18 Mediæval Military Architecture in England.
the mound form a part of the exterior defences of the place.
The mound and base court, though the principal parts, were not always the whole work. Often there was on the outside of the court and applied to it, as at Brinklow and Rockingham, a second enclosure, also with its bank and ditch, frequently of larger area than the main court, though not so strongly defended. It was intended to shelter the flocks and herds of the tenants in case of an attack. At Norham, the castle ditch was used for this purpose as late as the reign of Henry VIII. There are several cases in which the mound is placed within a rectangular enclosure, which has given rise to a notion that the whole was Roman. Tamworth is such a case, and there, fortunately, the mound is known historically to have been the work of Æthelflaed, as is that of Leicester, similarly placed. From this and from the evidence of the earthworks themselves a like conclusion may be drawn as to the superadded mounds at Wareham, Wallingford, and Cardiff. At Helmsley, as at Castle Acre, Brougham, and Brough, the earthworks stand upon part of a Roman camp, and at Kilpeck and Moat Lane, near Llanidloes, part of the area may possibly be British.
East Anglia contains some fine examples of these moated mounds, combined with rectangular encampments. Castle Acre is an excellent example, as are, in a less degree, Mileham and Buckenham.
When the English lord took up his abode within a Roman camp or station, he often turned the Roman works, whether of earth or masonry, to account, and threw up his bank in one corner, altering the contiguous banks and ditches to suit his new arrangements. Thus at Pevensey, Leicester, Cambridge, Lincoln, Southampton, Winchester, Chichester, Caerleon, Chester, English mounds and base courts are placed within Roman enclosures which either are or were walled. At Auldchester, near Bicester, the Roman Alauna, is a mound of later date than the camp. At Plessy, Tamworth, Wallingford, Wareham, Cardiff, are found mounds decidedly of later date than the enclosing works. There are also cases where the mound is placed within an earthwork with something of a tendency to the rectangular, though scarcely to be pronounced either Roman or Romano-British. Such are Clare, Hereford, Eaton-Socon, where the mound is very small indeed ; and Lilbourne. Tempsford is very peculiar : it is a small rectangular enclosure close to the Ouse, and in one corner, upon the bank, is a small mound.
The group of works, of which the mound was the principal
