and
the oblique ones decisive in proportion to their steepness.

A few close studies will soon teach you this: the only thing you need to
be told is to watch carefully the lines of _disturbance_ on the surface,
as when a bird swims across it, or a fish rises, or the current plays
round a stone, reed, or other obstacle. Take the greatest pains to get
the _curves_ of these lines true; the whole value of your careful
drawing of the reflections may be lost by your admitting a single false
curve of ripple from a wild duck's breast. And (as in other subjects) if
you are dissatisfied with your result, always try for more unity and
delicacy: if your reflections are only soft and gradated enough, they
are nearly sure to give you a pleasant effect. When you are taking
pains, work the softer reflections, where they are drawn out by motion
in the water, with touches as nearly horizontal as may be; but when you
are in a hurry, indicate the place and play of the images with vertical
lines. The actual _construction_ of a calm elongated reflection is with
horizontal lines: but it is often impossible to draw the descending
shades delicately enough with a horizontal touch; and it is best always
when you are in a hurry, and sometimes when you are not, to use the
vertical touch. When the ripples are large, the reflections become
shaken, and must be drawn with bold undulatory descending lines.

I need not, I should think, tell you that it is of the greatest possible
importance to draw the curves of the shore rightly. Their perspective
is, if not more subtle, at least more stringent than that of any other
lines in Nature. It will not be detected by the general observer, if you
miss the curve of a branch, or the sweep of a cloud, or the perspective
of a building;[231] but every intelligent spectator will feel the
difference between a rightly drawn bend of shore or shingle, and a false
one. _Absolutely_ right, in difficult river perspectives seen from
heights, I believe no one but Turner ever has been yet; and observe,
there is NO rule for them. To develope the curve mathematically would
require a knowledge of the exact quantity of water in the river, the
shape of its bed, and the hardness of the rock or shore; and even with
these data, the problem would be one which no mathematician could solve
but approximatively. The instinct of the eye can do it; nothing else.

If, after a little study from Nature, you get puzzled by the great
difference