Abstract
A theoretical discussion is given of the sharp drop in the Knight shift observed by Sasaki, Ikehata and Kobayashi (1973) at the metal-insulator transition in Si: P. It is suggested that a drop of this kind is to be expected in solids, but not in liquids, at any metal-insulator transition of Anderson type due to disorder. Some discussion is also given of the negative magneto-resistance in this system.