In Celebration

CONTENTS

1      PREFACE

3      INTRODUCTION

7      The Library of Congress and The National Union Catalog
            William J. Welsh

14     The National Union Catalog and Research Libraries
            Gordon R. Williams

21     Editing the NUC
            David A. Smith

28     Publishing the NUC
            John Commander

32     Antiquarian Booksellers and The National Union Catalog: A Survey
            Bernard Rosenthal

         Scholarly Uses of The National Union Catalog:

39         An International Perspective, by Nicolas Barker

45         A Bibliographic Saga, by William B. Todd

LIBRARIANS AND SCHOLARS from the United States and abroad gathered at the Library of Congress on January 27-28, 1981, to cele- brate the completion of the 754-volume National Union Catalog, Pre- 1956 Imprints, a monumental, fourteen-year publishing project of great importance for libraries and the world of scholarship. A sym- posium featuring papers by both those who developed and those who use The National Union Catalog (NUC) was the major event. The symposium sponsor, the Center for the Book in the Library of Congress, is pleased to make the principal papers available to a wide audience.

During the two-day meeting, many well-deserved tributes were offered to the project's staff and to members of the American Library Association's National Union Catalog Subcommittee, which shep- herded the project to completion. Seven members of the subcom- mittee were present: Gordon R. Williams, chair, National Union Catalog Committee, and former director, Center for Research Libraries; Douglas W. Bryant, former university librarian, Harvard University; Ralph E. Ellsworth, former director, University of Colo- rado Libraries; Warren J. Haas, president, Council on Library Resources, Inc.; Rutherford D. Rogers, Yale University librarian; Frederick H. Wagman, former director, University of Michigan Library; and William J. Welsh, Deputy Librarian of Congress. Four were unable to attend: John W. Cronin, Charles David, Herman Fussler, and George Schwegmann. One, Verner W. Clapp, is deceased.

In addition to thanking the participants whose remarks appear in this volume, the Center for the Book is grateful to the other sym- posium speakers who helped make the occasion a lively mixture of sentiment and scholarship: Henriette Avram, director for processing systems, networks and automation planning, Library of Congress; Johannes Dewton, head of the project from 1967 to 1975; Robert B. Downs, dean of library administration emeritus, University of Illi- nois at Urbana-Champaign; Sir Frank Francis, director and principal librarian of the British Museum, 1959-68; Joseph H. Howard, Assistant Librarian for Processing Services, Library of Congress; and

Henry Snyder, editor and director, Eighteenth-Century Short Title Catalog/North America, chair of the National Union Catalog Com- mittee and one of the principal organizers of the project. Thanks also go to William Matheson, chief, Rare Book and Special Collections Division, for his help in organizing the symposium and to Margery Maier for her assistance in preparing this book for publication.

Established by an act of Congress in 1977, the Center for the Book exists to "keep the book flourishing" by stimulating interest in books, reading, and the printed word. The center works closely with organizations outside the Library of Congress to increase the public's book awareness, to use other media to promote reading, and to stim- ulate the study of books. It pursues these goals primarily by bringing together members of the book, educational, and business commu- nities for symposia and projects. In addition to reading promotion, the center's major interests include the educational and cultural role of the book-nationally and internationally; the history of books and printing; the future of the book and the printed word; authorship and writing; the printing, publishing, and preservation of books; the use of books and printed materials; and literacy.

The Center for the Book's symposia and publications are made possible by gifts from individuals and organizations. Contributions are tax-deductible. Further information is available from the Center for the Book, Library of Congress, Washington, D.C. 20540.

A UNION CATALOG is, essentially, a centralized finding list of books in more than one library. Such catalogs traditionally have served several useful purposes: they facilitate interlibrary loans, they permit coordinated acquisition efforts, and they are a source for cataloging records. Each of these functions strengthens cooperation among libraries, whether the catalog is regional, national, or international, and no matter what its size. But if cooperation is an intrinsic part of union catalogs, so are patience, perseverance, and-especially- idealism. It is the idealistic notion that one ought to be able to put one's own hands on all the research materials in a certain region or country that lies at the root of the union catalog idea. The ideal predates the invention of printing. The first attempts at a union catalog are credited to the monk John Boston de Bury, who visited English monasteries to gather information about their manuscripts for his Catalog scriptorum ecclesiae, which appeared about 1410. The holdings of 195 monastic libraries were identified, and the continuing dream of a bibliographical utopia-a "complete" listing- has fueled the imaginations of scholars and librarians ever since.

The first prominent librarian to advocate a national union cata- log for the United States was Charles Coffin Jewett, librarian of the Smithsonian Institution from 1847 to 1854. Jewett felt that the newly created Smithsonian Institution should be the American national library and his proposed catalog, which would make the Smithsonian library the national bibliographic center, was an essential feature of his plan. His particular scheme came to naught, for his national library hopes were in direct conflict with the goals of Smithsonian Secretary Joseph Henry, who fired his ambitious librarian in 1854. However, Jewett's general plan was the foundation, half a century later, for Librarian of Congress Herbert Putnam's union catalog at the Library of Congress. Like Jewett, Putnam recognized the neces- sity of taking advantage of the latest technological developments in order to establish the catalog. Both men also perceived the impor- tance of the union catalog function for a library that had national ambitions.

Jewett's goal, described in his 1850 annual report, was a general

printed catalog of all books in American libraries so "every student in America would have the means of knowing the full extent of his resources for investigation." The next step would be a "universal catalog," for "if the system should be successful in this country, it may eventually be so in every country in Europe." Since the printed book catalogs were awkward, expensive, and quickly outdated, he proposed that the entries in the Smithsonian catalog be made from stereotyped plates that could be used by other libraries. Partici- pating libraries could prepare their own entries and plates for titles not in the Smithsonian, building their own catalogs and contributing to the central catalog in Washington. Uniform cataloging rules, essential for such an endeavor, were part of the idealistic Smith- sonian librarian's plan: "Nothing, so far as can be avoided, should be left to the individual taste or judgment of a cataloguer."

Card catalogs had replaced book catalogs by 1901 when Herbert Putnam began his union catalog at the Library of Congress, but it was a technological innovation, an updated version of the stereo- typed plate, that enabled Putnam to carry out Jewett's general plan. The innovation was the printing of the Library's catalog cards.' Dis- tribution of the cards to American libraries was the next step and the exchange of catalog cards soon brought a national union catalog into existence. Putnam thus fulfilled his hope of finding a way the "National Library" could reach out from Washington and perform "a service to the country at large." The Library of Congress-not the Smithsonian Institution-had become the de facto American national library.

Today the computer has greatly accelerated the impact of technology on all forms of bibliographical control, perhaps especially the union catalog. As Deputy Librarian of Congress William J. Welsh points out in the first paper in this volume, The National Union Catalog, Pre-1956 Imprints is probably the last large biblio- graphic catalog that will be compiled manually. Recognition that the era of nonautomated bibliographical control is behind us was one of the reasons the Library of Congress wanted to mark the com- pletion of the project. We wanted to bring those responsible for the catalog's planning, editing, and publication together with the librar- ians, bibliographers, booksellers, and scholars for whom it was intended. We wanted to learn about the catalog's actual uses-who, why, how-and about its potential uses. We sought and received opinions about two important questions: (1) How can we make the best and most efficient use of the information in this catalog that has been gathered, edited, and published so diligently over the past

fourteen years? and (2) From the standpoint of the user of the Pre- 1956 Imprints volumes, how can we best use the computer and other new technologies in planning the future of The National Union Catalog and the other Library of Congress book catalogs?

The future development of The National Union Catalog is not, however, the focus of this particular volume. These papers tell the story of how The National Union Catalog, Pre-1956 Imprints came to be published and how it is now being used. It is a tale of determi- nation in which librarians, publishers, and scholars can take pride. It also tells us something about the changing nature of the Library of Congress as a national institution.

During Herbert Putnam's librarianship, which spanned four decades (1899-1939), the Library of Congress tended to dictate cataloging and bibliographic policy to American libraries. And the union catalog could not have gotten started in any other way. In recent decades, as emphasized in Gordon Williams's paper, the rela- tionship between the Library of Congress and American libraries has become one of mutual support and cooperation. The 754-volume National Union Catalog, Pre-1956 Imprints, completed without any government or foundation funds, is a cooperative triumph eminently worth celebrating.

"An estimated 610 volumes to be published over 10 years."
"The most massive venture in catalog publicationyet undertaken."
"All research libraries will need this monumental bibliograph-
ical tool."

These are a few of the exuberant predictions made by librarians and scholars in 1967 when publication of the pre-1956 portion of The National Union Catalog was announced. We are now celebrating the success of this remarkable cooperative endeavor which, on the road to completion, extended itself from the "estimated 610 volumes in 10 years" to 754 volumes in fourteen years.

Rooted in the last decade of the nineteenth century, the union catalog became possible with the development of standardized cata- loging rules and a uniform-size catalog card. In October 1901, under the aggressive leadership of Librarian Herbert Putnam, the Library of Congress began to print and distribute its catalog cards. Three months later Librarian Putnam, never shy about his goals for the Library of Congress or for American librarianship, announced the creation of what was, in effect, the national union catalog.

It is fully recognized by the Library of Congress that next in importance to an adequate exhibit of its own resources, comes the

________

7

ability to supply information as to the resources of other libraries. As steps in this direction may be mentioned: First. The acquisition of printed catalogs of libraries, both American and foreign. Second. An alphabetic author catalog on cards of books in department and bureau libraries in Washington. Third. A similar catalog of books in some of the more important libraries outside of Washington.

The Library of Congress expects to place in each great center of research in the United States a copy of every card which it prints for its own catalogs; these will form there a statement of what the National Library contains. It hopes to receive a copy of every card printed by the New York Public Library, the Boston Library, the Harvard University Library, the John Crerar Library, and several others. These it will arrange and preserve in a card catalog of great collections outside of Washington.

From being a record of the holding of only a few distinguished libraries, with the Newberry Library and the libraries of the Uni- versities of Illinois and Chicago joining those mentioned by Putnam, the union catalog grew only gradually until the 1920s. Then, with a grant of $50,000 a year for five years from the Rockefeller Founda- tion, the project progressed rapidly from 1927 to 1932. Sought by the American Library Association, the Rockefeller gift resulted in "Project B," which, under the direction of Ernest Gushing Richard- son, added over six million cards to the catalog with the goal of locat- ing at least one copy of every important research book in American libraries. When "Project B" came to an end, the Union Catalog Divi- sion was established in the Library of Congress in September 1932.

