The fortress has the distinction of being the tallest tower keep in England (125') and consists of the keep, set within a curtain walled bailey, with an outer ditch.  All is set partially within the city defences which lie to the east.  The fortifications dominate the point where the Roman Watling Street - originally the main road between Canterbury and the Tower of London - crosses the River Medway.  Once more occupation of the castle site began in the Roman period, showing again that strategic concerns rarely changed across the centuries.  The western curtain wall of the bailey overlies the remains of the earlier Roman wall, making it likely that the area of the castle was once within the Roman town of Durobreve.

The earliest references to the castle are in the Domesday Book of 1086.  Here it is stated that the bishop of Rochester had been given land elsewhere  'in exchange for the land on which the castle stands'.  This shows that a castle existed in 1086.  The first fortification of the site in stone is generally accredited to Bishop Gundulf after the siege of 1088, although there is not a shred of solid evidence to back up this claim.  What the evidence, taken from a charter made between 1087 and 1089 and later copied into the Book of Rochester, does state is that the bishop agreed to build a stone castle as he was good at such things.

episcopus Gundulfus, quia in opere caementarii plurimum sciens et efficax erat, castrum sibi Hrosense lapideum de suo construeret.
Because Bishop Gundulf knew much of masonry work and was effective at it, his men would construct the castle of Rochester for him in stone. 

This wording does not imply that an earlier castle, not built of stone, originally occupied the site.  Indeed it proves nothing other than unspecified stonework was planned at Rochester in the first two years of the reign of Rufus.  Elsewhere in the charter it states that Gundulf built this ‘castle' at a cost of £60 as against the expected £40.  As the cost of the castle could not have been known before it was built, it is obvious that the charter itself was made in hindsight and is not a reliable source for the building of the castle bailey.  Further, it is a pity that we cannot compare what he built at this cost with work elsewhere, although 'Gundulf's Tower' at Rochester cathedral is often attributed to his work. 

After Easter 1088 the castle was seized by Bishop Odo of Bayeux and subsequently the city was besieged by King William II and eventually forced to surrender.  The idea that the keep was not built until long after this in 1127, is deduced from a notification of Henry I.  The text appears to state that the archbishop could build a castle, fortress or even a garrisoned site or indeed a tower if he so wished.  Such a structure could well be one of the mural towers and not the keep which may already have been standing.  However, there is no evidence that any such structure was ever built.  As we do know that various towers were built in other castles for the habitation of the knights who were to defend their fortresses, this is hardly secure evidence that the great tower was only built after 1127.  Despite this, writing in the late thirteenth century, ie. two hundred years after the alleged building of the tower, Gervase of Canterbury noted that Rochester castle had been confirmed to Archbishop William of Canterbury (1123-36), who 'built a splendid tower there'.  Similarly it cannot be denied that the tower is highly ornate and befitting for the pretensions of a the greatest cleric in England.  Gervase also recorded in the September of 1141 that Earl Robert of Gloucester was held prisoner in turre Roffensi by William Ypres for King Stephen.  Again, this late source implies the keep was standing by this time.

The city walls to the north and east of the castle are said to be thirteenth or fourteenth century mainly because of the surviving bastions, one to the south-east and one to the north-east.  The latter is two storeys high and has three lancet loops in the upper stage.  Much of the wall incorporates Roman remains, which makes the traditional dating difficult.  Certainly the city wall was repaired in 1193 and murage (a tax for building urban defences) was granted in 1225 and 1262.  This strongly suggests a twelfth century masonry wall existed.

Various works were recorded on the castle in 1167 and 1171.  Further, the castle itself was strengthened during the reign of Richard I (1189-1199).  During the siege of 1215, the curtain wall and keep were undermined by King John's engineers in a well known siege which was made into the modern film Ironclad.  Subsequently, repairs were made to the keep and the curtain wall.  The turret on the south-east angle was rebuilt probably in the 1220s to a circular plan against the older turret's rectangular form.  In 1237 mention is made of a southern gateway to the castle wall and the construction of a drawbridge for it.  This has now disappeared above ground.  In 1264 the castle was besieged by earls Simon Montfort and Gilbert Clare.  Earl William Warrene and Roger Leybourne held the fortress for the king.  The barons breached the city wall and the outer defences of the castle, but the great tower held out and the attackers were eventually forced to withdraw.  Little effort was made to repair the damage caused by this onslaught, and in the 1340 survey made for Edward III, it was reported that there were 'dilapidations over the whole extent of the castle'.  It was only in 1367 that a programme of rebuilding was begun, mainly due to French threats to the south coast.  By 1370 the programme was complete, although, between 1378 and 1383, a new tower was built on the north angle of the curtain wall.

The main feature of the castle is now the 6 storeyed stone keep, which was certainly in existence by 1138.  This is one of the largest keeps in England, being some 70' square with walls up to 12' thick and 113' high with the corner turrets taking it to 125'.  The main entrance into the keep was protected by a portcullis set in the doorway and operated from the forebuilding chapel above.  Throughout the tower decoration was sparingly applied and can be seen on the exterior of the principal doorways and upper floor embrasures.  Internally the decorations are mainly chevrons with some shafting, while the central arcade has scalloped capitals.  There are also early fireplaces, one being similar to that found in the keep at Portchester.  The work somewhat resembles that seen in the cathedral just outside the castle defences.

To the north of the keep is the irregular bailey, being some 400 feet north to south.  This is still partly defined by the curtain wall which once enclosed it.  That to the west is probably the oldest and is clearly built on top of the remains of the Roman city wall.  Above this a section of wall in herringbone masonry, topped by a wall with blocked Romanesque windows.  The dating of the intermediate wall is therefore of great interest.

A long section of wall to the east still contains two rectangular towers of rectangular design.  The southern tower appears to be totally the work of Edward III in the late 1360s, while the northern one was built simultaneously on the foundations of an earlier structure.   This probably explains their different designs.   In the south-east corner is a rounded tower which probably belongs to the thirteenth century.
 This is further defended by ground floor crossbow loops.  Such loops also appear at possibly Adare and certainly Dunamase in Ireland as well as at Beaumaris, Flint, Grosmont and Rhuddlan in Wales and in the early work at Goodrich and the supposed 1220s work at Aguilar in France.    Beyond the curtain wall on the landward side, but only now visible to the east and north of the castle itself, are the remains of the castle ditch. 

For more detailed descriptions of the castle see Kent Archaeology, the Archaeological Journal and Archaeologia Cantiana:

Why not join me at Rochester and other British castles this October?  Please see the information on tours at Scholarly Sojourns.


Copyright©2016 Paul Martin Remfry