Skip to main content

The Case Of The Vanishing Records

March 2024
19min read


In Springfield, Illinois, a historian doing research on the Pullman strike of 1894 asks to see the letters of Governor John Altgeld for that year. When he opens the packet, the letters crumble in his hands: some of the most important documents relating to the historic strike are unusable. In New York a biographer writing what may be the definitive life of the great painter John Singer Sargent begins going through the public library’s rare collection of contemporary reviews of Sargent’s work and the catalogues of his exhibitions. The material disintegrates when it is handled: the pieces of paper that could prove which of the paintings attributed to Sargent are authentic—a matter of increasing controversy—are lost to future generations. A writer preparing a history of photography asks the Museum of Modern Art in New York to show him the early work of the pioneering photographer Edward Steichen. The museum replies that it does indeed have some early Steichen pictures, but that his most important early photographs—those taken between 1890 and the mid-1930*5—were lost in his Connecticut studio before World War II, when thousands of them exploded from spontaneous combustion. A midwestern college giving a special seminar on American life in the twenties writes to various film archives asking for Theda Bara movies. The reply comes back that only two of the popular star’s more than twenty films still exist. The college then asks for Dorothy Gish films and is told that a number of her movies are available, but none of the comedies that made her a star.

These are all warnings of an astonishing fact : at a time when we pride ourselves on the volume of the record we are leaving to future generations—and indeed the volume will be impressive—parts of that record are nevertheless in danger of disappearing before our eyes.

The groundwork for this very serious problem was laid in the eighteen seventies and eighties, when wood-based paper became the most common material upon which man printed his theories and thoughts, and when cellulose nitrate film became the substance upon which he most often captured his image.

It was the highly civilized Moors who, in the twelfth century, brought papermaking to Western Europe, and there are official documents of the Holy Roman Empire on paper dating from 1228, though three years later the Emperor Frederick II forbade the use of paper for vital records, on the ground that it was not long lasting, and insisted on vellum. In the twelfth and thirteenth centuries the art of papermaking spread throughout Spain, Italy, Germany, France, and England; by the fifteenth century paper was superseding vellum. The appearance in the middle of the fifteenth century of Gutenberg’s movable-type press encouraged this change; it brought about an enormous increase in the number of books produced and a consequent demand for a plentiful substance on which to print them. The change did not go unmarked by the scholars of the day. At the end of the fifteenth century Johannes Tritheim, the abbot of the monastery of Sponheim in Germany, lamented: “Truly if writing is set down on vellum, it will last for a millennium. When printing is on paper, however, how long will it last? It would be surprising if printing in a paper volume were to survive for two hundred years.”

The good abbot need not have worried, for the paper used in his time was remarkably long-lasting. It was made of ground new linen rags or clippings from garment making, stiffened and bleached by alternate rinses of sour milk and an extract of wood ashes, and then exposed to sunlight. The result was a mildly alkaline paper with a life of from three hundred to eight hundred years. Indeed, perhaps the finest paper ever produced was that used for the Gutenberg Bible. In the extant paper copies (some of the copies are on vellum), the paper is in nearly perfect condition after more than five hundred years. The first paper made in what is now the United States—that produced by William Rittenhouse and his son in Germantown, Pennsylvania, in 1690 and used by the pioneer American printer William Bradford on his press in Philadelphia—shows few signs of deterioration.

A major change in papermaking came in the second half of the eighteenth century, when the increasing demand for paper outstripped the supply of good new linen rags. To supplement the supply, yellowed and deteriorated rags were mixed in with the pulp; to make the discolored rags acceptable, various strong bleaches were used. The most common of these bleaches was chlorine, which can combine with moisture to form hypochlorous acid, a substance that ultimately weakens the fibers of the paper. Thus, by the early nineteenth century the movement toward low-quality papers had begun.

The invention of the cotton gin in 1793 had a beneficial effect on the quality of paper in the first decades of the nineteenth century. The sudden availability of new rags made of cotton—the purest natural cellulose fiber—ended for a time the need for harsh chlorine bleaches; and the paper manufactured from 1800 to 1850 was, on the whole, better than that manufactured at the end of the eighteenth century.

