42 



MR. O. W. RICHAIM)S()N <>N THE TOXISATION 



The oxygen iu the first experiment was prepared by electrolysing caustic potash, 

 in the second hy heating potassium permanganate. The difference in the values for 

 oxygen is prohahly " accidental." 



The disparity hetween the two leaks is considerably smaller than that calculated 

 from the table, but it shows that nitrogen has a distinct inhibiting effect on the 

 positive ionisation produced by oxygen. It may do this by combiniug with the free 

 oxygen ions and so reducing their concentration, or by associating with the platinum 

 and so reducing the amount available for combination with oxygen. The latter view 

 would give a satisfactory quantitative explanation of the results. 



VI. 13. THE IONISATION IN HELIUM. 



A few experiments were made in this gas chiefly with the object of finding out if 

 hot platinum would liberate positive ions in an atmosphere in which the possibility of 

 chemical action in the ordinary sense was excluded. The helium used was obtained 

 from Messrs. TYREB AND Co., Sterling Chemical Works, Stratford, E. After 

 admitting it into the apparatus it was dried over phosphorus pentoxide and purified 

 by means of a discharge from a cathode of sodium potassium alloy. The tube used 

 for this purpose is shown in fig. 13. After the apparatus had been completely 

 exhausted, the liquid alloy was admitted through the stop-cock A. Care has to 

 be exercised in doing this, as the alloy is liable to stick in the 

 tube, and, when it begins to move again, to do so with such 

 violence that it is projected into other parts of the apparatus. 

 The electrical connection to the cathode was made by the 

 platinum wire B sealed into the glass. The anode C was of 

 aluminium. The side tubes D, E led to the pentoxide bulb and 

 testing vessel and to the pump and McLeod gauge respectively. 

 MEY* has shown that the discharge from a cathode of sodium 

 potassium alloy given by an induction coil rapidly absorbs all 

 gases except the inert ones. The discharge used was found to 

 run down 3 millims. of air to a pressure comparable with that 

 due to the residual argon in about 20 minutes. The discharge 

 was allowed to pass for several hours before measurements of the 

 leak were made. The helium must have been very pure to 

 start with, as no decrease in pressure was observed with the 

 McLeod gauge. The spectrum was also examined from time 

 to time, and no lines belonging to any gas other than helium 

 were noticed. 



There is every reason to believe that the helium used in these 

 experiments attained a very high degree of purity, and that the leak observed is really 

 due to helium and not to traces ot'smne ..iher u ,is present. The leak resembled that 



* 'DuuuE's Ann.,' vol. 11, p. 138. 



Fig. 13. 



