The following is a paper by H. Aspden published in Journal of Electrostatics, v. 13, pp. 71-80 (1982).
Abstract: It is argued that precision measurements which rely on the use of conductive housings for electrical screening may well be subject to spurious error owing to charge induction within the housing. If the momentum transfer processes associated with the absorption of thermal radiation act selectively upon electrons, a residual charge may be induced on surfaces internal to the housing and held in place by radiation pressure. The effects of such action upon the measurement of G is discussed in relation to anomalies reported in such experiments. The phenomenon is supported by the existence of the charge on the Earth's surface. A method of testing for the presence of induced charge and eliminating its effects is indicated.
Commentary: The author's interest in this theme has developed as a peripheral question when the author wrote his 1972 book 'Modern Aether Science'. It was later, when the author was at a conference on precision measurements in Gaithersburg in U.S.A., organized by the National Bureau of Standards, that he became aware of anomalies in the certain measurements of the constant of gravitation G. These involved tests performed in Faraday cages at different levels in a deep mine, the test mass inside the cage being assumed to be grounded electrically. The author's theory of gravitation was not able to explain the variations of G that were reported and so the test itself warranted scrutiny. Later, when the author was at the University of Southampton he performed experiments which verified the charge induction explanation of the subject paper. These experimental findings were presented at a conference organized by the U.K. Institute of Physics and held in Oxford, but because the results challenged orthodox belief as to the effectiveness of the Faraday cage, the paper was not included in the reported proceedings.
Special Note: The following author's note was prepared shortly after the paper had issued and was communicated by the author to recipients of the offprints of the paper.
"My attention has been drawn to the early measurements of the effect of temperature upon G, for example by L. Southerns, [Proc. R. Soc. Lond., A78, pp. 392-403 (1906)] and P. E. Shaw & N. Davy [Proc. R. Soc. Lond., 102, pp. 46-47 (1922)]. The measured effects are far less than might seem to be indicated by my paper.
The reason for this is that these experiments involve heating the apparatus by processes other than those involving incident radiation and then making the measurements with the gravitating body radiating energy at a higher temperature but absorbing incident radiation from an environment still at room temperature.
It is only the radiation absorbed by the layer of atoms at the surface that results in the charge induction below the surface and, though displacing charge and perhaps being reflected from there, it does not account for any residual surface charge effect. If the body is heated there is transfer of heat to the surface atoms by conduction and any radiation from the layer of atoms at the surface is directed equally inwards and outwards and produces a balanced radiation reaction force on this surface layer and so induces no surface charge. Half the radiation goes inwards and asserts a radiation pressure on the atoms below the surface and so causes charge displacement but no residual surface charge induction.
Experiments of the kind suggested on page 78 of the author's paper aimed at verifying this phenomenon seem, therefore, warranted notwithstanding the early studies of temperature effects on G."