The Repository Stone Record

Sudeley Castle's Roofs Outlived the Order to Slight It

October 16, 2026 · Runtime 9:38 · Watch on YouTube

In March 1658 five masons were sworn at the George Inn, Winchcombe, and said they had pulled parts of Sudeley Castle down — damages, they said, of four to ten thousand pounds. In 2026 thirty cores drilled from its outer-courtyard roofs dated the oak to 1569–72, nearly eighty years before the order to slight it.

Chapters

  1. 0:00 Five masons, one oath
  2. 1:20 The order of 3 April 1649
  3. 2:24 The fine and the claim
  4. 3:03 The George Inn, 1658
  5. 4:13 Fully destroyed, or parts?
  6. 4:49 The Bucks' engraving
  7. 5:27 Three roofs, drilled
  8. 5:45 How a ring date works
  9. 6:33 Felled 1569 to 1572
  10. 7:11 The window dated 1572
  11. 8:01 What a ring date cannot say
  12. 8:30 How to read a damage claim

Transcript

Five masons, one oath 0:00

On the thirtieth of March, sixteen fifty-eight, five masons sat in the George Inn at Winchcombe and swore an oath about Sudeley Castle. They said they had been hired to pull parts of it down. And that the loss ran into thousands of pounds.

This year, someone drilled thirty cores out of the castle's roofs and counted the rings. The trees were felled between fifteen sixty-nine and fifteen seventy-two. That is nearly eighty years before the demolition.

So was Sudeley destroyed in sixteen forty-nine, or not? It depends on which kind of evidence you trust. A sworn statement and a roof beam are both evidence. They were made for very different reasons, and only one of them was asking for money.

Here is the version you will read in most places. Sudeley, in Gloucestershire, was the seat of Lord Chandos, and the burial place of Catherine Parr, the last wife of Henry the Eighth. In the Civil War it changed hands, and in sixteen forty-four it was bombarded.

In sixteen forty-nine, Parliament ordered it slighted. The work is said to have included taking off the roofs. It lay ruinous until two glove-makers, the Dent brothers, bought it in eighteen thirty-seven.

The order of 3 April 1649 1:20

The paper trail is better than most. Emma Dent, of the family that restored the castle, printed it in her Annals of Winchcombe and Sudeley, in eighteen seventy-seven. On the third of April, sixteen forty-nine, the Council of State ordered Sudeley slighted: rendered untenable as a military post.

Lord Chandos was told to see it done. He raised objections, and won a short delay. On the twenty-fourth of April he gave up opposing, and asked for compensation instead. It was promised for when the work was finished.

The job went to Sir William Constable, the governor, who sat on the Council himself. On the third of August, Chandos took a pass to go abroad with his wife. While he was away, the castle was demolished.

On the twenty-ninth of September, the Council recorded that the work had been done to its satisfaction. That is the order, and that is the receipt. Neither one says what was taken down, or how much.

The fine and the claim 2:24

What we have is what people said afterwards, and they said it for a reason. Chandos had been fined as a royalist, and he was still paying. In May, sixteen fifty, the Council set aside one thousand pounds of what he owed, until his real loss from the demolition could be worked out.

Lord Chandos died in sixteen fifty-five. His widow Jane married twice more, and her third husband was George Pitt. In sixteen fifty-seven and fifty-eight, the two of them petitioned. They asked for the rest of the fine to be remitted, and they listed the damage done to Sudeley.

The George Inn, 1658 3:03

A commission was named to inquire into their claims. On the thirtieth of March, sixteen fifty-eight, witnesses were sworn at the George Inn, Winchcombe. Five masons gave their names: Reeve and Hall, of Didbrooke, and Bell, Bloxam and Baswell, of Winchcombe.

By Dent's summary, they swore they had been employed by the soldiers or agents of Parliament to pull down parts of the castle, so as to make it untenable. And they swore the demolition had cost Lord Chandos between four and ten thousand pounds.

Set those numbers side by side. The Council had set aside one thousand pounds against the loss. The masons put it at four thousand, and up to ten. They were describing a demolition. In the same breath, they were pricing a claim.

Here is the first problem with that oath. The men did the work, so they are good witnesses to what they pulled down. But the oath was gathered to win a remission, and a damage claim is written to be large. It is evidence for the claim. It is not a survey of the castle.

Fully destroyed, or parts? 4:13

The second problem is smaller, and stranger. The Castle Studies Trust says the five depositions stated the castle had been fully destroyed, and that two Gloucester judges, Edward Slaughter and Robert Megge, signed them off. Dent, who had copies at Sudeley, writes that the men pulled down parts.

Neither deposition has been read for this episode, only the two summaries. They differ by one word, and that word is the whole question. Fully destroyed, or parts. A roof beam does not have that ambiguity.

The Bucks' engraving 4:49

There is one more witness, from eighty years later. In seventeen thirty-two, Samuel and Nathaniel Buck engraved the castle from the south-west. They dedicated the print to George Pitt, proprietor of these remains. The inscription tells the family story, from Lord Chandos to his widow Jane.

