Mar 1933 to Apr. 1945, The Roosevelt Years 2nd Edition, part 9
Station CAST was the United States Navy signals monitoring and cryptographic intelligence fleet radio unit at Cavite Navy Yard in the Philippines, until it was captured by the Japanese forces in 1942. It was an important part of the Allied intelligence effort, addressing Japanese communications as the War expanded from China into the rest of the Pacific theaters. As Japanese advances in the Philippines threatened CAST, its staff and services were progressively transferred to Corregidor in Manila Bay, and eventually to a newly formed US-Australian station, FRUMEL in Melbourne, Australia.
Station CAST had originally been located at Shanghai but had been evacuated to Cavite in early 1941 as part of the US Navy's disengagement with China.
Prior to the war, CAST was the US Navy's Far East cryptographic operation, under the OP-20-G Naval Intelligence section in Washington. It was located at the Navy Yard in Manila and moved into the tunnels on Corregidor, as Japanese attacks increased. Station CAST possessed one of the Purple machines.
An early decision by OP-20-G divided responsibility for Japanese cryptanalysis amongst its various stations. Station CAST (at Manila in the Philippines), Station HYPO (Pearl Harbor, Hawaii), and OP-20-G itself in Washington, shared cryptanalytic duties. Other Stations (on Guam, in Puget Sound on Bainbridge Island, Guam, etc.) were tasked and staffed for signals interception and traffic analysis.
Furthermore, were rather capriciously distributed to high level officials in Washington, and in general, poorly used. SIS was able to build several PURPLE machine equivalents and the distribution of those machines has since been thought controversial. One was sent to Station CAST. After the US entered the War, two went to Bletchley Park, the center of British cryptographic work. One was sent to the Far East Combined Bureau (FECB) at Singapore, and was lost during the fall of Singapore 
Prior to the attack on Pearl Harbor, the amount of available traffic was low, and little progress had been made on the most important Japanese Navy system, called by US analysts JN-25. JN-25 was used for high level operations: movement and planning commands, for instance. It was a super-encrypted code, eventually a two-book system, and joint cryptanalytic progress was slow. JN-25B was introduced on December 1, 1940, but was broken immediately by the US. Most references cite about 10% of messages partially (or sometimes completely) decrypted prior to 1 December 1941, at which time a new edition of the system went into effect and sent all the cryptanalysts back to the beginning.
After December 7, 1941, there was considerably more JN-25 traffic as the Japanese Navy operational tempo increased and geographically expanded, and progress against it went better.
Station CAST and its personnel and equipment were moved from Manila to the tunnels on Corregidor as the Japanese approached and spent the next months working there. Eventually, they destroyed their equipment (some IBM punched card machines are said to have been among the gear shoved into the harbor) and were evacuated by submarine to Australia, for service with FRUMEL. Some personnel also worked at the Central Bureau, supplying signals intelligence to MacArthur's South West Pacific Area (command).
OP-20-G or "Office of Chief Of Naval Operations (OPNAV), 20th Division of the Office of Naval Communications, G Section / Communications Security", was the US Navy's signals intelligence and cryptanalysis group during World War II. Its mission was to intercept, decrypt, and analyze naval communications from Japanese, German, and Italian navies. In addition OP-20-G also copied diplomatic messages of many foreign governments. The majority of the sections effort was directed towards Japan and included breaking the early Japanese "Blue" book fleet code. This was made possible by intercept and High Frequency Direction Finder (HFDF) sites in the Pacific, Atlantic, and continental US, as well as a Japanese telegraphic code school for radio operators in Washington D.C.
The Code and Signal Section was formally made a part of the Division of Naval Communications (DNC), as Op-20-G, on July 1, 1922. In January 1924, a 34-year-old US Navy lieutenant named Laurance F. Safford was assigned to expand OP-20-G's domain to radio interception. He worked out of Room 2646, on the top floor of the Navy Department building in Washington DC.
