According to the recent researches of Mr. Townsend, the phenomenon is more complex when the pressure of the gas is low. At first the current appears to approach to a constant limiting value with increasing difference of potential; but after a certain point has been reached, the current begins again to increase with the field, and with very great rapidity. Mr. Townsend ascribes this increase to a new ionisation produced by the ions themselves when, under the action of the electric field, they acquire a velocity such that a molecule of gas encountering one of them becomes broken down into its constituent ions. A strong electric field and a low pressure are favourable to the production of this ionisation by ions already present, and, as soon as the action is set up, the intensity of the current increases uniformly with the field between the plates. The limiting current could, therefore, only be obtained under conditions of ionisation of which the intensity does not exceed a certain value, and in such a manner that saturation corresponds to fields in which, from multiplicity of ions, ionisation can no longer take place. This condition has occurred in my experiments.
The order of magnitude of the saturation currents obtained with uranium compounds is 10^{ –11} ampères for a condenser in which the plates have a diameter of 8 c.m., and are at a distance of 3 c.m. Thorium compounds give rise to currents of the same order of magnitude, and the activity of the oxides of uranium and thorium is very similar.
_Radio-activity of the Compounds of Uranium and Thorium._
The following are the figures I obtained with different uranium compounds. I have represented the intensity of the current in ampères by the letter _i_:—
Metallic uranium (containing a little carbon) 2·3 Black oxide of uranium, U_{ 2}O_{ 5} 2·6 Green oxide of uranium, U_{ 3}O_{ 4} 1·8 Hydrated uranic acid 0·6 Uranate of sodium 1·2 Uranate of potassium 1·2 Uranate of ammonium 1·3 Uranium sulphate 0·7 Sulphate of uranium and potassium 0·7 Nitrate of uranium 0·7 Phosphate of copper and uranium 0·9 Oxysulphide of uranium 1·2
The thickness of the layer of the uranium compound used has little effect, provided that the layer is uniform. The following illustrate this point:—