Chapter V

The Typescript Book or Memorandum: Reproduction Techniques

Word count: 17100

I. Carbon Paper Copying

∣📄 p.41 Carbon paper offers one of the oldest and cheapest means of multiplying typescript in a very few copies. It is constantly used for thesis work. With ordinary skill and an ordinary typewriter, two carbons can be made with high legibility and at a maximum price of $.05 per page per extra carbon. The number of legible copies can be increased by increasing the force of the typewriter stroke or diminishing the thickness of the paper. The Electromatic typewriter, which uses electric power in making the impression, exhibits great efficiencies here. The following table indicates the number of carbons that can be made with one typing and the corresponding weights of carbon paper and copy paper.

The thin paper will increase the number of legible copies, but it is useful only for multiplying ephemeral material; it will not make a book paper. The limit to the number of carbon copies that can be made into typescript books is set by the paper. A paper thin enough to receive with fair legibility the seventh copy from an ordinary typewriter, or the twenty-fifth from an Electromatic, would be too thin for book use.

Of the two elements of first cost, the composition costs are the labor costs of the fair copy typing; the area costs, which vary with the number of pages and hence with the number of words put on a sin∥ gle page, are those of the paper on which the fair copy is made; the running costs are those incurred in producing additional carbons.

For purposes of convenience in estimating composition costs for ordinary typing of manuscripts, a normal cost has been estimated on the basis of labor at $.40 per hour, paper for the fair copy typing at $1.25 per 1000 sheets, onionskin at $2.50 per 1000 sheets, and carbon paper at $2.50 per 100 sheets. Quotations received by the Joint Committee show that prices and costs are highly variable, as indicated by Table XIII on the following page.

A normal cost may be taken as $.40 per 1000 words and an additional charge of $.02½ per page per extra carbon, with the usual 20% discount given on materials for manuscripts of over 10,000 words and no charge for the materials required to produce the fair copy and the first carbon. In both cases the format chosen is the usual 300-word page of double-spaced pica typescript.

Before the technological limit of carbon paper copying is reached, the border is crossed to the field of the hectograph.

∣📄p.[42a] (blank) ∣📄 p.[42b]

II. The Hectograph Process

In this report the term hectographing is used to designate any process by which approximately 100 to 150 copies of a document may be reproduced in methyl-violet ∣📄 p.42 ∥ dye from a single master sheet.

∣📄p.[42c] (blank) ∣📄 p.[42d]

∣📄 p.43

a) Gelatin Hectograph Process

When a sheet of paper (which we will call a master sheet) bearing a text written or typed in methyl violet dye is pressed down upon a gelatin surface, the methyl-violet image transfers itself to the gelatin. If fresh sheets of paper are then successively pressed against the gelatin, the methyl-violet image transfers itself to them. Thus a number of copies of the original are produced.

Equipment for the gelatin process ranges from the little gelatin pan, on which the menu of the thirty-five cent lunch restaurant is made, to a power-run machine with a rotary drum resembling the mimeograph. The Ditto, the Speedograph, the Vivid Duplicator, and the Standard Rotary Duplicator are among the best-known gelatin process machines. They vary in price from $15.00 for a flat-bed machine to $475.00 for a power-operated rotary one. The gelatin for the rotary models is furnished in the form of films 9″ x 15″ in size, which are fastened to the drum of the machine as stencils are fastened on the mimeo∥ graph cylinder; the larger, flat-bed equipment employs a roll with a cloth or fiber backing, 9″ to 18″ wide and about 15′ long.

b) Liquid Hectograph Process

A second type of hectograph equipment dispenses with the gelatin. A piece of hectograph carbon paper is placed behind the master sheet, with the carbon side toward the master paper while the text is being written or typed. A reversed carbon image is thus produced on the back of the master sheet. The master sheet is fastened to the cylinder of the machine and the copy is reproduced by bringing the side bearing the carbon image in contact with fresh sheets of paper, which are moistened with a special duplicating fluid as they are fed into the machine.

There are three kinds of liquid process equipment: the Ormig (Figure XX), made in Germany and sold for $160.00 to $475.00; the Standard New Process Duplicator (Figure XXI), which sells for $187.50 to $440.00; and the Ditto Liquid Duplicator, fixed at $200.00 for the hand-run model and $265.00 for the power-run machine.

∣📄 p.44 The Joint Committee has experimented with both the liquid process and the gelatin machines. The result of extensive comparison in the practical working of the two processes is definitely in favor of the liquid process as against the gelatin for the following reasons:

1) Quality of Product: In the gelatin method, copies in the same run are never uniform in shade and are always lighter than the corresponding copies done by the liquid process. The age, condition, and preparation of the gelatin have a much more important effect on the quality of the copies done by the gelatin method than do the same factors of the carbon in the liquid process.

2) Permanence: Copies produced by the liquid process are more permanent than those produced by the gelatin method, because the liquid practically dyes the ink into the fibers of the paper, while the gelatin merely deposits the dye on the surface of the paper.

3) Speed: The rotary liquid machine is capable of producing many more copies per minute than the flat-bed or rotary gelatin equipment. The time for preparing for the gelatin run is almost equal to the actual running time of the liquid machine.

4) Simplicity of Operation: With the liquid method, only three operations are necessary—typing the original, clipping it in place, and turning the handle, or pressing the foot pedal, to produce the copies. The gelatin machine requires much more preparation.

5) Reprint Economies: Reprints from the original master sheet are feasible with the liquid process (but not recommended for a very long run); they are impossible with the gelatin method.

6) Gelatin Supply: It is necessary to have on hand as many fresh gelatin films as there are master sheets to be run in close succession, for a used film must be allowed to stand from six to twenty-four hours before the dye is sufficiently absorbed and dissipated to allow a new master image to be imposed upon it. If a 100-page ∥ document were to be hectographed by the gelatin process in a single day, a supply of at least fifty films would be necessary.

The variable elements in the gelatin process are so much greater than the variables in the liquid process that exact cost comparisons are unreliable. There are no decisive reasons for expecting the gelatin process to operate at a substantially lower cost than the liquid method.

The prices of special supplies for the gelatin process are listed below. Where the gelatin process equipment is already available, it can be used; but the calculation of the place of hectographing in the reproduction of typescript is based on the use of the liquid process.

