104 0. Barus — Apparent Hysteresis. 



respectively between 45°, 90°, 180°, and 360° both sides of 0°. 

 In figure 8 the line aa shows that the initial magnetization has 

 produced a decided decrement of rigidity ; line bb that the 

 persistent effect of longitudinal magnetization is an increment 

 of rigidity. The series of zigzags c (except the first which is 

 reached from zero or no apparent strain) bear this out very 

 fully. The first magnetization produces a deflection in a 

 direction opposite to the acting stress. On breaking the cur- 

 rent this deflection is increased (rigidity further diminished). 

 The second and subsequent magnetizations increase rigidity 

 but the original reading is not again reached. 



Figure 9 for twists between +90° and —90° is similarly 

 explained. The original magnetizations are indecisive, some- 

 times corresponding to increase sometimes to decrease of 

 rigidity. On the whole the initial slopes, aa, are nil. Slopes 

 bb correspond to marked increase of rigidity as usual. Full 

 account is given by the zigzags c. 



Figures 10 and 11 apply for twists between +180° and 

 -180°, and between +360° and -360°, respectively. Both the 

 initial and the subsequent magnetizations produce marked 

 increments of rigidity, as shown by the lines aa and bb for 

 each case and by the lines c. The scale of the phenomenon 

 increases as the limits of twisting increase. The displacements, 

 however, do not increase ; in other words the vertical breadth 

 of the cycles is not much greater for 360° than for 45°. 

 There appears to be a pronounced tendency for the displace- 

 ments to pass through maxima of slip for intermediate angles. 

 Finally, so far as can be discerned the slopes of the cycles are 

 the same throughout, rather an unexpected result. This 

 means that the mean increments of rigidity are proportional to 

 the twists, caet. par. 



Repetitions of the experiments with the same sample of 

 wire and small angles reproduced the original results as shown 

 for instance in figures 12 and 13. 



In view of the small effects produced in fields for (7=2, and 

 C=4: amperes, respectively, it seemed advisable to investigate 

 the effect of further doubling the strength of the longitudinal 

 field. The data are given on the same plan as in §6, in figures 

 14, 15, 16, IT. The displacements (vertical width of the 

 cycles) have remained about as large as before and tend to 

 pass through a maximum for intermediate angles of twist. 

 The slopes of the cycles have definitely increased, so that for 

 mean values, rigidity increases proportional to twist at a faster 

 rate in stronger fields. 



A rough summary of the principal data will be in place 

 here. 



