Abstract
The authors review briefly earlier work leading to the idea that random potentials are important in discontinuous metals. They then describe experiments that yield direct information about the energetics of island charging in discontinuous metal films. Comparison with results of a theoretical analysis taking into account disorder potentials, the distribution of island sizes and the effect of temperature indicates that the islands in the films are subject to disorder potentials of order 10 mV which is comparable with the charging potentials. They assert that progress in understanding conduction in discontinuous metals requires development of new models in which disorder potentials are an essential feature.