Two spacecraft are flying toward each other at nearly eight kilometers per second. Each is sealed, self-sufficient, and built never to open outward into space—except for a spacewalk. And now they need not merely touch, but fuse into a single whole, so that a person can pass from one to the other as from one room to the next. The problem is that the rooms were designed on different continents, breathe different air, and even the locks on their doors turn in opposite directions.
When engineers in Houston and Podlipki began seriously considering a joint flight in the early seventies, the first thing that stopped them was not politics or the difference in rockets. It was air. Apollo flew in an atmosphere of pure oxygen at a pressure of 5 psi—roughly one-third of Earth’s. The decision had been born back in the Mercury program: pure oxygen was simpler from an engineering standpoint, lighter in mass, and allowed for thinner cabin walls. The Apollo 1 tragedy in 1967 forced a revision of launch procedures, but in orbital flight the oxygen atmosphere remained. Soyuz lived differently—an oxygen-nitrogen mixture at 14.7 psi, essentially Earth air. The Soviet school chose that path deliberately: a familiar environment reduced physiological risks, while the spacecraft’s stronger hull could withstand full pressure.
A difference of nearly threefold is no small thing. If a cosmonaut from Soyuz simply stepped into the Apollo cabin, his blood would begin behaving like soda opened at altitude: dissolved nitrogen would turn into bubbles. Decompression sickness—the bends, familiar to divers. To avoid it during an ordinary transition from a nitrogen environment into pure oxygen requires hours of preliminary oxygen breathing—prebreathing—which flushes nitrogen from the tissues. Hours in a spacesuit, hours of immobility, hours of lost flight time. For a short joint mission, that would have been fatal: the crews would have spent half the shared flight simply preparing to cross over.
The solution came in a compromise that demanded concessions from both sides. Soviet engineers agreed to lower the pressure in Soyuz to roughly 10 psi for the duration of the joint flight. The Americans, in turn, raised the pressure in their cabin and diluted the pure oxygen. Oxygen content was limited to forty percent—beyond that lay the zone of unacceptable fire risk, and after Apollo 1 that boundary was treated with religious seriousness. To operate in the altered atmosphere, the Soviet side modified its onboard life-support system, based on potassium superoxide—a substance that simultaneously absorbs carbon dioxide and releases oxygen. In addition, Soyuz was fitted with an emergency repressurization system in case the seal at the docking interface proved imperfect.
Apollo’s standard docking system—probe-and-drogue—worked on the principle of a pin and a cone. One spacecraft extended an active probe, the other presented a receiving cone, the drogue. The system was reliable, proven on lunar missions, but fundamentally asymmetrical: the roles of “active” and “passive” were assigned in advance and forever. Soviet Soyuz spacecraft used their own docking system, also asymmetrical, also with rigidly fixed roles.
For the joint flight, something fundamentally different was needed—a mechanism in which either side could be both “pin” and “cone” at once. An androgynous docking interface, to use a biological analogy, is a hermaphrodite. Both halves are identical. Each can capture the other; each can be captured. Elegant in concept, but far more difficult in engineering: it had to provide initial capture, alignment, drawing together, and sealing—and do it all in a symmetrical structure where neither side was “in the lead” by default.
The system was designed by Soviet specialists from NPO Energia and designated APAS-75—the Androgynous Peripheral Attach System. Peripheral because capture and sealing occurred at the edge, around the perimeter of the docking ring, rather than in the center, as in probe-and-drogue. That freed the central section for crew passage—the hatch was wide enough for a person in a flight suit. The design used guide petals and latches arranged in a ring. As the spacecraft approached, the petals of one interface entered the slots of the other, providing rough alignment; then the latches locked the connection, and the draw mechanisms pulled the rings together until the sealing gaskets were fully compressed. What still amazes me is that both sides agreed on a design that gave neither of them an advantage—at the height of an era when every bolt had a nationality.
Even with matching docking interfaces, one problem remained: where exactly were they supposed to put this interface on Apollo? The Apollo command module had its own standard docking mechanism—the very probe-and-drogue through which it connected to the lunar module. Rebuilding the entire command module would have been madness: too expensive, too risky, too late. The answer came in an additional element—the Docking Module, built by North American Rockwell at its plant in Downey, California.
The Docking Module was a cylinder about three meters long, clamped between two worlds. On one side: the standard probe-and-drogue for connecting to Apollo. On the other: APAS-75 for docking with Soyuz. But it was not merely an adapter, not merely a coupling between two incompatible connectors. It functioned as an airlock. Inside it, air pressure and composition could be changed, creating an intermediate environment between Apollo’s oxygen atmosphere and Soyuz’s compromise mixture. An astronaut entered the module from the Apollo side, closed the hatch behind him, the pressure and gas composition changed, and only then did the hatch toward Soyuz open. In some ways, the procedure resembled a submarine lock: you could not open both doors at once, because beyond them lay different worlds.
In January 1975, assembly was already under way at the Kennedy Space Center on the Saturn IB launch vehicle, the last of its family, which was to carry this entire construction into orbit. The docking module sat beneath the payload shroud, above the second stage, and was extracted in orbit by the command module—much as Apollo had extracted the lunar module from beneath the shroud on lunar missions. Familiar choreography, but with an entirely new partner.
