44 



MR. O. \V. RICHARDSON ON THE IONISATION 



of a very complex question. So far as the negative leak is concerned, the ground 

 has been previously examined, and valuable results obtained by H. A. WILSON.* 



The hydrogen used in the experiments was prepared for the most part by the 

 action of pure zinc on pure hydrochloric acid with a little ferric chloride added to 

 make the action go. It was passed over solid potash and cotton wool before being 

 admitted into the apparatus. In some cases, when only a little hydrogen was 

 required, it was admitted into the apparatus by diffusion through the walls of a 

 platinum tube heated in a spirit-lamp flame. This method is a very convenient one 

 for obtaining a small quantity of hydrogen in a high degree of purity. As is well 

 known, the negative ionisation produced by hot platinum in an atmosphere of hydrogen 

 is very big, and it was found impossible to use the electrometer in the ordinary way, 

 with a condenser attached to the quadrants, in order to measure the currents 

 obtained. These currents, which amounted to as much as 10~ 4 ampere in some cases, 

 were measured by taking the steady deflection when the electrometer quadrants were 

 connected by a high resistance which could be varied from one to one million ohms. 

 This was a very convenient arrangement, as it enabled the same instrument, by 

 simply changing a few plugs, to be used for measuring any current from 10" 1 to 

 10~ 14 ampere. 



Interesting effects are observed when a wire, which has previously been heated in 

 oxygen only is first heated in hydrogen. Such a wire gives rise to an abnormally 

 high positive ionisation which gradually decays with time, whilst the negative 

 ionisation increases simultaneously. The kind of thing that occurs is typified by the 

 numbers in the subjoined table. The wire, which was O'l millim. in diameter and had 

 an effective area of 0*214 sq. ceiitim., had previously been heated in oxygen at 

 2 millims. pressure at a temperature of 800 C. Under these conditions it was 

 giving about 8'5 x 10~ 13 ampere positive leak, the negative leak being small compared 

 with this value. The wire was then allowed to cool, the oxygen pumped out, and 

 hydrogen let in to a pressure of 27 '5 millims. The currents, with potentials of 

 40 volts on the filament, were then measured at different times at a constant 

 temperature of 800 C., with the result shown by the table. 



It will be noticed that after the wire had been charged negatively, the positive 

 leak was abnormally high. This may be a spurious effect due to the insulation 



* ' Phil. Trans.,' A, vol. 202, pp. 243 et xtq. 



