| Articles
previous version

A Kushan-period Sculpture
from the reign of Jaya Varman, A.D. 185
Kathmandu, Nepal.
by Kashinath Tamot and Ian Alsop

July 10, 1996

This is the latest printed version of our article, which appeared in Orientations, September 2001. It is considerably shorter than the previous version, which has more detail and many more photos and maps; we encourage the reader to visit the previous version as well. The main difference of interpretation between this and the previous is our acceptance of elements of R. Garbini's reading of the inscription (see previous version), with the exception of his reading of the date as 207/ AD 285. We still feel the inscription may be subject to further refinement of interpretation.

1. Discovery

In early May 1992 workers digging a trench for the foundation of a house in Maligaon in Eastern Kathmandu stumbled across the most important art historical discovery in the Kathmandu Valley for many years. Lying face-down at a level of about three feet they found a life size (171 x 49 cm) standing male figure carved in pale sandstone. ( fig. 1, 2 and 3)

1.Figure of Jayavarman, or of Jayavarman as a yaksha

The sculpture is the donation of an early Licchavi or pre-Licchavi monarch, named Jaya Varman in an inscription on the pedestal. Although the identity of the figure is contested, it is our opinion that it is likely Jaya Varman himself who is portrayed. Shortly after it was discovered the image was moved to the archeological garden in Patan. It spent several years in the Conservation Laboratory there before being moved in March 1998 to the National Museum in Chhauni, where it has been well installed in the room containing the earliest stone sculptures of the museum's collections.

2. Jayavarman (side view)

This statue, carved in a clearly Kushan style and from a stone type favored by Kushan sculptors, joins a very few others of the same general type found in the Kathmandu valley. These few images, which have been dated on stylistic grounds to the early centuries of the Christian era, have long posed a challenge to art historians who have had to analyze their presence in the valley without any documentary evidence. The inscription on this sculpture, clearly dated to samvat 107, most likely corresponding to AD 185, provides this previously missing evidence, and pushes back the epigraphical documentation of royal rule in the Kathmandu valley nearly three hundred years.

3. Jayavarman

The sculpture was recovered in four large pieces: body from neck to ankles, head, right arm (with hand missing), and pedestal with inscription and feet. The present reconstruction adequately restores what must have been the original figure, minus the missing right hand. The sculpture is carved in the round, but the rear of the image is flat and rather cursorily carved. The figure is shown in a firm stance, fully frontal in disposition, with feet equally spaced and planted without any dehanchement or turning of the body. The left arm is held akimbo at the hip, with the left hand held as a fist against the body; the right arm is broken off but appear to have been held with the elbow tucked back against the body and the forearm; as the right hand is missing it is impossible to ascertain its position. Under the armpit of the right arm is what appears to be the handle of some object that may have been held by that hand, although what exactly the object was cannot be ascertained.

5. Maitreya

4. Attendant or Bodhisattva

This disposition of the two arms is typical of Kushan sculpture of various subjects (fig. 4, 5) The lower body, from the waist down, is clothed in a dhoti with a central gathering falling straight between the legs. The waist is secured by a belt with ribbons hanging down the figure's right side. The upper body is unclothed except for a scarf across the chest, which appears to continue down over the figure's left shoulder and thence between the arm and body. Jewelry on the chest and upper body consist of a flat necklace folded across the chest and a torque-like thick necklace at the neck, armlets on the upper arm and two bracelets at the left wrist. The ears are adorned with heavy spiral earrings.

The head is adorned with an unusual turban or cap; in one place we can see traces of braiding that suggest a turban. The features of the face are somewhat eroded but are clear enough to show a wide, open and relatively flat face, with wide, open eyes and fleshy lips (fig 6). The figure is barefoot.

