must introduce it by referring to one of the papers read at the Manchester meeting of the English Library Association. This paper urged a somewhat novel method of arrange

ment.

-an ex

Starting with the ordinary objections to an elaborate filosofical scheme, which looks very well on paper, but does not fit your books, the writer advocates not having any scheme at all at first, but putting each book as it comes in wherever it seems at the time to go best, and so letting an order grow up; an order which would be the one of all others best fitted for the particular library in which it originated, act fit, like the covering of a shell-fish, he might say, or skin of a man; whereas the ordinary premeditated system would be like a suit of ready-made clothing. Or he might claim for his way the superiority of the English constitution, developed through the ages as the exigencies of each generation required, over the French constitutions, carefully and fully drawn up, with an immense show of completeness in one generation, only to be found unsatisfactory and pushed aside by the next. But it seems to me that there is a via media, which is a better way than either. Do not make your scheme out of your own inner consciousness merely; that is unpractical. Do not decide any individual case on its own merits as it comes up, with no general ideas to go upon; for the result of that will be inconsistency, contradiction, confusion. That is equally unpractical. Make your plan beforehand, but make it from books. A scheme suggested by and made so as to contain all the books now in the world would not require much stretching to embrace those that shall hereafter be written. Of course no one can actually get at all these books themselves; but catalogs will supply their place. If you wish to prepare a frame for art literature, examine carefully the largest art library you can find, and make a skeleton arrangement that will suit that; then consult all the art catalogs you can lay your hands on, and see if you have omitted anything in the first sketch; then consider the subject itself, the relations of its parts, and the possibilities of future discussion. The scheme you will form in this way will be far superior to the hap-haz

ard order which our English friend suggests. It is true that the covering of the mollusk and the skeleton of the vertebrate grows with its growth, and is altered by its environment and the accidents of its life; but for all that it grows in accordance with a prearranged plan. Now this use of large special libraries to found a classification on is exactly what has been done in the third work to which I desire to call your notice.

The Bureau of Education has a library of over 20,000 volumes and pamflets on educational subjects. This collection has been classed by Mr. S: R. Warren, whom we all know as the editor and part author of the famous report of the Bureau on Libraries; and he has drawn up a systematic list of the divisions, which I understand is to be presented to us today, and afterwards printed by government. I shall say nothing about it, for in truth I have not had time to examine it; but I wish to point out the peculiar value to the general classifier of these full special schemes. It is true, in adapting them to a general library we have to make some changes, because for the purposes of such a collection various books are brought in, which for general use belong in other departments. But this is a trifle compared with the advantage of getting a conspectus of the subject at once comprehensive and minute.

And now I have only to urge upon any library that is about to rearrange, or is thinking of going into a new building, not to tie down the volumes to a particular shelf of a particular alcove, bringing books into an incongruous relation with the accidents of architecture; but to designate their location in terms taken from their own nature, to mark them, that is, as belonging to a particular branch of this or that subject. The first method must be at best temporary; the second is permanent.

Of course anything that is human is liable to error, and even in this movable location there may be need occasionally of change, to correct mistakes in placing individual works, and sometimes to improve a detail of the classification, or to provide for some new development of science. But in its main outlines, and in the

far greater part of its details, the movable location never needs any change; while the fixed location inevitably requires, sooner or later, unfixing and refixing.

Buildings become too small, become antiquated, decay, are abandoned; but geografy

does not become history; the natural sciences are not metamorphosed into the social sciences; mathematics will never be theology; fiction remains fiction, the drama the drama, poetry poetry, as long as literature and libraries last.

THE

CHE subject of library architecture came up for consideration at the last meeting of this Association in Boston by having our attention directed to the construction of larger buildings than we have had experience with, of which several will be built in this country during the next five years. There was no time for a thorough discussion, and it was, by common consent, agreed that the more deliberate consideration of the subject should be resumed at the Washington meeting.

