302 Mr. S. J. Plimpton on the Scattering 



is comparable with that in photographs 3, 4, and 5, of the 

 longer series, and there is reasonable correspondence between 

 the two sets of values, although the first series was taken 

 some years previously, and under different conditions. 



The results therefore indicate that the relative values of 

 the constants of the dispersion formula do depend on the 

 temperature of the vapour. Also that the change appears 

 as a decrease in the value of ai/a t with increasing tempera- 

 ture and not an increase, as would be expected if different 

 specialized atoms are engaged in the absorption of the 

 different lines. 



It is to the kindness of Dr. A. W. Porter, of University 

 College, London, that I owe the opportunity of completing 

 this experiment and generous advice with regard to it. 



Feb. 25, 1921. 



XXXVIII. On the Scattering of Rays in X-ray Diffraction. 

 By S. J. Plimpton, Worcester Polytechnic Institute, Mass., 

 U.S.A.* 



IN the theory of X-ray diffraction, it has always been 

 assumed that the secondary rays which give rise to the 

 diffraction effects are due to forced vibrations of the elec- 

 trons in the crystal molecules and should therefore be of the 

 same hardness as the incident rays which gave rise to them. 

 The hardness of the scattered rays has usually been compared 

 with that of the primary beam by observing the decrease in 

 intensity of the scattered rays when an aluminium screen is 

 placed first in the incident and then in the scattered beam. 

 This screening effect of a given piece of aluminium has 

 always been found to be greatest when placed in the scat- 

 tered beam, owing to the fact that in general longer waves 

 are more efficiently scattered by the crystal molecules so that 

 they become relatively accentuated in the secondary radiation. 

 If, however, the incident ravs could be made strictly mono- 

 chromatic, there should be no difference in the measured 

 absorption when the screen is changed from the incident to 

 the scattered beam. Numerous attempts have been made 

 to test this point by rays rendered monochromatic by filtering, 

 and with gamma rays and a crystal containing only light 

 atoms. In all cases the reduction of intensity has continued 

 to appear much greater with the screen in the scattered beam. 



* Communicated by Sir W. H. Bragg, K.B.E., F.R.S. The cost of 

 this research was covered by a grant from the Scientific and Industrial 

 Research Department, 



