Held at:





Internet – Archaeology Data Services [University of York]


Earthwork Castles of Gwent and Ergyng AD 1050-1250: Ewyas Harold Castle

Place name:

Ewyas Harold


1050 - 1250


This summary concerning Ewyas Harold Castle is an extract from a Doctoral thesis by Dr N Phillips, University of Sheffield (2005), entitled ‘Earthwork Castles of Gwent and Ergyng, AD 1050-1250’. This thesis is concerned with the earthwork and timber castles built in the southern March of Wales, addresses the presence of the castles and discusses their roles as weapons of conquest and structures of administrative control.


EWYAS HAROLD - Grid: SO 38502 28699




The village of Ewyas Harold is located to the west of the A465, Hereford to Abergavenny road, 31km north of Abergavenny. The castle occupies the south-east end of a spur of high ground to the north west of the village.


The visible remains of the castle today are quite substantial: a large heavily wooded motte with bailey, ramparts and ditch (Vol. 2. plate 97). Streams surround the motte on all sides except the north-west where the end of a ridge has been modified by the creation of a defensive ditch which cuts the motte off from the rest of the hill (Vol. 2. plate 98). The west, north-west and north sides of the motte have evidence of quarrying which has seriously destroyed the motte top on those sides. There are also discernable earthworks on the top of the motte which suggest sub-surface remains of buildings. To the south and east of the motte is the bailey which consists of a flat area of land itself some height above the surrounding valley and town. Running from the north-east of the motte towards the bailey edge are the remains of a cross-rampart (Vol. 2. plate 99). A similar rampart may once have existed to the south-west of the motte but this has been almost completely destroyed by farm buildings (Vol. 2. plate 100). A modern bungalow with landscaped garden occupies an area of high ground to the north of the motte.

Topographic survey: (Survey 17)

The most interesting detail revealed during the survey was that the 16.36m of height attributed to the motte only managed to raise it some 6m above the hill from which it was cut; by the excavation of the defensive ditch. Consequently, the large motte has only about 6m of height resulting from artificial modification, the other 10m being natural hillside. Survey of the motte top revealed that the present plan area was a sizable 384.839. However there is considerable damage to the motte rim resulting in angular and concave cuts made by quarrying. By predicting the original motte rim by rounding out the damaged areas it is possible to estimate an earlier plan area of 498.017 m².

The bailey area is also quite large with a surface area of 6924.617m² which is raised above the south-east fields by 9.45m, again using the natural hill rather than raising an artificial mound.

Strategic position:

The site probably offered obvious natural defence possibilities with the end of a steep ridge which the site has used to good advantage. There is, however, a possibility that the site was re-used as will be discussed below. If this was the case then the strategic advantage would be linked to such re-use.

Documentary evidence

Primary reference:

An early record for the castle is in the ASC (Peterborough Mss) which records for the year 1051 that ‘The foreigners had then built a castle in Herefordshire in Earl Swein’s province and inflicted every injury and insult they could upon the king’s men thereabouts’ (Swanton 2000. 174-5). Doubt as to the accuracy of the term castle was discussed above, (see chapter 3), however, the account does suggest that Ewyas Harold was fortified in some way.

Db.H records three records for the castlery of Ewyas and one mention of the castle itself: Of these nine hides, one part is in Alfred’s castlery of Ewyas’ (Thorn and Thorn 1983. 181c), ‘In the castlery of Ewyas Earl William gave four carucates of land to Walter de Lacy’ (Thorn and Thorn 1983. 184a), ‘In the castlery of Ewyas (Harold) Roger holds from Henry three churches’ (Thorn and Thorn 1983. 185b) and ‘Alfred of Marlborough holds the castle of Ewyas from the King. The king himself granted him the lands which Earl William, who had refortified the castle had given him’ (Thorn and Thorn 1983. 186a). William fitz Osbern’s refortification must have taken place between 1067-1071, (see chapter 3).

The earliest mention of the village of Ewyas is: ‘Elcon, situate on Dulais’, comes from the Lib. Land (Rees 1840. 451). How or when Elcon became Ewyas Harold is not known and the origin of the name Harold has always been cause for speculation by antiquaries, see below.

Antiquarian reference:

Some of the main theories identifying which Harold were recorded by Bull in 1869:

‘The son of ‘Kynge Harold’ (Leland).

The son of fitzOsborne Earl of Hereford (Dugdale).

The son of Ralph Earl of Hereford (Gough).

The son of Drogo fitzPonz (Freeman).’

(Bull 1869. 29).

