3 oo SCIENCE PROGRESS. 



gaseous laws shows that the osmotic pressure effects are 

 proportional to the number, and independent of the nature, 

 of the dissolved molecules. When, then, we find that 

 solutions of electrolytes have abnormally great osmotic 

 pressures of such a magnitude as to be proportional to the 

 number of ions into which the electrolyte can be resolved, 

 it is natural to test if the hypothesis that the substance 

 is dissociated into its ions, in such a manner that the ions 

 become physically independent molecules, cannot express 

 the facts of the case. 



That substances which show great chemical stability 

 when solid are largely dissociated when dissolved is, at 

 first sight, rather a startling idea. We must remember, 

 however, that it is precisely those bodies which show the 

 greatest amount of chemical activity in solution, that is to 

 say, most readily exchange their ions with those of other 

 molecules. The fact that a solution of potassium chloride 

 does not show any of the properties of the elements, potas- 

 sium and chlorine, though it has been urged as an objection, 

 is not a conclusive argument against the theory of dissocia- 

 tion, for the ions are certainly under chemical and electrical 

 conditions very different from those under which the ele- 

 ments exist in their usual forms. Another objection which 

 has been brought forward is that the two ions would diffuse 

 at different rates, and be therefore separable. But, since a 

 definite charge of electricity is associated with each ion, it is 

 obvious that, as soon as an ion got separated, it would leave 

 the rest of the substance oppositely electrified, so that electric 

 forces would be brought into play, tending to draw the ion 

 back, and these would increase as more ions wandered away, 

 till they prevented further diffusion. Still, some separation 

 would occur, and a volume of water, in contact with the 

 solution of an electrolyte, is found to take a potential of 

 the same sign as that of the more diffusible ion, leaving the 

 solution to assume a potential corresponding to that of the 

 less diffusible ion. 



Further evidence is given by the behaviour of semi- 

 permeable membranes. A membrane of copper ferrocy- 

 anide can be prepared which will allow potassium chloride 



