Vapour Pressure and Affinity of Isotopes. 177 



At low temperatures //,-is large and one may write 

 , p x 7t 4 / 1 1\ a r^ w e-^f 1 2 



or if //, is large enough 



P 9 



/'2 O 



Since // = 1 at 95° absolute the vapour pressures at these 

 temperatures have little practical interest. At higher tem- 

 peratures //, is small and one must use another expansion. 

 If B w are the Bernouilli numbers as above, 



X<2 * 3 8 60 5040 ^ 



A? 



9/- N /• . N 



+ w(^l— W+Ol — * 2 ) 



o 



= 31o »£ + I) ( ^~^)- 2M> ^-^ + • •• • +(! 'i- »'»)• 



Whether this equation holds beyond the melting-point it 

 is difficult to say, as the latent heat of fusion of the two 

 isotopes might differ. A more serious matter is the value of 

 the chemical constant, which has not yet been taken into 

 account. 



A number of: attempts * have been made within the last 

 seven years to derive the chemical constant from the element 

 of action li. Without entering into details a simple dimen- 

 sional consideration enables one to predict its variation with 

 the atomic weight, provided such a connexion exists. 



The dimensions of the chemical constant are those of the 

 logarithm of a pressure. In a m on atomic gas one may 

 suppose that it only depends upon the mass m, Boltzmann's 



constant k= », Planck's constant /z, and some temperature #, 



* O. Sacker, Nernst Festschrift, p. 705 (1912) ; H. Tetrode. Ann. d. 

 Phys. xxx^iii. p. 737 (1912). 



'fhil. Mag. S. 6. Vol. 38. No. 223. July 1919. N 



