328 



indication of too great a resistance to conduction (or "frictional" 

 resistance, as I have elsewhere called it*), being a true indication of 

 resistance or reaction of inertia to the commencement of the current 

 in the electro -magnetically -loaded branch j*. In such cases it is im- 

 possible, either by electrodynamic balance or in any other way, to 

 obtain a measurement of resistance without keeping the battery ap- 

 plied for the few seconds required to produce sensibly its final 

 strength of current undiminished by inductive reaction, over and 

 above the time required to get an indication from the galvanometer. 

 But, as already remarked, in all ordinary cases, the inductive reaction 

 becomes insensible after a very small fraction of a second, and the 

 operator may press the double key home to its second contact almost 

 as sharply as he pleases. With such a galvanometer as I have de- 

 scribed, he need not hold it down for more than 7 of a second (the 

 time of the simple vibration of the needle ) to test the balance of 

 resistances. The order of procedure will therefore generally be this : 

 The operator will first strike the key sharply, allowing it to rise 

 again instantly, adjust resistances in the balance-circuit according to 

 the indication of the galvanometer ; strike the key sharply again, 

 readjust resistances ; and so on, until the balance is nearly attained. 

 He will go on repeating the process, but holding the key down 

 rather longer each time. At the last he will press the key gently 

 down, hold it pressed firmly for something less than a second of 

 time, and let it rise again ; and if the spot of light reflected from the 

 mirror of the galvanometer does not move sensibly, the resistances 

 are as accurately balanced as he can get them. 



* " Dynamical Theory of Heat, Part VI., Thermo-electric Currents," Trans- 

 actions of the Boyal Society of Edinburgh, 1854 ; and Phil. Mag. 1856. 



t It is probable that a Wheatstone's balance, perfectly adjusted for equili- 

 brium of resistances to conduction, and used with the galvanometer circuit con- 

 stantly made, so as to show the whole effect of the inductive impulse, may afford 

 the best means for making accurate metrical investigations on electro-magnetic 

 induction, and especially for determining " electrodynamic capacities " in abso- 

 lute measure. 



J The mirror galvanometers commonly used in Germany have all much 

 longer periods (ten or twenty times as long in many cases) for the vibration of 

 their needles, and want proportionately longer contacts to obtain full advantage 

 of their sensibility, in each case a contact during a time equal to that of the 

 vibration of the needle one way being required for this purpose. 



