36 PROFESSOR SWAN ON THE GRADUAL PRODUCTION 
become blended by their persistence on the retina. The wheel,B alone would 
likewise be insufficient, for the disc D revolves so fast compared with B that 
the sector S would have returned to its position for transmitting light before the 
hole I had time to get out of the line of vision. 
It is obvious that the arrangement might be extended to any number of wheels, 
the only remaining practical difficulty being that of driving a train of toothed 
wheels at very high velocities. If this should render it impossible to observe im- 
pressions so short as may be desired, there are two methods which may be 
adopted. The disc itself may be driven directly by a reaction machine, or Barker’s 
mill, the train of wheels then serving merely as screens and registers of the num- 
ber of revolutions made by the disc. Impressions of very short duration might 
also be obtained by substituting for the disc a small revolving mirror, like Wheat- 
stone’s. The mirror attached, with its plane parallel to the axis on which it re- 
volves, receives the light of a flame which has passed through a narrow slit 
parallel to the axis, and at a considerable distance from it. The light reflected 
by the mirror falls upon the surface, whose brightness is to be observed, and 
illuminates each portion of it for a short interval of time, depending on the width 
of the slit and the velocity of the mirror. Impressions of extremely short duration 
could thus be obtained, for a small mirror could be driven at a much higher 
velocity than a disc, which, from its considerable size, experiences great resist- 
ance from the air; and the narrow slit would become equivalent to a sector of 
corresponding width, cut in a very large disc, revolving with twice the angular 
velocity of the mirror, and having a radius equal to the distance of the slit from 
the axis of revolution. 
I have designed an instrument on the principles now explained, which has 
been constructed for me by Messrs James MILNE and Son of Edinburgh, and which 
is represented in fig. 2. 
The instrument consists of a solid cast-iron frame A A, carrying a train of 
wheels B, C, D, E, by which the disc F is driven. Motion is communicated to 
the wheels by a weight suspended from a cord coiled round the barrel G. Each of 
these wheels B, C, D, E has 120 teeth ; the pinion 0 has 30 teeth, and the remaining 
three pinions, of which only d and ¢ are shown in the figure, have each 12 teeth. 
The wheels B, C, D and the disc F are each 6 inches, and E is 3 inches in 
diameter. } 
From the numbers of teeth in the wheels and pinions, it is obvious that for each 
revolution of the wheel D the wheel E makes 10, and the disc F 100 revolutions; 
and by means ofa train of gas-meter index-wheels H K, to which motion is commu- 
nicated by the wheel I on the axis of the wheel D,—I and H having the same num- 
ber of teeth—the number of revolutions made by the disc F is readily ascertained. 
To facilitate counting the revolutions, the frame carrying the wheels H K turns 
on a pivot about half-way between H and K, and is pressed by a spring near H 
