The Gupta-Era Vishnu Temple at Deogarh, Dated to the Early Sixth Century CE, Stands on a Five-Shrine Platform 55.5 Feet Square and Carries a Relief of Vishnu Asleep on the Serpent Ananta

At Deogarh, in Lalitpur district of Uttar Pradesh, a temple stands on a moulded stone platform 55.5 feet (about 17 metres) square and 9 feet high, with a flight of steps in the middle of each side and a small shrine at each corner. India’s dossier for UNESCO places it at the beginning of the sixth century CE, and each of its three outer walls holds a deep niche with a large relief. The south niche shows Vishnu asleep on the coils of the serpent Ananta. Popular usage calls it the Dashavatara temple, the temple of the ten avatars, though the surveyor who excavated it in 1917 and 1918 reported finding no evidence for that dedication. The UNESCO entry is a Tentative List entry, a preparatory step toward a nomination, and it confers no World Heritage status.

What India’s UNESCO dossier records

India’s Permanent Delegation to UNESCO submitted a serial nomination called Gupta Temples in North India, which UNESCO lists with the date 11/02/2025 and the reference 6805. It has 20 components across Madhya Pradesh, Uttar Pradesh, Rajasthan and Bihar, cites criteria (i) and (iii), and puts the Deogarh temple fifth in its list. UNESCO’s page notes that the content of each Tentative List is the sole responsibility of the State Party, so the figures below are India’s. The Udayagiri caves near Vidisha are another component, and this article covers them.

According to the dossier, the Betwa river encircles the site on three sides. The temple is a panchayatana, a five-shrine complex: a main shrine in the middle of the platform and a subsidiary shrine projecting from each corner. The platform panels, set between pilasters, show episodes from the Ramayana and from the life of Krishna. Facing west, the main shrine measures 10 feet square inside and 18.5 feet square outside. Its plan is triratha, meaning each wall steps out in three planes, with a doorway on the west and a central projection on the other three sides. Five carved bands frame the door, with yakshas (nature spirits), ganas (attendants), rosettes and amorous couples alternating with pairs of dwarfs, and Ganga and Yamuna in the extended flanks above the lintel. On the lintel, the central panel shows Vishnu seated on the coils of the serpent Adishesha.

Cunningham called it the Gupta temple in 1880

Alexander Cunningham of the Archaeological Survey described Deogarh in his report on tours made in 1874 to 1877, printed as volume 10 of the Survey’s reports. No name for the temple had come down to him, he wrote, and he proposed to call it the Gupta temple because it had the characteristics of that period’s style. Of its dedication he had no doubt: it was a temple of Vishnu.

His plan divides the platform into nine equal squares, with the temple on the middle one. He counted 20 sculptured panels on each of the four sides of the platform, 80 in all: 16 along the parapet and four on the walls flanking each staircase. Four were still in place when he saw them, two on the south face and two on the west. One showed a man and woman each holding a child, with two bullocks behind. Another he identified, following his colleague Beglar, as Lakshmana cutting off the nose and ears of Shurpanakha in the presence of Rama and Sita. Terraces with sculptured walls were known to him in Burma, he wrote, but not elsewhere in India.

Cunningham thought the sleeping Vishnu easy and graceful in attitude and dignified in expression, and from the excellence of the sculpture he inferred a date well before the Khajuraho temples of the tenth century. He placed the temple between 600 and 700 CE, because its pyramidal spire marked it as one of the latest examples of the Gupta style. No inscription turned up on the temple in his survey. His reading of the east relief has not survived: he took the four-armed ascetic for Shiva, and later scholars read the scene as Nara and Narayana at their penance.

The 1917 to 1918 excavation and the name Dashavatara

In the Director-General’s report for 1917 to 1918, Daya Ram Sahni is quoted from the Northern Circle on what the dig at the Gupta temple found. Sahni reports that the lower platform survived in fairly good condition and the upper platform had gone entirely. Each side had its flight of steps, and remains of smaller shrines came up at the corners of the basement. Among the sculptures dug up were many reliefs of domestic scenes, images of Vishnu, and a well-preserved relief of Rama and Lakshmana visiting the hermitage of Shabari, which the report prints as its first plate. Sahni wrote that the temple was most probably dedicated to Vishnu, but that he found no evidence for dedication to the ten incarnations, as P. C. Mukherji had proposed.

Puratattva, a heritage website that summarises the published literature, says Mukherji’s report of 1899 called it the Dashavatara temple, or Sagar Marh, because of ten incarnations of Vishnu he took to be depicted on it and on a nearby rock-cut well. Puratattva adds that Madho Sarup Vats, who excavated after Sahni and published his Memoir for the Survey in 1952, found panels of Vishnu’s incarnations such as Balarama, Narasimha and Vamana, yet still called it the Gupta temple. The scanned text of Vats’s Memoir was unreadable to tuput, so that account rests on Puratattva. UNESCO’s dossier acknowledges the split. The name comes from ten incarnations said to be depicted, it says, but many archaeologists and experts refer to the monument as the Gupta temple.

Avatar figures are present. Alexander Lubotsky and the Impart Encyclopedia of Art both describe Narasimha, the man-lion, and Vamana, the dwarf, beside the seated Vishnu on the lintel. Some scholars, Impart notes, argue that the label Dashavatara may have been given by mistake, since no actual panel of the ten exists. Puratattva adds that Vishnu seated on the coils of Shesha, as on this lintel, is a rare form, and that a similar figure in very large proportions is found in a cave at Badami. The Vaishnava cave at Badami has such a figure at the east end of its verandah.

