Bibliography, thus understood, may be defined as the science of books, having regard to their description and proper classification. Viewing books simply as vehicles of learning, it would undoubtedly be correct to extend our inquiry to the period when the only books, so called, were manuscripts. And such is, in fact, the view adopted by bibliographers like Peignot, Namur, and Hartwell Horne. But a survey so extensive is open to practical objections. In the first place, bibliography as a science was unknown until long after printing had laid its first foundations, and indeed made it a necessity, with requirements increasing with the multiplied productions of the press. The materials for comparative study were wanting in an age when books were regarded as isolated treasures, to be bought at prices corresponding with their scarcity. In the second place, the critical study and comparison of ancient manuscripts, their distribution into families deduced from one or more archetypes, and the investigation of ancient systems of writing, embrace a subject so wide in its scope and special in its character, that convenience of treatment, confirmed as it is by the facts of history, would alone suggest the propriety of distinguishing between manuscript and printed bibliography. This distinction it is here proposed to observe, the subject of MSS. being reserved for the article Palaeography, the name which in its maturity it received.

Amid much variety of treatment in detail, two main divisions underlie the general study of bibliography, viz., material and literary, according as books are regarded with reference to their form or their substance. The former belongs chiefly to the bookseller and book-collector; the latter to the literary man and the scholar. Material bibliography treats of what Savigny terms the “äussere Bücher wesen,” or the external characteristics of books, their forms, prices and rarity, the names of the printers, the date and place of publication, and the history of particular copies or editions. It involves a knowledge of typography, not, indeed, as a mechanical process, but in its results, and, in fact, of all the constituent part of books, as a means of identifying particular productions. Its full development is due to the gradual formation of a technical science of books. Considerations of buying and selling, which were first reduced to a system in Holland, and afterwards advanced to their present complete form in France and England, gave an impetus to this branch of bibliography. The growth of private libraries, especially during the last century in France, promoted a passion among rich amateurs for rare and curious books; and literary antiquarians began to study those extrinsic circumstances, apart from the merit of their contents, which went to determine their marketable value, and to reveal the elements of rarity.

Literary, or, as it is sometimes called, intellectual bibliography treats of books by their contents, and of their connection in a literary point of view. It has been subdivided into pure and applied, according as its functions became more complex with the spread of printed books and the increasing requirements of learning. Catalogues expanded into dictionaries, whose object was to acquaint literary men with the most important works in every branch of learning. Books were accordingly classified by their contents, and the compiler had to distinguish between degrees of relative utility, so that students might know what books to select. This duty, which devolved in mos* cases on men of learning, has led French writers in particular to exaggerate the province of bibliography. “La bibliographie,” says Achard, “étant la plus étendue de toutes les sciences, semble devoir les renfermer toutes; ’* and Peignot describes it under his proposed title of Bibliologie, as “la plus vaste et la plus universelle de toutes les connaissances humaines.” We know of no excuse for such pretensions beyond this, that books represent, in its transmissible form, the sum total of all kinds of knowledgo. The bibliographer has to determine the genuineness, not the authenticity of a book; its identity of authorship or publication, not the correctness of its contents. When he pronounces judgment on its intrinsic merits he usurps the office of the critic. Some works, indeed,—like Bailiet’s Jugemens des Savans, sur les Principaux Ouvrages des Auteurs, augmentés par Μ. de la Monnoye, 8 vols., Amst., 1724; Blount’s Censura Celebriorum Auctorum, London, 1690; Morhof’s Polyhistor Literarius, Philosophicus, et Practicus, the best edition of which is that of Fabricius in 1747; the 0nomasticon Literarιum of Saxius, Utrecht, 7 vols., 1759-90; and the Censura Literaria of Sir Egerton Brydges, 10 vols., 1805-9,—are collections of critical bibliography of extreme value to the literary historian; but there is a wide difference in design between compilations even of this kind and works devoted to original criticism. In like manner the proper objects of classification have been neglected by many bibliographers, who have indulged in refinements of method, not as a means of facilitating reference, but for the purpose of illustrating a philosophical system of learning. Pretensions such as these, have, unfortunately, done much to discredit bibliography as a science of practical application, by investing it with a false air of mystery, and exposing it to the charge of empiricism. Its real value, in a literary aspect, depends on the recognition of its purpose as ancillary to the study of literature; not, in short, as an end, but as a means to the attainment of knowledge, by the investigation of its sources.

