Chapter V: Part 5

It will be observed that the eastern wall, though very thick, is pierced by no galleries, and, with the exception of two windows, is absolutely solid from base to summit. Probably the object was, by placing this mass of firm masonry on the solid ground, to give support to the other three sides, and thus prevent them from sliding down the slope of the mound, as was the case with some much lighter and later buildings in a similar position at Cardiff.

It has been stated that the keep stands upon the south-eastern slope of the mound, consequently there is to the west, and in front of its entrance, nearly the whole table summit. This was enclosed by a circular wall, like a shell-keep, about 25 feet high, which, springing from the south-eastern angle of the keep, seems to have been carried round the mound, commencing at about half its height, until it reached the north-east angle of the keep, at which junction there seems to have been a gateway. Of this circular wall about one-half, either actual or in foundations, remains. The fragment of wall, about 5 feet thick and 20 feet high, is evidently of the date of the keep, the same thin red tiles being used in its chalk masonry. Also in this wall is an original garderobe, apparently of three stages: one at the ward level, one half-way up, and one on the battlements; the three seem to have united in a common shaft, the vent of which is seen outside the base of the wall.

The space thus enclosed by the keep and the circular curtain was the inner ward. How the main door of the keep was reached does not appear. There was not the usual fore-building, as at Norwich or Rochester; the circumscribing wall and the lofty position of the keep rendered this unnecessary. A row of small holes at the door level may have belonged to a sort of timber landing, to be reached by a flight of steps of the same material. There is said to have been a well in this ward, about 6 feet from the west wall of the keep, south of the door.

There remain two wards to be described. The line of the _enceinte_ of the castle seems to have been much as follows: Commencing at the angle of Castle Street and Quarry Street, where the site of the present “King’s Head” inn was probably marked by a tower, the west front took the line of Quarry Street, past the great gateway, to the tower containing a postern, which still marks the south-west angle of the enclosure. Thence the wall passed east till it reached the boundary of the extra-parochial ground, whence, in the line of that boundary, it probably took the curve of the counterscarp of the main ditch, in the direction of the Bowling Green Cottages. There, on the platform at the end of the causeway, was no doubt a barbican covering the direct approach to the keep. Thence the wall seems to have been continued along the line of the ditch, shown by the curve of Castle Street, until it again reached Quarry Street, thus enclosing what corresponds tolerably well to the area of the castle at the sale of 1612, which is described as 5 acres, 3 roods, 10 perches; or nearly six acres. Of this area, the part north of the great gateway was shut off by a curtain, parts of which remain, and which seems to have run up the mound to the _enceinte_ of the inner ward. In this, the middle ward, stood the hall and principal buildings, as is clear from the considerable, though fragmentary, Norman walls still to be seen, two of which, forming two sides of a large chamber, are very perfect, and one is pierced by a very perfect Norman window recess and loop.

What stood in the area south of the gateway, into which the postern led, is not known. It long contained the gardens of the governor of the gaol, when the castle was employed for that ignoble use, and in it is the celebrated well connected with the caverns.

The great gate in Quarry Street, though large, is at present a mere opening, perhaps of the age of Henry III., in an older, and probably Norman, curtain. Whether it was connected with the gatehouse is uncertain, but there is, as already stated, an indication in the masonry that such was the case. Also the portcullis groove is large, and so heavy a grate could not have been worked without a chamber above carrying a winch. Outside the gate are two buttresses, which have a late Norman aspect. One is nearly perfect, the other has been replaced in brick, but probably upon the old base. The south-west angle of the Quarry Street front is marked by the not inconsiderable remains of the postern tower, adjacent wall, and a large buttress, all pretty clearly late Norman. Towards the north and east the walls are entirely removed, but in these quarters the line of the ditch affords a clue to the original boundary. The _enceinte_ thus laid down measures about 535 yards; its greatest north and south diameter, 170 yards; and east and west, 140 yards. The Quarry Street front is straight and 138 yards long, with the gate nearly in the middle. From the postern tower to the end of the keep causeway is 213 yards, and thence to the “King’s Head” angle, 184 yards.

Captain James has detected traces of a line of wall parallel to, and about 30 yards south of, the High Street, which may not improbably have been the boundary of an enclosed area appended to the castle, as may the extra-parochial lands on the east and south-east, but the actually defended area of the castle seems to have been as above described.

It may further be observed that Quarry Street, which runs along the foot of the west wall of the castle and lay between it and the river, seems to have been defended on that side by the low cliff and talus already mentioned, supported probably by the retaining wall, of which traces and the jamb of a gate remain; while there is a tradition of a gate crossing the street near the postern of the castle, which probably guarded the approach to the town from the south, the only quarter from which a hostile approach would be apprehended.

