THE CODEBREAKERS

by DAVID KAHN

  They have grown more complicated as readers have grown more sophisticated. In 1932, Dorothy Sayers’ urbane Lord Peter Wimsey solved a Playfair in Have His Carcase—and quite an elegant solution it is, too. Undoubtedly the most complex cryptogram ever to appear in novel form is the mixed-alphabet polyalphabetic in Helen McCloy’s Panic, published in 1944, when interest in cryptology was high. The author displays a fair knowledge of cryptology and its literature, listing some of the standard works in describing the contents of her cryptologist victim’s bookshelves. The cryptologist’s niece solves the cipher somewhat intuitively, though with a fair amount of analysis, and the solution leads to the killer. In 1957, Ian Fleming gave his hero, James Bond, the task of capturing the cipher machine Spektor, whose workings were, unfortunately, not described; the book, From Russia, With Love, was one of the favorites of President John Kennedy.

  Cryptology has insinuated itself into motion pictures as well. In Dishonored, Marlene Dietrich rolled out a few magnificent chords in which was supposed to be concealed some secret message with the notes representing letters. During the 1930s, when Saturday afternoon movie serials were so popular, Paul Kelly starred in a 15-part thriller entitled The Secret Code, which had very little to do with cryptology but displayed an awful lot of cliff-hanging action. Rendezvous, the film based on Yardley’s The Blonde Countess, of course had a touch of cryptology, as did the film version of From Russia, With Love, which flashed the Spektor—which resembled no cipher machine in existence—upon the screen for the briefest moment. The wartime documentary The House on 92nd Street had a brief scene showing German spies being instructed in cryptography at a Hamburg spy school, complete with a Vigenère-like tableau.

  Cryptology played a big role during daytime radio serials in the late 1930s and early 1940s, before television. Such favorites as Captain Midnight and Little Orphan Annie would send their faithful young listeners cipher disks or secret codes to decipher secret messages about tomorrow’s adventure. The numbers or letters were read by an announcer with a portentous voice.

  Even music has a touch of cryptology. About 1898, composer Sir Edward Elgar, best known for his Pomp and Circumstance march, wrote Variations òn an Original Theme, in which he musically depicted in each variation a member of his circle of friends, his wife, and, to end the piece, himself. Elgar labeled the basic theme in G minor, on which the individual portraits were the variations, “Enigma,” and said that it was itself a variation on another piece of music—which he never disclosed. “The Enigma I shall not explain—its ‘dark saying’ must be left unguessed,” he wrote, adding, “… the principal theme never appears.” Many persons have tried to guess what the Enigma theme might be: a phrase from Parsifal, one from Pagliacci, or the theme of Auld Lang Syne. None has won acceptance. But it is possible that a clue to the Enigma lies hidden in a cryptogram that Elgar sent to one of the “varia-tionees” in 1897—Miss Dora Penny, the Dorabella of Variation X. As a girl in her twenties she spent much time with Elgar, and when she asked him about the Enigma, protesting that she simply could not figure it out, she was told by the composer, “I thought you, of all people, would guess it.” He would say no more. The cryptogram consists of 87 characters consisting of one, two, or three curves in various positions and looking as a whole rather like a flock of sheep. Nobody has solved it, and so nobody knows whether it will shed any light on the Enigma. But if it does, it may help resolve one of the oddest mysteries in the musical domain.

  Finally, there is the case of the painting with its title in code. This might perhaps be more understandable if the painting were abstract, but in fact it is a powerful representational image of two conspirators, one whispering into the ear of the other. Part of a series of 12 × 16-inch tempera panels depicting the birth of the United States and its struggle for freedom, the picture, by Jacob Lawrence, bears part of Benedict Arnold’s dictionary-code message to John André in its title: “120.9.14. 286.9.33.ton 290.9.27. be at 153.9.28 110.8.19. 255.9.29. evening 178.9.8…—An Informer’s Coded Message.”

