CHARLES J. MENDELSOHN AND WHY I ENVY HIM
by David Kahn
This is expanded from a talk 15 April 2003 at the opening of an exhibition for the Mendelsohn collection of books on cryptology at the Rare Books and Manuscript Department of the Van Pelt-Dietrich Library of the University of Pennsylvania. The collection has been newly catalogued by Regan Kladstrup, who selected the books for the exhibition and wrote the descriptions and accompanying texts. The exhibit was designed and mounted by Andrea Gottschalk. The collection＊s catalogue may be retrieved on line at <htto://www,franklin.library.upenn.edu/>
and by then doing an ※author§ search.
Charles Jastrow Mendelsohn built the largest collection of antiquarian books on cryptology in the world. He became one of the greatest scholars of the history of cryptology. Those are two of the reasons I envy him. But there are many others.
His father Samuel Mendelsohn, was the rabbi of the Temple of Israel in Wilmington, North Carolina. His mother, Esther Jastrow of Philadelphia, was the cousin of two distinguished graduates of the University of Pennsylvania. Joseph Jastrow, class of 1884, was awarded what is thought to be the first Ph.D. in psychology in the United States 每 in 1886 from Johns Hopkins. His older brother, Morris, Penn 1881, in 1892 was appointed professor of Semitic studies at Penn in 1892 and university librarian in 1898. He held the librarianship until 1919, the professorship until his death in 1921. So one of the other reasons I envy Charles Mendelsohn is his distinguished academic lineage.
When Rabbi Mendelsohn＊s congregation heard of his impending marriage, they showed their affection for him by carpeting and furnishing his house, on Chestnut Street between Fourth and Fifth Streets. Charles was born 8 December 1880 in Wilmington. By age 7, he was captaining a baseball team, the Blue Lights, which on Monday, 8 August 1887, defeated the White Stars, 30-20. But he also grew up in a scholarly atmosphere. His father published The Criminal Jurisprudence of the Ancient Hebrews in 1891 and in 1893 the state university conferred an honorary doctorate of laws upon him.
Charles attended the Episcopal Academy in Philadelphia, starting at 15 and graduating the following year, winning the prizes in Latin and Greek in 1896. Of course he went to Penn, entering at age 16. In his freshman year he won the Class of 1880 prize in entrance mathematics, the B.B. Comegys prizes in entrance Greek and Latin and the faculty prize in sight reading of Greek and, in his sophomore year, honorable mention in sight reading of Latin. He was a member of the Pennsylvania Debating Union and managing editor of the Pennsylvanian in his junior year and assistant editor-in-chief in his senior year. In 1900, he graduated Phi Beta Kappa and twice in a row won the university＊s fellowship in classical languages. In 1904, Mendelsohn was awarded the Ph.D. in classics from Penn. His dissertation was on word-play in Plautus, the Roman writer of comedies one of whose plots was used by Shakespeare in ※The Comedy of Errors§ and another by Moli豕re in ※L＊Avare.§ It was published in 1907.
After graduation, Mendelsohn was appointed to the faculty of the College of the City of New York as a tutor in Greek. When the United States entered World War I, he joined the censorship division of the Post Office Department. He may have been interested in codes and ciphers from childhood -- perhaps his motivation for moving to the army＊s codebreaking agency, section 8 of military intelligence, or MI-8. This was headed by a 28-year-old former telegrapher and State Department code clerk, the capable, energetic, charismatic Herbert O. Yardley. This is another reason I envy Mendelsohn. He knew and worked with Yardley, whom I would have loved to meet.
Mendelsohn was commissioned a captain, and Yardley placed him in charge of solving German codes. The messages Mendelsohn worked on were not from the Western Front; those were handled by codebreakers in France. Rather they were diplomatic messages that MI-8 obtained from a government radio station in Houlton, Maine. Most passed between Berlin and Madrid, Spain being neutral in favor of Germany. The messages were in four- and five-digit groups. Mendelsohn and his team, which included Victor Weiskopf, a former Department of Justice agent, and Edith Rickert, a University of Chicago English department professor, were greatly helped by Britain＊s having given MI-8 a partial reconstruction of German diplomatic code 13040 and others. When a large number of typists was unexpectedly and temporarily assigned to MI-8, Mendelsohn energetically took advantage of this windfall to have them to prepare statistics about the messages. MI-8 eventually solved six German diplomatic codes, which Mendelsohn and his co-workers determined derived from an unknown original that they called XX. His technique resembled the generation of stemmata 每 family trees of manuscripts tracing them back to what appears to be their source. It is an scholarly piece of work.
