Refortification at Rochester in the 1220s: a Public/Private Partnership?

REFORTIFICATION AT ROCHESTER IN THE 1220s: A PUBLIC/PRIVATE PARTNERSHIP? DEREK RENN Reassessing their excavations in and around the Deanery Garden, Flight and Harrison pointed out significant changes in Rochester's topography caused by shifting relationships. The ancient urban integrity expressed by the Roman town wall had been breached severally by the Norman castle and monastery, both latterly controlled by the archbishop, the city being controlled by a royal official. After the siege of 1215, the castle was controlled by the king and the citizens acquired a measure of autonomy by royal charter in 1227. The defences now encompassed both castle and town and 'expressed a new configuration of political relationships ... a partnership of sorts between king ... and the citizens of Rochester' .1 Crown interest in town defences Although burh-bot, the ancient obligation to work on public fortification, was sometimes diverted after the Norman Conquest to provide labour for royal castle-building (as at Ely in 1071 and London in 1097), a distinction was still made in 1215, when all the men of Berkshire were to be summoned 'to repair the ditch of our castle and town of Wallingford as they were accustomed to do in time past'. In the same year, separate orders to repair Hereford castle and to make a payment towards fortifying the town were enrolled consecutively.2 During the twelfth century, extraneous payments were made from the Exchequer towards the cost of fortifying just six towns, each with a royal castle. Five of the towns had Roman walls, and the sixth (Hereford) Saxon defences, already.3 The payments for Chester were by far the largest; the rest are a medley of odd small amounts. This suggests that the Exchequer met a specific bill, symbolically recognising communal work which was in the Crown's military or economic interest. There is only one concurrent group of payments, namely in 1189-91 for gates at London, Carlisle (three gates) and 343 DEREK RENN Hereford (four, plus a castle gate). Dr Hilary Turner has identified the Hereford town charter of 1189 as a precursor of change, the citizens being given the town at fee farm provided that they attended to its fortifications. 4 The earlier of the two payments to Rochester also mentions gates, the control points which are the most obvious element of any enclosure, but the later payment in 1192-3 differs from all the others listed. The local items are grouped under a bold heading ROUECESTRE, the farm is accounted for by a vintner and a clerk (and not by the sheriff) and continues: Et pro civitate Roffecr' esforcianda muro et fossata ix Ii.et x s.et iijd. per breve R.et per visum et testamonium Reginaldi de Cornhill' et Godardi contratalliatoris. It is unusual for such a small amount to be certified thus, and the unique reference to wall and ditch also hints at some larger operation. Rochester was the only non-county town among the six, and its castle was usually controlled by the archbishop. 1192-3 was a very unquiet time in England: King Richard was a prisoner of the Emperor, and Count John, Philip II of France and Llywelyn ab Iorwerth were circling. With the deposed chancellor Longchamp, still the papal legate, back in Dover and an invasion from Flanders feared, there was very good reason for the men of Rochester to look to their defences while the castle was being strengthened. 5 In 1202-3 the men of Southampton were given £100 a year for two years to enclose their town. 6 In 1204, King John enjoined his justiciar in Ireland 'to create a castle, if need be for the defence of the city [of Dublin] but first to build a tower' and separately commanded the citizens to fortify their city. Two years later an aid towards the town defences was granted but progress seems to have been slow both on the castle and on the town wall. 7 In 1212 the men of Shrewsbury were granted timber ad villam claudendam. 8 As well as trees, cash gifts, loans, remission of debts, customs, merchant fines or fractions of rent receipts, were made by the Crown to towns for their defence from 1215 onward. An innovative - and cost-free - grant was royal permission for town authorities to levy a toll on goods brought in for sale, the proceeds to be spent on walling (murage). The first enrolled murage grant was that to Shrewsbury (26 June 1220) immediately followed by an identical grant to Bridgnorth nearby.9 After a gap of four years, other towns received similar permits, perhaps significantly after the potential political bomb of the custody of royal castles had been skilfully defused. 10 344 REFORTIFICATION AT ROCHESTER IN THE 1220s Developments at Rochester The great ditch of 1225 Rochester never received a murage grant, and Canterbury and Dover had to wait until the fourteenth century for theirs. However, in February 1225 the king ordered the payment of wages to those labouring in the city ditch of Rochester, and Exchequer payments of some £300 for enclosing and strengthening the town were enrolled that year.11 A fourteenth-century chronicle states that the great ditch about the city was begun in 1225, 12 but there seem to be no later Exchequer enrolments, so the work may have been completed either very swiftly or eventually at the citizens' expense. The only hint of walling is the payment for lime-kilns for the work of the castle 'et ville nostre Roffe'. Where did this (presumably) new ditch run? Excavations outside the Roman wall, particularly south-east of the High Street, have revealed undated ditches.13 The west wall of the north gate was rebuilt late in the thirteenth century and the P-plan towers of the east gate perhaps later still. The butt-end of an adjoining ditch had a tripod pitcher in the filling, but the exact stratification is not clear: further north, the ditch seems to have adjoined the Roman wall. 14 The pavement line of The Common and of Free School Lane, together with plot boundaries further west, shown on Canon Livett's map of 1894-5, 15 are about 30m away from the line of the Roman town wall and may mark the counterscarp of the 1225 ditch; the ditch between castle and cathedral, cutting through that wall, was of similar width. 16 Boley Hill This part of the high ground adjoining the river has been occupied since at least Roman times: coins, urns and lachrymatories were found there in the eighteenth century, and the 'fag-end of a Saxon cemetery' in the nineteenth. 