book I have dylygently oversen and duly examyned to thende that it

be made acordyng unto his owen makyng. For I fynde many of the sayd bookes whyche wryters have abrydgyd it and many thynges left out. And in somme place have sette certayn versys that he never made ne sette in hys booke, of whyche bookes so incorrecte was one brought to me vj yere passyd whyche I supposed had ben veray true and correcte. And accordyng to the same I dyde do enprynte a certayn nombre of them whyche anon were sold to many and dyverse gentyl men of whome one gentylman cam to me and said that this book was not accordyng in many places unto the book that Gefferey chaucer had made. To whom I answered that I had made it accordyng to my copye and by me was nothyng added ne mynusshyd. Thenne he sayd he knewe a book whyche hys fader had and moche lovyd that was very trewe and accordyng unto hys owen first book by hym made; and sayd more, yf I wold enprynte it agayn he wold gete me the same book for a copye, how be it he wyst wel that hys fader wold not gladly departe fro it. To whom I said, in caas that he could gete me suche a book trewe and correcte yet I wold ones endevoyre me to enprynte it agayn. And thus we fyll at accord. And he ful gentylly gate of hys fader the said book and delyverd it to me, by whiche I have corrected my book."

Besides revising his text, Caxton added illustrations. There are twenty-four of these, but several are made to do duty twice over, a common custom with early printers. Thus the "poor parson" and the "doctor of physick," the "somnour" and the "Franklin," are represented by the same cuts; while the large illustration depicting the pilgrims sitting at supper at a round table does duty in some later publications for the "Assembly of the Gods."

As might have been expected, such a book became very popular, and is now consequently very rare. Besides a few more or less imperfect copies, only one perfect one is known, now in the library of St. John's College, Oxford, which unfortunately has the cuts rudely daubed with colour. When perfect, the book should contain 312 leaves, the first being blank. Chaucer's _Troilus and Creside_ and _Hous of Fame_, as well as a little tract of six leaves called the _Curial_, were also printed about this time.

September 2, 1483, is another definite date, for on that day Caxton finished an edition of Gower's well-known poem, the _Confessio Amantis_, or _Lover's Confession_, written by command of Richard II., who, meeting the poet rowing on the Thames, near London, invited him into the royal barge, and after much conversation requested him to "book some new thing." The book is a folio of 222 leaves, of which no less than four are blank, and only one copy is now in existence in which these blank leaves have been preserved. Otherwise the book is not uncommon, though nearly every copy is imperfect. An extremely fine copy, wanting only the blank leaves, is in the library of Shrewsbury School, and is mentioned here to correct an error of Blades, who goes out of his way to state, "The copy ascribed by N. Carlisle to Edward VI. Grammar School, Shrewsbury, is not from the press of Caxton."

The _Life of the Holy and Blessed Virgin Saint Winifred_ is a small folio of 16 leaves, printed about this date. Caxton states that he "reduced" this book into English, but there is some difficulty as to the source from which he took it. The life of the celebrated Welsh saint was written in Latin in the twelfth century by Robert, Prior of Shrewsbury, and this Caxton may have translated, but as no copy of the manuscript is now in existence the point cannot be determined. Only three copies of the printed book are known.

The book called _Caton_ was translated by the end of December, 1483, and must therefore have been printed at the beginning of 1484. The _Catonis Disticha_ was the best known school-book of the middle ages, and with the _Donatus_, was the groundwork of Latin learning over Europe. About 1480 a certain Daniel Church added a few Latin precepts to the original book, which acquired the name of _Parvus Cato_, and after his time the two are generally found together as _Cato, parvus et magnus_. In the second half of the fifteenth century it was, as Caxton himself tells us, "translated in to Englysshe by Mayster Benet Burgh, late Archdeken of Colchestre and hye chanon of saint stephens at westmestre, which ful craftly hath made it in balade ryal for the erudicion of my lord Bousher, sone and heyr at that tyme to my lord the erle of Estsex." Of this version Caxton printed three editions, which have already been noticed, but the present is a different and considerably larger work. It contains, besides the "disticha" and moral maxims, very extensive glosses or commentaries containing "histories and examples," translated by Caxton from a French original.

It is a folio of 80 leaves, of which four are blank and usually wanting. About twelve copies are known, and a good example is in the Lenox Library, New York.

