New Testament, connected with the name of the Abbé de Barne- | foreigners resident in or near London, while of the remaining ville, a priest of the Oratory at Paris. Impressed by the popular thirty, half are members of the Church of England, and half ignorance of the Scriptures, he himself translated, or caused others to translate, the New Testament into French from the Vulgate, and are members of other Christian denominations. formed an association to distribute copies systematically at low prices. The prefaces to his various editions contain details as to the methods of this association, and repeatedly insist on the importance of reading the Scriptures. (On this Société biblique catholique française see O. Douen, Histoire de la société biblique protestante de Paris, Paris, 1868, pp. 46-51.)

it has its own agencies or secretaries in twenty-seven of the chief | Schleswig-Holstein Bible Society, the Hamburg-Altona Bible Society cities of the world, and maintains depots in 200 other centres. It and others, together 56,000. employs 930 Christian colporteurs abroad, who sold in 1905-1906 During 1905, nine cantonal Bible societies in Switzerland circuover 2,250,000 volumes. It supports 670 native Christian Bible-lated altogether 71,000 copies; the Netherlands Bible Society women in the East, in connexion with forty different missionary reported a circulation of 54,544 volumes, 48,137 of which were in organizations. The centenary festival in 1904 was celebrated with Dutch; the Danish Bible Society circulated 45,289 copies; the enthusiasm by the Reformed Churches and their foreign missions Norwegian Bible Society circulated 67,058 copies; and in Sweden throughout the world. Messages of congratulation came from the the Evangelical National Society distributed about 110,000 copies. rulers of every Protestant nation in Christendom, and a centenary In Italy, by a departure from the traditional policy of the Roman thanksgiving fund of 250,000 guineas was raised for extending the Church, the newly formed "Pious Society of St Jerome for the society's work. During the year 1905-1906 the society expended Dissemination of the Holy Gospels " issued in 1901 from the Vatican £238,632, while its income was £231,964 (of which £98,204 represented press a new Italian version of the Four Gospels and Acts. By the receipts from sales). Up to the 31st of March 1906 the society had end of 1905 the society announced that over 400,000 copies of this expended altogether £14,686,072, and had issued 198,515,199 copies volume had been sold at 2d. a copy. of the Scriptures-of which more than 78,000,000 were in English.

The Bible in Spain (London, 1849); W. Canton, The History of the
British and Foreign Bible Society (London, 1904 foll.); J. Ballinger,
The Bible in Wales (London, 1906); T. H. Darlow and H. F. Moule,
Historical Catalogue of the Printed Editions of Holy Scripture (London,
vol. i. 1903, vol. ii. 1908).
(T. H. D.)
BIBLIOGRAPHY AND BIBLIOLOGY. The word Bißoypapia
was used in post-classical Greek for the writing of books, and as
late as 1761, in Fenning's English Dictionary, a bibliographer
is defined as one who writes or copies books." The transition
from the meaning a writing of books" to that of "a writing
about books," was accomplished in France in the 18th century
-witness the publication in 1763 of the Bibliographie instructive
of de Bure. In England the new meaning seems to have been
popularized by the Rev. Thomas Frognall Dibdin early in the
19th century, while Southey preferred the rival form bibliology,
which is now hardly used. Present custom inclines to restrict
the province of bibliography to printed books as opposed to
manuscripts, and on the other hand recognizes as coming within
its scope almost everything in which a book-loving antiquary
can be interested, including the history of printing (see
TYPOGRAPHY), book-binding (q.v.), book-illustration (see ILLUS-
TRATION) and book-collecting (q.v.). The present article is only
concerned with bibliography as the art of the examination,
collation and description of books, their enumeration and
arrangement in lists for purposes of information, and further
with the literature of this subject, i.e. with the bibliography
of bibliography.

of the New World. Bad forgeries of this kind can be detected by the tendency of all photographic processes of reproduction to thicken letters, and exaggerate every kind of defect, but the best of these imitations when printed on old paper require a specific knowledge of the originals and often cause great trouble. The type-facsimile forgeries are mostly of short pieces by Tennyson, George Eliot and A. C. Swinburne, printed (or supposed to have been printed-for it is doubtful if some of these "forgeries" ever had any originals) for circulation among friends. These trifles should never be purchased without a written guarantee.

