CHAP. I.] HADLEY'S SEXTANT 7 



from which it must never be moved until the equal altitudes 

 or other observations are complete. No matter what depth of 

 shade is then used by shifting the dark eye-pieces, the two 

 images will be of the same tint. 



The darker the shade used the better. Beginners are very 

 apt to use too bright suns. 



If in observing with the sun the observer can accustom him- 

 self to use one eye for taking the observation, and the other for 

 reading and setting the vernier, he will find it very convenient, 

 and it will tend to keep both his eyes in good order. 



3. It is very convenient for picking up the images in the Position 

 artificial horizon, if the up-and-down piece is so placed as to ^own 

 enable the observer to look over it into the horizon glass. Piece. 



In many sextants the up-and-down piece is placed so close 

 to the index glass that this is not possible, and regard should 

 be had to this point. 



4. An interrupted thread, to screw the telescope into the 

 collar of the up-and-down piece, is a great convenience. 



5. An extended vernier, i.e. a vernier whose divisions are 

 twice the distance apart of those on the arc, will be found 

 convenient for accurate observing. 



6. A steel tangent screw will be found to last longer and 

 work more evenly than a brass one. 



The methods of ascertaining the index and other errors of 



Hadley's sextant, and correcting them, are so fully entered into 



by Heather, that they are here omitted, with the exception of 



the following remarks on the centring error : — 



This very important error of the sextant cannot be corrected Centring 



. Error 



in the instrument, and it requires a considerable amount of 

 labour to settle its quantity, which in an indifferent instrument 

 may be quite sufficient to vitiate the result of any observations 

 on one side only of the zenith. 



The centring error, pure and simple, arises from the non- 

 coincidence of the centres of the index arm and of the graduated 

 arc, so that the vernier does not move truly along the arc, and 

 the angle read off will not be correct. This error varies with 

 the angle, and is generally greater as the angle increases, but 

 the same result of error appears from the index arm becoming 

 bent ; from any part of the frame receiving a blow which alters 

 its shape ; from the flexure of the instrument from varying 



