AN ELATED GENERAL, A SMILING BUDDHA
IN THE LATE 1960s, Lop Nur, Semipalatinsk, and Novaya Zemlya were not the only nuclear testing sites that commanded the attention of the men and women who selected the targets for the Corona and Gambit reconnaissance satellites.* In contrast to the sites in the Soviet Union and China, others were not located within a vast expanse of territory. Instead, they could be found on two tiny atolls in the South Pacific. And rather than belonging to a sworn adversary, they belonged to an ally, albeit a troublesome one.
FRANCE’S FIRST STEPS toward an atomic bomb can be traced back to a meeting held in the spring of 1939. One participant, Edgar Sengier, was the managing director of Union Minière, which controlled access to the uranium ore in the Belgian Congo. Another participant, along with several professional colleagues, was Frédéric Joliot-Curie, whose work had ultimately led Otto Hahn and Fritz Strassman to discover fission. Joliot-Curie and his fellow scientists proposed developing and then exploding a uranium bomb in the Sahara Desert. Sengier agreed in principle, promising to provide the ore and assist in the bomb’s development. He would soon provide eight tons of uranium oxide.1
In addition to uranium oxide from the Congo, Joliot-Curie and his colleagues wanted heavy water. They believed they had identified two means of producing an unrestrained chain reaction—with U-235, and with natural uranium mixed with heavy water to slow down neutrons and increase the rate of fission—and that the second way was the easiest. In November 1939, Joliot-Curie, who had been recalled to military service and appointed head of the army’s Group 1 of Scientific Research, recommended to minister of armaments Raoul Dautry that France acquire 880 pounds of uranium metal from the United States and Norway’s entire stock of heavy water.2
The second task was entrusted to Lt. Jacques Allier, an officer in the Deuxième Bureau, at that time still the elite French intelligence service. Allier, whose civilian career was in banking, had been dealing with Norwegian affairs since 1923. On February 28, 1940, he boarded a train from Paris to Amsterdam to begin his journey. Upon arrival in Norway he discovered that the Germans had tried to purchase the existing heavy-water supply and order more. But while France could not best the Germans on the battlefield, they did win this battle in the secret war. On March 9, Allier completed negotiations for Norway’s entire heavy-water supply, about 407 pounds, which would be turned over to France as a wartime loan, in exchange for 15 percent of the profits from any technical developments obtained from its use. Exactly a week later the entire supply was in Paris.3
With uranium oxide and heavy water in hand, the French physicists prepared to begin reactor-related experiments. But just as the German invasion of the Soviet Union in 1941 delayed that nation’s effort to develop atomic weapons, so the German invasion of France in June 1940 stopped the “clock of French research” on atomic weapons. After the German invasion the primary concern was keeping the raw materials that France had acquired out of Hitler’s hands. The heavy water went to Britain along with Joliot-Curie and his colleagues Hans Halban and Lew Kowarski. Some of the uranium oxide went to Morocco, where it would remain hidden in a mine for six years—keeping it out of the hands of first the Germans and then the Americans and British. Joliot-Curie went back to Paris, where he would later tell Kurt Diebner that he had no idea what happened to the uranium oxide or heavy water.4
In the fall of 1945, with the war over, another trip by Allier to Norway produced another commitment. The Norwegian foreign minister and Norsk Hydro-Electric agreed to provide France with the first five tons of heavy water produced after the war. The purchase arrangements would be completed in May 1946. That same year, Gen. Bloch Dassault, the brother of airplane builder Marcel Dassault, oversaw the quiet repatriation of the uranium oxide.5
France had not only acted to obtain the raw materials for atomic weapons, but also tried to determine what the Germans knew. At the end of the war, the First Army’s Operational Intelligence Service managed to snatch two or three scientists from the Russians. They also received the help of a “Captain Durand” in assessing recovered documents concerning nuclear physics. One day, an assistant to Col. Leon Simoneau, head of the intelligence service, told his boss that it was “astonishing” how closely Durand resembled Joliot-Curie. Considering that Durand and Joliot-Curie were one and the same, it was not astonishing at all.6
On one of the occasions in the spring of 1945 when Joliot-Curie was not pretending to be anyone other than himself, he met with Gen. Charles de Gaulle, leader of the Free French Forces who would become president of the provisional French government. Also in attendance was Pierre Auger, who, along with Halban, Kowarski, Jules Gueron, and Bertrand Goldschmidt, worked in the Manhattan Project’s Montreal laboratory. The duo insisted that it was time to set up a French atomic energy organization. De Gaulle, who during a trip to Canada had learned from the French physicists there of the atomic bomb effort, accepted their recommendation and told them to “take Dautry with you”—in reference to Raoul Dautry, who had become the minister for reconstruction.7
Their efforts culminated in a October 18, 1945, decree establishing the Commissariat à l’Énergie Atomique (CEA), the French Atomic Energy Commission, reporting directly to the president of the provisional government. No program to build nuclear weapons was established along with the commission. The CEA was independent of the military establishment, and its first objective, the construction of a small reactor, was to be accomplished in the light of day. Joliot-Curie became the high commissioner, and at the beginning of January 1946 his six-member executive committee—Kowarski, Goldschmidt, Gueron, Irène Joliot-Curie, Pierre Auger, and Francis Perrin, who had published the first approximate formula for determining the critical mass of uranium—was appointed. Dautry was appointed administrator-general.8
The CEA’s initial workplan involved five components. In the vicinity of Paris, a nuclear physics research center would be created, and the first two reactors established. An intense uranium-prospecting effort would be required. Offices and laboratories needed to be built. Private corporations and a gunpowder factory at Le Bouchet, twenty-five miles south of Paris, would process minerals and prepare extremely pure materials. The last step would involve the construction of a 100,000-kilowatt reactor, which could produce 5 percent of the electricity that France had consumed in 1938.9
That spring Dautry and Joliot-Curie agreed that “the wide-open and windy plateau” of Christ-de-Saclay would be the future site of CEA’s Saclay Nuclear Research Center. To serve as temporary quarters, they used an old fortress at Fort de Chatillon, on the Paris outskirts. By the end of 1947, the plant at Le Bouchet was completed and began refining the uranium oxide that had been returned from Morocco, which proved to be of much higher quality than the oxide retrieved from Le Havre after the war. The purified material it produced was used to feed France’s first reactor, at Chatillon, EL-1 or ZOE (Zero power, uranium Oxide, Eau lourde [heavy water]), which went critical in December 1948. ZOE would be employed for research and training and produce minute amounts of plutonium and radioisotopes.10
Before the end of 1948 the prospecting effort was underway and producing results. There was an exploratory mission to Madagascar, but more important and more promising were prospecting discoveries within France. Uranium was discovered at La Crouzille, in the Limousin, as well as in the Autun region. As a result, France would not have to rely on foreign sources of uranium.11
A second plant was built at Le Bouchet during the spring and summer of 1949, this one a laboratory-scale extraction facility. The uranium oxide fuel rods of the ZOE reactor were processed there to extract small quantities of plutonium. The first milligram, in the form of a purified salt, was produced in November 1949. By the end of 1951 one hundred milligrams had been extracted, which was sufficient for research purposes.12
That year had proved significant in France’s progress toward an atomic weapons capability. Early in the year Robert Spence, a British colleague of Bertrand Goldschmidt, told him that Glenn Seaborg’s proposition—there was no “best solvent for [extracting] plutonium, the best is the one that one knows how to use best”—had been invalidated. Spence, however, would not reveal what solvent the Americans had discovered was unequivocally superior. But he had said enough. There was a secret waiting to be uncovered; all that was necessary was to know it existed. Goldschmidt assigned two of his assistants to compile a complete bibliography of all recent U.S. publications concerning solvent extraction in mineral chemistry. A few weeks later they informed him that tributyl phosphate (TBP) had been successfully employed for the difficult separation of adjoining elements. Shortly afterward, one of Spence’s colleagues happened to visit. During their drive to Le Bouchet, Goldschmidt casually asked him if the British were working on TBP as well. Goldschmidt recalled that “he could not keep from replying: ‘Oh. You know!’ We thus ‘discovered’ the exceptional properties of this solvent.”13
In April 1951 Francis Perrin replaced Joliot-Curie as high commissioner of the CEA. Joliot-Curie, a Communist, had become increasingly hostile to the notion of France developing atomic weapons, fearful that they might be employed some day against the Soviet Union. In August, Félix Gaillard became the secretary of state for atomic energy. Over the next few months he appointed Pierre Guillaumat to become administrator-general of the CEA, and requested preparation of a five-year plan. Completed by the end of 1951, and approved by the National Assembly the following July, it called for the construction of two plutonium production reactors and a plutonium extraction facility at Marcoule on the Rhone River. The plan was silent about the real purpose of the facility, the production of atomic weapons, but an infrastructure would be required when the time arrived.14
Also during 1951 the army’s chief of staff created the Committee on Special Armaments (Commandement des Armes Spéciales). Among the first studies were those exploring the use of atomic weapons in combat. But while such theoretical studies increased military interest, it was a real-world military defeat that helped foster the notion that France should possess the bomb. In May 1954 French troops were defeated by the Viet Minh at Dien Bien Phu, signaling the end of French colonialism in Indochina and dealing a severe blow to French prestige. Development of an atomic bomb was seen as a means of restoring France’s status, and ensuring a greater voice among the Western allies. That the United States considered using atomic weapons in support of the French effort, but decided against such a dramatic action may have further spurred on the French.15
Before the end of the year, on December 26, Prime Minister Pierre Mendès-France, who had presided over the Dien Bien Phu debacle, met with his cabinet and other officials to discuss the possibility of joining the nuclear club. They decided that the idea should be studied, and Gen. Albert Buchalet was appointed to head the Bureau of General Studies, subsequently renamed the Department of New Techniques (in 1956) and then the Military Applications Directorate (in 1958), within the CEA.16
In May 1955 the first funds for bomb studies were covertly transferred from the Ministry of the Armed Forces to the CEA, and Buchalet’s bureau began work. The CEA’s nuclear development plan called for an extension of the basic nuclear infrastructure, including a graphite-moderated reactor and a chemical-processing plant to produce plutonium. And just as, two years later, the Taiwan Straits crisis of 1958 would reinforce Mao’s desire for nuclear weapons, so the Suez Crisis of 1956 did for France. The United States pressured its allies to withdraw their forces, while the Soviet Union threatened nuclear attack if they failed to do so. The event, according to one observer, “demonstrated to the French military . . . that strategic dependence on the United States might prove worse than futile.”17
ON JANUARY 7, 1956, G-1, the plutonium production reactor at Marcoule, went critical and in September reached full power. By that time, Col. Charles Ailleret, the most outspoken advocate of nuclear weapons in the armed forces, had been elevated to the rank of general and placed in charge of the Committee on Special Armaments. A November 30 memorandum instructed the CEA to conduct preliminary studies for a test explosion, prepare the scientific aspects of the test, provide the required plutonium, and build a facility for enriching uranium. Ailleret’s group was assigned responsibility for preparing for the test, including selection of the site. The next month, the Committee on Military Applications of Atomic Energy, which included military personnel as well as the most senior officials from the CEA, was established.18
The search for a suitable site began in 1957. There would be only two serious candidates—the Sahara Desert in the French colony of Algeria, and the Tuamotu islands in Polynesia—although a number of additional islands, including the Kerguelen Islands, Réunion, and New Caledonia, were the subject of preliminary studies. In July the Sahara was selected. Polynesia was rejected, at least in the short term, because of the distance from France and lack of an airport. Ailleret wrote at the time that the rejection should last “as long as we are not setting off megaton-size thermonuclear devices.” He also noted that “from a technical point of view, and without any evaluation or prediction regarding future political developments, it appears that only the [Sahara] is suited for the construction of an atomic testing complex.” The Sahara was still a provisional choice, apparently because the problem of establishing a water supply had not yet been solved.19
On April 11, 1958, Prime Minister Félix Gaillard signed a directive ordering the construction of a device, to be tested in the first quarter of 1960. By July the supreme authority over France’s nuclear program passed to Charles de Gaulle. The Fourth Republic had been a casualty of the Algerian war, as the French military tried to maintain control of the colonial outpost and the Algerian National Liberation Front (FLN) sought to drive them out. De Gaulle was invited to step in and assume leadership of a Fifth Republic. On July 22 de Gaulle confirmed Gaillard’s April 11 decision.20
