686 [June 18, 



the object to be tested. This projecting arm is surrounded by a 

 hollow cylinder of pumice soaked in sulphuric acid, for the purpose 

 of drying the air which enters the jar, and thus preserving insula- 

 tion. The insulation thus obtained is so good that the jar loses, on 

 the average, only 2 or 3 per cent, per diem. 



5. The force of repulsion between the needle and the repelling 

 plates depends not on the potential of the jar absolutely, but on the 

 difference between this and the potential of the cage, being for any 

 given position of the needle proportional to the square of the differ- 

 ence of potentials. In taking an observation, the force of repulsion 

 is ascertained by applying torsion to the glass fibre until the near 

 end of the needle is brought into a line between two sights, one of 

 which is on the plate of glass which covers the jar, and the other on 

 the bottom of the cage. The amount of torsion applied is read off, 

 and is assumed to measure the force of repulsion in the sighted posi- 

 tion of the needle ; hence the square root of the number of degrees 

 of torsion measures the difference between the potential of the inner 

 coating of the jar and the potential of the cage (the latter being the 

 same as that of the body tested) in terms of a unit which is constant 

 for any one electrometer. In accordance with a convention which 

 has been adopted by Professor Thomson, 1 have always multiplied 

 the degrees of torsion by 10 before extracting the square root. Thus, 

 if E denote the number of degrees of torsion required to bring the 

 needle to the sights when the cage is connected with the earth, and 

 A the number required when the cage is connected with a conductor 

 to be tested, the potential of the inner coating of the jar will be 

 VlO E, and the potential of the conductor tested will be VlO A; 

 hence the potential of the body tested will be VlOE VlO A. 

 It is by this last formula that the numbers under the heading 

 " Electricity of Air " in my tabulated results have been calculated. 

 If A is greater than E, the formula still holds, and the negative sign 

 indicates that the electricity of the conductor tested is of opposite 

 kind to that with which the jar is charged. E is called the earth- 

 reading, and A the air-reading. 



6. The portable electrometer, which is more convenient for use 

 when the electricity of the air is very strong, also contains a needle 

 and repelling plates, surrounded by a cage which is insulated from 

 them, the whole being enclosed in a Ley den jar; but the needle, 



