Transactions: 65 
invalidates, if it does not destroy, much of the reasoning by which 
Roy nad sought to identify so many of our northern rectilinear 
earthworks with Agricola, and seems to leave their Roman origin 
more doubtful than ever. 
From the statement and review now given, the following 
inferences regarding Birrens and Birrenswark appear to be 
legitimately deducible :—(1) Birrens as shown by the inscribed 
stones found there was almost to a certainty an important 
Roman settlement. Its earthworks may also be Roman. But 
the belief that they are, mainly rests at present on the sculptures 
found in their neighbourhood and on their quadrangular form. 
As we know that Roman Camp defences were sometimes imitated 
by tribes witn whom they came into hostile contact, and who 
might even modify them to suit their own ideas of a stronghold, 
it appears to be necessary to have additional proof of a connection 
in time between the Roman antiquities found at Birrens and the 
mounds to be seen there, before we can affirm that the latter are 
also certainly Roman. The proximity of two objects of antiquity 
is not sufficient evidence that they belong to the same people and 
age. (2) Since it is conceivable that a Roman garrison at Birrens 
would establish a post of observation at Birrenswark, the camps 
on the latter may be Roman. Their form, irregular as it is, so 
far favours the supposition. They have a certain resemblance to 
some of the camps figured in the Plates of Napoleon’s Histoire de 
Jules César, and said to be Roman. It is not to be supposed 
that Roman Camps were always laid out with the geometrical 
regularity assigned to them by certain writers. The nature of 
the ground must have often determined their outline. At the 
same time, we know far too little about the social and military 
arrangements of the ditferent peoples that successively occupied 
Annandale in post-Roman times to enable us on the evidence at 
present available to say with confidence when or by whom the 
Birrenswark encampments were raised. Further investigation is 
required before it can be held as beyond dispute that they are 
the work of the Romans. General Roy’s arguments, while 
ingenious, are by no means satisfactory when critically examined. 
It may now be asked, Have we then any means of obtaining 
the desired evidence? Ancient history is all but silent about both 
Birrens and Birrenswark. But there still remains one source of 
information to which we can go with some chance of success—the 
mounds themselves. Within them the secret of their origin and 
¥) 
