The Tamil-Brahmi Inscriptions at Pugalur Name Three Generations of a Chera Royal Family, Cut at a Jain Monk's Rock Shelter About the Second Century CE and Read by Iravatham Mahadevan in 1965

Two inscriptions cut into the rock of Arunattarmalai hill at Pugalur, near Karur in Tamil Nadu, name three generations of a Chera royal family. The better preserved of the two says that a rock shelter was carved for a senior Jain monk, Cenkayapan of Yarrur, when Ilankatunko, son of Perunkatunkon, son of King Atan Cel Irumporai, became heir apparent. The script is Tamil-Brahmi, the early script of the Tamil country, and the language is Tamil.

Iravatham Mahadevan, whose 2003 book Early Tamil Epigraphy edits the early Tamil inscriptions, dated both to about the 2nd century CE. The Archaeological Survey of India gave the same reading and date in a post on 4 June 2025. This article sets out the words, how they came to be read correctly, what the Sangam poems add, and where the dating still varies.

A hill above the Kaveri, north-west of Karur

Mahadevan places Pugalur, ancient Pukaliyur, on the south bank of the Kaveri about 15 km north-west of Karur. Overlooking the river stands Arunattarmalai, which has a Subrahmanya (Murugan) temple at its summit. On the southern slope are two caves with Tamil-Brahmi inscriptions on the brow and on stone beds, and a third cave on the northern side has one more bed inscription. A later annual report of the Government Epigraphist calls the same site Velayudhampalayam, the name of the village at the foot of the hill.

The hill holds twelve inscriptions, numbered 61 to 72 in Mahadevan’s corpus. That is 12 of the 21 Tamil-Brahmi inscriptions he lists from the western, Chera region of the old Tamil country. The stone beds are beds carved into the granite, with a raised pillow end in the words of the Jaina Inscriptions in Tamilnadu catalogue. The Kerala Council for Historical Research notes that rock shelters across the region have such beds, donated to Jain monks. For another Jain cave in Tamil Nadu, with a repair inscription of its own, see our article on Sittanavasal.

Mahadevan says the Irumporai line of the Chera dynasty ruled from Karur, and one stone-bed inscription at Pugalur names the city as Karu-ur. The Tamil Nadu archaeology department’s 2011 report says scholars hold different views on identifying Karuvur with Vanci, the Chera capital. It adds that the department’s own excavation at Karur in the 1990s found Tamil-Brahmi inscribed potsherds, assignable to the beginning of the Common Era, together with Roman amphorae and rouletted ware, a pottery of Mediterranean origin.

Misread for almost four decades, then read in an afternoon

The Government Epigraphist’s report for 1927 to 1928 listed the Pugalur inscriptions and called them mostly damaged. Of the longest, it said the rock was too weather-beaten to be read in full, and it took a name in it for that of a Buddhist monk. Mahadevan notes that the report’s number for the record, 348, was a misprint for 349, and that the one word it read correctly, Atan, was wrongly taken for a monk’s name. A second record found in 1963 to 1964 was also misread, as the cutting of an overhanging rock by one Kadungon Elangon Kolan.

Mahadevan, whom Prof. K. A. Nilakanta Sastri had urged in 1961 to take up the cave inscriptions, first came to Pugalur in February 1965. He reached the western cave late in the afternoon, when the low sun lit the brow of the cave from the side and picked out the cut letters. He had come expecting a damaged or fragmentary record. The inscription, in his account, was weathered but otherwise undamaged, and as he read the words ko, atan and irumporai he saw that it was the record of a Chera king of the Irumporai line that ruled from Karur in the Sangam age.

The upper inscription on the same brow is fainter. With some effort nearly all of it could be read, and it turned out to repeat the lower one, so the few letters lost to flaking could be restored from the lower text. A preliminary report by Mahadevan appeared as “Chera Inscriptions of Sangam Age” in The Hindu Weekly Review on 15 March 1965. It aroused much scholarly interest, he writes, because no inscriptions of the Sangam age were then known to exist. Fuller readings followed in his Corpus of the Tamil-Brahmi Inscriptions (1966, published 1968) and, revised, in Early Tamil Epigraphy in 2003. That book reproduces tracings, an estampage (an ink impression of the stone) of number 62 and a computer-enhanced colour photograph of number 61, whose own estampage is too faint to read, so readers can check the letters without climbing the hill.

What the four lines say

The lower inscription, number 61, runs over four lines in Mahadevan’s transliteration: muta amannan yarrur cenkayapan urai-y, ko atan cel-l-irumporai makan, perunkatunkon makan, lan-katunko ilanko aka arutta kal. He translates it as: the abode of the senior Jaina monk, Cenkayapan of Yarrur; the rock shelter was carved when Ilankatunko, the son of Perunkatunkon, the son of King Atan Cel Irumporai, became the heir apparent.

A few words need a gloss. Amannan is the Tamil word for a Jain monk, and muta marks him as senior. Ko means king, and added to the end of a name it means prince, so ilanko is the young prince, the heir. Kal is the rock, and the verb arutta means carved. Yarrur, Mahadevan suggests, is the modern Attur near Karur. The Jaina Inscriptions in Tamilnadu catalogue gives the same sense for both records and remarks on the three generations of grandfather, father and son.

