376 Prof. Owens on Thorium Radiation. 



A similar phenomenon was observed with thorium radiation. 

 Placing over a thick layer of thorium oxide a layer of paper 

 and on the paper a layer of aluminium foil, a certain con- 

 duction-current was obtained ; on reversing the layers the 

 current was reduced to nearly one half. A number of other 

 combinations were tried which also showed the effect in a 

 marked manner. 



6. Effect of Suspended Particles in the Path of the Con- 

 duction-Current. — A thick layer of thorium oxide covered 

 with one layer of aluminium foil was placed on the lower 

 plate in fig. 1, and the conduction-current measured when it 

 had become constant. 



The enclosing box was then iilled with tobacco-smoke and 

 the current again measured, and found to be about one 

 quarter of its former value. A similar result was obtained 

 when the sulphate was used instead of the oxide. As the 

 smoke gradually settled or disappeared, the current steadily 

 rose to its first value. The ions in their passage between 

 the plates giving up their charges to the smoke particles 

 encountered would explain the observed diminution of current. 



7. Variation of Conduction- Current ivith Pressure of Gas. 

 — The relation of conduction-current to pressure has been 

 studied for Rontgen rays by Perrin *, and for uranium 

 radiation by Becquerel f> de Smolan and Beattie t, and 

 Rutherford §. 



In general, the results show that the relation varies with 

 the potential gradient between the plates and the kind and 

 density of gas in which the radiations are absorbed. The 

 current at any particular distance from the active material 

 depends upon the number of ions produced in the gas up to 

 that point, the number which have recombined in the same 

 distance, and the electromotive force acting. 



If the volume ionization were uniform and so small as not 

 to appreciably affect the electrostatic intensity between the 

 plates, then, as the motion of translation of the ion under the 

 action of the electromotive force is small as compared with its 

 motion of agitation, the current should vary directly as the 

 potential gradient up to a point where the ionic velocity is 

 such that practically all the ions travel from plate to plate 

 before recombination takes place. After this the current 

 should remain constant. 



But if the volume ionization is not uniform, or is sufficient to 



* Comptes Hendus, cxxiii. p. 878. 

 t Ibid. p. 438(1897). 

 X Phil. Mag. xliii. p. 418 (1897). 

 § Ibid. Jan. 1899. 



