PART II

BOOK STORAGE

The several rooms will be treated separately, also different methods of shelving. The phrase “book rooms” is not used herein as in England, where book store or book room means only book storage, as distinguished from staff rooms and reading rooms, but will include all kinds of shelving, whether used for book storage only or combined with handling and reading.

In an article on Book-storage by H. Woodbine in a recent number of The Library Association Record,[257] he states the factors of past development as,—

1. Economy of space. 2. Economy of cost. 3. Expansibility. 4. Adjustability. 5. Safety from fire. 6. Protection of books (from pests, dirt, damp, etc.) 7. Convenience in service.

It is well to bear all these in mind when planning any library, though I should put the last first, and add cleanliness. They would serve as comprehensive tests of all kinds of shelving, wooden or metal; wall, floor, or stack. They are such important details in library service that I will take up the different forms of shelving in considerable detail.

Shelving, Generally

General rules in shelving are: (1) No book should be above reach of hand from floor. This means about 6½ feet (less in children’s rooms) or 7½ feet to cornice, or top of top space. Don’t use steps or ladders, they are obstructive and troublesome to use.

(2) Uprights should not be more than three feet apart, to avoid sagging, and weight in handling. Somewhat less is sometimes advised, never more.

(3) All shelves should be of the same measurements and interchangeable, for obvious reasons, throughout the library. Unadvised architects are apt to fill nooks and spaces with shelving to suit. This may not be so objectionable in fixed shelving, but is fatal with movable shelves.

(4) Shelving should be movable as well as adjustable. Private libraries and very small libraries can get along for a while with fixed shelving, but when books of different sizes accumulate, and close classification is adopted, movable shelving is necessary.

(5) Edges and corners of shelves and supports should be rounded. If hands or books strike sharp edges roughly, they suffer.

(6) There should be no projections to catch clothing. Watch this, especially in stacks.

(7) In shelving or supports, do not leave projections to catch dust. This is often a fault of carved end-uprights.

(8) Have both upper and lower shelves accessible and well lighted for easy inspection. Wherever there is ample room, use of only the breast-high shelves is more convenient both for inspection and for handling.

(9) The old-fashioned ledge is not needed, except in a few instances. It unnecessarily widens the aisle above, interfering with close storage. Wide books can be stored elsewhere; and space to lay books down in handling can be provided near by.

(10) The average dimensions of shelves[258] are well settled by custom; _e.g._, _Length_ (as above), not over three feet; _Depth_, eight inches, except for special sizes of books (see later); _Thickness_; for wooden shelves, ⅞ inch finished, (1 inch stuff, planed); _Interval_, Wood or metal 10 inches (11 inches top to top of wooden shelves) for octavos and duodecimos, though one advantage of movable shelves is the possibility of variation if desired anywhere.

(11) No doors of any kind are used in modern library bookcases, except where dust is to be excluded from delicate books, or thieves are to be excluded from rare books. Doors are an impediment to use.

=Shelf-bases.= To save books in sweeping, a four-inch solid base is usually provided in all lands of shelving. In unusually high shelves, this base projects as a step, but it is unsightly thus, and just so much as it projects it narrows the aisles and promotes stumbling.

See _Fletcher_, Public Libraries.[259]

=Fixed or Movable.= As stated above, fixed shelving is somewhat cheaper and more easily made, and will serve well in very small libraries. In setting up movable shelving a row of shallow holes an inch apart is bored an inch from the front and from the rear edge of the inside uprights. To support the shelves, projecting pegs of various kinds are inserted in these holes at any desired intervals. There are several patents, the most popular one being a metallic pin with shoulder, which may be turned over for slight alteration of interval. Plain picture screw-eyes, with the eyes turned flat, are favorites in some libraries, and are cheap. Accuracy is necessary in boring the holes, and experiments are advisable as to the fit and steadiness of the pins, so that the shelves will not be liable to tip or fall.

=Wood or Metal.= In small libraries there is no need at all of metallic cases or shelving and it is absurdly wasteful to buy them too soon. Wooden shelving is cheaper, easier put up by local builders, and though it may occupy a trifle more space, is serviceable and strong enough until superimposed stories of shelving become necessary. Even two stories of wood can be easily managed. If you want more than two stories to use as a stack, you must have iron or steel. There are, of course, many advantages in metal when you have to come to it, though it is more costly. It saves a certain amount of space; it does not obstruct light or ventilation so much as thicker material; it is more fireproof; shelves are more easily moved.

Metal in stacks is universal in larger libraries in America, so is wood in small libraries. In England wood seems much more used in large libraries than with us.

Hard wood is not necessary for shelving, the cheaper kinds of soft wood will do, and are easier set. No backing is necessary in any form of book case, except as a brace, or for appearance, or against a brick or stone wall.

“Use no paint, but varnish and rub thoroughly.”—_Poole._[260]

“Few village libraries need spend money for steel shelving. It costs twice as much as oak; four or five times as much as some woods. Wooden cases are movable, steel not; with wood you can shift and add. You would not prefer steel in your home.... For libraries of less than 30,000 volumes, wood is better.”—_Eastman._[261]

In planning small buildings do not let manufacturers lead you into the expense of putting in metallic shelving or fixtures. Wood answers every need as well, and often better, and is much cheaper. Miss Marvin says,[262] “No stack should be included in a building costing under $20,000.” I should put the limit higher, and say “No metallic stack is either necessary or desirable while wooden wall shelving and floor shelving will hold the books in the library.”

=Ledges.= In the early wooden shelving for libraries, ledges, “counter ledges,” so called from their being the height of an ordinary “counter,” were considered essential. Dewey[263] says: “These have a double use. They give a greatly needed shelf on which readers may lay books for consultation or while reaching others, and for the pages in getting and putting back books.”

These ledges do not appear so much now in floor-cases or stacks. They still survive, however, in wall-shelving.

But they served serious needs in handling books and have been seriously missed since they disappeared from use. See an article on a proposed substitute in stacks, under the title “Carrel,” p. 286, later. This feature might also be used with wooden floor-cases when lighted by “true stack windows.”

=Labels=, =Pins=, see articles in Library Notes.[264]

=Head-room.= It is best not to build floor-shelving, even in low rooms, quite up to the ceiling, but to leave some room over the tops of the books on the top shelf for free ventilation. But Dewey said at the 1887 Conference, “Why not leave it out—use all space for shelving, with artificial ventilation?” This might apply to the head-room usually left at the top of stack rooms. But how about heat? And in most libraries there is no effective artificial ventilation or forced draft. And in many rooms outside the stack, it will not be necessary to shelve quite up to the roof.