17 The long mound, once conical and flattopped, in the garden of Satis House contained an early thirteenthcentury occupation layer but has been much altered.18 Canon Livett recorded an early wall in the same garden, and Mr Tatton-Brown has drawn the writer's attention to the longer wall of Kentish rag on the west side.19 A long bank to the south of the house stands above a deep cutting of comparable width to the castle and town ditches (Fig. 1, a, b, c). The usual assumption is that the east boundary of Boley Hill followed St Margaret's Street, meeting the castle ditch at the corner where the Moot Tree once stood. But it is striking how the ground 345 'JOO, DEREK RENN . C 􀀁 Fig. l Site plan of Rochester castle (above) and Boley Hill (below). Ditches marked by double-headed arrows. 346 REFORTIFICATION AT ROCHESTER IN THE 1220s falls away towards the cathedral: might the ditch once have looped further east, either across College Green, before the erection of the bishop's palace in an extra-mural precinct, or even across the site of the later cathedral but enclosing the earlier church?20 If Boley Hill formed part of the eleventh-century castle enclosure,21 the twelfthcentury donjon would have been centrally placed, symbolically straddling the Roman town wall. However, Flight and Harrison argue that the early curtain wall near the donjon stood on a bank thrown up against the outside of the Roman town wall, so putting Boley Hill outside the early Norman defences. 22 Boley Hill could (should?) have been used during the sieges of 884, 1088, 1215 and 1264. 23 That Le Boleye remained in a separate jurisdiction for over 400 years suggests that it either became, or was created as, a faubourg or barbican added to the castle, perhaps in the 1220s when a gate was added towards the south. 24 If the Great Ditch included Boley Hill, this might explain some of the payments to the citizens. How the defences linked here is a puzzle, since the needs of the citizens, the garrison, the monks and the bishop would have differed. Livett's map shows a curious narrow strip of land between the palace and the 'site of the second south gate'; might this have been a 'bishop's postern' in the flank of a castle/town entrance (Fig. 1, d)?25 The siege of 1215 Briefly, the local situation after the sealing of Magna Carta was that the rebel barons who had seized London agreed to a compromise whereby the Tower of London was to be held by the archbishop of Canterbury. But, by the end of September 1215, Stephen Langton had left the country and King John had gone down to the coast, awaiting mercenary forces at Dover. 26 The rebel barons siezed Rochester to block any advance on London. Reginald of Cornhill II, the archbishop's constable, handed over the castle and joined the rebels. Royal forces broke Rochester bridge to prevent any relief, surprised the town on 11 October and besieged the castle. A ceaseless barrage of arrows and bolts from bows and of stones from 'engines' began.27 One chronicler states that the 'engines' breached the bailey wall, but another says that they did little damage. Both agree, however, that suffossatores had to be sent in before one side of the donjon was breached, and the defenders held out for a time on the far side of the spine wall. 2 8 Suffossator translates literally as 'under-digger', but were the walls tunnelled under or through? The evidence is equivocal. The picks ordered on 14 October, just after the siege began, might have been for 347 DEREK RENN ditching the siege-camp and digging assault trenches: Wendover speaks of vallavit.29 The trench found running towards the NE tower and interpreted as a siege-work was cut in natural gravel, probably the berm of the castle ditch. 30 Tunnelling in gravel requires close shoring and roofing. Further south, the castle ditch was cut into solid chalk, found about 6m down near the round tower in making a cesspit.31 Tunnelling is always hard work, and the shortest would be the best, perhaps approached from a zig-zag of surface trenches, of which this trench might be one. A ground-penetrating radar survey outside the south-west face of the donjon in 2001 found anomalies typical of a rubble-filled trench. 32 The full extent of the donjon 's foundations are unknown but are visibly deep, perhaps going down through made ground (a former motte?) but with the south-west wall standing partly on the Roman town wall.33 An enormous mine would have been needed to break such rigidity, whereas a surface sap would more easily make a hole large enough to cause a partial collapse.34 The need for bacon fat to fire props would be as necessary above ground as below, and the order for pigs was only sent four days before the final surrender.35 At Benington, Bridgnorth and Bungay, there are Norman donjons with one wall cleanly removed but only down to plinth level. At Bungay, an unfinished X-plan tunnel passes through the strongest corner and at Degannwy, Henry III's curtain wall was broken open at ground level in 1263 and collapsed outward. 36 What exactly was demolished in 1215 at Rochester? The curtain wall south and east of the donjon has been completely rebuilt, and the rounded southern angle of the donjon differs from the others. This angle was that most exposed to attack: the original square angle may have been riddled with rooms (facing the noonday sun), like that at the opposite comer, and hence easily sapped. Apart from the forebuilding, the donjon is remarkably symmetrical even now, with openings one above another on all sides, still representing lines of structural weakness. The generally accepted view is that one-half of the south-west wall fell, together with one-third of the south-east side.37 So either there was a deliberate major rebuilding on the century-old plan, or the initial collapse was far less extensive than has been previously thought. The argument against the latter rests on two largely-blocked openings at top-floor level, and a gap in the line of joist-holes beneath. However, at least half the greater two-order arch can be traced in the wall, and both its blocking, and that of the other opening, may have been deliberate covering-up of symbols of the archbishop's former control. The corner internal pilasters are carried up to the roof, and the other floor offsets run to the angle. 348 REFORTIFICATION AT ROCHESTER IN THE 1220s Externally, the narrow sloping plinth of irregular Kent Rag rubble runs consistently to the west side of the round turret, where there are Caen ashlars (?quoins) in the wall above, and the plinth resumes (as two courses of Caen ashlar) only a few feet beyond the east side of the turret, exactly where there are traces of an eccentric offset and return of the size and position of the flat pilaster buttresses elsewhere (Fig. 2, a). However, the offset might be due to an underlying Roman wall. The wall panel above this point is scarred by a deep horizontal slit about 2m up, and a large hollow in the wall face extending from the lowest broken-out window embrasure. The latter is said to have been a test cutting by a local pavior in 1738 to whom the stones had been offered. 