368 
DR. P. E. SHAW OH THE NEWTONIAN CONSTANT OF 
the glass blowers is open at top, A, and has the two side tubes welded on and the 
window flange prepared. First the inside is given a thick coat of silver up to about 
20 cm. above the window flange. This coating reduces radiation during the 
experiment and enables one to thoroughly “ earth ” that part of the vacuum 
surrounding the suspended system. The central portion, DW, of the main tube is 
wrapped externally with asbestos paper, and over that eureka wire is wound to act 
as a heating coil. The brass tube already mentioned, with platinum wire at one end, 
is inserted in tube C, and the glass sealing on tubes C and D is completed. 
A mercury seal t outside the platinum seal is shown. The copper inner sheath is 
lowered into the main tube and the rosette at the window opened ; the strap, U, 
which carries the sheath, is fitted to the brass tube in C to keep the sheath in position. 
The optically-true window is now sealed on to its flange. 
The torsion system is next prepared. From a frame holding some hundred quartz 
fibres of 60 cm. length and suitable diam. (i.e., from 21 /u to 1 b/j. diam.), four or five 
have been chosen several days before and all have been attached to a horizontal bar 
and loaded with a weight about 25 per cent, more than the actual load used in the 
experiment. It is impossible to predict to within 50 per cent, what a quartz fibre of 
known diameter will carry. A fibre may sustain a load for several hours or even a day 
and then break without external shock being applied. Hence every fibre is given a 
three or four days’ test before use. The finest fibre which has stood the test is fixed 
by Margot’s solder to the torsion head above and to the beam system below, which 
in the present case we will suppose has form 2, fig. 12. The whole suspended system, 
except the small mirror, is of the purest obtainable metal, say copper. After the 
beam system has been put together, but before the fibre is attached, the whole of 
it, except the mirror, is immersed in nitric acid, and then thoroughly washed, to 
remove any trace of iron which may have become attached by tools or by the hands 
during construction. 
It is important to see at this stage that the mirror is properly inclined to the 
vertical as no adjustment of it can be performed later in the vacuum. 
Next the torsion system, i.e., torsion head, fibre and beam system, are lowered 
with the utmost care to prevent breakage of the fibre, into the vacuum tube till the 
torsion head rests sochetted in a hole prepared in the brass tube, C (fig. 6). Several 
asbestos paper discs are placed on the top of the torsion head frame to keep the latter 
cool in the next stage. The top of the main glass tube is next sealed. This sealing 
of a 5 cm. tube requires two operators, one on each side with a blowpipe. Care is 
taken to keep heat from the torsion head just below. In later experiments a glass 
plate was waxed on the top (see figs. 9 and 10). The vacuum tube and contents 
is now cautiously removed and set on its gimbals in the supporting frame. The 
adjusting screws below are so set that the suspended system swings free of the tube 
walls so as to be ready for the experiment. The vacuum side tube, D, is sealed to 
the carbon tube, C a, and this to the pump (fig, 8). 