The decades from 1850 to 1870 were one of the great transitional periods in the art of papermaking, for the demand for more paper for books, magazines, and newspapers was causing a revolution. In the years just before the Civil War, the supply of rags—both linen and cotton—could not keep up with the demand for paper. (The search for rags reached such a frenzy that in the middle of the century the wrappings of Egyptian mummies were imported into the United States for papermaking.) The transition which took place at this time was described in 1891 by Rossiter Johnson, an editor, in an article that he wrote for the Library Journal :
… many American households kept themselves supplied with tinware by periodically exchanging the contents of their ragbags with the itinerant peddler who called at the back door with a spring balance in one hand and a shining new pie plate or milk pan in the other. … All that is changed. The constantly increasing demand for paper and the never-ceasing rage for cheapness stimulated the ingenuity of the inventor to try every possible substitute. The Civil War, with its blockade of Southern ports, made cotton so dear that in the second year (1862) common book paper rose to 22 cents a pound.

A substitute was at hand, and it was no accident that it came from the northern forests rather than from the southern cotton fields. The idea of making paper from wood had been suggested early in the eighteenth century by a French scientist, René Ferchault de Réaumur, who had noted that wasps use wood to make their nests and that the texture of the nests resembles paper. In 1800 Mathias Koops, an Englishman, demonstrated the feasibility of making paper from wood, but the suggestion was not adopted at that time, perhaps because of the increased supply of cotton resulting from the invention of the cotton gin. It was a German, Friedrich Keller, who read Reaumur’s observation and who, in 1840, patented a machine that ground wood into pulp for the making of paper. This groundwood paper quickly caught on in the United States and by the 1880’s became, both here and throughout the world, the substance upon which newspapers are printed. As early as January 14, 1863, the Boston Weekly Journal was using the new paper, and the next day it proudly announced that it consisted of “paper made of wood.”

Shortly after groundwood had replaced rag for news papers and other publications, a process was developed in England that allowed manufacturers to use wood as the base for a paper of much higher quality. This new process, which consisted of cooking wood chips in a solution of sodium hydroxide, was introduced into the United States in 1854, but it was not until the 1880’s, after Benjamin and Richard Tilghman of Philadelphia had made important improvements, that this chemical-wood paper was introduced on a commercial scale. With certain changes, this is still the paper used for most books, magazines, and scholarly journals.

Almost from the beginning, die short life expectancy of groundwood papers—as distinguished from chemical-wood papers—was noticed. In 1898 the librarian of Congress, John Russell Young, commented, ”… the question may well arise as affecting, not only our own, but all modern libraries, as to how much of our collections will become useless because of the deterioration and disintegration of the paper used in the cheaper forms of literature.” The problem was particularly acute for newspapers: a copy of the Western Advertiser, and Chambersburg Weekly , published on rag paper in Chambersburg, Pennsylvania, in the last decade of the eighteenth century, was in good condition, while a copy of the Chicago Tribune , published eighty years later and of far more historical value, was not. The New York Times recognized the seriousness of the problem and for many years published a rag-paper edition for library use; the practice was halted in 1953 by a combination of rising costs and a growing belief that microfilm was the answer to the problem of preserving newspapers. (Sears Roebuck for a while even printed a rag-paper edition of its catalogue.)

By the time that the Times ceased printing a daily run on rag paper, librarians, historians, and others who care about the printed record of mankind had discovered an alarming fact. Not only were newspapers and the cheaper forms of literature printed on groundwood paper disintegrating; much of the other matter printed on paper manufactured after 1870 was also in danger. The blame was promptly placed on wood papers of all kinds. But there was a puzzle. Some of the finest all-rag papers made since the 1870’s were also in a serious state of deterioration. Air pollution was blamed, as were the overheating and low humidity common in libraries, but none of the explanations was really satisfactory. Finding the real answer was the work of William J. Barrow, the documents restorer of the Virginia State Library, who had been studying the cause of paper deterioration since 1932.

In 1957 a grant enabled Barrow to conduct an investigation which showed for the first time what was actually happening to book paper in American libraries. He reported bluntly, “It seems probable that most library books printed in the first half of the aoth century will be in an unusable condition in the next century.” The situation was just as critical for quality magazines and journals. The maximum life expectancy of copies of many of the most important scientific journals, for instance, was discovered to be only fifty years.

