and far more visible to policy makers. None of us anticipated the work involved.

I returned to WHOI in late 1983. The World Ocean Circulation Experiment (WOCE) program was then taking shape under Carl Wunsch's leadership. In many ways it seemed a natural ally to the TTO program that had proceeded to the Equatorial, and South, Atlantic. But the latent conflicts between physics and chemistry had not disappeared: how many tracers were really necessary? Could a few sparsely placed samples yield adequate constraints? How solid were the boundary conditions? A rather fierce debate took place. The NAS Ocean Studies Board, with NSF support, provided the forum for this.

John Steele observed the emergence of WOCE, with roots in a desire to use an altimetric satellite combined with a global hydrographic survey, in order to study the global circulation. He was concerned that no program of similar scale existed to constrain the biogeochemical cycles of the ocean and that the promise of an ocean color satellite, hinted at by the Coastal Zone Color Scanner (CZCS) on Nimbus-7, might not be realized unless we took action. The descriptions of the major ocean biogeochemical cycles, which support life in the sea, rested on largely untested ground. He used the Ocean Studies Board to organize a major meeting, held in September 1984, at the Woods Hole Study Center.

The meeting (NRC, 1984) itself was plainly important, but confusing. Mixed together were satellites, primary productivity, higher organisms, sediment traps, radioisotopes, benthic instruments, and the sediment record. Linkages of this kind had been drawn in "horrendograms" by Francis Bretherton as visual drama but executing a coherent study was another matter alltogether. The CO₂ story was barely mentioned—would WOCE take care of this? John Steele, Jim Baker, Wally Broecker, Jim McCarthy, and Carl Wunsch kept a careful eye on proceedings. Ken Bruland had been selected to head the Planning Committee. The individual papers were good, but nothing seemed to gel—the topic was so broad, and almost no one had experience dealing with the National Aeronautics and Space Administration (NASA) and satellites. What to do next?

It was Neil Andersen who stepped forward and quietly asked a small group (Ken Bruland, Peter Jumars, Jim McCarthy, and me) to attend a meeting in Washington, ostensibly to edit the report. We met at the NAS building a week or two later. There Jim Baker walked us through the problem; getting a new start in NASA for an ocean color observing satellite would not be easy at all. And the political scene was fundamentally changed by the privatization passion of the current administration, so that commercial possibilities must be factored in. We did edit the GOFS report, and I urged that a far more prominent role for directly observing the controls on the oceanic carbon cycle be included.

A most difficult period then followed. Many participants had naively assumed that simply issuing a report would guarantee funds! The breadth of subject matter left room for a very large number of potential participants. And the review nature of the report, without any early crafting of tactics, left no road map with which to proceed. I was asked to chair the group and instantly felt these problems. A proposal to proceed was submitted through the National Research Council (NRC), and the first "pitch" was made to Burt Edelson at NASA. It went well.

The first step had to be consensus building, and a tense set of small working meetings followed through 1985, touching on each of the subthemes in the GOFS report. I became the sole member of a "Planning Office," helped enormously by the astonishing rise of electronic mail (pioneered by Omnet for the ocean science community). Great credit at this point must go to Neil Andersen, who saw the end point of a powerful program through the clutter of early discussions and carefully guided science along. Again, the Ocean Studies Board meetings served as the debating ground.

In October 1986, plans were more advanced, and I attended a major WOCE meeting, again at NAS. There I addressed the science behind the ocean carbon cycle and received a very enthusiastic audience response (my presentation materials were promptly "borrowed" by a complete stranger!). We had by then conceived within GOFS planning of a three-part attack on the problem: establishment of time-series stations at Bermuda and Hawaii to observe seasonal cycles and secular trends, a set of carefully crafted process studies to illuminate the controlling functions, and a global survey of the CO₂ field. It was this latter component that we wished to see accomplished as a collateral program with the WOCE global hydrographic survey, for it would be the critical glue that would scientifically link the two principal ocean observing programs. WOCE was measuring ¹⁴C distributions, building on the GEOSECS legacy, and our point was that the full CO ₂ system, with its embedded biogeochemical content, naturally followed. We could not afford two global surveys.

But enthusiasm and practicality do clash. There were basic problems of space and funds, let alone the interdisciplinary science. A blunt compromise was quickly reached; WOCE would provide bunk and laboratory space and access to samples. And GOFS would provide trained people, instruments, funds, and data, and would represent the program to appropriate bodies. It was a deal.

It was at once clear that this forced some new steps. WOCE and the global survey were now formally international. GOFS was still national. Within 24 hours a proposal was drafted to the Scientific Committee on Oceanic Research to request its attention to this program and to propose a true international effort. It was immediately hand carried to the SCOR General Meeting in Tasmania and well received.

The first SCOR-sponsored international meeting was held in Paris, at ICSU headquarters, in February 1987. It was