52 STATIC ELECTRICITY 



strain is not then altogether normal to the surface, and the elect riti- 

 cation moves about until a distribution is attained such that the 

 strain is normal to the surface. Then there is equilibrium, for, 

 the conductor being surrounded by an insulator, the electrification 

 can move no further. In the case of a conducting circuit ram 

 a current produced, we may suppose, by a voltaic cell, equilibrium 

 does not exist, and the strain has a component along the surface 

 of the wire from the copper terminal towards the /.IMC terminal, 

 and therefore either positive electrification moves with it or negati \ c 

 electrification moves against it, this motion of electrification being 

 one of the chief features of the current. While the motion goes on 

 there is also strain in the conductor itself, but such strain can only 

 be maintained by perpetual renewal, the energy for the renewal 

 being supplied by the cell. 



The magnitude of electric strain just outside a con- 

 ductor. Usually we are practically concerned with the magnitude 

 of the strain only where it ceases at the MII 

 We shall take the amount of charge gathering per unit urea 

 the conductor as the measure of the strain t li- 

 on a given conductor there is at one time twin- the charge per 

 unit area that there is at another time, the strain jiM out-id- tin- 

 conductor in the former case is twice as much as in the latti r.* 



Definition of surface density. The charge pn unit 

 area on a conducting surface is termed the surface density. Tin- 

 surface density, then, may be taken as iiiea^urin^ t M in the 

 insulator just outside the conductor. If we wish to men 

 strain at a point not close to a conduct MIL bring a 

 small conductor there and note the charge gathering per unit area 

 on one side or the other. But in general the introduction of a 

 conductor into a system alters the strain in the part of tin 

 immediately around it, so that the charge }> r unit area docs not 

 in general measure the original strain, though it is probably | 

 portionate to it. But there is one case in which the introdm -tion 

 of a conductor does not appreciably affect the strain in it- 

 bourhood, viz. when the conductor is a small exceedingly thin 

 conducting plane held perpendicular to the direction of strain, and 

 this case may be more or less nearly approached in practice l>\ the 

 proof plane. By using a pair of exceedingly thin equal proof 

 planes of known area we may measure, at least in imagination, tin- 

 strain at a point by holding the planes in contact at the point and 

 normal to the strain, then separating them and measuring tin- 

 charge of either. The strain equals the charge per unit area. U 

 shall therefore take as a definition of the magnitude of strain tin- 

 following : 



Measurement of strain at any point. We measure the 



* In taking this definition, it may be noted that the analojry with elastic 

 strain fails, for elastic strain has zero dimensions, while electric traio has 

 dimensions charge /length*. 



