Meanwhile NASA agents had completed an extensive survey of potential sites for the new development and operations installation for manned space projects of the future. At its Langley Air Force Base domicile, STG was busy planning for its expanding role in manned space exploration. Its personnel were weighing persistent rumors that the new Manned Spacecraft Center might be located in Texas, somewhere near the booming city of Houston.
The first objective of all this simultaneous activity was Mercury-Atlas 4, the fifth flight of an Atlas-launched spacecraft. This mission had been planned and replanned many times before the unsuccessful launch of MA-3 back in April 1961, and the failure of that mission directly affected the MAŚ4 plans. During the early months of 1960, MA-3 had been scheduled for a suborbital flight, with a crewman simulator aboard. First plans called for the Atlas booster to be held 150 feet per second below orbital velocity, with capsule separation occurring at the normal 100-mile-orbital-insertion altitude. Forty seconds after separation, retrofire was to have produced a landing beyond the Canary Islands and about 100 miles short of the African coast. And when this test was completed successfully, MA-4 was to repeat MA-3, but with a chimpanzee in the cockpit. Spacecraft No. 9 was to be specially fitted for the MA-4 flight.
Toward the end of 1960, however, Walter C. Williams advised the commanding officer of the recovery force, Destroyer Flotilla Four, that MA-4 would try for three orbits with a crewman simulator aboard and that the targeted launch date was April 1, 1961. But the MA-3 launch, still scheduled for a  sub-orbital flight with its "mechanical astronaut," slipped to April 25. While many Americans worried over the Soviet space coup represented by Yuri Gagarin's one-orbit flight on April 12, Robert R. Gilruth and Williams already had made the decision to change MA-3 to a one-orbit mission.1
April 25, 1961, came, but the day's recorded results were far from heartening. The MA-3 launch vehicle failed to program over into the proper trajectory; after 40 seconds of flight straight upward the Air Force range safety officer destroyed the Atlas booster. So it was necessary on MAŚ4 to strive for the same one-orbit objective and to delay still further the nominal three-orbit Mercury mission.
Meanwhile, for various reasons, production of the spacecraft and booster for MA-4 fell behind schedule. Atlas No. 88-D, allotted for MA-4, did not receive its factory rollout acceptance inspection until June 29-30, 1961, and it was July 15 before it was delivered to Cape Canaveral. And spacecraft No. 9 was not used, though originally planned. Instead No. 8 was fished from the Atlantic after its ill-fated flight in MA-3 and shipped back to McDonnell in St. Louis on April 27 for extensive overhauling. That meant cleaning, installing new insulation, replacing the external portion of the hydrogen peroxide control system, making spot-weld repairs in the large pressure bulkhead, and replacing the heatshield, antenna canister, escape tower, tower clamp ring, adapter, main clamp ring, and the inlet and outlet air snorkels. The overhauled spacecraft, redesignated 8-A, was returned to the Cape, but G. Merritt Preston's crew still had plenty of work. A leak had to be repaired in a reaction control system fuel tank; the environmental control system and the automatic stabilization and control system had to be reworked. A fairing to reduce launch vibration, like the one used on the Little Joe 5-B flight on April 28, 1961, and similar to that used on Virgil I. Grissom's suborbital mission in July, was added to the adapter clamp ring.2
Because of all this modification and overhaul, it was August 3 before the spacecraft for MAŚ4 was delivered to the pad and mated with the booster, supposedly to be launched on August 22. The day before the scheduled flight the Air Force's Space Systems Division in California called Cape Canaveral and reported that solder balls had been found in some transistors of the same brand that had been installed in the MAŚ4 booster. Coordination of this information among the various Mercury-Atlas teams at the Cape brought to light the fact that these types of transistors also had been used in the spacecraft. There was nothing left to do but postpone the launch and give both vehicles a thorough going-over to replace the defective transistors. On August 25 the spacecraft was returned to Hangar S, when it became apparent that this work might encompass several days. After these labors in the hangar, spacecraft 8-A was mated with the booster again on September 1. This time the engineers conducting the prelaunch checkouts found nothing wrong. Although 8-A was a secondhand capsule, its landing bag had not been installed, it had ports instead  of the new window, and the explosive egress hatch had been omitted, it still passed inspection.3
Besides the problem with the defective transistors, the Mercury-Atlas booster had been proceeding along the same tortuous route as the capsule toward flight qualification. By September, the Atlas had undergone so many changes that had to be integrated into launch vehicle No. 88-D, and experienced so many setbacks, that a successful orbital mission was necessary for the sake of NASA and national morale and to forestall any new attacks on the Atlas as the Mercury launch vehicle. The year in which the Soviets had orbited a man now was in its ninth month, yet the United States was still preparing to orbit a box full of instruments. The Mercury-Atlas flight record had produced only one completely successful launch - the MA-2 reentry heating test - out of four tries.
