Catalogues.

Catalogues of libraries are useful to readers in direct proportion to their fulfilment of three conditions: (1) Quick and ready reference. (2) Arranging all authors' names in an alphabet, followed by titles of their works. (3) Subjects or titles in their alphabetical order in the same alphabet as the authors. This is what is known as a "Dictionary catalogue"; but why is it preferable to any other? Because it answers more questions in less time than any other.

The more prevalent styles of catalogues have been, 1. A list of authors, with titles of their works under each. 2. A catalogue of subjects, in a classified topical or alphabetical order, the authors and their works being grouped under each head. 3. A catalogue attempting to combine these two, by appending to the author-catalogue a classed list of subjects, with a brief of authors under each, referring to the page on which the titles of their works may be found; or else, 4. Appending to the subject-catalogue an alphabet of authors, with similar references to pages under subjects.

Each of these methods of catalogue-making, while very useful, contrives to miss the highest utility, which lies in enabling the reader to put his finger on the book he wants, at one glance of the eye. The catalogue of authors will not help him to subjects, nor will the catalogue of subjects, as a rule, give the authors and titles with the fullness that may be needed. In either case, a double reference becomes necessary, consuming just twice the time, and in a two-column catalogue, three times the time required in a dictionary catalogue.

The reader who wants Darwin's "Origin of Species" finds it readily enough by the author-catalogue; but he wants, at the same time, to find other works on the same subject, and all the author-catalogues in the world will not help him to them. But give him a dictionary catalogue, and he has, in the same alphabet with his Darwin, (if the library is large) dozens of books discussing the theory of that great naturalist, under species, evolution, Darwinism, etc.

Thus he finds that there is no key which so quickly unlocks the stores of knowledge which a library contains, as a dictionary catalogue.

The objections to it are chiefly brought by minds schooled in systems, who look askance on all innovations, and instinctively prefer round-about methods to short-hand ones.

Ask such an objector if he would prefer his dictionary of the English language arranged, not alphabetically, but subjectively, so that all medical terms should be defined only under medicine, all species of fish described only under fishes, etc., and he will probably say that there is no analogy in the case. But the analogy becomes apparent when we find, in what are called systematic catalogues, no two systems alike, and the finding of books complicated by endless varieties of classification, with no common alphabet to simplify the search. The authors of systems doubtless understand them themselves, but no one else does, until he devotes time to learn the key to them; and even when learned, the knowledge is not worth the time lost in acquiring it, since the field covered in any one catalogue is so small. Alphabetical arrangement, on the other hand, strictly adhered to, is a universal key to the authors and subjects and titles of all the books contained in the library it represents. The devotee of a bibliographical system may be as mistaken as the slave of a scientific terminology. He forgets that bibliography is not a school for teaching all departments of knowledge, but a brief and handy index to books that may contain that knowledge. A student who has once made a thorough comparative test of the merits, as aids to wide and rapid research, of the old-fashioned bibliographies and the best modern dictionary catalogues, will no more deny the superiority of the latter, than he will contest the maxim that a straight line is the nearest road between two points. Meantime, "while doctors disagree, disciples are free;" and the disciples who would follow the latest guides in the art "how to make and use a catalogue," must get rid of many formulas.

The reader will find in the chapter on bibliography, notes on some classes of catalogues, with the more notable examples of them. We are here concerned with the true method of preparing catalogues, and such plain rules as brevity will permit to be given, will be equally adapted to private or public libraries. For more ample treatment, with reasons for and against many rules laid down, reference is made to the able and acute work, "Rules for a Dictionary Catalogue," by C. A. Cutter, published by the U. S. Bureau of Education, 3d ed. 1891.

CONDENSED RULES FOR AN AUTHOR AND TITLE CATALOGUE.

Prepared by the Co-operation Committee of the American Library Association.

entry.

Books are to be entered under the:

Surnames of authors when ascertained, the abbreviation "Anon." being added to the titles of anonymous works.

Initials of authors' names when these only are known, the last initial being put first.

Pseudonyms of the writers when the real names are not ascertained.

Names of editors of collections, each separate item to be at the same time sufficiently catalogued under its own heading.

Names of countries, cities, societies, or other bodies which are responsible for their publication.

First word (not an article or serial number) of the titles of periodicals and of anonymous books, the names of whose authors are not known. And a motto or the designation of a series may be neglected when it begins a title, and the entry may be made under the first word of the real title following.

Commentaries accompanying a text, and translations, are to be entered under the heading of the original work; but commentaries without the text under the name of the commentator. A book entitled "Commentary on ...." and containing the text, should be put under both.