Cost Analysis of Hectographing

(a) Composition and Running Costs

Composition and running costs for hectographing require little analysis. Fair and format typings are merged, and composition costs for typing may be figured at the normal typing rates. Labor, copy paper, and liquid charges comprise the running costs. Liquid for the hectograph process is sold for $2.50 to $3.50 per gallon. A gallon will serve to moisten at least 6,000 sheets of copy paper or as many as 10,000 sheets; it makes no difference whether these are run as 25 copies each of 240 separate pages of text, or 150 copies each of 40 different masters. The highly calendered copy paper required for the best results may be purchased for $1.65 in quantities of 1000 sheets, with reductions in.price starting at 10,000 sheets. For less attractive but perfectly legible work, ordinary bond paper at a slightly lower price per 1000 sheets is satisfactory, but as it absorbs more fluid than the regular copy paper it is really not economical.

∣📄 p.[44a]

∣📄p.[44b] (blank) ∣📄 p.[44c]

∣📄p.[44d] (blank) ∣📄 p.[44e]

∣📄p.[44f] (blank) ∣📄 p.[44g]

∣📄p.[44h] (blank)

(b) Area Costs and Legibility

Area costs are those of the master sheets ($1.50 to $3.00 per 1000 sheets) and the hectograph carbon paper ($.04 to $.06 per sheet in quantities of 100). These ∣📄 p.45 costs are the ones most subject to variation and they are the most important on account of their relationship to legibility. If a saving in area costs is attempted by using single rather than double-spaced typescript to get more words on the master sheet, the legibility of the product diminishes accordingly, but a more important consideration is the efficient use of the hectograph carbon paper. For the most legible results carbon paper should be used only once. Of course if the typewriter is set for double-spacing, the carbon may be used the second time with no decrease in legibility by typing between the lines of the first copy. Carbon paper may be used as many as three times, but with greatly decreasing contrast and sharpness of definition in the letters produced. The following summary of the cost of the dye in relation to legibility may be made:

Length of run is also a factor in determining the legibility of hectographed matter: the first few gelatin copies are brighter than the later ones, the diminishing contrast being due to the gradual exhaustion of the supply of dye; with the liquid process the decrease in legibility begins after the 100 or 150 mark is reached. This diminution in contrast may be regulated somewhat, however, as the operation of the liquid process machines may be modi∥ fied to yield a few more brilliant copies or a greater number of paler ones, and as the carbon papers are sold in various grades—i.e., hard, medium, and soft—adapted for runs of varying length.

The efficiency range of the hectograph begins with the five or six copies where it crosses the carbon paper technique and runs up toward 200. Bound in place here are four samples of hectograph work. They are numbered to indicate their place in the run. Figure XXII is made with the gelatin process; note that the letters are smoother and more blurred than those made by the liquid process. Figures XXIIIa and XXIIIb are made with the Standard New Process Duplicator, one page being taken from an early place in a run, the other being selected from a later part of the run. A comparison of these two pages will show the difference in legibility of work taken from two different places in runs. Figure XXIV is made with the Ormig liquid duplicator. By comparing these four sample pages, the reader will be able to note the variations between the products of the gelatin and liquid hectograph processes, and to estimate the curve of diminishing legibility as the number of copies taken from the master sheet increases.

The table below shows the cost of hectographing by the liquid process a book of 100,000 words, “with typing figured at $.50 per hour, labor for running the duplicating machine figured at $.40 an hour, a fresh sheet of carbon used for each master sheet, and hectograph copy paper used for the run. The best possible hectographed product will result.

∣📄 p.46

c) Application of Hectographing to the Problems of Scholarship

When compared with the average quotation for typing, these figures indicate that the hectograph will distribute the documents of internal circulation among scholars at prices so nearly within the range of those that authors are accustomed to assume in connection with the preparation of their typescript manuscripts, that the cost of giving a limited distribution in hectographed form could be assumed by the writers themselves with little more trouble than they now take in distributing reprints of their articles from periodicals.

An experiment in the use of this method as a means of reproducing doctoral dissertations has been sponsored by the Joint Committee. Dr. Stanton Davis of the Case School of Applied Science in Cleveland prepared by the liquid hectograph process an edition of fifty of a 350-page dissertation, “Pennsylvania Politics, 1860–62, at a unit cost of $1.95. Additional expenses for binding amounted to $32.50, and mailing to eighteen libraries cost $6.72. The reader may examine this work at one of the following libraries:

• Carnegie Library of Pittsburgh
• Cleveland Public Library
• Columbia University Library
• Duke University Library
• Free Library of Philadelphia
• Harvard College Library
• Historical Society of Pennsylvania
• Historical Society of Western Pennsylvania
• Library of Congress
• New York Public Library
• Princeton University Library
• University of California Library
• University of Chicago Library
• University of Pennsylvania Library
• University of Pittsburgh Library
• University of Wisconsin Library
• Western Reserve University Library
• Yale University Library

A few copies are available for sale at the Western Reserve University Bookstore at $3.00 (paper binding) and $3.75 (cloth binding). The hectograph costs, exclusive of binding and mailing, for an edition of fifty were $47.10 more than the cost for a typescript copy and one carbon. ∥

The hectograph process is probably more serviceable, however, as a means of distributing the type of material that would otherwise be printed in pamphlet form or as articles in periodicals. No binding costs other than simple wire stapling need intervene to increase the charge upon the author; hectographed memoranda up to twenty or thirty pages in length would probably find their place of storage in a vertical file. An example of this use of the process is the reproduction by the Joint Committee office of Dr. Robert C. Binkley’s thirty-four-page memorandum, “New Tools, New Recruits for the Republic of Letter,” at a unit cost of $.17.

Hectographing is not recommended as a means of preserving original source material because of the impermanence of the dye, which fades rapidly if exposed to sunlight for any length of time. However, the kind of monographic work that becomes out of date in thirty or forty years may be circulated in hectographed form without loss to scholarship from its impermanence. If a single standard carbon copy is made while typing the hectograph master sheet, this copy can be deposited and held for permanence.

III. Mimeograph Techniques: Mimeographing and Direct-Typing Offset Printing

The mimeograph process multiplies typescript by means of a stencil printing surface on which the letters of the format copy are cut by the typewriter. The stencil is fastened on a cylinder and as sheets of paper are fed into the machine an impression is produced on the copy paper by the squeezing of ink through the abraded portions of the stencil. The equipment ranges in price from $30.00 to $375.00 for a hand-run machine and from $255.00 to $450.00 for an electrically operated model with an automatic feed. The preparation of the printing surface is almost as simple as ordinary typewriting, and the machine can be operated by the lowliest office boy, so that although this form of duplicating is commercialized by service agencies, almost every office of any size possesses a mimeograph of its own. Both costs and prices, therefore, must be analyzed for mimeograph work.

