EXPERIMENTAL STUDY OF LARGE IONS IN THE AIR. 59 



The results show that by passing air over hygroscopic 

 material the ionisation is reduced to an extent which 

 depends on the amount of moisture absorbed, though it 

 was not found possible to remove all the ions from the 

 air in this way. 



4. Rate of reproduction of the large ions.— When two 

 similar testing pipes are placed in series, separated by a 

 length of earthed tube, if a potential difference is set up 

 between the first pipe and its inner electrode just sufficient 

 to force all large ions on to the inner rod for the air stream 

 used, then any large ions detected in the second pipe must 

 have been reproduced in the interval of time required for 

 the air to travel from one pipe to the other. Observations 

 have been made with such an arrangement using different 

 lengths of brass tube between the two pipes. Finally, to 

 increase the interval still further, two galvanised iron tubes 

 each 2*25 metres long and 7*5 centimetres in diameter, 

 were employed. For intervals up to two minutes the air, 

 in its passage from the one testing pipe to the other, was 

 in contact with brass; for intervals greater than this with 

 a surface of zinc, and the rates of reproduction observed 

 include any effect due to the metal within which the air 

 was contained. 



As the number of large ions in the air is very variable, 

 it is not to be expected that the rate at which they are 

 reproduced will be constant. The observations, indeed, of 

 the interval required for the reproduction of a given num- 

 ber of ions per unit volume at different times are wholly 

 erratic. The measures may however be treated in a 

 different way, and the relation between the time given for 

 reproduction and the number of ions produced, expressed 

 as a percentage of the number originally present, is shown 

 in Figure 1. The results of this method of reduction, though 

 very variable, show some approach to regularity and it is 