The Bible, or any part of it (including the Apocrypha), in any language, is to be entered under the word Bible.

The Talmud and Koran (and parts of them) are to be entered under those words; the sacred books of other religions are to be entered under the names by which they are generally known; references to be given from the names of editors, translators, etc.

The respondent or defender of an academical thesis is to be considered as the author, unless the work unequivocally appears to be the work of the praeses.

Books having more than one author to be entered under the one first named in the title, with a reference from each of the others.

Reports of civil actions are to be entered under the name of the party to the suit which stands first on the title page. Reports of crown and criminal proceedings are to be entered under the name of the defendant. Admiralty proceedings relating to vessels are to be put under the name of the vessel.

Noblemen are to be entered under their titles, unless the family name is decidedly better known.

Ecclesiastical dignitaries, unless popes or sovereigns, are to be entered under their surnames.

Sovereigns (other than Greek or Roman), ruling princes, Oriental writers, popes, friars, persons canonized, and all other persons known only by their first name, are to be entered under this first name.

Married women, and other persons who have changed their names, are to be put under the last well-known form.

A pseudonym may be used instead of the surname (and only a reference to the pseudonym made under the surname) when an author is much more known by his false than by his real name. In case of doubt, use the real name.

A society is to be entered under the first word, not an article, of its corporate name, with references from any other name by which it is known, especially from the name of the place where its headquarters are established, if it is often called by that name.

References.

When an author has been known by more than one name, references should be inserted from the name or names not to be used as headings to the one used.

References are also to be made to the headings chosen:

Headings.

In the heading of titles, the names of authors are to be given in full, and in their vernacular form, except that the Latin form may be used when it is more generally known, the vernacular form being added in parentheses; except, also, that sovereigns and popes may be given in the English form.

English and French surnames beginning with a prefix (except the French de and d') are to be recorded under the prefix; in other languages under the word following.

English compound surnames are to be entered under the last part of the name; foreign ones under the first part.

Designations are to be added to distinguish writers of the same name from each other.

Prefixes indicating the rank or profession of writers may be added in the heading, when they are part of the usual designation of the writers.

Names of places to be given in the English form. When both an English and a vernacular form are used in English works, prefer the vernacular.

Titles.

The title is to be an exact transcript of the title-page, neither amended, translated, nor in any way altered, except that mottos, titles of authors, repetitions, and matter of any kind not essential, are to be omitted. Where great accuracy is desirable, omissions are to be indicated by three dots (...). The titles of books especially valuable for antiquity or rarity may be given in full, with all practicable precision. The phraseology and spelling, but not necessarily the punctuation, of the title are to be exactly copied.

Any additions needed to make the title clear are to be supplied, and inclosed by brackets.

Initial capitals are to be given in English:

Foreign languages.—Titles in foreign characters may be transliterated. The languages in which a book is written are to be stated when there are several, and the fact is not apparent from the title.

Imprints.

After the title are to be given, in the following order, those in [ ] being optional:

After the place of publication, the place of printing may be given if different. This is desirable only in rare and old books.

The number of pages is to be indicated by giving the last number of each paging, connecting the numbers by the sign +; the addition of unpaged matter may be shown by a +, or the number of pages ascertained by counting may be given in brackets. When there are more than three pagings, it is better to add them together and give the sum in brackets.

These imprint entries are to give the facts, whether ascertained from the book or from other sources; those which are usually taken from the title (edition, place, publisher's name, and series) should be in the language of the title, corrections and additions being inclosed in brackets. It is better to give the words, "maps," "portraits," etc., and the abbreviations for "volumes" and "pages," in English.

Contents, Notes.

Notes (in English) and contents of volumes are to be given when necessary to properly describe the works. Both notes and lists of contents to be in a smaller type.

Miscellaneous.

A single dash or indent indicates the omission of the preceding heading; a subsequent dash or indent indicates the omission of a subordinate heading, or of a title.

A dash connecting numbers signifies to and including; following a number it signifies continuation.

A ? following a word or entry signifies probably.

Brackets inclose words added to titles or imprints, or changed in form.

Arabic figures are to be used rather than Roman; but small capitals may be used after the names of sovereigns, princes, and popes.

A list of abbreviations to be used was given in the Library journal, Vol. 3: 16-20.

Arrangement.

The surname when used alone precedes the same name used with forenames; where the initials only of the forenames are given, they are to precede fully written forenames beginning with the same initials (e. g., Brown, Brown, J.; Brown, J. L.; Brown, James).

The prefixes M and Mc, S., St., Ste., Messrs., Mr., and Mrs., are to be arranged as if written in full, Mac, Sanctus, Saint, Sainte, Messieurs, Mister, and Mistress.

