ALUMINIUM HYDROXIDE BY ELECTROLYTES. 15 



66 hours 1.68 and after 22 hours more, 1.64. To what this 

 diminution is due can only be surmised. 



No particular attention was paid to the mode of mixing the 

 colloidal solution with that of the electrolyte. 



7. Coagulative Power of Various Electrolytes. 



In the foregoing investigations we have seen that the relative 

 viscosity tj of the colloidal solution is nearly independent of 

 temperature and not much influenced by time. On the other 

 hand it has been found to be a function of the concentration 

 of the colloidal substance, and it might well be affected by the 

 previous history of the solution. The aluminous solution employed 

 throughout the following experiments was of the same origin 

 and concentration (0.05 9^ of alumina) ; hence the values of vj 

 observed are strictly comparable. We are thus in a position to 

 study the effect of the nature and concentration of various elec- 

 trolytes on the coagulation of the colloidal aluminium hydroxide. 



Most of the substances employed were purified by recrystal- 

 lisation, but some could be obtained only in such small quantities 

 that they had to be used as they were. Of the alkalies, caustic 

 soda was made from metallic sodium, while the others were the 

 ordinary materials of the laboratory. The solutions of Ag(CN)2 K, 

 PO^Nas and the like were prepared by adding the components in 

 proper proportions. Mellithic acid was prepared from mellithrite. 

 The colouring matters were ordinary samples, while the solution 

 of egg-albumin was obtained by dialysing the white of hen's eggs. 



The results of the measurements arranged in the order of 

 the increasing valency of the negative ions of the electrolytes 

 are contained in the following table. The concentrations are 

 given in equivalent normal. 