When the edition to which a book belongs is known, further examination is needed to ascertain if it is perfect and in its original state. Where no standard collation is available, this can only be ascertained by a detailed examination of the quires or gatherings of which it is made up (see below). In the earliest books these are often very irregular. A large book was usually printed simultaneously in four or six sections on as many different presses, and the several compositors, if unable to end their sections at the end of a complete quire, would insert a single leaf to give more space, or sometimes leave a blank page, or half page, for lack of matter, occasionally adding the note " Hic nullus est defectus." A careful examination of the text, a task from which bibliographers often shrink, and a comparison with other editions, are the only remedies in these cases.

If a copy contains the right number of leaves, the further question arises as to whether any of these have been supplied from other copies, or are in facsimile. Few collectors even now are educated enough to prefer copies in the condition in which the ravages of time have left them to those which have beca "completed" by dealers; hence many old. books have been

Examination and Collation.-Books are submitted to examination in order to discover their origin, or to test statements concerning it which there is reason to doubt, or to ascertain if they are perfect, and if perfect whether they are in their original condition or have been "made up" from other copies. The discovery of where, when and by whom a book, or fragment" made up" with leaves from other copies, or not infrequently of a book, was printed, is the most difficult of these tasks, though as regards books printed in the 15th century it has been much facilitated by the numerous facsimiles enumerated under INCUNABULA (9.v.). In the article Book (q.v.) a sketch is given of the chief external characteristics of books in each century since the invention of printing. Familiarity with books of different ages and countries soon creates a series of general ideas as to the dates and places with which any combination of these characteristics may be connected, and an experienced bibliographer, more especially if he knows something of the history of paper, will quickly narrow down the field of inquiry sufficiently to make special search possible.

As regards the correction of mis-statements in early books as to their place and origin, glaring piracies such as the Lyonnese counterfeits of the octavo editions of the classics printed by Aldus at Venice, and the numerous unauthorized editions of works by Luther, professing to be printed at Wittenberg, have long ago been exposed. A different variety of the same kind of puzzle arises from the existence of numerous original editions with fictitious imprints. As early as 1499 a Brescia printer, in order to evade the privilege granted to Aldus, gave to an edition of Politian the spurious imprint " Florentiae," and in the 16th century many controversial books printed in England purported to have been issued in German towns, or with pleasant humour, " at Rome before the castle of S. Angel at the sign of S. Peter." Only a knowledge of the general characteristics which a book printed at such a place and such a time should possess will secure avoidance of these traps, but when suspicion has been aroused the whole story will often be found in such books as Weller's Die maskirte Literatur der älteren und neueren Sprachen (1856-1867), and Die falschen und fingirlen Druckörte (1864), Brunet's Imprimeurs imaginaires et libraires supposés (1866), de Brouillant's La Liberté de la Presse en France; Histoire de Pierre du Marteau, imprimeur à Cologne, &c. (1888); in the various bibliographies of Erotica and in Brunet's Manuel de l'Amateur and other handbooks for the use of collectors. A special case of this problem of piracies and spurious imprints is that of the modern photographic or type-facsimile forgery of small books possessing a high commercial value, such as the early editions of the letter of Columbus announcing his discovery

from other editions. These meddlings often defy detection,
but proof of them may be found in differences in the height and
colour of the paper, in the two corresponding leaves at either
end of a folio quire both possessing a watermark, or in their
wiremarks not corresponding, or (in very early books) by the
ornamentation added by hand being in a different style.
When it has been ascertained that a copy contains the right
number of leaves and that all these leaves are original, the last
point to be settled is as to whether it differs in any respect from
the standard collation. Owing to the extreme slowness of the
presswork for the first two centuries after the invention of
printing, there were more opportunities for making small correc-
tions while an old book was passing through the press than
there are in the case of modern ones, and on the other hand the
balls used for inking the type sometimes caught up words or
individual letters and these were replaced by the compositors
as best they could. The small variations in the text noticed
in different copies of the First Folio edition of Shakespeare, and
again of Milton's Paradise Lost, are probably to be explained
by a mixture of these two causes. Where a serious error was
discovered after a sheet had been printed off, the leaf on which
it occurred was sometimes cut out and a new leaf (called a
"cancel") printed to replace it and pasted on to the rest of
the sheet. Variations between different copies of the first edition
of Herrick's Hesperides which have puzzled all his editors are
due to the presence of several of such cancels. Lastly, a printer
when he had printed part of a book might wish to increase the
size of the edition, and the leaves already printed off would have
to be reprinted, thus causing a combination of identical and
different leaves in different copies. The famous 42-line Bible
of c. 1455, variously attributed to Gutenberg and to Fust and
Schoeffer, and the Valerius Maximus printed by Schoeffer in
1471, are instances of editions being thus enlarged while passing
through the press. As each book was set up simultaneously en
several different presses, the reprinted leaves occur at the
beginning of each of the sections.

