Rajendra Chola's Army Marched to the Ganges and His Fleet Caught the King of Kadaram in 1025. The Navy and the Land Survey Behind Him Were His Father's.

On the ninety-sixth day of the twenty-first year of Rajaraja’s reign, about 1006, a party of witnesses followed an elephant round the village of Anaimangalam in the Pattanakkarru district on the Tamil coast. The royal order told them to “surround the elephant and make it walk over the boundaries, planting stones and kalli shrubs and thus establish the boundaries.” Kalli is spurge, a milky bush much used for fencing. The village was being given to a Buddhist monastery at Nagapattinam, and a foreign king had built that monastery: Sri Maravijayottungavarman, “sprung from the Sailendra family and the lord of the Srivishaya country.” Srivishaya is Srivijaya, the sea power of Sumatra and the Malay coast.

About nineteen years later Rajaraja’s son Rajendra sent ships against that kingdom and captured its king, Sangramavijayottungavarman, whom K. A. Nilakanta Sastri, the historian of the Cholas, treats as a successor of the monastery’s founder, following the epigraphist Eugen Hultzsch. Tuput’s earlier piece, The Emperor Who Turned the Bay of Bengal Into a Chola Lake, is the shorter portrait of the father. This one goes deeper on the son, the Ganges expedition, the raid on Srivijaya and the administration that paid for them.

A Sumatran king’s monastery at Nagapattinam

The charter survives as two sets of copper plates, now back in India (the last section tells how). The Sanskrit part, in the 1886 translation published by James Burgess, traces the Chola line from Manu to Sibi, “preserving his life for others’ sake.” The larger set has 21 plates on a bronze ring that carries Rajendra’s seal. Sastri’s reading is that Rajaraja approved the gift and his son confirmed it after his death and had it engraved. The paperwork ran over two years. The king gave the order on the 92nd day of his 21st year, the record was made four days later, and the execution was completed on the 113th day of the 23rd year. Accountants of 27 villages signed for their assemblies, Sastri counts.

Nagapattinam made sense as a site. Sastri points out that the itineraries of the Chinese pilgrim I-tsing show it as the first mainland port reached by ships from the east. The annals of the Song dynasty record embassies from a country the Chinese called San-fo-tsi in 1003 and 1008. The French scholar Georges Coedès recognised in the Chinese spellings the names of two Srivijayan kings, Chudamanivarman and Mara-vijayottunga-varman, and Sastri accepts his conclusion that San-fo-tsi and the grant’s lord of Srivishaya and Kataha were one kingdom.

Twelve thousand islands, and a navy nobody describes

The Cholas were far older than Rajaraja. Ashoka’s second rock edict, in Dhammika’s translation, lists them with the Pandyas, Satiyaputras and Keralaputras “among the people beyond the borders” of his domain.

Rajaraja’s own list of conquests is cut into the walls of the Tanjore temple. He dedicated the copper pot for its finial on the 275th day of his 25th year, Sastri reports, and UNESCO dates that regnal year to 1009 and 1010. Hultzsch’s translation of an inscription of his 29th year, 1014, says his army won Vengai-nadu, Ganga-padi, Tadigai-padi, Nulamba-padi, Kollam, Kalingam, the country of the Sinhalas, “the seven and a half lakshas of Iratta-padi, and twelve thousand ancient islands of the sea.”

Sastri reads the islands as the Maldives and notes that they appear only in Rajaraja’s latest inscriptions, as the last of his conquests. He found no account of the voyage. He argued all the same that the navy Rajendra used so well was built under the father, and he compared the pair to Philip of Macedon and Alexander.

Sri Lanka is among Rajaraja’s conquests in his inscriptions of 993, the eighth year of the reign, in Sastri’s account. The Sinhalese chronicle never mentions that invasion. Sastri explains that its record of King Mahinda V loses its thread after about 991, when an army revolt drove the king into the hills of Rohana. The Cholas took the north and made it a province, Mummudi-Chola-mandalam, with Polonnaruwa as its capital, renamed Jananathamangalam.

Rajaraja made Rajendra heir in 1012, and Rajendra counted his regnal years from then. Sastri takes the 29th year, 1014, as the end of Rajaraja’s reign, since his Tanjore records stop there.

A survey, a surveyor and an audit

Sastri dates the start of the great land survey to Rajaraja’s 16th year, 1001. The man who carried it out was a general, Kuravan Ulagalandan, who apparently earned his surname, “one who measured the world,” from the work. The survey fed a register of land rights and dues, the varippottagam, which trained officials kept up to date. Sastri defended it as a careful record of rights and denied that it was a tool of extortion.

He also gave the other side. No inscription states the share of the harvest that went to the state. The many extra local levies, he wrote, left the impression that taxation “pressed the people rather hard,” though he found that the tax-payers’ consent was usually sought beforehand.

The king’s officials audited the finances of village assemblies and attended meetings on weighty business, such as a change of rules or an alteration of land rights that touched the revenue. Otherwise, Sastri writes, the assemblies were left to themselves.

