On the Decomposition of Light. 915 



narrative of which I shall pass over for the present; merely 

 observing, that Newton did liot make his division 6n a 

 spectrum thus narrowed, but on one much larger, obtained 

 in the usual way without a lens. Opt. I, part 2, prob. I. 



Lastly I shall observe, that the green colour in our spec- 

 trum did not extend quite to the middle of its length, whence 

 it follov/ed, that the shades between the green and red 

 were a little shortened, and those of the blue and violet 

 proportionally elongated. These effects were owing no 

 doubt to the nature of the flint glass, of which our prism 

 was made. We had no opportunity of procuring common 

 glass free from streaks. Having tried a hollow prism, 

 formed of glasses joined together, and filled with water ; 

 the faces of the glasses occasioned duplications of the spec- 

 trum, which rendered it confused ; so that we returned to 

 our English flint glass, which, while perfectly void of 

 colour, combined homogeneity of substance, and accuracy 

 of structure, with the finest polish ; in short it was to all 

 appearance free from defect. 



It now remains for nie to speak of the particular ex- 

 periments on the analysis of colours, which I had long 

 planned. 



The reader may recollect, that I had suspected the blue 

 to be merely the result of a combination of green and violet 

 rays; and that in like manner the yellow proceeded only 

 from a mixture of green and red. I reasoned then in this 

 manner: on the supposition that in reality there existed no 

 simple rays of blue, if we prevent the arrival of rays to 

 that part of the spectrum, either by a substance that suffers 

 only the green rays to pass, or by one that allows a passage 

 to the violet only, we shall find beyond these substances 

 only green or violet; otherwise, supposing the blue rays to 

 be simple, they will traverse neither of the substances I 

 have mentioned, as we shall find beyond them nothing but 

 black. We may reason in a similar way with respect to 

 the yellow, v^hich must he subjected to the tiial of a red 

 substance and a green. 



Thus we must be furniihed with three substances coloured 



in the requisite manner. For the violet I employed an 



O 4 aramoniacal 