6. Jayavarman

The inscription (fig 7) , some 42 cm. in length, is written in Kushan Brahmi script comparable to Kushan epigraphy found elsewhere, and can be read as follows :

sam*vat a7 gri- pa 7 d(i)va pka maha-ra-jasya jaya varm(m)a(n*ah*)

Admitting that the exact reading and thus meaning of the inscription is still subject to debate, we supply the following tentative translation:

(In) the (Shaka) year 107 (AD 185), (on) the 4th (lunar) day of the 7th fortnight of the summer (season) , of the great King Jaya Varman

This inscription is the earliest epigraphy of the Nepal Mandala. Prior to this discovery, a broken pillar inscription of the year 381 [A.D. 459] found in a Pashupati courtyard was the earliest known Licchavi inscription (Rajbanshi, Shyam Sundar 1990. P. 1-2) Before the discovery of the Pashupati inscription, Manadeva's Cangu Narayana pillar inscription of the year 386 [A.D. 464] was accepted as the earliest inscription for about a century. The earliest inscribed and dated stone sculptures were the pair of Trivikrama (Visnuvikranta) figures of A.D. 467, one from Dhobicaur, Lajimpat now in the National Museum and the other at Tilganga, Pashupati.

7. Inscription

It is quite natural that the appearance of an inscription of such unprecedented age would cause some controversy among scholars: the sculpture and its inscription have attracted considerable notice among the scholarly community in Nepal, and several articles in Nepali and English have been published. There is no debate among the Nepali scholars that the date inscribed is 107, consisting of the letter "a", representing 100, followed by the numeral 7. But a team of Italian scholars, Angelo Andrea Di Castro and Riccardo Garbini, have suggested the date is actually 207. It is our belief that this reading is due to damage that occurred to the inscription after it was discovered, but further testing and examination may be needed.

Most scholars have accepted this date (107) to be in the same era as other later inscriptions of the early Licchavi period, that is the Saka samvat which began in A.D. 77/78 and is said to have been started by the great Kushan emperor Kaniska. Almost all scholars except Regmi (1992) agree with that the date is meant to be in the Saka samvat, which would place the inscription in A.D. 185 - or 184 if certain interpretations are followed (Regmi believes the era is a later one, and places the inscription in the 4-5th century CE) (Regmi, 1992a. pp 42-46.)

It should however be noted that the Saka samvat in its earlier periods is fraught with uncertainty. Although the era of the Saka samvat as employed in later centuries has been determined with comfortable accuracy, it is not entirely certain that the samvat in use in the early Kushan period and employed during the period of Kaniska is the same samvat, and scholars still debate the era of Kaniska's samvat. Estimates of the beginning of that era range from the A.D. 77/78 of the Saka samvat to later beginnings, as late as 128 or 144 A.D. (see Rosenfield, 1967, pp. 254 ff)

There are other differences in readings of this inscription. The phrase we show here as gri- pa 7 d(i)va pka, following Garbini (Garbini, p. 13) has been variously interpreted as sri paramadevanka, sri paramadevapka, sri pancadevapka, sri pancadevanka (indeed we originally read it as paramadevapka). This has resulted in various interpretations, leading to an identification of the figure with a yaksa or lord of wealth, or another interpretation of the month and day of the date.

While important, the variants in readings and interpretations do not contest the date (with the exception of the Italian 207 reading) or the mention of the "great king Jaya Varman".


The area of Maligaon (Maligam or Maga in Newari) and Hadigaon (Narah in Newari) in Kathmandu, where the sculpture was found, is considered by most scholars to be the site of the ancient capital of Licchavi Nepal. The present discovery of course lends further evidence to support this view and to identify it further with pre-Licchavi Nepal. Several Licchavi inscriptions have been found there. Thakur Lal Manandhar believed that the words Magah or Magala (Newari) and Maligaon (Nepali) are derived from the first Licchavi Royal Palace Managrha (Manandhar, 1977, 86-87). Several hundred metres north of the find site of the Jaya Varman statue is the famous Satya Narayana temple at Had*igaon. Excavations carried out just outside of the temple compound by an IsMEO team in 1984-1988 revealed several archaeological objects of the Kushan period (Verardi,1988) An earlier excavation was also held in Hadigaon at the Manesvari- area in 1965, although no objects of the Kushan period were found at that site.