On the 31st of January, 1484, Caxton issued the _Booke whiche the Knyght of the Toure made to the enseygnement and techyng of his doughters_. This work was compiled about the year 1371 by Geoffrey de la Tour-Landry, a literary knight of celebrated family, and was translated by Caxton, "at the request of a noble lady which hath brought forth many noble and fair daughters, which be virtuously nourished." In his preface he advises "every gentleman or woman having children desiring them to be virtuously brought forth to get and have this book," though it would in these days be considered anything but suitable for young persons,--or for the matter of that, for their elders.

The _Fables of Aesop_ was issued on March 26th, the first day of the year 1484. This is certainly one of the finest and rarest amongst the books which Caxton printed. It begins with a large full-page frontispiece containing a figure of Æsop similar in treatment to those occurring in some foreign editions. This is found only in the copy at Windsor Castle. In the text there are no less than one hundred and eighty-five wood-cuts, the work of two or perhaps even three different engravers, one of whom apparently cut the illustrations to the second edition of the _Game of Chesse_. One illustration is engraved in quite a different manner from the rest, and was probably cut hurriedly to replace one accidentally lost or broken, and has an appearance much more resembling modern work than the others, which are simply the ordinary heavy black outline cuts of the period. A complete copy of the book should contain 144 leaves, the last two being blank, and the leaves are numbered. It was twice reprinted in the fifteenth century by Richard Pynson at London, and these two reprints are even rarer than the original, one copy of each being known, and both of them imperfect.

The only perfect copy known of Caxton's edition is in the King's Library at Windsor, and was one of the very few books retained when the Royal Library was handed over to the nation by George IV. A note on the fly-leaf shows the reason for this. "Left to his Majesty by the late Mr. Hewett of Ipswich in Suffolk and delivered to Mr. Allen by Philip Broke, Esq. and Sir John Hewett, Bart. to present to the King." It is in magnificent condition and uncut. The British Museum was fortunate enough to be able to purchase a copy in 1844, which, with the one imperfection of not having the frontispiece, is in as fine condition as the Windsor copy, and in an early sixteenth-century binding by John Reynes. The third and last copy is in the Bodleian, to which it was presented in 1680, with other Caxtons, by Moses Pitt, a London bookseller. It is imperfect, wanting in all about twelve leaves.

A curious broadside was published about this time, which is generally known as the _Death-bed Prayers_. It contains two prayers to be said by a priest at the bedside of dying persons, and the only known copy, which was formerly in the Spencer Library, was found bound up with a copy of the _Pilgrimage of the Soul_.

The _Order of Chivalry_, which was printed in the reign of Richard III., may be ascribed to 1484. The author of the book is not known, but it was translated from the French, and agrees exactly with a manuscript in the Bibliothèque Nationale, entitled _L'Ordre de Chivallerie_, beautifully illuminated, and written in Flanders for Edward IV. Caxton used in this book and the _Aesop_ a large floriated initial letter A, the only large ornamental capital which he seems to have possessed.

Five copies are known, two in the British Museum, two in the Spencer collection at Manchester, and one in a private library in America. The book, which is a small quarto, should contain 52 leaves, the first and last being blank. One copy in the British Museum and one at Manchester are complete as regards text, but neither has both blanks.

The _Golden Legend_, Caxton's most important work, was finished, so far as regards the translation, on the 20th of November, 1483. In the second prologue the printer tells us that when beginning the translation the magnitude of the task and the heavy expenses of printing made him "halfe desperate to have accomplissd it," and he proposed to put what he had already done to one side and leave the work. The Earl of Arundel, however, encouraged him to proceed, not only by promising to take a certain number of copies when finished, but by the offer of an annual gift of a buck in summer and a doe in winter. Thus assisted, Caxton finished his translation and printed the book, and some idea of the task involved may be gathered from the fact that the work consists of 894 printed pages, each page containing two columns of 55 lines. It is illustrated with a frontispiece, eighteen large and fifty-two small wood-cuts. The translator compiled his version from three sources, for he tells us that he had beside him "a legende in frensshe, another in latyn and the thyrd in englysshe."

The French edition which Caxton used has been clearly identified in a curious manner. In one or two places it contains bad misprints which Caxton translated blindly. In the life of St. Stephen the words "femmes veuves" have been misprinted "Saine venue," which Caxton renders "hole comen," in spite of the words making no sense. In the life of St. Genevieve "a name" occurs in place of "a navire," which appears in the English version as "at name" in place of "by ship." This French version is of great rarity, the only two copies known being in the British Museum and the Cambridge University Library.

