Ch. XXXIV.] OF THE ISLE OF AKRAK 731 



successively visited Arran, and they have carefully searched there, as 

 I have done myself, to find an exception, but in vain. The rounded 

 masses consist exclusively of quartz, chlorite-schist, and other mem- 

 bers of the metamorphic series ; nor in the newer conglomerates of 

 No. 3 have any granitic fragments been discovered. Are we then 

 entitled to affirm that the coarse-grained granite (No. 2), like the fine- 

 grained variety (No. 5), is more modern than all the other rocks of 

 the island ? This we cannot assume, but we may confidently infer 

 that when the various beds of sandstone and conglomerate were 

 formed, no granite had reached the surface, or had been exposed to 

 denudation in Arran. It is clear that the crystalline schists were 

 ground into sand and shingle when the strata No. 3 were deposited, 

 and at that time the waves had never acted upon the granite, which 

 now sends its veins into the schist. May we then conclude, that the 

 schists suffered denudation before they were invaded by granite? 

 The opinion, although not inadmissible, is by no means fully borne 

 out by the evidence. For at that time when the Old Red Sandstone 

 originated, the metamorphic strata may have formed islands in the 

 sea, as in fig. 755, over which the breakers rolled, or from which 



Fig. 755. 



Sea 



torrents and rivers descended, carrying down gravel and sand. The 

 plutonic rock or granite (b) may even then have been previously 

 injected at a certain depth below, and yet may never have been 

 exposed to denudation. 



As to the time and manner of the subsequent protrusion of the 

 coarse-grained granite (No. 2), this rock may have been thrust up 

 bodily in a solid form, during that long series of igneous operations 

 which produced the plutonic formations (No. 5), and some of the trap 

 dikes of the same age. 



We have shown that these eruptions, whatever their date, were 

 posterior to the deposition of all the fossiliferous strata of Arran. 

 We can also prove that subsequently both the granitic and trappean 

 rocks underwent great aqueous denudation, which they probably 

 suffered during their emergence from the sea. The fact is demon- 

 strated by the abrupt truncation of numerous dikes, such as those at 

 b, c, d, which are cut off on the surface of the granite and trap. 



The theory of the protrusion in a solid form of the northern nucleus 

 of granite is confirmed by the manner in which the hypogene slates 

 (No. 1) and the beds of conglomerate (No. 3) dip away from it on all 

 sides. In some places indeed the slates are inclined towards the 

 granite, but this exception might have been looked for, because trrse 

 hypogene strata have undergone disturbances at more than one geo- 



