During winter 1977–78 seismologist Selwyn Sacks and geochemist Fouad Tera at lunch discussed a problem that troubled them about the subduction of lithospheric plates. It was clear that the lines of volcanoes that arranged themselves about 100 km behind a deep oceanic trench, which marked the subduction, had some kind of causal relationship with the plate, but what connection went beyond geometry? More to the point, did the plate furnish material for the magma that surfaced in these volcanoes? It was not a new problem, having been posed as soon as the dynamic model of the Earth had gained acceptance.

Needless to say, two parties had formed in the dispute, one seeing evidence for the subducted material in the lavas, the other seeing none. The difficulty was that all the chemical elements and isotopes were subducted, all the elements and isotopes were present in the mantle wedge that lay over the subducting plate, and all the elements and isotopes were present in the erupting lavas, so conclusions based on analyses of lava turned on an interpretation of relative proportions. Tera made the observation that the cosmogenic isotope 10Be, produced in the atmosphere by cosmic rays, had been measured in deep ocean sediments and that its half life of 1.5 million years was sufficiently long for it to be present in the lavas, if they incorporated any of the sediment and if the sediment had been transported to the roots of the volcanoes at speeds attributed to plates.