by DAVID KAHN
Later the Decyphering Branch read correspondence between American ministers in London, The Hague, and Berlin—in the latter city a future President named John Quincy Adams—between July, 1798, and February, 1800. The letters, enciphered in a homophonic substitution, seem to have been solved in retrospective solutions. In 1841, Britain lifted the flap of a two-part U.S. nomenclator and peeked at the American minister to Spain reporting on the successful conclusion of financial negotiations with that country.
By then the Decyphering Branch had only two members. Of the three grandsons of Bishop Willes who had joined in the 1790s, just one, Francis Willes, carried on the family tradition. His brother William had assisted only briefly, and his other brother, Edward, had died in 1812, the last to hold the title of Decypherer. Francis became so overworked that he enrolled his nephew, the Reverend Mr. William Willes Lovell (apparently no relation to James Lovell) as assistant.
Nor was France idle. On September 26, 1812, the American minister, writing to President Madison, carefully enciphered the names of two French officials who were backing his claims against Napoleon and made a point of asking the President to keep them confidential. But the cryptologic descendants of the Rossignols had their own way of finding out for the Little Emperor that the two were Cambacérès and Talleyrand.
These were the dying gasps of the black chambers. The winds of change, stirred up in part by the example of the nation whose codes were being solved, in part by the machines of the Industrial Revolution, freshened into the political gales of the 1840s which blew down most of Europe’s remaining absolution and the totalitarian agencies that propped it up. Europe’s new birth of freedom tolerated no government opening of mail. In England, a tremendous public and parliamentary outcry over the surreptitious opening of letter forced the government to discontinue the interception of diplomatic correspondence in June of 1844. That October the government dissolved the Decyphering Branch, pensioning off Willes and Lovell. In Austria the Geheime Kabinets-Kanzlei closed its doors in 1848. In France, the Cabinet Noir, which had been withering ever since the Revolution, passed away as well in that convulsive year. And in that same decade, the same vast social forces that ended the era of the black chambers simultaneously fostered an invention that transformed cryptography.
Decyphering Branch solution of a 1798 diplomatic dispatch to John Quincy Adams, American Minister in Berlin
* Lynceus was an Argonaut whose glances were so piercing that they penetrated the bowels of the earth.
†
Bellona was the Roman goddess of war.
* One story about Rossignol should be deflated, however. This is that his solutions were made “in a fashion so marvelous to his contemporaries that the device with which a lock is opened when the key has been lost is still called in French a rossignol.” While the fact of the current usage is true, its implied origin is false. Unfortunately for so charming an etymology, this particular use of the term rossignol appears as criminal argot in police documents as early as 1406—almost two centuries before the cryptologist was born. Since the word also means “nightingale,” it may be possible that the thieves adopted it as slang for a picklock because its nighttime solos of clicks and rasps were music to their ears.
* This activity forms the legal precedent for the modern tapping of telephones, at least in Britain. Significantly, however, the source of the power to intercept communications has never been determined. The Crown simply exercised it and, despite occasional debate, has continued to do so, presumably with the tacit approval of the public as necessary for the safety of the state.
6
THE CONTRIBUTION OF THE DILETTANTES
THE TELEGRAPH made cryptography what it is today. Samuel F. B. Morse sent “What hath God wrought!” in 1844. The next year his lawyer and promotional agent, Francis O. J. Smith, published a commercial code entitled The Secret Corresponding Vocabulary; Adapted for Use to Morse’s Electro-Magnetic Telegraph, in whose preface he declared that “secrecy in correspondence, is far the most important consideration.” This was provided by a superencipherment. Nine years later, an article on telegraphy in England’s Quarterly Review likewise emphasized the primacy of security:
Means should also be taken to obviate one great objection, at present felt with respect to sending private communications by telegraph—the violation of all secrecy—for in any case half-a-dozen people must be cognizant of every word addressed by one person to another. The clerks of the English Telegraph Company are sworn to secrecy, but we often write things that it would be intolerable to see strangers read before our eyes. This is a grievous fault in the telegraph, and it must be remedied by some means or other…. At all events, some simple yet secure cipher, easily acquired and easily read, should be introduced, by which means messages might to all intents and purposes be “sealed” to any person except the recipient.