38 After the siege The alterations to the donjon and the rebuilding of part of the curtain wall are probably the consequence of the 1215 siege. 'Probably', since there are no closely dateable details and (i) substantial sums were spent on the donjon in the twelfth century;39 (ii) another siege in 1264 pushed the defenders back into the donjon and mined it;40 (iii) major works were needed in 1256 and 1367-83 ; 41 (iv) after the removal of all the timber and nearly all the ashlar stone, the repairs of 1826 and 1896-190 l were also extensive.42 The 'stitching' apparent on the exterior south-west wall panel could be of any date from the twelfth to the twentieth century. Louis' forces recaptured Rochester early in 12 I 6, but it was recovered a year later and by March 1219 it was able to host the negotiations for an extension of the Treaty of Kingston/Lambeth.43 Small amounts for repairs are recorded during Henry III's minority: breaches in the walls were to be mended in 1223, but arrears of siege expenses were still being paid in 1224, and the instruction to transfer 96 picks from the Tower of London to the constable of Rochester castle in 1225 may have been for works at the castle rather than for the town.44 An account of June 1225 includes a shipload of wheat landed at Rochester and carried in turrim. 45 Taken literally, the donjon would have been at least partly weathertight by then. Up to £100 was to be spent on the turris in 1226 and the next year the sheriff was ordered to complete work upon it. However, it was not until 1232 that the floors were laid and the roof leaded. £530 had been spent in repairing the tower and £150 on the rest of the castle but the austerity of the work is nevertheless apparent.46 349 DEREK RENN 10 /0)(. / / Fig. 2 Rochester castle, south corner at ground level (a); and tower only at first floor level (b). The ten years' delay was due to competing claims: Hubert de Burgh's personal experience was reason enough to give Dover priority and, to a lesser extent, London. Only after his recovery of Rochester from a neutral holding (by the bishop) in 1223-4 was reconstruction possible.47 Rochester's national importance lay in its bridge, the only one across the lower Medway which was the natural obstacle to 350 REFORTIFICATION AT ROCHESTER IN THE 1220s invasion through Kent. It was also the seat of Kent's other cathedral, and so a natural focus for secular control in Kent, bypassing Canterbury. Perhaps also a belated war memorial and civil reparations, the royal works began on a high note but then fell away, later to resume as part of Henry III's palace programme. The cylindrical south angle of the donjon Allen Brown suggested that the round angle at Rochester was due to the donjon being rebuilt in Poitevin fashion.48 Marie-Pierre Baudry attributes it specifically to Savaric de Mauleon, who had been present at the siege, before he left England late in 1216 after having been badly wounded in a skirmish near London.49 Although so rapid a redesign and rebuilding seems impossible, Savaric was back in England by 1226-7 or another Poitevin might have provided the inspiration. Mauleon castle itself has lost its donjon but still has a gate flanked by solid round turrets and lies between the round-buttressed donjons of Chateaumur, Pouzages and Tiffauges. But it is Niort, with its double-donjon of the last quarter of the twelfth century, and Savaric's base for much of his colourful career in Poitou (1205-24 ), which might be the source: both John and Henry III gave money to fortify the town, and others in Poitou and Gascony.50 The north-west angle of the otherwise square donjon at Newcastleupon- Tyne was rebuilt with faces at 30 ° to each other. Rochester and Newcastle, at opposite ends of England, were the royal castles at each end of King John's great march through rebel territory in the winter of 1215-16 and both these alterations might have been made to mark this triumph, facing possible invaders (compare the Percy lion at Warkworth). But, at Newcastle, it is perhaps more likely to be part of the major repairs of 1271. 51 Another suggestion is that Rochester was altered to resemble (from one direction) the White Tower at London. The south corner of the donjon at Rochester is now pierced by only three openings below the wall-walk, each of which has lost its external dressings. The unusual diamond plan and access through a bent mural passage opening off one side of a window embrasure would have both provided a good arc of view (and fire) of about 40 ° and have avoided silhouetting an observer or archer against the light. Between them, the pair of openings at first-floor level and the one above (Fig. 3, a, b) cover an arc from south-east round to west, that is, overlooking all the land south of the cathedral. 52 The diamond plan arrowloop occurs several times at Lincoln and possibly once at Niort. At about two-thirds of its height, the slightly-tapering cylinder sharply narrows in diameter by a flat offset which does not continue along the adjoining wall faces. This might be a design change, for 351 DEREK RENN C J(W. (OJ(. b / a Fig. 3 Rochester donjon south turret at first floor (a); second floor (b); and wall walk ( c) level. structural economy or efficiency, or perhaps to take supports for a strutted framework whose main beamholes can be traced round the wallwalk and radiating inside the Dplan corner turret at that level. There are two oblong openings in the turret above wallwalk level (Fig. 3, c). The apparent width of all five openings is unusual for 352 REFORTIFICATION AT ROCHESTER IN THE 1220s arrowslits but there seems to be insufficient space behind them to work an 'engine'. The massive cylinder at Rochester created a powerful image, enhanced by the scarcity of openings, so functionally combining maximum passive strength with a modicum of active defensive capability, linked to the remarkable archery battery in the round tower below. The curtain wall and round tower The city's counter-sea! is a remarkable piece of engraving, showing not only the triangular frame of the hourd on the turrets of the donjon but the east face of the castle apparently before the fourteenthcentury reconstruction.53 The battlemented wall is fronted by waves: on the left is a tower higher than the central pair (which each have a squinch arch to either side). On the right is an opening flanked by square towers approached by a wooden bridge on stone arches, like those found in 1888.54 Behind these square towers are a pair of round ones, linked by an arch with a portcullis. This could be another gatehouse near the river bridge, perhaps buried under Richard II's great 'new tower' (today's 'NW Bastion') of 1378-83.55 A slight salient might be marked between the high tower and the next one which would confirm Flight and Harrison's analysis of the foundations of the latter, and be the reason for the survival of the berm here but not further north. 