ROCHESTER CASTLE.

ROCHESTER—Fortress, Cathedral, and City—is a very remarkable place—in some respects the most remarkable place in the South of England. In each of its triple capacities it claims a high antiquity. Its ecclesiastical history commences with Augustin and Æthelbyrht, the founders of its see, over which Justus, the friend of Augustin, was the first to preside, and to the endowments of which a long succession of Kentish and Mercian princes contributed. Its secular history, though often obscure, ascends to a yet more remote period, and its material evidences are still to be read in the form of works either in earth or masonry, showing Roman, Saxon, Danish, and Norman occupations.

Rochester is built upon a low cape or promontory formed by a grand bend of the Medway, the central part of which is occupied by the city, and the most prominent and northern point by the castle. The great Roman way from Canterbury to London traversed the length of the cape, and crossed the river near its apex, in the line of the ancient and nearly of the present bridge.

The position of the fortress is very noble. Rising from a base of hard ground high above the stream, it forms a grand feature upon the shore of the Medway, here 1,000 feet in breadth, and which almost vies with the Thames in the volume of its waters. From the turrets of the keep is seen outspread a view of great extent and exceeding variety, including the steep and verdant slopes of the chalk, and the broad and fertile meads across which the great river flows in graceful folds, concealed and betrayed by the inequalities of the hills and the scattered masses of forest trees. The cathedral, indeed, is scarcely worthy of these advantages. Far from asserting, as at Durham, Canterbury, or York, its equality with, or even superiority over, its temporal neighbour, the cathedral of Rochester, venerable as it is, and honoured as are the ashes that lie within its precincts, is of small dimensions, and in position low. The castle, on the other hand, is a masterpiece of Norman skill, and the first and last object that strikes the eye of the visitor.

The castle of Rochester occupies the northern termination of a chalk ridge which forms the right bank of the Medway for many miles above the city, and intervenes between the river and the cathedral, upon the margin of the one and but a short distance from the other. It thus covers about half of the western front of the city, and extends into what was the south-western angle of its walled enclosure.

Though of ancient date, the castle proper, that is the structure in masonry, has evidently been preceded by the works in earth, in close connection with a part of which it has been laid out. These are on a large scale, and, though much degraded by time and obscured by buildings and enclosures, are still tolerably apparent. They seem to have been composed of an oblong space included within a ditch, which commenced near the bridge foot, and was carried eastwards for about 130 yards, when it turned to the south, and ran for about 270 yards roughly parallel to the river, towards which it was again returned. This oblong area was subdivided into two unequal parts, the southern being the smaller, by a cross ditch, and the latter part was occupied by a large flat-topped conical mound, known as Boley Hill. The northern part contains the castle. Along the east or cathedral side, this ditch is in part a bold natural depression. Along the west side it is superseded by the river, here very broad, deep, and rapid. The area thus included is about 7¼ acres. The mound is of large size, though reduced by modern operations. It is in part natural, in part formed from the adjacent ditches. Like the mounds at York, Wallingford, and Wareham, and those formerly existing at Hereford, Buckingham, Worcester, Hertford, Nottingham, and Stamford, it is so placed as to watch the river, which must always have been an open and dangerous highway.

These lines of earthwork, though the principal and most evident now to be traced, are by no means the earliest among the defences of Rochester. Though the discovered Roman remains are not indicative of a large town, the Watling Street here crossed the Medway, and there was certainly a fortified post on the site of the mediæval city. Its rectangular form, about one-third longer than its breadth, the cruciform arrangement of its main streets, its four gates opposite to each other, and the termination of its name, all point to a Roman origin, and the Roman defences are thought to have been composed of a wall and ditch, represented by the line of the later wall, only that they are supposed to have passed from the south gate towards the river, about 45 yards west of the site of present keep.

The area measured about 470 yards by 160 yards, and therefore covered about 15½ acres, and of the gates Childegate stood to the north, and to the south was a gate removed only in 1770, the road between them passing close to the west end of the cathedral. Roman coins and sepulchres have been found in the castle area, and that Rochester was the Roman Durobrivis seems generally admitted.

The Roman lines were in part adopted by the later inhabitants, and fortified in the English manner. The position of the new city laid it open to frequent and severe attacks by both land and water, and its works suffered much from the inroads of the Danes, especially towards the close of the ninth century, and to that people, though perhaps on insufficient grounds, has been attributed the mound.

The earthwork fell into the hands of the Normans at the Conquest, and it became necessary to fortify the position according to the improved manner. The usual custom with the Normans, when dealing with an English stronghold, was to erect a shell keep upon the mound, or, where there was no mound, a rectangular tower, and to surround it with a strong curtain wall built upon the bank within the scarp of the existing ditch. Probably, in this case, the older area would have required too large a garrison for its defence, and the mound was at a rather inconvenient distance from the bridge or its preceding ferry. The mound, therefore, was excluded from the new enceinte, as it was at first from the area of the city. Probably it was occupied as an outwork, or it may have been reduced to a rather lower level. The new curtain was built on the scarp of the ditch of the northern division of the earthwork on the three landward sides. On the fourth it was placed upon the crest of a low cliff, about 40 feet above the river, and rather less above the quay, which has been widened in modern times, and in part added to the shore.

