How extraordinary to think that President Butler in 1926 already had such a clear vision of the library that would open on South Field eight years later, even to the extent of having an accurate notion of the building's final costs. The story seems such a confirmation of the exceptional command of University affairs for which Columbia's president of 43 years has often been remembered. Unfortunately, the evidence suggests the President had no such vision, indeed that the conversation never took place as described. Instead, the rise of the library that would eventually bear Nicholas Murray Butler's name - from the choice of a site to the building's design and construction - was a far more complex enterprise and far less the product of President Butler's mind than he seems to have recalled during his last years.

Surely Butler did already know in 1926 that Columbia needed a new library. But the story of its evolution actually appears to have started in August 1927, when Charles Williamson, the recently appointed director of Columbia's library, addressed a thirteen-page letter to President Butler: "During the past year," Williamson wrote, "I have given considerable thought to the problem of providing an adequate central building for the University library. The need for a modern building grows more acute each year. Already a condition has been reached which threatens to hamper the growth and development of the University..."²

The pantheon that Charles McKim had built thirty years earlier in tribute to Seth Low's father had always been a better monument than a library. Williamson, in his letter to Butler, ticked off Low Library's sins : the public service spaces, lofty though they might be, were cramped; the reference collection had all but taken over the rotunda, crowding out the readers; there was no place to accomodate the growing card catalogues; above all, Low lacked adequate shelf space. "Any building erected now," Williamson told Butler, "should provide for the growth of forty years at least, which would therefore mean shelving for not less than 4,000,000 volumes. Moreover, in any building plan adopted now some thought should be given to the still more distant future."³

Williamson's proposed solution didn't actually entail the construction of a new building but only the completion of McKim's never-finished University Hall. Located just north of Low Library, it was a building in which the campus' original architect had planned to house a theatre, a student dining hall and the university's administrative offices. Only the lower stories had ever been completed in permanent form, housing the old gymnasium and swimming pool, facilities which remain there even today, buried in the foundations of the Business School, their curved, north facade largely concealed by the new Schapiro. But in 1927 the completion of University Hall seemed the next step in the fulfillment of McKim's master plan for the campus, and Williamson's proposal for a University Hall library envisioned a building whose scale exceeded anything even McKim had imagined for Columbia.

Williamson's proposal actually involved not only the completion of University Hall but its physical merger with Low Library, and the fact that the latter building was to serve as an enormous vestibule to the new facility gives some sense of the scale of the librarian's vision. A researcher would climb the steps of Low, passing into the rotunda, emptied of reference books and reading tables, newly devoted to the ceremonial functions it serves today. On the north side of the rotunda, in the space that is now Low's faculty room, a grand staircase would be built, leading to a bridge. Spanning the space between Low and University Hall, the five-story bridge would contain the main reading room, its windows looking east and west to Schermerhorn and Havemeyer. At the north end of the bridge, the researcher would enter University Hall. The loan desk and reference room would stand to the right and left, the book stack directly ahead, surrounded by a semi-circle of reading rooms and lecture halls. The stack core itself would rise through the center of the completed building's eight stories. In addition, excavation under the plaza separating Low and University Hall would provide a connection to the old library's considerable book stack and also allow underground expansion east and west, potentially increasing the library's four-million volume capacity to six and accomodating the University's collections beyond the century's end.

Seven years after Williamson's letter to President Butler, on November 30, 1934, the two men sat with over nine hundred guests in the main reading room of Columbia's new library, presiding over its dedication ceremony. But the room in which they sat was not the dramatic facility Williamson had described to Butler years earlier, spanning the plaza between Low and University Hall, nor could the building in which they found themselves be likened to the dramatic structure envisioned by Williamson in 1927. Built not on the upper campus but behind the playing fields on South Field, the new library's stacks held just under three million volumes, not the six million Williamson had planned. Low Library's 750,000-volume book stack was now too far removed from the new building to serve any useful purpose and so was lost to the library and in time converted to office space. Of Columbia's 37 departmental reading rooms, only ten fit into the new building, leaving Williamson with much the same scattered system he had inherited in 1926. The new building, prosaically styled South Hall, until it was renamed for President Butler in 1946, was the product of compromise - financial, architectural and operational. The story of its design and construction illustrates both the scale of Columbia's vision and the limits of its resources.

