50 THE EXCAVATIONS 



understand it aright, lie there represents an ashlar wall 

 one stone in thickness and 14 feet high, as serving as a 

 revetment to a bank of clay with some rubble at the 

 l)ottom, rising to within a few feet of the top of the wall. 

 Now a rough rule due to calculation and experience would 

 seem to show that ground of an average character can be 

 retained by a wall that is one-third or possibly one-quarter 

 as thick as it is high. It is practically certain that the 

 outer shell of masonry at Melandra coidd not have 

 sustained the pressure of a clay bank.®^ If we assume 

 that the wall at Melandra stood at the height (suggested 

 hj Mr. Garstang) of 14 feet, then a wall 5 feet thick, 

 which seems suggested by the remains still to be seen 

 south of the eastern gate would be sufficient to hold in a 

 clay bank, and the whole structure would thus resemble 

 ihat at the Saalburg.t 



5. Of course the question arises : What has become of 

 "this rubble wall? I think the 1905 excavations, which 

 Professor Conway has specially directed towards the un- 

 covering of the outer rampart, have materially assisted in 

 answering this question. Mr. Garstang said of the outer 

 wall : " The traces of this now remain near the chief 

 gateways only." We haA^e traced it more or less completely 

 on all sides, sufficiently to prove without a doubt that it 

 once extended round the enclosure. But the plan will 

 show how completely this wall has been stripped by those 

 in search of stone, so that sometimes for 20 or 30 yards 

 not even a trace of the footings remains. The rubble wall 

 (even if it was not carried away) being thus robbed of its 

 support and pressed by the clay bank, would fall outwards. 



61. It is most interesting to note how emphatic Vitruvius is on this 

 ■question of lateral pressure of earth. Thus {op. rit. i., 6) "In the con- 

 struction of ramparts . . . the wall must be of sufficient thickness to 

 resist the pressure of earth against it." And again (vi., 11) "the thick- 

 ness of the wall must be proportioned to the weight of earth against it." 



t Mr. Haverfield does not think a height of lift, probable. 



