Typography


From Encyclopedia Britannica (11th edition, 1910)

EncycloReader

Typography (i.e. writing by types) is the general term for the art of printing movable (cast-metal) types on paper, vellum, &c. It is distinct from writing, and also from wood-engraving or xylography, which is the art of cutting figures, letters, words, &c., on blocks of wood and taking impressions from such blocks by means of ink, or any other fluid coloured substance, on paper or vellum.

I

History Of Typography Although the art of writing and that of block-printing both differ widely from printing with movable metal types, yet this last process has apparently been such a gradual transition from block-printing,' and block-printing in its turn such a natural outcome of the many trials that were probably made to produce pictures, books, &c., in some more expeditious manner than could be done with handwriting, that a cursory glance at these two processes will not seem out of place, especially as a discussion on the origin and progress of typography could hardly be understood without knowing the state of the literary development at the time that printing appeared.

The art of printing, i.e. of impressing (by means of certain forms and colours) figures, pictures, letters, words, lines, whole pages, &c., on other objects, as also the art of engraving, which is inseparably connected with printing, existed long before the 15th century. Not to go back to remoter essays, there is reason to suppose that medieval kings and princes (among others William '' We do not deal here with copperplate engraving (chalcography), nor with the question, raised by some authors, whether this art preceded that of wood-engraving (xylography), or vice versa. The earliest known date of the former is 1446 on the small engravings of " the Passion " in the Berlin Royal Print Room, whereas the earliest known date of wood-engraving is 1418 (on the Brussels Mary engraving). Both arts were naturally dependent upon MSS. for the forms of their letters, but as to the question of transition from the art of writing to that of typography, xylography alone can be regarded as the intervening and connecting link between those two arts, and there are good reasons for assuming that the inventor of printing with movable types was a xylographer (see below).

the Conqueror) had their monograms cut on blocks of wood or metal in order to impress them on their charters. Such impressions from stamps are found instead of seals on charters of the i 5th century. Manuscripts, even of the 12th century, show initials which, on account of their uniformity, are believed to have been impressed by means of stamps or dies.' Before the invention of printing, say about 1436, bookbinders are known to have impressed names or legends or other inscriptions on their bindings in two ways: (1) by means of single, insulated letters engraved reversely downwards into a stamp of brass, whereby the letters appeared en relief on the leather or parchment of the binding; (2) by letters engraved reversely en relief on the brass stamp, whereby the letters sank into the binding. For this reason the term impressor, applied afterwards to the " printer," was, in the first instance, applied to the binder, whereas ligator was the proper word for him (see F. Falk, Der Stempeldruck, in " Festschrift," 1900, p. 73 sqq.; Zedler, Gutenberg-Forschungen, 1901, p. 6). But the idea of " multiplying " representations from one engraved plate or block or stamp, or other form, was unknown to the ancients, whereas it is predominant in what we call the art of blockprinting, and especially in that of typography, in which the same types can be used again and again.

Block-printing and printing with movable types seem to have been practised in China and Japan long before they were known in asia t ic Europe. It is said that in the year 175 the text of East . the Chinese classics was cut upon tablets, and that impressions were taken of them, some of which are supposed to be still in existence. Printing from wooden blocks can be traced as far back as the 6th century, when the founder of the Suy dynasty is said to have had the remains of the classical books engraved on wood, though it was not until the 10th century that printed books became common. In Japan the earliest example of block-printing dates from the period 764-770, when the empress Shiyau-toku, in pursuance of a vow, had a million small wooden toy pagodas made for distribution among the Buddhist temples and monasteries, each of which was to contain a dharani out of the Buddhist Scriptures, entitled " Vimala nirbhasa Sara," printed on a slip of paper about 18 in. in length and 2 in. in width, which was rolled up and deposited in the body of the pagoda under the spire. In a journal of the period, under the year 987, the expression " printed book "(suri-hon)is applied to a copy of the Buddhist canon brought back from China by a Buddhist priest. This must have been a Chinese edition; but the use of the term implies that printed books were already known in Japan. It is said that the Chinese printed with movable types (of clay) from the middle of the i ith century. The authorities of the British Museum exhibit as the earliest instance of Korean books printed with movable types a work printed in 1337. To the Koreans is attributed the invention of copper types in the beginning of the 15th century; and an inspection of books bearing dates of that period seems to show that they used such types, even if they did not invent them.2 From such evidence as we have, it would seem that Europe is not indebted to the Chinese or Japanese for the art of blockprinting, nor for that of printing with movable types.

In Europe, as late as the second half of the 14th century, every book and every public and private document was written by hand; all figures and pictures, even MS. playing cards and images of saints, were drawn with the pen or painted with a brush. In the r3th century there already existed a kind of book trade. The organization of universities as well as that of large ecclesiastical establishments was at that time incomplete, especially in Italy, France and Germany, without a staff of scribes and transcribers (scriptores), illuminators, lenders, sellers and custodians of books (stationarii librorum, librarii), and pergamenarii, i.e. persons who prepared and sold the vellum or parchment required for books and documents. The books supplied were for the most part theological, legal and educational, and are calculated to have amounted to above one hundred different works. As no book or document was approved unless it had some ornamented and illuminated J 'Passavant, Le Peintre-Graveur, i. 18 (Leipzig, 1860-1864); John ackson, Wood Engraving (London, 1839); Bruno Bucher, Gesch. der lechn. Kunste, I. p. 362 seq.

2 See Ern. Satow, " On the Early History of Printing in Japan," in Trans. Asiat. Soc. of Japan, x. 48 seq.; and Stan. Julien, " Documents sur l'art d'imprimer," &c., in Journ. Asiat., 4me ser., vol. ix. p. 505.

initials or capital letters, there was no want of illuminators. The workmen scribes and transcribers were, perhaps without exception, calligraphers, and the illuminators for the most part artists. Beautifully written and richly illuminated manuscripts on vellum became objects of luxury which were treasured by princes and people of distinction. Burgundy of the r5th century, with its rich literature, its wealthy towns, its love for art and its school of painting, was in this respect the centre of Europe, and the libraries of its dukes at Brussels, Bruges, Antwerp, Ghent, &c., contained more than three thousand beautifully illuminated MSS.

In speaking of the writing of the manuscripts of the 15th and preceding centuries it is essential to distinguish C '/ass f ' in each country between at least four different wr;tvg. classes of writing, two of which must be again subdivided into two classes.

I. The book hand, that is, the ordinary writing of theological, legal and devotional books, used by the official transcribers of the universities and churches, who had received a more or less learned education, and consequently wrote or transcribed books with a certain pretence of understanding them and of being able to write with greater rapidity than the ordinary calligrapher. Hence they produced two kinds of writing: (a) the current or cursive book hand, of which several illustrations are given in Wilh. Schum, Exempla Codicum Amplon. Erfurtensium; the volumes of the (London) Palaeogr. Society, &c. Quite distinct from this current writing, and much clearer and more distinct, is (b) the upright or set book hand, which was employed not only by writers who worked for universities and churches, but also by persons who may be presumed to have worked in large cities and commercial towns for schools and the people in general without university connexion. (2) In the church hand (Gothic or black letter) were produced transcripts of the Bible, missals, psalters and other works intended for use in churches and private places of worship and devotion. This writing we may again subdivide into two classes: (a) the ornamental or calligraphic writing, found exclusively in books intended for use in churches or for the private use of wealthy and distinguished persons, and (b) the ordinary upright or set church hand, employed for less ornamental and less expensive books. (3) The letter hand may be said to be intermediate between the set literary book hand and the set literary church hand, and to differ but little from either. It was employed in all public documents of the nature of a letter. (4) The court or charter hand was used for charters, title-deeds, papal bulls, &c.3 These different kinds of writing served again, in the first instance, as models for cutting the inscriptions and explanatory texts that were intended to illustrate and explain the figures in blockbooks, and afterwards as models for the types used in the printing of books and documents.

Dypold Laber (Lauber), a teacher and transcriber at Hagenau in Germany, is known to have carried on a busy trade in manuscripts about the time of the invention of printing. His prospectuses 4 in handwriting of the middle of the 15th Century century announce that whatever books people wish C to have, large or small, " geistlich oder weltlich, Books, hilbsch gemolt," are all to be found at Dypold Lauber's the scribe. He had in stock Gesta Romanorum, mit den Viguren gemolt; poetical works (Parcival, Tristan, Freidank); romances of chivalry (Der Witfarn Ritter; Von eime Getruwen Ritter der sin eigen Hertze gab umb einer schônen Frowen willen; Der Ritter unter dem Zuber); biblical and legendary works (A Rimed Bible; A Psalter, Latin and German; Episteln and Evangelien durch das Jor; Vita Christy; Das gantze Passional, winterteil and summerteil; devotional books (Bellial; Der `Selen Trost; Der Rosenkrantz; Die zehn Gebot mit Glosen; Small Bette-Bucher); and books for the people (Gute bewehrte Artznien-Bucher; Gemolte Loss-Bucher, i.e. fortune-telling books; Schachtzabel gemolt). The lower educational books consisted for the most part of the Abecedaria, containing the alphabet, the Lord's Prayer, the creed, and one or two prayers; the Donatus, a short Latin grammar extracted from the work of Aelius Donatus, a Roman grammarian of the 4th century, and distinctly mentioned in a school ordinance of Bautzen of 1418; the Doctrinale, a Latin grammar in leonine verse, compiled by Alexander Gallus (or De Villa Dei), a minorite of Brittany of the 13th century; the Summula logica of Petrus Hispanus (afterwards Pope John XXI.), used in the teaching of logic and dialectics; and Dionysius Cato's Disticha de Moribus, and its supplement called Facetus, with the Floretus of St Bernard, used in the teaching of morals. As helps to the clergy in educating the lower classes, and as a means of assisting and promoting private devotion, there were picture books accompanied with an easy explanatory text, for the most part representations of the mystic relation 3 See further Palaeography.

4 An original copy of one of them is in the British Museum (Addit. MS. 28752).

between the Old and New Testaments (typology). Among these books the Biblia pauperum 1 stands first. It represents pictorially the life and passion of Christ, and there exist MSS. of it as early as the 13th century, in some cases beautifully illuminated.' A richly illuminated MS. of it, executed in the Netherlands c. 1400, is in the British Museum (press-mark, King's 5), and also fragments of one of the ,4th century (press-mark, 31,303). A remodelling and development of this work is the famous Speculum humanae salvationis, of which we shall speak when dealing with the blockbooks and early printed books. It was written in rhymed prose before 1324, and represents, in forty-five chapters, the Bible history of the fall and redemption of mankind interwoven with Mariolatry and legend. Of this work alone more than 200 MSS., illuminated or without pictures, are known to exist in various libraries of Europe. The National and Arsenal Libraries in Paris each possess one written some time after 1324; the British Museum has sixteen MSS. of it (eleven of which are illuminated) of the 14th and 15th centuries, written in the Netherlands, Germany, France and England, one (press-mark, 16,578) bearing the distinct date 1379 and another (press-mark, Egerton, 878) that of 1436. A work of a similar nature is the Apocalypsis, of which at least two recensions with illustrations may be pointed out. One gives the text as we know it, with or without commentary, for which cf. Brit. Mus. 17,333 (French), 18,633 (French, but written in England), Reg. 2 D. xiii. and 22,493 (French) - all four early 14th century. Another is more a short history or biography of St John, but the illustrations follow those of the former work very closely; cf. Brit. Mus. 19,896 (15th century, German). It is this last recension which agrees with the blockbook to be mentioned hereafter. Other devotional works are the Ars Moriendi, the Antichrist and other works which will be mentioned below among the blockbooks.

-

When all this writing, transcribing, illustrating, &c., had reached their period of greatest development, the art of printing from wooden blocks (block-printing, xylography) on silk, cloth, vellum, paper, &c., made its appearance in Europe. This art was already a great advance on writing, in that it enabled any one with a few simple tools to multiply impressions from any block of wood with text or pictures engraved on it, and so produce a number of single (paper) leaves or sheets with text or pictures printed on them in almost the same time that a scribe produced a single copy of them.

It seems to have been practised, so far as we have evidence, on cloth, vellum and other stuffs as early as the 12th century (Weigel, Anfange, i. so); and on paper as far back as the second half of the 14th century; while it began to be largely employed in the early part of the 55th all over Germany, Flanders and Holland in the production of (I) separate leaves (called briefs, from breve, scriptum), containing either a picture (print, Arent, shortened from the Fr. em print, empreinte, and already used by Chaucer, C.T. 6186, six-text, D. 604, printe, prente, preente, and in other early English documents; also called in colloquial German Helge, Helglein, or Halge), or a piece of text, or both together; and of (2) whole sheets (two leaves), a number of which, arranged like the MSS. in quires or gatherings, formed what are called " blockbooks," sometimes consisting of half picture and half text, or wholly of text, or altogether of picture.

The earliest dated woodcut that we know of is the Mary engraving, discovered at Malines, and now preserved in the Brussels Royal Library. It bears the date mccccxviii. Some authors Early dated have asserted that an 1 has been scratched out between Wood= the fourth c and the x; that, therefore, the date is 1468. Engravings. But there is no ground for such an assertion (cf. H. Hymans, L'Estampe de 1481, Brussels, 1903). A slightly modified reproduction of it, on a reduced scale, which could hardly be placed later than 5460, is preserved in the St Gall Library. The next date is 1423 found on the St Christopher, preserved in the John Rylands Library (Spencer collection) at Manchester. In the third place comes the woodcut of 1437 preserved in the Imperial Library at Vienna, which was discovered in 1779 in the monastery of St Blaise in the Black Forest, and represents the martyrdom of St Sebastian, with fourteen lines of text. The date, however, is said by some to refer to a concession of indulgences. A woodcut, preserved in the same library in Vienna, which represents St Nicolas de Tolentino, has the date 1 This title is applied to at least three works: (I) the wellknown blockbook, of which we speak below, (2) a treatise " in qua de vitiis et virtutibus agitur," and (3) a work in rhyme by Alexander Gallus.

2 See Laib and Schwarz, Biblia pauperum (Zurich, 1867).

1440, but written in by hand; as the saint was canonized in that year it may refer to that event. Another in the Weigel collection, representing the bearing of the cross, St Dorothea and St Alexis, has the date 1443, also written in by hand, though the woodcut is considered to belong to that period. These are the only known woodengravings with dates ranging from 1418 to 5443. But there exist a good many woodcuts which, from the style of the engraving, are presumed to be of an earlier date, and to have been printed partly in the 14th and partly in the first half of the 15th century. J. D. Passavant (Le Peintre-Graveur, 1860-1864, i. 27 seq.) enumerates twenty-seven of them, all of German origin and preserved in various libraries in Germany; 154 are recorded in the Collectio Weigeliana (vol. i., 1866), and W. L. Schreiber (La Gravure sur bois, vols. i. and ii., 1891 and 1892) enumerates over 2000 of them, some of which may be ascribed to the Netherlands, exx.g. (I) representing the Virgin Mary, with Flemish inscriptions in the museum in Berlin; (2) representing the Virgin Mary (see above) in the library at Brussels; (3) representing St Anthony and St Sebastian, in the Weigel collection (now in the Brit. Mus.); (4) a St Hubert and St Eustatius, in the royal library at Brussels; (5) representing the Child Jesus, in the library at Berlin; (6) the Mass of St Gregory, with indulgence, in the Weigel collection (cf. I, 195), now at Nuremberg.

In these blocks, as in wood-engraving now, the lines to be printed were in relief. The block, after the picture or the text had been engraved upon it, was first thoroughly wetted with a thin, watery, pale brown material, much resembling distemper; then a sheet of damp paper was laid upon it, and the back of the paper was carefully rubbed with some kind of dabber or burnisher, usually called a frotton, till an impression from the ridges of the carved block had been transferred to the paper. In this fashion a leaf or sheet could only be printed on one side (anopisthographic); and in some copies of blockbooks we find the sides of the leaves on which there is no printing pasted together, so as to give the work the appearance of an ordinary book. Any one wanting to set up as a printer of briefs or books needed no apparatus but a set of woodblocks and a rubber. We know only three blockbooks which do not possess this characteristic, as the Legend of St Servatius in the royal library of Brussels, which may be called a xylo-chirograph (see below), in which the pictures occur on both sides of the paper (with some lines of text written underneath), but apparently impressed by hand from blocks without any rubbing, there being no traces of any indentures either on the rectos or the versos; Das Zeitglocklein in the Bamberg Library (cf. Falkenstein, P. 49); Das geistlich and weltlich Rom, in the John Rylands Library (Spencer collection) and at Gotha (cf. Falkenstein, p. 46); but these belong to the end of the 15th century, and therefore to a later period than the ordinary blockbooks.

Formerly it was the general opinion that playing cards had been the first products of xylography; but the earliest that have been preserved are done by hand, while the printed cards date from the 15th century, therefore from a Block period in which woodcuts were already used for other Printers. purposes. Some of the wood engravings and blockbooks are supposed to have been printed in monasteries. In a necrology of the Franciscan monastery at NOrdlingen, which comes down to the beginning of the 15th century, this entry occurs: " VII. Id. Augusti, obiit Frater h. Luger, laycus, optimus incisor lignorum "; and on some of the engravings we find the arms of certain monasteries, which may, however, merely mean that they were printed for, not in, those monasteries. The registers of Ulm mention several woodengravers (formschneider) - in 1398 a certain Ulrich; in 1441 Heinrich Peter von Erolzheim, Joerg, and another Heinrich; in 1442 Ulrich and Lienhart; in 1447 Claus (Nicolas), Stoffel (Christopher) and Johann; in 5455 Wilhelm; in 1461 Meister Ulrich, &c. In a register of taxes of NOrdlingen we find from 1428 to 1452 a certain Wilhelm Kegeler mentioned as brieftriicker; in 1453 his widow is called alt brieftriickerin; and in 1461 his brother Wilhelm is registered for the same craft. At Mainz there was a printer, Henne Cruse, in 1440. At Nuremberg we find in 1449 Hans (Spoerer?), a formschneider, while his son Junghans exercised the same industry from 1472 to 1490. Hans von Pfedersheim printed at Frankfort in 1 459; Lienhart Wolff, priefdrucker, is mentioned in the registers of Regensburg of 1463; Peter Schott at Strassburg in 1464. A certain George Glockendon exercised the same trade at Nuremberg till 1474, when he died and was succeeded by a son and afterwards by a grandson. In Flanders a Jan de Printere was established at Antwerp in 1417; and printers and wood engravers (haute bildsnyters) worked there in 1442 (Privileges of the Corporation of St Luke at Antwerp). At Bruges printers and beeldemakers (makers or engravers of images) were enumerated in 1454 among the members of the fraternity of St John the Evangelist. The printers of playing cards seem to have constituted a separate class.

All these entries show that long before the middle of the 15th century there were men who exercised the art of wood-engraving and printing as a trade or craft. It seems also certain that wealthy persons and religious institutions were wont to possess sets of blocks, and, when occasion arose, printed a set of sheets for presentation to a friend, or in the case of monasteries for sale to the passing pilgrim. A printer of briefs or blockbooks had no need to serve an apprenticeship; printed from wooden blocks; in others the text was chi chirography. written first, and woodcuts pasted or printed in spaces reserved for them. These books, combining woodengraving with handwriting, are now in technical language called xylo-chirographs (wood-handwritten books); they may also be called semi-blockbooks, and form an intervening stage between the manuscript book and the blockbook (xylograph) entirely printed from wooden blocks. They tend to show that xylography, after having been for some time confined to the production and multiplication of insulated pictures, was gradually applied to the printing of whole series of illustrations, to be added to written texts, or to have written texts added to them. It is not possible to assign definite dates to these xylo-chirographs; they could hardly be placed after, but may, for ought we know, be contemporaries of the blockbooks. We know nine of them; the years 1440 (which occurs in No. 5) and 1463 (found in No. 9) marking, for the present, the period within which they can be placed.

(1) Biblia Pauperum, in the Heidelberg University Library, German work, MS., Latin text added to engravings (cf. Schreiber, Manuel, iv. 90, c. 1460; photogr. pl. xlv.); (2) Anti-christus, one part of which is in the Paris Bibl. St Gen. (see Bernard, Orig. de l'impr. i. 102), another at Vienna, Alb. Bibl.; Bavarian work, MS., German text added to engravings (Schreiber iv. 231, pl. Iv.); (3) Vita et Passio Jesu Christi, 48 leaves, in the Vienna Hofbibliothek, German work, the woodcuts printed on the versos, Latin prayers written on the rectos (Schreiber iv. 321, c. 1450, pl. lxxxx.); (4) Septem planetae, seven xylographically printed plates in the Berlin K. K. Library, German work, with German explanatory text written on separate leaves facing the engravings (Schreiber iv. 417, c. 1470, pl. cxi.); (5) Pomerium spirituale, by Henricus de Pomerio (or Henri Vanden Bogaert), in the Brussels Royal Library, bearing the date 1440 in two places; its twelve engravings seem to have originally been published as a blockbook, without any text (see below); 1 in this copy they are cut up, pasted on other (contemporary), leaves of paper, and a Latin MS. commentary added to them (see Alvin, Documents iconogr.; Schreiber iv. 317, pl. lxiv.; Conway, Notes on the Exercitium super Pater Noster; Holtrop, Mon. typ. p. 9). Some bibliographers unreasonably contend that the engravings cannot be earlier than c. 1470, and that the year 1440 is the date of the original, now lost, which the transcriber of this copy inadvertently repeated. (6) Exercitium super Pater Noster (ascribed for good reasons to the same Henri Vanden Bogaert); imperfect copy (8 leaves) in the Paris National Library (Invent. D. 1581); woodcuts printed on the recto of each leaf, and an explanatory text (in Flemish) written underneath them (Schreiber iv. 245, pl. lxxxvii.; Conway, 1. c.); (7) the same Exercitium, with the same eleven engravings that were issued, some time before, as a complete blockbook (see below), a copy of which is preserved in the public library at Mons, in which the engravings are cut up and (after the Flemish verses of the blockbook had been cut away) pasted, with their versos, on the versos of other contemporary leaves, with an explanatory (Latin) text written on the recto of the leaf next to each engraving (Schreiber iv. 247, pl. lxxxviii.; Conway, 1. c.; (8) a MS. of the Speculum humanae salvationis, with the written date 1461 (Munich Hof.-u. Staatsbibl. cod. lat. 21543), in which the 192 illustrations, usually found in the MSS. of the Speculum, have been impressed from small wooden blocks in the spaces reserved for them in the MS.; (9) another MS. of a German version of the Speculum in the same Munich library (Cod. Ger. 1126), with the written date 1463, in which the 192 woodcut illustrations, impressed in No. 8, are again impressed in the spaces reserved for them.

Of blockbooks of probable German origin the following are known: - I. The A pocalypsis, or Historia S. Johannis evangelistae ejusque visiones apocalypticae (Germ. Das Buch der haymlichen Offenbarungen 1 Dumortier testifies to having seen a copy of the engravings unaccompanied by MS. (" Notes sur l'imprimerie," in Bull. Acad. Roy. de Belg., 1841, vol. viii.).

Sanct Johans)

Of this work six or seven editions are said to exist, each containing 48 (the 2nd and 3rd edition 50) illustrations, on as many anopisthographic leaves, which seem to have been divided into three quires of eight sheets each.

The first edition alone is without signatures. Cf. S. L. oofG Sotheby, The Blockbooks, i. 1. A copy of the 5th edition (according to W. L. Schreiber, Manuel, iv. 168), 48 leaves, is in the Cambridge University Library. A copy of the supposed 4th edition in the British Museum (C. 9, c. I), and one of the 6th edition (IB. 14); also a single leaf (with signature H) of the 5th edition (IB. 16).

2. Ars moriendi. - Although the origin of this work must be ascribed to the Netherlands, some authors think that there are early German editions, among others that spoken of below as the 2nd Dutch edition. Certainly German is the edition of Hans Sporer of Nuremberg (1473), in the public library at Zwickau, and a fragment of leaf 18, in the British Museum (IB. 20); another by Ludwig zu Ulm, in the Paris National Library, and the one described in Collectio Weigel. (ii. 16), where also other, but opisthographic, editions are described (see Sotheby i. 70; Schreiber iv. 253). A copy of one of these in the British Museum (IA. 24). A copy of an edition printed in a press and ascribed to Augsburg, in the British Museum (IB. 23).

3. Ars memorandi quatuor evangelia; 30 leaves, folio, printed on one side, 15 leaves being letterpress and 15 plates (Sotheby ii. 2; Schreiber iv. 135). Copy in the British Museum ('IB. 17). ' 4. Salve Regina, bears the name of its engraver, Lienhart czu Regenspurck; 16 leaves; 2 leaves (signature a) are wanting in the only copy known of it, which was in the Weigel collection (ii. 103) and is now in the British Museum (IB. I); Schreiber iv. 381.

5. Vita et Passio Christi (German); 32 leaves, small 8vo. Two copies in the Paris Library (Sotheby ii. 143; Schreiber iv. 320, who describes other issues in German and Italian).

6. The Ten Commandments for Unlearned People (Die Zehn Bott fur die ungelernte Leut)

Ten leaves in the library at Heidelberg bound up with MS. No. 438; see Joh. Geffcken, Bildercatechismus (Leipzig, 1855), 4to; Sotheby ii. 160; W. L. Schreiber iv. 234.

7. The Passion of our Lord; 16 leaves in the Weigel collection (Sotheby ii. 141; Schreiber iv. 320), now in the British Museum (IA. 25).

8. The Antichrist (Der Enndchrist); 26 leaves, small folio (Sotheby ii. 38; Weigel ii. III; Schreiber iv. 217). Copies in the Manchester Rylands Library (Spencer collection); Coll. Weig. No. 264, leaf 6 and the upper half of 7 now in the British Museum, where also a fragment of leaf 28 is preserved; four copies at Munich.

9. The Fifteen Signs of the Last Judgment; 12 engravings, usually bound up with the engravings of The Antichrist (Sotheby ii. 42; Schreiber iv. 217). Copies as of No. 8. An edition was also published at Nuremberg in 1472 by Jung hannss Priffmaler (copy at Gotha).

To. Symbolum A postolicum; small 4to, 7 leaves printed on one side only, containing 12 woodcuts. Cf. Sotheby ii. 148; also Schreiber iv. 239, who describes three editions: (I) at Vienna; (2) at Heidelberg; (3) with German inscriptions, at Munich.

H.. The Legend of St Meinrad; 48 leaves. Copies in the libraries at Munich and Einsiedeln (Sotheby ii. 150; Schreiber iv. 385).

12. The Acht Schalkheiten, of which 8 leaves were in the Weigel collection (i. 112; Sotheby ii. 1 54) .

13. The Fable of the Sick Lion; 12 leaves. Copies in the Berlin Museum, and in the Heidelberg Library (No. 438). Cf. Sotheby ii. 159, pl. lxxxvi.; Schreiber iv. 44414. Defensorium Inviolatae Virginitatis b. Mariae Virginis; 16 leaves, folio, with the initials of the printer F(riedrich) W(althern) and the date 1470 on the first leaf (Schreiber iv. 368; Sotheby ii. 63). Copies in the British Museum (IB. 2); two at Paris; three at Munich; one at Berlin; another at Stuttgart.

15. The same work, 27 leaves, large folio, 1471, with the imprint " Johannes eysenhi.it impressor (at Regensburg) Anno ab incarnacois dnice M° quadringentesimo septuagesimo j° " (cf. Sotheby ii. 72; Schreiber iv. 374). Copies in the British Museum (IC. 4). at Berlin, Gotha, Manchester.

16. The Dance of Death (Dance Macabre; der Doten Dantz); 27 leaves; two editions; one in the library at Heidelberg; another at Munich (cf. Schreiber iv. 432; Sotheby ii. 156).

17. Die Kunst Ciromantia of Dr Johan Hartlieb (Sotheby ii. 84; Schreiber iv. 428). Ten leaves of the edition of Jorg Schapff of Augsburg c. 1478 in the British Museum (IB. 8).

18. Der Beichtspiegel or Confessionale; 8 engravings (Sotheby ii. 145; Schreiber iv. 252). Copy in the royal library (Mus. Meerman) at the Hague.

19. Exercitium super Pater Noster, only one leaf (the first) preserved at Kremsmtinster, of a German edition (Schreiber iv. 247). For two xylo-chirographic issues of this Netherlandish work, see above, and below for a xylographic edition. 20. Biblia Pauperum, German text; copy in the British Museum (IB. 3); and a copy of another edition (40 leaves) with the device of Hans Spoerer, and the date 1471 (IC. 5).

21. The Apostles' Creed; 7 leaves, folio. Copy at Wolfenbiittel. 22. The Credo, in German; 12 leaves, 4to. Copy in the Munich Royal Library.

any neat-handed man could print for himself. We learn from the inventory of the possessions of Jean de Hinsberg, bishop of Liege (1 4 1 9 - 1 455), and his sister, a nun in the convent of Bethany, near Mechlin, that they possessed " unum instrumentum ad imprimendas scripturas et ymagines, " and " novem printe lignee ad imprimendas ymagines cum quatuordecim aliis lapideis printis." These entries would seem to indicate that people purchased engraved blocks of wood or of stone from the wood-cutter rather than books from a printer.

Concurrently with these single woodcuts, with or without written or xylographic text, arose a class of books, in some of which written texts were added to pictures 23. Pro13ugnacula, seu Turris sapientiae (Sotheby ii. 164). One sheet, piano, in the British Museum (IC. 30). It may have originated in the Netherlands.

Blockbooks of Netherlandish origin are: i. Apocalypsis S. Johannis. - Copy in the Haarlem Town Library. A copy of the 3rd (?) edition, of 50 leaves, in the British Museum Of Nether- (IC. 40), the leaves 36 and 38 having been supplied from O Net another copy. Leaf 21 of another copy in the same. library.

Origin. 2. Biblia Pauperum; 40 folio leaves (each bearing a signature: a to v; .a. to .v.). As many as seven editions have been distinguished by Sotheby (i. 43), Holtrop (Mon. typ. p. 3), and ten by Schreiber (iv. I), who likewise mentions a Latin edition of 50 leaves, besides the two editions with German texts of 1470 and 1471. The British Museum Catalogue of 15th-century books enumerates copies or fragments of copies of seven editions.

3. Speculum humanae salvationis

Of this work a blockbook must have existed, of which only 10 sheets (= 20 leaves) with woodcuts and texts, besides 12 isolated woodcuts (used in 1483), have come down to us. We speak of it at length below when dealing with the typographic editions known of this work.

4. Ars moriendi; 24 leaves, small folio, 13 containing text, II plates. See above (German) No. 2; Sotheby i. 69; Holtrop, p. 8; Schreiber iv. 253, who enumerates thirteen editions, some of which are German. 1 The theory, started a few years ago, that the engravings of this blockbook are imitations of the sketches by the master E. S. (see M. Lehrs, Der Kunstler der Ars moriendi, 1890; L. H. Cust, The Master E. S., 1898) is wholly inadmissible. Copy in the British Museum (IB. 18), and an imperfect one in the Haarlem Town Library.

5. A copy of another edition of 24 leaves in the British Museum (IA. 19).

6. Canticum Canticorum; Historia seu Providentia B. Virginis Mariae ex Cantico Canticorum; 16 leaves in folio, two editions (Sotheby i. 77; Holtrop, p. 6; Schreiber iv. 151). Copies in the Haarlem Town Library (wanting the leaves 3, 4, 7, II, 13, 15, 16); the British Museum (IB. 46), which possesses also a copy of another edition (IC. 47).

7. Liber Regum, seu Historia Davidis; 20 leaves, folio (Sotheby i. 120 b; Schreiber iv. 146). Some consider this to be a German work.

8. Exercitium super Pater Noster, by Henricus de Pomerio or Henry Vanden Bogaert; io leaves, small folio (Sotheby ii. 137; Holtrop p. io; Conway, Notes on the Exercitium, 1887; Schreiber iv. 245). For other editions see the two preceding sections.

9. Pomerium Spirituale, by the same author as No. 8; 12 leaves, having 12 woodcuts. This blockbook is now only known from a xylo-chirographic issue with the MS. date 1440 (see above), preserved in the Brussels Royal Library. See Conway, Notes on the Exercitium. 10. Temptationes Demonis temptantis hominem de septem peccatis mortalibus; a single large folio leaf printed on one side (Sotheby i. 122°; Schreiber ii. 249). One copy in the British Museum (IC. 29), another in the Wolfenbiittel Library.

11. Vita Christi, or The Life and Passion of Christ; 36 cuts, originally printed in a press on six anopisthographic leaves, in 8vo. Copy in the Erlangen Library (Campbell, Annales, 746). 12. Historia Sanctae Crucis; a fragment of one leaf (with signature g), formerly in the Weigel Collection (ii. 92), but now in the museum at Nuremberg; it seems to be only a proof-sheet.

13. Alphabet (grotesque) in figures (Holtrop p. I i; Sotheby i. 122; Schreiber ii. 324-327). - There is one copy in the British Museum and another in the Basel Library, the latter having the date 1464 engraved on the letter A, which is mutilated in the Museum copy. A similar alphabet preserved at Dresden seems to be a copy made in Germany.

14. Donatus (Aelius) de octo partibus orationis. Leaf 6 of an edition c. 1500 of 16 leaves in the British Museum (IA. 48). For other xylographic editions of this work cf. Holtrop, Mon. typ. Besides the works of Sotheby, Holtrop, Weigel, Schreiber, Lehrs, Cust, &c., quoted above, consult Sir W. M. Conway, The Woodcutters of the Netherlands in the 15th Century (Cambridge, 1884); Heinecken, Idee generale (Leipzig, 1771); J. Ph. Berjeau's Facsimiles of the Biblia Pauperum, Canticum Canticorum, Speculum (London, 18 591861), and idem, Catal. Illustre des livres xylogr. (London, 1865); Dodgson, Cat. of Early German and Flemish Woodcuts in the Brit. Mus. - When the art of writing, and that of printing from wooden blocks (xylography), and all the subsidiary arts of illuminating, Invented at decorating and binding manuscripts, books, pictures, Haarlem. &c., were at their greatest height, and had long passed out of the exclusive hands of the monasteries into the hands 1 Heinecken enumerates six editions, of which one has German inscriptions. See also an article by Guichard, in Bull. du Bibliophile (Paris, 1841).

xxvll. 17 of students and artisans, the art of printing with movable cast-metal types (typography) was invented. As to when, where and by whom this invention came about, a dispute has been waged for more. than four hundred years. It will be seen below that we must attribute it, as in our former edition, to Lourens Janszoon Coster, of Haarlem, and not to Johan Gutenberg, of Mainz.

In saying this, we are aware that in the year 1900 (exactly four hundred years after the Cologne Chronicle had publicly started the dispute by saying that Gutenberg had improved but not invented the art) Germany enthu- the claims p ) y of Germany.

siastically celebrated the supposed 50oth anniversary of his birthday. The speeches delivered on that occasion, after making faint allusions to the doubts and opposition of former times, all declared that, after the rediscovery of the Helmasperger document of 1455, which could not be found in 1880 (Hessels, Gutenberg, pp. 99-101), it was impossible for any unbiased person to dispute Gutenberg's claims to the honour of the invention any longer.

In the same year a Gutenberg Museum was erected at Mainz to be a repository for anything connected with Gutenberg and printing; also a Society (Gutenberg-Gesellschaft) founded with the view of publishing any book that related, however remotely, to Gutenberg and his invention, to which the whole civilized world was invited to subscribe, as its object was to honour the genius who had conferred such an inestimable boon on mankind by his invention. As a first result, a " Festschrift " was published containing an historical introduction by Professor Hartwig; and articles on the first steps to typography (Schreiber); stamp-printing before Gutenberg and the Psalters of 1 457, 1 459, &c. (Falk); 15th-century printing in France (Labande); German printers in Spain and Portugal (Habler); German printers in Italy (Marzi); the coloured initials in Fust and Schoeffer's Psalter (Wallau); the Turkkalendar for 1455 (Wyss); the earliest spread of typography (Velke); also an elaborate pedigree of the family Gansfleisch (Schenk zu Schweinsberg), and an equally full account (by Schorbach) of all the documents related to Gutenberg. This " Festschrift " was followed by publications of the " Gutenberg Society ": I. (1902) Die iilteste Gutenberg type (Zedler); II. (1903) Die Donatand Kalendar type (Schwenke); III. (1904) Das Mainzer Fragment vom Weltgericht (Schroder, Zedler, Wallau); IV. (1905) Das Mainzer Catholicon (Zedler); V., VI., VII. (1908) Das Mainzer Fragment vom Weltgericht (Schroder); Die B42 type im Schiifferschen Missale Mogunt. von 1493 (Zedler); Die Missaldrucke P. Scho fers and seines Sohnes Johann (Tronier); Zu den Biicheranzeigen Peter Scho fers (Velke).

We admit the great value of these learned and painstaking publications, and those who have the time and patience to study the mass of material here brought together in a somewhat bewildering fashion, will find their knowledge enriched on various subjects connected with early printing, but no proofs that Gutenberg invented it. It is clear from these books that their authors firmly believed from the outset that Gutenberg invented printing, and printed nearly every book that appeared or can be placed before his death in 1468. Under this impression they always speak of him as the " great master," the " great genius," &c., and represent him, not as inventing printing by accident, but as conceiving, somewhere about 1436 or earlier, the idea of inventing it, and meditating from that moment over the problems which he had to solve. Consequently, our authors read a good deal between the lines of their documents, which we fail to find there, and in this way the texts of the documents always show somehow that " the great master " is making or has already made his invention. For instance, the Strassburg lawsuit of1436-1439is to them an unimpeachable proof that Gutenberg was secretly working there at printing and trying to solve his problems; when he is paying there, during the same time, a considerable sum in duties for large quantities of wine, we are told that he was then in good circumstances; but when he borrows money in 1442, 1448, 1450 and 1452, and is summonsed in 1455 for not repaying the two last loans, and prosecuted in 1457 for not paying the interest due on his first debt, it is all owing to his difficulties in working out the problems of his invention, though the documents themselves never allude to any " invention " and may be interpreted in quite a different way.

jHISTOR Y

We proceed to examine the documents. The earliest mention and description of the new art is perhaps that in the Donatus issued by Peter Schoeffer at Mainz before 1456, which, according to its colophon, was finished " Arte nova imprimendi seu caracterizandi (from character = letter). .. absque calami exaratione." Fust and Schoeffer said of the Mainz Psalter of 1457 that it was formed by an " adinventio artificiosa imprimendi ac caracterizandi absque calami ulla exaratione. The colophon of the Catholicon of 1460 says that the book was printed " non calami, stili, aut pennae suffragio, sed mira patronarum formarumque concordia, proporcione, ac modulo." In 1462 Albrecht Pfister says that he had " gedrucket " the Four Histories. Fust and Schoeffer say of the Liber Sextus Decretalium, published in 1465, that it was completed " non atramento (" atramento communi in the Justinianus of 1468 and 1472), plumali canna neque aerea, sed artificiosa quadam adinventione imprimendi seu caracterizandi," which phrase they slightly varied in Cicero's Officia, issued in the same year: " non atramento, plumali canna neque aerea, sed arte quadam perpulcra." The edition of St Jerome's Epistles of 1470 is said to have been completed by an " ars impressoria," the Decretum Gratiani of 1472 by an " ars quaedam ingeniosa imprimendi," the Dyalogus of 1478 by an " ars magistra." We find further - " ars sancta or divin g, nova ars scribendi," novum exscribendi genus prope divinum," "sculptoria archetyporum ars," " ars mirifica formandi," " ars excusoria," " nova imprimendi ratio," " ars pressurae," " chalcotypa ars," "chalcographia " (1472 and later), " chalcographia excusoria impressoriaque," " libraria impressio," " empryntynge " (Caxton, 1482), " prenterei " (Schoeffer, 1492), " truckery " (1505), " impression des livres " (1498), and " prenten." The early printers called themselves, or were called by others, " librorum prothocaragmatici " (Gramm. Rhythm., 1468), " impres sores librorum," " exsculptor librorum " (Jenson, 1471), chalcographus " (1473; Hain 13036), " magister artis impressoriae," " boeckprinter "; and during the 16th century we find them still frequently called " chalcotypus " and " chalcographus." The types were at first designated more by negative than positive expressions. In 1468 they were called " caragma," later on " caracter " or " character," ' ` archetipae notae " (1473 Hain 13036), sculptoria archetyporum ars, chalcotypa ars," " formae," " artificiosissimae imprimendorum librorum formae." We soon hear also of the process and material by which they were produced. The Grammatica of 1468, published by Schoeffer, says that it was " cast " (sum fusus libellus). In 1471 " aeneae formulae " are spoken of; and Bernardus Cenninus and his son testify that they had printed the Virgil " expressis ante calibe caracteribus et deinde fusis literis " (with letters first cut into steel and then cast). In 1473 Friedrich Creusner at Nuremberg states that he had " cut " (sculpsit) the work of Diogenes (Hain 6192). Johan Zeiner of Ulm says in 1474 that he had perfected a book, not with the pen, but with letters of metal (stagneis caracteribus). In 1474 Joh. Ph. de Lignamine speaks of " metallicae formae." In 1476 Husner of Strassburg represents the Nider as being printed with " letters cut of metal (litteris sculptis artificiali certe conatu ex aere)." Nicolas Jenson printed in 1480 with letters " cut and cast " (sculptis ac conflatis).

The word typographus seems to occur for the first time in 1488, in the preface of P. Stephanus Dulcinius Scalae to the Astronomicon of Manilius, printed in that year at Milan by Antonius Word Zarotus; 1 in 1498 Erasmus uses it in a letter (dated Feb. 13) to Christianus, a Lubeck merchant;' and in 1517 Johan Schoeffer applies the word to himself in the colophon of the Aeneas Sylvius published by him. But of the use of the word typographia no earlier instance is known than 1520, in which year Gerardus Noviomagus (= Geldenhaurius) in his Lucubratiuncula de Batavorum Insula (pref. to Nicol. Buscoducensis, dated 1520) says: " inventa Germanorum. .. bombarda videlicet, typographia, pyxis chartaque nautica "; and Johan Schott, a printer of Strassburg, in the Geogr. Ptolem. published by him, describes his grandfather, Johan Mentelin, as " primes typographiae inventor." Gerardus, it may be added, borrowed the whole passage from Pet. Montanus (li. 1 Adag., published an. 1504), who has chalcographia instead of typographia. Meerman indeed a speaks of a use of the word typographia (or at least of typographus) earlier than 1520, and refers to the preface of Bernardinus Veronensis in the edition of Tibullus, Catullus and Propertius published at Venice in 1493 by Symon Bevilaqua, " at least," Meerman adds, " as it (the preface) is read in the Annal. typogr. of Maittaire, i. 560, 2nd ed." But on page 560 Maittaire quotes the first two lines of Bernardinus's preface (till dicit) and then adds: " Graecis characteribus destitutus, typo graphus necesse habuit hiatus in commentario hic illic relinquere, is evidently Maittaire's own remark, not that of Bernardinus. The present writer at least has been unable to find such a passage in the Tibullus.

