Prof. G. Wiedemann's Magnetic Researches. 65 



It follows from these experiments that the temporary tor- 

 sion of the wires, produced by the smaller weight after the action 

 of the greater weight, with a decreasing load, is in every case 

 greater than that produced by the action of the same weight 

 with increasing load, and before the action of the greater 

 weight. The behaviour is observed not only during the accom- 

 modation of the wire, but also after it has attained a perma- 

 nent condition, when a repetition of the cycle of experiments 

 gives the same results. It also is not altered by strong vibra- 

 tion of the wire. 



We may hence assume that, in the action of any torsional 

 forces whatever (which may also be zero), the molecules are 

 displaced and rotated in accordance with the deformation pro- 

 duced by these forces, and afterwards, upon the action of 

 other forces, pass from their first positions of equilibrium 

 into others, which, however, are conditioned by the first. If, 

 then, the molecules are greatly displaced and rotated by a 

 more powerful torsion, they still partially retain this displace- 

 ment and rotation upon the action of a feebler force ; w r hilst, 

 conversely, upon changing from a feebler to a stronger tor- 

 sional force, the latter is unable to produce so great a displace- 

 ment and rotation of the molecules which lie nearer their 

 neutral position. 



A similar behaviour in the case of wires is shown by an 

 experiment of H. Tomlinson's*, in which, however, only small 

 differences were observed between loading and unloading. 



§6. 

 The behaviour of iron bars in temporary magnetization, 

 by increasing and decreasing forces, is exactly similar to 

 the mechanical behaviour here described. Messrs. Righif, 

 Fr online J, and Warburg § have shown that if an iron bar 

 is successively magnetized by forces a, b, c, b, a, of which 

 a<b<c, the temporary moment M'j ; corresponding to the 

 force b with decreasing magnetization, is greater than the 

 temporary moment M 6 , corresponding to the same force with 

 increasing magnetization. Warburg has ingeniously deduced 

 from this the behaviour of a magnetic needle vibrating above 

 an iron plate. 



* Phil. Trans. Roy. Soc. Lond. 1883, Part I. p. 10 (Exp. V.). Tom- 

 linson, in his experiments, put on the weights by hand without observing 

 irregularities ; it seems to me that in the method now described, whicn. 

 was used by me in 1858, better security against accidents is obtained. 



t Mem. di Bologna, 20 Mai, 1880 ; Beibl. v. p. 62 (1884). 



\ Wied. Ann. iv. p. 102 (1878), xiii. p. 318 (1881). 



§ Wied. Ann. xiii. p. 141 (1881). 



Phil Mag. S. 5. Vol. 22. No. 134. July 1886. F 



