10 



is very decidedly less at 0° than at 90°. Soap solutions are very 

 much more colloidal at lower temperatures than at the boUing pomt, 

 and the presence of colloid extends into comparatively dilute solutions 

 atO°. 



In case of the higher soaps the relative osmotic activity is practically 

 independent of the concentrations above • 4 N at 0°, showing how 

 complete the formation of colloid has been. Again, potassium 

 octoate is distinctive in its behavoiur at 0° in that in lower concentra- 

 tions it is a typical electrolyte whereas in extreme concentrations 

 the relative osmotic activity rapidly diminishes. 



(c) Concentration and Composition of the Ionic Micelle. 



The evidence for the existence of the ionic micelle is based primarily 

 upon comparison of conductivity and osmotic data. Briefly it 

 amounts to this, that in concentrated solution of the higher soaps, 

 the osmotic activity is often only about one half that required to 

 explain the conductivity. The conductivity is, say, 2/3 of that of a 

 salt like the acetate, whereas the osmotic pressure is, say, half that 

 of a non-electrolyte such as sucrose, and yet both sets of data are 

 fully trustworthy. The osmotic activity, therefore, corresponds with 

 tliat of one ion only, and the other half of the current must be carried 

 by an ion that is colloidal so as not to exhibit appreciable osmotic 

 activity, and that nevertheless retains the sum total of the electriral 

 charges of the ions from which it was derived. This is the ionic 

 micelle. 



The osmotic activity measures the total concentration of crystalloidal 

 matter of all kinds. The conductivity measures the K " or Na ■ in 

 addition to the carrier of the equivalent negative charges. To 

 reconcile the experimental data, say in normal solution of potassium 

 laurate at 0°, it is necessary to ascribe the whole of the observed 

 osmotic activity to the potassium ion, and even when this is done, 

 about half of the conductivity is still to be accounted for and must 

 then necessarily be ascribed to colloid. It is notable that the 

 equivalent conductivity thus ascribed to the ionic micelle of potassium 

 laurate is about three times greater than the sum total of the 

 conductivity which the separate laurate ions that are groujaed in the 

 micelle, would have exhibited if they had retained an independent 

 existence. This, however, as the writer has indicated above, is only 

 what would have been predicted from Stokes' Law, since the resist- 

 ance offered to a particle increases directly with its diameter. When 

 a number of small particles coalesce to form larger particles, the 

 diameter of the larger particle does not increase by any means in 

 the same ratio, whereas the electrical driving force will be directly 

 proportional to the number of aggregated laurate ions if these latter 

 have retained their equivalent electrical charges. Therefore for 

 electrical forces the mobUity of such a large particle is very great, 

 whereas its diffusibility is only that of a colloid. 



It is necessary to make some sort of assumption as to the mobility 

 of the ionic micelle for a particular case in order to be able to evaluate 

 the conductivity observed. A safe provisional rule is to ascribe to 

 the ionic micelle, the mmimum conductance which will serve to 



