By G. Poulett Scrape, Esq., M.P. 61 



two feet in width, through which the hot air from the furnace was 

 intended to circulate. The covering-slabs laid over these openings 

 from pillar to pillar, are not of tile, but rough plank stone, and 

 their lower surfaces still retained a thick coating of charcoal-soot. 

 Upon these slabs was laid a layer of concrete, or terrass, composed 

 of lime mortar and pounded brick, about six inches in thickness. 

 This had been every-where reddened by the action of the fire 

 beneath. Upon it, in three of the chambers at least, rested a 

 pavement of well-jointed and squared stone-slabs from two to three 

 inches in thickness. In the other rooms the surface of the concrete 

 itself formed the floor. In some parts it had been destroyed. The 

 uppsr floors in all the rooms were on the same level and about 3 

 feet 6 inches above the basement one, the pillars being 3 feet in 

 height. 



Within the two hypocausts A and B (which measure respectively 

 14 feet by 7 with the recesses) there were found many hol- 

 low flue-pipes of terra cotta, from eight or ten to six inches 

 square, and from one foot to two in length. Some of these were 

 entire and stood in their proper position upright and in contact 

 with the walls, others were broken in fragments. These flue-tiles 

 have patterns rudely scored upon them on three sides as if drawn 

 with a comb or toothed instrument, the patterns generally varying 

 on the difierent sides, (see plate x.) Two opposite sides have 

 usually one or two square holes in them, evidently intended for 

 the entrance and exit of the hot air, which rose through these flues 

 no doubt into the bath-rooms above the hypocaust. And for their 

 admission through the floor there were open spaces of five or six 

 inches in width left in some of the rooms between their 

 pavement and the walls. It is possible that the scoring may 

 have been intended only to make the mortar adhere the better : 

 yet the ornamentation of only three sides of these flue-pipes seems 

 to indicate that they were carried above the pavement of the upper 

 bath-rooms high enough to be visible. Perhaps they were even 

 continued as high as the ceiling. It is not indeed easy to under- 

 stand how the smoke of the fuel burning in the furnace was con- 

 veyed away, as no other kind of flues were found. But as these 



