Arrangement.

The final shape of the publication of the record has to be borne in mind in all the progress of it. The arrangement of the plates must precede the writing of the details of the work. In past generations the ideal was to define in words the conclusions and speculations of an author, and, where unavoidably necessary, to illustrate them by some costly engravings. How inefficient such publication may be, is seen at once in Greenwell’s British Barrows, a work full of important detail, which has to be painfully understood from hundreds of pages of text, where plans—and little else—are needed. Indeed the only means of using the information is to reconstruct plans from the intricate text. As form can now be almost as cheaply expressed as words, the ideal is widely changed. The reader is to be put first of all in possession of all the facts and materials, and the author’s conclusions are only a co-ordination, presented to enable the reader to grasp the material, and to feel clearly the effect of it on his sum of ideas, or organised sense of the nature of things. Hence nowadays the main structure of a book on any descriptive science is its plates, and the text is to show the meaning and relation of the facts already expressed by form. The plates, therefore, are the first thing to prepare; and when they are complete it is time to put in words the conclusions which have been reached.

The orderly arrangement of the material in plates is the first duty. The drawings are each to be made with the final scale in view, so that the lines may be of proper thickness, neither faint nor coarse. The material must be classified according to its nature,—views, plans, inscriptions, sculpture, small objects, pottery, etc. In each class, the historical order must be followed, objects that are to be compared placed together, and the material arranged in an orderly shape, so that it gives a clear impression, and can be easily found again from memory. The details of the squareness and alignment of the various drawings on a plate are much more serious than might be supposed; needless irregularity confuses and disappoints the eye and starves the memory, distinctly detracting from the use and value of the work. Obviously every object on a plate must have a number for reference; and in a long series it is best for the numbers to run through several plates; so that “sealing 157” or “mark 642” is a complete reference and definition. A uniform scale should be used throughout a plate, or a series of the same class, and it should be stated in the heading of the plate. Every plate should have stated in its heading the source, nature, age, and scale of the objects; for these render reference far easier, and also give a value to loose plates apart from the volume. The use of double-page plates is often desirable, in order to show the whole of a large class at one view; the only drawback to them is that objects are more difficult to find in turning over the leaves. At every point it must be remembered that nearly all foreign students, and most English ones, will know the plates but not the text; that the plates will be the material practically used for comparison, and building up a view of the subject; and therefore that they should be as far as possible self-contained and self-explanatory, with full lettering upon them, and should comprise the main results of the work in diagram. To help reference to the text, the list of plates should have the page references to each plate stated, to show where it is described and dealt with in the text. The facility of using, remembering, and referring to the plates should be the first consideration. It is even well to remember to make the right-hand edge, or outer edge, of each plate the strongest part, with the most striking objects and best arrangement, and let the other edge be a residual, as this ensures the best eye-grasp in turning over the leaves.

The amount of plates must depend upon the subject; but it is none too much if the area of plates is double that of the text, or twice as many plates as there are pages. Folding-out plates should be avoided where possible; a double page on a deep guard, so as to lie flat when the book is opened, is the largest that should be ordinarily used. The most absurdly inconvenient shape is to have wide margins to a plate, and a fold at side and another at base, to make it fit the book. It is best to remedy such folly by taking the plates out, cutting them to book size if margins allow, and resetting without folds. The wild freaks of recent books in Egyptology are incomprehensible. We see some with plates which might be bound with text, yet printed with gigantic margins and issued in an entirely different size, so that they cannot be bound, or even stand on the same shelf with the text; some plates put on guards of tissue paper, so that they tear out of the book in turning over; one serial in parts with the plates starting fresh numbers with every separate paper, thus one part has half a dozen “Plate 1” in it, making printed references to the work quite impossible; other publications with the plates all renumbered and rearranged after printing, and double references throughout; others with scattered numbers of the plates issued, and intermediate numbers to appear later, after many years or never; some with plates without any numbers to the objects, and stray references in the text showing that they are usually counted by the author from the base upward, and from right to left. Every absurdity which want of design, forethought, and common sense could perpetrate, seems to be found in these monumental works.

