If the object is placed within the reflectors at any point =D=, Fig. 16, between their object end =O=, and their eye-end =E=, a perfectly symmetrical picture will obviously be formed from it; but the centre of this picture will not be at =O=, the centre of the luminous sectors, but at the point =D=, where the object is placed, or its projection _d_, so that we shall have a circular luminous field enclosing an eccentric circular pattern. Such a position of the object is therefore entirely unfit for the production of a symmetrical picture, unless the object should be such as wholly to exclude the view of the circular field, formed by the reflected images of the aperture =A O B=. As the point =D= approaches to =O=, the centre _d_ of the symmetrical picture will approach to =O=, and when =D= coincides with =O=, the centre of the picture will be at =O=, and all the images of the object placed in the plane =A O B= will be similarly disposed in all the sectors which compose the circular field of view. Hence we may conclude, that a perfectly symmetrical pattern cannot be exhibited in the circular field of view, when the object is placed between =O= and =E=, or anywhere within the reflectors. If the eye could be placed exactly at the angular point =E=, so that every point of the line =E O= should be projected upon =O=, then the images would be symmetrically arranged round =O=; but this is obviously impossible, for the object would, in such circumstances, cease to become visible when this coincidence took place. But independent of the eccentricity of the pattern, the position of the object within the mirrors prevents that motion of the objects, without which a variation of the pattern cannot be produced. An object between the reflectors must always be exposed to view, and we cannot restrict our view to one-half, one-third, or one-fourth of it, as when we have it in our power to move the objects across the aperture, or the aperture over the objects. [Illustration: FIG. 16.] Another evil arising from the placing of the objects within the mirrors, is, that we are prevented from giving them the proper degree of illumination which is so essential to the distinctness of the last reflexions. The portions of the mirrors, too, beyond the objects, or those between =D= and =O=, are wholly unnecessary, as they are not concerned in the formation of the picture. Hence it follows, that the effects of the Kaleidoscope cannot be produced by any combination of mirrors, in which the objects are placed within them.