240 



The fundamental formulae employed by Dr. Robinson 

 are the following : 



l=e — w+Ri; 

 l=e' — w' — rY; 



l being the difference of longitudes between an eastern and 

 a western station ; e the correction of a watch when leaving 

 the eastern, and w the correction when arriving at the western, 

 while w' and e' are the corrections when leaving the latter 

 and returning to the former respectively ; i, i', the intervals 

 of time expended in thus going and returning ; and R, r', 

 the losing rates corresponding. By supposing r'zir, he 

 obtains what he considers the true travelling rate, namely, 

 _ (e'— w') — (e— w) 



R - T+? [ 



and the resulting longitude 



L = i fE'-w'+E-W+R(l-lO$ . 



The errors are then examined to which these determinations 

 are liable, and the numerical elements are given, from which 

 are deduced fifteen values for the longitude of Armagh, 

 varying between the extremes 26 m 34 s , 67, and 26 m 36 s , 32, 

 and giving as their general mean, 26 m 35 s , 44, with a pro- 

 bable error less than s , 1. 



Eclipses and occultations had given 26 m 35 s , 47 ; lunar 

 transits, 26™ 35 s , 64 ; and a few comparisons, made under 

 unfavourable circumstances with a single pocket watch, con- 

 structed by the late Mr. Sharp, had appeared to give 26™ 

 35 s , 09. On the whole, Dr. Robinson is not inclined to 

 change the quantity which he gave some years ago to Mr. 

 Stratford, for insertion in the Nautical Almanac, namely, 



+ 26™ 35 9 , 50. 



as the west longitude of Armagh from Greenwich. 



The same chronometric comparisons appear to require 

 that the value of the longitude of the Dublin Observatory, 



