Phenomenological approach to "current noise"

There is no satisfactory theory of the noise additional to Johnson noise which is found when a steady current is passed through resistors made of any kind of semi-conducting material. Experiments by various workers have failed to show conclusively that it is an interface phenomenon. Rough estimates are given of the total fluctuation (all frequencies) and the number N of charge carriers present in three examples, and the relative fluctuations are shown to be much greater than 1/ N . The inverse-frequency spectrum cannot be a result of a particular pulse shape, but could possibly be due either to a complicated diffusion process or more probably to the law relating mean-square perturbation and time for the whole population of carriers which is not to be regarded as an equilibrium system.