The gatehouse is tolerably perfect on the ground floor. The upper story is in ruins. The portal, about 32 feet deep, is vaulted throughout at different heights and with arches of different curves. It was defended by a portcullis and a pair of gates. The outer portal is handsome and peculiar. It is of about 10 feet opening, and shoulder-headed, the shoulders being worked brackets. Above is a pointed arch of relief, and the tympanum between the two is composed of stones joggled together with great neatness. Above is a good flat-topped Decorated window of two lights trefoiled. Traces of the chain-holes for the bridge are seen in the spandrels of the portal. The gatehouse is evidently of two periods. All behind the portcullis groove is original, either late Norman or early English. The groove with all before it is late Decorated, probably of the age of the upper story of the keep. Much as the buildings of this castle have suffered, it is curious that the piers, counter-piers, and bridge-pits of the four bridges should remain quite perfect, and all of excellent ashlar. The inner, or pier from which the bridge dropped, is from 9 feet to 12 feet long, and the pit across which it dropped of 12 feet opening. The counter-pier, upon which the bridge dropped, is much longer, from 40 to 45 feet, and as this long and exposed causeway was but 12 feet broad, any body of enemies approaching by it would be placed at a great disadvantage. It is difficult to form an opinion upon the age of the earthworks of this castle. Either the Romans or the Normans might have laid out an earthwork on a rectangular plan, but when either people desired to construct a place of excessive strength, they employed masonry rather than earthworks. The Saxons and early English, on the other hand, though much given to employ defences of earth, and often upon an immense scale, are not known ever to have made them rectangular. What was the practice of the Romanised Britons, who, inheriting something of Roman arts and military rules, might also well have derived from their Celtic forefathers a taste for works in earth, is not known. Such a fortification as the present may possibly be in part their work. Of course, it is possible that the whole may have been the work of Robert Fursan, especially as, remarkable as it is, it is not named in Domesday nor any early record.