The hard facts were there for librarians to see. When skeptics said that the volumes were still on the shelves and that the terms that the librarians were using, such as “disappearing books,” were exaggerated, the rejoinders were quick in coming. Gordon Williams, director of the Center for Research Libraries in Chicago, a collection of some two and a half million volumes used for advanced research by thirty-nine universities from Harvard to the University of California, made the point quite clear.

You ask what is meant by a book that is “disappearing.” This is one the paper of which has now become so brittle that you cannot handle the book, or turn the pages, without the paper breaking and crumbling. Such a book is essentially unusable. If the pages are not already too broken to use, the next person to read it would destroy it.

In seeking the cause of the problem, Barrow looked first at the wood base of the paper, for the appearance of short-lived paper seemed to coincide with the change-over from rag to wood pulp. There was no doubt that this was the cause of the disintegration of groundwood newsprint. Chemically, groundwood papers are half cellulose and half non-cellulose, and after fifteen to twenty years the non-cellulose materials break down into acid compounds and eat themselves up.

But what about the high-quality chemical-wood papers and the rag papers made since 1870? What was causing them to disintegrate? Here the answer was a surprise. Barrow took a hard look at the sizing used in papermaking. All papers made for writing and printing are “sized,” a term which comes from Old French and means “to set” or “to fix.” Sizing is quite simply a substance that is added to paper to prevent the ink from being absorbed and thus feathering over the surface. (A blotter is an example of unsized paper.) Until the second quarter of the seventeenth century, the sizing agent generally used was gelatin or glue made from the hides and tendons of animals. This is what helped produce the mildly alkaline, long-lasting paper of the time of Abbot Tritheim. But in the second quarter of the seventeenth century, alum began to be added to the gelatin to make the paper more resistant to ink. Alum is aluminum sulfate and is highly acid. Since papers manufactured after the second quarter of the seventeenth century tended to disintegrate faster than those made earlier, Barrow thought that he had spotted the cause.

The next critical test era was the 1850*5, when a new method of sizing, which did away entirely with animal gelatins and glues and replaced them with rosin, came into general use in the United States. This new alum-rosin sizing could be more easily applied than the gelatins and glues, and so was quickly taken up by paper manufacturers anxious to increase their output. Barrow made a thorough study of this alum-rosin sizing. It was known that alum by itself was acidic, but an investigation of the new sizing revealed something extraordinary. When the alum combined with the sodium resinate of the rosin, the resulting mixture decomposed into components of sulphuric acid that could burn up a book in thirty years.

To study the effect of this alum-rosin combination, Barrow tested five hundred nineteenth-century books, fifty published in each decade, and divided these up into three broad groups, those of 1800 to 1849, those of 1850 to 1869 (when alum-rosin sizing was first being introduced), and those of 1870 to 1899, when the new sizing became almost universal in papermaking. Even though some of the books of the second period were printed on paper made of the same material as those of the first—linen and cotton rags—they were in much poorer condition than those in the first period. The books from the last period were the weakest of all. Though there was no question but that the use of wood fibers that had not been thoroughly washed and the use of short fibers in the chemical-wood papers contributed to the weakness of the books published since 1870, the main culprit in the destruction of most paper was the sizing. Barrow stated the fact baldly: “The introduction of alum-rosin size contributed more to the deterioration of paper than any other development in papermaking.” Thus rag paper sized with alum-rosin was not really much better than wood paper.

In the ten years that have passed since Barrow made his report, the full impact of the highly acidic alum-rosin-sized paper has hit libraries and the scholarly community. Frazer G. Poole, the assistant director for preservation at the Library of Congress, says “It is undoubtedly one of the most serious problems in the library world.” In a collection as large as that of the Library of Congress many books are used infrequently but must be preserved for those times when they are needed by scholars. It is impossible for the Library to keep a check on all the books on the shelves, and Mr. Poole admits that it is nearly impossible to predict which books will “disappear.” “We will know that answer in a hundred years,” he warns, “and then it will be too late.”

