a school at that point. A fourth, known as the Little Dunhill, is said to have stood in the 18th century near the Terrace and Gravel Walk on the boundary line of the parishes of St. George and St. Mary Bredin, i.e. about 300 yds. north-east of the Dane John and, like it, just within the City walls. Apart from the bronze axe, all the discoveries associated with these mounds seem to have been of Roman or later date. Thus Leland records that ‘many years sins men soute for treasor at a place cauled the Dungen, thar Barnhailes House is now, and thar yn digging thei fownd a Corse closed yn leade.’ Since the whole group appears to have been known as the Dungeon Hills, it is not clear to which of the mounds Leland refers. Somewhat more precise is a reference to the accidental uncovering of a Roman cremation burial in 1783 in ‘an eminence on which is an orchard to the south-east of the Dane or Don John‘—presumably the third of the mounds noted above. The observer records that ‘on inspection of the mound raised over the place of interment, I found it to contain many fragments of brick, pottery, oyster-shells and animal bones’ (Soc. Antiq. MS. Minutes, xxxiii, 15 Jan. 1789). These discoveries suggest that the whole group of mounds may have belonged to the well-known series of mound-burials, characteristic of certain parts of northern France and Belgium and south-eastern Britain in the 1st and 2nd centuries A.D., and best represented in this country by the Bartlow Hills on the Essex-Cambridgeshire border (Roy. Corn. Hist. Mon. Hunts, p. xxxiii; Ann. Soc. Archeol. Namur, xxiv (1900). p. 45; Cumont, Comment la Belgique fut romanisée, p. 88; below p. 146). If this suggestion is correct, however, it is evident that the Dane John itself has been altered at more than one subsequent period. The uppermost 16 ft. of it are a modern addition, little more than a century old, whilst a ditch, which is said at one time to have girdled the mound, may be thought to indicate adaptation by Norman castle-builders (see W. Gostling, A Walk in and about the City of Canterbury, 1825, p. 9). Be that as it may, it is a fair assumption from the evidence that this group of mounds formed, in origin, a Romano-British—perhaps one might say Romano-Gaulish-—cemetery. Faussett came to a somewhat similar conclusion on less evidence (Arch. Journ. xxxii, 370-1).
   (b) Other cremation-burials in Wincheap, etc. Brent (Cant. Olden Time, 14) states in general terms that many cinerary urns have come to light beneath the gasworks, a malthouse in Castle Street, Wincheap Street, and Wincheap Green, scattered here and there in no distinguishable order. Of individual discoveries scarcely any record survives. An Upchurch urn filled with burnt bones, black earth and rubbish was found in 1849, 12 ft. deep under the Castle Gasworks. About 1856. and later, the construction of a gasometer east of Castle Street revealed ‘many urns lying sideways on clay, surrounded with flints.’ One of these urns, held the bones of a small animal and a bronze utensil; another was a Samian straight-sided cup (Dragendorif type 3 i), stamped inside QVINTI M. The only ornaments noted were two flat-headed bone pins. For these finds see Good, Brit. Arch. Assoc. Journ. xii, 73; J. Brent, Cant. Olden Time, 14, and Arch. Cant. iv, 28—all rather brief. Good mentions a bronze object which he took to be a sword-hilt, found near a skeleton, but this may be post-Roman. In July 1861 an unsatisfactorily described find was made a few feet inside the south wall of the Castle and no more than 12 in. deep—an urn containing burnt bones and bits of white concrete, with bones of sheep, swine and fish lying close round it (Brent, ibid.: his two accounts do not quite agree). In 1861, also, the construction of a new Railway Inn at the corner of Wincheap Street and Green revealed, 6 ft. deep, in a steep-sided pit or trench of (presumably) Roman origin, several mortuary urns, a black patera, an Upchurch thumb-vase, and a coin of Constantine—the only one of all the coins found in the cemeteries of Canterbury of which the date is recorded (Brent, Arch. Cant. iv, 34, calling the urn Castor ware: it is in Canterbury Museum where I have seen it). This seems the same as a ‘grave’ said to have been found about 186o at Ivy House, which was cut down through ‘alluvial soil’ to the chalk and contained a black globular burial urn; with it, immediately above the chalk, were tiles, broken glass, potsherds and oyster-shells (J. B. Shepherd in Proc. Soc. Antiq. Ser. II, i, 185; hence apparently Dowker in Arch. Cant. xvii, 367). Brent records also an urn containing two Roman rings, found in the gasworks in 1770, and a burial urn from 8 Dane John Grove (Cant. Olden Time, 14, 42). A deep Samian saucer stamped PATERCLINIO, found in the gasworks in 1885 and seen by me in Canterbury Museum, may come from outside the walls, if we can argue from the date of finding.
   (c) Other inhumation-burials in Wincheap and Martyrs’ Field. Of these we possess only meagre and unsatisfactory accounts. In 1861 the railway works and also gravel-digging in Wall Field, south-west of the railway station, revealed numerous skeletons, mostly 3 ft. to 4 ft. deep. They lay with feet to the east and south-east, save one (or two) buried with feet to the west. With them lay many iron nails, 7 in. to 10 in. long, suggestive of wooden coffins. Grave furniture was noted only in two or three burials, and the accounts of the things found in individual graves do not agree. They consisted of three bronze armlets—two of twisted wire with hook-and-eye fastening and one decorated with an animal’s head, a little bronze box containing coins, an iron hook, green and purple glass beads cut in facets and a blue ribbed bead, a bone pin with a green opal (?) head, and