Chapter XII: Of the Edwardian or Concentric Castles (7)

There remained, however, an heir male, in John de Braose, by Maud de Clare: he was the grandson by Maud de Clare of that William who was famished with his mother at Windsor, and he recovered Bramber and Gower. His son was William de Braose, who held Bramber and Gower, and added considerably to the Sussex estates. He is said, in 1262, to have raised a large sum of money from his tenants by foregoing his right of murage. He died at Findon, 1290, 19 Edward I., and was succeeded by John de Braose of Bramber, who married Margaret, daughter of Llewelyn, Prince of Wales. He died at Bramber of a fall from his horse (16 Henry III., 1232), and was succeeded by William, who died 19 Edward I. William was a powerful baron, and active in the wars of Edward I. and II. He was, however, extravagant, and reduced to sell Gower, which led afterwards to disputes between the purchaser and the lawful heirs of the vendor, of whom Aliva, his elder daughter, married John de Mowbray, from whom descended, in the seventh generation, Margaret de Mowbray, who married Sir Robert Howard, whence comes the Duke of Norfolk, the present owner of Bramber Castle. The barony is in abeyance. Probably, with the failure of the elder male line of De Braose, the castle ceased to be a residence, but in 1644 it was strong enough to be held by Captain James Temple, for the king, against a strong Parliamentary party, and it was probably in consequence that the keep was blown up, and the castle reduced to its present condition. There is an engraving by Hollar, taken 150 years ago, representing it very much as it now is.

Another branch of the family of De Braose rose to the rank of Barons of Parliament, and held large Sussex possessions. The last male of this cadet branch, De Braose of Chesworth and East Grinsted, died in the reign of Richard II.

Bramber is, like Pontefract, Lewes, and Dudley, an example of a natural hill, scarped and defended by art, and crowned with an artificial mound; but, unlike those castles, it has both a mound and rectangular keep, a rare combination, found also at Guildford and Christchurch. A rectangular keep upon a natural isolated hill, unaccompanied by a mound, though not usual, is not unknown. Examples of it are seen at Hedingham, Corfe, and Bridgenorth. At Bramber, as indeed is the case in many other sites of early fortresses, the original earthworks have survived the Norman additions, and remain pretty much as they must have existed before the Conquest.

BRIDGENORTH, OLDBURY, AND QUATFORD, IN SHROPSHIRE.

The river Severn, in its course from Shrewsbury to Worcester, passes for several miles down a deep and rugged ravine, within or near to which lie the populous districts of Coalbrook Dale, Iron Bridge, Coal Port, and Broseley, early seats of the iron manufacture, and evidences of the wealth, though scarcely in harmony with the natural beauty, of the country. The ravine commences a little below the ivy-covered ruins of Buildwas Abbey, and terminates twenty to twenty-five miles lower down, about Bewdley and Stourbridge, where it opens out into a valley of a soft and smiling character.

About half-way down, between Pendlestone rock and the incoming of the Worf, the Severn receives upon its right bank the waters from a short but deep and broad valley, which descends obliquely from the north-west, and between which and the main valley intervenes the point of a steep and narrow ridge of rock, rising about 200 feet above the river, and upon the nearly level summit of which is placed the town and what remains of the Castle of Bridgenorth. The rock is more lofty, and the position far more striking, than that of Pontefract Castle, the defences of which were also in a great degree natural, and in these respects Bridgenorth may challenge comparison with Coucy, which it also resembles in the relation of the castle to the town. In both places the castle lay contiguous to the town, and their connected defences formed the common _enceinte_, while the castle had besides a ditch proper to itself.

At Bridgenorth the castle occupied the apex and south end of the platform, the broader and northern part of which was covered by the town, and the town walls abutted against those of the castle, while the castle ditch traversed the platform from one face to the other. Of the defences of the town only the north gate remains, and that in a very mutilated and disguised form, but the line of the walls may be traced, partly by the inequality of the ground and the arrangement of the streets, and partly by the existence of the cliff upon which they in part stood. An ancient fortified bridge, standing when Grose visited the place late in the last century, though now rebuilt, crossed the Severn east of the town, and was approached from it by a steep and narrow flight of rock-cut steps, and by a carriage-way cut in traverses almost as steep. This bridge defended the passage of the river, and connected the place with the suburb called the Lower Town upon the left bank. It is this bridge which is supposed to have given to the place its early name of Brug, or Breig, the distinctive “North” being an addition, probably in the thirteenth century, when there is said to have been an earlier bridge a little lower down. The town contains several old buildings, and among them, near the bridge, a very fine one framed with timber, in which was born Bishop Percy. There is also the fine parish church of St. Leonard’s, which stands at the north end of the town, east of the north gate, and just within the line of the old walls.

