Chapter 4
Type Material

The most important aspect of Gutenberg's invention lay not so much in the creation and utilization of the printing-press as in the art of manufacturing movable type. Printing consists essentially of reproducing texts by means of movable metal type which has itself been reproduced mechanically. Before describing how type was made in Plantin's time, it is necessary, however, to discuss what stocks of type he and his contemporaries required to carry on their craft.1.

From punch to matrix. Left to right: punch; matrix being struck; one unjustified matrix; two justified matrices.

Nomenclature of type forms. Left: ‘Venetian style’ (c. 1470); right: ‘old style’ (after c. 1495).

N.B. Most authors writing before 1925 used beard for that part of the type that descends from the face (printing surface) to the shoulder. Modern custom uses the word beard to mean front shoulder, or that part which is allowed for descenders.
Parts of metal type and their names.

A Front (of body)
B Back (of body)
C Foot
D Head
E Nick
F Counter
G Beard or bevel showing depth of ‘drive’
H Shoulder
I Hair line
K Main stroke
L Serif
M Type line

Kerns. Many types are said to be ‘kerned’. The kern is any part of the face that overhangs and rests on the shoulder of the adjacent type.

Typefaces and founts1.

The roman and italic are today the two basic letter forms used in printing - a state of affairs which had already been reached by Plantin's time. North of the Alps, however, the angular Gothic or black letter script of the later Middle Ages was still used for handwritten books, and because of this it continued in use in all its angularity as a printed type. For everyday purposes a quicker, cursive, handwriting was employed. In 1557 the Frenchman Robert Granjon adapted this cursive Gothic for printing. Plantin was among the first to use this new style of type. He referred to it as the ‘lectre françoise’ or ‘lectre d'escriture’. In technical literature it is termed ‘civilité’ type. Roman, italic, Gothic, and civilité, augmented by decorative fleurons or printer's flowers, were sufficient for the ordinary run of publications. For some of his scholarly works Plantin also had to have Greek and Hebrew alphabets available. He had a Syriac alphabet cut for the Polyglot Bible. His son-in-law Frans Raphelengius at Leiden had founts of Arabic, Ethiopic, and Samaritan type prepared. In addition to these exotica, music editions required special type.

To be properly equipped a printing press had to possess a whole range of type sizes. It was not enough to have one set of a particular alphabet. A huge folio antiphonary required much larger type than a 24 mo edition of a classical author, and type size also had to be varied within a work in order to attract the reader's attention at particular places - to chapter headings, for example. For each design of type which he used, the printer of Plantin's day, like his modern successors, had to have several complete alphabets in various sizes.

These supplies of type of various kinds are referred to as ‘founts’, and their sizes are expressed in points - according to the Didot system

on the Continent and the pica point system in the English-speaking world. But in the sixteenth century each fount had its own name. It might take its name from the type-cutter who had helped to create it (‘Garamonde’, for example) or from the works customarily set in it: this is how ‘Bible’ and ‘St. Augustine’ type came to be so called. Sometimes the aesthetic qualities of a type decided its name, examples being ‘jolie’ and ‘nonpareille’.

By way of illustration, here are some lines from the chapter devoted to the printing press (and probably written by Plantin himself) in Dialogues françois et flamands, 1567:1.

G: ...But how do you come to have so many kinds of types?

E: That is on account of the diversity of works that have to be printed, either in large or smaller letter. According to them the types have received different names.

G: Is it your opinion that, through being accustomed to make a book in a certain kind of type, they have called such type after it?

E: I understand it so, as in the composition of missals they called some missal types canon and petit canon de messel, glose de messel; lettre de Cicéro, lettre de S. Augustin, because they had been used to printing such authors with these types.

G: Where did the others get their names?

E: Some have taken them from nations which have used them commonly. Of this sort are some we call romain and gros romain or texte, ordinary romain, petit romain, and the italics, lettre françoise, and Greek type.

G: Have others been named for different reasons?

E: Oh yes. Because of their great beauty some are called mignonne, nonpareille and paragon. Others have taken their names elsewhere, such as gros and petit canon, texte, two line tourné letters, gros trait, grand and petit bourgeois, lettre bâtarde, lettre de somme or modern, and lettre de parchemin.2.

The names given to types varied to some extent at least from one region to another. When Albert Moretus was ordered by the French authorities in 1810 to provide them with inventories of his stocks of cast type, he was careful to stress the fact that the terms he used were different from those current in France.1. It is even quite possible that the names varied from one printing office to another, although the type-cutters and the foundries must have tended to exert a standardizing influence over considerable areas as the terms they used became current among their customers.

The uncertainty of usage noticeable in Plantin's earlier years shows that the practice of naming the various sizes of type must then have been of recent date. Different names were sometimes given to the same fount2. and two different founts might be referred to by the same term.3. There were also founts that had not yet been named. In an inventory of 1566 Plantin had to describe a fount as ‘Petit texte de Hautin entre la nonpareille et la Garamonde bréviaire ou bible’. Later he was to term this type the grosse nonpareille and the body on which it was cast as the coronel.4. The fact that the name of the French type-cutter Claude Garamond, who had died in 1561, was given to one of the founts is another indication that the nomenclature was a recent innovation, at least as far as Plantin and his colleagues were concerned. The terminology which the great printer adopted in his early years remained in use in the firm until the nineteenth century.

The names of founts used in the Officina Plantiniana are given in the table on p. 56.5.

Names of Type Sizes as used in the Officina Plantiniana

The equivalents to Didot and Pica points are approximated averages. The value ‘20 lines in mm’ is to be understood as measured from the first to the twenty-first line, on corresponding points of the printed image.

The size or body of a fount is not obtained by measuring the size of the printed image of any of its letters, but is taken from the depth of the shank of the cast letter. This measurement is always slightly greater than the depth of the largest printed image. For example,

(7) Opposite: The furnaces of the type foundry in the Plantin house, on the second floor above the gallery. They probably date from 1620-22 when Balthasar I Moretus had this wing added to his premises.

(8) Top: Punches from the Plantin-Moretus Museum collection, the work of some of the greatest sixteenth-century punch-cutters: above left Van den Keere; above right Le Bé; below left Granjon; below right Garamond. Each of these craftsmen had his own way of finishing off his punches, seen in the length of the shanks, in square-cut, pointed, or rounded ends, so that the sets are quite easily distinguishable. Note among Granjon's punches one with the shank partly cut through (above left) and a small punch (below right). The latter is an accent which could be fastened to the larger punch and struck together in a matrix.




(9) Bottom: Moulds. Left a closed mould seen from above, with spoon. Right an opened mould with a cast letter (extreme right, partly visible, top turned to the left) and, barely visible, a matrix. Below right two cast letters, one still with the jet, the other with it broken off. Clearly visible on both moulds is the clamp which held the matrix in place underneath while the molten lead was poured in. The mould on the left also has hooks which served to prise loose any cast letter that had stuck.

(10) Opposite, top: A set of matrices (Ascendonica roman by Robert Granjon, ma 7) in a wooden box: possibly one of the 39 boxes supplied by Hendrik van den Keere in 1576. In the front are two punch boxes with painted lids in German style, possibly bought at the Frankfurt Fair.




(11) Opposite, bottom: Matrices. Left in lead-(ma 9); right in copper (ma 78) of the same body size (‘Grosses Capitales Extraordinaires’ by Garamond, later completed by the addition of Greek capitals by Van den Keere). The lead matrices, which lack the Greek capitals, are more roughly finished: they may have been a cheap set, meant for sale at Frankfurt.

(12) Opposite: Two series of initial letters cut in wood and to be cast in sand, (st 1: ‘La plus grande romaine’, 73 pieces; st 78 ‘Gros Flamand’, 58 pieces; both by Hendrik van den Keere). Two examples of type cast in sand are the letters D and V, centre left. The part cast in sand (with the type face) is relatively thin, as can be seen in the V, but it was applied to a lead block to achieve the correct height-to-paper. In the foreground matrices of two similar sets of music notes, the top row unjustified (and thus out of shape), the lower row justified.

in the ascendonica as used by the Plantin press the depth of the lowercase x is about 3 mm (⅛ inch), of the lower-case p and h about 5.5 mm (7/32 inch), while the depth of the shank of the cast type is 7 mm (c.9/32 inch). It is this measurement, which also determines the minimum distance between the lines in a text, that is indicated by the term ascendonica (or nowadays by the number of points in either the Didot or the Pica points system). Generally speaking a particular size of printed image corresponds to a particular size of body. Quite often, however, a smaller type was cast on a given body size1. - this might be done to introduce more white space between the lines for the sake of better legibility - or the relative size of the printed image to the body might be increased - by casting a slightly larger type on a given body size - to produce a more compact effect.2. In the latter eventuality the ascenders or descenders (or both) of the lowercase letters usually had to be reduced in length.3.