9. Torso (rear view)

The area of Maligaon-Hadigaon is a rich source of sculptural remains. Lain Singh Bangdel identifies some 40 sculptures of the Kushan period in Nepal among which eight are from the Hadigaon -Maligaon area (Bangdel, 1982). Among these eight images is the one which most closely resembles the Jaya Varman sculpture, and which most clearly shows the Kushan antecedents of the early Nepalese material. This image - a sandstone torso - is also now in the Nepal Museum in Chhauni (fig. 8, 9). Until the discovery of the Jaya Varma, there had been varying views on the identification and dating of this sculpture, with most scholars referring to it either as a yaksha or a bodhisattva figure, and assigning a date within the first two centuries of the millenium.The discovery of the Jaya Varman sculpture introduces the possibility that the torso is also a royal portrait.

A comparison with Kushan figures of yakshas on the one hand and bodhisattvas on the other reveals the arguments for each (figs 4, 5). The lack of jewelry and the inclusion of the sash over the shoulder would argue for the bodhisattva identification. It is also worth noting that the yaksha as a sculptural type in the Indian tradition is generally earlier than the bodhisattva, which is relatively more common during the 2nd century, when we can surmise the torso was made; but as Bangdel has pointed out, Buddhist images of any type are strangely lacking in the earliest periods of Nepalese sculptural history, so this remark is valid primarily in the Indian context. (Bangdel, 1982b).

The torso was first seen by Ramesh Jung Thapa and Tara Nanda Mishra in 1965, laid in a ditch on the way to Sano Gaucar from Manes'vari-. Mr. Mishra in a recent communication informed the authors that the image was found at an area called Macagal in Maligaon some fifty yards from where the Jaya Varman statue was found. We confirmed the yaksha find spot with local people who remembered the statue, which, when Thapa and Mishra first discovered it, was being stoned by the children of the locality as a headless ghost. Locally the image is still remembered as Murkat*t*a Murti (headless image).

It is certainly noteworthy that the torso and the Jaya Varman sculpture were found so close together; as in addition they are both made of a similar pale sandstone and are stylistically linked, we may well assume that they belong to the same period and context of manufacture. The stone used is referred to in Newari as "sakva lvaham", that is "Sankhu stone", after the locality of Sankhu where it is found; it is not generally used in Nepalese sculpture of later periods, which is usually of a darker, denser stone known in Newari as "kotakha lvaham". That both of these sculptures were carved in a lighter sandstone might indicate that they were manufactured by or under the direction of carvers from the Indian Kushan tradition who favored sandstone as a material.

Previously the "yaksha/bodhisattva" torso held the peculiar fascination of an enigma. As Banerjee and Rijal put it:

" It is difficult indeed in the present state of our knowledge to account for its presence in Nepal, especially because of the existence of...a yawning gap in style and chronological affiliation between it and the earliest dated sculptural pieces known". (Banerjee, N.R. and B.K. Rijal 1968, p. 38 )

When Banerjee and Rijal wrote in 1968, little did they suspect that barely 50 meters from the ditch where the yaksha was found lay an even more impressive sculpture from the same tradition, The Jaya Varman sculpture, which was most likely in the same complex.

Now that we have the figure of Jaya Varman with which we can compare the torso, it seems possible that a very similar figure was intended, and we may call it a "royal torso" in the same mold as the figure of Jaya Varman. The jewelry and adornments, more worn in the torso, relate closely to the similar features in the Jaya Varman. The modeling of the torso is more pronounced than the full figure, showing a full and pronounced belly, similar is many ways to early Kushan figures of yakshas or bodhisattvas. We might surmise that this is another, later portrait of the same king, or a portrait of another king.

One of the fascinating possibilities raised by the discovery of the Jayavarman sculpture in a spot so close to the spot of the torso, is that both may have been part of a gallery of royal portraits or chitrasala (Misra 2000, p.2). We know that this was a Kushan custom. The possibility of the existence of such a gallery, originally suggested by Pal in 1974, cannot be ruled out, though we must agree with Di Castro and Garbini that this "intriguing hypothesis" still lacks "any definitive proof". (DiCastro and Garbini, 1996, p. 17).

