521 
is about 23 mm., so that light is allowed to pass during 1/20000 
second. 
An image of the stationary slit S,, projected on the photographic 
plate of the camera C, when the dise rotates, is reproduced in 6, 
on the Plate, fig. 1: the two cross-wires are seen, and on the 
righthand side of the slit the rim of the circular plate R, the 
centre of which lies on the left side. The dark circles are caused 
by reflections, but are of no consequence. 
When S, moves and the disc revolves, after every thousandth 
of a second the light is let through the slit in #, and S, is photo- 
graphed by means of the lens JZ,. 
The images of S, assume an oblique position, because during the 
displacement of the beam, the slit in the dise gets continually 
higher in its movement. 
It is easy to see that the slope of the image is determined by 
the ratio of the velocity of the slit to that of the beam, or 
rather to that of the image on the disc. From the distances of S, 
and A to L,, and from LZ, to R and the plate, tbe reduction which 
the velocity of the beam undergoes in the image, can be immediately 
estimated, or it can be directly measured by photographing a divided 
scale in the plane S, S,. The amount of obliqueness of the slit 
image (Plate fig. 2) immediately gives an approximate value of the 
velocity of the beam, which can, indeed, be found in a still 
simpler way from the distance of corresponding points of 6’, 6"). 
For a more accurate determination of the velocity the second 
slit S,, which is at 4,15 em. distance from S,, can be of service. 
Let us suppose that about 1000 em./sec. have been found for the 
approximate value of the velocity of the beam, then S, has shifted 
about 1 em. after every thousandth of a second, and S, gets about 
to the first position of S, after four thousandths of asecond. Hence 
an image of S,, viz a,", will appear 
in general on the photo, from which the position of the slit images 
between the images of S, 
can be accurately derived, and then we know that in order for the 
beam to move a distance 4,15 em., + + a fraction thousandths of 
seconds are required, which can be measured from the relative 
positions. In order to distinguish the slits, a cross-wire has been 
stretched only over S,. 
We only get two images of each of the slits on the photo, because 
it is only possible that images are formed by the different lenses 
within a limited cone. 
As it is our intention to determine the velocity of the beam with 
a delinite direction of motion, the shutter (Ll § 4), which as a rule is 
