160 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



before, bundles composed of from one to twenty thin steel 

 plates, the dimension of each plate being the same as be- 

 fore. The agfreement of the formulas with the observa- 

 tions has been perfectly satisfactory ; and, as he remarks, 

 " we are now enabled to construct to order a magnet contain- 

 ing any given quantity of magnetism, or affording at its ex- 

 tremity any desired intensity." We can also calculate the 

 position of the poles and the magnetic moment of such a 

 ma2:net, and can determine the maximum limit of its mao^- 

 netism. There remains, however, still to be examined the 

 question of the variation of magnetic power with tempera- 

 ture, and with the chemical or physical structure of the met- 

 al. 6 i?, LXXXI. 



01^ MAGNETS COMPOSED OF COMPEESSED POWDER. 



In connection with his researches on magnetism, Jamin 

 states that he has repeated some interesting observations 

 published in 1836 by De Plaldat. He has, viz., by com- 

 pressing strongly, by means of a small hydraulic press, some 

 iron filings in a tube, found that when they commenced to 

 solidify the polarity increased considerably, and continued 

 increasing with the pressure. In order to be sure of the 

 nature of the original powder, he prepared some from pure 

 soft iron which had no appreciable magnetic force, but the 

 resulting phenomena were as before. We have, then, the 

 case of a metal which has no coercitive force when it is con- 

 tinuous, but which acquires such, and that to a great degree, 

 when it is reduced into discontinuous small fragments, and 

 when these are joined by pressure. Is it not, then, to this 

 discontinuity that it is necessary to attribute the observed 

 polarity, and does not this experiment explain the magnetic 

 force of steel ? The question thus put reminds one of 

 Jamin's previous experiments, which have shown that the 

 distribution of magnetism within a magnet can only be ex- 

 plained by considering the latter as composed of very small 

 filamentary magnetic elements having opposite poles, and 

 acting upon each other. The present pressure experiments 

 seem to show that these filaments may possibly be identical 

 with the small fra^-ments of iron filino-s, or with the small 

 crystals agglomerated together in the case of steel. A 

 further investigation of this subject, by means of a very 