The works of an author are to be arranged in the following order:

The order of alphabeting is to be that of the English alphabet.

The German ae, oe, ue, are always to be written as ä, ö, ü, and arranged as a, o, u.

Names of persons are to precede similar names of places, which in turn precede similar first words of titles.

A few desirable modifications or additions to these rules may be suggested.

1. In title-entries, let the year of publication stand last, instead of the indication of size.

2. Noblemen to be entered under their family names, with reference from their titles.

3. Instead of designations of title, profession, residence, or family, to distinguish authors, let every name be followed by the chronology, as—

James (Henry) 1811-82.
James (Henry) 1843-

It is highly desirable to give this information as to the author's period in every title-heading, without exception, when ascertainable. If unknown, the approximate period to be given, with a query.

4. All titles to be written in small letters, and printed in lower case, whether in English, German, or any other language, avoiding capitals except in cases named in the rule.

5. Works without date, when the exact date is not found, are to be described conjecturally, thus:

[1690?] or [about 1840.]

6. In expressing collations, use commas rather than the sign + between the pagings, as—xvi, 452, vii pp.—not xvi+452+vii pp.

7. Forenames should be separated from the surnames which precede them by parenthesis rather than commas, as a clearer discrimination: as—

Alembert (Jean Baptiste le Rond d')—not
Alembert, Jean Baptiste le Rond d'.

The printed catalogue of the British Museum Library follows this method, as well as that in the preceding paragraph.

8. All books of history, travels, or voyages to have the period covered by them inserted in brackets, when not expressed in the title-page.

9. All collected works of authors, and all libraries or collections of different works to be analysed by giving the contents of each volume, either in order of volumes, or alphabetically by authors' names.

Of course there are multitudes of points in catalogue practice not provided for in the necessarily brief summary preceding: and, as books on the art abound, the writer gives only such space to it as justice to the wide range of library topics here treated permits.

Probably the most important question in preparing catalogue titles, is what space to give to the author's frequently long-drawn-out verbiage in his title-page. There are two extremes to be considered: (1) Copying the title literally and in full, however prolix; and (2) reducing all title-pages, by a Procrustean rule, to what we may call "one-line titles." Take an example:

"Jones (Richard T.) A theoretical and practical treatise on the benefits of agriculture to mankind. With an appendix containing many useful reflections derived from practical experience. iv, 389 pp. 8°. London, MDCCXLIV." As abridged to a short title, this would read: "Jones (Richard T.) Benefits of agriculture, iv, 389 pp. 8°. Lond. 1744." Who will say that the last form of title does not convey substantially all that is significant of the book, stripped of superfluous verbiage? But we need not insist upon titles crowded into a single line of the catalogue, whether written or printed. This would do violence to the actual scope of many books, by suppressing some significant or important part of their titles. The rule should be to give in the briefest words selected out of the title (never imported into it) the essential character of the book, so far as the author has expressed it. Take another example:

"Bowman (Thomas) A new, easy, and complete Hebrew course; containing a Hebrew grammar, with copious Hebrew and English exercises, strictly graduated: also, a Hebrew-English and English-Hebrew lexicon. In two parts. Part I. Regular verbs. Edinburgh, 1879."

This might be usefully condensed thus:

Bowman, (Thomas) Hebrew course: grammar, exercises, lexicon, [&c.] Part I. Regular verbs. Edinburgh, 1879.

One objection brought against the dictionary catalogue is that it widely separates subjects that belong together. In the Boston Athenaeum catalogue, for example, the topic Banks is found in Vol. 1, while Money is in Vol. 3; and for Wages, one must go to Vol. 5, while Labor is in Vol. 3. But there are two valid reasons for this. First, the reader who wants to know about banks or wages may care nothing about the larger topics of money or of labor; and secondly, if he does want them, he is sent to them at once by cross-reference, where they belong in the alphabet; whereas, if they were grouped under Political Economy, as in classed catalogues, he must hunt for them through a maze of unrelated books, without any alphabet at all.

It is often forgotten by the advocates of systematic subject catalogues rather than alphabetical ones, that catalogues are for those who do not know, more than for those who do. The order of the alphabet is settled and familiar; but no classification by subjects is either familiar or settled. Catalogues should aim at the greatest convenience of the greatest number of readers.

It is noteworthy that the English Catalogue (the one national bibliography of the current literature of that country) has adopted, since 1891, the dictionary form of recording authors, titles and subjects in one alphabet, distinguishing authors' names by antique type. It is hoped that the American Catalogue, an indispensable work in all libraries, will adopt in its annual and quinquennial issues the time-saving method of a single alphabet.

