In 1476 Caxton returned to England and took up his residence in the precincts of Westminster Abbey, at a house with the sign of the "Red Pale" in the "almonesrye." This locality is thus described by Stow: "Now will I speake of the gate-house, and of Totehill streete, stretching from the west part of the close.... The gate towards the west is a Gaile for offenders.... On the South-side of this gate, King Henry the 7. founded an almeshouse.... Near unto this house westward was an old chappel of S. Anne, over against the which, the Lady Margaret, mother to King Henry the 7. erected an Almeshouse for poore women ... the place wherein this chappell and Almeshouse standeth was called the Elemosinary or Almory, now corruptly the Ambry, for that the Almes of the Abbey were there distributed to the poore."

In the account roll of John Estenay, sacrist of Westminster from September 29, 1476, to September 29, 1477, we find, under the heading "Firme terrarum infra Sanctuarium," the entry "De alia shopa ibidem dimissa Willelmo Caxton, per annum X^s." Another account-book, still preserved at Westminster, shows that in 1483 Caxton paid for two shops or houses, and in 1484 besides these for a loft over the gateway of the Almonry, described in 1486 as the room over the road (Camera supra viam), and in 1488 as the room over the road at the entrance to the Almonry (Camera supra viam eundo ad Elemosinariam). This latter was perhaps rented as a place to store the unsold portion of his stock.

The neighbourhood of the Abbey seems to have been a place much favoured by merchants of the Staple and dealers in wool, and this may have had something to do with Caxton's choice. He always continued to be a member of the Mercers' Company, and many of his fellow-members must have formed his acquaintance, or learned to esteem him, while he held his honourable and responsible post of Governor of the English nation in the Low Countries. Like himself, many were members of the Fraternity of our Blessed Lady Assumption and benefactors to the church of St. Margaret. The abbots of Westminster themselves were in the wool trade, and according to Stow had six wool-houses in the Staple granted them by King Henry VI. Some such special causes, or perhaps certain privileges obtained from Margaret, Henry VII.'s mother, who was one of the printer's patrons, must have made Caxton fix his choice on Westminster rather than on London, the great centre for all merchants, and which might have been supposed more suitable for a printer.

The first book with a date issued in England was the Dictes or Sayengis of the Philosophres, which was finished on the 18th of November, 1477. That Caxton should have allowed more than a year to elapse before issuing any work from his press seems improbable, especially considering the untiring energy with which he worked. On this point a curious piece of evidence is to be found in the prologue to the edition of King Apolyn of Tyre, printed by Wynkyn de Worde in 1510. Robert Copland, an assistant of De Worde and the translator of the book, says: "My worshipful master Wynken de Worde, having a little book of an ancient history of a kyng, sometyme reigning in the countree of Thyre called Appolyn, concernynge his malfortunes and peryllous adventures right espouventables, bryefly compyled and pyteous for to here, the which boke I Robert Coplande have me applyed for to translate out of the Frensshe language into our maternal Englysshe tongue at the exhortacion of my forsayd mayster, accordynge dyrectly to myn auctor, gladly followynge the trace of my mayster Caxton, begynnynge with small storyes and pamfletes and so to other."

Now, taking all the books printed by Caxton before the end of the year 1478, in number twenty-one, and considering that the first dated book was not issued until almost the end of 1477, and that Caxton had then presumably been in England for over a year, there does seem some reasonable ground for believing the statement of Copland, especially as there are amongst these early books a number which exactly answer to the description of "small storyes and pamfletes."

An exactly analogous case occurs in regard to the introduction of printing into Scotland. The first printer, Andrew Myllar, while preparing for the publication of the Aberdeen Breviary, which was issued at Edinburgh in 1509-10, published in 1508 a series of small pamphlets, consisting of stories and poems by Dunbar, Chaucer, and others. As might naturally be expected, such small books were especially liable to destruction, both on account of their size and the popularity of their subjects. It is not surprising to find that the majority have been preserved to us in single copies only. All the ten Edinburgh books are unique, and almost all the early Caxton quartos, so that it is impossible under these conditions to estimate what the output of Caxton's first year's working may have been. In writing of these earliest books, it will be perhaps best to take the folios first, and then the numerous small works, since, as they all agree so exactly as regards printing, they cannot be arranged in any definite order.

The first of the folios issued was most probably the History of Jason, translated by Caxton himself from the French version of Raoul Le Fevre immediately after he had finished those of the Recueil and the Game of Chess. The translation was undertaken under the patronage of Edward IV., with a view to the presentation of the book when finished to the ill-fated Prince of Wales, afterwards Edward V., "to thentent he may begynne to lerne rede Englissh." The book has every appearance of having been one of the very earliest issues of the Westminster press, and at the end of 1476 or beginning of 1477 the young prince would have been about four years old, a very suitable age to begin his education.

