The Invention of Printing/Chapter 26

by Theodore De Vinne

XXVI

The Tools and Usages of the Early Printers.

Punches made by Goldsmiths … Styles of Types imitated from Manuscripts … Popularity of the Gothic … Moulded Matrices … Types made without any System … From an Adjustable Mould. Appearance of Early Types … Large Fonts made … Importance of Mould … Rudeness of Early Composition … Method of Dictation … Faults of Compositors … Slowness of Improvement … Construction of the Hand-Press, with illustration … Inking Balls, with illustration … Slowness of Pressmen … Printing in Colors … Printing Ink … Ingredients used by the Ripoli Press … Moxon's Complaints about Ink … Neglect of Engraving on Wood … Peculiarities of Paper … The Degradation of Engraving … Proof-reading at Weidenbach … Faults of First Editions … Superiority of Printed as compared with Manuscript Books … Permanence of Gutenberg's Method.

---

All invention is progressive. … When a new machine is produced, we do not say, Why, it only consists of a number of wheels and cylinders, therefore, surely there is nothing new in it! All the parts may be old, and yet the combination be quite new. To analyse an invention into its several parts, would be equivalent to finding that a poem was only composed of the letters of the alphabet, or the words in a dictionary.

Dircks.

---

The first processes in the practice of typography—the cutting of punches and making of moulds—demanded a degree of skill in the handling of tools and of experience in the working of metal rarely found in any man who undertook to learn the art of printing. They were never regarded as proper branches of the printer's trade, but were, from the beginning, set aside as kinds of work which could be properly done by the goldsmith only. Jenson, Cennini, Sweinheym and Veldener seem to have been the only printers of the fifteenth century who had the preliminary education that would warrant them in attempting to cut punches with their own hands.

Not every goldsmith^[1] could do this work with neatness, and for this reason, as well as for the sake of economy, many beginners bought their matrices from the printers who owned punches. In some cases the types were bought outright, but matrices which gave the means of renewing a worn-out font must have been preferred. That there was a trade in matrices before type-foundries for the trade were established is shown by the appearance of the same face of type in many offices. The Round Gothic types cut by Jenson were frequently used by printers in France and Germany. Certain faces of types used by Caxton and by Van der Goes, by Leeu and Bellaert, by Machlinia and Veldener, are identically the same, and must have been cast from matrices struck from the same punches.

The styles of the early types were not invented by printer or punch-cutter. The Pointed Gothic letters of Gutenberg's Bibles and of the Psalter of 1457 are like those of the choice ecclesiastical manuscripts of that period. The Round Gothic letters of the Catholicon and of the Letters of Indulgence are of the form then used by German copyists in popular books. In Italy, the first types were cut in imitation of the popular form of Roman letters, or in the southern fashion of Round Gothic; in the Netherlands, they present the peculiarities of Flemish writing; in France and Burgundy, they were, for the most part, in the favorite French style of Bâtarde ancienne. In no instance did the printer invent a new style: he did no more than direct his punch-cutter to imitate, as closely as he could, the letters of a meritorious manuscript. In this matter, as well as in the arrangement of types, he followed the fashion set by an approved copyist or calligrapher. The peculiar characters^[2] of different languages were produced as they were required, somewhat slowly and of unequal merit, by different printers. The limitations of typography were not fully perceived, and many unsuccessful attempts were made to produce types and sectional wood-cuts that could be used in the construction of maps, ornaments and pictures.^[3]

The Gothic character was more popular than the Roman, but there were mechanical reasons why many printers preferred it. It was not so quickly cut, but its broad face, free from hair-lines, was more readily founded. It could be inked with facility and printed with more evenness of color, and it would not show wear as soon as the Roman. Early printers, who had no Roman, were loud in their praises of the Gothic.^[4] It was preferred by Verard, Pigouchet, Kerver, add nearly all French and Flemish printers. It did not entirely go out of fashion in Southern Europe nor in France until the close of the sixteenth century. It might have been supplanted by Roman characters in Germany, if there had not been at this time a strong prejudice against Roman customs and fashions of all kinds. Attempts at change were frequently made, but they were always unsuccessful.

The steel bought for the type-foundry of the Ripoli Press was probably intended for punches. The use of this metal in other type-foundries may be inferred from the sharpness, when new, of many fonts of early types. That the moulds were of brass is indicated by the allusions of early writers and printers to types made in brass. The matrices were of copper, but it is not probable that they were struck in cold metal, for it required great force and still greater discretion to strike the punch truly, and the risk of breaking it had to be hazarded. For the matrices of the large types of Gutenberg's Bibles and the Psalter of 1457, copper softened by heat^[5] should have been, and probably was, provided.

When the secrets of type-making had been divulged, the printers who found difficulties in making or buying matrices tried to evade its necessary conditions and cheapen its processes. The types of wood with holes for wire, described by Specklin and others, must have been punches of wood which had been made in the belief that it would be cheaper to cast words than to cast and compose single letters. The matrices of lead noticed by Enschedé were probably made by striking the punch of wood in half-melted metal, after the process described by Didot. The punch of wood, burned by contact with hot metal, was repaired, altered and renewed; the matrix of lead,^[6] clogged by the adhesion of metal, became defaced, and was soon worn out. Every change in punch or matrix produced a corresponding change in the cast type.

The types of the fifteenth century were made without system. The dimensions of each body and the peculiarities of each face were determined chiefly by the manuscript copy which had been selected as the model. No printer had any idea of the advantages to be derived from a series of regularly graduated sizes, nor of the beauty of a series of uniform faces, nor of the great evils they would impose on themselves and their successors by the use of irregular bodies.^[7] A classification by scale of the types of any printer of this period will show that there are often wide gaps between the larger, and confusing proximities between the smaller, bodies.^[8]

As the size of every body is determined by the mould in which it is cast, it would seem that there must have been a separate mould for every distinct body.^[9] But this inference is encumbered with fatal objections. The type-mould of hard metal is, and always has been, a very expensive tool, and it cannot be supposed that any early printer made two or four moulds for one body when one mould would have served. It is much more probable that he tried to make one mould serve for two or more bodies. The inventor of the mould may have thought that it should be constructed with adjustments, so that it should cast different bodies as well as different widths of types. The practicability of a mould of this description is properly demonstrated by the old-fashioned adjustable mould for irregular bodies, or by the mould used for casting leads, which can be so enlarged or diminished that it will cast many bodies or thicknesses. If we suppose that this mould was used by Gutenberg for casting the two bodies of the Letters of Indulgence, and by the unknown printer of the Netherlands for his four bodies of English, and that it was, of necessity, newly set or adjusted each time a new font was cast, we shall at once have a precise explanation of irregularities which are unaccountable under any other hypothesis. Casting types without the system, standards and gauges which modern type-founders use, it is not surprising that the first printers made types with differences of body. It was the impracticability of casting in this primitive mould, at different times, types of uniform body, that compelled later type-founders to discard it, and to use instead a mould for each body.