The Norman enceinte, the general substance of which is probably of the eleventh century, is roughly four-sided, about 130 yards east and west by 160 north and south, and includes about 4¼ acres. The eastern front is opposite to, and about 60 yards from, the cathedral. The south front faces the old mound of Boley Hill, and the north front, of rather greater length and set obliquely, points towards the bridge. It is probable that the ditch outside the south wall represents that of the Roman work. The gatehouse stood at the north-east angle, and the steep causeway leading up to it, and which has superseded the drawbridge, is known as the “Vennell.” The gatehouse is gone. At the north-west angle are the remains of a bastion tower, standing in 1735, and containing a postern. In it was a shaft for lifting stores from the river, an arrangement seen in another form above the Wye at Chepstow. This tower commanded the bridge. The south-east angle is still capped by a bold drum tower, of 30 feet diameter and three-quarters projection. It has a ground and upper floor, each pierced by an arcade with loops in narrow lancet recesses. The floor was of timber. The gorge is broken down, but may have been always open. This tower does not rise above the curtain, and seems to be an insertion in the early English period. Upon the east wall are two rectangular towers, one of 35 feet breadth and 25 feet length, wholly of exterior projection, and the other 30 feet by 25 feet, and partly within the wall. These seem insertions of the Perpendicular period. The curtain is much broken down, and part altogether removed; 40 feet of it remain next the south-east angle, on the south front. This portion is 7 feet thick, 30 feet high, and seems decided Norman. On its outer face are traces of two flat pilasters. It is pierced by three loops with splayed round-headed recesses, 3 feet broad and 8 feet from the ground, inside. There must, therefore, have been a platform against the wall. From the same angle, along the east face, the curtain is tolerably perfect all along. Near the rectangular tower it is seen to be built upon arches, intended to be covered up, as in the walls of York and of Southampton Castle. The top of one arch is seen inside the wall, and the whole of two arches outside. A long piece of the north wall remains, and is a good example of early masonry, probably Norman. It is very thick, and built of layers of rough undressed stone laid in very thick beds of mortar. As at Cardiff, there is no foundation, the wall being laid on the surface of the ground, as was not uncommon with very, thick walls. About 260 feet of the river front remains. At its south end was formerly to be seen some herring-bone work, and in other parts of it is shown a tendency to this mode of building. It is about 40 feet high. Upon its inner face are seen the outlines of two pointed arches, as though two vaults had abutted on the wall, only the arch stones do not project for toothing. There was an upper floor, and in the wall, but outside, are seen the outlines of two rather large round-headed arches. It is in this wall that Roman masonry has been suspected. Much of it seems late Norman. There was a cross wall, now removed, which ran close north of the keep—and divided the Norman castle into a north and south ward.

How the castle was connected with the defences of the city does not appear. The city wall is said to have been built about 1225, on the Roman site, as far as the south gate, from which it seems to have left the Roman line, and to have included the subsequent earthwork of Boley Hill. When this was done the old ditch was widened in some places to 85 feet and made deeper. Part of this wall is said to stand upon arches, like that of the castle; if so, there was probably an earlier wall, of Norman date. The city wall is so blocked in, that little is visible to an ordinary visitor. There is thought to be a trace of Roman masonry in the east wall, and there is a small door of Henry III.’s time, near the east end of the cathedral, where the wall is 4 feet thick and 30 feet high, and a tower of that date at the north-east angle, said originally to have been upon the town wall. This tower is a very fine work, 30 feet diameter, outside a cylinder and within an octagon, with three loops, and a door in the gorge, from which pass two mural galleries, one leading to a well-stair and the other to a garderobe. The upper floor is at the level of the top of the curtain. The line of the south side of the city wall was altered by the monks in 1290, by permission from the king. They were to rebuild it 5 perches and 5 feet to the south or outside of the old line, and it was to be 16 feet high and embattled. Moreover, the old ditch was to be filled up to the length of 54 perches 14½ feet, and in width 5 perches and 5 feet. As this involved prolonging the east wall, they turned an arch over the old ditch. This expansion of the old area lay to the east of the south gate of the city, a little east of which was the gate of the priory. The old wall is thought to have included Boley Hill.

The Keep, for which Rochester Castle is justly celebrated, is placed within, and very near to the south-east angle of the castle area, and upon the highest ground within it, about 60 feet above the river: a point probably specially selected because it is opposed to Boley Hill. It does not, like Porchester, form a part of the enceinte, but, like Carlisle, stands just within it, being from 16 feet to 33 feet from the eastern, and from 10 feet to 15 feet from the southern wall. Though not one of the largest, Rochester is one of the loftiest Norman Keeps in England, and one of the most worthy of note for its history, position, the boldness of its decorations, and its substantially perfect condition. Of additions, with the exception of the early rebuilding of its south-east angle, it shows no trace, and its dilapidations are not greater than may be attributed to seven centuries of age, of which the three latter have been periods of neglect.

The keep is a square of 70 feet at its base, and the walls are 12 feet thick, reduced at the summit to 10 feet, so that its exterior dimensions are there 66 feet. This reduction is made by a very slight exterior batter, not perceptible to the eye. There is no set-off or string-course, outside, and no regular shelf for the floors inside. This uncommon thickness is intended to allow, as at Dover and Hedingham, of an unusual number of mural chambers and galleries. The height is 113 feet from the ground to the coping of the parapet, and the turrets rise 12 feet higher. The two western angles are flanked by broad flat pilasters, with hollow angles or nooks of 8 inches. They rise from a common plinth, and each is carried up clear of the parapet to form the outer face of a square turret, the height of which adds much to the commanding aspect of the building. The north-east angle is capped in a similar way, save that the pilasters are rather bolder, have no plinth, and the angles are solid until clear of the forebuilding. The south-east angle has neither plinth nor pilaster, but is capped by a rounded projection, which, though early, is not original. This also terminates in a turret, which is rounded towards the field, but flat on its two inner faces. On each face of the keep, near its centre, is a single pilaster of moderate width and projection. These also rise from the plinth and ascend to the parapet, two of them actually to its coping. On the north face is a second narrower pilaster, formed by continuing the wall of the forebuilding upward.

