240 Dr. Mar Cuilocli o» the 



one centre, or the consequence of a process similar to that 

 occurring in the columns of Leptis. 



But in examining other cases of granitic desquamation, a 

 different appearance was observed. In these it was found, that 

 a single block desquamated in a manner so complicated that 

 no parallelism was maintained between the surfaces of the stone 

 and the crust ; and as, in some cases, such blocks were so tho- 

 roughly softened as to admit of being cut with a spade, it was 

 easily seen that more than one, or even two centres of desqua- 

 mation existed in a single mass ; the surfaces of the different 

 spheroids interfering with and compressing each other where they 

 came into contact, and the vacuities still required to fill up the 

 solid, consisting of deficient portions of crusts respecting one or 

 other of the approximate imbedded or internal spheroids. It is 

 evident, that in these instances, the effect could not have resulted 

 from that exterior action, be it of whatever nature, which pro- 

 duced the crusts on the columns already described, but must 

 have been determined by other causes depending on an interior 

 spheroidal and concretionary structure ; the existence of such 

 a structure being further supported by other well known ana- 

 logous circumstances in rocks, as well as by those already men- 

 tioned which occur in basalt artificially fused, and those which 

 are occasionally found in glass under particular circumstances. 



Thus, therefore, it is proved, that granite is susceptible of 

 desquamation from two causes, namely, from an internal con- 

 cretionary structure of a spheroidal and laminar nature, and in 

 consequence of the exposure of its surfaces to the atmosphere 

 independently of any such structure. 



But wherever, iu blocks of granite of acuboidal form, the whole 

 surface desquamates, it has already been shown that it cannot 

 be determined to which of these causes this effect must be 

 assigned ; and the same difficulty occurs in those cases where 

 the desquamation takes place in straight laminae. This is the 

 case of schistose granite, as it has been termed, and it requires 

 a further consideration. As the Island of Arran affords the 

 most accessible examples of schistose granite with which I am 

 acquainted, I shall describe the appearances which it there pre- 



