Valency of the Radio-elements. 599 



diffusion, and is partly due to the fact that i£ one portion of 

 a radioactive substance is in the ionic condition and the other 

 part is in a colloidal state, then the velocity of diffusion 

 calculated from the diffusing layers is not the resultant of 

 the two diffusing velocities, but shows deviations from that 

 which vary from layer to layer. For example, if a colloid 

 with a velocity of 0"2 diffuses in a mixture with an electro- 

 lyte which has a velocity of 0*4 and if three colloidal particles 

 are present for every ion, then after the diffusion has lasted 

 three days we shall obtain the following constants : — 



L = 0-29; IL = 0-24; IV. = 0'25. 



In a N/10 hydrochloric acid solution colloidal polonium 

 particles are no longer detectable and the polonium diffuses 

 with a velocity corresponding to a double charged ion. 



Diffusion of polonium from N/10 hydrochloric acid solution 

 into N/10 hydrochloric acid solution plus 5 per cent, 

 ethyl alcohol. 



-D150.5. JDi 8 o. 



0*57 

 L 2-83X0-32O °' 63 °' 67 



n - s*sSsis °- 63 0>67 



IV - fsSsh °' fi3 067 



After introducing the correction of 13 per cent, for the 

 retarding influence of the alcohol, we have a diffusion con- 

 stant of 0*77 sq. cm. per day, which corresponds to a double 

 charged substance PoCl 2 . More than 80 per cent, of this 

 substance is dissociated into Po" and CI' ions in N/10 hydro- 

 chloric acid. Iladium E takes up a mean position between 

 polonium and radium D, and in N/1000 hydrochloric ^acid 

 there are still colloidal radium E particles present although 

 the ionic part of the substance predominates. 



On the Separation of Chemically identical Elements 

 by means of Diffusion. 



The diffusion constants obtained for thorium B and 

 radium D and for thorium N and radium, &c. show small 

 differences (see table) which are ascribed to experimental 

 errors. This raises the question whether identical elements 

 of different atomic weights have the same diffusion co- 

 efficient, or whether it is possible to separate them from one 

 another by long continued diffusion. To be able to answer 



