towards a Dynamical Theory of Solutions. 63 



What occurs at the transition is this : as p 2 diminishes below 

 0*143 and p 2 increases above 0'857, the water molecules 

 surround the double molecules of acid so freely and make 

 destructive conjunctions with each separately so often, that 

 the special conjunction of six molecules of water with two 

 double molecules of acid sinks into comparative insignificance. 

 Since in these theoretical considerations we have introduced 

 pp the mass per unit volume, it is desirable to amend the 

 equation (58) to the following form : 



(\ V f=ap sP , (61) 



in which a has the value 0*242 for acetic acid, 0216 for 

 propionic, and 0'235 for butyric. For formic acid the 

 corresponding equation is 



(\7 ? )*=0-254(j9 2 + 0-25)p (62) 



The close approach of 0*254, 0*242, 0*216, and 0*235 to 

 equality is noteworthy, being due probably to the fact that 

 in each case the formation of the double molecule is due to 

 the junction of COO with COO by the bonds £b between 

 O atoms. It is interesting to notice that formic acid, which 

 is exceptional for its large molecular conductivity in dilute 

 solutions and its large K in Ostwald's formula, is also 

 exceptional for its larger molecular conductivity in solutions 

 which are not dilute and in having (62) in place of (61). 

 This continuity in the exceptionality of formic acid confirms 

 the theoretical attempt above to explain the nature of the 

 continuity connecting the two types of equation for molecular 

 conductivity. This continuity is evidence in support of the 

 view that throughout the solutions of the fatty acids we have 

 to do with double molecules. In the pure acids we have to 

 do with such, probably therefore with a preponderance of 

 such in their aqueous solutions containing 80 per cent, of 

 acid. The nature of the continuity is most simply explained 

 by the assumption that in an 80 per cent, solution of a fatty 

 acid, as in its solutions of all other strengths, the molecules 

 are all double. Otherwise, according to the current theory 

 of weak electrolytes we need a law for the dissociation of the 

 double molecules and a law for the ionization of the resulting 

 single molecules. To give the broad similarities and sim- 

 plicities we have just been discussing, these two laws would 

 need to be brought into simple relations. In view of all the 

 facts it seems to me that the better hypothesis is that the fatty 

 acids in their aqueous solutions exist as double molecules, the 

 Ostwald dilution formula giving the law of the equilibrium 

 between the double molecules and the ions derived from 



