36 MR. 0. W. RICHARDSON ON THE IONISATION 



IV. 11. THE IONISATION IN NITROOEN. 



The nitrogen uSed in the experiments was prepared from air and therefore 

 contained the inert gases in the proportion in which they occur to nitrogen in the 

 atmosphere. So far as was possible without using very elaborate refinements, care 

 was taken to free it from oxygen, hydrogen, carbon dioxide, dust and organic 

 impurities. The air of the laboratory was allowed to bubble very slowly through 

 a wash bottle containing alkaline pyrogallate and then passed over (l) a red-hot tube 

 about 60 centims. long containing copper and copper oxide, (2) a long tube packed 

 with cotton wool, and (3) a tube 30 centims. long packed with fused potash and 

 plugged with glass wool, before it was admitted into the apparatus. Inside the 

 apparatus it was finally dried over P 2 5 and further purified (1) at low pressures by 

 exposure to sodium amalgam from a cathode of which a discharge could be passed 

 from time to time, (2) at high pressures by heating a copper spiral red hot in it. 

 The copper spiral had previously been Seated in a vacuum to drive off occluded 

 hydrogen. With regard to the discharge from the sodium amalgam cathode it was 

 found advisable not to pass it too frequently, as it absorbs nitrogen rather quickly, 

 but not so quickly as it absorbs oxygen. All the purifying arrangements were 

 placed close to the testing bulb and between it and the rest of the apparatus so 

 as to be as efficient as possible. After the nitrogen had been prepared, the copper in 

 the hot tube was always found to be quite bright except at the front end. 



The positive ionisation in nitrogen is small compared with that in oxygen, especially 

 at low pressures. As we have seen, a small quantity of oxygen has a very big effect 

 on the positive leak, and it might therefore be supposed that the leaks observed in 

 nitrogen were due to traces of oxygen which had not been removed by the purifying 

 proceases. This, however, is rendered improbable by the fact that the ionisation in 

 nitrogen differs in character in certain important respects from that in oxygen. The 

 current requires a greater electromotive force to saturate it under similar conditions, 

 and it also showed time changes, after altering the applied E.M.F., similar to those 

 observed in a new wire. These changes were never observed in a wire which had 

 been heated in oxygen for a long time. 



On letting in fresh nitrogen the leak was always found to have a much bigger 

 value than that at which it finally settled down. The rate of decay was very big at 

 first, but after a time became extremely slow, so that there still seemed to be a slow 

 decrease after several hours' heating at 900 C. This decrease may have been due to 

 the gradual absorption of oxygen. No decrease in pressure could be observed, but 

 this does not prove anything, as the decrease in the pressure required to account for 

 the observed effects, supposing them due to oxygen, would fall within the probable 

 experimental error. The measurements of the relation between the current and the 

 applied E.M.F., &c., refer to the state in which the leak was varying very slowly, it 

 at all, with the time. 