The rotaprint and multilith have been briefly mentioned in Chapter III in ∣📄 p.47 connection with the discussion of photo-offset as a reprinting method. The press plates for reprinting are, of course, prepared photographically. But there is an alternative way of putting on such plates the format copy of the text to be reproduced: the metal sheets that are used on these small offset presses can be inserted in a typewriter as if they were sheets of paper, to receive typescript directly. The making of typescript books from press plates prepared in this manner is called “direct-typing” offset, as compared with photo-offset. Direct-typing offset is much less common than photo-offset but also much cheaper, and it is, therefore, worthy of consideration by any institution possessing sufficient resources to purchase a rotaprint or a multilith for either commercial or non-commercial use. For the benefit of owners and non-owners, both costs and prices will be quoted.

From the standpoint of cost behavior, mimeographing and direct-typing offset are very closely related. They will, therefore, be referred to henceforth in this report as the “mimeograph” techniques, and their costs will be analyzed together.

Composition Cost Analysis

It will be seen that the cost of the printing surface (or the area cost) for the mimeograph processes, whether of stencil or metal sheet, is considerably more than the cost of the printing surface used with the hectograph, and that it differs from ordinary paper in its physical character. There is, therefore, a greater likelihood in mimeographing than in hectographing that there will be three typing processes. The author will have a fair copy made from his draft copy, and then a format typing will be necessary to transfer the text to the metal sheet or stencil. This is not technologically necessary, for the typist could work directly from a draft copy to a mimeograph stencil; but this extra format typing is so often done that most of the quotations presuppose it.

Precisely because the area cost is Significant in mimeographing, the problem of format must be studied; for it becomes important to lay down as much as possible on a given area, without unduly sacrificing legibility and beauty. But it goes with- out saying that any format differing from ∥ the usual one-column, six-line-to-the-inch, pica or elite typescript format will be more expensive to prepare. In estimating the average mimeographing costs, therefore, format typing will be charged at $.80 per 1000 words in order to provide for the slowing up that is necessary to produce an unusual typescript page. The specifications for satisfactory mimeograph formats will be discussed in connection with area costs.

Area Costs and Format

Mimeograph stencils are of five kinds and are sold in two standard sizes, 8½″ x 11″ and 8½″ x 14″. The prices as standardized in the trade are given be- low. Bootleg dealers sometimes sell stencils manufactured in violation of patent rights at half these prices. The patent rights will expire within the next decade, but by that time technology may have moved far beyond the mimeograph. Large government orders are filled at discounts up to 50% in competitive bidding.

The wax stencil is used for runs ∣📄 p.48 not exceeding 500. It cannot be saved, and more care is needed to produce a good copy with the wax stencil than with the mimeotype stencil. A sample of work done on a wax stencil is seen in Figure XXVII. The Cellotype stencil, a white stencil with a carbon back, is used when very high visibility is desired; the dermatype, a wet process stencil, is seldom used and is almost obsolete; the mimeotype stencil, referred to as the “universal stencil,” is used for every purpose except the photographic work. The dermaprint stencil has been discussed at length in Chapter III.

There are two kinds of metal sheets that can be used directly in the typewriter—the rotaplate: and the multilith metal sheet. The claim is made that the multilith sheet can be used over again for about three times; the rotaprint people say that ∥ they have not found any process whereby the rotaplate can be used over again with satisfactory results.

∣📄 p.[48a]

∣📄p.[48b] (blank) ∣📄 p.[48c]

∣📄p.[48d] (blank) ∣📄 p.[48e]

∣📄p.[48f] (blank) ∣📄 p.[48g]

∣📄p.[48h] (blank) ∣📄 p.49

In designing a format for the maximum utilization of this surface five factors must be considered:

By combining these factors in various ways, there might be constructed over fifty different typescript formats. The difference in word-carrying capacity between the two extremes is very great. A hundred thousand words would go on 123 pages of 6½″ x 14″ size, elite type, single spaced, 6½″ lines, narrow-margin format; but would require 562 pages of 8½″ by 11″ size, pica type, double-spaced, 5″ lines, and a wide-margin format.

By a combination of experiment and guesswork, the writers have reached the conclusion that the formats shown in Table XIX are advisable for the number of words per page indicated.

Explanation of Sample Pages and Cost Table

In order to exhibit as many as possible of the different varieties of the mimeograph process, the sample pages following have been run off on mimeograph and bond paper of different weights, from rotaprint and multilith metal sheets and from wax and mimeotype stencils (Figures XXV to XXVIII). The reader must make allowance for the fact that paper stock, format, and process are all on display; and he must decide for himself which differences in the exhibits are due to format variables, which to paper, and which to the reproduc∥ tion methods. He may also judge from these samples whether pica type with special spacing of five lines to the inch is not just as legible as pica with standard double spacing. From a theoretical standpoint, five lines to the inch should be wholly satisfactory. This would mean that nine vertical inches of mimeograph printing surface could just as well take forty-five lines at five lines to the inch, or forty lines at four and one-half lines to the inch, as to take the twenty-seven lines that ordinary double spacing offers.

The five-lines-to-the-inch spacing would let twelve mimeograph stencils do the work of twenty stencils; the four-and- one-half-lines-to-the-inch spacing would let fourteen stencils do the work of twenty. The price of the special spacing equipment can be earned, therefore, by the economies in cost of stencils in a mimeographing job of less than 50,000 words.¹

The cost and prices of mimeographing a manuscript of 100,000 words in each of these formats are estimated in the following chart. But in order to make the comparison of formats very clear, the estimates in this case are based on the use of the same printing surface, namely, the ordinary 8½″ x 11″ mimeotype stencil, except with formats requiring 600 or 700 words to a page, which require an 8½″ x 14″ stencil. In each case the cost figure is given under two alternative ∣📄 p.50 hypotheses regarding composition cost: first, that the format copy is typed directly from the draft copy without an intervening fair copy; and second, that the fair copy is made on paper and then retyped as format copy on the mimeograph stencil. Labor for operating the duplicating machine is estimated at $.40 per hour. The choice of format would control a variation in area costs of 350% from the mini∥ mum and a corresponding variation in running costs. (See Table XX on the following page.)

These are costs based on average estimates for office work; prices vary greatly from this calculated cost basis. The comparison of cost and price, i.e., of work done in an office and work sent out, can be seen in one instance in the comparison as shown in Table XXI, page 52. ∣📄 p.51

∣📄 p.52

The following table shows the costs of producing a book by each of the “mimeograph” processes in a 400-word format page, as displayed in Figure XXVII. Price esti∥ mates are based on several quotations and represent an average from which there may be a variation of 35% for the maximum and 25% for the minimum.