Throughout 1958 and 1959, in France and in Africa, the French effort progressed. The CEA’s industrial directorate, which had built the Marcoule reactor, supervised the production of plutonium at the site, the separation plant having become operational in 1958. The Military Applications Division was responsible for weapons design and building the experimental device.21
Concerns about the water supply notwithstanding, the moratorium on atmospheric testing that began in late 1958, and the potential internationalization of the Algerian war, did not prevent the construction of the test center in the Sahara. Several potential underground sites in France were investigated, as were possible sites in the Pacific because of fears that international agreements or events in Algeria would preclude testing there. Ground zero was established about forty-five miles south of the oasis at Reggane, with a base camp and a landing strip about seven miles to the east, at the edge of the Tidikelt plateau.22
At 7:00 a.m. Paris time on February 13, 1960, the plutonium device, resting on a 344-foot tower, was detonated on Aillert’s orders in the remote Saharan site, marking France’s entrance into the nuclear club. Code-named Gerboise Bleue, the test produced a blast of between 60 and 70 kilotons, making it the most auspicious entry into the nuclear club. The blast “lighted the desert and paled the full moon still visible in the morning sky.” President de Gaulle expressed his elation with a “Hurrah for France!” reaction to the news. He went on to note that “since this morning [France] is stronger and prouder.” Other nations were less thrilled. Not surprisingly, the Soviet Union was not pleased. The U.S. reaction was one of resignation and regret. Several African nations strenuously objected to being exposed to the fallout from the French test. Ghana froze the assets of all French concerns “until such time as the effects on the population of Ghana . . . become known.”23*
Before the year was over, France conducted three more tower tests in the Sahara, although of much lower yields, with only the Gerboise Blanche test of April 1 producing a blast close to 20 kilotons. The last of the three tests, Gerboise Verte, occurred on April 25, 1961, and produced a yield of less than 1 kiloton. The device was detonated “hastily and prematurely” because three days earlier, Gen. Maurice Challe, the former commander-in-chief of French forces in Algeria, initiated the “The Revolt of the Generals,” in opposition to de Gaulle’s plan to disengage from Algeria. Detonating the device ahead of schedule ensured that it would not fall into the hands of Challe and his associates.24
U.S. CONCERN over French involvement with atomic energy had begun long before France had a nuclear weapons program. Operation Harborage had been intended to keep Werner Heisenberg, his colleagues, and their documents out of French as well as Soviet hands. Joliot-Curie inspired suspicion because of his Communist politics. Leslie Groves believed he first collaborated with the Germans, only joining the resistance when it became clear that the Third Reich was doomed.25
Suspicion extended beyond Joliot-Curie. A February 1946 memo from Selby Skinner of the Strategic Services Unit to Groves’s representative, W. R. Shuler, noted that “a reliable source” conveyed a rumor that “French scientists have the formula and techniques concerning atomic explosives, and that they are now willing to sell this information.” They were not targeting Allies or their own government, the memo reported, but were supposed to be interested in “selling the discovery to one of the smaller nations.” The six scientists named included Joliot-Curie, Madame Curie, and four professors identified as Bertrand, Pascal, Coupard, and Cappard.26
Several additional, more conventional reports or studies about French atomic activities were produced before the year was out. In their February 28 overview of foreign nuclear capabilities, W. R. Shuler and David Gattiker devoted one section to France, focusing on Joliot-Curie and mineral resources. The professor was attempting to form a European scientific bloc to “counteract the Anglo-American bloc.” One of those approached was an unsympathetic Paul Scherrer. The second section noted Joliot-Curie’s claim that France had two hundred tons of uranium which had not been discovered by the Germans.27
In August one of Groves’s representatives in London transmitted a study titled “Atomic Energy Research in France,” which ran a bit over six single-spaced pages and was classified Top Secret. The author addressed the CEA and its functions, other research organizations, personalities, raw materials, political considerations, and capabilities. It noted the plan to locate the CEA headquarters in Fort de Chatillon, and identified the facilities being employed at the time. The section on other organizations covered nineteen institutions, from the National Center for Scientific Research to the Cancer Institute. Fifteen different individuals, mostly scientists, were briefly discussed, including the usual suspects: the Joliot-Curies, Kowarski, Goldschmidt, and Gueron. The report also identified the principal bottleneck for the CEA as being “the lack of large stocks of fissionable material with which to carry on large-scale experimentation.” France, it continued, did not have access to uranium supplies, a comment perhaps made in the absence of knowledge about the uranium ore hidden in Morocco.28
Henry Lowenhaupt made a contribution to the analysis of the French program in November when he reported on the significant elements in a speech Joliot-Curie gave to the CEA. The Top Secret designation of his report may indicate that the speech was obtained from clandestine sources or the reports about atomic energy developments were politically sensitive, or both. One item of undoubted interest was Joliot-Curie’s statement that France had available a uranium supply that had been built up before June, and enough uranium “throughout the empire” to conduct the first stage of their work. He also suggested that through close French cooperation with Britain, “England could be aided in getting ‘out of the grasp of the United States.’”29
Various branches of the U.S. intelligence establishment continued to keep tabs on who was who in the French program during the early 1950s (and certainly beyond). In April 1951 the Division of Biographic Information, in what was then the State Department’s Office of Intelligence and Research, sent the department’s intelligence chief a short, secret report on new CEA chairman Francis Perrin. The report focused on his background (professor of atomic and molecular physics at the Collège de France since 1946), his politics (leftist), and his “reputation for honesty, reliability, and sound judgment among U.S. representatives in Berlin.”30
In December 1952 Garrison B. Coverdale of Army Intelligence sent a memo to the CIA and to the State Department’s special assistant for atomic energy, apparently based on reporting from the army attaché in Paris and focusing on CEA personnel developments. Jacques de Courlon, who had been invited to join the CEA, was described by one of the attaché’s sources as “an excellent geophysicist, but politically leftist.” There was also discussion of what role a particular individual would play in the commission’s work, and on the unannounced replacement of the commission’s secretary-general.31
What other intelligence was collected about French atomic activities and personnel during the first part of the 1950s, and what conclusions were drawn from it, is not evident. However, in the late 1950s and early 1960s, as France advanced toward membership in the nuclear fraternity, a variety of reports and studies were completed. An article in the May 29, 1957, issue of the Central Intelligence Bulletin noted that the French government that replaced defeated Prime Minister Guy Mollet’s might be unable to renew his assurances that France would forgo such a program if an early disarmament agreement was reached. “Pressures,” it stated, “seem to be mounting in France in favor of a national nuclear weapons program.”32
The following month, the intelligence community estimate titled Nuclear Weapons Production in Fourth Countries: Likelihood and Consequences reported the belief of community analysts that France could produce its first nuclear weapons in 1958. The authors explained that plutonium “in weapons quantities is now beginning to become available” and planned production facilities would be developed to the point that weapons with yields in the range of 20 to 40 kilotons could be produced at an annual rate of 3 in 1958, 50 in 1962, and 110 in 1967. The estimate also reported that France was “on the verge of deciding a nuclear weapons production program” but that there were two impediments—opposition from much of the public and the concern of some French officials that French production of atomic bombs could lead other nations, including West Germany, to build their own bombs.33
A little over a year later, another intelligence community assessment of the likelihood and impact of new nations joining the nuclear club informed its readers, “We believe (although we have no reliable evidence to demonstrate) that France has probably conducted a fairly significant amount of theoretical weapons research during the past few years.” In addition, the estimators concluded that it would take France another five years to develop an independent capacity to produce significant quantities of U-235, so that until then any independent French capability would be based on plutonium. The American analysts also believed that France could produce and test its first fission weapon of 20 to 40 kilotons by late 1958 or early 1959, noting that the French were establishing a test site in the Sahara.34
Then, in September 1958, the CIA’s Current Intelligence Weekly Summary reported that France “may have the capability of exploding a nuclear device . . . at any time, and intensive preparations for a test have been reported recently”—although such preparations would not begin in earnest until 1959. It surmised that de Gaulle’s government might want to conduct a test prior to the beginning of the testing moratorium that was scheduled to begin on the last day of October, overestimating the general’s concern for world public opinion. The weekly summary explained that a French nuclear test had not previously been expected to occur before 1959, although France had enough plutonium. What was not expected to be available before 1959 were the instruments used to obtain scientific data from a test. However, the CIA reported, it appeared that a German scientist who headed the joint French-German Research Institute in Alsace was “working feverishly” in midsummer on instrumentation for an impending test and had made several trips to the Reggane test site.35
But whatever the German scientist was doing, he was not preparing for a test in 1958. When Dwight Eisenhower visited France in September 1959, the nuclear club still only had three members, but the American president told de Gaulle of his concern that France had decided to develop an atomic arsenal. About two months later, the November 13, 1959, report by the CIA’s Office of Scientific Intelligence, The French Nuclear Weapons Program, told readers that France would soon be a member and that a U.S-Soviet testing moratorium would not stop it. The study indicated that the CIA had learned a great deal about French progress, some of that knowledge having been acquired from a U-2 overflight of France as well as from commercial airliners with covertly placed cameras. Among the targets were the facilities at Marcoule.36
In addition to examining the politics and history of the program, the paper provided an assessment of key facilities, the production of fissile material, weapons research and development, the test center, and the expected timing and nature of the test. France, it noted, was able to obtain most of its uranium from domestic sources, although small quantities were imported from Madagascar. The study also credited France with a reserve of ten thousand tons and an active exploration program in Africa.37
The paper reviewed French plutonium production and extraction. It summarized the history, capabilities, and performance of the three reactors at Marcoule. It provided details on the beginning of construction of the chemical separation plant, as well as the chemicals used, including tributyl phosphate, in the extraction process. According to OSI, there had been a considerable delay in getting the separation plant into operation, but U.S. intelligence agencies had been unable to determine the nature of the trouble. There were also delays in the production of weapons-grade plutonium, largely due to problems with the separation process and “dirty plutonium” (plutonium unsuitable for weapons purposes due to its having too high a concentration of plutonium-240 or having been contaminated by the separation process). Nevertheless, OSI was able to conclude that kilogram quantities of plutonium did not become available for weapons development purposes until the summer of 1959. However, when the Marcoule center was in full operation, probably in late 1960, it would be able to produce approximately 220 pounds of plutonium per year. By 1965, with completion of a planned nuclear reactor program, that total could rise to 1,210 pounds.38
The scientific intelligence report also reviewed developments in French uranium enrichment capabilities, noting that research had started at Saclay in 1955, and construction on two pilot plants had begun two years later—the first being used to test gaseous diffusion barriers. Ground was broken for a full-scale plant at Pierrelatte, fifteen miles south of Montélimar, in the fall of 1958. According to the study, plans called for the plant to be in partial operation within three years and in full production in four. The initial capability would enrich uranium to only about 3 percent U-235, enough to improve the efficiency of French reactors but not nearly enough to be used for weapons. Plans were being considered, although no decision had been made, to build additional stages, to produce weapons-grade material.39
