C. Barns — Apparent Hysteresis. 109 



360° than at 90°, for instance; This is the natural explanation 

 of the greater displacement corresponding to the greater stress 

 and of the change of sign of the displacement with the pre- 

 existing stress. Again the residual effect at zero is greater if 

 the twist has been accompanied by magnetization before 

 removal. Cf. JV, JV\ Combined magnetization and twist 

 (say at 360°, for instance) correspond to a greater value of 

 stress in the absence of magnetization ; for the magnetization 

 wipes out the lagging or unstable configurations, bringing the 

 strain at 360° in correspondence with the stress. 



Why, it may next be asked, is the displacement at zero after 

 removal of twist 360°, greater than the displacement at 3tf0°, 

 for instance. We meet here, I think, the usual occurrence in 

 viscous deformation.* For example, if a strained or hard 

 metal is tempered at 100° C, it is then molecularly stable or 

 shows no deformation or change of temper at temperatures 

 below 100° or for smaller values of strain, even when the 

 deformation at the original temperature (100°) and strain still 

 continue. 



If the wire is twisted to 360° (say) not only is strain stored 

 up in the wire, but at the same time new instabilities are 

 evoked by the process of twisting. When the stress is with- 

 drawn or the wire untwisted to zero, instabilities are not 

 encountered at nearly the same rate ; for the preceding larger 

 strain has wiped them out for the diminishing smaller strains. 

 Hence the wire is less viscous when twisted and more viscous 

 when untwisted for like stages or angles of twist. Now any 

 twist no matter how small or within the elastic limits stores 

 strain in the wire, which action possibly, is but another expres- 

 sion for the configurations broken during the process. This, 

 as I understand it, is the explanation of figures 18-20, in 

 which the slip at zero following a reduced strain is so much 

 larger than the slip at 360° following a growing strain. 



12. If the interpretation given be correct, then on shaking 

 out a strain with longitudinal magnetization, the wire should be 

 free from slip for circular magnetization ; and vice versa. This 

 is the case as shown in figure 21, p. 100, where the successive 

 scale readings are given (vertically) at equal intervals (abscissas) 

 apart, the current being alternately made and broken begin- 

 ning with no field. L denotes a longitudinal field, (7, a circu- 

 lar field if first in action ; L r and C\ the corresponding fields 

 if they succeed the action of other fields. The twist is alter- 

 nately + 180° and —180°. Mere inspection shows that from 

 the cases L' and C slip has practically been wiped out. It is 

 markedly present in cases L and C, the first deflection being 



*Barus and Strouhal, Bulletin U. S. Geolog. Survey, No. 14, p. 57, 1885; 

 Wied. Ann., xi, p. 965, 1880. 