Japan was of course a prime target for radio interception and cryptanalysis, but there was the problem of finding personnel who could speak Japanese. The Navy had a number of officers who had served in a diplomatic capacity in Japan and could speak Japanese fluently, but there was a shortage of radiotelegraph operators who could read Japanese Morse code communications sent in kana. Fortunately, a number of US Navy and Marine radiotelegraph operators operating in the Pacific had formed an informal group in 1923 to compare notes on Japanese kana transmissions. Four of these men became instructors in the art of reading kana transmissions when the Navy began conducting classes in the subject in 1928.
By June 1940, OP-20-G included 147 officers, enlisted men, and civilians, linked into a network of radio listening posts as far-flung as the Army's. OP-20-G did some work on Japanese diplomatic codes, but the organization's primary focus was on Japanese military codes. The US Navy first got a handle on Japanese naval codes in 1922, when Navy agents broke into the Japanese consulate in New York, cracked the safe, took photographs of pages of a Japanese navy codebook, and left, having put everything back as they had found it.
Before the war, the Navy cipher bureau operated out of three main bases:
- headquarters in Washington DC
- Station Hypo, a section at Pearl Harbor in Hawaii
- Station Cast, a section in the fortified caves of the island of Corregidor, in the Philippines. The codebreakers were backed up by a far-flung network of listening and radio direction finding stations.
The result was that much of the MAGIC was wasted. There was no efficient process for assessing and organizing the intelligence, or getting it to its proper end users. This was a dangerous problem as the time was rapidly approaching when that data would be a matter of life and death.
In the dark hours of the morning of December 7, 1941, the U.S. Navy communications intercept station at Fort Ward on Bainbridge Island, Washington, picked up a radio message being sent by the Japanese government to the Japanese embassy in Washington DC. It was the last in a series of 14 messages that had been sent over the previous 18 hours.
The messages were decrypted by a PURPLE analogue machine at OP-20-G and passed to the SIS for translation from Japanese, early on the morning of December 7. Army Colonel Rufus S. Bratton and Navy Lieutenant Commander Alvin Kramer independently inspected the decrypts.
They both became alarmed. The decrypts instructed the Japanese ambassador to Washington to inform the US Secretary of State, Cordell Hull, at 1:00 PM Washington time that negotiations between the United States and Japan were ended. The embassy was then to destroy their cipher machines. This sounded like war, and although the message said nothing about any specific military action, Kramer also realized that the sun would be rising over the expanses of the central and western Pacific by that time. The two men both tried to get in touch with Army Chief of Staff General George C. Marshall.
SIGINT after Pearl Harbor
As Japanese advances in the Philippines, the possibility of an invasion of Hawaii, and the increasing demand for intelligence, OP-20-G undertook two courses of action: the staff and services of CAST were progressively transferred to a newly-formed US-Australian-British station, FRUMEL in Melbourne, Australia. Another signals intelligence center, known as NEGAT was formed in Washington, using elements of OP-20-G headquarters.
Secretary of War Stimson called upon a prominent Chicago lawyer, Alfred McCormack, to examine the implications of the problem. McCormack recommended the creation of a branch within the Military Intelligence Service to deal with the processing of communications intelligence. The Special Branch was established in May 1942 with Col. Carter W Clarke as its head and with McCormack, now commissioned as a colonel, as his deputy. To acquire the necessary high caliber personnel to staff the new organization, McCormack drew heavily on lawyers from elite firms, who were given reserve commissions.
In late April 1943, Col. William Friedman, (US Army Signal Security Service) Col. Alfred McCormack, and Lt. Col. Telford Taylor (US Military Intelligence Service, Special Branch) traveled to Great Britain to meet with British cryptologists which lasted until May 12, 1943. It was in the same period that the 1943 Travis-Strong Agreement on SIGINT cooperation was negotiated and it appears this US delegation was part of that process. This agreement was made by Major General George Strong, US Army, G-2 (Intelligence).
Colonel Alfred McCormack, Special Assistant to the Secretary of State for Research and
Intelligence, took on the dual task of reassembling the OSS programs as Department of State
entities and of providing advice on the interagency intelligence planning mandated by the President.
