by DAVID KAHN
After the Elector of Hanover succeeded to the English throne as George I in 1714, retaining the rule of the German state, the Decyphering Branch collaborated with the black chamber maintained at Nienburg by the Hanoverian government. Cryptanalysts Bode, Lampe, and Neubourg had even been imported from there—an ironic development in view of Wallis’ refusal to divulge his techniques to Hanover a few years earlier. Mail opening became habitual. George and his successors took a constant personal interest in the work, often encouraging talent with royal bounty. Correspondence was closely watched for cribs that were passed to the Decyphering Branch.
During the 1700s, the branch’s output averaged two or three dispatches a week, and sometimes one a day. Its cryptanalysts solved the dispatches of France, Austria, Saxony and other German states, Poland, Spain, Portugal, Holland, Denmark, Sweden, Sardinia, Naples and other Italian states, Greece, Turkey, Russia, and, later, the United States. The record of French interceptions covers two centuries and comprises five volumes of intercepts totaling 2,020 pages plus three volumes of keys. Perhaps more typical is the Spanish dossier—three volumes of intercepts from 1719 to 1839 totaling 872 pages. Not all of the messages were solved at the time of their interception. Many were held either until enough had accumulated for a successful attack or until a need arose for their solution.
The solutions were read by the king and a few of the top ministers. They warned the government of the intrigues of foreign rulers and ambassadors and of impending war. An intercepted message between the Spanish ambassadors in London and Paris clearly suggested that Spain had allied herself with France against England in the Seven Years’ War. It was read at the British cabinet meeting of October 2, 1761. The Great Commoner, William Pitt the Elder, cited it as support for his proposal that England take the initiative, declare war before Spain did, and capture the fleet of treasure ships then transporting gold to Spain from her American possessions. His counsel was rejected, and he resigned. The war came anyway—after the immense cargo of bullion had been unloaded at Cadiz.
Solution of a 1716 French dispatch by England’s Decyphering Branch
The success of the cryptanalysts of France and England was undoubtedly due in large measure to their skill. But, as always, there was another side to it which François de Callières pointed out in his superb little work on diplomacy. The cryptanalysts, he said, “owe the Esteem they have gain’d solely to the negligence of those who give bad Cyphers, and to that of Ministers and their Secretaries, who make not a right use of them.”
He omitted the important factor of economics. In England in the 1700s, the Decyphering Branch at first tested and, after 1745, prepared England’s diplomatic nomenclators. These generally had four-figure codegroups and numerous homophones; they were printed on large sheets and pasted on boards for the cipher clerks’ use. The cryptanalysts, who should have known, thought that it was “little less than impossible to find them out.” But their initial strength, which was due to the extent of the lexicon and the many homophones, eventually proved a weakness: the foreign service was reluctant to change a nomenclator that, in the late 1700s, cost £150, or to order separate nomenclators for separate countries. Thus some remained in use for a dozen years or more, and some simultaneously served several embassies. For example, in 1772 and 1773, Paris, Stockholm, and Turin had Ciphers and Deciphers K, L, M, N, O, And P; Florence had K, L, O, and P; Venice, K and L for use with Florence and M and N for other purposes; Naples, M and N; and Gibraltar, O and P.
But Callières correctly remarked failures to make a “right use” of ciphers. Time and again, diplomats enciphered documents handed to them by the governments to which they were accredited, giving those governments’ cryptanalysts ideal cribs. They repeated in clear dispatches sent in cipher. Because of language difficulties, they used foreigners in secret work. And often their chiefs simply did not want to believe in cryptanalysis because it meant more work for them. In 1771, for example, the French ambassador to England complained that he had only two old ciphers and that there was in London “a bishop [Willes] charged with the decipherment of the dispatches of foreign ministers who succeeds in finding the key of all ciphers.” His superior replied that not even a bishop could translate French ciphers, “which have no kind of system and of which it is not possible to find the key because one does not exist.” Then, after rather heavy-handedly pointing out that ciphers are sometimes compromised through the indiscretions of those who hold them, he added, “I do not believe in decipherers any more than in magicians.”
