Andersen — System Anorthite-Forsterite- Silica. 425 



up. The quenchings showed that this mixture corresponds 

 exactly with the composition of the point as appears from 

 Table 3. The temperature of the point of intersection was 

 1444°±2°. 



The point was determined in a similar way to be An 54, 

 Fo 46 at the temperature 1466° ± 4°. 



The highest point determined on the forsterite liquidus was 

 at 1620° in the mixture An 40, Fo 60. Between this mixture 

 and pure forsterite, whose melting point was previously found 

 (in the iridium furnace) to be 1890°, no determination could 

 be made. Nevertheless the liquidus of forsterite may be 

 considered rather sharply fixed by the determinations stated 

 in Table 3, so much the more as it appears that all points are 

 consistent with each other. 



The liquidus of spinel was determined by the points B and 

 C, and three intervening points as appears from Table 4 and 

 fig. 3. It need not be further explained that the maximum 

 shown on this liquidus has no relation whatever to the maxi- 

 mum of ordinary two-component systems with a binary com- 

 pound. The spinel, MgAl 2 4 , can, of course, not be considered 

 a binary compound in a system where MgSi0 4 and CaAl 2 Si 2 8 

 are components. 



The anorthite liquidus was determined by the points A 

 (melting point of An) and B, and the liquidus point of the 

 intervening mixture An 90, Fo 10. 



The mixtures between A and B crystallize with anorthite as 

 primary phase. At the point B spinel begins to. separate and 

 at the same moment the system ceases to be binary, that is to 

 say, the composition of the liquid can no longer be expressed 

 in terms of the two-component system anorthite-forsterite. 

 The further course of the crystallization, therefore, belongs 

 in the quaternary system MgO-CaO-Al 2 O s -Si0 2 and can not 

 be discussed on the basis of the present data. Mixtures 

 between C and D behave in a similar way. Forsterite is the 

 primary phase and at the point C spinel begins to crystallize. 

 In the mixtures between B and C spinel is the primary phase, 

 and in consequence, the relations of these mixtures can not be 

 expressed in terms of the binary system. At temperatures 

 immediately below the liquidus B C spinel is the only solid 

 phase. At lower temperatures other phases occur together 

 with spinel in equilibrium with the liquid.* 



*It is obvious that the temperatures of the points B and C have no signifi- 

 cance whatever for any of the mixtures between B and C. These temper- 

 atures fix points in the binary system and the mixtures mentioned belong 

 in the quaternary system for all temperatures below the liquidus. 



