27 



months, and in h per cent, solution for days. In the case of ferric 

 oxide, the optimum concentration was I per cent, soap, and in the 

 case of potters' clay, 1/32 per cent. soap. Alumina showed remarkable 

 periodic optima in 1/4, 1/8, and 1/16 per cent, soap, with a similar 

 numerical periodicity of coagulation. Zhuknov and Shestakov found 

 that in laundry practice the best results were obtained with • 2 to 

 0-4 per cent, soap solution. 



Donnan and Potts found a similar optimum concentration in the 

 emulsification of paraffin oil by soap solutions in a concentration 

 of N/300. They emphasize that the excess of soap in the interface 

 which is a necessary corollary of the lowered surface tension and 

 the viscous natiu-e of the surface film must both contribute to the 

 stability of the emulsions. Pickering (Jorir. Chem. Soc, III., 86) 

 agrees both with Donnan and Spring, but considers also that oil 

 and other substances are soluble in the presence of soap. 



Jackson (Jour. Soc. Arts, 55, 1101 and 1122 (1908), Cantor 

 Lectures) called attention to the influence which soap exerts upon 

 the state of subdivision of the dirt, and he observed under the micro- 

 scoj)e dirt particles and fibres of linen being brought first into oscilla- 

 tion and then completely loosened by a soap. This spontaneous 

 action was best exhibited by an alkaline oleate. He also, pointed 

 out that the presence of glycerine in soaps had only an unimportant 

 influence upon their detergent action. 



Shorter (loc. cit) adduces experiments from which he concluded 

 that acid soaps exhibit no surface activity and that detergent action 

 is due mainly to undecomposed soap. Excess of alkali enhances 

 the detergent action. Developing the consideration of the electrical 

 effects introduced by Donnan and Potts (loc. cit), he points out that the 

 effect of the alkali is to increase the negative charge both on the 

 particles of oil and dirt and on the surface to be cleansed, which tends 

 to prevent coagulation and redeposition of those particles. 



It emerges from aU this discussion that there are a number of 

 definite factors in detergent action; first, the necessity of having 

 the soap in solution; second, power of emulsification which goes 

 ])ai-allel with low surface tension and the formation of surface films; 

 third, wetting power which like the last, is ascribable to the unde- 

 composed soap itself; foux'th, the action of soap in forming non- 

 adhesive coUoidal sorption compounds wdth tissue and impurities 

 due sometimes to acid soap, but more often to soap itself and capable 

 of remaining in stable suspension; fifth and lastly, it is an essential 

 in all cases that the soap should be in colloidal form. 



It is evident that comprehensive quantitative work is necessary 

 to complete and co-ordinate the existing fragmentary work in any 

 one case. Each of these factors are perfectly cajjable of simultaneous 

 determination and quantitative evaluation. 



IX. — Retrospect. 



We have seen that all the phenomena of soap solutions point to 

 the existence of a highly conducting heavily hydrated ionic micelle 

 of the general formula (NaP)j; • (P')„ ' (HoO),„ ; and that from 



