Length by Magnetization in Nickel and Iron. 137 



elongation is greater in the descending than in the ascending 

 curve, until field 25 is reached. At this point the wire 

 begins to shorten very abruptly. In zero field the wire is, 

 nevertheless, longer than at the first maximum. When the 

 field is reversed, the length continues to decrease, but only 

 slightly _, until field —15 is reached. Here the minimum 

 occurs, and the wire begins again to lengthen. The rate 

 of increase is comparatively slow, and reaches a maximum 

 in field —70. Beyond this the wire shortens quite rapidly, 

 and nearly in the same way as in ascending positive fields. 

 With decreasing magnetizing current, the wire again shows 

 hysteresis and the part I m goes below kl: The two branches 

 cross in field —110, and the wire goes on elongating till 

 it reaches a maximum atp in field —25. The course of 

 changes from I till the field reaches its former maximum 

 value is nearly similar to the curve d ef g 7ikl, as will be 

 seen from inspection of the curve Impqrst (the return 

 curve beyond q is given in dotted lines to avoid confusion). 

 Although the curve of hysteresis during a magnetic cycle 

 is very complex/ it is symmetrical with respect to the line 

 of no magnetizing force. It is evident that the maximum 

 and minimum points in the curve are due to the lagging of 

 the maximum elongation. 



The measured elongations of the wire are nearly the same 

 as those given by Bidwell. The field at which the wire 

 recovers its original length will probably be about 400. On 

 extrapolating from the curves given by Bidwell *, I find that 

 when the wire has no longitudinal stress it will show no 

 elongation at about the same strength of the magnetizing 

 force. 



The course of the curve is somewhat changed when the 

 limiting field is only a little greater than that of maximum 

 elongation. This case is elucidated in fig. 3 (PI. II.), from 

 an experiment made on a wire 10*5 centim. long and 4*75 

 millim. thick. On account of the shortness of the wire com- 

 pared to its thickness, the first maximum is only reached in 

 field 230 approximately, and, moreover, the slope of the curve 

 in higher fields is very gradual. After describing the curve 

 abed, the curve d e f g hklmnpqrs is traced during a 

 cycle. The peculiarity observed in this case is that in place 

 of having the maxima as in fig. 2, there are curious " humps " 

 in the curve at/ and n. Had the magnetizing current been 

 sufficiently strong to cause a large decrease of elongation, the 



* Loc. cit, 



