A stop on the Tales of Laugharne heritage trail · Laugharne / Talacharn, Carmarthenshire
The castle rose to a Tudor palace and fell in a single autumn week — and its wall is still, quite literally, the town.
The town's birth certificate still exists. You could hold it. And it will not tell you when the town was born.
This is a real object, not a legend — the borough charter under which the Laugharne Corporation still claims to hold its lands and run its affairs, more than seven centuries on. And the first, plainest, most maddening thing about it is this: there is no date on it, and no signature. Nowhere on the whole instrument does it say the year.
That single blank shapes everything the town can honestly say about its own beginning. A dated charter settles the question in one line. An undated one turns it into detective work — and that detective work has been running, in antiquarians' notebooks and archaeologists' footnotes, for the better part of two hundred years. It is still not closed.
And here is why there is no easy shortcut. Laugharne's liberties were not granted by a king. They were granted by the lord — one of the de Brian lords who also raised the green-stone castle above the estuary. Because the gift came from a marcher lord in his own private march, and not from the Crown, there is no royal record of it filed away at Westminster to go and check. It never went on the King's rolls, and its absence there proves nothing at all. The founding act of the town happened in the private word of a lord, and left no central paper trail behind it.
So to date the birth of Laugharne, there is only one place to look: the charter itself — the two words and the nine names written on it. That is where the next stops of this trail go hunting.
The document is late thirteenth century. Exactly how the town narrows it — from the phrase "the Lord Edward" and a list of witnesses — is the detective story told next.
The record: Guardrail: the charter is given as "late thirteenth century," never pinned to a single year. Granted by the marcher lord, not the Crown — no royal enrolment is to be expected, and its absence proves nothing.
A little tidal town put on the robes of a city — and then, quietly, wore them longer than the city did.
When the lord granted Laugharne its liberties, he did not invent them. He borrowed a cut that already fitted a far grander place. The charter dresses the burgesses of Thalacharn in exactly the laws and customs the burgesses of Carmarthen — the county town, the head port of the shire, twelve miles up the estuary — had enjoyed "in the time of King John." Stand in Laugharne and look up-water, and you are looking at the town whose civic clothing this small place put on and never took off.
And these were real, generous liberties, not flattery. We can read them because the charter was translated into print in 1839. It freed the townsfolk from every kind of field-work for the lord — no ploughing, no harrowing, no hay-making, no reaping or binding of his corn — from carriage-duties, from mending the mill, from being marched out to serve in his army or stand guard over the township. It guarded their market: no rival place of trade was to be set up "to their prejudice," and business was done by tally-stick. And it added a merchant's touch of fairness — buy something openly in the market, find later it was stolen, and swear you never knew, and the law would shield you.
Read them together and you can see the whole intention of the place in a single breath: this was built, from the charter outward, as a town of burgesses and tally-sticks — traders meant to trade — and not of labourers bound to a lord's fields.
And here is the quiet irony the town gets to keep. The very head port that lent Laugharne its civic clothing in the thirteenth century was stripped of its own medieval corporation in the nineteenth — reformed away — while little Laugharne kept hers. The borrowed robes outlasted the tailor.
How that survival happened — how a small town slipped through the reform that swept away Carmarthen's — is a story waiting further up the trail.
The record: The charter is modelled on Carmarthen's, granting the customs its burgesses held "in the time of King John"; the liberties are as translated in The Gentleman's Magazine, July 1839. Carmarthen was reformed out of its medieval corporation in the nineteenth century (1835) while Laugharne was not — bridge to.
The charter gave Laugharne's burgesses not just liberties but a share of the land — and told them how to walk its edge.
Read the grant and you can almost hear the old country in the names. Free common in the lord's forest of Coed Bâch. Common pasture in the marsh called Menetors. Free common of Makerells (said in the record, tier: Referenced). Grazing, timber and turf — the plain endowment on which a small trading town actually stood. And these are no dead words: they are the ancestors of the Laugharne Corporation Lands, a medieval gift carried unbroken down the centuries into the age of the Charity Commission (the story picks up again in C-19 and C-22).
But the treasure in this part of the record is the boundary itself. Adcock's 2024 translation of the charter reproduces the perambulation — the named walk around the common — and it reads like a map spoken aloud. Up from the rivulet Mackorellis to Greensladeshead; eastward over Eynons-down toward Brangweys; on to Corranshead, up to Horilake, to the top of Tadhill; down to Passenant's Lake, and east to the bounds between Moldhill and the old carucate that had once belonged to Rice. Most of those names have fallen off the modern map. But the act of naming them in a line has not fallen anywhere: it is the direct medieval ancestor of the town's tradition of walking its bounds. The charter did not only give the town its edge — it wrote down how to trace it.