Not only are the remains of the domestic buildings of a late Norman character, but among the repairs of the hall, in the reign of Henry III., two of the piers are mentioned as out of the perpendicular, a tolerably conclusive evidence that the hall resembled Oakham and Leicester and was of Norman date. Altogether, it is sufficiently evident that the whole area of the castle was enclosed by its original founder, and is not later than the middle of the twelfth century, the detached fragments of walls and buildings being in substance of the same date with the keep. The castle is in St. Mary’s parish, bordered on the north-east by Trinity and on the east by the extra-parochial plot, the origin of which is not known. Probably the _enceinte_ took in the whole of the residence of the royal Saxon owners.

Those who have supposed that the _enceinte_ of the castle extended to the present High Street have regarded the two well-known crypts remaining there as proofs of this extent. One of these, on the south side, the writer has not been able to visit; but the other, exactly opposite the former, and about 160 yards from the keep, he has examined, and it is said that the two are of the same age and dimensions and very nearly alike.

The north crypt, beneath the Angel Hotel, is a rectangular chamber, 31 feet north and south, by 19 feet east and west, and divided into two aisles and six bays by two central piers. The piers are plain cylinders, 18 inches diameter and 5 feet 6 inches high. The bases are now concealed; the piers are without caps and quite plain. The roof is vaulted, and 10 feet 3 inches from the floor to the cornice, groined and ribbed. The arches are drop and pointed, the ribs chamfered and springing from carved corbels in the wall. At the south end of each aisle is a window recess, converging and rising to a loop at the street level. The entrance is by a narrow drop-arched door, opening from a rising passage, vaulted, with hanging ribs. This opens into a small chamber, north of, and 4 or 5 feet above, the vault, whence another narrow door probably led up to the ground level. The date of the crypt seems to be of the thirteenth century, and it is quite clear that it never was prolonged southwards towards the other crypt, and was always and only lighted, as now, from the street level. In all probability this was the cellar of some considerable hostel, situate, as now, in the High Street, which, in the twelfth and thirteenth centuries, as now, was probably the main thoroughfare of the town.

In the immediate neighbourhood of the castle, and, indeed, running under the southern edge of its _enceinte_, are the celebrated caverns which have recently been explored, and for the first time correctly planned by Capt. James, whose excellent account of them is well known. These caverns are excavated in the chalk, which forms a cliff south of the town, and from the base of which they are entered. The chalk here dips northerly, at about ten degrees, and the hardest bed and the most suitable for building purposes lies at the base of the cliff and is about 6 or 7 feet thick. This bed has been largely quarried by open “patching” in the broken ground south of the cliff, which, indeed, is apparently artificial, and produced by these excavations, and it was only when the bed became too deep for that mode of working that the quarrymen had recourse to mining operations. These later works are, in general plan, composed of a gallery parallel to, and a few feet within, the face of the cliff, from which, at a right angle, eight parallel stalls are carried north-eastwards. The extreme points of the excavation are a little over 55 yards north-west and south-east, by 32 yards north-east and south-west; but the area actually excavated is only about 1,150 yards, and the cubical contents about 2,330 yards. The plan alone would show that they were opened for quarries. But, besides this, although much solid chalk has been removed, the excavation is nearly choked up with the immense quantity of “small” produced by unskilful working, and through which narrow paths are left to get at the face of the work. It is evident that this is not _débris_ brought in, nor caused by the fall of the roof, which is remarkably sound. It is simply broken chalk which has been thrown back as the miners proceeded, and remains undisturbed. The character and presence of this rubbish not only shows that the excavation was a quarry, but that it never was used for anything else, neither as a granary, nor a cellar, nor for human habitation, for nowhere has it been cleared away so as to set any part of the cavities free for such purposes.

Nearly in the deepest part of the working, about 60 feet below the surface, the caverns have been pierced by a well, sunk, it is said, in the garden of the governor of the old gaol. This well has been carried through the caverns, if to water, probably to a depth of another hundred feet, but it has subsequently been covered over with plank, at the level of the cavern floor, and so now remains. The pipe of the well above is very rugged, as though sunk by unskilled workmen—perhaps convicts,—and is stained, as though used as a cesspool. Also, for many yards around, the rubbish, elsewhere of pure white, is dark and foul, as though, failing to reach water, or afterwards disused for that purpose, the well had been employed as a receptacle for all the filth of the prison.

As to the age of these quarries, it is not easy to form a sound opinion. They have been supposed to be British, and various uses have been found for them, quite at variance with the appearances they present. The only argument for such an origin has been overlooked. The town is in Woking hundred, and Woking, like Wokey in Mendip, may be a corruption of the British “Ogof” (fovea), a cave. But Woking was never the name of the town, and the material evidence of the caverns does not favour this theory. They are certainly not British: the plan of the workings excludes this view. No doubt they might be Roman, but there are no traces of Roman buildings in the neighbourhood, and chalk, even hard chalk, is too plentiful all along the ridge to be carried hence to any great distance. The most probable supposition seems to me to be that they were opened by the builders of the Norman castle, who used chalk largely for their inner, and, indeed, for much of their more exposed, work. The quarries have no communication with any part of the castle. Where they infringe upon its borders, they are far too deep to have been employed against it during a mediæval siege.

