I will address Marty's concern about priority in the publication of 3He-based estimates for the flux and isotopic composition of mantle carbon (“Who published what, and when?”). I will also discuss our views on the history of mantle/crust exchange.

Our analyses of MORB carbon occurred from 1980 to 1986, in collaboration with H. Sakai, J.G. Moore and J. Delaney, and J. Blank. We reported that a 13C-depleted component, released by combustion below 500°C, had been added to the samples after their eruption [Des Marais et al., 1983; Des Marais and Moore, 1984; Sakai et al., 1984]. Essentially all of the carbon released above 500°C derived from the original magma and was isotopically similar, both to the value measured earlier for CO2 in MORB vesicles [Pineau et al., 1976; Moore et al., 1977], and also to the average value for the crust. While isotopic comparisons between mantle and crust have indeed “focused the attention of many workers, starting in the 1950s,” the early estimates of mantle carbon were either based on unusual minerals or rocks (e.g. diamonds or carbonatites) whose isotopic composition could not be proved to represent the average upper mantle value, or they were based on analyses that implied (incorrectly) that basalts contain two major, magma-derived, isotopically distinct carbon components ( −4 to −9 permil, and −20 to −30 permil). Because midocean ridge (MOR) volcanism provides most of the mantle carbon flux, it offers the most representative sample of the upper mantle. Our work, together with H. Craig's measurements of MOR hydrothermal fluids, first showed that carbon isotope discrimination during eruption was less than a few permil, and that, at several ridge sites, the indigenous carbon (not only in MORB vesicles but also in MORB glass) was isotopically quite similar to that of the crustal carbon reservoir.