Bamburgh Castle

Bamburgh Castle stands on the north-east coast of England, above the village of Bamburgh in Northumberland. It occupies a black crag of volcanic dolerite that belongs to the Whin Sill, also called the Great Whin Sill. One account places the escarpment 150 feet above the North Sea. Another places the castle 45 metres above sea level.

A survey of the outworks says the dolerite rises some 30 metres above the ground outside, that the plateau is about 3 hectares, and that the inner ward at the south-east reaches 45 metres above ordnance datum while the west ward by St Oswald’s Gate lies some 22 metres above sea level and the ground around it about 13 metres. The length is given as about a quarter of a mile. The area is given as 8 acres, as nine acres, and, for the plateau surface, as about 3 hectares. The rock has been a defended place for a very long time.

One summary of the early digging says occupation ran from the first century BC to the end of the Middle Ages, beginning in the pre-Roman Iron Age, with Roman activity that included a possible beacon on the seaward end of the rock.

The oldest names are not one name. The place is called Din Guarie, Din Guaydri, Din Guaire, Din Guardi, and Din Guyardi. It is also Bebbanburh, Bebba’s Burgh, and, later, Bamborough as well as Bamburgh. The Celtic Britons held a fort here before the Angles.

It may have been the capital of Bernicia from about 420 until 547. Bernicia was an Anglo-Saxon kingdom planted by Anglian settlers of the sixth century in what is now south-eastern Scotland and north-east England, roughly from the Forth to the Tees, taking in Northumberland, the lands that became Tyne and Wear and County Durham, and Berwickshire and East Lothian. Its British name was Bryneich or Berneich, and the English forms were Bernice and Beornice. The Brittonic kingdom of that coast has been read as the southern lands of the Votadini, perhaps part of a division around 420 of a supposed northern realm remembered in Welsh tradition as the Old North.

The Angles are thought to have taken the whole region by about 604, though that date is a likelihood rather than a recorded day.

The Anglo-Saxons were the Old English-speaking peoples of the early medieval kingdoms. Their period in Britain is placed from about 450 to the Norman conquest of 1066. The Normans were the people formed in Normandy from Norse settlers and the Franks of West Francia, and from 1066 their dukes were also kings of England. At Bamburgh the sequence is told as a Brittonic fort, then an Anglian royal seat, then a Norman and English royal castle, then a private house.

Ida of Bernicia is the first king of Bernicia who can be named. The Anglo-Saxon Chronicle puts the start of his reign in 547, calls him the son of Eoppa, and says he ruled for twelve years and built Bamburgh. He is also said to have died about 559, so a reign from about 547 to 559. He is given one queen, Bearnoch, and twelve sons, some born to his wife and some to concubines.

The lists do not agree name for name. Glappa, the next king in one sequence, is not among those sons. A later chronicle confuses his land with Northumbria and has Ælla of Deira succeed him, though Northumbria did not exist until Bernicia and Deira were joined. The castle’s own account calls him Ida the Flamebearer and says that in 547 he laid the first timbers of a wooden stockade, and that on his death the place passed to his grandson Æthelfrith.

Another account says the citadel was captured by Ida in 547, which assumes a fort already standing, and that it passed between the Britons and the Anglo-Saxons three times before it was in Anglo-Saxon hands in 590. Those statements do not describe the same act. Building, laying the first timbers, and capturing an older fort are three ways the start has been told, and the twelve-year reign ending about 559 does not by itself hand the rock to Æthelfrith.

Hussa is the seventh known ruler of Bernicia, given seven years from about 585 to about 592. Those years are a guess. He has been called a son of Ida, and he has been left off the list of Ida’s twelve sons and read instead as the leader of a rival Anglian party. During his reign a coalition of Rheged and the Brittonic kingdoms of Strathclyde and Elmet, with Bernicia named among the besiegers in one telling of that alliance, nearly drove the Anglian kings of Bernicia out of Britain.

The effort is said to have failed because the British leaders quarrelled, and Urien of Rheged was killed around 590 by his former ally Morcant. The castle account says the Britons briefly retook the fort from Hussa in the war of 590 and that it was taken back later the same year. Hussa’s own record does not name the rock in that siege, so the recapture of 590 and the siege of Hussa’s reign sit side by side rather than as one proven event. After Hussa, the chronicle tradition has a split: Hering, son of Hussa, is said to have led the forces of Áedán mac Gabráin against Æthelfrith at Degsastan in 603.

