Mutual Solubiliti/ of /liquids. 651 



showing a transient turbidity precedinf^ the settling out of 

 the liquids, and in view of the values, given by Guthrie him- 

 self, of the refractive indices his explanation of the non- 

 appearance of the milkiness is obviously insufficient. A 

 more probable explanation would be the comparatively small 

 difference in density and composition between the two phases, 

 as shown by the narrow pointed shape of the solubility-curve 

 (t-x). Still some doubt remains, and what Guthrie saw 

 may have beeu due to unequal heating; if the liquid had 

 been hotter than the space above, distillation would have set 

 in and recondensation, with the result of a temporary second 

 laver settling down on top of the original liquid with a very 

 faint line of demarcation. 



General Conclusions. 



It was first noticed by Lehfeldt * that of all partially 

 miscible pairs of liquids at least one component is an abnormal 

 substance ; no exceptions to this rule have yet been found. 

 We find, therefore, that those molecular conditions — probably 

 to be represented by a high molecular attraction — which in 

 a liquid by itself produce certain deviations from the laws of 

 corresponding states, characteristic of abnormal liquids, at 

 the same time seem to make it impossible for that liquid to 

 dissolve more than a limited amount of a second substance 

 without entering upon an unstable condition. A molecular 

 explanation of this fact cannot yet be given, except in so far 

 as we can suppose the foreign molecules to be squeezed out 

 by the high attraction of the others. 



The observations recorded under e, f, and (j show that, as 

 we ascend the alcohol series, the 'temperature below which 

 two layers are formed becomes lower : this is in accordance 

 with the theory that the abnormality of the alcohols plavs a 

 part in the phenomenon of partial miscibility, for we know 

 that the higher the alcohol the less abnormal it is. 



The experiments show the influence of the hydrocarbon to 

 be more complicated ; no doubt the phenomena depend to a 

 large extent on the gradual rise of the vapour-pressure and 

 the approach of the critical point of the hydrocarbon as we 

 come down the series. 



As regards the influence of pressure on the critical point of 

 the liquid layers, this we have found to be sometimes in the 

 direction of increasing and sometimes of diminishing the 

 mutual solubility ; we know from thermodynamics that this 

 difference is associated with a difference in ttie volume-change 



* Lehfeldt, Phil. Sl'^rr. (o) yol. xlvii. p. 285. 



