with Augustus or Tiberius; one or two Republican denarii, and one or two each of Augustus, Tiberius, Germanicus, Nero, Titus, Marcus, and Severus and his contemporaries; those of Tiberius and Nero are signalized by Battely as being fresh and unworn.²⁶  The second consists of a vast multitude of late coins, from Probus, Carausius and the Constantines to the end of the fourth century.
   Lastly, the access to the site. Physical conditions suggest that this must have been from the west or south-west, and we can trace a suitable route which may date from Roman times. This route runs north-eastwards from Canterbury past Sturry; then, turning in a more northerly direction, it climbs up and down hill past Maypole, Ford, Sweech Hill and Hillborough Church; finally it again turns eastwards and quickly reaches Reculver. Most of it is still in use as a road; part of it is a parish boundary, and the section nearest to Reculver is known as Old Lane.
   Such was Roman Reculver, so far as we know it to-day. We need not hesitate to infer its character. It was not a town : of that our evidence gives no suggestion. Its ramparts, its size, its strategic position watching the channel between Kent and Thanet, all show that it was a fort, and the coins found in it prove that it was occupied during the fourth century. Plainly it was one of the caste//a of the Saxon Shore, and its name—Recuif or Racuif or Raculfcestre in pre—Conquest days ²⁷ —identifies it with the castellum Regulbium, garrisoned (as the ‘Notitia’ tells us) by the Cohors I Baetasiorum. This regiment was quartered on the northern frontier of Britain during the second century. When the forts of the Saxon Shore were organized, it was apparently called south to assist in the coast defence.
   So far, our knowledge of Reculver, if slight, seems certain. A harder problem arises with respect to its history in earlier days. The occurrence of British and early Imperial coins suggests that the site was occupied long before the fourth century, and various structural details, such as the nearly square shape of the fort, and the absence of bonding tiles and of external towers or bastions, have been adduced by Mr. Fox to prove further that the actual ramparts date from a period definitely previous to A.D. 300. The question is not easy to decide without more evidence than we at present possess. The structural details seem to me indecisive. We have not the right to assert, for instance, that fourth-century forts had always external towers or bastions. Reculver itself witnesses to the contrary. For, whenever built, it was certainly occupied in A.D. 300—400. If bastions had then been thought indispensable, they would have been added, as they were added to the walls of certain of our Romano-British towns, such as London and Caerwent, or to earlier forts, as at Turn-Severin on the Danube, and (perhaps) at El-Leggun in Arabia. But it is safer at present to confine our view to the coins and pottery. They suggest an early occupation of Reculver, apparently in the first century of our era. Coins of Severus, or even of Marcus, might conceivably appear
   ²⁶ For the British and Republican coins see Battely, especially plate vii, and Evans, Ancient British Coins, indices; for 2 Augustus, 1 Germanicus, 1 Nero, 1 Titus, 1 Marcus, 2 or 3 Severus and ‘several hundreds of the Lower Empire‘ see R. Freeman, Regulbium, pp. 60, 65 unworn bronze of Tiberius and Nero, Battely, sec. 43, p. 79; base silver of Severus and contemporaries, Battely, sec. 51, 52, pp. 91, 94. Harris mentions also lead coins and’ unstruck globules of brass.’ B. Mus. Add. MS. 6626, fo. 8, mentions 14 Constantinian coins found in Dec. 1727 owing to a fall of cliff.
   ²⁷ For the Early English name see the Anglo-Saxon Chronicle 669, Bede, Hist. Ecci. v, 8, and early charter of 679, etc. The identification of Reculver and Regulbium has been accepted by almost all writers since Camden.