∣📄 p.53

IV. Photo-Offset from Typescript

The photo-offset technique has already been described in connection with the reprinting of books out of print. The same technique is available for the reproduction of typescript. Why should it be used, however, when typescript can be multiplied without an intervening photographic process?

The difference between the direct and indirect making of a typescript printing surface can be illustrated in comparing the two possible ways of using the multilith. By the first method the typist takes a metal sheet, coats it with a preparatory solution, puts it into her typewriter exactly as if it were a piece of paper, types the text on it, clips it to the drum of the multilith and reproduces it. By the second method the typist makes a copy with a strong black ribbon on paper. An operator photographs this typescript page, securing a negative on film or paper—in this case usually paper. Then the photographic negative is laid on the metal, which has meanwhile been sensitized, and strong arc lights burn in the text from the photographic negative on to the metal plate. Then at last the metal plate is put on the drum, and the multiplying of copies can begin. The additional work increases not only the labor time but also the labor rate, because the additional photographic skill is necessary. Why should the costs of all these intermediate steps be borne? Why should the area cost be so greatly increased when the end product is still a typescript book?

Justification for Increased Area Cost

If the text includes illustrations or characters that must be filled in by hand, or certain kinds of tabular or charted data, the reason for using the photographic step is clear. If, however, the text is simply typescript and nothing else, ∥ the justification of the extra cost must be found elsewhere.

There are three reasons for incurring this additional area cost: first, the plates of the larger offset presses are too big to be put into the typewriter; second, the better photo-offset presses will distribute ink more evenly and make finer letters than the mimeograph techniques (including direct-typing offset) will produce; third, the standard typescript relations of type size, length of line, interlinear spacing, and size of page are not related to the normal 10″ spacing of library shelving in a way that results in the most efficient use of page areas and bookshelf footage.

Photographic Reduction of Typescript: (a) Economies in Type Size

Typescript is wasteful of page areas because of the large size of the type. Photo-offset permits a photographic reduction of typescript in any degree.² Therefore, by the photo-offset process a typescript can be brought down to the equivalent of 8- and 10-point type as used in books, or to the equivalent of 7-point newsprint. A table of equivalents was suggested in Chapter IV, by which various reductions of typescript could be correlated with corresponding point sizes of book type by a photographic process.

The typewriting is done on thin paper backed with carbon so that a heavy black letter, reinforced from the back, is made. The copy is then transferred to the photographic negative paper directly, without using a camera. The National Process Company uses this method among others. From the standpoint of design of a typescript book format, its importance is only that it brings the photolithographic metal sheet nearer the typewriter. In this respect it is related to the rotaprint or multilith processes of direct typing on the metal.

There has been one direct study of the legibility of reduced typescript by Professor Greene of Michigan. Professor Greene came to the conclusion that what he called “seven-point typewriter type” (i.e., pica typescript reduced to 70% of full size so that it laid down fourteen letters to the inch instead of ten) was just as legible as full-sized pica typescript.³

The reader should be given an opportunity to test this judgment and to compare the legibility of typescript in various reductions with the legibility of book-face type. To make the comparison possible, six ∣📄 p.54 areas of four square inches each are filled with reading matter (Figure XXIX). The first square is filled with text in 10-point type, and typescript that will lay down the same number of words in the same area is set beside it; the second pair of squares is filled with pica typescript reduced to 80% of normal size, and with ordinary printing in 7-point type, which fills the same area with the same number of words; the third pair of squares shows that pica typescript must be reduced to 70% of full size to put the same number of words in the area that will be put there by ordinary newsprint. The reader may judge the comparative legibilities.

Whatever conclusions the reader may draw from these illustrations and from the exhibits of standard photo-offset typescript formats as to the permissible amount of reduction which pica typescript will take without sacrifice of legibility, the essential fact stands that reduction in size of typescript increases very rapidly the number of words that can be laid on a given page area. While the linear dimensions of the typescript are falling by an arithmetical ratio, the number of words per page is increasing by a geometrical ratio. This is the same situation that is encountered in any photographic reproduction process in which a reduction of the size of the writing is involved.

Photographic Reduction of Typescript: (b) Length of Line and Economy

There is an additional feature that makes this consideration especially important in the case of typescript, and that is the effect of reduction on length of line. there is very little reason to carry out a reduction of typescript size if a single column of typescript is still to be carried on a sheet of paper 8½″ x 11″, because so much of what is gained in the use of space by diminishing the type size is simply turned over to wider margins. But here two possibilities intervene: the reading matter may be typed in two or more columns and reproduced on the standard typewriter size page, or the size of the page itself may be reduced to something more nearly resembling that of an ordinary octavo book.

But since an octavo page, 6″ x 9″, with lines of maximum length (3.9″ for 6– to 10–point type, according to Vernon) ∥ sacrifices 35% of its surface to right- and left-hand margins, an 8½″ x 11″ page, with two columns each 3 ¼″ wide, which sacrifices only 23½% to right- and left-hand margins, is preferable. The difference is increased by the fact that a larger side margin usually calls for larger top and bottom margins. Since the cost of printing surface is an area cost and the margin space costs just as much per square inch as the part of the page that is utilized for text, it is especially important in planning an economy format for a photo-offset typescript book that the margins shall not be made any larger than necessary. It is interesting to note in this connection that Edwards Brothers propose, for an 8½″ x 11″ page, a two-column format with columns 3½″ wide, and the Donnelley & Sons Company suggests a similar format.

With these facts in mind, one may lay down the conditions under which it may be profitable to reproduce typescript by photo-offset rather than by the mimeograph process: by the increased utilization of paper and printing surface owing to reduction in size of type, and the diminution of the proportion of area given over to margins, the photo-offset process will at some point reach a cost figure that will make up for some of the additional costs of the photographic element in the process. If, for instance, the area cost of making a photo-offset page is three times the area cost of making a mimeograph page, the area cost per word will be the same only if six times as much reading matter can be put on the photo-offset page.

Specifications for Reproducing the Same Book at the Same Cost by Mimeographing and by Photo-Offset

Cost comparisons between photo-offset and mimeographing as means of reproducing typescript are complicated by the fact that photo-offset is usually shopwork sent out on contract, and, therefore, costs to the publisher must be calculated on the basis of actual price quotations by firms engaged in the printing business; while the mimeograph process is so much a part of office routine that its costs can be calculated in terms of labor and materials. The most appropriate comparisons, therefore, are those that stand between commercial mimeographing as quoted by office service shops ∣📄 p.[54a] and commercial photo-offset quotations.

