282 sosMAN AND hostetter: a vacuum furnace 



produced only 1100° without the tube, gave 1308° when the tube 

 was in place. 



The uniformity of temperature was not tested by direct meas- 

 urement, as a complicated arrangement of apparatus would have 

 been necessary to permit of moving the thermo-element up and 

 down in the completely sealed furnace. An equally sensitive 

 test was made, however, by taking a series of measurements of 

 the oxygen pressures produced by 0.5 gram charges of Merck's 

 ferric oxide, heated under similar conditions but at various levels 

 in the furnace. In our preliminary measurements of dissocia- 

 tion pressures the thermo-element junction was always placed 

 3 to 5 mm. above the top of the charge of oxide, so that the 

 change in weight of the charge could be determined without 

 uncertainty due to oxide adhering to the wires. It is obvious 

 that near the top of the furnace the thermo-element, under 

 these conditions, will be colder than the bottom of the charge, 

 whereas near the bottom of the furnace the reverse will be true. 

 Hence if the temperature of the thermo-element is set to the 

 same value in both positions, the temperature of the charge, 

 and therefore the pressure obtained, will be greater near the top 

 than near the bottom of the furnace. If two or more pressures 

 obtained near the middle of the furnace agree within the range 

 of reproducibility of these pressures, proof is afforded that there 

 is a space of uniform temperature equal to the distance between 

 the highest position of the thermo-element and the lowest posi- 

 tion of the bottom of the charge. Measurements can there- 

 fore be made on charges set in this part of the furnace with the 

 assurance that the temperature of the charge is uniform and 

 is represented accurately by the temperature of the thermo- 

 element. This condition is further assured by packing deep 

 charges into the crucible with a polished steel plunger which 

 shapes the surface with a central well, so that the thermo-element 

 can be placed actually within the charge without coming into 

 contact with it. 



The result of the uniformity tests is shown in Table I. The 

 distance from the top of the furnace tube to the top of the 

 crucible in millimeters is given in the third column, and the 



