The character of the space effort of the late 1940s, in all the services, was best described by a section in the first annual report of the new Defense Department, issued at the close of 1948:

The Earth Satellite Vehicle Program, which was being carried out independently by each military service, was assigned to the Committee on Guided Missiles F of the Research and Development Board] for coordination. To provide an integrated program with resultant elimination of duplication, the committee recommended that current efforts in this field be limited to studies and component designs; well-defined areas of such research have been allocated to each of the three military departments.

The limitation "to studies and component designs" was particularly galling to the Air Force. In December 1947, in the letter which had ultimately led to the Vandenberg policy statement, the Engineering Division had specifically recommended- - on the basis of the earlier RAND studies- -that a satellite project should be established and a start made on component development. In the minds of Air Force engineers and scientists there was no doubt of the feasibility of the RAND approach and of the satellite itself. The problem was essentially that other and more critical programs were suffering from monetary anemia, and in such an environment there was slight chance of obtaining funding support needed for an active space program. In an era of relative abundance, the Air Force might have been able to overcome the skepticism of the civilian scientists who advised the Defense Department, or the enthusiastic support of a group of recognized scientists might have served to loosen departmental purse strings. But in the absence of one or the other, nothing could be done. The key factor, it was early apparent, was the absence of a clearly recognizable military requirement. In so many words, the skeptics could ask "what can a satellite do that an airplane can't do? " The answer of the time was that the satellite could do many things beyond the capacity of an airplane, but none seemed to serve any demonstrable military purpose.

In point of fact, it was precisely toward a proof of military utility that the Air Force had begun moving in 1947. Although disappointed in the fact that no specific development program had been approved, the Air Force was making reasonably steady progress through the studies to which it was essentially limited. The basic feasibility of satellites, from the standpoint of rocket performance, had been examined and accepted. By virtue of the 1946 and 1947 studies (and the subsequent Engineering Division analysis of their findings), engineers and scientists had gained assurance that a useful rocket vehicle to launch a satellite could be developed with but minor and entirely attainable advances over existing technology. Second, they had decided that the payload would have to be relatively slight--probably less than 2,000 pounds--until better rockets were available. Third, it was apparent that a recoverable vehicle would be a complication of the basic problem.

With this indication that a payload would be restricted to instrumentation and communications equipment, the question became one of what equipment, and with what utility. Between 1947 and 1951, RAND devoted considerable effort to an analysis of military usefulness, particularly to reconnaissance, a field "in which a satellite may well show advantages over other types of vehicles."

In those same years, the Air Force continued its tenuous progress toward acquiring authority to conduct a development program as opposed to a study effort. The subdued controversy between Navy and Air Force interests had flared into an open conflict in December 1947, when the Navy formally submitted to the Research and Development Board a claim for exclusive possession of rights to satellite development. After several weeks of acrimony, the Navy on 16 January 1948 (the day after the Vandenberg "position paper" on Air Force space interests) withdrew its claim, and the two services again set about their separate approaches. For the Air Force, all that could immediately mean was continuation of the RAND work. In February 1948, Air Force headquarters, through the Engineering Division, had asked RAND to undertake further detailed studies, and shortly thereafter obtained the concurrence of the Research and Development Board in that approach.

For practical purposes, the Navy discontinued satellite studies at that point. Under contract to the Navy, the Glenn L. Martin Company had been doing work similar to that of RAND since early 1946. The Martin efforts had resulted in a proposal for a 1,450-pound satellite that, said the researchers, could be orbited in the near future. (Much later, in the aftermath of the Sputnik affair, the then-president of the Martin Company told a group of reporters that the Navy-Martin program could have put a satellite into orbit "before the Korean War.")

Army ordnance was in roughly the same situation, a group at White Sands Proving Ground having designed a space flight experiment in the fall of 1947. By virtue of a general agreement with the Research and Development Board, however, the Air Force became the only service authorized to expend Defense Department funds on studies of satellite vehicles. The Air Force assigned the work to RAND under its regular contract, and the Research and Development Board subsequently (mid-1948) confirmed that RAND was solely responsible for such studies.

In November 1950, RAND submitted definitive recommendations to Air Force headquarters covering extension of research into specific aspects of the reconnaissance mission for satellites. Major General D.L. Putt, Air Force Director of Research and Development, endorsed the proposal and saw that it received necessary support. Its product was a pair of brief reports submitted in April 1951--reports which for the first time categorically and in considerable detail stated the engineering feasibility of a military-purpose satellite.

