Three elements enter into the construction of any building: (1) Fitness; (2) permanence; (3) beauty. In library construction that is the order of their relative importance. The ques tion of fitness is essentially the concern of the librarian. Here he should he supreme, for un less the architect has a correct knowledge of the problems of library administration he is not competent to decide questions of arrange ment, etc., that so intimately affect the opera tions of the library when it is completed. The architect does not have to work in the building after completion, but the librarian does. This doe,s not, of course, mean that the librarian shall meddle with purely constructional details hut it does mean that either he or the trustees should determine the policy or program of con struction; the size of the building, its cost, ma terials, arrangement of rooms, etc., etc. The librarian and the architect should work together, and in order to do so most successfully the sphere of each should he so delimited that no possibility of misunderstanding can arise. In order to attain this happy result, the librarian or building committee should prepare a state ment setting forth the characteristics of the proposed building so clearly and fully that the architect can have no excuse for presenting a plan at variance with the proposals. This, in effect, will constitute an order on the architect and its nature will be determined by data pos sessed only by the librarian or building com mittee; i.e., the amount of appropriation, the nature and purpose of the library, the site, etc. The question of permanency is next in order of importance and involves not only the type of material to be used, such as stone, terra cotta, brick, concrete, etc., but also the prob lem of providing for future expansion. Failure in the latter respect has been the cause of more distress to librarians than anything else. Ap parently a library does not grow by arithmetical progression, but by geometrical, and as soon as an adequate home is constructed it attracts ma terial from undreamed-of sources and in a steadily increasing volume. Hence any esti mates of future development are generally much below the mark. A notable example of this was the case of the Library of Congress building. When planned, it was anticipated that it would accommodate the additions for several decades. This proved so mistaken that a new stack providing for a million volumes was under construction some 10 years after the building was first occupied. It is clear, there fore, that the growth of a library cannot be foreseen with any great accuracy. That it will be greater than any apparently liberal estimate may be taken for granted. Building for the fu ture, therefore, does not mean a structure vastly larger than the needs of the present and immediate future, but it does mean that no building shall be erected that cannot be en larged without injury to its design or without needless expenditures for demolitions and modifications. In short the additions should be considered in the original plans. With this in view, it is evident that a library building should not have ornamental façades on all sides, ex pensive to construct, and a complete loss if ex tensions arc made. The question of material should 1.c determined by the appropriation, by local conditions, etc. It is needless to say that it should he as nearly fireproof as possible. So called absolute fireproof construction costs one-third more )hall a itlathe 1ifcllrooi con struction and ordinarily the latter will present a sufficient factor of safety. Stone, of course, is favored for the greater library buildings, but is the most expensive type of material. Excel lent effects have been attained with much sav ing by the use of terra-cotta, brick and con crete. Frame construction, on account of its great inflammability, is not desirable, yet it would be better for a community to construct such a building rather than none at all. At least, it will be a nucleus for something better. The element of beauty, as far as it concerns a library, is essentially a by-product. Failure to recognize this point has been the prime cause of unfit library buildings. The designers of the public schools have learned this lesson and now library architects should recognize it Uni versally and thereby avoid experiments and sad mistakes. It is difficult to fix the responsibility for the common error that a library should be something ornate and monumental. Sometimes it is due to the desire of the building committee to supply an architectural wonder to the com munity; again it may be that the librarian either favors or fails to oppose with sufficient energy an ornate structure. Generally, how eves, it would appear that the chief sinner is the architect who has seized upon an oppor tunity to express himself in the grand style. Another fruitful cause of unfit buildings is the predilection of architects for certain styles of architecture. Thus, for instance, the follower of classic traditions will build nothing but Greek temples, although the Greek temple is un doubtedly the very worst thing wherein to house a collection of books; a lover of the Gothic will construct lofty apartments, difficult to heat and expensive to build, ignoring, more or less, the most important element of floor space. The craze for libraries built after the Romanesque pattern, which sprang up owing to the genius of Richardson, certainly resulted in an advance in library construction, but their thick walls and deficient fenestration give them a tomb-like quality thoroughly depressing to the average reader.
