Page 219

and five parts of verdigris, , composed a dun colour like that of a m6use; but there is
nothing in all this which militates against the explanation here given of the cause of
the coloured shadows of bodies; for even supposing that there did not exist in nature
any two bodies of such colours as to form perfect whiteness by their mixture; or, to
go still further, supposing that no two prismatic colours of the sun could form a
compound perfectly white; still the facts and reasonings here stated respecting the
mixtures of such colours as are called contrasts, are so near the truth, that they
furnish a satisfactory account of the appearances of the colours of the shadows which
we have been considering. The terms by which we are accustomed to denominate
colours, have not a very accurate or precise meaning, and particularly those terms
which denote colours that are known to be mixtures of others, as green, purple, and
orange: neither the prismatic green, nor the colour of any known green body, may,
perhaps, combine with red so as to make actually an accurate white, and yet the
existence or composition of such a green may not be impossible. The philosophical
reader will clearly perceive, that no argument of any weight can be drawn from
considerations of this sort against this theory of coloured shadows.
,3. Every one knows that red colours and yellow colours mixed together, in
different proportions, produce orange colours of various kinds; also that reds and
blues produce purples and violets; and, lastly, that blues and yellows produce greens
in great variety; but it is not so generally known that green, purple, and orange
colours, are as it were almost annihilated by mixture, and much improved by conti-
guity with red, yellow, and blue colours respectively.
The little diagram suggests all these things to the memory, and a great many
more of the same kind, and, therefore, must be extremely useful to the artist who is
endeavouring to produce certain effects by contrast, harmony, &c. but it should
always be carefully remembered, that it contributes nothing to the proof of any of
the truths here advanced; the proof rests upon the reasons given for each of them
,respectively."