Magnetic Hysteresis loss in Straight Iron Strips. 27 1 



The above-mentioned law as to the distribution of the 

 induction enables us, then, to assign the corresponding value 

 of the maximum induction density. 



Unless this can be done, the mere mean hysteresis loss mea- 

 surement is no practical use *. 



The operation of making the mean hysteresis measurement 

 is effected by a sensitive mirror bifilar electrodynamometer. 



This was constructed in the following way : — 



The fixed or series-coil consists of 90 turns of No. 1 6 S.W.G. 

 cotton-covered copper wire wound in live layers on a rect- 

 angular block of wood. The movable coil, or shunt-coil, 

 consists of a rectangular coil of 20 turns of No. 24 S.W.G. 

 wire, the turns being fastened together with silk and shellac. 

 This movable coil embraces, without touching, the fixed coil, 

 as in a Siemens dynamometer. The movable coil is suspended 

 by a bifilar suspension consisting of two silver wires *002 of 

 an inch in diameter and 37 centim. long, fixed about 3 millim. 

 apart. These wires are attached to the ends of the suspended 

 coil, and serve to convey the shunt-current in and out of the 

 coil. The movable coil carries a concave silver mirror by 

 means of which a sharp image of an incandescent lamp- 

 filament is formed upon a scale at a fixed distance of 100 

 centim. from the mirror. A torsion-head carries the upper 

 ends of the bifilar suspension, and serves to adjust the zero. 



In series with the shunt-coil is a high non-inductive 

 resistance of 1000 ohms. 



The resistance of the thick-wire fixed coil of the wattmeter 

 is *133 of an ohm, and the resistance of the shunt-coil and 

 suspension, without the associated inductionless resistance, is 

 11*7 ohms. 



When such a bifilar wattmeter is employed to measure 

 electrical power, it can be shown that the deflexions of the 

 movable coil, when not large, are strictly proportional to the 

 power passing through the wattmeter, so that each scale- 

 division has an assignable power value, which in our case 

 was almost exactly *01 of a watt per millimetre of scale. The 

 arrangement of circuits was then as shown in the diagram on 

 p. 264. The long magnetizing coil above described was con- 

 nected in series with the fixed coil of the wattmeter, and also 

 with a sensitive electrodynamometer or ammeter suitable for 



* Methods of measuring hysteresis loss by one observation of an instru- 

 ment have been suggested by other observers ; but the practical value of 

 these methods has been very small, because the investigators have not 

 shown how the corresponding maximum value of the effective induction 

 is to be obtained. The rule above given supplies this deficiency. 



