Page:Collected Physical Papers.djvu/103

Rh failed by turns, and it was impossible to find out which was really at fault.

From a series of experiments carried out to find the causes of the erratic behaviour of the receiver, I was led to suppose that the uncertainty in its response is probably due to the following:—


 * (1) Some of the particles of the coherer might be in too loose a contact against each other, whereas others might be jammed together, preventing proper response.
 * (2) The loss of sensibility might also be due to the fatigue produced on the contact surfaces by the prolonged action of radiation.
 * (3) Since the radiation was almost entirely absorbed by the outermost layer, the inner mass, which acted as a short circuit, was not merely useless but might introduce complications.

For these reasons I modified the receiver into a spiral-spring form. Fine metallic wires (generally steel, occasionally others, or a combination of different metals) were wound in narrow spirals and laid in a single layer on a groove cut in ebonite, so that the spirals could roll on a smooth surface. The spirals are prevented from falling by a glass slide in front. The ridges of the contiguous spirals made numerous and well-defined contacts, about one thousand in number. The useless conducting mass was thus abolished, and the resistance of the receiving circuit almost entirely concentrated at the sensitive contact surface exposed to radiation. If any change of resistance, however slight, took place at the sensitive layers, the galvanometer in circuit would show strong indications. The pressure throughout the mass Rh