Lost Worlds · Poompuhar, Cauvery delta, Tamil Nadu, India

Poompuhar (Kaveripattinam)

The drowned Chola port of Tamil epic — and an offshore horseshoe of stone that became a flashpoint in the debate over Ice Age lost cities.

Mainstream: c. 300 BC – AD 300 (Sangam-era Chola port), with offshore finds into the early medieval periodAlternative: c. 9600 BC (Hancock's proposed age for a submerged 'U-shaped' structure)11.14°, 79.86°

Lost Worlds Myth Catastrophe

At a glance

Poompuhar (Kaveripattinam) — Photo: Ssriram mt (Wikimedia Commons) · CC BY-SA 4.0

Poompuhar, also called Kaveripattinam or Puhar, sat where the Cauvery River met the Bay of Bengal and was the celebrated port capital of the early Cholas, immortalised in the Tamil epics Silappatikaram and Manimekalai as a wealthy, cosmopolitan city that was ultimately consumed by the sea. Both onshore and offshore archaeology confirm a genuine ancient port here, and the National Institute of Oceanography (NIO) has recovered structures from the seabed a few kilometres out. The site sits at the heart of two overlapping stories: a solidly attested Sangam-era trading city, and the much larger claim — tied to the mythic lost land of Kumari Kandam — that some submerged remains here are many thousands of years old.

See it on the globe →

The mainstream view

What archaeology says

Mainstream archaeology recognises Poompuhar as a real and important early historic port. Excavations on land and in the intertidal zone have yielded wharves, brick structures, ring wells, Roman and Chola artefacts, iron objects and Tamil-inscribed pottery, with pottery offshore dated to around the fourth century BC. This fits the literary and historical picture of a flourishing Chola emporium of the Sangam age (roughly 300 BC to AD 300) that traded with Rome, Southeast Asia and beyond, and that later declined as the coastline and river mouth shifted.

The NIO's marine surveys from the late 1980s onward, led by figures associated with S. R. Rao, mapped submerged objects offshore, including a much-discussed 'U-shaped' or horseshoe-shaped structure lying at about 23 metres depth several kilometres out. Marine geologists interpret such features cautiously: some are man-made harbour works or building foundations of the historic port drowned by erosion and sea-level change, while others may be natural.

Crucially, the mainstream rejects any Ice Age dating for the structures. Standard sea-level curves for the Bay of Bengal do put the 23-metre depth contour above water thousands of years ago, but that alone does not date a structure — a later building can stand in deep water if the seabed subsided or the object slid, and no material recovered from the U-shaped feature has been independently dated to the terminal Pleistocene. For orthodoxy, Poompuhar is a genuine drowned port measured in centuries and low thousands of years, not a relic of a vanished antediluvian continent.

Key evidence cited

Onshore and intertidal excavations of wharves, brick structures, ring wells and Roman/Chola artefacts
Offshore pottery dated to around the fourth century BC, consistent with a Sangam-era port
Literary attestation in Silappatikaram and Manimekalai of a real Chola port lost to the sea
NIO marine surveys mapping submerged structures interpreted as drowned historic harbour works
Sea-level and subsidence arguments showing depth alone cannot date an underwater structure

The alternative view

What the skeptics propose

The alternative case, most prominently argued by Graham Hancock, seizes on the offshore U-shaped structure and on Tamil traditions of Kumari Kandam — a vast southern land said to have been lost to successive floods. Hancock read about the NIO finds while in London, sought out S. R. Rao, and in 2001 arranged diver-led investigations funded by Channel 4 and the Learning Channel. Reasoning from the depth of the structure (about 23 metres) and from Bay of Bengal sea-level curves, he argued the feature would last have stood on dry land roughly 11,000 or more years ago, making it a candidate for a genuine Ice Age construction — potential support for his lost-civilisation thesis in Underworld.

Supporters emphasise that the structure is a real, surveyed object of apparently regular, horseshoe-like plan, singled out by professional oceanographers rather than invented by enthusiasts, and that its depth genuinely implies great antiquity if it was built on land. They fold in the Kumari Kandam tradition and other Tamil flood legends as cultural memory of drowned coastlines, arguing that South India's submerged shelf deserves the kind of sustained survey rarely funded.

The vulnerability of the argument is the same as at other Hancock sites: the depth-implies-age reasoning assumes the object was constructed on the surface at that spot and never moved or subsided, and no dateable material has tied the U-shaped feature to the Pleistocene. Sceptics also note the feature could be natural or a later structure. Even sympathetic marine scientists have treated the 11,000-year claim as unproven. Poompuhar thus stands as a real drowned port whose fringe reading rests on an inference about sea level that the recovered evidence has not yet confirmed.

Key evidence cited

A surveyed 'U-shaped' / horseshoe structure at ~23 m depth several kilometres offshore
Bay of Bengal sea-level curves placing that depth above water many thousands of years ago
The Kumari Kandam tradition and Tamil flood legends read as memory of drowned southern land
The feature was singled out by NIO oceanographers, not by amateurs (Hancock's emphasis)
The near-total absence of deep, sustained survey of South India's submerged continental shelf

Genuinely open questions

Is the offshore U-shaped structure man-made, and if so from what period?
Can any material recovered from the deep structures be independently dated?
How much of the historic port drowned through erosion and delta shift versus tectonic subsidence?

Worth knowing

Poompuhar's destruction is written into Tamil literature itself: the epics describe the great port being swallowed by the sea, so the archaeology here is, unusually, chasing a catastrophe that classical poets had already recorded.