MERIDIAN ON THE COAST OF COROMANDEL. 183 



ed, if necessary. That being- done, the degrees 

 and minutes, &c. on the arc were noted down, as 

 was also the particular division on the micrometer 

 scale, at which the index stood, and the fractional 

 part of a division in case there were any. In this 

 state every thing remained to within fifteen or 

 twenty minutes of the time the star was to pass, 

 when I repaired to the tent, and again examined 

 whether the wire bisected the dot; if it did not, the 

 instrument was again adjusted to the same dot, and 

 the horizontal axis also examined by the upper mi- 

 croscope, all this being done, the sector was placed 

 in the meridian. 



When the star entered the field of view, the mi- 

 crometer was moved gently till the star was near 

 the horizontal wire, but not bisected till it came 

 near the vertical, that the micrometer might not 

 be turned back, but continue moving in the s-ame 

 direction. This I did to avoid any false motion 

 in the micrometer screw, and I was led to this 

 precaution by the repeated experiments I had made 

 in examining the divisions on the arc, for it some- 

 times happened after moving the arc ovTr one of 

 the divisions till the wire bisected the next dot ; 

 and then turning it back again, that the index of 

 the micrometer was not at the same second, but 

 had passed over it perhaps one, and sometimes two 

 seconds ; but by moving over the next five minutes 

 in the same direction, the number of revolutions 

 and seconds were always what they ought to be, 

 to some very small fraction. This anomaly, how- 

 ever, only happened in some situations of the screw, 

 and to avoid any errors arising therefrom, I adopt- 

 ed the above method. 



The zenith distance of the star being now had, 

 on one part of the arc or limb, after the same pro- 

 cess had been gone through the next night, with 

 regard to the adjustment, the zenith distance was 



N4 