=Shelves High or Low.= The rule is, as stated, 7½ feet in height. In many old libraries, and in a few newer ones, higher cases are used, in order not to waste upper space in a high room, wherever this space is not needed for ventilation or diffused light. This is very unfortunate in inspecting or handling the books. To overcome the difficulty of seeing and getting at the highest shelves, various forms of steps or step ladders, or base steps and high handles on the uprights are in use which can be investigated and adopted when occasion requires, as it never should arise in a new building. If such shelving is inherited, or must be used, it would be best to use these shelves, too high to reach by hand, for storing sets of books or magazines rarely wanted. Or a gallery can be built half way up to avoid the awkward use of ladders.

As books to be inspected are best nearly opposite the eye of a reader standing or sitting, live books would better not be stored on lower shelves in any open-access cases. These shelves nearest the floor might be used, therefore, for similar sets not often needed.

Miss Marvin[265] advises uniform height for wall-shelving all over the building.

Low bookcases, “dwarf bookcases,” both in wall-shelving or floor cases, are often used, for different reasons, especially to serve as partitions, and have not the disadvantages of cases too high. In floor-cases, the top can be used as a convenient ledge. In this form, low cases can be set anywhere on the floor without seriously obstructing light, ventilation, or supervision, and low cases can be used against the wall when high-set windows are needed to throw light further across a room.

=Unusual Shapes or Sizes of Books.= Minimos, (sizes under the ordinary duodecimos) are so unusual that they can be shelved at the ordinary intervals; and if a set or lot of such small books come together, movable shelves can be closed together, without much waste of depth (or by doubling back, with no waste).

Folios and quartos occur in all libraries, in the smallest as books of reference, like dictionaries and atlases; in larger libraries they may come anywhere. Formerly, the lower shelves in all cases were made wider, with a ledge above, but this made the aisles so much wider than was necessary for shoulder room above, that ledges are not now much used in floor-shelving or stacks. Instead, special shelving is provided not far off on each floor, and slips or dummies put on the shelves to indicate where the larger volumes ought to come in the regular classification, and where they can be found when wanted.

This special shelving is often put along the walls, but in late stacks I have found it convenient at both ends of each story. The necessary ledge can be widened without much sacrifice of space, into a shelf at table height, which can be put to many purposes, part of it at one end being cut into to give room for the stack stairs, which usually rob either books or users of more room elsewhere. In other rooms, with wooden shelving, there is almost always a convenient recess or end, where quarto and folio shelving can be put without crowding the other cases. Indeed, when designing a library building, one thing to watch for is, where such shelving can be stowed away near at hand, with the most economy of space. In floor-cases, wooden or metal, occasional large books can be laid across two adjoining shelves.

As to dimensions, Mr. Poole’s recommendations in 1876[266] still hold good: a ledge about 34 inches high, with two shelves below, 18 and 16 inches high for folios, 16 inches deep, and as many shelves as the case will allow above, 12 inches high and 10½ inches deep. Burgoyne says,[267] 21 inches high for folios, 13 high for quartos. These are extreme. Dewey recommends 12 × 10 inches for quartos; for folios just double octavo measurement; large folios to be laid on their sides.[268]

If movable shelving is installed, it will be possible to shelve the exceptional books upright or flat, as their size and character requires.

Burgoyne[269] advises padding flat folio shelves. The British Museum uses cowhide; other libraries, canton flannel (bad) with falls.

Elephant folios will require special roller shelves.

Shelves in Reading Rooms

“The books most used should be stored around the walls of the reading-rooms.”—(_Miss Marvin._[270]) This has been a common custom, but Mr. Dana has suggested that such shelving is out of place in reading-rooms. So H. T. Hare, in 8 The Lib. Asso. Record:[271] “The placing of books around the walls wastes floor space otherwise available for readers.” In this opinion I concur,[272] for the double reason that it bars out just so many readers, and also it necessitates movement which interferes with serious reading. As to the former objection, take a room 30 × 40 with a perimeter of 140 feet, less say 10 feet for doors, 130 feet net: If this is shelved all around, the shelving with the usual ledge, and the three feet space in front of it needed for access, inspection and passing, four feet in all, will take up 456 square feet, out of a total area of 1200, nearly two-fifths. Without the wall shelving, the room would hold tables for that many more readers—the use for which it is intended. As to the latter consideration, to get at the books every attendant fetching or returning or cleaning them, every reader consulting them, has to pass before or beside or close back of some other reader who is trying to abstract himself at a desk. If stored somewhere else in floor shelving or in a stack close by, the books would not take up more space, would be more accessible, and less in the way.

If a serious reading room can open directly into an open-shelf floor of a stack, no wall-shelving will be necessary.

The second objection would, of course, not apply so much to rooms for light reading where more or less motion and noise are expected, and less serious study is usual.

=Class and Study Rooms.= Here wall-shelving for reference books permanently or class books temporarily required, and sometimes floor shelving also, or a combination of wall-shelving with occasional projecting cases, like shallow alcoves, opposite good light, will be required. The purpose of each room defines its needs in arrangement and shelving, as also in staff-rooms and all special rooms. In libraries of sufficient size, each such room should have telephone connection with the staff, and if possible separate lifts or corridor railway service.

Wall-Shelving

The earliest book storage was in cupboards or alcoves, the latest is in floor cases, but the persistent form between and even now is that of shelving around the walls of rooms. Mr. Dana and I object to it around reading rooms, but it now prevails, and perhaps it will still prevail even there. Certainly it will always be serviceable in most of the rooms of a small or large library. It was formerly continued even in combination with floor-cases or stacks, but it is vanishing from such book rooms to maintain its position sturdily wherever floors are not for shelves, but for tables.

In this form, the old-fashioned shelf-ledge survives, with folio or quarto shelving, or sometimes cupboards or bins below, and narrower octavo shelving above. The ledge is found serviceable in temporary examination of books and for resting them in transit.

“Every available foot of wall space should be utilized for shelving, between the windows and under the windows.”—_Marvin._[273] [But not unless light comes from the other side. See below. And where there is steam heat, the space under the windows is best for radiators.]

Wall-shelving ought always to be opposite and not next to windows, because direct light in the eyes blinds the reader so that he cannot distinguish the books. But if light comes from both sides of the room, both sides can have wall cases.

=Closed Cases.= In private libraries and in some rare book collections in public libraries, bookcases have locked sliding doors, either glazed or with strong wire mesh (for ventilation), too small a mesh to slip books through.

