DAYS of discussion during July left the Stewart Committee divided about whether the Orbiter using the Army's Redstone missile or the NRL satellite scheme based on the Viking rocket would best answer IGY purposes. The Air Force submitted a plan but only as a proposal that might be adopted if neither alternative were acceptable. The Air Force paper, containing nineteen pages of text and twenty-one of drawings and charts, was an elaborate dissertation on the scientific information attainable from a 150-pound satellite if launched by the Atlas rocket then under development. Lieutenant Colonel R. F. Lang of the Air Research and Development Command in making the presentation explained that his service had chosen 150 pounds as a minimum payload; the Atlas would be powerful enough to put hundreds of pounds, "even thousands of pounds." into orbit. The rocket held every promise of success and of growth possibilities, would use proven components and only two stages, would possess a low g factor, and would offer the advantage of simplicity of design. The Air Force, moreover, could supply full logistical support, and a preliminary estimate put the over-all cost at $16,350,000. But, Lang admitted, even if a minimum satellite were made part of the Atlas program, interference with the ICBM development would be inescapable because of competition for facilities, propulsion sources, and skilled personnel. Furthermore, the first launchings of Atlas-B were scheduled for January, February, and March 1958, by which time four launching stands and four assembly buildings would be available. At best that date would leave an uncomfortably narrow time margin for the IGY, and were the Air Force to take on responsibility for a satellite, flight tests of the rocket might have to be further postponed.1 Homer Stewart remarked in 1963 that such caution had in actuality proved needless, but, at the time, the Air Force had every reason to fear delays in its ICBM developments. Under the circumstances the committee shelved the Atlas-B scheme.2

The Army-ONR Orbiter proposal, dated l July 1955, offered a design considerably modified from the original scheme von Braun had tendered as a "Minimum Satellite Vehicle" in September 1954. As an alternative to the clusters of Loki rockets, von Braun and his associates had adopted a suggestion of the Jet Propulsion Laboratory to use Sergeant solid-fuel rockets reduced in size to power the second, third, and fourth stages. Either configuration should be satisfactory. Other changes recommended by JPL and incorporated in the new version consisted of refinements in the engineering of the upper stages. Equally important, tacitly recognizing the "million-in-one chance" of locating a small body in the vast expanses of space by relying solely on optical equipment, the Orbiter team had added a provision for electronic tracking: it might employ either NRL's "light, low-powered transmitter" and large radar directional antennas or else a device, upon which the Army's Diamond Ordnance Fuze Laboratory was working. which would use Lincoln Laboratory radar and dipole antenna modified with a central coil. The Orbiter scheme still called initially for a satellite of only five pounds, although the launcher would be powerful enough to carry a much heavier payload. Estimated costs ran to $17,700,000, of which $6,400,000 would be for eight Redstone missiles for the first stage.3 However inadequate this and the still smaller Air Force figure would look by 1958, both were less unrealistic than the $9,734,500 Kaplan mentioned to Waterman as sufficient for ten satellites and ten launchers.4 In the spring and summer of 1955 only guesses were possible about expenditures.

The NRL presentation made no attempt to estimate total costs. While the proposal was explicit about the scientific advantages its originators envisaged, and in explaining the mathematical formulas upon which they based their calculations, data about the launching vehicle were more general than detailed. The plan called for a three-stage carrier capped by a twenty-inch-long instrumented cone as the satellite, but the text did not describe the mechanism for spinning and firing the second and third stages in flight or the device for separating the satellite from the third stage. After discussing the importance of analyzing the flight path in order to minimize possible errors in projection, the written proposal offered two alternative vehicle configurations. One comprised a M-l0 Viking first stage and two solid-propellant stages, the other a liquid-liquid-solid combination consisting of the two-stage M-l5 rocket and a solid-propellant third stage.5