It is not claimed that the dictionary catalogue possesses fully all the advantages in educating readers that the best classed catalogues embody. But the chief end of catalogues being to find books promptly, rather than to educate readers, the fact that the dictionary catalogue, though far from perfect, comes nearer to the true object than any other system, weighs heavily in its favor. Edward Edwards said—"Many a reader has spent whole days in book-hunting [in catalogues] which ought to have been spent in book-reading." It is to save this wasted time that catalogues should aim.

Nothing can be easier than to make a poor catalogue, while nothing is more difficult than to make a good one. The most expert French bibliographers who have distinguished themselves by compiling catalogues have been most severely criticised by writers who no doubt would have been victimized in their turn if they had undertaken similar work. Byron says

"A man must serve his time to every trade,
Save censure;—critics all are ready made."

When De Bure and Van Praet, most accomplished bibliographers, published the catalogue of the precious library of the duke de La Vallière, the abbé Rive boasted that he had discovered a blunder in every one of the five thousand titles of their catalogue. Barbier and Brunet have both been criticised for swarms of errors in the earlier editions of their famous catalogues. The task of the exact cataloguer is full of difficulty, constantly renewed, and demanding almost encyclopaedic knowledge, and incessant care of minute particulars.

The liability to error is so great in a kind of work which, more than almost any other, demands the most scrupulous accuracy, lest a catalogue should record a book with such mistakes as to completely mislead a reader, that rules are imperatively necessary. And whatever rules are adopted, a rigid adherence to them is no less essential, to avoid misapprehension and confusion. A singular instance of imperfect and misleading catalogue work was unwittingly furnished by Mr. J. Payne Collier, a noted English critic, author, and librarian, who criticised the slow progress of the British Museum catalogue, saying that he could himself do "twenty-five titles an hour without trouble." His twenty-five titles when examined, were found to contain almost every possible error that can be made in cataloguing books. These included using names of translators or editors as headings, when the author's name was on the title-page; omitting christian names of authors; omitting to specify the edition; using English instead of foreign words to give the titles of foreign books; adopting titled instead of family names for authors (which would separate Stanhope's "England under Queen Anne" from the same writer's "History of England," published when he was Lord Mahon); errors in grammar, etc. These ridiculous blunders of a twenty-five-title-an-hour man exemplify the maxim "the more haste, the worse speed," in catalogue-making.

That our British brethren are neither adapted nor inclined to pose as exemplars in the fine art of cataloguing, we need only cite their own self-criticisms to prove. Here are two confessions found in two authors of books on catalogue-making, both Englishmen. Says one: "We are deficient in good bibliographies. It is a standing disgrace to the country that we have no complete bibliography of English authors, much less of English literature generally." Says another: "The English are a supremely illogical people. The disposition to irregularity has made English bibliography, or work on catalogues, a by-word among those who give attention to these matters."

An American may well add, "They do these things better in France and Germany," while declining to claim the meed of superiority for the United States.

Too much prominence should not be given to place-numbers in library catalogues. The tendency to substitute mere numerical signs for authors and subjects has been carried so far in some libraries, that books are called for and charged by class-numbers only, instead of their distinctive names. An English librarian testifies that assistants trained in such libraries are generally the most ignorant of literature. When mechanical or mnemonical signs are wholly substituted for ideas and for authors, is it any wonder that persons incessantly using them become mechanical? Let catalogue and classification go hand in hand in bringing all related books together, and library assistants will not stunt their intellects by becoming bond-slaves to the nine digits, nor lose the power of thought and reflection by never growing out of their a b c's.

There are two forms of catalogue not here discussed, which are adjuncts to the library catalogue proper. The accession catalogue, kept in a large volume, records the particulars regarding every volume, on its receipt by the library. It gives author, title, date, size, binding, whence acquired, cost, etc., and assigns it an accession number, which it ever after retains. The shelf catalogue (or shelf-list) is a portable one divided into sections representing the cases of shelves in the library. It gives the shelf classification number, author, brief title and number of volumes of each book, as arranged on the shelves; thus constituting an inventory of each case, or stack, throughout the library.

To check a library over is to take an account of stock of all the books it should contain. This is done annually in some libraries, and the deficiencies reported. All libraries lose some books, however few, and these losses will be small or great according to the care exercised and the safe-guards provided. The method is to take one division of the library at a time, and check off all books on the shelves by their numbers on the shelf-list, supplemented by careful examination of all numbers drawn out, or at bindery, or in other parts of the library. Not a volume should be absent unaccounted for. Those found missing after a certain time should be noted on the shelf-list and accession book, and replaced, if important, after the loss is definitely assured.