The keep has many openings, mostly of small size. A tier of round-headed loops marks the basement, and of square-headed ones, a trifle larger, the first floor. Next a line of small round-headed windows shows the level of the second or main floor, and one of larger windows the gallery of the same. Above all, a line of rather larger windows, also round-headed, indicates the uppermost floor. The two upper lines were ornamented, and had Norman half-piers and mouldings, and in some of the openings the windows were coupled, but the ashlar dressing has been much broken away. There are besides several loops placed near the angles, indicating garderobes, mural chambers, and staircases, and each turret has a window on each of its outer faces.

Internally, the keep is divided into two nearly equal chambers by a cross wall, east and west, 5 feet 6 inches thick throughout, and rising to the roof, where it carried the main gutter. Thus each floor was composed of a north and south chamber, one 21 feet and one 20 feet broad, and each 46 feet long. In the cross wall, near its centre, thickened by a pilaster strip on each face, is the well, said to be 60 feet deep from the ground level, and to contain usually 10 feet of water. It has an ashlar pipe, 2 feet 9 inches diameter, which seems to have been carried up to the roof, and was accessible by four small round-headed doors, one on each floor. It is unusual to see the well-pipe carried higher than the first floor: at Dover it reaches the second. To the north face of the Keep is appended the forebuilding covering the main entrance, which, as usual, was at the first-floor level.

The basement is reached only by a well-stair in the north-east angle, which stair, as in the Tower of London, ascends from the base to the summit. This staircase is 11 feet diameter, and contains 133 steps. It communicates with each floor, with several galleries and mural chambers, with the chapel, and with the roof of the forebuilding. The basement is at the average level of the ground outside, and there is no underground chamber.

The north chamber is entered at the north-east angle by the well-stair from above, and aired rather than lighted by three round-headed loops placed in its east, west, and north sides. These are of 4 inch opening, set in splayed recesses, and these again in flat-sided recesses, 6 feet broad and 3 feet deep, all round-headed, and the larger descending to the floor. The north loop, which opened into the bridge-pit of the main entrance, has been converted into a rude doorway, but the head of the loop remains. In this wall are two narrow doors, of which one descends 8 feet by nine steps into the basement of the forebuilding, and the other ascends by a slight slope into the first floor of the same. Besides these, near the west end of this wall, is a recess, 6 feet wide and 6 feet 9 inches deep. This recess, the depth of which seriously weakens the wall, can here produce no bad effect, the wall having been supported outside by the solid steps and lower gate of the forebuilding. In the west wall, besides the loop, is a broken doorway opening into a mural chamber, 4 feet 6 inches by 10 feet, which occupies the north-west angle of the building. In the cross wall, besides the opening to the well, are two doorways of 4 feet 7 inches and 5 feet opening.

The south chamber has single loops to the east and west, and two in the south wall in recesses, all of the pattern already described. There is also a mural chamber 3 feet 10 inches by 10 feet, in the south-west angle, entered by a door in the south wall, and near the east end of the same wall is a recess, 6 feet broad and 3 feet deep, but without a loop. The masonry shows this angle to have been rebuilt. Besides these cavities there are two, one in the west wall, 2 feet 10 inches by 3 feet 7 inches, entered by a rugged opening, 2 feet 7 inches broad, and the other in the south wall, 2 feet 10 inches by 4 feet 3 inches, with an opening much broken away. These cavities are shafts, which ascend vertically in the wall, and above are divided, showing that they are the vents of the garderobes in the first and second floors. The openings into them are not original, and probably the shafts were sunk a few feet below the floor, and ended either in cesspools or a drain. As there are other garderobes in the upper floors, there are, no doubt, other shafts, not broken into. There must be at least two, one at each end of the cross wall in the north chamber. It is curious that these vents, which in Norman buildings often open, as at Ludlow, upon the outer face of the wall, should have vertical shafts, while the turrets of the fire-flues vent in the face of the wall. At Coningsborough the arrangement is the reverse; the flues are vertical, the garderobe vents horizontal. The basement was evidently a store-room. It was about 15 feet high, and from it 20 steps led to the first floor.

Wyman & Sons, Gᵗ. Queen Sᵗ. London.

KEEP OF ROCHESTER CASTLE.

The first floor, also of two chambers, was lofty, gloomy, and perfectly plain. It was probably the soldiers’ lodging during a siege.

The North Chamber. In the north wall, near the east end, is the main entrance, 15 feet from the ground, and opening from the vestibule of the forebuilding. This is a bold, full-centred arch, of 6 feet 2 inches opening, with flat sides, in a wall here 10 feet 6 inches thick. The outer face is flanked by nook shafts, and the architrave has a bold chevron moulding. The inner architrave has a similar but rather plainer moulding, but the jambs are broken away. Just within the outer face is a square portcullis groove, and behind it a rebate for a door, expanding the passage to 7 feet. Near the inner face, in each side, is a small round-headed recess, 2 feet 6 inches broad by 1 foot 4 inches deep, probably to hold the keys or a lamp. The door was secured by a stout bar, and the grate, which seems to have been of iron, was worked from a mural passage above. This portal led direct into the north chamber, about 7 feet from the well-stair door. In the north-west angle is a mural chamber, called “Gundulph’s room,” vaulted and groined, 10 feet 3 inches by 11 feet 7 inches, entered by a door in the west wall, and lighted by two loops to the west. In its rounded north-west corner is a fireplace with a small Norman hood of 2 feet 6 inches projection, having a segmental arch. The tunnel is conical, and ends in a couple of openings in the hollow angle of the south end of the adjacent pilaster, in the face of the wall, above 12 feet above the hearth. Near the north-east angle of this chamber is a small blocked-up doorway, which, as at Middleham, opened upon the roof of the outer gate of the forebuilding. The ashlar rings of the loop recesses, and the hood of the fireplace, are of excellent workmanship. Besides these openings there are in the north chamber four windows, one at each end, and two in the north side. These are set each in a recess, 5 feet 6 inches opening, and 9 feet 6 inches deep, 15 feet high, and round-headed, the loops being square-headed, about 1 foot 6 inches high by 12 inches broad. These recesses, from their great height, produce a good effect, but were probably found to weaken the wall, for they have been partly built up and reduced at their lower 6 feet to a depth of 3 feet. The blocking is of inferior masonry, executed in haste. Between the two northern loops is a fireplace of 5 feet 6 inches opening, round-headed, having no hood, and with a round back. It is strengthened by a pilaster of 9 inches projection and 10 feet breadth, from the inner wall. The tunnel from the fireplace is conical, and has a lateral vent about 15 feet up. In the east wall, near the entrance from the well-stair, is a recess, which, besides the loop, contains a postern and a garderobe. The latter is a mural chamber, 5 feet 10 inches by 13 feet, which opens from the southern jamb of the recess: the postern, 2 feet 6 inches broad, is cut obliquely through the wall, and opens in the face of the keep, 15 feet above the ground, and about 30 feet from the rampart of the curtain, with which it must have communicated by a light plank bridge. There is such a door in the Norman keep at Adare, and in the later Wakefield tower in London, to which a stone bridge has recently been fitted. In the cross wall, besides the well opening, are two doors corresponding to those below.

