120 PAPERS, ETC. 
Their form is not that adopted by the Romans, who chose 
their encampments more in the open ground, and generally 
at the junction of two rivers ; but the earlier inhabitants 
or occupants of the country on the hill tops, or on a pro- 
montory, which they cut off from the remaining portion of 
the hill by a deep ditch. Mr. Skinner traces intermediate 
encampments, and observes that “by carrying on this line 
of connection between the camps on the hills facing the 
Avon, and extending it with intermediate stations and out- 
posts to Farley Castle, above the Frome river, it prevented 
all possibility of invasion by the Silures, who used to pass 
over the Severn from the opposite side in light coracles, 
made of hides, which drew so little water that they could 
ascend high up the river and pass through the interval 
between the hill camps to invade and plunder the distriet.” 
“The vallum of Antonine against the Caledonians in Scot- 
land, and that of Hadrian in Northumberland, were ar- 
ranged after the same manner.” 
Having lately had the satisfaction of examining the 
barrier of Hadrian, which stretches between Carlisle and 
Newcastle, and which is, perhaps, the most interesting 
Roman work existing in this country or in any other, I am 
compelled to come to a very different conclusion to Mr. 
Skinner. In the barrier of thelower Isthmus there are most 
unequivocal marks of Roman occupation, and I can discover 
scarcely any in the camps mentioned by Mr. Skinner, at 
least in Maes Knolland Stantonbury ; and the form of that 
on Leigh Down and the Borough Walls appears not to be 
Roman. 
The Vallum of Hadrian, between Newcastle and Car- 
lisle, has been most ably investigated by Dr. Bruce, whose 
researches are contained in a most interesting volume, 
which has now come to a third edition. The form of each 
m 
= Ro uw 
