Long, long ago an anonymous Roman wrote, ‘The first provision of any civilised society, after a code of law, is a reliable source of clean water’. Personally, I think the phrase ‘legalised bureaucracy’ in Latin was mistranslated to ‘code of law’. Whichever, planetary and life scientists might well like the adage for themselves: the sentiment applies nicely to active planetary tectonics and to the origin and survival of all conceivable life forms. The Earth has plenty of water at the surface and deep in the mantle. Without the second, the main mantle mineral olivine would be too stiff for the mantle to convect. Heat would build up within until magma formed in great abundance and emerged with a dreadful growl, as it did on Venus about 750 million years ago to repave the entire planet. It simply isn’t possible to think of answering the questions, ‘When did plate tectonics begin and life emerge?’ – let alone ‘How?’ – without first addressing where the Earth’s water came from and when our home world become so richly endowed.

In a very practical sense, these are the most important issues in geochemistry. Francis Albarède, of the École normale supérieure de Lyon, President of the European Association for Geochemistry and the first geochemist to deploy a multicollector, inductively coupled, plasma-source mass spectrometer, is a fitting person to review where the subdiscipline stands on them. (An MC-ICPMS is a tool for which many still yearn hopelessly.) His views appeared as a ‘Progress’ (a rare kind of Nature article) in the 29 October 2009 issue of Nature (Albarède, F. 2009. Volatile accretion history of the terrestrial planets and dynamic implications. Nature, v. 461, p. 1227-1233). The article casts doubt on the long-held views that when the Moon formed after a giant impact on the Earth, both bodies lost huge masses of volatiles, including water, and that Earth’s water-rich nature stemmed from repeated bombardment by volatile-rich comets up to about 3800 Ma.

Geochemical data are now available from a comet (Hyakutaki) and it contains twice the amount of deuterium relative to hydrogen that is in terrestrial seawater. The D/H ratio of carbonaceous chondrite meteorites is more Earth-like, and these primitive objects seem a more likely water source than comets. But did cataclysmic formation of the Earth-Moon system dehydrate both bodies and drive off other volatile matter? Planets and smaller bodies formed by gravitational accretion of solids that condensed from the initially hot gas or nebula that dominated the proto-solar system. Experiments show that condensation of the elements occurs in three discrete temperature ranges, separated by ranges in which few elements condense. Above around 1300 K the most refractory elements condensed, including oxides of some elements (Ca, Fe, Mg, Si) that now make silicate minerals, including the dominant mantle mineral olivine. Between 900-1200 K the alkali metals and some of the elements (chalcophile) that readily combine with sulfur emerged in solid form. In the third step from 500-800 K the more volatile chalcophile elements, including lead, and halogens condense, leaving four (Hg, O, N, C) that can take on solid form only below about 300 K. Interestingly, the proportions of volatile elements relative to refractory ones in the Earth, Moon and Martian meteorites are very low compared with those in carbonaceous chondrites. It is likely that volatile elements only accreted to the Inner Planets in small amounts before being swept to the outer reaches by an intense solar wind as the Sun was powering up, i.e. before nebular temperatures had fallen below about 1000 K. From that stems the inescapable conclusion that none of these planets were endowed with much water in their earliest forms.

Proportions of the lead isotopes ²⁰⁶Pb and ²⁰⁴Pb from terrestrial sulfide mineral deposits define a near-perfect linear relationship with the ages of mineralisation, from which an age can be estimated for the time the element lead appeared on Earth. That age is 4400 Ma; about 110 Ma younger than the actual age of the planet, and matches apparent ages derived from I-Xe and Pu-Xe decay schemes; iodine and xenon are volatile elements. This strongly supports the idea that 500-800 K condensates arrived late, and other evidence indicates that they and water ice were delivered by carbonaceous chondrite material falling towards the Sun from far beyond the orbits of the giant planets, once the early solar wind had lessened. That is, the Earth’s oceans formed very early in its history, and the mantle gained its water from them once hydrated lithosphere could founder deep into the evolving mantle by subduction. Albarède also summarises fascinating new ideas about the different course followed by Venus and Mars from essentially the same starting point. His ‘Progress’ is not difficult to read, and by marking the start of a new consensus in planetary evolution is of vital interest to all Earth scientists

Extraterrestrial water is also the subject of a Great Quest by NASA and other space agencies, though sadly an attempt on 9 October to prove that there is ice on the lunar surface, by hurling a US$79 million spacecraft at an obscure polar crater, produced no sensible results. Ironically, a couple of weeks later, three papers appeared in Science that document passive remote sensing evidence that the Moon contains a lot more water than long assumed (the most revealing is: Pieters, C.M. and 28 others 2009. Character and spatial distribution of OH/H₂O on the surface of the Moon seen by M³ on Chandrayaan-1. Science, v. 326, p. 568-572). The Apollo samples  astonished geologists when they proved to be almost completely anhydrous, any signs of minor hydration being ascribed to contamination after collection. The Moon Mineralogy Mapper (M³) aboard India’s first lunar mission Chandrayaan is a hyperspectral imaging device that operates in the visible to SWIR range of EM wavelengths (0.4 – 3.0 mm). That range includes SWIR wavelengths beyond 2.4 mm where OH^–, water and water ice have large absorption features that are masked in terrestrial remote sensing by the high moisture content of Earth’s atmosphere. Pieters et al. attempted to model hydroxyl and water content in the lunar surface, and discovered significant amounts (a few tenths of a percent) in the polar regions. That they got results when the Moon was fully illuminated by the Sun suggests that this is not due to ice hidden from heating in shadows, but to minerals that contain molecularly bound water and hydroxyl ions. That begs the question of how the water got there. One possibility is the late arrival of volatile condensates as above, another that it is due to hydrogen (protons) from the solar wind reducing iron in silicate minerals to metallic iron and combination with the oxygen released. Expect loud hurrahs from devotees of Star Trek and NASA because one prerequisite of civilised society seems to be there on the Moon. But judging from the bureaucracies involved in space, getting the funds to use it will not be easy.