a tendency towards reversal had already taken place after an exposure of about 9 seconds. In the third, response is feebler still. In the fourth, during illumination, the normal response is extremely weak, and lasts for only 3 seconds; there is then a reversal in the response which is fairly strong. On the stoppage of light the effect continues for some time. Hitherto the recovery commenced immediately on the cessation of light. After the fourth, the responses are of opposite signs. They also get feebler and feebler.
It will thus be seen that both electric radiation and light produce similar conductivity and electromotive variations. In both, two opposite effects are observed. The curves of effect in both are similar. Under the action of continued radiation both exhibit a limiting effect. Under too long continued action, both exhibit a tendency towards or an actual reversal.
In the next paper I shall adduce further evidence tending to show that these effects are due to molecular strain.
I have in the previous paper described the various interesting effects due to visible and electric radiation.
Different considerations were adduced which tended to show that these effects were due to molecular strain, produced in the substance by the action of radiation. The whole history of the change produced by radiation—both the direct and after-effects—was graphically recorded in the various curves given. It was supposed that the strain effect on a substance was attended with conductivity or electromotive variation. This supposition can be further verified by observing whether undoubted strain, which can be produced by mechanical means, does give rise to a conductivity or electromotive variation.
As regards the conductivity variation due to mechanical strain, it is well known that in the construction of standard resistance coils the effect of winding the wire on a spool is to produce a distinct variation of resistance, and that this strain effect can only be removed by annealing. The difference between the resistance of a substance when strained and after it is annealed is sometimes very considerable.
The effect of electric radiation is very great in changing the conductivity of a mass of discontinuous particles. It is to be borne in mind that the effect of electric radiation is only skin-deep. As the action is one of surface, the larger the surface the greater is the effect produced. We see that in loose particles the effective surface acces-
