treatment of jewellery and lotus pedestals suggests that they belong to the same artistic tradition as many of the objects attributed to the eighth and considered characteristic of the final Densatil style. The essential difference lies in their height, which is roughly half of the others. They could be a clue to the existence of a ninth but smaller tashi gomang that may have been located in the small square extension erected northwest of Densatil's Tsuglakang. It could also be the tashi gomang that Chokyi Gyatso says he observed on the first floor of the Tsuglakang, [66] and which in fact would necessarily have been small in size.

Fig. 340

Fig. 341

Fig. 342

A Vajrapani (Fig.343) and an Acala (Fig.344) which could have been placed in tier 3 were also photographed in the Tsuglakang of Jokhang by Ulrich Von Schroeder. They are also significantly smaller than all those observed on previous tashi gomang and could therefore have come from a smaller monument. The lotus pedestals, jewellery and scarves still evoke the sculptures of the final Densatil style.

Fig. 343

Fig. 344

Even if their dimensions are about the same, a group of offering goddesses plates summarily treated in repoussé in a decadent style but clearly in line with what we observed on the eight tashi gomang stupas could have adorned such a monument. The iconography no longer follows the archetypes, and the side plates are no longer decorated with lotus scrolls, but with other goddesses (Fig.345). Most of them are dressed with bone aprons which was never the case before. These aprons are no longer in the ancient 'garland' style but are a development of what we had observed on the Vajrayogini figures of the fifth and sixth monuments (Fig.346). All these goddesses rest on a large lotus pedestal with a single row of seven to eleven petals.

Fig. 345

Fig. 346

A standing bodhisattva accompanied by a fierce deity kneeling near a vase worked in repoussé was photographed in situ in the ruins of Densatil by Michael Henss (Fig.347). It must have once been plated on a tashi gomang. Since we have seen that on the eighth and last large monument, these deities were cast but positioned in tier 3 instead of tier 4, a repoussé version could therefore only come from a posterior tashi gomang of lesser artistic quality. The simple row of large petals that constitutes the lotus pedestal of this small group is to be compared to the one observable on the offering goddess plates that we have just mentioned. There is therefore every reason to believe that these fragments may have come from the same monument.  This petal shape can also be related to the new style that appeared on the last two tashi gomang stupas.

Fig. 347

Fig. 348

Fig. 349

Two plates also treated in repoussé representing a Jambhala (Fig.348) and a Mahakala (Fig.349) could be the type of work that made up the sixth tier of such a monument. The Mahakala has been cut out around its silhouette, but the Jambhala is leaning against a background with a generally circular shape, which reminds us of the shape of the plates of the sixth tier of the great tashi gomang stupas.

Such a monument should be considered as the end of Densatil's art. It could be that it was through this command of Ngagi Wangpo that the Lang/Phagmo Drupa clan made their last efforts to try to regain their lost greatness.

Abbot 15 / Stupa 16. Ngag-gi dbang-po (1439-1491)

The anarchy of 1434 had left Densatil without an abbot from 1434 to 1454. From 1454 to 1458, Ngagi Wangpo occupied the throne of the monastery before leaving his place to return from 1473 to 1491. He also occupied the throne of Neudongtse from 1481 to 1491.

Sensing his impending death, Ngagi Wangpo organized the protection of his clan and the monastery of Densatil, foreseeing that after him, the monastery would come under the governance of the fourth Zhva dmar sprul sku Chos graggs ye shes (1454-1524). The latter was at the time one of the most influential figures in Tibet with an ascendancy over the Rinpungpa.

A commemorative stupa had to be erected for Ngagi Wangpo in Densatil since we know from multiple sources that there would have been 18 of them. Unfortunately, no information has reached us about this stupa.

Abbot 17 / Stupa 17. 4th Zhva dmar Chos graggs ye shes (1453-1524)

He occupied the throne of Densatil from 1493 to 1524. Unfortunately, we know nothing of the possible commemorative stupa that may have been built there for him.

