48 
the more the colours are blended : they, however, are never perfectly 
unblended, if I understand the term right. 
"Examine also with the prism the coloured fringes of a white 
rectangular bod}', a little large, and well illuminated, you will remark 
on one side a red stripe, falling into the yellow abruptly, which, in 
its turn, loses itself in the white; and on the other hand some blue, 
distinctly enough between the white and a violet stripe* ; from this 
it would follow that there are only four sorts of colours." 
"Again, if we look at a white and narrow body, placed upon a 
black ground, such as a strijj of paper, a thread of silk, or a metallic 
needle, there will only appear, for one of these objects, if we are 
properly placed, three colours, red, green and violet, and scarcely 
can we perceive, either on the one or the other, any softening of the 
shades." See my Tab. 4. These answer in a great measure to a narrow 
line of light of Dr. Wollaston, which had blue and a little yellow. 
See Tab. 4. d. betwixt 3 and -1. 
Thus much resembles my Tab. 2. Jig. 3. made narrower, and the 
violet is more or less perceptible, as it is more or less relieved by a 
strong light or a dark back ground, especially if widened by the 
prism .And, upon examination, it will be found to be a mixture of 
binary tints, and accords well with Tab. 4, d. The lower part, if at 
all relieved or shining, will, according to precise laws, show brighter 
tints ; and 1 have further to observe that this is a mixture of the red, 
from the nature of the black ground above, with the blue and yellow 
of the lower black back ground : thus the red is purplish, and the 
yellow below the red, mixing with the other portion of the blue, 
forms the green &,c. It is not a little curious that C. A. Prieur, having 
so nearly hit upon the true mixture, should have let it pass. The con- 
cordance of my descriptions with his would almost look like a pla- 
giarism, as well with regard to him as with some parts of other de- 
* The '. !ue side of the white rectangular body was relieve 1 by some darkness, else no 
distinct violet stripe. How near to the true order of colours ! 