The print draws towers, ranges with windows and chimneys, and a chapel. And it calls all of it remains. It settles nothing. But it is a picture of a castle with a great deal still standing, eighty-odd years after the order to slight it.

Three roofs, drilled 5:27

Now the roofs. In twenty twenty-six, the Castle Studies Trust funded a dendrochronologist, Dr Andy Moir, to record the outer courtyard of Sudeley. Three ranges: west, thirty-six metres long; north, thirty-eight; east, forty-six.

How a ring date works 5:45

Tree rings work like this. A tree adds one ring a year, and the width of the ring follows the weather. A pencil-thin core gives you a barcode of good years and bad ones. Two timbers cut from trees that grew at the same time share the same barcode.

Slide the barcodes against each other until they lock, and you can join the timbers into one master pattern. Slide that pattern along a dated chronology, and the last ring gets a calendar year.

Thirty cores were taken from the roofs. Twenty of the timbers matched each other, and together they built a mean chronology one hundred and fifty-nine years long, from fourteen thirteen to fifteen seventy-one. Ten did not join it, so the result rests on twenty timbers, not thirty.

Felled 1569 to 1572 6:33

Where the last growth under the bark survives, a date gets a season. Here the trees were felled between the spring of fifteen sixty-nine and the winter of fifteen seventy-one to seventy-two. Timber was usually worked while still green, so the roofs were most likely built between fifteen sixty-nine and fifteen seventy-two.

The report names the likely builder: Edmund Brydges, the second Lord Chandos. A Lord Chandos of the same house, it seems, built these roofs. And nearly eighty years later, another Lord Chandos was ordered to see his castle slighted.

The window dated 1572 7:11

Dent gives a reason to believe it. A ruined window in the quadrangle carries the initials E. C., and the date fifteen seventy-two. The inquisition after his death says Edmund died on the eleventh of September that same year.

That overturns what the Trust calls the previous history: that the castle's roofs were destroyed in the Civil War. The oak in all three ranges is mainly Elizabethan. These frames were cut a lifetime before the order, and they were still there to be drilled.

Some histories already hinted at it. The castle's Wikipedia entry says the work took some five months, largely dismantled the inner courtyard and the royal apartments, and, strangely, left much of the outer courtyard intact. The rings give that strangely a date.

What a ring date cannot say 8:01

A ring date has limits, and it helps to say them. It tells you when a tree was felled. It does not tell you whether the roof was stripped of its covering, or lost a few bays, and then put back on the same frame.

What the rings do say is hard to argue with. These roof frames were cut under Elizabeth, and they still carry the roofs. Whatever the masons took down in sixteen forty-nine, these frames were not part of it.

How to read a damage claim 8:30

So read a damage claim like this. Ask who was asking, and for what. A paper written to win money will say how much was lost. It will not say what was left. Then look for the witness that was not asking for anything.

Left open: if the frames stood, what covered them? Lead, slate, stone tiles? The rings cannot say, and the summaries of the depositions do not say either. That is a question for the roof itself, and for the archive.

Sudeley has always been told as a demolition. The paper that tells it is an order, a receipt and an oath, and each was written by someone with a reason. The best witness was older than all three, and it had never asked for a penny.

At Bedford, the order was carried out and the ground showed it. At Sudeley, the oath said one thing and the oak said another. Next time: a king orders a chimney, in twelve twenty-nine. And what that order says about the cold, dark castle of every guidebook.

Description and sources

The belief: Parliament ordered Sudeley slighted in 1649, and the work included taking off the roofs. The paper trail is good — Emma Dent printed it in her Annals of Winchcombe and Sudeley (1877): the Council of State's order of 3 April 1649, Chandos's petition for compensation, Sir William Constable entrusted with the work, the Council satisfied on 29 September.

Two problems with reading the oath as a survey. The five masons (Reeve and Hall of Didbrooke; Bell, Bloxam and Baswell of Winchcombe) were sworn on 30 March 1658 to support a petition by Chandos's widow Jane and her third husband, George Pitt, to have the rest of his fine remitted — and the sum they named, four to ten thousand pounds, is four to ten times the £1000 the Council had set aside in 1650 against his real loss. And the two summaries of the depositions differ by a word: the Castle Studies Trust says “fully destroyed”; Dent says “parts of the Castle”. Neither deposition was read for this episode.

The roofs: in 2026 Dr Andy Moir drilled 30 cores from the west, north and east range roofs of the outer courtyard. Twenty timbers matched into a 159-year chronology (1413–1571); the trees were felled between the spring of 1569 and the winter of 1571/2, probably for Edmund Brydges, 2nd Lord Chandos, who died on 11 September 1572 — the year on a window in the inner quadrangle that carries his initials. What a ring date cannot say: whether the roof was stripped of its covering and put back on the same frame.

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

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

... and 13 more plates, each credited on screen in the video and listed in this project's manifest.

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

Every image is public-domain or CC BY material from Wikimedia Commons or the Internet Archive, credited on screen — nothing is AI-generated. The diagrams and set-type quotations are code: drawings of the argument, not measured surveys. The dating diagram shows the general method, not Sudeley's own ring data. The Dent and Castle Studies Trust passages are summaries of depositions the episode did not read.

#castles #history #archaeology