HARLECH CASTLE, MERIONETH.

DESCRIPTION.

The Castle of Harlech occupies a bold and rugged headland of rock which juts forward upon the coast-line of Merioneth over the broad alluvial plain known as Morfa Harlech, near to its southern and narrower extremity. Six centuries back, when the Traeth was an estuary, and the waves may have washed the foot of the rock, Harlech, as now Criccaieth, was probably accessible by water,—a circumstance likely to have governed its founder in his selection of the site. Although scarcely 200 feet above the sea level, and connected with a much higher background, the rock of Harlech is nevertheless a very striking object, and by the extreme boldness of its outline and its almost isolated position does justice to its very significant appellation. It commands one of the most remarkable prospects in Britain. Before it is the Bay of Caernarvon with its vast sweep of sandy shore, contained on the right by Snowdon and its subordinate peaks; whence the high land, after rising into the elevations of Carn Madryn, Carn Bodfuan, and Yr Eifl, gradually subsides into the Bay of Aberdaron and the Sound and Isle of Bardsey. Caernarvon and Conway are fortresses more ornate in character and of larger area; but they are not equal to Harlech in natural strength and in grandeur of position; nor is, in these respects, Beaumaris itself, though placed in the very eye of the Snowdon group, by any means its superior.

Harlech is a concentric castle of the Edwardian type, and of that type a simple and excellent example. It is composed of a central four-sided ward contained within four lofty curtains, and capped at each angle by a drum-tower of three-quarter projection. In the centre of the landward or eastern side is the great gatehouse; opposite to which, built against the curtain, are the remains of the hall and domestic buildings; and contiguous to them, against the north side, is the chapel.

The main or _Inner Ward_, thus composed and occupied, stands within the second or middle ward, which resembles it generally in plan, save that the four corners are not symmetrical, one being merely rounded, two others capped by more or less of three-quarter bastions, and the fourth rounded on one face and fashioned as a bastion on the other. In the centre of the south side is a half-round smaller bastion, corbelled out from the retaining wall below; and in the centre of the north side are two others, also small, between which is the postern of this middle ward. In the east face, opposite the great gatehouse, are two “tourelles,” or round bartizan turrets, corbelled out from the wall; and parts of a small low gatehouse, which contained the outer gate.

Wyman & Sons, Gᵗ. Queen Sᵗ. London.]

This middle ward is narrow and of unequal breadth, varying from 8 feet to 30 feet. It is rather below the level of the inner ward, and the ground outside it is from 10 feet to 15 feet lower still; and its walls are revetments crested with a parapet which seems to have ranged from 6 feet to 12 feet in height; in the latter case having a rampart-walk reached by open steps. The several bastions seem to have risen a little higher than the parapet, and to have contained each a low chamber, probably with a flat roof. This ward is protected on the east and south sides by a broad and deep dry ditch, quarried in the rock, and running out until it ends on the cliff. The other two sides are covered by an outer ward of considerable breadth, but composed for the most part of steep slopes and abrupt ledges of rock. A part of this ward towards the west or sea front contains a long passage which ascends by a lower traverse from a water-gate at the foot of the rock, resting partly upon a shelf of rock, and which by a second and upper traverse reaches the postern of the middle ward.

Passing into details, the court of the inner ward is about 164 feet north and south, by 132 feet east and west. The opposite sides are not quite equal, nor are its angles right angles, though nearly so. The curtains are about 40 feet high; that to the west is 10 feet thick, the others are 11 feet. The parapet was 3 feet thick, and the rear wall 2 feet, leaving 5 feet to 6 feet for the walk. The two western towers are circular, and 34 feet diameter, having three-fourths of their circumference exposed outside. Within, the gorge wall fills up the angle of meeting of the curtains and contains the entrance-door. The basement chamber is below the inner ward level and circular. The first floor, at the ward level, is polygonal, as are the two upper floors. None are vaulted, and the basement has neither loops nor stairs of access. Each of these two towers has a well-stair at its junction with the western curtain, lighted by five loops placed one over the other in the hollow angle between the tower and the curtain, outside. The stairs ascend 20 feet above the tower, in a round turret, battlemented on small corbels. Each turret has a door upon the tower roof. The staircases commence at the first floor, on or level with the inner ward, and open on each floor, but not upon the ramparts of the curtain. The upper floor has fireplaces with hoods.

