18 
three to seven, like those of a rainbow *. As I admit more light, of 
make tlie marks broader and larger, the colours become brighter 
and more separated, till white is admitted in the broadest, when it 
will appear that the white, with yellow, red and blue, only remain, 
and perhaps a little mixture, between the bright red and yellow, of 
orange-1-, and the yellow with the light or white, naturally blend 
into each other, and the blue is also softened into the white termi- 
nating at the lower edge. Having thus produced the coloured rays 
by a little light and much black, it may be proper to try a contrary 
process, by a little black and much white. We should therefore 
look at the middle of a piece of white pajier, or any white object, 
when if smooth, and free from spots or marks, we shall perceive no 
colour; but if a fine stroke of black be produced, pale tints or colours 
will be seen ; but if more black is added, the colours will be more 
distinct in a certain proportion. The order of the colours will com- 
mence as it were with the penumbra of the former experiment oa 
light, as light blue above passing downwards into indigo and violet, 
the white immediately beneath partaking of the red and yellow, as in 
the upper part of the white line, and the blue will not recommence 
till something terminates the space of white, as a wrinkle or shadow; 
the margin of the paper or book, a stroke with a pen, &c. producing, 
(and we have power of continuing) the same phaenomena at plea- 
sure. It was just observed, that if we added more black the colours 
would be more distinct in a certain proportion ; I may add, that they 
will be more vivid and apparently more perfect. 
Ultimate perfection not being in the power of mortals, it is well 
if we are allowed to attain to a certain degree of it ; and it is surely 
a hi<Th satisfaction to arrive at any certainty within our limited 
spliere. If proportion therefore of black will naturally or artificially 
discover the quality or quantity of the coloured rays, finding this 
* The rainbow mostly shows the same set of tints more or less brilliant. 
■\ The three tints or colours may be made to mix and even to form browns, by turn- 
hig the prism on its axis so much as to show a lengthened range of colours. 
