5°. On the UNEQUAL 



detection of this lingular property of the acid of fea-falt, be- 

 caufe, in making the fame experiment before, this inverfion of 

 the order of the colours had entirely efcaped me. I was then 

 examining it, to find whether it difperfed the feveral orders of 

 rays, in the fame ratio in which glafs does ; and being fatisfied 

 that it did not, from obferving the green and purple fringes, as 

 in other combinations, a circumftance fo little looked for, as 

 the inverfion of the order of the colours, did not ftrike me. 



This obfervation affords a remarkable exception to what I 

 had begun to confider as very probably a general law of na- 

 ture. In the refraction which takes place in mediums of the 

 leaft difperfive kind, the green rays, or rather perhaps the rays 

 in the confine of green and blue, are the mean refrangible, and 

 thefe fame rays, in the more difperfive mediums, were always 

 found among the lefs refrangible rays ; and hence when, by a 

 proper combination of two fuch mediums, the red and violet 

 rays are united, thefe united red and violet rays conftitute the 

 leaft refrangible rays, and the green conftitute the moft refran- 

 gible rays, as before explained. 



But in the muriatic acid, the cafe is juft the reverfe of this. 

 Then the green rays, which in mediums that difperfe the leaft, 

 are the mean refrangible, and which in effential oils and metal- 

 lic impregnations are found among the lefs refrangible, appear 

 amongft the more refrangible. Whence in fuch a combination 

 of the muriatic acid and an indifperfive medium as (hall, unite 

 the red and violet rays, thefe united red and violet rays emerge 

 moft refrangible, and the homogeneal green xays emerge leaft 

 refrangible, being juft the reverfe of what takes piace in com- 

 binations of crown-glafs with flint-glafs, or with effential oils, 

 or faturated metallic folutions. 



This unufual property of the marine acid does not, how* 

 ever, feem to admit of any immediate application to the im- 

 provement of optical inftruments. It is true that, inftead of 

 having recourfe to a compound concave for correcting the fe- 



condary 



