REFRANGIBILlTr of LIGHT. 49 



concave lens to remove the colour entirely. The colour may be 

 totally removed, and the aberration from the figure corrected, 

 by a concave which lengthens the focal diftance of the convex 

 only one third. 



From what hath been explained refpecting the total correc- 

 tion of colour, it will be underftood, that if the concave 

 lengthens the focal diftance beyond what is required, fringes 

 of green and purple ought to begin to appear in an inverted 

 order. This, which may be ftyled the expcrimentum cruets in 

 this matter, I now had it in my power to try without difficulty. 

 The refult turned out exactly as I expected. Upon applying a 

 compound concave, which nearly doubled the length of the 

 compound convex, a fringe of green appeared within the focus, 

 and a fringe of purple beyond it, which fets the theory of the 

 correction of this fecondary colour in the raoft fatisfactory 

 light. 



The compound concave in this and all the preceding experi- 

 ments, was formed of glafs and an eflTential oih 



I now happened, merely with a view of diverfifying the 

 experiment, to apply a compound concave, formed of glafs 

 combined with the muriatic acid, which has been mentioned aS 

 a fluid pofTefling a coniiderable degree of difperfive power. 

 This opened a new and unexpected fcene. The colours ap- 

 peared in the fame order as in the laft experiment, but the 

 fringes were fo very broad as greatly to furprife me, and create 

 a fufpicion that every thing was not as I had hitherto taken for 

 granted. Without delay I included fome of the marine acid 

 between two convex lenfes, whofe radii were duly proportioned 

 to the difperfive power of that fluid, for the purpofe of cor- 

 recting the colour. Upon applying an eye-glafs I found my 

 fufpicion verified. The fringes of green and purple appeared 

 nearly of the ufual breadth, but in an inverted order, there 

 being now a green fringe within the focus, and a purple fringe 

 beyond it. I was the better pleafed at being thus led to the 



Vol. III. G detection 