The casting of the types, which was always done in the printing office, was then adjudged a proper part of a printer's trade. The earlier chroniclers said the first types were made of lead and tin. The Cost Book of the Ripoli Press specifies these metals, and obscurely mentions another which seems to have been one of the constituents of type-metal. If this conjecture can be accepted, types were probably made in the fifteenth century, as they are now, of lead, tin and antimony.^[10] Not one of the millions of types founded during the fifteenth century has been preserved, nor is there in any old book an engraving or a description of a type. This neglected information has been unwittingly furnished by a careless pressman in the office of Conrad Winters, who printed at Cologne in 1476. This pressman, or his mate, when inking a slackly justified form, permitted the inking ball to pull out a thin-bodied type, which dropped sideways on the face of the form. The accident was not noticed; the tympan closed upon the form, and the bed was drawn under the platen. Down came the screw and platen, jamming the unfortunate type in the form, and embossing it strongly in the fibres of the thick wet paper, in a manner which reveals to us the shape of Winters' types more truthfully than it could have been done even by

A Type of the Fifteenth Century.^[11]
[From Madden.]

special engraving. The height^[12] of this type is a trifle less than one American inch. The sloping shoulder, or the beard, as it was once called, was made to prevent the blackening of the paper, for it would have been blackened if the shoulder had been high and square.^[13] The circular mark, about one-tenth of an inch diameter, on the side of the type, was firmly depressed in the metal, but did not perforate it. As this type had no nick on the body, it is apparent that the circular mark was cast there to guide the compositor. When the type was put in the stick with the mark facing outward, the compositor knew, without looking at the face, that it was rightly placed. There is no groove at the foot Duverger says that the early types had no jet or breaking-piece; that the superfluous metal was cut off, and the type made of proper height by sawing.^[14] These details may seem trifling, but they are of importance: they show that, in the more important features, the types of the early printers closely resembled ours.

There is a disagreement among bibliographers about the quantity of types ordinarily cast for a font by the early printers. Some, judging from appearances which show that one page only was printed at an impression, say that they cast types for two or three pages only; others maintain that they must have had very large fonts. That the latter view is correct seems fully established after a survey of the books known to have been printed by Zell, Koburger, Leeu, and others. It would have been impossible to print these books in the short period in which we know they were done, if the printer had not been provided with abundance of types.^[15] As the types were made in the printing office, by a quick method, from an alloy which could be used repeatedly for the same purpose, the supply was rarely limited by fear of expense.

The trades of compositor and pressman, and possibly that of type-caster, were kept about as distinct then as they are now. There were more compositors than pressmen, and the compositors, says Madden, in the heroic age of printing, were not boys, but men of education and intelligence. The early printers who were taught the business that they might become masters had to pay a premium for their education.^[16] In the brief time that they gave to the work, their education must have been more theoretical than practical. As the branch of composition required the largest number of workmen, and more intelligence, and less manual labor than any other, it was usually selected by the pupil for practice. Of type-casting and presswork he learned no more than was sufficient to enable him to direct the labors of his future workmen. The knowledge of the trade which the pupil coveted was the ability to practise it on his own account, and this knowledge was, in most instances, satisfactorily acquired when he got a theoretical knowledge of its secret processes.

The frequent specification of the formen in the earliest notices of printing shows that the mould, with its accompanying matrices, was regarded as the key to the knowledge and practice of the art. As the moulds were made by master mechanics, not bound to secrecy, and as the earlier compositors had some knowledge of the process of type-casting, it was not difficult for a journeyman to become a master printer. When he had bought a type-mould and matrices, he could go to any city and begin to print books. He could cast types and mix ink as he needed them; he could buy paper and the constituents of type-metal in any large town; properly instructed, any joiner could make the press.^[17]

The annexed illustration, a fac-simile of one of Amman's engravings of a printing office, is from his book dated 1564. The case for the type is of one piece and is resting on a rude frame. All the boxes are represented as of the same size, but this is probably an error, for it is an error which is frequently made by designers of this day.^[18] In this, and in many other early illustrations of type-setting, the compositors are seated on stools. In Italy and in Paris, women were employed as
Presswork and Composition as done in 1564.
[From Jost Amman.]compositors. In the wood-cut used by Jodocus Badius^[19] for a trade- mark, we see a hard-featured dame before a narrow case, composing types with judicial deliberation. She has in her left hand a narrow composing stick, made to hold but two or three lines of small types. The early stick was not like the neatly finished iron tool of our time, with steel composing rule and an adjustable screw and knee adapting it to any measure. It was a real stick of wood, a home-made strip of deal, with the side and end-piece tacked on. For every measure, a new stick or a retacking of the movable piece was required. The date of the introduction of the stick cannot be fixed, but it was used, without alteration for many years, by the printers of all countries. It is possible that some of the early printers had no sticks. The peculiar workmanship of the unknown printer and of Albert Pfister shows that the types were taken direct from the case and wedged in the mortised blocks of wood which served for chases. Blades attributes the uneven spacing and irregular endings of lines in the early printed books of Caxton and of other printers, to their ignorance of the advantages of a composing rule, without which types could not be readily moved to and fro, and adjusted.^[20]

In the following illustration, the compositor has the copy before her in the shape of a book, but Conrad Zeltner, a learned printer of the seventeenth century, said that this was not the early usage; that it was customary to employ a reader to read aloud to the compositors, who set the types from dictation, not seeing the copy. He also says that the reader could dictate from as many different pages or copies to three or four compositors working together.^[21] When the compositors were educated, the method of dictation may have been practised with some success; when they were ignorant, it was sure to produce many errors. Zeltner said that he preferred the old method, but he admits that it had to be abandoned, on account of the increasing ignorance of the compositors.

