Instrumental Error oj hh Horizon-Sector . 101 



the other, their object-glasses being nearly in contact. A cer- 

 tain division of the pearl slip of the square telescope, as seen 

 through the round telescope, being made to appear to bisect 

 a dot or minute particle of dust adhering to the vertical wire 

 of the latter, (which was effected by moving gradually in the 

 proper direction a piece of thin paper introduced between the 

 Y stand of the round telescope and the plank beneath it,) the 

 exact position of the level-bubble of each telescope was marked 

 on its tube by a hair-pencil dipped in white paint of the tem- 

 perature of the room. 



Measurement by the Sector of the Inclination of the Lines of 

 Collimation of the Proof Telescojies. 

 In the first attempt, which brought out the instrumental 

 error 25", the sector was placed at the south end of the plank, 

 and the round and square telescopes were stationed in suc- 

 cession at the other extremity with a lamp beyond. However, 

 as neither the division of the slip nor the dot on the wire could 

 be satisfactorily bisected by the horizontal wire of the sector, 

 the latter, deprived of its eye-tube, was removed to the north, 

 and white pasteboard set up beyond it in the place of the 

 lamp. One of the proof telescopes, for instance the square 

 one, being placed to the south in a line with the sector, their 

 object-glasses nearly in contact, and the pearl slip apparently 

 bisected longitudinally by the vertical wire of the sector, the 

 bubble of Fortin's level was brought to its mark by the re- 

 quisite alteration of the inclination of the square telescope on 

 which it was mounted. The hoi'izontal wire of the sector was 

 now made level or in a line with the (horizontal) division of 

 the pearl slip by the rack-work of the sector-stand. This 

 effected, the great level of the sector then uppermost was 

 carefully read off, and the cylinder turned halt-round within 

 its Ys, when the wire and division of the slip were again made 

 coincident, and the other great level read off. With these 

 data and the reversing points of the great levels, the inclina- 

 tion of the line of collimation of the square telescope could be 

 readily comjiuted. 



The round telescope being substituted for the square one. 

 Mid moved laterally until its vertical wire aj)peared a few se- 

 conds to the right or left of that of the sector, its inclination 

 was varied until the bubble of its level stood at its marks. In 

 the next place, the horizontal wire of the sector was made to 

 bisect the dot on the vertical wire of the round telescope be- 

 fore and after the cylinder of the sector had been inverted. 

 These operations being followed by tiic corresponding read- 

 ing of the great levels of the sector, the inclination of the line 

 o\ collimation of the round telescope could be obtained. 



