THE EXCAVATIONS 51 



Melandra, as we liappen to know, lies in a very bleak and 

 exposed situation. It forms, as it were, a focus for every 

 wind that blows. If we add to the wholesale pilfering 

 that has taken place there the effects of frost, rain, springs, 

 the roots of vegetation, and the dampness of the soil (which 

 would materially assist the frost in its work), and remember 

 that the disintegrating influences which we have actually 

 seen work such havoc in a single season have had free play 

 for many hundreds of years, during which time the wall 

 has been frequently exposed, the wonder will be not that 

 so little but that so much remains. Let us end as we 

 began, by saying that the mode of construction of the 

 Melandra rampart remains an unsolved problem. But I 

 have examined all the sections very many times, both 

 when they were fresh and (which is instructive) at frequent 

 intervals during the winter, when the various forces of 

 <i'>nudation have had their way, and taking into considera- 

 tion ail the arguments, and especially remembering how 

 completely the ashlar wall has been stripped, and how 

 exposed the situation is, there seems to me fair ground 

 for supposing that the Melandra defences were of a similar 

 foTTn to those at the Saalburg, though the masonry of the 

 wall may possibly not have been so good, and that at the 

 Saalburg seems to have had two faces, and to have been 

 the chief defence. 



One final question arises. Is there any evidence to show 

 whether the wall was built later than the clay rampart? 

 I think anyone who has studied the remains and realised 

 how much they have suffered from destruction and decay 

 will feel how impossible it must be to answer this question. 

 In making his sections into the rampart Professor Conway 

 thought he detected in several places a line of boulders, 

 marking what he thought might have originally served as 

 a drain to the outer face of the bank. If this line could 



