66 J. A. POLLOCK. 



the humidity experiments for the average of the values of 

 the meteorological conditions which existed during the 

 mobility observations. Owing to ionisation being caused 

 by phosphorus it is not advisable to use phosphorus pent- 

 oxide as the material to dry the air in the first instance ; 

 when it was employed as in some cases the values of the 

 mobility obtained were abnormal. 



The first experiments with dried air showed little or no 

 reduction of mobility. In searching for an explanation it 

 was thought that perhaps when the humidity was changed 

 the ions did not at once reach a state of equilibrium with 

 the new vapour pressure conditions ; the air after flowing 

 over the calcium chloride was therefore passed through 

 galvanised iron pipes before being directed into the testing 

 pipe, as in this way the time between the drying of the air 

 and the measurement of the mobility could be varied. In 

 the first part of Table II are given the values of the mobili- 

 ties of the ions at the end of different intervals of time 

 after the air had been in contact with the hygroscopic 

 substance, the humidity being sufficiently constant for the 

 purpose of the comparison. The positive and negative signs 

 in the table indicate the class of ions to which the respective 

 observations refer. 



The mobility decreases as the time increases from the 

 moment of contact of the air with the drying material. In 

 the case given the minimum value seems to be reached in 

 about 7 minutes ; in all succeeding experiments an interval 

 of 13 minutes was allowed to elapse between the drying of 

 the air and the determination of the mobility of the ions. 



In the latter part of Table II are given the further 

 results against which nothing is known. In many cases 

 observations have appeared normal, but owing to a change 

 in the ionisation were unable to be repeated ; only those 

 results have been retained which depend on experiments 



