In urging the conviction, that the hanging floors of these Roman buildings were meant to produce a comfortable warmth, rather than to generate steam, by having water sprinkled upon them, attention may be drawn to the thickness of their substance. At present, the floor of the principal apartment is nine inches thick, and when its upper surface was overlaid, as it no doubt was, with a tasteful concrete or mosaic pavement, it would be an inch or two more. It would require a very powerful furnace to raise this mass of matter to a considerable temperature. On the other hand, if the production of a genial and uniform warmth were the object in view, no contrivance could be more suitable. The heated air from a small furnace permeating the underground flues and the walls of a suite of apartments, and not passing off until, in its lengthened passage, it had given out the larger part of the warmth it had derived, would, in the lapse of some hours, give to the whole building a comfortable temperature, which it would not readily lose. Any inattention to the furnace, either by causing it to burn too fiercely or too feebly, would not be felt. The thickness of the floors would prevent the air from being scorched, and producing that disagreeable sensation which is experienced in rooms that are heated by the stoves in common use. It is not improbable that we may return to this method of warming our churches and public halls, even if we do not adopt it in our private buildings.[89] [Sidenote: METHOD OF WARMING BUILDINGS.] The door-ways of some of these apartments have been provided with double doors, probably for more effectually maintaining the warmth of the room. The masonry of those portions of the walls which are standing, is in an excellent state of preservation. In the angle near the buttress (A), the action of the trowel in giving the finishing touch to the pointing may be perceived. The walls rest upon two strong basement courses, the angle of the uppermost being bevelled off with a neat moulding. Some of the quoins of the door-ways consist of very large stones; one is six feet long, and is probably a ton in weight. This proves that it was not from lack of mechanical means that the interior buildings and walls of the stations were composed of small stones. More than one of the thresholds have a groove very roughly cut in them, apparently to allow of the egress of water. This has probably been done after the departure of the Romans and the general demolition of the buildings, by some houseless wanderers, who, having ‘camped’ in the ruin, were incommoded by the lodgement of rain on the floor.