L

ADIES AND GENTLEMEN:- -We are to be congratulated on coming together in Washington; and we come here opportunely. We are glad to draw nearer at last to the National Bureau of Education, remembering how it signalized the centennial year for us by the publication of that encyclopædic report about the institutions we represent.

We are glad to find ourselves face to face with that ardent friend of bibliography, the librarian of the Surgeon-General's office, who has shown, not only us, but the experts of the older world, how the highest results of that science can be reached by a rare intelligence and a comprehensive energy.

We are proud to be, as it were, the guests of the Librarian of Congress. We come in full recognition of a merit that well befits his official dominance among us. We hope the cheer he gives us only foreruns the pleasure which is due to him, when he shall see the treasures of a national library spread in all their amplitude through a spacious depository, worthy of a great nation and worthy of him.

Before our sessions are over we shall have opportunity to inspect the plans which have been proposed for this great national library. Whatever the disposition to make it every way worthy of our needs and worthy of our resources, there must still be, in the construc

tion of it, errors to be escaped as well as merits to be embodied. The problem, it must be confessed, is not an easy one. It will not solve itself, like some political ones. This is to be confronted successfully only by a thorough understanding of the possibilities of the future. The mechanical devices for annihilating time and space present, in these days, the question of library construction in a changed light.

The new significance of libraries as the necessity of the many, as well as the essential home, as it may be, of the few, widens the field of observation, and makes the institution both a monument and an engine. The library has grown to have eminently a practical bearing upon our general education and upon our training as citizens. I think of it sometimes as a derrick, lifting the inert masses and swinging them round to the sure foundations upon which the national character shall rise. who have had daily dealings with the work of libraries know this to be something more than a piece of rhetoric. We may discuss the many recurring mooted points in our economy, the fiction question, or any other, - assuming hat we tread upon a vantage ground; we may peer through vistas of our own making, and think we see the universe; we may be uncircumspect; we may go on floundering, without

You

lead or compass, and while we are doing it the library has grown; men and women have come up to it, and taken something better than homilies. The beneficence of the world of books has been spread about, and the wheat has choked the tares.

I would not be blinded to the fact that mischief, and enough of it, may lurk in books. It will do its work in spite of us; but, if we would keep it at its minimum, we do not wisely make this mischief prominent. Our emphasis should be upon the wholesome, and upon

that which healthfully stimulates. I would put more trust in one such educational catalogue as the term is—like that, for instance, of the Brooklyn Library, — an admirable boon to all of us, than in scores of narrow visionaries, who do not know that it is the motes in their own eyes which become the blotches on the playful page. I must decidedly differ from those who, for the common good, take to the method of magnifying an evil the better to eradicate it. I believe that under cultivation the weeds succumb.

HE Secretary of the French Asiatic

THSSE

Society, the late Jules Mohl, for twentyseven years prepared annually a report detailing and estimating what had been done throughout the world during the year in the study of Oriental subjects. It would not be a bad plan for us to have reporters, who should at these meetings give an account of what has been accomplished in the various departments of our profession. At any rate what I have to present to you now is not so much a paper as a report on three efforts in the section of shelfclassification, made by myself, Mr. Arnold, and Mr. Warren.

The two problems that we all have to deal with are: (1) to make an arrangement that will suit the needs of our particular kind of readers, whatever it may be; and (2) to express the classification in such a sign language that we shall be able to preserve it, and keep all the books of a class together while receiving the additions by which our libraries are growing from beginnings of a few thousand volumes to storehouses of it may be a million. In a paper read at the Boston Conference1 I explained the new notation for arranging authors alfabetically in each sub-section (so as to allow unlimited intercalation of new books without

1 Library journal, 4: 234-243.

disturbing the arrangement), the new class notation in which both numbers and letters are employed (so that the largest possible number of classes may be marked with the fewest characters), and some of the main features of the classification. This we have begun to apply at the Athenæum ; the plan for Literature has been finished, and History and Geografy nearly so. Some details have been improved, and some difficulties overcome. The more I work over the scheme the harder it seems to me to make a satisfactory classification, and the more I am convinced that even an imperfect classification is better than none, and that a notation which admits of indefinite extension is better than one which compels the arrangement of the books to be torn to pieces every fifteen or twenty years and built up all over again,

a wasteful and disheartening practice. The idea of publishing the classification in form similar to Mr. Dui's Amherst system has been given up for the present, as I prefer to subject my ideas to the test of actual use before fixing them by type. Copies of the parts which are complete have been made for the use of several other libraries.

