upon the Dimensions of Iron and Steel Bars. 239 



Experiment 27. Same hardened steel wire. Permanent 

 magnetism to begin with +1'3. Tension 1030 lbs. 



From the above experimentswe find that the induction of per- 

 manent magnetism produces no sensible effect on the length of a 

 bar of perfectly hardened steel, and that the temporary shorten- 

 ing effect of the coil is proportional to the magnetism multiplied 

 by the current traversing the coil. The shortening effect does 

 not in this case sensibly increase with the increase of tension. 

 I have made an experiment, in which a hard steel pillar was 

 subjected to a pressure of 80 lbs., which I have not inserted, 

 because the coil not being the same as I had hitherto em» 

 ployed, that experiment was not strictly comparable with the 

 rest. Its result however showed that the hard steel pillar 

 suffered a diminution of length equal to 0"1 of a division of the 

 micrometer, with a current capable of giving a magnetic polar- 

 ity of 1*7. This accords very well with the results obtained 

 with a tension of 1030 lbs. 



Copper is the only non-ferruginous metal which I have ex- 

 perimented on. In the trials made with wires of this metal, 

 pressure and tension were successively applied, and very power- 

 ful currents transmitted through the coil; but I could in no 

 case detect the slightest alteration in their dimensions. 



I shall not prolong this paper by a discussion of the pro- 

 bable cause of the remarkable facts I have pointed out. The 

 law oi elongation naturally suggests the joint operation of the 

 attractive and repulsive forces of the constituent particles of 

 the magnet as the cause of that phecnomcnon. On the other 