At the New York Public Library, where the city’s dirty air appears to be hastening the deterioration of the acid paper, the situation is particularly acute. One of the world’s great centers for original research, the library has a vast and spectacular collection of pamphlets, propaganda leaflets, and the like. Recently a librarian going through photocopies of its unsurpassed collection of anti-Semitic literature—a collection that is an invaluable key to the psychology of ethnic bias—took a look at the originals. Without exception, they had disintegrated to the point where they could not be read. “If someone had not by chance copied them ten years ago, they would have all been lost,” says James Henderson, head of the reference department.

The problem of paper treated with highly acid alumrosin sizing has shown up in unexpected places. Libraries, already threatened with the loss of many of their books, are faced as well with the prospect of having to replace their disintegrating card catalogues, also the victims of chemistry. Art dealers have discovered that water colors and drawings by American artists like Maurice Prendergast and James Whistler are on paper which may not last through the century.

In New York, the matter of paper decay has even entered the civil rights arena. John Henrik Clarke, a Negro historian, recently charged that the Schomburg Collection in Harlem, probably the country’s finest archive of writings by and about Negroes, is deliberately being allowed to decay by the white-dominated New York Public Library that owns it. The collection, gathered by Arthur Schomburg, a banker, in the early years of this century, is indeed in bad shape, but its letters, diaries, speeches, and poems are deteriorating because of poor paper rather than because there is a conspiracy to erase the memory of Negro achievements. Though inadequate library budgets have been a contributing cause, the laws of chemistry deal with Negro and white with an equal hand.

One of the most startling developments in the last few years is the discovery that acids readily migrate from acidic papers to acid-free ones. The results of this have been surprising. The Library Company of Philadelphia, which possesses one of the great collections of early American books and documents, found that its rare Revolutionary War broadsides had been affected by the alum-rosin-sized folders in which they had been kept. A leading New England museum recently acquired a collection of old-master etchings and discovered that the acid matting and backing had damaged them. The Library of Congress has found that a collection of rare American maps given to it in 1903 by the Department of State had been mounted on highly acid pulp board and was disintegrating as a result. Even more startling is the warning by Barrow, in the American Archivist , on the acidity of repair materials:
… acidity is a prevalent deteriorative property of both paste and mending tissue. “Repaired” documents, therefore, are not necessarily “saved” documents because the migration process may slowly continue over a long period of time.

America’s most famous fading document—the Declaration of Independence—has been helped on its way to oblivion by uninformed “preservers.” During the Second World War, when the Declaration was kept at Fort Knox, a close inspection revealed that it was mounted on heavy pulp board, and, according to a government librarian, “A strip of tissue paper, about 3/4 inch wide, had been pasted with an adhesive, apparently part glue and part paste, on the mount in the form of a rectangle.” Undoubtedly it has been “treatment” of this sort which has done much to damage our most valued document, although fading ink is a principal cause for alarm.

While acid was doing its worst to all things written or printed on paper, people interested in the motion picture found that the nitrate-based film in use since the invention of cinematography in the i88o’s could have as short a life expectancy as the cheapest acid paper, and moreover, that the aging process of such film was notoriously and dangerously unpredictable. Eastman Kodak developed a much more stable safety film, using an acetate base, for home movies as early as World War I, but nitrate-based film remained the standard for all theatrical motion pictures until about 1950. Cellulose nitrate used as a film base has good optical and physical properties, but like all nitrogen compounds related to nitrocellulose (guncotton), it is not only unstable but highly inflammable. It ignites easily (about 266° F. when new) and burns very quickly—in large quantities, explosively. More serious is the fact that even at normal temperatures it begins to decompose from the moment of manufacture. The decomposition proceeds very slowly and may take many years, but eventually and inevitably every image recorded on nitrate film will have to be transferred to acetate if the image is to be preserved.

Many of the masterpieces of the first sixty years of cinematography have already disappeared. “By no means does the nonpreservation of a film reflect on its artistic or financial success,” says Gary Carey, the assistant curator of films at the Museum of Modern Art, which has been a leader in film preservation.

Very few of the highly popular films made by Norma and Constance Talmadge are extant. Most of the films made by Laureue Taylor and Jeanne Eagels have been lost and with them the record of two of the most famous stage stars of our century. Even the continuing popularity of stars such as Fred Astaire and Ginger Rogers does not guarantee preservation; for there seem to be no prints of the 1935 Roberta they made together.

