Chapter XII: Of the Edwardian or Concentric Castles (2)

From hence the curtain trends to the north-west, being in great part Norman, but repaired at what is called the “Bloody Gap.” Beyond this is a garret bartizan, called Hotspur’s Chair. It was the gorge of a half-round tower, now removed, but shown in the plan of 1650. Beyond this the curtain is again mixed Norman and Decorated of 1312-15, to the Constable’s Tower. This is a strong half-round tower, capping a flattish salient to the north-east; it has an exterior entrance by stairs to each of its three floors. From this a short curtain, 1150, leads to the Postern Tower, 1312-15, rectangular, pierced below by the vaulted and portcullised passage of the postern, and vaulted also on the first floor. This tower is a fine example of a Northumbrian military building of the Decorated period. The arches are round-headed or segmental, and it has in its rear wall a remarkable two-light window with a transom, and in the head a foliated circle. There is also a very perfect garderobe with external shoot. From this tower a curtain, now removed, ran upwards to the bow of the hall tower of the keep, and thus completed the defences of this ward. The original postern passage is filled up, or nearly so, with earth, but by the side of the tower a way leads to a terrace walk between the keep and the river, and extending to Falconer’s Tower.

The eastern ward contained the castle chapel and the conduit, supplied with water by lead pipes from an exterior spring called Howling Well. The chapel, a considerable building, stood detached near the Bloody Gap, and was removed in 1755. The ecclesiastical establishment seems to have been on a respectable scale, and included a chantry for the performance of the obits of the family. In this quarter also stood the brew, bake, and slaughter-houses for the use of the garrison.

Alnwick Castle is probably the finest extant example of a Norman castle of this type, having an open keep and a complete _enceinte_; for although most of the present buildings are either of the fourteenth or the nineteenth century, the plan is certainly Norman, and certain detached portions of the construction. It seems also that the keep was never a mere shell, like Cardiff or Arundel, but was always set about with towers and provided with a handsome gatehouse. A notable feature is the use of the round-headed and the segmental arch in the Decorated period. This is not uncommon in the North of England. The very free use of stone warriors upon the parapets, carried to an absurd extent in the repairs of the last century, is also remarkable. They are seen at Bothal, and in Edwardian works both at Caernarvon and Chepstow, but by no means so freely distributed as here. They were obviously intended for ornament only, but, of all figures, that of the eagle at Caernarvon is the only one at all appropriate. No archer would or could have stood on the crest of the parapet. Most of the later figures were very properly removed by Mr. Salvin.

There is found upon the battlements of both walls and towers, in various parts of the castle, a convenient arrangement for hanging a movable wooden shutter in the embrasures, so as to defend the warders from a Scottish shaft, and from the scarcely less keen edge of the bleak winds of the Border. The shutter was suspended horizontally, like a port-lid, but from trunnions, of which one rested in a round hole in one merlon, and the other in a similar hole terminating in a groove in the other, so that the shutter hung freely, and could be lifted in and out if necessary. The arrangement is precisely that applied to the roller of a round towel. Traces of this arrangement remain in various parts of the castle. A perfect example is seen on the barbican. It may also be seen on the east wall of Goderich, at Chepstow, and elsewhere.

The officers forming the staff of this castle as a civil residence, in 1567, were the constable or governor; the porter of the outer gate; the greive, or executive officer or bailiff; the receiver or auditor; the feodary, who looked up the services and tenures; the steward, learned in the law, who administered justice; the clerk of the courts, who engrossed the rolls and kept the records; and the foreign or outer bailiff, who collected the castle-guard and cornage money, and summoned the tenants and suitors. The whole of the annual payment to these officers was £58. 18s.

Looking to the character of the country, so charged with traces of early military earthworks, and to the strong and well-defined natural position of Alnwick, it seems probable that it was occupied as a camp by some of the tribes who, from a very remote period, made this Border their battlefield, and whose defences are still visible in eleven distinct earthworks within a very short distance of the town. If so, they would necessarily have placed their defences to the north and east upon the lines of the present castle. The interior eminence would certainly have been their citadel, and the trench, completing their security, would most conveniently have been carried along the general direction of the western front, so as to connect the head of the Bow Burn with the Alne, and thus complete the seclusion of the peninsula. Such a site, so defended, was not uncommonly adopted or constructed by the Northmen and Saxons when they became settled, and they would have placed the timber and palisaded mansion of their thane upon the central entrenched knoll.

Probably the Norman Gilbert Tison, of cloudy memory, who is the reputed pioneer of the Conquest in these wild regions, found and contented himself with some early kind of timber fortress, for the earliest traces of masonry that remain _in situ_ or have been extracted from the walls, though Norman, are of late character, and attributable to Eustace Fitzjohn, who married Beatrix, daughter and heir of Ivo de Vesci, who is thought to have married Tison’s daughter. Eustace, called De Vesci, flourished under Henry I. and Stephen, and died in 1157. He was a likely man to have constructed a great castle, being a baron of considerable power, sheriff of Northumberland, and founder of the abbeys of Alnwick and, in Yorkshire, of Malton. Also he must have felt the want of a strong place, for in his days, in 1135, Alnwick Castle was taken by David I., of Scotland, in the interest of the Empress Maud.

