Page:Popular Science Monthly Volume 22.djvu/173

Rh (k) slides freely on the crutch-rod shown back of it in the figure, but is held by the screw on the end of the spindle (l) which bangs from the nut (m) at the crutch-axis. By turning the nut (m) the weight (k) can be lowered or raised, and this makes the clock gain or lose.

But the nicety of the correction of variations due to changes of temperature has brought to light variations due to another cause commonly quite overlooked; it has been found that the pendulum is affected by changes of barometric pressure. A change in the barometer of an inch and a half will sensibly alter the rate of the pendulum. The difficulty might be avoided by placing the clock in a vacuum, but this is evidently impracticable. In the Greenwich clock the method shown in Fig. 4 has been adopted to counteract the effects



of barometric changes. To the pendulum-bob are attached two vertical bar-magnets, one in front (a) with the north pole down, the other at the back (and therefore not shown in the figure), with the south pole down. Below these and normally at a distance of 3$3/4$ inches from them is a horseshoe magnet (b) which hangs on one end of a lever (c) nicely balanced on knife-edges at A; the other end of the lever (c) rests by means of a rod (d) on a float (e) in the shorter leg of a siphon