The Chapel and the Latrine Were the Same Building

Chapters

1. 0:00 A siege that's been argued about for 800 years
2. 0:52 What Richard actually spent
3. 1:59 The siege begins
4. 3:00 The outer wall comes down
5. 4:07 The eyewitness, word for word
6. 5:15 The same author, writing it twice
7. 6:22 A second witness, in a drawing
8. 7:37 The bridge nobody mentions
9. 8:59 The transferable rule
10. 10:03 Next: the fortress nobody knew was there

Transcript

A siege that's been argued about for 800 years 0:00

On the sixth of March, twelve hundred and four, a castle fell that its own builder had boasted he could hold if the walls were made of butter. It had taken less than two years to put up. It took six months to break. And in the eight centuries since, the argument about that day has not been about the mines, or the engines, or the winter. It has been about a hole in a wall.

Five or six men got inside through one opening, and opened the gate for an army. Two answers to what that opening was have been in print for centuries. The first says it was the castle's latrine. The second says the latrine is a stubborn legend, and that they came through a low window of the chapel. Both sides are quoting the same eyewitness. In the next few minutes that eyewitness is going to settle it — and not by picking a side.

What Richard actually spent 0:52

Start with what was being defended. Château Gaillard was built by Richard the First of England, on a chalk spur above a bend in the Seine, to shut the river road to Rouen. The work ran from eleven ninety-six to eleven ninety-eight — under two years, where a stone castle of that size normally took the better part of a decade. Historians put the bill at about twelve thousand pounds. Across his entire reign, Richard spent roughly seven thousand on every castle he held in England put together. He spent more on this one hill than on a kingdom's worth of stone.

And the site was meant to do most of the work. The rock stands about ninety metres above the river, with a cliff on one side and ravines around the rest. Richard is supposed to have said he could hold it if the walls were made of butter. But the ridge runs more than two hundred metres from end to end, in a thin line of separate enclosures — and to the south-east it is overlooked by a plateau. Flat, higher than the castle, and wide enough to stand siege engines on. That is where the French camped.

The siege begins 1:59

Philip the Second of France arrived at the end of October, twelve hundred and three. He did not attack the rock. He took the island in the river, the town below it, and every outwork around it, until nothing could reach the garrison. One of his men, Gaubert, tied clay pots of fire to his body, swam under water to the palisades on the side nobody was watching, and burned them. Then the French sat down and waited for the winter to do the rest.

What happened over that winter belongs in the record, even though it is the hardest part of it. The castellan, Roger de Lacy, had let townspeople shelter inside, and they were eating the stores. So he put them out. The first groups were let through the French lines unharmed. Then Philip forbade it, to keep the mouths inside. Several hundred people spent the winter trapped between the two armies at the foot of the walls, and most of them died there before the king relented and had the survivors fed.

The outer wall comes down 3:00

In February, twelve hundred and four, the French came up the ridge. The outermost enclosure stood on rock too steep for their ladders — so they cut into the rock itself, with daggers and swords, making holes for their hands and feet. They climbed to the foundations of the main tower. They mined underneath it, packed the cavity with tree trunks, set the trunks alight, and walked away. The tower came down. The garrison then burned everything left in that enclosure, so the French could not use any of it.

That got them one third of the way along the ridge. In front of them now was a very wide ditch, and beyond it a second wall with towers on it, with the survivors shut inside. And here the chronicle does something chronicles rarely do: it names the men who went looking for a way in. Bogis, Eustache, Manassé, Auricus, Grenier. Sergeants, not knights. They walked the outside of that wall in the dark, looking for anything at all. What they found is the thing everyone has argued about ever since.

The eyewitness, word for word 4:07

Here is the eyewitness, as close to word for word as translation allows. On the top of the hill, John had built, the year before, a certain house against the wall, on the right-hand side of the castle, the side that faces south. The lower part of it served for things that want to be put away and kept in a closed room. The upper part, serving the office of a chapel, stood open for the sacraments of the Mass. It had no door on the outside — only two inside, one going up and one going down. And in the lower part, facing outwards, there was a window.

