Alumina with Silica, Lime and Magnesia. 323 



were actually formed above 130°. — At ordinary temperatures 

 tridymite, having inverted in the solid state, is intricately 

 twinned and intergrown, and its optical examination is not as a 

 rule satisfactory, especially on artificial crystals. The birefring- 

 ence is weak and the average refractive index about \£77* 



Crystal aggregates formed out of pure melts of Si0 2 or from 

 Si0 2 glass or by inversion of quartz heated to a high tempera- 

 ture, show the above properties except that the refractive 

 index is slightly higher, about 1* 484 ±' 003 instead of 1*477. 

 In the description of the tridymite from the lime-silica series, 

 the writer noted this higher refractive index but was unable 

 to account for it. Through the courtesy of Professor Lacroix 

 of Paris, however, to whom specimens of the artificial crystals 

 had been sent, this difference can now be explained. Pro- 

 fessor Lacroix, after examination of the material, pronounced it 

 to be in all probability cristobalite, and not tridymite, and sub- 

 sequent examinations here have confirmed Professor Lacroix' s 

 determinations. Cristobalite has been found in nature only 

 rarely, and then usually together with tridymite. Its crystals 

 are octahedral in habit, but, like tridymite, are intricately 

 twinned and very weakly birefracting, so that the optical 

 examination is not satisfactory. Its refractive index is slightly 

 higher than that of tridymite, about 1-49. f The optical 

 behavior of cristobalite was first studied by Mallard,:): who 

 found that at about 175 u the crystals became abruptly iso- 

 tropic, and remain so at higher temperatures. On cooling the 

 reverse process takes place, /3-cristobalite changing back to 

 the a form abruptly, the minute birefracting patches flashing 

 up throughout the entire slide as the inversion temperature is 

 reached. The volume change on this inversion is apparently 

 very slight. The fact that this change is reversible and that 

 natural crystals of cristobalite are octahedral in habit indicates 

 that they were in all probability formed above 175°, the inver- 

 sion temperature. 



In the irregular crystalline aggregates obtained in labora- 

 tory preparations, tridymite and cristobalite can best be dis- 

 tinguished by heating in the thermal microscope. At about 

 130° tridymite becomes uniaxial and remains so at higher 

 temperatures, while in cristobalite no change occurs until 

 about 175°, when the interference colors disappear completely, 

 the material becoming isotropic and remaining so at higher 

 temperatures. The refractive index of tridymite (1*477) is 

 slightly lower than that of cristobalite (about 1'484), but the 

 difference is not great, and ordinarily would not, perhaps, be 

 relied on to distinguish the two in very fine powder. 



* Mallard, Bull. Soc. Min., xiii, 169, 1890. 

 fGaubert, Bull. Soc. Min., xxvii, 244, 1904. 

 {Bull. Soc. Min., xiii, 175, 1890. 