The book contains 150 leaves, of which the first and last are blank, and a full page has 29 lines. Like all early Caxtons, it has no signatures, which were not introduced until 1480; no headlines, which were rarely used; no numbers to the pages, which occur still more rarely; and no catch-words, which were never used at all.

As in all other early printed books, spaces were left for the insertion of illuminated initials at the beginnings of the chapters. Now, while in contemporary French, Italian, and Low Country books such spaces were often filled with the most gracefully designed and beautifully illuminated initials, rich in scrollwork and foliage, and ornamented with coats of arms or miniatures, there is not, so far as I know, any early English book in existence containing any attempt at such decoration. As a rule, the spaces were left blank as they came from the printer. In some cases, where the paragraph marks have been filled in by the rubricator, he has roughly daubed in the initial with his brush, making no attempt at ornament, or even neatness in the letter itself.

Seven copies of the Jason are still extant, the majority imperfect. By far the finest copy known was that sold at the Ashburnham sale in 1897, and which is now in a private collection in America. It is in the original leather binding as it issued from Caxton's workshop, and is quite uncut. This copy has generally been considered the finest Caxton in existence, and its various changes of ownership can be traced back for over two hundred years.

The great admiration which Caxton had for the work of Chaucer would no doubt make him anxious to issue it from his press as soon as possible, and we may therefore ascribe to an early date the publication of the Canterbury Tales and the translation of Boethius. The Canterbury Tales is a small folio of 374 leaves, with 29 lines to the page, and so rare that it is believed that no genuine perfect copy is in existence. Blades, in his account of the book, censures Dibdin for describing the copy at Merton College, Oxford, as imperfect, which, however, in Dibdin's time it certainly was, though through the kindness of Lord Spencer the missing leaves were afterwards supplied. One other copy, complete as regards text, is in the British Museum, having formed part of the library of George III. The Boethius contains 94 leaves, and is a much more common book. One copy is worthy of special mention, as it was the means of bringing to light the existence of three books printed by Caxton which up to that time were unknown. It was found by Mr. Blades in the old grammar-school library at St. Alban's, and he has left us an interesting account of its discovery. "After examining a few interesting books, I pulled out one which was lying flat upon the top of others. It was in a most deplorable state, covered thickly with a damp, sticky dust, and with a considerable portion of the back rotted away by wet. The white decay fell in lumps on the floor as the unappreciated volume was opened. It proved to be Geoffrey Chaucer's English translation of Boecius de Consolatione Philosophiae, printed by Caxton, in the original binding, as issued from Caxton's workshop, and uncut!" "On dissecting the covers they were found to be composed entirely of waste sheets from Caxton's press, two or three being printed on one side only. The two covers yielded no less than fifty-six half-sheets of printed paper, proving the existence of three works from Caxton's press quite unknown before." These fragments came from thirteen different books, and though other examples of one of the unknown works have been found, two, the Sarum Horae and Sarum Pica, are still known from these fragments only.

The Dictes or Sayengis of the Philosophres, though most probably by no means the first book printed in England, must still hold the important position of being the first with a definite date, November 18, 1477. The book was translated from the French by Lord Rivers, who had borrowed the original while on a voyage to the shrine of St. James of Compostella from a fellow-traveller, the famous knight Lewis de Bretaylles. Having finished his translation, he handed it to Caxton to "oversee" and to print, and the printer himself added a chapter "touchyng women." To this a quaint introduction is prefixed, in which it is pointed out that the gallant Earl had omitted the chapter, perhaps at request of some fair lady, "or ellys for the very affeccyon, love and good wylle that he hath unto alle ladyes and gentyl women." "But," continues Caxton, "for as moche as I am not in certeyn wheder it was in my lordis copye or not, or ellis peradventure that the wynde had blowe over the leef at the tyme of translacion of his booke, I purpose to wryte tho same saynges of that Greke Socrates, whiche wrote of tho women of grece and nothyng of them of this Royame, whom I suppose he never knewe."

It is curious that with one exception no copy of this first edition has a colophon. The copy in which it occurs was in Lord Spencer's library and is now at Manchester, but beyond this small addition, it varies in no way from the other copies. All the examples of the second edition, which was issued a few years later, contain a reprint of this colophon.

The Dictes when perfect contained 78 leaves (not, as stated by Blades, 76), of which the first and last two are blank, and though more than a dozen copies of the book are known, not one is quite perfect. In the library of Lambeth Palace is a manuscript of this work on vellum, copied from Caxton's edition, and dated December 29, 1477. It contains one poor illumination showing Earl Rivers presenting the copy to the Prince of Wales, afterwards Edward V. By the side of the Earl is an ecclesiastic, probably "Haywarde," the writer of the manuscript, and this figure has by some been considered, quite erroneously, to be intended for a portrait of Caxton.