The castle platform is in plan somewhat of an equilateral triangle, each side being about a furlong. Leland says its area is about one-third that of the town. The two sides of this area were protected by a cliff so steep as to render a ditch unnecessary, and the face of which, where it needed support, was, and still is, revetted, the wall and edge of the cliff having been, no doubt, crowned by a parapet. The cliff rises out of a steep talus, or slope. The base of the area was defended by a wall within a ditch, upon which was a great gatehouse, standing in Leland’s time and for a century later. The ditch has been filled up and built over, and the wall removed. Just within its line still stands what remains of the keep, and a few yards to the east of that was the castle chapel, now the church of St. Mary Magdalen, and, until recently, a peculiar with a special jurisdiction. The chapel was collegiate with an endowment for certain prebendaries, disendowed at the Reformation. The present building, constructed in 1796, is a large and distressing example of Telford’s church architecture, in what the great engineer was pleased to regard as the Grecian style. The two buildings stand on the highest part of the castle area, which falls 30 feet to 40 feet towards the southern point. A modern wall has taken the place of the old _enceinte_. The view thence is very noble, nor does any town in England possess a finer promenade than that with which corporate care has encircled the area.

The fragment of the keep, long known as “the leaning tower of Bridgenorth,” seems to be the only masonry remaining of the castle, for the revetment wall of the cliff looks as though it had been replaced. The keep was a regular rectangular tower of the usual Norman pattern, but in dimensions very unworthy of the powerful earl who built it, or of the celebrated fortress of which it was the citadel. It was 45 feet square, and from 60 feet to 70 feet high to the base of the parapet. On each face were two pilaster strips, 8 feet broad by 6 inches projection, placed close up to but not covering the angles of the tower, which are thus converted into nooks, or hollow angles, of 6 inches in the side, and which, instead of, as usual, terminating above and below in a flat square, end in a sloping surface, as though to receive a column. The pilasters rise from a common plinth, and ascend to the parapet. Whether they were continued upwards so as to form the usual angle-turrets does not appear. Each had two sets-off on the face and outer edge only, reducing the breadth to 7 feet 6 inches and 7 feet. The sets-off were continued round the building. In the west face was also another pilaster of the same breadth and 15 inches projection. This died into the wall some feet below the summit, and seems to have been connected with the entrance door, which probably opened in its face, for it is broken away below and a part of the rough backing of an arch is seen. The walls at the base are 9 feet thick, and about 7 feet at the summit. The building is of three stages, a basement, at the ground level, 12 feet high; a first floor, 25 feet; and a second floor rather more, perhaps 30 feet. The floors were of timber. The first floor seems to have rested on a ledge, the upper and the flat roof upon joists, those of the lower lying east and west; those of the upper, north and south. The joist holes in the north wall have been closed by early work, and above is a slight set-off or shelf in the wall, as though the level of the floor had been altered. In the west wall, also, new joist holes, smaller, have been cut above the old ones.

The original roof was very steep, having two slopes and a central gutter, as at Porchester, and the reverse of the arrangement at Ludlow, where the ridge was central, and the two gutters lateral. The weather table remains perfect in the north wall, with a hole 2 feet high by 1 foot broad, to carry the beam which supported the gutter and the feet of the rafters. The table is seen continued horizontally upon the west wall, where it was laid as a flashing to cover the upper edge of the tiling. The walls were brought up to a level line all round, so as to conceal the roof. This arrangement, as at Ludlow, Richmond, Porchester, and the gatehouse of Sherborne, shows that there was no original intention of using the roof as a platform for mangonels and such like heavy machines. The flat roof, of lead, was apparently of later introduction. No traces remain of any mural staircase in the north or west wall, nor of any mural passages.

The basement was, probably, a dark store or cellar, reached only by a trap in the floor above. The entrance seems to have been on the first floor in the west wall, in which also are traces of a loop or small window. The north wall remains perfect. It was most exposed, and is without openings of any kind. In what remains of the south wall is one jamb of an original fireplace, of which is seen the sloping back, and part of a lateral nook and Norman abacus above it. The flanking shaft is gone. In the part of the east wall still remaining is the northern half of a small full-centred window, deeply splayed inwardly. The east and south walls above the first floor are gone. In the west wall, upper floor, is seen the north jamb of a small full-centred window set in a bold splay of hourglass section. North of it, in the same wall, is a small recess, probably for a lamp, and which seems to have been round headed. There, probably, was a fireplace in the south wall.

Projecting from the outside of the south wall, bonded into, and of the same age with it, is a fragment of curtain 7 feet 6 inches thick, in which, as at Kenilworth, is seen the jamb of a doorway, defended outside by a portcullis, the groove of which, square and 5 inches deep by 3 inches broad, shows that the grate was of iron. The groove, as at Kenilworth, stops about 3 feet from the ground, the door having been reached by steps. The groove is not open at bottom, but runs up behind a covering wall, as usual, and was evidently worked from the rampart, as is still seen at the Fishergate postern, at York. The door jamb is about 6 feet from the keep. This was evidently the entrance into the innermost ward, in which, or rather upon the wall of which, stood the keep.

A few feet to the west of the keep is a mass of masonry, clearly a part of the forebuilding which covered the entrance. Its face towards the keep is 13 feet long, and towards the south 16 feet. It varies from 3 feet to 6 feet thick, and is at present about 10 feet high. At present it is distant from the keep 3 feet 6 inches, but the two faces were evidently once in contact, and were displaced by an explosion. The mine by which the keep was destroyed seems to have been placed here.

