as taken, in consequence of a breac aving been effected by means which requires some faith to credit on e testimony of local traditions id manuscripts. According to ich accounts, a battering-ram was orked by one thousand men, ad suspended to a frame, composed twenty large oaks. The breach as made in the depth of a dark ight, and king Edward escaped in te habit of a Welsh peasant. The ore effectually to disguise himself, e assisted with great eagerness to ile wood on the tremendously large res that lighted the besiegers in attering the castle. Local authoities assert, probably with some deree of poetical amplification, that ne hundred teams were employed > supply wood for those vast fires. The Welsh are said to have assisted he besiegers from all quarters, at a roper opportunity. Edward made is escape from every danger, and hrough the dark and stormy night, went on till he came to the parish of Llangonoyd, twenty miles westward, where he hired himself as a cowherd or shepherd, at a farm, still known by so singular a circumstance. After having been there for some time, but how long is not precisely ascertained; the farmer, finding him but an awkard and ignorant fellow, dismissed him. Such is the colouring of one account: but another story, in manuscript, relates, that the farmer knew who he was, and befriended him as long as he could. From Llangonoyd he went to Neath Abbey, whence he issued a proclamation, ordering his subjects to take the queen, with other particulars, which are to be found in Rymer's Fœdera. The Spencers were taken in their castle, where prodigious quantities of salt and fresh provisi

Of live cattle,

ons were found. there were lodged within the castle walls, two thousand fat oxen, twelve thousand cows, twenty-five thousand calves, thirty thousand fat sheep, six hundred draught horses, and a sufficient number of carts for them; two thousand fat hogs, two hundred tons of French wine, forty tons of cyder and wine, the produce of their own estates, with wheat enough to make bread for two thou. sand men for four years. In one of the towers, every apartment was crammed full of salt. Under this tower was a furnace for smelting iron, hot masses of which had been thrown by engines on the be siegers, who, when they had got possession of the castle, let out the fused iron from the furnace, and threw water upon it. This occasioned a most dreadful explosion, that rent the tower in two, and destroyed the salt. What stands of the tower at present, is that which overhangs its base. The subsequent fate of the two Hugh Spencers, father and son, is too well known to need a record on this occasion. Hugh Spencer, the grandson, however, with his faithful garrison, found means to destroy, very unexpectedly, a considerable number of the besiegers, and leading his men to the breach, was able to prevent others from entering. Presuming on this success, young Spencer succeeded in destroying his enemies within, and procuring tolerable terms, by which he was permitted to remain in possession of his castle and his estates, together with the lordship of Glamorgan. His son, Thomas Spencer, succeeded him. The next in the catalogue was a second Thomas Spencer; the last, and if possible, the worst, of this tyrannical

marshes, with ruinous walls of wonderful thickness, and a tower kept up for prisoners.

In the first volume of the Archæ ologia, published by the antiquarian society, there is a paper by the late Daines Barrington, at that time one of the judges on the circuit of North Wales. In this communication, he offers some reasons for supposing that Caerphilly castle was built by Edward I. on the ground of the probability, that as he had thought it necessary to construct the castles of Conway and Caernarvon, for the purpose of controlling the northern inhabitants of the principality, he might also have erected other castles in South Wales for the same purpose. I believe that the reputation of the author, and the ingenious reasoning of the paper, are generally considered as having set the question at rest; for it is attributed to Edward I. in most modern publications, on this authority specifically, without the slightest hint of suspicions or uncertainty. But I apprehend that a closer inquiry into the subject would have led that acute and learned antiquary into a train of observation, not altogether consistent with his hypothesis, and have induced him at least to doubt. Glamorgan was one of those petty sovereignties, called lordship's marchers Its lords were its sovereigns. They had their parliaments, their courts of justice, and their other offices, executive and jurisprudential, in which they, and not the king of England, were supreme. They exercised jura regalia, and did not hold of the crown, but per gladium, as their term was. They were generally, for their greater safety, in close alliance with the king of England, but not his subjects. This VOL. XLVIL