56 The round tower at the south angle of the curtain wall has two tiers of arrow loops, all made in rubble. The lower tier consists of five simple embrasures each with a double-coursed, slightly depressed, reararch, offering a 30° traverse for archers. To each side of the five is a taller and wider semi-circular rear-arch. That facing south has a central pier, separating two near-parallel slits. 57 The other is entirely blocked: internally only one-half is visible, but there are external traces of two slits. The jamb of the northmost slit is just visible at the junction with the rebuilt east curtain wall, which is featureless, apart from traces of its parapet, which also survive on the south-west curtain wall (Fig. 2, b). One pair of double-slits look towards the approach from the town centre to both castle and cathedral, and the other pair, towards the middle of Boley Hill. However, the exterior of the round tower has been largely refaced and, if the piers are insertions, then the openings are large enough to have once been flanking sallyports, like those in towers added to Dover, Lincoln, London and Winchester castles.58 The re􀂈entrant on the west side was perpendicular to the rebuilt curtain wall, with traces of a barrel-vault, probably for a stair to the upper floor of the tower whose joist-holes remain. Only five arrow- 353 DEREK RENN loops remain at a higher level of the round tower, with round-headed rear-arches shorter but wider than those below. Although the embrasures are above each other, their axes differ (Fig. 2, b). A late seventeenth-century sketch by Francis Place in the Hasted Collection shows the round tower with three tiers of arrow loops below a crenellated parapet.59 The curtain wall is drawn as extending north-west as far as a square tower, the section of wall next to which has higher and wider battlements; arrow loops are only shown in that part of the wall which survives today. Part-way along the wall was a small flattopped doorway, not shown on the bird's-eye map surveyed in I 689, but which does show eight embrasures in what then remained of the wall. 60 The present fragment of curtain wall ends in a vertical face with a double-chase as if for a doorway, perhaps that towards the south mentioned in 1237 or for the turning bridge of 1221-2.61 Four short but high-set embrasures remain, the southmost being offset to defend the right-angled re-entrant (Fig. 2, a), also covered by one in the upper level of the round tower (Fig. 2, b). So many arrow loops in one tower (nine at ground level and perhaps as many on each of the three upper levels) is remarkable, particularly as they seem to cover only a semicircle, the north-east/south-west arc, apart from the one just mentioned. Most castle wall towers had at most four arrow loops at any one level, although the Curfew Tower at Windsor has up to seven, Cobb Hall at Lincoln, up to six, and the Wakefield Tower at London now five, but possibly designed for more. These also all date to the 1220s, but do not connect donjon and curtain wall like Rochester. However, there is such a visual link, despite a separation of about 100m, at Kenilworth. Round the top of the donjon there runs a gallery from which two or more altered arrowslits open from each wall face. One angle tower of the outer curtain wall is polygonal, with pilaster buttresses which taper downwards like the Chateau Gaillard donjon to exaggerate the spreading plinth. Arrowslits similar to those of the donjon open alternately through the buttresses and wall panels of this round tower; ten can be traced (over four levels) but alterations make it difficult to decide whether some were genuine openings. This tower is the only one not fronted by the great artificial lakes, and it faces the unfortified borough, with the great donjon behind. Large amounts were spent on Kenilworth between 1211-19 and 1223-25.62 Both Kenilworth and Rochester may have been redesigned for effect and 'over-engineered' with so many arrow loops. The 1264 siege of Rochester was to show that its defensive deficiencies had not been overcome, but Kenilworth's siege of 1266 was a very different story: its great lakes performed better than did Rochester's great ditch. 354 REFORTIFICATION AT ROCHESTER IN THE 1220S The defences of Rochester castle in the 1220s Apart from the surviving donjon and the round tower alongside, we have only the partial view on the city's counter-seal as a guide to what mjght have been Rochester's other defences after 1215 but before the major works in the fourteenth century. The approaches were controlled by man-made bottlenecks at the river bridge, Boley Hill and the cathedral precinct. A peaceable visitor was thus guided along a route which showed off the castle to advantage. The steep river cliff ruled out any approach - certainly any mass attack - from the west, and an advance from the north required capture of the river bridge. An attack from the east would involve street-fighting through the city, past the cathedral precinct. The castle walls may not have reached that bridgehead by 1220, but certainly later on anyone crossing the bridge would be channelled between the second 'great tower' and the chapel, either into the town or along the counterscarp of the ditch round to the eastern gateway, with the old 'great tower' looming beyond a further crosswall. From any of these directions, an attacker would be within easy arrowrange from higher ground within the castle before he could mass his forces, with a need to protect his rear against a concealed counterattack. The most open (and higher) approach is from the south, whether by friend or foe, from which direction the castle today is at its most impressive. Unless Boley Hill was first captured, attackers would be funnelled into a narrow zone, facing fire from many arrowslits in round tower, curtain wall(s) and donjon, as well as from the parapetted wall-walks of all three. A stair probably linked the four levels of the round tower, which might also have had direct access into the donjon. Other contemporary 'partnerships' (and non-partnerships) Concurrent additions of large multi-loopholed single towers to castles and grants for adjoining town defences can be identified elsewhere, especially at royal castles which had been hard-pressed during the civil war of 1215-17. At Lincoln, a vaulted tower of horse-shoe plan was added to the castle, and the town given new gates and towers.63 New towers and a new town ditch defended London, while at the Tower of London, a great tower was built next to the royal lodgings. 