388 PEOE. J. W. JTJBD ON THE TEETIAET AND 



which are operating simultaneously upon the deep-seated rock- 

 masses below. 



The circumstance of the crystals of such rocks being full of 

 numerous cavities, many of which contain saturated solutions of the 

 alkaline chlorides and sulphates, or are filled with solid substances 

 like woUastonite and other silica-compounds, and of such cavities 

 also enclosing carbonic acid in a liquid condition, all bear witness to 

 the presence of these solvents and to the potency of their action 

 under pressure. 



The study of these deep-seated rocks at different depths shows 

 that the water and other solvents which permeate the whole of the 

 crystals under the enormous pressure attacks certain silica-compounds 

 more readily than others ; probably the compounds of silica with iron 

 and the alkalies are among the first to pass into solution. Eventu- 

 ally the whole of the compound silicates yield to the solvent and are 

 broken up ; but this does not take place unifonnly through the 

 crystals. In the direction of certain planes within crystals the mole- 

 cules are in a state of less perfect stability with respect to chemical 

 agencies than others, and along these planes the solvents eat out for 

 themselves hollows (negative crystals) which become filled with the 

 hydrated silica, the hydrated ferric oxides, and other products of 

 decomposition. At increasing depths new planes within the crystal 

 become susceptible to the action of solvents, and fresh enclosures are 

 formed along them. Thus, at moderate depths, the only planes along 

 which augite-crystals are attacked by the solvents, and along which 

 they have enclosures (infilled negative crystals) formed, are the 

 planes parallel to the orthopinacoid ; but at greater depths planes 

 parallel to the clinopinacoid and the basal plane are similarly 

 attacked. The planes of chemical instability are not necessarily 

 identical in position with those of imperfect cohesion (cleavage-planes); 

 indeed in many cases, as we have seen, they are wholly diff'erent. 

 But in many cases the position of these planes of easy solubility in 

 a crystal are clearly modified by the twinned condition of the 

 crystal. 



In some cases Schillerization consists only in a redistribution of 

 the matter within the crystals. Thus diallage, as a general rule, 

 differs in composition from augite only by the presence of a per- 

 centage of water which, as we have already seen, is probably 

 combined with the materials which fill the negative crystals. But 

 in many cases the dissolved material may be carried away from the 

 crystal and deposited within the cavities of neighbouring crystals of 

 different species. In this way ferric oxide, probably derived from 

 the pyroxenes, olivines, and magnetite, comes to be deposited within 

 the negative crystals of labradorite. In the end this process of 

 Schillerization must result in the blending together in the most inex- 

 tricable manner of materials derived from different crystallized 

 minerals, and the whole characters of the rock may be completely 

 altered. 



It has long been known that the faces and cleavage-planes of crys- 

 tals are attacked by appropriate solvents in an unequal manner, so as 

 to give rise to the characteristic forms of two dimensions known as 



