Divelling Places of Prehistoric Mem 33 



through the roof or sides, by the action of rabbits or other causes. 

 It occurred to me that by clearing away the silted covering of the 

 original floor level, remains might be found that would add con- 

 siderably to our knowledge. Anyone familiar with souterrains 

 will realize the difficulty of carrying away the silted soil through 

 the small passages that connect the chambers. 



The accompanying plan and section of Knock Dhu souter- 

 rain, kindly prepared by Messrs. Fennel & Clarke, will assist in 

 my description. There are seven chambers all connected, while 

 50 feet N.E. of the present entrance may be seen remains of an- 

 other chamber of a souterrain. I think a separate one, but pos- 

 sibly originally part of this one. It is now impossible to say, as 

 quarrying for road metal had formerly removed a great part of the 

 intervening surface, consisting chiefly cf broken stones, the accu- 

 mulation of centuries, fallen from the cliffs above. 



When we c(Mnmenced work, the entrance and much of the 

 first chamber were almost blocked up with loose stones, of which we 

 removed many tons. The difficulty of this work may be realised 

 when I say that at first we had only a space (;f from 18" to 24" 

 high to work in, and the difficulty of carrying heavy stones along 

 in so confined a space may be easily imagined. Under the space 

 occupied by the stones and a little clay we immediately came upon 

 a layer of from 6" to 12" deep, composed chiefly of burned wood 

 and bones mixed with the stiff glutincms blue clay prevalent in 

 the district. Tliis had been tramped into a solid brick-like sub- 

 stance by the feet of the many visitors who for generations had 

 been visiting the souterrain. Patient toil, however, enabled us to 

 work into this layer of charcoal, bone and clay, and we found that 

 it contained a great quantity of broken pottery and fragments of 

 iron, the latter for the most part so affected by corrosion that when 

 exposed to the air they crumbled to pieces. There were also 

 found lumps of iron ore; the only piece of manufactured iron 

 recognizable was a fine javelin head. Corrosion had greatly in- 

 jured it, but it still remains a good specimen, although the blade 

 broke off from the socket in removing it from the extremely stiff 

 clayey matter in which it lay. 



