With a longer time of exposure to the emanation, the ratio of the increase after removal is much less marked. For a day’s exposure, the activity after removal begins at once to diminish. In this case, the increase of activity of the matter deposited in the last few hours does not compensate for the decrease of activity of the active matter as a whole, and consequently the activity at once commences to decay. This increase of activity with time explains the initial irregularity in the recovery curve, for the active matter deposited during the first few hours takes some time to reach its maximum activity, and the initial activity is, in consequence, smaller than would be expected from the equation. The increase of activity on a rod exposed for a short interval in the presence of the thorium emanation has been further investigated by Miss Brooks. The curve _C_ in Fig. 65 shows the variation with time of the activity of a brass rod exposed for 10 minutes in the emanation vessel filled with dust-free air. The excited activity after removal increased in the course of 3·7 hours to five times its initial value, and afterwards decayed at the normal rate. The dotted line curve _D_ represents the variation of activity to be expected if the activity decayed exponentially with the time. The explanation of this remarkable action is considered in detail in section 207. [Illustration: Fig. 65.] =181. Effect of dust on the distribution of excited activity.= Miss Brooks[274], working in the Cavendish Laboratory, observed that the excited activity due to the thorium emanation appeared in some cases on the anode in an electric field, and that the distribution of excited activity varied in an apparently capricious manner. This effect was finally traced to the presence of dust in the air of the emanation vessel. For example, with an exposure of 5 minutes the amount of excited activity to be observed on a rod depended on the time that the air had been allowed to remain undisturbed in the emanation vessel beforehand. The effect increased with the time of standing, and was a maximum after about 18 hours. The amount of excited activity obtained on the rod was then about 20 times as great as the amount observed for air freshly introduced. The activity of this rod did not increase after removal, but with fresh air, the excited activity, for an exposure of 5 minutes, increased to five or six times its initial value.