94 MELANDRA CASTLE. 



towers, as some of those along the " Limes,'' where it is con- 

 sidered that these turrets were provided, in many cases, with a 

 useful chamber only in the upper storey which might be entered 

 directly from the sentry-walk upon the rampart. 



The " camp " is similar in area and shape to other Roman forts 

 of the smaller type. Being nearly square its entrances are found 

 about the centre of its sides : each was flanked on either hand by 

 a tower, and each seems to have been spanned by a double 

 arch, with the exception of that behind the praetorium, that is to 

 say, to the south of it, which was probably simpler in form 

 Of the latter, not much remains, nor indeed has it been 

 sufficiently excavated as yet. Of the others, that to the east, 

 that is the right hand when looking down the enclosure from the 

 praetorium, may best be taken as the type ; for while it is well 

 preser^'ed in almost every feature of its plan, the other two retain 

 only such points as render it probable that they resembled it in 

 every way. 



The foundations disclosed in the exca\ations were those of 

 the towers which guarded the entrance on either hand, and their 

 junctions with the main wall and rampart, with the bases of the 

 pilasters which supported the outer and inner arches, were all 

 in position. The bed of the central " spina," which supported 

 the weight of the double span in the centre, alone was difficult 

 to locate. By reference to the plan of this entrance, it may seem 

 that the gateway was recessed a little from the frontage of the 

 camp-wall. a, b, c, d, are the bases of the pilasters abutting 

 against the masonry of the towers, from which the arches 

 sprang on either side. The elevations represent the stones as 

 actually found on excavation. At c is an interesting little 

 point, the superposition of two joins in the lowest and second 

 courses of the wall, at a point exactly in line with the outer face 

 of the tower wall on that side, thus showing, to some extent, how 

 the design was carried out. 



Of greater interest are the worked stones that for some years 

 had been noticeable, lying about near this entrance. Several 

 are voussoirs from the main arches of the gate and from a 

 smaller door or window. Others, again, from their chamfer and 