This was scarcely an enviable record. Many hours, days, and months had been spent by special committees and working groups in ferreting out the sources of trouble. The STG, Space Technology Laboratories, Convair, and Air Force engineers who had reviewed the failure of MA-1 had concluded that the forward end of the Atlas was not designed to withstand the flight dynamic loads fed through the adapter section, that the adapter was too flexible, and that stiffeners were needed. MA-2 had confirmed the controversial "fix" of the adapter section. MA-4 would be the second of the "thick-skin" Atlases. Reviewing the MA-3 abort, the engineers assumed that the programmer's failure to pitch the booster into a proper trajectory was due to a transient voltage. Also, some two years previously, another anomaly caused the Big Joe Atlas to fail to stage, and even in MA-2 there had been some propellant sloshing in the booster. To correct the programmer problem, Convair modified the autopilot controls to give the gimbaling engines of the Atlas a preventive counteraction capability. One objective of MA-4, therefore, was to assess this innovation.4 In September the NASA-Air Force- contractor engineering team that had been beset with Atlas problems for two years felt that the ICBM-turned-space-launcher was ready to do its part in Project Mercury. In the words of Scott H. Simpkinson, STG's liaison man at the Convair factory, "MA-4 just had to work."
Not only would a successful orbital mission on MA-4 provide the necessary data on the performance of systems and components, but the Mercury tracking network crews and Department of Defense recovery forces would receive valuable training for supporting a manned orbital circumnavigation by an American. Many components, elements, procedures, and flight maneuvers had to be watched and assessed before one of the "Mercury seven" could be committed to an orbital mission around Earth.
Of the manifold segments of an orbital flight, reentry was perhaps the most critical. As it dropped back into the heavy atmosphere, the capsule would be subjected to searing temperatures of about 2,000 to 3,000 degrees F for six or seven minutes, or about eight times longer than on the previous Mercury suborbital shots. Retrofire between Hawaii and Guaymas, Mexico, would bring about a  gradual descent over the North American continent. About 345 miles east of Savannah, the first contact with atmospheric resistance would begin, at an altitude of 55 miles. At this point the appearance of the .05-g light on the panel would telemeter a signal that reentry was coming up. Peak aerodynamic heating would come when the spacecraft had descended to an altitude of 37 miles and was traveling at 15,000 miles per hour. Braking would be dramatic. Between 46 and 12 miles high, traveling over a slant range of 460 miles, the capsule's air speed would be reduced from about 17,000 to 1,350 miles per hour. Aerodynamic stresses in this region would provide a severe test of the spacecraft's structural strength, particularly the heatshield and the afterbody shingles.
Perhaps the second most critical segment of the orbital mission would come during the powered phase of the flight. The Space Task Group, supported by the DOD and industry, would also monitor carefully the vibration levels to ascertain if they would be tolerable for an astronaut. Even more important as the capsule was rocketed toward orbit was a reliable escape system, to wrench the capsule clear if the launch vehicle failed to perform. Also it was necessary to judge the ability of the Atlas to release the spacecraft, to evaluate the abort sensing and implementation system, to determine if the launch vehicle could withstand the aerodynamic loads of max q, and to demonstrate the capability of the Mercury network to perform its intended flight-control and data-collection functions.5 If all went well, MA-4 would provide data proving the validity of years of engineering calculations.