And tucked in the same passage, a small phrase with a long reach: the weights and measures used "in the time of Gwydo de Brione the elder." Even the charter looks back over its shoulder to an earlier de Brian lord — a settled point of reference in a document that will not date itself (a knot we open in C-05).
The lord gave them grazing, timber and turf — and a route around the whole of it, named stone by stone. Walk out past the marsh and you are walking the oldest map the town still owns.
The record: Free common in Coed Bâch, pasture in Menetors, free common of Makerells, and the named perambulation and "Gwydo de Brione the elder" weights-and-measures clause (per Adcock 2024). The perambulation is told strictly as the historical ancestor of the boundary-walking tradition; the living, present-day Common Walk belongs to the Corporation's own story and is deferred to the.
The charter carries no year — the town's true age hides inside two words and a line of nine names.
If nothing on the document tells you when the town was born, how can anyone answer the question at all? You do what a detective does. You read what the writing does not mean to tell you.
Start with the two words. The charter names its overlord not as "King Edward" but as "the Lord Edward." For a long time older writers leapt on that: Lord, not King, so surely this was written before Edward became king in 1272, while he was still only the king's son. A neat idea — and, it turns out, a wrong one.
The reading that took it apart is recent, and it is the strongest the document has ever had. Watch how it turns: Edward's own brother Edmund held the lordship of Carmarthen from 1265, and it came back to Edward only in November 1279 — so a charter that calls Edward the lord of this fee has to come after that year, not before his crown. Add one more hard fact: the Welsh revolt of 1257 had burned Laugharne's castle and town to the ground, and they were not rebuilt until the 1270s. There was, for those lost years, barely a town to charter at all.
Which leaves the nine names. With no date on the page, the men who witnessed the grant are the clock — five knights and four others, including a clerk, whose lives can only have overlapped in one narrow span. Read together, the two words and the nine names close like a vice on a window of about two and a half years: November 1279 to March 1282, the grant made by the Guy de Brian who lived 1251–1307.
So the honest, remarkable answer is not a guess and not a round number. Not 1278, not 1290, not 1307 — those tidy dates are undocumented, and the town does not need them. The charter is of the late thirteenth century, narrowed by scholarship to a span you could almost name to the season, reasoned out of two words and nine names on a page that never gave its own age away.
And there is still a trap in that grantor's name — because Laugharne's lords ran a long line of men all called Guy. Which Guy? That is the next stop.
The record: The dating (Adcock 2024) — the "Lord Edward" phrase, the Nov 1279 Carmarthen reversion, the post-1257 restoration, and the witness list — is Referenced; the Nov 1279–Mar 1282 window is the best current scholarship, not a charter-roll date. 1290 and 1307 are undocumented and not stated as fact.
The town would love its charter to come from the famous Garter knight. It didn't — and the true story is better.
Here is the trap, and it is worth naming before anyone falls into it. The lord who gave Laugharne its charter was a Guy de Brian. And the most dazzling man ever to carry that name — Sir Guy de Bryan, Knight of the Garter, Admiral of the West, who died in 1390 and lies in splendour at Tewkesbury — was also a lord of Laugharne. So it is the easiest thing in the world to hand the charter to him. He belongs a lord or two later. He is not the grantor. That famous Guy has his own place in this story; the charter is not it.
Worse, the earlier Guy who did grant it cannot even be given a clean number. Look him up and you will find rival schemes that flatly disagree — one record calls the grantor "Guy IV, first of the lords," another numbers the Guys differently, and the counts simply cannot be reconciled into a single tidy line. The truthful answer is the uncomfortable one: the grantor cannot be pinned to a numbered Guy at all. In that undated generation the lord was either the Guy who died in 1307 or his own father — and the record, with no date on the charter, will not tell us which.
But the family's shape we do know, and from a hard document: the inquisition taken after Guy de Brian died in the summer of 1307. In a single breath it names three Guys in a row — the dead man's father, the dead man himself, and his son and heir, Guy, aged twenty-three — with old Joan, the grandfather's widow, still holding her dower third of the barony. Three living Guys inside one family memory. No wonder the numbering breaks.
