Mr. J. A. Wanklyn on Fractional Distillation. 131 



Such, I believe, is the real march of the process of fractional 

 distillation ; and these formulas oiFer an explanation why by fre- 

 quent repetitions and judicious management the chemist is able 

 to effect almost perfect separations by means of fractional distil- 

 lation. The coefficient of volatility, however, is something 

 more than the tension of the vapour at the boiling-point of the 

 mixture. Ten years ago I showed that the density of the vapour 

 had to be taken into account (and Berthelot has also indepen- 

 dently, but, I believe, subsequently, insisted on the necessity of 

 regarding vapour-density in judging of relative volatility). The 

 coefficient of volatility comprises at least the tension of the vapour 

 and the density of the vapour. 



The heat rendered latent during evaporation must also be re- 

 presented in this coefficient. Possibly, too, adhesion between 

 the liquids will require representation in certain classes of in- 

 stances; and this will cause difficulty and complication. I am 

 of opinion, however, that in the vast majority of cases the con- 

 joined influences of tension, density, and latent heat will ade- 

 quately express the coefficient of volatility. If t be put for the 

 tension and d for the vapour-density, then I propose v = td 2 

 for the value of the coefficient of volatility. This will be more 

 fully explained on a future occasion. At present my main object 

 is to insist on the fact that, when distilling in mixture, different 

 liquids have different coefficients of volatility, and that the com- 

 position of successive portions of distillate is governed by these 

 coefficients and by the original proportions of the liquids present 

 in the mixture, in the manner set forth. 



I have made some experiments on mixtures of A and B 

 wherein the quantity of A is overwhelmingly greater than that 

 of B. This disposition of the quantities offers certain advantages 

 for investigating the general march of fractional distillation. 



When a is immense compared with b, 7 is undistinguish- 



able from -. And putting v=l, when a is immense compared 



bv' ,. ,. .-,,,„ bv } 





D av 



+ bv' 



auguiBuauioui 



a ' 



four 



experiments, 



wherein 



a + 



b 





Exp. 



I. . . 



1000000 + 1000 





Exp. 



II. . . 



1000000 + 



1-00 





Exp. 



III. . . 



1000000 + 



050 





Exp. 



IV. . . 



1000000 + 



0*20 



I have made 



In these cases the quantity of b in a unit of the original mix- 



K2 



