Cristobalite and Tridymite. 423 



1705° . 



30 min. 



34 Glass. 



1714° 



30 min. 



Trace of crystals + glass. 



1714° 



10 min. 



Trace of crystals -f- glass. 



7. Cristobalite made by heating clear quartz crystals for 144 

 hours at temperatures ranging from 1300 to 1400°. Two lots 

 were made at the same time, one from lumps a few mm. in diam- 

 eter, the other from powder 0-5 mm. and finer. At the end of 

 100 hours there was present a few percent of unconverted quartz 

 in each lot, and at the end of 144 hours traces of quartz. Also 

 during the final selection of material free from quartz, some 

 of the larger grains were found to have hard, clear nuclei of 

 tridymite. The cristobalite was very white and friable. 



Temperature Sample Quenched 



(corrected). 



after : 



Results. 



1695° 



22 min. 



No melting detected. 



1707° 



15 min. 



A few grains showed traces of 

 melting ; others were melted 

 into clear beads. 



1715° 



18 min. 



All melted to one mass of clear 

 glass. 



After heating No. 4, a calibration at the Pd point gave 

 16040 microvolts, after No. 6, 16025 microvolts, and after 

 No. 7, 16080 microvolts. The high value after No. 7 was 

 the result of the cutting out of the most contaminated 

 portions of the thermoelement prior to the No. 7 experi- 

 ments. A standard element gives 16140 microvolts. 



Discussion of Results on Cristobalite. 



The fluxing of cristobalite crystals above the liquidus by 

 a ternary melt very high in silica appears to be slow com- 

 pared with the growth of cristobalite crystals in a simi- 

 lar melt the same number of degrees below 7 the liquidus. 

 A possible explanation of the phenomenon may be that 

 during fluxing the crystal is surrounded by a thin film of 

 greater silica content and during crystallization a thin 

 film of less silica content than the average composition of 

 the melt. The rate mentioned W'Ould be a function of the 

 viscosity of this liquid film. 



The silica-glass tube (No. 2 above), although supposed 

 to be very pure material, carried enough impurity to give 

 a large temperature interval of melting. The highly re- 

 fracting specks were evidence of impurity, and possibly a 

 similar explanation may account for Fenner's results. 



