Almost all of the Ceramic Building Material (CBM) from the Manor Farm and Temple sites at Marcham has been roofing material (tegulae and imbrices). The few fragments of floor, wall and hypocaust tile that have been recovered are typical of the background scatter of CBM found on most Romano-British sites, probably introduced as hard core for construction projects. It is nevertheless curious that there is no evidence for a bathhouse with hypocausts as one might well have been expected on such an important site.
With one exception, all of the roofing material excavated from Manor Farm was of mixed production and dispersed around the site making it very unlikely that it was the remains of any roof from the site and most probably again represents a typical background scatter of CBM hard core. The exception was the western end of the large Trench 2 building where a homogeneous assemblage of third/fourth century tegulae were encountered. These had two very distinctive features: they had two finger channels down the side of each flange where they had been smoothed by the tile-maker where normally one or none would be observed and they had a distinctive (Type 16) cutaway to assist the tiles meshing on the roof. Identical tegulae have been observed at Dorchester-on-Thames. There were few imbrices in this assemblage and so they do not represent the collapse of a roof. However mixed into the assemblage were animal jaw bones, pottery and lots of extremely small, almost certainly votive, coins which taken together have the appearance of a ritual deposition.
In contrast to Manor Farm, 0.8 tonnes of tegulae and imbrices, much of it of similar manufacture, was excavated on the Temple site which is almost certainly the collapse of the temple roof. However it was well churned and broken up and had probably moved from its original location. Although this seems a considerable quantity of tile it only represents 2% of the total amount that would have been required to tile the whole of the temple roof. Some of this tile appeared to have been deliberately coloured normally with a coloured slip (similar example were found on Manor Farm) showing that the roof would have been patterned with different colours to make it more impressive.
Most of the Temple tile came from four contexts. 6003, 6010 and 7015 contained homogeneous tegulae probably made in the late second century and 6008 tegulae made in the mid-third century demonstrating that there were at least two phases of roof construction or reconstruction on the temple. Interestingly, apart from two isolated examples occurring in Trench 26, none of these sorts of tegulae were found on the Manor farm site despite there being a good scatter of second and third century tiles across the site. This would suggest that the roof of the temple remained in situ until after the requirement for hard core had passed, in other words probably until after the main Romano-British occupation of the site had ended.
Peter Warry – Roman tile specialist