110 Mr. S. Bidwell. On the Changes produced hy 



when the magnetising force ceased to act. Using a soft iron wire a 

 quarter of an inch in diameter, Joule found that when it was loaded 

 with a weight of 408 lbs. the effects were the same in direction (though 

 smaller in degree) as when the wire was unstretched ; its length 

 increased when it was magnetised, and diminished to the same extent 

 when it was demagnetised. When, however, the load was increased 

 to 740 lbs. the effects were reversed, and magnetisation produced 

 temporary retraction. After describing this experiment Joule 

 expresses his belief that with a tension of about 600 lbs. (roughly 

 the mean of 408 and 740), " the effect on the dimensions of the wire 

 would cease altogether in the limits of the electric currents 

 employed," i.e., that currents which produced on his tangent galvano- 

 meter deflections ranging from 6 to 58 degrees would neither increase 

 nor diminish the length of his quarter inch wire when stretched with 

 a weight of 600 lbs. If he had actually made the experiment he 

 would perhaps have found that the length of the wire was increased 

 by a weak current, that a current of medium strength would have had 

 no effect whatever, and that one of his stronger currents would have 

 caused the wire to retract. 



Joule's experiments have many times been repeated, and his results 

 generally confirmed. In particular Professor A. M. Mayer of the 

 United States, carried out a series of very careful experiments with 

 apparatus of elaborate construction and great delicacy.* The con- 

 clusions at which he arrived were in accord with those of Joule so far 

 as regards iron ; but in the case of steel there is apparently some 

 discrepancy. Mayer found that (after the first magnetisation) the 

 steel rods with which he worked, whether soft or hard, were invariably 

 shortened when the circuit was made and lengthened when it was 

 broken, the same current being used for the first and for the 

 subsequent magnetisations. This result is, however, not necessarily 

 inconsistent with Joule's, because the conditions of the experiment 

 were not the same, the second current which Joule applied being 

 stronger than the first, and the third stronger than the second. 

 Again, while in the case of Joule's "soft steel" the movements were 

 in the same direction as those observed with iron (though smaller in 

 degree), Mayer's " soft steel " behaved in exactly the opposite manner, 

 the movements (after the first magnetisation) being in the same 

 direction as those which occurred when harder steel was employed. 

 This difference may be accounted for, as Mayer himself suggests, by 

 supposing that his so-called " soft steel " was harder than Joule's. 

 Possibly too there was a sufficient difference in the magnetising 

 forces employed in the two cases to affect the results of the experi- 

 ments. More will be said on this point further on. The effects 

 resulting from the first action of the magnetising current are 

 * "Phil. Mag.," vol. xlvi, p. 177. 



