The Wizards of Langley

by Jeffrey T Richelson

  On December 12, 1966, Deputy Defense Secretary Vance, Bureau of the Budget chief Charles Schultze, Helms, and presidential science adviser Donald Hornig met to consider the alternatives. Over Helms’s objection, they suggested terminating the OXCART program, leading the DCI to request that the Air Force share the SR-71 fleet. Helms asked Duckett to prepare a letter to the President stating the CIA’s reasons for wishing to continue the OXCART effort.39

  Four days later, Schultze handed Helms a memo for the President requesting a decision either to turn part of the Air Force SR-71 fleet over to the CIA or to terminate the OXCART program entirely. Helms, having just received new information that he believed demonstrated the A-12’s superiority, asked for another meeting after January 1 to review the data and requested that the memo to the President be withheld pending the meeting’s outcome. Helms believed that the SR-71 could not match the photographic coverage provided by the A-12—since only one of the three SR-71 cameras, its Operational Objective System, was working near specification. It could photograph only a swath twenty-eight miles wide with a maximum resolution of twenty-eight to thirty inches when the target was directly underneath (at nadir).40

  In contrast, the A-12’s Type I camera could photograph a seventy-two-mile swath with a maximum resolution of twelve to eighteen inches at nadir. Oblique images had a resolution of fifty-four inches. Thus, the A-12 camera covered over twice as much territory, with better resolution. In addition, the A-12 could fly 2,000 to 5,000 feet higher and was faster, with a maximum speed of Mach 3.1.41

  On December 27, a memo from deputy director of central intelligence Vice Adm. Rufus Taylor to national security adviser Walter Rostow noted that the memo that had been received the day before on the SR-71/ OXCART issue “did not quite fully reflect Helms’ opinion.” Taylor noted that the SR-71 could not yet be considered interchangeable with the OXCART, because the OXCART had been fully operational for a year and demonstrated greater performance than the SR-71, “which has not yet achieved operational capability.” In spite of Helms’s request for delay, the Bureau of the Budget memorandum was submitted to President Johnson. On December 28, he approved the mothballing of the OXCART fleet and its phaseout by January 1968.42

  The CIA had to develop a schedule for the phaseout of the A-12, an effort it code-named SCOPE COTTON. The agency informed Vance on January 10, 1967, that the A-12s would gradually be placed in storage, with the process to be completed by the end of January 1968. In May, the Deputy Defense Secretary directed that SR-71s would assume responsibility for Cuban overflights by July 1, 1967, and for Southeast Asian over-flights by December 1, 1967. Until those capabilities were developed, OXCART was to remain on call, capable of conducting overflights of Southeast Asia (on a fifteen-day notice) and of Cuba (seven-day notice).43

  In the midst of planning for termination of the program, the CIA continued advocating that an A-12 be employed for a special mission targeted on the Soviet Union. In May 1967, the 303 Committee received a CIA proposal to employ an OXCART, in conjunction with a U-2 carrying ELINT gear, to solve the mystery of the Tallinn system.44 Although the CIA, Navy, and State Department had concluded that the system was airdefense oriented and had little ABM capability, the Defense Intelligence Agency and Army Intelligence suggested that considerable uncertainty remained, while Air Force Intelligence chief Jack Thomas argued that the Tallinn system “probably was designed for and now possesses an area anti-ballistic missile . . . capability.”45

  Photointerpreters insisted that twelve-to-eighteen-inch-resolution imagery was needed to determine the size of the missile, the antenna pattern, and configuration of the engagement radars associated with the Tallinn system. Unfortunately, attempts to photograph it using the high-resolution GAMBIT satellite had been defeated by cloud cover. In addition, ELINT analysts needed data about the Tallinn radars, but there were no U.S. intercept sites that could monitor emanations when the radars were being tested. The Soviets also never operated the radars in tracking and lock-on modes, preventing analysts from determining the frequency or other performance characteristics of the radars.46

  To settle the question, the Office of Special Activities suggested a mission that would employ the OXCART’s high-resolution camera along with a U-2 flying a peripheral ELINT mission. The highly classified proposal had an unclassified designation—Project SCOPE LOGIC—and a classified code name—Project UPWIND.47