Eustace no doubt built in the first half of the twelfth century a polygonal clustered keep upon the knoll, gave it the gateway we still see, and placed his residence within. Traces of his walls are said by Mr. Tate to have been discovered when the last rebuilding was being executed. No doubt also he cleared out the moat round the keep. To him also must be attributed the general wall of the _enceinte_, and possibly the ditch outside it; and this would have been strengthened by mural towers, many of which must have stood where their successors are now placed. De Vesci’s work is indicated by the stones being mostly square blocks of moderate size, laid in courses, but in beds more or less wavy, as though the mason used neither line nor level. The joints are open. Beyond question De Vesci constructed a castle in keeping with his wealth, and worthy of the chief baron of the Border.

In July, 1174, William the Lion, on his way back from an invasion of Cumberland, found himself, to his surprise, before Alnwick. William, son of Eustace de Vesci, attacked him. He was unhorsed, captured, and sent into England, and beyond sea, to prison. Eustace, son of William, succeeded in 1190, and was visited by King John, 12th February, 1201, and 24th April, 1209, when the king received at the castle the homage of Alexander, king of Scotland. Four years later, May 14th, John ordered Philip de Ulecote to demolish the castle of Alnwick—a mandate which could scarcely have been obeyed, seeing the king himself was there 28th January, 1213, and 11th January, 1216, no doubt unwelcome visits, for Eustace was a Magna Charta baron. He met his death from an arrow before Barnard Castle in the last year of King John. Henry III. visited Alnwick 23rd September, 1256, and Edward I. was the guest of John de Vesci there 30th April and 1st May, and 16th and 17th August, 1291, and 16th August and 13th and 18th December, 1292; and again 22nd, 23rd, 24th, 25th, 26th, and 27th September, 1296; and 26th and 29th June, 1298.

The Barons de Vesci became extinct in 1297, by the death of William, seventh Baron, when the castle and barony were acquired, it is said, to the fraudulent exclusion of the natural son, by Antony Bec, the warlike Bishop of Durham, by whom, in 1309, 3 Edward II., they were sold to Henry de Percy, the representative of a warlike family, whose advent forms an important era in the history of the Border. Percy, as the leader of the Northern barons, made Alnwick his residence, and although in possession only for five years, seems to have rebuilt much of the fabric, the rest being completed by his son of the same name.

The Percy Castle, laid out nearly upon the Norman lines, presented very nearly the appearance of the present structure. The authorship of the Edwardian part of the inner gatehouse is established by the escutcheon of Clifford on its walls, the second Henry de Percy having married a lady of that house. To the first half of this fourteenth century may be attributed, as has already been pointed out in detail, nearly all the leading features of the castle, as it stood at the incoming of the first Duke of the present family. The Percies, though they maintained the reputation of Alnwick as the great Border fortress during nearly four centuries, do not appear to have materially altered the fabric of the two earliest lords. They received here Edward II. in 1311 and 1322, and Edward III. in 1335; but the later earls were much at Petworth and in Yorkshire; and upon the death of the eighth earl, in 1537, and the attainder of his brother, the family ceased to reside at Alnwick, and the castle was neglected. The Percy line ended in Elizabeth, daughter of Jocelyn, the eleventh earl, who, 30th May, 1682, married Charles, Duke of Somerset. Of their children, two had issue, Algernon and Catherine, who married Sir Wm. Wyndham, and eventually conveyed to that family the Percy estates at Petworth, Egremont, and Leconfield.

Algernon Seymour, Duke of Somerset, and by creation Earl of Northumberland, left one child, Elizabeth Seymour, who inherited Alnwick, and married Sir Hugh Smithson, created Duke of Northumberland, and ancestor of the present family.

A survey of Alnwick in 1567 shows the decay then to have been very considerable, and as the Seymour lords preferred their paternal residence, Alnwick became almost a ruin. From this it was redeemed by the first duke, who, under the advice of Adam, restored, and in part rebuilt, the keep; and although he fitted up the interior with plaster and frippery, made the exterior sound and good, and, on the whole, in keeping with the character of the place, and with what remained of the ancient buildings.

Matters so remained until the accession of Duke Algernon, better known as Lord Prudhoe, a naval officer and a good man of business, who had travelled much and possessed a cultivated taste, and was of a truly noble and magnificent disposition. While foremost in works of public usefulness connected with his estates, county, and profession, and careful to drain his lands, rebuild the cottages of his labourers, restore the local churches, and provide lifeboats for his dangerous coast, he, under the sound advice of Mr. Salvin, almost rebuilt the castle, preserving with scrupulous care all that admitted of preservation, and adapting his new work to the period of the first and second Percy, the founders of the later castle. Having thus restored the great fortress of the Border with strict regard to the rules of military architecture, he proceeded, under the advice of Canina, to fit up the interior in the style of an Italian palace. The contrast afforded is certainly extreme, and the attempt on so costly a scale was hardy; but the adaptation of the fittings to the irregular plan of the rooms is so well conceived, the materials employed are so rich, and the execution of the details is so skilful, that it is difficult to regard even so great an incongruity as other than a distinguished success.