Abbé 18 / Stupa 18. Spyang nga Grags pa ‘byung gnas (1508-1570)

Grandson of Ngagi Wangpo, he was abbot of Densatil probably from 1524 or a little later, until 1570. He was probably enthroned at the death of the Zhva dmar sprul sku and then remained on the abbot's throne. Densatil's eighteenth commemorative stupa was therefore most likely made for him in 1570.  However, the state of decay of the Lang/Phagmo drupa clan certainly did not allow anything magnificent to be achieved.

Thus logically ends the long history of Densatil's eighteen stupas that began just four hundred years earlier with the death of Dorje Gyalpo Phagmo Drupa.

The fifth, sixth and seventh zhva dmar would then have served as abbots at Densatil, but no source has yet shed any light whatsoever on their possible commemorative stupas in the monastery.

Divided, the Lang/Phagmo Drupa clan finally disappeared for good in the second half of the seventeenth century after participating in the advent of the fifth Dalai Lama. The latter is said to have acquired religious objects which once belonged to the Lang/Phagmo Drupa and which were of religious, historical and political importance to him [67]. It is therefore possible that on this occasion some objects may have been removed from Densatil.

Proposed location of the eight tashi gomang stupas in Densatil's gtsug lhag khang

If Densatil's great Tashi gomang stupas are essentially known to us from P.F. Mele's photographs, it is because they were arranged in the tsuglakang around a rectangular zenithal opening allowing daylight to penetrate. If our stylistic attributions are correct, we can consider that P.F. Mele took photographs of the tashi gomang stupas 1, 2, 3 and 8 and that the 4 and 5 appear in the background on two views, the 5 being in a strong dark area. We can envisage that if 6 and 7 do not appear anywhere, it is because they were not sufficiently exposed to the light to allow a picture to be taken.

Fig. 350 Professor Tucci's team reportedly left for Densatil in the morning.[68] This meant that the photographs had to be taken in the late morning or afternoon, when the west side of the sanctuary was in shadow and only the monuments to the north, south and east could benefit from natural light. We mentioned above the two east-west stupas "covered with gold", probably the tashi gomang of 1267 and the stupa of 1280, which had to be leaned against the north wall. The third and fourth tashi gomang are clearly visible, leaning against the east wall, on either side of the central altar supporting the five Kadampa-style stupas. Considering the location of these stupas, the second tashi gomang that appears in several photographs could only be found leaning against the south wall, probably separated from the hut of Dorje Gyalpo Phagmo Drupa by a span of architectural pillars in which one can see the prabha to which a great deity facing the 1280 stupa must have been leaning. The second tashi gomang erected in 1360 must have been opposite the first one erected in 1267. Considering that this symmetry of the first two tashi gomang could be original, it is possible to envisage that part of the work ordered by Changchub Gyaltsen after 1360 may have concerned a reorganisation of the eastern wall and of the western part. Next to the second tashi gomang, in the shadow is probably the fifth. The eighth tashi gomang has good lighting, which would tend to make it situate against the north wall, probably next to the first, thus facing the fifth. This positioning could explain why the lokapalas on this stupa were placed on the lotiform circle, as they had to be on the first one as well. The sixth and seventh tashi gomang stupas that could not be photographed due to lack of light find a natural position leaning against the west wall, as a logical continuation of the fifth. (Fig. 350)

Other treasures of Densatil and attempt to define a Densatil style

In addition to the extraordinary tashi gomang stupas, P.F. Mele's photographs provided us with a view of almost the entire eastern wall of the tsuklakang (Fig. 1) with, in the centre, in front of the five kadampa-type stupas, large sculptures representing the lineage of the masters and two large peacocks a priori in embossed copper sheets. Shelves support portraits of lamas above bodhisattvas executed in the style derived from Indian Pala style. On the left of the photograph one can observe two of the five large trilobed lotus petals arranged in front of the tashi gomang of 1370, representing the five jinas.  Four of them and one side of the fifth are best seen in the blurred photograph by P.F. Mele (Fig. 351).

