by DAVID KAHN
The time and place of the writing of that “Traitté,” the author’s failure with polyalphabeticity, and his allegiance to Spain make it probable that he was a cryptanalyst named Martin, who figured in an incident that shows how rare and fortuitous was the solution of a polyalphabetic. The Cardinal de Retz, that liberal and popular French prelate-politician, narrated in his Mémoires how he escaped from the chateau of Nantes on August 8, 1654, after two years of political imprisonment. He digressed to discuss ciphers:
I had one with Madame La Palatine, which we called The Indecipherable, because it always seemed to us that no one could penetrate it without knowing the word that had been agreed upon. We placed such complete confidence in it that we never hesitated to write freely and to send the most important and the most confidential secrets by ordinary courier. It was in this cipher that I wrote to the Premier President [of the Parlement of Paris] that I would escape on August 8 … The Prince [of Condé], who had one of the best decipherers in the world, named, it seems to me, Martin, held this cipher six weeks with me in Brussels.* And he told me that Martin had confessed to him that it was indecipherable…. It was broken down sometime afterwards by [Guy] Joly [a counselor to the Châtelet tribunal in Paris and one of Retz’s followers], who, though not a professional decipherer, hit upon its key while reflecting on it and brought it to me at Utrecht, where I was at the time.
Retz was trying to show “how little confidence one can place in ciphers,” but the fact that it took a lucky guess by an intimate to effect the only solution in six years seems rather to enhance the cipher’s value.
The most interesting polyalphabetic solution of the nomenclator years came a century later. Its interest derives from a cryptanalyst who has become a very prototype in a field utterly removed from cryptanalysis, and whose obsession with that field was such that he even turned cryptanalysis to account in it.
It all happened in 1757, when he was talking about magic, alchemy, and chemistry with his friend, the wealthy Madame d’Urfé. She showed him a cipher manuscript describing the transmutation of baser metals into gold, and told him that she did not need to keep it locked up because she alone held the key. She gave it to him, remarking that she did not believe in cryptanalysis. “Five or six weeks later,” he stated in his memoirs, “she asked me if I had deciphered the manuscript which had the transmutation procedure. I told her that I had.” But Madame d’Urfé, still skeptical, replied:
“Without the key, sir, excuse me if I believe the thing impossible.”
“Do you wish me to name your key, madame?”
“If you please.”
I then told her the word, which belonged to no language, and I saw her surprise. She told me that it was impossible, for she believed herself the only possessor of that word which she kept in her memory and which she had never written down.
I could have told her the truth—that the same calculation which had served me for deciphering the manuscript had enabled me to learn the word—but on a caprice it struck me to tell her that a genie had revealed it to me. This false disclosure fettered Madame d’Urfé to me. That day I became the master of her soul, and I abused my power. Every time I think of it, I am distressed and ashamed, and I do penance now in the obligation under which I place myself of telling the truth in writing my memoirs.
But this did not stop him at the time from amazing the lady with some hocus-pocus in producing the keyword (NABUCODONOSOR, an Italian spelling of “Nebuchadnezzar”), and then taking his leave “bearing with me her soul, her heart, her wits and all the good sense that she had left.”
The cryptanalyst? Casanova.
Less dramatic solutions of polyalphabetics occurred early in the 1800s before a retired German infantry major published the general solution in 1863. It may seem that so many solutions should have dispelled the myth of polyalphabetic unbreakability. But they were isolated instances, scores of years apart, so unusual that standard works on cryptology do not mention them. Polyalphabetics remained freaks of cryptologic usage. The professionals avoided them. Their very unpopularity protected them. Had they been used more, perhaps the coincidences that lit the way to the general solution would have forced themselves upon cryptologists. But the world fixated upon the nomenclator, and so the legend of unbreakability flourished.
It was fed by the lesser writings of the time. These books shed no new light on polyalphabetics and none on the political cryptology of their day. They are divorced from the realities, and generally content themselves with commentaries on earlier works, chiefly Trithemius, and with describing a few trivial inventions. Neglect justly entombs most. A few are of minor interest.