Much attention has of late years been paid, and by very competent persons, to the history of this castle. Grose gives some particulars, now very valuable; but this and the castles of Warkworth and Prudhoe have been illustrated by the late Mr. Hartshorne, and are treated of also by him with great success in the Northumberland volume of the Archæological Institute. More recently, Mr. Tate has handled the subject of Alnwick Castle with both skill and accuracy in his admirable history of the barony of Alnwick. And finally, in the “Life of William Rufus,” Mr. Freeman has touched lightly but very effectively upon the connexion of Alnwick and its lords with Northumbrian history. In the above sketch free use has, to some extent, been made of the above materials; but the object of the writer has been to treat solely of the fabric of the castle, and that from a military point of view.

REFERENCES.

1. Donjon.
2. Inner Ward.
3. Outer Ward.
4. Terrace.
5. Stable Courts.
6. Kitchen.
7. Estate Offices.
8. Ravine.
9. Ditch.
10. Tower.
11. Inner Gate.
12. Middle Gate.
13. Barbican.
14. Hall.
15. Withdrawing-room.
16. Music-room.
17. Ante-room.
18. Library. Prudhoe Tower.
19. Chapel.
20. State Bedroom.
21. Entrance.
22. Spur Gallery.
23. West Garret.
24. Abbot’s Tower.
25. Falconer’s Tower.
26. Postern Tower.
27. Constable’s Tower.
28. Ravine Tower.
29. Hotspur’s Chair.
30. East Garret.
31. Warder’s Tower.
32. Auditor’s Tower.
33. Clock Tower.
34. Avener’s Tower.
35. Garden.

THE CASTLE OF ARQUES, NEAR DIEPPE.

Arques is one of the earliest examples of a Norman castle, for which reason, though not an English fortress, it has been thought convenient to include an account of it in these pages. This grand castle crowns and occupies the head of a steep and bold cape or promontory, in this case a spur from the great chalk table-land of the “Pays de Caux.” On the west it is flanked by a short but deep combe or dry valley, and on the east by the deeper and far wider valley of the Bethune and Varenne—streams derived from different sources, but which here meander across a broad and level bottom, above half a mile wide, until, a little below the castle, uniting, they receive the tributary Aulne, and, thus combined, under the name of “la Rivière d’Arques,” fall into the sea at the port of Dieppe.

The castle thus stands above the left bank of the principal valley. It is about 4 miles from Dieppe; and immediately below, and to its north-east, is the village whence it takes its name, remarkable for a church of unusual size, and a most elegant example of the style of the latter part of the sixteenth century. Beyond, upon the right bank, are the remains of the ancient Forest of Arques, a part of the spacious domain of the ancient lords of the fee, and upon the skirts of which, within shot of the castle, was fought, in 1589, a very celebrated battle.

The castle in its present form is composed of a rectangular keep, standing in the south-west corner of an inner ward, in plan something less than a half-circle, having its chord to the west, and contained within an _enceinte_ wall, strengthened by towers and buttresses along its sides and at its southern end, and capping its angles.

Applied to the north end of this is an outer ward, of later date, four-sided, and having drum towers at its four angles.

The main entrance, approached by a steep and winding road from the town, is in the north end, or at the point of the cape, between the two towers. Entering, is a second gatehouse, opening from the outer into the inner ward, also between two towers. A third gatehouse, at the other end of the fortress, leads direct from the exterior into the south end of the inner ward, and thus opens a communication with the root of the cape. There is also a lateral postern with vaulted passages in the west wall of the outer ward.

Outside the wall, encircling it closely, is the ditch, the most striking feature in the whole fortress. This is in general plan not unlike the long section of a pear, the northern end being the smaller, and the western side flattened so as to be nearly straight. The counterscarp of this ditch includes an area of about 5 acres. The ditch itself, measured from the level of the foot of the wall crowning its scarp, is about 60 feet deep, with slopes of 1 foot horizontal to 2 feet vertical, and about 70 feet broad. It is only just not too steep to be covered with short turf. The crest of the counterscarp is a ridge about 6 feet broad, and about 20 feet below the level of the foot of the wall. From it descends another slope, equally steep, but much deeper; on the west side descending about 150 feet to the bottom of the valley, and on the east to a rather less depth, as here this “glacis” is succeeded by a sort of broad terrace of pasture land, which falls gently towards the river, the level of which may be 250 feet below the platform of the castle. In many respects these defences resemble those of Bramber, in Sussex.

At the north-west end the ditch is traversed by a very modern causeway of earth, which supersedes the earlier drawbridge. At the south end, where the ground is highest and the ditch about 50 feet deep and 80 feet broad, there remain two engaged piers upon the scarp and counterscarp, and between them two detached piers, of which the outer has fallen against the inner. All are rectangular; and the inner of the detached two is considerably the larger, and probably carried a tower for the protection and working of a double drawbridge. These piers are of flint rubble, cased with ashlar, of which a small part only remains.

