The walls of the fort are found on examination to consist partly of the old rocks before enumerated, and partly of that which I have now mentioned. Gneiss, quartz, granite, mica-slate, clay-slate, pudding-stone and pyritical slate are seen entangled together, with a very small proportion of the particular rock on which the fort itself is founded. The source whence these rocks were derived is evident, with the exception of the pyritical slate, which I could not trace in the neighbourhood.
I have now to enquire what motive could induce the builders of this work to reject the stone which lay at their feet, and to fetch from such a distance the large quantities required to raise their walls. It is particularly remarkable, that although the plain and shore are covered with fragments, yet these are almost entirely fragments of the primary rocks. I state this for the purpose of obviating a supposition that may be adduced to nullify the argument which I am about to derive of a previous intention in the builders to vitrify their work, from their having neglected to use that rock on which the building was erected, and which was not adapted for the purposes of vitrification. It might otherwise be suggested that they collected the loose rolled stones of the plain, as being ready broken to their hands. But besides that the pudding-stone is rare among these fragments, the pieces of the wall which have not felt the fire are angular and not rolled stones, showing pretty clearly that they were not collected on an alluvial plain, but broken from the rocks where they were formed.
Now, in the walls, the pudding-stone which we shall presently find to be the only vitrifiable ingredient, predominates to such a degree as to occupy the greater part of it. Hence it appears at least a probable conclusion, that the builders were acquainted with the effect of fire in destroying limestone, and that intending to erect a
Vol. II.2 l
