by DAVID KAHN
The French sent many of their naval solutions to London, but Room 40 reciprocated as minimally as possible. Hall apparently never sent the Magdeburg nor any other in-force naval codebooks to the French. His motives were understandable. England depended for her very existence on control of the seas, and every additional person who knew of the German solutions added to the danger of loss of this supremely valuable intelligence and, consequently, of the nation’s maritime mastery. But, in the opinion of Colonel François Cartier, head of the French cryptologic service, Hall exceeded all decent bounds in his jealous hoarding of his cryptanalytic secrets.
Once when Cartier was visiting Hall, he told the director of naval intelligence that his bureau was cryptanalyzing the German naval codes but had only progressed to partial solutions. Hall suggested that Cartier leave the naval traffic to the British, who had an actual copy of the German code, could read the German messages with ease, and would apprise the French of anything of importance to them. Cartier replied by telling Hall how one of the fragmentary French decryptments had enabled them to save one of their auxiliary cruisers from possible torpedoing; the English must have known of the danger from the same intercept, but they had not warned the French. The intelligence chief explained that it was better to lose the ship than to take precautionary measures that risked disclosing the cryptanalysis to the Germans. “Would you feel the same way if the cruiser had been English?” Cartier asked coldly. Hall dodged that one, and a change of code ended the negotiations.
Mutual need sometimes overrode these differences, however, and cryptanalytic collaboration continued among the Allies. England, for example, read the Berlin-Madrid diplomatic messages in both the Spanish and the German codes and offered them to the French; France managed to solve a superencipherment in this traffic the very day that it was put into service, and then sent the solution to Hall. It was in this code that the German naval attaché in Madrid radioed to Germany several times to ask for funds and instructions for agent H-21. The agent, a beautiful dancer better known by her stage name of Mata Hari, was ordered to Paris. But the French had read the messages, which were the first concrete evidence that they had been able to collect that Mata Hari was a German spy. They picked her up, and though she fiercely contended that the money was a payment from her lovers, the messages convicted her. A few months later, courageously refusing a blindfold, she was executed by a twelve-man firing squad.
The French also solved an Austro-Hungarian code, which they later got from Hall, and a naval code of the same country, which was superenciphered in a way that gave rise to such peculiar codewords as PLESDEPOTS, CODY-FIGARO, and OGNISEXUAL. The French discovered in May of 1916 that the first four digits of the ten underlying codenumbers were enciphered in two groups of two and the last six digits in two groups of three. The solution proved of great value to the Italians. The Austrians apparently later changed this code, for in the autumn of 1917, Hall learned that the Italians were having little success in obtaining information about the movements of the Austro-Hungarian fleet, despite its extensive use of wireless. He dispatched three of Room 40’s staff as a “special secret information service” to study the Austro-Hungarian signals, and the British ambassador to Italy later wrote the foreign secretary that this service “has been of great value to us to obtain rapid and sure information of what was going on on the other side of the Adriatic, and I do not think either we or the Italians would have had much if it had not been for the system which he [Hall] devised and induced the Italians to work.”
Just as the reading of secret naval and diplomatic messages was not restricted among the Allies to Room 40, so it was not restricted among the belligerents to the Allies. The Germans had finally set up a cryptanalytic section, with an intercept and transmission post at Neumünster. They succeeded in penetrating the British naval codes (whether by capture or by cryptanalysis is unknown), and during Jutland they read Jellicoe’s order massing his destroyers to his rear to shield from a torpedo attack. Neumünster passed this order to Scheer. This, together with other information, confirmed his position well astern of the British battle fleet. He therefore thought it was safe enough for him to cross his enemy’s wake—and he did, running safely for home without encountering the superior British dreadnoughts.
Room 40 intercepted and read this message. Whether it specifically motivated them to change or improve their naval systems is unknown, but it is certain that they ended the war with unquestionably the era’s finest code. This is Cypher SA, apparently invented by one J. C. F. Davidson, who received £300 for it. It went into force at noon August 1, 1918, replacing Cypher W.
