SOME authors have a general idea of how a book is manufactured, but more have none. Even in the case of experienced writers, every printing-office could tell surprising stories to illustrate the unreasonableness born of a lack of knowledge of the ordinary mechanics of manufacture, or of a confidence born of too little knowledge. And unreasonableness on the part of the author means extra and unnecessary expense either to the printer or to the publisher. One of the most unfortunate features of the publishing business is that the exact cost of manufacturing a book can rarely be estimated in advance: typography, electrotyping, engraving, designing, presswork, paper, and binding can be figured closely, but the "extras," resulting from the author's carelessness, lack of knowledge of the book's mechanics, or change of heart as the manuscript goes into type, in many cases so increase the cost beyond the publisher's expectations that the publication can only show a loss instead of a profit.

These "extras" result from different causes: the manuscript may be carelessly prepared, with poor punctuation and clumsy expressions, which the author corrects in the proof. The author frequently boasts, "I know nothing about punctuation," but would an artist admit that he was ignorant of how to mix his colors? There is no question that authors sometimes take advantage of their "temperament," and lie down upon it in a manner most unfair to their co-partners in the enterprise.

Changes in the manuscript cost nothing, changes in the type cost one dollar per hour. To correct vital points after the book is in type is warranted; to correct blunders in punctuation or expression is needless expense, and is a reflection upon the intelligence of the author. Genius may be erratic, but it is more respected when it is not made to carry the responsibility of ordinary carelessness or ignorance. The writer recalls a case where the author of a story changed the name of one of the characters after the book was in type; it cost the publishers over eighty dollars. Frequently an author changes the name of his story, necessitating resetting the running-heads, the title-page, and recutting the brass dies, all of which adds expense beyond the publisher's original estimate. Countless other examples might be cited, but the main point is that all vital details should be discussed and settled while the story is still in manuscript, and after it has been placed in the printer's hands further changes should be only those which are of serious moment. Other "extra" expenses include the cost of proofs. The publisher usually receives from the printer two sets of galley-proof, two sets of page-proof, and two sets of foundry-proof. All proofs beyond these six sets are charged for as "extra," the usual rate being one cent per page.

If the author retains his proof longer than is necessary to read and correct it, this delay frequently forces the printer to work over-time to meet publication-day; this over-time work is charged for at double price.

An author would never have any difficulty in securing a letter of introduction from his publisher to some large printing-house, and the printer would gladly give him every opportunity to familiarize himself with the mechanical processes. This knowledge, together with a study of those elements which go into the manufacture of a book, would enable the author to avoid needless cost, or to incur intelligently such extra expense as became vitally necessary.

The following suggestions are important regarding the relations between the author and the printer:

It is always wiser to leave all questions of typography for the publisher to settle with the printer, unless there is some specific reason why the author wishes to accomplish a particular result by using certain type effects.

Copy should be typewritten, and revised carefully by the author, before sending it to the printer, to correct typewriter's errors. Interlineations and erasures which make the reading difficult should be avoided. It is always a simple matter to rewrite such portions without rewriting the entire chapter.

If the author has decided preferences regarding spelling or punctuation, this fact should be clearly stated on the manuscript; otherwise the printer follows his office style, which may or may not conform with the author's ideas.

In the preparation of copy, consistency of spelling and punctuation is strongly urged, as it not only simplifies the problem for the printer, but also prevents possible misunderstanding of copy and consequent necessity for resetting.

All paragraphs should be clearly indicated in the copy.

All directions written upon the manuscript, which are not intended as "copy," should be enclosed in a circle.

The author should punctuate each sentence as he writes it, for in this way the marks are indicative of the natural pauses, and better express his meaning.

Foot-notes should always be clearly indicated.

Unusual words, proper names, and figures should be written out with the greatest care and distinctness by the author. It is for the common advantage of the author, the publisher, and the printer that the author or the editor read all proofs promptly.

ESTIMATING THE MANUSCRIPT

The usual procedure in making a book is as follows: When the publisher sends the manuscript to the printer, a request goes with it for a sample page, set to a size and in a type which will make a volume of the desired number of pages. A novel is supposed to run from 320 pages to 400 pages. The first thing to be done is to estimate the number of words in the manuscript, and this is accomplished by averaging the number of words in say thirty lines, and then multiplying by the number of lines on a page. No allowance is made for fractional lines, as these also occur in the printed page. If the manuscript is carefully written, each page will contain the same number of lines, so the total number of words may be found by multiplying the number of words on the page, as arrived at above, by the total number of pages in the manuscript. This explains the importance of having a standard number of lines on each page.[32] No allowance is made for fractional pages at the end of chapters, as there are also fractional pages in the printed book, and it averages up.

The front matter has to be estimated separately, with allowance for the blanks on the reverse of bastard-title, dedication, etc.,[33] but the usual number of pages is eight. Then, again, an allowance of half a page for each chapter sinkage[34] has to be made. Suppose we have a manuscript of 90,000 words, with 24 chapters: A type page of 280 words gives us 322 pages, to which we add 8 pages for front and 12 pages for chapter sinkages, giving us a volume of about 344 pages. As the presswork is usually done in forms of 32 pages, an effort is always made not to exceed even forms by a small number of pages. Striking out the bastard-title will often save a form of press-work.

