Chateau Gaillard XIX: Actes du Colloque International de Graz, 1998 (2000)
Commanding the shortest sea crossing between England and the Continent, Dover Castle was a vital strategic and communication lynch-pin in the empire of the Angevin kings of England. They developed Dover’s fortifications on a spectacular scale from 1180 onwards and in 1216 this great castle successfully resisted a major siege directed personally by Prince Louis of France during his near-successful invasion of England. An unusually detailed account of this siege survives in the contemporary Histoire des Ducs de Normandie et des Rois d’Angleterre. This account is not only remarkable for its vivid description of the siege, but also for the light it throws on the architecture of the castle itself in 1216. It demonstrates that Dover’s outer defences, built by Richard I and John, were far more sophisticated than was hitherto believed and incorporated what was possibly the first twin drum-towered gatehouse in England.
As it stands today Dover is a development of a castle begun in 1180 by Henry II. He effectively demolished an existing castle on the site and began a vastly ambitious new set of fortifications within the U-shaped perimeter of what is probably the ditch of an Iron Age hill fort above the cliffs of Dover. At the centre of this hill fort site he completed a square keep surrounded by an inner bailey and, around the perimeter, began work on a great encircling wall in stone.1 This outer wall was almost certainly a replacement of an existing line of defence which, contrary to what some scholars have argued, extended right down to the cliffs at the southern end of the castle enclosure.2
Henry II died before the completion of this great castle but the work was continued by his sons Richard I and King John, both of whom are recorded to have spent considerable sums of money here.3 Unfortunately the precise nature of their work is largely undocumented.4 What it is clear however from the evidence of the extant buildings from the events of the siege itself and from the records of repairs effected afterwards, is that between them they continued the replacement of the outer defence to the castle in stone. This work included the construction of a great gatehouse at the northern tip of the castle enclosure, and the extension of the stone defences in an anti–clockwise direction around the entire northern end of the castle at least as far as Peverill’s gate. Judging by the sums of money known to have been spent on Dover by these two kings it seems likely that John undertook the lion’s share of this work, including the gatehouses.5 As is discussed below, the architecture of the defences further substantiates this attribution.
Henry II’s castle at Dover has been much discussed by scholars: its architecture is of seminal importance in the development of fortification in the late twelfth-century.6 So too is the thirteenth-century work of Henry III, who substantially remodelled parts of the castle after the siege and erected the two principal gatehouses which still serve it today.7 But the construction undertaken by John and Richard, between the reigns of these two King Henrys, remains virtually unknown and has usually been relegated to the status of a footnote in the castle’s history. The principal reason for this is that very little of it survives: the siege and the later medieval and modern rebuilding campaigns – most notably during the reign of Henry III, the Napoleonic Wars and the two World Wars – have carried away all but the smallest fragments of it.
But these lost northern defences, constructed by two kings renowned for their castle building, are of exceptional importance and interest. To reconstruct them, however, from the fragmentary physical evidence alone presents serious difficulties. These may be overcome in part by inferring information from the documented events of the great siege in 1216.
The siege is of interest not only as a case-study in the warfare of this period, but also because it was one element in a coherent military campaign by Prince Louis to assert his control over England. Nothing could more clearly illustrate the role of a great castle like Dover in a national political crisis than the events of 1216-17.8 Although the following narrative is necessarily very concise, it describes the siege in the broader context of national events. It is illustrated by reconstruction drawings prepared by Peter Dunn for English Heritage. To my knowledge no one has tried accurately to depict the progress of a particular medieval siege before. Despite the fact that this approach involves a great deal of invention it is, I believe, still of value as a means to visualizing and understanding the physical realities of siege warfare. It may be temerarious to present drawings of this kind in an academic journal, but the text below seeks to justify – as far as is possible – the principal reconstructed elements.* One feature which requires explanation at the outset is the representation of the siege works. This is based on the tentatively identified set at Berkhamstead, built by Prince Louis in 1216 when he invested that castle.9
In 1216 the fate of the Angevin Empire was hanging in the balance. King John, whose vast inheritance had already been greatly diminished by the conquests of Philip Augustus, was embroiled in an English civil war which threatened his very throne.10 A group of rebel barons, provoked at what they described as John’s tyrannical rule, had risen against the king and established themselves in London, from where they invited Prince Louis of France, the eldest son of Philip Augustus, to assume the crown. When Louis landed in England on the 21st May 1216 John fled before him; Canterbury opened its gates to his army: Rochester Castle fell after a short siege, and the French prince entered London in triumph on 2nd June.