It is better to back wall-shelving with wood whenever placed against brick or stone walls, to protect the books from damp and stain.

I have known buildings where the architect put a dado of expensive wood around rooms where wall-shelving was to be put up at once or was sure to come soon. This was, of course, a willful waste, as plain sheathing, to serve as a back for the shelving, would have been far better.

Floor-Cases

Floor-cases, as we use them, first appeared apparently in Leyden about A.D. 1600.[274] Their use in America can be traced to the pressure for space in the old libraries, just before the birth of the stack, which is only floor-cases built up into stories. As the term “floor-case” is used, it covers all bookcases set out from the wall across the floors, usually in parallel rows perpendicular to the windows, but sometimes radial or irregular. The cases are always double, back to back, their dimensions in each front being just those of wall-cases. The backs are usually open for light and ventilation, but are sometimes wired or wainscoted with wood. If backs are not used in floor-cases, some bracing is needed to make them rigid. The aisles between vary in width from three feet for service to six feet for open access, though service is possible in narrower spaces than three feet, and open access, with good light, does not absolutely require six. It is recommended by the authorities that cases should not exceed fifteen feet in length. Whenever longer rows are wanted, cross aisles at about that interval should interrupt, so that an attendant or reader should not have to walk too far if he needs to get quickly to the other side of a case.

Radial Cases

“In small libraries and branches, supervision is ensured by placing floor-cases as radii of a semi-circle whose centre is the desk.”—_Bostwick._[275]

Duff-Brown[276] says that this method of shelving secures oversight and ease of working.

The advantages and disadvantages of this arrangement are well summed up by Eastman,[277] who thinks it of doubtful value.

In small libraries, when set symmetrically in a true semi-circle, radial or concentric cases certainly have a pleasing effect. The building costs more, either in semi-circular or octagonal form, than in rectangular (more in stone or brick than in wood), and there is certainly waste of space in the widening of the wedge-shaped intervals, which, however, can be partially utilized by tables or short intervening floor-cases at their widest part.

This radial shelving has invariably, I believe, been built on the rear of the building. In many lots it has occurred to me that putting it in front, or on one side toward a street, could be made an agreeable feature, and would do more than any other thing could do toward attracting passers-by, and thus “advertising” the library far more effectively than many publicity schemes recently suggested.

As to supervision, I have seen in a recent discussion the reminder that one person blocks the narrow end toward the desk, and effectively hides disorder, mutilation, or theft beyond.

Sometimes the projection from the building is rectangular, and the shelving concentric, an arrangement likely to cast shadows. In some American libraries long rows of slanting floor-cases, not true radii, point toward the desk. So good a librarian as Mr. Wellman of Springfield, has adopted this arrangement in a large rectangular room. See also the Law Library at Rochester, N. Y. But does not this arrangement block light rather than facilitate its penetration into the room to the lowest shelves? I should doubt whether the advantage in supervision would counterbalance this interference and the waste of space. Champneys[278] (an architect) thinks there may be danger of “overestimating police methods.” It seems to me that in sizeable rectangular rooms, supervised entrance and exit at the desk, with rectangular arrangement of the shelves either perpendicular to the deskline or even athwart the room, thus trusting the public, would be better.

In small libraries, as in branches, this arrangement is worth considering, but should not be adopted, it seems to me, without very careful balancing of arguments _pro_ and _con_. Economy in construction and space and difficulties in enlargement are against; many considerations of cheerfulness and usefulness are in its favor. Where the library is so small, however, that only three or four floor-cases will hold all its stock of books, these in a rectangular projection back of the desk, will give most of the effect of the radial form, rather cheaper.

Librarians who have operated both forms could give points to any one in doubt, and many floor plans, English as well as American, with many interior views, are accessible to show different arrangements.

If adopted, it seems to me that the semi-circular plan with true radii, is better than the octagonal or rectangular walls, with obliquely placed floor cases. These may be arranged for good supervision, but their slant disturbs one’s sense of symmetry. Besides, the basement beneath may be devoted to a class or lecture room, for which such a semi-circular shape gives good light and cheerful effect.

The semi-circular plan has been adopted for alcove rooms in many places, such as the Library of Parliament at Ottawa, Princeton University, and so on, but these do not have radiating cases and need not be discussed here.

Shelf Capacity

To calculate shelf capacity, it has been usual to take ten volumes to a running foot, a figure which has been verified in some libraries. But books vary in thickness in different kinds of literature, and the exigencies of growth require gaps to be left in closely-classified libraries, at the end of each subject. These facts have tended to vary estimates, which do not now agree. In “Library Rooms and Building,” I said,[279] “For these reasons, it is prudent to calculate about eight volumes to a foot for octavos and under, and still less, say five volumes to the foot, for reference books, law books, medical books, and other bulky literature.” I have seen no reason since to change these figures for estimates, though planners should bear in mind the different classes and sizes of books to be stored in each room or on each case.

The English authorities still set the average number of volumes to a linear shelf foot rather higher, eight and a half to nine and a half for lending libraries or fiction shelves. See also, “Stack Capacity.”

The Poole Plan

This seems to be the best place to allude to the scheme which Dr. Poole proposed as an alternative of the stack. As Fletcher says, the principal objection to the stack plan was as to opportunities for readers to get at the books on the shelves. To place readers and books in close contact, Dr. Poole proposed dividing a building mainly into large rooms, in each of which readers should have tables near the windows, while opposite the windows the inner portion of the room should have floor-cases filled with some special class of books. He got the chance to embody this idea in the building of the Newberry Library of Chicago. As far as I know this plan has not been adopted elsewhere as a whole, but every large library since built has included rooms arranged more or less on this plan, which is indeed the idea of the department library in a college; or special rooms, such as Art and Patents, in a public library. So far as Dr. Poole advocated his plan he furthered library efficiency and should deserve credit and remembrance.

“In the Providence Public Library, for instance, two-fifths of the books are shelved outside of the stack.”—_Foster._[280]

But the stack plan has “won out” as a system, and has established itself as a factor in modern American library building. Further changes, developments and improvements are doubtless coming, but so far as administration and architecture are concerned, the stack must be reckoned as the distinctive difference between libraries and other buildings.

See description and criticism of the Poole plan, with vindication of the stack system, in B. R. Green’s article in the Library Journal.[281]

Dr. Poole was a sturdy fighter in his day, but he was an excellent, practical librarian. If he had lived to see the stack as now improved, and had also seen its combination with the department library or special library in large buildings, I think he would have conceded the merits of the new system.

