454 
MR. W. CROOKES OK THE SUPPOSED 
at the end, and passing over a pulley for convenience of raising and lowering the 
glass. A similar arrangement is attached to the other glass windows. Cotton-wool 
is put at the bottom of each glass window and round the support M, to keep out ah’ 
draughts. 
The cylinder is attached to the end of the fibre with a loop and hook, so that it can 
be rapidly changed for other cylinders. The fixed pillar screws to the arm, so that it 
also is easily changed. Ivory, ebonite, glass, and metal have been used for the 
cylinders, and ivory, ebonite, brass, and wood for the pillars. The pillars have also 
been made square, round, and wedge-shaped in section, and the surfaces have been 
bright and lampblacked. The apparatus was fitted up in a room free from draughts 
and quick changes of temperature, and during the course of the experiments no one 
but myself entered the room. 
The cylinder mostly used was of ivory, 5 mm. in diameter and 25 mm. long. The 
first pillar experimented with was also ivory, 7 mm. diameter and 30 mm. long. The 
mode of experimentation was the following:—The cylinder being at rest, I sat down 
in front of the apparatus with my face 8 inches from the cylinder and pillar, taking 
precaution to keep the breath as much as possible away from cylinder and pillar. The 
pillar was always placed on the right of the cylinder. On raising the front glass the 
cylinder commenced to rotate in the opposite direction to the hands of a clock, the side 
nearest me moving to the right.* It made 4'5 complete revolutions, the maximum 
speed being one revolution in 12 seconds. 
In several succeeding experiments a four-ounce flask of boiling water was used as 
the source of heat. It was coated externally with lampblack, and was placed exactly 
one inch from the pillar and cylinder. 
In other experiments the ivory pillar was replaced by one of hollow brass, 9 mm. in 
diameter and 38 mm. long, the surfaces being brightly polished ; by a wedge-shaped 
pillar of boxwood, 38 mm. high, 22 mm. broad, 9 mm. at the thickest end, and tapering 
oft’ to a blunt edge at the thinnest part; by an ebonite cylinder 9 mm. in diameter and 
38 mm. long ; and by a brass parallelogram 38 mm. long by 7 mm. square. 
The results are given most conveniently in a tabular form. The first column of 
the following Table gives the numerical order of the experiment, the second column 
shows the material of which the pillar consists. In the third column is given the 
maximum speed in seconds of one revolution ; the fourth column gives the number of 
revolutions performed by the cylinder before being stopped by torsion of the fibre ; 
and the exciting agent, the face or hot-water flask, is shown in the last column. In 
all cases the rotation was negative, namely, in the opposite direction to the hands of a 
clock. 
* I will call this the negative direction, and when the rotation is clockwise I will call it positive. 
