MERIDIAN ON THE COAST OF COROMANDEL. 181 



During the time I was at Trivanckpor^umy near 

 Cuddalore, the weather was settled and serene, and 

 the nights perfectly clear, so that I had an unli- 

 mited choice of stars, but having been successful 

 with Aldtbaran, I chose tliat star for determining 

 the length of the arc. 



As I consider the celestial arc more likely to be 

 erroneous than any terrestrial measurement, I have 

 thought it necessary to give some account of the 

 manner of observing and of adjusting the instru- 

 ment, for after two years experience, I have found, 

 that notwithstanding the great powers of the zenith 

 sector, extreme delicacy and attention are requi- 

 site to render the observations satisfactory. The 

 following method of adjustment I have always prac- 

 tised. After having brought the vertical axis 

 nearly to its true position by the adjusting screw at 

 the bottom, or so that the wire of the plummet 

 would bisect the same dot when the telescope was 

 moved to the opposite side, or half round on the 

 axis, I then examined whether the dot at the cen- 

 tre of the horizontal axis was bisected, and whether 

 the wire moved in the vertical plane clear of the 

 the axis ; for unless it be perfectly free, all the ob- 

 servations will be false. When I had bisected the 

 dot, I either took out the microscope and looked 

 obliquely, or did the same by a magnifying glass, 

 and by that means I could discover the smallest pa- 

 rallax. If it admitted being brought nearer to the 

 axis, it was done; but I found from experience, 

 that it was more eligible to leave the wire at a sen- 

 sible distance than to bring it very near. Having 

 satisfied myself in this particular, I examined with 

 the microscope again in front, moved the wire freely 

 in the vertical plane, and then bisected the dot. 

 The telescope was then moved, so that the wire 

 was brought over the dot zero on the arc, and the 

 same precaution used with respect to the wire mov- 



N 3 