∣📄p.[54b] (blank) ∣📄 p.55

A further complication increasing the difficulty of cost comparison is the fact that any format devised to secure a maximum use of the paper surface is likely to require a special format typing; the fair-copy typing operation cannot so easily be merged with the format-copy typing as in the processes which use a simpler format. In both cases, therefore, the cost of a format typing from the fair copy will be figured into the estimates made.

An extra operation required to justify the right-hand margin also calls for calculation, whether the method used should be that of a second format typing or the special paper invented by Spielvogel (see Chapter IV). The cost of this operation will not be figured into the estimates.

Taking all of these factors into account, it seems that mimeographing 50,000 words in a format that permits 300 words to the 8½″ x 11″ page would cost about the Same as reproducing the same number of words by photo-offset in a format that reproduces 1786 words to the 8½″ x 11″ page.

As in all problems of reproducing research materials, the rule holds that between certain limits an increase in attractiveness of text and format accompanies an increase in cost. Above the level at which photo-offset typescript costs the same as mimeographing, there are other photo-offset formats available at costs that increase until they come to exceed the cost of efficiency formats in printing.

Photo-Offset Typescript Format: Full Utilization of Printing Surface

In planning formats for the photo-offset book there must be considered a fixed set of dimensions not those of the typewriter page–namely, the dimensions of a metal sheet of the offset press. Though Commercial quotations often ignore this factor, it actually enters into the cost in avery material way, for whenever a format is so designed that it fails to use the full available printing surface of the photo-offset metal sheet, energy and material are wasted. The sensitizing and preparation of the metal sheet are substantially the same whether all or merely a part of the surface is used; the waste that occurs ∥ when the page is not planned to fit the metal sheet has been discussed in Chapter III in connection with reprinting.

The ideal typescript format, from the standpoint of printshop efficiency, would be based on a page size that would divide evenly into the area of one of the larger press plates. The plates most commonly in use will take paper sheets of the following dimensions:

• 22″ x 34″
• 28″ x 42″
• 35″ x 45″
• 36″ x 48″
• 38″ x 50″
• 44″ x 64″
• 59″ x 72″

The printers will sometimes say that they are controlled by the standard sizes in which book paper is available, and that it will not pay to print in any format that calls for cutting off and wasting any considerable portion of their standard paper sheets, which are usually:

• 17″ x 22″
• 25″ x 38″
• 28″ x 42″
• 32″ x 44″
• 35″ x 45″
• 38″ x 50″
• 42″ x 56″
• 36″ x 48″

This consideration is of great importance in the case of a long run, but if only 200 or 300 copies of a book are desired, the cost of the paper is so slight an element in the total cost that paper economy may in some cases be sacrificed in the interest of printing surface economy.

The relative importance of economies in utilization of metal sheet area and economies in utilization of paper is established by the fact that the first is a saving in area cost, the second a saving in running cost. The smaller the edition, the larger is the proportion of area cost and the smaller the proportion of running cost, and hence of cost of paper. In the small editions—around 100—the area cost so far preponderates over running cost that some firms give the same quotation for 100 copies as for 50 copies. In such editions wasteful use of paper is ∣📄 p.56 justified if it will bring equivalent economies in the use of the photo-offset metal sheet. The exact point at which it will cease to pay to waste paper for the sake of saving metal sheet printing surface can always be calculated by estimating the total price of the paper wasted and comparing it with the area cost saved.

What then is an ideal format for the photo-offset typescript book? The standard octavo printing formats (according to published figures on binding) of 6″ x 9″ and 6 ¼″ x 9 ¼″ are ill adapted to offset work because they lead either to an undue extent of margin, an undue length of line, or an inadequate use of the reducing power of the photographic process. A column of maximum width (3.9″) can be printed in the 5½″ x 8½″ page with only 42.4% of the area surrendered to margins, but this volume is too small for convenient library shelving. The best format for library shelving convenience and low margin ratio would be just short of 10″ high, and the narrowest page that would accept two columns would be 7 ¾″ wide. Such a page would take two columns of minimum width (2.9″) with ½″ between them, and external margins of about ¾″, The percentage of area given over to margins would be about 26%. The book page would have an unusual appearance—more nearly square than an ordinary octavo page. Yet it would be very efficient (Figure XXX).

Now the curious fact emerges that despite all the variety of sizes available in photo-offset press, there is none that will take without waste of printing surface a page 9 ¾″ x 7 ¾″. Thus the limitations of equipment impose themselves in planning a model photo-offset page.

Planning a photo-offset page is more intricate than planning a mimeograph page because, while all the variants of length of line, interlinear spacing, margins, size of type, and number of columns are present, there are added the additional factors of a photographic reduction ratio and a layout on a photo-offset metal sheet. The format copy is typed on one size of page, the book appears in another size of page. Moreover, while the use of book type in planning a page has been made the subject of much study, the use of typescript with photo-offset has been left to rule of thumb. ∥

To illustrate the planning of a photo-offset page, an exercise is here carried out in calculating a format for a book of 50,000 words to meet two different cost situations: first, to cost as much in an edition of 500 as low-cost printing; second, to cost as much in an edition of 500 as medium-cost mimeographing. By consulting the tables in Chapter III, it appears that an edition of this size, in an “economy format,” can be printed at $365.00 (6 ¾″ x 9½″ page size, two columns); and the same text, mimeographed in a 400-word format, will cost $158.00 (see page 51).

The question may then be asked, is photo-offset as expensive as printing? Is it as cheap as mimeographing? The answer is that it is both, and that the cost depends on the format. The format is itself the determinant of legibility, and anyone with a typewriter and a supply of paper can plan a format that will meet any price level from the lowest to the highest.

The general formula for finding a photo-offset format that will reproduce a given number of words for a given sum of money in an edition of a given size can be worked out from data already presented in this volume.

Given N (a number of words) and D (a sum of money available), from the information in Chapter IV on costs of typing, calculate the cost of making a format copy of this number of words, which will be D’ (the composition cost). D minus D’ will be available for photo-offset work—the area and running cost. From the tables in Chapter III, calculate the number and size of photo-offset pages that can be produced in the required number of copies for D minus D’ dollars. Then lay off on these pages the desired margins, make a decision on the number of columns of print, and conclude with the dimensions in inches of space to be filled with text. Let L be the length of the column, W the width of the column, P the number of pages, and C the number of columns to a page.

N/CP is the number of words per column. The type of text may lead to the conclusion that the number of keystrokes per word is greater or smaller than six, but six can be taken as a good average. To keep the formula perfectly general, however, let K be the number of keystrokes per word (including allowance for inter-word spacing ∣📄 p.[56a] and loss of keystrokes at the uneven right-hand margin). Then NK/CP is the number of keystrokes per column.