Outside, these towers rise from the ground without slope or cordon; two string-courses, however, mark the level of the two upper floors. The stairs are broken away, and the upper rooms inaccessible; but certain exterior loops show the existence of two tiers of small chambers (no doubt garderobes) in the north and south curtains, where they join the towers. Moreover, on the outside of each of these curtains, next to the tower, is a broad, flat buttress, thrown out to give space and support to these chambers and to contain the sewer-shaft from them. On the north wall, the buttress is of good ashlar of the age of the tower. On the south wall, it is of rude, inferior work, as though an addition. It may have been rebuilt. In the north curtain there seems to be a third chamber at a lower level. The drain here is not seen; on the south face it is open. Where these towers meet the rampart-walk they block it up; a sort of gallery is, therefore, thrown out on corbels, across the angle, and thus, as at Conway, the rampart-walk is carried on.

The two eastern towers resemble the others in general features and dimensions, but differ in details. Their basements have one loop towards the middle ward, and their first floor, at the inner ward level, is an irregular pentagon in plan, one angle being square. The doors are in the gorge wall, but do not lead direct into the tower, only into the staircase. In the south-east tower, a stair ascends in the northern wall, curving with it, and forks, the right branch leading to the second floor of the tower, from which alone, by a trap and descending ladder, the first floor and basement were accessible. This floor, like all the rest, was of timber, and from it, on the west side, a second stair commences, and curving with the wall, and having a small garderobe by the way, ascends to the ramparts of the south curtain. Reverting to the lower stair, the branch to the left opens upon the inner face of the east curtain, and ascends by a narrow, open stair, supported on corbels, across the gorge wall of the tower, and up the inner face of the south curtain to its ramparts. The roof and ramparts of the tower are reached by an exterior stair from the rampart of the east curtain. A loop in the hollow between the junction of this tower with the south curtain marks the place of the garderobe already mentioned. Above it was a second upon the battlements of the tower, and at the base of the wall is a large flat-topped sewer descending from the two. The south-east tower bears the name of Mortimer, the south-west that of Bronwen, the fair-bosomed, sister of Brân the Blessed.

The north-east, the debtors’ or armourers’, tower has a door in the gorge opening on the left upon a well-stair, 8 feet diameter, which ascends to the second floor only, from which the first floor and basement were reached by a trap and ladder. The second floor is seven-sided, those below cylindrical. As in the south-east tower, an independent stair led from the second floor to the ramparts of the curtain, and upon this curved stair is a garderobe, the loop of which is seen at the junction of the tower with the north curtain, and the mouth or vent at the ground level. The roof of this tower, like the other, is reached from the walls by an external stair. These two towers, having no well-stairs to the roof, have no subordinate turrets. That all these four towers had flat roofs is pretty clear from the position of two corbels in each, evidently intended to carry hammer beams or struts to the one main beam which crossed the aperture, and was thus rendered capable of carrying great weight.

The great gatehouse is 80 feet broad and 54 feet deep, besides which it has two half-round projections in the front and two three-quarter projecting stair-turrets 24 feet diameter at the outer angles of the rear, the former flanking the entrance, the latter communicating with each floor and the ramparts. The entrance passage, 54 feet long by 8 feet broad, is much mutilated, but seems to have had an exterior drawbridge, two grates, folding doors, and a grate at the inner front. The entrance portal has within it a “machecoule,” or _meurtrière_,—that is, an opening from the chamber above—and behind this a portcullis. Then follows a passage 11 feet long, crossed by two ribs, a second portcullis, and a portal arch, upon which rests the west wall of the chapel. Then follows another passage, 20 feet long, entered by gates opening towards the inner ward, and crossed by five broad ribs, with four open spaces. At the end of this is a third portcullis, the groove for which is now closed above at a level too low to allow the grate to be lifted to the height of a cart, while in the arch above is a square cavity or “machecoule.” It would seem that while the wall was rising it was decided not to use these grooves, and that the hole was intended to take the place of the grate as a defence. Beyond this is the inner portal, which, like the outer, has no rebate for a door. In the front division of this long entrance, between the two outer grates, are two loops from the side lodges, which are entered by two doors placed near to the inner end. This passage was covered over with boards, the flooring of the rooms above, and which rested upon the stone ribs. Here, as is often the case, the portcullis groove stops from 1 foot to 18 inches above the door sill, showing that the spikes at the lower end of the grate were of this length. This long entrance passage is further lengthened by the addition of two unequal piers to its internal face. They are blocks of masonry 10 feet thick. That on the south or left had a door whence a narrow staircase of two flights ascended to the front floor. The pier on the right is of less breadth, and was only an abutment to support the arch which connected the two and contained and continued the entrance passage, and on which was the landing at the stair-head.

