Scientists have cast doubt on a major part of the case for the Moon having once held abundant water.
A US team studied a mineral called apatite, which is found in a variety of lunar rock types.
Apatite, whose name comes from a Greek word meaning deceit, may have misled scientists into thinking the Moon is wetter that it actually is.
Lead author Jeremy Boyce said: 'We thought we had a great indicator, but it turns out it's not that reliable.'
Initial analysis of the lunar rocks brought back to Earth by the Apollo missions, suggested the Moon was 'bone dry'.
Apatite's name comes from the Greek word 'apat' meaning 'deceit'. And this is the last time we're going to let it trick us'
End Quote Dr Francis McCubbin University of New Mexico
But in the last decade, studies of volcanic glasses and apatite in lunar rocks have revealed them to be hydrogen-rich, building a compelling case for significant water having been present on the Moon as different minerals crystallised from the cooling magma.
Dr Boyce, a Nasa Early Career Fellow at the University of California, Los Angeles, presented his results at the Lunar and Planetary Science Conference (LPSC) in The Woodlands, Texas, this week. The work has also just been published in the prestigious journal Science.
The UCLA geochemist, along with collaborators Francis McCubbin, Steve Tomlinson, James Greenwood and Allan Treiman, simulated the formation of apatite minerals containing different amounts of volatile elements - hydrogen, chlorine and fluorine.
They demonstrated that it was possible to start with any water composition in the magma and by varying only the degree of crystallisation and the chlorine content, reproduce all the features seen in a diverse range of apatite from the Moon.
'We used to think it was a simple proportionality, that the more hydrogen was in the apatite, the more hydrogen in the magma,' Dr Boyce explained.
'Then we figured out... that it's a competition between hydrogen and mostly fluorine. Fluorine is the element that apatite most wants.'
To illustrate the complex chemical process involved in the formation of lunar apatite, Dr Boyce used a dating analogy, in which fluorine is apatite's ideal partner and chlorine is attractive, but not quite as much fluorine.
Lunar water: A brief timeline
- Early 1970s - Trace amounts of water found in rocks brought back by Apollo astronauts, but it is assumed to be contamination from Earth
- 1978 - Soviet scientists claim to detect water in samples returned to Earth by the robotic Luna 24 spacecraft
- 1994 - Results from the US military's Clementine probe's radar instrument suggest the presence of water-ice at the Moon's south pole
- 2008 - American scientists detect water in beads of volcanic glass from Apollo samples
- 2009 - Data from India's Chandrayaan spacecraft suggest the presence of hydrogen over large parts of the Moon's surface
- 2010-2011 - Studies by Boyce, McCubbin, Greenwood, Erik Hauri and others detect up to thousands of parts per million of water in apatite and lunar melt inclusions
He continued: 'Then the last apatite comes and there's nothing left but hydrogen. So it says: 'Okay, want to go out'?
'So all the apatites are taking all the fluorine and hiding it from the melt. Then the melt forgets that it had all that fluorine and the apatites get more chlorine-rich and more hydrogen-rich.' In this way, apatite may have produced a misleading indication of the original abundance of water in the Moon's interior.
The abundance of water in lunar rocks has important implications for the prevailing theory of the Moon's formation - known as the Giant Impact Model. According to the theory, several billion years ago, a planet-sized object called Theia collided with Earth, blasting rock into Earth orbit.
This material then coalesced to form the Moon. But this fiery origin story requires that volatile elements were boiled off, leaving the Moon depleted of water relative to Earth. So a less watery Moon ties in better with this theory.
The result generates new uncertainty about how much water the Moon started with. And the researchers point out that other sources of hydrogen, such as the solar wind, could have been incorporated into the apatite - further complicating the picture.
Dr Francis McCubbin, senior research scientist at the University of New Mexico, who's a co-author on the new study, told the BBC: 'There is some hydrogen that's coming in from the solar wind and getting stuck on the surface. But there is some amount - definitely seems to be less than on Earth - that the Moon started with.'
But what's also notable is that some of the authors of this study previously published some of the papers that built the case for a watery Moon.
'Clearly, we did the best we could at the time. But that's the progress of science - there are course corrections,' Dr Boyce explained.
'Definitely, there is still water on the Moon. Those rocks are not completely anhydrous. There's a really interesting record of heavy chlorine and hydrogen isotopes. But the abundances, we've demonstrated, are difficult to interpret.'
Dr McCubbin commented: 'Forty years ago, the Apollo astronauts built a building and the elevator was on a floor where [the water abundance] was one (ppb) part per billion.
'We took it up to where we were near terrestrial abundances, and then we realised we were on the wrong floor. We've taken it back down, but not all the way down to where we were 40 years ago.'
Dr Everett Gibson, from Nasa's Johnson Space Center (JSC) in Houston, who was not involved with the research, told BBC News found the results eminently plausible.
He said the scant evidence for hydrated minerals in lunar rocks had always been a problem for models of a water-rich Moon.
'The worry is, everything we measure is being modified and concentrated in a way that makes it impossible to get back to where it started,' explained Dr Boyce.
Dr McCubbin commented: 'Apatite's name comes from the Greek word 'apat', meaning 'deceit'. And this is the last time we're going to let it trick us.'
The latest results represent a personal landmark for Dr Boyce, who was hospitalised in 2012 with a viral infection and spent several weeks in a coma. The researcher had to learn to walk again afterwards.
'We're celebrating as only scientists celebrate, by publishing papers and getting back into our normal lives,' he explained.
Jeremy Boyce said that he, Francis McCubbin and James Greenwood of Wesleyan University, had been publishing papers on lunar apatite at the same time and could have spent the rest of their careers as scientific rivals.
However, the three scientists subsequently formed an alliance at a scientific meeting, deciding to collaborate, rather than compete, on the problem.
Paul.Rincon-INTERNET@bbc.co.uk and follow me on Twitter