21 



ings for the slight alterations in capacity produced by the 

 change in position of the gauzes, the curves for the different 

 width chambers were found to practically coincide for all 

 except very weak fields. 



When, therefore, the bottom gauze is negative, we may 

 conclude that for fields stronger than 50 volts per cm. very 

 few, if any, positive ions penetrate far into the bottom cham- 

 ber through the lower gauze ; otherwise, effects due 

 to recombination in that chamber should have been observed 

 clearly when the depth of the lower chamber was varied. 

 The smaller values of current, obtained when the bottom 

 gauze was negative and the top positive, were due to the 

 action of the field in the top chamber, more particularly its 

 arrangement in the immediate vicinity of the wires of the 

 middle gauze. 



Fig. 6. 



In fig. 6 are shown the results obtained with different 

 gauzes for the bottom electrode, and varying fields applied 

 to the lower chamber; the field in the top chamber being 

 maintained, in all experiments, at 100 volts per cm., arranged 

 in such a direction as to prevent the escape of the ions 

 through the middle gauze. The depth of the lower chamber 

 was 11 mm.; the dotted curves are for bottom gauzes of 

 crossed strands of 50, 40, and 20 meshes to the inch re- 

 spectively. The full line curves are for bottom gauzes made 

 of copper-wire 016 in. in diameter, set parallel, and equi- 