The reason for writing and printing all catalogue titles in small letters, without capitals (except for proper names) is two-fold. First, there can be no standard prescribing what words should or should not be capitalized, and the cataloguer will be constantly at a loss, or will use capitals in the most unprincipled way. He will write one day, perhaps, "The Dangers of great Cities," and the next, "The dangers of Great cities"—with no controlling reason for either form. Secondly, the symmetry of a title or a sentence, whether written or printed, is best attained by the uniform exclusion of capitals. That this should be applied to all languages, notwithstanding the habit of most German typographers of printing all nouns with capitals, is borne out by no less an authority than the new Grimm's Deutsches Wörterbuch, which prints all words in "lower case" type except proper names. Nothing can be more unsightly than the constant breaking up of the harmony of a line by the capricious use of capitals.

To discriminate carefully the various editions of each work is part of the necessary duty of the cataloguer. Many books have passed through several editions, and as these are by no means always specified on the title-page, one should establish the sequence, if possible, by other means. The first edition is one which includes all copies printed from the plates or the type as first set; the second, one which is reprinted, with or without changes in the text or the title. First editions often acquire a greatly enhanced value, in the case of a noted author, by reason of changes made in the text in later issues of the work. For though the latest revision may and should be the author's best improved expression, his earliest furnishes food for the hunters of literary curiosities. Every catalogue should distinguish first editions thus [1st ed.] in brackets.

In the arrangement of titles in catalogues, either of the various works of the same writer, or of many books on the same subject, some compilers follow the alphabetical order, while others prefer the chronological—or the order of years of publication of the various works. The latter has the advantage of showing the reader the earlier as distinguished from the recent literature, but in a long sequence of authors (in a subject-catalogue) it is more difficult to find a given writer's work, or to detect its absence.

The task of accurately distributing the titles in a catalogue of subjects would be much simplified, if the books were all properly named. But it is an unhappy failing of many writers to give fanciful or far-fetched titles to their books, so that, instead of a descriptive name, they have names that describe nothing. This adds indefinitely to the labor of the cataloguer, who must spend time to analyse to some extent the contents of the book, before he can classify it. This must be done to avoid what may be gross errors in the catalogue. Familiar examples are Ruskin's Notes on Sheep-folds (an ecclesiastical criticism) classified under Agriculture; and Edgeworth's Irish Bulls under Domestic animals.

The work of alphabeting a large number of title-cards is much simplified and abbreviated by observing certain obvious rules in the distribution. (1) Gather in the same pile all the cards in the first letter of the alphabet, A, followed in successive parallel rows by all the B's, and so on, to the letter Z. (2) Next, pursue the same course with all the titles, arranging under the second letter of the alphabet, Aa, Ab, Ac, etc., and so with all the cards under B. C. &c. for all the letters. (3) If there still remain a great many titles to distribute into a closer alphabetic sequence, the third operation will consist in arranging under the third letter of the alphabet, e. g., Abb, Abc, Abd, etc. The same method is pursued throughout the entire alphabet, until all the title-cards are arranged in strict order.

Too much care cannot be taken to distinguish between books written by different authors, but bearing the same name. Many catalogues are full of errors in this respect, attributing, for example, works written by Jonathan Edwards, the younger, (1745-1801) to Jonathan Edwards the elder, (1703-58); or cataloguing under Henry James, Jr., the works of his father, Henry James. The abundant means of identification which exist should cause such errors to be avoided; and when the true authorship is fixed, every author's chronology should appear next after his name on every card-title: e. g. James (Henry, 1811-82) Moralism and Christianity, New York, 1850. James (Henry, 1843- ) Daisy Miller, N. Y. 1879.

The designation of book sizes is a vexed question in catalogues. The generally used descriptions of size, from folio down to 48mo. signify no accurate measurement whatever, the same book being described by different catalogues as 12mo. 8vo, crown 8vo. &c., according to fancy; while the same cataloguer who describes a volume as octavo to-day, is very likely to call it a duodecimo to-morrow. Library catalogues are full of these heterogeneous descriptions, and the size-notation is the bête noir of the veteran bibliographer, and the despair of the infant librarian. Yet it is probable that the question has excited a discussion out of all proportion to its importance. Of what consequence is the size of a book to any one, except to the searcher who has to find it on the shelves? While the matter has been much exaggerated, some concert or uniformity in describing the sizes of books is highly desirable.

A Committee of the American Library Association agreed to a size-notation, figured below, adopting the metric system as the standard, to which we add the approximate equivalents in inches.