This sentiment found its most pointed expression in Voltaire’s remark that “those who boast of deciphering a letter without being instructed in the affairs of which it treats, and without having any preliminary help, are greater charlatans than those who boast of understanding a language which they have not even studied.” For once, one of his epigrams rang false.
Across the Atlantic, cryptology reflected the free, individualistic nature of the people from which it sprang. No black chambers, no organized development, no paid cryptanalysts. But this native cryptology, which had much of the informal, shirtsleeve quality of a pioneer barn-raising, nevertheless played its small but helpful role in enabling the American colonies to assume among the powers of the earth their separate and equal station. Indeed, the first incident occurred even before those colonies had declared their independence.
It started in August of 1775. A baker named Godfrey Wen wood was visited in Newport, Rhode Island, by a girl whom he had formerly known intimately. She asked him for help in getting in touch with some British officers so that she could give them a letter. Wenwood, a rebel patriot, grew suspicious. He persuaded her to give him the letter for delivery and to depart before his fiancée learned of her visit. But he did not forward it.
Instead, he consulted a schoolmaster friend. The friend broke the seal and found inside three pages covered with line after neatly printed line of small Greek characters, odd symbols, numbers, and letters. Unable to penetrate the mystery, he handed the missive back to Wenwood, who tucked it away while he considered the matter further. But soon thereafter he received a letter from the girl, who complained that “you never Sent wot you promest to send.” His suspicions now thoroughly aroused, Wenwood went up through channels and at the end of September was standing in the headquarters of Lieutenant General George Washington, showing him the letter.
The commander in chief could not read the cryptogram either, but he could question the girl. She was brought in that evening, and though, Washington said later, “for a long time she was proof against every threat and persuasion to discover the author,” intensive interrogation wore her down. The next day, she finally revealed that the letter had been given to her by her current lover, Dr. Benjamin Church, Jr. Washington was astounded. Church was his own director general of hospitals. A prosperous Boston physician who was a leader in the Massachusetts Congress and a colleague of Samuel Adams and John Hancock in the new House of Representatives, he had just asked to resign as hospitals chief. Washington had turned down his request because of his own “unwillingness to part with a good officer.” Could so distinguished a man be engaged in a clandestine and possibly traitorous correspondence? But he was brought in under guard.
Last lines of the cipher message of the Tory spy, Dr. Benjamin Church
The letter was his, he readily admitted, intended for his brother, Fleming Church, who was in Boston—though it was addressed to “Major [Maurice] Cane in Boston on his majisty’s service.” If deciphered, it would be found to contain nothing criminal. But though he repeatedly protested his loyalty to the Colonial cause, he did not offer to put the contents of his letter into plain language.
Washington cast about for someone who could solve it. He located the Reverend Mr. Samuel West, 45, a rather absentminded pastor who was, ironically, a Harvard classmate of Church’s. West, who later served as a delegate to the Constitutional Convention of 1787, was interested in alchemy and became convinced that prophetic portions of the Bibl
e predicted the course of events of the American Revolution.
When Washington’s need for a cryptanalyst became known, Elbridge Gerry, 31, chairman of the Massachusetts committee of military supply, volunteered his help. Gerry went on to greater fame as fifth Vice President of the United States and concocter of the political grotesquery known today as the “gerrymander.” He also suggested the name of Colonel Elisha Porter of the Massachusetts militia, who had been a year ahead of him at Harvard. Gerry and Porter teamed to attack the message, and West worked by himself through the night.
Washington received the two solutions of what proved to be a monalphabetic substitution on October 3. They were identical. Church was reporting to Thomas Gage, the British commander, on American ammunition supply, on a plan for commissioning privateers, on rations, recruiting, currency, a proposed attack on Canada, artillery that he had counted at Kingsbridge, New York, troop strength in Philadelphia, and the mood of the Continental Congress. It ended: “Make use of every precaution or I perish.”
