Masons at Dhamnar in Madhya Pradesh Cut 51 Caves Into a Laterite Hill and Dug a 30-Foot Pit to Leave a Whole Temple Standing in the Same Rock, and India Has Put the Site on UNESCO's Tentative List

A hill near Dhamnar village in Garoth tahsil (sub-district) of Mandsaur district, Madhya Pradesh, holds 51 rock-cut caves on its southern face and, a short way north of them, a temple that masons made by digging a pit into the hilltop and leaving a block of rock standing in the middle. India’s submission to UNESCO counts 14 larger caves and 37 smaller ones across 5.2 hectares. The Archaeological Survey of India (ASI) lists the caves under Act No. LXXI of 1951, dated 28 November 1951. UNESCO’s page carries the site as “The historic ensemble of Dhamnar” under criteria (ii) and (iv), with India’s submission dated 2024. A place on the Tentative List is the first step toward a nomination and is not World Heritage status.

A hill of laterite, hollowed from the south

Alexander Cunningham of the ASI surveyed the hill in 1864 to 1865 and described it as flat-topped coarse laterite, a rusty, crumbly stone, with a cliff 20 to 80 feet high on the south side and a slope of fallen rock below. The principal caves sit in a bay in that cliff. UNESCO’s text says the entrances face south around a crescent. Cunningham thought laterite poor stone for sculpture and well suited to quick excavation, which suits the plain look of most cells.

The better-finished caves copy timber buildings. According to UNESCO’s table, the hall called Badi Kacheri has pillars and stone beams that resemble wooden members. Cunningham gives the details. It is a flat-roofed hall 20 feet square on four pillars, with three rooms on each side and, at the back, a small shrine holding a stupa (a dome-shaped Buddhist monument). The side openings of its front portico are closed by a railing cut from the rock to imitate a wooden one with upright bars. The neighbouring Chota Kacheri is a chaitya, a worship hall built around a stupa, 23.5 feet by 15, with a stupa 16.5 feet high, and its roof is domed and ribbed to imitate a wooden structure. The “Kacheri” names mean court house and are local.

The largest excavation is Bhim Bazar. A chaitya with a ribbed roof stands inside a cloister of cells, an open passage runs all round it, and two small free-standing stupas, each 5 feet across, stand before the door. Puratattva, summarising Cunningham, gives the whole as 115 feet by 80. Cunningham judged the pairing of a worship hall with monks’ cells, which Fergusson had not seen elsewhere, to be peculiar to Dhamnar.

The small hall with fifteen Buddhas

Cunningham thought the most important cave was a small one, the Chota Bazar or “small market”, also called the child’s cave. It holds 15 Buddhist statues, he counted, which in a site without inscriptions gave him his best evidence for working out what the caves were. An open court with a small stupa leads to a shrine where a seated Buddha 8 feet high teaches, with a standing Buddha 10 feet high on a lotus at each side of the door. On the east wall lies a reclining Buddha 15 feet long, shown at his death. UNESCO’s submission calls that reclining figure rare.

Local people gave the figures names from the Mahabharata: the seated Buddha is Bhima, the reclining one is Bhima’s baby, and three seated figures are Pandu and two of his sons. James Tod, who visited in December 1821 and whom Cunningham credits with first making the caves known, was told by his Jain guru that the statues were Jain teachers. Cunningham found none of the symbols that identify Jain teachers and read the postures instead: meditation under the Bodhi tree, teaching and death, the three great events of the Buddha’s life.

He also saw plaster still stuck to the rock in sheltered corners and under the statues’ arms. He concluded that the donors’ records had been written on the plaster and had gone with it.

Counting the caves, and dating them without an inscription

The count has moved. Tod said he counted 170 caves. Fergusson, who visited in 1845, put the total at 60 to 70 once mere scratchings in the rock were included, and Cunningham agreed that there were no more than 70. India’s submission settles on 51, split between the two sizes. The Outlook Traveller feature repeats that figure.

None of the authors found an inscription that dates the caves. Cunningham assigned them to the 5th, 6th and 7th centuries, chiefly because the tall stupas resemble a great stupa near Benares, from which he had taken an inscription in letters of the 6th or 7th century. By Cunningham’s account, Fergusson placed the Chota Bazar cave, which he thought the latest of the group, in the 7th century by its likeness to Cave 27 at Ajanta. When Fergusson and Burgess published The Cave Temples of India in 1880, they quoted Cunningham’s range and judged that the larger caves could hardly be dated before the 8th century. India’s submission uses 5th to 7th, and the ASI’s page, which names the colossal seated Buddha and the reclining Buddha as the conspicuous sculptures, gives the 7th century.

A figure of 408 to 475 CE circulates online as Fergusson’s date for the caves. In the sources opened for this article it has a different origin: Cunningham cites gold coins of three Roman emperors, who ruled from 408 to 474, found in a stupa at Hidda, as a comparison for the shape of stupas. It does not date Dhamnar.