Fortunately, the _Golden Legend_ is one of the commonest of Caxton's books, though every copy is more or less imperfect. The finest is that which formerly belonged to Lord Spencer, which was made perfect as regards the text with leaves from other copies, and is, with the exception of these leaves, very large and in fine condition. In 1577 it belonged to Robert Hedrington, who appears to have owned many Caxtons.

The three books which follow the _Golden Legend_, and which are all dated 1485, are of very great interest. These are the _Morte d'Arthur_, the _Lyf of Charles the Great_, and the _History of Paris and Vienne_, all printed in folio.

Sir Thomas Malory's _Morte d'Arthur_, that cycle of stories connected with King Arthur and the Knights of the Round Table, and as Sir Walter Scott called it, the best of all English romances, is perhaps the most interesting volume that Caxton ever printed. Two copies only, one perfect, the other wanting a few leaves, are now known. The first has a long and interesting pedigree. It first appeared at Dr. Bernard's sale in 1698, when it produced the magnificent sum of two shillings and ten pence, passing into the vast library of the Earl of Oxford. Osborne, the bookseller who bought that library _en bloc_, sold the volume to Bryan Fairfax for five pounds, and in 1756 it passed with the whole of his library to Mr. Child, the banker, at a valuation of two pounds, twelve shillings, and six pence. While safely preserved at Osterley at the beginning of last century, it raised the most covetous feelings in the breasts of the two great Caxtonian collectors, Lord Spencer and his nephew, the Duke of Devonshire, who both made overtures for its acquisition. It had almost been ceded to the latter in exchange for some work of art, when it was discovered that it could not legally be parted with, and it remained at Osterley until 1885, when Lord Jersey's library was sold. At this sale it was purchased by Mr. Pope for a sum little under two thousand pounds, and left England for America, where it still remains. The second copy was obtained by Lord Spencer in 1816 at the sale of the library of Mr. Lloyd of Wygfair. Both copies are in very fine condition. The complete book consists of 432 leaves, the first being blank. There are 38 lines to a page, and as these run straight across, instead, as is so often the case, being made up into two columns, the effect of the whole, with the wide margins, is very striking. Sir Thomas Malory's translation from the French was finished in the ninth year of King Edward IV.,--that is, about 1470,--but apparently no manuscript of it is now in existence.

The _Life of the noble and christian prince, Charles the Great_, was translated by Caxton from an anonymous French version compiled at the request of Henry Bolomyer, Canon of Lausanne. In it the various stories and legends relating to Charlemagne have been gathered together from various sources. Caxton finished his translation on the 18th of June, when he had nearly finished the printing of the _Morte d'Arthur_, and the printing of the book was finished on the 1st of December.

The only copy known, which is perfect with the exception of the last blank leaf, is in the King's Library in the British Museum.

The moment Caxton had finished the translation of _Charles the Great_ he set to work on another short romance, the _History of the knight Paris and the fair Vienne_. This he finished on the 31st of August, and the book was printed by the 19th of December. Like the last, only one copy, and that quite perfect, is known; and it is also in the King's Library in the British Museum.

It seems very probable that at an early date these two books were bound together, but either before or on their coming into the possession of the Earl of Oxford they were bound separately. They agree entirely in size and typographical particulars, both having 39 lines to the page in two columns.

The _Paris and Vienne_ was reprinted in 1492 by Gerard Leeu at Antwerp in small folio, with illustrations. He reprinted also in the same month the _History of Jason_, and in the year following the _Chronicles of England_. The apathy in book production which seems to have immediately succeeded Caxton's death may have encouraged him to attempt printing for the English market, but his own death while his edition of the _Chronicles_ was passing through the press put an end to the trade. He printed one other English book, the _Dialogue of Salomon and Marcolphus_, of which one copy exists. This, like the rest, may have been copied from an edition printed by Caxton, but if so, all traces of it have disappeared.

