1882.] The Principles of Magnetism . 27 
the same time, the astatic tendency between the surface and 
the central groups of magnets would be assisted. Con- 
sequently, in any such magnetic system, the special relations 
which tend to establish themselves between contiguous 
magnets must be overcome by general relations between 
the various sections of the group, most probably resulting 
in the above condition of astatic relations between the 
superficial and the central sections of the system. As 
already said, such a relation could only partially display 
itself in any group of mass magnets, on account of their 
resistance to reversal of magnetism. But it could readily 
appear in groups of molecular magnets, from their ability to 
shift their axial directions. We are, therefore, not without 
warrant for the conclusion that in every magnet this rela- 
tion of its molecules exists, and that the magnetism of the 
central portions of the magnet is opposite in direction to 
that of its surface portions. 
But still another conclusion must arise from these 
premises. The conditions here assumed could only exist in 
the equatorial plane of the magnet. If we take the some- 
what parallel case of a globular mass, every particle of which 
is drawn by attraction towards the centre, we know that 
this is not because of a force of attraction existing in that 
centre, but because the attractions of all the remaining 
particles have their greatest sum in the line passing through 
the centre. But in the case of a magnetic mass the forces, 
though in one sense general, are in another sense special. 
The magnetic particle on the equator of the magnetic mass 
would be equally influenced by the northern and southern 
sections, and would thus obey on equally balanced axial 
influence. But a magnetic particle out of the equator, and 
nearer one pole than the other, must feel a preponderating 
influence from the greater mass in the direction of the more 
distant pole. It would not arrange itself, therefore, in true 
astatic parallelism with the axial particles ; but, if it be 
nearer the North Pole, it must be affeCted most strongly by 
the greater mass towards the South Pole, and thus be 
thrown out of its reverse parallelism with the axis. As it 
comes nearer and nearer to the North Pole this reversing 
influence from the South Polar region must grow more 
effective, until, when the North Pole is reached, the 
magnetic particle must become fully reversed, and in direCt 
parallelism with the axis. All this may seem a far reaching 
deduction from the astatic tendency of any two magnets. 
But it can be shown that it is in exaCt agreement with the 
ordinary phenomena of magnetism. 