At the outer end is the earthwork of a _tête du pont_, or ravelin, of triangular plan, the passage from which was a little to the east side of the apex. This work was evidently constructed when artillery was in use, and is attributed to Henry IV., during the campaign of 1589. It no doubt represents an earlier barbican, also of earth and timber. There are no traces of masonry beyond the bridge. This work opens upon the ridge of the promontory, which widens and rises somewhat higher to the south. The ground is scarred with banks and ditches, the remains of fieldworks of various dates, both of attack and defence.

The design of this castle ditch is peculiar. The more obvious plan would have been to place the walls upon the edge of the hill, and scarp its sides down to the valley with such steepness as suited the ground. Instead of this, the upper 40 feet of the hill, being chalk rock, was scarped vertically, and then faced or revetted by a wall, upon which was placed the _enceinte_ wall of the inner ward. At the foot of the revetment was then excavated the ditch just described, the material being thrown outwards so as to form an artificial scarp, which thus became a sort of advanced banquette or earthwork beyond the ditch, representing the crest of the glacis, and capable of being held by a line of soldiers, but which, when taken, was too exposed and too narrow to allow of its being held, or of cover being constructed upon it. This banquette was at a rather lower level than the foot of the opposite wall.

Such an arrangement is found in other castles in Normandy, and notably, as pointed out by M. Deville, at Molineaux, De Longueville, Bec de Montagne, and in the later work of Château-Gaillard, where, however, it is less marked.

The keep is rectangular, about 80 feet north and south by 70 feet east and west, and at present about 60 feet high. It stands in the south-west corner of the inner ward, close to the _enceinte_ wall, of which its south-west angle forms a part. Its walls at the ground level are about 13 feet thick. It presents three buttresses on the north face and two on the south—the third being a sort of cap thickening and enveloping the south-east angle. These are of the unusual breadth and projection of 9 feet, and they rise to the present, which cannot be above 10 feet below the original summit. At the north-east angle the adjacent buttresses are set square, leaving the angle free. The south face is plain, or nearly so, excepting the cap at its south-east angle, which extends southwards, and is connected with the adjacent _enceinte_ wall. The west face is plain, outside of which was the entrance.

This is composed of a flight of steps, beginning upon the north face, passing by a doorway through its most westerly buttress, and which then, turning, is continued along the west face, until at its south end it lands in the usual square appendage or forebuilding common in these keeps. This staircase was guarded by an exterior wall, and had gateways at its foot and its summit. It was covered over, as appears from marks upon the wall; and above it was no doubt the usual platform for defence. In the basement of the forebuilding, which forms also part of the _enceinte_ wall, was a vaulted chamber opening into the basement of the keep, as at Rochester, either a store or a prison. The landing story was barrel-vaulted, having at one end a loop towards the field, and at the other a door in the wall of the keep.

Entering this door, the staircase is continued southward in the wall of the keep, up a roughly-vaulted, round-headed, mural gallery, until at the angle it reaches the level of the first floor. The gallery now turns the angle, and is continued on the level half-way along the south wall, when it is stopped abruptly. A door on the right leads upon the _enceinte_ rampart, and one on the left probably led into the keep.

The interior of the keep is composed of a basement and an upper or, perhaps, two floors, divided by a north and south wall into two chambers on a floor. This division-wall is said to be original, and ought to be so in a keep of this size, but it looks of the sixteenth century, and may represent an older one. The west basement is much choked up with rubbish. The east chamber is tolerably clear, and shows an exterior aperture in its east wall, near the south end, which communicates with the adjacent south gateway. This may be original, but it is now a mere hole. In the north wall is a short mural gallery, entering a well-stair in the north-east angle, which ascends to the first floor and chapel only. The lower stage was not vaulted. The first floor has four windows on the north side, two in each room, and one on the east side. The second floor had also similar windows on its north, and a vaulted chamber on its east side; no doubt a chapel. This chapel is formed by throwing a vault from buttress to buttress for its floor, and at a higher level for its roof, and above this were the leads. There are traces of similar chambers on the north front. The upper story has been vaulted in six bays, three on each side, duly groined and ribbed, as is shown by the springers. The material and the workmanship, no less than the section of the ribs, show this to be a late addition, probably of the sixteenth century.

The supposed two upper floors were very possibly intended for one floor of state, with two tiers of windows and a chapel above. The chapel seems to have had a barrel round-headed vault, probably groined. The accounts show this eastern side to have been the royal chamber in the fourteenth century. The fireplaces seem to be confined to the upper floors. As now seen, they are of the date of the vaulting.

In the south-west angle of the keep, very near the wall, is the well, of which the pipe was continued at least to the first floor, as in the additions at Richmond. It is about 6 feet diameter, lined with ashlar, and in 1768 was choked up at 254 feet deep, or about the level of the river; a depth now reduced to 30 or 40 feet.

Outside, between the buttresses, are traces of walls, as though the spaces between them had been turned to account below as well as above; but these walls are thin, and do not seem original.

M. Deville cites the public records for the existence in 1318 of four turrets on the keep, roofed with lead.

M. Le Duc, in his Dictionary, Art. “Donjon,” gives a great variety of very curious detail connected with this keep, detail unknown to M. Deville, and for which there should be some authority other than the traces actually existing, which are very unsatisfactory.