The division began to receive the steady flow of incoming reports and to stimulate even more American and Canadian libraries to participate in the program. In the 1940s, the project benefited by the creative work of John Cronin of the Library's Processing Depart- ment, a remarkable man who developed and executed the idea of publishing in book form the Library of Congress's own catalog of printed cards, a 167-volume publishing project completed in 1946 by Edwards Brothers. The book catalog, which had been abandoned by most American libraries for half a century, was reborn.

In 1948 the union catalog was officially designated the National Union Catalog, and in the early 1950s the American Library Asso- ciation established a subcommittee on the National Union Catalog headed by Frederick H. Wagman. As a result, the catalog was divided into two parts. For imprints after 1956 the National Union Catalog was edited for publication by the Library's Catalog Publica- tion Division and has since been appearing on a regular basis, with monthly or quarterly issues and various annual and quinquennial

cumulations. The pre-1956 portion of the catalog, called the "retro- spective NUC," was maintained and added to as before, but its pub- lication was still distant.

Librarians are a persistent lot, however. Breakthroughs in the 1950s in bibliographical control of current acquisitions and catalog- ing, along with the demonstrated value of the post-1956 portion of The National Union Catalog in book form, made the need for editing and publishing the pre-1956 portion increasingly evident to librar- ians and others. In 1959 the ALA subcommittee on the National Union Catalog sponsored a pilot project to edit for publication all cards with imprint dates of 1952-55, inclusive, which, with Johannes L. Dewton as supervisor and editor, was completed in 1961. The result was a thirty-volume catalog that demonstrated that publica- tion of the entire pre-1956 file was possible.

In June 1964 the American Library Association and the Library of Congress signed a formal agreement in which the ALA agreed to procure funds to enable the Library to edit the pre-1956 imprints catalog for publication. In December 1964, subcommittee chairman Gordon Williams of the Center for Research Libraries reported that several publishers were definitely interested in bidding for the right to publish the pre-1956 portion in book form and were prepared to advance the editorial and printing costs. For the next two years the committee continued to discuss methods of publication, finally con- cluding that publication of the present catalog in book form was desirable even if it would eventually become available also in machine-readable form. Two different invitations to bid were issued in 1966. At the second deadline date, August 1966, three bids were submitted with sample pages. The bids were based on the price at which the publisher would make the printed volumes available to libraries and research institutions and, after due consideration, the subcommittee chose the bid of Mansell Information/Publishing Ltd. as the one providing the lowest sales price and the most satis- factory format. Under the terms of the agreement, Mansell under- took to pay all costs and expenses of publication and to make available to the American Library Association funds to finance the cost of the editorial work at the Library of Congress.

And how did this British firm get involved? John Cronin, in helping to bring the third edition of the Union List of Serials into being, had become acquainted in London with Mansell Publishing and in particular with John Commander. Mansell was the publisher of many works of interest to the library world, perhaps especially the British Museum General Catalogue of Printed Books. A Mansell-

related company, working in the field of optics during World War II, had developed certain camera techniques relevant to publishing, techniques found useful for solving many of the problems posed by publishing a massive work. The heart of the system involved the sense-marking of each card, best described as a shorthand instruction to the camera directing it to film only portions of the card rather than the entire card. The cards then could be filmed very rapidly and without individual handling. In short, a workable system-not computer-based, but rapid and efficient-gained the bid.

So the bargain was struck and the American Library Association assumed the responsibility of overseeing all phases of the project through its subcommittee on the National Union Catalog, still chaired by Gordon Williams. The Library of Congress, not a direct party to the contract, was designated the editing agent. And after only one more major delay, to sort out various problems involving copyright, work was ready to begin.

Early in 1967 a staff was formed, headed by Johannes Dewton, working under John Cronin's direction, and editing began. The Library of Congress was to supply the publisher with enough edited cards to furnish subscribers five volumes a month, total volumes to be 610, each with 698 pages, the whole to be completed in ten years. It is remarkably fortunate that we started when we did. The dark- ening economic picture since 1967 would surely have diminished enthusiasm for such a venture-indeed it has made its completion exceedingly difficult for all parties.

Mansell Publishing is a relatively small part of a very large British concern known as the Bemrose Corporation, a venerable firm whose publishing interests date back to the days of the earliest British railroad timetables and whose other printing specialties include calendars and printed checks. The support provided by Bemrose throughout the project has been of major importance, since the catalog itself has yet to show a profit. Mansell has borne an extraordinary financial burden. The editorial costs associated exclusively with the work at the Library of Congress have thus far exceeded twelve million dollars. The original fifteen-dollar-per- volume cost to subscribers seems as remote as the days of the penny postcard. That the overwhelming majority of the approximately 1,350 subscribers have continued to find funds as the costs have risen is a testimony to the catalog's value.

Johannes Dewton headed the project until his retirement in 1975, when David A. Smith assumed the helm. And what did editing the catalog entail? A remarkable effort to sift through about twenty

million cards that made up a catalog best described as overweight, unsightly, uncoordinated, and sluggish. The goal was to take this accumulation, built around changing catalog codes, idiosyncratic schemes, and tremendous differences in degree of completion and accuracy and, within the constraints of the money at hand and the time available, to raise its level to that of a "well-edited catalog." The story is one of compromise, simplification, negotiation, and increas- ing flexibility. The editorial process, once fully under way, involved twenty-five to thirty professional editors who every week syste- matically examined each card in a tray of approximately fourteen hundred cards. The burden borne by every editor was the project's unyielding publication timetable: the requirement to forward to Mansell almost twenty thousand finished cards every single week over an unbroken fourteen-year period. This requirement related directly to the publisher's own commitment to provide five printed volumes each month to each subscriber. The clock dominated the editorial process. Only rarely was there a cushion of finished work; the real race was to make each Friday's Pan American Airways Flight 106 to London. This shipping routing was carried out over six hundred times, and not a card was lost.

The editors twice marched through the alphabet. Editing on main sequence entries was finished in June 1979. Still to go were the 3.25 million cards that had cumulated since work began in 1967. The final job was the production of a supplement to integrate these entries with those in the 685-volume main sequence. Following a plan designed by the project's assistant head, Maria Laqueur, during the last year of main sequence editing, project editors began anew at Alpha. The seventy volumes of the supplement will include newly received reports from contributing libraries, a multitude of added entries and references, and a register of added locations, designed specifically to augment the number of locations identified for the less widely held items in the main sequence. Editing of the last part of the supplement was completed by the project's senior editors on Monday, January 12, 1981, once again in time for the Pan Am 106 to London.

Despite the obstacles I have mentioned, the quality of the pre- 1956 imprints catalog is high and the publication has already proved to be even more useful than the planners had envisioned. The rea- son, and a great source of pride for each of us associated with this undertaking, is the project's staff. They were a group of highly motivated men and women who somehow successfully contended with a unique set of trials, tribulations, and trays.

Now we have an educational task ahead of us. Like the post- 1956 portion of the NUC, the Pre-1956 Imprints portion assists librarians in acquisitions, cataloging, bibliography, interlibrary loan, reference, and research. But its potential uses go far beyond the walls of any single institution, whether it be library, university, book- store, or think-tank. As Gordon Williams commented in 1968 when the first volumes appeared, the publication of the retrospective volumes of the national union catalog vastly increases the historical and scholarly resources available for research in this country and around the world. Their publication enables any "library, wherever it is, to locate promptly and to provide its patrons with access to the millions of volumes it could never afford to acquire and house in its own collection."

It is unlikely that any single, traditional publication will carry forward the pre-1956 imprints project. Since late 1977, when the project stopped incorporating newly received reports, all such cards have been stored in a separately maintained file in the Library's Catalog Publication Division. Decisions about how and where to publish the pre-1956 cards-that is, whether to mix them with the earlier entries, to maintain them as a separate section within each issue of the current NUC, or to publish them in some other recurring sequence under the current NUC umbrella-have not yet been made. How could the planners of the catalog have estimated the catalog's own role in stimulating further cataloging of backlogged pre-1956 material by participating libraries? How could they have known that the automation of cataloging would result in libraries submitting yet another generation of cards for a vast number of items already reported? Traffic in pre-1956 reports is still heavy; a large number of new titles continue to surface.

Automation has now revolutionized union catalog production, so it seems safe to say that the Pre-1956 Imprints NUC will be the last of the monumental, traditional book catalogs. Future catalogs of this size, if there are any, will surely be issued in a different for- mat, perhaps in microform only, as is likely for this catalog once the remaining sets are subscribed. The ability we now have to store data generated locally and to share machine-readable data has already greatly simplified the manual exercise of producing catalogs of national holdings. It has also broadened coverage capacities. In fact, to make efficient use of the new technology, the Library of Congress is planning to automate The National Union Catalog and publish it in a register/index format expanding its coverage of its other library book catalog programs, such as the Chinese Cooperative Catalog,

the Monthly Checklist of State Publications, New Serial Titles, and the National Register of Microform Masters. This plan will require the cooperation of many other libraries and the bibliographic util- ities. Standardization is the key, for standardization of bibliographic records permits cooperative contributions to a national union catalog without loss of quality control, ensuring the usefulness of a record for many libraries. Given the fiscal restraints and limited resources of our times, we must continue to work together to make The National Union Catalog, in all its forms, a useful and affordable product for all its users.

Librarian of Congress Archibald MacLeish, in his preface to the 167-volume catalog of the Library's printed cards published in the 1940s by Edwards Brothers, speaks directly to the spirit of the achievement we are now celebrating:

What will touch the imagination of imaginative users (readers there will be none) is the fact that this enormous work is not merely a catalog of books, but a source book for the study of catalogs. It is indeed, if I may be permitted the respectful use of the metaphor, a kind of kitchen midden of American librarianship. Among the aca- demic clam shells here are the meaningful artifacts-the hopes and ambitions, the failures and the successes, of some very great Ameri- cans-Americans who are no less great because few of their fellow countrymen have heard their names. Charles Coffin Jewett, Ains- worth Rand Spofford, Richard Rogers Bowker, Melvil Dewey, and Herbert Putnam are not, perhaps, household names in the United -States, but they have done far more for the enduring life of their country than many whose first names and photographs are familiar around every wood-burning stove in the forty-eight states.

