Behaviour of Iron under Small Magnetizing Forces, 241 



The relation between fi and Bj for the different series is 

 shown in PL VII. fig. 1 (upper curve). It is seen that the value 

 of fi is very little affected by the frequency, the four sets of 

 points lying on practically the same curve for values of B x 

 up to 300 or so ; but beyond this the curves seem to separate, 

 the value of fi becoming smaller as the frequency is increased. 

 On the same diagram are plotted the points for the statical 

 permeability, and these also lie on the same curve. 



/-to decreases as the induction is lowered, this decrease 

 becoming more rapid as the value of J$i becomes smaller, 

 and the value of /jl appears to tend to a definite limit ot 

 about 157 for very low values of B r This limit is better 

 shown by plotting ^ against H^ If this be done, it appears 

 that up to about H=1'0 the relation between fi and H t can 

 be well represented by a straight line, as Baur and others 

 have already shown for the statical permeability. 



The values of 6 for the four series are plotted against B, 

 in fig. 1 (lower curve), and here, although there is not so 

 close an agreement between the four series as there was for 

 the jjb — Bj curves, yet the points all lie near a mean curve. 

 The value of 6 is seen to decrease rapidly as B x diminishes, 

 and the rate of increase becomes more and more rapid as the 

 induction becomes lower, the value of being about 1° when 

 B x is as low as 2, and there seems no reasonable doubt that 

 in the absence of eddy-currents 6 would become exceedingly 

 small for very low values of B : . 



The curves for b z (the ratio of the amplitude of the third 

 harmonic of the flux- wave to that of the first) seem to show 

 that this quantity decreases slowly as B : decreases, and finall}- 

 takes up a value of about *030 for the weakest fields used. 

 The behaviour of b 3 for the period T = *020 is somewhat 

 different, for there, in the weaker fields, b% increases. It 

 will be seen from Table I. that for this series h z also increases 

 in the weaker fields, and has a considerably higher value 

 than in the other series, though whether this is the cause or 

 an effect of the increase of b z cannot be said. yfr 9 , the phase- 

 lag of the third harmonic of the induction-wave behind the 

 first, shows the same general behaviour for all four series ; 

 gradually increasing as the induction decreases, it appears 

 to tend to a value of about 60° for exceedingly low values 

 of Bj. 



Of b 6} all that can be said is that it has a value somewhere 

 below *01, and of o/r 5 that it increases as the induction 

 diminishes ; but little dependence can be placed on the 