The discussion of uranium enrichment also provided some evidence of where the CIA was getting its information, and what intelligence about the French program it had been unable to acquire. Information on French research on gaseous diffusion had come from at least two sources: papers presented by the French at the same September 1958 Geneva conference that had been a source of intelligence on Tomsk in Siberia and the debriefing of U.S. scientists who had visited the Saclay pilot plant. But when it came to the gaseous diffusion barriers to be used in the operational plant, the CIA was in the dark.40
The report’s examination of weapons research and development largely focused on the organizations involved. While the Department of New Techniques was officially subordinate to the CEA, it appeared, according to the report, to be a joint CEA–Ministry of National Defense organization. Three organizations were identified as “probably” doing research and development on nuclear weapons under the new techniques department: the Centre d’études de Bruyerele-Chatel, which reportedly conducted theoretical and applied studies of the critical mass required for nuclear explosions and prepared weapons models; the Army’s Armament Research and Manufacturing Directorate, which was reported to have been involved in the research and development of detonators; and the Centre d’études de Vaujours (Research Center at Vaujours), located east of Paris. It also reported that the head of the army directorate’s atomic section, Prof. Paul Chanson, was reputed to be “one of the guiding lights” for atomic bomb construction.41
The study’s final presentation of facts concerned the Reggane test center, which it noted was scheduled for completion in the fall of 1958. It correctly identified Charles Ailleret as the chief of Special Arms as well as the commander of the test center. The actual test area was believed to be almost sixty miles due south of the test center headquarters, “in a remote part of the Sahara.”42
The test itself was predicted to be a 300-foot tower shot with a yield of about 20 kilotons, a significant underestimate of the actual yield of 60 to 70 kilotons. But the analysts were correct in their prediction that the device would probably be a plutonium implosion bomb, an estimate based on their conclusion that the French would not be able to produce sufficient amounts of highly enriched uranium for a bomb until 1965, unless they received it from the United States or the United Kingdom or enlarged the Pierrelatte facility.43
About two weeks before the test, the CIA Current Intelligence Weekly Summary noted that there had been a number of postponements, primarily due to technical difficulties, which included producing the proper plutonium isotope, problems of weapons research and development, and a shortage of trained personnel. It was speculated that the most recent delay might have been the result of France’s wanting to install test measuring equipment, much of it purchased from American firms, to obtain maximum diagnostic data.44
Another issue of the weekly summary, published about two weeks after the test, correctly informed its readers to expect another three tests, involving lighter, smaller devices than the one detonated on February 13. It noted that the second test was scheduled before May and might occur sooner, with the third and fourth shots planned for the fall, and possibly including an underground test. France’s interest in underground testing may have been the result, the summary suggested, of “widespread international condemnation of its first test.”45
THE REACTION, both in Africa and around the world, to the first four French tests made it clear to de Gaulle that atmospheric testing in the Sahara was not a viable long-term option. The premature detonation of April 1961 would be France’s last atmospheric test in Africa, but not its last test. Geological conditions forced subsequent underground tests to be conducted in horizontal galleries rather than deep vertical shafts, and thirteen were conducted between November 7, 1961, and February 16, 1966, at In Ecker, in southern Algeria, as France sought to develop operational weapons for its Mirage IV bombers. Most of the tests were less than 20 kilotons, but the Rubis test of October 20, 1963, exceeded 50 kilotons and the Saphir/Monique test of February 27, 1965, produced a blast equivalent to 110,000 tons of TNT.46
Long before the last test in Algeria, on March 18, 1962, France and Algeria had signed the Evian Agreement. France recognized its former colony’s independence and agreed to turn over control of the Sahara in five years. It had been clear in 1961 to all but the hard-core supporters of a French Algeria that the colony would become independent and would prohibit any French testing, even underground. France would need a new test site, a site that would have to accommodate high-yield thermonuclear tests. In July 1961 Gen. Jean Thiry, who had assumed responsibility for test site selection, listed four possibilities. Three were located in the South Pacific—New Caledonia, the Marquesas, and Réunion—and one in the South Atlantic, the Kerguelen Islands.47
Other sites were considered over the next year, and when the selection commission made its choice on March 22, 1962, it chose to locate the Centre d’Expérimentations du Pacific (Pacific Test Center) at Gambier–South Tuamotu in French Polynesia, which had been the favorite of French leaders from the start. Actual detonation points could be set up on nearby Mururoa, and perhaps at Temoe, Fangataufa, Maria, and Marutea. The choice was made official by the Defense Council on July 4, and the orders were confirmed before the end of the month. Work was to start in early 1964 so that the center would be operational at the end of 1966. The desire to reduce the amount of time to complete a test site meant the sites would be designed for atmospheric testing, since a surface site could be completed a year earlier than one for underground testing.48
Mururoa is a coral atoll, a ring-shaped coral reef enclosing a lagoon that is the visible rim of an extinct underwater volcano. The atoll is about 6 by 18 miles, and the average depth of its lagoon is between approximately 100 and 130 feet—deep enough for large ships. Its coral ring is between 650 and 985 feet wide, except for a gap about 2 ½ miles wide that connects the lagoon to the Pacific Ocean. Mururoa is located about halfway between Australia and South America, about 720 miles from Tahiti, in the extreme southeast corner of the Tuamotu Archipelago. One of five archipelagos making up French Polynesia, Tuamotu consists of about eighty atolls. In May 1963 the first detachment of engineers assumed control of the planned test site, followed in September by the first group of Polynesian workers. By January 1964 five hundred men were working on the atoll.49
Fangataufa, about 25 miles southeast of Mururoa, became the secondary Pacific Ocean test site. Measuring about 3 by 5 miles, it was originally a closed atoll, requiring the French army to blast a 250-foot gap through the coral ring to permit access from the ocean. A third atoll, Hao, located about 280 miles northwest of Mururoa, initially served as a base where the test devices were assembled. Device components arrived on planes, which were able to land on one of the longest runways in the South Pacific, built by the French military. Eventually, the nuclear assembly facility would be transferred to Mururoa.50
By early July 1966 France began atmospheric testing in the Pacific. On July 2, a specially equipped cruiser, the De Grasse, kept a safe distance from Mururoa. On board were Gen. Jean Thiry, then director of the French test program, Adm. Jean Lorrain, commander of the task force for the Pacific tests, and Jean Viard, technical director for the tests. Presumably it was Thiry who ordered the detonation of the pure-plutonium AN 52 warhead on a barge at the test site. The test, code-named Aldébaran, produced a blast of 30 kilotons. Seventeen days later Fangataufa was initiated into the nuclear era. Test Tamouré was an airdrop, with a Mirage IVA aircraft ejecting an AN 11 bomb that exploded with a force equivalent to about 60,000 tons of TNT.51
Another three tests, in September and October, followed before the year concluded. The September 11 test had the supreme French VIP in attendance: President Charles de Gaulle. The device, a prototype of a planned intermediate-range ballistic missile warhead, was carried into the air by a balloon, where it detonated with a yield of 120 kilotons. The next-to-last test of the year, on September 24, was code-named Rigel and represented a step toward a French thermonuclear capability. A boosted fission device containing plutonium and small quantities of thermonuclear material was detonated on a barge at Fangataufa and produced a yield of 150 kilotons.52
De Gaulle might have been most interested, not in the test he attended, but the one on September 24. In 1966 the temperamental general reversed his previously relaxed attitude toward French attainment of a thermonuclear capability. While in 1962 he was willing to wait until 1970 for France to advance beyond fission weapons, in 1966, with China on its way to developing a fusion device, de Gaulle wanted results and he wanted them quickly. The French president told Alain Peyrefritte, who had recently been named minister for research and atomic and space affairs, “I want the first experiment to take place before I leave! Do you hear me? It’s of capital importance. Of the five nuclear powers are we going to be the only one which hasn’t made it to the thermonuclear level? Are we going to let the Chinese get ahead of us?” De Gaulle gave Peyrefritte a new deadline—1968.53
One step toward beating that deadline was taken in April 1967, when the uranium enrichment plant at Pierrelatte became operational. The uranium produced was used in the three tests conducted on Mururoa during June and July, all of which involved research on the use of U-235 as a fissile material, apparently in the fission primary of a thermonuclear weapon.54
But the most significant developments occurred back in France. Work by physicists Pierre Billaud, Luc Dagens, and Michel Carayol, and possibly help from a friend, made 1968 a realistic goal. In January, Billaud, following up on work by Dagens, completed a paper that advanced French weapons designers’ thinking toward the same concept that U.S. and Soviet weapons designers had: a two-stage weapon using the principle of radiation implosion first worked out by Ulam and Teller. His work, “while not solving the entire problem,” unleashed a new round of reflections. In early April, Carayol produced a brief paper that presented, and justified mathematically, his architectural idea for such a two-stage fusion device.55
Then, in late September, according to Billaud, André Thoulouze, the French military attaché in London, arrived in Paris, bringing along information from Sir William Cook, who had been appointed director of the thermonuclear research, development, and testing program at the British Atomic Weapons Research Establishment at Aldermaston in 1954. Cook allegedly advised Thoulouze that in developing a hydrogen bomb, French scientists should not “look for complications” but should “try a simple design.” More specifically, the British scientist reportedly suggested that using X-rays to compress the thermonuclear fuel, rather than more complicated techniques, was their best bet. Cook’s information provided reassurance that Carayol’s design was correct. Billaud notes that “had this outline not already been in existence we would have had a difficult time understanding this information, and may have suspected an attempt at misleading us. In fact, there was a kind of reciprocal validation: Carayol’s sketch authenticated the seriousness of the source, while the latter confirmed the value of Carayol’s ideas.”56*
Those ideas bore fruit well before de Gaulle’s deadline expired, but not before China tested its first hydrogen bomb. Billaud traveled to the South Pacific to watch France’s first thermonuclear test, code-named Canopus, on August 24, 1968, at Fangataufa. The U-235 fission primary was used to ignite the lithium-6 deuteride secondary. The three-ton device was carried into the air by balloon and was detonated at approximately 1,970 feet, producing a blast of 2.6 megatons. De Gaulle proclaimed it a “magnificent scientific, technical and industrial success, achieved for the independence and security of France by an elite of her children.” The contamination was apparently so extensive that the atoll was declared off-limits for the next six years. Two weeks later, on September 8, the second test, Procyon, took place on Mururoa and produced a yield of 1.2 megatons.57
The French testing program took a holiday in 1969, which French officials claimed was due to budgetary limitations, while others suspected that contamination of Mururoa from the September 8 test was the real reason. Between 1970 and September 15, 1974, France conducted twenty-nine tests, all atmospheric, with twenty-eight on Mururoa and one on Fangataufa. The most notable of the eight 1970 tests was Licorne, a 1-megaton test on Mururoa on July 3. Six hours after its conclusion, French officials returned to the island, and minister of defense Michel Debré, who would die in 1996 at the age of eighty-four, swam in the lagoon in an attempt to neutralize criticism of the environmental impact of the tests.58
The thirteen tests conducted in 1971 and the two succeeding years primarily focused on the development of two warheads: the 500-kiloton MR41 boosted-fission warhead that was to be deployed on the M1 and M2 submarine-launched ballistic missiles, and the 1-megaton TN 60 warhead destined to be carried by the M20 submarine-launched ballistic missile. There was a single test apparently designed to test the warhead for a tactical missile. The French finally bowed to international pressure and ended their atmospheric testing program with the eighth test of 1974, on September 15. Those tests involved a small tactical atomic bomb and an airdrop from a Jaguar A aircraft.59
THE FRENCH NUCLEAR TESTS in Algeria and the South Pacific from 1961 to 1974 were of sufficient interest to U.S. officials that the U.S. nuclear intelligence establishment went to significant lengths to monitor them. On August 4, 1961, a group of midlevel officials from the State Department and air force met in the office of Philip J. Farley. Farley, Dean Rusk’s special assistant for atomic energy and outer space matters, was not there and Col. L. B. Williams presided over the meeting. The catalyst for the meeting was a July 26 letter from Air Force Technical Applications Center commander Jermain P. Rodenhauser, who had inquired about the possibility of establishing Atomic Energy Detection System stations in countries close to Algeria, to monitor the underground tests that “a reliable intelligence source” claimed would take place there from October 1961 through 1962.60