  OSA proposed flying an OXCART from the United States to the Baltic Sea, where it would rendezvous with a U-2. The A-12 would fly north of Norway and then turn south along the Soviet-Finnish border. Shortly before Leningrad, it would head west-southwest down the Baltic Sea, skirt the coasts of Estonia, Latvia, Lithuania, Poland, and East Germany, and then head west, returning to the United States. The entire flight would cover 11,000 miles, take eight hours and thirty-eight minutes, and require four aerial refuelings.48

  Although the OXCART would not intrude into Soviet airspace, it would appear to Soviet radar network operators to be headed directly over Leningrad. OSA hoped that the OXCART’s journey would provoke Soviet air defense personnel to activate the Tallinn system radars in order to track the aircraft. As the OXCART made its dash down the Baltic, its Type I camera would be filming the entire south coast, including Tallinn. If CIA analysts were correct and the system was designed to counter high-altitude aircraft at long ranges, then OXCART would be in jeopardy during its dash down the Baltic. However, the weapons experts in the Office of Scientific Intelligence believed that the A-12’s speed and electronic countermeasures would protect it from standard Soviet SAM installations. The more vulnerable U-2 would be flying farther out to sea, beyond the range of the SAMs. CIA and Defense Department officials supported the proposal, but Secretary of State Dean Rusk strongly opposed it, and the 303 Committee never forwarded the proposal to Johnson.49

  At the time Project UPWIND was under consideration, the CIA suggested another use for the OXCART—to determine if surface-to-surface missiles had been introduced into North Vietnam. Johnson asked for a proposal on the matter. The CIA briefed the 303 Committee, arguing that OXCART’s camera was far superior to the cameras on drones or on the U-2, and the plane was far less vulnerable. While State and Defense were examining the proposed political risks, Helms raised the issue at the President’s regular “Tuesday lunch” on May 16 and received approval for the deployment, which was dubbed BLACK SHIELD.50

  The agency wasted no time, and the airlift of personnel and equipment to Kadena Air Base in Japan began the day after Johnson approved the project. In less than two weeks, three OXCART aircraft arrived. By May 29, the CIA contingent was ready to fly operational missions, and the detachment was alerted to be ready the following day.51 There was no action that day, but an A-12 got its first taste of action May 31.

  A torrential rainstorm blanketed Kadena Air Base that day. Paul Ba-calis was there to witness the first launch and later recalled pilot Mel Vo-jvodich sitting in his plane on the runway awaiting orders. He also recalled aide Col. Slip Slater’s question—“What do you say, boss?”—as well as his response—“Launch him!” Bacalis watched “orange balls coming out of the rear of the aircraft” and the plane “disappearing into the overcast.”52

  Vojvodich then headed for Vietnam, flying just off the East China coast, without difficulty, at Mach 3.0. He prepared to enter hostile territory, flying at Mach 3.1 and at 80,000 feet, through the “front door.” He first flew over Haiphong and Hanoi, departed North Vietnam near Dien Bien Phu, refueled over Thailand, and penetrated enemy airspace near the demilitarized zone. Clear weather over the target area permitted photography of ten priority target categories as well as 70 of the 190 known SAM sites. In addition to SAM sites, the A-12’s cameras photographed Haiphong/Cat Bi Airfield, the port at Haiphong, a military training area, an army barracks, and a segment of the Hanoi/Lao Cai railroad. Three hours and thirty-nine minutes after takeoff, Vojvodich arrived back in Kadena, where it was still pouring—a downpour that forced him to make th
ree instrument approaches before landing.53

  Over the next six weeks, the detachment was alerted for fifteen missions and flew seven. Four detected radar tracking signals but no hostile action resulted. By mid-July, the imagery obtained from those missions provided enough evidence for analysts to conclude that no surface-to-surface missiles had been deployed in North Vietnam. Those missions also provided imagery of airfields, naval bases, port areas, railroad segments, iron and steel works, and supply depots. Thus, June imagery showed a variety of fighter aircraft at Phuc Yen airfield as well as several different classes of ships. From August 16 to December 31, twenty-six missions were alerted, and fifteen flown. Not until October 28 did a North Vietnamese SAM site fire at an OXCART, an event that was captured on film. The photos showed missile smoke around the SAM site and pictures of the missile and its contrail.54

