The Invention of Printing/Chapter 15

by Theodore De Vinne

XV

The Works and Workmanship of an Unknown Printer.

The Speculum not the Work of an Experimenter … Improbable that this was his only Typographic Book … Twelve Books, Eight Faces of Types and Forty-two Editions attributed to him or his Successors … Hessel's Classification of these Types … Fac-simile of the Types of the Speculum. Fac-simile of the Fables of Lorenzo Valla … Fac-simile of the Peculiarities of Criminal Law. Fac-simile of the Epitaphs of Pope Pius II … The Donatus … Fac-simile of the Abecedarium. The Eight Faces of Types were made by the same Printer … An Indication that he Wore out Types rapidly … That he Sold many Books … Trivial Character of the Books … His Types not Made of Wood … Illustrations of Types of Wood … Their Impracticability Demonstrated … Books not made from Cut Types … Cause of the Dissimilar Appearance of the Types … Were Founded. The Press of the Unknown Printer … Its Defects … Indications of the Use of a Frisket.

---

If any shall suggest, that some of the Enquiries here insisted upon (as particularly those about the Letters of the Alphabet) do seem too minute and trivial for any prudent man to bestow his serious thoughts and time about, such persons may know that the discovery of the true nature and cause of any the most minute thing both promote real knowledge, and therefore cannot be unfit for and Man's endeavours.

Bishop Wilkins, 1668.

---

If the printer of the Speculum was the rightful inventor of typography, his workmanship, as shown in the different editions of the book, clearly proves that he had passed the shoals of experiment, and was on the broad sea of successful practice. We can see, even without the help of the legends or chronicles, that he cut punches, made moulds and founded types of different faces and bodies; that he compounded ink in a proper manner, and printed his types upon a press constructed for the needs of his work; that he was successful both as a publisher and a printer. He practised printing not for amusement, nor in the way of scientific experiment, but as a business. Rude as his workmanship may appear, it fairly included all departments of the art: it was not experimental, but practical typography.

With these facts before us, it would seem proper to pass at once to the examination of the statements that have been made about the supposed printer of the book. But an examination at this point would be premature, for we have not, as yet, all the facts that are required. The four editions of the Speculum do not furnish enough evidence. It is not reasonable to suppose that two or three distinct fonts of type were made for no other purpose than the printing of four editions of this book. It is probable that the printer printed other books. But the early chronicles of Dutch printing tell us very little about these books. They are not only meagre in their recital of the more important facts connected with the invention, but are notoriously incorrect in their description of the minor details. They are unsafe guides. The books themselves, which reveal, to some extent, the process by which they were printed, are now regarded as of higher authority. We can accept the chronicles only so far as they corroborate the internal evidences of the books. It is proper that the books should be examined first.

The number of these books is greater than has been supposed, even by those who have favored the Dutch version of the invention of typography. Forty-three editions of twelve different works, printed from eight faces of types, are now attributed to the unknown printer of the Speculum or to his successors. In eleven works, the types resemble those of the Speculum, but the books are different as to character. They are in the form of small quarto or octavo, and are entirely destitute of illustrations. They are without name or place of printer, and, with one exception, are without date; they have no literary and no historical value; they differ but little, in a mechanical point of view, from numerous undated works of similar nature that have been assigned by bibliographers to the latter part of the fifteenth century. The places where these books or their fragments were found, and some of their peculiarities of workmanship, furnish evidences of value in an inquiry concerning their printer.

These books have been carefully classified, according to their types, by J. H. Hessels, the translator in English of Van der Linde's Haarlem Legend, from which work the classification following has been copied. The types have been specified by numbers, and have been arranged according to the order in which they are described by Holtrop in his Monuments typographiques. It is not pretended that the order of these numbers indicates the order in which the types were made; numbers have been assigned to them only for convenience in reference and for the purpose of accurate classification.

Type I. In this character^[1] the four notable editions of the Speculum were printed. In the same character were found the relics of six editions of the Donatus. The single leaf by which one edition of this book was identified, was pasted in a volume which once belonged to Sion Convent, at Cologne, and which contained several treatises printed by Ulric Zell, of Cologne. One of these treatises is dated 1467. Another leaf, now in the city hall of the city of Haarlem, was found in the original binding of an account book for the year 1474, which book was kept in the cathedral of that city. The account books of this church for the years 1476, 1485 and 1514, contain cuttings of leaves from the same edition. The first entry in the record of 1474 is to this effect: "Item. I have paid six Rhine florins to Cornelis the binder, for the binding of books."^[2] Fragments of other little books printed in the types of the Speculum have been found:

An abridgment of the Liturgy, then known as the Little Book of the Mass,^[3] a small quarto, with pages of twelve lines.

A Dutch version of the Seven Penitential Psalms, in the form of a very small quarto, containing but eleven lines to the page, printed on vellum, on one side only of the leaf. The only known copy of this work was found in Brussels.

Fragments on vellum of three editions of the Doctrinal of Alexander Gallus, a Latin grammar in rhyme, noticed by Van der Linde as the shabby compilation, by a priest of Brittany who lived in the thirteenth century, of the old Latin grammar of Priscianus. One of these fragments was found within the lining of a book printed at Deventer in 1495.

Four leaves of the Couplets of Cato, a small quarto which was then very popular in the schools.