No feature of early printing is more unworkmanlike than that of composition. Imitating the style of the manuscript copy, the compositor huddled together words and paragraphs in solid columns of dismal blackness, and sent his forms to press without title, running-titles, chapter-heads and paging-figures. The space for the ornamental borders and letters of the illuminator seems extravagant when contrasted with the pinched spaces between lines and words. The printer trusted to the bright colors of the illuminator to give relief to the blackness of the types, not knowing that a purer relief and greater perspicuity would have been secured by a wider spacing of the words and lines. The obscurity produced by huddled and over-black types was increased by the neglect of simple orthographical rules. Proper names were printed with or without capitals, apparently to suit the whim of the compositor. The comma, colon and period, the only points of punctuation in general use, were employed capriciously and illogically. Crooked and unevenly spaced lines and errors of arrangement or making-up were common. Madden has pointed out several gross blunders, caused by the transposition of lines and pages and an erroneous calculation of the space that should be occupied by print Words were mangled in division, and in the display of lines in capital letters, in a manner that seems inexcusable. But no usage of the early compositor is more annoying than his lawless use of abbreviations. Imitating the example of Procrustes, he made, the words fit, chopping them off on any letter or in any position, indifferent to the wants of the reader or to the proprieties of language.^[22] Whatever opinion may be entertained concerning the deterioration of printing in other branches, it is, beyond all cavil, certain that in the art of arranging types so that the meaning of the author shall be made lucid, the modern compositor is much the more intelligent mechanic.

Improvements were made slowly. The method of spacing out lines so as to produce a regular outline at the right side of every page had been practised before, but it was not in general use even as late as 1478. Arabic figures, instead of Roman numerals, were first used by Ter Hoorne of Cologne, and by Helye of Munster in 1470. Signatures to guide the binder in putting together in order the different sheets of a book were first used in printed books by Zarot of Milan in 1470. As the alphabetical letters of these signatures often had to be doubled, and sometimes quadrupled in thick books, it became necessary to print a full list of the signatures at the end of every book as an additional guide to the binder. This list, registrum chartarum, seems to have been first used by Colonna at Venice in 1475. The clumsiness of doubled alphabetical letters should have led to the use of Arabic figures for signatures, and should have suggested paging, but these reforms were not adopted for many years afterward.^[23] A table of errata, two pages folio, was exhibited by Gabriel Peter of Venice in 1478. The first full title, if a few lines in compact capital letters can be so called, was made by Ratdolt of Venice in 1477, but his example was not rapidly followed by rival printers. Running-titles and open chapter-headings are innovations of the next century. The printers of the fifteenth century who wished to free themselves from dependence on the illuminator filled up the white spaces about chapter-headings with bits of engraving on wood or metal.

Galleys, or trays of wood to keep in place the composed types, were not known; the types were placed line after line, perhaps letter by letter, in the mortised block of wood which served for the chase. Nice justification was impossible. If two pages were put in one mortise, one of these pages would often be out of square—an irregularity which has led some bibliographers to think that each page was separately printed from a separate form. The locking-up or tightening of the types, which was roughly done, often made the types crooked, springing them off their feet and making the spaces work up.^[24]

The neglect of the early printers to praise their presses is remarkable when contrasted with their frequent praises of the marvelous art of type-making. It is inferential evidence that the press was then regarded as an old contrivance, and not worthy of notice, but this conclusion cannot be unreservedly accepted. The principle of pressure was old, and for that reason, was undervalued by printers, but the mechanism of the press was new. That the printing press was an invention of merit will be perceived at a glance when it is compared with the screw press which is supposed to have served as the basis of construction.^[25] That a proper method of doing presswork was devised in the infancy of the art may be inferred, not only from the permanency of the primitive form of press, all the important features of which are still preserved in the modern hand-press, but from the meritorious presswork of the first books. The Bibles of Gutenberg were certainly printed on a press which quickly gave and quickly released its pressure, and which had the attachments of a movable bed, tympan and frisket, and contrivances for neatly inking the types and for keeping the paper in position.

Jodocus Badius of Paris was the first printer who published engravings of the printing press. It cannot be asserted that they are minutely accurate representations of the press then in use, but they will serve to show its general construction. Two features provoke hostile comment Contrary to modern usage, the piles of white paper and printed paper are unhandily placed on the off-side of the press, and the stalwart pressman pulls home the bar with both arms. The platen

Presswork and Composition as done in 1520.
[From Blades' fac-simile of the print of Badius.]

Two upright beams, or cheeks, supporting a thick cross-piece, or cap, made the frame-work. The cap held in place the screw and spindle which gave the impression, and the descent of the spindle was steadied by the large square collar, or till, which was supported by the cheeks. The point of the spindle pressed against the impressing surface, or platen, which was held in place by iron rods connecting it with the collar. The bed of the press and the form of types are concealed by the tympan drawer, which, with tympan and frisket, have been folded down and run under the platen. See illustration on page 307, and explanation on page 280, for the uses of these parts. The bed was of stone, but every other large piece was of wood. Iron was used only for the spindle, the core of the bar-handle, for nuts and bolts, and the minor pieces for which no other material would serve.

seems altogether too small when contrasted with the great screw, the heavy frame, and the two-handed pull of the pressman. The smallness of this platen was not an error of the designer. Moxon, who has minutely described the press of his time, says that the platen of an ordinary press should be of the size 9 by 14 inches, and that the coffin, or trough in which the bed was placed, should be 28 inches long and 22 inches wide. In other words, the platen was purposely made so that it could impress less than half the surface of the bed; it could print only one-half of one side of the sheet.^[26] Small as this platen may seem, it was large enough for the frame-work of wood. It gave great resistance under pull, and severely taxed the strength of the pressman. A platen of double size would have defied the pressman; it would have sprung under pressure and have broken the bed of stone.

