SOSMAN AND HOSTETTER: A VACUUM FURNACE 277 



PHYSICAL CHEMISTRY.— A vacuum furnace for the measure- 

 ment of small dissociation pressures. R. B. Sosman and 

 J. C. HosTETTER, Geophysical Laboratory. 



The study of those sUicate systems which contain iron requires 

 the measurement of a wide range of oxygen pressures. The 

 first problem in this connection is the study of the dissociation 

 pressures of the oxides of iron themselves. For this purpose 

 the apparatus described below has been developed. 



Vacuum furnaces in considerable variety have already been 

 described, but none of these was exactly suited to the work in 

 hand. Furnaces like those of Arsem, Ruff, and numerous others, 

 in which the heating element is of graphite or carbon, are of 

 course out of the question when iron and oxygen are to be studied. 

 Tungsten and molybdenum furnaces, like that of Birnbrauer, 

 cannot be used with oxygen. Slade's platinum tube furnace is 

 designed for a horizontal heating tube, which is not suitable for 

 the quenching of silicate melts by dropping them into mercury. 

 Another disadvantage of Slade's furnace is the softness of plati- 

 num at high temperatures, permitting the heating tube to be 

 easily deformed. 



Our furnace and accessory apparatus are shown in somewhat 

 diagrammatical section in figure 1. The design could no doubt 

 be improved, as some parts are the product of evolution rather 

 than of original design, but the description covers the apparatus 

 as it is actually being used for the measurement of small disso- 

 ciation pressures. In principle, the furnace is similar to that 

 of Slade.^ It consists essentially of two parts: (1) the furnace 

 tube, which serves both as the furnace wall inclosing the "inside 

 vacuum" and as the heating element; (2) the water-cooled jacket, 

 which surrounds the furnace tube and incloses the ''outside 

 vacuum." ^ 



An alternating current at low voltage is sent through the 

 furnace tube, which is made of an alloy of 80 parts platinum and 



^ R. E. Slade, An electric furnace for experiments in vacuo at temperatures 

 up to 1500°. Pr. Roy. Soc. London. 87 A : 519-524. 1912. 



