1 86 Dr. G. H. Bailey on 



the mass. No considerable differences were found, except 

 that in some parts the iron was entirely oxidised, whereas 

 in others it still remained, for the most part, in the ferrous 

 condition, and in addition to this there seemed to be rather 

 more alkalies in the denser parts of the mass. The typical 

 sample gave on analysis the following results, the sample 

 analysed by Dr. Angus Smith being placed alongside for 

 the sake of comparison, though I have no evidence that 

 they are identical specimens, and indeed the analyses them- 

 selves would certainly indicate that they were not. 



Mass in Mass from 



Manchester Museum. Glen Nevis. 



There may have been originally more alkalies present ; the 

 stone is of such a porous nature that these would be partially 

 dissolved away by exposure to atmospheric conditions. 



There is then a very low proportion of alumina, prac- 

 tically no lime or magnesia, and little alkali. No common 

 mineral or rock substance, as far as I am aware, shows such 

 a composition. 



It has been suggested by previous writers on this sub- 

 ject that basalt was added to clay, or some such substance, 

 to form a mass which could be fused at such temperatures 

 as were likely to be at the command of the builders of these 

 forts. Wood, it has been suggested, was the fuel used, and 

 indeed in one case wood has been actually found in situ 

 between the layers of stone constituting the wall. 



In the case of the specimen examined, however, no 

 basalt can have been added, for though the large proportion 