From the day he hired Williamson to run the library, President Butler had a donor in mind to fund the new building. Edward S. Harkness, who had inherited a significant portion of his father's 15% share in Standard Oil and spent his life giving the money away, focused his largesse on educational and health care institutions, funding their buildings while spurring them to reform their operations. It was Harkness who had engineered the merger of Columbia's College of Physicians and Surgeons with Presbyterian Hospital, and he funded development of the two institutions' vast complex on Washington Heights between 1921 and 1928.⁴ His concern to decentralize America's growing universities into more humane, residential units had put some $13 million into Harvard's "house plan" and similar funds into the residential colleges of his alma mater, Yale. When first approached by Butler in 1926 to pay for a new library, Harkness expressed a preference for funding a "more useful" building. But in time Butler and Williamson convinced him that housing Columbia's books was just as important as housing students at Harvard and Yale.⁵

Along with Harkness' money came his favorite architect, James Gamble Rogers. Over the previous decades Rogers - whose friends called him Gamble - had made himself a premier force in academic architecture, designing major portions of the campuses at Yale, the University of Chicago and Northwestern, often with Harkness picking up the bills. A gentleman's gentleman, valued as much for his dinner table repartee as for his architectural skills, Rogers had already worked with Butler in designing the Harkness-funded Columbia-Presbyterian hospital complex. Just as Butler was approaching Harkness in 1926, Rogers was undertaking the design of Yale's new Sterling Library. He seemed the inevitable choice to convert Williamson's University Hall dream into a brick and stone reality.

  Active correspondence between Williamson and Rogers began in the summer of 1928, as Rogers arranged for the library director to examine the drawings of Yale's Sterling and Williamson sent the architect a "tentative" program document for the building he believed Columbia required.⁶ The fruits of these discussions are manifest in Avery Library's drawing collection, where an array of Rogers drawings ranging from rough sketches to detailed blueprints present at least two options for executing Williamson's proposal. One set of drawings is a nearly precise reflection of the building described in Williamson's 1927 letter to President Butler and probably represents those early discussions (see Figure 3). Another set varies, moving the main reading room out of the bridge and into University Hall, perhaps drawn up after Rogers' receipt of the program document in the Summer of 1930 (see Figures 4 and 5).

But the University Hall design ran into resistance almost immediately. On October 1 of 1930 Williamson wrote to Rogers: "It seems clear to me as a result of the work you have done on it that within the precise limits of area and height of the original plan for University Hall, it would not be possible to construct a building that would be adequate for the needs of the University even at the present time."⁷ This can hardly have been a surprise to Williamson, whose proposals since 1927 had never envisioned a building corresponding to McKim's original plans for University Hall. The library director lamented: "Perhaps the proposal I originally made and still think the wisest solution of our problem, namely, a union of the present building and the new one, is too radical to secure the approval of the Trustees." He suggested a compromise solution, widening University Hall and increasing its height, thereby allowing the entire library to fit within its confines and eliminating the great merger. But even this limited proposal offered little prospect of success with the trustees, as Williamson lamented, "we are baffled in every attempt to find some suitable location for a library building off this block." The plan's final failure is represented in yet another set of drawings for completion of University Hall, dated in 1930, a version of the building that no longer contains a library or any connection to Low at all.⁸

  The available correspondence doesn't indicate the nature of the trustees' objection to Williamson's grand scheme, except that they wanted University Hall to remain a separate building and so would not agree to its merger with Low. The trustees might have had aesthetic objections. The building Rogers had laid out, even in the blueprints, is a daunting structure stretching almost 600 feet from the steps of Low to the north end of University Hall, rivaling the Cathedral of St. John the Divine in acreage. The arched first story of the bridge might have lightened the design somewhat, but the library would have dominated the campus in an extraordinary manner. Rogers, in a January 31 note to Williamson, made reference to the concerns of Prof. Joseph Hudnut of the School of Architecture, saying he appreciated the professor's "feelings and the question he brings up of locating the Library where we have it," and noting that Hudnut took a "great interest in the General Plan."⁹ Assuming the reference to the "General Plan" meant McKim's master design for the campus, Hudnut may have been a force in squelching the University Hall designs.

Yet it seems likely the plan's doom was spelled as much by cost as by aesthetics. The gymnasium and pool would have to remain in University Hall, until plans for a South Field athletic facility could be realized. So the book stack would have to be constructed, at least initially, atop these large open spaces, a difficult and expensive engineering challenge. At this very time President Butler was struggling to convince Edward S. Harkness to pick up the entire cost of the new library, an agreement reached in December of 1930.¹⁰ Looking north to Yale, whose Sterling library was nearing completion at a cost of over $8 million, Harkness may have balked at the Williamson/Rogers designs.