Stacks

=Generally.= These have been adopted in this country, in nearly all libraries which have got beyond the size where floor cases will serve. They come into use with us much earlier in the growth of a library than in England, where they seem not so much in favor.

The notion of the stack was first suggested by the modern revival in America, about 1850, of the floor-case system, exemplified two hundred years before in the Leyden University Library. The first modern mention of this system I can find is Winsor’s description (1876)[282] of the arrangement of his new Roxbury branch of the Boston Public Library. In his description of the floor-cases, then only floor-cases, he suggested the idea of providing for growth another story of superincumbent cases, apparently of wood, with “dumb-waiters,” and “spiral stairs.” In 1877, Winsor outlined plans for a similar shelving of several stories with iron framework and iron floors.[283] About this time (Winsor left the Boston Public Library and went to Harvard as librarian in 1877), the first metallic stack (with wooden shelves) was developed and installed in the addition to the Harvard library building. The idea seems due to Winsor, the practical embodiment of it in full stack form to the architects Ware and Van Brunt. The latter described it soon after in the Library Journal,[284] saying, “I am in part responsible for it.”

This pregnant idea, which, as developed, has done more to change library administration and library architecture than any other device, was evidently born in the brains of a librarian as a result of his thought and experiments, and developed into practicability by good architects, as all great problems of library building should be worked out. The original stack contained all essential ideas, but great improvements in details have since then been effected by librarians, architects, and constructors.

Stacks were at first stoutly opposed by many librarians. As described by Fletcher,[285] “The stack, as usually built, consists of a series of iron bookcases [_floor cases_] running from bottom to top of a high room divided at intervals of about seven feet [7½] by light [_iron_] openwork or glass floors [_decks_]. The stack undoubtedly offers the most compact storage of books with great ease of access to every part.” He then enumerates the objections to the stack, the principal of which he thinks is, “little or no provision can be made for the access of readers to the shelves, the idea of the stack being that of a place to keep the books when not in use.”

Since the first stack was installed at Harvard, remarkably serviceable even then as a new idea, some of our most inventive genius has been constantly at work in trying to perfect the advantages of the system, and overcome its acknowledged defects. Construction, ventilation, heating, lighting, communications, ease of operation, have been gradually improved, and recently Dr. Poole’s and Mr. Fletcher’s principal objection, difficulty of use by readers, has been so greatly overcome that a later chapter has been devoted to this subject. There are several good patent stacks in the market, which deserve study and a chance to submit bids in every new building project, large or small.

The best method of planning is for the librarian to calculate how many volumes he will have to provide for, and how large a stack he needs (floor area, and number of “decks”); to lay out, with the assistance of the architect, a floor plan for one story, with the number and width of gangways he wants, and a specification of stairways, lifts, folio-shelving, and other peculiarities.

It is better not to wait for working drawings and specifications for main building, or even for the stack shell (or building), but to ask for two bids for a stack of size described, one for the cheapest form and material each maker can supply, and another for the best form he would recommend, with his cheapest price for that. This alternative is suggested, because each make claims certain advantages over the other, which might overbalance a difference in price. The invitation to bid should reserve the right “to reject any bid for cause,” and the final decision should be reserved for the building committee, under recommendation of librarian and architect. The considerations for determination can be: cost, strength, lightness, compactness, adjustability, cleanliness (including lack of projections to catch dust); convenience of stairs, lifts, floors; details of heating and lighting; and pleasing design.

After the bid has been assigned, and before the makers have begun on construction, I advise calling their expert into consultation, and asking him if he can suggest any change or improvement in any point which will increase the usefulness of the stack, without increasing its cost. There is such a keen competition between stack builders, that any of them would welcome such a conference, in the hope of getting ideas from librarian or architect which might help him improve his patent.

The stack thus bid for is to be self-supporting, deriving its solidity from its own uprights, without depending in any degree on the shell, with which the architect will only cover it and protect it from the weather.

=Location.= A stack may be installed inside the building; for instance, all along the rear,[286] or side or front. A small stack is often a feature of a large department room. But generally it occupies an ell or wing of the building, of light construction, projecting from the rear, or from one side.

Where the building must face a noisy street there seems to be no reason why the stack, rather than reading rooms, should not be located there. Why could it not be designed, even if “true stack windows” would make it look like an organ front, as a distinctive architectural feature?

“The stack may be as refreshing a problem for the hard-witted architect to struggle with as he is liable to meet. It may be that the reading rooms will be within, shut off from every noise, and the stack arranged along the exterior.”—_Russell Sturgis._[287]

The reading room is now often put just over the stack, as a top-story, separated from it by a solid floor, but connected with it by service tubes, telephones and lifts. But in colleges, is it not better to use such a location for seminar rooms, and in many libraries could it not be used as part of an exhibition and special library or special study floor?

=The Stack Shell.= That is to say, the addition in which the stack is housed. As has been said, it usually projects from the rear (but sometimes from the side) of the main building, as an ell or wing. It can be of lighter, simpler and plainer construction than the rest, for it needs no other strength than is necessary to support its own walls and roof. Indeed, it has not yet been the victim of architectural ostentation. On the exterior, true stack windows usually run up and down the whole height, although they may be interrupted by cross sections at the level of the floors or decks, or rather just above them.

From recent experiments I have made in a stack, I am led to think that here, as elsewhere, top light from windows is ten times more valuable for penetration than bottom light, hence such a cross-section of wall, about a foot wide, if it has any binding power, strengthens the wall, gives space inside for heating pipes, or looks better, would not abstract any illumination from the interior. Perhaps, however, the piers do not need such binding. That is a question for the architect, and depends largely on their construction. If they are re-enforced by iron or steel T-beams, the piers need not be massive or be strengthened otherwise.

Some authorities (Champneys,[288] for instance) recommend solid floors every three decks, as guard against spread of fire, but this extra expense, not needed for support, seems to me unnecessary as protection.

The material of stacks must be iron, or better, steel, to support so much weight. The construction, indeed, is much like that of a “sky scraper,” whose steel frame stands alone, without help from the walls.

=Use by Readers.= It does not seem either possible or desirable to plan for continuous use of any space in stacks by readers. The temperature both in summer and winter is usually not so equable as in other rooms. The main object of the stack, which is book storage, is just so much frustrated by surrender of shelf space to readers. But there is much inconvenience in excluding them entirely.