The material of the keep was rubble-stone faced with excellent fine-jointed ashlar. The exterior face of the forebuilding seems to have had a similar casing.

It is not easy to obtain accurate measurements of the keep, so much has been removed, so much injured, and what remains is so obscured with ivy. Moreover, the interior is fitted up for two dog-kennels, kept in a very filthy condition, and with putrid carrion suspended from the walls. The ruins also stand in three distinct enclosures, all locked up. A mine has been sprung between the keep and the forebuilding, and the explosion has removed all the upper part of the latter, and so tilted the keep that it leans at an angle of fifteen degrees eastward from the vertical, and the upper part of the east and south walls are gone. Moreover, the keep seems to have been lifted bodily about three feet towards the east, and the north wall has a large open crack. About fifteen yards from the south-east angle stands a huge ivy-covered mass of masonry, probably the detached angle of the tower. What remains of the keep is held together by the excellence of the cement. The ruin is in a state of great filth and neglect, and it is much to be regretted that the whole area is not converted into a public garden. Judicious excavation would probably throw much light upon the details of the keep, and show the line of its contiguous curtain wall.

The masonry and details of the keep answer very well to the date of 1101 to 1102, to which history assigns it. It is certainly not earlier. The curtain of the inner ward was clearly of the same date, and enclosed a court in the north-west quarter of the general area, of which the keep probably formed the north-east angle, and which was entered on the east side close south of the keep. All else is gone; the “mighty North gate” of which Leland speaks is no more. The very ruins have perished, and the last trace of them, a good Norman arch, discovered while pulling down some houses in 1821, has since been destroyed by local Vandals.

The early history of Bridgenorth is exceedingly obscure. It is stated in the Saxon Chronicle that when, in 896, Alfred stranded the Danish ships in the Essex Lea, the Danes left them, traversed England, passed the winter at Quatbridge on the Severn, and there threw up a work. Three of the four original texts are thus rendered. The fourth makes them rest at “Brygce,” or Bridge, on the Severn. Florence of Worcester supports Quatbridge, and mentions the work or fortress. “Brygce,” in the Chronicle, is thought to be an interpolation, both where appended to Quat, and where it stands alone, it being probable that the Severn was not bridged at that time. There are at present two parishes into the names of which Quat enters on the left bank of the river, below Bridgenorth, Quat and Quatford, and upon the river is Danesford. Quat is regarded by Eyton as a corruption of the British “Coed,” a wood, the whole district having been a forest.

In the same Chronicle it is recorded that Æthelflæda, the great lady of the Mercians, a mighty burgh-builder in her day, and called by Henry of Huntingdon “Terror virgo virorum,” built, in 912, a burgh at Bricge, to which Florence adds, “on the western bank of Severn.” Bricge could scarcely be Bridgenorth, which is not even mentioned in Domesday. We ought, however, to find near the river, about Bridgenorth, earthworks thrown up by the Danes and by Æthelflæda, and it will be seen that there remain at the least three distinct works, any one or all of which may be of the ninth or tenth centuries. These are Oldbury, Quatford Castle, and Quatford.

Bridgenorth is not mentioned in Domesday. The Norman castle did not then exist, and there is no reason, strong and tempting as is the site, for supposing that it was occupied either by the Danes or by Æthelflæda. Mr. Eyton is of opinion that the site of the later town and castle is included within a certain two hides of land which in the survey constituted the demesne lands of the Norman Earl of Shrewsbury, within his great manor of Morville. Quatford, not Quatbridge, is mentioned in that record in conjunction with Ardintone. “Ibi,” that is in Ardintone, “Molendinum de iij oris et nova domus et burgum Quatford dictum nil reddens,” “there is a mill worth three ounces (5s. per annum), and a new house, and the borough called Quatford, paying nothing.” In 1085, therefore, it may be accepted that the earl had a new house at Quatford, where, indeed, it is known that at the request of his second wife Adelais, and in acknowledgment of her escape from shipwreck, he founded, about 1086, a collegiate church. The foundation charter of this alludes to the mount nigh to the bridge; the latter, probably an appendage to the new house, the former possibly part of the older earthwork of what is now known as Quatford Castle.

Earl Roger was succeeded in his English Honour and estates by his second son Earl Hugh, who was slain in Wales in 1098, and left the succession open to his elder brother, Robert de Belesme, who had already inherited his father’s estates in Normandy, and was Count of Ponthieu in right of his wife. Robert, who thus became Earl of Shrewsbury, though a cruel tyrant, was a man of great ability and energy, not only a great soldier, and “princeps militiæ,” or “Commander of the Forces” to Rufus, but a great military engineer. He selected the site and planned the works of the celebrated castle of Gisors on the Franco-Norman frontier. His brother’s death found him beleaguered in his castle of Balaon by Fulk, Count of Anjou, and the siege was raised by Rufus, who granted him, or confirmed him in, his brother’s Honour. When he came to England is uncertain, probably not before the end of 1099.

