THE COLLOIDAL STATE 109 



dispersion are similarly charged, they wUl mutually repel each 

 other and thus prevent settling. Should they be near collision, 

 the layer of ions surrounding them serves as a protective coat 

 and prevents coalescence. The presence of a protective layer, 

 whether an electric covering of ions, a water mantle, or a mem- 

 brane, is apparently necessary for the stability of colloidal par- 

 ticles. Experimental proof of it for colloidal metals is given by 

 Pauli, whose work also indicates the chemical constitution of 

 the ionic covering. 



As it is necessary to observe absolute cleanliness in preparing 

 colloidal solutions, because metallic suspensions are usually very 

 sensitive to electrolytes, workers were led to believe that a gold 

 suspension can exist only when pure gold is dispersed in pure 

 water; from this opinion, others dissented. Duclaux suggested 

 that metallic colloidal particles are surrounded by a field of ions 

 from salts of the metal. Pauli explained and interpreted this 

 idea by work which proved that pure gold dispersed in pure water 

 does not remain in suspension. When the conductivity of the 

 water is below 3 by 10~^ (reciprocal ohms), and when an abso- 

 lutely clean gold-plated vessel and pure gold electrodes are used, 

 no good dispersion results; but if a trace of salt or acid is added 

 to the water before the metal is dispersed, the suspension stays 

 up. It is possible to make metallic dispersions without the 

 addition of a trace of electrolyte only because the average 

 laboratory-distilled water and the glass vessels used supply 

 impurities enough to give an environment of adsorbed ions to 

 the colloidal particles. The manner in which the electrolyte 

 functions is as follows: When gold is electrically dispersed in a 

 very weak solution of hydrochloric acid, the acid and gold com- 

 bine at the arc to form gold chloride ions and hydrogen ions. 

 The former adhere closely to (are adsorbed by) the gold particles 

 and give to them their characteristic negative charge. The 

 hydrogen ions are free in the water (Fig. 160). Whether the 

 adhering ions are AuCU" or AuCl2~ was not readily determined, 

 but Pauli now believes them to be AuCU", enveloped in a cloud 

 of free H+. In an alkaline medium, the adhering ions are 

 Au(0H)2~, and the free ions K+. In the case of an iron hydroxide 

 dispersion, the situation is similar (Fig. 80) but more complex, 

 for it is not certain whether the core of the particle is Fe(0H)3 or, 

 more likely, (FeOCl)^ (see pages 364, 367). 



