College and Research Libraries Twenty-Five Years of Academic · Library Building Planning David Kaser The modular design concept, introduced into American library building planning following World War II, was fully accepted by 1960. In the early years the usual modular design was simple and uncluttered, lending itself well to efficient library operation. Some felt that the ap- pearance of the early modular buildings was too plain. As a result, the last quarter century has seen increased use of atria, designer lighting, monumental effects, unusual shapes, and other devices to relieve the perceived monotony of the unadorned early module. Many of these embel- lishments work to the detriment of sound library service. Colleges and universities should take greater care to assure that architects understand the imperative nature of their library build- ings' functional requirements, even if the resulting structures are less imposing in appearance. ore academic library buildings have been constructed in the United States in the last quarter century than in any other pe- riod in history. One might expect that the quality of libraries built at the end of the period would be unquestionably better than of those built at the beginning, but many librarians question the fulfillment of this expectation. This paper will review li- brary planning developments, identify sa- lient strengths and weaknesses, and com- ment upon problems and prospects in this important aspect of academic library work. THE CLASSIC MODULE By 1960 the simple modular design con- cept had totally superseded the fixed- function character that had dominated American academic library building plan- ning during the first half of the twentieth century.* The compelling virtues of the new style were (1) it was readily adaptable to post-World War II library service con- cepts requiring that readers have direct physical access to books on open shelves, (2) it lent itself readily to the profession's contemporary concern for improving the ''time and motion'' efficiency of libraries, and (3) it was remarkably flexible. Readers were no longer restricted to · massive reading rooms with floor struc- tures unable to support the weight of the books they wished to consult. Librarians could now fulfill their social contracts in modest structures rather than in palatial halls. Library interiors could now, as they could not previously, be rearranged easily *In a modular building the floor area is divided into equal rectangles, all of which are usually defined by structural columns at the corners, as well as by uniformity in ceiling heights, air and lighting treat- ment, and floor-loading capacities, thus allowing their use to be modified as needs change . A fixed- function building, on the other hand, lacks one or more of these features, thus tying the use of floor areas largely to the purposes for which they were originally designed. Before World War II, for exam- ple, almost all library buildings were engineered to carry the weight of the book collections only in multitiered structural stacks, thereby dictating a permanent separation of reader areas from stack ar- eas. David Kaser is professor, School of Library and Information Science, Indiana University, Bloomington, Indiana 47405. Dr. Kaser has been involved in planning more than eighty academic library buildings in North America, Asia, Africa, and the Middle East . 268 J and cheaply. It is small wonder that mod- ular academic library buildings were so quickly adopted by the librarians who were to work in them. Proposed initially by architect and stack manufacturer Angus Snead Macdonald in the mid-1930s, the full application of the modular concept of library building plan- ning was delayed for more than a decade, in part by the usual forces of conserva- tism, but also by the depression and by World War Ir_l With the encouragement, however, of its first principal proponent, Ralph Ellsworth, director of libraries of the ·University of Iowa, modular planning was used increasingly in the years immedi- ately after the war, first at Hardin- Simmons College (1947) and then at Princeton (1948), North Dakota State (1950), and Iowa (1951) universities. 2 The 1950s saw further consolidation of its strengths and further resolution of most of its principal weaknesses, so that by 1960 the modular academic library building was as close to perfection as it was fated to come. Because of its unadorned severity and stark simplicity, its uncompromising ad- herence to architect Louis Sullivan's doc- trine that ''form must follow function,'' and the almost brutal economy of its struc- tural style, classicists found the very ten- ets of their professional and esthetic University of Iowa: unadorned modular simplicity. Twenty-Five Years 269 creeds embodied within the design of early modular buildings. Among build- ings built in this early tradition were the li- braries of Louisiana State (1958), Colgate (1958), Brigham Young (1958), and St. Louis (1959) universities. However, some felt that this early mod- ular library building lacked visual interest. Its reliance on uniformly rectangular com- ponents was boring and cried out for re- lief. It was ugly, they said, and -looked ''like a box,'' although a box was indeed exactly what library