depth of 4 ft. or 5 ft., ‘a fine open space of Roman concrete, the bottom of a system of baths which, entering at the south-west, crossed the nave and both aisles and passed out into the churchyard on the north side.’ Again, in 1897, during excavations carried out at the base of the west tower in connexion with the reparation of the structure, a hypocaust with soot and wood-ashes was discovered.⁶⁶
   (3) About 50 yards south-west of St. Mary’s church is a site which was once the yard of the Royal Oak Hotel and has since been built over successively for the Metropole garage and the Plaza cinema. Here Mr. Amos reports the discovery, near the rear of the Metropole Hotel, of a huge chunk of Roman masonry, built of ragstone, flint and tufa with white mortar. The masonry lay on a thin layer of Roman made-ground overlying the clay, and along the south side of the site on this level were Roman sherds (mostly third or fourth century), tiles, etc., and much burnt clay. At the south-west corner of the former yard, a petrol-pit was cut into a tufa-faced wall with a 2 in. offset, but did not expose its depth or width. The mass of the wall was a jumble of Roman oddments, and there were bits of coloured wall-plaster, Roman mortar, grooved tiles and the like, including one tile with ^CL/_BR (Classis Britannica), the CL reversed as on a tile from Pevensey. The insertion of a petrol-pump near the New Street entrance showed that the Roman debris reached the clay here at a depth of 9 ft. or 10 ft. Subsequently, when the cinema was built, its sloping floor was carried down to the back of the tufa wall, but this was not then adequately observed. Mr. Amos was able, however, to recover from the spoil upwards of a dozen tiles with the CL.BR or CL .BR stamp; in one case the stamp occupied a circular field on the tile, somewhat like another found in 1924 in the Roman villa at Folkestone.
   (4.) Market Place, north side. Little is recorded from this side. The most important discovery—an early grave-slab with Runic inscription, possibly of the seventh century—lies outside our scope. But about 30 yards up Cannon Street, on the east side, under the premises of Messrs. Goulden and Winds, the process of deepening a cellar brought to light much mortar, Roman potsherds and rubbish and, at one place, ‘flint boulders in white mortar, almost like a floor.’ Mr. Amos adds that he has been told of something like this at the corner-house a little farther up the street, by St. Mary’s church.
   (5) In 1881, when the great collegiate church of St. Martin-le-Grand, on the west side of the Market Square, was demolished and the Carlton Club built, remains of a Roman building were found and were thought (without reason or probability) to have formed a part of the baths discovered beneath St. Mary’s, 130 yards away. Rooms with concrete flooring and good flint walls and hypocausts were uncovered, and a few yards farther south, on the western fringe of the Market Square, a tessellated pavement is recorded at a depth of 10 ft. An interesting bit of sculpture (P1. IX, No. 3), now in Dover Museum, was found lying on one of the floors first mentioned. Thisisan undraped female figure, about three-quarters life-size, in oolite, with face, arms, and feet destroyed. It stands leaning slightly forward with legs crossed and arms (apparently) stretched out. The head is wreathed, and drapery hangs round
   ⁶⁶  J. Lyon, Arch. V, 325, and Hist. of Dover (1813), i, ii; Puckle, Arch. Cant. xx, 120; and information from the verger, Mr. Mathews, who saw the discoveries in 1897.