When we, for the moment, leave out of sight the question as to when, where, and by whom the art was invented, and 1 Maittaire, Annales Typogr. i. 508, note I.

Opp. iii. col. 24.

Origg. Typogr. i. 32, note cx. take our stand on well-authenticated dates in such printed documents as have been preserved, we find that the first printed date, 1454, occurs in two different editions of the same letter of indulgence issued in that year by Pope Nicholas V. in behalf of the kingdom of Cyprus.

These two editions bear no printer's name, nor the place of printing, but are distinguished respectively as the 3 i-line and the 30-line Indulgence. The one with 31 lines claims priority,' from a chronological point of view, over the one with ndu gence 30 lines, because one of the sold copies that has been ndu g. preserved was issued at Erfurt on the 22nd of October 1 454 (in the possession of Herr Ernst Fischer at Weinheim, Centralbl., 1909, p. 30); a second (in the Hanover Archives; Veroffentl. II. tab. i.) at Fritzlar on the 12th of November 1 454; a third (in the Mus. Meerman, at the Hague) at Erfurt on the 15th of November 1 454, &c., whereas of the 30-line Indulgence the earliest sold copy that has as yet come down to us was issued at Cologne on the 27th of February 1455, though it has the printed date mccccliiii., which was altered with the pen to mcccchuij. In the 31-line Indulgence occur (a) a large church type used for the headings and commencing words of the absolutions, for the first word in the document and for the Christian name of the pope's legate; (b) a smaller text or brief type for the text; (c) a large initial V and two large initials M, which slightly differ from each other. In the 30-line Indulgence occur (a) a large church type, used as in the 31-line Indulgence; (b) a smaller text or brief type for the text; (c) a large initial U, and two large initials M differing from each other.

These two different editions are usually regarded as having been printed at Mainz; and, in the absence of any evidence to the contrary, we assume that such really was the fact. But we must at the same time conclude that about October 1454 there were at least two rival printers at work there: (1) the printer of the 3 i-line Indulgence, who may have been Johan Gutenberg, perhaps subsidized by Johan Fust; (2) the printer of the 30-line Indulgence, who was no doubt Peter (Schoeffer) de Gernssheym, as this Indulgence is connected with one of 1489 printed by him. Four written copies of this 1454 Indulgence are known to exist which respectively bear the dates: Frankfurt, 10th April 1454 (in the possession of Herr Lais, Wiesbaden); Frankfurt, 11th April 1454 (Frankfurt Archives); 11th July 1454 (place unknown; Darmstadt archives); Lubeck, 6th October 1454. As their dates precede by a few weeks only the earliest known date (Oct. 22, 1 454) on a printed copy, they mark, perhaps, the exact time when printing made its appearance at Mainz, in an already advanced state of perfection.

Basing ourselves on the above Indulgences with their printed date, and four different types, we subjoin two lists of the books which the German bibliographers of the present day regard as having all been printed by Johan Gutenberg at Mainz, in the types or " developments " of them, employed for these Indulgences. They are arranged in two columns (A and B) according to types, but without regard to strict or supposed chronology. For further details cf. Hessels, Gutenberg (1882), p. 150 sqq.; Schwenke, Berlin Festschr. (and in the T7ereentl. of the Mainz Gutenberg-Gesellsch.); Zedler (Gutenberg-Forsch. and in the Vereentl.), &c.

A.

Types: I (large church type, also called the 36-line Bible type) and II (smaller brief type), used by an unknown printer, not later than October 1454.

i. 31-line Indulgence; three different issues (A, B, C), with the printed year mccccliiii., and one issue (D) with the printed year mcccclv. All printed on vellum. Of issues A and B no sold copies have yet come to light; but three unsold copies of each are preserved at Brunswick, Wolfenbiittel and Hanover (Culemann coll.). Of issue C ten sold copies are known to exist in various libraries with dates ranging from the 22nd of October 1454 to April 1455, besides three unused copies. Of issue D ten sold copies with dates from the 7th of March 1 455 to the 30th of April 1455 and four unused copies are known.

No inferences can be drawn from this priority, as it merely rests on the date of a sold copy that has come to light.

B.

Types. III (large church type, somewhat smaller than Type I, also called the 42-line Bible type) and IV (a smaller brief type), used by Peter Schoeffer de Gernssheym (1454-1455).

i. 30-line Indulgence; one issue (A) with the printed year mceccliiii., and two issues (B, C) with the printed year mcccclquinto. All printed on vellum. Of issue A only one copy has been discovered (now in the RylandsSpencer Library), which was sold atCologne on the27th of February 1 455, the printed date mccccliiii having been altered with the pen to mcccclu u i. Of issue B two sold copies, with dates April i 1 and 2 9, 1 455, are in the Berlin Library and the British Museum. Of issue C a sold copy with date April 2 4, 1 455 is at Wolfenbiittel.

A (contd.). Type I continued; for type II. (of which no further trace is found) see below.

ii. Poem on the " Weltgericht." Fragment of one leaf (paper), discovered at Mainz about 1892, preserved in the Gutenberg Museum at Mainz; presumed to have been printed c. 1443-1444.

iii. Donatus, 27 lines. Fragments of 4 vellum leaves (4, 5, 8, 9) recently discovered in the Heiligenstadt Library, and now preserved in the Berlin Royal Library.

iv. Donatus, 27 lines. Two rubricated vellum leaves (5 and to) of an edition of 14 leaves, usually called the Donatus of 1451, preserved in the Paris National Library.

v. Donatus, 27 (?) lines. Two strips of vellum leaves, containing the remains of 3 lines and about 30 mutilated letters, discovered in the Heiligenstadt Library, and now in the Berlin Royal Library.

vi. Astronomical Kalendar, said to be for the year 1448, therefore supposed to have been printed at the end of 1447. Fragments of two large vellum rubricated sheets, printed on one side, discovered in 1901 in the binding of a MS. belonging to the monastery of Schonau, near Mainz, now preserved in the Wiesbaden Landesbibliothek.

vii. Donatus of 18 leaves, 26 lines, on vellum; of which 2 rubricated sheets (4 leaves, I, 2, 9, to) are preserved in the Berlin Royal Library; probably issued between 1447 and 1450 (Centralbl. xxvii. 65 sqq.).

viii. Manung widder die Durken. An almanac for January 1 455, in 4to, 5 paper leaves, 20 and 21 uneven lines. A unique copy, discovered at Augsburg, now in the Munich Hof Library.

ix. A German translation of the bull of Pope Calixtus III., dated XII. Kal. Julii (=Jun. 20) 1456. Fourteen rubricated leaves 4to, in the Kalendar type, except that two of the capital E's belong to the B36 type (13b and 14 blank), preserved in the Berlin Royal Library; not to be ascribed to P. Schoeffer (Centralbl. xxvii. 63).

x. Conjunctiones et oppositiones solis et lunae (now called by German bibliographers Laxier-Kalendar). A calendar for 1457, a broadside paper sheet, printed on one side, of which the upper half of the only copy known, discovered at Mainz, is in the Paris Library.

xi. Der Cisianus (not Cislanus) zu Dutsche. A broadside paper sheet, 36 lines, printed on one aide, with separate headline. The Tross-copy mentioned in suppl. to Brunet's Manuel (1878, sub voce " Cislanus ") was bought in 1870 for the Cambridge University Library.

xii. Donatus, 27 lines, 14 vellum leaves, of which the British Museum possesses the leaves 4, to and II (entire) with fragments of the leaves 2, 6-9 and 13. A fragment of 62 lines B (contd.). Type III continued (till about 1 457; of Type IV no further trace is found).

ii. Donatus, of 35 lines, folio, printed, according to the colophon, " per Petrum de Gernssheym. in urbe Moguntina cum suis capitalibus." iii. Bible of 42 lines (also called Mazarine Bible and referred to below as B 42), printed before the 15th of August 1456, as the binder of the paper copy in the Paris Library states that he finished its rubrication on that day. Two volumes folio, 641 leaves in2 columns of 4211neseach, though in somecopies thecolumns of pp. I to 9 contain 40 lines only, while the 10th page has 2 columns of4 'lines each,thedifference in the number of lines making no difference in the space which they occupy. For other copies see Hessels, Gutenberg, p. 170; Dziatzco, Beitr. zur Gutenbergfrage (Berlin, 1889); Schwenke, Festschr., who has drawn up a list of all the copies known to be still in existence. The copy known as the Klemm copy, which was bought by the Saxon Government in 1886, and presented to the " Deutsches Buchgewerbemuseum " at Leipzig, has the year " 1 453 " written in small Arabic numerals of 15th-century form at the bottom of the last leaf of the second volume. But this date is highly suspicious, for Klemm, who must have known its importance and high value, never mentioned it, though he described his copy three times, in 1883 and 1884.

iv. Donatus of 33 lines. Vellum fragment at Oxford, without printed initials.

v. Donatus of 33 lines. Vellum fragment at Paris, without printed initials; also three rubricated leaves (5, 6 and 8) in the Berlin Royal Library (Centralbl. xxvii. 68).

vi. Donatus of 33 lines. Leaf t (defective) on vellum, mentioned in Ludw. Rosenthal's Cat. 105, No. 3, and purchased by the Berlin Royal Library, which has also acquired the leaves t and II (Centralbl. xxvii. 69.). The large Psalter initials are used for the initials of chapters.

vii. Donatus of 33 lines. Leaf (vellum) discovered in the Berlin Royal Library.

viii. Donatus of 33 (?) lines. Small fragment, discovered in the library at Giessen, of a vellum leaf, which Schwenke thinks may be the 10th of an edition which differs from Schoeffer's 35-line edition, and also from the Paris 33-line edition.

ix. Donatus of 26 lines. One defective vellum leaf, discovered in a Munich private library, and now in the Mainz Gutenberg Museum.

x. Donatus of 26 lines. One vellum leaf at Mainz, another at Hanover, a third in the British Museum.

xi. Donatus of 24 (?) lines, between 1470 and 1477 (Schwenke).

A (contd.). in the Bodleian Library two small fragments discovered in the library at Heiligenstadt.

xiii. Donatus, 27 lines, which Schwenke calculates to have consisted of 14 vellum leaves, of which the leaves 6 to 9 are now in the Berlin Royal Library.

xiv. Donatus. 27 lines. Three strips of a rubricated vellum leaf 5 discovered in the Karlsruhe Hot-Bibhothek.

xv. Donatus. 27 lines. One rubricated vellum leaf (6). in the Kalendar type. in the Berlin Library (Centralbl. xxvii. 62) xvi Donatus. 27.28 or 30 (?) lines. Fragments of two vellum leaves of an edition of 12 (?) leaver discovered in the binding of a book (printed at Milan in 1476) which formerly belonged to the Episcopal Library at Salzburg, and is now in the Munich Hof-Bibliothek.

xvii. Donatus, 27 (or 30?) lines. Vellum fragments of an edition of 12 (?) leaves in the British Museum (C. 18. e. i No. 5). Leaves t and 2 are in the Bodleian Library, and leaf 8 in the Mainz Town Library.

xviii. Donatus, 27 lines. Fragment of a vellum leaf (3?) discovered in the binding of a MS. in the Munich Hof-Bibliothek.

xix. Donatus, 27 lines. Two vellum fragments of the leaves 6 9, the upper part of which is preserved in the Bodleian Library (Auct. 2 Q infra I. 50 No. 6), the lower part in the Bamberg Royal Library (VI. F 1).

xx. Donatus, 28 (?) lines. One defective vellum leaf, showing 25 lines, formerly in the possession of Jacq. Rosenthal (Incun. typ. ii. No. 2154), afterwards in the Amherst collection (Handlist No. 5). Another leaf in the Mainz Gutenberg Museum.

xxi. Bible of 36 lines (referred to everywhere as B 36), 2 vols., folio, 882 leaves, with 2 columns of 36 lines each on a page. Some bibliographers, assuming that Pfister printed it, call it the Pfister Bible. A paper copy of it is in the Paris Library, and also a separate copy of the last leaf, which bears the MS. date 1461. Other copies are preserved in the Rylands-Spencer Library, in the British Museum, at Jena, Leipzig, Antwerp, &c. (Hessels, Gutenberg, p. 160; Bernard, Origine, ii. 31).

B (contd.). xii. Cantica ad Matutinas; only known from one vellum leaf (the first) in the Paris Library, considered to be the remains of a Psalterium. for the printing of which Humery may have furnished (!) the type (Schwenke Untersuch. p. 72 seq.). Judging from the leaf preserved, the work corresponds in every respect to the 42-line Bible, having double columns 42 lines. &c.

Type V. - The " first stage " of Type VII., supposed by Otto Hupp (Ein Missale Spec.) and others to have served for printing (1) a Missale speciale, in the possession of Ludw. Rosenthal at Munich; (2) a Missale abbreviatum discovered in 1900 in the Benedict Church of St Paul in the Lavantthale.

Type VI. - The large type for the Psalter of 1457.

Type VII. - The small type for the same Psalter (" second stage " of Type V). Types VI and VII were also used for the " Canon Missae " of 1458, a copy of which is preserved in the Bodleian Library.

Type VIII used for (t) Joannis de Balbis Catholicon of 1460. Large folio, 373 leaves, with two columns of 66 lines each on a page; (2) Matth. de Cracovia, Tractatus racionis, 22 leaves with 30 lines to the page, 4to; (3) and (4) Thomas de Aquino, Summa de articulis fidei, two 4to editions, one of 13 leaves with 34 lines to the page; the second of 12 leaves with 36 lines to the page; (5) an Indulgence of 1461 of 15 lines (see Hessels, Gutenberg, p. 171 sqq.).

The above eight types and the books printed with them (besides a few others printed by Albrecht Pfister at Bamberg) are the only ones that bear, more or less closely, on the question regarding the introduction, or possible invention, of printing at Mainz.

Till recently the church type 1, of the 31-line Indulgence, had always been regarded as identical with that of B 36, and the church type 3, of the 30-line Indulgence, with that of B 4V. But, as the capital P of Indulgence 30 seems not to occur in B 42, and on examination minute differences show themselves in other respects, identity between the two types cannot be accepted. The use of the brief type 2 of Indulgence 31 seems to have been limited to printing this one document, as its great resemblance to the type employed at Eltville in 1472 for printing a Vocabularius ex quo, and Thomas Aquinas' Summa de articulis fidei, amounts not to identity. Nor has any further trace been found of the brief type 4 of the Indulgence30, so that the four types used for the two Indulgences were, perhaps, specially manufactured for them and discarded afterwards or melted down for other types.

Hence there is nothing to connect these two broadsides with any locality or any printing-office, except that one of the initial M's of the Indulgence 30 re-occurs as the initial M of the second absolution of a 33-line Indulgence of 1489, which was unquestionably printed by Peter Schoeffer at Mainz, for " Raymundus Peyraudi archidiaconus Alniensis in ecclesia Xanton," who issued it at the order of Pope Innocent VIII., " pro tuicione orthodoxe fidei contra Turchos. " For this reason types 3 and 4 and the books printed with them, including B 42, must all be ascribed to him, all the more as he printed, with the type of B 42, the 35-line Donatus, which bears his name in the colophon. As Schoeffer, in the colophon of this Donatus (ii.) which bears his name, says that it was printed " cum suis capitalibus," and as these capitals gradually disappear after 1 459 and the type of the 42-line Bible is no longer found after 1456, we must presume that some of the twelve incunabula mentioned above (in col. B) were printed by Peter Schoeffer alone before he entered (in 1457) into partnership with Johan Fust (see Hessels, Gutenberg, p. 166 seq.).

During the last two decades, however, the two types (3 and 4) and most of the books mentioned above in column B, including B 42, together with the two types (1 and 2), and several of the books in column A, including B 36, have been attributed by German bibliographers to Gutenberg. This singular proceeding is chiefly owing to the late Dr Dziatzko's treatises (Beitrcige zur Gutenbergfrage, 1889; Gutenberg's fruheste Druckerpraxis, 1890) on Gutenberg's supposed work as a printer. This author, noticing that the two types of B 36 and B 42, their signs of contraction, marks of punctuation, &c., though differing in size, closely resemble each other in form, concluded that they were manufactured in one and the same office, by one and the same printer, that is, Gutenberg. He thought his conclusion confirmed by the two Bibles being printed on the same kind of paper showing the same watermarks, and arranged in quires in the same way, and divided off into parts at the same place. Finally, from a misprint in B 42 being rectified in the Stuttgart copy of B 36 by a cancel (Druckerpraxis, p. 95), he concluded: (a) that B 36 was a reprint of B 42; that the latter was printed by Gutenberg during his partnership (1450-1455) with Fust, who supplied the money and the material, while he himself superintended the manufacture of the type, instructed the compositor and printer, and therefore was its printer; and that the type came afterwards into Schoeffer's possession; (b) as B 42 was Gutenberg's first work, and had been begun in 1450, B 36, a reprint of it, could not be dated before this year; but as its type already existed in 1454 (in the Indulgence 31), Gutenberg, foreseeing his quarrels with Fust, must have been preparing it since 1453, and have printed with it, first, some Donatuses, the Indulgence 31, &c., and finally B 36, with the technical and financial assistance of Albrecht Pfister who, shortly before 1458, acquired its type and printing-material (see further, Hessels, " A Bibliogr. Tour, " in The Library, July 1908). Dr Dziatzko, noticing also a " resemblance " between the types and the workmanship of the two Indulgences, attributed both these broadsides likewise to Gutenberg.

His conclusions, and the method of research by which he reached them, the German bibliographers of the present day have adopted and amplified into a bibliographical and typographical " system," which professes to examine minutely the form and size of every letter, capital or small; the combined letters like do and de cast on one type; the signs of contraction above, or by the side of or through certain letters, the marks of punctuation, the habits and workmanship of the printer, the arrangement of the quires, the paper and its water-marks, &c.

The " system " divides the Gothic or Church types with which B 36 and B 42 and the other books mentioned above are printed into " chief " and " by-forms," (Hauptand Nebenformen). The tops and bottoms of the former are ornamented with minute protruding tags, angles and points, while the " by-forms " miss most of these ornaments, their limbs being straight on the left or right, so as to be easily joined to the protruding tags, angles and points of the " chief forms, " whenever the two come together. For instance, if a u or a t follows an e, the " by-form " of u with straight limbs was to be used, while the t was to be without its crossbar protruding on the left.

The bibliographers who deal with the incunabula enumerated above, in accordance with this " system," regard the books in which they find these chief and by-forms used in their proper places as the earliest, and therefore as the products of Gutenberg's " creative genius and skill," while they ascribe the books which bear evidence of the misuse of those forms to other printers, but their types to him. But this is an uncertain guide, as by errors in the distribution of the types after the printing of the first or second pages this misuse may already occur in the third and further pages of a book. In this way, however, the " system " arranges the books enumerated above in the following approximately chronological order: 1 4432 444. " First phase " of the Gutenberg type (=the Donatus type). The numbers ii., iii., iv. (with the suspicious date 1451) and v.

1447 (end of) till 1457(?). " Second phase " of the same type (=the Kalendar type). The numbers vi. to xiv.

1450-1453. B42 presumed to have been finished in or before 1453, taking this year, written in the Klemm copy, as genuine.

1453. " Third phase " of Gutenberg's type, B 36 (xviii., of which the earliest known date is 1461).

1 454. The two Indulgences with their types (1, 3; 2, 4). 1457. The two Psalter types.

1461, 1462 till (?). Pfister, who is said to have acquired the type of B 36 from Gutenberg, is known to have issued a book with the date 14 February 1461, and another with the year 2462. Hence, Schwenke says that the 36-line Bible type, which he regards as a " continuation. " of the Donatus, and the Kalendar types, had a life of nearly 20 years (Veroffentl. ii. I). Type v. is thought to be Gutenberg's earliest (before 1443!) by the few who regard the " Missale speciale " and the " Missale abbreviatum " as his work.

The " Donatus type " is so called from the Paris Donatus, on one of whose leaves the year 1451 is written. Zedler, somewhat unreasonably, considers this date to be a forgery of Professor Bodmann, though he is known to have forged other Gutenberg documents. This type is regarded as the same as that of the Astronomical Kalendar, but in an earlier, more imperfect stage. As this Kalendar calculates the ephemerides of the sun, moon and stars, either for the year 2429 or for 1448 or 1467, it is presumed to have been printed for 1448, that is at the end of 1447, and as its type looks new and almost perfect, the Paris Donatus is placed considerably earlier because its type looks old. The poem on the " Weltgericht " i No. ii.) is said to show all the forms of the Donatus type, but as ts workmanship looks primitive, it is dated back to 1443-1444The Heiligenstadt Donatus (No. iii.) is placed after the " Weltgericht " (ii.), but before the Paris Donatus (iv.) and the other Heiligenstadt Donatus (v.). Some German bibliographers do not feel sure that Gutenberg manufactured types v., vi. and vii., though they have no doubt as to the remaining. Others are of opinion that Pfister printed some of the books in the type of B 36; Schwenke thinks this Bible could not have been begun before 1457, but all agree that every book in the above lists must have been printed either by Gutenberg himself, or in his office, or with his type, or under his superintendence.

Though the church type 1 cannot be said to be identical with that of B 36, and no further trace of the brief type 2 has been found, we see no reason for separating Indulgence 3 ' from Mainz printing. And assuming that it was printed there, its printer may have been Johan Gutenberg, who was at Mainz in 1454.

A peculiarity of the above-mentioned " system " is that it ascribes two types, so different in size, shape and form, as those of B 36 and B 42, to one and the same printer, merely because they " resemble " each other. This shows that the " system " takes no account of the fact that the inventor of printing, and all the early printers who came after him, in manufacturing their types necessarily imitated the forms of the written characters of their time. Hence if two printers simultaneously erected their presses in one town, their types, though cut and cast independently, were apt to resemble each other, as appears from various examples. The printers of B 3 ' and B 42 are no exception to this rule; they each took a MS. as their model, and the types which they produced are simply imitations of the Gothic or Church hand, which, from its first beginnings in the 10th century, if not earlier, can clearly be traced down to, and reached its greatest development in, the 15th century.' The written characters of all ages and countries resemble and yet differ from each other in various respects; and as their resemblances and differences are closely reproduced by the metal printing types of every country, we are able to ascribe MSS. as well as incunabula to definite countries, some manuscripts even to " schools, " a few even to definite scribes. But when two types differ in size and form, however slightly, and there is no evidence that they belonged to one and the same printer, some of their characteristics may justify us in ascribing both to the same country or town, but not to the same printer. It is, moreover, not safe to ascribe incunabula to one and the same printer on account of their similarity of the quires and divisions into volumes, their paper or water-marks (which Dziatzko observed in the two Bibles), as these particulars are nothing but a continuance of the MSS.

1 The Cambridge University Library possesses two folio volumes (press-mark; Dd. 7. I, 2), the writing of which, ascribed in the catalogue to 1490, resembles the types of B 42 with all its chief and by-forms so much, that at first sight they might be mistaken for copies of this Bible.

Nor is his evidence for saying that B 36 is a reprint of B 42 conclusive. The types of B36 and B 42 may be ascribed to Germany, but as both are used for the printing of a Bible and editions of Donatus, it is improbable that the printer of B 42 and one set of Donatuses should manufacture, about the same time, another type for another Bible and another set of Donatuses. We have shown above that B 42 must, on bibliographical grounds, be ascribed to Peter Schoeffer at Mainz, and as he used its type for a book which actually bears his name, all the other books in the same type must be ascribed to him. It follows that B 36 and every other book in column A must be assigned to some other printer or printers.

Type v. is a Church type and resembles those of B 36 and B42, but it can have nothing to do with Gutenberg or the invention of printing, as it is not earlier than 1480-1490. Types vi. and vii., which are nothing but imitations of the written Psalters of the time, are employed for a work, the colophon of which distinctly mentions Fust and Schoeffer as the printers; hence they cannot be claimed for Gutenberg. Of the Catholicon type we speak below. Therefore the books numbered i. to xxi. in column A of the above list are the only ones about which there can be any doubt or discussion.

Here we encounter another peculiarity of the above-mentioned " system," which treats the three different types detected in these twenty-one works not as different, but as " phases " or " developments " of one and the same type, while the differences between them, and the absence or presence of certain forms of letters, are taken as guides for approximately dating the books, and for subdividing the type, hitherto known as the 36-line Bible or Gutenberg type, into three or more varieties. For instance, Schwenke (Centralbl., 1908, p. 74) explains that " the types b, c, i, s, t enable us to distinguish the earliest from the later elements in the Donatus type; the ` Weltgericht ' shows, at least of i and s, the old forms still unmixed. But in the Paris Donatus, the new forms appear by the side of the old forms, though the latter are already to a great extent superseded. The new (Heiligenstadt) Donatus comes between these two works; it has chiefly the old b, which begins to a great extent to be absent in the Paris Donatus." As we cannot regard types which differ in form as " developments " of one type, we must deal with three types in column A, that is (1) the so-called Donatus type; (2) the Kalendar type; (3) the 36-line Bible type, besides the two employed for the Indulgence 31. Gutenberg's career, and the straightened circumstances in which he appears to have lived, so far as they are known to us, make it difficult to ascribe them all to him.

More than thirty documents have come to light which enable us to trace Johan Gutenberg from 1420 to 1468. Dr Carl Schorbach has published nearly all their texts, with elaborate explanations, in the Festschrift zum goo jdhr. Geburtstage von J. Gutenberg (suppl. to Centralbl. f. Biblioth., 1900, p. 163 sqq.), and they are further explained by Hessels (Gutenberg, was he the Inventor of Printing ? 1886; idem, The so-called Gutenberg Documents, 1911) .

At least six of them are known to be forgeries, among them the " relics " of a printing-press with the date " 1441 " which were accidentally(!) discovered in 1856 in the " Hof zum Jungen " which had always been supposed to have been Gutenberg's first printingoffice at Mainz, but which we now know not to have been the case. Assuming that the Gutenberg mentioned in the remaining documents is no other than Henne (= Hans or Johan) Gensfleischcalled Gutenberg from his mother (whose maiden name was Elsa Wyrich) having lived in the " Hof zum Gutenberg " at Mainz, where he is supposed to have been born about moo - he appears to have lived at Strassburg from 1436 (?) till the 12th of March 1444, in easy and somewhat luxurious circumstances, at least during the first three years, as he was then paying duties for large quantities of wine (about 1924 liter). But this prosperity does not seem to have continued, for on the 17th of November 1442 he borrowed 80 pounds Strassburg denarii (=about 4800 marks) from the Strassburg St Thomas Chapter, a Strassburg citizen, Martin Brechter, being his surety. From the 12th of March 1444 till the 17th of October 1448 there is no trace of him, but on the latter day he again borrowed, this time at Mainz, 150 gold guilders. Both these loans he never redeemed, nor is it known whether he ever paid any interest on his Mainz loan. But the account books of the Thomas Chapter, still preserved in the Strassburg Public Archives, show that the interest of 4 pounds per annum on his loan of 1442 was regularly paid, by him or his surety, till 1457The interest due in the latter year was also paid, but difficulties appear to have occurred before the Chapter received it, as there is an item in their account book for1457-1458of two shillings for expenses, incurred by them for arresting Gutenberg and his surety. In and after 1458 no further payments were made; the Chapter had recourse to law, and made various efforts to arrest the defaulters, but in vain; and in 1474, six years after Gutenberg's death, the debt is no longer recorded in the Chapter's accounts. He can be traced at Mainz from 1450 (when he borrowed money from Fust) till the 21st of June 1457, when he is a witness at the conveyance of property in Bodenheim near Mainz. After this date we hear no more of him until the 17th of January 1465, when the archbishop of Mainz appointed him as his servant and courtier for life on account of the " grateful and willing service which he had rendered to himself and to his Stift, and will and may render in future. " The nature of this " service " is not stated. It has always been supposed that he was then residing at Eltville, the residence of the archbishop, and that he died there about or before the 26th of February 1468, on which day Dr Kunr. Humery received from the archbishop some " printing apparatus which belonged to him, and which he had lent to Gutenberg." But recent researches seem to have shown that Gutenberg remained at Mainz till his death, and was buried there.

Apart from the six forgeries, about which there is no dispute, Bockenheimer, a Mainz magistrate, istrate, explains (Gutenberg-Feier, Mainz, 1900) as forgeries also (I) the document of the 14th of March 1434, which represents Gutenberg as having at Strassburg arrested and released the secretary of Mainz for a debt which this city owed him; (2) a document of 1437 recording a breach of promise case between Gutenberg and a Strassburg lady; (3) the records of a Strassburg lawsuit between Gutenberg and some Strassburg citizens in 1 439; (4) the Helmasperger notarial instrument of the 6th of November 1455, recording a lawsuit of Joh. Fust against J oh. Gutenberg. The last two, and a third dated the 26th of February 1468, mentioned above, are the only documents that can be said to connect Gutenberg with the art of printing. Various external and internal circumstances throw serious doubts on the genuineness of the 1439 documents; but suppose they were genuine, they only show that Gutenberg had been engaged, with other Strassburg citizens, in " polishing stones " and " manufacturing looking-glasses," and promised to give instruction in " new arts." A " press," however, is mentioned, and a clause reports that one of Gutenberg's witnesses, Hans Diinne, a goldsmith, had testified that he had earned nearly 100 guilders from Gutenberg, " merely for that which belonged to printing " (alleine das zu dem trucken gehoret). The document contains nothing to connect Gutenberg with the art of printing, except this line, which has clearly been added (as an afterthought) by a different hand from the one that wrote the two first lines of this witness's testimony, a circumstance which makes the whole document more than suspicious. Several theories, however, as to Gutenberg printing at Strassburg in or before 1439 have been built upon this document, and German bibliographers are even now expressing their hope of finding some day evidence of Gutenberg having printed Donatuses and other works in that town.

As to the notarial instrument of 1455, Bockenheimer suggests that as it contains absurdities which are contradictory to all the legal usages of the time, it may be a forgery of the Faust family, perhaps of Joh. Fr. Faust von Aschaffenburg (who pretended to descend from Joh. Fust, whom he called " Faust "), who appears to have possessed, in or about 1600, an " original " of the instrument. From this " original " are derived all the texts published before 1741. In that year, however, J. D. Kehler (Ehren-Rettung Joh. Guttenberg's, Leipzig) printed the text again from an " original " which is now in the Göttingen University Library (republished by Dziatzko, Beitrage, Berlin, 1889), and is perhaps identical with Faust von Aschaffenburg's " original." Though an analysis of the text brings out various incongruities as to the business relations between Fust and Gutenberg, it is difficult to look upon the Göttingen document as a forgery, and we deal with it here as genuine.

It is dated the 6th of November 1455, and records some of the proceedings in the lawsuit between Johan Fust (q.v.) and Gutenberg, which had taken place on that day in the convent of the Barefooted Friars at Mainz, whereby the former sought to recover from Gutenberg 2026 guilders in repayment of 1600 guilders which he had advanced to him (800 about August 1450, and another 800 about December 1452), with the interest thereon. The document first relates that, on some previous day (not stated), Fust had testified (1) that by a written agreement between them, Gutenberg was to " finish the work " (line 24) with the 800 guilders to be advanced to him at 6%; Fust being unconcerned whether it cost more or less. (2) Gutenberg had not been content with these 800 guilders, and Fust, wishing to please him, advanced him another 800 guilders at 6%. (3) He had himself borrowed this money, and as Gutenberg had never paid any interest, the principal sum and the interest thereon amounted to 2026 guilders (=between 15,000 and 16,000 marks), which he now demanded from him.

(4) On the same occasion Gutenberg had replied that Fust should have furnished him with 800 guilders, wherewith to make his " tools " (or apparatus; Germ. Geczuge), and he should be content with this money, and might devote it to his own use. (5) Such tools should be a pledge to Fust. (6) The latter should also give him (lines 37 to 40) annually 300 guilders for maintenance and furnish workmen's wages, house-rent, parchment, paper, ink, &c. (7) If they did not agree further, he should return Fust his 800 guilders, and his tools should be free; but it was to be well understood that he should finish " such work " (line 41) with the money which Fust had lent him on his pledge, and he hoped that he had not been bound to Fust to spend such 800 guilders on " the work of the books " (line 41). (8) Fust had told him that he did not desire to take interest from him; nor had these 800 guilders all, and at once, come to him in accordance with the agreement. (9) Of the additional 800 guilders he wished to render Fust an account; hence he allowed Fust no interest, nor usury, and hopes not to be legally indebted to him.

We assume, though it is nowhere stated, that these clauses relate to the " printing of books," to be executed by Gutenberg with the money which Fust advanced to him. But as he was already in debt at Strassburg since the 17th of November 1442 (and had to pay annually interest on this debt), and at Mainz since the r7 th of October 1448 (also against interest), it is not surprising that when he contracted this fresh loan in 1450, at the high rate of 6%, he (by not giving any security except tools which he had still to make) practically admitted that he was penniless, and stipulated that Fust should give him also an annual sum for maintenance, and besides furnish workmen's wages, house-rent, parchment, paper, ink, &c., in fact everything required for setting up a printing-office and keeping it going. Fust seems not to have complied with these demands, otherwise he would have mentioned them in his account and at the trial. But he advanced another 800 guilders in December 1452, barely two years after his first advance, merely to please Gutenberg, who had not been satisfied with the first 800.

It is argued that Gutenberg must have been able to show Fust some specimens of his work to induce him to lend him so much money, and we have seen above that German bibliographers attribute to him a poem on the " Weltgericht, " which they date c. 1 4431 444, and the Paris Donatus which they date a little later, both printed, it is said, in the " first phase " of the " Gutenberg type," but showing already some traces of wear and tear; and thirdly, an Astronomical Kalendar (a broadside of 4 leaves) which they ascribe to the end of 1447, and regard as a " masterpiece " printed in a new type, said to be a " development " or " second phase " of the Gutenberg type, which must have been used for several years afterwards, till a fresh or " third phase " was cast of it (for B 36) with the alteration of some of the letters. But if Gutenberg had printed these three works in the years ascribed to them, however small they may be, he must be supposed to have had, from 1443 to 1448, types for printing them, and patrices and matrices for making his types, besides a press and various other tools for printing. Yet the notarial instrument of 1455, if it is genuine, reveals him as borrowing money, not so early as 1443, but so late as 1450, for " preparing his tools," and as having, at the time, nothing to offer his creditor as security except the tools which he still had to make(!). But, says one theory, Gutenberg, intending to print a Bible, and finding the type in his possession too large for it, manufactured a smaller one with the aid of Fust's money, while another theory would have it that he wanted to begin with the printing of a Missal, and for this purpose casted two types, one large and the other smaller. Difficulties, however, arose which induced him to use the smaller type for B 42, which was finished about the beginning of 1 453, and Dziatzko places the type of B 36 also in the year 1453, while Schwenke assigns a life of nearly twenty years (1443-1462) to this type.

If, however, Gutenberg had cast all these types, and printed all these books, and sold them, straight from 1443 to 1450, and from 1450 straight on to, say, 1455, he could not have done this without Fust, his money-lender, becoming aware of it, especially as Fust, for his first advance of 800 guilders, was to have received, as security, the " tools " which Gutenberg had to make before he could begin to print. Yet in 1455, fully five years after Fust had entered into such close financial relations with Gutenberg, he claimed, in spite of what he must have known of Gutenberg's supposed activity, the whole of the money which he had advanced, with interest and compound interest on it. And Gutenberg, instead of pleading on the first day of the trial that he had from 1450 to 1455 printed two large folio Bibles and a considerable number of other books, merely refers to the initial stages of his work, to " tools " to be prepared by him as a future pledge for Fust; he tells the judges that he had expected Fust to supply him with various necessaries for printing and his own existence, without saying whether Fust had complied with his demands or not, and finally declares that he had not felt called upon to devote the first 800 guilders to the " work of the books "; that he was ready to account for the second 800, but did not feel indebted to Fust either for interest or anything else, while, on the second day of the trial, he absented himself, and merely sent two of his workmen to hear what was going on (!). This does not look as if he had performed much from 1450 to 1455, but rather the reverse. Anyhow, if the Helmasperger instrument of November 1455 is not a fabrication, it shows that Gutenberg could not have begun to print before 1450; that in this year, 1450 (about August), when he borrowed money from Fust, he had no property such as a printing-office, presses, types, patrices, matrices, &c., which he must have possessed if he had been printing since 1443, to offer his creditor as security; had not a penny to maintain himself; besides being already in debt at Strassburg since 1442, and at Mainz since 1448.

The remainder of the instrument records the verdict given on the first day of the trial which decided (1) when Gutenberg shall have rendered his account of all receipts and disbursements paid out by him on the " work for the use [or profit] of them both " (i. 49), whatever less 1 money he then has received and taken in above it, that shall be reckoned in the 800 guilders; (2) but if the account should show that Gutenberg had paid out more for Fust than 800 guilders which had not come in their common good [or use] (line 60) Gutenberg shall return it to Fust; (3) and if Fust adduces by oath or by reasonable evidence that he has borrowed the above money on interest, and not lent it of his own money, then Gutenberg shall also pay such interest according to the tenor of the schedule.

The verdict is followed by Fust's sworn declaration regarding the amount of his claim, which he had been ordered to make in Gutenberg's presence, but which he now made in his absence, declaring (4) that he had taken up 1550 guilders which Gutenberg had received and which also had gone on " our common work " (line 60); (5) that he had annually given interest and loss, part of which he still owed; six guilders for every 100 guilders which he had thus taken up; (6) of all that Gutenberg had received of this borrowed money, which has not gone on the " work " of them both, which is found in the account, he claimed from him the interest in accordance with the verdict.

Gutenberg appears not to have produced the account which he was expected (clause 1) to render, as Fust's allusion to an account (in clause 6) must refer to his own account. Hence we know not whether he made any " disbursements. " The " receipts " seem to mean nothing more than the instalments of the first 800 guilders which he acknowledged to have received from Fust, though some authors think that allusion is made to things (printed books or broadsides ?) from which he might have received money by sale or otherwise.

It is to be noticed that Fust speaks here (for the sake of accuracy?) of having taken up 1550 not 1600 guilders, as in his first account. On the whole the wording of the verdict and the sworn declaration is obscure, and open to different interpretations, but it is impossible to ascribe to Gutenberg, on the strength of this document, the manufacture of the types and the printing of all the books in column A above, especially when we have regard to his own inexplicable silence at the trial, when it was incumbent on him for his own sake to show what he had done with Fust's money, and still more when we have regard to the pecuniary difficulties in which he had been placed at least eight years before he contracted these heavy new loans with Fust. Within the space of two years after the trial he was bankrupt, unable to pay either his loans or the small interest thereon, and might have ended his days in prison if the Strassburg St Thomas Stift had been able to have him arrested.

Certain circumstances point to Albrecht Pfister of Bamberg as the printer of the numbers vii., viii., ix., xviii. and perhaps those that come between them in column A. Even in former years when the church type of the Indulgence 3l (1454) was believed to be identical with that of B 36, it was the general opinion that, though Pfister could not have printed the Indulgence, he had acquired its church type from Gutenberg for printing B36 Now that a closer examination has shown that the type of B36 need not be dated so early as 1454, the known dates of Pfister (1461, 1462) harmonize with the approximate date (1460) of B36. It is admitted that the types of vii., viii. and ix. differ from that 1 The instrument says: " was er dan men gelts dar uber enpfangen. .. hait. " Senckenberg, Kohler, Van der Linde, &c., printed nun for the correct reading men. This latter word has hitherto been interpreted as meaning more (see Dziatzko, Gutenbergfrage, p. 34, note 1; Schorbach, in Festschr. of 1900, p. 259). Zedler (Gutenbergforschungen, p. 65, note) thinks that it is a dialectic by-form of the Mid. H. German mein found in mein-kouf, meinrdt, mein-swern, mein-tdt, and still preserved in the Mod. H. German Meineid; he translates it therefore as " widerrechtlich " (unlawfully). But men is the same as the Mid. Dutch min (see Verdam's MiddelNederl, Woordenb, in voce) = New Netherl. minder, and means less, the only meaning which can give sense to this clause.

of B 36 in the form of certain capitals. But Pfister issued on the 14th of February 1401 at Bamberg, with the B 36 type, an edition of Boner's Edelstein (88 leaves fol., with wood-engravings), and at least eight other works (Hessels, Gutenberg, p. 161, seq.), one of which bears the date 1462, the seven others none.

Most of the copies of the 36-line Bible now known to us were at one time or another preserved in the libraries of Bavaria, and several fragments have been found in monasteries of that country, even in a register of the year 1460 of the abbey of St Michael at Bamberg. Moreover, a transfer or sale of type from Gutenberg to Pfister is contrary to all analogy in the infancy of printing, when every printer started with a type of his own making.

It is alleged that, in consequence of the lawsuit between Gutenberg and Fust, the former was deprived of all tools, &c., The which he had made, or is supposed to have made, Catholicon with the latter's money, and that afterwards a cer Type. tain Dr Homery or Humery, a syndic of Mainz, lent him fresh money to enable him to set up another printingoffice.

This allegation is made on the strength of a letter of obligation (dated Feb. 26, 1468) referred to above, and given by Dr Homery to Adolph, the archbishop of Mainz, by which he acknowledges to have received from the said archbishop " several forms, letters, instruments, implements and other things belonging to the work of printing, which Johan Gutenberg had left after his death, and which had belonged and still did belong to him (Dr Homery)." It is to be observed that Homery, though willing to assist or oblige Gutenberg, had been cautious enough to reserve to himself all rights to this printing apparatus, in somewhat the same way as Fust in 1450 demanded, or was promised, to receive Gutenberg's " tools " as pledge for his advances. The Homery apparatus could hardly have been of large dimensions, seeing that it was readily passed on first from him to Gutenberg, then from the latter to the archbishop and returned again to its owner. But it is presumed that with these types, which appear in the above list as type VIII., Gutenberg had printed (1) Joannis de Balbis, Catholicon of 1460, copies of which exist in the Cambridge University Library, three in the British Museum, two in the Paris Library, in the Spencer collection of the Rylands Library, in the Wolfenbiittel and Mainz libraries, &c.; (2) Matthaeus de Cracovia, Tractatus rationis, 22 leaves, of 30 lines, 4to, three copies of which are in the British Museum, one in the Rylands, one in the Cambridge, two in the Paris Library, &c.; (3 and 4), two editions of Thomas Aquinas, Summa de articulis fidei, in 4to., the first of 13 leaves and 34 lines (two copies of which are in the British Museum, one in the Rylands and one in the Cambridge Library, &c.); the second of 12 leaves and 36 lines (copies in the British Museum and the Paris Library); and (5) an indulgence of 1461 of 15 lines.

