400 Mr. R. A. Lehfeldt on the 



We will now introduce the symbols t for the ratio of the 

 quantities of vapour, and s for the ratio of numbers of mole- 

 cules present in the vapour. So that 



and 



mf _ B m' _ B 



^=i+iP, ^=7+1*' 



Making these substitutions, differentiating (9) with respect 

 to q 9 and subtracting (8) from it, we get 



RT f 1 Bs _ 1 -bp 



Z- + 



} 



q + 1 A l*(s + l)dg r p"dq 



2 + 1 B Is + IB? 2>3£ J ' 



(10) 



or 



or 



or 



(_2 L_\ If . (i + JAl §2 -o 



VA<* + 1) B(« + l); 3? + VA + B^ ^ -^ 



/___% A n^ i3^.- n 



Vs(s + 1)(B? + A) (8 + l)(Bq + A)J-dg pbq~ ' 



VB^ + A s + ljs'dq^p'dq K } 



This, then, is the thermodynamic relation connecting/*, q, 

 and s. Experiments may give the vapour-pressure p as a 

 function of the composition q of the liquid ; or the compo- 

 sition s of the vapour as a function of q. But if only one of 

 these data exist, the equation gives the means of finding the 

 other. 



Equation (11), however, is not integrable in general: to 

 make it integrable it is necessary to assume some particular 

 relation between s and q. It follows, then, that if the form 

 of the function s is known p can be calculated : but a know- 

 ledge of p does not suffice to calculate s. 



(ii.) Resume* of Experimental Data. 



Several physicists have measured the vapour-pressure over 

 a mixture of two liquids whose volatility is comparable. 

 The best measurements are, firstly, by Wullner*, whose 

 object was chiefly to investigate the variation of the vapour- 

 pressure with temperature. The liquids he used were (i) water 

 * Pogg. Ann. cxxix. p. 353 (66). 



