Chapter 1

ONE DAY OF MAGIC

At 1:28 on the morning of December 7, 1941, the big ear of the Navy's radio station on Bainbridge Island near Seattle trembled to vibrations in the ether. A message was coming through on the Tokyo-Washington circuit. It was addressed to the Japanese embassy, and Bainbridge reached up and snared it as it flashed overhead. The message was short, and its radiotelegraph transmission took only nine minutes. Bainbridge had it all by 1:37.

The station's personnel punched the intercepted message on a teletype tape, dialed a number on the teletypewriter exchange, and, when the connection had been made, fed the tape into a mechanical transmitter that gobbled it up at 60 words per minute.

The intercept reappeared on a page-printer in Room 1649 of the Navy Department building on Constitution Avenue in Washington, D.C. What went on in this room, tucked for security's sake at the end of the first deck's sixth wing, was one of the most closely guarded secrets of the American government. For it was in here -- and in a similar War Department room in the Munitions Building next door -- that the United States peered into the most confidential thoughts and plans of its possible enemies by shredding the coded wrappings of their dispatches.

Room 1649 housed OP-20-GY, the cryptanalytic section of the Navy's cryptologic organization, OP-20-G. The page-printer stood beside the desk of the GY watch officer. It rapped out the intercept in an original and a carbon copy on yellow and pink teletype paper just like news on a city room wire-service ticker. The watch officer, Lieutenant (j.g.) Francis M. Brotherhood, U.S.N.R., a curly-haired, brown-eyed six-footer, saw immediately from indicators that the message bore for the guidance of Japanese code clerks that it was in the top Japanese cryptographic system.

This was an extremely complicated machine cipher which American cryptanalysts called PURPLE. Led by William F. Friedman, Chief Cryptanalyst of the Army Signal Corps, a team of codebreakers had solved Japan's enciphered dispatches, deduced the nature of the mechanism that would effect those letter transformations, and painstakingly built up an apparatus that cryptographically duplicated the Japanese machine. The Signal Corps had then constructed several additional PURPLE machines, using a hodgepodge of manufactured parts, and had given one to the Navy. Its three components rested now on a table in Room 1649: an electric typewriter for input; the cryptographic assembly proper, consisting of a plugboard, four electric coding rings, and associated wires and switches, set on a wooden frame; and a printing unit for output. To this precious contraption, worth quite literally more than its weight in gold, Brotherhood carried the intercept.

He flicked the switches to the key of December 7. This was a rearrangement, according to a pattern ascertained months ago, of the key of December 1, which OP-20-GY had recovered. Brotherhood typed out the coded message. Electric impulses raced through the maze of wires, reversing the intricate enciphering process. In a few minutes, he had the plaintext before him.

It was in Japanese. Brotherhood had taken some of the orientation courses in that difficult language that the Navy gave to assist its cryptanalysts. He was in no sense a translator, however, and none was on duty next door in OP-20-GZ, the translating section. He put a red priority sticker on the decode and hand-carried it to the Signal Intelligence Service, the Army counterpart of OP-20-G, where he knew that a translator was on overnight duty. Leaving it there, he returned to OP-20-G. By now it was after 5 a.m. in Washington -- the message having lost three hours as it passed through three time zones in crossing the continent.

The S.I.S. translator rendered the Japanese as: "Will the Ambassador please submit to the United States Government (if possible to the Secretary of State) our reply to the United States at 1:00 p.m. on the 7th, your time." The -- "reply" referred to had been transmitted by Tokyo in 14 parts over the past 18' hours, and Brotherhood had only recently decrypted the 14th part on the PURPLE machine. It had come out in the English in which Tokyo had framed it, and its ominous final sentence read: "The Japanese Government regrets to have to notify hereby the American Government that in view of the attitude of the American Government it cannot but consider that it is impossible to reach an agreement through further negotiations." Brotherhood had set it by for distribution early in the morning.

The translation of the message directing delivery at one o'clock had not yet come back from S.I.S. when Brotherhood was relieved at 7 a.m., and he told his relief, Lieutenant (j.g.) Alfred V. Pering, about it. Half an hour later, Lieutenant Commander Alwin D. Kramer, the Japanese-language expert who headed GZ and delivered the intercepts, arrived. He saw at once that the all-important conclusion of the long Japanese diplomatic note had come in since he had distributed the 13 previous parts the night before. He prepared a smooth copy from the rough decode and had his clerical assistant, Chief Yeoman H. L. Bryant, type up the usual 14 copies. Twelve of these were distributed by Kramer and his opposite number in S.I.S. to the President, the secretaries of State, War, and Navy, and a handful of top-ranking Army and Navy officers. The two others were file copies. This decode was part of a whole series of Japanese intercepts, which had long ago been given a collective codename, partly for security, partly for ease of reference, by a previous director of naval intelligence, Rear Admiral Walter S. Anderson. Inspired, no doubt, by the mysterious daily production of the information and by the aura of sorcery and the occult that has always enveloped cryptology, he called it MAGIC.

When Bryant had finished, Kramer sent S.I.S. its seven copies, and at 8 o'clock took a copy to his superior, Captain Arthur H. McCollum, head of the Far Eastern Section of the Office of Naval Intelligence.

He then busied himself in his office, working on intercepted traffic, until 9:30, when he left to deliver the 14th part of Tokyo's reply to Admiral Harold F. Stark, the Chief of Naval Operations, to the White House, and to Frank Knox, the Secretary of the Navy. Knox was meeting at 10 a.m. that Sunday morning in the State Department with Secretary of War Henry L. Stimson and Secretary of State Cordell Hull to discuss the critical nature of the American negotiations with Japan, which, they knew from the previous 13 parts, had virtually reached an impasse. Kramer returned to his office about 10:20, where the translation of the message referring to the one o'clock delivery had arrived from S.I.S. while he was on his rounds.

Its import crashed in upon him at once. It called for the rupture of Japan's negotiations with the United States by a certain deadline. The hour set for the Japanese ambassadors to deliver the notification -- 1 p.m. on a Sunday -- was highly unusual. And, as Kramer had quickly ascertained by drawing a navigator's time circle, 1 p.m. in Washington meant 7:30 a.m. in Hawaii and a couple of hours before dawn in the tense Far East around Malaya, which Japan had been threatening with ships and troops.