We have seen above that on the 17th of January 1465 Adolph II., archbishop of Mainz, had appointed " Johan Gudenberg, his servant and courtier. " It has always been inferred from this that Gutenberg had quitted Mainz and gone to Eltville (Elfeld) to reside at the archbishop's court, and that, his dignity as courtier preventing him from printing himself, he passed the Catholicon types on to Henry Bechtermuncze at Eltville. It seems certain that in 1467 the Catholicon type with some additions (already found in the Indulgence of 1461) was at Eltville near Mainz, in the hands of Henry and Nicholas Bechtermuncze and Wigandus Spyes de Orthenberg, who issued on the 4th of November of that year (vi.) Vocabularius ex quo (a Latin-German vocabulary) in 4to, 166 leaves, 35 lines, the only known copy of which is in the Paris Library, and (vii.) Vocabularius ex quo, 2nd edition, with colophon dated the 5th of June 1469, 4to, 165 leaves, 35 lines, copies of which exist in the Rylands, the Blenheim, and the Paris libraries. It is therefore asked how the Bechtermunczes could have been using the Catholicon type in 1467, if we assume that it was this type to which Homery refers in his letter of obligation as being in his possession. Some, therefore, conclude that the Catholicon and the four other works in the same type were printed at Mainz by Henry Bechtermuncze, who may afterwards have transferred his printing office to Eltville. In that case it is difficult to see what type Homery could refer to, unless it were type II, a close imitation of which, if not the actual type, was used by Nicholas Bechtermuncze at Eltville in printing (March 12, 1472) a 3rd edition of the Vocabularius ex quo, 166 leaves, 35 lines, copies of which are preserved in the Paris and Hamburg libraries, and an edition of Thomas Aquinas, Summa de articulis fidei, 12 leaves, 35 lines (Munich Library).

It would seem, however, that Fust and Schoeffer were the printers and publishers of the Catholicon, and the other three works mentioned above, as the latter advertised them for sale in a list which he printed and circulated in1469-1470(see Konr.

Burger, Buchhiindleranzeigen des 15 Jahrhunderts, Leipzig, 1907, No. 3). Schoeffer may of course have purchased the stock of these books from Gutenberg or acquired it after his death from Homery, but as nothing compels us to attribute the printing of these books to Gutenberg, there is still less reason to deny that Fust and Schoeffer printed them, as the much discussed colophon of the Catholicon is found, almost verbatim, in three books published by them in 1465 and 1467. Hence the numbers i. to vi. are the only ones that could be ascribed to Gutenberg.

Even this number, involving the manufacture of four different types (apart from the alterations in the forms of certain letters which involved the making of new patrices and matrices) would be large for a man who, after having lived in luxury for some years, practically subsisted from 1442 to 1455 on money which he borrowed from various parties and never repaid. But the poem on the " Weltgericht," printed on paper, could scarcely be placed at the head of a list which includes and, but for this poem, begins with vellum printed works. Moreover, as it can hardly be regarded as a specimen of primitive printing, it takes a more natural place by the side of the paper-printed Turkkalendar, Cisianus and Conjunctiones, which all show that printing on paper was beginning to supersede that on vellum. It is asserted that its type is the same as that of the 1451 Donatus, but this is doubtful.

That the Astronomical Kalendar calculates the ephemerides for 1448 is no evidence of its having been printed at the end of 1447, as kalendars of this kind seem to have been printed without any regard to time and circumstances. Some years ago the Cisianus was ascribed to Gutenberg and to the year 1444, because some of the saints and movable feasts mentioned in it were thought to relate to that year. But as the same saints and feasts occur in the same way in Cisianus editions printed long after 1500, this notion was abandoned. The Astronomical Kalendar in question lays down rules for blood-letting at certain times of the year, and was evidently intended to be hung up in houses as guides for this purpose. It is admitted that it contains mistakes if we apply its calculations to 1448, and it has not yet been proved that these rules required a special kalendar for each year in particular. Removing, therefore, Nos. ii. and vi. to somewhat later dates in the list, the Donatus No. iii. and that of 1451 (No. iv.) with another edition (No. v.) of the same school-book remain at the head of the column A, together with the Indulgence 31, as the only works that could be ascribed to Gutenberg. They bring us down to the time (c. 1451) when he, according to the Helmasperger document, may be supposed to have been in a position to exercise the new art of printing.

It is necessary to point out that eight books - (1) Prognostication or Calendar; (2) Hermann de Saldis, Speculum sacerdotum; (3) Tractatus de celebratione missarum; (4) a work in Berman treating of the necessity of councils; (5) Dialogus inter Hugonem Cathonem et Oliverium super libertate ecclesiastica; (6) Sifridus de Arena, Determinatio duarum quaestionum; (7) idem, Responsio ad quatuor quaestiones; (8) Klagspiegel, or New geteutscht Rechtbuch - have been ascribed to Gutenberg on the strength (a) of the date 1460, which was said to be found in a Prognostication in the Darmstadt library, and (b) of a so-called rubrication alleged to be in a copy of the Tractatus de celebratione missarum, in which " Johannes dictus a bono monte " and Johannes Numeister are represented as offering this work on the 19th of June 1463 to the Carthusians at Mainz. But the date in the Prognostication has been falsified from 1482 into 1460, and the rubrication in the Tractatus is a forgery (Hessels, Gutenberg, pp. 107-114). The eight books are now considered to have been printed by Erhard Reuwich.

Apart from these disputed points there is no further difficulty as regards the history of Mainz printing. Fust and Schoeffer worked together from 1457 to 1466, starting in August 1457 with an edition of the Psalterium, printed in large missal types, which, as far as we know, is the first printed book which bears a date, besides the place where it was printed and the name of the printers. It was reprinted with the same types in 1459 (the second printed book with date, place and name of printer), in 1490, and in 1502 (the last work of Schoeffer, who had manufactured its types). In 1459 Fust and Schoeffer also published Gul. Durantus, Rationale divinorum of iciorum, with the small type (usually called Durandus type) with which they continued to print long afterwards. In 1460 they published the Constitutiones of Pope Clement V., the text printed in a type (Clement type) about a third larger than the Durandus. This type was, however, in existence in 1459, as the colophon of the Durandus is printed with it.1 The Invention Controversy. - Now that we have traced the art of printing from the moment (14J4) that it made its 1 See further Bernard, Origine, i. 21 6 seq.

appearance in a perfect state at Mainz, and have seen that none of the particulars known to us of the life and career of Johan Gutenberg, who is alleged to have invented it, nor any of the books said to have been printed by him, afford us any basis for ascribing that honour to him, we will examine what has been said during a period of more than four hundred years on the question of the invention. For this purpose we will gather up into a chronological sequence (a) a few of the most important expressions used by the earliest printers in their colophons, (b) whatever documentary evidence there may be on the subject, and (c) some accounts of the earliest authors on the question. The Roman numerals i., ii., &c., are for the sake of convenient reference.

The earliest 1 testimony (i.) is the notarial instrument, dated the 6th of November 1455, of the lawsuit between Fust and Gutenberg, already mentioned above, which records transactions between the two men from August 1450 to November 1455, Fust speaking of the work and of " our common work "; Gutenberg of " tools " which he wanted to prepare, of " workmen's wages, house-rent, vellum, paper, ink, &c.," of " such work " and of " the work of the books, " whereas the judges speak of " the work to the profit of both " and " their common use. " (ii.) In the first 2 book published with a date (the Mainz Psalter, issued the 14th of August 1457 by Fust and Peter Schoeffer) it is said that it was perfected at Mainz by an " adinventio art ficiosa mprimendi ac caracterizandi absque calami Colophons, ulla exaratione, repeated and varied later, by the same printers in their colophons of the years 1459 to at least 1470. (iii.) In 1460 the colophon of the Catholicon published at Mainz without the printer's name, after stating that " the book was printed at Mainz, the genial city of the renowned German nation, which town God's mercy had deigned to prefer and adorn above the other nations of the earth by such an exalted light of genius and spontaneous gift, " adds that the book was printed and completed " non calami, stili, aut pennae suffragio, sed mira patronarum formarumque concordia, proporcione, et modulo. " This work (which is to be ascribed to Peter Schoeffer) is considered to have been printed by Gutenberg, and the mention of God's mercy, &c., is regarded as an allusion to the invention of printing. The phrase is, however, also found, with some variations, in the Liber sextus Decretalium, in the Summa of Thomas Aquinas, and in the Clementinae, published respectively on the 17th of December 1465, the 6th of March and the 8th of October 1467, by Fust and Schoeffer. (iv;) On the 17th of January 1465 Adolph II., archbishop of Mainz, by a public decree, appointed Gutenberg as his servant in reward for " his services, " but he does not say what kind of " services " he had rendered, nor does he speak of him as the inventor of printing, nor as a printer.

(v.) In the Grammatica rhythmica, published in 1466 by Fust and Schoeffer, the third line of the colophon runs: " Hinc Nazareni sonet oda per ora Johannis, " which was formerly regarded as an allusion to Johann Fust or Johann Gutenberg, but which more probably refers to Johann Brunnen or Fons, the author of the grammar.

(vi.) On the 26th of February 1468 Dr Homery wrote to the archbishop of Mainz the letter quoted above, from which it may be inferred that Gutenberg had been a printer, though nothing is said as to his being the inventor of printing. (vii.) In 1468 Schoeffer reprinted Fons's Grammatica, in the colophon of which it is said: " At Moguntina sum fusus in urbe libellus meque (the book) domus genuit unde caragma venit. " (viii.) Schoeffer published on the 24th of May 1468 the 1st edition of Justiniani Imper. Institutionum juris libri VI., cum glossa. To this were added by way of colophon some verses commencing: " Scema tabernaculi, " &c., in which it is said that (the ornament of the church) Jesus " hos dedit eximios sculpendi in arte magistros ... Quos genuit ambos urbs Moguntina Johannes, librorum insignes prothocaragmaticos," which is regarded as an allusion to Johann Gutenberg and Johann Fust as first or chief printers. (ix.) In the same year (1468) Johannes Andreae, bishop of Aleria, says, in the dedication of his edition of St Jerome's Epistles, published in that year (Dec. 13,) at Rome, to Pope Paul II., that " Germany is to be honoured for ever as having been the inventress of the greatest utilities. Cardinal Cusa wished that the sacred art of printing, which then (under Cardinal Cusa, who died on the rah of August 1464) seemed to have arisen in Germany, were brought to Rome. " (x.) In 1470 Guil. Fichet, in an octastichon inserted in the Paris edition of 1470 of the Letters of Gasparinus of Bergamo, exhorts Paris to take up the almost divine art of writing (printing) which Germany is acquainted with (see below No. xiii.). In the same year Erhard Windsberg writes to the same effect in an epigram inserted in the Epistolae Phalaridis published at Paris about 1470. (xi.) In 1471 Ludov. Carbo, in the dedication of the Letters 1 The earliest would he the records of the Strassburg lawsuit of 1439, in which the word " trucken " is used, but we cannot accept them as genuine.

Earlier is perhaps the Donatus issued by Peter Schoeffer, possibly before 1456, the colophon of which says that it was finished Arte nova imprimendi seu caracterizandi. .. absque calami exaratione (by a new art of printing or making letters ... without the writing of a pen).

of Pliny to Borso, duke of Modena, speaks of the Germans having invented printing; Nicolaus Gupalatinus (Venice, 1471) of a German being the inventor of printing, and Nicolaus Perottus of the art which had lately come from Germany. (xii.) On the 21st of May 1471 Nicolas Jenson published an edition of Quintilian, edited and revised by Ognibene de Lonigo (Omnibonus Leonicenus), who in the preface speaks of its printer as "librariae artis mirabilis inventor, non ut scribantur calamo libri, sed veluti gemma imprimantur, ac prope sigillo, primus omnium ingeniose demonstravit. " (xiii.) About 1472 the first three printers of Paris published Gasparinus Pergamensis's Orthographiae liber, to`which is prefixed (in the copy of the university library of Basel) a letter, dated the 1st of January, from Guillaume Fichet (see above No. x.), prior of the Sorbonne, to Robert Gaguin, in which he says that " it is rumoured that in Germany, ` not far from the city of Mainz,' a certain Johann Gutenberg (Johannes, cui cognomen Bonemontano) first of all invented the art of printing (impressoriam artem), by means of which books are made with letters of metal, not with a reed (as the ancients did), nor with the pen (as is done at present). " (xiv.) On the 14th of July 1 474 Jo h. Philippus de Lignamine published at Rome Chronica summorum pontificum imperatorumque, in which, between two entries, relating one to the 14th of July 1459 and the other to the 1st of October 1459, an undated paragraph is found saying that Jacobus with the surname of Gutenberg of Strassburg and a certain other one named Fustus, " imprimendarum litterarum in membranis cum metallicis formis periti, trecentas cartas quisque eorum per diem facere innotescunt apud Moguntiam Germaine civitatem. " It says the same of Mentelin, and (under 1464) of Conrad Sweynheym, Arnold Pannarts, and Udalricus Gallus. (xv.) On the 23rd of May 1476 Peter Schoeffer issued the 3rd edition of the Institutiones of Justinian, with the same imprint as in the edition of 1468 (see testimony viii.), but with the addition that Mainz is the " impressoriae artis inventrix elimatrixque prima. " (xvi.) In the Fasciculus temporum, issued at Cologne in 1478, it is stated under the year 1457 that the printers of books were multiplied on earth, deriving the origin of their art from Mainz. The earlier editions merely stated that the printers of books were multiplied on earth. (xvii.) In 1483 Matthias Palmer of Pisa, in the Chron. Euseb. published at Venice, stated under the year 1457 that students owe a great debt to Germany, where Johannes Gutenberg zum Jungen, knight of Mainz, invented the art of printing in 1440. (xviii.) In the same year, 1483, Jac. Phil. Foresta of Bergamo, in the Supplementum chronicorum, says under the year 1458 that the art of printing books was first discovered in Germany, according to some by Guthimberg of Strassburg, according to others by Faust (see xiv.), according to others by Nicolas Jenson (see xii.). (xix.) On the 6th Of March 1492 Peter Schoeffer published the Niedersdchsische Chronik of Conrad Botho, saying in the colophon that it was " geprent ... in. .. Mentz, die eyn anefangk is der prentery." (xx.) At the end of 1494 two Heidelberg professors, Adam rom Docu- Wernher and Joh. Herbst, composed some Latin verses in honour of Johannes Gensfleisch (Gutenberg's family Evidence. name turned into the Latin Ansicarus), whom they called "primus librorum impressor " and " impressoriae artis inventor primus. " (xxi.) In 1499 Jacob Wimpheling (born at Schlettstadt 1450, died 1528) published (at Mainz, by P. Friedberg [?]) an Oratio in Memoriam Marsilii abinghen(d. 1396), in which he, on leaf 22 a, praises Joannes Ansicarus in Latin verse for his invention at Mainz. (xxii.) These verses are preceded by a Latin epitaph on Johann Gensfleisch, " artis impressoriae inventor " and " repertor, " written by Adam Gelthus, a relative of Gutenberg, adding that his remains rest in the Franciscan Church at Mainz. (xxiii.) In the same year (1499) Polydore Vergil (De inventoribus rerum, Venice, lib. ii. cap. 7) says that a certain Peter [Schoeffer ?], a German, invented in 1442 the art of printing at Mainz in Germany, as he had heard from the latter's countrymen; this statement was repeated in a Venice edition of 1503. In later editions " Peter " was altered to " Joh. Gutenberg. " (xxiv.) In the same year Koelhoff, printer at Cologne, published Cronica van der hilliger Slat van Coellen, in which on fol. 311 b, the following statements occur: (1) The art of printing was found first of all in Germany at Mainz about the year 1440; (2) from that time till 1450 the art and what belonged to it were investigated; (3) and in 1450, when it was a golden year (jubilee), they began to print, and the first book that they printed was the Bible in Latin, in a large letter, resembling that with which at present missals are printed. (4) Although the art was found at Mainz, as aforesaid, in the manner in which it is generally employed now, yet the first prefiguration was found in Holland from out the Donatuses which were printed there before that time, and from and out of them was taken the beginning of the aforesaid art, and it was found much more masterly and exact (subtilis) than that other manner was, and has become more and more artistic. (5) Omnibonus wrote in a preface to Quintilian, and in some other books, too, that a Walloon 3 These verses were not published at the time, but in the 19th century by F. J. Mone, Quellensamml. der bad. Landesgesch. iii. 163, from the contemporary MS. of Adam Wernher, preserved in the archives of Carlsruhe.---We pass over here a few books which merely say that the invention was made at Mainz: a Chronyk der landen van Overmaas, written by an inhabitant of Beek, near Maastricht, in the 15th century; the Chronycke van Hollandt (Leiden, 1517), &c.

from France, named Nicol. Jenson (see xii.), discovered this art; but that is untrue, for there are those still alive who testify that books were printed at Venice before Nicol. Jenson came there and began to cut and make letters. (6) But the first inventor of printing was a citizen of Mainz, named Junker Johan Gudenburch. (7) From Mainz the art was introduced first of all into Cologne, then into Strassburg, and afterwards into Venice. (8) The origin and progress of the art were told to the writer verbally by Ulrich Zell of Hanau, still printer at Cologne (anno 1499), through whom the said art came to Cologne. (xxv.) In 1501 Jacob Wimpheling (see xxi.), who stated in his Oratio querulosa contra Invasores Sacerdotum, &c. (published at Delft, c. 1495) that chalcography had been invented at Mainz, says on p. 43 of his Germania (Strassburg, Joh, Priiss, 1501), that the invention was made at Strassburg by Johann Gutenberg of Strassburg, and that it was perfected at Mainz. (xxvi.) In 1503 Johann Schoeffer (the son of Peter Schoeffer and the grandson of Johann Fust) published an edition of Hermes Trismegistus, in which he represents himself as one of the most distinguished citizens of Mainz (nobili vrbe maguntina artis impressoriae inventrice illuminatriceque prima), descended from the most fortunate race who invented the art of printing. (xxvii.) In 1504 No Wittig, a canon and the keeper of the seal of the St Victor Cathedral near Mainz (of which Gutenberg had been a lay member), erected in the house " Zum Gutenberg " a memorial stone and an epitaph (missing already in 1700) to Joh. Gutenberg of Mainz, " qui primus omnium litteras aere imprimendas invenit." (xxviii.) In 1505, in the German translation of Livy published by Johann Schoeffer (see xxxii.) the dedication to the emperor Maximilian, probably written by No Wittig (see xxvii.), speaks of Johan Giittenbergk as inventor of printing (1450) and Johan Faust and Peter Schoeffer as improvers and perpetuators of the art. This work was reprinted at least eight times (in 1514, 1523, 1529, 1530, 1 533, 1 55 1, 1 553, 1 557) with the same dedication; but in 1509 the Breviarium Moguntinum says that it was printed at the expense and labour of Johann Schoeffer, whose grandfather (i.e. Johann Fust) was the first inventor and author of the art of printing (see xxvi.). (xxix.) In 1505 Jacob Wimpheling, in his Epithoma Germanorum (Strassburg, 1505), asserts (on leaf xxxviii. b and xxxix. a) that in 1440 Johann Gutenberg of Strassburg invented there the art of printing. And in 1507, in his Catal. episcoporum Argent. (Strassburg, 1507), he says that the art was invented, though in an imperfect manner, by a certain Strassburger, who afterwards went to Mainz and joined others working and trying the same art, where it was, under the guidance of Johann Gensfleisch, perfected in the house " boni montis " (Gutenberg). This he repeated in 1515. (xxx.) About1506-1511Johannes Trithemius wrote his Chronikon of Spanheim, published at Frankfort in 1601, in which he says (p. 366), under the year 1450, that the art of printing books was discovered afresh (a novo) at Mainz by a certain citizen said to be Johan Gutenberg, who, after having spent all his property in accomplishing the new invention, perfected it by the advice and assistance of Johan Fust and others. The first propagator of the new art was, after the inventor, Peter Schoeffer. (xxxi.) In 1515 Johann Schoeffer published Joh. Trithemius's Compendium sive Breviarium historiae Francorum, and said in the colophon that the book was published at Mainz (the first inventress of the art of printing), by him, the grandson of the late Johann Fust, the first author of the said art, who finally from his own genius commenced to excogitate and to investigate the art in 1450, and in 1452 perfected it and commenced printing, assisted by many necessary inventions of Peter Schoeffer von Gernsheim, his servant and adopted son. Johann Fust and Peter Schoeffer kept this art secret, binding all their servants and domestics by oath never to reveal it; but in 1462 it was spread by the same domestics into divers countries. The same statements were repeated in the Breviar. eccles. Mindensis of 1516. (xxxii.) On the 9th of December 1518 the emperor Maximilian accorded to Johann Schoeffer the privilege of printing Livy (1518-1519), saying that " he has learnt and been advised on the faith of worthy testimonies that the ingenious invention of chalcography was effected by the printer's grandfather." Erasmus, in his preface to this book, says that great praise is due to the inventors of the almost divine art of printing, the chief among whom is rumoured or said to be Joan Faust, the grandfather of Joan Scheffer; and Nicolaus Carbachius, in a final notice of the edition, speaks of " Joan Scheffer Chalcographus," whose grandfather first invented and exercised this art in Mainz. (xxxiii.) In 1519 Joh. Thurmayer Aventinus (1474-1534) wrote that " in 1450 Joannes Faustus, a German, a citizen of Mainz, invented a new kind of writing, called chalcography, and completed it in two years; it was kept secret by him and Peter Schoeffer, his son-in-law, but divulged in Germany ten years afterwards by Faust's servant, Johannes Guttenberger, a Strasburger." (xxxiv.) In a pedigree of Lourens Janssoen Coster of Haarlem and his descendants, preserved in the Haarlem Town Library, it is asserted that " he brought the first print into the world 1446." It would seem that an attempt was made, at some time or other, to alter the date 1446 of this document into 1440, otherwise its genuineness is beyond doubt; in its present state it was probably first drawn up about 1559, but its first four divisions including the above statement were evidently copied from some earlier document, xxvii. 17 as they are all written by one hand, in Roman or Karoline minuscules, and, of course, this earlier document may be assumed to have existed long before 1502-1560, the period usually assigned to this pedigree, and to go back to the time of L. J. Coster himself. There is some doubt as to whether the year 1446 refers to Coster's bringing the first print into the world, or to the marriage of his daughter. In the latter case the " first print " must be earlier. (xxxv.) In 1520 Johan Schott, a printer at Strassburg and grandson of Johan Mentelin, the first printer of that town, published an edition of Ptolemy, and printed at the end the arms of his grandfather with the following inscription: " insigne Schottorum Familiae ab Friderico Rom. Imp. III. Joan. Mentelio primo Typographiae Inventori ac suis concessum: Anno Christi 1466." Apart from the assertion that Mentelin was the inventor of printing, we may remark that the emperor Frederick III. raised Mentelin to the rank of a nobleman in 1466 and granted him new arms. (xxxvi.) In 1524 Johan Schoeffer speaks again (at the end of S. Prosperi libellus) of his maternal grandfather Joan " Faust " and his father Peter Schoeffer, citizens of Mainz, who first of all invented and practised metal printing. (xxxvii.) in 1531 No Schoeffer, the son of Johan Schoeffer, speaks of his great grandfather Johan " Faust " having invented chalcography, and " Faust " continues for many years afterwards to be spoken of as the inventor, sometimes in connexion with Peter, once or twice even with No Schoeffer. (xxxviii.) About 1 533 the Neapolitan Mariangelo Accorso, who had resided at the court of Charles V., wrote on the first leaf of a vellum Donatus (in the possession of Aldus Manutius, jun.) that " Joh. Faust of Mainz first discovered the art of printing with metal types which afterwards he made of lead;; his son Peter Schoeffer added much afterwards to polish the said art. This Donatus and Confessionalia were printed first of all in 1450. Faust derived the suggestion from a Donatus printed before in Holland from an engraved block." This statement is found on p. 411 of the Biblioth. Apost. Vaticana of Angelo Roccha (Rome, 1591), who saw the leaf. Some consider its latter part to have been derived from the Cologne Chronicle (xxiv.) and it seems probable that it was a mixture of some of the above testimonies. (xxxix.) In 1536 Johan Schott (see xxxv.) published Historien Handt-Buchlein (Strassburg, 1536), in which on leaf b' and b 2 he says that " Hans Mentlin of Strassburg invented the art, which, through infidelity, was brought to Mainz." On the strength of this and other statements (xxv., xxix., xxxv.) the bicentenary of the Strassburg invention was celebrated there in 1640. (xl.) In 1541 Joh. Arnold (Bergel or) Bergellanus, who had settled as press-reader at Mainz two years previously, published his Encomium chalcographiae (Mainz, in the St Victor Stift, Fr. Behem, 1541, 4to), in which the lawsuit between Fust and Gutenberg (i.) is alluded to for the first time. Bergel had read Tritheim's books (xxx.), in which the invention is ascribed to Johan Gutenberg with two coadjutors, Johann Faust and Peter Schoeffer, which he (Bergel) had heard confirmed in conversations with Mainz citizens; he had also seen some old tools prepared for the work by the originators which were still in existence. Gutenberg invented it in 1450. (xli.) About 1561 Jan van Zuren (born at Haarlem in 1517) and Dirk Volkerts Coornhert (born at Amsterdam in 1522) established a printing-office at Haarlem. Of the former it is alleged that he had compiled a work on the invention of printing, which is presumed to have been lost during the siege of Haarlem in 1573. This work was not publicly mentioned before 1628, when Peter Scriverius published his Laurecranz voor Laurens Coster, in which he says that he had only found the title, preface and introduction, in which Van Zuren contended that the first foundations of the art were laid at Haarlem, and that it afterwards accompanied a foreigner to Mainz. In this introduction he does not mention the name of the inventor, nor a date, but points in indefinite terms to the house of the inventor as still existing. (xlii.) In the same year (1561) Van Zuren and Coornhert published an edition of the Officia Ciceronis, in which the latter, in a dedication to the magistracy of Haarlem, refers to the rumour that the art of printing books was invented first of all at Haarlem, and was brought to Mainz by an unfaithful servant and much improved there. He adds that very old and dignified persons had often told him, not only the family of the inventor, but also his name and surname, and had explained the first crude way of printing, and pointed out to him the house of the first printer. (xliii.) In 1566 Luigi Guicciardini, a Florentine nobleman, who had visited the Netherlands and had resided many years at Antwerp, finished a description of the Netherlands (published in 1567), in which, alluding to Haarlem, he speaks of the invention there according to the assertions of the inhabitants, the evidence of some authors, and other remembrances; the inventor died before the perfection of his art; his servant went to Mainz, where he perfected the art, and hence the report that it was invented there. (xliv.) About 1568 (it is calculated) Hadrianus Junius wrote his Batavia, published at Leiden in 1588, with two prefaces, dated, the one from Leiden, the 6th of January 1575, the other from Delft ad annum salutis 1575. On p. 253 he says: (a) the opinion that the forms of the letters whereby books are printed were first discovered at Mainz is very inveterate, but old and eminent inhabitants of Haarlem had assured him that they had heard from their ancestors (b) that there lived at Haarlem, more than 128 years before, in a decent house then existing, near the market-place, opposite the royal palace, Lourens (son of) Jan, surnamed Coster, who, while walking in the wood near Haarlem, began to shape beechen bark first into figures of letters, by which, reversely impressed one by one on paper, he composed one or two lines to serve as an example for the children of his son-in-law. (c) When this succeeded, he began to contemplate greater things, and first of all invented, assisted by his son-in-law Thomas (son of) Peter, a more gluey and substantial kind of ink (as the ordinary ink was found to blot), with which he printed whole tablets with pictures, with the letters added. (d) Junius had seen books of this kind printed by Coster (the beginnings of his labours) on the rectos of the leaves only, not on both sides; the book was written (in Dutch) by an anonymous author, and entitled Speculum nostrae salutis, in which care was taken that the blank versos could be pasted together, so that the blank pages should not present any unsightliness. (e) Afterwards (Coster) changed the beechen characters into leaden, and the latter again into tin ones. Very ancient wine-pots cast of the remains of these types were still to be seen in the house of Lourens, which was afterwards inhabited by his great-grandson Gerard (son of) Thomas, who had died an old man a few years before. (f) When the new merchandise attracted purchasers everywhere, workmen were added to (Lourens') household, among whom was a certain John (whether, as was suspected, Faust, or another of the same name, Junius did not inquire), who was bound to the work of printing by oath. But, when he thought he knew the art of joining the letters and of casting the types, &c., he stole away, when everybody had gone to church, the whole apparatus of the types and the tools prepared by his master, and hastened to Amsterdam, thence to Cologne, until he arrived at Mainz, where he could remain in safety, and, having opened a workoffice, issued within the space of one year, about 1442, the Doctrinale of Alexander Gallus and the Tracts of Petrus Hispanus, printed with the same types which Lourens had used at Haarlem. (g) Junius recollects that Nicolaas Gaal, his tutor, a man of firm memory and venerable old age; had told him that as a boy he had often heard a certain bookbinder, Cornelis (a man of more than eighty years of age, who had been an under-workman in the same office) narrating the story of the invention (as he had heard it from his master), the polishing and increase of the crude art, &c., and cursing those nights which he had passed, during some months, with the culprit in one bed. (h) The burgomaster Quirinus Talesius admitted to Junius that he had formerly heard nearly the same from the mouth of the same bookbinder.

(xlv.) Natalis Comes, in his Universa historia sui temporis (Venice 1581; the edition of 1572 contains only books 1 to so), lib. xxiv. 521, says that Haarlem is memorable on account of the almost divine invention of printing books first contrived by John Cutenberg in the year 1453; who, when he had invented the rudiments of it, had a rather cunning servant, observant of his master's art, who, after the death (see xliii., xlvi., xlvii.) of Johan went to Mainz and there perfected the art, and hence the report that it was invented in that city. (xlvi.) Geo. Braunius, in the second volume of his Civitates orbis terrarum (Coln, 1575?), says of Haarlem, that in this town and the whole province of Holland, there was a fixed tradition that the art of typography was first invented there. But before it was perfected and brought to light, the inventor died (see xliii., xlv.) and his servant went to Mainz, and made it known there. (xlvii.) Mich. Eyzinger on p. 75 of his Niederldndsche Beschreibung (Coln, 1584) says that the art of printing, as it was then done, with letters and characters on paper or otherwise, was invented by some one at Haarlem, but, on the death of his master (see xliii., xlv., xlvi.), was brought to light in perfection by his servant. (Repeated by Matthias Quadus Pictor Juliacus in Compendium Universi, sive Geographicae narrationes, lib. iii. c. 38, Colon. 1600.) (xlviii.) Chronicon Sublacense, per P. D. Cherubinum Mirtium Trevirensem monachum Sublacensem laboratum anno. .. 1629. A MS. in 4to, on p. 150 of which is read: Non egre ferat, quaeso lector, si inseruero ratione temporis rem non plane ab instituto nostro alienam, nempe laudabile studium monachorum Sublacensium teutonicorum. .. Nempe, quod nobilissima librorum typographia paucis ante annis in inferiori Germania enata est et in lucem producta (with a note by Mirtius: Hollandia A.D. 1 453 in civitate Haarlem per Joannem Cutenbergam, quae tamen ars, postea Moguntiae per dicta inventoris famulum in meliorem redacta fuit excudendi formam). It is supposed that xlv. to xlvii. are derived from Test. xliii., but this seems impossible as regards xlviii.

(xlix.) In 1628 Scriverius in his Laurecranz (see xli.) placed the date of the Haarlem invention as far back as 1428, and mentioned as its inventor Lourens Janszoon, sheriff of Haarlem. He asserts that the art of printing appeared, " not in the manner as it is used now, with letters cast of lead and tin, but a book was cut leaf for leaf on wooden blocks," and the Haarlem inventor was robbed in 1440 by Johan Gutenberg. Scriverius based the date 1428 upon a Hebrew Chronicle compiled by Joseph ben Meir (1496-1575?),1575?), and published in 1554 at Sabionetta by Cornelius Adelkind, where, under the year of the Jewish era 5188 (=1428), the author mentions a book (without giving the title) printed at Venice and seen by him. Scriverius, being convinced that this could only refer to a book printed at Haarlem, applied the entry to a xylographic Biblia pauperum, of which he gave a description, together with several other blockbooks and early printed books.

(1.) In 1639 Boxhorn pushed the date of the Haarlem invention back to 1420, referring, as his authority, to the same Chronicle of Rabbi Joseph. Since that time the date of the Haarlem invention has been variously placed between 1420 and 1430.

Later testimonies are mere repetitions of earlier statements.' We need not discuss the story of Antonio Cambruzzi, who asserted that Pamfilo Castaldi invented printing at Feltre, in Italy, in 1456, and that Fausto Comesburgo, who lived in his house in order to learn the Italian language, learnt the art from him, and brought it to Mainz; the story, however, found so much credence that in 1868 a statue was erected at Feltre in honour of Castaldi. Nor need we speak of Kuttenberg in Bohemia, where John Gutenberg is asserted to have been born and to have found the art of printing. Nor is it necessary to speak of Jean Brito, who printed at Bruges c. 1477-1488, and is asserted to have invented printing there. We may also pass over Johann Fust, later on called Faust (testimonies xiv., xviii., xxvi., xxviii., xxxi., xxxii., xxxiii., xxxviii.), as we know from the Mainz lawsuit of 1455 that he had simply assisted Gutenberg with loans of money. We may also pass over Johann Mentelin of Strassburg (testimonies xxxv., xxxix.), only remarking here that he had already printed a Bible in 1460, and that he is mentioned in Strassburg registers as a chrysographer or gold-writer from 1447 to 1450; but of his whereabouts between 1450 and 1460 there is no record. That he had gone, or had been called, after 1450 by Gutenberg to Mainz has been asserted but not proved, though there is no reason why he should not be one of the two Johannes alluded to as the prothocaragmatici of Mainz in the Justinian of 1468 (testimony viii.). That Nicolas Jenson came to be regarded in certain circles and for a time as the inventor of printing is owing to testimony xii. being misunderstood.

There remain, therefore, to be considered the testimonies which bear on the rival claims of Haarlem and Mainz. So far as we know, the controversy between Germany and Holland was publicly started as early as 1499 by the Cologne Chronicle (testimony xxiv.), that between the two towns mentioned not publicly before 1561 (testimony xli.); while the name of the Haarlem inventor was not mentioned publicly in print earlier than 1588 (testimony xliv.).

The claims of Germany and Mainz, as centred in the person of Johann Gutenberg, have been discussed above while treating of the early printing at Mainz. A few more words about these claims are necessary. Though some of the documents relating to him connect him with the art of printing, they say nothing of him as the inventor of it; nor do any of the books ascribed to him.

The first document that connects him with the art of printing, the notarial instrument of the 6th of November 1455 (testimony i.), says nothing of an invention or a new mode of printing. And yet the occasion was such as to make it almost imperative on Gutenberg to speak of his invention, if he had made any, for he had spent 1600 guilders of Fuses money for making " tools," apparently without printing anything, 2 and was on the point of being robbed by the latter and having taken away from him all that he is supposed to have made and done to give effect to his idea or invention. The next testimony (ii.) i.e. the earliest Mainz books with printed dates (1457 to 5467), shows that the art of printing was not treated as a secret at Mainz; it is openly proclaimed; its importance fully realized and appreciated, but it is distinctly advertised as a " by-invention of printing," and still more distinctly as a " new art of printing "; the public were informed that books were now no longer produced by means of the pen, but by a new art of forming characters and printing. Such advertisements are natural and appropriate if we assume that the new art of printing had recently (say about 1450 to 1455) become known at Mainz, but not when we assume that Gutenberg had been printing there devotional and school books and folio Bibles since 1443. But, though the new art is so distinctly described and advertised, in none of these advertisements is there one word of a " Mainz invention " or an " inventor." In testimony iii. (the Catholicon of 1460) there is an allusion to Mainz being favoured by God, but again not one word about an invention or an inventor. If Gutenberg had printed the Catholicon, it would be incredible that he, who had been wronged and robbed by his. two rivals (Fust and Schoeffer), should join in with them in defining and proclaiming the new art, but never with one word assert his claim to the honour and profit of the invention, if he had made any, and should even omit his name, whereas he saw ' Over a hundred of them have been collected by Ger. Meerman, Origines typogr. ii. 58 seq.

In line 42 Gutenberg distinctly declares that " he hoped he was under no obligation to Fust to devote the first 800 guilders to the work of the books "; and, as Fust, by advancing the second 800 guilders in 1452, had practically become Gutenberg's partner, it seems clear that the former claimed in October or November 1 455, when the trial may be said to have commenced, his money and interest because Gutenberg had as yet not printed anything.

his two rivals never neglect to print their names in full on every book which they published. Those who believe that Gutenberg was the inventor of printing suggest that he kept silent, as otherwise his creditors would have seized his copies and his printing-office. But this explanation cannot be accepted, as we have seen that Gutenberg was practically bankrupt at that time, and prosecuted as a defaulter; and the verbose colophon at the end of a gigantic folio book like the Catholicon, published at a time when there were perhaps not more than three printing offices in the world, would be calculated to draw attention to its printer and his residence, not to conceal him. Testimony v. (1466) can no longer be regarded as having any reference to Gutenberg or the invention of printing; vii. (1468) was formerly thought to mean: " I, the book, am cast (i.e. its types are cast) in the Mainz city, and the house whence the type came (= where the type was invented) produced me." But of late years it has been shown that the author of the book, Johann Fons, was Peter Schoeffer's press-corrector. And, as he no doubt resided in Schoeffer's house, the two lines evidently mean: " I am a little book cast in Mainz, and I was born (=written) in the same house whence the type comes 1 (=where I am printed)." Testimony viii. (also of 1468) speaks of two Johannes (Gutenberg and Fust) as the " prothocaragmatici librorum quos genuit urbs Moguntina." But this means, not that the first printers of books were born at Mainz, but that the two Johannes (born) produced at Mainz were the chief printers of books.

When we now place together the clear documentary testimonies (i. to viii.) of the first fourteen years of printing (1454 to 1468) at M Mainz, we see that they all come from Mainz itself.

Everybody connected with the art when speaking of

does so in the most public and unreserved manner; .

its importance was as fully realized and advertised then as it is now; the German nation is even congratulated on possessing it; there is never any secrecy about it; but from the moment that it begins to be mentioned there (say about 1456) it is called a new art. In the midst of all this publicity, however, the new art which Mainz and Germany possess is never spoken of as having been invented at Mainz or anywhere else in Germany. The supposed Mainz inventor (Gutenberg) even speaks himself on two occasions (certainly in the lawsuit of 1455, and presumably in the Catholicon of 1460) but never says that he made an invention. The archbishop of Mainz, too, speaks publicly of Gutenberg in 1465 (testimony iv.), and rewards him for services, but does not speak of him as the inventor of printing, nor even as a printer. Nor does Dr Homery, in his letter to the archbishop (testimony vi. of 1468), in which he refers to Gutenberg's printing apparatus, call him the inventor of printing.

In 1468 we enter on a new phase in the history of the invention. Even if we set aside testimony viii. as bein g merely local, testimony ix. (1468) speaks of the art of printing as having arisen in Germany. This testimony, however, does not come from Germany, nor. from Mainz, but from Italy, and is supposed to have been inspired by the two German printers who had established a printing-office at Subiaco in 1465, and in 1467 at Rome, and who most likely learned their craft at Mainz.

As the two printers are mentioned in the testimony, and as it does not speak of Gutenberg, nor of Mainz, it is far more likely that it was merely derived from the colophons of Fust and Schoeffer, or from something that Cardinal Cusa had heard during his embassies in Germany. To the Mainz colophons we must also ascribe (a) the two testimonies of 1470 (x.) and (b) the three of 1471 (xi.), all five of which come from France and Italy. At last, in 1472 (testimony xiii.), the invention of printing is ascribed to Gutenberg of Mainz, but as a rumour, and the testimony comes from France. Guil. Fichet of Paris, who gives it, is supposed to have heard the rumour from the three German printers who commenced printing at Paris in 1470. And as two of them had resided, immediately before they came to Paris, in the university of Basel, and are supposed to have learnt their art there, the rumour is traced to " Bertolff von Hanauwe," who appears in the lawsuit of 1455 as Gutenberg's servant and who was printing at Basel in 1468. But it came more likely from information which Fichet obtained from the St Victor Cathedral, near Mainz (of which Gutenberg had been a lay member), as he speaks of the art having been invented " not far from that town." Testimony xiv. (1474) again comes from Italy, from Rome, and was, 'perhaps derived from one of the German printers settled there at that time. It merely speaks of Gutenberg, Fust and Mentelin as printers, but says not a word which even touches upon the invention of the art. In testimony xv. (1476) we have the first definite mention of Mainz as the inventress of the art; it is given as an addition to the Mainz colophon of 1468 (see viii.). In 1478 Mainz is again mentioned in a Cologne testimony (xvi.) which gives evidence of research, as it is an amplification of an earlier one in which Mainz was not mentioned. Germany, Gutenberg and Mainz are again mentioned in the Venetian testimony xvii. (1483), which gives (under the year 1457) for the first time 1440 as the date of the invention. In the same year we have ' Venit (comes), the present not the perfect tense (has come).

two earlier testimonies (xiv. and xii.) worked into one (xviii.), to the effect that printing was invented either by Gutenberg or by Fust or by Jenson. Testimony xix. (1492), which states that printing commenced at Mainz, is practically equivalent to xv. In 1494 and 1499 we have three German testimonies (xx., xxi., xxii.) as to Gutenberg being the inventor of printing; these, however, come, not from Mainz, but from Heidelberg; xxii. is given by a relative of Gutenberg, Adam Gelthus, and, as the latter resided at Heidelberg, it is clear that he was the real source of the other two Heidelberg testimonies (xx. and xxi.). Two years later, when Wimpheling, the author of testimony xxi., had left Heidelberg, he ascribed (xxv.) the invention of printing to Strassburg, though stating that Gutenberg was the inventor. Testimony xxiii. is recorded above to show the confusion that reigned in people's minds about 150o regarding the invention. We must add to these testimonies those of 1504 (xxvii.) and 1505 (xxviii.), which are owing to No Wittig, a canon and the keeper of the seals of the St Victor Cathedral, near Mainz, of which, according to its liber fraternitatis, Gutenberg had been a lay member.

Thus the Helmasperger document, the two Indulgences of 1454 and the 42-line Bible tell us, that in the period from August 1450 to 1456 the art of printing had commenced and been perfected at IVIainz; but not a word is heard as to how it arose, or what its nature was. In the period from 1456 (if we place Schoeffer's 35line Donatus in this year) to 1468 various books were printed at Mainz with colophons in which the art of printing is proclaimed as a by-invention of printing; more especially as a new art; its mechanism is fully described and said to be quite different from the mode of producing books by means of the pen; but, no one says that it was invented at Mainz, or mentions the name of a Mainz inventor. In the period from 1468 to 1505, however, we have (1) several vague statements made in Italy and France as to the art of printing being known or practised or invented in Germany, statements which arose from the books and colophons published at Mainz; (2) one item of rumour in 1472 that Gutenberg invented it near that town; (3) two Mainz statements, of 1476 and 1492, and one Cologne statement, of 1478, that it was invented at Mainz; (4) three German statements, of 1 49 2, 1 494 and 1499, that Gutenberg had invented it; and (5) two Mainz statements, of 1504 and 1505, to the same effect. But it is to be particularly noticed that the statements (2, 4, 5), which speak distinctly of Gutenberg being the inventor, can be clearly traced to the St Victor Cathedral, that is, to Gutenberg himself and one of his relatives.