Three niches and a doorway

On the north wall, Vishnu rides his mount Garuda to free the elephant king Gajendra. In Cunningham’s description the elephant stands amid lotuses, with a serpent king and queen below him in adoration, and four flying figures above hold a canopy over Vishnu. Lubotsky counts two serpents with folded hands and an elephant offering a lotus to Vishnu in gratitude.

The east niche has two ascetics in meditation, Nara and Narayana, with deer and a lion at their feet. Brahma sits on a lotus in a separate register above. Lubotsky credits an earlier scholar, Shastri, with showing that the Vishnudharmottara Purana’s description of Nara and Narayana matches this relief closely. That Sanskrit manual dates to about the seventh century CE in Lubotsky’s reckoning, and its third section covers images and temples.

The south niche holds the relief that gives the temple much of its fame. In the World History Encyclopedia’s description, Vishnu reclines on the serpent Ananta and floats on the waters of oblivion, with Brahma seated on a lotus above him. Cunningham saw four arms, a bent right knee and Vishnu’s head resting on one of the left hands, under the canopy of Ananta’s seven hoods. Above, gods sit in a row. Lubotsky names Brahma at the centre, flanked by Shiva with Parvati on the bull Nandi and by Indra on his elephant, with two more gods whose identity is debated; he takes them for Kartikeya and Surya. Below is a slab of six standing figures, five male and one female. Cunningham judged that the slab had been fitted in to fill space. Lubotsky and Puratattva identify four of the figures as Vishnu’s weapons in human form, fighting two demons whom Puratattva names as Madhu and Kaitabha. One of them, the woman, is the personified mace.

Details remain open. Impart and the UNESCO dossier describe Brahma on a lotus that rises from Vishnu’s navel, while Lubotsky and Puratattva say the stalk rises from behind the reclining god. At Vishnu’s feet sits a woman, Lakshmi in the World History Encyclopedia and in the reading Puratattva attributes to Vats, and Bhudevi in the reading it attributes to T. A. Gopinatha Rao.

How Lubotsky reads the four Vishnus

A 1996 article by Lubotsky in Ars Orientalis argued that the Vishnudharmottara Purana describes the Deogarh temple under the name Sarvatobhadra, and that its author must have seen it. Lubotsky notes that the text places a square platform with four staircases, four corner shrines and a band of panels around the platform, and that the plan fits what is left at Deogarh, where the superstructure and the small shrines on the platform are ruined. In his account the doorway and three reliefs are the only parts still intact.

Next, Lubotsky reads the doorway lintel and the three niches as four forms of Vishnu in the Pancharatra tradition, the vyuhas, or emanations: Vasudeva on the lintel, then Sankarshana at Gajendramoksha, Pradyumna at Nara-Narayana and Aniruddha in the sleeping Vishnu. Impart records that reading and another, in which the four scenes stand for progressive enlightenment toward liberation. On the first, it says, a devotee would walk clockwise round the temple, and on the second, anticlockwise.

In the same season of 1917 to 1918 the Survey also uncovered the Varaha temple mound in the fort at Deogarh. The report says panels of Gajendramoksha and of Nara and Narayana came out of that dig, and cites a suggestion by Mr. Hargreaves that the Varaha temple imitated the Gupta temple. Lubotsky notes that the Varaha temple faces east, the reverse of the Gupta temple, and that opinions on its date differ.

The date: no ruler named

None of the sources consulted for this article ties the temple to a named Gupta ruler, and Cunningham found no inscription on it. Gupta here names the period and the style, and the empire is usually dated to about 320 to 550 CE. India’s UNESCO dossier places it at the beginning of the sixth century. Lubotsky writes that it was finished in the first half of the sixth century and refers readers to Joanna Williams’s 1982 book on Gupta art for the dating discussion. Impart gives the early sixth century, while the World History Encyclopedia has the fifth to sixth century.

Puratattva records the earlier proposals. Cunningham argued for 600 to 700 CE. R. D. Banerji suggested about 575 CE on his theory of how the tower form developed, and Percy Brown placed the temple at about 500 CE, a date Puratattva says Sahni and Vats accepted. Puratattva also reports an inscription on a detached pillar at the site, in Gupta-period script, which in its transcription names a devotee called Govinda and carries no date or ruler. Sahni took the script as pointing to the early Gupta period.

What remains of the tower

Of the shikhara, the tower over the sanctum, the dossier says only four base courses survive. Its lower two courses are plain and the higher ones are carved. The upper course carries early northern-style (proto-Nagara) square units with arched window motifs, preserved on the east face only. Above that, the dossier says, only the core remains with its facing stones gone, so the exact design cannot be certain. Loose fragments at the site suggest a stepped pyramid. Cunningham, writing in 1880, described the spire as long in ruins, with the lower part still standing on the east side and amalaka stones, the ribbed discs that crown a Hindu tower, lying about.

Vats’s conjectural restoration, as Puratattva reports it, was a straight-edged pyramid in three receding tiers with amalakas at the corners. Impart gives an original height of about 40 feet, and a learning-material sheet on gcgldh.org gives about 12 metres, describing the original appearance as largely conjectural. Impart describes Deogarh as the earliest fully structural temple in north India to have a shikhara. The dossier says it is one of the few surviving early Gupta temples with one, and Puratattva calls it probably the only surviving early Gupta temple with a tower. Each of the three sources words the claim differently.

The Lalitpur district website says the Archaeological Survey of India protects the monuments at Deogarh. Of the 80 plinth panels Cunningham counted, Puratattva says only two remain in place, and that most of those excavated are in the National Museum in New Delhi. Impart says some of them are in the museum.