Despite its name, it was a two-part code, bound in two volumes in the standard lead covers so it would sink when jettisoned in time of danger. The encoding section ran 341 pages and gave five-digit codenumbers for everything from A to Zwyndrecht, with up to 15 homophones for many plaintext expressions. Ship, for example, had that many, but the effective number was even larger because of the 35 phrases containing the word ship, as ship will be, which itself had three homophones, plus the separate entries for ships, shipping, shipped, and so on. The code included two pages of nulls, tables of digraphs and single letters for spelling words not in the vocabulary, separate sections for numbers, dates, message references, senior officers’ names, British navy warships, and names of foreign men-of-war, as well as indicators to shift to a separate “code index” with names of important merchantmen and steamship companies. The 536-page decoding section ran from 00100 (for Vathy) to 53698 (for Nought one four five), but many numbers were skipped in the codegroup series; at one point, for example, it ran 07401, 07403, 07404, 07406. The instructions called for the use of at least 25 per cent nulls in every message—which had to start with one.
The code’s major feature, however, was its extensive use of the polyphone, a codegroup that has multiple meanings. Obviously, if codegroup 07640 can mean either eight, or fifth April, or then North-ward, the task of the cryptanalyst becomes substantially more difficult. This situation prevailed in Cypher SA with a large percentage of the groups. How, then, could the legitimate decoder keep the meanings separate, so that he would not inadvertently select eight when fifth April is meant? The code distinguished between the three meanings by tagging the three polyphones with an A, a B, or a C both fore and aft of the codenumber. In encoding, the code clerk had to pick the codegroup that had the same letter in front of it as the codegroup preceding had behind it. In other words, a group ending with a B must be followed by one beginning with a B. The code was so constructed that wherever the clerk had to make this selection, a choice of codegroups was provided. All the polyphones, in other words, were homophones (but not vice versa). The code clerk dropped the letters before transmitting the cryptogram. The decoder could pick up the thread with the first group, a null, because all the nulls and many plaintext groups were prefixed with a dash. This meant that they did not have to follow any particular letter, and so could serve as the free end of a chain. The letter at their tail, however, forged the first link in this chain, which the decoder tracked through his codebook.
Polyphones are a powerful weapon for confusing a cryptanalyst, for a codegroup may not always be what it seems. This is not to say that Cypher SA was unbreakable; but it undoubtedly would have demanded considerably more time, more traffic, and more corollary information than others. The connoisseur may also revel in its exquisite ingenuity.
Hall’s supersession of Ewing roughly coincided with the end of the great sea battles. This was largely due to Room 40. “Without the cryptographers’ department there would have been no Battle of Jutland,” Churchill wrote, and Jutland bottled up the German fleet so effectively that it never ventured forth again. The closing of this phase of the war reduced the need for tactical intelligence, and Hall, as aggressive as any man, shifted the emphasis to the strategic. He gained access to the larger and more exciting arena of international affairs through Room 40’s diplomatic decryptments. This was usurpation of power, for his
province was nominally just naval intelligence, and indeed, while the Foreign Office appreciated his information, it grudged him his powers. But it was helpless to stop him, for his control of Room 40 made him absolute master of the vital information it produced, from disclosures of the far-flung subversions and conspiracies and aggrandizements of the Central Powers to the coded squeakings of a minor spy. Though Hall did pass this information to other governmental departments (usually in a form that concealed the source), he also stuck his fingers into more than one political pie. Fortunately for Britain, he nearly always came out with plums.
A page of the encoding section of Admiralty Cypher SA, showing homophones
He was doing this even before Ewing left. There was, for example, the German plan for a revolt in Persia, bared by Room 40’s cryptanalysis of the plotters’ messages. In another case, Trebitsch Lincoln, an embittered former member of Parliament, sent military information to the German consul in neutral Rotterdam in one dictionary code and two jargon codes that were solved by Room 40. In one of the jargon systems, family names meant ships or ports; in the other, various petroleum products stood for them. A message that read CABLE PRICES FIVE CONSIGNMENTS VASELINE, EIGHT PARAFFIN really meant [At] Dover [are] five first-class cruisers, eight sea-going destroyers. Lincoln, unfortunately, evaded the British authorities and escaped to New York.
