"Cabinet-grade walnut," replied the director with pride in his knowledge of woods.

"No, sir," corrected the concealment specialist, "that is volume in a cellulose wrapping. And we can put anything we choose underneath the wrapping as long as it doesn't exceed the volume."

Once the "volume" of the bug package, consisting of microphone, transmitter, switch receiver, power cells, and antenna were reduced to six cubic inches or less, relatively small blocks of wood could encase all the system's components. The wood block became the workhorse for "quick plant" audio operations. "RF transparent" wood could be cut into almost any configuration with hand tools and then screwed, bolted, glued, or wedged into place. Small blocks could be fashioned to blend in with furniture, the molding in an office, or a picture frame by matching wood types, grains, and finishes.

For twenty years after the introduction of the SRT-3, each successive SRT model saw either the size of the transmitters decrease or improvements in performance or security.2 Transmitter models in the mid-1960s also marked the introduction of signal masking systems to defeat audio countermeasures. Without masking, a technical sweep team could inspect a facility with electronic and magnetic equipment that scanned the RF signal spectrum and detected foreign objects to locate, lock onto, and expose the secret audio transmissions. Masking reduced the vulnerability of bugs by making their signal harder to isolate and identify as clandestine transmissions. Transmitter models in the mid-1960s also marked the introduction of signal masking systems to defeat audio countermeasures. Without masking, a technical sweep team could inspect a facility with electronic and magnetic equipment that scanned the RF signal spectrum and detected foreign objects to locate, lock onto, and expose the secret audio transmissions. Masking reduced the vulnerability of bugs by making their signal harder to isolate and identify as clandestine transmissions.

One masking technique commonly used by both the United States and Soviets buried the transmission in the signal's subcarrier. RF transmissions were designed to broadcast in two parts, much like stereo. The first part, a clear signal resembling white noise, was pa.s.sed over as benign by someone scanning the radio spectrum. Then, just to the left or to the right on the dial-up or down the spectrum-was the subcarrier with the clandestine message. By tuning to the right frequency and tuning out the white noise, it was possible to hear the covert transmission. In principle, the use of subcarriers worked like hiding a piece of clear gla.s.s in a container of water. The gla.s.s remains invisible until the water is drained.

Other techniques for using subcarriers sent the audio signal along the existing AC power lines where it was collected and retransmitted to a listening post. Signals could be encrypted, masked, or both.

True to the nature of espionage, each technological advance was inevitably met by an effective countermeasure. In time, KGB counterintelligence teams began tuning in on the white noise in search of subcarrier transmissions. OTS responded and advanced the technology to the next level. "Concealing signals was an area I felt very strongly about," said an OTS manager who oversaw the program. "I wanted to come up with new modulation schemes, every year I wanted at least four or five brand-new ones to hide our transmissions. For a while we got into a pattern of using certain types of subcarriers almost exclusively and, unfortunately, the Russians knew what to look for in our 'offsets.'"

Eventually these techniques of hiding transmissions came to include a frequency-hopping technique in which short transmission bursts bounced up and down the radio spectrum in no apparent order. Without a receiver coordinated to the changes in transmissions, these frequency hops proved particularly difficult to identify and intercept since it was nearly impossible for sweep teams to antic.i.p.ate the signal's pattern.

The complexities and opportunities presented by clandestine audio seemed endless. Installing an audio bug always put the techs at personal risk of discovery and arrest when entering, leaving, or working at a target. Building reliable miniature components for covert systems that could withstand extreme environments challenged the best engineering minds. Configuring the system to operate within the available concealment s.p.a.ce required mastery of craftsmanship and design, but no matter how sure the tradecraft and skilled the engineering, none of that mattered without access, and some targets were virtually inaccessible.

The problem of access led TSD and its partner, the Office of Research and Development (ORD), to experiment with an array of exotic audio surveillance delivery systems.3 In the early 1960s, Soviet diplomats in one Central American capital city often conferred in their emba.s.sy's courtyard on matters they believed were too sensitive to risk discussing in their offices, which they believed were possibly bugged. The courtyard, while surrounded by a security fence, was not walled and CIA officers observed that one bench seemed to be a favored gathering place for Soviets of particular interest. Adjacent to the bench was a large shade tree. Station officers had no means to gain access to the bench inside the emba.s.sy compound so the DDP turned to TSD to devise a means to bug the conversations that occurred around the bench. The open security fence surrounding the emba.s.sy led to the idea of shooting a bullet containing the microphone and transmitter into the tree just above where the diplomats usually conferred. In the early 1960s, Soviet diplomats in one Central American capital city often conferred in their emba.s.sy's courtyard on matters they believed were too sensitive to risk discussing in their offices, which they believed were possibly bugged. The courtyard, while surrounded by a security fence, was not walled and CIA officers observed that one bench seemed to be a favored gathering place for Soviets of particular interest. Adjacent to the bench was a large shade tree. Station officers had no means to gain access to the bench inside the emba.s.sy compound so the DDP turned to TSD to devise a means to bug the conversations that occurred around the bench. The open security fence surrounding the emba.s.sy led to the idea of shooting a bullet containing the microphone and transmitter into the tree just above where the diplomats usually conferred.

For the concept of the "bullet bug" to work, TSD would need an audio device small enough to fit into a projectile, a means to clandestinely shoot the package into the tree, and components that would tolerate the velocity and impact necessary to bury a projectile far enough into the tree to escape notice.

