In what city and from what printer Caxton received his earliest training in the art of printing has been a much debated question amongst bibliographers. The only direct assertion on the point is to be found in the lines which form part of the prologue written by Wynkyn de Worde, and added to the translation of the De proprietatibus rerum of Bartholomaeus Anglicus, issued about 1495.

As Wynkyn de Worde was for long associated with Caxton in business and became after his death his successor, it seems impossible to put aside his very plain statement as entirely inaccurate. William Blades, in his Life of Caxton, utterly denies the whole story. "Are we to understand," he writes, "that the editio princeps of Bartholomaeus proceeded from Caxton's press, or that he only printed the first Cologne edition? that he issued a translation of his own, which is the only way in which the production of the work could advance him in the Latin tongue? or that he printed in Latin to advance his own interests? The last seems the most probable reading. But though the words will bear many constructions, they are evidently intended to mean that Caxton printed Bartholomaeus at Cologne. Now, this seems to be merely a careless statement of Wynkyn de Worde; for if Caxton did really print Bartholomaeus in that city, it must have been with his own types and presses, as the workmanship of his early volumes proves that he had no connexion with the Cologne printers, whose practices were entirely different."

The meaning which Mr. Blades has read into the lines seems hardly a reasonable one. Surely, the expression "hymself to avaunce" cannot apply to the advancement of his own interests, but rather to knowledge; nor can we imagine a sensible person who wished to learn Latin entering a printing-office for that purpose. It must rather apply to the printing itself, and point to the fact that when at Cologne he printed or assisted to print an edition of the Bartholomaeus in Latin in order to learn the practical details of the art.

It must also be borne in mind that in 1471, when Caxton paid his visit to Cologne, printing had been introduced into few towns. Printed books were spread far and wide, and some of Schoeffer's editions have inscriptions showing that they had been bought at an early date, within a year of their issue, at Bruges; but Cologne was the nearest town where the press was actually at work, and where already a number of printers were settled.

Blades adds as another argument the fact that no edition of a Bartholomaeus has been found printed in Caxton's type, but when starting as a mere learner in another person's office he could hardly be expected to have type of his own. But there is an edition of the Bartholomaeus, which, though without date or name of place or printer, was certainly printed at Cologne about the time of Caxton's visit. It is a large folio of 248 leaves, with two columns to the page and 55 lines to a column. It is described by Dibdin in his Bibliotheca Spenceriana (Vol. III., p. 180), though with his usual inaccuracy he gives the number of leaves as 238. There is little doubt that the words of Wynkyn de Worde refer to this edition.

Cologne, as might be expected from its advantageous position on the Rhine, was one of the earliest towns to which the art of printing spread from Mainz. Ulric Zel, its first printer, was settled there some time before 1466, when he issued his first dated book, and by 1470 several others were at work. The study of early Cologne printing is extremely complex, for the majority of books which were produced there contain no indication of printer, place of printing, or date. Some printers issued many volumes, and their names are still unknown, so that they can only be referred to under the name of some special book which they printed; as, the "Printer of Dictys," the "Printer of Dares," and so on.

M. Madden, the French writer on early printing, who had a genius for obtaining from plausible premisses the most utterly preposterous conclusions, was possessed with the idea that the monastery of Weidenbach, near Cologne, was a vast school of typography, where printers of all nations and tongues learned their art. He ends up his article on Caxton, as he ended up those on other early printers, "Je finis cette lettre en vous promettant de revenir, tÔt ou tard, s'il plaÎt À Dieu, sur William Caxton se faisant initier À la typographie, non pas À Bruges, par Colard Mansion, comme le veut M. W. Blades, mais À Weidenbach, par les frÈres de la vie commune."

As we know from Caxton's own statements, he had when at Cologne considerable leisure, which was partly employed in writing out his translation of Le Recueil, and like all literary persons, must have felt great interest in the new art. It was no longer a secret one, and there would be little difficulty for a rich and important man like Caxton to obtain access to a printing-office, where he might learn the practical working and master the necessary details.

The mechanical part of the work was not at that time a complicated process, and would certainly not have taken long to master. Caxton no doubt learned from observation the method of cutting and the mechanism of casting type, and by a little practical work the setting up of type, the inking, and the pulling off the impression.

At the close of 1471 Caxton returned to Bruges, and presented to the Duchess of Burgundy the manuscript of the Recuyell of the Historyes of Troye, which he had finished while at Cologne. This work, which had been undertaken at the request of the Duchess, proved to be exceedingly popular at the court. Caxton was importuned to set to work on other copies for rich noblemen. The length of time which the production of these copies would take reminded him of the excellent invention which he had seen at work at Cologne, that art of writing by mechanical means, "ars artificialiter scribendi," as the earliest printers called it, by which numerous copies could be produced at one and the same time.

