The pottery is divided into four groups according to the stratigraphical evidence, which is given in detail in the descriptions of the pottery on pp. 66–67. With one exception (group 1, no. 1), which is a stray of earlier date, all the pottery may be regarded as contemporary, since even the latest, group 4 from pit C1, was sealed by the old turf-line over the Norman Bank. There is no historical evidence for the construction of this Bank, but since it is not in alignment with the Castle Keep, built c. 1076–86, it is believed to be slightly later in date. It may be said at once that no appreciable difference in date can be detected between the pottery of groups 2, 3, and 4. Comparison with dated material found elsewhere in England suggests that the pottery is not later than the early part of the twelfth century, but closer dating is provided by comparison with material from Colchester Castle itself. This consists of four sherds, including the angle of a sagging base, obtained from the foundation trench of the Castle Keep in 1931–2, which Mr. Hull kindly sent to me for examination. Although so few in number, the ware of these sherds is consistent and agrees precisely with that of groups 2, 3, and 4 as described below. The date of these groups, and therefore of the Norman Bank, is thus as close to the date of the Keep as the sequence of construction will allow. On balance, the available evidence points to a date within the period c. 1080–1100 for the Norman Bank.