Sizes.  Size abbreviations.  Centimetres
outside height.		  Inches.
Folio, F°.F4016
Quarto, 4°.Q3012
Octavo, 8°.O2510
Duodecimo, 12°.D208
Sixteen mo., 16°.S17.57
Twenty-four mo., 24°.T156
Thirty-two mo., 32°.Tt12.35
Forty-eight mo., 48°.Fe104

It will be understood that the figure against each size indicated represents the maximum measure: e. g. a volume is octavo when above 20 and below 25 centimetres (8 to 10 inches high).

As this question of sizes concerns publishers and booksellers, as well as librarians, and the metric system, though established in continental Europe, is in little use in the United States and England, it remains doubtful if any general adherence to this system of notation can be reached—or, indeed, to any other. The Publishers' Weekly (N. Y.) the organ of the book trade, has adopted it for the titles of new books actually in hand, but follows the publishers' descriptions of sizes as to others. Librarian J. Winter Jones, of the British Museum, recommended classing all books above twelve inches in height as folios, those between ten and twelve inches as quartos, those from seven to ten inches as octavos, and all measuring seven inches or under as 12mos. Mr. H. B. Wheatley, in his work, "How to Catalogue a Library," 1889, proposed to call all books small octavos which measure below the ordinary octavo size. As all sizes "run into each other," and the former classification by the fold of the sheets is quite obsolete, people appear to be left to their own devices in describing the sizes of books. While the metric notation would be exact, if the size of every book were expressed in centimetres, the size-notation in the table given is wholly wanting in precision, and has no more claim to be adopted than any other arbitrary plan. Still, it will serve ordinary wants, and the fact that we cannot reach an exact standard is no reason for refusing to be as nearly exact as we can.

And while we are upon the subject of notation may be added a brief explanation of the method adopted in earlier ages, (and especially the years reckoned from the Christian era) to express numbers by Roman numerals. The one simple principle was, that each letter placed after a figure of greater equal value adds to it just the value which itself has; and, on the other hand, a letter of less value placed before (or on the left of) a larger figure, diminishes the value of that figure in the same proportion. For example:

These letters—VI represent six; which is the same as saying V+I. On the contrary, these same letters reversed represent four; thus—IV: that is V-I=4. Nine is represented by IX, i. e., X-I, ten minus one. On the same principle, LX represents 60—or L+N: whereas XL means 40—being L-X. Proceeding on the same basis, we find that LXX=L+XX=70; and LXXX or L+XXX is 80. But when we come to ninety, instead of adding four X's to the L, they took a shorter method, and expressed it in two figures instead of five, thus, XC, i. e. 100 or C-X=90.

The remarkable thing about this Roman notation is that only six letters sufficed to express all numbers up to one thousand, and even beyond, by skilful and simple combinations: namely the I, the V, the X, the L, the C, and the M, and by adding or subtracting some of these letters, when placed before or after another letter, they had a whole succession of numbers done to their hand—thus:

I,1XX,20CC,200
II,2XXX,30CCC,300
III,3XL,40CCCC,400
IV,4L,50D,500
V,5LX,60DC,600
VI,6LXX,70DCC,700
VII,7LXXX,80DCCC,800
VIII,8XC,90CM,900
IX,9 C (centum),100 M, (mille),1,000
X,10

Now, when the early printers came to apply dates of publication to the books they issued, (and here is where their methods of notation become most important to librarians) they used precisely these methods. For example, to express the year 1695, they printed it thus: MDCVC, that is—1000+500+100+100-5. But the printers of the 15th century and later, often used complications of letters, dictated by caprice rather than by any fixed principles, so that it is sometimes difficult to interpret certain dates in the colophons or title-pages of books, without collateral aid of some kind, usually supplied to the librarian by bibliographies. One of the simpler methods of departure from the regular notation as above explained, was to substitute for the letter D (500) two letters, thus—IƆ, an I and a C inverted, supposed to resemble the letter D in outline. Another fancy was to replace the M, standing for 1,000, by the symbols CIƆ—which present a faint approach to the outline of the letter M, for which they stand. Thus, to express the year 1610, we have this combination—CIƆ IƆ CX, which would be indecipherable to a modern reader, uninstructed in the numerical signs anciently used, and their values. In like manner, 1548 is expressed thus: MDXLIIX, meaning 1000+500+40+10-2. And for 1626, we have CIƆ IƆ C XXVI.

