VOL. LXXXI.] PHILOSOPHICAL TBANSACTIONS. 15 



have had an opportunity of making observations. The focus in different instances 

 may be seated at a different distance from the surface ; but none are probably less 

 than several miles at least deep. 2. The currents of granitic lava in the Pontian 

 isles leave little room to doubt of the power of subterraneous fire to produce this 

 substance. To suppose them to be rocks of granite fused, but otherwise un- 

 changed, and that even fissile rocks may be made to flow without losing their 

 laminated structure ; is as bold an assumption as can easily be taken up. In the 

 great igneous processes of Nature, fire need not be imagined to act otherwise than 

 in our small experiments ; we actually see it producing glass and cellular spongy 

 scoriae: when the products are of a different character, we must have recourse to 

 accessory circumstances, and not violate the plainest rules of philosophizing by 

 attributing different effects to the same cause. The latent motive for such an 

 extraordinary hypothesis may easily be divined ; the observer took it for granted, 

 that all granite is of aqueous formation ; hence he was obliged to reason back- 

 wards from the unknown, that of the Alps for instance, to the known, instead of 

 proceeding from the palpable effects of subterraneous fire by easy steps to a gene- 

 ral theory of granite. When it is taken for granted, before examination, that 

 granite cannot he formed by fire, there remains no resource but to say, that gra- 

 nitic lavas are granite rocks fused, but not altered in the arrangement of their 

 constituent parts. Though the heat of volcanos be sometimes and in some places 

 moderate ; in others we have good reason to believe, that it exceeds any degree 

 we can produce, except by means of factitious air ; we are certain that it forms 

 molten currents of petrosilex and flint exactly the same as our gun flints. If we 

 admit this reasoning, the appearance of granite in the bosom of volcanic desola- 

 lation may, if duly examined in all its circumstances, afford strong evidence of its 

 production by fusion ; and it is reasonable to conclude, that it was once covered 

 to a considerable depth by erupted matters, which the course of time, and the 

 injuries of the atmosphere, have removed ; though he by no means denies that a 

 volcano may force its way through pre-existing rocks of granite. 



There is still another analogy between basaltes and granite, more important to 

 the theory of the earth, and less liable to controversy than either of the preced- 

 ing. In their situation, with respect to other rocks, we may observe the same 

 law. The general rule of super-position, reckoning from below upwards, is, 

 1. granite; 2. schistus ; 3. lime-stone. This rule has been found to hold good 

 by so many mineralogical travellers that, though it may not be absolutely uni- 

 versal, it must be allowed to prevail very extensively. Now, in this island there 

 are numerous instances where basaltes is substituted in the series instead of 

 granite, and where it seems to alternate with granite as the substratum of other 

 rocks. On the road from Dolgelly in Merionethshire, by Mallwhyd and Cann's 

 Office, through Llanfair to Welchpool, schistus appears always incumbent on 