The keep is built of large chalk flints grouted copiously in mortar, and cased outside with ashlar, now mostly stripped off and removed. Within, the flints are occasionally laid herring-bone fashion. The ashlar was a calcareous tufa, known in the country, and formed by the trickling of calcareous springs over moss and similar vegetation. It was much used in the earlier French castles. The later ashlar of the vaulting ribs and inserted door-cases seems to be a fine hard limestone, approaching Caen stone in appearance, and perhaps actually that material. Where the ashlar is wanting, the putlog holes are seen, placed with exceeding regularity. The joints of the original ashlar are large, those of the later fine. The new and old ashlar can readily be distinguished; but one flint wall is very much like another.

The inner ward, in length about 160 yards, and in its greatest breadth about 70 yards, is a natural chalk platform, revetted all round by a wall about 8 feet thick, which on the east side is reduced to a parapet, but on the west rises about 20 feet higher, probably its original height. In 1708 this court contained the apartments of the governor and the staff of the garrison, a well, and a chapel. These were probably of the sixteenth century or later, and have now entirely disappeared.

The _enceinte_ wall, which girdles this inner ward, deserves attention, as most of it is of early date. Setting aside the four northern towers with their curtains, which are of later date, we have about 380 yards of curtain broken by five mural towers and three rectangular buttresses. Nearly the whole of the wall is faced with flint, with three bands of ashlar. Much of the flint is laid in herring-bone fashion, the repairs, where of brick, being of much later date. The ashlar bands are of tufa. Of the towers two, half round, are on the east front. Of these one, though probably original, has been cased with brick. The other has had an ashlar base, and the upper part, of flint, shows herring-bone work. Each is about 15 feet diameter, with walls 5 feet thick.

Between these towers are three rectangular buttresses; two of 15 feet breadth and 12 feet projection, and one about 7 feet square. The two former contain no herring-bone work, and are probably early additions, perhaps by Henry I.; the latter is original.

On the west face are now no towers, but in 1708 there were two, half round, and of small size, traces of which remain. They were, no doubt, original.

The remaining three towers capped the three salient angles of the south end, the central containing the gateway, and the others flanking it. All are one quarter engaged.

The flanking towers are alike, about 22 feet diameter and 55 feet high from the exterior base, with walls 7 feet thick. The bases are either solid or pierced by steps leading down to the galleries. There is a regular basement story, and above it a floor on the level of the inner ward. They are not vaulted and show no exterior herring-bone work, though one has a little inside. A modern summer-house has been built upon that to the south-east.

The central or gate-tower is 24 feet diameter. It is pierced at the level of the scarp by a portal which opens upon the drawbridge, the piers of which have been described; and there is a stage above this.

The portal ascends towards the keep; its details are much broken down, and little can be made of them.

Of the _enceinte_ of the inner ward there remains to be noted the _northern gatehouse_. This, the original and, probably, the only entrance to the Norman fortress, though much ruined, does not appear to have been materially altered. It consists in a rectangular building, 40 feet deep by 20 feet broad, set in the centre of the curtain, with which its outer face is nearly flush. It is crossed by an outer, middle, and inner wall, each pierced by an arch of 12 feet opening, through which lies the passage. There remain the rectangular grooves of a portcullis, and a few years ago there was evidence of a second, and in the wall is herring-bone work. The arches are plain, without moulding or chamfer. The inner one is round-headed, and springs from a flat abacus, chamfered below. The joints of the ashlar work are about 1 inch wide. The curtains on either side of this gateway, though much repaired, seem to be original, and there are traces of the old round gate-towers. The pit of the drawbridge remains in front of the gate, upon the original line of fosse.

The outer ward no doubt occupies the site of an earlier outwork. It is built against the narrow and north end of the inner ward, is four-sided, about 250 feet north and south, by a mean of 110 feet east and west. Its west side is straight, being the continued line of that face of the old fortress. The east face has a slight re-entering angle, caused apparently by the shape of the ground. At the two southern angles are two large drum towers, which connect the old and newer work, and probably replace two smaller and older towers. These stand in the line of the old ditch, and flank the Norman gateway. That to the south-west, rather the larger of the two, and slightly oval in plan, has a mean diameter of 60 feet; it is of two stages. Both are vaulted, or rather domed. That to the south-east, of 50 feet diameter, is nearly circular, but has a remarkable spur or keel-shaped projection towards the field. It is of three stages, the two lower being domed. These towers are in fact casemates, having embrasures for small culverins towards the field. Each has a well-stair in its southern side, and is entered from the gorge.

The two other towers cap the northern angles of the ward, and flank the main gateway. They are of irregular form, semicircular to the field and angular within. In diameter they are about 40 feet, and about the same height. They are of two stages, which have been domed. The walls of these four towers are from 14 feet to 16 feet thick. They are of flint faced with brick.

Between the gate-tower is the gateway, composed of a larger and smaller portal, the latter very narrow. The present work is modern, but no doubt in this double entry represents the earlier openings. The ditch, two detached piers standing in it, and with these any traces of the drawbridge, are concealed by the modern causeway.

In the west curtain, near the south-west tower, a flight of steps beneath hanging arches of brick descends from the ward level in the direction of the foot of the wall. This is much encumbered with ruin, but seems to have been a postern, opening upon the ditch.

