38 Proceedings of the Royal Irish Academy. 



above it. In most of our experiments the potential difference between these 

 plates was 20 volts. The apparatus is contained in a metal box, which can 

 be rendered air-tight. The alternating E.M.F. was obtained from a trans- 

 former, the primary of which was connected to the city supply. An Ayrton 

 and Mather electrostatic voltmeter was connected across the secondary of 

 the transformer. In order to obtain small potentials, a potentiometer 

 arrangement was used, as shown in the diagram. The E.M.F. obtained in 

 this way was remarkably steady. Except on very rare occasions the volt- 

 meter showed no iiuctuations. Electrometer readings of great accuracy and 

 consistency were possible. 



The theory of the method is well known. Ions of one sign diffuse 

 through the perforated plate, move downwards under the influence of the 

 field, and give up their charge to the lower plate if they have reached it 

 before the field is reversed. If they have not reached it, they are carried 

 upwards and are discharged at the perforated plate, the lower plate being 

 unaffected. It is clear, therefore, that if we keep the frequency of alterna- 

 tion constant and gradually increase the field, no charging of the lower plate 

 will occur until the field is reached, which is just sufficient to carry the 

 fastest ion across the space during half the periodic time. For higher 

 voltages the current will increase rapidly. If more than one class of ion is 

 present, each class will show itself by a sharp upward inclination of the 

 current-voltage curve at a voltage inversely proportional to its mobility. 

 The relation between this critical voltage and the other quantities, assuming 

 that the E.M.F. is of sine form, is given by the expression 



Vu = — ^ . ncP, 

 v/2 



V being the voltage reading on the electrostatic voltmeter, u the mobility, 

 n the frequency, and d the distance between the plates. In order to make 

 sure that ions which just escape capture at the lower plate are discharged at 

 the end of their upward journey, it is necessary to make the alternating field 

 slightly asymmetric, so that the upward field is greater than the downward. 

 This was effected in our case by introducing a small steady voltage between 

 the alternating voltage terminal and the apparatus. This is represented in 

 our diagram as a single cell, but it was more often a fraction of the voltage 

 of a cell. 



In the first set of experiments the apparatus was not rendered air-tight. 

 The air was, therefore, the ordinary moist air of the atmosphere, and the 

 results are consequently in some degree comparable with those obtained for 

 saturated air by the other method. Some examples of the current- voltage 