On the death, in 1100, of Rufus, Earl Robert took part with Duke Robert, whose claims, however, were not at first brought forward. It was probably while preparing for their open assertion that he decided to fortify the strong position which rose unoccupied scarce a mile from his father’s church and residence. His decision was prompt, and followed at once by his acts. He transferred the “Burgus” of Quatford to a new town on the hill, and with it his father’s house and bridge, which he also rebuilt. The result was the borough town, castle, and bridge of Bridgenorth, the latter structure giving name to the whole as Bridge or Bruge, the distinctive “north” not being added till the reign of Edward II. or III.

With Bridgenorth, Earl Robert also founded the castle of “Caroclove” in Wales, and such was his need that the works were carried on day and night. He also fortified Arundel, Shrewsbury, and Tickhill. His exertions in 1100 and 1101, when he seems to have built the castle, must have been excessive. King Henry, however, was not less active. He despatched Robert, Bishop of Lincoln, to lay siege to Tickhill, while he himself, having commenced with Arundel, proceeded to Bridgenorth. He took it, after a three weeks’ siege, in September, 1102, and this brought to a close Earl Robert’s short tenure of power in England. The earl fled to Normandy, his earldom of Shrewsbury was forfeited, and Bridgenorth was after a time granted to Hugh de Mortimer of Wigmore, the son of one of Henry’s most trusted supporters. In 1126, Waleran, Earl of Mellent, was here imprisoned, as was Meredith ab Llywarch in 1128. In 1130 wine was sent hither for the king’s use, so that Mortimer was probably rather constable for the Crown than the owner in fee.

Mortimer, in the new reign, took the part of Stephen, at whose death he held both Wigmore and Bridgenorth. As he was in rebellion against Henry II., the king took the field against him, and in April, 1155, the castle a second time stood a royal siege. Cleobury, one of Mortimer’s castles, surrendered in July, and Wigmore and Bridgenorth followed. Henry was for some time before the place, and his charter to Stoneley Abbey is dated “apud Brugium in obsidione.” It was at this siege that Hubert de St. Clair is said to have stepped forward to receive the arrow aimed at his sovereign, a romantic but unfounded tale. Henry retained the castle for the Crown, and used it largely as a prison for his Welsh hostages. In 1173-4, when Prince Henry rose against his father, Bridgenorth was victualled at a cost of £22. 5s. 2d. In 1175-6 the king dated a Wenlock Abbey charter from hence. The frequent charges for repairs between 1166 and 1189 show the importance attached to this castle by Henry II. In 1176 the Pipe Roll gives a charge of 1d. per day for the castle porter.

King John was six times at Bridgenorth, passing there about fourteen days. He confirmed a charter of incorporation to the town. Both he and his predecessor, Richard, kept up the castle, as is attested by frequent charges for repairs during sixteen years. In 1198 there was paid 6s. 3d. for the hire of the barge in which the wife of Griffith ab Rhys was conveyed from Bridgenorth to Gloucester. In 1203 John presented to a prebend in the castle chapel. In 1209 a stag from the adjacent forest entered the castle through a postern and was captured, and no doubt converted into venison by the castellans. For this they were prosecuted by the verdurers, and the proceedings imply that the castle was then in charge of five persons only, of whom one was the constable. No doubt in time of peace the royal castles were left almost to themselves, and often not even kept in repair. When a war arose they were repaired, garrisoned, and victualled in all haste. So far as repairs went, Bridgenorth, however, seems to have fared well. We read of repairs on the king’s house, on the basement of the castle, on the “Barbe-kana et Pons-tornalis” or barbican and drawbridge; on the turret of the outer wall, the chimney of the great chamber, the castle walls, the tower, the well, the glass windows in the hall, the queen’s oriel, and the chapel.

These charges are continued through the reign of Henry III., and well into that of Edward I., from 1218 to 1281. In 1232 the sheriff was to cause to be repaired the castle stable, and the kitchen within the barbican of the tower, and in 1244-5 was a charge for covering the tower at Brug with lead. This was probably the keep. In 1267, Henry III. and his queen were at Bridgenorth, and it would seem that the Mortimers were still connected with the castle, for, in 1273, on the death of Hugh de Mortimer, Edward I. continued his successor Ralph in the offices of sheriff of the county and constable of the castle.

In 1281 an inquisition was held upon the state of the building, which had latterly been neglected. The timbers of the great tower were rotten, the leaden covering having been carried away; also the castle bridge was broken down, so that carriages could not cross it.

In common with most other castles held by the Crown and therefore not inhabited, Bridgenorth seems to have been allowed to fall into decay from the reign of Edward I., so that probably little remained beyond the walls. At any rate, it played no part in English history until the great rebellion, when the town declared for the king, for whom the castle was garrisoned. The tower was attacked by the Parliamentary forces on the north side, and was entered by a breach near St. Leonard’s Church. Upon this the town was burned up to the castle, which still held out. The Parliamentary batteries are said to have been posted upon the Oldbury earthwork, which is probable enough, though there are no traces of parapets or breastworks of that date, which, however, from the distance from the castle and the deep valley between, were probably not needed. The garrison held out three weeks, and capitulated on honourable terms. The castle was blown up, and the materials probably sold.