After securing the capital, Louis marched west to take Winchester. By this time his success was attracting many important figures from John’s allegiance as well as the support of Alexander, the king of Scotland, and Llywelyn, Prince of North Wales. Winchester fell within a month, along with several important neighbouring castles. With England’s two principal cities under his control, Louis now turned his attention to capturing the stronghold which governed his communications with France: Dover Castle.
Dover’s Constable at this time was a staunch supporter of King John, Hubert de Burgh, the earl of Kent and Justiciar of England.11 He was a veteran soldier who had led the heroic defence of the castle of Chinon in 1205, and his garrison, comprising more than 140 knights and. a great number of men at arms, was plentifully supplied.12 To have taken Dover would have placed Louis in a very powerful position and the success or failure of the castle’s defence was of great importance. It was indeed in relation to the events of 1216 that Matthew Paris famously described Dover Castle as “Clavis Angliae”, the key of England.13
Several contemporary accounts of Louis’ siege of Dover survive but most are brief and conflate events, or are simply unreliable. The striking exception to this is the account in Histoire des Dues de Normandie et des Rois d’Angleterre, an anonymous work in French which, through its verisimilitude of detail, is generally regarded as having been written or informed by an eyewitness of the events themselves.14
According to the History, when Louis arrived in Dover he spent several days with his army billeted or camping in and around the town.15 The castle, set high on the hill above, was not under siege at this time and the garrison took the opportunity to parade itself in full armour on several occasions outside the barbican of the main castle gate.16 All evidence of this barbican has disappeared today but it is described in the account as being enclosed with palisades of oak and surrounded by a wide ditch.17 During these parades the crossbowmen of Louis’ army went to shoot at the garrison and, in a curious touch of detail, the History recounts that on one occasion during this bizarre ritual a crack crossbowman named Ernaut came close enough to the garrison’s line-out to be run upon and captured.18
The siege proper began in mid-July 19 when Louis divided his forces, one part remaining in the town, and the remainder moving to an encampment on the hill in front of the castle.20 He also sent his fleet to sea, in order, as the chronicler explains, to close in the castle on all sides – one of several intimations given that the castles defences reached down to the cliffs at this date.21 Perriers and mangonels were then set up to bombard the walls and gate, and a wattle siege tower – described as “a very high castle made of hurdles” – was erected. Given the subsequent events of the siege and the geography of the site, this must have stood before the north gate. Louis then, using a cat to protect his miners (perhaps in this case a covered gallery driven across the ditch), undermined the barbican’s timber palisade. The breach was stormed and the barbican fell.22 We are told that Huart de Paon, a horse soldier who bore the banner of the Lord of Bethune was the first to mount the breach, and that the captain of the gate and barbican, Pierre de Creon, was mortally wounded in the fighting.23
Louis now pressed his attack, sending miners to dig beneath the castle gate. “They mined,” the chronicler relates, “so that one of the towers fell, of which there were two.”24 Large numbers of Louis’ men managed to enter the castle through this breach but, after a sharp encounter, they were repulsed and the gap in the defences successfully blocked with great timbers, including crossbeams, which must have been stripped from the castle’s buildings.25
It deserves mention that all the tactics described in the siege up to this point are analogous to those employed by Philip Augustus at Chateau Gaillard in 1204. Is this a text-book Capetian siege of a major stronghold?
