A few days later the basic mission directive document appeared in its third revised edition; in turn it was superseded by a fourth edition and by a technical information summary. At the end of January, Robert Seamans and Abe Silverstein of Headquarters accepted Gilruth's STG recommendation to fly MA- 2. Before the middle of February preparations were complete. NASA had become convinced, but the Air Force was not sure MA-2 should fly yet. This was a hazardous and complex decision, shared by a number of people in Washington, at Langley, St. Louis, Los Angeles, and San Diego.48 On February 17, Seamans called Rhode at Convair, asking his technical judgment as to MA-2's chances for success with its "belly-band fix." Rhode replied that MA-2 was structurally  ready within acceptable wind velocities at launch.49 George M. Low reported to the new Administrator, James Webb, that MA-2 was scheduled for launch on February 21 at 8 a.m.:
Atlas 67-D will be the launch vehicle for this test. This is the last of the "thin-skinned" Atlases to be used in the Mercury program. It differs from the booster used in the MA-1 test in that the upper part of the Atlas has been strengthened by the addition of an 8-inch-wide stainless steel band. This band will markedly decrease the stresses of the weld located just below the adapter ring on top of the Atlas; the high stress region is shifted by about 8 inches, to a point where the allowable stresses are considerably higher. In addition to this strengthening of the top section of the Atlas, the bracing on the oxygen vent valve, which fits into the top of the Atlas tank, has been changed. The adapter between the Atlas and the capsule has also been stiffened.Webb and Seamans, pressed by Air Force worries over the technical, political, and public effects if MA-2 should fail, decided to trust the judgment of Rhode and Gilruth and to back NASA's commitment to accept all the blame if the worst should happen. Timely decisions by NASA had been required to permit deployment of the recovery forces to maintain the scheduled launch date.
* * * * * *
The Atlas will be cut off prematurely at a velocity of about 18,000 feet per second. The resulting trajectory will yield the most severe reentry conditions that could occur during an abort in an orbital launching.50
There was so much concern over the Atlas-Mercury compatibility problem that many people almost forgot the first of several first-order objectives for the capsule and its booster. That was to test the integrity of the structure, ablation shield, and afterbody shingles of the capsule for reentry from the most critical abort situation. A second first-order objective required the Atlas abort sensing and implementation system (ASIS) to be operated "closed-loop" on the Mercury-Atlas configuration for the first time. But because MA-2 had already been made into a Federal test case, with the President, Congress, and top echelons in the Pentagon and NASA Headquarters vitally interested in its outcome, the engineers at the working levels were more anxious than ever to make this one go. Its specific results were politically less important than its general appearance of success.
The preflight checkouts had ticked off nicely the last several days before capsule No. 6 was to undergo its ordeal. And spirits were rising with the Sun on the morning of February 21, 1961. The Mercury crew for launch operations was much the same as that for MR-2, but just as Atlas was an entirely different vehicle from the Redstone, so was its military/industrial launch operations crew quite different. From the factory in San Diego had come most of the senior engineers in the Mercury booster program office, including Philip E. Culbertson, Charles S. Ames, Howard Neumann, Joseph A. Moore, and Richard W. Keehn, as well as the same machinists, welders, and test supervisors who had made the "horse collar" work in bench and tunnel tests in California. At the Cape they  worked alongside the executive agent for Mercury-Atlas launchings, the 6555th Aerospace Test Wing of the Air Force, and with Thomas O'Malley and Calvin D. Fowler, who had the industrial responsibility for actual launch operations of the Atlas. The Air Force Ballistic Missile Division representatives, Lieutenant Colonel R. H. Brundin and Major C. L. Gandy, together with Aerospace engineers Bernhard Hohmann and Ernst R. Letsch, were also on hand, watching final preparations to make this "bird" fly. Their special concern with the design and implementation of the chief reliability component of the Atlas, namely the abort system or "ASIS," also brought Charles Wilson and J. W. Schaelchlin of Convair/ Astronautics, and D. R. White of Space Technology Laboratories, into the blockhouse of Launch Complex 14 on this special morning. John J. Williams was the Mercury-Atlas test conductor presiding there.
Engineers and workers at lower levels in the industrial and military hierarchy were beginning to call all these senior men "tigers" and to speak of them collectively as "tiger teams." They were the senior designers and the old-line specialists on Atlas subsystems who came out to the launch site to help the field engineers actually doing the work of final preparation for a launch.51 Walter Williams and Christopher Kraft, in the Mercury Control Center about three miles southwest of the beach-side launch pad, watched the lights turn green one by one as the gantry backed away and "loxing" commenced about 7:30 a.m. The weather was perfect at the Cape, but 1200 miles downrange in the recovery area there were scattered squalls, which delayed the launch for one hour. Outside the Control Center that day stood Gilruth, Low, and Major General Ritland of the Air Force Ballistic Missile Division, waiting and watching for the liftoff. Each had prepared press releases in his pocket making this shot a NASA "overload" test in case of failure.
MA-2 roared off its pad at 9:12 a.m., and for the next 2 minutes the tiger teams and the managers of Mercury hardly dared breathe. An audible sigh of relief spread through the Control Center and blockhouse about one minute after liftoff, when it was announced that the "horse-collared" booster had gone through max q intact and was accelerating. At that point, said Low, "Gilruth became a young man again." Telemetry verified "BECO" and the staging of the booster engines, escape tower separation, a good trajectory, capsule separation, capsule retrofire attitude, retrorocket firing, and retropackage jettison.52 Capsule entry attitude looked excellent at the time tracking and telemetry were lost, because of extreme range, about 9:22 a.m. Three minutes later, lookouts aboard the uprange destroyer Greene reported observing the reentry of both capsule No. 6 and Atlas booster No. 67-D.
