AT the revival of the arts, some evil genius, who was determined to retard the progress of painting, dictated this rule. “A picture should always have its horizon the height of the eye that looks at it—in nature, the eye being always the height of the horizon; therefore a picture will be most like nature that has its horizon the height of the natural eye.” One of the falsest rules that ever was founded on a false principle! and this is the more lamentable, as it has spoiled, in point of perspective, three parts of the historical pictures that have ever been painted.

As it is very difficult to destroy a rooted error, and as this is a most pernicious one, it is necessary to be full and particular.

When I say eye and horizon—the natural eye and horizon are meant. When the terms artificial eye and artificial horizon are used, the eye and the horizon represented in painting are to be understood. We must be clear in this distinction, for it is the confounding of the ideas expressed by these terms which has occasioned the mischief.

The eye, and the horizon, are always of the same height—therefore

The artificial eye and the artificial horizon must always be so—but

There is no connection between the real eye, and the artificial horizon.

In every picture the artificial eye, or point of sight, is supposed to be at a certain height from the base-line; as high as a human figure would be, represented as standing there. To this point every thing in the picture tends, as every thing in a real view tends to the natural eye. The picture then, as far as this circumstance is concerned, is perfect, if the artificial eye and the artificial horizon go together; for these always bear the same relation to each other, let the picture be placed any where.

Let A be the eye, B the picture (in section) and c the horizon of the picture.—The eye is always the apex of the cone; there is constantly the same relation between the parts in every position. It must be observed that there is a defect in this illustration which it was impossible to avoid—for tho’ I have considered A as the eye, yet upon paper, it is artificial as well as the picture B. If you cannot make this distinction, I propose the following demonstration.——Take a landscape and stand it upon a table—hang it up the height of the eye—above the height—put it upon a chair—upon the floor—it still, perspectively considered, is seen equally well—for

The real eye is always the height of the artificial eye, whether the picture be fixed in the cieling or laid upon the floor.

Indeed if this was not so, how would it be possible to hang one picture over another? and yet this is done, and with the greatest propriety.

I have often lamented the shifts to which painters are reduced, who have followed this rule in opposition to their senses. Laresse was so thoroughly possessed with it, that his idea of fitting up a room with pictures, was to have those which were below the eye to contain nothing but ground, and those which were above, the sky and clouds. But though he was convinced of the rectitude of his principle, he was struck with the oddity of the practice—he therefore recommended that there should be but one picture from the floor to the cieling, in which there might be a perfect coincidence of the natural and artificial horizon.

A portrait-painter sets the person he is to draw generally the height of his eye.——Suppose it to be a whole-length with a landscape in the back-ground: the artist considers his picture is to hang above the eye, and for that reason makes his horizon low, about the height of the knees. The consequence is, that there are two points of sight, which supposes an impossibility; for the eye cannot be in two places at the same time. If the eye be supposed on a level with the head of the figure, as it was on drawing the face, then the back-ground is too low; if equal to the horizon of the back-ground, then the figure is too high, unless we suppose it on an eminence, or ourselves in a pit; in that case, instead of seeing the face in front, we must have looked under the chin—but as we do not, the figure always appears to be falling forward.

Raffaele’s horizon is most commonly the height of his figures, so that they stand properly, and seem to be, whether in a print or a picture, the size of human creatures;—on the contrary, when the horizon is low, the figures always appear gigantic. When I was a boy, I had formed so very exalted an idea of the size of running horses, from seeing them drawn with the distant hills appearing under their bodies, that the first time I was at a course, it appeared but as a rat-race.

Every whole length picture will furnish you with an instance of this false principle, which would appear more disagreeable, if custom had not in some measure reconciled us to it. I am aware that the practice of so many great men is a strong objection to my argument; but as I conceive, with due submission to such authority, that there is demonstration on my side, I cannot easily retract what I have advanced.