A TSD engineer took the concept to the president and chief scientist of America's leading hearing aid company, asking that they build a microphone small enough to fit into a .45 caliber bullet and rugged enough to function after hitting a tree. The problem of small size appeared solvable, but nothing in the company's inventory would tolerate the shock. As the technical discussion progressed, problem after problem arose. It seemed apparent the idea had no future, until the president suddenly interjected. "Well, it's a really good challenge, yes, let's do it." A team of engineers was formed to create a one-of-a-kind microphone with no manufacturing markings or signature.

After obtaining a similar commitment from a company specializing in small transmitters, TSD began testing and evaluation. Within three months, a 400 MHz transmitter, battery, and microphone small enough to fit into the projectile, somewhat larger than a .45 caliber bullet, were delivered. Battery life was limited to less than a day due to size limitations.

The antenna was a simple wire that trailed behind the projectile after it left the barrel of the gun, but presented a problem since it caused the projectile to wobble in flight and hit the target broadside. Over time, the techs found that by adjusting the antenna length the projectile would fly true, embed itself at the proper angle, and maintain the audio link to the listening post.

A vintage World War I rifle became the test weapon. Its long-rifled barrel enhanced accuracy by building up projectile speed and stabilizing the bullet and antenna before leaving the barrel. Test firings into three one-inch plywood targets clamped together were conducted at an abandoned rock quarry near Baltimore, Maryland.

Opting for safety, as they were using an old gun and unconventional ammo, the techs mounted the rifle to a table, placed sandbags around it, and attached a cord to the trigger for firing. After a few test shots, when the rifle did not fly apart, the more courageous techs fired it from a shoulder position. Repeated firings determined the correct amount of powder needed to limit the projectile penetration to no more than two inches, the maximum depth from which the microphones and transmitter could operate.

The techs found it impossible to use a standard silencer on the weapon so, to quiet the report, they jerry-rigged a fifty-gallon steel drum filled with acoustic baffles. Both ends of the drum were cut away and a center area of free s.p.a.ce was created through which the weapon could be sighted. When fired from within the makeshift acoustic chamber, the sharp firing noise was reduced to a ba.s.s boom. Because the weapon was still too loud for operational use and they were without a technical solution, the planners envisioned a scenario in which two loud motorcycles would start at precisely the time the weapon was fired, masking the gunshot for anyone who might be within hearing distance.

From the first testing, the transmitter and battery components proved both reliable and functional. The microphones required several adjustments, in part because prior to the requirement, magnetic microphones were designed to withstand drop impacts only, a significantly lower stress than a bullet impact. Eventually, the microphones and other components proved consistently durable when the projectile traveled at approximately 500 miles per hour over distances up to 50 yards. The microphones picked up sounds of a portable radio sitting next to the plywood target and transmitted quality audio up to 250 feet.

In the next round of field tests, the "audio bullets" were fired into live trees to simulate an operational scenario. Once fired into the tree, two people seated nearby carried on a conversation at normal voice levels. Surprisingly, audio quality was poor compared to the plywood tests. a.n.a.lysis showed no damage to the device, but the live tree wood proved different from plywood. Tree fibers when hit by the projectiles formed cones similar to the design of an echo-free anechoic chamber and swallowed up the audio.

Additional a.n.a.lysis determined that if the transmitter's size was increased, the necessary audio amplification could be attained. This would require, however, a larger bullet, increased firing noise, and a redesign of the weapon itself. The hole in the tree would also become larger and more noticeable. In the end, the DDP judged the value of the potential information insufficient to justify TSD's cost in time and dollars for additional development and the bullet bug was filed away.

While the project did not result in an audio "silver bullet" to bug the Soviets, technologies for high-reliability miniature microphones did emerge. Based on data obtained from the tests, TSD produced a series of very small microphones that could withstand high impact and high heat stress. This new generation of rugged microphones could endure rough handling, be installed in almost any wet or dry material, and perform at near zero failure rates regardless of where they were buried. Commercially, the research and design effort by the contractor produced shockproof microphones that enabled the size of hearing aids to shrink along with improving the microphones' performance in varied temperature and high moisture environments.

Animals as well as technology played starring roles in the quest to prove that any target could be "hit." When CIA operatives sought a means to penetrate the private meetings of an Asian head of state, reports reached Headquarters that during the target's long strategy sessions with his aides, cats wandered in and out of the meeting area. Feral cats were common to the region and generally ignored. Whether the concept of an "Acoustic kitty" came from a case officer or a tech is lost to memory, but the idea launched a research project that generated unwarranted ridicule and accusations after public disclosure.4 In fact, absent from the Acoustic kitty project were both cruelty and mutated, grotesque creatures from horror movies.5 From the beginning, the techs recognized that the concept, undertaken jointly between OTS and the Office of Research and Development, fell into the high-risk category. At the time, embedding electronics inside animals or people was not a routine medical procedure. From the beginning, the techs recognized that the concept, undertaken jointly between OTS and the Office of Research and Development, fell into the high-risk category. At the time, embedding electronics inside animals or people was not a routine medical procedure.

The implant could not affect any of the natural movements of the cat nor could the cat experience any sense of irritation or the presence of the device lest it induce rubbing or clawing to dislodge components or disturb performance. The audio system components would include a power source, transmitter, microphone, and antenna.

Working with their prime audio equipment contractor, the techs produced a three-quarter-inch transmitter for embedding at the base of the cat's skull where loose skin and flesh provided a natural pocket. Implanting the transmitter proved viable, once a device was packaged to withstand the temperature, fluids, chemistry, and humidity of the body. Microphone placement presented a more difficult problem since flesh is a poor conductor. Eventually, the ear ca.n.a.l became the preferred location. An antenna of very fine wire was attached to the transmitter and woven into the cat's long fur. The cat's size permitted only the smallest batteries, a factor that restricted the amount of hours the audio could transmit.