This, to Washington, refuted Church’s protestations that he had deliberately transmitted the information to the redcoats to impress them with patriot strength and so deter them from attacking just when American ammunition was low. It also convinced most Colonial leaders of his guilt. “… what a complication of madness and wickedness must a soul be filled with to be capable of such perfidy!” ejaculated an angry Rhode Island delegate. And the paymaster general of the Continental forces commented, “I have now no difficulty to account for the knowledge Gage had of all our Congress secrets, and how some later plans have been rendered abortive.” It also developed that information furnished by Church caused Gage to send troops to Boston to capture American stores at Concord—a move that resulted in the historic clash at Lexington that began the American Revolution.
Church was imprisoned. The Massachusetts legislature expelled him. When he was paroled briefly, a mob assailed him. Congress rejected a British proposal to exchange him. Finally, in 1780, Massachusetts exiled him to the West Indies under pain of death should he return. But the small schooner in which he sailed was never heard of again, and the first American to have lost his liberty as a result of cryptanalysis evidently lost his life because of it as well.
Cryptology served another traitor much better. No mere monalphabetic substitution for ambitious Benedict Arnold. He played for much higher stakes and his systems excelled in security. The correspondence between Arnold, in charge at West Point, and John Andr6, an engaging young British major whose gallantry caused some to call him the “English Nathan Hale,” was conducted in several types of code. Arnold apparently handled his own cryptographic duties, but encoding and decoding at the Tory end devolved largely upon Jonathan Odell, a Loyalist clergyman of New York, and upon Joseph Stansbury, a Philadelphia merchant also partial to the Crown.
At first they employed a book code based on volume I of the fifth Oxford edition of the legal classic, Blackstone’s Commentaries. “Three Numbers make a Word,” André instructed Stansbury, “the 1st is the Page the 2d the Line the third the Word.” Words not in the book were to be spelled out, and these codenumbers distinguished from the others by drawing a line through the last number, which then represented the position of a letter in that line instead of a word.
They promptly ran into unsuspected practical difficulties. Only a very few of the encoded words (the messages were encoded only in part) could be found whole, such as general (35.12.8) and men (7.14.3). Arnold managed to find the word militia, but he had to search to page 337 to find it, whereas the other words in his message of June 18, 1779, came from pages 35, 91, and 101. Most of the words and the proper names had to be spelled out in an enormously cumbersome fashion that required tedious counting for each letter and then the writing of four digits as its ciphertext equivalent. Sullivan, for instance, became (with a stroke through the final number of each group) 35.3.1 35.3.2 34.2.4 35.2.5 35.3.5 35.7.7 35.2.3 35.5.2. Arnold consequently abandoned the system after sending one message in the Blackstone code, and receiving only one from Stansbury to Odell.
The conspirators switched to the best-selling Universal Etymological English Dictionary of Nathan Bailey as a codebook; the words, being listed alphabetically, were considerably easier to locate. Then they turned to a small dictionary, which has not been identified. Through its pages sifted the bulk of the clandestine correspondence relating to Arnold’s betrayal of West Point to the British in return for money, security, and honor. Both sides enciphered their codenumbers by adding 7 to each of the three figures—including the middle digit which, representing the column, always appeared as 8 or 9 in what would have been a giveaway to the system. But the security of the system was never put to the test of Colonial cryptanalysis, for the attempted betrayal was blocked by the capture of André before any of the missives were intercepted. He was hanged; Arnold escaped—to a life of ignominy.
