This effect is very simply shown by constructing a small apparatus like a radiometer. Two covered glasses are attached to the end of a glass fibre about 6 cms. long, the surfaces lying in the same plane. The apparatus is free to rotate on a pivot. The two vanes are coated on alternate faces with radium bromide, and the whole apparatus contained within a glass receiver. If an electrified rod of ebonite or sealing wax is brought up close to the receiver, a rotation is communicated to the vane which increases as the pressure of the air is lowered to 5 or 6 cms. of mercury. By placing the apparatus between parallel plates connected with the terminals of a Wimshurst machine, a steady rotation is communicated to the vanes. The rotation is always in such a direction that the radium coated surface is repelled from the electrified body. This action was examined still further by attaching the vanes to the glass beam of a Coulomb’s balance. A metal sphere, which could be charged from without, was fixed facing the side coated with radium. A repulsion was always observed except when the charge was very strong and the vane near the sphere. If, however, the two vanes were connected by a light wire and a similar sphere placed exactly opposite the other, an attraction was observed if one sphere was charged, but a repulsion if both were charged. These effects were observed whether the vanes were of aluminium or glass. Joly found that the effect could not be explained by any direct action due to the movement of the ions in an electric field. The recoil, due to the expulsion of α particles from one side of the vane, is far too small to account for the movement observed.