534 



Mr. G. Gore on 



of chemical change, hot platinum is strongly positive in a strongly 

 acid solution, or strongly negative in a very alkaline one ; the nearest 

 approach to this was in the case of aqueous selenious acid. A law 

 without an apparent exception is an unusual phenomenon, and, as a 

 single real exception will invalidate a general law, if the above ex- 

 ceptions are real ones, they prove that the theory that acids are 

 thermo-electro-positive, and alkalies thermo-electro-negative, is not an 

 exact one. 



Some strongly acid solutions give only feeble currents ; for instance, 

 acid sulphate of potassium (Nos. 48 and 128), and some strongly 

 alkaline ones behave similarly, except when persistently heated ; for 

 example, potassic hydrate (No. 112), sodic hydrate (No. 142), and 

 carbonate of rubidium (No. 149). On the other hand, some neutral 

 solutions yield currents stronger than the chemical theory would lead 

 us to expect ; for instance, chloride of potassium (Nos. 66 and 130). 



One rather common effect observed has been a very great increase 

 in the magnitude of the current by continuance of the heat without 

 increased rise of temperature, as if molecular inertia or. some other 

 slowly yielding resistance had to be overcome. This occurred in 

 nearly all solutions which were strongly alkaline, and was at one time 

 suspected by me to be wholly due to the action of the liquid upon the 

 washers ; it occurred freely, however, in experiments with the bent 

 tube apparatus in which no washers were employed. The following 

 are examples of such an effect: — with carbonate of sodium (Nos. 39, 

 40, and 41), carbonate of potassium (Nos. 70, 71, 72, and 121), caustic 

 potash (Nos. 108 and 112), iodide of potassium (No. 132), sodic hypo- 

 sulphite (No. 135), cyanide of potassium (No. 145), carbonate of 

 rubidium (No. 149), fluoride of potassium (No 151), and fluoride of 

 rubidium (No. 153). It also took place, though less frequently and to 

 a less degree, in liquids of acid reaction ; for instance, an acidified 

 solution of acid chromate of potassium (No. 38), iodate of potassium 

 (No. 131), an acidified solution of cupric nitrate (No. 138), bromide 

 of cobalt (No. 140), argentic nitrate (No. 147), and sulphate of man- 

 ganese (No. 148). In some other cases the quantity of the current 

 diminished by continuance of heat. 



The whole of the results obtained with this apparatus further con- 

 firm the conclusion obtained from the results of the experiments made 

 with the cylinder one, viz., that, provided chemical action is not present, 

 it is the nature of the liquid, and not so much that of the metal of the 

 electrodes, which determines the existence and direction of the cur- 

 rent.* If also we class under the heads of their respective acids all the 

 salts employed in the experiments, and then class them similarly under 

 their respective bases, we find that both the direction and the magni- 



* No. 109 is not a real exception to this statement. Compare Nos. 109, 145, and 

 167. 



