Magnetism of Iron Exhibited by Rotation. 15 



and, however the plate might be made to revolve, the parti- 

 cles of the fluid would revolve with the plate, and the en- 

 tire action of the plate would be reversed by reversing its 

 position. On the other hand, if we could find a plate of per* 

 fectly soft iron, then the magnetic particles of the fluid would 

 maintain their direction independently of the rotation of the 

 plate, and the effects observed by Mr Christie would not take 

 place. 



But iron cannot be obtained perfectly soft : There will be 

 certain points in the best manufactured iron, which are harder, 

 or which offer a stronger coercive power to the magnetic 

 fluid than others ; and, therefore, the magnetic particles will not 

 revolve with the plate as in the first case, nor remain constant 

 in position as in the second. The iron in the case in ques- 

 tion being, however, nearly soft, the greater part of the fluid 

 will maintain its oi'iginal polarized direction, but other par- 

 ticles, meeting with the obstructions alluded to during the ro- 

 tation, will be carried to the right or left hand with the plate, 

 and produce all the observed phenomena ; differing in effect, 

 according as the position of the plate is more or less advan- 

 tageously situated on the magnetic sphere, for the develope- 

 ment of its magnetism. This, at least, appears to be the 

 most simple explanation of the phenomena, and it has the ad- 

 vantage of requiring the acknowledgment of no principle of 

 magnetic energy that is not already admitted by most of the 

 philosophers of the present day, and which has been found 

 sufficient to explain all the phenomena. We ought to 

 observe, that these experiments appear to have been made 

 about three or four years back. They are very extensive and 

 varied, but as an abstract of them has already appeared in 

 this Journal, we shall not enter farther upon them in this 

 place. 



One of the most striking peculiarities in the preceding ex- 

 periments is, that no increase of effect is produced by multi- 

 plying the number of rotations, so that there was nothing to 

 conduct us to the inference that velocity had any essential in- 

 fluence. From some other considerations, however, which he 

 states, Mr Barlow was led to conceive that magnetism which 

 i« produced by various processes with iron, might even be ex- 



