26 DPt. FARADAY’S EXPERIMENTAL RESEARCHES IN ELECTRICITY. (SERIES XXII.) 
recede from the magnet. The whole was covered with a glass shade, and when left 
to become of uniform temperature, and at rest, the needle magnet was found to take 
up a constant position under the torsion force of the susperiding filaments. Further, 
any rotation of the glass and copper wire support did not produce a final change in 
the position of the magnet ; for though the motion of the air would carry the magnet 
away, it returned, ultimately, to the same spot. When removed from this spot, the 
torsion force of the silk susj)ension made the system oscillate; the time of a half 
oscillation, or a passage in one direction, w'as about tliree minutes, and of a whole 
oscillation therefore six minutes. 
2566. When a crystal bismuth was fixed on the support with the magnecrystallic 
axis in a horizontal direction, it could be placed near the lower pole of the magnet 
in any position, and being then left for two or three hours, or until by repeated exa- 
mination the magnetic pole was found to be stationary, the place of the latter could 
be examined and the degree and direction in which it was affected by the bismuth 
ascertained. Extreme precaution was required in these observations, and all steel 
or iron things, as spectacles, knives, keys, &c., had to be dismissed from the ob- 
server before he entered the place of experiment ; and glass candlesticks were used. 
The effect produced was but small, but the result was, that if the direction of the 
magnecrystallic axis made an angle of 10°, 20°, or 30° with the line from the mag- 
netic pole to the middle of the bismuth crystal, then the pole followed it, tending 
to bring the two lines into parallelism ; and this it did whichever end of the magne- 
crystallic axis was towards the pole, or whiehever side it was inclined to. By 
moving the bismuth at successive times, the deviation of the magnetic pole could be 
carried up to 60°. 
2567. The crystal of bismuth therefore is able to react upon and affect the magnet 
at a distance. 
2568. But though it thus take up the character of a force acting at a distance, 
still it is due to that power of the particles which makes them cohere in regular order, 
and gives the mass its crystalline aggregation ; which we call at other times the attrac- 
tion of aggregation, and so often speak of as acting at insensible distances. 
2569. For the further explication of the nature of this force, I proceeded to examine 
the effect of heat on crystals of bismuth when in the magnetic field. The crystals 
were suspended either by platina or fine copper wire, and heated, sometimes by a 
small spirit-lamp flame applied directly, sometimes in an oil-bath placed between 
the magnetic poles ; and though the upward currents of air and fluid were strong in 
these cases, they were far too weak to overcome the set caused by magnecrystallic 
action, and helped rather to show when that action was weakened or ceased. 
2570. When the temperature was gradually raised in the air the bismuth crystal 
continued to point, until of a sudden it became indifferent in that respect, and turned 
in any direction under the influence of the rising currents of air. Instantly removing 
the lamp flame the bismuth revolved slowly and regularly, as if there were no ten- 
dency to take up one position more than another, or no remains of magnecrystallic 
