Ohservatixyiis on the Annular Solar Eclipse, 127 



pocket chronometer of the usual construction, making five beats 

 in two seconds. This train is not so convenient for estimating 

 fractions of seconds as a clock beating exact seconds, or a chro- 

 nometer making exactly two or three beats in a second, and 

 consequently I was unable to mark parts of seconds satisfacto* 

 rily. The chronometer was set to mean solar time a little be- 

 fore the commencement of the eclipse, and lost two seconds, as 

 nearly as could be estimated, during the continuance of the ob- 

 servations, when compared with a clock whose error had been 

 ascertained that day by direct observation. On allowing a pro- 

 portional part of this rate to each observation, fractional parts 

 have been introduced whose accuracy is not warranted to nearer 

 than half a second of the truth of the time of the phase intend- 

 ed to be recorded.* 



Having made these preliminary observations and explanatory 

 remarks, I shall now state the final results in mean solar time. 



u. M. s. 



The first contact of limbs took place at . 1 33 9.75 



Annulus completely formed, . . 2 57 21.12 



Annulus broke . . . . 3 1 11.03 



Termination of the eclipse. . . . 4 19 22.40 



From a mean of the times of the commencement and termi- 

 nation of the annulus, I infer that the time of greatest obscura- 

 tion took place at 2^ 29"* 16^07, and from the difference of 

 those times that the annulus lasted 3™ 49^91. It may be add- 

 ed, that all the observations here recorded having been taken 

 with equal care, and no oversight having occurred or accident 

 happened in the management of the instruments, they are all 

 entitled, in the opinion of the observer, to equal confidence. 



These observations, when compared with others made on dif- 

 ferent meridians, will give the difference of longitude between 

 those meridians with great accuracy, though for ordinary prac- 

 tice the calculations are rather long and difficult to persons not 

 much habituated to such operations. Taking the sun's and 

 moon's places as given in the Nautical Almanac, which are de- 

 rived from tables known to be liable to small errors only, the 

 longitude of places may be obtained to a considerable degree of 

 precision. On this principle, by the rules and formulae which 



• Different observers even at the same place will occasionally, in these 

 kinds of observations, differ some seconds. 



