of small black pebbles from the beach. Bonding tiles, such as appear in the other Roman forts of the Saxon Shore, and in much other Roman walling, seem to be absent. Boys, noticing much tile about the Roman debris of the place, suggested that bonding courses had existed in the upper parts of the walls, which are now destroyed. But these courses, if used at all, would hardly have been omitted entirely from the lower 8 ft. or 10 ft. of walling. And they should at any rate occur on the inner face of the wall, which Mr. G. Dowker in 1877 and Major Gordon Home in 1927 found fairly perfect. Apart from this seeming lack of bonding tiles, the masonry is not unlike that of the other forts of the Saxon Shore. Major Home’s excavation showed that the wall was backed by a sandy bank, contemporary with it, but of unascertained size. Little is known as to the gates. A slight inward trend of the south wall and a gap about half-way along it were taken by Dowker to indicate a southern entrance; and a discovery in 1931 proved the existence of a buried gate here. Others have conjectured eastern and western gates, and at the supposed site of the eastern is a large, loose jambstone. But without excavation no certainty can be attained. Gough records a surrounding ditch. There are now no surface vestiges of it, but a great part of the land-surface outside the walls has been lowered several feet since the Roman period.
   Of the buildings of the fort, next to nothing is known. The ruins of the Saxon church are built wholly or largely of Roman material, but the two tall columns taken from the church, and now preserved in the cathedral precincts at Canterbury, are Saxon rather than Roman.²²  Apart from the defences, the only Roman masonry now visible, or seen recently, in situ consists of two lengths of rough and rather indeterminate stone walling found by Major Home a short distance to the north-west of the presumed south entrance of the fort; and a well found in 1923 near the centre, a short distance south of the churchyard. This well was 3 ft. 4 in. in diameter, was lined with flint rubble (in which foot-holes with brick lintels were placed at regular intervals on opposite sides), and to the depth of upwards of 15 ft. was filled with sand containing few and insignificant Roman remains. Roman buildings seen by our predecessors and now destroyed add little to the story. Three items can alone be cited, and none of them is quite satisfactory. (1) Battely (p. 35, sec. 30) mentions that, fifteen years before he wrote his book—that is, towards the end of the seventeenth century—a fall of cliff revealed brick substructions and arches, and also vestiges of mosaic flooring. This naturally suggests hypocausts with dwelling-rooms above them. Unfortunately, the position of the find is uncertain. Battely does not state it precisely, and merely treats the remains as generally connected with the fort. Harris, writing a little after him in 1719, describes the spot as ‘half a mile westerly of Reculver towards Hearn.’ Mr. Fox, writing on Reculver, thought it was near the fort but outside it, since it was discovered before the sea had attacked the fort. He conjecturally places it between the north wall of the fort and the shore, and suggests that it may have been the bath-house which often occurs outside Roman forts. It is not perhaps necessary to consider it an external building. The sea had already attacked the northern rampart in Battely’s time, and if
   ²² For the church see C. R. Peers, Arch. lxxvii, 241. The columns have evoked much discussion: see C. R. Smith, Coll. Ant. vi, 222; Richborough, etc., p. 197, and Proc. Soc. Antiq. 2nd ser. i, 369; Gent. Mag. 1861 (i), 148; Fox, Builder, 20 Oct. 1900, and Arch. Journ. liii, 353; Baldwin Brown, Arts in Early England, vii, 258; A. W. Clapham, Saxon Archit. (Oxford, 1930).