Mr. Blades, in common with almost every writer, assumes that printing was introduced into Bruges at a very much earlier date than there is any warrant for supposing. He speaks of Colard Mansion as having "established a press shortly after 1470 at Bruges." Other writers put back the date as much as three years earlier, confusing, as is often the case, the date of the writing of a book with the date of its printing. Colard Mansion's name does not occur in a dated colophon before 1476, in his edition of the French translation of a work of Boccaccio, and we have no reason to suppose that he began to work more than two years at the outside before this date. In the guild-books at Bruges he is entered as a writer and illuminator of manuscripts from 1454 to 1473, so that we are certainly justified in considering that he did not commence to print until after the latter date. Other writers have brought forward a mysterious and little known printer, Jean Brito, as having not only introduced the art into Bruges, but as being the inventor of printing. An ambiguous statement in one of his imprints, where he says that he learned to print by himself with no one to teach him, refers more probably to some method of casting type, and not to an independent discovery, and his method of work and other details point almost certainly to a date about 1480. Some of his type is interesting as being almost identical with a fount used a few years later in London.

Now, there is one very important point in this controversy which appears to have been quite overlooked. Caxton, we may suppose, learned the art of printing about 1471 at Cologne, the nearest place to Bruges where the printing-press was then at work. But, say the opponents of this theory, his type bears no resemblance to Cologne type, so that the theory is absurd. It must, however, be remembered that in the interval between Caxton's learning the art and beginning to practice it printers had begun to work in Utrecht, Alost, and Louvain. If he required any practical assistance in the cutting or casting of type or the preparation of a press, he would naturally turn to the printers nearest to him,—Thierry Martens, with John of Westphalia at Alost, or to John Veldener or John of Westphalia (who had moved from Alost in 1474) at Louvain.

Caxton's preparations for setting up a printing-press on his own account were most probably made in 1474. His assistant or partner, Colard Mansion, by profession a writer and illuminator of manuscripts, is entered as such in the books of the Guild of St. John from 1454 to 1473, when his connexion with the guild ceases. This may point to two things: he had either left Bruges, perhaps in search of printing material, or had changed his profession; and the former seems the most probable explanation.

If Caxton was assisted by any outside printer in the preparation of his type, there can be little doubt that that printer was John Veldener of Louvain. Veldener was matriculated at Louvain in the faculty of medicine, July 30, 1473. In August, 1474, in an edition of the Consolatio peccatorum of Jacobus de Theramo, printed by him, there is a prefatory letter addressed "Johanni Veldener, artis impressoriae magistro," showing that he was by that time a printer. He was also, as he himself tells us, a type-founder, and in 1475 he made use of a type in many respects identical with one used by Caxton.

In body they are precisely the same, and in most of the letters they are to all appearance identical; and the fact of their making their appearance about the same time in the Lectura super institutionibus of Angelus de Aretio, printed at Louvain by Veldener, and in the Quatre derrenieres choses, printed at Bruges by Caxton, would certainly appear to point to some connexion between the two printers.

Furnished with a press and two founts of type, both of the West Flanders kind and cut in imitation of the ordinary book-hand, William Caxton and Colard Mansion started on their career as printers.

Unlike all other early printers, Caxton looked to his own country and his own language for a model, and although in a foreign country, issued as his first work the first printed book in the English language. Other countries had been content to be ruled by the new laws forced upon them by the revival of learning. Caxton then, as through his life, spent his best energies in the service of our English tongue. The Recuyell of the Hystoryes of Troye, a translation by Caxton from the French of Raoul Le Fevre, who in his turn had adapted it from earlier writers on the Trojan war, was the first book to be issued.

The prologue to the first part and the epilogues to the second and third contain a few interesting details of Caxton's life. That to the third contains some remarks about the printing. "Therefore I have practysed and lerned at my grete charge and dispense to ordeyne this said booke in prynte after the maner and forme as ye may here see, and it is not wreton with penne and ynke as other bokes ben to thende that every man may have them attones. For all the bookes of this storye named the recule of the historyes of troyes thus enprynted as ye here see were begonne in oon day, and also fynysshed in oon day."

The wording of this sentence, which is perhaps slightly ambiguous, has caused several writers to fall into a curious error in supposing that Caxton meant to assert that the printed books were begun and finished in one day. His real meaning, of course, was, that while in written books the whole of a volume was finished before another was begun, in printed books the beginnings of all the copies of which the edition was to consist were printed off in one day, so also the last sheet of all the copies would be printed off in one day, and the whole edition finished simultaneously.

The Recuyell is a small folio of 352 leaves, the first being blank, and each page contains 31 lines, spaced out in a very uneven manner. The second leaf, on which the book begins, contains Caxton's prologue, printed in red ink. The book is without signatures, headlines, numbers to the pages, or catchwords.

Although a considerable number of copies—some twenty in all—are still in existence, almost every one is imperfect. The very interesting copy bought by the Duke of Devonshire at the Roxburghe sale in 1812 for £1,060 10s. which had at one time belonged to Elizabeth, wife of Edward IV., wanted the last leaf; Lord Spencer's wanted the introduction. Blades, it should be noticed, in his lists of existing copies of Caxton's books, uses the word "perfect" in a misleading way, often taking no notice of the blank leaves being missing, which are essential to a perfect copy, and often also omitting to distinguish between a made-up copy and one in genuine original condition.

