1064 



The objection might be made tiiat the splitting up of the CO 

 under tlie influence of electric discharge is accompanied already by 

 a decrease of volume and that e.g. when Pt-electrodes are used, the 

 metal film formed by disintegration occludes also gases. This is 

 true and we have controlled it experimentally. Still there is an 

 enormous difference in the velocity with which the gases are 

 fixed when phosphorus is present or not. 



We have also investigated more closely the behaviour of phos- 

 phorus with respect to nitrogen and hydrogen and found quite 

 analogous phenomena. We must only draw the attention to the fact 

 that the platinum disintegrated by the discharge can occlude relatively 

 large quantities of hydrogen, so that in this case already without 

 phosphorus a rapid decrease of the gas pressure will be observed. 



In fig. 1 we have represented our measurements graphically. The 

 ordinary lines lepresent the decrease of the gas pressure when 

 phosphorus is present, the dotted ones show the behaviour when 

 this agent is absent. 



Fig. 1. 



Fig. 1 clearly shows, that phosphorus can practically fix all gases very rapidly, 

 when glow-discharges are used. In this respect its action surpasses even that of 

 potassium, the phosphorus being able to fix even metal vapours. Moreover it is of 

 course easier to work with red phosphorus than with potassium. In fig. 2 three 

 spectrograms are given: the first one shows the spectrum of air, the second one 

 that of a mixture of argon and nitrogen, witli 12 ° q nitrogen. In the discharge 



