Cristobalite and I rid y mite. 



411) 



melting- point. In the ternary system diopside-fosterite- 

 silica lie found the cristobalite liquidus rose too steeply 

 from the boundary curve separating the silica field from 

 the pyroxene field to warrant such a low value, and sug- 

 gested that the melting' point probably was at a higher 



Fig. 2. 



Section A- A. . 



Sec+*on B-B 



■ Ward burned magnesia. 

 I I A\\_mdunr\^ 



\ 1 Marquord-V porce\avr\ , 



temperature even than that assigned to it by Endell and 

 Eieke. 



Our work on the system CaO-MgO-Si0 2 has confirmed 

 Bowen's results and has led to our undertaking the rede- 

 termination of the melting point of cristobalite. 



Such an investigation required a furnace capable of 

 maintaining a charge at temperatures above 1700° for 

 periods of time up to an hour in length. The iridium- 

 tube furnace will reach these temperatures but is 