And these were never only Laugharne men. The same Guy who died in 1307 held Walwyn's Castle in Pembrokeshire and the family's Devon seat at Tor Brian; "lord of Laugharne" was one title on a longer roll.
So the story to tell is not a neat pedigree. It is an honest knot — a founding charter from a Guy de Brian we can set in his generation but not in a numbered slot, and emphatically not the Garter knight the town might wish had signed it. The truth is harder to hold, and far more human, than the legend.
The famous Sir Guy de Bryan gets his own telling further on — the man who was castle and church and Garter star all at once. Follow him there.
The record: Standing name-trap
The confident date "Laugharne 1278–82" traces back to one man's book — a former Portreeve of the town, writing without a date himself.
Every confident record has a citation, and a citation is a door. Open the one behind Laugharne's founding date and watch where it leads.
The trail starts respectably. The Dyfed Archaeological Trust — the body that keeps the historic-landscape record for this corner of Wales — states the town's grant of borough status "in 1278–82," lands bestowed by Sir Guy de Brian. Solid enough. Except beside it sits a single, quiet reference: (Williams, n.d.). Follow it through the Trust's own list and "Williams" opens into a name: E. V. Williams, Laugharne Corporation — Narberth, no date, probably 1963. A study of the town written by one of its own former Portreeves. A local history. Not the charter. Not a royal roll. One man, close to his subject, undated himself, standing between the whole modern record and a document that never gave its age.
This is no slight on Williams. He is a fine guide, and his book is the only reason the date circulates at all. But he is a single guide — one thread, and everything precise about the town's birth year is hung from it.
Then watch the tidy horror of what happens next, because it explains a date you'll find everywhere. Williams writes "1278–82." The archaeological record cites Williams. And a national heritage database, as the next story shows, lops the range down to its first year and prints a bare "1278" — as though a clerk on a charter roll had written it. Four steps from one former Portreeve's undated book to a hard date on a government record.
So the honest reading is the modest one: "1278–82" rests on a single undated authority, and belongs here as local tradition, told as tradition. The charter itself still carries no date on its face. The town does not need the tidy number. What it has is better — a date you can trace, name, and see for exactly what it is.
And the next stop is where that quiet mistake goes national — a single record that gets the town wrong three times on one page.
The record: "1278–82" is carried as tradition, resting on the single undated authority E. V. Williams, Laugharne Corporation (Narberth, n.d. prob. 1963), a former-Portreeve local history and not a primary source; the app and site read "late thirteenth century." No charter-roll date is asserted.
One page of the national record gets three of the town's own facts wrong — its age, its fame, its government. Here is the correction.
The record is Coflein's NPRN 33066, "the Medieval and Later Borough of Laugharne," kept by the Royal Commission. On most of what it says, it is excellent. It calls Laugharne a formerly walled medieval borough, set on the landward side of its early castle; it keeps the old names, Abercorran and Talacharn; it explains why an English-speaking town sits in a Welsh-speaking country — the castle stood on the Landsker, the line of frontier strongholds; and it carries aerial photographs of the whole town, year by year, from 1946. On the fabric of the place, trust it.
But turn to the town's own constitutional facts, and the same page slips three times — and because it is a national record, the errors travel far.
Take the date first. The page says the borough charter was granted in "1278," and names the wrong Guy de Brian as grantor. But that "1278," as the trail behind it shows, was never on a charter roll — it is a local historian's range, "1278–82," quietly clipped to its first year until it looked official. The charter has no date on its face. It is late thirteenth century, and no more precise honestly can be said.
Take the fame next, and this is the one that matters most. The record calls Laugharne "the last surviving medieval corporation in the UK." It is not the last — the very Act that once saved Laugharne named Malmesbury in the same breath. The town holds the claim more carefully, and smaller: Laugharne is one of a tiny handful of medieval corporations still meeting, alongside the City of London. "The last" is the one phrase this town must never use about itself — and there it sits, in the national record.
Take the office last. The page says the charter provides for "two Portreeves." There is one Portreeve. The pair often mistaken for a second are the two Common Attorneys — a different office entirely. One Portreeve; two Common Attorneys; never two Portreeves.
A wrong age, an overclaimed fame, a miscounted officer — three slips on a single page. Naming them isn't point-scoring. It is the whole reason a project like this earns its keep: the charter has no date, the town is one of a rare few still meeting, and the Portreeve is one man — and those three true things are firmer ground to stand a town on than the tidy record that got them wrong.
Getting a heritage database to say the right thing is quiet work. But it is exactly the work that lets the rest of the town's story be told without a single crack in it.