In 1981, as we celebrate the completion of the NUC in the fifty states, new names should be added to MacLeish's honor roll, including those of John Cronin, Gordon Williams, John Commander, Johannes Dewton, and many, many others. We are proud of you.

IN 1901, WHEN Librarian of Congress Herbert Putnam started the union catalog, the implicit assumption was that-except where librar- ies had pursued specialized interests-the main differences between collections were determined by the age and size of the individual libraries in which they were housed. It was assumed that all com- parable research libraries held in common virtually the same core collection and that it was essentially the older and larger libraries that were the repositories of books not generally to be found in the younger and smaller ones. This belief is implicit in Putnam's view, expressed in his annual report for 1900, that with the completion of the filing of cards from Harvard, the Boston Public Library, the New York Public Library, and a few others, the union catalog would "constitute the closest approximation now available to a complete record of books in American libraries."

The following facts, which many librarians still find difficult to believe, did not become clear until much later. Research library collections, even those of about the same age, size, and purpose, hold many fewer titles in common than everyone thought. Far more titles and editions are held by only one or very few of them. And, anything even approaching a complete record of books in American libraries requires a union catalog based on the holdings of hundreds of libraries.

When Putnam began the union catalog, the collaborative aspects of librarianship which we now take for granted were largely

________

Gordon R. Williams, former director of the Center for Research Libraries in Chicago, became chairman of the National Union Catalog Subcom- mittee of the Resources and Technical Services Division, American Library Association, in 1964. His unfailing guidance, support, and con- cern for the venture were vital to its successful completion.

undeveloped. Interlibrary borrowing had barely begun, and then only on a very restricted basis and with rules that were intended to keep it that way. In fact, the principal use made of the union catalog during the first twenty-five years of its existence was by Library of Congress catalogers looking for help in cataloging the Library's own acquisitions. Only a very limited search service was offered to other libraries wanting to know the location of books they needed to borrow on interlibrary loan. But by 1926 the need for interlibrary borrowing was accelerating, although the volume of this traffic was still too low to create pressure to spread the burden of lending through knowing more than one location. The inability of the union catalog to locate even one copy of needed titles, however, was of wide and growing concern.

Even in 1926, when the American Library Association sought and got a grant from the Rockefeller Foundation to expand the union catalog by adding cards from more libraries, the stated purpose was still only to locate as rapidly as possible at least one copy of every important research book in American libraries. During the five years of the Rockefeller grant, 1927-32, almost 6.5 million cards from other libraries were added to the two million already there, bringing the total to about 8.5 million cards-a more than fourfold increase in size.

The union catalog still proved to be inadequate. Over a decade later, in 1945-by which time the catalog had grown to nearly twelve million cards-it was impossible to provide a source for 30 percent of the titles libraries asked it to locate, and this rate did not even improve after another decade and more of additions. Indeed, as late as 1966, it was not possible to locate from the catalog some 22 per- cent of the titles about which inquiries were received. In addition to the inability to locate even one copy of a large percentage of needed books, by 1954 other problems relating to the catalog were beginning to assume serious dimensions. One was the growing volume of interlibrary borrowing-at least double that of 1945. Another was that the National Union Catalog, as it was by then called, could be consulted only at Library of Congress. This meant that other libraries needing to know the location of a required book had to send their requests for information to Washington. What with the time taken in the mail and in processing the inquiries in conditions of increasing pressure at the Library of Congress, a wait of two weeks or more for a reply was inevitable.

The obvious answer to these problems included bringing the holdings of still more libraries into the NUC and publishing it in a form that libraries could consult right in their own building. But the

idea of treating the historic record of American libraries' holdings this way was at that time too much to ask of library cooperation, and it was decided to start with newly published works. In 1954 the American Library Association Board on Resources of American Libraries formed a Subcommittee on the National Union Catalog chaired by Frederick W. Wagman. The subcommittee was to meet with the Library of Congress to see if other libraries' catalog cards of holdings of books with imprints of 1956 and later could be incor- porated into the cumulative book catalog of the Library of Con- gress's own printed cards, which the Library was then publish- ing on a regular monthly basis. At the same time more libraries were to be encouraged to report their cataloged acquisitions more comprehensively.

The Library of Congress proved to be agreeable to the wedding if the bride would provide enough dowry to pay for the increased cost. The ALA was able to do so, and in 1956 what had been the Cumulative Catalog of Library of Congress Cards, arranged by author, became The National Union Catalog, a Cumulative Author List of books printed in 1956 and later. To the individual entries under author were added symbols indicating which reporting librar- ies held the books described.

Consequent upon this publication of The National Union Catalog of recent and current imprints, the growth in libraries' reports of their acquisitions was astounding, reflecting both the increase in their rate of acquisition of new books and their more comprehensive reporting of these to the NUC. From 103,000 reports in the first year (1956), the number more than tripled to 326,000 in the next year. By 1960 about seven hundred thousand reports were being received annually and by 1966 the astonishing total of over 1.5 million reports a year were flooding in.

Meanwhile, and occupying an inexorably expanding area of the Library of Congress, the great National Union Catalog of books published before 1956 was growing also at an almost equal rate of increase in numbers of reports received. Attempting to satisfy read- ers' research needs, libraries were acquiring material more rapidly than ever before and reporting vast numbers of older titles that they had missed or been unable to find or to afford in the past. In 1959 the NUC subcommittee, then chaired by Charles W. David, decided to try publishing a part of the retrospective National Union Catalog. After much work, means were found to publish two years later about five hundred thousand entries for imprints from 1952-55 as reported on cards submitted by more than five hundred North American

libraries, including the Library of Congress. The success of this pub- lication indicated that given sufficient library cooperation and a strong and imaginative publisher, the whole National Union Catalog file of pre-1956 imprints-containing not just a few hundred thousand cards but, as was then underestimated, about thirty-two million cards representing about thirteen million entries-might be published in book form.

This great quantitative difference alone posed more difficult problems than any similar project previously attempted. First was the problem of financing, not only by the libraries who would ulti- mately have to pay for it by their purchase of the catalog but also by some publisher who would have to finance large editorial and production costs well in advance of any return income from sales. Next was the problem of the editorial work itself on so huge and so heterogeneous a file of cards. The cost of this led the NUC Com- mittee to consider carefully the possibility of publishing the cards without editing. But the variations in style and standards of the cata- loging inevitable in a record accumulated over more than fifty years from hundreds of different sources meant that works by the same author, and even different copies of the same title, could be so widely scattered through the alphabet that the usefulness and authority of the catalog for either bibliographical or location purposes would be greatly diminished. Further, duplicate and otherwise redundant entries for the same book would so increase the size of the published catalog that the cost of the additional printing and binding would, by itself, nearly cancel out any savings in editing.

As to who could do the editing, there was no question but that this could be done to an acceptable standard only by the Library of Congress. The Library, stimulated in part no doubt by the desire to get what was becoming an incubus out of the building, expressed willingness to take on the editing, but with the proviso that it would (and, by its constitution, could) only do so if the committee arranged to have the Library paid on a current basis for all of its expense in doing the job. It may be said here that the whole project, from beginning to end, was done without benefit of either govern- ment or foundation grants, though we badly wanted and tried hard for such funding.

A third factor arising from the quantitative difference in scale of the project was the long time over which the publication process was going to have to extend. This was inevitable because of the limits both on the speed with which the editorial work could be accomp- lished and on what libraries could reasonably afford to pay annually

for a publication which was going to make a noticeable dent in the purchase funds of many of them for whom the catalog would be most useful. This ten-year publication period meant that very careful consideration had to be given to the possibility of significant future changes. The committee was well aware of comparable long-term publishing projects where failure to foresee possible changes clearly enough had resulted either in the projects foundering before com- pletion or in their completion being delayed by many years-catas- trophies that the NUC Committee was determined to avoid. In retrospect the committee did remarkably well.

On the question of form of publication the committee had to make an immediate decision. Should the catalog be published in printed book form, in microform, or in machine-readable form for computer manipulation? After much debate we were forced to the same conclusion with respect to the machine-readable form as every- one else, then or since: we could not afford the cost of converting such an enormous back-file of cards. In spite of the astonishing advances made in the technologies and methods of computerized bibliographic control in the past fifteen years, the soundness of the committee's decision has been proved. The catalog exists, and were the decision being made today, the conclusion would still be that, for the foreseeable future, major existing files of historic material would be frozen in their present form and only new acqui- sitions, or entries newly made, would be put into machine-readable form.

The committee was also painfully aware in the mid-1960s of the lack of any generally accepted standard for a machine-readable bibliographic record, and especially of an international standard. This was critically significant for the NUC since the content and value of the catalog related to the needs of the library and scholarly communities worldwide. It was also clear that the economics of the project would require the support not only of American libraries through their purchase of the catalog but also of foreign libraries, for whom a machine-readable form would, it was judged, be unus- able. This judgment about sales has in the event been more than borne out for the number of North American sales of the catalog to date has been substantially less than was anticipated and indeed than they should have been. Only the larger-than-expected sale to foreign libraries has kept the cost to all libraries of our national union catalog down to an acceptable level and has permitted its completion on time and without lowering editorial or production standards.

As this brief history indicates, what was begun as a union cata- log intended only to serve the bibliographic needs of one library was gradually extended in both scope and availability as more and more libraries increasingly found that they could not meet all the needs of their patrons on their own and that a national union catalog was essential to effective interlibrary cooperation. In addition, publica- tion of the catalog has revealed a largely unrecognized and certainly underexploited mine of bibliographic information whose richness is only now becoming fully apparent to those accustomed to working it-whether librarians themselves in all their specialties or the users of research libraries in all theirs.

