14 ART. 8. — S. KAWAMTJEA : COAGULATIOlSr OF COLLOIDAL 



expected, but varies more slowly. Even when the viscosity of 

 pure water instead of the respective aluminous solution is taken 

 as the standard of comparison, strict proportionality does not 

 obtain. But this is after all not very strange, because the ap- 

 parent viscosity observed in the these experiments can not be a 

 very simple function of the number of suspended particles. 



The simplest empirical equation, which rejoresents the rela- 

 tion, will have the form : 



where c stands for the concentration of the colloidal solution, 

 and a and ß are constants. In the jiresent case it has been 

 found by trial that « = 10 and ß=0.9. 



Concentration c. 0.05 0.025 0.0325 0.005 



"-yiim-l (found). 0.G8 0.36 0.21 0.13 



■^iim-1 (calculated). 0.68 0.36 0.20 0.09 



6. The Effect of Time. 



As the colloidal solution is essentially an unstable system 

 and its coagulation is generally a change to a more stable state, 

 time must exercise some effect on the phenomena. This has been 

 confirmed in the case of arsenious sulphide. But in the present 

 case no marked effect was observed. This is well shown by the 

 constancy of the viscosity of various solutions, partially or com- 

 pletely coagulated, during the time in which repeated measure- 

 ments of the time of flow were made. Yet when a completely 

 coagulated solution is allowed to stand for a comparatively long 

 time the viscosity gradually diminishes. For example, in a 

 solution whose relative viscosity was initially 1.76, it became after 



