256 Intelligence and Miscellaneous Articles, 



This observation is accurate but imperfect. I have repeated it 

 by forcibly compressing with a small hydraulic press the iron filings 

 in the tube. When they begin to aggregate, the polarity is seen 

 to augment considerably, and continues to increase with the pres- 

 sure. I exhibit to the Academy tubes of from 8 to 10 centims. 

 length and 3 centims. diameter, which attract at least as much 

 steel filings as would be attracted by pieces of good steel of the 

 same dimensions. 



As the filings I employed were of unknown origin, I had some 

 prepared in my presence from very soft iron, perfectly reduced and 

 without appreciable coercive force. The results were quite equal 

 to the former. Here, then, is a metal which has no coercive force 

 when it is continuous, but acquires a coercive force equal to that 

 of steel when we reduce it to minute fragments and then make 

 them contiguous by pressure. Must not the polarity observed be 

 attributed to this discontinuity ? and is not the coercive force of 

 steel to be explained by the same cause ? 



The distribution of the force in a magnet cannot be accounted 

 for without regarding it as composed of rows of minute magnetic 

 elements with opposite poles reacting on each other at a distance ; 

 and it is verified that the quantities of separate magnetism in each 

 of them are increased by this reaction, from the extremity to the 

 middle line (Lame, Physique, vol. iii. p. 100). Up to the present 

 it seems to be assumed that these elements are the molecules them- 

 selves ; the preceding experiment seems to show they are formed 

 either by fragments of iron in proximity to one another, or by 

 minute crystals agglomerated as in steel. 



When, before pressing the filings, we interpolate among them 

 substances which render the mass more homogeneous, the same 

 polarity cannot be given as when they are unmixed. For example, 

 by making a paste of chloride of iron and the filings and pressing 

 it, we obtain, after a few days, a subchloride continuous in appear- 

 ance, capable of being filed and polished like pure iron, but hardly 

 magnetizable. 



Iron reduced by hydrogen, and scales of oxide of iron, behave 

 like iron filings ; but magnetic or diamagnetic substances mixed 

 with the filings alter notably the faculty they possess of being mag- 

 netized. The study of all these circumstances promises interesting 

 researches. Hitherto I have had at my disposal only insufficient 

 apparatus and a small laboratory press. It is probable that by greatly 

 increasing the compression of the powders the coercive force will 

 be found to increase to a maximum, and afterwards diminish when 

 the approximation of the fragments has restored a sufficient degree 

 of continuity to the mass. 1 shall soon be in a position to commu- 

 nicate to the Academy the result of these new researches. — Comptes 

 llendnsde VAcademie des Sciences, Aug. 2, 1875, vol. lxxxi. pp. 205-207. 