It is a hindrance to investigation to have to make inquiries, or selections, through the medium of an application at a desk. A large number of serious readers want to glance at all the books bearing on the point they are investigating, often to “taste” books by dipping into them here and there; and to make choice directly from the shelves, of books they want to examine more thoroughly or copy from, to be carried to a public or private reading room and used there undisturbed at leisure. They want free access to the stack for ten minutes only at a time, but they want it badly. See Fletcher.[289]

“It is fortunate for those who have the use of a library if they can be admitted to the shelves and select their books by actual examination.”—_Cutter._[290]

For this, several devices have been used. One is to leave the space in stacks next to windows for tables and chairs, to be used by readers. “Or alcoves on one side, as in Iowa College.”—(_Marvin._[291]) A variation of this takes the form of “cubicles,” little glassed-in rooms next the windows, as in the new Harvard Law School stack, or as proposed for the Harvard University Library. But before using this form generally, it would be better to calculate, first, how much space this will abstract from the storage capacity of the stack; second, how much it affects the penetration of daylight into the stack; third, how often any one reader will want to use any one section of the library so long as to make this arrangement worth while; fifth, the expense of construction and provision of equivalent stack room elsewhere; and sixth, the problems of heating and ventilation, for readers who require reading-room conditions.

Another favorite device is to shorten the outer ends of ranges of shelves, say by one three-foot section, in every other case on every floor, where a tiny desk can be set into the range, with a chair or stool underneath for the use of a reader. This furnishes room for reading but _pro tanto_ less space for books.

=Open Access Stacks.= Can wider aisles be left in stacks so that readers may stand well back or stoop to inspect books, and pass each other easily? Yes, stack cases five feet “on centres” will allow fairly free movement, as this means 3-feet-6-inch or even 3-feet-8-inch aisles. But no such width could well be allowed as is called for with open-access floor cases, _i.e._, six feet clear between. The present methods of stack construction would not apparently lend themselves well to wide spaces on the ground floor and narrow spaces above, because the uprights would not directly support each other. A building might have, indeed, two or more different stacks, one open access for readers, the other close storage for books, but this seems rather wasteful. Is there no way to provide, in a stack which will give the maximum storage, some facility for such inspection and handling as is needed both for staff and readers?

=A Suggestion.= In reading “Clark’s Use of Books,” I came across an old expedient of mediæval days which will give a good name for the device I had already thought of. (See next section.) His quotation[292] is as follows:—

“In the north Syde, the Cloister was all fynely glazed. And in every wyndowe iii Pewes or Carrels, where every one of the old Monks had his carrell, severall by himselfe, and there studied upon there books. From one stanchell of a window to another, and in every one was a deske to lye their bookes on.” “These were devices to provide a certain amount of privacy for literary work.”[293]

=Carrels.= While thinking of this conflict between the desired use by readers and the close storage which is the proper use of a stack, I tried to find some wasted space which might serve the one use without infringing upon the other. While searching I noticed that window ledges were thus wasted. Look through Koch’s floor plans,[294] or any others, and you will notice that window frames, usually set midway between the outer and inner surfaces of the wall, were sometimes set flush with the inner surface, thus leaving outside a window “stool” nearly the full width of the wall. But why leave it outside where it would be only useful for pigeon-roosts or flower-boxes, neither strictly necessary? Why not set the window-frame flush with the outer wall and so leave the whole ledge inside, both sill and stool? In the Salem Public Library stack, as the architect saw no structural reason against it, this has been tried. In each stack window on every floor a thin shelf has been run across, table high. The setting back allows this shelf to be twelve inches deep and three feet long without projecting into the aisle, and without materially interfering with light. Set a stool near and here is provision, close to the books, and without cutting into the stack, for just as many choosers of books as there are windows on each floor. When no readers need them, here is a ledge for attendants to use in assembling or dispersing books.

This device does not suit permanent reading, for which the stack is not intended,—but why does it not perfectly meet the needs of casual inspection, and choice?

It has been gradually tried out. In the John Hay Memorial Library at Brown, rather narrow window-shelves were tried; then wider sloping desks at the Episcopal Theological School; and recently, the wider Salem carrels, where the windows are set quite flush with the exterior of the piers.

There is still an opportunity for experiment and development. Is such a shelf better, fixed or hinged? What would be the simplest form of hinging and fastening? Is it better, in view of its temporary and intermittent use, to have it at desk height, for a standee? How thin can it be, and of what wood, cheapest and least liable to splitting? Might not metal shelves, furnished with the stack, be better, and about as cheap?

As finally improved with these carrels we could bring the whole stack back to the narrowest intervals consistent with moving books, and thus avoid resort to underground stacks and sliding cases, until much later.

[Webster’s International Dictionary gives only the spelling “carol,” but the old records call it “carrell.”]

At Durham, the carrels were 2 feet 9 inches wide. At Gloucester there were twenty carrels, each 4 feet wide, 6 feet 9 inches high, and 19 inches deep.[295]

The modern Salem Public Library carrel is wider than the one at Durham, and about as high and deep as those at Gloucester Cathedral.

=Stack Details.= _Dark Interiors_ are discussed elsewhere; having the library built around a stack, to be lighted by electricity, open to daylight only by way of the roof, and opening to outer corridors or rooms on each floor. This is mainly an architectural problem, though its administrative aspects would have to be considered by the librarian.

_Height._ The height of each stack floor is generally set at seven feet to seven and a half. I favor seven and a half, of the two, so that a tall man need not stoop under the deck beams and electric bulbs. In order to get the ground floor of building and stack coterminous, the lower story of the stack must correspond with that of the building, which is not usually higher than ten feet. As it is most convenient to have the basement floors of stack and building also coterminous, the unusual height, for this case only, may be accepted, and the inconveniently high shelves used for some kind of slow or dead books.

It is usual to leave several feet above the top shelves, just under the roof, for ventilation.

_“Broken” floors_ are used in some libraries, the Massachusetts State Library, for instance; one stack floor being three and one-half feet higher and the next one three and one-half feet lower than the corresponding building floor, on the idea that it is easier to go up or down half a flight than a whole flight, for anyone wanting to get books. But isn’t the average the same? In this form, the very great convenience of moving books by trucks is sacrificed, so that the almost universal custom is to have the ground floor, and every second floor above, level in the stack with floors in the building, thus fixing the height of the latter at fourteen or fifteen feet, except the top floor, which is free, and the basement, usually determined by other exigencies.

The material used for “decks” may be openwork iron, marble, or more usually translucent ground glass.

The floor of the stack as well as of the building basement, is generally cemented, with special provisions for excluding dampness.