Seeing then how slender the basis is for the assertion that printing was invented by Gutenberg at Mainz, that even this slender basis was not laid till fourteen years after the art had been fully established and proclaimed in that city, and that it may be traced to Gutenberg himself, we cannot be surprised to find it promptly contradicted, not in Holland, but in Germany itself.

This contradiction was made in 1499 (testimony xxiv.) in a Chronicle published at Cologne. To facilitate the understanding of this testimony it is divided above into eight sections. The first (taken from Hartmann Schedel's Chronicle, 1493), second, sixth, seventh and eighth are no doubt due to the compiler of the Chronicle, and must not be connected with the third, fourth and fifth, which, according to the compiler, are due to Ulrich Zell, a printer at Cologne, who had probably settled there about 1463, and had most likely learnt his art at Mainz, as he called himself " clericus moguntinus." As Zell's testimony leaves to Gutenberg nothing but the honour of having perfected the art, various attempts have been made to explain away this account. As long as no typographically printed Donatus had been found that could be fitted into Zell's account it was argued that he meant a Donatus printed from wooden blocks; and this argument is brought forward even at the present time. But a practical printer like Zell must have been able to express himself to that effect if he had really meant to say so; and, as block-printing was not less practised in Germany than in Holland, we could hardly assume that blockbooks printed in Holland would have inspired the German inventor rather than the same books printed in Germany. That testimony xxxviii. speaks of a Donatus printed from wooden blocks may be ascribed to the notion arising at that time (c. 1533) that block-printing had given rise to typography. It has also been remarked that unless we take Zell to refer to a Donatus xylographically printed in Holland, the passage in the Chronicle would be contradictory, as it says in its first and sixth section that the art of printing was found first of all at Mainz about 1440, by a Mainz citizen, Junker Johan Gudenburch, and then in its fourth that the art had already been found before that time in another place. But if the fourth section is read in accordance with its punctuation in the Chronicle itself, it says clearly that the art was found at Mainz, as aforesaid in the manner in which it is generally employed now, that is, more masterly, more artistic than in the Donatuses printed in Holland. It has further been asserted that Holland in the Chronicle means Flanders; but the _Chronicle is usually correct in geographical matters, and is therefore not likely to have gone astray in this particular case. It has also been suggested that Zell most likely learnt his art in Fust and Schoeffer's office and invented the passage to injure the reputation of Gutenberg, who had been their enemy. Finally it has been said that Zell did not suggest or write the passage at all; but it is hard to see how this can be maintained in face of the compiler's own statement to that effect.

As, therefore, all these suggestions do not weaken or invalidate Zell's testimony, we must see how far it harmonizes with other circumstances and the testimonies xxxiv., Claims. xli. to xlix., which claim the honour of the invention for Haarlem in Holland Testimony xxxiv. (the Pedigree) is sufficiently clear as to the invention of printing at Haarlem, the supposed date and the name of its inventor. Testimonies xli. and xlii., though coming from Haarlem, do not mention the name of the inventor. But xli. is a mere introduction destined for a complete book that seems to have been lost during the siege of Haarlem in 1573 before it was printed; we are, therefore, not justified in saying that Van Zuren did not know the name; xlii. may have omitted the name, because the publication of Van Zuren's work was in contemplation at the time that it was written. That Guicciardini (testimony xliii.) in 1566 did not mention the name of the reputed Haarlem inventor cannot be considered as an indication that it was not known or had not yet been " invented " when he wrote, as his accounts of the cities of the northern Netherlands are all rather meagre and for the most part derived from correspondence. He and other authors coming after him (testimonies xlv. - xlvii.) state that the Haarlem inventor had died before the art was perfected, and that thereupon his servant had brought it to perfection at Mainz. We do not find any such statement in Junius. The latter's account (xliv.), however, gives various particulars as regards the inventor and his invention. He begins by referring to the difficulty of vindicating the honour of the invention for Haarlem on account of the deep-rooted and general opinion that it took place at Mainz. He then mentions that Lourens (son of Jan) surnamed Coster resided at Haarlem " more than 128 years ago," and gives us to understand that in the year indicated by that phrase he invented the art of printing. Junius's book was not published till after his death, in 1588, but its two prefaces are dated 1575 (he died June 16, 1575), hence the number 128 is supposed to go back from the date when he actually wrote his account, which he is calculated to have done about 1568. Thus we get the year 1440 as the supposed date of the Haarlem invention, though, if we based our calculation upon the date of the preface, the year 1446 or 1447 would have to be assumed. But, as Junius adds that Coster's types were stolen by one of his servants, who fled with them to Mainz, and, establishing there a printing-office, printed within a year's time, in 1442, two books, he must, if this latter date is correct, have meant 1440. By testimonies xlix. and 1. we see that in the 17th century the date of the Haarlem invention was first put back as far as 1428, then to 1420; and since then it has usually been regarded as 1420-1423, especially after it was discovered that the Haarlem wood where Coster is said to have cut his wooden letters was destroyed during a siege in 1426.

The researches regarding the reputed Haarlem inventor have hitherto been made in an inadequately scientific manner, and it appears that, after Scriverius (1628) had pushed back, in spite of Junius, the date of the invention to 1420-1428, he and later Dutch authors on the subject mixed up two Haarlem citizens (a) Lourens Janszoon, who never bore the surname Coster: he is proved to have been sheriff, wine merchant and innkeeper from 1404 to 1439, and to have died in the latter year; (b) Lourens Janszoon Coster, authenticated by official documents as a chandler and innkeeper from 1436 to 1483, leaving Haarlem in the latter year. The name of this person and some genealogical particulars known of him seemed to agree with Junius's account and the Coster pedigree.

But recent investigations at Haarlem and elsewhere tend to show that there have been two, if not three, persons of this name living at Haarlem about the same time. Though this superabundance of namesakes shows that van der Linde and those who accepted his conclusions were rather hasty in declaring L. J. Coster to be a myth, it is somewhat perplexing to the historian, and it would seem that the Dutch people prefer to make speculations and guesses on this point, rather than search in some systematic way the original documents and registers from which they draw haphazard extracts. The result of the latest inquiries (so far as they may be called inquiries) is that L. J. Coster, who would agree with Junius's account and the Haarlem Coster pedigree, was a member of a Christmasgild in 1436, is mentioned in the Haarlem registers as a dealer in candles and oil till 1454, and seems to have died before 1460 (see Fruin, De huidige stand van het Costervraagstuk, 1906; Enschede, Laurens Jansz. Coster, 1904); so that his business as printer was probably continued by one of his relatives, and finally broken up about 1481, when the Speculum cuts are in the hands of Veldener.

Junius's account of the Haarlem invention is based on three books: (1) a Dutch edition of the Speculum humanae salvationis; (2) the Doctrinale of Alexander Gallus; and (3) the Tracts of Petrus Hispanus (Pope John XXI.). The first work, he said, was printed by Coster as a first specimen of his art, and it would seem from his words that the tradition believed it to be printed with wooden types; the second and third books, he declares, were printed at Mainz with Coster's types, stolen from him by his workman. Of the Hispanus Tracts no edition answering to Junius's description has as yet come to light. Of the Doctrinale and the Speculum we possess editions which fit into his account, though, of course, it will be impossible to say whether any of the Doctrinale editions were printed at Haarlem or at Mainz. Various editions of the Latin grammar of Aelius Donatus, printed in the same types, link Junius's independent testimony regarding Haarlem and Coster on to that of Ulrich Zell, who declares in the Cologne Chronicle of 1 499 that editions of this school book printed in Holland were the models (prefiguration) for the printing at Mainz, which commenced about 1450.' As the evidence for Haarlem's claims has been obscured by various adverse and not always intelligent criticisms, and no less by imperfect and incorrect descriptions of the books on which they rest, we describe here, from autopsy, the types and books that have always been and still may be, on solid grounds, attributed to Coster, and which, for this reason, we continue to call Costeriana.

The Costeriana. Xylographic Printing. Of the Speculum humanae salvationis, a folio Latin blockbook (that is, an edition printed entirely from wooden blocks) must have been printed several years before 1471, consisting, like the later type-printed Latin editions, of at least 32 sheets = 64 leaves, all printed on one side of the leaf only, alternately on the versos or rectos (therefore 64 printed pages). The sheets were, no doubt, arranged in the same number of quires (a 3 for the preface; bcd7, e s = 29 sheets for the text) as in the later editions; the first leaf was perhaps blank, the preface occupied the leaves 2 to 6, and 58 leaves remained for the 29 chapters of text, each occupying two opposite pages of two columns each. We may further assume that the upper part of each printed page of the text was occupied by one of the woodcuts, which we know from the later editions, and which are divided each into two compartments or scenes by a pillar, with a line or legend below each compartment explaining, in Latin, the subject of the engraving; and that underneath the woodcut was the text, in two columns, corresponding to the two divisions of the engraving above.

This blockbook has already been alluded to above among the Netherlandish blockbooks, but we give here further details, as various circumstances make it clear that it was the work of the same (Haarlem) printer who issued the other editions of the Speculum, together with xhe several incunabula described below, and to whom a Haarlem tradition ascribes the invention of printing.

All the Speculum editions which concern us contain, so far as we know, 29 chapters. But previous to the above blockbook another one of more than 29 chapters (may be 45, like most of the MSS.) must have existed, as may be inferred from Johan Veldener's 4to edition of a Dutch version of the Speculum, published in 1483, in which all the 58 blocks of the old folio editions reappear cut up into 116 halves to suit this smaller edition, besides twelve additional woodcuts for three additional chapters (the 25th, 28th and 29th) not found in any of the old folio editions. As these additional woodcuts appear to be also cut-up halves of six larger blocks, they point to the existence, at some earlier period, of a folio edition (xylochirographic or xylographic?) of at least 32 chapters, at present unknown to us.

Of the blockbook as is here assumed we know now only 10 sheets or 20 leaves, which, in combination with 22 sheets or 44 typographically printed leaves, make up an edition, called, on account of this mixture of xylography and typography, the mixed Latin edition. These twenty xylographic leaves are (counting the 6 leaves of the type-printed preface) 7+20, 8+19, 10+17, 11+16, 12+15, 13+14 (in quire b); 22 +33, 2 3+3 2, 27+28 (in quire c); 52 +61 (in quire e). Copies of this mixed Latin edition still existing: (I) Bodleian Library, Oxford (Douce collection, 205), perfect; (2 and 3) Paris National Library, 2 copies, one perfect, the other wanting the first (blank) leaf; (4) John Rylands Library at Manchester (Spencer collection), wanting the first (blank) leaf; (5) Colonel Geo. Lindsay Holford, London, wanting the first (blank) leaf; (6) British Museum (Grenville collection), wanting the leaves 1 (blank) and 21 (this being supplied in facsimile); (7) Royal Public Library at Hanover, wanting the leaves 19 (xylogr.) and 24 (typ.), but having duplicates of the (xylogr.) leaves 35 and 28; (8) Museum Meerman-Westreenen, the Hague, wanting the leaves I to 36, and portions of the text of leaves 37 and 38; (9) Berlin Royal Library, wanting the leaves t (blank), 58, 59, 62, 63, 64, while in place of the (xylogr.) leaves 52 and 61 it has the same (type-printed) leaves of the second Latin edition; several of the other leaves are bound in a wrong order; (so) Pembroke library at Wilton House, wanting the leaves i to 7 and 64, while the leaves 9 -}- 18 have been supplied from the second (type-printed) Latin edition; (I I) Copy, represented now by the leaves 15 -}- 28, which appear as duplicates in the Hanover copy (above, No. 7); (12) Ottley (Invention of Printing, p. 287) mentions another copy as having belonged to Mr Singer, which wanted three or four leaves, but has since been taken to pieces and dispersed. See further Holtrop, Cat. bibl. reg. Hag. 560; idem, Mon. typ. p. 22 and facs. pls. 20, 21; Bernard, Orig. i. 13 sqq.; Sotheby i. pl. xxxii.; Campbell, Ann. No. 1570 (who wrongly states that the two copies in the Paris National Library belong to the unmixed Latin edition).

Efforts have from time to time been made to account for the unusual mixture of xylography and typography in this one book, and to assign a date to it and the other editions, with the further view of ascertaining the date of their printer, as for him the honour of the invention of printing is claimed. Bernard (1853) was uncertain as to the chronological order to be assigned to the various editions, but, without stating his reasons, concluded that at least six or seven must have been issued, and that the xylographic leaves of the mixed Latin (his edition A), are the remains of a first complete, entirely xylographic edition. As there is a close resemblance between the letters of the xylographic and typographic texts, and both texts agree, with a few exceptions, word for word with the corresponding texts of the other Latin. edition (which, being wholly typographical, is called the Ottley unmixed Latin), Ottley in 1816 concluded that the xylo graphic pages were facsimiles from those of the typographically printed unmixed Latin edition, which the publisher caused to be made after having lost, through some accident in his office, not only those sheets already typographically printed, but also his types. In support of this theory he pointed to some defects or breakages in the pillars, dresses, &c., of the woodcuts of the xylographic pages which he did not find in the same woodcuts in the unmixed Latin edition; so that he thought the latter must be the first edition. Secondly, as the scrolls in the last vignette (Daniel interpreting the handwriting on the wall) are black in the Inglis copy of the unmixed Latin edition, but white in all the copies of the mixed Latin and the other editions, he concluded that the former must have been printed before the woodcutter had cut away the piece of wood which produced the black scroll, which was to him an additional proof that the unmixed Latin edition was the first. These theories were adopted by Sotheby in 1858 and again by Schreiber in two treatises on xylography (in Centralbl., 1895, p. 20 sqq.; in the Gutenberg-Festschrift, Centralbl., 1900, p. 46 sqq.; and in his Manuel de la gravure sur bois, 1902, iv. 114 sqq., vii. pls. 48, 49, viii. pls. 79, 80). The latter author is of opinion that xylography was not employed for the multiplication of books till about 1468-1470, and that about that time printing with movable metal types was almost unknown in the Netherlands. Hence he thinks that the woodcut illustrations in the various editions of the Speculum were printed somewhere in the Netherlands, and the sheets afterwards sent to Germany, most likely to Cologne, for the purpose of having the texts added by typography. These proceedings, he fancies, were successful twice, once with what he calls the first (unmixed) Latin, secondly with the first Dutch edition, but on the third return journey a part of the material of the second (mixed) Latin edition, that is the ten sheets in question, all packed in one parcel, were lost, and the publisher, in a hurry to sell his copies, had these sheets replaced by xylography.

As a careful examination of the mixed Latin and other editions clearly shows their real condition and the order of their issue, we do not discuss Schreiber's improbable theories. As to those of Ottley and Sotheby, some of the lines which they regarded as broken in the copy or copies of the mixed Latin edition which they examined, are intact in other copies of the same edition, so that no reliance can be placed on these defects and breakages, which are clearly due to printing from wooden blocks, a process which admittedly causes more defects in the impressions than printing from types. Of the black and white scrolls we speak below.

It is to be noticed first. of all that the legends underneath the woodcuts are in Latin, so that they were no doubt engraved for a Latin edition. But, unless we take the twenty xylographic leaves as remains of a complete xylographic edition issued (at Haarlem) before the invention of printing, there would be no Latin edition to connect the woodcuts with in the first instance, as the primitive types and workmanship of one, if not two, of the Dutch editions described below show that these must have been printed before the 44 type-printed leaves of the mixed Latin edition, and also before the wholly type-printed unmixed Latin edition, the types of which are new and far better cast.

Incidentally, this fact that the types of the mixed Latin edition are later than those of the Dutch editions disposes also of another theory favoured by some authors, viz. that during the progress of the xylographic edition its printer invented the movable type, and thereupon stopped his xylographic work to complete the book by means of type, so that in this mixed Latin edition we were to see the transition from xylography to typography.

The priority of the xylographic over the typographic leaves is proved by the Pembroke (No. io) and Berlin (No. 9) copies. In the former the third sheet of quire (= the leaves 9 -{- 18 with the figures 5, 6 and 23, 24), the only type-printed sheet in this quire in the other copies (1 to 7 and 9), is not the same as in the other copies, but belongs to the unmixed or second Latin edition.' A somewhat similar but still more important manipulation we observe in the Berlin copy, in which the fourth sheet of quire =the leaves 52 and 61), the only xylographic sheet in this quire in the other copies, is replaced by the corresponding type-printed sheet of the unmixed or second Latin edition.

All this makes it clear that the printer of the Speculum, some time after having become a type-printer instead of a block-printer, replaced gradually (or by one operation), forty-four xylographically printed leaves of his first edition by type-printed leaves, for the purpose of issuing the Latin edition, now known as the mixed Latin edition; then, at a later stage, prepared a new Latin edition, wholly printed in movable type (now known as the unmixed Latin edition), and afterwards used sheets of this latest edition, not only to replace more of his old xylographic sheets (as in the Berlin copy), but even (as in the Pembroke copy) some of the forty-four sheets which he had printed (evidently for no more copies than he calculated to have left of the old xylographic stock), in the first instance, for issuing the mixed Latin edition. We shall see below that he proceeded in a somewhat similar way in completing copies of his Dutch editions.

Hence the sequence of the Latin editions was thus: (1) The xylographic edition of 64 (?) or more leaves in 29 (?) or more chapters, of which we have only 20 leaves remaining, which was issued before the invention of printing with movable types, and was probably preceded in its turn by a xylographic or xylo-chirographic edition of at least 32 or more chapters; (2) another issue of 20 leaves of the preceding edition, in combination with 44 typographic leaves (the mixed Latin edition) printed for the purpose of replacing the corresponding xylographic leaves of the preceding edition, considered unfit for further publication, or discarded for other reasons; (3) the wholly typographically printed edition known as the unmixed Latin edition.

This clear sequence of the Latin makes it easy to explain that of the other editions of the Speculum. Typographic Printing. (Speculum type 1). - First edition of a Dutch translation of the Speculum, with the title Spieghel der menschliker behoudenisse, hitherto called the first, or the unmixed, Dutch edition, or the Dutch edition in one fount of type. First issue entirely printed in type 1.

Judging by this and the third, the editions of the Dutch version of the Speculum must have had the same number of sheets, arranged (woodcuts and text) in the same way, as the mixed and unmixed Latin editions, with the exception of the preface, which required only 2 sheets (=4 leaves). Hence complete copies consist of the quires a 2 (prefatory matter), bcd 7, e 8 = 31 sheets or 62 leaves.

Holtrop, who gives a facsimile of one of its pages (Mon. pl. 22), regarded this edition as the last of all the Speculum editions, because he thought the type to be identical with that employed for the other editions, only here more used up. Bernard, however, saw that it was a different fount, and there can be no doubt that it is; it differs in form and size from Speculum type 2 as well as from type 3, though it has all the characteristics and the family likeness of the two. Most of the letters might even be regarded as identical with those of type 3, if they were not slightly smaller. That it looks old and battered seems to be owing to bad ink having been used for the printing; it was, however, badly engraved and badly cast, for not one line in the book runs straight. For this reason alone this edition is to be placed before the next two, which are printed with a better type, especially the third. There are, however, more reasons for doing this. First of all, leaf 46 (with the figures 83: Semey, and 84: Rex amon) of Lord Pembroke's copy belongs to the 3rd edition (in Speculum type 3), so that the present edition, to which the Pembroke copy belongs, must have existed earlier. It appears from Holtrop's facsimile (Mon. pl. 22) that leaf 46 was duly printed in type 1 like the other leaves of this edition. But the leaf 46, from which he took his, facsimile, is an isolated one which found its way into the Meerman Museum at the Hague, but is wanting in the copy of the Communal Library at Lille. Hence this particular leaf is, perhaps, a cancel meant to be replaced (in the Lille copy) by another one of the 3rd edition, as in the Pembroke copy. The corresponding leaf of this sheet (33, with the figures 57: Cristus fleuit, and 58: Jeremias) is wanting in the Pembroke copy, so that we can obtain no further information. Another reason for placing this edition before the 3rd is found in the Haarlem copy (No. 5), the leaves 24 -{- 27, 25 + 26 of which also belong to the (3rd) edition, and were apparently meant to replace in that copy the corresponding leaves of this edition, which 1 The Pembroke copy has this additional peculiarity that these leaves 9 + 18 consist each of two separate slips, one having the engraving, the other the text, the latter being pasted on to the bottom part of the former slip.


may have been lost, or the stock of which had become exhausted. Similar manipulations we have noticed above, type-printed leaves having been used to replace earlier xylographic leaves, and again below in the 3rd edition leaves of another edition are found.

Hence we must distinguish between at least three issues of this edition; the first, with the whole text printed in Speculum type I; the second, with sheet 46 of the 3rd Dutch edition; the third, with the leaves 24 to 27 of the 3rd Dutch edition. Copies of the first issue: (I) Communal Library at Lille, wanting the leaves 33 and 46 (which latter are probably now in the Meerman Museum at the Hague), and showing several peculiarities; 1 (2) Haarlem Town Library (No. 4), wanting the leaves 2 and 3, besides the woodcuts (figures 7, 8 and 21, 22) belonging to the leaves 8 and 15. The sheets of this copy have all been cut up into halves, mounted on other larger sheets, and so bound in one volume, together with a copy of the Liber Alexandri Magni, printed at Utrecht by Ketelaer and De Leempt, and of Pet. Scriverius' Laurecrans, both mounted in the same way. There is no rubrication. Second issue: (3) Lord Pembroke's copy, which was completed by leaf 46 of the 3rd Dutch edition. Besides wanting the original leaf 46 in type 2, this copy also wants the leaves 3 2, 33, 54 and 55. It shows, moreover, these peculiarities, that on the recto of leaf 7 and the verso of the corresponding leaf 16 (therefore, on the verso of the third sheet of quire b) are illegible sets-off of the texts of two other pages, or, perhaps, they are faulty impressions of the leaves 8 and 15, which, in the Haarlem copies, seem to be reprints. Third issue: (4) Haarlem Town Library (No. 5), wanting the leaves 20 + 31, 21 + 30, 22 + 29, 23 + 28, while its leaves 24 + 27, 25 + 26 belong to the third (formerly called second) Dutch edition (in Speculum type 3). It has, moreover, this peculiarity that the fourth sheet of quire (=the leaves 8 + 15, with the figures 7, 8 and 21, 22), consists of two separate slips of paper, one containing the impression of the engravings, the other that of the text, the latter slip being pasted on the former, while underneath the figures 7 and 8 are still visible the blind impressions of the two top lines (on the corresponding leaf even 3 lines) of the old discarded letterpress. Seeing that the other copy at Haarlem has the text of these leaves, but not their engravings, it would seem that the letterpress had failed, that is, it had been impressed on the paper without its having been inked.

It is clear from all these manipulations in the copies of this edition, that its printer was inexperienced; moreover, considering its defective type, &c., it is necessary to give it precedence to all the other types and to place this edition immediately after the xylographic edition.

(Speculum type 2).-Second (?) edition of the Dutch version of the Speculum, at present only known from one sheet (the 26th) =the two leaves 49 (with the figures 89: X pus crucifixus and 90: Inventores artis) and 60 (with the figures III: Exitus Tone and 112: Lapis reprobatus), that is, the third sheet of quire e, found in all the existing copies of the Dutch edition (in the Speculum type 3), called the mixed Dutch edition, on account of its having these two leaves, printed in a different type, bound up with the others.

The type (on which see Holtrop, Mon. pl. 19, and Ottley, Inquiry, i. 249) used for these two leaves is slightly smaller than the Speculum type 3, and differs from it and from Speculum type 1 in several respects, though there is a great family likeness between all three. We place it before type 3 because the letters ba, be, ha, he, he, ho, pe, pe, ve, &c., are cast in pairs on one body of type, which combinations appear no longer in type 3. Moreover it looks so primitive, uneven and used up that its proper place would almost seem to be before Speculum type I, although the latter's uneven, wobbling condition suggests its priority. Further, its look and " ductus litterarum " bear such a singular likeness to the Valla type (mentioned below) that it seems reasonable to place it as near to that type as possible. Under ordinary circumstances these two leaves might be regarded 1 The fourth sheet of quire b (leaves 8 and 15) consists of two separate slips of paper, one containing the engravings, the other the text, the latter being pasted on the former. The fifth sheet of quire c (= leaves 23 and 28) is in the same condition. But these slips are not, like the former, pasted one on the other, but the pieces of leaf 23 are pasted on a small, apparently old, slip of paper, another newer piece of paper having been pasted on to the outer margin to strengthen the old piece. The slips of leaf 28 are pasted together by a slip of modern paper on the back, from which it would appear that they had been left loose when the volume was issued. Further the 7th or centre sheet of this same quire c (leaves 25 and 26) is bound wrongly in the place of the first sheet of quire d (leaves 33 and 46), which is wanting in this copy, so that leaf 25 follows after leaf 32, taking the place of the missing leaf 33, while leaf 26 follows after leaf 45, taking the place of the missing leaf 46 (now at the Hague). But on leaf 25 a, which should be blank, is an impression of the text belonging to leaf 62 a, but not of the figures (115, 116), while on leaf 26, which should also be blank, is now the text belonging to leaf 47, but without its figures (85 and 86). Hence the text of these pages (62 a and 47 b) occurs twice in this copy, first on the leaves 25a and 26 b, and secondly in their proper place. These peculiarities seem to show that the letterpress was printed first, and that in this case a mistake was made in the first instance, but discovered when the figures were printed.

as later impressions for completing the edition in which they occur. Ottley and others regarded them as replacing earlier leaves which, by some accident in the printing-office, had got lost or spoiled. But why should a printer use an old, quaint-looking type for printing and reprinting, with differences, one sheet for a book which he had printed entirely with a new and better type employed for many other works ? We rather assume that the leaves are the remains of a complete (the second Dutch ?) edition in Speculum type 2 and were used on this occasion as substitutes for the two corresponding leaves of the third edition, which had become defective or momentarily unavailable.

Differences in the text of the second column of leaf 60 between Meerman's copy and the Spencer Rylands and (Enschede) Crawford copies (see Meerman, Origg. typ. i. 121, note cl., and facs. on pl. vi. 3rd div.; also Holtrop Mon. pl. 19, sec. col.) point to another edition printed in this same type. We therefore distinguish between one edition represented by the Meerman-Westreenen copy, and another represented by the two other copies, without being able to say which of the two is the earlier.

No other trace of this type has hitherto been found, but as it looks old and used up, it seems reasonable to suggest that it must have been employed not only for printing one or more editions of the Speculum, but for other books not yet known to us. It bears a singular likeness to the Valla type mentioned below, and some of the capitals seem almost identical.

(Speculum type 3). (I) The (second, or third, but) first typeprinted Latin edition of the Speculum, or rather of 22 of its sheets (= 44 leaves), printed on one side only, in a type which is newer, and therefore later than the above types 1 and 2, and, for that reason, here called Speculum type 3. It has hitherto been called the Speculum type, as it was thought that all the editions of the Speculum were printed in one and the same type; type I being considered identical with 3, while of type 2, regarded as a stray one, no account was taken. The 22 type-printed sheets of this edition are only found in combination with the 10 sheets (zo leaves) printed entirely (figures and text) from wooden blocks, described above; and the edition so made up is, on account of this mixture of xylography and typography, called the mixed Latin edition. The type-printed leaves are I (blank) + 6, 2 + 5, 3 + 4 (quire a, preface); 9 + 18 (of quire b); 21 + 34, 2 4 + 3 1, 2 5 + 30, 26 + 29 (of quire c); the whole quire d (leaves 35 + 48, 3 6 + 47, 37 + 46, 3 8 + 45, 39 + 44, 40 + 43, 41 + 42); and the leaves 49 + 64, 5 o + 63, 5 1 + 62, 53 + 60, 54 + 59, 55 + 5 8, 5 6 + 57 of quire e. The copies of this edition, still in existence, with all the particulars related to them, have been enumerated above.

(2) The third (hitherto called the second) Dutch edition; also called the mixed Dutch edition, or the Dutch edition in two types, two of its leaves (49 and 60) being printed in a different type (see above, Speculum type 2). This edition is arranged in the same way as the first and second, and consists therefore of 62 leaves. Copies: 1. John Rylands Library, Manchester (Spencer collection), perfect; (2) Lord Crawford's library, perfect; (3), Museum Meerman, the Hague, perfect; (4) Geneva Public Library.

(3) The (third, or fourth, but) second type-printed Latin edition, usually called the unmixed Latin edition, it being printed throughout in one type (3). It contains 64 leaves, printed on one side and arranged in the same number of quires as the mixed Latin edition. But under figure loo (column loo) it has a line (5th) which is wanting in the first (mixed) Latin edition, and the final word of line 4 is correctly printed corporali, not spirituali as in the mixed Latin. Moreover, line 10 in col. 104 has the final word egipti, which is wanting in the mixed Latin, and line 6 in col. 62 has the correct final word terrestris instead of celestis as in the mixed Latin. (See also Holtrop, BRH. 561; Sotheby, i. 145; Bernard, i. 17; Facs. in Holtrop, Mon. pls. 17, 19; Sotheby, i. pls. xxix. and xxx.). Copies: (1) The Hague, Museum Meerman-Westreenen, wanting the first six leaves of the preface. A separate impression of the engraving (Semey mo;ledicit + Rex amon) of leaf 48 is pasted on the lower part of the same cut, which had been printed with the text in the first instance, but defectively (Holtrop, Mon. p. 20, and pl. 17); (2) Florence, Royal National Library, formerly in the Pitti Palace, wanting the first (blank) leaf and having alsoa separate impression of the engraving of leaf 48, but here the text seems to have failed and is pasted on the engraving; (3) Stuttgart, Landesbibliothek, wanting the first (blank) leaf; (4) Munich, Hofbibliothek (pressmark Xyl. 4to No. 37) wanting the first (blank) leaf; (5) Vienna, Hofbibliothek (pressmark Inc. 2 D 19) wanting the first (blank) leaf; (6) [John B. Inglis, bought by Mr Quaritch, and now in] the Lennox library; (7) Haarlem, Town Libra y (No. 8), wanting the preface (leaves 1 to 6); (8) Brussels, Royal Library, wanting, besides the first (blank) leaf, the second and third sheet of quire b (leaves 8 + 19, 9 + 18), and the second half of the fourth sheet (leaf 31) of quire c; (9) Hanover, Royal Library, wanting 18 leaves, that is, the first four, and the whole quire d (leaves 34 to 48); (Io) Munich, University Library (pressmark Xyl. to), wanting the four leaves I (blank), 54, 55 and 59. In this copy Schreiber (Centralbl. 1895, p. 208) discovered the date 1471, in old arabic numerals in rubrics, underneath the blind impression of some line after the last line of the Prohemium. The date is repeated by a hand of the 18th century, in modern arabic numerals, underneath the old date, by way of explanation; (i 1) Library of the Royal Gymnasium at Freiberg in Sachsen, where the 14 leaves of quire c are said to be preserved, but which in June 1908 could not be found.

In the Florence, Munich (University Library), Vienna and (Inglis) Lennox copies, all four belonging to this (unmixed) Latin edition, the three scrolls on the last vignette of the book (over col. 116), representing Daniel before Belshazzar, and the " handwriting on the wall," appear black (see Sotheby, Principia typogr. i. p1. xxx., xxxvii., xxxviii.), but blank in all other copies of this and the other editions. From this fact some authors have concluded that the unmixed Latin edition, here called the last, was, in reality, the first, as the black scrolls show that the pieces of wood which caused these black impressions had not yet been cut away when the copies were printed off. But as its type and other circumstances connected with this unmixed Latin edition make it impossible to regard it as the first, we have to look for another explanation of these black scrolls. First of all, scrolls, especially scrolls proceeding from the mouth of some individual, were already common in the pictures or illustrations of the manuscriptand block-printing periods, just as they are now. They were then, as they are now, intended in all cases to convey to the reader some memorable saying, quotation, inscription or motto. As black scrolls, therefore, could have had no object, we should have to assume that the practised engraver of the Speculum had prepared this last engraving carelessly and only saw his mistake after some copies had been printed off, which yet he allowed to pass into circulation. In some copies the Bible words Mane thecel phares have been written in the blank scrolls, as was to be expected; other copies vary this by adding the Latin interpretations, numerus, appensio, divisio. But in one of the Haarlem copies the scrolls have been coloured yellow with a brush, and it would seem that to some such operation the black scrolls are due; the colour in none of the impressions looks exactly like that of the vignette. It is, however, more than probable that, for some purpose or another, some of these scrolls were intended to be black, and that, while they were printed, something was placed in the block in the hollow of the scrolls to produce a black impression.

Sotheby, in his Principia typogr. p. 178 sqq., calls attention to an imitation of this Speculum vignette by Jacobus de Breda, who began printing at Deventer about 1483. This imitation (having one scroll which proceeds from the mouth of a figure supposed to represent Jacobus himself) he used for the first (?) time in Matthaei Bossi Sermo, c. 1491, the scroll being blank. But when he uses the engraving for the second (?) time, in P. Ovidi Naso. metamor. Liber Secundus, c. 1493 (copy in the Cambridge University Library), his name, " Jacob' de Breda " appears in the scroll (upside down when reading from right to left). A third time the vignette appears in his edition of Pub. Ov. Nas. Metamorphoseos lib. tertius (copy in the Cambridge Library) with his name in the ordinary way. A fourth time it is on the title-page of Seneca de quattuor virtutibus, c. 1495 (also in the Cambridge Library), with the name " Seneca " in the scroll. Sotheby shows that it occurs a fifth time on the titlepage of a Donatus published by J. de Breda, again with his name in the scroll. A sixth time (says Sotheby) the engraving occurs on the title-page of a tract Dominus que Pars, again with his name in the scroll. And finally (says Sotheby) it is on the title-page of Secunda Pars Doctrinalis Alexandri, with the date 1511 and the name " Joanes Bergis " in the scroll. Seeing then the use made of this imitation till 1511, Sotheby, not unreasonabl y, suggests that the original scroll in the Speculum was from the + beginning meant to contain the name of the printer (the inventor of printing). See also Dibdin, Bibliographical Decameron, ii. 285-296. One thing seems certain, the scrolls in the Speculum were not intended to be black in all cases, but to contain something or other, and not always the words Mene, &c., as in that case it would have been as easy for the engraver to cut them on the block as any other words or figures, pillars, &c. The printer probably wished to leave the choice to his purchasers. Incidentally the use made by Jacobus de Breda of his scroll points to his having been aware of the use for which the original scroll, which he imitated, was intended; and as the printer of the Speculum was undoubtedly the first printer of Holland, it is not improbable that Jacobus learnt his craft from him.

The above descriptions and explanations, based on bibliographical and typographical facts, deal exclusively with the editions and issues of editions of the Speculum now known to us. They by themselves make it clear (1) that their printer began as a xylographer and block-printer; (2) that the six editions which he published of this one work cannot be placed later than 1471, as this date is written in a copy of the latest of them; (3) that, for the printing of his five type-printed editions (Dutch and Latin), he manufactured no less than three different types.

When round these editions and types we now group the various other incunabula which must be ascribed to him, as being printed with the same types or others related to them by a striking family likeness and other circumstances, we obtain the following sequence for this printer's work.

A

The Xylographic Period. i. One or two folio editions of the Speculum in Latin, printed (pictures and text) from wooden blocks, and consisting most likely of 32 if not more chapters, but of which only ten sheets (twenty leaves), and six separate woodcuts (cut up into twelve halves, for the Veldener 4to edition of 1483) have come down to us. Of one of these xylographic editions, at least of ten sheets of it, three issues are known to have been made in combination with type-printed leaves (see below).

2. As various circumstances compel us to regard the printer of the Speculum as having been a xylographer before he invented printing with movable types, it is necessary to mention here a small block of wood which is known to have been preserved for nearly 300 years at Haarlem as a remnant of Coster's printing-office. On it is engraved part of an Horarium; its first lines beginning with Servu [m] tuum in pace Quia viderunt oculi mei Salutare, &c.,of the hymn of Simeon. About 1628 it was in the possession of Adriaen Rooman, printer to the Haarlem Corporation, who had obtained it from one of Coster's descendants, a man of great age. Rooman gave it to Dr Johan Vlasveld, of Haarlem, at whose death, in 1684, it came into the hands of his children; in 1734 it was bought by Jan Maas -of Haarlem, who left it at his death to his son-in-law the Rev. Jacobus Mandt, a pastor at Gorinchem; at whose death it was bought by Jacobus Koning, the well known author on the invention of printing, and after his death it was acquired by the Haarlem Town Library where it now is (see A. de Vries, de Uitvinding der Boekdrukkunst, 1862, p. 35).

B

Printing with movable Metal Types. Type I., also called the A becedarium type, with which were printed: (i) The Abecedarium, 4 leaves, 16mo, on vellum, now preserved at Haarlem (Town Library), where M. Joh. Enschede discovered it in 1751, in a MS. Breviarium of the 15th century; (2) An edition of Donatus, 31 lines, 4to, two vellum leaves, printed on one side, discovered in 1844, in the ancient binding of a Dutch Book of Hours, printed at Delft in 1484; it is now preserved in the Hague Royal Library.

Type II. (Speculum type 1; see p. 525; hitherto erroneously regarded as identical with Speculum type 3): (1) First Dutch edition of the Speculum, of 31 paper sheets (62 leaves) printed on one side, folio, hitherto known as the first or unmixed Dutch edition. Two issues: (a), printed entirely in this type, represented by copies at Lille and Haarlem (No. 4); (b), having some of its leaves replaced by leaves of the third Dutch edition, represented by the Pembroke and Haarlem (No. 5) copies. (2) 1 An edition of Donatus, 28 lines, 4to; two vellum leaves in the Haarlem Town Library, found pasted in the original binding of an account book of 1474 of the cathedral of that town, in which an entry testifies that this account-book was bound by Cornelis the bookbinder, whom Junius asserts to have been the servant of Lourens Janszoon Coster (Meerman, Orig. typ. Tab. VI.*). (3) Another Donatus of 28 lines, two leaves of which are in the Haarlem Town Library, and were discovered in the original bindings of account-books of the Haarlem Cathedral Church of 1476, also bound by " Cornelis the bookbinder." Fragments of this same edition are also in the Paris National Library, and in various other public and private collections. (4) Donatus, 28 lines, 4to, one vellum leaf, in the Hague Royal Library (BRH. 2; Ca. 612; Holtrop, Mon. pl. 13), discovered in the binding of a book belonging formerly to the Sion Convent at Cologne containing several treatises printed by Ulrich Zell, one being dated 1467. (5) Donatus, 30 lines, 4to. Two unrubricated vellum leaves in the Cambridge University Library (Inc. 4. E. 1. I), discovered in the binding of a copy of J. Mile's Reportorium, Louvain, 1475, now also in the same library. The first leaf contains the chapters xiv. 11 to xvi. 4, the second chapter xxvi. 6 to xxix. to. The text is abridged, having amabamus, batis, bant, &c., where other editions have amabamus, amabatis, amabant, &c. (6) Donatus, 30 lines, abridged edition, 4to, one unrubricated vellum leaf, cut into halves. Wrongly described by Holtrop (BRH5) and Campbell (614) as part of No. 7 (below). (7) Donatus, 30 lines, 4to; two rubricated vellum leaves, in the Paris National Library (Van Praet, Velins, No. 8; now 1040). (8) Donatus, 30 uneven lines, 4to. Two rubricated vellum leaves, in the Hague Royal Library (BRH 5; Ca. 614). (9) Donatus, 30 lines 4to. Two vellum leaves in the Haarlem Town Library, discovered in 1750 by M. Joh. Enschede at Haarlem in the binding of a MS. (Handvesten. .. van Kennemerland, 1330 to 1477). (to) A liturgical book, containing rules for saying Mass, in 16mo (12 lines to a page [Holtrop, Mon. pl. 14] 2 vellum leaves, pp. 3-6), in the Brussels Royal Library. (11) Alex. Galli Doctrinale, on vellum, 32 lines, The present writer is certain that Speculum type 1 differs from Speculum type 3 in size, and somewhat in form too. But he is still uncertain whether the Donatuses (2 to 7) here enumerated are in the same type as the first Dutch Speculum, though he travelled twice to the places where they are preserved to examine them. It would seem that the Donatuses are in a different type, more compact, regular and better cast than that used for this edition of the Speculum. But if there is any difference between the types it is so minute that it is well nigh impossible to detect it.

4to; one leaf and fragment of a second, in the Ghent University Library (Res. 1409). (12) Alex. Galli Doctrinale, on vellum, 32 lines, 4to. Two leaves (forming one sheet) in the Cologne Town Library.' Type III. (Speculum type 2): (I) Second Dutch edition of the Speculum, which probably consisted of 31 paper sheets (62 leaves) printed on one side in folio, like the first and third. Only known from one sheet (leaves 49 and 60) which forms part of all the copies of the mixed or third Dutch edition preserved to us. (2) On account of differences in the setting up of the second column of leaf 60, another edition in this type may be supposed to have existed. There is no further trace of this type,' which greatly resembles type IV.

Type IV., also called the Valla type: (I) Laur. Vallae Facetiae morales et Franc. Petrarcha de salibus virorum illustrium ac faceciis tractatus, 24 paper leaves, small 4to. No other books printed in this type' are known to exist. But four of its capitals (B, H, L, and M) have been used in printing the edition of the Singularia of Ludovicus (Pontanus) de Roma, which otherwise is entirely printed in type VI.

Type V. (Speculum type 3, hitherto wrongly called The Speculum type): (I) The [second or third, but] first type-printed Latin edition of the Speculum, for which only 22 paper sheets (44 leaves) seem to have been printed to replace the same sheets of the earlier xylographic edition A, and to make up, in combination with the ten remaining xylographic leaves, a folio Latin edition of 64 anopisthographic leaves, called, on account of this mixture of xylography and typography, the mixed Latin edition. Some copies (the Berlin and Pembroke) of this mixed edition were still further mixed with sheets of the second type-printed Latin edition. (2) Third Dutch edition of the Speculum, hitherto known as the mixed Dutch edition, as having two leaves, (49 and 60) printed in a different type (Speculum type 2); 31 paper sheets (62 leaves) printed on one side, folio. (3) A Dutch version of the Seven Penitential Psalms, one vellum sheet (= 2 leaves), 4 pages 16mo, II lines to the page, printed on one side; copies in the Royal Library of Brussels (where it was discovered) and the Hague. (4) An edition of Donatus, of 27 lines, fragments of which are in the British Museum and the Bodleian Library. (5, 6, 7) Three editions of Donatus, of 30 lines, all on vellum (Holtrop, Mon. pl 14b; Meerman, Origg. iv.). (8) A French translation of Donatus, on vellum, 29 or 30 lines to a page; four leaves, now in the Utrecht University Library, discovered by Dr Samuel Muller, the Archivist of Utrecht, in a Utrecht MS. Cartulary of the first half of the 16th century. (9, io) Two different editions of Alexandri Galli Doctrinale on vellum, 32 lines to a page (Holtrop, Mon. pl. 15). (II) Catonis Disticha, imperfect copy of four vellum leaves, 8vo, 21 lines to a page (Holtrop, Mon. pl. 16) in the John Rylands Library (Spencer Collection). (12) The [third or fourth, but] second type-printed Latin edition of 32 sheets (64 leaves), printed entirely in this Speculum type (3), and therefore known as the unmixed Latin edition (Holtrop, Mon., pl. 17). For the use of sheets of this edition to complete copies of the earlier edition, see above V.I. The Munich University Library copy has the rubricator's date 1471.