As every considerable library has early printed books, a librarian must know these peculiarities of notation, in order to catalogue them properly, without mistake as to their dates. In some books, where a capricious combination of Roman numerals leaves him without a precedent to guide him to the true date, reference must be had to the bibliographies of the older literature, (as Hain, Panzer, etc.), which will commonly solve the doubt.

As to the mechanics of catalogue-making, widely different usages and materials prevail. In America, the card or title-slip system is well nigh universal, while in England it is but slowly gaining ground, as against the ledger or blank book catalogue. Its obvious advantage lies in affording the only possible means of maintaining a strict alphabetical sequence in titles, whether of authors or subjects. The title-cards should be always of uniform size, and the measure most in vogue is five inches in length by three inches in breadth. They should not be too stiff, though of sufficient thickness, whether of paper or of thin card board, to stand upright without doubling at the edges. They may be ruled or plain, at pleasure, and kept in drawers, trays, or (in case of a small catalogue) in such paste-board boxes as letter envelopes come in.

The many advantages of the card system, both for catalogues and indexes, should not lead us to overlook its palpable defects. These are (1) It obliges readers to manipulate many cards, to arrive at all the works of an author, or all the books on any subject, instead of having them under his eye at once, as in printed catalogues. (2) It can be used only in the library, and in only one place in the library, and by only one person at a time in the same spot, while a printed catalogue can be freely used anywhere, and by any numbers, copies being multiplied. (3) It entails frequent crowding of readers around the catalogue drawers, who need to consult the same subjects or authors at the same time. (4) It requires immeasurably more room than a printed catalogue, and in fact, exacts space which in some libraries can be ill afforded. (5) It obliges readers to search the title-cards at inconvenient angles of vision, and often with inadequate light. (6) It is cumbersome in itself, and doubly cumbersome to searchers, who must stand up instead of sitting to consult it, and travel from drawer to drawer, interfering with other searchers almost constantly, or losing time in waiting. (7) To this is added the inconvenience of constant insertion of new title-cards by members of the library staff, and the time-consuming process of working the rods which keep the cards in place, if they are used, and if not used, the risk of loss of titles, or misplacement equivalent to loss for a time.

Says Mr. H. B. Wheatley: "I can scarcely imagine anything more maddening than a frequent reference to cards in a drawer." But it is to be considered that all systems have defects, and the problem of choosing the least defective is ever before us. Most of the suggested defects of the card catalogue, as concerns the readers, can be obviated by making a two-fold catalogue, the type-written titles being manifolded, and one set arranged in card-drawers for the use of the library staff, while another is mounted on large sheets in bound volumes for use of the public. This would secure the advantages of a printed catalogue, with no more expense than the manuscript titles would cost. If desired, a number of copies could be bound up for reading-room use. Accessions of new books could be incorporated from month to month, by leaving the right-hand pages blank for that purpose. This would be near enough to alphabetical order for most readers, with the immense advantage of opening at one glance before the eye, any author or subject. It would go far to solve the problem how to unite the flexibility and perfect alphabeting of the card system, with the superior comfort, safety, and ease of reference of the book. It would also be a safe-guard against the loss or displacement of titles, a danger inherent in the card system, as they could be replaced by copying missing titles from the catalogue volumes.

While the undoubted merits of the card system have been much overrated, it would be as unwise to dispense with it as the complete official catalogue of the library, as it would be to tie down the public to its use, when there is a more excellent way, saving time and patience, and contributing to the comfort of all.

To print or not to print? is a vital question for libraries, and it is in most cases decided to forego or to postpone printing, because of its great expense. Yet so manifest are the advantages of a printed catalogue, that all public libraries should make every effort to endow their readers with its benefits. These advantages are (1) Greater facility of reading titles. (2) Much more rapid turning from letter to letter of the catalogue alphabet. (3) Ability to consult it outside of the library. (4) Unlimited command of the catalogue by many readers at once, from the number of copies at hand, whereas card catalogues or manuscript volumes involve loss of time in waiting, or interfering with the researches of others. A part of these advantages may be realized by printing type-written copies of all titles in duplicate, or by carbon paper in manifold, thus furnishing the library with several copies of its catalogue: but why not extend this by multiplying copies through the ingenious processes now in use, by which the printing of titles can be effected far more cheaply than in any printing office? Might not every library become its own printer, thus saving it from the inconvenience and risk of sending its titles outside, or the great expense of copying them for the printer?

The titles thus manifolded could be combined into volumes, by cutting away all superfluous margins and mounting the thin title-slips alphabetically on paper of uniform size, which, when bound, would be readily handled. All the titles of an author's works would be under the eye at a glance, instead of only one at a time, as in the card catalogue. And the titles of books on every subject would lie open, without slowly manipulating an infinite series of cards, one after another, to reveal them to the eye. The classification marks could be readily placed against each title, or even printed as a part of the manifold card titles.

