478 MEASUREMENT OF CAPACITY. DIELECTRICS. 



the same potential, and have no action on each other. It is clear 

 that, if the plate B is brought nearer, the induction will be increased 

 on this side, the leaf b will become positive, the leaf a negative, and 

 attraction will be observed. 



Putting the plates again in position, a dielectric plate of thick- 

 ness e is interposed between C and B. If the distance of the plates 

 is small in comparison with their dimensions, the plate introduced 



is equivalent (122) to a layer of metal of thickness e I i j , and 



produces the same effect as if the plate had been brought nearer 

 B by the same amount. An attraction of the gold leaves is observed, 

 and, measuring the quantity d by which the armature B must be 

 removed so as to restore equilibrium, we have 



d-e( i-- ) or 



e-d 



This arrangement has the drawback that the want of equilibrium, 

 on whatever side it takes place, always produces attraction of the 

 gold leaves. If the plates A and B are connected with the quad- 

 rants of an electrometer, the needle of which is connected with 

 the plate C, it is no longer necessary first to connect these plates 

 with the ground. The needle stays at zero if the induction is the 

 same on each side ; and, whatever be the sign of the electrification, 

 the deflection always shows on which side is the greatest induction. 



1068. Gordon* has used a more complete induction balance, 

 the general idea of which is due to Sir W. Thomson and Maxwell. 



In order to eliminate the influence of external bodies and to 

 realise as much as possible the case of unlimited parallel surfaces, 

 the plates A and B of Faraday which communicate with the 

 quadrants of an electrometer (Fig. 227) are placed in the intervals 

 of the three plates C, A l and B x of greater dimensions ; the external 

 plates A 1 and B t are connected with each other, while the central 

 plate is connected with the needle of the electrometer. 



When once equilibrium is established the needle should be sta- 

 tionary, whatever be the difference of potential V which is established 

 between the conductors C and AjBj. This is what would be the 

 case if, for instance, the system is quite symmetrical in respect 

 of the intermediate plate C. One of these extreme plates A : is 



* J. E. H. GORDON. Trans, Roy. Sac., 1879, p. 417. Physical Treatise on 

 Electricity, Vol. I., p. HO. 

 f 