The south chamber has a loop at each end, and three in the south wall, all in recesses similar to those in the north chamber, and these also have been partially blocked up to strengthen the wall. There is a fireplace in the south wall corresponding to that described above, and in the south-west angle a door leads into a well-stair, 11 feet diameter, which commences at this level, and communicates with the floors above, and with the battlements. There are four mural recesses in this chamber. One, a garderobe, opens in the south jamb of the recess of the west window, and another in the east jamb of the central window in the south wall. The shafts from these recesses descend into the wall of the lower chamber, as described. The other galleries open right and left from the window recess in the east end. That to the north is a passage, 3 feet 6 inches by 8 feet, with a garderobe in the end of it. That to the south is a gallery, 20 feet long by 3 feet broad, which ends in the south-east angle, where it has two windows to the south and east. This is in the rebuilt angle. This floor rested upon timber joists. From it forty steps lead up the north-east staircase 30 feet to the main floor, and at thirty-six steps branched off a narrow stair to the chancel of the chapel.

The main, state, or second floor, 32 feet high, rested on each side upon eighteen joists, whole timbers. It contains two tiers of windows, the lower opening directly from the chamber, the upper with the intervention of a mural gallery. Between the two is a plain string of Norman pattern. The roof was flat, being the joists and planks of the upper floor. The window recesses of the lower tier range from 3 feet 2 inches to 5 feet 10 inches opening, and those of the gallery tier nearly correspond to them. All are flanked with nook shafts, but the lower architraves have a bold roll and hollow moulding, those above have the chevron pattern. The ashlar of the windows is mostly gone. The lower seem to have been plain and small, the upper richer and rather larger, but none at all large. The projections in front of the well, the fireplaces, and the two staircases, are beaded at the angles. There are no marks of corbels for struts to stiffen the joists and give support to the ceiling.

The north chamber was entered by a sort of lobby, 5 feet 6 inches square, and groined, into which the staircase opened. The recesses of the east and west windows each have mural garderobes opening from the southern jambs. The two recesses in the north wall have no side openings. Between them is the fireplace, similar to that below, but with a handsome chevron architrave. Over the great entrance is a doorway of 7 feet opening, leading to the chapel. In it is the chase for working the grate, and beyond this is a rebate reducing the passage to 3 feet to form the chapel door. In the north-west angle is a small chamber.

Wyman & Sons, Gᵗ. Queen Sᵗ. London.

KEEP OF ROCHESTER CASTLE.

Arch Moulding.
String-course, second floor.
Window Moulding, second floor.
Window Moulding, third floor.
Fireplace Arch, second floor.
Fireplace Arch, fourth floor.

The south chamber has three windows in the south wall, a fireplace, and a door opening into the south-west well-stair. In each end is a loop. In the recess of that to the west, on its north side, is a passage to a garderobe, and there is probably a corresponding place in the opposite, or east window, but the window recesses are inaccessible.

The cross wall, instead of two doors, is pierced by four arches of unequal size, two on each side of the well pier, which has its small and rather ornate door. Between each pair of arches is a heavy Norman pier, a plain cylinder with fluted capital, and opposite to which are two half-piers, or responds. The arch has been ribbed and worked in bold roll and hollow, varied with the chevron pattern, but the central rib has fallen away. The capitals of this arcade range with the general string-course, and the arches rise into the upper division of the apartment. There is rather more ornament on the north than on the south face. The arcade was filled up with a stone screen, about 10 feet high, of which one division remains and a part of its doorway. The screen is original, and has been rather richly worked, but it had no bond into the piers, being merely built against them.

From this floor twenty steps led up 17 feet to the gallery. This threads the whole wall, passing all round the building. It is from 3 feet to 3 feet 6 inches broad, barrel-vaulted, from 8 feet to 9 feet high, and is placed 4 feet from the inner face of the wall, and 5 feet from the outer. It is generally level, save that it rises at the south-west angle to accommodate itself to the staircase, and in the centre of the east front to clear the vaults of the garderobes below, and possibly to meet the thrust of the arch in the cross wall.

In the north division there are in the west end three recesses, 3 feet 7 inches, 5 feet 10 inches, and 4 feet 1 inch, which open into the gallery, and correspond to three windows in its outer side. In the north wall are also three windows with recesses corresponding, each of 5 feet 10 inches in width. In the east end is only one recess of 5 feet 9 inches opening. At this point the gallery is rather above 3 feet wide, and the corresponding window has a recess, 5 feet by 4 feet. Most of these expansions of the gallery opposite to the windows are groined, as in the corresponding gallery in the Tower of London.

