- Home
- DAVID KAHN
THE CODEBREAKERS Page 9
THE CODEBREAKERS Read online
Page 9
The communication-security precautions paid off. Whether or not the messages slipped by the American radio monitors in Hawaii mattered little. Mere interception would not have helped much. The messages bore no external indication of their intended recipient, and they could not have been read. Rochefort’s attack on Japanese naval codes had achieved some minor successes in late October and November, but he could read only about 10 per cent of the naval traffic, and much of this consisted of weather and other minor systems. The information obtained, Rochefort said, “was not in any sense vital.” Cavite was spottily reading JN25 messages—which revealed nothing about Pearl Harbor—until December 4, when the superencipherment was suddenly changed. As a message that moved on the COPEK channel put it: “Five numeral intercepts subsequent to zero six hundred today indicate change of cipher system including complete change differentials and indicator subtractors X All intercepts received since time indicated checked against all differentials three previous systems X No dupes.” Corregidor was not to get the initial break into the new superencipherment until December 8. And the only other system in which the Yoshikawa messages might have been forwarded—the flag officers’ system—remained unsolved.
A possibility of warning was opened at the source, however, when Yoshikawa’s original messages became available to Rochefort’s unit. Mayfield had picked up another batch of cables in the surreptitious fashion from Street on Friday morning and immediately sent them down to Rochefort’s unit by messenger. Solving them was not part of its duty,* but when a superior officer and colleague asks one to do a favor, it is hard to say no. Rochefort assigned the messages to Chief Radioman Farnsley C. Woodward, 39, who had had some experience with Japanese diplomatic codes at the Shanghai station from 1938 to 1940. He had some help from Lieutenant Commanders Thomas H. Dyer, Rochefort’s senior cryptanalyst, and Wesley A. Wright, Dyer’s assistant. Although the unit was not working on the diplomatic systems, it had information on them in the Navy’s R.I.P.s, or Radio Intelligence Publications, with which all radio intelligence units were supplied. The R.I.P. gave, however, only the PA code list, leaving the onerous reconstruction of the current K2 transposition to the cryptanalyst. The half-dozen or so dispatches, plus some in LA, reached Woodward about 1:30 or 2 p.m. Friday, and he immediately began the first of a series of 12- and 14-hour days to read them. He had no difficulty with the LA messages, which were translated into English by Marine Corps Captain Alva Lasswell, but these yielded “nothing but junk.” The K2, however, eluded him, and he worked on it far into the night.
At about 5 p.m. that day, a trans-Pacific telephone call came through to Mrs. Motokazu Mori, wife of a dentist prominent in Hawaii’s Japanese community. She was the Honolulu correspondent for the militaristic Tokyo newspaper Yomiuri Shimbun. Mrs. Mori had received a wire from her editor the previous day asking her to arrange a telephone interview with a prominent Japanese on conditions in Hawaii. She cabled an acknowledgment but, unable to get anyone, she took the call herself.
Yomiuri: Hello, is this Mori?
Mrs. Mori: Hello. This is Mori.
Yomiuri: I am sorry to have troubled you. Thank you very much.
Mrs. Mori: Not at all.
Yomiuri: … I would like to have your impression on the conditions you are observing at present. Are airplanes flying daily?
Mrs. Mori: Yes, lots of them fly around.
Yomiuri: Are they large planes?
Mrs. Mori: Yes, they are quite big.
Yomiuri: Are they flying from morning till night?
Mrs. Mori: Well, not to that extent, but last week they were quite active in the air.
There ensued Q-and-A about the number of sailors, relations between Japanese and Americans, factory construction, population growth, whether the airplanes carried searchlights, Hawaii weather, newspaper comment, and comparison of impressions made during stopovers in Hawaii by two ambassadors to the United States, Kurusu of Japan and Maxim Litvinoff of Russia. The interview continued:
Yomiuri: Do you know anything about the United States fleet?
Mrs. Mori: No, I don’t know anything about the fleet. Since we try to avoid talking about such matters, we do not know much about the fleet. At any rate, the fleet here seems small. I don’t [know if] all of the fleet has done this, but it seems that the fleet has left here.
