Bible Encyclopedias
Bibliography and Bibliology

1911 Encyclopedia Britannica

Search for…

Enter query below:

or

A B C D E F G H I J K L M N O P Q R S T U V W Y Z

Prev Entry
Bible Societies

Next Entry
Bibliomancy

Resource Toolbox

Print Article

The word f fiXcoypacia 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, book-illustration (see Illustration) and book-collecting. 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.

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 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. In the article BooK 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 dlteren and neueren Sprachen (1856-1867), and Die falschen and fingirten Druckorte (1864), Brunet's Imprimeurs imaginaires et libraires supposes (1866), de Brouillant's La Liberte de la a 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 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 been " completed " by dealers; hence many old books have been " made up " with leaves from other copies, or not infrequently 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 corrections 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 on 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 numbers were required, e.g. the Paraphrases of Erasmus, the First Prayer-book of Edward VI., and the " Songs and Sonnets " known as Tottell's Miscellany, each forme was set up two or more different times. The formes were then used at haphazard for printing, and both at this stage and when the printed sheets came to be stitched almost any number of different combinations might be made. The books named were all printed in the middle of the 16th century, but probably later instances could be produced.

Description

The ideal towards which all bibliographical work should be directed is the provision in an accessible form of a standard description of a perfect copy of every book of literary, historical or typographical interest as it first issued from the press, and of all the variant issues and editions of it. When such standard descriptions shall have been made, adequately checked and printed, it will be possible to describe every individual copy by a simple reference to them, with a statement of its differences, if any, and an insistence on the points bearing on the special object with which it is being re-described. Only in a few cases has any approach been made to a collection of such standard descriptions. One instance which may be cited is that of the entries of the 15th century books in the Repertorium Bibliographicum of Ludwig Hain (1826-1838), which the addition of an asterisk marks as having been examined by Hain himself in the copies in the Royal library at Munich. The high standard of accuracy of these asterisked entries (save for the omission to note blank leaves at the beginning or end) has been so well established, and the Repertorium is so widely known, that in many catalogues of incunabula the short title of the book together with the number of Hain's entry has been usefully substituted for a long description. Books printed at Oxford up to 1640 can be equally well described by their short titles and a reference to Mr Falconer Madan's Early Oxford Press published in 1895. At present the number of works which can thus be taken as a standard is only small, owing partly to the greater and more accurate detail now demanded, partly to the absence of any system of co-operation among libraries, each of which is only willing to pay for catalogues relating exclusively to its own collections. It may be hoped that through the foundation of bibliographical institutes more work of this kind may be done.

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 advantageously 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 statement 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 " gvingve," " qveen," " Evrope," due to an unintelligent transliteration of the forms Qvinqve, Qveen, 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.

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

A troublesome question often arises as to what notice should be taken in reproducing the misprints which frequently occur in the original titles. Bibliographers who have satisfied themselves (and their readers) of their own accuracy may reproduce them in silence, though it will need constant watchfulness to prevent the printer from " setting them right." Transcribers of only average accuracy will consult their happiness by indicating the misprint in some way, 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 a note at the end stating that the misprints in question occur in the original.

(b ) The " size " of a book is a technical expression for the relation 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 once, so that each sheet has made two leaves. In a book in-quarto, each sheet has been folded twice so as to make four leaves. In an octavo another fold has produced eight leaves, and so on for books in 16mo, 32mo and 64mo. For books in twelves, twenty-fours, &c., the pap;:r has at some stage to be folded in three instead of in two, and there will be some difference in form according to the way in which this is done. The size of a book printed on handmade paper " is very simply recognized by holding up a page to the light. Certain white lines, called wire-lines, will be noticed, occurring as a rule about an inch apart, and running at right angles to the fine lines. These wire-lines are perpendicular in a folio, octavo, 32 mo, and horizontal in a quarto and 16mo. In a 12mo, as the name implies, the sheet is folded in twelve; and in the earlier part at least of the 16th century 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, while the width is divided into six instead of into four as in an octavo. The later habit has been to fold the sheet differently, the height of the sheet forming the width of four pages, and the width of the sheet the height of three pages, consequently the wire-lines are horizontal " (E. G. Duff, Early Printed Books, pp. 206-207).

