Kallanai, a 329 m Weir of Rough Stone on the Kaveri Credited by Tradition to Karikala Chola, Still Splits the River Into Four Channels for the Delta and Was Recognised by ICID in 2021

The Grand Anicut at Kallanai, on the Kaveri (Cauvery) near Tiruchirappalli in Tamil Nadu, is a weir of rough stone about 329 m long. It was in service before the British arrived, and the International Commission on Irrigation and Drainage (ICID) recognised it as a World Heritage Irrigation Structure at its 72nd International Executive Council meeting in Marrakesh in 2021. Tradition credits it to the Chola king Karikala and places it in the second century CE.

That date is tradition. No source we could open quotes an inscription or a dated excavation that records the building of this weir, and the evidence for what happened to it afterwards is much better than the evidence for how it began. This piece sets out both.

What the weir does

An anicut is a dam across a river, and Bijker notes that the word comes from the Tamil anaikattu. The Kaveri splits near Tiruchirappalli into two branches around the island of Srirangam, the Kollidam to the north and the Kaveri to the south. In Wiebe Bijker’s 2007 account of Chitra Krishnan’s doctoral study of the weir, the Kollidam is the wider, steeper and faster branch, a flood carrier that delta farmers barely used. Almost all of the roughly 600,000 acres the river irrigated in 1800 lay south of the Kaveri branch, so that branch was the farmers’ lifeline.

Bijker says Krishnan reconstructed the Kallanai as a flood-diverting weir. When the river rose above its crest, a large part of the water went over into the Kollidam through a short connecting stream and ran to the sea, and the delta’s fields were spared. Hemingway’s 1907 gazetteer for Trichinopoly district describes the same arrangement in a colonial official’s words: the Grand Anicut sits across the channel that would otherwise carry Kaveri water into the Kollidam and starve the Tanjore delta.

Bijker’s figure caption puts the weir 28 km downstream of the point where the river divides. ICID’s entry says the purpose is to divert the Kaveri’s water across the Thanjavur delta by canals, and DT Next’s 2018 report says the weir splits the river into four branches, the Kollidam, Cauvery, Vennaru and Pudhu Aaru. Further downstream the Kollidam also feeds the Chola-period Veeranam tank through the Lower Anicut.

How big it is, in two sets of units

Hemingway’s Tanjore and Trichinopoly gazetteers give the old dam as a mass of rough stone set in clay, 1,080 feet long, 40 to 60 feet wide and 15 to 18 feet deep, running across the outlet in a serpentine curve. In metric that is about 329 m long, 12 to 18 m wide and 4.6 to 5.5 m deep. DT Next and one paragraph of ICID’s entry give 329 m long, 20 m wide and 5.4 m high, which matches the gazetteer on length and height but not on width.

ICID’s page also carries a second set of figures, “more than 300 m long, 20 m wide, and 4.5 m high”, the wording in Bijker’s article. Both heights, 4.5 m and 5.4 m, sit on the same page. Raised and re-faced more than once, the weir has had different dimensions at different moments.

Who is said to have built it

Tradition names Karikala, the Chola king of the Sangam age, the period of the earliest Tamil poems (see our piece on Sangam literature). K. A. Nilakanta Sastri’s standard history of the Cholas, in its 1955 edition, takes as a working hypothesis that the Sangam age falls in the first three or four centuries of the Christian era. The Trichinopoly and Tanjore gazetteers, both by F. R. Hemingway, say Karikala’s reign “has been assigned” to 50 to 95 CE. ICID, DT Next and Bijker give the second century.

What the gazetteers say about the weir itself is more cautious. Hemingway writes that the Grand Anicut was built “perhaps by the Chola kings” and that its date “is not exactly known”. On Karikala, he notes that sluices and canals from the Kaveri existed in his time and that it was probably he who built them and raised banks against floods, because the embanking of the river is mentioned in the Leiden copper-plate grant of Rajaraja I (reigned 985 to 1013) and in records of the Telugu Chodas.

Those records are late, and Hemingway says so in a footnote: the inscriptions show the tradition was known as early as the eleventh century, but “cannot be represented as contemporary history”. Sastri places the first mention a little earlier, in the Malepadu plates of Punyakumara, a Telugu Choda king of the seventh or eighth century. He lists the raising of the Kaveri’s flood banks among the stories that gathered around Karikala in later centuries, and says others in that group, such as the state elephant choosing him as king, find no support in the earliest authorities. Sastri adds that Karikala is said to have promoted the clearing of forest land and added to the country’s prosperity through irrigation tanks.

So the record has three layers. A Sangam-age king of that name is attested in early Tamil poetry. A later tradition says he embanked the Kaveri, first found in inscriptions of the seventh to eleventh centuries. Modern writing makes the step from there to this particular weir at Kallanai. British engineers hedged too. In Lady Hope’s 1900 biography of Cotton, her summary of the argument between Madras and Bengal engineers over weirs on sandy riverbeds has the Madras side say the Tanjore work was built by the natives “in the second century, as is supposed”. Elsewhere in the book she calls the weir “of unknown age”.

