AUG 14 1916 



THE 



AMERICAN JOURNAL OF SCIENCE 



[FOURTH SERIES.] 



Art. XIII. — The Melting Points of Crist obalite and 

 Triclymite; by J. B. Ferguson and H. E. Merwin. 



I. Crist obalite. 



While investigating the system CaO-MgO-Si0 2 at 

 high temperatures, we were impressed with the need for 

 further work on the melting point of crist obalite, the high 

 temperature form of silica, and wish in this paper to 

 present some new results. 



In 1906 Day and Shepherd 1 upon heating finely pow- 

 dered quartz in an iridium furnace at different tempera- 

 tures observed a partial melting even at 1650°, and noting 

 the sluggish nature of this phenomenon estimated the 

 melting point of silica to be at 1625 °. 2 The experimental 

 conditions were such, however, that it is uncertain to 

 which crystalline form of silica this melting point 

 belongs. 3 



In 1912 Endell and Bieke 4 determined the melting- 

 point of cristobalite to be 1685° ±10°. They heated 

 crist obalite powder in an iridium furnace and measured 

 their temperatures by means of an Ir:Ir-Ru thermoele- 

 ment. They carefully calibrated this thermoelement at 

 the gold, palladium and platinum points, but such an 

 element under their working conditions could not remain 

 unchanged even during one experiment. The amount of 

 the resulting contamination and the direction of its ther- 



1 A. L. Day and E. S. Shepherd, this Journal, 22, 265, 1906. 



2 Recalculated on the basis of the temperature scale given by R. B. Sosman, 

 ibid., 30, 1-15, 1910. 



3 C. N. Fenner, ibid., 36, 381, 1913. 



4 K. Endell and R. Rieke, Z. anorg. Chem., 79, 239, 1913. 



Am. Jour. Sci.— Fourth Series, Vol. XLVI, No. 272.— August, 1918. 

 14 