No dated book of 1486 is known, but several may be ascribed to this date. First the _Directorium Sacerdotum_, or _Pica_, a work compiled by Clement Maydeston, containing the rules for adapting the calendar to the services of each week in accordance with the thirty-five varieties of the Almanac. Of this book, so interesting to liturgical students, but one copy is known, now in the British Museum, a library, however, to which it should not rightly belong. The volume formed part of the collection bequeathed to the Cambridge University Library by Dr. Holdsworth in 1648, but it was stolen from there in or shortly before 1778, and soon afterwards "bought of a man introduced by Dr. Nugent" by William Bayntun, Esq., of Gray's Inn, after whose death it came into the possession of King George III., and passed with the rest of the King's Library into the British Museum. At the beginning of the book a single leaf containing a large wood-cut has been inserted which does not really belong to the volume. In the centre is a half-length figure of Our Lord with the hands crossed. Behind the head and shoulders is the cross, and on either side the spear and the reed with the sponge. Below is the text of an _Indulgence_, which in this case has been cut out, while round the whole is a framework composed of twenty-eight small square compartments, each containing some emblem of the crucifixion. These early English prints, several of which exactly similar in treatment are known, go under the name of the _Image of Pity_.

The _Directorium_ is a folio of 160 leaves, the first, which is wanting in the only known copy, having been most probably blank.

About this year the first edition of Bonaventura's _Speculum Vitae Christi_ was issued, remarkable for its illustrations. These, though not large, are much more graceful in design and better in execution than any which preceded them, and are clearly the work of a new engraver. It is a curious fact that in neither edition which he printed did Caxton use the full series of these cuts, for odd illustrations appear in later books which clearly belong to the set, but which had not been made use of before. Besides the regular series, a few smaller cuts occur, much ruder in execution.

These belong to a set cut for an edition of the _Horae ad usum Sarum_, but the early editions of this book are known only from fragments, so that we cannot ascertain how many there were in the original series.

Several of these _Speculum_ cuts reappear in the _Royal Book_, a translation of _La Somme des vices et vertus_, published very shortly after. This book at present enjoys the distinction of having brought the highest price hitherto paid for a Caxton, a copy having been sold (March 20, 1902) for the sum of £2,225. The history of this particular copy is an interesting one. It belonged early in the seventeenth century to Thomas Archer, parson of Houghton Conquest, Bedfordshire, who bequeathed it to the Church Library of St. John's, Bedford. This library was afterwards moved to St. Paul's Church, Bedford, and about 1840 transferred to the Bedford Literary Institute. The council of this Institute, entirely ignoring their moral obligations, determined to make money out of so valuable a book, and not only did they do so, but they also did their best to destroy one of the very few existing evidences of Caxton's work. The book when I examined it several years ago was in its original binding, tooled with stamps which we have many reasons for believing belonged to Caxton himself. This Bedford book afforded the strongest proof of all, for the boards of the binding were lined with unused copies of one issue of Caxton's 1481 _Indulgences_. Of these there had originally been four, two at each end, but two had been abstracted. When the book was sold the remaining two were taken out and sold separately, thus destroying for ever a most valuable piece of evidence. This book, together with one of the _Indulgences_, is now in a private library in America.

A few years ago Mr. Robert Proctor, working in the library of New College, Oxford, found in the binding of a book two small slips of vellum with some printing upon them in Caxton's type No. 5. These turned out to be portions of a leaf of a hitherto unknown Caxton, an edition of the _Donatus melior_, revised by Mancinellus, printed in folio. They are also the earliest specimens of Caxton's use of vellum. The date of the book would be about 1487.

In May, 1487, Caxton finished the printing of the _Book of Good Manners_, which he had translated from the French at the request of one of his friends, William Pratt, a mercer of London who had lately died. The original book was written by Jacobus Magnus or Jacques Legrand, the author of the _Sophologium_, and was evidently popular, for it was frequently reprinted, no less than four other English editions having been issued in the fifteenth century. Caxton's edition is a small folio of 66 leaves, and three copies, all in public libraries, are known. The finest is in the Cambridge University Library, and another, also perfect, is in the Royal Library at Copenhagen. This latter, which was sold by auction in 1682 for the sum of two shillings, was purchased by the Copenhagen Library in 1743 for one guinea. The third copy, wanting some leaves, is at Lambeth.

So far Caxton had worked entirely with his own materials and without any assistance from outside. His work had been confined to the most ordinary kind of printing, which required no special trouble and no great variety of type or ornament. The close of the year, however, saw a change in this respect, and the first influences of the French press, which were gradually later on to assume such large proportions, began to make themselves felt.