636 Miss J. M. W, Slater on the 



(g) The effect of a secondary radiation was tested by en- 

 closing the active disk while in the discharge-tube in a little 

 cylindrical box of aluminium, with a closely-fitting lid in 

 which was a window o£ very thin aluminium-foil ; the active 

 surface was immediately behind the window, on to which a 

 strong beam o£ cathode rays fell. There was no loss of 

 activity in this case. The effect of Rontgen rays was also 

 directly tested by placing the active disk just outside a 

 Rontgen-ray bulb, where a fluorescent screen showed that 

 the rays were strongest : an exposure of |- hour caused no 

 change. 



The only causes left capable of explaining the observed 

 effect are, therefore, the variations of potential occurring 

 during the discharge, together with some direct action of the 

 cathode rays. To obtain, if possible, further light on this 

 point, a number of experiments were made in which the active 

 disks were enclosed in earthed metal boxes. The lid of the 

 box was generally of fine copper gauze. It was thought at 

 first that this arrangement would prevent any considerable 

 variation of potential on the enclosed disk, but in practice it 

 is hardly possible to get the earth connexions good enough 

 to prevent changes of potential with the very rapid variations 

 occurring with a coil-discharge, especially when (as was 

 generally the case) the gauze lid faced the cathode, and the 

 cathode rays were able to enter the box. When the metal 

 bottom was towards the cathode, the loss of activity did not 

 occur. When the box was made cathode, the enclosed disk 

 lost 15 to 20 per cent, in a short time. 



The similarity of the effects obtained with rays falling on 

 the active and inactive sides of the disks is rather curious. 

 The incidence of cathode rays must cause very large momen- 

 tary variations of potential, which will spread below the 

 surface struck to a depth varying inversely as the conduc- 

 tivity of the material. Thus with thin mica disks, it is not 

 surprising that there should be sputtering of the active matter 

 from the back as well as from the front. Aluminium also 

 is not one of the metals of highest conductivity, and the thin 

 disks generally used (*12 mm. in thickness) may allow of 

 considerable variations of potential reaching the side away 

 from the cathode. To test this point, disks *34 mm. thick 

 were used. When cathode rays fell directly on the inactive 

 side, these disks also lost their activity, though not to the 

 same extent as before, the diminution observed being about 

 10 per cent. In some cases, it was evident that the far side 

 of the disk (on which was the active matter) was behaving 

 as a secondary cathode, as the glass stopper behind it glowed 