The second inscription, number 62, differs in small ways. It spells the first king’s name Irumpurai instead of Irumporai, and it calls the prince Katunkon Ilankatunko. The Irumpurai spelling also occurs on some of the copper coins of the dynasty found at Karur, Mahadevan notes, and R. Nagaswamy, writing about one such coin, reads the two spellings as showing that porai and purai were both in use at the time. In Mahadevan’s judgement the upper inscription was cut later than the lower one, when the drip ledge above the cave mouth was raised. Another bed inside the cave carries Cenkayapan’s own name, which he takes to mark the monk’s seat.

Three names, three kings of the poems

The Patirruppattu, the Sangam anthology of ten sets of ten poems with each set praising one Chera king, survives with eight of its ten sets. The introduction to the Auvai Duraisamy Pillai edition, by T. V. Sadasiva Pandarathar, says the first and last are lost and that the surviving eight praise five kings of the Utiyan line and three of the Irumporai line. Our overview of Sangam literature explains where the anthology sits among the eight collections.

The Irumporai line fills the seventh, eighth and ninth sets, and the introduction makes the three heroes a grandfather, his son and his grandson: Selvakkadungo Vazhiyathan, Perunceral Irumporai, who is called the conqueror of Takatur, and Ilanceral Irumporai. R. Panneerselvam, in a paper of 1968, set the inscription’s three names beside these three. Mahadevan followed him in 1971 and sets the match out in his 2003 book:

• Ko Atan Cel Irumporai in the inscription corresponds to Selvakkadungo Vazhiyathan of the seventh set.
• Perunkatunkon corresponds to Perunceral Irumporai of the eighth set.
• Ilankatunko corresponds to Ilanceral Irumporai of the ninth set.

His argument is that Atan, though common in the Chera family, is borne by only one ruler of the Irumporai line, and that the prefixes peru, meaning great, and ila, meaning young, appear in the poems’ names exactly as in the inscriptions. On his reading each prince took the royal title, Perunceral Irumporai and then Ilanceral Irumporai, on coming to the throne. The case is cogent, he writes.

The inscriptions give names and a family relation, and the poems give deeds. Perunceral Irumporai is famous, in Mahadevan’s summary, for defeating Atiyaman Neduman Anji of Takatur, whose own inscription at Jambai is number 59 in the same corpus. Mahadevan also retells a poem in which a poet climbed unaware onto the platform of the royal drum and fell asleep, and the king fanned him instead of having him punished. The match gives him one of the most important fixed points for dating the poets and princes of the Sangam age, from about the middle of the 1st century to the middle of the 3rd century CE.

Others named on the same hill

Most of the other Pugalur records are gifts by townspeople and traders. One reads, in Mahadevan’s translation, “the seat of Natti, the gold merchant from Karu-ur”, and he marks that reading of the town’s name as uncertain. Another records the seat of Venni Atan, the oil merchant. Madras Musings reports that Mahadevan was delighted to find the trade word for an oil merchant here, since the Jain beds at Alagarmalai near Madurai had already shown salt, jaggery, ploughshare and cloth traders as donors.

Numbers 65 and 66 record gifts by Kiran Korri, daughter of Pittan (Pittantai in the second) of Nalliyur, and number 67 names Ilavan, son of Korrantai, as donor. The Sangam poems know a Pittan, chieftain of Kutiraimalai and commander of the Chera army, and his son Korran, Mahadevan notes. He calls the match circumstantial, since the inscriptions give only the names, and observes that these three records, like the royal ones, show three generations of one family.

Nagaswamy has also tied the hill to a coin. A copper coin bearing the Tamil legend Kollip-purai, which he reads as the Chera ruler of Kolli hill, was found at Karur, and he argues that it was issued by the Perunkatunkon of the Pugalur record. Mahadevan dates such coins from the Amaravati river bed at Karur to about the 1st century CE and calls them Irumporai issues, but says that without more specific evidence they cannot be assigned to a particular ruler. The trade road that joined Karur to the west coast through the Palghat gap is described in our article on Kodumanal.

How old are the inscriptions

Mahadevan dates Pugalur numbers 61 to 66, which include both royal records, to about the 2nd century CE on palaeographic evidence, meaning the shapes of the letters. He dates numbers 67 to 72 to the 3rd century. His book divides Tamil-Brahmi into an early period and a late one, and puts these records in the late one, which he places at about the 2nd to 4th centuries CE.

Mahadevan also criticises the dates in the government’s annual reports, which put the 1963 to 1964 finds at Pugalur in about the 1st century CE. In his view those dates were assigned without discussion of the letter forms. Other sources give different figures:

• The Archaeological Survey of India’s 2025 post says 2nd century CE, the same as Mahadevan.
• The Tamil Heritage Foundation’s page on the scripts of the Tamil country also says 2nd century CE.
• The Kerala Council for Historical Research’s field note of 2012 gives about the 1st to 2nd centuries CE for the shelters in general.
• Nagaswamy writes that the inscription is assignable to the beginning of the Common Era.
• The Jaina Inscriptions in Tamilnadu catalogue gives the 3rd to 4th centuries CE for both records.

Published sources therefore range from the beginning of the Common Era to the 4th century. Mahadevan’s century rests on his stated palaeographic reasoning, and the Survey’s 2025 post gives the same one.

The hill today

The rock shelters sit on the sides of the hill, with the modern temple at the top. A team from the Kerala Council for Historical Research visited the site in May 2012 and recorded that the shelters are protected by the state department of archaeology.