All the works of this outer ward are of flint, rubble-faced with brick, which material forms the lines of the embrasures. The quoins are sometimes of ashlar, as are the extensive string-courses and bands, and the dressings of the openings. Traces of herring-bone work in its west curtain, outside, will be accounted for afterwards.

As the original castle was confined to the inner ward, its ditch everywhere encircled it. When the outer ward was added, the intercepted portion of the ditch was partially filled up, but the new work was included in a new ditch, which was an accurate prolongation of the old one, of equal depth and breadth, and continued in the same direction. Accurate observation will, however, detect two slight shoulders in the counterscarp, showing where began the curve of the old ditch.

Very remarkable in this castle are the subterranean galleries, driven in the chalk rock beneath the lines of the original wall, and behind the scarp of the ditch. These are now in part blocked up, but there still remain two or three hundred yards of them of which the direction is known. They are entered by a passage in the inner ward, in its north-west quarter, and by a descent of fifty-four steps near the keep, and possibly from other points now lost. The galleries are about 7 feet high and 6 feet wide, rudely cut, and somewhat singularly laid out. They lie within, without, and beneath the wall, and give off frequent spurs or short passages intended to occupy any space in which a mine was likely to be opened. At this time they have been broken into at three or four points in the scarp of the ditch, about half-way down. They were intended as a system of permanent counter mines to meet any attempt at mining on the part of the besiegers. In one place a large central pier is surrounded by a gallery, whence the branches go off,—an arrangement intended probably to check the progress of those who might break in. So far as is known, these galleries, of which a survey was made in 1708, are confined to the southern half, or four-fifths of the old castle. There are none under the outer ward.

Outside, and to the north-east of the outer gate, are some semicircular platforms, which seem to have played a part in the defence of the castle; but whether before or after the use of artillery is uncertain.

The Bel.—A curtain wall, which originally was about 5 feet thick, and from 15 feet to 20 feet high, commences abruptly upon the crest of the counterscarp in two places; one on the east opposite to the mural tower, which marks about the centre of that front of the castle, and thence descends towards the river; and the other north, close outside of the drawbridge of the main entrance. This latter wall is continued down the hill, and makes a bold sweep towards the town, and, finally reaching the river bank, joins the river wall. The enclosure thus formed contains about twice the area included by the ditch of the castle, and has long been known as Le Bel or La Baile, a form evidently of the Norman-Latin _Ballium_, called by us the “Bailey.” This enclosure is traversed by the road from Dieppe to Martigny, which passes through the two gates bearing those names. A third, or water-gate to the east, opened upon the river. Upon the north front appear to have been two small half-round mural towers, of which one remains. The river has somewhat encroached upon the lower part of the Bel, and has undermined part of the wall.

The Dieppe gate, which is also that from the town of Arques, was in 1433 called “La Première Port du Bel de Château d’Arques,” and, as such, was the subject of a tenure by castle guard, the tenant being bound to defend it for forty days in time of war. From this gate the approach ascends to the castle, having the curtain-wall as a protection on its right. Where this wall approaches the castle, at either end, it is covered by a broad ditch, continued down the slope.

A flint wall may be of any age, but the remains of the gates, which a few years ago showed round-headed arches, prove this inclosure to be of early date, probably one of the earliest additions to the castle, and made by Henry I. No doubt, before the construction of the outer ward, the wall of the Bel was produced so as to unite with those of the castle. M. Deville has discovered a part of this wall worked into the great curtain of the outer ward, which lies in its line, and may still be seen.

Looking to the history of this castle, and to the evidence afforded by its remains, it seems probable that the keep is the oldest part of the masonry, and the work of the Conqueror’s uncle, Guillaume d’Arques, between the years 1039-1043, and it is supposed to be one of the earliest, if not the earliest, of the rectangular Norman keeps known. The chronicle of Normandy, cited by M. Deville, says of William, “Si fist faire une tour moult forte, audessus du chastel d’Arques”; as though there had been an earlier castle, which the aspect of the earthworks renders highly probable.

To the Conqueror or his immediate successor must be attributed the _enceinte_ of the inner ward and the formation of the galleries. The ditch may be much older than the western masonry. The inner ward no doubt formed part of the original plan, and it is only the occasional appearance of round turrets upon the wall that leads to the opinion that any time intervened between the actual construction of the keep and of its surroundings.

The southern entrance, with its gate and two flanking towers, and one or two of the other mural towers or buttresses, seem to be additions, but of the Norman period, probably the work of the Conqueror’s son, King Henry I., who, about 1123, seems also to have enclosed the Bel. Robert de Thorigny—called also “Du Mont,” from his abbacy of Mont St. Michael—a Norman chronicler of the twelfth century, says that King Henry I. “fortified admirably the castle of Arques with walls and a tower.” This has been held to show that the whole structure was the work of Henry, who reigned from 1105 to 1135, and the extreme boldness of the buttresses and superincumbent constructions of the keep no doubt favour this view; but, as M. Deville remarks in the same passage, similar reference is made to Gisors, Falaise, and other castles, known to be of earlier date.

