560 



RICE 



ART. L 



r which were about 80 times as great as those calculated on the 

 basis of Gibbs' equation. The sodium salt of congo red, methyl 

 orange and sodium oleate were also tried and exhibited a similar 

 though less marked discrepancy. Despite the experimental 

 difficulties of the tests, there was no possibility of ascribing 

 these results to experimental errors or to the use of the simple 

 form of the equation. The excessive adsorption was almost 

 certainly a characteristic of the semi-colloidal solutes employed. 

 Subsequently Lewis used a solute of much simpler constitution, 

 and one truly soluble in the aqueous medium, viz. aniline, and 

 measured the adsorption at a mercury-water interface (Zeit. 

 physik. Chem., 73, 129, (1910)). The calculated and observed 

 adsorption values now showed agreement as regards order of 

 magnitude, both being small multiples of 10"^ grams per sq. cm. 

 A still more successful test was carried out by Donnan and 

 J. T. Barker {Proc. Roy. Soc, 85 A, 557, (1911)) who measured 

 the adsorption of nonylic acid at an air-water surface. The T 

 was evaluated from the expression { — c/Rt) (dcx/dc) and cal- 

 culated, first, on the assumption of non-ionization of the acid 

 and, second, on the assumption of complete ionization. The 

 table gives the observed and calculated values. 



Adsorption of Nontlic Acid at Air-Water Surface 



Donnan and Barker also measured the adsorption of the 

 glucoside saponin at an air-water surface; this forms very stable 

 foams and viscous films at the bounding surface of air bubbles. 

 There was agreement between the orders of magnitude of r 

 observed and calculated, but from a substance of this character 

 little more could be expected, and the results with nonylic acid 

 are of greater value. 



Patrick (Zeit. physik. Chem., 86, 545, (1914)) investigated the 