Research to determine the performance of the various individual components and the most effective placement areas was conducted first on dummies and then live cats. Doc.u.mentation of reactions of the cats to the "foreign materials" and to nerve stimulation refined the research and eventually produced an integrated audio system suitable for dress rehearsal. Agency officials reviewed questions of humane treatment of the animals and the potential negative publicity should the activity become publicly known. After those factors were weighed against the operational value of the project's success, the techs received authorization to go forward.

A small crowd stood behind the vet who conducted an hour-long procedure on a full-grown, anesthetized gray-and-white female cat in a clean, brightly lit animal hospital. The TSD chief audio engineer, seeing the first incision and a trace of blood, asked to sit down. No other complications arose, and after the cat awakened, she was put into a recovery area for further testing. Technically the audio system worked, generating a viable audio signal. However, control of the cat's movements, despite earlier training, proved so inconsistent that the operational utility became questionable. Over the next few weeks, Acoustic kitty was exercised against various operational scenarios, but the results failed to improve.

Acoustic kitty demonstrated that transmitters could be embedded in animals without damage or discomfort. The experimental animals could be directed to move short distances to target locations and people in a known environment. However, outside the experimental laboratory, Acoustic kitty had a mind of its own. Eventually, deployment of Acoustic kitty in a foreign environment over which the "handler" would have no a.s.sured control was judged impractical and the project was closed.6 Exotic detours aside, OTS's most productive audio ops followed a disciplined formula. Identify an operational requirement, select a target, survey the target, a.s.semble the right equipment, establish a listening post, make the entry, install the device, test the system, restore any damaged area to original condition, dispose of any evidence of being there, and get out without getting caught. Audio techs improvised when it came to tools, combining an a.s.sortment of commercially available hardware store implements with specially fashioned gear made either in the lab or of their own design.

"Acoustic kitty" was TSD's attempt to implant a clandestine listening device in a cat, mid-1960s.

In one operation, a tech used standard well-drilling equipment- configured to operate horizontally-to drill more than a hundred feet from the listening post to the target site on the opposite side of a major thoroughfare. "We bugged every room in that building, and then brought all the wire leads down to the bas.e.m.e.nt," the tech who led the operation remembered. "I surveyed in on where I wanted to enter the bas.e.m.e.nt, then drilled a hole using just the azimuth and elevation data from the post to the target. I came out a foot away from target hole. The case officer said, 'You missed.' I said 's.h.i.t, a hundred and nine feet underground, in a foreign city, I think I did pretty good,' and we figured out a way to compensate for the other twelve inches."

Some techs excelled at solving problems with their own inventions to meet specific operational needs. Although of little use beyond covert operations, the devices were invaluable for making installations. The Nail Pusher, Nail Pusher, or or Silent Hammer Silent Hammer, was used for restoration work on baseboards and molding. Essentially the device was a hollow tube with a plunger-type mechanism to reinsert nails silently without leaving traces of a hammer mark.7 One innovation that earned its inventor a unique, if dubious, reputation among his fellow techs was a new microphone housing. Techs had long been beleaguered by the challenge of securing a mic into position within the hole drilled to reach the target's wall. Too often after the tech carefully positioned a microphone in the hole against the pinhole, it slipped slightly away from the tiny air pa.s.sage before being firmly anch.o.r.ed. If unnoticed at the time, the smallest misalignment produced a degraded sound. The tech's clever solution encased the mics in a sheath of pilable latex that fit snugly into the three-eighths-inch-diameter hole leading to the pinhole. Because of the phallic appearance, techs named it the Peter Mic Peter Mic.

As intelligence flowed through audio's reliable equipment, so did the audio techs' confidence in their tradecraft skills. Among the techs, and even case officers, the thinking became, "If access could be obtained, almost any target was vulnerable." In some respects, this was "spiral development." Hard targets required greater tradecraft skills and, as those skills were acquired, they were applied to even harder targets.

The increased sophistication of bugs and a willingness to take on the tough operations required better equipment. For instance, drilling holes represented a core skill for audio techs. Holes for bugs were drilled down from ceilings, up from underneath floors, and horizontally in walls. When the techs could not physically get inside a room to install a bug, they drilled through a common wall. The danger of such an operation lay in the fact that the techs were literally blind to what or who was on the target side of the wall.

These drilling operations had two major security risks: noise and unintended breakthrough. Electric drills were fast, but so noisy they were not an option for use in the middle of the night or with the target room occupied. Hand-turned drills were slow and difficult with harder construction materials. To drill quietly usually meant drilling so slowly an installation could require days, especially if multiple bugs were being installed.

In a typical operation, techs preferred to start with a three-eighths-inch drill bit (the hole had to be large enough for the circ.u.mference of the microphone) until reaching the final half-inch of material in the target's wall. At that point, depending on the size of the microphone head, drill bits of less than .050 inches were used to drill a breakthrough pinhole. The tiny hole created enough of an air pa.s.sage for clear audio pickup while virtually invisible to normal observation.

With blind drilling techs never knew how close they were to the breakthrough point. Even for the best drill tech, it was a matter of guess, estimate, feel, and experience. Wrongly judged, the drill's breakthrough would leave a noticeable hole in the target's wall and debris on the floor. "If we don't know the thickness of the wall, then we don't know for sure when we are close to the other side," explained one tech. "So if we punch through a wall with a three-eighths-inch hole, somebody is going to notice. Sometimes when we did inadvertently drill through, we joked that our audio operation became a video operation."

