34 Mr. W. Sutherland on Weak Electrolytes and 



infinite dilution A//? 2 reaches the same or nearly the same 

 limit with formic acid as with the next three acids of the 

 fatty series. Probably the chief cause of the difference 

 between the action of excess of formic acid on water and 

 that of the other three acids is the fact that the dielectric 

 capacity of formic acid is much nearer to that of w r ater than 

 is the case with the others. The dielectric capacities are 

 about 80 for water, 60 for formic acid, 10 for acetic, 5 for 

 propionic, and 3 for butyric. When a dilute solution of 

 water is made with acetic, propionic, or butyric acid, the 

 water is surrounded by substance of very different dielectric 

 capacity from itself, and this difference seems to promote the 

 splitting of dihydrol and trihydrol into hydrol. In formic 

 acid the water is surrounded by a substance of more nearly 

 its own dielectric capacity, and so dihydrol and trihydrol 

 have more nearly the stability which they possess in water. 

 In regard to ethyl alcohol similar considerations apply, as 

 its dielectric capacity is 25. Of course dielectric capacity 

 is only one of many factors controlling stability in such 

 mixtures. By (25) for formic acid, if we neglect the small 

 term /oA 2 /2, we have A proportional to pip 2 p, which, except 

 for the entrance of p, is the law deduced in " The Mol. Const. 

 of Aqueous Solutions" (Phil. Mag. [6] xii. p. 1, 1906) for 

 the contraction that occurs in the formation of aqueous electro- 

 lytic solutions. As solutions of formic and the other fatty 

 acids at the dilutions we are studying are almost completely 

 non-electrolytic, it is interesting to find a similar law for 

 part of their contraction to that which holds for electrolytic 

 solutions. It was shown in that paper that positive ions 

 change trihydrol into dihydrol while negative ions change 

 dihydrol into trihydrol. A gramme-equivalent of positive 

 ion changes between 1'3 and 2*5 gramme-molecules of 

 (H 2 0) 3 into (H 2 0) 2 , while a gramme-equivalent of negative 

 ion forms out of (HoO) 2 a number of gramme-molecules of 

 (H 2 0) 3 which ranges from 0'25 for CO s to 1-57 for CH s COO 

 the acetic ion. From (15) for acetic acid, neglecting the 

 small terms pA 2 /2 and 0*045pi, it is easy, neglecting hydrol, 

 to calculate the number of gramme-molecules of (H 2 0) 3 

 converted into (H 2 0) 2 by a gramme-molecule of (CH 3 COOH) 2 , 

 namely 1-014, say 1/0. Thus the action of (CH 8 COOH) 2 in 

 changing tri- into di-hydrol is nearly equal to that of a 

 monovalent positive ion. We see clearly the importance of 

 molecular and ionic electric fields in the causation of chemical 

 change. 