Æthelfrith, son of Æthelric and grandson of Ida, succeeded Hussa around 592 or 593. He is the earliest Bernician king of whom any substantial story survives. His capital is placed at Bamburgh. Genealogies attached to the Historia Brittonum give him twelve years in Bernicia and twelve more in Deira, which can be read as Bernicia alone from about 592 until about 604, when he also took Deira.

That joining is the first making of Northumbria. Bede compared his success against the Britons to Saul, and recorded that he was pagan. He may have been the victor at Catraeth, the fight remembered in the poem of the Gododdin, around 600. The Britons called him Flesaur, the twister.

In 603 Áedán of Dál Riata brought a great army against him and was beaten at Degsastan. Æthelfrith was killed around 616 at the Battle of the River Idle, against a coalition led by Rædwald of East Anglia, who then set Edwin of Deira on the throne. Around 600, in the castle account, Æthelfrith passed Bamburgh to his wife Bebba, and the name Bebbanburh comes from her. His life does not repeat that gift, so the naming stands as the castle’s account of him.

His children Oswald and Oswiu were sent to Iona. Oswald was born about 604 and was killed on 5 August in either 641 or 642. He was king of Northumbria from 634. He had been in exile, and he returned after defeating Cadwallon of Gwynedd at Heavenfield, joining Bernicia and Deira again and favouring Christianity.

The castle account dates that return to 634 and says he brought Aidan from Iona, gave him Lindisfarne, and so began the monastery later famous for the Lindisfarne Gospels. Bede, writing less than a century after Oswald’s death, is the main narrative source for him and treats him as a saint. The inner ward is said to have held St Peter’s basilica, and in it a silver shrine with Oswald’s arm. Another account says the remains of Oswald were kept in that basilica.

A Peterborough story says monks later took the uncorrupted arm from Bamburgh by night and housed it in a chapel in their abbey, with a narrow stair where a monk stood watch. One of the claimed heads of Oswald lies at Durham with Cuthbert. The arm at Peterborough and the arm or the remains at Bamburgh are both in the record, and they are not the same moment.

The excavators of the west ward read Bede as saying that Penda, king of the Mercians, piled a great stack of timber against the foot of the rock and failed to burn the defensive wall. They take traces of earlier timber defences near the later wall as the kind of work that story implies. No year for that attack is fixed in their account.

In 867 a Viking army broke the Northumbrian forces at York and the united kingdom came apart. North of the Tyne a de facto kingdom continued, called Bamburgh after this stronghold and ruled by high-reeves. The lands between Tyne and Tees went to the community of Cuthbert. In 927 Ealdred of Bamburgh accepted the overlordship of the West Saxons, though the country north of the Tees stayed outside the West Saxon shire system until after the Normans.

In 973 Edgar of England acknowledged Scottish sovereignty over the northern part of old Bernicia, by then called Lothian. The castle’s own telling of the later Anglo-Saxon centuries is harsher and shorter. It speaks of attacks from 700 to 993, and of a Viking sack in 993 in which buildings were burned, booty taken, monks killed with axes, and, in return, flayed Danish skins nailed to church doors. It then says the fortress fell into disrepair.

That sack is not given the same detail in the other accounts used here.

The Normans built a new castle on the site, and that work is the core of what stands. Robert de Mowbray, a Norman, was Earl of Northumbria from 1086 until 1095. He joined the rebellion of 1088 for Robert Curthose against William II and was pardoned. He later led the army that killed Malcolm III of Scotland at Alnwick.

In 1095 he joined a plot to move the crown from the Conqueror’s sons to Stephen of Aumale. Most of the conspirators drew back. The crisis broke when Mowbray seized four Norwegian ships in the Tyne, ignored summons to the king’s court, and shut himself in Bamburgh. William II, also called William Rufus, king from 1087 until he was killed by an arrow while hunting on 2 August 1100, brought an army and built a temporary siege castle beside Bamburgh, called Malvoisin, the evil neighbour.

A later note places that siege castle near St Oswald’s Gate. During the siege Mowbray left with a few knights, was pursued to Tynemouth, and after six days was wounded in the leg and captured. He was brought back to Bamburgh, where his wife was still holding out. She surrendered when the besiegers threatened to blind him.

