406 TRANSACTIONS OF SECTION A. 
excluding the light from the insulation the day-rate of leak was reduced 
to practically the same as the night-rate. : 
The theory that this change of rate of leak was due to a rapid and 
violent change in the rate of ionisation of the air in the vessel was early 
abandoned, for it was found that by merely causing the sunlight already 
shining into the vessel to fall directly on the sulphur insulation a great 
increase in the rate of leak was obtained, although there was no more 
sunlight entering the electroscope than formerly. The theory that sulphur 
is affected in much the same way under sunlight as zinc is under ultra- 
violet rays was also abandoned, for it was found that positive and negative 
charges leaked equally fast under similar conditions. An electroscope was 
constructed with a guard-ring about the sulphur insulation which was on 
the exterior of the vessel, and thus completely exposed to the light. 
In daylight, when the guard and the leaf carried charges opposite in 
sign, the leaf lost its charge, and in time acquired a charge similar in 
sign to that on the guard; but when both guard and leaf carried the same 
kind of charge, the charge on the leaf increased if initially less than 
that on the guard, but decreased if initially greater. On the other hand, 
at night either kind of charge leaked away from the leaf, even with the 
guard charged, which was undoubtedly due to the ionisation of the air in 
the vessel. 
The conclusion is that sulphur in the presence of light becomes to a 
slight degree a conductor of electricity, and the greater the intensity of the 
light, the greater the conductivity; and, further, that the leak due to 
ionisation is less than that due to the increase in the conductigity of 
sulphur, exposed even to ordinary daylight. It is necessary, therefore, that 
great care should be taken when measuring small electrical changes with . 
instruments having sulphur insulation to keep light from falling on the 
sulphur. 
A series of preliminary investigations in which amber and ebonite were 
used as insulators indicated a slight increase in the conductivity of ebonite 
in strong sunlight, but amber did not show any such effect. The importance 
of the effect of light on insulation was so great that the writer purposes 
investigating thoroughly with these and other materials in order to find 
its extent, and also, if possible, to discover the exact nature of the change 
produced by light on sulphur. 
9. The Charge on Gaseous Ions. 
By 'T. Franck and Dr. W. Wrstpuau. 
J. S. Townsend! has shown that by X-rays doubly charged ions are 
generated. Recent experiments made by the author ? lead to the conclusion 
that these doubly charged ions are only a small part of the total ionisation, 
and that generation in an electric field, contrary to Townsend’s view, has 
nothing to do with their formation. The coefficient of diffusion of the ions 
generated by X-rays, as measured by the method of Townsend,* may be 
reduced to half the value of singly charged positive ions by placing some 
wire gauze in the way of the ions, which causes a kind of fractionated 
diffusion, more of the easily diffusing singly charged ions being lost in 
the holes than of the double ones. If we assume that practically all singly 
charged ions have been removed by the wire gauze, the coefficient of 
diffusion of the double ions is thus found to be half that of the single ones. 
From the point of view of Rutherford’s cluster theory of the ions, this 
1 Proc. Roy. Soc. May 1908. 
2 Verh. Deutsch. Phys. Ges. 11, 146, 1909; 11, 276, 1909. 
° Phil. Trans. (A.) 193, 129, 1900, 
