9lt' 



rect vision after a half turn of the instrument; the index 

 fixed at 0, and the back horizon glass shifted, till the images 

 are brought to coincide, whichsoever of them be viewed 

 directly. The quadrant is to be supported on a moveable 

 stand, on the points of two erected pins fixed on the stand, 

 inserted into two conical holes made in the middle of the 

 heads of the screw pins in the back of the instrument, which 

 fasten the central pins supporting the index and the back 

 horizon glass j the placing the respective glasses alternately 

 on these points, in the manner represented in fig. ]., will 

 veveree the quadrant, by giving it just a semicircular mo- 

 tion in its own plane. The manner of performing this ad- 

 justment has been fully described b}' Mr. Liicllam, to whom 

 I refer; but as I have seen no demonstration of its accuracy, 

 1 give the following proof of it ; assuming the established 

 optical principles. 



Let A P Q (fig. 1.), be the octant, fixed on two points 

 ^mder the centres of the index glass A, and the back hori- 

 zon glass B, or any two other fixed points ; and let C and c 

 be the two candles or objects by which the glass B is to be 

 adjusted. The image of the object C will be seen by the 

 eye E, looking through the sight vane, coincident with the 

 object c, when the stand of the quadrant is properly placed, 

 by the ray E B, parallel to A C, if the glasses are at right 

 angles. Let the quadrant now be turned half round, and 

 placed on the points in the position a p q ; and if the back 

 horizon glass is properly adjusted, then the eye -looking 