The basement of the gatehouse is at the ground level. On each side of the passage are two chambers, those in front occupying the half-round projection and looped to the field. They are entered from the chambers in the rear, which are rectangular, having shoulder-headed doors from the passage and into the well-stairs. The northern chamber has a fireplace in the south-east angle. The two southern chambers communicate through a large arch, the northern through a doorway only. There are also two upper floors, divided as these below, and reached by the two large well-stairs. There are spacious and handsome rooms, two on each floor, with large windows of two lights in the western or larger rooms, and in all are fireplaces with stone hoods. The eastern rooms below are half circles in plan; and above, are polygonal. Between the lateral rooms and over the entrance passage are two narrow chambers unequally divided by a cross wall. The eastern is an oratory, with a small, pointed east window over the entrance gate of the castle, and near it, in the south wall, is a piscina, which is in the cill of a small window opening into a small mural chamber, a vestry. There is a similar chamber, but without the window, in the north wall. Both rooms are entered from the oratory. As at York, Chepstow, and elsewhere, this oratory served also as a portcullis chamber, and the floor was of wood, with traps to allow the passage of the grates when lifted. The grates were suspended from the vault above, as is still seen. The other and larger chamber, placed over the western part of the passage, had also a wooden floor. It had a west window of two lights over the inner portal, and north of this a round-headed doorway. The portcullis, if lifted, would have blocked this entrance, and therefore when the door was opened it was stopped. The machecoule is seen in the window seat. The upper chambers are not accessible, but they seem similar to those below, and there is a second oratory above the first, with a smaller east window—a very unusual arrangement. This floor communicates laterally with the ramparts of the curtain, and at the junction on each side is a mural garderobe. On the south side a mural stair descends to two chambers at different levels, both in the curtain wall. On the north side the arrangement is rather different. There, the mural garderobes are supported in part by a projection at the first-floor level, corbelled out in the angle between the gatehouse and the curtain, outside, and the vent was probably between the corbels. Above, at the rampart level, half the thickness of the wall is occupied by a garderobe chamber, of which the side is broken down. Several of the chimney shafts are collected in a central group, each shaft having a bold capital with a plain roll moulding.

The domestic buildings were placed against the curtain on the west side of the inner ward. The kitchen is thought to have been at the north end, including within its limits the basement of the north-west tower. It is, however, more probable that this was the withdrawing room, placed between the hall and the chapel. A gloomy corner, no doubt, but the state rooms were evidently in the gatehouse. The kitchen would scarcely have been designed originally between the hall and the chapel. The cross wall, still standing, but which looks either modern or rebuilt, formed the north end of the hall, and the recesses in the west wall of the curtain carried the hammer beams of its open roof. In this wall are the remains of a large fireplace, of which the hood is gone, and the lower part has recently been rebuilt. On either side are the broken apertures for two windows, and in the wall, near its south end, a segmental-headed door, now walled up, but evidently a postern. There are also near this two small windows, one of which seems to have lighted the gallery and the other the space below it. Of the position of the gallery there can be no doubt, but the wall behind it, forming the south end of the hall, and now removed, had no bond either into the curtain or into the east wall. Most of this east wall, the inner wall of the hall, is gone. The hall was 30 feet broad. The roof seems to have been lofty, and part of the weather moulding of its gutter remains along the west wall. On the floor, in the north-west corner of the hall, has been built a large oven of stone, the lining of which is much burnt. It probably was inserted when the castle was used as a prison.

South of the hall is a considerable space extending to the gorge wall of Bronwen Tower, and in the east wall of this space are remains of a door and two windows. It is probable that the kitchen was here, in the rear of the gallery, and that a row of corbels outside the east wall carried a lean-to building attached to it and near this; against the south wall is a rectangular pit, the underground storey of some building now removed. If the kitchen was at this end, the hall fireplace was a little below the dais, a very probable position.

The chapel, a later building, was placed against the north wall. Its east wall and pointed window remain. The south wall is gone. In the centre of the north curtain is a segmental-arched doorway, evidently a postern, and nearly opposite to that of the middle ward. It is much mutilated, and does not seem to have had a portcullis. The wall east of it is pierced by three loops, 4 feet above the ground level. There was at least one loop westward of the postern. The well was in the north-east angle of the court. It has recently been opened a few feet down.

The _Middle Ward_ contains little of interest. On the north side it is 15 feet broad, and hence, between its two roundels, 10 feet apart, opened the postern, 8 feet wide, now walled up. On the west front, the ward is 27 feet broad and forms a noble terrace overlooking the sea and commanding the approach from the water-gate. The hall had windows looking this way, and upon it opened the hall postern. Towards the south end a few steps descended about 10 feet into the south-west bastion. Probably there was a cross wall here with a doorway. Turning the south-west corner, the ground again rises to a door in a wall which crosses the south terrace near its west end. This side of the wall has a central half-round bastion, the broken parapet of which shows traces of a loop and of a garderobe. On the remaining or eastern side is the great entrance. Here the gateway, which crowns a low salient, is flanked by two roundels. The portal is broken down, and it does not now appear how this was connected with the inner gatehouse. Probably the short distance between the two was arched over and had lateral doorways into the middle ward. From the inner gate, twenty steps descended to the bridge, so that no horse or carriage could have entered this way.

