Secondary y Radiation. 239 



aluminium, but sufficient experiments have not been made to 

 show how far the quantity of: /3 rays set free depends on the 

 quality of the secondary radiation. This does not appear to 

 be sufficiently marked to make it an effective method for 

 distinguishing radiations. According to Bragg's results*. 

 tin and carbon would be the best materials to use for 

 contrasting this production of /3rays. A combination of two 

 plates of tin and carbon lias recently been tried. The 

 ionization current is always slightly greater when tin is the 

 inside plate. The contrast is not as marked as when a com- 

 pound plate of aluminium and lead is used. 



The emanation was then screened with another piece of 

 lead 1*14 cm., and the distribution examined with aluminium 

 and then lead as the side of the electroscope. The absorption 

 coefficients in this lead screen have already been included in 

 Tables III. and IV., and they illustrate the corresponding 

 alteration in distribution. 



Whatever radiator has been so far employed and whether 

 the emanation was screened with lead *084 cm. thick or with 

 lead ('084 + 1' 14 cm. thick) there has always been a pre- 

 ponderance of the emergent over the returned radiation. 

 This increases the nearer the electroscope is brought to the 

 direct primary beam on the emergent side of the radiator. 



First of all let us examine the returned radiation when 

 the side of the electroscope consists of: aluminium. The 

 results obtained with a side of lead need not be considered, 

 as a large correction is required for the absorption of the 

 radiation in the plate. The distribution of the returned 

 radiation from aluminium, zinc, and lead apparently corre- 

 sponds to the distribution anticipated from the ordinary theory 

 of scattering. There is always a greater proportion returned 

 normally to the plato than at right angles to the primary 

 beam. 



On the other hand, carbon appears to be an exception. 

 There is evidence of a greater scattering of the radiation at 

 right angles to the primary beam when the proper corrections 

 are made for absorption. 



In the case of the emergent radiation the same pecu- 

 ities are observed as with X rays. There is a marked 

 dissymetry in the returned and the emergent radiation and 

 the quantity emerging in any given direction is dependent 

 on the material of the radiators. 



It is important in these experiments to know to what 

 extent the ionization current depends on the penetrating 

 power oi' the radiation examined. It would appear that the 

 * Bragg and Madsen, Phil. Mag. December 1908. 



