Si 



Proceedings of the Royal In's/i Academij. 



cubic centimetres, and by suitably varying the pressme a very steady air- 

 stream of the required amount eould be obtained. The aii- was drawn into 

 the gas-holder through the appai-atus, so that all the lai^ ions were removed 

 during its passage through the electric field, the saturation cuirent c being 

 observed. It was stored for various intervals in the gas-holder and again sent 

 through the electric field, the saturation current c being again taken. It was 

 found that for different values (At, when the deionised air was stored in the 

 gas-holder for the same intervai,the values of the ratio of c ' to c were much more 

 concordant than the absolute values of t/. The results of experiments over 

 different intervals are shown in the curve (fig. 9), Where the abscissae denote 









i I 1 1 



^"~^. 









1 



\ 

















\ 













■ 



\ 















\ 













X^T" 













1^^--^ 













i 







- — ' ____ 













-J,50O . 2.000 2,500 3,0DD 



Tane. Inter^aZs in J-Mnutes. 



the time-intervals during which the air was stored, and the ordinates repre- 

 sent the values of the ratio of e' to e. It wiU be seen that the value of the 

 ordinate has almost reached its masimum of 086 in about an horn-, remains 

 more or less constant for a considerable time, and then falls, somewhat rapidly 

 at first, and then very slowly towards zero. It seems, therefore, that large 

 ions may be reproduced in air from which they have been removed by an 

 electric field, though not t» the same extent as before, and that these ions 

 very slowly disappear again. 



To see how long the process of reproduction can go on in a given mass of 

 air from which the large ions are removed at intervals by an electric field, the 

 following eicperiments were carried out. The air was first drawn into the 

 gas-holder A through the electric field, all the large ions being in this way 

 removed. After an interval of ten minutes it was sent through the electric- 

 field again into another similar gas-holder JB, the saturation current being 

 oteerved. It was sent from .B to ^ after an interval of ten minutes, the 

 saturation cuirent being again taken; and this process was repeated for several 



