QUADRANT ELECTROMETER. 189 



one coating of a Leyden jar, the dielectric of which is the bell jar ; 

 the other coating, the tinfoil which covers part of the outside, and 

 which is connected with the earth. The uncovered portion of the 

 bell is usually so moist that we may suppose it also is at zero potential. 

 We may thus keep the needle at a very high potential, and the insides 

 of the bell jar are so well insulated that the loss of electricity is very 

 small. 



The two pairs of quadrants are supported by glass rods, which 

 the atmosphere of the jar keeps perfectly dry, and are connected to 

 the rods on the outside, which serve as electrodes. These rods are 

 insulated from the lid by means of ebonite. 



In order to charge the .jar, and keep the potential of the inside 

 constant, a rephnisher and a gauge of appropriate sensitiveness are 

 used. These two parts, which are also fixed to the cover of the 

 instrument, may be detached from the internal coating, so as to avoid 

 useless losses. 



The needle itself weighs 0-07 grammes, and with the mirror does 

 not exceed 0-12 grammes; it is supported by two independent silk 

 fibres, the distance of which may be varied at pleasure. In ordinary 

 conditions of sensitiveness a difference of potential of three to four 

 volts between the quadrants sends the image beyond the scale. 



The sensitiveness can be easily lessened without altering the 

 charge of the bottle or the distance of the fibres. When one of the 

 pairs of quadrants is insulated, one of the bodies whose potential is 

 to be determined being then connected with the case, the deflections 

 are ten to fifteen times less, while still proportional. We may go 

 further by means of a metal plate called the inductor^ placed above 

 one of the quadrants. When the four quadrants are connected in 

 pairs, the inductor has no action, whether it be connected to the case 

 or to the adjacent quadrant ; but if the quadrants are entirely or 

 partially insulated, and one of them is replaced by an inductor, a 

 series of different degrees of sensitiveness may be obtained. 



818. The construction of this electrometer seems to present 

 great difficulties. The quality of the glass is very important, for 

 the jar ought to retain its charge, without appreciable loss, at any 

 rate for a day. On the other hand, the fittings of ebonite alter 

 rapidly, and it may happen that the quadrants are not sufficiently 

 insulated for the investigation of bodies which are not in connection 

 with sources of electricity. Lastly, the deflections of the needle are 

 not exactly symmetrical for equal differences of potential of opposite 

 signs, which may in some cases be inconvenient, as in observations 

 on atmospheric electricity. 