The solution began to dawn in the first weeks of 1931. On January 29 Williamson wrote to Rogers regarding the previous day's meeting of the Library Council at which President Butler "spoke in a guarded way of the idea of a building on South Field".¹¹ The new location was a logical alternative, being a site on which the master plan itself had envisioned a student center and/or gymnasium. A 1923 drawing by McKim reveals a building in this position that might well have inspired Rogers' eventual design: a renaissance palazzo with second-story colonnade and an inspirational freize - a passage from Dante rather than the current set of classical authors.¹² Hudnut apparently favored the new location, as Williamson proposed bringing him to see Rogers, and the initial plans for South Hall apparently arose from those meetings directly after the January 28 Library Council.

Those initial plans are reflected in the first of six sets of South Hall blueprints and in a rendering of the north facade by Rogers, both in the Avery Library drawings collection (see Figure 1). These documents present a building both like and unlike the one that eventually rose on South Field. The exterior displays certain notable differences: the familiar frieze containing the names of Greek and Roman authors is missing, and in its place is a more prominent string course, providing greater unity between the building's central block and the wings east and west; the attic story is brick and limestone rather than solid limestone; most notably, the building has a three-door entrance, an element that many who observe the current library's diminuitive front door would find an attractive change. But the differences, and with them the greater cost of this original design, are yet more apparent in the blueprints. A look at the lobby is a simple illustration: the three doors are evident here, and they are reflected in a double row of columns, creating a virtual nave and aisles. A lecture hall to the right and a reserve reading room to the left flank the lobby with busy public spaces rather than the sedate reading rooms that took those locations in the final plans; the grand staircase does not disappear right and left into enclosed tunnels, as it does now, for the walls separating it from the ground floor corridors are opened in three archways on each side. A similarly dramatic change is notable in the plans for the floor above: the great colonnade across the building's north facade is composed of free-standing columns, giving the facade a greater depth and texture than the less expensive pilasters, built into the reading room wall of the building as constructed.

But once again Williamson and Rogers ran into opposition, and this time the obstacle was purely financial. On April 9, 1931 the construction firm of Mark Eidlitz & Sons provided a cost estimate on Rogers' initial design, and the price tag was well over $5 million.¹³ Edward S. Harkness had agreed to spend approximately $3 million on the new building. President Butler did his best to convince the philanthropist to support the increased cost, still courting the donor's concern to build a "useful building", writing to Harkness' assistant, Malcolm Aldrich, that the library would be "the central workshop of every part of the University, thronged by students and professors both by day and by night."¹⁴ But the depression was on, and even the heir to Standard Oil was not immune. The cost of the building had to come down.

On April 30 Rogers wrote to Williamson with a description of the revised building: it would be 264 feet wide, 167 feet deep, with a main reading extending through the second, third and fourth floors and a book stack rising in 15 tiers and capable of holding 2.9 million volumes.¹⁵ Surely Williamson must have cringed, thinking of his own 4-million volume minumum. Rogers assuaged the library director's concerns a bit, noting that his plan entailed erecting the stack "in such a way that six additional tiers may be added for the future, which will give additional capacity of 1,100,000 volumes." This would accomodate the forty years' growth Williamson thought essential in his original letter to Butler. But there was no longer a provision for the "still more distant future" to which he had so sensibly referred in his 1927 letter.

The price tag had settled down to a workable $3,678,500, but dispute over the new library was not finished. The ground breaking ceremonies on June 4 were hardly over, when Prof. Helmut Lehman-Haupt took Rogers and Williamson to task for their failure "to set an example of spiritual independence and courage in breaking away for the first time from an obsolete custom."¹⁶ Lehman-Haupt, who taught in the School of Library Service and was director of the Rare Book Department, found Rogers' design "a very dry and uninspired product of the Tacademical' taste in the very bad sense of the word." He recognized the need to build something that harmonized with the general spirit of the campus architecture, but he was convinced this could be done with modern concepts and without imitating "a historical style which belongs definitely to the past."