∣📄p.[56b] (blank) ∣📄 p.57

The formula for making NK/CP keystrokes on a reduced size (or enlarged size) photo-offset column of given dimensions L x W inches can be worked out for any typewriter.

Let A be the number of keystrokes per inch laid down by the typewriter (so that with a pica machine, A equals 10; with an elite machine, A equals 12; with a micro-elite machine, A equals 14; with a “nine-pitch” machine, A equals 9); and let B stand for the number of lines to the inch (which with ordinary single spacing is six, with double spacing is three, and with special spacings may be four, four and one-half, five, and five and one-half).

Given chosen values for A and B, it remains to find W’ and L’, the width and length of the original typescript copy of a column from which the photo-offset page is to be photographically copied, in a reduction ratio or enlargement ratio which will have the value R.

It is known that \({W \over L} = {W' \over L'}\); and \({AW' \times BL'} = {NK \over CP}\). \({L'} = {NKL \over CPABW }\); \(W' = {NKW \over CPABLE }\); therefore, \({L \over L'} = {W \over W'} = R\).

∥ The composition costs of typescript books, it has been proved, are highly variable; the principal levels for a book of 50,000 words are determined by the five situations shown in Table XXIII.

To simplify cost calculations, composition cost will be taken for the third situation only—i.e., for typing done by a typing bureau at an average price. The typescript composition cost for photo-offset will therefore stand at $62.50 for 50,000 words. Although the differences in format will react upon this composition cost in some measure, this influence will be ignored. Substracting, then, the composition cost from the two total sums suggested as alternatives (i.e., price of a mimeographed and price of a printed book), the following amounts remain for the production of a photo-offset book, exclusive of typing:

∣📄 p.58 In Chapter III area costs of photo-offset work are found. These indicate that the sums listed above will purchase the following amounts of page area in a photo-offset edition of 500 copies:

For $95.50 can be bought 500 copies of twenty-eight pages, 8½″ x 11″, which will require 1786 words to a page to reproduce 50,000 words; or, $95.50 will buy 500 copies of forty-four pages, 6″ x 9½″ (twenty-two printings from a metal sheet 9½″ x 12″), which will require 1136 words per page to reproduce 50,000 words.

How can the typescript copy be prepared to lay this number of words on a page in a format satisfactory for photo-offset in these sizes of metal sheet or paper?

The arrangement of 6″ x 9½″, less ½″ margin on top, bottom and outside, and with a ¾″ margin on the bound edge, can be made in this manner: two columns, each column 8½″ high, carrying 568 words, in column width not greater than 2 ⅛″, leaving ½″ between columns. Such a column will roughly resemble newsprint in width. Counting 6 keystrokes per word to allow for spacing and loss on the right-hand margin, the number of keystrokes must be 3408 in a column 8½″ x 2 ⅛″. An interlinear spacing of five lines to the inch with pica type would be used.

In full pica size, 8½″ will be filled with forty-two lines at five lines to the inch, and 2 ⅛″ with 21 keystrokes; at full size, therefore, this area will accept 882 keystrokes. But since it must accept 5408 keystrokes in reduced size, the reduction must be such that four times this number of keystrokes will go on this area. The square root of four is two. The pica typescript must, therefore, stand at a reduction of two diameters. Since 42 keystrokes reduced two diameters will fill the 2⅛″ column, the typewriter should be set to make columns of this width (42 keystrokes, or 4.2″) and carried to a column length of 17″. When this is reduced photographically, it is a reduction of approximately 50%. (Figure XXXI.)

In the 8½″ x 11″ size of page the necessary 1786 words would probably best be set in three columns. Making margin allowances of ¾″ inside margin, ½″ between columns and for outer margin, and 1″ on the top and bottom, the effective width of the columns would be 2.08″ and the ∥ height would be 9″. Each column would be required to take 595 words, or 3570 keystrokes. A reduction ratio of 50% will be necessary. Pica typescript reduced to 50% of original size will suffice in this format. In the format copying, the typist should make each line 43 keystrokes long (4.3″) and the column should be typed to 18″ in height. (Figure XXXII.)

Thus it appears that there are at least two formats in which a photo-offset copy of a 50,000-word text could be made for a price equivalent to the cost of a mimeograph copy. The choice of an 8½″ x 11″ format would involve reduction of pica typescript by 50%; the choice of the 6″ x 9½″ format would also mean a reduction of 50%. The typescript format copy for the 6″ x 9½″ page could be prepared on a typewriter with a 12″-carriage (two columns, each column 4.2″, with ½″ margin between them); the format copy for the 8½″ x 11″ page would call for a typewriter carriage that would make three columns, each 4.5″, with two internal margins of ½″, and hence a 15″-carriage would be necessary. Of course the typing could be done one column at a time on the typewriter, and the columns then mounted together to make the format copy.

The essential significance of this whole demonstration is, first, that photo-offset formats can be designed to bring photo-offset costs down to the level of mimeographing; and second, that when formats are so designed to bring the photo-offset costs to this level, the legibility diminishes below that of the corresponding mimeograph copy. Pica typescript reductions of 50% are close to the lower limit of legibility without a reading glass.

If the composition cost is set at a lower level by basing it upon a wage rate in an office, the number of 8½″ x 11″ pages upon which the 50,000 words would be spread would be increased from twenty-eight to thirty-two, because there would be more money left over from typescript composition to pay for the photo-offset work. It would then be necessary to put only 1560 words on a page instead of 1786, only 520 words in a column instead of 595. This could be accomplished by reducing the typescript only 47%, to 53% of its full size. ∣📄 p.[58a]

∣📄p.[58b] (blank) ∣📄 p.[58c]

∣📄p.[58d] (blank) ∣📄p.[58e] (blank) ∣📄 p.[58f]

∣📄p.[58g] (blank) ∣📄 p.[58h]

∣📄p.[58i] (blank) ∣📄 p.[58j]

It often happens that a manuscript is submitted to a photo-offset shop and a ∣📄 p.59 printing shop for quotations, and that the quotations for a printed product are about the same as those for a photo-offset book from typescript. This is invariably the result of the choice of format for the photo-offset book. The printing prices, as has been proved in Chapter III above, cannot respond as actively to change in format as the photo-offset prices. To ask “is photo-offset as expensive as printing?” is to avoid the essential element in photo-offset pricing, namely area cost and reduction ratio of the typescript. By calculations corresponding to those used above, it is easy to determine a format in which a typescript book, reproduced in 500 copies by photo-offset, will cost as much as the same book reproduced in print.

