198 CHEMISTRY OF THE EARTH. 



been described in §23. While the crystalline stratified rocks are but 

 slightly porous the unaltered strata hold large quantities of water in their 

 pores. The mean of thirty-six determinations npon sandstones, shales, 

 limestones, and dolomites from twenty-five different localities among 

 the nnaltered paleozoic sediments of Canada shovred that 7.75 volumes 

 of water were held in 100 volumes of the thoroughly-moistened rock. 

 The proportion varied from less than 1.0 per cent, iu the more com])act 

 limestones, to 10.0 and even 21.0 per cent, in the sandstones, an amount 

 which is greatly exceeded in some more recent limestones.* A large 

 proportion of the ocean's waters is thus imprisoned in the vast volume 

 of unaltered sediments, and set free when these become metamor])hosed, 

 a process which is attended with a corresponding reduction of volume. 

 In addition to this, moreover, the clays and other hydrated silicates lose 

 a large part of their chemically-combined water during metamorphism, 

 and become changed into crystalline compounds of increased density. 

 This becomes obvious when Ve compare the specific gravity of such 

 species as garnet, epidote, chloritoid, staurolite, andalusite and kyanite 

 with that of the unaltered sediments in the midst of which they are gen- 

 erated. From this condensation, then, as well as from the mechanical 

 contraction consequent upon the expulsion of water, the metamorphism 

 of sediments is attended with a very considerable diminution of bulk, 

 which is not without geological significance. It results from the exper- 

 iments of Sorby (§ llfthat such chemical changes as are accompanied by 

 condensation or diminution of volume are favored and accelerated by 

 pressure, which may thus become a direct agent in promoting meta- 

 morphism as well as solution. 



§ 35. The crystallization which takes place in sedimentary rocks not 

 uufrequently effaces more or less completely the traces of their stratified 

 and sedimentary origin, as is seen, for example, in many gneisses, which 

 are scarcely distinguishable from granite. The study of such rocks, 

 moreover, affords abundant proof that tliis alteration has been attended 

 with such a softening that the material has been molded by pressure, 

 forced into fissures or openings in less fusible or less heated strata, and 

 thus taken the form of what is designated as eruptive rocks. The action 

 of heat upon sedimentary rocks is not, however, confined to condensa- 

 tion, crystallization, and softening; strata in which carbonates, sulphates, 

 chlorides, and carbonaceous substances are mingled with sdicious and 

 argillaceous matters, will, at a sufficiently-elevated temperature, in the 

 presence of water, undergo such changes as must liberate carbonic acid, 

 hydrochloric acid, and sulphuretted hydrogen, which are the common 

 gaseous accompaniments of volcanic action. From these considerations 

 we are led to a rational theory of volcanic and eruptive rocks, which we 

 conceive to have their seat, not in an uncongealed portion of the once 

 liquid globe, but in the more deeply-buried portions of that disinte- 

 grated crust whose origin has been explained in § 14. 



§ 36. The history of this theory forms an interesting chapter in 

 geology. As remarked by Humboldt, a notion that volcanic phenomena 

 have their seat in the sedimentary formations, and are dependent on the 

 combustion of organic substances, belongs to the infancy of geology. 

 To this period belong the theories of Lemery and Breislak, {Cosmos, v. 

 443; Otte's translation.) Keferstein, in his NaturgescMclite des ErdVOr- 

 lyers, published in 1834, maintained that all crystalline non-stratified 

 rocks, from granite to lava, are products of the transformation of sedi- 

 mentary strata, iu part very recent, and that there is no well-defined 



* Geologica 1 Report of Canada, 1866, p. 283.— American Journal of Science, [2,] 

 xxxix, 183. 