  The missile fired that day did not endanger the spy plane, but several missiles fired two days later resulted in the closest call an OXCART would experience. During the plane’s first pass over North Vietnam, pilot Denny Sullivan detected that he was being tracked by radar and that two SAM sites were prepared to launch, although neither did. During his second pass, the North Vietnamese tried to bring down the plane with a barrage of at least six missiles. In addition to seeing the vapor trails, Sullivan witnessed three missile detonations near the rear of the A-12, which was traveling at Mach 3.1 at about 84,000 feet. After he returned, an inspection of the aircraft revealed that a piece of metal had penetrated the underside of the right wing, passed through three layers of titanium, and lodged against a wing tank support structure. The fragment was not a warhead pellet but probably debris from one of the detonations.55

  At least three of the missions conducted between August and December returned not only overhead photographs of North Vietnam but also oblique images of mainland China. The first leg of the September 17 OXCART mission took the plane to the northern reaches of North Vietnam and thus just south of China. During that portion of the flight, the plane’s camera captured images of military installations, railroad segments, storage areas, and urban complexes. An October 4 mission produced photography of Chinese air installations, a barracks, electronics/ communications sites, and naval and port facilities. Another mission later that month also yielded images of military facilities.56

  As a result of the capabilities OXCART demonstrated in its missions, high-level presidential advisers and congressional leaders began to question the decision to terminate the program, and the issue was reopened. The CIA continued to argue that the A-12 was a superior aircraft because it flew higher and faster and had better cameras. The Air Force maintained that its two-seat SR-71 had a better collection of sensors, with three different cameras (search, high-resolution, and mapping), infrared detectors, side-looking radar, and ELINT collection gear.57

  A series of missions, designated NICE GIRL, were conducted in an attempt to settle the issue. The two routes posed no risk to the pilots, since they flew over the continental United States. The first two missions were flown October 20 and October 25, 1967. The final missions were flown November 3, when an A-12 and an SR-71 flew identical flight paths, separated in time by one hour, from north to south, roughly above the Mississippi River. The data collected during the missions were evaluated by representatives of the CIA, DIA, and other Defense Department intelligence organizations.58

  The results proved inconclusive, with both photographic systems providing imagery of sufficient quality for analysis. The A-12’s Type I camera with its seventy-two-mile swath width and 5,000-foot film supply proved superior to the SR-71’s Operational Objective camera with its twenty-eight-mile swath and 3,300-foot film supply. On the other hand, the SR-71’s infrared, imaging radar, and electronic and communications intelligence equipment provided some unique intelligence not available from the A-12. Air Force planners admitted that some of this equipment would have to be sacrificed to provide the SR-71 with electronic countermeasures gear when it flew over regions of the world more hostile than the central United States.59

  Although the flyoff did not settle the question of which aircraft was superior, OXCART won a temporary reprieve in late November 1967 when the Johnson administration decided to keep both fleets for the time being. But with war costs rising, a challenge to maintaining competing reconnaissance programs was bound to occur again.60

  A December 29, 1967, memo from Deputy Defense Secretary Paul Nitze to Gen. Earle G. Wheeler, Chairman of the Joint Chiefs of Staff, and Alexander Flax, director of NRO, announced the decision to maintain an OXCART capability, consisting of five operational aircraft, through June 30, 1968. The memo also called for planning for the introduction of the SR-71 into reconnaissance operations in North Vietnam “as rapidly as ECM [electronic countermeasures] implementation and other program considerations will permit.”61

  BLACK SHIELD missions continued during the first three months of 1968, with six missions flown of the fifteen alerted—four over Vietnam and two over North Korea. The last OXCART overflight of Vietnam occurred on March 8. On January 26, an A-12 overflew North Korea for the first time, in response to the North Korean seizure three days earlier of the USS Pueblo, a signals intelligence ship. The objective was to discover where the Pueblo was being held, and whether North Korea was preparing any large-scale hostile move in the wake of the incident.62 The A-12 pilot, Frank Murray, recounted his flight:

