Magneto-electric Phenomenon. 219 



Since the moment of inertia of a disk is also proportional to 

 the fourth power and to the thickness, so long as the thickness 

 is small, it will appear that the velocity imparted to a disk of 

 any size or thickness, or to a ring, during a change in a mag- 

 netic field will be the same. To what extent a correction 

 should be applied to these results for self-induction between 

 different parts of the disk I am not prepared to say ; their 

 calculation would give trouble. 



For comparing one field with another, disks or rings of 

 metal may be used ; but for absolute measurements, as it 

 would be impossible to measure the exact conductivity of a 

 disk, a coil is preferable. By employing disks of different 

 metals their conductivities can be compared without the 

 trouble of drawing into wire or cutting into long strips. 



It will be found that a coil of area A and resistance R will, 

 under a torsion T ? experience a throw 



Q= A 2 H 2 sin2a 

 4R^MT~* 



If, instead of a disk, a sphere be used, no twisting should be 

 experienced if the conductivity in different directions is the 

 same. If, however, there is a plane of greatest or least con- 

 ductivity, it should be possible to discover it. Crystallization 

 or mechanical treatment might give rise to such planes in 

 metals; no definite results could be expected in any thing else. 



I have referred to the apparent repulsion and attraction of 

 a disk of metal by a pointed pole at the making and breaking 

 of the magnetizing current. As the lines from such a pole 

 radiate outwards, they are not normal to the metal except in 

 the middle. On their passage inwards or outwards they give 

 rise to circular currents tending to move each part of the disk 

 normally to the lines of force. There is therefore a longitu- 

 dinal component away from the point of radiation during an 

 increase in the field, and towards it during a diminution of the 

 field. A closed coil of wire is subject to the same forces. If a 

 coil be made of uncovered copper wire in the form of a double 

 helix with the ends joined together, and if the convolutions 

 are separate so as nowhere to touch one another, the growth 

 of the magnetic field can be watched by placing the coil nearly 

 over one pole. On making the current the field begins to 

 grow, at first quickly and afterwards more slowly. The coil 

 will receive a push and will extend itself. As the push 

 diminishes in amount, owing to the diminishing rate of growth 

 of the magnetic field, the coil will gradually regain its former 

 shape. It might be thought that the slow recoil is simply due 

 to the damping action of the field ; but this is not sufficient 



