The Thermocouple Revisited: The Thomson Effect

Abstract

Abstract Measurements of Thomson (later Lord Kelvin) coefficients are made in an evacuated region on a silver sample with a novel transducer in temperature intervals T A ± 20 K {T_{A}}\pm 20\hspace{0.1667em}\text{K} , where the T A {T_{A}} are very carefully controlled ambient temperatures. This is the first systematic examination of Thomson coefficients in these temperature intervals. The Thomson coefficients when plotted against T, the temperature of measurement, are found to be discontinuous precisely at T A {T_{A}} . When the Thomson coefficients are multiplied by a transformation involving T and T A {T_{A}} , a linear curve in T results. Examinations here of measurements of Thomson coefficients produced by others show multiple values at T in some cases and odd behavior in other cases. Multiplying the results of others by the transformation discovered here almost always produces linear curves. The conclusions are the following. (1) Thomson’s explicit assumption that the Thomson effect involves no energy exchange with surroundings was wrong. (2) Any non-equilibrium thermodynamic approach to deriving the Thomson effect must take into account the energy exchange with the surroundings and consequently must be made in three dimensions. (3) From the work here and that of others, the energy exchange with the environment is probably mostly thermal radiation.