France must be regarded as the real mother of bibliography. Italy was the field in which book-collections first began on a large scale, and that country can boast of names like Magliabecchi, Apostolo Zeno, Bandini, Audiffredi, Mazzuchelli, and Morelli, besides provincial works like Moreni’s Bibliografia della Toscana, and Gamba’s Serie di Testi. But the labours of French bibliographers, especially after Naudé, converted a study, more or less desultory, into a science and a systematic pursuit. In Germany, poor in public and almost destitute of private libraries, bibliography has been studied almost exclusively in its literary aspect. Belgium has shown much recent activity; but neither Holland, Spain, nor Portugal can show any modern work of importance. In England the paucity of bibliographers is the more to be regretted from the wealth of her resources. Richard de Bury, in his Philobiblion, had descanted on the charms of book-collecting as early as the 14th century; but Blount’s Censura, published in 1690, was the only regular treatise on bibliography up to that date. Oldys, wh03e British Librarian first appeared in 1737 but was never completed, was among the first in this country to divert the public taste from an exclusive attention to new books, by making the merit of old ones the subject of critical discussion; and Maittaire, who was second master of Westminster School, and who died in 1747, first established the study of bibliography in England on a solid basis. The labours of Dibdin we shall have occasion frequently to refer to; they mark a new phase of bibliography in England which followed the opening up of the Continent after the great war with France. The science in America has been cultivated only recently; but the names of Cogswell, Ticknor, and Jewett are already well-known to bibliographers.

I The Constituent Parts of Books, and Differences of Editions.

The history of the materials used for early manuscripts— a subject fruitful in research—lies outside the limits we have proposed for bibliography as the study of printed literature. Fortunately for the spread of books, in the modern sense of the term, the invention of printing was preceded by the important discovery of the art of making paper from linen rags. The precise date of this discovery is not known, nor are writers agreed as to the country in which it was made; but it seems to be ascertained that this kind of paper was in general use in Europe before the end of the 14th century. Caxton and the other early English printers appear to have used paper of foreign manufacture. Such questions, among others, as the relative priority of different editions, or the productions of different presses, are frequently to be determined by a comparison of the constituent elements of the books themselves; but the subject is too technical to be noticed in detail. The question as to the origin of printing belongs strictly to a consideration of that art; but as its history and its progress are illustrated by the productions of different presses, the bibliographer will find much matter of interest in the principal works devoted to the subject. Prominent among these are the Monumenta Typographies of Wolfius, Hamburg, 1740; Meerman’s Origines Typographical; Prosper Marchand’s Histoire de l , origine et des premiers progrès de V Imprimerie, 1740,—a valuable supplement to which was published by Μ. Mercier, Abbé de Saint Leger, in 1773, and republished in 1775; and Lambinet’s Iiecherches historiques, littéraires, et critiques sur l'origine de Γ Imprimerie, first published at Brussels in 1799.

An accurate knowledge of the different forms of books is necessary to the bibliographer, as without it no book can be correctly described; and however easy such knowledge may appear, it is yet certain that errors in this respect have been committed even by experienced bibliographers, and that doubts have been entertained as to the existence of editions, owing to their forms having been inaccurately described. These mistakes generally proceed from this, that there are different sizes of paper comprehended under the same name. But the water-lines in the sheets afford a test, as they are uniformly perpendicular in the folio and octavo, and horizontal in the quarto and duodecimo sizes. In the infancy of printing the sizes were generally folio and quarto, and some have supposed that no books were printed in the smaller forms till after 1480; but Μ. Peignot instances many editions in the smallest forms of an earlier date; as may be seen in the article “Format” of the supplement to his Dictionnaire de Bibliologie. The subject of water-marks is treated at length in Sotheby’s Principia Typographies.