The record: NPRN 33066's fabric facts (walled borough, Landsker line, alt-names, 1946+ aerials) are Verified. Three Refuted formulations held as live guardrails: the "1278" charter date (charter given as "late thirteenth century"); "last surviving medieval corporation" (corrected NOT to "the last" and NO LONGER to "one of only two," but to the softened, defensible class "one of a tiny handful of medieval corporations still meeting, alongside the City of London" per Dominic's 7 Sep 2026 ruling; precise class open — Mererid check pending, Malmesbury shares the 1883 statutory saving); "two Portreeves" (one Portreeve; two Common Attorneys). Nothing added; the corrections are the town's own held positions. Claim-softening (7 Sep 2026), per Dominic's ruling: the fame correction reworded from "one of only two… alongside the City of London" to "one of a tiny handful still meeting, alongside the City of London," and "the town has a twin" to "one of a rare few still meeting." "The last" is still reported-and (the record's error, named as such). No other fact, tier or guardrail moved. Back to Karen for the fidelity check.
The old books call this a walled borough — right about the borough, loose about the wall. What the record gives instead is firmer ground.
A borough did sit on the castle's landward side. The gazetteers and the national town-record agree that an enclosure was established there, the gatehouse standing "at its northern apex facing into the walled borough." But "walled" is doing gentle work in that phrase. The circuit is read from the town's shape and its early core of just over thirty dwellings, not from any surviving masonry. No upstanding town wall stands here, as one still does at Carmarthen or Caernarfon. What ringed Laugharne was far likelier a bank and a ditch than a wall of stone.
The firmer fact the record does hold comes out of the same century. In the Calendar of Patent Rolls of Edward IV, Laugharne appears as a forfeited lordship: "the castle, town, manor and lordship of Telagharn in South Wales," late of James Butler, earl of Wiltshire and Ormond, a leading Lancastrian. It was seized into the king's hands by an act of forfeiture in Parliament at Westminster on 4 November 1461, with William Herbert commissioned to take Butler's South Wales castles. On the eve of the Tudor century the town belonged to an attainted earl, and passed through the Crown by forfeiture. That is a stranger and a steadier truth than a wall the ground never kept.
And the directories bite back at the very end. Samuel Lewis's careful 1833 dictionary — so good on markets and fair-dates — hands down, in the same breath, pure legend as history: the castle "besieged by Cromwell," holding out "for three weeks" under "General William Laugharne." Every clause is wrong. The 1644 siege lasted about a week at the end of October, not three weeks; the commander was Major-General Rowland Laugharne, not William; Cromwell was nowhere near it; and — the correction that never sleeps — the town is not named after the general, the family took its name from the town. It is the perfect close: the same page that fixes the price of a fair will, one line later, march a besieging Cromwell up a road he never travelled. Take the trade lists as evidence. Test the history at the top of the page every single time.
The record: A walled borough existed on the castle's landward side (Gatehouse Gazetteer + Coflein NPRN 33066). "Walled" inferred from topography/documents/~30-dwelling core, not upstanding masonry; consistent with an earthwork circuit. Laugharne the 1461-forfeited lordship of James Butler, earl of Wiltshire (attainder 4 Nov 1461; Herbert commissioned). Lewis 1833's "Cromwell / three weeks / General William Laugharne" siege tradition, not adopted; correct account: siege ~a week end-Oct 1644, Maj.-Gen. Rowland Laugharne, town not named after him. Charter kept "late thirteenth century."
The lords who built this castle were nearly all called Guy de Brian, and the books number them three different ways — but one document of 1307 puts them in order.
First, feel the size of the confusion, because it is real. Three respectable authorities number these lords three different and flatly incompatible ways: Coflein makes the charter-granter "Guy IV," the Community Council makes him "Guy V," and the Corporation's own website counts "Guy 1" through "Guy 4." No wonder the honest question has felt unanswerable — everyone was counting, and no two counts agreed.
Now read the primary record, and watch the fog lift — not into a tidy numbering, but into something better and firmer: a spine of death-dates. The Calendar of Inquisitions post mortem, volume 5, number 64 — the 1307 inquest taken after a Guy de Brian, lord of the barony of "Tallauhern," had died — does the whole job in two sentences. It records that "Joan late the wife of Guy de Brian, grandfather of Guy who now is, has a third part in dower," and that "Guy his son, aged 23, is his next heir". Read it slowly: one document, three Guy generations at a stroke — the grandfather, dead before 1307, his widow Joan still holding her dower third; the lord who died in 1307; and the heir aged 23, born about 1283–84. Then the Complete Peerage carries the line on: that heir died in 1349, and his son — Sir Guy de Bryan, Knight of the Garter — died on 17 August 1390.
