[ 34 ] 



that of the central rays ; hence the incidence of the exterior 

 rays on the fecond furface may be made lefs than that of the 

 central rays, and therefore the refradion of the former ray 

 into air may be lefs than that of the latter ; that is (fince thefe 

 refraflions caufe a convergence) the divergence of the exterior 

 rays is lefs diminifhed than that of the central, therefore 

 the exterior rays will for both reafons diverge more than the 

 central. 



This, it will be faid, is equally cffedled by a double concave; 

 but if the curvature of the firft furface in the concavo-convex 

 be confiderable in refpe£t of that of the fecond furface, it may 

 be brought to pafs that the incidence of the exterior rays on the 

 fecond furface fhall be of a different affedion from that of the 

 central rays, and therefore that the refraflion of the fecond 

 furface may augment the divergence of the former, while it 

 diminiflies the divergence of the latter, and thus caufe a greater 

 divergence of the exterior rays than can be efFedted by the double 

 concave, while the divergence of the central rays caufed by both 

 is the fame. 



These expedients feem capable of correding the unufual 

 errors of fhort fight, that cannot be removed by any double- 

 concave glafs, if they fhould be occafioned by the caufes that 

 I have fuppofed, i. e. by confiderable errors of fphericity. 



And it feems probable, from what has been faid, that fuch 

 dcfcd is occafioned by thofc caufes, where perfons of the above 

 delcription are affifted in vifion by looking through a pin-hole, 



and 