The types were inked by balls, an appliance which is not more than fifty years out of fashion. These balls were made of untanned sheepskin, stuffed hard with wool, and mounted with handles. The gluey ink was evenly distributed by forcibly rocking their curved surfaces against each other. This done, the balls were then beaten upon the types in the form.

When we learn that the early presses were made almost entirely of wood, and put together by ordinary joiners, we may infer that many were unscientifically built,^[27] shackly. All the materials for presswork were imperfect The types, cut to length by a saw, were of uneven height; the paper was usually of very rough surface and of irregular thickness; the platen of wood, rarely ever truly flat, must have given unequal pressure at different corners. It was necessary that some substance should be put between the platen and the white sheet which would compensate for these irregularities.
Early inking Balls.
[From a Playing Card of Sixteenth Century.]This substance was a woolen blanket, in two or more thicknesses, which spread or diffused the impression. The wetting of the paper, which made it soft and pliable, materially aided the pressman, but his great reliance seems to have been on strong impression. All the old cuts of presses represent the pressman tugging at the bar with a force which seems out of all proportion to the size of the form.

The early press was rude, and the method of printing was unscientific, but in many offices the pressman. was superior to his press and his method. By doing his work slowly and carefully he often did it admirably. It was always done slowly, with a waste of time which, if allowed in the modern practice of printing, would make books of excessive price. Some notion of this waste may be had after an examination of the letters of the Psalter of 1457, in which exact work was produced by painting, not by printing proper. That the performance of the press even on ordinary black work was slow, is indicated by the great number of presses used by the early printers, and is proved by the plain statement of Philip de Lignamine, who said that the printers of Mentz printed three hundred sheets a day. This seems a small performance.^[28]

The accurate register of the first books was produced by placing the white sheet on four fixed points which perforated the four corners of the leaf when the first side was printed. In printing the back of the page, the half-printed sheet was hung on the same points, from the same point-holes, and was impressed in the same position. Blades notices the four point-holes in some of Caxton's books, and it is probable that the mysterious pin-holes in other books are the marks of points. It was soon discovered that register could be had with two points, which were placed in the centre of the sheet where the marks would be hidden by the binder.^[29]

The printing ink of the fifteenth century, as we now see it, is of unequal merit. In the books of Jenson it appears as an intense, velvety, glossy black; in the Bibles of Gutenberg it is a strong, permanent black, without gloss; in the Psalter of 1457 it appears in some places as a glossy black, and in others as a faded color which had to be retouched with the pen; in the works of the unknown printer it is a dingy and smearing black; in the book of some printers it is a paste color which can be rubbed off with a sponge; in nearly all, it is uneven, over-black on one page and gray on another.^[30]

The general impression that early printing ink is blacker and brighter than modern ink is not always correct. Early ink seems blacker, because it is shown in greater quantity, for the early types were larger, of broader face, without hair lines, and could be over-colored without disadvantage.^[31] The same ink applied to the small thin Roman types of our time, would seem dull and gray. The microscopic examination of any early ink will show that the black is not fine and not thoroughly mixed with proper drying oil. But this imperfection is comparatively unimportant. It is a graver fault in some early inks that they are not firmly fixed to the paper.^[32]

Ingredients of Printing Ink used by the Ripoli Press.
Ingredients.	Tuscan Currency.	American Currency.
Linseed Oil, bbl.	lir. 3 10 0	$3. 17
Turpentine, lb.	4 0	.18
Pitch, Greek	4 0	.18
Pitch, Black	1 8	7	½
Marcassite	3 0	.13	½
Vermilion	5 0	.22	¾
Rosin	3 0	.13	½
Varnish, hard	8 0	.36
Varnish, liquid	12 0	.54
Nutgalls	4 0	.18
Vitriol	4 0	.18
Shellac	3 0	.13	½

There is no trustworthy account of the invention of printing ink, but the types and the inks were undoubtedly invented together. One was the proper complement of the other. It may be supposed that Gutenberg acquired the knowledge of the newly found properties of boiled linseed oil^[33] from German painters. It is certain that he used oil as the basis of his ink, and that it was also used by his pupils and successors. And it has been in use ever since, for there is no substitute.

We have not been told how the ink was compounded. Our nearest approach to this knowledge is through the Cost Book of the Ripoli Press for 1481, which specifies and prices the materials. As no mention is made of smoke-black, we have to infer that pitch was burnt to make this black. Linseed oil, as the most bulky ingredient, very properly occupies the first place. The real value of nutgalls and vitriol is not so apparent: they were important ingredients in writing ink, and the Italian printer may have thought them indispensable in printing ink. Shellac and liquid varnish were used to give a glossy surface.

Printers soon discovered that printing was an art of too many details, and that the manufacture of printing ink was its most objectionable duty. There was risk of fire in the boiling of linseed oil; there was discomfort and dirt connected with the manipulation of the ingredients; and in inexpert hands there was waste and often entire failure. In all large cities, ink-making was set apart and practised as a distinct trade. As a necessary consequence, the quality deteriorated through the competition that followed. Moxon's criticism of ink made in England in 1683 could be applied without any injustice to much of the ink of the fifteenth century.^[34]

Gutenberg, Schœffer, Zell, Mentel and many early printers of France and Italy neglected engraving on wood.^[35] It may be that this neglect originated in the difficulties of printing

Reduced Fac-simile of a large Wood-cut, said to be of the Fifteenth Century.
[From Jackson.]

types and wood-cuts together,^[36] or in a despisal of the rude productions of the block-printers,^[37] and in the intention of the typographers to make emphatic the superiority of their branch. Wood-cuts were freely used by typographers in the heart of Germany and in the Netherlands, the districts where we find the earliest notices of block-printing, but they are generally of a low order. Many of them are barbarous, as faulty in cutting as in drawing, and pleasing only to uncultivated tastes. It is probable that, about this time, many of the more skillful engravers and designers^[38] abandoned the practice of xylography, attracted, no doubt, by the superior advantages offered by the newly invented art of copper-plate printing. The art of engraving on wood, although it afterward enlisted the services of artists like Durer and Holbein, could not compete with this formidable rival. It suffered a long eclipse, from which it did not emerge until the days of Bewick.

