The castle of Wark on Tweed, the defensive Work (hence, Wark) of Carham, lies at the west end of the modern village of Wark on the south side of the River Tweed, which at this point forms the border between England and Scotland, where it occupies part of an elongated glacial mound or ‘kaim’, a mini-escarpment rising abruptly from the south and extending east and west for some 300 metres each side of the motte.
The greater part of the castle and modern village are located on the north side of the kaim, where the land dips gently away to the north before rising slightly to the riverbank, ending in a cliff up to 15 metres high. Sitting opposite Wark on the north flood plain of the river Tweed is the farmstead and hamlet of Fireburn Mill, bounded on its north side by the Kelso-Coldstream road which forms a junction here with the A697, an ancient route to Edinburgh. Fyreburne Mylne forde is, with Saynte Gillyforde on its west side, one of two fords mentioned on the immediate west side of Wark castle - one of 10 between Cornhill and Carham - in a Border Commissioners report of 1541.
Wark Castle was, with Norham and Berwick, one of three English castles on the banks of the Tweed and, with Norham, one of the two to be taken by James IV in September 1513. Unlike Norham, however, there is no known documentary information available on James IV’s conquest of Wark Castle, leaving archaeological methods as the only means by which to approach questions about the layout of English defences at the time and the direction of the Scottish attack, which is thought most likely to have been from Fireburn Mill on the north side of the Tweed or, alternatively, from west of the castle on the south side.
The landscape of Wark on Tweed site is dominated by the earthwork castle mound, raised on the centre of the kaim, with the Middle and Outer Wards positioned to the north on gently undulating ground between the kaim and river cliff. Much of the inner bailey is now overgrown and waste ground, while the Outer Ward contains modern buildings and open spaces. Stonework remains of the castle are very fragmentary and, apart from sections of heavily overgrown curtain walling, there are no surviving standing medieval structures. Indeed, the covering of vegetation is so complete on the south and east sides that visitors from the direction of Cornhill see only a green mound that it not immediately identifiable as a castle site. Visitors approaching by car from the west may catch glimpses of stonework within the grassed-over west elevation of the castle mound, but the view is impeded and confused by the presence of the kaim, which, although perhaps modified in places, is substantially a natural feature.
In addition to the area bounded to the north by the river Tweed between the former ferry and ford crossings, and to the south by the present Cornhill-Carham road, the castle site also includes, or is associated with, a gently undulating terrace above the river Tweed which extends from the west side of the motte to a point just east of St Giles’s chapel cemetery, which lies on slightly lower ground. A break in the kaim at this point gives access to St Giles’s chapel site from the south, but a short distance to the west the kaim tapers and disappears.
Sources and History
The principal published accounts, which summarise the known history of the castle derived from the corpus of chronicles and documentary sources, include those provided by Bates in his work Border Holds of Northumberland (1891, 331-59), by Vickers in Volume XI of the Northumberland County History (1922, 44-74) and by Hunter Blair in the Proceedings of the Berwickshire Naturalists Club (1935-7, 76-103). The latter published for the first time two 16thcentury plans of the castle. Volumes II (Colvin et al. 1963, 852-3) and IV (Colvin et al. 1982, 688-94) of The History of the King’s Works provides the most authoritative analysis of the works undertaken in the castle in the periods when it was in royal hands during the Middle Ages and more importantly with regard to the complex sequence of renovation, modernisation and repair undertaken by the Crown during the 16th century.
While prehistoric settlement activity in the vicinity of Wark is attested by lithics and cropmark sites and the topography of the castle site itself suggests that it may well have been settled prior to the establishment of the Norman castle, no evidence for occupation prior to the early 12th century has yet been found. The recorded origin of the castle site is as a motte and bailey earthwork castle, presumably with timber keep, built during the reign of Henry I (1100-35) by Walter L’Espec (d.1153), who had been granted the honour of Carham by Henry I. After being destroyed in 1138, the castle was taken into royal hands by Henry II and rebuilt in 1158-61, perhaps at least partly in stone. Thereafter it passed in and out of Royal hands, although even when in private hands both Henry III and Edward I made temporary use of the castle on several occasions in the 1250s and 1290s/1300s. After briefly falling into Scottish hands in 1318 the castle was described as ‘ruined and broken’ in 1329 when the Crown again relinquished control. In the turbulent 14th and 15th centuries there are repeated references to the defences being slighted and its walls beaten down but stone castles were difficult to destroy completely before the widespread availability of gunpowder and some of the damage may have been less extensive than the chroniclers might imply. The overall impression is that the castle was not subject to any major remodelling after the work of Henry II, although continuous repair would undoubtedly have been necessary prior to the works subsequently undertaken by the Crown during the 16th century.