Kramer immediately directed Bryant to insert the one o'clock message into the reddish-brown looseleaf cardboard folders in which the MAGIC intercepts were bound. He included several other intercepts, adding one at the last minute, then slipped the folders into the leather briefcases, zipped these shut, and snapped their padlocks. Within ten minutes he was on his way.

He went first to Admiral Stark's office, where a conference was in session, and indicated to McCollum, who took the intercept from him, the nature of the message and the significance of its timing. McCollum grasped it at once and disappeared into Stark's office. Kramer wheeled and hurried down the passageway. He emerged from the Navy Department building and turned right on Constitution Avenue, heading for the meeting in the State Department eight blocks away. The urgency of the situation washed over him again, and he began to move on the double.

This moment, with Kramer running through the empty streets of Washington bearing his crucial intercept, an hour before sleepy code clerks at the Japanese embassy had even deciphered it and an hour before the Japanese planes roared off the carrier flight decks on their treacherous mission, is perhaps the finest hour in the history of cryptology. Kramer ran while an unconcerned nation slept late, ignored aggression in the hope that it would go away, begged the hollow gods of isolationism for peace, and refused to entertain -- except humorously -- the possibility that the little yellow men of Japan would dare attack the mighty United States. The American cryptanalytic organization swept through this miasma of apathy to reach a peak of alertness and accomplishment unmatched on that day of infamy by any other agency in the United States. That is its great achievement, and its glory. Kramer's sprint symbolizes it.

Why, then, did it not prevent Pearl Harbor? Because Japan never sent any message saying anything like "We will attack Pearl Harbor." It was therefore impossible for the cryptanalysts to solve one. Messages had been intercepted and read in plenty dealing with Japanese interest in warship movements into and out of Pearl Harbor, but these were evaluated by responsible intelligence officers as on a par with the many messages dealing with American warships in other ports and the Panama Canal. The causes of the Pearl Harbor disaster are many and complex, but no one has ever laid any of whatever blame there may be at the doors of OP-20-G or S.I.S. On the contrary, the Congressional committee that investigated the attack praised them for fulfilling their duty in a manner that "merits the highest commendation."

As the climax of war rushed near, the two agencies -- together the most efficient and successful codebreaking organization that had ever existed -- scaled heights of accomplishment greater than any they had ever achieved. The Congressional committee, seeking the responsibility for the disaster, exposed their activity on almost a minute-by-minute basis. For the first time in history, it photographed in fine-grained detail the operation of a modern codebreaking organization at a moment of crisis. This is that film. It depicts OP-20-G and S.I.S. in the 24 hours preceding the Pearl Harbor attack, with the events of the past as prologue. It is the story of one day of MAGIC.

The two American cryptanalytic agencies had not sprung full-blown into being like Athena from the brow of Zeus. The Navy had been solving at least the simpler Japanese diplomatic and naval codes in Rooms 1649 and 2646 on the "deck" above since the 1920s. Among the personnel assigned to cryptan alytical duties were some of the Navy's approximately 50 language officers who had served in Japan for three years studying that exceedingly difficult tongue. One of them was Lieutenant Ellis M. Zacharias, later to become famous as an expert in psychological warfare against Japan. After seven months of training in Washington in 1926, he took charge of the naval listening station on the fourth floor of the American consulate in Shanghai, where he intercepted and cryptanalyzed Japanese naval traffic. This post remained in operation until it was evacuated to Corregidor in December, 1940. Long before then, radio intelligence units had been set up in Hawaii and in the Philippines, with headquarters in Washington exercising general supervision.

The Army's cryptanalytical work during the 1920s was centered in the so-called American Black Chamber under Herbert O. Yardley, who had organized it as a cryptologic section of military intelligence in World War I. It was maintained in secrecy in New York jointly by the War and State departments, and perhaps its greatest achievement was its 1920 solution of Japanese diplomatic codes. At the same time, the Army's cryptologic research and code-compiling functions were handled by William Friedman, then as later a civilian employee of the Signal Corps. In 1929, Henry L. Stimson, then Secretary of State, withdrew State Department support from the Black Chamber on ethical grounds, dissolving it. The Army decided to consolidate and enlarge its codemaking and codebreaking activities. Accordingly, it created the Signal Intelligence Service, with Friedman as chief, and, in 1930, hired three junior cryptanalysts and two clerks.

The following year, a Japanese general suddenly occupied Manchuria and set up a puppet Manchu emperor, and the government of the island empire of Nippon fell into the hands of the militarists. Their avarice for power, their desire to enrich their have-not nation, their hatred for white Occidental civilization, started them on a decade-long march of conquest. They withdrew from the League of Nations. They began beefing up the Army. They denounced the naval disarmament treaties and began an almost frantic shipbuilding race. Nor did they neglect, as part of their war-making capital, their cryptographic assets. In 1934, their Navy purchased a commercial German cipher machine called the Enigma; that same year, the Foreign Office adopted it, and it evolved into the most secret Japanese system of cryptography. A variety of other cryptosystems supplemented it. The War, Navy, and Foreign ministries shared the superenciphered numerical HATO code for intercommunication. Each ministry also had its own hierarchy of codes. The Foreign Office, for example, employed four main systems, each for a specific level of security, as well as some additional miscellaneous ones.

Meanwhile, the modern-style shoguns speared into defenseless China, sank the American gunboat Panay, raped Nanking, molested American hospitals and missions in China, and raged at American embargoes on oil and steel scrap. It became increasingly evident that Nippon's march of aggression would eventually collide with American rectitude. The mounting curve of tension was matched by the rising output of the American cryptanalytic agencies. A trickle of MAGIC in 1936 had become a stream in 1940. Credit for this belongs largely to Major General Joseph O. Mauborgne, who became Chief Signal Officer in October, 1937.

Mauborgne had long been interested in cryptology. In 1914, as a young first lieutenant, he achieved the first recorded solution of a cipher known as the Playfair, then used by the British as their field cipher. He described his technique in a 19-page pamphlet that was the first publication on cryptology issued by the United States government. In World War I, he put together several cryptographic elements to create the only theoretically unbreakable cipher, and promoted the first automatic cipher machine, with which the unbreakable cipher was associated. He was among the first to send and receive radio messages in airplanes. As Chief Signal Officer, he retained enough of his flair for cryptanalysis to solve a short and difficult challenge cipher. He was also talented in other directions: he played the violin well and was an accomplished artist, exhibiting at, among others, the Chicago Art Institute.

