The position of Montgomery Castle is formed by nature, and needed only to be seen to be recognised as very suitable for a border fortress. A narrow and lofty ridge of rock, lying nearly north and south, rises abruptly between two valleys; that to the east very suitable for the town which has sprung up within it; that to the west, narrower and equally steep, but rising on its further side to far loftier elevations, one of which is crowned by the encampment of Frydd-Faldwin, and the other, rather lower and more to the south, is occupied by a much smaller work of very different, and, probably, English type. The ridge is in length about 500 yards, and that part of its summit occupied by the castle and its works about 330 yards. Its greatest breadth is about 60 yards, of which the castle may occupy about 40 yards. The ridge is reached from the south-east, or town quarter, by a steep road, but towards the north it terminates in a sharp point, whence a very steep slope falls to the top of a cliff of rock, the whole height being perhaps 250 feet. The contiguous sides are also very steep, so much so that a stone set rolling from the top does not stop till it reaches the gardens and orchards 150 feet or more below. The parish church stands within the town upon an eminence opposite to, but lower than the castle. The view from the ridge extends over the plain of the Severn to the Welsh mountains in one direction, and to the Shropshire hills in the other. Below, in the direction of Chirbury, Offa’s Dyke may be discerned about a mile distant, the castle being upon its outer or Welsh side. The ridge is traversed by three ditches quarried in the rock, and dividing it into four platforms, which formed the four wards of the castle, each of which appears to have been enclosed by walls or palisades, and connected with the others by bridges of timber. The northern and strongest ward is also by some feet the highest ward, and formed the donjon, or citadel; keep there was certainly none. This ward was about 52 yards north and south, by 32 yards east and west. It is nearly rectangular, and its four faces are scarped and revetted, that portion of the wall being about 15 feet high. Upon this stood the curtain and the outer walls of the contained buildings. A considerable building stood on the west face, along the southern half of which a wall remains about 9 thick, which supported a basement and two upper floors, of which holes for the joists are seen. In the first floor is a recess with a window and two garderobes, with shafts in the wall. The window was pointed, and that is all that can be ascertained. In the upper floor is also a garderobe. Connected with this fragment of wall is half a horseshoe tower, 30 feet in diameter, projected 25 feet from the wall. This also had a basement, and two upper floors, but little of it now remains. In the outer side of its wall, to the north, are the remains of another garderobe shaft, and, high up, part of a straight mural staircase. From hence, northwards, is a mere curtain. On the three other faces only the revetment, or traces of it, remain. A heap of rubbish at the south end may be the foundation of the gatehouse. There is no trace of a well, hall, or chapel. The only ashlar remaining is a sort of quoin in the substance of the great wall, and in a window jamb in the horseshoe tower. The masonry is rude but coursed rubble.
Between this and the third ward was a ditch about 20 yards broad, dammed up at its lower or west end to form a pond, probably for the benefit of the castle cattle. There is no trace of a bridge of masonry here. The third ward is a regular oblong, carefully scarped, and probably revetted all round. Only the eastern face is visible, and there the wall is about 20 feet high, with the remains of three half-round solid buttresses, 6 feet in diameter, and something like traces of two more. There is no masonry above the surface of the platform, which is about 40 yards north and south by 35 yards east and west. A ditch of about 8 yards broad divided it from the second ward.
The second ward is roughly oval, about 50 yards long by 30 yards broad, and its southern end is occupied by a rocky mound. There is no trace of masonry connected with this ward, which may have been defended with timber. The ditch dividing this from the first ward is not above 4 yards or 5 yards broad, and very irregular, the depression being apparently in part natural.
The first ward is smaller than the rest, irregular and rocky. It bears traces of dry walling, and upon its platform are the foundations of a rectangular building, and at the south end of a sort of tower, indicated only by a heap of earth. It may have been about 25 yards by 15 yards. It may be that the first and second wards were merely natural platforms palisaded with timber brétasches, as they are called in the close roll, but the third and fourth wards were certainly walled and must have been strong. To the north, 20 feet to 30 feet beyond, below the north ward, is a level triangular platform of turf about 60 yards in the side, and protected by a light bank on which may have been a fence. This platform, by nature so strong, was probably intended for a pasture for cattle. At the other or south end is also a platform, 20 feet or 10 feet below the level of the first ward, and now occupied by a cottage. Here was the entrance, which probably was covered by some sort of tower, protected in advance by a palisade.
Scanty as are the remains of this castle, it may safely be concluded that they present no masonry of Norman date, whether early or late. The plan of the works is, of course, dictated by the natural outline of the rock, and it therein resembles Bere, though that is a little earlier than Henry III., and Dolforwyn, probably a work of that king’s reign. Henry III.’s border castles had no keep. They were mostly mere enclosures, the curtains being set rather thick with towers.
Altogether this castle, as it now stands, seems to be the new castle referred to by Henry III., and built early in his reign, nor is there any trace of any earlier work, although there is every reason to believe that such there was, and that it or they stood on this site. Moreover, as to these earlier works, tempting as is the position, there is nothing in the way of earthwork upon it which can be safely attributed either to the Welsh or the English. That Earl Roger constructed a castle of some kind at or near the present site is certain, and it is very improbable indeed that with a position so convenient, and made so strong by nature, he should have selected any other, nor is there in the immediate neighbourhood any trace of any other work likely to have been constructed later than the Conquest.
It has been mentioned that the borough of Montgomery had licence, in the reign of Henry III., to enclose the town within a wall and ditch, and although gates and ditches do not necessarily imply the more costly addition of a wall of masonry, it is certain that the town was fortified, and Mr. Sandford’s view of 1610 shows a wall including an area somewhat wider than the present town, and, what is very unusual, this wall includes the castle, instead of the castle forming a part of the circuit. No traces of any wall of masonry are known to exist, but outside it, to the west, and at the foot of the slopes of the castle, is a bank with a ditch, looking very much like a local dyke of the age of that of Offa, but which is reputed to be what remains of the town wall. The curious thing is that this dyke, which, while opposite to the castle is commanded and therefore strengthened by the castle rock, passes southwards along the steep slope of an opposite hill, and is there commanded from the outside, and would be utterly untenable. This is often the case with Offa’s dyke, which, though laid out with a bank and outer ditch, was really rather a boundary than intended at every point for defence. Its defence was obviously the general fear of the Mercian power, rather than the apprehension of an armed force at every point along its line. It may be that the town wall was placed upon this bank, but the bank itself seems older and part of an earlier boundary.
For the plan and section of the castle and castle rock which accompany this paper, and add materially to any little value that it may possess, the author and his readers are indebted to the skill and kindness of Mr. Mikleburgh, of Montgomery.