Over the years, more than one tech accidentally broke through a wall, then looked into the hole only to see a curious eye peering back. During one operation, a tech drilled a larger than intended hole into a Soviet apartment. Several minutes pa.s.sed, then without knocking, the Russian diplomat burst into their room and furiously began berating his "neighbors" for their carelessness in punching a hole through his wall while hanging pictures. The case officer apologized and a.s.sured the diplomat that his workers would be more careful in the future.

Targets were not the only ones displeased with tech mistakes. On a seemingly routine operation, the techs made an uneventful entry into a commercial building and began drilling the starter hole into the common wall with a Soviet trade mission. Suddenly, the drill bit broke through, creating a gaping hole in the adjoining room. With no way to repair the damage, the best that the techs could manage was to patch their side of the wall and retreat to the local chief of station's office to report their problem.

"We have a nice hole in the wall for audio," the head tech reported.

"Good," the COS replied, "very, very good."

"Well, what I mean," explained the tech, "is that we broke through with a really nice big hole."

The chief went ballistic. "Get out of my country and never show your face here again!" He spoke the order loudly enough that the techs heard each word and didn't even think of putting up an argument.

"I guess that tells us something about the lack of a sense of humor some of these guys have," whispered one of the team members.

A few weeks later, after the chief calmed down, another visiting OTS officer suggested that, given the importance of the target, another attempt was merited. The chief demanded and received a.s.surances a similar mistake would not be made. The installation went off without a problem.

Listening to the live audio a few days later, the techs heard sounds of a work crew coming into the room. Not clear from the conversation was whether this was a sweep team or construction crew, but they were obviously looking carefully at the walls. The techs held their breath, waiting for what would happen next.

"Look at this," one guy said.

"d.a.m.n, what's that hole? That shouldn't be here," came the reply. "Well, we better get rid of it."

With relief, the techs listened as the conscientious construction crew repaired the wall with the three-eighths-inch hole, never noticing the pinhole a few feet away. This time even the chief of station was amused.

To avoid the disasters of noise or breakthrough, techs would often drill just a few revolutions per minute. To create the pinhole, they would slowly twist a six-inch cylinder shaft that held the tiny bit with thumb and forefinger, applying little, if any, pressure, and letting the bit pull itself through the final fraction of an inch.

The problem of measuring the thickness of the remaining wall between the end of a drill bit and "breakthrough" was partially solved by one of the OTS's cleverest tools, the Backscatter Gauge. Backscatter Gauge.8 While the basic technology employed was not new, its covert application was a model of ingenuity. The principle behind backscatter technology is nuclear science. A tiny radioactive source emits a steady pulse of gamma rays, bouncing them off an object while a reader contained in the unit measures the number of pulses that bounce back. A thick material will repel more gamma rays than thin material. The gauge calculates the percentage of the returning rays against the number emitted. For example, an object that bounces back 50 percent of the pulses is twice as thick as one that returns only 25 percent. A more sophisticated version of the technology later evolved into security devices used to scan baggage and individuals at airport checkpoints. In this configuration, with advanced signal processing, the returned gamma rays actually paint a picture, similar to that of an x-ray. While the basic technology employed was not new, its covert application was a model of ingenuity. The principle behind backscatter technology is nuclear science. A tiny radioactive source emits a steady pulse of gamma rays, bouncing them off an object while a reader contained in the unit measures the number of pulses that bounce back. A thick material will repel more gamma rays than thin material. The gauge calculates the percentage of the returning rays against the number emitted. For example, an object that bounces back 50 percent of the pulses is twice as thick as one that returns only 25 percent. A more sophisticated version of the technology later evolved into security devices used to scan baggage and individuals at airport checkpoints. In this configuration, with advanced signal processing, the returned gamma rays actually paint a picture, similar to that of an x-ray.

When OTS adapted the technology for clandestine purposes in the 1970s, backscatter was widely used in industrial applications, primarily for quality control. "It was being used in paper mills to keep the thickness or consistency of paper constant," explained Martin Lambreth, the engineer who helped design the system. "By using the backscatter technique, the thickness could be measured continuously as it moved through production. As long as the radiation remained the same, the product was good. If it changed, they'd stop production. We wanted to use the same principle for measuring the distance from our drill bit to the surface of the wall we couldn't see."

With the OTS version of the technology, engineers developed a unit attached to a probe that fit into the three-eighths-inch drill hole. Techs would drill a short distance into a wall, then withdraw the drill and insert the probe to take a measurement. This drill-and-probe process continued as the electromechanical counter mounted on the unit recorded a wall's thickness at the deepest point of the drilled hole. In time, techs became so proficient that some abandoned consulting the small mechanical readout altogether, preferring to judge depth by the clicks of the counter. The faster the clicking sounds, the thicker the wall. Since drilling sometimes occurred in near darkness, this also reduced the need for illumination and added a measure of security to nighttime jobs.

"I just listened for the clicks. I'd be on a ladder, the gauge is on the floor attached by a long cable going click-click-click," said Martin, an experienced tech. "Then, after awhile, I'd hear click . . . click . . . click . . . and say, something's about to happen. I better be careful from now on." The reading, while not precise, was close enough to eliminate virtually all breakthroughs and earned praise from audio techs around the world.

