by DAVID KAHN
Then there was the time a mysterious gentleman offered the Austrians a handwritten copy of the Serbian code, copied in snatched moments at the risk of life by his nephew, who worked in the Serbian code room. To prove its validity, he said, he would leave it with the Austrians, who would test it on the next two Serb telegrams. They came through the very next day and were duly decoded with the new key; they dealt with some matters of custom duties—the dull routine of embassy affairs. The gentleman got his 10,000 kronen and the Austrians a warm glow of self-satisfaction, especially when they reminded one another how a copied code was less likely to be changed than a stolen one. Soon more Serb messages trickled in. “We took our cipher out of the safe,” recalled one former staff member, “arranged our dispatches on the desk, and set to work. And we kept on working—perspiring, groaning, cursing—and could not get an intelligible sentence out of them—not a letter, a syllable, a punctuation point. With the exercise of some imagination we made one of them read, The male mother of the warship has been built.
“Then we had a bright idea. We composed a dispatch to the ambassador and put it into cipher with our 10,000-kronen key.” They marked it Urgent and sent it off. Soon an angry little Serbian secretary bounced into the Vienna central telegraph office and demanded that three hopelessly garbled telegrams be repeated. They were, of course, the two that the Austrians had first decoded with their new key and the one that they had made up with it. What had happened was that the code was a pure fake, written out by the mysterious gentleman himself, who had an accomplice send two telegrams in them to the Serbian embassy. The Serbs let them lay, undecoded, until the urgent wire shook them into action.
But the Austrians were not always on the losing end. Indeed, they showed no little ingenuity of their own at one time. Through cryptanalysis, they had identified about 150 words of the Italian code used between Rome and Constantinople. Then they bogged down. So they inserted a tidbit of military information into an Italian-language paper published at Constantinople. As they had hoped, the Italian military attaché picked up the item—verbatim, as it turned out—and sent it encoded to Rome. The Austrians began phrasing their paragraphs so as to enlarge their vocabulary, and within just three months they had a fairly workable Italian code of about 2,000 words.
As the tides of history flowed toward war, nations tightened their alliances and intensified their mobilization efforts. At a high-level meeting in London in 1911 to arrange for Anglo-French intercommunications in the event of war, it was decided, among other matters, to prepare an English-French codebook. Cartier, who was at the meeting, later returned for conferences with Britain’s Lieutenant Spiers on the code’s lexicon and its rules of service; in 1913, he checked the final draft before printing. Soon thereafter, Spiers brought three copies of the codebook to Paris—one for the French G.H.Q., one for the French Army that would have the English on their flank, a third for the French cryptologic section.
The minuteness of these preparations reflected that of War Plan w, which the British and French general staffs had completed by the spring of 1914 and which specified the billet of every battalion in the British Expeditionary Force, down to the places where the troops were to drink their coffee. The Central Powers were no less lax, and finally, in an obscure corner of the Balkans, someone helpfully slew an archduke, and the nations leaped recklessly into the bloody cockpit of war.
* The earliest known printing cipher machine appears to be one invented before 1874 by Émile Vinay and Joseph Gaussin, but no description of it is known.
* Whether Bazeries, who had been lent to the ministry in September, was serving in the cipher bureau at the time, existing records do not say. His discussion of the Panizzardi telegram in his book is so sketchy that it seems unlikely that he participated in its solution.
9
ROOM 40
BEFORE DAWN on the morning of August 5, 1914, the first day of a world war that was to convulse country after country and to end the lives of millions, an equipment-laden ship slid quietly through the black and heaving waters of the North Sea. Off Emden, where the Dutch coast joins the German, she dropped some grappling gear overboard with a dull splash, and shortly there rose dripping from the sea great snakelike monsters, covered with mud and seaweed. Grunts of men, chopping sounds—and soon they were returned, severed and useless, to the depths. These were Germany’s transatlantic cables, her chief communications lifelines to the world, and the vessel was the British cable ship Telconia. Though the Committee of Imperial Defence never dreamed of it when it planned the move in 1912, the cutting of these cables, England’s first offensive action of the war, forged the first link in a chain that helped to end it.
