The watermark of this paper is an eight-pointed star in a circle. The supply of this paper does not appear to have been kept up for long, for I have only found it in two other English books. The Bartholomæus contains some very good woodcuts, finer than others of the period, and the press-work seems rather more regular than usual, so that perhaps we may accept the statement of Dibdin that “Of all the books printed in this country in the fifteenth century, the present one is the most curious and elaborate, and probably the most beautiful for its typographical execution.” It is only fair to say, however, that the copy described by Dibdin was a very exceptional one. In 1496 also came out a reprint of the well-known Book of St Alban’s, as it is generally called, a treatise on hunting, hawking, and heraldry, with the addition in this issue of the delightful chapter on fishing with an angle, our earliest printed treatise on the art. There is a woodcut of the angler at the beginning, and we see him busily at work with a large tub beside him, just like the German fisher of to-day, into which he may put his fish and keep them alive.

This book would naturally appeal especially to the richer class, and De Worde not only took especial pains with it, but struck off copies upon vellum, some of which have come down to our own day. From a typographical point of view the book is of great interest, for it is printed throughout in a foreign type which made its appearance in England on this occasion only. It was used at Gouda by Govaert van Os, but he seems to have discarded it about 1490 when he removed to Copenhagen. Besides acquiring this fount De Worde also obtained a number of woodcut capital letters, which are used in all his earliest books, and one or two woodcuts, which he used frequently until they were broken and worn out. It has always been a puzzle to me why, if De Worde had had this fount of type beside him for several years, he never used it before, and why, having used it this once, he never used it again. Not a single letter ever appears in another book, and yet the type is a handsome one.

1498 saw the issue of three fine folios: the Morte d’Arthur, of which the only known copy is in the John Rylands Library at Manchester, the Golden Legend, of which the only known perfect copy is in the same library, and lastly the Canterbury Tales of Chaucer. The only perfect copy of this book was sold lately in the Ashburnham sale for £1000, and is now in a private library in America. The first of these three books, the Morte d’Arthur, is a reprint of Caxton’s edition, but it differs from it in having illustrations. These are no doubt of native workmanship, and might be justly described as the worst ever put into an English book, being coarsely drawn, badly designed, and incompetently engraved. The Golden Legend is a mere reprint of the earlier editions, but is interesting for two points in the colophon. The first is an example of the carelessness of the printers. The words in the earlier editions run, “Thus endeth the legend named in latin legenda aurea, that is to say in Englysshe the golden legende, for lyke as golde passeth all other metals, so this legende exceedeth all other books, wherein be contained all the high and great feasts of our lord” and so on. In this edition a line has been omitted, and the words run, “For like as golde passeth all other metalles, wherein ben contained all the highe and grete festes of our lord.” Now although the omission makes nonsense of the whole sentence, it is reprinted exactly the same in the later editions issued by De Worde and Julian Notary.

The other point is the date in the colophon, which runs, “Fynysshed at Westmynster, the viii day of Janeuer, the yere of oure lorde Thousande . cccc. lxxxxviii. And in the xiii year of the reygne of kynge Henry the VII.” Now as the 13th year of Henry VII ran from August 22, 1497, to August 21, 1498, it is clear that De Worde in speaking of January 8, 1498, meant 1498 as we would calculate, and not 1499, and therefore that he began his year on the first of January and not on the 25th of March, a most important point to be settled in arranging dated books. Another later proof as to De Worde’s dating may be mentioned. In the tracts which he printed between January 1 and March 25, 1509, he speaks of himself as printer to the king’s mother, but after Henry VIII succeeded in 1509 he styles himself printer to the king’s grandmother, so that he clearly used our method of dating.

About the year 1498, De Worde introduced his second device, the largest of the three used in the fifteenth century. It is almost square, with a broad border, and having Caxton’s mark and initials above a flowering plant. Between July and December, 1499, a series of small nicks was cut all round the outside edge, and this gives us a useful clue to checking the dates of several books.

In 1499, De Worde brought out an edition of Mandeville’s Travels. It was not the first edition published in England by a year or two, but it was the first with illustrations, and most realistic illustrations they are. No doubt it was a very popular book, and the two copies known, one in the Cambridge University Library, the other at Stonyhurst, are both imperfect. Fortunately by means of the two we can obtain an exact collation. This year seems to have been a very busy one. While the dated books in the other years of the fifteenth century never rise above four, in this year there are ten, and a considerable number of undated books can be assigned to this year as well. Among them are a number of small poetical pieces by Lidgate, reprints of Caxton’s editions. One of these reprints shows how careless a printer W. de Worde was. He reprints the Horse, the Shepe, and the Ghoos, from a copy of Caxton’s wanting a leaf, but never noticing anything wrong prints straight ahead, making of course nonsense of the whole.

