424 Ferguson and Merwin — Melting Points of 



The superheating of quartz is perhaps better illus- 

 trated by experiment 3 than by any experiment that has 

 heretofore been made. In the case of pure substances 

 this phenomenon is probably due to the rigidity of the 

 crystals themselves and, unlike the solution of a crystal 

 in a melt of different composition, is independent of the 

 viscosity of the melt. 



Cristobalite formed at a temperature of 1350° behaved 

 in a similar manner to that formed at higher tempera- 

 tures. If Fenner's 9 conception of the molecular struc- 

 ture of cristobalite is correct, the molecular readjustment 

 must take place in the crystals themselves. 



The melting point 10 of cristobalite is placed at 

 1710° ± 10°. This is as narrow a limit as the observa- 

 tions will permit, although the temperature measure- 

 ments were more accurate. (See results on tridymite.) 



II. Tridymite. 



In an attempt to duplicate the observations of Le Cha- 

 telier 11 upon the tridymite-cristobalite inversion we were 

 fortunately able to obtain and examine microscopically 

 specimens of silica bricks from two types of glass fur- 

 naces which had been in continuous operation for a long 

 time. The one specimen from the crown of a furnace of 

 the regenerative type in which the direction bf the flue 

 gases is reversed every 15 or 20 minutes contained good 

 crystals of tridymite while the other specimen from the 

 crown of a furnace of the recuperative type contained 

 cristobalite. There could be no question but that this 

 latter specimen was subjected to a higher temperature 

 than the former since the flames continuously played 

 upon it and it was not affected by the cold air which 

 leaked into the furnace through the crevices about the 

 furnace door. 



These observations do not agree with those made by 

 Le Chatelier, and we therefore decided to determine the 

 melting point of tridymite which if obtained would 

 clearly indicate the relation this form of silica bears to 

 cristobalite. 



9 N. L. Bowen, this Journal, 38, 218, 1914; C. N. Fenner, ibid., 36, 366, 

 1913. 



10 The softening points of silica bricks were found by C. W. Kanolt to be 

 1700, 1705, 1700 respectively. Technologic paper No. 10, Bur. Standards 

 1912. 



11 Henry Le Chatelier, Bull. Soc. Fr. Min. 1917, p. 44. 