The Florentine Jacopo Silvestri published the second printed book on cryptology at Rome in 1526. His Opus novum … begins with a Dantesque scene of the author fleeing the plague at Rome to a small country estate near the Tiber. There he received a visit from an Etruscan friend, who, discoursing with him on ancient modes of writing, discussed cryptology. Silvestri’s friend begged him to write down his knowledge of it for universal advantage. But most of the 88 pages of the Opus novum are merely given over to a vocabulary that can serve as the basis for a small code.
In 1624, Augustus II, Duke of Braunschweig-Lüneberg (afterwards Hanover) in Germany, issued his Cryptomenytices et Cryptographiae libri IX under the pseudonym Gustavus Selenus. This was a play on his name, GUSTAVUS being an anagram (with the interchangeable u and v of the time) of Augustus, and Selene, the Greek goddess of the moon, which is “luna” in Latin, standing for Lüneberg. The duke, who was cousin to the grandfather of George I of England, is probably the highest ranking author of a book on cryptology; both he and the present queen of England descend from Ernest the Confessor, of the house of Guelph. He prefaced the almost 500 small-folio pages of his volume with 17 pages of tributes from his courtiers (“As, what night in dusty cloak conceals, bright Cynthia soon with torch full-flaming shows,/So, too, Gustavus now, Selenus called, uncovers things that time has long in shadow held”). One such, a particularly laudatory one entitled a “Sportive Poem,” was contributed to this volume of the supposedly unknown Selenus by none other than the gracious Duke Augustus himself! But the work, while containing some cipher systems, mainly defends the occultism of Trithemius.
The most celebrated scholar of his day, the Jesuit Athanasius Kircher, who had won fame by his “solution” of hieroglyphics and by his having been lowered into the crater of Vesuvius to study underground forces (a feat that, with a book on the subterranean world, won him the title of Father of Vulcanology), published his Polygraphia nova et universalis at Rome in 1663. The book contains chiefly processes of encipherment, as well as a multilingual, cross-indexed code which is one of the earliest essays at a universal language. Two years later, his student, Gaspar Schott, a Jesuit physicist, brought out Schola steganographia at Nuremberg. Schott’s book, like his teacher’s, is largely a compilation of cipher systems.
Only two English works of the period merit attention. The first book in English on cryptology appeared anonymously in 1641, but Mercury, or the Secret and Swift Messenger was the offspring of John Wilkins, a “lustie, strong growne … broad shouldered” young chaplain who later married Oliver Cromwell’s sister and became Bishop of Chester and a founder and first secretary of the Royal Society. A succinct volume, very well grounded in the classics, Mercury introduced the words cryptographia (defined by Wilkins as “secrecy in writing”) and cryptologia (“secrecy in speech”) into English. The author reserved the term cryptomeneses, or “private intimations,” for the art of secret communication in general. In addition to summing up the knowledge of the time, Wilkins depicted three kinds of geometrical cipher, a mystifying system in which a message is represented by dots, lines, or triangles. The letters of the alphabet, in normal or mixed order, were written out at known spatial intervals; this served as the key. This line of letters was held at the top of a sheet of paper, and the message was spelled out by marking a dot for each plaintext letter underneath tha
t letter in the key alphabet, each dot lower than its predecessor. The dots could then be connected by twos to form lines, by threes to form triangles, or all together to form what would look like a graph—or they could be left as dots. The receiver, who had an identically proportioned key, noted the positions of the dots, the ends of the lines, or the apexes of the triangles against the alphabetical scale to read the plaintext.
The second English book on the subject excelled. Cryptomenytices Patefacta was written by John Falconer, about whom nothing is known except that he was a distant relative of the Scottish philosopher David Hume, was reportedly entrusted with the private cipher of the future King James II, and died in France while following James into temporary exile there. The book came out posthumously in 1685, with its author listed only as “J.F.” It proved so popular that it was reissued in 1692 with a new title page that clearly indicates just what its 180 pages comprise: Rules for Explaining and Decyphering all Manner of Secret Writing….
