320 Alexander Scott — Saturation of Minerals. 



siliceous material which remains partly undissolved, while the quartz 

 basalts of California ' have been explained as due to the intratelluric 

 mixing of a dacite and basalt. This explanation has not been fully 

 accepted/ although the presence of a reaction-rim round the quartz 

 crystals indicates their probable xenocrystic origin. 



There are two possible Avays, however, in which primary quartz 

 (or other silica mineral) may be formed in unsaturated rocks in small 

 quantities. If water be present in a magma consisting mainly of 

 metasilicates, the hydrolitic action of the water vapour or its 

 dissociated constituents at high temperature may cause the meta- 

 silicates to partly break up into orthosilicate and free silica, and the 

 latter may finally crystallize as quartz or tridymite. 



The other possibility is that the reaction 



olivine + silica — > enstatite 

 may be appreciably reversible. It is most probable that all chemical 

 reactions are theoretically reversible, though, in many cases, under 

 suitable conditions a reaction can proceed so far in one direction 

 that the reverse reaction cannot be detected. Since practically all 

 chemical reactions involve an absorption or evolution of heat, it 

 follows that one of the actions in a reversible reaction is, in general, 

 endothermic and the other exothermic, and we can deduce, thermo- 

 dynamically, that a rise in temperature displaces the equilibrium 

 position in such a way that the endothermic reaction receives the 

 greater acceleration. The formation of a metasilicate by the inter- 

 action of silica and an orthosilicate is exothermic, and the reverse 

 reaction is therefore endothermic and will proceed to a greater extent 

 the higher the temperature. Thus the possibility of silica minerals 

 separating from a part-saturated rock will be greater when the 

 crystallization takes place at a high temperature than when it occurs 

 at a lower oiie.^ 



The disadvantages of a criterion such as this empirical one of 

 'saturation', and indeed of all similar empirical ones, become more 

 manifest when the stability relations of the minerals in a rock are 

 considered. So many factors, external and internal, influence the 

 cooling of a rock, that in many cases the resultant solid is not in 

 a condition of maximum stability, and one function of a rational 

 classification should be to indicate in what way, and to what extent, 

 there is deviation from a state of equilibrium. Por instance, many 

 rocks show evidence of undercooling, not only with regard to the 

 passage from the liquid to the solid state, but also with regard to 

 transformations in the solid state. Thus glass is not a true solid but 

 merely an undercooled liquid, and, as such, is a metastable phase at 

 ordinary temperatures. Devitrification, i.e. the transformation to the 

 more stable crystalline form, follows, in accordance with the Principle 

 of Maximum Work, the glass having a higher potential than the solid. 



^ Diller, Bull. U.S. Geol. Surv., 1891, No. 79, pp. 21-33. 



'\ Cf. Iddings, Bull. U.S. Geol. Surv., 1890, No. 66, pp. 20-32. 



■' Since this was written, Bowen & Andersen (Am. Journ. Sci., ser. rv, 

 xxxvii, pp. 487-500) have described experiments which show that on cooling 

 a melt of the composition Mg 0-Si O2, some forsterite separates out first and 

 then the residual liquid crystallizes as a mixture of silica and clinoenstatite. 



