Radiation 'produced by slowly moving Cathode Rays. 3G5 



of leak increased at one stage very rapidly with the current. 

 With small currents there was hardly any leak; but as the 

 current was gradually increased the rate of leak rapidly in- 

 creased, at one stage much more rapidly than the current. I 

 attribute this to the current when small taking almost exclu- 

 sively the shortest path between the electrodes: when the 

 current increases it spreads out and many more cathode-rays 

 travel towards the perforated box. 



Some interesting points were established by this apparatus. 

 One was that the radiation which produced the leak was not 

 due to the light coming from the luminous negative column, 

 but arose from the impact against the windows of negatively 

 electrified particles travelling away from the cathode. • This 

 was proved by the effect of a magnet on the rate of leak: the 

 negative glow follows the lines of magnetic force, so that 

 when placed in a strong magnetic field, with the lines of force 

 parallel to the face of the plate containing the windows, the 

 luminous glow had a sharply-defined boundary parallel to the 

 plate. The plate was then lowered down on to the top of this 

 luminous negative glow; there was, however, absolutely no leak 

 from the disk, though, as soon as the magnet was taken away 

 and the negative particles struck against the windows, there 

 was a rapid leak. If the leak bad been due to radiation coming 

 from the luminous gas in the negative glow, the leak when 

 the magnet was on would be greater than when it was off; for, 

 in consequenco of the concentration of the luminosity along 

 the lines of magnetic force, the luminosity of the glow is 

 increased by the magnetic field. Another very interesting 

 point is that, although the rate of leak at first increases as the 

 windows approach the cathode, yet when they get close 

 enough to the cathode to enter the dark space, the rate of 

 leak from the disk rapidly diminishes as the window moves 

 nearer to the cathode, until when the window gets well within 

 the dark space the leak practically ceases. The preceding 

 experiment shows that it is not primarily due to the greater 

 luminosity of the negative glow. To measure the way the 

 rate of leak varied with the distance from the cathode, the 

 arrangement was modified ; since with the form of apparatus 

 shown in fig. 2, the cathode-rays reaching the window form 

 only a kind of bye-stream, the proportion which this bears to 

 the main stream varying rapidly with the intensity of the cur- 

 rent : it seems, therefore, desirable to place the tube containing 

 the windows directly in the main stream. The arrangement 

 adopted is shown in fig. 3. The cathode was attached to a 

 glass float, floating on the top of a mercury column; by 

 altering the level of the mercury in the cistern, the cathode 