Bogis saw the window. He climbed onto his companions' shoulders, threw himself bodily through it, dropped a rope, and hauled the others up one at a time. Inside, they burned the doors off their hinges to get out. The garrison heard the noise, piled wood against the building and set fire to it to seal them in — and the French came out through the flames with drawn swords, and the whole enclosure caught. By the end of that night, the poem says, Roger had barely one hundred and eighty men still able to fight.

The same author, writing it twice 5:15

Now. That phrase — things that want to be kept in a closed room — is a euphemism, and the man who wrote it knew exactly what he was doing with it. Because he wrote this siege up twice. The version in verse is the celebrated one, the one everybody quotes. He also wrote a prose history of the same reign, and the editor of the standard French edition footnotes the difference: in the prose, he says, the author calls it the latrines outright, and adds that this seemed contrary to religion, given how close the chapel was.

Which dissolves the argument rather than deciding it. The two rival answers are not rivals. They are two floors of one house. John built a single block hard against the curtain wall: the privy on the lower storey, the chapel on the upper, one internal stair joining them, no way in from outside at all except a single window in the lower room. Bogis went through the lower window. Everyone who says the chapel, and everyone who says the latrine, is pointing at the same building, about three metres apart, vertically.

A second witness, in a drawing 6:22

And there is a second witness to that, in a drawing. When Viollet-le-Duc published his measured plan of the castle in the eighteen fifties, he keyed that block into the legend with one letter and one line: H — chapel and latrines, built by John. Not two structures listed side by side. One entry, for one building, from an architect who had walked the ruin with a tape.

So why has the argument survived anyway? Because the poem is what gets read. It is the famous source, and it is coy in precisely the place where being coy causes damage. The prose is blunt and almost nobody quotes it. A polite phrase in a Latin poem became a popular story, and then a correction of the popular story, and both of them were describing one small building that nobody had put back together.

And now the part that ought to be the story, and almost never is. The window is the famous detail of this siege. It is also close to the least important thing that happened in it. A castle of three enclosures had just lost its second one. The strongest position on the rock was still standing, behind its own ditch, and the fight had months of stone left in it, on paper.

The bridge nobody mentions 7:37

It lasted days. Because at the foot of the ditch guarding that last enclosure there was a bridge cut out of the living rock. Not a drawbridge that could be raised. A solid, permanent span, left there deliberately when Richard had the ditches cut deeper. The French slid a siege machine onto it, and then worked underneath it — mining the foot of the wall in shelter, out of reach of everything the garrison could drop or shoot.

The garrison mined back, broke through, and drove the sappers off. But their own tunnel had hollowed the wall from the inside. When a stone-throwing engine put a heavy shot into it, the wall split and a breach opened. The chronicle says this without any softening at all — the work the defenders had done on the inside turned against them. The French climbed the rubble. The castle was taken on the sixth of March.

How many were left is one more place where the sources will not line up. The poem says barely one hundred and eighty men were still able to fight when the second enclosure fell. A modern French account of the surrender lists thirty-six knights and one hundred and seventeen sergeants and crossbowmen. The standard English account says twenty knights and a hundred and twenty other soldiers. Three counts, one garrison, and no honest way to reconcile them from here.

The transferable rule 8:59

There are two rules worth carrying out of this, and they work on any siege. First: before deciding which of two accounts is wrong about a place, check whether they are describing the same structure. Half of the arguments in castle studies are two people standing in one room calling it different names. Second: a fortress does not fail at a point. It fails in sequence, at whatever happens to be weakest when the assault arrives there. The modern archaeological reading of this siege puts it exactly that way — errors in the design of the fortress itself, appearing one after another as the attack progressed.

Three months after the rock fell, Normandy was gone. A duchy held from Rouen since nine hundred and eleven was in French hands, and it did not come back. And the thing almost everybody remembers about the day it started is a window in a toilet. The thing that actually took the castle was a bridge of solid rock, cut by the man who built the place, in the one spot on that ridge where he needed it not to be there.

