Margaret Hill is the continuation of Borough High Street to St. George's Church; the name may mark a military mound. In Domesday it appears that Southwark had been subject to the Confessor and Godwine.[115] The men of Southwark testified that in King Edward's time no one took toll on the Strand or in the Water Street save the king. Godwine had a house here, and he must have held the burh. In the dispute of 1051-52 between the Confessor and Godwine, the earl carried his forces up the river to Southwark, the burghers of which followed his cause and supported him by land. The king's navy and land force faced him from the north. The Londoners sympathised with the earl, but officially it was a case of Southwark against the city.[116] It would probably be possible even now to lay down the course of the "walls" (of earth, like Wareham and Wallingford) by comparing the boundary of the old manor or "town" with street lines and names and other evidence.[117] Godwine's holding seems to have coincided with the gildable manor which extended along the river from St. Mary Overie's dock to Haywharf in the east, and southward nearly to St. Margaret Hill. Two other adjoining manors were included in the parliamentary area. Even the site of the great earl's manor house can, with some probability, be pointed to.[118] Excavations have shown that before Saxon days there was a considerable Roman settlement on the site of Southwark, and that the present High Street lies over the Roman approach to London. Roach Smith says that substantial remains of Roman houses have been found, particularly on both sides of the High Street up to the vicinity of St. George's Church, in which district the wall paintings and other evidence indicated villas of a superior kind. Nearer the river, where the ground had been subject to inundation, the houses were built upon piles. In 1016 Cnut, to turn the flank of the bridge, dug a "mickle dyke" on the south, and dragged his ships to the west side of the bridge. Sir W. Besant has shown that quite a little dyke a few yards long would go round the bridge end and take a Danish ship, but he has not considered the preliminary forcing of the South-work which would have been necessary. As to the probable course of the dyke, see Allen's _History of London_, vol. i., and Faithorn's map, 1658, which shows a considerable stream flowing into St. Saviour's dock. It was required more for the investment of the stronghold than for the ships (which, as at Constantinople, could have been dragged over land), as shown by the complete passage: "They dug a great ditch on the south side, and dragged their ships to the west side of the bridge, and then afterwards ditched the city around, so that no one could go either in or out." * * * * *