[Illustration: PLATE 6. Half Tone reproduction of a photograph by Mr. W. Rowan.
Part of a shingle beach shewing plants of Sea Blite (_Suaeda fruticosa_) and a ring plover's nest with four eggs.]
Before passing on it may be mentioned that screens with patterns other than that represented in Fig. 4 are sometimes employed; for instance, the wavy-line screen gives the impression of coarse collotype.
The preparation of the blocks may now be briefly dealt with.
A negative of the picture, using a screen suitable for the purpose, is taken on a special dry gelatine plate ("process" plates) or on some other form of negative, _e.g._, wet collodion which is most commonly employed. This negative requires very careful development in order to get the dots right.
From the negative a positive is made upon a copper or zinc plate, suitably coated with a sensitive film. The usual practice is to coat the polished metal plate with a mixture of water, albumen, fish glue, ammonium bichromate, chromic acid and ammonia; the plate is then dried and, when cooled, exposed under the negative. The action of the light on such a film, the essentials of which are the albumen, the glue or gelatine and a chromate, has already been described. The mixture becomes more or less insoluble in water, according to the intensity of the light falling upon it.
The positive is now rinsed in water, and is sometimes stained with an aniline dye in order to render the film more visible. Next it is developed in a stream of water until the surface of the metal is visible between the dots, the last traces of the soluble gelatine being removed with warm water. After drying, the plate is evenly heated over a Bunsen burner until the dots of gelatine mixture turn chocolate colour, when the plate is allowed to cool gradually. This is known as burning in. The plate, if necessary, is now touched up and the back, sides and margins varnished in order to protect them from the acid: when the varnish is dry, the plate is etched in a weak solution (about 2-1/2 per cent.) of nitric acid if the metal be zinc; if the plate be copper, it is usually etched with a solution of iron perchloride.
On taking a proof, there is almost certain to be a lack of contrast, the plate is then fine etched, by which means a considerable improvement can be made; and, by covering certain parts with an acid-resisting substance ("stopping out"), it is possible to fine etch locally.