Fig. 351

Fig. 352

Fig. 353

After the tashi gomang stupas, these five petals are undoubtedly among the greatest artistic achievements of the monastery. The large size of these large petals, combined with the quality of the details and the depth of the embossing, unequivocally place them among the most remarkable examples of the art of repoussé in Tibet.  They are probably the work of great Nepalese artists who are masters in this art.
The Vairocana is today kept in a private collection (Fig. 352 & Fig. 353).

Examination of the jewellery and other visible details on this petal suggest a dating circa 1360, during the reign of Changchub Gyaltsen. The Tibetan historical literature is unanimous on the fact that Changchub Gyaltsen, who came to power over the whole of Tibet in 1354, was at the origin of Densatil's second tashi gomang erected in 1360 and that he also commissioned major redevelopment work on the Tsuklakang. We can therefore propose that this remarkable set of five petals may originally have been part of an arrangement of the east wall. The organization of the works on this wall was necessarily strongly modified between 1370 and 1390 with the installation of the third and fourth tashi gomang on either side of the five kadampa-style stupas. It is therefore likely that this ensemble was preserved close to its original location, but simply placed in front of the tashi gomang of Dragpa Sherab who was abbot from 1361 to 1370, possibly during the period of the redevelopment ordered by Changchub Gyaltsen.

Professor Tucci mentions a large statue of Buddha from which he was able to note the inscription that would have mentioned Changchub Gyaltsen as the commissioner of this work. The inscription would also have provided the names of the artists and indicated that the statue had been carved according to the models of Nepalese art. On both sides of the Buddha were statues of bodhisattvas. [69] Such an object had not been photographed at the time, and it has not yet been found.

Chinese scholars have recently published a large statue of the Buddha found in the ruins of Densatil. The inscription indicates the name rgyal name mtshan dpal bzang po, which may designate several persons in the context of Lang/Phagmo Drupa but is probably the fourteenth abbot of Densatil, especially since the inscription would date the work to 1433. (Fig. 354 & Fig. 355).

Fig. 354

Fig. 355

Fig. 356

Other sculptures not from the tashi gomang stupas, exhumed from the ruins of the monastery and now kept in the Lhasa Museum or the Capital Museum in Beijing, have also been published in China. Although executed in different periods, they all reflect a common style derived from Nepalese art with remarkable ornaments inlaid with stones of various colours.  Among these, we will retain a Manjushri from the Lhasa museum (Fig.356) who wears a type of armbands that seems to appear on the twin tashi gomang of 1407-1408, which should thus make it possible to propose for him a dating on the end of the fourteenth and the beginning of the fifteenth century.

Despite some small differences, a fourteenth century Vairocana (Fig. 357) exhumed from Densatil's ruins can be compared with another Vairocana today at the Rietberg Museum in Zurich (Fig. 358). This suggests that the latter may also come from the monastery. A Manjushri, also today preserved in the Rietberg Museum (Fig. 359), and a vajrasattva from the Tibet Museum in New Delhi (Fig.360) also displays all the stylistic elements characteristic of the sculptures observed on the tashi gomang stupas in their jewellery and ornaments.

Fig. 357

Fig. 358

Fig. 359

Fig. 360

These few examples among many others show that the artists who worked on the ornamentation of the tashi gomang stupas also produced a lot of works that must have adorned altars in Densatil, and probably elsewhere too. It can therefore be considered that this particular style, which will always remain attached to Densatil, is in fact the most perfect expression of a court art linked to the Lang/Phagmo Drupa dynasty, which in the fourteenth and fifteenth centuries ruled Tibet from its palace in Neudongtse.

