C '20 3 



already well adjufled by a diftant object, and you do- 

 not chufe to alter it by adjufting it by a near one, 

 move the index, and bring the image of the near 

 direct object to coincide with the fame feen directly, 

 and the number of minutes by which (o) of the 

 index ftands to the right hand of (o) of the quadrant 

 upon the arch of the excefs is the correction, which 

 added to the angle meafured by the arch of the qua- 

 drant between this direct object and any other object 

 feen by reflection will give the true angular diftance 

 between them reduced to the centre of the index- 

 glafs. 



To find the angular diftance between two near objects 

 by the back-obfervation. 



It is fuppofed that the-horizon-glafs is truly 

 adjufled ; if it is not, let it be fo. Obferve the 

 diftance of the objects by the back-obfervation, and 

 take the fupplement of the degrees and minutes 

 ftanding upon the arch to 180 degrees, which call 

 the inftrumental angular diftance of the objects ; this 

 is to be corrected as follows. Keep the centre of the 

 quadrant or index-glafs in the fame place as it had in 

 the foregoing obfervation, and obferve the diftance 

 between the near object, which has been juft taken 

 as the direct object, and fome diftant object, twice; 

 by making both objects to be the direct and reflected 

 ones alternately, holding the divided arch upwards in 

 one cafe and downwards in the other, ftill preferving 

 the place of the centre of the quadrant. The 

 difference of thefe two observations will be the cor- 

 rection, which added to the inftrumental angular 



diftance, 



