Like so many other castles there is no evidence for when or by whom Elmley castle was founded.  The castle may have been first built by the Abetot family and Urse Abetot (d.1108) was directing the construction of Worcester castle in 1069.  Tradition says the site was first fortified in the middle ages by his brother Robert Despenser Abetot, who died childless before 1100.  At some point after 1095 his lands had passed to his brother, Urse Abetot (d.1108).  At Domesday Robert had held the unnamed Elmley as a part of Crophorne and Netherton from the bishop of Worcester.  At least half of his lands passed to Urse and possibly then his brother Osbert, although after 1113 Urse's son Roger inherited the family estates.  In 1114 Roger was banished and his lands and office of sheriff passed to his sister, the wife of Walter Beauchamp (d.bef.1133), possibly disinheriting other younger brothers of Urse.  Before 1115, King Henry I informed the bishop, sheriff and barons of Worcester that none were allowed to poach Walter Beauchamp's pheasants at his manor Elmley (Almega) or elsewhere.  Even though Elmley castle is not actually mentioned until 1216 its design indicates that it was probably founded before 1100 and likely soon after the Norman Conquest to help dominate Worcestershire.  It therefore is distinctly possible that Robert Despenser was its founder.

After 1114, the Beauchamp family would appear to have made Elmley castle their caput and as such became known as Beauchamp of Elmley, to differentiate themselves from many other cadet families, like the Scottish Beauchamps, the Bedford Beauchamps and the Beauchamps of Hatch.  Under their ownership the vill and therefore castle was recorded under the names Almega, Aumel, Elmeleg, Elmesleg, Hammele, Hammeleg and Haumel.  During the Anarchy (1136-54), William Beauchamp (d.1170), the son of Walter (d.bef.1133), was lord of Elmley.  Initially he supported the cause of King Stephen (d.1154) and was rewarded with the constableship of the rebel Miles Gloucester (d.1143) on his defection to the Empress Matilda in 1139.  However, William proceeded to argue with the king over Count Waleran of Meulan (d.1166) being made earl of Worcester and being granted Worcester castle whose custodianship had previously been with the hereditary sheriffs of Worcester, the Abetots and then the Beauchamps.  As a consequence in July 1141 at Oxford, the Empress made William the hereditary constable of Worcester castle and sheriff of the county just as his father Walter had been.  William also pledged allegiance to the Empress against all mortals and especially against Count Waleran of Meulan (d.1166). 

William became my liege man against all mortals and particularly against Count Waleran of Meulan, in such a manner that neither Count Waleran or those aforesaid may make a fine with me whereby William should not hold of me in chief, unless of his own free will and consent he wishes to hold of the count.

Further, for his allegiance he was granted the honour of Tamworth to hold just as freely as Robert Despenser and Urse Abetot had held it.  At this point Count Waleran was in London and then Kent with Queen Matilda (d.1152), the wife of King Stephen, in opposition to the Angevins.  He later made his peace with the Empress in March/April 1142 and was recognised by her as earl of Worcester.  However, Waleran now withdrew to Normandy and there is no evidence he ever returned during the disturbances, but left William Beauchamp in charge of Worcester, nominally in his name.  In early 1146 he sent an order to Beauchamp, which rather suggests that this was the agreement reached between the 3 at Devizes.  In 1149 it was confirmed that Walter Beauchamp and William his son had granted a tithe of their wine at Elmley, which in itself suggests that this was already a caput of the family. 

In 1150 King Stephen attacked Worcester city and burned it, although the castle, under the men of Earl Waleran Beaumont of Worcester, held out.  The king then ‘overran the territories of the hostile lords and, no one resisting them, carried off an immense booty'.  Presumably one of these hostile lords was William Beauchamp of Elmley.  The next year, 1151, Stephen returned;

having been unable to reduce the castle the year before, he now assaulted it with the utmost determination.  As the garrison made a manly resistence, he constructed two castles to expel them.

