706 GLASS-CAVITIES IN GRANITE. [Ch. XXXIH. 



tained even in the laboratory in a crystalline state by fusion. Mr. 

 Sorby, who has devoted much time and talent to the solution of this 

 and kindred problems, has come to the conclusion that it can be so 

 obtained. He informs me that he is convinced, by the examination 

 of quartz fused by Mr. David Forbes, that silica can crystallize in the 

 dry way, and he has found in quartz forming a constituent part of 

 some trachytes, both from Guadaloupe and Iceland, glass-cavities * 

 quite similar to those met with in genuine volcanic minerals, which 

 prove most conclusively that this quartz crystallized out from a fused 

 material like obsidian. 



By "glass-cavities" are meant those in which a liquid, on cooling, 

 has become first viscous and then solid without crystallizing or under- 

 going a definite change in its physical structure. Other cavities 

 which, like those just mentioned, are frequently discernible under the 

 microscope in the minerals composing granitic rocks, are filled some 

 of them with gas or vapor, others with liquid, and by the movements 

 of the bubbles thus included the distinctness of such cavities from 

 those filled with a glassy substance can be tested. 



Mr. Sorby admits that the frequent occurrence of fluid cavities in 

 the quartz of granite implies that water was almost always present in 

 the formation of this rock ; but the same may be said of almost all 

 lavas, and it is now more than forty years since Mr. Scrope insisted on 

 the important part which water plays in volcanic eruptions, being so 

 intimately mixed up with the materials of the lava that he supposed 

 it to aid it in giving mobility to the fluid mass. It is well known that 

 steam escapes for months, sometimes for years, from the cavities of 

 lava when it is cooling and consolidating. 



As to the result of Mr. Sorby's experiments and speculations on 

 this difficult subject, they may be stated in a few words. He con- 

 cludes that the physical conditions under which the volcanic and 

 granitic rocks originate are so far similar that in both cases they com- 

 bine igneous fusion, aqueous solution, and gaseous sublimation — the 

 proof, he says, of the operation of water in the formation of granite 

 being quite as strong as that of heat.f 



When rocks are melted at great depths water must be present, for 

 two reasons: First, because in a state of solid combination water 

 enters largely into the composition of some of the most common 

 minerals, especially those of the aluminous class ; and, secondly, be- 

 cause rain-water and sea-water are always descending through fissured 

 and porous rocks, and must at length find their way into the regions 

 of subterranean heat. But the existence of water under great pres- 

 sure affords no argument against our attributing an excessively high 

 temperature to the mass with which it is mixed up. 



Bunsen, indeed, imagines that in Iceland it attains a white heat at 



* See Quart. Geo!. Journ., vol. xiv. p. 465. 

 f Ibid., p. 488. 



