Is there really a lake under the Paris Opera?

Where does the lake in The Phantom of the Opera come from?

From a newspaper serial. Gaston Leroux began publishing the novel in instalments in a Parisian paper in 1909, and opened it with a sly preface: the narrator swears the Phantom was real, body and all, and offers as proof a corpse unearthed in the theatre's cellars while workmen were digging a place to bury recordings of singers' voices.

Beneath the stage of the novel lie a lake and a boat. From the water a secret passage rises to the rue Scribe, ending behind an iron gate whose key belongs to Christine, the soprano at the heart of the story. When Raoul, the young man who loves her, goes searching frantically for that gate and ends up asking the concierge, she answers quite calmly that she knows perfectly well the Opera has a lake beneath it, though she's never been down there. And she laughs.

That reply says it all: in the fiction, the lake is common knowledge. The novel was leaning on something real. There was water under the building, and its story had begun almost half a century earlier.

What's really under the Palais Garnier?

Sand and water. Garnier won the 1861 competition at thirty-five, with no major building to his name, and the ground gave him his first headache before a single stone was laid: firm sand shot through by a large body of water that, to make matters worse, was on the move. Some took it for the old Ménilmontant stream. Garnier said no: it was the groundwater that supplied the wells of the surrounding houses. In his book about the Opera, published in 1878, he recorded that there had been water down there and that there still was.

To give the pit beneath the stage the depth it needed, the foundations had to sit about five metres below the average level of that water. If it got in, Garnier warned, the cellars would turn into mushroom beds and bogs. So he devised a double-walled container, like one bathtub set inside another and held apart by supports: anything that seeped through would gather in the space between, where pumps could draw it off. He named it the cuve, French for a vat or tank, and even bothered to defend the word: he knew a vat normally holds its water inside, whereas his existed precisely to keep it out.

Building it took months of non-stop pumping, day and night. The pumps lowered the neighbours' wells, and Garnier had to install fountains fed by the municipal supply to silence the complaints. He also made several quiet visits to the houses round about, afraid the ground might sink beneath them. By his calculations, the water removed would have filled a block with the footprint of the Louvre's courtyard and twice the height of the towers of Notre-Dame.

Why did Garnier fill the tank with water?

To cover up a failure. He admitted as much himself two decades later. On the day the pumps were switched off, as evening fell, he went to the site on his own and, peering over the edge of the excavation, made out puddles on the concrete floor. He writes that a violent shock ran through him. He drove off in a carriage to fetch the contractor, and first thing next morning the two of them found that the water had risen. There were leaks.

In Paris there were people only too keen to broadcast his setback. So Garnier had the pumps restarted, except that the water drawn from the wells, instead of ending up in the sewers, went straight into the tank itself, in front of everyone, with the site gates thrown open. He told all and sundry, his assistants included, that the trial had been a triumph and that the hydraulic lime needed at least a year under water to gain strength. Only his friend Louvet knew the truth. The upshot, he recounts with some amusement, was that everyone praised his cleverness.

While the water hid the problem, Garnier could inspect every joint and plug every leak at his leisure, until the tank was perfectly watertight. And he confesses it plainly: he did it "pour cacher une défaite partielle", to hide a partial defeat.

There's a nuance that hardly ever gets mentioned. According to Garnier, even if the test had gone well he'd have flooded the tank, at least partly, as a precaution. The sand showed that the water had once climbed three metres above its average level, and water inside helped to counter the push from outside. The difference, he adds, is that he'd have done it after confirming success, not to disguise a setback.

What is the tank used for?

It's worth keeping apart what the architect said and what's said now. Garnier wrote that if any water remained in the tank, he'd let people believe it was being kept deliberately for putting out fires. And he granted that it wouldn't be entirely far-fetched: the idea had often been raised, and he didn't rule out the tank one day serving as a great reserve in case the city's water supply failed.

There's a widespread claim that the tank is still full, that firefighters draw on it and even that fish live in it. We haven't found a reliable source to back that up. Watch the dates too: one official tourism website has the lake coming into being in 1881, when the groundwater was pumped out in 1862.

And the phantom? What's fiction

The Phantom is Leroux's invention, and there's no evidence of any masked man living in the theatre's cellars. Still, the novel likes to latch on to real events and inflate them. Its famous chandelier plunging onto the audience grew out of a genuine accident: in May 1896, when a cable broke, one of its counterweights, around seven hundred kilos, smashed through the ceiling of the highest boxes and killed a member of the audience, Madame Chomette. The chandelier stayed where it was. Leroux brought the whole thing down.

So if anyone asks you about the phantom's lake, you can answer like the concierge in the novel: of course there's water under the Opera. It's just that the first tall tale about it was signed by the man who shut it in down there.

Fact or legend.

• Fact: a layer of groundwater lies beneath the Opera, and Garnier gave the foundations a double-walled tank, which he filled with water.
• The architect's own word: the failed test, the lime excuse and the cover-up. Garnier wrote it down some twenty years later, and he's the only source.
• Fiction: the lake with its boat, the passage to the rue Scribe and the Phantom, all from Leroux's novel (1909–1910).
• Unconfirmed: that the tank is still full, used by the fire brigade and home to fish.
• Myths doing the rounds: the lake doesn't date from 1881, and it isn't the old Ménilmontant stream, which Garnier himself ruled out.

Where to hear it

Podtour tells this story on the route "From Vendôme to the Opera", which ends by circling Garnier's building. The novel is told by the building's side gates, at the stop "The Opera gates where the phantom began", and the real water in front of the main facade, at "The lake and Garnier's story". Madame Chomette and the chandelier get a stop of their own: "The accident that became a falling chandelier". If the Paris beneath the streets appeals, read how the Paris Catacombs began. And if you're planning your trip, our Paris guide has the practical side covered.

Further reading

• Charles Garnier, Le Nouvel Opéra de Paris, vol. I, Paris, Ducher et Cie, 1878 (chapter "De la construction de la cuve", pp. 217–228). https://archive.org/details/lenouveloprade01garn
• Gaston Leroux, Le Fantôme de l'Opéra (serialised 1909–1910; 1926 edition on Wikisource). https://fr.wikisource.org/wiki/Le_Fantôme_de_l’Opéra
• Musée d'Orsay, "Parcours: L'Opéra de Charles Garnier". https://www.musee-orsay.fr/fr/ressources/parcours-lopera-de-charles-garnier-80
• Bibliothèque municipale de Lyon, Guichet du Savoir, answer on the fall of the Opera chandelier, 2015. https://www.guichetdusavoir.org/question/voir/48586