Two of the solved messages exposed German intrigues in Mexico. The first was sent from Nauen, the German transmitter in an exurb of Berlin, to Mexico at least 64 times between 23 January and 2 February 1918. It discussed a plan for providing Mexico with arms and machinery and with technicians for manufacturing weapons and airplanes. Nothing came of the idea. The second, likewise from Nauen and intercepted in February, authorized the German minister in Mexico to offer the Mexican government 10 million pesetas as a ※preliminary amount§ ※on supposition that Mexico will remain neutral during war.§ But the minister, who had been urging that Germany loan Mexico 20 times that amount to resist American pressures, never even mentioned the proposal to the Mexican government, perhaps as too insulting, and no significant German capital ever passed to Mexico. Though the United States merely watched these developments, the intercepts deepened its knowledge of Germany＊s machinations in a neighbor.
A radiogram from Germany＊s Madrid embassy to Berlin in Code 9700 revealed a plot to infect horses that the Allies were to buy. MI-8＊s reading of a 1,500-word prewar message validated its cryptanalyses, for the solution proved identical to a memorandum by the German ambassador to the secretary of state that was in the files. And MI-8 felt good about that, for while it found German diplomatic codes better than those of any other government it studied during the war, it broke them.
After the war, Mendelsohn returned to CCNY, now to teach history, and to its associated high school, Townsend Harris. But World War I had shown the United States the importance of codebreaking. and Yardley＊s executive ability had dazzled his government. He was placed in charge of America＊s first permanent peacetime codebreaking agency. This was set up in New York, at first in a brownstone on 38th Street, then in one on 37th Street, and finally in a tall office building at 52 Vanderbilt Avenue, on the corner of 47th Street. The agency came under military intelligence but was funded jointly by the War Department and the State Department. It was formally known as the Cipher Bureau but later became known as the Black Chamber, from the title of Yardley＊s best-selling book of 1931. Mendelsohn worked there part time apparently in the afternoons. He also joined with Yardley in a commercial codebook endeavor, in which Yardley got 49 shares, Mendelsohn, also 49, and the lawyer two. The Commercial Code Company compiled and sold a commercial code, the Universal Trade Code. Commercial codes were not secret but condensed a phrase or a whole sentence into a single codeword, thereby saving on cable tolls. For example, in the Yardley-Mendelsohn code, account-s subject to discount (of) was represented by ACKWO, which was cheaper to transmit than five words.
What cryptanalytic work Mendelsohn may have done does not appear in the records of the Cipher Bureau. It may not have been much, since no German codes seem to have been solved. Indeed, after the solution of Japanese diplomatic codes, which helped the United States win a diplomatic victory at the Washington naval disarmament conference of 1921-22, the Cipher Bureau languished, solving few foreign intercepts and having no influence on America＊s foreign policy, which was, in any event, isolationist. Yardley spent most of his time during the real estate boom of the mid- and late ＊20s buying and selling property in the New York City borough of Queens, where he lived. He and his wife transacted two property matters with Mendelsohn on 4 December 1928.
Yardley＊s Cipher Bureau was closed down in 1929 by Secretary of State Henry L. Stimson on the ground, as he said later, that ※Gentlemen do not read each other＊s mail.§ Yardley wrote a best-selling book, The American Black Chamber, about the work of MI-8 and the Cipher Bureau in which he cited only one collaborator, Dr. John M. Manly of the University of Chicago. Mendelsohn was not mentioned. Yardley was always broke, and at one point he sold his share of the code-making company to Mendelsohn. Sales of the one code that it had produced, the Universal Trade Code, were handled by another code company, C. Bensinger Co. at 15 Whitehall Street, New York.