A second drilling innovation adapted from industrial technology was the Grit Drill. Grit Drill. Smooth plaster walls presented a particularly difficult problem for the audio techs. It seemed that every diplomat who happened to be an operational target also had an office or home with plastered walls. To cut through the plaster, some pressure was needed on the drill, but no matter how careful the tech tried to be, that pressure was just enough to spall the other side of the wall. Small chips of plaster from spalling were a dead give-away to anyone doing a security inspection. Smooth plaster walls presented a particularly difficult problem for the audio techs. It seemed that every diplomat who happened to be an operational target also had an office or home with plastered walls. To cut through the plaster, some pressure was needed on the drill, but no matter how careful the tech tried to be, that pressure was just enough to spall the other side of the wall. Small chips of plaster from spalling were a dead give-away to anyone doing a security inspection.

OTS management dispatched an engineer, working under an alias as well as commercial cover to obscure CIA interest, on a nationwide tour to find a solution. Lugging dozens of circular samples of brick, concrete, ceramic tiles, terrazzo, and other materials neatly packaged in plastic sleeves, he began a cross-country journey in search of a better drill.

He visited more than a dozen companies, large and small. Anyone who knew anything about drilling through hard materials was fair game. Those who agreed to meet with the engineer were told that his company was looking for a drill so fine and so tough it could penetrate every one of the circular samples with a clean one-millimeter hole.

He visited precision drilling companies that cut holes in circuit boards and scientists in labs working with microwave energy. In upstate New York, he paid a call on a company that excavated concrete and was anxious to help. "I think we can do it," said one of the mining engineers. "We would use a very small controlled explosive with great care." The tech found the concept intriguing but the idea of using explosives, no matter how small, was not something he could likely sell to OTS.

The engineer eventually found a research company in the South that said it had a scientist with a reputation for innovative engineering. So far, the search had been futile, but he went through the requirements yet again. There was little reaction except the scientist asked that some of the material samples be left behind. Because it was the last stop on the circuit, the tech was happy to discard the weight from his luggage. Back at OTS, he reported little progress on the problem.

Nothing happened for a few weeks, and then, unexpectedly, the OTS engineer received a call from the scientist who had kept the samples. "I've got the solution," offered the caller, and the engineer was on the next plane going south.

At the lab, the scientist rigged up an old thermal drill, a type used for a brief time by dentists. The drill utilized a very fine nozzle and air pressure to shoot a thin, high-velocity stream of extremely small aluminum oxide particles, essentially eroding the tooth's enamel to create a hole rather than drilling it out. While erosion eliminated some of the discomfort of pressure during drilling, patient complaints about the taste of the particles made the technology unacceptable.

The OTS engineer pulled out samples of every possible material and the two went to work. They drilled holes in gla.s.s, concrete, plaster, stucco, and ceramic tile. No material was spared from testing and a clean hole appeared in each sample. Both engineer and scientist admired what the drill had accomplished.

"Works great, you solved the spall problem. And it's no good to us," the engineer told the stunned scientist.

A lengthy conversation followed as the engineer pointed out that despite the precise, clean holes, the drill would not be usable in a clandestine operation since it sent a fine spray of particles through the hole at breakthrough. If a target room was on the other side of the hole, the drill would deposit a fine coating of dust on carpet, furniture, and files that would surely alert the room's occupant.

The scientist listened intently and asked questions about the nature of operations and the special tools that were required. "Can I keep working on this?" he asked.

The OTS engineer readily agreed. Here was a scientist with demonstrated creativity and who was clearly hooked by the challenge of clandestine requirements. Had he been a case officer, the engineer would have claimed a "recruitment."

A few days later, an OTS secretary took a cryptic phone message from the scientist. "Tell my friend to come on down, he'll know what it's about."

The following day the OTS engineer watched as the scientist attached a sleeve to the device that was inserted into the hole being drilled. The drill went into the sleeve, which was sealed around the outside of the hole. Extending from the collar was a hose that ran from a filtering system to a clear Plexiglas tube. "He blew the grit back into part of a vacuum cleaner bag, so he could filter the stuff out. Then coming out of the vacuum bag was a clear tube and in that tube was a Ping-Pong ball and a photoelectric cell on the side. That was the on/off switch," the engineer explained. "So, when the gas came on for drilling, it created air pressure in the tube and lifted the Ping-Pong ball. While drilling he had positive gas flow and ball stayed up. As soon as the drill tip broke through, the pressure in the tube dropped, the Ping-Pong ball fell, triggering the photoelectric cell, and the d.a.m.ned thing turned off."

The operational dust problem was solved along with the problem of cutting a clean pinhole through smooth plaster. OTS engineers reconfigured the device for portability and named it the Grit Drill. Helium was subst.i.tuted for compressed air for its higher exit velocity. The entire Grit Drill and its accessories could be squeezed into a standard briefcase.

Once the Grit Drill Grit Drill kit was certified for deployment, the well-traveled engineer received orders to demonstrate the system to the overseas OTS tech bases. There, skeptical audio techs a.s.sembled for a demonstration of the "latest solution from Headquarters." For field types, the Headquarters' show-and-tells had a reputation for bringing more hype than practical value. kit was certified for deployment, the well-traveled engineer received orders to demonstrate the system to the overseas OTS tech bases. There, skeptical audio techs a.s.sembled for a demonstration of the "latest solution from Headquarters." For field types, the Headquarters' show-and-tells had a reputation for bringing more hype than practical value.

The engineer described the system, explained why it worked and demonstrated how to set it up. He then began punching tiny holes through materials the most susceptible to spalling. Each tiny hole was perfect, with no spall appearing on the target side of the sample materials. But before the demonstration was finished the chief of audio operations interrupted. "I've seen enough," he declared. "I'm doing an operation tomorrow; bring that thing along because you're going with me." The engineer was stunned. Not only had he never been on a clandestine operation, the Grit Drill Grit Drill had never been used operationally. had never been used operationally.