His marriage was dissolved, his lands were given to the new husband of his former wife, and he was imprisoned for life. He died in 1125, having become a monk. The castle then belonged to the English king. One summary says it was built on the Northumbrian defended site in 1095, when it was taken from the Earl of Northumberland, and that apart from a short time in the hands of the Earl of Huntingdon, the son of King David of Scotland, it stayed a royal fortress until the English and Scottish crowns were joined.

The keep is the oldest surviving part of the stone castle. One account says Henry II, king from 1154 until his death on 6 July 1189, probably built it, and that it was complete by 1164. Another says the keep is mentioned in 1164 and that building work was done from 1220 to 1237. The castle’s own account puts a mighty keep in the eleventh and twelfth centuries and calls it the heart of the place.

The walls are described as eleven feet thick on the front and nine feet thick elsewhere, standing on a massive plinth meant to stop miners firing the base. A bottle-shaped doorway let a mounted man enter the keep hall at a gallop. The keep hall is said to have an Anglo-Saxon well sunk through 145 feet of rock. The excavators describe the early medieval well still visible in the keep as circular and about 2 metres across.

Those are a depth and a diameter, not two rival depths. The official list of the building, first made on 4 January 1952 and written both as grade I and as grade 1, calls it a twelfth-century castle of squared sandstone and ashlar, divided into apartments, with a twelfth-century keep and three wards. The grid is given as NU 18319 35114 and also as NU 18373 35053.

In 1138 a Scottish army raided the place. The siege is called unsuccessful, and a wall at the foot of the rock is said to have collapsed, with many casualties. Further unsuccessful Scottish sieges are placed in 1328 and 1333. After the siege of Acre, which ended in July 1191, the castle account says Richard I made Sir John Forster the first governor, as a reward for service.

Richard’s own life, as read here, does not repeat that appointment. Philip of Oldcoates, who died in 1220, was sheriff of Northumberland in 1212 and was given custody of Bamburgh and of Newcastle. He had supported John against Richard, lost Tickhill for it in 1194, and was restored when John became king in 1199. He was captured at Chinon in 1206 and later helped administer the vacant bishopric of Durham.

In the civil war at the end of John’s reign the castle is said to have been under his control, and in early 1216 he held Durham Castle for the king against the rebels. John himself, king from 1199 until 19 October 1216, is named among the monarchs who stayed at Bamburgh, together with Henry III, Edward I, Edward II, and Edward III. Henry VI is named as well, in the time of the roses. Henry III is credited with a great hall and with glass windows, and during his struggle with Simon de Montfort this is said to have been the only royal castle that stayed consistently loyal to him.

Hostages and prisoners were kept in the keep, among them the wounded David Bruce.

David II of Scotland was born on 5 March 1324 and died on 22 February 1371. He became king in 1329 at the age of five, on the death of his father Robert the Bruce, and was crowned at Scone in November 1331. After the English victory at Halidon Hill in 1333 he was sent to France and did not return until 1341. In 1346, under the Auld Alliance, he invaded England to draw Edward III away from France after Crécy.

He had some success at Hexham and was then beaten at the Battle of Neville’s Cross on 17 October 1346, half a mile west of Durham. A Scottish army put at 12,000 met an English army of about 6,000 to 7,000 under Ralph Neville. David took two arrow wounds in the face and was captured by Sir John de Coupland. He was taken to Wark on Tweed and then to Bamburgh, where barber-surgeons came from York to treat him.

In January 1347 he was moved to the Tower of London, later to Windsor, and then to Odiham. He was a captive in England for eleven years, not all of them here. A treaty at Berwick on 3 October 1357 set his ransom at 100,000 marks, at 10,000 marks a year. The castle account says he was held prisoner at Bamburgh after Neville’s Cross, and the family account calls him the wounded king kept in the keep.

The itinerary shows a stop for surgery, then London.

The Wars of the Roses ran from 1455 to 1487 between the houses of York and Lancaster. Richard Neville, 16th Earl of Warwick, was born on 22 November 1428 and was killed at Barnet on 14 April 1471. He is the Kingmaker. In October 1462 Margaret of Anjou landed with French troops and took Alnwick and Bamburgh.

