EH VI, Northeast Extension

(Nancy Proctor and Daniel Roschinotti 1995)

Objectives

The extension was designed to link EH VI to the castle proper, and to investigate a clear cut in the rock between the sanctuary and the castle wall.

Dimensions

The trench measured 2 x 11m., terminating to the northeast on top of the outer wall of the castle. Excavation was completed down to Roman levels.

Stratigraphy

The earliest feature in the trench was an almost vertical cut in bedrock, 60cm. deep (210). This runs parallel to the Roman alignment, 2.95 meters northeast of wall 6, and fairly clearly in relationship to it (it is almost precisely 10 Roman feet away). The cut delimits the leveled area of the sanctuary. Outside of this area the bedrock was also leveled, and seems to have been covered by a deposit of find gray earth (209). This may represent a road, for it covers an area exactly 6.2m. wide, which is a fairly standard width for the roads of Cosa (Brown 1951:27).

On the northeast side of the possible road was constructed a house, whose south corner is visible under the medieval wall (218). This was built of dry stone, faced on the outside. Within the building was found a coarse black and white mosaic pavement (219) with a mortar and tile preparation (201). The external surface contemporary with the building was of earth with a very compacted surface (211). We have no good evidence for the date of the structure, as no pottery was recovered from these levels. Its destruction is marked by 200, an orange-clay layer which seems to consist of decomposed pis» and which covers the walls and surfaces. There is no further trace of occupation in the Roman period, and the first medieval activity in the trench is the construction of the outer wall of the castle. This was built of roughly coursed rubble masonry, with much whitish-gray mortar. It thus resembles the second phase of the outer walls as seen in EH II, rather than the first, of which there is no trace in this trench. Somewhat later (although there is no stratigraphic build-up between the two) the defenses were strengthened by the construction of a berm and the digging of a ditch. The berm consisted of a single course of blocks certainly taken from the polygonal walls of the town (203), laid with their faces outward on top of a rubble preparation layer (223) (see pl. 30). Behind this revetment three layers of earth and rubble were deposited (180, 179, 177), compacted on the surface. In front of the wall the Roman layers were cut away (220) down to bedrock parallel to the Roman cut 210. No effort seems to have been made to cut the ditch any deeper, so that it was only 70cm. deep overall, and around 5m. wide. If the height of the berm is included, the new defenses constituted an obstacle of around 1.50m. in front of the main external wall of the castle.

The latest layers in the trench (178, 229) consisted of the tumble from the tower and the castle wall, containing, like the corresponding layers in EH I, far more stones than earth. These fell directly onto the bedrock at the bottom of the ditch, which suggests that very little time intervened between the cutting of the ditch and the destruction of the castle.

Dating

The construction technique of the berm, which uses a single course of stones removed from the polygonal walls to revet an earthen bank, is very similar to that of the catapult base excavated in EH in 1990 and it seems reasonable to assume that it had a very similar date, perhaps just before the fall of the castle in 1328.