Type VI., also called the Pontanus type: (I) Ludov. (Pontani) de Roma Singularia juris (in type VI.) and Pii Secundi Tractatus de mulieribus pravis et ejusdem Epitaphia (in type VII.), 60 paper leaves, folio, of which the Pontanus occupies the leaves I (blank) to 45 recto, and the Pius, the leaves 45 verso to the end. Various differences are found in the copies of the Pontanus known to us, and we may assume two if not three issues. This type VI., therefore, is linked on to type VII. by the two being used in one and the same book, while it is inseparably connected with type IV. by the capitals B, H, L and M of this latter type being employed in printing the Singularia. Copies in the British Museum, Cambridge University Library, John Rylands Library (Spencer Collection), Hague Royal Library. (2, 3, 4, 5) Four different editions of Donatus, each of 24 lines, fragments of which are preserved in the Hague Royal Library, Haarlem Town Library, Paris National Library, Cologne Town Library, &c.

Type VII., also called the Saliceto, or the Pii Secundi Tractatus type. (I) Pii Secundi Tractatus et Epitaphia, mentioned above under type VI. as being printed with the Pontanus in one volume. (2) Guil. de Saliceto De salute corporis. Fragments of two vellum leaves of this edition, discovered in the binding of a copy of the Formulae Noviciorum, printed at Haarlem by Joh. Andreae, in 1486, are now in the British Museum. The fragments are printed on one side only, and their texts correspond to the leaves 3 and 5 of It may be that some of the works enumerated under type v. are really printed in the first Speculum type, but it is almost impossible to come to some certainty as to the difference between types i and 3, unless the books are together.

The present writer has recently purchased from Herr Jaques Rosenthal, of Munich, two leaves of a Donatus, which were said (in Herr Rosenthal's catalogue) to be printed in the Valla type (IV.). On examination this proves not to be the case. At first sight it seemed to him to be type III (Speculum type 2). But this is not the case either. It has, however, the peculiarities of both these types combined, so that he does not hesitate to call these fragments a unicum, and its type provisionally type III.* another edition (see below) in the same type, to which treatises of Turrecremata, Pius Secundus, &c. have been added. It is not clear why these fragments were printed on one side only; the versos have not been scraped as was asserted by Holtrop and Campbell, nor are they printer's waste, as they are rubricated. It is not known whether the treatises added to the other edition formed also part of this one. (3) An edition of Donatus minor, or abbreviatus, 26 lines. (4, 5, 6, 7, 8) Five different editions of Donatus of 27 lines. (9) An edition of the Doctrinale, of 26 lines. (io) A Doctrinale of 28 lines. (II) Doctrinale of 29 lines. (12) Doctrinale of 32 lines. (13) Catonis disticha, 21 lines. (14) [Incerti auctoris, vulgo Pindari Thebani] Iliados Homericae Epitome abbreviatum (metrice), cum praefatione Pii II. in laudem Homeri, in folio, io leaves (first blank), 35 lines; first edition having, on fol. 9a, as last line 35: " intecio homeri in precedeti poemate est describere," as in the copy in the Cambridge University Library (Inc. 3 E. 1. I). (15) Guil. de Saliceto De salute corporis; De Turrecremata De salute corporis; Pii II. Tractatus de amore; (Pindari) Iliados Homericae epitome abbreviatum, cum praefatione Pii II.; added are three additional pages, the first contains " Hectoris ... Epitaphium "; the second " Homonee ... Epitaph."; the third is blank. In folio, 24 leaves (first blank), divided into two quires of six sheets each; 34, 35 and 36 lines (second edition of the Saliceto, and of the Yliada; but first of the Turrecremata, and the Tract. de amore of Pius II.). This edition is represented by the copy in the Hague Museum Meerman, in which a MS. note records that it was bought between 1471 and 1474 (Campbell Ann. 1493), which still has in the Yliada: " in precedeti poemate est describere." (16) Second edition of the Yliada, having as last line (35) on folio 9a a more correct reading: " intecio homeri in hoc opere est describere troiana." This edition is represented by the British Museum copy (pressm. 8814) and the three additional pages (3rd blank already found in No. 15) " Hectoris.. Epitaphium " and " Homonee.. Epitaph." (17) Another edition of No. 15 (that is third edition of the Saliceto, second of the Turrecremata and Tract. de amore of Pius II., third of the Yliada and third of the additional pages), but with the line in the Yliada (on 22a): " intecio ... troiana." This edition is represented by the British Museum copy (C. 14. b io). (18) Another issue of the Saliceto; Turrecremata; Pii Tract. de amore et epitaphia, 26 leaves, with various additions or omissions and differences in the setting up not in the former editions. Copy in the Darmstadt Hof-Bibliothek (S 4705), which has the rubricator's date 1472 written in two places. (19) Another issue of the Yliada with the Pii Tract. de amore et epitaphia, again with additions, omissions and differences in the setting up, not in the Darmstadt copy or in the earlier editions. Represented by 17 loose leaves in the Museum Meerman at the Hague (see Holtrop, Mon. typ., pp. 3 2, 33).

An eighth type, hitherto regarded as a Costerian, is type VI. in Hessels's List of Costeriana (Haarlem not Mentz, p. 31 seq.), where two editions of Donatus in this type are mentioned, one of 26 lines, four leaves of which are in the Catholic Gymnasium at Cologne (Campbell, 629), another of 27 lines, of which leaf i i is in the Museum Meerman at the Hague, some fragments in the Haarlem Town Library and two leaves (formerly in the Weigel Collection) in the British Museum (IA 47028). Holtrop (Mon. typ. pl. 21) and Meerman (Opp. pl. II.) give a facsimile of the type. Campbell, in his Annales (No. 629, 631), referring to pl. 31 of Holtrop's Mon. for a facsimile of both these editions, says that they are printed with the types of the Pii II. Tractatus (the Saliceto type), but that, by the size and form of the P, this edition is distinguished from the other books in this type. Hessels (1. c. p. 24) repeated this; but Campbell's assertion proves to be an error, as the two types differ, in spite of a great likeness between them (the C, F, I and V being almost identical). That of the two Donatuses is an early Gothic, and has some of the characteristics of the Costerian types, as the t with perpendicular stroke to its cross-bar, the marks of contraction connected with the letters above which they appear, but only a few pairs of letters cast on one body, and no r with a curl; so that it seems somewhat later than those mentioned above.

A .ninth type (facsimile in Holtrop's Mon. pl. 32a), hitherto regarded as a Costerian, is No. VII. in Hessels's List (l.c.). It resembles much that of the Saliceto, and has served for a Donatus of 27 lines, fragments of which representing two copies, were found in the binding of a Durandi Rationale, printed at Strassburg, 1493, belonging to the Convent of the Holy Cross, at Uden in North Brabant. This type again bears a great likeness to the Saliceto and also to the above type 8, but it differs from both.

Setting aside for the moment the types viii. and ix. as doubtful Costerians, we must also point out that there is no direct evidence that type i. is connected with the other seven, or that it is the first of them. But it is a primitive one; it has all the characteristics of the Speculum and other Costerian types, and could hardly be placed later than the earliest of them; the Donatus printed with it is printed on one side of the leaf only; it shows, moreover, in other respects that it must be dated before 1470. The Abecedarium printed with the same type, and discovered at Haarlem in a 15th century manuscript belonging to a Haarlem family, looks as the work of an inexperienced printer. The types II., III. and V. (the Speculum types I, 2 and 3) are inseparably connected with each other; they must have been in one and the same office; their workmanship shows that their founder step by step simplified and improved his work, and in what order they are to be placed; the most perfect of them (V.) was in existence not later than 1471 (see above), and the three, together with the xylographic leaves in the mixed Latin Speculum (from which they cannot be separated) take us back to a period which could not possibly be extended beyond 1470, but which may reasonably be said to have begun as early as, say 1440.

Therefore these three types, and the books printed with them in combination with the xylographic leaves, and various circumstances pointing to Haarlem as their birthplace, would alone suffice to support and vindicate the Haarlem claims to the honour of the invention of printing. It could, however, serve no useful purpose to separate the types I., IV., VI. and VII. from those of the Speculum, as they have all a great family likeness and three distinctive peculiarities common among them: (1) a perpendicular stroke to the cross-bar of the t; (2) a small curl attached to the top of the r found in no other Netherlandish type; it goes backward in types 1 and 3; and in type 2 another curl is added to the first, bending to the right again; (3) a minute perpendicular link connecting the marks of contraction with the letters above which they appear (a peculiarity common also in the Dutch MSS. of the time). A copy of the latest issue of the Saliceto, preserved at Darmstadt, printed in type VII., has the rubricator's date 1472 in two places; another book in the same type (in the Meerman Museum) was bought between 1471-1474, and as this type is used for a tract printed together with the Pontanus treatise printed in type VI., and the Pontanus type is supplemented with capitals of type IV. (the Valla type), it follows that these three types (IV., VI. and VII.) must have been in use in one and the same office, and that the latest of them (VII.) cannot be placed later than 1472. Again, it must be said that there is no direct evidence that these three types were used by the printer of the Speculum, but as fragments of Donatuses in the Saliceto type have been found in account-books of the Great Church at Haarlem, all presumably bound by the same Cornelis the bookbinder (the reputed servant of the Haarlem inventor), who also used fragments of Donatuses in the Speculum types, Haarlem may be regarded as their common birthplace.' Hence these seven types may be grouped thus: (a) the Abecedarium type; (b) the three Speculum types; (c) the Valla, Pontanus and Saliceto or Pius types; the (a) group cannot be dated later than 1470; (b) (three types) not later than 1471; (c) (three types) not later than 1472 and perhaps not before 1458.

Here then we have a printer who, before 1472, had manufactured and extensively used at least seven (if not eight or nine) different and primitive looking types; three of the seven must have existed long before 1471, as with them he had printed before that year no less than five folio editions of one book (the Speculum), besides several editions of Donatus and the Doctrinale of Alex. Gallus and other smaller books. This work may be supposed to have extended over a number of years, and before he printed any of these type-printed books he had already engraved, printed and issued at least one large folio blockbook (the Speculum). And yet the catalogues of the present day, which profess to arrange the incunabula chronologically, under their respective countries, towns, printers, types and dates - according to some " historical " or " natural history method " suggested in 1870 by an eminent bibliographer, and intended to show the " development of printing " - assign this primitive Dutch printer, and his primitive types and books, to what is presumed to be their " chronological " place, after the productions of Germany, Italy, Switzerland and France; that is, they are placed in a period when printing presses had been established in nearly every large town of Europe, and the art of printing was already so fully developed and vulgarized, that the books of that period show, on comparison with the Costeriana, that the latter must have preceded them by at least two or three decades.

Apart from this anachronism, the same catalogues assign this printer and his books no longer to Haarlem in North Holland, to which they had always been attributed in conformity with the tradition that printing had been invented in that town and the Speculum and other books printed there; but they locate them at Utrecht, the capital of the province of the same name, although the types of the Costeriana show that they are imitations of the handwritings indigenous to the province of Holland, not to those of Utrecht.

' The Cambridge University Library possesses two sheets of two different editions of Donatus, one (unrubricated) printed in Speculum type 1, the other (rubricated) in the Saliceto type, both found pasted by the binder on the wooden boards of a copy of J. Mile's Reportorium, printed at Louvain in 1475, which is also in the same library.

This bibliographical calamity dates from the year 1870, when Dr Anton Van der Linde published his book The Haarlem Coster Legend. After it had become known to him that for years past the " Lourens Janszoon Coster " mentioned by Junius as the inventor of printing had been confused by some authors with another inhabitant of Haarlem, whose name was " Lourens Janszoon, " but who had never borne the surname " Coster, " he, after an inadequate investigation in the Haarlem archives and elsewhere, professed to prove from documents (1) that the Haarlem tradition was nothing but a " legend, " the kernel of which. was " Jacob Bellaert, " who published in 1483 the first Haarlem book with a date; (2) Lourens Janszoon Coster was a " myth ";. (3) Cornelis the bookbinder, Junius's chief witness for the Haarlem tradition, had been Bellaert's servant, and, telling his story in his second childhood, magnified the first Haarlem printer of 1483 into the first printer of the world; (4) the " Spiegel " and the Donatuses could not have been printed before 1470-1474, &c. As Van der Linde's book was apparently based on documents, it was generally thought to have put an end to the Haarlem claims. It seems to have struck nobody at the time that this Haarlem tradition or legend, if it had originated in or after 1483, could not have been so strangely distorted and altered that, within a few decades, " Jacob Bellaert " its hero, according to Van der Linde, was forgotten, while his " servant, " in his second childhood, substituted for him another person of an entirely different name and of a much earlier period; whose descendants all appear in Haarlem's history, and one of whom records him in a genealogy; who is himself mentioned again and again in the Haarlem registers of the time, but who is finally, in 1870, declared to be a " myth." Nor did it strike anybody at the time that if Cornelis the bookbinder had been Bellaert's servant or binder, and his story of the inventor related to him, and to no other printer, this bookbinder must have used fragments of Bellaert's productions for strengthening his bindings, instead of which he employed fragments of the Costeriana, which are admittedly not printed by Bellaert.

These are two of the many points which might have arrested Van der Linde in his sweeping denunciation of the Haarlem tradition if he had given more attention to the subject. As no reply invalidating the main part of his criticism emanated from Haarlem, Henry Bradshaw, the librarian at Cambridge, who had been studying the Dutch incunabula for some years, accepted Van der Linde's conclusions, and published, in 1871, his List of the founts of type used by printers in Holland in the 15th century, in which he explained that he was compelled to place the printer of the Speculum at Utrecht because " it is there that the cuts of the old folio editions first appear cut up into pieces in a book (Epistelen ende Evangelien) printed by Veldener at that place in 1481. Without further information he would have found it necessary to place the printer of the Speculum last among the Utrecht presses and to affix as his date (before 1481). But as the types of the Yliada (VII.) and of the Ludovicus de Roma (VI.) bear a close resemblance to those of the Speculum, they could not be separated from the latter, and a note in the Hague copy of the Tractatus de salute corporis in the same type VII. makes it clear that it was bought between 1471 and 1474, this was the only date which he could accept, and it compelled him to place the. printer of the Speculum at the head of the Dutch printers, just as the Speculum compelled him to place him at Utrecht." It is clear that Bradshaw's system of classifying the incunabula, so inflexible as regards dates and places of printing, that he would admit any stray statement on these points if it be found in the books themselves, rather than go outside the books for further information, is yet elastic enough to ascribe the Yliada and the Pontanus to the printer of the Speculum, merely on account of a close resemblance between the types of these books. As he knew that the early printers shaped their types according to the handwritings indigenous to the places where they settled, it must have escaped him that in locating the printer of the Speculum at Utrecht, he placed him among printers whose types bore no resemblance to those of the Costeriana. This system, therefore, so rigorous on the one hand and so flexible on the other, can only be applied with safety to books whose country, printer and date are known, not to such as the Costeriana, which have neither date nor printer's name, nor place of printing, and might, therefore, be ascribed to France, Italy, Germany or any other European country, if it were not that some of them were printed in the Dutch vernacular.

As to the Speculum cuts being in Veldener's hands in 1481 (and 1483), various circumstances show (see Holtrop, Mon. p. 110 sqq.) that he could not have possessed them, nor acquired them from other printers at Utrecht, until he used them cut up into halves and already considerably worn out. It is also known that ten years at least before he employed them, the cuts had been used intact as illustrations in a book which could not be ascribed to him. In such cases bibliography is bound to inquire where they could have been so used before ascribing them to the place where they are used in 1481. The statements of the Cologne Chronicle (1499) and of Junius (1568) when examined together with the types and workmanship of the Costeriana give satisfactory answers on this point. The fact that fragments of a French translation of Donatus, printed in Speculum type 3, and of a treatise of Ludov. Pontanus on Canonical Law in the Pontanus type, were discovered at Utrecht, cannot be set against the finding of many more fragments of Donatuses, &c. at Haarlem.

Bradshaw lived to see some result of his system in Campbell's Annales, published in 1874, where all the Costeriana are ascribed to a Prototypographie neerlandaise a Utrecht, and he regretted it. Unhappily, his untimely death prevented him from testing his system more closely; those who adopted it were unable, or considered it unnecessary, to repeat his explanations and reservations, so that the Costeriana are now, in almost every catalogue, placed at Utrecht,' without any sign of doubt or hesitation, though all the particulars connected with them prove that they could not have originated there.

To ascertain the probable date of the Haarlem invention, we have at our disposal: (A) some historic statements and documents, namely (a) two entries of 1446 and 1451 in the Diary of Jean Le Robert (Abbat of Cambray); (b) the Helmasperger Instrument of 1455; (c) Ulrich Zell's account of the invention of printing in the Cologne Chronicle of 1499; (d) the Coster pedigree; and (e) Junius's narrative of the Haarlem tradition; (B) a collection of nearly, if not more than, fifty incunabula, known as Costeriana, the printing of which must have involved the manufacture of seven types, four of which (the Abecedarium, and three Speculum types) cannot be placed later than 1471, the other three (the Valla, Pontanus and Saliceto types) not later than 1472. With these types were printed five folio editions of the Speculum, twenty-three of Donatus, eight of the Doctrinale, besides several other important books.

A. Historic Statements

Junius, saying that Coster invented printing in 1440, and that Johan, who stole his types, printed with them at Mainz in 1442, probably knew, or had heard, nothing more definite about a date than that Coster's types were used at Mainz within a year after the theft. The year 1440 as that of the invention was first mentioned, it seems, in 1483, in testimony xvii.; a second time by the Cologne Chronicle in 1499 (but only as the year in which the art began to be " investigated," whatever that may mean), and again in 1505 and later (testimonies xxix., xxxix.). Junius, therefore, may have derived 1440 not from the Haarlem tradition, nor from the Coster pedigree (which gives 1446, and may imply a still earlier date), but from other sources, and hence fixed the commencement of printing at Mainz in 1442 (first mentioned, it seems, in 1499 by Polyd. Vergil, testimony xxiii.). Be this as it may, the Helmasperger instrument of 1455, if it is genuine, shows that G, uten berg could not have begun printing before the end of 1 45 0, if so early, as in that year, about the middle of 1450. August, he borrowed money for " making his tools," and was then, moreover, destitute of everything necessary for printing, as parchment, paper, even ink. This year 1450 agrees with the date (1451) written in the Paris Donatus, which, on insufficient grounds is considered to be a forgery. It also agrees with Ulr. Zell's statement in the Cologne Chronicle that printing and all that belonged -to it were " investigated " from 1440 to 1450, and that in the latter year they began to print. And it likewise agrees with the testimonies xxviii., xxx., xxxi., xxxiii., xxxviii. and xl. quoted above, which all come from persons who may be supposed to have known something about the date of early Mainz printing, namely, Johan Schoeffer, the son of Peter Schoeffer, Joh. Trithemius (who was personally acquainted with both Peter and Johan Schoeffer), Joh. Thurmayer Aventinus (who lived from 1474 to 1534), Mariangelo Accorso (who wrote c. 1533), while No. xl. is that of Joh. Bergellanus, the first author, so far as we know, who mentioned the lawsuit of 1455, in his Encomium, printed and published in the very St Victor Stift of which Gutenberg had been for some years a laybrother till his death, so that this testimony points to Gutenberg's own version of the " beginning " of Mainz printing.

Therefore the Mainz date 1450, derived from documents and testimonies which cannot be lightly set aside, is much later than the latest date (1446) of the Haarlem claims, and those who accept the Haarlem tradition, as we do, may reasonably conclude that Fust was induced to advance money to Gutenberg about August 1450, not by seeing anything printed by the latter, but by having some ' It is pleasant to be able to record some exceptions. Voullieme and Gunther in their Catalogues still mention Haarlem.

of Coster's types and tools, and a type-printed Donatus, shown and explained to him.

We are, however, now asked to disregard this date 1450 and all documents that indicate, and have hitherto always been relied on as fixing, the beginning of printing at Mainz in that year, and to believe that the Astronomical Kalendar, said to be for 1448, was printed at Mainz in 1447. If this year could be accepted for the printing of this Kalendar, its value would of course be greater than any written or printed statement. It is, however, far from certain, and its assumed date, though not interfering with the Haarlem dates, as it falls after 1446 of the Coster pedigree, is incompatible with the Helmasperger instrument, which shows that so late as August 1450 Gutenberg had not printed anything, and had not even made his apparatus for printing. There remains the Poem on the " Weltgericht," also ascribed to Gutenberg and said to be printed in the same type as the Donatus of 1451, with the exception of certain letters the form of which represents, it is thought, a still earlier stage. Hence the Poem is dated back, apparently for no tangible reason, to 1443-1444, and the Donatus placed between it and the Kalendar, the type of which is said to be a " development " of the Donatus type. This date, which is even more speculative than that assigned to the Kalendar, militates entirely against the Helmasperger instrument; it can hardly be said to go against the Coster pedigree, while it does not interfere with, but rather favours, Junius' dates.

Among the historic statements also come the two entries of the Abbot of Cambray, on folio 161 a of his Diary, preserved in the Archives at Lille, in which he records having bought in January ( 445, o.s. =) 1446 and in 1451, at Bruges and Valenciennes, printed Doctrinalia (on vellum ' and on paper). Even if printing could be said to have begun at Mainz in 1450 or earlier, no Doctrinalia printed there have ever come to light, unless we accept the Haarlem tradition, that those printed with Coster's types were printed there. Hence these entries can only be applied to the Doctrinalia printed in Holland in the same types as the Speculum (on which Junius based the tradition of the Haarlem invention) and the Donatuses which fit into Zell's historic statement (in the Cologne Chronicle of 1499), that the Donatuses printed in Holland were the models for the Mainz printing. Therefore there is no certainty as to any Mainz printing having been done before 1450, and, if the Helmasperger instrument has any value, it is certain that it could not have begun there before that year; Ulrich Zell unreservedly places the printing done in Holland before that of Mainz; Jean Le Robert's statements make it certain that printing was exercised before January 1446; the Coster pedigree fixes no later date than 1446 for the invention at Haarlem; Junius' years (1440-1442) are, perhaps, his own guess. Anyhow, if historic statements and documents have any value, the invention must have been accomplished within the six years from 1440 to 1446 (also indicated by Zell).

B. The Costeriana. - It has been pointed out above that we have nearly 50 Costeriana, for which seven types have been employed, four of which cannot be placed later than 1471, the remaining three not later than 1472; and that with these types five folio editions of the Speculum were printed, 23 of Donatus, 8 of the Doctrinale, besides several other important books. With such an abundance of material, for the greatest part of which we have the year1471-1472as an undoubted terminus ad quem, we need not inquire too anxiously whether Junius placed the invention in 1440, or whether the Haarlem Coster pedigree fixes it at 1446 or earlier. For, by placing intervals either between the seven types or between the several editions of the Speculum, Donatus, Doctrinale, &c., we can easily reach any terminus a quo which may be found to agree with the historic statements explained above. Such intervals, however natural and necessary they may be to arrange the Costeriana in some chronological order, must always be more or less arbitrary, as it is impossible to say whether the editions followed each other within two months or within two or more years, or whether the types became used up within six months or within six, seven or more years. Therefore, only such intervals need be suggested as may show that the Costeriana, or some of them, may reasonably be placed before Mainz dates which are certain (that is c. 1450, derived from the Helmasperger instrument, and 1454, the date of the Indulgences), or speculative (as1443-1444for the " Weltgericht," and 1447 for the Astronomical Kalendar). The first products of the art of printing were intended to be faithful imitations of the manuscripts, and no material deviations from the general plan become observable till about 1473-1477. Nowhere are the features of the MSS. of the 15th century so faithfully imitated as in the productions of the three earliest printingoffices of Coster, Gutenberg (?) and Schoeffer. They are all without 2 The abbot speaks of Doctrinalia " gette " or " jettez en molle," and the phrase is, as Bernard (Origine, i. 97 seq.) shows by eight examples from 1474 (the year when printing is first officially spoken of in France) to 1593, and down to the present day, applied to typographically printed books only; see also Fred. Godefroy, Dictionaire, in voce mole (which he interprets as caractere d'imprimerie, where he gives six quotations showing the same meaning.

The abbot does not mention the word vellum, but states that the Doctrinale which he had bought at Valenciennes was full of mistakes wherefore he had bought one on paper. signatures, without printed initial directors,' without printed catchwords; in short, without any of those characteristics which we see gradually, one after the other, come into almost general use when printing becomes more developed, that is from 1473 (if not earlier) to 1480. Hence a comparison of the Speculum, Donatus and Doctrinale editions, printed in the Speculum and other types, with the Gutenberg and Schoeffer Donatuses and their other books enumerated above, shows that the types, mode of printing and workmanship of all these books stand on nearly the same primitive stage. Yet there is a considerable difference between the productions of the three offices, those of the Haarlem office being more primitive than any of the other two. First of all the types of the Costeriana (which have nothing in common with any of those used in the Netherlands after 1471), show by their t with the perpendicular stroke attached to its cross-bar, the r with a curl, and the signs of contraction connected by a fine link to the tops of the letters over which they stand, that they were manufactured during the MS. and block-printing periods of Holland. Secondly, none of the Costeriana have any hyphens, which, in the Gutenberg and Schoeffer incunabula appear already from the beginning. Thirdly, the five editions of the Speculum are all printed anopisthographically (that is, on one side), the woodcuts at the top of the pages as well as the explanatory text underneath, which would hardly be the case if they had been printed after 1471, when the printing of woodcuts, together with text in movable types, on both sides of the leaf, was no longer a novelty. None of them have any colophon (except such a word as explicit), which would, for a collection of nearly 50 books, be incompatible with a period after 1471, but not with the earlier period of the blockbooks and MSS. Moreover, of the 50 no less than 38 are printed on vellum, which is incompatible with a period after 1471 and even earlier, when printing on paper had become universal, but not with the earlier period of the MSS. Therefore, those who wish to date the Donatuses, ascribed to Gutenberg, before 1450, or before 1447, must not forget that the more primitive editions of the Speculum, Donatus and Doctrinale printed in types I. and II. &c. can also be dated before 1450 or 1447; and when once so much is admitted, there is no reason to reject Zell's statement that the Donatuses printed in Holland served as models to Mainz printing.

In addition to the above considerations, there is the remarkable fact that the chief productions of the three earliest printing-offices are editions of Donatus, all printed on vellum. This fact has become more conspicuous by the discovery in recent years, in various parts of Holland and Germany, of a multitude of fragments of different editions of this schoolbook. Of the Haarlem office we know 23 editions; 13 are ascribed to Gutenberg; 9 we have in the Schoeffer or B 42 type. The production of so many editions, all about the same time in the infancy of printing and in two different places, so widely apart from each other as Haarlem and Mainz, cannot have been an accident or coincidence, but suggests some connexion, some links 2 between the three or more offices that produced them. One link we find in Ulrich Zell's statement that the Donatuses printed in Holland were the models for Mainz printing, another in the Haarlem tradition, as narrated by Junius, that one of Coster's workmen, taking his master's types and tools, went with them to Mainz and settled there as a printer. These two statements go far to explain not only how the art of printing was transferred from Haarlem to Mainz, but how, at the latter place, it was thought expedient to continue the printing of Donatuses begun at Haarlem. Bearing this obvious connexion between the three earliest offices in mind, and also that the books of the printer of the Speculum show that he could not have learnt his art at Mainz or any other place, the only question really is: Can the Costeriana, or some of them, by placing an interval between them, be dated so far back that they may be placed before the certain or speculative dates now attributed to books or broadsides printed at, or ascribed to Mainz. In our former edition, when only 20 Costerian editions of Donatus were known, and no earlier final date than 1474, we suggested an interval of 18 months between each of them, giving about 30 years, from 1474 back to 1445, for the issue of all the Donatuses. We now know 23 editions, and 1472 as final date for the existence of all the types, though, of course, some of the editions may have appeared after this year. Therefore, our interval need not be longer than about 15 months, which makes a. stretch of nearly 29 years from 1472 back to 1443. As to an interval between the types, an eminent type-founder, Dr Ch. Enschede of Haarlem, when dealing with Coster's types (in his treatise Laurens Jansz. Coster de uitvinder van de boekdrukkunst, ' An exception is to be noticed in the Costerian Yliada (see above type VII., no. 14-17) in which on the recto of the second leaf the initial director i is printed.

Schwenke has, to some extent, observed this connexion, and suggested that the texts of the Donatuses should be studied, as the differences between them might show whether those of Mainz were printed from the Haarlem editions or vice versa. Such a study may be useful, but could hardly lead to a definite result, as the types of these schoolbooks, like those of other incunabula, were imitations of the respective handwritings of the places where they were printed, and the texts were no doubt taken from the same MSS. in the first instance, though it is possible that the types were cast for other books and used afterwards for the Donatuses. Haarlem, 1904, p. 28), reminds us of three printers (Eckert van Homberch of Delft, Govaert Bac and Willem Vorsterman, of Antwerp), who used one type all the time that they were printing (which means 23 years for the first and 19 for the second), and declares that we could not possibly put a shorter interval than 6 years between each type. As there are seven Costerian types, such an interval would mean a period of 42 years, from 1472 back to 1430, hence only four and a half years (= 312 years) between each type would suffice to reach the year 1440.

These calculations, however, include the Abecedarium (i.), Volta (v.), Pontanus (vi.) and Saliceto (vii.) types. and, as has been pointed out above there is no absolute proof that these four also belonged to the printer of the Speculum. Types v., vi., and vii. cannot be separated, and two circumstances, mentioned above, make it more than probable that they did belong to him. But the Abecedarium type can be ascribed to the Speculum printer on no other grounds than that it has all the characteristics of the Costerian types; that it is too primitive to be attributed to any later Dutch printer, so far as we know them, and that the Abecedarium printed with it, was discovered at Haarlem in a Dutch MS. which belonged to a Haarlem family.

Hence a computation based on the five Speculum editions (all printed and issued at least before 1471), the 12 editions of Donatus and four editions of the Doctrinale printed in the same types might be more convincing to the opponents of Haarlem's claims. Apart from the final date (1471) for them there is also evidence that the Speculum type 1 existed a considerable time before 1474, as in that year the bookbinder Cornelis used fragments of a Donatus printed in that type in the binding of an account book of the cathedral church at Haarlem. Their types and workmanship, moreover, compel us to place them before the Valla, Pontanus and Saliceto (or Pius) types. The last two, employed together in one book, cannot have been used for this book before 1458, as it bears the name of Pope Pius II., who was not elected till that year, but it is certain that it cannot have been printed after 1472. The Valla type, however, existed before the Pontanus and Saliceto types, as four capitals of the former were used to supply the want of such capitals in the Pontanus type.

If then, as suggested by Enschede, the type-founder, an interval of six years is placed between the three Speculum types, it would mean 18 years, or a period from 1471 back to 1453. A similar number of years we obtain by intervals of 18 months between each of the 12 editions of Donatus printed in type 1. Even this moderate calculation makes it plain that the printer of the Speculum must have begun printing at least about the same time that printing began at Mainz. But we have seen above that this printer did not hesitate to make up complete copies of his books by mixing sheets of a later edition, printed in a different type, with those of an earlier edition, and even mixed type-printed with xylographically printed sheets. A printer so carefully and economically husbanding his stock of sheets is not likely to have printed new editions of his books before the old ones were fully sold off, or to have manufactured new types till his old ones were used up. Moreover, Haarlem, a quiet provincial town, could not have been a favourable market for a rapid sale of books, especially not for books in the vernacular, like the Dutch versions of the Speculum. Hence we should not put too short an interval either between his editions or his types.

As (e.g.) Gerard Leeu 3 printed at Gouda, during the six years 1477 to 1482, 17, mostly bulky, volumes, together consisting of 2968 leaves, or nearly 6000 folio pages, all in one type, we need not hesitate to place at least eight or nine years between each of the three Speculum types, that is together 24 or 27 years from 1471 back to 1447 or 1444. It is true, the types manufactured after, say 1477, may have been more enduring than the earlier types, as being, perhaps, cast of better material and by a more perfect process than those of Coster, but the number of pages printed by the latter with the three Speculum types, barely amounts, so far as we know, to a tenth part (boo pages) of Gerard Leeu's work. Our calculations are, of course, liable to modification or alteration; earlier dates may yet be discovered in the Costeriana or in other documents; more editions of Donatus in the same types may be found, which would shorten the intervals. But we have shown that, without straining chronology, bibliography or typology, the Costeriana can be dated back so as to harmonize with any historical date, Dutch (1440, 1446) or German (1450), known at the present time, or so as to precede even the speculative dates (1447 or 1444) assigned to some Gutenberg products.

There is therefore no reason to discredit Zell's statement in the Cologne Chronicle of 1499, that the Donatuses printed in Holland were anterior to, and the models for, the art Necessity of of printing at Mainz, or that of Hadrianus Junius in anEartier his Batavia, that printing was invented at Haarlem by Lourens Janszoon Coster, and that the Speculum nz. was one of his first productions. The two statements were made independently of each other. But even without 3 These examples might easily be multiplied. Ulr. Zell, for instance, printed more than 80 books in his first type.

them, the existence of a group of nearly fifty primitively printed books of undoubtedly Dutch origin, the printing of which must have taken a number of years before 1471, would suggest serious doubts as to the priority of Mainz printing. Zell's statement is all the more weighty, as it is not one made at random but meant to be a direct contradiction of the vague rumours and statements about an invention of printing at Mainz by Gutenberg, which had gradually crept into print since 1468 in Italy and France, and had found their way back into Germany about 1476, after Mainz and Germany had given the greatest publicity, during twenty-two years, to the existence of the new art in their midst; while all those who might, and would and could, have told the public that the invention had been made at Mainz, if it had come about there, preserved a profound silence on this particular point, even the supposed inventor himself. And, though Zell accords to Mainz and Gutenberg the honour of having " improved " the art and having made it more artistic, he denies to them the honour of having " invented " or " begun " it, and this latter honour was never claimed by that town before 1476. Junius's account, on the other hand, is the embodiment of a local tradition at Haarlem, the first written traces of which we have in a pedigree (testimony xxxiv) of the family of the reputed Haarlem inventor, which, as regards its central part, may have existed at least as early as 1520, whereas its first part may be dated much earlier. His account is indirectly confirmed by the finding of several fragments at Haarlem, all belonging to the groups of books mentioned above, but still more by the discovery of several fragments of the Donatuses printed in the Speculum type and 3, some of which had been used as binder's waste by Cornelis, the bookbinder, the very man whom Junius alleges to have been the servant of Coster.

As the case stands at present, therefore, we have, after careful and impartial examination, no choice but to repeat that the invention of printing with movable metal types took place at Haarlem between the years 1440 and 1446 by Lourens Janszoon Coster.

That the Haarlem inventor of printing was, as we have shown, a block-printer before he printed with movable types, helps us to understand what the tradition, as chronicled by Junius, says of him (Testimony xliv. b) : that he, while walking in the wood near Haarlem, cut some letters in the bark of a tree, and with them, reversely impressed one by one on paper, he composed one or two lines as an example for the children of his son-in-law. Junius does not say it, but clearly implies that, in this way, Coster came to the idea of the movability (the first step in the invention of typography) of the characters which, hitherto, he had been cutting together on one block. He perceived the advantage and utility of such insulated characters, and so the invention of printing with movable types was made. The questions as to whether he continued to print with movable " wooden " types, or even printed books with them, cannot be answered, because no such books or fragments of them have come down to us. Junius's words (Test. xliv. § d) on this point are ambiguous, and no Dutch edition of the Speculum printed, figures and text, from wooden blocks or movable wooden types, is known.

By the middle of the 19th century the claims of Coster and Haarlem had steadily gained ground, owing to the researches of Joh. Enschede (1751), Meerman (1765), Koning (1815), Young Ottley (1816), Bernard (1853), Sotheby (1858) and others. But in 1870 they were wellnigh destroyed by a criticism which afterwards proved to be partly groundless, partly a distortion of facts. At the time, however, it was, without further research, accepted as decisive; the claims were regarded to be a fiction, and a system of classifying the incunabula started with the unfortunate result that Utrecht came to be adopted as the birthplace of the Costeriana and Coster and Haarlem almost obliterated from all our catalogues. Since then many things have come to light, all tending to confirm Haarlem's claims, and showing how unjustifiably they were attacked in 1870. An examination of the incunabula on which they rest is, far from easy or inexpensive, as the books are scattered not only over Europe but now also over America, and therefore not easy of access. We have, however, made it, sufficiently to be able to prove that the claims are based on good grounds. Our evidence, though still circumstantial, is not based on guesses; we assert nothing except on bibliographical or historical grounds; nor do we accept one statement unless it is corroborated by other statements, or by the rules of bibliography and history. Hence we should not accept Zell's evidence or that of Junius, or of any one else, if the books to which they refer did not corroborate them to the fullest extent, or if the claims of Mainz to the honour of the invention could be said to have any substance of fact. The great efforts made in Germany since 1882 to strengthen the case for Gutenberg, which culminated in the celebrations of 1900 and the publication of valuable and learned books, have enriched our knowledge of early Mainz and German printing, but at the same time conclusively shown that it requires great courage to maintain that Gutenberg was the inventor of printing.

How long Coster or his successors continued the first printingoffice at Haarlem we cannot say; it seems to have come to an end in or before 1481, as the cuts of the Speculum had evidently then passed into John Veldener's hands, and the Haarlem tradition says that wine-pots had been cast of the remains of the types. In 1483 Jacob Bellaert was printing at Haarlem, and Jan Andrieszn in 1485; their types are imitations of the writing of their time, but already differ from those of the Speculum and the other Costeriana in various respects, and show many features of a later period. The question as to whether they learnt their craft from the first Haarlem printer, or from other masters, has been asked but not yet answered.

Spread of Typography

Having explained the early printing of Haarlem and Mainz, in so far as it bears upon the controversy as to where and by whom the art of printing was invented, and shown that the testimony of Ulrich Zell (in the Cologne Chronicle of 1499) as to Mainz having learnt the art of printing from Holland through the Donatuses printed there, and that of Hadrianus Junius, as to the tradition of its Haarlem origin, are confirmed by bibliographical and historical facts, we can follow its spread from Haarlem to Mainz, and from the latter place to other towns and countries.

1460; Strassburg. - First printers: Johann Mentelin, who completed a Latin Bible in that year, according to a rubrication in a copy at Freiburg in the Breisgau; Adolph Rusch de Inguilen, who is presumed to be the printer of the undated books with a singularly shaped R,' c. 1464; Henricus Eggestein, 1471; George Husner, &c. 1461; Bamberg. - First printers: Albrecht Pfister, who in 1461 published Boner's Edelstein, though it is still doubtful whether he did not print earlier, while he has always been regarded as the printer of B36 (see above); Joh. Sensenschmidt, c. 1480.

1465; Subiaco. - First and only printers: Conrad Sweynheym and Arnold Pannarts, who completed in that year an edition of Cicero, De Oratore, and Lactantius, and removed to Rome in 1467.1466; Cologne. - Earliest printers: (1) Ulrich Zell, who published in that year Chrysostom, Super Psalmo quinquagesimo liber primus, though it is presumed that he printed already in 1463; (2) Arnold Ther Hoernen, 1470; (3) Johannes Koelhoff of Lubeck, 1470, who printed the Cologne Chronicle in 1 499; (4) Nicolaus Gotz, 1474; (5) Goiswinus Gops, 1 475; (6) Petrus de Olpe, 1476 (not 1470); (7) Conradus Winter of Homburg, 1476; (8) Joh. Guldenschaaf, 1 477; (9) Henricus Quentel, 1479, &c.2 1467; Eltville. - First printers: Nicolas and Henry Bechtermuncze and Wygandus Spyes de Orthenberg, who completed in that year a Vocabularius ex quo. 1467; Rome. - First printers: Conrad Sweynheym and Arnold Pannarts from Subiaco, who published an edition of Cicero's Epistolae ad familiares; Ulrich Hahn or Udalricus Gallus, who issued on the 31st of December 1467 Turrecremata's Meditationes. 1468; Augsburg. - First printer: Gunther Zainer or Zeyner. Same year at Basel (first printer Berthold Rot of Hanau) and at Marienthal (Brothers of the Common Life).

1469; Venice. - Printers: (1) Johannes of Spires; (2) his brother Vindelinus of Spires; (3) Christopher Valdarfer; (4) Nicolas J enson, &c. The further spread of typography is indicated by the following dates: 1470 at Nuremberg (Johan Sensenschmidt, Friedr. Creusner, Anton Koberger, &c.), Berona or Beromunster in Switzerland (Helyas Helye alias De Llouffen), Foligno (Emilianus de Orfinis and Johannes Numeister), Trevi (Johann Reynard), Paris (first printers the three partners Ulrich Gering, Michael Friburger, Martin Krantz); 1471 at Spires, Bologna, Ferrara, Florence, Milan, Naples, Pavia, Treviso, Savigliano (Hans Glim?); 1472 at Esslingen, Cremona, Mantua, Padua, Brescia, Parma, Monreale (Mondovi), Fivizzano, Verona, Iesi, St Ursino (?); 1473 at Lauingen, Ulm (perhaps as early as 1469), Merseburg, Alost, Utrecht, Lyons, Messina, Buda-Pest, Santorso; 1474 at Louvain, Genoa, Como, Savona, Turin, Vicenza, Modena, Valencia; 1475 at Lubeck, Breslau, Blaubeuren, Burgdorf, Trent, Cracow (?), Reggio (in Calabria), Cagli, 1 M. Philippe, Origine de l'imprimerie a Paris, p. 219, mentions two books printed in this type, which contain manuscript notes, to the effect that they were purchased in 1464 and 1467, so that Inguilen is to be placed before Eggestein.

2 Johann Veldener, who is said to have printed at Cologne, was never established there, but at Louvain (1473-1477), Utrecht (1478-1481), and Culenborg or Kuilenburg (1483-1484); see Holtrop, Mon. typ., PP. 4 2, 47, 109.