The recognition of what is meant by the size of a book has been obscured by the erroneous idea that the quires or gatherings of which books are made up necessarily consist of single sheets.' If this were so all folios would be in gatherings of two leaves each; all quartos in gatherings of four leaves; all octavos in gatherings of eights. In the case of books printed on handmade paper, this is generally true of octavos, but to reduce the amount of sewing the earliest folios were usually arranged in tens, i.e. in gatherings of five sheets or ten leaves, while in Shakespeare's time English folios were mostly in sixes. In the same way quartos are often found made up in eights, and on the other hand the use of a half-sheet produces a gathering of only two leaves.

When a manuscript or early printed book was being prepared for binding, it was usual for the order in which the quires or gatherings were to be arranged to be indicated by signing them with the letters of the alphabet in their order, the alphabet generally used being 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 4. 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, 2 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 carefully 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 4 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 signatures the leaf-number may be given, the letters " a " 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.

The foregoing details are all directed to showing which leaves of 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 detected. They give little or no indication of the dimensions of the book. In the case of modern editions this may be done by adding one of the trade epithets, pott, foolscap, crown, &c., to the name of the size, which when thus qualified denotes paper of a particular measurement (see Paper). As, however, these measurements are not easily remembered, it is better to give the actual measurements in inches or millimetres of a page of an uncut copy. In old books uncut copies are not easily found, and it is useful instead of this to give the measurement in millimetres of the printed portion of the page (technically called the " type-page "), although this is subject to a variation of about 3% in different copies, according to the degree to which they were damped for printing. To this is added a statement of the number of lines in the page measured. The character of the type (roman, gothic or italic) is next mentioned, and in the case of 15th-century books, its number in the sequence of founts used by the printer (see Incunabula). Finally a reference to any authoritative description already printed completes this portion of the entry. Thus the description of the collation of the first-dated book printed at Augsburg, the Meditationes of S. Bonaventura, printed by Gunther Zainer in 1468, should read: Folio (a 10, b-d 8, ag 10, h 8) 72 leaves. Type-page (1) 202 X 120 mm.; 35 lines. Type 1 (gothic letter). Hain 3557.

(c ) While many books, and this is especially true of early ones, contain little or nothing beyond the bare text of a well-known work, others are well provided, not only with commentaries which are almost sure to be mentioned on the title-page, or in the colophon (which the editor himself often wrote), but also with dedicatory letters, prefaces, complimentary verses, indexes and other accessories, the presence of which it is desirable to indicate. In these cases it is often convenient to show the entire contents of the book in the order in which they occur, noting the leaves or pages on which each begins. Thus in the first edition (1590) of the first three books of Spenser's Faerie Queene, the literary contents, their order, and the space they occupy can be concisely noted by taking the successive gatherings 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 2 -00 8, text of books i.-iii.; Pp 1, letter dated the 23rd of January 1589 [ 1590] to Sir Walter Raleigh expounding the intention of the work; Pp 3 verso, commendatory verses signed W. R[aleigh], Hobynoll (Gabriel Harvey), R.S. H.B., W.L. and Ignoto; Pp 5 - 81 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 "; Qql-4, fifteen other sonnets.

Some bibliographers prefer to reverse the order of notation, (title, A,, recto; dedication, A 1, verso, &c.), and no principle is sacrificed in doing so, though the order suggested usually works out the more neatly.

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.

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 recondite. 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 titles, 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 information about individual books, a strictly alphabetical arrangement " by authors and titles " (i.e. by the names of authors in their alphabetical order, and the titles of their books in alphabetical sequence under the names) will be the most useful, because it enables the student to obtain the information he seeks with the greatest ease. But while such an alphabetical arrangement offers the speediest access to individual entries, it has no other merit, unless the main object of the bibliography be to show what each author has written. If it is desired to illustrate the history and development of a subject, or the literary biography of an author, the books should be entered chronologically. If direction in reading is to be given, this can best be offered by a subject-index, in which the subjects are arranged alphabetically for speedy reference, and the books chronologically under the subject, so that the newest are always at the end. Lastly if the object is to show how far the whole field has been covered and what gaps remain to be filled, a class catalogue arranged according to what are considered the logical subdivisions of the subject has its advantages. It is important, however, to remember that, if the bulk of the bibliography is very large, a principle of arrangement which would be clear and useful on a small scale may be lost in the quantity of pages over which it extends. An arrangement which cannot be quickly grasped, whatever satisfaction it may give its author, is useless to readers, the measure of its inutility being the worn condition of the alphabetical index to which those who cannot carry a complicated " logical " arrangement in their heads are obliged to turn, in the first instance, to find what they want. It should be obvious that any system which necessitates a preliminary reference to a key or index rests under grave suspicion, and needs 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 noteworthy 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 Bibliograplticum, 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 visible, would have been obscured by the fuller entries causing it to extend over many pages.