What Cotton found when he cut through it

Arthur Cotton was given charge of the Kaveri irrigation as a captain. His daughter’s biography says work began on 1 January 1830, when his younger brother, Lieutenant Fred Cotton, cut through the Grand Anicut to build sluices. Hemingway’s gazetteer dates the sand-scouring sluices to the same year.

What they found is the best physical description of the original method. In Cotton’s own words, as Lady Hope prints them, the mass “consist[ed] of nothing but loose stone in mud, with the upper course only laid in chunam, and plastered with concrete”. Chunam is lime mortar. He added that the weir had never needed more than occasional renewals of the concrete and had never given cause for alarm. ICID’s entry, which reuses Bijker’s wording, calls the weir one of “unhewn stone” and says it was built on “interlocking technology without cementing material”. That sits uneasily with Cotton’s plastered, mortared top course. Closer to Cotton’s account is the gazetteers’ “rough stone in clay”.

Cotton drew on the method for the Godavari. In a letter on his weir at Dowlaiswaram, reproduced by Lady Hope, he wrote that this was the mode of construction “originally used” at the Grand Anicut, “which has stood for so many centuries”, and he proposed rubble and concrete in place of cut stone, which the district had too few masons to dress. Lady Hope adds that the four anicuts he built across the Godavari are surface coatings of stone over river sand, and says the idea was taken “evidently” from the Grand Anicut. The Godavari works are her father’s, not the Chola-era weir’s, so this shows the lineage of a method, not the date of its first use.

Cotton also used it as an argument. Lady Hope prints his recollection that when he first urged an anicut across the Kaveri, the idea was called visionary. He had thought one built on the Coleroon would end such talk, since people would see that what the natives did hundreds of years ago “with their little science and poor means” could be done by the British too. The Hindu reported in 2021 that a Tamil Nadu government publication on the delta’s history includes his description of the Grand Anicut.

Repairs and rebuilds since 1804

What stands in the river today is a repaired weir. Hemingway says British engineers raised it in 1804, fitted sand-scouring sluices in 1830 and made it the base of a road bridge in 1839. ICID gives the 1804 work as repairs by an officer it names Captain Coldwell, who added dam stones 0.69 m high along the crest and raised the river bank above it, and adds that in 1829 Major Sim proposed under-sluices into the Kollidam to stop silt building up in the upper reaches.

Raising the crest had a side effect. The gazetteer says the weir stopped Kaveri water running into the Kollidam but silted the Kaveri and scoured the Kollidam’s bed, and by 1804 it was clear that silting where the Kollidam leaves the Kaveri was the greater danger to the delta. For some 25 years the bed was dug out and embankments were thrown across the Kollidam’s head, to little effect. Cotton then built the Upper Anicut across the head of the Kollidam at Mukkombu in 1836 to 1838, and Hemingway records that the Grand Anicut was later remodelled with shutters so it could help pass surplus water into the Kollidam.

Bijker’s reading of Krishnan’s work is that the original shape did much of the work: a curved masonry section, a sloping crest and an irregular drop from front to back. She conjectures that the first British changes after 1801, made without grasping how the weir handled sediment, made silting worse. Her reconstruction was tested in hydraulic scale models, which Bijker says supported her ideas on the benefit of curved anicuts over straight weirs.

The acreage figures

ICID says the structure initially irrigated about 69,000 acres, a number DT Next repeats, and that it now irrigates “close to 1 million acres”, then lists the area as 13,20,116 acres, which is 1.32 million. Bijker’s 600,000 acres is the delta’s irrigated land in 1800, before the British works. These figures count different things, and none comes with a survey behind it that we could open.

What passes the weir now

DT Next reported on 16 June 2025 that the Chief Minister opened the Grand Anicut on 15 June, three days after Mettur dam’s release on 12 June, and that the water was expected to serve about 13 lakh acres (1.3 million) of paddy, the kuruvai and samba crops, across the delta. Releases began at 1,500 cusecs each into the Cauvery and Vennar, 500 into the Grand Anicut canal and 400 into the Kollidam. A cusec is a cubic foot a second, so 1,500 cusecs is about 42 cubic metres a second. We could open no operating report later than June 2025. How the river’s water is shared between states is covered in how India settles river water disputes.

Public Works Department engineers quoted by the Deccan Chronicle in 2018 put the capacity of the Cauvery and Vennar at 12,000 cusecs, the Grand Anicut canal at 3,500 to 4,000, and the Kollidam, as flood carrier, at up to four lakh cusecs, about 11,300 cubic metres a second. That paper adds that the state installed a bust of Arthur Cotton at the Grand Anicut in 2005, and its photo caption places a statue of Karikalan in a memorial hall at the site.