Even under optimum storage conditions, nitrate film of any age may unpredictably begin to disintegrate. First the silver image begins to undergo a brownish discoloration and serious fading. This is followed by stickiness of the emulsion leading to a partial softening of the film and the appearance of blisters. A pungent odor signals the next stage, when the entire film congeals into one solid mass. The film base then disintegrates into a brownish powder which has a very low ignition temperature and is highly explosive.

The International Federation of Film Archives has advised its members to inspect nitrate film regularly, whatever its vintage, and to transfer it to acetate at the first sign of discoloration and fading. The I.F.F.A. reports that millions of feet of artistically and historically irreplaceable nitrate film are disintegrating in the vaults of major studios at this moment. So little attention has been paid to the problem that the Academy of Motion Picture Arts and Sciences has discovered that it cannot mount a complete retrospective of Academy Award-winning movies.

In Washington, the newly organized American Film Institute is working feverishly with the Library of Congress to establish the National Film Collection, to be a comprehensive and permanent collection of American films, before it is too late. The library has been collecting films since the advent of cinematography, when the first films were deposited for copyright protection, but films are only a small part of the library’s concerns, and the congressional appropriations have never been adequate to collect and preserve more than a token number of those produced.

The American Film Institute’s Archive Division is now scouring the nation to locate deposits of nitrate film not yet in institutional hands, and is warning those institutions holding nitrate that they may have a potential explosive or that—in case they haven’t looked lately—they may be storing worthless blobs of gelatin and silver salts. To dramatize the situation the institute has drawn up a list of 250 important American films which it believes are either lost or in imminent danger of decay because they are still on nitrate. This “rescue list” includes works by D. W. Griffith, Charlie Chaplin, Mack Sennett, and Erich von Stroheim, along with others featuring stars like John Barrymore, Laurel and Hardy, Douglas Fairbanks, and W. C. Fields. The institute also reports that there are no known prints of Cecil B. de Mule’s first film, The Squaw Man , one of the milestones of film making. And almost all the historic Edison films have disappeared.

Recently the film companies themselves have become interested in the preservation of their product. Metro-Goldwyn-Mayer, Hollywood’s largest studio, is converting all of its remaining titles from nitrate to acetate; Universal, one of the oldest film companies, now keeps its negatives in special vaults in New Jersey and regularly inspects the ones still on nitrate for any signs of deterioration. When deterioration is spotted, the films are transferred to acetate. Whether or not all of the thousands of films which have been made are worth saving is open to question, but there can be no doubt that the loss has been a tragic one. Last winter the Museum of Modern Art mounted an exhibition of promotion photographs from lost films; among the films included in the exhibition were The Divine Woman (1928) starring Greta Garbo; The Devil’s Passkey (1920) directed by the great Von Stroheim; and One Glorious Day (1922), starring Will Rogers. It is clear, as Sam KuIa, the archivist of the American Film Institute, says, that “the disintegrating nitrate film has carried away a significant part of the American film heritage.”

The destruction wrought on the record of man by the instability of nitrate film has not been limited to motion pictures. When, in the eighteen nineties, still photographers began switching from the almost literally permanent glass-plate negatives to nitrate sheets, they were risking consigning their work to oblivion, for the new sheets had the same highly flammable, self-destructive base as the film used by movie makers.

“From its inception,” warns Thomas Barrow, assistant curator of Eastman House in Rochester, New York, which has one of the most important collections of still photographs and movies, “nitrate-base photographic film was dangerous. It is highly flammable, and eventual disintegration is inevitable.” A large part of the production of Lewis Hine, the famous chronicler of America’s twentieth-century technological society, has been lost because he worked with nitrate-based film. A pointed example of the impermanence of nitrate is the story of what happened to part of the work of the great turn-of-the-century Staten Island photographer Alice Austen. When her photographs were rediscovered in the 1940'$, all of her early work, which was on glass plates, was in perfect condition; many of the negatives that she had taken later, after switching to nitrate film, had become sticky sheets from which prints could not be made.

