36 On the UNEQUAL 



is only a partial . correction of colour ; the differently refrangi- 

 ble rays cannot all be converged to one focus." 



The next method that occurred to me of determining the 

 point in queftion was more decifive. This was to obferve whe- 

 ther any of this green and purple colour appeared through the 

 mod perfect kind of achromatic object-glafs above defcribed, 

 and reprefented in the ninth figure, in which there is only one 

 refraction. This I found to be the cafe ; and therefore confi- 

 dered myfelf as in poiTeffion of fufficient authority for conclu- 

 ding, that the theory advanced by Mr Dollond, and gene- 

 rally received, was defective. For with the large aperture and 

 high power made ufe of in thefe experiments, the colour that 

 appears in viewing a bright object is not weak and hardly fen- 

 fible, but a beautiful bright purple inclining to crimfon and a 

 ftrong full green, and thefe in fuch a quantity as evidently to 

 be the obflacle to increafing the aperture of the object-glafs. 



This was the conclufion I was then led to, and which I have 

 found confirmed by numerous experiments made fince. But 

 before entering farther on the fubject, it will be neceffary to 

 explain what is meant by different mediums not difperfing the 

 heterogeneal rays of light proportionally. 



Let A B "and CD (Figures 14. and 15.) reprefent the fur- 

 faces of two mediums, both of which equally refract the mean 

 refrangible ray. This we fhaH fuppofe to be the green ray, 

 though, in this explication, it is not material which is called 

 the mean refrangible ray. The angles of incidence KGL, 

 MRN, will then be equal, and the angles of refraction of the 

 green ray HG^, PRy, will alfo be equal in both thefe me- 

 diums. 



Let one of thefe mediums CD exceed the other A B fo 

 much in difperfive power, as to make the difference of the 

 angles of refraction of the green ray, and extreme violet ray,, 

 in the medium C D, double of what it is in the medium A B > 

 that is to fay, the angle vKy double the angle vG^. Then 



if 