The arrangements of the south division are very similar. In the east end is one recess of 5 feet, and in the west three of 3 feet 2 inches, 5 feet, and 3 feet 7 inches, the central being the largest. That at the south end represents a short window, intended to conceal the gallery, which here rises to the staircase. In the south wall are three recesses, with windows corresponding. There is no chamber in the south-east angle, the gallery merely turns at a right angle, but in the north-west turret is a groined chamber, 6 feet square, with which the gallery communicates, and which has windows in its two outer faces. The north-east and south-west angles are occupied by the staircases. It is evident from the character of the masonry that with the south-east angle have been rebuilt the two adjacent windows in the south face and of that in the east face. The new work is inferior to the older, and there is no attempt at ornament. The arrangements of this grand floor, like those of the corresponding rooms called the Council Chamber in the Tower of London, must have been inconvenient in the extreme. The rooms must have been very cold and very public, with windows, doors, and garderobes on every side, and a gallery in which a whole household might have listened, unseen, to what went on below. There is nothing to distinguish between the uses to which these two noble chambers were applied. Probably the northern, having the more ornate entrance, was the hall, and the southern a room of a more private character.

From the gallery level twenty-three steps lead up 17 feet to the upper floor, and on the way a small passage opens upon the battlements of the forebuilding. This floor contains two very cheerful and handsome rooms, 25 feet high, with larger windows than the floors below, and a finer view. Here, also, the window recesses and fireplaces are ornamented, and we may suppose this floor to contain the private apartments. The roofs were open and low pitched, and seem to have been supported by rafters placed near together, and without principals. The arrangement of the mural galleries here is peculiar. They do not run round the rooms, but are short, connecting two, or at most three window recesses. Where the wall has been rebuilt they are altogether omitted. They do not pass behind the fireplaces, the vents from which pass off horizontally, and so block the way.

The north chamber is entered, like the rest, from the north-east angle, and from the staircase passes a gallery which pierces the east wall for 20 feet, and traverses the recesses of two windows which light the east end of the room. In the west end are also two windows, in like manner pierced by a short gallery, entered only from the chamber. In the north wall, besides the fireplace, are four windows, two on each side of it, each pair connected by a short gallery, also entered from the chamber. In the cross wall are two doorways, and near the centre the small door of the well pipe.

The arrangements of the south chamber are somewhat different. In the west end are three window recesses, having a mural gallery in common, but not communicating with the staircase. In the east end is but one window, the jambs of which are solid. In the south wall is the door from the south-eastern staircase and also three windows, with a fireplace. There are no galleries in this wall. This chamber has suffered much from the destruction of the south-east angle, which it is evident carried with it about 28 feet of the south and 15 feet of the east wall. In the untouched part of the east wall remains embedded the half pier, capital, and half the architrave of a large original late Norman arch, of about two-thirds the span of the room, more or less ornamented, and evidently the recess for a large east window. It is just possible, as this was an archiepiscopal fortress, that this was a chapel. In the south wall, close east of the fireplace, is the jamb and half the arch of a window recess, also ornate, of which the other part has been removed. The restoration has been effected in a very hasty and slovenly manner. The masonry is of inferior quality, and the new recesses are perfectly plain, although the older ones in the same floor are rather highly ornate. Whatever, in designing this chamber, may have been the intention of the builder, it is very certain that the re-builder thought only of making the tower defensible in the cheapest and quickest way.

From this floor thirty steps ascend 25 feet to the battlements. On the way a passage from the north-eastern staircase opens into a gallery in the east wall, 16 feet long and 5 feet wide. It is traversed by a window, the recess of which, 4 feet wide, opens on the gable of the north room of the upper floor. The staircase passes up to the base of the north-east turret, in which are two doorways opening on the ramparts of the north and east walls. The rampart walk is 4 feet broad, with a parapet of 2 feet and a rere-wall, now gone, of 3 feet. The walk is carried through the north-west turret, which contains a chamber 9 feet square, and through the south-eastern turret, which also contains a chamber. From hence the passage is continued along the south wall, but is stopped by the wall of the south-west turret, the stair contained in which only opens on the west wall. The turret chambers were about 8 feet high and had flat roofs, reached possibly by wooden steps. Each turret had two entrances, one on each face. The parapets of the curtains between the turrets were 8 feet high, and contained five embrasures on each face. At this level the profile of the double roof may be traced. It was composed of two ridges, one over each chamber, with a central gutter resting upon the cross wall. Both parts were low pitched, and the ridges did not rise above the battlements: the northern portion sprung a little the lowest and had rather the steeper pitch. The southern, springing higher, delivered half its water from eaves 2 feet above the common gutter, whence it was carried away by an enormous stone shoot, a sort of gurgoyle. Another such, taking the contents of the north gutter, crosses the head of the staircase and projects eastward. At the rampart level the walls and turrets are pierced by a row of holes 9 inches or 10 inches square, intended, evidently, for the horizontal beams of a brattice. There are no corbels or cavities, as usual, below, to receive struts, the beams being probably strong enough without them. These holes seem not to be original, and being rough, were probably made in haste when the south-east angle was rebuilt: they are of importance enough to be shown in the tower which figures in the corporation seal. Besides these, in the inner face of the north wall, above the gutter, are two rows of pigeon-holes, probably original, and even now accommodating a few birds.

