52 Proceedings of the Royal Irish Academy. 



galvanometer G or the electrometer E with extra capacity 0, being in use 

 through key 1, according to the conductivity of the layer under examination. 

 Sometimes, also, the conductivity was obtained by observing the drop of 

 potential along a standard megohm. The E.WF. employed in battery B x was 

 usually two volts, but in later experiments larger voltages were used. The 

 plate P was insulated, and could be raised to a high potential by battery B 2 

 or earthed through key 2, the tinfoil terminals being to earth during the 

 charging and discharging of plate P. 



2. The first observations were made with graphite layers, and a typical 

 experiment is as follows : — The powder was rubbed on the paraffin surface 

 until, with a capacity of 10~ 5 farad, the electrometer charged at the rate of 60 

 scale divisions per minute, the E.M.F. of battery J5, being 2 volts. This 

 corresponded to a current of 3'3 x 10" 13 ampere, the electrometer having a 

 sensitiveness of 3000 divisions per volt. The resistance of the powder layer 

 was therefore 6 x 10 1 - ohms. The plate P was then charged to a high poten- 

 tial — ±00 volts in this experiment — and discharged, and the conductivity again 

 measured. The current was now 3 x 10~ s ampere, so that the conductivity 

 was 100,000 times its original values. Even larger relative increases of con- 

 ductivity can be obtained if the initial conductivity of the layer is very small. 



This remarkable increase of conductivity suggests the coherer effect. 

 The conductivity is not, however, always destroyed by tapping, as in the case 

 of the coherer. When the layer is formed of graphite on a paraffin surface, 

 even vigorous tapping does not reduce the conductivity ; but in the case of 

 some other powders — bronze metal, for example — very vigorous tapping does 

 destroy the conductivity. We shall see later other reasons for distinguishing 

 the effect from a simple coherer action. 



3. Several experiments have been carried out with a view to making clearer 

 the nature of the effect, and these we shall briefly describe. It may be 

 mentioned that the direction of the applied electric force between the powder 

 and the metal plate is immaterial. It should also be mentioned that 

 throughout these experiments care was taken not to include effects due to 

 residual charge, which in most cases were negligible in comparison with the 

 main effect. 



Other powders than graphite were tried, and the effect could be obtained 

 with any powder which gave slight conductivity when spread over the 

 paraffin in a very thin layer. One powder in particular, a bronze or Dutch 

 metal powder, was found to act very well, and was used for many of the 

 subsequent observations. With this powder greater relative increases of 

 conductivity were obtained than with the graphite. With this powder the 

 conductivity could be destroyed by sharp tapping. 



