抄録
In the part I of this article, it has been shown that the thermoelectric power of monovalent metals at high temperatures is interpreted, at least semi-quantitatively, based on the spherical energy band model. However, it has also been shown that the model gives no possibility of the positive sign of the thermoelectric power of pure metals at low temperatures. Therefore, in this part II, the thermoelectric power at low temperatures is calculated for an arbitrarily anisotropic energy band. The results show that there is a possibility to give positive thermoelectric powder for pure metals at low temperatures if the shape of the energy surface is appropriate. The relation with the spherical band model used in I is discussed also.