Any account of Bridgenorth Castle would be very imperfect that did not take notice of the very remarkable earthworks seen in its neighbourhood, and which are evidently connected with the events referred to as of the ninth and tenth centuries. These are three in number—Oldbury, Quatford Castle, and Quatford; and first of Oldbury.

On the left bank of the Severn, about a quarter of a mile below and south-west of the castle, and on the opposite side of the deep dry valley that forms its western defence, the high ground of Oldbury is broken by three deep combes which descend to the river, and between which are two high ridges or knolls, steep towards the Severn and the combe on either side, and on the west connected by a neck with the higher land. The larger of these, that nearest the castle, is known by the inelegant but most descriptive name of “Pan Pudding Hill.” It does, in fact, much resemble in figure a beef-steak pudding just turned over out of the pan in which it was boiled. Naturally oblong, it has been scarped and rounded. The circular flat top is 150 feet diameter. In the centre is a slight nipple-like mound 3 feet high, and a raised bank now about 4 feet high crests its circumference. Towards the river the slope is steep for 50 feet or 60 feet, towards the land it is protected by a cross trench about 50 feet broad and 10 feet to 12 feet deep. Half-way down the slope on the eastern side is a narrow ledge or path which may have been protected by a stockade. Towards the south this ledge expands into a stage or shelf from 90 feet to 100 feet broad, and which is excavated so as to carry a ditch. The arrangements are very simple, and the mound is mainly natural, though scarped and fashioned by art. It was probably here that the Parliamentary guns were posted in the seventeenth century.

Close south of this hill is a second work, lower, smaller, and less clearly defined. Its summit is also circular, and about 100 feet diameter. These two works are of one general type, and probably of one date, and if not the work of Æthelflæda are no doubt of her period. They are either English or Danish, not British. That the main work is older than the Norman fortress is evident from a document of 1299, in which it is called “the old castle.” Moreover, it is the burgh which gives name to Oldbury, the parish in which it is situated.

Quatford Castle is on the left bank of the Severn, one and a quarter miles below Bridgenorth, and a furlong from the river, the intervening ground being a strip of meadow, while Danesford is still the name of an adjacent ford and village. A short steep combe descends from the high ground to the north-east, and, branching below, includes a knoll of rock perhaps 150 feet above the valley and 200 feet above the river. The soft red rock has been pared and scarped, and a part of the material employed to give an artificial top to the hill. This is somewhat of an oval, and seems to have had a sort of mound at its east end, now occupied by a modern castellated house. The slopes are steep, especially towards the west, and they are broken by narrow terraces, now walks, but which may have been ditches. The approach is by a sort of causeway on the north-east or least steep side. The summit and sides of the work are converted into a house and gardens, but the general arrangement of the original hill can readily be detected. It must have been very strong, and resembles generally Devizes and similar works of English origin. Probably this is the site of the “nova domus” of Earl Roger, as it was the seat of his English predecessor. It is a very curious work, and deserves to be surveyed on a large scale by the officers of ordnance. It may be mentioned that in the courtyard is a small passage cut in the rock, at an angle of forty-five degrees, and which descends by ninety-four steps to a well, whence the house is supplied. The passage is evidently an addition, the original well shaft descending, as now, vertically from the surface.

A little north-west of the castle, towards Bridgenorth, the tail of a piece of detached highish ground has been cut off by a trench, near a place called, in the inch ordnance, “Dog in the Wall.” It seems to have been a light temporary work for the accommodation of a small body of men.

On the same road, two miles south of the town, is Quatford, close to the parish church of that name, where the road crosses a steep ridge in deep, but probably modern, rock cutting. The ridge abuts upon the Severn, in a bold rocky promontory about 70 feet high. Upon it is thrown up a mound about 30 feet high, and mainly artificial. It is circular, and about 60 feet diameter on the top, which has been much cut about, probably for modern purposes. The sides are steep, about three quarters to one in slope. This mound is divided from the root of the promontory by a trench cut in the rock about 12 feet deep and 12 feet broad, which extends from cliff to cliff, and includes about three-fourths of the mound. Outside this ditch and to the east of it is an area of irregular figure, governed by the outline of the ground. Its north and south sides are defended by a ditch, which to the south is deep and wide. This probably included the east side, but is now superseded by the hollow road. The area is not very large, and would perhaps accommodate about two hundred men.

West of and beyond the road is the church, a building with some Norman remains. It stands rather higher than the camp, and its churchyard would, with a little care, have formed a part of and doubled the area of the camp. Whether it ever did so it is difficult to say, probably not. The ford, which gives part of its name to the village, is still in use when the river is low. It crosses the Severn below the camp.

This camp is omitted in the ordnance map, and of the earthwork of Burf Castle, placed a mile and a half to the west on that record, but one side remains.

