The jutting rock on which this relic is situated is called "Pen-y-cnwc"; a name probably having reference to the tumulus which prominently occupied its summit, the word "cnwc" signifying a bump or knob. The tumulus itself was styled "Barclodiad-y-Gawres"; a phrase well understood, in this part of Wales, to mean "the giantess' apronful". Strangers may find the spot more readily by inquiring for "Careg-yr-enwau" (the stone of names),—a modern appellation assigned to it in consequence of the names and initials carved by visitors on the remaining capstone of its central chamber. Whether this early receptacle of the dead was covered over by a mound of earth, or by a carnedd¹ of stones, has not been ascertained ; probably by a combination of both, in which stones predominated. Judging from existing traces, its base had a circumference of 240 feet; and although supposed to have been proportionably high, it has been reduced to its foundation. A wall, built for agricultural purposes, touches its northern boundary, and bears evidence of having derived much of its materials from the curious stonework of cells and passages recklessly destroyed. One slab amongst others, set edgewise in the face of this wall, was measured, and its dimensions were ascertained to be 6½ ft. by 4 ft. The entrance was from the north; and the passage connecting it with the interior was at least 35 ft. long by rather more than 3 ft. wide, its position and extent being marked by a depression in the ground, and by seven stones, which are all that remain of its side-walls and roof-supports. Of these, six are situated in a line to the left of the entrance; and on the opposite side, one, as if designedly spared to denote its width. Fronting the inner and southern extremity of this passage was the central chamber, the extent of which is not exactly traceable. The solitary roof-stone which remains measures 10 ft. in length by 5½ ft. in its widest part; its eastern and broadest end resting on two low but substantial supports, whilst its narrower extremity, towards the west, is sustained by the earth and debris of the tumulus. This ill-formed slab is at present the prominent feature of the remains, and having the characteristics of a cromlech on a small scale, has been noticed as such on the Ordnance Map. It is also the "Caregenwau" of the neighbouring peasantry. South of this central compartment was a curiously small pentagonal cell, measuring diagonally 3 ft. 3 ins. by 3 ft. 9 ins., the side-stones or walls of which are in their original state.

Note 1. It has been suggested as probable that the greater number of our cromlechs in Wales were covered over by heaps of stones or "carneddau", which would account, in some degree, for the disappearance of these coverings in so many instances. A pile of stones would be more useful and tempting to a farmer, when fencing and reclaiming his waste land, than a mound of earth. Treasure-seekers would commence the work of destruction. Remains of these sepulchral " carn- eddau" are to be seen on some of the highest of the Carnarvonshire hills, such as " Carnedd Dafydd", " Moel Siabod", the hills above Drws-y-Coed, etc.; but even on spots so remote as these I have not met with any which have not been partially scattered, and the roof- stones of their chambers turned over.

These tumuli, so prominently situated, must have been conspicuous objects when viewed off the coast, and might well have served as the last resting-places of vikings or rovers, whose remains, deposited here, would have been near to the element which had conveyed them to battle, and perhaps to fortune. Although "Barclodiad-y-Gawres" stands some 70 or 80 feet above the waters which restlessly chafe on the rocks beneath, the greensward or the stones which here covered the dead, whether natives or rovers, must have been wet with ocean spray whenever a storm set in from the southwest. It is, perhaps, doubtful whether there is anything in the construction of this tomb to mark it more particularly as the grave of a stranger, or to distinguish it especially from others of a somewhat similar character, which are met with in retired mountain districts twenty miles from the sea ; such as the highly interesting¹ one at Capel Garmon, above the Vale of Llanrwst. (Arch. Camb., April, 1856, p. 91.) It should be observed, however, that on the opposite side of a small creek or bay which extends inland on the south-eastern side of "Peny-cnwc", there is a small camp, such as, in Pembrokeshire, is called a "rath", designed probably for no other purpose than for the protection of one or two small vessels which might have been hauled ashore immediately under its defences. This inlet is open to storms and heavy seas from the south-west; but the rock on which the earthwork is situated projects into the little bay, and serves as a natural breakwater. On all sides the rock is precipitous, excepting where it was connected with the mainland, at which point a low rampart (now from 5 to 8 ft. high, and measuring transversely, at its base, about 24 ft.) has been thrown up, strengthened outwardly by a perpendicularly sided fosse, 9 ft. wide by 8 ft. deep, which completes its isolation and its defences. The almost triangular space within this enclosure measures, on two sides, from 50 to 58 yards; and on the third side, 33 yards. Towards the centre of it is a small mound of earth or stones, the original purpose of which has not been ascertained. This work may have been contrived by some adventurous rover for the double purpose of protecting his ship and booty whilst away himself in the interior of the island on some errand of pillage and devastation, and as a provision for his safe retreat if overtaken by adverse fortune. It is also possible that it may have served as the fortified residence of a native chieftain in post-Roman or mediaeval times. If supposed to have an early British origin, it must not he forgotten that apparently it contains no traces of circular hut-foundations. After all, an underground investigation might bring to light indications of a Roman occupation,—a research which at the commencement of the present year I fully hoped to have accomplished. The adjacent farm is called "Trecastell", Anglicé castle-town or castle-hamlet. Although "tref" sometimes signifies no more than a home or homestead, it may perhaps, in this instance, be accepted as implying that in early times a town or village existed near to this spot, which has long since been converted into stone walls.

Hugh Prichard.

Note 1. Interesting in many respects, but more particularly, as Mr. Barnwell has recently observed, because, in its present transitional state, half a tumulus and half a cromlech of large dimensions, with a capstone 14| ft. by upwards of 12 ft., it is so thoroughly illustrative of the sepulchral origin of cromlechs. The still incredulous should by all means visit it, distant about two miles and a half from Bettws-y-Coed.