According to one chronicle, William Beauchamp was captured at this time by Ralph Manderville [probably of Marshwood, Dorset].  In any case, during this year, 1151, Earl Robert Beaumont of Leicester (d.1168), Earl Waleran's younger twin brother, appeared on the scene and led his brother's forces in ‘demolishing the 2 castles' and breaking the siege.  This was despite the fact that he was still nominally a royalist.  It is possible that one of these ‘siege' castles may have been Elmley as it was only 13 miles away.  The other castle may have been on the Herefordshire Beacon some 10 miles away.  In 1136 it was thought that it was unnecessary to build a siege castle at Harptree as the royal garrison at Bath was sufficient to play that part.  Harptree castle is 13 miles from Bath.

The fall of the 2 siege castles would appear to have left Worcester in the hands of the men of the nominally Angevin Earl Waleran.  In 1153 Waleran's knights of Worcester castle were accused of seizing and imprisoning Sheriff William Beauchamp ‘by trickery'.  If this is correct it shows that the unlucky William had been captured twice in 2 years.  It is uncertain exactly what happened, but Beauchamp and Earl Roger of Hereford (d.1156) had worked together for the Angevin cause since the early 1140s.  Count Waleran was at least nominally pledged for the Empress' son, Duke Henry Plantagenet of Normandy (1133-89) as too by this point may have been his brother, Earl Robert of Leicester (d.1168).  The Angevin Earl Roger of Hereford now approached King Stephen and offered to join him in reducing Worcester in ‘a pact of inviolable peace and friendship' and to rescue his friend, Sheriff William Beauchamp.  King Stephen agreed to the proposal and the allied forces of king and earl of Hereford besieged Waleran's men within Worcester castle.  In an act of duplicitous treachery Earl Roger then informed Duke Henry of Normandy (1133-89) of his actions and asked him to come over to England at once.  King Stephen, who was apparently aware of Earl Roger's actions, therefore left the siege, although he left part of his army to aid the earl.  Worcester castle was duly reduced, but after this Earl Roger promptly declared again for Duke Henry, breaking his pact with King Stephen.  This act appears to have finally extinguished the Beaumont earldom of Worcester, though once again the status of William Beauchamp is unknown, although presumably he was released from his captivity within Worcester castle.  However in 1153, King Stephen took Tamworth castle from Beauchamp and restored it to Robert Marmion.  King Henry II (d.1189) did not overturn this action of his enemy once he became his heir and then successor.

William Beauchamp (d.1170) remained loyal to the Angevins and in 1155 was with the royal army besieging Hugh Mortimer (d.1181) at Bridgnorth castle.  During the Anarchy William had married Bertha, the daughter of William Braose (d.1180, of Radnor, Buellt and Bramber) and Bertha Hereford the daughter of Earl Miles of Hereford (d.1143) and lord of Abergavenny.  On his death he was buried at the door of Worcester cathedral chapter house.  His son died young in 1197 in Normandy and his grand uncle, William Braose (d.1211) bought his custody.  William Beauchamp came of age in 1207 and died in 1208, possibly in the Braose war of that year.  At this point his Uncle Robert bought the custodianship of William's younger brother, Walter (d.1236).  In 1212 he came of age and married Joan (d.1225), the daughter of Roger Mortimer (d.1214) of Wigmore.   During the civil war of 1215-17, Walter was appointed sheriff of Worcester on 19 August 1215 in the aftermath of the Magna Carta negotiations.  Later that year he obviously fell away from John after 2 February 1216 and joined the rebellion.  On 2 August 1216 Walter received a safe conduct to visit the king.  This meeting obviously took place and on 8 August the king wrote to William Cantilupe (d.1239) informing him that he placed Walter Lacy (d.1241, of Longtown, Ludlow), Hugh Mortimer (d.1226, Wigmore), Walter Clifford (d.1263, Clifford, Bronllys) and John Monmouth (d.1236, Monmouth) to the custody of Walter Beauchamp's castle of Elmley and his lands until the excommunicated Beauchamp can get to the papal legate and receive absolution.  At this time the aforesaid barons would return his castle and lands to him.  As John sent this letter from Oswestry after his Marcher campaign, it would seem likely that Beauchamp had crumbled before the king's show of force in the Welsh borders, Elmley castle being an obvious target of the king's ire.  On 19 August 1216, Walter's seneschal, Richard Vincent was in the custody of Constable Hugh Vivon of Berkeley castle while the king sojourned at the fortress.  From here onwards the Beauchamps proved loyal to King John's son and heir, King Henry III (1216-72).