The M-l0 was a modification of the Navy's sounding rocket which had reached to over one hundred miles altitude in seven flights between 1950 and June 1954. Stripped of fins and equipped with the General Electric Company's Hermes power plant, the new version. NRL asserted, could attain an altitude of 216 miles with a tangential velocity of 5,060 feet a second. Much smaller than the Redstone. the M-l0 would have a four-foot diameter, a forty-foot length, and a dry weight of 2,250 pounds. The Glenn L. Martin Company had spent two years in studying the design and had prepared detailed drawings of what would be the smallest available vehicle that could serve as a first stage for a satellite. The small size made the M-10 easily transportable by motor vehicle and reduced to a minimum the amount of logistic support needed. The Atlantic Research Corporation had designed the two solid-propellant stages, and both the Martin Company and NRL had scrutinized the plans and considered them sufficiently "conservative" to permit developing and testing within the required time. This combination would put a forty-pound instrumented payload into orbit at a perigee of 216 miles unless an error in projection of the solid-propellant stages occurred; an 0.88° error would reduce perigee to 150 miles. A postcutoff control system similar to that already in use in the Viking would tilt the carrier over to a predetermined angle to enter the circular orbit.

Adoption of the second configuration would enable the vehicle to carry a forty-pound instrumented satellite into an elliptical orbit with a 303-mile perigee. The M-15 was the M-l0 Viking with an Aerobee-Hi liquid-propellant second stage. As early as 1949 the Army's Bumper-Wac, a liquid-fueled second-stage rocket fired by a V-2 in flight, had established the practicability of that type of propulsion system. Under NRL scientific direction, the Aerojet General Corporation had spent two years in developing the Aerobee-Hi; the first flight was scheduled for August 1955. Since the angular precision required to produce an orbit varies inversely with the altitude of projection, one major advantage this combination promised was the higher altitude of projection. At 200 miles, analysis showed, the tolerance would be 0.67°; so that a very precise orienting mechanism would be necessary; at 300 miles the tolerance would be 2.0° and thus permit use of simple and more reliable orienting equipment. With the more efficient flight path made possible by the Aerobee-Hi, the two liquid stages would burn in succession; the second stage would coast to orbital altitude, be oriented, and then spin and fire the third stage. A disadvantage of the M-l5 configuration, however, lay in its needing more time than would the M-l0 for design and test of the second-stage controls. Whereas the latter would be ready to launch a first satellite two years after work began on the program, the M-l5 would have to have an additional six months. "Both configurations," the statement noted, "could be carried forward simultaneously, since the M-l0, the most expensive stage, is common to both. The liquid-liquid-solid combination appears to offer more in the way of growth potential."

As for contractors to produce the vehicle, the authors drew attention to the Glenn L. Martin Company's nine years of experience in the design and production of Vikings and the company's "many design and performance studies of satellite systems and components." The Aerojet General Corporation was ready to supply the Aerobee-Hi for the second liquid-propellant stage, and the Atlantic Research Corporation had submitted designs for ARC solid-propellant second- and third-stage rockets. For "system contractor, the agency which would have primary responsibility for the entire program, the recommendation was that the Naval Research Laboratory take charge because of its long-term interest in and familiarity with atmospheric phenomena, radio and radar propagation, optics, radio astronomy, and upper-air research using rockets. The Army Map Service should handle the geodetic measurements and provide the optical tracking instruments, while the National Committee for the IGY should coordinate the geophysical measurements. NRL was ready to supply radio tracking equipment called "Minitrack" which John Mengel and Roger Easton of the Laboratory had devised.

The description of the scientific devices for the satellite and the means of relaying data to recording stations on earth were the most impressive features of the NRL memorandum. Miniaturization, today a commonplace of technology, was a novelty in 1955. The Laboratory's proposal, however, hinged on it. A satellite casing weighing eight pounds would carry miniaturized instruments weighing ten pounds for accumulating scientific data, tiny batteries weighing twelve pounds, Minitrack equipment weighing two pounds and consisting of a miniature electron-beam vacuum tube and a crystal-controlled radio receiver with a hearing-aid-type amplifier to respond to instructions sent by powerful radars at the ground stations, and two pounds of telemetering equipment to transmit information back to earth, bringing the total weight to thirty-four pounds. Although solar cells instead of conventional batteries would save weight, the NRL team discarded the solar-cell source because its dependability was not as yet established. In actuality, solar cells were later incorporated into Vanguard I. Investigation of the possibility that the temperature of the satellite might interfere with the proper operation of the instruments indicated that the temperature change would not exceed 10° C as the satellite moved from the day to the night side of the earth and that equilibrium temperature could be kept to 10° C by such simple means as coating half the casing with thick lead-base white paint and leaving the other half of the surface unpolished aluminum.