The end of Yardley＊s bureau did not end Mendelsohn＊s interest in cryptology. He was on good terms with William F. Friedman, a brilliant cryptologist who, with the demise of the Black Chamber, headed all Army codemaking and codebreaking. Friedman had served as technical consultant to the American delegation to the International Telegraph Conference at Brussels in 1928, one of whose major rulings dealt with the cable rates to be charged for the various kinds of codewords used in commercial codes. He joined with Mendelsohn in a paper that utilized Mendelsohn＊s mathematical ability. It sought to determine the maximum number of codewords permitted under the rules of the International Telegraph Union in force starting 1 October 1929. Codewords could be ten letters long but had to contain at least three vowels to make them ※pronounceable§ for the telegraph operators, who used not teletypewriters but Morse keys. To keep codebooks to a manageable size, codemakers used codewords five letters long, which they combined in pairs. But each five-letter codeword had to have at least two vowels to ensure that, when any two were joined, they would have at least three vowels. This reduced the number of usable five-letter codewords from the theoretical maximum of 26 5 26 5 26 5 26 5 26, or 11,881,376.
In addition, Mendelsohn and Friedman, who had experience with the practical problems of transmitting the gibberish of code, imposed two additional safety restrictions on codewords. One was to have them differ from one other not by just one letter but by at least two. In other words, if AAAAA were a codeword, AAAAB would not be allowed, but AAABB would be. So if a garble mangled one letter, the resultant five letters would not appear in the code word list. This would signal an error and the cipher clerk could search the five letters to find the one that would produce a coherent plaintext. The second restriction was to prohibit codewords that differed from one another by the transposition of adjacent letters. Such switching is a common psychological error 每 FEIND for FIEND, for example. Mendelsohn and Friedman sought to determine how many five letter codewords they could produce with these restrictions. Their calculations showed that an alphabet with an odd number of letters would make it possible to avoid transposition of alternate letters. But an alphabet of 25 letters would, by eliminating one letter, reduce the number of possible words. So they cleverly added a dummy letter, l, lambda, to create a 27-letter alphabet. This would produce more combinations than a 25-letter alphabet. They then deleted all words with the l, leaving codewords made up of only the letters A to Z. With a 25-letter alphabet, they would have had 390,625 words. With the 27-letter alphabet without the l, they had 440,051. They published this article in the August-September 1932 issue of the American Mathematical Monthly, alongside articles on ※Quasi-Cyclotronic Polynomials§ and ※Real Roots of a Class of Reciprocal Equations.§ It had little effect because the advent of teletypewriters impelled what was now called the International Telecommunications Union in 1932 to drop the pronounceability requirement and permit five-letter codewords without any restriction. Still, the article showed Mendelsohn＊s breadth of knowledge. He was truly a polymath. I envy him that ability.
He and Friedman were good friends. They passed personal information back and forth in their letters, and lots of chit-chat. Mendelsohn, telling him about a bookplate by Yardley, said, ※If you don＊t laugh at the following you＊re really sick.§ They lent material to each other. They visited. When Mendelsohn planned to pass through Washington, Friedman asked him to make sure to let him know at what time he would arrive as he wanted to meet him at the station. Mendelsohn told Friedman about cryptology books that were available. After he bought a 1726 work by Solbrig for $10.50, he wrote Friedman that ※I ＃ am now offered another copy. If you want it for the files let me know at once, as it may be gone. It will cost $9.§ On 29 September 1936, he wrote Friedman, ※I have been offered H. O. Yardley, ＆Le Cabinet Noire＊ [sic] at 20 francs. Do you want it?§ Friedman asked him to buy it. This is another reason I envy Mendelsohn. He got to know well and to work with Friedman, another of the giants of American cryptology. I would have loved to do that.
After the closure of Yardley＊s Cipher Bureau. Friedman, who had worked alone in the 1920s, hired and trained cryptanalysts. Mendelsohn＊s World War I report on the German codes would help teach them. Friedman asked Mendelsohn to prepare the report for confidential Army publication. Mendelsohn accepted, arranging to pick up the documents in Washington on his way to vacationing with his mother in Wilmington (his father had died in 1922) and to work on them during the summer. Friedman planned to meet him at the rail station, saying, ※It will be a great pleasure to go over the stuff with you.§ It gave me a very nice feeling to see that indication of their personal and intellectual friendship.
