APPENDIX. 



inftrument may be relied on, when there is no other 

 obfervation to check or corroborate. 



It may alfo be proper here to mention, that the time 

 by the watch was not obferved at the inftant that the fun 

 had returned to the vertical wire, and at which the pen- 

 dulum was obferved to fhow 4 i feconds, my attention 

 being wholly engaged in obferving the pendulum. The 

 watch was found to have loft 77" | by the pendulum, in 

 twenty- three hours, twenty-fix minutes, forty-two feconds 

 and a half. An allowance according to this rate for 

 34^ 4''^ (the fupplement of the laft obfervation by the 

 watch to the time of the fun's pafTage when the pendulum 

 fhewed 4"!) amounts to i"i. 



From whence it follows, that the Weft limb of the fun 

 touched the Eaft fide of the vertical hair at five hours, 

 twenty minutes, thirteen feconds and a half, by the 

 watch ; which had therefore loft four feconds in twenty- 

 four hours. 



As the comparifon of the watch and the pendulum in 

 this one inftance is not from adual obfervation, at the i?t'- 

 J}a?ttj but fuppofes that the watch had kept for thirty- 

 four minutes to the fame rate of lofing at which it had 

 been obferved to lofe for nearly twenty-four hours 

 immediately preceding ; the time by the watch thus 

 found is inferted in the table of obfervations within 



C c 2 hqoks 



169 



