The thermoelectric effect of pyrite from copper deposits in Japan is mainly discussed. Prof. Smith measured the thermoelectric potential of pyrite specimens from many localities and as a result, and deduced that there are some qualitative relations between the thermoelectric potential and the temperature at which they are formed. A. D. Mutch reported that the value for the temperature of formation obtained by the pyrite geothermometer, appears to be reasonable and in relations to geological problem, to vary in the right sense. The writer measured the thermopotential of pyrite from many localities in Japan by Smith's method and discussed. The conclusion is reached that these results differ somewhat from those mentioned by F. G. Smith and it seems to be difficult to expect qualitative difference in thermoelectric potential for indication of the temperature of ore deposition.