1885.] 



Experimental Researches in Magnetism. 



Gl 



netism to varying load in a constant field is examined at great length, 

 both in soft annealed wire, and in wire hardened by stretching, and 

 the effects, which are very complex, are exhibited graphically. They 

 also afford an easy criterion by which a stretched piece can be distin- 

 guished from an annealed piece, and, both in stretched and annealed 

 metal, they exhibit static hysteresis, which can be more or less com- 

 pletely wiped out by mechanical vibration. The connexion of the 

 author's present work with the earlier results of Villari, and the 

 extensive researches of Sir W. Thomson, is pointed out. 



Numerous curves are given, showiug the relation to J to p when 

 the process of magnetisation is conducted while the piece under test 

 is maintained in a constant state of stress. These show very clearly 

 the position of what Thomson has called the " Yillari critical point," 

 or point at which the sign of the effect of stress on magnetism 

 becomes reversed; and it is shown that this reversal is associated 

 rather with a particular value of magnetisation than with a particular 

 value of magnetising force, the same reversal being found to occur 

 when the magnetism dealt with is wholly residual. 



Instances are given of very curious molecular reminiscences of 

 previous loads, which were found to exist even when the previous 

 loading had been performed at a time when the metal was entirely 

 destitute of magnetism, and although the loads whose effects were 

 found to remain were much less than loads previously applied to the 

 same wire, and were consequently unable to produce mechanical 

 change of any ordinary kind. These residual effects of former loads 

 became visible on the next magnetisation of the piece. It is shown 

 that these effects of previous loads can be almost obliterated by sub- 

 jecting the wire to vibration before beginning to magnetise it. 



The ratio of residual to induced magnetism in soft and hardened 

 iron and steel has been examined for various degrees of magnetisation, 

 with the result of showing that the ratio passes a maximum when 

 the magnetisation is moderately strong, and then falls off as the state 

 of so-called saturation is reached. The influence of the presence of 

 stress on this ratio and on residual magnetism generally has also 

 been investigated. 



The last part of the paper deals more briefly with the effect of tem- 

 perature on magnetism, a subject already largely treated by G. Wiede- 

 mann and others. The present experiments were specially directed to 

 discover whether there is hysteresis in the relation of magnetism to 

 temperature when the temperature suffers cyclic changes. Careful 

 experiments on both induced and residual magnetism have led the 

 author to conclude that within the range of temperature dealt with 

 by him, there is no perceptible hysteresis in the relation of magnetism 

 to temperature. Curves are also given exhibiting the process of mag- 

 netising the same piece at different temperatures. These cross each 



