236 



Sir David Salomons, Bart. [June 21, 



FIG. 2. FIG. 2A. 



creasing the current the bands are driven towards the ends of the 

 tube. A point is reached when the rod is clear of bands, only one 

 appearing at each end of the rod. Still further increasing the current 

 the bands are farther driven back until at last the tube appears filled 

 with light without anj trace of striae. From these two experiments 

 I conclude that, in a large number of cases, the phenomenon of bands 

 is masked in consequence of too much current being employed. 

 Similar tubes, with glass discs at the end of the rod were used, and 

 with same effect, see fig. 2A. Some of the experiments which follow 

 clearly explain how this masking effect is produced. 



Experiment 3. 



It is well known that a bright band is formed at any little projec- 

 tion placed within a vacuum-tube. For instance, if a rod of glass is 

 placed within such a tube along its axis, and has upon it little beads 

 of glass or any other material, a bright band will be produced at 

 these places. It appeared to me that, from Experiment 2, it was not 

 improbable when considering the " repulsion effect " that these bright 

 bands really consist of a pair in close contact. 



In order to examine this question more closely I use a tube (fig. 3) 

 which contains two thin glass disks placed upon a glass rod at one 

 end of the tube. On passing the current, sufficient to light up the 