So the record hands us a spine you can trust — a Guy dead in 1307, a Guy dead in 1349, a Guy dead in 1390 — and no absolute ordinal survives the crossing. These lords held Laugharne as one barony among several, alongside Walwyn's Castle in Pembrokeshire and Tor Bryan in Devon, their family seat. And here is the detail to carry away from the stone itself: it was in the mid-fourteenth century, during the building campaign of the lord who would become the Garter knight, that much of the castle was refaced in its distinctive green stone. Then the line failed — the Garter knight's own son and heir died before him in 1386, the barony passed to granddaughters and fell into abeyance in 1390 — and Laugharne was left with no resident lord, and began the long decline that ended only with the Tudor grant to Perrot (Verified; the causal link to Laugharne's own decline Referenced-pending).
So — which Guy granted the borough charter? In the generation that matters, it was either the Guy who died in 1307 or his father, Joan's husband; and the record will not choose between them without the charter's own text. And so the charter stays exactly what it must: late thirteenth century — no year, no numeral, no false certainty. The wonder here is not a name pinned down. It is a family whose sameness of name became its own kind of monument, and a single 1307 parchment steady enough to hold three generations in place.
The record: Calendar of Inquisitions Post Mortem, vol. 5, no. 64 (the 1307 inquest: Joan's dower, the heir aged 23, the barony of "Tallauhern"); The Complete Peerage, vol. 2, "Bryan or Briene" (the deaths of 1349 and 1390, the heir's death in 1386, the barony's abeyance); R. Avent, Laugharne Castle (Cadw, 1995): the mid-fourteenth-century green-stone campaign. The three numbering schemes compared are those of Coflein, Laugharne Township Community Council and the Laugharne Corporation website.
The little gate between the Town Hall and the castle is the seam where the chartered town met the lord in his walls.
It is easy to walk past. But look again: this is the Outer Gatehouse to Laugharne Castle, listed Grade I in its own right (Cadw 9652, 25 September 1986), and the listing places it precisely where you are standing — "at the bottom of the street near the town Hall." A rubble, two-storey, once-crenellated gate of the late thirteenth or early fourteenth century: a passage between rectangular guardrooms, a chamber above with its own little garderobe, arrowslits with round oillets punched below them, spur buttresses — and, for a flourish, a dragon turning on the roof as a weathervane.
Now feel why it matters. The medieval castle was a small inner court set inside a wide, diamond-shaped outer ward, and the great gate stood at that ward's northern apex, facing straight into the walled borough — the town laid out on the castle's landward side. This gate is that apex. Its defences were rebuilt in stone at the end of the thirteenth century, then rebuilt again in the mid-fourteenth in the de Brians' distinctive green stone — the same campaign, the same lord, that refaced the castle itself. Perrot altered it with everything else in 1587–92; its north side was smashed open in the 1644 siege.
So walk the sequence the record keeps — Town Hall, gate, castle — and you cross from the self-governing borough into the lord's ward through the one structure built to hold both together. One caution before you go, because the two are so easily muddled: this is the free-standing two-storey outer gate. The tall, four-storey, bay-windowed gatehouse you will meet further in is a different building — the inner gate, part of the castle's own listing, which Perrot raised to its present height. Two gates, two listings. This one is the threshold — and its whole meaning is where it stands.
The record: Cadw LB 9652; Avent, Laugharne Castle / Cadw Scheduling Report via the Gatehouse Gazetteer.
The most photographed thing here is a Tudor bay window over the gatehouse — finished the very year its builder was condemned to die.
The man was Sir John Perrot (1528/9–1592): Elizabethan statesman, three times mayor of Haverfordwest, first president of Munster, Lord Deputy of Ireland — and, at the last, a prisoner who died in the Tower after being attainted for treason. At Laugharne he did what Renaissance grandees did to old castles: raised the medieval curtain and gatehouse, threw ranges of fine lodgings around the inner court, put new windows in the hall, and laid the cobbled courtyard the excavators would later uncover.
