MAGMATIC DIFFERENTIATION OF IGNEOUS ROCKS 663 



a and b will be displaced only very inconsiderably by the presence 

 of some dissolved water.^ 



We have here supposed that H2O enters as an independent com- 

 ponent (H2O). But it may also be thought that H2O, CO2, HCl, 

 etc., under certain conditions, partly also may form special combina- 

 tions (for instance HjSiOj, H4Si04, H Al Si04, etc.), which may be 

 broken up during the run of the crystallization. Hereby there is 

 a possibility for complications which will be very difficult to explore. 



In a binary system (under high pressure) of a light volatile 

 component, as, for instance, H2O, and a silicate, as Ab, An, Or, 

 Ca Mg Si204, etc., even a little H2O will effect a relatively consider- 

 able depression of the melting point of the silicate, and this depres- 

 sion is nearly proportional to the quantity of H2O, etc. We refer 

 to the explanation given by Niggli (1920), pages 34-35, and illus- 

 trated by his Figure 2, II and III. And in general, the light volatile 

 components (H2O, CO2, HCl, etc.) will effect a tolerably consider- 

 able depression of the crystallization interval. 



From theoretical reasons it must be presupposed that the light 

 volatile components may effect a very considerable increase of the 

 thin-fluidity. This is, as far as I know, not verified by laboratory 

 experiments, but that it is the case, may inter alia be proved by the 

 fact that the magma of granite-pegmatite dikes, in spite of the low 

 crystallization interval (about 8oo°-7oo°) must have been extremely 

 thin-fluid. 



With regard to the quantity of the light volatile components in 

 the various eruptive magmas, we may present the following general 

 consideration. 



By crystallization-differentiation in a parent-magma with a 

 certain percentage of light volatile components (as, for instance, 

 H2O), the first separated crystals will, in most cases because of their 

 increased density, sink into a deeper-lying and higher-heated 

 magma-layer where they are dissolved (or resorbed) . As we shall 



I In this connection we will cite a portion of the conclusion in the treatise of P. 

 Niggli, "Die Gase im Magma" {Centralhl. f. Min., etc. [1912], pp. 337-38). "Es 

 ergiebt sich aus den (physikalisch-chemischen) Versuchen dass in vielen Fallen (hoher 

 Druck) die Gasmineralisatoren wie eine andere Komponente behandelt werden 

 konnen, dass also Vogt's Ansicht, dass einfache Schmelzfliisse schon viele petro- 

 graphische Probleme erleuchtern, richtig ist." 