The idea of establishing stations in Mali or Niger was soon dismissed. Two State Department officials observed that neither government was likely to give its consent, and even if one did it would be “almost impossible to conceal such activities from the French.” The third nation whose proximity to Algeria made it a subject of discussion was Libya, where the United States had a military presence at Wheelus Air Base. State Department official Richard St. F. Post noted that “certain factors” might make it possible to place detection equipment in Libya.61
There were two possible cover stories for a collection effort. Post pointed out that the 64th Engineer Battalion was engaged in a mapping survey project in the Ghat-Sebha-Ghadames triangle, the precise location to establish a station. In addition, the detection equipment could be brought into Libya without difficulty under the guise of being for the survey team. At the same time, the similarity of the detection equipment to that used in oil exploration created another option, since most of the territory of interest was covered by oil concession land, and most of that land had been rented to American oil companies. Post continued that it might be possible to conduct the operation without informing either the Libyan government or the nation’s chief of state, King Idris.62
In a letter dated August 15, Howard Furnas, the acting special assistant for atomic energy and outer space, informed Rodenhauser of the conclusions reached at the August 4 meeting. Furnas also suggested that Rodenhauser determine whether the possible covers for a detection site were truly feasible. In addition, he noted that the State Department was weighing the foreign policy implications of the proposal and that the U.S. ambassador to Libya had been asked for his views. Furnas also mentioned the damage to the U.S. position in Africa if the existence of such a covert collection operation were revealed.63
It is not clear from the declassified record whether a station was established in Libya during 1961, or after.* It is clear that U.S. leaders and their intelligence support apparatus remained interested in French nuclear progress. In the summers of 1962 and 1963, special national estimates titled French Nuclear Weapons and Delivery Capabilities were completed. In April 1963 the CIA reported indications of technical difficulties at the Pierrelatte gaseous diffusion plant. Later that month or early the next, President Kennedy was made aware of reports from Bonn, Rome, and Brussels alleging that France had requested financial and technical assistance from West Germany to complete the Pierrelatte facility. In a May 7 national security action memorandum, national security adviser McGeorge Bundy informed key officials that Kennedy wanted CIA chief John McCone and Atomic Energy Commission chairman Glenn T. Seaborg to prepare an appraisal of the gaseous diffusion project.64
In addition to responding to the president’s request, CIA analysts also spent part of May completing a study, The French Nuclear Strike Force Program, which concluded that France’s initial operational nuclear arsenal would consist of Mirage IV bombers equipped with 50- to 60-kiloton fission warheads, which would be followed by a fleet of submarines carrying intermediate-range missiles armed with thermonuclear warheads. At the very end of the month, Glenn Seaborg personally followed up on John McCone’s assertion that France was sharing its nuclear weapons know-how with West Germany. On his return from a visit to unclassified Soviet nuclear facilities and laboratories, he stopped in Paris to have lunch with Bertrand Goldschmidt. “His vehement denial of any such activity,” Seaborg recalled, “was completely convincing to me and I had no trouble convincing President Kennedy and the CIA director that such collaboration in nuclear weapons was not occurring.”65
In late July, the U.S. Intelligence Board approved a completed special national intelligence estimate, The French Nuclear Weapons Program. Most of the estimate focused on delivery systems, the cost of the nuclear program to the French economy, and political considerations. Only a few pages of the estimate focused on the size of France’s stockpile or its ability to produce U-235. The intelligence community’s analysts had concluded that France would not be able to produce weapons-grade U-235 until 1967, when they expected the gaseous diffusion plant under construction at Pierrelatte to be completed.66
A CIA study on the French nuclear energy program that was completed in early 1964 examined the availability of uranium, research facilities and their functions, nuclear weapons research institutions, production of plutonium and U-235, and the nine nuclear weapons tests that had taken place through October 20, 1963. It concluded with a brief examination of French plans to establish a new testing area in the South Pacific.67
The desire to closely monitor French progress in producing fissile material and in completing the new test center led the intelligence community to conduct space and aerial overflights of key targets over the following years. Overflights by satellites carrying KH-7 cameras between January 22, 1966, and June 12, 1967, produced ten images of Pierrelatte, four of which were obtained during conditions of good visibility. On eight occasions between December 8, 1966, and June 12, 1967, Marcoule was photographed, with five of those occasions producing good-quality images. In the mid-1960s, at least, the CIA also had a couple of human sources inside the French government who provided intelligence on France’s nuclear program—sources that the agency’s operations directorate was “pretty proud” of, according to a former senior intelligence official.68
Interest in developments at the test center spurred what would become the only operational U-2 missions launched from an aircraft carrier. On several occasions between 1957 and 1967, the navy tried, but failed, to obtain a joint agreement between the CIA and air force to develop such a capability. In mid-1963, Kelly Johnson, the Lockheed airplane designer who had conceived of the U-2, deputy CIA director Marshall Carter, and Brig. Gen. Jack Ledford, head of U-2 operations for the CIA, met at the officers’ quarters at Edwards Air Force Base in California, where the CIA U-2 detachment was located. Their discussion of the advantages of being able to launch a U-2 without securing suitable foreign bases led to Project Whale Tale, the modification of some U-2s to permit carrier operations.69
As a result, in May 1964 a U-2G, a U-2C with the necessary modifications, was aboard the USS Ranger when it set sail for the mid-Pacific, and its ultimate target was Mururoa. The images the U-2 produced, from the only two operational missions the spy plane ever flew off a carrier on May 19 and May 22, were available to the authors of a 1965 CIA study that focused solely on the test center. After providing a brief history of French nuclear testing and the selection of the test center, the The French Pacific Nuclear Test Center reviewed the different installations in the region, including the Mururoa testing area, the Hao support base, and the Tahiti headquarters. The report noted that while Fangataufa had been designated an observation post by General Thiry, the atoll would probably be used as a test site at some point in the program. The CIA analysts also summarized the history of the construction effort, noting that “construction of a test site . . . halfway around the world has placed a tremendous logistics burden on France.” Nevertheless, they concluded that testing would begin on Mururoa in 1966 as scheduled. They also correctly predicted that the testing program would include barge and balloon tests, and pinpointed 1968–1970 as a period during which France would probably conduct tests of a thermonuclear warhead with a megaton yield.70
The Mururoa test center, and the atmospheric tests conducted there through September 1974, were the target of a wide variety of U.S. intelligence collection efforts. Monitoring the test center was done through open sources, including French newspapers and government publications, and more secret sources, including Corona and Gambit satellites. On sixteen occasions between June 30, 1966, and August 20, 1968, Corona satellites (carrying KH-4A or KH-4B cameras) targeted the French South Pacific test site. And on at least two occasions during the KH-7 portion of the Gambit program—on March 13, 1966, during Mission 4016 and on May 26, 1967, during Mission 4037—the United States obtained high-resolution images of the area, images of the site before the first test of each year. Undoubtedly, the KH-8 component of the Gambit program also contributed to the monitoring of the test center from its beginnings in July 1966 through September 1974, a period during which forty KH-8 cameras were shot into orbit.71
Dino Brugioni, a former official at the National Photographic Interpretation Center (NPIC), recalls that the French were “sneaking down there” and not eager to acknowledge their use of U.S. construction practices, but the imagery from the KH-7 missions, as well as other sources of intelligence, allowed the United States to monitor their activities. In addition, he concluded that despite the image of “beautiful Tahiti” associated with French Polynesia, the places where the French were working were “not paradise.”72
Beginning in June 1971 Brugioni and other photointerpreters at NPIC had an additional asset at their disposal: the first of eighteen Hexagon satellites carrying the KH-9 camera was placed in a 114- by 186-mile orbit at an inclination that took it over the entire planet. After the final Corona mission of May 1972, the KH-9 cameras in orbit would provide the intelligence community with the wide-area imagery that had been Corona’s responsibility. The images covered four times more territory (80 by 360 miles in a single frame), were far sharper (with a ground resolution of one to two feet), and were far more numerous, since Hexagon satellites carried four film return capsules in contrast to the two carried by later Corona models.73
While imagery satellites could monitor test site construction, test site preparation, and the destructive impact of a detonation, the United States needed a whole variety of systems to detect and evaluate the French testing program in the South Pacific. To increase the capability of the Dawn Star acoustic network to monitor those tests, U.S. Army Signal Research Unit No. 15, reporting to the Army Security Agency, was established at Brisbane, Australia, in the fall of 1967. That unit, along with a temporary station at Fiji, detected the French tests during the summer of 1968. The Fiji station was deactivated that September.74
Monitoring the 1968 tests was also a responsibility of the Strategic Air Command’s 55th Strategic Reconnaissance Wing. The wing’s effort, designated Burning Light, employed two KC-135R aircraft—modified tankers whose primary sensor measured and recorded the electromagnetic pulse from nuclear detonations while secondary sensors photographed the density and opacity of the resulting nuclear cloud. The data gathered would help the United States predict the effects of low-altitude nuclear detonations—information of particular interest to the Pentagon’s Defense Nuclear Agency, which was the primary customer for Burning Light data. While the plane was flown by crews from the reconnaissance wing, the sensors were installed and operated by personnel from AFTAC.75
The 1968 tests resulted in fifteen Burning Light missions, which began and concluded at Hickam Air Force Base in Hawaii. While the French government did not provide SAC and AFTAC with advance warning of the tests, the eavesdroppers at the National Security Agency did. Communications intercepts allowed NSA to notify SAC of the approximate time of a test. Before that time a KC-135 would begin orbiting in the vicinity of Mururoa or Fangataufa until there was a detonation. (At some point, possibly as early as 1968, the advance notice that NSA was providing SAC may have originated with a far smaller eavesdropping operation—that of New Zealand. In the late 1960s or early 1970s the New Zealand Government Communications Security Bureau, through its NR [Navy Receiver] 1 station at Waiouru, began monitoring communications, including telexes, between France and the Pacific test center. The Joint Intelligence Bureau also used the intercepts to prepare reports on the French program that were provided to the United States and other allies.)76
The French did not test in 1969, so the Burning Light aircraft had a year off, but between May 15 and August 6, 1970, thirteen Burning Light missions collected data on the eight atmospheric tests France conducted in the South Pacific, six at Mururoa and two at Fangataufa. Those missions were only one part of Nice Dog operations, Nice Dog being the overall designation for long-range collection against French nuclear testing (each year’s effort also had its own code name). Another component of the 1970 effort was ship-based monitoring of the tests, conducted by the 455-foot-long USNS Wheeling, a missile-range instrumentation ship, also known by the code name Pock Mark.77
The small French test program for the summer of 1972, three tests, was reflected in the smaller Burning Light effort of nine sorties as part of the Nice Dog/Dial Flower collection effort. The following year’s test series was slightly larger, with five tests between July 21 and August 27, 1973, tests which led to an antibomb rally in Papeete that attracted five thousand people and protests from several Latin American countries. The United States continued monitoring the French tests, an activity designated Hula Hoop that year—which included both Pock Mark and Burning Light missions, the latter performed by two NC-135A aircraft provided by the Air Force Systems Command. On board were personnel from the Defense Nuclear Agency, U.S. Air Force Security Service, AEC, and AFTAC. Also employed to monitor the tests was the USNS Corpus Christi, an aircraft repair ship operated by the Military Sealift Command and code-named Pot Luck. Both ships carried SH-3A helicopters that were used in data collection. A third ship, the USNS Huntsville, also participated in the effort, with drones equipped for nuclear sampling being launched from its deck.78