This “Densatil style” finds a remarkable expression in an important series of representations of the crowned Buddha. These Buddhas themselves are of a classical Nepalese style with great unity. The crowns are all decorated with five finials and end on each side of the head with two blooming lotus flowers from which are hung scarves ending with pendants reminiscent of those that were placed at the ends of the scarves of the deities on the tashi gomang stupas. Finials and ornaments are usually richly inlaid with coloured semi-precious stones. The analysis of the variations in the treatment of these crowns shows that they follow very closely the evolution of the jewels observed on the deities of the tashi gomang stupas. Most of them seem to have been produced during the golden age of the Lang/Phagmo Drupa dynasty, which was the second half of the fourteenth century and the beginning of the fifteenth. (Fig.361, 362, 363)
At least one of these numerous Buddhas would have been found in the ruins of Densatil and should date from the first third of the fifteenth century (Fig.364). One made of silver (Fig.365) has the headband of its crown arranged to fix a row of fine pearls that has now disappeared. This type of ornament seems to have been common on sculptures from the tashi gomang stupas during the last third of the fourteenth century.

Fig. 361

Fig. 362

Fig. 363

Fig. 364

Fig. 365

Still in silver, a small Tara (Fig.366) with characteristic jewels and presenting a kind of textile loop that we have already observed on two goddesses of the fifth tier of the tashi gomang of 1370 (Fig.103 and Fig.105) constitutes a perfect example defining the Densatil style. It was adapted, probably at a more recent date, on a gilded pedestal in the style of those supporting images of Chinese origin at the beginning of the fifteenth century.

Fig. 366

Fig. 367

Fig. 368

Fig. 369

Three other small representations of Tara are also perfect examples to define the Densatil style (Fig.367 - Fig.368 - Fig.369). All three wear five-shell buns, each surmounted by an inclusion of coloured stones as seen on the offering goddesses of the tier 5 of the tashi gomang stupas as well as jewellery and dhotis also richly inlaid. This profusion of inlays of semi-precious stones on the dhotis to represent the flowers that are usually engraved there is undoubtedly one of the best markers of the great Densatil style. Note that the third is surrounded by a scarf that forms three arches above the head and shoulders, as frequently seen in the sculptures attributed to the tashi gomang of 1386 and also in those of 1407-1408.

Another of the most perfect examples of this Densatil style is undoubtedly a Chakrasamvara embracing Vajrayogini of too modest dimension to have been able to decorate the top of a tashi gomang (Fig.370). The particular style of the bone apron of the goddess and the specific movement of the richly decorated scarf that surrounds the god are very close to the Chakrasmavara and Vajravarahi figures of the Tier 1 of the tashi gomang of 1370. Spaces in the hollow of the scarf, the god's armbands, and the loops of the Vajrayogini bone apron were once filled with rows of fine pearls that have now disappeared.

A Vajradhara (Fig.371) could also be considered a good example of the Densatil style around 1430-1435. Indeed, its earrings and armbands are very similar to those found on the objects attributed to the seventh and eighth tashi gomang stupas.

Fig. 370

Fig. 371

Fig. 372

Fig. 373

Many portraits of monks scattered throughout the collections seem totally characteristic of the Densatil style, without it being possible to determine whether they come from a tashi gomang or not. Thus, a presumed portrait of Milarepa without lotus pedestal (Fig.372) presents all the characteristics of the style, and the quality of gilding that allow us to envisage that it could come from the first tier of one of the tashi gomang stupas of the fourteenth century, probably that of 1370. One llama portrait among many others is resting on a lotus pedestal characteristic of the style (Fig.373) but probably does not come from a tashi gomang. Like most of these portraits, for lack of inscription, it remains unidentifiable with certainty.