In an appendix recommending scientific experiments to undertake in a first satellite the NRL team listed five conditions it had imposed upon itself: (l) the information should be significant and obtainable in no other way; (2) the instrumentation required should be of proven design and (3) it should weigh less than forty pounds; (4) the experimental data must be communicable by telemetry; and (5) the experiments must be applicable to a satellite with an equatorial orbit. On the basis of those criteria NRL scientists proposed for the two first experiments to use instruments that could determine the distribution of hydrogen in outer space and could show whether or not a ring current encircles the earth in an equatorial plane beyond the ionosphere.

Two Lyman-alpha detectors in the satellite would serve for the first experiment, one detector to measure the intensity of the ultraviolet radiation emitted by atomic hydrogen from interstellar space, the other less sensitive detector to measure radiation coming directly from the sun. The two together were expected not only to reveal the density of neutral and ionic hydrogen in space within ninety million miles of the earth but also to furnish data on the motions, densities, and sizes of streams of particles ejected from the sun. Physicists and astronomers suspected that the streams affect cosmic ray intensity, the aurora, and magnetic storms. For the second experiment a highly sensitive magnetometer, an instrument for measuring magnetic elements, would serve. Placed in the satellite, it could detect the presence of a ring current and gauge the intensity of the earth's magnetic field, provided the altitude of the satellite were known within 1.7 miles. The Minitrack system could probably meet that proviso. But to interpret the results of the hydrogen density experiment and to correct the roll and tumble rates for the magnetometer, the attitude of the satellite had to be known for any given moment of flight. The method outlined for obtaining that information was to measure by means of a miniature electron-beam vacuum tube the angle between an axis in the satellite and the earth's magnetic field. The device would be actuated from the ground, allowed a minute for warm-up, and would then transmit for three minutes when the satellite was above the ground receiving station. A second signal from a ground-based transmitter or else a timer could turn the instrumentation off and thus lengthen its useful life.6

The appendix of the memorandum wound up with the statement that the instrumentation was already available and would need only slight modification to fit into the container, the "pot," as it came to be called. A total of $110,000 should cover the costs, including about $10,000 for instrumentation. Construction of the telemetering ground station would come to about $30,000, and salaries of the men at the receiving station would add about $ 100 a day. The actuating station would be the most expensive item-apart, that is, from the launcher. The powerful "classified" radar would cost "a few million dollars," unless it were possible to borrow the unit. The Orbiter cost analysis had slid over these elements.

Accompanying the NRL text with its careful exposition and mathematical calculations were graphs and a photograph of "The Double-Axis Phase-Comparison Angle-Tracking Unit," a forerunner of the Minitrack. The Minitrack in fact would be essentially identical with the unit pictured except for operating frequencies and antenna configurations. A phase comparison angle tracking system in effect determines the angle of arrival of a radio signal by measuring the difference in length between two radio paths from the signal source in the satellite to each of two receiving antennas located on the ground at a known distance apart. The system would provide three coordinates of satellite position and three vectors of satellite velocity, plus accurate time of transit at each ground station once during each pass of the satellite. Since radio transmissions from a subminiature transmitter in the satellite would be feasible at any time of day and under all normal weather conditions excepting severe local thunderstorms, the Minitrack would furnish complete tracking and position information throughout the life of the equipment. A diagram showing the layout of a single Minitrack ground station from the antennas to the recorder clarified the textual explanation.7

