Oxygen 



251 



that in the bell- jar were in communication with one 

 another via this neck. The retort was then heated 

 cautiously by means of the furnace to a temperature 

 just short of that at which mercury boils. After some 

 hours the surface of the mercury became covered with 

 reddish specks, and a simultaneous rise in the level of 

 the mercury within the bell-jar took place. The heating 

 was continued for a considerable time, and at length 

 no more of the red substance seemed to be formed and 

 the mercury rose no further in the bell- jar. To make 

 quite certain that no more of the air could be caused to 



Fig. 80. 



disappear, the heating was kept up for over eleven days. 

 The apparatus was then allowed to cool and the decrease 

 in volume of the air contained in the bell- jar was 

 measured as carefully as possible. All the red powder 

 was next removed from the surface of the mercury in 

 the retort, placed in a smaller retort, and heated 

 strongly. A gas was evolved and this was collected in 

 a graduated measuring jar. Mercury alone was left 

 in the smaller retort as a result of this treatment and 

 the volume of the gas given off was found to be exactly 