  * It may not be coincidence that in Ulysses an inventory of Mr. Leopold Bloom’s locked private drawer at 7 Eccles Street included, among other things, “3 typewritten letters, addressee, Henry Flower, c/o P.O. Westland Row, addresser, Martha Clifford, c/o P.O. Dolphin’s Barn: the transliterated name and address of the addresser of the 3 letters in reversed alphabetic boustrephodontic punctuated quadrilinear cryptogram (vowels suppressed) N.IGS./WI.UU.OX/W.OKS.MH/Y.IM:….” “Quadrilinear” meant to set the cipher in four lines; “reversed alphabetic” indicated the key of a = Z, b = Y, etc.; “boustrephodontic,” an adjective concocted from the adjective “boustrephodon,” a technical term in paleography referring to writing that runs left and right in alternate lines, indicating that the lines of the cryptogram were to be read in that way. Unfortunately, Joyce or Bloom forgot about this in the fourth line, which incorrectly reads left to right. The cryptogram and its solution thus are:

  * And perhaps not even by her. The system described is virtually the same as that apparently used by the Carbonari at the same period except that the superscribed differentiating numbers are omitted. It may well have been Berryer’s biographer who omitted them, either because someone did not give him all the details of the system or because he misunderstood. It is true that the Duchess of Berry was a royalist and the Carbonari were antimonarchical, but both were opposed to the line then occupying the throne upon the Duchess’ return to France in 1832. They might have shared the Carbonari system, or, more probably, it might have just been in the air at that time and therefore widely used. Thus the keyphrase cipher may owe its existence to Poe’s creating it as a result of a mistake!

  22

  RUMRUNNERS, BUSINESSMEN, AND MAKERS OF NONSECRET CODES

  THE STAR PROSECUTION WITNESS in the New Orleans Federal Court on May 2, 1933, was a new kind of detective. Instead of tailing suspects through the mazes of the underworld, she tracked down letters through the tortuous labyrinths of code and cipher. Instead of dusting surfaces for fingerprints, she applied sensitive analytical tests that developed traces of plaintext. Yet her evidence could be fully as incriminating as the gumshoe methods of ordinary police work. Mrs. Elizebeth Smith Friedman, a cryptanalyst for the Coast Guard, was about to testify to her solutions of coded messages of the Consolidated Exporters Company, Prohibition’s largest and most powerful bootlegging ring—messages that at last connected the ringleaders to the actual operations of the rumrunning vessels.

  It was not the first time Mrs. Friedman had done this. As the national thirst grew during year after year of Prohibition, as speakeasies sprang up and disregard for the law became rampant, geniuses of crime battened on the illegal demand. Small-time hoodlums like Al Capone burst forth as big businessmen. Whole syndicates, rivaling in intricacy and geographic dispersion the industrial giants of America, came into being just for smuggling. Crime became organized—and the foundations laid in those days support the Mafia and the Cosa Nostra of today.

  The lessons of organization learned by the bootleggers on land were taught to the rumrunners at sea, who brought in from foreign countries the flow of liquor without which the whole criminal operation would dry up. While the hoodlums ashore had to contend with Prohibition agents of the Department of Justice, those afloat had to contend with the United States Coast Guard—charged, during Prohibition no less than before or after, with preventing smuggling into the United States. As the rumrunners became more numerous and better organized, they turned increasingly to radio to control their offshore fleets. Messages between ship and shore stations warned of Coast Guard countermeasures, told ocean-going ships where to meet the small fast craft that would run the liquor in to some secluded cove, ordered decoy tactics by one ship to let another slip past a picketing Coast Guard patrol, reported that a Coast Guard ship was tailing a rumrunner and advised that no speedboats be sent out, and in general coordinated rumrunning activities in a highly efficient and effective manner.

  Naturally
their messages were coded. But although Coast Guard radiomen had long been intercepting them and forwarding them to headquarters, no law-enforcement agency could break them down. By April of 1927, hundreds of intercepts had accumulated in the Coast Guard intelligence office files.