The Latin alphabet has twenty-six letters. For each of these characters the printer needs lower-case letters and upper-case or capitals; he usually requires an additional set of slightly smaller capitals, known as small caps. Some of the letters in the sixteenth century had more than one form. The lower-case s, for example, nearly always occurred in the short (s) and the long (f) variety. On the other hand, certain letters might be missing. The w and capital J did not figure in Roman type, which was mainly used for printing texts in Latin, were w does not occur and J was rendered by I. In addition to these basic sorts there are also several punctuation marks, figures, letters with diacritical marks, and the many ligatures (tied letters) so beloved in the Renaissance period. A complete roman and italic alphabet con-

sisted of an average of 120 to 150 signs, the number being occasionally much greater than this.1. The heavy black letter alphabets or ‘flamandes’ were less richly provided with ligatures and diacritical marks, and had only one size of capital letter (no small caps), so that the total number of signs was considerably less than in the roman and the italic. Civilité type came halfway between the black letter and roman in this respect. A good fount might contain about 120 different items. Far and away the greatest number of signs was found in Greek alphabets. Because of their extensive, if not excessive use of ligatures, hardly any founts of Greek letters had less than 250 characters while some approached 500.2.

The illustrations (pp. 59-62) show typical numbers of signs.

Punches and matrices3.

The metal type needed for printing was cast one letter at a time in matrices. These matrices obviously had to be shaped so as to produce the desired letter and this meant that a number of processes had to be carried out, and a number of implements prepared before the actual casting of type could begin.

These processes, which lie at the basis of the whole craft of printing, and on which its aesthetic results depend, are described in the chapter of the Dialogues françois et flamands already quoted above.

Ma 96. ‘Texte flamand’ (16.6 Pica points): Hendrik van den Keere, 1570; 113 matrices.

Ma 20 a and b. Texte (16.6 Pica points): Claude Garamond. Mentioned for the first time in Plantin's 1563 inventory; 176 matrices:

Ma 15. Parangonne (18.7 Pica points): Robert Granjon. Mentioned for the first time in Plantin's 1563 inventory; 162 matrices.

G: Let us proceed then and begin with the types, since you have put them foremost. How are they made?

E: First the punch is made. This is a long piece of steel, on the end of which is engraved the desired character.

G: What becomes of that?

E: When it is made it is struck into copper and a matrix is made, which is nothing but the impression of the character struck, exactly as when a seal is impressed in wax.

G: What is the purpose of the character thus struck into copper?

E: Into this matrix the type-metal, such as lead or tin, with which they wish to make the type, is poured, in a mould.

Ma 108. ‘Gros texte’ (16.6 Pica points): Robert Granjon, 1567; 180 matrices.

This exposition is brief in the extreme and does no more than indicate what the first stages in the manufacture of type were, and states that a punch and a matrix were the result.

The first step was to cut the letter in relief on the end of a steel shank. This was the punch. Sometimes the punch-cutter had to use additional small punches to deal with details as the counters of the letters a, e or g. There are fifteen of these counterpunches in the Plantin-Moretus Museum, the only sixteenth-century examples known.1.

Ma 142 and 143. Mediane (11.3 Pica points): Pierre Haultin. Mentioned for the first time in Plantin's 1561 inventory; 407 matrices.

The punch was then struck in a copper block by means of a hammer. This rough treatment knocked the copper out of shape and the ‘strike’ as it was termed (the officina's French texts called it ‘une frappe [creue]’, the Flemish texts ‘een [rauwe] afslagh’) had to be further worked upon. The block itself had to be filed until it once more formed a regular parallelopiped and the impression of the character struck had been perfected. This ‘justified’ matrix was then ready for use.

Very occasionally punches were struck in lead, thus producing lead matrices. The twenty-three lead matrices of the ‘Grosses Capitales Romaines Extra-ordinaires’ preserved in the Museum are the only known instance of this happening in the Plantinian firm.1. These were capitals of very large dimensions that could be more easily impressed and justified in lead than in copper. Moreover, only a small number had to be cast from these matrices and this obviated some of the risks entailed in casting lead in lead. A set of copper matrices was also struck with the same punches.2. It should not be concluded from this that the lead matrices proved unsatisfactory in use and had to be replaced with the copper set - from the records it would seem that the lead matrices were actually prepared after the copper ones.3. It is very likely that the lead set constituted a relatively cheap strike for sale at the Frankfurt Fair. However, for one reason or another there were no customers and the set returned to Plantin's personal collection.

Equally exceptional were the punches cut in wood instead of in steel. They were not used to produce matrices but were impressed in sand, in which the type was subsequently cast. The ‘lectres en bois de Grosse Romaine extraordinaire [pour jecter en sablon]’ were made by Hendrik van den Keere in 1575. When he delivered them he also provided some sample characters he had cast in sand.4. A similar

series, this time a black letter, was sent in 1580.1. These were extraordinarily large characters and could hardly have been produced with the traditional steel punch, struck in copper matrices, nor could they have been cast in the rather small portable moulds normally in use.2.

For a given set there were usually fewer punches than matrices.3. It was possible, for example, to combine a punch for a diacritical mark with one for a letter and so produce a whole range of modified vowels. One small additional punch of this kind greatly increased the number of possible matrices. All that was required was accurate placing of the two punches when striking them in the matrix.

In theory it was possible to strike an unlimited number of matrices with one punch: that is to say, provided it did not break or get knocked out of shape. And with one matrix, always providing that it was not damaged in any way, a theoretically unlimited number of letters could be cast. It was therefore the punch which determined the quality of the type and it was the manufacture of punches that demanded the greatest amount of time and skill.

In the pioneering days of printing the typographer had to cut his own punches, strike matrices, and cast the type - or at least have these operations carried out in his own workshop under his direct control. By the end of the fifteenth century, however, specialization had begun to develop and professional punch-cutters and type-founders appeared. At first one craftsman would practise both skills and it remained normal for punch-cutters to have a foundry for casting type. However, there were already type-founders in the sixteenth century who hardly ever created their own type designs but were content to work with matrices prepared by their more skilful colleagues.

In Plantin's time it was therefore already possible to stock up with lead type from specialist firms. Many of Plantin's contemporaries

and competitors were quite happy to do this, but not the master of the Gulden Passer:1. he did not want to be dependent on what type-founders happened to have in stock. He wanted to have the most beautiful types available and as far as possible they had to be exclusively for his own use. He could only do this by buying up punches and matrices and with these producing type to meet his requirements. Plantin carried out this policy - and on a scale that seems far to have exceeded the actual needs of his press. He seems to have had a veritable mania for collecting punches and matrices.

In 1556, scarcely a year after he had started printing, Plantin owned at least four sets of matrices.2. In 1561 the number had already risen to twenty-two, five of the sets being unjustified strikes.3. A year later Christophe Plantin had to flee Antwerp and all his possessions were officially auctioned in the Vrijdagmarkt.4. The goods that came under the hammer included Plantin's stock of cast type,5. but not his greatest typographical asset: his collection of matrices had been taken to safety in time. He made use of his exile in Paris to buy new sets of punches and matrices.

The list drawn up at the end of 1563, after his return to Antwerp, mentioned twenty-nine sets of justified matrices, six sets of strikes and eight sets of punches.6. When Plantin put his signature to the contract with members of the Van Bomberghen family, he was not entering the partnership empty-handed. His collection of punches and matrices remained his personal property but it was put at the disposal of the new enterprise and was regarded as the equivalent of a capital investment of 1,200 fl. In addition Plantin was paid an annual sum of 60 fl. under the terms of the contract for the use of this equipment.7.

Plantin continued to build up his collection during the years of the partnership. The list compiled in 1566 of the ‘utensiles d'imprimerie acheptées par moy et mon frère [i.e., Pierre Porret in Paris] depuis la compagnie faicte’ notes the acquisition of 13 sets of punches, 17 sets of matrices, and a number of moulds, the total value of these items being 1,358 fl.1. The partnership was dissolved in 1567 when Plantin's associates had to flee from the persecuting Alva. At least four sets of Hebrew matrices belonging to Cornelis van Bomberghen remained behind in the Plantin press.