To sum up, the aesthetic requirements of a library are satisfied by a construction marked' by taste, dignity and repose, with no excess of decoration and ornamentation. In fact, warmth, homelikeness and restfulness are the charac teristics that good psychology would recom mend for a building dedicated to the conserva tion and distribution of knowledge. Should these principles be adhered to there can be no doubt that a purely library type of architecture will develop, if it has not already done so. Like the modern skyscraper, the library will find a type of construction all its own, sufficiently lovely and none-the-less a library. To gain this end, only one thing is necessary: to plan libra ries from within, for all beautiful buildings be came such because first of all they were fit. From the viewpoint of interior efficiency the following are the minimum requirements: (1) Sufficient floor space; (2) compactness; (3) light; (4) heat and ventilation; (5) storage. 1. Floor space. The amount of floor space as signed to the various departments will neces sarily depend upon the nature of the library, but taking the public library as the norm, the assignment of floor space would average as follows: Reading-room, delivery-room, etc., 50 per cent; administration, 20 per cent; lecture hall, 10 per cent; stack, 20 per cent. There will of course be variations, particularly with regard to the stacks, which of course ex pand more rapidly and easily than the other groups. 2. Compactness. This is closely re lated to the preceding requirement, and means that the library be planned for economical administration. What is known in military science as lines of communication)) should be adhered to, so that the various processes in handling the book will link up without loss of space or time. Every step that ilk- wasted and every needless operation caused by constructional exigencies stamp the building as faulty to that degree. All portions of the building from the basement to the attic should be available, and waste spaces formed by un necessary passages and stairways should be avoided. 3. Light. The eye is the organ used almost exclusively in a library, hence the illumi nation should be abundant in quantity and dif fused in quality. Therefore any form of con struction that decreases light, such as over hanging porticoes, massive walls or dark tones in decoration, are distinctly bad. Direct sun light and glares are bad, but otherwise there is little danger of too much light. This applies equally to artificial lighting, which should be supplied by electricity where available. In
direct systems of illumination are considered preferable, but they are not economical, and certainly are not recommended for stacks and situations requiring concentration of light. 4. Heat and Ventilation. Mental work is most efficiently performed where the temperature is neither too warm nor too cool, 68° to 70° Fahrenheit, and where there is an abundance of fresh air, without drafts. In general the direct indirect method of heating appears to be the most satisfactory, for by means of it not only is the building heated but it is also supplied with fresh air. Furthermore the space-wasting and unsightly radiators are avoided, air-ducts in convenient positions in the walls or at po sitions being substituted. There should, of course, be ventilating exits for impure air and, in addition, some libraries use forced or artificial ventilation. Such a system has its advantages, but it is expensive to install and to maintain, and in certain cases has not been wholly a success. It might, however, be used to advantage in the reading-rooms and other assembly halls. The nature of the collections and the uses to which they are put will neces sarily have a determining effect upon the type of building demanded. Thus, for instance, a special or associational library will require different arrangements from a public library, and the public library from a university library. For a technical collection, a small reference library or university seminar collection, the alcove system has very decided advantages, which disappear, however, as soon as the col lection exceeds, say, the 50,000-mark. These points apply, though with less force, to theo logical, medical and law libraries. The lat ter, in particular, requires special treatment, for a law library is a reference library in the strictest sense, not partially but as a whole. The entire collection, sometimes very exten sive, should be immediately accessible. In many cases, such as that of the Social Law Library, Boston, the problem has been solved by placing the stacks down one side of a long apartment, the other side of the room being arranged for readers, who have direct access to the books on the shelves. A similar arrange ment would be satisfactory for the larger medical, technical and scientific libraries. Theo logical, historical and document collections, par taking more of the characteristics of general libraries, could make use of stack system. Gen eral reference libraries are subject to the con ditions determining any library building, yet there are certain differences that require special attention. The university library is perhaps the best type of this class. This has some vee special problems that influence the construc tion of the building. One of these is