Sittanavasal's Arivar Koil, a Jain Cave in Pudukkottai With Ceiling Murals That Scholars Rank Next to Ajanta, Carries a Pandya-Era Repair Inscription and a Builder Still Argued Over

Arivar Koil, the rock-cut Jain cave on the western face of a hill at Sittanavasal in Pudukkottai district, Tamil Nadu, is about 24 feet long, 12 feet wide and 11 feet high, and its painted ceiling is why Swaminathan and the Tribune rank it next to Ajanta in importance. S. Paramasivan, a chemist at the Government Museum in Madras, wrote in Nature in 1937 that the paintings are the best Jain murals of the classical Ajanta school that survive in India. Three different things need dates here, and the sources disagree on all three: the cave itself, the inscriptions, and the paintings.

Recent writing spells the site Sittanavasal and older scholarship Sittannavasal. Both mean the same hill, and the name has its own dispute. Impart, an online encyclopedia of South Asian art, notes one reading as “abode of great saints” from Sanskrit and another as a Tamil form of “smaller Annalvayil”, after a nearby town.

What is cut into the rock

The cave faces west. According to Prof. Subramanian Swaminathan, in a heritage study written in 2008, it began as a sanctum (the inner shrine room) and a front hall, both carved from living rock. A pillared veranda in front is a twentieth-century addition, put up at the instance of a British administrator named Tottenham, with pillars brought from the ruins of the temple at Kudumiyamalai. Kudumiyamalai, in the same district, is also where a rock-cut musical score is cut into a wall.

Two seated figures sit in niches on the side walls of the hall. One is Parshvanatha, the twenty-third Tirthankara (a liberated teacher whom Jains revere), shown under a five-headed serpent hood. The other sits under a single umbrella, and a label on the nearby pillar reads Sri Tiruvasiriyam, “revered teacher”. In Swaminathan’s reading, that makes the figure a teacher and not a Tirthankara. Three more seated figures are carved on the back wall of the sanctum, according to Saurabh Saxena in the heritage site Puratattva.

Paramasivan wrote in 1937 that the paintings survive only on the ceilings, the tops of the pillars and the beams. The largest is a lotus tank on the ceiling of the hall, full of fish, birds, buffaloes, elephants and men gathering flowers. T. N. Ramachandran and the art historian Stella Kramrisch read it as the “khatika bhumi”, the tank region a believer passes on the way to the samavasarana, the hall where a Tirthankara preaches. I. Job Thomas, also cited by Puratattva, calls that identification thinly supported, since a samavasarana is normally painted whole, and points to Tamil Jain literature that links lotus offerings with the Jina, the conqueror. On the pillars are two dancing celestial women and a royal couple with a monk.

How the paintings reached the scholars

The first publication was an inscription, not a painting. Saxena writes that the paintings came to scholars’ notice after that inscription appeared in the Annual Report of Epigraphy for 1904. Venkataranga Raju, curator of the Pudukkottai Museum, made water-colour copies of the murals in 1910. S. Radhakrishna Iyer, a local historian, described the cave in his General History of the Pudukkottai State of 1916, though Swaminathan notes the book’s mainly regional readership kept it from wider notice.

The wider notice came from G. Jouveau-Dubreuil. T. A. Gopinatha Rao, the iconographer, told him in 1919, according to Puratattva, that the paintings were as old as the shrine and well enough preserved to copy in full. According to Swaminathan, the two worked at Sittanavasal from 1918 to 1920 and put it before the archaeological world. Impart gives the title of their 1920 monograph as Fresco Painting at Sittanavasal.

The Maharaja of Pudukkottai had the paintings cleaned in 1937 to 1939, Swaminathan writes, with a preservative coat applied and loose plaster strengthened, without retouching. In 1942 Paramasivan and K. R. Srinivasan, who were cleaning the paintings, found a patch of older carpet-pattern paint beneath a later layer.

The cave: Pallava, Pandya or unattributed

No foundation inscription survives on the cave, Swaminathan says, so its date rests on style: the plain plan of a sanctum and front hall, and the pillars. Several writers read those as Pallava work. Paramasivan’s 1937 paper assigns the paintings to the time of the Pallava king Mahendravarman I, whom he dates to about 600 to 625. Saxena reports that Jouveau-Dubreuil attributed the cave to Mahendravarman I (given there as 580 to 630), probably on the style of the pillars, and that K. R. Srinivasan agreed about the Pallava features. Cuboid pillars like these are characteristic of early Pallava architecture, Impart adds. The Tribune in 2005 wrote that the cave temple is said to have been completed in the seventh century by Mahendravarman, and the Pudukkottai district administration’s tourism page ties the paintings to the same king. Tradition holds that Mahendravarman was a Jain before he turned to Shaivism under the saint Appar, a point Puratattva and Impart both mention.

Doubt starts with geography. Mahendravarman’s territory is said not to have reached beyond Tiruchirappalli, Swaminathan writes, and that caves of similar style stand in the Pandya country, where Pallava power was unknown, such as a Shiva cave with painting remains at Tirumalaipuram near Tirunelveli. Saxena adds a conditional argument. If the Sanskrit verse at Mandagappattu, which says Mahendravarman made a shrine without brick, timber, metal or mortar, marks the king’s first rock-cut shrine, as he says K. R. Srinivasan reads it, then a cave of this type at Sittanavasal would more probably be Pandya work, by Maran Sendan (654 to 670) or Arikesari Maravarman (670 to 700). That verse’s summary in South Indian Inscriptions does not use the word “first”. Saxena’s own estimate for the cave is the seventh or eighth century, probably.