Warwick organised the recovery, which was done by January 1463. Another account says the castles were back in Yorkist hands within three months, and a third says the Yorkists took Bamburgh twice in 1462. Ralph Percy and other rebel leaders were pardoned and left in charge of the recovered castles. In the spring of 1463 Percy besieged Norham.

Warwick relieved Norham in July, but the Lancastrians kept Northumberland for a time. In 1463 French and Scottish forces are also said to have taken Bamburgh for the Lancastrians. In 1464 the Yorkists took it with guns. The castle account calls that a nine-month siege by Warwick, marked by heavy use of artillery, and the first British castle taken by gunfire.

The same event is called the first English castle to succumb to artillery, the first castle in England to be destroyed by gunfire, and the first castle in the world to fall to gunpowder. Henry VI is said to have made Bamburgh his home during those wars. The long sieges of Harlech and Bamburgh were later singled out as fights in remote country, unlike the quicker battles farther south. A tower in the cross wall between the east and west wards is named for Warwick, and a small half-round turret on that wall is said to survive from the older fortress.

The Forster family then provided governors for the crown. After the dissolution of the monasteries, in the mid-1500s, the crown is said to have granted the church and the castle, or perhaps only a lease, to another Sir John Forster. By the sixteenth century the place was in bad repair. In 1610 James I gave it to Claudius Forster, called the last royal keeper, for long service, and it passed into private hands after more than 1,200 years as a royal dwelling.

It soon became an uninhabitable ruin, the family being unable to keep a castle. Those two Forster grants, the mid-sixteenth century and 1610, are both in the record. William Forster died in 1700 and was declared bankrupt after his death. His estates, including the castle, were sold under an Act of Parliament in 1704 to Lord Crew, bishop of Durham, the husband of William’s sister Dorothy.

Another account says that in 1701 what remained passed to Dorothy as the last Forster heir, and that she then married Nathaniel Crewe, and that after her death he founded a charity in her memory to restore the castle and help the village. The marriage is therefore placed both before the 1704 sale and after a 1701 inheritance. Crewe put the castle in the hands of trustees chaired by Thomas Sharp, archdeacon of Northumberland. After Thomas died, his son John Sharp, also archdeacon, repaired the keep and the court rooms and set up a hospital.

The building record says the castle was ruinous when Crewe acquired it in 1704 and was made habitable after his death by Dr Sharpe, trustee of the charity in his will. The spelling is Sharp and Sharpe. The castle’s own dates put Dr John Sharp’s direction of the trustees’ restoration in 1757, after Crewe’s death had left a large sum for the upkeep. The castle then served as a pharmacy, an outpatient surgery, a hospital, and a free school.

Meat, milk, and coal were given to the poor, and their corn was ground at the castle windmill.

In 1786 John Sharp is credited with a coast watch thought to be the first of its kind. A gun was fired from the ramparts in fog, patrols watched the beach, and heavy iron chains were kept to haul ships off. In that year he launched what is called the first lifeboat at Bamburgh. The trustees later ran into money trouble again and offered the castle for sale.

William Armstrong, 1st Baron Armstrong, was born William George Armstrong in Newcastle upon Tyne on 26 November 1810 and died at Cragside on 27 December 1900, aged ninety. He was buried at Rothbury beside his wife Margaret, who had died in September 1893. They had no children. His heir was his great-nephew William Watson-Armstrong.

Armstrong trained as a solicitor, married Margaret Ramshaw in 1835, and left the law for hydraulics. In 1847 his company took land at Elswick. He invented a weighted hydraulic accumulator, and in the Crimean War he designed a lighter rifled breech-loading gun. He gave the patents to the government, was knighted in 1859, and served as engineer of rifled ordnance.

After the government returned to muzzle-loaders in 1862, Elswick sold guns abroad and then, with Charles Mitchell, built warships. Cragside, begun in 1863 with the architect Richard Norman Shaw, is called the first house in the world lit by hydroelectricity. He was raised to the peerage in 1887 as Baron Armstrong of Cragside. His last great project, begun in 1894, was the purchase and restoration of Bamburgh.

The castle account says he bought a semi-derelict building for £60,000, meaning to make a convalescent home for retired men, and that he died in 1900 without seeing the work finished. The restoration is said to have cost more than a million pounds and to have been completed in 1901 by the second Lord Armstrong, the great-nephew, who kept Bamburgh as a family house. The building record dates the high-quality rebuilding by C. J.