64 A new tower was added to Winchester castle in 1222, through whose base a tunnel runs into both town and castle ditches; the city wall towers were repaired and reordered.65 A similar tower was built in the ditch of Dover castle, and its inner earthwork, walls 355 DEREK RENN and gatehouses expensively reorganized. 66 Windsor castle's defences were remodelled, beginning with an odd pair of towers below the motte and a great towered west fa􀁾ade.67 But here, and at Dover, the citizens had to shift for themselves: the whole royal budget had gone on castle-works. Other towns whose defences were aided by the Crown often had a tower rebuilt or added to the royal castle simultaneously (Bridgnorth, Exeter, Hereford, Northampton, Shrewsbury). Bridgnorth and Exeter also had new barbicans, as did Bristol, Gloucester, Lincoln and Oxford. One thousand marks were credited to expenditure on Scarborough castle in 1216-20, and an aid to fortify the town was granted in 1225. Clifford's Tower at York (1245-62) might be regarded as a very belated contribution to such a 'partnership'. Hubert de Burgh, as justiciar, organised the foundation of New Montgomery, the castle with its outsize D-plan tower begun in 1223 and the settlement below was to have a wall and ditch in 1227, when Hubert secured a grant of both for life.68 However, although he rebuilt his long-held Three Castles of Gwent and began a new castle at Hadleigh under a licence of 1230, he did not fortify any settlements outside them. Neither did the earl of Chester (at Beeston, Bolingbroke and Chartley) nor did the individual recipients of new licences to crenellate (Beverston, Redcastle, Tattershall and Whittington). The castles of the Welsh princes are sui generis.69 Henry IIl's regents, however, seem to have linked town and castle defence works on their own estates long before John's death. William Marshal I probably built much of Pembroke castle, including the great round donjon and part of the town walls. 70 The bishop of Winchester, Peter des Roches, had added three new towers to Taunton castle and ditches round both castle and town. Siege damage at Farnham castle can still be seen, and the town ditch was dug before 1218.71 Finally, two fortified cathedral towns were separated from their royal castles after the death of King John. New Salisbury was created a mile away from Old Sarum castle and town, and most of the site of Worcester castle was returned to the monks of the cathedral priory where John was buried, and then a long sequence of murage grants were given to the city.72 CONCLUSION Clearly, the works at Rochester (including the town wall and ditch of c. 1192) are part of a long history. Concurrent expenditure on castle and town defences, on an ad hoe basis, can be traced back elsewhere to the twelfth century, and perhaps more systematically from the 356 REFORTIFICATION AT ROCHESTER IN THE 1220s beginning of John's reign (Southampton, Dublin), when two magnates at least (William Marshal I and Peter des Roches) were fortifying both town and castle on their principal estates long before they undertook the regency. However, Hubert de Burgh, who succeeded them, only linked such operations on behalf of the Crown and not for himself, apart from Montgomery (afterward). Who initiated payment from Crown to town? Charles Coulson has demonstrated how, at Canterbury and elsewhere, town authorities lobbied for grants in the later middle ages. 73 Earlier royal exhortations to citizens to fortify their towns (Dublin, Hereford, Shrewsbury, Stafford) may have been diplomatic refusals of similar requests, but appeals which were often later granted. A draft charter of 1215 prepared by London, includes aid for its defences 'as of right' at a time when royal grants of timber, money and a loan were promised, and the rebels were repairing the town walls and gates with stone from plundered Jews' houses.74 Whether King John's various grants of 1215 amounted to a deliberate policy to ally himself with the towns to counter the rebel barons is arguable. Like Lincoln, London and Winchester, Rochester suffered heavily in the war of 1215-17. Richard Bales has shown how the royal castle policy during the minority of Henry III operated by negotiation rather than confrontation, and how the sensitive issue of the redistribution of the control of royal castles was skilfully managed. 75 Each of the protagonists had a separate agenda. The 'partnership policy' between Crown and town, although not ubiquitous, certainly continued throughout the 1220s in England whilst being transmuted into the murage grant. NOTES 1 C. Flight and A.C. Harrison, 'The Southern Defences of Medieval Rochester', Ar• chaeologia Canriana, cm (!986), 26. 2 H.M. Colvin (gen. ed.), The Historyofrhe King's Works: The Middle Ages (London, 1963), vol. I, 6-8, 24-25; Rotuli Litterarum Clausarum (ed.) T.D. Hardy (London, 1833) vol. I, 231. 3 Pipe Rolls: Canterbury (13 Henry JI, 196,201; 14 Henry II, 153; 20 Henry II, 2): Carlisle (31 Henry I, 140; 11 Henry II, 54; 2 Richard I, 49): Chester (7 Henry n, 35: 8 Henry II, 20-21): Hereford (2 Richard 1, 45, 49): London (I Richard I, 223; 3 Richard l, 136): Rochester (4 Henry II, 180; 5 Richard I, 166). 4 H.L. Turner, Town Defences in England and Wales: an architectural and documentary study, A.D. 900-1500 (London, 1971), 30; D.A. Whitehead, 'The historical background to the city defences', in R. Shoesmith (ed.), Hereford City Excavations 2: Excavations on and close to the defences, CBA Research Report 46 (London, 1982), 18-20. For an early thirteenth-century timber wall-tower there, seep. 56. 5 Pipe Roll, 5 Richard I, Introduction and 136. 357 DEREK RENN 6 Pipe Rolls, 4 John, 78-9; 5 John, 145. 7 Rotuli Lltterarum Clausarum (note 2), vol. 1, 6b; Rotuli Litterarum Patentium (ed.) T.D. Hardy (London, 1835), 69, 165b. A. Lynch and C. Manning, 'Excavations at Dublin Castle 1985-7' in (ed.) S. Duffy, Medieval Dublin, II, (2001), 169-204; James Lydon, 'Dublin Castle in the Mjddle Ages', in ibidem, III (2002), 115-27. A grant of customs to fortify the city was made in 1221, and other grants in 1225: Patent Rolls 1216-25, 297; Rotuli Lillerarum Clausarum, vol. 11, 40b, 48b. 8 Pipe Roll, 14 John, 90. 9 Patent Rolls 1216-25, 238-9 repeated 445, S 14. 10 Turner, op. cit. (note 4), 23-26 and appendix C; Richard Bales, 'Castles and Politics in England 1215-1224', in (eds) P.R. Coss and S.D. Lloyd, Thirteenth Century England, II (1988), 23-43; C.L.H. Coulson, The Castles of Communities (in preparation), Appendix N. 11 Rotuli Litterarum Clausarum (note 2). vol. 11, 12, 19, 27b, 29b, 30b, 38, 56, 58b, 98b, transcribed and translated by W.H.St J. Hope, 'The Architectural History of the Cathedral Church and Monastery of St. Andrew at Rochester. II - the Monastery', Archaeologia Cantiana, XXIV (1900), 12-16. Colvin, op. cir. (note 2), vol. II, 808, note 9. 