SURFACES OF DISCONTINUITY 573 



attained a maximum when the mol fraction of the alcohol was 

 about 0.25, and the value there corresponded to an area of 24 

 sq. A per alcohol molecule, which indicates a close-packed uni- 

 molecular layer of these molecules. Thereafter the surface ex- 

 cess rapidly fell, and when the mol fraction was 0.75 the surface 

 excess was apparently no greater than it was when the mol 

 fraction had a value well under 0.1; this value of siu-face excess 

 was apparently maintained for mixtures still richer in alcohol 

 right up to alcohol itself. Exactly similar results were obtained 

 for the surface excess of pyridine at the interface between mer- 

 cury and mixtures of pyridine and water, care being taken to 

 neutralize the electric charge which is known to exist normally 

 at a surface between mercury and water. Now it is highly 

 improbable that there is really a decrease in the surface excess 

 with increase in the proportion of alcohol or pyridine, and the 

 situation shows how troublesome the interpretation of Gibbs' 

 equation may become in particular cases. We have seen that 

 it does definitely point to the existence of a unimolecular layer, 

 and there is also evidence, which we shall touch on later in this 

 commentary, that at least partial orientation of the molecules 

 occurs as well (just as in the case of insoluble films). Now it 

 might happen that with increasing concentration of the alcohol, 

 the more polar water molecules being replaced by weaker alco- 

 hol molecules, there would be a decrease in orientation with an 

 increase in area occupied, caused by each alcohol molecule lying 

 flatter in the surface. But a more probable explanation has 

 been given by Rideal and Schofield, viz., that there is formed 

 below the outer layer of alcohol, a second layer of water. "In 

 the derivation of Gibbs' equation, the mathematical dividing 

 membrane X Y was so placed as to make the adsorption of the 

 water zero — that is, so that the average concentration of 

 water in volumes above and below XY were exactly equal to 

 those in the vapor and the liquid at a distance from the sur- 

 face. If there is a layer of water below the outermost layer 

 of pure alcohol, this will involve placing the dividing surface, 

 not below the alcohol molecules, but some distance above the 

 average level of their lowest points, perhaps more than half-way 

 up the molecules (owing to the thermal agitation this refers to 