By the early 16th century the castle had grown to consist of three wards, visible on 16th century views of the castle by Rowland Johnson, Antonio da Bergamo and an anonymous bird’s eye view reproduced by Hunter Blair (Wardle and Nolan 1997). These comprise an inner ward at the southwest corner of the castle complex, defined by the extent of the Ring, with middle and outer wards occupying the area between keep and river cliff as far as the west side of the western of two access routes into the village from the main southern through road. The boundary between the middle and outer wards is defined by the current access roadway to the gamekeeper’s cottage, formerly the Salmon Inn above the old ferry landing (formerly Gillies ford). It appears likely, however, that the rest of the area defined by the current village, which occupies the area between kaim and river cliff from the castle motte as far east as a traditional fording point next to Wark Farm, served in the medieval period as an extra outer ward, perhaps occupied by a permanent civilian population (thus equivalent, perhaps, to a Roman vicus) but providing an enclosed area for use as refuge by the wider population and its stock during times of military threat, or for billeting large numbers of men mustering in advance of campaigns to the north. The area to the west of the motte has been suggested as serving a similar purpose and is the subject of fieldwork investigations described below.
Successive spells of royal tenure mean that works on the castle are recorded in a much more copious body of documentation than would be the case for an equivalent baronial castle. This material includes some highly detailed reports outlining the state of the castle or listing work recently accomplished, notably a paymaster’s account book for works at Wark between 12 February and 10 November 1543, of which only a one page abstract has hitherto been published (Bates 1891, 349-50; cf. Colvin et al. 1982, 691).The period of current interest is covered principally by the Letters and Papers, Foreign and Domestic, of the Reign of Henry VIII, preserved in the Public Record Office (London, 1862-1932). The castle also figures in a succession of overall surveys of the state of the Border defences undertaken by crown officials in 1415, 1541, 1550 and 1584 reproduced by Bates and Hodgson (1415: Bates 1891, 14; 1541: Bates 1891, 30, 347-9, and Hodgson 1828, 179-82 fn; 1550: Hodgson 1828, 201-3, Bates 1891, 350-52; 1584: Bates 1891, 72), and is also unusually well-represented by 16th century plans and illustrations. The principal early 16th century descriptive surveys with relevance to the Flodden campaign and its aftermath are Lord Dacre’s report of 1519 (Bates 1891 342), Hart’s Inspection of 1522 (L&P Hen VIII Addenda 837), a 1523 Survey (L&P Hen VIII iii, 8286), Sir Robert Bowes and Sir Ralph Ellerker’s Border survey of 1541 (in Hodgson 1828, 179-82) and Sir Robert Bowes Book of the State of the Frontiers and marches betwixt England and Scotland, 1550 (Hodgson 1828, 201-3).
Repairs and alterations carried out to the castle in the aftermath of the Flodden campaign began in 1517-19 with works supervised by Lord Dacre when the keep was rebuilt to house artillery. In the period 1522-23 further works were carried out, modifying the original design to give the keep a flat roof so that artillery could be mounted upon it, and in September 1523, in preparation against Scottish attack, the Earl of Surrey ordered new bulwarks and other defences with earth and timber to be erected, again presumably to support artillery pieces. Dacre supervised further repairs in 152426 following an unsuccessful two-day siege by Albany in November 1523. The period 1542-50 saw the construction of the artillery fortification, first referred to in a survey of 1561 as ‘the Ring’, around the motte [Hunter Blair, p76], transforming it into a raised and revetted, level platform for artillery, approx. 7.3 m high and 1.83 m thick, with twelve embrasures for cannon which stood on a platform inside the wall. At the centre of The Ring stood the earlier keep. Work was also begun at this time on the building of an inner wall, parallel with the south and east walls of the middle court, extending as far as the cross wall separating the outer and middle courts/wards, with the intervening space being infilled to provide a level, 20ft wide artillery platform along these sides of the middle court, similar to the Ring itself. The raised artillery batteries along the inner face of the north, riverside wall may also have originated in this period. In 1592-97 further repair works were carried out, notably to the north, ‘water wall’ above the Tweed, but by 1633 what remained of the artillery had been removed, and in 1639 the castle was described as ‘ruinated’.
Relatively little archaeological work has been carried out at Wark, though limited excavations were undertaken on top of the motte in 1862, revealing the remains of a flight of steps on its east side (Mearns 1863-8, 61-6). More recently a survey of the castle earthworks was undertaken by the RCHME in 1992 (Welfare et al. 1999, 53-5, fig. 19) and Newcastle City Archaeology Unit undertook archaeological recording whilst monitoring conservation works on the 16th century artillery fort and the north curtain wall of the middle bailey, as well as the construction of a retaining wall to support the north face of the castle mound (Wardle & Nolan 1997). Although limited in scope, this work led to a better understanding of structural phasing, notably that differences in the corework of The Ring and a very small section visible on the south curtain suggest that the latter is not part of the 1543 refortification and may be a remnant of an earlier phase of construction.
