New measurements of the high-pressure volume of oxygen make plausible old, controversial diamond anvil cell (DAC) observations of a high-pressure reaction between liquid Fe and Fe-bearing silicate in which FeO dissolved into liquid Fe. If the Earth's D″ layer between core and mantle is the waste product of this reaction proceeding forward with oxygen transfer to the core, then D″ would form an effective lid preventing escape of siderophile elements from the core. On the other hand, it is at least as plausible that the core is now oxygen saturated and that core cooling and internal crystallization drive this reaction backward with transfer of Fe- and O-rich material out of the core into D″. The 186Os anomalies in some plume basalts are more consistent with this D″ as a siderophile element pump scenario than the D″ as-lid scenario. Contributions of siderophile elements to the mantle from the core could assist in balancing the persistently enigmatic budget of these elements.