ON A QUARTZ THREAD GRAVITY BALANCK. 233 



would be most noticeable after extreme temperature changes. When we come to 

 discuss the observations we shall see that there appears to have been some effect of 

 the kind. 



Thermal Insulation. 



It has already been mentioned that the copper tube, which forms the inner case of 

 the balance, is surrounded by a packing of paper, about half-an-inch thick, over which 

 the outer brass tube fits. It is clear that heat is able to enter or leave through the 

 metallic ends, the sextant part being especially difficult to insulate. We, therefore, 

 put the whole affair into a box of thin copper sheet and packed the interspace with 

 cotton wool. It is, however, obvious that the sextant arm and the microscope must 

 project, and, consequently, it is quite a question whether it is really worth while to 

 insulate the other part so carefully as we have tried to do. 



Packing and Transport. 



The balance in its copper box is lifted by its movable handles on to a tray provided 

 with mild steel handles, extending upwards above the top of the copper lx>x. The 

 tray carrying the balance is then lifted into a pine box which it just fits. Rigid 

 connection between the machine and the box is secured by means of two hard wood 

 strips which slip and dovetail into notches at the top of the box. When the copper 

 box is in position the wooden strips lie exactly above it, and there are brass screws 

 which pass from the strips into the frame of the balance inside the thin copper box, 

 which is not strong enough in itself to form a proper connection. In this way all 

 relative motion of the instrument and the box is avoided. The riding level is taken 

 off the balance and screwed to the box by special screws. The pine box containing 

 the balance is supported on a set of sofa springs, which are attached to a false bottom, 

 and some side support is given to the box by certain iron rods which are attached to 

 the false bottom and extend upwards, so surrounding the box by a kind of iron frame- 

 work. Connection between the box and the framework is secured by means of rubber 

 buffers, and in this way . the box is prevented from swinging about on its spring 

 bottom. 



The balance is handled by means of two handles screwing into the framework 

 through two holes in the copper lid of the outer box. 



THEORY OP THE BALANCE. 



The tension of the spring which stretches the thread is so large in comparison with 

 the weight of the lever that the thread is very nearly straight ; we will suppose that 

 it is exactly so. 



Let HI (fig. 1, p. 218) represent the plan of a thread lying in a horizontal plane with 



VOL. CXCHI. A. _ ii 



