of the Coefficients of Induction. 



459 



b. 



02- 



e r 



L 2 - 



A. 







688 



508 



•0637 



•220 



1 



627 



462 



•0634 



•200 



5 



444 



321 



•0606 



•147 



10 



367 



261 



•0577 



•110 



20 



230 



160 



•0547 



•073 



30 



164 



111 



•0517 



•055 



100 



49-5 



29-75 



•0419 



•020 



130 



38-0 



22-7 



•0414 



•016 



200 



24 



135 



•0381 



•0105 



230 



20-5 



110 



•0360 



•0092 



corresponding values obtained by Maxwell's method three 

 months previously. This is probably due to the fact that 

 the electromotive forces of the cells used were only approxi- 

 mately determined. The differences, however, are small, and 

 may have been due to alterations in the state of the iron. 

 The values for L 2 obtained for small currents are seen to be 

 much lower than those given by the equation 



L 2 = -2A + -0425. 



They are, however, very much higher than that of the coil 

 with the iron core removed. The coefficient in this case was 

 •0028 secohms. 



6. The coefficient was very much increased by keeping a 

 piece of soft iron pressed against the poles of the electro- 

 magnet. As, however, under these circumstances, a large 

 fraction of the magnetism remained on removing the mag- 

 netizing current, the values obtained depended upon the 

 previous history. Thus the value of L 2 obtained for a mag- 

 netizing current of about *04 ampere varied between '337 

 secohm and "130 secohm, according as the flings were ob- 

 tained on reversing the current, on breaking the battery- 

 circuit, or on making it so as to include or exclude the 

 permanent magnetism. The numbers obtained showed, 

 moreover, that the curve of magnetization was not quite 

 symmetrical with respect to positive and negative currents. 



The following table gives the values of L 2 obtained for 

 different currents by the method of reversals. The resistances 

 and electromotive forces used to obtain the last five obser- 

 vations were quite different from those used to obtain the 

 first six. 



