then, printers could afford to be careless, and speedily began to avail themselves of their new license. In the early days of the art no such license was possible, and the striking similarity in the appearance of the printed books and manuscripts produced contemporaneously in any given city or district, is the best possible proof of the success with which the early printers competed with the most expert of the professional scribes.

All this is trite enough, but we are somewhat less

frequently reminded that, after some magnificent experiments by Fust and Schoeffer at Mainz, the earliest printers deliberately elected to do battle at first with the scribes alone, and that in the fifteenth century the scribes were very far, indeed, from being the only persons engaged in the production of books. The subdivision of labour is not by any means a modern invention; on the contrary, it is impossible to read a list of the medieval guilds in any important town without being struck with the minuteness of the sections into which some apparently quite simple callings were split up. Of this subdivision of labour, the complex art of book-production was naturally an instance. For a proof of this, we need go no further than the records of the Guild of St. John the Evangelist at Bruges, in which, according to Mr. Blades's quotation of the extracts made by Van Praet, members of at least fourteen branches of industry connected with the manufacture of books joined together for common objects. In the fifteenth century a book of devotions, commissioned by some wealthy book-lover, such as the Duke of Bedford, might be written by one man, have its rubrics supplied by another, its small initial letters and borders by a third, and then be sent to some famous miniaturist in France or Flanders for final completion. The scribe only supplied the groundwork, all the rest was added by other hands, and it was only with the scribe that the early printers competed.

The restriction of their efforts to competition with the scribe alone, was not accepted by the first little group of printers until after some fairly exhaustive experiments. The interesting trial leaves, preserved in some copies of the 42-line Bible, differ from the rest not only in having their text compressed into two lines less, but also in having the rubrics printed instead of filled in by hand. Printing in two colours still involves much extra labour, and it was easier to supply the rubric by hand than to be at the pains of a second impression, even if this could be effected by the comparatively simple process of stamping. Except, therefore, in the trial leaves, the rubrics of the first Bible are all in manuscript. Peter Schoeffer, however, when he joined with the goldsmith Fust in the production of the magnificent Mainz Psalter of 1457, was not content to rely on the help of illuminators for his rubrics and capitals, or, as the disuse of the word majuscules makes it convenient to call them, initial letters. Accordingly, the Psalter appeared not only with printed rubrics, but with the magnificent B at the head of the first psalm, which has so often been copied, and some two hundred and eighty smaller initials, printed in blue and red.

Schoeffer's initial letters appear again in two editions of the Canon of the Mass attributed to 1458, in the Psalter of 1459, in the Rationale of Durandus of the same year, and in a Donatus printed in the type of the 1462 Bible. As Mr.

Duff has pointed out, in some sheets of this Bible itself the red initial letters are printed and the outline of the blue ones impressed in blank for the guidance of the illuminator in filling them in. Thereafter Schoeffer seems to have kept his initials for special occasions, as in the 35-line Donatus issued c. 1468, perhaps when he was starting business for himself, and in the antiquarian reprints of the Psalter in and after 1490. Doubtless he was sorry when he could no longer print in the colophon of a book that it was 'venustate capitalium decoratus, rubricationibusque sufficienter distinctus,' but while illuminators were still plentiful, handwork was probably the least expensive process of decoration. It is noteworthy, also, that Mr. Duff's discovery as regards the 1462 Bible brings us down to the beginning of those troublous three years in the history of Mainz, during which Fust and Schoeffer only printed 'Bulls and other such ephemeral publications.' When they resumed the printing of important works in 1465 with the Decretals of Boniface VIII. and the De Officiis of Cicero, Schoeffer was content to leave decoration to the illuminator. The firm's expenses were thus diminished, and purchasers were able to economise in the amount of decoration bestowed upon the copy they were buying. It is noteworthy, indeed, that even in 1459, when he was habitually using his printed initial letters, Schoeffer did not refuse customers this liberty, for while one of the copies of the Rationale Durandi at

the Bibliothèque Nationale has the initials printed, in the others they are illuminated by hand.

Very little attention has as yet been devoted to the study of the illumination and rubrication of printed books, and much patient investigation will be needed before we can attain any real knowledge of the relation of the illuminators to the early printers. Professor Middleton, in his work on Illuminated Manuscripts, had something to say on the subject, but the pretty little picture he drew of a scene in Gutenberg's (?) shop seems to have been rather hastily arrived at. 'The workshop,' he wrote, 'of an early printer included not only compositors and printers, but also cutters and founders of type, illuminators of borders and initials, and skilful binders, who could cover books with various qualities and kinds of binding. A purchaser in Gutenberg's shop, for example, of his magnificent Bible in loose sheets, would then have been asked what style of illumination he was prepared to pay for, and then what kind of binding, and how many brass bosses and clasps he wished to have.' What evidence there is on the subject hardly favours the theory which Professor Middleton thus boldly stated as a fact. The names we know in connection with the decoration of the 42-line Bible are those of Heinrich Cremer, vicar of the Church of St. Stephen at Mainz, who rubricated, illuminated, and bound the paper copy now in the Bibliothèque Nationale, and Johann Fogel, a well-known binder of the time,

whose stamps are found on no fewer than three of the extant copies of this Bible. We have no reason to believe that either Cremer or Fogel was employed in the printer's shop, so that as regards the particular book which he instances, it is hard to see on what ground Professor Middleton built his assertion.

As regards Schoeffer's practice after 1462, the evidence certainly points to the majority of his books having been rubricated before they left his hands, but the variety of the styles in the copies I have seen, especially in those on vellum, forbids my believing that they were all illuminated in a single workshop. A copy in the British Museum of his 1471 edition of the Constitutions of Pope Clement V. presents us with an instance, rather uncommon in a printed book, though not infrequently found in manuscripts, of an elaborate border and miniatures, sketched out in pencil and prepared for gilding, but never completed. The book could hardly have been sold in this condition, and would not have been returned so from any illuminator's workshop. We must conjecture that it was sold unilluminated to some monastery, where its decoration was begun by one of the monks, but put aside for