472 Mr. W. C. M. Lewis : Experimental Investigati 



ion 



been exhausted and the total number of drops estimated 

 from the time taken in emptying the reservoir, the solution 

 which had been thus treated was withdrawn and analysed. 

 For this particular substance the silver only was estimated 

 before and after the experiment by precipitation with hydro- 

 chloric acid, the Gooch crucible method being employed. 

 An extremely small difference in weight was observed, 

 namely 1'6 milligrams of silver chloride. The total surface 

 area of the oil was 7000 cm. 2 



Hence the adsorption of the silver is 



T H = 1-7 X 10- 8 o-nn./cm. 2 

 A repetition of the determination gave 



T Ag =: 3'4xl0- 8 grm./em. 2 

 The mean value is therefore 



T Ag = 2-5xlQ- 8 grm./cm. 2 



It may be mentioned that the measuring vessels used in 

 these and subsequent determinations were recalibrated for 

 15° C, and all solutions were cooled to this temperature in 

 a bath before measuring out the volume. 



It will be noted that the value obtained above is much 

 smaller than for the previously mentioned substances (except 

 caustic soda). One might therefore expect a closer approxi- 

 mation to the value calculated on Gibbs's expression. 



Calculation of the Adsorption of Silver Nitrate. 



As in previous cases the oil-solution tension curve was 

 obtained by means of the drop-pipette. The oil employed 

 for these measurements was not quite identical with that 

 used with caustic soda. The values obtained (corrected for 

 density) are given in the following table. 



Oil 



Table IT. 

 ■Silver- Nitrate Solutions. 



Concentration, 

 in gram-moles per litre. 



Pipette 

 Drop-number. 



Tension 

 dynes, cni. 







135 



52 



01 



140 



50 



0-02 



145 



48-4 



0-04 



149 



473 



