MR. T. ROYDS ON THE CONSTITUTION OF THE ELECTRIC SPARK. 
335 
so that the lines which appear at both the top and bottom, of the spectrum can be 
measured and the means taken, Schuster and Hemsalech showed that the velocity 
of the vapour between two points is equal to 
where 
x 2 Xi 
K is the reciprocal of the magnification of the system (= 1'16) ; 
v is the velocity of the film at the centre of the spectrum ; 
x 1 , x 2 , the co-ordinates of the points measured parallel to the spectrum ; 
y 2 , the co-ordinates of the points measured at right angles to the spectrum. 
Even if the line can only be measured at either the top or the bottom, this 
expression may generally still be used, the differences from the more accurate one 
being less than the errors of experiment. If, however, it should be required to 
calculate velocities from the more accurate formula of Schuster and Hemsalech, it 
will be necessary to know the distance of any line in the spectrum from a vertical 
through the centre of the disc. This has been obtained by measuring on the disc itself 
the distance of a vertical through the centre of the disc from where the air line 
A. 5004 falls on the disc. From this, the distance for any line in the spectrum can 
be found. 
Method of Making the Measurements .—A number of fine lines were drawn parallel 
to each other, at equal distances of about 8 millims., on a sheet of paper, and reduced to 
the necessary size by photographing on a glass plate. In this way rulings have been 
made whose spaces were 0’0503 centim. and 0‘0200 centim. 
The photograph of the spectrum was placed flat on the ruling with the two 
emulsions together, and clamped on the travelling microscope stage of Kayser’s 
measuring machine, the pitch of whose screw is ^ millim. The stage is moved by 
turning the screw until the vertical wires in the eye-piece are on the point where a 
streamer intersects a line of the ruling, and the reading of the screw head then 
recorded. Thus the ruling gives the y readings, and the screw the x readings. 
The x readings have to be corrected for the curvature of the spectral lines which is 
introduced by the use of prisms. 
It is necessary that the spectral lines and the vertical wires in the eye-piece should 
both be at right angles to the direction of travel of the stage, and the lines of the 
ruling parallel to this direction. Since the lines in the stationary photograph are 
finer than the air lines in the rotating one, the adjustments are first carried out by 
means of the stationary photograph. The photograph is clamped to the ruling with 
spectrum parallel to the lines of the ruling. The whole is then adjusted on the stage 
until a point in the eye-piece remains on a line of the ruling as the screw turns. 
The ruling and the spectrum are then parallel to the direction in which the microscope 
stage travels and the spectral lines at right angles to it. The wires in the eye-piece 
