The pair then donned each other's mask and clothing, with padding as needed to match each other's body shape. They directed six volunteers (who were fans of the show) to approach them, starting from 90 feet and ending at 5 feet, while using an unrelated task to mask the nature of the test. Three worked with "Jamie", three with "Adam", and all six quickly detected the deception as soon as the imposter spoke. Six more volunteers then took part in a test in which the disguised hosts did not speak; none of them suspected anything until they came face-to-face with the imposter.

Grant and Kari, both thoroughly familiar with Jamie's appearance and behavior, took part in a similar test, with subtle changes in "Jamie's" wardrobe as distractions. Neither of them identified him as a fake until they were 5 feet away, though Kari expressed some doubt at 70 feet. Adam and Jamie classified the myth as plausible, depending on the circumstances under which the deception is carried out.

The team added steel plates to the rails in an attempt to stop the bullets and force them to transfer their energy to the merry-go-round. After making this change, they tried another deer slug and managed to turn it by 1.5 inches; a shot from an Armalite AR-50 rifle turned it 8 inches and knocked one of the plates off. At this point, the team declared the myth busted and set about replicating the results, with Tory redesigning the merry-go-round's bearing system to reduce friction and Kari and Grant attaching more bullet-stopping plates. They were finally able to keep the device turning with steady streams of 12-gauge deer slugs and .416 Barrett rifle fire.

For full-scale testing, they built a replica of Jason Bourne's apartment, as seen in the film, and set up the toaster, magazine, and gas supply. Their first attempt, which exactly replicated the scene from the film, had no effect whatsoever, with the magazine completely burning away well before the gas concentration could reach explosive levels. On their second trial, they placed a firestarter log in the middle of the room and allowed the gas to gradually build up to explosive levels. The mixture did eventually ignite, creating a small explosion that partly blew out a wall and gave way to an intense fire that destroyed most of the furniture. Adam and Jamie declared the myth busted and made one last attempt, using enough methane to achieve a 9% concentration (the center of the flammability range) and a set of fans and diffuser hoses to mix the gas and air thoroughly, before lighting the gas mix with a specialized device. This time, they were able to get one wall to blow out and set the apartment on fire.

Grant and Tory visited a NASA research center and placed the rig in a tunnel designed to replicate the high wind speeds and cold temperatures typical of airplane cruising altitudes. When the toilet was dumped all at once, the liquid quickly atomized in the wind, leaving only a thin film to freeze on the fuselage. However, a slow leak allowed the ice to build up into a large mass that did not break loose until the rig "descended" to 12,000 feet.

To determine the ability of such a chunk to survive a fall to earth, Kari and a skydiver flew up to that altitude, threw a 35-pound ice block out of the plane, and jumped out after it as Grant and Tory tracked it from the ground. The block did not suffer any significant melting or damage during the fall or impact, prompting the team to declare the myth confirmed but unlikely due to the number of safeguards that would need to fail simultaneously to achieve the result.

After studying film footage of a basilisk running on water, Adam and Jamie decided to build separate footwear rigs to duplicate its leg movements. They called in Jessica Fortunato, a trained acrobat and gymnast, and had her try the rigs: Adam's with hinged foot platforms and a long tail, and Jamie's with concave foot cups and an outrigger frame held in front. She was unable to stay above water, whether unaided or using either of the rigs.

At this point, Adam and Jamie declared the myth busted and built a submersible bridge to replicate the results. After camouflaging it and setting up the camera at a particular angle, both were able to run across the surface until they went off the end of the bridge.

In order to eliminate these wall effects, Adam and Jamie built a 4-by-16-foot bubbler to place at the bottom of a swimming pool. After they added weights to keep the rig from floating up, Adam tried to swim across the pool and back through the bubbles. The trip proved difficult at 25% power and impossible at 100%. Adam and Jamie classified the myth as plausible, but for a different reason from the one expected – water currents holding the swimmer at the surface, rather than a density reduction causing him/her to sink.

Declaring the myth busted, the team did some small-scale tests with different explosives and placement techniques. A TNT charge drilled into the trunk shredded it at the blast point, while an ANFO charge laid in grooves cut along the wood grain caused some degree of splitting.