The respective merits of different editions can bo ascertained often only by minute inquiries. It is a principal object of the bibliographical dictionaries, to be afterwards mentioned, to point out those editions of important works which such inquiries have discovered to be the best. There are many particulars in which one edition may differ from or excel another. There may be differences or grounds of preference in size, in paper, and in printing. Later revision by the author may give his work, when it comes to be reprinted, a complexion differing largely from what it had at the first; while the first edition exhibits his orginal thoughts as they came fresh from his pen. One edition may derive its superiority from being furnished with notes, an index, or a table of contents. Plates make great differences in the value of editions, and even in the value of copies of the same edition. In the beautifully engraved edition of Horace by Pine, a small error in the first impressions serves as a test whether any copy contains the best engravings of those elegant vignettes which illustrate that edition. The medal of Augustus, on page 108 of the second volume, has in the first copies the incorrect reιding Post Est instead of Potest; this was rectified in the after impressions; but as the plates had meanwhile sustained some injury, the copies which show the incorrect reading are of course esteemed the best. Dibdin, in his Bibliomania, points out this as an instance of preference founded on a defect; but the real ground of preference is the superiority of the impressions, ascertained by the presence of this trifling defect. There are sometimes differences between copies of the same edition of a work. Walton’s Polyglot Bible is a celebrated instance. The printing of that great work, for which Cromwell liberally allowed paper to be imported free of duty, was begun in 1653 and completed in 1G57, and the preface to it in some copies contains a respectful acknowledgement of this piece of patronage on the part of the Protector; but in other copies the compliment is expunged, and replaced by some invectives against the republicans,— Walton having on the Restoration printed another preface to the copies which had not by that time been disposed of.

II. Early Printed Books.

The first productions to which the name of Books has been applied, were printed, not with movable types, but from solid wooden blocks. These consisted of a few leaves only, on which were impressed images of saints and other historical pictures, with a text or a few explanatory lines. The ink was of a brownish hue, and glutinous quality, to prevent it from spreading. These are known by the name of Image Books, or Block Books, and are generally supposed to have succeeded the earlier impressions for playing cards, which are dated back to the end of the 14th century. Strictly speaking, they were the immediate precursors, rather than the first specimens of typography; in fact, they mark the transition to that art from engraving. Peignot puts their number at seven or eight, but others have extended it to ten. They belong chiefly to the Low Countries, and were often reprinted, as is generally thought, during the first half of the 15th century, and, indeed, after the discovery of printing, properly so called. One of the most celebrated is the Biblia Pauperum, 1 consisting of forty leaves, printed on one side, so as to make twenty when pasted together, on which passages from Scripture are represented by means of figures, with inscriptions. It appears to have been originally intended for the use of those poor persons who could not afford to buy complete copies of the Bible. Some fugitive sheets still attest the primitive attempts at printing, in the modern sense of the word. The Letters of Indulgence of Pope Nicholas V., two editions of which, on a small sheet of parchment, were printed in 1454, fix the earliest period of the impression of metal types, with a date subjoined. 2 The earliest known book, however, of any magnitude, and probably the first thus printed, was the undated editio princeps of the Bible, commonly known as the Mazarin Bible, from a copy having been found by De Bure in the library of the Cardinal. It is undated, but authorities generally concur in ascribing it to a period between 1450 and 1455. The work is usually divided into two volumes, the first containing 324, and the second 317 pages, each page consisting of two columns. The characters, which are Gothic, are large and handsome, and resemble manuscript. No fewer than twenty copies are known to be extant. 3 The first printed book with a date is the Psalter of Fust and Schöffer, printed at Mentz in 1457, as a somewhat pompous colophon announces. It was found, in 1665, in the Castle of Ambras, near Innsbruck, where the Archduke Francis Sigismund had collected a quantity of MSS. and printed books, taken chiefly from the library of Corvinus. A few other copies are in existence, one of which was bought under Louis XVIII. for the Royal Library at Paris for the sum of 12,000 francs. Whether the types employed were wooden or metallic has been disputed between Van Praet and Didot. As a specimen of early printing the work is magnificent; it contains richly embellished capitals in blue, red, and purple.