functional require- ments warranted, with rectangularity dic- tated by the real library module-the book itself-of which the building must accom- modate hundreds of thousands or even millions. Nevertheless, it soon became clear that the ''box'' was going to have to be dressed up so that it looked less like a box. It is difficult to determine today where this movement to disguise the box origi- nated, although it does not seem to have been among librarians. Librarians were very aware of how well off they were with the box and preferred to remain with it. It is easy to hypothesize that donors were among the first to want more than a box. They, after all, were in some cases to have their names memorialized by these build- ings and might be indisposed to favor just a box. Or it may have been university 270 College & Research Libraries presidents and trustees who wanted something more elaborate in their library buildings than boxes. Some administra- tors began calling for their new library buildings to serve not only as libraries but also as symbols, statements, signatures, or embodiments of institutional style, ac- tual or imagined. DECADENCE AND THE ROMANTIC MODULE As soon as these dual expectations be- gan to arise, the overall quality of aca- demic library building design in the United States began slowly to decline. Li- brary buildings were no longer allowed to serve one simple, straightforward func- tion that would determine their form. Now library buildings were expected to serve two functions, one bibliothecal and the other symbolic. Just as a slave cannot serve two masters, form cannot serve two functions. Most architects appear to have been just as proud of the boxes they designed in the 1950s as their first owners were to have them.lt must be acknowledged, however, that architects generally were quick to fol- low the wishes of their clients when irr the early 1960s they began to abandon the box in favor of more esoteric forms. Thereaf- ter, many architects were found in the vanguard of the movement away from simplicity. An art historian might say that begin- ning about 1960 a period of classicism in academic library building design gave way rapidly to a strong romantic move- ment. Contorted shapes, unusual tex- tures, artistic license, emphasis upon ef- fect, revulsion against constraint and discipline, return to tradition (or what was perceived as traditional) and to nature (or what was perceived as natural)-all the standard paraphernalia of romanticism pervaded much of the work that was done during the next two decades. As a result, the modular planning so consistent with classic theories of design, and so much ap- preciated by librarians, became unfash- ionable. These romantic elements were also partially responsible for the rapid in- crease in building costs during this period. Boxes are simply cheaper to build than July 1984 more complex containers. It is ironic that this period of romanti- cism set in at almost the very moment that the classical modular form was attaining its zenith, so much so that both styles can be seen in a single watershed building. The library at Washington University in St. Louis, planned in the late 1950s and opened in 1962, has been described by many informed observers as the modular building brought to its highest level of suc- cess. It is a square structure with entry in the middle of the third of five stories, pro- viding the most efficient access to all loca- tions within, square bays sized to library modular equipment, low ceilings, uni- form air and light treatment throughout, and distribution of activities based solely upon the functions themselves. These qualities led Ellsworth to call it simply ''the best.' ' 3 Few have departed markedly from that assessment. Yet the beginning of decline is also evi- dent in Washington University's fine modular library. Two features were incor- porated into the design primarily to keep it from looking too much like a box; both cost extra money to construct, and one of them impaired function (although only slightly). One of these features was a wide deck that circumscribed the building at the second level above grade. The other was a tree court notched into one corner of the building. The deck was claimed to be nec- essary to shade a collar of glass used at grade level to make the building appear to "float" on the campus, thereby diminish- ing its apparent bulk. The tree court was rationalized as a device to bring natural light to the main stair, reserve reading area, and rare book suite. But a deck twenty-seven feet wide completely sur- rounding the building was certainly not needed just for shading, and since the in- terior of the building was amply and am- biently lit artificially, natural light was not needed. The court used up space needed for library activities, especially at grade level, thereby obtruding upon function. Both were extremely pleasant features, however, and they did relieve the per- ceived monotony of a box-shaped build- ing. Not all academic library architecture Washington University: Ellsworth calls it "the best. " broke so immediately during the early 1960s for the more euphoric and emo- tional style of the romantic movement. Some buildings designed during this pe- riod even won AlA/ ALA awards for