In one last full-scale test, the team chose a 100-foot Ponderosa pine and fitted it with 25 pounds of ANFO, laid in grooves cut near the core. The resulting blast tore the tree into hundreds of pieces.

Adam and Jamie built several full-scale torpedoes, took them to a lake, and set up a target ship 800 ft (244 m) from shore. Two accuracy tests – first with the full propellant charge (as a rocket motor), then with a reduced one – caused the torpedo to go airborne and miss the target completely. For the next two trials on payload delivery, they used a tether to guide the torpedo, but were unable to score a hit even after they moved the ship to within 200 ft (61 m) of shore. Finally, they removed the tether and reduced the rocket power further, allowing them to hit the ship and set it on fire with the incendiary payload. They classified the myth as plausible, since they could find no record of the device being used in combat.

After visiting a local winery to find out about what could make a wine bottle blow its cork, the Build Team decided to investigate wine/bottle type and temperature in the shop. They placed each bottle on a burner and measured the distance each cork flew. Chilling the bottles increased the distance due to the higher pressure rise caused by the larger temperature change. The best performers of the still and sparkling wines were Riesling (30 ft (9 m)) and champagne (50 ft (15 m)), respectively, with the difference due to the high pressure under which champagne is bottled.

For a full-scale test, the team loaded a semi trailer with 1,000 bottles of champagne and some Riesling, then set fire to it. All three agreed that the sound of the exploding corks resembled machine guns or firecrackers; however, no corks flew farther than 50 feet. Based on this result, the team declared the myth busted, then set up a Gatling gun-like rig to rotate the bottles over a burner and launch their corks straight ahead. Grant, dressed in a fire suit, stood in front of the rig and took several direct hits.

With smaller sails and higher prop speeds, the cart rolled backward once air began to spill around the sail edges. The combination of a large sail and an elevated fan did result in the cart moving forward; Grant theorized that this was caused by the sail "reflecting" a portion of the prop's thrust backward (in fact, the air hitting the sail flows outwards, towards the top, bottom and sides, but, due to the sail's curvature, the flow is angled slightly backwards). A larger-scale test, with a small jet engine mounted on a skateboard and using a fire-resistant sail, gave the same effect.

Finally, the team set up a full-scale test on a lake, using a swamp boat with a 40-horsepower fan. The boat reached a speed of 20 mph (32 km/h) with the fan facing backward and the sail not in use. After the fan was reversed and the sail was hoisted, the boat traveled forward at 3 mph (5 km/h) once the team fine-tuned the fan's aim.

All tests used a basic square sail design.

Jamie then built a new device from a short pipe the same diameter as the hubcap, with the free end sharpened into blades. This design tore up Adam's bodywork, popped one tire, and pulled the other one off its rim, leaving the car sitting on its chassis.

Next, Adam mounted a gun and a full ammunition hopper on his hood and chased Jamie, who drove a target car with the rear windshield removed. He was able to riddle the rear end and hit Jamie's head and seat a few times with the fixed mount; in the aimed-mount test, he got so many hits in such a short time that Jamie called an early end to the run.

After re-checking their own footage, the team heard a sound that might have been a spinning bullet. Once winter came, they visited Caples Lake in the Sierra Nevada and dug out a target area on its frozen surface for further work. Testing gave no results until they aimed at the center of their target site, allowing for a backward ricochet instead of a forward one. The team found the bullets from these last tests to be spinning in place on their side, leading them to declare the myth confirmed.

They then watched Dave fire standard blank cartridges from various distances and found that they could not see his muzzle flash at all past 200 yd (183 m). When he switched to Hollywood-style blanks with much heavier gunpowder loads, they could easily see the flash out to 1,200 yd (1,097 m). Finally, they set up a blank-firing rifle at 200 yards, wired to a timer and paintball gun; when one man pulled the trigger, a paintball would be fired directly at the other's chest after 231 ms. Neither was able to dodge any shots until the rifle was moved to 500 yd (457 m) (600 ms delay) and loaded with Hollywood blanks. Adam and Jamie declared the myth busted, since an actual sniper would take precautions to ensure that the target would not see the muzzle flash (including a flash suppressor).