Ferguson to 1894–1904. A third account says Armstrong himself completed the restoration in 1905. The finish is therefore 1901, 1904, and 1905, and the man who finished it is either Armstrong or his heir.

The peerage of 1887 ended with him. A new barony was created in 1903 for his great-nephew, William Watson-Armstrong. William Watson-Armstrong, 3rd Baron Armstrong, was William Henry Cecil John Robin Watson-Armstrong, born at Jesmond Dene House on 6 March 1919 and dead on 1 October 1987. He was the only son of the second baron and Zaida Cecile Drummond-Wolff, and he was educated at Eton and at Trinity College, Cambridge.

He succeeded on 6 July 1972. He chose to live at Bamburgh and gave Cragside, with 911 acres, to the government in place of death duties. In 1977 the house passed to the National Trust through the National Land Fund, and he gave an endowment. On 16 August 1947 he married Maria-Teresa Chiodelli-Manzoni.

They had no children of their own and adopted a son and a daughter. At his death in October 1987 the peerage ended again, and the Bamburgh estate passed to the adopted son, Francis Watson-Armstrong. The castle account calls Francis the son of the third baron and the present owner, without the word adopted. The family still lives there and opens the castle to visitors.

In the Second World War, pillboxes were set in the dunes against invasion, and in 1944 a Royal Navy corvette was named HMS Bamborough Castle, using the older spelling. After the war the castle was listed. The description of 1952, repeated in the later record, says it was twelfth-century, ruinous in 1704, made habitable by Dr Sharpe, and rebuilt by Ferguson for Lord Armstrong. The east entrance is largely twelfth-century, with round towers beside a tunnel-vaulted arch.

The inner gateway has a twelfth-century vault and is largely nineteenth-century above. Ferguson’s east-ward ranges include the Captain’s Lodge and the King’s Hall, in an elaborate Perpendicular manner, with other work in a quieter early Tudor style. A twelfth-century chapel with an apse stands in ruins in the ward. The keep has a tall moulded plinth, a twelfth-century ground-floor doorway of two round arches, angle pilasters that rise into turrets, and many round-headed windows that are mostly eighteenth-century.

Inside, the ground floor has two parallel tunnel vaults on huge square piers, a mural stair, and an armoury that may have begun as a chapel with a vaulted apse. The Captain’s Lodge has a fourteenth-century barrel-vaulted basement of ten chamfered ribs, now a shop, and above it Ferguson’s stair with Art Nouveau balusters. The King’s Hall and the Cross Hall stand on the medieval hall, whose three service doors remain. The hall has two large oriels, a fireplace with a joggled lintel, and a false hammer-beam roof of six bays, carved in teak.

The castle account says that teak came from Thailand. Pantry and buttery keep high pointed tunnel vaults. The kitchen has three great segmental fireplaces and four pointed medieval doorways, two of them blocked. The Smith Gate is largely nineteenth-century.

The Neville tower keeps medieval vaulting. Nineteenth-century apartments run along the south side and end in a round clock tower that is medieval in its lower parts. The west side has late nineteenth-century stables of five bays with octagonal corner towers. An outer ward to the west holds the windmill and pieces of medieval curtain wall.

The King’s Hall is called a Victorian masterpiece on the site of the medieval great hall. The armoury in the keep shows pikes, halberds, muskets, a fifteenth-century composite bow, and a seventeenth-century Flemish target crossbow. The cross hall has a copy of The Card Players by Theodor Rombouts, also spelled Theodoor Rombouts. The library holds books of the mid-seventeenth and mid-eighteenth centuries, added to by Armstrong.

The inner ward’s battlements enclose the site of the Anglo-Saxon basilica, and the ruins of a twelfth-century church remain on that spot. The battery terrace, above the gate used by horse and ox carts, was armed with cannon against Napoleon, and medieval arrow slits remain in the gate towers. The Victorian stables in the east ward were built for Armstrong.

Brian Hope-Taylor was born on 21 October 1923 and died on 12 January 2001. He was an artist, a wartime model-maker for aerial photographs, and then an archaeologist. He took a doctorate at Cambridge in 1961 on Yeavering, without having first taken a degree, and he published that excavation in 1977. He dug early Anglo-Saxon sites at Lindisfarne, at Doon Hill near Dunbar, and at Bamburgh, where he found the Bamburgh Sword in 1960.