But the utility of the National Union Catalog does not stop there. It is essential to the answering of two other problems which cannot much longer be ignored without catastrophe. One of these is the problem of the rapid physical deterioration of an ever larger proportion of libraries' present collections. Unless coordinated action is taken soon to preserve these vast quantities of printed matter, most of it of relatively recent publication, much will be lost forever. The other problem is the urgent need to expand the resources avail- able to every library by a greater coordination of acquisitions, by the optimum use of purchase funds, and by a wider and more efficient sharing of material, both nationally and internationally. The develop- ment of coordinated acquisitions policies requires that each partici- pating library know what its colleagues already have from the past, what new publications they have acquired, and where the particular strengths and weaknesses of their holdings lie.

The committee was also acutely aware of the need to coordinate preservation activity and knew that successful coordination would depend on the ready availability of a national union catalog of pre- served books. Knowing that in the not-too-distant future such a catalog would-or at least should-embrace every book in the present NUC, the thought that their successors might have to go through the ordeal of publishing all the information in the seven hundred and more volumes of the pre-1956 NUC all over again appalled the com- mittee. To avoid this necessity, the committee arranged that every title recorded in the NUC should be given a short but unique identi- fying number by which it could be cited and easily and quickly located in the catalog. The NUC of preserved books, once that much-delayed process of preservation gets under way, could then be reduced to a compact catalog of simple NUC numbers with designation of location and preserved form, instead of a full biblio- graphic description. The unique identifying number for each biblio-

graphic entry, particularly when it has been applied to so large a body of data, has other potential uses as well, which have not yet been much explored.

It will, of course, take far more than The National Union Catalog to resolve preservation and the many other problems that face those trying to maintain the standards, integrity, and traditions of research libraries. But no practical solution of common problems is possible without a national union catalog. Its publication is timely and we are fortunate that, with the catalog now complete and widely avail- able, the basis for effective action is at hand. The continuation of the NUC for current and future imprints lies almost certainly in the domain of the computer and probably in the technology of the laser- readable disc. But whatever form it may take, this continuing catalog will depend, in meeting the needs of research libraries and their users, on the base of The National Union Catalog, Pre-1956 Imprints.

ON MANY OCCASIONS I have had the opportunity of talking about the background and functioning of the National Union Catalog Publication Project: its prehistory, the various steps leading to our trans-Atlantic bond with Mansell, the launching pad and early flight problems, the settled but always unsettling fourteen-year routine of editing with one eye on the cards and one on the clock. I have especially had occasion to give somewhat detailed descriptions of the editing process itself.

But on those occasions never did it occur to me that our work would really one day be done, that the project would be past history and not the continuing, frustrating, and yet profoundly satisfying activity that held us in willing thrall for so long. But it is done-the last "Z" (or "Zed") supplement card, a work by Zyndram- Koscialkowska, Wila, was turned on January 12-and no longer can I speak as one immersed in a gigantic effort, one with just a touch of uncertainty nipping at the fringes.

That uncertainty had to do with many things, ranging from our own abilities to cope with the ever-increasing deviousness lurking in the cards themselves to a strong concern for the economic health of our paymaster, our agency, our country, and our subscribers worldwide. It served us well in the end, for the further we got the more determined we became that, by God, we'd do our part to see that "Z" was reached, and twice at that.

________

Since 1975, when he succeeded Johannes Dewton as head of the National Union Catalog Publication Project, David A. Smith has devoted most of his waking hours to the catalog's successful completion, in the process becoming the Library of Congress's closest equivalent to James Murray, the longtime editor of The Oxford English Dictionary. Before joining the project staff in 1967, Mr. Smith worked at the National Library of Medi- cine and in the Descriptive and Shared Cataloging Divisions of the Library of Congress.

Now, rather than rehearsing again the details of the project, I would like to concentrate on some of the more compelling aspects of the project's inner editorial workings, at least as reflected in my own experience.

The most beautiful room in Washington is surely the Main Reading Room of the Library of Congress. I need not rhapsodize about it; its beauty is apparent. But how can one fully appreciate it without having seen it before the restoration done in 1964-65? Less than twenty years ago, the glorious interior of the dome was obscured by the accumulated grime of six and a half decades. This is my analogy for the work we have just finished. How can one fully understand and appreciate the finished catalog without having experienced it in its earlier state?

The national union catalog concept, from the first, was a majestic one. Yet the very growth processes which kept enriching and enlarging it were simultaneously begriming it. Certain problems were present almost from the start. Changes in the cataloging rules (the first of which occurred in 1908) were, for us, a bete noire trans- formed into a raison d'etre. I hesitate to complain about this because, like tooth decay for a dentist, such changes legitimated and even necessitated our existence. Also present from the outset were the bewildering number of idiosyncratic schemes, conforming to no known or recognizable descriptive cataloging rules, of the earliest major contributors. Princeton's "one-line slugs" and the American Antiquarian Society's "Enter under subject, neglect author entirely, omit at least first three words of title" cards spring all too readily to mind. (Be it noted that later cataloging from these sources was of high merit indeed.) And beyond rule changes and idiosyncratic schemes, the degree of completeness of description and accuracy of cataloging among the various contributors differed tremendously. Cards ranged from exquisite to abysmal, from descriptive bibliog- raphy in full flower to subminimal impressionism at its most laid back. (As Johannes Dewton so well put it, cards giving you their approximate rather than exact telephone number.)

The more subtle factors leading to disparateness arose from using competent catalogers around the country, who started from the same base but who viewed and recorded identical materials with quite different eyes and hands. In these same broad arenas we con- tinue to do battle today, striving to get things right at last, to put the definitive cataloging rules in place. Always the problem of choice and form of entry recurs: author against title, title against editor or corporate body, real name versus pseudonym, complete name versus

abbreviated name, original author versus commentator or adapter, or Latin form of name versus vernacular form. Basic problems arise from differences in the rendering of collation and imprint or con- ceptual differences in the treatment of musical works and the works of graphic artists. Particularly problematic are works resulting from legal actions, proceedings of meetings, works involving the praeses/ respondent dichotomy, and treaties. In fact, anything other than the very simplest case-a single, uniquely named known author of a single simply titled monograph-yields a great variety of catalog entries, reflecting the seeming determination of catalogers every- where to imaginatively and creatively list and describe their libraries' treasures. And, adding further confusion, libraries over the years submitted generations of cards for the same item, as either catalog codes or local emphases changed. The problem of multiple cards from the same library for the same item stayed with us right to the end and even increased as many libraries, in tune with the new era, contributed large batches of machine-generated records to replace their earlier handcrafted ones.

These then, were some of the major components of the grittiness obscuring a good view of the National Union Catalog. The frame- work was in place, the materials at hand, but what a job to be done! The Library's Main Reading Room was shut down for over a year, bright red scaffolding extended to the highest reaches of the dome, and everyone marvelled when it came down at last. To those of us who know the catalog well, the transformation has been as remark- able. Perfection, of course, was never the goal, and as its compilers and continuous users we are certainly more aware of the catalog's blemishes than are its other subscribers. Our charge was to produce a well-edited catalog, given the monies at hand, and we feel we have accomplished that end. The catalog's beauty, its detail and intellec- tual sharpness, the accuracy of its unique listings according to the old-fashioned principles, the structure of its alternative approaches (i.e., its added entries and cross references), the integrity of the sup- plement sequence of sixty-nine volumes with the 685 main sequence volumes-these things our editors, perched on their editorial scaf- folds, have persistently and tenaciously and skillfully brought about.

The technical problems we encountered were certainly far more difficult than had been foreseen. I have often wondered if the project would have been undertaken had the true state of affairs been grasped on all sides. The early production projections were very optimistic. John Cronin could edit four times as fast as the fastest editor, through the generous use of rubber bands, paper clips, and

notes to subordinates suggesting appropriate action. During the frenetic period of the project's earliest days, editors trooped to the Processing Department office, cradling a tray and other necessary implements, to sit at John Cronin's side, nervously flipping cards, cigar smoke enshrouding the scene. From these early sessions came the policies and procedures needed to accomplish realistic produc- tion goals.

As great and difficult to assess as the technical problems were, the very early correspondence explaining them to John Commander is in retrospect amusing. We tried hard to account for the unex- pected difficulty of the patch of the alphabet just done and to give assurances, not always heartfelt, that the coming patch would be more agreeable. Even more difficult to understand initially was the importance of the staff, especially the editors themselves. Unlike the staff of The Oxford English Dictionary, another purported ten-year project (which, however, extended to more than forty-eight), our personnel were many. I read K. M. Elisabeth Murray's wonderful biography of her grandfather James Murray, Caught in the Web of Words, with an immediate appreciation and empathy few can have.

Our enterprises were close kin, but the union catalog had a total staff of about fifty-five, of whom twenty-five were assistant editors and three associate editors. Together with the project's head and assistant head, the associate editors quickly reviewed the work of the assistant editors as the catalog's final quality check. The unit of work for each editor each week was a tray of about fourteen hun- dred cards, narrowed down from a much higher number during a series of pre-editorial processes designed to remove obvious duplica- tion and call attention to potential conflicts. The content of each tray was provided by the random bounty of the alphabet. Unlike so much cataloging-related work at the Library of Congress, this was work for generalists rather than specialists. A tray might offer cards in any Roman alphabet language and cover all kinds of printed formats- books, maps, music, serials, and so on. Editors had to deal with everything-always against the clock, and there was no putting a card aside for future inspiration. Cards in an editor's hands one week were in London the next.

An editor, in confronting each card, carried on an internal monologue: Does the choice of entry appear to be correct? Does the form of entry appear to be correct? Are there other places in the catalog where cards for this author or item might lie? Are there cross-references or added entries that should be made linking these other possibilities with the card in hand? Is the filing appropriate

and correct? Will the card photograph or must it be retyped? And so on.

The art, of course, was to answer these questions quickly and accurately and move on, to recognize the troublemakers but not trouble over the vast majority of perfectly solid citizens. Even in a catalog as dirty at the outset as ours, I should add, most of the entries were correct "as is." Once editors had recognized an offender, and we had various rules of thumb to help do this, the art again was to find the quickest possible path to an acceptable solution, or, failing that, to know when to let go, to pass the problem on to one of the senior editors. Thus, the ideal editor would possess good tech- nical skill in all phases of descriptive cataloging, strong language abilities, a thorough knowledge of bibliographical and general reference sources, an active imagination, a sense of appropriate compromise, and lots of that most precious and elusive commodity, common sense.

