SOSMAN AND HOSTETTER: A VACUUM FURNACE 



283 



pressure in millimeters of mercury in the fourth column. The 

 degree of reproducibility of the pressures is shown by the second, 

 third, fourth, and fifth measurements^. The pressures at depths 

 of 85, 92, and 95 mm. are practically constant. Since the depth 

 of the crucible is 18 mm. there seems therefore to be a range 

 of about 25-30 mm. in which the temperature is uniform within 

 one degree, judging from the variation of pressure, which is of 

 the order of magnitude of 0.01 mm. per degree under the con- 

 ditions of the tests. For subsequent measurements the crucible 

 was always placed in the middle of this zone, and the thermo- 

 element brought near the center of gravity of the charge as 

 explained above. The charges varied in depth from 5 to 10 



mm. 



TABLE I 



Pressures obtained from Merck's FeiOz at 1100° at various levels in furnace 



Glass stopcocks are the most troublesome part of a vacuum 

 apparatus. We have found Gundelach's large right-angle stop- 

 cocks (Schiff's form) satisfactory for the present work. A 

 stiff rubber-parafRne-vaseline stopcock grease must be used, 

 however, to prevent the atmospheric pressure on the large sur- 

 face from causing the cock to stick fast. The grease should 

 be renewed occasionally, also, as it becomes gummy after use 

 for a long time in a vacuum. 



As Professor Morley has well said, a stopcock is usually nothing 

 more than a located leak. It is necessary, then, to shut off the 

 located leaks in the connections to pumps and oxygen supply 

 by a mercury cutoff, shown at C in figure 1. Its construction 