The capsule passed directly overhead and was lost in the sun at 09:37. Reentry was clearly visible and the capsule could be seen ahead of the booster tankage. The capsule was not glowing but a distinct smoke trail was seen streaming behind it. The booster tankage was glowing with an intense white glow. Several fragments appeared to be traveling along with the tankage  and tumbling at a high rate. One of the ship's observers tracked the reentry on a gun mount and indicated a separation distance between the capsule and tankage of 50 miles when it passed overhead.53The landing ship dock Donner, almost at the center of the 20-by-40-mile elliptical dispersion area, also sighted the reentry but lost sight over the horizon northeastward before the parachute descent. Within 10 minutes, however, radio signals from the sarah beacon pinpointed the floating capsule's locations, and helicopters were dispatched to pick it up after only 24 minutes in the water. It was returned to the LSD less than one hour after launch.
MA-2 was a magnificent flight, "nominal in nearly every respect." This second mission followed a flight path essentially the same as that for MA-1. The Atlas-Mercury compatibility problem had been resolved, the sequence system for the booster/capsule combination had worked perfectly, and the tracking and real-time data transmission had given immediate and excellent impact prediction from the computers at Goddard to the control centers at the Cape and on Bermuda and to the recovery forces at sea. The capsule was in extremely good condition, its ablation heatshield being charred no worse than that for Big Joe, its afterbody shingles neither burned nor warped. The Space Task Group was pleasantly surprised to find the jettisoned antenna canister and to learn, even more surprisingly, that the "mousetrap" aerodynamic destabilizing flap had not, as expected, burned away.54
At a press conference later that day, Gilruth beamed as he announced that this was "a very successful test" that "gives us new confidence in the integrity of the system, although I would like to caution you all that there are still a number of critical tests that have to be made before we contemplate manned orbital flight." Asked if a man could have survived this flight, Gilruth said yes. When asked whether this flight also would aid the Mercury-Redstone program, Gilruth again gave an affirmative answer, stressing the identical nature of the capsule electrical, power, abort, and parachute systems. The Earth-fixed maximum velocity of the MA-2 capsule had been approximately 12,000 miles per hour, the highest velocity achieved by a Mercury launch since Big Joe had demonstrated the boiler plate model of the Mercury concept. As a cap stone for this happy occasion, Gilruth read a statement announcing that three out of the seven astronauts, namely "Glenn, Grissom and Shepard, in alphabetical order," had been selected to begin concentrated preparations for the initial manned Mercury space flights. The nominees had known about and been in training for their missions since January, but most Mercury engineers did not know who was assigned to which flight.55
47 Purser, "Notes on Capsule Review Board Meeting," Jan. 20, 1961. The conception of Mercury Mark II (or what was named Project Gemini almost a year later) was taking place at this time. See memo, Purser to STG Dir., "Atlas Modifications, Cost, and Scheduling," Jan. 17, 1961. Message, Hohmann and Robert H. Brundin to Philip E. Culbertson, re tests of the restraining band to reduce the discontinuity stresses in the M/A station 502 area, Jan. 16, 1961.
48 Seamans interview; "MA-2 Mission Directive," NASA Project Mercury working paper No. 140, June 24, 1960, rev. Aug. 11, 1960, Jan. 29, 1961, and Feb. 9, 1961; Donald T. Gregory, "Technical Information Summary of Mercury-Atlas Mission No. 2 (Capsule No. 6)," Feb. 10, 1961.
49 Rhode interview. Owing to airline engineers' strike, Rhode flew to the Cape via a routine Air Force logistics flight, arriving just in time to climb the gantry and personally inspect the "fix."
50 Webb interview; memo, George M. Low to Administrator, "Mercury-Atlas 2 Launch," Feb. 18, 1961; "Calculated Trajectory Data for MA-2," NASA Project Mercury working paper No. 163, Dec. 7, 1960.
51 "Proceedings of the Mercury-Atlas Booster Reliability Workshop," San Diego, July 12, 1963, passim.
52 Low, interview, Houston, Sept. 15, 1965; Ritland, interview, Andrews AFB, Dec. 30, 1964; Gilruth, interview, Houston, Mar. 18, 1964; P. E. Culbertson, comments, Aug. 16, 1965; Paul P. Haney, comments, Sept. 15, 1965; Purser, notes on MA-2 launch as relayed from Mercury Control Center, Feb. 21, 1961.
53 "Post Launch Report for Mercury-Atlas No. 2 (MA-2)," STG, March 13, 1961, 161. An unidentified ship, a tanker flying a hammer-and-sickle flag, but apparently without any unusual radar antennas, also was able to see the unusual reentry. Memo, Donald C. Cheatham to Assoc. Dir., "Russian Ship in MA-2 Primary Landing Area," March 8, 1961.
54 Memo, North to Administrator, "Preliminary MA-2 Flight Results," Feb. 23, 1961. Many NASA engineers and managers think of MA-2 as being "the day Mercury won its spurs" from the Air Force because in retrospect it represented the only potentially serious difference of opinion with the military services throughout the program; see Low comments.
55 "Press Conference; Mercury-Atlas No. 2," Cape Canaveral, Feb. 21, 1961. See also John H. Glenn, Jr., "We're Going Places No One Has Ever Traveled in a Craft No One's Flown," Life, L (Jan. 27, 1961); Loudon Wainwright, "Chosen Three for First Space Ride," Life, L (March 3, 1961). For the Atlas manufacturer's postflight analysis, see A. F. Leondis, "Project Mercury Structural Dynamic Analysis (Atlas 67D; MA-2)," Convair/Astronautics report No. AE 61-0743, Aug. 10, 1961.