538 Mr. N. R. Campbell on Radiation 



the volumes are small and the probable error great. The 

 matter is complicated by the influence of the corners of the 

 box ; at the edges of any wall part of the rays, unless they 

 proceed normally from the surface, will be cut off by the 

 adjacent wall and fail to exert their full effect. When 

 account is taken of this, f(x } X) becomes too complicated for 

 further progress to be made in this direction. It might 

 appear at first sight possible to assume that the rays were 

 emitted equally in all directions ; for the properties along 

 any line in an infinite slab of material — and for this purpose 

 the thicknesses used may be considered infinite — should be 

 independent of the direction of that line. But the values of 

 X obtained on this assumption are so utterly discordant with 

 those found in the experiments described below that the 

 assumption must be rejected. Probably the inequalities of 

 the surface have an influence which cannot be neglected. 

 Much better agreement is also found between the calculated 

 and plotted curves on the assumption that all rays are 

 projected normally. 



For these reasons the attempt to deduce values of X from 

 the curves has been abandoned. But it is readily seen that 

 the values of X cannot be the same or nearly the same, for 

 different materials. The distance from the origin at which 

 the curve becomes appreciably straight w T ill clearly be some 

 measure of the penetration of the rays : it is evident then 

 that X for carbon and tin is less than for lead and aluminium. 



Experiments are in preparation in which the complicated 

 effects of the corners will be eliminated, and, it is hoped, 

 greater accuracy obtained. It may then be possible to deter- 

 mine I and X, and to ascertain whether v Y is or is not 

 accurately zero. 



§ 8. Observations of the same nature were subsequently 

 made in which the boxes were surrounded by some 150 kilos, 

 of lead forming a layer about 3 cms. thick ; this screen 

 should reduce by 2/3 the intensity of Cooke's penetrating 

 radiation. 



The resulting curves are given in fig. 2, PI. VII., and the 

 values of s and v 2 deduced from them in Table I., columns 4 

 and 5. It is to be noted that the value of s is reduced in the 

 cases of tin, carbon, zinc, and platinum, but remains unaltered 

 in lead and aluminium. That is to say, in the former materials, 

 the intensity of the radiation proceeding from the walls 

 themselves is reduced by a screen outside the walls. We 

 have thus strong confirmation of the conclusion reached by 

 Wood *, that part of the spontaneous radiation from ordinary 



* A. Wood, Proc. Camb. Phil. Soc. vol. xii. p. 477 (1904). 