the neces sity for making it conform with the other build ings in the academic group, for more and more the universities are planning all new buildings according to predetermined architectural styles. The necessity for conformity on the part of the library could conceivably result in an un satisfactory building, unless great care were taken in its design. Another serious problem of the university library is the necessity for providing for the special collections demanded by the different academic departments. In some institutions the university library is little more than a collection of separate special li braries, the general library containing only a minority of the total collections. However, the difficulties of administration and oversight, the danger of losses and changes of position, more than offset the advantages of this system and has led to its general abandonment. A compromise has been adopted either in the nature of a rigid restriction of the books in a department's collection to its particular sub ject, or by providing seminar or special rooms for the various departments in the central li brary, where they can be administered by the staff. In some cases there is direct access to the stacks and the main body of the literature from the seminary room. This system, in the main, follows the principles advocated by Dr. Poole and followed out by him in the New berry Library. Practically all of the university libraries have adopted it, notably Harvard, Johns Hopkins, Chicago, Illinois, Michigan and California. In most cases provision is made for research work in the stacks, table alcoves being installed for the use of professors and advanced students. This policy of course requires the construction of wider aisles along the stack ranges. In general no dis tinction is made in the reading .and refer ence rooms between the serious student and the casual reader or those who make use of these rooms as study halls between classes. The last group presents a very serious problem in large university libraries and in certain buildings an effort is made to solve it by providing study halls. The constnictional prob lems of a large public library are infinite, but as such a library, owing to its size and the variety of its interests, tends to split up into very distinct departments, these problems can best be solved by considering each department as a unit. By determining its particular needs first, and then defining its relations to other departments and co-ordinating the whole, a satisfactory basis for a plan could be attained. With regard to the site a main public library should be situated as near to the centre of population as possible without establishing it on a traffic artery, with the accompanying hi conveniences of noise, dust, etc. The building should he of durable material, preferably stone or brick; and as nearly fireproof as possible. As great public libraries are both reference and circulating libraries the plans should provide for a separation of these activities. This prob lem has been most admirably solved in the plans of the New York Public Library. It appears to be traditional that a public library should possess a large auditorium. This is more or less of a luxury and as it is only occasionally used it would seem wiser to hire a hall when lectures are offered by the library, rather than carry so much unused space. The suggestions already presented regarding light, ventilation and provision for expansion apply with particular force to the public library. In the larger cities the central library cannot keep in touch with the whole community. For this reason branch library systems have been estab lished. Special buildings have been erected or rooms acquired. The constructional problems of these, however, are those of the small li brary, the only differences being extreme com pactness permitting a small administrative force, absence of cataloguing rooms, etc.; small stacks, preferably open to the public; and pro visions for the reception and return of hooks forwarded by the parent library. The stack question is, of course, a most vital one in li brary planning, yet it is one that has been fairly well answered by American constructive genius. The modern stack to be found in the libraries of our country constitute a most adequate adap tation of construction to the demands of the situation. There is nothing to compare with it in foreign countries. One step is yet to be taken, which is for the architect to frankly accept the stack and, instead of hiding it away behind screens of walls, often unnecessary, to concern himself with its architectural pos sibilities. The admirable results obtained in the Denver Public Library and in some smaller li brary buildings point the way to a new and most interesting field of development in library construction.
Bibliography.— Burgoyne, 'Library Con struction' (The Library Series, II, London 1905) ; Champneys, 'Public Libraries; a Treatise on Their Design, Construction and Fitting' (London 1907) ; Koch, 'A Book of Carnegie Libraries' (New York 1917) ; Snead and Company, 'Library Planning, Bookstacks and Shelving' (Jersey City 1915) ; Soule, 'How to Plan a Library Building' (Boston 1912) ; id., 'Points of Agreement among Librarians as to Library Architecture' (Brochure Series, architectural illustrations, 1897).