OF TREES.

What outlines are seen in trees at a distance against the sky which
serves as their background?

The outlines of the ramification of trees, where they lie against
the illuminated sky, display a form which more nearly approaches the
spherical on proportion as they are remote, and the nearer they are
the less they appear in this spherical form; as in the first tree
_a_ which, being near to the eye, displays the true form of its
ramification; but this shows less in _b_ and is altogether lost in
_c_, where not merely the branches of the tree cannot be seen but
the whole tree is distinguished with difficulty. Every object in
shadow, of whatever form it may be, at a great distance appears to
be spherical. And this occurs because, if it is a square body, at a
very short distance it loses its angles, and a little farther off it
loses still more of its smaller sides which remain. And thus before
the whole is lost [to sight] the parts are lost, being smaller than
the whole; as a man, who in such a distant position loses his legs,
arms and head before [the mass of] his body, then the outlines of
length are lost before those of breadth, and where they have become
equal it would be a square if the angles remained; but as they are
lost it is round.

[Footnote: The sketch No. 4, Pl. XXVIII, belongs to this passage.]

The cast shadow of trees (452. 453).