270 Sir E. Rutherford and Dr. E. N. da C. Andrade : 



method were more definite and reliable, while the exposures 

 required for the photographs were relatively much less. 



Table II. 

 Penetrating y rays from radium B and radium C. 



Transmission method. 











Eeflexion method. 



Wave-length 

 (mean). 



Absorption. 



Reflexion. 



42' 



43' 



44' 



■71 X 10 " 9 cms. 



1° 0' 



1° 0' 



1° 0' 



•99 



1°10' 



1°10' 



1°11' 



T15 



1°24' 



1°25' 



1°24' 



1-37 



fl°37'* 



fl°36' 

 1 1° 44' 



f 1° 37' 

 1 1° 44' 



fl-59 

 \l-69 



\ 1° 44' 





2° 0' 



2° 0' 



1-96 



2° 20' 



2° 20' 





2-29 







2° 28' 



2-42 







2° 40' 



2-62 





3° 0' 



3° 0' 



2-96 



3° 18' t 





3° 18' t 



324 







4° 0't 



3-93 







4° 22' 



4-28 



* See note to Table I. 

 t Possibly second order. 



On p. 265 we gave evidence that the lines reflected in the 

 neighbourhood of 1° belong mainly to radium C, and those 

 at greater angles mainly to radium B. In examining the 

 transmission photographs, strong evidence was obtained that 

 the absorption band at 1° 40' must consist of a close double. 

 In experiments with some crystals of rock-salt, this was 

 more marked than with others. The absorption line was 

 considerably wider than would be expected from the dimen- 

 sions of the source if only a single frequency were involved. 

 This observation has been indirectly confirmed by some 

 experiments now being carried out in the Manchester 

 Laboratory by Mr. H. Richardson, M.Sc. 



In examining the absorption by lead of the y rays from 

 radium B and radium C separately, he has found evidence 

 that each of these substances emits two similar types of 

 y radiation, which are absorbed far more easily in lead than 

 the very penetrating radiation from radium C. We should 

 thus anticipate that the spectra of the soft radiation from 

 radium C and the corresponding radiation from radium B 

 should partly overlap. It is, however, impossible from the 

 photographs to determine differences in frequency of possible 

 close doubles at these small angles. 



