400 Mr. K. H. Kingdon on some 



These figures clearly indicate that the carbide was slightly 

 radioactive. Any trace of: radium emanation present would 

 make itself very apparent in the gas ganerated from the 

 carbide, and it is to the' emanation that the high value of 

 the residual ionization previously obtained was due. 



II. On the possibility of a Portion of the Residual Ionization 

 in Gases being due to the Collisions of Thermal Agitation. 



Introduction. — In the first part of this paper the three 

 possible components of the residual ionization in gases were 

 stated to be : — (1) a radioactive impurity in the gas, (2) a 

 radioactive impurity in the walls of the ionization-chamber, 

 (3) ionization by the collisions of thermal agitation. Now, 

 for the gases air, carbon dioxide, hydrogen, and nitrous oxide, 

 McLennan and Treleaven * have shown that if a clean zinc 

 ionization-chamber is used, and if the gases are properly dried 

 and filtered before admitting them to the chamber, uniform 

 and reproducible values of the residual ionization are obtained 

 for each gas. Hence it does not seem probable that the 

 residual ionization is in general due to the presence of traces 

 of radioactive emanation in the gas. The part played by the 

 second of the above-mentioned components has been quite 

 fully investigated. Recently McLennan and Murray f have 

 shown that by constructing the ionization-chamber of ice, it 

 is possible to obtain a very low value for the residual 

 ionization in air. Their experiments show the great effect 

 of radioactive impurities in the walls of the chamber on the 

 ionization, for with one ice-chamber the number of ions made 

 per c.c. per second was 2' 6, while with another the number 

 was 5'5. Although from these experiments it appears pos- 

 sible that the residual ionization, in air at least, may be due 

 entirely to radioactive impurities in the walls of the ionization- 

 chamber, yet it seemed worth while to make some experiments 

 to test for the presence of ionization due to the collisions of 

 thermal agitation. The method was to vary the temperature 

 or the density of the gas in the ionization-chamber, and from 

 the resulting changes in the ionization to see if it was possible 

 to detect the presence of any ionization produced by the 

 collisions of thermal agitation. The results of the expe- 

 riments in which the density of the gas was varied show 

 that only a part of the residual ionization can be due to the 

 collisions of thermal agitation. The results of the expe- 

 riments in which the temperature of the gas was varied give 



* Loc. cit. 



t J. C. McLennan and H. G. Murray, Phil. Mag. xxx. Sept. 1915. 



