needle. A theory of the quadrant electrometer, accounting for this action, has been recently given by G. W. Walker6. The effect appears to be due to the presence of the air space that necessarily exists between adjoining quadrants.
Suppose that it is required to measure with an electrometer the ionization current between two horizontal metal plates A and B (Fig. 14) on the lower of which some active material has been spread. If the saturation current is required, the insulated plate A is connected with one pole of a battery of sufficient E.M.F. to produce saturation, the other pole being connected to earth. The insulated plate B is connected with one pair of quadrants of the electrometer, the other pair being earthed. By means of a suitable key K, the plate B and the pair of quadrants connected with it may be either insulated or connected with earth. When a measurement is to be taken, the earth connection is broken. If the positive pole of the battery is connected with A, the plate B and the electrometer connections immediately begin to be charged positively, and the potential, if allowed, will steadily rise until it is very nearly equal to the potential of A. As soon as the potential of the electrometer system begins to rise, the electrometer needle commences to move at a uniform rate. Observations of the angular movement of the needle are made either by the telescope and scale or by the movement of the spot of light on a scale in the usual way. If the needle is damped so as to give a uniform motion over the scale, the rate of movement of the needle, i.e. the number of divisions of the scale passed over per second, may be taken as a measure of the current through the gas. The rate of movement is most simply obtained by observing with a stop-watch the time taken for the spot of light, after the motion has become steady, to pass over 100 divisions of the scale. As soon as the observation is made, the plate B is again connected with earth, and the electrometer needle returns to its original position.
In most experiments on radio-activity, only comparative measurements of saturation currents are required. If these measurements are to extend over weeks or months, as is sometimes the case, it is necessary to adopt some method of standardizing the electrometer from day to day, so as to correct for variation in its sensibility.