That may have been the correct formula, but it was a prescrip- tion that was difficult to fill. Oddly enough, some editors with the strongest technical backgrounds in cataloging had a difficult time indeed, chiefly because of their hard-to-quench desire for, if not perfection, at least certainty. In addition, most were accustomed to a very structured setting, where every cataloging contingency must have a corresponding action plan, often fiendishly complex. By con- trast, we relied on only a very few basic ground rules and encour- aged editors to fashion their own working procedures. Some of the most successful editors, given our rather unusual requirements, were, rather than the cohort of retired catalogers John Cronin had envi- sioned, free-spirited but mostly well-behaved youngsters not yet fixed by a more ster regime. Still, there were successes and a few disappointments whatever the background. What became clear, and what I think is important for those conducting or considering future similar ventures, was the necessity of having at least a few passion- ately inspired and talented people to keep things moving.

Several important points about the catalog's construction and make-up should be emphasized. First, the catalog, at heart, is only as good as the information presented to it permits. The intervention of editors could often improve the catalog, but ultimate responsi- bility for questionable entries must reside with the libraries supply- ing them. One of my very frustrating problems, incidentally, was to prevent over-intervention at the editorial level. How often editors thought they knew better than the library that had supplied the card! One might change a perfectly good heading for the shakiest of

reasons, albeit with a sometimes stunning certainty and finality. (One of the most frequent reasons being, "But it's listed that way in the British Museum"-though that august catalog marched to a different tune.) Though the overwhelming number of alterations made were valid, I begged editors always to ask: "Is this change really necessary?," to put themselves in the place of the original cataloger and divine that individual's intentions in preparing his record. Especially dangerous were changes made to records of uniquely held items, since a change here would divorce the record as it appears in our volumes from the same record as it appears in the unique contributor's catalog.

A related and obvious point, but one sometimes overlooked, is that, except for a few cases involving Library of Congress items, we saw only cards, not the books themselves. Nor was there time or money for extensive correspondence or telephone calls, which could have resolved many puzzles.

Second, the catalog is traditional: a given author is represented in a unique fashion at one place in the file and all his works are listed at that point. Though American and Canadian cataloging rules changed over the years, this principle at least had been maintained. Only now, with AACR 2, has a quantum change taken place. I have difficulty in seeing how we would have coped with the AACR 2 generation of cards, with their apparently profound differences. It is true that we daily reconciled the incompatible. Whether we could also have taken in this additional layer of mechanically and philo- sophically divergent records is a problem with which I'm glad we did not have to grapple. The catalog was started-and finished-just in time.

Finally, the catalog, though well-edited and, we hope, internally consistent, is replete with necessary compromise. No preconceived notion remained sacred-the sanctity of the Library of Congress printed card, the authority of the ALA cataloging rules, even the principle of unique listing-these and all other rules and principles were sacrificed to the simple need to get things done well but on time. When faced with progressive serial editing and unalterable deadlines, one must and does find immediate solutions, which are often compromises. It is my hope that our compromises were intelli- gent, helpful ones.

I have often wondered how the final compilers of the great cata- logs of the past-those we leaned on heavily in our work, such as Lorenz, Kayser, Pagliaini, and Palau-felt as their work was finished. Surely there is a sense of the usefulness, value, and magnitude of the

accomplishment. But I wonder if they too felt the inability to convey how things really went. Who, using the "New York" file, can know of the family illness experienced by the editor involved and the result- ing staff effort to overcome an unexpected and critical loss of time? Who, using the fourteen-volume "United States" file, can understand the enormous problems it posed or know that Hans Dewton worked so diligently on it following his retirement in 1975? Who can com- prehend the tremendous accomplishment of Maria Laqueur, the project's assistant head, in designing and starting the editing of the supplement while the last of the main sequence was simultaneously being roped in?

And finally, most difficult to express, who can know the inner beauty experienced in bringing order to the chaos? There was a serenity and joy here that overcame the frenzy and frustration that so often seemed to dominate. On occasions when the related but extraneous cares of the day could be discarded and when one was, to use Arthur Ashe's phrase, "in the zone," when making the catalog was all that mattered, the reward was great and sufficient.

Our work is done, the scaffolds are down, the results are avail- able for scrutiny, and, we hope, a little admiration. The project staff is indeed proud of its role in transforming the vision of others into a grand structure, the finest effort yet to document man's written record.

As THE PUBLISHER of The National Union Catalog, Pre-1956 Imprints and as one who has been closely involved with its publication from earliest days, I can only acknowledge astonishment at how fully and how closely the hopes and intentions of fifteen and more years ago have been fulfilled. We assumed that it would prove possible to set up an editorial unit at the Library of Congress that would per- form a task whose complexities were only guessed at and for which, as nothing comparable had previously been attempted, no precedent methodology existed. We also assumed that fully edited copy would be sent in weekly, air-freighted consignments to London at a rate sufficient to permit publication of five large volumes each of 704 pages each month. It happened. No shipment from the Library of Congress was missed over the whole period and, in spite of the hazards of commercial and industrial life throughout the 1970s, no year passed without a further sixty volumes being issued on time and in good order.

Of course this was not achieved without a fair share of crises- human, financial, and organizational. Even when the whole process had settled to routines which on the face of it made it all seem easy, day-to-day problems continued to tax the resilience, ingenuity, exper- tise, and sheer professionalism of the interrelated, but 3,500 miles- distant, ends of the operation. That this gap was bridged and never

________

At the symposium, members of the NUC subcommittee paid tribute to publisher John Commander for his vision, persistence, and scorn "for any expediency that would diminish the scholarly integrity or usefulness of the catalog." Mr. Commander, whose involvement with bibliographic publishing began with the British Museum General Catalogue of Printed Books (1961-66), has been associated with the Pre-1956 Imprints NUC for more than fifteen years, first as managing director of Mansell Pub- lishing, subsequently as head of Bemrose Publishing, and currently as the catalog's publisher.

seemed in much danger of opening into a gulf was due largely to the trust and confidence in which each held the other. And from the base of good professional relationships, friendships developed which have, for once, validated the now rather old-fashioned and some- what tarnished belief that ventures in international collaboration can prove constructive, fruitful, and even pleasurable. For once, not even the barrier of a shared language was sufficient to disrupt the single-minded identity of purpose and endeavor which from the beginning informed the whole project.

This collaboration element may be seen as one of the more remarkable manifestations of Anglo-American interaction in the fields of bibliography and library cooperation that has been devel- oped, particularly over the last quarter century. The pioneering work of the Library of Congress in instigating, under John Cronin's direc- tion, the rapid publication of major bibliographic records in book form pointed the way to the production between 1961 and 1966 of the British Museum General Catalogue of Printed Books and stimulated its publication. As the time, its 263 volumes were rightly seen as a major step forward in bibliographic control. The application of new and ingenious techniques specially developed for the project both permitted and required that the Catalogue be released from the stranglehold of editorial perfectionism and expense, which had reduced the progress of its publication to a crawl. The will to do this resided in the Museum Library. The means of its accomplishment were conceived and created by Balding and Mansell, a firm of English printers whose interest at this stage was in securing a substantial and continuing flow of work for its printing machines. However, Balding and Mansells initiative in developing the new equipment and techniques quickly led to a deeper involve- ment with the professional needs of libraries and with the concern of librarians to manage the exponential growth of both bibliographical information itself and the demands of its users. In this we were actively encouraged by John Cronin, who managed the application of the unique processes developed for the British Museum Catalogue to the increasingly urgent needs of North American libraries.

In particular, the possibility of putting the pre-1956 NUC into publishable form was raised and addressed. By about 1965 we had established to our own satisfaction the realistic possibility of convert- ing the sixteen million or so cards of the pre-1956 NUC into edited, book form. Prototype cameras, related equipment, and a feasible methodology of a much more sophisticated nature than that used for the Catalogue had been defined and successfully demonstrated. By

1967, through the good offices and under the control and guidance of the ALA subcommittee on the NUC, Mansell Publishing-specially formed to produce and publish the catalog-and the Library of Congress had struck the agreement. The juggernaut had begun its, as it was then thought, ten years' progress. In fact, the journey has taken fourteen years and its successful completion is due in no little measure to the thoroughness with which the ALA subcommittee plotted the route.

All of which would have been unavailing had the project not received the support of the library community at large. In many senses this was self-help: indeed, it was evidently desirable to trans- late into more usable and readily accessible form the vast resource of bibliographic information that had been accumulated over the years from the reports of many hundreds of contributing libraries. But even in the relatively affluent days of the late 1960s, it argued belief in the worthwhileness of the catalog and some faith in the likelihood of its completion to subscribe at a price that, even at that time, was likely to make a perceptible dent in the purchasing budgets of most libraries.

Fortunately, such conviction was lnot lacking and the hopes of the publisher and those associated with the project received early endorsement. At the end of 1968 some seven hundred libraries in North America were subscribing-many to more than one set-while from the rest of the world a further one hundred and thirty subscrip- tions had been received. But the pattern soon began to alter and it is, perhaps, a wry comment on the way the world has changed in the last decade that at the end of 1980 North American subscriptions still stood at fewer than a thousand while the rest of the world's had risen to three hundred and sixty-one. This relative decline in the propor- tion of North American subscribers-from 84 percent to 73 percent of the total-over the years is the one major disappointment that has attached to the project. It is evident that an unreasonably large number of North American libraries for whom the catalog provides a resource basically relevant to their own and their users' needs remain unequipped with a prime and vital tool of their trade. What- ever the reasons for this, it remains for the publisher a challenge, an opportunity, and a necessity to see that the remaining sets of this not misnamed National Union Catalog should find homes in North American libraries where they will best fulfill their purpose. At the same time, we as publishers together with the North American library community at large must be grateful to libraries, institutions, and enlightened colleagues in forty-nine other countries throughout

the world who have recognized the scope, authority, and utility of The National Union Catalog, Pre-1956 Imprints. Their subscriptions, paid very often from increasingly exiguous and hard-pressed resources, have afforded essential support in meeting the very heavy costs and risks of publication.

