between the Terminals of a Vacuum Tube. 157 



The discs were placed at the same distance, about 2 cm., from the 

 electrodes in each tube. By the switch K, the current could be 

 sent through either tube at will, and from the reading of the spark 

 micrometer S, the potentials of the tubes could be inferred. The 

 apparatus was attached to the pump at T and exhausted until 

 a good phosphorescence was obtained when the discharge passed. 

 The coil was adjusted to give a spark of from 4 to 6 cm. in air at 

 atmospheric pressure. 



In the first experiment the electrodes were of iron and the 

 discs in both tubes were either all of aluminium or of copper. 

 Then the reading of the spark micrometer was the same for each 

 tube. Next, aluminium discs were inserted in one tube and copper 

 discs in the other. The spark lengths were as follows : 



or approximately Al : Cu = 3 : 1. 



The sealing-wax joints gave considerable trouble owing to the 

 tube becoming so hot that the wax melted, and therefore the tube 

 shown in the figure was devised. 



The stopper was hollow and had a small aluminium cap 

 attached to the platinum wire. Into this cap different electrodes 

 could be fitted. Some little indentations were made in the side of 

 the tube so that the various discs could be placed at the same 

 distance from the cathode. The discs were really shallow caps 

 turned to fit the tube snugly, the hole being of course of the same 

 size, 0'4 cm. each in diameter. 



The tube without a disc was used as a standard with which to 

 compare the potential of the other tube. The use of a second 

 disc close to the anode was found to exert but little effect, and 

 therefore was not used further. 



The mode of conducting an experiment consisted in exhausting 

 the tubes until a good phosphorescence was obtained, the dark 

 space extending up to or beyond the disc. Then the standard tube 

 being in connection with the coil the micrometer was adjusted 

 until the sparks passed between the knobs and the discharge 

 through the tube. The micrometer reading was then taken. The 

 current was then switched from the standard to the one with the 

 disc, and a micrometer reading taken as before. The ratio of 

 the spark lengths was taken to be the ratio of the potential 

 differences. 



VOL. XII. PT. III. 11 



