452 Andersen — System Anorthite-Forsterite-Silica. 



In the same way we conclude that if a fourth component 

 were added to a ternary system containing forsterite and silica,* 

 there would be a reaction field of forsterite in the resulting four- 

 component system. In general, therefore, the relations of for- 

 sterite may be expressed as follows : In any system, of any 

 number (n) of components among which are forsterite and 

 silica, there will be a reaction field of forsterite. This reaction 

 field may extend over a part of the n-component forsterite 

 fieldf only, but may also embrace all this field and parts of the 

 adjoining fields. 



In the preceding discussion we have only considered pure 

 forsterite, which is a rare modification (if occurring at all) 

 among natural olivines in rocks, most of which contain a con- 

 siderable amount of iron silicate, although the magnesium 

 silicate is by far the predominating one. We know, however, 

 from the relations of olivine minerals in rocksj and from their 

 physical and chemical properties,§ that there is a continual 

 series of solid solutions between the forsterite (Mg 2 Si0 4 ) and 

 the iron silicate fayalite (Fe 2 Si0 4 ). In other words, there is 

 no boundary between a forsterite field and a fayalite field, as 

 there is only one olivine field in all systems where olivine 

 occurs. What is stated about the possible field relations of 

 forsterite, therefore, also holds true for olivine in general, and 

 thus we may draw the following conclusion : In a system con- 

 taining the components of an olivine- and pyroxene-bearing 

 rock there will always be a field (reaction field), the melts of 

 which solidify in such a way that the olivine is resorbed during 

 the cooling. 



It now remains to discuss whether it is probable that the 

 compositions of any actual rocks fall within the reaction field 

 of olivine. We consider the system anorthite-forsterite-silica. 

 By adding different amounts of iron oxides, alkalies and some 

 minor constituents to certain mixtures of this system we come 

 very near the compositions of some actual rocks like gabbros 

 and anorthosites. If we, for example, added between 5 and 

 10 per cent iron oxides and alkalies to mixtures containing 

 70-90 per cent anorthite and lying within the reaction field of 

 forsterite, and made some minor changes in the relations of the 

 other constituents, we would obtain the composition of actual 

 olivine-bearing anorthosites. In the same way small changes 



* The reaction consists in formation of MgSi0 3 by dissolution of forsterite 

 and cannot appear in systems which do not contain MgSi0 3 , that is, forster- 

 ite and silica, or MgO and Si0 2 . 



f By field is here simply meant all compositions of melts from which for- 

 sterite separates as the primary phase. In other connections the term has a 

 similar meaning. 



XJ. H. L. Yogt, Tsch. Min. petr. Mitt., xxiv, 481, 1905. 



§ H. Backhand, Trav. Mus. Geol. Pierre le Grand pres l'Ac. Imp. Sc, 

 St. Petersbonrg, 1909, iii, 77. 