When he became head of the Signal Corps, he immediately set about augmenting the important cryptanalytic activities. He established the S.I.S. as an independent division reporting directly to him, enlarged its functions, set up branches, started correspondence courses, added intercept facilities, increased its budget, and put on more men. In 1939, when war broke out in Europe, S.I.S. was the first agency in the War Department to receive more funds, personnel, and space. Perhaps most important of all, Mauborgne's intense interest inspired his men to outstanding accomplishments. More and more codes were broken, and as the international situation stimulated an increasing flow of intercepts, the MAGIC intelligence approached flood stage.

Mauborgne retired in September, 1941, leaving an expanded organization running with smooth efficiency. By then the Japanese had completed the basic outline for a dawn attack on Pearl Harbor. The plan had been conceived in the fertile brain of Admiral Isoroku Yamamoto, Commander-in-Chief Combined Fleet, Imperial Japanese Navy. Early in the year, he had ordered a study of the operation, contending that "If we have war with the United States, we will have no hope of winning unless the United States fleet in Hawaiian waters can be destroyed." By May, 1941, studies had shown the feasibility of a surprise air attack, statistics had been gathered, and operational planning was under way.

In the middle of that month, the U.S. Navy took an important step in the radio intelligence field. It detached a 43-year-old lieutenant commander from his intelligence berth aboard U.S.S. Indianapolis and assigned him to reorganize and strengthen the radio intelligence unit at Pearl Harbor. The officer was Joseph John Rochefort, the only man in the Navy with expertise in three closely related and urgently needed fields: cryptanalysis, radio, and the Japanese language. Rochefort, who had begun his career as an enlisted man, had headed the Navy's cryptographic section from 1925 to 1927. Two years later, a married man with a child, he was sent, because of his outstanding abilities, as a language student to Japan, a hard post to which ordinarily only bachelor officers were sent. This three-year tour was followed by half a year in naval intelligence; most of the next eight years were spent at sea.

Finally, in June of 1941, Rochefort took over the command of what was then known as the Radio Unit of the 14th Naval District in Hawaii. To disguise its functions he renamed it the Combat Intelligence Unit. His mission was to find out, through communications intelligence, as much as possible about the dispositions and operations of the Japanese Navy. To this end he was to cryptanalyze all minor and one of the two major Japanese naval cryptosystems.

His chief target was the flag officers' system, the Japanese Navy's most difficult and the one in which it encased its most secret information. From about 1926 to the end of November, 1940, previous editions had provided the U.S. Navy with much of its information on the Japanese Navy. But the new version -- a four-character code with a transposition superencipherment -- was stoutly resisting the best efforts of the Navy's most skilled cryptanalysts, and Rochefort was urged to concentrate on it. The other major system, the main fleet cryptographic system, the most widely used, comprised a code with five-digit codenumbers to which were added a key of other numbers to complicate the system. The Navy called it the "five numeral system," or, more formally, JN25b -- the JN for "Japanese Navy," the 25 an identifying number, the b for the second (and current) edition. Navy cryptanalytic units in Washington and the Philippines were working on this code. Rochefort's unit did not attack this but did attack the eight or ten lesser codes dealing with personnel, engineering, administration, weather, fleet exercises.

But cryptanalysis was only part of the unit's task. The great majority of its 100 officers and men worked on two other aspects of radio intelligence -- direction-finding and traffic analysis.

Direction-finding locates radio transmitters. Since radio signals are heard best when the receiver points at the transmitter, sensitive antennas can find the direction from which a signal is coming by swinging until they hear it at its loudest. If two direction-finders take bearings like that on a signal and a control center draws the lines of direction on a map, the point at which they cross marks the position of the transmitter. Such a fix can tell quite precisely where, for example, a ship is operating. Successive fixes can plot its course and speed.

To exploit this source of information, the Navy in 1937 established the Mid-Pacific Strategic Direction-Finder Net. By 1941, high-frequency direction-finders curved in a gigantic arc from Cavite in the Philippines through Guam, Samoa, Midway, and Hawaii to Dutch Harbor, Alaska. The 60 or 70 officers and men who staffed these outposts reported their bearings to Hawaii, where Rochefort's unit translated them into fixes. For example, on October 16, the ship with call-sign KUNA 1 was located at 10.7 degrees north latitude, 166.7 degrees east longitude -- or within Japan's mandated islands.

These findings did not serve merely to keep an eye on the day-to-day locations of Japanese warships. They also formed the basis of the even more fruitful technique of traffic analysis. Traffic analysis deduces the lines of command of military or naval forces by ascertaining which radios talk to which. And since military operations are usually accompanied by an increase in communications, traffic analysis can infer the imminence of such operations by watching the volume of traffic. When combined with direction-finding, it can often approximate the where and when of a planned movement.

Radio intelligence thus maintains a long-range, invisible, and continuous surveillance of fleet movements and organization, providing a wealth of information at a low cost. Of course it has its limitations. A change of the call-signs of radio transmitters can hinder it. The sending of fictitious messages can befuddle it. Radio silences can deafen it. But it cannot be wholly prevented except by unacceptable restrictions on communications. Hence the Navy relied increasingly on it for its information on Japanese naval activities as security tightened in Japan during 1941, and almost exclusively after July, when the President's trade-freezing order deprived the Navy of all visual observations of Japanese ships not on the China coast.

It was in July that a Japanese tactic set up a radio pattern that was later to deceive the Combat Intelligence Unit. The Nipponese militarists had decided to take advantage of France's defeat and occupy French Indochina. The naval preparations for the successful grab were clearly indicated in the radio traffic, which went through the usual three stages that preceded major Japanese operations. First appeared a heavy flurry of messages. The Commander-in-Chief Combined Fleet busily originated traffic, talking with many commands to the south, thereby indicating the probable direction of his advance. Then came a realignment of forces. In the lingo of the tranalysis people, certain chickens (fleet units) no longer had their old mothers (fleet commanders). Call-sign NOTA 4, which usually communicated with OYO 8, now talked mostly with ORU 6. Accompanying this was a considerable confusion in the routing of messages, with frequent retransmissions caused by the regrouping: Admiral Z not here; try Second Fleet. Then followed the third phase: radio silence. The task force was now under way. Messages would be addressed to it, but none would emanate from it.