In 1234 King Henry III sent deer to Walter to stock the park he had created at Elmley.  Thirty years later war nearly came to Elmley castle.  During the Barons' War William Beauchamp took the part of the king.  On 5 May 1263, he was ordered to join the royal army forming at Worcester and in 1265 his sons took part in the battle of Evesham which was fought within sight of Elmley castle, although the fortress is not recorded as taking any role in the fighting.  However, the castle's time as caput for this middling baronial family came to an end in 1268 when William Beauchamp (d.1269) became earl of Warwick and transferred the family seat to Warwick castle.  When his son, Earl William Beauchamp of Warwick, died on 9 June 1298 Aumelegh castle was found to be in need of much repair.  In 1308 a chantry was founded in the castle chapel that was to be served by 8 chaplains and 4 clerks endowed with £20 from the manor of Childs Wickham.  In 1311 the rectory of Elmley had been added to the endowment, but it still proved too expensive to run so was reduced by a chaplain and 2 clerks.  In 1312 the captive Piers Gaveston was brought here before being moved to Warwick castle.

When Earl Guy Beauchamp died in 1315, Aumeleye castle was only valued at 6s 8d although there was a garden within the fortress and fisheries in the vicinity.  A second valuation found the castle as worth nothing.  As Guy's heir was only 1½ years old the king granted the castle to his executors on condition that they did not grant it to any others without the king's express licence.  These descriptions show that the castle was not actually derelict, but that there was no income to support it.  This implication is proved by what happened next.

On 1 December 1316, the king granted Elmley castle to the executors of Earl Guy Beauchamp's will during the minority of the heir, but had on 2 August 1317 transferred the custody to the senior Hugh Despenser (d.1326) in part payment of the debt owed him, the castle and manor having been extended at £49 12s 10½d per annum.  On 1 November 1317, Hugh was ordered to have the castle securely guarded and then on the 22nd to place 20 fencible footmen in Elmelegh castle and retain them at the king's wages until further orders, paying the wages out of the farm of the castle.  Presumably these men were still at the castle on 4 June 1321, when the sheriff of Worcester was ordered to take the body of the castle of Elmley into his hands from those of the senior Hugh Despenser and to have the fortress safely guarded.  This was on the exile of the Hugh Despensers as was forced on the king that summer.  The sheriff was also to make an indenture of the arms, victuals and other goods found there before a jury of 12 men.  However, the king soon brought the Despensers back with the result that after Christmas 1321 the kingdom was rocked by an anti-Despenser riot which caused the king on 15 January 1322, to write to Constable Richard Lovel of Bristol castle ordering him to arrest and imprison Earl Humphrey Bohun of Hereford and Essex (d.1322), Roger Mortimer of Wigmore (d.1330), Roger Mortimer of Chirk (d.1326), Roger Damory (d.1322), Hugh Audley the elder (d.1325), Hugh Audley the younger (d.1347), Bartholomew Badlesmere (d.1322 [Leeds]), John Giffard of Brimpsfield (d.1322 [Carreg Cennen, Clifford]), Maurice Berkeley the elder (d.1326), Henry Tyeys (d.1322), Roger Clifford (d.1322 [Appleby, Brough, Brougham, Pendragon, Skipton]), John Wylington (d.1340+), John Mautravers (d.1364), Henry Wylington the younger (d.1337+), Robert Watteville (d.1330), Gilbert Ellesfield, Gilbert (d.1346) and Richard Talbot (d.1356) and all their company as they had taken by night and burned Bridgnorth town and attacked the king's servants there who were there purveying for him... and who afterwards took in a like manner the castles of Elmley and Hanley which were in the king's hands, beat and wounded his men and servants therein, took some away and slew others and burned the gates and houses of the said castles, detaining the town and castles against the king and they stole garments, jewels, beasts and other goods and chattels of the king's men, slaying certain men and detaining others in prison until they made grievous ransoms. 

