International cooperation is supposed to be the salvation of big science, but a four-country effort to design and build the next-generation fusion test reactor shows that the outlook for such collaborations is murky as well. This week, 18 months behind schedule, the United States, European Communities (EC), Russian Federation and Japan were expected to sign an agreement for a six-year, $1-billion engineering design phase of the International Thermonuclear Experimental Reactor (ITER). The pact, signed in Washington, will permit the three design centres, located in San Diego, California; Garching, Germany; and Naka, Japan to begin hiring what are expected to be as many as 180 scientists and engineers and 120 support staff. The fact that three centres were required, instead of the one venue first envisioned, suggests that even bigger problems loom when the partners try to decide where to build the actual reactor. Paul-Henri Rebut, the project's director-designate, wants to settle on a site in four years. Even then, the reactor could not operate until 2005. The partners had expected the engineering phase to flow relatively quickly from an earlier and smaller conceptual design completed in late 1990. Although an agreement was reached in principle last Augusteight months behind schedule-it has taken almost a year to gain the required approval of the member governments. ITER's official cost is $5.8 billion, although its supporters admit that the eventual price may be close to the $8.25-billion budgeted for the Superconducting SuperCollider. US officials are fond of citing that giant particle accelerator, under construction in Texas, as big science conceived and carried out the wrong way, without full international participation. Great care has been taken to ensure that the three design centres are seen as coequals, although Rebut, now head of the EC's Joint European Torus (JET) project in Culham, England, will work in San Diego. The project's oversight and design integration functions will also be there, although they may rotate as part of a continuing effort to keep the sites equal. The Garching site, outside Munich, will be responsible for designing the reactor vessel and its contents. That group will be headed by Ronald Parker of the Massachusetts Institute of Technology. The Japanese centre, led by Michel Huguet, also from JET, will design components outside the vessel, including the five-storey-high superconducting magnets that will shape and confine the fusion plasma within the reactor. Yasuo Shimonura of Japan's Atomic