ANTIQUARIAN BOOKSELLERS are a minuscule fraction of the constit- uency served by the NUC. There are about fifteen hundred of us here in the United States, Canada, and Europe (including Great Britain). If we consider only those who consistently use bibliograph- ical reference works other than American Bookprices Current or similar pricing tools and who do a bit more than take an occasional hasty glance at Howes's U.S.iana or Brunet's Manuel, the number shrinks quite dramatically, and I think we end up with about 300 to 350.

________

An outstanding representative of the tradition of the scholar-bookseller, Bernard Rosenthal of San Francisco heads one of America's leading anti- quarian firms. Mr. Rosenthal is a member of the national advisory board of the Center for the Book.

States owns a set-one of them wistfully observed that "weight and space (rather than cost) considerations keep us from acquiring the NUC." As a matter of fact, the only privately owned set I know of is in England.

Next came a question aimed at determining my colleagues' chief reasons for using the NUC. Was it to determine proper bibliographi- cal entry, to check a collation, to locate the nearest copy, to get an idea of rarity by checking the number of recorded copies, to clarify edition or issue points, or some other reason?

Then, I was curious to know whether there is a tendency to use the NUC as a kind of last resort, after every other avenue-meaning consultation of the books in one's own reference library-has been tried. And I followed this up by asking whether the NUC generally provided the information one was after. Up to this point, I suppose my questions could have been asked of any user, not only of book- sellers. But the next two were directed specifically to the antiquarian bookseller: (1) If you find that a library doesn't have a book which, in your opinion, it should have, do you then quote it to them and mention that you are doing so on the basis of consultation of the NUC? (2) If you find that a book is "not in the NUC," do you raise the price?

At the end, I left a lot of blank space for "other comments," and I invited comparison of the NUC with other, similar large-scale bibliographical projects, particularly the British Museum General Catalogue of Printed Books.

I am going to outline the replies I received and give you some broad possible conclusions. But first a few figures will reflect the degree of response which I elicited. In the United States, where I mailed out 166 questionnaires, almost 75 percent of my colleagues replied-if nothing else, this proves the power of a self-addressed, stamped envelope. In Great Britain, about 50 firms got the question- naire, and slightly more than half sent in replies. In continental Europe, of the 90 dealers who were queried, also about half replied. I will comment primarily on results gathered on this side of the Atlantic and only occasionally refer to replies from abroad.

Regarding frequency of use, the largest contingent by far are the Occasional Users, i.e., the booksellers who, like myself, check the NUC less than once a week. The practical thing is not to consult the NUC just for a single title but to accumulate a number of problems and take a shopping list to the library at some point when we can absent ourselves from the shop. Right here, by the way, we have a key difference between the scholar and academician on one side and

the bookseller on the other. The former generally spend a great deal of their time in libraries or on campuses where consultation of the NUC is no different from normal library research routine. For book- sellers, especially those of us who have small staffs or none at all, consultation of the NUC means interrupting our normal shop routine and, in some cases, it gives us the choice of either closing the shop or working on a weekend-an awful choice, really.

After the Occasional Users come the Frequent Users, about 20 percent-and all of these, of course, live within easy reach of a set. I don't quite know what to do about the category who say they never use it. On paper, it looks like about 20 percent. Quite a few firms, both specialists and generalists, are amply served by exist- ing bibliographies (early music, for instance, or contemporary authors) or have a type of operation in which there is no time or need for bibliographical research. What I found rather astonishing was the fact that on the Continent, a number of prominent col- leagues of mine were not aware of the existence of the NUC or, when they were aware of it, did not realize that it was available at, say, the Bibliotheque nationale. The slightly guilty feeling some of us have for not consulting the NUC is beautifully summed up by a bookseller in London who confesses: "We never look at it-perhaps we should?" A little public relations work, in the form of articles in our trade journals, for instance, would go a long way toward alerting the members of our trade to the many possibilities which the NUC opens for us.

The replies to my next question, about the proximity to the nearest set, prove that Mr. Commander has done an outstanding sales job: the overwhelming majority of my colleagues, both here and in Britain and Europe, have a set of the NUC within a forty-five- minute or twenty-mile range. This is perhaps not all that surprising if we bear in mind that our profession tends to be concentrated in urban areas, where large institutional libraries are located. By the way, I note that almost all those lazy colleagues of mine who never look at the NUC do have one within easy reach.

If we look at the question about reasons for using the NUC, the replies tend to be rather predictable for the trade. At the top of the list are those whose chief motive is to get a sense of rarity by check- ing if a title is in the NUC and, if it is, how many copies are recorded. (This is both predictable and amusing, because one of the features of the NUC which my friends consistently criticize is the record of copies, which they say is weak and undependable.) Almost equal enthusiasm is voiced by those who need to clarify edition or

issue points which, of course, greatly influence the price. Rather further down the line are those who need a collation, and still further down are those who want to check proper entry. That checking an entry is near the end of the list is, of course, typical for us book- sellers: most of us have never heard of Anglo-American Cataloging Rules 1, let alone 2, and we have our own ideas about entries-I defy anyone to find a bookseller who puts Breviarium under C (Catholic Church. Liturgy and Ritual. Breviary.), as he would find it in the NUC.

At the very bottom of the list are those who are interested in finding the location of the nearest copy-again not surprising when we bear in mind that we are interested in books we can sell, not read. We now come to the question of whether we tend to use the NUC as a last resort. Antiquarian booksellers with bibliographical expertise and well-defined specialties tend to have their own refer- ence libraries. I do, and others like myself tend to exhaust all their "in-house" means, including the Encyclopaedia Britannica (eleventh edition, of course), before leaving their four walls for a pilgrimage to the NUC. The replies I received show that there is a sharp division between the Frequent Users and the Occasional Users. Over half the Frequent Users do not consider the NUC a last resort, and let it be said, in their honor, that these are, by and large, also the booksellers who have excellent reference libraries of their own. Among the Occasional Users, the majority (65 percent) do consult the NUC as a last resort, after their own tools have proved inadequate.

Coming now to the next question, whether the NUC provides the information sought, it is heartening to note that those who con- sult it most frequently are also those who are most satisfied. A large majority, close to 80 percent, answered "yes," and the rest gave answers which boiled down to about half the time; not one gave an outright "no." Among the Occasional Users, the answers were more lukewarm-fewer than half are fully satisfied, some said no, they are not satisfied (8 percent), and the remainder are in the fifty-fifty category. I think these results can be given a very optimistic inter- pretation. Having looked through all the two hundred-or-so replies which I received, I found that the most frequent users of the NUC were also those possessing-at least in my view-the highest degree of bibliographical expertise. In other words, they are the dealers who know what questions to ask of the NUC, who know how and when to use it, and who also have the best perception of its strengths and weaknesses-an 80 percent satisfaction rate in this hard-to-please and highly opinionated constituency is quite a compliment!

So much, then, for the questions of general validity. Now we come to the two which are applicable to the trade only, and here I find that fewer than half of the NUC users among my colleagues- about 40 percent both in the United States and abroad-us it as part of their sales technique. To some extent, this may be a result of our perception that the location of copies is a rather undependable fea- ture of the NUC and also because there simply hasn't been enough time yet to get to know this giant well enough to exploit it fully. The NUC could be a most valuable tool in collection building if the dealer had a good profile of a specific library's needs and policies, and could then quote books he or she felt the library should have. On a practical level, I think that such a collection-building program could best be worked out by librarian and dealer getting together for a thorough exchange of views. As a dealer, I would like to know, for instance, what the librarian can tell me about the dependability of the location entry as it relates to his or her own library-the greater the dependability, the smaller would be my risk of wasting my time in needless quotes.

The final, and somewhat brutally frank question-"Do you tend to increase the price of a book if it's not in the NUC?"-elicited some comments very much worth quoting. First, a German colleague of mine says, with a definite tone of disapproval: "We never raise our prices. Do you?" Second, a dear friend in New York says: "Yes-I raise my prices on the slightest pretextl" Third, a comment from Switzerland: "This is a good idea. I hadn't thought of it before. Thank youl" My severe statistics show that, in the United States, 45 percent say yes, they do raise their price, and about an equal number say they don't. Allow me to give these figures my own inter- pretation: the 45 percent who say that they raise their prices are telling the truth, and those who say they don't, aren't.

We now come to the relative merits of the NUC and the British Museum General Catalogue of Printed Books (BMC), or other com- parable large-scale works. Booksellers expressed all sorts of opinions, and it's quite amazing how passionate people get about such matters. Not surprisingly, the frequent users have the most strongly held opinions-opinions which are often diametrically opposed, some pro- claiming that the NUC is incomparably superior, others defending their beloved BMC with equal fervor. Quite a few of my colleagues have expressed the opinion that it's really unfair to compare the two, because of their obviously different purpose and nature, and I tend to agree with them.

A German colleague finds that the greater consistency of con-

tent in the BMC is, obviously, due to the more consistent acquisitions policy of a single, large library. An English friend of mine tells me he finds the NUC exciting because of its newness ("It has added zest to my life," he says!), and another prefers it to the BMC because of its "fuller entry formula, more logical indexing, and indication of rarity." This same dealer, incidentally, says he uses the NUC as a first, not last resort, even though he has a very fine reference library of his own. I suppose the truly wise men are those of my colleagues who state that, for best results, the NUC and BMC should be used together, one of them calling them "both superb" and then quickly adding that "the NUC, however, is more comprehensive."

In one case, the reply came in the form of poetry, and since it seems to sum up the feelings of so many of my respondents, I might as well quote it in full. Its author is my cousin Fiammetta Olschki- Witt, who runs her business with her husband Mario from a rather pastoral setting, a remodelled farmhouse in the Toscana, close to Arezzo but very far from the NUC:

NUC and BMC Are puzzling to the likes of me Amateur in bibliography. But of the two, if I must choose (A fate one often can't refuse) I find for practicality There's none that beats the BMC.