  Then Commander Charles S. Root conferred with the Bureau of Prohibition, and Prohibition employed Mrs. Friedman and established her in the Coast Guard office in Washington. At the same time, Prohibition furnished the personnel and the Coast Guard the equipment for two intercept stations, one in San Francisco, one in Florida, to assure a continuing flow of material. Within two months Mrs. Friedman had solved the bulk of the messages; she then began concentrating on intercepts from Coast Guard stations on the West Coast that dealt with current activities. Most of it emanated from two rival rum fleets—the giant Consolidated Exporters Corporation, whose messages she was to confront again and again in her work, and the so-called Hobbs interests, both operating out of Vancouver. She solved the messages in Washington and forwarded the results to the Pacific, by telegraph if they were of immediate value, by airmail otherwise. To save time and make intercepts immediately available in plain language at the point where they were needed, Mrs. Friedman went to the West Coast in June of 1928 to teach C. A. Housel, of the office of the Coordinator of the Pacific Coast Details, how to decrypt rumrunners’ messages in systems that Mrs. Friedman had already solved. Housel proved himself apt and industrious, and in the next 21 months he handled 3,300 messages between four or five shore stations and two dozen vessels.

  As a result of the information obtained from cryptanalysis and from direction-finding, the Coast Guard put increasing pressure on the smugglers’ activities. Evidently the bootleggers discovered the weakness of their wireless operation, particularly their codes and ciphers, for in two years their radio and cryptographic organizations ramified at an enormous rate. Whereas in 1927-28 only two general systems were in use, changed only every six months, in mid-1930 practically every rum boat on the Pacific Coast had its own code or cipher. In May of 1930, for example, the Consolidated Exporters Corporation, with three shore stations, employed a different cryptosystem from its headquarters to each of its “blacks,” or rumrunning craft, while the mother ship corresponded with these blacks in an entirely different system. In the fall of 1929, this giant, which had gobbled up most of its competition in the Pacific, established a branch in Belize, British Honduras. Traffic in this Gulf Coast branch rapidly climbed to several hundred cryptograms a month. On the Atlantic side of Florida, 25 cryptograms a day were intercepted, while in the New York region, in only five days in February of 1930, a radio inspector heard no fewer than 45 unlicensed stations from within ten miles of New York. They were involved in operations from Nova Scotia to the Bahamas. It was reported that one syndicate paid its radio expert $10,000 a year—this during the Depression! A retired lieutenant commander of the Royal Navy devised the systems for Consolidated Exporters’ Pacific operation, though its Gulf and Atlantic groups made up their own as needed.

  His name was unknown, but his cryptologic expertise was apparent. The smugglers’ systems grew increasingly more complicated. “Some of these are of a complexity never even attempted by any government for its most secret communications,” wrote Mrs. Friedman in a report in mid-1930. “At no time during the World War, when secret methods of communication reached their highest development, were there used such involved ramifications as are to be found in some of the correspondence of West Coast rum running vessels.” One such system, employing two different commercial codes, passed through five steps: The clerk (1) encoded the plaintext in the commercial ABC Code, 6th edition, (2) added 1000 to the numerical codegroup, (3) looked up this codenumber in another commercial code, the Acme, (4) transcribed the codeword opposite that codenumber, and (5) enciphered that codeword in a monalphabetic substitution. Much of this complexity, however, was vitiated by the clerk’s habit of only partially encoding messages and enciphering the rest in a monalphabetic substitution that appears to have been the same as for the code. Mrs. Friedman illustrated the process with an actual message (which may have some slight errors in it):

  Mrs. Friedman solved the system. “In this case,” she wrote, “an inspection determined that the system employed came under the general classification of Enciphered Code. Then began what seemed endless experimentation to determine the particular type of enciphered code. There are hundreds of public codes any one of which might have been used, and in order to discover which, it was necessary to solve the cipher applied. With enormous difficulty the cipher alphabet was built up, by which the groups actually appearing in the messages were resolved into code groups of the Acme Code. But as this resulted in no intelligible meaning, it was obvious that further steps were necessary in order to reach clear language. The processes of experiment continued, the search among hundreds of code books was again prosecuted, and finally the whole laborious process was revealed.”