In the most important of the April 1951 studies, RAND reported that "pioneer reconnaissance (general location and determination of appropriate targets) and weather reconnaissance are suitable with the resolving power presently available to a satellite television system." In the interval between 1947 and 1951, of course, it was precisely that sort of intelligence which had become vitally important to the Air Force; the obvious prospective foe was the Soviet Union, its vast spaces and totalitarian political structure giving it relative security from conventional approaches. Moreover, in that interval the Soviet Union had demonstrated a largely unsuspected scientific competence by detonating atomic weapons years before experts anticipated that event, the Soviet sphere of influence had extended over the whole of continental China, and for the first time since 1939 the Soviet world had attempted armed assault on a bordering state- -the Republic of Korea.

In the case of robot reconnaissance, the researchers considered the basic problems involved in developing, assembling, launching, operating and profiting from the device. The analysis certainly was more comprehensive than anything previously attempted.

And the conclusions were quite encouraging, though not markedly different from those of l946:

The various components constituting a satellite vehicle to be utilized for reconnaissance . . . [are] individually feasible to various degrees. To combine these parts into a reliable operating whole will require considerable basic scientific and engineering effort. No radically new developments are indicated, however; rather, a reconstitution of known theory and art in rocketry, electronics, engines, and nuclear physics.

Specifically, the researchers concluded that a two-stage rocket (as opposed to the three- or four-stage vehicle originally considered in 1946) weighing about 74,000 pounds and carrying a 1,000-pound payload could satisfactorily conduct general reconnaissance, resolving objects with a maximum dimension of 200 feet. Reliability- - largely a matter of refining electronic components- -would generally determine the duration of useful activity. With improvements in television components to a stage then attained under laboratory conditions, it seemed entirely possible to reduce the resolvable dimension requirement to 100 feet while still providing continuous coverage from a single satellite on a basis of every target surveyed every other day. A further improvement (to a resolvable dimension of 40 feet) would theoretically permit virtually all military reconnaissance to be performed by satellite. Obviously, useful weather information could be obtained by even less demanding techniques: resolution on the order of 500-foot dimensions probably would prove entirely adequate. On the basis of sketchy experience with interpretation of weather trends from photographs taken from probe rockets, weather prediction also seemed feasible.

As was sometimes done for particularly significant or potentially sensitive subjects, RAND both preceded and followed the formal published studies and reports by presenting data to specialized groups, particularly at Wright Field and in the Pentagon. Among a great many experts who had been desperately puzzling with the reconnaissance problem, there resulted considerable enthusiasm for RAND's findings. The Research and Development Board, which had only recently rejected another Navy proposal for a small-package scientific satellite, fully sanctioned further studies. In the wake of the 1951 studies, the Air Force authorized RAND to make specific recommendations for the start of development work in the reconnaissance satellite program- -then called Project Feed Back.

Submission and consideration of the April 1951 RAND studies coincided, quite by accident, with the activation of an autonomous Air Research and Development Command and with increased stature for the recently created headquarters Air Force staff agency, the Deputy Chief of Staff, Development. Both organizations were guided by officers who were firmly convinced that far too much emphasis had been placed on procurement and production aspects of the materiel function in postwar years. They proposed to re-emphasize the research and development aspects of the Air Force mission, and they promptly set about their task.

By November 1951, the Air Force had arranged for the Atomic Energy Commission to begin work on small reactors suitable for use as power sources in satellite vehicles. RAND planned to subcontract major portions of the next phase of basic research, starting with the study of an orbital sensing and control subsystem (subcontracted to North American Aviation in March 1952). By June of that year, preliminary results of the reactor analyses were available; all were favorable to the feasibility of the proposal. A contract between RAND and the Radio Corporation of America followed, in mid-June; the electronics firm was to study optical systems, television cameras, radiation, recording devices, presentation techniques, and reliability aspects of a reconnaissance subsystem for a satellite. Concurrently the Communication and Navigation Laboratory at Wright Air Development Center contracted with North American Aviation for a study of a pre-orbital guidance system for a satellite (July 1953).

Most of this work was financed under a special supplement to the existing contract with RAND, effective for fiscal year 1953 and specifically designed to support the satellite research. The Atomic Energy Commission acceded to an Air Force proposal that it fund the study aspects of the reactor work, at least to the point of proving theoretical feasibility.

In the first two years following the establishment of an autonomous Air Research and Development Command, a minor difference of opinion involving RAND and the new organization occurred. The command decided early in its existence that the Air Force rather than the corporation should have management responsibility for the several subcontract studies being monitored by RAND. Nothing came of the proposal initially, and RAND continued to control study efforts ("the research phase") under the philosophy of turning the work over to the Air Force "as soon as development work can be started."