Caselle or Casale, Pieve (Piove) di Sacco, Perugia, Piacenza, Saragossa; 1476 at Rostock, Bruges, Brussels, Angers, Toulouse, Polliano (Pogliano), Pilsen; 1477 at Reichenstein, Deventer, Gouda, Delft, Westminster, Lucca, Ascoli, Bergamo, Tortosa, Palermo, Seville; 1478 at Oxford, St Maartensdijk, Colle, Schussenried (in Wurtemberg), Eichstadt, Geneva, Vienne, Trogen (?), Chablis, Cosenza, Prague, Barcelona; 1479 at Erfurt, Wurzburg, Nijmegen, Zwolle, Poitiers, Toscolano, Pinerolo, Novi, Lerida, Segorbe, Saluzzo; 1480 at London, St Albans (or in 1479), Oudenarde, Hasselt, Reggio (in Modena), Salamanca, Toledo, Nonantola, Friuli (?), Caen; 1481 at Passau, Leipzig, Magdeburg, Treves, Urach, Casale di San Vaso, Saluzzo, Albi, Antwerp, Rougemont; 1482 at Reutlingen, Memmingen, Metz, Pisa, Aquila, Promentoux, Zamora, Odense, Chartres, Wien, Guadalajara, Munchen, Erfurt; 1483 at Leiden, Kuilenburg (Culenborg), Ghent, Chalons-sur-Marne (?), Gerona, Stockholm, Siena, Soncino, Salins; 1484 at Bois-le-Duc, Eichstatt, Novi, Sangermano, Chambery,. Udine, Winterberg, Klosterneuburg, Rennes, Loudeac, Tarragona; 1485 at Heidelberg, Ratisbon, Pescia, Vercelli, Treguier or Lantreguet, Brunn, Salins, Burgos, Mallorca, Hijar, Palma, Xeres; 5486 at Munster, Stuttgart, Chiavasco, Voghera, Casal Maggiore, Abbeville, Schleswig, Toledo; 1487 at Ingolstadt, Gaeta, Rouen, Murcia, Besancon; 1488 at Stendal, Viterbo, Gradisca, Faro, Constantinople, Lantenac; 1489 at Hagenau, Kuttenberg, San Cucufat (near Barcelona), Portesio, Coria, Pamplona, Tolosa, Lisbon; 1490 at Embrun, Orleans, Grenoble, Dale; 1491 at Hamburg, Kirchheim, Norzano, Goupillieres, Angouleme, Dijon, Narbonne; 1492 at Marienburg, Cluni, Zinna, Valladolid, Leiria; 1493 at Luneburg, Freiburg (in Breisgau), Urbino, Cagliari, Lausanne, Nantes, Copenhagen, Rieka; 1 494 at Oppenheim, Tours, Macon, Monterey, Braga; 1495 at Freisingen, Freiberg (near Leipzig), Scandiano, Forli, Limoges, Schoonhoven (monastery Den Hem), Pamplona, Wadstena, Cettinje; 1496 at Offenburg, Provins, Barco, Valence, Granada; 1497 at Carmagnola, Avignon; 1498 at Tubingen, Perigueux, Schiedam, Gripsholm; 1 499 at Danzig, Olmutz, Montserrat, Madrid; 1500 at Pforzheim, Sursee, Perpignan, Valenciennes, Jaen.

Printing seems to have begun in Scotland after September '507, when King James IV. granted a patent to Walter Chepman and Andrew Myllar (also printed Millar) for the establishment of a printing press at Edinburgh. Their first book (The Maying or disport of Chaucer) appeared on the 4th of April 1508. Myllar, however, appeared to have been established there as a bookseller already in 1503 and to have published there his first book, Joh. de Garlandia Interpr. vocabulorum equivocorum (printed for him abroad) in 1505, his second Expositio Sequentiarum (also printed abroad) in 1506. (See Rob. Dickson and John Ph. Edmond, Annals of Scottish Printing from 1507 to the 17th century, Cambridge, 1890; Harry G. Aldis, List of Books printed in Scotland before 1700, Edinburgh 1904). Printing was introduced into Ireland at Dublin in 1551 by Humfrey Powell, who published in that year a verbal reprint of Whitchurch's edition of the Common Prayerbook of 1549. Printing in Irish types was brought into the kingdom in 1571 by N. Walsh and John Kearney, the first book printed in that type being A Catechism, written by Kearney.

Above we have stated that printing was established at Avignon in the year 1497. But during the last two decades various treatises have been published endeavouring to show that printing had already been exercised there more than half a century earlier.

In 1890 the Abbat Requin discovered at Avignon, in three notarial registers, five Latin notarial Protocols of the years 1 444 and 1446, which, though they mention only the arts of " writing artistically," and painting different colours on stuffs, he and others interpreted as showing that, during those years, certain artisans had exercised the art of printing with movable types at Avignon; so that, if the art was not invented there, one of those artisans must have learnt the secret from Gutenberg, said to have been en- 'gaged in printing at Strassburg from 1436 to 1439And hence Avignon, hitherto regarded as the 60th town where printing was introduced, was to take the second place, if not the first, in the history of the invention of printing, between Strassburg and Mainz (Requin, L'Imprimerie a Avignon en 1444, Paris, 1890; id., Origines de l'imprimerie en France, Avignon, 1444, Paris, 1891).

From Requin's first document (dated July 4, 1 444) it appears that a silversmith, Procopius Waldfoghel, of Prague, residing at Avignon, had received from a magister Manaudus (also called Menaldus Vitalis, born at Dax, in the Departement des Landes, baccalaureus in decretis, and student at Avignon) two alphabets of steel, two iron forms (frames?), one steel screw, 48 forms of tin, and divers other forms belonging to the art of writing (duo abecedaria calibis et duas formas ferreas, unum instrumentum calibis vocatum vitis, quadraginta octo formas stangni necnon diversas alias formas ad artem scribendi pertinentes), and promised to return these instruments (ad usum scribendi pertinencia) the moment Manaudus asked for them. The second document (dated August 2 7, 1 444) makes no mention of tools or instruments, but is Procopius's bond for two sums of money (io to 27 florins) which he had borrowed from Georgius de la Jardina; for the first he promised to instruct the said George in the art of writing well and seemly, and to do the necessary and suitable things for one month (pro quibus promisit instruere dictum Georgium in arte scribendi bene et condecenter, et administrare necessaria et opportuna, hint ad unum mensem), on condition that neither of them should instruct anyone else in the said art of writing, without the consent of the other (fuit tamen de pacto quod nullus non debeat instruere aliquem in dicta arte scribendi, nisi de licentia alterius). The third document (March io, 1446) is an agreement between Procopius and a Jew of Avignon named Davinus de Codarossia, who had advanced money to him and held property from him as security. The Jew had promised to teach Procopius to paint stuffs in different colours, and the latter had promised the Jew to make for him and to deliver to him " twenty-seven prepared Hebrew letters, well and properly cut in iron according to the science and practice of writing, which, two years ago, the said Procopius had shown and taught the Jew, together with instruments of wood, tin and iron (Procopius promisit. .. judeo facere et factas reddere et restituere viginti septem litteras ebreaycas formatas, scisas in ferro bene et debite juxta scientiam et practicam scribendi, sunt duo anni elapsi ipsi judeo per dictum Procopium ostensam et doctam, ut dixit, una cum ingeniis de fuste, de stagno et de ferro). It was also agreed that the Jew should pay for the tin and wood for the instruments of the Hebrew writing (fuit de pacto quod idem judeus solvet stagnam et fustes artificiorum sive ingeniorum scripture ebrayce). And Procopius further promised to give the Jew, the following week, ten florins to recover certain pledges or utensils which the Jew had in pawn from him, the latter binding himself not to reveal the science or teach the art to any one as long as Procopius should remain at Avignon or in the neighbourhood (promisit eidem judeo dare decem florenos per totam hebdomadam proxime futuram et restituere sibi certa pignora sive ustensilia que ipse judeus habet in pignora a dicto Procopio). The fourth document (April 5, 1446) shows that Procopius had made for the above-named Menaldus Vitalis and Arnaldus de Coselhaco (and Girardus Ferrosis?) and delivered to them several instruments or tools of iron, steel, copper, latten, lead, tin and wood for writing artistically; he had instructed them in the said art of writing artistically, and all the tools belonged to them in common. But Menaldus, wishing to sell his share in the said tools to the others and to retire from the association, twelve florins were paid to him in two instalments, but at the request of Procopius he testifies under oath ths't the said art of writing, taught him artistically by the said Procopius, was real and most proper, and also easy, practicable and useful to any one wishing and choosing to work it (Cum dictus Procopius super arte scribendi artificialiter fecerit venerabilibus viris. .. Menaldo Vitalis et Arnaldo de Coselhaco. nonnulla instrumenta sive artificia causa artificialiter scribendi tarn ferro de callibe, de cupro, de lethono, de plumbo, de stagno, et de fuste. .. dictamque artem scribendi artificialiter eos docuerit, instrumentaque ipsa omnia et singula sins. communia inter eosdem studentes. .. Cumque dictus. .. Vitalis cupiat. .. par- tern suam dictorum instrumentorum sive artificiorum. vendere et a communione eorum recedere. .. vendidit dicto Procopio et Girardo presentibus. partem suam. .. precio duodecim flore - norum. .. Ibidem Vitalis. .. medio suo juramento. dixit dictam artem scribendi per dictum Procopium artificialiter eidem doctam, esse veram et verissimam, esseque facilem, possibilem et utilem laborare volenti et diligenti earn). The fifth document (April 26, 1446) shows that Procopius had recovered from Davinus all the pledges which he had pawned with him, except one mantle and 48 letters engraved in iron, that Davinus had not yet carried out his part of the agreement as to teaching Procopius the painting of different colours on stuffs, whereas Procopius had delivered to the Jew all the arts, tools and instruments pertaining to writing artistically in Latin letters, as he had promised to do on the 10th of March last. (Procopius confessus fuit se ab eodem judeo recepisse ... omnia pignora sua per eum penes dictum judeum impignorata, excepto uno mantello et quadraginta octo litteras gravatis in ferro. Et. .. dictus judeus confessus fuit. .. recepisse a dicto Procopio omnia artificia, ingenia et instrumenta ad scribendum artificialiter in litera latina, &c.) Again the compact is that Davinus shall not reveal the science to anyone, at least so long as Procopius should reside at Avignon or within 30 m. in the neighbourhood. (nemini mundi dicere, notificare nec quovismodo revelare, per se nec per alium ullomodo, presentem scientiam in teorica nec pratica, et nulli mundi earn docere neque revelare earn fuisse ostensam per quemvis). It is difficult to find the art of printing with movable types, or the art of casting types in these documents. The Abbat, however, says they prove the establishment of a printing-office at Avignon in 1444, and he reads " matrices," " caracteres d'imprimerie," une " imprimerie," and " tout un materiel d'imprirnerie " in them, although the documents themselves do not mention such things; they only allude to the " art of writing," the " practice " or " exercise of writing "; the " art of writing well and seemly "; the " science and practice of writing "; the " art of writing artistically." And there is, apparently, no reason to think that these precise documents, while speaking exclusively of this art, should always mean another art which they do not mention. Procopius, indeed, seemed to have known an art of writing, in which he instructed others (second document) and which he and his associates wished to keep secret, while the " letters," tools, &c. of which they speak were no doubt " movable." But Procopius himself appears to have possessed neither letters nor tools nor instruments or forms at the beginning of these proceedings; it was Menaldus Vitalis, a bachelor of law and student at Avignon, who entrusted to him the " two steel alphabets, two iron forms, one steel screw, and forty-eight tin and other forms," mentioned in the first document of 1444. Procopius, however, appears to have seen no permanent value in these letters, forms, &c. as he, of his own accord, promised to return them at the first request of Menaldus, who had handed them to Procopius without asking for a receipt. The third document, however, makes it plain that Procopius engraved for Davinus the Jew, not for himself, twenty-seven Hebrew letters (therefore a complete alphabet, including the five final letters) in iron, in accordance with the art of writing which he had taught Davinus two years ago, together with tools of wood, tin and iron, in return for which the Jew would teach Procopius the art of painting stuffs. The fourth document shows that Procopius had made tools of iron, steel and other metals for writing artistically, but again not for himself but for two other men one of whom was Menaldus who, two years ago, had entrusted him with two alphabets and some tools; Procopius, however, had this time reserved to himself a share in these tools, and Menaldus sold his share in the tools for twelve florins to the other associates, so that the value of all these tools cannot have amounted to more than about 36 florins of Avignon currency.

Therefore, the precise descriptions in the documents of the letters, tools and instruments required for Procopius's art of writing artistically, and the absence of all allusions to paper, ink and other things necessary for printing with movable types, show that there is no reference to this art, even in its infancy. That art means the multiplication of books or documents by means of an adequate quantity of single types for composing a whole page of text, and capable of being taken asunder and used again for a second, a third and a multitude of other pages, and so produce a number of copies of a book in the same or a shorter time than a scribe with his pen could produce one copy. But two Latin alphabets (of steel) and one Hebrew alphabet (of iron) would not suffice for composing and printing more than two or three words on any one page at a time, so that a person with such a small quantity of letters at his command would, in several respects, be worse off than a scribe. Hence the documents which only refer to the art of writing, mean nothing more serious than an art of taking impressions of certain letters (perhaps initials or capitals) in a more regular and steady fashion than even trained scribes could produce them by hand. For pressing in such (ornamental) initials or capitals here and there in MSS., after the scribes had done their ordinary work of writing, the insulated alphabets of Menaldus and Davinus would be a great help and save a deal of time and labour, but useless for the art of printing with movable types. If the two steel alphabets, and the one Hebrew alphabet of iron, and the 48 letters engraved in iron had been patrices, and the 48 forms of tin had been matrices, the documents, no doubt, would contain some expressions to show this, in spite of the endeavour not to divulge this art of writing. What the nature of this writing was, and why all these forms and instruments, even a screw, were required, we cannot say. It has been pointed out that the art of printing was also described as an art of writing, which is true; but when it is so described we learn at the same time that typography is meant. But we must bear in mind that Davinus the Jew was engaged on the painting of colours on stuffs and that Procopius desired to become acquainted with this industry. No doubt tools were much more required for this work than for writing. However, this writing association seems to have come to an end in 1446, and the parties departed from Avignon, without leaving there or anywhere else any trace of themselves and their interesting operations. See also Zedler, GutenbergForsch., p. 10 sqq.

As for non-European countries and towns, printing was established in Mexico in '544, at Goa about 1550, at Tranquebar in 1569, Terceira in the Azores 1583, Lima 1585,1585, Manila and Macao (China) 1590, in Hayti in the beginning of the 17th century, at Puebla in 1612, Cambridge (Mass.) 1638, Batavia 1668, Tiflis 1701,1701, German-town 1735, Ceylon 1737, Halifax (Nova Scotia) 1766,1766, Madras 1772,1772, Calcutta 1778, Buenos Aires 1789, Bombay 1792,1792, in Egypt (at Alexandria, Cairo, and Gizeh) in 1798, at Sydney 1802,1802, Cape Town 1806, Montevideo 1807, Sarepta 1808, Valparaiso 1810,1810, Astrakhan 1815,1815, in Sumatra and at Hobart Town and Santiago (in Chile) in 1818, in Persia (at Teheran) in 1820, and at Chios about 1821.1 On the introduction of printing in various towns, consult Henry Cotton, A Typog. Gazet., 8vo, Oxford, 1831 and (second series, 8vo, Till the moment (say 1477) that printing was practised in almost all the chief towns of the Netherlands, Germany, Italy, Switzerland, France, Spain, England, not a single printer carried away with him a set of types or a set of punches or moulds from the master who had taught him, but, in Li' 8an1' setting up his printing office, each man cast a set of types r t for his own use, always imitating as closely as possible the handwriting indigenous to his locality, or of some particular manuscript which he or his patron desired to publish. When we compare Schoeffer's 30-line Indulgence of 1454 with a manuscript copy of the same Indulgence dated the 10th of April 1454, now in the hands of a private collector at Wiesbaden, we see that the types used in printing that document were specially cast for the purpose after the model of the handwriting employed for the written copies. We know also that the types of the 36-line and 42-line Bibles and those of the Psalter of 1457 are the closest possible imitations of the ornamental church handwriting customary at the time of their production. Also, when we compare the 31-line Indulgence of 1454 with the German blockbook called the Enndtchrist, and both in their turn with the German MSS. of that period (especially the manuscript portions in the printed copies of the Indulgences), we see that the cutter of the text type of the Indulgence, as well as the engraver of the blockbook, formed his characters according to some German handwriting (book hand) of the period. This imitation extended, not only to the shape of the individual letters of the alphabet, but likewise to all those combinations of letters (double p, double f, double s, st, ti, tu, re, cu, ct, si, de, co, ci, te, ce, or, y e, po, fa, he, be, &c.) and contractions (for pro, -urn, -em, -en, the-, uer, -bus, -bis, sed, am, tur, qui, quae, quod, secundum, &c.) which were then, and had been for many centuries, in use by scribes. In most, if not all cases, the MSS. which the printers imitated were, as has been remarked above, indigenous to the place where they settled. Thus the first printers of Subiaco, though they were Germans and had most probably learnt the art of casting types and printing at Mainz, yet cut their types after the model of some Italian MS. which was free from any Gothic influence, but written in a pure Caroline minuscule hand, differing but slightly from the Caroline minuscules which the same printers adopted two years afterwards at Rome. The first Paris printers started in 1470 with a type cast entirely on the model of the Caroline minuscule handwriting then in vogue at Paris. John de Westphalia, who introduced printing into Belgium, used from the beginning a type which he calls Venetian. Therefore a great similarity (without absolute identity) between the types of two printers (e.g. Schoeffer and Ulr. Zell), should be attributed to the similarity of the handwritings which the printers followed, not to any attempt on their part to imitate each other's types. To this universal system (clearly discernible in the first twenty-five years of printing) of each printer setting up business with a new type cast by himself, there would be, according to the conjectures of some bibliographers, only two exceptions; one is Albrecht Pfister (see above); the other is the Bechtermunczes of Eltville (see above).2 Another important feature in the earliest books is that the printers imitated, not only the handwriting, with all its contractions, combined letters, &c., but all the other peculiarities of theU MSS. they copied. There is in the first place the un- of even evenness of the lines, which often serves as a guide to the approximate date of an early printed book, especially when we deal with the works of the same printer, since each commenced with uneven lines, and gradually made them less uneven, and finally even. The unevenness was unavoidable in manuscripts as well as in blockbooks; but in the earliest printed books it is regarded as evidence of the inability of the printers to space out their lines. If this theory be correct, this inability was perhaps owing to the types being perforated and connected with each other by a thread, or to some other cause which has not yet been clearly ascertained. In some incunabula we find some pages with uneven lines, and others quite straight in the same book. It is not impossible, however, that the unevenness was simply part and parcel of the system of imitating MSS., and that only gradually about 1473 or 1474, but in some cases later) printers began to see that even lines looked better than uneven. This seems clear when we observe that the imitation of MSS. was carried so far that sometimes things which deviated from the work of the scribe, but had accidentally been printed in, were afterwards erased and altered in conformity with the MS. The Paris Library, for instance, possesses two copies of the Liber Epistolarum of Gasparinus Pergamensis (printed at Paris in 1470), in both of which the initial G of the first line and the initial M of the fourth line were printed in, and, whilst they have been allowed to Oxford, 1866); (P. Deschamps) Diet. de geogr. a l'usage du libraire, (8vo, Paris, 1870); R. C. Hawkins, Titles of the First Books from the Earliest Presses Established in Different Cities in Europe, (4to, New York, 1884); Rob. Proctor, Early Printed Books in the British Museum, (1898), &c.

2 In recent years Dr Dziatzko, overlooking the relation between MSS. and typography in its infancy, has attempted to show that the types of the 36-line Bible were imitations of those of the 42-line Bible.

remain in one of the copies, in the other they were regarded as a fault and replaced by a rubricated L and M.

In the second place the initials of books or the chapters of books in MSS., and again in blockbooks and the earliest products of printing, Inifials, were always, or at least in most cases (they are printed in the Indulgences of 1454), omitted by the scribe and the printer and afterwards filled in by the rubricator. As the latter artists were sometimes illiterate and very often filled up the gap by a wrong initial, we find in many MSS. as well as early printed books small letters written either in the margin or in the blank left for the initial, to guide the rubricator. In most cases where these letters (now called initial directors) were written in the margin, they were placed as much as possible on the edges of the pages in order that they might be cut away by the binder as unsightly; but in many incunabula they have remained till the present day.' Later on these initial directors were in many books printed in (in lower-case type) with the text. In all cases, whether written or printed, they were meant to be covered by the illuminated initial; but, as a matter of fact, the latter very seldom covers the initial director so completely as to make it invisible, and in various cases the intended illumination was never carried into effect. With respect to the hyphens, which were used in the 1454 Indulgences and the 36-line and 42-line Bibles, always outside the printed margin, some of the earliest printers did not employ them at the moment that they started their presses, and in the case of some printers the nonuse or use of hyphens, and their position outside or inside the printed margin, serve as a guide to the dating of their products. After about 1472 they become more uniform in their shape and more generally used.

The use of signatures was confined in MSS. mostly to mark the quires (with a numeral or a letter of the alphabet), sometimes also the leaves; in many cases they were written close to the Signatures. bottom of the leaf, so that they might be cut off by the binder, which has happened in many cases, wholly or in part, as may be seen in many MSS.; in blockbooks they are usually printed with the picture on each sheet or page; they are not printed in incunabula close to the bottom line of the page before 1472 (at least in no earlier book with a date), when they appear in Joh. Nider's Praeceptorium Divinae Legis, published by Johan Koelhoff at Cologne. Caxton did not adopt them till 1480. In the books printed before 1472 they were written by the rubricator or the binder, in the same way as in the MSS.

Catchwords (custodes) were used for the first time about 1469 by Catchwords. of Spires, at Venice, in the first edition of .

Tacit us.

Pagination or rather foliation was first used by Arn. Ther Hoernen, at Cologne in 1471, in Adrianus's Liber de remediis fortuitorum. casuum, having each leaf (not page) numbered by figures placed in the end of the line on the middle of each right-hand page.

The practice among early printers of imitating and reproducing MSS. was not abandoned till many years after the first dated docuinent (1454) made its appearance; and, looking at the books printed, say from 1454 to 1477, from our present Progress . standpoint, the printing of that period may be said to First have been almost wholly stagnant, without any improve ment or modification. If some printers (for instance, Sweynheym and Pannarts at Subiaco and Rome, and Nicolas Jenson at Venice) produced handsomer books than others, this is to be attributed to the beauty of the MSS. imitated and the paper used rather than to any superior skill. Generally speaking, therefore, we shall not be far wrong in saying that the workmanship of Ketelaer and De Leempt's first book, published at Utrecht c. 1473, and that of Caxton's first book issued at Westminster in 1477, exhibit almost the same stage of the art of printing as the 1454 Indulgences. If, therefore, any evidence were found that Ketelaer and De Leempt and Caxton had really printed their first books in 1454, there would hardly be anything in the workmanship of these books to prevent us from placing them in that year. And conversely, if the Indulgences of 1454 had been issued without a date or without any names to indicate their approximate date, their workmanship might induce bibliographers to ascribe them to c. 1470, if not somewhat later. Even after 1 477 alterations in the mode of printing books came about slowly and almost imperceptibly. It was no longer a universal system for printers to begin business by casting a type for themselves, but some received their types from one of their colleagues. And, though there were still many varieties of types, one sort began to make its appearance in two or three different places. The combinations of letters were the first to disappear; but the contractions remain in a good many books even of the 17th century.

Some theories have been based on, and others have been considered to be upset by, the supposition that the early printers always required as much type as printers of the present day, or at any rate ' The university library of Basel possesses a collection of the earliest Paris books still bound in their original binding, in which these initial directors are written not only on the outer edges, but on the inner sides of the pages, and so close to the back that they can only be seen by stretching the books wide open.

so much as would enable them to set up, not only a whole quire of 4 or 5 sheets (= 8 or io leaves =16 or 20 pages), but even two quires (=40 pages). Consequently calculations have been made that, for instance, the printer of the 42-line Bible required a fount of at least 120,000 characters. See Bernard, Orig. de l'impr. i. 164, who was a printer himself and speaks very strongly on this point. But there are numerous proofs that many early books were printed page by page, even when in small 4to. For instance, in some books it has been observed that portions of the types with which the text of the first, second or third pages of a quire had been printed, were used to " lock up " the types employed for the later pages of the same quire, as is evident from the blank impressions of such portions being found on these later pages. Again, in some small books, two, three or four blank leaves are found at the end, showing a miscalculation of the printer at the commencement. Moreover, numerous itinerant printers of the 15th century established a press for a shoat time wherever they went, which proves that the furniture of the earliest printing-offices cannot have been of any great extent.

Early Types and their Fabrication

We must now take notice of two theories or traditions which have been current for a long time as to some intervening stage between the art of block-printing and the art of printing with movable cast metal types.' One theory or tradition would have it that the inventor of printing, after the idea of single, individual, movable types had arisen in his mind, practised his new invention for some time with wooden types, and that he came only gradually to the idea of movable types cast of metal.

Junius gives us to understand that the Dutch Speculum was printed with such wooden types. Of Johann Gutenberg it was asserted that he printed his first Bible with wooden Woode types. The Mainz psalter, printed in 1457 by Joh. Fust Typ es. and Peter Schoeffer, was alleged to have been printed with wooden types, in which case the 4th edition, published in 1502, and even the 5th edition of 1516, would be printed with wooden types, the same being used for them as for the editions of 1457 and 1 459. Theod. Bibliander was the first to speak (in 1548) of such types and to describe them: first they cut their letters, he says, on wood blocks the size of an entire page; but, because the labour and cost of that way was so great, they devised movable wooden types, perforated and joined one to the other by a thread.' Bibliander does not say that he had ever seen such types himself, but Dan. Speckle or Specklin (d. 1589), who ascribed the invention to Mentelin, asserts that he saw some of these wooden types at Strassburg.4 Angelo Roccha asserted in 1591 that he had seen at Venice types perforated and joined one to the other by a thread, but he does not say whether they were of wood or of metal.° In 1710 Paulus Pater asserted that he had seen wooden types made of the trunk of a box-tree, and perforated in the centre to enable them to be joined together by a thread, originating from the office of Fust at Mainz.° Bodman, as late as 1781, saw the same types in a wormeaten condition at Mainz; and Fischer stated in 1802 that these relics were used as a sort of token of honour to be bestowed on worthy apprentices on the occasion of their finishing their term.

Besides those who believed in these wooden types from the fact that the letters (especially in the Speculum) vary among themselves in a manner which would not be the case had they been cast from a matrix in a mould, there were authors and practical printers who attempted to cut themselves, or to have cut for them, some such wooden types as were alleged to have been used by the early printers. Some of them came to the conclusion that such a process would be quite practicable; others found by experiment that it would, in the case of small types, be wholly impossible. Nearly all the experiments, however, were made with the idea that the inventor of printing, or the earliest printers, started, or had to start, with as large a supply of type as a modern printer. This idea is erroneous, as it is known that, for a good many years after the first appearance of the art, printers printed their books (large or small) not by quires (quaternions or quinternions) but page by page? Therefore, all considerations of the experimenters as to the impracticability of such wooden types, on account of the trouble and length of time required for the cutting of thousands of types, fall to the ground in face of the fact that the earliest printers required only a very small quantity of type, in spite of the peculiar forms (combined letters, letters with contractions, &c.) which were then in vogue. Up to ' We do not allude to Tritheim's assertion that the Catholicon of 1460 was printed from wooden blocks; for this story, which he declares he had heard from Peter Schoeffer, if it were true, would belong to the history of block-printing. Nor need we speak of Bergellanus's verses (1541), in which he distinctly alludes to carved blocks.

Commentatio de ratione communi omnium linguarum et literarum, p. 80 (Zurich, 1548).

Chron. Argent. MS. ed. Jo. Schilterus, p. 442.

° De Bibliotheca Vaticana, p. 412 (Rome, 1591).

s De Germaniae miraculo, p. 10 (Leipzig, 1710).

7 See, for instance, W. Blades, Life of Caxton, i. 39.

the present time no book or document has come to light which can be asserted to have been printed from single, movable, wooden types. But we have seen above that the Haarlem tradition, as told by Junius, distinctly points to such types having been used for, among other things, the first edition of the Dutch Spiegel, and no one examining this edition (of which two copies are preserved at Haarlem) would deny that there are grounds for this belief; the dancing condition of the lines and letters making it almost impossible to think that they are metal types. For how long and to what extent such types were employed, if at all, we cannot say.

The other theory would have it that between block-printing and printing with movable cast types there was an intermediate stage Sculpto> of printing with " sculpto-fusi " types, that is, types Types, of which the shanks had been cast in a quadrilateral fusi mould, and the faces, i.e. the characters or letters, engraved by hand afterwards. This theory was suggested by some who could not believe in wooden types and yet wished to account for the marked irregularities in the types of the earliest printed books.

Gerardus Meerman, the chief champion of this theory, based it, not only on the words of Celtes (Amores, iii. 3), who in 1502 described Mainz as the city " quae prima sculpsit solidos aere characteres," but on the frequent recurrence of the word sculptus in the colophons of the early printers (for Jenson and Husner of Strassburg, see p. 514 above). Sensenschmid in 1475 said that the Codex Justinianus was " cut " (insculptus), and that he had " cut " (sculpsit) the work of Lombardus, In Psalterium. Meerman also interpreted the account of the invention of printing by Trithemius 1 as meaning that, after the rejection of the first wooden types, the inventors discovered a method of casting the bodies only of all the letters of the Latin alphabet from what they called matrices, on which they cut the face of each letter; and from the same kind of matrices a method was in time discovered of casting the complete letters of sufficient hardness for the pressure they had to bear, which letters they were before - that is, when the bodies only were cast - obliged to cut. 2 In this way Meerman explained that the Speculum was printed in sculptofusi types, although in the one page of which he gives a facsimile there are nearly 1700 separate types, of which 250 alone are e's. Schoepflin claimed the same invention for Strassburg, and believed that all the earliest books printed there were produced by this means. Meerman and Schoepflin agreed that engraved metal types (literae in aere sculptae) were in use for many years after the invention of the punch and matrix, mentioning among others so printed the Mainz psalter, the Catholicon of 1460, the Eggestein Bible of 1468, and even the Praeceptorium of Nider, printed at Strassburg in 1476. But the difficulty connected with the process of first casting the shanks and afterwards engraving the faces of the types has become apparent to those who have made experiments; and it seems more probable that the terms sculpere, exsculpere, insculpere, are only a figurative allusion to the first process towards producing the types, namely, the cutting of the punch, which is artistically more important to the fabrication of types than the mechanical casting - all the more as Schoeffer in 5468 makes his Grammatica vetus rhythmica say, " I am cast at Mainz," an expression which could hardly be anything but a figurative allusion to the casting of the types.

Granting that all the earlier works of typography preserved to us are impressions of cast-metal types, there are still differences of Types Cast opinion, especially among practical printers and type Sand. founders, as to the probable methods employed to cast In them. It is considered unlikely that the inventor of printing passed all at once to the perfect typography of the punch, the matrix and the mould. Bernard 3 thought that the types of the Speculums were cast in sand, as that art was certainly known to the silversmiths and trinket-makers of the 15th century; and he accounts for the varieties observable in the shapes of various letters on the ground that several models would probably be made of each letter, and that the types, when cast by this imperfect mode, would require some touching up or finishing by hand. He exhibits a specimen of a word cast for him by this process which not only proves the possibility of casting types in this manner, but also shows the same kind of irregularities as those observable in the types of the Speculum.

But here again it is argued that in types cast by this or any other primitive method there would be an absence of uniformity in what founders term " height to paper." Some types would stand higher than others, and the low ones, unless raised, would miss the ink and not appear in the impression. The comparativerarity of faults of this kind in the Speculum leads one to suppose that, if a process of sand-casting had been adopted, the difficulty of uneven heights had been surmounted either by locking up the forme face downwards, or by perforating the types, either at the time of casting or afterwards, and holding them in their places by means of a thread or wire.

Annales Hirsaugienses, ii. 421: " Post haec inventis successerunt subtiliora, inveneruntque modum fundendi formas omnium Latini alphabeti literarum, quas ipsi matrices nominabant, ex quibus rursum aeneos sive stanneos characteres fundebant, ad omnem pressuram sufficientes, quos prius manibus sculpebant." 2 Origines typographicae, app. p. 4 7 (The Hague, 1765).

Origine de l'imprzmerie, i. 40.

To this cause Ottley attributed the numerous misprints in the Speculum, to correct which would have involved the unthreading of every line in which an error occurred. And, as a still more striking proof that the lines were put into the forme one by one, in a piece, he shows a printer's blunder at the end of page 42 in the unmixed Dutch edition, where the whole of the last reference-line is put in upside down, thus: Nee viuia 6rs.pUt slam be rube Met unrtru il 1. -ta; l itiu 3q s}sattaso A " turn " of this magnitude could hardly have occurred if the letters had been set in the forme type by type.

A second suggested mode is that of casting in clay moulds, by a method very similar to that used in the sand process, and resulting in similar peculiarities and variations in the. types.

Ottley, who was the chief exponent of this theory, Types Cast suggested that the types were made by pouring melted in Clay lead or other soft metal into moulds of earth or plaster, Moulds. after the ordinary manner used from time immemorial in casting statues of bronze and other articles of metal. But the mould thus formed could hardly avail for a second casting, as it would be scarcely possible to extract the type after casting without breaking the clay, and, even if that could be done, the shrinking of the metal in cooling would be apt to warp the mould beyond the possibility of further use. Ottley therefore suggests that the constant renewal of the moulds could be effected by using old types cast out of them, after being touched up by the graver, as models - a process which he thinks will account for the varieties observable in the different letters, but which would really cause such a gradual deterioration and attenuation in the type, as the work of casting progressed, that in the end it would leave the face of the letter unrecognizable as that with which it began. It would, therefore, be more reasonable to suppose that one set of models would be used for the preparation of all the moulds necessary for the casting of a sufficient number of types to compose a page, and for the periodical renewal of the moulds all through the work, and that the variations in the types would be due, not to the gradual paring of the faces of the models, but to the different skill and exactness with which the successive moulds would be taken.

It is evident that the sand and clay methods of casting types above described would be slow. The time occupied after the first engraving of the models in forming, drying and clearing the moulds, in casting, extracting, touching up and possibly perforating the types required for one page, would exceed the time required by a practised xylographer for the cutting of a page of text upon a block. But he that has gone through the trouble of casting separate movable types has a clear gain over the wood-block printer in having a fount of movable types, which, even if the metal in which they were cast were only soft lead or pewter, might be used again and again in the production of any other page of text, while the wood block can only produce the one page which it contains. Moreover, only one hand could labour on the xylographic block; but many hands could be employed in the moulding and casting of types, however rude they might be. Bernard states that the artist who produced for him the few sand-cast types shown in his work assured him that a workman could easily produce a thousand such letters a day. He also states that, though each letter required squaring after casting, there was no need to touch up the faces.

A third suggestion was made as to the method in which the types of the rude school may have been produced. This may polytype. be described as a system of what the founders of about 1 800 called polytype, which is a cast or facsimile copy of an engraved block, matter in type, &c.

Lambinet, 4 who is responsible for the suggestion, based upon a new translation of Trithemius's narrative, explains that this process really means an early adoption of stereotype. He thinks that the first printers may have discovered a way of moulding a page of some work - an Abecedarium - in cooling metal, so as to get a matrixplate impression of the whole page. Upon this matrix they would pour a liquid metal, and by the aid of a roller or cylinder press the fused matter evenly, so as to make it penetrate into all the hollows and corners of the letters. This tablet of tin or lead, being easily lifted and detached from the matrix, would then appear as a surface of metal in which the letters of the alphabet stood out reversed and in relief. These letters could easily be detached and rendered mobile by a knife or other sharp instrument, and the operation could be repeated a hundred times a day. The metal faces so produced would be fixed on wooden shanks, type high, and the fount would then be complete. Lambinet's hypothesis was endorsed by Firmin-Didot, the renowned type-founder and printer of Lambinet's day. But it is impossible to suppose that the Mainz psalter of 1457, which these writers point to as a specimen of this mode of execution, is the impression, not of type at all, but of a collection of " casts " mounted on wood.

Yet another theory has been proposed by Dr Ch. Enschede, head of the celebrated type foundry of the same name at Haarlem, who says (pp. 15 sqq. of his Technisch onderzoek naar de uitvinding van de Boekdrukkunst, 1901), that the principle of a printing surface 4 Orig. de l'imprimerie i. 97 (2 vols. 8vo., Paris, 1810).

composed of separate pieces was known to the block-printer, but he would have found it impossible to use small insulated blocks of wood Enschede for printing, or to manufacture them for that purpose Theory. with the necessary mathematical precision. Hence the idea of separate movable characters was not the invention of printing, but the art of casting them, and this was a work not for the block-printer, but for another industry, for a foundry.

From the types of B36 and B 42 Enschede concludes that Gutenberg's punches (patrices) were made, like the bookbinders' stamps, of yellow copper (brass, Germ. Messing). With such patrices only leaden matrices could be made, but the latter could be produced in two ways: the lead can be poured over the patrix, or the patrix be pressed into cold lead. The first mode is somewhat complex, but the matrix would have a smooth surface, and need no further adjustment. The second mode is more simple, but requires great force, although lead is a soft metal. Moreover, the surface of the matrix has to be trimmed, as the impression forces the lead downwards and sidewards, which makes the surface uneven, though by this pressure the lead becomes firmer and more compact, tothe advantage of the type-founder. Enschede thinks that Gutenberg's letters must have been sharp, and that he obtained his matrices by the second mode; he had each letter engraved on a brass plate, 2mm. thick, therefore a mere letter without anything underneath it. This letter (patrix) was pressed, by means of a small flat plate, so far into the metal that its back formed one surface with the top part of the lead, and then removed. After the patrix and matrix had been made in this way, the letter was to be cast, and Enschede believes that for this work Gutenberg used what in Germany is called the Abklatsch-method, which, after having been gradually improved, was at last superseded by more perfect machinery. By this method the letter was cast in two tempos. First the letter itself on a small plate; then the plate placed underneath a castingform, to fix it to a small shank, which was to be cast into the form and would make, with the plate, the exact height of the letter. The letter on the plate was made not by pouring the metal into the matrix, but by beating the latter into the molten metal. When lead is heated so as to be a soft mass it easily assumes the form of any object which falls on or in it, therefore also of the matrix, which is the image of the engraved type. When the metal is not overheated it will immediately cool down by contact with the cold matrix, so that the latter will not be injured, although it consists of the same substance as the molten metal. In this way a great many letters can be cast from one matrix. Enschede describes various difficulties connected with this method, and tells us that only large letters, like those of B 36 and B 42, could be made by it, as the operation of adding the shank to the letter becomes impossible in the case of smaller letters. Hence Gutenberg, having conceived the idea of printing from seeing (!) the Dutch Donatuses, chose this large size of type for his work; for the smaller types of the 1454 Indulgences a copper matrix was required, which, in its turn, necessitated the use of a steel patrix, the introduction of which he ascribes, as others have done before him (e.g. Bergellanus), to Peter Schoeffer.

As to the Costerian types, their bad and irregular condition shows, he thinks, that they were produced from leaden matrices, and the latter from brass patrices, though wooden patrices are also possible, but not probable. All the tools, however, were imperfect, and the workmen inexperienced, and therefore bound to produce such imperfections as he finds in the Abecedarium and Donatus types. But the types were cast in one tempo; the Abklatsch-method would have been out of the question for them on account of their small size. In this way Enschede thinks Coster, not having learnt his art from anybody, invented the type cast with the staff, in one tempo, while Gutenberg, having had a Costerian Donatus as his model, cast his large types in two tempos by the Abklatsch system till Peter Schoeffer, by means of his steel patrices, was able to cast smaller types such as those of the 1454 Indulgences, with staff and all.

Enschede warns us that he is merely making suggestions as a type-founder, that he is not a bibliographer, and leaves the interpretation of documents to others. We quote his theories as coming from such a qualified type-founder, and because they have made some impression in certain quarters, but they lead us away from the real points connected with the invention of printing. First of all the " casting of metal types " is not, as he thinks, the first stage in the invention; its beginning, its essence is, and has always been thought to be, the movability of the characters. This movability, and the accidental way in which it was discovered, form together the pith of the Haarlem tradition as told by Junius. He indicates it, without using the word " movable," by saying that Coster, while walking in the Haarlem wood, cut some letters in the bark of a tree, and with them, " reversely impressed one by one on paper," composed one or two lines. Nothing seems more natural than that a block-printer (as the printer of the xylographically printed Speculum must have been) should cut such separate letters, and thereupon perceive that they could be used over and over again for a variety of words, on different pages, while those which he used to cut in a block only served him for one page and for one purpose. It is equally clear from the Haarlem tradition that the art of casting metal types was the second stage in the invention, a development or outcome of the primary idea of " movable letters," and the realization of their advantage, for Junius says that Coster " afterwards changed the beechen characters into leaden, and the latter again into tin ones." This also shows that the discoverer of the insulated movable wooden letters - after realizing, perhaps, that they could not endure much pressure, or missed (as Enschede says) the mathematical precision necessary for his purpose - transformed himself from a woodcutter into a letter-founder, and had no recourse (as Enschede would have it) for casting his types to a foundry apart from his own. As this transformation is possible and probable there seems to be no reason for departing from the simple but clear Haarlem tradition as we read it in Junius.

In the infancy of printing every printer, in different countries and different towns, starts with his own types; hence we may conclude that he had learnt the art of engraving and casting them himself, and so combined the art of type-founding with that of printing. This points back to a combination of the two or three arts in the first printing-office. It would be strange if the inventor of the movable letters, whom we have shown to have been a blockprinter, and therefore acquainted with the art of engraving letters, and other mechanical contrivances connected with printing, had lacked the ability, which his immediate followers possessed, of imparting to his movable characters, by some means or another, that firmness and precision which he required for the realization of his invention. How long Coster had been a block-printer before he invented, and how long and to what extent he continued to use, the movable wooden letters, we cannot tell.

That Enschede ascribes to Coster the invention of casting metal types with a shank (as they have been manufactured for centuries afterwards), and that of another mode of manufacturing types (the Abklatsch-met hod) to Gutenberg, suggested to the latter by seeing (!) the Donatuses printed at Haarlem, looks like an amiable attempt to get over the unpleasant tradition of the theft of Coster's types, but his theories are irreconcilable with the Haarlem tradition, with Zell's account of the relation between Dutch and Mainz printing and with bibliography in general.

