4. DOVER.⁶⁴

   Dover lies by the sea in the mouth of a long deep valley of the east Kent uplands. Chalk downs tower above it on either side. To the east are the cliffs of the precipitous Castle Hill. Opposite, with a slightly gentler ascent, rise the Western Heights. In the middle of the valley the tiny stream of the Dour makes its way to the sea. The shore is a bar of sand and shingle heaped through long years by the Channel currents, but never heaped high enough wholly to block the river mouth. Behind this bar, fed by the little river, spreads a lagoon, which, till the great expansion of recent years, provided a harbour for Dover. The mouth of the lagoon now opens beneath the seaward cliff of the Western Heights. In earlier days, before the shingle had gathered thick and pushed the river westwards, the lagoon and its mouth alike lay under the Castle Hill near the Burlington Hotel. The harbour can never have been deep or spacious. But it was sheltered by tall hills from northern and western winds. It could be approached from the sea without fear of shoal or quicksand. It was the one natural haven in the long range of chalk cliffs between Walmer and Hythe. It was the nearest point to the continent of Europe. It commanded the narrow seas at their narrowest point. All this was recognized, but perhaps not fully recognized, in Roman times. The true value of Dover was first seen in the Middle Ages. Indeed, its importance has steadily grown as the years have gone by. To-day Dover counts for very much more than we shall find it counting in Roman days.
   The Roman remains of Dover consist of a few traces of inhabitation beside the harbour in the valley, some cemeteries, and the remains of two lighthouses—the one still standing and the other almost vanished—on the two heights which overhang the valley (Fig. 11).
   (i) The traces of building in the valley represent no doubt the Roman settlement at Dover. They do not tell us much about it, but they are well worth recording and mapping (P1. VIII).
   (1) Near the parish church of St. Mary, walling has been noted at and close to the site of Biggin Gate. It contained ‘bricks of an extraordinary size, stone and other materials, singularly hard and firm.’ Something of the same sort was observed about 1831—2 a little east of this at Eastbrook.⁶⁵  Whether this wall was Roman or medieval, the description does not indicate.
   (2) At various times the remains of a Roman bath-building have been found under the west end of St. Mary’s church, and extending both north and south of it. In I 77 8—9 the Rev. John Lyon saw parts of a narrow passage and four rooms, one or perhaps two provided with hypocausts and flue-tiles and another with ducts for water. His very summary plan suggests two periods of Roman work. Some of the tiles bore the stamp CL. BR (P1. IX, No. 5), showing that the building (or at least its material) was officially connected with the classis Britannica or the Channel Fleet. During the restoration of the church in I 843, Canon Puckle found under the three western bays of the nave, at a
   ⁶⁴  For the structural remains of Roman Dover, see Wheeler, Arch. Journ. lxxxvi, 29; and Amos and Wheeler, Ibid., p. 47. Much information has kindly been supplied by Mr. E. G. J. Amos.
   65 Puckle, Arch. Cant. xx, 129; Batchellor, Guide to Dover (Dover, 1845), pp. 12—14. A solid wall found on the sea-front, between Butchery Gate and Snargate, must have been the sea-wall of the medieval town. A fragment of this wall can still be seen in Townwall Street, where it is carried on an old arch over the Dour.