the light triplet pillars, with arches that go round the room. The appearance of mortice holes in the walls for the ends of beams, at the height of about the middle of the windows, led Camden to suppose that the cieling was projected from thence, and that an apartment above was lighted by the upper portion of the windows; but surely at a time when symmetry in building was so well cultivated, and where it appears to have been so successfully applied, such a ridiculous contrivance could not have taken place: more probably, as I conceive, from those mortices a support was derived for a lofty arched roof, or a gallery. {175} Eastward of the hall, is the curiosity of a leaning tower, a bulky fragment of the ruin between seventy and eighty feet in height, whose walls are of a prodigious thickness: it hangs nearly eleven feet out of the perpendicular, and is only held together by the strength of its cement. How or when this phenomenon happened no legend informs us; but it has remained in this state many centuries. As the adjoining towers, and all the standing parts of the ruin, remain perpendicular, the cause must have arisen from a local failure of the foundation: hence I am of opinion, that a solution of the phenomenon may be found in the effects of a mine, and which probably took place during the long siege which Hugh le Despenser sustained in this castle in the time of Edward the Second. Near this part of the ruin a place is shewn as the mint, with two furnaces for melting metal. From this chamber we ascended a spiral staircase to the corridor, still in very good preservation, which, lighted by small windows, and passing round the principal court, formed a communication with the different apartments. The external view of the western entrance of the ruin, with its ponderous circular towers venerably shaded with ivy, is remarkably striking; and, with the remains of its drawbridge and defending outwork, may be considered as the most entire part of the ruin. An artificial mound some distance off, but within the works of the castle, was most likely used for exploratory purposes. From the great plan of this castle, and there being no direct evidence to the contrary, its foundation has been attributed to the Romans; and some ingenious arguments have been adduced to prove, that it was their Bullaum Silurum. But it sufficiently appears, that no considerable part of the present fortress was built by them, as the predatory army of Rhys Tycan took and rased Caerphilly castle in 1221. The best supported opinion is that of the Hon. Daines Barrington, who attributes the present erection to Edward the First.—Caerphilly has lately increased from an obscure village to a well-built little town; and the respectable appearance of its two inns may be in a great measure dated from the great increase of the visitants of the castle. {177}