The defences beyond the middle ward are the ditch, the outer ward, and the water-gates and passage. The ditch covers only the east and south, the two landward sides. It is quarried in the rock, and is about 60 feet broad and was 20 feet deep, with vertical sides. Its scarp is the revetment wall of the middle ward, and the counterscarp, where the rock was broken, is also lined with masonry. The ditch runs out at either end upon the shelving face of the rock. Across it, to the main entrance, led a bridge upon which it is said there were two openings with drawbridges. The whole is now a solid causeway.

Although the castle stands upon a promontory of rock, there is a broken shelving space between its wall and an actual cliff in which the rock terminates below, and it is this space, which lies to the west and north, which has been enclosed as the outer ward, the containing wall of which crowns the cliff and where necessary is supported by a revetment. This outer wall begins below the north-east bastion of the middle ward, whence a door with steps seems to have led down about 10 feet to its ramparts. It is at that point a very stout wall, about 14 feet high, with a parapet on the western face, thus defending the ditch and main bridge from an enemy who might be in possession of the outer ward and be disposed to turn the eastern flank. It is probable, however, that the wall had a double parapet, for lower down, where the wall faces the north, the parapet is on that face. Near the bastion there seems to have been a door in this wall, giving a passage from the outer ward to the ditch. Lower down, where the wall stands on the cliff, it is thinner, and in parts much broken away. Still lower it is more perfect and much stronger, and where it turns the north-west corner of the rock, opposite the railway station, it is of great thickness, and has a rampart wall and parapet towards the sea, above the level of which it is about 30 feet; near this point is the lower water-gate, a regular postern, in a small rectangular shoulder in the wall. A roadway of about 5 yards or 6 yards long, cut in the rock, rises from the marsh 10 feet or 12 feet, and upon it, in front of the portal, was a drawbridge with a pit 12 feet deep, and within the portal a short shoulder-headed passage closed apparently by a door, but without any portcullis. Beyond this a flight of open stairs niched in the curtain ascended to an embattled platform over the gate. From the lower gate, the road leads up a rather steep passage formed partly by taking advantage of a shelf and partly by quarrying the rock, the outer side being protected by a wall 8 to 10 feet high and from 2 to 3 feet thick, and looped at about every 20 feet. As the inner side of the roadway is the irregular face of the cliff, it varies much in breadth, from 6 to 12 feet or more. This road, continually ascending, thus covers the whole seaward face of the castle rock, and at about 70 feet or 80 feet in height it terminates in the middle gate, which is about 20 feet below the base of the south-western bastion of the middle ward. Here, a shoulder in the rock is occupied by a second gatehouse, fortified as the first, with a drawbridge and a deep pit, which below has two arches, one for the discharge of water from the pit and the other, which may be merely to support the side wall of the gatehouse, but which may also be a sewer from the castle. Outside this gate is a platform which rakes the face of the wall of the passage below, while above and within the gate is a broad bastion, whence commences the second traverse. At this point, the end of the main ditch lies just below the bastion wall, and was reached from it by a small door and some steps, now gone.

The road now makes a complete turn, and commences a new traverse which rises much more gently than that below. When abreast of the mid-front of the castle it is supported by a retaining wall and two small square buttresses or buttress turrets, traces of which are seen upon a ledge of rock. Passing these, where the road comes opposite to the north-west bastion of the middle ward, it was crossed by a wall and doorway, of which traces remain, which divided the outer ward into two parts. Above this, the way turned eastward and ascended to the centre of the north front, where it reached the postern of the middle ward and there ended.

These are the whole of the works proper to the castle, but a few yards to the north of the rock a steep road has been cut by which men and horses could be led up from the castle landing-place to the village without entering the _enceinte_, though commanded from it.

No one acquainted with Caerphilly can visit Harlech without observing the close resemblance between the two castles, so far as regards the plan of the interior and middle wards. The court, rectangular, or nearly so, the absence of a keep, the drum-towers capping the four angles, the general character of the gatehouse and its position in the centre of one side, and the domestic buildings placed against the wall of the inner court, are peculiarities common to both. In each also the gatehouse is the grand feature of the building. Further, there is to be observed in both the excessive narrowness of the middle ward, its revetment rendering more than a parapet unnecessary, its slender and subordinate gatehouse, and its lateral postern opening direct through both wards. As Harlech did not need the outworks and exterior gate of Caerphilly, nor Caerphilly the water-gate of Harlech, here the resemblance ceases, but it is such as to justify the conclusion that Henry of Elfreton, who was the architect of Harlech, had studied Caerphilly, if indeed he was not also its architect.

The defences of Harlech seem calculated for protection against a surprise by the Welsh, who were probably as active as they were fearless. Hence the very lofty curtains, the long entrance bridge, the ascending steps to the main entrance, and the dimensions of the middle ward, too narrow to allow any considerable body of men to effect a lodgment there for an attack upon the inner ward, and the water-gates and covered way, in the construction of which the natural strength of the rock was enhanced by the occupation of its various points of vantage. Whether in the reign of Edward I. Morfa Harlech was more than a marsh is a question for a geologist to solve; but either by the shallow sea or by a canal cut across the low ground, it seems certain that in planning the castle Edward counted upon the means of reaching it by a quarter quite independent of the Welsh.