After the failure of the assault Louis struck a truce on 14th October.26 Almost immediately afterwards, on 18th October, King John died with only a nine year old boy, Henry Ill, to succeed him. An attempt was made to persuade the garrison to acknowledge the French prince as king instead of the young boy,27 but this offer was rejected and Louis left for London after nearly three months spent before the gates of Dover. Curiously, the siege encampment appears to have been left as it stood because the History also relates that it was destroyed and the guards killed a few months later in 1217.28
This reverse was to be of decisive importance in Louis’ bid for the English throne. The Dover garrison kept its truce badly, harrying his communications and preventing the landing of reinforcements. So discomforting was this that in 1217 he was forced to invest the castle again. According to the History, the second siege began on 12th May when Louis again encamped on the hill in front of the castle. He set up a trebuchet, which proved ineffective, and busily began the construction of “maisons”, possibly fortified positions of some kind.29 Louis’ trebuchet may have been the first ever seen in England and the History itself remarks on its rarity.30 Mathew Paris may have been refering to it when he mentions a great catapult at the siege called Malvoisin, the bad neighbour.31
The investment of Dover necessitated the division of Louis’ forces across England and in his absence one half of his army was destroyed at the battle of Lincoln on 20th May 1217. When news of this reached Louis he disassembled his trebuchet at Dover and, shortly afterwards, moved to London.32 The disaster at Lincoln was compounded on 24th August by the battle of Sandwich, a sea battle in which an English fleet led by Hubert de Burgh inflicted a crushing defeat on a new incoming force of men and supplies as they crossed the Channel.33 In its aftermath Louis gave up his attempt on the English throne and tried to negotiate an honourable withdrawal from England.
Having considered the siege itself it is now possible to return to the fabric of Dover Castle to look at the fragmentary remains of its Angevin outer defences. In the space available this article will focus specifically on the northern defences and great gate – the area where the siege was contested.
The fabric of the northern tip of the castle is an impossibly complex hotchpotch of different periods of documented and undocumented building work, to such an extent indeed that, in the absence of a full archaeological survey, much informed guess-work is involved in establishing its development. Externally the medieval castle has been smothered by eighteenth- and early nineteenth-century accretions. In this period the backs of the walls were buttressed with earth and backed by casemates; the towers cut down for the sighting of artillery; and an elaborate fortified spur, incorporating a series of medieval tunnels, con–structed to the north of the walls.34
If, in the mind’s eye, all this modern work is removed, the 13th-century fabric is exposed. This is essentially of two periods. The first belonging to the pre-1216 fortress, and the second to Henry III’s remodelling of the castle after the siege.
Some details of these two periods of construction are easy to distinguish. For example it has long been known that a curious trinity of towers, known as the Norfolk Towers, at the northern tip of the castle’s curtain, incorporate the remains of the original gate to the castle. Today these towers, truncated late in the eighteenth century to serve as a gun battery, comprise a beaked tower set between two flanking round towers. The central beaked and eastern towers are solid masonry and have been constructed over and against the western tower. This latter structure contains an interior vaulted chamber on the ground floor and the remains of two tiers of arrow loops.
In the siege, as has been described, one of the two gateway towers to the castle was collapsed by Louis’ miners. What appears to survive here is the remaining tower, and Henry III’s repair of its breached defences: essentially two solid towers, one replacing the collapsed gate-tower, and a second blocking the old gateway. Some excavation work has actually revealed the old gateway itself, still incorporated within the solid masonry of Henry III’s work, and what appears to be the road surface of the original entry.
It is possibly a reflection of the difficulty of access to this area that no one has tried to piece together any details of this gate. The task, however, is a rewarding one and the ghost of a substantial gatehouse can be readily identified. Incorporated within the eighteenth-century work are two parallel projecting walls constructed in fine Kentish Rag ashlar which formed the sides of this building. The surviving chamber of the western gate-tower gives an impression of the high quality of this sadly mauled building. It is plainly vaulted and its floor level has been lowered, as is shown by the roughly-hewn faces of the walls beneath the arrow loops. These loops are themselves finely constructed in Kentish Rag ashlar blocks and interestingly their stonework has been burnt crimson in some places, presumably evidence of the fire-damage inflicted by Louis’ mine in 1216. They are arranged at regular intervals round the tower and originally had steeply slanted sills which have been cut away to make them serviceable after the floor was lowered. All are inclined at different angles, a feature which suggests that some thought went into their positioning.35
From the evidence of the excavations the gateway appears to have comprised a segmental opening-arch set beneath a two-centred arch, the latter serving to decorate the facade, and to receive the drawbridge. No physical evidence for the form of the gate-passage survives today.