The quality of the paper in early books is as unequal as the printing. In the Bible of 36 lines, the paper is thick and strong, of coarse fibre, yellowish, apparently made from sun-bleached flax; in the books of Schœffer, and of the later German printers, the paper is thinner, but dingy and harsh; in the books of the Venetian printers, it is often very thin, usually of smooth surface and a creamy white tint that seems to have been unchanged by time. Different qualities are often noticeable in the same book. There were many paper-mills from which the printers drew their supplies, and every mill made different qualities. Blades says that it was the practice to sort the paper before printing, separating the rough from the smooth, and the thin from the thick, and to print and bind together sheets of similar quality. The sizes required by printers were small. The books first made were printed on sheets about 16 by 21 inches, one leaf of which was as large as could be printed by one pull of the press. The sizes 15 by 20, 14 by 18 and 12 by 15 inches were common, and in request for quartos and ocatvos. The largest size seems to have been royal, about 20 by 25 inches. The Cost Book of the Ripoli Press gives names and prices to nine distinct qualities or sizes of paper, but it does not define the weights and measurements. The smallest size and cheapest quality, possibly a pot foolscap, was put down at the price of 2 lire 8 soldi (about $2.18) per ream; the largest and best, probably royal, at 6 lire 8 soldi (about $5.80) per ream.^[39]

The Fall of Lucifer, as shown in Zainer's Edition of the Speculum Salutis.
An Illustration of the Degradation of Engraving on Wood.
[From Heineken.]

The paper made for the Bibles of Gutenberg and for the earlier books was the ordinary writing paper of the period. Made from linen rags that had not been weakened by caustic alaklies or by steam-boiling and gas-bleaching processes, and strongly sized by the dipping of each sheet in a tub containing a thin solution of glue, it was strong and of hard surface. But the qualities which commended the paper to the copyist were objectionable to the printer. The hard surface caused harsh impression, and strong sizing made the damp sheets stick together. It was soon discovered that unsized paper, which, according to Madden, was about half the price of the sized, was easier to print. It would take a clearer impression, and more thoroughly imbibe the oily ink. These advantages could not be overlooked, and, consequently, hard-sized papers went out of fashion. By far the largest part of the books printed during the last quarter of the fifteenth century were of unsized or half-sized paper.

The early printer tried to gratify luxurious tastes by printing copies on vellum, but its inordinate price, and the great difficulties then encountered in printing, obliged him to give it up as an impracticable material. When book-lovers found that able printers like Kerver and Pigouchet printed paper more neatly and evenly in color, vellum^[40] went out of fashion.

We do not know what system or method was observed in early proof-reading. Madden has pointed out many curious errors in three distinct copies of a book printed at Weidenbach about 1464, which seem to show that the compositor of each copy read the proof of his own work, and read it badly. Possibly this was the method of many of, the amateur printers of that century, whose books, according to Schelhorn, bristle with horrid and squalid errors. It could not have been the method of Gutenberg, whose Bibles, although not free from faults, were obviously read with care. Nor was it the method of careful printers, for there is evidence that many of them enlisted the services of eminent scholars as proof-readers or correctors of the press.^[41] These correctors did a double duty; they corrected the errors of the compositors and those of the

A Print of 1475, probably the work of an amateur engraver.
[From Heineken.]

manuscript copy. ^[42] From the frequency and earnestness of the complaints then made concerning faulty manuscript texts, it seems that the copyists needed correction more than the compositors. But the correctors were not always equal to the task. Some of them were grossly incompetent, and still further corrupted the texts they undertook to improve.^[43] Considering the difficulties the early printers encountered in getting correct copies and competent readers, it is surprising that their books are not more full of faults. The errors of early printed books have been frequently commented on, but the remarks of Prosper Marchand are, perhaps, the most emphatic:

It is a prejudice altogether too common, a prejudice which dealers in old books have kept alive and profited from, to think that the editions of the fifteenth century are more accurate because they were printed from manuscript copies. Many of these editions were printed from faulty texts, picked up by chance, or selected without judgment by printers who were unable to see their faults, and were still further corrupted by the ignorance and rashness of their editors and correctors. I know that this is a kind of literary blasphemy, but it is warranted by respectable authority. … They are deceived who think that books are accurate in proportion to their age. For the most part, the older they are, the more inaccurate they are.^[44]

Inaccurate as early printed books may have been, they were more correct than those of the copyists. The errors of a faulty first edition were soon discovered and the faulty editions were supplanted by the perfect. It is not the least of the many benefits of printing that it has effectually prevented the accidental or intentional debasement of texts.

The inferiority of the tools of the early printing office could be plainly exhibited by contrasting them with, those of our time—the early hand-press with the modern cylinder printing machine—the entire collection of types made in the fifteenth century with the specimen book of any reputable modern type-founder. But the pride of the young printer in improvements which have been most largely made by the men of this century should be modified by the reflection that there has been no change in the theory, and but few changes in the elementary processes of printing. The punch, matrix and mould, the tympan, frisket and points, the use of damp paper and oily ink, of curved surfaces for applying the ink, and of blankets for diffusing the impression, are still in fashion. Printing is done quicker, cheaper, with more neatness and accuracy, with more regard for the convenience of the reader, with many new features of artistic merit, and in varieties and quantities so vast that there can be no comparison between early and modern productions—but it is the same kind of work it was in the beginning. It has not been made obsolete by lithography or photography, nor by any other invention of our time. The method invented by Gutenberg still keeps its place at the head of the graphic arts.