_Passages._ Those running lengthwise may be called gangways, those across between cases, aisles. The number of gangways varies with the size and use of the stack. Although it might be built without a center gangway, and have one on each side, or only on one side—it would then be a very narrow stack—the usual construction is to have a gangway about four feet wide down the center, and one of less width (just enough to allow passing around, say two feet,) at each outer end. But if it is desired to have very close packing, these side gangways may not be necessary. In building the new Salem stack, Mr. Jones decided that he could so run the classification of the books from the center around back to the center, in every aisle, that there would be little need of passing around the outer ends, and he could omit them and so gain that much more for books.

The center gangway may be any width desired, but should of course be wide enough to serve as thoroughfare for men, book-trucks, and boxes. Although four feet seems the average width, it varies from three feet to six feet in existing libraries. Good, large windows on each floor should light gangways at the far end.

The length of aisles varies with the width of the stack building, though limited by the belief that no bookcase should be more than 15 or 18 feet long, which requires other gangways at that interval. The width of the aisles has varied. The original Harvard width, 2 feet 4 inches, appears to be the very narrowest which will allow passage of two persons, or stooping to the lower shelves; 2 feet 8 inches is very common; 3 feet is so roomy that the stack becomes convenient for limited open-access; while 5 feet “on centers” (3′ 6″ or 8″ aisle) is the maximum in stacks at present.

Many stacks have wide intervals at the sides of the “deck” in each aisle—so wide as to have to be wired to prevent books falling through—“for ventilation, diffusion of light, and communication,” but such wide spaces are not needed for light or ventilation, and are much handier for dropping pencils than for passing books, so that I prefer wider decks with small rims for protection, and much narrower spaces along the cases.

_Stairs._ Stack stairs need not be wide, for they are so short that two people never need to pass. Two feet wide is enough. When first adopted, circular stairs were used, as supposed to occupy less space, but they were found to be inconvenient and dangerous, and since measurement has shown that straight stairs need occupy no more space, the “cork screws” have been entirely superseded. Eight-inch risers and 9-inch treads are recommended by Champneys,[296] who thinks, by the way, 2 feet 4 inches the right width, iron with rubber treads being the material.

Stairs should be put in wherever they will be most convenient, and where they interfere least with book storage and passing. One flight certainly should be next the entrance on each floor, and one flight generally at the other end. If they be set sideways in the folio shelving there, which is not always all needed, they seem to interfere least. (See paragraph on circular or winding stairs.)

_Lifts._ Light lifts for single books, or few books at a time, are needed for all stacks (See that title, on page 228.) In large libraries and high stacks, elevators large and strong enough to carry trucks and boxes, are also necessary. For lifts, hand operation will serve, or electricity; for freight elevators, some sort of power is better.

Every such carrier should run from basement to top, with opening on every floor. A speaking tube should run beside it, with mouthpiece also on each floor.

_Ledges._ (See under Shelving, p. 265.) As a ledge on both sides of each case would greatly narrow the aisles for passage and diminish the capacity for storage, these have disappeared from the modern stack. Their place has been taken in some stacks by sliding shelves (to be drawn out when wanted), which do not appear to be entirely satisfactory. But the need for some substitute, for the use of which Dewey speaks, has suggested ledges for folio shelving on each floor and for the new device of carrels, which may at least partially replace ledges without diminishing storage capacity or easy passage.

_Shelves._ The shelving of stacks follows the rules already described under the title “Shelving,” except as dimensions are varied by the use of steel, which is less bulky. Movable shelves also allow more variety in intervals to suit the average size of books in any part of the stack. It is usual to maintain the 10-inch height for intervals between shelves, all over the stack, except as thus modified here and there to suit exigencies and except for folio shelving at the ends (or sides) of each floor.

Different patents offer much choice in stack shelving. Avoid especially projections, likely to catch dust or tear clothing or injure books. Test very carefully all forms of “clutch” or detachable shelves.

=Stack Lighting.= _Natural._ North light is the best, but the choice is not often open. The location of the stack is determined usually by other considerations than aspect. Unless it runs along the rear or side of the main building; if it projects, that is, it will naturally have two sides lighted, one of which in any location would have to be south or west, and thus sunny. If wired glass is used as a protection against fire it will be more or less opaque and thus will temper glare. Shades can, of course, be used on the worst exposure, and some contrivance can be used, like that at the Library of Congress, to work all these curtains at once to save time.

Overhead light will penetrate one glass floor of a stack fairly well, not more.[297]

“If daylight is on the whole better and more wholesome, as it is certainly cheaper than electric light, then a well windowed stack room is better than a dark one.”—_Russell Sturgis._[298]

Light penetrates stack aisles effectively only about twenty feet, hence a stack lighted on both sides may be forty feet wide, plus width of centre aisle.

_Artificial._ The best light is, of course, electricity, and here the expert of the stack to be installed can give valuable advice. The question of the location of the bulbs, their power, their direction (transverse or perpendicular), their frequency, their wiring, their switches, such questions must be determined. As a great deal might depend on the particular structure of the stack, one bid for the stack, another for the lighting, with specifications from each bidder, might be invited.

Hand bulbs at the end of cords have not been found satisfactory. Various devices have been used, but good systems of fixed lights (bulbs with reflectors and shades), worked well by means of switches, have been perfected.

_Reflective Colors._ To help diffusion and local effectiveness of both natural and artificial light, inner walls and the whole stack would well be painted some agreeable light tint of enamelled paint. This is a question of taste for the architect, with approval by librarian and committee.

Stack Windows

As stack windows must be high and narrow, they introduce a new and imperative architectural feature on the exterior of the stack fronts. The usual form is a continuous window from foundation to eaves. This may, however, be broken for a foot up from every floor, by a cross band of iron or stone, for effect or for any interior convenience, like continuous hanging of steam pipes, without real diminution of daylight inside, provided that the windows run quite to the ceiling in each deck, to give full top light. If the windows are glazed with wire glass, they will afford some protection from outside fire, and being opaque, would temper the glare of sunlight. Factory ribbed glass is also used, as both tempering and intensifying daylight.

_True Windows._ To give full effect the piers between windows should be only as thick as the depth of the double book cases, sixteen inches, and directly opposite them. They have only to support themselves and the roof, as the stack floors are independent and self-supporting. Re-enforcement with a steel T-beam will render them stiff enough with sixteen inch width, and even allow flaring from the windows to admit more light.

With this construction, each window can have the full width of the aisle it fronts and be so framed and glazed as not to intercept any light, thus throwing illumination as far as possible down the aisle, with oblique rays from the side of the window to the other side of the aisle, reaching both rows of books to the far end.

