The principal discovery was a building 217 ft. long (fig. 23), and for the most part about 50 ft. wide, containing a range of rooms, with a corridor 8-9 ft. broad on each side, and a projecting room at each end of the eastern corridor. The walls were of flint, ragstone and tufa, roughly set in mortar, and about 22 in. thick. The rooms were mostly small; their walls were unpaved, and no signs of hypocaust or mosaic were noticed. The exterior room (B) at the north end of the east corridor was more carefully examined than the rest of the building. It yielded coloured wall-plaster, tiles and potsherds, including embossed Samian, and was considered by the excavators to be the only apartment which showed signs of inhabitation. Outside its north wall was a rubbish pit with tiles, fragments of glass vessels and of window glass, Upchurch and Castor ware, much Samian, some of embossed and other pieces stamped SECVNDINI, OFPARO, OFSECVND, OFCEN, two bits of white marble, some bronze objects, including part of a ligula, hair pins, tweezers, nail-cleaners and the like, and two ‘small brass’ coins of Domitian. A hole dug at A also yielded potsherds, tiles and a bronze stylus. This was not the whole building. About 6o yds. eastwards traces of two

Fig. 25. Plan etc ., of supposed Mithraic Chamber discovered at Burham, Kent
(From Proc. Soc. Antiq. xvi, 109)

 other rooms were found, one of which contained a tessellated pavement in sandstone and chalk; in it many tesserae, potsherds, a bone spindle - whorl, and a ‘middle brass’ of Vespasian were picked up. These rooms were thought to be connected with the north end of the excavated building, but the intervening foundations were not at all clear. Further small remains, including a bronze finger-ring and hairpin, were found in a well 12 ft. deep, 30 yds. south-west of the tessellated pavement, while many tiles and potsherds, a mortarium stamped. . . NVS, iron horseshoes, a fluted blue glass bead and ‘large brass’ coins—one each of Domitian, Pius, Marcus, and Lucilla—occurred in the immediate neighbourhood. Lastly, an isolated piece of strong 

walling, 6 ft. long and 4 ft. thick, was found to the south of Boxted Farm, on the other side of the Upchurch Road and east of Breech Lane, perhaps a bit of boundary wall or outbuilding.¹⁵
   It is impossible to reconstruct the scheme of these various buildings. The block, which was more or less completely excavated (fig. 23), is certainly not the whole, and probably not even the principal part, of the villa. The tessellated floor which lay to the east of it seems to indicate a better residence, not yet uncovered. But this was not traced further; even its site has not been accurately recorded, and its precise connection with the other remains is not clear. If the excavators were right in thinking it continuous with them, we must suppose a range of rooms extending over 200 ft. But it has all the appearance of an independent building; the traces of masonry between it and the
   ¹⁵ Payne, Arch. Cant. xv, 104, Coll. Cant. 61—69, and Soc. Antiq. Proc. ix, 1882, 162—3; roof tiles and fragments of tessellation in Maidstone Museum.