ON A QUARTZ THREAD GRAVITY BALANCE. 219 



E is a microscope attached to the frame of the machine so as to occupy an invariable 

 position with respect to the supports A and B. 



The microscope is provided with a riding level showing 2" of arc per division of 

 2 '5 millims. at ordinary temperatures. 



The whole apparatus is provided with levelling screws, so that the axis of collimation 

 of the microscope can be brought into an invariable position by means of the riding 

 level, which invariable position is very nearly or quite horizontal. Also the thread 

 itwelf can be adjusted to lie in a horizontal plane by means of a subsidiary level, which 

 is j>ermanently attached to the frame work, and is at right angles to the riding level. 



Tho appiratus is thermally insulated so far as possible, and a platinum wire 

 thermometer wound upon very thin glass lies alongside the thread. 



There is also an arrangement for arresting the end of the lever for security during 

 transport. 



The gravitational moment of the lever about the thread requires to be adjusted by 

 the addition of small drops of fusible metal solder. The moment is so adjusted that 

 about three whole turns of each end of the thread are required to keep the lever 

 horizontal. This adjustment is made once for all ; during the process of observing 

 the thread is twisted from one end only, viz., the end attached to the sextant arm. 

 It will be shown directly (in the section dealing with the theory of the instrument) 

 that the equilibrium of the lever becomes unstable when its centre of gravity rises a 

 few degrees above the horizontal plane passing through the thread. 



The microscope is so arranged with respect to the lever that, when the riding level 

 is horizontal, the image of the end of the lever lies on the cross wire ; and the lever 

 itself is almost, but not quite, in the position of instability. 



By turning the vernier arm it is possible to increase or diminish the twist of the 

 thread from I up to the lever. When the thread is increasingly twisted, the side of 

 the lever on which is the centre of gravity rises until the position of instability is 

 reached and the lever upsets ; it is caught, however, by the arrester, and is not 

 allowed to fall right over. When the intensity of gravitational force increases, the 

 centre of gravity of the lever falls, and the thread has to be twisted by the vernier 

 arm in order to bring the lever back to its former position. 



The immediate subject of observation is the amount of twisting or untwisting 

 necessary to bring the lever to its sighted position. 



In order to calibrate the instrument, it is necessary to know the total twist of 

 the thread, that the thread is uniform, or at all events of a definite shape, the 

 position of the centre of gravity of the lever, and the exact position occupied by 

 the lever with respect to the ends of the thread. As these quantities are not exactly 

 ascertainable, we prefer to calibrate the instrument by observing the change of 

 reading which occurs when it is taken from Sydney to Melbourne the gravitational 

 data of both of these places being sufficiently well established. 



It appears from the theory below that this information, together with the vernier 



2 F 2 



