Finding the Longitude from Lunar Distances. 91 



tables. The corrections for excentricity have not been applied ; 

 the results, therefore, are those given by a good sextant used 

 in the ordinary way, with careful reduction. The local time 

 has been taken from a chronometer watch, compared with 

 the mean time clock of the Observatory. On a few 

 occasions, it was taken from a clock beating seconds, whose 

 error was determined in the same manner, and in only 

 one instance is there a doubt about this element. The 

 clock was used in the observation of October 14, 1888 — it 

 will be noted as the most discrepant but one of the series — 

 it was reduced soon after it was observed and the discrepancy 

 remarked, and on examining the clock, I found that in that 

 part of the dial where the minutes were read, the zeros of 

 the second and minute hands did not correspond, and I have 

 reason to believe that a mistake of a minute was made in 

 the local time; as I have, however, a great dislike to cooking 

 observations, I have not changed the original entry of the 

 observation. As the altitudes are in ever}^ case computed, a 

 mistake in the mean time would affect them also. 