Blowback - America's recruitment of Nazis, and its disastrous effect on our domestic and foreign policy by Christopher Simpson Collier / Macmillan, 1988
Information Warfare: Impacts and Concerns by Col James W. McLendon, USAF
Colonel, later Brigadier General, W. Preston Corderman commanded the Army's Signal Security Agency, a central code breaking organization. Colonel William F. Friedman, the creator of American Army cryptology, was its director of research.
Terminology was only one aspect of the new procedures. The actual dissemination of ULTRA in the field was handled by Special Security Officers (SSOs) selected and trained by the Military Intelligence Service, Special Branch and operating under its direct command using special cipher systems. This was both faster and more secure than the usual practice of sending intelligence through successive layers of command channels. Again, this followed British practice. The first three American SSOs went out to commands in the Pacific in the fall of 1943. The British agreement that ULTRA supplied to American commanders in the European theater would henceforth be disseminated through American channels led to the procurement of eighty additional special security officers in December 1943.
At first, the SSOs were attached only to the highest level of commands. They had the dual mission of securing the vital ULTRA material and of explaining its significance to commanders and intelligence officers unfamiliar with the uses and limitations of high-grade communications intelligence. However, by July 1944 the decision was made to disseminate ULTRA directly to field armies and down to independently operating Army corps.
The Army's own increasing successes in communications intelligence, combined with the new availability of the British COMINT product, helped bring about a general reorganization of the Military Intelligence Service separate from the Military Intelligence Division. Because of the growth of its responsibilities, the Special Branch expanded to the point where it constituted the largest component of the MIS Intelligence Group.
Historian and CIA intelligence analyst Ray Cline later described the situation, the "G-2 (staff) was so little geared into high-level strategic decisions that it was engaged in a colossal struggle for office space when D-Day for the Normandy invasion came along: all of G-2's files were locked up sitting in safes in the halls waiting for moving crews when frantic requests for data on the landing zone situations began to descend on the hapless Army intelligence officers, who hardly knew each others' phone numbers, let alone what was in the files."
In addition, certain activities previously carried out at dispersed locations in the United States were now centralized in Washington, D.C. The American Intelligence Service operating from Miami, Florida, had been terminated in January 1944. Now that the war had been carried to the shores of Europe, America was no longer worried about Axis subversion in Latin America. Similarly, the branches previously established in San Francisco, New Orleans, and New York City had lost most of their usefulness as the war progressed and other sources of intelligence became available.
Since 1943 communications intelligence, Military Intelligence Service's main source of intelligence had been under the control of the chief signal officer or of theater commanders. In December 1944 MIS at last secured operational control over the Signal Security Agency, and was able to give direction to the Army's most important COMINT asset. This step also positioned MIS to take over this whole field of intelligence when the war came to a close.
In 1943, the Army Signal Intelligence Service (later the Army Security Agency) began intercepting Soviet intelligence traffic sent mainly from New York Cityassigning the code name VENONA to the project. By 1945, some 200,000 messages had been transcribed, a measure of Soviet activity. On December 20, 1946, Meredith Gardner made the first break into the VENONA code, revealing the existence of Soviet espionage at the Los Alamos National Laboratory.
"After intense study of the Japanese language, I was stationed at Arlington Hall as part of the 2nd Sig Serv Bn. I was given training in cryptography at Vint Hills Farms before the stint at Arlington Hall. We were reading the 1944 Japanese Army General Codebook. There were always glitches, both from faulty reception and failure to break the monthly cyphers. It was our job to fill these in (guess at them) and get the translated messages to McArthur's HQ in the Pacific. We were aided by several two-story barracks filled with IBM computers (then called punch-card tabulators). In January, the new 1945 codebook was published by the Japanese and we were pretty much back to square one. Myself and another expendable Lt. were sent to the Pacific looking for captured code equipment. Our knowledge of what it looked like would supposedly qualify us to find such. No luck on Saipan. 1st Marines did find the code books in the catacombs under Shuri Castle. Back at AH we translated and decrypted, until one day I got a message with a new word: Genshi Bakudan...atom bomb. War ended. From there it was to Japan to interview civilians and military (USSBS). Then discharge and back to Texas."