Mendelsohn edited and expanded the report and returned the papers to the Army on 10 August 1931, saying, ※It has been a source of real pleasure to renew acquaintance with this code material. Besides some miscellaneous odds and ends, I have written a study of some of the matter, which I am sending direct to Major Friedman.§ (Friedman, a civilian employee of the War Department, held a reserve commission.) But publication was delayed for one reason or another and the Army did not print it as a technical paper until December 1936. Studies in German Diplomatic Codes Employed During the World War had three sections: I, ※Code 18470 and its Derivatives,§ which traced the relationships of the codes back to the unknown XX code; II, ※The ＆Fuenfbuchstabenheft,＊§ dealing with a five-letter code; and III, ※German Methods of Code Encipherment.§ He later supplemented that third part with An Encipherment of the German Diplomatic Code 7500, another Army technical paper.
He and Friedman then tackled the cryptanalytic problems involved in the greatest intelligence coup of all time. This was the Zimmermann Telegram 每 a cablegram sent in 1917 by the Kaiser＊s foreign minister, Arthur Zimmermann, to the president of Mexico. Zimmermann, believing that Germany＊s unrestricted submarine warfare would bring America into the war against Germany, proposed that Mexico join Germany in the war. When they won, Mexico would get back the territories it had lost in 1848 每 Texas, Arizona, and New Mexico. Britain intercepted and solved this message. Realizing that this was the propaganda weapon that would finally bring America into the war on her side, Britain gave it to President Wilson. He made it public. America exploded in outrage. Six weeks later, Congress declared war on Germany. Two German codes were involved in this 每 the modern 7500 in the message from Berlin to Washington, and the older 13040 for the relay from Washington to Mexico City, whose German legation did not hold 7500. Mendelsohn and Friedman examined and partially reconstructed the German codes involved and clarified the cryptology of this most important cryptogram in history.
The pair planned to write a scholarly history of cryptology. The professor of Latin would deal with the works written in that international lingua franca in the cradle years of the science. The active modern cryptologist would discuss the later evolution. The book was never written, but Mendelsohn translated two fundamental items as a warm-up. One was the 1466 manuscript of the father of Western cryptology, Leon Battista Alberti, an architect and a model of what came to be called a Renaissance man. His manuscript, of about 26 pages, included the first Western exposition of cryptanalysis, the first polyalphabetic substitution, and the first enciphered code. The other item was the world＊s first tract devoted entirely to cryptanalysis. By Cicco Simonetta, a secretary of the dukes of Milan, on two narrow strips of paper on 4 July 1474, it comprised 13 rules for solving monoalphabetic substitutions. Friedman published it in the Signal Corps Bulletin
These two translations served as forerunners to Mendelsohn＊s four scholarly articles on early cryptology, primarily of the Renaissance. Scholarly articles present fresh information, they＊re interesting, they deal with a single topic, and they＊re short. I like to read them and to write them. This is another reason I envy Mendelsohn. He did what I like to do, and he did it well.
The first of his articles appeared in 1939 in a journal called Scripta Mathematica. It dealt with the brief sections on cryptography in two works in Latin by Girolamo Cardano, an Italian physician, mathematician and polymath of the 1500s. Mendelsohn, who claims Cardano＊s style is not the easiest, says that he gives what he calls a lazy man＊s example of the autokey. It was in fact a defective form of the cryptosystem, carelessly printed, and it is not surprising that only a few of the many post-Cardano writers on it whom Mendelsohn cites understood it. Mendelsohn elucidates it and sets it into its place in the evolution of cryptology. He quotes Cardan＊s observation about the number of alphabets that a monoalphabetic cipher would make possible and remarks that ※Trust in large numbers has been an ignis fatuus of cryptographers throughout the centuries.§ He describes Cardano＊s mask cipher, or grille. This developed into the turning grille, a transposition, which later had a vogue and which mathematicians analyzed to determine how many different grilles can be made with a fixed number of squares.