Type II. The Dutch edition of the Speculum, which is described in this book as the third, contains, on pages 49 and 60, types which resemble those of other editions, and which

Type II. Fac-simile of the Small Types in the Third Edition of the Speculum.
[From Holtrop.]

seem to be the workmanship of the same letter-cutter. As these types are of a smaller face and body, they must have been founded in another mould. No fragments of any book in this smaller type have been found. Type III. The types of this face are newer, but they resemble those of Type II; some capitals are identical, but others have differences which establish it as a distinct face. As it is of a larger body, it must have been founded in a different mould. A book which contains the Fables of Lorenzo Valla and the Witty Speeches of Great Men, two little works of some popularity in the fifteenth century, is the only known specimen of this type. The paper of this book, which is like that of the Speculum, contains many of the strange blemishes, previously described, of useless letters embossed in the white

Type III. Fac-simile of the Types of the Fables of Lorenzo Valla.
[From Koning.]

lines and near the margins. As the written preface of the author is dated May, 1438, it is apparent that the book must have been printed subsequently to this date.

Type IV. Of this face, the fragments of four copies, and presumably of four distinct editions, of the Donatus have been found. This type, which does not closely resemble the faces previously described, was founded on a body a little larger than Paragon. The largest book in this type is a treatise on the Roman Law, apparently an abridgment of the fifth book of the Pandects of Justinian. It is described in the preface as The Peculiarities of Criminal Law, by Lewis of Rome. This treatise, which consists of forty-four pages, is printed in the form of small folio, twenty-six lines to the page. It was the largest book and contains the largest type of the unknown printer.

Type V. The forty-fifth page and all subsequent pages of the book previously described are devoted to a Treatise and Epitaphs by Pope Pius ii, and a Eulogy on Lorenzo Valla. In these names we find sure indications of the probable age of the book: Cardinal Piccolomini

Type IV. Fac-simile of the Types of the Peculiarities of Criminal Law.
[From Koning.]

or Æneas Sylvius was made Pope Pius ii in the year 1458; Lorenzo Valla died in 1457. The book must have been written and printed after these dates. The workmanship of this part of the book is of superior character: the types were fairly founded on a body about the size of Great-primer; they were decently printed in good black ink and on both sides of the paper, but the remarkable defect of embossed letters which has been noticed as one of the blemishes of the Speculum is also noticeable in this book.

This Type V seems to have been more frequently used than any other type in the list, but it was always on petty books or pamphlets. One book printed in it has only twenty-four pages, but it is made up of four distinct tracts: William of Saliceto on the Health of the Body; Torquemada on the Health of the Soul; A Treatise on Love, etc., by Pope Pius ii; The Iliad of Homer, or more definitely, a commendation of the Iliad. Two editions of this book have been discovered. A fragment of one edition was found in the binding of a work printed by Jan Andrieszoon, of Haarlem, in the year 1486. Another book in the same type, which consists of ten leaves, contains an abridgment or an epitome of the Iliad, with a preface by Pius ii in praise of Homer. Of this book two editions were printed. Six editions of the Donatus, four editions of the Doctrinal of Alexander Gallus, and one edition of the Couplets of Cato were also printed in this type.

Type V. Fac-simile of the Types of the Epitaphs of Pope Pius ii.
[From Koning.]

Type VI. An edition of the Donatus, twenty-seven lines to the page, is the only known book in this type, which was founded on Great-primer body.

Type VII. Four leaves of a Donatus on vellum, taken from the binding of a book printed in Strasburg in the year 1493, and belonging to a convent in North Brabant, are all that is known of this type, which closely resembles the character described as Type V.

Type VIII.^[4] Impressions from this face of type have been found in the fragments of only two books. Two broad bands of parchment printed upon one side only with the text of a Donatus, which were discovered in the cover linings of a manual of devotion, printed at Delft in 1484, are the only known relics of one of these books. The types are barbarous, of singularly ungraceful cut, of uneven height and out of line, evidently founded by a man who had no skill in type-founding. They are printed in pale ink which is readily removed by the application of water. The presswork is as slovenly as the type-founding, but the composition was done with some care and intelligence. The lines of type are nearly even as to length, and the words, when broken, are properly divided in syllables. It is evident that the compositor knew how to space and divide words, but the font of type that he used was not provided with hyphens or marks of punctuation. The fashion of the letter is in the Dutch style as may be seen in the final t with the perpendicular bar.

The other fragment in this type is a little pamphlet of eight pages, printed on parchment and upon one side only. It is described by some as a Horarium, or a little book of prayers; by others as an Abecedarium, or a child's primer. It contains the Alphabet (all the small letters but not the capitals), the Lord's Prayer, the Ave Maria, the Apostles' Creed, and two prayers. The Alphabet has the k, a letter that was not used in the Latin language; it has no w, this letter being formed by the union of the two characters v. Holtrop says that the types seem to have been made for the Dutch language.

The "turning upside down" of four letters on the second page of this little work proves that the letters are impressions from movable types.

Line 2.	Paue should be Pane.	Line 5.	uobis should be nobis.
Line 3.	Cotidiaun should be Cotidianu.	Line 6.	uostra should be nostra.