Fig. 374 A Vajravarahi kept in the Tibet Museum in Lhasa is a fine example of the expression of the Densatil style (Fig.374). At first glance, it evokes those that adorned the first tier of the tashi gomang of 1386 or those of 1408-1410. However, a careful examination of the treatment of the scarf and the ornaments, would incite to situate it later in time. Indeed, the style of the bone apron still marks an evolution from those observed on the deities of the first two tashi gomang of the fifteenth century.  The rows of pearls here are punctuated by more square pieces at the intersections and the shape of this apron now tends towards that of those worn by lamas. This form seems to appear as much in sculpture as in painting and as a ritual object in the fifteenth century. It should be noted, however, that the artists who worked on the seventh and eighth tashi gomang stupas did not use this new form, but returned to the "garland" type observed on the first four monuments.  This impressive Vajravarahi should therefore not predate the second quarter of the fifteenth century, and possibly much later. This type of sculpture is a good indicator of the recurring problem in Tibetan art posed by the many more or less recent replicas of older originals.

A number of repoussé objects with very strong stylistic similarities can logically be attributed to Densatil's style, and probably to the monastery. Two Nagarajas present ornaments identical to those found on the objects of the tashi gomang stupas of the fourteenth century (Fig.375, Fig.376). A prabhamandala presents around the space on which the deity was leaning a decoration that evokes the one found on the plates of the rank of the Buddhas on the fourteen century tashi gomang stupas (Fig.377). The ornaments surrounding the arcature are inlaid with red stones. A row of blue circular inlays decorate the border. We find the same blue pearls on the ornaments of one of the mahasiddhas of the tier 1 of the tashi gomang of 1370 (Fig.141). It would therefore be logical to date this prabhamandala from approximately the same period.

Fig. 375

Fig. 376

Fig. 377

The Densatil style, an expression of the court art of the Lang/Phagmo Drupa dynasty, should therefore be able to be defined both stylistically and chronologically in relation to the sculptures that decorate the eight stupas tashi gomang, and thus know its term around 1435-1450.

The photographs on the eastern side of the fourth tier of the eighth tashi gomang of 1434-1435 (Figs.315, 316, 317) show us a fine series of representations of Amitabha resting on lotus pedestals of a style as yet unobserved in Densatil's stylistic chronology. These lotus pedestals can be compared to those associated with a workshop of which at least one artist's name, Sonam Gyaltsen, has come down to us thanks to the inscription engraved on a large Avalokiteshvara (Fig.378 & Fig.379).

Fig. 378

Fig. 379

No additional information has so far shed more light on this artist. It is true that Sonam Gyaltsen is a name that was given to a large number of monks. The most famous for the period that interests us here is the fourteenth abbot of Densatil, sponsor of the seventh tashi gomang, and recipient of the eighth after his death. However, even though many great Tibetan abbots have remained famous in history for having also been talented artists, there is no evidence to support such a hypothesis.

We know that the Avalokitshvara figure was made for Zhonnu Gyalchog at the Byams chen chos sde monastery near Shigatse (founded in 1427) at the request of two brothers, Norzang and Palzang. These details make it possible to determine without any doubt that the Norzang mentioned is none other than Norbu Zangpo (Nor bu bzang po -1403-1466) of the Rinpungpa clan. 

We have seen above that the Rinpung clan was an ally and vassal of the Lang/Phagmo Drupa and that Norbu Zangpo, already very influential at the court of Neudongtse, took advantage of the chaos of 1434 to further increase his influence and become the true strongman of Tibet. Without the Lang/Phagmo Drupa being effectively removed from their situation, the Rinpungpa were the true masters, first of their Tsang area and then of the whole Tibet at least until the beginning of the sixteenth century.

The particular style of this great Avalokiteshvara is certainly associated with the workshop of which Sonam Gyaltsen was undoubtedly one of the masters, but also with the region of the Tsang (gtsang) which was the stronghold of the Rinpungpa rulers.

