What are the principles on which such a form is admissible? First. It is very rarely indeed that the eye contemplates any landscape without elevating or depressing itself. In all mountain and architectural scenery it is raised; in all prospects of distant country, depressed. In this case the cone of rays enters the eye obliquely, upwards or downwards. But the plane of the picture is always vertical to the eye. Consequently we have the section, by a vertical plane, of a cone whose axis is inclined. This is an ellipse whose major axis is vertical. Similarly: it is seldom that the eye includes the thirty degrees on each side of its legitimate point of sight. There is always something more attractive on one side than on the other, and it directs itself to the attractive side *[We have not space to prove this more directly; but it is always acknowledged, practically, by the artist's placing his horizontal line high, or low in the picture, as the eye is depressed or elevated.], including, perhaps, forty degrees on one side; twenty degrees on the other. We have then the section of an oblique cone by a parallel plane, or an ellipse whose major axis is horizontal. Here, then, we have a most valuable modification of the monotonous circle; we have a figure susceptible of as much variety of form as the rectangle, and whose sides, where they cut the axes, very nearly correspond to straight lines. We have the power of increasing apparent elevation of architecture, by using the vertical ellipse; or of diminishing an overwhelming mass of sky, by taking the horizontal one. All this is of infinite practical advantage.
