26 THE EXCAVATIONS 



taken away and tlie superincumbent earth removed to a 

 considerable deptb before the first trace of the foundation 

 was discovered. When, liowever, tbe outer line of the 

 stone rampart bad been struck on botb sides, the position 

 of the gate was located and gradually the foundations of 

 the structure were uncovered. The excavations raised a 

 number of interesting points, which it will be well to put 

 on record. 



Beginning at the western side of the gate the stone 

 rampart was found to terminate in a stone 3 ft. square, 

 wider than the rest of the course, and beyond this ap- 

 peared a large boulder, apparently placed in position to 

 protect the angle of the gateway. This stone is embedded 

 in a considerable quantity of dark cement. An analysis 

 of this cement by Mr. Francis Jones, M.Sc, has shown that 

 it contains ferric oxide, traces of other metals, and sand. 

 It may be mentioned here that in his section of the wall 

 of the Roman fort at Manchester, Mr. Charles Roeder 

 marks a course of " brownish-black Roman mortar." ^ 



The plan shows that this gate was just as deeply recessed 

 as that on the east, but though the masonry is of excellent 

 character, what remains is not quite so massive. The 

 general plan appears to have been the same at both en- 

 trances. The foundations of the western guard-chamber 

 (if such it be) are nearly complete. Immediately to the 

 west of it, instead of the clay rampart, was found a mass 

 of charcoal about two feet deep, containing fragments of 

 pottery, and the floor of the chamber also showed traces 

 of charcoal. This is, however, a common feature of these 

 chambers.^ The natural inference is that we have here 



7. Boman Manchester, p. 8. 



8. See Ward : 77! e Roman Fort of Gellygaer, p. 40. (I have to thank 

 Mr. Ward for kindly giving me permission not only to quote from his 

 book, but also to make free use of his illustrations). See especially also 

 on this point Mr. J. P. Gibson's account of his excavation of the 

 Mucklebank Turret. Arch. Aelian., vol. xxiv., p. 16. 



