32 Proceedings of the Royal Irish Academy. 



plate are observed. The eiurent in general is made up of two parts — (1) that 

 due to ions produced throughout the space between the plates by the un- 

 screened part of the radiation, and (2) that due to ions produced by the 

 intense radiation at what we may regard as the " point-source," close to one 

 plate and captured on the other plate before they had been carried away by 

 the air-current. K a current-voltage curve is plotted, (1) in the absence 

 of (2) would give a smooth curve something like an ordinary saturation curve. 

 The ions, however, from the " point-sotixce " will show up (if they are all of 

 one kind) as an abrupt upward step on the curve at a certain voltage. If 

 there is more than one class of ion present, each class will show up as a 

 distinct step on the curve. The mobilities are readily calculated from the 

 formula Vu = QajLb, where V is the " critical voltage," u the mobility, 

 Q the volume of air passing per second through the apparatus, a the distance 

 between the plates, L the length of the receiving plate, and h the breadth 

 of the apparatus. As in the work pre\-iously described, the air was passed 

 through the apparatus from one gasometer into another. These gasometers 

 were of identical dimensions^ and were connected together so as to move at 

 the same rate. The greatest care was taken to secure steady working con- 

 ditions, as the special object of these particular experiments was to determine 

 accurately the form of the current-voltage curve, and to obtain exact values 

 for the mobilities of the different types of ions already recognized. 



Results. 



In fig. 1 a number of the current-voltage curves indicating the ions of 

 high mobility are shown. The curves selected for reproduction are such that 

 the value of the product Vti, (critical voltage x mobility) is very approximately 

 the same for all. The current values are plotted to the same scale, but to 

 various arbitrary zeros in order to show all the curves without overlapping. 

 These curves illustrate the method that has been adopted in this work, i.e., to 

 work over a limited range with small voltage-steps and great attention to 

 accurate electrometer readings. The agreement between the curves is very 

 satisfactory. In ordinary circumstances the most mobile ion found produces 

 a peak or step at about 2 volts, and has therefore a mobility value of 12 

 approximately. This ion is found both with positive and negative charges. 

 On two occasions, however, distinct evidence was obtained of a negative ion 

 having twice this mobility. The curves in question have been reproduced. 

 Apart from this ion of mobility, approximately 24, no diflerence between 

 positive and negative ionisation has been noticed. The mobility values 

 deduced from all reliable observations are given below. These mobilities have 



