130 Mr. M. Ishino on the Scattering and 



For instance, the value of the coefficient depends to some 

 extent upon the thickness of the absorbing materials, and 

 also on the relative position o£ the absorber with respect 

 to the source of gamma rays and the testing vessel. The 

 discrepancies are generally ascribed to the different amount 

 of scattered rays which enter the testing vessel under the 

 experimental conditions. The loss of intensity of a primary 

 beam of gamma rays in its passage through an absorbing 

 material, under the condition that the scattered rays are com- 

 pletely prevented from entering the testing vessel, can be 

 conveniently divided into two parts. One part results from a 

 true absorption of the rays in the material where the energy 

 of the rays is transformed into energy of a different form. 

 The other part is due to a true scattering of the primary beam. 

 Let fi and a be the coefficients of the true absorption and 

 of the true scattering respectively. Suppose a fine beam of 

 homogeneous gamma rays penetrate into a plate normally 

 to its surface. Let the intensity of the beam at any point 

 in the plate, distant x from the incident surface, be I ; then 

 the change of the intensity d\ in the distance dx should be 

 the sum of — fildx due to true absorption and — aldx due 

 to scattering : i. <?., 



dl = —I(fjb + a)dx. 



Hence we have, for the intensity of the emergent beam, 



I = I Q e-^ + ^, ... . . (1) 



where I is the initial intensity of the beam, and t the 

 thickness of the plate. Consequently, the coefficient of 

 the loss of energy is the sum of the coefficients of true 

 absorption and of true scattering. In the case where a 

 part of the scattered rays enters the testing vessel, the 

 coefficient of the loss of energy should give a value inter- 

 mediate between ^ and ytt + cr, and this value is usually 

 taken as the coefficient of absorption. It is for this reason 

 that the observed values of the absorption coefficients found 

 by different observers show divergences often greater than 

 the experimental error. 



The present researches were undertaken to determine the 

 two coefficients \x and a separately, and to find the relation 

 between a and the scattering materials. 



During the course of the present experiments a paper has 

 been published by Hull and Marion Rice * on the law of 



* Hull and Marion Eice, Phys. Keview, viii. p. 326 (1916). 



