Figures 24-28. Lunar and planetary probes. These were the deep-space analogs of the observatory-class satellites: Ranger, above left; Surveyor, above right, on the lunar surface; Lunar Orbiter, lower right; Mariner, opposite at top; Viking, lower opposite.

science was concerned, the program was to investigate the magnetosphere, to probe the origin and evolution of the moon and planets, to understand the solar processes, etc. NASA managers agreed, of course, and indeed from the very outset NASA people set forth the scientific objectives, not the space hardware, as the purpose of the space science program. 34 But the shorthand was too convenient, and the practice persisted even among the scientists.

HARD-LEARNED LESSONS

The rockets and spacecraft that won NASA an image of success did not come easily at first. During NASA's first two years the launch vehicles especially produced their share of grief. There were important lessons to learn, and experience proved to be a stern teacher. Atlas-Able and Pioneer provide a good illustration.

The Atlas-Able--a launch vehicle using the Atlas missile as the first stage combined with a second stage from the International Geophysical Year's Vanguard-had been brought into the spaceflight program by Abe

Silverstein, head of NASA's Office of Space Flight Development. NASA hoped to fill in and perhaps even to steal a march on the Soviet Union in lunar and planetary exploration while the agency pressed the development of its own space launch capability.35 Contracting with Space Technology Laboratories for three deep-space probes-Pioneer spacecraft-to be launched by a space vehicle yet to be proved, NASA hoped to move faster in its science program than would otherwise be possible with the smaller rockets available to the agency. At first intended to launch a Venus probe, Atlas-Able was switched to lunar missions when the planetary flight appeared to be too much to attempt at the start. Hopes ran high for AtlasAble and a number of scientists vied to help outfit the probe with scientific instruments. They had reason to be unhappy when their hard work went for naught.

Pioneer spacecraft on the Atlas-Able vehicles carried instruments to investigate interplanetary space and the moon. The Pioneer program, begun by the Air Force, was taken over by NASA when the new agency assumed responsibility for the nation's space science program.36 Between 11 October 1958 and 15 December 1960, eight attempts were made to launch Pioneers into space. 17 Six, including all three Atlas-Able firings, were failures. Pioneer 4, riding a Juno II launch vehicle, achieved a limited success.

Figure 29. Apollo lunar module. Six LMs descended from lunar orbit to land men on the moon, three of the craft carrying a lunar roving vehicle. Above, Apollo 16 LM and Rover on the moon in 1972 (with lunar dust on the camera lens showing in streaks).

Only Pioneer 5, launched by a Thor-Able on 11 March 1960, could be called an unqualified success. Although instruments on Pioneer 5 provided a great deal of information on cosmic rays and the space environment, the success could hardly erase the gloomy picture of six outright failures.

The significance of the Pioneer and Atlas-Able failures was all too clear at the time. In retrospect that significance stands out in high relief. For scientists it emphasized the sad plight that could await the experimenter who put his instruments on an as yet unproved vehicle. Long months, even years, of hard work could prove fruitless if the vehicle did not do its job properly. John Simpson of the Fermi Institute at the University of Chicago had reason to lament this hard fact of life. His group had developed new instruments to investigate the composition and energy of the solar wind and had accepted an invitation to put their instruments aboard the Pioneers, of which only Pioneer 5 had been satisfactory.

The fiasco precipitated a long series of exchanges between Simpson and NASA in which the frustrated investigator explored ways of recouping his losses by gaining access to other NASA flights. On 16 December 1960, the day after the last Atlas-Able attempt, Simpson was on the phone urgently reviewing his situation with the author.38 A half year later Simpson wrote that his group had participated in at least eight launchings with only two successes, a circumstance he attributed to having the Chicago experiments flown on unproved rocket systems or being assigned the role of secondary objectives.39

NASA had to face the issue of backups for important experiments, an issue that would always be at the back of an experimenter's mind when he

signed up for a long difficult period of preparing for a space mission.40 It soon became apparent that the time and effort required to conduct an experiment on a satellite or a space probe was far greater than that required to perform experiments in the laboratory. For the laboratory, one normally thought in terms of months, whereas space experiments could require years of hard work. Scientists began to point out that taking part in one or a few space science experiments could consume an appreciable fraction of a person's productive career, and in a "publish or perish" world the failure to get results because a rocket or a spacecraft didn't function properly could seriously affect that career. NASA was early moved by such considerations to adopt a policy of rescheduling experiments which, through no fault of the experimenter, did not succeed.11