The Dictes or Sayengis was followed shortly by another dated folio, the Morale Proverbes of Cristyne, issued on the 20th of February, 1478. It contains only four printed leaves, and three copies are known. The two verses added at the end of the book tell us of the author, translator, and printer, and are interesting as being the earliest printed specimen of Caxton's poetical attempts.

The author, Christine de Pisan, wife of Étienne Castel, was one of the most famous women of the middle ages. Left early a widow, with but narrow means, she had three children and her own parents to provide for. Being a woman of high attainments and considerable learning, she took up the profession of literature, and for many years worked incessantly. Les proverbes moraulx was written as a supplement to Les enseignemens moraulx, an instructive work addressed to her young son, Jean Castel, who was for some time in England in the service of the Earl of Salisbury.

Another point to be noticed about this book is the date, which here, fortunately, is quite clear. Among the early printers there is very considerable variation as to the day on which the new year began. Putting on one side the foreign and considering only the English printers, the dates narrow themselves to two, January 1st and March 25th, so that any date falling between these two may be in two different years, according to the habit of the printer. For instance, March 1, 1470, will really mean 1470 if the printer began his year on January 1st. If, on the other hand, he did not begin it until March 25th, the real date will be 1471.

Fortunately, Caxton frequently added to his dates the regnal year, which gives at once a definite solution. For instance, his edition of the Cordyale was begun the day after Lord Rivers handed him the manuscript, on February 3, 1478, and finished on March 24th following, in the nineteenth year of Edward IV. Now, the nineteenth year of Edward IV. ran from March 4, 1479, to March 3, 1480, so that Caxton's 1478 was really 1479, and his custom was, therefore, to begin his years on the 25th of March.

As has been said earlier, it is probable that Caxton began his printing in England with small pamphlets, and of these a considerable number have come down to our time, but as the majority are unique, it is impossible to conjecture how many may have utterly perished. The most considerable collection is in the University Library, Cambridge, which owns a series, originally bound in one volume, which was in the collection of Bishop Moore presented to the University in 1715 by George the First. This library was peculiarly rich in early English books; indeed, the great majority of those now at Cambridge formed part of it, and their acquisition was mainly due to the exertions of that much maligned person, John Bagford, whom Moore employed to search for such rarities, and who did so with conspicuous success.

Amongst these priceless volumes one stands out pre-eminent. It was until recently in an old calf binding, lettered on the back, "Old poetry printed by Caxton," and contained eight pieces, the Stans puer ad mensam, the Parvus Catho, The Chorle and the Bird, The Horse, the Shepe and the Goose, The Temple of Glas, The Temple of Brass, The Book of Courtesy, and Anelida and Arcyte. Five of these are absolutely unique; of the others a second copy is known.

These books must have caught the popular taste, for of several we find second editions issued almost at once. A second issue of the Parvus Catho is known from a unique copy belonging to the Duke of Devonshire. York Cathedral possesses the only known copy (with the exception of a few leaves at Cambridge) of the second edition of The Horse, the Shepe and the Goose, and a unique second edition of The Chorle and the Bird.

All these little poetical pieces agree typographically. They contain nothing but the bare text, and are without signatures, headlines, or pagination. Probably they were all issued at intervals of a few days, and not many printed, so that the second editions may have been issued only a few months after the first.

There are three other early quartos to be noticed, which are of quite a different class from those just mentioned. These are the Sarum Ordinale, the Propositio Johannis Russell, and the Infancia Salvatoris.

The Sarum Ordinale, or Pica, was a book giving the rules for the concurrence and occurrence of festivals, containing an explanation for adapting the calendar to the services of each week, in accordance with the thirty-five varieties of the almanac. This book would be in very considerable demand amongst those officiating in services, and would be a good method of attracting the attention of the priests to the new art, so that no sooner had the book been printed than Caxton struck off a little advertisement about it. "If it plese ony man spirituel or temporel to bye ony pyes of two and thre comemoracions of salisburi use enpryntid after the forme of this present lettre whiche ben wel and truly correct, late hym come to westmonester in to the almonesrye at the reed pale and he shal have them good chepe. Supplico stet cedula." The quaint Latin ending, "Pray don't tear down the advertisement," was then perhaps a customary formula attached to notices put up in ecclesiastical or legal precincts, but it might naturally be supposed that those most likely to damage or tear down advertisements would be uneducated people, who would be ignorant of Latin.

When the advertisement first came before the notice of writers on printing, the existence of the Ordinale was unknown, and it is amusing to read the various conjectures as to the buying of "pyes" hazarded by them. One of the most ingenious occurred in a letter from Henry Bradshaw to William Blades, which was that the syllable "co" had dropped out by accident, and that the word should read "copyes," and this appeared all the more probable, as the word "pyes" comes at the end of the first line, which is slightly shorter than the rest. This is the only specimen of an early English book advertisement known, though foreign examples are not uncommon.