Here is the date that gives the story its edge — and it is Cadw's own, the only one anchored in the standing fabric: "refenestration and reconstruction by Sir John Perrot 1587 to 1592." A building campaign whose window closes in the year of his fall. Perrot was attainted on 17 April 1592 and dead the same year. His palace was, quite literally, built to end when he ended. You will hear a neater tale — that he "acquired the castle in 1575 and was granted it outright in 1584." It fits his known lease-then-grant pattern at Carew, so it may well be true; but no record actually ties those years to Laugharne, so the honest telling holds them lightly, not stated as fact.
Two more honesties belong beside the romance, and the app tells them plainly. The colourful claim that Perrot was a secret son of Henry VIII is Refuted by the current scholarship — his mother was no royal favourite, and the myth has no foundation; it keeps resurfacing, and it should be named and corrected, never repeated. And this was never his true home: Carew was the seat he largely rebuilt and made his principal house; Laugharne was the second palace, the grand estuary retreat. When the axe fell, the castle did not come back to his line — his son Sir Thomas was restored in blood and given Haroldston, "but not Carew and Laugharne castles, which reverted to the Crown and the earl of Northumberland respectively". Barely finished, the palace passed out of the builder's family the moment he was condemned. Its next chapter — a single autumn week of cannon fire — would be its last.
The record: History of Parliament — Perrot; DWB — Perrot family; Cadw LB 9653.
This ruin is picturesque because the town carried it home — a single week in 1644 broke it, then centuries quarried it into houses.
The war set a strange stage. When it broke out the castle belonged to Sir William Russell, a Royalist — yet it was held for Parliament. Geography decided that: Laugharne sits on the sea, and Parliament held the navy, so this garrison could be fed and relieved by water while the King's inland forces could not. For a moment the advantage flipped. In June 1644 a Royalist army under Sir Charles Gerard swept through south-west Wales — Kidwelly, Carmarthen, Cardigan, Newcastle Emlyn, and Laugharne Castle with them. But the Royalist ruin at Marston Moor that July pulled Gerard's field army back out of Wales, leaving the men he'd installed here stranded, with no hope of rescue.
The counter-blow came at the end of October 1644. Major-General Rowland Laugharne, Parliament's commander in the south-west, came up with a force of about 2,000 men and several guns, and sent in a summons to surrender on 29 October. It was refused (Verified; the round figure of 2,000 and the summons date come through Avent's Cadw account, flagged for a direct read; the exact battery positions are single-sourced and Referenced-pending). What followed ran about a week — roughly 28 October to 3 November: the town taken, the castle pounded, the Outer Ward stormed, the front of the thirteenth-century outer gatehouse blown open by cannon. The garrison surrendered on Sunday 3 November 1644 after a night attack, and was let march away to the King's men at Carmarthen.
Then came the deliberate part. The castle was slighted — reduced on purpose so it could never be defended again — and it was never rebuilt as a fortress. In time it went back to Russell, and the ruin slid gently into the ornamental garden of Castle House. And here the story runs straight into the streets around you: the broken castle was pillaged for stone to build the houses of Laugharne. It is picturesque because the town took it home.
Two things this stop exists to set straight, gently but firmly. The siege lasted about a week — six or seven days — not "five days"; you'll meet the tidy "five-day" version, but the same account dates the approach to late October and the surrender to Sunday 3 November, and the honest span is anchored to those dates. And the town is not named after the general. It's a seductive slip, because the commander who took the castle was himself a Laugharne — but the place is on record as Abercorram and Talacharn through the twelfth and thirteenth centuries, five hundred years before Rowland Laugharne drew breath. The man took his surname from the place; the place owes nothing to the man.
The record: Avent, Laugharne Castle (Cadw, 1995) — flagged for direct read; Castles, Forts and Battles; Laugharne Township Community Council history; BCW Project.
Four years after the siege, the general who took this castle turned against Parliament himself — and what Laugharne did in that second war, the record has not yet told.
His name was Rowland Laugharne — Major-General, Parliament's commander in the south-west, the soldier who had battered this fortress into surrender. In 1648 he turned. With John Poyer and Rice Powell he rose against the Parliament they had all fought for, was beaten at the battle of St Fagans on 8 May 1648, and watched the rising collapse into the sieges of Pembroke, Tenby and Carmarthen (Referenced — the standard narrative of the Second Civil War, not yet read at object level for Laugharne itself). What the town and its broken castle did in that second war — garrisoned, passed, or simply ignored — the register cannot yet say. On 1648, here, it has almost nothing (Verified as a state of the record).
