606 Dr. Manne Siegbahn on 



as above. For the sake of comparison we may mention the 

 following values of other authors : Millikan * from his 

 precise measurements on fundamental physical constants 

 deduces the value 



^ = 3-030 + 0-001. 10" 8 ; 

 S. Gorton t gives the value 



^ = 3-028. lO" 8 ; 

 and, finally, A. H. Compton J finds 



d = 3-0279. 



v '. 2 he lattice- constant of potassium ferrocyanide. 



Potassium ferrocyanide has a special interest on account 

 of the fact that Moseley, in his classical investigations on the 

 X-ray spectra of the elements, used this crystal. The syste- 

 matical determinations which have been made at the laboratory 

 of Lund show, when compared with the measurements of 

 Moseley, a constant deviation of about 0'5 per cent., which 

 considerably exceeds the limits of experimental error. As a 

 probable cause of this deviation I have suggested that the 

 lattice-constant which Moseley used in the calculations of the 

 wave-lengths was not in accordance with the assumed value 

 for rock-salt. The latter crystal has been used in our inves- 

 tigations, and the value d= 2*814 given by Moseley in his 

 earlier paper was taken as its lattice-constant. For the 

 longer waves selenite was used, the lattice-constant of which 

 was referred to that of rock-salt. 



For the control measurement of the lattice-constant for 

 potassium ferrocyanide a home-made crystal specimen was 

 used. The crystal was not very good, but good enough to 

 allow sufficiently accurate measurements. From a spectro- 

 gram on CuKa of the second order the reflexion angle was 

 determined at : 



CuK ai : 2 = 1O° 32' 14", 

 which gives as value for d, 



d= 8-407. lO" 8 cm. 



* Millikan, ' The Electron.' 



t S. Gorton, Phys. Rev. Feb. 1906. 



§ A.fl. Uompton, Phys. Rev. June 1916. 



