TRANSACTIONS OF SECTION C. 571 



point, and tlie water Had passed into a liquid, supersaturated _ witli_ alkaline 

 clilorides, wliicli have crystallised out as small cubes in the fluid-cavities, just as in 

 the case of minerals in some of the blocks ejected from Vesuvius. 



Confining our attention, then, to extreme cases, we thus see that rocks of the 

 granitic type difier in a most characteristic manner from the products of artificial 

 igneous fusion, both in the structure of the crystals and in containing liquid water, 

 inclosed at the time of their formation. The question then arises, whether these 

 differences were due to the presence of the liquid water, or whether its presence 

 and the characteristic structure were not both the effects of the great pressure of 

 Buperincumbent rocks. I do not see how this can be decided in a perfectly satis- 

 factory manner, but must confess that I am inclined to believe that, wliilst great 

 pressure was necessarily the reason why the water did not escape as vapour, the 

 presence of liquid water during final consolidation must have had a very consider- 

 able influence in modifying the structure of the rock, and had a great share in 

 developing what we may call the granitic type. 



It would be very instructive to follow out the gradual passage from one 

 extreme type to another far more completely than is possible on the present 

 occasion. The most interesting examples of rocks intermediate between the 

 granitic and volcanic types that I have been able to examine in adequate detail, 

 are the various Comish elvans and other quartz felsites, which furnish all but a 

 complete passage from pitchstone to granite. Some specimens prove that quartz 

 may crystallise out from and inclose a perfectly glassy base, without a trace of 

 liquid water ; and at the same time other specimens prove equally well that, as we 

 approach the granitic type, the quartz was not deposited from a glassy solvent, but 

 inclosed more or less water. In the few intermediate cases there appears to be 

 evidence of the conjoint presence of uncombined water and melted stony matter. 

 On the whole, if we take into consideration only the external form of the larger 

 crystals, rocks of the granitic type are very much as though the crystals met with 

 in trulj^ volcanic rocks had been strained out from the glassy or fine-grained base, 

 and the intermediate spaces filled with quartz. The internal structure of the 

 crystals is, however, very different, the cavities in one class containing glass, and 

 in the other water. This most essential and characteristic difl'erence proves that 

 rocks of the true granitic type cannot have been formed simply by the more com- 

 plete crystallisation of the general base of the rock. If the crystals in granite were 

 analogous to those developed in volcanic rocks, and the only essential difl'erence 

 was that the residue crystallised out more slowly and completely, so as to give 

 rise to a more coarsely crj-stallised base, the crystals first formed ought not, as I 

 think, to differ so essentially as that in one case they should inclose only glass, and 

 iu the other only water. Taking all into consideration, we can therefore scarcely 

 suppose that the crystals in granitic rocks were deposited from a truly-melted, dry, 

 glassy solvent, like those in volcanic rocks or in slags. 



General Remits. 



I have, I trust, now said enough to show that the objects here described may 

 be conveniently separated into three well-marked groups, viz., artificial slags, 

 volcanic rocks, and granitic rocks. My own specimens all show perfectly well- 

 marked and characteristic structures, though they are connected in some cases by 

 intermediate varieties. Possibly such connecting links might be more pronounced 

 in other specimens that have not come under my notice. I must, however, base niy 

 conclusions on what I have been able to study in an adequate manner, by examining 

 my own preparations, and leave it for others to correct any errors into which I may 

 have been led from lack of more numerous specimens. In any case the facts seem 

 abundantly sufficient to prove that there must be some active cause for such a 

 common, if not general, difference in the structural character of these three difierent 

 types. The supposition is so simple and attractive, that I feel very much tempted 

 to suggest that this difierence is due to the presence or absence of water as a gas 

 or as a liquid. In the case of slags it is not present in any form, Considering how 

 large an amount of steam is given oft' from erupted lavas, and that, as a rule, no 



