726 Dr. J. A. Orowther on U J" Radiation. 



Table II. 



Coefficients of absorption in Aluminium of primary and 

 fluorescent radiations. 



■Rarlinfnr X V Wave-length X 1CT 8 . 



radiator. A prfmary . X Becondary . p rimary< Secondary. 



Paraffin-wax ... 2'22 4-8 "38 '43 



Aluminium 2"8 6"1 -41 



•51 

 20 GO '36 



Copper 2-1 3'7 36 -46 



The average wave-lengths given in the last two columns 

 are deduced from the measured coefficients of absorption of 

 the radiations. 



Discussion of Results. 



Assuming that the secondary radiations whose wave- 

 lengths have been calculated are true fluorescent radiations, 

 it seems clear that they cannot all belong to the same series. 

 In the case of all lines of the K and L series the frequency 

 of a given line varies approximately as the square of the 

 atomic number of the demerit from which it is emitted, and 

 there is no reason to suppose that the law would be different 

 for a " J " series of radiations. This criticism applies 

 equally to the absorption bands discovered by Barkla. The 

 J absorption band for carbon occurs at a wave-length of 

 0-42 xlO~ s cm. and that for aluminium at 0*37 xl.0~ 8 cm , 

 while the atomic number of aluminium is more than twice 

 that of carbon. It may be noted in pnssing that the wave- 

 length of the radiation obtained in the present experiments 

 from paraffin-wax (0 43xl0~ 8 cm.) agrees closely with 

 Barkla' s absorption band for carbon, and affords confirmation 

 of a fluorescent radiation from carbon of about this wave- 

 length. 



Some light may be thrown on the problem by a study of 

 the 7 radiation. It has been shown by Rutherford and 

 Robinson (Phil. Mag. (6) xxviii. 1914, p. 263) that the /3 rays 

 excited by the 7 radiation from radium C consist of a series 

 of groups each with a characteristic and very definite 

 velocity. It is assumed that each group of /3 rays is pro- 

 duced by y radiation of definite wave-length, and that the 

 7 ray spectrum thus consists of a considerable number of 

 characteristic lines. These conclusions are strongly con- 

 firmed by recent experiments of Ellis (Proc. Roy. Soc. A, 

 1921). A series of no less than twelve lines are tabulated 