Any docking mechanism is a device that has to work the first time. There may be no second chance in orbit: a miss, a jammed latch, an incompletely compressed seal—and the mission turns into an expensive flyby. The APAS-75 test program was therefore multistage and, unusually for the Cold War, bilateral.
In Moscow, tests were conducted on scale models—two-fifths full size. These models made it possible to test the kinematics of capture: how the petals found the slots at different approach angles, with lateral displacement, with angular misalignment. A scale model is not a toy; it preserves geometric proportions and makes it possible to catch problems that look harmless on a drawing but turn into binding or skewing in metal. In Downey, at the North American Rockwell plant, they separately tested the interface seals—the very rubber rings that were expected to hold vacuum on one side and pressure on the other. The seal of a docking interface is not just a gasket; it is the boundary between life and death, and the tolerances for its compression are measured in fractions of a millimeter.
Dynamic tests took place on the DDTS rig at the Johnson Space Center (JSC) in Houston. Here they simulated the moment of contact itself: impact, rebound, damping, recapture. Two spacecraft weighing tons, closing at relative speeds of fractions of a meter per second—this is not a gentle touch. The mechanism had to absorb the energy of impact without throwing the spacecraft apart or damaging the seals. Joint equipment checks, when Soviet and American components were physically connected for the first time, became a genre of engineering diplomacy all their own: each side wanted to make sure the other half of the mechanism behaved exactly as the documentation promised.
Apollo’s prime crew consisted of Thomas P. Stafford (commander), Vance D. Brand (command module pilot), and Donald K. “Deke” Slayton (docking module pilot). Slayton is a story of his own: one of Mercury’s original “Magnificent Seven,” grounded in 1962 because of an irregular heartbeat and restored to flight status only a decade later. ASTP became his first and only flight into space. The backup crew—Alan L. Bean, Ronald E. Evans, and Jack R. Lousma—trained with the same intensity, knowing that historically the odds of a backup crew flying in the Apollo program were slim.
On the Soviet side, the commander of Soyuz-19 was Alexei Leonov—the man who first walked in space in 1965 and nearly died trying to get back into his spacecraft when his inflated spacesuit would not fit through the hatch. Beside him was flight engineer Valery Kubasov. The USSR’s program leadership was handled by Boris N. Petrov and Konstantin D. Bushuyev; on the American side, ASTP was managed by Glynn S. Lunney. For years, these people met in working sessions, coordinating not only hardware but terminology: what to call the stages of docking, which commands to give, in what language, in what order. Stafford studied Russian, Leonov English, and each tried to speak the other’s language; Leonov eventually joked that three languages were heard in orbit—Russian, English, and “Oklahoman.”
I like the detail that Slayton—a man who had waited thirteen years for his flight—was given the position of “docking module pilot,” meaning the master of that very airlock cylinder through which the whole handshake passed. The man who had stood the longest before a closed door ended up responsible for the door in the most literal sense.
On July 17, 1975, Apollo and Soyuz-19 closed in on each other in orbit. The docking took place that same evening, and over the Elbe—the river where Soviet and American troops met in 1945—the hatches opened and the handshake took place. The coincidence was so symbolic that people still say the moment must have been scheduled; Leonov later insisted that it had been “some kind of providence.” Stafford piloted Apollo, bringing the docking module with APAS-75 toward the corresponding interface on Soyuz. The petals entered the slots, the latches engaged, the draw mechanisms compressed the seals. Airtightness confirmed.
Then the transfer procedure began. Pressure in the docking module was equalized with the pressure in Soyuz. The hatches opened. Stafford extended his hand through the Soyuz hatch, and Leonov shook it. That moment was broadcast to the entire world, but what remained off camera was the fact that between those two hands lay three sealed volumes with three different atmospheres, two types of docking mechanisms, thousands of hours of testing, and several years of negotiations over which way a locking lever should turn.
The crews crossed over to each other several times during the two days of joint flight. They exchanged souvenirs, conducted joint experiments, ate together. Then the spacecraft undocked, docked again—this time Soyuz played the active role, confirming the system’s androgyny—and undocked for good. Soyuz-19 landed on July 21, Apollo splashed down on July 24, 1975—the last landing of an Apollo capsule, the end of an era.
APAS-75 did not become a museum piece. Its descendant, APAS-89, was used on the Mir station for dockings with American shuttles in the nineties. The principle of androgynous peripheral capture, perfected for one symbolic handshake, proved so successful that it became the foundation of docking systems still operating today. The international docking system IDS, used on the ISS, was built under the noticeable influence of the same school: in form and spirit, its mechanism echoes APAS-95—a direct descendant of that very design Soviet engineers created for a mission lasting less than a week.
Put bluntly, the entire history of ASTP is the story of how opening a door between two sealed worlds proved harder than building each of those worlds separately. Pressure, gas composition, lock mechanics, the shape of seals, safety protocols, the language of commands, even the direction in which handles turned—all had to be agreed upon. None of those questions was the main one. All of them were the main one at once. And perhaps most astonishing of all—both sides really did reach an agreement, even though in 1975 the Cold War was far from over and the level of mutual distrust was measured in megatons.
The docking module had a separate fate of its own: it was jettisoned a day before the return, freeing the command module, and for several days it flew alone in orbit—an empty three-meter cylinder, no longer needed by anyone—until it burned up in the atmosphere on August 2. The only artifact that had physically been “between” was destroyed routinely, according to plan, as befits a transitional space: first it ceased to be needed, then it ceased to exist.