His next article, an important one, was published in 1939 in the Signal Corps Bulletin 每 almost certainly through Friedman＊s influence. Mendelsohn showed how another Renaissance cryptologist, Giovanni Battista Porta, had, in the 1602 edition of his De Furtivis Literarum Notis, come within a hair＊s breadth of finding the general solution for a repeating-key polyalphabetic, which had just completed its development. This would have anticipated the Prussian Major F. W. Kasiski＊s discovery of its solution by 250 years and would have destroyed the polyalphabetic＊s reputation as ※the indecipherable cipher§ almost before it had gained it. Mendelsohn says elsewhere of Porta, ※He was, in my opinion, the outstanding cryptographer of the Renaissance. Some unknown who worked in a hidden room behind closed doors may possibly have surpassed him in general knowledge of the subject, but among those whose work can be studied he towers like a giant.§ Only someone who had read widely and deeply in the literature of the time could offer so confident an opinion.
Mendelsohn＊s most impressive article was published in 1939 in the prestigious Proceedings of the American Philosophical Society, submitted apparently at Manly＊s suggestion. In it he showed how the most famous form of the polyalphabetic -- the Vigen豕re, named for the 16th-century French diplomat Blaise de Vigen豕re -- had been misunderstood for hundreds of years since its publication in Vigen豕re＊s 1586 volume. Mendelsohn says that this book may properly be called ※a part of the cryptographer＊s Bible,§ adding acidly that ※Like the Bible, it is often spoken of but generally unread.§ Vigen豕re had devised a far better cipher than the one ascribed to him by the 23 authors from 1624 to 1935 whom Mendelsohn cites. Mendelsohn ascertained this by reading Vigen豕re＊s original text 每 a feat that, astonishingly (or perhaps not so astonishingly) 每 almost nobody else had done for the previous 350 years. He traces polyalphabetic substitution from its source in the Alberti manuscript, through other cryptologists＊ additions and refinements in obscure publications of the 1500s, up to Vigen豕re＊s book. He then sets out what Vigen豕re really devised, which was the perfected form of the autokey.
Mendelsohn＊s last article, ※Bibliographical Note on the ＆De Cifris＊ of Leone Battista Alberti,§ was published in Isis, the journal of the history of science, in 1947, after the hiatus of World War II.
I found all three of his articles of value and used them extensively to improve the accuracy of The Codebreakers. His reference in the Vigen豕re article to the cabalistic work ※Sefer Yezirah§ helped me later while I was tracing the origin of polyalphabeticity back past Alberti to the philosophical disks of the Catalan mystic Ramon Lull.
Mendelsohn was the first real scholar of the history of cryptology. Other authors 每 such as Johann Ludwig Kl邦ber in his excellent 1809 Kryptographik -- included historical material. But none approached Mendelsohn in the range, precision, and soundness of his scholarship. I envy him that as well.
Still, the activity with which Mendelsohn＊s name will be forever associated is less his scholarship, less his monographs and articles, than his collection of antiquarian books on cryptology. He was a bachelor and thus perhaps had more time and money to devote to this pursuit than married men, who often have to support a wife and children. During the 1930s, Mendelsohn was collecting the antiquarian books on cryptology that form the basis of his fabulous collection. I have not checked a list of his books against the collections of the British Library in London or of the Biblioth豕que Nationale in Paris or of the remarkable one of David Shulman in the New York Public Library. But I have no doubt that Mendelsohn＊s compares well with those of the two centuries-old libraries and may even exceed them, even though those institutions were buying books decades before Mendelsohn was born. The reason is that he worked at collecting on that subject. For example, he may have many more works in German than they have, because British and French libraries may not have paid as much attention to works in a foreign language on a curious but unimportant subject as he did. Moreover, Mendelsohn had the means, the knowledge, and the energy to pursue these works. Of course, he collected before the explosion of interest in cryptology, occasioned by the rise of computer security, the revelation of the World War II Ultra secret, and the invention of public-key cryptography. So his collection on cryptology does not include any works on those subjects 每 though the library containing his collection must surely include some. But in earlier ※classical§ periods, it is, if not the greatest in the world, certainly one of the greatest.
