appear in the middle oft, through which the object C is 

 seen directly. As the whole card covering the mirror a, is 

 seen equidistant from the extremities of b, every point in the 

 surface of a, and consequently every ray reflected from such 

 point, must be in the same manner seen to preserve their re- 

 lative positions, and as the picture of a seen in b, should 

 occupy nearly its whole surface ; the extremities of a, or any 

 rays reflected from such extremities, could not be seen in the 

 centre of b; but if the objects C and c, being small, could 

 not subtend at the eye so great a space as the whole mirror 

 a, the image of c would cover but a small part of the image 

 of a ; and if that image proceeded by reflection not from 

 the centre, but the extremity of a, it would be visible in the 

 extremity of the image of a as seen in b; i. e. at a distance 

 from the centre of b, (and consequently remote from the 

 image of C ;) if it were seen in the centre of 6, it must be 

 reflected from the centre of a; but if the whole surface of 

 a were not apparently coincident with that of b, this might 

 not be the case. 



Hence appears the justness of Mr. Ludlam's direction^ 

 that the centres of both mirrors should be seen to coincide 

 in the horizon glass with the object seen directly ; for the 

 images can appear thus, in both positions of the quadrant 

 coincident only under a certain and invariable position of 

 the specula, though their whole surfaces were uncevered ; 

 it is hard however to distinguish by the eye what is the 



