of Magnetic mid Non-jnagnetic Bodies. 425 



in hardened iron and steel ; the cause of which is, that in soft 

 iron the particles are much more mobile around their position 

 of equilibrium. Soft iron and steel magnetized by the action 

 of a magnet or a second helix, exhibit weaker vibrations when 

 the external current tends to magnetize them in the direction 

 in which they are already magnetized, and stronger in the con- 

 trary case. The passage of a continuous current through an iron 

 wire modifies the sound which the intermittent action of the 

 external current of the helix causes it to emit, provided that 

 the transmitted current is very strong and the iron very soft. 



I shall not enlarge more on this first case, which 1 think is 

 now well explained ; I pass to that of the transmitted current. 



An electric current transmitted intermittently through 

 a wire of very soft iron and of small diameter (of 1 to 3 

 millimetres), determines in it vibrations as strong as those 

 which are produced in it by the same current acting extei'- 

 nally under the most favourable conditions. If the iron wire 

 becomes larger, or if it is more hardened, or if it is a steel 

 wire, the effect of the transmitted current is less than that of 

 the external current*. With rods of soft iron, of 4 to 5 

 millimetres in diameter, the transmission of the continuous 

 current does not completely extinguish, but only diminishes, 

 the effect of the discontinuous current. If the rods are very 

 large, the diminution becomes less perceptible, unless very 

 strong batteries are employed. In the preceding experiments, 



1 made use of two of Grove's batteries of five cells, one for 

 the continuous current, the other for the discontinuous one. 



The same experiment made on steel rods and wires gave 

 me precisely contrary results. The sound produced by the 

 transmission of the discontinuous current is weak ; but it is 

 increased, instead of being diminished, by the passage of a 

 continuous current passing in the same direction as the dis- 

 continuous one. It is a curious fact, that this augmentation 

 remains some instants after the continuous current has ceased 

 to pass ; and that it disappears, not all at once, but by degrees 

 and by fits. 



I performed these experiments with steel wires of 1 and 



2 millimetres in diameter, with rods, both tempered and not 

 tempered, of 3 to 4 millimetres in diameter. The results 

 were the same; the strengthening due to the continuous cur- 

 rent is more decided with the rods than with the wires f. 



* This experiment of the iron wire covered with silk its whole length, 

 and which gives all the same sounds as when its surface is perfectly naked, 

 would seem to indicate that external causes, such as friction, enter less 

 than is imagined into the production of these sounds: this point deserves 

 to be more closely examined. 



f The steel rods are magnetized by the simple passage of these disconti. 