By that time Plantin's punches and matrices already constituted an impressive and unique possession. He was very proud of them and on 19th December 1566 he wrote to de Çayas:2. ‘Quant aux charactères je les ay tous taillés et en ordre et les ay par le moyen de mes amis recouverts et acheptés de longue main, à tels frais, travail et nombre d'argent qu'on n'y pourroit bonnement mettre prix: d'autant que je ne pense pas qu'il s'en trouvast encores autant ensemble de si beaux et bons en aucunne partie de toute l'Europe, ainsi comme plusieurs des principaux imprimeurs et gens à ce cognoissants de la France, de l'Alemagne et de l'Italie l'ont rescript et maintes fois confessé en mon absence, et puis après en ma présence à la foire de Francfort.’

Plantin may be suspected of some exaggeration in this letter in which he was trying to arouse the enthusiasm of Philip II's secretary for the Polyglot Bible project. Nevertheless the printer was able to provide the Spanish king and his advisers with evidence in support of his claims. In 1567 he published his famous Index sive specimen characterum Christophori Plantini, some copies of which he sent to Madrid.3. Three copies survive today, all of them to be found in the Plantin-Moretus Museum. This Index was issued in two slightly different versions, one showing 41, the other 42 types. It presented

a selection of the principal founts then at Plantin's disposal - 12 roman, 10 italic, 3 black letter, 3 civilité, 6 Greek, and 7 Hebrew.1.

Plantin did not rest on his laurels. The ‘Registre de touts les poinsons, matrices et instruments [i.e., moulds] appertenants à Christoffle Plantin’ compiled in 1572 showed that his collection had been appreciably enlarged.2. The inventory lists 12 strikes (‘frappes creues’), 56 sets of justified matrices, and 31 sets of punches. Seven sets of matrices and one of punches must have been left out, while ‘toutes les matrices de l'Hebrieu et Syriac tant creues que justifiées, qui sont tout ensemble en une boite’ were mentioned, but without being itemized.

Before 1570 the sources from which Plantin obtained his punches and matrices were quite numerous and varied. Through Cornelis van Bomberghen he acquired at least four sets of matrices for Hebrew type. These matrices (or at all events the punches with which they had been struck) dated back to the beginning of the century. They had been used at Venice by Cornelis's uncle, Daniel van Bomberghen, for his famous Hebrew editions.3. In 1565 Plantin had in his service Jacques Sabon, a type-cutter and founder who later went to live in Germany; as head of the Egenolff type-foundry, in Frankfurt, he played an important part in the development of the craft in Germany. Yet while he was at Antwerp Jacques Sabon had only worked one incomplete set of large capitals.4.

At this time there were two type-cutters and founders working in Antwerp who compare favourably with other memorable sixteenth-century practitioners of the art. François Guyot, a Frenchman by birth, was active in the city from 1539 until his death in 1570.5. His importance lay in being one of the first to adapt elegant French type faces and introduce them into Southern Netherlands typography.

Ameet Tavernier was his contemporary in Antwerp and died in the same year as the Frenchman. Tavernier was a pupil of Joos Lambrecht, the Ghent type-cutter who was so deeply imbued with the Renaissance spirit: he was at the same time printer, seal-engraver, schoolmaster, and poet. Tavernier's work was superior to that of Guyot and he ranked high among the great typographical artists of the second half of the sixteenth century.1.

The young Plantin used material from these two craftsmen. The first works that left his officina were largely set in Guyot and Tavernier types. At first, however, he was content, or was constrained to buy ready-cast type. Only one set of Guyot matrices2. and one set that can be attributed to Tavernier3. are listed in the inventory of 1561. The Plantin-Moretus Museum possesses other sets of matrices by these two type-cutters,4. although two of the Tavernier sets are not mentioned until the inventory of 1588, where they are listed as strikes5. - and it is in this unworked and unused state that they have been preserved. The other sets cannot be identified in the inventories of Plantinian types (up to 1652). They must have been bought by one of the Moretuses at a later date, probably when the effects of a typefounder or printer were put up for auction;6. these also remained unused.

Plantin did not buy punches or matrices from these two outstanding Antwerp craftsmen after 1561. He remained on friendly terms with Tavernier and he made use of Guyot's services as a type-founder,7. so it is clear that he did not stop buying their products for any personal

reason. He may not have been completely satisfied with the quality of Guyot's work, but even the most difficult client could not have faulted that of Tavernier. The explanation is probably that Plantin did not want to use types which any of his Antwerp colleagues could obtain, or already had in their possession. He wanted to have more exclusive designs and he found them in France. From 1561 until 1570 he obtained most of his punches and matrices from French craftsmen.

Claude Garamond is regarded by many experts as the greatest type-cutter of the sixteenth century and one of the greatest of all time.1. His roman alphabets are among the most beautiful ever made. In 1556 Plantin already possessed a ‘Garamont petit romain’, and a ‘mediane Garamont rommain’ is added in the inventory of 1561.2. Garamond died in Paris in 1561. His stocks and materials were publicly auctioned. Plantin was able to acquire the matrices for a Hebrew alphabet, but this set had actually been made by Guillaume Le Bé,3. not by Garamond. All that Plantin managed to obtain of Garamond's own work were a number of ‘varia’ and uncompleted sets.4. Guillaume Le Bé did better, getting the punches of at least four and possibly six roman alphabets - Garamond's last important creations. Le Bé bought most of these directly from the widow, not at the auction. It was at this time that Le Bé ran into financial difficulties. On the specimens of his ‘tres gros hebrieu’ (cut in 1559) he sadly noted that in 1562 ‘à cause des troubles’ (i.e., the beginning of the French religious wars) he had been obliged to sell the punches, moulds, and matrices of this set to Plantin. He sold the punches of Garamond's augustine and

bible to the printer at the same time.1. Between 18th April and 4th July 1573 Plantin bought the matrices of a further Garamond petit romain and a mediane romain in Paris.2. This was probably from Le Bé again, for between 20th September and 6th October 1573 the type-cutter received 5 pond 10 schellingen for the capitals of a garamonde romain and between 6th October and 29th November 1573 a further sum of money for a Garamond parangon romain.3. Plantin in fact made these various purchases in 1573 on behalf of the Ghent type-cutter Hendrik van den Keere, whose activities will be described in more detail below.4. After Van den Keere's death these sets, together with the bulk of the rest of his typographic material, passed into Plantin's collection.5. The printer left them unused.

Guillaume Le Bé also ranked among the great masters of the craft. His speciality was Hebrew. The nineteen alphabets which he cut between 1545 and 1591 were some of the finest of their kind. They had a decisive influence on the development of Hebrew type faces.6. As the Polyglot Bible project took shape, Plantin began to feel the need of more Hebrew alphabets to augment Van Bomberghen's Hebrew types, the Lectre hébraique de Bé pour Garamont (the vray Texte Hébreu à la façon de Venice, as he usually termed it), which he had purchased at the Garamond auction of 1561 in circumstances outlined above,7. and Le Bé's tres gros hebrieu that he had acquired in 1562.8. Plantin turned once more to Le Bé, who supplied him with a double parangonne and a double mediane about 1566 and a coronelle in March 1570.9.

Pierre Haultin (died 1587?) was another important French master

of this period.1. He had a bookshop in Paris but, being a Huguenot, he regularly left the capital in times of tension for the greater safety of La Rochelle or Lyons. Plantin had quite a lot of contact with him in the early years of his career. The texte romain in the 1556 inventory may possibly be Haultin's work.2. The next inventory, that of 1561, lists no fewer than seven sets of matrices as having been supplied by him.3. However, the transactions stopped as suddenly as they had begun. The 1563 inventory reports a Haultin bible grecque and in 1565 a coronelle romaine was delivered,4. but these were the last. It is true that in July 1567 Plantin tried, through Pierre Porret, to obtain strikes of a Greek alphabet, but this was with the intention of reselling the set at Frankfurt.5.

Another French type-cutter had replaced Haultin in Plantin's favour. This was Robert Granjon, also one of the great masters of the craft, who was born and raised in Paris but lived also in Lyons.6. He devised the civilité in 1557, and although he cut all the other typefaces as well, his importance lies in his italic alphabets. Granjon's work helped to determine the future development of the italic: his influence is comparable to that of Garamond and Le Bé in their respective fields.

