Studies of isothermal formation cells by the electronic commutator method gave the following equations for
E°
of formation of silver halides. normalAgCl )(478°–731°C:E°=0.9081−0.280×10−3 )(t−455+0.110×l0−6 t−4552 normalAgBr )(447°–743°C:E°=0.7997−0.288×10−3 )(t−432+0.097×10−6 t−4322 normalAgI )(558°–730°C:E°=0.5720−0.227×l0−3 )(t−558+0.220×l0−6 t−5582 The E°'s and derived formation constants ΔG°, ΔH°, ΔS°, and ΔC°
p
are compared with values obtained by others from calorimetric data and by use of reversible formation cells. Nonisothermal data reveal that large errors in reversible cell measurements will result from small temperature changes produced at the cathode by adding halogen from outside the cell. Combination of the iso‐ and nonisothermal data gives the following values, in mV/deg, for the thermoelectric power of molten silver halide thermocells. normalAgCl )(500°–700°C:θnormalAg=−0.378 θCl2=−0.725+2.4×10−4 )(t−500 normalAgBr )(500°–700°C:θnormalAg=−0.417 θBr2=−0.776+3.5×10−4 )(t−500 normalAgI )(600°–700°C:θnormalAg=−0.432 θI2=−0.859 These values agree with direct determinations within the present rather large error limits. However, consistencies in the observed variations suggest that in formation cells the concentration gradients of solutes resulting from solution of silver and halogen at the electrodes may produce a significantly greater potential drop in the thermal gradient of a nonisothermal formation cell than in the corresponding thermocell, having a smaller gradient of only one solute.