The lettering must next be attended to. The individual figures are usually designated by numbers; this is a bad method, since it involves referring to the description of the plate. The best way is to use a number, and after it to add the name of the plant or animal, and, if needs be, a description as short as may be. If the author can "print" or write reasonably well, well and good; if not, it is better to attach a slip to the plate with full directions relating to lettering, and to write in pencil on the plate the titles, etc., required in the proper places for the guidance of the craftsman. The typewriter is sometimes employed for this purpose by authors; it is purely a matter of taste, but some readers feel a slight shock when this method is resorted to. In some cases a key to the plate printed on tough translucent tissue paper and having the necessary information, guide lines, etc., is inserted with the plate.[A] [Footnote A: See Kerner and Oliver: _Natural History of Plants_ (First Edition) London, 1894.] In the case of glossy chromolithographs this practice is best avoided, for the key is apt to stick to the plate if too much pressure is used when the book is bound. With regard to the "catch letters" used to indicate different parts: these should be as obvious as possible, and the guide lines should be either in black or in white ink, according to the general tone of the illustration. These lines should be conspicuous without being heavy. Not infrequently they, together with the lettering, are printed on the plate by a second impression in red ink. The foregoing is primarily the business of the author; with regard to editors and publishers, all plates should be mounted in a manner to facilitate reference and should be printed on suitable paper; the former is seldom or never done. All plates which must be constantly referred to in reading the text should have a selvedge as broad as the book, so that when unfolded the whole plate is visible, no matter what page is being read. This would, no doubt, prove an additional expense, but this should not militate against the suggestion here made, not by any means an innovation, for in many cases it would save the expense of mounting on guards, and, further, the additional expenditure could be saved several times over in other ways. With regard to paper, this generally is satisfactory; unfortunately, highly glazed paper, mis-termed art paper, with an enamelled or chromo surface, and consisting chiefly of china clay and size, is generally used for printing the best half-tone reproductions. For this purpose a paper with a suitable surface, obtained by means other than those mentioned and not too costly, is highly desirable, since art paper has the reputation of being not at all permanent, owing to the deleterious action of moisture, and is somewhat brittle. When used, art paper, if folded, should have a proper paper hinge along the fold.