Edward I. had

distinction, however, is to be imderstood in reference to these lordships only; for with respect to their ba ronies and estates in England, they were to all intents and purposes subjects. King Edward had no jurisdiction at that time in Glamorgan. He could not possess an acre of land there, but as a subject to the lord of the country. It happened, indeed, in subsequent ages, that in consequence of intermarriages, the lordship of Glamorgan devolved on the king of England, and he in that case granted it to others on such terms as he thought proper, till, in the time of Henry VII. it was united to the crown of England, as were most of the other lordship's marchers in the same manner: and this assumption enabled Henry VIII. to incorporate the whole of Wales with England. united North Wales, by conquest, with the crown of England. He had done the same by that part of South Wales, which had been subject to the house of Dinevowr, and its princes; but those most powerful of the lordship's marchers, Glamorgan and Pembroke, in South Wales, with those of Denbigh and Flint, in North Wales, part of the lordships belonging to the carls of Chester, that of Shrewsbury, and possibly some others, continued independent of the crown of England till the time of Henry VIII, when the incorporation took place. These circumstances go to prove, that it could not have been Edward I. who built Caerphilly castle. We have already seen from historical documents, deduced from the Welsh authors, that John de Bruse built it in 1221; that after it had been taken, and of course partly ruined, it had afterwards been rebuilt in greater strength 3 L

than

than it had before possessed, by Ralph Mortimer; and in process of time was again greatly augmented and strengthened, by Hugh Spencer, the son, whose wealth appears, by all the accounts we have of him, to have been fully equal to such an undertaking and it may be supposed, with sufficient probability, that it was as great, and very possibly greater, than that of Edward I.

The present appearance of Caerphilly, fully accords with the ideas which ancient records inspire, of its strength, magnificence and extent. The area is entered between two dilapidated towers, and the interior view of this great gateway, between its mighty bastions, is as striking as any of the venerable structure. The circuit of the outer-works incloses a very large tract of ground, though the circumference, great as it is compared with that of fortresses in general, scarcely renders credible, the enormous provision related to have been thrown in by the younger Spen

cer.

The wall of the celebrated leaning tower, though but a fragment, is still between seventy and eighty feet, and of a prodigious thickness. It hangs eleven feet and a-half out of the perpendicular, and seems only to rest on one part of its south side. It appears as if held together, principally by the strength of its cement, which is of a tenacity unknown to the experience of modern masons. Its singularity is best observed by an interior examination, or from the moat underneath, whence the effect of its apparently falling mass is indeed stupendous. They show the mint close by this interesting part of the ruin, arched in a curious manner, with two furnaces for melting metal. These furnaces, likewise, dealt out dreadful vengeance on besiegers,

and were the means, according to the most plausible as well as best authenticated accounts, of placing the adjoining tower in that singular situation, to account for which, has given rise to so many conjectures of fancy, and so many tales of super. stition. From the mint there is an ascent to a long gallery, which communicated with the different apartments, and afforded a ready intercourse between the guards, who occupied the embattled towers. This corridor remains entire for the extent of from ninety to one hundred feet on the south side, except where the staircases have been destroyed, which circumstance prevents its being traversed: but the view from the extremity, along the vaulted passage, darkening as it recedes, realizes the awe inspired by the irra tional sublimity of chivalrous romance. The descent of the sallyport is tremendously steep. When once the force of the castle began to pour down, the alternative rested between victory and death. The declivity impelled the steps of those, whose fears might have paused on even ground: and there could be no retreat for the foremost, while the ranks behind were rushing to the conflict. The great hall is large and complete. It exhibits an august example of gothic grandeur, united with a considerable degree of elegance. This room is about seventy feet by thirty, and seventeen in height. It has large windows, and an ornamented chimneypiece, in masterly and scientific proportions, with clusters of pillars along the side walls. The north window of the chapel is not only perfect, but uncommonly light and elegant. The window of what my guide, in the spirit of modern re

finement,