Although the general plan of Harlech is evidently the work of one mind and its execution generally of one date, there are some appearances in the work which show that alterations and additions were introduced affecting, not the general plan, but certain of its parts. It is evident that parts of the curtain have been thickened about 2 feet,—the north and south walls by additions inside; the west, on the outside. Also this thickening seems to have been decided upon when the walls were 30 feet high, as above that level they are of one mass and date. The exterior stair on the inner face of the great gatehouse was also an afterthought, and the doorway at its head clearly was not originally introduced. Besides this, the six windows on that front of the gatehouse, in the two upper floors, have been reduced in height by the insertion of a segmental arch between 2 feet and 3 feet below the original head; but the pattern is the same, and the masonry filling up the space seems of the date of the window, or very nearly so. These windows are of a peculiar pattern. Their two lights are trefoiled; and in the spandrels are also trefoils pierced. The mouldings are concave; and one is a small hollow, as in the early Perpendicular style. They must, however, be original.

The inference from these alterations seems to be, that Edward visited the castle when the works were far advanced, and the hall, gatehouse, and the lower part of the north, south, and west curtains built. The gatehouse curtain was probably always intended to be of its present height, as at Caerphilly. He ordered the other three curtains to be thickened and raised to the full height of the gatehouse curtain; to obey which order, the thickening was applied where possible on the inside, but where the hall prevented this, on the outside. The upper part of the walls so raised would, of course, be of one date, and solid. At the same time it was decided to make the rooms of the upper floors of the gatehouse those of state; and as the ways up the well-staircases were not thought suitable, a new and more direct staircase was built and a new door opened in the wall. The chapel in the inner ward seems a still later addition.

The character of the masonry throughout is exceedingly rough, as though hastily executed. It is rubble, and some of it very poor rubble indeed. The towers are of far better work than the curtains. The stones are larger, and their interstices filled in with more care. The ashlar is very good, but is sparingly used, and confined to the dressings, window-cases, chimney-hoods and heads, and a few of the more important doorways. The ordinary doors are mere openings to the walls, without rebates or chamfer, with shouldered heads of a rude character; and the sewer-openings, seen under the garderobes, have merely long stones for lintels. The masonry of the covered way and water-gates is also very inferior, and much of the side wall has, in consequence, slipped away from the rock.

The turret-heads of the gatehouse and two western towers have parapets projecting upon a corbel-table about 6 inches. There are no traces of holes for brattices; but upon the exterior of these two towers the putlog-holes are arranged in a spiral ascending form, east to north. In the north-west tower, on its east face, at the height of the old curtain, is a row of round holes about a foot apart, and from this level the spiral commences. It is pretty clear that having built the curtain, the masons here threw out a platform, and that the spiral round, by which the materials were raised for the upper part of the tower, began here. The tower of Coucy was scaffolded in the same way. There is throughout the building a remarkable absence of vaulting. It was confined to the oratory and to parts of the entrance passage.

The bird’s-eye view here given is taken from the east or landward side, and shows the entrance, with its great gatehouse, here, as at Caerphilly, seen to cross and close the narrow outer ward. The northern postern is also shown.

The castle seems to have escaped the usual dismantling that followed upon the civil wars, and no part has been blown up. It has, however, been freely used as a quarry by the people around; and, with its iron and timber, much of its ashlar has been rudely detached and stolen. There is but little evidence of any material addition to, or alteration in, the work of Edward I., which is singular, seeing that the place was long the seat of an assize, and the judges lodged here. It was then also a prison, and the windows were heavily barred, the bars forming shallow cages in front of the windows, as in some of the Italian palaces. Any later work introduced for the judicial or prison arrangements has either fallen down or been removed. The quarry whence the castle was built is pointed out on the hill-side, a short distance to the south-east. Although the present castle certainly is not older than the reign of Edward I., probably about 1280, the Welsh claim to have been the founders of an older fortress on the same spot, called by them Caer Gollwyn, from Collwyn ab Tangno, a Welsh chief who lived A.D. 877. Possibly a spot so inviting might have been occupied by a camp; but all that is now seen, whether of earthwork or masonry, is evidently not older than the thirteenth century. In 1404 the castle is said to have been taken by Owen Glyndwr; and Margaret of Anjou was sheltered here in 1460, in memory of which event the south-east tower for some time bore her name. There does not seem to be any detailed account of the siege of 1468, when the governor was Dafydd ab Ievan ab Einion, the same who had received Queen Margaret, and whose boast it was, that as he had held a castle in France till all the old women in Wales had heard of it, so he would hold his Welsh trust till it had become equally well known in France. He seems to have redeemed his pledge by standing a long siege, and yielding at last, on honourable terms, to Sir Richard Herbert, the commander for Edward IV. Harlech was held for Charles I., and surrendered on articles to General Mytton in 1647. The borough seal represents a castle triple towered, but the design is evidently conventional. The first constable was Hugh de Wonkeslow, appointed about 1283 by Edward I.; the last was the late W. W. E. Wynne, Esq., of Peniarth, as good a man as he was an eminent antiquary.