So this is an honest stop. It marks the gap, and it names the source that will one day fill it. The great printed collection for the war in Wales is John Roland Phillips's Memoirs of the Civil War in Wales and the Marches, 1642–1649 — Longmans, 1874, printed at Carmarthen — and it is the right kind of source, because its second volume prints the original documents, drawn from the King's Pamphlets at the British Museum and the State Papers at the Record Office. A fact Phillips quotes from a dated document can stand as evidence, not just as story. That is the rule this history has taken from it: what Phillips prints from a document may reach the highest tier; his own connecting narrative stays Referenced (Verified as method).
There is a beautiful trap in that book, and it is worth catching before anyone else falls in. In his preface Phillips writes that he began the work to learn "whether the Castle, which forms its principal attraction, had been garrisoned during the Civil War" — the castle of his "native parish." A book printed at Carmarthen, subscribed across the county, opening on a castle that is a parish's chief glory: it reads exactly like Laugharne. It is not. Phillips was born at Cilgerran, in Pembrokeshire, in 1844, and the castle behind the Memoirs is Cilgerran Castle. Neither the man nor his motive belongs to Laugharne — a near-miss caught at the door.
One small, true thing the source did give up on this pass. Phillips's own 1874 list of subscribers names a "W. Norton, Esq., Laugharne Castle" — a glimpse of who was living at or beside the ruin in the Victorian town — and a "George Lewis, Esq., Victoria-street, Laugharne," which carries the name Victoria Street back as an attested Laugharne street to 1874. The war's Laugharne chapter waits. But the source that holds it is proven, free and named — and the thread stays open, on purpose, for the day someone reads it through.
The record: J. R. Phillips, Memoirs of the Civil War in Wales and the Marches, 1642–1649 (Longmans, 1874): vol. II prints original documents from the King's Pamphlets (British Museum) and the State Papers; the 1874 subscribers' list names "W. Norton, Esq. Laugharne Castle" and "George Lewis, Esq. Victoria-street, Laugharne". Dictionary of Welsh Biography: "Phillips, John Roland (1844–1887)", born at Cilgerran. The 1648 rising and the battle of St Fagans (8 May 1648) from the standard narrative of the Second Civil War, not yet read for Laugharne itself.
The "boundary walls" by Castle House are the castle's own thirteenth-century curtain — the fortress that learned to look like a garden.
Three separate Grade II listings, all made on the same day — 25 September 1986 — describe what is really one continuous circuit, split three ways. Read them apart and they are garden walls. Read them together and they turn back into a ring of defence.
The first stretch (Cadw 9654) runs south-west from the Outer Gatehouse, a part-crenellated rubble wall holding up the raised outer ward, Wogan Street below it, before it turns a sharp right angle. And at its far end sits the whole story in a single Cadw sentence: "a ruined probably C18 garden wall on top of and incorporating the late C13 curtain wall adjoining SW castle tower." There it is — genuine late-thirteenth-century curtain-wall fabric, surviving inside what the register files as a boundary wall. The medieval de Brian castle and the town-edge wall are not two things here. For this run they are one wall, wearing a Georgian cap.
The second stretch (Cadw 9655) runs east from the same gatehouse — rubble at the west end, then mostly late-C18 or early-C19 Flemish-bond brick, built hard against the older rubble, pierced with arched openings, one with classical detail, that lead through into the Castle House garden. The shared listing text then follows the line on: a rubble wall with ruined crenellations climbing along the cliff edge to the gazebo, and down to the foreshore. The ring does not politely stop at the garden gate — it goes up to the little tower on the sea-edge and drops to the tide, the very line the medieval defences once held.
The third listing (Cadw 9658), the wall west of Castle House, is the odd one out, and it gives the game away by being so. It is classed Domestic, not Defence, and — unlike its two neighbours — is not said to lie within the Scheduled Ancient Monument. A C18–C19 rubble wall lining the road along the castle area's northern edge, and at its north end, something genteel: paired decorative ironwork gates and trellis-work iron piers, "probably contemporary with the ca 1810 remodelling of Castle House". Here the wall stops being a fortress and becomes a front garden — and these are the same gates you enter Castle House by.
So the whole circuit maps the castle's afterlife in stone. South and east, a defensive ring still carrying thirteenth-century curtain fabric, running to the gazebo and the sea. West, a Georgian garden wall with pretty gates, dated to the house's Regency gentrification. The boundary of the castle enclosure is, in part, the castle's own wall — and where it isn't, it is the polite wall a Regency gentleman raised to keep a ruin for a view. The town never quite let the fortress go. It just gave it a coping stone and a gate.
