Intelligence and Miscellaneous Articles* 253 



part due to the action of currents, and perhaps also to the reaction 

 of gas-particles in highly rarefied media. In any case they are 

 real facts, since they have been verified by several experimenters ; 

 and they appear to have led to the realization of an apparatus in 

 some sort thermometric, endowed with extraordinary sensitiveness 

 in regard to radiant heat. — Bibliotheque Universelle, Archives des 

 Sciences Phys. et Nat. Jan. 15, 1876, pp. 84-88. 



ON THE ACTION OF HEAT IN MAGNETIZATION. BY L. FAVE. 

 It has long been known that the magnetic state of a steel bar 

 changes with the temperature. Coulomb, Kupifer, and other physi- 

 cists have studied the very complex laws of the diminution of 

 magnetic • intensity ; and the most marked consequence resulting 

 from their experiments is, that by raising the temperature of steel 

 to a certain point we cause it to lose definitively the magnetism it 

 has received. 



M. Jamin has recently demonstrated * that steel is capable of 

 receiving considerable magnetization at a temperature at which it 

 loses almost entirely that which it received when cold, while the 

 magnetism of a bar magnetized when hot diminishes very rapidly, 

 and in a very short time almost completely disappears. 



"We have investigated the variations of magnetic intensity by the 

 method of Van Eees, which permits a speedy determination of the 

 total quantity of free magnetism in a bar. 



If a helix, formed of a few turns of copper wire wound round 

 the bar, be slid quickly from the middle to a distance at which its 

 influence is no longer sensible, the initial deflection of the galva- 

 nometer-needle, produced by the very feeble induced current re- 

 sulting from the movement, measures sensibly the total quantity 

 of free magnetism of the bar. We are assured that the earth's 

 action is not sufficient to give an induced current that could disturb 

 the result. 



"We operated on bars previously annealed, and cooled slowly, so 

 that we could regard the changes of state as absolutely temporary, 

 and the variations of temper did not complicate the phenomenon. 



The bar to be examined is suspended horizontally to a copper 

 rod, above a tube pierced with a series of apertures forming gas- 

 jets close enough to heat the bar in a manner sensibly uniform 

 throughout its length. This arrangement permits the magnetizing- 

 coil to be introduced without the bar ceasing, even when inside, to 

 be under the action of the source of heat — an important condition 

 for following from its commencement the phenomenon of loss. The 

 current from a pile of ten Bunsen elements is caused to pass round 

 the bar ; then the coil is withdrawn ; the helix of copper wire, 

 protected by a non-conducting envelope, is afterwards introduced 

 as far as the middle of the bar, and then suddenly withdrawn ; the 

 initial deflection of the needle of a mirror-galvanometer produced 

 by this last movement is measured. The temperature of the bar 

 was given approximately by a thermoelectric couple. 

 * Comptes Mendus, December 22, 1873. 