During all this, however, not only were no messages heard from the aircraft carriers, none were sent to them, either. This blank condition exceeded radio silence, which suppresses traffic in only one direction -- from the mobile force -- not in both. American intelligence reasoned that the carriers were standing by in home waters as a covering force in case of counterattack, and that communications both to and from them were not heard because they were being sent out by short-range, low-powered transmissions that died away before reaching American receivers. Such a blank condition had obtained in a similar tactical situation in February. American intelligence had drawn the same conclusions then and had been proven right. Events soon confirmed the July assessment as well. Twice, then, a complete blank of carrier communications combined with indications of a strong southward thrust had meant the presence of the carriers in Empire waters. But what happened in February and July was not necessarily what would happen in December.

During the summer and fall of 1941, the pressure of events molded America's two cryptanalytic agencies closer and closer to the form they were to have on December 7. The Signal Intelligence Service, which had 181 officers, enlisted men, and civilians in Washington and 150 at intercept stations in the field on Pearl Harbor Day, had been headed since March by Lieutenant Colonel Rex W. Minckler, a career Signal Corps officer. Friedman served as his chief technical assistant. S.I.S. comprised the Signal Intelligence School, which trained Regular Army and Reserve officers in cryptology, the 2nd Signal Service Company, which staffed the intercept posts, and four Washington sections of the S.I.S. proper: the A, or administrative, which also operated the tabulating machinery; the B, or cryptanalytic; the C, or cryptographic, which prepared new U.S. Army systems, studied the current systems for security, and monitored Army traffic for security violations; and the D, or laboratory, which concocted secret inks and tested suspected documents.

The B section, under Major Harold S. Doud, a West Point graduate, had as its mission the solution of the military and diplomatic systems not only of Japan but of other countries. In this it apparently achieved at least a fair success, though no Japanese military systems -- the chief of which was a code employing four-digit codenumbers -- were readable by December 7 because of a paucity of material. Doud's technical assistant was a civilian, Frank B. Rowlett, one of the three original junior cryptanalysts hired in 1930. The military man in charge of Japanese diplomatic solutions was Major Eric Svensson.

The Navy's official designation of OP-20-G indicated that the agency was the G section of the 20th division of OPNAV, the Office of the Chief of Naval Operations, the Navy's headquarters establishment. The 20th division was the Office of Naval Communications, and the G section was the Communication Security Section. This carefully chosen name masked its cryptanalytic activities, though its duties did include U.S. Navy cryptography.

Its chief was Commander Laurence F. Safford, 48, a tall, blond Annapolis graduate who was the Navy's chief expert in cryptology. In January, 1924, he had become the officer in charge of the newly created research desk in the Navy's Code and Signal Section. Here he founded the Navy's communication-intelligence organization. After sea duty from 1926 to 1929, he returned to cryptologic activities for three more years, when sea duty was again made necessary by the "Manchu" laws, which required officers of the Army and Navy to serve in the field or at sea to win promotion. He took command of OP-20-G in 1936. One of his principal accomplishments before the outbreak of war was the establishment of the Mid-Pacific Strategic Direction-Finder Net and of a similar net for the Atlantic, where it was to play a role of immense importance in the Battle of the Atlantic against the U-boats.

Safford's organization enjoyed broad cryptologic functions. It printed new editions of codes and ciphers and distributed them, and contracted with manufacturers for cipher machines. It developed new systems for the Navy. It comprehended such subsections as GI, which wrote reports based on radio intelligence from the field units, and GL, a record-keeping and historical-research group. But its main interest centered on cryptanalysis.

This activity was distributed among units in Washington, Hawaii, and the Philippines. Only Washington attacked foreign diplomatic systems and naval codes used in the Atlantic theater (primarily German). Rochefort had primary responsibility for the Japanese naval systems. The Philippines chipped away at JN25 and did some diplomatic deciphering, with keys provided by Washington. That unit, which like Rochefort's was attached for administrative purposes to the local naval district (the 16th), was installed in a tunnel of the island fortress of Corregidor. It was equipped with 26 radio receivers, apparatus for intercepting both high- and low-speed transmissions, a direction-finder, and tabulating machinery. Lieutenant Rudolph J. Fabian, 33, an Annapolis graduate who had had three years of communication intelligence in Washington and the Philippines, commanded. The 7 officers and 19 men in his cryptanalytic group exchanged possible recoveries of JN25b codegroups with Washington and with a British group in Singapore; each group also had a liaison man with the other.

Of the Navy's total radio-intelligence establishment of about 700 officers and men, two thirds were engaged in intercept or direction-finding activities and one third -- including most of the 80 officers -- in cryptanalysis and translation. Safford sized up the personnel of his three units this way: Pearl Harbor had some of the best officers, most of whom had four or five years of radio intelligence experience; the crew at Corregidor, which in general had only two or three years' experience, was "young, enthusiastic, and capable"; Washington -- responsible for both overall supervision and training -- had some of the most experienced personnel, with more than ten years' experience, and many of the least: 90 per cent of the unit had less than a year's experience.

Under Safford in the three subsections most closely involved with cryptanalysis were Lieutenant Commanders George W. Welker of GX, the intercept and direction-finding subsection, Lee W. Parke of GY, the cryptanalytical subsection, and Kramer of GZ, the translation and dissemination subsection. GY attacked new systems and recovered new keys for solved systems, such as PURPLE. But while it made the initial breaks in code solutions, the detailed recovery of codegroups (which was primarily a linguistic problem as compared to the more mathematical cipher solutions) was left to GZ. Four officers in GY, assisted by chief petty officers, stood round-the-clock watches. Senior watch officer was Lieutenant (j.g.) George W. Lynn; the others were Lieutenants (j.g.) Brotherhood, Pering, and Allan A. Murray. GY had others on its staff, such as girl typists who also did the simple deciphering of some diplomatic messages after the watch officers and other cryptanalysts had found the keys.

