a timber stair-way within the tower, suggesting that part of the upper hall remained to need access. Two pieces of walling (w and ww), north-west of the Hall, one of which remains, with a channel running into the older walling of the solar undercroft, and both founded on tile-debris, suggest a timber-framed building in this corner. Some of the clay mixed up with the tile-debris, especially in this area, may be the daub from this building.

D. The Dismantling
   Tile-debris was everywhere, deep in places, and is visible on every section in Figs. 4 and 5. It included ridge-tiles and at least one hip-tile from the west end of the Hall. It was always associated with pottery of the same types, just beneath it, mixed up with it, or, frequently, just above it. The sherds were often large, with fresh breaks, but no vessel approached completeness. Other accompanying rubbish included a little bone and much shell, chiefly oyster. The picture of the devastation, almost certainly that of 1312, that it presented was vivid and immediate: it was no mere break-in but a veritable Poltertag, The scale of the destruction suggests a larger party than the few named in the plea, as though the whole village had vented its feelings against the intruding judge. What is debatable is how much of the material broken had been stored in the castle, perhaps for a decade, and how much was brought in at the time. In contrast with earlier middens, jugs were very numerous, testifying, it seems, that copious drinking lubricated the work, with barrels of oysters to hand. Just as the plea claims, doors were broken down and their wrenched hinges were found thrown upon the pile. Most of the other small finds, belt-fittings and the like, were lost in the same confusion, but nothing very valuable seems to have been destroyed. The small quantity of glass from the fine windows, sifted from so much debris, almost echoes a shout to stop after the first stone was thrown, and take it down carefully for use elsewhere.

II. THE EARLIER PHASES
   These are considered in the light of our total knowledge of the Castle's evolution. The starting-point is the lower part of the curtain, which is now seen as the first deposit on the site. Since everything else is seen from sectioning, not from stripping, there must be an element of conjecture in any attempt to envisage the earlier stages in plan.

W.  The Primary Construction
   Preconditions of Excavation. In establishing the floor-levels inside the Hall (see below, Phases X and BB), it appeared that these had been raised after a fire, by nearly 1 m. in the main undercroft and by much