As the most exciting invention of the first half of the century, the telegraph stirred as much interest in its day as Sputnik did in its. The great and widely felt need for secrecy awakened the latent interest in ciphers that so many people seem to have, and kindled a new interest in many others. Dozens of persons tried to dream up their own unbreakable ciphers. Nearly all were amateurs, the professionals (except for a few code clerks) having lost their jobs when the black chambers were abolished. A surprising number of these dabblers were intellectual and political leaders of the day who beamed their powerful and original minds on the engrossing field of cryptology. Their contributions enriched it with dozens of new cipher systems.
As businessmen and the public used the telegraph more and more, they found that their fears about lack of privacy were exaggerated. The clerks dealt impersonally with the messages. The telegraph companies respected their confidentiality. And commercial codes like Smith’s, which replaced words and phrases by single codewords or codenumbers to cut telegraph tolls, afforded sufficient security for most business transactions by simply precluding an at-sight comprehension of the meaning. The brokers and traders soon realized that the main advantage of these codes was their economy.
Smith’s pioneering venture was followed by dozens, then scores, then hundreds of commercial codes. Though a few had as many as 100,000 entries and some specialized ones only a few hundred, considerations of optimum manageability and selling price concentrated most in the neighborhood of Smith’s 50,000-entry size. They improved on his in two ways. They provided dictionary words as codewords instead of the letter-and-number groups that he had used. It was easier, cheaper, and more reliable to send ALBACORE to mean alone than the A. 1645 of Smith’s Vocabulary, And they greatly increased the number of phrases, thereby raising their toll-saving potential. Smith listed only 67 phrases, collected on a single page, compared to his 50,000 words; later codes had 10 or 20 times as many phrases as individual words.
Government ministries used the telegraph, too. At first they must have encoded with their nomenclators. But although secrecy was paramount for them, they liked the telegraphic economy of a large code—especially as they telegraphed more and more. So when the time arrived to compile a new nomenclator, they abandoned that form, copied the commercial form, and produced a full-fledged code. The nomenclators had had their 1,-or 2,000 codenumbers in mixed order, but the war and foreign ministries balked at the expense of drawing up a 50,000-entry code in two parts, and they had no professional cryptanalysts to warn them of the danger of the one-part format. They relied for security upon small editions, big safes, extensive lexicon (large codes are harder to break than small ones, other things being equal), and superencipherment, retaining codenumbers to facilitate this instead of switching to codewords. This evolution was essentially complete by the 1860s. The large, one-part code had replaced the small, two-part nomenclator in high-level military and diplomatic cryptography.
Meanwhile, the telegraph, author of this development, was creating something new in war—signal communications, or voluminous command and reconnaissance messages. Of course such messages had existe
d before, with torches, pigeons, and couriers, but in so rarefied a form that they were not even called “signal communications.” The telegraph enabled commanders, for the first time in history, to exert instantaneous and continuous control over great masses of men spread over large areas. It filled a need, for universal military conscription had begun to raise such armies to fight the nationalistic warfare of democracies (as contrasted with the small, professional armies that fought the dynastic wars of kings), the new railroad transported these large forces rapidly over great distances, and the industrial society supported them. These developments, together with the breech-loading gun, brought about the era of modern warfare.
No longer could a general sit horseback atop a hill and survey the battle, like Napoleon or Hannibal, sending messengers to hand-carry instructions to wheel or to counterattack. The forces engaged were too numerous, the field too vast. He had to work from a command post far in the rear, following the progress of the battle by telegraph on maps that showed more than his naked eye could ever see. He could issue orders by telegraph that would coordinate the movement of one out-of-sight wing with that of another, bring up reserves to block an enemy charge, order up food and ammunition in a hurry. The number of messages grew correspondingly. The command post became virtually a communications center.
