Ch. VI.] PRIMARY ROCKS. 93 



the granite of Goatfell presents solid and continuous faces of 

 rock ; while in others, it offers an irregularly laminated struc- 

 ture, very much resembling that which occurs in certain trap 

 rocks, including the syenites and porphyries. The fine- 

 grained granite, on the western side of this mountain, in the 

 Glen of Catcol, is very often even schistose. The laminae, 

 into which it exfoliates on decomposition, vary in thickness, 

 but may be found so thin, as not to exceed the tenth of an 

 inch. This granite is principally composed of an intimate 

 and minute mixture of felspar and quartz, with small crystals 

 of hornblende, and occasional scales of mica, sparingly dis- 

 persed throughout. The schistose structure, therefore, de- 

 pends in no wise on the mica ; and presents no analogy to 

 that of gneiss. And it must not be omitted to notice, that this 

 laborious and observant geologist has remarked, " that the 

 blocks which seem likely to undergo this change, and even 

 those in which it has actually commenced, show no symptoms 

 of future and similar exfoliations ; and that all mechanical 

 attempts to produce a new, or prolong an old, fissure, are 

 unavailing ; the rock breaking before the chisel or wedge in 

 the ordinary and irregular jnanner." This circumstance is 

 quoted, because it is at variance with the author's experience : 

 it will, however, be referred to shortly, when the cleavage of 

 granite comes under consideration. 



The developement of the structure of rocks, by a partial 

 and incipient decomposition, appears to be dependent on the 

 same principle as that by which the crystalline texture of a 

 saline or metallic mass is disclosed by the action of a weak 

 chemical solvent : but the laws, by which this action is go- 

 verned, have not been sufficiently investigated. We find, 

 wherever a large section of granite is exposed to the atmo- 

 sphere, that its surface becomes divided by two sets of fissures, 

 which preserve an exact parallelism among themselves, and 

 which, crossing each other at right angles, separate the mass 

 into blocks of a quadrangular form. Whilst this process is 

 proceeding, the large square surfaces of the granite will be 