It is not surprising that Enschede's theories called forth others from Zedler (Verofentl. i. 34), who argues as follows: Enschede says rightly that the type of the Hague Dutch Donatus is more defective than that of any other 15th-century book, more than even that of the Paris Donatus. Such types could not have been cast from a copper matrix. But a printer who had derived his art of casting types from Gutenberg or one of his pupils, would hardly, after the introduction of the steel stamp and the copper matrix (necessary for manufacturing the small types of the 1454 Indulgences), have returned to the casting of a small type from a leaden matrix, and used, moreover, a process which remained, in its consequences, behind that of Gutenberg. Zedler then points to a peculiarity of the earliest Dutch incunabula already mentioned above, namely, the sign of contraction connected with some letters by a fine stroke, which he says is not (!) found in the Dutch blockbooks, or in the Dutch MSS. He thinks, therefore, that this stroke was required by the method of casting this type. The stamp for making the matrix cannot have been a sta f f, on the lower end of which the reversed letter was cut, but a mere letter without any footing. Consequently, it must have consisted of lead not wood, and have been manufactured in the same way as Gutenberg's type was made, according to Enschede. Every sign of contraction had to be one whole with the letters to which they belonged to prevent their being shifted during the process of printing. The letters cast from the matrix made in this way had as foot a thin square plate which enclosed the letter but no staff, owing to the mode of making the stamp and the matrix. If the Dutch printer had intended to cast a type with a staff by means of a casting tool, however primitive, he would not have required the thin plate. But his letters, with a thin plate as their foot, required to be pasted on a sheet of strong paper, so as to be firmly connected in words and sentences for the purpose of printing. Hence the printer could regulate the spaces between the words, without using, like Gutenberg, spaces of a definite width for this purpose, so that he had no trouble in making the lines end evenly. From such a printing-surface with a firm footing, it was possible, after the ground had become hard, to obtain impressions just as from movable types enclosed in the forme. Zedler was told by an expert that, technically, there was nothing against such an explanation, but, he says, if it were correct, it would not solve the question, not yet satisfactorily answered, as to what we have to understand by the printed Dutch Donatuses. The " doctrinal jette en molle " of Jean le Robert and the libri impressi, mentioned under the year 1450 in the Memorial of the monastery Weidenbach in Cologne would then be books printed from such printing plates with separately cast letters. In this way Zells' account in the Cologne Chronicle would be confirmed (!). We should also understand why the Dutch, though knowing the art of casting types, only printed Donatuses and similar small schoolbooks, for which there was much demand, for in the present day, stereotype-printing is likewise used for books which, when editions follow each other rapidly, have to be printed unaltered. In this case Gutenberg would not be the inventor of the cast letter. But the Dutch could not claim, with Enschede, the honour of the invention of movable metal types. They invented the casting of letters, but it would be Gutenberg's merit to have invented the movable cast types. At any rate he would be the inventor of the casting instrument whereby the letter with the staff became independent, that is movable. The early Dutch printing letter, which could only be used by being firmly footed on a plate, would have missed its real value for printing, its free movability.

Zedler, for want of data, cannot say where and when Gutenberg learnt the technics of early Dutch printing, though the Cologne Chronicle tells us that from this printing his work began. But he thinks that tl .e secret arts which occupied Gutenberg at Strassburg, and which, when the documents are impartially (!) considered, can be regarded as nothing but experiments in the printing of books, are ea-lier than 1440. He will not decide whether Gutenberg has been in Holland, or whether this historical kernel is the foundation o the Coster legend (!) of Adrianus Junius which is independent the Cologne Chronicle. Anyhow, Gutenberg still required ten years of hard work and troublesome experiments, before he, basing himself on the early Dutch printing, whatever this may have been, could become the inventor of the present mode of printing books.

We here see how Enschede's theories give rise to Zedler's structure of theories. When the former says that Gutenberg chose for his first work the large letters of B36 and B 42, because the Abklatsch-method (invented [?] by him) was only fit for large letters, he forgets that the printers of these Bibles, wishing to apply their new art to the production of copies of the Bible in a speedier way than the scribes of their time were able to do, had, of necessity, to design their types from the large ornamental church-hand then in vogue for Bibles, Psalters, Missals, &c. For the same reason they prepared different, much smaller, types for the Indulgences of 1454, as the manuscript copies of these Indulgences, handed to them as " copy," were written in the bastard Roman book-hand, used for such documents. When the arts of casting types and of printing with them found their way to Mainz they were new in that city, but they came there already well-developed, and the printers, whoever they were, knew how to prepare themselves for any book or document which it was thought desirable to print. But of these questions Enschede takes no account. He ascribes the two Bibles to Gutenberg, because Dziatzko has done so, without inquiring whether Gutenberg (not Pfister) had, after all, anything to do with Bas.

Zedler's theories, partly developments, partly corrections of those of Enschede's, are based on the misapprehension that a peculiarity in the Costerian types, i.e. the connexion of the signs of contractions by a fine stroke with the letters over which they stand, does not occur either in the Dutch blockbooks or in the Dutch MSS. This connexion, however, far from being not found, is a conspicuous feature, in the Dutch blockbooks and MSS., and being faithfully reproduced in the Costerian types, shows how near these types stand to the block-printing and MS. periods. Zedler does not explain how he would print with the plate-footed types, pasted on strong paper, which he ascribes to Coster. Nor does he say whether he ever examined the Costerian editions of the Speculum, Donatuses, &c., to see whether they showed any traces of such awkward contrivances.

After having done justice, we hope, to these latest theories, which, in spite of their great length, leave many things unexplained, it is a pleasure to read once more Junius's unvarnished account of the Haarlem tradition, which contains no intricate theories, but a simple explanation of the rise and progress of printing with movable (metal) types in that city. The reading of it shows that real facts can be explained in a few words, while theories require long explanations, first for explaining away the real facts, and then for explaining the theories, which after all lead us astray.

The shape and manufacture of the types used as early as c. 1470 do not seem to have differed materially from those of the present types. This is evident (I) from the shape of the old types which were discovered in 1878 in the bed of the . river Saone, near Lyons, opposite the site of one of the Type 15th-century printing-houses of that city, and which there is reason to believe belonged once to one of those presses, and were used by the early printers of Lyons; (2) from a page in Joh. Nider's Lepra moralis, printed by Conrad Homburch at Cologne in 1476, which shows the accidental impression of a type, pulled up from its place in the course of printing by the ink-ball, and laid at length upon the face of the forme, thus leaving its exact profile indented upon the page; (3) from an entirely similar page (fol. 4 b) in Liber de laudibus ac festis gloriosae Virginis (Cologne, c. 1468). From the small circle appearing in the two last-mentioned types, it is presumed that the letters were pierced laterally by a circular hole, which did not penetrate the whole thickness of the letter, and served, like the nick of modern types, to enable the compositor to tell by touch which way to set the letter in his stick. The fact that in these two cases the letter was pulled up from the forme seems to show that the line could not have been threaded.

Vinc. Fineschi, Notizie Storiche sopra la stamperia di Ripoli, P. 49 (Florence, 1781), gives an extract from the cost-book of the Ripoli press, about 1480, which shows that steel, brass, copper, tin, lead and iron wire were all used in the manufacture of types at that period.' ' On the above theories and types consult T. B. Reed, Old English Letter Foundries, pp. 3-26.

History of the Earliest Types

The history and nomenclature of the earliest types are practically a continuation of the history and nomenclature of the characters figured in the earliest blockbooks, wood-engravings and MSS. For instance, Gothic type was first used, say, about the year 1445; but Gothic writing, of which that type was an imitation, was already known and used about the second half of the 12th century and can be traced still farther back (see above). Again, the pure Roman type, which appeared about 1464, is nothing but an imitation of what in palaeography is called the Caroline minuscule, a handwriting which was already fully developed towards the end of the 8th century (see Palaeography).

The broad outlines of the history of the earliest types are as follows: Gothic type, of the angular or pointed kind, was first used by the Haarlem printer of the Speculum, Donatus, &c. (see specimen No. I, taken from the British Museum copy of the Speculum humanae salvationis, mixed Latin edition), presumably c. 1445. An entirely similar but larger type (No. 2, taken from the British Museum copy of Ludovicus [Pontanus] de Roma, Singularia) was used, presumably by the same printer, c. 1465-1470. Gothic type appeared in Germany as a church type in 1454, in the 31-line Indulgence, presumably printed by Johan Gutenberg at Mainz (No. 3, from the Göttingen copy), and in the 30-line Indulgence (No. 4, taken from the British Museum copy), printed by Peter Schoeffer at Mainz. Type No. 3 was also used about the same time for the 36-line Bible, and type No. 4 for the 42-line Bible. Two much larger Gothic types appeared in the Psalter of 1457, published by Fust and Schoeffer (see Bernard, Origine, pl. vii.). In Italy Gothic type appears in 1468 (No. 5, taken from the British Museum copy of Cicero, De oratore, published at Rome by Ulr. Hahn, the 15th of December 1468, in small Roman type, with imprint in Gothic), but in a more rounded form; it is practically the ordinary Italian writing influenced by the Gothic. In France Gothic began to be used in 1 473; in England it appears first in Caxton's type about the year 1480.2 It was employed extensively in a great many of the earliest presses all over Europe, and continued to be used largely at all times, especially for Bibles, law books, royal proclamations, &c., and even to this day it is the national character of Germany. It is now usually called lettre de forme, black letter or English in English-speaking countries, lettre flamand in Holland, and fractur in Germany.

Bastard Italian or bastard Roman was introduced in 1454 at Mainz in the 31-line (No. 6) and 30-line (No. 7) Indulgences. It is also called lettre de somme, some think from the Summa of Thomas Aquinas, printed in the type of the Bible of 1462 by Fust and Schoeffer. Varieties of this kind Malian or of type were, like the Gothic, much used by the earliest Roman. printers, as, for instance, the printer of the 1460 Catholicon, Mentelin of Strassburg, c. 1460, and Ulrich Zell at Cologne, c. 1466, &c. In England it appeared in the first three books printed (1478, 1479) at Oxford (No. 8, taken from the British Museum copy of Jerome's Expositio in Simbolum Apostolorum wrongly dated 1468 for 1478) .

Roman type, the Caroline minuscule of palaeography, was first used in Germany about 1464, Strassburg, by the printer whose fount of type is known by a peculiarly shaped R, and who on that account is usually called " the R printer " (No. 9, taken from the British Museum copy of Durandus, Rationale, of which the Basel library possesses a copy which was bought in 1464).' In Italy it appears in 1465 at Subiaco (see Bernard pl. xii. No. 19), at Rome in 1467 (op. cit. p1. xii. No. 20), but in all its purity at Venice in 1469, used by Johannes of Spires (op. cit. pl. xii. No. 25), and at Paris in 1470 (op. cit. pl. xiii. No. 25). In England it was not used before 1518, when Richard Pynson printed Pace's Oratio in Pace nuperrima (see facsimile in Reed's Type Foundries ,p 92) .

Burgundian type, or gros batarde or secretary, was first used about1470-1472by Colard Mansion at Bruges (No. Io, taken from the British Museum copy of La Controversie de Noblesse, c. 1471-1472). With a somewhat similar type (No. I I, taken from the British Museum copy of the Recuyell) William Caxton is presumed to have printed, likewise at Bruges, a set of five books, of which the Recuyell of the History of Troye, a translation of a work by Raoul le Fevre, is the best known and was probably printed c. 1471.4 To this same class belong the first type (No. 12, from the British Museum copy of the Dictes) used in England by William Caxton for the printing of Dictes and Sayings of the Philosophers (Nov. 18, 1477), and that used by the printer of St Albans (No. 13, taken from the Cambridge University Library copy of Aug. Dactus, Elegancie). It was an imitation of the manuscript hand of the English and Burgundian scribes of the 15th century, and, after having figured for a long time in several of the early London and provincial presses, was about 1534 entirely superseded by the English black letter. To this class of type 2 See Blades, Life of Caxton, p1. xvii.

3 See Jules Philippe, L'Imprimerie a Paris, p. 219.

4 Cf. Blades, Life of Caxton. belong also the later lettre de civilite (c. 1570), the script (lettre coulee, lettre de finance, Dutch, geschreven schrift), set court, base secretary, and running secretary types.

roue 0:811tio peeb e: eo 8e& ro fuu: tr¢mou quo McmrnatCe poi; r No. 13. - Aug. Dactus, Elegancie, 1479.

On the types, illustrations, initials, &c., before 150o, consult also the facsimiles in Holtrop's Mon. typ. des Pays-Bas (the Hague, 1868); R. C. Hawkins, First Books and Printers of the Fifteenth Century (New York, 1884); William Blades, The Life of Caxton (London, 1861-1863); Bernard, Origine de l'imprimerie, vol. i. pls. iii. - xiii. (Paris, 1853); Placidus Braun, Notitia de libris ab artis typogr. inventione usque ad annum 1479 impressis (Augsburg, 1788); H. Noel Humphreys, Hist. of the Art of Printing, fol. (London, 1867); Veroffentlichungen der Gesellsch. fiir Typenkunde des 15. Jahrhunderts. Edd. Isak Collyn, Rud. Haupt, H. O. Lange, K. Haebler, V. Madsen, E. Voullieme, vol. i, &c. (220 facs. published, Leipzig, 1907 -); The Woolley [Geo. Dunn], Photographs of Early Types (400), designed to supplement published examples with references to the British Museum Index 18 991904, 5 pts., folio; K. Burger, Deutsche and italienische Inkunabeln, in getreuen Nachbildungen herausgeg., pts. 1-8 (200 pls.), folio (Berlin, 1892 -); E. Gordon Duff, Early English Printing, a series of facs., folio (London, 1896); Ch. Enschede, Fonderies de caracteres et leur materiel dans les Pays-Bas du Isme au ?9me siecle, fol. (Haarlem, 1908); Horace Hart Notes on a Century of Typography at the University Press, Oxford, 16 931 794, folio (Oxford, 1900); Olgar Thierry-Poux, Premiers monuments de l'imprimerie en France au 15 me siecle, fol. (Paris, 1890); British Museum (Facsimiles from early printed books in the), (1897), 32 pls. folio; Type Facsimile Society, folio (Oxford, 1900 -).

The types after 1500 can best be learned from the catalogues of type-founders, among which those of Messrs Enschede of Haarlem occupy a foremost place. Of others we may mention: Indite dei caratteri nella stampa Vaticana, 4to (Rome, 1628); Epreuves des caracteres qui se trouvent chez Claude Lameste, 4to (Paris, 1742); Epreuves des car. de la fonderie de Claude Mozet, 8vo (Nantes, 1754); Les Car. de l'imprimerie par Fournier le Jeune, 8vo (Paris, 1764); Proef van Letteren, Bloemen, £&c., van Ploos van Amstel, 8vo (Amsterdam, 1767); Epreuve de car. de Jacques Francois Rosart, 8vo (Brussels, 1771); Schriften.. bey J. H. Prentzler, 4to (Frankfort-on-Main, 1 774) i Epreuves des car. de la fond. de J. L. Joannis, 8vo (Paris, 1776); Epreuves des car. de la fond. de J. L. de Boubers, 8vo (Brussels, 1 777); Proeve van Letteren welke gegooten worden door J. de Groot, 8vo (the Hague, 1787); Pantographie, by Edmund Fry, 8vo (London, 1799); and Manuale typographico, by G. Bodoni, 4to (Parma, 1818).

Printers after 150o. - Though the Cologne Chronicle of 1499 denies to Mainz the honour of the invention of the art of printing, it was right in asserting that, after it had been brought there from Holland, it became more masterly and exact, and more and more artistic. During the first half-century of printing a good many printers distinguished themselves by the beauty, 'excellence and literary value of their productions. We may mention as such: Johan Fust and Peter Schoeffer at Mainz; Johan Mentelin and Heinrich Eggestein at Strassburg; Ulrich Zell at Cologne; Sweynheym and Pannarts at Subiaco and at Rome; Nicolas Jenson at Venice; Anton Koberger at Nuremberg; Ketelaer and De Leempt at Utrecht; Johan Veldener at Louvain, Utrecht and Kuilenburg; Gerard Leeu at Gouda; Johan of Westphalia at Louvain; and William Caxton (q.v.) at Westminster.

Very soon the demand for books increased, and with it came a reduction in their prices. This caused a decline in the execution of printing, which begins to be appreciable about 1480 in some localities, and may be said to have become general towards the end of the 15th century. At all times, however, we find some printers raise their art to a great height by the beauty of their types and the literary excellence of their productions. Among the later printers we may mention the Aldi of Venice (1490 to 1597); G. B. Bodoni of Parma (1768-1813); John Amerbach at Basel (1492-1516); John Froben at Basel (1496-1527); John Baskerville at Birmingham (1750-1775);1775); the house of Weichel, first at Paris (c. 1530-1572), afterwards at Frankfort; Christopher Plantin at Antwerp (1554-1589); the Elzevirs, fist at Leiden, afterwards at Amsterdam (1580-1680); Antoine Verard at Paris (1485-1513); Josse Bade or Badius at Paris (1495-1535); and the Estiennes at Paris (1502-1598).

The Italic type 1 is said to be an imitation of the handwriting of Petrarch, and was introduced by Aldus Manutius of Venice for the purpose of printing his projected small editions of the classics. The cutting of it was entrusted to Fran cesco da Bologna, an artist who is presumed to be identical with the painter Francesco Francia or Raibolini. The fount is a " lower case " only, the capitals being Roman in form. It contains a large number of tied letters, to imitate handwriting, but is quite free from contractions and ligatures. It was first used in the Virgil of 1500. Aldus produced six different sizes between 1501 and 1558. It was counterfeited almost immediately in Italy, at Lyons and elsewhere. Originally it was called Venetian or Aldine, but subsequently Italic type, except in Germany and Holland, where it is called " cursive." The Italians also adopted the Latin name " characteres cursivi seu cancellarii." In England it was first used by Wynkyn de Worde in Wakefield's Oratio in 1524. The character was at first intended and used for the entire text of classical works. When it became more general, it was employed to distinguish portions of a book not properly belonging to the work, such as introductions, prefaces, indexes, notes, the text itself being in Roman. Later it was used in the text for quotations, and finally served the double part of emphasizing certain words in some works, and in others, chiefly translations of the Bible, of marking words not rightly forming a part of the text.

Greek type (minuscules) first occurs in Cicero, De officiis printed at Mainz in 1465 by Fust and Schoeffer. The fount used is rude and imperfect, many of the letters being ordinary Latin. In the same year Sweynheym and Pannarts used a good Greek. Greek letter for some of the quotations in their edition of Lactantius (see, for instance, leaves I ia, 19a, 36a, 139, 140); but the supply was evidently short at first, as some of the larger quotations in the first part of the book were left blank to be filled in by hand. The first book wholly printed in Greek minuscules was the Grammar of Lascaris, by Paravisinus, at Milan in 1476, in types stated to have been cut and cast by Demetrius of Crete. The fount contains breathings, accents and some ligatures. The headings to the 1 These paragraphs on the various types are for the most part taken from T. B. Reed's History of the Old English Letter Foundries, p. 50 seq. (London, 1887).

No. 1. - Speculum type 1 445 (?)3 6 Nos. 3 and 6. - Mainz 31-line Indulgence, 1454.

No. 5. - Cicero, De oratore, 1468.

ft No. 8. - Jerome's Expositio (1468), 1478. No. ii. - Recuyell of the Hist. of Troye, c. 1 47 1. No. 2. - Pontanus type, C. 1470 (?).

lariat 7 Nos. 4 and 7. - Mainz 30-line Indulgence, 1454.

No. io. - Controversie de Noblesse, c. 1 47 1-14 7 2. 'de f ice ' No. 9. - Durandus, c. 1464.

a No. 12. - Dictes and Sayings, 1477.

c.

chapters are wholly in capitals. The Anthologia graeca of Lascaris was printed at Florence in 1494 wholly in Greek capitals (litterae majusculae), and it is stated in the preface that they were designed after the genuine models of antiquity to be found in the inscriptions on medals, marbles, &c. But as late as 1493 Greek type was not common, for in that year the Venice printer Symon Bevilaqua issued Tibullus, Catullus and Propertius with blanks left in the commentary for the Greek quotations. In England Greek letters ap W ared for the first time in 1519 in W. de Worde's edition of Rob. hittington's Grammatica, where a few words are introduced cut in wood. Cast types were used at Cambridge in Galen's De temperamentis, translated by Linacre, and printed by Siberch in 1521, who styles himself the first Greek printer in England; but the quotations in the Galen are very sparse, and Siberch is not known to have printed any entire book in Greek. The first printer who possessed Greek types in any quantity was Reginald Wolfe, who held a royal patent as printer in Greek, Latin and Hebrew, and printed in 1543 two Homilies of Chrysostom, edited by Sir John Cheke, the first Greek lecturer at Cambridge. In Edinburgh, in 1563, and as late as 1579, the space for Greek words was left blank in printing, to be filled in by hand.

The Oxford University Press, re-established in 1585, was well supplied with Greek types, which were used in the Chrysostom of 1586. About 1607 Sir Henry Savile introduced Greek types (vulgarly called on account of their beauty " the silver letter ") into Eton College, for printing his edition of St Chrysostom (8 vols., 1610-1613, John Norton), and other Greek authors. He afterwards presented this type to the university of Oxford. In 1632 Cambridge applied to Oxford for the loan of a Greek fount to print a Greek Testament, and the same university made an offer in 1700 for the purchase of a fount of the king's Greek at Paris, but withdrew on the French Academy insisting as a condition that every work printed should bear the imprint " characteribus Graecis e typographeo regio Parisiensi." It should not be forgotten that the large number of ligatures in the Greek of that day made the production of a fount a serious business. The Oxford Augustin Greek comprised no fewer than 354 matrices, the great primer 456, and Fournier's fount showed even 77 6 different sorts. The Dutch founders effected a gradual reduction of the Greek typographical ligatures. Early in the 19th century a new fashion of Greek, for which Porson was sponsor and furnished the drawings, was introduced, and has remained the prevailing form to this day. Cf. Rob. Proctor, The Printing of Greek in the XVth Century, folio (Oxford, 1900).

The first Hebrew types are generally supposed to have appeared in 1475 in Petrus Niger's Tractatus contra perfidos Judaeos (leaf to),. printed by Conrad Fyner at Esslingen. De Rossi states that a Hebrew work in four folio volumes entitled Arba Turim, of Rabbi Jacob ben Asher, was printed in 1475 at Pieve di Sacco in Austrian Italy, while in the same year, a few months earlier, Salomon Jarchi's Comment. on the Pentateuch appeared at Reggio in Italy, printed in the Rabbinical character. Numerous other Hebrew works followed before 1488, in which year the first entire Hebrew Bible was printed, with points, at Soncino, by a family of German Jews. The first English book in which any quantity of Hebrew type was used was Dr Rhys's Cambro-Brytannicae Cymraecaeve linguae institutiones, printed by Thomas Orwin in 1592, though already in 1524 Hebrew characters, but cut on small blocks of wood, were used by W. de Worde in Rob. Wakefield's Oratio. The Hebrew fount made use of in Walton's Polyglott in 1657 was probably the first important fount cut and cast in England, though there were as yet no matrices there for Rabbinical Hebrew. In the beginning of the 18th century Amsterdam was the centre of the best Hebrew printing in Europe.

The first book printed in Arabic types is said to be a Diurnale Graecorum Arabuni, printed at Fano in Italy in 1514.1 Two years. later P. P. Porrus's Polyglott Psalter, comprising the Arabic version, was printed at Genoa; and two years later a Koran in Arabic is said to have been printed at Venice. In 1505 an Arabic Vocabulary at Granada had the words printed in Gothic letters with the Arabic points placed over them; and in other presses where there were no Arabic types the language was expressed in Hebrew letters or cut in wood. De Guignes and others mention a fount of Arabic used by Gromors in Paris in1539-1540to print Postel's Grammar. In England some Arabic words were introduced in Wakefield's Oratio of 1524, but apparently cut on small blocks of wood. In Minsheu's Ductor in linguas, 1617, the Arabic words are printed in Italic characters. Laud's gift of Oriental MSS. to Oxford in 1635, and the appointment of an Arabic lecturer, were the first real incentives to the cultivation of the language by English scholars. Previous to this it is stated that the Raphelengius Arabic Press at Leiden had been purchased by the English Orientalist, William Bedwell; but, if it was brought to England, it does not appear to have been immediately made use of. The Arabic words in Thomas Greave's Oratio de linguae Arabicae utilitate, printed at Oxford in 1639, were written in by hand.

Syriac type, probably cut in wood, first appeared in Postel's Linguarum XII. Alphabeta, printed in Paris in 1538; but the characters are so rude in form and execution as to be scarcely legible.

1 See Panzer vii. 2.

In 1555, however, Postel assisted in cutting the punches for the Syriac Peshito New Testament, printed at Vienna in 4to, the first portion of the Scriptures, and apparently the first book, Syriac printed in that language. In1569-1572Plantin at Antwerp included the Syriac New Testament in his Polyglott, and reissued it in a separate form in 1574. In England Syriac was usually expressed in the earlier works in Hebrew characters. But in 1652, when the prospectus and preliminary specimen of Walton's Polyglott were issued, we find Syriac type in use.

Of the Armenian character the press of the Vatican possessed a good fount in 1591, when Angelo Roccha showed a specimen in his Bibliotheca Apostolica Vaticana. A psalter is said to have been printed at Rome in 1565, and Rowe Mores mentions doubtfully a liturgy printed at Cracow in 1549. Armenian printing was practised in Paris in 1633; but the Armenian bishops, on applying to France for assistance in printing an Armenian Bible, in 1662, were refused, and went to Rome, where, as early as 1636, the press of the Propaganda had published a specimen of its Armenian matrices. The patriarch, after fifteen months' residence in Rome, removed to Amsterdam, where he established an Armenian press, and printed the Bible in 1666, which was followed in 1668 by a separate edition of the New Testament. In 1669 the press was set up at Marseilles, where it continued for a time, and was ultimately removed to Constantinople. In England the first Armenian type was that presented by Dr Fell to Oxford in 1667. The alphabet given in the prolegomena of Walton's Polyglott was cut in wood.

Of Ethiopic the earliest type appeared in Potken's Psalter and Song of Solomon, printed at Rome in 1513. The work was reprinted at Cologne, in 1518, in Potken's Polyglott Psalter. In 1548 the New Testament was printed at Rome by some Ethiopic. Abyssinian priests. The press of the Propaganda issued a specimen of its fount in 1631, and again in Kircher's Prodromus Coptus in 1636. Erpenius at Leiden had an Ethiopic fount, which in 1626 was acquired by the Elzevirs. Usher attempted to procure the fount for England; but, his attempt failing, punches were cut and matrices prepared by the London founders for the London Polyglott, which showed the Psalms, Canticles and New Testament in the Ethiopic version.

Of Coptic the press of the Propaganda possessed a fount, and a specimen was issued in 1636, in which year also Kircher's Prodromus Coptus appeared from the same press. In England Coptic. David Wilkins's edition of the New Testament was printed in 1716 from Coptic types cast with matrices which Dr Fell had presented to Oxford in 1667. The alphabets shown in the introduction and prolegomena to the London Polyglott of 1655 and 1657 were cut in wood.

Of Samaritan the press of the Propaganda had a fount in 1636, and the Paris Polyglott, completed in 1645, contained the entire Pentateuch in type, the punches and matrices of which had been specially prepared under Le Jay's direction.

The fount used for the London Polyglott in 1657 is admitted to have been an English production, and was probably cut under the supervision of Usher.

With Slavonic type a psalter was printed at Cracow as early as 1491, and reprinted in Montenegro in 1495. The only Slavonic fount in England was that given by Dr Fell to Oxford, and this, Mores states, was replaced in 1695 by a fount of the more modern Russian character, purchased probably at Amsterdam. The Oratio Dominica in 1700 gives a specimen of this fount, but renders the Hieronymian version in copper-plate. Modern Slavonic, better known as Russian, is said to have appeared first in portions of the Old Testament, printed at Prague in 1517-1519. Ten years later there was Russian type in Venice. A Russian press was established at Stockholm in 1625, and in 1696 there were matrices in Amsterdam, from which came the types used in Ludolph's Grammatica Russica, printed at Oxford in that year, and whence also, it is said, the types were procured which furnished the first St Petersburg press, established in 1711 by Peter the Great. Mores notes that in 1778 there was no Russian type in England, but that Cottrell was at that time engaged in preparing a fount. It does not appear that this project was carried out, and the earliest Russian in England was cut by Dr Fry from alphabets in the Vocabularia, collected and published for the empress of Russia in 1786-1789. This fount appeared in the Pantographia in 1799.

A fount of the Etruscan character cut by William Caslon about 1733 for Swinton of Oxford was apparently the first produced. Fournier in 1766 showed an alphabet engraved in metal t or wood. In 1771 the Propaganda published a specimen of their fount, and Bodoni of Parma in 1806 exhibited a third in his Oratio Dominica. Runic types were first used at Stockholm in a Runic and Swedish Alphabetarium, printed in 1611. The fount, which was cast at the expense of the king, was afterwards acquired by the university. About the same time Runic type was used at Upsala and at Copenhagen. Voskens of Amsterdam had matrices about the end of that century, and it was from Holland that Francis Junius is supposed to have procured the matrices which, in 1677, he presented to Oxford. This fount appears in the Oratio Dominica of 1700, and in Hickes's Thesaurus (1703-1705), and it remained the only one in England.

Matrices of Gothic type were presented to Oxford by Francis Junius in 1677, and a fount of them was used for the Oratio Dominica of 1700 and in Hickes's Thesaurus. A different fount was used for Chamberlayne's Oratio Dominica, printed at Amsterdam in 1715. Caslon cut a fount which appeared in his first specimen in 1734. This and the Oxford fount were the only two in England in 1820.

Founts of Icelandic, Swedish and Danish were included in Junius's gift to Oxford in 1677, and were, perhaps, specially prepared in Holland. The first-named is shown in the Oratio Dominica of 1700 and in Hickes's Thesaurus. Printing had been practised in Iceland since 1531, when a Breviary was printed at Hoolum, in types rudely cut, it is alleged, in wood. In 1574, however, metal types were provided and several works produced. After a period of decline, printing was revived in 1773, and in 1810 Sir George M'Kenzie reported that the Hoolum press possessed eight founts of type, of which two were Roman, and the remainder of the common Icelandic character, which, like the Danish and Swedish, bears a close resemblance to the German.

For the Anglo-Saxon language the first type was cut by John Day in 1567, under the direction of Archbishop Parker, and appeared Anglo- in ?Elfric's Paschal Homily in that year and in the 1Elfredi res gestce of Asser Menevensis in 1574.. AngloSaxon type was used by Browne in 1617, in Minsheu's Ductor in linguas; and Haviland, who printed the second edition of that work in 1626, had in 1623 made use of the character in Lisle's edition of lElfric's Homily. The first fount of Irish character was that presented by Queen Elizabeth to O'Kearney in 1571, and used to print the Catechism. which appeared in that year in Dublin, from the press of Franckton. But the fount is only partially Irish, many of the letters being ordinary Roman or Italic. It was used in several works during the early years of the 17th century, and as late as 1652 in Godfrey Daniel's Christian Doctrine, printed in Dublin. The Irish seminaries abroad were better supplied with Irish type. A new type was cut by Moxon, and appeared in 1681 in Boyle's New Testament, printed by Robert Everingham.

The earliest specimen of music type occurs in Higden's Polychronicon, printed by De Worde at Westminster in 1495. The Musk. square notes appear to have been formed of ordinary Mus quadrats, and the staff-lines of metal rules imperfectly joined. In Caxton's edition of the same work in 1482 the space had been left to be filled up by hand. The plain chant in the Mainz psalter of 1490, printed in two colours, was probably cut in wood. Hans Froschauer of Augsburg printed music from wooden blocks in 1473, and the notes in Burtius's Opusculum Musices, printed at Bologna in 1487, appear to have been produced in the same manner; while at Lyons the missal printed by Matthias Hus in 1485 had the staff only printed, the notes being intended to be filled in by hand. About 1500 a musical press was established at Venice by Ottavio Petrucci, at which were produced a series of mass-books with lozenge-shaped notes, each being cast complete with a staff-line. In 1513 he removed to Fossombrone, and obtained a patent from Leo X. for his invention of types for the sole printing of figurative song (cantus figuratus). Before 1550 several European presses followed Petrucci's example, and music type was used, among other places, at Augsburg in .1506 and 1511, Parma in 1526, Lyons in 1532 and Nuremberg in 1549. In 1525 Pierre Hautin cut punches of lozenge-shaped music at Paris. Round notes were used at Avignon in 1532. In England, after its first use, music-printing did not become general till 1550, when Grafton printed Marbecke's Book of Common Prayer, " noted " in movable type, the four staff-lines being printed in red and the notes in black. There are only four different sorts of notes used - three square and one lozenge. About 1660 the detached notes hitherto employed began to give place to the " new tyed note," by which the heads of sets of quavers could be joined. But at the close of the 17th century music-printing from type became less common, on account of the introduction of stamping and engraving plates for the purpose. Cf. Rob. Steele, The Earliest English Music Printing, folio (London, 1903); Andr. Deakin, Mus. Bibliogr., 8vo. (Birmingham, 1893).

Printing for the blind was first introduced in 1784 by Valentin Haiiy, the founder of the asylum for blind children in Paris. He made use of a large script character, from which impressions were taken on a prepared paper, the impressions being so deeply sunk as to leave their marks in strong relief and legible to the touch. Haiiy's pupils not only read in this way, but executed their own typography, and in 1786 printed an account of their institution and labours as a specimen of their press. The first school for the blind in England was opened in Liverpool in 1791, but printing in raised characters was not successfully accomplished till 1827, when Gall of the Edinburgh asylum printed the Gospel of St John from angular types. Alston, the treasurer of the Glasgow asylum, introduced the ordinary Roman capitals in relief, and this system was subsequently improved upon by the addition of the lower-case letters by Dr Fry, the type-founder, whose specimen gained the prize of the Edinburgh Society of Arts in 1837. Several rival systems have competed in England for adoption, of which the most important are those of Lucas, Frere, Moon, Braille, Carton and Alston (see Blindness): The trouble and cost involved in the use of the initial director early suggested the use of woodcut initials, and Erhard Ratdolt of Venice, about 1475, is generally supposed to have been the first printer to introduce the literae florenies, called also lettres tourneures, or typi tornatissimi, which eventually superseded the hand-painted initials. Caxton introduced one or two kinds in 1484. Among the earliest to be used are the so-called Lombardic initials or capitals. The more elaborate initials, such as those used in the Mainz indulgences and psalter, by Aldus at Venice, by Johann Schoeffer at Mainz in 1518, by Tory and the Estiennes at Paris, by Froben at Basel, and by the other great printers of their day, were known as lettres grises. Besides these, the ordinary " twoline letters " or large plain capitals came into use; and these were generally cast, whilst the ornamental letters were for the most part engraved on wood or metal.

Type ornaments and flowers began, like the initials, with the illuminators, and were afterwards cut on wood or metal. The first printed ornament or vignette is supposed to be the scutum or arms of Fust and Schoeffer in some copies of their 1457 Psalter, and of their edition of the Bible of 1462. There is no vignette in the Subiaco Lactantius of 1465 (as stated by Mr Reed, Letter Foundries, p. 82). In Holtrop's Monum. typogr. des Pays-Bas may be seen borders used by some of the earliest printers of Holland (1475-1490), which would not look bad even in the present time. Caxton in 1490 used ornamental pieces to form the border for his Fifteen 0's. At the same time the Paris printers engraved still more elaborate border pieces. At Venice entire frames were engraved in one piece, while Aldus as early as 1495 used tasteful head-pieces cut in artistic harmony with his lettres grises. Early in the 16th century we observe detached ornaments and flourishes which have evidently been cast from a matrix.

B1sL10GRAPHY

Besides the works of Berjeau, Bernard, Blades, Hawkins, Hessels, Holtrop, Noel Humphreys, Koehler, Jules Philippe, T. B. Reed, Sotheby, Steele, Weigel, &c., already mentioned, consult also Bigmore and Wyman, A Bibliography of Printing (London, 1880); Geo. Wolfg. Panzer, Annales typog. (Nuremberg, 1 793, &c.); Lud. Hain, Repertorium bibliog. (Stuttgart, 1826-1838), with indices by Conr. Burger (1891 and 1908); suppl. by W. A. Copinger (1895-1902), and Appendices ad Hainii-Copingeri repertorium, by Diet. Reichling (1905-1909); Holtrop, Cat. librorum sec. xv° impressorum in bibl. Regia Hagana (the Hague, 1856); M. F. A. G. Campbell, Ann. de la typog. neerlandaise au xve sii cle (the Hague, 1874); Rob. Sinker, A. Cat. of the X V. Century Printed Books in the Library of Trinity College, Cambridge (Cambridge, 1876); W. Th. Lowndes, Bibliographer's Manual, ed. by H. G. Bohn (London, 1858, &c.) J. C. Brunet, Manuel du libraire (Paris, 1860; four earlier editions;- Th. F. Dibdin, Bibliotheca Spenceriana (London, 1814, &c., and his other works); Ennen, Katalog der Incunabeln in der Stadt-Bibliothek zu Koln; Schoepflin, Vindiciae typog. (1760); Meerman, Origines typog. (the Hague, 1765); Dupont, Hist. de l'impr. (Paris, 1869); Firmin-Didot, Hist. de la typog. (Paris, 1882); E. Duverger, Hist. de l'invention de l'impr. (Paris, 1840); P. Lambinet, Origine de l'impr. (Paris, 1810); Ch. Ruelens, La Legende de St Servais (Brussels, 1873); J. P. A. Madden, Lettres d'un bibliographe (Paris, 1868-1878); Wetter, Krit. Gesch. der Erfindung der Buchdruckerkunst (Mainz, 1836); A. de Vries, Eclaircissemens sur l'histoire de l'inv. de l'impr. (the Hague, 1843); Jos. Ames, Typogr. Antiquities (augmented by W. Herbert; London, 1785-1790); T. C. Hansard, Typographia (London, 1825); Thomas, Hist. of Printing in America (Albany, 1874); Th. L. Devinne, The Inv. of Print. (London, 1877); W. Skeen, Early Typography (Colombo, 1872); Sam. Palmer, A General Hist. of Print. (London, 1732); W. Young Ottley, Inquiry concerning the Inv. of Print. (London, 1863); Henry Bradshaw, A Classified Index of the 15th Century Books in the Collection of the late M. J. de Meyer (London, 1870); idem, Hist. of the Founts of Type and Woodcut Devices used by Printers in Holland in the 15th Century (London, 1871); idem, The Printer of the Historia S. Albani (Cambridge, 1868); A. Van der Linde, Haarlem Legend (London, 1870) idem, Gutenberg (Stuttgart, 1881); idem, Gesch. der Erfind. der Buchdruckerkunst (Berlin, 1886); Schaab, Gesch. der Erfind. der Buchdruckerk. (Mainz, 1830); K. Falkenstein, Gesch. der Buchdruckerk. (Leipzig, 1856); Lorck, Handb. der Gesch. der Buchdruckerk. (Leipzig, 1882); K. Faulmann, Illustr. Gesch. der Buchdruckerk. (Vienna, 1882) M. Denis, Wiens Buchdruckergesch. bis 1560 (Vienna, 1782); C. R. Hildeburn, A Century of Printing - The Issues of the Press in Penns y lvania, 1684-1784 (Philadelphia, 1887); J. Garcia Icazbalceta, Bibliografia Mexicana del siglo xvi. (Mexico, 1887); Bibliographica (3 vols., London, 1895-1897); W. Blades, Bibliographical Miscellanies (London, 1890); C. M. Briquet, Les Filigranes. Dict. hist. des marques du papier (4 vols., Geneve, 1907); Konr. Burger, Beitrlige zur Inkunabelbibliogr. (Leipzig, 1908); Catal. of Books printed in the 15th Century, Brit. Mus. pt. i. (London, 1908); Catal. of MSS. and early Printed Books in the library of J. Pierpont Morgan (3 vols., London, 1907); Arth. Christian, Origines de l'imprimerie en France (Paris, 1900); A. Claudin, Monum. de l'imprimerie d Poitiers (Paris, 1897); idem, Histoire de l'imprimerie en France au xv et au xvi siècle (Paris, 1900-1904); W. P. Courtney, A Register of Nation. Bibliography (2 vols., London, 1905); J. P. Edmond, Catal. of Early Printed Books in the library of the Society of Writers to the Signet (Edinburgh, 1906); Ehwald, Handschr. u. Inkunabeln der Gymnasialbibliothek zu Gotha, 4to (Gotha, 1893); Will. J. van Eys, Bibliogr. des Bibles et des Nouv. Testaments en langue franc. (Geneve, 1900); John Ferguson, Some Aspects of Bibliography (Edinburgh, 1900); G. Fumagalli, Lexicon typographicum Italiae; Diction. geogr. d'Italie (Florence); Gravures sur bois tirees des livres francais du siècle (Paris, 1868); Konrad Haebler, Typenrepertorium der Wiegendrucke (Halle, 1905, 1908); idem, Typographie iberique du xv siècle (La Haye, Leipzig, 1902; 87 plates); idem, Bibliografia iberica del siglo xv. (La Haye, Leipzig, 1903); Otto Gunther, Die Wiegendrucke der Leipziger Sammlungen (Leipzig, 1909); Alb. Hubl, Die Inkunabeln der Biblioth. des Stifles Schotten in Wien (Vienna and Leipzig, 1904); L'Imprimerie hors l'Europe, par un bibliophile (Paris, 1902); Ad. Knutgen, Incunabeln im kon. kathol. Gymnasium zu Heiligenstadt (Heiligenstadt, 1888); Paul Lacombe, Livres d'heures imprimes au xv et au xvi siècle (Paris, 1907); Ad. Lange, Peter Scheer (Leipzig, 1864); H. O. Lange, Analecta bibliographica (Copenhagen, 1906); F. Madan, The University Press at Oxford (Oxford, 1908); Baron F. del Marmol, Diction. des filigranes (Namur, 1900); Joh. Jac. Merlo, Ulrich Zell, ed. Otto Zaretzky (Cologne, 1900); Henri Monceaux, Les Le Rouge de Chablis, calligraphes et miniaturistes, graveurs et imprimeurs (Paris, 1896); R. A. Peddie, Printing at Brescia in the 15th Century (London, 1905); Marie Pellechet, Catal. general des incunables des bibliotheques publiques de France (Paris, 1897); M. A. Pericaud, Bibliogr. lyonnaise du siècle (Lyons, 1861); J. Philippe, Origine de l'imprimerie a Paris (Paris 1885); Guillaume Fichet, Introduction de l'imprimerie a Paris (Paris, 1892); Henr. R. Plomer, Hist. of English Printing,1476-1898 (London, 1900); G. R. Redgrave, Erhard Ratdolt and his work at Venice (London, 1894); Fr. Reiber, De primordiis artis imprimendi ac praecipue de inventione typographiae Harlemensi (Berol., 1856); Ph. Renouard, Bibliographie des impressions et des oeuvres de Josse Badins A scensius,1462-1535 (Paris, 908); Seymour de Ricci, A Census of Caxtons (London, fol., 1909); Duc de Rivoli, Bibliogr. des livres a figures venit., 1469-1525 (Paris, 1892); Paul Schwenke, Untersuch. zur Geschichte des ersten Buchdrucks, herausgeg. von der konigl. Bibl. zu Berlin (1900); L. C. Silvestre, Marques typographiques (Paris, 1853); Day. E. Smith, Rara arithmetica, in the library of Geo. Arthur Plimpton of New York (Boston and London, 1908); Henri Stein, Manuel de bibliographie gen. (Paris, 1897); C. H. Timperley, Encyclopaedia of Literary and Typographical Anecdote (2nd ed., London, 1842); Tijdschrift voor boek-en bibliotheekwezen (Antwerp, Ghent, 1903); Leon Vallee, Bibliogr. des bibliographies (Paris, 1897); Herm. Varnhagen, Eine Sammlung alter italien. Drucke der Erlanger Universitdtsbibliothek (Erlangen, 1892); Ernst Voullieme, Der Buchdruck Kolns bis zum Ende des XV. Jahrhunderts (Bonn, 1903); idem, Die Incunabeln der kon. Universitdts-Bibl. zu Bonn (Leipzig, 18 94); W. H. J. Weale, Bibliographia Liturgica (London, 1886).