This little tract was discovered in 1751 by the celebrated type-founder Enschedé, of Haarlem, in a manuscript breviary of the fifteenth century, among the books of the descendants of John Van Zuren, a printer of Haarlem in 1561.

If barbarous type-founding and shabby printing could be accepted as conclusive evidence of the superior antiquity of the book in which these faults occur, the Abecedarium should be the oldest piece of printed matter. One cannot imagine a printed book with more slovenly workmanship.
First Page.
The Enschedé Abecedarium.
[From Holtrop.]Its types present all the irregularities of the Donatus previously described. The pages have but nine lines of types to each page, yet they are very crooked. This crookedness was partially produced by an unskillful fastening, or locking-up of the types, but it is plain that the types were of irregular size as to body, and that the letters were badly adjusted upon the bodies. Some types are high and others low to paper, and there are types that are legible at one end of the face and not at the other. The presswork is wretched: we see the evidences of too weak and badly distributed ink and of uneven impression. The text shows many faults of composition in the division of syllables.
Second Page.
The Enschedé Abecedarium.
[From Holtrop.]To the observer who is not an expert in typography, the workmanship of the book seems that of a man who had no experience in any department of printing: the faults do not appear to be those of a badly taught printer, but those of an experimenter.

For this reason the Abecedarium has been claimed by the Dutch historians of typography as the first production of the inventor of the art. They say that it was printed before any edition of the Speculum, and probably in the first quarter of the fifteenth century. A closer examination of the book does not lead to this conclusion: the printer of the book was, no doubt, a careless workman, but he had been taught the trade. The fragments of the tract are in four pieces, but they were printed in one form of eight pages, and by one impression. This artificial arrangement of the pages, in the arbitrary position which allows them to be folded together in regular order, reveals an expertness in little technicalities on the part of this early printer which is somewhat unexpected. The method of printing sheets imposed in forms of eight pages was not in fashion before it was adopted by Aldus Manutius, of Venice, in his edition of Virgil dated 1501. It is not an invention of the first, but of the last quarter of the fifteenth century, to which period this book belongs.^[5]

The types of the book were not set up by an experimenter or ignoramus. The comparatively even outline to the right of every page shows that the compositor tried to space out his lines and to give every page an appearance of uniform squareness. As full and even-spaced lines are not to be found in any edition of the Speculum, nor in any of the first books of the early printers, we may conclude that the Abecedarium was printed at a later date, when this improvement had been adopted by all printers.

It has been maintained that the book must be very old, because it is printed on one side only, after the fashion of the block-printers. This is an improper inference, for each fragment has the appearance of a spoiled impression which was rejected before the sheet had been perfected by printing on the other side. The unfilled space for the initial letter shows that the work on the sheet was never completed.

The eight faces of types show their relation to each other, not only by common features, but by the occasional appearance of two faces in one book. That they were never used by any printer of Germany, nor by any known printer of the Netherlands, is acknowledged even by those who dispute their age. That they were founded and used in the Netherlands, and probably in Holland, may rightfully be inferred from the language of two editions of the same book, from the Dutch fashion of the letters in all the books, and from the fact that all existing copies or fragments of works in these types have been discovered in the Netherlands. That they were the work of one printer, or of the successors of that printer, is highly probable. But this admission involves difficulties. These eight faces of types were founded on as many different bodies: four of these faces are on bodies nearly the size of English; two of them are on bodies about the size of Great-primer. The modern printer is at a loss to imagine why his unknown predecessor should have cut so many punches and made so many fonts of types with faces closely resembling each other, yet so unlike that they could not be used together. His perplexity is increased when he discovers, after careful measurement, that each face on English body and each face on Great-primer body was cast in a new or different mould. It would seem that the unknown printer of the Speculum not only incurred the needless expense of cutting new punches and making new moulds for every new font of types, but that he intentionally introduced in his printing office bodies so nearly alike that they could not, in the shape of single types, be distinguished apart.

The questions at once arise, Why were so many faces and bodies of types that could be readily mistaken for each other, and were so liable to be mixed together, allowed in one office? Why were so many punches cut for such trivial differences of face, and so many moulds made for such slight differences of body? These questions can be answered only by conjectures fairly derived from the remarkable workmanship of the books. The harsh indentation of the types in the paper shows very clearly that the types were roughly used, and that they wore out rapidly. We can see, also, that the method of making types was as imperfect as the method of obtaining impression. It is possible that the matrices and moulds wore out as fast as the types, but they could not have been renewed if they had not been made by a much quicker and cheaper method than that of modern type-founders. It is not at all probable that these different types were in use together. We may suppose that as soon as a font of types was worn out, it was replaced by another font, which may have been cast from new matrices and a new adjustment of mould. A new font made in imitation of the old one, but made without scientific method, and without regard to exact accuracy, would show the difference in face and body which seems so strange to the modern printer.

These eight fonts of type seem all the more unnecessary when we consider the trivial nature of the unknown printer's works.^[6] The Speculum is the only book of respectable size; the others are so diminutive that they could be classified as pamphlets. They were cheaply made, adapted, apparently, to the wants of school-boys, and were probably sold for small sums. It is evident that the books met with ready sale. We find four editions of the Speculum in two faces of type and in two languages; nineteen editions of the Donatus in six faces of type; six editions of the Doctrinal in two faces; and twelve editions of other books.