His campaigns here are also dated 1959 to 1961 and 1970 to 1974, and more loosely to the 1960s and to the late 1960s and early 1970s. Most of the work was in the west ward. He found the Bamburgh Beast, called a gold plaque and also a gold mount with interlaced animals, pattern-welded swords, and hundreds of styca coins. A re-excavation of his 1960 trench says that trench produced two pattern-welded swords and an axe, and also Roman and prehistoric pottery and Neolithic flint.

He published only two short interim notes. The records and many of the finds stayed unavailable until he died. Plastic sheet and fertiliser bags laid in 1974 protected what he had not dug to the bottom. His papers went to Edinburgh, and his site huts beside the trench still held tools, soil samples, and boxes of bone.

Black plastic, grid pegs, and an early mortar mixer he called the gin-gang were still there when the later diggers opened his cuttings. The mixer is evidence of mortared stone buildings in the early medieval period, and stone buildings of that age were found in the inner ward in 2004 and 2007.

The Bamburgh Research Project was formed in 1996 by archaeologists in the north-east, among them Graeme Young, with permission from Lady Armstrong. A heritage summary dates a Bamburgh Project, under a different practice, to 1997, and a survey note dates the research project to 1996–97. The first aim was the burial ground Hope-Taylor had not found. An early map marked a Danish burial ground.

A storm in 1817 had already shown skeletons to antiquaries. The cemetery, the Bowl Hole, lies in the dunes south of the castle, and a later note places it 300 metres south of the castle gate. Digging from 1998 to 2007 found the remains of 120 people of the seventh and eighth centuries. Charlotte Roberts led that research from Durham.

She was born on 25 May 1957, was professor of archaeology at Durham from 2004 until she retired in October 2020, and is professor emeritus. The bones were read as people who had come from Ireland, Scotland, Scandinavia, the Mediterranean, and North Africa. In 2016 they were moved into the crypt of St Aidan’s church in the village, where visitors can see the crypt through a small gate. Post-excavation is said to have lasted ten years under Charlotte Roberts and Durham.

In February 2026 a monograph of 310 pages was published, The Bamburgh Bowl Hole Anglian-period Cemetery, edited by Sam Lucy, Charlotte A. Roberts, and Graeme Young. Sam Lucy of Cambridge was advertised to speak on it to the Society of Antiquaries of London on 26 February 2026.

Inside the castle, trench 3 in the west ward began in 2001 as a trial trench 30 metres long, looking for Hope-Taylor’s edges. It was finished in 2020. The sequence runs from modern soil through a later medieval midden into timber buildings on rubble foundations, halls and workshops, cobbled paths, and thick ash from industry. A stone-founded timber building of the later ninth century held a hoard of stycas, with more coins scattered outside, and iron objects nearby included two pattern-welded swords and small pieces of decorated gold.

Hammerscale showed high-temperature welding of the kind used for weapons. A metalworking building is placed broadly in the ninth century. Animal bone from the layers is the evidence for diet. In the deepest part of that trench, in 2020, the diggers found part of the stone footing of a roundhouse more than 10 metres across, with some floor surfaces left and an entrance facing broadly south-west, which they read as Romano-British from the objects and the sequence.

Samples for radiocarbon dating and a block of the floors for microscopic study were still to be processed when that account was written. The castle’s news of the same kind of discovery says volunteers, at the end of a summer dig, went a few inches farther and found a substantial roundhouse about 12 metres across, which the position of the rubble suggested could be more than 2,000 years old and of the Roman occupation. The director called it one of the most important finds at the castle since digging began in the 1960s, and a rare northern roundhouse with a well-preserved run of floors. He placed Bamburgh in a military zone north of Hadrian’s Wall, where client leaders were paid to keep the peace, and he said the rock’s story goes back at least 3,000 years, with almost no written record before the kings.

Further sampling and carbon tests were to be paid for by the Armstrong family. The two diameters, more than 10 metres and about 12 metres, and the two datings, Romano-British and possibly more than 2,000 years old, are both published. No laboratory date for the roundhouse is stated here.