The titles of other works on the invention, progress and process of printing, &c., may be learnt from the lists of books on such subjects in the works already quoted. Also the catalogues of secondhand booksellers, as Jos. Baehr (Frankfurt), Harrowitz (Berlin), Leo S. Olschki (Florence), Bern. Quaritch and W. M. Voynich (London), Jaques Rosenthal, Ludw. Rosenthal (Munich), &c.

(J. H. H.) II. - Modern Practical Typography The printing surfaces used in the production of books and newspapers, apart from woodor process-blocks and casts, and apart also from such surfaces as are obtained by means of the Linotype and kindred machines, are made up primarily of an aggregation of separate types, each representing a letter, mark or sign, though the actual surface employed on the printing press is frequently a duplicate copy made by a process of stereotyping or electrotyping.

Material Characteristics of Type

A fount consists of a proportioned quantity of each of these letters and signs of any one particular body and face. It therefore contains single letters, both capitals (" upper case ") and small letters (" lower case "), diphthongs, ligatures, such as ff, fl, accents, points, figures, fractions, commercial signs such as @, £, " peculiars " such as *, t and leaders (...), together with quads (pieces of metal which do not print, but are used to compensate for the shortness of occasional lines, as at the end of a paragraph), and spaces which separate words. A fount may thus have about 275 characters or sorts, about loo of them consisting of italic letters, points and figures.

The numbers of the different sorts vary with different languages, and even with the style of different writers, the works of Charles Dickens, for instance, making unusually heavy demands on the vowels, while the writings of Lord Macaulay run with like persistence on the consonants. Type-founders determine the proportions of the different sorts according to a bill of type, or scheme, either numerically, when the basis of the computation is the number of lower case m's (or of A's, in the case of display type used for headings) or by weight. In the second method a fount of 125 lb of Roman type includes, on one scheme, 8 oz. of E, M, C; 9 oz. of T; 8 lb of e; 5 1b each of a, b, n, o, t; and so on down to 3 oz. of z. A fount of body-letters, that is those used for the reading matter of books and newspapers, as made up by one British type-founder, contains capitals 9% by weight, small capitals 4%, figures 6%, lower case letters, points and leaders 56%, spaces 15% and quads io %; rules, accents and fractions not being supplied except in new complete founts or when specially asked for. A rule for estimating the quantity of type required for a page is to divide the number of square inches it contains by 4, when the quotient represents approximately the weight of type in lb. But for large founts 25% and for small ones 4q% should be allowed in addition, on account of unused type in the cases, which cannot be completely set.

For many years it was a favourite idea with inventors, especially those who were not practical printers, that great economy might be gained in composition by the use of word characters, Logotypes. or " logotypes," instead of single letters. The constant recurrence of certain words such as " the," " and," " is," suggested that they, as well as affixes and suffixes like ad-, ac-, -ing, -ment, should be cast in single pieces instead of being set up with their component letters. Such logotypic printing was used in 1785 in the London Daily Universal Register, which three years later became The Times, but it has never found general favour. The chief practical objection is that it involves the use of cases with an inconveniently large number of boxes. The greater the variety of characters the more " travel " of the compositor's hand over the cases is necessary for picking them up, and by so much is the speed of his work retarded.

Each of the parts of a type has a technical name. In fig. 1, representing the capital letter M, the darkest space a, a, a, a, is called the face; and only that part of the type touches the paper in printing. The face is divided into the stem, T y p e.

2

marked I, which comprises the whole outline of the type M; the serifs, or the horizontal lines marked 2, which complete the outline of the letter; the beard, consisting of the bevel or sloping part marked b, b, and the shoulder or flat portion below b. The shank is the entire body of the letter d, the front part (that shown) being known as the belly and the corresponding part behind as the back. The spaces at h and h are the counters, which regulate the distances apart of the stems in a line of type. The hollow groove extending across the shank at e, e is the nick, which enables the workman to recognize the direction of the type and to distinguish different founts of the same body. The absence of this simple expedient would retard the operation of hand-setting up by fully one-half. The earliest type-founders did not know the use of the nick. If a part of the face overhangs the shank, this part is called the kern, but kerned letters are avoided as much as possible. The groove g divides the bottom of the type into two parts called the feet. An impression from that part of a type on which it stands would be as Types must be perfectly rectangular, the minutest deviation rendering them useless. Any roughness at the sides is called burr, and any injury to the faces a batter. Types which have the face cast in the middle of the shank, as a, c, e, m, &c., and thus leave an open space above them corresponding to that below, caused by the beard, are known as short Species o f letters. Those whose stem extends to the top of the shank, as b, d, f, &c., are called ascending letters. Those that have a stem extending over the shoulder, as g, p, &c., are called descending letters. Those that are both ascending and descending, and extend over the whole of the shank, as Q and j, are long letters. Small letters and figures cast upon the upper part of the shank, as I a , are called superiors; those very low down on the shank are inferiors, as H3. Types that are very heavy and massive in appearance are called fat faced; those that are fine and delicate, lean faced. A type whose face is not in proportion to the depth of the shank (e.g. a small pica cast on a pica body) is a bastard type. Types of are various sizes, from those used for the smallest pocket bibles to those used for large placards, and the sizes are classified according to the dimensions of their ends or bodies. In a given fount the length of the end of the type which bears the face is the same for all characters, but the Type. width varies, an i for example being narrower than a w. Each body has a distinctive name, but it used to be a confusing and inconvenient anomaly that types made by different founders, though called by the same name, were not of precisely the same size. The long primer of one maker, for example, was 89 lines to the foot, of another 89k, and of a third 92. This inconvenience was remedied in America by the founders agreeing to adopt a uniform pointsystem; the pica of 0 . 16604 in. was taken as a standard, six picas being o996 in., and w,is divided into twelve parts or points of 0 . 013837 in., other types being cast as multiples of one of these points, and specified according to the number of them they contained. This system, with the same basic unit, has been adopted by British FIG. I. - Finished Type.

typefounders, though not to the exclusion of older sizes, and it has been extended to regulate the thickness or set of types, and also the position of the faces on the bodies as regards alignment. The Didot point-system, used in France, is based on a point of 0.376 mm., the English point being o35145 mm. The following are specimens of the principal bodies of ordinary British and American types, with their corresponding appellations on the point-system, the fast five being now mainly for display purposes: The Encyclopaedia Brit The Encyclopaedia Britannica, The Encyclopaedia Britannica, The Encyclopaedia Britannica, i rth The Encyclopaedia Britannica, i ith The Encyclopaedia Britannica, ttih edition The Encyclopaedia Britannica, nth edition The Encyclopaedia Britannica, ttth edition The larger type used in the body of this work is to-point, and the smaller 8-point.

The height of types is i z in. Those lower than the standard dimensions are said to be " low to paper," and if surrounded by higher types will not give perfect impressions. Spaces and quads are 4-in. high for direct printing, but for stereotyping are cut rather higher (0.83 in.).

According to the purpose for which they are used, types are divided into two classes - book type, including Roman and Italic; and job type, including a multitude of fanciful forms of. letters, chiefly founded on the shape of the Roman and Italic letters, and intended to be more prominent, delicate, elegant, &c. It is impossible to enumerate all the varieties of the latter class, as additions are being constantly made and once popular styles always going out of fashion. The leading varieties are the antiques, which are Roman letters with strokes of nearly uniform thickness, as M; sanserifs or grotesques, which have no serifs, as M; blacks, as Ai; and scripts, which represent the modern cursive or Italian handwriting, as, Black letter is now only a jobbing type in English-speaking countries, although it was the first character used in printing. It is still used in Germany, with certain modifications, as the principal text-letter for books and newspapers. A comparison of the numerous reproductions that have been issued of Caxton's works with any modern line of black letter will show how greatly the form and style have been altered. The present style of Roman type dates only from about the first quarter of the 18th century. Previously the approved shape was as follows Printing has been defined to be the act, art, or practice The use of this type was revived by Charles Whittingham, nephew of the founder of the Chiswick Press, about 1843, and it has since become a favourite form, under the name of old style. Some of the punches cut by the first notable English type-founder, William Caslon (1692-1766), have been preserved, and types are being constantly cast from them. Nearly all founders now produce modernized old style.

In this connexion reference may be made to the modern revival of artistic book printing in England by William Morris and others influenced by him. This development took definite form in the founts and books of the Kelmscott Press, which is distinguished by the use of three founts designed by Morris. The Troye and Chaucer founts, both Gothic, are best fitted for ornamented medieval works, while the Golden or Roman fount is without the exaggerated contraction of form laterally, the exaggerated use of thick and thin strokes, and the vicious stroke-terminations common to modern founts. It is a type of full body, designed in careful relation to the up and down strokes, and resting upon solid serifs, as with Jenson's fount, for instance, but in detail more allied to fine penmanship or black letter. The Vale books, often classed with the Kelmscott, may be counted with them so far as they also are controlled by one designer, from the important matter of type, decoration and illustration, to that of " build " and press work. The first Vale book in which these conditions were achieved is Milton's MinorPoems (1896). In this is employed the Roman type, known as the Vale fount, designed by Charles Ricketts, which differs from the Kelmscott fount in a greater roundness or fullness of body, and in a modification of details by the conditions of type-making. The second fount used in the Vale issues, first employed in The Plays of Shakespeare (1896), is less round in body, more traditional in detail and lighter in effect.

Manufacture of Type

Type is made of an alloy, known as type-metal, which consists chiefly of lead, with smaller amounts of antimony and tin. The exact proportions vary in different countries and foundries and with the size and quality of the type, but in general more than 60% is lead and the antimony predominates over the tin. Sometimes small quantities of other metals, such as copper and iron, are added. Large letters, such as are used for bills and posters do not come within the province of the type-founder; they are made of wood, chiefly rock maple, sycamore, pine and lime, planed to the right size and engraved by special machinery.

The earliest printers made their own types, and the books printed from them can now be distinguished with almost as much certainty as handwriting can be identified. The modern printer matrices for has recourse to the type-founder. The first step in T ype. the making of type, according to the old method, is the production of a matrix. The letter is cut on the end of a piece of fine steel, forming the punch (fig. 2), which is afterwards hardened. A separate punch is required for each character in every fount of type, and the making of them requires great care and delicacy in order that the various sorts in a fount may be exactly uniform in width, height and general proportions. During the process of its Gt. FIG. 2. - Punch. FIG. 3. - Drive. FIG. 4. - Matrix.

manufacture, the punch is frequently tested or measured by delicate gauges to insure its accuracy, and from time to time it is examined by means of a smoke-proof, that is, an impression obtained by holding it in a flame and stamping it on paper. When the letter is perfect, it is driven into a piece of polished copper, called the drive or strike (fig. 3). This passes to the justifier, who makes the width and depth of the faces uniform throughout the fount. They must then be made to line exactly with each other. When completed, the strike becomes the matrix (fig. 4), wherein the face of the type is made. But matrices are now commonly produced by the aid of an engraving engine which copies a standard drawing of each letter on any desired scale, and they may be obtained from existing founts by electrotyping.

Until well into the 19th century types were cast from the matrices in small hand-moulds, the output from which with a skilful worker was about 400 letters an hour. The mould Type consisted of two portions fitting closely to each other and containing the matrix with a space to receive the metal for the shank; holding it in his left hand the operator poured in the metal with his right, and after jerking it at arm's length, to bring the metal well up against the matrix, opened the two halves and threw out the type. In 1838 David Bruce, Junr., of New York, a Scotsman, who had migrated to America, invented a machine to perform substantially the same operations; this increased the rate of production to about too a minute for ordinary sizes, and with improvements and modifications remained a standard appliance for 40 years after its introduction. The metal, kept molten by a small furnace, was injected by a pump into the mould, which at every revolution of the axle came up to the spout of the pump, received a charge of metal, receded, opened, and discharged the type. But neither the hand mould nor the Bruce machine produced finished type. To the bottom of each there was attached a wedge-shaped jet (fig. 5), somewhat similar to that on a bullet cast in a hand mould. This had to be picked off by hand; the burr on the shoulder of the types had also to be rubbed off, and a groove had to be cut in the bottom to form the feet. Many efforts were made to devise machines which should perform these operations and produce finished type, one of the most satisfactory being that patented by Henry Barth, of Cincinnati, in 1888, but the principle of the divided mould which opened to discharge the type was generally retained. A new principle, however, was adopted by Frederick Wicks (1840-1910) in his rotary type casting machine, which FIG. T e was developed into a practical apparatus in London 5' yp just at the end of the 19th century, and which with jet.

is able to produce finished types, ready to be despatched to the printer without any inspection or treatment beyond packing, at a Points.

2-line small pica. 22 Great primer. 18 English

14 Pica 1 2 Small pica.. it Long primer 10 Bourgeois 9 Brevier 8 Minion 7 Nonpareil. . 6 Rub Pearly ipm a?l ??? ? 10 IOW ?C 6 III _ nnrl ?? uu 1 8 ? ?

IIIIII ?„ "..,: ,!, i UVN. .

continuous rate of 60,000 an hour. It consists of a horizontal mould wheel, 20 in. in diameter, contained in the casing D (fig. 6), in which are cut loo radial slots, each having a matrix at its inner end. These slots thus form moulds, and are of varying width according to the letter each has to cast. Each wheel can only produce type of the particular body for which it has been cut, but by changing the matrices the moulds can be made to cast any description of face capable of being received upon the body. The wheel is rotated once in every six seconds, so that the slots are successively presented to a jet of molten type-metal, which_is pumped from the FIG. 6. - Wicks Rotary Type-casting Machine.

metal reservoir A by a pump B of special construction, and forced out at high pressure through a nozzle under the shield C. As soon as any particular slot has passed the jet and been filled with metal, a cam-action comes into play and gradually pushes out the formed type. This operation is completed in half a revolution, the ejected type being taken up by carriers mounted on a continuous chain E, which is moved along exactly in step with the wheel. The carriers, which are of different sizes according to the particular letters they have to hold, are raised by a cam-action as they come opposite the slots to receive the types, but fall again at the point F, depositing the letters at the end of the race G. Each successive type thus dropped pushes its predecessors farther along the race until when the row contains zoo types - the product of two revolutions of the wheel - an attendant lifts the whole series off and places them on the plate H, one row below the other. Since the sequence of the letters is of course the same in each revolution, the result is that each vertical line on the plate consists of the same character, and each sort can be easily removed and packed in any required form for despatch to the printer. As soon as each slot has been emptied of its type, another cam begins to draw in the matrix towards the centre of the wheel, so that it is in as far as it can go by the time the slot is again opposite the jet. To prevent a type from being drawn back with the matrix, the bead-cam K engages with the nicks which have already been formed on the front of the type-bodies by the operation of the machine. To ensure trueness and accuracy in the product, the conditions under which casting is conducted are maintained as uniform as possible. The composition of the type-metal alloy is kept constant; the temperature of the molten metal is carefully regulated by the aid of a pyrometer to about Boo° F., so as not to volatilize the antimony it contains; the pumps work up to a pressure of 900 lb to the square inch, and by the interposition of a reducing valve deliver the metal at the nozzle at a constant pressure of 200 lb; and the moulding slots are maintained at an equably cool temperature by an elaborate system of water circulation.

Type-setting by Hand

The types, received from the foundry in the packages called pages, containing about 8 lb, are placed in shallow trays called cases. These contain compartments T or boxes, each of which is appropriated to some particular sort or character. The cases when in use stand on frames or sloping desks. The case at the top is the upper case, and that below the lower case. The former contains 98 equal-sized boxes, appropriated principally to the capital and small capital letters; the latter has 53 boxes of various sizes, appropriated to the lower case sorts. The difference in the size of the boxes corresponds to the difference of quantity of letters in a fount, the lower-case e, for instance, having the largest box. As a man picks out from the boxes seldom less than 1500 letters an hour and distributes or replaces on the average about ©%?S/?,?-?/ /? '? -?- ®O`L? O?, ?. /? /0 /e %? /O ??? ?% w/?/??esT/ ? ? ' / O/?' ??/ ? ?[? %, -x/ ??p??/p/?

/?? ?o'??. ???i?

rya /? ?? ie ° H ? ui FIG. 7. - Type-case.

5000 an hour, it is necessary that the most economical allocation of the boxes should be adopted. The system of allocating the various types is called the lay of the case; one plan is illustrated in fig. 7.

The types when taken from the cases are arranged in lines (composed or set up) in an instrument called a composing stick, made of iron, brass or gun metal. The slide in the middle is movable so as to accom modate varying lengths of lines. The compositor fixes the " copy " or document which he has to repeat in type, in a convenient place before his eye. In his left hand h., holds the composing stick, and with the thumb and first finger of the right hand lifts the letters from the boxes, and arranges them in the composing stick, every letter, point or sign being picked out separately. In this operation he is much assisted by the use of a setting-rule, a thin brass or steel plate which, being removed as successive lines are completed, keeps the type in place. When so many words and parts of words as will nearly fill the line have been composed, it is made the exact length required by increasing or diminishing the space between the several words. This is called justifying the line and is effected by means of the spaces already mentioned. If the work is not " solid " - that is, if the lines are not close together - the strips of metal called leads or brasses are inserted between each. When the composing stick is filled, the type is lifted upon a galley, a shallow tray of wood or metal, two or three sides of which are flanged, for the purpose of supporting the type when the galley is slightly inclined. Stickful after stickful of type is placed on the galley until it is full. The matter is then fastened up, a proof taken at the proof press, and the work of the reader or corrector of the press begins (see Proof-Reading). The proof, marked with the necessary corrections, is given back to the compositor, in order that he may make the required alterations in the type.

The type, being duly corrected, is made up into pages of the required length (unless the author has desired to see proof in slip). It is then imposed, that is, the pages are arranged in such a manner that, when printed and the sheet folded, they will fall in due numerical sequence. The impression from any arrangement of pages will be the reverse of that in which they are laid down. If a four-page newspaper be opened and spread out with the first page uppermost, it will be found that on this side the order of pages is 4, 1; when turned the pages are 2, 3. The type pages must be ranged in the reverse way, as I, 4; 3, 2. Thus the fourth page is placed alongside the first, because both must be printed together on the outside; the third page is to the left, and the second to the right, because in books the odd page - the verso - is always to the right. For a quarto a sheet of paper is folded twice, that is once across its breadth and then once in a perpendicular direction down the middle. It contains four leaves, and if these are printed on both sides eight pages. The two sides of a sheet are called the outer and inner formes respectively. A sheet of octavo is folded three times, making 8 leaves or 16 pages. The size of a book depends not only upon the number of times the sheet has been folded, and described accordingly as 4to, 8vo, 12mo, &c., but upon the size of the sheets. The dimensions of the papers commonly used in bookprinting are: imperial, 22 X30 in.; super royal, 201 X271; royal, 20 X 25; medium, 19 X 24; demy, 171 X 221; double crown, 20 X 30; double foolscap, 17 X 27; post, 151 X 19 2. Hence to say merely that a book is a quarto gives no precise indication of its dimensions, as a quarto of one size of paper may be smaller than an octavo of another; it is also necessary to know the size of the sheets of which it is composed.

When a printed book is opened, it will be found that at the foot of certain pages there is usually a letter and at the foot of another a letter and a figure, as B, B 2; farther on another letter and another letter and figure. On going through the book it will be seen that the letters are in regular alphabetical order, and occur at regular intervals of eight, twelve, sixteen, &c., pages. These designate the several sheets of which the book is composed and are called signatures, so that a sheet may be designated B, and the pages of which it consists are thereby sufficiently indicated. (Occasionally, numbers are used instead of letters.) These signatures assist the binder in folding, as they occupy a certain specified place in each sheet; hence to ascertain if the sheet has been folded properly it is only necessary to examine the position of the signature. The binder also is thus assisted in gathering or collating together the sheets of a volume in proper order. Signature A is omitted, because it would be on the title or first page, and would be both unnecessary and unsightly. By old custom J, V and W are discarded, I and J, U and V being originally used indiscriminately, by printers, while W was written UU or VV. When the alphabet is exhausted, a new one is begun, distinguished by a figure precedent, as 2 B, 2 C, &c.

The pages of types are arranged in proper order on a flat table, covered with stone or metal, called the imposing stone, and are then. ready to be made into a forme, that is, into such a state that they can be securely fastened up and moved about.

The forme is enclosed in an iron frame or chase, sub-divided by a cross bar. The portions of the type are separated by furniture, which may be of metal or wood or both. It is of the same height as the chase, but lower than the type, and therefore does not print, but forms the margin of the printed pages. As the sides of the two sections of the formes are pieces of furniture of a tapering shape, called side-sticks, and at the top and bottom corresponding pieces, called foot-sticks. Small wedges, called quoins, are inserted and driven forward by a mallet and a shooting-stick, so that they gradually exert increasing pressure upon the type. Other mechanical means for locking up are also occasionally adopted. When sufficiently locked up, the whole is quite as firm and portable, however many thousands of pieces of metal it may consist of, as if it were a single plate, and is ready for use on the printing press, either directly or in the form of a stereotyped or electrotyped copy.

After use the type undergoes the operation of distributing, which is the converse of composing; it is de-composing the forme and returning the several letters to their proper boxes in. the case. The forme is first washed over with an alkali or other detergent to remove the ink from its surface, and then laid down on the imposing surface, unlocked and damped; this assists the cohesion of the type, after the chase, furniture, sidesticks, &c., are removed. The compositor then takes in his left hand, supported by a setting rule, a portion of type in lines, and with the right hand takes a word or so between the finger and thumb, letting each letter drop separately into its proper box. The types are held upside down, that is, with the nicks uppermost; hence the letters of each word are read from left to right like ordinary matter when printed, but the words are of course dealt with in the inverse order.

Type-setting by Machine

The above method of producing a printing surface depends entirely upon hand labour, but it has long been an object of inventors in connexion with printing to perfect a mechanical system by which hand-work may be done away with both in setting type and in distributing it after use. The first step in this direction was the construction of composing machines in which the compositor put together types in the required order, not by lifting them one after another from his " boxes " and placing them by hand in his " stick," but by operating a keyboard which liberated them from magazines and assembled them in the order in which the keys had been struck. Such machines were followed as a natural correlative by distributing machines which performed the converse operation. Then the idea occurred of avoiding distribution altogether, by returning the printing surface to the melting-pot and using the metal over again to produce an entirely new printing surface as required, instead of sorting the types into their various kinds to be set up again either by hand or by machine. There are two main solutions of this problem. One is to manufacture ordinary movable types at a cost that is less than that of distribution, when it obviously becomes advantageous to treat the formes, after use, as old metal and return them directly to the melting pot without distribution. In 1900 The Times xxvii. 18 began to adopt this method, thus securing the advantage of fresh new type for each issue. In its offices for several years type made by the Wicks casting machine was set up by composing machines, and after being used in making the necessary stereotype plates was returned to the foundry to be melted and recast. The other solution depends upon the employment of apparatus which are in effect combinations of type-setting and type-casting machines, and may be divided into two broad classes: (a) those in which, by the operation of a keyboard, letters are translated into metal types which appear as a product for use in the printing-press, not singly, but cast into cornplete bars or lines of type; and (b) those in which the final product is separate types, delivered made up into lines of the required length. The former class is exemplified by the Linotype, the Typograph, and the Monoline machines, the latter by the Lanston Monotype, the Tachytype and the Goodson. In machines of the Linotype class, which have come into extensive use, especially for newspaper printing, it is impossible to make corrections or alterations in the line of type after it has been cast. The smallest change, such as the addition of a comma, involves the resetting and recasting of a whole line, while, if two or three words have to be added or removed, the compositor may have to recast a considerable number of lines, perhaps a whole paragraph. Machines of the second class, like the Monotype, which has been employed for setting up the present edition of the Encyclopaedia Britannica, appeal rather to the book printer, though the Monotype is used by such newspapers as The Times (London) and the Sun (New York). They have the advantage that corrections can be made as with hand-set type; but for newspaper work the fact that the manipulation of the keyboard does not, as with the Linotype, directly produce a printing surface but merely a punched strip of paper, which has then to be passed through a separate casting machine, inevitably introduces some delay. This is a matter that must be taken into account in the hurried conditions under which a daily paper is produced, when the shortest possible interval must elapse between the time when the latest news is received and the actual printing is begun. A machine invented by Mr H. Gilbert-Stringer is designed to combine the advantages of the Linotype and Monotype machines by casting at a single operation separate types properly arranged in lines and uniformly spaced. Up to the point where the matrices are ranged in a line ready for the bar of type to be cast, the mechanism may be identical with that of the Linotype; from that point each matrix is separately pushed into a mould which is automatically varied in size to suit the size of the particular letter it is casting, and also casts the spaces between the words (determined by the use of a modified Schuckers wedge-space), so that when all the individual types and spaces in the line are assembled after casting they exactly fill the line. The machine requires only one operator, and while one line is being cast the matrices which have formed the preceding one are being distributed to the magazine, as in the Linotype, and the following one is being set up. The matrices differ from those of the Linotype in that the face is impressed on their broad flat surface, not on the thin edge.

Composing Machines

An early attempt to make a machine for setting up ordinary foundry type was patented in England by Dr William Church in 1822. In the machine of Young and Delcambre, which was used in London for composing the Family Herald in 1842, and was the forerunner of the Kastenbein machine adopted in The Times office in 1869, the types were arranged in tubes placed either vertically or horizontally, and the lowest or endmost letter was, when wanted, ejected from the tube by a pusher actuated by a finger-key. It then passed down the channels of a guide-plate to a common point, whence it was pushed forward by a reciprocating motion to the line of previously composed matter and divided into lines of the required length. To the same group belong the Fraser machine, the Hattersley and the Empire, also known in America as the Burr. Another group of machines developed from the rotary composer was invented by Alexander Mackie of Warrington in 1871, and used in the office of the Warrington Guardian. In this the types were arranged in vertical tubes round a rotating disk, and the letters were automatically selected by a strip of paper previously punched with holes through which feelers passed and caused the desired type to be ejected upon a travelling band. This device of using a paper strip perforated in different positions to correspond to different letters was patented by Felt in 1860 (U.S. Patent Spec. No. 28,463), and he also utilized it for effecting distribution, the " dead " or used type being dealt with by another machine through which the paper strip was run in the reverse direction. This quality, however, was not so valuable as it might appear at first sight, since any correction, however simple, of necessity made the perforated paper ineffectual as a guide in distribution. The Thorne machine, exhibited in the Paris Exhibition of 1878, was a development of the principle of a rotating disk, but the types, which were contained in a vertical cylinder, were selected by touching keys in the ordinary manner. When liberated they fell upon a rotating table, whence they were deflected by a finger upon a travelling band and delivered into the composing race. The American Simplex machine resembles the Thorne very closely. The Wicks composing machine, again, adopts a different principle from both the above groups. The types are ejected upon a straight race set at an angle of 45°. Thus each has to travel a different distance from the other - a result which the inventors of the Delcambre group of machines were at pains to avoid; and when several keys are struck together so as to give a combination like " and," the several types delivered to the race follow each other in proper succession to the point of assembly, the letter whose key is nearest to the left side of the keyboard preceding those whose keys are more to the right.

The Paige composing, justifying and distributing machine - an American invention - is one of the most remarkable pieces of mechanism ever put together. It contains 18,000 parts, and the patent specifications form an imposing volume. It is operated by keys in the ordinary way, but automatic mechanism advances the ejected letters in words, spaces them and inserts the lines in the " galley " with " leads " if desired; at the same time other mechanism automatically distributes dead matter and refills the tubes which contain the supplies of types. Two machines were made, and are said to have done good work, but the cost of construction and the complicated nature of the mechanism made the apparatus impracticable commercially, and the two that were made are now on view as mechanical curiosities, the one in the Columbia Institute and the other in Cornell University. The Paige machine dispensed with the guide-plate of the Delcambre group, the letters being ejected on a plane along which a driver passed at intervals and swept the type into a receiving race on the left of the machine. The Dow composing and justifying machine, a later American invention, adopts this characteristic of the Paige, but has two drivers meeting at the centre of the plane which receives the letters. The types having been swept to the centre by these, a vertical driver forces them downwards into a vertical receiver. When a line has been set a justifying key is touched, the vertical line passes to a horizontal position, and is driven forwards to a point where apparatus measures it, and having removed temporary brass spaces replaces them with others selected from a series of ten different thicknesses.

Distributing Machines

There are two main classes of distributing machines. One, which is exemplified by the Delcambre or the FIG. 8. - Nick System of Distribution (Simplex Machine).

Fraser machine, is operated by a keyboard; the compositor strikes the keys corresponding to the letters of the printed matter he wishes to distribute, and thus opens gates through which the types pass and find their way down a guide-plate to their proper tubes. The other comprises a number of machines which agree in requiring the type to be specially nicked for their use. Each type has its own particular combination of nicks, and the receptacles in which the type is collated are provided at their entrances with wards corresponding to these nicks, so that each type can only enter the one receptacle for which its nicks are arranged (fig. 8). In some cases, as in the Empire and the Dow, the distributor is a separate machine; in others, as the Thorne and the Simplex, it is combined with the composing machine in such a way that the two work simultaneously: Linotype. - An enormous amount of ingenuity has been expended on the Linotype, which was developed into a practical machine by FIG. 9. - Diagram of Linotype Machine.

Ottmar Mergenthaler, of Baltimore, though two of its elements - the solid bar of type and the wedge space - were invented by others, the former by T. W. Smith, of the Caslon Foundry, and the latter by Jacob W. Schuckers, of Washington. The following will give a general idea of its working: In the magazine A (fig. 9) are a series of matrices, formed with the characters in intaglio on one edge, which are discharged by gates, operated from the keyboard D into the chutes E, and thence upon the travelling belt F; this delivers them upon a revolving pusher wheel by which they are set up in proper order in the assembler block G. Above the assembler block is a space magazine, and from this the space key J releases a space bar, when desired, which drops into place in the line. As the matrices are forced into the assembler block they move to the left against the resistance of a sliding abutment, thus being held compactly in place in the line. As soon as a complete line is set up, the compositor operates a hand lever by which the assembler block and matrices are raised to the level of a horizontal slide, where the line is grasped between two jaws and carried to the left, and lowered into position opposite the mouth of the mould wheel K. Here the justification of the line is effected by means of an upwardly moving plunger which drives the wedge-shaped spaces, seen in fig. Io, into the line, and thus expands it to the exact length required. The matrices are then locked firmly in a vice with the characters opposite the mouth of the mould. At this time the pump plunger in the melting pot M (fig. 9) is forced downwards by mechanism actuated by suitable cams on the driving shaft, and a jet of molten type metal is ejected into the mould and against the characters on the matrices, thus casting the bar or " slug." The cast bar is next forced, by a revolution of the wheel K, between a pair of knives, by which it is trimmed, and into a galley, where it is pushed along by a packer arm and placed beside its fellows in a column ready for use.

It is next necessary to distribute the matrices to the magazines, in order that the operation of the machine may be carried on continuously. The matrices and spaces are raised from the vice and brought opposite a bar R, which carries on its under side a series of undercut ribs corresponding to the teeth which are shown at the edges of the V-shaped notch in the top of the matrices (fig. Io) and the matrices are pushed on to this bar so as to be suspended by the - -=--=_T= FIG. Io. - Line of Matrices with Spaces. of its own magazine chan nel, where they are inter rupted in such a manner that the matrix is unsupported and drops into the magazine for further use. It will thus be apparent that there is a constant circulation of the matrices through the machine, and the composing of one line, the casting of another and the distribution of a third are all carried on at the same time, which adds greatly to the speed of the operation. The machine may be fitted with double magazine, which with double-letter matrices gives 360 characters or four faces ready for use, or even with three magazines, which provide for 540 characters or six faces, the movement of a hand lever bringing the desired magazine into use.

c 0 h o o

K °

o b

° u p.

"CE 0

0 1 ° 0

° s ?

0 ° p ° o 0

0

Lanston Monotype

In the Lanston apparatus there are two distinct machines, a ribbon-punching machine and a type-casting and composing machine. The first of these is a small device resembling a typewriter, having a number of keys, 257 in all, corresponding to all the characters used in a fount of type, with some additions representing certain movements to be performed by the composing machine. These keys, when depressed, admit compressed air to a plunger or combination of two plungers working punches, whereby perforations are made in a strip of paper fed step by step through the machine. Most of the keys make two perforations, though some a single one only. These perforations stand in a 0 r o 0 5 I o o 0 0 t o I 0 i 0 0 n 0 0 O 0 0 o o o [[Spac[ Fig. Ii]]. - Perforated Strip. which the number of units of space still vacant in the line bears to the number of space-types which the line contains. For example, if there are ten space-types and i l ii of an inch of space remains to be filled, each space-type must be increased in thickness just i o o of an inch completely to fill the line. It is not necessary, however, for the operator to make this calculation, for he has only to consult the scale provided for this purpose, and is referred at once to the proper keys to punch the justifying perforations in the strip. Each time the space key is depressed a pointer rises one step against a cylindrical scale placed vertically in front of the machine, and when the operator has finished setting a line he presses a special key which causes the cylinder to rotate until it automatically stops with the required number at the end of the pointer. This number is in the form 4, and to complete the justification of the line the operator has only to depress the appropriate keys in the top two rows of the keyboard, in this case No. 3 of the top row and No. 4 of the second.

The ribbon thus prepared in the punching machine is used to control all the movements of the casting or composing machine. The matrices for making the type faces are formed in a plate about 3 in. square, and any character is brought opposite the casting point by the movement of the matrix-carrier in two directions, or rather by the resultant of two such independent movements. As the perforations for controlling the galley movements and those for justifying the line are necessarily made after the others in the perforating machine, and these operations must be provided for in the composing machine before the line is set up, the latter machine is so organized that the ribbon is passed through and the types are set in the reverse order to that in which the strip was punched. The perforated ribbon is wound from one wheel off to another, passing over the edge of a tracker board in which there are a number of holes corresponding to those which may occur in the ribbon, and each of these holes communicates by a tube with a small piston which controls some device for performing one of the various operations of the machine. As the ribbon passes over the tracker board, a jet of compressed air passes to the appropriate operating device whenever a ribbon perforation or any combination of them coincides with the proper holes. The two perforations on each transverse line control two stop pins which limit the movements of the matrixcarrier to bringing the proper matrix to the casting-point, while the justifying perforations set in motion devices which open the space mould to cast space type of the exact size to effect the proper justification of the line, and the galley perforation starts the feeding device which moves the galley for the next line of type. The matrixcarrier may be readily removed and another carrying a different style or size of type substituted therefor.

In modern printing it is often the case that the printing surface actually used in the press (see Printing) is not the original forme of type, whether consisting of separate type set up by machine or by hand, or of Linotype slugs, but a reproduction of it made by electrotyping or by stereotyping. Of these two processes the former is the slower and the more costly, but it produces the better results, since electrotyped plates are capable of yielding a larger number of sharp impressions than are stereotypes.

Electrotyping

In making an electrotype, a moulding composition consisting mainly of wax with a little blacklead, is poured when molten into a shallow metal tray, and, when it has set, its surface is brushed over with blacklead and polished. An impression of the forme, which is also blackleaded, is next taken in the wax while it is still warm, often by the aid of a hydraulic press, and the mould thus obtained, after being separated from the forme, undergoes a process of building up, which consists of dropping heated wax upon those portions which require to be more deeply sunk in the finished type, that is, upon those places where " whites " are to appear in the print. The face of the finished mould is then carefully covered with blacklead, which is a conductor of electricity, and the whole is immersed in an electrotyping bath, where copper is deposited on the blackleaded portions by means of the current from a Smee's battery or a dynamo machine. When the deposit, or shell, is sufficiently thick, it is disengaged from the wax mould, backed with a metal which resembles type-metal but contains a larger proportion of lead, and trimmed and planed. For use in rotary presses curved electrotypes may be produced.

Stereotyping

The great advantage of stereotyping is in connexion with the production of newspapers, where the desideratum is the printing off of a large number of copies in a short time. For this purpose, in the first place, rotary machines must be employed, and stereotyping affords a ready means of obtaining curved printing surfaces to fit their cylinders. It is true that stereotyping is not absolutely necessary for rotary printing, since it has been found possible to print from movable type clamped on the cylinders in curved frames known as " turtles." But to set up duplicate formes of type is impracticable, and, therefore, this device does not permit the utilization of more than one press. Herein lies the second great advantage of stereotyping, for it enables the printer to obtain as many replicas of each forme as he desires, and thus not only to employ a number of machines simultaneously, but also to " dress " each of them with several duplicates of the same forme, as is required in the later developments of high-speed presses.

o

°

°"0 °

° A 0

D 0

c s o

°

0

° t 0 0

0

° 0 0

0

° °0

0

0

t o 0

The first attempt at making stereotypes was by means of moist clay into which, after it had been impressed with the type and baked, molten type metal was poured; but this method did not yield a curved plate. Later the clay was replaced by papier-mâché, which being flexible can be bent to the required shape. This papiermache, known as flong and composed of several sheets of paper united by a paste capable of withstanding a high temperature without burning, is moistened and laid over the forme of type, into which it is well pressed either by beating with a long-handled brush or, according to the more modern and expeditious method, by being passed through a moulding press. The flong is next dried, for which ribs. They are next pushed still farther towards the right of the machine into a box having ribs engaging the notches in the side of the matrices, but with downwardly inclined grooves crossing these ribs, by which the shoulders at the upper end of the space bars are allowed to descend, and the spaces are thus dropped out of line and fall through a chute into the space-box from which they originally came. The matrices are pushed still farther to the right, where their teeth slide along the distributor bar T, being carried by two screws which engage opposite sides of the matrices and keep them separated so that they hang loosely from the distributor bar. The ribs of the distributor bar are so arranged as to support each matrix by one or more pairs of teeth until it arrives opposite the mouth 0 o 0 o s 0 I o transverse line across the strip, as shown in fig. I I, and their relative position in the line varies with the particular key operated. At the end of each word a spacing key is struck, and suitable perforations are made in the strip, and as the end of a line is neared, a bell rings to warn the operator, who, by looking at a line scale facing him on the machine, is enabled to see how many determine whether another word or syllable can be set If not, it then becomes necessary to provide units of be filled, and can then space remain to proper space-type to justify or fill out the line, which is done by increasing the width of the normal space-types already provided for in the proportion purpose it is either placed with the type in a heated chamber covered with blankets which absorb the moisture, or is removed from the type and heated separately. Sometimes these two methods are used in combination; processes have also been devised for pressing the Hong dry upon the type, when [subsequent drying becomes unnecessary. For casting a plate the matrix thus prepared is fastened in a casting mould or box curved to the circumference of the cylinder of the press, and molten stereo-metal (a softer form of type-metal) is poured upon it. During this process the box stands upright, but while the matrix is being placed in position it lies horizontally, a swivel mounting enabling it to be readily turned. After time has been allowed for solidification, the cast is taken out, stripped from the matrix and adjusted on a " finishing saddle," where a machine cuts off the superfluous metal from its upper end and forms a bevel by which it can be clamped on the press. It is then placed face downwards in another machine which shaves out and smooths its interior surface, and finally it is set face upwards, while men with chisels remove protruding pieces of metal that might take ink and print.

Up to the end of the 19th century the general method of stereotyping was as outlined above, though of course there were variations in different establishments. The time required to produce a plate, as distinct from making the matrix, was about 1 or 1 z minute, and the process was expensive in labour since it required the employment of half-a-dozen men. This time may seem short enough, but when plates are needed by the score, as may be the case with a paper having a large circulation, the delay entailed by the preparation of the whole number by this method becomes of serious importance. Means were therefore sought to reduce it by the adoption of automatic mechanism. In the Autoplate machine, invented in America by Henry A. Wise Wood, and first used by the New York Herald in 1900, the operation of casting is performed automatically from the time the matrix is put in position until the finished plate is ready to be placed on the printing press, and from a single matrix four plates z in. thick, or seven or eight 4 in. thick, can be produced every minute, by the aid of three men only. The casting is done against a horizontal cylinder or core, the interior of which is cooled by water. Below it is a frame or " back " carrying the matrix. This back has an up and down movement of about six inches, and when it is in its topmost position there is a semicircular space between it and the core equal in length, breadth and thickness to the plate which has to be cast. Molten metal having been injected into this space by a pump, there is a pause of a few seconds to permit of solidification, and then the back falls, bringing away the matrix with it. Immediately afterwards the cylinder makes a half turn, and presents what was previously its upper half to the matrix for another cast. The first cast is taken with it as it turns, and is then pushed along from the top of the core against two rotating saws which trim its edges. Next it comes under a shaving arch, where it pauses while its interior surface is smoothed to proper thickness, and finally water is directed against its back, to cool it without wetting its printing face. The Junior Autoplate is a simpler machine which does not perform so many operations. In it the casting core is vertical, not horizontal, but the matrix is still automatically stripped from the plate, the casts are made alternately on the two halves of the cylinder, and as one plate is being removed another is being cast. The machine also automatically cuts off the sprue which is left on the top of the plate as it stands in the casting box. About three plates a minute are produced, but they are not delivered completely finished, and have to undergo several further operations before they are ready to be placed on the press. The Double Junior consists of two Junior Autoplates served from a common melting-pot, and its capacity is six plates a minute, with two matrices; and another machine, the Autoshaver, has been devised which can shave, cool and deliver that number of plates automatically, no labour being required except to take the plate from the casting machine and place it on the Autoshaver.

See Practical Printing, by John Southward and Arthur Powell (5th ed., London, 1900); Modern Printing, by John Southward (London, 1898) The American Handbook of Printing, by Edmund G. Gress (New York, 1907); History of Composing Machines, by John S. Thompson (Chicago, 1904); Traite de la typographie, by Henri Fournier (4th ed., Paris, 1904); " Type Casting and Composing Machines," by L. A. Legros, Proc. Inst. Mech. Eng. (London, 1908); " Modern Stereotypy and the Mechanics of the Newspaper," by Henry A. Wise Wood, Journ. Franklin Inst. (Philadelphia, 1910). (J. So.; H. M. R.)