The briefing given to the crew of the Pock Mark vessel stressed the classified nature of the mission, even though U.S. and French spokesmen had noted the presence of U.S. ships in the vicinity of the testing site from the very first South Pacific test. The crew was also assured that the ship’s positioning was based on both the data to be collected as well as safety factors. At the time of detonation there would be a slight jolt, followed by a wind of up to eighteen knots per hour. “Thermal output will be detected as a slight warming of the skin by topside personnel,” they were told, but “no nuclear radiation will be experienced because of the ship’s upwind position.”79*
Successful monitoring of the tests using ships and aircraft equipped with specialized nuclear intelligence gear depended on more traditional intelligence disciplines—human intelligence, imagery, and communications intelligence to provide warning of the upcoming tests. Burning Light aircraft, based in Hawaii, were not dispatched until notice of an upcoming detonation was received. After flying 2,600 miles and arriving in the vicinity of the test range, they orbited the area before, during, and after the detonation. The planes refueled before entering their orbit, so they had just enough to circle for about two and a half hours and then return to Hawaii. In some cases, accurate intelligence was negated by unforeseen delays. During the 1973 tests, both Burning Light aircraft were launched to cover each test, but the detonation of July 28 was delayed due to technical difficulties. By the time it went off, the two NC-135As were more than 1,500 miles north of the test area, on their way back to Hawaii.80
Burning Light missions also experienced problems in 1974, the final year of French atmospheric testing. The first test, on June 16, went uncovered by the single NC-135A that had been assigned to the mission because intelligence sources failed to warn that a test was imminent. There were also repeated postponements and cancellations of tests in June and July. The problem was further complicated in August by fifteen days of continuously bad weather, not the sort of problem expected of a tropical paradise, which complicated traditional testing arrangements. Rather than being launched on the basis of solid data, NC-135A missions largely depended on “guess work.” As a result of the 1974 experience, the Defense Nuclear Agency and the AEC jointly decided to terminate the Burning Light mission on August 16, about a month before the last French atmospheric test.81
Of course, the United States had assets that did not require advance notice of a planned test but were constantly on watch. In 1965 the first of four hydroacoustic stations was established, and another five were in operation by the end of 1966. The stations were adjuncts to the navy’s Sound Surveillance System (SOSUS) network of hydrophones, ocean-bottom listening devices that could detect the sound emanating from underwater events. Development of the SOSUS network began in 1954, based on Maurice Ewing’s discovery of an underwater sound channel, with the primary objective of detecting and characterizing Soviet submarines. By the late 1960s, networks of hydrophones and associated ground stations received the SOSUS data (via cable) off the coasts of the United States as well as in the Atlantic and Pacific oceans.82
In addition, the Vela satellites circled 67,000 miles above the earth. The two final pairs of advanced Velas had been launched in 1969 and 1970. While their design life had been short, the satellites would exceed expectations and one, 6911, was operational well into the 1980s. The Defense Support Program (DSP) satellites also contributed to intelligence. Whereas the first DSP satellite, launched in November 1970, went into an elliptical orbit, subsequent launches in May 1971, March 1972, and June 1973 successfully delivered the spacecraft into their intended geostationary orbits 22,300 miles above points on the equator. In that orbit the speed of their rotation allowed them to to keep the same, substantial portion of the earth under constant surveillance as the earth turned on its axis.83
The primary mission of the satellites was the detection of Soviet missile launches, whether fired for test purposes or with more sinister intent. Their data was transmitted back, in the early years of the program, to the appropriate ground station—either the one at Nurrungar, Australia, or Buckley Air Force Base in Colorado. To gather the data, each of the twenty-three-foot-long, ten-foot-wide, two-thousand-pound DSP spacecraft was equipped with a twelve-foot-long Schmidt infrared telescope containing an array of detectors. In addition to monitoring missile launches, the telescope could sense a number of other infrared events, including the heat from an atomic explosion as well as from the resulting fireball.84
The satellites also carried a variety of additional sensors—bhangmeters, an atmospheric fluorescence detector, an X-ray locator system—to help in its secondary mission of detecting nuclear explosions in the atmosphere or space. The fluorescence detector could register nuclear detonations that took place between 31 and 1,240 miles, recording the optical time history of the nitrogen fluorescence signals produced when X-rays from a nuclear detonation outside the atmosphere (an exoatmospheric detonation) excited air molecules at the top of the atmosphere. The X-ray locator measured the direction and arrival of X-rays from near-earth exoatmospheric detonations. In addition to estimating the location and time of a blast, the locator sensors made it possible to estimate a blast’s yield and yield-to-mass ratio, as well as establish additional characteristics of the explosion.85
During the summer of 1973 there was only one DSP satellite, sent into orbit in March 1972, that could detect a test at Mururoa. Although it was located over the equator at 70° west longitude, over Latin America, its foot-print extended well into the South Pacific. And it successfully detected all five French tests conducted that summer. By the following summer another DSP satellite had joined the operational constellation. Stationed above the Pacific, at 134° west longitude, it ensured that two DSP satellites could provide data on the tests of that summer and the tests during the summer of 1974.86
The last test in the summer of 1974, on September 15, would be not only the final French atmospheric test of 1974 but also the final French atmospheric test. Mururoa would be the site of repeated tests in the coming decades, but those tests would take place underground. Although the DSP’s sensors would be of little use in monitoring those tests, the U.S. intelligence community would continue to gather intelligence through other means, in response to its mandate to monitor nuclear testing across the globe and to permit analysts to estimate the future capabilities of France’s nuclear arsenal.
FRANCE’S CESSATION of atmospheric testing after the detonation of September 1974 represented a victory for governments and activist groups seeking a halt to such tests. But just a few months earlier, another test, although conducted underground, meant defeat for those, including the U.S. government, who had hoped that no new nations would join the nuclear club—ever.
In mid-May 1974, two-dozen scientists and engineers, including Dr. Rajagopala Chidambaram, were encamped near the Indian village of Pokhran. Pokhran is in the northwestern Indian state of Rajasthan, bordered on the north and northwest by Pakistan, in a desert area where winter temperatures can reach 82 degrees while summer can see the thermometer hit 115, a far less pleasant environment than Mururoa.87
Chidambaram and his colleagues were hoping that their preparations to make India the sixth nation to test a nuclear device would not be detected by a U.S. spy satellite. While it is doubtful that the names Hexagon and Gambit would have meant anything to the Indian scientists, they certainly knew that the United States had such capabilities. They looked up at the night sky, hoping not to see the light of an American reconnaissance satellite passing overhead.88 The idea of assisting India in obtaining nuclear weapons had been proposed a dozen years earlier, but rejected, and the U.S. policy was to discourage nations—whether allies, neutrals, or enemies—from joining the nuclear club. Life would be easier for Indian leaders if they could present the world with a fait accompli, rather than testing in the face of diplomatic pressure.
Like France, India’s path to an atomic weapons capability was an incremental and prolonged one. Homi Bhabha, the father of the Indian bomb effort, moved in the same circles as Joliot-Curie and other atomic physicists of the pre–World War II era. In 1927 he left India to study engineering at Cambridge University. Eight years later Cambridge awarded him a doctorate, in physics rather than engineering. The focus of his work had been in the area of cosmic rays, a subject that, as two historians of twentieth-century physics explained, became an Indian specialty because “India’s high mountains and deep mines enabled the physicists to do important research with inexpensive instruments.” Bhabha went on to become an important contributor to the meson theory of nuclear forces, which centered on a particle (called the meson, as Bhabha had suggested) first discovered during research on cosmic rays.89
Before returning to India in 1939, Bhabha visited the institutes and laboratories of the top physicists working on the Continent, including Neils Bohr, James Franck, and Enrico Fermi. While in India the Second World War erupted, stranding him in his homeland. With nowhere else to go, he accepted the position of “reader” in theoretical physics at the Indian Institute of Science in Bangalore. In 1941, at the age of thirty-one, he received two honors. He was promoted to professor of cosmic ray research and elected a Fellow of the Royal Society. Three years later he wrote to the Sir Dorabji Tata Trust, requesting that the philanthropy provide funds for a nuclear research institute to help develop indigenous expertise, so that when nuclear energy for power production became feasible “in . . . a couple of decades,” India would not need to rely on foreign assistance.90
Before the end of 1945 the Tata Institute of Fundamental Research opened, with Bhabha as its director. The following year Bhabha took on the additional job of chairman of the newly formed Atomic Energy Research Committee, created to promote nuclear physics education in Indian colleges and universities. Then, in 1948, Prime Minister Jawaharlal Nehru, who had written that “the future belongs to those who produce atomic energy,” submitted legislation to create India’s Atomic Energy Commission (AEC). The Atomic Energy Act established the legal framework under which the Commission would operate. Surpassing the restrictions in British and U.S. legislation, the act imposed a veil of secrecy over atomic energy research and development and established government ownership of uranium, thorium, and all other relevant raw materials. The secrecy, Nehru argued, was necessary to protect the nation’s raw materials and knowledge from colonial exploitation as well as to ensure countries with which India might cooperate that their nuclear secrets would be safe in Indian hands.91 The legislation would be the first portent of the extreme secrecy that would envelop the Indian nuclear weapons effort in the years that followed.
When the dust cleared from the legislative debate, India had its Atomic Energy Commission, which, along with the Atomic Energy Research Committee, reported to the Department of Scientific Research. The three-member commission was established on August 10. Bhabha became its first chairman, while S. S. Bhatnagar, a chemist with a sideline in Urdu poetry, and K. S. Krishnan, the director of the National Physical Laboratory, were named his fellow commissioners.92
The 1950s saw further bureaucratic developments, creation of plans, and attempts to acquire the resources needed for an atomic energy program. A 1951 nuclear cooperation agreement with France was followed the next year with Nehru’s unveiling of a four-year plan to move India toward a nuclear capability; the plan included surveys for atomic materials and extracting thorium from monazite. Bhabha began gathering technical information on reactor theory, design, and related technologies. In 1954 another bureaucratic entity was created, the Department of Atomic Energy, and Bhabha became its secretary. The next year, ground was broken for the first Indian reactor, a research reactor named Aspara, at Trombay, on the Indian west coat, just north of Bombay. In January 1957 Trombay became the headquarters of the Atomic Energy Establishment, which was assigned the mission of directing Indian nuclear research and development. Not surprisingly, Bhabha was named director.93
In 1958, Nehru’s government adopted Bhabha’s plan to employ atomic energy to stimulate economic development, which he had outlined in a November 1954 address to the Conference on the Development of Atomic Energy for Peaceful Purposes. The first stage involved the construction, with Canadian assistance, of natural uranium–fueled reactors to produce power and plutonium. A second set of reactors, fueled by the recycled plutonium and thorium, would then be built. The plutonium-thorium fuel, when fissioned in the second-generation reactors, would produce U-233 as a by-product. The U-233, along with thorium, would then be used in yet another set of reactors, to produce more U-233 than consumed by fission. Owing to India’s extensive supply of thorium, an unlimited supply of thorium—U-235 fuel would be created. At least that was Bhabha’s expectation.94
Turning the multistage plan into reality required India to master nuclear technology, a process started with the construction of the Aspara reactor. Of greater significance was the offer Canada made that year to build the 40-megawatt, heavy water–moderated CIRUS (Canadian-Indian, U.S.) research reactor, which burned natural-uranium fuel. Also of importance for the future Indian atomic weapons program was Canada’s failure to attach significant restrictions on the use of the plutonium produced by CIRUS beyond a promise, contained in a secret annex to the agreement, that the reactor and its product would only be used for peaceful purposes. Then, in 1958, while CIRUS was under construction, Bhabha decided to build a plutonium extraction plant at Trombay. Ground was broken for the plant, which was given the name Phoenix, in April 1961. It would become part of the establishment complex at Trombay, which by 1961 included over one thousand scientists and engineers and their offices, and eventually Phoenix, nuclear reactors, and a uranium enrichment plant.95
