CONTRIBUTIONS FROM THE CHEMICAL LABORATORY OF 



HARVARD COLLEGE. 



ON THE TEMPERATURE COEFFICIENT OF THE POTEN- 

 TIAL OF THE CALOMEL ELECTRODE, WITH SEVERAL 

 DIFFERENT SUPERNATANT ELECTROLYTES. 



By Theodoue William Richards. 



Presentea May 12, 1897. Received May 18, 1897. 



The well known Gibbs-Helmholtz formula for the temperature coeffi- 

 cient of a galvanic cell has been often tested,* with results which have 

 usually been confirmatory. In 1885, however, among a number of satis- 

 factory experiments, Gockel obtained a series of results t which he was 

 at a loss to explain. He measured the temperature coefficient of calomel 

 electrodes containing various supernatant electrolytes by setting up cells 

 of the general form Hg-HgCl-MCl,-HgCl-Hg, and maintaining the two 

 ends at different temperatures. His results were as follows, each of the 

 solutions being of half normal strength. 



Gockel's Results. 



Cell. Temperature Coefficient. 



Hg-HgCl-ZnC1.2-HgCl-Hg +0.000675 



Hg - HgCl - CdCl., - HgCl - Hg +0.000680 



Hg - HgCl - BaClo - HgCl - Hg +0.000440 



Hg - HgCl - SrCl, - HgCl - Hg +0.000287 



Hg - HgCl - NaCl - HgCl - Hg +0.000119 



These variations are not easily explicable. It is manifest that the usual 

 equation for the temperature coefficient of a galvanic cell involving uni- 

 valent ions — -JL = ^ — -jL, in which tt represents the potential, Tthe 

 a T T Cg 1 



absolute temperature, Q the total change of energy (the heat of the reaction), 



* Czapski, Wied. Ann., Vol. XXI. p. 239 (1884) ; Jahn, Wied. Ann., Vol. XXIV. 

 pp. 21, 491 (1886) ; Loven, Z. Phys. Chem., Vol. XX. p. 456; Bugarszky, Z. Anorg. 

 Chem., Vol. XIV. p. 145; and otliers. 



t Wied. Ann., Vol. XXIV. p. 618; table on p. 638. 



