GRAVITATION AS AFFECTED BY TEMPERATURE. 
369 
Preparation of the Vacuum .—-A heating coil of eureka wire is pushed tip over the 
lower section, WB (fig. 6), and another put on the upper section, AC, of the main 
tube. The middle section, as stated, is already wired. Also the whole of the glass 
tubing connecting to the pump and McLeod gauge, and the oxygen plant are 
similarly wired. The large carbon tube has a special heating coil of several layers 
designed to attain a high temperature. The whole system of heating coils is put in 
series with a lighting circuit and a current of 3'5 amperes is developed. The injector 
pump and Gaede pumps are started. The temperature attained by various parts is 
arranged according to requirement. The lower part of the vacuum tube, WB, 
attains a temperature of 130° C. The upper tube containing the soldered joints is not 
raised much above 100° C, while the sealed-on plates are kept as a rule below 60° C. 
Any small condensation of vapour on the cool window can be carefully removed by 
placing a carbon glow-lamp in front of it for a short time. The carbon tube is 
raised to 340° C., as shown by a platinum thermometer placed between it and the 
heating coil. The double process of heating and exhausting proceeds for five days 
(say 50 working hours), during which time the McLeod gauge should show steadily 
improving vacuum. At night, when the evacuation is stopped, the taps have to be 
shut in such order that the rapidly cooling main tube shall not receive vapour from 
the hot carbon. On the last day of exhaustion the vacuum would be, say, 10/*, when 
the temperature is full everywhere. Then the vacuum is washed out, once or twice, 
with oxygen to remove traces of the less absorbable gases, nitrogen, helium, argon. 
Finally on cutting off the heat the pressure will drop to the smallest readable by the 
McLeod gauge, say, 0'05/x. The carbon will then be about 200° C. The taps are 
closed till, when temperature everywhere is normal, the taps t x and t 2 are opened, and 
all three taps have warm siegelwachs run over them. The glass between tap t 3 and 
the pump is sealed off. 
Evacuation being finished, the apparatus has to be prepared for the experiment. 
Adjustments.— First, the torsion head is turned so that the beam mirror, S x (figs. 6 
and 7), exactly faces the window ; the electro-magnet, E, is lowered until its poles are 
level with the torsion head magnet, and by rotation of E the torsion head is brought 
to the correct azimuth. By raising E to the level of the clamping magnet and again 
suitably rotating the former, the inside magnet is brought down to bear hard on the 
torsion head and thus fix it. Under this condition the mirror, S 2 , fixed to the window 
will exactly stand for the position of the top of the fibre. Flotation of one always 
accompanies that of the other, both being rigidly attached to the main tube. 
Next, the loose heating coils are stripped off and the whole vacuum system, including 
the connecting tubes wrapped in several layers of cotton wool. Before, however, the 
cotton wool is put on two platinum thermometer wires are wrapped on the glass of 
the vacuum tube, one above, and one below the window, and compensating copper wires 
are arranged. 
In providing a water-jacket to screen the vacuum tube from the hot masses, M, M, 