Kramer was in an odd position. Though he worked in OP-20-GZ, he was formally attached to OP-16-F2 -- the Far Eastern Section of the Office of Naval Intelligence. This arrangement was intended in part to throw off the Japanese, who might have inferred some measure of success in codebreaking if a Japanese-language officer like Kramer were assigned to communications, in part to have an officer with a broad intelligence background distribute MAGIC so that he could answer the recipients' questions. Kramer, 38, who had studied in Japan from 1931 to 1934, had had two tours in O.N.I. proper before being assigned full time to GZ in June, 1940. An Annapolis graduate, chess fan, and rifle marksman, he lived in a world in which everything had one right way to be done. He chose his words with almost finicky exactness (one of his favorites was "precise"); he kept his pencil mustache trimmed to a hair; he filed his papers tidily; he often studied his MAGIC intercepts several times over before delivering them. Included in this philosophy was his duty. He performed it with great responsibility, intelligence, and dedication.

The first task of OP-20-G and of S.I.S. was to obtain raw material for the cryptanalysts. And in peacetime America that was not easy.

Section 605 of the Federal Communications Act of 1934, which prohibits wiretaps, also prohibits the interception of messages between foreign countries and the United States and territories. General Malin Craig, Chief of Staff from 1937 to 1939, was acutely aware of this, and his attitude dampened efforts to intercept the Japanese diplomatic messages coming into America. But after General George C. Marshall succeeded to Craig's post, the exigencies of national defense relegated that problem in his mind to the status of a legalistic quibble. The cryptanalytic agencies pressed ahead in their intercept programs. The extreme secrecy in which they were cloaked helped them avoid detection. They concentrated on radio messages, since the cable companies, fully cognizant of the legal restrictions, in general refused to turn over any foreign communications to them. Consequently, 95 per cent of the intercepts were radio messages. The remainder was split between cable intercepts and photographs of messages on file at a few cooperative cable offices.

To pluck the messages from the airwaves, the Navy relied mainly on its listening posts at Bainbridge Island in Puget Sound; Winter Harbor, Maine; Cheltenham, Maryland; Heeia, Oahu; and Corregidor and to a lesser degree on stations at Guam; Imperial Beach, California; Amagansett, Long Island and Jupiter, Florida. Each station was assigned certain frequencies to cover. Bainbridge Island, which was called Station S, copied solid the schedule of Japanese government messages between Tokyo and San Francisco. Its two sound recorders guarded the radiotelephone band of that circuit; presumably it was equipped to unscramble the relatively simple sound inversion that then provided privacy from casual eavesdropping. Diplomatic messages were transmitted almost exclusively by commercial radio using roman letters. The naval radiograms, however, employed the special Morse code devised for kata kana, a syllabic script of Japanese. The Navy picked these up with operators trained in Japanese Morse and recorded them on a special typewriter that it had developed for the roman-letter equivalents of the kana characters. The Army's stations, called Monitor Posts, were: No. 1, Fort Hancock, New Jersey; No. 2, San Francisco; No. 3, Fort Sam Houston, San Antonio; No. 4, Panama; No. 5, Fort Shafter, Honolulu; No. 6, Fort Mills, Manila; No. 7, Fort Hunt, Virginia; No. 9, Rio de Janeiro.

At first both services airmailed messages from their intercept posts to Washington. But this proved too slow. The Pan-American Clipper, which carried Army intercepts from Hawaii to the mainland, departed only once a week on the average, and weather sometimes caused cancellations, forcing messages to be sent by ship. As late as the week before Pearl Harbor, two Army intercepts from Rio did not reach Washington for eleven days. Such delays compelled the Navy to install teletypewriter service in 1941 between Washington and its intercept stations in the continental U.S. The station would perforate a batch of intercepts onto a teletype tape, connect with Washington through a teletypewriter exchange, and run the tape through mechanically at 60 words per minute, cutting toll charges to one third the cost of manually sending each message individually. Outlying stations of both the Army and Navy picked out Japanese messages bearing certain indicators, enciphered the Japanese cryptograms in an American system, and radioed them to Washington. The reencipherment was to keep the Japanese from knowing of the extensive American cryptanalytic effort. Only the three top Japanese systems were involved in this expensive radio retransmission: PURPLE, RED (a machine system that antedated PURPLE, which had supplanted it at major embassies, but that was still in use for legations such as Vladivostok), and the J series of enciphered codes. The Army did not install a teletype for intercepts from its continental posts until the afternoon of December 6, 1941; the first messages (from San Francisco) were received in the early morning hours of December 7.

The intercept services missed little. Of the 227 messages pertaining to Japanese-American negotiations sent between Tokyo and Washington from March to December, 1941, all but four were picked up.

In Honolulu, where a large Japanese population produced nightmares of antlike espionage and potential sabotage, the 14th Naval District's intelligence officer, Captain Irving S. Mayfield, had long sought to obtain copies of the cablegrams of Consul General Nagao Kita. If Rochefort's unit could solve these, Mayfield figured, he might know better which Japanese to shadow and what information they sought.

His intuitions were sound. On March 27, 1941, not two weeks after Mayfield himself took up his duties, a young ensign of the Imperial Japanese Navy, 25-year-old Takeo Yoshikawa, who had steeped himself in information about the American Navy, arrived in Honolulu to serve as Japan's only military espionage agent covering Pearl Harbor. Under the cover-name "Tadasi Morimura," he was assigned to the consulate as a secretary. He promptly made himself obnoxious -- and drew suspicion upon himself within the consulate staff -- by coming to work late or not at all, getting drunk frequently, having women in his quarters overnight, and even insulting the consul himself on occasion. But he managed to tour the islands, and within a month was sending such messages as: "Warships observed at anchor on the 11th [of May, 1941] in Pearl Harbor were as follows: Battleships, 11: Colorado, West Virginia, California, Tennessee...." These were sent in the consulate's diplomatic systems, not in naval code.

But Mayfield's hopes of peering into these secret activities through the window of a broken code were stymied by the refusal of the cable offices to violate the statute against interception. His desires grew more intense as another source failed to yield any information of counterespionage activity. For months one of his enlisted men, Theodore Emanuel, had tapped half a dozen of the consulate's telephone lines, recorded the 50 or 60 calls made on them each day, and turned the recordings over to Lieutenant Denzel Carr for translation and summarization. But this eavesdropping produced at best some juicy items about bachelor Kita's sex life (such as his chasing a maid around a bedpost one night after a sake-soaked Japanese wedding); there was little to help Mayfield.

