PROIMVKI* I!Y HOT PLATINTM IN 1HKFKKKNT CASKS. 



in nitrogen in showing too big a value after tin- wire had leen allowed to remain 

 cold and only gradually settling down t<> a steady \alne. It is very difficult to 

 explain this effect as being due to the absorption of impuritiea, a* it appeared to be 

 much the same whether the discharge had lx?en run before the leak was tested or 

 not. It seems far more likely to he due to a change taking place in the platinum. 

 The following measurement*) at 907 C. of the leak at 2'4 millims. pressure with 

 + 11 volts on the filament show the way it varied with time : 



The values of the steady leak under +11 volts at the same temperature and 

 different pressures are as follows : - 



The positive ionisation in helium at 2 millims. pressure and 907 C. appears to be, 

 roughly speaking, about three times that in nitrogen and one-fortieth that in oxygen 

 under like conditions. 



The above experiments are not complete enough to enable us to deduce the law 

 according to which the leak in helium varies with the pressure. The experiments 

 were not pushed up to high pressures owing to the difficulty of ensuring the purity 

 of the gas. They are chiefly of interest as showing that hot metals can produce 

 ionisation in the inert gases when chemical action is out of the question. 



VII. 14. THE IONJSATION IN HYDROGEN. 



A large number of experiments have been made with this gas, but the phenomena 

 are very complicated, so that it is difficult to be quite certain of the interpretation of 

 the results. This is largely owing to our ignorance of the absorption and diffusion of 

 hydrogen in platinum at low pressures. Some of the effects observed seem to be 

 of considerable interest and importance, and they will be described in order to give 

 greater completeness to this account of the subject. Discussion of the theoretical 

 liearing of the results will be avoided except in so far as it is necessary to render the 

 purpose of the experiments intelligible, and generally speaking the writer is of 

 opinion that this jwrt of the work ought to be regarded as a preliminary exploration 



G 2 



