Page:Collected Physical Papers.djvu/204

184 electric waves. The second difficulty is, however, far more serious for owing to the diffuse action of the comparatively long electric waves it is impossible to shield one plate while exposing the other. If both the plates are equally acted upon, there would then be no electromotive variation. It was only after the conclusion of another line of investigation on the electromotive variation produced by mechanical stimulus that a clue was obtained to overcome the difficulty. I then learnt that the effects of the same stimulus on two pieces of the same metal, forming a voltaic element, are not the same if the molecular conditions of the two are different. Under such a condition a P. D. exists between the two, and stimulation of both causes a variation of the existing electromotive force.

I therefore expected to detect the effect of electric radiation by an induced variation of the original electromotive force. And if the effects are at all parallel to those found in the conductivity variation method (as diminution or increase of resistance) the corresponding effects might be observed by a diminution or an increase of the existing electromotive force.

In carrying out experiments to verify the above suppositions, I found my anticipations to be justified. I at first made a cell by taking two varieties of silver. A piece of cotton wool moistened with amylic alcohol was placed in a glass tube. Ag and Ag′ were placed on opposite sides of the moistened cotton, thus forming a voltaic element, the amylic alcohol acting as the electrolyte. Two electrodes compressed the powder, till a current was observed to flow. As in