ARRAN. 'GEOLOGY. SECONDARY ST11ATA. 387 



A nan actually overlying or traversing the secondary 

 rocks of that island. Any portion would be sufficient 

 for this purpose ; since, as the overlying syenite is proved 

 to be connected with masses inferior to the secondary 

 strata, it would be indifferent, in establishing the claims 

 of any analogous rock, whether the larger visible portion 

 of it was superior or inferior to those strata. The want 

 of this criterion is doubtless a most important, but it 

 is not an unsurmoun table, objection. I have formerly 

 shown that in Sky there are detached portions of syenite, 

 often of a granitic texture and composition, lying on 

 the secondary strata, frequently limited to a very small 

 space, and now separated by a considerable interval 

 from the larger masses with which they were probably 

 once connected. Since it was also proved that this 

 rock was very subject to decomposition, to such a 

 degree indeed as even to render its annual waste 

 sensible, there is little doubt that it once covered many 

 parts of the secondary strata from which it has long 

 since disappeared. Should circumstances remove the 

 very few portions still displaying its actual superposition 

 to these strata, there would be no criterion by which 

 to judge of its overlying nature, and it might therefore 

 be mistaken for a mass of granite of an origin prior 

 to the secondary rocks ; a supposition that would be 

 strongly confirmed by its granitic composition. On 

 such a supposition the syenite of Sky becomes the 

 granite of Arran : it only remains to imagine that 

 those changes which time has not yet completed in 

 the former have been accomplished in the .latter, and 

 the granite of Arran becomes the syenite of Sky. 

 This question is not visionary : but it must not be 

 pursued with visionary arguments, and I shall therefore 

 leave it for future and increased experience. Geologists 

 will have no more reason for surprise if the former 

 continuity of the granite, the porphyry, and the trap 

 of Arran, should be established, than by finding that 



