required to produce Sparks in Air and other Gases. 109 



Adjustments and Corrections. 



The balance of the electrometer was carefully adjusted to 

 equilibrium before beginning the experiments. The guard- 

 ring and movable disk were made parallel to the lower plane 

 by means of a piece of plane parallel glass, which was moved 

 from place to place ; the lower plate being raised or lowered by- 

 three set-screws, until the glass just touched both surfaces in 

 its different positions. This adjustment, which served also to 

 give the zero-reading of the electrometer, was made several 

 times in order to avoid constant errors. 



A correction to the distance between the plates, which is 

 rendered necessary when the guard-ring and suspended disk 

 are not in the same plane, was avoided in this case by so ad- 

 justing the latter that in its zero position it was exactly in 

 the plane of the guard-ring, and prevented by a set-screw 

 under the balance-arm from moving further away from the 

 lower plate. The extent of its downward motion under the 

 action of the electrical forces need not be considered ; inas- 

 much as the quantity to be measured, for a given difference of 

 potential, is that distance between the plates at which the 

 suspended disk can just be moved. As the equilibrium is 

 unstable, a force which is sufficient to move the disk at all 

 will move it downward as far as the supports permit. 



The interval between the guard-ring and disk was 1 millim., 

 and a correction was applied to the area of the latter, as cal- 

 culated from the formula (Maxwell, ' Electricity and Mag- 

 netism,' i. p. 307, footnote) 



A=| w {k 2 + R 1 2 -(R 1 2 + K)- [j ^}, 



where Ri — R is the thickness of the annular space between 

 the suspended disk and guard-ring, and 



a=0-220635(R 1 -R). 



The weights used in the balance-pan varied from 0*2 to 5 grms., 

 a different weight being employed for every 20 or 25 readings. 

 To find the difference of potential or the electrostatic force 

 from observations on an absolute electrometer, we have only 

 to apply the formula 



Vx— V is the difference of potential to be measured; D is the 

 distance between the plates of the electrometer ; A is the 

 area (corrected) of the suspended disk ; w the weight in the 



