Prof. G. Wiedemann's Magnetic Researches. 69 



That this is also the case with an unchanged distribution of 

 the magnetizing forces is seen from the above experiments. 



§8. 



In close connection with these phenomena is the thermo- 

 electric behaviour of stretched wires towards the same un- 

 stretched, studied by Cohn *, which is certainly one of the 

 most sensitive modes of detecting changes in structure. 

 Here, also, the thermoelectric forces which correspond to 

 equal tensional forces are different with increasing and with 

 decreasing tension. There could be no doubt that the in- 

 vestigation of the thermoelectric behaviour of wires twisted 

 to and fro would give similar results, and this is confirmed 

 by experiment ; as also in the investigation of the same 

 behaviour with iron and steel bars, which are magnetized by 

 increasing and decreasing magnetizing forces. In so far as 

 the torsion or detorsion affects the electric conductivity of 

 the wires, it must show analogy to the magnetic behaviour. 

 This much may, however, safely be concluded from these 

 investigations, that if the molecules of the bodies are dis- 

 placed or rotated from their former positions of equilibrium, 

 before as well as after the molecules have accommodated 

 themselves by the repeated action of the force, they will 

 always, to some extent, also retain this displacement or rotation 

 upon gradual alteration of the force. 



This is seen in various ways. Thus, for example, if the 

 force influencing the body is reduced to zero, a permanent dis- 

 placement and rotation remains ; again, in order to reduce a 

 permanent torsion or magnetization to zero, a smaller force is 

 necessary than is required for the production of the same. 

 Further, a body which has accommodated itself by the re- 

 peated action of increasing and decreasing forces, and is then 

 again repeatedly exposed to the action of the same forces, 

 shows each time, more or less, the position of the molecules 

 corresponding to the greater or smaller force preceding that 

 allowed to act last. 



Since it is not possible, in the present condition of our 

 knowledge, to express these phenomena of elasticity in the 

 widest sense of the word by mathematical laws under appro- 

 priate theories, further experimental researches on this subject 

 are required, in which the time during which the forces act, 

 and the time which elapses after they cease to act, must be 

 taken into account. A thorough study of the magnetic pro- 

 perties of bodies, from which we can infer, up to a certain 



* Wied. Ann. vi. p. 385 (1879). 