Fieldwork in 2013 and 2015
An appraisal of the potential of the site carried out for the Flodden 500 project in 2012 (AP Ltd. 2012), followed by test pitting and geophysical survey, led to the identification of sites for excavation which aimed to gain insights into the methods used to besiege and defend the castle in 1513. Exploratory test pitting was carried out in the Summer of 2013 within the village but largely outside the recognised boundaries of the castle, followed by geophysical survey on the large, slightly undulating area of land to the west of the castle which, it was suggested by Flodden 500 archaeological project manager Chris Burgess, may have served as an additional ward of the castle when troops - either English or Scottish, depending on who held it at the time - massed ahead of campaigns.
This area, which is entirely covered by the remains of broad rig & furrow cultivation features of presumed medieval or early post-medieval date, is bounded to the south by the kaim, to the north by the river and to the west by an apparent linear earthwork running north-south, beyond which are the remains of a medieval chapel to St Giles’s.
The latter, now surviving as a triangular-shaped cemetery with the foundations of a building towards it north side, is said to be the chapel associated with the castle occasionally referenced in medieval accounts, but its position a considerable distance west of the castle leaves this association open to doubt. Test-pitting in the village uncovered few finds of medieval or 16th century origin, although a 2 x 1 m trench on the south side of the playground revealed fragments of medieval and early post-medieval pottery, probably resulting from the discard of domestic waste, below a considerable build-up of topsoil. Geophysical survey in the field west of the castle identified a number of anomalies, notably an apparent linear feature associated with the apparent earthwork at the west end of the target area, which were selected for further investigation. Accordingly, trenches were sited to investigate a number of possible features suggested by the geophysical survey, while a number of additional trenches were placed in positions thought promising for reasons of topography or association with built features, including the chapel of St Giles’s. While no significant remains were found in two trenches close to the kaim at the east end of the area under investigation, where shallow deposits of topsoil lay above the natural gravels and cobbles of the glacial kaim, trenching in the northern half of the area revealed a considerable depth of topsoil overlying medieval remains. Trench X, positioned close to the river cliff over a linear geophysical anomaly uncovered a hearth associated with medieval pottery and a number of slots and shallow gulleys which, although they failed to constitute a meaningful pattern, may represent the remains of temporary shelters. In general the nature of these remains supported the notion that they could have resulted from episodes of camping by troops, although only a single iron ‘bodkin’ arrow-head was found alongside quantities of animal bone and pottery, interpreted as domestic waste from food preparation.
The most interesting and enigmatic remains, however, were found at the west end of the area, on the fringes of and outside the area subject to geophysical survey in 2012. Two main sets of features were revealed in this area. First, the putative earthwork feature forming the west end of the survey area was found to be a substantial stone-reveted embankment. In the area investigated it was formed of two drystone walls spaced some 3.1 metres apart (between inner faces), that on the east side being up to eight courses and 1.3 m high and battered, or sloping slightly inwards, from bottom to top, while the east wall was less-clearly defined. Between the two walls was a core of dense, silty material containing abundant stones of natural origin, presumably upcast from groundworks associated with the remodelling of ground surfaces associated with the revetment walls. Although the ground level steps or slopes down from the western edge of the earthwork towards St Giles’s chapel, there is no clear indication of a ditch on that side, not on the east side where the ground surface above the wall continues eastwards at about the same level (into what was until recently as the Football Field, giving some indication of its topography). Although undateable, the east face of the east wall contained reused medieval dressed stone and is on that basis thought likely to be of late medieval origin. Intriguingly, a survey of 1541, produced a generation after the battle of Flodden recommended, in the light of a continuing threat from the north, the extension of the castle westwards as far as it already existed to the east which, if the eastern extent is taken to include the village as far east as the former ford crossing, equates roughly to the site of the remains in question, suggesting that the survey recommendations were enacted. The south end of the earthwork embankment, which terminates on top of the kaim on the east side of Gilly’s Nick, was also investigated by means of a single, small trench but, other than the earthen bank itself, no structural or artifactual remains were found. It could not be ascertained whether the bank was contemporaneous with the original creation of Gilley’s Nick, which is a gap in the kaim providing access to St Giles’s Chapel from the south, or is secondary, as seems more likely, as St Giles’s chapel is likely to be medieval in origin while the earthwork, at least in its central section, appears to be later.