For scientists eager to get reliable data on the ionosphere and interstellar space beyond, the NRL presentation could hardly fail to have a strong appeal. The Orbiter and the Air Force propositions also discussed the scientific benefits they could offer, but neither was as specific about how its measuring and tracking schemes were to work. Furthermore, the five-pound satellite of Orbiter was manifestly a far less useful research tool than NRL's bigger, more elaborately instrumented payload. The relative merits of the launchers were another matter. The Stewart Committee, several of whose members were interested in basic research as well as in engineering technology, faced a difficult choice, its difficulty doubtless heightened after the presidential announcement and the attendant publicity. The meeting on 3 August, at which the committee prepared its formal recommendations, took place without Professor McMath of the University of Michigan Observatory. Illness kept him from attending, a fateful circumstance if, as rumored, he later declared that he would have voted with the minority.8 Had he been on hand to do so, the minority might easily have become the majority, for when three men endorsed the NRL proposal and two the Orbiter, the remaining two, explaining that they were not guided missile experts, chose to go along with the numerical majority; if the split had been three to three, the fence-sitters might have landed on the other side. Homer Stewart admitted privately in 1960 that some of the ad hoc Group disliked the idea of using a booster that was a modification of a Nazi Vengeance missile developed by German engineers; an American IGY satellite launcher should be an American product. But, Stewart added, that line of reasoning had had little bearing on the majority's decision.9

The report sent to Secretary Quarles on 4 August laid down first the general conditions that would have to be observed to attain success no matter which design was selected. These conditions included a satellite kept below fifty pounds and a perigee of the orbit of not more than one hundred fifty to two hundred miles above the earth. Whatever the launching system adopted, some development work would be necessary and so would present the risk of interference with military programs. Yet if properly carried out, the project would produce long-term military as well as scientific benefits. Only clear, undivided administrative responsibility could fulfill the objective. "Great caution is imperative to insure that existing techniques, existing contractors, group skills and facilities be used." To forestall diversion of resources, top-level control would be essential: "otherwise additional and unnecessary delays will be inevitable." The cost would probably total about $20.000,000, and would be more were "full advantage…[not taken] of existing programs, facilities, and reasonable logistical support." The over-all undertaking should embrace two phases, the first realizable before the end of 1958, the second a long-term scheme which could ensure a higher orbit and a payload of as much as a ton. For the second phase an ICBM booster, such as the Air Force Atlas-B, should be used and the Air Force put in charge of the program, but. in view of the uncertainty about whether the ICBM development could keep up to schedule, the Stewart Committee regarded specific recommendations about Phase II a matter beyond its competence. For the IGY program the choice narrowed down to Orbiter or the modified Viking and a payload.

In arriving at recommendations for the first phase, the committee, after agreeing upon the practicability of putting up "anything" during the IGY, considered nine factors: (l) the minimum payload and altitude that could provide "something useful"; (2) duration of the orbit; (3) tracking requirements; (4) the growth potential of the equipment, meaning its chances of leading to more sophisticated scientific and military devices; (5) maximum use of available facilities and skills; (6) minimum delay to military projects; (7) maximum scientific utility; (8) broad national interest; and (9) over-all economy during a five-year period. The group tacitly equated "broad national interest" with success in a launching achieved without interference with ballistic missile programs and at a cost the economy could stand and over which the public-or at least Congress-would not boggle. The relative dollar price of one proposal as over against the other and the effects on the American economy did not occupy the committee long. The estimates submitted were, after all, only estimates and those given in the NRL memorandum covered only the Minitrack system. Still, part of the committee believed the bigger, heavier Orbiter vehicle would cost considerably more than the Viking combination, and excessively high costs for a first satellite would imperil the chances of a continuing program.10 Nor did duration of orbit net much committee attention. Ironically enough, long orbiting life would prove to be one of Vanguard I's notable features; not only is it still circling the earth and expected to remain in orbit for at least two centuries, but its telemetry system transmitted radio signals for over seven years. In August of 1955 the remaining six factors posed more important and more controversial questions.