Thirty wards, a boy and a pot

Uttaramerur, a village in what is now Kancheepuram district, has two resolutions of its Brahmin assembly, the sabha, cut into stone in the reign of an earlier Chola king, Parantaka I. Sastri says we know the working of no other sabha in such detail. The Cholas did not invent village assemblies, and Sastri notes that a sabha was always the assembly of a Brahmin village, while the ur was the simpler assembly of a village’s residents and a nagaram that of a merchant town.

The first resolution is dated in Parantaka’s 12th year, 919, and the second, a revision two years later by Sastri’s reading, in his 14th year, 921. A royal officer sat with the assembly each time under the king’s letter. The village was divided into thirty wards. The people of each ward wrote on tickets the names of men who qualified, and a boy drew the tickets from a pot. The 919 text says the boy must be one “who cannot distinguish any forms.” The 921 text says one “who cannot see the difference (between things) even by day.”

The age limits changed between the two. In Sastri’s 1932 translation the first resolution admits men “below sixty and above thirty years of age.” The revision says “below seventy and above thirty-five,” which is also the figure in R. Nagaswamy’s translation. ThePrint’s summary, drawing on Sastri, gives 30 to 60, so both pairs circulate. The first pair comes from the 919 resolution and the second from the 921 revision.

Candidates in 921 had to own more than a quarter veli (a land measure) of taxed land, live in a house on their own site, know the Mantrabrahmana, a Vedic text, and have experience of teaching it. A man with only an eighth of a veli could stand if he knew one Veda and could expound one of four commentaries. The rule bars anyone who had served on a committee and failed to submit accounts, together with a long list of his relatives, and also thieves, the violent, “those who rode on asses, and those who forged documents.” Members served 360 days, and any who were found guilty were removed at once.

The Mantrabrahmana rule meant the pool of candidates was learned Brahmins, which fits Sastri’s point that a sabha belonged to a Brahmin village. The 919 resolution promised to hold “as long as the sun and the moon last.” It was rewritten in 921 because, in Sastri’s words, the attempt to give both wards and streets a place on the committees “did not work.”

The pillage of Lanka, in the chronicle’s words

Rajendra completed the Chola conquest of Lanka in about 1017. The Sinhalese chronicle Culavamsa, in Wilhelm Geiger’s translation, gives the account of the people who lost. A horse-dealer from the opposite coast “told the Cola King on his return about the conditions in Lanka,” and the king sent troops. In the 36th year of Mahinda V’s reign the Cholas seized his chief queen, his jewels and his regalia. The king himself had fled to the jungle, and they “captured alive, with the pretence of making a treaty.” The chronicle says they broke open the relic chambers and “violently destroyed here and there all the monasteries, like blood-sucking yakkhas,” a yakkha being a demon. Geiger titles the chapter “The Pillage of Lanka.”

Mahinda “dwelt twelve years in Cola land,” the chronicle says, and died there. Rajendra’s inscriptions, Sastri notes, say nothing of pillage and boast instead of the crown of Ceylon’s king, the queens’ crowns and the Pandya regalia left on the island. Sastri also reports that the Sinhalese rallied round Mahinda’s son Kassapa, who after a six months’ war made the south-east, Rohana, independent again and ruled it as Vikramabahu I from 1029 to 1041.

Water from the Ganges, fetched by a general

Rajendra’s inscriptions of his 12th year name the northern rulers. In the translation hosted by the Dharmalekha project, they are Dharmapala, from whom he took Dandabhukti, Ranasura, whom he attacked for Takkanaladam, Govindachandra, who fled from Vangaladesa “having lost his fortune,” and Mahipala, whom he was pleased to “frighten in a hot battle.” Sastri identifies Mahipala as the powerful Pala king of Bengal.

The Sanskrit Tiruvalangadu plates tell it in verse. In Hultzsch’s translation, the king “wished to sanctify his own country with the waters of the Ganga.” His general went, not Rajendra. The general “reached the celestial river,” and the conquered chiefs on its banks carried “the most sacred waters” back, while Rajendra himself “reached the river Godavari.” The king then built a tank in his own dominions “as a pillar of victory,” called Cholagangam, “composed of the waters of the Ganges.”

The two records disagree on the order of the victories. Sastri prefers the Tamil record, which grew with each conquest until the 13th year and was cut almost at once, and notes that the court poet Narayana wrote the Sanskrit account after all the conquests were over. He judged the march “a hurried raid across a vast stretch of country,” and the story of defeated kings carrying the water a boast, though he had no doubt of its essentials. He rejected the older view of the epigraphist Venkayya that it was a pilgrimage. For Sastri it was a show of power in the tradition of the conqueror’s tour of the quarters, the digvijaya.

He dates the march to about 1022 and says it lasted less than two years, since inscriptions of Rajendra’s tenth year omit it and one of the eleventh mentions it. Britannica gives 1023. Rajendra took the title Gangaikonda Chola, meaning the Chola who took the Ganges, and gave it to a new capital, Gangaikondacholapuram. His own records mention the city rarely and not before his 17th regnal year. UNESCO says the temple there was completed in 1035 and gives its tower a height of 53 metres.