From the character of the books, one might judge that they had been printed for the use of some school, and at the suggestion, or under the direction, of the authorities of the church. The Abecedarium was a primer for small children. The books most frequently published, the Donatus and the Doctrinal, were those most needed by very young scholars. The Couplets of Cato, the curt treatise on the Roman Law, and the Praise of the Iliad, are, in size and subject, the books that would be suitable for a boy's school in the middle ages. The Treatises of Saliceto and Torquemada, the Witty Sayings of Great Men and the Eulogy of Pope Pius ii,
1. Experimental Letters Drawn on Wood.
[From De la Borde.]may also be included in the list of books that were intended to be used in schools for the teaching of morals. The character of these works is more juvenile than that of any other typographic printer of that century. Whoever compares them with the ponderous theological works that were printed by Mentel, Gutenberg and Schœffer, and by numerous printers in Germany, and subsequently in the Netherlands, will at once see that this unknown printer made books for boys where other printers made books for men. Probably he could secure no other buyers. His workmanship
2. Experimental Letters Drawn on Wood.
[From De la Borde.]was so rude that it could not be sold to an intelligent or critical reader. His process was suitable only for the cheapest work and the simplest tastes.

It is unnecessary to prove that the types of these books, like the types of the Speculum, were founded in a mould. They show the same features, and must have been made by the same process. It is, however, necessary to show that neither these types, nor any types made in the infancy of the art, could have been cut on wood or metal. There is a tradition, which has found its way in many popular treatises on typography, and even in encyclopædias, that the first types were cut or sawed out of wood. We are told that separate letters, drawn at graduated distances, were engraved on blocks of wood, and that a saw cutting through the intervening spaces separated the fixed letters and made movable types. According to Meerman, the uncouthness of the types of the Abcedarium is fully explained by the acceptance of this tradition. It is necessary, at the outset, to show the impracticability of these imaginary types of wood. This can be done in no better way than by presenting the illustrations of Leon De la Borde, one of the most eminent defenders of the theory. In these engravings, we see how the letters were drawn on the blocks, how lines were marked out to guide the saw that cut them apart, and how the dissected letters were recombined in new positions.

3. Types made from the Experimental Letters.^[7]
[From De la Borde.]

But this illustration really proves the reverse of what was intended: it proves that types may be cut out of wood, but that they cannot be used after they have been cut. In this third illustration, the lines of type are separated by leads,^[8] but the types stand more unevenly in line than the letters of any xylographic book. It is obvious to every printer that they could not have been printed at all, if they had not been leaded. As an imitation, the illustration is of no value, for it illustrates a method of arranging types which was never practised by the unknown printer, whose types were always composed without leads. This pretended demonstration must be put aside as a complete failure.^[9]

Those who have written in defense of types of wood have failed to see that the cutting of the faces is the least difficult part of the work. The real difficulty is in the cutting of the bodies—in making bodies so accurate that they can be interchanged with facility, in all kinds of combinations, without showing distortion in the line of the face. In small types made of wood this accuracy is not possible. Even if it were possible to cut them, it would be impossible to use them. No care could keep them from warping. Types must be wet with ink, and they must be cleansed with lye or water; they must be exposed to changes from heat to cold, from dampness to dryness. Under these influences, the little skewers of wood, for so they must be regarded, would soon be twisted out of shape, and unfitted for future service. It is in this liability to warp that types of wood fail most signally. It is not enough that they can be made to serve for one experiment; the only demonstration of practicability that a printer can accept is that of repeated distribution and recomposition, a feat which has never been done. That types of wood were tried by the inventor of typography is probable; that single leaves were printed, experimentally, is possible; but the statement that any printer used them repeatedly in the printing of books, cannot be admitted. No book was ever printed in Europe with small types of wood. It is time, says Van der Linde, that criticism made a bonfire of these imaginary types.^[10]

The hypothesis of types of wood has been given up reluctantly. It was considered that the singular variety of letters, so noticeable in all the books of the unknown printer, and so contrary to the usage of the modern type-founder, could have been produced only by engraving the types. A demonstration of the impracticability of bodies of wood seemed to destroy with it the only reasonable explanation of the greatest peculiarity of these types. To place this imaginary method of making types on unassailable ground, Meerman offered a modification of the theory. He supposed that the first printers of Germany founded little cubes of metal, with truly squared bodies, upon one end of which the faces were subsequently engraved. The misconstruction of the language of a chronicler of the sixteenth century—who, in trying to explain the process of making types, carelessly placed the cutting of the punch after the founding of the type—seemed a full warrant for this conjecture. It is, however, but a conjecture: there is no credible authority for the statement that the printers first cast the bodies and then cut the faces. Cut types, if made at all, were made only in the way of preliminary experiment. The method is as impracticable as it is absurd. "He must have been an imbecile," says Bernard, "who could not see that the process of founding in a mould which made the body would also make the face."

The allusions to letter-cutting that are so frequent in all the earlier notices of type-making can be readily explained. The cutting is not that of types used for printing, but of the punches by which the printing types were made. The types of the early printers were made by two classes of workmen: he who poured the melted metal was the founder; he who made the model letters was the cutter. Performing the more artistic and the more difficult part of the work, the punch-cutter was properly regarded as the maker of the types.