The dispute was nothing new to Rogers. It hardly compared with the response to Sterling Library at Yale. In 1930 a Yale student magazine, The Harkness Hoot, had published an article by an undergraduate, Harlan Hale, entitled, "Art Vs. Yale University."¹⁷ The article's essence was expressed in a pair of photographs: one of Sterling's book tower under construction, its steel superstructure still fully exposed - the other of the completed library in all its neo-gothic grandeur. Under the former photo Hale wrote: "It might have been made into a monumental modern building - with the structural and decorative ideas evolved by American skyscraper designers newly adapted to a splendid and living institutional structure." He called the finished Sterling "a monument of lifelessness and decadence". The author received encouraging letters from the likes of Henry Russell Hitchcock, Lewis Mumford and Hugh Ferriss, and Frank Lloyd is said to have had the article read aloud in his drafting room.¹⁸

But collegiate eclecticism was not quite dead yet. The trustees of the Yale Corporation were not inclined to heed Mr. Hale's criticism, and Lehman-Haupt's comments at Columbia fell on equally deaf ears. Williamson passed the professor's letter on to Charles H. Kent, the project architect for South Hall at the Rogers' firm, apparently not considering the comments worthy of Rogers' own consideration.¹⁹ The library director simply said he did not pretend to know much about architecture and that he always told his critics, "of which there are unfortunately a good many around the University - that it is a matter in which I do not feel any more concerned than the rest of the community."

Despite Williamson's declared ignorance of architecture, he in fact took the most intimate interest in the continuing design and construction of South Hall. His requests for changes yielded a new set of blueprints on June 10, yet another set on June 19 and another on July 6. Charles Kent wrote to the library director on July 16 acknowledging that "minor changes" would still be possible but saying it was necessary to give a final set of drawings to the engineers for their approval.²⁰ Yet Williamson's requests did not cease, though agonizing sciatica put him in and out of the hospital all during the design and construction phases. From his bed at Presbyterian Hospital and from the room he took at the King's Crown Hotel on 116th Street, when recuperation didn't allow his commute from Hastings-on-Hudson, he continued to oversee the design of his building. On July 16 he asked Kent whether the windows in the Main Reading Room might be extended all the way to the floor, "or at least to within 2' or 3' of the floor", to improve the view of South Field and Low Library.²¹ It didn't happen. But two weeks later he informed Rogers that the Reference Room lacked adequate shelving and patron space. He requested the addition of a mezzanine with shelves, and that change did shape the room Columbia students use today.²² An extensive correspondence in August and September concerned the Browsing Room.²³ Williamson even sent his own sketches, moving the room's door to the south end and adding two alcoves. In June 1932 the director wrote to Angus McDonald of Snead & Co., contractors for the book stack, rejecting plans for "rolling shelving" on the upper tiers, predicting that graduate students and officers would have stack access and could not use what apparently was an early version of compact shelving.²⁴ A seemingly endless correpondence concerned the extensive pneumatic tube system, and a detailed letter of Williamson to Charles Kent described the director's preferred design of the "annunciator system", the number board above the circulation desk.²⁵ The last great change actually came after the building's completion. The book stack facade, protruding above the sixth-floor roof, had been constructed of brick. But in the fall of 1934, when Williamson, Rogers and Butler stood back and looked at their new building, they decided the bricks didn't work, and several thousand dollars were spent to rebuild the facade entirely in limestone.²⁶ Williamson's comments continued even after the library's official dedication, as in 1935 he complained to Eidlitz & Sons that he didn't care for the chandeliers in the Main Reading Room.²⁷

And so the finished building might easily be called a collaboration between James Gamble Rogers and Charles Williamson. It was unmistakably a Rogers building. The facade easily recalls the Shelby County Courthouse Rogers built in Memphis in 1905-09, and the inspiration for South Hall's Main Reading Room can be seen in the court room Rogers designed for New Orleans in 1908-1915.²⁸ The beauty of the library's public spaces is also a Rogers hallmark, exhibiting his noted mastery of detail work in the Browsing Room's panelled alcoves and the wonderful ceilings in the two north reading rooms of the ground floor. But the building is also suffused with Williamson's concern that it be a workable library. If it could not hold the six million volumes he might have wanted, the library director could insure that the book stack was air conditioned to preserve the collections. Though it did not gather all or even most of the library system's many reading rooms under one roof, as Williamson had hoped, the public and staff spaces were arranged to accomodate the efficient operations that had been impossible in Low.