Let me take up the theme of practicality at this point. The BMC is available in the Readex Microprint Edition. For conceiving the idea of issuing this microprint, Albert Boni deserves an honored place in the annals of bibliography, quite aside from the tall pedestal he has earned for himself in the history of publishing. And since this microprint edition is available at a price which is within reach of a number of booksellers, quite a few have it on their shelves. We use it constantly, regardless of our specialties. We have become familiar with it, we know its weaknesses and strengths, and we know pre- cisely when to curse it and when to hug it. The effect of this Readex Microprint Edition of the BMC on the accuracy of our catalogs and on the dependability of our descriptions is difficult to quantify, but I daresay that it may be comparable to the effect of Hain, Brunet, or Sabin in their day. You see, the existence of bibliographical informa- tion alone is one thing, but easy access to it on one's own shelves is quite another. Not surprisingly, then, many colleagues have expressed the desire of seeing the NUC in a similarly accessible-and affordable-format. If it were technically possible to produce it in

microfiche or micro-whatever, as long as it is reasonably easy to use and costs less than, say, three thousand dollars, there would be a fairly sizable market for it in the antiquarian book trade.

An often-voiced criticism was a purely physical one: the gold- tooled spine titles, so goes the complaint, are hard to read, especially when seen from an angle. Since in most libraries, including the Uni- versity of California at Berkeley, half the NUC is below hip-level, I find myself quite often in the same undignified position as my col- league who must search for a particular volume on all fours, "using a hand as a light screen." Two American dealers, both enthusiastic users of the NUC, would like to see it used as a basis for a new edi- tion, with elimination of duplication and more standardized catalog- ing. To which I would add that, if such an immense project were indeed contemplated, more libraries should participate more consistently.

I would like to conclude on a note of enthusiasm commensurate with the completion of this awesome undertaking, and I will quote verbatim what two of my most respected colleagues, both of them in New York, and neither of them given to hyperbole, have written on their questionnaire: "I enthusiastically support the idea of the project (despite its shortcomings) and congratulate its editors on completing the most voluminous reference work in the history of bibliography in such an incredibly short time," and "I consider the NUC the greatest bibliographical achievement in my time." To which I can only add, Amen.

FOR MANY YEARS NOW, European libraries and those who frequent them have become used to the American scholar, better equipped than the home-grown article both as regards time and money and almost alarmingly familiar with the printed catalog of the institution in question, or at least with the printed literature on that part in which he happens to be interested. The two first characterizations are now less true than they used to be, but the haphazard way in which English scholars acquire their familiarity with sources is still put to shame by the overall efficiency of the American academic information retrieval apparatus. But as the great green buckram wall of the NUC grows brick by brick on European library shelves, European scholars are becoming more and more aware of the vast size and usefulness of the sources now revealed to them. In part, these are specifically American sources. That is, with all the biblio- graphical resources open to us, we have failed to realize the existence of many American books and periodicals dealing with subjects that interest us that have been published over the last century. Equally, however, the NUC is a vast testimonial to the wide and catholic interest in American libraries in a whole range of subjects of more European than American interest, stretching back well beyond the discovery of America. This resource is fascinating enough as it stands, and most European scholars will accept what is offered in

________

Nicolas Barker is head of Conservation at the British Library, editor of the Book Collector, and author of many distinguished works on the history of books and printing, among them Stanley Morison (1972), Biblio- Lindesiana (1977) and The Oxford University Press and the Spread of Learning (1978).

To me, both these questions are fascinating. The history of book collecting in America still remains to be written, although a formida- ble amount of source material is already in existence. I restrict myself to three examples. First, there is Edwin Wolfs history of the formation and catalog of the library of James Logan of Pennsylvania, remarkable enough in its own right and doubly remarkable as that of a man who was determined to remain American while passionately curious about the latest intellectual movements in Europe. Secondly, there is the collection that William Rainey Harper, first president of the University of Chicago, bought, in the teeth of opposition at home and abroad, from the great Berlin booksellers Calvary. It repre- sented a cross-section through the learned publications of four cen- turies of European scholarship in almost every discipline, whose quality is only now coming to be noticed. Thirdly, there is the Library of Congress itself. Incorporating the papers and libraries of the men who made the new republic, the Library of Congress has been built up on the lines of the great deposit libraries in Europe, and with the addition of special collections of older books such as the Rosenwald Collection, it may be said to typify the book collect- ing instincts of the United States.

This is to leave out collections of such first-rate importance to European scholars as the Harvard and Yale libraries, which are as old as many European institutional libraries, the John Carter Brown Library, itself the earliest as well as a model thematic collection, and the great fonds of modern literary material now at the Humanities Research Center at the University of Texas. All these resources are represented in the NUC. All are, to some extent, exceptional. But they may be held to typify the extraordinary number of libraries all over the United States of America that have been collecting books old as well as new, from Europe as well as America, over the last century and in some cases even longer. European scholars are now beginning to learn, thanks to the NUC, how widespread as well as how rich the bibliographical resources of North America are. The importance of Kansas to the student of the English eighteenth cen- tury, the astonishing growth of Brigham Young University, the spe- cial contribution of Canada-all this is reflected in the NUC. It will even tell you that the Folio Plautus annotated by John Milton is at Loras College, Des Moines, Iowa. What can be read is now revealed to trans-Atlantic readers and the relative cheapness of the dollar against most European currencies and the huge reduction in air fares make it much easier than it used to be for them to cross the Atlantic.

What will their impact be upon the natives? How will these new conquistadores mine the newfound gold?

The first of these questions is rather easier to answer than the second. Speaking as a rather lately joined member of the corps of the British Library, whose own catalog, in a modest 263 volumes, came out relatively recently, I find it possible to offer a little advice which may save, if not the lives, at least the time of the future Incas and Aztecs who are next on the list. To be received into an American institutional (or private) library is, for the European scholar, to go back to the nineteenth century. Material is made easily available to the scholar. He himself is made comfortable, allowed to come and go, given (if he is a long-stay resident) a table and a locker of his own, even in some cases allowed to search the shelves himself. More than all this, the time and knowledge of the staff is made generously available, so generously that the scholar is all too frequently apt to forget that the staff have anything else to do. The British Museum was already far from this primal state when the publication of its catalog opened its resources to the hordes who came not merely from the Western Hemisphere but from Europe and even Japan. What had in effect been a club of some thousands of members became an international society measured in millions. Books which the club had allowed to slumber gently for two hundred years were pressed into active service. Staff members from the director down to the library assistants found themselves facing a new volume of work generated by an ever-increasing number of new readers.

There are four lessons to learn from this. First, make sure that the external and internal services of the library are clearly defined and performed if not by different people (which would be absurd), then in a way which will enable the same people to perform both functions adequately. If a library has two main duties, to conserve its collections and to make them available, the neglect of the essential internal business of keeping the organization going will be damaging in the long run to both, as damaging as the preference of either duty to the other. Second, make certain there is a procedure for dealing with readers' wants, not just a generous willingness to fit in with their requirements. Some provisions are general, from adequate security arrangements (which mean much more than a casual inspection of briefcases at the door) to a regular updating of open- access reference books. Each library will have its own recipe here. More important, however, is specific provision: catch the reader first and find out in detail what he wants, rather than deal with the much more complicated problems he may generate if he tries to extract

what he wants without assistance. Some sort of reception service, even perhaps a detailed questionnaire, is a valuable precaution.

Third, watch the orders for photography. To publish your cata- log is to invite orders for microfilm from afar. It is not unknown for a German scholar to order a microfilm of a German fifteenth-century book from the British Library, writing from a town which possesses a copy of the same edition of the same book. Finding out why the order has been placed and if need be rewriting it to suit the library's convenience rather than the applicant's becomes increasingly impor- tant. Otherwise, your books will become reduced to the present piti- ful state of the British Museum first copy of Foxe's Actes and Monu- ments (1563), of which single pages were photographed over and over again for the woodcut illustrations, a piece of vandalism which if anticipated could have been prevented by a complete microfilm of the whole book. Lastly, there are the ever-present and ever- increasing demands of conservation. Books that are more used decay more rapidly: a book that is only slightly the worse for wear can be put in good order at a tenth of the cost in time and money that will be involved if it is allowed to stay in heavy use for only a year or two more. So, before the hordes arrive, check your stocks and see that they are in good shape: time and money spent now will save an infinity of woe and expense later.

All this would be so much easier if the direction of the demands were predictable. What will the marauders from across the Atlantic want? Here I am afraid I can offer no counsel or consolation. Succes- sive, persistent, and prolonged inquiries at the British Library into the requirements and reading habits of readers has merely reinforced the pretty general conclusion that every reader wants something different and that the past is no guide to the future, even on an indi- vidual, let alone a general, basis.

So I will end on a personal note by recounting my own exceed- ingly happy pilgrimage through a small section of the NUC. As is well known, the scope of Carter and Pollard's famous Enquiry into the Nature of Certain Nineteenth Century Pamphlets, published in 1934, was materially enlarged by the discovery, shortly afterward, that Harry Buxton Forman, the distinguished editor of Shelley and Keats, a man seventeen years older than Thomas Wise, had been the knowing accomplice in the crime of forging first editions of the minor works of major nineteenth-century writers. The discovery of a pamphlet by Richard Henry Home, Galatea Secunda, bearing the imprint "Melbourne, 1867," but with part of the edition at least printed on paper watermarked "1873," put Forman's role in a new

light. Forman had befriended Home on his return from Australia in 1869, had helped him get some of his later works through the press, and had become his literary executor on his death in March 1884. Indeed, the bequest of Home's library had filled his house overfull, and he was forced to dispose of "a portion" of his library at Sotheby's the following November. In this sale, lotted with five other items, and sold for no more than two shillings, was a copy of Galatea Secunda. Forman, then, was already involved in questionable deal- ings before he met Wise (1886) and before the emergence of the first forgery with which Wise's name can be associated (1888). Was Forman then the true originator of the fraud, and Wise the second- ary accomplice? It became imperative to study Forman's literary contacts and other publications, especially his relationship with Home.

The pages of the NUC devoted to Home provided the answer. Home occupies some five pages altogether and a substantial number of entries. Among these, defined by common location symbols, are a group of Home's pamphlets which were clearly associated with Forman. Letters to the libraries concerned soon revealed that all their holdings had come from a common source, Messrs. Elkin Matthews Ltd., who purchased the residue of the library of For- man's son Maurice Buxton Forman in 1946, issuing a catalog for which in several cases they were able to supply multiple orders from multiple stock. The deductions drawn from this were important. The absence of any trace of an intervening sale made it clear that Forman's enterprise in this case was not intended, as was the later partnership with Wise, to make money. Significantly, presentation copies of all the other Horne pamphlets have emerged, but none of Galatea Secunda. The implication, that he was ashamed of the mendacious imprint and did not wish Home to see it, was clear, and conveyed a vivid early impression of a conscience at work, which is conspicuously absent from Wise's dealings.