Germany was now forced to communicate with the world beyond the encircling Entente by radio or over cables controlled by her enemies. She thus delivered into the hands of her foes her most secret and confidential plans, provided only that they could remove the jacket of code and cipher in which Germany had encased them. It was an opportunity for which England was unprepared, but of which she promptly availed herself.
On that first day of the war, the director of naval intelligence, Rear Admiral Henry F. Oliver, walked to lunch with the only man at the Admiralty to take any interest in cryptology, the director of naval education, Sir Alfred Ewing. A few months before, Ewing had devised what he later called a “futile” ciphering mechanism, and he had spoken to Oliver about new methods of constructing ciphers. Oliver mentioned that some naval and commercial radio stations were sending to the Admiralty some messages in code that they had picked up and that these were accumulating on his desk. The Admiralty had no department to deal with enemy cryptograms, he said. Ewing was at once interested, and when he saw the messages that afternoon he recognized that they were probably German naval signals and that their solution could be of great value. He at once undertook the task.
Ewing was then 59, a short, thickset Scot with blue eyes beneath shaggy eyebrows, a quiet voice, and the manner of a benign physician. He had been knighted three years before for his contributions to science, which included pioneering studies of Japanese earthquakes, of magnetism, and of mechanical lagging effects in stressed materials (now known by a word he coined, “hysteresis”), and for his public services, notably his naval education directorship. He was to become president of the British Association for the Advancement of Science and perhaps his country’s greatest living expert on mechanical science. And now he was about to found a cryptanalytic bureau that was to become almost legendary and to exert a direct and noticeable effect upon the course of history.
He began by boning up on ciphers in the stacks of the British Museum library and on the construction of codes at Lloyd’s of London and at the General Post Office, where commercial codebooks were on file. He called in four teachers at the naval colleges at Dartmouth and Osborne, A. G. Denniston, W. H. Anstie, E. J. C. Green, and G. L. N. Hope, all friends of his with a good knowledge of German, and, sitting together around the table in his office, they inspected the incomprehensible lines of letters and numbers with only the feeblest general idea on how to begin.
Among those first messages was one which, had they been able to solve it, might have affected the entire course of the war. It may have been among the first batch that Oliver showed Ewing, for it had been issued at 1:35 a.m. August 4 by the German naval high command and transmitted by the powerful radio station at Nauen outside Berlin to Admiral Wilhelm Souchon, commander in the Mediterranean. Message 51 read: “Alliance with Turkey concluded August 3. Proceed at once to Constantinople.” Souchon started eastward from the central Mediterranean at once in the battle cruiser Göben and the light cruiser Breslau. The British Mediterranean squadron, convinced that Souchon would try to force the Strait of Gibraltar, patrolled the waters west of Sicily while Souchon coaled at Messina. When a British cruiser finally spotted him heading east out of the Strait of Messina, the squadron made a frantic effort to catch and destroy him, but he eluded them among the isles of Greece. On Sunday, A
ugust 10, Göben steamed into the Dardanelles, bringing, as Winston Churchill later acknowledged, “more slaughter, more misery and more ruin than has ever before been borne within the compass of a ship.” For Göben’s strength and its bombardments of Russian Black Sea ports brought Turkey into the war and sealed off Russia from her allies, contributing to her eventual capitulation and all that that would mean. Had the Admiralty been able to read Souchon’s orders then as they did retrospectively later in the war, England might have won the fateful game of hide-and-seek, with consequences perhaps greater than any other single exploit of the war.
Of all this, nothing was foreseen at Whitehall, where on that very Sunday Ewing was, he wrote to his family, “in the thick of office work, special work quite outside my ordinary lines.” The codebooks of several German commercial firms were being rounded up, but they proved of no help. Not much better was a mercantile signal-book used by German outpost vessels that had been taken from a German merchantman in Australia at the outbreak of war. Meanwhile, Russell Clarke, a barrister and a radio ham, set up the first low-frequency intercept station at Hunstanton to bring in the raw material, and soon, with the help of Clarke and another ham, Commander B. Hippisley, intercepts were trickling in by direct land line from 14 coast intercept stations to “Ewing Admiralty.”