So when David Sarnoff, president of the Radio Corporation of America, vacationed in Hawaii, Mayfield spoke to him. It was subsequently arranged that thenceforth R.C.A.'s Japanese consulate messages would be quietly given to the naval authorities. But the consulate rotated its business among the several cable companies in Honolulu, and R.C.A.'s turn was not due until December 1.

In Washington, however, intercepts overwhelmed GY and S.I.S. The tiny staff of cryptanalysts simply could not cope with all of them expeditiously. This difficulty was resolved in two ways.

One was to cut out duplication of effort. At first, both services solved all their Japanese diplomatic intercepts. But beginning more than a year before Pearl Harbor, messages originating in Tokyo on odd-numbered days of the month were handled by the Navy, those on even days, by the Army. Each began breaking the messages sent in from its own intercept stations until it reached the Tokyo date of origin; it would then retain them or send them over as the dates indicated. The cryptanalysts utilized the extra time to attack as-yet-unbroken systems and to clean up backlogs.

The other method was to concentrate on the important intercepts and let the others slide, at least until the important ones were completed. But how can a cryptanalyst tell which messages are important until he has solved them? He cannot, but he can assume that messages sent in the more secret systems are the more important. All dispatches cannot be transmitted in a single system because the huge volume of traffic would enable cryptanalysts to break it too quickly. Hence most nations set up a hierarchy of systems, reserving the top ones for their vital needs.

Japan was no exception. Though her Foreign Office employed an almost bewildering variety of different codes, resorting, from time to time, to the Yokohama Specie Bank's private code, a Chinese ideographic code list, and codes bearing kata kana names, such as TA, JI, or HEN, it relied in the main on four systems. American cryptanalysts ranked these on four levels according to the inherent difficulty of their solution and the messages that they generally carried. Intercepts were then solved in the order of this priority schedule.

Simplest of all, and hence the lowest in rank and last to be read (excluding plain language), was the LA code, so called from the indicator group LA that preceded its codetexts. LA did little more than put kata kana into roman letters for telegraphic transmission and to secure some abbreviation for cable economy. Thus the kana for ki was replaced by the code form CI, the kana for to by IF, the two-kana combination of ka + n by CE. Its two-letter codewords, all of either vowel-consonant or consonant-vowel form and including such as ZO for 4, were supplemented by a list of four-letter codewords, such as TUVE for dollars, SISA for ryoji ("consul"), and XYGY for Yokohama. A very typical LA message is serial 01250 from the Foreign Minister to Kita, dated December 4, which begins in translation: "The following has been authorized as the year-end bonus for employee typists of your office." This sort of code is generally called a "passport code" because it usually serves for messages covering the administrative routine of a mission, such as issuance of passports and visas. LA was a particularly simple one to solve, partly because it had been in effect since 1925, partly because of the regularities in its construction. For example, all kana that ended in e had as code equivalents groups beginning with A (ke = AC, se = AD), and all that began with k had code equivalents beginning or ending with C. Identification of one kana would thus suggest the identification of others.

One rung up the cryptographic ladder was the system known to the Japanese as Oite and to American codebreakers as PA-K2. The PA part was a two- and four-letter code similar to the LA, though much more extensive and with codegroups disarranged. The K2 part was a transposition based on a keynumber. The letters from the PA encoding were written under this keynumber from right to left and then copied out in mixed order, taking first the letter under number 1, then the letter under number 2, until the row was completed. The process was repeated for successive rows.

For example, on December 4 Yoshikawa wired the Foreign Minister that "At 1 o'clock on the 4th a light cruiser of the Honolulu class hastily departed -- Morimura." In romaji (the roman-letter version of the kata kana) this became 4th gogo 1 kei jun (honoruru) kata hyaku shutsu ko -- morimura. In PA, with the parentheses getting their own codegroups (OQ and UQ), it assumed this form: BYDH DOST JE YO IA OQ GU RA HY HY UQ VI LA YJ AY EC TY FI BANL, with FI indicating use four-letter code. (The code clerk made two errors. After encoding kata by vi, he encoded an extra ta into LA and an unnecessary re into TY.) This was then written under the keynumber from right to left, with an extra letter I as a null to complete the final five-letter group:

Transcribed line by line according to the numbers (S under 1 first, D under 2 second, etc.), prefixed with system indicator GIGIG and key indicator AUDOB, the message number, and the telegraphic abbreviation of Sikuyu ("urgent"), the message (with three more errors: the Y under 13 became the J in CJYHH, the F under 2 became the E in IYJIE, and the T under 9 became the I in AUIAY) became the one actually sent over Kita's name:

PA-K2 did not pose much of a problem to experienced American cryptanalysts. Rochefort estimated that his unit could crack a PA-K2 message in from six hours to six days, with three days a good average. The transposition was vulnerable because each line was shuffled identically; the cryptanalyst could slice a cryptogram into groups of 15 or 17 or 19 and anagram these simultaneously until the predominant vowel-consonant alternation appeared on all lines; the underlying code could then be solved by assuming that the most frequent codegroups represented the most frequent kana (i, followed by ma, shi, o, etc.) and filling out the skeleton words that resulted. Since the system had remained in use for several years, this reconstruction had long been accomplished by the Washington agencies. Hence solution involved only unraveling any new transposition and, with luck, might take only a few hours. It could also take a few days. Primarily because of PA-K2'S deferred position in the priority list, an average of two to four days elapsed between interception and translation.

The code clerk in Honolulu enveloped Yoshikawa's final messages in PA-K2 only because higher-level codes had been destroyed December 2 on orders from Tokyo. Normally, espionage reports of shipping movements and military activities, sent routinely by Japanese consuls from their posts all over the world, were framed on that next level of secrecy. Here prevailed a succession of codes called TSU by the Japanese and the J series by Americans. These were even more extensive and more thoroughly disarranged than PA, and they were transposed by a system of far greater complexity than the rather simple and vulnerable K2. Furthermore, the code and the transposition were changed at frequent intervals. Thus J17-K6 was replaced on March 1 by J18-K8, and that in turn by J19-K9 on August 1.

The transposition was the real stumbling block. Like the K2, it used a keynumber, but it differed in being copied off vertically instead of horizontally, and in having a pattern of holes in the transposition blocks. These holes were left blank when the code groups are inscribed into the block. For example, letting the alphabet from A to V serve as the code message:

The letters were transcribed in columns in the order of the keynumbers, skipping over the blanks: BJMV EHKT NW CGORX AFILQU DPSY. This would be sent in the usual five-letter groups.