There is not enough room here to embark on a comparison of this sculpture with the other early sculptures from the Kathmandu Valley which appear to be datable to the first two or three centuries of the first millenium. Lain Singh Bangdel has already embarked on this in a recent study of the early iconography of the valley (Bangdel, 1996), and hopefully more will be done in this area in the future to more firmly fix the chronology of this earliest period of Nepalese art

Most of the Nepali scholars who have discussed this image have assumed that the image is that of the king Jaya Varman himself, based obviously on the inscription. In a recent publication, however, Mohan Prasad Khanal vigorously challenged this view, claiming the image to be "indisputably" that of a yaksha (Khanal, 1995, p. 261-8). He cites strong stylistic evidence for this view, noting that the jewelry, stance and coiffure of the image are comparable with yaksha figures from Pre-Kushan periods and the Kushan period in India (see figs 3 and 4) . Khanal's stylistic analysis is well-taken, as comparison with Kushan yaksha figures do show many similarities. Furthermore, the present sculpture shows none of the royal accoutrements found in other known Kushan portrait figures (fig 10). In these portrait figures, the sculpture is normally fully clothed and booted, and bears royal arms. This figure shows none of these royal characteristics.

Khanal's interpretation is based partially on his interpretation of a reading of the inscription which varies somewhat from ours, by substituting "sri pancadevapka" for the seasonal dates first suggested by Garbini, which we have now accepted as the most likely reading. Khanal explains the pancadeva and maharaja of his reading of the inscription as references to Kubera, Lord of the Yakshas, with pancadeva interpreted as a variant of pacika (pancika). We find his interpretation somewhat tortured, and flawed by his failure to discuss or explain the reference to the name Jaya Varman; in addition, we do not agree with his reading of the inscription. But his stylistic arguments for identifying the figure as a yaksha are reasoned and not to be dismissed out of hand. In this context it should be noted that in many respects the figure bears an equally strong resemblance to Kushan bodhisattva images. A brief examination of a few Indian Kushan images of yakshas and bodhisattvas show the obvious similarities to the Jaya Varman sculpture (figs 3, 4).

10. Statue of Kaniska

Despite Khanal's arguments to the contrary, it is still possible - in our opinion likely - that the image is indeed intended to represent the king Jaya Varman, so clearly mentioned in the inscription. It is however unusual to portray a king - when a true portrait is intended - barefoot and without some royal accouterment, often a sword. The famous sculpture of Kanishka, is shown with a long sword, clad in royal robes, and impressively booted (fig 10). But in weighing the dissimilarities to other known Kushan royal portraits we should consider the dangers inherent in using prior Indian traditions to interpret contemporary conventions in Nepal. Certainly the boots and greatcoats of most of the known Kushan royal portraits were themselves foreign in the Indian context and equally so in Nepal, and we should not necessarily be suspicious of the idea of a barefoot monarch unencumbered by the heavy clothing of these other portraits.

It is not unknown for a king to pose in the form of a god, particularly in Nepal. Mary Slusser and Gautamvajra Vajracharya, in one of their ground-breaking articles on Nepalese art within the historical context, showed an image of Visnu with two attendant figures on the grounds of Pashupatinath to be a portrait of the early 7th century Lichhavi-period monarch Visnugupta with his two sons (Slusser and Vajracarya, 1973). It is worth noting that this portrait clearly shows the king in divine guise, and barefoot. So this may be Jayavarman in guise of a Yaksha.

Who was Jaya Varman?

There is no inscriptional evidence in Nepal between this second century inscription and the fifth century inscription of Pashupati, a span of some 274 years. So to interpret this inscription, we must turn to later documentary references to this earlier, previously undocumented period of Licchavi or pre-Licchavi rule. Among the most important of these later references is a genealogy in Jayadeva II's Pashupati inscription of A.D. 733. In that inscription another Jayadeva ("Jayadeva I") is described as twelve generations before king Vrisadeva, the great-grandfather of the historical king Manadeva.

Dhanavajra Vajracarya surmised that Jayadeva I was the king who established the Licchavis as the ruling house in Nepal as there is a qualifying 'vijayino' (victorious) before Jayadeva's name in this genealogy (although as we shall see below the most important Nepali chronicle lists several predecessors to this king) (Vajaracarya, 1973, no. 148, p.548 ff). It seems likely that the Jaya Varman of our inscription is in fact the same king mentioned by the later Jayadeva.

