THE EXCAVATIONS 59 



mentioned elsewhere (p. 91), is a rather unique relic. In 

 a number of cases the fragments of pottery found have 

 been successfully pieced together, so that fairly complete 

 specimens may be seen of the "Samian" bowl, the am- 

 phora, the mortarium, the patera, and glass bottles (see the 

 List of Miscellaneous Remains, infra.). 



The soil of Melaiidra has a deteriorating influence on 

 the pottery, which is quite soft when found, though it 

 hardens on exposure. On the other hand, the glass is well 

 preserved. Exactly the opposite is, I believe, the case at 

 Wilderspool, where the soil is sandy. All objects of lead 

 found at Melandra are thickly coated with the double 

 hydrate and carbonate of lead which is usually produced 

 when lead is left in contact with water. The coating has 

 been analysed by Mr. Francis Jones, who finds that it 

 contains no unusual features. 



MATERIALS. 



Some reference has been made in an earlier paper to 

 the materials of which the walls are built. On this point 

 Professor Boyd Dawkins writes me in answer to a ques- 

 tion : "All the sandstones at Melandra come from the 

 millstone grit, the light coloured flags as well as the 

 massive blocks. They might very well have come from. 

 Mouselow, or even nearer. . . . The Roman tiles were 

 probably made from boulder clay, but not necessarily 

 from any of the clays in the immediate neighbourhood." ®^ 

 As is indicated above, the gritstone varies greatly in 

 quality. Broken pieces of the upper beds, which have 

 poor weathering qualities, have been used for the founda- 



69. Vitruvius (De Arch., i., 5) declines to dilate on the question of 

 materials "because those which are most desirable cannot, from the 

 situation of a place, be always procured. We must, therefore, use what 

 are found on the spot." 



