Ostorius Scapula is said to have occupied with a camp the Castle Hill. He, Suetonius Paulinus, and Agricola were busied mainly with the midland and northern parts of the island, and the southern province, Britannia Prima, seems to have been at peace. Roads were laid out, towns built, the metals were smelted, and agriculture prospered. Dubris (Dover), Durobrivis (Rochester), Rutupiæ (Richborough), Lemanis (Lymne), Regulbium (Reculver), and Anderida (Pevensey), came afterwards into notice as towns or havens. The Watling Street, which ran from Canterbury by Rochester northwards, seems to have been commenced at Dover. There were indeed roads from Canterbury to Lymne and to Richborough, but Dover would be the port reached by the production of the road in a straight line from Canterbury. In the reign of Valentinian, A.D. 364–7, a cohort of the Second Legion of 1,100 men was stationed at Dubris, where stamped tiles show them to have built a bath, and which is mentioned as a port in the Iter of Antoninus and as a town (civitas) by the geographer of Ravenna. In the time of Constantine, Dubris was one of the six ports south of Thanet, under the Count of the Saxon shore, the others being Rutupiæ, Regulbium, Lemanis, Anderida, and Adurnus (Portsmouth). It is, however, not included in the list of the twenty-eight towns existing when the Romans retired from Britain, of which Rutupiæ was one. Still, even if Dubris were one only of three heads of the Watling Street, its importance under the Roman sway was considerable. To the Roman period is to be referred the burial-ground laid open near the edge of the cliff in 1797, and the bath discovered on the brook west of St. Mary’s Church. No mention is made of a Roman fortress, nor was it in accordance with the practice of that people to place a permanent camp, still less a military station, upon a height so inaccessible as the Castle Hill. The existing earthworks show no traces of Roman outline, nor, when they had possession of the whole district, was there any need to fortify the lighthouse. The lighthouse alone, of the works upon the hill, can with certainty be pronounced to be Roman, but this, of course, implies the existence and employment of the port. All that the topographer can affirm is that the earthworks do not now present, and, so far as description may be relied on, do not appear ever to have presented, anything of a Roman character.