ground behind him, which Veronese has painted first, and then
when it was dry, or nearly so, struck the locks of the dog's white hair
over it with some half-dozen curling sweeps of his brush, right at once,
and forever. Had one line or hair of them gone wrong, it would have been
wrong forever; no retouching could have mended it. The poor copyists
daub in first some background, and then some dog's hair; then retouch
the background, then the hair, work for hours at it, expecting it always
to come right to-morrow--"when it is finished." They _may_ work for
centuries at it, and they will never do it. If they can do it with
Veronese's allowance of work, half a dozen sweeps of the hand over the
dark background, well; if not, they may ask the dog himself whether it
will ever come right, and get true answer from him--on Launce's
conditions: "If he say 'ay,' it will; if he say 'no,' it will; if he
shake his tail and say nothing, it will."

Whenever you lay on a mass of colour, be sure that however large it may
be, or however small, it shall be gradated. No colour exists in Nature
under ordinary circumstances without gradation. If you do not see this,
it is the fault of your inexperience; you _will_ see it in due time, if
you practise enough. But in general you may see it at once. In the birch
trunk, for instance, the rosy grey _must_ be gradated by the roundness
of the stem till it meets the shaded side; similarly the shaded side is
gradated by reflected, light. Accordingly, whether by adding water, or
white paint, or by unequal force of touch (this you will do at pleasure,
according to the texture you wish to produce), you must, in every tint
you lay on, make it a little paler at one part than another, and get an
even gradation between the two depths. This is very like laying down a
formal law or recipe for you; but you will find it is merely the
assertion of a natural fact. It is not indeed physically impossible to
meet with an ungradated piece of colour, but it is so supremely
improbable, that you had better get into the habit of asking yourself
invariably, when you are going to copy a tint,--not "_Is_ that
gradated?" but "_Which way_ is it gradated?" and at least in ninety-nine
out of a hundred instances, you will be able to answer decisively after
a careful glance, though the gradation may have been so subtle that you
did not see it at first. And it does not matter how small the touch of
colour may be, though not larger than