It should be mentioned that there are books of which it is difficult to find two copies in exact agreement. Either to quicken presswork or to comply with trade-regulations made in the interest of compositors, in some books of which large

sired to retain the V forms the words should be printed in majus-
cules. If minuscules are used, the words should be transliterated
printers themselves.
as quinque, queen, Europe, according to the practice of the old

numbers were required, e.g. the Paraphrases of Erasmus, the First Prayer-book of Edward VI., and the "Songs and Sonnets known as Toltell's Miscellany, each forme was set up two or more different times. The formes were then used at haphazard A troublesome question often arises as to what notice should be for printing, and both at this stage and when the printed sheets taken in reproducing the misprints which frequently occur in the came to be stitched almost any number of different combinations original titles. Bibliographers who have satisfied themselves (and might be made. The books named were all printed in the though it will need constant watchfulness to prevent the printer their readers) of their own accuracy may reproduce them in silence, middle of the 16th century, but probably later instances could from" setting them right." Transcribers of only average accuracy be produced. will consult their happiness by indicating the misprint in some way, Description. The ideal towards which all bibliographical and the frequent use of (sic), more especially when printed in italics, or of the German (!), being ugly, probably the simplest plan is to add work should be directed is the provision in an accessible form of a note at the end stating that the misprints in question occur in the a standard description of a perfect copy of every book of literary, original. historical or typographical interest as it first issued from the (b) The" size" of a book is a technical expression for the relation press, and of all the variant issues and editions of it. When such of the individual leaves to the sheet of paper of which they form a part. A book in-folio means one in which the paper has been folded standard descriptions shall have been made, adequately checked once, so that cach sheet has made two leaves. In a book in-quarto, and printed, it will be possible to describe every individual copy each sheet has been folded twice so as to make four leaves. In an by a simple reference to them, with a statement of its differences, octavo another fold has produced eight leaves, and so on for books in if any, and an insistence on the points bearing on the special 16m0, 32m0 and 64mo. For books in twelves, twenty-fours, &c., the paper has at some stage to be folded in three instead of in two, and object with which it is being re-described. Only in a few cases there will be some difference in form according to the way in which has any approach been made to a collection of such standard this is done. The size of a book printed on handmade paper" is very descriptions. One instance which may be cited is that of the simply recognized by holding up a page to the light. Certain white entries of the 15th century books in the Repertorium Biblio-lines, called wire-lines, will be noticed, occurring as a rule about an graphicum of Ludwig Hain (1826-1838), which the addition of inch apart, and running at right angles to the fine lines. These wire-lines are perpendicular in a folio, octavo, 32 mo, and horizontal an asterisk marks as having been examined by Hain himself in in a quarto and 16mo. In a 12mo, as the name implies, the sheet is the copies in the Royal library at Munich. The high standard folded in twelve; and in the earlier part at least of the 16th century of accuracy of these asterisked entries (save for the omission to this was done in such a way that the wire-lines are perpendicular, the height of the sheet forming two pages, as is the case in an octavo, note blank leaves at the beginning or end) has been so well while the width is divided into six instead of into four as in an established, and the Repertorium is so widely known, that in octavo. The later habit has been to fold the sheet differently, the many catalogues of incunabula the short title of the book height of the sheet forming the width of four pages, and the width together with the number of Hain's entry has been usefully of the sheet the height of three pages, consequently the wire-lines are horizontal" (E. G. Duff, Early Printed Books, pp. 206-207). substituted for a long description. Books printed at Oxford up The recognition of what is meant by the size of a book has been to 1640 can be equally well described by their short titles and a obscured by the erroneous idea that the quires or gatherings of which reference to Mr Falconer Madan's Early Oxford Press published books are made up necessarily consist of single sheets. If this were in 1895. At present the number of works which can thus be so all folios would be in gatherings of two leaves each; all quartos taken as a standard is only small, owing partly to the greater In the case of books printed on handmade paper, this is generally in gatherings of four leaves; all octavos in gatherings of eights. and more accurate detail now demanded, partly to the absence true of octavos, but to reduce the amount of sewing the earliest of any system of co-operation among libraries, each of which is folios were usually arranged in tens, i.e. in gatherings of five sheets only willing to pay for catalogues relating exclusively to its own or ten leaves, while in Shakespeare's time English folios were mostly collections. It may be hoped that through the foundation of eights, and on the other hand the use of a half-sheet produces a in sixes. In the same way quartos are often found made up in bibliographical institutes more work of this kind may be done. gathering of only two leaves. A standard description of any book must, as a rule, consist of the following sections, though in the case of works which have no typographical interest, some of the details may be advantage ously omitted:-(a) A literal transcript of the title-page, also of the colophon, if any, and of any headings or other portions of the book serving to distinguish it from other issues; (b) Statements as to the size or form of the book, the gatherings or quires of which it is made up, with the total number of leaves, the measurement of an uncut copy or of the type-page, a note of the types in which different parts of the book are printed, and a reference to any trustworthy information already in print; (c) A statemert of the literary contents of the book and of the points at which they respectively begin; (d) A note giving any additional information which may be needed.