The variety of faces in the types of the unknown printer can be explained in a much more satisfactory manner than by attributing them to the accidental slips or deviation of the graving tool. The letters of the manuscript books of that century were not uniform; it was not necessary that printed letters should be uniform. The fashion of the day did not require it. On the contrary, it did seem desirable that the letters should be printed with the variety of shapes to which readers were accustomed. Whether this variety of shape in type was the result of design, of accident, or of necessity need not now be considered; in this place it is enough to say that all the early printers made many varieties of the letters which they most frequently used.^[11] It should, however, be noticed that this apparent taste for variety of form was confined to the small or lower-case letters. Two forms of a capital letter are rarely found in the same book, but the same form of capital is occasionally used with two faces of lower-case types that are decidedly different.

The dissimilarity of the small types has been made greater by faults of type-founding and of presswork. In all copies of the Speculum the careful observer will see the impressions of types with imperfect faces. There are many half-formed letters, with little peculiarities of appearance which can be satisfactorily explained only by the conjecture that the types in leaving the mould, carried with them the impress of defects in the matrices. We can see that the types were unequal in height, and that the over-high types have been flattened out under impression. This flattening-out of the soft metal has produced a strange appearance of compactness, making letters that were really separate seem connected. The ink, which was sometimes thin and gray and sometimes thick and strong black, was applied by an imperfect method which has filled the counters of some letters until they are almost illegible, while it has not fairly covered the faces of other letters. The singular irregularities of a collection of types, apparently new on one page and worn-out on another, which have provoked the astonishment of many critics, are chargeable, not to the condition of the types, but to faulty methods of inking and impression. Few persons have a proper notion of the changes that can be given to the appearance of the best modern types by substituting wet for dry paper, hard for light impression, and thin for thick ink.^[12]

How the types of these and of other early books were founded cannot be learned from the vague descriptions of the early chroniclers of typography. We have to conjecture the process from the workmanship of the books. The discrepancies in the bodies and the imperfections of the faces indicate that the process was rude and unscientific, and that the mould was not of metal. It is possible that the maker of these types followed the example of other founders in metals, and made types in moulds of sand.^[13] There are some peculiarities in his types which almost confirm this conjecture. The difficulty encountered in fitting matrices to these moulds, or in adjusting the mould of the face of the letter in proper position on the body, a difficulty that calls for no explanation, may be the reason why the types are so often out of line, crookedly set on body and of irregular height to paper. The feebleness of the sand mould, its liability to damage, and the necessity for its frequent renewal are, possibly, the reasons why we find in the impressions of the unknown printer types of so many bodies, and with such singular defects.^[14] The rounded edges, spotted stems and deficient lines of many of the letters seem the faults of types unskillfully founded in moulds of sand, from metal insufficiently hot, poured in without the force that is needed to make it penetrate all the finer lines of the matrix.^[15]

Koning, the author of a prize essay on the invention of typography by Coster, expresses his belief in the theory that the types of the Speculum were made from punches of wood and were founded in matrices of lead. His belief in the use of these rude implements is based on the well known fact that matrices of lead were frequently used by the earlier German and Dutch printers. Enschedé of Haarlem had in his type-foundry matrices of lead, which he claimed were used by Peter Schœffer in the fifteenth century. Firmin-Didot, the eminent type-founder of Paris, says that punches of wood and matrices of lead were used in his type-foundry for the casting of large ornamental types even as late as the beginning of the present century. His description is as curious as it is instructive.

…I have often made use of this process, … which is to sink in lead, a character cut on wood, at the instant when, melted by heat, the lead is about to harden. Matrices of lead made by this process are subsequently justified for height and for lining, like other matrices, Then, by the ordinary process of stereotyping, one may take from this matrix, a duplicate in metal, which, after having been dressed, is replaced in the matrix in lead, and fitted up to a mould. The melted metal poured in this mould, not only makes the body of the type, but at the same time solders itself to the stereotype [nested in the matrix] which makes the face of the type. By this process one may take from a matrix in lead, a type as perfect as that which is obtained in the ordinary manner. But these matrices in lead will only make a limited number of stereotypes. … By taking the precaution to cool occasionally a matrix in lead, one can obtain from sixty to eighty types, without being obliged to re-enter the old matrix with the punch of wood, or to make a new matrix from the same punch. For vowels, and for the letters that are more frequently used, it is necessary to increase the number of matrices. But whenever the punch re-enters the matrix, the form of the punch undergoes some alteration from the effects of the pressure and the heat. It often happens that the punch is burned during the little time that it is buried in the hot metal. It then becomes necessary to re-engrave the punch. These are the reasons why differences in shape are to be found in the letters that are most frequently used.^[16]

Whether the types of the unknown printer were founded entirely in sand, or in matrices of lead, cannot be positively determined from the appearance of the letters, for it seems that either method of founding would produce types showing similar defects. It is probable that the punches were cut on wood, and sunk in hot metal as described by Didot, and that the types of the Speculum were not only cast in lead matrices, but that the matrices were sometimes conjoined, and that two or more letters were cast together on one body. There is a closeness of fitting in some of the words which cannot be explained with entire satisfaction by the hypothesis that this closeness is the result of flattening out under pressure. One is strengthened in this belief when he discovers that it was not an uncommon practice in the type-foundries of the fifteenth century to join the matrices. Six of the matrices owned by Enschedé, and by him attributed to Schœffer, were made to be combined. These leaden matrices were pierced through their sides with a gimlet-hole, in which an iron wire was inserted to bind them together, and keep them securely on the mould. The method was faulty, for it could not keep the matrices in proper position; it could not produce types uniform as to height and true as to line.^[17]