M. Deville is disposed to attribute the southern gate to Charles V., as he finds a record of 1367, charging cost of transport of 6 “nances” of stone, each of 16 to 18 “tonneaux,” from the river to the castle, for the masonry of the new bridge and the new gate of the castle. This material was taken by the king’s direction from the dismantled “manoir” of Veules or Weulles, at St. Valery-en-Caux. The accounts of 1378-80 mention the tower on the bridge behind the keep, its drawbridge, axles, ties, “vergues” or levers 18 feet long, and its beams of 9 feet. This was probably the southern drawbridge and gate, including the opening of the communication between this and the basement of the keep. These works are attributed to Charles V. about 1378-80. He probably only pierced the existing central tower, not otherwise altering or rebuilding it.

The next considerable work was the outer ward, which may be attributed to the fifteenth century, subsequently to the general use of brick and the introduction of siege artillery. It is singular that no record of this very considerable work should be preserved, for it included not only the outer ward, a castle in itself, with its enormous towers and massive curtains, but the extension of the very formidable ditch, the repair of the older walls and towers, and finally the fitting up and vaulting of the keep. All this is supposed to have been the work of Francis I., and it is said that the date of 1553 was inscribed upon some of the additions to the keep.

Henry IV., during his occupation of the castle in 1589, may have constructed quarters in the inner ward and repaired what was amiss in the old building; but more probably his traces are to be found in the field works which crown the adjacent hills, and along the high ground towards Dieppe.

It has been thought that the ditch of the castle is a remain of an older fortification, such a work as the early Northmen or still earlier Celts might have constructed. No doubt this was usually the case with the sites of the great Norman castles, both in Normandy and in England, and the position of Arques is a tempting one. There is, however, no positive evidence of an earlier encampment.

It will be observed that the keep is so placed as to command both the inner ward and the most exposed side of the castle,—that along the level ridge of the promontory. It was perfectly capable of holding out when all else was taken, and finally, if threatened with fire or starvation, its garrison had a possible escape by the galleries.

This castle is the triumph of Norman skill. Often attacked, it was never taken by storm. Without being a royal residence it was visited in peace or in war by our Norman kings, from the Conqueror to John, and by most of the kings of France, from Henry I. to Henry IV.; and, after a lapse of 800 years, its oldest parts are still those best worth attention, and are at least as well preserved as the additions of far later date.

ARUNDEL CASTLE, SUSSEX.

Arundel, in its position, magnitude, and history, is one of the foremost of the mediæval military structures of England. English in its origin, Norman by adoption, it is not only among the rare castles recorded in Domesday, but is the only one therein specifically mentioned as in existence before the coming in of Duke William. It was held by Alfred and by Harold, and was granted with its surrounding and dependent lordships by William to the head of the great house of Montgomery as an acknowledgment of kinship, and a fitting reward for his distinguished services at Hastings. Three members and two generations of that family held the castle for about thirty-two years. They were followed by Henry I. and his Queen, who held it for half a century, and it has since descended in succession by the lines of D’Albini, Fitz-Alan, and Howard, through seven centuries of inheritance, to its present lord. Though less extensive in area, it is equal to Windsor in antiquity and position, and resembles it closely in type, having an upper and lower ward, and a lofty mound partially interposed between the two, girdled by a deep fosse and crowned by a circular keep. Within its precinct is the Chapel of St. George, and far beyond the walls, but in one direction only, extends its broad demesne. Just beyond the wall of the Norman castle, but probably within the earlier English outwork, stands the fine parish church, a Fitz-Alan gift, and attached to its east end is the chapel of the college founded by Richard, Earl of Arundel, in 1386, and now, by a grant or purchase in the reign of Edward VI., vested in the descendants of its founder, and the private property of the lord of the castle. The buildings of the college are mostly removed, and the chapel contains many tombs of great magnificence and considerable historic interest, beneath which rest the remains of many members of the great house of Fitz-Alan, who loved full well and gave full largely to that church to which their Howard descendant has shown himself a not less liberal benefactor.

The Arun, a principal river of the south, has its sources in Surrey. It traverses the whole breadth, here about twenty miles, of the county of Sussex, and in its passage to the sea cleaves the southern ridge of the chalk by a very striking ravine or dell, which, with the river, gives name to the castle and the town, as well as to the Rape of Arundel, within which they stand. The castle is placed upon a bold bluff of chalk, which rises on the right or western bank of the river, here flowing in easy curves across a broad tract of low and level land, which intervenes between the downs and the sea, from which the castle is distant about four miles. Upon and at the foot of the slope has sprung up the town, which for centuries paid allegiance to the castle and received from it protection and support. Six miles westward of the Arun is Bognor, rich in Roman remains, and as far to the east is the mouth of the Adur, where Shoreham contests with Portsmouth the representation of the ancient Portus Adurni. A few miles north of the castle, Staneway marks the line of the old Roman road between Chichester and London. Various villages, the names of which show their English origin, are thickly posted along the lower course of the Arun; but the name of the river may be British, as are no doubt the hill camps here and there scattered over the downs. The Rape is supposed to be a Jutish division. It includes five Hundreds, and was formerly nearly all forest. The courts were held beneath an oak-tree at Madehurst.

