some half an inch wide, in the
sheet of cardboard, and a series of small circular holes in a slip of
cardboard an inch wide. Pass the slip over the square opening, and match
each colour beside one of the circular openings. You will thus have no
occasion to wash any of the colours away. But the first rough method is
generally all you want, as after a little practice, you only need to
_look_ at the hue through the opening in order to be able to transfer it
to your drawing at once.

[240] If colours were twenty times as costly as they are, we should have
many more good painters. If I were Chancellor of the Exchequer I would
lay a tax of twenty shillings a cake on all colours except black,
Prussian blue, Vandyke brown, and Chinese white, which I would leave for
students. I don't say this jestingly; I believe such a tax would do more
to advance real art than a great many schools of design.

[241] I say _modern_, because Titian's quiet way of blending colours,
which is the perfectly right one, is not understood now by any artist.
The best colour we reach is got by stippling; but this not quite right.

[242] The worst general character that colour can possibly have is a
prevalent tendency to a dirty yellowish green, like that of a decaying
heap of vegetables; this colour is _accurately_ indicative of decline or
paralysis in missal-painting.

[243] That is to say, local colour inherent in the object. The
gradations of colour in the various shadows belonging to various lights
exhibit form, and therefore no one but a colourist can ever draw _forms_
perfectly (see "Modern Painters," vol. iv. chap. iii. at the end); but
all notions of explaining form by superimposed colour, as in
architectural mouldings, are absurd. Colour adorns form, but does not
interpret it. An apple is prettier, because it is striped, but it does
not look a bit rounder; and a cheek is prettier because it is flushed,
but you would see the form of the cheek bone better if it were not.
Colour may, indeed, detach one shape from another, as in grounding a
bas-relief, but it always diminishes the appearance of projection, and
whether you put blue, purple, red, yellow, or green, for your ground,
the bas-relief will be just as clearly or just as imperfectly relieved,
as long as the colours are of equal depth. The blue ground will not
retire the hundredth part of an inch more than the red one.

[244] See, however, at the close of this letter, the notice of one more