The thick faces and flattened lines of the types in many of the unknown printer's books show that his types were of very soft metal, probably of pure lead. To satisfy his doubts on this subject, Enschedé cast in some of his antique moulds types composed almost entirely of lead. The experiment succeeded: he was convinced that practical types of lead could be founded in matrices of lead.^[18] Blades carried this experiment to a more successful conclusion, for he put the types to practical use. He had cast for him a collection of types in "unmixed lead," with which he printed five hundred impressions on rough and dry paper. He says that the types showed no appreciable wear; but this is not surprising, for we have evidences that they were printed by an expert pressman on an iron press provided with every appliance requisite for a nice adjustment of the impression.

It is not at all probable that the press of the unknown printer had these handy appliances. All the printing presses made before the nineteenth century had wooden frames, with beds of slate or stone, and platens or pressing surfaces of wood. Impression was given by the direct action of a screw, the force applied being regulated only by the discretion of the pressman. Knight, in his essay on Caxton, says the press of that printer was a modification of the cheese-press, provided with an attachment that permitted the form of types to be moved in and out of the press. German authors say that the first printing press was a modification of the wine-press. Bernard says it was, probably, an improved form of coining or stamping press. But these are only conjectures. We can find no engraving nor any verbal description of the form of the printing press in use during the fifteenth century. The general neglect by all artists and writers of this important auxiliary to printing is an indication that no importance was attached either to the mechanism of the press or to the principle of impression. It seems to have been generally understood that, whatever merit there might have been in the invention of printing, no noteworthy inventive skill had been shown in the construction of the press. It was not only a rude but an old contrivance.

We have many evidences that the press of the unknown printer was of the rudest construction. Some pages have the marks of strong pressure in one corner and of weak impression in another—manifestly the result of the printer's inability to regulate or control the force he exerted. The margins of the Speculum are of unequal width; the type-work is rarely ever parallel with the engraving at the head or at a proper distance from it. On some pages, the types overlap or bite on the wood-cuts; on other pages they are too near or too far from them. One of the reasons why the Speculum was printed on one side only was the deficiency in this press of any contrivance for determining the proper position of the sheet before the impression was taken. The pressman could not print one page truly and squarely on the back of another page. Koning says that the printer did not have the least idea of the means to be used for accomplishing this result.^[19] This defect of the press can be seen in the pages of the small books without illustrations: they were printed on both sides, but the modern printer would condemn the work as seriously out of register.

The most remarkable peculiarity in the presswork of the Speculum is the embossed letters at the ends of the short lines.^[20] They are most noticeable in the two Latin editions, which contain lines of unequal length. To the modern printer the purpose to be accomplished by the use of the old and worn types that produced these embossed letters is apparent at a glance. They served as bearers or guards to shield newer and better types in exposed positions from an impression which could not be regulated. This exposed position was at the ends of the long lines; the types that projected beyond their fellows received the hardest impression, and the printer knew no better method of shielding them than by the insertion of worn types at the ends of the shorter lines above and below.^[21]

This expedient was insufficient. On the margins of many copies of the Speculum can be detected (for the grain of wood is unmistakable) the marks of impressions against wood. It seems that the pages of types were fastened in a mortised block of wood of the same height as the types. This block of wood not only served as a chase to hold the types, but as a bearer to shield the types from uneven impression. It steadied the descent of the platen, and diffused the impression equally over the entire surface. These bearers shielded the types from undue impression, but they made a new difficulty, for they were of the same height as the types. The inking of a form so constructed must have blackened with equal impartiality the types of the text, the worn types used as bearers, and the wooden chase. To lay a sheet of white paper over such a form would smear and blacken it at the ends of short lines and in the margins where no color was required. It became necessary to put a mask over these bearers, so that the ink on the bearers would not be transferred to the paper.

This mask was substantially the same contrivance which modern printers call the frisket. It shielded the white sheet from contact with ink where ink was not required, but could not shield it from impression. It really strengthened and deepened the impression, producing the embossed letters in the short lines and the marks of wood in the margins. On some pages the slipping or displacement of this paper mask caused the false letters to be printed in black; on one other page the mask slipped so trivially that one-half of the false types was printed in black, while the other half was embossed in white on another page the mask slipped over the text type, and obscured the end of the line. These were exceptional errors; the general execution of this part of the work shows that the printer was a man of some intelligence, and that with imperfect materials he performed a very difficult task.

The Frisket, Tympan and Bed of a European Hand Printing Press.

The operation of presswork begins with inking the form on the bed of the press, which, in this illustration, is supposed to contain a form not unlike that of the Speculum, nested in a chase type-high. The sheet is laid on the tympan against guides that keep it in place. The frisket, containing the paper masks cut out to sink the irregularities of the form, is folded down in the line A B, partially covering the paper on the tympan. The tympan is then folded over on the line C D, which operation brings the paper down on the face of the form, ready to receive the impression. These are the appliances of a modern press. The frisket of the unknown printer was of much simpler construction, probably nothing more than a mask of paper laid on the form of types by hand.