In 1556 Plantin already possessed two sets of Granjon matrices for

cursive alphabets, namely a cicero italique and a petite italique.1. The 1563 inventory lists six more sets - 4 italic, 1 Greek, and 1 mediane faceon d'escriture à la main Granion, this being the new civilité which Granjon had ‘invented’.2. In November 1564 Granjon arrived in Antwerp. He remained there until August 1567 and later returned for a shorter stay, from November 1569 to April 1570. It is not known whether Granjon worked for other Netherlands printers while he was in Antwerp, but he certainly received many orders from Plantin. In this period he supplied a very large proportion of the punches and matrices that are today counted among the treasures of the Plantin-Moretus collection. Granjon's final departure for France in 1570 virtually put an end to his dealings with Plantin, except for a few occasional orders - punches and matrices for a nonpareille cursive (1573-74) and a jolye Grecq (1574-75)- for which the printer paid either at the Frankfurt Fairs, or through his son-in-law Egidius Beys in Paris.3. The Plantin-Moretus Museum possesses no fewer than 40 sets of punches and matrices that are wholly or partly the work of Granjon (16 italic, 10 roman, 5 civilité, 5 Greek, 1 Syriac, and 3 of music type).4.

These bare figures cannot fully convey Granjon's importance to the building up of Plantin's collection of types and to the evolution of the Plantinian typefaces. As far as can be made out, Plantin in his dealings with Garamond, Le Bé, Haultin, and even with the Antwerp masters Guyot and Tavernier, had to be content with what they happened to have ready in stock (the three Hebrew alphabets which he ordered from Le Bé for use in the Polyglot Bible were an exception). This was also the case with Granjon before 1564 and after 1570. From 1564 to 1570, however, contacts between printer and punch-cutter were much closer, and from the agreements that were drawn

up it appears that the latter worked largely to order.1. By commissioning appropriate work from Granjon, Plantin was able to fill the many gaps in his collection and also supplement, and sometimes even remodel the sets already in his possession.2.

After Granjon's close association with the Plantin press had ended in 1570 his place was immediately taken by another punch-cutter. This was Hendrik van den Keere the Younger, of Ghent, who was also known by the French form of his name, Du Tour. He was the son of Hendrik the Elder who himself had continued the business of his old master Joos Lambrecht. The younger Van den Keere was the greatest Flemish punch-cutter of the sixteenth century.3. Although his roman alphabets never quite equalled the elegance of his French models, they were nevertheless strongly designed, easily legible, and at the same time economical, because of their smaller ascenders and descenders. His speciality, however, was the ‘flamande’: his black letter alphabets were among the most beautiful ever designed. On 7th January 1568 he supplied Plantin with 21 matrices for fleurons and on 16th June 1569 he contracted to deliver strikes of a nonpareil gothic within five to six weeks.4. Orders did not become really frequent until after Granjon's departure, but from 1570 until his death in the summer of 1580, Van den Keere supplied Plantin with punches and matrices with unrelenting regularity, and supplemented

or modified existing sets, greatly extending the Plantinian typographical collection. Altogether the Ghent craftsman delivered 44 sets of punches and matrices (14 roman, 14 Gothic, 1 cursive italic, 1 civilité, 2 Greek, and 12 music types), and also a number of fleurons and various signs.1.

Hendrik van den Keere died between 11th July and 4th October 1580 in the prime of life, probably of blood poisoning following an injury to his leg.2. In addition to the material he had supplied to Plantin he had a private collection consisting partly of his own work and partly of that of other masters.3. On 4th October 1580 Van den Keere's widow offered to sell Plantin this collection.4. The letter spoke of ‘all the punches, matrices, and moulds’ but it is clear from other documents that the widow and Thomas de Vechter (Hendrik's foreman, who continued the business) kept back quite a number of these. Twenty-two sets of punches and 26 sets of matrices, valued at 2,250 fl. were offered to Plantin for 2,000 fl. Letters were exchanged and the bargaining continued until finally, on 15th February 1581, a contract was signed whereby Plantin acquired 20 sets of punches and 12 sets of matrices for the sum of 1,400 fl.5. It was stipulated that the children of Hendrik van den Keere and Thomas de Vechter should be able to buy back the material at the same price. They did not exercise their right and the punches and matrices - the work of Van de Keere, Granjon, Garamond, and Tavernier - remained Plantin's property.6.

This was Plantin's last major acquisition. He had built up one of the largest collections of typographical material ever assembled by a

single printer, and he expressed his justifiable pride in it in many of his letters. As with all collections, not all the items were of the same high quality. Some punches and matrices remained unused - sometimes for no ascertainable reason - while other sets were used for a time and then discarded as typographical fashion, or Plantin's tastes changed with the passing years and as new alphabets became available.

At no stage in his career did Plantin have more than a relatively limited number of founts in use, but he could, if he so wished, have a range and variety of types cast from his matrices that was not equalled by any printer of the Renaissance and Baroque periods. The inventory of cast type in stock in May 1575 (Inventaire des lectres de l'imprimerie de C. Plantin) lists no fewer than 67 founts: 26 roman, 13 italic, 11 black letter, 8 Hebrew, 6 Greek, 2 civilité, 1 Syriac, and 7 music.1. For a time in 1585 Plantin was considering winding up the business. With this prospect in mind he had a type specimen set in the various alphabets and founts then in his possession.2. This document catalogued a total of 101 founts of type: 33 roman, 24 italic, 18 black letter, 4 civilité and bastarde, 8 Greek, 10 Hebrew, 1 Syriac, and 3 music - and even then the list was not complete.3.

After Plantin's death in 1589 the collection was divided up, part going to Jan Moretus in Antwerp and part to Frans Raphelengius in Leiden. It was probably already smaller by a few sets. At that particular moment it was in fact housed in three different places. At Antwerp4. there were 36 sets of justified matrices, 28 sets of strikes and a comparatively small number of punches (647 altogether, 364 of them being for civilité alphabets and most of the rest for music types). In 1583 Plantin had taken a quantity of typographical material with him to Leiden. In 1585 he left behind there a total of 1,191 punches

(10 sets of black letter, 100 lettres syriennes, i.e. Syriac, 172 hebraiques de diverses sortes), 30 sets of justified matrices, and 9 sets of strikes.1. The rest, that is to say most of the punches for his roman and italic alphabets (1,736 altogether), six sets of justified matrices, and 14 of strikes, were in Plantin's warehouse at Frankfurt,2. for reasons that will be discussed below.3.

Naturally Jan Moretus and Frans Raphelengius kept the material Plantin had left at Antwerp and Leiden respectively. The stock at Frankfurt was divided between them. As his share Raphelengius received 290 punches (a jolie romaine, a colineus cursive, a petite musique and a number of fleurons) and 5 sets of strikes.4. The rest passed to Jan Moretus. If the figures are totalled it appears that Plantin left his heirs 3,574 punches, 72 sets of justified matrices, and 51 sets of strikes.5. Jan Moretus's share of the total collection was 2,093 punches, 42 sets of justified matrices, and 37 sets of strikes;6. Raphelengius received 1,481 punches, 30 sets of justified matrices, and 14 sets of strikes.

The Leiden and Antwerp branches of Plantin's family remained on good terms with one another. They lent each other matrices7. or had strikes made from their punches for their kinsmen.8. Between 1590 and 1601 a number of punches were exchanged for matrices at the request of the Raphelengii.9. When the latter began to dispose of their material the Moretuses were given first option. In 1613 and

1619-20 the bulk of the Plantinian collection returned to Antwerp from Leiden, together with a number of additions which the Raphelengii had commissioned. These included punches and matrices for an Ethiopic and Samaritan alphabet that had been made for Scaliger's Opus de emendatione temporum, 1593.1. The Moretuses, however, were unable to acquire the Raphelengian punches and matrices for an Arabic alphabet. These had been sold in about 1612 to the English Arabic scholar William Bedwell.2.

The Plantinian collection was largely intact again. Thereafter it was to undergo no major alteration. The type-founders who worked for the Moretuses occasionally cut a punch to replace one that had been lost or damaged, or struck and justified a matrix.3. Very few new sets of punches and matrices were ordered or purchased. In the seventeenth century there was only the nonpareille romaine achaptée de la vesve de Th. Strong. Thomas Strong was an Irish type-founder who worked for the Moretuses from 1600 to 1624.4. This set was probably bought largely as a gesture of goodwill to the widow; in fact it corresponds with a Haultin nonpareil, for which the house had possessed justified matrices since 1561.5.