According to the History, this gatehouse was protected by a substantial barbican enclosed by an oak palisade and a wide ditch. In the past scholars have simply ignored this reference to a barbican and, as a result, have assumed that all the recorded medieval earthworks and structures to the north of the gatehouse post-date the siege. Early surveys of the castle show two such features, both of which have been attributed in their entirety to Henry III’s rebuilding work of the 1220-30s: a barbican and, connecting this with the castle itself, a series of tunnels running under the gatehouse. That Henry III was responsible for extensive work in this part of the castle is without doubt, but the documentation recording this work is not very clear about what he actually did and a tentative case may be made that these features existed in some analogous form prior to the siege.36
The evidence for this is two-fold. First, with regard to the barbican. If there was a substantial earth and timber fortification in this area during the siege, its logical position would have been the site of the present spur, which is a remodelling of an area of high ground adjacent to the castle with obvious strategic value. It seems reasonable to argue therefore that the northern barbican, which formerly stood here, was substantially a pre–1216 structure. In the reconstruction drawings this barbican is represented as being D-shaped – modelled in plan on that built in 1216 at Oxford Castle 37 – but, since there is no evidence that Henry III altered its shape, it might equally have been shown in the circular form recorded in early maps of the site.
Second, regarding the tunnels. Since Henry III blocked the existing gate and made the gatehouse a redundant structure, it seems remarkable that the underground works seem to respect its lay-out. Not only does the main passage leading to the barbican pass underground between the gateposts of the blocked gateway, but its entrance curves right round to one side as if to avoid obstructing the old gate-passage entrance. At one point this passage is bisected by what appears to be an earlier, but incomplete, tunnel running at higher level towards the moat. Although the case is possible to prove, these details would make sense if Henry III actually greatly enlarged and extended a system of tunnels, possibly the remains of permanent counter–mining galleries, that had first been constructed in relation to the gatehouse.
There may be corroborative evidence for this suggestion too in the fact that John is documented to have employed miners in his building work at Dover,38 and also that there exists a two-tiered system of passages running through the walls of the northern curtain and its towers. The passages are blocked in those areas repaired, or likely to have been repaired, by Henry III – such as the faces of the mural towers or at the base of the collapsed gatehouse. It therefore seems probable that these are the remains of a more extensive pre-1216 network damaged by the siege. The construction of such mural passages and (presumed) counter-mining galleries would be entirely consonant with castle architecture from the 1180s onwards.39 From the evidence of these passages and one high set mural arrow loop, it may also be inferred that at least one of the mural towers (and therefore probably all) stood in 1216, and that a line of buildings stood against the curtain at this point.
