17 



ceptible angle'; which Mr. Ludlam says will be, when thej 

 are removed at least half a mile oft": * and for the same rea- 

 sons, the adjustment cannot be made by the observer's turn-; 

 ing himself half round with the instrument, without revers- 

 ing it, unless the objects arc at a distance as great as this, 

 if it be not fixed on the same points, as above directed ; by 

 which alone the parallelism of the glasses is preserved, and 

 also the same incidences and reflections, which are only ex- 

 changed one for the other by a half turn of the instrument ; 

 so that when the horizon glass is rightly adjusted, the direct 

 and reflected images are reciprocally visible and coincident. 



By this mode of adjusting the back horizon glass, by placing 

 the quadrant on two fixed points between two near objects, a 

 contrivance is made practicable, of using with full advan- 

 tage the excellent method proposed by Mr. Blair of adjust- 

 ing it at all times, by placing it parallel to a reflecting plane 

 perpendicular to the index glass : for ascertaining which per- 

 pendicularity, the above mode of adjustment is necessary; as 

 Avithout knowing and making allowance for any deviation 

 from it, in all observations taken, they would all be erroneous ; 

 wdiich circumstance, as also this adjustment being the test of 

 the accuracy of the addition, which I am to propose to the 

 furniture of the quadrant, is the reason why 1 have been so 

 diffuse in the explanation of this method. 



The reflecting plane Mr. Blair proposed to be formed of 



t This depends on the magnifying power of the telescope, and the snjallness of the 

 ang. it will render disccrnable. 



VOL. XII. D 



