452 
MR. W. CROOKES OK THE SUPPOSED 
raised or lowered without sudden jerks which might rupture the fibre. The apparatus, 
in a word, is a small pendulum which hangs freely over the centre of a level table in 
the middle of a room having; all the windows closed to avoid draughts. 
“ When thus arranged, if the cylinder is left to itself, after oscillating and rotating 
for some time, it becomes almost motionless. If desired, it can be steadied more 
quickly by lightly touching it with something. When steady, if a second ivory 
cylinder is gently brought about a millimetre from the first cylinder, vertical and 
parallel to it, as shown in the accompanying figure, the hanging cylinder is observed to 
acquire a movement of rotation, accelerating, and apparently only limited by the 
torsion of the fibre. 
“ This rotation always takes place in the direction of the hands of a watch when 
the second cylinder is on the left of the first in relation to the observer, who is 
supposed to face the apparatus ; and in the contrary direction when the second cylinder 
is on the right. This double movement always takes place, whatever be the position of 
the observer- round the table, when he approaches the second cylinder. 
“ This law is of remarkable constancy, for during the course of my numerous 
experiments I have never met with a single failure when I carefully took all precautions 
to avoid interfering influences. 
“ The nature of the substance of the two cylinders has no influence on the produc¬ 
tion of movement, and the same may be remarked as to their mass. With liquids or 
solids, full or empty, the rotation is always the same. The second cylinder may even 
be replaced by a single stretched hair, or by a single fibre of silk, which is still more 
fine, without there being any sensible modification. 
“The speed of rotation is a function (l) of the length of the two cylinders; (2) of 
their nearness ; (3) of the diameter of the first cylinder. It is in direct proportion to 
this length ; it appears to vary inversely as the diameter of the first cylinder, and to 
diminish much faster than would be required by the law of inverse squares. 
“ Flat screens interposed between the experimenter and the apparatus, or placed the 
other side (at least when 20 centims. from the cylinders), interfere with the movement. 
When, on the contrary, they are arranged laterally on the right or left, or placed 
above or below, they are without action. A hemicylindrical screen placed behind the 
cylinders, as shown in fig. 2, so that the observer is opposite the opening, has the 
singular action of reversing the direction of rotation. 
“ I have ascertained that light is without action, whatever its nature, its intensity, 
or its direction. It is the same with heat. Neither can the action be attributed to 
electricity, for the cylinders, after immersion in water, are as active as before. The 
poles of a magnet are inert. The action of gravity is balanced by the tension of the 
suspending fibre. Lastly, air-currents cannot explain the remarkable uniformity of 
direction in which the rotation always takes place, nor its inversion in the case of the 
h emicy 1 in dri cal screen. 
“It is seen that the direction of rotation is closely connected with the position of 
