in Nickel and Cobalt. 169 



in iron and cobalt, make a prediction by analogy as to what 



would prove to be the character of the curves for ^- in these 



metals. We should say that they would be sensibly straight 

 for a much longer distance than the curve for nickel, and that 

 in fact it might be difficult to carry the magnetization far 

 enough to detect any marked departure from a straight course. 

 So great a difference in behaviour as is indicated by a reversal 

 of the sense of the transverse effect, however, makes any such 

 predictions hazardous. 



This difference of sign in the rotational coefficients of the 

 magnetic metals is so anomalous and so important a fact, that 

 one returns again and again to its consideration. Quite 

 recently the determination of this sign for all three metals has 

 been made anew. I have now tested, in all, four plates of iron 

 (three of them having been cut from the same sheet, but the 

 fourth being of a different thickness and probably of a some- 

 what different character), two plates of nickel (certainly very 

 different from each other in condition), and one specimen of 

 cobalt. With all these the record is perfectly consistent. 

 Nevertheless it would be desirable to examine more specimens, 

 and those differing widely in character. Different experimen- 

 ters have observed many peculiar effects in iron under the 

 influence of magnetism and the electric current, magnetism 

 and mechanical strain, or the combined influence, which in a 

 certain form we have here, of all three; and these effects 

 appear to differ greatly, and sometimes to be of different signs, 

 in soft iron and hard iron or steel. Thomson has found* 

 that, under conditions of the above character, soft iron and 

 nickel are, in certain apparently very important particulars, 

 opposed in behaviour. I have looked in vain through all the 

 facts of this kind with which I am acquainted for any plausible 

 explanation of the fundamental phenomenon of the transverse 

 action, nor can it be said that any clue has been found to the 

 cause of the diversity observed. Nevertheless the opposition 

 which Thomson has found in the behaviour of soft iron and 

 nickel, under conditions of magnetism and mechanical strain, 

 furnishes an analogy which should not be lost sight of. 

 Thomson has moreover noticed that the effect which he was 

 studying in soft iron became reversed in this metal at a 

 very moderate value of the magnetizing force. It might be 

 well to test the direction of the transverse effect also with 

 very small intensities of the magnetic field. 



An extended examination of the effect in iron and cobalt, 

 similar to that which has been made in the case of nickel. 



* Phil. Trans. May 1878. 