These tactical messages required protection: telegraph wires could be tapped. Neither the old nomenclator nor the new code would do. They were too easy to capture in combat, too hard to reissue quickly and frequently to the numerous and widespread telegraph posts. Signal officers turned away from them. They looked instead to that neglected child of cryptography, the cipher. Ciphers could be printed cheaply on a single sheet of paper and distributed with ease. Secrecy was based upon variable keys, so capture of the general system and even of one of the keys would not compromise all an army’s secret messages. Solutions would be prevented by rapid key changes. Ciphers were ideal for battle-zone messages, and the first of the modern wars, the American Civil War, used them for just that. Thus was born a new genre in cryptography: the field cipher.
The first one was waiting in the wings. This was polyalphabetic substitution, in the form of the straight-alphabet Vigenère with short repeating keyword. The old objections to its use, which boiled down to the impossibility of correcting a garbled dispatch quickly enough, vanished with the telegraph. It fulfilled the requirements of noncompromisability of the general system and of ease of key changes. Moreover, it had the reputation of being unbreakable—which, if its cryptograms were not divided into words, it largely was. The military adopted it at once.
Then, in 1863, a retired Prussian infantry major discovered the general solution for the periodic polyalphabetic substitution. At one stroke he demolished the only impregnable structure in cryptography. Signal officers, compelled to provide secure communications, hunted frantically for new field ciphers. They found many good ideas in the writings of the dilettante cryptographers who had proposed ciphers for the protection of private messages. Soon some of these systems were serving in the various armies of Europe and the Americas. More ideas came from army officers who had studied cryptography in the courses in signal communication that the national military academies, such as St. Cyr, had added in the mid-1800s. Inevitably, crypt-analysts—who were either amateurs or soldiers with a professional interest, for full professionals there were none—replied with new techniques for breaking the new ciphers. From the slow crawl of nomenclator days, when the introduction of a special group meaning Disregard the preceding group would constitute a remarkable technical advance, the race between offense and defense in cryptology accelerated to its modern pace.
The history of cryptology from the decade that saw both the death of the black chambers and the birth of the telegraph to World War I is thus a story of internal development. Without Rossignols or Willeses, and without major wars or diplomatic struggles, cryptology could not influence world events, and, except for one or two unusual cases, it did not. The telegraph launched this evolution of cryptology. It broke the monopoly of the nomenclator. The nomenclator had reigned for 450 years as a general, all-purpose system, but it could not meet the new requirements either of high-level diplomatic or military communications or of low-level signal communications, which the telegraph had engendered. Each called for its own kind of cryptosystem, a specialized one. Signal officers ranked these systems in a hierarchy, rising from the simple and flexible and easily solved to the extensive and hard to solve. The telegraph thus stimulated the invention of many new ciphers and, by reaction, many new methods of cryptanalysis, and compelled their arrangement in a scale of complexity.
Many of these ciphers and techniques have become classic and are in use today. Moreover, cryptography still functions through a hierarchy and employs a multitude of special systems. The telegraph thereby furnished cryptography with the structure and the content that it still has. It made it what it is today.
All these things have antecedents, and just as the telegraph itself did, so were there precursors of the cryptographic systems that it engendered. What may be the earliest printed forerunner of the codes of today appeared at Hartford in 1805. A Dictionary; to Enable Any Two Persons to Maintain a Correspondence, With a Secrecy, Which Is Impossible for Any Other Person to Discover was a small book listing words and syllables in alphabetical order; these were to be numbered serially by the correspondents, omitting one number in every ten so that no two sets of correspondents would have the same code equivalents.