This I call a true stack window. In looking over modern plans, you will see that many libraries have them as to position, though the entire available width is not always used.

If you have Clark’s “Care of Books,” see how true the alcove windows were in the Queen’s College, Cambridge, library as long ago as A. D. 1472.

_Defective Windows._ In other stacks, you will find windows too short (even if there is a cross band, it should not be more at the most than eighteen inches in height, leaving a window on each deck, six feet full down from the deck above), but oftener windows narrower than the aisle, giving too little light to reach the inner ends of the cases. There is no excuse for these. As has been said above, there is no structural need to build the piers between windows wider than the book cases inside, and just so much as they encroach upon the windows they commit the unpardonable sin of darkening the stack.

Many modern plans show this defect.

_False Windows._ By these I mean windows which outside take the gridiron stack form, but do not come truly and fully opposite every aisle inside.

“The rear elevation of the New York Public Library plainly shows that the architects wilfully omitted to place a window at the end of each aisle. All the beauty of the elevation will not make good the want of light in the lower floors of the stack.”—_Oscar Bluemner._[299]

The falsity of this arrangement, which is found in many modern libraries, lies in using an exterior scheme which does not meet inside conditions. The excuse is that sufficient diffused light is provided for the whole stack. But if this is true (which I cannot concede), any other equal window area could be used in any other form, which would not give outer promise of inward excellence. They are only a sham, and can therefore be called false stack windows.

=Heating.= The best form developed for stacks is by hot water or steam pipes along the walls just above the floor of each story clear of the books, with coils in the windows. Overhead pipes are very bad, as they concentrate heat at the top of each story, where it is most oppressive to those walking or working below.

=Ventilation.= There should be an air space above the top shelves in a stack. Good ventilation can be provided there by end windows and through the side windows. Some writers have advised sealed windows so as to be dust proof. In that case some system of forced draft would have to be installed.

The ventilation of a stack, where use by staff and public is only intermittent, is perhaps not so important as that of reading rooms constantly crowded, but the open construction and height of the stack differentiate the problem rather than avoid it.

=Underground.= In England, Burgoyne says[300] four stories is the rule. But in America, every library builds its stack, in all dimensions, according to its wants and space. Four-story stacks are common, but by no means the limit.

The impending exigencies of storage have not only brought suggestions of dark stacks in the interior of a building, but they have already carried stacks under ground. Even the Bodleian Library in England has installed a two-story subterranean stack, mechanically lighted and ventilated, under its front lawn. Plans are on foot for stacks many floors below ground-level, to be lighted and aired by electricity. See p. 222.

=Upward.= Ten “decks” is the maximum height now, but why is it not possible to build further up into the air before we burrow under ground? Are there any structural difficulties? Would it cost more to have a “sky-scraper” stack than a dungeon?

It is a question how underground cases will affect the books. It is claimed that forced draft will avert the evils of dampness, but Dr. Thwaites reports that he has found trouble from mould deposited on the backs of books as the warmer air from the surface above comes into contact with the cooler walls of the cellar. Would not books packed in sliding cases, away from the moving air, be more apt to develop inside rot and insects?

It does not appear to me that cellars for book storage have got beyond experimental stage. Some years of test seem needed to prove their perfect availability.

=Stack Towers.= B. R. Green says[301] “the stack might be in the center, and rise from the roof as a tower. It would be a simple thing to make a stack of twenty or more stories.” Why not? and why not so rise from an ell, as well as from the center? Why not build it as a sky-scraper, any number of stories upward, supporting itself, with a shell plastered on the exterior? The structural objections would seem no greater in a stack than an office building. The operating objections are surely no weightier going up than going down. The daylight would be better, the dampness less. It might be easier to flood cellars than towers, in case of fire, but the certainty of water is even a worse foe to books than the possibilities of fire.

Why is not here a chance to develop a new type of architectural beauty? If towers are fine features in churches and abbeys, why not in libraries? Before digging catacombs for our books, why not set our inventive faculties on hanging gardens of literature reached by elevators like the levels of the Eiffel Tower?

=Capacity.= Various ways of calculating capacity have been suggested, but most of them disregard the fact that stacks vary in measurement, and only two whose interior dimensions are exactly alike can be safely compared.

Capacity of an average stack can be roughly calculated at twenty volumes to a square foot on each deck. Thus a 30 × 40 stack, three stories high, will hold about 72,000 vols.

I prefer to calculate the capacity of every new stack independently, when planning it.

Taking folio shelving separately and adding its figures in later, I take one floor by itself. It has so many double cases, such and such length, on each side of the central gangway. One case 15 or 18 feet long, multiplied by 2 for the two sides, and 7 or 8 for such shelves as the librarian thinks he can use, then multiplied by 8 volumes to each foot, will give the “practical capacity” in volumes for octavos and duodecimos. Multiply by the number of cases on both sides, plus your calculation for folios, and you have the capacity of that deck. Multiply again by number of decks, and you have the practical capacity of the stack.

If you wish to get the “full capacity,” as it is reported in many plans, make your volume-multiplier ten instead of eight, or add twenty-five per cent to your first calculation, which amounts to the same thing. But eight to the foot is practically full capacity for closely classified libraries, where frequent gaps must be left for growth, at the end of each subject.

Sliding Cases

We can wisely borrow from England the “sliding presses” which Dr. Richard Garnett brought to the attention of the Library Association of the United Kingdom at its annual meeting of 1891, having previously described them in Dewey’s Library Notes and elsewhere in 1887.

Adapted from the Bethnal Green library in 1886, they were put on trial in the British Museum in 1887, and have since been in operation, regarded apparently as an invention quite as valuable as the stack appears to us. “I think enough has been said,” to quote Dr. Garnett’s words, “to convince librarians of the expediency of taking the sliding-press, or some analogous contrivance, into account in plans for the enlargement of old libraries, or the construction of new ones.”

The British Museum press is described as “an additional bookcase hung in the air from beams or rods projecting in front of the bookcase it is desired to enlarge, working by rollers running on metal ribs, and so suspended as not to touch the ground anywhere.” In other words, it is a movable bookcase parallel to a fixed case, and sliding to and from it by wheels above. It may be distinctively called a hanging case or press. It is better suited to the arrangement of aisles and construction of floors in the British Museum than to most American libraries, and so far as I know has not been copied here.

[See illustration in Library Notes,[302] and also in Burgoyne.[303]]

Another double press used at the Museum is called by Dr. Garnett the pivot press. It is apparently a second case, kept front to front close to the fixed case and swung out from it when wanted, by a door-motion hinged on a perpendicular pivot; overhung, I gather, at the Museum, but elsewhere running by wheels on metal semi-circular tracks laid on or in the floor. Such were early experiments in Trinity College, Dublin, twenty-five years ago. These might be called folding bookcases. They have not yet been copied in America.

