1894.] On some Phenomena in Vacuum-tubes. 247 



Experiment 12. 



It has already been said that the glass itself appears to assist in 

 the formation of the bands. This may be due to the electrification 

 of the glass or to some other circumstance. In any case the bands 

 have a tendency to stick to the glass with some pertinacity. The 

 most convincing way to show this is by means of a tube lit up by 

 induction in a manner similar to the type of tube employed in the 

 experiment last described. A tube is employed with inductor elec- 

 trodes consisting of glass balls, but the main tube consists of one 

 contracted near the glass balls and then expanded, in the centre of 

 its length, into a large sphere (see fig. 12). When such a tube is lit 

 up the results must be looked for in the large glass sphere. It will 

 be noticed that bands are formed at the entrances to the sphere, 

 somewhat parabolic in shape, i.e., the centres of the bands appeal- 

 driven towards the centre of the sphere, while the edges of the bands 

 seem to stick to the glass and to be retarded. On increasing the 

 current the bands expand somewhat, and become slightly more 

 numerous, and eventually a few of the most forward ones suddenly 

 leave the glass sides and agree with the equatorial plane. (Figs. 

 12A and 12s give a rough idea of the two stages.) 



Experiment 13. 



To further illustrate the action of the glass in the formation of 

 bands, a long tube is employed with a number of bulbs blown along 

 the length of the tube at short distances apart. (See fig. 13.) On 

 passing the current, bands are formed along the whole length of the 

 tube, excepting at those places where the spheres are blown. At 

 these places the bright bands disappear. If the current is consider- 

 ably increased, the bright bands enter a certain distance into the 

 spheres. These are not independent bands, but those driven out of 

 the straight portions of the tube (the bands having become expanded), 

 which is very easily observed when adjusting the current and watching 

 the tube. 



Experiment 14. 



Professor Fleming suggested that, in order to make some of the 

 experiments more convincing, the open tube and tube with a disk 

 should be combined. 



I therefore employed a tube shown in fig. 14 for this purpose. 

 The tube is a plain tube, and at one end, 40 mm. from the electrode, a 

 glass disk is hinged from the side of the tube. This disk may lie flat 

 on the tube or stand vertical (as shown in the diagram) at pleasure, 

 by tilting the tube or by turning it upon its axis. This tube must be 

 placed horizontally in order to raise or lower the disk. 