Yomiuri: Is that so? What kind of flowers are in bloom in Hawaii at present?
Mrs. Mori: Presently, the flowers in bloom are fewest out of the whole year.
However, the hibiscus and the poinsettia are in bloom now.
The editor seemed a little confused about the hibiscus, but the interview continued with discussions about liquor and the number of first- and second-generation Japanese. Finally the editor thanked Mrs. Mori. She asked him to hold on for a moment, but he had already hung up.
Unknown to both of them, someone had been listening. And that someone thought that the talk about hibiscus and poinsettias sounded mighty suspicious—especially on an expensive transoceanic telephone connection, and especially at a time of extraordinarily tense relations.
In Tokyo it was a little after 1 p.m. on Saturday, December 6. The Japanese reply to Hull’s note of the 26th had recently been sent to the cable room of the Foreign Ministry for transmission to the embassy in Washington. Kazuji Kameyama, the cable chief, broke it into fourteen approximately equal parts to facilitate handling and ordered these enciphered on the 97-shiki O-bun In-ji-ki. He also enciphered a shorter “pilot” message from Togo alerting the embassy that the reply was on the way and instructing it “to put it in nicely drafted form and make every preparation to present it to the Americans just as soon as you receive instructions.” At 8:30 p.m., the pilot message was telegraphed from the cable room to Tokyo’s Central Telegraph Office, from where, 45 minutes later, it was radioed to the United States. Bainbridge Island intercepted it and relayed it to OP-20-G. By five minutes past noon on Saturday, December 6 (Washington time), OP-20-G had delivered the teletype copy to S.I.S., which promptly ran it through the PURPLE machine. By 2 p.m. Bratton had it, translated and typed. An hour later it was in the hands of the Army distributees. S.I.S. had officially closed at 1 p.m. and was not due to reopen until 6, when it was to go on 24-hour status. But this notification of the imminent receipt of the long-awaited reply to Hull’s note of the 26th led to telephoning employees Mary J. Dunning and Ray Cave about 2:30 and asking them to report to work. By 4 both were there.
In Tokyo, Kameyama had released the first 13 parts of the Japanese note to the Central Telegraph Office. Following the instructions of the American bureau, he retained the crucial 14th part, which broke off negotiations. Shortly after 10 p.m., commercial radio began sending the 13 parts to Washington. Most of them took less than ten minutes to transmit, but even though two transmitters were used, it was not until two minutes before 2 a.m. that the tail of the last part had gone. Bainbridge, of course, was listening, and it picked the parts up in this order: 1, 2, 3, 4, 10, 9, 5, 12, 7, 11, 6, 13, 8. One batch arrived by teletype at OP-20-G at eleven minutes before noon, Saturday, December 6, Washington time, and the other at nine minutes of 3 that afternoon. Though it was Saturday, December 6, an even date and hence an Army date of responsibility, the Navy handled the dispatches because it knew that S.I.S. was not expected to work that afternoon, and it considered the intercepts of great importance. Decryptment did not go very smoothly, however. Something seemed to be in error. GY knew the key, but it was producing garbles every few letters. The cryptanalysts tried to correct them.
Meanwhile, a decode into Japanese of the long PA-K2 message that Yoshikawa had sent concerning Kühn’s visual-signal system for Hawaii was placed on the desk of Mrs. Edgers in GZ. “At first glance,” she said, “this seemed to be more interesting than some of the other messages I had in my basket, and so I selected it and asked one of the other men, who were also translators working on other messages, whether or not this shouldn’t be done immediately and was told that I should and then I started to tr
anslate it. Well, it so happened that there was some mistake in the message that had to be corrected and so that took some time. That was at 12:30 or perhaps it was a little before or after 12:30; whatever time it was, we were to go home. It being Saturday, we worked until noon. I hadn’t completed it, so I worked overtime and finished it, and I would say that between 1:30 and 2 was when I finished my rough draft translation.” Mrs. Edgers left it in the hands of Chief Yeoman Bryant. But the message was still not entirely clear, and she had not yet had enough experience for her translations to be sent out without further checking. Kramer, busy with the 13 parts, did not examine it in detail.
