ABSTRACT
In the rectilinear ‘playing-card’ plan of the normal Roman fort the position of the gates in the longer sides is not always easy to determine exactly since it is usually to one side of the mid-point. But here again the modern street plan provides the clue. The dominant element, recognisable in the light of much later knowledge, in the layout of the fort area as it was at the time of the bombing was the irregular cross-pattern made by Wood Street with the east-to-west line of Addle Street and Silver Street (with Falcon Square) (Fig. 4). There are of course abundant analogies for the survival, inevitably distorted, of the features of Roman forts or settlements in the plans of the towns or villages that have succeeded them; and it has already been demon strated that Wood Street is the descendant of the north-to-south street of the fort. It was not unreasonable to guess therefore that Addle Street-Silver Street might be its east-to-west equivalent, though Addle Street at its east end must have moved a good deal off its original course. If this were indeed so, hope of finding the east gate was foredoomed to disappointment, for the cellars where the street crossed the line of the east wall were too deep for anything to have survived. At the west end on the other hand what was known of the relative levels suggested that some part of the gate might survive there. When these possibilities were being deliberated conditions on the site were such that they could not be put to the test because all the relevant cellars were full of bomb-rubble. The matter had therefore to be postponed until 1956, when the City Engineer kindly agreed that as part of the preliminary work of site-clearance for the building of the new Route 1 1 (later named London Wall) from Aldersgate Street to Moorgate Street the obstructions should be removed in sufficient time to enable an excavation to be conducted before the road-works in this area commenced. The result was to confirm the above diagnosis in the most satisfactory
fashion. It is not too much to say that the Roman remains were wrapped about with modern brickwork: their solid quality was such that the modern builders had not thought it necessary to remove them but had wherever possible incorporated the earlier work in their own. Much therefore of the northern half of the gate had survived within the area available for excava tion and it had not suffered unduly from the presence of one or two stone-or brick-lined cesspits of the eighteenth-nineteenth century, which had not destroyed the Roman work at any significant point. The gate followed one of the normal patterns for a fort gate (Fig. 5 and
Plate 10). It consisted of a double roadway of hard-rammed gravel divided by a central ‘spine’, made up of two detached piers. These piers carried the arches which spanned the roads and supported the gallery over them to pro vide access to the wall-walk from each side without descending to streetlevel. The roads on each side of the spine were 8J feet wide. The roadways were flanked to the north and south by square turrets. The northern turret was available for excavation at this stage; that to the south lay under Falcon Square, but part of its north side became accessible later when the Square was cut back to make way for new works. The face of the north turret was set slightly in front of that of the fort wall and its lower part at least was built of massive blocks of purple sandstone, some of them as much as 4 feet long: the generally rugged appearance of the frontage would have been enhanced by the heavy rock-dressing of the stone faces. The walls were fairly uniformly about 3 feet thick but the foundations of the front and back walls were deeper than those on the sides. A curious feature of the rear wall was that it did not rest conformably on its foundation, oversailing it by several inches externally, while the foundation was offset from the faced wall by a corresponding amount on the inside. The variation was presumably due either to a mistake in laying the foundation or to a change in plan while building was in progress. The turret is about 15 feet square overall. The doorway, at the south-east angle-that is, at the inner end of the gatepassage-had had its sill renewed at least once, probably as an adjustment to the raising of the road-level, and the stones showed signs of much wear. The floor of the turret was made up of mixed materials, but displayed no special features and produced no finds of significance. In the rebates of the outer arch both on the turret and on the spine the
sockets for the door-pivots had survived, that on the turret being the better
