294 KEPOET— 1893. 



wheels, P and Q, with V-shaped grooves, which turn stiffly on their axles. 

 This part of the instrument will be referred to as the ' frame.' The thin 

 silver wire which supports the mirror passes partly round each of these 

 wheels, and is fastened by a small screw in their edges. The wire hangs 

 down in a loop just below the frame, and is stretched tight when the 

 mirror is hooked on. The north portion of the wire is attached to the 

 upper wheel, and the south portion to the lower. A convenient vertical 

 adjustment is given to the mirror by turning the pulleys and winding 

 some of the silver wire on to them ; the friction with which they turn 

 prevents them running back. 



When the mirror is suspended in this manner the frame and mirror 

 are inserted into the brass tube, the lower end of the frame resting on 

 the hollow conical surface, R, of the passage from the tube to the mirror- 

 box, so that the mirror hangs within the latter. A rectangular opening 

 in a circular brass plate, d (fig. 2), at the top of the tube admits the 

 frame and allows it to move without rotation in an east and west direc- 

 tion. A thin rod, which passes through a hole in the south-east side 

 near the top of the cap of the tube, is kept pressing by a spiral spring 

 against the bar H (fig. 3), and this holds the frame against the north 

 side of the rectangular opening, and also against the end of the screw, e 

 (figs. 2 and 3), on the west side. Thus, all movement of the frame 

 inside the tube is prevented, except that in an east and west direction, 

 produced by turning the screw, e. Another and much stronger spring 

 outside the cylinder keeps the screw e pressed in one direction in order 

 to prevent any backlash in the screw. 



Before the plate, d, and the cap are screwed on, the mirror-box and 

 tube are filled with paraffin oil, so that the pendulum and supports are 

 entirely immersed in it. The effect of this is to make the pendulum 

 absolutely dead-beat and capable of registering slow earth-tilts only. It 

 is therefore unaflTected by the rapid tremors such as would be produced 

 by the rattling of passing vehicles. 



The two west or front levelling-screws being equidistant from the 

 east and west line through the centre of the instrument, it is evident 

 that the inclination of the pendulum along the latter line is not changed 

 if one of the screws be raised and the other depressed by equal amounts. 

 The sensitiveness of the pendulum can be increased or decreased by 

 turning the screws in this manner. At present this is done by hand, 

 but it might be desirable to put a slow motion for their adjustment 

 which could be worked from a distance. The inclination of the pen- 

 dulum in the east and west direction can be altered by the east or back 

 levelling-screw (called the 'back-leg' in the Report of 1881) without 

 affecting the sensitiveness, and a movement of the same kind but of far 

 greater delicacy can be given to the frame by the screw, E. This can be 

 done by hand, or a small turn of the screw can be made by the lever, f. 



The distances between the levelling-screws being known, and also 

 the number of threads to the inch, it is evident that the tilt given to the 

 instrument by one turn, or fraction of a turn, of the back-leg could be 

 estimated. In the present form of the instrument the tilt is, however, 

 given to the frame by turning the screw, e, through a known angle. If 

 a be the distance between the bottom of the frame and the point at 

 which the screw acts, h the breadth of the frame, and a the semi- 

 vertical angle of the hollow cone, E, on which the frame rests, a move- 

 ment of the screw, E, through a distance, h, will tilt the frame through an. 