One small find bears on the name. The ASI’s review Indian Archaeology 1960-61 reports that A. H. Khan, while the monuments were being conserved, found a clay seal reading Chandanagiri-mahavihara in letters of the 5th to 6th century. Puratattva reads this as the old name of the Buddhist establishment and notes, citing D. R. Patil, that the name Dhamnar is not known from any historical or literary source. Puratattva also records the view of K. C. Jain, who thought the modern name comes from Dharmanatha, the name of the linga in the temple. Madhya Pradesh Tourism attributes the caves to the Gupta period. Cunningham found no inscription naming a patron, and none of the other sources opened for this article names one.

The temple cut downward

North of the caves, Cunningham found a different kind of excavation, which he judged more curious than the Buddhist caves because it is rarer. Masons dug a rectangular pit into the top of the hill, left a block of rock standing in its middle and carved that into a temple, so that, as Fergusson and Burgess put it, the building cannot be seen from outside at all. Cunningham gave the pit as 104 feet long and 30 feet deep at its west end. Fergusson and Burgess gave 104 feet by 67. UNESCO and Puratattva, which credits Henry Cousens’ report of 1909, give 170 feet long, 66 feet broad and 30 feet deep. The sources do not explain the gap in length.

A crooked passage 282 feet long leads from the face of the hill to the pit and also drains rainwater from it, as Cunningham explains. Steps cut along the long sides lead down from the hilltop. At three points the passage runs through tunnels, two of them 17 feet long and one 23 feet long, and the longest has a niche on each side, with Bhairava on the south. Cunningham found three letters of an inscription on the pedestal of the figure on the north side, in characters he placed in the 8th or 9th century.

The central temple is 48 feet long and 33 feet across its portico, according to Fergusson and Burgess. It has the usual northern plan of porch, hall, ante-chamber and sanctum, and a tower in the Nagara style. Seven smaller shrines surround it. Inside the sanctum Cunningham saw a four-armed Vishnu in basalt, and local people call the temple Chaturbhuj, “the four-armed one”. Puratattva adds that ten incarnations of Vishnu are carved around the image and that the sanctum doorway was made separately and installed, which Puratattva suggests was because fine carving would not hold in the crumbly rock. Among the smaller shrines Cunningham records a panel of Vishnu asleep on the serpent Shesha, with the demons Madhu and Kaitabha at his feet, and another of Vishnu’s incarnations in which the figure that usually stands for the Buddha is replaced by a four-armed Vishnu. Puratattva also records a dancing Shiva with four mother goddesses.

A Shiva linga now stands before the Vishnu image. Cunningham took it as a later addition to a temple that was Vishnu’s from the start, and Fergusson and Burgess agreed. UNESCO’s nomination uses the name Dharmarajeshwara, with Dharmanatha as an alternative.

When the temple was cut

Fergusson and Burgess placed the temple in the 8th century, at about the age of the temple at Barolli, which it resembles from outside. Cunningham put it in the 8th or 9th century on the strength of the three letters. Puratattva reports that D. Dayalan, in his 1995 study of the monolithic temples of Madhya Pradesh, dated it to the later half of the 9th century on architecture, sculpture and letter forms. On those dates the temple follows the caves by a century or more.

Fergusson and Burgess set the Dhamnar pit beside the Kailasa temple at Ellora, which was begun at about the same time, they thought, and they read both as a Brahmanical answer to Buddhist caves. The reading is theirs and dates from 1880. UNESCO’s nomination reads the same hill differently, as a place where two religions kept sanctuaries side by side. Like the Kailasa at Ellora, which was also cut downward from a slope, the Dhamnar temple was shaped by removing rock, and a Sahapedia article on the Masrur temples in Kangra lists it among the monolithic rock-cut temples of India. That article places Dhamnar in Rajasthan, an error: the ASI lists it in Mandsaur district of Madhya Pradesh.

Protection and the nomination

The ASI’s 1969 to 1970 review records a rubble wall built along the top of the rock-cut temple to prevent accidents. UNESCO’s submission says parts of the caves are in poor condition from water damage and vandalism and that the site is otherwise undisturbed apart from excavation. It compares Dhamnar with Ellora and Bhimbetka at home and, among others, with Lalibela in Ethiopia and Dambulla in Sri Lanka abroad. The ASI’s own page for the caves, in an archived copy of 7 October 2024, records a temple and a house built inside the protected area, a case it describes as pending in court.

Cave sites in the same state have inscriptions that fix their dates, such as the Udayagiri caves near Vidisha with a Gupta-year inscription of 401 CE and the Bagh caves in Dhar district, which have a copper-plate grant naming the monastery. Mandsaur itself, ancient Dasapura, has its own written record in a weavers’ guild inscription of about 473 CE.

As of 5 November 2026, UNESCO’s page still lists Dhamnar as Tentative List reference 6736, submitted by India’s Permanent Delegation, under criteria (ii) and (iv).