Affixed to the north front is the forebuilding, composed of a gatehouse, staircase, drawbridge, and vestibule tower. These formed and protected the entrance to the keep. The approach is thought to have commenced on the west front, near the north-west angle, by a flight of steps, which turned the angle, as at Dover, and were continued as a broad staircase against the wall of the building. Upon this, flush with the west front, was a small low tower, about 12 feet square, through which the stairs passed. This portal had a barred door, but no grate, and its roof was reached from Gundulf’s chamber in the first floor of the keep. From this gate the steps rose to the bridge pit, a cavity 15 feet deep by 9 feet opening. This was crossed by a drawbridge, which fell from the gate of the vestibule tower. The stairs were 8 feet broad, and protected by an outer parapet 2 feet thick and 6 feet high, and probably looped. The parapet crossed the bridge pit on an arch, so that persons ascending the stairs or crossing the bridge were protected.

The vestibule tower is 19 feet broad by 36 feet long, and two-thirds of the height of the main building, against the north face of which, at the east end, it is built. Its walls are 6 feet thick; they have no plinth or pilaster steps, and though probably of the age of the keep the work is of an inferior character. There is a basement, first floor, vestibule floor, and chapel. Each floor is entered from the keep. The basement is 14 feet by 23 feet, vaulted and groined, and it has two air-holes near the vault, ascending obliquely through the wall. The floor is on the ground level, here low, and from it a passage, 2 feet 7 inches wide and 12 feet long, leads up nine stairs into the basement of the keep. The present door from the exterior is evidently not original. This chamber was a prison.

The first floor measures 13 feet by 23 feet; it also has two air-holes, rather larger and less oblique than those below. A recess, perhaps a doorway, perhaps a cupboard, has been opened in the east wall and fitted with a Tudor door-case. This chamber had a flat timber ceiling, forming the floor of the vestibule. It was entered by a passage, 2 feet 10 inches broad, slightly ascending to it from the basement of the keep. This also was probably a prison, but of a less severe character.

The second or vestibule floor is of the same dimensions. In its west wall is the outer door of the keep, of 6 feet opening, round-headed, and flanked inside and out with engaged columns, with plain caps, and round the arch a band of chevron moulding. In the rubble work inside, above its head, is seen a pointed arch of relief. The portal was closed by a barred door only. On the right, on entering the vestibule, is the door of the keep. The vestibule has five windows: a small one in the west wall commands the staircase; in the north wall are three windows, each formed of a pair, coupled; in the east wall is one. They are all round-headed, with Norman ornaments: one-third of the east end of this chamber is covered with a barrel vault.

The third or upper floor of the forebuilding is occupied by the Chapel, the dimensions of which correspond nearly to the rooms below. It is divided into two parts, a nave and chancel, separated by a plain, bold, round-headed arch of 12 feet span. The nave is entered from the main floor of the keep by a side door, close to which, in the wall, is a recess, now inaccessible. The chancel has a stone floor, resting on the vault below, and its roof is three sides of an octagon, also vaulted, a sort of apse, but covering a rectangular space. There are two east windows, and a large stone drain, which has led to the notion that this room was the kitchen, and no doubt it would have made a very good one. The nave had an open roof of low pitch, which was replaced by one perfectly flat, but at the level of the original ridge. If the large upper room in the keep was a chapel, a second would scarcely be necessary. Probably both were not in use at the same time. A kitchen is a very uncommon adjunct to a Norman keep, as the inmates seem to have broiled their meat at an open fire, whereas a chapel is not uncommon; and at Middleham, Dover, Newcastle, and probably at Corfe and Kenilworth, it was in the forebuilding.

The material of the keep is chiefly the rag-stone of the country, with ashlar coigns and dressings of Caen stone. The ashlar is rather sparingly used, but is of sound quality and close jointed. The rubble is poor, imperfectly coursed, and held together by the excellence of the mortar, or rather of the lime, for the mortar is carelessly mixed and is full of sea shells, which have been brought with the sand and left uncrushed. It is clear from appearances, both inside and outside the building, that the south-east angle has, at some period soon after its construction, fallen down, and been rebuilt soon afterwards clumsily and in haste. No mere battering by a ram, no strokes from the missiles of a catapult, however ponderous, would have brought down both turret and wall. Such ruin must have been produced by a mine, and this is the more probable since the south-east is the most exposed angle for such an attack, the scarp of the ditch being here steep, the soil soft chalk rock, and the keep but a few feet from the ditch. A mine opened in the scarp, just outside the angle of the curtain, and driven 40 feet in the chalk, would reach the foundations of the turret, and thus undermined, its fall would bring down more or less of the adjacent walls, and falling outwards, would crush the angle of the curtain. This would account for the large mass of the angle which has fallen, and for the circular bastion tower, evidently an insertion, at the angle of the curtain. Supposing the keep, built in 1126, to have been mined and breached in the memorable siege of 1215, and the damage soon after made good, this would account for present appearances. It is true that the bastion tower is decidedly early English, while the restoration of the keep is Norman, but it is of a period when round and pointed arches were both in use, and when the character of the older part of the keep would naturally lead to the employment of the former; so that the restoration might have been effected as late as 1225, ten years after its probable destruction, and when an entry in the Pipe rolls shows that the sheriff of Kent repaired the tower.

There is a good deal in the character of the forebuilding that looks as though it was an addition to the keep; and yet this cannot be, seeing that the difference in its masonry includes that of the adjacent staircase turret, which could scarcely be an addition. Also the remains of the lower gatehouse, on the north face of the keep, are clearly as old as the keep, of which they are part.

