The contract negotiated with McDonnell had compromised between the company's bid of $17,583,717, which was far from the lowest, and the more liberal STG estimate, to settle on a price of $18,300,000 for manufacturing 12 capsules. In view of this compromise upward, NASA officials were unprepared for the sudden acceleration of costs that the contractor claimed was necessary for spare parts, ground support, and checkout equipment. Before the ink was dry on the prime contract, the scope of research and development work was found to have mushroomed. In March, when McDonnell advised NASA that spares and test equipment would more than double the total contract costs, Abe Silverstein applied counterpressure, saying indignantly, "I will not tolerate increases such as those above in the contract for any reasons - utterly unreasonable to increase an $18,000,000 contract to $41,000,000 by these devices."46
Meanwhile STG and McDonnell representatives held a meeting at the working level to consolidate and condense the requirements for spare parts and equipment. Savings effected here were eventually greatly overridden by costs arising elsewhere. No one could yet foresee that the basic contract for 12 spacecraft would have an evolutionary history of its own.47 Cost accounting for a development program was recognized as a hazardous occupation, but just how hazardous and where to look for particular pitfalls took time to learn.
Whereas cynics might expect that the private-enterprise contractor for the capsule might have underbid to gain the contract, the civil servants in STG were  more surprised to learn that the public enterprise of furnishing the Nation's ballistic missile defense systems should also have underestimated costs by approximately one third. Informed by the Air Force Ballistic Missile Division in January that each Atlas booster would cost $3.3 million instead of $2.5 million, George Low tried for two months to get a satisfactory explanation of this sudden inflation.48
When in May, however, the STG learned of an increase by $8 million in the amount the Army Ballistic Missile Agency proposed to charge for the Redstones and Jupiters, the time had come for a thoroughgoing review of cost effectiveness and program requirements. Gilruth and Purser learned by investigation that the Ballistic Missile Agency was billing NASA a "burden" surcharge for the benefit of laboratory overhead costs at Huntsville. Purser's considered reaction to this was to threaten cancellation of the Jupiter program. If NASA must pay for research and development at the Redstone Arsenal, he said, then NASA, and STG in particular, must be more frugal in the estimation of their needs.
The Jupiter rocket had been selected to boost a full-scale capsule to about 16,000 feet per second, a velocity midway between the capacities of Little Joe and Redstone (6000 feet per second), and of Atlas (25,000 feet per second). But rather than insist on this step, Purser argued that the Atlas should be harnessed to duplicate the mission of the Jupiter flights. Since "the cost now equals or exceeds the cost of an Atlas for the same mission" and the Jupiter system would not be a "true duplicate of the Mercury capsule system," Purser recommended that the two Jupiter shots be canceled.49
After further consideration and more negotiations, Purser's recommendation was adopted by NASA Headquarters; the Jupiter series was eliminated from the Mercury program. In the aftermath of this episode, Glennan made an official complaint to the Secretary of Defense about the necessity to curtail proposed launchings to control costs, describing the situation with some chagrin:
Members of the staff who have visited Redstone Arsenal report that exceptionally high overhead rates apparently result from the necessity of supporting a large technical staff with a limited approved work program. The net result to us has been the increased costs of a Jupiter launching to more than that of an Atlas, whereas a Redstone launching is about $200,000 less than that of an Atlas. The prices being 2.7 and 2.9 million respectively.50At the same time Mercury engineers who were looking for an alternative to the balloon flight program discovered that the altitude wind tunnel, the biggest physical installation at Lewis Research Center, could be used to simulate environmental conditions up to 80,000 feet. Therefore the balloon flight test program, primarily designed to "soak" the capsule at comparable altitudes, was in effect canceled by May. DeMarquis D. Wyatt and other NASA Headquarters staffers preparing the budget requests for fiscal year 1960 now had evidence of STG's cost consciousness. The cancellations of the Jupiter series and the balloon program greatly simplified the program buildup toward manned space flight. STG engineers were pleased by the resulting concentration of effort.51
 One reason STG shed no tears over cancellation of Jupiter and the balloon tests was that the Little Joe program was making good progress. Blueprint work for the Little Joe airframe had begun early in 1959. North American had assigned A. L. Lawbaugh as project engineer; Langley Research Center had appointed Carl A. Sandahl as its representative for support of this test booster program; and William M. Bland, Jr., was managing Little Joe for the Space Task Group. Throughout the year 1959 these three men were primarily responsible for Little Joe.