Falconer’s cryptanalytical bias sharpened his comments on the standard systems, and led him to make a praiseworthy assault on that old bugbear, polyalphabetic substitution. He suggested guessing at the short words in a cryptogram, deducing the keyletters (these were standard alphabets), and seeing whether “they can be joyned to make up part of the Key.” Knowing the number of letters in the key is a great help, he says, “since thereby you have the several Returns of each Alphabet.” The technique is quite valid for cryptograms with word divisions, and bespeaks an acute mind. Falconer also gave what seems to be the earliest illustration of keyed columnar transposition, a cipher that is today the primary and most widely used transposition cipher, having served (with modifications) for French military ciphers, Japanese diplomatic superencipherments, and Soviet spy ciphers.
These five books, plus the even less important ones that were also published at this period, have—with the possible exception of Falconer’s—a certain air of unreality about them. There is good reason for this. The authors borrowed their knowledge from earlier volumes and puffed it out with their own hypothesizing, which seems never to have been deflated by contact with the bruising actuality of solving cryptograms that they themselves had not made up. The literature of cryptology was all theory and no practice. The authors did not know the real cryptology that was being practiced in locked rooms here and there throughout Europe, by uncommunicative men working stealthily to further the grand designs of state.
* In Alberti’s disk, the outer capital letters are the plaintext and the inner lower-case letters are the ciphertext. This contradicts the convention of this book, and is being used in the section on Alberti only to avoid altering his text. The difference is signalized by not using italic for the lower case.
* In 1564, Bellaso (now spelling his name with a double l) published a third edition of his booklet in which he described a form of autokey without these obstacles. It keys the first letter of the message with the first alphabet and successive letters of the first word with successive alphabets. Then it keys the first letter of the second word with the first letter of the first word, successive letters being keyed with the succeeding alphabets. This procedure—partly autokey, partly progressive—is repeated to the end of the message.
* The key in full is actually IN PRINCIPIO ERAT VERBUM. The third-to-last letter of the ciphertext, C, should be I.
* A term meaning polyalphabetics, derived from the two “keys” needed—one the cipher alphabet, the other the keyword or keyphrase. It survives in modern French usage as “double-key cipher.”
* Retz visited Condé in Brussels in 1658.
5
THE ERA OF THE BLACK CHAMBERS
RÉALMONT was under siege. The royal army, under Henry II of Bourbon, Prince of Condé, had invested it at dawn Wednesday, April 19, 1628. But the Huguenots, inside the battlements of the little town in southern France, were putting up a stiff defense. They cannonaded Condé from a tower and contemptuously rejected his demands that they surrender, saying that they would die instead. Condé brought up five big cannon from Albi, a dozen miles away, and on Sunday ranged them in an ominous line facing Réalmont.
That same day his soldiers captured an inhabitant of the town who was trying to carry an enciphered message to Huguenot forces outside. None of Condé’s men could unriddle it, but during the week the prince learned that it might be solved by the scion of a leading family of Albi who was known to have an interest in ciphers.
Condé sent him the cryptogram. The young man solved it on the spot. It revealed that the Huguenots desperately needed munitions and that, if they were not supplied, they would have to yield. This was news indeed, for despite the destruction of a number of houses by the Catholic batteries, the town was continuing to resist stoutly with no sign of surrender. Condé returned the cryptogram to the inhabitants, and on Sunday, April 30, 1628, though its fortifications were still unbreached and its defenses still apparently adequate for a long siege, Réalmont suddenly and unexpectedly capitulated. With this dramatic success began the career of the man who was to become France’s first full-time cryptologist: the great Antoine Rossignol.
When word of the incident reached Cardinal Richelieu, the astute and able Gray Eminence of France, he at once attached this useful talent to his suite. Rossignol proved his worth almost immediately. The Catholic armies under Richelieu surrounding the chief Huguenot bastion of La Rochelle intercepted some letters in cipher, which the young codebreaker of Albi read with ease. He told His Eminence that the starving citizens were eagerly awaiting help that the English had promised to send by sea. When the fleet arrived, the primed guardships and forts so intimidated it that it stood off the port’s entrance and made no serious attempt to force a passage. A month later, the city capitulated in full sight of the English vessels—and the great French tradition of expertise in cryptology had been founded.
