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



the secondary coil gave the R.M.S. value of the induced 

 electromotive force set up in the exploring coil. The form 

 of the curve of the primary current was very nearly a true 

 sine curve, and the variation of induction density in the iron 

 at any point closely followed the same law. 



In any case, if/ is the form factor * of the secondary 

 voltage curve, which in these experiments was 1*13, and if 



JST = the number of windings in the secondary coil, 



S=the total cross-section of the iron, 



B = the maximum value of the induction density in the 

 iron in C.G.S. units at that point in its length at 

 which the secondary coil is placed, and 



n = the frequency of the induction cycle ; then if e 2 is the 

 R.M.S. value of the secondary electromotive force 

 in volts as measured by the voltmeter, we have 

 always 



10 8 e 2 = 4/'NSBrc. 



Hence the maximum value of the induction density B can 

 be calculated from the known value of all the other quantities. 

 In this manner a large number of measurements of the 

 value of the induction density B were made at different points 

 in the lengths of various bundles of iron strips of different 

 lengths and total cross-sections. 



These values were then laid down in a series of induction 

 density curves so drawn that the ordinates represented the 

 maximum value of the induction density in the bar at dif- 

 ferent transverse sections along the half-length of the bar, 

 taken as corresponding abscissae. 



It has been shown by Steinmetz and fully confirmed by 

 others that for most varieties of iron used for electrotechnical 

 purposes the hysteresis loss in the iron, per cycle of induction, 

 per cubic centimetre varies as a power of the maximum value 

 of the induction density, which is very nearly 1*6, when that 

 maximum value of the induction density has any value below 

 about i>000 C.G.S. units. Hence if the ordinates of the 

 above-mentioned induction curves are all raised to the l'6th 

 power and a new curve plotted showing the variation of 

 B 1 ' 6 along the half-length of the bar, we have another curve 

 which represents the variation of the hysteresis loss p'er unit 

 of volume along the half-length of the bar from point to 

 point. A large series of such curves showing the variation 

 of B and B 1 ' 6 were drawn for bundles of iron strips and steel 



* See ' The Alternate Current Transformer,' Fleming, vol. i., new. ed., 

 p. 583, for a reference to this term. 