12 British Library: Cott. MS, Nero D II, f. 132. 13 Flight and Harrison, op. cit. (note I), alternative scenarios at pages 19-26. Thefossata of 1192-3 (note 5) may be a further complication, although dismissed by them. 14 P.J. Tester, 'Northgate B', in A.C. Harrison and C. Flight, 'The Roman and Medieval Defences of Rochester in the light of recent Excavations', Archaeologia Cantiana, LXXXIII (1968), 70-73, 85, 93-4; A.C. Harrison, 'Rochester East Gate, 1969', Archaeo/ ogia Cantiana, LXXXVII (1972), 125-8, 149-50, figs. 4, 5, 16. 15 G.M. Livett, 'Medieval Rochester', Archaeologia Cantiana, XXI (1895), plan facing 17. 16 C. Flight and A.C. Harrison, 'Rochester Castle, 1976', Archaeologia Cantiana, XCIV (I 978), fig. 2, sections 3 and 4. 17 G. Payne, 'Roman Rochester', Archaeologia Cantiana, XXI (1895), 13 citing E. Hasted, The History of Kent, octavo edition IV (Canterbury, 1798), 163; G .L. Gomme, 'Boley Hill, Rochester', Archaeologia Cantiana, xvu (1887), 181-88; A.A. Arnold, 'On Roman Remains discovered at Rochester', Archaeologia Cantiana, xvm ( 1889), 193-5; Canon S.W. Wheatley, 'Boley Hill, Rochester', Archaeologia Cantiana, XXXIX (1927), 159-64; idem, 'Boley Hill, Rochester, after the Roman Period', Archaeologia Cantiana, XLI (1929), 127-41. 18 G.T. Clark, 'Rochester Castle', The Archaeological Journal, XXXII ( 1875), 206,224; E.S. Armitage, The Early Norman Castles of the British Isles (London, 1912), 196-7; A.J.F. Dulley, 'Rochester, Boley Hill', Archaeologia Cantiana, LXXIV (1961), 197-8. 19 Livett, op. cit. (note 15), 55. Personal communication, 26 April 2002. 20 Livett, op. cir. (note 15), 44-47, plate III, no. 2 facing 45. Clark, op. cit. (note 18), 228 reported in a postscript that J .T. Irvine had found a ditch 12ft wide and 14+ft deep, containing Roman material and running N•S under the south transept. This does not quite agree with what Irvine wrote in 1888: Payne, op. cit. (note 17), 13. If correctly measured, it sounds more like a pit (or pits) than a ditch. 21 Gundulf's stone castle of 1087 x 89 may be that mentioned in Domesday Book, since David Roffe has demonstrated that the Book contains references to post-1086 events: Domesday: the Inquest and the Book (Oxford, 2001). 22 Flight and Harrison, op. cit. (note 16), fig. 2, section 2. 23 For the first two sieges, see The Anglo-Saxon Chronicle (ed.) G.N. Garmonsway (London, 1954), 79 and The Ecclesiastical History of Ordericus Vita/is, V. (ed.) M. Chibnall (Oxford, 1975), 210-11. 358 REFORTIFICATION AT ROCHESTER IN THE 1220s 24 Armitage, op. cit. (note 18), 200; Colvin, op. cit. (note 2), vol. II, 808 note I, 810. 25 Livett, lac. cit. (note 15). 26 W.L. Warren, King John (Harmondsworth, 1966), pp. 255, 263-4 citing T. Rymer, Foedera (Record Commission, 1816), i. i. 133. 27 R. Allen Brown, Rochester Castle, Kent: Official Guidebook (London, 1969); I. F. Rowlands, 'King John, Stephen Langton and Rochester Castle, 1213-15', in (eds) C. Harper-Bill, C.J. Holdsworth and J.L. Nelson, Studies in Medieval History presented to R. Allen Brown (Woodbridge, 1989), 267-79; J. Bradbury, The Medieval Siege (Woodbridge, I 992), I 39-40. 28 Walter of Coventry, Memoria/e (ed.) W. Stubbs (Rolls series 58, 1873), I, 227; Roger of Wendover, Flores Historiarum (ed.) H.G. Hewlett (Rolls series 84, ! 887), U, 149-50; Ralph of Coggeshall, Chronico11 A11g/icanum (ed.) J. Stevenson (Rolls series 66, 1875), 176. An anonymous information panel erected to the south of the castle [2002] states that the first tunnel from the south-west was baulked by the Roman wall foundations but that a second tunnel from the south-east was successful, being inside the Roman wall-line (Fig. 1, e. f). 29 Rotuli Litterarum Clausarum (note 2), Vol. I, 231 b. Wendover, op. cit. (note 28), 147. Military equipment was requisitioned very shortly after: Rotuli Litrerarum Clausarum, vol. I, 234b. 3° Flight and Harrison, op. cit. (note 16), 38, figs.4, 5. Had it been in the ditch proper, some evidence of the latter's profile should have been found. A trench running obliquely towards the Constable's gate of Dover Castle was similarly interpreted: T. Blash ill, 'The Castle of Dover', Jo11rnal of the British Archaeological Association, 40 (1889), 373-8. An alternative explanation is that both were part of the defences. 31 G. Payne, 'Rochester Notes', Archaeologia Cantiana, XXJX (1911), lxxxiv. The gravel found when the berm was repaired recently may overlie this (CAT, interim report, Archaeologia Canriana, CXVI (1996), 323). 32 Charles Bristow and Gerald Wait in 'Castle' presented by Marc Morris (Lion TV), broadcast by Channel Four on 15 May 2003. 33 Livetl, foe. cit. (note 15). The two fragments drawn west ofthedonjon, which do not seem to be described in his meticulous account, are presumably not those later reported by G. Payne, 'The Reparation of Rochester Castle', Archaeologia Cantiana, XXVIJ (1905), 189. A possible alternative explanation is that an outward-fallen Roman or Norman wall underlies the loopholed curtain and round tower, the squat buttress on the east face of the latter marking part of an older wall. The watching brief during the repaving has provided evidence for the sites of the south gates (CAT, interim report, Archaeo/ogia Cantiana, cxxn (2002), 357. At the British Archeological Association Conference at Rochester in July 2002, it was mentioned that a watching brief in 1996 had found evidence for a possible matte at the do11jo11 (Fig. I, f). 34 J. Heyman, The Stone Skeleton: strnctural engineering of masonry architecture (Cambridge, 1995), 23-26, 109-11. A simjlar surface sapping is reconstructed in John Goodall, 'Dover Castle and the Great Siege of 1 216', CMteau Gaillard, 19 (1999), 95-6, Pigs.4, 5. 35 Rotuli Litterarum Clausarum (note 2), vol. I, 238b. 36 D.F. Renn, 'The demolition of the keep at Benington', Antiquaries Journal, 41(1961), 96-7; H.S. Braun, 'Some notes on Bungay Castle'; 'Bungay Castle - report on excavations', Proc. Suffolk Inst. Archaeology and Natural History, XXII ( 1934-6 ), 113, 118, Figs, II, VIII and 211. Leslie Alcock, 'Excavations at Degannwy Castle, Caernarvonshire', The Archaeological Journal, CXXIV (I 965), I 90-201, figs. 3, 4. 37 Allen Brown, op. cit. (note 27), plans inside back cover: I 986 edition, 27. 38 J.C. Taylor, Guide to Rochester Castle and Keep (Rochester, 1951). 5. 359 DEREK RENN 39 Pipe Rolls, 19 Henry II, 88; 7 Richard I, 2; 8 John, 47. 40 Allen Brown, op. cit. (note 27), 18: 1986 edition, 15. D.A. Carpenter, The Minority of Henry Ill (Cambridge 1990), 27. 41 Colvin, op. cit. (note 2), vol. II, 811-12. 