(a) In transcribing the title-page and other parts of the book it is desirable not to omit intermediate words; if an omission is made it should be indicated by three dots placed close together. The end of a line should be indicated by an upright stroke. It is a considerable gain to indicate to the eye in what types the words transcribed are printed, i.e. whether in roman, gothic letter, or italic, and in each case whether in majuscules or minuscules (" upper or lower case "). To do this, however, adds greatly not only to the cost of printing, but also to the liability of error. If roman minuscules are used throughout, or roman for the text and italic for the imprint of colophon, the method of transliteration which the printer himself would have used should be adopted. Many of the best modern catalogues and bibliographies are disfigured by the occurrence in them of such forms as qvinqve," "qveen,' Evrope," due to an unintelligent transliteration of the forms QVINQVE, OVEEN, EVROPE, as they occur on title-pages at a date when V" was the majuscule form of both V and "u." If it is de

"

1 Some bibliographers prefer to use double strokes to avoid confusion with the old-fashioned long commas. Others use a single stroke to indicate the space between two lines and increase the number of strokes where the space left is wider than this.

When a manuscript or carly printed book was being prepared for binding, it was usual for the order in which the quires or gatherings of the alphabet in their order, the alphabet generally used being were to be arranged to be indicated by signing them with the letters the Latin, in which I stands for both I and J; V for both U and V, and there is no W. If more than twenty-three letters were needed the contractions for et, con, rum and (less often) that for us, were used as additional signs, and for large books minuscules were used as well as majuscules, and the letters were doubled. In 1472 printed signatures came into use. If the quires or gatherings in the book to be described are signed in print, the signatures used should be quoted without brackets. If they are not signed, the order of the gatherings should be noted by the letters of the alphabet in square brackets. In each case the number of leaves in each gathering should be shown by index-figures. Thus, six gatherings of eight leaves followed by one of four should be represented by the symbols A-F G. The "make-up "of an old book in original binding is usually sufficiently shown by the strings in the middle of each quire. In books which have been rebound help may sometimes be obtained from the fact that between (roughly) 1750 and 1850, a period during which there was much rebinding of early books, the gatherings before being put into their new quires were mostly separately pressed, with the result that the outer pages of each gathering are much smoother than the rest. But the only safe guide to the make-up of an old book without printed signatures is a collation by means of the watermarks, i.e. the devices with which the papermaker as a rule marked each sheet (see PAPER). In a folio book one of every pair of leaves should have a watermark in the middle of the paper. in a quarto some pairs of leaves will have no watermark; in others it will be found divided by the fold of the paper. As the great majority of books without printed signatures are in folio or quarto,