1. For a fac-simile (from Holtrop) of this face of type see page 277.
2. A fuller notice of Cornelis the binder will be given in the chapter on the Legend of Coster, in which his relations to early printing will be described. Attention may be called to the significance of the fact that no fragments of any book in the types of the Speculum have been found in the covers or binding of any manuscript book of earlier date than 1467.
3. This work was in use as late as the reign of Charles v. It was enjoined by him that a printer should furnish without alteration "the little book commencing with the alphabet, the little book which directs how to bless the table (grace at meals), and the little book which directs how to answer at the holy mass." Van der Linde, Haarlem Legend, p. 2.
4. Hessels does not describe this as Type VIII, but as the Type of the Enschedé Abecedarium. He thought it "advisable to separate these two little works [the Donatus and the Abecedarium, which are printed in this face], to a certain extent, from the others" but he admits that the types of these books bear the family likeness and cannot be omitted.
5. Berjeau, who accepts this Abecedarium as one of the first products of the invention, says that impositions of eight pages seem more complex than they really are—that the printer had but to fold a sheet, to mark the pages and then unfold the sheet, to see the method at a glance. This reasoning is specious, but it is inconclusive. It was the argument of the courtiers with Columbus after he had stood the egg on its end. Anybody can do it. Simple as the process may seem, the imposition of eight pages of type in one form was not done by any of the early printers, and we have to infer that they did not know how to do it.
6. Caxton, who printed thousands of pages in folio, made use of but eight fonts. Blades, Life and Typography of Caxton, vol. ii, p. xxvii. Gutenberg, who practised printing for thirty years, did his work with not more than six fonts of type, Schœffer, who was a printer and publisher for forty-three years, made use of but six fonts.
7. Leon De la Borde, Debut de l'imprimerie à Strasbourg, pp. 70, 72.
8. Leads are very thin pieces of metal which are inserted between the lines of types to increase the distance between the lines, and to give the printed page a more open and inviting appearance.
9. This apparently easy method of demonstrating the practicability of types of wood has been attempted by many writers. Wetter, the author of a valuable history of printing, published in his book a page printed from types of wood, which he offered as conclusive evidence that types of wood could have been made and were made by the early printers. But his types of wood are larger than those of the Speculum, and they are also provided with leads to keep them in line. Notwithstanding these precautions, they are more out of line than the types of the Speculum. Meerman, in his Origines Typographicæ, printed a few words from types of wood with a similar result; but he showed a practical disbelief in his own theory, by engraving all the fac-similes of the alleged types of wood upon plates of copper. The substitution of copper for wood was, virtually, an acknowledgment of the impracticability of wood types. Schinkel, a Dutch printer, was more successful than either Meerman or Wetter in obtaining a good impression from small types of wood, but he subsequently admitted that his success was but a trick, and that it did not prove that they could be used in the ordinary practice of printing. Léon De la Borde afterward conceded that types of wood would be impracticable.
10. The impracticability of types of wood is cleverly stated by Enschedé:

    "I have exercised printing for about fifty years, and I have cut letters and figures for my father's and my own printing office in wood of palm, pear, and medlar trees; I have now been a type-founder for upwards of thirty years; but to do such things as those learned gentlemen [Junius and Meerman] pretend that Laurens Coster and his heirs have done, neither I nor Papillon [the most clever wood-engraver of France] are able to understand, nor the artists Albrecht Durer, De Gray, and Iz. Van der Vinne either; but such learned men who dream about wooden movable letters make Laurens Janzoon Coster use witchcraft, for the hands of men are not able to do it. To print a book with capitals of the size of a thumb, as on placards, House and Ground, which are cut in wood, and which I have cut myself by hundreds, would be ridiculous; to do it with wooden letters of the size of a pin's head is impossible. I have made experiments with a few of a somewhat larger size. I made a wooden slip of Text Corpus [a body about the size of Long-primer], and drew the letters on the wood or slip; thereupon I cut the letters. I had left a space of about the size of a saw between each letter on purpose, and I had no want of fine and good tools; the only question now was to saw the letters mathematically square off the slip. I used a very fine little saw, made of a very thin spring of English steel, so cleverly made that I doubt whether our Laurens Janszoon had a saw half as good; I did all I could to saw the letters straight and parallel, but it was impossible; there was not a single letter which could stand the test of being mathematically square. What now to do? It was impossible to polish or file them. I tried it, but it could not be done by our type-founder's whetstones, as it would have injured the letters. In short, I saw no chance, and I feel sure that no engraver is able to cut separate letters in wood, in such a manner that they retain their quadrature, for that is the most important part of the work of type-casting. If, however, I wished to give my trouble and time to it, I should be able to execute the three words, Spiegel onzer Behoudinis, better than the Rotterdam artist has done in the Latin works of M. Meerman; but it is impossible, ridiculous, and merely chimerical, to print books in this manner." Van der Linde, Haarlem Legend, pp. 72, 73.