Trenches 1 and 2 were the project’s first cuttings inside the castle, at the north end of the west ward. Trench 2, at the tip, had been eroded before the modern wall. Trench 1 grew into an open area from the wall to St Oswald’s Gate and was finished in 2017. A monastic notice of 774, copied by Symeon of Durham in the twelfth century, says the steps to St Oswald’s Gate were the early entrance.

The gate is also called the principal, and perhaps the only, entrance in the eighth century, until the main gate moved to its present place in the twelfth century. Late eighteenth-century flags were lifted and showed at least two earlier surfaces, with bedrock worn smooth beneath. Two phases of a hall stood by the gate, a timber building dated broadly to the seventh century and a smaller stone hall abandoned by the eleventh century, when its stone was robbed. The place would suit a gate-warden.

West of the halls were traces of timber defences.

Trench 8, dug in 2006, reopened Hope-Taylor’s 1960 cutting, the one with the two swords, down to deposits above the bedrock. Below the early medieval layers were Romano-British features and Roman pottery, and below those, metalled prehistoric surfaces. In 2017 the Royal Archaeological Institute paid for the post-excavation of that trench. Trench 9 reopened his cutting C, which ran from the gate steps toward the windmill.

In one standing section a partial horse and a cow had been buried over the robbed medieval steps in the early post-medieval period. The diggers note that the Forsters held the estate from the start of the seventeenth century to the middle of the eighteenth and leave open whether the animals belong to that time.

The present digging is at St Oswald’s Gate and the outworks beyond it, a cleft that has been a way up from prehistory. The phase began in the summer of 2021 and centres on the Tower of Elmund’s Well, named in thirteenth-century records, with a cottage built into it in the late eighteenth century. In 2022 two splayed window openings appeared in the walls. Four builds show in walls that reach almost two storeys: the lower two are read as the medieval tower, the upper as the cottage.

Cartographic evidence put the well in the basement. In the 2024 season the stairs were cleared and the well itself was found. It is not a round wishing well. It is a square stone-lined cistern taking up much of the basement floor, comparable in size to a square well at Rainhill that measures about 2 metres each way.

A short flight of two steps drops about half a metre, and more fill remained. The south side is rough stone, read as a porous wall letting water in. The east and north sides are neater and later. A stain marks an older waterline, a little below the level at the time of a wet winter dig, and a circular hole runs more than a metre under the platform, read as a drain to the seaward side.

Large squared timbers lie on three sides. On the east they run under the stone wall, so they are not only a lining. They may be an earlier timber well, and they may be older than the tower. They sit under the water table, in conditions that can preserve wood.

A later look, when the water was lower, showed that the timbers continue under masonry thought to be of the thirteenth or fourteenth century, on a different alignment, and so of another phase. A ninth-century date has been suggested because the name Elmund is most at home in Old English sources of that time. No tree-ring date is stated. The money for that dating was still being raised, and the diggers hoped to sample in July 2025.

The well is thought to have served a small port immediately to the north, now silted, and the outworks to have enclosed both the water and the way in. A wall at the base of the rock is mentioned from at least 1138. The stone outworks now visible are placed mostly in the thirteenth to fifteenth centuries, with changes in the eighteenth and nineteenth when the cottage was built. The postern toward the village passes through a thick medieval wall, several metres high in places and about 3 metres high where the present gate goes through.

A partly blocked hole for a large squared timber bolt survives. The path, some of the door jambs, and the outer stairs are late post-medieval, and the wall may have been lowered on purpose to open the view to the gate. A second postern, found walled up, has neat courses and was narrowed at least once. The gap between one pair of faces is a little more than a metre.

Square postholes about 30 centimetres each way held timber uprights. Outside, the threshold can stand as much as 2 metres above the ground, which implies an outer stair, perhaps of timber that could be taken away in a siege. The eastern opening is read as the older, high enough to need a timber stair, and the western one, still in use, as the later and lower, though its present stairs look late post-medieval. Much of what is exposed is late, and the medieval wall is expected lower down.

A narrow wall west of the living postern, once thought to run on, is now suspected to be a Victorian folly. Between the walls a rubble foundation of large dolerite slabs and beach pebbles matches the footing of an earlier cross-wall cut by the wide medieval wall, and a round-headed arch toward the tower and perhaps the port belongs to that earlier phase.

