\ 



I 



\ 



THE EXCAVATIONS 43 



of tlie buildings was not unlike that of the southern fort. 

 One point in which the two have a striking resemblance, 

 is the central position of the Via Principalis. 



As the details of the interior of Melandra have still to 

 be obtained by excavation, the numbered squares (of 20 ft. 

 side), into which the area has been divided, have been laid 

 upon a separate sheet, so that, as excavations proceed, the 

 results may be added from time to time, pending the pub- 

 lication of a more complete plan of the fort. 



THE RAMPART. 



We now arrive at one of the most interesting questions 

 which the excavation of Melandra has raised. In his 

 interim report, referred to above, Mr. Garstang said : 

 " The rampart surrounding the fort is a feature of great 

 archaeological interest, and apparently of unique type." 

 In his paper on Melandra he describes it as "a form of 

 rampart unusual in Roman works." Nothing has tran- 

 spired that would tend to qiialify this description, and in 

 entering upon a short discussion of the subject it is better 

 to state at the outset that the mode of construction of the 

 Melandra rampart remains an unsolved problem. So far 

 no other fort fully excavated shows a similar defence, 

 though Mr. Haverfield kindly tells me (under date 

 December 27th, 1905) that " the rampart now uncovering 

 at Newstead, near Melrose, seems to have had a stone 

 facing, some rubble, and a lot of clay, but its details are 

 not yet clear." ** 



Mr. Garstang's description of the Melandra defence is 



44. The excavations at Newstead are not yet completed. Dr. 

 Anderson has, however, kindly sent nie the information that this 

 station, the largest as yet investigated in Scotland, was "defended by a 

 great earthen mound some 40 feet in width, faced with a wall 8 feet 

 thick, with three parallel lines of ditches." 