In May 1953, this process went one step farther. Air Force headquarters first directed the research and development command to investigate the feasibility of starting development work on an auxiliary nuclear power plant for the satellite, and then added instructions that the agency was to begin "active direction" of the entire Feed Back program by 1 June. One of the prime motives was the obvious fact that the reconnaissance satellite program had to be carefully integrated with the recently re-activated Atlas ballistic missile effort. Staff planners clearly foresaw that Atlas was the logical boost vehicle for the satellite. (At the time, the Air Force was also specifying reconnaissance versions of its major development systems, including Snark and Matador missiles. The reconnaissance package seemed to offer potential for a similar reconnaissance payload for the Atlas.

Air Research and Development Command representatives who began arranging a transfer of custody emerged from their initial contact with the RAND group in a state of high enthusiasm. Lieutenant Colonel V.L. Genez returned from his initial visit to the RAND satellite office with the firm conviction that an immediate effort should be made to orbit a satellite, regardless of the availability of the reconnaissance subsystem. He considered the psychological advantages of such a program to far outweigh any disability arising from limited operational utility. One month later, in September 1953, RAND itself flatly recommended letting a system design contract within a year and proceeding to a full system development program "perhaps immediately following the completion of experimental component tests."

Endorsement of the RAND recommendation by the research command headquarters and preliminary steps toward the start of component development marked the closing months of 1953. Although there were objections to the proposed acceleration of work (notably from the command's atomic energy program manager, who felt that at least another nine months of study should be devoted to the auxiliary power source before development began), the threads gradually began to draw together once more.

At that point, the Air Research and Development Command decided to pull together the proliferating aspects of the satellite work into a single project, thus making its unified management more feasible. Tentatively identified as Project 409-40, "Satellite Component Study," the program was also given unofficial possession of a system number (Weapon System 117 L) to cover the ultimate system development effort. On 3 December 1953, the program received new direction; headquarters of the Air Research and Development Command ordered Wright Air Development Center to redocument the effort under newly adopted management procedures (80-4) and to direct activity toward a demonstration of the feasibility of major satellite components. The television-optical reconnaissance subsystem, attitude and guidance control equipment, and the auxiliary power plant were specific goals. By the end of the year, the entire satellite "program" had made a semiofficial transition from a planning project to a proposed system. Given a new project number, it was transferred to the custody of the Bombardment Missiles Branch in Wright Air Development Center's Systems Management Organization.

In January 1954, while RAND was in the final stages of preparing a summary report on Project Feed Back, Project 1115 acquired the unclassified title "Advanced Reconnaissance System" and an MX (engineering project) number: MX-2226. Apart from the fact that the new names and codes were rather more prosaic than the "Feed Back" nomenclature earlier used, their adoption served to distinguish the proposed Air Force program from the RAND studies, which were rather well known throughout the services. However, the several items of code numbering, system number, and project number had not yet received confirmation or approval from Air Force headquarters. Although the work was progressing, it still lacked the authorization required for a fully effective program. Such authorization was to come in the trail of the long-awaited summary report from RAND on Project Feed Back.

Refinement of engineering data, intensive investigations of individual aspects of the reconnaissance satellite proposal, and highly detailed analysis of technical, fiscal and political (international) requirements and repercussions were complete by early 1954. Over a period of more than two years, RAND had subcontracted studies to a variety of highly qualified research and industry groups. Several hundred scientists and engineers had a part in the contributory studies and in the final report. In consequence, that report (dated 1 March 1954) contained the validated findings of some of the most highly regarded individuals and organizations in the nation. On the basis of such work, RAND specifically recommended that the Air Force undertake "the earliest possible completion and use of an efficient satellite reconnaissance vehicle" as a matter of "vital strategic interest to the United States." Additionally, RAND urged that the satellite project be "considered and planned" at a high policy level and that it be conducted under elaborate secrecy wraps to prevent dangerous international repercussions. On such a basis, it seemed possible to RAND that the development and initial operation of the satellite could be completed in about seven years and at a total cost "on the order of $165 million"- -although the researchers cautioned that uncertainties inherent in the prediction of development trends might double or treble that cost. (RAND also remarked, with considerable foresight, that "it may be possible to attain the end goal of the program from one to two years earlier at a considerable increase in cost.")

There was an element of finality to the concluding paragraph of the summary:

RAND has been working on the satellite vehicle for 8 years. During this period the metamorphosis from a feasibility concept to a useful reconnaissance purpose has occurred. Cognizance is now being turned over to the Air Force with the recommendation that the program be continued on a full-scale basis.