subsequently, arguing the continued existence and prosperity of the community. Lastly, the dyke was constructed, the earliest part of it not before the reign of Maximinus II, in the beginning of the fourth century, as is shown by a coin of that emperor found beneath the rampart on the old surface-line. At or after the departure of the Romans, a change, probably to render it more defensible at this point, was made in the direction of a portion of the dyke; and we may perhaps be permitted to surmise that the renewal of troubles, which this alteration indicates, led to the final destruction or abandonment of the settlement.
Before these excavations were begun not a trace of the village was to be seen, and its very existence had been forgotten. In the Itinerary of Antoninus the name of Vindogladia occurs on this line of road, and the distance between it and Sorbiodunum (Old Sarum) is put down as twelve Roman miles. Where Vindogladia was has hitherto been a matter of conjecture. General Pitt-Rivers suggests that it was precisely Woodyates, the distance from Sorbiodunum answering the requirements as nearly as possible. And he points to the fact that preceding antiquaries, though unaware of the existence of Woodyates, have interpreted the name to mean the White Rampart, from two Celtic words, vint, white, and gladh, a ditch or rampart—a name very suitable to Bokerly Dyke when the chalk out of which it was cut was fresh.
As in the case of the former villages, the number and size of the drains are one of its most impressive features. No wells were uncovered like those at Woodcuts and Rotherley; but the drains alone bore witness to a much heavier rainfall than at present. Some of them seem to have been afterwards filled up while the occupation of the site continued. This was found to have been the case also in the other villages, and the excavator has been much puzzled to account for it. Whether these particular drains became unnecessary owing to a diminished rainfall, or whether the