The choice of the committee majority was the M-l5 with an Aerobee-Hi liquid-propellant second stage, even though that configuration would require six months more to develop than the M-10 with two solid-propellant stages. The proponents of the M-l5 argued that, despite its small size and less than 30,000-pound thrust, it offered better performance and more reserve margin than Redstone with its 75,000-pound thrust. Doubts about the efficiency of the latter as a satellite booster in fact took George Clement to Huntsville on 28 July to look for himself. He concluded that the Redstone was too heavy for the purpose and, even if stripped of some of its boilerplate, would probably still be relatively unsatisfactory.11 Instead of Orbiter's four stages, the second and third consisting of multiple clusters and deemed proportionately less reliable than single-rocket stages, the three-stage design of the M-l5 gave the NRL proposal another advantage. The smaller launcher would also require less logistic support and thus prove more economical for continued satellite use after the IGY was over. The modifications needed for the Viking appeared to be well within engineering capabilities and, because of the probable availability of a new General Electric rocket engine, would avoid any interference with weapon projects. Unless an ICBM rocket motor replaced that of the Redstone in the near future, Orbiter, in the opinion of three committee members, was less likely to succeed than the Viking combination.

Stewart and Furnas took exception to that interpretation. In their view the Redstone booster had more power and flexibility and fewer development problems; as part of an active weapon program it was already under test and had range facilities at its disposal which would minimize interference with military programs. Viking was a greater risk, if only because the margins of error allowed at each stage were so narrow as to be "at the limit of current engineering knowledge"; to correct malfunctions would take precious time. Since the IGY was to end in December 1958, the main question was, in Stewart's words, "what could be done with things that already had some development history and could work at a small level without starting from scratch?"12 The critics did not emphasize the fact that the Navy would have to beg for space at the Florida missile center, if, as NRL proposed, the Viking were to be projected eastward in order to take advantage of the earth's rotation to heighten the vehicle's velocity. Certainly neither NRL nor the Stewart Committee majority when making its decision envisaged fully the delays and general wear-and-tear that would spring from the Navy's having to negotiate for a launching pad and blockhouse, and then fight for testing time at Cape Canaveral, not to mention the design changes in the rocket which the safety rules at the Air Force base required.

Viking's opponents did, however, point out that the adequacy of the second- and third-stage fuels was not firmly established. Increased performance of the second-stage Aerobee-Hi by use of unsymmetrical dimethylhydrazine, UDMH for short, had not yet been put to the test, although in the Air Force's Bomarc motor the Aerojet Corporation had mixed UDMH with JP-4, the standard kerosene fuel for jets and had run one test of a twenty-five percent UDMH and seventy-five percent analine-furfural mixture. Even the time schedule for fuel tests was uncertain. The design of an attitude stabilization system was on hand, but the attitude control system would have to undergo careful testing with the new fuel. The third-stage solid fuel called for ammonium perchlorate dispersed and fused in a polyvinyl chloride matrix which the Atlantic Research Corporation had tested in small motors but not in large. To get an efficient structural design in an end-burning configuration with a long burning time might take longer than the project could afford. For the spin-stabilized third stage, moreover, NRL had not presented any analysis of how to forestall low frequency oscillations; those might prevent attainment of an accurate orbit.

To these objections defenders replied that equal uncertainty applied to Redstone's fuel, and the UDMH system looked like "a very straightforward engineering procedure." Inasmuch as the Glenn L. Martin Company had a feasible approach with commercially available components for an attitude control system, the development presented no greater difficulty than that for a conventional missile autopilot system. Spin stabilization had to be worked out for Orbiter also; the problem was as hard for one as for the other. Although Redstone testing facilities already existed at Patrick Air Force Base, using them for testing the Orbiter satellite vehicle would interfere with ballistic missile programs as much as Viking satellite tests would. The military and scientific projects should be divorced from each other as completely as possible. The Viking was a research rocket, not a weapon. Without saying so in so many words, supporters of the NRL plan apparently felt that under the aegis of a research laboratory manned by civilians, the satellite venture would avoid much of the military flavor likely to permeate it were it directed by the Army Ballistic Missile Agency. NRL had had long experience in atmospheric research, had a deserved reputation for meeting time schedules, and at the moment had no high-priority work afoot.