If the $365 that would be required to print 50,000 words is devoted to making a photo-offset reproduction from typescript, $62.50 of it must be allotted to typescript composition, leaving $302.50 to pay for the photo-offset work itself. This will pay for 500 copies of 69 pages, 9½″ x 12″, or 138 pages, 6″ x 9½″. It would be necessary to put no more than 725 words on each 9½″ x 12″ page, or 368 words on each 6″ x 9½″ page. Leaving wide margins and typing in a single column with five-lines-to-the-inch spacing, a reduction of 24%, to 76% of full size, would take care of the 9½″ x 12″ page. The typist would simply set the stops to write 85 keystrokes to a line (8.56″), type 60 lines on a page, and ask the photo-offset printer to reduce the page to 76% of full size. On the 6″ x 9½″ page, the typist would set the stops to write 54 to 55 keystrokes, type 47 lines at four-lines-to-the-inch Spacing, and ask the printer to reduce it 33% to 67% of full size. (See Figure XXXIII.) The resulting pages would be very neat, very legible, and a very expensive way of using the photo-offset process.

If it should happen that the format copy were then sent for quotation to a photo-offset printer who was using a press with a 17″ x 22″ metal sheet, his quotation for the job would be substantially higher than the figure of $302.50 that is the basis of this calculation, because he would have to run the whole 17″ x 22″ to get four pages 6″ x 9½″, and the wastage of his metal sheet printing surface would be almost 40%. This fact is brought into the ∥ picture to emphasize the importance of considering, in the planning of a format, not only the arrangement of typescript columns that will lay the required number of words on a page, but also the relation of the page size chosen to the dimensions of the metal sheet of the press upon which it is to be run.

In the first edition of this manual, the writer gathered quotations on the manufacture of photo-offset typescript books and fitted them to a chart in which they were compared to the manufacturing prices of printed books. It has become evident, as a result of further study, that this way of presenting the comparison is completely misleading because it ignores the vital factor of format. Therefore, in this analysis of the use of photo-offset to reproduce typescript, several formats are designed, and the manufacturing prices of a book in each format are calculated. These prices are sufficiently checked with actual quotations to make them indicative of the general price levels at which photo-offset can be used. (See Figures XXXIV and XXXV.)

Photo-Offset from Normal Fair Copy, without Additional Format Typing

It will often happen that a fair copy of a manuscript has been made in the ordinary typescript format, 300 words to the page. This question then arises: shall the text be retyped as a special format copy for photo-offset reproduction, or shall the fair copy be used as a format copy? This is a special case of the problem of controlling typescript composition costs by merging fair and format typing operations. It can now be given a specific answer.

A manuscript of 50,000 words, typed in the normal 300-word format, fills 166 pages. To retype this manuscript in a special format for photo-offset reproduction would cost $62.50. Now this $62.50, if diverted from the payment of composition cost to the payment of area cost, will buy nineteen extra pages, 8½″ x 11″, or 28 extra pages, 6″ x 9½″, in an edition of 500. If this $62.50, instead of going to composition, is actually paid to the offset printer, the text can be spread over this additional amount of page area without changing the total cost of the edition. The additional page area released to the photo-offset by an elimination of the retyping ∣📄 p.60 process is the same, regardless of the total amount spent for the publication. Whether the book is to be published for $158.00 or for $365.00, it will still be true that the money diverted from composition cost will buy exactly the same amount of additional photo-offset surface.

Will it pay to make this diversion? Will retyping so greatly increase the efficiency in the use of photo-offset page area as to justify reducing the number of available pages upon which the text is to be reproduced? This special case of the fair copy-format copy problem is particularly important, because the 300-word format is the normal one in which typescripts are prepared, and in which they are presented to editors or bound as typescript dissertations.

It has been shown above that the 50,000 word text can be retyped and reproduced in a 6″ x 9½″ or an 8½″ x 11″ book for $158.00 (the same price as mimeographing) by reducing the pica typescript to 50%. of full size.

If the entire $158.00 is spent for photo-offset work, it will buy six full sheets 22″ x 34″ ($153.66). These six sheets will divide into 168 pages of 5½″ x 4 ⁶⁄₇″. This particular way of folding the sheet will cause an additional expense in gathering, but the extra expense will probably not exceed the difference between $153.66 and $158.00. The normal typescript pages, reduced to go on pages of this size, must pass through a reduction of 45%, leaving them 55% of full size. The page would then have a ½″ margin top and bottom, and side margins of nearly an inch. This reduction ratio would be the same if the typescript pages were reduced to go on pages 5½″ x 4 ¼″, made by folding 8½″ x 11″ sheets into four, and costing, for the edition, $133.40. The difference between $133.40 and $153.66 would be an unnecessary expense for needless width of side margins. The alternatives in the “near-print” reproduction of a fair copy of normal typescript are therefore:

1. To retype and mimeograph ($158.00)
2. To photo-offset without retyping, reducing to 55% of full size ($133.40) ∥
3. To retype and photo-offset, reducing to 50% of full size ($158.00).

Another standard price level is suggested by the dimensions and word count of the normal typescript page, because of the importance of utilizing the full area of the photo-offset metal sheet. This is the size 8½″ x 5½″, obtained by reducing a full page of typescript to a half page of photo-offset. Metal sheets that will take an 8½″ x 11″ page without wastage will also take the 8½″ x 5½″ without wastage. If top and bottom margins of ¾″ are left on the photo-offset page, the reduction of the typescript will be 23%, to 77% of full size, and the cost of the edition of 500 will be $265.00. This is a price level at which the innumerable bound theses that fill the shelves of graduate schools could be brought conveniently into print without retyping.

If the sum of $265.00 is to be spent to reproduce typescript in a retyped format, it can be made to yield a two-column page 8½″ x 11″, with each column 3.6″ wide, and ½″ margins top and bottom, in exactly the same reduction, namely 23%.

Therefore the cost of reproducing a typescript by photo-offset in a reduction of 23% would be the same, whether or not the additional format typing were done.⁴ The saving of page area involved in shifting from one column to two, and from normal double spacing to special spacing four and one-half-to-the-inch, and from the ¾″ to the ½″ top and bottom margin, would exactly pay the cost of the format typing.