There was more activity in this field in the eighteenth century. Johan Michael Smit worked in the Plantinian press from November 1732 to April 1736 arranging the typographical material and re-justifying matrices which had been badly spoiled. He also brought a few Garamond and Granjon founts up to date by cutting new letters.6. The well-known Belgian type-cutter Jacques-François Rosart was given a more important task. In February 1758, one year before he

left Haarlem to settle in Brussels, he sent Franriscus Joannes Moretus 182 matrices, with punches, of a garmond gros oeil.1. In his covering letter he wrote ‘I will send you the second in three or four weeks’. The ‘second’ referred to was a colineus roman, the 196 matrices and 134 punches of which have also been preserved.2. At this time Franciscus Joannes seems to have been seriously thinking of having all his roman alphabets modernized. In about 1760, Jan Baptist van Wolsschaten, a member of a prominent Antwerp family of punch-cutters and type-founders,3. cut an augustin roman for him.4. A few years before, in 1757, the famous French craftsman Pierre-Simon Fournier had got in touch with the Moretus firm and announced that he had bought the type-foundry belonging to the Mesdemoiselles Le Bé. He offered to sell a number of strikes of which he possessed duplicates, but the Moretuses do not seem to have taken the matter any further.5.

That was as far as the modernization went.6. For the rest there are a number of punches and matrices in the collection of very heterogeneous origin that the Moretuses must have bought in the seventeenth or eighteenth century and which they certainly did not use. They probably acquired these when the estate of a printer or typefounder was put up for auction.7. The material includes three sets of Tavernier matrices, one set by François Guyot, and matrices for two sets of capitals which date back to the fifteenth or early sixteenth century.8.

In the course of the centuries the Plantinian typographical collection suffered a certain amount of inevitable depletion. Some entire sets

were lost; individual punches or matrices are missing from others. That the collection should have been preserved so largely intact, however, is nothing less than miraculous. In the Plantin-Moretus Museum there are 4,477 punches, 15 counter-punches, 15,825 justified matrices, and 4,681 strikes. The bulk of the collection consists of material that Plantin himself brought together. The Imprimerie nationale in Paris, the house of Enschedé in Haarlem, the Oxford University Press, and the Vatican all have important collections of punches and matrices, but in none of them is the sixteenth century so plentifully represented, and by the work of the greatest type-cutters of the age. It is only in this Museum that the work of Claude Garamond, Robert Granjon, Guillaume Le Bé, Hendrik van den Keere, and other craftsmen of their period can be studied comprehensively and in detail. Here their individual techniques of punch-cutting can be examined, impressions can be taken of the characters they engraved from the metal punches, and type can still be cast in their matrices. The Plantin-Moretus collection is therefore unique. It is one of the chief treasures of the Plantinian house and one of the most splendid relics of the golden age of printing.

Some of the matrices were found wrapped in paper. Others were preserved in oak boxes which still house them today. A number of these boxes undoubtedly date back to Plantin's time. They may be the boictes nouvelles of which Van den Keere delivered thirty-nine in February 1576, and for which he received the sum of 3 fl.1. For some years past the punches have been stored by sets in suitable wooden boxes. The Museum also has eleven pinewood boxes which must have been specially made to hold punches. Their lids are decorated with various paintings in a typically German style. They appear to date from the sixteenth century and were probably bought by Plantin or Jan Moretus at the Frankfurt Fairs.

Plantin himself designed no type, and he seems to have given the punch-cutters who worked for him a free hand. Only one instance is known of his specifying what he wanted. Plantin's instructions

for the coronelle Hebrew used in the Polyglot Bible and his sketches for the letters aleph and beth can be seen in one of Guillaume Le Bé's albums in the Bibliothèque nationale in Paris.1. There is only one occasion recorded on which the printer is known to have consulted an expert about an alphabet. This was when his friend Guillaume Postel, the great French orientalist, gave him detailed advice as to how Granjon was to cut a Syriac fount.2. Through his choice of types, however, and through his contact with the greatest type designers of the Renaissance, Plantin exercised an important influence on the evolution of the printed letter in Western Europe. He introduced the French roman and italic alphabets in their most elegant form into the Southern Netherlands and familiarized the Flemish type-cutters with them. Undoubtedly it was Plantin's orders that led Hendrik van den Keere to take up the roman alphabet. The types used in the Leiden Officina Plantiniana, together with the Van den Keere material which Thomas de Vechter took to that university town after 1582, became a source of inspiration to Dutch type-cutters in the seventeenth century. These Dutch craftsmen in turn exercised a decisive influence on succeeding generations of Western European type-cutters. Modern roman and italic types can be said to have derived from sixteenth-century French alphabets by way of the Plantin House and seventeenth-century Holland.

These historical considerations give the typographical collection of the Plantin-Moretus Museum a significance that far exceeds its intrinsic value, great though this is. So the fact that the entire collection has at last been fully catalogued, and the makers of the various sets identified, is of the utmost importance for the history of typography in general and for the study of the evolution of the printed letter in particular. It was a herculean task and it made the greatest demands on the knowledge, skill, and patience of the experts involved. These were the English scholar Harry Carter, of the Oxford University Press, from

(13) Opposite: Fount scheme for three formes of a roman garamond on colineus (8.6 pica points), supplied by Hendrik van den Keere, 1571 (Arch. 153, folio 125). The type-founder used 261 lb. of type-metal and delivered a total of 101,857 letters.

(14) Opposite: Fount scheme for three formes ofa gothic nonpareille (5.8 pica points), supplied by Hendrik van den Keere in 1569 or 1570 (Arch. 153, folio 271), a total of 204,331 types. The Dutch text, below ‘Somma Sommarum’, reads: ‘which for casting at 8 stuivers per thousand comes to 81 fl. 14 st. And I do not wish to profit from the waste material nor still less to lose from it, seeing that I have taken more care over the work than over my own. Also supplied 23 lb. of lead for the quadrats, coming to 1 fl. 3 st. Also 16 lb. of type-metal for 2 fl. 2 st. And for the baskets, cords, mats, and packing 10 st. Therefore everything together concerning this type comes to the sum of 85 fl. 9 st.’

whom the first impulse for the project came; the American expert Mike Parker; and Dr. H.D.L. Vervliet, then Assistant Curator of the Plantin-Moretus Museum. They were assisted by Matthew Carter, Harry Carter's son, and by K. Melis of the Museum staff.1. The result made all their efforts worth while and it has considerably advanced present knowledge of Renaissance typography.

The Plantinian archives afford few details of the technical aspects of cutting punches and striking matrices,2. except for a little information about the speed - theoretical or actual - at which the work was carried out. In a letter of 16th January 1576, Van den Keere dealt at length with a number of orders which he was then executing for Plantin and discussed the time that they would take.3. In connexion with a very large music type he stated that he would not be able to deliver one punch per day4. - which implies that this would have been the tempo with more normal sizes. Justifying the matrices of a jolie roman alphabet would require three weeks as he could not complete more than five or six matrices a day, and even to maintain this rate he would need to have help.5. For another fount he specified three months for cutting, two weeks for justifying matrices and two weeks for the moulds and the preparation of the steel and copper. This added up to four months, but Van den Keere pointed out that, allowing for other work that was likely to arise, the printer must reckon on half a year. 6. In a letter of 13th February 1579, he estimates

the time needed for justifying the matrices of a philosophie roman at three weeks.1.

The speed at which Granjon cut punches can be gauged from some interesting data. On 3rd February 1565 he contracted to supply Plantin with a parangon Greek. On 9th February he delivered the first fifteen punches, and the whole set was ready on 29th June. Granjon had thus completed 200 punches in a period of 156 days, making, including Sundays and holy days, an average of 1½ punches per day.2.

Naturally there is more information about the financial aspects of all this in the Plantinian archives. It appears that Granjon reckoned his prices on the basis of the punch plus one matrix struck with it. This probably meant that the cost of the copper for the matrices was included in the price. For example, on 3rd July 1565 he agreed to supply Plantin with a garamond italic and a mediane italic at 2 fl. for each punch with one justified matrix.3. On 21st November 1569 the 108 punches of a Syriac aphabet were entered at a rate of 2 fl. 5 st. per punch with matrix.4. Granjon was also expected to deliver the matrices for the parangon Greek ordered on 3rd February 1565, but the provisional price entered in the accounts was for the punches only, at 1 fl. each.5. In the case of the italic sur la grosse ascendonica it was agreed, however, on 18th April 1570 that Plantin should pay 1 fl. 5 st. per punch and should receive four matrices to every punch ‘dont ie luy payeray le cuivre’. There were 43 Flemish pounds of this copper according to an entry of 22nd April. At 4 st. per pound it came to 8 fl. 12 st. altogether.6.

