186 Barbury Castle. 
Barbury has no such complete tradition as surrounds Old Sarum, 
though the vicissitudes of its tenure and its first origin were probably 
not dissimilar. The clump of beeches that stand, weird sentries, 
just outside its trenches, bears witness to the presence of a thin 
capping of tertiary sand and clay overlying the chalk of the hill, 
and suggests the probability that in ancient times a grove of trees 
may have sheltered the bleak spit of down. Cesar tells of the 
oppida of the Britons as he knew them. The oppzdum of Cassivell- 
aunus was protected by woods and marshes, and would on occasion 
hold large numbers of people with their flocks and herds. Protected 
by vallum and foss, and strongly defended by nature as well as by 
art, a British oppidum, hidden when possible in the depths of a 
forest or crowning some commanding knoll, offered security from 
the incursions of foemen, and the inhabitants gathered there from 
their villages and huts, together with their moveable property, at 
the approach of danger. The Roman method of attack on such a 
British camp was simple; it consisted in a furious simultaneous 
onslaught on the entrenchment in two different places. The Roman 
castellum (a diminutive of castrwm), when planted in the neighbour- 
hood of a British oppidum, was a fortified position generally at some 
distance—sometimes as much as two or three miles away from it. 
Such an oppidum was Barbury Castle, and such a casted/um probably 
was the intrenchment first described in this notice. The term castle 
seems, in its application to a hill-fort, to have been transferred in 
later parlance from a neighbouring Roman settlement to what 
was originally a British Dun or Caer and afterwards an English 
Burh; which, indeed, not unfrequently (as at Marlborough and 
Ludgershall) became the site of a Norman castle. It has been 
supposed that the earliest form of these hill-forts was that of a single 
agger, with its foss—/ossa et terreus agyer—like the old (supposed) 
Belgian boundary, the Wansdyke, and many other so-called walls 
and dykes; and that the double mound and ditch and the searping 
of hill sides were the work of later and, probably, of Anglo-Saxon 
engineers. This may have been so with some of the great entrenched 
works of ancient Wessex like Sidbury and Old Sarum; but it is not 
likely that many of these doubly-mounded entrenchments were 