42 Date on rebuilt wellhead and Payne, op. cit. (note 33), 177-92. 43 Annales Monastici (ed.) H.R. Luard (Rolls series 36, 1865), II, 285; Histoire des Dues de Normandie et des Rois d'Angleterre (ed.) F. Michel (Paris, I 840), 171; Carpenter, op. cit. (note 40), 44, 104. 44 Pipe Rolls, 4 Henry Ill, 156; 5 Henry III, 20 I; 6 Henry ill, 57. Rotuli Litterarum Clausarum (note 2), vol. I, 447b, 495b, 577b; vol. II, 7b, 19 (quoted by Hope, op. cit. (note II), 13, 14, 84, 89. 45 Rolls of Divers Accounts for the early years of the reign of Henry Ill (ed.) F.A. Cazel, Pipe Roll Society, new series 44 (1982), xi, 4. 46 Colvin, op. cit. (note 2), vol. II, 807-8. 47 Carpenter, op. cit. (note 40), 328. 48 Allen Brown, op. cit. (note 27), 1986 edition, 11; idem, 'Les ch§.teaux feodaux', Cahiers de civilisation medievale, XXIX (1986), 39, citing A. Chil.telain, Donjons romans de pays d'ouest (Paris, 1973) for the restricted distribution of round-buttressed rectangular donjons in France. 49 M-P. Baudry, Les fortifications des Plantagenets en Poitou 1154-1242 (Paris, 2001), 105ff; ibidem, Les fortifications dans les domaines plantagenets Xlle-XJ!le siecles: actes de colloque international tenu a Poitiers du 11 au I 3 novembre, 1994; Civilisation Medievale X (2000), 23-39. 50 Royal and other historical letters illustrative of the reign of Henry Ill ( ed) W. W • Shirley (Rolls Series 27, I 862), I, 383-4 and passim. 51 Colvin, op. cit. (note 2), vol. II, 747 note 3. 52 These openings are inaccessible at present, and the description is based upon Payne, op. cit. (note 33), I 78 and Allen Brown, op. cit. (note 27). 53 Flight and Harrison, op. cit. (note I), 7 note 23. Although Allen Brown, op. cit. (note 27), inside front cover of 1969 edition, dated it to the fourteenth century, John Cherry prefers a mid thirteenth-century date: 'Imago Castelli; the depiction of castles on medieval seals', Chtiteau Gaillard, XV (1992), 85-8.The donjon flying a flag of the royal arms can be compared with the scene after the capture of Bedford castle in 1224, as drawn by Matthew Paris: Chro11ica Majora, Parker Ms. 16, Corpus Christi College, Cambridge. The round donjon shown there may be wrong: Paris drew the White Tower as round, but he got the square proto-Beauchamp Tower right. 54 A.A. Arnold, 'Medieval Remains at Rochester', Archaeologia Cantiana, xvn (1889), 196-200. 55 Colvin, op. cit. (note 2), vol. II, 811-13. 56 Flight and Harrison, op. cir. (note 16), 32-36. 57 A variant of the earlier styles at Dover, where there are multiple slits in a single embrasure (inner ward towers) and multiple embrasures to a single slit (Avranches traverse). 58 Page 355. 59 British Library, Add. Mss. 32,370 folio 183. The writer owes this reference to Jeremy Ashbee. 60 Arnold, op. cit. (note 54), facing 200, reproduced in Allen Brown, loc. cit. (note 27). 61 Rotuli Litterarum Clausarum (note 2), vol. II, 98b; Calendar of Liberate Rolls 360 REFORTIFICATION AT ROCHESTER IN THE 1220s 1226-40, 258. But Livett, op. cit. (note 15), 31 argued that the overhang and 'foreign' mortar adhering to the chase was due to it being built up against an earlier (i.e. presiege) wall. 62 M.W. Thompson, Kenilworth Castle, Warwickshire: Official Guidebook (London 1985); Colvin, op. cit. (note 2), vol. II. 683; Pipe Rolls, 13 John, 188, 190, 191; 14 John, I 36; 17 John, 28; 4 Henry III, 30; Rotuli Litterarum Clausarum (note 2), vol. I, 422b, 534; vol. II, 18. 63 Although the city changed hands several times between 1214 and 1217, the castle was not captured. Over £500 was spent on repairs shortly thereafter: probably Cobb Hall, the horseshoe plan vaulted tower with multiple arrowloops, was added at this time; Pipe Roll, 2 Henry III, 34; Colvin, op. cit. (note 2), 705. In 1224-5 the Lucy Tower and eastern gate were mended and a barbican made for the latter. The broken gates were to be surveyed and the tower of the east gate completed in 1227, £20 being spent on the gates in the following year: Calendar of Liberate Rolls 1226-40, 26-7, 46; Cazel, op. cir. (note 45), 44, 62; D.F. Renn, 'Cobb Hall Tower' [Lincoln Castle Conferenc.e 1995], Occasional paper 13 of the Lines. Soc. for Archaeology and History, forthcoming. £200 spent on enclosing the town in John's reign was refunded in 1225, when mu rage grants began. A new south gate is mentioned 1217 x 1220 and in 1228 £ I O was expended on walls and towers at Wigford, possibly including the round tower excavated at Brayford Pool. Rotuli Litterarum Clausarum (note 2), vol. I, 580: vol. II, 28b, 31; Patent Rolls 1216-25, 518-9; C. Johnson and A. Vince, 'The South Bail Gates of Lincoln', Lincolnshire History and Archaeology, 27 ( I 992), 13; Calendar of Liberate Rolls 1226-40, 63; C. Colyer, 'Excavations at Lincoln 1970-72: the Western Defences of the Lower Town: an interim report', Antiquaries Journal, LV (1975), 260-2, fig. 9, plates XLII, XLIII. 64 According to John Stow, writing at the end of the sixteenth century (A Survey of London (ed.) H.B. Wheatley (London, 1956), 19, 250) the city ditch was begun in 1211 and finished in 1213, while in 1215-16 the barons 'brake the Jews' houses, rifled their coffers and, with the stones of their houses, repaired the gates and walls of the city'. Kingsford's edition (I, 38) records Hebrew inscriptions being found in the rebuilding of Ludgate. Royal orders to provide timber and gifts and loans of money for strengthening the city were enrolled in the same year. One entry refers to the wall 'between the Tower and the town wall'. That town ditch, as well as its predecessors and successors, has been excavated. The hollow wall towers west of Moorfields are probably medieval: the pottery found above and below the floor of Bastion 14 indicates a date of c. 1225 and Bastion 21 is mentioned in a 1235 lease with other turrets on the same wall. Rotuli Litterarum Clausarum (note 2), vol. I, 198 (bis); Close Roll 1215-16 (Pipe Roll Society, new series 31, 1955), nos. 15, 16, 38; Calendar of Patent Rolls 1232-47, 106-7; Coventry, op. cit. (note 28), 11, 212; J. Maloney, 'Recent work on London's defences', in (eds) J. Maloney and B. Hobley, Roman Urban Defences in the West, CBA Research Report SI (London, 1983), 105-11; G. Milne, Excavations at Medieval Cripplegate, London: Archaeology after the Blitz 1946-68 (London, 2002), 28-9. As well as continuing repairs to the Tower of London, two new towers were begun, one being the Wakefield Tower (by 1226). This is a large round tower occupying the angle between the river wall and the loopholed Main Guard of the Inmost Ward. The lowest storey is a regular octagon internally, each face having a similar embrasure. Five still form arrowloops and the others may represent changed plans. Of the latter, one embrasure would have opened behind the Great Hall, another is blocked by the Bloody Tower's guardroom and the third forms a side passage to the pastern gate below the royal apartments of which the stylistically later upper floor of the Wakefield Tower formed part. P.E. Curnow, 'The Wakefield Tower, Tower of London', in (eds) M.R. Apted, R. Gilyard-Beer and A.D. Saunders, Ancient monuments and their interpretation: essays prese11ted to A.J.Taylor (Chichester, 1977), 155-89. 