A 20-Metre Ancient Wall Under Paris May Shed Light on Lutetia
Archaeologists have uncovered a stone-and-wood wall beneath the Hôtel-Dieu AP-HP on Paris's Île de la Cité, near Notre-Dame Cathedral.
The wall beneath the hospital
The structure dates to the Gallo-Roman period and was found under a hospital site on the island, according to CBS News. It survives along about twenty metres, Inrap reports. The City of Paris describes the remains as a stone wall with burnt wood.
The discovery was made in summer 2026. It came during preventive excavations, work done before construction or renovation, carried out since February 2026 by the Ville de Paris archaeological unit and Inrap in two courtyards of the Hôtel-Dieu, the City reports.
Lutetia and the island digs
The structure is believed to be a fragment of the ramparts of Lutetia, the ancient name for Paris. The ramparts were built by the Gallic Parisii, according to The Guardian. French archaeological authorities describe Lutetia as an oppidum, a fortified town, of the Parisii people on an island in the Seine and mentioned in Caesar's Gallic Wars.
The City of Paris says the Hôtel-Dieu remains could add a great deal to understanding of Lutetia. Gallic defensive walls still in place are rare on the island, like pages in a book that are seldom preserved. Buried layers are also hard to reach beneath hospital, religious and court complexes used and rebuilt over centuries.
The find sits inside a concentrated 2026 program on the Île de la Cité. Excavations on the parvis, the open square in front of Notre-Dame, began in January 2026. Teams are also digging beneath Notre-Dame Cathedral itself to explore remains dating back 2,000 years to Roman Paris. The City has scheduled the final excavation phase at the Cour du Mai of the Palais de Justice to begin in fall 2026.
The broader context here is island topography and control of the Seine crossing. A fortified edge on the Île de la Cité would fit the oppidum model drawn from ancient texts and earlier scattered observations. A continuous twenty-metre line offers measurable detail on stone courses, timber supports, foundation work and later burning. Comparison with sequences from the parvis and subsoil layers under the cathedral will be central.
In my view, the perimeter cannot yet be redrawn. Twenty metres is substantial in this setting, but it remains a fragment. Phasing, dating and the link to later Roman masonry still need to be established through dig records, finds and lab dating such as radiocarbon or tree-ring dating where wood survives. The next openings at the parvis and Cour du Mai may test for continuity, turns or gates. Until then, the Hôtel-Dieu rampart is best treated as a fixed point for hypothesis, not a closed circuit.