Two significant design changes for Little Joe early in 1959 undoubtedly delayed the program slightly but contributed greatly to its eventual success. The first change, decided upon by Gilruth and Faget in January, required a switch from straight to canted nozzles on all the forward-thrusting rocket motors. Little Joe had no guidance system, and such a redesign would minimize any upset from unsymmetrical thrust conditions. The other departure from the original design was the addition of a so-called "booster destruct system." In the interest of range safety there should be some provision to terminate by command the thrust of the main motor units. Therefore Charles H. McFall and Samuel Sokol of Langley devised a booster blowout system, which North American and Thiokol Chemical Corporation, the manufacturers of the rocket motor components, added to the forward end of each rocket combustion chamber.52
By mid-February it was apparent that a development program for rocket hardware, even of such limited scope and relative simplicity as the Little Joe booster, demanded a far more sophisticated management organization than either Langley or the Task Group had envisioned. Although informal arrangements had sufficed to get the program started, funding allocations, personnel expansion, and contract monitoring problems began to weigh heavily. Carl Sandahl lamented in one weekly progress report that the transfer of Caldwell C. Johnson from Langley to the Space Task Group could "just about break up the Little Joe Project." Langley's loss was STG's gain in this respect, however, and cooperation continued to be encouraging. Indeed, in May, Bland reported that the delivery of the first Little Joe booster airframe could be expected approximately two weeks earlier than scheduled.53
Parallel to the development of the Little Joe test booster, STG and Langley engineers continued work on what now was called the Scout, the multistage, solid-propellant research rocket being designed since the previous year for sounding, probe, or small satellite missions. Langley had maintained its responsibility for designing the Scout for the Air Force after NACA became NASA; and early in 1959, Robert O. Piland and Joseph G. Thibodeaux came to work with William E. Stoney on the staging principles for the long, slim rocket. Although the Scout, as a Langley project, was not an integral part of STG's activities in Project Mercury, the Task Group held open the possibility of using this simple and relatively inexpensive rocket to launch scale models of the Mercury configuration and to probe for further critical data on heat transfer and stability. Thus the  Scout's capability could fill research gaps that might arise in the manned satellite project.54
Since January, when it had become apparent that the Army would not soon relinquish to NASA its rocket development team at Huntsville, NASA Headquarters had encouraged the Space Task Group to proceed full speed on personnel recruitment. The exact status of the organization and authority of STG was left unspecified, while Headquarters felt its way toward the establishment of the "space projects center" at Beltsville, just outside Washington. Although NASA had a "hunting license" as a result of its enabling legislation, STG's managers could not, without full support from President Eisenhower or Administrator Glennan, know how far or how hard to push the Space Task Group toward a permanent semi-autonomous establishment.55
STG's need for acquiring competent people without raiding established NASA research centers was met in large degree by a fortuitous accident that dramatized Anglo-European complaints about the "brain drain" of their scientific-technological manpower to the United States. A group of over 100 Canadian and British aeronautical engineers, who had been employed on a fighter-plane project for the British A. V. Roe (AVRO) Company near Toronto, Canada, were out of work. AVRO tried to find new jobs for them when the CF-105 Arrow project was canceled as a result of the Commonwealth's decision that the Bomarc missile made the Arrow obsolescent. Twenty-five of these engineers, led by James A. Chamberlin, a Canadian, were recruited by STG and immigrated to work at NASA's Virginia colony in mid-April. They were assigned jobs as individuals with the existing teams wherever each could be most useful, and they quickly proved themselves invaluable additions to making Mercury move.56
At the same time, the chief business administrator of the new NASA center at Beltsville, Michael J. Vaccaro, was planning to accommodate a complement of 425 people for fiscal year 1960 should Gilruth and his manned satellite team move to Maryland. On the first day of May 1959 the "space projects center," growing out of Naval Research Laboratory's Vanguard team, was renamed the Goddard Space Flight Center, and Gilruth's second hat, as the Center's Assistant Director for Manned Satellites, was reaffirmed. The Mercury program was specified as one of the six divisional offices at Goddard.57
While many questions of personnel, network management, and contract procedures for the capsule were still pending, Glennan made his first visit to the Space Task Group at Langley on May 18, 1959. He was impressed by the enormity of Project Mercury, by its working-level complexities, and by the extraordinarily fine morale in STG. Glennan returned to Washington resolved not to tamper with the esprit of STG. But he was also determined that NASA as a whole should not become a "space cadet" organization.58 The Administrator's resolution that NASA must not be overwhelmed by the complexities of manned space flight led to a Headquarters policy of minimal interference with the Task Group. During the next year, however, the weight of pressure from  the public press and the scope of intragovernmental coordination related to Mercury was to strain this policy.