Room 40 also read coded messages involving Sir Roger Casement, the former British consul who, after failing to recruit an anti-English battalion among Irish prisoners-of-war in Germany, sought to raise rebellion in Ireland. Several of these cryptograms were passed between Berlin and German diplomatic posts in the United States. One urged German military support of the rebellion by “troops, arms, and ammunition”; another dealt with the transmission of $500 to Casement by John Devoy, an Irish agitator in America who had arranged for Germany’s delivery of 20,000 rifles and ten machine guns for the uprising. Another message read by Room 40 reported to Devoy that Casement’s sailing on a submarine was imminent and arranged that the codeword OATS would be cabled if the U-boat left with Casement aboard as scheduled and the codeword HAY if there was a hitch. On April 12, 1916, among the day’s usual batch of intercepts appeared one containing the word OATS. Ten days later, Casement landed near Tralee Bay—and was promptly arrested by waiting authorities. He remained cool, giving a false name and saying he was a writer, but on the way to Adfert Barracks he tried to discard a piece of paper on which was written a small code of phrases he might need, such as send more explosives. The police saw it and confiscated it for evidence. He was tried and convicted of high treason. Hall deflated the strong public pressure for a reprieve by surreptitiously circulating through London clubs and the House of Commons specimen pages of Casement’s homosexually-inclined “Black Diaries.” Casement was hanged on August 3.
A page of the decoding section of Cypher SA, showing poly phones
Not all Hall’s activities were so nefarious. Spy scares were rampant, so much so that when a bird flew up from near where a foreign-looking individual stood, a hysterical bystander called police, convinced that the “alien” was sending messages to the enemy by homing pigeon. One day a self-described “code expert” from the London financial district came to tell Hall that he had solved secret messages relating to the movement of troops that had been concealed as personal advertisements in newspapers. The head of naval intelligence listened attentively and invited him to return when he had further proofs. Then Hall, who was not without a sense of humor, composed a suspicious-sounding message and inserted it in the personal column of The Times. Next day the expert arrived, highly agitated, with a “solution” that disclosed that certain battleships were about to sail from the naval ports of Chatham, Portsmouth, and Plymouth. His reaction when Hall told him what had happened is, regrettably, not recorded.
At about half-past ten on the morning of January 17, 1917, the Reverend William Montgomery, a thin, gray-haired scholar of the early church fathers who was serving as a cryptanalyst in the diplomatic section of Room 40, came to tell Hall of what looked like an important message. Montgomery’s instincts were right. The cryptogram that he and a youthful colleague, Nigel de Grey, had partially read was to become the single most far-reaching and most important solution in history.
The message was a long one, consisting of about a thousand numerical codegroups. Dated at Berlin January 16, it was addressed to the German ambassador in the United States, Count Johann Heinrich Andreas von Bernstorff, and the two cryptanalysts recognized that it was encoded in a German diplomatic code known as 0075, upon which they had been working for six months. Room 40 knew from its analyses that 0075 was one of a series of two-part codes that the German Foreign Office designated by two zeros and two digits, the two digits always showing an arithmetical difference of 2. Among the others, some of which Room 40 had solved, were 0097, 0086, which was used for German missions in South America, 0064, used between Berlin and Madrid and perhaps elsewhere, 0053, and 0042. Code 0075 was a new code that the German Foreign Office had first distributed in July of 1916 to German missions in Vienna, Sofia, Constantinople, Bucharest, Copenhagen, Stockholm, Bern, Lugano, The Hague, and Oslo. Somehow the British obtained copies of enough of the telegrams in this code to enable Montgomery and de Grey, whose assignment it probably was, to make a start in breaking it. In November, Room 40 began intercepting messages to the German embassy in the United States in the same code, and if Hall guessed that the code and the keys to the superencipherment that it sometimes used had been sent across the Atlantic on the second voyage of the cargo U-boat Deutschland, which docked at New London on November 1, 1916, he would have been right.