The record: Cadw statutory listings LB 9654, LB 9655, LB 9658 (all).
Stand still for a moment. Wherever you are in old Laugharne, you are almost certainly standing on something one man built.
Look up at the castle: that green stone facing the wall above the water is his. The gatehouse at the edge of the borough, rebuilt in the same distinctive green — his. The great cross-shaped church on the hill, three-quarters of it still the fourteenth-century fabric he raised over ground already a thousand years holy — his too. Castle, church and lordship, all meeting in a single life. He is the one person in the whole story of this town who touches all three of its founding things.
His name was Sir Guy de Bryan, and he was as high as an Englishman could climb. A Knight of the Garter — that small, jealous order of the king's own chosen. Admiral of the West. Summoned to Parliament in his own right. Laugharne was only one of his seats, but it was here he cut the deepest mark: he refaced the fortress in green, threw a bridge across its ditch, and rebuilt St Martin's on its ancient hill.
And here is the town's oldest mistake — the one nearly every guidebook makes — so learn it now and you will always have the truth. He did not grant the borough charter. That belongs to an earlier Guy de Brian, a generation or two before him, back in the late thirteenth century. This Guy is the grandson, the Garter knight. Merge the two and you hand a fourteenth-century admiral a charter he could not possibly have sealed. Keep them apart, and the town makes sense again.
His story ends in a quiet, fateful failure. His only son died before him, in 1386. So when the old lord died full of years in 1390, his heirs were granddaughters, and the lordship passed for good out of resident de Brian hands — into long-distance ownership, and then into decline, until Perrot came to raise the castle a second time. The greatest lord Laugharne ever had was also its last to live here. The whole shape of the next two centuries was set on the day his son died first.
He himself lies a hundred miles away, under an alabaster canopy in Tewkesbury Abbey. But everything your eye can reach from this spot, he left behind.
Now climb to the castle, or up to St Martin's — and see his two hands on the same town.
The record: Load-bearing name-trap held: the Garter knight is NOT the charter-grantor — the charter belongs to the earlier, late-C13 Guy de Brian.
That fine Tudor bay window over the gatehouse arch was finished just before the year its builder lost everything.
His name was Sir John Perrot, and the ruin above the estuary looks the way it does because of him. The raised curtain wall, the ranges of lodgings thrown round the inner court, the cobbled Tudor courtyard whose stones the ground still gives up — all his, in one furious campaign of building. Cadw dates the work to 1587 to 1592. There is a terrible logic in that closing year. The works stop in 1592 because in 1592 Sir John Perrot died in the Tower of London, condemned for treason. The palace and the man ended in the same twelvemonth.
He was an Elizabethan of the very first rank — sheriff, three times mayor of Haverfordwest, a Member of Parliament, and Lord Deputy of Ireland, the Crown's chief officer in a hard posting. Elizabeth I granted him Laugharne in 1575. And here, in the last five years of a very large life, he poured his energy into turning a battered marcher fortress into something a great man could live in — and did not live to finish it.
Two tales cling to him, and both are false, so meet them here and let them go. He was not a secret son of Henry VIII — a piece of Tudor gossip the record firmly refutes. And this was not his true home: his principal seat was Carew, over in Pembrokeshire. Laugharne was the improvement project, the grand ambition, the place he was making magnificent when the axe fell.
There is one more thread, and it belongs to the sea below you. In 1573 the Queen set Perrot to hunt down those along this coast who had dealings with pirates. In that same decade the most notorious Welsh pirate of the age was selling his plunder in the villages around Laugharne. The Queen's commissioner and the coast's most wanted man, circling the same small harbour — a tempting scene. But keep it honest: no document yet places the pirate inside Perrot's net. Two true facts, not yet joined. The lead is real; the story stays open.
He rose here, in green ambition and cut stone. And a single autumn week, a lifetime later, this same castle would fall — but that is the next chapter.
Walk into the inner ward and stand on his cobbles. This is where the ruin was, for five short years, almost a palace.
Rest at the Globe → Open the interactive guide →
Every walk in Laugharne finishes at the Globe — bar, restaurant and bakery. theglobe.wales · globebakery.wales
On these walks: The Whole Town · The Estuary Side · Laugharne in an Hour · A Turn Around the Town
Every fact here is checked against the record and sourced; where the record goes quiet, we say so. Part of the House of Culture, Laugharne.