Abstract
Attention is called here to the fact that polyvalent impurities of the silver period of the periodic table give rise to a smaller residual resistivity in both copper and silver than do polyvalent impurities belonging to the copper period. An explanation of this experimental result is presented. The explanation is based on a modification of the Friedel sum rule suggested by Harrison by which one may take lattice distortion into account. The calculated results reflect in every detail the qualitative features of the experimental data and, moreover, lead to better quantitative agreement between theory and experiment than previous calculations based on the free-electron model. The occurrence of a similar period dependence of the thermoelectric power of these alloy systems is suggested and is currently being investigated.