540 Mi*. N. K. Campbell on Radiation 



15 x 15 x 6 cms., In one side of which was a window 12 x 12 

 cms. covered with a sheet of thin aluminium-foil ; the thick- 

 ness o£ the foil used in most of the experiments was 

 "00033 cm.; similar results were obtained with a thickness 

 of -00053 cm., and also with a sheet of very thin paper 

 rendered conducting with graphite. 



The observations consisted in ascertaining the leak through 

 the vessel when no solid body was within a distance of one 

 metre from the window, and then noting the effect of 

 bringing up to the window a plate of the material the 

 radiation from which was to be investigated. At first the 

 vessel was charged to a high negative potential, but troubles 

 occurred which were traced to the deposition on the surface 

 of the " excited activity i3 .which is known to collect on all 

 negatively charged bodies exposed to the atmosphere. The 

 amount of the excited activity, the rays from which could 

 penetrate the window, varied with the freedom of access of 

 the outside air to the vessel, and hence the leak was constantly 

 changing. In the later experiments the vessel was always 

 charged positively. 



§ 10. If our conclusion is correct that the influence of the 

 walk on the ionization consists of rays projected from them 

 absorbable by some 6 cms. of air, we ought to be able to 

 detect the influence of the rays through the aluminium foil, 

 which is equivalent in surface density to '69 cm. of air. 

 Accordingly when plates of lead, tin (foil or plate), aluminium, 

 zinc, iron, brass, copper, graphite, wood, or glass were 

 brought up to the window, an unmistakable increase in the 

 leak was detected — in the case of tin the increase was as 

 much as 30 per cent. It is clear that the increase caused by 

 any material will rise with the penetrating power as well as 

 with the intensity of the rays given off from it ; it is, then, 

 in no way inconsistent with Table I. that tinfoil gave a larger 

 effect than lead, and carbon an effect nearly as great. With 

 some materials, such as paper and ebonite, the increase was 

 so small that it cannot be positively asserted to exist. 



Strutt* has found large differences between the effects 

 obtained with different specimens of the same material, but 

 though I have tried several samples of each of the materials 

 mentioned, appreciably the same value was obtained for 

 each. Strutt must have been extraordinarily unfortunate in 

 securing samples contaminated with radioactive impurities, 

 for McLennan, Righi, and Wood have found remarkable 

 constancy in the value given by different samples of the 

 same metal. 



* Strutt, < Nature,' vol. lxvii. p. 369. 