In July 2023 a masonry survey, photogrammetry, and geophysical survey were added. St Oswald’s Gate itself is a vaulted tunnel of two medieval builds, refaced inside and out, probably in the late eighteenth century. The wing wall is narrower than the later defences, has at least four phases, and contains an arched doorway of early form that may be reused. The south wall is at least two builds and contained two posterns.

On the cricket ground to the west, magnetometry, resistivity, and ground-penetrating radar point to a large ditch some 45 metres across, where a castle ditch would be expected. A rock-cut ditch in sandstone can be followed at the modern entrance, the site of the twelfth-century gate, as far as the sandstone ridge. The radar signal dies before a true depth. Tracks on an 1865 map reappear: one to St Oswald’s Gate, one to a cleft called the Miller’s Nick, used in the nineteenth century to scramble up to the west ward, an S-shaped path that may skirt the ditch, and a way parallel to the village road called the Wynding, which records suggest was medieval.

Early in the 2024 season, outside the tower on the seaward side and in a relatively modern layer, the diggers found a silver penny of the tenth century. The cross is small and appears on both faces, with no royal portrait, which suggested a date before Edgar’s reform of the coinage in 975. The letters barely form words. A coin specialist read it as irregular, a forgery or the work of a moneyer outside the regulated mints, where bad coin was punished severely.

A matching piece comes from a hoard buried on Iona around 990 and now kept in Edinburgh. Another notice calls it an early to mid-tenth-century penny and says some scholars would value a penny of that age at about ten pounds or even twenty pounds in modern money, with a sheep at five pence and a cow at twenty. Because the coin was not in its original layer, it dates activity in the area only loosely. In the well fill, just above the water, was a cast-iron fragment with the name Boston cast into it, identified with Boston and Skirbeck, makers of water pumps in the nineteenth century, and likely part of a pump used when the cottage still drew water here.

Two cannonballs lay close together. Corrected a little for corrosion, the smaller is read as a 9-pound shot and the larger as an 18-pounder, sizes common in the eighteenth and early nineteenth centuries. Records say the castle was equipped with two 18-pounders, but only one of those guns could be traced, and no 9-pounder is recorded, nor any firing on the castle in that period.

A field school was advertised for 19 July to 14 August 2026, the thirtieth year since the project began, to go on with the gate and the well. Mounds beyond the outworks were flagged as possible remains of the siege castle of 1095, to be looked at if time allowed. No result of that season is stated in the accounts read here. In the summer of 2026 a youth group visited the well, which was not open to the public, and noted old graffiti, freshwater still in it, and animal bone and shell from the trenches, including bone from a ditch they were told had later been filled with rubbish.

That visit does not publish a new date or a new structure.

The castle was its own civil parish. It had been a township in Bamburgh parish, was a parish from 1866, and in 1951 had a population of 18. On 1 April 1955 the parish was abolished and merged with Bamburgh.

Stories have collected on the rock. The ballad of the Laidly Worm of Spindleston Heugh is placed around 1270. Thomas Malory identified Bamburgh with Joyous Gard, also spelled Joyous Garde, the castle of Lancelot. Bernard Cornwell, born on 23 February 1944, used the old name Bebbanburg for the home of Uhtred in The Saxon Stories, thirteen novels published from 2004 to 2020, after he learned that a historical Uhtred of Bebbanburg stood in his own ancestry.

The series follows the making of a single English kingdom under Alfred and his heirs. Television adapted it as The Last Kingdom. The castle also appears under its own name in a modern crime novel and has been used as a filming place, but those uses are later than the fortress.

What the place was for changes with the centuries, and the sources support each use only for its own time. It was a Brittonic stronghold and perhaps a capital, an Anglian royal seat and a centre of metalworking, a shrine of Oswald, a Viking target, the seat of the high-reeves of Bamburgh, a Norman earl’s castle, a crown fortress and palace, a prison for a wounded king, a gun-ruined Lancastrian hold, a Forster house, a charity hospital and school and coast watch, and then the Armstrong family’s residence and a place the public pays to enter. The digging has not closed the gaps. The roundhouse still waits on a laboratory date.

The well timbers still wait on tree rings. The 1464 guns, the 1095 blinding, and the name Bebba are as secure as the pages that state them, and where the pages disagree the disagreement is part of the history.