For a final test at terminal velocity, roughly 120 miles per hour, Tory threw two pig carcasses out of a helicopter at 600 ft (183 m), after which they were X-rayed to determine injuries. The pavement drop resulted in 17 fractures, a shattered pelvis, and a decapitation, while the water drop yielded seven fractures and a broken neck. Since no water landing produced the same level of impact force or injury as a fall from the same height onto pavement, the team declared the myth busted.

For full-scale testing, the team outfitted a go-kart to match the mass and top swimming speed of a typical sailfish and hooked their line to it from a star-drag reel. Kari drove it at 68 mph, while Tory applied drag and Grant measured the temperature. The first two runs gave a peak of 530 °F and smoked without any fire, even after Tory used larger amounts of flammable lubricant on the second run.

Declaring the myth busted, the team brought in a professional motorcycle racer and his sidecar rig to achieve higher speeds. Tory changed the lubricant out for even more flammable engine starting fluid. Runs at 140 and 180 mph gave temperatures over 700 °F, with the second of these causing the line to melt without burning. Finally, Tory set the reel on fire by exposing it to an open flame.

They set up six mirrors to bounce light back and forth down the length of the shop, using a spotlight as the source. Polished metal mirrors gave 1.13 lux, but caused the beam to spread out after only a few reflections. When modern glass mirrors were used instead, the light level registered at 0.49 lux and the beam stayed focused longer. Adam and Jamie realized that the light needed to scatter in order to illuminate the room.

At Treasure Island in San Francisco Bay, they set up a hangar area as a full-scale tomb and brought in six glass mirrors and a 7,000-watt spotlight. After adjusting the mirrors, they measured a light level of 2.3 lux and could easily see their way around. Once the sky cleared, they reflected sunlight into the tomb and found a peak of 2.5 lux until the sun's position shifted, throwing off the mirrors’ alignment. Finally, Jamie stood in the light beam, which scattered in the air when it hit his white shirt and gave 8.6 lux. The need to keep adjusting the mirrors, and the unlikely prospect of finding several of them ready to use after thousands of years underground, led them to declare the myth "plausible but ridiculous."

Next, Grant and Tory tried to sandwich Kari's car from either side and managed to stop her, though they began to spin out somewhat. The two men then built car-stoppers: side paddles to attach to two rescue cars’ front bumpers and maneuver to block in the runaway (Grant), and a hood-mounted spear to hook the rescue and runaway cars together so the rescuer can slow them both down from behind (Tory). Both rigs worked without endangering either Kari or her rescuer(s).

For full-scale tests, they used a mast fitted with sensors at depths of 6 in (15 cm) and 2.5, 15 and 25 ft (0.8, 4.6 and 7.6 m), with a fifth at the surface. The first two depths were intended to represent the depth at which a swimmer's torso would be submerged if he were floating or treading water. A 10 lb (5 kg) charge of TNT was set off at a depth of 15 ft (5 m) and a set distance from the sensors (30, 70 or 150 ft (9, 21 or 46 m)).

Using a pressure threshold of 87 psi (corresponding to a 50% mortality rate from the shock wave), Adam and Jamie found that a floating person would survive at 30 ft (9 m), while greater depths led to pressures likely to cause death. At longer distances, the readings fell below 87 psi. Based on the 30-foot results, they declared the myth confirmed. Jamie attributed these results to the fact that much of the energy from the underwater pressure wave dissipated into transverse waves when it reached the surface.

Using an armor sample (1⁄2 in (13 mm) paper vs. 1⁄32 in (1 mm) steel) placed over a block of clay, they tested resistance to blunt force, swords, and arrows. The paper did as well as steel in the sword and arrow tests, failing only the blunt-force test, so the team built a full suit of paper armor to match against a period-accurate steel counterpart.

Each team member ran one of three timed courses in both armor types to evaluate speed, endurance, and agility; paper outperformed steel in all three. Finally, they attacked the suits with arrows, swords, and two firearms — an 18th-century flintlock pistol and a 19th-century .45 revolver. The armor types resisted every attack except the .45, leading the team to classify the myth as plausible. They pointed out, though, that the paper armor could quickly begin to disintegrate if it got wet or took repeated blows (which happened during the full-scale tests).

