one good study,
than twenty left unnecessarily inaccurate,--take your pen, and put a
fine outline to all the boughs, as you did to your letter, taking care,
as far as possible, to put the outline within the edge of the shade, so
as not to make the boughs thicker: the main use of the outline is to
affirm the whole more clearly; to do away with little accidental
roughnesses and excrescences, and especially to mark where boughs cross,
or come in front of each other, as at such points their arrangement in
this kind of sketch is unintelligible without the outline. It may
perfectly well happen that in Nature it should be less distinct than
your outline will make it; but it is better in this kind of sketch to
mark the facts clearly. The temptation is always to be slovenly and
careless, and the outline is like a bridle, and forces our indolence
into attention and precision. The outline should be about the thickness
of that in Fig. 4, which represents the ramification of a small stone
pine, only I have not endeavored to represent the pencil shading within
the outline, as I could not easily express it in a wood-cut; and you
have nothing to do at present with the indication of foliage above, of
which in another place. You may also draw your trees as much larger than
this figure as you like; only, however large they may be, keep the
outline as delicate, and draw the branches far enough into their outer
sprays to give quite as slender ramification as you have in this figure,
otherwise you do not get good enough practice out of them.

[Illustration: FIG. 4.]

22. You cannot do too many studies of this kind: every one will give you
some new notion about trees. But when you are tired of tree boughs, take
any forms whatever which are drawn in flat color, one upon another; as
patterns on any kind of cloth, or flat china (tiles, for instance),
executed in two colors only; and practice drawing them of the right
shape and size by the eye, and filling them in with shade of the depth
required.

In doing this, you will first have to meet the difficulty of
representing depth of color by depth of shade. Thus a pattern of
ultramarine blue will have to be represented by a darker tint of gray
than a pattern of yellow.

23. And now it is both time for you to begin to learn the mechanical use
of the brush; and necessary for you to do so in order to provide
yourself with the gradated scale of color which you will want. If you
can, by any means, g