the Constant of an Electro-dynamometer, 113 



If, as is desirable, the instrument is set up so that the term 

 depending on the earth's magnetism vanishes, we have simply 



and the value of icrk being obtainable from one reading, there 

 is no error due to variation of the current. 



Having thus obtained the value of iak, the coil S is sus- 

 pended within the coils F so that its axis of suspension coin- 

 cides with its previous axis of rotation; and by means of sights, 

 properly adjusted, it is brought into the same position relatively 

 to F as it occupied when making contact with M and N. 



If now an unknown current C be made to flow through the 

 instrument, and the electrical turning moment of the sus- 

 pended coil be opposed by an equal and opposite mechanical 

 moment T in the suspension-springs, we have 



T = C 2 i<Tk + KCak'; 



and on commutation of the current, 



T=C 2 ^-HC(7^ 

 from which 



where T is the moment due to unit twist on the springs, and 

 t the actual twist on them. 



Here, as before, the instrument can be set up so that TLCak' 

 vanishes, and a value of the current can be obtained in one 

 reading. 



The absolute value of T can, of course, be found accurately 

 by taking the time of vibration of a mass of known moment 

 of inertia when suspended in the place of the coil S; and thus 

 the constant, K, of the instrument can be determined. 



There are two points in the practical carrying out of this 

 method which require special attention — the prevention of 

 thermal and other extraneous currents in the circuit of the 

 revolving coil, and the keeping its speed uniform throughout 

 each revolution. Success in both these cases depends chiefly 

 on the manner of making contact between the revolving coil 

 and the points M and N. 



In my first experiments with the instrument I allowed the 

 two ends of the revolving coil to touch, in passing, two fixed 

 wires in connexion with M and N. After calculating a con- 

 stant, I compared the instrument with the very accurate tangent- 

 galvanometer belonging to University College, and found a 



Phil. Mag. S. 5. Vol. 17. No. 104. Feb. 1884. I 



