272 Prof. J. A. Fleming on a Method of determining 



measuring small alternating currents of about an ampere or 

 less in value. 



The shunt-coil of the wattmeter was joined across the ends 

 of the long coil. Suitable resistances beyond controlled the 

 current. 



The process of measurement was then as follows : — A 

 bundle of straight iron strips was fitted with the secondary 

 coil at the effective point, viz. "56 of the half-length of the 

 bar from the centre. The bundle was slipped into the mag- 

 netizing coil, and an alternating current sent through the 

 coil of such a magnitude (generally about '2 of an ampere) 

 as to give an induction density of not more than about 

 6000 C. G. S. units at the effective point. The wattmeter 

 deflexion was then read. To interpret this reading the iron 

 strips were removed from the coil and another wattmeter 

 measurement made on the copper coil alone. As the removal 

 of the iron sends up the current strength by decreasing the 

 inductance of the circuit, it is necessary to readjust the current 

 to the same value as before, and to know that value in amperes. 

 The resistance of the coil being known, and also the current 

 passing through it, we can calculate the power being taken 

 up in the copper coil alone for this standard current ; and 

 hence know the power in watts producing the observed 

 wattmeter deflexion. 



The wattmeter is therefore calibrated on the copper circuit 

 of the magnetizing coil, and from the known copper loss in 

 this coil and the proportionality of its deflexions we are 

 enabled to assign at once the value in watts of its deflexion 

 when used to measure both the iron loss and copper loss 

 together when the iron strip is inserted in the coil. 



If the induction is carried up as far as 4000 or 5000 G.Gr.S. 

 units at the effective point, and if the strips have a thickness 

 over '01 inch, then it becomes necessary to apply a small 

 correction for the eddy-current loss set up in the iron. This 

 can be at once done by using the following simple practical 

 formula given by the author * for this loss in rectangular 

 sectioned strips. 



If t is the thickness of the strip in mils (1 mil = *001 inch), 

 n is the frequency of the current, and B is the maximum 

 value of the induction density during the cycle at any point 

 on the bar, then it can be shown that the eddy-current loss in 

 watts per cubic centimetre at that point in the rectangular 



* See ' The Alternate Current Transformer,' Fleming, vol. ii. p. 490, for 

 proof of this formula. 