All these facts and insights derive from the scrutiny of a few pages of the NUC. The same scrutiny, and the realization that NcD and MH had the most extensive holdings of the suspect pamphlets, brought an extended correspondence with Duke University and Harvard, which has brought new friends and an abundance of other facts extending far beyond the narrow spectrum of Forman's rela- tions with Home. (I cannot help adding that the same five pages add a splendid piece of gratuitous information, in the twenty-seven cards listing that part of the notable Leigh Hunt manuscript collection at the University of Iowa at Iowa City that relate to Home.) I cannot

begin to be sufficiently grateful for the rapidity and thoroughness with which what would otherwise have been a lengthy and laborious search was concluded.

The NUC has brought me new friends as well as new knowl- edge. I do not think I am exceptional in this respect. The NUC will continue to absorb me and many others, to provide the essential background to a whole new range of scholarly research into every subject, period, and language imaginable. I am sure that others will be as happy in their experience as I have been. I only hope that the libraries who have given so generously to provide this resource will feel correspondingly grateful for the extension of knowledge which their visitors may bring them.

AS WE CONSIDER the manifold uses of the NUC, we may well be assured that this magnificent instrument of research will carry us far beyond the hesitant gropings of present scholarly endeavor. The incessant quest for knowledge impels the literary scholar to seek out the truth, first as it may partially reside within some book already possessed, then as it is further divulged in a local library, and eventually as it might be fully manifest in a vast encyclopedia of the printed word. Now that this seemingly unattainable ideal has been realized, essentially, in the NUC and is presented in a form immedi- ately accessible, the catalog stands, as David Smith declared, as "the best record we shall ever have of the first 500 years of man's written history."

To appreciate the immediate uses and limitless range of this ultimate catalog consider the constraints of an earlier time, from which my own frustrating experience dates. Long ago as an under- graduate I arrogantly declared that the three hundred thousand volumes then in my college library were quite inadequate for my purpose, whereupon I was advised to proceed at once to the Library of Congress and, if what I sought was not there, to examine the cumulative card file of other holdings. And so began my early initia- tion to the perils of scholarship, looming in mountains of cards at

________

the Library, many of them obviously deranged alphabetically or chronologically, and some quite hieroglyphic in whatever message they conveyed.

Given these and other deterrents, the only sustained research I then attempted was that of identifying all early editions of Matthew Lewis's The Monk (1796-1800), a notorious gothic romance not to be found at the Library of Congress but readily available in all variants only at the University of Virginia. So informed, I was soon on my way there and in 1949 issued my first publication under the auspices of that institution. That early I perceived that a national union catalog could direct me to various distant locales where some percipient librarian or dedicated professor had gathered together for intensive study the very books I was eager to examine.

Somewhat later the Library of Congress was able to expand its service, and so enlarge the scope of scholarly enquiry, by provid- ing on request a continuous 8-mm film of the entire file of any given author. The earliest film record I acquired, produced I was told by a navy war-surplus camera, remains among my memorabilia as a considerable advance over what had gone before: the reproduction of 961 cards on Edmund Burke, which I could view repeatedly if I desired, but always of course in an inalterable sequence usually at variance to my own intent. Nonetheless, within the narrow confines of an 8-mm film, one constantly falling out of the 16-mm projectors I used, my investigations could proceed with much greater dis- patch-so much so that later I ordered more and more film, the last roll on Mark Twain, representing 2,650 cards.

With all these awkward expedients now happily behind us we may observe in the expansive format of the NUC the marvelous opportunity of conducting "contextual" explorations. Here we are allowed immediately a certain panoramic view around whole ranges of books, perhaps forty or fifty at a time, all widely divergent in origin, form, or present locale but all classed together and now here assembled, typographically, for direct investigation. To exemplify this greatly enlarged perspective of the printed word let us wander afield momentarily to contemplate the productions of two nineteenth- century educators, both hardly recognized in the usual surveys yet, as measured intermittently in the NUC, having an influence quite surpassing those more often acclaimed.

The several manuals compiled by Lindley Murray, my first candidate, are duly noted in both the Dictionary of National Biog- raphy and the Dictionary of American Biography as having long been used in schools "to the exclusion of all other grammar books,"

totaling altogether, according to the best estimates, between 1.5 million and two million copies. That count, large as it may seem, can now on the warrant of NUC be regarded as a considerable devalu- ation, since it fails to account for the rampant piracies in this coun- try, all of which were beyond the author's purview. Of Murray's English Grammar and its subsequent Abridgement, the British Library Catalogue has only 82 listings of the authorized issues, the last a 54th "edition" of 1846, and the NUC several regular issues of still later date, the last a 110th "edition" of 1881, known only by the single copy at Columbia University. Besides these legitimate London and York editions, however, the NUC also records of the original Grammar about 170 irregular issues-most of them described as abridged, adapted, corrected, improved, or simplified-and of the later Abridgement no fewer than 154 entries, deriving from forty- one different cities in the United States. Of these the Boston version went through at least eighteen so-called editions. Certain others, from Bennington, Brattleboro, Burlington, Concord, and Pittsburgh, are said to derive from the twentieth English edition, the Utica issue from the thirtieth, and the Buffalo issue from the fifty-second. Obvi- ously this extensive though still partial record, as now made evident in the NUC, will confront the bibliographer with many complexities in classification and the sociologist with as many imponderables in measuring the impact of such a work upon the educational process.

The second grammarian I would bring to your attention is of no less consequence, though now he seems completely ignored. Henry Butter was once everywhere recognized as the author of the Etymo- logical Spelling-Book and Expositor, a book rarely located today and then only in single copies: editions 1, 43, 111, and 238 in the British Library Catalogue and 4,120, 209, and 353 in the NUC, with edition 209 there cited under two dates, 1856 and 1857. To this vestigial record I would now add my own unique copy of the 429th edition, of 1884, this with a lengthy preface lamenting piracies in America ranging over a million copies. Even among the authorized issues we are here prima facie missing 418 editions, enough to demonstrate that even the NUC cannot conduct us unerringly, book by book, through all the trackless wastes of the past. The unpredictable dic- tates of destiny as here inscribed may suddenly disclose everything we may desire about the literary activities of one person and as sud- denly deny us all but a fleeting glimpse of his equally important contemporary.

I refer to the Catalogue of the British Library as well as to the NUC in these matters, partly from my habitual practice of consulting

all compendious references and partly in the conviction that what is reported on one side of the Atlantic, whether of an English or of an American writer, needs to be supplemented by reports from the other side. Indeed the two accounts at times may be perfectly com- plementary. For Haydn's Dictionary of Dates, a standard reference among historians and even among a few professors of English, the British Library's Quarterly recently listed as on deposit the first edi- tion of 1841 but "no copy traced" of the second. The NUC indicates no copy of the first but two copies of the second, at Harvard and the American Antiquarian Society. For the first American edition, the Quarterly lists G. P. Putnam's issue of 1867, the NUC one by D. Appleton in 1866. The Quarterly reports nothing further on American issues, but the NUC goes on to cite from Putnam eighteen later editions and from Harper Brothers eight more in a rival series apparently of different content, designed "for the use of American Readers." Assuredly, as this combined report testifies, in the bound- less realm of knowledge some portion of it must be sought beyond the national frontier.

Given the hazards attending any venture beyond the limits of one's present expertise, I imagine that most scholars will be content to use the NUC not for escapades of the kind just recited but rather to confirm and further amplify a known circumstance in book pro- duction. As multitudes of imprints are registered in the Eighteenth- Century Short Title Catalog its compilers freely acknowledge con- tinuous recourse to the NUC. For similar controls over a more limited operation, the six thousand volumes issued by the nineteenth- century Leipzig publisher Bernhard Tauchnitz, the bibliographers thus concerned also resort constantly to the same record. Even for a single work it may be enlightening to specify at length the data con- tained in the NUC, as witnessed, for instance, in the elaborate account of Washington's Farewell Address (1796) in a catalog just issued from Atlanta, Georgia. Here one may quickly discover which among the twenty-four libraries there cited possess all three imprints (Huntington, John Carter Brown, Library of Congress, New York Public) and be further warned that, despite the varying declarations of Eames, Evans, Paltsits, and Vail, the priority of these imprints still remains undetermined.

Finally, in all such evaluations it is well to remember that, among the eleven million separate reports now represented, the NUC can usually record only the outward and visible signs of the book, not its inner construction. Many books appearing to be the same in all respects, and so enrolled with numerous location symbols in a single

entry, will eventually be disclosed to be three or four issues, one genuine, the others counterfeit. And as many more, here set apart according to variant titles, imprints, or dates, upon further examina- tion will be revealed to be a single edition repeatedly reissued. Occa- sionally pagination alone will indicate something approaching an international conspiracy, as observed, for example, in an 1876 Lon- don issue of Forster's Life of Goldsmith or in several 1889-90 New York issues of Jane Austen-all, I would now assert, deriving from stereotype plates prepared years before in Leipzig and all, I must add, known only by copies at Harvard. Without further analysis one may readily accept a report, from a major research library, that through 1955 Melville's Moby Dick ranged through 118 "editions." Upon proper investigation, however, one must conclude, with G. Thomas Tanselle, that all these NUC entries actually make up only thirty-five editions. Quite obviously, then, mere appearances can be very deceptive and, if unquestioned, may grossly distort the literary event under consideration.

In this brief survey of a reference work extending to 754 vol- umes, I am now rather startled to find, at least, that amid all my airy allusions to thousands and millions of copies, the books specifically mentioned usually appear only in solitary specimens. This circum- stance I can only proclaim as yet another and possibly the most dis- tinctive service of the NUC: that it often rescues and identifies from times long past certain isolated artifacts still to have some immeasur- able effect upon time present. What this and other effects may be, whether occasioned by one or by many books, should now be the constant concern of scholars everywhere. Allowing this as a primary rationale for research, it may be agreed that henceforth all academics still posing as scholars yet now admitting ignorance of NUC should be dismissed as imposters.