the Thermo-Electric Properties of Liquids. 



533 



deflection was 18°. A film of brown oxide was fonnd on that electrode 

 only which had been heated, and it was probably this which made that 

 electrode first neutral and then negative. (Compare Nos. 77, 94, and 

 160.) 



No. 162. A deep blue solution of sulphate of chromium, probably 

 containing much free sulphuric acid. Cold platinum positive 7J° at 

 180° F,, and 5^° at 200° F. On cooling the liquid the needles returned 

 to zero. 



No. 163. An aqueous solution of iodide of cadmium. Hot platinum 

 was feebly positive 1° at 200° F. 



No. 164. A nearly saturated aqueous solution of pure carbonate of 

 sodium. Copper ribbons, 30 inches long and inch wide. Hot 

 copper positive 70J at 200° F. (Compare No. 6.) 



No. 165. The same solution. Iron ribbons, 14 inches long and 

 11 inch wide. Hot iron positive 70° at 200° F. No signs of corrosion 

 of the iron. Remarks. A cheap liquid thermo-electric battery might 

 perhaps be formed by means of this combination. It is well known 

 that solutions of alkalies protect iron from corrosion, and lime water 

 is used for that purpose. 



No. 166. The solution of sodic hydrate of No. 142. Iron ribbons. 

 Hot iron positive 751° at 200° F. (Compare Nos. 87, 110, and 142.) 



No. 167. The solution of potassic cyanide of No. 109. Iron ribbons. 

 Hot iron positive 44° at 200° F. (Compare No. 109 and 145.) 

 Remarks. As an illustration of the absence of corrosive action of this 

 liquid upon iron, I took two thin wires of equal dimensions and 

 perfectly clean and bright, one of iron and the other of pure gold, 

 twisted them together, and immersed them in a solution of potassic 

 cyanide in a closed bottle kept at ordinary atmospheric temperature. 

 After a few weeks the gold had entirely dissolved and disappeared, but 

 the iron remained unaltered in appearance. I also found iron to be 

 electro-negative to gold in such a solution. Iron employed also as an 

 anode in such a liquid offers a much greater degree of resistance 

 than gold to the passage of an electric current. 



Remarks and conclusions. A review of the results obtained with the 

 improved apparatus confirms, in nearly all cases, the general conclu- 

 sions (see p. 528) arrived at by means of experiments with the cylinder 

 one. The most conspicuous exceptions yet met with have been obtained 

 with a solution of sulphate of manganese (No. 148), chrome alum 

 (No. 156)> sulphate of nickel (No. 155), and argentic nitrate (No. 

 J 47), probably also of sulphate of iron (No. 159); cupric sulphate also 

 (Nos. 15 and 113) appears to be a feeble exception. In some acid 

 solutions, also, hot platinum is feebly positive, for instance, dilute 

 selenious acid (No. 157), an alcoholic solution of chloride of chromium 

 (No. 129), and an aqueous one of ammonic alum (Nos. 96 and 136). 

 No instance, however, has yet been met with in which, in the absence 



