Kerner: _Pflanzenleben_, Leipzig, 1888. This contains some excellent engravings by Winkler and others.
Le Maout et Decaisne; _Traité général de Botanique_, Paris, 1876. This work contains splendid examples by Riocreux and Steinheil (see Fig. 8).
Oliver: _First Book of Indian Botany_, London, 1869. This contains some characteristic work of W. H. Fitch.
It does not appear to be generally known that excellent reproductions in colour may be obtained from wood blocks by superposed printing in a manner comparable to that followed in chromolithography although, of course, in the present instance, the blocks are in relief (Fig. 1).
From the foregoing account it is obvious that the engraving even of a small illustration, except it be in mere outline, involves a considerable amount of labour; in fact, if the subject were large it was usual to cut it up into areas and distribute between several engravers, the finished blocks finally being joined together to make the block of the whole picture. Hence it is not surprising to find that when the photo-mechanical processes were perfected, the older methods of reproduction were ousted by the newer, more especially since they are much less expensive; these, therefore, may next be considered.
THE HALF TONE PROCESS.--For the making of a relief block by photo-mechanical means, the main difficulty is the proper rendition of the tones intermediate between black and white; this has been solved, at any rate in part, by the discovery of the half-tone process.
If an ordinary photographic negative be highly magnified, it will be seen that the high lights, the low lights, and the intermediate tones are made by the varying density of the reduced silver. In the lighter parts the small black particles are surrounded by colourless areas, whilst in the dark regions small colourless patches are surrounded by black areas owing to the closeness of the particles of silver (Plate 5, Fig. 2).
What is required, therefore, is a relief block which will print a number of dots of equal density but of unequal size. Vervasser illustrates the point in an ingenious way: a plate, covered with a number of cones, is supposed to be acted upon by light in such a way that the cones are truncated in varying degrees according to intensity of the light falling upon them. The section of such a plate would therefore shew a curve (Fig. 3); now if the truncated cones be brought down to one level and a print taken from them, the high lights would be represented by black dots surrounded by white areas and so on.