Transactions between Plantin and Granjon usually involved detailed and elaborate contracts. A number of them have been preserved.1. They all contained clauses which gave Plantin a virtual monopoly of the types. In the contract of 3rd July 1565, for the supply of two italic alphabets, Granjon agreed not to cut any identical type for himself or any other party, except on payment to Plantin of the prohibitive sum of 200 escus d'or as compensation. The type-cutter was allowed to keep the matrices ‘pour mon seul et particulier usage et proffict’, although again there were certain conditions: any type that he cast from these matrices must be solely for his own use and he was to keep Plantin informed as to which works he intended printing with them (Granjon was also a printer and bookseller). The contract of 7th December 1566, in which Granjon undertook to cut a garamond civilité for 200 fl., also reserved a strike for the maker ‘laquelle frappe il sobligera de nalliener, vendre, prester ne faire fondre dessus pour personne du monde que pour sen servir soymesmes en sa maison ou demeure et non ailleurs ne par autre qui que ce soit’. The agreement of 3rd February 1565 for the paragon Greek also permitted Granjon to reserve a set of matrices for himself, but the conditions under which he could use them were so restricted that in practice he could do little except store them away. The only way he could make money from them was to sell them to Plantin at the price laid down in the contract.2. Plantin was probably able to dictate these stringent terms because in this case he had advanced Granjon money against delivery.3.

Hendrik van den Keere's letters, and above all his accounts, afford

interesting particulars of the financial side of his dealings with Plantin. In his letter of 16th January 15761. he stated his terms for a set of matrices and punches for music type: 2 fl. 10 st. per punch, 6 st. for each justified matrix; Plantin was to pay for the copper needed. Van den Keere's estimate was 8 to 10 pounds of copper at 10 st. per pound. If Plantin was agreeable to these proposals, then he must buy a 50-pound lump of copper as soon as possible. In the same letter Van den Keere warned Plantin that he would have to pay as much as 3 fl. per punch for another set, in this case for a large music type.2.

Van den Keere calculated his prices on a somewhat different basis from Granjon. Punches and matrices (more precisely: the justifying of the matrices) were priced separately, while the printer had to supply the copper or else pay for it separately. The prices quoted above would suggest that the Ghent master charged more for his products than Granjon. However, they refer to two fairly difficult, or at least extensive commissions and included the cutting of one of Plantin's largest music types. Van den Keere's prices for ordinary alphabets were roughly the same as Granjon's, and in general even slightly lower. For example, in October 1570 he sent in the following bill for a parangon black letter:3.

The account for a canon flamande on 23rd July 1570 was as follows:4.

In the case of the lettre castillane, a type which was later known in the Plantin press as the Spaanse canon (Spanish canon), Van den Keere charged 2 fl. 10 st. per punch and 5 st. for each justified matrix.1. This came to 167 fl. 10 st. for 53 punches and 15 fl. 10 st. for 62 matrices. It had been intended to print the great choir books for Spain in this large, rounded black letter type, but as a result of the events of 1576 the project was abandoned before any printing was put in hand.

Matrices were not always justified by the craftsman who had cut the punches. Jacques Sab on in 1565 received 17 fl. ½ st. for perfecting the 227 matrices of a Granjon paragon Greek, a rate of 1½ st. per matrix.2. In 1569 (?) Herman Gruter sent a bill for 3 fl. 11 st. for justifying 41 matrices of a Hebrew alphabet- 1¾ st. each.3.

Something that does not emerge so clearly from the various prices charged is the relationship between justified matrices and sets of strikes. When Granjon and Van den Keere submitted their accounts they were in fact charging for their actual work on justifying the matrices; for the strike itself they asked virtually no more than the value of the copper block. Taken by themselves, however, strikes were in reality worth a great deal more than the copper they were made of. For those who did not have the corresponding punches at their disposal, a set of strikes was the starting point for type-founding. When strikes were sold separately by a punch-cutter he normally priced them so as to receive a reasonable return for the skill he had exercised in making the original punches; if a third party sold them, the aim was usually to recover part of the purchase price of the punches. Strikes were

therefore sold at much higher prices than were asked for equal amounts of unworked copper. On the other hand it was usual for sets of strikes to depreciate in value in the course of time, either because the novelty or exclusiveness of the typeface had gone, or because punches or matrices had been lost or damaged.

The punch-cutters' and type-founders' accounts do not give so clear a picture of these matters as some of the Plantinian inventories. Unfortunately only three of them record prices and estimates: those of 1561, 1563, and 1566. It is not always possible to extract figures for each individual series: punches and matrices, or different sets of matrices, were often grouped together. The 1563 inventory is also peculiar in that the sums of money quoted there are frequently lower than their equivalents in the 1561 and 1566 lists.1. Apart from this anomaly,2. prices or estimates correspond more or less with the bills submitted by Granjon and Van den Keere. It may be assumed that the 1561 and 1566 figures are a reasonably accurate record of what Plantin actually paid for his punches and matrices, whereas depreciation should possibly be taken into account when interpreting those for 1563. Another interesting point about these inventories is the fact that they enable the prices of strikes and justified matrices to be compared: the former generally figure at only half the price of the justified ones.3.

The 1563 and 1566 inventories record a number of abnormally low prices paid by Plantin, but there is an explanation for these.

They refer to acquisitions which he made in Paris: Guillaume Le Bé stated that in 1562 he had sold some sets of punches and matrices to the printer for no more than a nominal sum; and the material which Plantin bought at the auction of Garamond's estate was probably knocked down to him at less than its full value.1.

These inventories of 1561, 1563, and 1566 enable the growth of Plantin's typographical collection to be expressed in money terms. The 17 sets of justified matrices and the five sets of strikes which were in the printer's possession before his flight to Paris in 1561 were valued at 900 fl. The 1563 inventory gives the position when the printer went into partnership with the Van Bomberghen family. We have, however, already seen that in general the noted prices were much lower than in 1561 and 1566. The sums quoted here add up to 817 fl. 10 st. Prices for some items were not entered - comparison with the 1561 and 1566 inventories suggests that these would have totalled 250 fl. - and no particulars are available for three sets of matrices. The total value must have been in the region of 1,200 fl. - and this is the amount quoted in the deed of October 1563 as the contribution Plantin made to the partnership by bringing in his punches and matrices.2.

The inventory of 1566 puts the total value of Plantin's acquisitions ‘depuis la compagnie faicte’ at 1,358 fl. In fact this amount also included the purchases made in Paris in 1562, before the setting up of the partnership - they had already figured in the 1563 inventory. If these are deducted, Plantin's purchases between 1563 and 1566 work out at around 700 fl. This would make his collection in 1566 worth about 2,258 fl.3.

After 1566 no further totals are available until 1589, when Plantin's estate was divided among his heirs.4. To make the calculations easier

average figures were used in the assessment. Punches were valued at 15 st. each, sets of justified matrices at 30 fl., and sets of strikes at 15 fl. Thus the 3,574 punches were taken to be worth 2,680 fl. 10 st., the 72 sets of matrices 2,160 fl. and the 51 sets of strikes 765 fl., making a total value of 5,605 fl. 10 st.1. This was a considerable sum, but Plantin had in fact paid much more than this for his collection: a reasonable estimate would be 10,000 to 15,000 fl. The assessment at 5,605 fl. 10 st. took depreciation fully into account, but it did not allow for the marked devaluation of the currency which had taken place. This had reduced the purchasing power of the guilder in 1589 to half of what it had been in 1561-66. If the purchases made between 1561 and 1576 are expressed in terms of 1589 values, the cost to Plantin of building up his typographical collection could be put as high as 20,000 fl.

Plantin's punches and matrices represented a large investment of capital, but as this was spread over the thirty-four years of his printing operations it did not bear down too heavily on the business. In 1566, for example, he spent 103 fl. 12 st. on these items, compared with a total expenditure of 13,041 fl.:2. barely 0.8 per cent and less than he had paid for ink in that year.3. In some years the percentage might be much higher; on the other hand there were years in which Plantin made no additions to his collection. Although punches and matrices were by no means cheap, the amount of money spent on them was small in comparison with what Plantin had to pay to have his type cast. What should be stressed is the fact that this investment in equipment was not strictly necessary, at all events not on this scale. Plantin could have obtained all the cast type he needed from a typefounder, or from a punch-cutter who practised both branches of the craft. It might have been slightly more expensive to buy type in this

way, but the difference in cost between casting type with his own matrices and having it done with someone else's material certainly did not warrant the amount of money invested in punches and matrices. The extensive purchase of these materials was undertaken largely to secure a monopoly of certain type faces for the Plantin press.

