Fig. 104.
The conductor, of which the loss of charge was to be measured, was placed near the centre of the vessel A. It consisted of a narrow strip of metal with a gold-leaf attached. The strip of metal was fixed to the upper rod by means of a small sulphur bead. The upper rod was connected with a sulphur condenser with an Exner electroscope B attached to indicate its potential. The gold-leaf system was initially charged to the same potential as the upper rod and condenser by means of a fine steel wire which was caused to touch the gold-leaf system by the attraction of a magnet brought near it. The rate of movement of the gold-leaf was measured by means of a microscope provided with a micrometer eye-piece. By keeping the upper rod at a slightly higher potential than the gold-leaf system, it was ensured that the loss of charge of the gold-leaf system should not be due in any way to a conduction leakage across the sulphur bead.
The method employed by Wilson in these experiments is very certain and convenient when an extremely small rate of discharge is to be observed. In this respect the electroscope measures with certainty a rate of loss of charge much smaller than can be measured by a sensitive electrometer.
Both Geitel and Wilson found that the leakage of the insulated system in dust-free air was the same for a positive as for a negative charge, and was independent of the potential over a considerable range. The leakage was the same in the dark as in diffuse daylight. The independence of leakage of the potential is strong evidence that the loss of charge is due to a constant ionization of the air. When the electric field acting on the gas exceeds a certain value, all the ions are carried to the electrodes before recombination occurs. A saturation current is reached, and it will be independent of further increase of the electric field, provided, of course, a potential sufficiently high to cause a spark to pass is not applied.
C. T. R. Wilson has recently devised a striking experiment to show the presence of ions in dust-free air which is not exposed to any external ionizing agency. Two large metal plates are placed in a glass vessel connected with an expansion apparatus similar to that described in [section 34]. On expanding the air, the presence of the ions is shown by the appearance of a slight cloud between the plates. These condensation nuclei carry an electric charge, and are apparently similar in all respects to the ions produced in gases by X rays, or by the rays from active substances.
Wilson found that the loss of charge of the insulated system was independent of the locality. The rate of discharge was unaltered when the apparatus was placed in a deep tunnel, so that it did not appear that the loss of charge was due to an external radiation. From experiments already described, however ([section 279]), it is probable that about 30 per cent. of the rate of discharge observed was due to a very penetrating radiation. This experiment of Wilson’s indicates that the intensity of the penetrating radiation was the same in the tunnel as at the earth’s surface. Wilson found that the ionization of the air was about the same in a brass vessel as in one of glass, and came to the conclusion that the air was spontaneously ionized.
Using a brass vessel of volume about 471 c.c., Wilson determined the number of ions that must be produced in air per unit volume per second, in order to account for the loss of charge of the insulated system. The leakage system was found to have a capacity of about 1·1 electrostatic units, and lost its charge at the rate of 4·1 volts per hour for a potential of 210 volts, and 4·0 volts per hour for a potential of 120 volts. Taking the charge on an ion as 3·4 × 10-10 electrostatic units, this corresponds to a production of 26 ions per second.
Rutherford and Allan[[440]] repeated the results of Geitel and Wilson, using an electrometer method. The saturation current was observed between two concentric zinc cylinders of diameter 25·5 and 7·5 cms. respectively and length 154 cms. It was found that the saturation current could practically be obtained with a potential of a few volts. Saturation was however obtained with a lower voltage after the air had remained undisturbed in the cylinders for several days. This was probably due to the gradual settling of the dust originally present in the air.
Later observations of the number of ions produced in air in sealed vessels have been made by Patterson[[441]], Harms[[442]], and Cooke[[443]]. The results obtained by different observers are shown in the following table. The value of the charge on an ion is taken as 3·4 × 10-10 electrostatic units: