Instrumental Error of his Horizon-Sector. 343 



cond telescope at the northern extremity of the plank as a 

 counterpoise to the one at the opposite end, and introducing 

 a narrow bar of wood between the table and the plank, on 

 which the latter would altogether rest without touching the 

 table, the cross of the threads of the telescope, which had 

 been previously set to bisect the horizontal wire of the cy- 

 linder, was found to have started from it to the amount of 

 several seconds. 



P7-ocess of Measurement. — The line of coUimation of the 

 proper object-glass of the cylinder having been made pa- 

 rallel to the axis of the latter (which was easily and most 

 accurately effected by the cross threads of the telescope), it 

 was then made parallel to that of the telescope by elevating 

 or depressing the cylinder by means of the rack-work of the 

 stand, rather than by the more expeditious, yet uncertain ap- 

 plication of the micrometer, which, with a view to insure its 

 steadiness, had been long undisturbed. In the next place, 

 by reversing the cylinder within its Ys, the additional object- 

 glass was brought close to that of the telescope, when its line 

 of collimation was found to be so nearly parallel to that of the 

 telescope, that it could be rendered level with it by unscrewing 

 a little the additional object-glass, which was not well centred, 

 within its cell. This effected, the cylinder was inverted within 

 its Ys, and on looking through the telescope, the intersection 

 of the wires of the cylinder appeared above the level, and con- 

 siderably to the right of the point of intersection of the threads 

 of the telescope. To measure the difference of level of these 

 two points of instersection, evidently equal to the quadruple of 

 the instrumental error, both the great levels were read off; 

 then the horizontal wire of the cylinder being brought by the 

 rack-work of the stand to the level of the intersecting point 

 of the threads of the telescope (by which it appeared to be 

 bisected), the reading off of the levels was repeated. Ex- 

 ample, February 9th, 1830. 



Cylinder reversed ; additional object-glass and that of tele- 

 scope together. 



Bubble. 



Right-hand level (scale, 1° = l"-75) 46^°; 136°^ 



Left-hand level (scale, 1° = 1 -92) 50 ; 141 



The line of collimation of the telescope level with the hori- 

 zontal wire of cylinder. Bubble. 



Right-hand level 19^^; 109°| 



Left-hand level 25| ; 116^ 



To confirm the measurements the cylinder was again re- 

 versed, which brought the proper object-glass of the cylinder 



close 