It may be noted that some confusion is caused in descriptions of books by the word "sheet," which should be restricted to the original sheet of paper which by folding becomes folio, quarto, &c., being applied also to the double-leaf of four pages. A word specially appropriated to this is greatly needed, and as gatherings of two, three, four, &c., of such double-leaves are known technically as duernions, ternions, quaternions, &c., the double-leaf itself might well be called a " unit."

the sequence of watermarked and un-watermarked leaves, if care-
fully worked out, will mostly reveal the "make-up "of the successive
gatherings.
After the size and sequence of the gatherings has been stated, the
total number of leaves should be noted, with a mention of any
numeration of them given in the book. Any discrepancy between
the total of the leaves assigned to the successive gatherings and the
total as separately counted of course points to an error, and the
reckonings must be repeated till they tally. Errors in the printed
enumeration of the leaves of old books are common, and it is seldom
necessary to point them out in detail. When reference has to be
made to a particular page of an old book, the printed signatures
offer the readiest means, an index number placed below the letter
indicating the number of the leaf in the gathering and the addition
of recto or verso marking the upper or under page of the
leaf. Thus "X, recto (some bibliographers prefer the rather
clumsier form "X 4 recto") stands for the first page of the fourth
leaf of the gathering signed X. Where there are no printed signa-
tures the leaf-number may be given, the letters "
and "b"
above the numeral taking the place of recto and " verso
(leaf 99). Where some leaves of a book are numbered and others
not, if the reference is to the printed numeration this should be stated.
Printed leaf numeration is found as early as 1470, and became
common about ten years later. Printed pagination did not become
common till nearly the middle of the 16th century.

"

"

"

"

