If then thou art shading the white mantle, it must not be shaded with so
dark a colour as the red, for it would be impossible for a white thing
to yield so dark a shadow as a red. Neither could they be compared one
with another, save that in total absence of daylight everything is
black, seeing that colour cannot be recognised in darkness. Though,
therefore, in such a case, the theory allows one, without blame, to use
pure black for the shadows of a white object, yet this can seldom
come to pass.

Moreover, when thou paintest anything in one colour--be it red, blue,
brown, or any mixed colour--beware lest thou make it so bright in the
lights that it departs from its own kind. For example, an uneducated man
regardeth thy picture wherein is a red coat. "Look, good friend," saith
he, "in one part the coat is of a fair red and in another it is white
or pale in colour." That same is to be blamed, neither hast thou done it
aright. In such a case a red object must be painted red all over and yet
preserve the appearance of solidity; and so with all colours. The same
must be done with the shadows, lest it be said that a fair red is soiled
with black Wherefore be careful that thou shade each colour with a
similar colour. Thus I hold that a yellow, to retain its kind, must be
shaded with a yellow, darker toned than the principal colour. If thou
shade it with green or blue, it remaineth no longer in keeping, and is
no longer yellow, but becometh thereby a shot colour, like the colour of
silk stuffs woven of threads of two colours, as brown and blue, brown
and green, dark yellow and green, chestnut-brown and dark yellow, blue
and seal red, seal red and brown, and the many other colours one sees.
If a man hath such as these to paint, where the surface breaketh and
bendeth away the colours divide themselves so that they can be
distinguished one from another, and thus must thou paint them. But where
the surface lieth flat one colour alone appeareth. Howbeit, if thou art