136 H, Nagaoka on Hysteresis attending Change of 



we get a similar curve of hysteresis (fig. 6, PI. I). The 

 only difference between the two is that the field which gives 

 the minimum contraction is higher in the second case. 



With still greater range of the magnetizing field very 

 similar results are obtained, as shown by the curves (figs. 6 

 and 7, PI. I.). An examination of these curves will illustrate 

 the character of the hysteresis better than a mere verbal 

 description. 



The contraction of the magnetized nickel wire does not 

 show any maximum so far as the experiment goes, but the 

 curves have two points of inflexion, one in weak and the other 

 in tolerably large fields. The measured changes of length 

 are nearly the same as those given by Bidwell *. (See Table II. 

 at the end of the paper.) 



Before entering into the description of the hysteresis 

 observed in iron, I will describe a few experiments on the 

 changes observed in iron wires of different lengths. The 

 curves A, B, C (fig. 1, PL II.) represent the elongations 

 observed in iron wires whose lengths were 66, 22, and 10*6 

 times the diameter respectively. In the longest, the elonga- 

 tion reaches a maximum in field 70 nearly, and shows an 

 inflexion-point before reaching the maximum. In curves B 

 and C the inflexion-point has not been reached. On con- 

 tinuing the curve B (given in fig. 3, PL II.), the maximum 

 is found in field 230, which is very large compared to the 

 former. This relation of the elongation to the length of 

 the magnet is very analogous to the well-known relation of 

 magnetization to length. 



The changes wrought by a cycle of magnetization on the 

 length of iron wire is so complex as to require a detailed de- 

 scription. An experiment made on a wire 19*4 centim. long 

 and 2 '83 millim. thick will illustrate the nature of the hys- 

 teresis. (For measurements, see Table III.) The curve 

 given in fig. 2 (PL II.) shows that the elongation in weak 

 fields increases gradually, but that, beyond a certain strength 

 of field, it increases rapidly until it reaches the " wendepunkt.''' 

 The rate of change then gradually lessens, and the elongation 

 reaches a maximum in field 70. Beyond this point, in 

 increasing fields, the elongation diminishes steadily. The 

 curve thus traced up to field 305 is abed. With decreasing 

 current the wire again elongates, but shows lagging, so that 

 the length of the wire for the same strength of field is shorter 

 on the return journey until field 120 is reached. Here the 

 branches of the curve cross. In lower fields accordingly the 



* Proc. Roy. Soc. 1890. 



