( 345 ) 



XXI. Notice regarding a Time-Keeper in the Hall of the Royal 

 Society of Edinburgh. By JOHN ROBISON, Esq. Sec. R. S. Ed. 



(Read 1th February 1831.) 



J-HERE being some peculiarities in the construction of the 

 Clock lately set up in this room, which may prove to have consi- 

 derable influence on the performance of such instruments, and 

 also on their cost and duration, it is presumed that a short notice 

 of them, together with a few preliminary observations, may not 

 be deemed uninteresting to the Society. 



An eminent philosopher, in a work recently published, has 

 defined a clock to be " nothing more than a piece of mechanism, 

 for counting the oscillations of a pendulum." This definition 

 cannot be considered as complete, as besides having to register 

 the oscillations of its pendulum, a clock has to communicate suc- 

 cessive impulses to it, to enable it to overcome the friction of its 

 suspension, and the resistance of the air. If a maintaining power 

 were not exerted by the clock, the pendulum would soon be 

 brought to a state of rest. 



There are therefore two principal points which require to be 

 attended to in the construction of a good time-keeper. One, that 

 the pendulum shall perform all its oscillations in equal times, in 

 spite of the variations of temperature it may be exposed to ; and 

 the other, that the clock or mechanism shall communicate un- 

 varying impulses to the pendulum during long periods of time. 



