I have call'd it the retiring shade, because by the successive, or continual change in its appearance, it is equally instrumental with converging lines[13], in shewing how much objects, or any parts of them, retire or recede from the eye; without which, a floor, or horizontal-plane, would often seem to stand upright like a wall. And notwithstanding all the other ways by which we learn to know at what distances things are from us, frequent deceptions happen to the eye on account of deficiencies in this shade: for if the light chances to be so disposed on objects as not to give this shade its true gradating appearance, not only spaces are confounded, but round things appear flat, and flat ones round. [13] See p. 17. The two converging lines from the ship, to the point C, under fig. 47, plate I. [Illustration: Fig. 94] But although the retiring shade hath this property, when seen with converging lines, yet if it describes no particular form, as none of those do in fig. 94, on top of plate II, it can only appear as a flat-pencil'd shade; but being inclosed within some known boundary or out-line, such as may signify a wall, a road, a globe, or any other form in perspective where the parts retire, it will then shew its retiring quality: as for example, the retiring shade on the floor, in plate II, which gradates from the dog's feet to those of the dancer's, shews, that by this means a level appearance is given to the ground: so when a cube is put into true perspective on paper, with lines only, which do but barely hint the directions every face of it is meant to take, these shades make them seem to retire just as the perspective lines direct; thus mutually compleating the idea of those recessions which neither of them alone could do. [Illustration: gradated shade] Moreover, the out-line of a globe is but a circle on the paper; yet, according to the manner of filling up the space within it, with this shade, it may be made to appear either flat, globular, or concave, in any of its positions with the eye; and as each manner of filling up the circle for those purposes must be very different, it evidently shews the necessity of distinguishing this shade into as many species or kinds, as there are classes or species of lines, with which they may have a correspondence. In doing which, it will be found, that, by their correspondency with, and conformity to objects, either composed of straight, curved, waving, or serpentine lines, they of course take such appearances of variety as are adequate to the variety made by those lines; and by this conformity of shades we have the same ideas of any of the objects composed of the above lines in their front aspects, as we have of them by their profiles; which otherwise could not be without feeling them. Now instead of giving engraved examples of each species of shade, as I have done of lines, I have found that they may be more satisfactorily pointed out and described by having recourse to the life.