To speed processing of the 13 parts, GY, learning that some people were in S.I.S., sent over parts 1 and 2. But when Major Doud of S.I.S. ordered Miss Cave to OP-20-G to help in the smooth typeups, the two parts were returned to GY for solution there, probably because of the garbles. But other messages also coming in were retained by S.I.S.
At 3 o’clock, Kramer, in GZ, had checked with GY to find out whether any more Tokyo traffic had come in before releasing his translators for the day. Since the critical matter of a diplomatic note is often found in the last sentences, GY broke down the last part intercepted for him. The first part of the first line indicated in Japanese that this was part 8 of a 14-part message. After about three lines of Japanese text in the preamble, the message came out in English, just as the Foreign Office had sent it. Kramer could let his translators go home. Interspersed throughout the English text were many of the three-letter codewords indicating punctuation, paragraphing, and numbering, but these posed no problem since they had been recovered long ago.
At 4 o’clock, when Linn took over the GY watch, the garbles still had not been cleared. He decided to start from the very beginning, to check the key, find what was wrong, and redecrypt the messages rather than to try to guess at the garbled letters and possibly make serious errors that would distort the sense. Discarding all the previous work caused a serious jam on the Navy’s one PURPLE machine, and about 6 p.m. GY again called on S.I.S. for help. Parts 9 and 10 were sent over; an hour later, the decrypts came back in longhand. By 7:30, the last of the 13 parts was being decrypted.
Not all the garbles had been scrubbed out. Part 3 had a 75-letter smudge that could not be read at all, Part 10 a 45-letter blur, and Part 11 one of 50 letters. Part 13 went awry in two patches. One deciphered as andnd and the other as chtualylokmmtt; GY thought the first should be and as and the second China, can but.*
In the Japanese embassy, about a mile away, the code clerks had completed deciphering the first seven or eight parts of the message by dinnertime. Then they all repaired to the Mayflower Hotel for a farewell dinner for Hidenari Terasaki, head of Japanese espionage for the western hemisphere, who had been ordered to another post.
While they were enjoying themselves, American code clerks at the Department of State were at work encoding a personal appeal for peace from the President of the United States to the Emperor of Japan. This had been off again, on again since October, Roosevelt apparently wishing to save it for a last resort. Now he decided that the time had come. The message was on its way by 9 o’clock. It traversed the 7,000 miles to Tokyo in an hour. But it took ten hours to get from the Central Telegraph Office to the American embassy.
As the President was addressing a message of peace to the Emperor, the men of the Japanese strike force were listening to a message of war. Shortly before, Admiral Nagumo had topped off the fuel tanks of his combat ships for the final dash. His crews waved farewell to the slow-moving tankers. Now the officers read a stirring message from Yamamoto to all hands: “The moment has arrived. The fate of the empire is at stake. Let every man do his best.” Banzais rent the air. Up the mast of Akagi fluttered the very flag that had flown at Japan’s great naval victory over Russia in 1905. It was a moment of great emotion. Nagumo altered course to due south and bent on 26 knots. Through a mounting sea, the battle force plunged toward its target.
Lovely, peaceful, that target lay “open unto the fields, and to the sky,” oblivious to the onrushing armada of destruction. But many people were seeking clues to Japanese intentions, particularly concerning sabotage, which was regarded as a serious threat. Among these was Robert L. Shivers, special agent in charge of the F.B.I.’s Honolulu office and the man who, under authority of the Attorney General, had ordered the tap on the overseas phone that picked up Mrs. Mori’s interview. By noon he had received a transcript in English of the call, and soon after 4 p.m. was conferring about it with May-field and the Army assistant G-2, Lieutenant Colonel George W. Bicknell, in his office on the sixth floor of the Alexander Young Hotel in Honolulu. Mayfield consulted with Lieutenant Carr, who had translated the Navy telephone taps and who happened to be duty officer that afternoon at the District Intelligence Office; both thought that Carr should listen to the original recording to see if any hidden meaning was concealed in the intonations. Shivers said he would have it by 10 the following morning. Bicknell, whose job included heading the Army’s counterintelligence in Hawaii, was convinced that the hibiscus and poinsettias smelled of espionage. He telephoned his boss, Colonel Kendall J. Fielder, the G-2, and said he wanted to see him and General Short immediately on a matter of importance.