Three days later the engineer's anxiety had turned to excitement. The operation went smoothly with the engineer personally operating the Grit Drill Grit Drill during its first use in an audio installation. Now with an operational success as part of his briefing, he visited other techs in the region demonstrating the drill to interested audiences. But it was the thrill of the clandestine work that stayed with him. Like his scientist friend, he had been hooked by the excitement of clandestine operations. Back at Headquarters, after completing the show-and-tell TDY, the engineer transferred from his lab a.s.signment to the cadre of audio tech officers. during its first use in an audio installation. Now with an operational success as part of his briefing, he visited other techs in the region demonstrating the drill to interested audiences. But it was the thrill of the clandestine work that stayed with him. Like his scientist friend, he had been hooked by the excitement of clandestine operations. Back at Headquarters, after completing the show-and-tell TDY, the engineer transferred from his lab a.s.signment to the cadre of audio tech officers.

As the audio surveillance equipment improved and inventories of transmitters, power sources, microphones, and installation tools grew, the techs themselves required additional training to configure and test equipment. The era of the "tinkerer" had ended. The days of on-the-job, trial-and-error audio training gave way to more rigorous and formal instruction aimed at greater professionalism and broader knowledge of increasingly complex equipment.

"If we weren't doing this, we'd be robbing banks," said Antonio J. "Tony" Mendez, one of the three OTS technical officers honored in 1997 as a CIA "Trailblazer."9 No element of an audio operation created a more intense adrenaline rush than a surrept.i.tious entry into a secure and guarded target. Surrept.i.tious entry, defined as "an entry by stealth," accompanied almost every clandestine audio installation. No element of an audio operation created a more intense adrenaline rush than a surrept.i.tious entry into a secure and guarded target. Surrept.i.tious entry, defined as "an entry by stealth," accompanied almost every clandestine audio installation.10 In most instances, the techs entered the premises or property of the target and then made additional entries into luggage, mail, or vehicles. Due to the risk involved, surrept.i.tious entries were thoroughly scripted and rehea.r.s.ed prior to the operation. In most instances, the techs entered the premises or property of the target and then made additional entries into luggage, mail, or vehicles. Due to the risk involved, surrept.i.tious entries were thoroughly scripted and rehea.r.s.ed prior to the operation.

During World War II, the OSS created a surrept.i.tious entry capability by enlisting former second-story men with prior experience in burglary, lock picking, and safe cracking. OSS created and issued a small "lock picking knife" containing "picks" instead of blades, that could be conveniently carried in the operative's pocket for quick access when needed. The original OSS surrept.i.tious-entry manual cited the purpose for their work as helping the agent solve his problem: .

He wishes to obtain access to secret doc.u.ments, copy or memorize their contents, and leave the premises in the same condition as he found them. To arouse suspicion that the entry had been made, would in many cases be as fatal as being caught in the act of rifling the safe. The agent should therefore learn thoroughly the technique of surrept.i.tious entry, so as to adapt it, as the occasion requires, to a similar job in enemy territory.11 .

The reasons for conducting entry operations did not change after World War II, only now the "enemy territory" became the guarded official missions of America's Cold War adversaries scattered in countries throughout the world. While all the audio techs were trained in the basics of surrept.i.tious entry, a handful of officers specialized in the work.12 These techs were skilled at climbing ladders, bypa.s.sing alarm systems, picking locks, cracking safes, and performing room searches as well as installing devices. These entry specialists demonstrated that, if given sufficient time and resources, virtually any lock could be opened and any alarm system bypa.s.sed, though there were always limits on time and the amount of equipment that could be deployed at the job site. These techs were skilled at climbing ladders, bypa.s.sing alarm systems, picking locks, cracking safes, and performing room searches as well as installing devices. These entry specialists demonstrated that, if given sufficient time and resources, virtually any lock could be opened and any alarm system bypa.s.sed, though there were always limits on time and the amount of equipment that could be deployed at the job site.13 Regardless of how easy lock picking was represented to be on television and in the movies, the techniques of manipulating locks required skill and practice. At best, picking remained more art than science. One TSD tech remembered sitting at home in a European capital in the late 1960s working through the weekend to "get the feel" for opening a new brand of foreign lock. The first successful attempt required more than twelve hours, but, once he had acquired the feel, he could open the lock in less than five minutes.

Target sites were usually well protected, and the more valuable the information inside, the more layers of protection surrounded it-locks, secured doors, gates, windows, file cabinets, vaults, safes, and even the alarm systems. A lock specialist had to be proficient in dozens of mechanisms since locks in different parts of the world varied in type and method of operation. The techs discovered that German locks were particularly difficult compared to those in South Asia.

The grab bag of different locks and security architecture the techs found in countries from Ghana to Paraguay ranged from early colonial to state-of-the-art. With narrow time windows for covert installations, the techs had to know how many minutes were required to break through locks and security barriers and then restore and rearm the security systems. Information on all of these factors was obtained in a detailed preinstallation clandestine survey of the target conducted by techs and included in the operational proposal. Only after headquarters approval of the survey could an installation operation commence.

Under the best of circ.u.mstances, the CIA would obtain advance information that a Soviet intelligence officer planned to move into a new apartment or that the Chinese government was renting an office suite for a new trade mission. If possible, local support agents were recruited to rent or even buy office s.p.a.ce, apartments, houses, or property adjacent to target buildings.