Kadaram, 1025

Rajendra’s own account of the sea raid is cut into the Tanjore temple wall, dated the 242nd day of his 19th year, about 1030. Sastri puts the raid itself at about 1025. In Hultzsch’s translation, he, “having dispatched many ships in the midst of the rolling sea,” caught “Samgramavijayottungavarman, the king of Kadaram, along with (his) vehicles, (viz.) rutting elephants.” The loot included a Vidyadhara-torana, an arch at the enemy city’s war-gate. The inscription then names places one after another: Pannai, Malaiyur, Mayirudingam, Ilangasogam, Talaittakkolam, Ilamuri-desam, Manakkavaram and Kadaram itself.

For decades nobody could place them. Hultzsch, who published the text in 1891, first looked for Kadaram in the Madura district, and for some years the raid was read as a war on Pegu in Burma. Burmese archaeologists announced two granite pillars near Pegu as Rajendra’s victory pillars. In 1918 Coedès showed that the places belonged to Srivijaya, whose centre was Palembang in Sumatra. Pannai lay on Sumatra’s east coast and Malaiyur at the southern end of the Malay peninsula. Ilamuri-desam was Lamuri in northern Sumatra, Manakkavaram the Nicobar Islands and Kadaram Kedah, on the west coast of the peninsula. Burma’s archaeology department later took the pillars off its list of protected monuments.

Sastri suggests the Cholas called the Srivijayan ruler the king of Kadaram because Kedah was the first place their ships reached, as I-tsing’s itineraries imply. He doubted that the raid achieved more than a vague acknowledgement of Rajendra’s overlordship, and found no sign that the Cholas tried to rule the lands. In 1068, Sastri reports, Rajendra’s son Virarajendra claimed to have conquered Kadaram again for a ruler who begged for his protection.

Why the fleet sailed

No Chola record gives a reason, Sastri observes. He offers two, either that Srivijaya obstructed Chola trade with the East or, more probably in his view, that Rajendra wanted a digvijaya across the sea. He adds that Rajendra confirmed the Anaimangalam grant at the start of his reign, so relations began friendly.

In Kulke’s summary, George Spencer reads the raid as one more case of the Chola politics of plunder, a thesis he set out for Ceylon in a 1976 article arguing that organised plunder was a more important tool of the traditional Indian state than historians had allowed. Tansen Sen argues for trade. In his chapter in the 2009 volume Nagapattinam to Suvarnadwipa, Sen writes that Rajendra attacked Srivijayan ports twice, in 1017 and more widely in 1025, in answer to Srivijayan interference in the direct trade between southern India and Song China. Kulke adds that Sen points to Srivijaya’s habit of forcing foreign ships into its ports.

The Chinese court kept its own record of the trade. Johannes Kurz notes that the Song huiyao, the Song dynasty’s collected documents, dates a mission from Zhunian, the Chinese name for the Chola country, to 17 October 1015. The chief envoy was Suolisanwen, and on 18 December 1015 he sent birthday greetings to Emperor Zhenzong. Sastri dates the first Chola embassy to the same year and says the Chinese named its king as Rajaraja, so the first mission went out in the father’s name. Later missions followed in 1033 and 1077. The 1077 embassy of 72 men went home with 81,800 strings of copper cash for tribute goods that included glassware, camphor, ivory and cloves. Sastri called it “clearly a trading venture.”

Where the plates went

The two sets of plates reached Leiden University in 1862. The Colonial Collections Committee, whose investigation Leiden had requested, found it likely that the Dutch East India Company dug them up between 1687 and 1700, while building Fort Vijf Sinnen and redeveloping the site of the “Chinese” pagoda at Nagapattinam, and that the Camper-Kettinghs family took them to the Netherlands in 1712. Florentius Camper was a pastor in Batavia. India asked for the plates back in the summer of 2023. The committee concluded that removing them from Nagapattinam without the consent of the rights holders was “an involuntary loss of possession” and advised unconditional return. The British side of the story, thousands of Indian objects in British collections, is in what Britain took home from India.

Leiden’s catalogue numbers are Or.1687, the 21 plates with Rajendra’s seal, and Or.1688, three plates with the seal of Kulottunga I. Sastri explains the smaller set. About 1090 the king of Srivijaya sent two envoys to Kulottunga and asked for a charter listing the villages that Chola kings had given to the two monasteries the Sumatran line had built at Nagapattinam. One was named after Rajaraja and carried the alternative name Sri Sailendra-Chudamanivarma Vihara, so the monastery of 1006 was still being endowed, and its grants confirmed, some 65 years after Rajendra’s raid.

The Sanskrit grant of the larger set ends with a wish: “So long as Sésha, the king of all other serpents, holds this entire earth so long may this Vihara stand with power on the earth.” Leiden presented the plates to India on 16 May 2026, with the prime ministers of India and the Netherlands present, and its release records that by 7 September the university had handed them over. They go to the Archaeological Survey of India in New Delhi, and together the two sets weigh 30 kilos.