The Book Nobody Read

by Owen Gingerich

  Whose could the third hand be? Several days later, still feeling exuberant about the identification and recovery of the missing Mercator copy, I shared my discovery with a colleague, showing him both the photocopies and the lavish Gerhard Mercator volume. In describing what a polymath Mercator was, I turned to pages showing his astrolabes and then to those with a sample of his new italic letters. Suddenly I got an idea. What about comparing the italic letters with the third hand in the book? Almost immediately the match became clear. The capital £, h, and p were particularly distinctive. The third hand was also Mercator's but written with his new letter forms. Even though in the end his notes themselves didn't seem particularly illuminating, identifying Mercator's copy brought a satisfying closure to one of the principal remaining mysteries of the Copernicus chase.

  ONE MORNING as I was hard at work on the Census, I got a call from Jonathan Hill, a leading book dealer in New York City who had taken a keen interest in the project and who had given me many helpful pointers on the proper way to describe books as physical objects. "I've just obtained a nice copy of the second-edition Copernicus," he informed me.

  "Well, have I seen it? Is it in the census?"

  "How in the world would I know?" was his candid reply, so I suggested that he send the book up to Cambridge so I could inspect it, which he promptly did.

  This was by no means an unprecedented request. I had once had a first edition in my office for nearly six weeks in late 1980, on loan for inspection from the San Francisco dealer Warren Howell, and I didn't even catch on that I had seen it before. Nearly two years after I had sent the copy back to California I got a call from Howell, who asked me to compare my notes on his copy with those from the copy I had earlier seen in the John Crerar Library in Chicago. Because all my census notes were in an extensive set of file folders in a row on my desk, I offered to handle his request in real time while he stayed on the line. I propped the receiver on my shoulder and dug into my notes. The Chicago copy, like Howell's, was censored in the standard way specified by Rome in 1620. But the Chicago copy had seventeen extra manuscript lines on folio 4 verso. Amazingly, so did Howell's copy. And both had exactly the same page size, to the millimeter.

  "Why, these are the same book!' I blurted out with some surprise. "I didn't know the Crerar Copernicus was missing."

  "That's what I was afraid of," Howell said rather sadly. "I didn't know either that it was missing. Don't tell anybody just yet. I've got to figure out what I'm going to do."

  Howell, an eminent bookman whose specialties included rare science, had been approached by a man named Joseph Putnam claiming to be an emigre from Eastern Europe who had brought the books from there in a clandestine manner, and who had on offer some wonderful classics in the history of science, including De revolutionibus and the even rarer De motu cordis by William Harvey. Unknown to Howell, Putnam had ingratiated himself with the staff of the John Crerar Library in Chicago, so that he had regular unsupervised access to the rare book vault and had helped himself to very rare books. He went in with a large briefcase, sealed the books he wanted into large addressed envelopes as if they were finished projects ready to be mailed, and walked out of the library with only cursory inspections. While he was at it, he removed all the relevant cards from the card catalog as well. His clandestine operation began to unravel when someone inquired about a medical manuscript known to be uniquely in the Crerar Collection, but which Howell sold to another dealer and which ended up in the Staatshibliothek in Berlin. In parallel the Crerar Library discovered its rare Harvey volume couldn't be found, and eventually they took an inventory and found their losses were much greater. But only when the FBI showed Howell a list of the missing books did he begin to appreciate that he was the victim of a major scam. That's when he called me to check if the Copernicus book really came from the Crerar. A few weeks later, in mid-January 1983, he called back to report that Joseph Putnam had been arrested in Milwaukee, and that he had more than 300 rare books from the Crerar Library in his house.

  Howell tried his best to get the books back from his customers, with partial success, but he died a broken man in the following year—a tragedy, I felt, because he was such a kind and unmalicious person. Putnam was found guilty, and sentenced to two years in federal prison.

  The second-edition De revolutionibus that Jonathan Hill had on consignment was a lot easier to recognize. Someone had taken a black felt-tip pen to delete one identifying word from the Latin inscription on the title page, but a computer search quickly picked up the rest of the phrase. The marked-out word was Brunae, Latin for Brno. I had inspected the book in the State Scientific Library in Brno, but after communism faded and Czechoslovakia fissioned, the library had returned the book to its former owner, the Augustinian Monastery outside Brno. It was a wellknown establishment, where the renowned geneticist Gregor Mendel had been the abbot.

  Now the question was: Had the monastery deaccessioned the book to gain some needed cash? Hill had the book on approval from a reputable German dealer, who was in turn convinced that his source was legitimate. The Web and e-mail having come into ascendancy, I could confirm the status of the book within thirty-six hours, and the news was bad. The monastery had not sold the book. It had disappeared along with a temporary employee. I assume the book is now back in the Brno monastery, and that someone lost money to the vanished thief. Whether it was the German dealer or his insurance company I do not know, but I am aware that nowadays many dealers carry insurance policies for precisely this sort of situation.

  Unfortunately, it seems that in the majority of cases, unlike these two, stolen books do not find their way home.

