This limitation is however •necessary, for very different reasons. When the eye is placed near C, the rays are inci dent almost perpendicularly. upon the. mirrors; and it is well known, from the experiments of Bouguer, that the loss of ligldby reflection, is in this case extremely great both in'the case of metals and polished glass. From this cause, the intensity even of the first reflected image is greatly infe rior to that of the sector A013 seen by direct vision ; but if the number of reflections is 16, 18, or 20, or even above 8 or 10, the last reflected image is scat ccly if at all percepti ble, even when a pretty strong light is thrown upon the aperture. If the picture consequently had been symmetri cal and agreeable to the eye, in so far as the arrangement of its parts was concerned, it would have lost all its beauty from the extreme inequality in the light of the reflected images of which it is composed. When the reflectors arc made of glass, or of glass covered on one side with black varnish, this difference is so very striking, that the circular field can scarcely be completed. By very powerful
indeed, the last image may be rendered visible ; but the difference of the intensities of the sectors remains the same, and therefore the imperfection of the picture cannot be corrected even by the application of the strongest lights.
As the eye advances from C to E, the angles of incidence increase, the loss of light diminishes, and the difference in the intensity of the reflected images is the least possible when the eye arrives at E. • In the case of blackened glass, the last reflected images are sufficiently bright, when the number of sectors is 12 or 16. Hence it follows, that in order to obtain a perfectly symmetrical picture from the images formed by reflection, and toprocure as much equality as possible in the light of the different images, the eye should be placed in the planes of both the reflectors, ur as near as possible to the angular point at E.