A third kind of movable bookcase, which may more properly be called the sliding case, is used in the Patent Office Library, London. This apparently also swings from the top. Duff-Brown[304] describes it: “These presses are swung closely side by side, and drawn out, one at a time, as required.” He does not say drawn out endwise, however.

This idea is developed in The Librarian[305] by James Lymburn, who suggests “a store-room of any length, 22 feet wide by 35 feet high, in three stories, lighted from the roof through iron grating floors; with center passages of 9 feet, and sliding cases 6 feet long, closely packed in on each side.” He calculates that such a room 40 feet long would hold 100,000 volumes; its advantages being close storage and shelter from dust and sunlight.

See for illustration, Champneys.[306]

Jenner, in the Library Chronicle,[307] claims for the sliding case these merits: Cheapness, as compared with enlarging the building; possibility of gradual installation as needed; nearness to other shelves in a classification; absence of obstacle to light(?) or motion.

I have also received from a dealer in Oxford, England, a small pamphlet hinting at rather than describing, a room laid out after Lymburn’s idea. The pamphlet calculates it will save about half the space taken by stack storage. These cases, and Mr. Lymburn’s, are evidently double.

See also H. Woodbine in The Library Association Record.[308]

_Per contra_, H. M. Mayhew says in The Library,[309] “The drawback of the ordinary sliding or hanging or extension case is the difficulty of moving so great a weight whenever one book is wanted.”

I cannot figure out much from these English descriptions about problems of mechanism, repairs, lighting, or cleaning.

In America, the general idea of sliding cases has been discussed since Dr. Garnett’s description of the British Museum device in Library Notes, and since Mr. Gladstone called attention to it in the Nineteenth Century of March, 1890.

Mr. Gladstone describes what he calls these “book cemeteries” thus, as he has seen the “tentative and initial processes”:—

“The masses represented by filled bookcases are set one in front of the other, and in order that access may be had as required, they are set on trams inserted in the floor (which must be a strong one), and wheeled off and on as occasion requires.”

The masses which he thinks ought first “be selected for interment” are Hansard’s Debates, the Gentleman’s Magazine, and the Annual Register.

So far as I know only two trials of this idea have been made here; several years ago by Dr. Little at Bowdoin College, more recently by Mr. Lane at Harvard University. Both of these are wooden single cases, side by side, pulled out by the end, and locked or lockable. Both slide, not hang.

Mr. Lane has now a line of twenty-three in a row, sliding on ball-bearing wheels at the bottom, which in turn run on rails countersunk in the floor. At the top, the cases are held erect and guided, but not supported, by small wheels along the sides of a T-rail. He uses his cases entirely for rare books in an exhibition room on the ground floor, and finds them very satisfactory for the purpose, although he utters a warning that provision should be made for free access to all the mechanism, which occasionally needs repair.

Dr. Little submitted a paper describing his cases to the A. L. Institute at its New York meeting in 1911. By reference to a photographic view accompanying I see that he has a double-decker,—two stories of five single wooden cases each; each case “about six feet high and three feet long.” “These cases can be made of either wood or metal, for either octavos or quartos, supplied with either fixed or movable shelves.” [At Harvard the middle shelf is fixed as a brace, the others are movable.] “They must be mounted _at the center of the base_ on small ball-bearing trucks which run on metal rails sunk in the floor. Their tops are at the same time guided and kept securely in place by a slot and a T-iron, the friction against which is reduced to a minimum by rollers, placed horizontally. If properly constructed and placed upon level rails, a slight pull with one hand will bring one forth. The increased storage is estimated at 100 per cent.... We also have the Patent Office Gazette on six wooden sliding cases like these, on either side of the door of the room in which they are stored.... This method of storage is especially economical in case a depository library desires to keep its sheep-bound set of Congressional Documents as a unit, arranged by their serial number.... The cost of these cases and their installation varies greatly with the material, finish and location. My first cost less than $15 each, my last about twice that amount.”

I suppose Dr. Little means this for the cost of each separate bookcase, fully equipped and mounted. Mr. Lane’s figures I have not been able to put my hands on.

So far for the statement of facts. I must confess to having approached the subject with some prejudice against the mechanism of these cases, founded on an experience of sliding doors in dwelling houses, which slide or not, as they feel like it, and whose machinery is most difficult to get at and repair. But machinery can be got under control by mechanics. I yield my prejudices in view of the evident advantages of this system, and am prepared to make definite suggestions as to its use in future repairs or building in this country.

In alterations of those architectural extravagances which have wasted so much perpendicular capacity in high rooms and corridors, I see a way to use the style of cases experimented on by Dr. Little and Mr. Lane, rather than any of the English styles. Either as a single story along a wall anywhere, or in the double story style, swung out anywhere on the vacant floor of any room or any unnecessarily wide corridor, there will be relief in the storage of any books not required for open access or frequent reference;—as Dr. Little says, “for compact storage of less used books.”

In planning new buildings I hardly think it would be necessary to set up such cases at first, except perhaps in the case of rare books as at Harvard, where locked cases and protection from sunlight were wanted, with infrequent access; or in equipping rooms for rapidly growing sets, such as Congressional or State Documents, Patent Office Reports, sets of periodicals or publications of societies, or any similar sets whose titles and volume numbers can be labelled on the ends of the cases; or for “dead” books. The Oxford pamphlet sketches a room somewhat after the “Poole plan,” equipped with tables and chairs toward the windows and a row of sliding cases along the blind wall opposite the window light. This seems to me good for many departments.

But except in rooms evidently adapted to such treatment, I would not install sliding shelves anywhere, but would most certainly leave space, in a perfectly dry basement if nowhere else, for possible future installation whenever need may arise.

One reason for this postponement is this: that several details must be studied, experimented on, and perfected before fully equipped rooms of this kind can be considered as tried out and permanently satisfactory. Lymburn’s scheme seems good, but the plans presented by Champneys and the dealers do not work out well on examination as regards space, light or handling. I suggest as problems to be investigated,—

Smooth and sure working of the mechanism. Easy access to top and rear for repairs. Access for cleaning and ventilation. Incidence of weight (this is not even on floors as in a stack; but is moving, as on bridges). Lighting (most important) on each face of each case. Floors sure to remain true. Width of center aisles for all emergencies.

See Bookworms, p. 222.