Having myself discarded the "fixed location"- that is location by shelves—and adopted the "movable "— that is location by subjects I was pleased to see that Harvard University

Library had found the advantages of the latter so great as to adopt it in regard to one class of books, the publications of the British Record Commission, and of the Master of the Rolls, and of societies like the Camden Society, Early English Text Society, etc. "The advantages," Mr. Arnold says in his report," are that an indefinite number of new volumes can be inserted at any point, and that if the collection should outgrow its present position it can be removed without any change of the numbering." Now, as these advantages do not apply one whit more to these few publications than to all the other classes of the books, it is difficult to see why the movable location should not be used throughout the whole library, unless there is some great counterbalancing advantage in the fixed. I have used the latter for twenty years in three different libraries, and I see no such superior merit in it; and I am glad to see that our national library, which no doubt is destined to remain the greatest in the country, and which we all hope is going to be in its new building the exemplar of all that is best in library architecture and library management, I am glad to see that this library is not hampered by any traditions of the antiquated unexpansive system, but is arranged solely by subjects; and I hope that for its own sake it will adopt some method of noting the exact places of the books in their several classes, which a very short stay there this week showed me would greatly facilitate the work of all connected with the library.

The Boston Public Library, I was told the other day by one of its officers, is experiencing considerable embarrassment from its fixed location. Some of the classes are crammed full; in others there is room. But books cannot be drafted from the overcrowded districts into the less populous, because that would confuse the classification. A little relief has been obtained by carrying off certain sets not much used into upper rooms, etc., and some others, chiefly bibliografical, into the working-rooms of the library. But this expedient has its drawbacks. In a printed catalog the shelf-marks cannot be changed. The boy sent for one of these books comes where it ought to be, and it is not there. In its place he finds a dummy, a

thin block of wood bearing the shelf-mark of the missing volumes, and a direction to their present habitation. So the unfortunate "runner" (as those boys are called), who has already, perhaps, come some distance to get his book, has to go another journey in search of it, while the reader waits patiently. In a large building each journey may be long; and here, no doubt, is the origin of some of those half-hours of waiting, into which the public magnify their four or five minutes of actual delay. Then, again, I am told a dummy is more easily misplaced than a book, and the absence of a dummy, tho not an insuperable obstacle, is a serious one.

I wish now to call your attention to a new classification, whose author, I was sorry to find, could not be here himself to explain it. At Harvard University Library, Mr. G: F. Arnold, who has charge of the book arrangement, has introduced a method, not, indeed, new in its theory, but which, so far as I know, has never been carried out in just the same way in its details. If the plan is not new, the execution is original. The theory is the one followed at many college libraries, of dividing the books according to the professorial departments, rather than according to any formal, preconceived system of human knowledge. So far as access to the shelves is allowed at all in a college library, it is allowed to the professors and to students working under them. It is, therefore, entirely in accordance with the fitness of things that the books, for example, which the professor of the classics and his pupils will want to use together more than any other professor or his pupils, should be put in the classical alcove. Part of these books, according to the systems, would go somewhere else. No matter; in this library they are not wanted so much somewhere else as they are here; we will put them here. So with the professor of history, of political economy, of law, of art, of mathematics. Now, it goes without saying that this plan will not do in a library where there are no professors, nor anything corresponding to them; where the general public is to be suited, with its multifarious, continually intercrossing demands. For a special library, special arrangements, and, therefore, for a college library,

which is, properly, a collection of special libraries; but, for a general library, a general sys

tem.

You may ask, indeed, why a general library, in which (as a rule) the public have no access to the shelves, should have any arrangement whatever. If all books are to be asked for on call-slips made out from the catalog, and are to be brought by runners, who know nothing of books but their numbers, why waste the brain and time of the librarian in devising, and the brain and time of him and all his assistants in applying, a system of classing books which, for all practical purposes, must be a dead letter? I answer that, in the first place, there are a number of general libraries, like my own, where the patrons themselves go to the shelves and hunt up their own books. In such libraries, the importance of a good, and, especially, of any easily explained, system is incalculable. No catalog, however well made, can compare, for educational power, with the sight of the books themselves; or, for convenience, with a thuro and minute shelf-arrangement. In the second place, there is no library so exclusive that it does not permit some favored persons to go to the shelves. In the third place, the librarian needs it daily, many times a day, to assist him in recommending books to his readers. Mr. Green has shown us, in several papers, at successive conventions, what a librarian can do to double the consultation, and quadruple its value, by judiciously pointing out to people the best books of reference, and showing them how to use such aids. Would Mr. Green's work have been possible if his reference-library had been entirely unarranged, as his circulating department is? Perhaps he can carry his library in his head, — as do many custodians of a reasonable number of books, and so has no absolute need of classification. Very well; but suppose Mr. Green should be offered the librarianship of the great public library of Tokio, and, fired with the idea of introducing among the Japanese his theories of the proper relation of the librarian to his public, should desert Worcester for a new field, how soon would his successor get a working knowledge of a collection of 20,000 unarranged books? Subject arrangement is worth making merely for the

assistance it gives the librarian in the scientific working of his library.

ence.

