CHAP. XII.] LATITUDE BY STARS 279 



from a direction as nearly at right angles to the plane of the 

 sextant as possible, to avoid parallax. 



A small electric light fitted to the sextant, as described on 

 p. 11, obviates the necessity for using a lantern, and is a great 

 convenience. 



If the horizon stand is raised off the ground by about a foot, 



and placed on a firm foundation, thus bringing the artificial 



horizon closer to the telescope, faint stars are more easily 



observed, and the movement of the sextant necessary to keep 



the star in the field, owing to its motion in the heavens, will 



be lessened. A lantern placed on the ground behind, or a 



little on one side of, the observer, and faintly showing on the 



artificial horizon, will sufficiently illuminate the wires of the 



telescope on a dark night. 



In taking the next observation, turn the tangent screw on Resetting 



. . for each 



or back alternately before commencing to bring the images observa- 



together again, so as to be entirely free from bias as to whether ^^°^- 

 the star is rising or falling, and also to compensate for small 

 centring errors which are liable to be introduced by the action 

 of the tangent screw pressing the centre, upon which the index- 

 bar turns, in opposite directions, according to the way in 

 which the screw is moved. 



The amount of time to observe before and after culmination Time from 

 varies with the position of the star and the latitude of the 

 place, as before mentioned ; but, as a rule, commencing six 

 minutes before the calculated time of meridian passage, and 

 continuing for a like time afterwards, will be ample, as we do 

 not wish the correction eventually to be applied to the mean 

 of the observed altitudes to bring it up to the meridian altitude 

 to be more than one minute if we can manage it. 



Observing as close to the meridian as is recommended above, Decimals 

 the decimal parts of seconds need not be recorded in taking gary. 

 the time. 



The observed altitude of a heavenly body can be corrected Reduction 

 to the meridian when the hour angle and the latitude are Meridian. 

 known ; but if the hour angle is large, the calculation of the 

 quantity to be added to the altitude is a complicated process, 

 and it is only when the hour angle is small, and very little in 

 error, that the formula assumes a simple and practical form. 

 The error introduced by working with an assumed latitude also 



