From the keep a spur wall descends the slope, and was intended to cover the approach, as at Hawarden and Coningsborough. It evidently crossed the ditch, and formed part of the barbican. Thus the keep could be reached rapidly and directly by three ways, all narrow and well defended,—one from the outside by a postern, another from the main gate, and a third from the west ramparts. In substance the masonry and arrangement of this keep is clearly Norman, but the whole has been refitted, and no doubt refaced in the Perpendicular period.
Mention must be made of a very curious and early excavation in the main ward. On the surface, a few feet from the king’s seat, a flight of rock-cut steps descends nearly north-west, and at 70 feet distant is the mouth of a square shaft, lighting the passage below. Descending, thirty-three steps lead steeply down a gallery, 4 feet broad, with a hanging roof. A little way down, on the right, are traces of a cylindrical staircase, no doubt the original way in, but now destroyed, with the tower, in the base of which it no doubt was contained. At the foot of the stairs is a plain round-headed door-case, apparently of late Norman date. Beyond this the stairs recommence, and ten steps lower the descent ceases and the passage forks, a short branch running north, and another, a trifle longer, east. Before the fork, part of the passage is vaulted in fine-jointed ashlar, with two plain round-headed ribs. In the wall, on the right, is a round-headed recess for a lamp, and the commencement of another passage, also round-headed, but left as a mere recess. Above the fork opens the shaft, here seen to be a truncated pyramid, about 6 feet by 12 feet, and 30 feet deep. At the fork the salient is occupied by two small oblong cells, with pointed roofs. They communicate with each other and the passages by narrow lancet doorways. The excavation is now called the magazine, and no doubt was so used at the siege; but it is of Norman and early English date, and probably was intended for a cellar. The arrangements of the cells are scarcely suitable for a prison. The present entrance is clearly an addition, opened when the well staircase was disused.
There remain some exterior points to be noticed. Leaving the keep by its postern, and going north-west along the foot of the west face, the wall is seen evidently to be Norman, and near the centre of this front is the original Norman postern. There are upon the face of the wall two broad shallow pilaster strips, 8 feet broad by 6 inches projection, between which is a plain round-headed relieving arch, and below it a segmental-headed doorway, of 4 feet opening, without portcullis, but with a rebate for a door, and holes for two stout bars. This opens into a straight vaulted passage, about 5 feet broad, lofty, also round-headed, of excellent ashlar, and clearly Norman. It runs about 15 feet, and is then choked up. It possibly ends in a well-stair, and might readily be excavated. In later, probably Perpendicular times, this postern has been disused, and the door converted into a loop, and blocked with the usual window-steps within.
Following the base of the cliff along the north front, it is seen to have been carefully made good with masonry; and at the north-west angle, under what was Queen’s Tower, a large rift in the rock has been lined with ashlar, and spanned by a round-headed arch in good masonry. It looks like a large cavalry postern, but is merely a recess. At the foot of the talus on the west front, and about 180 yards outside the wall, are the remains of Swillington Tower, an outwork built by Thomas Earl of Lancaster, and in which he is said to have been imprisoned. About half of the basement remains. The tower was 46 feet square, with walls 10 feet 6 inches thick. It was intended to command the approach from the north, and was of great use during the siege as a flanking defence. Doubtless a double wall connected it with the main ward; but of this there is no trace.
It would seem that at Pontefract, as at many inland castles, a dam was thrown across a valley below the place, and thus provision made for defence and for the working of a mill. This seems to have been the case here below the northern front. The valley was converted into a lake, employed to feed two mills, of one of which traces remained in 1806, and the other, the lower mill, was removed in 1766, when the dam was levelled, and the pool converted into a meadow. Bubwith Bridge, no doubt, crossed this pool, at what is still called “the Wash.”
Nearly all traces of the Barbican are gone, but its memory and site are preserved in Barbican House, Row, and Garden, and there remains a fragment, probably of the lower gate, between Ass Hill and the Castle chain. There were two approaches, one from the town and one from the great church, which met in the outer ward of the Barbican. In front of the north entrance there still remains a good but late Tudor House, into the front of which has been inserted a grand old stone heater shield, bearing the three lions of England and a label of three points, carved in bold relief, a relic probably of the royal occupation of the castle. The style of the shield is early, and the blazon points to the eldest son of a king of England, before Edward III. introduced the lilies of France.
Boothroyd’s bird’s-eye view gives a general notion of the castle before it was destroyed. There were eight mural towers,—the Keep, the Red Tower, the Treasurer’s or Pix Tower, Swillington Tower in advance of the wall, the Queen’s Tower, the King’s Tower, Constable’s Tower, and the Gatehouse. All, save the keep, were rectangular, perhaps Norman. Of these only the keep and the ruins of Pix Tower are traceable: the rest, with the great hall, kitchens, and lodgings, were carefully removed by the Parliamentary contractor, though probably a few pounds spent in excavation would still show the basements, and establish a general plan.
Looking to the general evidence afforded by the remains of this castle, it is clear that it was a strong place in pre-Norman times; those who fortified it placing the mound at what was naturally its weakest point.
The greater part of the remaining masonry is Norman, and not improbably early. The enceinte wall, the buildings connected with it on the west platform, the rear wall of the platform, the old postern, the interior of the keep, and the magazine, all seem to be in substance Norman. Of the early English and Decorated periods very slight traces are left visible; but it is clear that under the House of Lancaster, in the Perpendicular period, much was added. Probably, the buildings on the north-east platform were constructed. St. Clement’s Chapel was rebuilt, Swillington Tower added, the keep refaced, and much done in repairing the chambers and staircases within.
Boothroyd, whose history, with all its imperfections, should be mentioned with respect, gives a copy of the account rendered for the destruction of the works, a plan of the siege operations, and a bird’s-eye view of the castle.