Rossignol very quickly established himself in the royal service. By 1630, his solutions had made him rich enough to build a small but elegant chateau at Juvisy, 12 miles south of Paris, later surrounding it with a charming informal garden designed by Le Nôtre, the gardener of Versailles. Here Louis XIII stopped to visit the young cryptanalyst in 1634, 1635 and 1636 on his returns to Paris from Fontainebleau.
In the swashbuckling court of that monarch, and then in the resplendent one of Louis XIV, Rossignol served with an extraordinary facility. The stronghold of Hesdin surrendered a week sooner than it otherwise would have because he solved an enciphered plea for help, and then composed a reply in the same cipher telling the townspeople how futile their hopes were. How many other towns he compelled to surrender, how many diplomatic coups he made possible, how many betrayals he uncovered among the great nobles in those days of shifting allegiances, he never discussed. This reticence caused some at the court to charge that he never actually solved a single cipher, and that the cardinal spread inflated rumors about his abilities to discourage would-be conspirators. But in fact Richelieu was frequently telling his subordinates such things as, “It is necessary to make use, in my opinion, of the letters of the man who has been arrested by the civil authorities at Mézières, that is to say, have them put into Rossignol’s hands to see if there is something important in them.” Or, eight years later, in 1642, writing to Messieurs de Noyers and de Chavigny: “I saw, in some extracts, that Rossignol sent me, a truce negotiation of the King of England with the Prince of Orange; I do not think that it can have any effect, but … it is up to you, gentlemen, to keep your eyes peeled.”
Louis XIII, on his deathbed, recommended Rossignol to his queen as one of the men most necessary to the good of the state. Two years later, on February 18, 1645, Richelieu’s successor, Cardinal Mazarin, named him a master of the Chamber of Accounts and a counselor of state. Like Richelieu, Mazarin himself sometimes sent him intercepts. In 1656, for example, he forwarded a letter of the Cardinal de Retz instructing Rossignol to solve it. Under Louis XIV, Rossignol often worked in a room next to the king’s study at Versailles. Fr
om here issued the streams of solutions that helped the Sun King direct the polity of France.
Rossignol had, at 45, improved his social position by marrying 23-year-old Catherine Quentin, the daughter of a nobleman and the niece of a bishop. Their marriage was a happy one, full of playfulness and endearments, and they had two children, Bonaventure and Marie.
One of their best friends was the poet Boisrobert, who originated the idea of the Académie Française. He loved to hold forth at the excellent Rossignol table, which he liked for its fine wines and Madame Rossignol’s charm as a hostess. (In a 13-line poem to her, he declared her friendship “sweeter than sugar with cream.”) When he found himself out of favor at court, he complained about his unhappiness in a poem to his influential cryptologist-friend. Rossignol showed it to Mazarin, who singled out Boisrobert at the next audience and praised the poem loudly. Boisrobert, delighted at this sign of favor, addressed a paean of thanks to Rossignol. Perhaps out of gratitude, he later praised him extravagantly in the first poem ever written to a cryptologist. Some of the choicer of the 66 lines of the untitled Épistre 29 in his Épistres en Vers read:
II n’est plus rien dessous les Cieux
Qu’on puisse cacher à tes yeux;
Et crois que ces yeux de Lyncée*
Lisent mesme dans la pensée.
Que ton service
est éclatant
Et que ton Art est important!
On gagne par luy des Provinces,
On sçait tous les secrets des Princes,
Et par luy, sans beaucoup d’efforts,
On prend les
villes & les forts.
Certes j’ignore ton adresse,
Je ne comprends point la finesse
De ton secret; mais je sçay bien
Qu’il t’a donné beaucoup de bien;
Tu le mérites, & je gage
Qu’il t’en
donnera davantage;