In the light of this evidence a picture of a spectacular northern defence to Dover in 1216 emerges: a twin-towered gate–house with a line of flanking towers, connected by passages above and below ground, that commanded a deep-cut ditch along the head of the spur on which the castle stands. Beneath this and separated from it by a wide ditch was a barbican. Presumably access to this was across a steep wooden bridge, and some idea of its scale can still be gleaned from the existing geography and the plan of the castle as it was recorded in 18th-century surveys.40
Dating this much altered line of northern defences without the help of good documentation is problematic. There is nothing to exclude the possibility that Richard I had a hand in this work, but on balance there are three pieces of evidence which point to John’s responsibility for most, or even all, of it. First, John spent a great deal more on the castle than Richard.41 Second, the drum shaped towers are reminiscent of architecture attributed to him elsewhere – for example at Scarborough.42 Finally, the twin drum-towered gatehouse at Dover, although it may be compared in some details to the gatehouse of Pevensey Castle (possibly of the 1190s)43 – would find more convincing parallels in the architecture of the following decade: it is closely comparable to developments in French design, as exemplified at Philip Augustus’ Louvre, and also with military architecture in Ireland, particularly at Trim and the royal castles at Limerick and Dublin.44
Whatever the precise dating of the work, the existence of the fully-developed gatehouse form at Dover in 1216 challenges the orthodoxy that such structures did not develop in England until the 1220s. This means that the suggested late twelfth or early thirteenth century dating of gatehouses at castles such as Warkworth and Pevensey must be taken seriously. If these buildings are early they pose a serious challenge to the accepted typological development of gateway defences by com–pressing the variety of their forms into an improbably small period of time.45 Military architects around 1200 were not necessarily building sequentially better-designed gateway defences to ever more elaborate plans in an evolutionary struggle between attacker and attacked, but executing a wide variety of designs at the same time: gate-towers as well as twin-towered and single towered gateways and gatehouses; incorporating round, square and polygonal forms. All this would be further proof that the strictly military interpretation of these structures is inadequate and that their builders should also be credited with a real interest in, and flair for, the construction of architecturally magnificent fortifications.
* Please note that several illustrations are included in the original version, and are not reproduced here.
I am particularly grateful to Jonathan Coad, Peter Dunn and Kevin Booth for their help with this article. Also to English Heritage for enabling me to research and present it.
Beeler, J. H., “Castles and strategy in early Angevin England”, Speculum 31 (1956), p.581-601
Brown, R.A., “Royal Castle Building in England, 1154-1216”, The English Historical Review v.70 (1955), p. 353-398
Brown, R.A., amd Colvin, H.M., “Dover Castle”, The History of the King’s Works: The Middle Ages vol.1 (London, 1963), p.629-641
Cannon, H.L., “The Battle of Sandwich, and Eustace the Monk”, The English Historical Review v.27 (1912), p. 649-70
Carpenter, D.A., The Minority of Henry III (London, 1990)
Coad, J.G., and Lewis, P.N., “The later fortifications of Dover”, Post Medieval Archaeology v.16 (1982), p. 141-200
Coad, J., Dover Castle (London, 1995)
Colvin, H.M., The Building Accounts of King Henry III (Oxford, 1971)
Curnow, P.E., “Some developments in military architecture, c.1200: Le Coudray-Salbart”, Anglo-Norman Studies v.2 (1979), p. 42-69
Eales, R., “Castles and politics in England, 1215-1224”, Thirteenth-Century England v.2 (1988), p. 23-43
Ellis, C., Hubert le Burgh (London, 1952)
Goodall, J.A.A., “Dover Castle”, Country Life (1999), p. 44-47 & p. 110-113
Goodall, J.A.A., Pevensey Castle. English Heritage Guidebook (1999)
Goodall, J.A.A., Scarborough Castle. English Heritage Guidebook (2000)
Gransden, A., Historical Writing in England I, 550-c.1307 (London, 1996)
Hardy, T.D. (ed.), Rotuli Litterarum Clausarum (London, 1833)
Hassall, T.G., “Excavations at Oxford Castle, 1965-73”, Oxoniensia v.41 (1976), p. 232-254.
Kendall, J.M., “The Siege of Berkhampstead Castle in 1216”, The Antiquaries Journal v.3 (1923), p.37-48.
Kenyon, J.R., Medieval Fortifications (Leicester, 1990)
King, D.J.C., The Castle in England and Wales (London, 1988)
Knight, J.K., “The road to Harlech: aspects of some early thirteenth-century Welsh castles”, Castles in Wales and the Welsh Marshes. Essays in honour of D.J. Cathcart King, eds. J.R., Kenyon and R. Avert (Cardiff, 1987), p. 75-88.