Various short-cuts have been suggested for estimating the number of words in a printed page, but the old-fashioned method of counting is the safest. Here is a table which is as accurate as any short-cut can be:

In cases where the number of lines to the inch of certain sizes of type is desired, the following table may be employed up to 18-point body:

THE SAMPLE PAGE

With these details settled, the sample page is next in order. Knowing that the book is to be a 12mo (size of leaf 5? × 7?) or a 10mo (size of leaf 5½ × 8¼), the printer must "lay out" the page so as to leave margins of proper size and proportion. A 12mo type page may vary from 3 × 5¼ inches to 4 × 6¾ inches. Somewhere within this area, in the given example, the page must contain about 280 words. If the manuscript is long, then the type page must be large, the type itself small (never smaller than long primer[36] nor larger than pica[36]), the leads[36] reduced or omitted altogether. This is where the printer's taste and skill is given an opportunity for expression: he is the architect of the book, and must not combine types or decorations which are inharmonious, and his proportions must be kept correct.

For his sample page for the given novel, the printer would select from these standard faces:

Type sizes in the present day are determined by the point system, the fundamental unit of which is the point. This is obtained by dividing a length of 13? inches into 996 equal parts, each one being called a point. One point is therefore .0138 of an inch or 72.46 points are equal to 1 inch.

For purposes of convenience, a point is expressed as being 1/72 of an inch. Thus 6-point type occupies 6/72 of an inch of space, 12-point 12/72 and so on. This does not mean, however, that the actual printed face occupies six points on the paper, but that it is six points from the base to the top of the body carrying the face.

In other words, one may say that it is 12 points from the bottom of one line of 12-point type to the bottom of the next line of 12-point type, etc.

The pica is the standard of measurement of the old system, and is equal to 12 points of the new system; thus six picas are equal to 1 inch or 72 points. Printers still estimate the length and width of a page or a column by the pica; thus a page 4 inches wide is 24 picas.

The "em" is the square of a type body. Thus a "12-point em" is 12 points wide and 12 points long, or 1 pica long and 1 pica wide. A "10-point em" is a 10 point square, etc. The em used in measuring newspaper column widths, magazine columns, etc., is known as the em pica, which is 12 points square.

In using larger faces for headings and display, or smaller faces for footnotes or quoted matter, the printer will select from the same family to which the type belongs, or from some family which combines with it harmoniously. Old-style faces should not be used with modern faces, but the Scotch face, which is a cross between old-style and modern, combines well with either.

As to leading, this volume is leaded with a 1-point lead; between the first and second lines of the preceding paragraph there is no leading (technically, "set solid"); between the second and third, a 2-point lead, and between the third and fourth, a 3-point lead.

In technical volumes and schoolbooks the Old Style Antique type is largely used for subject-headings and side-notes:

THE TYPESETTING

With the sample page accepted by the publisher or author, or both, the printer is authorized to proceed with the typesetting. Setting type by hand is now almost entirely superseded by machine-composition, except for the display pages (such as the title) and where the type itself runs larger than the English (14-point) size, this being the limit of the machines. Linotype[37] composition is cheaper than monotype,[37] but as the type is cast all in one line, instead of in separate characters, the cost of corrections is much higher. To change even a mark of punctuation requires recasting the entire line. If the manuscript is reasonably final in its form, the publisher is likely to order linotype composition; otherwise, monotype will be selected. Both machines carry the standard faces and sizes of type.

THE PROOFS

The first proofs sent out by the printer are called "galley-slips," or "galleys."[38] These are supposed to give the author opportunity to make such changes as are absolutely necessary. When returned to the printer, these galleys are made up into "page-proofs,"[38] and frequently go again to the author, or the type may be "cast" (made into electrotype plates) at this point. When page-proofs are submitted to the author, the publisher expects him to revise them, making sure that all his galley corrections have been properly made, rather than to make further corrections, as changes in the pages are still more expensive than in the galleys. If changes must be made, the author should endeavor to have the correction occupy exactly the same space as the matter cut out, or to cut out further matter to make room for the addition. Otherwise, the page so corrected will contain more than the standard number of lines, which must be thrown forward, and the make-up of each page changed to the end of the chapter.

Competent proofreaders in the best offices frequently call the attention of the author to errors in dates, figures, or proper names, but this should always be regarded by the author as a courtesy rather than as something which the printer is expected to do. The proofreader, on the other hand, is supposed to have corrected every typographical error, and for the author to mark corrections which have been overlooked is a courtesy on the part of the author. The fact that the author or editor has passed over typographical errors in no way relieves the proofreader of his responsibility.

The proofreader is expected to correct any obvious error without hesitation, but to make no other changes. If he thinks a change should be made, it will take the form of a query in the margin to the author. The author should carefully note all such queries, and answer them or strike them out, bearing in mind that if he accepts the query the change necessitated in the type becomes an author's correction, the expense of which falls upon the publisher.

Any marks on the proofs for correction should be made distinct by drawing a short line through the letter to be changed, etc., placing in the margin the recognized sign indicating the change, exactly opposite the line in which the change is to be made, and in the order in which the necessary alterations occur. In doing this be sure to write legibly, and do not cover the proof with lines and counter-lines.

The author should familiarize himself with the standard proofreading marks, and employ these in marking all corrections upon the proofs which are sent him. These marks are as follows:

THE above marks are the ones most generally used in proofreading. There are many others that are required in different classes of work, but these are in the main self-explanatory. This display of proof marks and their meanings has been prepared for The Graphic Arts and endorsed by the Boston Proofreaders Association.