Events of the spring of 1943 would reshape the entire structure of Military Intelligence. In April Army cryptanalysts scored their first success against Japanese military codes. A month later, a party of officers from MIS and from the Signal Security Agency visited the British Government Code and Cypher School at Bletchley Park. For the first time, American Military Intelligence became aware of the dimensions of the British success against high-level German communications. British efforts in breaking the German Enigma and other ciphers used on command links had laid bare many of the most important secrets of the Nazi high command. The intelligence derived from this source, known as ULTRA, was disseminated by the British under rigidly controlled conditions. Although such intelligence had been provided to Eisenhower during the invasion of North Africa, U.S. Army Intelligence had not been fully aware of its origins. Now the British agreed to share this intelligence with the U.S. Army on an unrestricted basis, in exchange for reciprocal access to American communications intelligence on the Japanese.
These twin developments confronted Army Intelligence for the first time with the problem of disseminating communications intelligence to the field. The previous American success against the Purple machine, although providing valuable background information to Washington during the course of the war, had not had tactical implications. In contrast, ULTRA could be of immediate operational value. The problem was now to transmit this extremely sensitive combat intelligence from central processing centers in the United States and Great Britain to theater commanders thousands of miles away in a fashion that would avoid any compromise of the source of intelligence.
In response, the Special Branch adopted the existing British system of handling communications intelligence. New security classifications were introduced. At first, high-level communications intelligence was termed ULTRA DEXTER, lower level material designated DEXTER. Later, ULTRA was reserved for the results of high level cryptanalysis, while intelligence derived from breaking simpler systems was termed CIRO PEARL or PEARL, and information produced from radio-direction finding was labeled THUMB. Ultimately, in order to mesh Army and Navy practice, PEARL and THUMB were merged into a single category, PINUP. Meanwhile, in April 1944 the U.S. Army at last adopted a "top secret" classification to provide a satisfactory equivalent to the British "most secret."
Terminology was only one aspect of the new procedures. The actual dissemination of ULTRA in the field was handled by special security officers (SSOs) selected and trained by the MIS Special Branch and operating under its direct command using special cipher systems. This was both faster and more secure than the usual practice of sending intelligence through successive layers of command channels. Again, this followed British practice. The first three American SSOs went out to commands in the Pacific in the fall of 1943. The British agreement that ULTRA supplied to American commanders in the European theater would henceforth be disseminated through American channels led to the procurement of eighty additional special security officers in December 1943.
At first, the SSOs were attached only to the highest level of commands. They had the dual mission of securing the vital ULTRA material and of explaining its significance to commanders and intelligence officers unfamiliar with the uses and limitations of high-grade communications intelligence. However, by July 1944 the decision had been made to disseminate ULTRA directly to field armies and equivalent AAF commands and even down to independently operating Army corps. This necessitated recruiting 172 more SSOs in August 1944. In addition, 65 enlisted men were brought into the system to operate communications, thus relieving SSOs from the necessity of deciphering their own messages. By the end of the war, the elaborate dissemination system was headed in each theater by a senior special security representative.
The Army's own increasing successes in communications intelligence, combined with the new availability of the British COMINT product, helped bring about a general reorganization of the Military Intelligence Service. Because of the growth of its responsibilities, the Special Branch expanded to the point where it constituted the largest component of the MIS Intelligence Group. This created a situation in which much of the analysis performed at the War Department level was undertaken by individuals without access to the single most important intelligence source exploited in the war. Only a "very select few" top production officials had been granted access to the whole picture.
The United States' entry into World War II naturally imposed new demands on the Army's Signal Intelligence Service. Up to this point, the SIS had achieved its main successes against intercepted diplomatic communications provided by its 2d Signal Service Company, which manned seven small fixed sites located at Fort Hancock, New Jersey; Fort Hunt, Virginia; Fort Sam Houston, Texas; the Presidio of San Francisco, California; Corozal, Canal Zone; Fort Shafter, Hawaii; and Fort McKinley, Philippine Islands. At the time of the attack on Pearl Harbor, the Signal Intelligence Service had the strength of 331, almost equally divided between the field sites and its headquarters in the old Munitions Building, a World War I-vintage structure in downtown Washington, D.C.
