vcith a Dead Beat Scapement and Deal Pendulum. 37 



I shall now advert to the valuable series of observations made 

 by Col. Beaufoy in 1819, 1820, 1821 and part of 1822, on the 

 going of a clock with a straight-grained deal pendulum rod. 

 From these observations I am led to believe that the small 

 and sudden irregularities in the rate of this clock are not re- 

 ferable to changes in the length of the pendulum, but to those 

 in the arc of vibration : for had the changes of the atmo- 

 sphere affected the length of the pendulum, whilst the arc of 

 vibration remained constant, the clock would have gradually 

 gained upon its mean rate as the summer advanced, and lost 

 as the winter approached, which does not seem to have been 

 the case from these observations. So that we may conclude 

 this pendulum rod to have undergone little or no change du- 

 ring the period of three years and upwards that the daily 

 rate of this clock was registered. 



But if we take the differences of each succeeding daily rate, 

 as given in the tables for the deal pendulum, and compare them 

 with like differences of the daily rate, as found by a subsequent 

 series of observations made also by Col. Beaufoy in 1824 and 

 1825, with a teak pendulum, we shall find that in the latter 

 year, though the clock gradually gained from the beginning 

 of March, when its daily rate was about two seconds, to the 

 6th of August, when its daily rate was 9-5 seconds ; yet the 

 viaxmiim difference of the rates with the teak pendulum, only 

 exceeded that of the deal pendulum '31 of a second. 



From the great increase of the daily rate (as observed above) 

 durino- the spring and summer months, we must infer that 

 there was a contraction of the teak pendulum, and which came 

 to its maximum about the beginning of August : nevertheless 

 the rate was not much affected by this cause, from day to day, 

 as the second differences are not found to be more dissimilar 

 than those obtained from the observations made with the deal 

 pendulum. 



As it is possible to render the deal pendulum peifectly con- 

 stant under all changes of the atmosphere, — which is all that is 

 wanted in this respect, for the purpose of accuracy ; it is there- 

 fore unnecessary to have recourse to the more expensive mode 

 of compensation, which from the uncertainty in the ratio of 

 expansion of the different metals of which the gridiron and si- 

 milar pendulums are composed, together with the nicety re- 

 (juired in the adjustment of the different parts, so as to insure 

 a perfect balancing of effects in the centre of oscillation, must 

 be a work that recjuires more time and attention for insuring 

 a recjuisite degree of accuracy than is usually bestowed on 

 such occasions: whilst after all, if the arc of vibration be in 

 a stale of ffuctuation, a doubt may arise whether lo ascribe 



the 



