32 DR. FARADAY’S EXPERIMENTAL RESEARCHES IN ELECTRICITY. (SERIES XX.) 
the influence of the magnetism, and such continues to be the case until it is placed 
just over the edge of the terminal face of the core, where it has no motion at all 
(here, by another arrangement of the experiment, it is known to tend in what is at 
present an upward direction from the core). If carried a little further outwards, the 
magnetism then makes the bismuth ball tend to go outwards or be repelled, and 
such continues to be the direction of the force in any further position, or down the 
side of the end of the core. 
2299. In fact, the circular edge formed by the intersection of the end of the core 
with its sides, is virtually the apex of the magnetic pole, to a body placed like the 
bismuth ball close to it, and it is because the lines of magnetic force issuing from it 
diverge as it were, and weaken rapidly in all directions from it, that the ball also 
tends to pass in all directions either inwards or upwards, or outwards from it, and 
thus produces the motions described. These same effects do not in fact all occur 
when the ball, being taken to a greater distance from the iron, is placed in magnetic 
curves, having generally a simpler direction. In order to remove the effect of the 
edge, an iron cone was placed on the top of the core, converting the flat end into a 
cone, and then the indicating ball was urged to move upwards, only when over the 
apex of the cone, and upward and outwards, as it was more or less on one side of it, 
being always repelled from the pole in that direction, which transferred it most 
rapidly from strong to weaker points of magnetic force. 
2300. To return to the vertical flat pole : when a horizontal bar of bismuth was 
suspended concentrically and close to the pole, it could take up a position in any 
direction relative to the axis of the pole, having at the same time a tendency to move 
upwards or be repelled from it. If its point of suspension was a little excentric, the 
bar gradually turned, until it was parallel to a line joining its point of suspension 
with the prolonged axis of the pole, and the centre of gravity moved inwards. When 
its point of suspension was just outside the edge of the flat circular terminating face, 
and the bar formed a certain angle with a radial line joining the axis of the core and 
the point of suspension, then the movements of the bar were uncertain and wavering. 
If the angle with the radial line were less than that above, the bar would move into 
parallelism with the radius and go inwards : if the angle were greater, the bar would 
move until perpendicular to the radial line and go outwards. If the centre of the 
bar were still further out than in the last case, or down by the side of the core, the 
bar would always place itself perpendicular to the radius and go outwards. Ail these 
complications of motion are easily resolved into their simple elementary origin, if re- 
ference be had to the character of the circular angle bounding the end of the core; 
to the direction of the magnetic lines of force issuing from it and the other parts of the 
pole ; to the position of the different parts of the bar in these lines ; and the ruling 
principle that each particle tends to go by the nearest course from strong to weaker 
points of magnetic force. 
2301. The bismuth points well, and is well repelled (2296.) when immersed in water, 
