1 06 An Account of the Bepeating Circle, 



which makes one piece with it. This is the simplest action 

 that an instrument can possess ; the forces in opposition to 

 each other are quite inconsiderable ; there is no rubbing of 

 indices ; and the angles are read off immediately after obser- 

 vation without any thing having changed place. The error 

 of the level, mentioned before, will also affect this instrument ; 

 but, by judicious management, as will be shown presently, may 

 be almost wholly counteracted. The method indeed would 

 apply to R, but at an expense of time that could not be ad- 

 mitted. It is this : reversing the position of the circle in azi- 

 muth, by the level, place the axis truly vertical, which then 

 becomes a substitute for the level during the short time a star 

 is bisected ; after this observation is read off and the telescope 

 set for another star, I would note the position of the level, and 

 again reverse the instrument: when, if any deviation is seen, 

 I would bring the level by the foot screw half-way to the point 

 where it stood before. This operation places the axis vertical 

 again, whether a deviation of the level or of the whole instru- 

 ment has taken place. To do this in the easiest manner, is 

 to make a series of observations with the divided side of the 

 circle to the east and west alternately; this way saves half the 

 trouble of frequent reversion, renders observations independent 

 of each other, and makes it resemble in its use the zenith sec- 

 tor, as nearly as one instrument with a level can do a better 

 with a plumb-line. 



When A is used for ascertaining the value of a celestial 

 arc, corresponding to a measured distance on the surface of 

 the earth, man}- stars should be marked for observation, none 

 of which ought to be more than 60° from the zenith; and to 

 avoid hurried observation, they should be picked from a cata- 

 logue at nearly equal distances of right ascension. In dohig 

 this there can be no difficulty, because those of the fourth or 

 fifth magnitude are for this purpose as good as any others. 

 As to the time required for observing a star, different ob- 

 servers will want more or less, but surely the space of five 

 minutes is enoii";h for any one. If he has previously written 

 down, from his catalogue, the degree and first figure of the 

 minutes, he has no more to do in reading off than, when the 

 star is bisected, to put dov.'u the last figure of the minutes and 

 the seconds as he takes them from the micrometers. For the 

 next star he should set to what his catalogue gives, then re- 

 verse the position of the instrument by the divisions of the 

 azimuth circle, aftei'wards correct for any deviation of the level 

 and verticality of the axis, and then wait for the appulse of the 

 star to the meridian. In every pair of observations, the jilace 

 of a star is read off on four points of the circle ; and, where 

 those marked lor observation extend to 60° of zenith distance 



both 



