For our chromatic exhibitions we can, however, divide objects generally into _primary_ and _secondary_. The expressions of themselves denote what we understand by them, but our meaning will be rendered still more plain by what follows. 221. Primary objects may be considered firstly as _original_, as images which are impressed on the eye by things before it, and which assure us of their reality. To these the secondary images may be opposed as _derived_ images, which remain in the organ when the object itself is taken away; those apparent after-images, which have been circumstantially treated of in the doctrine of physiological colours. 222. The primary images, again, may be considered as _direct_ images, which, like the original impressions, are conveyed immediately from the object to the eye. In contradistinction to these, the secondary images may be considered as _indirect_, being only conveyed to us, as it were, at second-hand from a reflecting surface. These are the mirrored, or catoptrical, images, which in certain cases can also become double images: 223. When, namely, the reflecting body is transparent, and has two parallel surfaces, one behind the other: in such a case, an image may be reflected to the eye from both surfaces, and thus arise double images, inasmuch as the upper image does not quite cover the under one: this may take place in various ways. Let a playing-card be held before a mirror. We shall at first see the distinct image of the card, but the edge of the whole card, as well as that of every spot upon it, will be bounded on one side with a border, which is the beginning of the second reflection. This effect varies in different mirrors, according to the different thickness of the glass, and the accidents of polishing. If a person wearing a white waistcoat, with the remaining part of his dress dark, stands before certain mirrors, the border appears very distinctly, and in like manner the metal buttons on dark cloth exhibit the double reflection very evidently. 224. The reader who has made himself acquainted with our former descriptions of experiments (80) will the more readily follow the present statement. The window-bars reflected by plates of glass appear double, and by increased thickness of the glass, and a due adaptation of the angle of reflection, the two reflections may be entirely separated from each other. So a vase full of water, with a plane mirror-like bottom, reflects any object twice, the two reflections being more or less separated under the same conditions. In these cases it is to be observed that, where the two reflections cover each other, the perfect vivid image is reflected, but where they are separated they exhibit only weak, transparent, and shadowy images. 225.