Atlas-Able made plain that success on space missions would be neither automatic nor cheap. There was a price to pay, and part of that price was failure of some missions. This price NASA management would never find comfortable. The trouble with a philosophy of accepting a certain number of failures as normal and inevitable, was that even “learning failures” in an open program like NASA's, conducted under the watchful eye of a whole world, looked to the public and Congress like absolute failures. The press treated them as failures. It didn't matter that previous development projects like the V-2, Atlas, Thor, Polaris, and almost any other major rocket development one could name, had had their share of unsuccessful early firings, and that this had been accepted as a necessary growing pain. Those difficulties had been hardly visible under the cloak of military secrecy. But space program difficulties were highly visible and distressing.

The first notable application of a double standard came even before NASA was formed, with the spectacular Vanguard explosion in December 1957.42 Before the Vanguard development ever got under way, it had been agreed at the National Academy of Sciences that if only one International Geophysical Year satellite out of six made it to orbit, that would be taken as a successful outcome of the project, so difficult was it considered. In fact, when the number of launches for the program was cut back from 12 to 6, the scientists argued strongly that they couldn't reasonably expect more than two successful flights. 43 Actually three were successfully launched before the end of the program. Moreover, out of the Vanguard development came the liquid- and solid-fueled upper stages that made the Delta launch vehicle the tremendous success that it became. In addition, the Vanguard program contributed the Minitrack and Baker-Nunn tracking nets to help get the new space program off to a good start.

By any reasonably objective measures—certainly by previously accepted standards-Vanguard was a successful development. Yet the early launch failures made the entire program a symbol for failure in the public mind.

The lesson of Vanguard was plain. NASA could not afford to regard failure as acceptable under any guise. Success had to be sought on the first

try, and every reasonable effort bent toward achieving that outcome. This philosophy was epitomized in the "all-up" approach adopted by George Mueller, who had taken over direction of NASA's manned spaceflight programs in September 1963.44 In Apollo the all-up philosophy-which called for assembling a complete launcher and attempting to carry out a complete mission even on the early test flights-was intended to produce economies as well as to preserve an image of success.

But the all-up approach had already been applied to other NASA projects, particularly in the space science and applications areas. Explorer satellites were simply expected to succeed, and did. The Orbiting Solar Observatory was fully instrumented for space science on its first launch in March 1962. The flight to Venus in August 1962 was Mariner's maiden voyage. Later, Associate Administrator Seamans did not permit the Office of Space Science and Applications to try for only a part of the project's objectives on early launches of Surveyor to the moon. He insisted on a full-scale mission on the first lunar flight attempt. It became customary to plan for only one or two spacecraft for a project, expecting those to meet the desired objectives. While this policy kept the question of backups always in view, and was a source of some uneasiness, the detailed attention required to make the policy of success-on-the-first-try work was important in building confidence in NASA's ability to do its job.

And that was the third issue highlighted by the ill-fated Atlas-Able project. While getting on with the business of systematically building up the national launch capability, NASA had more or less turned over the handling of Atlas-Able flights to the industrial contractor. It was not a procedure that the agency would follow very often. Rather, with its open program-operating in a goldfish bowl, as it were-the agency would prefer to monitor its contractors very closely, often more closely than the contractors thought necessary. Indeed, a great deal of NASA's management time would be taken up in overseeing the work of contractors on rockets, spacecraft, and other equipment needed for the space program.

Finally, the Atlas-Able missions served to emphasize the key role played by the rocket in space operations. Successful rockets opened up space to human exploration. Without rockets, space must remain ever remote and inaccessible. Moreover-and this was the point behind the Atlas-Able attempts—some rockets could do more than others. Neither the Jupiter C, which had launched Explorer 1, nor the Vanguard launch vehicle could send probes to the moon or planets. With such vehicles one would have been restricted essentially to small artificial satellites of the earth. The limitations of these first American launch vehicles were further emphasized by comparison of the 8-kg Explorer 1 and the 28-kg Vanguard 3 with the 84- and 508-kg weights of Sputnik 1 and 2.45 Indeed, even more disturbing than the Soviet Union's launching of the first satellite was its obvious superiority in launch vehicles and implied superiority in long-range missile capa