On the Excavations at Rotherley, Woodeuts, and Bokerly Dyke. 289 
great states, for their defence, because the great armies of our time 
are encumbered with large supplies of food and ammunition, that 
have to be drawn from the rear, and for which it is necessary to keep 
open lines of communication with the base of their operations, and 
the frontier fortresses of an invaded state serve for the defence of 
that state, because it is impossible four an invading army to pass 
between them without exposing its lines of communication. Such 
fortresses also serve as fortified magazines for an invading army. 
But in barbarous times, such impedimenta did not exist in con- 
nection with invading forces; their objects were for the most part 
predatory, and their wants were few, they could penetrate between 
the fortified places, and subsist by plunder in the country surrounding 
them, and the defenders of the fortresses, if they kept on the 
defensive, and remained shut up in them, would only have to look 
on. Wherever, therefore, we find such isolated eneampments on 
the tops of hills, in prehistoric times, we may be sure that they were 
simply places of refuge for some local tribe inhabiting their vicinity, 
to which they resorted when attacked by a neighbouring tribe, 
They imply a low state of civilization, before the inhabitants of any 
large district had attained to such organization as was necessary for 
combined defence. 
When the people decineed to a higher state of civilization, and 
several tribes combined for the defence of a district, it was not by 
detached forts, but by continuous entrenchments, that they accom- 
plished that object. They threw up continuous lines of ditch and 
bank, the latter probably surmounted by a stockade, running for 
miles along the open country, from an inaccessible position on one 
flank to some other natural defence on the other flank ; and although 
it may be true—as has often been said in support of the opinion 
that these long entrenchments could not be defensive works—that 
they would be difficult, or impossible, to defend at all points, yet we 
know as a fact that this was the system adopted, and that the 
Romans used it, not only in the north of Britain, as a defence 
against the Picts and Scots, but also in the more extended defence 
- 
{ 
of their German frontier, by means of the Pfahlgraben, joining 
the Rhine and Danube. When these continuous barriers were 
