182 



is not such firm or good work as the rubble grouted ; because 

 the mortar only penetrates part of the way into the interstices, 

 instead of forming a solid mass by completely filling thei)i. 

 The reader will see that I mention such herring-bone work 

 occurring at Neiningek's on the opposite side of the East 

 Gate from my own premises ; and that it also exactly matches 

 the building in the walls of Caei'went, 



Caerwent was a supporting station to Caerleon, which from 

 the testimony of Tacitus, confirmed by Ptolemy, we know 

 to have been the head-quarters of the second Augustan legion 

 during the first centwy. Caerwent^ from its position, must have 

 been occupied by the legion before Caerleon ; for it lies nine 

 miles east of it, and is the key to its approach either from the 

 Cotteswold Camps across the Severn, or from Gloucester by 

 way of the Forest of Dean. We are then certain that as the 

 second legion was the one which held it, from the very outset, 

 its ivalls ivere built by that legion. That they were built at an 

 early period is shewn by the flanking towers not being bonded 

 into the wall. They were added afterwai'ds. Now in several 

 of the stations on Hadrian's Wall there ai-e memorials of 

 the second legion, showing that its men shared largely in the 

 consti'uetion. 



The reader can put these facts together for himself, and I 

 think he will see what I am driving at, Hadrian's Wall, 

 partly built by the second legion, presents the exact types of 

 masonry, and the same materials in the mortar, which I find 

 in the north and south sides of the Decuman Gate of Gloucester ; 

 and these ii/^^es and materials together, without brick, are not 

 generally found elseiohere in the south of Britain. 



There is another point in the Gloucester wall which is not 

 infrequent in those of Roman build, but which I must not pass 

 over. All along one of the courses, at irregular intervals of 

 from three to five feet, there are square holes running trans- 

 versely through the wall. They are four or five inches in 

 diameter; and 1 found them filled with fine black mould. My 

 achseological visitors have given many explanations to account 

 for them ; but not such as are satisfactory. The general idea 