---

---

1. Gutenberg's employment of the goldsmith Dünne at Strasburg, and the payment to him of a big sum for work connected with printing, can be most satisfactorily explained by the conjecture that Dünne was hired to cut punches and make a mould. I find no mention of punch-cutting or mould-making at Mentz, but there is, in the accounts of the Ripoli Press, an unequivocal notice of one John Peter of Mentz, who was selling matrices to the printers of Florence in 1476. It is evident that this John Peter had experience in this branch of typography. The Ripoli Press bought of him, in 1477, the matrices of a full font of Roman, for 10 florins in gold. John Peter was not the only punch-cutter. In 1478, the Ripoli Press paid the goldsmith Benvenuto 110 livres for the punches of three fonts—two of which were of Roman and one of Gothic face. In 1481, another goldsmith, Banco, made a sale to the manager of the Ripoli Press, of "100 little letters, 3 big letters, and 3 vignettes on copper."
2. Square notes of music, partly written, partly printed, are seen in the Psalter of 1457. Greek letters were made by Schœffer and Sweinheym, but the first book in Greek was printed by Paravisinus at Milan in 1476. Hebrew types were made at Soncino in 1488. At the close of the century, a German printer at Paris made an imitation of writing, but the letters were not connected, and the only penmanlike features were in the capitals. About 1500, Manutius had the engraver Francis of Bologna cut punches for Italic types, in imitation of the handwriting of Petrarch.
3. Jacob Bellaert of Haarlem combined isolated engravings, cut for the purpose, in the belief that each combination would seem a new engraving. Kerver tried to give variety to his pages by varying combinations of detached pictorial borders. But it was guickly demonstrated that typography could deal successfully with letters only. The large ornamental initial letters of books were not cast, but cut, sometimes on wood, oftener on metal. Small and ornamented capital letters were cast by Mentel of Strasburg, and by Ratdolt of Venice in 1477.
4. Colonna and Manthen at Venice said that their Gothic was a "sublime letter." John Herbort, in 1483, said his was "a most captivating letter, unquestionably excelling all others." Nicholas Prevost said his book was printed "in types the most beautiful and most becoming for polite literature." Chevalon said his Gothic was "the polite and fashionable letter."
5. In France, the punches are struck in hot copper to prevent their breakage.
6. I know by experience that the ordinary metal used for types can be cast in a matrix of lead to the number of 125 or 150 types before the matrix will be destroyed. After 50 or 60 castings, there will be an alteration in the mould; the finer lines will disappear and ruder lines be presented. This will account for the differences that the same letters present on every page. Magazin Encyclop. de Millin, 1806, vol. i, p. 74, as quoted by Bernard, vol. i, p. 299.
7. Gutenberg's larger bodies were irregularly graduated and of Pointed Gothic face; his smaller bodies were not separated at proper distances, and were of Round Gothic face. The unknown printer had four faces and four bodies of the size English. Caxton had two faces and two bodies each of the sizes Paragon, Great-primer and English. The types of many printers at Paris and Venice show irregularities of body which seem remarkable and inexplicable to the modern printer.
8. The smallest sizes which I have met in any book of the fifteenth century are in the Decretals of Gregory, printed in black and red by Andrew Torresani at Venice in 1498, in which book the text is in Bourgeois and the surrounding notes are in Brevier. Nonpareil was first made by Garamond of Paris about the middle of the sixteenth century. Diamond was made by Jannon of Sedan about 1625. Nothing smaller was attempted until 1827, when Henry Didot, then 66 years old, cut a font on the French body of 2½ points—a body known to American printers as Brilliant, or Half-nonpareil—about twenty-five lines to the American inch.
9. It has been suggested that these distinct bodies were founded in sand moulds; that a new pattern for the body was made every time a new font was cast; and that the irregularities in body are the results of unintended or undetected variations in the pattern. But this hypothesis cannot be accepted. The small bodies, the sharp edges, close fitting-up and even lining of the types, are peculiarities which could not have been produced by a sand mould, nor by a mould of any plastic material.
10. See page 66 of this book. Was this obscure metal antimony? The text books say that antimony was, for the first time, set apart as a distinct metal in 1490, by Basil Valentine, a monk of Erfurt. But Madden says that a book supposed to have been printed at Cologne, before the year 1473, plainly describes antimony as a metal frequently used and much abused by many monks of the thirteenth century in their pharmaceutical preparations. Lettres d'un bibliographe, 4th series, p. 115.
11. Lettres d'un bibliographe, 4th series, p. 231.
12. It agrees exactly with the old French standard (of 1723) for height of type, which was 10½, geometric lines, or, by modern French measure, 24 millimetres. Fournier, Manuel typographique, vol. i, p. 125.
13. The sloping shoulder, which was in general use in the first quarter of this century, was discarded to meet the requirements of the new art of stereotyping. It was found that these sloping shoulders made projections in the plaster mould, which imperiled the making of an accurate cast. The blackening of the sheet from square shoulders was prevented by altering the mould and placing the shoulder lower on the body.
14. See page 399 of this book.
15. Bernard believes that Gutenberg cast for the Bible of 42 lines at least 120,000 types, or enough for two sections, or forty pages. He supposes that twenty pages were perfected, and ready for press or under press, while the succeeding twenty pages were in the compositor's hands. This would be the method adopted by the modern printer, and it may have been the method of Gutenberg, but it is probable that the difficulties connected with the new art compelled him to print the book more slowly, and with imperfect system. But the printers who followed him certainly used quick methods.
16. Caxton said that he had "practysed & learned at [his] grete charge and dispense to ordeyne this said booke in prynte."
17. Many of the early master printers practised their trade for a few years in one place, and a few years in another, roving about from town to town with a seeming indifference to change which seems unaccountable to the modern printer, who knows how expensive it is to move a printing office. The roving habits of the masters will not seem so strange when it is known that the equipment of the early office was simple, and that the more expensive tools could be carried with little difficulty.
18. The engravings of cases shown by Moxon have boxes of unequal size. No doubt, they were so made from the beginning, for a day's experience would teach any compositor that his case must have a larger box for the letter e than for the letter x.
19. See page 528.
20. Bernard says that sticks of wood were used by Christopher Plantin, "king of printers." It is characteristic of the taste of his time, that Plantin had sticks of wood, although he boasted that some of his types were cast in [matrices of] silver.