More significant, the Redstone satellite plan offered relatively little growth potential for future space exploration, whereas NRL's opened up a variety of possibilities, a committee appraisal that was to prove sound. While that consideration doubtless loomed larger to some members than to others, no one dismissed it lightly. On the other hand, the very innovations in the Viking-based design presented greater risks of delay than did the more orthodox features of Orbiter.13

Unanimity prevailed on one point: the superiority of the NRL tracking system and indeed all the NRL satellite instrumentation. What the committee would have liked to recommend, Clifford Furnas said later, was the use of the Army's rocket and the Navy's "pot" of instruments. The hitch about that procedure was the intensity of service rivalries. No military service was willing to give "personnel, money-or even information, at times-to a project for which some other branch would get the most credit. We finally decided," wrote Furnas. "that breaking the space barrier would be an easier task than breaking the interservice barrier." That conclusion reached, it was a "toss-up between the Army and the Navy plans."14

The committee was at pains to suggest needed improvements to both plans. Orbiter ought to include better satellite instrumentation; eventually the Redstone motor should be replaced with a liquid-oxygen and gasoline motor, and, in the interest of reliability, possibly the second stage should also use a liquid propellant. NRL should schedule a far larger number of tests of components and rocket stages preliminary to attempting a satellite launch and should consider substitution of a Sergeant-type, solid-propellant third-stage rocket for the Arcite-type proposed. Those recommendations, if reflecting committee doubts about the adequacy of both plans as they stood, implied that modifications might make either one satisfactory. Both the Redstone and the NRL teams were ready to adopt the committee's suggestions. Quarles, faced with a five to two decision in favor of the NRL proposal, put the matter up to his Policy Council. The council, after listening to the arguments of its Army and Navy members, voted to postpone a final choice for two weeks.15

During that interval, in response to vigorous protests from Major General Leslie Simon of the Army Ordnance Corps, who insisted that misinterpretation of facts had prejudiced the case for Orbiter, Quarles asked the Stewart Committee to reexamine the Navy-Redstone plan. General Simon's memorandum, dated 15 August 1955, asserted:

The substitution of the 135,000-pound North American rocket engine for the current 75,000-pound engine in the Redstone missile is a less complicated operation than the design of a new Viking missile. There is greater assurance the Redstone with 135,000-pound motor will be available within a 2-year period. Actually the first orbital flight of this improved Redstone motor can take place in August 1957. Using three scaled Sergeant high speed stages, a payload of 162 pounds can be placed in an orbit with a perigee of 218 miles. Payload can be traded for excess velocity. There is sufficient excess velocity to place a 100-pound payload on the moon.16

In actuality the launcher thus described was never built.17 Simon continued:

The development problems confronting the Viking development make it obvious that the probability of success within the IGY is low. This conclusion is reinforced by looking at the development times of major missile programs already completed. Such programs are rarely completed on the originally predicted dates and require 5 years at the minimum and usually run for approximately 8 years.

The improved Redstone 75,000-pound performance permits improved payloads at orbitable altitudes. The following table was computed with 900 feet per second excess velocity and with existing propellants:

Perigee altitude (miles) Payload (pound,)

300 6

216 18

The first orbital flight for this configuration can be scheduled for January 1957 if an immediate approval is granted. Since this is the date by which the U.S.S.R. may well be ready to launch, U.S. prestige dictates that every effort should be made to launch the first U.S. satellite at that time. Although this time scale is dependent on Sergeant or Loki clusters, the engineering feasibility has been approved by four competent agencies.

The satellite missile does not interfere with the Redstone missile program because the program is in process of being turned over to the Chrysler Corp, and because the designers and planners are completing their work with the Redstone missile. A new program is therefore necessary for adequate utilization of the talent available. If a new and challenging project so not soon placed at Redstone Arsenal, the loss of key personnel will jeopardize the successful completion of the Redstone missile project. Therefore the scientific satellite program will strengthen rather than weaken the Redstone missile project.

