56 Mr. W. Sutherland on Weak Electrolytes and 



acids which on being added to water form a surface film 

 of their own pure substance. This remarkable tendency 

 amongst the alcohols and fatty acids confirms the theoretical 

 deduction that the surface film of water at low temperatures 

 from 0° G. to 40° consists of trihydrol. Otherwise this 

 investigation of the exceptional surface tensions of our 

 mixtures does not advance the immediate subject of our 

 inquiry except that it shows a contrast between the simple 

 course of the equation (50) applying over a wide range of pi 

 to the pure trihydrol of the surface film and the frequently 

 double course of the equations required to express the 

 changes of other properties in the bulk consisting of both 

 trihydrol and dihydrol. 



4. Conclusions to be drawn from these Results. 



In the study of contraction it was found that its course in 

 mixtures of water and alcohol could be fairly well separated 

 into two processes, that of conversion of the trihydrol of 

 water into dihydrol, which is completed when the percentage 

 of alcohol has risen to 46. At this point the experimental 

 contraction is 0*0408, and that calculated from conversion of 

 trihydrol into dihydrol 0*0393. When p 1 is 0*46, the con- 

 traction gives 0'873 for the density of liquid trihydrol in 

 place of 0"88 deduced from the density of ice. When p^ is 

 0*99 and p 2 is 0*01 the contraction for the change of 0*0 L 

 water into hydrol of density 1*26 would be 0*00206, while 

 the experimental contraction is 0'00230. In other words the 

 hydrol produced by alcohol would need to have a density 

 1*31 in place of 1'26 deduced for liquid hydrol from the 

 density of solid hydrol in water of crystallization. With 

 acetic acid and water contraction does not proceed so rapidly 

 with increasing amount of acid as in the case of alcohol. 

 Both processes seem to go on together. But when p 2 is 

 small the contraction is such as would give a density 1*443 to 

 hydrol instead of l - 26. 



This confronts us with the difficulty of having to regard 

 hydrol as a highly variable substance. When we turn to 

 the case of 0"01 water in 099 of formic acid, we find a 

 contraction only half of that occurring with 0*99 of acetic, 

 propionic, and butyric acid. All through, the action of 

 formic acid on water is less vigorous than that of the next 

 three fatty acids. With formic, propionic, and butyric acids 

 the experimental data for small enough values of p 2 are 

 lacking to enable us to ascertain whether the density of 

 hydrol in an excess of these acids is the same as in excess 