  THE COMPUTER typesetting for the 112 first and second editions now in the United States was relatively easy, since with few exceptions the descriptions and annotations were quite straightforward, even though the block of American holdings was the largest single group. By and large the American copies have neither the extensive annotations nor the extended provenances that some of the European copies boast; precisely why this is true is hard to divine. The most conspicuous exception is the Yale Beinecke Library copy, whose description goes for eight pages including five illustrations. No other American description exceeds an entire page except for a first edition at the Linda Hall Library in Kansas City, whose brief description includes a full-page plate to illustrate a small paper instrument laid in the book. There are, of course, besides Yale's "three-star" copy, a few "two-star" exemplars, including the first-edition Offusius copies at the University of Oklahoma and the University of Michigan.

  Eventually, I reached the final U.S.A. entry, a second edition at the University of Virginia, perhaps the copy ordered for the library by Thomas Jefferson. Because many of the volumes of the library were destroyed by fire in 1895, it is now impossible to know whether the copy is the original one or a replacement.

  Since Yugoslavia had been replaced by Croatia, it had lost its final place in the alphabetic rankings, and since Vatican City had been grouped with Italy, the sequence closed with "USA: Texas—Virginia." This left just the appendixes, an assortment of miscellaneous tables such as a chronological list of auction records (which nicely documented the ever-upward price of the two editions), and lists of locations of the related Copernican books, namely, Rheticus' Narratio prima and his 1542 edition of the trigonometric section of De revolutionibus. Though I had long been collecting locations of this latter book, I belatedly noticed that I had never systematically organized a list. Disconcerting as this last-minute realization was, I decided that I would include such an appendix if within two days I could get up a list of approximately forty libraries holding copies of the De lateribus et angulis triangulorum.

  I had located over two dozen copies of this work as I visited libraries in search of De revolutionibus itself, and I got almost a dozen from the NUC (National Union Catalog of the Library of Congress) and the OCLC (On-line Computer Library Center). The richest haul came from the Karlsruhe Virtual Catalog, the main European computer
database, enough to put me well over my goal. Another source was the Italian bibliographer Giovanna Grassi's published list of early books in European astronomical observatory libraries. Glancing at her listings for Copernicus, my eyes wandered from De lateribus to De revolutionibus, and I was aghast! Though her book had been on my shelf for more than a decade, and was a reference I had frequently consulted, for some reason I had never checked it for locations of the magnum opus itself. Grassi listed fifteen copies of the first edition, fourteen of which I had seen, but one that was new to me. Similarly, of the fifteen copies of the second edition, there was also a new location. It appeared that the observatories in Naples and in Athens had copies of De revolutionibus I had overlooked.

  With long-distance telephone and e-mail, the confirmations were in hand almost instantly. But there was still more news. The Naples observatory had not only the first edition but the second as well. The three additional copies, now recorded in an addendum to the Census, brought the total number of entries to 601. That total included three German copies destroyed in World War II, for which some limited information about their early provenances has survived. There were as well a few far-flung copies that I did not personally inspect, in Lithuania, Corsica, Croatia, and the Philippines, for example, although I received information about all these copies. And the Vatican librarians have not been able to find one of their copies, though for completeness I listed it without any details.

  How many extant copies did I fail to list? An unanswerable question, of course. As the survey had gone on, the discovery of new locations had dwindled to a small trickle, yet if I extrapolated, the number always went to infinity because there was no evidence that I had come to the end of previously unknown copies. Infinity has to be the wrong answer, of course, since the likely number of first-edition copies printed lies between four and five hundred, and for the second edition between five and six hundred. Since the Census was printed, I have located one more copy of the first edition and six more of the second. If I had to pick numbers, I would estimate that there may be as many as a dozen undiscovered copies of the first edition and two dozen of the second. Only time will tell how cloudy my crystal ball really is.

  From the earliest days of my Copernican census I had conferred with my Polish colleagues and had agreed with Paul Czartoryski, editor in chief of the Studia Copernicana series of the Polish Academy of Sciences, that the Census should be published in their series. As the project grew in scope, I realized that the volume would be of interest to librarians and book collectors all over the world, and that it would be highly desirable to copublish with a Western press to guarantee its ready availability. Eventually, Czartoryski organized an agreement with Brill Academic Publishers in Leiden for a Western Studia Copernicana series, and thus they became the publisher. Brill did a splendid job, printing the book on high-quality paper that works well with the sixty-three glossy images it contains.

  Brill published the Census in an edition of only four hundred, just matching the lower estimate of the first edition of De revolutionibus, and guessed that it would take twenty years to sell that many copies, again approximately matching the time it took Petreius' edition of the Copernicus to sell out. These estimates missed rather badly, because the volume is already almost out of print.

  Only a handful of books have been targets for a complete census. Of course bibliophiles are keen to keep a census of all the locations of Gutenberg Bibles. As recently as 1996 a previously unlisted fragment of 177 pages was found in Rendsburg, Germany, but that seems unlikely to be repeated. It was comparatively easy to compile a census of copies of John James Audubon's Birds of America—not just because the giant elephant folios are hard to miss but because there was a list of original subscribers. A majority of all known copies of the so-called First Folio of Shakespeare's plays resides in the Folger Shakespeare Library in Washington, and a century ago a census of this title was made. None of these efforts includes the sort of detailed physical descriptions and lists of provenances found in the Copernican Census, although such a project is currently under way for the Shakespeare First Folios.

