36 A MANUAL OF TOPOGRAPHIC METHODS. 



means reduce all these astronomical determinations to one point, thus ob- 

 taining for this point a number of astronomical determinations each having 

 a different station error. The mean of these gives for this point a position 

 from which — in part, at least — station error has been eliminated, and this 

 mean position can be transferred back by means of the triangulation to the 

 several astronomical stations, thus giving each of them a position similarly 

 comparatively free from station error. 



OBSERVATIONS FOR AZIMUTH. 



The initial direction from which the directions of other lines in primary 

 triangulation and in primary traversing are computed is obtained by means 

 of astronomic observations. Such observations should be taken not only 

 upon the initial line, but at intervals throughout the work for its verification. 

 Such intervals should not exceed in the primary triangulation 100 miles, and 

 in primary traversing 10 to 20 miles. 



Azimuth observations are made with the theodolite used in primary 

 triangulation or traverse. The observations consist in the measurement of 

 the horizontal angle between some close circumpolar star, usually Polaris, 

 and a terrestrial mark, generally a bull's-eye lantern set at a distance of half 

 a mile to a mile from the observing station. The time of observation on the 

 star should be noted by a chronometer or a good watch. As the star is at 

 a much higher angle of elevation than the lamp it is necessary not only to 

 level the instrument carefully but to measure the error of level and to cor- 

 rect for it. It is therefore essential that the value of a division of the level 

 bulb be known. These observations for azimuth may be made at any time 

 of the night, but preferably they should be made at or near the time of 

 elongation of the star, as it is then moving most slowly in azimuth, and any 

 error in the time of observation has the least effect upon the resulting azimuth. 

 If such observations In- taken at elongation, no record of time need be made, 

 and the reduction of the observations is simplified. When such observations 

 are made at any other time than at elongation, the time must be noted, as 

 it forms an element in the reduction. The error of the clock or watch used 

 may be obtained by comparison with railroad time, and corrected for the 

 difference in longitude between the station and the meridian of the railroad 

 time. A form of observation and record is appended. 