All De Worde’s quarto tracts were got up in the same style, the title at the top of the first leaf printed in one of Caxton’s types, below this a woodcut not always very apposite to the subject of the work. There were two stock cuts of masters with large birches and their pupils seated before them, one of these being among the material obtained from Govaert van Os. These of course were suitable for grammars and school-books. Caxton’s cuts for the Sarum Horae, the Crucifixion, The tree of Jesse, the three rioters and three skeletons, the rich man and Lazarus, and David and Bathsheba, came in very useful for theological books. The only special cut, that is, one specially cut for the particular book and not belonging to a series, that I have found, is that on the title of the Rote or mirror of consolation, which depicts seven persons kneeling before an altar, above which two angels hold a monstrance.

At the end of the year 1500, De Worde moved from Westminster into Fleet Street at the sign of the Sun, the earliest book from the new address being dated May, 1501. This from the point of view of the bibliographer was an extremely well-timed move, for we can at once put all books with the Westminster imprint as before 1501, and all with the London one after 1500, thus dividing clearly the fifteenth and sixteenth century books. At the time of his moving he seems to have got rid of a considerable portion of his stock; some seems to have been destroyed and some sold, for many cuts which had belonged to De Worde or to Caxton are found afterwards in books printed by Julian Notary. De Worde seems to have been a successful business man, for when he moved into Fleet Street he occupied two houses close to St Bride’s Church, one his dwelling-house and the other a printing-office, for which he paid the very high tithe rent of sixty-six shillings and eightpence.

The number of books printed by Wynkyn de Worde in the fifteenth century, counting in different editions of the same book, is 110, and of a considerable number of these only a single copy is known. It would seem probable that the printer, when issuing a small book, printed only a small number of copies, preferring to set up the type for a new edition rather than burden himself with much unsaleable stock. And it is curious how these various editions have been accidentally preserved. Only two copies are known of a book called the Rote or mirror of consolation, printed by De Worde in the fifteenth century, one of them is in the Pepysian library, the other in Durham Cathedral. Yet these two are of quite different editions, the one at Durham being certainly about 1496, the other certainly after the middle of 1499. Of the Three Kings of Cologne we have two editions, though only three copies are known. Indeed, for some time it was thought that each copy represented a different edition, as the copy in the British Museum, evidently bound up separately out of a volume of tracts, had had the last page of the tract preceding it bound in in place of the correct title-page.

Looking at the very large number of small books which De Worde printed between the end of 1496 and 1500, it is surprising how many are known from single copies. I have kept for many years a register of all the copies of early English books which are to be found anywhere, and taking the quartos printed by W. de Worde, which number altogether 70, I find that out of that number 47, that is more than two-thirds, are known to us now from single copies or fragments. And I feel certain that we owe the preservation of the majority of these to a cause we are now doing our best to destroy. A few worthy people centuries ago made collections of these tracts and bound them up in immensely stout volumes, which gave them an air of importance in themselves, and tended to preserve the tracts inside in a much better manner than if bound separately. I do not think I am exaggerating when I say that a hundred and fifty of the rarest that De Worde printed during his whole life would have been found a hundred years or so ago bound up in about twelve volumes. Some twenty-two of the rarest of W. de Worde’s in the Heber Library came to him in one volume. Thirteen unique tracts which sold at the Roxburghe sale for £538, were in a single volume when the Duke purchased them fourteen years before for £26. I need only refer you to the University Library, a large number of whose unique Caxton and De Worde tracts came in three or four volumes. Then again, when so many are known only from fragments or single copies we may imagine what a large number have absolutely disappeared.

Some have been lost of late years or have disappeared since they were described. Three unique W. de Worde books of the fifteenth century were supposed to have perished in a fire in Wales in 1807 but fortunately they had been sold by the owner of the library a short time before the fire. Others seem to have drifted into libraries whose owners know nothing about them. There is a unique De Worde printed before 1501, entitled the “Contemplacyon or meditacyon of the shedynge of the blood of our lorde Jhesu Cryste at seven tymes.” This was seen and described by Herbert, who very likely saw it when it was sold at the Fletewode sale in 1774. Since then we have no record of the book, and though every year more information about private collections is published I can come upon no trace of it.

Beside the genuine books which have disappeared, by this I mean books which have been described by a trustworthy bibliographer, there are others which may reasonably be supposed to have existed, and one clue to these is afforded by the woodcuts. W. de Worde for example had certain series of cuts, specially made for certain books; but when he wished to decorate the title-page of a small tract, which was not itself to be otherwise illustrated, he used an odd cut out of his sets. Now when we can trace in different tracts odd cuts, manifestly belonging to a series, we may reasonably suppose that the book for which the series was engraved must have been printed.

