8 Mr Harris on Magnetic Intensity by the 
has been effected, sometimes by a single weight, at others by a 
shifting point of suspension, and occasionally by mechanical con- 
straint, all of which may cause important differences in the re- 
sults of experiment. 
13. The chances of error with bars whose points of suspen- 
sion are made to slide on them, are so manifest, that it seems es- 
sential to preserve in the bar a fixed point, so as to render this 
condition invariable. This point, as is evident, should be as 
nearly as possible in a line passing through the centre of gravity, 
the centre of motion, and the point of magnetic neutrality. In 
this case the bar may be conceived to vibrate without error, by 
the action of the two terrestrial resultants, in the same way as 
either of its polar halves would oscillate separately, if acted on by 
one of them alone, since the directive force of the earth may be 
conceived to constitute in the bar an attractive and repulsive sys- 
tem of parallel forces, each of which is resolvable into a single re- 
sultant, whose direction, intensity, and point of application, may be 
determined according to the elementary principles of mechanics. 
14. The following method of preparing and suspending a 
magnet for vibration, will be found to possess considerable advan- 
tage. The bar being at first roughly forged, a coarse hole is 
drilled through the middle of it. The upper part of this hole is 
then bushed with a small piece of brass, through which an ex- 
tremely fine hole is drilled, and which is subsequently intended 
to pass through the centre of motion, as also the centre of mag- 
netic neutrality. If the bar is to be suspended occasionally from 
the opposite side, then that side must undergo a similar prepa- 
ration; only, in this case, a second large hole is drilled at right 
angles to the former, also through the centre, Fig. 4; the fine holes, 
if requisite, may be drilled immediately through the steel. When 
the bar has been regularly worked, but before tempering it, equal 
distances are accurately measured off on each side the fine cen- 
tral hole above mentioned, so as to make the arms exactly equal, 
and two shallow slits are cut in the extremities, perpendicu- 