The first step in solving a columnar transposition like this, but without blanks, is to cut the cryptogram into the approximately equal segments that the cryptanalyst believes represent the columns of the original block. The blanks vastly increase the difficulty of this essential first step because they vary the length of the column segments. The second step is to reconstruct the block by trying one segment next to the other until a codeword-like pattern appears. Here again the blanks, by introducing gaps in unknown places between the letters of the segments, greatly hinder the cryptanalyst.

The problems of solving such a system are illustrated by the fact that J18-K8 was not broken until more than a month after its introduction. The cryptanalysts had to make a fresh analysis for each pattern of blanks and each transposition key. The key changed daily, the blank-pattern three times a month. Hence J19-K9 solutions were frequently delayed. The key and pattern for November 18 were not recovered until December 3; those for November 28, not until December 7. On the other hand, solution was sometimes effected within a day or two. Success usually depended on the quantity of intercepts in a given key. About 10 or 15 per cent of J19-K9 keys were never solved.

This situation contrasts with that of PURPLE, the most secret Japanese system, in which all but 2 or 3 per cent of keys were recovered and in which most messages were solved within hours. Did the Japanese err in assessing the security of their systems? Yes and no. PURPLE was easier to keep up with once it was solved, but it was a much more difficult system to break in the first place than J19-K9. The solution of the PURPLE machine was, in fact, the greatest feat of cryptanalysis the world had yet known.

The cipher machine that Americans knew as PURPLE bore the resounding official Japanese title of 97-shiki O-bun In-ji-ki. This meant Alphabetical Typewriter '97, the '97 an abbreviation for the year 2597 of the Japanese calendar, which corresponds to 1937. The Japanese usually referred to it simply as "the machine" or as "J," the name given it by the Imperial Japanese Navy, which had adapted it from the German Enigma cipher machine and then had lent it to the Foreign Ministry, which, in turn, had further modified it. Its operating parts were housed in a drawer-sized box between two big black electrically operated Underwood typewriters, which were connected to it by 26 wires plugged into a row of sockets called a plugboard. To encipher a message, the cipher clerk would consult the thick YU GO book of machine keys, plug in the wire connections according to the key for the day, turn the four disks in the box so the numbers on their edges were those directed by the YU GO, and type out the plaintext. His machine would record that plaintext while the other, getting the electric impulses after the coding box had twisted them through devious paths, would print out the ciphertext. Deciphering was the same, though the machine irritatingly printed the plaintext in the five-letter groups of the ciphertext input.

The Alphabetical Typewriter worked on roman letters, not kata kana. Hence it could encipher English as well as romaji -- and also roman-letter codetexts, like those of the J codes. Since the machine could not encipher numerals or punctuation, the code clerk first transformed them into three-letter codewords, given in a small code list, and enciphered these. The receiving clerk would restore the punctuation, paragraphing, and so on, when typing up a finished copy of the decode.

The guts and heart of the machine were the plugboards and the coding wheels. They diverted the current flowing along the connections from the input typewriter to the output one so that when the a key was depressed on the input keyboard an a would not be typed on the output machine. The diversion began with the plugboard connections. If the coding box were not present, a plugboard wire would take the electric impulse from the a key of the plaintext typewriter and bring it directly to, say, the R typebar of the ciphertext machine. Other wires would similarly connect the plaintext keys to noncorresponding ciphertext typebars. This would automatically produce a cipher, though a very elementary one. Each time plaintext a was depressed ciphertext R would appear. So simple a system affords no security. The plugboard connections can be changed from message to message, or even within a message, but this does not noticeably augment the system's strength.

Here is where the four coding wheels came in. Interposed between the plugboard of the plaintext typewriter and that of the ciphertext machine, they were shifted constantly with respect to one another by their supporting assembly. The enciphering current had to traverse their winding wire paths to get from one typewriter to the other, and the constant shifting continually set up different paths. Thus impulses from a given plaintext letter were switched through the box along ever varying detours to emerge at ever differing ciphertext letters. Plaintext a might be represented in a long message by all 26 letters. Conversely, any given ciphertext letter might stand for any one of 26 plaintext letters. Switches on the coding wheels could be flicked one way or the other; this constituted part of the key and was done by the code clerk before enciphering. Usually the plugboard connections were changed each day.

These factors united to produce a cipher of exceptional difficulty. The more a cipher deviates from the simple form in which one ciphertext letter invariably replaces the same plaintext letter, the harder it is to break. A cipher might replace a given plaintext letter by five different ciphertext letters in rotation, for example. But the Alphabetical Typewriter produced a substitution series hundreds of thousands of letters long. Its coding wheels, stepping a space -- or two, or three, or four -- after every letter or so, did not return to their original positions to re-create the same series of paths, and hence the same sequence of substitutes, until hundreds of thousands of letters had been enciphered. The task of the cryptanalysts consisted primarily of reconstructing the wiring and switches of the coding wheels -- a task made more burdensome by the daily change of plugboard connections. Once this was done, the cryptanalyst still had to determine the starting position of the coding wheels for each day's messages. But this was a comparatively simple secondary job.

American cryptanalysts knew none of these details when the Japanese Foreign Office installed the Alphabetical Typewriter in its major embassies in the late 1930s. How, then, did they solve it? Where did they begin? How did they even know that a new machine was in service, since the Japanese government did not announce it?

The PURPLE machine supplanted the RED machine, which American cryptanalysts had solved, and so probably their first clue to the new machine was the disconcerting discovery that they could no longer read the important Japanese messages. At the same time, they observed new indicators for the PURPLE system. Clues to the system's nature came from such characteristics of its ciphertext as the frequency of letters, the percentage of blanks (letters that did not appear in a given message), and the nature and number of repetitions. Perhaps the codebreakers also assumed that the new machine comprised essentially a more complicated and improved version of the one it replaced. In this they were right.

