of Sound in the Bell Telephone, 133 



being contracted in diameter when the current reinforced the 

 magnetism of the pole, becoming enlarged when it passed in 

 the opposite sense. 



Knowing from the experiments of Joule and De la Rive 

 that a portion of iron, when magnetized in a particular direc- 

 tion, grows longer in that direction and shorter in its trans- 

 verse dimension, let us deduce what the effect will be on 

 the diaphragm of a telephone of these two species of mag- 

 netization. If the magnetization were radial, the tendency 

 would undoubtedly be, supposing the disk clamped circum- 

 ferentially, to thrust the middle point of the disk backwards 

 towards the magnet, and to give it a conical shape. If the 

 magnetization, on the other hand, were lamellar, the tendency 

 would be to make the diaphragm thicker, and to contract it 

 over the area thus magnetized. In the actual case where the 

 magnetization partakes of both characters, the two distribu- 

 tions being separated by a neutral zone, the tendency to each 

 form would exist over the regions respectively affected. But 

 the extent of these regions varies with the varying induction 

 of the currents in the coil. Hence, while the total attraction 

 varies, giving rise to oscillations of the diaphragm as a whole, 

 the neutral annular line will also be continually shifting its 

 position and predisposing the diaphragm to take up new nodal 

 forms of vibration, thereby rendering the timbre corresponding 

 to the complicated undulations of the currents arriving from 

 the transmitter. 



The result obtained may be regarded from another point of 

 view. If a slight displacement of the iron disk, though unable 

 to affect to any appreciable extent the strength of the magnetic 

 field as a whole, alters its strength at any one point or in any 

 narrow region, or if, even without altering the average num- 

 ber of lines of force in any part of the field, such a displace- 

 ment shifts the position of some of the lines of force across 

 a narrow region of the field, it may still exercise a consider- 

 able inductive action on a closed coil of wire lying in the 

 region where the amount of shifting is greatest. For, since 

 the induced electromotive force in a closed circuit is not pro- 

 portional to the strength on an invariable magnetic field in 

 which it lies, nor yet to all changes in its strength, but only 

 to such changes as cause a greater or less number of lines of 

 force to pass through the area within the closed circuit, it is 

 evident that the inductive action will be strongest in coils of 

 wire which lie in the region where there is most change in the 

 direction of the lines of force. We have here the rationale of 

 the empirical practice of the constructors of the telephone 

 alluded to above — namely, that of using only a small coil of 



Phil. Mag. S. 5. Vol. 8. No. 47. Aug. 1879. L 