At this point, Adam and Jamie decided to try to improve a motorcycle's aerodynamic performance so that it would use less fuel and give lower emissions. They built a teardrop-shaped shell to cover the entire vehicle, with a tubular steel frame and heat-shrink plastic sheeting. Jamie then drove simulated city and highway courses at Naval Air Station Alameda on a modern motorcycle, with the shell on and off. The shell gave the best fuel efficiency and lowest CO2 emissions, but did not perform better than the cars on the other three pollutants. They classified the myth as busted, and Jamie attributed this result to the fact that car pollution technology has advanced faster than that for motorcycles.

With help from Institute personnel, they built a rig to launch an RPG directly toward a remote-controlled gun of the same caliber as that used in the film (a Smith & Wesson Model 460 revolver). After a failed first test, they were able to hit the RPG with a bullet and see the former explode in midair. Separate shots of the two weapons on an 80-foot test range, the same distance as the film scene, led them to calculate that the bullet and RPG would collide 16 feet in front of the RPG shooter. However, the particular rounds used for these tests had fuze systems that would not arm until they had flown roughly 60 feet.

The team set up two dummies 80 feet apart, outfitted with pork-belly abdomens and foil burst gauges to evaluate injuries. Assuming that the film RPG might have had a defective fuze, the team set up a target 16 feet in front of the RPG dummy, backed the launcher up to ensure a detonation, and fired. Both dummies survived without injury, though molten metal flew through the air just to each side of the handgun dummy and his survival was chalked up to sheer dumb luck. The need for a bad fuze, the forward direction of the explosion, and the fact that the intended target survived only through luck led the team to call the myth busted.

To simulate miniature wrecking balls, Adam and Jamie filled five 6 in (152 mm) hollow steel balls with plaster that had properties similar to concrete. This version gave only 63% efficiency, but Jamie suggested putting a steel plate through the center of each ball to reduce energy transfer into the concrete. They then built five 2,000 lb (907 kg) balls by cutting hollow naval buoys in half, sandwiching a steel plate in each, and filling the space with rebar and concrete. When the rig was assembled at Mare Island Naval Shipyard and hung from an I-beam frame, it gave very poor efficiency and quickly stopped swinging, even after each ball was hung from two cables rather than one as shown in the video. Adam and Jamie declared the myth busted, noting the increased potential for energy losses at large scales, and later learned that the original video had indeed, been faked confirming their initial thoughts.

With these two still in the car, birds were brought in and allowed to perch on the hood. Two pigeons (1 lb (454 g) each), two hawks (2.5 lb (1.1 kg) each), and an eagle owl (7 lb (3 kg)) were tried separately; none were able to do more than wobble the car slightly. Grant commented that although a weight near the front bumper gave a large amount of leverage, the car itself had enough weight to counteract the effect.

The team located a model helicopter and modified it to weigh 20 lb (9.1 kg), the same as a California condor (the largest flying bird in North America). When this too failed to tip the car, the team classified the myth as busted and set out to find out how much weight it would take. The car only tipped after Kari put 80 raw, 1.5 lb (0.7 kg) game hens onto the hood and Tory added a raw, 20 lb (9 kg) remote-controlled turkey for a total of 140 lb (64 kg). Tory additionally commented that for the feat to be accomplished they would need a pterodactyl to land on the car's hood.

This myth was revisited in Bird Balance Limo in 2012.

For a real-world situation, they then decided to investigate the ability of a person to navigate a straight course if lost in the woods. With no landmarks or destination in view, they tried to follow separate headings for 30 minutes and succeeded by using the sun's position to stay on track. However, with buckets on their heads to simulate reduced visibility at night or in a snowstorm, Adam did poorly while Jamie stayed on track by carefully pacing around obstacles, drawing on his wilderness survival experience. Finally, they attached themselves to opposite ends of a long ladder with hip belts, thinking that each could feel the other's veering and correct it, but failed the open-field walking test again.