45 Memo, MacDougall to Project Manager, "Estimated Cost of Manned Space Capsule Contract," Dec. 15, 1958; MacDougall, interview, Feb. 5, 1965.
46 Silverstein, marginal notes on memo, Low to Dir. for Space Flight Development, March 12, 1959; MacDougall, interview, Sept. 13, 1965.
47 Memo for files, Meyer, "Visit of McDonnell Representatives to Discuss Spare Parts and Ground Support Equipment," March 10, 1959. See also Ms., G. F. Bailey and S. A. Armstrong for Project Mercury Technical History Program, "Outline of the History of the Mercury Contract," April 8, 1963.
48 Low, "Status Report No. 8 - Project Mercury," March 4, 1959. Cf. Low, "Status Report No. 4." The extent of redesign work to "man-rate" the Atlas was more quickly recognized by its fabricators than by its new customers; Williams interview.
49 Memo, Purser to Gilruth, "Analysis of Army Ordnance Missile Command Revised Funding Estimate for Redstones and Jupiters," June 5, 1959.
50 Letter, Glennan to Neil H. McElroy, July 14, 1959. Administrator Glennan began to keep a desk diary in December 1958 [not available to this author], which carefully noted each day's transactions thereafter.
51 Low, "Status Report No. 14 - Project Mercury," May 22, 1959. De Marquis Wyatt, interview, Washington, Sept. 1, 1965; Bond interview.
52 Ms., William M. Bland, Jr., for Project Mercury Technical History Program, "The Birth of Little Joe Booster"; memo for files, Charles H. McFall, Jr., "Project Little Joe: Ground Instrumentation Required," April 15, 1959.
53 Memo for files, Sandahl, "Progress on Little Joe," Feb. 16, 1959; memo for files, Bland, "Results of Trip," May 19, 1959.
54 Purser, log for Gilruth, Jan. 12, 1959; memo, Charles B. Rumsey to Assoc. Dir., "Meeting to Discuss Project Mercury Problems to Which PARD Rocket Firings Might Contribute Information at an Early Date," Feb. 24, 1959. William E. Stoney, interview, Houston, Feb. 13, 1964. For Scout's capabilities, see "Considerations Affecting Satellite and Space Probe Research with Emphasis on the 'Scout' as a Launch Vehicle," NASA Technical Report R-97, Washington, 1961.
55 Purser log, Jan. 12, 1959. Cf. NASA Appropriations, Hearings, testimony of Dryden, 83-115. Wesley L. Hjornevik, interview, Houston, Feb. 17, 1964.
56 "Canadian Personnel Chart: Duty Assignments, Need to Know, Travel Requirements," STG, April 1959; Purser, log for Gilruth, April 21, 1959. Bringing fresh insight and seasoned experience to aid STG, this group included two, John D. Hodge and Jack Cohen, of rank equivalent to civil-service rating GS-14. Several more Anglo-Canadian engineers later joined STG.
57 Purser, log for Gilruth, April 27, 1959. See also documents signed by Glennan, Silverstein, and T. E. Jenkins, Administrative Officer of Goddard Space Flight Center, published as Exhibits 10, 11, and 12 of Appendix D in Alfred Rosenthal, The Early Years: Goddard Space Flight Center, Historical Origins and Activities through December 1962 (Washington, 1964), 35.
58 Glennan, interview with Eugene M. Emme, Cleveland, April 6, 1965; Purser, log for Gilruth, May 20, 1959.