None of the small band of pioneers had had any real previous knowledge of cryptanalysis, and they made only antlike progress in those first weeks. But Ewing was exhilarated by the job, and it was not until October 25 that he took a Sunday off. By then, England had had a stroke of fortune that gave such an impetus to its cryptanalytic work that it remained far ahead of its enemies through the rest of the war. What happened has best been told in his own style by the minister who then headed the Admiralty, the First Lord, Winston Churchill:
At the beginning of September, 1914, the German light cruiser Magdeburg was wrecked in the Baltic. The body of a drowned German under-officer was picked up by the Russians a few hours later, and clasped in his bosom by arms rigid in death, were the cypher and signal books of the German Navy and the minutely squared maps of the North Sea and Heligoland Bight. On September 6 the Russian Naval Attaché came to see me. He had received a message from Petrograd telling him what had happened, and that the Russian Admiralty with the aid of the cypher and signal books had been able to decode portions at least of the German naval messages. The Russians felt that as the leading naval Power, the British Admiralty ought to have these books and charts. If we would send a vessel to Alexandrov, the Russian officers in charge of the books would bring them to England. We lost no time in sending a ship, and late on an October afternoon Prince Louis [of Battenberg, First Sea Lord] and I received from the hands of our loyal allies these sea-stained priceless documents.
The date was October 13. Russell Clarke, who was as adept with a camera as with a crystal set, copied it by photography at his home. But even the astounding windfall of the Magdeburg codebook—perhaps the luckiest in the whole history of cryptology—did not enable Ewing’s team to read the German naval messages, for the four-letter codewords in that book did not appear in the dispatches. Finally, Fleet Paymaster Charles J. E. Rotter, a principal German expert, discovered that the code had been superenciphered with a monalphabetic substitution. Solution of such a superencipherment is not too difficult a problem with the codebook in one’s possession. As in ordinary plaintext, certain codewords recur more frequently than others and in familiar clusters, letters in one codeword reappear in others in different arrangements, and the codewords themselves possess some structural regularities: in the case of the German naval code, consonants alternated with vowels in the four-letter codewords. When these characteristics are known, the cryptanalyst can spot them almost as well as the more pronounced ones of ordinary language, and can exploit them to solve the superencipherment.
So green were the British cryptanalysts that it took them almost three weeks before they began reading portions of some German naval messages. These, Churchill says, “were mostly of a routine character. ‘One of our torpedo boats will be running out into square 7 at 8 p.m.,’ etc. But a careful collection of these scraps provided a body of information from which the enemy’s arrangements in the Heligoland Bight [bordering the northwest German coast] could be understood with a fair degree of accuracy.”
By this time, Ewing’s staff had grown to such an extent that they crowded his office, and they were continually irked by having to put their work out of sight when he had visitors on educational subjects. So about the middle of November the entire cryptanalytic group moved to Room 40 in the Old Buildings of the Admiralty. This was a large room with a small room adjoining, with a camp bed for tired staffers. Room 40, O.B., had the advantage of being out of the main stream of Admiralty traffic, yet being relatively handy to the Operations Division, which received its output. Though the cryptanalysts were later designated as I.D. 25 (section 25 of the Intelligence Division), “Room 40” was so convenient and so innocuous a name that it soon became the common identification for the organization. The name stuck even when I.D. 25 moved into larger quarters.
For it expanded rapidly. At the end of December an English trawler brought up a heavy chest containing a number of books and documents that had been jettisoned by one of four German destroyers which had been sunk in an action in the Heligoland Bight October 16. In it was found an important German codebook that was missing from the Magdeburg find. The cryptanalysts immediately used it to read signals to German cruisers harassing English shipping, but they did not discover for several months that it also served to encode messages between Berlin and German naval attaches abroad. The increase in traffic required five new men, which brought watches to two-man strength. As additional codes were discovered, more staff was enrolled, frequently in a casual British manner.
