Therefore it must be crystal-clear that such views offer no test of accurate perspective drawing.
Then we have this rule:
Parallel lines that do not recede never have the appearance of meeting anywhere.
Look at the diagram of a village street.
The nature of the ground prevents the building of a long straight street, and the houses are dotted about. Although they present a diminishing effect (they are receding from the vision), yet they diminish to the horizon at four different points. Observe this well; it will demonstrate that:
If there are ten retiring lines and all parallel, there will be only one vanishing point for all; but if among the ten there are not two parallel lines, there will be no vanishing points.
perspective comes to our aid when we are perplexed with curves of arches, bridges, and doorways—beautiful objects that tempt the pencil and deceive the eye.
First sketch the arch or window, marking the direction of the base, the thickness of the wall; then, if you are in doubt about the rightful position of the curve, and the highest point of the arch, enclose the base lines in a square, drop a line from each corner, and at the intersection (or meeting-place) draw an upright line; that should find the centre of the arch.
In other words, enclose curves shapes in rectangular shapes.
Although a single arch, or even a couple of arches, might be sketched fairly correctly without such aid, a cluster of arches presents a more complicated problem, and we should feel justified in using this method of checking perspective.
Circles, we know, are exceedingly difficult to draw correctly. An artist, of course, should draw circles without resorting to mechanical means, but a beginner, on occasion, may wish to check his drawing of a circle by enclosing it in a square.
In Fig. 57 we have an upright circle in a square, also a circle enclosed in a square and in perspective—i.e, receding from the spectator.
Strictly speaking, many perspective problems belong to geometry and not to art, and provided that we understand a