Luard, H.J. (ed.), Mathaei Paresiesis Chronica Majora vol. 2 and 3 (Rolls Series 57) (London, 1874-1876)
Michel, F. (ed.), Histoire des Ducs de Normandie et des Rois d’Angleterre (Paris, 1840)
McNeill, T., Castles in Ireland (London, 1997)
Mesqui, J., Chateaux et Enceintes de la France Medievale (Paris, 1991)
Parfitt, K., “A lost earthwork near Dover Castle”, Kent Archaeological Review (1995), p. 121, 10-11
Renn, D.F., “The Avranches Traverse at Dover Castle”, Archaeologia Cantiana v.84 (1969), p. 79-92
Renn, D.F., “An Angevin gatehouse at Skitpon Castle”, Chateau Gaillard v.7 (1975), p. 172-182.
Stenton, D.M., et al. (eds.), “Pipe Rolls 1189-1214”, Pipe Roll Society New Series 1-37 (1925-61)
Stevenson, J. (ed.), Radulphi de Coggeshale Chronicon Angliacanum (Rolls Series v.66) (London, 1875)
Warren, W.L., King John (London, 1978)
Weiss, M., “The Castellan. The early career of Hubert de Burgh”, Viator v.5 (1974), p. 235-52.
1. The medieval development of the castle is described in detail in BROWN/ COLVIN 1963, p. 629-641 and CORD 1995. p. 12-52. Henry II’s work discussed in RENN 1969, p. 79-92 and considered in the context of the cult of St Thomas Becket and Canterbury Cathedral in GOODALL 1999, p. 44-47 & 110-113
2. Only by extending to the top of the cliffs can the castle control the harbour. This arrangement is implicit in the events of the siege in 1216 (see below).
3. The documented total spent by these kings on Dover is in excess of £1,800. For a statistical breakdown of royal expenditure on castles, and for a discussion of the documentation, see BROWN 1955, p. 353-398.
4. Aside from work to domestic buildings the only exceptions are £76. 11 s. spent on munitions and “repairs to the walls” in 1195-6: Pipe Rolls 1195. in: ed. H.M. STENTON, Pipe Roll Society New Series 7 (1930), 281.; £18 for repairs “to the doors of the castle” in 1200-1 : Pipe Rolls 14 (1936). 284.; and record of mining activity (see note 38 below). For fuller printed details of expenditure see RENN 1969, 90-92.
5. Over £725 and £ 1082 of documented expenditure respectively. BROWN 1955, 393-4.
6. See for example KING 1988, p. 71 and 91-2. But note too that some aspects of this design owe as much to political symbolism as to advances in military science. See GOODALL 1999a.
7. See BROWN/COLVIN 1963, 633-8.
8. For discussion of the wider role of castles see BEELER 1956, p. 581-601 and EALES 1988, p. 23-43.
9. KENDALL 1923, p. 37-48.
10. The events described below are outlined in many histories. Two readable and authoritative accounts are CARPENTER 1990, p. 5-49 and WARREN 1978, p. 241-59.
11. For de Burgh’s early career see WEISS 1974, p. 235-52. or ELLIS 1952.
12. MICHEL 1840, p. 170. 13. LUARD 1874-1876, p. 28.
14. From his version of events it would seem the author was attached to Louis’ army. Meyer, in his introduction to the transcription of the History, suggests that the author was a Fleming: MICHEL, 1840, ii-iii. The composition has loosely been connected with Marshall’s family; see GRANSDEN 1996, p. 518. In my work on the Histoire I am much indebted to a translation by Eric Poole.
15. MICHEL 1840, p. 177.
16. “Devant cele barbacane venoient souvent chil dou castiel tout arme, si que chil de Post les veoient plainement.” MICHEL 1840, p. 178.
17. “…il avoient une barbacane defors la porte…qui estoit close de moult boin roulleis de caisne, et si of boin fosse tout entour.” MICHEL 1840, p. 177-78.
18. “…si les aprocha tant qu’il li coururant sus, et il les atendi ; si I remest pris, car mauvaisement fu secourus.” MICHEL 1840, p. 178.
19. Ralph of Coggeshall, in: STEVENSON 1875, p. 182, says the siege began around the feast of St Mary Magdalene (22nd July).
