One of the oddities of the deep Earth is the presence of zones of the order of 1 to 10 km thick close to the boundary between the lower mantle and the outer core that have seismic wave speeds well below those expected at such depths. Because wave speed is inversely proportional to density, the chances are that they are “ponds” of extremely dense solid materials. Denser in fact than basalt might become in the form of eclogite, even compressed appropriately to these extreme depths. The zones have been a puzzle, but that seems to have been resolved by work from University College, London (Dobson, D.P. & Brodholt, J.P. 2005. Subducted banded iron formations as a source of ultralow-velocity zones at the core-mantle boundary. Nature, v. 434, p. 371-374). The densest materials found commonly at crustal levels are iron oxides and hydroxides, but today they are disseminated through much larger volumes or quartz-rich sediments. Up to about 1.8 billion years ago, they were produced in huge abundance in sedimentary rocks, along with interbedded cherts, to form banded iron formations (BIFs). That is widely agreed to have been a phenomenon only possible when the ocean was oxygen free so that iron could be dissolved in the oceans, and that they were precipitated when that Fe(II) came into contact with oxygen being produced by photosynthesising blue-green bacteria in shallow water. Without any shadow of doubt, BIFs are the densest sediment that the Earth has ever produced, with a 50:50 mix of iron oxide and chert having a density of 3900 kg m-3 at near-surface pressures, compared with 3100 for the upper mantle. Long ago, Bob Newton of the University of Chicago reckoned that they “didn’t oughta be around still”: Precambrian BIFs are so vast and so dense that they are even more likely to be subducted than oceanic basalt converted to eclogite. And they would not even need to be metamorphosed to do that. So, it has taken a long time for someone to cotton on to Newton’s typical prescience. Quite possibly, BIFs were a tectonic driving force at a time when the basalt-eclogite transformation was thermodynamically unlikely. Dobson and Brodholt observe that BIF density can only get larger (much larger; 6600 kgm-3 at CMB pressure) if they sink This is a nice hypothesis, for BIFs fit the bill exactly for the ultra-low velocity zones, and carries some interesting corollaries. BIFs contain a great deal of oxygen, in fact probably the entire productivity of the early Precambrian biosphere: that would have a biogenic isotope signature. Could that be added to any plume material emanating from the CMB? Equally, BIFs contain unusually high concentrations of transition metals, and there is another possibility for deep-mantle geochemists to juggle with. The authors also observe that iron-oxides have high electrical conductivity compared with silicates, and ponder on the electromagnetic consequences of that so close to the core. One thing seems certain; iron oxides probably would not melt, but, depending on the amount of oxygen in the core, they might dissolved in the molten outer core.