It may be observed that the character of the surface of the country hereabouts is very favourable for the construction of these earthworks with mounds. There are scores of natural rounded hillocks of red sandstone that have an artificial aspect, and that, with a little scarping, would be strong. There is one, especially, close east of the road between Quatford and Dudmaston Park, that looks very like an English earthwork, and wants nothing but a ditch to make it perfect. Besides the earthworks above described are others in the district which appear to be of the same type. Such are Castle Hill, nine miles south-east from Bridgenorth, and the isolated knoll called “the Devil’s Spittle Dish,” two miles south-east of Bewdley.

BRONLLYS TOWER, BRECKNOCKSHIRE.

Bronllys tower, on the left bank of the Llyfni, a tributary of the Wye, is situate in the parish of the same name, close north of the town of Talgarth, on the regular and ancient way between Hereford and Brecknock.

The tower occupies the summit of a mound or knoll of earth, in great part artificial, which crowns the steep bank of the adjacent river; rising, perhaps, 60 feet above the stream, and 30 feet or so above the ground to the west of and behind the building. The mound is placed at the apex of an earthwork of rather a pear-shaped outline, of which the river-bank forms the steep east side, while to its base, or north face, has been applied a vallum nearly rectangular, and which may or may not be a Roman addition to a Celtic camp. The mound, which has borne the very considerable weight of the tower in safety, must be of considerably earlier date; and altogether the work resembles much one of those numerous instances in which advantage has been taken of an earlier mound to give elevation to a Norman or early English keep.

There are, however, it is said, remains of masonry, of the character and probable age of the tower, still standing upon a part of the vallum, and indications that, as is known to have been the case, the tower did not stand alone, but was within a base court. These walls are not now of any extent, and seem to be included within a modern house built upon the old enclosure.

This tower is at its base 37 feet in diameter, and batters inwards to 12 feet high, when it is girt by a bold cordon or string-course, much eroded, but apparently of a half-round section, with a water-groove on its under side. Above this the tower is cylindrical, and 34 feet diameter, or very nearly so, to the summit, which is at present about 60, and may have been 70 feet high. Besides the battering base there is a slight rough set-off, apparently part of the foundation, above ground on the southern face. Round the whole is a walk of about 18 inches broad, so that the mound is about 40 feet across at its top.

The tower is composed of a basement and three floors, above which was the battlement, now completely destroyed. The basement within was cylindrical, 18 feet diameter, with walls 9 feet 6 inches thick, and covered by a pointed vault, the ridge of which runs about north-east and south-west. This room was aired rather than lighted by a stepped recess, terminating in a small loop or hole at a considerable height, through which nothing could be seen. It was entered on the opposite or west side by a trap door in the first floor, which lifted within a window recess, and disclosed a flight of eight very steep stairs, 2 feet 7 inches broad, terminating in a doorway rather above 7 feet from the floor; below which, therefore, was probably a wooden ladder. The rebate shows the door to have opened inwards, and to have had bolts on its outer side. The floor is on the level of the top of the mound, and in its centre is a depression which may indicate a well.

Two openings have been broken into this chamber from the outside, on the east and west sides. Why nine feet of masonry should have been twice pierced, at immense labour, it is difficult to say; nor is there any trace of door or loop which might have made the task easier, or have suggested these openings. The broken walls do, however, show at the very base of the structure, on each side, a horizontal or nearly horizontal square drain, of very rough construction, in the substance of the wall. These holes have been the subject of much speculation. They were evidently drains from the upper floors of the tower, collected to fall into one outlet. Such drains may be seen in the heart of an overthrown solid tower at Corfe. It is not improbable that one of the breaches may have contained a recess or garderobe, which communicated with the drain, and suggested the penetration of the wall in that direction.

The first floor, of 17 feet 9 inches clear diameter, was entered by an exterior door in the wall, at the level of the cordon, or 12 feet above the ground, on the east side. There must have been exterior steps; but they did not bond into the wall, and may have been of wood. The door has 3 feet 3 inches opening, with a drop arch and plain chamfered moulding. It was defended by an interior door; but there was no portcullis or other defence. The door-recess is 4 feet broad, and has a drop arch.

This floor had two windows towards the north and south-west, the openings of which are about 18 inches broad, with plain equilateral heads. The former opens from a recess 7 feet 4 inches broad, having stone side-seats; between which, in the floor, is the trap descending into the dungeon. The other window has a recess 6 feet 10 inches broad, with a stone seat on its left side. In its right, or west jamb, is a door 2 feet 5 inches wide, square headed, beneath a drop relieving arch; from which rises a mural stair 2 feet 5 inches broad, lighted by two exterior loops, and with a flat covering, leading to the second floor. The window recesses are segmental, and are formed of excellent limestone tufa ashlar.

The second story had a timber floor resting on twelve corbels, and is cylindrical, like the first, and of the same diameter, but higher. It was the best room. Besides its entrance door on the west, it has on the south-west a fireplace under a flat segmental arch with plain chamfer, above which are two slender octagonal corbels, which evidently supported a hood, probably of timber. There are also two windows beneath drop-arch recesses, towards the south-east and north-east. The former has stone seats; and the latter a light 2 feet broad, with a cinquefoil head, of which the central foil is an ogee. The head is made of two stones only. The moulding is plain, having a shutter rebate inside; and outside, rounded jambs in place of the usual chamfer. This recess has a stone seat on the west side only. In the east jamb is a small square-headed door opening upon a mural stair of 2 feet 5 inches broad, of which nineteen steps remain, and which led to the third floor.

