CASTLE NEROCHE. 35 
could be found, we often find the extreme point of a hill 
separated from the rest by one or more deep trenches and 
ramparts, cut quite through the ridge from side to side, 
and fortified in the manner before described. It is also 
frequently the case, that a part of the area within the 
rampart is divided in this way from the rest, and fortified 
with peculiar care, forming what is almost analogous to 
the keep of a medival castle. 
The distinetion, then, between Roman and British en- 
trenchments is so marked as almost to obviate the possi- 
bility of mistake, where both have been kept separate; and 
even where the Romans afterwards occupied the British 
works, as at Hamdon Hill, the site of their camps is easily 
to be distinguished from the British fortifications. But 
with regard to Saxon and Danish works, the case is very 
different, nor am I able to give any criterion by which they 
may be distinguished from those of the Britons. Kenny 
Wilkins’s castle near the British fortified station at Pen, 
(which is probably the camp occupied, and perhaps con- 
structed by the West Saxons, in the year 658, when 
Kenewalch, King of Wessex, defeated the Bretwallas at 
Pen, and drove them beyond the Parret), has unfortu- 
nately been planted; and owing to the thickness of the 
covert, it is not very easy to trace its original plan. It ap- 
pears, however, to have consisted of one deep trench between 
two raised mounds, enclosing a considerable area, having two 
entrances into the trench through the external rampart, 
and one through the internal agger ; a higber point of 
ground within the area being separated from the rest by a 
trench and rampart. The earthworks of the Danes, whose 
rapid and predatory incursions did not admit of any 
elaborate fortifications, appear generally to consist of a 
rude and irregular entrenchment, of no great strength, 