The Propositio Johannis Russell is one of the very few pieces printed by Caxton dealing with current affairs or politics. It is the oration delivered at Ghent, early in 1470, on the occasion of the investiture of the Duke of Burgundy with the Order of the Garter. It has often been considered as one of Caxton's very earliest pieces,—perhaps printed at Bruges. Blades writes, rather vaguely: "To me it appears most likely that it was issued at Bruges at no long period after its delivery, and before Caxton's final departure for England. At that town, both with the subjects of the Duke of Burgundy and the 'English nation' there resident, it would secure a good circulation; not so if issued seven years after its delivery in another country."

It could not have been printed anywhere by Caxton before 1475, and everything seems to point to its having been printed at Westminster in 1476-1477, perhaps at the instance of the author himself, then Bishop of Rochester.

It is a little quarto tract of four leaves, and two copies only are known, one belonging to the Earl of Leicester at Holkham, the other, formerly in the Spencer Library, now at Manchester. This latter was originally bound up, apparently by mistake, amongst the blank leaves of a note-book used for miscellaneous manuscript treatises of the fifteenth century, which run on over the first and last blank pages of the tract itself. It appeared, unrecognized, at the Brand sale in 1807, and was described amongst the MSS., "A work on theology and religion, with five leaves at the end a very great curiosity, very early printed on wooden blocks, or type." It was bought by Lord Blandford for forty-five shillings, and purchased at his sale in 1819 by Lord Spencer for £126.

Blades speaks of it as in its original binding, a quite inexplicable mistake, for it was bound between the years 1807 and 1819 in resplendently gilt morocco, double, with gauffered gilt edges! The copy at Holkham, which used to be in an old vellum wrapper, has also been rebound, and the two inner leaves, by some unfortunate mistake, transposed.

Of the Infancia Salvatoris, a version of one of the smaller treatises among the apocryphal books of the New Testament, but one copy is known. It was in the celebrated Harleian Library, which was bought entire by Osborne in 1746. The Caxton collectors of the period seem to have passed it over, for it did not get sold, even at its very modest price, until three years later, when it was bought for the University Library of GÖttingen. It is still in its old red morocco Harleian binding, with Osborne's price—15—on the fly-leaf. Another note records, "aus dem Katalogen Thomas Osborne in London d. 12 Maij 1749 (No 4179) erkauft." Blades, in his description of the book, which he had not examined, conjectured that it was made up in three quires, the first of eight leaves, the second and third of six each, making in all twenty leaves, including a blank both at beginning and end. An examination of the water-marks of the paper shows that this was not the case, and that it consisted of two quires, the first of eight leaves, the second of ten, and that there were no blank leaves.

This tract, and the Compassio lamentationis Beate Marie Virginis, are the only two unique Caxtons in libraries outside England.

Some time towards the end of 1478 Caxton recast his fount No. 2, in which almost all the books so far mentioned were printed, and added a few extra types. With this new fount he printed the Margarita Eloquentiae of Laurentius de Saona, Saona being the earlier form of Savona, the birthplace of Columbus, a city not far from Genoa. At the end of the book, which contains neither name of printer nor place, is a notice that the work was completed at Cambridge on the 6th of July, 1478.

In an old catalogue of books bequeathed by Archbishop Parker to the library of Corpus Christi College, Cambridge, the entry occurs, "Rethorica nova impressa Canteb. fo. 1478." Strype, in writing his life of the Archbishop, came across this notice and communicated it to Bagford, who reported it in his turn to Tanner, the antiquary. Ames, from their information, placed it at the head of Cambridge books in his Typographical Antiquities, and Herbert, in his reprint, merely reproduced the account. Dibdin does not mention it, and it was not until 1861 that Henry Bradshaw, coming across it by accident, discovered that it was a genuine production of Caxton's press.

The book is a folio of 124 leaves, and besides the copy at Cambridge, one other is known, now in the University Library at Upsala.

On the 24th of March, 1479, was issued the Cordyale, a translation from the French Quatre derrenieres choses, by Earl Rivers. The translation, as the colophon tells us, was handed to Caxton on the day of the Purification (February 2d), and the printing was begun "the morn after the saide Purification of our blissid Lady, which was the daye of Seint Blase, Bisshop and Martir: And finisshed on the even of the annunciacion of our said bilissid Lady fallyng on the wednesday the 24 daye of Marche."

The Cordyale contains 78 leaves, with a blank at each end, and is not very uncommon. The second edition of the Dictes or Sayengis was issued this year, and is considerably rarer than the first, only four copies being known. Its collation is exactly the same as the first, and Blades has fallen into the same mistake, and gives it two leaves too few.