Their first essays at breaking into the cipher both accompanied and supplemented their attempts to determine the type of cipher. Their previous success with the RED machine and with the lesser systems had given them insight into the Japanese diplomatic forms of address, favorite phrases, and style (paragraphs were often numbered, for example). These provided the cryptanalysts with probable words -- words likely to be in the plaintext -- that would help in breaking the cipher. Opening and closing formulas, such as "I have the honor to inform Your Excellency" and "Re your telegram," constituted virtual cribs. Newspaper stories suggested the subject matter of intercepts. The State Department sometimes made public the full texts of diplomatic notes from Japan to the American government, in effect handing the cryptanalysts the plaintext (or its translation) of an entire dispatch. (State reportedly did not pass the texts of confidential notes to the cryptanalysts, though this would have helped them considerably and was done by other foreign ministries.) Japan's Foreign Office often had to circulate the same text to several embassies, not all of which had a PURPLE machine, and a code clerk might have inadvertently encoded some cables in PURPLE, some in other systems -- which the cryptanalysts could read. A comparison of times of dispatch and length, and voilà! -- another crib to a cryptogram. Errors were, as always, a fruitful source of clues. As late as November, 1941, the Manila legation repeated a telegram "because of a mistake on the plugboard." How much more common must errors have been when the code clerks were just learning to handle the machine! The sending of the identical text in two different keys produces "isomorphic" cryptograms that yield exceedingly valuable information on the composition of the cipher.

The cryptanalysts of S.I.S. and OP-20-G, then, matched these assumed plaintexts to their ciphertexts and looked for regularities from which they could derive a pattern of encipherment. This kind of work, particularly in the early stages of a difficult cryptanalysis, is perhaps the most excruciating, exasperating, agonizing mental process known to man. Hour after hour, day after day, sometimes month after month, the cryptanalyst tortures his brain to find some relationship between the letters that hangs together, does not dead-end in self-contradiction, and leads to additional valid results. "Most of the time he is groping in the darkest night," one solver has written. "Now and again a little flicker of light gleams across the darkness, tantalizing him with a glimpse of a path. Hopefully he dashes to it only to find himself in another labyrinth. His knowledge that night is inevitably followed by day keeps his waning courage up, and he steers his course towards where the morning sun is soon to appear. Except that sometimes he is engulfed in an interminable polar night."

It must have seemed like that interminable night to the cryptanalysts who began attacking the new Japanese machine. The codebreakers went just so far -- and for months could not push on further. As William Friedman recalled, "When the PURPLE system was first introduced it presented an extremely difficult problem on which the Chief Signal Officer [Mauborgne] asked us to direct our best efforts. After work by my associates when we were making very slow progress, the Chief Signal Officer asked me personally to take a hand. I had been engaged largely in administrative duties up to that time, so at his request I dropped everything else that I could and began to work with the group."

Friedman was (and is) the world's greatest cryptologist. Then in his late forties, he was a quiet, studious man, well liked by his associates, of average height and build, and a natty dresser given to bow ties. Trained as a geneticist, he had become interested in cryptology in 1915 at a research institution in Illinois called the Riverbank Laboratories. He served as a cryptanalyst with the American Expeditionary Forces in World War I, and returned to Riverbank to write an 87-page tract that revolutionized cryptanalysis by introducing statistical methods for the first time. Hired by the Signal Corps in 1921, he applied these methods to a cipher-machine solution that placed America in the forefront of world cryptology. During these years, his wife, the former Elizebeth Smith, whom he had met and married at Riverbank, was solving rumrunners' codes for the Coast Guard. He wrote textbooks in cryptanalysis that are models of clarity. He became head of S.I.S. when it was founded and continued to exercise his extraordinary cryptanalytic abilities. His genius soon manifested itself in the attack on PURPLE.

Lighting his way with some of the methods that he himself had developed, he led the cryptanalysts through the murky PURPLE shadowland. He assigned teams to test various hypotheses. Some prospected fruitlessly, their only result a demonstration that success lay in another direction. Others found bits and pieces that seemed to make sense. (OP-20-G cooperated in this work, with Harry L. Clark making especially valuable contributions, but S.I.S. did most of it.) Friedman and the other codebreakers began to segregate the ciphertext letters into cycles representing the rotation of the coding wheels -- gingerly at first, then faster and faster as the evidence accumulated. The polyalphabetic class of ciphers, to which PURPLE belonged, is based ultimately upon an alphabet table, usually 26 letters by 26. To reconstruct the PURPLE tables, the cryptanalysts employed both direct and indirect symmetry of position -- names only slightly less forbidding than the methods they denote. Errors, caused perhaps by garbled interceptions or simple mistakes in the crypt-analysis, jarred these delicate analyses and delayed the work. But slowly it progressed. A cryptanalyst, brooding sphinxlike over the cross-ruled paper on his desk, would glimpse the skeleton of a pattern in a few scattered letters; he tried fitting a fragment from another recovery into it; he tested the new values that resulted and found that they produced acceptable plaintext; he incorporated his essay into the over-all solution and pressed on. Experts in Japanese filled in missing letters; mathematicians tied in one cycle with another and both to the tables. Every weapon of cryptanalytic science -- which in the stratospheric realm of this solution drew heavily upon mathematics, using group theory, congruences, Poisson distributions -- was thrown into the fray.

Eventually the solution reached the point where the cryptanalysts had a pretty good pencil-and-paper analog of the PURPLE machine. S.I.S. then constructed a mechanism that would do automatically what the cryptanalysts could do manually with their tables and cycles. They assembled it out of ordinary hardware and easily available pieces of communication equipment, such as the selector switches used for telephones. It was hardly a beautiful piece of machinery, and when not running just right it spewed sparks and made loud whirring noises. Though the Americans never saw the 97-shiki O-bun In-ji-ki, their contraption bore a surprising physical resemblance to it, and of course exactly duplicated it cryptographically.

S.I.S. handed in its first complete PURPLE solution in August of 1940, after 18 or 20 months of the most intensive analysis. In looking back on the effort that culminated in this, the outstanding cryptanalytic success in the whole history of secret writing up to its time, Friedman would say generously:

Naturally this was a collaborative, cooperative effort on the part of all the people concerned. No one person is responsible for the solution, nor is there any single person to whom the major share of credit should go. As I say, it was a team, and it was only by very closely coordinated teamwork that we were able to solve it, which we did. It represents an achievement of the Army cryptanalytic bureau that, so far as I know, has not been duplicated elsewhere, because we definitely know that the British cryptanalytic service and the German cryptanalytic service were baffled in their attempts and they never did solve it.