  An extensive list has been made of locations of Isaac Newton's Prin­dpia, without attempting to be definitive. That highly significant volume would be a good target for a survey comparable to the Copernican census, but as my own experience has shown, such a labor cannot be undertaken lightly. An excellent listing of locations of all of Johannes Kepler's titles recently appeared, but with no attempt to locate copies in private collections. His Astronomia nova of 1609 would be another candidate for a survey of annotations, but my own preliminary search suggests it would not be nearly as interesting as the Copernicus chase because there seem to be many fewer annotated copies.

  ONE MIGHT examine the series of provenances that the Census lists for each book in order to find which copy was once owned by a movie star, which by a saint, or which by a heretic. Or which copy has the longest list of provenances. Many copies have only a single provenance simply because the previous owner is completely unknown. In a few cases, of course, the copy is still in the original institution that acquired it in the sixteenth century. But some copies have records of remarkable journeys.

  Title page of De lateribus et angulis triangulorum, the trigonometric section of Copernicus' book that was edited by Rheticus in Wittenberg in 1542.

  I am amazed at the travels of one of the first editions now at the Paris Observatory. The earliest evidence for its ownership dates from the second half of the sixteenth century when an unidentified owner copied out the notes from Jean Pierre de Mesmes' copy; De Mesmes had in turn during the late 1550s copied most of his notes from his teacher, Jofran­cus Offusius. That anonymous owner included some of his own notes, including a citation to Maestlin's Ephemerides, which were published in 1580. The next record comes from Caspar Peucer, who had been a student of Erasmus Reinhold in Wittenberg and who succeeded him as professor of astronomy there in 1554. Despite his insider's position— he had married Philipp Melanchthon's daughter in 1550—he was jailed for his crypto-Calvinism for a dozen years between 1574 and 1586, during which time he made ink from his own blood in order to continue writing. He was subsequently rehabilitated and played host to Tycho Brahe when the Danish astronomer was on his way from Denmark to Prague. Peucer must have got this De revolutionibus after he was released from jail, though how it got from Paris to Wittenberg remains a blank spot in its history. The next known owner of this copy was Joseph-Nicolas Delisle, a French astronomer who founded the original observatory in St. Petersburg in the eighteenth century and who amassed a substantial personal library, which went to the library of the Depot de la Marine and subsequently to the Paris Observatory, where it has been ever since. We can imagine that Delisle found the book in Germany, took it along to Russia, and eventually brought it back to Paris, but that is of course pure speculation. Precisely where and when he acquired the book is again an unknown. In any event, I still find it astonishing that the book returned to France after such a journey.

  A more fragmentary provenance, but nevertheless one that I was particularly smug about establishing, links the first edition now in the Scheide Collection at Princeton with the large Biblioteca Nazionale Braidense in Milan. Three generations of Scheides have formed this private collection, which is housed in the Princeton University Library. John Hinsdale Scheide, the founder of the collection, bought his copy in 1928 for #785 from the colorful, pace-setting American dealer A. S. W Rosenbach. It contained some illegible initials, the manuscript date 12 March 1710, and an old but unrelated shelf mark, S.15.11, but otherwise no identifiable provenance. When I visited the National Library in Milan, not only did I record its two second editions—one censored according to the Inquisition's decree, the other totally unannotated—but I also noticed in the old manuscript catalog of the library that at one time this institution had a first edition with the shelf mark S.XV.II. One day in my office, much later, when I was idly sifting through some of the census data, it occurred to me to try
to compare this shelf mark against some of the old numbers I had recorded on books in other collections, and almost immediately I found the match. Surely, the Scheide copy was once in the Milan library, but how and when it got out will probably forever remain a deep mystery. One can only wonder if Rosenbach, a truly remarkable wheeler-dealer, had any clue about its origins.

  Perhaps the most heavily annotated title page known, in a Paris Observatory first edition of Copernicus' book including an attribution of notes to Caspar Peucer, Reinhold's successor at Wittenberg (upper right), and the hand of Joseph-Nicolas Delisle, who brought the book back to Paris (left center).

  I'm always curious about published catalogs of old collections, and when I get the chance, I examine them to see if they include Copernicus' book. By and by I encountered Biblioteca Firmiana, a multivolume library catalog published in Milan in 1783. The collection, established by Carlo Giuseppe Firmian, included a first-edition De revolutionibus, presumably the one that subsequently entered the Braidense collection. That chance encounter with the Firmiana catalog established a still earlier provenance for the Scheide copy, but neither the Scheide Copernicus nor the Paris Observatory copy come close to holding a record number of provenance steps. The winners in the Census each have nine listed ownerships: a second edition in the Ossolineum in Wroclaw, and a first edition at the University of Michigan in Ann Arbor.