After the defeat and in many cases the execution of the last rebels in March 1322, on 13 April 1322, the constables of all the royal castles, including Elmley, were ordered:

To remove from that castle the munition of men that the king lately caused to be put therein by reason of the late disturbances in the realm and to keep the castle in the same way as before the disturbances and to cause the king's victuals therein to be kept safely and to cause the victuals that will not keep to be sold and cause more to be bought in their place...

The castle apparently played no further part in Edwardian politics, although on 28 May 1322, it was recorded that John Wyard, Roger Wassebourn with Robert Harley [Brampton Bryan] and others had lately entered the manors of Sheriffs Lench, Elmley, Comberton and Salwarp in Worcestershire, then in the king's hands and committed divers trespasses there.  Elmley then remained as a secondary fortress of the Beauchamp earls of Warwick until 1446 when the male line came to an end.

Elmley castle then passed to the Neville earls of Warwick, Duke George of Clarence and finally, in 1487, was subsumed by King Henry VII (1485-1509).  During this period the fortress was maintained with slight repairs to the structure  recorded in 1413 and 1425 under the Beauchamps.  In 1463 Earl Richard Neville of Warwick (d.1471), otherwise known as Warwick the Kingmaker, obtained a royal licence to grant land to the value of 20m (£13 6s 8d) to the warden of the chantry in Elmley castle towards hiring an additional chaplain.  For similar purposes his son in law and successor at Elmley, Duke George of Clarence (d.1478), before meeting his end in the Tower of London in a butt of Malmsbury wine, gave the manor and advowson of Naunton Beauchamp.  Further repairs under the Crown were carried out in 1480, during the minority of Earl Edward of Warwick and in 1492.   Walter Walshe was appointed royal constable in 1528 and held the office for 10 years when Urian Brereton took over.  Soon after John Leland rode by and found:

Ther stondithe now but one tower, and that partly broken.  As I went by I saw carts carienge stone thens to amend Persore (Pershore) Bridge about ii miles of.  It is set on the tope of a hill full of wood, and a townelet hard by.

In 1544 a survey discovered that Elmley castle was strongly sited upon a hill and was surrounded by a ditch and wall, but the castle was now completely uncovered and in decay.  Since then most of the remaining stonework has been removed.

Elmley castle stands on a ridge on the irregular northern slopes of Bredon Hill at about 600' above sea level.  The hill summit, less than a mile to the north-west, is some 980' high and is enclosed by a true hillfort of 2 or 3 phases and which has returned C14 dates of second century BC down to about 100BC.  These defences consist of a scarp on the 2 steepest sides to north and west with a simple rampart and ditch forming a rectangular enclosure to make the other 2 sides.  A second rampart and ditch virtually doubles the enclosed area of the fort.  A third 2 phase hill fort lies just over a mile south of Elmley at Conderton on a south running ridge.  This rectangular ditched enclosure has been excavated and has provided good Middle Iron Age (400-100BC) dating material.  Elmley castle is therefore dissimilar to both these sites, lying as it does on a shelf half way down the northern scarp of Bredon Hill.  It is also considerably larger and much better defined.  However, it also shows evidence of being a multi phase structure.  Consequently, although Elmley castle is reckoned to be a hillfort, no finds have been made from this era.  It is therefore quite possible that all the current earthworks are medieval.  These consist of 3 or more distinct structures. 