It was long the custom to attribute this keep to Gundulph, making it contemporaneous, or nearly so, with the Tower of London, which it resembles in some of its arrangements, especially its mural gallery and fireplace. Attention to its other details shows, however, that this cannot be so, and that its more probable builder was Archbishop Walter Corboil, about 1126. Those who desire to see examples of Gundulph’s work, and to compare his masonry with that of Rochester keep, should examine St. Leonard’s Tower at Malling, some parts of the nave and transept of Malling Abbey Church, and the north tower of Rochester Cathedral. The White Tower in London has been so often repaired and refaced that it is difficult to be sure of its original masonry. Part of the curtain of the enceinte of Rochester Castle may also be Gundulph’s work. The north wall looks very early, as does the east wall, which is of excellent though rudely coursed rubble, the stones being large and the joints broad, though there are no layers of flat stones as at St. Leonard’s.

Rochester much resembles Hedingham, a very pure and very perfect Norman keep, with three floors, the remains of a forebuilding, and the upper gallery in the main floor. In each ornamentation is effectively employed, and great use made of the chevron moulding. The two buildings are, probably, of about the same date.

The history of the castle of Rochester, in the proper sense, begins with the Norman Conquest, but there is something to be said of the preceding earthwork, subsequent to the Roman period. The bishopric, of which the city is the metropolis, was founded in A.D. 600, and Justus, its first bishop, was consecrated in 604. The church, dedicated to St. Andrew, was founded by Æthelbyrht, King of Kent, and his charter of 604 mentions certain land as extending from the west gate and the wall of the city, one boundary being “fram Suthgaete west, and langes wealles.” Also, other land granted is said to be, “extra murum civitatis versus aquilonem.” In 673 a synod was held here, says Bede, attended by Putta, “Episcopus castelli Cantuariorum quod dicitur Hrofescæstir;” so that it was then a strong place, and the bishop’s residence, and this whether “castellum” relates to the city, or, which is at least equally probable, to the adjacent fortress.

In 765, Ecgberht, King of Kent, granted by charter to Eardulf of Rochester, his faithful minister and bishop, “terram intra castelli mœnia supranominati, id est Hrofiscestri;” and in 781, Æthelberht, King of the West Saxons, says, “Concedo Hrofensis ecclesiæ ... aliquantulum terræ juris mei intra mœnia supradictæ civitatis in parte aquiloni”; and in 788, Offa, King of the Mercians, granted “terram ... sex aratrorum” at Trotescliff, “ad ecclesiam beati Andræ apostoli et ad episcopium castelli quod nominatur Hrofescester.” Also, in 789, King Offa, in another charter to the same Bishop Weremund, mentions the wall (murus) of the city, and says, “Ecclesia quæ sita est in castro quod nominatur Hrofesceaster.” Further, in 850, Æthelwulf, King of the West Saxons, mentions “murum civitatis Hroffi;” and in 855 he granted land “Dunne ministro meo ... in meridie castelli Hroffi.”

In 839 and in 885, the city was attacked by the Danes. On the former occasion they took and sacked it, on the latter it was relieved by Alfred. In 986, the city was attacked, but unsuccessfully, by Æthelred, and in 998 with more success by the Danes, who came up the Medway. Boley Hill is said to have been thrown up by the Danes in 884, but the greater part of the hill is natural, though, no doubt, with a very considerable addition; but even the addition would have required more time than was allowed them by Alfred, and a work of such a character would be far more useful for defence than for attack. The towers used in an attack were generally of timber, and so constructed as to admit of being pushed up to the walls. It is more probable that the mount was thrown up as a strong residence for the bishop, to give such personal security as was found necessary at Sherborne.

It is not known in whose hands the fortress of Rochester rested in the times preceding the Conquest, but Earl Godwin had certainly encroached upon the episcopal property; and what more probable than that he should have laid his hands upon a place which, with Dover, would give him the command of the great road from the South to London? We read in Domesday of the episcopal manor of Estockes or Stoke. “Hoc manerium fuit et est de Episcopatu Rofensi, sed Godwinus comes T. R. E. emit illud de duobus hominibus qui eum tenebant de Episcopo et eo ignorante facta est hæc venditio.” So when the term of the tenancy expired the bishop would find that instead of two nameless men ready to give up possession, he would have to deal with a powerful noble claiming the freehold. Stoke passed with Godwin’s Earldom of Kent at the Conquest to Bishop Odo of Bayeux, and was recovered by the archbishop, as Domesday goes on to relate, “Postmodum vero regnante W. Rege diratiocinavit illud Lanfrancus Archiepiscopus contra Baiocensem Episcopum et inde est modo saisita Rofensis Ecclesia.”

It is probable that in some such way Odo received the city with the earldom, and with it the fortress, which the Bishop of Rochester, however, may have recovered possession of in the way shown, and which he certainly afterwards exchanged with the king for land in Aylesford. “Episcopus etiam de Rouecestria pro excambio terræ in qua castellum sedet tantum de hac terra tenet quod xvii solidos et iiij denarios valet.” There never was a castle at Aylesford. The bishop gave up the site of Rochester Castle, receiving for it as much land as was worth 17s. 4d.

Gundulf, the follower and friend of Lanfranc, and consecrated Bishop of Rochester in 1077, was a great architect, and a very remarkable man, well deserving the panegyric bestowed upon him upon his promotion, “Eratque Gundulfus religione plenus, literarum non nescius, in rebus forensibus acerrime elimatus et qui putatus sit divina potissimum electione hunc honorem meruisse.” As to his architectural skill, and his work at Rochester Castle, it is said, “Gundulfus, quia in opere cœmentarii plurimum sciens et efficax erat, castrum sibi, Hrofense lapideum de suo construxit.... Igitur, hoc pacto coram rege inito, fecit castrum Gundulfus Episcopus de suo ex integro totum costamine, ut reor, LX librarum.” This transaction between the bishop and the king occurred about 1076. The king was to restore to the see the manor of Hedenham, in Bucks, and the bishop to employ his skill and spend £60 in building a castle, that is, a tower of some sort. What Gundulf certainly built is the tower which still bears his name, and which seems originally to have been upon the city wall. On the death of William the castle was a very strong place, and was seized upon and held by Odo for Duke Robert. Odo (Judas Iscariot, as he was called by the friends of Rufus), associated himself with Eustace of Boulogne and Roger de Belesme, and garrisoned the castle, storing there the spoil drawn from the adjacent country. Rufus, whose activity was a strong contrast to the indolence of his brother, attacked Rochester, while Odo retired to Pevensey to await succour from Normandy, which never came. Rufus took Tunbridge and Pevensey, and forced Odo to demand the surrender of Rochester by his allies. The garrison understood the sincerity of the demand, sallied out, captured their royal escort and their bishop, and installed him, nothing loth, in the government of the castle. Rufus converted the siege into a blockade of both castle and city, and finally, under the pressure of a pestilence, both were surrendered, and Odo was finally banished. Gundulf died in 1108. This first siege of Rochester is one of the great military events of a very stirring period, and one deserving far more notice than has hitherto been bestowed upon it.

