The Roman fort at Lympne covered an area of some 10 or 11 acres. Its shape is irregular, and may be likened to a rectangle with a triangular extension on its highest side. On the east, north and west the walls still survive. But they have been strangely distorted in scattered ruin. The slope of Lympne hill is full of springs. These, causing great landslips, have thrown the walls down, probably at a very early date, and it is not easy now, except by excavation, to recover the lines of the original enceinte. The south wall of the fort, which is also the water front, presents a different problem. It has disappeared from the surface, and, as at Richborough, some writers have argued that it never existed, and that some now vanished cliff or lake supplied the defence. But the analogy of other Roman fortifications requires us to reject this idea, and, indeed, some traces of the missing wall ‘were discovered by the spade twelve years ago (P1. XI).⁸⁸
   The construction of the walls is that usual on the Saxon Shore—a core of ‘mortar and local sandstone rubble, a facing of the same limestone in neatly coursed blocks usually about 5 in. to 6 in. high and 7 in. to 8in. long, and regular :rows of bonding tiles running through from face to face. Their thickness is 12 ft. to 14 ft., and their height is still in places 20 ft. For the most part, no ‘special foundations seem to have been supplied; the walls were planted simply on the ground, with a set-off on each face. But Sir Victor Horsley found under the tower at the north-west corner, beneath the set-off course, at least 9 ft. of concrete containing stones of 12 in. to 18 in. diameter.
   External bastions strengthened the defence. They vary slightly in size and shape, but are generally semicircular and project 15 ft.; and with one exception they are tied into the main wall, and match it exactly in their courses of stone and tile. They are solid, but three of them at least have small internal chambers, 6 ft. high and 6 ft. to 8 ft. wide and long.
   The chief gate was a little below the middle of the east wall. If its distorted ruins

Fig.. 12.—LYMNE: THE CHIEF GATE
(From C. Roach Smith, Richborough, Reculver and Lynme, p. 252)

have been rightly understood by its discoverers, it had a single entrance 11 ft. wide, flanked with two solid towers or walls of masonry, which projected like semicircular bastions alike in front of the main wall and behind it. The whole was supported by a substantial platform of heavy stonework in two layers, which extended right across the gate.⁸⁹ We may compare the large gateway at Richborough where also

   ⁸⁸ Excavations by Sir Victor Horsley in 1894, briefly noticed in Athenaeum, 22 Sept. 1894, but apparently very properly described.
   ⁸⁹   C. R. Smith, Report, p. 12 and plate iii by Fairholt. The walling shown in the plate between the bastions has puzzled some critics, including Mr. Fox; it appears, however, to be due to landslip (Richborough, etc., p. 253). Fairholt’s original drawing is in the Soc. of Antiq. scrapbooks: it shows a gatepost hole not visible in the engraving.