ters upon the sky, the history of the choice that had been
directed by a drop of rain, and of the balance that had been turned by a
grain of sand.

[Illustration: FIG. 63.]

Sec. 24. Such are the principal laws, relating to the crested mountains,
for the expression of which we are to look to art; and we shall
accordingly find good and intelligent mountain-drawing distinguished
from bad mountain-drawing, by an indication, first, of the artist's
recognition of some great harmony among the summits, and of their
tendency to throw themselves into tidal waves, closely resembling those
of the sea itself; sometimes in free tossing towards the sky, but more
frequently still in the form of _breakers_, concave and steep on one
side, convex and less steep on the other; secondly, by his indication of
straight beds or fractures, continually stiffening themselves through
the curves in some given direction.

[Illustration: FIG. 64.]

Sec. 25. Fig. 63 is a facsimile of a piece of the background in Albert
Durer's woodcut of the binding of the great Dragon in the Apocalypse. It
is one of his most careless and rudest pieces of drawing; yet, observe
in it how notably the impulse of the breaking wave is indicated; and
note farther, how different a thing good drawing may be from delicate
_drawing_ on the one hand, and how different it must be from ignorant
drawing on the other. Woodcutting, in Durer's days, had reached no
delicacy capable of expressing subtle detail or aerial perspective. But
all the subtlety and aerial perspective of modern days are useless, and
even barbarous, if they fail in the expression of the essential mountain
facts.

Sec. 26. It will be noticed, however, that in this example of Durer's, the
recognition of straightness of line does not exist, and that for this
reason the hills look soft and earthy, not rocky.

So, also, in the next example, Fig. 64, the crest in the middle distance
is exceedingly fine in its expression of mountain force; the two ridges
of it being thrown up like the two edges of a return wave that has just
been beaten back from a rock. It is still, however, somewhat wanting in
the expression of straightness, and therefore slightly unnatural. It was
not people's way in the Middle Ages to look at mountains carefully
enough to discover the most subtle elements of their structure. Yet in
the next example, Fig. 65, the parallelism and rigidity are definitely
indicated, the crest outline being, howev