The prices, compared with those of today, make one green with envy. He refers to a book by Bazeries (he does not say which one) as costing 38 cents! He ※lost out on the Thicknesse [a nice English book of 1772] at $3.75.§ Even at those prices, Mendelsohn did not throw his money around. When Colonel Parker Hitt＊s very little, very elementary The ABC of Secret Writing came out in 1938, Mendelsohn asks Friedman, ※Is Hitt＊s new book worth the dollar it costs?§ Friedman replies that ※Hitt＊s new book is worth at most 25 cents. ＃ I bought one, but that is only because I am collecting.§ Mendelsohn comments interestingly about the books in his correspondence with Friedman. ※ ＆Allgemeine Schrift＊ by David Solbrig, Saltzwedel, 1726, is one of the earliest codes I know, and elaborate and interesting. I do not aim to have a complete set of codes, but this book, with its introduction and method I am very glad to own. I should call it almost a ＆must have＊ book. I paid $10.50 for mine.§ On 21 January 1938, he wrote, ※Today I received what I regard as a prize 每 DuCarlet, ＆La Cryptographie,＊ 1644. In its contents the book is not great, but it seems to be a rarity. I know of no copy in America; do you? It was offered me a year ago by Maggs (morocco bound, for which I don＊t give a damn) at $78.00 (sic!), and I just forgot about it. Through a friend I came on the track of the present copy (calf bound) for $15.50 (sic!).§ He has been unable to find the Kasiski that he has been looking for 每 a fundamental work of 1866, explaining for the first time how to solve polyalphabetic substitution with a repeating key, regarded for centuries as unbreakable (though it is now in the collection, perhaps from another source). He and Friedman agreed that the 566-page Étude sur la cryptographie, son emploi 角 la guerre et dans la diplomatie, by the Belgian army officer, A. Collon, which appeared first as articles in the Revue de l＊Armee Belge in 1899, is one of the best works on the subject.
In November 1938, six weeks or so after the war scare before France and Britain capitulated in Munich to Hitler＊s demands on Czechoslovakia, the U.S. Army＊s Signal Intelligence Service, whose technical head was Friedman, sought to bring Mendelsohn back into active cryptanalytic service. The World War I cryptanalyst was ※looking forward to this work with pleasure,§ but warned on 23 November that ※A certain amount of red tape must be unwound in connection with any application for leave from an academic institution.§ In Washington, he would ※require nothing more than a room and my own bath, preferably about fifteen minutes＊ walk from the Munitions Building, with street car or bus facilities available in bad weather. Quiet, especially within the place where I live, is balm to my general well-being.§ The army told him it wanted him because ※We have a 4-letter code in the language with which you are familiar [German, not Latin], and on which we are making fair progress. The final filling-in of the code is the work for which we are going to depend upon you. In any case, this job will have progressed pretty well by the time you come. We also expect to have done quite a little preliminary work upon a more difficult code in the same language, and shall hope for your help in that. Also, if time permits, we hope you can go through our files of World War material and assist in classifying it.§ He began work on 24 January 1939 for six months at $400 a month, being paid by confidential voucher because ※disclosure of Dr. Mendelsohn＊s name in connection with this work would indicate the exact nature of the work.§
On 6 September, a few days after Nazi Germany had invaded Poland and Britain and France had declared war, Friedman told Mendelsohn that ※We are most anxious to continue with where you left off here.§ Could he ※obtain another extended leave of absence?§ The red tape began again to be unwound, with a letter to the acting president of the City College. Then, on Saturday, 23 September, George and Alice Bernheim, friends whom he was visiting in Port Chester, a suburb of New York, found him, as Alice, a doctor, wrote to Friedman, ※limp, cold, sitting on the edge of his bed. He had had a cold and a cough, had gone to bed the night before feeling chilly. I first thought he had had a stroke. I got him warm as fast as possible then sent him in an ambulance to the Mt. Sinai Hospital. He seemed to recover somewhat that evening 每 but on Sunday morning he showed signs of a full-blown meningitis§ and he died Wednesday, 27 September, aged 58.
This meant, bitterly, that the articles of the greatest historian of early modern cryptology were -- with the possible exception of the one in Scripta Mathematica -- all published posthumously. Friedman edited the ones on Porta and Vigen豕re for their publication. Mendelsohn＊s mother was his sole survivor, and the Chief Signal Officer, General Joseph O. Mauborgne, who had known him since World War I, expressed his condolences. The Army, worried that its cryptologic expert had been poisoned on the eve of his coming to work, asked the Justice Department to investigate. Evidently, it soon discovered the unsuspicious truth. But what was to be done with his things, particularly the books. He had left everything to his mother.
