Electrification at Liquid-Gas Surfaces. 313 



that in the static type of experiment the charge at a water- 

 gas surface is very slightly affected by the nature of the gas. 

 Let us suppose for simplicity that the gas has no effect and 

 that — following Prof. J. J. Thomson's view — the liquid in 

 the surface layer is chemically unsaturated. The adsorbed 

 gas, oxygen or hydrogen for example, would become attached 

 to water molecules, but in such a way as not to affect the 

 value of the charge in the electrical double layer, which 

 would have the same sign and value as if the gas were 

 absent. When, as in a waterfall experiment, a separation of 

 the gas and water molecules occurs, the electrons are shared 

 between them in a different w r ay. Oxygen, having an 

 affinity for negative electrons, will carry away an excess of 

 them, giving the gas a negative charge. Hydrogen will 

 leave an electron behind in the water and go off positively 

 charged. In other words, the dynamical experiment brings 

 out the difference in the chemical, i.e. the electrical, nature 

 of the ga.«. 



This simple view may be slightly modified in an actual 

 case. Oxygen might attract negative ions into the surface 

 and give it a larger negative charge than hydrogen would, as is 

 shown, for example, in the comparison readings for these two 

 gases in A1 2 (S0 4 ) 3 . In the static experiments the ions in the 

 liquid will determine the surface charge, and since in water 

 a surface is always negative, one may suppose the existence 



of a selective adsorption of OH ions. In support of this is 



+ 

 the fact that the surface is electrically neutral when H ions 

 are in excess, as in acid solutions (leaving polyvalent ions 

 out of consideration). When dissolved salts are present, 

 their adsorption modifies the charge again, but the gas in 

 contact still shows scarcely any effect. This neutrality was 

 observed by Metcalf (Zeit. Phys. Chem. Hi, p. 1, 1905) on the 

 formation of films or crusts upon water by methyl violet, 

 peptone, &c. He could detect no difference in the films 

 when oxygen or hydrogen was in contact with them. A 

 similar behaviour is observed in the change of sign which 

 occurs when a bubble of hydrogen or oxygen is slowly 

 absorbed into a solution of thorium nitrate as described above. 

 When first introduced into the liquid the bubble is negatively 

 charged. As it diminishes in size it gradually becomes 

 neutral and then slowly takes on a positive charge. The 

 effects are exactly those that would be produced if the con- 

 centration of the salt were slowly increased, as shown in the 

 table above for thorium nitrate. It is natural to conclude 



