36 president's address — section a. 



From the theory of electrostatics to that of magnetism is a 

 very easy step. Each is founded on exactly similar laws, and 

 the results must in the two cases be similar. Here we have 



a magnetic potential 12 and a magnetic force -r-, usually 



called H : we have a " magnetic inductive capacity," more 

 often called " permeability," and denoted by the symbol fx, and 



B 



a " magnetic displacement " -j— : B is generally called the 



magnetic induction. These three quantities are connected 

 by the law (analogous to that of electrostatics) 



37— = -r- • -T— or B = mH- Word for word the general 



problem of magnetism and its particular solutions may be 

 copied from those of electrostatics, if only we substitute for 



D, K, and E f or j the symbols -r—, /i, and H. The 



genera] problem is tlien the same in the two cases. But there 

 are modifying circumstances which render the particular 

 cases different in the two theories. In electrostatics the 

 quantity K varies only in magnitude between certain narrow 

 limits. In those media where it is highest it is only two or 

 three times the value it has in those media in which it is 

 lovvest. It is true we may look upon conductors as parts of 

 the medium in which K is inhnite, but we may also consider 

 them as bounding surfaces over which V is always constant. 

 Strictly speaking this is the jn-oper method to take, because a 

 conductor is not a medium in which K is carried to an 

 extreme, an infinite value, but a medium possessing a new 

 property, that of conductivity. Olass, for example, possesses 

 both properties, has two constants, one of specific inductive 

 capacity, another of conductivity. In magnetism the range 

 of fi is much greater. In iron, for example, it may be 

 thousands of times as great as in air. In magnetism, too, 

 there is no jmrallel to the conductor in electricity. The sur- 

 face integral of B over any closed surface is always zero. We 

 may, in fact, consider our medium to extend through all space. 

 There is no need to partition off certain portions of space by 

 boundaries. But we must remember also that in certain 

 bodies ^ is variable. It depends (1) on the value B has at 

 the time, (2) on the value B has had in former times, i.e., on 

 the substance's retentiveness. This last peculiarity renders 

 possible the existence of lines of induction, which return into 



