294 MEASUREMENT OF CURRENTS. 



in which R is the resistance of the dielectric, and r the resistance 

 of the connecting wires. 



As the ratio is generally very small, we have sensibly 

 K. 



= E i-<r . 



The change may be considered instantaneous, if the quantity O 

 is infinitely small compared with the duration of the contact, which 

 ought to be so small that the principal current I has not had time to 

 undergo an appreciable change. It must finally be assumed that 

 the quantity of electricity removed from the principal circuit by the 

 electrometer or the condenser does not appreciably modify the 

 current between the points A and B. These conditions are usually 

 pretty easy to realise. 



The connections at A and B should be made and broken 

 simultaneously, unless the conditions of the experiment do not 

 allow of one of the points A being kept at constant potential 

 by connecting it to earth, for instance. It is sufficient then to 

 arrange the contact at the point B. 



The electrometer may still be replaced by a galvanometer. The 

 swing imparted to the needle will be proportional, other things being 

 equal, to the principal current I. 



907. The sudden opening of the current in the principal cir- 

 cuit S would produce, in a closed circuit S' of resistance R', an 

 induced current the quantity m of which, being proportional to the 

 coefficient M of mutual induction, would satisfy equation (541) 



From this would be derived a means of determining I by the 

 swing of a ballistic galvanometer situated in the circuit S'. If the 

 principal circuit S contains neither resistances nor capacities of 

 considerable magnitude, and the intensity I of the current is small, 

 the duration of the spark on breaking may be regarded as negligable. 



The time / at which a variable current is observed should be 

 counted from a starting-point which is in relation with the nature 

 of the phenomenon. In the case of induced currents, this origin 

 is naturally when the induction begins. The interval / will be de- 

 duced from the mechanical arrangements employed, either directly 

 or by the aid of an auxiliary current as in Pouillet's method (895). 



