fort had a single gateway; but to reach this it was necessary to penetrate the outer defence at a point further to the north or south, and to pass round behind the double ditch cut in front of it. Thus, no assailant could avoid presenting his uncovered side to the defenders on the rampart. Hyginus prescribes a somewhat simpler, but equally effective, device for the protection of the gates of a camp.[1] It was known as the clavicula or 'bolt.' As the wall or rampart approached the entrance from the left side it suddenly curved inwards, as if to form a semicircle with the wall or rampart coming to meet it from the opposite side. But the semicircle was not completed. A gap wide enough to allow of a gateway remained open. The result was that the attacking party were forced to operate in a comparatively narrow space, and to expose their right sides to the concentrated fire of those on the rampart above them. This contrivance was employed by Caesar at all five gates of his camp on the Aisne,[2] and it is also to be seen at Kneblinghausen, where indeed it is an indication of the early period to which the camp belongs. Something of the same idea underlies the Newstead plan. The effect of the way in which the entrances were arranged was to narrow down the space in which the enemy could operate, and to make the attacking force pass immediately in front of the defenders with its right flank turned towards them.

The Abandonment of the Early Fort

The early fort seems to have been abandoned after a period of occupation which, though perhaps of no very great length, probably covered several years, to judge by the amount of material which had gathered in the ditches. That on the west side contained an accumulation of some four to five feet of black silt, from which a varied collection of broken pottery, metal objects and leather was obtained. A curious feature here was the occurrence of regular black lines in the filling. A typical section was one taken at the south-west angle where the ditch bends to the south, and where it underlies the rampart of the later fort. At this place the total depth from the modern surface was twelve feet two inches. The uppermost five and a half feet consisted of the clay of the later rampart. Within the next three feet, nine parallel black bands, about four inches apart, crossed the deposit horizontally. Below these, for a thickness of three feet eight inches, was the usual brown vegetable silt. The black bands at first gave the impression that they had been caused

1 Liber de Mun. Castr. c. 55.

2 Napoleon III., Histoire de Jules Céesar, Atlas, pl. ix.