324 



opinion, wliicli he supports by the presence of vegetable mould 

 found in those holes. I leave it to the professional man to 

 decide whether these holes were not merely meant to fix the 

 scaffolding, and have been left empty after having fulfilled their 

 purpose. The structure of the wall of Hadrian, well known 

 through Bkuce's excellent treatises, (1) being identical with the 

 Gloucester one, the author puts the date of the walls of Glevum 

 at the time of the Emperor Hadrian. But this is decidedly 

 not a conclusive argument. The mode of construction in itself 

 dates back to time immemorial, and was undoubtedly used 

 whenever the want, or the inferior quality of the material, or 

 shortness of time, recommended it in preference to the costly 

 massive free-stone. The measurements too of the facing-stones of 

 the wall of Hadrian, are entirely different. (2) The two modes of 

 construction agree, but in this one point, that there are not, as is 

 the case with the Roman structures of northern Italy, and other 

 provinces, and in some instances in the south of England, rows of 

 tiles between the facing-stones. All these considerations seem to 

 have induced the author to consider the two as identical. But the 

 absence of tiles has been noticed in many other Roman structui-es 

 in Britain and elsewhere, and may partly be accounted for by the 

 want of the proper clay for the manufacturing of tiles, and the 

 more abundant supply of stones; partly also by the quicker way in 

 which the latter can be obtained. I cannot, it is true, state with 

 certainty, whether roughly hewn sandstones can always be more 

 quickly prepared than tiles, when carefully made. In very few 

 places only we find, instead of the rough conglomerate, that mode 

 of construction which the English call herring -bo7ie work: evenly 

 hewn stones are placed in straight rows, their points in an 

 oblique direction, and set in mortar ; thus giving an appearance 

 like fish-bones. The portions of the walls of Glevum built in this 

 manner (pp. 19 & 29) the author considers as of more modern 

 origin ; for my part, however, I am rather inclined to think that 

 the reverse is the case, and that the more careful, although less 



(1) Vide my notice in the Jenaer Litteraturzeitung, 1875, p. 808. 



(2) "Bruce's Roman AVall" (3rd edit. London, 1867. 4) p, 81. They are mostly 

 1.5 to 20 inches long, 8 to 9 inches thick, and 10 to 11 inches wide. 



