ALTAZIMUTH WIRES, COLLIMATION, LEVEL ERROR, AZIMUTH ZERO. 415 



spirit-level at every observation of azimuth. The value of the graduations 

 engraved on the glass bubble, determined by the makers, was 60 to one 

 minute of arc. The Error of Level is corrected for the inequality of the 

 pivots of the horizontal axis. This inequality was determined by Mr. Sid- 

 greaves, by reversing the axis eight times on 1875, January 2, to be 0"'52, 

 and by an equal number of reversions on January 5, 0"'48, the smaller pivot 

 being on the same side as the lamp. The correction applied to the level- 

 error on this account is 0"'50, positive with lamp left, negative with lamp 

 right. 



Mr. Sidgreaves, by a number of suitable levellings, determined that the 

 pillar supporting the axis on the lamp side was shorter than the pillar on 

 the side of the micrometers by O00030 of an inch, causing a constant 

 difference in the level-error lamp R. and lamp L. of 12" % 6, a circumstance 

 which caused much trouble when observing with the instrument. After 

 leaving the makers' hands it had been carelessly taken to pieces. 



Nearly the whole of the observations of the Moon in azimuth, for the 

 determination of the longitude of the station, were made by Mr. Sidgreaves, 

 who, whenever possible, made one or more observations of stars with large 

 zenith distances. The local sidereal time determined by one observer with 

 the transit instrument was transferred to the Altazimuth-clock, and used 

 generally for the azimuth observations by another observer ; but the adoption 

 of a zero of azimuth determined by a star at the same time and with the 

 same clock-error practically eliminates a systematic error that might thus 

 arise, and the observations become strictly differential. 



The azimuth observations of the Moon should be divided into two classes 

 those accompanied by a star, and those without. The former present no 

 difficulty ; the zero of azimuth derived from the star (or stars) by an observer 

 is applied to his observations of the Moon, thus eliminating systematic error. 

 In the treatment of the latter class, which cannot be made perfectly satis- 

 factory, the observations of the distant azimuth-mark have been taken as 

 affording the best zero of azimuth obtainable ; the absolute azimuth of 

 the mark being taken as 235. 41'. 8"'7, which is the mean of 23 deter- 

 minations. The distant mark was a sharply-defined edge of rock 

 projected against the sky. It was always observed by Mr. Sidgreaves 

 himself. 



The horizontal transits of the Moon and stars were taken over the middle 



3 H 