their classical commitment to functionalism. Among such award winners were the un- dergraduate library at the University of South Carolina; completed in 1960; and the libraries of Lafayette College (1963), the University of Miami (1964), and Le- Moyne College in Memphis (1964). Al- though each of these buildings had some functional problems (as virtually any li- brary will), none resulted from a lack of modularity. All of these buildings were square or rectangular, and all evinced re- straint in design. As the years passed, however, these qualities became increas- ingly rare in newly designed academic li- brary buildings. PROBLEM AREAS What were some of the problems that in- creasingly insinuated themselves into American library design during the subse- quent ye~us? Few were new problems; most instead were old problems that had grown out of hand. Most were already ob- vious to Keyes Metcalf in the early 1960s when he wrote his comprehensive vol- ume, Planning Academic and Research Li- brary Buildings. Among those which he enumerated were (1) irregular shapes, (2) Twenty-Five Years 271 interior or exterior courts, (3) monumen- tality, and (4) too much or too little glass. 4 These four potential flaws will be dis- cussed here with a view to their promi- nence in buildings that have been con- structed during the last two decades. Irregular Shapes Other things being equal, simple squares or rectangles that can be entered near the center of the long side, especially at the building's middle level, lend them- selves best to economical library use and operation. They create minimal exterior wall area requiring expensive cosmetic treatment. Of course, other things are sel- dom if ever equal, and legitimate factors, - usually relating to irreconcilable site con- siderations, frequently impinge upon a building's design. In too many cases, however, this economy of construction and operation has been forgone for no ap- parent reason except to attain desired es- thetic effect or, more frequently, to keep the box from looking like a box. Some of these irregularly shaped build- ings have been round, although that cer- tainly was not a new shape for libraries. Semicircular libraries were built as early as Roman times, and the first completely cir- cular library was the Radcliffe Camera built at Oxford University in 1749. Many others have been built since then. Since books and off-the-shelf library furniture 272 College & Research Libraries are rectangular, round buildings (or in- deed buildings of any other nonrectangu- lar form) are profligate in their use of floor space. Radial stack layouts, for example, require a certain minimum distance (per- haps thirty inches) between ranges at the hub end simply to allow persons to pass through. The farther the stack ranges radi- ate from the center, however, the wider the distance between ranges becomes, creating large areas of unusable floor area which reduce the efficiency and drive up the cost of the building. Building contrac- tors, moreover, can construct buildings with ninety-degree angles cheaper than they can contend with curvilinear or other nonrigh t -angular structures. Although round libraries have never been found to be efficient, they continue to be built with unabashed frequency. The University of Corpus. Christi built one in 1963, as did Chabot College in 1966, St. Peter's College in 1967, and St. Michael's College in 1968. One architect experi- mented unsuccessfully with radial stacks at Wells College in 1968 but then pro- ceeded to design three round stack towers with the same deficiencies for Northwest- ern University a year later. Other institu- tions built libraries that were only partially curved, such as Nevada Southern Univer- sity and Worcester Polytechnic Institute, both in 1967. It is ironic that at the same time that some architeCts were trying to disguise rectangular boxes, others were trying to disguise these "hat b_oxes" by Worcester State College: ratchet-shaped. - ---------~ July 1984 .J modifying their roundness. Thus Mary- wood College opened a gear-shaped li- . brary in 1967, and Worcester State College built a ratchet-shaped building in 1970. Some other institutions, while eschew- ing rectangles, at least opted for rectilinear forms. Oral Roberts University built a hexagon in 1966, and Drexel University (1959) and Marymount (1967) built octa- gons. Widener College (1969), the Univer~ sity of Toronto (1973), and Sangamon State University (1976) settled for trian- gles, and the University of Texas built a parallelogram in 1978. Western Illinois University opened a library shaped like a pinwheel in 1978, with each level turned forty-five degrees from the floors above and below. The shape of the University of Chicago Library building (1970), while re- taining a semblance of rectangularity, nonetheless defied simple description, and the library at the University of Califor- nia at San Diego (1970) took the shape of a mushroom cloud. Still others failed to as- sume any discernible shape, such as the Wells College Library, which, in Ellsworth's words, "oozes down the hill- side" toward Cayuga Lake. 5 Interior and Exterior Courts There was a time, before modern light- ing and air treatment became available, when interior and exterior courts were