Ostensibly, India’s effort was devoted to the peaceful purpose of creating energy. Nehru repeatedly pledged that the Indian nuclear activities were for “peaceful purposes only,” although he was also willing to use Indian nuclear know-how for deterrent purposes, noting in January 1958 that India had the capability to build a bomb “in three or four years if we divert sufficient resources in that direction.” In contrast, in 1958 Bhabha privately told an English friend that he wanted India to develop nuclear weapons. A French colleague, Bertrand Goldschmidt, recalled, “Bhabha always wanted the bomb.”96
That desire was enhanced by China’s march toward the bomb. In 1959, in reaction to concern over China’s atomic weapons program, Bhabha told a parliamentary committee that India could build its own bomb without foreign assistance. But while India would continue to advance toward building a nuclear weapon in the first part of the 1960s, it did so without an explicit decision. In August 1960 Nehru told his parliament’s lower house, the Lok Sabha, that the first Indian nuclear power station would be built at Tarapur, north of Trombay, and that the nation would also move forward with the proposed plutonium extraction facility at Trombay. Then, in 1963 Canada reached agreement in principle to build a natural-uranium power reactor in Rajasthan, the Rajasthan Atomic Power Station Unit 1 or RAPS-I, while the U.S. government approved American construction of two light-water power reactors at Tarapur, to the north of Trombay. India agreed to accept IAEA safeguards on the U.S.-supplied fuel for the Tarapur reactor.97
Early in 1964 Bhabha presented a paper to a conference on “Current Problems of Disarmament and World Security” in India, in which he argued that nuclear weapons represented one means of a smaller country deterring a larger, more powerful country, such as China. Then, in May 1964 Nehru died. His death was followed by Lal Bahadur Shastri’s accession to the position of prime minister and, coincidentally in June, by the first spent fuel from the CIRUS reactor entering the Trombay plutonium-reprocessing facility.98
During the summer and fall, as the Chinese test grew closer, there was increasing pressure for Indian development of a bomb. Less than a week after Dean Rusk’s warning of the upcoming Chinese test, Bhabha began public and private efforts to push Shastri and his government to approve additional work on bomb development. During a visit to London he claimed that India’s scientists could develop and test an atomic bomb within eighteen months if permitted, but said, in an attempt to generate domestic support in favor of a bomb, that he did not expect “such a decision will be taken.” Eight days after the test Bhabha, on All-India Radio, told his audience that “atomic weapons give a State possessing them in adequate numbers a deterrent power against attack from a much stronger State.”99
On November 27 the issue was taken up in parliament when the opposition party, the right-wing Jana Sangh, introduced a resolution calling for India to develop and deploy atomic weapons. The resolution was defeated by a voice vote, as Shastri wished. The prime minister believed that India should press for nuclear disarmament rather than become part of a nuclear arms race, and feared the impact on the economy of diverting resources to build a bomb. But that was not the end of the story. In his speech Shastri reminded the legislators that India was still able to produce a bomb within “two or three years” if necessary. Most importantly, while he reaffirmed India’s commitment to peaceful nuclear activities, he extended the concept beyond the production of nuclear energy to include peaceful nuclear explosives that could be used for tunneling through mountains, canal construction, and deepening and widening ports, a concept that formed the basis for the U.S. Plowshare program.100
Shastri’s revised formulation of peaceful nuclear activities left the door open for Bhabha’s pursuit of nuclear weapons, since the physics of peaceful and military nuclear detonations were identical. In addition, as Bhabha was aware, India’s claim of peaceful intent allowed nuclear cooperation with the United States and Canada to continue, whereas an acknowledged atomic bomb program would have led those countries to terminate their assistance to the Indian nuclear program.101
There is some evidence that in 1965 Bhabha sought to obtain a U.S. Plowshare device or blueprints that would, he believed, reduce the time required for India to build one from eighteen to six months. If he did try, he failed. Had he been successful, he might have lived to see India detonate an atomic device. But on January 24, 1966, Bhabha’s Air India flight slammed into the highest mountain in western Europe, the 15,800-foot-tall Mont Blanc in the French Alps. The plane crashed in almost the same spot where another Indian plane, the “Malabar Princess,” did in November 1950. The 1966 crash killed the fifty-six-year-old Bhabha and all 116 of his fellow passengers.102
His death required Indira Gandhi, who succeeded Shastri as prime minister after a heart attack killed him two weeks before Bhabha’s death, to fill the numerous positions that Bhabha had occupied. Others had to assume the role of advocating Indian development of atomic weapons. Dharma Vira, cabinet secretary and AEC member, became interim commission chairman, and was replaced several months later by another commission member, Vikram Sarabhai, an opponent of nuclear explosions of any kind. Like Bhabha, the new chairman had received his physics training at Cambridge University. Homi Sethna, a chemical engineer who had supervised the construction of the plutonium separation plant at Trombay, replaced Bhabha as director of the Atomic Energy Establishment. Sethna and Raja Ramanna, another leading scientist at Trombay, would lead the effort to develop an atomic device.103
Sarabhai’s opposition to the development of nuclear explosives led him to order a halt to work on the topic. That limited authorized development of an atomic bomb, but did not stop it. The separation facility at Trombay continued to extract plutonium from the fuel rods used in the CIRUS reactor, although at a much slower rate than expected, while the nuclear establishment developed the expertise to transform the plutonium metal into bomb cores.104
During 1968, at a time when the government’s leadership, including Indira Gandhi, was focused on the proposed nonproliferation treaty* rather than development of an Indian nuclear capability, a small group of scientists at the Bhabha Atomic Research Centre (BARC), as the Atomic Energy Establishment had been renamed in January 1967 in homage to their late leader, began a concerted effort to move India into the nuclear club. In late 1967 or early 1968 Ramanna, then the chief of the physics group at BARC, instructed Chidambaram to develop the equation of state for plutonium, which required both theoretical analysis and experiments with shock waves. The thirty-two-year-old scientist, who had received his doctorate in physics from Madras University, had been assigned a task that was fundamental to determining how much high explosive was needed to compress plutonium to a specified density, as well as the explosive yield of a device.105
Chidambaram then began recruiting and supervising physicists and engineers, from BARC and the laboratories of the Defence Research and Development Organization (DRDO), who were asked to design the components of the chemical high-explosive device that would be needed to implode the plutonium core of a bomb. Ramanna and Chidambaram, along with senior experimental physicist and Ramanna deputy P. K. Iyengar, selected contributors to the effort and coordinated their activities, in cooperation with DRDO director B. D. Nag Chaudhuri.106
While Sarabhai eventually became aware of the efforts taken in defiance of his instructions, he did not seek to halt them. As a result, Ramanna, Chidambaram, and their associates could proceed until they were able to build and ready to test a device—provided they received approval from the prime minister, whomever he or she might be when the time arrived.107
The new decade would bring India increasingly close to that time. Construction of a new research reactor began in 1970 at Trombay. While the size and costs associated with the Plutonium Reactor for Neutron Investigation in Multiple Assemblies (given the approximate acronym Purnima, the Indian word for “festival”) were small, it would provide crucial data for the design of India’s first atomic device. The reactor allowed Indian scientists to evaluate the behavior of plutonium-based chain reactions as well as study the system when it just exceeded critical. The data could then be employed to identify critical parameters and determine how to achieve optimum yield. While much of the information obtained from the reactor was already published by the United States and other nations, Indian scientists wanted to be sure they would not become the victims of deception. At the same time, the reactor facilitated Indian deception, for India’s plans for atomic power stations employing fast-breeder reactors using plutonium cores provided cover for the reactor’s utility in bomb development.108
Throughout 1972 India moved toward the day when it would detonate a device. The Purnima reactor began operations in May, allowing Indian physicists to refine their understanding of fast neutrons and fast fission. That same month, Homi Sethna, who had succeeded Sarabhai as AEC chairman in February after Sarabhai’s death in late December 1971, gave Ramanna and his colleagues the green light to begin work on a device. In early September Indira Gandhi visited Trombay for a tour. Afterward, Sethna invited her to his office, where he showed her a model of the device he had designed. Sethna soon asked her, “Should we do it or not?” According to Sethna, the prime minister told him, “Get it ready. I will tell you whether to do it or not.” Toward the end of the year a search for a possible test site began.109
Once the decision was made to proceed with construction of a device, commonly referred to as the Smiling Buddha, work accelerated on each phase of the project: the electrical system in the device, the neutron initiator (code-named Flower), the charges that would implode the plutonium, and the diagnostic equipment and instruments. Ramanna, now the director of BARC, supervised the effort in conjunction with the head of the DRDO, B. D. Nag Chaudauri. Chidambaram and Satinder Kumar Sikka, who would go on to play a leading role in India’s pursuit of a hydrogen bomb, worked together on bomb design while the Terminal Ballistics Research Laboratory in Chandigarh produced the conventional high-explosive system to implode the plutonium.110
In January 1973 Ramanna, after being informed that the plutonium extraction plant at Trombay, which had been shut down in 1970 due to its erratic performance, would not become operational again until late 1973 at the earliest, ordered the Purnima reactor shut down so that its plutonium oxide fuel rods could be melted down to obtain enough plutonium for the test device. In March the DRDO reportedly tested the high-explosive system to determine if it would produce the necessary symmetrical detonation. By the middle of the year Chidambaram had concluded that the shaft for the explosion would have to be 350 feet deep to prevent poisonous fumes from being vented.111
The 61 Engineer Regiment, stationed in Jodhpur in Rajasthan, was assigned responsibility for digging the shaft for the device. The regiment’s commander, under instructions from Ramanna to tell only those with a need to know what the shaft was for, informed his engineers that they would be digging a deep well to supply water to the range. Digging began in October 1973 but ran into a series of problems, including hitting water in January. Despite the cover story, the project had been given the code name Operation Dry Enterprise, which reflected the need for a dry shaft, since water or humidity could seriously damage the device. Digging would resume in February, when the test had been envisioned to take place, and continue into May.112
On May 13 Iyengar, Ramanna’s deputy, and four others began assembling the device, whose components had been transported to the test site by a number of means. A regular Indian Airlines flight flew the trigger, concealed in a thermos to prevent the leak of radioactivity and accompanied by Iyengar and T. S. Murthy, a key member of the team that developed the bomb’s neutron initiator. An army convoy transported the plutonium sphere, placed in a specially designed box, along with measurement equipment, from Trombay to Pokharan, a 560-mile journey through hills, plains, and desert. During the entire three-day journey Chidambaram and P. R. Roy, head of the core fabrication team, ate and slept next to the box and took it with them whenever they left their truck.113
On May 15, with assembly completed, the device was lowered into the shaft, which was then sealed with sand and cement.114 For the next three nights Chidambaram and his colleagues would wonder if their plans would be or had already been uncovered by a U.S. spy satellite, and then be obstructed by American pressure.
CHIDAMBARAM NEED NOT have worried, despite the CIA, NSA, and other intelligence organizations having monitored Indian nuclear activities for decades. Those agencies had relied on a combination of open sources, diplomatic reporting, communications intelligence, and satellite photography to track Indian nuclear developments. The AEC also monitored India’s nuclear progress, given the U.S. role as a nuclear supplier. The intelligence gathered helped inform policymakers throughout the 1950s, 1960s, and early 1970s about India’s nuclear status. A 1958 assessment by the CIA’s Office of Scientific Intelligence provided reassurance that India’s nuclear intentions were peaceful. The section on possible military applications was a single paragraph in length, which noted Nehru’s pledge that his government and future Indian governments would use atomic energy for peaceful purposes. The section closed with the observation that “there is no indication in government or scientific circles of a change from the traditional Indian pattern of passivity and mediation.”115 In later years, U.S. decisionmakers would worry that India’s leaders would decide that neutralizing the threats from China and Pakistan, as well as an Indian desire for international prestige, required a nuclear program that produced bombs as well as energy.
