towards a Dynamical Theory of Solutions, 19 



is added an isomeric form of itself having the same freezing- 

 point. Here it is not to be expected that the rule of 

 the molecular lowering of freezing-point will apply. The 

 difference between the molecule of solute and of solvent 

 must reach a limit of some sort before the rule applies. 

 Similarly in another direction we must expect a limit to 

 the applicability of the rule of molecular lowering of freezino- 

 point. Imagine that in (CH 3 COOH) 2 each of the CH 3 COOH 

 o-roups is so loosely attached to the other that they are just 

 on the verge of parting company, then in relation to the 

 molecules of solvent each CH 3 COOH group has almost as 

 much freedom of motion as if it were detached from the 

 other. In such a case, until we have dynamical proof to 

 the contrary, we may expect that the method of the lowering 

 of the freezing-point will give the molecular mass as that of 

 the nearly free CH3COOH instead of its double. The 

 difficulty seems to me analogous to that which arises in 

 acidimetry when working with acid of strength comparable 

 with that of the indicator. The lowering of freezing-point 

 is generally a good indicator of molecular mass, but near 

 the limits of its applicability, its results need careful col- 

 lating with those from other sources. It seems to me that 

 the theory of the physical chemistry of electrolytes has gone 

 astray by accepting for a foundation the simplest chemical 

 hypothesis which would account for Ostwald's dilution law 

 for weak electrolytes, and also for the molecular lowering 

 which they cause in the freezing-point of water. Originally, 

 the primary object of the present paper was to test critically 

 m y suggestion as to the meaning of OstwakTs dilution law 

 with weak electrolytes, but in the course of the work it 

 seemed better to widen the scope of the inquiry, and to 

 investigate a few typical solutions from the molecular 

 dynamical point of view as far as seemed likely to be 

 profitable in the absence of a kinetic theory of liquids. The 

 solutions chosen are mixtures of ethyl alcohol and water to 

 represent non-electrolytes, and mixtures of water with acetic 

 and other fatty acids to represent weak electrolytes, all of 

 which have been elaborately investigated by experimenters 

 whose results have not received a correspondingly thorough 

 study from the dynamical point of view, so that the theory of 

 physical chemistry has been kept too narrow for the facts. 



To these solutions the supporters of the hydrate theory 

 of solutions have devoted some attention, tracing their 

 salient peculiarities to the existence of hydrates, such as 

 C 2 H 5 OH + 3H 2 in the case of alcohol, and CH 3 COOH-f 

 H 9 in that of acetic acid. It has been suggested that the 



2 