Next: the fortress nobody knew was there 10:03

Next time, a fortress nobody knew existed. In twenty sixteen, a remote-sensing survey flown over a forest in eastern Poland showed a ring of ramparts under the tree canopy, in the middle of a marsh. It took until the spring of twenty twenty-four before anyone walked it. Two hundred and forty-four objects came out of the topsoil — knives, sickles, arrowheads, a silver ring — from a stronghold that stood from the tenth century to the thirteenth, and was then left entirely alone.

Description and sources

This channel's first entry in Siege Forensics reads one of the most argued-over details in medieval siege history: the opening five or six French sergeants climbed through at Château Gaillard on the sixth of March, 1204, which let the whole second enclosure of Richard the Lionheart's castle fall in a single night.

Half the popular literature says it was the castle's latrine. The other half calls that a stubborn legend and says the men came through a low window of the chapel. Both readings quote the same source — Guillaume le Breton's Philippide, an eyewitness verse chronicle — and this episode reads that source directly, word for word, rather than picking a side. The eyewitness describes one building: a house John built the year before against the curtain wall, a privy below, a chapel above, one internal stair joining them, and a single window in the lower room. The two rival answers are not rivals — they are two floors of one house, and the argument survived for eight centuries because the same chronicler wrote the siege up twice: coy in the famous verse, blunt in his own prose history, where a footnote in the standard 1825 edition says outright that he called it the latrines.

Then the episode turns to the part almost nobody tells: the window is nearly the least important thing that happened in that siege. What actually broke the castle's last enclosure was a bridge cut into the living rock at the foot of its ditch — built deliberately by Richard himself when he deepened the ditches — under which French sappers mined in shelter, and a countermine dug by Richard's own garrison that hollowed the wall from the inside, so that a French siege engine's stone broke through exactly where the defenders had weakened it. The modern archaeological reading puts it in one sentence: a fortress doesn't fail at a point, it fails in sequence.

PRINT-READY, FROM THIS CHANNEL

How to Read a Castle, Vol. 01 — the method, the measured drawings and the documents behind the first two episodes

https://therepository.gumroad.com/l/read-a-castle

PRIMARY SOURCES

• Château Gaillard — Wikipedia (construction cost and dates, Richard's total castle spending in England, the siege timeline, the surrender numbers, citing Gillingham 2002, McLynn 2012, Oman 1991 and Cothren 1986). https://en.wikipedia.org/wiki/Ch%C3%A2teau_Gaillard
• Siège de Château-Gaillard — Wikipédia (FR) (the siege timeline, the winter expulsion of non-combatants, Gaubert's swim, citing McGlynn 1998 and Pitte 2007). https://fr.wikipedia.org/wiki/Si%C3%A8ge_de_Ch%C3%A2teau-Gaillard
• Château-Gaillard (Les Andelys) — Wikipédia (FR) (published dimensions, Viollet-le-Duc's plan and its known error in the plan legend). https://fr.wikipedia.org/wiki/Ch%C3%A2teau-Gaillard_%28Les_Andelys%29
• Guillaume le Breton, Philippide, Book VII — Latin text with Guizot's 1825 French translation and notes, including note 14 on the latrine/chapel passage. https://remacle.org/bloodwolf/historiens/guillaumelebreton/philippide7.htm

ARCHIVE PLATES — public-domain documents, with author, date, archive and licence.

• Paul Dujardin — 1896 | Wikimedia Commons | Public domain

... and 10 more plates, each credited on screen in the video and listed in this project's manifest.

Stone Record reads castles as documents — what the surviving paperwork and the standing masonry actually say, against what gets repeated about them. Primary sources for this episode are linked above.

Every historical image in this video is public-domain material from Wikimedia Commons, credited on screen — no image in this video is AI-generated. Engravings and drawings of Château Gaillard by Paul Dujardin (1896), Victor Adolphe Malte-Brun (1882), Gourdault (1893), Faure (1820) and Thiénon (19th c.); the measured plan is Eugène Viollet-le-Duc's, c. 1856. The site plan, every diagram, the timeline and the two quoted documents are original drawings and typography made for this channel — not photographs, not AI-generated.

#castles #medievalhistory #france #archaeology #siegewarfare