  There was much excitement at Columbia upon the opening of South Hall. Guests came from all over the world for the dedication, where Mr. John Buchan, British publicist and MP for the Scottish universities, delivered a stirring oration on the value of great libraries. The Department of Buildings and Grounds hired 22 cleaners to maintain the building's public rooms and corridors, and a full-time electrician and two plumbers were devoted to the building.²⁹ But Williamson cannot have avoided some disappointment at the compromises built into his new library. Its architecture didn't meet with rave reviews. Even the special issue of Alumni News published to celebrate the building's dedication is filled with impressive facts but quite devoid of the adulation one might expect. Indeed simple words of praise are difficult to find. Within hardly more than a decade there was discussion of altering the building's layout, and detailed plans for its complete redesign were drawn up for the first time in 1969.³⁰

South Hall was a state-of-the-art library building, when it opened in the fall of 1934. Its pneumatic tubes and conveyor belts, its air conditioned book stack, its non-glare lighting were all the finest technology available at the time. Williamson cannot have known how soon all of these wonders would become obsolete - how the advent of browsable stacks would render the pneumatic tubes and conveyor belts useless, how the growing polution of New York's air would overwelm the stack air conditioning, how the building's corridors and reading rooms would come to seem dark and gloomy by modern standards. But from time to time Williamson's eye must have wandered north to University Hall, which remained unfinished long after his retirement, and he must occasionally have regretted the compromises and fiscal realities that had given him South Hall.

NOTES

^*This article would not have been possible without the assistance of many of my fellow librarians, especially the staffs of Columbia's Rare Book and Manuscript Library and the Avery Architectural and Fine Arts Library. I owe special debts to Bernie Crystal, who first suggested I look at the Library Office Files, to Hollee Haswell, with whom I have spent many profitable hours examining Columbia's history, and to Janet Parks and Daniel Kany, whose encyclopedic knowledge of their Avery Drawings Collection led me to discoveries no finding aid could have yielded. The illustrations in the this article are courtesy of Avery's Drawings and Archives.

1. "South Hall Renamed," Columbia Alumni News, June 1946, p. 11.

2. Williamson to Butler, 16 August 1927. Library Office Files, Correspondence Wa-Z, Rare Book and Manuscript Library, Columbia University.

3. Ibid.

4. Albert R. Lamb, The Presbyterian Hospital and the Columbia-Presbyterian Medical Center 1868-1943 (New York: Columbia University Press, 1955), 75-91, 161-64.

5. Aaron Betsky, James Gamble Rogers and the Architecture of Pragmatism (New York, 1994), 203.

6. Williamson to Rogers, 2 July 1928, "Tentative Statement of Requirements for a New Library Building for Columbia University", date-stamped 30 Jl. '30. Library Office Files, Correspondence of Charles Clarence Williamson, 1928-1942, Rare Book and Manuscript Library, Columbia University.

7. Williamson to Rogers, 1 October 1930. Library Office Files, Correspondence of C.C. Williamson, 1928-42.

8. Drawings and Archives, Avery Architectural and Fine Arts Library, Columbia University.

9. Rogers to Williamson, 31 January 1931, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

10. Betsky, 203.

11. Williamson to Rogers, 29 January 1931, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

12. Papers of McKim, Mead & White, New York Historical Society.

13. Betsky, 203.

14. Ibid.

15. Rogers to Williamson, 30 April 1931. Library Office Files, Correspondence of C.C. Williamson, 1928-42.

16. Lehman-Haupt to Williamson, 12 June 1931. Library Office Files, Correspondence of C.C. Williamson, 1928-42.

17. William Harlan Hale, "Art Vs. Yale University," The Harkness Hoot, 13 November 1930, 17-32.

18. Betsky, 60.

19. Williamson to Kent, 13 June 1931, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

20. James Gamble Rogers, Inc. (signed by C.N. Kent) to Williamson, 16 July 1931, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

21. Williamson to James Gamble Rogers, Inc., 16 July 1931, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

22. Williamson to Rogers, 39 July 1931, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

23. Williamson to Kent of James Gamble Rogers, Inc., 21 August, 11, 14 September 1931, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

24. Williamson to MacDonald, President, Snead & Co., 8 June 1932, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

25. Williamson to Kent, 13 July 1933, Library Office Files, Correspondence of C.C. Williamson, 1928-42.

26. Columbia Alumni News, Special South Hall Issue, 23 November 1934, 6.

27. Memorandum of C.C. Williamson, 1 May 1935. Library Office Files, Correspondence of C.C. Williamson, 1928-42.

28. Betsky, 82, 85.

29. Columbia Alumni News, Special South Hall Issue, 23 November 1934, 9.

30. "Plans for the Renovation of Butler Library," by Prentice & Chan, December 1969, Butler Library Floor Plans & Furnishings, Library Office Files, Correspondence of C.C. Williamson, 1928-42.