In previous work with the Gopalaraja vams'avali- (GV), Nepal's treasured and unique early chronicle, Mary Slusser and Gautamvajra Vajracarya have shown its unexpected accuracy in matters of ancient history (Slusser and Vajracarya, 1973). It is thus of interest to note that GV records the twelfth king preceding vis'vadeva (for vrisadeva) as Jaya Varman, (Vajracarya and Malla, 1985, p. 28) while Jayadeva II's geneology in the Pashupati stele separates Jayadeva I from Vrisadeva by the same number of kings. It is also of interest that the GV uses Varma as the name of many of the kings before Vrisadeva; the name is also found in the later Bhasa vams'avali-s.

This interpolation of fourteen monarchs between Jaya Varman/ Jayadeva I and Manadeva of Changu Narayan, a span of some 280 years, results in a reign span of twenty years per king, a not entirely unreasonable figure, although it is a longer reign period than the succeeding period between Manadeva and Jayadeva II, when 22 kings reigned in a period of some 270 years, only some 12 years per king. It is perhaps worth mentioning that the GV assigns reigns of varying length to the early Licchavi monarchs, the least being 37 years. It is also worth noting that adopting the Garbini reading of the inscription reduces the reign period to a perhaps more reasonable 13 years per king, which is in line with the firmly historical succeeding period. But since the list of fourteen kings is unsupported by any other evidence and may be incomplete, this is conjectural.

We cannot say for sure whether Jaya Varman was a member of the later Licchavi dynasty, or another ruling clan entirely. It is worthy of note that the GV does include him as a Licchavi, and we have learned to respect this almost uncanny -if late - chronicle.

It should be emphasized that name and dynasty are often not too closely linked. Most of the post 5th c. Licchavi monarchs had the name -deva and some -gupta. The dynasty name Licchavi is mentioned only occasionally in their epithets, and the first use of "Licchavi" in an inscription occurs in the early 6th century.

Tarananada Misra has recently published and extensive study of the earlier cultural relations between Nepal and the Kushana empire. Drawing on epigraphical, numismatic, taxinomic and archeological evidence, he makes a powerful case for a strong relationship between Kushan Indian and Nepal. (Misra, 2000). He further analyzes the name Varma to suggest that the term denoted a family other than the Licchavi. He shows the presence of various high personages in Licchavi inscriptions, who "played very important roles in the society" but "were a different… people than the Licchavis". One of those he cites is the great king Amsuvarma of the early 7th century. Could it be that this king represents a resurgence of the ancient Varman clan, reimposing their power briefly in the midst of another clan's ascendancy?

Misra also surmises that the "Licchavis of Vaisali might have entered the Nepal Valley around 410 AD", citing evidence against an earlier entry. But this evidence is uncertain and seems to be based on the assumption that the Licchavis, who were well known in India at the time of Jayavarman, could not have moved to Nepal without dying out in India.

It will take more research to determine the political and dynastic identity of the Varmas and Jayavarman, but it is clear that the culture this king ruled over was strongly allied to the great Kushan empire then on the wane in India. Whether they were a clan distinct from the later Licchavi kings, or whether they were the earliest of the Licchavis, is still uncertain.

The implications of this sculpture for our assessment of the artistic traditions of the Kathmandu Valley are profound. Previously, the various clearly pre-Gupta sculptures that have been found in the valley were considered as puzzling precursors to the voluminous sculptural remains from the fifth century onwards. Early Nepalese art was considered largely in the context of the art of the Guptas, and rightly so, for the majority of the early sculpture which has come down to us clearly shows a relationship to the Gupta style.

