831. One-mirror Stereoscopes. L. Pigeon (1904) made an instrument, somewhat analogous to the one described in 184.9 by Brewster and in 1851 by H. W. Dove. Closed, this instrument appears like a book ; when opened, the two side panels form an angle of about a median frame joined to the sides by linen or metal strips bisects the angle formed by the side panels. On one of the panels is placed at A'B' (Fig. 212) a normal image of the right-hand picture, and on the other panel at AB is placed a reversed image of the left-hand picture. The two pictures can be join «i by means of a paper or cloth hinge the whole slides under a bridge in the lower edge of the bisecting panel. A small mirror M is fixed 83o. Stereoscopes Having Two Reflectors in Front of One Eye. It is clear that one of the images of a stereoscopic pair can be viewed with the naked eye, and the other after reflection in two parallel mirrors suitably placed to enable fusion of the two images. This fusion is only possible on the condition that the path of the rays between the two reflecting surfaces is negligible compared with the distance of the stereograun from the eyes. Such instruments have been designed, notably by J. Duboscq (2857), for viewing two images placed one above the other, and by T. Brown (1895) for viewing two images side by side. On the same principle, at the end of the bisecting panel D,11 in such a position that the left eye 0 can be placed very near to it ; this eye sees at A ,B, the reflected image of A B, whilst the right eye 0' sees A' /3' directly.' 832. Stereoscopes with Reversing Prisms. Many devices have been worked out to produce a stereoscope for viewing stereograms printed without transposition (§ 8i6) from a pair taken on a single plate. Unfortunately, their use has not become general, but the principles employed will be briefly indicated.
A stereoscope suggested by Brewster in 1849 has two mirrors back to back in a plane per pendicular to that of the pictures examined, and situated half-way between the principal points. Each eye then sees the reversed image of its picture. The same inventor pointed out the possibility of reversing the images by means of a direct-vision total-reflecting prism in front of each eye.
The stereogram can be placed above or below the eyes and seen by means of a mirror held in front of the eyes (T. Brown, 1899), or placed in a plane parallel to the plane containing the two axes of vision, and seen by reflections at (G. Balmitgere, 1909). In another model by the latter inventor the stereogram is placed in a vertical plane at about 45° to the axes of the eyepieces ; two reflectors suitably placed produce virtual images normal to the direction of vision. The inequality in the two optical paths is then compensated by placing in the path of the more distant image a prism of appropriate thickness or an eyepiece constructed differently from that in the other path.
Stereoscopes with reversing eyepieces com prising total-reflecting prisms were worked out by Duboscq (1857) and by E. Colardeau (19n) ; in this model the prisms, producing three reflec tions, are placed between the eyepieces and the transparency to be examined ; they are held in a movable carrier which can be removed from the field to allow of pseudoscopic examination of the same stereogram, or stereoscopic viewing of a transposed stereogram.
833. Anaglyphs. An anaglyph consists of the two pictures of a stereoscopic couple super imposed on the same support, the two pictures being printed in two complementary colours. The result, though meaningless to the naked eve, appears in relief if examined through a double eyepiece such that each eye sees through a filter of colour complementary to the colour of its corresponding image (L. Ducos du 1891).
The left-hand image is generally printed in blue-green or bluish-green, and the right-hand image in red, slightly orange. Under these conditions the left eye through a red filter sees the green image in black on a red ground, and cannot see the red image on this red ground ; it therefore sees only its corresponding image. In the same way, the right eye through a green filter sees the red image in black on a green ground and cannot see the green image (or only very feebly). 2 The reconstructed object therefore appears black on a ground of colour not far removed from the tint resulting from the mixture of coloured lights transmitted respectively by the two filters, but with momentary predominance of one colour or the other, according to the state of fatigue of the eyes.
When examining these anaglyphs with the two-colour eyepieces, those parts of the object the corresponding points of which are printed in coincidence appear to be in the plane of the anaglyph, the other elements of the recon structed object appearing in front or behind this plane. The elements whose corresponding points have a separation equal to (or slightly smaller for easy viewing) the separation of the eyes will appear at infinity.
Anaglyphs have several advantages : they do not impose any limit to the dimensions of the pictures and only require for examination an instrument very easily and cheaply made by means of coloured gelatines (manufactured on the large scale) fixed in a light card, which can be punched out. After falling into disuse soon after their invention, anaglyphs have been em ployed to illustrate some scientific publications,3 and since 1923 have received many applications (projections, catalogues, illustrated journals), the prints usually being made by a photo mechanical process.
834. Parallax Stereograms. Imagine in front of a plate F (Fig. 213) in a parallel plane a grating 7' formed by opaque vertical and trans parent bands, having the same width I over the entire length of the grid. Place at two points Of), separated by a distance b (mean separation of the eyes) and at a distance d from the plate, the two lenses fitted with reflectors for reversing each image and at the time of recording the images adjust the separation e between the grating and the plate to satisfy the relation e . Under these conditions the bands GG'G" . . of the sensitive plate only receives light from the lens 0,,, whilst the intermediate bands DD'D" . . . only receive light from the lens 0,,. The sensitive plate can thus only register half the total area of each of the images of the stereoscopic pair, but, if the bands are sufficiently narrow (about 125 to the inch), the discontinuity of each image will not be as con spicuous as the discontinuity of photo-mechan ical reproductions (see for example the plates in Chapter XVII, obtained by means of cross-line screens of about 150 lines per inch).