Early Possible Ringwork
The possibly original fortification at the site consists of what may have been an early ringwork at the northernmost apex of the site, it being about 130' in diameter.  At the Apex was a rectangular masonry structure pressed hard against the curtain wall standing on a rampart and making a right angled turn at this point.  Much rubble and a wall lying on its side can be made out here as well as great views of the plain to the north.  Possibly this is the site of the castle hall.  The possible ringwork rampart has been overlain to north, east and west by the rampart and curtain of the main ward, but to the south what appears to be the mostly rubbed out line of the ringwork can be followed.  At its southern extreme are the slight remains of a great tower keep which would appear to be projecting from the rampart of the original ringwork.  The western corner of the original ringwork can still be made out with the later ringwork rampart running away from it to the main gate to the south-west.  If this is an early ringwork, then the rest of the ‘ringwork' enclosure to the south is the outer bailey.  Presumably the early castle was massively expanded at this point and the old southern ringwork rampart may have been taken down when the keep and rest of the enclosure to the south was built in stone.  Certainly there appears to have been a masonry structure between the keep and the main rampart to the north-west.

The Main Ward
In the northern half of the main enclosure or ward the defences consist of buried curtains running south-east and south-west from the ‘hall' at the northern apex of the site.  Both curtains are relatively straight and end where causeways cross the surrounding ditch.  The southern part of the ward, about a third of its size, consists of an irregular semi-circular return linking the 2 gates.  The whole has an approximately 10' deep ditch surrounding the rampart, while the interior of the main ward is some 5'-10' higher than the ground level in the outer bailey.  Together, including the possible early ringwork, the inner ward forms an eye drop shaped enclosure, the more circular southern portion being about 320' in diameter with the possible ringwork added to the north to make a north to south distance of some 440'. 

The Outer Ward
A sub rectangular bailey covers the main ward to the south and some half of its east and west sides, being some 300' deep north to south.  This may once have been an abortive town settlement.  On its 800' long southern side the outer ward is covered by a ditch and rampart with curved corners to south-east and south-west.  Roughly central along the long south front is a fishpond lying just south of the outer ditch.  To the west of this a causeway crosses the dip the fishpond lies in and passes over the ditch and in through a slightly interned gateway.  Possibly this entrance is prehistoric.  The northern front of the outer ward, west of the main ward, uses the scarp of the hillside for much of its protection, although again a ditch and rampart is present.  A second ditch and rampart lies further down the hill to the north and links to near the north apex of the inner ward ditch.  This is also claimed as prehistoric without substantive evidence.  To the south-east of the inner ward the outer ward defences make a straight line to the north-east and then perform a right angled turn at the start of the main hillside scarp, a powerful ditch running back to, but not penetrating, the inner ward ditch just north of the east gateway.  The dog legged turn is in turn covered by an ear-shaped earthwork, about 70' east to west, but 140' north to south.  From the north-west corner of this a fine ditch and rampart runs off to the north-west, following the line of the inner ditch, but about 60' out.  This curves around the north apex of the inner ringwork and, using the scarp of the hillside, curves back to join the north-west corner fo the outer ward, covering the north side of the inner ringwork.  The trackway leading from here back down to the village and parish church is most likely medieval.

The Masonry Remains
The main ward is protected by a rampart up to 10' internally and spreading some 25' deep.  Whether this is an original earthen rampart or is merely the debris spreading out from a collapsed curtain wall is uncertain, but as the tower to the south has been partially excavated to its basement level the latter seems more likely.  It has been claimed that the entire inner ward is prehistoric as the western entrance is slightly interned in Iron Age hillfort style.  However, it is equally possible that the defences here have been damaged when the site was quarried for stone in the late sixteenth century.  Certainly the causeway on this side is probably post medieval as too is the earthwork forming a possible barbican before it and running along the scarp of the hill to the west. 

All around the inner enclosure the remnants of the main curtain wall can be seen protruding through the turf, showing that originally this was a powerful ringwork castle surrounding a large great keep, some 80' by 60' (with a powerful 30' deep forebuilding), set somewhat off centre to the east.  The keep was investigated as were the south tower and northern apex of the site in the 1930s.  This has shown that much masonry still exists just beneath the surface throughout the site.  Little can be said though of the design or date of these structures, though comparisons may be made with other ringwork castles with large great towers, viz. Castle Acre, Castle Rising, New Buckingham, Duffield, Hedingham and Rochester.  There was once a rectangular building backing against the south tower and the clump of trees between it and the keep is reckoned to be site of the castle chapel and chantry.


Copyright©2021 Paul Martin Remfry