The stair is lighted by a small hole below, and above by a square-headed loop of 9 inches in a recess splayed to 3 feet 7 inches opening. The loop, though about 60 feet from the ground, was closed by one vertical and three horizontal bars. This floor may have been used as a prison.

The third stage has walls 8 feet thick, and had a wooden floor. The stair from below opened into it on the south-east side, but seems to have been continued in the south wall, so as to reach the battlement platform, now entirely gone. This floor has a small mural chamber, no doubt a garderobe, on its west side, the door into which is narrow, and has an arch of two stones, which seems to be four-centred, or of Tudor pattern. This door is placed between a window on the north-west, the recess of which has a flat drop arch; and another on the south-west, of which the recess is broken away. There is also a small fireplace on the north side, and another window to the north-east.

In the wall close south of the mural chamber are two small square shafts, one of which was no doubt a chimney and the other perhaps a garderobe vent from the battlements.

Bronllys Tower presents divers peculiarities. Though of rude masonry, its door and window dressings are excellent. In general design it resembles early English work; but its doors, recesses, fireplace, and corbels, seem of early Decorated, and perhaps, in parts, of Perpendicular work. It is altogether superior in detail to Penrice, which it resembles in dimensions, and it is inferior to Tre-Tower. The walls may be safely assigned to the first quarter of the thirteenth century; but it was no doubt inhabited as a place of defence, and afterwards as a dwelling, for two centuries and a half after this; and from time to time it received certain alterations, of which the present fireplace-front, the cinquefoiled and other windows, and the entrance to the mural chamber in the upper floor, may be cited as instances. The vault of the basement is possibly original, but may be an addition.

THE CASTLES OF BROUGH AND BROUGHAM, WESTMORELAND.

“Anne Dorset, Pembroke, and Montgomery,” Baroness Clifford, Westmoreland, and Vesci, hereditary sheriff of Westmoreland, and Lady of the Honour of Skipton, in Craven, was in every way a remarkable woman: she was of high birth, held large estates, was the widow of two considerable peers, and had received and largely profited by an excellent education. To a strong and copious memory she added a sound judgment and a discerning spirit. She was a person of great firmness of character, and passed her life amidst events that exercised and strengthened that quality. Among the many subjects upon which she was informed, and which ranged, says Dr. Donne, from “predestination to slea silk,” was included a very close knowledge of the particulars of her own estates, and a very thorough determination to maintain her houses and castles in good repair. She found the castles of her Clifford and Vipont ancestors, Appleby, Brougham, and Brough, in ruins; she restored and made them habitable, and, though time and the hand of the spoiler have again brought two of them, Brougham and Brough, to decay, their walls still exhibit much of the amending hand of the great Countess, as well as of the original work of her remote ancestors.

BROUGH CASTLE.

Brough Castle covers the whole of a steep knoll which rises 60 feet on the left bank of the Swimdale or Helle Beck, and is about 50 yards from the water. The beck receives the Augill from the south-east, just above the castle, and their combined waters, at times of considerable volume and force, fall into the Eden about a mile and a half lower down. The castle itself is 630 feet above the sea-level, and the encircling fells of Westmoreland and Yorkshire rise to elevations of from 1,000 feet to 2,000 feet. About five miles to the east are the sources of the Greta and the Balder, and a little further off and towards the north the head springs of the Lune, not the noble stream that gives name to Lancaster, but a tributary to the Tees.

Even in this wild and almost impenetrable country are traces of Roman civilisation. The road from Lavatræ (Bowes) to Luguvallium (Carlisle) ran through Brough, which is identified with the station Verteræ, and by Appleby and Brougham, joining at Penrith the main road, the work of the same people, from the south. Five and seven miles to the east of Brough are two Roman camps, and there are others upon the same line of road at Redlands and Kirkby-Thore, and a very perfect one at Brougham. The Roman road at Brough runs generally east and west, and crosses the beck at Market Brough, so called in distinction from Church-Brough, which lies three furlongs to the south of the river, and contains the castle. The collective parish is named “Brough-under-Stainmore.” Verteræ, if identified with the existing camp and castle, stands off the main road, and south of the river, as at Brougham.

Verteræ is represented by a rectangular camp, of which the castle covers the northern and higher end, that next to the river. This camp is 157 yards north and south, by 113 yards east and west, of which the platform within the ditch covers 134 yards by 90 yards. Of this area there is cut off at the northern end by a cross-ditch a plot, 90 yards east and west by 50 yards, which is occupied by the castle. This portion is further defended by some additional earthworks, perhaps Norman, to the east and west. The southern fragment of the camp seems to have been used as a sort of outwork to the castle, probably for the protection of sheep and cattle.