The castle, thus taken in 1088, seems to have remained in the Crown for thirty-eight years, when, in 1126, Henry II. granted “to Archbishop William de Corboil and his successors the perpetual charge and constableship of the castle of Rochester,” thus leaving it in the Church, but detaching it finally from the see of Rochester. The archbishop, we are told, built, in consequence, a handsome tower, “egregiam turrim,” which is, no doubt, the existing keep. William died in 1139, between which date and 1126 the keep was, therefore, built.

Henry II., also alludes to the castle in his confirmation charter to the church of Rochester, “volo etiam et ipsi et homines sui sint liberi et absoluti ab omni opere castelli, et expeditione archisve constructione;” so that, probably, the arx, or citadel, was then in contemplation or in progress. It may be that Gundulph’s Tower was removed to make way for the new keep, but in this case its materials would have been made use of, and some trace of them would be almost certain to be detected. But there is no such trace, so that, probably, the new keep did not supersede the older tower.

The castle, especially with its splendid and very strong keep, was far too important a military post to remain unchallenged in the possession of the see of Canterbury, and both when the see was vacant and at other times the Crown got possession, and its repairs were then entered on the Pipe roll. In 1141, William of Ypres, a Fleming, was its governor for Stephen, and when the Earl of Gloucester was taken he was confined here for a short time—from September to November,—until he was exchanged for Stephen. Various sums were spent upon the castle between 1167 and 1202, when it was again given up to the archbishop, then Stephen Langton. Towards the close of John’s reign the castle was placed by Langton in the hands of William d’Albini, to be held in the interest of the barons, whose army proposed to march from London to its relief. John, however, interposed, and in 1215, after a severe siege of three months, it was surrendered. Wendover expressly states that the military engines employed on this occasion produced but little effect, but that the place was taken by the efforts of the miners, who first undermined and threw down the walls, and then applied the same method of attack against the tower. It was, without doubt, on this occasion that the south-east angle was destroyed. In May, 1216, it was taken, apparently without any difficulty, by the Dauphin, but on John’s death it fell, with other Crown possessions, into the hands of Henry III. Under Henry III. considerable sums were spent upon it in repairs, especially in 1225, when, probably the broken angle was made good. In 1226–7 a bretasche and drawbridge were repaired, and the gutters of the hall in the keep. Unfortunately, here as elsewhere, the repairs were commonly those of buildings in the courtyard, all which are now destroyed. The keep was not used save for stores. In 24 Henry III., the tower was ordered to be whitewashed in those places which had not been so washed before. In 31 Henry III., both chapels were ordered to be wainscotted. One of these, at the least, was pretty sure to be in the outer ward, for the use of the garrison.

In 1264, the castle was in charge of the celebrated Roger de Leybourne, who had just taken part with the king. As de Montfort was supposed to meditate an attack on Rochester, Leybourne defended it with a strong garrison, and stored it with ample provisions. In April, the attack being expected, Warren, Earl of Surrey, Henry’s brother-in-law, came to the castle. The barons laid siege to it just before Easter, and remained before it nearly a week, but without any result, and on their retreat most of the garrison, with Leybourne, joined the king at Lewes, and took part in the battle. The loss of that battle was followed by the surrender of the castle, but after the death of de Montfort and the fall of Kenilworth, Leybourne resumed his governorship. A century later, in 1367–8, extensive repairs were undertaken by Edward III., under Prior John of Rochester as chief clerk of the works. Stone was imported from Beer, Caen, and Reigate, with copings and crests for battlements, probably for buildings in the court. Since that period it has played no part in the transactions of the kingdom, nor is its military history of any special interest. An extant drawing, taken in 1588, shows the turrets domed over and capped with vanes, like those of the White Tower.

Excellent papers have been written upon Rochester Castle by Messrs. Beal Poste, Blaauw, Hartshorne, and Burtt, the latter in the form of an account of Roger de Leybourne’s share in the Barons’ war.

P.S. Since the above account was written, Mr. Irvine, the very able and acute superintendent of the late Sir Gilbert Scott’s works at Rochester Cathedral, has made a discovery which is not in accord with any of the existing opinions as to the particulars of the Roman settlement. It appears that the south transept is built across a ditch, which runs north and south, and, no doubt, underlies the whole breadth of the cathedral. The ditch is 12 feet broad at the top, and 7 feet at the bottom of the excavation, now 14 feet deep, but was probably much deeper. It is cut through sound old gravel, and was filled up with black soil, with occasional dressed stones of Roman workmanship, Roman tiles, and bits both of black and red pottery, one of the former having a stamped interior ornament. Mr. Irvine, it is said, is in a condition to show that this was the ditch of a wall standing to its west, or towards the river. If so, the original station may have occupied the site of the present castle, at the west end of the accepted place of the Roman town, of which it may have been the citadel.