296 Notes on the Discovery of Romano-British Kilns and Pottery 



the clay lining measured Sin., and in others under 2in., showing 

 great inequality. 



The urn had apparently been imperfectly baked, and fell to pieces 

 when it was removed. I have since been able to restore it, and it 

 is by far the best example, as it is quite complete except a small 

 piece of the rim.^ 



These kilns were apparently in a row and connected with each other 

 by flues, or narrow channels. Pursuing the last kiln northwards it was 

 found to get narrower as the excavation proceeded ; 4ft. from the en- 

 trance, or rather, the side which I first came upon it, it measured 

 32in. across, and two feet further on SSin., and further on still it ulti- 

 mately contracted to 14in. It therefore was not so globular as the 

 previous kiln, but shaped more like an egg set lengthways. I did 

 not excavate further, as it appeared that at this point we had reached 

 the end of the series of kilns in that direction. I, however, made 

 another excavation 19ft. in a direct line from this kiln, and found a 

 small flue or channel which seemed to be a continuation of the small 

 passage leading from the kiln. It was filled with black earth and 

 bits of pottery, and was about 16in. below the surface. A similar 

 excavation 10ft. further northward produced similar results. An 

 excavation at another 10ft. to the north led to no results. The 

 growing crops prevented an extensive system of trenches. 



In addition to the urns and numerous pieces of pottery, masses of 

 a fine clayey mixture, turf fibres, and pieces of burnt clay were 

 found. 



The upper greensand stratum at Broomsgrove contains horizontal 

 bands of sandstone and cherty rock, probably corresponding to the 

 ** Potterne Rock " in the neighbourhood of Devizes and elsewhere, 

 as described by Mr. Jukes-Brown in Wilts Mag., vol. xxv., p. 3Sd2. 

 These horizontal layers would greatly facilitate the construction of 

 the globular-shaped kilns, as they would to a considerable extent 

 prevent the falling in of the looser sand. 



Taking into consideration the discovery as a whole, together with 

 its situation, we may safely draw the following conclusions :— the 



' This is the smaller of the two urns in the accompanying illustration. 



