54 Transactions of the Canadian Institute. [Vol. IX. 



which had been gradually denuded of its chlorine content, and found 

 good evidence of the charge being due to the presence of the chlorine in 

 the particle. The purpose of the present paper is to examine this theory 

 from another point of view. 



Duclaux's decisive experiments are those determining the coagulative 

 power of various salts on a given sample of colloid and the coagulative power 

 of the same salt on various samples of the colloid. Undoubtedly the surest 

 way of testing the charge on a colloidal particle is to measure its velocity 

 in a given electric field. This method has been used by the writer and 

 leads to an extension of our knowledge of the charge possessed by the 

 ferric hydroxide particle. 



The experiments may be divided into three sections : 



a. Continued dialysis of ferric hydroxide against conductivity water 



to find if the charge on the particle decreases as the chlorine is 

 eliminated : 



b. Test of the action of the addition of electrolytes on the size of the 



charge on the particle : 



c. Test of the coagulating power of ions of different valency on a 



selected sample of the dialysed colloid. 



a. Relation of the Charge to the Chlorine Content. 



Ordinary commercial ferric hydroxide solution sold by druggists 

 tmder the name of ' ' Dialysed Iron" was used for the work. It is prepared, 

 according to the method indicated above, by the action of a base on ferric 

 chloride and is dialysed by the manufacturer, I presume against ordinary 

 water. When analysed it gave proportions of chlorine and iron corres- 

 ponding to the formula, Fe^Cl^. 13.6 Fe^O^. The original solution was 

 very dense in colour and had a very high electrical conductivity. The 

 velocity with which the particles moved in a given electric field was 

 measured by the direct method with a U-tube similar to that used in 

 previous work (4). If the conductivity of the liquid under consideration 

 is very high the surface, the displacement of which we observe, does not 

 remain distinct and so the method fails. In order to overcome this 

 difficulty, whenever a velocity measurement was to be taken about ^ ccs. 

 of the colloidal solution was mixed with 50 ccs. of conductivity water 

 and the velocities of the particles in these samples was measured. Of 

 course it is quite apparent that we are not then dealing with the exact 

 solution which we are dialysing, but, as each sample from time to time was 



(4) Phil. Mag. S. 6, Apr. 1906. (Burton). 