361 DEREK RENN 65 Winchester, like London, changed hands several times during 1215-17, and both royal and episcopal castles needed extensive repair. The tower of the royal castle was being mended in 1220-4 while a new tower was begun in 1222, the ditch having been cleaned out previously. The base of that round tower is pierced by a stepped passage from the razed Norman donjo11, a passage which forks to doorways either side of the town wall, giving onto the castle ditches both within and without the town wall. Rot11li Litterarum Cla11sarum (note 2),v ol. II, 226,240; Patent Rolls 1216-25, 44, 52; Colvin, op. cit. (note 2), vol. II, 857-9; Martin Biddle, 'Excavations at Winchester, I 970: tenth and final interim report, Part I', Antiquaries Joumal, LV (1975), 102, fig. 3 and XLlX (1969), plate LIX; L (I 970), 289-92, fig. 2, plate XLb. In 1215 the town had a grant of timber to floor its towers and alures, and in 1217 the north gate was to be used as an official weighing office. . Contemporary work on a mural tower was found in Tower Street: Martin Biddle 'Excavations at Winchester, 1964: fourth interim report', Antiq1wries Journal, XLV (1965), 238-9, plate LXXIIb. 66 Colvin, op. cit. (note 2), vol. II, 633-7; R. Allen Brown, Dover Castle, Kent: Official Guide (London, 1966). H.M. Colvin, Building Acco1mts of King Henry Ill (Oxford, 1971), 1-87; Goodall, loc. cit. (note 34). A tunnel runs under the blocked north gatehouse passage, through the chalk to a four-storey round tower and onwards to an earthwork barbican. The tower, protected by drawbridge and portcullis, has a doorway on each side opening into the bottom of the moat. Only one arrowslit is now visible, pointing back into the castle, although another may once have existed at the same level but facing outward, and a third at a lower level between the inner doors. The curtain walls were extended to the sea cliff, new gates inserted and a new inner earthwork created round the Saxon church and Romanpharos: (eds) D.M. Wilson and D.G. Hurst, 'Medieval Britain in 1962 and 1963', Medieval Archaeology, VIII ( 1964), 254. The Norman and later town of Dover lies in the river valley well below the castle; the earliest mention of town walls is in 1231 (Dr S. Sweetinburgh, pers. comm. 15 April 2002: see also pp.183-207 of this volume). 67 The two towers occupying re-entrant angles south of the m otte are dated to 1223. A possible explanation of their plan and purpose will be offered elsewhere. The great west front, with three large D-plan vaulted towers, begun 1227-30, is an extension of the castle overlooking the unfortified borough. Sir William St John Hope's great two-volume study of the castle (London 1913) is being updated by Drs Stephen Brindle and Stephen Priestley in the light of later research. A short resume of work on the defences is given by Virginia Jansen: 'Henry ID's Windsor: Castle-Building and Residences', in (eds) Laurence Keen and Eileen Scarff, Windsor: Medieval Archaeology, Art and Architecture of the Thames Valley, British Archaeological Association Conference Transactions, XXV (2002), 98-100. 68 Work on the new stone castle began in the spring of 1224; the first work was the inner ward, entered through a twin-towered gatehouse with a single surviving arrowloop. A D-plan tower occupies the far end and a much larger one (rebuilt and now ruined) the western re-entrant. Calendar of Charter Rolls 1226-57, 10, 40, 101; Patent Rolls 1216-25, 414,501. J.K. Knight, 'Excavations at Montgomery Castle, Part l', Archaeologia Cambrensis, CXLI ( 1992), 100. Charters to the burgesses,a llowing them to enclose the town with a wall and ditch, were granted in 1227. B.H.St.J. O'Neil and A.H. Foster-Smith, 'Excavations at Montgomery Town Wall', ibidem, XCV (1940), 217-9; A.J. Taylor, 'Montgomery Town Wall', ibidem, XCIX (1947), 281-3. 69 Richard Avent, Cestyll Tywysogion Gwynedd/Castles of the Princes of Gwynedd (London, 1983). 10 D.J.C. King, 'Pembroke Castle' Arcliaeologia Cambrensis, CXXVIl (1978), 75-121; idem and Mark Cheshire, 'The town walls of Pembroke', ibidem, CXXXI (1982), 77-84; N. Ludlow, 'Pembroke castle and town walls', Fortress, 8 (1991), 25-30; I.W. Rowlands, 'William Marshal, Pembroke Castle and the Historian', ClzlJ· teau Gaillard, XVII (1996), 151-5. 362 REFORTJFICATION AT ROCHESTER JN THE l220s 71 [Taunton] N.R. Holt, Pipe Roll of the bishopric of Winchester 1210-11 (Manchester, 1964), 170-1; T.J. Hunt, 'Taunton Castle in the year of Magna Carta', Notes on the History of Taunton Castle (County Museum, 1954), 9-12 and idem, 'The medieval customs of the manors of Taunton and Bradford on Tone', Somerset Record Society, LXVI (l 952); A.W. Vivian-Neal and H.St. George Gray, 'Materials for the History of Taunton Castle' Proceedings of the Somersetshire Archaeological and Natural History Society, LXXX VI (1940), 45-78; C.A.R. Radford and A.D. Hallam, 'The History of Taunton Castle in the Light of Recent Excavations', ibidem, XCVII (1953), 55-96; A.D. Hallam, 'Further Excavations at Taunton Castle', ibidem, CIX (1965), 98-103; P. Leach, The archaeology of Taimton: excavations and fieldwork to 1980, Western Archaeological Trust monograph 8 (1984), 11-12, 30, 59, 69-73. [Farnham] (ed.) R. Poulton, 'Archaeological investigation of historic Surrey towns: Chapter 4: Farnham', Surrey Archaeological Collections, 85 (1998), 102-4, 125, 147-51. 72 New Sarum had been planned well before 1219, and the charters of 1225 and 1227 said that it was to be 'enclosed with adequate ditches': The City of Salisbury, Royal Commission on Historic Monuments, England (London, I 980), I, xxxii, 50-5 I. The grant of most of Worcester castle to the monks was backdated to the death of King John: Calendar of Charter Roils 1226-57, 154. Aids for enclosing city 1224-6: Rotu/i Litterarum Clausarum (note 2), vol. I, 426; vol. II, I0!b, 131. 73 C.L.H. Coulson, 'Battlements and the Bourgeoisie: Municipal Status and the Apparatus of Urban Defence in Later-Medieval England', in (eds) S. Church and R. Harvey, Medieval Knighthood, V (Woodbridge, 1995), 133-7. 74 British Library: Add. Mss. 14252, f.124: 'vi. De claustura civitatibus et adiuuaretur': Mary Bateson, 'A London Municipal Collection of the Reign of John', English Historical Review, XVII (I 902), 726. Also note 64 above. 15 Loe. cit. (note 10). 363