inclosure surrounded by high walls, which on three sides follow the natural boundary of the hill. The principal buildings now traceable are over against the keep above the river, and also towards the angle formed by the turn in the course of the stream. The keep forms part of the wall towards the town, and the entrance to the castle appears to have been between it and the river to the left, so as to be completely commanded. Along the wall to the left, as you enter from the town above the Swale, are a few traces of turrets and other small buildings, now appropriated as pigeon and rabbit-houses; but in the angle over the river before mentioned, is a large room commanding a delightful view over both reaches of the stream, the weir below, and the opposite woods and hills; a window placed here indicates the good taste of the builders. Nearer the keep, the chapel is recognised by a trefoliated piscina, and between this and the last-mentioned window is a horribly dark and gloomy square cavity, intended I presume for a dungeon; its depth to my surprise is stated at only about thirteen or fourteen feet. Opposite the keep along the line of the river is a large building, entered by a semicircular Norman arch. The ground floor, which is lower than the present surface, appears from the corbels which remain to have been vaulted; it was lighted by square-headed loop-holes. In the room above are double Norman windows towards the river-wall, with little central pillars. This building appears to be of about the same date as the keep. At the further angle of this river-wall is a watch-tower. There do not appear to have been any buildings along the curve from hence to the keep.
The keep, the chief object of interest, is of oblong form, and of great height. Its dimensions are stated at fifty-four feet by forty-eight on the exterior, with an elevation of ninety- nine or a hundred feet. As however the walls are about eleven feet thick, and the interior dimensions of the ground- floor are thirty-four feet by nineteen, a length and breadth of about fifty-six and forty-one would probably be nearer the truth. The walls batter slightly, and are well built; from the style of execution of the doors and windows I should suppose the tower to be about thirty years earlier than that of Coningsburgh. The breadth on the side next the main entrance is less than that on the opposite side, by reason of a recession in the plan at the angle, as shewn in the view of the keep, given in Whittaker's Richmondshire, and in the