necessary in order to make central spaces in large buildings usable. Atria and light wells are thus among the oldest architec- tural refinements known to the human race, and they have been used effectively in libraries since before the Christian era. High vaulted ceilings and clerestory light- ing were essential standard features of the eighteenth- and nineteenth-century al- coved book halls used as libraries follow- ing their introduction by Sir Christopher · Wren at Trinity College, Cambridge, in the 1690s. By the end of World War II, however, good ambient artificial light and efficient ventilating and air-cooling systems were available at low cost, so that open wells, courts, and high ceilings were seldom if ever thereafter needed for any functional purpose. With the need for them gone, their great inefficiency soon became ap- parent. It was immediately recognized that four kinds of problems, none suscep- tible to easy resolution, resulted from the use of wells. The first was that they in- creased construction cost by swelling the bulk of the building, thus increasing the amount of exterior skin that needed to be finished off. The second was that they in- creased maintenance costs by creating large blocks of unusable interior cubage that had to be heated in winter and cooled in summer. Third, they usually blocked more direct and efficient circulation, or they were in locations that could have served better for assignable library pur- poses. And fourth (and doubtless most annoying to patrons), they permitted the transmission of noise vertically from floor to floor, impairing the libraries' essential acoustical ambience. Also, floor areas be- low atria and in high-ceilinged rooms were difficult to light, and they reduced flexibility in the future use of a building. Obviously, exterior courts experienced only some of these problems, but they were nonetheless deserving of concern. For these reasons few atria were used in the early (i.e., pre-1960) modular build- ings, although occasional mezzanines and other high ceilings made appearances, largely as vestiges of the fixed-function li- braries that had preceded them. Law school libraries, especially, continued to favor mezzanines, but some turned up Twenty-Five Years 273 also in general academic libraries. The La- mont Library at Harvard (1949), the library at Georgia Institute of Technology (1955), the St. Louis University Library (1959), and the Colorado College Library (1962) can be counted among the early modular buildings that utilized mezzanines. High ceilings tended in this early period to be limited in use to entryways, as at South- ern Illinois University (1955) and Clemson University (1966). Despite their functional deficiencies, however, atria and wells, open and inte- rior, have become used widely during the last two decades. In fact, for a time follow- ing the opening in 1967 of Atlanta's Re- gency Hyatt Hotel, resplendent with a cavernous atrium, libraries like bagels seemingly all came with holes in the mid- dle, the assumption apparently being that if an atrium is good for the hotel business it must be good for the library business as well. Not only did wells increase in popu- larity, but they also increased in size. But- ler University's library atrium (1963) was relatively modest in scale, as was the one at Providence College (1968), while those at the Countway Medical Library (1966), the undergraduate library at Stanford . University (1967), and the undergraduate library at the University of Washington (1972) yawned ever larger and higher. The trend to bigger interior wells had to end somewhere, and it appears to have done so with the opening in 1970 of New York University's library that boasted a gaping 10,000-square-foot maw that loomed upward fully twelve stories from its entrance on Washington Square. There are two ironies in this building. The first is that for almost a century the largest inte- rior well in an American library had been that of the Peabody Library in Baltimore (1878), which because of its size had sig- naled the end of the earlier book-hall style of library architecture. The second irony is that on the outside the NYU Library looks like an unmitigated box. By this time, however, some of the more deleterious effects of atria in libraries were becoming apparent, and efforts were made to render them less obtrusive. In the first place, they were all scaled back in 274 College & Res~arch Libraries size. The University of Utah (1967) and Le- high University (1984) glassed in theii: atria from the start to reduce sound trans- mission, and, after more than a decade, Stanford University was able to obtain funds to encase in glass the atrium in its undergraduate library so as to diminish the decibel level in the building's center. Some newer libraries, such as the Western illinois University Library (1978), reduced the acoustics problems of their atria by lin- ing them at all levels with utilities or other functions less likely to be disturbed by sound. Others with existing wells some- times attempted to mask the obtrusive sounds transmitted through them by in- stalling a bubbling fountain at the lowest level. Called "sound perfume" by some, this step was considered at Delta State College (1968) and Illinois Wesleyan Uni- versity (1968), and it was tried at Butler University (1963) with little success. Despite uniformly poor experience with wells and high ceilings in academic li- . braries, they continue to be built in ever- growing numbers. The University of Chi- cago (1970), Clark University (1968), Hamilton College (1972), and St. Mary's College at Notre Dame (1982) are only a handful of the many libraries built in the United States in the last twenty years that have been plagued to a greater or lesser degree by one or more of the four prob- lems of atria or wells enumerated above . Monumentality For more than four millennia prior to the Renaissance, libraries served primarily temple and palace function~, and they were therefore almost always housed in monumental structures appropriate to their status. Even the alcoved book-hall li- braries, used almost universally in the United States until1880, were adapted di- rectly from cathedral architecture, com- plete with narthex, nave, aisles, and apse, bringing library monumentality close to the present era. Considering this very re- cent palatial heritage of library architec- ture, it is not surprising that the fixed- function libraries built between 1910 and 1940 continued to use monumental ele- ments. Monumentality in library buildings can July 1984 take many forms. In general, the term re- fers to almost any building element that exceeds in size or cost what is necessitated by its function. Thus much of what was discussed earlier in this paper on irregular building shapes and open wells also can be classed as monumentality. Over-broad entries, elevated podiums, sumptuous building materials, grandiose staircases, as well as conspicuously crafted accesso- ries, can, and in library buildings usually do, constitute monumentality. High ceilings are often said to be monu- mental, and in many cases they are, but in the early decades of this century they were also often essential to good functional li- brary design. The handy 71 /z-foot floor-to- floor dimension of the multitier structural steel stacks used during that period al- most dictated that multiples of that di- mension be used elsewhere in the build- ing (if floors were to meet), encouraging high ceilings, especially in reading rooms. Large windows, moreover, could supple- ment the limited illumination available from incandescent lamps and, in the ab- sence of air conditioning, could be used to exhaust the heat build-up. The modular style of the last three de- cades, however, when used in conjunc- tion with modern lighting and air treat- ment, eliminated all functional need for ceiling heights above 81 /z feet, thus uncou- pling library design from its princely ori- gins and allowing the development of util- itarian library structures appropriate to their present-day egalitarian societal role. Yet, as with atria, some vestigial monu- mentality in ceiling heights continues to survive. Examples may be seen at St. John's University (1966) and at Colgate University (1982), where high ceilings were used for esthetic rather than for func- tional reasons. High ceilings, as with atria, necessitate larger areas of exterior skin to enclose the building, and they create sonic, illumination, and air-treatment problems that today' s library users should not have to endure. Other kinds of monumentality, how- ever, have declined in use since the ad- vent of the module. To be sure, a softer material with better acoustical properties · would have been more practical in the vast J l NYU library lobby than its pretentious, ex- pensive, and resounding marble. Curva- cious stairs can often be very fetching and achieve a monumental effect, but they also are conducive to vertigo and are likely eventually to be eliminated by building codes. Nonetheless they continue to be widely used in libraries, as at Adelphi Col- lege (1963), Scarritt College (1968), Fisk University (1969), and Rosary College (1970). In other efforts to achieve monumental- ity, some library buildings designed in the last two decades gave up the flexibility of the module, so much appreciated by li- brarians, and returned to the fixed- function constraints of the earlier style. The libraries at Clark University (1968) and at St. Mary's College at Notre Dame (1982), for example, both incorporated very low ceilings where stacks were ini- tially installed, and very high ceilings or wells and fixed task lighting in some other locations, rendering any future effort to revise layout difficult, if at all possible, to accomplish. The greatest impairment of Clark University: a fixed-function "box." Twenty-Five Years 275 future flexibility has been the relinquish- ing in many buildings of the principle of modular ceiling lighting. A desire to "play," "bathe," and "landscape" with light, rather than to use it for library pur- poses, led to the construction of some buildings, as at Wells College (1968) and the School of Business at Indiana Univer- sity (1981), that are virtually unalterable because it would be too costly to change the lighting. Other libraries, meanwhile, lost flexibility because they lacked ade- quate floor-loading capability to permit stack relocation into certain areas, as into the cantilevered extensions of the upper levels of the Sangamon State University building (1976). Too Much or Too Little Glass Only one of the problems reported by Metcalf twenty years ago has been eased; that is the use of too much or of too little glass. Direct sunlight has never been good for reading purposes, and indirect natural light is available for reading during only one half the hours that a library is open. 276 College & Research Libraries Although the full measure of the deleteri- ous effects of sunlight on book-paper has only recently become clear, it has long been known to bleach bookbindings. For these reasons, librarians have long es- chewed windows on any but the north fa- cades of their buildings. · Prior to the availability of modern flores- cent lighting, libraries had to rely heavily upon natural light, and the extensive fen- estration used in many of the older fixed- function buildings was carried over into the early modular libraries. Furthermore, as a recently developed building material, glass enjoyed unusually wide popularity that, although perhaps warranted in some kinds of structures, caused severe prob- lems for libraries. Among libraries from this period suffering from the overuse of glass were those at Grinnell College (1959), Butler University (1963), Clafin College (1967), and the University of Cali- fornia at San Diego (1970). Overuse of glass in the 1950s and early 1960s brought on a modest revulsion against it, resulting in some buildings be- ing built with few if any windows. To be faddish, it seemed that buildings built around 1965 had to be faced either entirely of glass or completely without it. Among . "under-windowed" library buildings constructed during this period were those at Oral Roberts University (1966), Roches- ter Institute of Technology (1967), and In- diana University (1968). This ambiguous situation prevailed at the time Metcalf commented upon the misuse of glass in li- braries. For the most part, relatively good ·sense has been used in the employment of glass in library buildings since that time. The in- troduction of mirror glass for building ex- teriors in the mid -1970s led for a time to its overuse. The Benedict College Library (1976), for example, was built of mirror glass, and the addition to the Vassar Col- lege Library (1976) was originally planned in mirror glass, although it was changed prior to construction because of its energy inefficiency. In most recent library build- ings, however, concern for function has resulted both in limited fenestration, based upon the needs of library users rather than upon some counterproductive July 1984 esthetic effect, and in the proper use ot'ex- terior sunscreens rather than reliance upon drapes or indoor blinds requiring frequent replacement. Nevertheless, be- cause of the large number of poorly fenes- trated library buildings already in use, li- brarians justifiably study with a jaundiced eye all proposals for windows in new space. NEW DIRECTIONS Among the many restricting realities that academic library building planners had to face during the period under dis- cussion were increasingly high real estate costs, limited construction space, andre- luctance to obtrude a building unnecessar- ily upon remaining vistas. As a result, a number of high-rise and below-grade li- brary buildings were constructed. If these had been the only reasons for their use, high-rise and below-grade .buildings might have rested comfortably in their ac- complishment, but unfortunately other, nonfunctional, considerations sometimes also came into play, limiting their success. Up ... High-rise library buildings were not an invention of the last quarter century. The New York Mercantile Library and the first John Crerar Library buildings were high- rise buildings, both built well over sixty years ago, but they were not academic li- braries. On the other hand, the twenty- eight-story Cret building at the University of Texas (1934), the library tower at Fisk University (1936), and the nineteen-story Hoover Institution at Stanford University (1940) were academic libraries, so univer- sity librarians in 1960 were not wholly lacking in experience with them. Thus when the University of Notre Dame built its present fourteen-story library building (1963), many old heads questioned its wis- dom. Rather than being dictated by library functional factors, this building's height was determined by the university's desire for a symbol of its academic excellence that would outshine its reputation for football prowess. The library fulfilled that sym- bolic purpose admirably, with its mosaic- bedecked facade visible from 60 percent of the seats in the stadium. Happily, also,. Notre -Dame: irreverent un.der- gradua1es be1ie..ve Christ signals J'Touchdown.1" the building worked .quite well .as a ii- br-ary1 since the assignable area on each tower floor w_as large enough to permit-the effective .dEployment of an ample admix- ture of staCk--and Teader facilities, a cond.i- tion nol: 1:-t"ue uf tbe ~artier buildings at Texas, Fisk, and Stanford. This first high-rise library in a quarter cent-ury iRSpired nthe-r.s to try similar seh.emes, hut perhaps none has been so SU£Cessful, and some have been abysmal .failur-es. Were it not f.or its unique pro- gt"am requirement til bring -departmental libraries un~r a single roof and yet emible them to