8 ANCIENT FORTIFICATIONS in 



which they lie. All that I am at prefent concerned to fliew, is, 

 that, from the fuperficial or external appearance of this hill, 

 there is no reafon for fuppoling that it ever contained inteftine 

 fire. 



The ftone, of which the whole of this hill, and molt of the 

 neighbouring hills are compofed, is a mixed mafs of round 

 water-worn pieces of different coloured granite, greyifh or 

 fpeckled quartz, and the common white quartz. This compound 

 ftone, which is well known to miners, has, from its appear- 

 ance, been termed plum-pudding ftone. Thofe who have en- 

 tertained the notion of Craig-Phadrick's being an extinguifhed 

 volcano, have maintained, that this compound ftone is of the 

 nature of the volcanic tufas. This, however, will be acknow- 

 ledged to be a miftake, by all who have examined and compared 

 the two fubftances. The volcanic tufas are all compofed of 

 materials which have undergone a change by fire ; the plum- 

 pudding ftone has undergone no fuch change. Sir William 

 Hamilton defcribes tufa to be a foft ftone, compofed of pu- 

 mice, afhes and burnt matter, its colour often tinged with grey, 

 green and yellow. It is formed, fays he, by water making up 

 thefe materials into a fort of clay, which afterwards hardens. 

 The plum-pudding ftone, on the contrary, contains no burnt 

 materials. Its component parts, fo far from being already 

 burnt, when expofed to fire, undergo a total change, and the 

 whole ftone fuffers an imperfect vitrification. Upon the whole 

 furface of this hill, and amidft all the detached fragments, both 

 of the natural ftone and of the vitrified matter, there is not, fo 

 far as I could obferve, any thing that bears the appearance of a 

 pumice ftone. The burnt matter, indeed, is often full of fmall 

 holes or honey-combed ; but it ftill retains a glafly appearance 

 and a confiderable weight, both which circumftances fufEcient- 

 ly diftinguifh it from pumice. Bafaltes are, I believe, conftant- 

 ly found, in fome form or another, upon all volcanic hills ', but 

 neither on the rock of Craig-Phadrick, nor on any of the neigh- 

 bouring 