MA-4 would be launched from complex 14 at the Cape on a true azimuth heading of 72.51 degrees east of north. Following engine ignition, after being held to the pad for three seconds to ensure smooth combustion, the Atlas booster engines would propel the spacecraft within two minutes to a speed of about 6,500 miles per hour and an altitude of 35 miles over a downrange distance of 45 miles. The sustainer engine would continue to burn. A gradual pitch program would begin to tilt the Atlas toward the sea about 20 seconds after liftoff. Seconds after booster engine cutoff (called "BECO" by the various Mercury-Atlas working teams at the Cape), or at about 41 miles' altitude and a slant range of 56 miles from the pad, the launch vehicle programmer would trigger a greater pitch-over maneuver to put the Mercury-Atlas combination on a course parallel to Earth's surface. At this time the escape tower would be jettisoned. After capsule separation, orbital insertion would occur about 498 miles downrange from the pad at an altitude of about 100 miles. The nominal inertial velocity at this point was supposed to be 25,695 feet per second, increased to 25,719 feet per second by the ignition of the posigrade rockets, which separated the spacecraft from the booster. Within 50 seconds, the spacecraft should have drifted some 790 feet from the booster. The Atlas, rather than falling away, would trail the orbiting spacecraft around Earth at an altitude of about 100 miles, and should complete each circle about once every 90 minutes for an estimated three days.6
Instrumentation affixed to the spacecraft would provide data from nearly  every conceivable point about the capsule. Noise levels in the vicinity where an astronaut's head would rest would be measured and recorded on magnetic tape. Excess vibration, a problem during early Mercury-Redstone flights, would be monitored closely by seven strategically placed sensors, mostly in the area where capsule and adapter joined. To determine what radiation dosages a pilot would encounter, four standard and two special film packs would be carried. The standard packs were placed on the sides and at the top and bottom of the couch. Carrying a heavier emulsion, the two extra packs would measure the radiation spectrum - the range of all kinds of radiation to which the capsule would be exposed - as well as penetration levels. Flight data other than radiation would be transmitted by two separate telemetry links, each providing essentially the same information.
The flight would be well covered photographically. Located on the left side of the capsule cabin was the instrument panel camera, which would start operating at liftoff, provide about 20,000 frames of panel information during the mission, and cease five seconds after impact. Placed near the right-side port, the Earth-sky camera was loaded for about 600 frames of pictorial data, which would be exhausted somewhere over the Indian Ocean. A third camera, affixed to the periscope, was loaded with about 10,000 frames of film for the mission. This camera would provide especially useful information on the spacecraft's orbital attitude reference to Earth at points where landmarks were recognizable.
Five recorders aboard the spacecraft would tape most of the mission data. Three were seven-track systems to record all telemetry outputs, vibration levels, noise, and shingle strain. The two others were single-track recorders, to be operated in tandem and used to check the reliability of the tracking network communications system.7
Plans for spacecraft operations after the powered phase were essentially the same as those for the suborbital flights, only on a much larger scale. Retrofire was scheduled at 1 hour, 29 minutes, and 4 seconds after launch, with the three rockets firing at five-second intervals in order: top-left, bottom, top-right.8 Recovery plans for orbital missions were considerably more complicated than they had been for the suborbital flights, since many more contingency areas, including abort and overshoot, had to be considered. Besides the nominal landing area off the coast of Bermuda, five secondary landing areas were selected. Providing that the launch was nominal and proceeded according to the preflight calculated trajectory, the abort recovery areas were spaced as follows: Area A began about 13 miles from the launch pad and continued along the track for 2,200 miles. For the first 550 miles the coverage extended 30 miles to each side of the track. This area covered the first 72 seconds after launch, or through booster staging. The remainder of Area A, accounting for the period up to 298 seconds after launch, narrowed to about 15 miles on either side of the track. Areas B and C were small elliptical blips on the track, 4 and 8 degrees of longitude beyond A. These were designated for a possible abort at 298 or at 301 seconds, respectively.  The third contingency site, Area D, was a longer ellipse (20 by 122 miles) beginning about 7 degrees of longitude past C. At this point the "go/no go" flight decision would be made. The last, Area E, an ellipse 24 by 231 miles along the track, covered aborts up to 304 seconds after the go/no go decision.9
The MA-4 capsule also was fitted with a number of aids to assist the DOD forces in their recovery task. Two one-pound sofar bombs, one set to eject upon main parachute deployment and the other set to detonate at 4000 feet of hydrostatic pressure if the spacecraft sank, were carried. A flashing light with a life of about 24 hours was set to activate upon impact. Fluorescein dye, ejected at touchdown, would be visible for about six hours. Navy recovery forces were asked to attempt the recovery of the drogue and main chutes and the spacecraft antenna canister. Balsa wood blocks and Styrofoam had been attached to these components for flotation.10
As the launch date of the Mercury-Atlas 4 combination neared, weather problems began to threaten this attempt to orbit a "mechanical astronaut." Not one but two hurricanes thrashed the Mercury tracking areas. "Carla" raked the Corpus Christi tracking station, while "Debbie" moved in a northerly direction on the day before the launch, menacing and causing the ships to get rather a "rough ride" in the prime recovery zone. The equipment at the Texas site withstood the storm without damage. The STG-Air Force-Navy recovery planners at the Cape felt that Weather Bureau support predictions had given them a sufficient margin of safety in the Atlantic to allow the mission to proceed.11
1 Message, Walter C. Williams to Cdr., DesFlotFour, Dec. 8, 1960; "Project Mercury Status Report No. 9 for Period Ending Jan. 31, 1961," 40, 41, 43; Paul E. Purser, log for Robert R. Gilruth, April 17, 1961; "Project Mercury Status Report No. 10 for Period Ending April 30, 1961," 33. For a complete discussion of the MA-3 mission, see pp. 335-337. Counting MA-4 as the fifth Mercury-Atlas combination launched includes Big Joe.
