Two independent types of contact electrification are discussed; the one requiring different surfaces, but not needing friction; the other requiring asymmetric rubbing, but not a difference in the surfaces. It is suggested that the combination of these two effects which occurs when different surfaces are rubbed together (usually asymmetrically) is a contributory factor to the notorious uncertainty of such experiments. It is shown that almost all rubbing processes involve local asymmetry, with the result that like surfaces rubbed together, though each neutral as a whole, are covered with a pattern of opposite charges. A similar effect accounts for the different triboelectric behaviour of surfaces of the same material but different degrees of roughness, and occurs as a complication in the rubbing together of different materials. The part played by asymmetric rubbing between two solids in observed transference of charges to the air is discussed.

Experiments are described in which all possible pairs of ten different materials are brought into contact by a method involving the minimum of rubbing. It is shown that the results, with very minor exceptions, are in accordance with the existence of a self-consistent electrostatic series.

The separation of charge by the asymmetric rubbing of like surfaces is discussed, and the hypothesis suggested that it is due to the thermal gradient across the interface, and that it is analogous to thermal diffusion and to the Thompson thermoelectric effect in metals. As yet little experimental evidence exists to test this hypothesis, but suggestions are made as to how it might be obtained.