The castle was thus placed cross-wise in the camp, and parallel to the river, the action of which has carried away the outer half of its ditch, and converted the slope into a precipitous bank, at the top of which is the curtain-wall. The eastern outwork is composed of the end of the knoll, or ridge, on which the castle stands, and which is scarped into a triangular platform, the base of which, 57 yards long, covers the end of the fortress, and projects 38 yards. The ditch cutting off this work from the body of the place is 23 yards broad, and very deep. The earthworks westward are two banks and ditches, across the tail of the knoll, one 57 yards long, and 47 yards in advance of the main ditch, and the other 84 yards further in advance, and 94 yards long. Both are intended to cut off approaches along the river bank. The cross-ditch covering the south front of the castle is about 30 yards broad, and up it, from the east, came the main approach. These ditches, on the south-east and west fronts, are wholly artificial. There is also a trace of a bank and ditch along the east front of the camp, about 30 yards in advance of the main ditch, and about 60 yards long. A road, which may be Roman, comes up from the south, and crosses the Augill by a bridge, 250 yards above or to the east of the castle, to join the main road over another bridge in Market Brough. Upon this stream is the castle mill. The defences, in masonry, seem to have been confined to the castle proper. There is no trace of such upon the outworks, which probably were stockaded. The ditches were at far too high a level to have been fed from the river; but the soil is retentive, and they seem to have been filled with rain water.

The castle is composed of one ward, a trapezium in outline, the east, north, and west sides being at right angles, and respectively 77, 90, and 57 yards, and the south or oblique side, 94 yards. It is, in fact, a right-angled triangle, with the acute angle truncated. The keep is placed at the truncated end, and forms the south-west angle of the ward, its south and west sides being exterior, and in the line of the curtain. The domestic buildings were along the south side, and about the south-east angle, and built against the curtain. Near the centre of that side was the gatehouse, and attached east of it the hall, beyond which a large three-quarter drum tower capped the south-east angle. This and the gatehouse are the only mural towers. The kitchen and chapel, and some later buildings, probably rested against the east wall, and there are slight foundations between the gatehouse and the keep. The north curtain, towards the river, seems to have been free from buildings. Upon it are two buttresses, and in one a garderobe, entered probably by a side door and passage in the wall. This curtain is 3 feet to 5 feet thick, and from 12 feet to 15 feet high inside.

The _Gatehouse_ is placed near the centre of the south side of the castle. It was an oblong building, and formed the west end of a block, of which the hall and withdrawing-room formed the eastern part and remainder. It was composed of a passage between two walls, of which one remains, and is 6 feet 6 inches thick and 45 feet long; the outer 10 feet being outside the curtain. The portal is gone, but the spring-stone remains of the inner doorway, 3 feet 7 inches broad, and recessed 8 feet 6 inches within the inner front. The vault of the passage, about 10 feet long, rested upon three bold ribs, of which the springing-stones remain, and beyond which was the outer gate, of which a part only remains. There was an upper floor, and if the rubbish were cleared away the plan of the gatehouse would be visible. The walls seem Norman, but the ribs, vault, and fittings are probably insertions of the Decorated age.

The _Hall_ was poor. It was on the first-floor level, and had a floor above it, and, therefore, a flat ceiling. The basement is composed of three vaults placed transversely, with flat, slightly-pointed arches. The doors are towards the court, and one chamber has a small Tudor fireplace, in a corner. Each had a loop in the outer end, and in the ends of two are mural garderobes in the substance of the curtain. The hall was not above 12 feet or 14 feet high. In its north side is a round-arched recess, probably the original entrance, by an outer stair; and near it a fireplace. In the south or curtain side are two good late Decorated windows of two lights, rather flamboyant in tone, with plain exterior drips. Above these is a step, or ledge, for the floor of the upper room, and two windows, not directly over those of the hall.

There is no very decided evidence of a _Chapel_. The _Kitchen_, probably, was near the north-east end of the hall, and against the east end of the curtain. The withdrawing-rooms were at the east end of the hall, and extended into _Clifford’s Tower_, a fine bold drum of 30 feet diameter, which caps the south-east angle of the ward. This tower seems to be of Decorated date, and the base and part of the wall original; but it has been almost rebuilt, probably with the old cut stones, in the Tudor period, to which belong its numerous square-headed windows. Part of it was taken down in 1763.

The _Keep_ stands upon rather the highest part of the enclosure. It is rectangular, 43 feet east and west, by 51 feet north and south, and stands upon the curtain, with which its west and south faces are continuous. As it does not quite cover the whole end of the ward, this is closed by a low curtain, 17 feet long and 3 feet thick, which extends from the keep to the north-west angle of the ward. The keep is composed of a basement and three floors, and is about 60 feet high. The parapet is gone. It has a plinth only on the two exterior faces, where the ground is low, and there are two sets-off which indicate the level of the second and upper floors. At the end of each face is a pilaster, 7 feet broad, and of 6 inches projection, and those adjacent meet and form a solid angle. These pilasters rose clear of the wall to form angle turrets, of which parts remain. From the upper set-off, on the north and south faces, rises an intermediate pilaster, 3 feet broad. The walls at the base are 10 feet thick, and, at the top, 6 feet. The basement is at present nearly filled up with earth and rubbish, concealing much of the east side. Part of the south-east angle fell in in 1792, and obscures the details of the main entrance.