11. This taste for variety in the shape of letters was more clearly exhibited in Greek and German than in Roman types. The Greek types of the sixteenth century are so full of ligatures and variants, that they are undecipherable to the scholar who has been taught the language only in modern text books. So far from trying to make letters readable, the literati of that period tried to make them obscure: they were evidently determined not to make the acquisition of the language easy for their successors. When Francis i of France established the royal printing office, he engaged a skillful Greek penman to design additional varieties of contractions. Two centuries afterward, Pierre Fournier, the younger, a type-founder of Paris, commended the Greek types of his own manufacture as much less complicated than any Greek types then in use. But I count 776 characters in the font. More than 300 of Fournier's contractions, once esteemed as admirable graces, have been rejected by modern type-founders. Blades, who has made a careful analysis of the characters used by Caxton, shows that in the face described by him as 1 there are at least 167 distinct characters. But 24 of these are capitals and 81 are double letters. In faces 2 and 2* there are 380 characters, exclusive of figures, spaces and marks of punctuation.
12. Blades, in his Life and Typography of William Caxton, has given a practical illustration of these changes in Plate ix b, which also illustrates the feasibility of types of pure lead, for a notice of which see next page.
13. The most approved process in the modern art of stereotyping is that in which the mould is made of calcined gypsum or plaster. The same material is used by type-founders in the manufacture of the largest types of metal. The cheapness of sand, and the ease with which it can be worked, make it the most serviceable of materials for all founders who wish to produce cheap castings.
14. To satisfy his own doubts as to the feasibility of casting small types in moulds of sand, Bernard, of Paris, gave to a brass-founder the types of a few Roman capital letters as the models from which he requested founded duplicates. He charged the founder not to dress nor finish the face of the founded letters, nor to give them more than ordinary care. The founded letters so made were printed by Bernard in his history as practical illustrations of the feasibility of sand moulds. They lack the finish of types made by the professional type-founder; they look like badly worn types, but they are legible. The brass-founder assured Bernard that a workman could make one thousand similar types in one working day. Bernard then gave to this founder separate types of a word in Gothic letters and requested him to furnish duplicates of these types founded on one body. The duplicates returned showed the very defects of the types of the Speculum; the thick lines were spotted, and the letters were out of line. Bernard's impression shows that the movable types which made the word were jostled or trivially disturbed at the instant of moulding. A disturbance of this nature would explain the irregularity of line and the rounding of the edges. The spotted and ragged edges of the founded word were probably caused by the roughness of the moulding sand, or by the sticking fast to the mould of bits of metal. It is a proper inference that in both cases the defects were the imperfections of the same process. The experiment of Bernard fully proved the feasibility of making small types in sand moulds.
15. In the sand mould, the hot metal is poured in; in the metal mould, whether worked by hand or machine, the hot metal is forced or cast in. The phrase "casting type," which implies a sudden throw or violent jerk, has entirely supplanted the older phrase of "founding type."
16. Didot, Essai sur la typographie, p. 607.
17. The process seems impracticable, but whoever carefully studies the British and American patent reports, will find specifications of inventions in typography that are much more absurd. There can be no doubt of their use. Koning cites one M. Fleischman, who had not only seen conjoined matrices in the type-foundry of C. Hardwich, of Nuremberg, but had experimentally cast types from them in an old mould that appears to have been made for this express purpose. Speckelinus, Paul Pater, Meerman, Schoepflin, Spiegel, and other early chroniclers, have specifically mentioned types pierced with a hole, and bound together with wire. These so-called types were either punches or matrices. Koning, l'Origine, etc., de l'imprimerie, p. 12.
18. Benjamin Franklin, in his autobiography, has given a curious description of his attempt to supply his defective printing office with types cast in matrices of lead:

    "Our printing house often wanted sorts, and there was no letter-foundry in America; I had seen types cast at James's in London, but without much attention to the matter; however, I contrived a mould, and made use of the letters we had as puncheons, struck the matrices in lead, and thus supplied in a pretty tolerable way all deficiencies. I also engraved several things on occasion; made the ink; I was warehouseman, and, in short, quite a factotum."

19. Dissertation sur l'origine, l'invention, etc., de l'imprimerie, p. 18.
20. It has been shown that book types must be on square bodies. As a necessary consequence every form of types must be squared. If the lines of types in any page are not of uniform length in the metal, and the page is not truly squared, the form cannot be handled nor printed. But although the lines are of uniform length in the metal, they do not always appear so in print. The last line of a paragraph is frequently short; lines of poetry are always of an irregular length. To make the form square, and yet produce this desired irregularity at the end of every short line, the compositor inserts metal blanks, technically known as quadrats. As these blanks are about one-third shorter than the letters, they are not touched by the inking roller; they receive no ink and take no impression, and are consequently invisible to the reader. Quadrats are now regarded as an indispensable part of every font of types, but the appearance of the Speculum shows that the printer of the book had to do his work without them. That he knew the utility of quadrats is apparent, for he used low types as spaces between words. His imperfect press compelled him to reject quadrats at the end of short lines, and to fill the blanks with bearers.
21. To protect types in places similarly exposed, stereotypers insert at the extreme ends of short lines types of flat face expressly designed for this object, which are usually known as guards. When the plates have been made perfect in other points, the guards are no longer needed, and are cut away. When books were printed on hand presses during the first half of this century, pressmen sometimes pasted on or tacked on thin strips of wood around the forms of types to shield the ends of lines from injury. It is a strange surprise to encounter this modern method of protecting types from injury in one of the earliest books.