Declaring the myth busted at this point, the team traveled to the New Mexico Institute of Mining and Technology to use the on-site rocket sled apparatus. After setting off a 100 lb (45 kg) load of explosive to determine the appearance of the actual blast, they set a target car on the track, with the same size load, and fitted a pickup truck's front end onto the sled. Enough rocket motors were attached to accelerate the sled to a top speed of 300 mph (483 km/h), but the impact only disintegrated the target car without triggering a detonation. A cloud of undetonated Tannerite was visibly sprayed in all directions by the force of the impact.

The tests suggest that Tannerite is difficult to detonate unintentionally, the myth being that it was detonated unintentionally. Other binary explosives, however, may not necessarily be as safe.

For sensitivity tests, they rigged a full, 7 lb (3.2 kg) cooking pot to fall on a piece of burning C-4 from 3 ft (91 cm), then set up a robotic leg in a combat boot to stomp on it. Neither of these trials produced an explosion, so they dropped a 90 lb (41 kg) anvil on the material, without success. Shooting at the C-4 with different .308 rifle cartridges – standard, tracer, incendiary – failed to trigger an explosion as well. Finally, Adam and Jamie set up a thermite charge to drip molten material at 4,500 °F (2,482 °C) onto the C-4, thinking that extreme heat might make it sensitive enough to detonate. This test failed as well, and Jamie noted that C-4 was formulated to be stable unless set off by a blasting cap. He and Adam judged the first half of the myth as confirmed, but the second half as busted. Jamie added that stomping on burning C-4 is still a bad idea, but more because it burns at a very high temperature and tends to stick to the bottom of shoes, rather than any risk of it exploding.

After a few refinements to his design, he was able to take the head off and have it retained within the device. Kari and Grant set up a combat situation, involving one moving enemy dummy and a second that could pop up from behind a screen. Tory hit the pop-up enemy with a glancing blow and knocked it down, but could not land his device on the moving one. Based on these results, the team decided that the flying guillotine was plausible as an assassination weapon, but not in combat.

They next built a 40 ft (12 m) sewer in a trench using concrete culvert sections, installed three manholes, and placed rolls of chain-link fencing in it to simulate debris. After burying it under 3 ft (91 cm) of dirt, they pumped in the methane and set it off. The resulting explosion threw the covers 150 ft (46 m) skyward and slightly shifted the dirt over the sewer. High-speed footage showed fire coming from only one manhole, indicating that a deflagration had taken place rather than a detonation. Judging the myth confirmed at this point, Adam and Jamie noted that a detonation might have sent the covers even higher.

Fatal Flashlight

On the bomb range, they set up a charge of C-4 and a high-speed camera to measure the delay between triggering and detonation. A control run with an ambient-temperature bomb gave a delay of 3.3 milliseconds, while pouring the LN2 directly onto the charge increased it to 15 milliseconds. When the LN2 was poured over both the charge and the battery powering the trigger, the bomb would not detonate at all until the entire rig had warmed up to ambient temperature, 15 minutes later. Adam and Jamie commented that the technique was actually more effective in real life than in the film; the characters could have simply stood up and walked out of the bathroom.

To test the third part of the myth, they built a bathroom set with a heavy cast-iron tub as seen in the film, put Buster in the tub under a blanket, and fitted the tub with sensors to measure the pressure inside and outside. When a 1 kg charge of C-4 was set off under the toilet, the entire set was destroyed except for the tub. The peak pressure outside the tub was a lethal 180 psi (1.2 MPa), while the inside sensor recorded 8 psi (55 kPa), meaning that a person would survive but have some chance of suffering hearing damage. Adam and Jamie declared the myth confirmed.

At Tracy Municipal Airport, the Build Team met with a 9-member acrobatic flight club to test out flying formations: V, wingtip to wingtip, single file, and a control run with no formation. A tight V reduced fuel consumption by 3-5% compared to the control, while the other two formations either gave no reduction or increased the fuel usage. The team declared the myth confirmed at this point, but noted that the close formation could pose a danger to the pilots. A second run, with the planes spaced farther apart, gave a fuel savings for only Tory due to the difficulty of staying in the lead planes’ vortices. Nevertheless, the team commented that even a small fuel savings could lead to a sizable financial advantage at a large scale.