The Bibliography of Bibliography

The zeal of students of early printing has provided the material for an almost exhaustive list (see Incunabula) of the books printed in the 15th century still extant. Of those printed in the years 1501-1536 there is a tentative enumeration in the continuation of Panzer's Annales Typographici (1803), and materials are gradually being collected for improving and extending this. But the projects once formed for a universal bibliography have dwindled in proportion as the output of the press has increased, and the nearest approaches to such a work are the printed catalogue of the library of the British Museum, and that of the Bibliotheque Nationale at Paris, now in progress. Of books of great rarity unrepresented in these catalogues a fairly sufficient record exists in Brunet's Manuel du libraire, the bibliographical collections of Mr W. C. Hazlitt, the Bibliographer's Manual by Lowndes, and the other bibliographical works enumerated in the article on book-collecting. When a universal bibliography was recognized as an impossibility, patriotism suggested the compilation of national bibliographies, and the Bibliotheca Britannica of Robert Watt (Edinburgh, 1824) remains an extraordinary example of what the zeal of a single man could accomplish in this direction, Querard's La France littdraire (Paris, 1827-1839), while it gives fuller titles, is much less comprehensive, embracing mainly books of the 18th and early 19th centuries, and only such of these as appeared to the compiler to be written by " savants, historiens, et gens de lettres." In the works of Heinsius ( Allgemeines Bilcherlexikon, 170o-1815, Leipzig, 1812-1817), and Kayser ( Bilcherlexikon, 1750,. c., Leipzig, 1834, &c.) Germany possesses a fine record of her output of books during the last two centuries, and since the organization of the book-trade, contemporary lists of books, with resumes and indexes issued at intervals, exist for most European countries. For the period before these became of importance in England much bibliographical material has been collected in the Catalogues of English Books printed up to the end of the year 1640, issued by the British Museum in 1884, by the John Rylands library, Manchester, in 1895, and by the University library, Cambridge, in 1900-1906. A similar record of the rich English collections in the Bodleian library, Oxford, remains a great desideratum. While these substitutes for a universal author catalogue have gradually been provided, similar contributions to a universal subject catalogue have been made in the form of innumerable special bibliographies compiled by students or bookmen interested in special subjects or departments of literature. The most important of these are enumerated in the bibliographical notes appended to articles in this Encyclopaedia, but many attempts have been made to compile separate catalogues of them.

The most recent of these bibliographies of bibliographies naturally take over all that is of any value in their predecessors, and it may suffice therefore to make special mention of the following : - Bibliotheca bibliographica. Kritisches Verzeichniss der das Gesammtgebiet der Bibliographie betreffenden Litteratur des Inand Auslandes, in systematisches Ordnung bearheitet von Dr Julius Petzholdt. Mit alphabetischen Namen and Sachregister (Leipzig, 1866), 8vo, pp. xii. 940; Manuel de bibliographie gendrale, par Henri Stein (Paris, 1898), 8vo, pp. xx. 896; Manuel de bibliographie historique, par Ch. V. Langlois (Paris, 1901), 12mo, pp. xi. 623; A Register of National Bibliography. With a selection of the chief bibliographical works and articles printed in other Countries, by W. P. Courtney (London, 1905), 8vo, pp. viii. 631.

It should also be noted that the List of Books of Reference in the Reading-Room of the British Museum, first published in 1889, and the Subject-index of the Modern Works added to the Library of the British Museum in the y ears .1881-190o, edited by G. K. Fortescue (supplements published every five years), include entries of a vast number of bibliographical works, and that an eclectic list, with a valuable introduction, will be found in Professor Ferguson's Some Aspects of Bibliography (Edinburgh, 1900). (A. W. Po.)