In view of the fact that Army Ordnance can provide heavier orbital payloads with shorter time scales and with greater assurance of success, and that the Naval Research Laboratory is already heavily committed to the Aerobee-Hi development for the IGY. it is obvious that the Naval Research Laboratories will be better employed instrumenting properly the large payloads (100 pounds or more) which can be made available.18

The allusion to Aerobee-Hi development referred to improvements in sounding rockets which the IGY committee was anxious to use in continuing probes to supplement data from satellite experiments.

That General Simon's calculation of the perigee and payload in relation to the velocity attainable in the new high-powered version of Orbiter would correspond almost exactly to those achieved in January 1958 by the Army's Explorer satellite demonstrates the soundness of most of his predictions.19 His gloomy prophesy that rejection of Orbiter would cause "the loss of key personnel" at the Arsenal and thus jeopardize completion of the Redstone ballistic missile, on the contrary, turned out to be erroneous. To some ears it sounded like a piece of special pleading smacking of high-pressure political maneuvering. His dictum about what NRL should be doing did not sit well with Navy men who had faith in the Laboratory's special talents. And NRL and the Martin Company were to prove him mistaken in declaring that the development of any missile required "5 years at the minimum": Vanguard cut that time in half.

When Captain Samuel Tucker, the director of NRL, learned that the Stewart Committee was again to review Orbiter, he discussed the situation with Milton Rosen. Rosen suggested enlisting the aid of Vice Admiral John Sides of the Office of the Chief of Naval Operations in getting a second hearing for NRL also. The sympathetic Admiral Sides advised them to put their case to Admiral Robert P. Briscoe, Deputy Chief of Naval Operations. Briscoe at once volunteered to talk to Paul "Red" Smith, Assistant Secretary of the Navy for Research and Development. Indignant at the Army's attempt to snatch from the Navy its fairly won victory, Smith arranged for a second NRL presentation to the Stewart Committee. By then cost estimates for the Viking-based plan were ready, putting the figure for the launchers at $10.4 million and the over-all cost at $12 million.20

For the second NRL hearing Milton Rosen prepared a concise summary of the new features the Laboratory had introduced into its initial proposition. Substitution of a Sergeant-type third-stage rocket in keeping with the committee's recommendation necessitated a reduction in the weight of the satellite from 40 to 21.5 pounds for a 303-mile-perigee orbit, but specifications and performance data furnished by the Thiokol Chemical Corporation showed that a twelve-inch-diameter scale of its T-65 rocket which was then in production would meet NRL requirements. Laboratory calculations, moreover, indicated that the launcher thus modified could achieve a final velocity of 27,730 feet per second with a ten-pound payload at a 200-mile-perigee orbit, while computations undertaken by the Glenn L. Martin Company put the velocity at 28,350 feet per second, twelve percent above the required speed. The test schedule was now to include three firings of the first and second stages separately, a "large number" of the new third stage, and three of second and third stage combined before attempting a satellite launch at all. "The Laboratory is confident," read the final sentence of the summary, "that the first satellite can be launched eighteen months from the start of the program."21

Rosen had originally put the time needed at thirty month-an exactly accurate figure as events later showed-but the Martin Company believed a year and a half sufficient. Under pressure to pare down his estimate, especially as "we were fighting for our lives against a competitor who confidently said he could do the job in eighteen months," Rosen succumbed, accepting against his better judgment the more optimistic figure.22

To fortify faith in the Laboratory's capacity to meet so tight a time schedule, Rosen appended to his memorandum copies of telegrams and a letter from the four major industrial firms with whom NRL would expect to deal. A wire from the Thiokol Chemical Corporation on 22 August promised delivery of an enlarged solid-fueled T-65 rocket in nine months from receipt of a contract. A similar nine-month delivery guarantee on the power plant for the first-stage booster came from the General Electric Company. The Aerojet General Corporation wired that it was prepared to make a first delivery of Aerobee-Hi engines for the launcher's second stage eleven months after signing a contract, but as heat transfer difficulties were to be expected if UDMH fuel seas to be used, "some development firings" would be necessary to test the propellant.

