artists, both in portrait and in landscape, now avail themselves of photography, both as an auxiliary and a guide in their profession; but there are certain difficulties and imperfections in the art itself, and so many precautions required in its right application, whether we use its pictures single, as representations on a plane, or take them binocularly, to be raised into relief by the stereoscope, that we must draw from the principles of optics the only rules which can be of real services to the arts of design. [58] _Modern Painters_, vol. iii., Preface, pp. 11, 12. In painting a landscape, a building, a figure, or a group of figures, the object of the artist is to represent it on his canvas _just as he sees it_, having previously selected the best point of view, and marked for omission or improvement what is not beautiful, or what would interfere with the effect of his picture as a work of high art. His first step, therefore, is to fix upon the size of his canvas, or the distance at which the picture is to be seen, which determines its size. His own eye is a camera obscura, and the relation between the picture or image on its retina is such, that if we could view it from the centre of curvature of the retina, (the centre of visible direction,) a distance of half an inch, it would have precisely the same apparent magnitude as the object of which it is the image. Let us now suppose that the artist wishes to avail himself of the picture in the camera obscura as received either on paper or ground-glass, or of a photograph of the scene he is to paint. He must make use of a camera whose focal length is equal to the distance at which his picture is to be seen, and when the picture thus taken is viewed at this distance (suppose _two_ feet) it will, as a whole, and in all its parts, have the same apparent magnitude as the original object. This will be understood from Fig. 47, in which we may suppose H to be the lens of the camera, RB the object, and H_y′_ the distance at which it is to be viewed. The size of the picture taken with a lens at H, whose focal length is H_y′_, will be _b′r′_, and an eye placed at H will see the picture _b′r′_ under an angle _b′_H_r′_, equal to the angle RHB, under which the real object RB was seen by the artist from H. In like manner, a larger picture, _byr_, taken by a camera the focal distance of whose lens at H is H_y_, will be an accurate representation of the object RB, when viewed from H, and of the same apparent magnitude. If either of these pictures, _b′r′_ or _br_, are viewed from greater or less distances than H_y′_, or H_y_, they will not be correct representations of the object RB, either in apparent magnitude or form. That they will be of a different apparent magnitude, greater when viewed at less distances than H_y′_, H_y_, and less when viewed at greater distances, is too obvious to require any illustration. That they will differ in form, or