Francis Toye, a young war-prison administrator and interpreter who had worked in the British censorship and who later became a widely known music critic, attended a dinner one Thursday evening at the London home of financier Max Bonn. Among the guests was one of Room 40’s brighter members, Frank Tiarks, a partner in the banking firm of J. Henry Schroder & Co. and a director of the Bank of England.
“We talked a good deal,” Toye recalled, “and after dinner he took me aside and asked me whether I should like to come to the Admiralty. Registering proper—and wholly genuine—surprise, I answered that I could not see what use the Admiralty is likely to have for my services.
“‘Max has just told me that your German is very good,’ he replied; ‘you are obviously intelligent and presumably, from your record, trustworthy. There are hundreds of people with one of these qualifications, a number with two, but very few with all three. What about it?’
“‘What about my job, the War Office and all that?’ said I.
“‘If you’ll come you can leave everything in our hands.’
“‘But of course I’ll come if I’m really wanted.’
“‘Very well, I’ll check up on you and you’ll hear in due course.’… About a fortnight later the Commandant sent for me and silently handed me a War Office telegram: ‘Lieutenant-Interpreter Toye is to report as soon as possible to the Admiralty for special duty.’ So omnipotent and expeditious is the British Admiralty when its mind is once made up; think of the yards of red tape that must have been cut in those two weeks!”
Meanwhile, naval intelligence was building up activities concomitant to cryptanalysis. Major radio direction-finding stations were—largely thanks to Oliver’s foresight—set up at Lowestoft, York, Murcar, and Lerwick; they fed their readings into Whitehall, where they proved of immense help in locating the German fleets and the movement of the U-boats. There was no way of avoiding a fix except by maintaining radio silence. This fact was of course known to the Germans, and in view of it England made no attempt to keep its direction-finding activity secret, using it as a smokescreen for its less obvious and more valuable cryptanalytic work. Two other sources of radio intelligence were the identification of
ships’ radio call-signs and the recognition of a radio operator’s “fist,” or characteristic way of sending Morse code. If the Admiralty knew that a call-sign heard under way in the North Sea belonged to the 12-gun battleship Westfalen, it would pursue tactics different from those if the call-sign was assigned to U-20. This radio intelligence, plus cryptanalysis, plus other information streaming into the Admiralty was correlated and interpreted by Admiral Sir Arthur Wilson, a former First Sea Lord and naval elder statesman who was charged by Churchill with advising the top war officials of its substance.
Thus it was that, at about 7 p.m. on December 14, 1914, Wilson came to Churchill to report that intelligence indicated a sortie of German vessels, possibly against British coasts. Less than three hours later the Admiralty ordered units of the British fleet to proceed at once to a “point where they can make sure of intercepting the enemy on his return.” So while the German First Cruiser Squadron hurled high-explosive shells into the seacoast towns of Hartlepool and Scarborough early on the morning of the 15th, four British battle cruisers and six of the most powerful battleships in the world were standing 150 miles to the eastward, cutting off their return. As the Germans headed back toward their base in the Jade River at Wilhelmshaven on December 16 after the bombardment, the weather thickened and heavy squalls reduced visibility. But the Admiralty intelligence had placed the light cruiser Southampton so precisely in the path of the German vessels that, at 10:30 a.m., Commodore W. E. Goodenough saw their shapes driving through the fog. He could not be sure that they were not British ships on station, however, so he flashed his recognition signal at them. They failed to reply; he opened fire, but soon lost contact. Two hours later, the heavy British forces sighted the enemy. But when the commander of the German light cruisers saw the giant forms of the British battleships looming up through the drizzle, he, with great presence of mind, blinked the recognition signal that Goodenough had made to him shortly before. Then he turned away and escaped behind the curtains of mist before the deception could be discovered and the 13.5-inch guns could blow him out of the water.