In the course of my remarks on that occasion, in which I made some suggestions as to the construction of this class of buildings, I said: "I know of no better rule to be observed in the library architecture of the future than this: Avoid everything that pertains to the plan and construction of the conventional American library building.'' My present purpose is to explain and illustrate what I then could treat only in outline, and do some construction on my own account. I am convinced that the conventional style of library architecture is very faulty, and that we shall never have a general reform until better principles are applied to the construction of the largest buildings. The smaller libraries are constantly copying and perpetuating the confessed faults and worst features of the large libraries.

By the "conventional American library building" I mean the style of which the Boston Public Library, Boston Athenæum, Astor Library, Cincinnati Public Library, Baltimore Peabody Institute, Congress Library, and others which I might mention, are the representative types. All these buildings have lofty rooms, and a large, open space surrounded with

alcoves and galleries which are used for the storage of books. Although these buildings have a variety of detail in other respects, this is the conventional style of which I speak. I might illustrate what I have to say by exhibiting the interior view of any one of them. I have selected, however, for this purpose, the latest, the best, and the most carefully planned of all these buildings, that of the Peabody Institute, of Baltimore. Here some of the objectionable features of the older buildings have been avoided, and useful appliances and devices have been introduced. It is, however, with the general plan we are now concerned.

The main library hall, of which I show you a ground plan and an interior view, is 84 feet long, 70 feet wide, and 61 feet high. On the front is the reading-room, 72 × 36 feet, and in the rear a work-room, 384 × 20 feet, and the librarian's room, 15 X 20 feet. The alcoves are six stories high; they project 18 feet from the walls, and there is a passage-way two feet wide next to the wall, for access between the alcoves, which are 12 feet apart. A skylight in the roof and two small windows in each alcove furnish ample light. The present shelving capacity of the room is 150,000 volumes. It is certainly a stately and imposing structure; and if we will banish from the mind all consideration of convenience, utility, and economy, and regard its architecture simply as an æsthetic recreation, we may pronounce the picture before us beautiful. It is the nave and aisles of a Gothic church of the Middle Ages, with the classic associations of five centuries about it, brought down to the practical uses of a modern library structure.

There are some objections to this venerable and conventional arrangement, and I will

mention:

1. The wastefulness of space in this central portion of the building. Books are shelved only in the aisles; the nave is empty, and serves no other purpose than contributing to the architectural effect. Is not this an expensive luxury? Here is a solid block of vacuity, 84 feet long, 34 feet wide, and 61 feet high,more than half the capacity of the room, which can be applied to no possible use in the storage of books. The floor can be used, and is used, in most of the libraries of this class, as a reading-room, and as a general promenade for tramps and sight-seers. It is unfit, however, as we shall presently see, for a readingroom; and the trustees of the Peabody Institute have had the good sense to provide another and suitable room for this purpose. The storage of books, therefore, is the only practical use to which this room is applied, and half its capacity is wasted in order to secure architectural effect. 2. The second objection I will mention is the difficulty and expense of heating such a room as this. In our northern climate fires are kept for six or seven months of the year, and, for four of these months, large fires. Hot air from a register or radiator rises to the ceiling like a balloon, and the upper strata become intensely heated before the lower stratum, in which we live, has a comfortable temperature. This arrangement is a wasteful expenditure of heat. In the Cincinnati Public Library the unequal distribution of heat is partially obviated by warming the marble floor, by means of steampipes beneath the floor, and drawing off the heated air of the upper galleries by ventilation, or cooling it in the lantern of the roof, which in winter serves as a refrigerator. This is done, however, at an enormous expense for fuel. The librarian informs me that 500 tons of coal are consumed in the library furnaces in an average season. He has sent to me tests of the temperature in different parts of the library which he made on December 29, when the thermometer outside indicated 3° below zero, and also on the evening of January 4, when 120 gas-lights were burning, which indicate that the temperature on both occasions

was fairly equalized. Four years ago a friend of mine visited this library, and, observing the intense heat in the upper galleries, procured a thermometer and ascertained the temperature near the floor and in the upper gallery. Six feet from the floor it was 65°; and in the upper gallery, 124°. Mr. Dyer, librarian of the Mercantile Library of St. Louis, writes to me, under date of Feb. 3, 1881, that the temperature of his library hall on that day, one foot above the floor, was 64°; 10 feet above, 74°; 19 feet above, 820,- indicating that the increase of heat was about one degree for every foot of elevation. He adds that, during the summer, the mercury, two feet below the ceiling, frequently reaches 140°.

