64 Transactions. 



which did not come into Britain till the reign of Hadrian, and 

 whose headquarters were at fork. He further supposes that 

 soon after Agricola's recall, the Romans lost the greater part of 

 thej country between the two isthmuses, and that Hadrian in 

 consequence fixed the boundary of the empire in Britain on 

 the southern isthmus. While, however, the wall was being 

 built, he posted a detachment of his army at Birrenswark to 

 watch the enemy's motions, especially if they advanced in any 

 great body from the north to interrupt the work. The first 

 halting place of Agricola, in the west of which any trace 

 remained in his day, was, Roy thought, a camp on Torwood 

 Moor, near Lockerby. Adopted in the main by succeeding 

 antiquaries, Roy's views on the Roman occupation of Southern 

 Scotland may be said still to hold the field. 



Roy, it must be acknowledged, made an earnest attempt to 

 grapple with his subject. His method has all the appearance of 

 being strictly scientific. He seeks to plant his foot firmly before 

 taking another step in advance, and to remove any obstacles that 

 seem to stand in his way. Tlie most fatal blot on his work is his 

 acceptance of the De Situ Britannim as genuine. This not only 

 vitiates his " rectification of the ancient geography of North 

 Britain," but leads him far astray in other matters, although it 

 only indirectly affects what he says of Birrens and Birrenswark. 

 Moreover, he, too, readily fell in with Captain Melville's opinion 

 as to the Strathmore " camps." Under its influence he saw 

 resemblances between them and those of the normal Roman type 

 that it may be safely said would never have otherwise occurred to 

 him. The wish became father to the thought. But after his work 

 was finished his own confidence in his conclusions must have been 

 shaken. In 1787 a "camp" was discovered near the sources of 

 the Ythan, in Aberdeenshire, with characteristics as Polybian 

 as those of Strathmore and^ Torwood Moor ; yet, it is situated 

 where, on any interpretation of Tacitus' words, Agricola could 

 hardly have been. Its existence is said to have been made 

 known to him ; and a plan of it with particulars is the last plate 

 in the " Military Antiquities." The insertion of this plate, how- 

 ever, is probably due to his editor or editors. Roy could hardly, 

 without some explanation, have sanctioned the statement made 

 on it that this Aberdeenshire "camp" resembles " the camps which 

 are supposed to be Roman on the south side of the Grampian 

 hills." It is not too much to say that the discovery of this camp 



