Nolan an]> Hauris — lonisaiion in Moist and Dry Air. 35 



occasions the ion was negative and not very well marked. The mobilities 

 calculated were 2'24 and 2'13. 



The following table gives the values deduced from all good determinations 

 of the slower groups : — 



Table of Slower Ions. 



(Corrected to 76 cm. pressure.) 

 Negative. Positive. 



(2-24) 



1-92 



1-72 



1-62 



1-37 





f 



2-07 



1-79 



1-60 



1-37 



(2-13) 



2-07 



1-80 



1-51 



1-37 





2-07 



1-79 



1-54 



1-35 





2-09 



1-70 



1-60 



1-36 





1-95 



1-79 





1-38 





2-00 





1-50 







2-01 











2-07 





















-* 



"Weighted 



means ; 



204, 



1-79, 



1-52, 



and 1-37. 







In the previous paper some uncertainty was expressed as to whether the 

 mobility of the slowest of these ions was the same for positive as for nega- 

 tive. The present work shows a very exact agreement for this ion as for 

 the others. The. best agreement between repeated observations and the 

 most distinct indications are obtained as far as the negative ions are con- 

 cerned for the ion of mobility 1"51, and for the positive for the ions of 

 mobilicies 1'79 and 1.37. It is interesting to note, that Zeleny's values for 

 moist air are I'ol for negative and 1'37 for positive. 



Examination of the Ionisatio7h in COo ly the Air- Current 3Iethod. 



Having completed the examination of moist air, we decided to undertake 

 the examination of other gases. Hydrogen would obviously be of the 

 greatest interest, but we were unable to obtain this gas in the large volumes 

 demanded by our method. Carbon dioxide was available, and was submitted 

 to the test. The greatest difficulty was experienced in keeping the gas even 

 moderately pure. It vi'as necessary that the whole system, including two 

 gasometers, connecting pipes, and the ionisation chamber, should be air-tight. 

 After much trouble this was approximately secured ; but very minute leaks 

 were still present, and after the lapse of a few days much air had found its 

 way into the system. All the results obtained therefore apply to CO2 mixed 

 with various quantities of air. The percentage of CO2 in the gas was deter- 

 mined before and after each experiment, so that in each case we know the 

 degree of contamination. The results obtained apply only to positive ions. 



As will be seen from the table which follows, the general character of 

 the ionisation resembles very much tliat found in the case of air. For each 