The processes used for plates vary greatly in cost and quality. The cheapest is photolithography from line drawings; but only black and white can be given thus, without any half-tones, and the illustrations must be all together on a plate, and cannot be placed in the text. Yet as it can be done at less than 2d. a square inch for 250, or 6d. for 2000, it enables a much larger quantity of illustration to be given than would be possible otherwise. Relief process from line drawings costs 4d. a square inch for the blocks alone, without printing on paper; but as it can be placed with the text and printed together, it has a great advantage, especially for small subjects.

Collotype is next in cost, being 6d. a square inch for 250, or 2s. for 2000, but less than this cost in Germany. It has the same disadvantage in being restricted to whole plates, and not mixable with text, but it gives the half-tones well from photographs, and in fine examples is almost as good as a silver print. The finest I have seen were from Berlin. It is best to supply glass positives to the collotyper, and leave him to make such negatives as may suit him. If negatives are sent they are often destroyed. Net process gives half-tones, though with too coarse a grain for very delicate details. The cost is about double that of relief blocks, but as it reproduces photographs which can be mixed with the text it has an enormous use now, from cheap newspapers up to art publications. A disadvantage is that it requires a highly glazed paper to print upon, such as is unpleasant to read, heavy to hold, and liable to decay. Its duration therefore is distinctly ephemeral.

For special subjects the more costly processes are requisite. Chromo-lithography may be expected to cost about half as much again as photolithography for each colour used. As seldom less than four colours are efficient it costs at least six times as much as the line plates; thus the cheapest colour plate begins at the cost of the best net process; and it may easily come to three or four times that amount. But probably the three-colour photography will soon abolish chromo-lithography, and work much cheaper, perhaps at three or four times the price of collotype.

The autotype, platinotype, heliogravure, Swan electric engraving, and other processes all have their place for special subjects, but seldom come into the general run of archaeological illustration.

A very successful policy for costly works of research is to issue a magnificent edition for libraries, book-collectors, and rich amateurs; and then to have a much larger edition, deficient in a few of the most costly and least necessary plates, sold at a cheap rate for students and the general public. Thus one great work of coloured folio plates costs £20 or 3s. a plate for the complete edition; whereas with a few plates deficient it is only £6 or 1s. a plate. Thus the cost of production is borne by those who demand magnificence, and the results are yet within reach of students.

Another useful arrangement is to issue a public edition for general reading, and an appendix of extra plates for students, which would overweight a general edition. Thus a 2000 edition of the popular half of the plates may cost £400, and a 250 edition of the students’ half of the plates may cost £100, so saving £300, which would be uselessly spent on 1750 copies that are not wanted, and which would only be a dead-weight to the main work.

In arrangement of the text the main necessity is ready reference, and a form which can be remembered. The way to this is by classifying the material, dividing into chapters and paragraphs, each with a title, and above all making a good index, which ought to be about a tenth of the length of the work. A list of plates should have page references for each plate. Remember that all smaller type, footnotes, and tables are far more expensive than straightforward printing.

The general nature of the record of results has been already dealt with under the recording; and the need of giving an organic handling of the whole has been pointed out.

As to publication, if any publisher will undertake to issue a work of research at his own risk, well and good. If the author gets a gradually increasing royalty after the first 100 copies, that is as much as can be expected from this class of literature. But in no case have any profit-sharing agreement. Usually such a work will have to be issued at the author’s risk, and a few of the pitfalls of such arrangements may be noted. Let the manuscript really be in final condition, down to every stop, before it goes to the printer; consider the details of headlines, paragraphing, insertion of illustrations, arrangements of any tables or lists, (counting the letters), and in short leave nothing undefined. Have an agreement with the printer for terms, including an average of, say, two author’s alterations in every page, none to alter the length of any page: this allows for inevitable small improvements, without leaving an entire uncertainty in the charges. Correct the proofs in red for the author’s alterations, in black for the printer’s errors. If alterations exceed the allowance, reckon on paying for the resetting of the worst pages, so as to bring the average to the allowance on the rest. Beside the contract for printing and binding, have a contract with the lithographer, another with the collotyper, and another with the bookseller, for his terms of commission on sales. Thus the author knows exactly where he is, and no unpleasantness can arise from unexpected charges.

After publication, the binder and plate-printer should be asked for any blocks used; and to send up any “overs” or spoilt plates; as such are often valuable afterwards to cut up for special uses, and may save spoiling copies of the book. All photographs and drawings supplied to the plate- or block-maker should also be asked for if not returned at once.

---