REPORT OF MEETINGS FOR 1902 247 



ascents from the river. The north side was rendered almost 

 equally strong by a morass (DD) in a hollow now drained, 

 which formerly covered the whole front, except a small 

 portion at the west end, where a narrow neck (E), level 

 with the plateau, connected it along the top of the bank 

 with the hills beyond. This approach was far too narrow 

 to permit of a serious attack, as the assailants would be 

 hemmed in between the steep bank and the morass. On 

 the east front the ground, although easier, was by no means 

 unfavourable to the defence. The southward trend of the 

 morass contracted the width of the access from the east, and 

 the ground fell away from the plateau in a hollow (F), which, 

 bending southward between the plateau and the "moraine" 

 (H), opened on the haugh of the river. Thus the only level 

 approach from the east was by a narrow space (G) between 

 this hollow and the morass. The "moraine" (H) 100 yards 

 dis^ant, and quite detached, may have been used as an outlying 

 defence, but no doubt, if taken by an enemy, would be a 

 source of weakness to the garrison. It will be seen how 

 skilfully the fortifications were designed, for the trenches, 

 marked by dark lines on the plan, are the most important 

 features in the defence. 



Before the recent excavations, almost the only fortifications 

 visible were the rfmains of the rectangular work (A), set 

 with its back on the western edge of the plateau ; but it 

 did not occupy the whole width of the plateau, and thus 

 two strips of level ground, one on the north (B), the other 

 on the south (0, 1), were left, upon which an enemy attacking 

 in force might effect a lodgment. The excavations proved, 

 however, as was to be expected, that the Romans had not 

 been unmindful of this risk, and had provided against it by 

 constructing the two wings or annexes (B, C) at the west end 

 of the ])lateau, thus occupying its full width at that end. 

 The north annex took in practically the whole of the level 

 ground on that side, and although the south annex occupied 

 only a part of the level ground on its side, it flanked (I) 

 the remaining part. 



On the strong side of the main work, that is the east, 

 Mr Ross showed us charts, one of which gave a section of 

 that part of the ground on which we were at the moment 



