Fiff. 2. 



240 Mr. J. P. Joule on the Effects of Mag7ietismy ^c. 



hand, the fact that the shortening effect is proportional to the 

 magnetic intensity of the bar multiplied by the current tra- 

 versing the coil, seems to indicate that, in this case, the effect 

 is produced by the attraction of the magnetic particles 

 by the coil. But then it will be asked, why so re- 

 markable an augmentation of the effect is produced 

 by the increase of tension in the case of the soft iron 

 bars ? When we are able to answer this question in a 

 satisfactory manner, we shall probably have a much 

 more complete acquaintance with the real nature of 

 magnetism than we at present possess. 



Postscript. 

 I have already, in the former part of this paper, 

 described an experiment which indicated that no altera- 

 tion in the hulk of a bar of soft iron could be pro- 

 duced by magnetizing it. I thought, however, that 

 it would be interesting to confirm the fact by an ob- 

 servation of the alteration of the dimensions of the 

 iron at right angles to the direction of its polarity. 

 For this purpose I took a piece of drawn iron gas- 

 piping one yard long, -/c^ths of an inch in bore, and 

 y^ths of an inch in thickness, A piece of thick 

 covered copper wire was inserted into this tube, and 

 bent over the outside of it in the manner indicated 

 by fig. 2. The lower extremity of the iron tube 

 being fixed, and the upper end being attached to the 

 micrometrical apparatus described in the first section 

 of this paper, each division of which corresponded to 

 _^_i^-^th of an inch, 1 obtained the following re- 

 sults: — 



Experiment 28. Drawn iron tube. 



