744 Dr. C. G. Barkla and Mr. C. A. Sadler on 



D 1 and D 2 was 177 : 100, after correction for the radiation 

 from air and the normal ionization. Thus, according to 

 the simple theory of the superposition of the homogeneous 

 secondary radiation on the untransformed transmitted 



radiation, ^'_ of the ionization in position D 1 was produced 



by the transmitted copper radiation, while ^= was the 



result of the radiation from the iron plate. But the 

 absorption of the radiation entering the electroscope in 



position D 1 was found to be such as would be obtained 



if the mixture had contained p^ of iron radiation. Thus 



the ratio of intensities of the iron radiation in the two 

 directions BD X and BD 2 was 96 : 100. 



This is as near equality as could be expected from the 

 nature of the experiment. We conclude, therefore, that 

 the observed phenomena are due simply to the homogeneous 

 radiation uniformly distributed, superposed on the trans- 

 mitted radiation, and that the radiation proceeding in the 

 original direction is untransformed. 



Now, as in experiments on the absorption of such radia- 

 tions, beams of considerable cross-section must be used to 

 produce measurable effects, an electroscope placed behind 

 the absorbing plate, unless very distant from that plate, 

 must receive a portion of the secondary radiation from that 

 plate. Thus without correction the transmitted beam would 

 appear more intense than it actually is. The correction 

 might have been determined experimentally in the manner 

 shown above. A simple calculation, however, gives us the 

 constitution of the beam received by the electroscope ; and 

 consequently the correction that must be applied to determine 

 the true diminution in energy of the primary beam in terms 

 of quantities which may be easily determined. 



Let I be the intensity of a parallel beam of homogeneous 

 radiation incident normally on a plate of absorbing material 

 of thickness t, 

 then I = I <?- x i 2 gives the intensity I at a depth of x, 



where \ is the coefficient of absorption in the substance. 



An electroscope placed in the path of the transmitted beam 

 beyond the absorbing plate (fig. 2) receives the beam, together 

 with a portion of the radiation scattered by the plate, and of 

 the secondary homogeneous radiation characteristic of the 

 material of the plate. 



It may easily be shown that the former — the scattered 

 radiation — may be neglected in most, and possibly in all, of 

 the cases considered. 