Techs infrequently picked a lock, preferring to find other methods to gain entry. Picking took too long, the results were unpredictable, scratches from the picking on the mechanism or housing were detectable, and once a lock was picked open, it had to be picked closed at the end of the operation. Sometimes, when the location was known far enough in advance, the techs could wire the vacant property before its occupants moved in. Architectural s.p.a.ces, such as the attics of row house buildings, were particularly inviting, since their design offered a contiguous common s.p.a.ce over each unit. Once the tech gained access into the attic, he had un.o.bstructed movement to any top-floor unit in the row. Buildings might also provide common bas.e.m.e.nt areas with several outside entrances that enabled the tech team to avoid being seen coming and going through the front door. Given complete access and unlimited time to do the installation, the techs planted multiple microphones and transmitters as well as ran wires, precluding the need to pick any locks.

Depending on the relationship with security services of the host country, known collectively as "liaison," the CIA could a.s.sign the task of performing an entry to liaison. In many countries, the internal security service already had duplicate keys to all rooms in every major hotel, and master keys for apartment houses and important commercial buildings.

The techs also made duplicate keys. By the early 1960s, master keys for every popular hotel in more than one major European city hung neatly on a rack in the techs' shop in local stations. When possible, case officers would "beg, borrow, bribe, or steal" master or original keys. OTS had key-cutting machines at its bases and the traveling techs carried portable key-making devices that fit inside a briefcase. From the properly sized and configured blank, the techs could cut a duplicate within a few minutes.

For keys that were briefly available, OTS developed a portable key-impressioning kit. The kit consisted of a small mold with two halves filled with plastic modeling putty.14 The purloined key was placed on the putty and the two halves of the mold pressed together to capture a three-dimensional model of the key. Later, the tech could pour Wood's Metal into the mold and create an exact copy of the key. The purloined key was placed on the putty and the two halves of the mold pressed together to capture a three-dimensional model of the key. Later, the tech could pour Wood's Metal into the mold and create an exact copy of the key.15 The last choice was to attempt to pick an unknown lock. For this contingency, OTS issued leather lock pick kits that were small enough to fit in a jacket pocket, but provided the necessary tools for picking and raking the tumblers of many commonly found locks.16 These portable kits were most useful for opening luggage, desk drawers, and other smaller locks. These portable kits were most useful for opening luggage, desk drawers, and other smaller locks.

In the late 1960s, a TSD engineer developed a concept that mechanisms within key-operated locks could be measured and characterized remotely by marrying emerging ultrasonic measurement technology with an oscilloscope. Portable oscilloscopes had just been introduced and when combined with a small ultrasonic device, the techs would have a tool they could carry easily to the target, use to measure the lengths of the pins in a lock, and thereby acquire the precise data to make a key.

Once the engineer produced a prototype device that produced accurate calculations, OTS contractors refined the design for a field deployable unit. A year later, after the device proved itself by enabling several surrept.i.tious entries into previously inaccessible targets, Cord Meyer, the a.s.sociate Deputy Director for Plans, recognized the engineer with a special award that included a $5,000 stipend. In his presentation, Meyer said he could not mention what was acquired from the entries, but added, "This is the largest stipend the DDP has ever awarded for a technical development. This gadget is right out of the James Bond movies."

Not all entry operations involved breaking into rooms or safes. HTLINGUAL, for example, was the CIA's controversial Cold War program that intercepted and examined U.S. mail to and from the Soviet Union.17 Covert mail intercept required skills in "flaps and seals" to open and reseal envelopes, cartons, and packages thought to contain intelligence. "Surrept.i.tious opening" and surrept.i.tious entry shared the common objectives: to get inside a protected area and copy or steal the contents. Covert mail intercept required skills in "flaps and seals" to open and reseal envelopes, cartons, and packages thought to contain intelligence. "Surrept.i.tious opening" and surrept.i.tious entry shared the common objectives: to get inside a protected area and copy or steal the contents.18 The two primary methods for opening a sealed gummed flap, such as on an envelope, were "dry openings" and "steaming." The two primary methods for opening a sealed gummed flap, such as on an envelope, were "dry openings" and "steaming."19 From 1940 to 1973 the FBI, and later the CIA, conducted covert activities to open and photograph suspect mail in the United States. The earliest techniques of chamfering (mail opening) were taught to the FBI by a friendly Allied intelligence service during World War II. Information obtained from these programs was sanitized to protect against revealing sources and was disseminated to the intelligence agencies, the Attorney General, and the President of the United States.20 As the Cold War intensified, the CIA initiated its mail-opening project in New York to target mail from the Soviet Union. The HTLINGUAL operation was conducted by the Counterintelligence Staff and the Office of Security with TSD's a.s.sistance. Over a twenty-year period, more than 215,000 letters to and from the Soviet Union were opened and photographed in New York.

The New York mail project originated in 1952 with a proposal to scan exteriors of all letters to the Soviet Union and record the names and addresses of the correspondents as a means of identifying U.S. contacts of Soviet intelligence. The project expanded when James Angleton, chief of the Counterintelligence Staff, advised Richard Helms, then Acting Deputy Director for Plans, of a need to open and examine a selected portion of the letters. He advised Helms that there was no capability for "searching for secret writing and/or microdots, or to determine whether items have been previously opened, and to open items sealed with the more difficult and sophisticated adhesives."21 TSD set up a lab in New York in 1961 to test letters for secret-writing chemicals and to study Soviet censorship techniques. Because technical examination was time consuming, only a small percentage of the letters opened and photographed were actually tested. The Manhattan field office of the CIA's Office of Security handled most of the opening and photographing. Those who opened the mail attended a one-week course called "flaps and seals" conducted by TSD at CIA Headquarters.