In late June 1961 a State Department message conveyed the JAEIC’s interest in Indian nuclear developments. Sent to American embassies and consulates in North America, Europe, and Asia—including Bonn, Paris, Karachi, New Delhi, and Bombay—the secret cable passed on a series of technical questions about the Indian nuclear program. Its recipients were told that beyond technical matters, they should also report on political and economic considerations that might influence a decision to embark on an expanded program, and the relationship of any expanded program to India’s capability and intentions to develop nuclear weapons.116
Any useful information that had been provided in response to a similar message, sent out early in the previous month, would have been employed in writing the September 1961 national intelligence estimate titled Nuclear Weapons and Delivery Capabilities of Free World Countries Other Than the US and UK. The analysts noted that India “is deliberately improving its over-all capabilities in the nuclear field, possibly in anticipation that a future decision to develop an operational capability may be required.” In addition, India had three nuclear reactors in operation, including one (CIRUS) that could produce enough plutonium for about one or two weapons a year; preliminary construction had started on a plutonium separation plant, which was unlikely to begin operations before 1964–1965; and India was seeking to develop its own sources of uranium. In the event of decision within the “next year or two” to develop nuclear weapons, India would probably have a “modest capability, using aircraft and fission weapons, by 1968–1969,” the estimate concluded. It also judged, however, that India’s leaders would not make such a decision unless they were convinced that no disarmament agreements were possible and China’s foreign policy “was clearly growing more truculent.”117
Those conclusions were affirmed about two years later by a national estimate on the likelihood and consequences of nuclear proliferation, which addressed the likely impact of a Chinese nuclear detonation on India. Such an event would probably not lead India to respond in kind, the CIA analysts believed. However, India was likely to continue its current nuclear efforts to the point where a crash weapons program could be instituted to produce a bomb in a relatively short time.118
As predicted in the September 1961 national estimate, despite plans that the Phoenix separation facility at Trombay would be completed in 1963, it did not begin operations until 1964. At 5:00 p.m. on March 31, in an informal ceremony attended by Homi Sethna and other scientists and engineers involved in building the facility, an inactive uranium fuel rod was pushed into the plant by Homi Bhabha. It did not require secret intelligence efforts to discover those and other facts about the new plant. They were reported in a Department of Atomic Energy press release, and forwarded to the Department of State by the American consul in Bombay.119
Beyond monitoring the growth of the Indian nuclear infrastructure, U.S. intelligence continued to address the issue of whether China’s entrance into the nuclear club would cause India to pursue a nuclear weapons capability. During the time between the 1961 estimates and Phoenix commencing operations, India had fought and lost, in 1962, a limited war with China. Then in October 1964 China joined the nuclear club. A national intelligence estimate on the prospects for nuclear proliferation, completed shortly after the Chinese test, noted a number of factors that would affect India’s decision, including the scope and pace of the Chinese program, Sino-Soviet relations, and guarantees from other nations. However, the analysts concluded that the chances were “better than even” that within a few years India would seek to join the nuclear club. They also noted that India had the basic facilities needed for a modest program, including the Phoenix facility, and estimated that by 1970 the country could have an arsenal of about a dozen 20-kiloton weapons.120
While the estimators were predicting “within a few years,” the CIA’s spies, or at least one of them, was telling the agency that Prime Minister Shastri and other Indian leaders were not yet ready to commit to a nuclear weapons program. On October 22, 1964, the day after the new estimate on proliferation was approved by the U.S. Intelligence Board, a CIA Intelligence Information Cable, based on reporting from the field, was disseminated. Titled “Indian Government Policy on Development of Nuclear Weapons,” it claimed that although India had the ability to produce an atomic bomb quickly, it did not plan to do so “as yet.” The author explained that India’s government was convinced that “the Chicoms [Chinese Communists] will not have an offensive nuclear capability for at least five years.” In the interim, if the situation changed, the Indians were relying on the assurance from President Johnson that the United States would come to the aid of any nation threatened by China.121
A week later the American embassy in New Delhi transmitted its assessment of the situation to secretary of state Rusk, which echoed that of the CIA’s source in India. “Our current estimate,” the diplomats wrote in cableese, “is that in foreseeable future India’s leaders will continue stand on position of responsibility and adherence to no-bomb policy.” It also reported on a conversation between an embassy official and a member of the Indian Ministry of External Affairs, who stated that pressure to build a bomb was growing and that Bhabha was the “leading advocate . . . and . . . was actively campaigning for India to go down the nuclear road.” While the External Affairs official felt that India would not reverse course, he did reveal that the matter was under “active consideration” and that Shastri had authorized Bhabha to produce an estimate of what would be required for India to conduct an underground detonation.122
Not surprisingly, assessment of the Indian nuclear program continued within the CIA and other agencies. In early November, OSI completed a secret study of the Indian nuclear energy program which focused on infrastructure rather than the “will they or won’t they” question. It addressed the functions of the Atomic Energy Establishment, the existence and capabilities of the three reactors in operation (including Aspara and CIRUS), the state of India’s uranium reserves, and the capacity of the heavy-water plant at Nangal. In addition, it reported on plans to build three nuclear power stations in India—at Tarapur, in Rajasthan, and in Madras. The study also noted that a plant for the production of plutonium metal, necessary for weapons manufacture, was underway and could be in operation in the fall of 1965.123
The following month, Harry Rowen, of the Defense Department’s International Security Affairs office, completed a draft titled “The Indian Nuclear Problem,” which focused not on India’s nuclear resources, but what the country was likely to do with them and the consequences. He concluded that India might soon begin a weapons program since “the pressures for a weapon are likely to be irresistible after the Chinese test their next device in the absence of some better alternative.” The consequences of such a program would include there being “one more national state [that] could some day attack the United States.” It would also add to the states that could start nuclear actions “with a fair chance of spreading and involving the United States.”124
Whether India would, sometime in the future, decide to join the nuclear club was a mystery to analysts for years—a question that no one, not even the Indian government itself, could answer. Collection systems such as Corona and Gambit were of no use in trying to unravel such mysteries. They could, however, provide significant intelligence on nuclear developments in the world’s second most populous country.
On April 29, 1965, a Corona satellite carrying a KH-4A camera blasted off from Vandenberg Air Force Base. Whereas the second film capsule ejected went into a higher orbit rather than down to earth, the first capsule, ejected on May 4, was snatched out of midair over the Pacific. In addition to the photographs it contained of targets in the Soviet Union and China, it included a direct overhead image of the Trombay complex. Later that year more detail may have been accumulated when a KH-7–carrying Gambit satellite orbited the earth for five days, beginning on September 30.125
Whatever information NPIC photointerpreters extracted from any KH-7 photos of Trombay, along with the intelligence derived from the Corona images, would have been accessible to analysts in OSI and other portions of the intelligence community who prepared the October 21, 1965, special national intelligence estimate titled India’s Nuclear Weapons Policy. The report reviewed the same basic facts and repeated some of the judgments that had appeared in the October 1964 estimate on proliferation: that India had all the facilities necessary to produce plutonium, could quickly transition from a peaceful to a military nuclear program, and could build an arsenal of a dozen bombs by 1970.126
The estimate also warned its readers that India’s joining the nuclear club within months was not out of the question. The American intelligence analysts believed India probably had enough plutonium on hand to produce a nuclear device, and if its leaders had decided in late 1964 or early 1965 to develop nuclear weapons, a test might occur within months. However, they noted, such an event would require weapons design to be well advanced and establishment of a test site in the immediate future. The CIA, NSA, and other intelligence collection efforts had provided “no evidence that such activities are well advanced.”127
Work on weapons technology and design, which in its early stages is “easy to conceal and difficult to identify,” was probably underway. And intelligence, either secret or open source, did indicate that India had expanded significantly the electronic facilities at its nuclear establishment and may have begun to set up a high-explosive test site—both of which could be, but were not necessarily, related to an imminent test. It was possible, then, that even if India didn’t join the nuclear club in a few months, it could before the end of 1966.128
Looking ahead, the estimators noted that if India did decide to construct an atomic device and test it underground, it might claim that it was exploring the potential of nuclear explosions for peaceful purposes. A decision to build a bomb is one the analysts thought “unlikely that we would immediately learn of,” although they expected the intelligence community would be able to detect “advance indications of the first detonation.”129
The search for advance indications relied in part on what the foreign service officers at the American embassy in New Delhi could uncover. In late March 1966 the embassy received an airgram from the State Department providing what had become a boilerplate description of the Indian program, noting the lack of information on a decision by the Indian government to develop nuclear weapons as well as the capability of the CIRUS reactor to produce the necessary plutonium if it did so. The same message also tasked the embassy to report information in five areas: signs of activity in remote areas that might indicate test site construction, indications of the covert establishment of nuclear research facilities or tightened security at known installations, evidence of continued use of the CIRUS reactor to produce “clean uranium,” Indian procurement or development of small electronic neutron generators and high-quality explosive detonators, and the testing of highly instrumented high-explosive shapes or sections.130
Whatever information the embassy’s political, scientific, and economic officers could come up with was of interest at the highest levels of the U.S. government. On June 9, 1966, a little over five weeks after the State Department’s tasking to the embassy, President Johnson met with his National Security Council. Noting that in the aftermath of the third Chinese test there was an increased urgency in dealing with the Indian nuclear weapons issue, he directed the Defense and State Departments, along with the Arms Control and Disarmament Agency and other agencies, to examine a variety of issues concerning the Indian program. Johnson asked that the study address the extent to which the United States might use its economic leverage to discourage Indian development of a bomb, how far the United States should go in meeting Indian security concerns, how various arms control agreements might have an impact on Indian intentions, and “what price the U.S. should be prepared to pay for such agreements.”131
The report delivered to the president on July 25 identified two key issues. One was the political and prestige factors that might propel India toward a nuclear capability. The other was the more tangible problem of India’s need to be able to deter or neutralize Chinese nuclear blackmail. The report’s authors did not believe a decision was “imminent,” indicating that the New Delhi embassy’s collection efforts produced no smoking gun. Nor did they expect a “go nuclear” decision to be made within the year.132
One means of curbing India’s desire for nuclear weapons, the report suggested, was sharing U.S. intelligence analysis of China’s programs, which might “without falsely discounting ChiCom progress, make clear difficulties and limitations still confronting the ChiCom nuclear weapons program and aid in keeping the potential ChiCom nuclear threat in strategic perspective as far as India’s interests are concerned.” At the same time, in order to obtain as much warning as possible about a shift in “India’s present no-bomb policy,” an “increased priority should be assigned to the collection and analysis of relevant intelligence data.” Johnson approved the recommendations on August 1.133
The gathering of information about Indian nuclear activities through the remainder of the decade illustrated the varied sources employed. Corona missions in November 1966, May 1967, and November 1967 produced imagery of the Trombay complex. There was also KH-7 imagery, of unknown quality, from a February 1966 mission. The assorted papers presented by senior Indian scientists were also undoubtedly gathered up by the CIA and other agencies, such as the paper Homi Sethna coauthored on the fuel reprocessing plant at Trombay, prepared for the November 1967 New York conference titled Recent Advances in Reprocessing Irradiated Fuels.134
The American embassy continued to dig out data on topics such as uranium exploration in India. While some information was available in annual reports and other authorized publications from the Department of Atomic Energy, detailed technical data was withheld for reasons of security. As the embassy noted in a May 1968 cable to Washington, current estimates of Indian uranium reserves were not available, and “the overall scope of field exploration and development programs may only be assessed through interpretation from a variety of disconnected chemical analysis reported to be indicative of ores contained in certain areas under exploration, and from personal observation and verbal inquiries by Embassy staff personnel.”135
U.S. foreign service personnel in India also kept an eye on the new power plants that were under construction in the late 1960s and early 1970s, including the ones forty miles south of Madras and in Rajasthan. In 1971 the consulate in Madras provided a five-page report, including a map courtesy of the Indian atomic energy department, which focused on progress of the Madras Atomic Power Project, slated for completion in 1973—a facility not subject to safeguards that could be used to produce plutonium. The 1971 report updated earlier State Department findings from 1968 and 1969. It was followed by an embassy report in August 1972, which stated that the Rajasthan station had become operational earlier in the month.136