The bluff occupied by the castle is the end of a ridge of high ground, which is specially steep towards the east and south, and ascends to still higher land towards the north and west. In this direction lies the park, which covers about 1,100 acres, and includes, between two ridges of the chalk, a deep valley, in the lower part of which, just below the castle, is the lake called Swanbourne, the pool from a very remote period for the use of the mill.

The history of Arundel has been written by Dallaway in his Western Sussex, and by Tierney in a specific work. Whatever may be the value of either production for the history and descent of the lordship, neither gives a full plan of the earthworks of the castle, nor enters at all scientifically into its details from either the military or architectural point of view. Unfortunately, the proximity of the castle to Brighton and some lesser watering-places has caused the exclusion of visitors, save under restrictions so very narrow that it is impracticable to examine the works at all in detail. It appears from Tierney’s history that the castle is protected towards the north or open country by a double line of defence, composed of a bank and ditch, of which the outer includes the present as well as the older park, and is, in places, at least two miles from the castle. The inner line is of much smaller extent, but the two seem intended to protect the town and the port, as well as the site of the castle, and to be rather lines of defence than the remains of a camp, whether British or English. Within both of these, and probably of later date, are the earthworks upon a part of which the Norman castle was founded. These resemble the works common to England with Normandy, and may probably be attributed to the Northern invaders in the eighth or ninth centuries. The high ground is occupied by an oblong inclosure, of which more or less of the original earthbank remains, and outside of which to the south and east was the natural steep, and to the west a natural hollow, deepened and extended round the north front by art. The inclosure is in length 317 yards, and in breadth 83 yards, and includes about 5½ acres. Near the centre of the western side, and forming a part of the _enceinte_, is a large circular mound, almost wholly artificial, having its proper ditch, which on the outer face is also the ditch of the general _enceinte_. The mound is about 90 feet diameter at its table summit, and about 230 feet at its base. As the ground rises towards the north, the height is on that side about 50 feet, and on the opposite side about 70 feet from the bottom of the ditch. The Windsor mound is about the same height, but rather larger, being 125 feet across, at its summit. By its projection this work materially narrows the main area, and with the addition of a short cross-ditch, bank, and wall, now gone, divided it into an upper or northern and lower or southern ward, in the latter of which are the domestic buildings. These two wards were the parts of the earlier fortification taken possession of by the builder of the Norman castle, but there is also an inclosure opposite to the mound, and outside the ditch, which runs up to a narrow end towards the north-west. Whether this was fortified with masonry is uncertain, but it certainly was a part of the earlier fortress, and shows its extent and importance. The ditches everywhere being in the chalk, and at a high level, were no doubt always dry. Earl Roger, probably, found along the crest of these earthbanks the same sort of defences that he left behind him, on similar banks and mounds, at his paternal castle in Normandy; but whether these were of timber or of masonry is uncertain. They might very well have been of either. It is, however, certain that the earl, or his immediate successors, inclosed the main area within a curtain of masonry, which ran up the mound and placed half of it, as at Lincoln and Tonbridge, outside the inclosure. Upon this curtain, a little south of the mound, was placed the gatehouse, and along it at various points were mural towers, probably square, of which one, on the curtain north of the mound, and known as Bevis Tower, still remains, though more or less altered. The domestic buildings were then, as now, along the sides of the lower ward, and upon the mound was placed a shell keep.

As the public are only permitted to inspect the lower part of the gatehouse, the adjacent curtain, and the interior of the keep, it is impossible to form a correct opinion as to the extent of the original earthworks, or as to the age of the general curtain, or the basement of the domestic buildings, parts of which are said to be of Norman date. That the earthworks were extensive, and in part remain, may be seen from the summit of the keep, and from the upper gate of the park. Dallaway gives a large, but not quite correct, plan of the keep and its mound, the gatehouse and Bevis Tower, and the lower ward; but his plan does not include the higher ward nor the earthworks, nor is it accompanied by any sections. Dallaway and Tierney were not strong in architectural details.

The _Gatehouse_ stands on the line of the curtain, projecting inwards from it, and upon the southern edge or counterscarp of the ditch of the mound. It opens into the lower ward. It is of the early Norman type, resembling that at Tickhill. It is a square of 32 feet, having in the outer and inner faces a large round-headed doorway, quite plain, without even a chamfer, and with a plain sloped abacus. The chamber is, of course, square, and but little broader than the doorways. That towards the ward seems to have been closed by a pair of doors only. The other or outer doorway has a broad portcullis groove, which may or may not be original; more probably not. The covering—the floor of the upper room—was no doubt of timber, and flat. The present vaulting is an insertion. There was always an upper floor, but this has been much altered, chiefly in the Decorated period. The old loops, now blocked, may be traced. Some of the inserted windows are very late. In the south part of the inner doorway was a door opening into a mural stair leading to the upper floor and to the rampart-walk of the curtain, which seems to have been continued through the gatehouse, over the outer portal. The present staircase is exterior, and modern. This is the whole of the original gatehouse, which, with the adjacent and very thick curtain, is probably the work of Earl Roger.