252 Dr. G. Gore, delations of Heat to I 'olUiic and 



immersed in the liquid, and then connected with the galvanometer. 

 Thermometers were supported in the liquid of the two cups, and the 

 temperatures of the two portions of liquid and pieces of metal were 

 in every case (except otherwise stated) 60 and 160 F. respectively. All 

 the metals and compounds employed were the purest I could obtain, 

 and liquids were avoided which would yield a metallic deposit upon 

 the metals. A series of metals and one liquid were examined at a 

 time. The most frequently recurring difficulty in the research was 

 polarisation. In all the cases the thermal, electrolytic, and polarisa- 

 tion effects, produced in the liquids at the electrodes by the passage 

 of the current, influence and modify the results by secondary 

 action. When also the current is sufficiently strong, the positive 

 metal is corroded. 



By these means the thermo-electric couples found with one liquid 

 and a series of metals were first classed into two groups, viz., those 

 in which the metal was thermo-electro-positive, and those in which it 

 was negative. To determine the series from the two groups, two 

 similar apparatuses, as already described, were simultaneously used, 

 one containing two pieces of one of the metals A, and the other, two 

 of another metal B, from the same thermo-electric group, and so 

 connected in single series, that the two currents opposed each other, 

 and the one alone of greatest E.A1.F. deflected the needle of the 

 galvanometer. The following table exhibits the results, and the 

 position of each metal in order of electro-motive power to the one 

 next below, and to that next above it in the series, was thus in every 

 case experimentally ascertained. The top metal in each column is 

 the most positive in the particular liquid ; and the metals above the 

 dividing lines are those from which the current proceeds from the 

 hot piece through the liquid to the cold one, and the ones below those 

 lines the reverse. 



The proportions of substances dissolved in each ounce of water to 

 form the solutions were as follows : of potassic cyanide (containing 

 93*52 per cent, of the actual substance), fluoride, bromide, nitrate, 

 iodide, chloride, sulphate and hydrate, sodic diphosphate, chloride, 

 sulphate, potash alum and ammonia alum, 10 grains each, potassic 

 carbonate 12 grains, anhydrous magnesic sulphate 4'8 grains, oxalic 

 acid 2 grains, dextro-tartaric acid 1 grain, chloric acid 6 minims, 

 hydrochloric, formic (sp. gr. 1'18), sulphuric, and nitric acids 

 1^ minims each. The liquids were selected of such a strength as 

 was considered to be the most suitable, especially as regards rate of 

 corrosion and strength of current. It was necessary in these experi- 

 ments to continually remember that the results vary considerably 

 with different specimens of the same metal, even when the metal is 

 very pure and the pieces of it are cut from different parts of the 

 same sheet. 