The basic method of opening the mail was simple. First, the glue on the envelopes was softened by steam from a kettle, and with the aid of a narrow stick, the flap was pried open and the letter removed. One of the agents who opened the mail testified "you could do it with your own teapot at home." It took approximately five to fifteen seconds to open a letter. At one point, the CIA developed a type of "steam oven" capable of handling about one hundred letters simultaneously, but its performance was judged inadequate and the agents soon returned to the kettle-and-stick method.

The original letters, which had been opened, photographed, and possibly subjected to TSD examination, were resealed and returned the next morning to the airport for reinsertion into the mail stream. Translations and summaries of the letters' contents were disseminated within the Agency and to the FBI.22 As the sophistication of the technology and number of audio installations expanded, OTS developed an intense, year-long training program for the audio techs designed around the lessons and mistakes from the early years. In addition to learning the ins and outs of the technology itself, novice techs were taught the basics of building walls, mixing plaster, matching paint, restoring wallpaper, and making repairs after a device was implanted. They learned how to pick locks, make key impressions, cut keys from blanks, manipulate combination locks, and do electrical and telephone wiring.

"There was every kind of reconstruction. One of our instructors was a master plasterer who, before retiring, worked in the White House and the Capitol," recalled one tech. "We had a dedicated facility, an old food warehouse in Alexandria, Virginia, where we'd learn how to mix mortar and lay brick. It didn't matter if you had a college degree or not. If you wanted to be a tech, you took that training."

For a typical lesson, the master plasterer a.s.signed trainees to build a wall, plaster it over, then knock holes in it to simulate burying audio equipment, and replaster it. Then came the tough part. Not the least impressed by the tech's CIA affiliation, the plasterer would shine his flashlight on the gleaming wall, silently studying the work, then invite the tech to join him as he pointed out a ripple here and another there. No ripple was acceptable. "Nope, not good enough" were dreaded words. With those, the tech knocked down his wall and started over.

The course work on walls and plastering alone lasted a month, and then came paint matching, which included training with special paints that OTS formulated to be fast-drying and odorless.

Specialized soldering courses followed along with instruction about glues, adhesives, tapes, and fasteners that hold things together. The techs had to learn how to open and close all types of materials-fabrics, leather, wood, concrete, and masonry-in preparation for burying bugs in any concealment that might be in a target environment.

The techs received training to operate laser surveillance systems that, by projecting a laser through a window, could pick up audio from minute vibrations of the windowpane in the target room. Although these officers were audio specialists, overseas stations would not hesitate to ask their help in all other disciplines of OTS, so familiarization was provided in the full range of agent communication including microdot, secret writing, photography, and short-range electronic systems.

"They didn't skimp on training at Headquarters. It was thorough and hands-on. And in the field, every new officer got a mentor-a more senior audio officer-for their first tour," said the tech. "You didn't do anything on your own. You traveled with somebody else and they showed you the ropes. You drank beer together, stayed in the tech hotels, and, if you wanted to succeed, you listened to the lore, no matter how long that lasted late at night."

These field mentors also provided valuable unofficial training. Junior techs learned how to economize on s.p.a.ce while taking the necessary tools for jobs that were never completely predictable. One tech always carried four types of tape wrapped around a No. 2 lead pencil. Individual rolls of tape added weight, required s.p.a.ce, and contained far more tape than was ever needed on most jobs. Duct tape, double-sided tape, electrical tape, and copper foil tape were standards. Duct tape held devices in place while the epoxy dried, double-sided tape was used to stick components to walls or ceilings in temporary configurations for testing, electrical tape insulated and repaired wiring, and the copper foil with sticky backing made good practice or emergency test antennas. There was always room on the pencil to wind several loops of solder wire and utility wire as well.

Epoxy was the tech's best friend. The strength of small amounts and its brief curing time were important to completing a permanent installation quickly. It could repair broken housings, fill cracks and holes, and hold equipment suspended at awkward angles in almost any location. Epoxy could also teach a lifelong lesson. A tech on his first job and anxious to please his mentor, was asked to quickly prepare a batch of epoxy. Without seeing a container to hold the mixture, the tech squirted the two components into his palm and stirred them together. His palm became warm, then very warm, then very hot. Yet, the tech's professional pride and urgency to get the installation finished overrode the burning pain. He said nothing until the team returned to base and the medics treated him for a first-degree burn that permanently scarred his palm.

While techs were officially a "service" that responded to DO operational requirements, they frequently became involved in defining the requirement and developing the operational proposal for Headquarters. When operational proposals required technical detail, the writing responsibility fell to the tech. All technical operations required formal approval from both OTS, weighing in on the technical feasibility, and the operational division, which evaluated intelligence value and counterintelligence risk. Therefore, the field proposal issued under the local chief's signature became all-important. The COS always had the final word on the proposal, but the techs established informal codes to communicate differing opinions to Headquarters without crossing the chief.

One effective method for informing Headquarters of what the tech really thought involved the length of the proposal. In drafting the cable, concise and clear language signaled the tech's confidence in the operation, while a lengthy, excessively detailed proposal, with pros and cons, conveyed the tech's doubts and gave Headquarters plenty of information to "pick at" and challenge. In this way, when an operation was turned down, the chief directed his displeasure at Headquarters, not the tech.

Interactions between the audio techs and station management mirrored family relationships more than commercial customer-supplier exchanges. This was because OTS had no compet.i.tors for the services that the stations needed, but, more significantly, the case officers and techs shared a commitment to the common mission. Even so, disagreements between case officer and tech were part of the everyday picture.