Leaving the prehistoric walls of the town, we pass to those of the later defensive work-the mighty stronghold which occupied over 30 acres in the north-eastern angle of Wallingford. There may still be seen its triple rampart and moats, defences against the open country on the north, and its strong protection against the inhabitants of the town itself: all suggestive of the great change brought to England by the Norman conquerors, whereby it became a question, not of defending an already existing town, as such, but of establishing a feudal lord in safe quarters, whence he could take his share of the work of controlling the Saxon inhabitants of England. As the Rev. J. E. Field tells the history of this Castle, it is unnecessary for me to touch on the important part it played in Norman and later days, and we may pass to the material evidences which remain.

Though many sections of the earthworks have been mutilated to accommodate the site to modern residential purposes, enough exist to show the general scheme.1 We see that the leading feature was a high, artificial mount,2 such as we are familiar with throughout England and Normandy, reared not to carry stone structures, but to be surmounted and encircled by wooden defences; strong palisades, or stockades, of timber perhaps wattled and plastered, for newly thrown-up earth cannot carry the weight of masonry.

Surrounding the mount was its deep fosse, or moat ;3 and beyond, on the north, was the principal court, or bailey, with its triple defences and complicated series of banks and ditches: except on the eastern side, where a high rampart, partly natural, overhung the waters of the

1 It is deeply to be regretted that the modern "Castle Lane" was made, cutting deeply through the heart of the works, and destroying their continuity. Why it was made it is easy to see, but not the less to be lamented.

2 Mr. Hedges, excavating the mount, found it to rest on a solid foundation of stone, sloping upwards towards the outside, and forming a saucer-shaped base for the earth.

It is possible that the fosse, or moat, did not extend round the north base of the mount, though probably it did, but has been filled in for pasturage.

river, which in early days touched the foot of the steep slope. This large court was doubtless well occupied with buildings, barracks, halls, and stables; while south-west of the mount was a smaller court, still containing some fragments of walls and towers of stone.

Stone succeeded timber in later days, when the long succession of years had solidified the earth of artificial ramparts and mounts, or was used earlier when a natural hard foundation was near the surface, as in some parts of Wallingford Castle.

Even while all else was constructed of earth and timber, the Norman military architect usually inserted a gateway or gatehouse of masonry where he pierced the wall, and threw his drawbridge over the moat. Such a well guarded entrance-way may have been on the west, but if so, it has been swept away by modern alterations. There is some trace of a passage-way to the water side at the south-east angle of the Castle, but all other entrances are modern.

Before concluding, something must be said as to the date of these Castle earthworks. It will have been gathered from my remarks that I look upon the works as appertaining to the Norman period, and I think that much of it was the work of Robert D'Oyley, completed in 1071, while other portions-extensions of the original-we may owe to Plantagenet days.

Domesday records the destruction of eight hagae (probably meaning houses, with their appurtenances) for the construction of the Castle

Pro castello sunt viii [hagae] destructae.

If by this we understand that a new castle was builtas I think we may-we are faced by the difficulty that thirty acres (the area of the Castle precincts) is a space far beyond the requirements of eight hagae of the average size of those recorded as existing in Wallingford.

The entry in Domesday Book states that there were 276 hagae in the town, of which eight were destroyed. As we know the space within the walls to have been 114 acres, a simple arithmetical calculation shows that eight average hagae would occupy about three acres,

which is the space covered by many of the early Norman castles, and probably D'Oyley's work was no larger, the remaining twenty-seven acres of the Castle works being added later.

Mrs. Armitage has ably championed the Norman origin of mount and court castles, including this; while our Associate, Mr. T. Davies Pryce-an earnest student of the earthworks of Wales and Ireland, as well as England -inclines to an earlier origin for many, and possibly for Wallingford.2

Leaving the question of the exact date of the Castle's construction, I cannot but conclude by thanking the Misses Hedges for affording facilities for a quiet examination of the earthworks, and congratulating them on the possession of a place beautiful by nature and so full of archæological interest.

1 "Early Norman Castles of England," English Historical Review, 1904.

2 "The Alleged Norman Origin of Castles in England." Ibid, 1905