  In her office—first in a building near the Bureau of Printing and Engraving, then in a building on Pennsylvania Avenue opposite the Willard Hotel—Mrs. Friedman solved 12,000 messages in just her first three years for the Coast Guard, the Bureau of Customs, the Bureau of Narcotics, the Bureau of Prohibition, the Bureau of Internal Revenue, and the Department of Justice. About the same number of messages were examined and discarded. In October and November of 1929, she spent a month in Houston, Texas, solving a mass of smuggling traffic subpoenaed by the United States Attorney. Among the approximately 650 messages in 24 systems were some that were to play a role in a case world-famous in international law.

  Complexity of the code organization of a bootlegging sundicate during prohibition. “Black” means a rumrunning ship

  Her solutions made it clear to the Coast Guard that obtaining this information on a current basis would enable the government agencies to take action that would prevent smuggling. It therefore embarked upon an experiment unique in the annals of cryptology and criminology: a floating cryptanalytic crime-detection laboratory. This was CG-210, a 75-foot patrol boat commanded by Lieutenant Frank M. Meals, a former telegraph operator and radioman who in 1924 had compiled, with civilian employee Robert T. Brown, the Coast Guard’s first codebook. CG-210 was specially outfitted with a battery of high-frequency receivers, direction-finders, and a cryptanalyst—none other than William F. Friedman, lent by the Army for two weeks of nautical codebreaking. Between September 14 and 27, 1930, Friedman solved the code used by a group of smugglers operating off New York and read the operating orders to their ships, completely preventing them from transferring any liquor to shore for several days. “The resulting confusion to this group of rum ships was more than all the efforts of the destroyer force and the other units combined have been able to effect in months—and it should be remembered that this was accomplished by a single patrol boat with nine men aboard which never went near ‘rum row,’ ” wrote Lieutenant Commander F. J. Gorman, head of the Coast Guard’s intelligence office. In addition, CG-210 located an unlicensed radio station in New Bedford, Massachusetts, used to control the rum ship Nova V; this was raided by Justice and Commerce department officials, and the operator, Joseph Travers, found guilty of illegal transmissions, largely on the basis of the cryptanalyzed evidence.

  These spectacular results led the Coast Guard to concentrate even more on the bootleggers’ communications, the weakest link in the criminal chain of operations. Intelligence office chief Gorman wrote: “This intercepted material contains much of the information that the investigative agencies of the Customs and Justice are after and practically all of the plans, including contact points, to obtain which the Coast Guard vessels cruise endlessly.” In 1930, the Coast Guard established a radio-intelligence unit under the command of Lieutenant Meals. The unit, attached directly to headquarters for freedom of action, covered the entire Atlantic Coast. At New York six commissioned and five warrant officers learned radio-intelligence work. Eventually four more 75-foot patrol boats, fitted out like CG-210, put
to sea with a commissioned officer in command, a cryptanalyst, and six radiomen to combat the rumrunners’ communications.

  The cryptanalysts who went to sea did not, however, have either the experience or the material with which to undertake the long and difficult solution of some of the systems employed by the rumrunners. Their work was primarily current, involving perhaps the stripping of a superencipherment from a rumrunner code. Unfortunately, the headquarters cryptanalytic unit—on which the whole radio-intelligence operation depended—consisted of only Mrs. Friedman and a clerk. She explained in a memorandum how this situation cramped Prohibition law enforcement:

  For the past several years intercept activities maintained both by the Coast Guard and by other agencies concerned with the enforcement of the law regarding smuggling have yielded a very large volume of communications passing between shore stations and ships engaged in smuggling. With the extremely limited personnel available for work in connection with the solution of this intercepted traffic little has been accomplished compared with what might have been and still might be accomplished were an adequate and trained force available for solution activities. For the most part the smugglers use extensive code books which they usually compile or have compiled for them by code firms. From a technical point of view the solution of code messages is much more difficult and requires much greater time and effort than does the solution of ciphers. Moreover, in the case of code the mere breaking down of the basic system is only the beginning of the work, because, unlike cipher systems, the solution of one message discloses very little about the remaining messages. The solution of code is a long-drawn-out process, which must be continued through the life of the code, if all messages are to be read. It may be stated that every system employed by the smuggling interests has been solved but in no case has it been possible to read all of the messages in view of the large amount of labor involved and the lamentable lack of personnel to accomplish the work.