The great barbican is a very remarkable work, not unlike a similar defence at Castle Rising, which there, however, is confined to an earthwork, whereas here the defences are both in earthwork and masonry. The work is irregular in its shape, concave where it rests upon the counterscarp of the main ditch,—on its other faces convex. In depth, or from north to south, it measures about 56 yards; in breadth, 74 yards. It is defended by a ditch of moderate dimensions, which communicates with the main ditch at each end, but is scarce a quarter of its depth. The curtain wall is very irregular in outline, the west side being made up of about four straight pans of masonry, meeting at various angles. The front is a long straight wall, and the east side, also straight, meets it at a right angle. The curtain of this outwork is about 5 feet thick, and from 15 feet to 18 feet high. Upon it are four mural towers; three, drum-shaped, are 10 feet in internal diameter, with entrances at the ground floor from the court. In height they rise a story above the curtain. They are looped to the field, have timber floors, and their staircases seem to have been of timber. One tower is angular, and has evidently been the kitchen, part of the oven and the outline of the large fireplace being still visible.
The outer gatehouse is placed at the south-eastern corner of the barbican, terminating its wall and resting upon the counterscarp of the main ditch. It is of rude construction, composed of two solid masses or cheeks of wall, 46 feet long and 9 feet thick, rounded at the outer end. Most of the northern wall is gone, and with it the roof or vault of the entrance passage, and the details of the gateway. This sort of entrance between the high and long walls is found at Conisborough and Berkhampstead. The other or inner end of the barbican wall, at the south-west corner, ends abruptly, and probably descended a little way down the counterscarp of the main ditch, so as to prevent an entrance. In the wall is a recess 6 feet 4 inches broad and 3 feet deep, which may have been a public garderobe, similar to that in the wall of the middle ward of Corfe Castle.
The lesser barbican covers the south entrance, the way to which does not, however, appear to have lain across it. It is of earth only,—that is, has no walls, but is of great strength; in shape, and, indeed, in function, it is a lunette. Its diameter, in the direction from the gate outwards, is 20 yards, and its breadth 100 yards. It rests upon the counterscarp of the main ditch, here about 20 yards broad. It has, also, on its front and sides, a ditch of its own about 20 yards broad, and communicating with the main ditch. In parts this work seems to have been revetted. It was connected with the main work or body of the castle by a narrow bridge, which sprang from its eastern angle, and of which traces remain. There does not appear to have been any outer gatehouse at this end; probably the approach was by a boat across the ditch, or by a plank bridge upon trestles.
Besides the above-described works there is another and much larger, though less strong, earthwork, which covers the whole east face of the castle, as well as the flanks of its barbicans. It is about 30 yards broad, and outside the main ditch, on which it rests, and upon the ditches of the barbicans. It had also, towards the field, a ditch of its own, now converted into a hollow road. The approach to either gate of the castle, from the east, lay across this outwork, which was probably stoutly palisaded, and intended to give a moderately safe refuge to the peasantry and their cattle, during invasions from the Welsh.
It is evident from the general design, position, and capacity of this castle, that it was not intended, like Caerphilly, for the accommodation of a great military baron, nor like Morlais or Castell Coch for a mere military port. Its central part is nothing but a fortified area, like Framlingham, at which place, however, there seems to have been, as at St. Briavels, a keep. Here certainly there was none. It is and always was a mere shell, composed of a strong and lofty curtain, mural towers, and a gatehouse. There do not appear to have been any of the usual buildings contained within a large castle. The accommodation within the walls seems to have consisted in timber-sheds, resting on stone foundations, and built against the walls. In this way a very large number of soldiers might have been accommodated.
Moreover, the capacity for lodging a large body of troops was not confined to the inner ward. The outworks are also spacious and strong, and could contain a small army, or the élite, or perhaps the cavalry, of a large one, besides affording accommodation for a considerable herd and flock of cattle and sheep. Such a place would be secure against any ordinary inroad of the Welsh, and supported, as it would be, by the equally strong, though less capacious, castles of Grosmount and Skenfrith, not to mention Chepstow and Caerleon, which were not always in friendly hands, not only would the entrance of a Welsh army into Gwent be difficult, but its retreat, once within the line, would be almost impossible.
As to the age of White Castle, it is difficult to speak positively. The earthworks are not at all of an English, still less of a Welsh, character. Nor, as a general rule, did the Normans ever throw up earthworks on so immense a scale. The masonry is probably of the age of John, or, perhaps, even of that of Henry III., contemporary with Skenfrith, or perhaps a trifle later, and the earthworks may be of the same date. They are certainly designed to suit a castle in masonry, not the mere palisaded residence of a Mercian Thane. They do not at all resemble the earthworks connected with Offa’s Dyke, or those known to be the work of Æthelflæd or Eadweard. On the whole, it seems more probable that, looking to the extreme importance of the position, the early Norman invaders exceeded the usual character of their earthworks, and designed and executed those which are still seen.
The name White Castle,—in Welsh, “Castell Gwyn,”—is said to be derived from Gwyn ap Gwaethvoed, a brother of the reputed founder of Skenfrith, and one of the Welsh lords of this district. If so, he must have held it during the troubles preceding and subsequent to the death of Harold, but any works that he may have found or thrown up were probably effaced by those of the present castle. The castle was held, with those of Grosmount and Skenfrith, sometimes by the De Breose family, sometimes by the Crown, and, finally, by the Dukes of Lancaster. No doubt it shared in the general neglect of the Border castles; that is, of such as were not the seat of any baronial family. In the time of James I. it was a ruin, and had been so beyond the memory of man.