However, the collection was an asset and not just so much dead capital. It served as collateral in certain transactions. It has already been seen that when Plantin went into partnership with the Van Bomberghens in 1563, his punches and matrices gave him a share in the business worth 1,200 fl. When the printer died in 1589 he had 1,736 punches, 6 sets of matrices, and 14 sets of strikes stored in the warehouse he had rented in Frankfurt. He had offered these - or a proportion of them - as security for a loan of 6,000 fl in 1585. The money was to be paid back with 4 per cent interest before 1590.1. In this way he was able to obtain money that helped him keep his officina going in the catastrophic years of 1585-89.

Plantin was also able to sell punches and matrices from his collection and thus recover some of the money he had invested in it - plus a little profit. At Frankfurt, in 1579, he sold strikes of Granjon's gaillarde (bourgeois) roman and granjonne italic to the successors of Dietrich Gerlach of Nuremberg.2. The inventory of c. 1572 begins with an enumeration of 15 frappes creues by Granjon and Garamond, followed by the brief remark ‘Tout cecy est envoyé à Francfort’.3. As no further trace of these has ever been found it must be assumed that they never returned from Frankfurt.

The fact that Plantin sometimes sold items from his typographical collection no doubt explains why from time to time sets of punches or matrices would cease to appear in the inventories. On the whole this does not seem to have happened very often: sales were not very extensive and were largely confined to strikes, for which Plantin

retained the punches so that he could have new matrices made if necessary. Sometimes he himself would buy a set of strikes specially for sale elsewhere. One example was the Haul tin mediane Greek which he ordered through Pierre Porret in July 1567.1.

‘Elsewhere’ in this case was Frankfurt once again: Plantin's sales of typographical material were largely limited to the fairs held in that city. The reason for this is not far to seek. In Paris, Plantin had little chance of selling at a worthwhile profit matrices for type faces that French printers could obtain locally. He did not want to sell them in Antwerp or elsewhere in the Netherlands as this would have lost him the exclusiveness of his types. In Germany, on the other hand, these French Renaissance alphabets were less widespread and were in some demand. Their use by German printers could do Plantin little serious harm. Plantin also sent a few sets of matrices to Spain - not on his own initiative in this case, but as part of transactions that were more or less forced on him by the Spanish authorities.2.

Finally in this section, some attention must be given to the question of how it was that, from 1580 onwards at least, so many sets of punches and matrices came to be stored in Plantin's warehouse in Frankfurt. Did the printer consider selling a large part of his collection

in the years of crisis after the Spanish Fury of November 1576? This idea may well have occurred to him, but Plantin was more likely to have been thinking of the safety of his collection than of its possible conversion into ready cash. But the fact that the material was in Frankfurt may have given him the idea - and the opportunity - of using it in 1585 as security against a cash loan: a course of action that did not require him to give up his valuable collection.

Cast type1.

The result of the type-cutter's work was a matrix from which the type needed for printing could be cast. The casting was done in a mould. The chapter entitled ‘L'Imprimerie’ in the Dialogues françois et flamands of 1567 contains a description of the mould and how it was used. The exposition, however, is not as clear as it might be:

E: Into this matrix the type-metal, such as lead or tin, of which they wish to make the type, is poured, in a mould.

G: I understand what you mean. However, it seems to me very difficult to make letters in that way so expertly proportioned that they all fit together exactly.

E: That is done by means of the mould, which is made of several pieces fastened together, by which all the types are made alike, being as they say of the same fount.

G: The mould may thus take the matrix of an A as readily as that of a B and so on, and the A and B are therefore proportioned alike?

E: That is right.

G: It is made of several pieces, you say?

E: Yes, necessarily so, for otherwise the type would not be able to have the things necessary to it. First the mould is mounted on a block (the wood) against which there is a little bow which lifts. Then there is a plate, the long pieces (carriage) and the wire (nick) fastened to the long pieces. There are the bodies, the gauges, the jets, the registers, the gallows, and the stool, essential to the complete mould.

G: This, then, is how the founts are cast inside the moulds, to which the matrices are attached...

The mould1. consisted of two halves that fitted together. When pushed together a small aperture was left in the centre. The matrix was pressed against the underside of this slot by means of the ‘bow’ or clamp. Molten metal was poured in through the top of the slot with a spoon until the aperture was filled. When the mould was opened, a casting appeared with the letter on one end. This letter was not quite ready for use: the jet had to be broken off and the foot of the shank planed smooth.

The castings made in a particular mould had all the same body and length, though they varied in width from i to m, as the lead had been poured in the same aperture. On the other hand, each metal type size (body) had to have its own mould. Sets of matrices were often supplied with the appropriate mould. They were carefully noted in the inventories of typographical material. That of c. 1572, for example, mentions twenty-three ‘instruments’ and specifies the bodies for which they were intended. The number of these had risen to thirty in the 1612 inventory and to thirty-three in that of 1652.2. Documents relating to the division of Plantin's estate in 1589-90 mention 36 moulds; as there is no further specification, it is possible that the figure is not accurate.3.

It is clear from these figures that the increase in the number of moulds did not run parallel with the growth of Plantin's collection of punches and matrices. It was, in fact, not altogether practical to use a separate mould for each type face. Attempts at rationalization were made quite early, with the idea of producing moulds which, with small adjustments, could serve for the various kinds of type (roman, italic, black letter, etc.) on a particular body and even for

more than one body. Plantin himself laid some stress on this in a letter to de Çayas,1. while certain details in Van den Keere's correspondence2. and in the inventories3. also point to these developments. Nevertheless, the 33 moulds mentioned in the 1652 inventory had to be supplemented at a later date. The Plantin-Moretus Museum now possesses 62 old moulds which were handed over with the rest of the typographical collection in 1876.4.

The width and depth of the shanks of cast type varied from body size to body size. All shanks, however, had to be of exactly the same height to paper in order to form a level surface for inking and printing: any piece of type projecting above the others would have gone through the paper, and any piece of type that was too short would not have printed.

This was the state of affairs in the sixteenth century and for a large part of the seventeenth. At the end of the seventeenth century, however, there was a slight change when a new technique was evolved for printing liturgical books in black and red.1. Type of normal height to paper was used for the parts of the text to be printed black, taller type for the parts to be printed red. The red was printed first, then the type used for this was removed and replaced by quadrats and spaces and the sheets put back in the press for the black to be printed. Compared with the method previously used, this was quicker and more accurate, although it meant that the printer had to have letters of two different heights. The introduction of this technique into the Plantin press can be dated fairly accurately. In the accounts for 31st January 1682 of the Van Wolsschaten family, the type-founders who supplied the officina from 1660 onwards, there appeared the first specific mention of a delivery of ‘high’ type (116 pounds of a bible roman).2. Thereafter, deliveries of this type alternated with ordinary type, for which no special term was used, until in 1701 the latter acquired the name ‘low letter’ by which it was subsequently known in the house.3. It is possible, however, that the technique had been applied there even earlier than this, and that founders working in the officina4. had cast type of this kind some years before 1660.5.

Moulds were valued at 5 to 6 fl. in the inventories;6. the thirty listed in 1612 were thus worth from 150 to 180 fl.7. As well as being

not too expensive they lasted for quite a considerable time before they needed to be replaced.

The same was not true of cast metal type. This had to be replaced or replenished fairly often, as it wore out, buckled, or broke. It may have been this factor which led Plantin to invest such large sums of money in comparatively unproductive punches and matrices. In this way he was at least sure of always having the type he liked when he needed it.

Even if the Plantinian archives give comparatively few technical details about punches and matrices, they do contain a wealth of such information about type-founding. The composition of the estoffe,1. the type-metal from which the type was cast, forms a good starting point. An analysis of a Plantinian lead type of about 1581 shows it contained 82.71 per cent lead, 8.99 per cent tin, 5.81 per cent antimony, 0.53 per cent copper and traces of iron.2. A text dated 1565 actually states how much of each ingredient Plantin needed to produce a given quantity of type-metal.3. It is detailed on p. 96.

The 773 pounds mentioned produced only 517 pounds of usable type-metal,4. which meant that about one third was lost or expelled during smelting.

The basic ingredients here are the same as in the type that was

analysed, but the proportions are different.3. To judge from eighteenth-century accounts - of which more below - it is mainly the antimony, with the copper and iron, that was expelled during smelting, and the proportion of lead and tin was correspondingly increased.