Not that the card catalogue system would be abolished: it would remain as the only complete catalogue of the library, always up to date, in a single alphabet. Daily accessions inserted in it would render it the standard of appeal as to all that the library contained, and it would thus supplement the printed catalogue.

Of course, large and increasing accessions would require to be combined in occasional supplementary volumes of the catalogue; and in no long number of years the whole might be re-combined in a single alphabet, furnishing a printed dictionary catalogue up to its date.

The experience of the great British Museum Library in this matter of catalogues is an instructive one. After printing various incomplete author-catalogues in the years from 1787 to 1841, the attempt to print came to a full stop. The extensive collection grew apace, and the management got along somehow with a manuscript catalogue, the titles of which (written in script with approximate fullness) were pasted in a series of unwieldy but alphabetically arranged volumes. To incorporate the accessions, these volumes had continually to be taken apart by the binder, and the new titles combined in alphabetical order, entailing a literally endless labor of transcribing, shifting, relaying and rebinding, to secure even an imperfect alphabetical sequence. In 1875, the catalogue had grown to over two thousand thick folio volumes, and it was foreseen, by a simple computation of the rate of growth of the library, that in a very few years its catalogue could no longer be contained in the reading-room. The bulky manuscript catalogue system broke down by its own weight, and the management was compelled to resort to printing in self defence. Before the printing had reached any where near the concluding letters of the alphabet, the MS. catalogue had grown to three thousand volumes, and was a daily and hourly incubus to librarians and readers.

This printed catalogue of the largest library in the world, save one, is strictly a catalogue of authors, giving in alphabetical order the names, followed by the titles of all works by each writer which that library possesses. In addition, it refers in the case of biographies or comments upon any writer found in the index, to the authors of such works; and also from translators or editors to the authors of the translated or edited work. The titles of accessions to the library (between thirty and forty thousand volumes a year) were incorporated year by year as the printing went on. All claim to minute accuracy had to be ignored, and the titles greatly abridged by omitting superfluous words, otherwise its cost would have been prohibitory. The work was prosecuted with great energy and diligence by the staff of able scholars in the service of the Museum Library. As the catalogue embraces far more titles of books, pamphlets, and periodicals than any other ever printed, it is a great public boon, the aid it affords to all investigators being incalculable. And any library possessing it may find, with many titles of rare and unattainable works, multitudes of books now available by purchase in the market, to enrich its own collection. It is said to contain about 3,500,000 titles and cross-references. It is printed in large, clear type, double columns, well spaced, and its open page is a comfort to the eye. Issued in paper covers, the thin folios can be bound in volumes of any thickness desired by the possessor.

It has several capital defects: (1) It fails to discriminate authors of the same name by printing the years or period of each; instead of which it gives designations like "the elder", "the younger", or the residence, or occupation, or title of the author. The years during which any writer flourished would have been easily added to the name in most cases, and the value of such information would have been great, solving at once many doubts as to many writers. (2) The catalogue fails to print the collations of all works, except as to a portion of those published since 1882, or in the newer portions issued. This omission leaves a reader uncertain whether the book recorded is a pamphlet or an extensive work. (3) The letters I and J and U and V are run together in the alphabet, after the ancient fashion, thus placing Josephus before Irving, and Utah after Virginia; an arrangement highly perplexing, not to say exasperating, to every searcher. To follow an obsolete usage may be defended on the plea that it is a good one, but when it is bad as well as outworn, no excuse for it can satisfy a modern reader. (4) No analysis is given of the collected works of authors, nor of many libraries made up of monographs. One cannot find in it the contents of the volumes of any of Swift's Works, nor even of Milton's Prose Writings. (5) It fails to record the names of publishers, except in the case of some early or rare books.

The printing of this monumental catalogue began in 1881, the volumes of MS. catalogue being set up by the printer without transcription, which would have delayed the work indefinitely, and it is now substantially completed. Its total cost will be not far from £50,000. There are about 374 volumes or parts in all. Only 250 copies were printed, part of which were presented to large libraries, and others were offered for sale at £3.10 per annum, payable as issued, so that a complete set costs about £70. One learns with surprise that only about forty copies have been subscribed for. This furnishes another evidence of the low estate of bibliography in England, where, in a nation full of rich book-collectors and owners of fine libraries, almost no buyers are found for the most extensive bibliography ever published, a national work, furnishing so copious and useful a key to the literature of the world in every department of human knowledge.

CHAPTER 23.