Transactions. 39 



15 to IG feet wide. On the south-west it again deepeus, and here 

 it measures 14 feet, anil so it continues to the south, where it 

 altogether disappears. The second fosse, or rather what remains 

 of it, commences at the north-east, being fairly marked on the side 

 next the ditch, and reaching an extreme elevation of 8 feet, and 

 sweeps to the north-west. At the west it is, however, barely 

 traceable. But on the south-west aspect it again becomes marked, 

 especially towards the ditch, and finally it ends at the south. 

 This outer rampart is fast disappearing under tlie plough, and in a 

 few years I fear no trace of it will remain. The gateway through 

 this rampart, as shewn on the ordnance survey, has disappeared, 

 but was situated about 70 feet north of the inner gateway. The 

 greatest diameter of what remains of the camp is 370 feet. This 

 runs from the south-west to the north-east. The ramparts are 

 constructed of large loose stones, on which there is neither the 

 mark of chisel nor trace of mortar. These boulders are covered 

 with earth, dressed into a military shape. Having thus described 

 the position, measurements, appearance, and construction of tho 

 camp, I pass to a consideration of its origin. Beyond all doubt 

 the camji is British. This is obvious for the following reasons : 

 First. It is circular. The fortresses of the ancient Britons are 

 always found to consist of concentric circles of stones, whereas 

 those of the Komans are invariably square. Hill Burton insists 

 on the sameness of Roman camps in all i^arts of the world (p. 73 

 of History of Scotland), and says that in construction they evince 

 " an extremity of immutability." And both he and Chalmers use 

 " circular and British " and " square and Roman " as synonymous 

 terms. Second. The ramparts are not of Roman construction. 

 According to Chalmers (Caledonia Vol. I., p. 25) the ramparts of 

 British forts " were composed of dry stones and earth, without any 

 appearance of mortar or cement." Maclagan deals largely on this, 

 but in a more extended form. This is the construction of " the 

 lines " at Wood Castle. Third. The gateways through the 

 different ramparts in a British strength are placed in a zig-zag 

 manner, and not as in a Roman fort directly opposite each other. 

 — F/de Hill Burton, p. 84-8G, Vol. I. Fourth. The ancient 

 Britons often chose lakes for fortresses. Now the position of 

 Wood Castle is very peculiar. At the time of its construction it 

 must have been a peninsular stronghold. On the soutli-cast is the 

 Mill Loch, on the south-west the Upper Loch, and stretching 

 away along the whole of the west to the north is still mossy 



