knoivn as Amhn'.shnrjj Jjanh.s, Kpp'uKj Forest. 57 



reader for all that has been known about these camps up to 

 the present time, as it appears to me undesirable to trench 

 upon them by quoting passages which might deprive them of 

 their proper place in the history of this investigation. Not 

 having been the result of actual excavations they are neces- 

 sarily speculative, but they have served to keep up interest 

 in the subject, and may therefore be said to have contributed 

 materially to the present results. I would, however, remark 

 with respect to the Ambresbury Camp that the somewhat 

 angular form given to it has not conveyed to my mind, as it 

 appears to have done to some, the impression that it was the 

 work of the Komans. Eeference to Mr. D'Oyley's new plan, 

 accompanying this paper (Plate III.), in Avhich, at my 

 suggestion, the shading of the slopes around the camp are 

 given, will show that the configuration of the ramparts is 

 adai3ted to the features of the ground.^ On the east side a 

 ravine (a) approaches the camp from the valley below, and 

 divides into two forks (ab, ac) as it nears the camp; the 

 rampart at this place is drawn across the points of these 

 forks so as to sweep down them. On the south side also 

 advantage is taken of another ravine (d) to strengthen the 

 fortifications on that side. These are points which, although 

 influencing the principles of defence which have prevailed at 

 all times, are more especially British as distinct from Koman. 

 The Romans, caring more for their internal discipline and 

 the position of their cohorts than for external defence, 

 arranged their camps on geometrically constructed lines, and 

 often disregarded natural features altogether. It is true that 

 at the northern corner (e) of Ambresbury Camp the rampart 

 turns at an abrupt angle, but this is owing to the fact that at 

 that particular spot there are no natural features to guide 

 camp builders : the ground is a dead flat, and as the turn had 

 to be made somewhere it was made abruptly, as so often 

 occurs in British camps. An example of this is seen in the 



1 [Levels were carefully taken by Mr, D'Oyley for the purpose of this 

 investigation around and through the camp at distances of 100 feet or 

 less. The results are given in the jjlan, the datum being, of course, one 

 of the Ordnance bench-marks. — Ed.] 