These figures make possible an interesting comparison with the actual price quotations offered by Edwards Brothers. Edwards Brothers offer to type and manufacture an edition of 500 copies of 50,000 words for $265.00. The calculations above and the format designed above to bring a retyped book to the same price of manufacture as a book reproduced directly from a normal fair copy of 8½″ x 5½™ photo-offset pages, are based on an average quotation for typing and on the photo-offset shop quotations made by Hoch. The agreement with the Edwards figures seems very near, for it would appear that Edwards ∣📄 p.61 Brothers would also find a point where the economies of a compressed format on a page would make up for the extra cost of a format typing. The reduction in both cases is not far from 25% (30%, 23%).

The prices charged by Edwards Brothers have been widely published in two ∥ pamphlets: Junior Models for Lithoprinting, and Manual of Lithoprinting. These quotations are listed herewith: the figures for price of edition and price per page have been calculated by the author from the Edwards figures.

∣📄 p.62 The Edwards’ model 700-D requires seventy-two pages for the 50,000 words; the Edwards! quotations for typing in this format are $.78 per 1000 words instead of the $1.25 which is the average figure. The composition cost is figured at $39.00 for the 700-D model with 50,000 words. The price of an edition of 500 copies typed by the author is $225.00; the price of an edition typed at the shop is about $39.00 more (or $264.00). Edwards quotes $265.00 for the shop-typed copy.

There are other contrasts to be examined. The Hoch quotations for 500 copies are page 6½″ x 11″ stand at $5,205 Edwards prints this amount of matter, collates, and includes a cheap binding for $3.68. But the Hoch quotations show a constant proportionate relationship between the prices of full metal sheets and half sheets. If Edwards Brothers sets up the 8½″ x 11″ area in the form of two pages, 8½″ x 5½″, the price is equivalent to $3.92 instead of Hoch’s $3.20. Of course the collating and binding cost counts for something in this comparison, but the figures reveal nonetheless a certain elasticity in costs per square inch of page (area costs) and in running cost behavior.

The buyer of photo-offset service can conclude from these illustrations that it is wise to secure quotations separated into the three parts:

1. Area costs (page area chosen to fit metal sheets)
2. Running costs (quoted as a constant figure per additional hundreds of copies)
3. Typing costs (quoted in thousands of words, or in pages of a specific format)

With these quotations in hand, he can divide the work and plan his format to produce maximum legibility at minimum cost.

Typing Instructions

Since these tables reveal a considerable difference in costs and prices dependent upon the character of a composition cost, instructions for the preparation of a format copy for the photo-offset typescript book are properly a part of this chapter.

Since the typescript copy must be a photographic object it should be as black as possible. Some companies recommend the ∥ use of a special bronze ribbon; in all cases the ribbon should be fresh and black. To avoid an uneveness in the impression produced by the fabric of the typewriter ribbon a special silk ribbon or one made of carbon paper may be used. The typewriter type must be perfectly clean and the touch of the typist even; the Electromatic typewriter may be used to great advantage here.

Format instructions as to line length, interlinear spacing, etc., are most easily followed out if paper with a printed margin line can be provided, clearly designating the typing area. Photo-offset printers often provide such sheets. With paper so marked, and with the right ratchet and ribbon on the typewriter, and the right paper on the roller, any standard typewriter and any normally efficient typist should be able to turn out superior work.

There are two ways of preparing typescript for the two-column page. The copy can be typed directly on a large sheet marked off into two columns or upon a long one-column galley, to be cut and fastened on the two-column page backing with rubber cement. The former is perhaps the better method in handling copy which is absolutely “straight”—no footnotes, no illustrations, no probability of author’s revisions. But if the design of the page requires so many letters per column that the two columns cannot be typed together on the normal 10″-carriage, or if there is a prospect of many corrections or of footnote material to be fitted in place, the galley strip method of typing is probably superior.

All corrections may be made with shears and rubber cement. The new matter may be placed on top of the old, or else the deleted matter may be cut out and the new matter pasted in its place.

Paste pot and shears will also serve to make page proofs of galley proofs. The paper for the page proof should have a strong line running down the middle if the final copy is to show the two columns divided by a line. Diagrams, photographs, and other non-typescript matter can be fitted into the pages at this stage in the preparatory process.

V. The Manuscript Book

Our habits of judgment have been so formed that we always expect to see certain kinds of thought going around in their ∣📄 p.63 working clothes; others we expect to see parading in formal dress. Typescript and newsprint are the clothes that thought wears in its daily work; the typography of book manufacture is its more formal attire. But the formal dress that thought acquires in book typography is only a ready-made suit after all: calligraphy offers to clothe thought in much richer and more beautiful tailor-made attire, which should have a higher prestige value than book typography. ∣📄p.[62a] (blank) ∣📄 p.[62b]

The beauty of the manuscript has been increasingly appreciated as the centuries have confirmed the triumph of letterpress. Precisely the same qualities that make printing more attractive than typescript would tend to make manuscript more attractive than printing. There is more chance for close adaptation of letter form to the spirit of the text, more chance for creative artistry.

The twentieth century has already developed its scribes. The influence of Edward Johnston, author of the foremost general book on illuminating and lettering in English,⁵ has been felt not only in the beautification of typographical design but also in the growth, especially on the Continent, of a school of artists in whose work calligraphy meets creative literature, pictorial, art and bookbinding in a product which adapts medieval craft to modern design.

The handwritten book, as made by a penman like Rudolf Spemann, is, however, a rarity; its costs are high and its editions small, with the result that the average person seldom sees and never possesses a manuscript book.

But now a new prospect opens: the manuscript can come back in current publishing; the graphic arts are prepared to make it available to the common man. Photo-offset will multiply any black on white for∥ mat copy, whether prepared by relief printing, typewriting or calligraphy. The inexpensive photo-offset press may make possible a production of manuscript books as a special small-scale craft; the art of the scribe can now return, relieved from the rule of constant costs which placed it at such a disadvantage in competing with diminishine costs of the printing craft. The illustration on the preceding page serves to show what might be done in this line. (Figure XXXVI.)

Cost Analysis: Composition Costs

The area and running costs of a manuscript book reproduced by photo-offset do not differ from those of any other kind of book reproduced by this technique. The difference comes in composition costs, which naturally are high as compared with those of typescript and linotype work. There are no set prices for handwritten master copy, but an estimate of 1200 words per day at $3.00 or thirty pages a week for $20.00 was made for the Joint Committee.

These prices would naturally be prohibitive for the 50,000-word book chosen as a standard for the other pages in this chapter. They would be reasonable, however, for the reproduction of current literature—say, a small volume of poetry. The reproduction of current literature in this form might have an interesting effect upon culture, for it might help to implement cultural decentralization away from the great metropolitan centers. It would have little bearing, however, upon the problem of reproducing research materials except as it would help to deprive book typography of its privileged place as an arbiter of intellectual quality.

Notes