236 



Prof. J. Patterson on the lonizatioii in 



but slightly with the temperature (pressure constant) up to 

 450° C, but above this it increased very rapidly, due to ions 

 being given off by the walls of the vessel, and the current 

 being much greater when the cylinder was charged negative 

 than when positive. The observations were not continued 

 beyond this. 



The air in the cylinder was at atmospheric pressure in all 

 the experiments, consequently it was necessary to determine 

 the variation of the ionization with pressure before deducing 

 its variation with temperature. This was done as follows : — 



The same cylinder and aluminium electrode were used, but 

 the arrangement was slightly modified as shown in fig. 4. 

 It was impossible to get the vessel air-tight, but the pressure 



Fi-. 4. 



did not vary more than one or two millimetres while taking 

 the readings. 



The c}dinder was connected to a water-pump and a mamo- 

 meter for measuring the pressure. Between the cylinder 

 and the w^ater-pump a tube filled with calcium chloride and 

 having a plug of glass wool at each end was inserted so that 

 the air in the cylinder was fairly dry and dust free. 



Experiments were also made with the cylinder filled with 

 hydrogen. The hydrogen was made from pure zinc. The 

 cylinder was first exhausted and then filled wdth hydrogen ; 

 this was repeated two or three times so that the hydrogen 

 would be fairty free from air. 



The method of observation was the same as before. For 

 pressures below 300 mm. of mercury the cylinder was charged 

 to a potential of 340 volts, and for pressures above this lOCO 

 volts were used. 



The results for a set of observations on air and hydrogen 

 are given in the following table. The results are plotted in 