HASTINGS CASTLE, SUSSEX.

The Rape of Hastings is the most eastern of the six divisions of the county of Sussex which are supposed to derive their somewhat peculiar name from the early occupation of the district by emigrants from Jutland. The Rape of Hastings, probably, is so called from its principal town, which was also the seat of its lords both before and after the Norman Conquest. Like the other rapes, it had its river and its forest, its castle and its castelry, often designated as an Honour. The river, the Rother, is the common boundary of Kent and Sussex, and joins the sea at Rye. The forest, long since disafforested and enclosed, is represented by frequent patches of woodland, scattered over the least fertile parts of the district, and by the numerous and well-timbered parks by which it is characterised. The town once possessed a small port, now silted up. It was situate at the mouth of a stream, which still brings down its inconsiderable tribute, flowing at the foot and to the west of the castle hill.

The origin and etymology of the name of Hastings are lost in obscurity. It is so uncommon, that it has been supposed to come from Haesten, the celebrated pirate and Danish Viking, who infested the southern coasts of England and the valley of the Seine in the ninth century,—a period when warriors of Northern descent gave their names to their possessions, instead of, like their descendants of two centuries later, reversing the practice. Mr. Lower, familiar with Sussex topography, has suggested that the small stream of the Asten, a few miles west of Hastings, may play some part in its etymology, and that Hasting, or the Haesten-Ceaster of the English, may have stood upon its margin. It has also been regarded as the seat of the Hastinges, a tribe said to have been warred upon by Offa of Mercia in 792. Another not less interesting point connected with the place is the fact that it gave name to the ancient family who long ago bore the title of Pembroke, and still bear that of Huntingdon.

The castle occupies the narrow, acute, and very steep extremity of a ridge of the Wealden formation, which here terminates abruptly to the seaward at a height of about 180 feet above the seashore, and seems intended by nature for a defensive position. The ridge is a spur from a range of larger and more elevated character which extends from the high ground of Battle, by Ore, south-eastwards to the sea, and has to its west and south a large tract of broken and highly-fertile picturesque land known as the hop garden of Sussex, and within which are the well-known seats of Hurstmonceaux, Ashburnham, Battle, and Crowhurst,—a tract which presents to the sea an open frontier of about twenty-two miles from Hastings to the marsh of Pevensey, defended anciently by the two castles of those names, and more recently by twenty or thirty martello towers, in which our fathers, three quarters of a century ago, placed an expensive confidence.

The ridge and promontory of Hastings remain unencumbered by modern buildings, and are occupied only by the castle and its outworks. The older part of the town, with its two parish churches of St. Clement and All Saints, occupies a deep valley to the east of the castle, while round the nose of the rock the remains of the old port have been superseded by the new town, a fashionable watering-place. Much of this town stands within the parish of St. Mary de Castro, the parish church having been the chapel of the castle, and collegiate. The town seems to have been partially walled, and certainly had four gates. Hastings, though now affording no accommodation for shipping, is still a cinque port, and its ancient consideration is attested by the fact that in 1229, Seaford, Pevensey, Bulverhythe, Hidney, Iham, Beaksbourne, Greenhythe, and Northeye, were its subordinate members, and _tamquam membra_ were Rye and Winchelsea.

The castle is composed of two wards, separated by a formidable ditch cut through the sandstone rock about 400 feet, across the ridge, with a breadth of 60 feet, and to a depth of 30 to 40 feet. The inner ward lies to the west of this ditch, between it and the point of the promontory. The outer ward lies to the east, and is again protected by an outer and considerable, though less clearly-defined, ditch of greater length than the former, and also crossing the ridge. There was also another ditch cut as a sort of step in the steep northern slope, and covering that side of the place, and chiefly intended for the defences of the inner ward. Between the east end of this ditch and the north end of the great cross ditch was the original entrance to the inner ward, the approach to which wound up the steep north-western slope of the rock.

The _Inner Ward_ is in figure nearly a right-angled triangle, containing about 1½ acres, the north side 154 yards, and the east side 87 yards in length, while the hypothenuse curves inwards, reducing the breadth of the ward at one point to 16 yards. The cliff which forms the boundary has been either scarped or has fallen away, so that it is at present precipitous, and its face has been patched with modern masonry and brickwork. It is said that this side was once straight, and has been cut away to make room for the gasworks and a crescent of houses below, so as to diminish the castle area. This may be so, but the encroachment cannot have been considerable. A wall for defence on this front could never have been needed: probably it was only a low parapet.