One cipher system invented before the telegraph was so far ahead of its time, and so much in the spirit of the later inventions, that it deserves to be classed with them. Indeed, it deserves the front rank among them, for this system was beyond doubt the most remarkable of all. So well conceived was it that today, more than a century and a half of rapid technological progress after its invention, it remains in active use.
But then it was invented by a remarkable man, a well-known writer, agriculturalist, bibliophile, architect, diplomat, gadgeteer, and statesman named Thomas Jefferson. He called it his “wheel cypher,” and it seems likely that he invented it either during 1790 to 1793 or during 1797 to 1800. During the first period he was America’s first Secretary of State, and the responsibilities of conducting foreign policy, the need to protect communications from England and France, the divided American cabinet, the spirit of invention that he found as administrator of the patent law, all spurred his own natural inventiveness; he was then also in contact with Dr. Robert Patterson, a mathematician of the University of Pennyslvania and vice president of the American Philosophical Society, who was interested in ciphers. During the later period, he was again in close contact with Patterson, who in 1801 sent him a cipher. Jefferson’s explanation of the wheel cypher is characteristically clear and economical:
Turn a cylinder of white wood of about 2. Inches diameter & 6. or 8.1, long, bore through it’s center a hole sufficient to recieve an iron spindle or axis of 1⁄8 or 1⁄4.I. diam. divide the periphery into 26. equal parts (for the 26. letters of the alphabet) and, with a sharp point, draw parallel lines through all the points of division from one end to the other of the cylinder, & trace those lines with ink to make them plain. then cut the cylinder crosswise into pieces of about 1⁄6 of an inch thick, they will resemble back-gammon men with plane sides, number each of them, as they are cut off, on one side, that they may be arrangeable in any order you please. on the periphery of each, and between the black lines, put all the letters of the alphabet, not in their established order, but jumbled & without order, so that no two shall be alike. now string them in their numerical order on an iron axis, one end of which has a head, and the other a nut and screw; the use of which is to hold them firm in any given position when you chuse it. they are now ready for use, your correspondent having a similar cylinder, similarly arranged.
Suppose I have to cypher this phrase. “Your favor of the 22d. is recieved.” and so on till I have got all the words of the phrase arranged in one line
, fix them with the screw, you will observe that the cylinder then presents 25. other lines of letters, not in any regular series, but jumbled, & without order or meaning. copy any one of them in the letter to your correspondent. when he recieves it, he takes his cylinder and arranges the wheels so as to present the same jumbled letters in the same order in one line. he then fixes them with his screw, and examines the other 25. lines and finds one of them presenting him these letters: “your favor of the 22 is recieved.” which he writes down. as the others will be jumbled & have no meaning, he cannot mistake the true one intended. so proceed with every other portion of the letter. numbers had better be represented by letters with dots over them; as for instance by the 6. vowels & 4. liquids. because if the periphery were divided into 36. instead of 26. lines for the numerical, as well as alphabetical characters, it would increase the trouble of finding the letters on the wheels.
When the cylinder of wheels is fixed, with the jumbled alphabets on their peripheries, by only changing the order of the wheels in the cylinder, an immense variety of different cyphers may be produced for different correspondents. for whatever be the number of wheels, if you take all the natural numbers from unit to that inclusive, & multiply them successively into one another, their product will be the number of different combinations of which the wheels are susceptible, and consequently of the different cyphers they may form for different correspondents, entirely unintelligible to each other….
Jefferson went on to say that if the cylinder be six inches long (“which probably will be a convenient length, as it may be spanned between the middle ringer & thumb of the left hand, while in use”) the number of wheels would total 36, and the number of ways in which they can be strung on the spindle to form different ciphers for different correspondents would come to 36 factorial, or 1× 2×3×… ×35×36, which Jefferson calculated almost exactly as “372 with 39 cyphers [zeros] added to it.” In fact, 36 factorial is 371,993,326,789,901,217,467,999,448,150,835,200,000,000.