In 2015, in addition to re-excavating the site of the earthwork first examined in 2013, a single exploratory trench was also opened in the area west of this, north of the remains of St Giles’s chapel in the extreme north-west part of the site. Here, two pits filled with dark, humic loam and containing abundant fragments of medieval pottery and associated animal bone, as well as a lead (loom) weight and bone pin, were revealed cut into the natural gravels. Subsequently, in September 2015 this area was revisited and a more extensive excavation undertaken, revealing several more pits of various sizes and depths, all filled with remains indicative of domestic waste disposal in the later medieval period. One of the pits investigated was up to three metres long and over a metre deep; this and another deep pit showed some signs of structural elements in the form of post holes and post-impressions against their internal sides, perhaps suggesting timber and post, or timber and wattle supporting structures to prevent collapse. It is thought that the pits may have originated as latrine and rubbish pits
for the large numbers of men billeted on this putative ‘temporary outer ward’ of the castle, although this is far from certain. What is certain, on the basis of pottery evidence and a coin of Edward I, is that the pits pre-date the battle of Flodden by at least two centuries, but are probably contemporaneous with the use of the adjacent St Giles’s chapel and cemetery. The interpretation of the pits as latrines may appear dubious given the proximity of a place of worship, but it is of course unknown whether the chapel was in use during the military gatherings inferred and, regardless, it is arguably inappropriate to impose modern sensibilities on a medieval military context. These arguments notwithstanding, other interpretations of these features remain distinctly possible, albeit elusive.
The survey and excavation work carried out within and west of Wark Castle between 2012-15 has for the first time proved that significant archaeological remains, suggestive of intensive military activity over a considerable period of time, exist in the area west of the castle. The suggestion made, therefore, by Flodden 500 archaeological director Chris Burgess at the outset of investigations, that this area should be considered as an additional, or auxiliary outer ward of the castle, much like the area similarly regarded on the south side of Norham castle, has been verified. However, although this conclusion applies to the medieval castle it has not been shown to apply to the early 16th century castle, since no diagnostic remains of that period were found during fieldwork, although it is suggested that the substantial boundary feature uncovered at the west end of the site was built as a result of recommendations for strengthening the castle following the Flodden campaign.
The work completed as part of the Flodden 500 project provides the foundation for future work in the same area which should aim to determine questions posed in this discussion, including the extent of pits and putative related structures in the area north of St Giles’s Chapel, the relationship between these and the chapel itself and the date of construction of the linear, reveted earthwork feature forming a boundary between St Giles’s chapel complex and raised platform of land to the east. Other questions also persist, such as the dedication of the current chapel remains to St Giles’s when Gilly’s Ford, noted above, is the easterly of the two fords mentioned in 1541 west of the castle, and must lie east of Fireburn Mill which, in turn, is east of the current St Giles’s chapel. Could it be that the dedication to St Giles’s refers to a chapel formerly situated within the castle complex itself, while the chapel currently known as St Giles’s, lying outside the western limits of the castle, is another establishment altogether, perhaps a hospital or monastic site?
Bates, C J, 1891, ‘Wark Castle’ (in) The Border Holds of Northumberland,. Archaeologia Aeliana 2, XIV, 331-369.
Bowden, M, RCHM(E) Survey of Wark-on-Tweed Castle, March 1992. RCHM(E).
Colvin H M, Allen Brown R and Taylor A J, 1963, History of the King’s Works; Vol. 1: The Middle Ages. HMSO; London.
Colvin H M, Summersen J, Biddle M, Hale J R and Merriman M, 1982, History of the King’s Works; Vol. 4: 14851660 (Part II). HMSO; London.
Hodgson J, 1828, History of Northumberland
Hunter Blair, C H, 1935, The Castle of Wark-upon-Tweed, Proceedings of the Berwickshire Naturalists’ Club 29, 76-103.
Kenyon, J. R. 1977, Wark Castle and its artillery defences in the reign of Henry VIII, Post-Medieval Archaeology 11, 50-60
Mearns, P, 1868, Wark Castle, History of the Berwickshire Naturalists Club, Vol. V, 1863-68.
RCHM(E). Survey of Wark-on-Tweed Castle, March 1992
TAP (The Archaeological Practice Ltd.), 2012, Wark on Tweed Castle, Northumberland: Options appraisal for community involvement and access improvements. Unpublished report prepared for Flodden 1513 Ltd.
Tate G, 1868, Wark Castle; in Berwickshire Naruralist’s Club 5, 427-37.
Vickers, K J, 1922, Wark Castle, Northumberland County History Vol.XI, 44-74
WARDLE S & NOLAN J, 1997, Wark Castle Archaeological Recordings. Unpublished Report by Newcastle City Archaeology Unit for Northumberland County Council.
Welfare, H., Bowden, M., and Blood, K. 1999 Fieldwork and the castles of the Anglo-Scottish Borders, in Pattison, P, Field, D and Ainsworth, S (eds) 1999, Patterns of the Past: Essays in Landscape Archaeology for Christopher Taylor, Oxford and Oakville, 53-60.