426 
MR. J. E. H. GORDON ON MEASUREMENTS OF ELECTRICAL CONSTANTS. 
Thus the electrifications, of the needle, the five plates, the dielectric, and the 
four quadrants were all reversed 12,000 times in each second, and yet the light 
spot was found to be absolutely steady, and completely under the control of the 
screw of a. 
A motion of a of ‘001 inch usually caused a deflection of from 1 to 2 millims. 
on the scale. I so arranged the connections that, to bring the light spot to zero, a 
had to be moved so as to follow the deflection. This was a very convenient way of 
remembering in which direction to turn the screw. 
The actual determinations were made as follows :—The dielectric was removed and 
the slide on which it stood being pushed in under the plates, the position of a was 
determined which brought the light spot to zero with only ah' in the balance. The 
slide was then drawn out, the dielectric placed upon it, and it was than pushed in so 
as to place the dielectric between a and b . The dielectric plates were all 7 inches high 
and 7 inches or 8 inches long, and varied from \ inch to 1 inch in thickness. A 
second reading was taken when the dielectric was in position. 
Before each set of measurements, preliminary experiments were made by means of 
which a was placed very nearly in its right position in each case before the primary 
circuit was closed. This prevented large deflections of the needle, which were found 
injurious, as after a very large deflection the light spot seldom returned exactly to the 
same zero. 
Before commencing a measurement the engines were started : the observer standing 
where he could see the scale, was able with his right hand to make and break the coil 
primary circuit, and with his left to work the screw of a (see fig. 5). As soon as 
equilibrium was established, contact was broken, the engines stopped, and the reading 
of a taken with the microscope. 
The formula used for calculating the results assumes that the lines of force through 
the dielectric are straight and perpendicular to its sides. Now the lines of force are 
very nearly straight over most of the space a b, but not quite straight close to either 
a or b. When the lines of force through the dielectric are perpendicular to its sides 
they will be at their shortest in it, and it will produce its minimum effect. By means 
of the mechanical slide the dielectric was in each case adjusted to give its minimum 
effect. The field of force was, however, found to be very nearly uniform except close 
to the plates. A cardboard box covered with tinfoil, and connected to earth, pro¬ 
tected the second stage which carried the electrometer connections. 
All w r ires leading from the quadrants to the balance were protected by wide card¬ 
board" tubes covered with tinfoil connected to earth. The metal portions of the cases 
of balance and electrometer were also connected to earth. 
* May, 1879.—Brass tubes ancl a brass box over the upper stage have since been substituted with 
advantage. 
