VOL. XLVII.] PHILOSOPHICAL TRANSACTIONS. 273 



end of the lever, and k the point where the screw acts on the longer end of the 

 lever, which being the place where it intersects the line ef, it is evident the ball 

 of the pendulum will be as much raised by the lever, as it would have been de- 

 pressed by the expansion of the iron ; but the triangle ihg is similar to the triangle 

 bgk ; and therefore as ih, the excess of the expansion of the brass, is to bg, the 

 whole expansion of the iron, so will hg, the shorter arm of the lever, be to gk, 

 the longer arm of the lever. Q. e. d. 



At Fig. 1 is placed a strong double spring, whose ends pressing against the 

 under edge of the ball, hinder it from bending the brass bar by its forcible action 

 at the point b, which, when the ball is of a considerable weight, it might other- 

 wise be very liable to do. 



The description here given is exactly agreeable to the original contrivance; and 

 the only alteration he afterwards made in it, consists in placing the screws g, g, 

 within the ball of the pendulum, as represented in fig. 4. But as the success of 

 this contrivance depended entirely on the supposition that metals were expanded 

 differently by the same degree of heat, before putting it in execution, he inquired 

 what experiments had already been made on this subject: when Mr. John Eames 

 put into his hands Mr. Graham's account of his quicksilver pendulum, as it is 

 now commonly called, published in the Philos. Trans. N° 3Q2, which account 

 was introduced by the following paragraph : 



" Whereas several, who have been curious in measuring of time, have taken 

 notice, that the vibrations of a pendulum are slower in summer than ip winter; 

 and have very justly supposed this alteration has proceeded from a change of 

 length in the pendulum itself, by the influences of heat and cold on it, in the 

 different seasons of the year; with a view therefore of correcting in some degree 

 this defect of the pendulum, I made several trials, about the year 1715, to dis- 

 cover whether there was any considerable difference of expansion between brass, 

 steel, iron, copper, silver, &c. when exposed to the same degrees of heat, as 

 nearly as I could determine; conceiving it would not be very difficult, by making 

 use of two sorts of metals diflfering considerably in their degrees of expansion 

 and contraction, to remedy, in great measure, the irregularities, to which com- 

 mon pendulums are subject. But though it is easily discoverable, that all these 

 metals suffer a sensible alteration of their dimension by heat and cold ; yet I found 

 their differences in quantity, from each other, were so small, as gave me no 

 hopes of succeeding this way, and made me leave off" prosecuting this affair any 

 farther at that time." 



The reading this paragraph proved at that time sufficient to make him lay 



aside all thoughts of succeeding in a contrivance founded on principles, which a 



gentleman of so great abilities, and known accuracy in making experiments, had 



after trial judged to be insufficient. And it was not till about the latter end of 



VOL. X. Nn 