3. I object, in the third place, to the shelving of books in galleries under any circumstances, and especially in this instance, where the alcoves are piled one upon another, six stories high. I may group my objections under three heads :

(a.) Because galleries are a wasteful expenditure of the physical strength of attendants in climbing stairs, and of the time of readers in waiting for their books.

(b.) Because the bindings of books in galleries perish from heat, and the higher the books are above the floor the more active is this destructive agency. Leather is an animal tissue, and will not, like linen, cotton, paper, and other vegetable substances, sustain, without injury, a higher temperature than we find agreeable to ourselves. Books cannot live where men cannot live. They are more nearly allied to us as cogeners than we are wont to suppose. In excessive heat the leather of bindings slowly consumes, and its life departs. If we put our friends in torment, they prove to us the doctrine of annihilation. Bindings perish from other causes, one of which is the presence of sulphuric acid in the leather. This acid is used in a process of the manufacture called "clearing," and, from haste or negligence, is not thoroughly extracted before the leather is finished. The sulphurous residuum of gas combustion is also said to be injurious to bindings. The burning of gas, I have no doubt, is very injurious to bindings in libraries of this construction, and chiefly because it raises the temperature in the galleries.

In libraries bindings have no such aggressive and destructive an enemy as excessive heat. All the large libraries in this country and in Europe are lamenting its ravages, and often without a suspicion of the real cause of the deterioration. A well-known architect of Boston recently called upon me, and, conversing upon this subject, which was new to him, said that he frequently went into the galleries of the Boston Athenæum to consult books, and when he came down found his clothes covered with a fine red powder. He asked if I knew what that powder was. I replied that I had often observed the same fact in the same locality, and I had no doubt that it was the ashes of the bindings which had been consumed by excessive heat.

-

Books should, therefore, be shelved in the coolest part of the room, where the air is never likely to be overheated, — which is near the floor, where we ourselves live and move. In the private libraries of our residences a mistake is often made in carrying the shelving of our bookcases so high that they enter the upper and overheated stratum of air. If any one be skeptical on this point, let him test, by means of a step-ladder, the condition of the air near the ceiling of his common sitting-room, on a winter evening when the gas is burning freely. The heat is simply insufferable.

(c.) Besides the reasons already given, I object to the shelving of books in galleries, because it is unnecessary. The 150,000 volumes, the present capacity of the Peabody Institute hall, can all be shelved near the floor, where convenience in reaching them and their preservation require them to be. In order to exhibit this fact to the eye I ask your attention to this scale-drawing of the floor, with the bookcases so inserted. The folios and quartos will be shelved in wall cases extending around the room, and the royal octavos and smaller volumes in double cases, open on both sides, three feet apart, the side alleys being three and a half feet wide, and the central alleys four feet wide. Instead of having two alleys four feet wide, the better arrangement for this room would be to have a central alley five feet wide, which would give direct communication with the reading-room from the work-room and

librarian's room. The cases will not be so high but that a person of average stature can reach any book without step or ladder.

The rule for estimating the shelving capacity of any room of considerable size, arranged in this manner, is to allow 25 volumes for each square foot of flooring. In this instance the capacity is 27 volumes per square foot, because the cases are longer than they are usually made. The shelving capacity of these cases is 160,050 volumes.

As I am to use further on, in some construction of my own, the estimate that each square foot of flooring will shelve 25 volumes, I will here explain how it is obtained. The double cases are 18 inches wide, and of any desired length, say 16 feet. The space which one case will require is a rectangle, of which the longer side is the length of the case plus the width of the alley (usually four feet), or 20 feet. The shorter side is the width of the case (18 inches), plus the distance between the cases (three feet), or 4 feet. One case, therefore, requires 90 square feet of flooring. The area of shelving on one side of the case is 16 X 7, or 120 square feet; on both sides, 240. The conservative rule which is usually adopted for estimating shelving capacity for books of all sizes which go to make up a general library is ten volumes for each square foot of front area. The capacity of the case requiring 90 square feet of flooring is, therefore, 2,400 volumes; and one square foot will shelve 26.6 volumes. Twenty-five volumes, therefore, to the square foot is a reasonable estimate.

