WINTER'S ELECTRICAL MACHINE. 597 



in winter better than in summer, because the air is drier in 

 winter than in summer. 



In any case it is well before the experiments to carefully rub 

 the glass pillars which carry the conductor and the axis with a dry 

 cloth which has been a little warmed before a fire. 



The wooden board on which the machine is supported must be 

 fixed to the table by a screw-clamp so as to ensure its remaining firm 

 and steady while the handle is turned. The further the machine 

 stands from other objects, the better; everything not absolutely 

 required for the experiments should therefore be removed from the 

 neighbourhood, especially bodies with rough or angular surfaces ; 

 also burning candles, for burning or glowing bodies act like points, 

 their effect being even stronger. A lamp-flame within a glass 

 cylinder has, however, not much more disturbing effect than the 

 lamp itself would have without the flame. The electroscope should 

 be placed, if possible, a few metres from the machine, to. prevent its 

 indications being influenced by the inductive action of the machine. 



When the machine has been worked a short time, let 

 the conductor be touched with the proof -plane, and 

 some of the electricity of the conductor be conveyed by 

 the proof- plane to the electroscope. The leaves will 

 diverge, and if a rubbed glass rod be brought near it, 

 the divergence will increase. The electricity of the 

 conductor is hence of the same kind as that of the 

 glass rod, that is, it is positive. Now let the proof- 

 plane touch one of the rubbers, and be brought near to 

 the positively charged electroscope, the divergence 

 diminishes, hence the electricity of the rubbers is 

 negative. 



If the divergence, on first applying the proof-plane, is at once so 

 great as to render its increase impossible, the quantity of electricity 

 in the electroscope may be diminished by alternately touching it 

 and the finger with the proof- plane. 



To prevent error the negatively charged proof-plane must be 

 brought near the positively charged electroscope slowly ; for if it be 

 brought near at once so as to touch the electroscope it may happen 

 that the divergence increases instead of diminishing. This would 

 take place whenever the quantity of negative electricity in the proof- 



