CONTEMPORARY ADVANCES IN PHYSICS 307 



normal to ds and the vector I. Magnetism in the amount I -ds- cos 6 

 is to be spread upon ds; magnetism is to be spread over the surface of 

 the elhpsoid with surface-density /-cos d. This film of "magnetism" 

 would produce, everywhere outside of the ellipsoid, the same field 

 as the poles or the continuous magnetization which we have imagined 

 as existing inside the ellipsoid. Furthermore, it has a firmer basis 

 in experience than do the poles. For, if a beam of polarized light is 

 directed against the surface of a magnetized ellipsoid, the reflected 

 beam is curiously altered; this effect, known by the name of its 

 discoverer Kerr, is sometimes extremely complicated, but in magnitude 

 it is always proportional to the value of the imagined quantity /-cos d 

 at the point where the reflection occurs; and by promenading a spot 

 of light over a magnetized piece of iron and analyzing at every point 

 the reflected beam, one can actually find how /-cos 9 varies all over 

 the surface. This property endows the vector / with a physical 

 reality. 



There is still one of the effects which a magnet produces outside 

 of itself, which requires our attention; did it not exist, magnets 

 would not play nearly so great a role as they do in the life of the world. 



Hitherto I have implied that one maps out the external field of a 

 magnet by exploring it with some one of the known field-measuring 

 devices, of which there are several: the magnetometer needle, the 

 bismuth wire which changes its resistance according to the field im- 

 pressed upon it, the plate of glass which rotates a traversing beam of 

 plane-polarized light to an extent proportional to the field. There is 

 another method essentially different from these, and capable of 

 measuring something which they cannot. One may set up a loop of 

 wire in the neighborhood of the to-be-magnetized piece of metal; 

 suddenly impress the magnetizing field; and measure the sudden 

 rush of charge around the loop. This rush of charge is proportional 

 to the mean value of the magnetic field thus suddenly created in the 

 region enclosed by the loop.* One might map a field in this manner; 

 but that is not the unique feature of the method. 



We consider a special and actual case. Take an unmagnctized 

 ring of iron; cut out a thin segment, leaving two nearly parallel 

 end-surfaces facing one another across a narrow gap; take a loop 

 slightly larger than the cross-section of the ring, suspend it in the gap, 



* The E.M.F. around the loop at any instant is equal to the time-derivative of 

 the surface-integral, over any surface bounded by the loop, of the component of the 

 magnetic field normal to the surface; in technical language it is equal to the rate 

 of change of the flux of magnetic field through the loop. The rush of charge is 

 equal to the quotient of the time-integral of this E.M.F., which is the difference 

 between the initial and final values of the surface-integral, by the resistance of 

 the loop. 