What would happen, above all, to his books? In a letter to Alice Bernheim, Friedman wrote: ※I cannot imagine anything more disquieting than to contemplate the dispersal of this collection by individual sale. So much time and labor of love has gone into the making thereof that it would be extremely unfortunate to have this happen. Furthermore, I doubt very much whether the sale of the books would bring in enough money to make a real difference in the well-being of Mrs. Mendelsohn, unless it is the case that she is much worse off financially than I have reason to suspect. It should be brought out that in my opinion nobody, in this country certainly, and perhaps nobody in the world, had made as serious a study of the history of cryptography as Charley did, and in making this study his collection was of course invaluable. For this reason it should be kept intact if for no other reason.§ (An Austrian historian, F. Stix, who had written some scholarly articles on the Austrian black cabinet of the 1700s and on other aspects of absolutist cryptology, was Mendelsohn＊s only rival in this field. But his papers were lost in World War II and, discouraged, he did not resume his studies.)
Friedman said he was astonished that ※so meticulous a person as Charley ＃ had made no provision in writing with respect to the collection of books in his library.§ He went on:
※I must tell you how vividly I recall the circumstances of his telling me of his wishes. At least two years ago, after he had spent a considerable sum of money and much time in gathering together his collection, I asked him what his intentions were with respect to the disposal of the collection after he passed on. He replied that he intended to will them to his alma mater. I pointed out to him that it would be in a way too bad to have the collection deposited there because they would not be easily accessible to the people most likely to want to consult them, namely, persons in the cryptographic services of the Government. This argument left him cold, however. About a year afterwards I repeated my question, simulating a poor memory on my part, but I received exactly the same answers and I did notice a small degree of irritation which was aroused by my query. Finally last winter I had the temerity, in a more or less jocular mood, to repeat my query, and this time Charley burst out with an oath, saying that if I asked him that question again he would brain me. Little did I suspect how near the day was that the question would be of immediate importance.§
George Bernheim, a lawyer, was Mendelsohn＊s executor. Friedman wrote him as well ※how unfortunate it would be to have Charley＊s collection of books on cryptography dispersed by sale and I hope that this will not be necessary.§ He also pointed out that Mendelsohn had purchased Yardley＊s share of their code publishing firm and thus became sole owner. ※Just how much business the Code Compiling Company was able to do in recent years I do not know, but possibly there is enough to make a material difference in the financial affairs of Charley＊s mother.§ (Its code had gone into a second edition in 1928.) Bernheim replied that he knew that Mendelsohn was sole owner and ※in due course I will have to wind up that company and see what I can get for it.§ As for the book, ※I feel quite sure that there will be enough money available for Mrs. Mendelsohn so that the books on cryptography will not have to be sold.§ A month later, on 8 November, he wrote Friedman that he had written Herbert Bluethenthal, whose wife was a first cousin of Alice Bernheim and a friend to the Mendelsohn family, to send him, Bernheim, the books and papers. Bluenthenthal apparently had custody of them, for Mendelsohn had lived for many years at the University of Pennsylvania Club, which probably did not allow him room for his collection and which moved four times just during the 1930s. The books may have been kept in Bensinger＊s office or in Wilmington. Presumably they went via Bernheim to Penn, giving it custody, as their owner had wished, of one of the greatest collections on cryptology in the world, in the hope and expectation that it would serve future scholars. As it has! And as it commemorates their collector!
Of course, we are all only custodians of the things of this world. We cannot take them with us. Still, it would have been nice to have enjoyed that collection for a while. And that is the final reason I envy Charles Jastrow Mendelsohn. He had all those books!
_________________________________The sources for this article lie in a study ※The Mendelsohn Family 每 Misc. History§ marked ※From the files of Bill Reaves, Wilmington, N.C. 1996§ and generously made available to me, in Mendelsohn＊s correspondence with Friedman, now in the Friedman collection at the George C. Marshall Library in Lexington, Virginia, and in the work of curator Regan Kladstrup.
Copyright © 2008 David Kahn. All rights reserved.