THE EXCAVATIONS 35 



used. Anyone who has visited other forts would expect 

 that this was the case. The presence of what might be a 

 small hearth in one of them points in the same direction. 

 Whatever may be the answer to this question, the space 

 inside must have been very limited. The outside measure- 

 ments of these towers at Melandra vary from 8 ft. 5 in. 

 to 9 ft. 11 in. Even if the walls Avere only two feet thick 

 (and at Gellygaer they are thicker than this), the inside 

 dimensions would be not more than 5 ft. 11 in. and 

 4 ft. 5 in. respectively, so that the rooms would be mere 

 cells. (As will be seen in a moment, this was not the case 

 at the southern gateway.) At Chesters, Gellygaer, 

 Borcovicium, and other places where guard chambers 

 actually existed, the inside measurements vary from 8 to 

 12 feet. 



There is one other point. If we may draw an 

 analogy from the angle turrets at Melandra, there seems 

 no doubt that the lower chambers of these had no entrance 

 from the outside, and can only have been used, if used at 

 all, as storerooms entered from above. Mr. Garstang 

 (who excavated the two best-preserved towers) says 

 expressly -'' that " in no case had a tower, whether in a 

 corner, or flanking a gate, a masoned floor at the ground 

 level, nor any definite appearance of an entrance;" and 

 he goes on to refer to similar cases on the German Limes, 

 where the turrets are conjectured to have been provided 

 with a useful chamber in the upper storey only, which 

 might be entered directly from the sentry walk on the 

 rampart. We need not, however, go so far afield as 

 the Limes for an illustration. The towers at Hard Knott, 

 with outside measurements varying from 13 ft. 3 in. to 

 8 ft. 8 in. had no entrance on the ground floor, but 



27. Proc. Derb. Arch. Soc, vol. xxiii., p. 92. 



