THE EXCAVATIONS 27 



the remains of a large fire,^ but the bank has not yet been 

 cut back sufficiently to show how far the charred remains 

 extend. As the section has weathered back during the 

 winter, the black layer has only come out more distinctly. 



The floor of the chamber consists of irregular stones and 

 clay, and there is no indication of an entrance on either 

 side. The faced stones of its shell that still remain are 18 

 inches long, set back six to eight inches on a flag founda- 

 tion. Of the outer of the two bases of the pilasters on this 

 side nothing remains but the flag foundation, which is 

 about 3 ft. 6 ins. square ; that is, much larger than at 

 some other forts, indicating what stately structures the 

 Melandra gates miist have been. The inner one has two 

 courses of dressed stones in situ (the upper recessed), 

 and the accompanying photograph, though taken 

 in an unfortunate light, will serve to show the nature of 

 the work. The photograph is taken looking inwards, 

 towards the camp, in a westerly direction. In the fore- 

 ground to the right, part of the flag foundation of the 

 outer pilaster can just be made out, and the masonry of 

 the inner pilaster is well shown, as well as the floor or core 

 of the chamber in rear. The first course of stones has 

 a depth of 1 ft. 1| ins., the second of 10 inches. The 

 pilaster is very well squared, and (just as would be done 

 in work of the present day) the straight joint has been 

 broken on both sides. The style of the work leaves no 

 doubt that both arches were of a substantial character, 

 though, as the plan shows, the inner part of the spina is 

 lost. It was not considered worth while to show in the 

 plan the irregular stones lying about between the 

 chambers. 



Near this pilaster, evidently embedded in the road, 



9. Roeder searched in vain for evidences of a conflagration at Man- 

 chester. Roman Manchester, p. 56. 



