of luminousness to over-dark shadows and variety to blank whites.
The almost forgotten process of Pouncy, and of Poitevin, now known
as the gum bichromate process, was rehabilitated in 1894 by M.
Rouille Ladeveze expressly to meet the needs of the pictorial
worker. Perhaps the best results that have been achieved by it
are those of M. Robert Demachy of Paris, though many English
workers have used it with remarkable success. In it paper of any
kind may be selected as the support. The power of the operator
to modify the printed image to almost any extent, even to introducing
and eliminating lights and shadows, and in other ways to
depart widely from the image given by the negative, depends upon
the fact that the coating of gum and pigment (which, being bichromatized,
becomes insoluble in proportion as it is acted upon by light)
holds the pigment but imperfectly, and yields it up upon a vigorous
application of water. According, therefore, to its application or
retention, the operator can lighten or deepen in tone any portion.
Numberless variations of other methods, such as brush development
and local toning or stopping, have been suggested with the
same object. Other workers have shown that by dexterously shutting
off and admitting the light to various parts of the negative whilst
printing, the disposition of the lights and shades in the print can
be modified to so great an extent as to alter the general contour of
the scene. Examples of an original unaltered print, and one
which has been thus modified, are shown in the accompanying
plate. Portions are shaded in by allowing the light to have access
to the print, either through the negative—in which case the image
with all its details, prints more deeply—or by removing the negative.
when the action of the light is to flatten and suppress both detail
and contrast Latterly some few have resorted to extensive
working on the negative, both on the back and on the film; drawing
by hand is practised on the film to render too prominent features less
obtrusive, and objects in the background are merged by an intricacy
of lines and cross-hatching. Many of the results are very pleasing,
although one hesitates to justify the means, however good the
end. On the other hand, to exclaim for purity of method and the
exclusion of extraneous aids is very like setting up an arbitrary
standard no less unreasonable than those conventions against
which pictorial photography has so long striven.
Authorities.—P. H. Emerson, Naturalistic Photography; H. P. Robinson, Picture-making by Photography; Art Photography; Pictorial Effect in Photography; Elements of a Pictorial Photograph; A. H. Wall, Artistic Landscape Photography (1896); A. Horsley Hinton, Practical Pictorial Photography (1898), and subsequent editions; C. Puyo, Notes sur la photographic artistique (Paris).
(A. H. H.)
PHOTOGRAPHY, CELESTIAL. The requisites for celestial photography are best explained by a comparison with ordinary photography in several essential points.
a. Illumination.—In taking a portrait artificial light is used, being thrown on to the face of the sitter either directly or by reflection. If the day is dull a longer exposure is required, and artificial light may be used when the daylight fails. In photographing the stars there is no question of illuminating them by artificial light, for the strongest searchlight which We could throw in the direction of the heavenly bodies would have no sensible effect. The light used is their own, and its feebleness renders it necessary to make long exposures, the length increasing as we attempt to get images of fainter objects. The invention of the dry plate, by making it possible to give very long exposures caused a revolution in celestial photography. With the wet plate, exposures were limited to the few minutes during which the film would remain wet; but the dry plate can remain in the telescope for days, weeks or even years if necessary. On the approach of daylight, the cap is put on the camera, or the plate removed into the dark room; but when night returns the plate is put back in the telescope, which is accurately pointed to the same stars, the cap is removed, and the exposure is resumed without any loss from the interruption.
b. Magnification.—In taking a portrait we can obtain a large or small size by placing the camera near the sitter or far away. But this method is not available for the heavenly bodies, since we cannot sensibly approach them. To magnify an image we must lengthen the focus of the camera, either directly or indirectly. The direct method is to construct a lens or mirror of long focus, the camera becomes similar in length to a telescope; and indeed resembles a telescope in other respects, except that we take away the eye-piece and put in a photographic plate instead. If, however, we already have a lens of short focus which we wish to use, we may lengthen the focus indirectly by using a secondary magnifier, that is by putting in another lens near the focus of the first. In either case the profitable magnification is limited, not only by the imperfections of the optical apparatus but by disturbances in the atmosphere. Air currents, either outside or inside the telescope, act as irregular lenses of varying shape, and produce such defects in the image that we gain nothing by enlarging it beyond a certain point. Such air disturbances do not trouble the ordinary photographer at all, or scarcely at all: he is only concerned with a few feet of air, whereas the celestial photographer cannot escape from the necessity of looking through many miles of it.
c. Steadiness.—In taking a portrait the photographer is only concerned to fix his camera firmly and to induce his sitter to remain still. The heavenly bodies are in constant motion, though their real and apparent movements are fortunately smooth, except for air disturbances above mentioned. If, therefore, it were possible to devise perfectly smooth clockwork, we could keep the camera or telescope continually pointed to the required star or stars. But human workmanship has not yet made clockwork of sufficient strength and accuracy to keep a large telescope satisfactorily pointed. The clockwork which had been found good enough for use with visual telescopes was soon found to be quite inadequate for photography. The first method adopted was to bind two telescopes, one visual and the other photographic, firmly together, and by looking through the visual one to keep some object steadily on the cross wires by using the slow motion screws; meanwhile the other telescope was kept properly pointed for taking a photograph. As it was sometimes found that extremely fine movements were required, electrical arrangements were devised, whereby the observer, on simply pressing a button, could accelerate or retard the rate of the clockwork by a minute amount, instead of actually turning the screws by hand. And about the same time the idea arose of making these corrections automatically. This automatic correction is based on the principle that a freely swinging pendulum, which has no work to do, will naturally keep much better time than the clockwork which has to drive a heavy telescope; and if such a pendulum is therefore arranged to send a current every second through certain electro-magnets, apparatus can be devised to detect whether the clockwork is going properly; and to correct it in the right direction, if it is not. One or more of these three methods, which may be called hand-guiding, electrical control, and automatic electric control, are used in taking all celestial photographs.
The Photographic Image.—The image of a star on the plate should be, theoretically, merely a point; but in practice it is a small patch on the plate which grows in size as the exposure is lengthened, while at the same time it becomes darker in the middle. One reason for this is that light is many-coloured, and when we attempt to focus it by a lens, we can only get a very few colours into even approximate focus; the other colours are not brought to focus at all, and form concentric patches of fainter light on the plate, which increase in size with the error of focus. Thus at best our focusing is only a compromise. When the exposure is short, those colours which have most nearly been brought to focus have an effect, while the faint light of the others may produce no sensible impression. It is natural to select for the colours to be brought most sharply to focus those which are most important photographically, viz. those at the violet end of the spectrum. As the exposure proceeds the faint light of the other colours affects the plate by accumulation, and hence the image spreads, while at the same time the central part naturally becomes blacker.
A reflecting telescope brings all colours to the same focus; and it might appear, therefore, that images formed with it will not spread in this way. There is, however, another cause of spreading besides that due to colour; neither the reflecting telescope nor the lens can focus all the light received by them for more than one particular star. It is just theoretically possible to construct a mirror which would focus all the light from a star seen in the direction of its axis; but the light from another star seen in a slightly different direction would not be truly focused, since directly we leave the axis, some parts of the mirror have a focus slightly different from other parts; and if the image
