I909.] 



The Dialysis of Colloidal Ferric Hydroxide. 



5^ 



Table III. 

 Action op K3PO4 on Colloidal Ferric Hydroxide Solution. 



These results are quite in accord with the action of the same salt on 

 the Bredig copper colloidal solution already reported by the writer. There 

 is a marked divergence in the fact that quite complete coagulation had set 

 in before the velocity of the particles is reduced to anything near zero. 

 However, in the case of the last two samples coagulation began by the 

 formation of flocculent masses in the body of the liquid while the measure- 

 ments were in progress. These flakes began to settle at the bottom of the 

 velocity tube so that the reading was really carried out on the finer particles 

 left in suspension in the liquid. As a result, the velocity numbers given 

 in these two cases are in the nature of the velocities of the particles bearing 

 the maximum charges at the time. 



There remains, however, no doubt of the keen action of the potassium 

 phosphate in discharging the particles and producing coagulation. 



c. Coagulating Power of Ions of Different Valencies. 



A mixture of 5ccs. of the dialysed ferric hydroxide and 250 ccs. of 

 conductivity water was made up for this test. The coagulative influence 

 of various salts on this sample was tried. The following salt solutions 

 were made up: Normal potassium chloride, 1/30 normal potassium 

 sulphate, i/ioo normal aluminium sulphate, and i/iooo normal potassium 

 phosphate. In this way was obtained a series of solutions containing 

 univalent, divalent, and trivalent negative ions in ratios which are the 

 inverse of those representing the coagulative powers of the various ions 