British spy cryptography was surpassed by that of the two most important American agents. Samuel Woodhull of Setauket, Long Island, and Robert Townsend of New York City supplied Washington with reams of information about the redcoat occupation of New York during 1779. They wrote their reports in a one-part nomenclator of about 800 elements that had been constructed by one of Washington’s spymasters, Major Benjamin Tallmadge of the Second Connecticut Dragoons. Tallmadge extracted the words he thought would be needed from a copy of John Entick’s New Spelling Dictionary, wrote them in columns on a double sheet of foolscap, and assigned numbers to them. Personal and geographic names followed in a special section. Thus, 28 = appointment, 356 = letter, 660 = vigilant, 703 = waggon, 711 = George Washington, 723 = Townsend, 727 = New York, 728 = Long Island. In addition, the following semimixed alphabet permitted the encipherment of words not in the code list:
Benedict Arnold’s dictionary-code message of July 15, 1780, to Major John André, reading, in part, “If I point out a plan of cooperation by which S[ir Henry Clinton] shall possess him self of West Point, the garrison &c &c &c, twenty thousand pounds sterling I think will be a cheap purchase for an object of so much importance….” Arnold’s code signature, 172.9.192, stands for his codename, MOORE.
Tallmadge provided copies of these pocket codes to both spies and to Washington, and kept one himself. A typical letter from Woodhull, dated at Setauket, August 15, 1779, began: “Sir: Dqpeu Beyocpu agreeable to 28 met 723 not far from 727 and received a 356, but on his return was under the necessity to destroy the same, or be detected….” (DQPEU BEYOCPU was Jonas Hawkins, a messenger.) The spies further masked their identity under codenames, Woodhull being CULPER SR. and Townsend CULPER JR.
The CULPERS used invisible ink extensively. Washington supplied them, getting it from Sir James Jay, who had been a physician in London and was the brother of John Jay, the American statesman who became the first Chief Justice. Sir James recounted the story of the ink in a letter he wrote years later to Thomas Jefferson:
When the affairs of America, previous to the commencement of hostilities, began to wear a serious aspect, and threatened to issue in civil war, it occurred to me that a fluid might possibly be discovered for invisible writing, which would elude the generally known means of detection, and yet could be rendered visible by a suitable counterpart. Sensible of the great advantages, both in a political and military line, which we might derive from such a mode of procuring and transmitting intelligence, I set about the work. After innumerable experiments, I succeeded to my wish. From England I sent to my brother John in Newyork, considerable quantities of these preparations….In the course of the war, General Washington was also furnished with them, and I have letters from him acknowledging their great utility, and requesting further supplies…. By means of this mode of conveying intelligence, I transmitted to America the first authentic account which Congress received, of the determination of the British Ministry to reduce the Colonies to unconditional submission; the ministry at the time concealing this design, and holding out conciliatory measures. My method of communication was th
is: To prevent the suspicion which might arise were I to write to my brother John only, who was a member of Congress, I writ with black ink a short letter to him, and likewise to 1 or 2 other persons of the family, none exceeding 3 or 4 lines in black ink. The residue of the blank paper I filled up, invisibly, with such intelligence and matters as I thought would be useful to the American Cause…. In this invisible writing I sent to [Benjamin] Franklin and [Silas] Deane, by the mail from London to Paris, a plan of the intended Expedition under Burgoyne from Canada.
By July, 1779, Washington was writing CULPER SR.: “All the white Ink I now have (indeed all that there is any prospect of getting soon) is sent in phial No. 1 by Col. Webb. The liquid in No. 2 is the counterpart which renders the other visible by wetting the paper with a fine brush after the first has been used and is dry. You will send these to C-----R, JUNR., as soon as possible, and I beg that no mention may ever be made of your having received such liquids from me or anyone else.” But though Washington urged the use of a cover-text in black ink, the CULPERS customarily wrote their message on a blank sheet of paper, inserting the sheet at a predetermined point in a whole package of the same letter paper.
Numerous letters in this “stain”—as Washington and the CULPERS generally called the secret ink—successfully eluded British inspection and transported considerable information to the American commander in chief. The reports of the CULPERS were filled with detail on such matters as how many troops were stationed where, what warships were anchored in New York harbor, what provisions were entering the town, and so forth. Washington found their reports “intelligent, clear and satisfactory” and said of CULPER JR. that “I rely upon his intelligence.”
