Experimental Comparison of Coefficients of Induction. 95 



standard measure. Therefore, in making the coils, there is no 

 necessity to count the number of turns exactly, or to lay them 

 with the utmost accuracy: they may be wound in the ordinary 

 way, and then a hundred machines or more connected together, 

 with the main circuits in series and with the derived circuits 

 in series, and a current sent from a suitable source through 

 each series ; then, if there is one meter which has been standar- 

 dized by careful experiment, all the rest can be regulated, just 

 as clocks are, by screwing down the weights X X of those 

 that are going fast, or screwing up the weights of those that 

 are going slow. 



If in the foregoing paper any of the apparatus is not as 

 fully described as it might be, I must plead as an excuse an 

 endeavour to occupy a reasonable space with an account of 

 what is essentially one invention. 



XI. The Experimental Comparison of Coefficients of Induction. 

 By Heebekt Babfieud, B.Sc* 



OF all electrical investigations, the experimental compa- 

 rison of coefficients of induction is one of the most dif- 

 ficult to carry out with accuracy ; and it is doubtless most 

 desirable that we should be able to compare such quantities 

 much more accurately than has been hitherto possible. The 

 usual methods, as far as their general arrangement is con- 

 cerned, are in all probability not capable of improvement ; 

 but, as carried out, there is one modification applicable to all 

 of them, by which the sensitiveness can be vastly increased. 

 This modification is in the use made of the galvanometer. 

 Hitherto it has been usual to observe the throw of the needle 

 due to the passage of the quantity of electricity in one tran- 

 sient current ; but if, instead of doing this, we pass through 

 the galvanometer a number of such quantities in rapid succes- 

 sion and (after the manner of Siemens and others) observe the 

 permanent deflection of the needle, we shall find this far 

 greater than any throw attainable by even the most suitable 

 apparatus. 



In fact, the permanent deflection due to a certain number 

 of transient currents per second may be compared with the 

 throw of the needle due to one of them, in the following 

 manner: — Let 



H represent the magnetic force acting on the needle, 



Gr the galvanometer-constant, 



M the magnetic moment of the needle, 



t its time of vibration, 



* Communicated bv the Author. 

 12" 