2 "Status Report No. 10," 33; James M. Grimwood, Project Mercury, A Chronology, NASA SP-4001 (Washington, 1963), 214; "Project Mercury Postlaunch Report for Mercury-Atlas Mission 4 (MA-4, Capsule 8A)," NASA Project Mercury working paper No. 213, Nov. 10, 1961.
3 Ibid.; message, NASA Hq. to STG, Aug. 25, 1961; memo, Morton Schler to Flight Dir., "Report on Test 1254," Oct. 3, 1961; Walter C. Williams, interview, Houston, Aug. 23, 1965; Bernhard A. Hohmann, interview, Houston, Sept. 16, 1965; memo, P. I. Harr, GD/A, to Members of Astronautics Reliability Policy Committee, "Minutes of Special 28 August 1961 Meeting on Transistors," Aug. 29, 1961.
4 "Project Mercury Status Report No. 11 for Period Ending July 31, 1961," 11, 12. As finally configured, Atlas No. 88-D had modifications in the sustainer engine liquid oxygen duct to improve performance, and the first four panels of the upper liquid oxygen tank area were of "thick-skin" materials designed to support high aerodynamic loads. "Postlaunch Report for MA-4." Moreover, a three second hold-down was programmed for MA-4. Tests conducted by the Rocketdyne Division, North American Aviation, indicated that a two-second hold-down was adequate for Mercury-modified Atlas engines. So for flights beginning with MA-5, STG officials planned to institute the two-second procedure.
5 NASA News Release 61-182, "Mercury-Atlas 4," Aug. 20, 1961; "Project Mercury Technical Information Summary of Mercury-Atlas Mission No. 4/8A (Capsule No. 8A)," NASA/STG, July 21, 1961. The Mercury ground tracking communications network at this time had 140,000 actual circuit miles, consisting of 100,000 miles of the teletype circuits, 35,000 of telephone circuits, and 5000 of high-speed telemetry circuits.
6 "Pre-release Draft on Launch Vehicle (MA-4)," STG, undated; "Mercury-Atlas 4"; "Project Mercury Calculated Preflight Trajectory Data for Mercury-Atlas Mission No. 4 (MA-4) (Capsule No. 8A, Atlas No. 88-D)," NASA Project Mercury working paper No. 204, Aug. 2, 1961. The nominal launch trajectory was computed by the Aerospace Corp. and Space Technology Labs. under the technical direction of the Space Task Group. The abort sensing and implementation system continued monitoring during the entire powered phase. If trouble developed, the clamp-ring released and posigrade rockets fired to separate the spacecraft, and the recovery gear was ready for action. Provided the powered flight phase went well, by about five minutes after launch the radio-inertial guidance system would be measuring speed, altitude, and flight course. If those factors anticipated a successful orbital insertion, the ground guidance computer, in operation shortly after booster engine cutoff, would initiate the shut-down command to the sustainer engine.
7 "Project Mercury Mission Directive for Mercury-Atlas No. 4 (Capsule No. 8A)," NASA Project Mercury working paper No. 203, July 28, 1961; "Project Mercury Addendum Data Report for Mercury-Atlas Mission 4 (MA 4, Capsule 8A)," NASA Project Mercury working paper No. 218, Nov. 29, 1961.
8 "Preflight Trajectory Data for MAŚ4."
9 Letter, Williams, STG, to Cdr., DesFlotFour, June 8, 1961, with enclosure, "Project Mercury, Mercury-Atlas No. 4 Recovery Requirements." The recovery forces consisted of 8 destroyers, 12 aircraft, a landing ship dock, and a utility vessel. Williams also stipulated secondary-zone recovery requirements and called for a nine-hour watch. In plotting contingency recovery areas, STG's planners had to allow for trajectory alteration resulting from the added thrust of escape rockets or retrofire in an abort.
10 Williams letter; "Mission Directive for MA-4." William T. Lauten, Jr., said of the sofar bombs that during the program they jokingly referred to one as the sofar bomb and to the other, which was set to detonate several thousand feet beneath the waves, as the "so-long bomb."
11 "Storms Hit 2 Mercury Trackers," Newport News Times-Herald, Sept. 12, 1961; "Postlaunch Report for MA-4"; Williams,interview.