which a mask was formerly attached, through which a stream gushed into the basin below. Two or three steps led down to this _baptisterium_, where the cold bath was taken in the open air. This court and portico were paved in mosaic. 21. Apodyterium. 22. Frigidarium. 23. Tepidarium. These two rooms, in neither of which was there a bathing vessel, show that frequently rooms thus named were not intended for bathing, but simply to preserve two intermediate gradations of temperature, between the burning heat of the caldarium or laconicum and the open air. In fact, no trace of any contrivance for the introduction or reception of water has been found in No. 22. It was simply a cold chamber, cella frigidaria. Nor was the little chamber, 23, large enough to receive conveniently a bathing vessel; but seats of wood were found there for the convenience of those who had quitted the bath, and who came there to undergo the discipline of the strigil, and a minute process of purification and anointing. This room is not above twelve feet by six: the bath, therefore, could not have been calculated for the reception of more than one, or, at most, of two persons at once. Here the great question relative to the use of glass windows by the ancients was finally settled. This apartment was lighted by a window closed by a movable frame of wood, which, though converted into charcoal, still held, when it was found, four panes of glass about six inches square. A more elaborate and curious glass window was found at a later period in the public baths. 24. Caldarium. It might, however, be employed at pleasure as a tepid or cold bath, when the weather was too cold for bathing in the open air. The suspensura caldariorum, as Vitruvius calls the hollow walls and floors raised upon pillars, are in remarkably good preservation. By means of these the whole apartment was entirely enveloped in flame, and might be easily raised to a most stifling temperature. We will, however, add that Vitruvius directs a bed of clay mixed with hair to be laid between the pillars and the pavement; and some tradition of this custom may be imagined to subsist, for the potters of the country, in some cases, work up wool with their clay, a practice unknown elsewhere, as we believe, in the art of pottery. The burning vapor passed out above the ceiling, gaining no entrance into the apartment. Air and light were admitted by two windows, one higher than the other. In one of these Mazois found a fragment of glass. The bathing-vessel, _e_, lined with stucco, and coated on the outside with marble, was fed by two cocks, which must have been very small, to judge from the space which they occupied. Hence, hot and cold water were supplied at pleasure; and it was only to fill the vessel with boiling water, and the whole apartment would be converted into one great vapor bath.