Tico Ions produced in (ruses hy Bout gen Radiation. 127 



good interrupter, which is always a very important factor, a 

 comparatively uniform emission of rays is obtained, together 

 with a steady unwavering fluorescence. 



The main advantage of this form appears to lie in the fact 

 that the cathode is within the contracted part B'C, which 

 during action remains quite free from surface electrification, 

 so that the flickering^ due to the surface discharges from this 

 part of the tube are absent. 



It has been found convenient in testing whether a tube is 

 giving out rays at a uniform rate, to employ two parallel 

 plates, one of which is connected to a constant source of po- 

 tential and the other of which is connected at the same time 

 to an electrometer, and to earth through such a high resist- 

 ance as a column of xylol. As rays are allowed to fall between 

 the two plates, conduction takes place between them, and the 

 plate that is connected with the electrometer-quadrants assumes 

 a certain potential such that the leak to earth through the high 

 resistance is equal to the current between the two plates. 

 Any variation in the intensity of the radiation is indicated by 

 a corresponding variation in the electrometer-reading, so that 

 by simply watching this it is possible to see the amount and 

 nature of the variations in the radiation. With such an 

 arrangement permanently set up one can conveniently com- 

 pare from time to time the strength of the rays which are 

 being used for any purpose; or if an electrometer can be 

 spared for this, the relative strength of the rays can be 

 directly observed at any instant. 



§ 4. Manipulation and Results. 



In the practical working of the method the readings were 

 taken in the following manner : — The pair of quadrants of the 

 electrometer, connected with the gauze T, was joined to earth 

 by means of the key H. The rays and blast were then 

 allowed to run a short time so as to establish a steady state in 

 the conducting gas between the two plates. The electrometer- 

 zero was now read, and the quadrants were disconnected from 

 earth. At this instant a stop-watch was started, and a defi- 

 nite interval afterwards, usually 20 seconds, the rays were 

 stopped by cutting off the current from the coil, and, finally, 

 the electrometer-deflexion was read. 



The following table shows how, for air, the quantities of 

 electricity which reach the gauze in a given time vary when 

 the air-stream is kept constant while the potential of Q is 

 varied in amount and sign. 



