Thermo-electric Action of Metals in Electrolytes. 289 



instance of this kind is silver and platinum in a strong solution of 

 potassic fluoride. The electromotive force of this combination was 

 found to be '0477 of a volt (see Table XII, p. 273). 



In order to test the inference " that at all times, whilst no heat or 

 cold is being applied to either of the junctions of the thermo-electric 

 couple, the molecular actions at the two points of union tend to 

 produce a current," I suspended, by means of a pure silver wire, 

 a sheet of pure silver, 2| inches wide and 3 inches deep, between two 

 sheets of platinum, 3 inches wide and 6 inches deep, in a glass cell, 

 containing a solution of pure potassic hydrate, 10 grains per ounce of 

 water, the platinum being immersed 4 inches, and the silver wholly, 

 in the liquid, and the two metals united outside by the wire. At the 

 outset, the platinum was positive to the silver, about 5 deflection 

 being produced of the needle of a galvanometer of 100 ohms resist- 

 ance, and after a few hours 1, and permanently half a degree ; it 

 then remained constant at that point, as shown by occasional testing 

 during about four months, when, by concentration of the liquid by eva- 

 poration, the current was feebly reversed. After diluting the liquid 

 to its original strength, the platinum was again positive half a degree, 

 and continued so during a further period of one year. In a similar 

 previous experiment with smaller plates similar results were obtained. 



To ascertain whether the platinum was corroded, two electrodes 

 one of sheet silver and one of platinum each about 1 inch wide and 

 3 inches deep, were partly immersed in a similar solution of the 

 hydrate, and a current from two Smee's cells passed through the 

 liquid, the platinum being the anode, during several weeks. The 

 plates were then reweighed ; the platinum had not perceptibly dimi- 

 nished in weight, and the silver had lost less than '01 grain. 



As an additional example, when gold and platinum produce an 

 electric current in a solution of neutral potassic sulphate, gold being 

 positive (see Table X, p. 266), the solution does not corrode them 

 (see p. 283), and the current is therefore not due to chemical corrosion. 



In each of such cases there are three points of heterogeneous con- 

 tact : 1st. Where the liquid A touches the metal B ; 2nd. Where it 

 touches the metal C ; and 3rd. Where the metals B and C touch 

 each other. At each of these points the substances are rendered 

 electro-polar by molecular action, and as a feeble current occurs the 

 polarities thus produced do not balance each other. As also a less 

 degree of electric potential is necessary to make a current pass 

 through a suitable electrolyte than to cause a refractory metal such 

 as gold or platinum to corrode, the current is produced without 

 corrosion of metal. This current at once begins to produce the usual 

 effect of raising obstacles to its own continuance, by causing polarisa- 

 tion at the electrodes, and becomes more feeble. As a portion of the 

 current, however, continues permanently, producing the usual effect 