By reducing the width of the alleys from 4 to 3 feet, and the distance between the cases from 3 to 24 feet (in the stack-room of Harvard College library the distance is 2 feet 4 inches), the shelving capacity could be considerably increased. I have preferred to allow liberal spaces between the cases, and not to force the principle of contraction to its utmost limit. The estimate of 25 volumes to each square foot applied to large rooms brings out such enormous results as to be almost incredible.

We have now, in theory at least, shelved all the books which these six tiers of alcoves will contain upon the floor, and have space for 10,000 volumes more. We have, also, over

head, 61 feet of air and light, which is more than we need. Sixteen feet is better than 60, for it is enough. Three other floors, each of the same capacity, the rooms being 15 feet high in the clear, would fill the 61 feet and about 6 feet more. In the three upper stories the space which on the lower floor is appropriated to the work-room and the librarian's room could be used for bookcases, and would shelve 76,800 volumes. The entire storage capacity of the building would therefore be 717,000 volumes. This arrangement, when the library comes to need so much shelving space, would allow of a classification of its books into four grand divisions or departments of knowledge, each one of which would have a floor and readingroom to itself. The reader, then, by means of a modern elevator, would go directly to the floor on which the books in his own range of study are stored.

4. Returning to my general series of objections to the conventional style of library architecture, I mention, in the fourth place, the difficulty of getting about from one part of the library to another. Not to speak further of the burden of climbing stairs, it is necessary, in order to move from one gallery to another on the opposite side, to travel on the outer edge of a parallelogram, when the economies of locomotion require that we move as nearly as we can in straight lines, and from the centre outwards. Observe the ease with which any case of books can be reached on this floor plan, and the difficulty in the conventional plan of passing from a lower alcove to one in a remote corner of the upper gallery. In a popular circulating library it is positively cruel to send attendants for books with such an arrangement for shelving as this; and to station them in the overheated and stifling air of galleries to answer calls for books is even more inhuman.

5. I object, in the fifth place, to this plan of construction, on account of its insecurity from fire. In an interior finished with wood, no arrangement could be more skilfully devised for favoring the destructive operations of fire than a series of alcoves piled one upon the other six stories high, with every facility for draft — unless it be a pile of empty packingcases. When a building of this kind takes fire

the work of the insurance adjuster is very simple, for it is a total loss of the whole library. Water, heat, and smoke are as fatal to books as fire itself. Congress Library has twice been burned; Harvard College library once; so also the Chicago Historical Society's library, in what was thought to be a fire-proof building; and the Birmingham Free Library, which several of us visited little more than three years ago, has since, with its great Shaksperian and Cervantes collections, been burned with fire, and nothing of its more valuable treasures saved. The class of library buildings which we are now considering will contain books, manuscripts, and public records of inestimable value which money cannot replace. To lose one of these libraries by fire would be a national calamity. After all that may be done in the way of external protection, there is still a . large risk from internal accidents.

On a summer evening, a few years ago, a fire broke out in one of the rooms of the Cincinnati Public Library after the building had been closed for the night. It was fortunately discovered and extinguished before much damage was done. The origin of the fire was at first a mystery; but it soon appeared that the painters, who had been finishing the wood-work of the room, had left their oiled rags on the ledge of one of the bookcases when they quit work at night, and they had ignited by spontaneous combustion, and had set the bookcases on fire. The Birmingham Library was set on fire in the daytime by the lamp of a careless plumber who was thawing out the gas-pipes. A fire may start in a large library at any time by accidents as unusual as these; and it were a shame if, from errors of construction, it be allowed to range through the whole building. Hence, buildings such as we are considering should be constructed in a series of fire-proof compartments, in order that the fire may be confined within narrow limits. I am not aware that this precaution ever has been taken. The principle, however, has been applied to the great ocean steamers, and many a ship has been saved by having its hull divided into several water-tight compartments. A practical method of securing this protection will be considered later in our investigation.