Under-floor heated rooms would be seen, along with mosaics and painted wall-plaster as a sign of 'status', something to which any upwardly mobile Romano-Briton would aspire.
Under-floor heating is by means of 'hypocausts'. These depend upon heated air from an external furnace being circulated under a floor and vented by flues up through the walls. This system means that the floor becomes warm, as do the walls. In bath-house systems, the need was for swift and efficient circulation of the hot air, so the floor was carried on small pillars.
Our hypocaust consists of large blocks of built masonry supporting the floor, with hot air channels between. Because of the pattern of channels, this type is sometimes known as a 'Union-Jack' hypocaust. It is suggested that they acted as a sort of 'night-storage' heater. The fire need not be as great as with the bath-house hypocausts, saving on fuel, and the masonry blocks would give off a more gentle radiant heat. This would seem to be appropriate for a use of the room either as a bedroom or as a winter dining room.
Heat was transferred from under the floor was by means of square 'box tiles' set into the wall. Photographs from the original Sparsholt records show the size and location of several of these flues and the Roman tendency to build symmetrically led to further locations being selected around the hypocaust.
Unfortunately, we have no evidence from Roman walls of how these flues exited the buildings. If the walls were of masonry to their full height, the box tiles could have continued up through the walls and exited out of the roof. The problem becomes greater if we are dealing with a half-timbered building, with flues exiting at the top of the walls. This would mean that they are in direct contact with the timbers of the roof structure, which may pose a fire risk. It seemed to us that another approach was needed.
The basis of this came from the box-tiles. Most box tiles have solid sides with holes at either end but, in many excavated examples, a number have been found with holes cut in the side. This suggests that the heat was meant to be able to travel two ways. So, we have constructed the flues so that they are joined together by a horizontal run of box tiles, to a common exit in the open air. Is it right? Well, we can't be 100% sure and just in case we are wrong the vertical flues can, with some simple unblocking, be vented out of the top of the walls, however that might be.
One different feature about the Sparsholt villa, on which the ground floor plan of the Butser Roman villa was based, was the presence of a 'fireplace'. The evidence for this was the presence of a recess in the wall lined with tiles and with a tile apron in front and 'cheeks' to either side. There was evidence of burning around this area. So what is it ? The most complete examples have evidence of a canopy over the apron area.. One possibility is that they might have been a recess for a shrine. The evidence that burning was involved would not contradict use as a shrine where offerings could be burned. However the positioning on the ground, the presence of a canopy, the apron in front and the large amounts of burning and associated food debris in some cases, all suggests that we are dealing with a fireplace or hearth. It could be a low level hearth burning charcoal or possibly wood, either directly on the apron or in a brazier, and could have been used for cooking or heating food or for warmth. This may or may not have needed a vent. Alternatively it might be a proper fireplace with a fire set at the back and a vent to act as a proper chimney.
Created 15 November 2003 - Updated 15 November 2003