Electromotive Forces from ThermocJiemical Data. 219 



the layers near the metallic mercury, depositing crystalline 

 nitrate on it and greatly increasing the resistance. But by 

 allowing the current to pass for not more than 10 to 20 minutes 

 in one direction this was avoided, and the rise of temperature 

 which was observed corresponded very nearly with the heat 

 due to c^rt, and was independent of the direction in which 

 the current passed ; showing therefore no absorption of heat 

 with direct currents, as might be anticipated from its large 

 excess of E.M.F., viz. -335 volt. 



This result was so plainly at variance with what was to be 

 expected, that a solution of mercurous nitrate was electrolysed, 

 using a platinum anode ; but in this case also there was no 

 indication of more chemical work being performed by the 

 current than supplied by it, i. e. no absorption of heat was 

 detected, but an evolution corresponding closely with the 



c'^rt 

 amount calculated from the values of — j-. Hence the re- 

 sult of these experiments clearly pointed to the conclusion 

 that mercury, if it be anomalous in its E.M.F., is not so on 

 account of any transformation of sensible heat into electric 

 energy ; and I was therefore led to examine the thermo- 

 chemical data of mercury upon which the calculated E.M.F.s 

 have been founded. 



Thermochemical Data of Mercury. 



In the first place Julius Thomsen* appears to be the only 

 experimentalist who has made determinations of the heats 

 of formation of mercury salts, and since he depends upon 

 the heat of formation of mercurous nitrate for the ther- 

 mal values of almost all the other salts, if that were in error 

 all the values derived from it would evidently be equally 

 inaccurate. But the method he employed for the determina- 

 tion of the heat of [Hgg, 0, N2O5 Aq] does not appear to me 

 entirely free from objection. He determined it by reducing 

 the mercury from a solution of its nitrate by sulphur dioxide, 

 and determined the heat of neutralization of [Hg20,2HN03Aq] 

 by decomposing the nitrate with sodic hydrate ; finding, as 

 above stated, 47990 calories for the heat of formation of the 

 nitrate in aqueous solution. Since, however, mercurous 

 nitrate requires the presence of free nitric acid to maintain 

 it in solution, he used a solution containing free nitric acid to 

 an extent variously stated as 1*079 and 3'079 HNO3 per 

 molecule of Hgg (^03)2 ; as the latter number is mentioned 

 twice, probably the former is a misprint. But the exact 



* Thermochemische Untersuch. Band iii. S. 355. 



