568 N. L. BOWEN 



definitely that a liquid saturated with a certain niember of a reaction 

 series is effectively supersaturated with all preceding members of 

 that series. It cannot dissolve such members but can only react 

 with them to convert them into the members with which it is satu- 

 rated. The reaction is not a simple subtraction from the liquid 

 of the material necessary for this transformation, but some precipi- 

 tation from the liquid itself is involved and the liquid ordinarily 

 maintains its position on the same saturation surface. The products 

 of crystallization from the liquid and the possible course of frac- 

 tional crystallization are thus unaffected. 



On the other hand, a liquid saturated with a certain member of a 

 reaction series is unsaturated with all subsequent members of the 

 series. Inclusions consisting of these later members can become a 

 part of the liquid by a sort of reactive solution, the heat of solution 

 of inclusions being supplied by the precipitation of their heat equiv- 

 alent of the member of the series with which the liquid is saturated. 

 It should be noted that the material that can by this reactive process 

 become a part of the liquid must consist of a later member of the 

 reaction series, that is, must be material toward which the liquid 

 could pass spontaneously by fractional crystallization. The net 

 effect upon the liquid is, then, to push it onward upon its normal 

 course. 



In Table II the products of crystallization of subalkaline mag- 

 mas are arranged as reaction series, as definitely as may be in such 

 complex series. The action of magmas upon foreign inclusions of 

 igneous origin may be deduced from this arrangement of the crystal- 

 line products as series by application of the principles developed 

 from the above study of simple systems. Thus we find that a 

 granitic magma saturated with biotite cannot dissolve oUvine, 

 pyroxene, or amphibole, but can only react with them to convert 

 them into biotite, the phase with which it is saturated. Or, stated 

 more generally, no saturated magma can dissolve inclusions con- 

 sisting of minerals belonging to an earlier stage of the reaction series 

 (usually more basic) . 



Saturated basic magma, on the other hand, will react with inclu- 

 sions belonging to a later stage of the reaction series (more acidic), 

 the reaction being of such a nature that the inclusions become a part 



