SURFACES OF DISCONTINUITY 575 



easy to see how this can be done in connection with the work of 

 McBain and his colleagues. There the difficulty is different from 

 that just dealt with. The experiments on amyl alcohol, for 

 example, show that the measured amount at maximum adsorp- 

 tion was so great that if it were packed in a unimolecular layer 



o 



the area was only 14 sq. A for each molecule; in the case of 

 sodium oleate only 1 1 sq. A. It is impossible for these molecules 

 to be packed so tightly in a layer one molecule thick. It may be 

 possible, as we have stated earlier, that there may be a uni- 

 molecular layer with the additional material forced out into 

 small droplets above it here and there, the unimolecular layer 

 being the true adsorption agreeing with the adsorption equa- 

 tion. But clearly these difficulties still await solution. It is 

 interesting to note that a somewhat similar situation exists in 

 connection with insoluble oil films. The evidence for uni- 

 molecular layers is strong, yet there can be no doubt that the 

 area of an oil film can be reduced until there is no longer room 

 for all the molecules at their closest possible packing. The 

 suggestion is that the film gives way under tangential squeezing, 

 buckles and expels enormous numbers of molecules to form local 

 ridges, the rest of the film being unchanged. Adam in his book 

 hazards the opinion that the cases of "polymolecular" films such 

 as those obtained by Harkins and Morgan {Proc. Nat. Ac. Sci., 

 11, 637, (1925)) are really examples of "partially collapsed uni- 

 molecular films, with the excess material collected into lumps 

 much thicker than the film itself." 



X. Desorption 



27. Unimolecular Layers and Negative Adsorption 



If a solute raises the surface tension of a solution above that of 

 the pure solvent, the Gibbs' equation shows that the calculated 

 value of r2 (Fi being made zero as usual) is negative. This is 

 interpreted by saying that the surface is poorer in the solute 

 than the bulk phase or (alternatively) richer in the solvent. In 

 the nature of things "negative adsorption" cannot reach such 

 large numerical values as the positive ; obviously it cannot exceed 

 the bulk concentration of the solute divided by the thickness 

 of the layer in numerical amount.* 



* See Gibbs. I, 274. 