To give a couple of instances. In the unique copy of Legrand’s Book of good manners in the University Library without date, but printed about the middle of 1498, are two cuts, which really belong to a series made to illustrate the Seven wise masters of Rome. These cuts are fairly accurate copies of those used by Gerard Leeu in his edition of 1490. At a considerably later date De Worde did issue an edition of the Seven wise masters, illustrated with the series of which the two mentioned above formed part, and showing at that time marks of wear. Now as De Worde had the series cut by the beginning of 1498, I think it most probable that an edition of the book was then issued, for it is unlikely that he would go to the trouble of cutting the set unless he was preparing to print the book.

Again, before the end of the fifteenth century De Worde had a series to illustrate Reynard the Fox. One cut is found on the first leaf of an edition of Lidgate’s The Horse, the sheep, and the goose, in the University Library, another on the title-page of Skelton’s Bowge of Court in the Advocates’ Library at Edinburgh. In the collection of the University Librarian is a fragment of an edition of Reynard, evidently printed by W. de Worde about 1515, and this contains a third cut agreeing absolutely in size, in workmanship, and in style with the other two.

In this case again it seems probable that an edition illustrated with these cuts appeared before 1500.

The last press at Westminster during the fifteenth century is that of Julian Notary, which while it started in London about 1496 and only moved to Westminster in 1498, is more suitably taken in this place on account of its connexion with Wynkyn de Worde.

The first book issued was an edition of Albertus de modis significandi, printed in a neat Gothic type, but containing no information in its colophon beyond that it was printed in London at St Thomas the Apostle’s, probably close to the church of that name, and not at a house with that sign. There is also a printer’s mark containing three sets of initials, I. N. for Julian Notary, I. B. for Jean Barbier, and I. H. for someone unidentified, but who there are some reasons for supposing to have been Jean Huvin, a printer at Rouen, who was associated in the production of books for the English market.

In 1497 the same printers issued an edition of the Horae ad usum Sarum, very neatly printed, and with delicate borders round the pages. All that remains of the book is a fragment of four leaves, rescued from a book-binding, but this luckily contains the colophon, telling us that it was printed at St Thomas the Apostle’s, for W. de Worde. This book also contains the device with the three sets of initials.

In 1498 appeared a Sarum Missal, the first edition printed in England, and though otherwise well got up, the musical parts have the drawback of being without notes, only the staves having been printed, though whether this was done by design or merely because the printers had no musical type remains unknown. From the colophon of the Missal we learn that the printers, Julian Notary and Jean Barbier, had settled at Westminster, and had printed the book at the command and expense of W. de Worde. On the last leaf is Caxton’s device, and on the title-page that of the printers. Of this book five copies are known, and of the four I have examined, the copy in the University Library is the only perfect one. About the fifth, belonging to the Duke of Sutherland, I have no information.

I. H. it is clear had left the firm, and though the printers use the same device as before, the initials I. H. have been cut out of it.

In 1499 Jean Barbier also disappeared, for in the edition of the Liber Festivalis and Quattuor Sermones which appeared in that year the printer’s mark has again been altered. All initials have been cut out and the name Julianus Notarii inserted in type. This form of the name suggests that he was not a notary as is generally stated, but the son of one. I have never been able to see a perfect copy of this book though Herbert describes one which he said was in the Inner Temple Library, but my inquiries there met with no success. Hain in his Repertorium Bibliographicum mentions a copy which seems not to be the one noticed by Herbert.

In April, 1500, Notary printed a most minute edition of the Horae ad usum Sarum, it is in 64s as regards folding, and a printed page measures an inch and a quarter by an inch. Only a fragment of it is known, a quarter sheet containing sixteen leaves, but that luckily contains the colophon. It was very likely copied from another edition of the same size, which was printed at Paris the year before, but this point cannot be determined, as the only copy of the latter which existed was burnt with the greater part of the Offor collection. All we know now of it is the meagre note in the auctioneer’s catalogue, “imperfect, but has end with imprint”—and he has not given the imprint!

The colophon of Notary’s Horae tells us that it was printed in King Street, Westminster. King Street is the short street at the bottom of Whitehall in a straight line between Westminster Abbey and the Foreign Office, though in Notary’s time it appears to have extended from Westminster to Charing Cross. Lewis, in his life of Caxton, says that Caxton’s printing-office was in King Street, but I do not know of any reason for his assertion.

The last of Notary’s books printed at Westminster is an edition of Chaucer’s Love and complaintes between Mars and Venus, with some other pieces. This rare little book, having passed through the collections of Farmer, the Duke of Roxburghe, Sir Mark Masterman Sykes and Heber, is now at Britwell. The colophon runs: “Thys inpryntyde in westmoster in kyng strete. For me Julianus Notarii.” In spite of the word For, I think the book was printed by Julian Notary himself. It contains two cuts, reversed copies of two of Caxton’s.

At what time Notary left Westminster cannot at present be settled, but probably almost immediately after W. de Worde. When his next dated book was issued in 1503 he had moved to London, and with his departure from King Street to Pynson’s old house near Temple Bar printing ceased altogether in Westminster.