induced or exhibited in Iron , by Rotation , 1 21 
pass placed near the same, at any azimuth, and on either side. 
Moreover, there were several holes in the table, which enabled 
the experimenter to screw down the frame itself in any azimuth. 
The ball could only revolve with its axis horizontal, but it could 
be made to revolve direct or reverse at pleasure. The platform 
might be depressed or raised to any height, and the needle thus 
placed over or under any proposed point of the shell. 
Things being thus prepared, the platform was first placed in 
the horizontal plane of the halfs axis, and the needle placed suc- 
cessively at every point all round. And it appeared, that what- 
ever might be the azimuth of the needle (provided it was neu- 
tralized by magnets from the directive power of the earth), on 
turning the ball, the north end of the needle approached to- 
wards it when the motion was towards the needle, but receded 
when the motion was from the needle. That is, when the upper 
part of the shell, by the revolution, descended towards the needle, 
the north end of the needle approached towards the shell ; but 
when the ball revolved in an opposite direction, then the south 
end of the needle approached, or the north end receded. 
When the needle was carried round the shell in a vertical cir- 
cle, ascending 10° each time, the following results were observed, 
(the needle being in every case neutralized, and placed parallel 
to the axis of rotation), viz. from the horizon to an altitude of 
about 54°, the needle placed itself perpendicularly to the axis, 
the north end passing the contrary way to that in which the mo- 
tion of the shell was made'; from 54° to 90°, or to the zenith of 
the ball, the needle also placed itself perpendicularly to the axis, 
but in a reversed position to what it took up before, viz. the 
north end passed in the direction of motion in the shell. It did 
the same on the other side of the vertical, to a like point or alti- 
tude of about 54° ; but from that point to the horizon, it ar- 
ranged itself as at first. Below the horizon it also preserved the 
same direction, till it amounted again to 54° ; it then changed 
its direction as when above the horizon. There were, therefore, 
four points of change in the direction of the needle (the motion 
of the ball remaining the same), viz. at 54° above and below the 
horizon of the ball, on each side of the zenith and radius. It will, 
of course, be understood, that, by reversing the motion of the 
ball, the needle also changed its direction, but the points of 