retain some semblance of their-in- dividual integrity {a dubiouspoliticat ded- sion r-ather than .a sound library decision), the fourteen-story sc-ience library at --Brown University -(1966) wouid rertainly have been disproportionately talL The Twenty-Five Years 277 ten--story height of -the library at Hofstra University (1967) may hav-e exceeded an "'ptimalTatio!o its breadth and depth, -and the stack tower at Memphis State U.niver~ .sity (1968} dearly did so .. The misbegotten twenty-e1.gh.t~leve11ibr.ary .structur-e at the UniverSity-ofMassacnusettst1972) has the single virtue !hat no one has yet exceeded it in height. If hist-ory can serve as a -guide" Jmwever., and if institutions continue tQ demand tha.tthe~riibrary buildings serv-e a symbolk fun.clion, then it -seems likely that the height-ev-en -of this towering zig- _gurat will-eventually b.e surpassed~ dOubt- less witb equally-disma1-r.esu1ts_ ••• -and Down Burying library buildings, or ,at -least lalge portions of them, below grade was -.an ·innovation. of lhe period be-ing -dis- 278 College & Research Libraries cussed here, and this was generally much more successful than the aforementioned efforts to go up. In the 1960s the principal reasons for constructing libraries below grade we.re (1) to keep their large bulk from overwhelming adjacent structures or areas, or (2) to keep an important site or view unobstructed. The first reason ac- counted for the location below grade of fully half of the floor area of the Washing- ton University Library in 1962 and more than three fourths of the Johns Hopkins University Library in 1964. Meanwhile, the second consideration accounted for the first library building to be constructed completely below grade at Hendrix College in 1967. Here the only satisfactory site was at the center of the one campus quadrangle where a conven- July 1984 University of Massachusetts : a "towering ziggurat" of twenty- eight floors. tional building would have blocked all views and circulation, so a two-level building was sunkinto the ground and the occasion used to relandscape the quad above. In 1968 two other libraries, the sci- ence library at Vanderbilt University and the undergraduate library at the Univer- sity of illinois, were built completely be- low ground for somewhat similar reasons. The former was sited below ground to en- able it to serve an engirdling ring of labora- tory and classroom buildings without vis- ually encroaching upon their quadrangle, and the latter to attain proximity to the main library without casting a profaning shadow upon the sanctity of the Morrow experimental corn plots, a compelling concern in the state of Illinois. In the early 1970s a third reason for Hendrix College: the first library completely below grade. building below ground became promi- nent. It was the energy efficiency of such a structure, and a number of underground or partially underground libraries were built in order to gain this economy. Some were set into hillsides, cropping out to the Twenty-Five Years 279 leeward, such as the libraries of Scripps Institute of Oceanography (1976) and St. Meinrad College (1983). Others continued to be built for reasons other than energy efficiency, although they enjoyed that benefit as well. Among them was the Pu- St. Meinrad College: underground floors crop out to the leeward. 280 -college & Research Libraries sey Library at Harvard, c-ompleted in 1976. Here the need to effect a juncture with the Lamont, Widener, and Houghton li- braries while preserving the integrity of the Yard led to the planning of a five-level below-grade "Structure, -although lor cost reasems its height was reduced to three levels before 1t was built. Likewise for site reasons, the -University of Michigan Law Library was expanded below -grade in 1981, with much more su<:cess than a pre- vious expansion in 1955, when a ghastly addition in glass and aluminum was per- formed upon the :graceful-Gothic splendor {')f the Cook law quadrangle. Considering1he 1ev:el-of popular appre- hension re_garding undergr-{)um:i libraries, surprisingly few problems -have been ~n­ amnter..ed wi.th l hem. No more water -comes 1nt-o t-hem than comes through sky- lights or, f-or that matterJ t-hrough ilat roofs. Since they require li-ttle~ if .any, -costly exterior facade, they can -often be cons-t-ructed .as cheaply as, or rhea_p~r than, conventional above-grade struc- tures, an.-d modern lighting .and -air tr-eat- ment can render them as ·habitable and graci0us as any olher int-erior space . Since they have no-exterior visible 'form., there is no-temptation to <:on tart them into 1rregu.- lar shapes or masses inimical to sound fi- br.ary function. One probl'em -common to alm-ost -all below-grade libraries built-thus far, now- every has been !heir reliance -en exteiior wells to bring natural1ight{and nsuaily · vegetation) down into their interiors .. Ob- viously rlone to mask over the troglodyte :charader of these spaces, -such wells nonetheless .create the same library -.ays- tunctions below ,grade as when {hey are used above grade .. The -light weHs cal the University of Illinois, Harv-ard, and Heinecke Library at Ya1e University (19~, l or-example, aU pierced t-he -service floors and th-e.refor,e .obtruded a.pon Hbr~ar_y nee.d-s; The