198 Prof. E. Bouty's Studies on Magnetism. 



spring. The bundle, tempered and magnetized all in one piece, 

 is afterwards divided. 



If the mother needle (the bundle) is saturated, we find that 

 the separate laminae are not. This is not surprising ; for the 

 superposed laminae react on one another in a direction the re- 

 verse of their magnetization, and consequently the permanent 

 magnetic moment which they can attain while united is less than 

 if they were separate; but what does at first appear astonishing 

 is that the sum of the magnetic moments of the separate laminae 

 is considerably greater than that of the primitive bundle. 



This last fact is well brought out by making the experiment 

 in the following manner. I take a bundle which is square in 

 section, formed by the union of four square bars of the same 

 length. This combination is tempered hard and immediately 

 magnetized ; I measure its magnetic moment, and then take it 

 to pieces and measure the moment of each bar separately. The 

 sum of the moments is found to amount to very considerably 

 more than the magnetic moment of the united bars. Joining the 

 bars together in twos, the sum of the moments of these partial 

 bundles is intermediate to those of the total bundle and of the 

 separate bars. And, finally, if we reconstitute the primitive 

 bundle, the magnetic moment returns to its first value. 



In this experiment the bundle, which, innocent of any pre- 

 vious magnetization, has only once undergone the action of 

 the magnetizing spiral, is in an absolutely normal condition 

 at the instant of the first separation ; and we do not see that 

 any new magnetic force interposes to which we might attri- 

 bute the observed augmentation of permanent magnetism. It 

 is true that by separating the bars we suppress their mutual 

 reaction, and we know that in each of them it acts in opposition 

 to the permanent magnetism ; but that can have no other 

 effect than to cause a certain amount of temporary magnetism 

 to disappear. Thus, in a normal bar, a certain quantity of 

 permanent magnetism is superposed to a temporary magnetism 

 opposite in direction, which latter the removal of a magnetic 

 layer parallel to the axis causes to vanish. 



This curious superposition may be observed in other circum- 

 stances. It is known, for instance, that if a magnetized bar of 

 steel be submitted to the action of a current of too little inten- 

 sity to demagnetize it entirely, during the action of the current 

 there is observed a diminution of the bar's magnetism, which 

 may proceed even to the reversal of the poles, while it is found 

 after the cessation of the current that the bar is magnetized in 

 the primitive direction. It is also known that, if the south pole 

 of a powerful bar magnet be presented to the south pole of a 

 magnetized needle, the repulsion which is manifested at a great 



