474 



Dr. G. Gore. Thermo-Electric Behaviour of [Dec. 11,. 



minute swing of the needle was obtained by repeated connexion 

 at proper intervals of time. B was now heated. About seven or 

 eight minutes was usually occupied in raising its temperature the 

 required number of degrees ; the deflections at the various tempera- 

 tures being noted. After an experiment had been made, the mercury 

 and solution of both vessels were mixed and cooled, and the mercury 

 filtered, before repeating the trial with the same liquid. 



Care was taken to select solutions which had the least amount of 

 chemical action upon the mercury ; those of salts of the heavy and 

 less positive metals were therefore not employed. ]STo liquid was used 

 which produced any visible film or dulness upon that metal when 

 heated with it. It was found necessary to heat each solution to 

 about 180° F., and keep it at that temperature some little time to 

 expel dissolved air, and then cool it previous to making experiments, 

 otherwise the results obtained in repetition experiments were not 

 quite uniform. In a previous research on the thermo-electric pro- 

 perties of liquids ("Proc. Roy. Soc," No. 188, 1878, pp. 513—543), 

 in which platinum electrodes were employed, such previous heating 

 was less necessary, probably because the thermo-electric currents 

 were then stronger.* After heating and cooling the liquid, it was 

 freely shaken with the mercury, and then filtered, previous to making 

 experiments with it. With every liquid, one or more repetition ex- 

 periments were made in order to obtain uniform results, and confirm 

 their reliability; the amount of difference of deflection of the needles 

 in such experiments did not exceed one degree. The galvanometer 

 employed was an astatic one, having two coils, each of 50 ohms re- 

 sistance, the two being connected and used as one of 100 ohms 

 resistance. 



I have ascertained by separate experiments of a different kind, that 

 mercury when sufficiently agitated with solutions, neutral to test 

 paper, of salts of the alkali metals, renders some of those liquids 

 feebly alkaline ; the** effect, however, is so slight, requires such 

 extensive and long-continued contact of the substances, that it 

 appears consistent with the view that chemical action is not the cause 

 of the currents in these thermo-electric experiments. 



The following are the results obtained with different liquids in the 

 above apparatus. 



Behaviour of different Solutions. 



Experiment No. 1. Ten ounces of water, and 2 \ grs. of potass! c 

 cyanide of a high degree of purity. Cold mercury was positive 2° 



* The greater strength of the currents with the platinum electrodes was partly 

 due to the circumstance that each platinum electrode offered a stirface of 71,680 sq, 

 millims. of contact with the liquid, or fourteen times greater than that of the above- 

 mentioned mercurial surface. The solutions employed were also much stronger. 