a

classification of human knowledge is a question for philosophers and men of science, that the knowledge of chemistry and of its history needed to make a good bibliography of chemistry is altogether extrinsic to bibliography itself; that all, in fact, to which bibliography can pretend is to suggest certain general principles of arrangement and to point out to some extent how they may be applied. The principles are neither numerous nor record.te. To illustrate the history of printing, books may be arranged according to the places and printing-houses where they were produced. For the glorification of a province or county, they are sometimes grouped under the places where their authors were born or resided. For special purposes, they may be arranged according to the language or dialect in which they are written. But, speaking generally, the choice for a basis of arrangement rests between the alphabetical order of authors and tities, a chronological order according to date of publication, a “logical " or alphabetical order according to subjects, and some combination of these methods. In exercising the choice the essential requisite is a really clear idea of the use to which the bibliography, when made, is to be put. If its chief object be to give detailed The foregoing details are all directed to showing which leaves of information about individual books, a strictly alphabetical a book would be printed by the same pull of the press, how it was made up for binding, and how imperfections in any copy may be arrangement "by authors and titles" (i.e. by the names of detected. They give little or no indication of the dimensions of the authors in their alphabetical order, and the titles of their books book. In the case of modern editions this may be done by adding in alphabetical sequence under the names) will be the most one of the trade epithets, pott, foolscap, crown, &c., to the name of useful, because it enables the student to obtain the information the size, which when thus qualified denotes paper of a particular he seeks with the greatest case. But while such an alphabetical measurement (see PAPER). As, however, these measurements are not easily remembered, it is better to give the actual measurements arrangement offers the speediest access to individual entries, it in inches or millimetres of a page of an uncut copy. In old books has no other merit, unless the main object of the bibliography uncut copies are not easily found, and it is useful instead of this to be to show what each author has written. If it is desired to give the measurement in millimetres of the printed portion of the illustrate the history and development of a subject, or the page (technically called the " type-page "), although this is subject to a variation of about 3% in different copies, according to the literary biography of an author, the books should be entered degree to which they were damped for printing. To this is added chronologically. If direction in reading is to be given, this can a statement of the number of lines in the page measured. The best be offered by a subject-index, in which the subjects are character of the type (roman, gothic or italic) is next mentioned, arranged alphabetically for speedy reference, and the books and in the case of 15th-century books, its number in the sequence chronologically under the subject, so that the newest are always of founts used by the printer (see INCUNABULA). Finally a reference to any authoritative description already printed completes this at the end. Lastly if the object is to show how far the whole portion of the entry. Thus the description of the collation of the field has been covered and what gaps remain to be filled, a class first-dated book printed at Augsburg, the Meditationes of S. Bona catalogue arranged according to what are considered the logical ventura, printed by Günther Zainer in 1468, should read: Folio subdivisions of the subject has its advantages. It is important, (alo, b-d3, e-go, h) 72 leaves. Type-page (1) 202 X 120 mm. ; 35 lines. however, to remember that, if the bulk of the bibliography is Type 1 (gothic letter). Hain 3557. (c) While many books, and this is especially true of early ones, very large, a principle of arrangement which would be clear and contain little or nothing beyond the bare text of a well-known work, useful on a small scale may be lost in the quantity of pages over others are well provided, not only with commentaries which are which it extends. An arrangement which cannot be quickly almost sure to be mentioned on the title-page, or in the colophon (which the editor himself often wrote), but also with dedicatory grasped, whatever satisfaction it may give its author, is useless letters, prefaces, complimentary verses, indexes and other accessories, to readers, the measure of its inutility being the worn condition the presence of which it is desirable to indicate. In these cases it is of the alphabetical index to which those who cannot carry a often convenient to show the entire contents of the book in the order complicated "logical" arrangement in their heads are obliged in which they occur, noting the leaves or pages on which each begins. It should Thus in the first edition (1590) of the first three books of Spenser's to turn, in the first instance, to find what they want. Faerie Queene, the literary contents, their order, and the space be obvious that any system which necessitates a preliminary they occupy can be concisely noted by taking the successive gather-reference to a key or index rests under grave suspicion, and needs ings according to their signatures and showing what comes on each page. Thus: A1, recto, title; verso, dedication, "To the Most Mightie and Magnificent Empresse Elizabeth": A-Oos, text of books i-iii.; Ppi, letter dated the 23rd of January 1589 [1590] to Sir Walter Raleigh expounding the intention of the work; Pp, verso, commendatory verses signed W. Raleigh], Hobynoll (Gabriel Harvey), R.S., H.B., W.L. and Ignoto; Pps, complimentary sonnets severally inscribed to Sir C. Hatton, the earls of Essex, Oxford, Northumberland and Ormond, Lord Ch. Howard, Lord Grey of Wilton and Sir W. Raleigh, and to Lady Carew and to the Ladies in the Court; and "Faults escaped in the print "; Qq-, Some bibliographers prefer to reverse the order of notation, (title, A, recto; dedication, A,, verso, &c.), and no principle is sacrificed in doing so, though the order suggested usually works out the more neatly.

fifteen other sonnets.

Enumeration and Arrangement.-In the 18th and early 19th centuries there was a tendency, especially among French writers, to exaggerate the scope of bibliography, on the ground that it was the duty of the bibliographer to appraise the value of all the books he recorded, and to indicate the exact place which each work should occupy in a logical classification of all literature based on a previous classification of all knowledge. Bibliographers are now more modest. They recognize that the 1 Here specify the page measured.

some clear counterbalancing gain to justify the loss of time which it entails. The main classification should always be that which will be most immediately useful to readers of the books. To throw light on the history of a subject and to indicate how far the field is covered are honourable objects for compilers, but should mostly be held subordinate to practical use. It is note worthy also that they may often be better forwarded by means of an index or table than by the main arrangement. The history of Hain's Repertorium Bibliographicum, which enumerates in an alphabetical arrangement of authors and titles some 16,000 books printed in the 15th century, is a good example of this. For sixty-five years it was of the utmost use for its accurate descriptions of individual books, but threw practically no light on the history of printing. In 1891 Dr Konrad Burger published an appendix to it containing an Index of Printers, since greatly enlarged in his index to Dr Copinger's Supplement to Hain (1902). The form of the index enables each printer's work to be seen at a glance, and the impetus given to the study of the history of printing was very great. But if the book had originally been arranged under Printers instead of Authors, it would have been far more difficult to use; its literary value would have been halved, and the record of the output of each press, now instantly