They were both on their way to dinner at Schofield Barracks, and Fielder asked if it couldn’t wait until tomorrow. Bicknell said it was too important; Fielder agreed to see him. Bicknell drove hurriedly out to Fort Shafter, where Fielder and Short had their homes side by side, and at about 6 p.m. the three men discussed the message for a while, but though they considered it “very suspicious, very fishy,” Fielder said, “we couldn’t solve it, we couldn’t make heads nor tails out of it.” The flower references seemed totally out of place, as if they were indeed conveying secret military information by open code, but, on the other hand, the Japanese spoke quite openly about airplanes and the fleet. The whole thing was very baffling, and they never did reach a conclusion about it.
They did not know that the Yomiuri Shimbun was then being hawked on the streets of Tokyo with an atmosphere feature on Hawaii based on the Mori interview—complete with reference to flowers. Nor, apparently, did they realize that the Japanese did not need so weak and dangerous a system. They could send much more detailed reports by cable in their diplomatic code. And, in one of the most ironical of situations at Pearl Harbor, they were doing precisely that at that very minute. While the three American army officers were standing on Short’s porch worrying about the hibiscus, the R.C.A. office was time-stamping “1941 Dec 6 PM 6 01” on a message from the consulate. It was signed “Kita” but it came from Yoshikawa. It was brief (only 44 groups) and cheap ($6.82), but it reported that “(1) On the evening of the 5th, the battleship Wyoming and one sweeper entered port. Ships at anchor on the 6th were: 9 battleships, 3 minesweepers, 3 light cruisers, 17 destroyers. Ships in dock were: 4 light cruisers, 2 destroyers. Heavy cruisers and carriers have all left. (2) It appears that no air reconnaissance is being conducted by the fleet air arm.” Yoshikawa was, as usual, partly right and partly wrong. He mistook Utah for Wyoming. His figure on the battleships was correct, but in harbor that afternoon were 6 light and 2 heavy cruisers, 29 destroyers, 4 minesweepers, 8 minelayers, and 3 seaplane tenders. With this message Yoshikawa completed his assignment. It was the last cable sent by the Japanese consulate in Hawaii for many years.
By 8:45 p.m. in Washington, the 13 parts had been typed in smooth copies and put up in folders. Kramer began telephoning the recipients to find out where they were so he could bring the MAGIC to them. He also called his wife, Mary, who agreed to chauffeur him during his deliveries. They reached the White House first, at about 9:15. The naval aide, Beardall, had told the President that some MAGIC would be delivered that evening, and at about 4 p.m. he had ordered his communications assistant, Lieutenant Lester R. Schulz, to stand by and bring it to the President. Schulz was waiting in Beardall’s small office in the corner of the basement mail room in the White House when Krame
r arrived. The Roosevelts had been entertaining at a large dinner party, but the President had excused himself. Schulz obtained permission to bring the MAGIC to the President, and an usher accompanied him to the oval study on the second floor and announced him. Roosevelt was seated at his desk. Only Harry Hopkins was with him. Schulz unlocked the briefcase with the key that Beardall had given him, removed the sheaf of MAGIC, and handed it to the President. He read the 13 parts in about ten minutes while Hopkins paced slowly up and down. Then Hopkins read them. The 13th part rejected Hull’s offer, and when Hopkins had passed the papers back to the President, Roosevelt turned to him and said, in effect, “This means war.” Hopkins agreed, and for about five minutes they discussed the situation, the deployment of Japanese forces, the movement towards Indochina, and similar matters. The President mentioned his message to Hirohito. Hopkins remarked that it was too bad that the United States could not strike the first blow and prevent any kind of surprise in the inevitable war.
On the eve of Pearl Harbor, Takeo Yoshikawa sends his final message over Consul Kita’s signature, using the PA-K2 code, to report that the U.S. fleet is still in port