  I left Kadena, topped-off, then entered northern airspace over the Sea of Japan via the Korean Straits. My first pass started off near Vladivostok, then with the camera on I flew down the east coast of North Korea where we thought the boat was. As I approached Wonsan I could see the Pueblo through my view sight. The harbour was all iced up except at the very entrance and there she was, sitting off to the right of the main entrance. I continued to the border with South Korea, completed a 180-degree turn and flew back over North Korea. I made four passes, photographing the whole of North Korea from the DMZ to the Yalu border. As far as I knew, I was undetected throughout the flight but when I got back to Kadena some folks told me that the Chinese had detected me and told the North Koreans, but they never reacted.63

  Murray’s mission, which showed no signs of any upcoming action by North Korea, was followed by two further missions on February 19 and May 8. To the extent those missions responded to DIA requirements, they included coverage of seven jet-capable airfields, naval facilities (including the Mayang Do submarine facility), and the North Korean–Chinese border. Dean Rusk had been reluctant to approve a second overflight for fear that the plane might be brought down in hostile territory, but was assured the A-12 would slice through North Korea in seven minutes and was unlikely to land in either China or North Vietnam.64

  In the midst of the overflights, the question of OXCART’s future had remained open, as national security adviser Walter Rostow, members of the PFIAB, key congressmen, and the President’s Scientific Advisory Committee questioned whether the program should be terminated. A new study was ordered and completed in spring 1968. Analysts considered four alternatives—transferring the entire OXCART fleet to SAC by October 31, 1968, and turning A-12 test aircraft over to the SR-71 test facility; transferring all OXCART aircraft to SAC and storing eight SR-71s; closing the OXCART home base and collocating the fleet with SR-71s at Beale AFB in California, but with the CIA retaining control and management; continuing OXCART operations at its own base under CIA control and management.65

  Not surprisingly, Helms preferred the last option, a position he conveyed to Nitze, Hornig, and NRO director Al Flax in an April 18 memo. He questioned the collocation option on the grounds of security and the lower cost figures associated with combining the two fleets. The key point, he argued, was the desirability of retaining a covert reconnaissance capability under civilian management.66

  On May 16, Clark Clifford, the new Secretary of Defense, reaffirmed the original decision, and President Johnson confirmed it on May 21. Two months earlier, in accord with Vance’s orders, SR-71s had
begun arriving at Kadena to take over the BLACK SHIELD mission. After the May 8 mission over North Korea, members of the Kadena detachment were advised to pack their bags and prepare to return home. By mid-June, all the OXCARTs were back in the United States, their espionage careers prematurely terminated.67

  CATS AND BIRDS

  While portions of the directorate focused on aerial and underwater reconnaissance, the Office of Research and Development was busy pursuing a variety of projects. By mid-1968, Robert Chapman, a geophysicist, was in his fourth year as head of ORD; he had become acting director in May 1964 and director in March 1965.68

  ORD continued experimentation on both humans and animals. In 1968, the office established a joint program, Project OFTEN, with the Army Chemical Corps at Edgewood Arsenal Research Laboratories (EARL) in Maryland to study the effects of assorted drugs on human and animal subjects. The Army not only assisted ORD in building a computerized database for drug testing but also supplied military volunteers for some of the experiments.69

  The research and development office paid EARL $37,000 to perform new tests on a compound that Edgewood had dubbed “EA 3167” and previously tested on volunteers. ORD’s Medical and Behavioral Sciences Division suspected that EA 3167, a powerful incapacitant, could be administered through a handshake or other casual contact. In 1971, it contracted with Edgewood for additional tests of EA 3167. The contract ended in 1973 before any tests could be performed, according to the EARL official who oversaw the contract.70

  Cats and dogs did not have the option of volunteering. Victor Marchetti, who served as executive assistant to the deputy DCI during the late 1960s, recalled an attempt to turn cats into mobile bugging devices—a project commonly referred to as “Acoustic Kitty.” The project was “more than just a goofy operation,” according to Marchetti. The problem with the microphones of the day was that they picked up all the sound in a room—from voices to tinkling glasses—often producing recordings in which conversation could not be filtered out from other noises. A bug placed in the couch of a Chinese diplomat in France proved ineffective because of the squeaking noises that drowned out conversations, not only when the diplomat was using it for his frequent sexual escapades but when visitors were simply sitting.71