the insulators used at various points should be made as small as possible, in order to avoid disturbances due to their electrification. In damp climates, paraffin, amber, or sulphur insulates better than ebonite. The objection to paraffin as an insulator for sensitive electrometers lies in the difficulty of getting entirely rid of any electrification on its surface. When paraffin has been once charged, the residual charge, after diselectrifying it with a flame, continues to leak out for a long interval. All insulators should be diselectrified by means of a spirit-lamp or still better by leaving some uranium near them. Care should be taken not to touch the insulation when once diselectrified.
In accurate work it is advisable to avoid the use of gas jets or Bunsen flames in the neighbourhood of the electrometer, as the flame gases are strongly ionized and take some time to lose their conductivity. If radio-active substances are present in the room, it is necessary to enclose the wires leading to the electrometer in fairly narrow tubes, connected with earth. If this is not done, it will be found that the needle does not move at a constant rate, but rapidly approaches a steady deflection where the rate of loss of charge of the electrometer and connections, due to the ionization of the air around them, is balanced by the current to be measured. This precaution must always be taken when observations are made on the very penetrating rays from active substances. These rays readily pass through ordinary screens, and ionize the air around the electrometer and connecting wires. For this reason it is impossible to make accurate measurements of small currents in a room which is used for the preparation of radio-active material. In course of time the walls of the room become radio-active owing to the dissemination of dust and the action of the radio-active emanations9.
63. Electrometer key. For work with electrometers of high sensibility, a special key is necessary to make and break from a distance the connection of the quadrants with earth in order to avoid electrostatic disturbances at the moment the current is to be measured. The simple key shown in Fig. 16 has been found very satisfactory for this purpose. A small brass rod BM, to which a string is attached, can be moved vertically up and down in a brass tube A, which is rigidly attached to a bent metal support connected with earth. When the string is released, this rod makes contact with the mercury M, which is placed in a small metal vessel resting on a block of ebonite P. The electrometer and testing vessel are connected with the mercury. When the string is pulled, the rod BM is removed from the mercury and the earth connection of the electrometer system is broken. On release of the string, the rod BM falls and the electrometer is again earthed. By means of this key, which may be operated at any distance from the electrometer, the earth connection may be made and broken at definite intervals without any appreciable disturbance of the needle.