Besides these monuments of infant typography, a special interest attaches to the productions of the 15 th century. They are usually known as Incunabula, a term applied to them by modern German writers. Brunet, following Santander, estimates their number at 18,000 or 20,000; but it is probable that many duplicates are included in this reckoning. They came into demand chiefly at the beginning of the last century, and especially about 1740, at the third centenary of printing. The passion for them at first was indiscriminate, but preference afterwards was given to the presses of Mayence, Bamberg, Cologne, Strasburg, Rome, and Venice.

As regards these early printed books, a knowledge of typography is necessary to the bibliographer, to enable him to verify their identity. A brief reference to some of their leading peculiarities must suffice here. The printer’s name, and the date 4 and place of printing were at first omitted, the printer imitating the reticence of the copyist, and the book being a mere fac-simile of the manuscript. In Germany and the Low Countries few dated books are found before 1476 or 1480. Title-pages appear to have come in a few years later; none of Caxton’s works, with one doubtful exception, have any. Titles to chapters were first used in the Epistles of Cicero, 1470. According to Palmer, the use of signatures, or letters at the bottom of the page to guide the bookbinder in the arrangement of the sheets, began with Zarot in a Terence printed by him at Milan in 1470. Marolles ascribes them to John of Cologne, who printed at Venice in 1474, and the Abbé Rive to Jo1id Koelhof, a printer of Cologne. They were in use in that city in 1475, and at Paris the next year, but were not employed by Caxton until 1480. Catch-words, which, like signatures, preceded the numbering of pages, are found in MSS. of the 11th century, and were first applied to printing by Vindelin de Spira at Venice. Their purpose, to direct the binder, had been previously supplied by Registers, or alphabetical tables of the first word of chapters, which were introduced about 1469. The earliest system of numbering was applied, not to pages, but to leaves, a large Roman figure being placed at the top of the recto in each leaf. The characters were uniformly Gothic—the foundation of our Black-letter—until 1467, when Gothic was supplanted by the Roman type, introduced in that year at Rome, and improved on by Jenson at Venice. It was first used in England by Pynson. Italics were first used by Aldus in his Virgil of 1501; they are said to have been suggested to him by Petrarch’s writing, and were employed to compress matter into his small octavos without the inconvenience of abbreviations. Hebrew characters began at Soncino, in the duchy of Milan, in 1482, and at Naples in 1487. The only points first used were the colon and full stop; but Alduβ improved punctuation by giving a better shape to the comma and adding the semicolon. With Caxton oblique strokes took the place of commas and periods. The form of the earliest books was chiefly folio and quarto. Almost every page abounded in abbreviations or contractions. Blank spaces were left for capitals and the first letters of periods, which were afterwards filled up by the illuminator. The Basel press was noted for its ornamental initials; and Calliergus at Rome and the Paris printers excelled in decorative printing of this kind. The taste for embellishment led to ornamental title-pages about 1490, the usual ornament at first being the “author at his desk.” The custom of coloured frontispieces appears to have prevailed until the end of last century. Decorated borders appear in the first page of some of Sweynheim and Pannartz’s productions; few ornaments, however, were introduced into the body of the text before the first Hebrew publications. 5 The Aulus Gellius of 1469 by the same printers is cited as the first book with a preface; and their Apuleius of the same year contains the earliest marginal notes. For further information on the characteristics of early printed books the reader will do well to consult Palmer’s General History of Printing (a work ascribed chiefly to George Psalmanazar); Jungendres, Be Notis Characteristicis Librorum a Typographioe Incunabulis ad annum 1500 impressorum ; and Maro lie’s Recherches sur l'Origine des Signatures et des Chiffres de Page. 1 So called first by Heinecken, Idée générale d'une collection complete d' Estampes, 8vo., 1771. Dibdin, inhis Bibl. Spenceriana, and Ottley, in his History of Engraving, have given fac-similes of the figures in several of the block-books. See also Falkenstein’s Geschichte der Buchdruckerkunst in ihrer Entstehung und Ausbildung, 4 to, 1840; Schelhorn’s Aman. Lit.; the works of Maittaire, D. Clement, Fournier, Meermann, Papillon, and De Bure; and J. P. Berjeau’s Catalogue illustré des livres xylographiques, 1865. Heinecken was the chief authority until recently, when his views, especially on the chronology of the block-books, have been much contested. Sotheby’s Principia Typo- graphica,...