418 Andersen — System Anorthite-Forsterite-Silica. 



tion of variations in the lower inversion points between a and 

 ft cristobalite. This hypothesis suggests that cristobalite con- 

 sists of two or more molecular modifications in one crystal and 

 that variable conditions of formation may bring about variable 

 molecular constitutions with resulting variations in inversion 

 point and melting point. 



Cristobalite exists* in two modifications, the a form and the 

 ft form, with an inversion point varying between 198° and 

 275°. The high temperature (ft) modification is regular, form- 

 ing distorted crystals and aggregates, usually with the octahe- 

 dron as the most common crystal form, sometimes with twins 

 after the spinel law. The low temperature (a) form is slightly 

 birefringent ; 7 == 1*487; a = 1*484. 



Tridymitef also forms two modifications, w T ith an inversion 

 region between 117° and 163°. The high temperature (ft) 

 form is probably hexagonal, the low temperature (a) form prob- 

 ably orthorhombic. The crystals generally form hexagonal 

 plates, often combined to characteristic wedge-shaped twins. 

 Optical properties (of a form), 7 = 1*473 ; a (and ft) = 1*469 ; 

 2Y = 35*8°. 



In the preparations of the anorthite-forsterite-silica system 

 the silica crystals, formed in the accessible temperature regions, 

 were so small that it was, in general, very difficult to dis- 

 criminate between the two phases. As the crystals were 

 always imbedded in glass, the refractive index could not be 

 determined, but it was always much lower than that of the 

 glass (which varied between 1*50 and 1*53). The crystals were 

 either isotropic or showed a very faint double refraction. In 

 some cases, at temperatures around 1500°, excellent crystals 

 with cube and octahedron in equal development were seen. 

 These crystals were perfectly isotropic and very probably repre- 

 sented the ft form of cristobalite. In most cases, however, the 

 crystals formed rounded grains without distinct faces, mostly 

 isotropic, but sometimes showing indication of double refraction. 

 At the lowest temperatures where silica could be obtained as 

 the primary phase, in mixtures near the ternary eutectic, anor- 

 thite-clinoenstatite-silica (1222° to 1300°) the crystals appeared 

 like irregular patches, more frequently showing double refrac- 

 tion than those at higher temperatures. These crystals were 

 probably tridymite. 



Glinoenstatite. 



ClinoenstatiteJ (MgSi0 3 ) breaks up at 1557°, forming for- 

 sterite (Mg 3 Si0 4 ) and liquid. The last of the forsterite melts 



at 1577°. 



*C. N. Fenner, loc. cit, p. 360. 

 f C. N. Fenner, loc. cit., p. 372. 

 X N. L. Bowen and Olaf Andersen, loc. cit., p. 495. 



