March io, 192 i] 



NATURE 



47 



V 



lysis, that the hydrolytic alkalinity of soap solu- 

 tions is for most purposes negligible, and hence 

 that the conductivity observed must be proper to 

 the soap itself. Incidentally, this result is of in- 

 terest in showing that the process of saponifica- 

 tion in the manufacture of soap could be much 

 more complete than was thought by such authori- 

 ties as Lewkowitsch. 



A further essential stage in the development of 

 this problem was attained through the study of the 

 osmotic activity of the soap solutions. This pro- 

 perty is, in such cases, surprisingly inaccessible 

 to trustworthy quantitative measurement. How- 

 ever, a development of Cumming's dew-point ap- 

 paratus gave a general method of securing data, 

 and the results were confirmed by cryoscopic 

 measurements upon the few soaps which could 

 be studied in solution at o°. The upshot is that 

 a mass of trustworthy data proves that soaps ex- 

 hibit osmotic activity comparable with that of an 

 ordinary crystalloid such as sugar. 



This at once exposed a fundamental difficulty 

 in interpreting the results according to any of the 

 other hitherto recognised theories of physical 

 chemistry. The conductivity is that of a highly 

 dissociated salt, whereas the osmotic activity is 

 scarcely equal to that of an undissociated crystal- 

 loid, and yet many years of work had been de- 

 voted to establishing the trustworthiness of each 

 of these facts. Examination of the results of the 

 concentrated solutions of the higher soaps showed 

 that, whereas the conductivity corresponded to 

 that of two good conducting ions, the osmotic 

 pressure was only that of one ion altogether. In 

 other words, the osmotic result proved that the 

 only crystalloidal constituent of such a solution 

 was the sodium or potassium ion, all the other 

 constituents, including whatever accounted for 

 quite half the conductivity, being colloidal. 



Hence we are driven to the conclusion that 

 there are present in these solutions colloidal par- 

 ticles, the "ionic micelle," possessing an actual 

 conductivity often several times greater than that 

 of the sum-total of the ions which are contained 

 in it, and which in so aggregating have retained 

 their electrical charges. These aggregates are 

 so large that they have little or no osmotic effect. 

 For suggestions that make plausible the proper- 

 ties and stability of such aggregates, reference 

 must be made to papers published by the Royal 

 Society and the London and American Chemical 

 Societies, where also it is shown how these con- 

 ceptions explain the various properties of soap 

 solutions. Direct measurements are now being 

 carried out to test even more directly the validity 

 of the explanations here advanced. 



For the sake of clearness it should be empha- 

 sised that conductivity is not identical with rate 

 of movement in an electric field, for it is a re- 

 markable fact that matter in all states of sub- 

 division from single atomic ions up to coarse 

 granules may move at roughly the same rate in 

 an electric field. This movement (cataphoresis) 

 in the case of a fine grain of sand might thus be 



NO. 2680, VOL. 107] 



equal in magnitude to that of one of the slower 

 ions, whereas the resulting equivalent conductivity 

 is only infinitesimal. The ionic micelle of soap 

 solutions is noteworthy in that its mobility in an 

 electrical field exceeds that of most true ions. 



It is probable that quite general laws underlie 

 the behaviour of colloidal particles together with 

 all surfaces of separation in which ionising 

 solvents are involved, thus including emulsions 

 as well as large continuous surfaces. 



In another respect, too, soap solutions afford 

 a particularly good example for the study of 

 a colloid in that the whole gamut of transition 

 stages between ordinary salts and colloids can be 

 illustrated by choosing the salts of the various 

 fatty acids, or even by a mere change in concen- 

 tration of a solution of any one of these. In 

 dilute solution the soaps are largely present as 

 simple salts, whereas in concentrated solutions of 

 the higher soaps we have the complete formation 

 of colloidal electrolyte. 



Having gained some insight into the properties 

 and behaviour of the slightly charged colloids and 

 the highly charged colloidal electrolytes, the 

 greatest need at the present time for the develop- 

 ment of colloid chemistry is the discovery of some 

 method of studying neutral uncharged colloids, 

 such as, for instance, rubber or nitrocellulose solu- 

 tions. No one has yet succeeded in developing a 

 general method for obtaining quantitative data 

 of direct significance, and a big advance is to be 

 hoped for in this direction. This would probably 

 lead to rational methods for the study of such 

 familiar but complicated structures as the textiles, 

 or paper, in which solvent is no longer present. 



Recent study of soap solutions in the Bristol 

 University laboratory has shown, further, that 

 they can exist in three distinct characteristic 

 forms — namely, clear, somewhat viscous, liquid 

 sols, transparent elastic gels, and white opaque 

 curds. Nearly all our previous knowledge of the 

 properties of jellies has been due to the study of 

 gelatin, usually containing admixed and partly 

 combined salts or acids. The simpler case of the 

 soap gels is, again, suited for study because 

 no extraneous substances are present, and, as 

 we have seen, the various constituents of the soap 

 solution are characterised by well-marked proper- 

 ties such as conductivity and osmotic activity. 



It has now been shown that the properties of 

 soap solutions are independent of whether the 

 solution is in the form of sol or gel except for 

 the distinctive mechanical properties of the latter. 

 In other words, the chemical equilibria, and 

 hence the colloidal particles, are identical in sol 

 and gel. This means that the gel structure must 

 be built up of the same colloidal particles as 

 were present in the sol. The possibilities as to 

 the nature of this structure are severely limited 

 by the fact that the conductivity remains un- 

 altered. Hence we must infer that the colloidal 

 particles are stuck together to form loose aggre- 

 gates, which may be fragments of irregular net- 

 work, or more probably innumerable filaments, 