21. Madden, in his first collection of Lettres d'un bibliographe,—the most curious piece of analytical criticism that has appeared in typographical literature—has demonstrated that the method of dictation was practised in the office at Weidenbach. In this series of letters he critically examines three books, printed at this office with the same types, and at the same time, and points out the peculiar errors of three different compositors, who, not seeing the copy, were misled by their misapprehension of the dictated words. He claims that these books were the practice work of three amateur compositors who were then learning the trade. Each compositor had copies of his own workmanship who, not seeing the copy, were misled by their misapprehension of the dictated words. He claims that these books were the practice work of three amateur compositors who were then learning the trade. Each compositor had copies of his own workmanship as a memento of his errors. Novel as they may seem, I am inclined to accept the conclusions of Madden. Many copies of early printed books, known to be of the same edition, or done at the same time, show variations in the typographical arrangement which cannot be explained by any other hypothesis than that of a double composition by compositors working from dictation.
22. The composition of Schœffer's edition of the Decretals has been injudiciously praised by Bernard. In the fac-simile on page 463, it will be noticed that the page is crooked, and that the justification and making-up are very faulty. In a copy of Torresani's edition of the Decretals, the frequent contractions make the work almost unreadable. This book has been highly commended for its even spacing; but it is a sufficient answer to say that any printer could space admirably, even in the narrowest measure, if allowed to mangle words to suit his convenience.
23. The statement made by Lacroix that one book was paged in 1469 does not prove that this was the usage. In some books printed at Venice during the last ten years of the fifteenth century, the leaves (not the pages) are numbered on every odd page. But this was not the common practice. In the Statius of Aldus, printed at Venice in 1502, and in the Italian translation of the Commentaries of Julius Cæsar, printed by Bernard Venetus of that city in 1517, neither leaves nor pages are numbered.
24. Some early chases held their types not with quoins, but by the pressure of screws. A German printer's hand-book, dated Leipsic, 1743, has diagrams of imposition in which the pages are fastened by screws perforating the chase. Quoins and bevels were not an early invention.
25. See page 395 for illustration of primitive screw press.
26. Mechanick Exercises, vol. i, pp. 52, 69. To the printer who has seen only the press in which the platen covers the bed this may seem an absurd method, but it was a method in general use even as late as the beginning of this century. Men are yet living who have printed books by the method shown in the cut—pulling down the bar when one-half of the form was under the platen—releasing the pressure—running the other half of the bed under the platen—and finishing the presswork of the other half of the sheet by a second pull.
27. There should have been a gradual improvement in the construction of the press, as there was in the making of the types, but there was no decided change for two centuries. Moxon, in 1683, commending the "new fashion" presses of Blaew, denounced the "old fashion presses as make-shift, slovenly contrivances practised in the minority of this art." Nor was Blaew's press perfect. To insure proper register, Jackson (who undertook, at Venice in 1745, to print wood-cuts in colors) was obliged to reconstruct the press of Blaew.
28. It must also be remembered that on the early printing press two pressmen were required for the work—one to beat or to ink, and one to pull or to print. The ordinary task of the hand-pressman of New-York in 1840 was rated at 1500 impressions, but these impressions were made by one man (working an inking machine) and one pull on forms of large size. Considering the surface printed, the performance of one hand-pressman in 1840 was about eight times more than that of one pressman in 1458.
29. Words and lines were sometimes printed in red in a text of black, with a nicety of register rarely equaled by any printer during the first years of this century. The early method of printing red with black, has been described by Moxon. The black form was first printed with quadrats in the places that should be occupied by the red words or lines. This done, the form remaining on press, the quadrats were taken out and the vacant space partially filled with "underlays" of reglet, about one-sixth inch thick. On these underlays the types to be printed in red were placed, which adjusting made them about one-sixth of an inch higher than the types of the black form. The bearers were then raised, the impression was readjusted, a new frisket was put on, and the pressman was ready to print red as he had printed the black. This method of printing red with black, a clumsy method at best, which can be practised only on small forms on the hand-press, has been out of fashion for many years.—The color work of the early printers has been overpraised. Superior, no doubt, to that of printers of the last century, who tried to do more work in less time, it cannot be compared with the color work of our time. The rubricated Book of Common Prayer printed by Welch, Bigelow & Co. of Cambridge, Massachusetts, the Specimen Book of Charles Derriey of Paris, the French-English Dictionary of George Bellows of Gloucester, England, maybe offered as specimens of modern color presswork which show an exactness of register and a purity of color and of impression not to be found in any early book.
30. This unevenness does not prove the use of two distinct inks. In some instances, it was caused by the negligence of the pressman who applied an unequal quantity of ink upon different pages. In many instances, it was produced by the variable qualities or conditions of the paper or vellum. If the paper laid out for one form differed from that used for other forms in being too coarse or too dry, or over-wet, or if the vellum had been polished too much or too little, or had not been entirely freed from lime and grease, it would take up from the types, during each condition, a variable quantity of color, and produce prints of a different degree of blackness. These variations in color are most noticeable in books of vellum. In a prayer book printed by Kerver in 1507, the ink is black wherever the vellum is smooth, and gray where it is rough. In another edition of the same book on paper, printed by Kerver in 1522, the ink is not so black as it appears on the smooth vellum, but the color is more uniform. Equal carefulness seems to have been taken with each book, and the ink was, no doubt, substantially the same. Some of the early printers sorted their sheets after printing, separating the under-colored from the over-colored and binding each together.
31. In trying to avoid the gloominess of early printing, modern printers have gone too far in the opposite direction. The fault of imperfect blackness which is justly censurable in many modern books is largely due to what Hansard calls the "razor-edged" hair lines and thin stems of modern types which give the printer no opportunity to show black color. Readers have been taught to prefer a feminine elegance in types, a weak and useless imitation of copper-plate effects, to the masculine boldness, solidity and readableness of the old-style letter of the last century.