In 1943 the War Department gave the British primary operational responsibilities for code breaking in Europe, which freed the SSA to concentrate most of its energies on the Japanese military problem.
The year 1944 saw the full maturation of SSA's activities. In January Australian forces captured the codes of the Japanese 20th Division at Sio on New Guinea. The find led to full exploitation of Japanese military communications. In the spring of 1944 the intercept facilities of SSA's 2d Signal Service Battalion were also extended when new fixed stations were established at New Delhi, India; Asmara, Eritrea; Fairbanks and Amchitka, Alaska; and Fort Shafter, Hawaii. In the fall the steady advance of American forces in the Pacific allowed another fixed site to be established on the island of Guam. Additionally, the 2d Signal Service Battalion assumed control over former Office of Strategic Services "listening posts" at Bellmore, New York, and Resada, California, converting them into security monitoring stations.
US Army collection efforts were supplemented by material forwarded to Arlington Hall by MacArthur's multinational center and by British, Canadian, and Indian sources. To process the material, which came by courier pouch and through forty six teletype lines at Arlington Hall, the Signal Security Agency's military and civilian work force of 7,000 was supported by a battery of 400 IBM punch-card machines.
The positions of chief of the Signal Security Agency and commander of the 2d Signal Service Battalion continued to be combined under one colonel Col. Preston W Corderman, who had commanded the agency since 1943, finally received a brigadier general's star in June 1945. throughout most of World War II.
All this was conducted under the tightest secrecy, which would be maintained for thirty years.
By mid1942 the Army had replaced its older M 134s and M134As with the M134C, more commonly known by its short title, SIGABA. Issued down to division level, the SIGABA served as the backbone of the Army's secure high-level communications throughout the war. The handy little M209, designed for use at lower command echelons, was in the hands of American troops before the first American landings in North Africa in November 1942.
By June 1943 the agency had developed a secure telephone apparatus used for transatlantic conversations between Roosevelt and Churchill and later employed to link theater commanders with Washington. The SIGSALY, as it was called, afforded high security although with the device weighing ninety tons, its use was obviously confined to major headquarters. Smaller voice "scrambler" devices offering less security were made available to lower echelons of command.
After Pearl Harbor the Signal Corps had transferred its Air Warning Service radar companies to the control of the Army Air Forces, ultimately turning over all responsibilities in radar development to the air arm. However, the Signal Corps continued to exercise staff supervision over Army radio countermeasures and "radio or radar deception" through a newly formed Protective Security Branch, which was reassigned to the Signal Security Service at Arlington Hall in December 1942.
Only theater commanders could authorize the use of radio or electronic countermeasures, although task force commanders might be delegated this authority
In practice, however, American forces did not attempt communications jamming to any extent in World War II, since it would risk interfering with the vital flow of communications intelligence. The Americans made one significant attempt to disrupt enemy communications circuits during the Ardennes campaign, using electronic jamming equipment mounted in aircraft to interfere with the radio transmissions from German tanks.
The 3103d Signal Service Battalion, activated in December 1943 for deployment in the European Theater of Operations (ETO), could simulate the communications nets of large formations. It played an important role in diverting the Germans from the real locations of American troop concentrations before the invasion of Normandy A smaller deception unit, the 3153d Signal Service Company, went to the Pacific in 1944.
The VENONA project was a long-running secret collaboration of the United States and United Kingdom intelligence agencies involving cryptanalysis of messages sent by intelligence agencies of the Soviet Union, the majority during World War II. There were at least 13 code words for this project that were used by the US and British intelligence agencies (including the NSA); "VENONA" was the last that was used. That code word has no known meaning. (In the decrypted documents issued from the National Security Agency, "VENONA" is written in capitals, but lowercasing is common in modern journalism.) It was not until 1995 that project materials were released by the government. Analysis supported some criminal spy cases, such as that against Julius Rosenberg for some of the charges, but cast doubt on the case against his wife Ethel Rosenberg