However, examination of the set of objects that we could hastily attribute to the studio in which Sonam Gyaltsen worked clearly shows that it is more likely a style that was followed by several different artists and studios at about the same time.
We can therefore probably conclude that if some of the sculptures adorning the last two tashi gomang stupas of Densatil erected around 1432-1435 can in some respects be attributed to this new school of sculpture, it is because this new style was then emerging at the same time as the power of the Rinpungpa and was destined to become characteristic of the court art of the new dynasty called to rule Tibet from the Tsang province during the fifteenth century. Perhaps it should then be called the Rinpung style.

At the end of this overview of the eighteen stupas formerly in place within the Tsuglagkhang of Densatil, it is clear that the artistic creation in this monastery perfectly testifies to the rise and fall of the Lang/Phagmo Drupa clan from 1208 to the end of the sixteenth century.

If this first attempt to classify the elements coming from the eight tashi gomang stupa of the Dorje Gyalpo Phagmo Drupa’s commemorative type will probably allow us to lay the logical bases for further development with a view to establishing a catalogue raisonné of the hundreds of objects that survived the destruction of the monastery, it also highlights all the gaps, both historical and material, the maximum filling of which will undoubtedly be the key to achieving an optimum result.

To this end, everyone's contribution to the Densatil Project by sending photographs of unpublished objects that may have come from this monastery would be greatly appreciated.

Appendices:
Plates illustrating the stylistic chronologies of certain objects from Densatil's tashi gomang stupas.

Plate A: Lokapalas

Plate B: Parnashavari Tier 5

Plate C: 8 arms Tara Tier 5

Plate D: Prajnaparamita / Tier 3

Plate E: Chakrasamvara / Tiers 2 & 3

Plate F: Vajravarahi / Tier 1

Plate G: Chakrasamvara / Tier 1

Footnotes

1) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014

2) Czaja O. and Proser A. Golden Visions of Densatil: A Tibetan Buddhist Monastery Asia Society New-York 2014

3) Henss Michael. The Cultural Monuments of Tibet 2014 pp 258,259 fig. 379, 380, 381.

4) Tsepon Wangchuk Deden Shakabpa. One Hundred Thousand Moons: An Advanced Political History of Tibet. Brill 2009 p.75

5) https://treasuryoflives.org/biographies/view/Chennga-Tsultrim-Bar/TBRC_P3473

6) Bell, Charles Alfred The Religion of Tibet, Oxford, Clarendon Press, 1931 p. 58

7) Tucci G. Gyantse and its monasteries part 1 & 3. Dehli: Aditya Prakashan, 1989 pp 100-101 fig. 14-18

8) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 pp 352,353

9) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 406

10) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 381

11) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 464

12) Tsepon Wangchuk Deden Shakabpa. One Hundred Thousand Moons: An Advanced Political History of Tibet. Brill 2009 p.75

13) Tucci G. To Lhasa and beyond. Roma: Ist. Poligr. dello Stato. 1956.    Das S.C. Lhasa and Central Tibet. Edited by the Hon. W. W. Rockhill. Delhi: Mehra Offset Press. 1902-1988.

14) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 563

15) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 354

16) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 82

17) For examples Himalayan Art Ressources #58301, #65205, #77201 #81410 #81411 #85965

18) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 379 and Luczanits C. Mandalas of Mandalas. The Iconography of a Stūpa of Many Auspicious Doors for Phagmodrupa In Luczanits and Lo Bue (eds.) 2010.

19) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 380,381

20) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 601

21) Luczanits C. Mandalas of Mandalas. The Iconography of a Stūpa of Many Auspicious Doors for Phagmodrupa In Luczanits and Lo Bue (eds.) 2010. Tibetan Art and Architecture in Context. PIATS 2006: Proceedings of the Eleventh Seminar of the International Association for Tibetan Studies, Königswinter 2006. Halle: IITBS, 281–310. -  Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014

22) Tucci G. To Lhasa and beyond. Roma: Ist. Poligr. dello Stato. 1956. p.128

23) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 382

24) Tucci G. To Lhasa and beyond. Roma: Ist. Poligr. dello Stato. 1956. p.142

25) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 363

26) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 pp 578-579

27) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 361

28) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 580

29) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 367

30) Weldon D. Defining the Style of a Group of 13th Century Tibetan Gilt Copper Statues. Asianart.com September 2018

31) Weldon D. Defining the Style of a Group of 13th Century Tibetan Gilt Copper Statues. Asianart.com September 2018. p. 10, 11, 12.

