278 On the stopping of Astronomical Clocks. 
scending eleven inches keeps it in motion a week. ‘The 
pendulum is supported by a strong upright bar of steel, 
which also supports the pully, over which passes the cord 
which suspends the weight. . 
This clock I found frequently to stop when the weight 
descended nearly opposite to the pendulum from which it 
was then distant about three fourths of an inch. This I 
attributed to the attraction of the weight and pendalum, 
and determined to remove the weight to a greater distance. 
Before this was done, however, on attentively observing 
the clock, | found that before it stopped, the weight acquir- 
ed a considerable oscillation, evidently communicated 
from the pendulum through the common support. 
To destroy these oscillations, I placed a perpendicular 
wire by the side of the weight, the whole line of its de- 
scent. On this a ring was fitted to slide, and was connect- 
ed with the weight by a delicate spring. This spring by 
constantly pressing the weight towards one side prevented 
the oscillations from taking place and completely effected 
the object, and the clock has since continued to go without 
any interruption. 
The experiment has been often repeated, of supporting 
two clocks on the same horizontal beam. If one be set in 
motion it will in a short time communicate its motion to 
the other, andif both pendulums be of the same length, 
they will continue their beats with perfect isochronism. 
This sympathy, which also exists between two watch bal- 
ances which are supported by the same plates, has been 
ingeniously used in practice by Breguet, who has con- 
structed clocks and watches, each including two distinct 
movements, having no connection together except by the 
plates which form the common support. In these time- 
keepers the pendulums and balances beat perfectly to- 
gether, and thus one pendulum or balance is made to cor- 
rect the irregularities of the other. 
If this experiment be made where the pendulums are of 
different lengths, the clock first in motion will be stopped. 
This effect can be transmitted through media apparently 
very solid, and probably has taken place in the instances 
when the clock’s stopping has been attributed to the at- 
traction of the pendulum and the weight; in such cases 
the weight suspended by its cord becomes a pendulum of 
