53 MEASUREMENT OF RESISTANCES IN ABSOLUTE VALUE. 



or making the circuit ; a commutator breaks the inducing circuit n 

 times in a second, and only closes the induced circuit by the gal- 

 vanometer during a single one of the variable phases of the inducing 

 current, making or breaking. A deflection a is thus obtained cor- 

 responding to the equation 



h 



i = nwi = tan a. 

 g 



By stopping the commutator, and putting the galvanometer in 

 the inducing current, we find 



k 



I = - tan 8 - 

 8 



from which is deduced 



, tan 8 



tana 



The time to be measured is thus that which flows between two 

 breaks. Graphical methods give this time with great accuracy. 



The method is thus reduced to the greatest simplicity as regards 

 the number of quantities to be measured ; but it introduces some 

 uncertainty, arising from the rapidity with which the breaks succeed. 

 It may be feared that, from the effects of polarization on the one 

 hand and of extra currents on the other, the intensity of the inducing 

 current on breaking does not differ from the intensity relative to the 

 permanent regime. It is, finally, possible that the break causes a 

 loss of part of the induced current. This latter source of error in 

 particular would have the effect of increasing the number found for 

 the resistance R, and therefore of diminishing the value of the unit. 



1120. METHOD OF DAMPING. This method was first used by 

 W. Weber in observing the oscillations of a magnetised bar in a 

 galvanometric frame, which was so good a conductor and so near 

 the bar as to produce a rapid damping. 



In calculating the resistance by the data furnished by experiment, 

 Weber did not bring in the term for the coefficient of self-induction, 

 which moreover is very small. 



Taking into account equations (845) 



HM *** 



