these chevaux de frise, as they may be called, only occur where the natural defences are weakest, while on the north-east side a single wall was considered sufficient, as the slope on those sides was so steep. On the other sides we find three walls and ditches; and at A two walls, which, by means of two cross ones, are converted into two defensive outworks. When these had been forced, the interior could only be reached by the entrance at B, the adjoining portions of the wall being of extra breadth, so as to accommodate a greater number of defenders standing on the top. There was a third entrance at c, in the middle of the side, protected by three walls and ditches. A projecting portion (D) prevents these defences being turned, while a fourth ditch, at some distance from the other three, encloses a space protected by these upright stones. The same arrangement occurs at A, where a similar space is also protected by an exterior ditch.

The whole arrangement shows no little skill on the part of the defence, as well as their faith in the efficacy of these pointed stones, found only in the enclosed spaces above described. Their small projections, and the sharpness of their points, would make the approach to them impossible to enemies, without carefully picking out their way while they were being attacked with missiles by the defenders standing on the tops of the walls.

Pennant, in his search for traces of Sarn Helen, discovered this British post as he terms it. He describes it as having the "usual fosses and vast ramparts of stones, with some remains of the facings of walls." But what struck him most "were two considerable spaces of ground thickly set with sharp pointed stones set upright in the earth, as if they had been meant to serve the use of chevaux du frise (sic) to impede the approach of an enemy." This was written more than a century ago; and with the exception of the greater dilapidation of the walls, it applies to the present condition of the work.

There is no other example in Wales of this kind of defence, unless Treceiri, in the same county, gives a similar example, but with a remarkable difference.

The original entrance to this strong city is well ascertained, and is strongly protected by a number of walls, for the particular arrangements of which the reader is referred to the Archeologia Cambrensis, vol. for 1867, pp. 66-78. But on the opposite side of this same mountain there are sheets (as they are best described) of flat stones placed at short intervals, en échelon, which can only be crossed with great care, as the stones are set wide enough from each other to admit and easily break a man's leg. In passing from one group to another a smooth, green path is traversed, which exposes the flanks to attack from above. At first sight these sheets may be mistaken for the ruins of walls; but the regularity with which they are placed at intervals, and their uniformity in size and arrangement, preclude this supposition. There cannot, in fact, be the smallest doubt as to their origin and design. It is true they have not been noticed before. Even our late lamented member, the Rev. W. Wynn Williams of Menaifron, who was probably as well acquainted with the same class of remains as Mr. Prichard of Dinam, was not aware of the existence of sheets of stones, and expressed his intention of examining them,--an intention, unfortunately, not carried out. One reason for their escaping notice may be that they do not occur on the usual route; and it was only by accident that the writer found his way thither; for on meeting a quarryman at the foot of the mountain, who was about to cross it, and being informed he could reach Treceiri by that way, he joined company with him until the man turned to the right on his way to his quarry. In wet weather (and wet may be considered the usual rule) the ascent is rendered much more difficult by the slippery surface of such smooth stones.

A still more remarkable example occurs in the Isle of Arran, off Galway, which has been well described by