The longest communication was from the Glenn L. Martin Company, producer of the Viking. The company's executive vice president stated: "We see no reason why it should not be possible to put a satellite in being in approximately 18 months provided the program is well defined and effectively managed." After reiterating that "the mission to be performed must be defined clearly." he added: "both government and industry must understand clearly the part each is to play in the program execution." A curiously admonitory tone pervaded the message, but it concluded on an encouraging, if slightly boastful, note: "We recognize that we have systems management experience in addition to the specific experience gained from Viking, but"-and this passage in Rosen's copy of the document was underlined in red pencil-"whether we are called upon to manage the program or to provide the airframe alone, you can be assured that we will support the program in the aggressive fashion necessary to achieve a satellite at the earliest practical date."23

These assurances from reputable industrial firms. particularly in regard to delivery dates, made an impression upon the Stewart Committee, since the time element, tellingly stressed by the Army general who spoke for Orbiter, now appeared to be about equal in both propositions. It left the strength of the pro-Viking arguments unimpaired-the sophistication of the instrumentation and the electronic tracking system, the miniaturization of parts, and the adaptability of the design to more elaborate spacecraft in the future. So the earlier verdict in favor of the NRL satellite stood.24

Oral word of the decision reached the Laboratory and Redstone Arsenal some seven weeks after the first presentations to the Group on Special Capabilities and more than a fortnight before the Deputy Secretary of Defense officially notified the Secretaries of the Army, Navy, and Air Force on 9 September that the Navy was to be in charge of a joint three-service program.25 The Army officers promoting Orbiter were incensed. General John Medaris of the Ordnance Corps privately labeled the rejection "a boondoggle." Commander George W. Hoover of ONR was incredulous that knowledgeable men who had listened to von Braun and examined the sheaves of detailed drawings prepared by him and his staff could have accepted an alternative consisting of "a blueprint of a pencil-shaped vehicle" many parts of which existed as yet only in the imagination of its authors. Frederick C. Durant III, one of the original backers of Orbiter, recounted that in the post-Sputnik era-after the failure of the first attempted Vanguard satellite launching and the subsequent success of the Army's vehicle-a National Academy official remarked ruefully that "one of the major reasons for the Army's losing out was von Braun's lousy presentation." Durant dubbed that explanation "odd" in view of von Braun's gifts of lucid exposition and persuasiveness.26 Conceivably in the course of his two-hour speech to the committee in July, the German rocket expert appeared so to exaggerate Orbiter's technical capabilities as to raise doubts in the minds of his audience.27

If only because several members of the IGY National Committee had served in the Navy during World War II and had acquired high respect for the caliber of Navy research, one might speculate as to whether they had greater confidence in the scientific environment at the Naval Research Laboratory than in that at Redstone Arsenal. Nothing, however, suggests that National Academy preferences, if known, influenced the Stewart Committee's choice. Outside the Department of Defense a number of scientists assumed that a big factor in the decision had been the relative security classification of the two boosters: Redstone, intended to be not only a test vehicle but a weapon in the American defense arsenal, would have to carry a secret tag, whereas the modified Viking would not; and secrecy would run counter to the IGY plan of sharing information with other nations. That idea was a misconception. The committee had known from the first that the guidance and control systems of any satellite vehicle would have to be military secrets; security considerations consequently had played little part in committee discussions.28

At NRL the staff was at once elated and frankly surprised. To those who knew most about the competing proposals the chances had looked minimal that the slim forty-foot Viking booster with a third stage of only partly determined design could win against the powerful sixty-nine-foot Redstone into which nearly four years of development work had already gone. Although the magnitude of the task to be accomplished by the Laboratory in three years might well have induced a touch of stage fright, excitement over the challenge submerged every doubt. Then and later, the NRL team attributed its victory to the quality of the scientific data the plan promised to produce and to the prospects it held out for future advances in rocket technology and space exploration.29