our experience. Yet there are cases where
experience is less decided, and where, consequently, we may regard any
particular line as advancing or receding. And it is found that when we
vividly imagine that the drawing is that of a convex or concave surface,
we see it to be so, with all the force of a complete perception. The
least disposition to see it in the other way will suffice to reverse the
interpretation. Thus, in the following drawing, the reader can easily
see at will something answering to a truncated pyramid, or to the
interior of a cooking vessel.

[Illustration: FIG. 5.]

Similarly, in the accompanying figure of a transparent solid, I can at
will select either of the two surfaces which approximately face the eye
and regard it as the nearer, the other appearing as the hinder surface
looked at through the body.

[Illustration: FIG. 6.]

Again, in the next drawing, taken from Schroeder, one may, by an effort
of will, see the diagonal step-like pattern, either as the view from
above of the edge of an advancing piece of wall at _a_, or as the view
from below of the edge of an advancing (overhanging) piece of wall at
_b_.

[Illustration: FIG. 7.]

These last drawings are not in true perspective on either of the
suppositions adopted, wherefore the choice is easier. But even when an
outline form is in perspective, a strenuous effort of imagination may
suffice to bring about a conversion of the appearance. Thus, if the
reader will look at the drawing of the box-like solid (Fig. 3, p. 79),
he will find that, after a trial or two, he succeeds in seeing it as a
_concave_ figure representing the coyer and two sides of a box as looked
at from within.[49]

Many of my readers, probably, share in my power of variously
interpreting the relative position of bands or stripes on fabrics such
as wall-papers, according to wish. I find that it is possible to view
now this stripe or set of stripes as standing out in relief upon the
others as a ground, now these others as advancing out of the first as a
background. The difficulty of selecting either interpretation at will
becomes greater, of course, in those cases where there is a powerful
suggestion of some particular local arrangement, as, for example, the
case of patterns much brighter than the ground, and especially of such
as represent known objects, as flowers. Yet even here a strong effort of
imagination will often suffice to bring about a conversion of the first
ap