Abstract
Attention is drawn to the fact that the environment of an ion, which changes on alloying, has a much greater effect on the form factors of the ion than has been hitherto considered. The importance of taking a proper account of this effect is demonstrated by carrying out a fresh calculation of the impurity form factors based on the Harrison pseudopotential and comparing the results with those of Harrison, where this effect has been considered only in a limited way. The modifications are found to change Harrison's residual resistivities by as much as a factor of 4 or more. The basic inadequacy of Harrison's treatment having thus been removed, it is found that, on incorporating a few further improvements (nonlocal screening, effective mass, etc.), the calculations for several pairs of metals agree with experiment satisfactorily.