354 Prof. J. J. Thomson on the 



changes in the gas by which it is surrounded or in substances 

 dissolved in it which give rise to the electrification. As, 

 however, the quantity of substance dissolved in the gas is 

 increased the mixture becomes more nearly chemically satu- 

 rated or neutral, and more incapable of producing the requisite 

 chemical change in the surrounding gas, or of putting the 

 substance dissolved in it into a state in which it is chemically 

 active. At a certain stage the diminution in the chemical 

 activity of the water produced by the addition of more of the 

 substance, more than counterbalances the effect due to the in- 

 crease in the number of molecules of the substance. When 

 this stage is reached, any increase in the strength of the 

 solution will diminish the electrical effect. We may put this 

 reasoning in a mathematical form : Suppose that A is the 

 measure of the electric effect due to the action of the dis- 

 tilled water on the air, and that when m gramme molecules of 

 a salt are dissolved in a litre of water, the mixture becomes 

 more nearly chemically neutral, and its power of originating 

 the chemical change which produces the electrical double layer 

 is enfeebled, so that the electric effect is now only equal to 

 Ae~^ m , where j3 is some constant. Suppose, too, that the dis- 

 tilled water could put a fraction B of the molecules of the salt 

 into the condition in which they are able to produce the 

 chemical changes which lead to electrification, but that the 

 solution is only capable of putting the fraction Be"?™ into 

 this state. Suppose, too, that each molecule in this state 

 produces an electric effect represented by C/B. On these 

 suppositions the electric effect of the drops, which is the sum 

 of the effects of the water and the salt, will be proportional to 



i Ae-^ + Cme-v™. 



If we represent this relation by a curve in which the ordinate 

 represents the electric effect, the abscissa the strength of the 

 solution, we get the curves of the type a, /3, y according as 

 C is of the same sign as A, of the opposite sign, or zero (fig. 3). 



These curves represent the general behaviour of the sub- 

 stances I have examined. Curves of the type a represent 

 the behaviour of solutions of phenol, eosine, fluorescene ; 

 curves of the type /5 that of solutions of potassium perman- 

 ganate, chromium trioxide, hydrogen peroxide, rosaniline, and 

 methyl violet ; and curves of the type 7 that of solutions of 

 zinc chloride, hydrochloric and hydriodic acids. 



The measurement of the electrification developed by the 

 drops does not enable us to find the potential difference 

 between the drop and the surrounding gas ; it only shows 

 that this difference is finite and enables us to determine its 



