THE TRANSFORMATION OF ELEMENTS 



83 



the reaction. The apparatus was modified by the removal of the 

 phosphorus pentoxide tube where water, or gases reacting with 

 phosphorus pentoxide, were to be treated. 



In all cases, by raising the reservoir h, the gases were forced 

 through the apparatus, past the stopcock k, into a measuring 

 tube, as in fig. 3, a. This contained a piece of blue glass, sealed 

 to the tube, and broken off at a very fine point a, to which 

 the mercury surface was accurately adjusted by raising or 

 lowering the reservoir m (fig. 2). The difference in height 

 between the mercury surface at a and that in the reservoir, 

 subtracted from the barometric pressure, gave after the neces- 

 sary temperature corrections the pressure exerted at o° by the 

 perfectly definite volume of gas enclosed. Readings of the 



r\ 



Fig. 3. 



pressure were taken daily. In each the tap was opened, 

 the mercury readjusted to the point a, and the tap then closed, 

 so that the gas was kept at constant volume throughout the 

 experiment. By calibrating the measuring tube with mercury 

 at the end of the experiment, the capacity was found. The 

 observed pressures then gave the actual volume of gas at o° and 

 760 millimetres pressure, at each time of observation. 



After three or four weeks, the amount of emanation re- 

 maining was practically negligible ; by this time also the 

 pressure became constant within the limit of error. When 

 water and gas were both present the point a was set to the 

 surface of the water as in fig. 3, b. In this way, while the tap 

 b was closed, the gas and water phases, c and d, were both kept 

 at constant volume. There is considerable reason to believe 

 that emanation decomposes in different media at the same rate. 