Monitoring developments concerning the Madras and Rajasthan projects was not restricted to America’s diplomats. NSA’s eavesdroppers also made a contribution, in August 1972. An NSA report, classified “Top Secret Umbra,” indicating it was based on sensitive communications intelligence, reported on French financing for the Madras project, specifically that “the Banque National de Paris, on 22 June reported that it had been asked to set up a credit agreement . . . to finance an Indian atomic energy project.” The same report also informed its readers that Sweden was supplying material for the project, and that a French company, in Orsay, had informed Indian nuclear officials that the Swedish material “would be shipped within a short time.”137
It did not take such secret intelligence to keep the Indian nuclear weapons problem in front of key decisionmakers such as President Richard Nixon, or his national security adviser, Henry Kissinger. Kissinger had been cautioning Gandhi against a test since 1970, when press reports suggested, prematurely, that India was considering conducting a nuclear test. The State Department, then under the command of William Rogers, informed India that employment of the plutonium from the CIRUS reactor for a test would be considered a violation of India’s pledge of peaceful uses of the heavy water that had been provided by the United States.138
On May 18, 1972, Kissinger, in Nixon’s name, commissioned another study of Indian nuclear developments. The resulting study, by an NSC interdepartmental group, again noted that India’s nuclear energy program afforded the country the ability to conduct a test on short notice and “of mounting a rudimentary weapons program on short notice.” The group also wrote, six days before Indira Gandhi’s visit to Trombay, that “there is no firm intelligence that Mrs. Gandhi has given a political go-ahead for detonating an underground nuclear device (which the Indians would undoubtedly label a peaceful nuclear device).” It further reported that “our intelligence assessment is that over the next several years the chances are about even that India will detonate a nuclear device.”139
EVEN HAD CHIDAMBARAM not been concerned about America’s spy satellites, he probably would not have gotten much sleep on the night of May 17. None of the other scientists there were able to get more than an hour’s worth, undoubtedly due to a combination of anxiety and oppressive heat. By 8:00 a.m., when the detonation was scheduled, Ramanna, Sethna, Nag Chaudhuri, and Iyengar were in place to observe the test, about three miles from the shaft.140
A slight delay was caused when a jeep carrying a member of the ballistics research laboratory broke down near the shaft, forcing him to walk over a mile rather than wait for the vehicle to repaired. At five minutes past eight, Pranab Revati Dastidar, BARC’s electronics expert, pushed the red button. The lack of an immediate reaction led Ramanna and Sethna to fear the device was not going to detonate, but then they saw a small mountain of sand rise from the ground before collapsing. Iyengar recalls thinking, “Now I believe all those mythological stories about Lord Krishna lifting a hill.”* Legend has it that Sethna called the prime minister’s office, spoke to P. N. Dhar, her principal secretary and one of the few individuals other than those working on the project who knew of its existence, and told him that “the Buddha is smiling.”141
Less than an hour later, everyone listening to All-India Radio was in on the secret. India’s version of the BBC interrupted its programming at 9:00 a.m. for a special announcement: “At 8:05 a.m. this morning, India successfully conducted an underground nuclear explosion for peaceful purposes at a carefully chosen site in western India.” Dhar had informed the U.S. ambassador to India, Daniel Patrick Moynihan, a half hour after the test.142
Later in the day, Gandhi publicly congratulated the scientists at a news conference with Sethna. “It is a significant achievement for them and the whole country. . . . We are proud of them. They have worked hard and done a good, clean job.” At the same time, she told newsmen, “There’s nothing to get excited about . . . we are firmly committed to only peaceful uses of atomic energy.” Ramanna reported that the blast created a 650-foot crater and an artificial hill, and called the test “a spectacular sight.” A Western diplomat characterized the prime minister’s pronouncement as “gobbledygook,” explaining that “it is gobbledygook for them to claim that now that they have the ability they won’t use it to make weapons.”143
THAT MOYNIHAN—and apparently, every other U.S. official—was surprised by the test meant that despite the prolonged concern about India “going nuclear” and the efforts of the CIA, National Reconnaissance Office, NSA, State Department, and other intelligence collection and analysis agencies, the intelligence community had failed to provide advance warning. In a Top Secret report, the CIA characterized the test as “a well kept secret” that “took the world by surprise.” The Indian drilling activities in Pokhran were apparently interpreted as involving the search for water, the Indian cover story, or oil.144
The failure was not an epic one though, and may have had more to do with the extraordinary secrecy surrounding the Indian program and the dual-use nature of the facilities, than with the inadequacy of America’s intelligence establishment. While thousands were involved in India’s nuclear program, only fifty to seventy-five scientists were actually part of the effort to design and build an explosive device. Work on the plutonium core was done by the scientists alone. One of them, C. Ganguly, recalled, “We had to do it ourselves, there were no technicians or helpers.” Knowledge outside the group of scientists was so tightly restricted that only three additional individuals, including Gandhi and her principal secretary, knew what was being planned. The minister of defense was informed only eight days before the test; the minister of external affairs, only two days ahead of time. The rest of the cabinet was kept in the dark until after the detonation. In the years after the test, Raja Ramanna was asked by several people about India’s ability to restrict prior knowledge of the test to so few individuals and the strict silence of those in the know prior to the detonation. In his autobiography he suggested that “it was the magnitude of the operation and the enormity of its implications that led us all to honour the oath of secrecy so diligently.”145
While the CIA and other agencies may have failed to provide advance warning, they were able to assess the test based on intelligence gathered from open sources, imagery, human sources, and various components of the Atomic Energy Detection System. A Top Secret Codeword article on the test that appeared in the May 20 issue of the Central Intelligence Bulletin reported the claim of Indian AEC chairman Sethna that the detonation occurred at a depth of slightly more than 325 feet and was completely contained—with no venting of radioactive substances. It also relayed Sethna’s claims that the device was entirely developed by India and relied on implosion. He was noncommittal, the article noted, as to whether India would conduct further tests.146
Beyond reporting Indian claims concerning aspects of the blast, analysts tried to evaluate some of those assertions concerning yield, the absence of venting, the geological conditions of the test site, and other aspects of the test. One of those analysts was Milo D. Nordyke of Lawrence Livermore National Laboratory, who produced two papers on the Indian test. In a May 29 paper he started by accepting the Indian claims concerning the depth of the test as well as a yield of 10 to 15 kilotons. If such numbers were accurate, then the apparent size of the crater created and the absence of venting indicated the device had been detonated in hard, dry rock—the type of rock that a peaceful nuclear explosion might be used to excavate. He concluded that “all the known facts appear to support the Indian statements that their nuclear test was carried out to further their PNE (peaceful nuclear explosions) program.” In September, Nordyke received information indicating that rather than harder rock, the explosion took place in shale and at a depth of 357 feet. Given the size of the crater, he concluded that the yield of the Indian test was probably around 10 kilotons.147
In June the State Department’s Bureau of Intelligence and Research focused on nontechnical issues related to the Indian detonation. It reported that the “euphoria that characterized the response of the Indian public” to the test “has been overtaken by uncertainty about the relationship between nuclear explosions and development needs and about the durability of India’s proclaimed status as a non-weapons nuclear state.” Much of the analysis focused on press and political reaction to the test and the question of whether India should develop nuclear weapons, the costs of future testing, the impact on the nonproliferation treaty, and the impact of the public on future decisions.148 The post-test political analysis in many ways echoed the pretest analyses noting the differing views and competing pressures over India’s future use of atomic energy—illustrating that the test, unlike the first tests in other nations, had far from settled the issue of whether India would become a nuclear weapons state.
SIX DAYS AFTER the Indian test, Lt. Gen. Daniel O. Graham, deputy to the director of central intelligence, William Colby, sent a memo to the directors of several intelligence agencies (NSA, Defense Intelligence Agency, and INR); the heads of the CIA directorates for intelligence, operations, and science and technology; and the chairman of the JAEIC. He informed them that he had requested the Intelligence Community Staff to assess the community’s performance with respect to the Indian nuclear test.149
The postmortem that Graham ordered for Colby was completed in July and ran fifteen pages in length. The executive summary, the only portion of the postmortem that has been declassified, begins by stating the obvious: “In the months prior to India’s 18 May nuclear test, the intelligence community failed to warn US decision makers that such a test was being planned. This failure denied the US Government the option of considering diplomatic or other initiatives to try to prevent this significant step in nuclear proliferation.”150
The summary then noted that the intelligence community had “estimated as far back as 1965 that India would ‘in the next few years’ detonate a nuclear device” and that “its inability to predict the actual event was due essentially to two factors,” one of which was the “inadequate priority against an admittedly difficult target”—not the first or last time that an intelligence failure would be attributed, at least partially, to a lack of attention. Fixing the problem would require “a more focused and dedicated effort by existing collection assets, chiefly in the HUMINT [human intelligence] area,” Graham informed Colby. There was also, according to the authors of the postmortem, a lack of adequate communications among those elements of the community, both collectors and producers, whose combined talents were essential to resolving the problem.”151
By the next January, there had been a number of intelligence community responses to the postmortem’s recommendations, which included a COMIREX review of why intelligence analysts had failed to ask the National Photographic Interpretation Center at the CIA to analyze images they had specifically requested be obtained by America’s spy satellites and the assignment of technical specialists abroad to support the CIA case officers involved in collecting proliferation intelligence. Intelligence reporting on proliferation subjects had increased tenfold and the AEC had established a “proliferation watch” effort to develop better indicators of proliferation activity. In addition, the Human Sources Committee of the U.S. Intelligence Board had established an ad hoc group to work with the JAEIC.152
___________
* Until 1967 selecting targets was the responsibility of the Committee on Overhead Reconnaissance. In July 1967, the Committee for Imagery Requirements and Exploitation (COMIREX) was established by director of central intelligence Richard Helms, and assigned two missions: making the targeting decisions for imagery overflights and deciding how the task of exploiting the imagery obtained would be divided among the National Photographic Interpretation Center, Defense Intelligence Agency, and military service interpreters at organizations such as the Air Force Foreign Technology Division and the Naval Scientific and Technical Intelligence Center. See Jeffrey T. Richelson, America’s Secret Eyes in Space: The U.S. Keyhole Spy Satellite Program (New York: Harper & Row, 1990), pp. 252–256.
* The story merited front-page coverage in the New York Times as well as the Washington Post and Times-Herald. But while the story received column-one treatment in the Times, the biggest headline in the Post concerned the snowstorm that had blown into the Washington area and threatened to deposit eight inches before it departed.
* Billaud believes that the British government approved Cook’s disclosure. However, it was treated as a very sensitive piece of information, so much so that some of the young engineers involved in the French program were not told why they were being instructed to work on one particular design, and not another that they believed superior. Some in Britain and France reacted with skepticism to the claim that Cook provided technical advice. See “Did UK Scientist Give France Vital Clues About H-bomb?” Nature, December 5, 1996, p. 392.
* The U.S. ability to detect French tests in Algeria in 1965 is illustrated by a Joint Atomic Energy Intelligence Committee statement of March 3, 1965, noting the detection of a French test on February 27 and giving a yield estimate of 125 kilotons. See ACDA to U.S. Mission Geneva, Amembassy Moscow, Amembassy Paris, “Soviet and French Underground Nuclear Tests,” March 4, 1965.
* It is not clear whether any aerial sampling aircraft were employed to gather the debris from the French tests in 1973, or during other testing campaigns. The WB-57 aircraft stationed at Yokota Air Base, Japan, could have been used for that purpose. There were also U-2Rs that conducted Olympic Race aerial sampling missions, including those that followed Chinese tests in July 1973 and June 1974. None of the SAC or Pacific Command histories that discuss the monitoring of French nuclear tests refer to such aerial sampling efforts, although they do with regard to China. See Strategic Air Command (SAC), History of SAC Reconnaissance Operations, FY 73 (Offut Air Force Base, Neb.: SAC, 1974), p. 84; Strategic Air Command, History of SAC Reconnaissance Operations, FY 74 (Offut Air Force Base, Neb.: SAC, 1975), pp. 65, 67.
* The treaty, formally the Treaty on the Non-Proliferation of Nuclear Weapons, and also known as the Nuclear Non-Proliferation Treaty, was signed before the end of 1968. It obligated the five acknowledged nuclear-weapon states of the time—the United States, the Soviet Union, France, the United Kingdom, and the People’s Republic of China—to refrain from transferring nuclear weapons, other nuclear explosive devices, or their technology to nonnuclear weapon states. Nonnuclear weapon states who signed the treaty agreed not to acquire or produce nuclear weapons or other nuclear explosive devices.
* Lord Krishna is a mythical Hindu deity who reportedly appeared in human form in 3328 BC and disappeared 125 years later.