(15) Opposite: fount scheme for a ‘french type’ (civilité) augustine (13.4 pica points) supplied by Hendrik van den Keere, 1580 (arch. 153, f^o 281). The founder delivered a total of 39,100 types, weighing, with quadrats and packing, 191 lb. The Dutch text below ‘Somma sommarum’ reads: ‘Which costs for casting at 10 st. per thousand the sum of 19 fl. 11 st. for casting 19½ lb. of quadrats at the same [rate] together with the basket, waste, and cords, and taking it to the ship 4 fl. Somma sommarum 23 fl. 11 st. Shipped the 10th December, weighing 191 lbs.’

(16) Opposite: fount scheme for a Greek type, paragon (18.7 pica points), supplied by Van Everbroeck, 1565 (arch. 153, f^o 47). The Dutch text below reads: ‘This type, metal only, without baskets, weighs 264 lb. Each pound costs 3 st. coming to 39 fl. 12 st.’ In a different handwriting, probably Jan Moretus's, is added in french: ‘Item 29 lb. de mouleurs [i.e., quadrats] à 3 st. la lb.’

Iron is missing in the instance detailed here, although it is clear from other accounts of the same period that Plantin regularly bought iron filings, old nails, and similar items for inclusion in his type-metal.1.

The percentages given above provide a general idea of the composition of Plantin's type-metal, but they are by no means absolute, or hard and fast. Purely empirical methods were applied and the relative proportions could vary greatly, partly in accordance with the body size of the cast type.2. Lead was always the basis of metal type in Plantin's time and to this were added appreciable quantities of tin, to make the metal more fluid in the molten state, and antimony to make it harder, together with a little iron and copper. as the intensity with which Plantin's stocks of type were used increased, so did the frequency with which they wore out or became damaged. It became customary to obtain some of the metal needed for casting new type by melting down worn or damaged type: from 1565-66 onwards this was regular practice.3.

No particulars have been found of the composition of type-metal in the seventeenth century, but there are a number of detailed accounts from the eighteenth century. In about 1760 franciscus Joannes Moretus drew up a ‘memorandum concerning the cost of making hard metal for casting type’ which gives examples taken from practice.4. Some are detailed on p. 98.

The amount of waste material - the lycagie as it was termed - was remarkably large when iron scraps were added to the molten metal. It is clear from the subsequent notes on type-founding which F.J.

Moretus wrote for his own edification and instruction that those engaged in the process were aware that most of this iron, and a considerable proportion of the antimony was lost during smelting.1. Notable too is the emphasis which was placed on the use of old type: ‘Observe that for good results old type should comprise half the mixture.’2. This remark follows the ‘memorandum given by Jan Michiel Smit’ on ‘the proportions that should be adhered to in preparing metal for casting type’, which also prescribes a goodly percentage of melted down old type. Smit's proportions were:

This emphasis on the use of old type must be seen in the light of

another factor: the absence of tin and copper as compared with 1566. It is certainly no coincidence that the missing metals were the two most costly ingredients. In 1565 Plantin was paying 5 fl. 10 st. per 100 lb. of antimony, and 3 fl. 18 st. for the same amount of lead, whereas the prices for these quantities of tin and copper were 15 fl. 10 st. and 17 fl. 10 st. respectively. This very likely would have been the main reason for the disappearance of tin and copper from type-metal. However, to obtain good metal, a certain percentage of tin was still needed - and this could be got from old type prepared to the standards applied in Plantin's time. as this process was repeated over and over again, the percentage fell until even the ‘old’ type going into the crucibles contained hardly any tin. Given the empirical methods of the time, however, it may be assumed that the type-founders went on regarding old type as essential for the improvement of their type-metal - which to some extent at least it was.

It was not only in the Plantin press or among antwerp type-founders that copper and tin ceased to be used. The tendency was general to Western Europe, as appears from an analysis of a number of old cast types.1. Whereas Plantinian type of around 1580 contained 8.99 per cent tin and 0.53 per cent copper, the percentages had dropped in 1647 (in London) to 2.9 and 0.15 per cent. The other examples for the seventeenth and eighteenth centuries contain hardly any traces of copper and only negligible amounts of tin (with a few exceptions, probably due to the presence of old type in the metal).2. The percentage of antimony rose proportionately. It was only 5.81 per cent in the Plantinian type, fluctuated around 7 to 8 per cent until 1700 and exceeded 14 per cent in subsequent years. Tin did not

figure again in appreciable quantities until the end of the eighteenth century. The proportion of antimony was maintained, which meant that again the percentage of lead in type-metal was considerably reduced.1.

Mention has already been made of the amount of lycagie (dross), the loss of material during the preparation of this type-metal. But not all metal produced could be turned into cast type: during this process too quite a lot was lost through oxidization, spilling, and from shavings. In 1575 Plantin must have expressed his surprise to Hendrik van den Keere at receiving so little type in comparison with the amount of material he had supplied, for in connexion with a delivery the type-founder noted tartly that ‘tout le laccage monte à 10 % sur toutes les sortes communes. Et à advenant de 20 % sur la petite nonpareille, 16% sur la jolye et coronelle, 12% sur la Byble’. This meant that only 2,713½ pounds of cast type had been made from 3,027¼ pounds of metal, a wastage of 313¾ pounds.2. Plantin must in fact have known this: figures that he recorded in 1563-64 give roughly the same proportions.3.

F.J. Moretus's memorandum gives a few interesting details of the technique of preparing type-metal.4. When the ingredients had been sufficiently mixed and smelted in the pot or crucible, a hole was bored in the vessel to allow the metal to run out. F.J. Moretus compares this with the method then in vogue in Holland, but without coming to any conclusion about the relative merits of the two techniques.

In the north the crucibles were after the manner of the silversmiths, who allowed the molten metal to become cool and then broke the vessels in order to get the metal out. In both methods the melting pots were destroyed. This must have happened in Plantin's time too; a note of 18th May 1565 gives the additional expenses incurred in the preparation of a quantity of type-metal, including ‘creusets, pièces: 10’, valued at 3 fl. 2 st.1.

Even a small text presupposes a large amount of cast type, and there had to be a sufficient quantity of each typeface and body available to meet the needs of the moment. The type was cleaned and distributed (‘dissed’) again after use and thus became available once more - but while the text was being printed, a quantity of type was obviously immobilized. The only way to feed an increasing number of presses with type was to add to the stock. In a letter of 1st august 1572 Plantin explained that, in order to speed up the printing of a work by Luis of Granada he had increased the number of presses engaged from two to four and had started to cast new stocks of type.2.

However, storage space and above all the cost of manufacture set a limit to increases of typographical stocks. In some instances, rather than immobilize quantities of type, or increase his stock of it, Plantin found it more expedient to print a limited number of copies. In this way the type was quickly available again. He would then have the text re-set later and make up the number of copies. He even resorted to this policy for the great Polyglot Bible.3.

The compositors worked by formes.4. The quantities of type to be

supplied were therefore calculated and expressed in this measure. The type-founders delivered the amounts of type needed to set as many formes as were required.1. But in these formes certain letters occurred more frequently than others: a text required a far larger quantity of letter ‘e’ than of ‘x’ or ‘z’. Nowadays type is supplied with a fixed prescribed proportion of each letter: this is known as the fount-scheme (french: police; Dutch: polis; German: Giesszettel).2. The principle was already known and applied in Plantin's time. The archives of the firm contain quite a number of fount-schemes as calculated in the sixteenth, seventeenth, and eighteenth centuries.3. It sometimes happened that a complete fount-scheme was not needed, only certain sorts; in the Plantinian accounts these deliveries were entered as defects or imperfections.4.

The number of letters that could be cast with a particular quantity of type-metal naturally varied according to typeface and body. These ratios are also specified many times in the archives. The entries given opposite date from 1563-65.5.

F.J. Moretus, in his memorandum of about 1760, expressed the relationships differently, giving the weight of type-metal per 1,000 pieces of type, as shown in the table on p. 104.6.

There is some information available about the time that the type-founders took to complete particular orders. In a letter of 13th february 1579, Hendrik van den Keere stated that Plantin could have

three formes of a philosophie roman within three to four weeks.1. This involved 340 pounds of type-metal, representing a little over 100,000 letters. It is not known, however, how many men the Ghent type-founder had working for him at that time. In his memorandum,

Weight of Type-metal per 1,000 Pieces