The finest copy is probably that formerly in the library of the Earl of Jersey, which was sold in 1885. It was described as perfect, and possessed the blank leaf at the beginning. Valued in 1756, when Bryan Fairfax's library was bought by Lord Jersey's ancestor, Mr. Child, at £8 8s., it produced the high price of £1,820.

The next book to appear from the Bruges press was the Game and playe of the Chess, "In which I fynde," as Caxton says in his prologue, "thauctorites, dictees, and stories of auncient doctours philosophres poetes and of other wyse men whiche been recounted and applied unto the moralite of the publique wele as well of the nobles as of the comyn peple after the game and playe of the chesse."

The original of the work was the Liber de ludo scacchorum of Jacobus de Cessolis, which had been translated into French by Jean Faron and Jean de Vignay, both belonging to the order of preaching friars, but who worked quite independently of each other. Caxton appears to have made use of both versions, part of his book being translated from one and part from the other.

It is a considerably shorter book than the Recuyell, containing only 74 leaves, of which the first and last were blank. Like the last, it is a folio, with 31 lines to the page. It is not a very scarce book, as about twelve copies are known, but of these almost every one is imperfect. The best copy known is probably that belonging to Colonel Holford, of Dorchester House, which still remains in its old binding, and another beautiful copy was obtained by Lord Spencer from the library of Lincoln Minster, the source of many rarities in the Spencer collection. The story has often been told how Dibdin, the well-known writer of romantic bibliography, persuaded the lax Dean and Chapter of Lincoln to part with their Caxtons to Lord Spencer. We must, however, give even Dibdin his due, and point out that he was quite ignorant of the transaction, which was carried out by Edwards, the bookseller. The letter from Lord Spencer to Dibdin is still in existence, in which he describes the new Caxtons he had acquired, carefully omitting to say through whom or from what source. This, however, Dibdin found out for himself some time after, and raided Lincoln on his own account. He issued a small catalogue of his purchases, under the title of A Lincoln Nosegay, and a few were bought by Lord Spencer, the remainder finding their way into the libraries of Heber and other collectors.

The last book printed by Caxton and Mansion in partnership at Bruges was the Quatre derrenieres choses, a treatise on the four last things, Death and Judgment, Heaven and Hell, commonly known under the Latin titles of De quattuor novissimis or Memorare novissima, and later issued in English by Caxton as the Cordyale.

In this book first appears Caxton's type No. 2, which bears so strong a resemblance to the fount used by Veldener. The book is a folio of 74 leaves (not 72, as stated by Blades), and has 28 lines to the page. There is a certain amount of printing in red, which was produced in a peculiar way. It was not done by a separate pull of the press, as was the general custom, but the whole page having been set up and inked, the ink was wiped off from the portions to be printed in red, and the red colour applied to them by hand, and the whole printed at one pull.

For long but one copy of this book was known, preserved in the British Museum, and bound up with a copy of the Meditacions sur les sept pseaulmes, to be described shortly. Some years ago, however, another copy wanting two leaves was found, and it is now in a private collection in America.

This was the last book printed abroad with which Caxton had any connexion, and the new type used in it was no doubt specially prepared for him to carry to England. It contained far more distinct types than the first, which had 163, for it began with 217, which were increased on recasting to at least 254.

Supplied with new type and other printing material, Caxton made his preparations to return to his own country. The exact date cannot now be determined, but it was probably early in the year 1476. It is curious that just about this time one of the Cologne presses issued the first edition of the Breviary for the use of the church of Salisbury, the use adopted by all the south of England, and it may be that Caxton, who had had dealings with the Cologne printers, may have been connected in some way with its production and publication in England.

After Caxton had left Bruges his former partner, Colard Mansion, continued to print by himself. In Caxton's first type, which had been left behind at Bruges, he printed three books, Le Recueil des histoires de Troyes, Les fais et prouesses du chevalier Jason, and the Meditacions sur les sept pseaulmes. All three are in folio, with 31 lines to the page. As they are often confused by writers with books really printed by Caxton, and as they are produced from type which was at one time in his possession, they may perhaps merit a short description.

The Recueil contains 286 leaves, of which two are blank. Six copies are known, of which by far the finest was sold at the Watson Taylor sale in 1823 to Lord Spencer. It was then in its original binding and uncut, but Lord Spencer, who, like most collectors of his day, despised old bindings, had it rebound in morocco, and the edges trimmed and gilt. Another very fine copy, probably "conveyed" from some continental library, was purchased from M. Libri by the British Museum in 1844.

The Jason contains 134 leaves, of which the first and last two are blank. A magnificent copy, the only one in England, is in the library of Eton College, and there are two other copies, slightly imperfect, at Paris.

Of the third book, the Meditacions sur les sept pseaulmes, only one copy is known to exist. It is in the British Museum, bound up with a copy of the Quatre derrenieres choses, and is quite perfect. It contains 34 leaves, the last being blank.

Mansion continued for some time onwards to print at Bruges in the workshop which perhaps he had shared with Caxton, over the church porch of St. Donatus, but later in life seems to have been unsuccessful and fallen on evil times. The books which he then printed with such little success are now by the chance of fate the most sought for and valuable amongst the productions of the early continental press.