THOMPSON AND LOMBARD. — NITROGEN IN CALCIUM CYANAMIDE. 2G3 



librium pressure. It was therefore decided to try an experiment in the 

 Heraeus furnace that had been used as an auxihary furnace for cali- 

 brating the Wanner pyrometer. Of course this could be used only at 

 comparatively low temperature. It was glazed only on the inside, and 

 it was anticipated that it would leak. 



The charge was contained in a graphite boat 1.5 inches long. It 

 consisted of 6.14 grams of a mixture of calcium carbide and calcium 

 chloride containing 10 per cent of the latter. The ends of the porce- 

 lain tube of the furnace were closed with rubber stoppers ; the stoppers 

 and ends of the tube were shellacked. A small glazed porcelain tube 

 for holding the thermo-electric couple extended through the larger 

 tube and the two stoppers. The j unction was pushed in until it was 

 opposite the graphite boat, in the middle of the furnace. 



August 10th, at 5 p. m., the furnace was pumped out to a pressure 

 of 1 millimeter and left till August 11th, 16 hours later, when it had 

 leaked to 13.1 millimeters. The furnace was then heated to about 

 1070° and pumped out a number of times, after which the gases re- 

 maining were swept out with nitrogen. While still hot, nitrogen was 

 admitted to 110 millimeters. Absorption was taking place rapidly, 

 and the following observations were taken. 



It seems that here the absorption of nitrogen at first overbalanced 

 the leakage. The following day, minima of 26.8, 30, and 31.5 milli- 

 meters were obtained, starting with pressures of 145, 165, and between 

 100 and 200 millimeters respectively ; and the best value of course 

 would be obtained by starting very little above the equilibrium pres- 

 sure, so that there would not be so much time for leakage to take place. 



Although these determinations were somewhat in error from the 

 leakage, this error was evidently not great, as the pressures agreed 

 well with those obtained in the vacuum furnace. The temperature 

 measurements were more reliable than the others, as they were made 

 directly with the thermo-electric junction. The lowest pressure, 21.5 