Montgomery and de Grey could read only parts of the long message. But they could see that it was a double-decker, consisting of Berlin’s messages Nos. 157 and 158 to Bernstorff. They could read the signature of the German Foreign Minister, Arthur Zimmermann. As far as they could extricate its sense on the basis of their partial solution of 0075, the second message read:
Most secret for your Excellency’s personal information and to be handed on to the Imperial Minister in (? Mexico) with Telegram No. 1 (…) by a safe route.
We propose to begin on the 1st February unrestricted submarine warfare. In doing so, however, we shall endeavor to keep America neutral. (?) If we should not (succeed in doing so) we propose to (? Mexico) an alliance upon the following basis:
[joint] conduct of the war.
[joint] conclusion of peace.
(…)
Your Excellency should for the present inform the President [of Mexico] secretly (? that we expect) war with the U.S.A. (possibly) (…) (Japan) and at the same time to negotiate between us and Japan. (Please tell the President) that (…) or submarines (…) will compel England to peace in a few months. Acknowledge receipt.
Zimmermann.
Montgomery handed this fragmentary solution to Hall, who stared down at the phrases that seemed to jump off the page at him: “unrestricted submarine warfare,” “war with the U.S.A.,” “propose … an alliance.” He realized at once that here was a weapon of enormous potentiality. He urged Montgomery to hurry the solution, ordered all copies except the original message and a single solution burned, and, without a word to the Foreign Office, sat down by himself to contemplate the situation.
It was as bleak as that winter’s day. The war that everyone had expected would last only a few weeks had now dragged into its third year. Nor was there any prospect of an end. France had expended half a million lives at Verdun and only succeeded in restoring the battle line to where it was ten months before. England, which had lost 60,000 men at the Somme in a single day, struggled to gain a few yards of shell-blasted earth, then fell back exhausted. The Hindenburg line remained unbreached. Rumania, a new ally, had been quickly overrun, and Russia, the colossus of the east, was virtually defeated. The stepped-up U-boat campaign increased the economic pressure on the Allies. Worst of all, despite the provocation of the Lusitania sinking and despite the tug of
ancient common ties, the United States, guided by a President who had just won reelection on the slogan “He kept us out of war,” remained obstinately neutral.
Things were no better in Germany. Her initial offensive had stalled at the Marne and her gray-coated troops had been locked in the futile trench slaughter ever since. Civilians were living on potatoes—a result of the stranglehold of the British blockade. Fifteen-year-olds were being conscripted. Greece and Portugal had recently entered the war against her. Like the Allies, she could see no immediate hope for victory.
Except one.
Unleash the submarines, the generals cried, and England would soon be “gasping in the reeds like a fish.” The blockaders would become the blockaded. For months the generals had hammered away on this theme, and, as the signs of exhaustion multiplied, they finally prevailed. Foreign Minister Zimmermann, who had long opposed the idea, fell in line. But this big jolly bachelor, the first to break the Junker barrier in the higher regions of the Kaiser’s officialdom, perceived that the repeated sinkings of American vessels would sooner or later torpedo American neutrality, and he bethought himself of a scheme to counter this danger. He proposed a military alliance with Mexico, then particularly hostile to the imperialistic Norteamericanos as a result of Pershing’s punitive expedition into Mexican territory. He sweetened the proposition with an offer of money and the possibility of support from Japan, standing at America’s back, and with still more anti-Yankee inducements.
Unable to deal through the Mexican ambassador, who was in Switzerland, Zimmermann sent his proposal to his minister in Mexico, Heinrich J. F. von Eckardt, by way of Washington. To ensure that it would get there, he routed it two ways, both monitored by Britain. The cruise of Telconia was paying off.
