THE EXCAVATIONS 49 



2. If there was such a wall, the mortar has dis- 

 appeared. Now, we know for certain that there was 

 good mortar at Melandra, as some can still be shown m 

 situ. But it has nearly all disappeared, even from the 

 gateway piers. The mortar has also so completely dis- 

 appeared from Hard Knott, that it was only by the most 

 careful examination that the presence of mortar was de- 

 tected at all,s6 and at Gellygaer it is reduced to a sandy 

 loam.^'^ 



3. There is one very possible reason for the disappear- 

 ance of the mortar at Melandra. The fort is built in the 

 midst of the gritstone country, and the difficulty of 

 obtaining lime (so far as I know, there are no limestone 

 beds within a radius of ten miles) may easily have 

 influenced the character of the mortar.^s I liave 

 dealt with this question later,^^ in the section headed 

 -:! Materials." 60 



4. But the point which seems to have been most fre- 

 quently lost sight of in the discussion of the Melandra ram- 

 part is the question of the lateral fluid pressure due to the 

 presence of a bank of clay, or an accumulation of loose 

 rubble. I must confess that, bearing this point in mind, 

 the conjectural sketch of the Melandra defences given by 

 Mr. Garstang on Plate I. of his valuable paper on Eoman 

 Military "Works seems to me to be an impossible one. If I 



56. lb., p. 413. 



57. Ward. Op. cit., p. 25. 



58. Moreover, lime from the carboniferous limestones is said to be not 

 as good for mortar as that from other formations. 



59. See p. 61. 



60. It is interesting to note that Vitruvius mentions the decay of walls 

 in Rome in his time through the perishing of the mortar. "We may see 

 this in several monuments about the city, built of marble or of stones 

 squared externally . . . but filled up with rubble run with mortar. 

 Time has taken up the moisture of the mortar, and destroyed its 

 efficacy. ... All cohesion is thus ruined, and the walls fall to decay." 

 {De Arch., ii., 8.) 