32) Karl-Heinz Everding  gSang phu Ne’u thog, Tibet’s Earliest Monastic School (1073). Reflections on the Rise of its Grva tshang bcu gsum and Bla khag bcu  in Zentral Asiatische Studien des Seminars für Sprach und Kulturwissenschaft Zentralasiens der Universität Bonn. 38/2009. International Institute for Tibetan and Buddhist Studies. Halle. P. 137 – 154.

33) Karl-Heinz Everding  gSang phu Ne’u thog, Tibet’s Earliest Monastic School (1073). Reflections on the Rise of its Grva tshang bcu gsum and Bla khag bcu  in Zentral Asiatische Studien des Seminars für Sprach und Kulturwissenschaft Zentralasiens der Universität Bonn. 38/2009. International Institute for Tibetan and Buddhist Studies. Halle. p. 149

34) Weldon D. Defining the Style of a Group of 13th Century Tibetan Gilt Copper Statues. Asianart.com September 2018. p. 7.

35) Ulrich von Schroeder, Buddhist Sculptures in Tibet, Hong Kong, 2001, Vol. II, p. 1047, pl. 263C

36) Luczanits C. Mandalas of Mandalas. The Iconography of a Stūpa of Many Auspicious Doors for Phagmodrupa In Luczanits and Lo Bue (eds.) 2010. Tibetan Art and Architecture in Context. PIATS 2006: Proceedings of the Eleventh Seminar of the International Association for Tibetan Studies, Königswinter 2006. Halle: IITBS p. 288

37) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 383

38) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 581

39) Tucci G. Deb t’er dmar po gsar ma. Tibetan Chronicles by bSod nams grags pa. vol.I, Tibetan Text, Emendations to the Text, English Translation and an Appendix to the Text containing two Minor Chronicles. Roma: IS.M.E.O. 1971 p. 209

40) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p 384

41) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 583

42) Huang Chunhe 西藏丹萨替造像研究 Capital Museum Beijing octobre 2009

43) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 386, 387

44) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 362

45) Ulrich von Schroeder, Buddhist Sculptures in Tibet, Hong Kong, 2001, Vol. II, p. 1044, 1045, pl. 261

46) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 856

47) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 385

48) Tucci G. Tibetan Painted Scrolls Rome 1949 pp 632-641
49) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 385

50) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 594

51) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 385

52) Luczanits C. Mandalas of Mandalas. The Iconography of a Stūpa of Many Auspicious Doors for Phagmodrupa In Luczanits and Lo Bue (eds.) 2010. Tibetan Art and Architecture in Context. PIATS 2006: Proceedings of the Eleventh Seminar of the International Association for Tibetan Studies, Königswinter 2006. Halle: IITBS p. 306

53) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 208

54) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 385

55) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 362

56) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 202

57) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 204

58) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 210

59) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 385

60) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 214

61) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 215

62) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 593

63) Roerich, George N. The Blue Annals. Dehli: Motilal Banarsidass, 1976 p 594

64) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 362

65) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 386

66) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 360

67) Czaja O. Medieval Rule in Tibet The Rlangs Clan and the Political and Religious History of the Ruling House of Phag mo gru pa Austrian Academy of Sciences Press 2014 p. 343

68) Tucci G. To Lhasa and beyond. Roma: Ist. Poligr. dello Stato. 1956. p.127

69) Tucci G. To Lhasa and beyond. Roma: Ist. Poligr. dello Stato. 1956. p.128