The son comes from frequent use of the term ‘Mab Harold’ in documents relating to him; Mab meaning son (Bull 1869. 29). The most likely contender would, however, be the 5 year old son of Ralf the Timid, Earl of Hereford (Marshall 1938. 144). The earliest antiquarian record of the castle comes from Leland who in 1530 seemed less than impressed with the site writing ‘nothing remains of it now but the mound and fosse’ (cited in Bull 1869. 32). In Leland’s time a church stood in the castle, probably on the bailey, ‘a large part of the castle still stands’ together with a chapel of St Nicholas within it’ (Chandler 1998. 228).

The dating of the castle here is central to one of best known arguments in the field of Norman castle construction. Vying for the claim to be the first castle built in Britain are Richard’s Castle outside Ludlow, Hereford Castle and Ewyas Harold Castle, all products of Norman followers of Edward the Confessor who were granted land after his accession in 1043 (Swanton 2000. 162). An early claim for Ewyas Harold’s construction came from Bannister, his History of Ewyas suggested that it was built on a former Saxon burgh constructed around 915 as a defence against the ‘Black Pagans’, the Danes (1902. 6).

Modern reference: HWCM1449

The modern record of the site can be best presented by two maps, one the RCHME 1931, (Vol. 2. figure 53) and the other by Kay from the 1940s, (Vol. 2. figure 54). The RCHME version shows the layout of the site in relation to the topography of the area, and includes detail of earthworks to the south of the bailey. It also includes a longer section of the east rampart than remains today and without the breach for the access road to a modern bungalow north-east of the motte, on the raised earthwork. The Kay sketch identifies more of the features of the site than the earlier map including the shell keep on the top of the motte and the earthworks to the south and east of the bailey. It should be noted that Kay’s compass bearing is about 260º out.

Additional references:

VCH 1908. 237-238.

Armitage 1912. 150-151.

Renn 1968. 184-185.

Hogg and King 1963. 111.

King 1983. 278.

Interpretation: Motte and bailey (Early)

Ewyas Harold is one of the best known motte and bailey castles in the country, cited as a standard by which to measure others. The site, however, is not as standard as it would first appear. The site occupies the end of a ridge where the Norman practice of cutting a ditch has been used to separate an area of ground for a motte. Along from the motte, the top of the ridge would normally be used as the bailey, thereby, providing added protection to the motte and added defence to the bailey by incorporating the natural defence of the ridge sides. In the case of Ewyas Harold the motte is adjacent to the top of the ridge and only 6m above it, putting the most secure part of the castle in a weak position. A misconception about Ewyas Harold is the enormity of the motte itself. The survey showed that of the 16m of height of the motte some 10m is probably natural.

The bailey was formed by levelling the ground to the east, south and west to make a platform, possibly not as large a task as may be thought since the natural hillside may have already been partially shelved due to glacial action which is common in the region. Spoil from this operation could have been used to create the steep defensive bank around the bailey. A problem with the site comes from the fact that the bailey is overlooked by the ridge, making it very open to attack. It is possible that the cross ramparts may have been designed to shield the bailey from the ridge. As they do not appear to form a curtain wall up the motte sides, they cannot have been used as a defensive earthwork to bar entry. Such a shielding structure would seem to be a remedy to an existing problem rather than a pre-conceived strategic device.

The usual layout for a motte and bailey castle would be to have the motte at the steepest part of the natural landscape with the bailey in front, in the weaker direction. At Ewyas Harold, therefore, the motte would have been to the south of the pre-existing ridge with the bailey to the north, on top of the ridge. The ditch could still be in the same position but a further ditch or rampart, or both would have been cut to secure the bailey. This apparent anomaly is worth noting for it may shed light on the origin of the site. Possibly the Norman rebuilding of the castle in 1052 is the key. The site is a motte and bailey castle now, but what was it originally? Does its unusual layout result from a use that dictated a different form, less dependent on the need for defence? It is possible that the Normans developed a pre-existing Saxon burh.

The interpretation of the site, based on actual remains, survey, location and documentary evidence leaves little doubt as to the function and date of this castle. There is no question as to the castle’s early date as documentary evidence tells us that fitz Osbern repaired the castle in 1067-1071, thereby, establishing its place in early conquest history. The unusual size of the top of the motte, similar to Dorstone, suggests that this castle also functioned as an administrative centre. The lack of any evidence of private defence is a problem that cannot be explained but it is possible that masonry structures atop the motte would have catered for such a function.



The material is copyright by the author, and is reproduced here from the Archaeology Data Service website of the University of York for research purposes under their terms of use


Top - Back

Ref: rs_ewh_0012