Other dates sit further out. Impart gives 500 to 800 CE and says no evidence definitively dates or attributes the cave. A Pudukkottai district page calls it a second-century Jain temple and gives no basis, while saying in the same paragraph that the paintings go back to Mahendravarman. Swaminathan’s own summary is that the builder cannot be ascertained without a foundation inscription, and that the cave must be older than a ninth-century inscription that mentions repairs.

The Tamil inscription of a Pandya repair

That repair record is the firmest anchor at Sittanavasal. The published series South Indian Inscriptions lists it as No. 45 of its Pandya volume, from the Annual Report of Epigraphy for 1904 (number 368). It is cut on the rock south of the cave, in Tamil verse, in early letter forms that mark a bare consonant with a dot. It names the Pandya king Srivalluvan, meaning Srivallabha, with the title Avanipasekharan. On his behalf, it says, Ilan Gautaman, described as an asiriyan (teacher) of Madurai, repaired the front hall of the temple at Annalvayil and built a new pillared hall in front of it. The stone is damaged, but the surviving lines suggest land was given to the temple’s priests. Swaminathan reads the temple’s name in it as Arivar Koil, the temple of the Arhats, the Jain term for those who have conquered their senses.

The corpus summary gives no year. Swaminathan and Puratattva both place the king at 815 to 862 as Srimara Srivallabha, the range K. A. Nilakanta Sastri gave that ruler in his 1929 history of the Pandyas, and Puratattva says the regnal year of the record is unknown. The name recurs, which leaves room for error: Nilakanta Sastri also describes a later Jatavarman Srivallabha, tentatively a contemporary of the Chola king Kulottunga I, who came to the throne about 1070. The corpus calls the letters early, and Puratattva’s list of the cave’s inscriptions dates this one to about 815.

The hall that Ilan Gautaman added did not last. It collapsed through neglect, says Swaminathan, and Puratattva adds that it was already gone when renovation began in 1942. The 17-line record fixes the latest date by which the cave existed, since it must be older than the repair, and gives no earliest date.

One layer of paint or two

The 1942 cleaning raised a second dating question, because the paintings are not all of one moment. In Swaminathan’s account, the later layer, probably Ilan Gautaman’s, spread into the sanctum and over the hall ceiling, pillars, brackets and beams. Puratattva sets out two views. Kramrisch holds that the two ceiling layers were painted together. K. R. Srinivasan holds that the first is as old as the cave and the second came with the ninth-century extension. Swaminathan’s opening paragraph dates the surviving paintings to the ninth century, whereas Paramasivan and the Tribune give the seventh. Impart’s range of 500 to 800 covers the cave and its murals together.

Accounts of the technique differ as well. Fresco-secco, paint laid on a dry lime-plastered wall, is Swaminathan’s name for it, and he lists Paramasivan’s pigments: lime for white, lamp black, ochres for yellow and red, and terre verte, a green earth, for green. He adds that the Archaeological Survey of India’s information board says vegetable dyes were used. Impart instead describes painting on wet lime plaster, with vegetable and mineral pigments.

The royal couple is disputed too. K. R. Srinivasan, per Puratattva, identifies the pair with Ilan Gautaman meeting Srivallabha and his queen. For Impart, the male figure is generally taken to be either Mahendravarman I, compared with a similar relief at Mahabalipuram, or the Pandya king Srimara Srivallabha.

On style, Swaminathan places the murals after Ajanta’s painted caves (which he dates 2nd century BCE to 6th century CE), the fifth-century paintings at Sigiriya in Sri Lanka and the sixth and seventh century paintings at Bagh in Madhya Pradesh, and before the eleventh-century Chola paintings at Thanjavur. He calls them the only early Jain frescoes known. The Tribune headlines the site as the Ajanta of Tamil Nadu.

The oldest writing on the hill

The hill’s earliest inscription is not in the cave. Eladipattam, a natural cavern on the hill, has 17 polished beds with raised stone pillows, where Jain ascetics practised austerities, according to Puratattva. Inscription No. 1 in K. R. Srinivasa Ayyar’s translated volume of Pudukkottai inscriptions is cut on one of those beds. The volume reproduces H. Krishna Sastri’s note assigning it, on the shapes of its letters, roughly to the end of the third century and the start of the second century BCE. Puratattva places it at the start of the third century BCE, and Impart dates Jain use of the site from about 100 BCE.

Even the language is open. The volume describes the characters as Asoka-Brahmi of a type found in southern caves, but in 1924 K. V. Subrahmanya Iyer argued the inscription is in Tamil and suggested that Citupocil, a name in it, was the old name of Sittanavasal. The same footnote quotes Subrahmanya Iyer saying that, with the single exception of the Sittannavasal epigraph, the Brahmi cave inscriptions of the Pandya country point to the third century BCE. The 1906 to 1907 epigraphy report, quoted in the volume, put the claim conditionally: if the inscriptions and the beds are of the same date, these are the earliest stone records of the Tamil country.

Later labels on the beds give names, probably of Jain ascetics, in characters that the volume dates to about the ninth century and Puratattva to the eighth or ninth. Swaminathan puts the span of Jain use of the hill at 1,200 years, from the third century BCE to the tenth century CE, while Impart gives about 100 BCE to about 900 CE.

Protection today

Puratattva counts two ticketed monuments at Sittanavasal under the Archaeological Survey of India, the cave and the Jain beds, and the Pudukkottai district page also places the cave temples under the Survey. Opening hours and ticket prices are left out here because no official Survey notice for them was found. The Hindu reported, in an article by Nahla Nainar, that the site draws young visitors and school groups and needs better upkeep. In February 2008 Swaminathan wrote that the paintings were barely visible, mainly because of vandalism over the previous 50 to 60 years.