But I have been somewhat diverted from Mr. Arnold. I was going to speak of his sub-arrangement. By Mr. Dui's Amherst scheme in each subdivision the books are put upon the shelves as they happen to come into the library, the first numbered I and the second 2, and so on. In mine, the final order under each subdivision is alfabetical, and a carefully devised notation was prepared to allow of new books being added, like cards to a card-catalog, without disturbing this order. But Mr. Arnold's prevailing arrangement is chronological, and rightly. I want a library fitted for ready referHe wants a library fitted for deliberate and careful study. English literature, for instance, I have divided into Poetry, Drama, Fiction, Miscellanies, Collected Works, and a dozen other less important classes; and under each I follow out the alfabetical plan, so that one can put his hand in an instant on Tennyson's poems, or Sheridan's plays, or George Eliot's novels, or Macaulay's essays, or Hobbes' works; and this is done because that plan will bring most convenience to my readers. But Mr. Arnold's readers are different. They want to study the whole Elizabethan age, or the writers of Q. Anne's time, and so he groups together all the literature of the Elizabethan age, poets, drama, miscellanies; and then the Q. Anne period, literature of all kinds; and then the Georgian era; and then the reign of Victoria. So Science, when he comes to it, he will undoubtedly throw into groups, and separate the early chemistry from the chemistry of Lavoissier and his contemporaries, and their works from what is now styled the " new Chemistry.” Not that Mr. Arnold despises the alfabet! By no means! When there is no reason for any other arrangement he adopts that. The history of Massachusetts towns he will undoubtedly arrange alfabetically by the names of towns. Nor do I despise chronology. On the contrary, in the History of England I have a separate class-mark for every reign. And even in Literature I intend to divide the belles lettres of each country into two parts: one, the modern, going back as far as one can read with ease and comfort, say in English to Shak

speare; and the other, that earlier literature, which a man not specially practised reads with difficulty and a glossary. In French I should divide just after Montaigne; in German at about the same date. So much division is all that the general reader needs; and so much occasions him no discomfort. The users of early English are a distinct and small class. It is an aid and no trouble to them to have the early authors put by themselves. Ordinary readers are little affected one way or the other by the segregation, and both know in most cases whether any author they are in search of is before or after Shakspere. Only a few books on the borders. give us pause. On the other hand, for the impatient man of business to have to cudgel his memory to determine whether the author he wants to refer to in a hurry is an Elizabethan, or a Georgian, or a Victorian, is felt by him to be an intolerable check; and, what is of equal importance, it is very hard to explain to him just how, why, and where you draw the lines. For him the more simplicity the better. The problem is to estimate the forces at work in different directions, some pulling towards minute classification, some towards larger division, some towards the alfabetical point of the compass, some towards the chronological, so as to determine accurately where in any given library is the point of equilibrium.

There is an example of this in a question which I had to decide the other day, where to put Fairy stories, Legends, Imaginary voyages. I had provided a place for them in my classification; but the books themselves were with the Fiction; had been so in the Athenæum from time immemorial. Should they stay mixed in with the novels in one alfabet, for the sake of ready reference, or should they each form a small collection by themselves? If you mix them you have merely to know the name of the author or collector of a volume of Fairy stories or Legends, and you find the book at once in the alfabet of novels: Grimm, for instance, between Gerald Griffin and Mme. Guizot; and Andersen's Fairy tales between Ames and Arblay. You do not have to stop and think whether the book you want is a colection of Legends or a collection of Fairy

Stories, as you do if the three classes are separated. On the other hand, if you have forgotten the collector's name it will take you a long time to pick out the comparatively few Fairy Stories among the hosts of novels, and find the one you want; and so it will if you are desirous of seeing at once all the legendary collections which the library possesses. Of course if there is a good classed catalog the difficulty is very much diminished; but the shelf classification by itself is evidently in this case insufficient. I do not undertake to say what would be generally best; but for our library I decided that the ready reference obtained by having the Fairy stories all by themselves was greater than the convenience of never having to think whether a given book was a novel or a fairy story, because in a majority of cases the readers would know well enough in which category to look, and in the few doubtful cases the worst that could happen would be to have to look in two places. Moreover, the shock to the classificatory sense of having an indiscriminate mixture of such widely different kinds of fiction" is worthy of some consideration.

In comparing these two methods I am not saying that one arrangement or the other is the best abstractly. You cannot say that any arrangement is the best. Everything depends on the end which you wish to reach. In fact, I have to repeat in regard to shelf arrangement what was said in my essay on cataloging in regard to classed catalog vs. dictionary catalog: the one is better for the thorough and leisurely scholar, the other for the hurried man of business. In Harvard College Library Mr. Arnold's elaborate arrangement very properly accompanies Mr. Abbot's thorough and minute catalog. When both are finished I have no doubt both will be found admirably suited to the atmosphere of the place, and to the needs of the students. In my own library, where I must provide not so much for the wants of those who are pursuing a connected course of study as for the desultory reader, more simple arrangements, which demand less previous knowledge on his part, must be sought for.

One other work has been done this year which falls under the scope of my report. I