“The critical years are from about 1890 to the Second World War,” Thomas Barrow says. “After that, most still-camera film [was made with] a non-inflammable (safety) acetate base, which is much more permanent than nitrate. By means of transferring them to acetate or the newer polymer-based film and keeping them at the optimum temperature and humidity, we can take care of the photographs that we have at Eastman House.” But Barrow points out that because of storage space and budget, Eastman House cannot even accept all the negatives that are offered to it for preservation. “The real danger” he says, “is the probable loss of those possibly splendid local photographers—the Alice Austens of this century—whose nitrate film is now stored in attics, basements, or the libraries of local historical societies. Most likely by the time someone knowledgeable gets around to looking at them, there will be nothing left.”

As the full seriousness of the situation became apparent, a variety of rescue plans for the printed word were suggested. William Barrow himself experimented with a number of techniques whereby books could be de-acidified by being soaked in or sprayed with solutions of calcium and magnesium bicarbonate. But these techniques are slow, expensive, and of uncertain effectiveness.

More widely publicized is the use of microfilm (on acetate) for preserving the contents of books and newspapers. It has been suggested that this is the answer. But in the early 1960’s, some signs of deterioration were found on microfilm negatives that were only twenty-five to thirty years old, and a special committee formed by the Association of Research Libraries stated: “Although apparently not widespread this deterioration is potentially serious enough to justify not placing reliance on negative microfilm as a means for long-term preservation of even the text of significant books.”

The makers of microfilm replied that the spots that had appeared had not impaired the film. They also stated that microfilm should be stored in tin cans—the acid in some paper cartons can migrate to film—at less than thirty per cent humidity, and that the temperature should not be allowed to go over seventy degrees. The microfilm should be kept free of dust and away from polluted air, should be handled by experts, and should be used only in clean viewing machines. These are requirements easily met by, say, a company that uses microfilm to preserve rarely consulted records, but an understaffed public library in the middle of a dirty city with thousands of people using its collection every day simply cannot enforce such standards. Moreover, librarians are unanimous in the opinion that the task of transferring enormous volumes of material to microfilm presents problems that are presently insurmountable.

The major over-all proposal for preserving the written word of our heritage is that put forth by a committee of the Association of Research Libraries. This would establish a central library that would assure the preservation for as long as possible of at least one copy of every significant document, by de-acidification as well as by cold storage—which, it has been found, sharply retards disintegration—and provide for use of these through microfilm or other photocopies. Research libraries and archives would be expected to give their deteriorating materials to the central library before they became completely unusable. Though this proposal has met with widespread approval, nothing has yet been done to implement it.

There is one hopeful development which, though it cannot do much about the past, may make life easier for scholars and librarians of future decades. In the course of his work William Barrow showed that if an alum-free sizing were used, if the wood in the pulp were thoroughly washed, and if the paper had long wood fibers, it was possible to make long-lasting chemical wood paper. As a result, many manufacturers, including Standard, S. D. Warren, Baton’s, Meade, and Oxford, are now producing acid-free papers. Standard, for instance, has made a paper that it says will last three hundred years compared to the thirty-to-fifty-year life of most book papers of ten years ago, and the company now claims a breakthrough to a paper that has a life expectancy of one thousand years. The new papers are being used by a number of presses, especially those connected with universities—Oklahoma, Yale, Harvard, Chicago, Ohio, and Indiana, for example—as well as by a number of scientific publications, including the Biophysical Journal and the Journal of Mathematics and Mechanics . More recently, commercial publishers such as McGraw-Hill and Barnes and Noble have begun to switch to acid-free paper. It seems safe to predict that since the new papers cost only slightly more than the acid ones, more and more paper manufacturers will make them and more publishers will use them.

In reviewing the damage done to the record of the past decades by acidic paper and nitrate film, it is only reasonable to point out that in an era when more was being printed and more images captured than ever before, a great deal will survive. It is also true that, at least temporarily, books which are in private libraries, and which therefore are more lightly used, stand a better chance of surviving than those in our great public collections. Moreover, many volumes that are in constant demand will survive because they are constantly being reprinted. It is possible to argue, finally, that there are many things recorded on both paper and film which are not worth saving. But who is to make the awesome decision about what is to be saved and what is not? Who is to say what will be important tomorrow? Who is to say which crumbling book or magazine or letter might hold the key to our understanding of some phase of our past? It would be a brave man indeed who would venture an opinion.

We hope you enjoy our work.

Please support this 72-year tradition of trusted historical writing and the volunteers that sustain it with a donation to American Heritage.

Donate