20. It has tentatively been suggested that a circular earthwork enclosure recorded in the eighteenth century to the north of the castle was the remains of Louis’ camp. PARFFIT 1995, p. 10-11.
21. “Et tost apries chou s’en ala Looys el mont o toute s’ost, si assist le castiel: une partie de ses gens fist demourer en la ville per cels dedens del tout avironner, et en la mer refurent ses ties ; et ainsi furent chil del castiel de toutes pars enclos.” MICHEL 1840, p. 178.
22. Continuing from note 21 : “Lors fist Looys drecier ses perieres et ses mangounais pour jeter a la porte et au mur; si fist j. castel de cloies moult haut et un cat por mener au mur; ses mineours fist entrer el fosse, qui minerent la piere et la tierre desous le roilleis. Puis les fist assailfr as chevaliers de Post; si fu tantost la barbacane prise.” MICHEL 1840, p. 178.
23. MICHEL 1840, p. 178.
24. “Puis mist Looys ses mineours a la porte ; si minerent tant que une des tours cai, dont deus I avoit.” MICHEL 1840, p. 179.
25. Continuing from note 24: “Lors entra une gratis partie des gens Looys on castiel ; mais chil dedens les remisent hors par grant vigour, et puis refrement la endroit a for murs estoit chaus, de grans mairiens et de grans baus travesains et de grant roilleis de caisne.” MICHEL 1840, p. 179.
26. MICHEL 1840. The date is given in STEVENSON 1875, p. 182.
27. Matthew Paris gives account of the parley in Chronica Majora. LUARD 1874–6, Vol. 3. 4-5.
28. MICHEL 1840, p. 189.
29. “…[Louis] se loga sour le mont devant le castiel; si fist drechier for trebouket, qui asses for fist poi de mal. Lors commencierent moult durement a faire maisons par tout.” MICHEL 1840, p. 192.
30. MICHEL 1840, p. 188.
31. Chronica Majora, LUARD 1874-6, Vol. 2, p. 664. Paris actually gives no account of the second siege and mentions Malvoisin in what appears to be a narrative of the first. His version of events however may well conflated their details.
32. “Lors fist abatre son trebucet et s’en apparella d’aler” MICHEL 1840, p. 195-6.
33. For an account of the battle see CANNON 1912, p. 649-70.
34. For an analysis of the later defences of Dover see CORD/LEWIS 1982, p. 141-200.
35. I am particularly grateful to Kevin Booth for his many suggestions and help in the analysis of this building.
36. The bulk of this documentation is printed in COLVIN 1971, p. 20-87.
37. HASSALL 1976, p. 233-35 and p. 243-55.
38. In July 1205 to work in the “boves” at Dover: HARDY 1833, 42a (bis).
39. See MESQUI 1991, p. 240-48 for a discussion of these features. Works at Domfront and Chinon deserve particular comparison with those at Dover.
40. The present moat must have been dug to its present depth at least by Henry III’s reign, and possibly before, because today there are medieval doorways opening from St John’s Tower (built shortly after the siege) onto the bottom of the ditch.
41. Including at least one new work: in 1208, 100 marks “for beginning the works at our castle of Dover” HARDY 1833,106b. Could this be connected with the gatehouse?
42. See GOODALL 2000. My thanks to Jonathan Clark in this.
43. Notably in the arrangement of its gateway opening which may be reconstructed from the physical evidence as having, like Dover’s, a segmental arch set beneath a two-centred arch. The dating of this building is a matter of conjecture but it evidently predates the late thirteenth-century works at the castle. See GOODALL 1999, p. 6-7.
44. For a discussion of these Irish buildings see McNEILL 1997, p. 24, 45-49 and p. 53-4.
45. The form of the Dover gatehouse and the northern defences as a whole add to the conclusions of recent discussions of developments in fortification during this period, notably CURNOW 1979, p. 42-69; RENN 1975, p. 172-182; KENYON 1990, p. 63-73; and KNIGHT 1987, p. 75-88.