7T0 Dr. Tycbo E:son Auren on Scattering and, 



it may be that this relation does not hold good for other 

 ranges of wave-lengths and is accidental in the present case. 

 However, in these experiments, the relations may be used to 

 find the mass scattering coefficient of a certain element. 

 Taking Alc u and ]\I a to denote the atomic weights of Cn and 

 the element (a) at issue and Z a the atomic number of the 

 element, then the following formula for the mass coefficient 



® - 



Ids good : 



3I Cu • Z c . q fl 



-a 



p"-M a (l + 29g) ' p 



Cn (5j 



"in — 



AsMca=63-6and«^Cu = 000609 (Table IV.), - at a greater 



P P 



wave-length, when 29 q may be neglected, can be expressed 

 thus : 



S a=~ T a . 0-387 (6) 



p M fl 



As for H. the value - 387 is obtained, which is somewhat 

 lower than what I found before (0*445), owing to the fact 

 that absorption of hydrogen could not. with sufficient accu- 

 racy, be ascertained. The value I have found now. however, 

 agrees very closely with what has been calculated according 

 to Thomson's theory and what Barkla has experimentally 

 found. For Al formula (6) gives the value - 187 instead of 

 the known value 0*2. We infer from formula (5), however, 



that - decreases with the wave-length, because q is con- 

 tinually increasing {cfr. Table IV.). At X = 0*166 A., 



- therefore attains the value of 0*157, which is in good 



p s 



conformity with the fact that - cannot have the same value 



P 

 at all wave-lengths. Already at \ = 0*16 A, the mass 

 absorption coefficient of Al is 0*18, and the mass scattering- 

 coefficient, being part of the whole absorption coefficient, 

 cannot be greater than the whole. Richtmever and Grant * 

 have recently published the results of experiments of ab- 

 sorption in about the same wave-length range and, as an 



* Pliys. Rev. xv. p. 547 (1920). 



