On the Hate of Clocks. 179 



of time, and might have done so /or hniv lovg I shall not pre- 

 tend to say ; had it not at last became necessary for me to put 

 an end to it. One of the clocks was a month one, with a re- 

 coiling scapement, and a compensation pendulum made after 

 Ward's construction. The other was an eight-day one. having 

 a detached scapement, and a zinc tube compensation pendulum, 

 both made some years ago, and of my own contrivance. 



During a period of four months previous to the middle of No- 

 vember 1813, they kept so constantly and close together, that 

 at no time could 1 perceive the tenth part of a second of difference 

 between them; indeed no sensible difference could be perceived, 

 either by myself or by another person who had occasion fre- 

 quently to see them expressly for this purpose. A pretty sharp 

 cold taking place on one of the nights in this November, the 

 month clock made a little deviation from the other: this was im- 

 puted to the infiuence of the cold upon the oil, as the arc of the 

 vibration of the pendulum was followed by being shortened a 

 little at this same time. However, very soon after, they again 

 went together, allowance being made for the difference or devia- 

 tion made by the one from the other, and this difference was 

 kept up to for more than a period of 1G4 days afterwards. The 

 mean daily rate of these clocks, during the four months previous 

 to the middle of November 1813, was -f 0-2 per diem. The 

 mean daily rate of them during a period of 164 days, viz. 

 from th'e 2d of February 1814 to the 16th of July, was also 

 -f 0-2_ per diem, the same as before. I shall give an abstract 

 more in detail, of their going during these 164 days, not with 

 the view of showing any great pe?fection either in their rates or 

 in the compensation ; yet at the same time it may be allowed 

 that in these respects, considering the incorrect state in which 

 they vyere, few clocks, even in their best and most correct 

 condition, have, if avij thing, much exceeded this. On the 2d 

 of February, the clock having the detached scapement was fast 

 by mean time l-" 35'. The montli clock was about 53 seconds 

 taster than it, and this difference it maintained during 164 days, 

 when the month clock was taken down to be corrected and to 

 get another pendulum. 



February 2. -f- 1-" 35' No.ofDays. Total Gain. ^^%" 'j"'^^ Mmnofthe 

 iv,f u "^ -r 1 ; X ^"''^- Thermo. 



March 7. -Hi 46 33 U .f.0-33 40 



^ 29. -1-1 50 22 4 .-fO-18 42 



May 11. -J- 1 55 48 5 -fOll 55 



June 10. +2 4 30 9 -j-0- 3 S"? 



July 16. -f2 10 36 6 -fO-16 GO 



The intermediate times of trial gave them for the greatest 



M 2 mean 



