102 



C. Barns — Apparent Hysteresis. 



The interesting feature of the experiments are thus the 

 lines, a, together with the corresponding displacements of the 

 tiducial zeros given by the zigzags, c. These phenomena can- 

 not be reproduced in brass* stressed within the elastic limits. 

 Hence I infer that they are viscous in character and due to the 

 breakdown of unstable magnetic configurations (in Maxwell's 

 view) resulting from the new magnetization. For although 

 the viscosity of a magnetic wire and of a non-magnetic wire 

 are identical, the period of breakdown which follows the first 

 magnetization must be rich in unstable configurations and, 

 therefore, temporarily low in viscosity. I conclude, therefore, 

 that the displacement of the zero is due to the molecular 

 instability which accompanies the act of magnetization when 

 first applied ; that in subsequent magnetizations the molecular 

 disturbance is less because the configurational effect of the first 

 magnetization is largely permanent. Fixed configurations are 

 merely deflected in their elements without being broken up 

 anew. Further discussion would be of little value because of 

 the heat discrepancy, occurring in spite of the submergence.f 

 The essential data are to be incorporated with the remarks 

 made for longitudinal fields. 



6. After these results it seemed necessary to reconsider in 

 detail the corresponding phenomena in a longitudinal field 

 since here the heat discrepancy can be quite eliminated. This 

 was done with the same precautions. The helix was supplied 

 with currents of about 2, 4, and 8 amperes, respectively, evok- 

 ing longitudinal fields of roughly 200, 400, and 800 c. g. s. 

 units. The identical annealed iron wires were each of about 

 35 centims. long and -024 cm in diameter. They were insulated 

 from each other so that a current could be passed through the 

 lower alone, which was submerged in flowing water while the 

 upper wire was suspended freely in air. Both wires were 

 stretched in the same vertical line. Cf. figure above. 



But one observation was made with the weak field for C~ 2 

 amperes, as shown in figure 7, the twists being alternately 4-90° 

 and —90°. The interesting features are the directions a and b, 

 corresponding respectively to the first and subsequent magne- 

 tizations, and, therefore, to the temporary and the persistent 

 effects. Both are increments while in the above cases of cir- 

 cular magnetism they were decrements. The displacement of 

 the zero, however, here and above has the same direction and 

 corresponds to diminished viscosity (high molecular instability) 



* I also tested the effect of a current on a submerged brass wire stressed a 

 little beyond the elastic limits. The results showed that whereas the rigidity is 

 decidedly modified by the presence . of a current, the viscosity of the wire is not 

 noticeably different in the two cases. The effects are due to heat. 



f See preceding paper in this Journal, IV, x, p. 417, 1900. 



