42 Mr. W. Sutherland on Weak Electrolytes and 



F line, which is near to the two alternative values for water 

 dissolved in an infinite volume of alcohol. This tends to 

 confirm our conclusion from contraction that water in an 

 infinite volume of acetic acid is changed into hydrol. 

 Returning to the minimum, we see that it is caused by the 

 conversion of trihydrol into dihydrol which lowers r/ until 

 the process of changing dihydrol into hydrol with larger r 2 ' 

 brings the diminution to an end and gives the minimum. 

 In the study of contraction we concluded that with acetic 

 acid the process of forming hydrol makes itself apparent 

 at much smaller values of p { than in the case of alcohol. 

 Refraction confirms this result. I shall not analyse existing 

 experimental facts in greater detail as I think we need, as 

 with alcohol, a special set of experiments for p and r' at 

 different temperatures, the data to be handled by means 

 of equations (31). 



Dielectric Capacity* 



Here the experimental and theoretical difficulties are well 

 known in connexion with mixtures. I shall make (K-— 1)//? 

 the subject of study with the supposition that for an ideally 

 simple mixture 



(K~l)/p=p 1 {K 1 -l)/ Pl +p. 2 (K 2 -l)/p,. . (32) 



In our exceptional mixtures we shall regard (K 2 — 1)//? 2 as 

 varying with p x on account of the changes occurring in 

 water as liquid 2, and shall calculate values of (K 2 — l)/p 2 

 from those of (K— l)/jO-by (32). For alcohol and water we 

 have the data of Nernst (Tabellen) from 19° C. to 20°, 

 say 20°, and those of Thwing (Ztschr. f. ph. Chem. xiv. 

 p. 286, 1894) at 15° C. Thwing's results for mixtures of 

 associating liquids and for water showed some remarkable 

 singularities, which in the case of mixtures he interpreted as 

 due to hydrates or similar compounds. In the case of water 

 he found a sharp maximum for K at 4° C. This maximum 

 was specially looked for by Yonwiller (Phil. Mag. [6] vii. 

 p. 655, 1904) but not found, nor have Thwing's singularities 

 for mixtures been found by other experimenters. Yet 

 Thwing's values for pure substances and for water not near 

 4° C. agree well with other determinations and the general 

 course of his values for mixtures of associating liquids 

 agrees with that of other observers. ISTernst's data give the 

 results : — 



lOOj^ 300 90 80 70 60 50 



K 26 29'3 33*5 38'0 43-1 4S-5 (about) 



(K a — l)/p a ............ ... 61-0 65-5 687 70'2 72-0 



