IT must be emphasized that the reality of a triboelectric series implies a single determinant property acting at both donor and acceptor surfaces. An `abundance effect'1, if real, may be a cause of minor anomalies, but it cannot be dominant since metals, which may be inserted in the series, contain incomparably more conduction electrons than do insulators. The Frenkel mechanism2 of thermal excitation alone is inadequate, since pyrolysis would occur, in the 100 µsec suggested by Frenkel, below 500° C, at which the population of conduction electrons might well be Only 100 per c.c. Since it is known that calendering may reduce the molecular weight of polymers, I suggest that free radicals, produced by mechanical stress, form a source of relatively high-energy electrons, easily excited to the conduction-level. The much-discussed charge transfer by non-rubbing contact remains improbable except between distant members of the series.