in the relative apparent size of their parts, has, so far as I know, not been conjectured. In order to shew this, let us suppose a man _six_ feet high to occupy the foreground, and another of the same size to be placed in the middle distance, the distance of the two from the artist being ten and twenty feet. The apparent magnitudes of these two men on the photograph will be as two to one; and if we look at it at any distance greater or less than the focal length H_y′_ of the lens, the same proportion of two to one will be preserved, whereas if we look at the original figures at a greater or less distance from them than the place of the artist, the ratio of their apparent magnitudes will be altered. If the artist, for example, advances five feet, the nearest man will be five feet distant and the other fifteen feet, so that their apparent magnitude will now be as three to one. [Illustration: FIG. 47.] The same observations apply to a portrait of the human face. In looking at a human profile let us suppose the breadth of the nose to be one inch, that of the ear one inch, and that we view this profile at the distance of _three_ feet from the ear, which is two inches nearer the observer than the nose. The apparent magnitude of the ear and nose will be as thirty-eight to thirty-six inches, whereas if we view the profile from the distance of _one_ foot the ratio will be as fourteen to twelve, that is, the ear will be increased in apparent size more than the nose. Hence it follows that all pictures should be viewed under the same angle of apparent magnitude under which they were seen by the artist as taken photographically, for if we view them at a greater or less angle than this we do not see the same picture as when we looked at the original landscape or portrait, under the same angle of apparent magnitude.