32. Mr. Ticheborne, a recent contributor to Chambers' Journal, says that the older printing inks are more easily saponified and washed off by alkalies than those of the last century. Some of the old inks he found so sensitive, that on introducing them to a weak solution of ammonia, the printed characters instantly floated off the surface of the pages. His explanation, that the oil had not been properly prepared by boiling, and was not changed into an insoluble varnish, and "resinfied," is, no doubt, correct. A practical ink-maker, in a series of papers to L'imprimerie (vol. I, p. 129), says that in many books of the fifteenth century, the adhesion of the color to the paper is very weak, and that the ink can be made pale or washed off with a moist sponge.
33. Lanzi refers to an Italian manuscript of 1437 in which it is asserted that the new method of painting in oil, as practised by the Germans, must begin with the process of boiling linseed oil. History of Painting in Italy. Bohn's edition, 1852, vol. I, p. 86.
34. Our Inck-makers to save charges, mingle many times Trane-Oyl among theirs and a great deal of Rosin; which Trane-Oyl by its grossness Furs and Choaks up a Form, and by its fatness hinders the Inck from drying; so that when the Work comes to the Binders, it Sets-off; and besides is dull, smeary and unpleasant to the eye. And the Rosin, if too great a quantity be put in, and the Form be not very Lean-Beaten, makes the Inck turn yellow: And the same does the New Linseed-Oyl.—Secondly. They seldom Boyl or Burn it to that consistence the Hollanders do, because they not only save labour and Fewel, but have a greater weight of Inck out of the same quantity of Oyl when less Burnt away than when more Burnt away; which want of Burning makes the Inck also, though made of good old Linseed-Oyl, Fat and Smeary, and hinders its Drying; so that when it comes to the Binders it also Sets-off.—Thirdly. They do not use that way of clearing their Inck the Hollanders do, or indeed any other way than meer Burning it, whereby the Inck remains more Oyly and Greasie than if it were well clarified.—Fourthly. They, to save the Press-man the labour of Rubbing the Blacking into Varnish on the Inck-Block, Boyl the Blacking in the Varnish, or at least put the Blacking in whilst the Varnish is yet Boyling-hot, which so Burns and Rubifies the Blacking, that it loses much of its brisk and vivid black complection.—Fifthly. Because. Blacking is dear, and adds little to the weight of the Inck, they stint themselves to a quantity which they exceed not; so that sometimes the Inck proves so unsufferable Pale, that the Press-man is forced to Rub in more Blacking upon the Block; yet this he is often so loth to do, that he will rather hazard the Content the Colour shall give, than take the pains to amend it: satisfying himself that he can lay the blame upon the Inck-maker. Moxon, Mechanick Exercises, vol. ii, pp. 76, 77.
35. No exception need be made for the initial letters of the Psalter of 1457. The thin curved lines of the ornamental portions of these letters could not have been cut on the flat boards then used by all engravers on wood. The absence of cracks and broken lines, after long service, in every print taken from these cuts is presumptive evidence that they were cut on metal. The ornamentation is unlike that of the professional engravers of block-books and at once suggests the thought that they were cut on brass or type-metal by the hand that cut the types of the text.
36. That the early printers did encounter serious difficulties in the use of wood-cuts in type forms is proved by their selection of blocks of smaller size. Full-page cuts are rare in the books of Koburger, Leeu and Veldener. Von Os of Zwoll cut up the blocks of the Bible of the Poor. Blades says that Colard Mansion printed the types and wood-cuts that appeared on the same page by two impressions. Sad experience in the warping and cracking of blocks of wood in forms of types was, no doubt, the reason for this extra labor. This difficulty seems to have been avoided by Pigouchet, Kerver and the printers of ornamental books, whose cuts have all the mannerisms of engraving on metal.
37. The disconnection between the arts of engraving on wood and typography is fairly indicated by the quarrel between the type-printers and block-printers of Augsburg.
38. Some engravers on wood who would not work with typographers undertook a new branch of printing—the making of prints, thirty or forty inches long, for the decoration of interior walls. Becker has published a collection of these large prints, taken from the original blocks, some of which he says were made before 1500. See cut on page 535.
39. If Florentine money had eight times the purchasing power of its American equivalent, these were high prices. They justify the observation of Keyser and Stol, printers at Paris in 1486, that the price of paper was out of all proportion to the price of printed books.
40. Vellum was made out of the dressed skins of very young kids and lambs; parchment from the skins of sheep and goats. The vellum was very thin, flexible and highly polished; the parchment was thick and horn-like; but each substance was prepared by nearly the same process. The skin, when freed from hair, was put in a lime-pit, until it was deprived of its fat. It was then stretched on a frame, pared with a knife, rubbed with lime and pumice-stone, and repeatedly dried and wet, and rubbed and stretched, until the surface was made faultlessly smooth.
41. See page 469 for the testimony of Schoeffer's proof-reader.
42. The copyists, underpaid by the stationers, did their work recklessly, abbreviating words so freely that it was often impossible to discover the meaning of the author. The faults of the calligrapher, who preferred beauty to accuracy, and of the young scholar, who rashly undertook to correct errors—tended to the same result. Fichet, a professor of the University of Paris, who seems to have been the first man of letters who esteemed printing, said, in a complimentary letter to Gering, Crantz and Friburger, that books were becoming barbarous through the faults of the copyists. Bouhier, a later president of the University, said that the books of the copyists were monstrous, and often unintelligible.
43. Marchand quotes at length an author who says that John Andrew, the corrector for Sweinheym and Pannartz, was a very presumptuous meddler with texts. When he met a word he did not understand, he printed it in Latin, or put in words at a venture, often making the text more unintelligible than ever. Another ecclesiastical reader, Bishop Nicholas Perotti, was quite as great an offender.
44. Marchand, Histoire de l'imprimerie, vol. i, pp. 97-103, and notes. In support of this assertion he cites the opinions of Schelhorn, Maittaire, Naudé, and other eminent bibliographers, and gives many specifications of the inaccuracies of the early printers from Fust and Schœffer to Froben. Not even Aldus Manutius escapes, for Marchand quotes at length the accusation of Erasmus that the Homer, Cicero, and Plutarch of Aldus were depravatissima. This criticism is hardly warranted by the errors of these editions, and is decidedly unjust in its reflection on a printer whose industry and carefulness as an editor have never been surpassed, and who, in his edition of Plato of 1513, offered a gold coin for every mistake that should be discovered. This damaging accusation would probably never have been made if Erasmus had not quarreled with Aldus, and had not thought it necessary to deny with much asperity that he had served as a corrector of the press in the Aldine office. As a corrector, Erasmus was not beyond reproach, as will be more clearly seen in his reading of the Greek Testament. Froben's lamentation over the two pages of errata in this book (published by him, but corrected by Erasmus) shows how much easier it is to discover errors after commission than it is to correct them in time. Stung by the taunts of critics, Erasmus said that if the Devil did not preside over typography, there must have been a diabolical malice on the part of the compositors.