A sculpture such as the headless sandstone torso of the National Museum could only be the subject of conjecture: clearly Kushan in style and materials, it had no closely related, similar examples with which it could be compared. The Jaya Varman sculpture now provides a previously missing context for the small and scattered group of sculptures that can clearly be seen to be pre-Gupta. Our conception of the sculpture of Nepal must now stretch back farther in time than we had once believed. The second century, not the fifth, must now be our starting point for charting the history of Nepal's sculptural traditions. The Kushan tradition, not the Gupta, must be seen as the tradition in which Nepalese sculpture first arose and developed.

all images © Orientations magazine


  • Banerjee, N.R. and B.K. Rijal, 1968
    "Three early sculptures in the National Museum of Nepal",
    Ancient Nepal, No. 4, pp 37-43
  • Bangdel, Lain Singh, 1982
    Pracina Nepali murtikalako itihasa
    Kathmandu : Royal Nepal Acaademy.
  • Bangdel, Lain S, 1982,
    The Early Sculpture of Nepal, Delhi, Vikas, 1982
    Pl. 1 and p. 12.
  • Bangdel, Lain Singh, 1996, "Iconography of the Early Sculptures of the Kathmandu Valley" pp 1-15 of Bangdel, Mishra and Mukund Aryal, May 1996, A Report on the study of Iconography of Kathmandu Valley and Their Preservation and Protection, Kathmandu, HMG Ministry of Youth, Sports and Culture, Department of Archeology.
  • Castro, Angel Andrea Di and Riccardo Garbini, 1996. "An Inscribed statue of year 207 from Maligron, Kathmandu."East and West (Rome:IsIAO), 46: 3-4 (December), pp.299-317.
  • Deo, Shantaram Bhalchandra 1968
    Archaeological Excavations in Kathmandu : 1965
    Kathmandu : Department of Archaeology, HMG.
  • Khanal, Mohan Prasad, 1995
    Nepali Kala
    Center for Nepal and Asian Studies, Tribuvan University, Kathmandu, V.S. 2052 - 1995
  • Manandhar, Thakur Lal 1977
    "Nepal in the Early Medieval period : Gleanings from the Bendall vamsavali",
    Journal of the Nepal Research Centre, Vol. I pp. 83-88.
  • Mishra, Tara Nanda, 2000 "Dated figure of king Jayavarma, the tradition of figure making and the historical importance of this discovery". Ancient Nepal, 146, (November), pp.1-23.
  • Pal, Dr. Pratapaditya, 1974,
    The Arts of Nepal, Part I, Sculpture,
    Leiden, E.J. Brill
  • Rajbanshi, Shankar Man 1993.
    "Maligaumma prapta nayam abhilekhako barema vicara vimarsa",
    Ancient Nepal, 135 (Oct.-Nov.), pp. 1-3.
  • Rajbanshi, Shyam Sundar 1990. "Pasupatinatha pranganama prapta aprakasita licchavi abhilekha", Ancient Nepal, 115 (Dec. 1989 - Jan. 1990
  • Regmi, Dinesh Chandra 1992a.
    "Jayavarma, vamsavali ra puralekhana",
    Madhuparka, 25:5 / Sept.-Oct. 1992 (VS 2049 Asoj), pp. 42-46
  • Rosenfield, John M., 1967,
    The Dynastic Art of the Kushans,
    Berkeley, Los Angeles, University of California Press
  • Slusser, Mary 1982
    Nepal Mandala,
    Princeton, Princeton University Press
  • Slusser, Mary Shepherd and Gautamavajra Vajracarya, 1973
    "Some Nepalese Stone Sculpture: A Reappraisal within their Cultural and Historical Context",
    Artibus Asiae, Ascona Switzerland, 1973, Vol XXXV, 1/2, pp 790-138
  • Tamot, Kashinath and Alsop, Ian, 1996 "A Kushan - period Sculpture from the reign of Jaya VarmaA.D. 184/185 Kathmandu, Nepal. Published in
  • Vajracarya, Dhanavajra 1973.
    Licchavikalaka Abhilekha.
    Kathmandu : Institute for Nepal and Asian Studies.
  • Vajracarya, Dhanavajra and Kamal P. Malla 1985.
    The Gopalarajavamsavali
    Wiesbaden : Franz Steiner Verlag Gmbh.
  • Verardi, Giovanni 1988
    Harigaon Satya Narayana, Kathmandu, A report on the Excavations carried out in 1984-1988.
  • ----------------- 1977
    Excavations at Harigaon, Kathmandu, Final report.
  • Zimmer, H, 1955,
    The Art of Indian Asia,
    Princeton, Bollin

previous version
| Articles